Since commit 7361c36c52 (af_unix: Allow credentials to work across
user and pid namespaces) af_unix performance dropped a lot.
This is because we now take a reference on pid and cred in each write(),
and release them in read(), usually done from another process,
eventually from another cpu. This triggers false sharing.
# Events: 154K cycles
#
# Overhead Command Shared Object Symbol
# ........ ....... .................. .........................
#
10.40% hackbench [kernel.kallsyms] [k] put_pid
8.60% hackbench [kernel.kallsyms] [k] unix_stream_recvmsg
7.87% hackbench [kernel.kallsyms] [k] unix_stream_sendmsg
6.11% hackbench [kernel.kallsyms] [k] do_raw_spin_lock
4.95% hackbench [kernel.kallsyms] [k] unix_scm_to_skb
4.87% hackbench [kernel.kallsyms] [k] pid_nr_ns
4.34% hackbench [kernel.kallsyms] [k] cred_to_ucred
2.39% hackbench [kernel.kallsyms] [k] unix_destruct_scm
2.24% hackbench [kernel.kallsyms] [k] sub_preempt_count
1.75% hackbench [kernel.kallsyms] [k] fget_light
1.51% hackbench [kernel.kallsyms] [k]
__mutex_lock_interruptible_slowpath
1.42% hackbench [kernel.kallsyms] [k] sock_alloc_send_pskb
This patch includes SCM_CREDENTIALS information in a af_unix message/skb
only if requested by the sender, [man 7 unix for details how to include
ancillary data using sendmsg() system call]
Note: This might break buggy applications that expected SCM_CREDENTIAL
from an unaware write() system call, and receiver not using SO_PASSCRED
socket option.
If SOCK_PASSCRED is set on source or destination socket, we still
include credentials for mere write() syscalls.
Performance boost in hackbench : more than 50% gain on a 16 thread
machine (2 quad-core cpus, 2 threads per core)
hackbench 20 thread 2000
4.228 sec instead of 9.102 sec
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Acked-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This reverts commit 0856a30409.
As requested by Eric Dumazet, it has various ref-counting
problems and has introduced regressions. Eric will add
a more suitable version of this performance fix.
Signed-off-by: David S. Miller <davem@davemloft.net>
Patch series 109f6e39..7361c36c back in 2.6.36 added functionality to
allow credentials to work across pid namespaces for packets sent via
UNIX sockets. However, the atomic reference counts on pid and
credentials caused plenty of cache bouncing when there are numerous
threads of the same pid sharing a UNIX socket. This patch mitigates the
problem by eliminating extraneous reference counts on pid and
credentials on both send and receive path of UNIX sockets. I found a 2x
improvement in hackbench's threaded case.
On the receive path in unix_dgram_recvmsg, currently there is an
increment of reference count on pid and credentials in scm_set_cred.
Then there are two decrement of the reference counts. Once in scm_recv
and once when skb_free_datagram call skb->destructor function
unix_destruct_scm. One pair of increment and decrement of ref count on
pid and credentials can be eliminated from the receive path. Until we
destroy the skb, we already set a reference when we created the skb on
the send side.
On the send path, there are two increments of ref count on pid and
credentials, once in scm_send and once in unix_scm_to_skb. Then there
is a decrement of the reference counts in scm_destroy's call to
scm_destroy_cred at the end of unix_dgram_sendmsg functions. One pair
of increment and decrement of the reference counts can be removed so we
only need to increment the ref counts once.
By incorporating these changes, for hackbench running on a 4 socket
NHM-EX machine with 40 cores, the execution of hackbench on
50 groups of 20 threads sped up by factor of 2.
Hackbench command used for testing:
./hackbench 50 thread 2000
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Lower SCM_MAX_FD from 255 to 253 so that allocations for scm_fp_list are
halved. (commit f8d570a4 added two pointers in this structure)
scm_fp_dup() should not copy whole structure (and trigger kmemcheck
warnings), but only the used part. While we are at it, only allocate
needed size.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Start capturing not only the userspace pid, uid and gid values of the
sending process but also the struct pid and struct cred of the sending
process as well.
This is in preparation for properly supporting SCM_CREDENTIALS for
sockets that have different uid and/or pid namespaces at the different
ends.
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Acked-by: Serge E. Hallyn <serge@hallyn.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Reorder the fields in scm_cookie so they pack better on 64bit.
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Acked-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
This cleanup patch puts struct/union/enum opening braces,
in first line to ease grep games.
struct something
{
becomes :
struct something {
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch removes an unused member (seq) scm_cookie; besides initialized
to 0 in the header file, it is not used.
Signed-off-by: Rami Rosen <ramirose@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Conflicts:
security/keys/internal.h
security/keys/process_keys.c
security/keys/request_key.c
Fixed conflicts above by using the non 'tsk' versions.
Signed-off-by: James Morris <jmorris@namei.org>
Wrap access to task credentials so that they can be separated more easily from
the task_struct during the introduction of COW creds.
Change most current->(|e|s|fs)[ug]id to current_(|e|s|fs)[ug]id().
Change some task->e?[ug]id to task_e?[ug]id(). In some places it makes more
sense to use RCU directly rather than a convenient wrapper; these will be
addressed by later patches.
Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: James Morris <jmorris@namei.org>
Acked-by: Serge Hallyn <serue@us.ibm.com>
Cc: netdev@vger.kernel.org
Signed-off-by: James Morris <jmorris@namei.org>
__scm_destroy() walks the list of file descriptors in the scm_fp_list
pointed to by the scm_cookie argument.
Those, in turn, can close sockets and invoke __scm_destroy() again.
There is nothing which limits how deeply this can occur.
The idea for how to fix this is from Linus. Basically, we do all of
the fput()s at the top level by collecting all of the scm_fp_list
objects hit by an fput(). Inside of the initial __scm_destroy() we
keep running the list until it is empty.
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This is the largest patch in the set. Make all (I hope) the places where
the pid is shown to or get from user operate on the virtual pids.
The idea is:
- all in-kernel data structures must store either struct pid itself
or the pid's global nr, obtained with pid_nr() call;
- when seeking the task from kernel code with the stored id one
should use find_task_by_pid() call that works with global pids;
- when showing pid's numerical value to the user the virtual one
should be used, but however when one shows task's pid outside this
task's namespace the global one is to be used;
- when getting the pid from userspace one need to consider this as
the virtual one and use appropriate task/pid-searching functions.
[akpm@linux-foundation.org: build fix]
[akpm@linux-foundation.org: nuther build fix]
[akpm@linux-foundation.org: yet nuther build fix]
[akpm@linux-foundation.org: remove unneeded casts]
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: Alexey Dobriyan <adobriyan@openvz.org>
Cc: Sukadev Bhattiprolu <sukadev@us.ibm.com>
Cc: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Paul Menage <menage@google.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The OPEN_MAX constant is an arbitrary number with no useful relation to
anything. Nothing should be using it. SCM_MAX_FD is just an arbitrary
constant and it should be clear that its value is chosen in net/scm.h
and not actually derived from anything else meaningful in the system.
Signed-off-by: Roland McGrath <roland@redhat.com>
Cc: "David S. Miller" <davem@davemloft.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
From: Catherine Zhang <cxzhang@watson.ibm.com>
This patch implements a cleaner fix for the memory leak problem of the
original unix datagram getpeersec patch. Instead of creating a
security context each time a unix datagram is sent, we only create the
security context when the receiver requests it.
This new design requires modification of the current
unix_getsecpeer_dgram LSM hook and addition of two new hooks, namely,
secid_to_secctx and release_secctx. The former retrieves the security
context and the latter releases it. A hook is required for releasing
the security context because it is up to the security module to decide
how that's done. In the case of Selinux, it's a simple kfree
operation.
Acked-by: Stephen Smalley <sds@tycho.nsa.gov>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch implements an API whereby an application can determine the
label of its peer's Unix datagram sockets via the auxiliary data mechanism of
recvmsg.
Patch purpose:
This patch enables a security-aware application to retrieve the
security context of the peer of a Unix datagram socket. The application
can then use this security context to determine the security context for
processing on behalf of the peer who sent the packet.
Patch design and implementation:
The design and implementation is very similar to the UDP case for INET
sockets. Basically we build upon the existing Unix domain socket API for
retrieving user credentials. Linux offers the API for obtaining user
credentials via ancillary messages (i.e., out of band/control messages
that are bundled together with a normal message). To retrieve the security
context, the application first indicates to the kernel such desire by
setting the SO_PASSSEC option via getsockopt. Then the application
retrieves the security context using the auxiliary data mechanism.
An example server application for Unix datagram socket should look like this:
toggle = 1;
toggle_len = sizeof(toggle);
setsockopt(sockfd, SOL_SOCKET, SO_PASSSEC, &toggle, &toggle_len);
recvmsg(sockfd, &msg_hdr, 0);
if (msg_hdr.msg_controllen > sizeof(struct cmsghdr)) {
cmsg_hdr = CMSG_FIRSTHDR(&msg_hdr);
if (cmsg_hdr->cmsg_len <= CMSG_LEN(sizeof(scontext)) &&
cmsg_hdr->cmsg_level == SOL_SOCKET &&
cmsg_hdr->cmsg_type == SCM_SECURITY) {
memcpy(&scontext, CMSG_DATA(cmsg_hdr), sizeof(scontext));
}
}
sock_setsockopt is enhanced with a new socket option SOCK_PASSSEC to allow
a server socket to receive security context of the peer.
Testing:
We have tested the patch by setting up Unix datagram client and server
applications. We verified that the server can retrieve the security context
using the auxiliary data mechanism of recvmsg.
Signed-off-by: Catherine Zhang <cxzhang@watson.ibm.com>
Acked-by: Acked-by: James Morris <jmorris@namei.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Instead of doing a memset then initialization of the fields of the scm
structure, just initialize all the members explicitly. Prevent reloading
of current on x86 and x86-64 by storing the value in a local variable for
subsequent dereferences. This is worth a ~7KB/s increase in af_unix
bandwidth. Note that we avoid the issues surrounding potentially
uninitialized members of the ucred structure by constructing a struct
ucred instead of assigning the members individually, which forces the
compiler to zero any padding.
[ I modified the patch not to use the aggregate assignment since
gcc-3.4.x and earlier cannot optimize that properly at all even
though gcc-4.0.x and later can -DaveM ]
Signed-off-by: Benjamin LaHaise <benjamin.c.lahaise@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Initial git repository build. I'm not bothering with the full history,
even though we have it. We can create a separate "historical" git
archive of that later if we want to, and in the meantime it's about
3.2GB when imported into git - space that would just make the early
git days unnecessarily complicated, when we don't have a lot of good
infrastructure for it.
Let it rip!
