proc: fix a race in do_io_accounting()
authorVasiliy Kulikov <segoon@openwall.com>
Tue, 26 Jul 2011 23:08:38 +0000 (16:08 -0700)
committerLinus Torvalds <torvalds@linux-foundation.org>
Tue, 26 Jul 2011 23:49:43 +0000 (16:49 -0700)
commit293eb1e7772b25a93647c798c7b89bf26c2da2e0
treedebd49012773091732ee1e1b2388462b759525a7
parentd2857e79a2ba7c155eaa1a7d3581c8d26b31e54e
proc: fix a race in do_io_accounting()

If an inode's mode permits opening /proc/PID/io and the resulting file
descriptor is kept across execve() of a setuid or similar binary, the
ptrace_may_access() check tries to prevent using this fd against the
task with escalated privileges.

Unfortunately, there is a race in the check against execve().  If
execve() is processed after the ptrace check, but before the actual io
information gathering, io statistics will be gathered from the
privileged process.  At least in theory this might lead to gathering
sensible information (like ssh/ftp password length) that wouldn't be
available otherwise.

Holding task->signal->cred_guard_mutex while gathering the io
information should protect against the race.

The order of locking is similar to the one inside of ptrace_attach():
first goes cred_guard_mutex, then lock_task_sighand().

Signed-off-by: Vasiliy Kulikov <segoon@openwall.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: <stable@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
fs/proc/base.c
This page took 0.026022 seconds and 5 git commands to generate.