lkml.org 
[lkml]   [2002]   [Apr]   [30]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectLOCKUP: 2.4.18 + crtypo pathc2 locking raid/loop/ext3 wait_o under stress test


/ is ext3 over /dev/md/0 (raid1)
swap is /dev/loop/0 with blowfish over /dev/md/1 (raid0)
/home is ext3 over /dev/loop/1 with blowfish over /dev/md/2 (raid1)

256 ram. duron, hpt370,
raw IDE speed 40MB/s,
raw raid1 speed 37MB/s, raid0 speed near 80.
raw crypto speed 20MB/s (cpu full)

stress test is
1. a.out swap stress seeking 512M memory, while (1) for (i =0; i < 256M ;
i+=256) intarray[i]=i, taking 1-3% processor, all io
2. dd on /home of 512MB chunks , write,read, taking sometimes all
processor all IO, if swap takes io, dd is low
3. while true; do ls /home; done ; takes all the rest of IO.
4. ping flood outside computer


any following ls anywhere does not block.
sync blocks. It is somewhat weird as one ls is blocking.

process states after several killall -9 a.out dd ls

F UID PID PPID PRI NI VSZ RSS WCHAN STAT TTY TIME COMMAND
100 0 1 0 9 0 1412 408 blowfi S ? 0:06 init
040 0 2 1 9 0 0 0 contex SW ? 0:00 [keventd]
040 0 3 0 19 19 0 0 ksofti SWN ? 0:00 [ksoftirqd_
040 0 4 0 9 0 0 0 wait_o DW ? 0:49 [kswapd]
040 0 5 0 9 0 0 0 bdflus SW ? 0:00 [bdflush]
040 0 6 0 9 0 0 0 wait_o DW ? 0:00 [kupdated]
040 0 7 1 9 0 0 0 down_i SW ? 0:00 [i2oevtd]
040 0 8 1 -1 -20 0 0 md_thr SW< ? 0:00 [mdrecovery
040 0 9 1 -1 -20 0 0 md_thr SW< ? 0:29 [raid1d]
040 0 10 1 19 19 0 0 md_thr SWN ? 0:09 [raid1syncd
040 0 11 1 -1 -20 0 0 md_thr SW< ? 0:01 [raid1d]
040 0 12 1 19 19 0 0 md_thr SWN ? 0:00 [raid1syncd
040 0 13 1 9 0 0 0 kjourn SW ? 0:02 [kjournald]
040 0 14 0 9 0 0 0 blowfi SW ? 0:00 [klids]
040 0 54 1 -1 -20 0 0 down_i SW< ? 0:35 [loop0]
040 0 376 1 9 0 0 0 rtl813 SW ? 0:00 [eth0]
040 0 452 1 9 0 1472 476 do_sel S ? 0:00 syslogd -m
140 0 457 1 9 0 2092 456 do_sys S ? 0:00 klogd -2
040 0 538 1 9 0 0 0 kjourn SW ? 0:00 [kjournald]
040 0 539 1 -1 -20 0 0 down_i SW< ? 0:00 [loop1]
040 0 541 1 9 0 0 0 wait_o DW ? 0:05 [kjournald]
040 0 592 1 7 0 1584 448 nanosl S ? 0:00 crond
040 2 628 1 9 0 1444 392 nanosl S ? 0:00 /usr/sbin/a
100 0 635 1 9 0 2320 412 wait4 S vc/1 0:00 login -- ro
100 0 636 1 9 0 2320 412 wait4 S vc/2 0:00 login -- ro
100 0 637 1 9 0 2320 428 wait4 S vc/3 0:00 login -- ro
100 0 638 1 9 0 2320 648 wait4 S vc/4 0:00 login -- ro
100 0 639 1 9 0 2320 668 wait4 S vc/5 0:00 login -- ro
100 0 640 1 9 0 2320 668 wait4 S vc/6 0:00 login -- ro
100 0 643 635 15 0 2540 1000 read_c S vc/1 0:04 -bash
100 0 695 636 9 0 2516 596 wait4 S vc/2 0:00 -bash
100 0 902 695 9 0 2364 412 wait4 S vc/2 0:00 su - elmer
100 500 903 902 9 0 2452 600 wait4 S vc/2 0:00 -bash
100 500 2166 903 9 0 2876 652 do_sel S vc/2 0:00 ssh 10.110.
100 0 2243 637 9 0 2508 640 wait4 S vc/3 0:00 -bash
000 0 2290 2243 9 0 501372 152928 wait_o D vc/3 0:49 ./a.out .
100 0 2310 638 9 0 2508 960 wait4 S vc/4 0:00 -bash
040 0 2359 2310 9 0 2508 960 wait4 S vc/4 0:00 -bash
000 0 2362 2359 9 0 525952 456 wait_o D vc/4 0:48 dd if /dev/
100 0 2365 639 9 0 2500 1124 wait4 S vc/5 0:00 -bash
100 0 2412 640 9 0 2500 1060 wait4 S vc/6 0:07 -bash
000 0 6444 2412 9 0 1708 592 wait_o D vc/6 0:00 ls --color=
000 0 6445 2365 19 0 2212 1120 do_sel S vc/5 0:02 top
000 0 6473 643 9 0 1632 488 wait_o D vc/1 0:01 sync
100 0 6506 643 17 0 3136 1200 - R vc/1 0:00 ps axl

-
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 13:25    [W:0.032 / U:0.864 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site