linux/fs/nfs/super.c

2875 lines
71 KiB
C
Raw Normal View History

/*
* linux/fs/nfs/super.c
*
* Copyright (C) 1992 Rick Sladkey
*
* nfs superblock handling functions
*
* Modularised by Alan Cox <alan@lxorguk.ukuu.org.uk>, while hacking some
* experimental NFS changes. Modularisation taken straight from SYS5 fs.
*
* Change to nfs_read_super() to permit NFS mounts to multi-homed hosts.
* J.S.Peatfield@damtp.cam.ac.uk
*
* Split from inode.c by David Howells <dhowells@redhat.com>
*
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
* - superblocks are indexed on server only - all inodes, dentries, etc. associated with a
* particular server are held in the same superblock
* - NFS superblocks can have several effective roots to the dentry tree
* - directory type roots are spliced into the tree when a path from one root reaches the root
* of another (see nfs_lookup())
*/
#include <linux/module.h>
#include <linux/init.h>
#include <linux/time.h>
#include <linux/kernel.h>
#include <linux/mm.h>
#include <linux/string.h>
#include <linux/stat.h>
#include <linux/errno.h>
#include <linux/unistd.h>
#include <linux/sunrpc/clnt.h>
#include <linux/sunrpc/stats.h>
#include <linux/sunrpc/metrics.h>
#include <linux/sunrpc/xprtsock.h>
#include <linux/sunrpc/xprtrdma.h>
#include <linux/nfs_fs.h>
#include <linux/nfs_mount.h>
#include <linux/nfs4_mount.h>
#include <linux/lockd/bind.h>
#include <linux/smp_lock.h>
#include <linux/seq_file.h>
#include <linux/mount.h>
#include <linux/mnt_namespace.h>
#include <linux/namei.h>
#include <linux/nfs_idmap.h>
#include <linux/vfs.h>
#include <linux/inet.h>
#include <linux/in6.h>
#include <net/ipv6.h>
#include <linux/netdevice.h>
#include <linux/nfs_xdr.h>
#include <linux/magic.h>
#include <linux/parser.h>
#include <asm/system.h>
#include <asm/uaccess.h>
#include "nfs4_fs.h"
#include "callback.h"
#include "delegation.h"
#include "iostat.h"
#include "internal.h"
NFS: Define and create superblock-level objects Define and create superblock-level cache index objects (as managed by nfs_server structs). Each superblock object is created in a server level index object and is itself an index into which inode-level objects are inserted. Ideally there would be one superblock-level object per server, and the former would be folded into the latter; however, since the "nosharecache" option exists this isn't possible. The superblock object key is a sequence consisting of: (1) Certain superblock s_flags. (2) Various connection parameters that serve to distinguish superblocks for sget(). (3) The volume FSID. (4) The security flavour. (5) The uniquifier length. (6) The uniquifier text. This is normally an empty string, unless the fsc=xyz mount option was used to explicitly specify a uniquifier. The key blob is of variable length, depending on the length of (6). The superblock object is given no coherency data to carry in the auxiliary data permitted by the cache. It is assumed that the superblock is always coherent. This patch also adds uniquification handling such that two otherwise identical superblocks, at least one of which is marked "nosharecache", won't end up trying to share the on-disk cache. It will be possible to manually provide a uniquifier through a mount option with a later patch to avoid the error otherwise produced. Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Steve Dickson <steved@redhat.com> Acked-by: Trond Myklebust <Trond.Myklebust@netapp.com> Acked-by: Al Viro <viro@zeniv.linux.org.uk> Tested-by: Daire Byrne <Daire.Byrne@framestore.com>
2009-04-03 15:42:42 +00:00
#include "fscache.h"
#define NFSDBG_FACILITY NFSDBG_VFS
enum {
/* Mount options that take no arguments */
Opt_soft, Opt_hard,
Opt_posix, Opt_noposix,
Opt_cto, Opt_nocto,
Opt_ac, Opt_noac,
Opt_lock, Opt_nolock,
Opt_v2, Opt_v3,
Opt_udp, Opt_tcp, Opt_rdma,
Opt_acl, Opt_noacl,
Opt_rdirplus, Opt_nordirplus,
Opt_sharecache, Opt_nosharecache,
NFS: add "[no]resvport" mount option The standard default security setting for NFS is AUTH_SYS. An NFS client connects to NFS servers via a privileged source port and a fixed standard destination port (2049). The client sends raw uid and gid numbers to identify users making NFS requests, and the server assumes an appropriate authority on the client has vetted these values because the source port is privileged. On Linux, by default in-kernel RPC services use a privileged port in the range between 650 and 1023 to avoid using source ports of well- known IP services. Using such a small range limits the number of NFS mount points and the number of unique NFS servers to which a client can connect concurrently. An NFS client can use unprivileged source ports to expand the range of source port numbers, allowing more concurrent server connections and more NFS mount points. Servers must explicitly allow NFS connections from unprivileged ports for this to work. In the past, bumping the value of the sunrpc.max_resvport sysctl on the client would permit the NFS client to use unprivileged ports. Bumping this setting also changes the maximum port number used by other in-kernel RPC services, some of which still required a port number less than 1023. This is exacerbated by the way source port numbers are chosen by the Linux RPC client, which starts at the top of the range and works downwards. It means that bumping the maximum means all RPC services requesting a source port will likely get an unprivileged port instead of a privileged one. Changing this setting effects all NFS mount points on a client. A sysadmin could not selectively choose which mount points would use non-privileged ports and which could not. Lastly, this mechanism of expanding the limit on the number of NFS mount points was entirely undocumented. To address the need for the NFS client to use a large range of source ports without interfering with the activity of other in-kernel RPC services, we introduce a new NFS mount option. This option explicitly tells only the NFS client to use a non-privileged source port when communicating with the NFS server for one specific mount point. This new mount option is called "resvport," like the similar NFS mount option on FreeBSD and Mac OS X. A sister patch for nfs-utils will be submitted that documents this new option in nfs(5). The default setting for this new mount option requires the NFS client to use a privileged port, as before. Explicitly specifying the "noresvport" mount option allows the NFS client to use an unprivileged source port for this mount point when connecting to the NFS server port. This mount option is supported only for text-based NFS mounts. [ Sidebar: it is widely known that security mechanisms based on the use of privileged source ports are ineffective. However, the NFS client can combine the use of unprivileged ports with the use of secure authentication mechanisms, such as Kerberos. This allows a large number of connections and mount points while ensuring a useful level of security. Eventually we may change the default setting for this option depending on the security flavor used for the mount. For example, if the mount is using only AUTH_SYS, then the default setting will be "resvport;" if the mount is using a strong security flavor such as krb5, the default setting will be "noresvport." ] Signed-off-by: Chuck Lever <chuck.lever@oracle.com> [Trond.Myklebust@netapp.com: Fixed a bug whereby nfs4_init_client() was being called with incorrect arguments.] Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2008-12-23 20:21:37 +00:00
Opt_resvport, Opt_noresvport,
Opt_fscache, Opt_nofscache,
/* Mount options that take integer arguments */
Opt_port,
Opt_rsize, Opt_wsize, Opt_bsize,
Opt_timeo, Opt_retrans,
Opt_acregmin, Opt_acregmax,
Opt_acdirmin, Opt_acdirmax,
Opt_actimeo,
Opt_namelen,
Opt_mountport,
Opt_mountvers,
Opt_nfsvers,
Opt_minorversion,
/* Mount options that take string arguments */
Opt_sec, Opt_proto, Opt_mountproto, Opt_mounthost,
Opt_addr, Opt_mountaddr, Opt_clientaddr,
Opt_lookupcache,
Opt_fscache_uniq,
NFS: Allow either strict or sloppy mount option parsing The kernel's NFS client mount option parser currently doesn't allow unrecognized or incorrect mount options. This prevents misspellings or incorrectly specified mount options from possibly causing silent data corruption. However, NFS mount options are not standardized, so different operating systems can use differently spelled mount options to support similar features, or can support mount options which no other operating system supports. "Sloppy" mount option parsing, which allows the parser to ignore any option it doesn't recognize, is needed to support automounters that often use maps that are shared between heterogenous operating systems. The legacy mount command ignores the validity of the values of mount options entirely, except for the "sec=" and "proto=" options. If an incorrect value is specified, the out-of-range value is passed to the kernel; if a value is specified that contains non-numeric characters, it appears as though the legacy mount command sets that option to zero (probably incorrect behavior in general). In any case, this sets a precedent which we will partially follow for the kernel mount option parser: + if "sloppy" is not set, the parser will be strict about both unrecognized options (same as legacy) and invalid option values (stricter than legacy) + if "sloppy" is set, the parser will ignore unrecognized options and invalid option values (same as legacy) An "invalid" option value in this case means that either the type (integer, short, or string) or sign (for integer values) of the specified value is incorrect. This patch does two things: it changes the NFS client's mount option parsing loop so that it parses the whole string instead of failing at the first unrecognized option or invalid option value. An unrecognized option or an invalid option value cause the option to be skipped. Then, the patch adds a "sloppy" mount option that allows the parsing to succeed anyway if there were any problems during parsing. When parsing a set of options is complete, if there are errors and "sloppy" was specified, return success anyway. Otherwise, only return success if there are no errors. Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2008-06-24 23:28:02 +00:00
/* Special mount options */
Opt_userspace, Opt_deprecated, Opt_sloppy,
Opt_err
};
static const match_table_t nfs_mount_option_tokens = {
{ Opt_userspace, "bg" },
{ Opt_userspace, "fg" },
{ Opt_userspace, "retry=%s" },
NFS: Allow either strict or sloppy mount option parsing The kernel's NFS client mount option parser currently doesn't allow unrecognized or incorrect mount options. This prevents misspellings or incorrectly specified mount options from possibly causing silent data corruption. However, NFS mount options are not standardized, so different operating systems can use differently spelled mount options to support similar features, or can support mount options which no other operating system supports. "Sloppy" mount option parsing, which allows the parser to ignore any option it doesn't recognize, is needed to support automounters that often use maps that are shared between heterogenous operating systems. The legacy mount command ignores the validity of the values of mount options entirely, except for the "sec=" and "proto=" options. If an incorrect value is specified, the out-of-range value is passed to the kernel; if a value is specified that contains non-numeric characters, it appears as though the legacy mount command sets that option to zero (probably incorrect behavior in general). In any case, this sets a precedent which we will partially follow for the kernel mount option parser: + if "sloppy" is not set, the parser will be strict about both unrecognized options (same as legacy) and invalid option values (stricter than legacy) + if "sloppy" is set, the parser will ignore unrecognized options and invalid option values (same as legacy) An "invalid" option value in this case means that either the type (integer, short, or string) or sign (for integer values) of the specified value is incorrect. This patch does two things: it changes the NFS client's mount option parsing loop so that it parses the whole string instead of failing at the first unrecognized option or invalid option value. An unrecognized option or an invalid option value cause the option to be skipped. Then, the patch adds a "sloppy" mount option that allows the parsing to succeed anyway if there were any problems during parsing. When parsing a set of options is complete, if there are errors and "sloppy" was specified, return success anyway. Otherwise, only return success if there are no errors. Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2008-06-24 23:28:02 +00:00
{ Opt_sloppy, "sloppy" },
{ Opt_soft, "soft" },
{ Opt_hard, "hard" },
{ Opt_deprecated, "intr" },
{ Opt_deprecated, "nointr" },
{ Opt_posix, "posix" },
{ Opt_noposix, "noposix" },
{ Opt_cto, "cto" },
{ Opt_nocto, "nocto" },
{ Opt_ac, "ac" },
{ Opt_noac, "noac" },
{ Opt_lock, "lock" },
{ Opt_nolock, "nolock" },
{ Opt_v2, "v2" },
{ Opt_v3, "v3" },
{ Opt_udp, "udp" },
{ Opt_tcp, "tcp" },
{ Opt_rdma, "rdma" },
{ Opt_acl, "acl" },
{ Opt_noacl, "noacl" },
{ Opt_rdirplus, "rdirplus" },
{ Opt_nordirplus, "nordirplus" },
{ Opt_sharecache, "sharecache" },
{ Opt_nosharecache, "nosharecache" },
NFS: add "[no]resvport" mount option The standard default security setting for NFS is AUTH_SYS. An NFS client connects to NFS servers via a privileged source port and a fixed standard destination port (2049). The client sends raw uid and gid numbers to identify users making NFS requests, and the server assumes an appropriate authority on the client has vetted these values because the source port is privileged. On Linux, by default in-kernel RPC services use a privileged port in the range between 650 and 1023 to avoid using source ports of well- known IP services. Using such a small range limits the number of NFS mount points and the number of unique NFS servers to which a client can connect concurrently. An NFS client can use unprivileged source ports to expand the range of source port numbers, allowing more concurrent server connections and more NFS mount points. Servers must explicitly allow NFS connections from unprivileged ports for this to work. In the past, bumping the value of the sunrpc.max_resvport sysctl on the client would permit the NFS client to use unprivileged ports. Bumping this setting also changes the maximum port number used by other in-kernel RPC services, some of which still required a port number less than 1023. This is exacerbated by the way source port numbers are chosen by the Linux RPC client, which starts at the top of the range and works downwards. It means that bumping the maximum means all RPC services requesting a source port will likely get an unprivileged port instead of a privileged one. Changing this setting effects all NFS mount points on a client. A sysadmin could not selectively choose which mount points would use non-privileged ports and which could not. Lastly, this mechanism of expanding the limit on the number of NFS mount points was entirely undocumented. To address the need for the NFS client to use a large range of source ports without interfering with the activity of other in-kernel RPC services, we introduce a new NFS mount option. This option explicitly tells only the NFS client to use a non-privileged source port when communicating with the NFS server for one specific mount point. This new mount option is called "resvport," like the similar NFS mount option on FreeBSD and Mac OS X. A sister patch for nfs-utils will be submitted that documents this new option in nfs(5). The default setting for this new mount option requires the NFS client to use a privileged port, as before. Explicitly specifying the "noresvport" mount option allows the NFS client to use an unprivileged source port for this mount point when connecting to the NFS server port. This mount option is supported only for text-based NFS mounts. [ Sidebar: it is widely known that security mechanisms based on the use of privileged source ports are ineffective. However, the NFS client can combine the use of unprivileged ports with the use of secure authentication mechanisms, such as Kerberos. This allows a large number of connections and mount points while ensuring a useful level of security. Eventually we may change the default setting for this option depending on the security flavor used for the mount. For example, if the mount is using only AUTH_SYS, then the default setting will be "resvport;" if the mount is using a strong security flavor such as krb5, the default setting will be "noresvport." ] Signed-off-by: Chuck Lever <chuck.lever@oracle.com> [Trond.Myklebust@netapp.com: Fixed a bug whereby nfs4_init_client() was being called with incorrect arguments.] Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2008-12-23 20:21:37 +00:00
{ Opt_resvport, "resvport" },
{ Opt_noresvport, "noresvport" },
{ Opt_fscache, "fsc" },
{ Opt_fscache_uniq, "fsc=%s" },
{ Opt_nofscache, "nofsc" },
{ Opt_port, "port=%s" },
{ Opt_rsize, "rsize=%s" },
{ Opt_wsize, "wsize=%s" },
{ Opt_bsize, "bsize=%s" },
{ Opt_timeo, "timeo=%s" },
{ Opt_retrans, "retrans=%s" },
{ Opt_acregmin, "acregmin=%s" },
{ Opt_acregmax, "acregmax=%s" },
{ Opt_acdirmin, "acdirmin=%s" },
{ Opt_acdirmax, "acdirmax=%s" },
{ Opt_actimeo, "actimeo=%s" },
{ Opt_namelen, "namlen=%s" },
{ Opt_mountport, "mountport=%s" },
{ Opt_mountvers, "mountvers=%s" },
{ Opt_nfsvers, "nfsvers=%s" },
{ Opt_nfsvers, "vers=%s" },
{ Opt_minorversion, "minorversion=%u" },
{ Opt_sec, "sec=%s" },
{ Opt_proto, "proto=%s" },
{ Opt_mountproto, "mountproto=%s" },
{ Opt_addr, "addr=%s" },
{ Opt_clientaddr, "clientaddr=%s" },
{ Opt_mounthost, "mounthost=%s" },
{ Opt_mountaddr, "mountaddr=%s" },
{ Opt_lookupcache, "lookupcache=%s" },
{ Opt_err, NULL }
};
enum {
Opt_xprt_udp, Opt_xprt_tcp, Opt_xprt_rdma,
Opt_xprt_err
};
static const match_table_t nfs_xprt_protocol_tokens = {
{ Opt_xprt_udp, "udp" },
{ Opt_xprt_tcp, "tcp" },
{ Opt_xprt_rdma, "rdma" },
{ Opt_xprt_err, NULL }
};
enum {
Opt_sec_none, Opt_sec_sys,
Opt_sec_krb5, Opt_sec_krb5i, Opt_sec_krb5p,
Opt_sec_lkey, Opt_sec_lkeyi, Opt_sec_lkeyp,
Opt_sec_spkm, Opt_sec_spkmi, Opt_sec_spkmp,
Opt_sec_err
};
static const match_table_t nfs_secflavor_tokens = {
{ Opt_sec_none, "none" },
{ Opt_sec_none, "null" },
{ Opt_sec_sys, "sys" },
{ Opt_sec_krb5, "krb5" },
{ Opt_sec_krb5i, "krb5i" },
{ Opt_sec_krb5p, "krb5p" },
{ Opt_sec_lkey, "lkey" },
{ Opt_sec_lkeyi, "lkeyi" },
{ Opt_sec_lkeyp, "lkeyp" },
{ Opt_sec_spkm, "spkm3" },
{ Opt_sec_spkmi, "spkm3i" },
{ Opt_sec_spkmp, "spkm3p" },
{ Opt_sec_err, NULL }
};
enum {
Opt_lookupcache_all, Opt_lookupcache_positive,
Opt_lookupcache_none,
Opt_lookupcache_err
};
static match_table_t nfs_lookupcache_tokens = {
{ Opt_lookupcache_all, "all" },
{ Opt_lookupcache_positive, "pos" },
{ Opt_lookupcache_positive, "positive" },
{ Opt_lookupcache_none, "none" },
{ Opt_lookupcache_err, NULL }
};
static void nfs_umount_begin(struct super_block *);
static int nfs_statfs(struct dentry *, struct kstatfs *);
static int nfs_show_options(struct seq_file *, struct vfsmount *);
static int nfs_show_stats(struct seq_file *, struct vfsmount *);
static int nfs_get_sb(struct file_system_type *, int, const char *, void *, struct vfsmount *);
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
static int nfs_xdev_get_sb(struct file_system_type *fs_type,
int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt);
static void nfs_kill_super(struct super_block *);
static int nfs_remount(struct super_block *sb, int *flags, char *raw_data);
static struct file_system_type nfs_fs_type = {
.owner = THIS_MODULE,
.name = "nfs",
.get_sb = nfs_get_sb,
.kill_sb = nfs_kill_super,
.fs_flags = FS_RENAME_DOES_D_MOVE|FS_REVAL_DOT|FS_BINARY_MOUNTDATA,
};
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
struct file_system_type nfs_xdev_fs_type = {
.owner = THIS_MODULE,
.name = "nfs",
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
.get_sb = nfs_xdev_get_sb,
.kill_sb = nfs_kill_super,
.fs_flags = FS_RENAME_DOES_D_MOVE|FS_REVAL_DOT|FS_BINARY_MOUNTDATA,
};
static const struct super_operations nfs_sops = {
.alloc_inode = nfs_alloc_inode,
.destroy_inode = nfs_destroy_inode,
.write_inode = nfs_write_inode,
.statfs = nfs_statfs,
.clear_inode = nfs_clear_inode,
.umount_begin = nfs_umount_begin,
.show_options = nfs_show_options,
.show_stats = nfs_show_stats,
.remount_fs = nfs_remount,
};
#ifdef CONFIG_NFS_V4
static int nfs4_get_sb(struct file_system_type *fs_type,
int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt);
static int nfs4_remote_get_sb(struct file_system_type *fs_type,
int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt);
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
static int nfs4_xdev_get_sb(struct file_system_type *fs_type,
int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt);
static int nfs4_referral_get_sb(struct file_system_type *fs_type,
int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt);
static int nfs4_remote_referral_get_sb(struct file_system_type *fs_type,
int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt);
static void nfs4_kill_super(struct super_block *sb);
static struct file_system_type nfs4_fs_type = {
.owner = THIS_MODULE,
.name = "nfs4",
.get_sb = nfs4_get_sb,
.kill_sb = nfs4_kill_super,
.fs_flags = FS_RENAME_DOES_D_MOVE|FS_REVAL_DOT|FS_BINARY_MOUNTDATA,
};
static struct file_system_type nfs4_remote_fs_type = {
.owner = THIS_MODULE,
.name = "nfs4",
.get_sb = nfs4_remote_get_sb,
.kill_sb = nfs4_kill_super,
.fs_flags = FS_RENAME_DOES_D_MOVE|FS_REVAL_DOT|FS_BINARY_MOUNTDATA,
};
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
struct file_system_type nfs4_xdev_fs_type = {
.owner = THIS_MODULE,
.name = "nfs4",
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
.get_sb = nfs4_xdev_get_sb,
.kill_sb = nfs4_kill_super,
.fs_flags = FS_RENAME_DOES_D_MOVE|FS_REVAL_DOT|FS_BINARY_MOUNTDATA,
};
static struct file_system_type nfs4_remote_referral_fs_type = {
.owner = THIS_MODULE,
.name = "nfs4",
.get_sb = nfs4_remote_referral_get_sb,
.kill_sb = nfs4_kill_super,
.fs_flags = FS_RENAME_DOES_D_MOVE|FS_REVAL_DOT|FS_BINARY_MOUNTDATA,
};
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
struct file_system_type nfs4_referral_fs_type = {
.owner = THIS_MODULE,
.name = "nfs4",
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
.get_sb = nfs4_referral_get_sb,
.kill_sb = nfs4_kill_super,
.fs_flags = FS_RENAME_DOES_D_MOVE|FS_REVAL_DOT|FS_BINARY_MOUNTDATA,
};
static const struct super_operations nfs4_sops = {
.alloc_inode = nfs_alloc_inode,
.destroy_inode = nfs_destroy_inode,
.write_inode = nfs_write_inode,
.statfs = nfs_statfs,
.clear_inode = nfs4_clear_inode,
.umount_begin = nfs_umount_begin,
.show_options = nfs_show_options,
.show_stats = nfs_show_stats,
.remount_fs = nfs_remount,
};
#endif
static struct shrinker acl_shrinker = {
.shrink = nfs_access_cache_shrinker,
.seeks = DEFAULT_SEEKS,
};
/*
* Register the NFS filesystems
*/
int __init register_nfs_fs(void)
{
int ret;
ret = register_filesystem(&nfs_fs_type);
if (ret < 0)
goto error_0;
ret = nfs_register_sysctl();
if (ret < 0)
goto error_1;
#ifdef CONFIG_NFS_V4
ret = register_filesystem(&nfs4_fs_type);
if (ret < 0)
goto error_2;
#endif
register_shrinker(&acl_shrinker);
return 0;
#ifdef CONFIG_NFS_V4
error_2:
nfs_unregister_sysctl();
#endif
error_1:
unregister_filesystem(&nfs_fs_type);
error_0:
return ret;
}
/*
* Unregister the NFS filesystems
*/
void __exit unregister_nfs_fs(void)
{
unregister_shrinker(&acl_shrinker);
#ifdef CONFIG_NFS_V4
unregister_filesystem(&nfs4_fs_type);
#endif
nfs: fix oops re sysctls and V4 support NFS unregisters sysctls only if V4 support is compiled in. However, sysctl table is not V4 specific, so unregister it always. Steps to reproduce: [build nfs.ko with CONFIG_NFS_V4=n] modrobe nfs rmmod nfs ls /proc/sys Unable to handle kernel paging request at ffffffff880661c0 RIP: [<ffffffff802af8e3>] proc_sys_readdir+0xd3/0x350 PGD 203067 PUD 207063 PMD 7e216067 PTE 0 Oops: 0000 [1] SMP CPU 1 Modules linked in: lockd nfs_acl sunrpc Pid: 3335, comm: ls Not tainted 2.6.23-rc3-bloat #2 RIP: 0010:[<ffffffff802af8e3>] [<ffffffff802af8e3>] proc_sys_readdir+0xd3/0x350 RSP: 0018:ffff81007fd93e78 EFLAGS: 00010286 RAX: ffffffff880661c0 RBX: ffffffff80466370 RCX: ffffffff880661c0 RDX: 00000000000014c0 RSI: ffff81007f3ad020 RDI: ffff81007efd8b40 RBP: 0000000000000018 R08: 0000000000000000 R09: 0000000000000000 R10: 0000000000000001 R11: ffffffff802a8570 R12: ffffffff880661c0 R13: ffff81007e219640 R14: ffff81007efd8b40 R15: ffff81007ded7280 FS: 00002ba25ef03060(0000) GS:ffff81007ff81258(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b CR2: ffffffff880661c0 CR3: 000000007dfaf000 CR4: 00000000000006e0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Process ls (pid: 3335, threadinfo ffff81007fd92000, task ffff81007d8a0000) Stack: ffff81007f3ad150 ffffffff80283f30 ffff81007fd93f48 ffff81007efd8b40 ffff81007ee00440 0000000422222222 0000000200035593 ffffffff88037e9a 2222222222222222 ffffffff80466500 ffff81007e416400 ffff81007e219640 Call Trace: [<ffffffff80283f30>] filldir+0x0/0xf0 [<ffffffff80283f30>] filldir+0x0/0xf0 [<ffffffff802840c7>] vfs_readdir+0xa7/0xc0 [<ffffffff80284376>] sys_getdents+0x96/0xe0 [<ffffffff8020bb3e>] system_call+0x7e/0x83 Code: 41 8b 14 24 85 d2 74 dc 49 8b 44 24 08 48 85 c0 74 e7 49 3b RIP [<ffffffff802af8e3>] proc_sys_readdir+0xd3/0x350 RSP <ffff81007fd93e78> CR2: ffffffff880661c0 Kernel panic - not syncing: Fatal exception Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com> Acked-by: Trond Myklebust <trond.myklebust@fys.uio.no> Cc: "J. Bruce Fields" <bfields@fieldses.org> Cc: <stable@kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-09-19 05:46:40 +00:00
nfs_unregister_sysctl();
unregister_filesystem(&nfs_fs_type);
}
void nfs_sb_active(struct super_block *sb)
{
struct nfs_server *server = NFS_SB(sb);
if (atomic_inc_return(&server->active) == 1)
atomic_inc(&sb->s_active);
}
void nfs_sb_deactive(struct super_block *sb)
{
struct nfs_server *server = NFS_SB(sb);
if (atomic_dec_and_test(&server->active))
deactivate_super(sb);
}
/*
* Deliver file system statistics to userspace
*/
static int nfs_statfs(struct dentry *dentry, struct kstatfs *buf)
{
struct nfs_server *server = NFS_SB(dentry->d_sb);
unsigned char blockbits;
unsigned long blockres;
struct nfs_fh *fh = NFS_FH(dentry->d_inode);
struct nfs_fattr fattr;
struct nfs_fsstat res = {
.fattr = &fattr,
};
int error;
error = server->nfs_client->rpc_ops->statfs(server, fh, &res);
if (error < 0)
goto out_err;
buf->f_type = NFS_SUPER_MAGIC;
/*
* Current versions of glibc do not correctly handle the
* case where f_frsize != f_bsize. Eventually we want to
* report the value of wtmult in this field.
*/
buf->f_frsize = dentry->d_sb->s_blocksize;
/*
* On most *nix systems, f_blocks, f_bfree, and f_bavail
* are reported in units of f_frsize. Linux hasn't had
* an f_frsize field in its statfs struct until recently,
* thus historically Linux's sys_statfs reports these
* fields in units of f_bsize.
*/
buf->f_bsize = dentry->d_sb->s_blocksize;
blockbits = dentry->d_sb->s_blocksize_bits;
blockres = (1 << blockbits) - 1;
buf->f_blocks = (res.tbytes + blockres) >> blockbits;
buf->f_bfree = (res.fbytes + blockres) >> blockbits;
buf->f_bavail = (res.abytes + blockres) >> blockbits;
buf->f_files = res.tfiles;
buf->f_ffree = res.afiles;
buf->f_namelen = server->namelen;
return 0;
out_err:
dprintk("%s: statfs error = %d\n", __func__, -error);
return error;
}
/*
* Map the security flavour number to a name
*/
static const char *nfs_pseudoflavour_to_name(rpc_authflavor_t flavour)
{
static const struct {
rpc_authflavor_t flavour;
const char *str;
} sec_flavours[] = {
{ RPC_AUTH_NULL, "null" },
{ RPC_AUTH_UNIX, "sys" },
{ RPC_AUTH_GSS_KRB5, "krb5" },
{ RPC_AUTH_GSS_KRB5I, "krb5i" },
{ RPC_AUTH_GSS_KRB5P, "krb5p" },
{ RPC_AUTH_GSS_LKEY, "lkey" },
{ RPC_AUTH_GSS_LKEYI, "lkeyi" },
{ RPC_AUTH_GSS_LKEYP, "lkeyp" },
{ RPC_AUTH_GSS_SPKM, "spkm" },
{ RPC_AUTH_GSS_SPKMI, "spkmi" },
{ RPC_AUTH_GSS_SPKMP, "spkmp" },
{ UINT_MAX, "unknown" }
};
int i;
for (i = 0; sec_flavours[i].flavour != UINT_MAX; i++) {
if (sec_flavours[i].flavour == flavour)
break;
}
return sec_flavours[i].str;
}
static void nfs_show_mountd_options(struct seq_file *m, struct nfs_server *nfss,
int showdefaults)
{
struct sockaddr *sap = (struct sockaddr *)&nfss->mountd_address;
switch (sap->sa_family) {
case AF_INET: {
struct sockaddr_in *sin = (struct sockaddr_in *)sap;
seq_printf(m, ",mountaddr=%pI4", &sin->sin_addr.s_addr);
break;
}
case AF_INET6: {
struct sockaddr_in6 *sin6 = (struct sockaddr_in6 *)sap;
seq_printf(m, ",mountaddr=%pI6", &sin6->sin6_addr);
break;
}
default:
if (showdefaults)
seq_printf(m, ",mountaddr=unspecified");
}
if (nfss->mountd_version || showdefaults)
seq_printf(m, ",mountvers=%u", nfss->mountd_version);
if (nfss->mountd_port || showdefaults)
seq_printf(m, ",mountport=%u", nfss->mountd_port);
switch (nfss->mountd_protocol) {
case IPPROTO_UDP:
seq_printf(m, ",mountproto=udp");
break;
case IPPROTO_TCP:
seq_printf(m, ",mountproto=tcp");
break;
default:
if (showdefaults)
seq_printf(m, ",mountproto=auto");
}
}
/*
* Describe the mount options in force on this server representation
*/
static void nfs_show_mount_options(struct seq_file *m, struct nfs_server *nfss,
int showdefaults)
{
static const struct proc_nfs_info {
int flag;
const char *str;
const char *nostr;
} nfs_info[] = {
{ NFS_MOUNT_SOFT, ",soft", ",hard" },
{ NFS_MOUNT_POSIX, ",posix", "" },
{ NFS_MOUNT_NOCTO, ",nocto", "" },
{ NFS_MOUNT_NOAC, ",noac", "" },
{ NFS_MOUNT_NONLM, ",nolock", "" },
{ NFS_MOUNT_NOACL, ",noacl", "" },
{ NFS_MOUNT_NORDIRPLUS, ",nordirplus", "" },
NFS: add "[no]resvport" mount option The standard default security setting for NFS is AUTH_SYS. An NFS client connects to NFS servers via a privileged source port and a fixed standard destination port (2049). The client sends raw uid and gid numbers to identify users making NFS requests, and the server assumes an appropriate authority on the client has vetted these values because the source port is privileged. On Linux, by default in-kernel RPC services use a privileged port in the range between 650 and 1023 to avoid using source ports of well- known IP services. Using such a small range limits the number of NFS mount points and the number of unique NFS servers to which a client can connect concurrently. An NFS client can use unprivileged source ports to expand the range of source port numbers, allowing more concurrent server connections and more NFS mount points. Servers must explicitly allow NFS connections from unprivileged ports for this to work. In the past, bumping the value of the sunrpc.max_resvport sysctl on the client would permit the NFS client to use unprivileged ports. Bumping this setting also changes the maximum port number used by other in-kernel RPC services, some of which still required a port number less than 1023. This is exacerbated by the way source port numbers are chosen by the Linux RPC client, which starts at the top of the range and works downwards. It means that bumping the maximum means all RPC services requesting a source port will likely get an unprivileged port instead of a privileged one. Changing this setting effects all NFS mount points on a client. A sysadmin could not selectively choose which mount points would use non-privileged ports and which could not. Lastly, this mechanism of expanding the limit on the number of NFS mount points was entirely undocumented. To address the need for the NFS client to use a large range of source ports without interfering with the activity of other in-kernel RPC services, we introduce a new NFS mount option. This option explicitly tells only the NFS client to use a non-privileged source port when communicating with the NFS server for one specific mount point. This new mount option is called "resvport," like the similar NFS mount option on FreeBSD and Mac OS X. A sister patch for nfs-utils will be submitted that documents this new option in nfs(5). The default setting for this new mount option requires the NFS client to use a privileged port, as before. Explicitly specifying the "noresvport" mount option allows the NFS client to use an unprivileged source port for this mount point when connecting to the NFS server port. This mount option is supported only for text-based NFS mounts. [ Sidebar: it is widely known that security mechanisms based on the use of privileged source ports are ineffective. However, the NFS client can combine the use of unprivileged ports with the use of secure authentication mechanisms, such as Kerberos. This allows a large number of connections and mount points while ensuring a useful level of security. Eventually we may change the default setting for this option depending on the security flavor used for the mount. For example, if the mount is using only AUTH_SYS, then the default setting will be "resvport;" if the mount is using a strong security flavor such as krb5, the default setting will be "noresvport." ] Signed-off-by: Chuck Lever <chuck.lever@oracle.com> [Trond.Myklebust@netapp.com: Fixed a bug whereby nfs4_init_client() was being called with incorrect arguments.] Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2008-12-23 20:21:37 +00:00
{ NFS_MOUNT_UNSHARED, ",nosharecache", "" },
{ NFS_MOUNT_NORESVPORT, ",noresvport", "" },
{ 0, NULL, NULL }
};
const struct proc_nfs_info *nfs_infop;
struct nfs_client *clp = nfss->nfs_client;
u32 version = clp->rpc_ops->version;
seq_printf(m, ",vers=%u", version);
seq_printf(m, ",rsize=%u", nfss->rsize);
seq_printf(m, ",wsize=%u", nfss->wsize);
if (nfss->bsize != 0)
seq_printf(m, ",bsize=%u", nfss->bsize);
seq_printf(m, ",namlen=%u", nfss->namelen);
if (nfss->acregmin != NFS_DEF_ACREGMIN*HZ || showdefaults)
seq_printf(m, ",acregmin=%u", nfss->acregmin/HZ);
if (nfss->acregmax != NFS_DEF_ACREGMAX*HZ || showdefaults)
seq_printf(m, ",acregmax=%u", nfss->acregmax/HZ);
if (nfss->acdirmin != NFS_DEF_ACDIRMIN*HZ || showdefaults)
seq_printf(m, ",acdirmin=%u", nfss->acdirmin/HZ);
if (nfss->acdirmax != NFS_DEF_ACDIRMAX*HZ || showdefaults)
seq_printf(m, ",acdirmax=%u", nfss->acdirmax/HZ);
for (nfs_infop = nfs_info; nfs_infop->flag; nfs_infop++) {
if (nfss->flags & nfs_infop->flag)
seq_puts(m, nfs_infop->str);
else
seq_puts(m, nfs_infop->nostr);
}
seq_printf(m, ",proto=%s",
rpc_peeraddr2str(nfss->client, RPC_DISPLAY_PROTO));
if (version == 4) {
if (nfss->port != NFS_PORT)
seq_printf(m, ",port=%u", nfss->port);
} else
if (nfss->port)
seq_printf(m, ",port=%u", nfss->port);
seq_printf(m, ",timeo=%lu", 10U * nfss->client->cl_timeout->to_initval / HZ);
seq_printf(m, ",retrans=%u", nfss->client->cl_timeout->to_retries);
seq_printf(m, ",sec=%s", nfs_pseudoflavour_to_name(nfss->client->cl_auth->au_flavor));
if (version != 4)
nfs_show_mountd_options(m, nfss, showdefaults);
#ifdef CONFIG_NFS_V4
if (clp->rpc_ops->version == 4)
seq_printf(m, ",clientaddr=%s", clp->cl_ipaddr);
#endif
if (nfss->options & NFS_OPTION_FSCACHE)
seq_printf(m, ",fsc");
}
/*
* Describe the mount options on this VFS mountpoint
*/
static int nfs_show_options(struct seq_file *m, struct vfsmount *mnt)
{
struct nfs_server *nfss = NFS_SB(mnt->mnt_sb);
nfs_show_mount_options(m, nfss, 0);
seq_printf(m, ",addr=%s",
rpc_peeraddr2str(nfss->nfs_client->cl_rpcclient,
RPC_DISPLAY_ADDR));
return 0;
}
/*
* Present statistical information for this VFS mountpoint
*/
static int nfs_show_stats(struct seq_file *m, struct vfsmount *mnt)
{
int i, cpu;
struct nfs_server *nfss = NFS_SB(mnt->mnt_sb);
struct rpc_auth *auth = nfss->client->cl_auth;
struct nfs_iostats totals = { };
seq_printf(m, "statvers=%s", NFS_IOSTAT_VERS);
/*
* Display all mount option settings
*/
seq_printf(m, "\n\topts:\t");
seq_puts(m, mnt->mnt_sb->s_flags & MS_RDONLY ? "ro" : "rw");
seq_puts(m, mnt->mnt_sb->s_flags & MS_SYNCHRONOUS ? ",sync" : "");
seq_puts(m, mnt->mnt_sb->s_flags & MS_NOATIME ? ",noatime" : "");
seq_puts(m, mnt->mnt_sb->s_flags & MS_NODIRATIME ? ",nodiratime" : "");
nfs_show_mount_options(m, nfss, 1);
seq_printf(m, "\n\tage:\t%lu", (jiffies - nfss->mount_time) / HZ);
seq_printf(m, "\n\tcaps:\t");
seq_printf(m, "caps=0x%x", nfss->caps);
seq_printf(m, ",wtmult=%u", nfss->wtmult);
seq_printf(m, ",dtsize=%u", nfss->dtsize);
seq_printf(m, ",bsize=%u", nfss->bsize);
seq_printf(m, ",namlen=%u", nfss->namelen);
#ifdef CONFIG_NFS_V4
if (nfss->nfs_client->rpc_ops->version == 4) {
seq_printf(m, "\n\tnfsv4:\t");
seq_printf(m, "bm0=0x%x", nfss->attr_bitmask[0]);
seq_printf(m, ",bm1=0x%x", nfss->attr_bitmask[1]);
seq_printf(m, ",acl=0x%x", nfss->acl_bitmask);
}
#endif
/*
* Display security flavor in effect for this mount
*/
seq_printf(m, "\n\tsec:\tflavor=%u", auth->au_ops->au_flavor);
if (auth->au_flavor)
seq_printf(m, ",pseudoflavor=%u", auth->au_flavor);
/*
* Display superblock I/O counters
*/
for_each_possible_cpu(cpu) {
struct nfs_iostats *stats;
preempt_disable();
stats = per_cpu_ptr(nfss->io_stats, cpu);
for (i = 0; i < __NFSIOS_COUNTSMAX; i++)
totals.events[i] += stats->events[i];
for (i = 0; i < __NFSIOS_BYTESMAX; i++)
totals.bytes[i] += stats->bytes[i];
#ifdef CONFIG_NFS_FSCACHE
for (i = 0; i < __NFSIOS_FSCACHEMAX; i++)
totals.fscache[i] += stats->fscache[i];
#endif
preempt_enable();
}
seq_printf(m, "\n\tevents:\t");
for (i = 0; i < __NFSIOS_COUNTSMAX; i++)
seq_printf(m, "%lu ", totals.events[i]);
seq_printf(m, "\n\tbytes:\t");
for (i = 0; i < __NFSIOS_BYTESMAX; i++)
seq_printf(m, "%Lu ", totals.bytes[i]);
#ifdef CONFIG_NFS_FSCACHE
if (nfss->options & NFS_OPTION_FSCACHE) {
seq_printf(m, "\n\tfsc:\t");
for (i = 0; i < __NFSIOS_FSCACHEMAX; i++)
seq_printf(m, "%Lu ", totals.bytes[i]);
}
#endif
seq_printf(m, "\n");
rpc_print_iostats(m, nfss->client);
return 0;
}
/*
* Begin unmount by attempting to remove all automounted mountpoints we added
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
* in response to xdev traversals and referrals
*/
static void nfs_umount_begin(struct super_block *sb)
{
struct nfs_server *server;
struct rpc_clnt *rpc;
lock_kernel();
server = NFS_SB(sb);
/* -EIO all pending I/O */
rpc = server->client_acl;
if (!IS_ERR(rpc))
rpc_killall_tasks(rpc);
rpc = server->client;
if (!IS_ERR(rpc))
rpc_killall_tasks(rpc);
unlock_kernel();
}
/*
* Sanity-check a server address provided by the mount command.
*
* Address family must be initialized, and address must not be
* the ANY address for that family.
*/
static int nfs_verify_server_address(struct sockaddr *addr)
{
switch (addr->sa_family) {
case AF_INET: {
struct sockaddr_in *sa = (struct sockaddr_in *)addr;
return sa->sin_addr.s_addr != htonl(INADDR_ANY);
}
case AF_INET6: {
struct in6_addr *sa = &((struct sockaddr_in6 *)addr)->sin6_addr;
return !ipv6_addr_any(sa);
}
}
return 0;
}
static void nfs_parse_ipv4_address(char *string, size_t str_len,
struct sockaddr *sap, size_t *addr_len)
{
struct sockaddr_in *sin = (struct sockaddr_in *)sap;
u8 *addr = (u8 *)&sin->sin_addr.s_addr;
if (str_len <= INET_ADDRSTRLEN) {
dfprintk(MOUNT, "NFS: parsing IPv4 address %*s\n",
(int)str_len, string);
sin->sin_family = AF_INET;
*addr_len = sizeof(*sin);
if (in4_pton(string, str_len, addr, '\0', NULL))
return;
}
sap->sa_family = AF_UNSPEC;
*addr_len = 0;
}
#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
static int nfs_parse_ipv6_scope_id(const char *string, const size_t str_len,
const char *delim,
struct sockaddr_in6 *sin6)
{
char *p;
size_t len;
if ((string + str_len) == delim)
return 1;
if (*delim != IPV6_SCOPE_DELIMITER)
return 0;
if (!(ipv6_addr_type(&sin6->sin6_addr) & IPV6_ADDR_LINKLOCAL))
return 0;
len = (string + str_len) - delim - 1;
p = kstrndup(delim + 1, len, GFP_KERNEL);
if (p) {
unsigned long scope_id = 0;
struct net_device *dev;
dev = dev_get_by_name(&init_net, p);
if (dev != NULL) {
scope_id = dev->ifindex;
dev_put(dev);
} else {
if (strict_strtoul(p, 10, &scope_id) == 0) {
kfree(p);
return 0;
}
}
kfree(p);
sin6->sin6_scope_id = scope_id;
dfprintk(MOUNT, "NFS: IPv6 scope ID = %lu\n", scope_id);
return 1;
}
return 0;
}
static void nfs_parse_ipv6_address(char *string, size_t str_len,
struct sockaddr *sap, size_t *addr_len)
{
struct sockaddr_in6 *sin6 = (struct sockaddr_in6 *)sap;
u8 *addr = (u8 *)&sin6->sin6_addr.in6_u;
const char *delim;
if (str_len <= INET6_ADDRSTRLEN) {
dfprintk(MOUNT, "NFS: parsing IPv6 address %*s\n",
(int)str_len, string);
sin6->sin6_family = AF_INET6;
*addr_len = sizeof(*sin6);
if (in6_pton(string, str_len, addr,
IPV6_SCOPE_DELIMITER, &delim) != 0) {
if (nfs_parse_ipv6_scope_id(string, str_len,
delim, sin6) != 0)
return;
}
}
sap->sa_family = AF_UNSPEC;
*addr_len = 0;
}
#else
static void nfs_parse_ipv6_address(char *string, size_t str_len,
struct sockaddr *sap, size_t *addr_len)
{
sap->sa_family = AF_UNSPEC;
*addr_len = 0;
}
#endif
/*
* Construct a sockaddr based on the contents of a string that contains
* an IP address in presentation format.
*
* If there is a problem constructing the new sockaddr, set the address
* family to AF_UNSPEC.
*/
void nfs_parse_ip_address(char *string, size_t str_len,
struct sockaddr *sap, size_t *addr_len)
{
unsigned int i, colons;
colons = 0;
for (i = 0; i < str_len; i++)
if (string[i] == ':')
colons++;
if (colons >= 2)
nfs_parse_ipv6_address(string, str_len, sap, addr_len);
else
nfs_parse_ipv4_address(string, str_len, sap, addr_len);
}
/*
* Sanity check the NFS transport protocol.
*
*/
static void nfs_validate_transport_protocol(struct nfs_parsed_mount_data *mnt)
{
switch (mnt->nfs_server.protocol) {
case XPRT_TRANSPORT_UDP:
case XPRT_TRANSPORT_TCP:
case XPRT_TRANSPORT_RDMA:
break;
default:
mnt->nfs_server.protocol = XPRT_TRANSPORT_TCP;
}
}
/*
* For text based NFSv2/v3 mounts, the mount protocol transport default
* settings should depend upon the specified NFS transport.
*/
static void nfs_set_mount_transport_protocol(struct nfs_parsed_mount_data *mnt)
{
nfs_validate_transport_protocol(mnt);
if (mnt->mount_server.protocol == XPRT_TRANSPORT_UDP ||
mnt->mount_server.protocol == XPRT_TRANSPORT_TCP)
return;
switch (mnt->nfs_server.protocol) {
case XPRT_TRANSPORT_UDP:
mnt->mount_server.protocol = XPRT_TRANSPORT_UDP;
break;
case XPRT_TRANSPORT_TCP:
case XPRT_TRANSPORT_RDMA:
mnt->mount_server.protocol = XPRT_TRANSPORT_TCP;
}
}
/*
* Parse the value of the 'sec=' option.
*
* The flavor_len setting is for v4 mounts.
*/
static int nfs_parse_security_flavors(char *value,
struct nfs_parsed_mount_data *mnt)
{
substring_t args[MAX_OPT_ARGS];
dfprintk(MOUNT, "NFS: parsing sec=%s option\n", value);
switch (match_token(value, nfs_secflavor_tokens, args)) {
case Opt_sec_none:
mnt->auth_flavor_len = 0;
mnt->auth_flavors[0] = RPC_AUTH_NULL;
break;
case Opt_sec_sys:
mnt->auth_flavor_len = 0;
mnt->auth_flavors[0] = RPC_AUTH_UNIX;
break;
case Opt_sec_krb5:
mnt->auth_flavor_len = 1;
mnt->auth_flavors[0] = RPC_AUTH_GSS_KRB5;
break;
case Opt_sec_krb5i:
mnt->auth_flavor_len = 1;
mnt->auth_flavors[0] = RPC_AUTH_GSS_KRB5I;
break;
case Opt_sec_krb5p:
mnt->auth_flavor_len = 1;
mnt->auth_flavors[0] = RPC_AUTH_GSS_KRB5P;
break;
case Opt_sec_lkey:
mnt->auth_flavor_len = 1;
mnt->auth_flavors[0] = RPC_AUTH_GSS_LKEY;
break;
case Opt_sec_lkeyi:
mnt->auth_flavor_len = 1;
mnt->auth_flavors[0] = RPC_AUTH_GSS_LKEYI;
break;
case Opt_sec_lkeyp:
mnt->auth_flavor_len = 1;
mnt->auth_flavors[0] = RPC_AUTH_GSS_LKEYP;
break;
case Opt_sec_spkm:
mnt->auth_flavor_len = 1;
mnt->auth_flavors[0] = RPC_AUTH_GSS_SPKM;
break;
case Opt_sec_spkmi:
mnt->auth_flavor_len = 1;
mnt->auth_flavors[0] = RPC_AUTH_GSS_SPKMI;
break;
case Opt_sec_spkmp:
mnt->auth_flavor_len = 1;
mnt->auth_flavors[0] = RPC_AUTH_GSS_SPKMP;
break;
default:
return 0;
}
return 1;
}
/*
* Error-check and convert a string of mount options from user space into
NFS: Allow either strict or sloppy mount option parsing The kernel's NFS client mount option parser currently doesn't allow unrecognized or incorrect mount options. This prevents misspellings or incorrectly specified mount options from possibly causing silent data corruption. However, NFS mount options are not standardized, so different operating systems can use differently spelled mount options to support similar features, or can support mount options which no other operating system supports. "Sloppy" mount option parsing, which allows the parser to ignore any option it doesn't recognize, is needed to support automounters that often use maps that are shared between heterogenous operating systems. The legacy mount command ignores the validity of the values of mount options entirely, except for the "sec=" and "proto=" options. If an incorrect value is specified, the out-of-range value is passed to the kernel; if a value is specified that contains non-numeric characters, it appears as though the legacy mount command sets that option to zero (probably incorrect behavior in general). In any case, this sets a precedent which we will partially follow for the kernel mount option parser: + if "sloppy" is not set, the parser will be strict about both unrecognized options (same as legacy) and invalid option values (stricter than legacy) + if "sloppy" is set, the parser will ignore unrecognized options and invalid option values (same as legacy) An "invalid" option value in this case means that either the type (integer, short, or string) or sign (for integer values) of the specified value is incorrect. This patch does two things: it changes the NFS client's mount option parsing loop so that it parses the whole string instead of failing at the first unrecognized option or invalid option value. An unrecognized option or an invalid option value cause the option to be skipped. Then, the patch adds a "sloppy" mount option that allows the parsing to succeed anyway if there were any problems during parsing. When parsing a set of options is complete, if there are errors and "sloppy" was specified, return success anyway. Otherwise, only return success if there are no errors. Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2008-06-24 23:28:02 +00:00
* a data structure. The whole mount string is processed; bad options are
* skipped as they are encountered. If there were no errors, return 1;
* otherwise return 0 (zero).
*/
static int nfs_parse_mount_options(char *raw,
struct nfs_parsed_mount_data *mnt)
{
char *p, *string, *secdata;
int rc, sloppy = 0, invalid_option = 0;
if (!raw) {
dfprintk(MOUNT, "NFS: mount options string was NULL.\n");
return 1;
}
dfprintk(MOUNT, "NFS: nfs mount opts='%s'\n", raw);
secdata = alloc_secdata();
if (!secdata)
goto out_nomem;
rc = security_sb_copy_data(raw, secdata);
if (rc)
goto out_security_failure;
rc = security_sb_parse_opts_str(secdata, &mnt->lsm_opts);
if (rc)
goto out_security_failure;
free_secdata(secdata);
while ((p = strsep(&raw, ",")) != NULL) {
substring_t args[MAX_OPT_ARGS];
unsigned long option;
int int_option;
int token;
if (!*p)
continue;
dfprintk(MOUNT, "NFS: parsing nfs mount option '%s'\n", p);
token = match_token(p, nfs_mount_option_tokens, args);
switch (token) {
NFS: Allow either strict or sloppy mount option parsing The kernel's NFS client mount option parser currently doesn't allow unrecognized or incorrect mount options. This prevents misspellings or incorrectly specified mount options from possibly causing silent data corruption. However, NFS mount options are not standardized, so different operating systems can use differently spelled mount options to support similar features, or can support mount options which no other operating system supports. "Sloppy" mount option parsing, which allows the parser to ignore any option it doesn't recognize, is needed to support automounters that often use maps that are shared between heterogenous operating systems. The legacy mount command ignores the validity of the values of mount options entirely, except for the "sec=" and "proto=" options. If an incorrect value is specified, the out-of-range value is passed to the kernel; if a value is specified that contains non-numeric characters, it appears as though the legacy mount command sets that option to zero (probably incorrect behavior in general). In any case, this sets a precedent which we will partially follow for the kernel mount option parser: + if "sloppy" is not set, the parser will be strict about both unrecognized options (same as legacy) and invalid option values (stricter than legacy) + if "sloppy" is set, the parser will ignore unrecognized options and invalid option values (same as legacy) An "invalid" option value in this case means that either the type (integer, short, or string) or sign (for integer values) of the specified value is incorrect. This patch does two things: it changes the NFS client's mount option parsing loop so that it parses the whole string instead of failing at the first unrecognized option or invalid option value. An unrecognized option or an invalid option value cause the option to be skipped. Then, the patch adds a "sloppy" mount option that allows the parsing to succeed anyway if there were any problems during parsing. When parsing a set of options is complete, if there are errors and "sloppy" was specified, return success anyway. Otherwise, only return success if there are no errors. Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2008-06-24 23:28:02 +00:00
/*
* boolean options: foo/nofoo
*/
case Opt_soft:
mnt->flags |= NFS_MOUNT_SOFT;
break;
case Opt_hard:
mnt->flags &= ~NFS_MOUNT_SOFT;
break;
case Opt_posix:
mnt->flags |= NFS_MOUNT_POSIX;
break;
case Opt_noposix:
mnt->flags &= ~NFS_MOUNT_POSIX;
break;
case Opt_cto:
mnt->flags &= ~NFS_MOUNT_NOCTO;
break;
case Opt_nocto:
mnt->flags |= NFS_MOUNT_NOCTO;
break;
case Opt_ac:
mnt->flags &= ~NFS_MOUNT_NOAC;
break;
case Opt_noac:
mnt->flags |= NFS_MOUNT_NOAC;
break;
case Opt_lock:
mnt->flags &= ~NFS_MOUNT_NONLM;
break;
case Opt_nolock:
mnt->flags |= NFS_MOUNT_NONLM;
break;
case Opt_v2:
mnt->flags &= ~NFS_MOUNT_VER3;
break;
case Opt_v3:
mnt->flags |= NFS_MOUNT_VER3;
break;
case Opt_udp:
mnt->flags &= ~NFS_MOUNT_TCP;
mnt->nfs_server.protocol = XPRT_TRANSPORT_UDP;
break;
case Opt_tcp:
mnt->flags |= NFS_MOUNT_TCP;
mnt->nfs_server.protocol = XPRT_TRANSPORT_TCP;
break;
case Opt_rdma:
mnt->flags |= NFS_MOUNT_TCP; /* for side protocols */
mnt->nfs_server.protocol = XPRT_TRANSPORT_RDMA;
xprt_load_transport(p);
break;
case Opt_acl:
mnt->flags &= ~NFS_MOUNT_NOACL;
break;
case Opt_noacl:
mnt->flags |= NFS_MOUNT_NOACL;
break;
case Opt_rdirplus:
mnt->flags &= ~NFS_MOUNT_NORDIRPLUS;
break;
case Opt_nordirplus:
mnt->flags |= NFS_MOUNT_NORDIRPLUS;
break;
case Opt_sharecache:
mnt->flags &= ~NFS_MOUNT_UNSHARED;
break;
case Opt_nosharecache:
mnt->flags |= NFS_MOUNT_UNSHARED;
break;
NFS: add "[no]resvport" mount option The standard default security setting for NFS is AUTH_SYS. An NFS client connects to NFS servers via a privileged source port and a fixed standard destination port (2049). The client sends raw uid and gid numbers to identify users making NFS requests, and the server assumes an appropriate authority on the client has vetted these values because the source port is privileged. On Linux, by default in-kernel RPC services use a privileged port in the range between 650 and 1023 to avoid using source ports of well- known IP services. Using such a small range limits the number of NFS mount points and the number of unique NFS servers to which a client can connect concurrently. An NFS client can use unprivileged source ports to expand the range of source port numbers, allowing more concurrent server connections and more NFS mount points. Servers must explicitly allow NFS connections from unprivileged ports for this to work. In the past, bumping the value of the sunrpc.max_resvport sysctl on the client would permit the NFS client to use unprivileged ports. Bumping this setting also changes the maximum port number used by other in-kernel RPC services, some of which still required a port number less than 1023. This is exacerbated by the way source port numbers are chosen by the Linux RPC client, which starts at the top of the range and works downwards. It means that bumping the maximum means all RPC services requesting a source port will likely get an unprivileged port instead of a privileged one. Changing this setting effects all NFS mount points on a client. A sysadmin could not selectively choose which mount points would use non-privileged ports and which could not. Lastly, this mechanism of expanding the limit on the number of NFS mount points was entirely undocumented. To address the need for the NFS client to use a large range of source ports without interfering with the activity of other in-kernel RPC services, we introduce a new NFS mount option. This option explicitly tells only the NFS client to use a non-privileged source port when communicating with the NFS server for one specific mount point. This new mount option is called "resvport," like the similar NFS mount option on FreeBSD and Mac OS X. A sister patch for nfs-utils will be submitted that documents this new option in nfs(5). The default setting for this new mount option requires the NFS client to use a privileged port, as before. Explicitly specifying the "noresvport" mount option allows the NFS client to use an unprivileged source port for this mount point when connecting to the NFS server port. This mount option is supported only for text-based NFS mounts. [ Sidebar: it is widely known that security mechanisms based on the use of privileged source ports are ineffective. However, the NFS client can combine the use of unprivileged ports with the use of secure authentication mechanisms, such as Kerberos. This allows a large number of connections and mount points while ensuring a useful level of security. Eventually we may change the default setting for this option depending on the security flavor used for the mount. For example, if the mount is using only AUTH_SYS, then the default setting will be "resvport;" if the mount is using a strong security flavor such as krb5, the default setting will be "noresvport." ] Signed-off-by: Chuck Lever <chuck.lever@oracle.com> [Trond.Myklebust@netapp.com: Fixed a bug whereby nfs4_init_client() was being called with incorrect arguments.] Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2008-12-23 20:21:37 +00:00
case Opt_resvport:
mnt->flags &= ~NFS_MOUNT_NORESVPORT;
break;
case Opt_noresvport:
mnt->flags |= NFS_MOUNT_NORESVPORT;
break;
case Opt_fscache:
mnt->options |= NFS_OPTION_FSCACHE;
kfree(mnt->fscache_uniq);
mnt->fscache_uniq = NULL;
break;
case Opt_nofscache:
mnt->options &= ~NFS_OPTION_FSCACHE;
kfree(mnt->fscache_uniq);
mnt->fscache_uniq = NULL;
break;
case Opt_fscache_uniq:
string = match_strdup(args);
if (!string)
goto out_nomem;
kfree(mnt->fscache_uniq);
mnt->fscache_uniq = string;
mnt->options |= NFS_OPTION_FSCACHE;
break;
NFS: Allow either strict or sloppy mount option parsing The kernel's NFS client mount option parser currently doesn't allow unrecognized or incorrect mount options. This prevents misspellings or incorrectly specified mount options from possibly causing silent data corruption. However, NFS mount options are not standardized, so different operating systems can use differently spelled mount options to support similar features, or can support mount options which no other operating system supports. "Sloppy" mount option parsing, which allows the parser to ignore any option it doesn't recognize, is needed to support automounters that often use maps that are shared between heterogenous operating systems. The legacy mount command ignores the validity of the values of mount options entirely, except for the "sec=" and "proto=" options. If an incorrect value is specified, the out-of-range value is passed to the kernel; if a value is specified that contains non-numeric characters, it appears as though the legacy mount command sets that option to zero (probably incorrect behavior in general). In any case, this sets a precedent which we will partially follow for the kernel mount option parser: + if "sloppy" is not set, the parser will be strict about both unrecognized options (same as legacy) and invalid option values (stricter than legacy) + if "sloppy" is set, the parser will ignore unrecognized options and invalid option values (same as legacy) An "invalid" option value in this case means that either the type (integer, short, or string) or sign (for integer values) of the specified value is incorrect. This patch does two things: it changes the NFS client's mount option parsing loop so that it parses the whole string instead of failing at the first unrecognized option or invalid option value. An unrecognized option or an invalid option value cause the option to be skipped. Then, the patch adds a "sloppy" mount option that allows the parsing to succeed anyway if there were any problems during parsing. When parsing a set of options is complete, if there are errors and "sloppy" was specified, return success anyway. Otherwise, only return success if there are no errors. Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2008-06-24 23:28:02 +00:00
/*
* options that take numeric values
*/
case Opt_port:
string = match_strdup(args);
if (string == NULL)
goto out_nomem;
rc = strict_strtoul(string, 10, &option);
kfree(string);
if (rc != 0 || option > USHORT_MAX)
goto out_invalid_value;
mnt->nfs_server.port = option;
break;
case Opt_rsize:
string = match_strdup(args);
if (string == NULL)
goto out_nomem;
rc = strict_strtoul(string, 10, &option);
kfree(string);
if (rc != 0)
goto out_invalid_value;
mnt->rsize = option;
break;
case Opt_wsize:
string = match_strdup(args);
if (string == NULL)
goto out_nomem;
rc = strict_strtoul(string, 10, &option);
kfree(string);
if (rc != 0)
goto out_invalid_value;
mnt->wsize = option;
break;
case Opt_bsize:
string = match_strdup(args);
if (string == NULL)
goto out_nomem;
rc = strict_strtoul(string, 10, &option);
kfree(string);
if (rc != 0)
goto out_invalid_value;
mnt->bsize = option;
break;
case Opt_timeo:
string = match_strdup(args);
if (string == NULL)
goto out_nomem;
rc = strict_strtoul(string, 10, &option);
kfree(string);
if (rc != 0 || option == 0)
goto out_invalid_value;
mnt->timeo = option;
break;
case Opt_retrans:
string = match_strdup(args);
if (string == NULL)
goto out_nomem;
rc = strict_strtoul(string, 10, &option);
kfree(string);
if (rc != 0 || option == 0)
goto out_invalid_value;
mnt->retrans = option;
break;
case Opt_acregmin:
string = match_strdup(args);
if (string == NULL)
goto out_nomem;
rc = strict_strtoul(string, 10, &option);
kfree(string);
if (rc != 0)
goto out_invalid_value;
mnt->acregmin = option;
break;
case Opt_acregmax:
string = match_strdup(args);
if (string == NULL)
goto out_nomem;
rc = strict_strtoul(string, 10, &option);
kfree(string);
if (rc != 0)
goto out_invalid_value;
mnt->acregmax = option;
break;
case Opt_acdirmin:
string = match_strdup(args);
if (string == NULL)
goto out_nomem;
rc = strict_strtoul(string, 10, &option);
kfree(string);
if (rc != 0)
goto out_invalid_value;
mnt->acdirmin = option;
break;
case Opt_acdirmax:
string = match_strdup(args);
if (string == NULL)
goto out_nomem;
rc = strict_strtoul(string, 10, &option);
kfree(string);
if (rc != 0)
goto out_invalid_value;
mnt->acdirmax = option;
break;
case Opt_actimeo:
string = match_strdup(args);
if (string == NULL)
goto out_nomem;
rc = strict_strtoul(string, 10, &option);
kfree(string);
if (rc != 0)
goto out_invalid_value;
mnt->acregmin = mnt->acregmax =
mnt->acdirmin = mnt->acdirmax = option;
break;
case Opt_namelen:
string = match_strdup(args);
if (string == NULL)
goto out_nomem;
rc = strict_strtoul(string, 10, &option);
kfree(string);
if (rc != 0)
goto out_invalid_value;
mnt->namlen = option;
break;
case Opt_mountport:
string = match_strdup(args);
if (string == NULL)
goto out_nomem;
rc = strict_strtoul(string, 10, &option);
kfree(string);
if (rc != 0 || option > USHORT_MAX)
goto out_invalid_value;
mnt->mount_server.port = option;
break;
case Opt_mountvers:
string = match_strdup(args);
if (string == NULL)
goto out_nomem;
rc = strict_strtoul(string, 10, &option);
kfree(string);
if (rc != 0 ||
NFS: Allow either strict or sloppy mount option parsing The kernel's NFS client mount option parser currently doesn't allow unrecognized or incorrect mount options. This prevents misspellings or incorrectly specified mount options from possibly causing silent data corruption. However, NFS mount options are not standardized, so different operating systems can use differently spelled mount options to support similar features, or can support mount options which no other operating system supports. "Sloppy" mount option parsing, which allows the parser to ignore any option it doesn't recognize, is needed to support automounters that often use maps that are shared between heterogenous operating systems. The legacy mount command ignores the validity of the values of mount options entirely, except for the "sec=" and "proto=" options. If an incorrect value is specified, the out-of-range value is passed to the kernel; if a value is specified that contains non-numeric characters, it appears as though the legacy mount command sets that option to zero (probably incorrect behavior in general). In any case, this sets a precedent which we will partially follow for the kernel mount option parser: + if "sloppy" is not set, the parser will be strict about both unrecognized options (same as legacy) and invalid option values (stricter than legacy) + if "sloppy" is set, the parser will ignore unrecognized options and invalid option values (same as legacy) An "invalid" option value in this case means that either the type (integer, short, or string) or sign (for integer values) of the specified value is incorrect. This patch does two things: it changes the NFS client's mount option parsing loop so that it parses the whole string instead of failing at the first unrecognized option or invalid option value. An unrecognized option or an invalid option value cause the option to be skipped. Then, the patch adds a "sloppy" mount option that allows the parsing to succeed anyway if there were any problems during parsing. When parsing a set of options is complete, if there are errors and "sloppy" was specified, return success anyway. Otherwise, only return success if there are no errors. Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2008-06-24 23:28:02 +00:00
option < NFS_MNT_VERSION ||
option > NFS_MNT3_VERSION)
goto out_invalid_value;
mnt->mount_server.version = option;
break;
case Opt_nfsvers:
string = match_strdup(args);
if (string == NULL)
goto out_nomem;
rc = strict_strtoul(string, 10, &option);
kfree(string);
if (rc != 0)
goto out_invalid_value;
switch (option) {
NFS: Allow either strict or sloppy mount option parsing The kernel's NFS client mount option parser currently doesn't allow unrecognized or incorrect mount options. This prevents misspellings or incorrectly specified mount options from possibly causing silent data corruption. However, NFS mount options are not standardized, so different operating systems can use differently spelled mount options to support similar features, or can support mount options which no other operating system supports. "Sloppy" mount option parsing, which allows the parser to ignore any option it doesn't recognize, is needed to support automounters that often use maps that are shared between heterogenous operating systems. The legacy mount command ignores the validity of the values of mount options entirely, except for the "sec=" and "proto=" options. If an incorrect value is specified, the out-of-range value is passed to the kernel; if a value is specified that contains non-numeric characters, it appears as though the legacy mount command sets that option to zero (probably incorrect behavior in general). In any case, this sets a precedent which we will partially follow for the kernel mount option parser: + if "sloppy" is not set, the parser will be strict about both unrecognized options (same as legacy) and invalid option values (stricter than legacy) + if "sloppy" is set, the parser will ignore unrecognized options and invalid option values (same as legacy) An "invalid" option value in this case means that either the type (integer, short, or string) or sign (for integer values) of the specified value is incorrect. This patch does two things: it changes the NFS client's mount option parsing loop so that it parses the whole string instead of failing at the first unrecognized option or invalid option value. An unrecognized option or an invalid option value cause the option to be skipped. Then, the patch adds a "sloppy" mount option that allows the parsing to succeed anyway if there were any problems during parsing. When parsing a set of options is complete, if there are errors and "sloppy" was specified, return success anyway. Otherwise, only return success if there are no errors. Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2008-06-24 23:28:02 +00:00
case NFS2_VERSION:
mnt->flags &= ~NFS_MOUNT_VER3;
break;
NFS: Allow either strict or sloppy mount option parsing The kernel's NFS client mount option parser currently doesn't allow unrecognized or incorrect mount options. This prevents misspellings or incorrectly specified mount options from possibly causing silent data corruption. However, NFS mount options are not standardized, so different operating systems can use differently spelled mount options to support similar features, or can support mount options which no other operating system supports. "Sloppy" mount option parsing, which allows the parser to ignore any option it doesn't recognize, is needed to support automounters that often use maps that are shared between heterogenous operating systems. The legacy mount command ignores the validity of the values of mount options entirely, except for the "sec=" and "proto=" options. If an incorrect value is specified, the out-of-range value is passed to the kernel; if a value is specified that contains non-numeric characters, it appears as though the legacy mount command sets that option to zero (probably incorrect behavior in general). In any case, this sets a precedent which we will partially follow for the kernel mount option parser: + if "sloppy" is not set, the parser will be strict about both unrecognized options (same as legacy) and invalid option values (stricter than legacy) + if "sloppy" is set, the parser will ignore unrecognized options and invalid option values (same as legacy) An "invalid" option value in this case means that either the type (integer, short, or string) or sign (for integer values) of the specified value is incorrect. This patch does two things: it changes the NFS client's mount option parsing loop so that it parses the whole string instead of failing at the first unrecognized option or invalid option value. An unrecognized option or an invalid option value cause the option to be skipped. Then, the patch adds a "sloppy" mount option that allows the parsing to succeed anyway if there were any problems during parsing. When parsing a set of options is complete, if there are errors and "sloppy" was specified, return success anyway. Otherwise, only return success if there are no errors. Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2008-06-24 23:28:02 +00:00
case NFS3_VERSION:
mnt->flags |= NFS_MOUNT_VER3;
break;
default:
goto out_invalid_value;
}
break;
case Opt_minorversion:
if (match_int(args, &int_option))
return 0;
if (int_option < 0 || int_option > NFS4_MAX_MINOR_VERSION)
return 0;
mnt->minorversion = int_option;
break;
NFS: Allow either strict or sloppy mount option parsing The kernel's NFS client mount option parser currently doesn't allow unrecognized or incorrect mount options. This prevents misspellings or incorrectly specified mount options from possibly causing silent data corruption. However, NFS mount options are not standardized, so different operating systems can use differently spelled mount options to support similar features, or can support mount options which no other operating system supports. "Sloppy" mount option parsing, which allows the parser to ignore any option it doesn't recognize, is needed to support automounters that often use maps that are shared between heterogenous operating systems. The legacy mount command ignores the validity of the values of mount options entirely, except for the "sec=" and "proto=" options. If an incorrect value is specified, the out-of-range value is passed to the kernel; if a value is specified that contains non-numeric characters, it appears as though the legacy mount command sets that option to zero (probably incorrect behavior in general). In any case, this sets a precedent which we will partially follow for the kernel mount option parser: + if "sloppy" is not set, the parser will be strict about both unrecognized options (same as legacy) and invalid option values (stricter than legacy) + if "sloppy" is set, the parser will ignore unrecognized options and invalid option values (same as legacy) An "invalid" option value in this case means that either the type (integer, short, or string) or sign (for integer values) of the specified value is incorrect. This patch does two things: it changes the NFS client's mount option parsing loop so that it parses the whole string instead of failing at the first unrecognized option or invalid option value. An unrecognized option or an invalid option value cause the option to be skipped. Then, the patch adds a "sloppy" mount option that allows the parsing to succeed anyway if there were any problems during parsing. When parsing a set of options is complete, if there are errors and "sloppy" was specified, return success anyway. Otherwise, only return success if there are no errors. Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2008-06-24 23:28:02 +00:00
/*
* options that take text values
*/
case Opt_sec:
string = match_strdup(args);
if (string == NULL)
goto out_nomem;
rc = nfs_parse_security_flavors(string, mnt);
kfree(string);
NFS: Allow either strict or sloppy mount option parsing The kernel's NFS client mount option parser currently doesn't allow unrecognized or incorrect mount options. This prevents misspellings or incorrectly specified mount options from possibly causing silent data corruption. However, NFS mount options are not standardized, so different operating systems can use differently spelled mount options to support similar features, or can support mount options which no other operating system supports. "Sloppy" mount option parsing, which allows the parser to ignore any option it doesn't recognize, is needed to support automounters that often use maps that are shared between heterogenous operating systems. The legacy mount command ignores the validity of the values of mount options entirely, except for the "sec=" and "proto=" options. If an incorrect value is specified, the out-of-range value is passed to the kernel; if a value is specified that contains non-numeric characters, it appears as though the legacy mount command sets that option to zero (probably incorrect behavior in general). In any case, this sets a precedent which we will partially follow for the kernel mount option parser: + if "sloppy" is not set, the parser will be strict about both unrecognized options (same as legacy) and invalid option values (stricter than legacy) + if "sloppy" is set, the parser will ignore unrecognized options and invalid option values (same as legacy) An "invalid" option value in this case means that either the type (integer, short, or string) or sign (for integer values) of the specified value is incorrect. This patch does two things: it changes the NFS client's mount option parsing loop so that it parses the whole string instead of failing at the first unrecognized option or invalid option value. An unrecognized option or an invalid option value cause the option to be skipped. Then, the patch adds a "sloppy" mount option that allows the parsing to succeed anyway if there were any problems during parsing. When parsing a set of options is complete, if there are errors and "sloppy" was specified, return success anyway. Otherwise, only return success if there are no errors. Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2008-06-24 23:28:02 +00:00
if (!rc) {
dfprintk(MOUNT, "NFS: unrecognized "
"security flavor\n");
return 0;
NFS: Allow either strict or sloppy mount option parsing The kernel's NFS client mount option parser currently doesn't allow unrecognized or incorrect mount options. This prevents misspellings or incorrectly specified mount options from possibly causing silent data corruption. However, NFS mount options are not standardized, so different operating systems can use differently spelled mount options to support similar features, or can support mount options which no other operating system supports. "Sloppy" mount option parsing, which allows the parser to ignore any option it doesn't recognize, is needed to support automounters that often use maps that are shared between heterogenous operating systems. The legacy mount command ignores the validity of the values of mount options entirely, except for the "sec=" and "proto=" options. If an incorrect value is specified, the out-of-range value is passed to the kernel; if a value is specified that contains non-numeric characters, it appears as though the legacy mount command sets that option to zero (probably incorrect behavior in general). In any case, this sets a precedent which we will partially follow for the kernel mount option parser: + if "sloppy" is not set, the parser will be strict about both unrecognized options (same as legacy) and invalid option values (stricter than legacy) + if "sloppy" is set, the parser will ignore unrecognized options and invalid option values (same as legacy) An "invalid" option value in this case means that either the type (integer, short, or string) or sign (for integer values) of the specified value is incorrect. This patch does two things: it changes the NFS client's mount option parsing loop so that it parses the whole string instead of failing at the first unrecognized option or invalid option value. An unrecognized option or an invalid option value cause the option to be skipped. Then, the patch adds a "sloppy" mount option that allows the parsing to succeed anyway if there were any problems during parsing. When parsing a set of options is complete, if there are errors and "sloppy" was specified, return success anyway. Otherwise, only return success if there are no errors. Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2008-06-24 23:28:02 +00:00
}
break;
case Opt_proto:
string = match_strdup(args);
if (string == NULL)
goto out_nomem;
token = match_token(string,
nfs_xprt_protocol_tokens, args);
switch (token) {
case Opt_xprt_udp:
mnt->flags &= ~NFS_MOUNT_TCP;
mnt->nfs_server.protocol = XPRT_TRANSPORT_UDP;
kfree(string);
break;
case Opt_xprt_tcp:
mnt->flags |= NFS_MOUNT_TCP;
mnt->nfs_server.protocol = XPRT_TRANSPORT_TCP;
kfree(string);
break;
case Opt_xprt_rdma:
/* vector side protocols to TCP */
mnt->flags |= NFS_MOUNT_TCP;
mnt->nfs_server.protocol = XPRT_TRANSPORT_RDMA;
xprt_load_transport(string);
kfree(string);
break;
default:
NFS: Allow either strict or sloppy mount option parsing The kernel's NFS client mount option parser currently doesn't allow unrecognized or incorrect mount options. This prevents misspellings or incorrectly specified mount options from possibly causing silent data corruption. However, NFS mount options are not standardized, so different operating systems can use differently spelled mount options to support similar features, or can support mount options which no other operating system supports. "Sloppy" mount option parsing, which allows the parser to ignore any option it doesn't recognize, is needed to support automounters that often use maps that are shared between heterogenous operating systems. The legacy mount command ignores the validity of the values of mount options entirely, except for the "sec=" and "proto=" options. If an incorrect value is specified, the out-of-range value is passed to the kernel; if a value is specified that contains non-numeric characters, it appears as though the legacy mount command sets that option to zero (probably incorrect behavior in general). In any case, this sets a precedent which we will partially follow for the kernel mount option parser: + if "sloppy" is not set, the parser will be strict about both unrecognized options (same as legacy) and invalid option values (stricter than legacy) + if "sloppy" is set, the parser will ignore unrecognized options and invalid option values (same as legacy) An "invalid" option value in this case means that either the type (integer, short, or string) or sign (for integer values) of the specified value is incorrect. This patch does two things: it changes the NFS client's mount option parsing loop so that it parses the whole string instead of failing at the first unrecognized option or invalid option value. An unrecognized option or an invalid option value cause the option to be skipped. Then, the patch adds a "sloppy" mount option that allows the parsing to succeed anyway if there were any problems during parsing. When parsing a set of options is complete, if there are errors and "sloppy" was specified, return success anyway. Otherwise, only return success if there are no errors. Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2008-06-24 23:28:02 +00:00
dfprintk(MOUNT, "NFS: unrecognized "
"transport protocol\n");
return 0;
}
break;
case Opt_mountproto:
string = match_strdup(args);
if (string == NULL)
goto out_nomem;
token = match_token(string,
nfs_xprt_protocol_tokens, args);
kfree(string);
switch (token) {
case Opt_xprt_udp:
mnt->mount_server.protocol = XPRT_TRANSPORT_UDP;
break;
case Opt_xprt_tcp:
mnt->mount_server.protocol = XPRT_TRANSPORT_TCP;
break;
case Opt_xprt_rdma: /* not used for side protocols */
default:
NFS: Allow either strict or sloppy mount option parsing The kernel's NFS client mount option parser currently doesn't allow unrecognized or incorrect mount options. This prevents misspellings or incorrectly specified mount options from possibly causing silent data corruption. However, NFS mount options are not standardized, so different operating systems can use differently spelled mount options to support similar features, or can support mount options which no other operating system supports. "Sloppy" mount option parsing, which allows the parser to ignore any option it doesn't recognize, is needed to support automounters that often use maps that are shared between heterogenous operating systems. The legacy mount command ignores the validity of the values of mount options entirely, except for the "sec=" and "proto=" options. If an incorrect value is specified, the out-of-range value is passed to the kernel; if a value is specified that contains non-numeric characters, it appears as though the legacy mount command sets that option to zero (probably incorrect behavior in general). In any case, this sets a precedent which we will partially follow for the kernel mount option parser: + if "sloppy" is not set, the parser will be strict about both unrecognized options (same as legacy) and invalid option values (stricter than legacy) + if "sloppy" is set, the parser will ignore unrecognized options and invalid option values (same as legacy) An "invalid" option value in this case means that either the type (integer, short, or string) or sign (for integer values) of the specified value is incorrect. This patch does two things: it changes the NFS client's mount option parsing loop so that it parses the whole string instead of failing at the first unrecognized option or invalid option value. An unrecognized option or an invalid option value cause the option to be skipped. Then, the patch adds a "sloppy" mount option that allows the parsing to succeed anyway if there were any problems during parsing. When parsing a set of options is complete, if there are errors and "sloppy" was specified, return success anyway. Otherwise, only return success if there are no errors. Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2008-06-24 23:28:02 +00:00
dfprintk(MOUNT, "NFS: unrecognized "
"transport protocol\n");
return 0;
}
break;
case Opt_addr:
string = match_strdup(args);
if (string == NULL)
goto out_nomem;
nfs_parse_ip_address(string, strlen(string),
(struct sockaddr *)
&mnt->nfs_server.address,
&mnt->nfs_server.addrlen);
kfree(string);
break;
case Opt_clientaddr:
string = match_strdup(args);
if (string == NULL)
goto out_nomem;
kfree(mnt->client_address);
mnt->client_address = string;
break;
case Opt_mounthost:
string = match_strdup(args);
if (string == NULL)
goto out_nomem;
kfree(mnt->mount_server.hostname);
mnt->mount_server.hostname = string;
break;
case Opt_mountaddr:
string = match_strdup(args);
if (string == NULL)
goto out_nomem;
nfs_parse_ip_address(string, strlen(string),
(struct sockaddr *)
&mnt->mount_server.address,
&mnt->mount_server.addrlen);
kfree(string);
break;
case Opt_lookupcache:
string = match_strdup(args);
if (string == NULL)
goto out_nomem;
token = match_token(string,
nfs_lookupcache_tokens, args);
kfree(string);
switch (token) {
case Opt_lookupcache_all:
mnt->flags &= ~(NFS_MOUNT_LOOKUP_CACHE_NONEG|NFS_MOUNT_LOOKUP_CACHE_NONE);
break;
case Opt_lookupcache_positive:
mnt->flags &= ~NFS_MOUNT_LOOKUP_CACHE_NONE;
mnt->flags |= NFS_MOUNT_LOOKUP_CACHE_NONEG;
break;
case Opt_lookupcache_none:
mnt->flags |= NFS_MOUNT_LOOKUP_CACHE_NONEG|NFS_MOUNT_LOOKUP_CACHE_NONE;
break;
default:
dfprintk(MOUNT, "NFS: invalid "
"lookupcache argument\n");
return 0;
};
break;
NFS: Allow either strict or sloppy mount option parsing The kernel's NFS client mount option parser currently doesn't allow unrecognized or incorrect mount options. This prevents misspellings or incorrectly specified mount options from possibly causing silent data corruption. However, NFS mount options are not standardized, so different operating systems can use differently spelled mount options to support similar features, or can support mount options which no other operating system supports. "Sloppy" mount option parsing, which allows the parser to ignore any option it doesn't recognize, is needed to support automounters that often use maps that are shared between heterogenous operating systems. The legacy mount command ignores the validity of the values of mount options entirely, except for the "sec=" and "proto=" options. If an incorrect value is specified, the out-of-range value is passed to the kernel; if a value is specified that contains non-numeric characters, it appears as though the legacy mount command sets that option to zero (probably incorrect behavior in general). In any case, this sets a precedent which we will partially follow for the kernel mount option parser: + if "sloppy" is not set, the parser will be strict about both unrecognized options (same as legacy) and invalid option values (stricter than legacy) + if "sloppy" is set, the parser will ignore unrecognized options and invalid option values (same as legacy) An "invalid" option value in this case means that either the type (integer, short, or string) or sign (for integer values) of the specified value is incorrect. This patch does two things: it changes the NFS client's mount option parsing loop so that it parses the whole string instead of failing at the first unrecognized option or invalid option value. An unrecognized option or an invalid option value cause the option to be skipped. Then, the patch adds a "sloppy" mount option that allows the parsing to succeed anyway if there were any problems during parsing. When parsing a set of options is complete, if there are errors and "sloppy" was specified, return success anyway. Otherwise, only return success if there are no errors. Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2008-06-24 23:28:02 +00:00
/*
* Special options
*/
case Opt_sloppy:
sloppy = 1;
dfprintk(MOUNT, "NFS: relaxing parsing rules\n");
break;
case Opt_userspace:
case Opt_deprecated:
dfprintk(MOUNT, "NFS: ignoring mount option "
"'%s'\n", p);
break;
default:
invalid_option = 1;
NFS: Allow either strict or sloppy mount option parsing The kernel's NFS client mount option parser currently doesn't allow unrecognized or incorrect mount options. This prevents misspellings or incorrectly specified mount options from possibly causing silent data corruption. However, NFS mount options are not standardized, so different operating systems can use differently spelled mount options to support similar features, or can support mount options which no other operating system supports. "Sloppy" mount option parsing, which allows the parser to ignore any option it doesn't recognize, is needed to support automounters that often use maps that are shared between heterogenous operating systems. The legacy mount command ignores the validity of the values of mount options entirely, except for the "sec=" and "proto=" options. If an incorrect value is specified, the out-of-range value is passed to the kernel; if a value is specified that contains non-numeric characters, it appears as though the legacy mount command sets that option to zero (probably incorrect behavior in general). In any case, this sets a precedent which we will partially follow for the kernel mount option parser: + if "sloppy" is not set, the parser will be strict about both unrecognized options (same as legacy) and invalid option values (stricter than legacy) + if "sloppy" is set, the parser will ignore unrecognized options and invalid option values (same as legacy) An "invalid" option value in this case means that either the type (integer, short, or string) or sign (for integer values) of the specified value is incorrect. This patch does two things: it changes the NFS client's mount option parsing loop so that it parses the whole string instead of failing at the first unrecognized option or invalid option value. An unrecognized option or an invalid option value cause the option to be skipped. Then, the patch adds a "sloppy" mount option that allows the parsing to succeed anyway if there were any problems during parsing. When parsing a set of options is complete, if there are errors and "sloppy" was specified, return success anyway. Otherwise, only return success if there are no errors. Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2008-06-24 23:28:02 +00:00
dfprintk(MOUNT, "NFS: unrecognized mount option "
"'%s'\n", p);
}
}
if (!sloppy && invalid_option)
return 0;
return 1;
out_invalid_value:
printk(KERN_INFO "NFS: bad mount option value specified: %s \n", p);
return 0;
out_nomem:
printk(KERN_INFO "NFS: not enough memory to parse option\n");
return 0;
out_security_failure:
free_secdata(secdata);
printk(KERN_INFO "NFS: security options invalid: %d\n", rc);
return 0;
}
/*
* Use the remote server's MOUNT service to request the NFS file handle
* corresponding to the provided path.
*/
static int nfs_try_mount(struct nfs_parsed_mount_data *args,
struct nfs_fh *root_fh)
{
unsigned int auth_flavor_len = 0;
struct nfs_mount_request request = {
.sap = (struct sockaddr *)
&args->mount_server.address,
.dirpath = args->nfs_server.export_path,
.protocol = args->mount_server.protocol,
.fh = root_fh,
.noresvport = args->flags & NFS_MOUNT_NORESVPORT,
.auth_flav_len = &auth_flavor_len,
};
int status;
if (args->mount_server.version == 0) {
if (args->flags & NFS_MOUNT_VER3)
args->mount_server.version = NFS_MNT3_VERSION;
else
args->mount_server.version = NFS_MNT_VERSION;
}
request.version = args->mount_server.version;
if (args->mount_server.hostname)
request.hostname = args->mount_server.hostname;
else
request.hostname = args->nfs_server.hostname;
/*
* Construct the mount server's address.
*/
if (args->mount_server.address.ss_family == AF_UNSPEC) {
memcpy(request.sap, &args->nfs_server.address,
args->nfs_server.addrlen);
args->mount_server.addrlen = args->nfs_server.addrlen;
}
request.salen = args->mount_server.addrlen;
/*
* autobind will be used if mount_server.port == 0
*/
nfs_set_port(request.sap, args->mount_server.port);
/*
* Now ask the mount server to map our export path
* to a file handle.
*/
status = nfs_mount(&request);
if (status == 0)
return 0;
dfprintk(MOUNT, "NFS: unable to mount server %s, error %d\n",
request.hostname, status);
return status;
}
static int nfs_parse_simple_hostname(const char *dev_name,
char **hostname, size_t maxnamlen,
char **export_path, size_t maxpathlen)
{
size_t len;
char *colon, *comma;
colon = strchr(dev_name, ':');
if (colon == NULL)
goto out_bad_devname;
len = colon - dev_name;
if (len > maxnamlen)
goto out_hostname;
/* N.B. caller will free nfs_server.hostname in all cases */
*hostname = kstrndup(dev_name, len, GFP_KERNEL);
if (!*hostname)
goto out_nomem;
/* kill possible hostname list: not supported */
comma = strchr(*hostname, ',');
if (comma != NULL) {
if (comma == *hostname)
goto out_bad_devname;
*comma = '\0';
}
colon++;
len = strlen(colon);
if (len > maxpathlen)
goto out_path;
*export_path = kstrndup(colon, len, GFP_KERNEL);
if (!*export_path)
goto out_nomem;
dfprintk(MOUNT, "NFS: MNTPATH: '%s'\n", *export_path);
return 0;
out_bad_devname:
dfprintk(MOUNT, "NFS: device name not in host:path format\n");
return -EINVAL;
out_nomem:
dfprintk(MOUNT, "NFS: not enough memory to parse device name\n");
return -ENOMEM;
out_hostname:
dfprintk(MOUNT, "NFS: server hostname too long\n");
return -ENAMETOOLONG;
out_path:
dfprintk(MOUNT, "NFS: export pathname too long\n");
return -ENAMETOOLONG;
}
/*
* Hostname has square brackets around it because it contains one or
* more colons. We look for the first closing square bracket, and a
* colon must follow it.
*/
static int nfs_parse_protected_hostname(const char *dev_name,
char **hostname, size_t maxnamlen,
char **export_path, size_t maxpathlen)
{
size_t len;
char *start, *end;
start = (char *)(dev_name + 1);
end = strchr(start, ']');
if (end == NULL)
goto out_bad_devname;
if (*(end + 1) != ':')
goto out_bad_devname;
len = end - start;
if (len > maxnamlen)
goto out_hostname;
/* N.B. caller will free nfs_server.hostname in all cases */
*hostname = kstrndup(start, len, GFP_KERNEL);
if (*hostname == NULL)
goto out_nomem;
end += 2;
len = strlen(end);
if (len > maxpathlen)
goto out_path;
*export_path = kstrndup(end, len, GFP_KERNEL);
if (!*export_path)
goto out_nomem;
return 0;
out_bad_devname:
dfprintk(MOUNT, "NFS: device name not in host:path format\n");
return -EINVAL;
out_nomem:
dfprintk(MOUNT, "NFS: not enough memory to parse device name\n");
return -ENOMEM;
out_hostname:
dfprintk(MOUNT, "NFS: server hostname too long\n");
return -ENAMETOOLONG;
out_path:
dfprintk(MOUNT, "NFS: export pathname too long\n");
return -ENAMETOOLONG;
}
/*
* Split "dev_name" into "hostname:export_path".
*
* The leftmost colon demarks the split between the server's hostname
* and the export path. If the hostname starts with a left square
* bracket, then it may contain colons.
*
* Note: caller frees hostname and export path, even on error.
*/
static int nfs_parse_devname(const char *dev_name,
char **hostname, size_t maxnamlen,
char **export_path, size_t maxpathlen)
{
if (*dev_name == '[')
return nfs_parse_protected_hostname(dev_name,
hostname, maxnamlen,
export_path, maxpathlen);
return nfs_parse_simple_hostname(dev_name,
hostname, maxnamlen,
export_path, maxpathlen);
}
/*
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
* Validate the NFS2/NFS3 mount data
* - fills in the mount root filehandle
*
* For option strings, user space handles the following behaviors:
*
* + DNS: mapping server host name to IP address ("addr=" option)
*
* + failure mode: how to behave if a mount request can't be handled
* immediately ("fg/bg" option)
*
* + retry: how often to retry a mount request ("retry=" option)
*
* + breaking back: trying proto=udp after proto=tcp, v2 after v3,
* mountproto=tcp after mountproto=udp, and so on
*/
static int nfs_validate_mount_data(void *options,
struct nfs_parsed_mount_data *args,
struct nfs_fh *mntfh,
const char *dev_name)
{
struct nfs_mount_data *data = (struct nfs_mount_data *)options;
if (data == NULL)
goto out_no_data;
args->flags = (NFS_MOUNT_VER3 | NFS_MOUNT_TCP);
args->rsize = NFS_MAX_FILE_IO_SIZE;
args->wsize = NFS_MAX_FILE_IO_SIZE;
args->acregmin = NFS_DEF_ACREGMIN;
args->acregmax = NFS_DEF_ACREGMAX;
args->acdirmin = NFS_DEF_ACDIRMIN;
args->acdirmax = NFS_DEF_ACDIRMAX;
args->mount_server.port = 0; /* autobind unless user sets port */
args->nfs_server.port = 0; /* autobind unless user sets port */
args->nfs_server.protocol = XPRT_TRANSPORT_TCP;
args->auth_flavors[0] = RPC_AUTH_UNIX;
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
switch (data->version) {
case 1:
data->namlen = 0;
case 2:
data->bsize = 0;
case 3:
if (data->flags & NFS_MOUNT_VER3)
goto out_no_v3;
data->root.size = NFS2_FHSIZE;
memcpy(data->root.data, data->old_root.data, NFS2_FHSIZE);
case 4:
if (data->flags & NFS_MOUNT_SECFLAVOUR)
goto out_no_sec;
case 5:
memset(data->context, 0, sizeof(data->context));
case 6:
if (data->flags & NFS_MOUNT_VER3) {
if (data->root.size > NFS3_FHSIZE || data->root.size == 0)
goto out_invalid_fh;
mntfh->size = data->root.size;
} else
mntfh->size = NFS2_FHSIZE;
memcpy(mntfh->data, data->root.data, mntfh->size);
if (mntfh->size < sizeof(mntfh->data))
memset(mntfh->data + mntfh->size, 0,
sizeof(mntfh->data) - mntfh->size);
/*
* Translate to nfs_parsed_mount_data, which nfs_fill_super
* can deal with.
*/
args->flags = data->flags & NFS_MOUNT_FLAGMASK;
args->rsize = data->rsize;
args->wsize = data->wsize;
args->timeo = data->timeo;
args->retrans = data->retrans;
args->acregmin = data->acregmin;
args->acregmax = data->acregmax;
args->acdirmin = data->acdirmin;
args->acdirmax = data->acdirmax;
memcpy(&args->nfs_server.address, &data->addr,
sizeof(data->addr));
args->nfs_server.addrlen = sizeof(data->addr);
if (!nfs_verify_server_address((struct sockaddr *)
&args->nfs_server.address))
goto out_no_address;
if (!(data->flags & NFS_MOUNT_TCP))
args->nfs_server.protocol = XPRT_TRANSPORT_UDP;
/* N.B. caller will free nfs_server.hostname in all cases */
args->nfs_server.hostname = kstrdup(data->hostname, GFP_KERNEL);
args->namlen = data->namlen;
args->bsize = data->bsize;
if (data->flags & NFS_MOUNT_SECFLAVOUR)
args->auth_flavors[0] = data->pseudoflavor;
if (!args->nfs_server.hostname)
goto out_nomem;
/*
* The legacy version 6 binary mount data from userspace has a
* field used only to transport selinux information into the
* the kernel. To continue to support that functionality we
* have a touch of selinux knowledge here in the NFS code. The
* userspace code converted context=blah to just blah so we are
* converting back to the full string selinux understands.
*/
if (data->context[0]){
#ifdef CONFIG_SECURITY_SELINUX
int rc;
char *opts_str = kmalloc(sizeof(data->context) + 8, GFP_KERNEL);
if (!opts_str)
return -ENOMEM;
strcpy(opts_str, "context=");
data->context[NFS_MAX_CONTEXT_LEN] = '\0';
strcat(opts_str, &data->context[0]);
rc = security_sb_parse_opts_str(opts_str, &args->lsm_opts);
kfree(opts_str);
if (rc)
return rc;
#else
return -EINVAL;
#endif
}
break;
default: {
int status;
if (nfs_parse_mount_options((char *)options, args) == 0)
return -EINVAL;
if (!nfs_verify_server_address((struct sockaddr *)
&args->nfs_server.address))
goto out_no_address;
nfs_set_port((struct sockaddr *)&args->nfs_server.address,
args->nfs_server.port);
nfs_set_mount_transport_protocol(args);
status = nfs_parse_devname(dev_name,
&args->nfs_server.hostname,
PAGE_SIZE,
&args->nfs_server.export_path,
NFS_MAXPATHLEN);
if (!status)
status = nfs_try_mount(args, mntfh);
kfree(args->nfs_server.export_path);
args->nfs_server.export_path = NULL;
if (status)
return status;
break;
}
}
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
#ifndef CONFIG_NFS_V3
if (args->flags & NFS_MOUNT_VER3)
goto out_v3_not_compiled;
#endif /* !CONFIG_NFS_V3 */
return 0;
out_no_data:
dfprintk(MOUNT, "NFS: mount program didn't pass any mount data\n");
return -EINVAL;
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
out_no_v3:
dfprintk(MOUNT, "NFS: nfs_mount_data version %d does not support v3\n",
data->version);
return -EINVAL;
out_no_sec:
dfprintk(MOUNT, "NFS: nfs_mount_data version supports only AUTH_SYS\n");
return -EINVAL;
#ifndef CONFIG_NFS_V3
out_v3_not_compiled:
dfprintk(MOUNT, "NFS: NFSv3 is not compiled into kernel\n");
return -EPROTONOSUPPORT;
#endif /* !CONFIG_NFS_V3 */
out_nomem:
dfprintk(MOUNT, "NFS: not enough memory to handle mount options\n");
return -ENOMEM;
out_no_address:
dfprintk(MOUNT, "NFS: mount program didn't pass remote address\n");
return -EINVAL;
out_invalid_fh:
dfprintk(MOUNT, "NFS: invalid root filehandle\n");
return -EINVAL;
}
static int
nfs_compare_remount_data(struct nfs_server *nfss,
struct nfs_parsed_mount_data *data)
{
if (data->flags != nfss->flags ||
data->rsize != nfss->rsize ||
data->wsize != nfss->wsize ||
data->retrans != nfss->client->cl_timeout->to_retries ||
data->auth_flavors[0] != nfss->client->cl_auth->au_flavor ||
data->acregmin != nfss->acregmin / HZ ||
data->acregmax != nfss->acregmax / HZ ||
data->acdirmin != nfss->acdirmin / HZ ||
data->acdirmax != nfss->acdirmax / HZ ||
data->timeo != (10U * nfss->client->cl_timeout->to_initval / HZ) ||
data->nfs_server.addrlen != nfss->nfs_client->cl_addrlen ||
memcmp(&data->nfs_server.address, &nfss->nfs_client->cl_addr,
data->nfs_server.addrlen) != 0)
return -EINVAL;
return 0;
}
static int
nfs_remount(struct super_block *sb, int *flags, char *raw_data)
{
int error;
struct nfs_server *nfss = sb->s_fs_info;
struct nfs_parsed_mount_data *data;
struct nfs_mount_data *options = (struct nfs_mount_data *)raw_data;
struct nfs4_mount_data *options4 = (struct nfs4_mount_data *)raw_data;
u32 nfsvers = nfss->nfs_client->rpc_ops->version;
/*
* Userspace mount programs that send binary options generally send
* them populated with default values. We have no way to know which
* ones were explicitly specified. Fall back to legacy behavior and
* just return success.
*/
nfs_remount oops when rebooting + possible fix Jeff, Trond, The commit 48b605f83c920d8daa50e43fc2c7f718e04c7bfa (NFS: implement option checking when remounting NFS filesystems (resend)) generate an Oops on my platform when rebooting while its root FS on an NFS share (NFSv3, TCP) : Unmounting local filesystems...done. Unable to handle kernel NULL pointer dereference at virtual address 00000000 pgd = c3d00000 [00000000] *pgd=a3d72031, *pte=00000000, *ppte=00000000 Internal error: Oops: 17 [#1] Modules linked in: cpufreq_powersave cpufreq_ondemand cpufreq_userspace cpufreq_conservative ext3 jbd sd_mod pata_pcmcia libata scsi_mod pcmcia loop firmware_class pxafb cfbcopyarea cfbimgblt cfbfillrect pxa2xx_cs pxa2xx_core pcmcia_core snd_pxa2xx_ac97 snd_ac97_codec ac97_bus snd_pxa2xx_pcm snd_pcm_oss snd_mixer_oss snd_pcm snd_timer snd isp116x_hcd soundcore rtc_sa1100 snd_page_alloc pxa25x_udc usbcore rtc_ds1307 rtc_core CPU: 0 Not tainted (2.6.26-03414-g33af79d-dirty #15) PC is at nfs_remount+0x40/0x264 LR is at do_remount_sb+0x158/0x194 pc : [<c00bbf54>] lr : [<c0076c40>] psr: 60000013 sp : c2dd1e70 ip : c2dd1e98 fp : c2dd1e94 r10: 00000040 r9 : c3d17000 r8 : c3c3fc40 r7 : 00000000 r6 : 00000000 r5 : c3d2b200 r4 : 00000000 r3 : 00000003 r2 : 00000000 r1 : c2dd1e9c r0 : c3c3fc00 Flags: nZCv IRQs on FIQs on Mode SVC_32 ISA ARM Segment user Control: 0000397f Table: a3d00000 DAC: 00000015 Process mount (pid: 1462, stack limit = 0xc2dd0270) Stack: (0xc2dd1e70 to 0xc2dd2000) 1e60: 00000000 c3c3fc00 00000000 00000000 1e80: c3c3fc40 c3d17000 c2dd1ebc c2dd1e98 c0076c40 c00bbf20 c01c61e4 00000001 1ea0: c2dd1ebc 00000001 c3c3fc00 c2dd1ef0 c2dd1ee4 c2dd1ec0 c008c6d8 c0076af4 1ec0: 00000021 00000040 c2dd1ef0 c3d77000 c3eaa000 00000000 c2dd1f6c c2dd1ee8 1ee0: c008d1bc c008c5f8 00000000 c2dd0000 c3c0c320 c3805b38 c002064c 0001f820 1f00: 0001f810 00000001 00000001 00000000 c2dd0000 00000000 c2dd1f34 c2dd1f28 1f20: c005ead8 c005e6f8 c2dd1f44 c2dd1f38 c005eaf8 c005ead0 c2dd1f6c c2dd1f48 1f40: c008ae3c 00000000 c3d77000 0001f810 c0ed0021 c0020ca8 c2dd0000 00000000 1f60: c2dd1fa4 c2dd1f70 c008d2d4 c008d0bc 00000000 0001f810 c2dd1f9c c3eaa000 1f80: c3d17000 00000000 00000000 be8b6aa8 be8b6ad0 00000015 00000000 c2dd1fa8 1fa0: c0020b00 c008d254 00000000 be8b6aa8 0001f810 0001f820 0001f830 c0ed0021 1fc0: 00000000 be8b6aa8 be8b6ad0 00000015 00000000 be8b6ad0 0001f810 be8b6aa8 1fe0: 0001f810 be8b6964 0000aab8 40125124 60000010 0001f810 00000000 00000000 Backtrace: [<c00bbf14>] (nfs_remount+0x0/0x264) from [<c0076c40>] (do_remount_sb+0x158/0x194) r9:c3d17000 r8:c3c3fc40 r7:00000000 r6:00000000 r5:c3c3fc00 r4:00000000 [<c0076ae8>] (do_remount_sb+0x0/0x194) from [<c008c6d8>] (do_remount+0xec/0x118) r6:c2dd1ef0 r5:c3c3fc00 r4:00000001 [<c008c5ec>] (do_remount+0x0/0x118) from [<c008d1bc>] (do_mount+0x10c/0x198) [<c008d0b0>] (do_mount+0x0/0x198) from [<c008d2d4>] (sys_mount+0x8c/0xd4) [<c008d248>] (sys_mount+0x0/0xd4) from [<c0020b00>] (ret_fast_syscall+0x0/0x2c) r7:00000015 r6:be8b6ad0 r5:be8b6aa8 r4:00000000 Code: 0a000086 ea000006 e3530003 8a000004 (e5923000) ---[ end trace 55e1b689cf8c8a6a ]--- ------------[ cut here ]------------ WARNING: at kernel/exit.c:966 do_exit+0x3c/0x628() Modules linked in: cpufreq_powersave cpufreq_ondemand cpufreq_userspace cpufreq_conservative ext3 jbd sd_mod pata_pcmcia libata scsi_mod pcmcia loop firmware_class pxafb cfbcopyarea cfbimgblt cfbfillrect pxa2xx_cs pxa2xx_core pcmcia_core snd_pxa2xx_ac97 snd_ac97_codec ac97_bus snd_pxa2xx_pcm snd_pcm_oss snd_mixer_oss snd_pcm snd_timer snd isp116x_hcd soundcore rtc_sa1100 snd_page_alloc pxa25x_udc usbcore rtc_ds1307 rtc_core [<c0025168>] (dump_stack+0x0/0x14) from [<c0032154>] (warn_on_slowpath+0x4c/0x68) [<c0032108>] (warn_on_slowpath+0x0/0x68) from [<c003531c>] (do_exit+0x3c/0x628) r6:0000000b r5:c3c3dc80 r4:c2dd0000 [<c00352e0>] (do_exit+0x0/0x628) from [<c0025004>] (die+0x2b0/0x30c) [<c0024d54>] (die+0x0/0x30c) from [<c00270bc>] (__do_kernel_fault+0x6c/0x80) [<c0027050>] (__do_kernel_fault+0x0/0x80) from [<c00272e0>] (do_page_fault+0x210/0x230) r7:c3fa7118 r6:c3c3dc80 r5:c3d166a8 r4:00010000 [<c00270d0>] (do_page_fault+0x0/0x230) from [<c00201ec>] (do_DataAbort+0x3c/0xa0) [<c00201b0>] (do_DataAbort+0x0/0xa0) from [<c002064c>] (__dabt_svc+0x4c/0x60) Exception stack(0xc2dd1e28 to 0xc2dd1e70) 1e20: c3c3fc00 c2dd1e9c 00000000 00000003 00000000 c3d2b200 1e40: 00000000 00000000 c3c3fc40 c3d17000 00000040 c2dd1e94 c2dd1e98 c2dd1e70 1e60: c0076c40 c00bbf54 60000013 ffffffff r8:c3c3fc40 r7:00000000 r6:00000000 r5:c2dd1e5c r4:ffffffff [<c00bbf14>] (nfs_remount+0x0/0x264) from [<c0076c40>] (do_remount_sb+0x158/0x194) r9:c3d17000 r8:c3c3fc40 r7:00000000 r6:00000000 r5:c3c3fc00 r4:00000000 [<c0076ae8>] (do_remount_sb+0x0/0x194) from [<c008c6d8>] (do_remount+0xec/0x118) r6:c2dd1ef0 r5:c3c3fc00 r4:00000001 [<c008c5ec>] (do_remount+0x0/0x118) from [<c008d1bc>] (do_mount+0x10c/0x198) [<c008d0b0>] (do_mount+0x0/0x198) from [<c008d2d4>] (sys_mount+0x8c/0xd4) [<c008d248>] (sys_mount+0x0/0xd4) from [<c0020b00>] (ret_fast_syscall+0x0/0x2c) r7:00000015 r6:be8b6ad0 r5:be8b6aa8 r4:00000000 ---[ end trace 55e1b689cf8c8a6a ]--- /etc/rc6.d/S60umountroot: line 17: 1462 Segmentation fault mount $MOUNT_FORCE_OPT -n -o remount,ro -t dummytype dummydev / 2> /dev/null The new super.c:nfs_remount function doesn't check the validity of the options/options4 pointers. Unfortunately, this seems to happend. The obvious patch seems to check the pointers, and not to do anything if the happend to be NULL. Tested on an XScale PXA255 system, latest git. Regards, M. Signed-off-by: Marc Zyngier <marc.zyngier@altran.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2008-07-17 11:21:55 +00:00
if ((nfsvers == 4 && (!options4 || options4->version == 1)) ||
(nfsvers <= 3 && (!options || (options->version >= 1 &&
options->version <= 6))))
return 0;
data = kzalloc(sizeof(*data), GFP_KERNEL);
if (data == NULL)
return -ENOMEM;
lock_kernel();
/* fill out struct with values from existing mount */
data->flags = nfss->flags;
data->rsize = nfss->rsize;
data->wsize = nfss->wsize;
data->retrans = nfss->client->cl_timeout->to_retries;
data->auth_flavors[0] = nfss->client->cl_auth->au_flavor;
data->acregmin = nfss->acregmin / HZ;
data->acregmax = nfss->acregmax / HZ;
data->acdirmin = nfss->acdirmin / HZ;
data->acdirmax = nfss->acdirmax / HZ;
data->timeo = 10U * nfss->client->cl_timeout->to_initval / HZ;
data->nfs_server.addrlen = nfss->nfs_client->cl_addrlen;
memcpy(&data->nfs_server.address, &nfss->nfs_client->cl_addr,
data->nfs_server.addrlen);
/* overwrite those values with any that were specified */
error = nfs_parse_mount_options((char *)options, data);
if (error < 0)
goto out;
/* compare new mount options with old ones */
error = nfs_compare_remount_data(nfss, data);
out:
kfree(data);
unlock_kernel();
return error;
}
/*
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
* Initialise the common bits of the superblock
*/
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
static inline void nfs_initialise_sb(struct super_block *sb)
{
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
struct nfs_server *server = NFS_SB(sb);
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
sb->s_magic = NFS_SUPER_MAGIC;
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
/* We probably want something more informative here */
snprintf(sb->s_id, sizeof(sb->s_id),
"%x:%x", MAJOR(sb->s_dev), MINOR(sb->s_dev));
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
if (sb->s_blocksize == 0)
sb->s_blocksize = nfs_block_bits(server->wsize,
&sb->s_blocksize_bits);
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
if (server->flags & NFS_MOUNT_NOAC)
sb->s_flags |= MS_SYNCHRONOUS;
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
nfs_super_set_maxbytes(sb, server->maxfilesize);
}
/*
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
* Finish setting up an NFS2/3 superblock
*/
static void nfs_fill_super(struct super_block *sb,
struct nfs_parsed_mount_data *data)
{
struct nfs_server *server = NFS_SB(sb);
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
sb->s_blocksize_bits = 0;
sb->s_blocksize = 0;
if (data->bsize)
sb->s_blocksize = nfs_block_size(data->bsize, &sb->s_blocksize_bits);
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
if (server->flags & NFS_MOUNT_VER3) {
/* The VFS shouldn't apply the umask to mode bits. We will do
* so ourselves when necessary.
*/
sb->s_flags |= MS_POSIXACL;
sb->s_time_gran = 1;
}
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
sb->s_op = &nfs_sops;
nfs_initialise_sb(sb);
}
/*
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
* Finish setting up a cloned NFS2/3 superblock
*/
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
static void nfs_clone_super(struct super_block *sb,
const struct super_block *old_sb)
{
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
struct nfs_server *server = NFS_SB(sb);
sb->s_blocksize_bits = old_sb->s_blocksize_bits;
sb->s_blocksize = old_sb->s_blocksize;
sb->s_maxbytes = old_sb->s_maxbytes;
if (server->flags & NFS_MOUNT_VER3) {
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
/* The VFS shouldn't apply the umask to mode bits. We will do
* so ourselves when necessary.
*/
sb->s_flags |= MS_POSIXACL;
sb->s_time_gran = 1;
}
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
sb->s_op = old_sb->s_op;
nfs_initialise_sb(sb);
}
static int nfs_compare_mount_options(const struct super_block *s, const struct nfs_server *b, int flags)
{
const struct nfs_server *a = s->s_fs_info;
const struct rpc_clnt *clnt_a = a->client;
const struct rpc_clnt *clnt_b = b->client;
if ((s->s_flags & NFS_MS_MASK) != (flags & NFS_MS_MASK))
goto Ebusy;
if (a->nfs_client != b->nfs_client)
goto Ebusy;
if (a->flags != b->flags)
goto Ebusy;
if (a->wsize != b->wsize)
goto Ebusy;
if (a->rsize != b->rsize)
goto Ebusy;
if (a->acregmin != b->acregmin)
goto Ebusy;
if (a->acregmax != b->acregmax)
goto Ebusy;
if (a->acdirmin != b->acdirmin)
goto Ebusy;
if (a->acdirmax != b->acdirmax)
goto Ebusy;
if (clnt_a->cl_auth->au_flavor != clnt_b->cl_auth->au_flavor)
goto Ebusy;
return 1;
Ebusy:
return 0;
}
struct nfs_sb_mountdata {
struct nfs_server *server;
int mntflags;
};
static int nfs_set_super(struct super_block *s, void *data)
{
struct nfs_sb_mountdata *sb_mntdata = data;
struct nfs_server *server = sb_mntdata->server;
int ret;
s->s_flags = sb_mntdata->mntflags;
s->s_fs_info = server;
ret = set_anon_super(s, server);
if (ret == 0)
server->s_dev = s->s_dev;
return ret;
}
static int nfs_compare_super_address(struct nfs_server *server1,
struct nfs_server *server2)
{
struct sockaddr *sap1, *sap2;
sap1 = (struct sockaddr *)&server1->nfs_client->cl_addr;
sap2 = (struct sockaddr *)&server2->nfs_client->cl_addr;
if (sap1->sa_family != sap2->sa_family)
return 0;
switch (sap1->sa_family) {
case AF_INET: {
struct sockaddr_in *sin1 = (struct sockaddr_in *)sap1;
struct sockaddr_in *sin2 = (struct sockaddr_in *)sap2;
if (sin1->sin_addr.s_addr != sin2->sin_addr.s_addr)
return 0;
if (sin1->sin_port != sin2->sin_port)
return 0;
break;
}
case AF_INET6: {
struct sockaddr_in6 *sin1 = (struct sockaddr_in6 *)sap1;
struct sockaddr_in6 *sin2 = (struct sockaddr_in6 *)sap2;
if (!ipv6_addr_equal(&sin1->sin6_addr, &sin2->sin6_addr))
return 0;
if (sin1->sin6_port != sin2->sin6_port)
return 0;
break;
}
default:
return 0;
}
return 1;
}
static int nfs_compare_super(struct super_block *sb, void *data)
{
struct nfs_sb_mountdata *sb_mntdata = data;
struct nfs_server *server = sb_mntdata->server, *old = NFS_SB(sb);
int mntflags = sb_mntdata->mntflags;
if (!nfs_compare_super_address(old, server))
return 0;
/* Note: NFS_MOUNT_UNSHARED == NFS4_MOUNT_UNSHARED */
if (old->flags & NFS_MOUNT_UNSHARED)
return 0;
if (memcmp(&old->fsid, &server->fsid, sizeof(old->fsid)) != 0)
return 0;
return nfs_compare_mount_options(sb, server, mntflags);
}
static int nfs_bdi_register(struct nfs_server *server)
{
return bdi_register_dev(&server->backing_dev_info, server->s_dev);
}
static int nfs_get_sb(struct file_system_type *fs_type,
int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt)
{
struct nfs_server *server = NULL;
struct super_block *s;
struct nfs_parsed_mount_data *data;
struct nfs_fh *mntfh;
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
struct dentry *mntroot;
int (*compare_super)(struct super_block *, void *) = nfs_compare_super;
struct nfs_sb_mountdata sb_mntdata = {
.mntflags = flags,
};
int error = -ENOMEM;
data = kzalloc(sizeof(*data), GFP_KERNEL);
mntfh = kzalloc(sizeof(*mntfh), GFP_KERNEL);
if (data == NULL || mntfh == NULL)
goto out_free_fh;
security_init_mnt_opts(&data->lsm_opts);
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
/* Validate the mount data */
error = nfs_validate_mount_data(raw_data, data, mntfh, dev_name);
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
if (error < 0)
goto out;
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
/* Get a volume representation */
server = nfs_create_server(data, mntfh);
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
if (IS_ERR(server)) {
error = PTR_ERR(server);
goto out;
}
sb_mntdata.server = server;
if (server->flags & NFS_MOUNT_UNSHARED)
compare_super = NULL;
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
/* Get a superblock - note that we may end up sharing one that already exists */
s = sget(fs_type, compare_super, nfs_set_super, &sb_mntdata);
if (IS_ERR(s)) {
error = PTR_ERR(s);
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
goto out_err_nosb;
}
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
if (s->s_fs_info != server) {
nfs_free_server(server);
server = NULL;
} else {
error = nfs_bdi_register(server);
if (error)
goto error_splat_super;
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
}
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
if (!s->s_root) {
/* initial superblock/root creation */
nfs_fill_super(s, data);
NFS: Define and create superblock-level objects Define and create superblock-level cache index objects (as managed by nfs_server structs). Each superblock object is created in a server level index object and is itself an index into which inode-level objects are inserted. Ideally there would be one superblock-level object per server, and the former would be folded into the latter; however, since the "nosharecache" option exists this isn't possible. The superblock object key is a sequence consisting of: (1) Certain superblock s_flags. (2) Various connection parameters that serve to distinguish superblocks for sget(). (3) The volume FSID. (4) The security flavour. (5) The uniquifier length. (6) The uniquifier text. This is normally an empty string, unless the fsc=xyz mount option was used to explicitly specify a uniquifier. The key blob is of variable length, depending on the length of (6). The superblock object is given no coherency data to carry in the auxiliary data permitted by the cache. It is assumed that the superblock is always coherent. This patch also adds uniquification handling such that two otherwise identical superblocks, at least one of which is marked "nosharecache", won't end up trying to share the on-disk cache. It will be possible to manually provide a uniquifier through a mount option with a later patch to avoid the error otherwise produced. Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Steve Dickson <steved@redhat.com> Acked-by: Trond Myklebust <Trond.Myklebust@netapp.com> Acked-by: Al Viro <viro@zeniv.linux.org.uk> Tested-by: Daire Byrne <Daire.Byrne@framestore.com>
2009-04-03 15:42:42 +00:00
nfs_fscache_get_super_cookie(s, data);
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
}
mntroot = nfs_get_root(s, mntfh);
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
if (IS_ERR(mntroot)) {
error = PTR_ERR(mntroot);
goto error_splat_super;
}
error = security_sb_set_mnt_opts(s, &data->lsm_opts);
if (error)
goto error_splat_root;
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
s->s_flags |= MS_ACTIVE;
mnt->mnt_sb = s;
mnt->mnt_root = mntroot;
error = 0;
out:
kfree(data->nfs_server.hostname);
kfree(data->mount_server.hostname);
NFS: Define and create superblock-level objects Define and create superblock-level cache index objects (as managed by nfs_server structs). Each superblock object is created in a server level index object and is itself an index into which inode-level objects are inserted. Ideally there would be one superblock-level object per server, and the former would be folded into the latter; however, since the "nosharecache" option exists this isn't possible. The superblock object key is a sequence consisting of: (1) Certain superblock s_flags. (2) Various connection parameters that serve to distinguish superblocks for sget(). (3) The volume FSID. (4) The security flavour. (5) The uniquifier length. (6) The uniquifier text. This is normally an empty string, unless the fsc=xyz mount option was used to explicitly specify a uniquifier. The key blob is of variable length, depending on the length of (6). The superblock object is given no coherency data to carry in the auxiliary data permitted by the cache. It is assumed that the superblock is always coherent. This patch also adds uniquification handling such that two otherwise identical superblocks, at least one of which is marked "nosharecache", won't end up trying to share the on-disk cache. It will be possible to manually provide a uniquifier through a mount option with a later patch to avoid the error otherwise produced. Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Steve Dickson <steved@redhat.com> Acked-by: Trond Myklebust <Trond.Myklebust@netapp.com> Acked-by: Al Viro <viro@zeniv.linux.org.uk> Tested-by: Daire Byrne <Daire.Byrne@framestore.com>
2009-04-03 15:42:42 +00:00
kfree(data->fscache_uniq);
security_free_mnt_opts(&data->lsm_opts);
out_free_fh:
kfree(mntfh);
kfree(data);
return error;
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
out_err_nosb:
nfs_free_server(server);
goto out;
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
error_splat_root:
dput(mntroot);
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
error_splat_super:
deactivate_locked_super(s);
goto out;
}
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
/*
* Destroy an NFS2/3 superblock
*/
static void nfs_kill_super(struct super_block *s)
{
struct nfs_server *server = NFS_SB(s);
bdi_unregister(&server->backing_dev_info);
kill_anon_super(s);
NFS: Define and create superblock-level objects Define and create superblock-level cache index objects (as managed by nfs_server structs). Each superblock object is created in a server level index object and is itself an index into which inode-level objects are inserted. Ideally there would be one superblock-level object per server, and the former would be folded into the latter; however, since the "nosharecache" option exists this isn't possible. The superblock object key is a sequence consisting of: (1) Certain superblock s_flags. (2) Various connection parameters that serve to distinguish superblocks for sget(). (3) The volume FSID. (4) The security flavour. (5) The uniquifier length. (6) The uniquifier text. This is normally an empty string, unless the fsc=xyz mount option was used to explicitly specify a uniquifier. The key blob is of variable length, depending on the length of (6). The superblock object is given no coherency data to carry in the auxiliary data permitted by the cache. It is assumed that the superblock is always coherent. This patch also adds uniquification handling such that two otherwise identical superblocks, at least one of which is marked "nosharecache", won't end up trying to share the on-disk cache. It will be possible to manually provide a uniquifier through a mount option with a later patch to avoid the error otherwise produced. Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Steve Dickson <steved@redhat.com> Acked-by: Trond Myklebust <Trond.Myklebust@netapp.com> Acked-by: Al Viro <viro@zeniv.linux.org.uk> Tested-by: Daire Byrne <Daire.Byrne@framestore.com>
2009-04-03 15:42:42 +00:00
nfs_fscache_release_super_cookie(s);
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
nfs_free_server(server);
}
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
/*
* Clone an NFS2/3 server record on xdev traversal (FSID-change)
*/
static int nfs_xdev_get_sb(struct file_system_type *fs_type, int flags,
const char *dev_name, void *raw_data,
struct vfsmount *mnt)
{
struct nfs_clone_mount *data = raw_data;
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
struct super_block *s;
struct nfs_server *server;
struct dentry *mntroot;
int (*compare_super)(struct super_block *, void *) = nfs_compare_super;
struct nfs_sb_mountdata sb_mntdata = {
.mntflags = flags,
};
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
int error;
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
dprintk("--> nfs_xdev_get_sb()\n");
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
/* create a new volume representation */
server = nfs_clone_server(NFS_SB(data->sb), data->fh, data->fattr);
if (IS_ERR(server)) {
error = PTR_ERR(server);
goto out_err_noserver;
}
sb_mntdata.server = server;
if (server->flags & NFS_MOUNT_UNSHARED)
compare_super = NULL;
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
/* Get a superblock - note that we may end up sharing one that already exists */
s = sget(&nfs_fs_type, compare_super, nfs_set_super, &sb_mntdata);
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
if (IS_ERR(s)) {
error = PTR_ERR(s);
goto out_err_nosb;
}
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
if (s->s_fs_info != server) {
nfs_free_server(server);
server = NULL;
} else {
error = nfs_bdi_register(server);
if (error)
goto error_splat_super;
}
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
if (!s->s_root) {
/* initial superblock/root creation */
nfs_clone_super(s, data->sb);
}
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
mntroot = nfs_get_root(s, data->fh);
if (IS_ERR(mntroot)) {
error = PTR_ERR(mntroot);
goto error_splat_super;
}
if (mntroot->d_inode->i_op != NFS_SB(s)->nfs_client->rpc_ops->dir_inode_ops) {
kernel BUG at fs/nfs/namespace.c:108! - can be triggered by bad server Hi Trond, I have discovered that the BUG_ON in nfs_follow_mountpoint: BUG_ON(IS_ROOT(dentry)); can be triggered by a misbehaving server. What happens is the client does a lookup and discoveres that the named directory has a different fsid, so it initiates a mount. It then performs a GETATTR on the mounted directory and gets a different fsid again (due to a bug in the NFS server). This causes nfs_follow_mountpoint to be called on the newly mounted root, which triggers the BUG_ON. To duplicate this, have a directory which contains some mountpoints, and export that directory with the "crossmnt" flag using nfs-utils 1.1.1 (or 1.1.0 I think) The GETATTR on the root of the mounted filesystem will return the information for the top exportpoint, while a lookup will return the correct information. This difference causes the NFS client to BUG. I think the best way to fix this is to trap this possibility early, so just before completing the mount in the NFS client, check that it isn't going to use nfs_mountpoint_inode_operations. As long as i_op will never change once set (is that true?), this should be adequately safe. The following patch shows a possible approach, and it works for me. i.e. when the NFS server is misbehaving, I get ESTALE on those mountpoints, while when the NFS server is working correctly, I get correct behaviour on the client. NeilBrown Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2007-11-01 05:50:20 +00:00
dput(mntroot);
error = -ESTALE;
goto error_splat_super;
}
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
s->s_flags |= MS_ACTIVE;
mnt->mnt_sb = s;
mnt->mnt_root = mntroot;
/* clone any lsm security options from the parent to the new sb */
security_sb_clone_mnt_opts(data->sb, s);
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
dprintk("<-- nfs_xdev_get_sb() = 0\n");
return 0;
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
out_err_nosb:
nfs_free_server(server);
out_err_noserver:
dprintk("<-- nfs_xdev_get_sb() = %d [error]\n", error);
return error;
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
error_splat_super:
deactivate_locked_super(s);
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
dprintk("<-- nfs_xdev_get_sb() = %d [splat]\n", error);
return error;
}
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
#ifdef CONFIG_NFS_V4
/*
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
* Finish setting up a cloned NFS4 superblock
*/
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
static void nfs4_clone_super(struct super_block *sb,
const struct super_block *old_sb)
{
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
sb->s_blocksize_bits = old_sb->s_blocksize_bits;
sb->s_blocksize = old_sb->s_blocksize;
sb->s_maxbytes = old_sb->s_maxbytes;
sb->s_time_gran = 1;
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
sb->s_op = old_sb->s_op;
nfs_initialise_sb(sb);
}
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
/*
* Set up an NFS4 superblock
*/
static void nfs4_fill_super(struct super_block *sb)
{
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
sb->s_time_gran = 1;
sb->s_op = &nfs4_sops;
nfs_initialise_sb(sb);
}
static void nfs4_validate_mount_flags(struct nfs_parsed_mount_data *args)
{
args->flags &= ~(NFS_MOUNT_NONLM|NFS_MOUNT_NOACL|NFS_MOUNT_VER3);
}
/*
* Validate NFSv4 mount options
*/
static int nfs4_validate_mount_data(void *options,
struct nfs_parsed_mount_data *args,
const char *dev_name)
{
struct sockaddr_in *ap;
struct nfs4_mount_data *data = (struct nfs4_mount_data *)options;
char *c;
if (data == NULL)
goto out_no_data;
args->rsize = NFS_MAX_FILE_IO_SIZE;
args->wsize = NFS_MAX_FILE_IO_SIZE;
args->acregmin = NFS_DEF_ACREGMIN;
args->acregmax = NFS_DEF_ACREGMAX;
args->acdirmin = NFS_DEF_ACDIRMIN;
args->acdirmax = NFS_DEF_ACDIRMAX;
args->nfs_server.port = NFS_PORT; /* 2049 unless user set port= */
args->auth_flavors[0] = RPC_AUTH_UNIX;
args->auth_flavor_len = 0;
args->minorversion = 0;
switch (data->version) {
case 1:
ap = (struct sockaddr_in *)&args->nfs_server.address;
if (data->host_addrlen > sizeof(args->nfs_server.address))
goto out_no_address;
if (data->host_addrlen == 0)
goto out_no_address;
args->nfs_server.addrlen = data->host_addrlen;
if (copy_from_user(ap, data->host_addr, data->host_addrlen))
return -EFAULT;
if (!nfs_verify_server_address((struct sockaddr *)
&args->nfs_server.address))
goto out_no_address;
if (data->auth_flavourlen) {
if (data->auth_flavourlen > 1)
goto out_inval_auth;
if (copy_from_user(&args->auth_flavors[0],
data->auth_flavours,
sizeof(args->auth_flavors[0])))
return -EFAULT;
}
c = strndup_user(data->hostname.data, NFS4_MAXNAMLEN);
if (IS_ERR(c))
return PTR_ERR(c);
args->nfs_server.hostname = c;
c = strndup_user(data->mnt_path.data, NFS4_MAXPATHLEN);
if (IS_ERR(c))
return PTR_ERR(c);
args->nfs_server.export_path = c;
dfprintk(MOUNT, "NFS: MNTPATH: '%s'\n", c);
c = strndup_user(data->client_addr.data, 16);
if (IS_ERR(c))
return PTR_ERR(c);
args->client_address = c;
/*
* Translate to nfs_parsed_mount_data, which nfs4_fill_super
* can deal with.
*/
args->flags = data->flags & NFS4_MOUNT_FLAGMASK;
args->rsize = data->rsize;
args->wsize = data->wsize;
args->timeo = data->timeo;
args->retrans = data->retrans;
args->acregmin = data->acregmin;
args->acregmax = data->acregmax;
args->acdirmin = data->acdirmin;
args->acdirmax = data->acdirmax;
args->nfs_server.protocol = data->proto;
nfs_validate_transport_protocol(args);
break;
default: {
int status;
if (nfs_parse_mount_options((char *)options, args) == 0)
return -EINVAL;
if (!nfs_verify_server_address((struct sockaddr *)
&args->nfs_server.address))
return -EINVAL;
nfs_set_port((struct sockaddr *)&args->nfs_server.address,
args->nfs_server.port);
nfs_validate_transport_protocol(args);
nfs4_validate_mount_flags(args);
if (args->auth_flavor_len > 1)
goto out_inval_auth;
if (args->client_address == NULL)
goto out_no_client_address;
status = nfs_parse_devname(dev_name,
&args->nfs_server.hostname,
NFS4_MAXNAMLEN,
&args->nfs_server.export_path,
NFS4_MAXPATHLEN);
if (status < 0)
return status;
break;
}
}
return 0;
out_no_data:
dfprintk(MOUNT, "NFS4: mount program didn't pass any mount data\n");
return -EINVAL;
out_inval_auth:
dfprintk(MOUNT, "NFS4: Invalid number of RPC auth flavours %d\n",
data->auth_flavourlen);
return -EINVAL;
out_no_address:
dfprintk(MOUNT, "NFS4: mount program didn't pass remote address\n");
return -EINVAL;
out_no_client_address:
dfprintk(MOUNT, "NFS4: mount program didn't pass callback address\n");
return -EINVAL;
}
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
/*
* Get the superblock for the NFS4 root partition
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
*/
static int nfs4_remote_get_sb(struct file_system_type *fs_type,
int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt)
{
struct nfs_parsed_mount_data *data = raw_data;
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
struct super_block *s;
struct nfs_server *server;
struct nfs_fh *mntfh;
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
struct dentry *mntroot;
int (*compare_super)(struct super_block *, void *) = nfs_compare_super;
struct nfs_sb_mountdata sb_mntdata = {
.mntflags = flags,
};
int error = -ENOMEM;
mntfh = kzalloc(sizeof(*mntfh), GFP_KERNEL);
if (data == NULL || mntfh == NULL)
goto out_free_fh;
security_init_mnt_opts(&data->lsm_opts);
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
/* Get a volume representation */
server = nfs4_create_server(data, mntfh);
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
if (IS_ERR(server)) {
error = PTR_ERR(server);
goto out;
}
sb_mntdata.server = server;
if (server->flags & NFS4_MOUNT_UNSHARED)
compare_super = NULL;
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
/* Get a superblock - note that we may end up sharing one that already exists */
s = sget(&nfs4_fs_type, compare_super, nfs_set_super, &sb_mntdata);
if (IS_ERR(s)) {
error = PTR_ERR(s);
goto out_free;
}
if (s->s_fs_info != server) {
nfs_free_server(server);
server = NULL;
} else {
error = nfs_bdi_register(server);
if (error)
goto error_splat_super;
}
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
if (!s->s_root) {
/* initial superblock/root creation */
nfs4_fill_super(s);
NFS: Define and create superblock-level objects Define and create superblock-level cache index objects (as managed by nfs_server structs). Each superblock object is created in a server level index object and is itself an index into which inode-level objects are inserted. Ideally there would be one superblock-level object per server, and the former would be folded into the latter; however, since the "nosharecache" option exists this isn't possible. The superblock object key is a sequence consisting of: (1) Certain superblock s_flags. (2) Various connection parameters that serve to distinguish superblocks for sget(). (3) The volume FSID. (4) The security flavour. (5) The uniquifier length. (6) The uniquifier text. This is normally an empty string, unless the fsc=xyz mount option was used to explicitly specify a uniquifier. The key blob is of variable length, depending on the length of (6). The superblock object is given no coherency data to carry in the auxiliary data permitted by the cache. It is assumed that the superblock is always coherent. This patch also adds uniquification handling such that two otherwise identical superblocks, at least one of which is marked "nosharecache", won't end up trying to share the on-disk cache. It will be possible to manually provide a uniquifier through a mount option with a later patch to avoid the error otherwise produced. Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Steve Dickson <steved@redhat.com> Acked-by: Trond Myklebust <Trond.Myklebust@netapp.com> Acked-by: Al Viro <viro@zeniv.linux.org.uk> Tested-by: Daire Byrne <Daire.Byrne@framestore.com>
2009-04-03 15:42:42 +00:00
nfs_fscache_get_super_cookie(s, data);
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
}
mntroot = nfs4_get_root(s, mntfh);
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
if (IS_ERR(mntroot)) {
error = PTR_ERR(mntroot);
goto error_splat_super;
}
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
error = security_sb_set_mnt_opts(s, &data->lsm_opts);
if (error)
goto error_splat_root;
s->s_flags |= MS_ACTIVE;
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
mnt->mnt_sb = s;
mnt->mnt_root = mntroot;
error = 0;
out:
security_free_mnt_opts(&data->lsm_opts);
out_free_fh:
kfree(mntfh);
return error;
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
out_free:
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
nfs_free_server(server);
goto out;
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
error_splat_root:
dput(mntroot);
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
error_splat_super:
deactivate_locked_super(s);
goto out;
}
static struct vfsmount *nfs_do_root_mount(struct file_system_type *fs_type,
int flags, void *data, const char *hostname)
{
struct vfsmount *root_mnt;
char *root_devname;
size_t len;
len = strlen(hostname) + 3;
root_devname = kmalloc(len, GFP_KERNEL);
if (root_devname == NULL)
return ERR_PTR(-ENOMEM);
snprintf(root_devname, len, "%s:/", hostname);
root_mnt = vfs_kern_mount(fs_type, flags, root_devname, data);
kfree(root_devname);
return root_mnt;
}
static void nfs_fix_devname(const struct path *path, struct vfsmount *mnt)
{
char *page = (char *) __get_free_page(GFP_KERNEL);
char *devname, *tmp;
if (page == NULL)
return;
devname = nfs_path(path->mnt->mnt_devname,
path->mnt->mnt_root, path->dentry,
page, PAGE_SIZE);
if (devname == NULL)
goto out_freepage;
tmp = kstrdup(devname, GFP_KERNEL);
if (tmp == NULL)
goto out_freepage;
kfree(mnt->mnt_devname);
mnt->mnt_devname = tmp;
out_freepage:
free_page((unsigned long)page);
}
static int nfs_follow_remote_path(struct vfsmount *root_mnt,
const char *export_path, struct vfsmount *mnt_target)
{
struct mnt_namespace *ns_private;
struct nameidata nd;
struct super_block *s;
int ret;
ns_private = create_mnt_ns(root_mnt);
ret = PTR_ERR(ns_private);
if (IS_ERR(ns_private))
goto out_mntput;
ret = vfs_path_lookup(root_mnt->mnt_root, root_mnt,
export_path, LOOKUP_FOLLOW, &nd);
put_mnt_ns(ns_private);
if (ret != 0)
goto out_err;
s = nd.path.mnt->mnt_sb;
atomic_inc(&s->s_active);
mnt_target->mnt_sb = s;
mnt_target->mnt_root = dget(nd.path.dentry);
/* Correct the device pathname */
nfs_fix_devname(&nd.path, mnt_target);
path_put(&nd.path);
down_write(&s->s_umount);
return 0;
out_mntput:
mntput(root_mnt);
out_err:
return ret;
}
/*
* Get the superblock for an NFS4 mountpoint
*/
static int nfs4_get_sb(struct file_system_type *fs_type,
int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt)
{
struct nfs_parsed_mount_data *data;
char *export_path;
struct vfsmount *root_mnt;
int error = -ENOMEM;
data = kzalloc(sizeof(*data), GFP_KERNEL);
if (data == NULL)
goto out_free_data;
/* Validate the mount data */
error = nfs4_validate_mount_data(raw_data, data, dev_name);
if (error < 0)
goto out;
export_path = data->nfs_server.export_path;
data->nfs_server.export_path = "/";
root_mnt = nfs_do_root_mount(&nfs4_remote_fs_type, flags, data,
data->nfs_server.hostname);
data->nfs_server.export_path = export_path;
error = PTR_ERR(root_mnt);
if (IS_ERR(root_mnt))
goto out;
error = nfs_follow_remote_path(root_mnt, export_path, mnt);
out:
kfree(data->client_address);
kfree(data->nfs_server.export_path);
kfree(data->nfs_server.hostname);
kfree(data->fscache_uniq);
out_free_data:
kfree(data);
dprintk("<-- nfs4_get_sb() = %d%s\n", error,
error != 0 ? " [error]" : "");
return error;
}
static void nfs4_kill_super(struct super_block *sb)
{
struct nfs_server *server = NFS_SB(sb);
dprintk("--> %s\n", __func__);
nfs_super_return_all_delegations(sb);
kill_anon_super(sb);
nfs4_renewd_prepare_shutdown(server);
NFS: Define and create superblock-level objects Define and create superblock-level cache index objects (as managed by nfs_server structs). Each superblock object is created in a server level index object and is itself an index into which inode-level objects are inserted. Ideally there would be one superblock-level object per server, and the former would be folded into the latter; however, since the "nosharecache" option exists this isn't possible. The superblock object key is a sequence consisting of: (1) Certain superblock s_flags. (2) Various connection parameters that serve to distinguish superblocks for sget(). (3) The volume FSID. (4) The security flavour. (5) The uniquifier length. (6) The uniquifier text. This is normally an empty string, unless the fsc=xyz mount option was used to explicitly specify a uniquifier. The key blob is of variable length, depending on the length of (6). The superblock object is given no coherency data to carry in the auxiliary data permitted by the cache. It is assumed that the superblock is always coherent. This patch also adds uniquification handling such that two otherwise identical superblocks, at least one of which is marked "nosharecache", won't end up trying to share the on-disk cache. It will be possible to manually provide a uniquifier through a mount option with a later patch to avoid the error otherwise produced. Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Steve Dickson <steved@redhat.com> Acked-by: Trond Myklebust <Trond.Myklebust@netapp.com> Acked-by: Al Viro <viro@zeniv.linux.org.uk> Tested-by: Daire Byrne <Daire.Byrne@framestore.com>
2009-04-03 15:42:42 +00:00
nfs_fscache_release_super_cookie(sb);
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
nfs_free_server(server);
dprintk("<-- %s\n", __func__);
}
/*
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
* Clone an NFS4 server record on xdev traversal (FSID-change)
*/
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
static int nfs4_xdev_get_sb(struct file_system_type *fs_type, int flags,
const char *dev_name, void *raw_data,
struct vfsmount *mnt)
{
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
struct nfs_clone_mount *data = raw_data;
struct super_block *s;
struct nfs_server *server;
struct dentry *mntroot;
int (*compare_super)(struct super_block *, void *) = nfs_compare_super;
struct nfs_sb_mountdata sb_mntdata = {
.mntflags = flags,
};
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
int error;
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
dprintk("--> nfs4_xdev_get_sb()\n");
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
/* create a new volume representation */
server = nfs_clone_server(NFS_SB(data->sb), data->fh, data->fattr);
if (IS_ERR(server)) {
error = PTR_ERR(server);
goto out_err_noserver;
}
sb_mntdata.server = server;
if (server->flags & NFS4_MOUNT_UNSHARED)
compare_super = NULL;
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
/* Get a superblock - note that we may end up sharing one that already exists */
s = sget(&nfs4_fs_type, compare_super, nfs_set_super, &sb_mntdata);
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
if (IS_ERR(s)) {
error = PTR_ERR(s);
goto out_err_nosb;
}
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
if (s->s_fs_info != server) {
nfs_free_server(server);
server = NULL;
} else {
error = nfs_bdi_register(server);
if (error)
goto error_splat_super;
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
}
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
if (!s->s_root) {
/* initial superblock/root creation */
nfs4_clone_super(s, data->sb);
}
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
mntroot = nfs4_get_root(s, data->fh);
if (IS_ERR(mntroot)) {
error = PTR_ERR(mntroot);
goto error_splat_super;
}
if (mntroot->d_inode->i_op != NFS_SB(s)->nfs_client->rpc_ops->dir_inode_ops) {
dput(mntroot);
error = -ESTALE;
goto error_splat_super;
}
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
s->s_flags |= MS_ACTIVE;
mnt->mnt_sb = s;
mnt->mnt_root = mntroot;
security_sb_clone_mnt_opts(data->sb, s);
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
dprintk("<-- nfs4_xdev_get_sb() = 0\n");
return 0;
out_err_nosb:
nfs_free_server(server);
out_err_noserver:
dprintk("<-- nfs4_xdev_get_sb() = %d [error]\n", error);
return error;
error_splat_super:
deactivate_locked_super(s);
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
dprintk("<-- nfs4_xdev_get_sb() = %d [splat]\n", error);
return error;
}
static int nfs4_remote_referral_get_sb(struct file_system_type *fs_type,
int flags, const char *dev_name, void *raw_data,
struct vfsmount *mnt)
{
struct nfs_clone_mount *data = raw_data;
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
struct super_block *s;
struct nfs_server *server;
struct dentry *mntroot;
struct nfs_fh mntfh;
int (*compare_super)(struct super_block *, void *) = nfs_compare_super;
struct nfs_sb_mountdata sb_mntdata = {
.mntflags = flags,
};
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
int error;
dprintk("--> nfs4_referral_get_sb()\n");
/* create a new volume representation */
server = nfs4_create_referral_server(data, &mntfh);
if (IS_ERR(server)) {
error = PTR_ERR(server);
goto out_err_noserver;
}
sb_mntdata.server = server;
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
if (server->flags & NFS4_MOUNT_UNSHARED)
compare_super = NULL;
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
/* Get a superblock - note that we may end up sharing one that already exists */
s = sget(&nfs4_fs_type, compare_super, nfs_set_super, &sb_mntdata);
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
if (IS_ERR(s)) {
error = PTR_ERR(s);
goto out_err_nosb;
}
if (s->s_fs_info != server) {
nfs_free_server(server);
server = NULL;
} else {
error = nfs_bdi_register(server);
if (error)
goto error_splat_super;
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
}
if (!s->s_root) {
/* initial superblock/root creation */
nfs4_fill_super(s);
}
mntroot = nfs4_get_root(s, &mntfh);
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
if (IS_ERR(mntroot)) {
error = PTR_ERR(mntroot);
goto error_splat_super;
}
if (mntroot->d_inode->i_op != NFS_SB(s)->nfs_client->rpc_ops->dir_inode_ops) {
dput(mntroot);
error = -ESTALE;
goto error_splat_super;
}
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
s->s_flags |= MS_ACTIVE;
mnt->mnt_sb = s;
mnt->mnt_root = mntroot;
security_sb_clone_mnt_opts(data->sb, s);
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
dprintk("<-- nfs4_referral_get_sb() = 0\n");
return 0;
out_err_nosb:
nfs_free_server(server);
out_err_noserver:
dprintk("<-- nfs4_referral_get_sb() = %d [error]\n", error);
return error;
error_splat_super:
deactivate_locked_super(s);
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
dprintk("<-- nfs4_referral_get_sb() = %d [splat]\n", error);
return error;
}
/*
* Create an NFS4 server record on referral traversal
*/
static int nfs4_referral_get_sb(struct file_system_type *fs_type,
int flags, const char *dev_name, void *raw_data,
struct vfsmount *mnt)
{
struct nfs_clone_mount *data = raw_data;
char *export_path;
struct vfsmount *root_mnt;
int error;
dprintk("--> nfs4_referral_get_sb()\n");
export_path = data->mnt_path;
data->mnt_path = "/";
root_mnt = nfs_do_root_mount(&nfs4_remote_referral_fs_type,
flags, data, data->hostname);
data->mnt_path = export_path;
error = PTR_ERR(root_mnt);
if (IS_ERR(root_mnt))
goto out;
error = nfs_follow_remote_path(root_mnt, export_path, mnt);
out:
dprintk("<-- nfs4_referral_get_sb() = %d%s\n", error,
error != 0 ? " [error]" : "");
return error;
}
NFS: Share NFS superblocks per-protocol per-server per-FSID The attached patch makes NFS share superblocks between mounts from the same server and FSID over the same protocol. It does this by creating each superblock with a false root and returning the real root dentry in the vfsmount presented by get_sb(). The root dentry set starts off as an anonymous dentry if we don't already have the dentry for its inode, otherwise it simply returns the dentry we already have. We may thus end up with several trees of dentries in the superblock, and if at some later point one of anonymous tree roots is discovered by normal filesystem activity to be located in another tree within the superblock, the anonymous root is named and materialises attached to the second tree at the appropriate point. Why do it this way? Why not pass an extra argument to the mount() syscall to indicate the subpath and then pathwalk from the server root to the desired directory? You can't guarantee this will work for two reasons: (1) The root and intervening nodes may not be accessible to the client. With NFS2 and NFS3, for instance, mountd is called on the server to get the filehandle for the tip of a path. mountd won't give us handles for anything we don't have permission to access, and so we can't set up NFS inodes for such nodes, and so can't easily set up dentries (we'd have to have ghost inodes or something). With this patch we don't actually create dentries until we get handles from the server that we can use to set up their inodes, and we don't actually bind them into the tree until we know for sure where they go. (2) Inaccessible symbolic links. If we're asked to mount two exports from the server, eg: mount warthog:/warthog/aaa/xxx /mmm mount warthog:/warthog/bbb/yyy /nnn We may not be able to access anything nearer the root than xxx and yyy, but we may find out later that /mmm/www/yyy, say, is actually the same directory as the one mounted on /nnn. What we might then find out, for example, is that /warthog/bbb was actually a symbolic link to /warthog/aaa/xxx/www, but we can't actually determine that by talking to the server until /warthog is made available by NFS. This would lead to having constructed an errneous dentry tree which we can't easily fix. We can end up with a dentry marked as a directory when it should actually be a symlink, or we could end up with an apparently hardlinked directory. With this patch we need not make assumptions about the type of a dentry for which we can't retrieve information, nor need we assume we know its place in the grand scheme of things until we actually see that place. This patch reduces the possibility of aliasing in the inode and page caches for inodes that may be accessed by more than one NFS export. It also reduces the number of superblocks required for NFS where there are many NFS exports being used from a server (home directory server + autofs for example). This in turn makes it simpler to do local caching of network filesystems, as it can then be guaranteed that there won't be links from multiple inodes in separate superblocks to the same cache file. Obviously, cache aliasing between different levels of NFS protocol could still be a problem, but at least that gives us another key to use when indexing the cache. This patch makes the following changes: (1) The server record construction/destruction has been abstracted out into its own set of functions to make things easier to get right. These have been moved into fs/nfs/client.c. All the code in fs/nfs/client.c has to do with the management of connections to servers, and doesn't touch superblocks in any way; the remaining code in fs/nfs/super.c has to do with VFS superblock management. (2) The sequence of events undertaken by NFS mount is now reordered: (a) A volume representation (struct nfs_server) is allocated. (b) A server representation (struct nfs_client) is acquired. This may be allocated or shared, and is keyed on server address, port and NFS version. (c) If allocated, the client representation is initialised. The state member variable of nfs_client is used to prevent a race during initialisation from two mounts. (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find the root filehandle for the mount (fs/nfs/getroot.c). For NFS2/3 we are given the root FH in advance. (e) The volume FSID is probed for on the root FH. (f) The volume representation is initialised from the FSINFO record retrieved on the root FH. (g) sget() is called to acquire a superblock. This may be allocated or shared, keyed on client pointer and FSID. (h) If allocated, the superblock is initialised. (i) If the superblock is shared, then the new nfs_server record is discarded. (j) The root dentry for this mount is looked up from the root FH. (k) The root dentry for this mount is assigned to the vfsmount. (3) nfs_readdir_lookup() creates dentries for each of the entries readdir() returns; this function now attaches disconnected trees from alternate roots that happen to be discovered attached to a directory being read (in the same way nfs_lookup() is made to do for lookup ops). The new d_materialise_unique() function is now used to do this, thus permitting the whole thing to be done under one set of locks, and thus avoiding any race between mount and lookup operations on the same directory. (4) The client management code uses a new debug facility: NFSDBG_CLIENT which is set by echoing 1024 to /proc/net/sunrpc/nfs_debug. (5) Clone mounts are now called xdev mounts. (6) Use the dentry passed to the statfs() op as the handle for retrieving fs statistics rather than the root dentry of the superblock (which is now a dummy). Signed-Off-By: David Howells <dhowells@redhat.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
2006-08-23 00:06:13 +00:00
#endif /* CONFIG_NFS_V4 */