Hi,

I'm having a repeatable bug, namely that arbitrarily processes claim a CPU and hang
indefinitely. They recover sometimes, be it after 5 minutes or after 30 minutes, and
sometimes not.

Latest example was libtool-wrap while compiling the jpeg library:

# emerge media-libs/jpeg
....

# top
16973 root 39 19 10860 608 444 R 100 0.0 113:59.68 libtool-wrap

# while true; do cat /proc/16973/stat; sleep 1; done
16973 (libtool-wrap) R 15444 30375 6434 34816 30375 4194304 178 0 0 0 10604 1 0 0 39
19 1 0 6722416 11120640 152 18446744073709551615 4194304 4906172 140734787246416
18446744073709551615 47039760684325 0 0 16777216 0 0 0 0 17 3 0 0 0 0 0
16973 (libtool-wrap) R 15444 30375 6434 34816 30375 4194304 178 0 0 0 10704 1 0 0 39
19 1 0 6722416 11120640 152 18446744073709551615 4194304 4906172 140734787246416
18446744073709551615 47039760684325 0 0 16777216 0 0 0 0 17 3 0 0 0 0 0
16973 (libtool-wrap) R 15444 30375 6434 34816 30375 4194304 178 0 0 0 10805 1 0 0 39
19 1 0 6722416 11120640 152 18446744073709551615 4194304 4906172 140734787246416
18446744073709551615 47039760684325 0 0 16777216 0 0 0 0 17 3 0 0 0 0 0
....
16973 (libtool-wrap) R 15444 30375 6434 34816 30375 4194304 178 0 0 0 431667 65 0 0
39 19 1 0 6722416 11120640 152 18446744073709551615 4194304 4906172 140734787246416
18446744073709551615 47039760684325 0 0 16777216 0 0 0 0 17 3 0 0 0 0 0
16973 (libtool-wrap) R 15444 30375 6434 34816 30375 4194304 178 0 0 0 431767 65 0 0
39 19 1 0 6722416 11120640 152 18446744073709551615 4194304 4906172 140734787246416
18446744073709551615 47039760684325 0 0 16777216 0 0 0 0 17 3 0 0 0 0 0
16973 (libtool-wrap) R 15444 30375 6434 34816 30375 4194304 178 0 0 0 431868 65 0 0
39 19 1 0 6722416 11120640 152 18446744073709551615 4194304 4906172 140734787246416
18446744073709551615 47039760684325 0 0 16777216 0 0 0 0 17 3 0 0 0 0 0

I tried different kernels (gentoo-2.6.25, gentoo-2.6.26, git-2.6.27-rc8) and tried
all three preemption models without noticing any difference (except that on
git-2.6.27-rc8 processes seem never to recover). Processes which I noticed hanging
are: sed, awk, rm, touch, libtool-wrap, ... It's not happening instantly, but it
doesn't take long to force it to happen.

It's happening on:

# /usr/src/linux/scripts/ver_linux

Linux quadux 2.6.27-rc8-git1 #1 SMP Wed Oct 1 08:53:31 CEST 2008 x86_64 Intel(R)
Core(TM)2 Quad CPU Q6600 @ 2.40GHz GenuineIntel GNU/Linux

Gnu C 4.1.2
Gnu make 3.81
binutils 2.18
util-linux 2.13.1.1
mount 2.13.1.1
module-init-tools 3.4
e2fsprogs 1.40.9
reiserfsprogs 3.6.19
Linux C Library 2.6.1
Dynamic linker (ldd) 2.6.1
Procps 3.2.7
Net-tools 1.60
Kbd 1.13
Sh-utils 6.10
udev 124
Modules Loaded

I was *not* yet able to observe this behaviour on a x86_64 Intel(R) Core(TM)2 Duo CPU.

What can I do to get down to this issue?

Thanks,
Reto



--
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/