Re: OOM killer in 2.6.31-rc2

From: Wu Fengguang
Date: Fri Jul 10 2009 - 09:25:15 EST


On Thu, Jul 09, 2009 at 05:03:06PM -0400, Gene Heskett wrote:
> On Thursday 09 July 2009, John Stoffel wrote:
> >>>>>> "Gene" == Gene Heskett <gene.heskett@xxxxxxxxxxx> writes:
> >
> >Gene> On Wednesday 08 July 2009, Wu Fengguang wrote:
> >>> On Wed, Jul 08, 2009 at 01:15:15PM +0800, Wu Fengguang wrote:
> >>>> On Tue, Jul 07, 2009 at 11:42:07PM -0400, Gene Heskett wrote:
> >
> >Gene> [...]
> >
> >>>> I guess your near 800MB slab cache is somehow under scanned.
> >>>
> >>> Gene, can you run .31 with this patch? When OOM happens, it will tell
> >>> us whether the majority slab pages are reclaimable. Another way to
> >>> find things out is to run `slabtop` when your system is moderately
> >>> loaded.
> >
> >Gene> Its been running continuously, and after 24 hours is now showing:
> >
> >Just wondering, is this your M2N-SLI Deluxe board?
> Yes.
> >I've got the same
> >board, with 4Gb of RAM and I haven't noticed any loss of RAM from my
> >looking (quickly) at top output.
>
> I am short approximately 500 megs according to top:
> Mem: 3634228k total, 3522984k used, 111244k free, 308096k buffers
> Swap: 8385912k total, 568k used, 8385344k free, 2544716k cached
>
> From dmesg:
> [ 0.000000] TOM2: 0000000120000000 aka 4608M <what is this?

That 4608M includes memory hole I guess.

> [...]
> [ 0.000000] 2694MB HIGHMEM available.
> [ 0.000000] 887MB LOWMEM available.
>
> The bios signon does say 4092M IIRC.
>
> >But I also haven't bothered to upgrade the BIOS on this board at all
> >since I got it back in March of 2008. No need in my book so far.
>
> I had been running the original bios, #1502, because 1604 and 1701 had very
> poor uptimes. 1502 caused an oops about 15 lines into the boot but that
> triggered a remap and it was bulletproof after that running a 32 bit 64G+PAE
> kernel. (I haven't quite made the jump to a 64 bit install, yet...)
>
> >> uname -a
> >
> >Linux sail 2.6.31-rc1 #6 SMP PREEMPT Wed Jun 24 21:40:33 EDT 2009 x86_64
> > GNU/Linux
>
> Linux coyote.coyote.den 2.6.31-rc2 #4 SMP PREEMPT Wed Jul 8 09:37:15 EDT 2009
> i686 athlon i386 GNU/Linux
> >> cat /proc/meminfo
> >
> >MemTotal: 3987068 kB
> >MemFree: 170608 kB
> >Buffers: 355272 kB
> >Cached: 2034416 kB
> >SwapCached: 0 kB
> >Active: 1836284 kB
> >Inactive: 1482444 kB
> >Active(anon): 857076 kB
> >Inactive(anon): 86112 kB
> >Active(file): 979208 kB
> >Inactive(file): 1396332 kB
> >Unevictable: 3972 kB
> >Mlocked: 3972 kB
> >SwapTotal: 0 kB
> >SwapFree: 0 kB
> >Dirty: 36 kB
> >Writeback: 0 kB
> >AnonPages: 933160 kB
> >Mapped: 141188 kB
> >Slab: 398124 kB
> >SReclaimable: 348212 kB
> >SUnreclaim: 49912 kB
> >PageTables: 30916 kB
> >NFS_Unstable: 0 kB
> >Bounce: 0 kB
> >WritebackTmp: 0 kB
> >CommitLimit: 1993532 kB
> >Committed_AS: 1570980 kB
> >VmallocTotal: 34359738367 kB
> >VmallocUsed: 116160 kB
> >VmallocChunk: 34359584603 kB
> >DirectMap4k: 4992 kB
> >DirectMap2M: 4188160 kB
>
> MemTotal: 3634228 kB
> MemFree: 114312 kB
> Buffers: 309088 kB
> Cached: 2541864 kB
> SwapCached: 72 kB
> Active: 1584988 kB
> Inactive: 1739508 kB
> Active(anon): 354584 kB
> Inactive(anon): 120072 kB
> Active(file): 1230404 kB
> Inactive(file): 1619436 kB
> Unevictable: 100 kB
> Mlocked: 100 kB
> HighTotal: 2759560 kB
> HighFree: 13020 kB
> LowTotal: 874668 kB
> LowFree: 101292 kB
> SwapTotal: 8385912 kB
> SwapFree: 8385344 kB
> Dirty: 52 kB
> Writeback: 0 kB
> AnonPages: 473576 kB
> Mapped: 111332 kB
> Slab: 143624 kB
> SReclaimable: 127820 kB
> SUnreclaim: 15804 kB
> PageTables: 8776 kB
> NFS_Unstable: 0 kB
> Bounce: 0 kB
> WritebackTmp: 0 kB
> CommitLimit: 10203024 kB
> Committed_AS: 1029032 kB
> VmallocTotal: 122880 kB
> VmallocUsed: 44180 kB
> VmallocChunk: 65924 kB
> HugePages_Total: 0
> HugePages_Free: 0
> HugePages_Rsvd: 0
> HugePages_Surp: 0
> Hugepagesize: 4096 kB
> DirectMap4k: 8184 kB
> DirectMap4M: 901120 kB
>
> Huge diffs it appears. ??

Most relevant ones:

- 300+MB >4G memory is not reachable by kernel and user space
- 2.7GB high memory is not usable for slab caches and some other
kernel users

Thanks,
Fengguang
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/