[BUG] 2.6.26-rc1-git4 - task blocked on powerpc for more than 120 seconds

[Nadia Derbey]

Kamalesh Babulal wrote:
> Adrian Bunk wrote:
> 
>>On Wed, May 07, 2008 at 06:52:53PM +0530, Kamalesh Babulal wrote:
>>
>>>While running the ltp over the powerpc with the 2.6.26-rc1-git4 kernel,
>>>task get blocked for more 120 seconds and does not proceeds future.
>>>
>>>The task msgctl08 is a ipc testcase, which test the msgget(2) msgctl(2)
>>>syscalls.
>>
>>Thanks for your report.
>>
>>I assume this is reproducible?
>>
>>If yes, what was the last kernel that worked for you?
> 
> 
> This is reproducible on 2.6.26-rc1 and 2.6.26-rc1-git(s). 2.6.25 is the last know kernel
> to work.
> 
> 
>>I've also added Pierre Peiffer and Nadia Derbey to the Cc since their 
>>recent ipc commits are my first suspects.
>>
>>
>>>Machine is stuck in the task, printing the following call trace
>>>more than 5000 times. The oom-killer invoked once in-between.
>>>
>>>INFO: task msgctl08:16248 blocked for more than 120 seconds.
>>>"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
>>>Call Trace:
>>>[c0000000762435a0] [0000000000000001] 0x1 (unreliable)
>>>[c000000076243770] [c000000000010acc] .__switch_to+0x128/0x16c
>>>[c000000076243800] [c0000000004b07a4] .schedule+0x7ac/0x890
>>>[c0000000762438f0] [c0000000004b2ba4] .rwsem_down_failed_common+0x260/0x2b0
>>>[c0000000762439b0] [c0000000004b2c60] .rwsem_down_read_failed+0x2c/0x44
>>>[c000000076243a60] [c0000000004b1b84] .down_read+0x44/0x5c
>>>[c000000076243af0] [c000000000295e10] .ipc_lock+0x34/0xe0
>>>[c000000076243b90] [c00000000029690c] .ipc_lock_check+0x24/0x78
>>>[c000000076243c20] [c0000000002972c0] .do_msgsnd+0xb0/0x3f8
>>>[c000000076243d10] [c000000000293ce8] .compat_sys_msgsnd+0x94/0xc0
>>>[c000000076243db0] [c000000000014418] .compat_sys_ipc+0x130/0x1f4
>>>[c000000076243e30] [c000000000008734] syscall_exit+0x0/0x40
>>>msgctl08 invoked oom-killer: gfp_mask=0x1200d2, order=0, oomkilladj=0
>>>Call Trace:
>>>[c00000001e1c7640] [c0000000000101bc] .show_stack+0x70/0x1bc (unreliable)
>>>[c00000001e1c76f0] [c0000000000c2c78] .oom_kill_process+0x78/0x230
>>>[c00000001e1c77a0] [c0000000000c3390] .out_of_memory+0x28c/0x320
>>>[c00000001e1c7870] [c0000000000c70ac] .__alloc_pages_internal+0x364/0x468
>>>[c00000001e1c7980] [c0000000000e83cc] .alloc_page_vma+0x120/0x14c
>>>[c00000001e1c7a20] [c0000000000e0224] .read_swap_cache_async+0x7c/0x160
>>>[c00000001e1c7ae0] [c0000000000e035c] .swapin_readahead+0x54/0xd4
>>>[c00000001e1c7ba0] [c0000000000d47e8] .handle_mm_fault+0x520/0x9d8
>>>[c00000001e1c7c80] [c0000000004b5054] .do_page_fault+0x440/0x624
>>>[c00000001e1c7e30] [c0000000000053fc] handle_page_fault+0x20/0x5c
>>>Mem-info:
>>>Node 0 DMA per-cpu:
>>>CPU    0: hi:    6, btch:   1 usd:   1
>>>CPU    1: hi:    6, btch:   1 usd:   1
>>>Node 1 DMA per-cpu:
>>>CPU    0: hi:    6, btch:   1 usd:   5
>>>CPU    1: hi:    6, btch:   1 usd:   5
>>>Active:31 inactive:2628 dirty:1 writeback:6 unstable:0
>>> free:156 slab:13071 mapped:548 pagetables:41109 bounce:0
>>>Node 0 DMA free:5312kB min:5760kB low:7168kB high:8640kB active:1024kB inactive:768kB present:3928832kB pages_scanned:3912 all_unreclaimable? no
>>>lowmem_reserve[]: 0 0 0
>>>Node 1 DMA free:4672kB min:4992kB low:6208kB high:7488kB active:960kB inactive:169792kB present:3404992kB pages_scanned:140140 all_unreclaimable? no
>>>lowmem_reserve[]: 0 0 0
>>>Node 0 DMA: 5*64kB 15*128kB 2*256kB 1*512kB 0*1024kB 1*2048kB 0*4096kB 0*8192kB 0*16384kB = 5312kB
>>>Node 1 DMA: 8*64kB 5*128kB 6*256kB 0*512kB 0*1024kB 1*2048kB 0*4096kB 0*8192kB 0*16384kB = 4736kB
>>>8185 total pagecache pages
>>>Swap cache: add 29545, delete 21888, find 5602/10373
>>>Free swap  = 803712kB
>>>Total swap = 2048128kB
>>>114688 pages of RAM
>>>766 reserved pages
>>>231218 pages shared
>>>7657 pages swap cached
>>>Out of memory: kill process 15371 (msgctl08) score 39223 or a child
>>>Killed process 15373 (msgctl08)
>>>INFO: task msgctl08:15385 blocked for more than 120 seconds.
>>>"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
>>>Call Trace:
>>>[c0000000e9a43290] [0000000000000001] 0x1 (unreliable)
>>>[c0000000e9a43460] [c000000000010acc] .__switch_to+0x128/0x16c
>>>[c0000000e9a434f0] [c0000000004b07a4] .schedule+0x7ac/0x890
>>>[c0000000e9a435e0] [c0000000004b0d90] .io_schedule+0x7c/0xe8
>>>[c0000000e9a43670] [c0000000002d7f44] .get_request_wait+0x15c/0x1e0
>>>[c0000000e9a43750] [c0000000002d8950] .__make_request+0x3ec/0x4d8
>>>[c0000000e9a43800] [c0000000002d6624] .generic_make_request+0x35c/0x3b0
>>>[c0000000e9a438e0] [c0000000002d81b0] .submit_bio+0x118/0x140
>>>[c0000000e9a439a0] [c0000000000dfa88] .swap_readpage+0x94/0xb4
>>>[c0000000e9a43a20] [c0000000000e02b8] .read_swap_cache_async+0x110/0x160
>>>[c0000000e9a43ae0] [c0000000000e035c] .swapin_readahead+0x54/0xd4
>>>[c0000000e9a43ba0] [c0000000000d47e8] .handle_mm_fault+0x520/0x9d8
>>>[c0000000e9a43c80] [c0000000004b5054] .do_page_fault+0x440/0x624
>>>[c0000000e9a43e30] [c0000000000053fc] handle_page_fault+0x20/0x5c
>>>INFO: task msgctl08:15405 blocked for more than 120 seconds.
>>>"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
>>>Call Trace:
>>>[c0000000b1757290] [0000000000000001] 0x1 (unreliable)
>>>[c0000000b1757460] [c000000000010acc] .__switch_to+0x128/0x16c
>>>[c0000000b17574f0] [c0000000004b07a4] .schedule+0x7ac/0x890
>>>[c0000000b17575e0] [c0000000004b0d90] .io_schedule+0x7c/0xe8
>>>[c0000000b1757670] [c0000000002d7f44] .get_request_wait+0x15c/0x1e0
>>>[c0000000b1757750] [c0000000002d8950] .__make_request+0x3ec/0x4d8
>>>[c0000000b1757800] [c0000000002d6624] .generic_make_request+0x35c/0x3b0
>>>[c0000000b17578e0] [c0000000002d81b0] .submit_bio+0x118/0x140
>>>[c0000000b17579a0] [c0000000000dfa88] .swap_readpage+0x94/0xb4
>>>[c0000000b1757a20] [c0000000000e02b8] .read_swap_cache_async+0x110/0x160
>>>[c0000000b1757ae0] [c0000000000e035c] .swapin_readahead+0x54/0xd4
>>>[c0000000b1757ba0] [c0000000000d47e8] .handle_mm_fault+0x520/0x9d8
>>>[c0000000b1757c80] [c0000000004b5054] .do_page_fault+0x440/0x624
>>>[c0000000b1757e30] [c0000000000053fc] handle_page_fault+0x20/0x5c
>>>INFO: task msgctl08:15406 blocked for more than 120 seconds.
>>>"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
>>>Call Trace:
>>>[c0000000eec53290] [0000000000000001] 0x1 (unreliable)
>>>[c0000000eec53460] [c000000000010acc] .__switch_to+0x128/0x16c
>>>[c0000000eec534f0] [c0000000004b07a4] .schedule+0x7ac/0x890
>>>[c0000000eec535e0] [c0000000004b0d90] .io_schedule+0x7c/0xe8
>>>[c0000000eec53670] [c0000000002d7f44] .get_request_wait+0x15c/0x1e0
>>>[c0000000eec53750] [c0000000002d8950] .__make_request+0x3ec/0x4d8
>>>[c0000000eec53800] [c0000000002d6624] .generic_make_request+0x35c/0x3b0
>>>[c0000000eec538e0] [c0000000002d81b0] .submit_bio+0x118/0x140
>>>[c0000000eec539a0] [c0000000000dfa88] .swap_readpage+0x94/0xb4
>>>[c0000000eec53a20] [c0000000000e02b8] .read_swap_cache_async+0x110/0x160
>>>[c0000000eec53ae0] [c0000000000e035c] .swapin_readahead+0x54/0xd4
>>>[c0000000eec53ba0] [c0000000000d47e8] .handle_mm_fault+0x520/0x9d8
>>>[c0000000eec53c80] [c0000000004b5054] .do_page_fault+0x440/0x624
>>>[c0000000eec53e30] [c0000000000053fc] handle_page_fault+0x20/0x5c

Kamalesh,

I'm wondering whether this is not related to a problem found in LTP 
after msgmni had been changed from 16 to a computed value.

Does your ltp test suite contain the patch I had sent in thread 
http://lkml.org/lkml/2008/2/21/123?

If you want to have a look at it, discussion started in thread 
http://lkml.org/lkml/2008/2/16/6

Regards,
Nadia