Sparc t2000 hangs under stress load
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
linux (Ubuntu) |
Invalid
|
Undecided
|
Unassigned | ||
linux-source-2.6.17 (Ubuntu) |
Invalid
|
Undecided
|
Unassigned |
Bug Description
Binary package hint: linux-source-2.6.17
Stress testing t2000 generates reproducable system hangs.
1. uname -a
Linux wgs94-206 2.6.17-
2. dmesg of sloft lockup
[36674.762366] BUG: soft lockup detected on CPU#4!
[36674.900712] Call Trace:
[36674.900737] [000000000043411c] smp_percpu_
[36674.900785] [00000000004109d4] tl0_irq14+0x14/0x20
[36674.900815] [0000000000494590] anon_vma_
[36674.900858] [000000000045150c] copy_process+
[36674.900890] [0000000000452060] do_fork+0x40/0x220
[36674.900915] [0000000000406c94] linux_sparc_
[36674.900954] [00000000000029f4] 0x29f4
3. ps in an uninterruptable sleep state
ps in an uninterruptable sleep state
S 1001 572 17225 0 78 0 1080 415 compat pts/2 00:00:15 WatchDog_foS
D 1001 10783 572 0 77 0 1320 408 access pts/2 00:00:00 ps
S 1001 10784 572 0 78 0 1072 409 pipe_w pts/2 00:00:00 grep
S 1001 10785 572 0 76 0 768 408 pipe_w pts/2 00:00:00 awk
4. Steps to reproduce
Please contact submitter
5. Changes to system
Included patches available (apt-get update etc.) 3/8/2007
Removed crashme from stress load.
6. Our planned next steps
a. build from ubuntu sources and check hang reproducable
b. check gettimeofday() source code investigating memory barrier instruction placements
thanks ted
Hi Ted, thanks for the 2 bug reports (I saw the other one too). It would be interesting for us to know if you can reproduce these problems with dapper and feisty too.
Also.. what kind of stress test did you perform in this setup? the other was crashme, here?
Thanks
Fabio