linux/ipc/shm.c

1096 lines
26 KiB
C
Raw Normal View History

/*
* linux/ipc/shm.c
* Copyright (C) 1992, 1993 Krishna Balasubramanian
* Many improvements/fixes by Bruno Haible.
* Replaced `struct shm_desc' by `struct vm_area_struct', July 1994.
* Fixed the shm swap deallocation (shm_unuse()), August 1998 Andrea Arcangeli.
*
* /proc/sysvipc/shm support (c) 1999 Dragos Acostachioaie <dragos@iname.com>
* BIGMEM support, Andrea Arcangeli <andrea@suse.de>
* SMP thread shm, Jean-Luc Boyard <jean-luc.boyard@siemens.fr>
* HIGHMEM support, Ingo Molnar <mingo@redhat.com>
* Make shmmax, shmall, shmmni sysctl'able, Christoph Rohland <cr@sap.com>
* Shared /dev/zero support, Kanoj Sarcar <kanoj@sgi.com>
* Move the mm functionality over to mm/shmem.c, Christoph Rohland <cr@sap.com>
*
[PATCH] Rework of IPC auditing 1) The audit_ipc_perms() function has been split into two different functions: - audit_ipc_obj() - audit_ipc_set_perm() There's a key shift here... The audit_ipc_obj() collects the uid, gid, mode, and SElinux context label of the current ipc object. This audit_ipc_obj() hook is now found in several places. Most notably, it is hooked in ipcperms(), which is called in various places around the ipc code permforming a MAC check. Additionally there are several places where *checkid() is used to validate that an operation is being performed on a valid object while not necessarily having a nearby ipcperms() call. In these locations, audit_ipc_obj() is called to ensure that the information is captured by the audit system. The audit_set_new_perm() function is called any time the permissions on the ipc object changes. In this case, the NEW permissions are recorded (and note that an audit_ipc_obj() call exists just a few lines before each instance). 2) Support for an AUDIT_IPC_SET_PERM audit message type. This allows for separate auxiliary audit records for normal operations on an IPC object and permissions changes. Note that the same struct audit_aux_data_ipcctl is used and populated, however there are separate audit_log_format statements based on the type of the message. Finally, the AUDIT_IPC block of code in audit_free_aux() was extended to handle aux messages of this new type. No more mem leaks I hope ;-) Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2006-04-02 21:07:33 +00:00
* support for audit of ipc object properties and permission changes
* Dustin Kirkland <dustin.kirkland@us.ibm.com>
*
* namespaces support
* OpenVZ, SWsoft Inc.
* Pavel Emelianov <xemul@openvz.org>
*/
#include <linux/slab.h>
#include <linux/mm.h>
#include <linux/hugetlb.h>
#include <linux/shm.h>
#include <linux/init.h>
#include <linux/file.h>
#include <linux/mman.h>
#include <linux/shmem_fs.h>
#include <linux/security.h>
#include <linux/syscalls.h>
#include <linux/audit.h>
#include <linux/capability.h>
#include <linux/ptrace.h>
#include <linux/seq_file.h>
#include <linux/rwsem.h>
#include <linux/nsproxy.h>
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
#include <linux/mount.h>
namespaces: move the IPC namespace under IPC_NS option Currently the IPC namespace management code is spread over the ipc/*.c files. I moved this code into ipc/namespace.c file which is compiled out when needed. The linux/ipc_namespace.h file is used to store the prototypes of the functions in namespace.c and the stubs for NAMESPACES=n case. This is done so, because the stub for copy_ipc_namespace requires the knowledge of the CLONE_NEWIPC flag, which is in sched.h. But the linux/ipc.h file itself in included into many many .c files via the sys.h->sem.h sequence so adding the sched.h into it will make all these .c depend on sched.h which is not that good. On the other hand the knowledge about the namespaces stuff is required in 4 .c files only. Besides, this patch compiles out some auxiliary functions from ipc/sem.c, msg.c and shm.c files. It turned out that moving these functions into namespaces.c is not that easy because they use many other calls and macros from the original file. Moving them would make this patch complicated. On the other hand all these functions can be consolidated, so I will send a separate patch doing this a bit later. Signed-off-by: Pavel Emelyanov <xemul@openvz.org> Acked-by: Serge Hallyn <serue@us.ibm.com> Cc: Cedric Le Goater <clg@fr.ibm.com> Cc: "Eric W. Biederman" <ebiederm@xmission.com> Cc: Herbert Poetzl <herbert@13thfloor.at> Cc: Kirill Korotaev <dev@sw.ru> Cc: Sukadev Bhattiprolu <sukadev@us.ibm.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-08 12:18:22 +00:00
#include <linux/ipc_namespace.h>
#include <linux/ima.h>
#include <asm/uaccess.h>
#include "util.h"
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
struct shm_file_data {
int id;
struct ipc_namespace *ns;
struct file *file;
const struct vm_operations_struct *vm_ops;
};
#define shm_file_data(file) (*((struct shm_file_data **)&(file)->private_data))
static const struct file_operations shm_file_operations;
static struct vm_operations_struct shm_vm_ops;
#define shm_ids(ns) ((ns)->ids[IPC_SHM_IDS])
#define shm_unlock(shp) \
ipc_unlock(&(shp)->shm_perm)
static int newseg(struct ipc_namespace *, struct ipc_params *);
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
static void shm_open(struct vm_area_struct *vma);
static void shm_close(struct vm_area_struct *vma);
static void shm_destroy (struct ipc_namespace *ns, struct shmid_kernel *shp);
#ifdef CONFIG_PROC_FS
static int sysvipc_shm_proc_show(struct seq_file *s, void *it);
#endif
void shm_init_ns(struct ipc_namespace *ns)
{
ns->shm_ctlmax = SHMMAX;
ns->shm_ctlall = SHMALL;
ns->shm_ctlmni = SHMMNI;
ns->shm_tot = 0;
ipc_init_ids(&shm_ids(ns));
}
/*
* Called with shm_ids.rw_mutex (writer) and the shp structure locked.
* Only shm_ids.rw_mutex remains locked on exit.
*/
static void do_shm_rmid(struct ipc_namespace *ns, struct kern_ipc_perm *ipcp)
{
struct shmid_kernel *shp;
shp = container_of(ipcp, struct shmid_kernel, shm_perm);
if (shp->shm_nattch){
shp->shm_perm.mode |= SHM_DEST;
/* Do not find it any more */
shp->shm_perm.key = IPC_PRIVATE;
shm_unlock(shp);
} else
shm_destroy(ns, shp);
}
namespaces: move the IPC namespace under IPC_NS option Currently the IPC namespace management code is spread over the ipc/*.c files. I moved this code into ipc/namespace.c file which is compiled out when needed. The linux/ipc_namespace.h file is used to store the prototypes of the functions in namespace.c and the stubs for NAMESPACES=n case. This is done so, because the stub for copy_ipc_namespace requires the knowledge of the CLONE_NEWIPC flag, which is in sched.h. But the linux/ipc.h file itself in included into many many .c files via the sys.h->sem.h sequence so adding the sched.h into it will make all these .c depend on sched.h which is not that good. On the other hand the knowledge about the namespaces stuff is required in 4 .c files only. Besides, this patch compiles out some auxiliary functions from ipc/sem.c, msg.c and shm.c files. It turned out that moving these functions into namespaces.c is not that easy because they use many other calls and macros from the original file. Moving them would make this patch complicated. On the other hand all these functions can be consolidated, so I will send a separate patch doing this a bit later. Signed-off-by: Pavel Emelyanov <xemul@openvz.org> Acked-by: Serge Hallyn <serue@us.ibm.com> Cc: Cedric Le Goater <clg@fr.ibm.com> Cc: "Eric W. Biederman" <ebiederm@xmission.com> Cc: Herbert Poetzl <herbert@13thfloor.at> Cc: Kirill Korotaev <dev@sw.ru> Cc: Sukadev Bhattiprolu <sukadev@us.ibm.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-08 12:18:22 +00:00
#ifdef CONFIG_IPC_NS
void shm_exit_ns(struct ipc_namespace *ns)
{
free_ipcs(ns, &shm_ids(ns), do_shm_rmid);
}
namespaces: move the IPC namespace under IPC_NS option Currently the IPC namespace management code is spread over the ipc/*.c files. I moved this code into ipc/namespace.c file which is compiled out when needed. The linux/ipc_namespace.h file is used to store the prototypes of the functions in namespace.c and the stubs for NAMESPACES=n case. This is done so, because the stub for copy_ipc_namespace requires the knowledge of the CLONE_NEWIPC flag, which is in sched.h. But the linux/ipc.h file itself in included into many many .c files via the sys.h->sem.h sequence so adding the sched.h into it will make all these .c depend on sched.h which is not that good. On the other hand the knowledge about the namespaces stuff is required in 4 .c files only. Besides, this patch compiles out some auxiliary functions from ipc/sem.c, msg.c and shm.c files. It turned out that moving these functions into namespaces.c is not that easy because they use many other calls and macros from the original file. Moving them would make this patch complicated. On the other hand all these functions can be consolidated, so I will send a separate patch doing this a bit later. Signed-off-by: Pavel Emelyanov <xemul@openvz.org> Acked-by: Serge Hallyn <serue@us.ibm.com> Cc: Cedric Le Goater <clg@fr.ibm.com> Cc: "Eric W. Biederman" <ebiederm@xmission.com> Cc: Herbert Poetzl <herbert@13thfloor.at> Cc: Kirill Korotaev <dev@sw.ru> Cc: Sukadev Bhattiprolu <sukadev@us.ibm.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-08 12:18:22 +00:00
#endif
void __init shm_init (void)
{
shm_init_ns(&init_ipc_ns);
ipc_init_proc_interface("sysvipc/shm",
" key shmid perms size cpid lpid nattch uid gid cuid cgid atime dtime ctime\n",
IPC_SHM_IDS, sysvipc_shm_proc_show);
}
/*
* shm_lock_(check_) routines are called in the paths where the rw_mutex
* is not necessarily held.
*/
static inline struct shmid_kernel *shm_lock(struct ipc_namespace *ns, int id)
{
struct kern_ipc_perm *ipcp = ipc_lock(&shm_ids(ns), id);
if (IS_ERR(ipcp))
return (struct shmid_kernel *)ipcp;
return container_of(ipcp, struct shmid_kernel, shm_perm);
}
static inline struct shmid_kernel *shm_lock_check(struct ipc_namespace *ns,
int id)
{
struct kern_ipc_perm *ipcp = ipc_lock_check(&shm_ids(ns), id);
if (IS_ERR(ipcp))
return (struct shmid_kernel *)ipcp;
return container_of(ipcp, struct shmid_kernel, shm_perm);
}
static inline void shm_rmid(struct ipc_namespace *ns, struct shmid_kernel *s)
{
ipc_rmid(&shm_ids(ns), &s->shm_perm);
}
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
/* This is called by fork, once for every shm attach. */
static void shm_open(struct vm_area_struct *vma)
{
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
struct file *file = vma->vm_file;
struct shm_file_data *sfd = shm_file_data(file);
struct shmid_kernel *shp;
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
shp = shm_lock(sfd->ns, sfd->id);
BUG_ON(IS_ERR(shp));
shp->shm_atim = get_seconds();
shp->shm_lprid = task_tgid_vnr(current);
shp->shm_nattch++;
shm_unlock(shp);
}
/*
* shm_destroy - free the struct shmid_kernel
*
* @ns: namespace
* @shp: struct to free
*
* It has to be called with shp and shm_ids.rw_mutex (writer) locked,
* but returns with shp unlocked and freed.
*/
static void shm_destroy(struct ipc_namespace *ns, struct shmid_kernel *shp)
{
ns->shm_tot -= (shp->shm_segsz + PAGE_SIZE - 1) >> PAGE_SHIFT;
shm_rmid(ns, shp);
shm_unlock(shp);
if (!is_file_hugepages(shp->shm_file))
shmem_lock(shp->shm_file, 0, shp->mlock_user);
else
user_shm_unlock(shp->shm_file->f_path.dentry->d_inode->i_size,
shp->mlock_user);
fput (shp->shm_file);
security_shm_free(shp);
ipc_rcu_putref(shp);
}
/*
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
* remove the attach descriptor vma.
* free memory for segment if it is marked destroyed.
* The descriptor has already been removed from the current->mm->mmap list
* and will later be kfree()d.
*/
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
static void shm_close(struct vm_area_struct *vma)
{
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
struct file * file = vma->vm_file;
struct shm_file_data *sfd = shm_file_data(file);
struct shmid_kernel *shp;
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
struct ipc_namespace *ns = sfd->ns;
down_write(&shm_ids(ns).rw_mutex);
/* remove from the list of attaches of the shm segment */
shp = shm_lock(ns, sfd->id);
BUG_ON(IS_ERR(shp));
shp->shm_lprid = task_tgid_vnr(current);
shp->shm_dtim = get_seconds();
shp->shm_nattch--;
if(shp->shm_nattch == 0 &&
shp->shm_perm.mode & SHM_DEST)
shm_destroy(ns, shp);
else
shm_unlock(shp);
up_write(&shm_ids(ns).rw_mutex);
}
static int shm_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
{
struct file *file = vma->vm_file;
struct shm_file_data *sfd = shm_file_data(file);
return sfd->vm_ops->fault(vma, vmf);
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
}
#ifdef CONFIG_NUMA
static int shm_set_policy(struct vm_area_struct *vma, struct mempolicy *new)
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
{
struct file *file = vma->vm_file;
struct shm_file_data *sfd = shm_file_data(file);
int err = 0;
if (sfd->vm_ops->set_policy)
err = sfd->vm_ops->set_policy(vma, new);
return err;
}
static struct mempolicy *shm_get_policy(struct vm_area_struct *vma,
unsigned long addr)
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
{
struct file *file = vma->vm_file;
struct shm_file_data *sfd = shm_file_data(file);
struct mempolicy *pol = NULL;
if (sfd->vm_ops->get_policy)
pol = sfd->vm_ops->get_policy(vma, addr);
mempolicy: rework mempolicy Reference Counting [yet again] After further discussion with Christoph Lameter, it has become clear that my earlier attempts to clean up the mempolicy reference counting were a bit of overkill in some areas, resulting in superflous ref/unref in what are usually fast paths. In other areas, further inspection reveals that I botched the unref for interleave policies. A separate patch, suitable for upstream/stable trees, fixes up the known errors in the previous attempt to fix reference counting. This patch reworks the memory policy referencing counting and, one hopes, simplifies the code. Maybe I'll get it right this time. See the update to the numa_memory_policy.txt document for a discussion of memory policy reference counting that motivates this patch. Summary: Lookup of mempolicy, based on (vma, address) need only add a reference for shared policy, and we need only unref the policy when finished for shared policies. So, this patch backs out all of the unneeded extra reference counting added by my previous attempt. It then unrefs only shared policies when we're finished with them, using the mpol_cond_put() [conditional put] helper function introduced by this patch. Note that shmem_swapin() calls read_swap_cache_async() with a dummy vma containing just the policy. read_swap_cache_async() can call alloc_page_vma() multiple times, so we can't let alloc_page_vma() unref the shared policy in this case. To avoid this, we make a copy of any non-null shared policy and remove the MPOL_F_SHARED flag from the copy. This copy occurs before reading a page [or multiple pages] from swap, so the overhead should not be an issue here. I introduced a new static inline function "mpol_cond_copy()" to copy the shared policy to an on-stack policy and remove the flags that would require a conditional free. The current implementation of mpol_cond_copy() assumes that the struct mempolicy contains no pointers to dynamically allocated structures that must be duplicated or reference counted during copy. Signed-off-by: Lee Schermerhorn <lee.schermerhorn@hp.com> Cc: Christoph Lameter <clameter@sgi.com> Cc: David Rientjes <rientjes@google.com> Cc: Mel Gorman <mel@csn.ul.ie> Cc: Andi Kleen <ak@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-28 09:13:16 +00:00
else if (vma->vm_policy)
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
pol = vma->vm_policy;
mempolicy: rework mempolicy Reference Counting [yet again] After further discussion with Christoph Lameter, it has become clear that my earlier attempts to clean up the mempolicy reference counting were a bit of overkill in some areas, resulting in superflous ref/unref in what are usually fast paths. In other areas, further inspection reveals that I botched the unref for interleave policies. A separate patch, suitable for upstream/stable trees, fixes up the known errors in the previous attempt to fix reference counting. This patch reworks the memory policy referencing counting and, one hopes, simplifies the code. Maybe I'll get it right this time. See the update to the numa_memory_policy.txt document for a discussion of memory policy reference counting that motivates this patch. Summary: Lookup of mempolicy, based on (vma, address) need only add a reference for shared policy, and we need only unref the policy when finished for shared policies. So, this patch backs out all of the unneeded extra reference counting added by my previous attempt. It then unrefs only shared policies when we're finished with them, using the mpol_cond_put() [conditional put] helper function introduced by this patch. Note that shmem_swapin() calls read_swap_cache_async() with a dummy vma containing just the policy. read_swap_cache_async() can call alloc_page_vma() multiple times, so we can't let alloc_page_vma() unref the shared policy in this case. To avoid this, we make a copy of any non-null shared policy and remove the MPOL_F_SHARED flag from the copy. This copy occurs before reading a page [or multiple pages] from swap, so the overhead should not be an issue here. I introduced a new static inline function "mpol_cond_copy()" to copy the shared policy to an on-stack policy and remove the flags that would require a conditional free. The current implementation of mpol_cond_copy() assumes that the struct mempolicy contains no pointers to dynamically allocated structures that must be duplicated or reference counted during copy. Signed-off-by: Lee Schermerhorn <lee.schermerhorn@hp.com> Cc: Christoph Lameter <clameter@sgi.com> Cc: David Rientjes <rientjes@google.com> Cc: Mel Gorman <mel@csn.ul.ie> Cc: Andi Kleen <ak@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-28 09:13:16 +00:00
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
return pol;
}
#endif
static int shm_mmap(struct file * file, struct vm_area_struct * vma)
{
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
struct shm_file_data *sfd = shm_file_data(file);
int ret;
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
ret = sfd->file->f_op->mmap(sfd->file, vma);
if (ret != 0)
return ret;
sfd->vm_ops = vma->vm_ops;
#ifdef CONFIG_MMU
mm: merge populate and nopage into fault (fixes nonlinear) Nonlinear mappings are (AFAIKS) simply a virtual memory concept that encodes the virtual address -> file offset differently from linear mappings. ->populate is a layering violation because the filesystem/pagecache code should need to know anything about the virtual memory mapping. The hitch here is that the ->nopage handler didn't pass down enough information (ie. pgoff). But it is more logical to pass pgoff rather than have the ->nopage function calculate it itself anyway (because that's a similar layering violation). Having the populate handler install the pte itself is likewise a nasty thing to be doing. This patch introduces a new fault handler that replaces ->nopage and ->populate and (later) ->nopfn. Most of the old mechanism is still in place so there is a lot of duplication and nice cleanups that can be removed if everyone switches over. The rationale for doing this in the first place is that nonlinear mappings are subject to the pagefault vs invalidate/truncate race too, and it seemed stupid to duplicate the synchronisation logic rather than just consolidate the two. After this patch, MAP_NONBLOCK no longer sets up ptes for pages present in pagecache. Seems like a fringe functionality anyway. NOPAGE_REFAULT is removed. This should be implemented with ->fault, and no users have hit mainline yet. [akpm@linux-foundation.org: cleanup] [randy.dunlap@oracle.com: doc. fixes for readahead] [akpm@linux-foundation.org: build fix] Signed-off-by: Nick Piggin <npiggin@suse.de> Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com> Cc: Mark Fasheh <mark.fasheh@oracle.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-19 08:46:59 +00:00
BUG_ON(!sfd->vm_ops->fault);
#endif
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
vma->vm_ops = &shm_vm_ops;
shm_open(vma);
return ret;
}
static int shm_release(struct inode *ino, struct file *file)
{
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
struct shm_file_data *sfd = shm_file_data(file);
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
put_ipc_ns(sfd->ns);
shm_file_data(file) = NULL;
kfree(sfd);
return 0;
}
static int shm_fsync(struct file *file, struct dentry *dentry, int datasync)
{
int (*fsync) (struct file *, struct dentry *, int datasync);
struct shm_file_data *sfd = shm_file_data(file);
int ret = -EINVAL;
fsync = sfd->file->f_op->fsync;
if (fsync)
ret = fsync(sfd->file, sfd->file->f_path.dentry, datasync);
return ret;
}
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
static unsigned long shm_get_unmapped_area(struct file *file,
unsigned long addr, unsigned long len, unsigned long pgoff,
unsigned long flags)
{
struct shm_file_data *sfd = shm_file_data(file);
return get_unmapped_area(sfd->file, addr, len, pgoff, flags);
}
int is_file_shm_hugepages(struct file *file)
{
int ret = 0;
if (file->f_op == &shm_file_operations) {
struct shm_file_data *sfd;
sfd = shm_file_data(file);
ret = is_file_hugepages(sfd->file);
}
return ret;
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
}
static const struct file_operations shm_file_operations = {
.mmap = shm_mmap,
.fsync = shm_fsync,
.release = shm_release,
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
.get_unmapped_area = shm_get_unmapped_area,
};
static struct vm_operations_struct shm_vm_ops = {
.open = shm_open, /* callback for a new vm-area open */
.close = shm_close, /* callback for when the vm-area is released */
mm: merge populate and nopage into fault (fixes nonlinear) Nonlinear mappings are (AFAIKS) simply a virtual memory concept that encodes the virtual address -> file offset differently from linear mappings. ->populate is a layering violation because the filesystem/pagecache code should need to know anything about the virtual memory mapping. The hitch here is that the ->nopage handler didn't pass down enough information (ie. pgoff). But it is more logical to pass pgoff rather than have the ->nopage function calculate it itself anyway (because that's a similar layering violation). Having the populate handler install the pte itself is likewise a nasty thing to be doing. This patch introduces a new fault handler that replaces ->nopage and ->populate and (later) ->nopfn. Most of the old mechanism is still in place so there is a lot of duplication and nice cleanups that can be removed if everyone switches over. The rationale for doing this in the first place is that nonlinear mappings are subject to the pagefault vs invalidate/truncate race too, and it seemed stupid to duplicate the synchronisation logic rather than just consolidate the two. After this patch, MAP_NONBLOCK no longer sets up ptes for pages present in pagecache. Seems like a fringe functionality anyway. NOPAGE_REFAULT is removed. This should be implemented with ->fault, and no users have hit mainline yet. [akpm@linux-foundation.org: cleanup] [randy.dunlap@oracle.com: doc. fixes for readahead] [akpm@linux-foundation.org: build fix] Signed-off-by: Nick Piggin <npiggin@suse.de> Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com> Cc: Mark Fasheh <mark.fasheh@oracle.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-19 08:46:59 +00:00
.fault = shm_fault,
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
#if defined(CONFIG_NUMA)
.set_policy = shm_set_policy,
.get_policy = shm_get_policy,
#endif
};
/**
* newseg - Create a new shared memory segment
* @ns: namespace
* @params: ptr to the structure that contains key, size and shmflg
*
* Called with shm_ids.rw_mutex held as a writer.
*/
static int newseg(struct ipc_namespace *ns, struct ipc_params *params)
{
key_t key = params->key;
int shmflg = params->flg;
size_t size = params->u.size;
int error;
struct shmid_kernel *shp;
int numpages = (size + PAGE_SIZE -1) >> PAGE_SHIFT;
struct file * file;
char name[13];
int id;
Do not account for the address space used by hugetlbfs using VM_ACCOUNT When overcommit is disabled, the core VM accounts for pages used by anonymous shared, private mappings and special mappings. It keeps track of VMAs that should be accounted for with VM_ACCOUNT and VMAs that never had a reserve with VM_NORESERVE. Overcommit for hugetlbfs is much riskier than overcommit for base pages due to contiguity requirements. It avoids overcommiting on both shared and private mappings using reservation counters that are checked and updated during mmap(). This ensures (within limits) that hugepages exist in the future when faults occurs or it is too easy to applications to be SIGKILLed. As hugetlbfs makes its own reservations of a different unit to the base page size, VM_ACCOUNT should never be set. Even if the units were correct, we would double account for the usage in the core VM and hugetlbfs. VM_NORESERVE may be set because an application can request no reserves be made for hugetlbfs at the risk of getting killed later. With commit fc8744adc870a8d4366908221508bb113d8b72ee, VM_NORESERVE and VM_ACCOUNT are getting unconditionally set for hugetlbfs-backed mappings. This breaks the accounting for both the core VM and hugetlbfs, can trigger an OOM storm when hugepage pools are too small lockups and corrupted counters otherwise are used. This patch brings hugetlbfs more in line with how the core VM treats VM_NORESERVE but prevents VM_ACCOUNT being set. Signed-off-by: Mel Gorman <mel@csn.ul.ie> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-02-10 14:02:27 +00:00
int acctflag = 0;
if (size < SHMMIN || size > ns->shm_ctlmax)
return -EINVAL;
if (ns->shm_tot + numpages > ns->shm_ctlall)
return -ENOSPC;
shp = ipc_rcu_alloc(sizeof(*shp));
if (!shp)
return -ENOMEM;
shp->shm_perm.key = key;
shp->shm_perm.mode = (shmflg & S_IRWXUGO);
shp->mlock_user = NULL;
shp->shm_perm.security = NULL;
error = security_shm_alloc(shp);
if (error) {
ipc_rcu_putref(shp);
return error;
}
sprintf (name, "SYSV%08x", key);
if (shmflg & SHM_HUGETLB) {
Do not account for the address space used by hugetlbfs using VM_ACCOUNT When overcommit is disabled, the core VM accounts for pages used by anonymous shared, private mappings and special mappings. It keeps track of VMAs that should be accounted for with VM_ACCOUNT and VMAs that never had a reserve with VM_NORESERVE. Overcommit for hugetlbfs is much riskier than overcommit for base pages due to contiguity requirements. It avoids overcommiting on both shared and private mappings using reservation counters that are checked and updated during mmap(). This ensures (within limits) that hugepages exist in the future when faults occurs or it is too easy to applications to be SIGKILLed. As hugetlbfs makes its own reservations of a different unit to the base page size, VM_ACCOUNT should never be set. Even if the units were correct, we would double account for the usage in the core VM and hugetlbfs. VM_NORESERVE may be set because an application can request no reserves be made for hugetlbfs at the risk of getting killed later. With commit fc8744adc870a8d4366908221508bb113d8b72ee, VM_NORESERVE and VM_ACCOUNT are getting unconditionally set for hugetlbfs-backed mappings. This breaks the accounting for both the core VM and hugetlbfs, can trigger an OOM storm when hugepage pools are too small lockups and corrupted counters otherwise are used. This patch brings hugetlbfs more in line with how the core VM treats VM_NORESERVE but prevents VM_ACCOUNT being set. Signed-off-by: Mel Gorman <mel@csn.ul.ie> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-02-10 14:02:27 +00:00
/* hugetlb_file_setup applies strict accounting */
if (shmflg & SHM_NORESERVE)
acctflag = VM_NORESERVE;
file = hugetlb_file_setup(name, size, acctflag);
shp->mlock_user = current_user();
} else {
/*
* Do not allow no accounting for OVERCOMMIT_NEVER, even
* if it's asked for.
*/
if ((shmflg & SHM_NORESERVE) &&
sysctl_overcommit_memory != OVERCOMMIT_NEVER)
acctflag = VM_NORESERVE;
file = shmem_file_setup(name, size, acctflag);
}
error = PTR_ERR(file);
if (IS_ERR(file))
goto no_file;
id = ipc_addid(&shm_ids(ns), &shp->shm_perm, ns->shm_ctlmni);
if (id < 0) {
error = id;
goto no_id;
}
shp->shm_cprid = task_tgid_vnr(current);
shp->shm_lprid = 0;
shp->shm_atim = shp->shm_dtim = 0;
shp->shm_ctim = get_seconds();
shp->shm_segsz = size;
shp->shm_nattch = 0;
shp->shm_file = file;
/*
* shmid gets reported as "inode#" in /proc/pid/maps.
* proc-ps tools use this. Changing this will break them.
*/
file->f_dentry->d_inode->i_ino = shp->shm_perm.id;
ns->shm_tot += numpages;
error = shp->shm_perm.id;
shm_unlock(shp);
return error;
no_id:
fput(file);
no_file:
security_shm_free(shp);
ipc_rcu_putref(shp);
return error;
}
/*
* Called with shm_ids.rw_mutex and ipcp locked.
*/
static inline int shm_security(struct kern_ipc_perm *ipcp, int shmflg)
{
struct shmid_kernel *shp;
shp = container_of(ipcp, struct shmid_kernel, shm_perm);
return security_shm_associate(shp, shmflg);
}
/*
* Called with shm_ids.rw_mutex and ipcp locked.
*/
static inline int shm_more_checks(struct kern_ipc_perm *ipcp,
struct ipc_params *params)
{
struct shmid_kernel *shp;
shp = container_of(ipcp, struct shmid_kernel, shm_perm);
if (shp->shm_segsz < params->u.size)
return -EINVAL;
return 0;
}
SYSCALL_DEFINE3(shmget, key_t, key, size_t, size, int, shmflg)
{
struct ipc_namespace *ns;
struct ipc_ops shm_ops;
struct ipc_params shm_params;
ns = current->nsproxy->ipc_ns;
shm_ops.getnew = newseg;
shm_ops.associate = shm_security;
shm_ops.more_checks = shm_more_checks;
shm_params.key = key;
shm_params.flg = shmflg;
shm_params.u.size = size;
return ipcget(ns, &shm_ids(ns), &shm_ops, &shm_params);
}
static inline unsigned long copy_shmid_to_user(void __user *buf, struct shmid64_ds *in, int version)
{
switch(version) {
case IPC_64:
return copy_to_user(buf, in, sizeof(*in));
case IPC_OLD:
{
struct shmid_ds out;
ipc64_perm_to_ipc_perm(&in->shm_perm, &out.shm_perm);
out.shm_segsz = in->shm_segsz;
out.shm_atime = in->shm_atime;
out.shm_dtime = in->shm_dtime;
out.shm_ctime = in->shm_ctime;
out.shm_cpid = in->shm_cpid;
out.shm_lpid = in->shm_lpid;
out.shm_nattch = in->shm_nattch;
return copy_to_user(buf, &out, sizeof(out));
}
default:
return -EINVAL;
}
}
static inline unsigned long
copy_shmid_from_user(struct shmid64_ds *out, void __user *buf, int version)
{
switch(version) {
case IPC_64:
if (copy_from_user(out, buf, sizeof(*out)))
return -EFAULT;
return 0;
case IPC_OLD:
{
struct shmid_ds tbuf_old;
if (copy_from_user(&tbuf_old, buf, sizeof(tbuf_old)))
return -EFAULT;
out->shm_perm.uid = tbuf_old.shm_perm.uid;
out->shm_perm.gid = tbuf_old.shm_perm.gid;
out->shm_perm.mode = tbuf_old.shm_perm.mode;
return 0;
}
default:
return -EINVAL;
}
}
static inline unsigned long copy_shminfo_to_user(void __user *buf, struct shminfo64 *in, int version)
{
switch(version) {
case IPC_64:
return copy_to_user(buf, in, sizeof(*in));
case IPC_OLD:
{
struct shminfo out;
if(in->shmmax > INT_MAX)
out.shmmax = INT_MAX;
else
out.shmmax = (int)in->shmmax;
out.shmmin = in->shmmin;
out.shmmni = in->shmmni;
out.shmseg = in->shmseg;
out.shmall = in->shmall;
return copy_to_user(buf, &out, sizeof(out));
}
default:
return -EINVAL;
}
}
/*
* Called with shm_ids.rw_mutex held as a reader
*/
static void shm_get_stat(struct ipc_namespace *ns, unsigned long *rss,
unsigned long *swp)
{
int next_id;
int total, in_use;
*rss = 0;
*swp = 0;
in_use = shm_ids(ns).in_use;
for (total = 0, next_id = 0; total < in_use; next_id++) {
struct kern_ipc_perm *ipc;
struct shmid_kernel *shp;
struct inode *inode;
ipc = idr_find(&shm_ids(ns).ipcs_idr, next_id);
if (ipc == NULL)
continue;
shp = container_of(ipc, struct shmid_kernel, shm_perm);
inode = shp->shm_file->f_path.dentry->d_inode;
if (is_file_hugepages(shp->shm_file)) {
struct address_space *mapping = inode->i_mapping;
struct hstate *h = hstate_file(shp->shm_file);
*rss += pages_per_huge_page(h) * mapping->nrpages;
} else {
#ifdef CONFIG_SHMEM
struct shmem_inode_info *info = SHMEM_I(inode);
spin_lock(&info->lock);
*rss += inode->i_mapping->nrpages;
*swp += info->swapped;
spin_unlock(&info->lock);
#else
*rss += inode->i_mapping->nrpages;
#endif
}
total++;
}
}
/*
* This function handles some shmctl commands which require the rw_mutex
* to be held in write mode.
* NOTE: no locks must be held, the rw_mutex is taken inside this function.
*/
static int shmctl_down(struct ipc_namespace *ns, int shmid, int cmd,
struct shmid_ds __user *buf, int version)
{
struct kern_ipc_perm *ipcp;
struct shmid64_ds shmid64;
struct shmid_kernel *shp;
int err;
if (cmd == IPC_SET) {
if (copy_shmid_from_user(&shmid64, buf, version))
return -EFAULT;
}
ipcp = ipcctl_pre_down(&shm_ids(ns), shmid, cmd, &shmid64.shm_perm, 0);
if (IS_ERR(ipcp))
return PTR_ERR(ipcp);
shp = container_of(ipcp, struct shmid_kernel, shm_perm);
err = security_shm_shmctl(shp, cmd);
if (err)
goto out_unlock;
switch (cmd) {
case IPC_RMID:
do_shm_rmid(ns, ipcp);
goto out_up;
case IPC_SET:
ipc_update_perm(&shmid64.shm_perm, ipcp);
shp->shm_ctim = get_seconds();
break;
default:
err = -EINVAL;
}
out_unlock:
shm_unlock(shp);
out_up:
up_write(&shm_ids(ns).rw_mutex);
return err;
}
SYSCALL_DEFINE3(shmctl, int, shmid, int, cmd, struct shmid_ds __user *, buf)
{
struct shmid_kernel *shp;
int err, version;
struct ipc_namespace *ns;
if (cmd < 0 || shmid < 0) {
err = -EINVAL;
goto out;
}
version = ipc_parse_version(&cmd);
ns = current->nsproxy->ipc_ns;
switch (cmd) { /* replace with proc interface ? */
case IPC_INFO:
{
struct shminfo64 shminfo;
err = security_shm_shmctl(NULL, cmd);
if (err)
return err;
memset(&shminfo, 0, sizeof(shminfo));
shminfo.shmmni = shminfo.shmseg = ns->shm_ctlmni;
shminfo.shmmax = ns->shm_ctlmax;
shminfo.shmall = ns->shm_ctlall;
shminfo.shmmin = SHMMIN;
if(copy_shminfo_to_user (buf, &shminfo, version))
return -EFAULT;
down_read(&shm_ids(ns).rw_mutex);
err = ipc_get_maxid(&shm_ids(ns));
up_read(&shm_ids(ns).rw_mutex);
if(err<0)
err = 0;
goto out;
}
case SHM_INFO:
{
struct shm_info shm_info;
err = security_shm_shmctl(NULL, cmd);
if (err)
return err;
memset(&shm_info, 0, sizeof(shm_info));
down_read(&shm_ids(ns).rw_mutex);
shm_info.used_ids = shm_ids(ns).in_use;
shm_get_stat (ns, &shm_info.shm_rss, &shm_info.shm_swp);
shm_info.shm_tot = ns->shm_tot;
shm_info.swap_attempts = 0;
shm_info.swap_successes = 0;
err = ipc_get_maxid(&shm_ids(ns));
up_read(&shm_ids(ns).rw_mutex);
if (copy_to_user(buf, &shm_info, sizeof(shm_info))) {
err = -EFAULT;
goto out;
}
err = err < 0 ? 0 : err;
goto out;
}
case SHM_STAT:
case IPC_STAT:
{
struct shmid64_ds tbuf;
int result;
if (cmd == SHM_STAT) {
shp = shm_lock(ns, shmid);
if (IS_ERR(shp)) {
err = PTR_ERR(shp);
goto out;
}
result = shp->shm_perm.id;
} else {
shp = shm_lock_check(ns, shmid);
if (IS_ERR(shp)) {
err = PTR_ERR(shp);
goto out;
}
result = 0;
}
err = -EACCES;
if (ipcperms (&shp->shm_perm, S_IRUGO))
goto out_unlock;
err = security_shm_shmctl(shp, cmd);
if (err)
goto out_unlock;
memset(&tbuf, 0, sizeof(tbuf));
kernel_to_ipc64_perm(&shp->shm_perm, &tbuf.shm_perm);
tbuf.shm_segsz = shp->shm_segsz;
tbuf.shm_atime = shp->shm_atim;
tbuf.shm_dtime = shp->shm_dtim;
tbuf.shm_ctime = shp->shm_ctim;
tbuf.shm_cpid = shp->shm_cprid;
tbuf.shm_lpid = shp->shm_lprid;
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
tbuf.shm_nattch = shp->shm_nattch;
shm_unlock(shp);
if(copy_shmid_to_user (buf, &tbuf, version))
err = -EFAULT;
else
err = result;
goto out;
}
case SHM_LOCK:
case SHM_UNLOCK:
{
struct file *uninitialized_var(shm_file);
lru_add_drain_all(); /* drain pagevecs to lru lists */
shp = shm_lock_check(ns, shmid);
if (IS_ERR(shp)) {
err = PTR_ERR(shp);
goto out;
}
audit_ipc_obj(&(shp->shm_perm));
[PATCH] Rework of IPC auditing 1) The audit_ipc_perms() function has been split into two different functions: - audit_ipc_obj() - audit_ipc_set_perm() There's a key shift here... The audit_ipc_obj() collects the uid, gid, mode, and SElinux context label of the current ipc object. This audit_ipc_obj() hook is now found in several places. Most notably, it is hooked in ipcperms(), which is called in various places around the ipc code permforming a MAC check. Additionally there are several places where *checkid() is used to validate that an operation is being performed on a valid object while not necessarily having a nearby ipcperms() call. In these locations, audit_ipc_obj() is called to ensure that the information is captured by the audit system. The audit_set_new_perm() function is called any time the permissions on the ipc object changes. In this case, the NEW permissions are recorded (and note that an audit_ipc_obj() call exists just a few lines before each instance). 2) Support for an AUDIT_IPC_SET_PERM audit message type. This allows for separate auxiliary audit records for normal operations on an IPC object and permissions changes. Note that the same struct audit_aux_data_ipcctl is used and populated, however there are separate audit_log_format statements based on the type of the message. Finally, the AUDIT_IPC block of code in audit_free_aux() was extended to handle aux messages of this new type. No more mem leaks I hope ;-) Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2006-04-02 21:07:33 +00:00
if (!capable(CAP_IPC_LOCK)) {
uid_t euid = current_euid();
err = -EPERM;
if (euid != shp->shm_perm.uid &&
euid != shp->shm_perm.cuid)
goto out_unlock;
if (cmd == SHM_LOCK &&
!current->signal->rlim[RLIMIT_MEMLOCK].rlim_cur)
goto out_unlock;
}
err = security_shm_shmctl(shp, cmd);
if (err)
goto out_unlock;
if(cmd==SHM_LOCK) {
struct user_struct *user = current_user();
if (!is_file_hugepages(shp->shm_file)) {
err = shmem_lock(shp->shm_file, 1, user);
if (!err && !(shp->shm_perm.mode & SHM_LOCKED)){
shp->shm_perm.mode |= SHM_LOCKED;
shp->mlock_user = user;
}
}
} else if (!is_file_hugepages(shp->shm_file)) {
shmem_lock(shp->shm_file, 0, shp->mlock_user);
shp->shm_perm.mode &= ~SHM_LOCKED;
shp->mlock_user = NULL;
}
shm_unlock(shp);
goto out;
}
case IPC_RMID:
case IPC_SET:
err = shmctl_down(ns, shmid, cmd, buf, version);
return err;
default:
return -EINVAL;
}
out_unlock:
shm_unlock(shp);
out:
return err;
}
/*
* Fix shmaddr, allocate descriptor, map shm, add attach descriptor to lists.
*
* NOTE! Despite the name, this is NOT a direct system call entrypoint. The
* "raddr" thing points to kernel space, and there has to be a wrapper around
* this.
*/
long do_shmat(int shmid, char __user *shmaddr, int shmflg, ulong *raddr)
{
struct shmid_kernel *shp;
unsigned long addr;
unsigned long size;
struct file * file;
int err;
unsigned long flags;
unsigned long prot;
int acc_mode;
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
unsigned long user_addr;
struct ipc_namespace *ns;
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
struct shm_file_data *sfd;
struct path path;
fmode_t f_mode;
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
err = -EINVAL;
if (shmid < 0)
goto out;
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
else if ((addr = (ulong)shmaddr)) {
if (addr & (SHMLBA-1)) {
if (shmflg & SHM_RND)
addr &= ~(SHMLBA-1); /* round down */
else
#ifndef __ARCH_FORCE_SHMLBA
if (addr & ~PAGE_MASK)
#endif
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
goto out;
}
flags = MAP_SHARED | MAP_FIXED;
} else {
if ((shmflg & SHM_REMAP))
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
goto out;
flags = MAP_SHARED;
}
if (shmflg & SHM_RDONLY) {
prot = PROT_READ;
acc_mode = S_IRUGO;
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
f_mode = FMODE_READ;
} else {
prot = PROT_READ | PROT_WRITE;
acc_mode = S_IRUGO | S_IWUGO;
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
f_mode = FMODE_READ | FMODE_WRITE;
}
if (shmflg & SHM_EXEC) {
prot |= PROT_EXEC;
acc_mode |= S_IXUGO;
}
/*
* We cannot rely on the fs check since SYSV IPC does have an
* additional creator id...
*/
ns = current->nsproxy->ipc_ns;
shp = shm_lock_check(ns, shmid);
if (IS_ERR(shp)) {
err = PTR_ERR(shp);
goto out;
}
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
err = -EACCES;
if (ipcperms(&shp->shm_perm, acc_mode))
goto out_unlock;
err = security_shm_shmat(shp, shmaddr, shmflg);
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
if (err)
goto out_unlock;
path.dentry = dget(shp->shm_file->f_path.dentry);
r/o bind mounts: filesystem helpers for custom 'struct file's Why do we need r/o bind mounts? This feature allows a read-only view into a read-write filesystem. In the process of doing that, it also provides infrastructure for keeping track of the number of writers to any given mount. This has a number of uses. It allows chroots to have parts of filesystems writable. It will be useful for containers in the future because users may have root inside a container, but should not be allowed to write to somefilesystems. This also replaces patches that vserver has had out of the tree for several years. It allows security enhancement by making sure that parts of your filesystem read-only (such as when you don't trust your FTP server), when you don't want to have entire new filesystems mounted, or when you want atime selectively updated. I've been using the following script to test that the feature is working as desired. It takes a directory and makes a regular bind and a r/o bind mount of it. It then performs some normal filesystem operations on the three directories, including ones that are expected to fail, like creating a file on the r/o mount. This patch: Some filesystems forego the vfs and may_open() and create their own 'struct file's. This patch creates a couple of helper functions which can be used by these filesystems, and will provide a unified place which the r/o bind mount code may patch. Also, rename an existing, static-scope init_file() to a less generic name. Signed-off-by: Dave Hansen <haveblue@us.ibm.com> Cc: Christoph Hellwig <hch@lst.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-17 06:31:13 +00:00
path.mnt = shp->shm_file->f_path.mnt;
shp->shm_nattch++;
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
size = i_size_read(path.dentry->d_inode);
shm_unlock(shp);
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
err = -ENOMEM;
sfd = kzalloc(sizeof(*sfd), GFP_KERNEL);
if (!sfd)
r/o bind mounts: filesystem helpers for custom 'struct file's Why do we need r/o bind mounts? This feature allows a read-only view into a read-write filesystem. In the process of doing that, it also provides infrastructure for keeping track of the number of writers to any given mount. This has a number of uses. It allows chroots to have parts of filesystems writable. It will be useful for containers in the future because users may have root inside a container, but should not be allowed to write to somefilesystems. This also replaces patches that vserver has had out of the tree for several years. It allows security enhancement by making sure that parts of your filesystem read-only (such as when you don't trust your FTP server), when you don't want to have entire new filesystems mounted, or when you want atime selectively updated. I've been using the following script to test that the feature is working as desired. It takes a directory and makes a regular bind and a r/o bind mount of it. It then performs some normal filesystem operations on the three directories, including ones that are expected to fail, like creating a file on the r/o mount. This patch: Some filesystems forego the vfs and may_open() and create their own 'struct file's. This patch creates a couple of helper functions which can be used by these filesystems, and will provide a unified place which the r/o bind mount code may patch. Also, rename an existing, static-scope init_file() to a less generic name. Signed-off-by: Dave Hansen <haveblue@us.ibm.com> Cc: Christoph Hellwig <hch@lst.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-17 06:31:13 +00:00
goto out_put_dentry;
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
r/o bind mounts: filesystem helpers for custom 'struct file's Why do we need r/o bind mounts? This feature allows a read-only view into a read-write filesystem. In the process of doing that, it also provides infrastructure for keeping track of the number of writers to any given mount. This has a number of uses. It allows chroots to have parts of filesystems writable. It will be useful for containers in the future because users may have root inside a container, but should not be allowed to write to somefilesystems. This also replaces patches that vserver has had out of the tree for several years. It allows security enhancement by making sure that parts of your filesystem read-only (such as when you don't trust your FTP server), when you don't want to have entire new filesystems mounted, or when you want atime selectively updated. I've been using the following script to test that the feature is working as desired. It takes a directory and makes a regular bind and a r/o bind mount of it. It then performs some normal filesystem operations on the three directories, including ones that are expected to fail, like creating a file on the r/o mount. This patch: Some filesystems forego the vfs and may_open() and create their own 'struct file's. This patch creates a couple of helper functions which can be used by these filesystems, and will provide a unified place which the r/o bind mount code may patch. Also, rename an existing, static-scope init_file() to a less generic name. Signed-off-by: Dave Hansen <haveblue@us.ibm.com> Cc: Christoph Hellwig <hch@lst.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-17 06:31:13 +00:00
file = alloc_file(path.mnt, path.dentry, f_mode, &shm_file_operations);
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
if (!file)
goto out_free;
ima_counts_get(file);
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
file->private_data = sfd;
file->f_mapping = shp->shm_file->f_mapping;
sfd->id = shp->shm_perm.id;
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
sfd->ns = get_ipc_ns(ns);
sfd->file = shp->shm_file;
sfd->vm_ops = NULL;
down_write(&current->mm->mmap_sem);
if (addr && !(shmflg & SHM_REMAP)) {
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
err = -EINVAL;
if (find_vma_intersection(current->mm, addr, addr + size))
goto invalid;
/*
* If shm segment goes below stack, make sure there is some
* space left for the stack to grow (at least 4 pages).
*/
if (addr < current->mm->start_stack &&
addr > current->mm->start_stack - size - PAGE_SIZE * 5)
goto invalid;
}
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
user_addr = do_mmap (file, addr, size, prot, flags, 0);
*raddr = user_addr;
err = 0;
if (IS_ERR_VALUE(user_addr))
err = (long)user_addr;
invalid:
up_write(&current->mm->mmap_sem);
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
fput(file);
out_nattch:
down_write(&shm_ids(ns).rw_mutex);
shp = shm_lock(ns, shmid);
BUG_ON(IS_ERR(shp));
shp->shm_nattch--;
if(shp->shm_nattch == 0 &&
shp->shm_perm.mode & SHM_DEST)
shm_destroy(ns, shp);
else
shm_unlock(shp);
up_write(&shm_ids(ns).rw_mutex);
out:
return err;
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
out_unlock:
shm_unlock(shp);
goto out;
out_free:
kfree(sfd);
r/o bind mounts: filesystem helpers for custom 'struct file's Why do we need r/o bind mounts? This feature allows a read-only view into a read-write filesystem. In the process of doing that, it also provides infrastructure for keeping track of the number of writers to any given mount. This has a number of uses. It allows chroots to have parts of filesystems writable. It will be useful for containers in the future because users may have root inside a container, but should not be allowed to write to somefilesystems. This also replaces patches that vserver has had out of the tree for several years. It allows security enhancement by making sure that parts of your filesystem read-only (such as when you don't trust your FTP server), when you don't want to have entire new filesystems mounted, or when you want atime selectively updated. I've been using the following script to test that the feature is working as desired. It takes a directory and makes a regular bind and a r/o bind mount of it. It then performs some normal filesystem operations on the three directories, including ones that are expected to fail, like creating a file on the r/o mount. This patch: Some filesystems forego the vfs and may_open() and create their own 'struct file's. This patch creates a couple of helper functions which can be used by these filesystems, and will provide a unified place which the r/o bind mount code may patch. Also, rename an existing, static-scope init_file() to a less generic name. Signed-off-by: Dave Hansen <haveblue@us.ibm.com> Cc: Christoph Hellwig <hch@lst.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-17 06:31:13 +00:00
out_put_dentry:
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
dput(path.dentry);
goto out_nattch;
}
SYSCALL_DEFINE3(shmat, int, shmid, char __user *, shmaddr, int, shmflg)
{
unsigned long ret;
long err;
err = do_shmat(shmid, shmaddr, shmflg, &ret);
if (err)
return err;
force_successful_syscall_return();
return (long)ret;
}
/*
* detach and kill segment if marked destroyed.
* The work is done in shm_close.
*/
SYSCALL_DEFINE1(shmdt, char __user *, shmaddr)
{
struct mm_struct *mm = current->mm;
struct vm_area_struct *vma;
unsigned long addr = (unsigned long)shmaddr;
int retval = -EINVAL;
#ifdef CONFIG_MMU
loff_t size = 0;
struct vm_area_struct *next;
#endif
if (addr & ~PAGE_MASK)
return retval;
down_write(&mm->mmap_sem);
/*
* This function tries to be smart and unmap shm segments that
* were modified by partial mlock or munmap calls:
* - It first determines the size of the shm segment that should be
* unmapped: It searches for a vma that is backed by shm and that
* started at address shmaddr. It records it's size and then unmaps
* it.
* - Then it unmaps all shm vmas that started at shmaddr and that
* are within the initially determined size.
* Errors from do_munmap are ignored: the function only fails if
* it's called with invalid parameters or if it's called to unmap
* a part of a vma. Both calls in this function are for full vmas,
* the parameters are directly copied from the vma itself and always
* valid - therefore do_munmap cannot fail. (famous last words?)
*/
/*
* If it had been mremap()'d, the starting address would not
* match the usual checks anyway. So assume all vma's are
* above the starting address given.
*/
vma = find_vma(mm, addr);
NOMMU: Make VMAs per MM as for MMU-mode linux Make VMAs per mm_struct as for MMU-mode linux. This solves two problems: (1) In SYSV SHM where nattch for a segment does not reflect the number of shmat's (and forks) done. (2) In mmap() where the VMA's vm_mm is set to point to the parent mm by an exec'ing process when VM_EXECUTABLE is specified, regardless of the fact that a VMA might be shared and already have its vm_mm assigned to another process or a dead process. A new struct (vm_region) is introduced to track a mapped region and to remember the circumstances under which it may be shared and the vm_list_struct structure is discarded as it's no longer required. This patch makes the following additional changes: (1) Regions are now allocated with alloc_pages() rather than kmalloc() and with no recourse to __GFP_COMP, so the pages are not composite. Instead, each page has a reference on it held by the region. Anything else that is interested in such a page will have to get a reference on it to retain it. When the pages are released due to unmapping, each page is passed to put_page() and will be freed when the page usage count reaches zero. (2) Excess pages are trimmed after an allocation as the allocation must be made as a power-of-2 quantity of pages. (3) VMAs are added to the parent MM's R/B tree and mmap lists. As an MM may end up with overlapping VMAs within the tree, the VMA struct address is appended to the sort key. (4) Non-anonymous VMAs are now added to the backing inode's prio list. (5) Holes may be punched in anonymous VMAs with munmap(), releasing parts of the backing region. The VMA and region structs will be split if necessary. (6) sys_shmdt() only releases one attachment to a SYSV IPC shared memory segment instead of all the attachments at that addresss. Multiple shmat()'s return the same address under NOMMU-mode instead of different virtual addresses as under MMU-mode. (7) Core dumping for ELF-FDPIC requires fewer exceptions for NOMMU-mode. (8) /proc/maps is now the global list of mapped regions, and may list bits that aren't actually mapped anywhere. (9) /proc/meminfo gains a line (tagged "MmapCopy") that indicates the amount of RAM currently allocated by mmap to hold mappable regions that can't be mapped directly. These are copies of the backing device or file if not anonymous. These changes make NOMMU mode more similar to MMU mode. The downside is that NOMMU mode requires some extra memory to track things over NOMMU without this patch (VMAs are no longer shared, and there are now region structs). Signed-off-by: David Howells <dhowells@redhat.com> Tested-by: Mike Frysinger <vapier.adi@gmail.com> Acked-by: Paul Mundt <lethal@linux-sh.org>
2009-01-08 12:04:47 +00:00
#ifdef CONFIG_MMU
while (vma) {
next = vma->vm_next;
/*
* Check if the starting address would match, i.e. it's
* a fragment created by mprotect() and/or munmap(), or it
* otherwise it starts at this address with no hassles.
*/
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
if ((vma->vm_ops == &shm_vm_ops) &&
(vma->vm_start - addr)/PAGE_SIZE == vma->vm_pgoff) {
size = vma->vm_file->f_path.dentry->d_inode->i_size;
do_munmap(mm, vma->vm_start, vma->vm_end - vma->vm_start);
/*
* We discovered the size of the shm segment, so
* break out of here and fall through to the next
* loop that uses the size information to stop
* searching for matching vma's.
*/
retval = 0;
vma = next;
break;
}
vma = next;
}
/*
* We need look no further than the maximum address a fragment
* could possibly have landed at. Also cast things to loff_t to
* prevent overflows and make comparisions vs. equal-width types.
*/
size = PAGE_ALIGN(size);
while (vma && (loff_t)(vma->vm_end - addr) <= size) {
next = vma->vm_next;
/* finding a matching vma now does not alter retval */
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
if ((vma->vm_ops == &shm_vm_ops) &&
(vma->vm_start - addr)/PAGE_SIZE == vma->vm_pgoff)
do_munmap(mm, vma->vm_start, vma->vm_end - vma->vm_start);
vma = next;
}
NOMMU: Make VMAs per MM as for MMU-mode linux Make VMAs per mm_struct as for MMU-mode linux. This solves two problems: (1) In SYSV SHM where nattch for a segment does not reflect the number of shmat's (and forks) done. (2) In mmap() where the VMA's vm_mm is set to point to the parent mm by an exec'ing process when VM_EXECUTABLE is specified, regardless of the fact that a VMA might be shared and already have its vm_mm assigned to another process or a dead process. A new struct (vm_region) is introduced to track a mapped region and to remember the circumstances under which it may be shared and the vm_list_struct structure is discarded as it's no longer required. This patch makes the following additional changes: (1) Regions are now allocated with alloc_pages() rather than kmalloc() and with no recourse to __GFP_COMP, so the pages are not composite. Instead, each page has a reference on it held by the region. Anything else that is interested in such a page will have to get a reference on it to retain it. When the pages are released due to unmapping, each page is passed to put_page() and will be freed when the page usage count reaches zero. (2) Excess pages are trimmed after an allocation as the allocation must be made as a power-of-2 quantity of pages. (3) VMAs are added to the parent MM's R/B tree and mmap lists. As an MM may end up with overlapping VMAs within the tree, the VMA struct address is appended to the sort key. (4) Non-anonymous VMAs are now added to the backing inode's prio list. (5) Holes may be punched in anonymous VMAs with munmap(), releasing parts of the backing region. The VMA and region structs will be split if necessary. (6) sys_shmdt() only releases one attachment to a SYSV IPC shared memory segment instead of all the attachments at that addresss. Multiple shmat()'s return the same address under NOMMU-mode instead of different virtual addresses as under MMU-mode. (7) Core dumping for ELF-FDPIC requires fewer exceptions for NOMMU-mode. (8) /proc/maps is now the global list of mapped regions, and may list bits that aren't actually mapped anywhere. (9) /proc/meminfo gains a line (tagged "MmapCopy") that indicates the amount of RAM currently allocated by mmap to hold mappable regions that can't be mapped directly. These are copies of the backing device or file if not anonymous. These changes make NOMMU mode more similar to MMU mode. The downside is that NOMMU mode requires some extra memory to track things over NOMMU without this patch (VMAs are no longer shared, and there are now region structs). Signed-off-by: David Howells <dhowells@redhat.com> Tested-by: Mike Frysinger <vapier.adi@gmail.com> Acked-by: Paul Mundt <lethal@linux-sh.org>
2009-01-08 12:04:47 +00:00
#else /* CONFIG_MMU */
/* under NOMMU conditions, the exact address to be destroyed must be
* given */
retval = -EINVAL;
if (vma->vm_start == addr && vma->vm_ops == &shm_vm_ops) {
do_munmap(mm, vma->vm_start, vma->vm_end - vma->vm_start);
retval = 0;
}
#endif
up_write(&mm->mmap_sem);
return retval;
}
#ifdef CONFIG_PROC_FS
static int sysvipc_shm_proc_show(struct seq_file *s, void *it)
{
struct shmid_kernel *shp = it;
#if BITS_PER_LONG <= 32
#define SIZE_SPEC "%10lu"
#else
#define SIZE_SPEC "%21lu"
#endif
return seq_printf(s,
"%10d %10d %4o " SIZE_SPEC " %5u %5u "
"%5lu %5u %5u %5u %5u %10lu %10lu %10lu\n",
shp->shm_perm.key,
shp->shm_perm.id,
shp->shm_perm.mode,
shp->shm_segsz,
shp->shm_cprid,
shp->shm_lprid,
[PATCH] shm: make sysv ipc shared memory use stacked files The current ipc shared memory code runs into several problems because it does not quite use files like the rest of the kernel. With the option of backing ipc shared memory with either hugetlbfs or ordinary shared memory the problems got worse. With the added support for ipc namespaces things behaved so unexpected that we now have several bad namespace reference counting bugs when using what appears at first glance to be a reasonable idiom. So to attack these problems and hopefully make the code more maintainable this patch simply uses the files provided by other parts of the kernel and builds it's own files out of them. The shm files are allocated in do_shmat and freed when their reference count drops to zero with their last unmap. The file and vm operations that we don't want to implement or we don't implement completely we just delegate to the operations of our backing file. This means that we now get an accurate shm_nattch count for we have a hugetlbfs inode for backing store, and the shm accounting of last attach and last detach time work as well. This means that getting a reference to the ipc namespace when we create the file and dropping the referenece in the release method is now safe and correct. This means we no longer need a special case for clearing VM_MAYWRITE as our file descriptor now only has write permissions when we have requested write access when calling shmat. Although VM_SHARED is now cleared as well which I believe is harmless and is mostly likely a minor bug fix. By using the same set of operations for both the hugetlb case and regular shared memory case shmdt is not simplified and made slightly more correct as now the test "vma->vm_ops == &shm_vm_ops" is 100% accurate in spotting all shared memory regions generated from sysvipc shared memory. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-20 21:57:53 +00:00
shp->shm_nattch,
shp->shm_perm.uid,
shp->shm_perm.gid,
shp->shm_perm.cuid,
shp->shm_perm.cgid,
shp->shm_atim,
shp->shm_dtim,
shp->shm_ctim);
}
#endif