lkml.org 
[lkml]   [2004]   [May]   [2]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
Date
From
SubjectRe: 2.6.6-rc3-mm1: some sort of deadlock occurs under heavy i/o
On Sun, 2 May 2004, Jurriaan wrote:

> When I do a _lot_ of I/O in 2.6.6-rc3-mm1, I've seen some sort of
> deadlock multiple times now. pinging from outside still works, num-lock
> turns the led on/off, I can switch consoles but otherwise, nothing
> works. The lot of I/O is for example: expiring two large newsspools, and
> reading a large folder in mutt. This worked fine in 2.6.6-rc2-mm2 and
> kernels before that (except that pdflush sometimes started using 100%
> cpu. I've not seen that this time, but this is worse, unfortunately).
>
> This is a single P4, using hyperthreading. In use are reiserfs and ext3,
> raid-1 and linear raid. Alt-sysreq-t produced a large list of processes,
> where the last 3 each had:

Try the following patch from Andrew;

Index: linux-2.6.6-rc3-mm1/mm/vmscan.c
===================================================================
RCS file: /home/cvsroot/linux-2.6.6-rc3-mm1/mm/vmscan.c,v
retrieving revision 1.1.1.1
diff -u -p -B -r1.1.1.1 vmscan.c
--- linux-2.6.6-rc3-mm1/mm/vmscan.c 1 May 2004 08:19:20 -0000 1.1.1.1
+++ linux-2.6.6-rc3-mm1/mm/vmscan.c 1 May 2004 21:39:36 -0000
@@ -156,7 +156,7 @@ static int shrink_slab(unsigned long sca
shrinker->nr = LONG_MAX; /* It wrapped! */

if (shrinker->nr <= SHRINK_BATCH)
- break;
+ continue;
while (shrinker->nr) {
long this_scan = shrinker->nr;
int shrink_ret;
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
\
 
 \ /
  Last update: 2005-03-22 14:02    [W:0.028 / U:0.492 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site