Thread (11 messages) 11 messages, 6 authors, 2015-05-14

Re: [PATCH][RT] xfs: Disable preemption when grabbing all icsb counter locks

From: Dave Chinner <david@fromorbit.com>
Date: 2015-05-04 00:51:53
Also in: lkml

On Thu, Apr 30, 2015 at 12:33:03PM -0400, Steven Rostedt wrote:
Running a test on a large CPU count box with xfs, I hit a live lock
with the following backtraces on several CPUs:

 Call Trace:              
  [<ffffffff812c34f8>] __const_udelay+0x28/0x30 
  [<ffffffffa033ab9a>] xfs_icsb_lock_cntr+0x2a/0x40 [xfs] 
  [<ffffffffa033c871>] xfs_icsb_modify_counters+0x71/0x280 [xfs] 
  [<ffffffffa03413e1>] xfs_trans_reserve+0x171/0x210 [xfs] 
  [<ffffffffa0378cfd>] xfs_create+0x24d/0x6f0 [xfs] 
  [<ffffffff8124c8eb>] ? avc_has_perm_flags+0xfb/0x1e0 
  [<ffffffffa0336eeb>] xfs_vn_mknod+0xbb/0x1e0 [xfs] 
  [<ffffffffa0337043>] xfs_vn_create+0x13/0x20 [xfs] 
  [<ffffffff811b0edd>] vfs_create+0xcd/0x130 
  [<ffffffff811b21ef>] do_last+0xb8f/0x1240 
  [<ffffffff811b39b2>] path_openat+0xc2/0x490 

Looking at the code I see it was stuck at: 

STATIC void
xfs_icsb_lock_cntr(
	xfs_icsb_cnts_t	*icsbp)
{
	while (test_and_set_bit(XFS_ICSB_FLAG_LOCK, &icsbp->icsb_flags)) {
		ndelay(1000);
	}
}

I'm not sure why it does the ndelay() and not just a cpu_relax(), but
Because the code was writtenlong before cpu_relax() existed, just
like it was written long before the generic percpu counter code was
added...

....
Now, when PREEMPT_RT is not enabled, that spin_lock() disables
preemption. But for PREEMPT_RT, it does not. Although with my test box I
was not able to produce a task state of all tasks, but I'm assuming that
some task called the xfs_icsb_lock_all_counters() and was preempted by
an RT task and could not finish, causing all callers of that lock to
block indefinitely.

Looking at all users of xfs_icsb_lock_all_counters(), they are leaf
functions and do not call anything that may block on PREEMPT_RT. I
believe the proper fix here is to simply disable preemption in
xfs_icsb_lock_all_counters() when PREEMPT_RT is enabled.
RT is going to have other performance problems that are probably
going to negate the scalability this code provides. If you want a
hack that you can easily backport (as this code now uses the generic
percpu counters) then have a look at fs/xfs/xfs_linux.h:

/*
 * Feature macros (disable/enable)
 */
#ifdef CONFIG_SMP
#define HAVE_PERCPU_SB  /* per cpu superblock counters are a 2.6 feature */
#else
#undef  HAVE_PERCPU_SB  /* per cpu superblock counters are a 2.6 feature */
#endif

You can turn off all that per-cpu code simply by:

-#ifdef CONFIG_SMP
+#if defined(CONFIG_SMP) && !defined(CONFIG_PREEMPT_RT)

Cheers,

Dave.
-- 
Dave Chinner
david@fromorbit.com
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help