Re: [PATCH net-next] af_unix: dont send SCM_CREDENTIALS by default

On Fri, 2011-09-09 at 07:06 +0200, Eric Dumazet wrote:

Le jeudi 08 septembre 2011 à 02:24 -0700, Tim Chen a écrit :

quoted

Looking forward to the patch.  This should improve the scalability of
af_unix.

Here it is, based on top on previous one [af_unix: Fix use-after-free
crashes]

Thanks

[PATCH net-next] af_unix: dont send SCM_CREDENTIALS by default

Since commit 7361c36c5224 (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.

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

4.224 sec instead of 9.102 sec


Signed-off-by: Eric Dumazet <redacted>

This speeds up the case where we're not passing SCM_CREDENTIALS nicely.

Acked-by: Tim Chen <redacted>