lkml.org 
[lkml]   [2020]   [Nov]   [13]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
Date
From
SubjectRe: [PATCH] kfence: Avoid stalling work queue task without allocations
On Thu, Nov 12, 2020 at 09:54:06AM -0800, Paul E. McKenney wrote:
> On Thu, Nov 12, 2020 at 05:14:39PM +0100, Marco Elver wrote:

[ . . . ]

> > | [ 334.160218] BUG: workqueue lockup - pool cpus=0 node=0 flags=0x0 nice=0 stuck for 15s!
>
> It might be instructive to cause this code to provoke a backtrace.
> I suggest adding something like "trigger_single_cpu_backtrace(cpu)"
> in kernel/workqueue.c's function named wq_watchdog_timer_fn()
> somewhere within its "if" statement that is preceded with the "did we
> stall?" comment. Or just search for "BUG: workqueue lockup - pool"
> within kernel/workqueue.c.

And I did get a small but unexpected gift of time, so here is an
(untested) patch.

Thanx, Paul

------------------------------------------------------------------------

diff --git a/kernel/workqueue.c b/kernel/workqueue.c
index 437935e..f3d4ff7 100644
--- a/kernel/workqueue.c
+++ b/kernel/workqueue.c
@@ -5792,6 +5792,7 @@ static void wq_watchdog_timer_fn(struct timer_list *unused)
pr_cont_pool_info(pool);
pr_cont(" stuck for %us!\n",
jiffies_to_msecs(jiffies - pool_ts) / 1000);
+ trigger_single_cpu_backtrace(cpu);
}
}

\
 
 \ /
  Last update: 2020-11-13 18:59    [W:0.122 / U:3.332 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site