Re: swapper: page allocation failure. order:2, mode:0x4020 in 2.6.27

netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed

* Re: swapper: page allocation failure. order:2, mode:0x4020 in 2.6.27
       [not found] <a06240403c58ae633aba5@[129.98.90.227]>
@ 2009-01-08  1:19 ` KOSAKI Motohiro
  2009-01-08  4:44   ` Evgeniy Polyakov
  0 siblings, 1 reply; 2+ messages in thread
From: KOSAKI Motohiro @ 2009-01-08  1:19 UTC (permalink / raw)
  To: Maurice Volaski; +Cc: kosaki.motohiro, linux-kernel, netdev

Hi

(cc to netdev)

> Just started with 2.6.27-gentoo-r7 (I think gentoo may have patches 
> from 2.6.28) on an x64 system and during an apparently idle time, I 
> saw this:
> 
> Jan  6 23:20:19 [kernel] [358105.255843] swapper: page allocation 
> failure. order:2, mode:0x4020
> Jan  6 23:20:19 [kernel] [358105.255851] Pid: 0, comm: swapper Not 
> tainted 2.6.27-gentoo-r7 #3
> Jan  6 23:20:19 [kernel] [358105.255854]
> Jan  6 23:20:19 [kernel] [358105.255854] Call Trace:
> Jan  6 23:20:19 [kernel] [358105.255857]  <IRQ>  [<ffffffff81076bbf>] 
> __alloc_pages_internal+0x41a/0x439
> Jan  6 23:20:19 [kernel] [358105.255872]  [<ffffffff8126575c>] ? 
> __netdev_alloc_skb+0x1d/0x39
> Jan  6 23:20:19 [kernel] [358105.255875]  [<ffffffff81076c4c>] 
> __get_free_pages+0x17/0x56
> Jan  6 23:20:19 [kernel] [358105.255879]  [<ffffffff8109632b>] 
> __kmalloc_track_caller+0x3b/0xbb
> Jan  6 23:20:19 [kernel] [358105.255882]  [<ffffffff81264d87>] 
> __alloc_skb+0x66/0x12c
> Jan  6 23:20:19 [kernel] [358105.255885]  [<ffffffff8126575c>] 
> __netdev_alloc_skb+0x1d/0x39
> Jan  6 23:20:19 [kernel] [358105.255899]  [<ffffffffa00d203c>] 
> tg3_alloc_rx_skb+0xd1/0x195 [tg3]

if stack dump has tg3_alloc_rx_skb(), that isn't bug. (maybe)
tg3 network card often reqire non order-0 pages under irq/softirq context.

but linux memory manager can't drop any cache under irq/softirq context.
therefore the allocation failure is relatively typical.

the allocation failure in packet receive (tg3_alloc_rx_skb) cause
packet dropping and network peer resend the same packet.
it decrease network performance a bit. but it doesn't cause any serious problem.

last problem is, why we should look this false-positive warning repeatedly.


net-guys, if I say false thing, please fix.



> Jan  6 23:20:19 [kernel] [358105.255906]  [<ffffffffa00d9eb2>] 
> tg3_poll+0x563/0xa68 [tg3]
> Jan  6 23:20:19 [kernel] [358105.255910]  [<ffffffff8104b567>] ? 
> clockevents_program_event+0x73/0x7c
> Jan  6 23:20:19 [kernel] [358105.255914]  [<ffffffff81268608>] 
> net_rx_action+0xd3/0x1d0
> Jan  6 23:20:19 [kernel] [358105.255917]  [<ffffffff81268597>] ? 
> net_rx_action+0x62/0x1d0
> Jan  6 23:20:19 [kernel] [358105.255921]  [<ffffffff81035222>] 
> __do_softirq+0x6d/0xdd
> Jan  6 23:20:19 [kernel] [358105.255925]  [<ffffffff8100d2cc>] 
> call_softirq+0x1c/0x28
> Jan  6 23:20:19 [kernel] [358105.255928]  [<ffffffff8100e405>] 
> do_softirq+0x34/0x72
> Jan  6 23:20:19 [kernel] [358105.255931]  [<ffffffff81034f54>] 
> irq_exit+0x3f/0x82
> Jan  6 23:20:19 [kernel] [358105.255933]  [<ffffffff8100e6ee>] 
> do_IRQ+0x146/0x168
> Jan  6 23:20:19 [kernel] [358105.255936]  [<ffffffff8100c541>] 
> ret_from_intr+0x0/0xa
> Jan  6 23:20:19 [kernel] [358105.255938]  <EOI>  [<ffffffff812dd3cc>] 
> ? __atomic_notifier_call_chain+0x0/0x87
> Jan  6 23:20:19 [kernel] [358105.255947]  [<ffffffff81012593>] ? 
> default_idle+0x2b/0x40
> Jan  6 23:20:19 [kernel] [358105.255950]  [<ffffffff8100b041>] ? 
> enter_idle+0x22/0x24
> Jan  6 23:20:19 [kernel] [358105.255952]  [<ffffffff8100b0d3>] ? 
> cpu_idle+0x90/0xd8
> Jan  6 23:20:19 [kernel] [358105.255956]  [<ffffffff812d4ce7>] ? 
> start_secondary+0x155/0x159
> Jan  6 23:20:19 [kernel] [358105.255958]
> Jan  6 23:20:19 [kernel] [358105.255959] Mem-Info:
> Jan  6 23:20:19 [kernel] [358105.255961] DMA per-cpu:
> Jan  6 23:20:19 [kernel] [358105.255963] CPU    0: hi:    0, btch:   1 usd:   0
> Jan  6 23:20:19 [kernel] [358105.255965] CPU    1: hi:    0, btch:   1 usd:   0
> Jan  6 23:20:19 [kernel] [358105.255967] DMA32 per-cpu:
> Jan  6 23:20:19 [kernel] [358105.255968] CPU    0: hi:  186, btch:  31 usd: 126
> Jan  6 23:20:19 [kernel] [358105.255970] CPU    1: hi:  186, btch:  31 usd: 148
> Jan  6 23:20:19 [kernel] [358105.255972] Normal per-cpu:
> Jan  6 23:20:19 [kernel] [358105.255973] CPU    0: hi:  186, btch:  31 usd: 181
> Jan  6 23:20:19 [kernel] [358105.255975] CPU    1: hi:  186, btch:  31 usd: 183
> Jan  6 23:20:19 [kernel] [358105.255978] Active:218402 
> inactive:819867 dirty:27016 writeback:0 unstable:0
> Jan  6 23:20:19 [kernel] [358105.255980]  free:44839 slab:151548 
> mapped:5449 pagetables:2110 bounce:0
> Jan  6 23:20:19 [kernel] [358105.255983] DMA free:5328kB min:4kB 
> low:4kB high:4kB active:0kB inactive:0kB present:4256kB 
> pages_scanned:0 all_unreclaimable? yes
> Jan  6 23:20:19 [kernel] [358105.255986] lowmem_reserve[]: 0 3480 5484 5484
> Jan  6 23:20:19 [kernel] [358105.255991] DMA32 free:169860kB 
> min:6008kB low:7508kB high:9012kB active:433812kB inactive:1876764kB 
> present:3563808kB pages_scanned:0 all_unreclaimable? no
> Jan  6 23:20:19 [kernel] [358105.255993] lowmem_reserve[]: 0 0 2004 2004
> Jan  6 23:20:19 [kernel] [358105.255998] Normal free:4168kB 
> min:3460kB low:4324kB high:5188kB active:439796kB inactive:1402704kB 
> present:2052096kB pages_scanned:0 all_unreclaimable? no
> Jan  6 23:20:19 [kernel] [358105.256000] lowmem_reserve[]: 0 0 0 0
> Jan  6 23:20:19 [kernel] [358105.256004] DMA: 2*4kB 3*8kB 5*16kB 
> 5*32kB 3*64kB 4*128kB 3*256kB 1*512kB 1*1024kB 1*2048kB 0*4096kB = 
> 5328kB
> Jan  6 23:20:19 [kernel] [358105.256012] DMA32: 42208*4kB 0*8kB 
> 12*16kB 3*32kB 0*64kB 1*128kB 0*256kB 1*512kB 0*1024kB 0*2048kB 
> 0*4096kB = 169760kB
> Jan  6 23:20:19 [kernel] [358105.256020] Normal: 914*4kB 1*8kB 2*16kB 
> 1*32kB 1*64kB 1*128kB 1*256kB 0*512kB 0*1024kB 0*2048kB 0*4096kB = 
> 4176kB
> Jan  6 23:20:19 [kernel] [358105.256029] 1015437 total pagecache pages
> Jan  6 23:20:19 [kernel] [358105.256031] 0 pages in swap cache
> Jan  6 23:20:19 [kernel] [358105.256032] Swap cache stats: add 0, 
> delete 0, find 0/0
> Jan  6 23:20:19 [kernel] [358105.256034] Free swap  = 0kB
> Jan  6 23:20:19 [kernel] [358105.256035] Total swap = 0kB
> Jan  6 23:20:19 [kernel] [358105.285015] 1572848 pages RAM
> Jan  6 23:20:19 [kernel] [358105.285015] 170874 pages reserved
> Jan  6 23:20:19 [kernel] [358105.285015] 986700 pages shared
> Jan  6 23:20:19 [kernel] [358105.285015] 411152 pages non-shared
> -- 
> 
> Maurice Volaski, mvolaski@aecom.yu.edu
> Computing Support, Rose F. Kennedy Center
> Albert Einstein College of Medicine of Yeshiva University
> --
> 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/




^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: swapper: page allocation failure. order:2, mode:0x4020 in 2.6.27
  2009-01-08  1:19 ` swapper: page allocation failure. order:2, mode:0x4020 in 2.6.27 KOSAKI Motohiro
@ 2009-01-08  4:44   ` Evgeniy Polyakov
  0 siblings, 0 replies; 2+ messages in thread
From: Evgeniy Polyakov @ 2009-01-08  4:44 UTC (permalink / raw)
  To: KOSAKI Motohiro; +Cc: Maurice Volaski, linux-kernel, netdev

Hi.

On Thu, Jan 08, 2009 at 10:19:23AM +0900, KOSAKI Motohiro (kosaki.motohiro@jp.fujitsu.com) wrote:
> if stack dump has tg3_alloc_rx_skb(), that isn't bug. (maybe)
> tg3 network card often reqire non order-0 pages under irq/softirq context.
> 
> but linux memory manager can't drop any cache under irq/softirq context.
> therefore the allocation failure is relatively typical.
> 
> the allocation failure in packet receive (tg3_alloc_rx_skb) cause
> packet dropping and network peer resend the same packet.
> it decrease network performance a bit. but it doesn't cause any serious problem.
> 
> last problem is, why we should look this false-positive warning repeatedly.
> 
> 
> net-guys, if I say false thing, please fix.

You are right. You repeatedly get that message likely because of jumbo
frames enabled and fragmented ram which does not allow to allocate more
than order-0 pages.

-- 
	Evgeniy Polyakov

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2009-01-08  4:44 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
     [not found] <a06240403c58ae633aba5@[129.98.90.227]>
2009-01-08  1:19 ` swapper: page allocation failure. order:2, mode:0x4020 in 2.6.27 KOSAKI Motohiro
2009-01-08  4:44   ` Evgeniy Polyakov

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).