linux.git/fs/eventpoll.c, branch v2.6.28

epoll: introduce resource usage limits

2008-12-02T03:55:24+00:00

It has been thought that the per-user file descriptors limit would also
limit the resources that a normal user can request via the epoll
interface.  Vegard Nossum reported a very simple program (a modified
version attached) that can make a normal user to request a pretty large
amount of kernel memory, well within the its maximum number of fds.  To
solve such problem, default limits are now imposed, and /proc based
configuration has been introduced.  A new directory has been created,
named /proc/sys/fs/epoll/ and inside there, there are two configuration
points:

  max_user_instances = Maximum number of devices - per user

  max_user_watches   = Maximum number of "watched" fds - per user

The current default for "max_user_watches" limits the memory used by epoll
to store "watches", to 1/32 of the amount of the low RAM.  As example, a
256MB 32bit machine, will have "max_user_watches" set to roughly 90000.
That should be enough to not break existing heavy epoll users.  The
default value for "max_user_instances" is set to 128, that should be
enough too.

This also changes the userspace, because a new error code can now come out
from EPOLL_CTL_ADD (-ENOSPC).  The EMFILE from epoll_create() was already
listed, so that should be ok.

[akpm@linux-foundation.org: use get_current_user()]
Signed-off-by: Davide Libenzi 
Cc: Michael Kerrisk 
Cc: 
Cc: Cyrill Gorcunov 
Reported-by: Vegard Nossum 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

epoll: avoid double-inserts in case of EFAULT

2008-10-26T19:09:49+00:00

In commit f337b9c58332bdecde965b436e47ea4c94d30da0 ("epoll: drop
unnecessary test") Thomas found that there is an unnecessary (always
true) test in ep_send_events().  The callback never inserts into
->rdllink while the send loop is performed, and also does the
~EP_PRIVATE_BITS test.  Given we're holding the mutex during this time,
the conditions tested inside the loop are always true.

HOWEVER.

The test "!ep_is_linked(&epi->rdllink)" wasn't there because we insert
into ->rdllink, but because the send-events loop might terminate before
the whole list is scanned (-EFAULT).

In such cases, when the loop terminates early, and when a (leftover)
file received an event while we're performing the lockless loop, we need
such test to avoid to double insert the epoll items.  The list_splice()
done a few steps below, will correctly re-insert the ones that were left
on "txlist".

This should fix the kenrel.org bugzilla entry 11831.

Signed-off-by: Davide Libenzi 
Signed-off-by: Linus Torvalds

epoll: drop unnecessary test

2008-10-16T18:21:32+00:00

Thomas found that there is an unnecessary (always true) test in
ep_send_events().  The callback never inserts into ->rdllink while the
send loop is performed, and also does the ~EP_PRIVATE_BITS test.  Given
we're holding the mutex during this time, the conditions tested inside the
loop are always true.  This patch drops the test done inside the
re-insertion loop.

Signed-off-by: Davide Libenzi 
Cc: Thomas Gleixner 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

fs/eventpoll.c: fix sys_epoll_create1() comment

2008-08-12T23:07:30+00:00

The `size' argument was removed.

Reported-by:  Geert Uytterhoeven 
Cc: Ulrich Drepper 
Cc: Davide Libenzi 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

flag parameters add-on: remove epoll_create size param

2008-07-24T17:47:29+00:00

Remove the size parameter from the new epoll_create syscall and renames the
syscall itself.  The updated test program follows.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
#include 
#include 
#include 
#include 
#include 

#ifndef __NR_epoll_create2
# ifdef __x86_64__
#  define __NR_epoll_create2 291
# elif defined __i386__
#  define __NR_epoll_create2 329
# else
#  error "need __NR_epoll_create2"
# endif
#endif

#define EPOLL_CLOEXEC O_CLOEXEC

int
main (void)
{
  int fd = syscall (__NR_epoll_create2, 0);
  if (fd == -1)
    {
      puts ("epoll_create2(0) failed");
      return 1;
    }
  int coe = fcntl (fd, F_GETFD);
  if (coe == -1)
    {
      puts ("fcntl failed");
      return 1;
    }
  if (coe & FD_CLOEXEC)
    {
      puts ("epoll_create2(0) set close-on-exec flag");
      return 1;
    }
  close (fd);

  fd = syscall (__NR_epoll_create2, EPOLL_CLOEXEC);
  if (fd == -1)
    {
      puts ("epoll_create2(EPOLL_CLOEXEC) failed");
      return 1;
    }
  coe = fcntl (fd, F_GETFD);
  if (coe == -1)
    {
      puts ("fcntl failed");
      return 1;
    }
  if ((coe & FD_CLOEXEC) == 0)
    {
      puts ("epoll_create2(EPOLL_CLOEXEC) set close-on-exec flag");
      return 1;
    }
  close (fd);

  puts ("OK");

  return 0;
}
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Signed-off-by: Ulrich Drepper 
Acked-by: Davide Libenzi 
Cc: Michael Kerrisk 
Cc: 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

flag parameters: check magic constants

2008-07-24T17:47:29+00:00

This patch adds test that ensure the boundary conditions for the various
constants introduced in the previous patches is met.  No code is generated.

[akpm@linux-foundation.org: fix alpha]
Signed-off-by: Ulrich Drepper 
Acked-by: Davide Libenzi 
Cc: Michael Kerrisk 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

flag parameters: epoll_create

2008-07-24T17:47:28+00:00

This patch adds the new epoll_create2 syscall.  It extends the old epoll_create
syscall by one parameter which is meant to hold a flag value.  In this
patch the only flag support is EPOLL_CLOEXEC which causes the close-on-exec
flag for the returned file descriptor to be set.

A new name EPOLL_CLOEXEC is introduced which in this implementation must
have the same value as O_CLOEXEC.

The following test must be adjusted for architectures other than x86 and
x86-64 and in case the syscall numbers changed.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
#include 
#include 
#include 
#include 
#include 

#ifndef __NR_epoll_create2
# ifdef __x86_64__
#  define __NR_epoll_create2 291
# elif defined __i386__
#  define __NR_epoll_create2 329
# else
#  error "need __NR_epoll_create2"
# endif
#endif

#define EPOLL_CLOEXEC O_CLOEXEC

int
main (void)
{
  int fd = syscall (__NR_epoll_create2, 1, 0);
  if (fd == -1)
    {
      puts ("epoll_create2(0) failed");
      return 1;
    }
  int coe = fcntl (fd, F_GETFD);
  if (coe == -1)
    {
      puts ("fcntl failed");
      return 1;
    }
  if (coe & FD_CLOEXEC)
    {
      puts ("epoll_create2(0) set close-on-exec flag");
      return 1;
    }
  close (fd);

  fd = syscall (__NR_epoll_create2, 1, EPOLL_CLOEXEC);
  if (fd == -1)
    {
      puts ("epoll_create2(EPOLL_CLOEXEC) failed");
      return 1;
    }
  coe = fcntl (fd, F_GETFD);
  if (coe == -1)
    {
      puts ("fcntl failed");
      return 1;
    }
  if ((coe & FD_CLOEXEC) == 0)
    {
      puts ("epoll_create2(EPOLL_CLOEXEC) set close-on-exec flag");
      return 1;
    }
  close (fd);

  puts ("OK");

  return 0;
}
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Signed-off-by: Ulrich Drepper 
Acked-by: Davide Libenzi 
Cc: Michael Kerrisk 
Cc: 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

flag parameters: anon_inode_getfd extension

2008-07-24T17:47:27+00:00

This patch just extends the anon_inode_getfd interface to take an additional
parameter with a flag value.  The flag value is passed on to
get_unused_fd_flags in anticipation for a use with the O_CLOEXEC flag.

No actual semantic changes here, the changed callers all pass 0 for now.

[akpm@linux-foundation.org: KVM fix]
Signed-off-by: Ulrich Drepper 
Acked-by: Davide Libenzi 
Cc: Michael Kerrisk 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

[PATCH] sanitize anon_inode_getfd()

2008-05-01T17:08:50+00:00

a) none of the callers even looks at inode or file returned by anon_inode_getfd()
b) any caller that would try to look at those would be racy, since by the time
it returns we might have raced with close() from another thread and that
file would be pining for fjords.

Signed-off-by: Al Viro

signals: use HAVE_SET_RESTORE_SIGMASK

2008-04-30T15:29:37+00:00

Change all the #ifdef TIF_RESTORE_SIGMASK conditionals in non-arch code to
#ifdef HAVE_SET_RESTORE_SIGMASK.  If arch code defines it first, the generic
set_restore_sigmask() using TIF_RESTORE_SIGMASK is not defined.

Signed-off-by: Roland McGrath 
Cc: Oleg Nesterov 
Cc: Ingo Molnar 
Cc: Thomas Gleixner 
Cc: Martin Schwidefsky 
Cc: Heiko Carstens 
Cc: "Luck, Tony" 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds