* Re: [BUG] 2.6.23-rc3-mm1 oom-killer gets invoked
[not found] <46CCA14A.6060103@linux.vnet.ibm.com>
@ 2007-08-22 21:22 ` Andrew Morton
0 siblings, 0 replies; only message in thread
From: Andrew Morton @ 2007-08-22 21:22 UTC (permalink / raw)
To: Kamalesh Babulal; +Cc: linux-kernel, Balbir Singh, linux-mm
On Thu, 23 Aug 2007 02:19:14 +0530
Kamalesh Babulal <kamalesh@linux.vnet.ibm.com> wrote:
> I see the oom-killer getting invoked many times on the
> 2.6.23-rc3-mm1 kernel and have attached the complete console
> log and config file.
>
> ========================================================
> cc1 invoked oom-killer: gfp_mask=0x280d2, order=0, oomkilladj=0
>
> [<c0147386>] out_of_memory+0x70/0xf6
>
> [<c01484c4>] <4>cc1 invoked oom-killer: gfp_mask=0x201d2, order=0,
> oomkilladj=0
>
> [<c0147386>] out_of_memory+0x70/0xf6
>
> [<c01484c4>] __alloc_pages+0x21b/0x2a8
>
> [<c0173168>] mntput_no_expire+0x11/0x6e
>
> [<c0149f29>] __do_page_cache_readahead+0xc8/0x13a
>
> [<c014535b>] filemap_nopage+0x164/0x30d
>
> [<c0150c7d>] do_no_page+0x91/0x2fb
>
> [<c0151201>] __handle_mm_fault+0x151/0x2bc
>
> [<c016000f>] do_filp_open+0x25/0x39
>
> [<c0115990>] do_page_fault+0x2a3/0x5f7
>
> [<c01156ed>] do_page_fault+0x0/0x5f7
>
> [<c03336c4>] error_code+0x7c/0x84
>
> [<c0330000>] packet_rcv+0xfd/0x2d7
>
> =======================
>
> DMA per-cpu:
>
> CPU 0: Hot: hi: 0, btch: 1 usd: 0 Cold: hi: 0, btch: 1
> usd: 0
>
> CPU 1: Hot: hi: 0, btch: 1 usd: 0 Cold: hi: 0, btch: 1
> usd: 0
>
> CPU 2: Hot: hi: 0, btch: 1 usd: 0 Cold: hi: 0, btch: 1
> usd: 0
>
> CPU 3: Hot: hi: 0, btch: 1 usd: 0 Cold: hi: 0, btch: 1
> usd: 0
>
> Normal per-cpu:
>
> CPU 0: Hot: hi: 186, btch: 31 usd: 67 Cold: hi: 62, btch: 15
> usd: 57
>
> CPU 1: Hot: hi: 186, btch: 31 usd: 170 Cold: hi: 62, btch: 15
> usd: 57
>
> CPU 2: Hot: hi: 186, btch: 31 usd: 126 Cold: hi: 62, btch: 15
> usd: 50
>
> CPU 3: Hot: hi: 186, btch: 31 usd: 50 Cold: hi: 62, btch: 15
> usd: 19
>
> HighMem per-cpu:
>
> CPU 0: Hot: hi: 186, btch: 31 usd: 21 Cold: hi: 62, btch: 15
> usd: 10
>
> CPU 1: Hot: hi: 186, btch: 31 usd: 23 Cold: hi: 62, btch: 15
> usd: 19
>
> CPU 2: Hot: hi: 186, btch: 31 usd: 7 Cold: hi: 62, btch: 15
> usd: 27
>
> CPU 3: Hot: hi: 186, btch: 31 usd: 20 Cold: hi: 62, btch: 15
> usd: 56
>
> Active:228916 inactive:238880 dirty:0 writeback:0 unstable:0
>
> free:12273 slab:14578 mapped:36 pagetables:10564 bounce:0
>
> DMA free:8076kB min:68kB low:84kB high:100kB active:2496kB
> inactive:1716kB present:16224kB pages_scanned:171850 all_unreclaimable? yes
>
> lowmem_reserve[]: 0 871 2011
>
> Normal free:40400kB min:3740kB low:4672kB high:5608kB active:358456kB
> inactive:358000kB present:892320kB pages_scanned:1427617
> all_unreclaimable? yes
>
> lowmem_reserve[]: 0 0 9123
>
> HighMem free:616kB min:512kB low:1736kB high:2960kB active:554712kB
> inactive:595804kB present:1167812kB pages_scanned:2215902
> all_unreclaimable? yes
>
> lowmem_reserve[]: 0 0 0
>
> DMA: 1*4kB 1*8kB 2*16kB 1*32kB 1*64kB 0*128kB 1*256kB 1*512kB 1*1024kB
> 1*2048kB 1*4096kB = 8076kB
>
> Normal: 70*4kB 25*8kB 1*16kB 1*32kB 1*64kB 33*128kB 7*256kB 0*512kB
> 1*1024kB 0*2048kB 8*4096kB = 40400kB
>
> HighMem: 49*4kB 20*8kB 4*16kB 6*32kB 1*64kB 0*128kB 0*256kB 0*512kB
> 0*1024kB 0*2048kB 0*4096kB = 676kB
>
> Swap cache: add 12954, delete 12954, find 2676/3551, race 0+1
>
> Free swap = 0kB
>
> Total swap = 8024kB
>
> Out of memory: kill process 27480 (make) score 2082 or a child
It's a bit harsh, sending 330kb emails to linux-kernel. Please try to
strip these things down to the important information. Also, it'd be nice
if you could find the source of those double-linefeeds and make it stop.
I don't know whwy the VM decided that you've run out of memory. After the
oom-killing, does
echo 3 > /proc/sys/vm/drop_caches
make the numbers in /proc/meminfo change significantly?
Which filesystem was being used here?
Thanks.
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
^ permalink raw reply [flat|nested] only message in thread