Skip to content

GitLab

Switch branch/tag
  1. 15 Jan, 2016 1 commit
    • Takashi Iwai's avatar
      ALSA: timer: Harden slave timer list handling · b5a663aa
      Takashi Iwai authored 
      
A slave timer instance might be still accessible in a racy way while
operating the master instance as it lacks of locking.  Since the
master operation is mostly protected with timer->lock, we should cope
with it while changing the slave instance, too.  Also, some linked
lists (active_list and ack_list) of slave instances aren't unlinked
immediately at stopping or closing, and this may lead to unexpected
accesses.

This patch tries to address these issues.  It adds spin lock of
timer->lock (either from master or slave, which is equivalent) in a
few places.  For avoiding a deadlock, we ensure that the global
slave_active_lock is always locked at first before each timer lock.

Also, ack and active_list of slave instances are properly unlinked at
snd_timer_stop() and snd_timer_close().

Last but not least, remove the superfluous call of _snd_timer_stop()
at removing slave links.  This is a noop, and calling it may confuse
readers wrt locking.  Further cleanup will follow in a later patch.

Actually we've got reports of use-after-free by syzkaller fuzzer, and
this hopefully fixes these issues.
Reported-by: default avatarDmitry Vyukov <dvyukov@google.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: default avatarTakashi Iwai <tiwai@suse.de>
      b5a663aa
  3. 14 Jan, 2016 1 commit
  5. 13 Jan, 2016 1 commit
  7. 27 May, 2015 1 commit
  9. 10 Mar, 2015 1 commit
  11. 02 Feb, 2015 3 commits
  13. 19 Jan, 2015 1 commit
  15. 12 Jun, 2014 1 commit
  17. 14 Feb, 2014 1 commit
  19. 10 Feb, 2014 1 commit
    • Takashi Iwai's avatar
      ALSA: Replace with IS_ENABLED() · 8eeaa2f9
      Takashi Iwai authored 
      
Replace the lengthy #if defined(XXX) || defined(XXX_MODULE) with the
new IS_ENABLED() macro.

The patch still doesn't cover all ifdefs.  For example, the dependency
on CONFIG_GAMEPORT is still open-coded because this also has an extra
dependency on MODULE.  Similarly, an open-coded ifdef in pcm_oss.c and
some sequencer-related stuff are left untouched.
Signed-off-by: default avatarTakashi Iwai <tiwai@suse.de>
      8eeaa2f9
  21. 11 Mar, 2012 1 commit
    • Paul Gortmaker's avatar
      device.h: cleanup users outside of linux/include (C files) · 51990e82
      Paul Gortmaker authored 
      
For files that are actively using linux/device.h, make sure
that they call it out.  This will allow us to clean up some
of the implicit uses of linux/device.h within include/*
without introducing build regressions.

Yes, this was created by "cheating" -- i.e. the headers were
cleaned up, and then the fallout was found and fixed, and then
the two commits were reordered.  This ensures we don't introduce
build regressions into the git history.
Signed-off-by: default avatarPaul Gortmaker <paul.gortmaker@windriver.com>
      51990e82
  23. 31 Oct, 2011 1 commit
  25. 08 Aug, 2011 2 commits
  27. 16 Mar, 2011 1 commit
  29. 24 Nov, 2010 1 commit
    • Kay Sievers's avatar
      ALSA: support module on-demand loading for seq and timer · 03cfe6f5
      Kay Sievers authored 
      
If CONFIG_SND_DYNAMIC_MINORS is used, assign /dev/snd/seq and
/dev/snd/timer the usual static minors, and export specific
module aliases to generate udev module on-demand loading
instructions:

  $ cat /lib/modules/2.6.33.4-smp/modules.devname
  # Device nodes to trigger on-demand module loading.
  microcode cpu/microcode c10:184
  fuse fuse c10:229
  ppp_generic ppp c108:0
  tun net/tun c10:200
  uinput uinput c10:223
  dm_mod mapper/control c10:236
  snd_timer snd/timer c116:33
  snd_seq snd/seq c116:1

The last two lines instruct udev to create device nodes, even
when the modules are not loaded at that time.

As soon as userspace accesses any of these nodes, the in-kernel
module-loader will load the module, and the device can be used.

The header file minor calculation needed to be simplified to
make __stringify() (supports only two indirections) in
the MODULE_ALIAS macro work.

This is part of systemd's effort to get rid of unconditional
module load instructions and needless init scripts.

Cc: Lennart Poettering <lennart@poettering.net>
Signed-off-by: default avatarKay Sievers <kay.sievers@vrfy.org>
Signed-off-by: default avatarClemens Ladisch <clemens@ladisch.de>
Signed-off-by: default avatarTakashi Iwai <tiwai@suse.de>
      03cfe6f5
  31. 22 Nov, 2010 1 commit
  33. 05 May, 2010 1 commit
    • Dan Carpenter's avatar
      ALSA: take tu->qlock with irqs disabled · bfe70783
      Dan Carpenter authored 
      
We should disable irqs when we take the tu->qlock because it is used in
the irq handler.  The only place that doesn't is
snd_timer_user_ccallback().  Most of the time snd_timer_user_ccallback()
is called with interrupts disabled but the the first ti->ccallback()
call in snd_timer_notify1() has interrupts enabled.

This was caught by lockdep which generates the following message:

> =================================
> [ INFO: inconsistent lock state ]
> 2.6.34-rc5 #5
> ---------------------------------
> inconsistent {HARDIRQ-ON-W} -> {IN-HARDIRQ-W} usage.
> dolphin/4003 [HC1[1]:SC0[0]:HE0:SE1] takes:
> (&(&tu->qlock)->rlock){?.+...}, at: [<f84ec472>] snd_timer_user_tinterrupt+0x28/0x132 [snd_timer]
> {HARDIRQ-ON-W} state was registered at:
>   [<c1048de9>] __lock_acquire+0x654/0x1482
>   [<c1049c73>] lock_acquire+0x5c/0x73
>   [<c125ac3e>] _raw_spin_lock+0x25/0x34
>   [<f84ec370>] snd_timer_user_ccallback+0x55/0x95 [snd_timer]
>   [<f84ecc4b>] snd_timer_notify1+0x53/0xca [snd_timer]
Reported-by: default avatarStefan Richter <stefanr@s5r6.in-berlin.de>
Signed-off-by: default avatarDan Carpenter <error27@gmail.com>
Signed-off-by: default avatarTakashi Iwai <tiwai@suse.de>
      bfe70783
  35. 13 Apr, 2010 1 commit
    • Takashi Iwai's avatar
      ALSA: core - Define llseek fops · 02f4865f
      Takashi Iwai authored 
      
Set no_llseek to llseek file ops of each sound component (but for hwdep).
This avoids the implicit BKL invocation via generic_file_llseek() used
as default when fops.llseek is NULL.

Also call nonseekable_open() at each open ops to ensure the file flags
have no seek bit.
Signed-off-by: default avatarTakashi Iwai <tiwai@suse.de>
      02f4865f
  37. 03 Mar, 2010 1 commit
  39. 14 Apr, 2009 1 commit
  41. 16 Mar, 2009 1 commit
    • Jonathan Corbet's avatar
      Rationalize fasync return values · 60aa4924
      Jonathan Corbet authored 
      
Most fasync implementations do something like:

     return fasync_helper(...);

But fasync_helper() will return a positive value at times - a feature used
in at least one place.  Thus, a number of other drivers do:

     err = fasync_helper(...);
     if (err < 0)
             return err;
     return 0;

In the interests of consistency and more concise code, it makes sense to
map positive return values onto zero where ->fasync() is called.

Cc: Al Viro <viro@ZenIV.linux.org.uk>
Signed-off-by: default avatarJonathan Corbet <corbet@lwn.net>
      60aa4924
  43. 18 Dec, 2008 1 commit
  45. 01 Nov, 2008 1 commit
    • Al Viro's avatar
      saner FASYNC handling on file close · 233e70f4
      Al Viro authored 
      
As it is, all instances of ->release() for files that have ->fasync()
need to remember to evict file from fasync lists; forgetting that
creates a hole and we actually have a bunch that *does* forget.

So let's keep our lives simple - let __fput() check FASYNC in
file->f_flags and call ->fasync() there if it's been set.  And lose that
crap in ->release() instances - leaving it there is still valid, but we
don't have to bother anymore.
Signed-off-by: default avatarAl Viro <viro@zeniv.linux.org.uk>
Signed-off-by: default avatarLinus Torvalds <torvalds@linux-foundation.org>
      233e70f4
  47. 15 Aug, 2008 1 commit
  49. 13 Aug, 2008 1 commit
  51. 10 Jul, 2008 1 commit
  53. 15 Feb, 2008 1 commit
  55. 31 Jan, 2008 2 commits
  57. 16 Oct, 2007 1 commit
  59. 20 Jul, 2007 1 commit
  61. 08 May, 2007 1 commit
  63. 12 Feb, 2007 1 commit
  65. 09 Feb, 2007 2 commits
  67. 23 Sep, 2006 2 commits