Bug 6648 - swapper: page allocation failure. order:0, mode:0x20
swapper: page allocation failure. order:0, mode:0x20
Status: REJECTED UNREPRODUCIBLE
Product: Memory Management
Classification: Unclassified
Component: Page Allocator
i386 Linux
: P2 normal
Assigned To: Andrew Morton
:
Depends on:
Blocks:
  Show dependency treegraph
 
Reported: 2006-06-05 04:25 UTC by Pawel Staszewski
Modified: 2009-07-11 06:52 UTC (History)
4 users (show)

See Also:
Kernel Version: 2.6.17-rc5-git7
Tree: Mainline
Regression: No


Attachments
cat /proc/net/fib_trie (874.39 KB, application/octet-stream)
2008-03-31 11:16 UTC, Pawel Staszewski
Details

Description Pawel Staszewski 2006-06-05 04:25:19 UTC
Most recent kernel where this bug did not occur: don't know
Distribution: Gentoo
Hardware Environment: Intel(R) Pentium(R) 4 CPU 3.00GHz cpu family:15 model: 4
stepping: 3  x86_64 2x intel e1000 1x acenic Tigon2
Software Environment: Quagga,  Zebra BGP
Problem Description: After 15 min of system running :

cat /proc/meminfo
MemTotal:       502684 kB
MemFree:        259560 kB
Buffers:          3048 kB
Cached:           5684 kB
SwapCached:      17788 kB
Active:         139980 kB
Inactive:        20452 kB
HighTotal:           0 kB
HighFree:            0 kB
LowTotal:       502684 kB
LowFree:        259560 kB
SwapTotal:      996020 kB
SwapFree:       974940 kB
Dirty:               4 kB
Writeback:           0 kB
Mapped:         152736 kB
Slab:            72716 kB
CommitLimit:   1247360 kB
Committed_AS:   163188 kB
PageTables:       1004 kB
VmallocTotal: 34359738367 kB
VmallocUsed:      2356 kB
VmallocChunk: 34359735819 kB
HugePages_Total:     0
HugePages_Free:      0
HugePages_Rsvd:      0
Hugepagesize:     2048 kB



Call Trace: <IRQ> <ffffffff8024abf7>{__alloc_pages+623}
       <ffffffff802145d4>{smp_local_timer_interrupt+35}
<ffffffff80214639>{smp_apic_timer_interrupt+65}
       <ffffffff802612f2>{cache_alloc_refill+695} <ffffffff802616d7>{__kmalloc+114}
       <ffffffff8038d148>{__alloc_skb+85} <ffffffff8034123e>{e1000_clean_rx_irq+740}
       <ffffffff80340e94>{e1000_clean+239} <ffffffff80392be2>{net_rx_action+171}
       <ffffffff8022d553>{__do_softirq+103} <ffffffff8020aaa6>{call_softirq+30}
       <ffffffff8020c678>{do_softirq+61} <ffffffff8020c631>{do_IRQ+62}
       <ffffffff80207e94>{mwait_idle+0} <ffffffff80209e04>{ret_from_intr+0} <EOI>
       <ffffffff80207ede>{mwait_idle+74} <ffffffff80207e75>{cpu_idle+98}
       <ffffffff805da79b>{start_secondary+1127}
Mem-info:
DMA per-cpu:
cpu 0 hot: high 0, batch 1 used:0
cpu 0 cold: high 0, batch 1 used:0
cpu 1 hot: high 0, batch 1 used:0
cpu 1 cold: high 0, batch 1 used:0
DMA32 per-cpu:
cpu 0 hot: high 186, batch 31 used:72
cpu 0 cold: high 62, batch 15 used:13
cpu 1 hot: high 186, batch 31 used:30
cpu 1 cold: high 62, batch 15 used:50
Normal per-cpu: empty
HighMem per-cpu: empty
Free pages:        4920kB (0kB HighMem)
Active:48439 inactive:4651 dirty:0 writeback:0 unstable:0 free:1230 slab:68845
mapped:39985 pagetables:255
DMA free:1992kB min:172kB low:212kB high:256kB active:0kB inactive:0kB
present:10604kB pages_scanned:0 all_unreclaimable? yes
lowmem_reserve[]: 0 481 481 481
DMA32 free:2928kB min:8016kB low:10020kB high:12024kB active:193756kB
inactive:18604kB present:492776kB pages_scanned:126 all_unreclaimable? no
lowmem_reserve[]: 0 0 0 0
Normal free:0kB min:0kB low:0kB high:0kB active:0kB inactive:0kB present:0kB
pages_scanned:0 all_unreclaimable? no
lowmem_reserve[]: 0 0 0 0
HighMem free:0kB min:128kB low:128kB high:128kB active:0kB inactive:0kB
present:0kB pages_scanned:0 all_unreclaimable? no
lowmem_reserve[]: 0 0 0 0
DMA: 0*4kB 1*8kB 0*16kB 0*32kB 1*64kB 1*128kB 1*256kB 1*512kB 1*1024kB 0*2048kB
0*4096kB = 1992kB
DMA32: 0*4kB 0*8kB 1*16kB 1*32kB 1*64kB 0*128kB 1*256kB 1*512kB 0*1024kB
1*2048kB 0*4096kB = 2928kB
Normal: empty
HighMem: empty
Swap cache: add 0, delete 0, find 0/0, race 0+0
Free swap  = 996020kB
Total swap = 996020kB
Free swap:       996020kB
128997 pages of RAM
3326 reserved pages
13499 pages shared
0 pages swap cached
swapper: page allocation failure. order:0, mode:0x20

Call Trace: <IRQ> <ffffffff8024abf7>{__alloc_pages+623}
       <ffffffff802612f2>{cache_alloc_refill+695} <ffffffff802616d7>{__kmalloc+114}
       <ffffffff8038d148>{__alloc_skb+85} <ffffffff8034123e>{e1000_clean_rx_irq+740}
       <ffffffff80340e94>{e1000_clean+239} <ffffffff80392be2>{net_rx_action+171}
       <ffffffff8022d553>{__do_softirq+103} <ffffffff8020aaa6>{call_softirq+30}
       <ffffffff8020c678>{do_softirq+61} <ffffffff8020c631>{do_IRQ+62}
       <ffffffff80207e94>{mwait_idle+0} <ffffffff80209e04>{ret_from_intr+0} <EOI>
       <ffffffff80207ede>{mwait_idle+74} <ffffffff80207e75>{cpu_idle+98}
       <ffffffff805da79b>{start_secondary+1127}
Mem-info:
DMA per-cpu:
cpu 0 hot: high 0, batch 1 used:0
cpu 0 cold: high 0, batch 1 used:0
cpu 1 hot: high 0, batch 1 used:0
cpu 1 cold: high 0, batch 1 used:0
DMA32 per-cpu:
cpu 0 hot: high 186, batch 31 used:72
cpu 0 cold: high 62, batch 15 used:13
cpu 1 hot: high 186, batch 31 used:30
cpu 1 cold: high 62, batch 15 used:50
Normal per-cpu: empty
HighMem per-cpu: empty
Free pages:        4920kB (0kB HighMem)
Active:48439 inactive:4651 dirty:0 writeback:0 unstable:0 free:1230 slab:68845
mapped:39985 pagetables:255
DMA free:1992kB min:172kB low:212kB high:256kB active:0kB inactive:0kB
present:10604kB pages_scanned:0 all_unreclaimable? yes
lowmem_reserve[]: 0 481 481 481
DMA32 free:2928kB min:8016kB low:10020kB high:12024kB active:193756kB
inactive:18604kB present:492776kB pages_scanned:126 all_unreclaimable? no
lowmem_reserve[]: 0 0 0 0
Normal free:0kB min:0kB low:0kB high:0kB active:0kB inactive:0kB present:0kB
pages_scanned:0 all_unreclaimable? no
lowmem_reserve[]: 0 0 0 0
HighMem free:0kB min:128kB low:128kB high:128kB active:0kB inactive:0kB
present:0kB pages_scanned:0 all_unreclaimable? no
lowmem_reserve[]: 0 0 0 0
DMA: 0*4kB 1*8kB 0*16kB 0*32kB 1*64kB 1*128kB 1*256kB 1*512kB 1*1024kB 0*2048kB
0*4096kB = 1992kB
DMA32: 0*4kB 0*8kB 1*16kB 1*32kB 1*64kB 0*128kB 1*256kB 1*512kB 0*1024kB
1*2048kB 0*4096kB = 2928kB
Normal: empty
HighMem: empty
Swap cache: add 0, delete 0, find 0/0, race 0+0
Free swap  = 996020kB
Total swap = 996020kB
Free swap:       996020kB
128997 pages of RAM
3326 reserved pages
13499 pages shared
0 pages swap cached
printk: 21353 messages suppressed.
swapper: page allocation failure. order:0, mode:0x20

Call Trace: <IRQ> <ffffffff8024abf7>{__alloc_pages+623}
       <ffffffff802612f2>{cache_alloc_refill+695} <ffffffff802616d7>{__kmalloc+114}
       <ffffffff8038d148>{__alloc_skb+85}
<ffffffff80341c62>{e1000_alloc_rx_buffers+119}
       <ffffffff803414ef>{e1000_clean_rx_irq+1429}
<ffffffff80340e94>{e1000_clean+239}
       <ffffffff80392be2>{net_rx_action+171}
<ffffffff803166dd>{add_timer_randomness+284}
       <ffffffff8022d553>{__do_softirq+103} <ffffffff8020aaa6>{call_softirq+30}
       <ffffffff8020c678>{do_softirq+61} <ffffffff8020c631>{do_IRQ+62}
       <ffffffff80207e94>{mwait_idle+0} <ffffffff80209e04>{ret_from_intr+0} <EOI>
       <ffffffff80207ede>{mwait_idle+74} <ffffffff80207e75>{cpu_idle+98}
       <ffffffff805da79b>{start_secondary+1127}
Mem-info:
DMA per-cpu:
cpu 0 hot: high 0, batch 1 used:0
cpu 0 cold: high 0, batch 1 used:0
cpu 1 hot: high 0, batch 1 used:0
cpu 1 cold: high 0, batch 1 used:0
DMA32 per-cpu:
cpu 0 hot: high 186, batch 31 used:129
cpu 0 cold: high 62, batch 15 used:57
cpu 1 hot: high 186, batch 31 used:30
cpu 1 cold: high 62, batch 15 used:60
Normal per-cpu: empty
HighMem per-cpu: empty
Free pages:        4976kB (0kB HighMem)
Active:40234 inactive:1675 dirty:8 writeback:0 unstable:0 free:1244 slab:79910
mapped:39985 pagetables:255
DMA free:1992kB min:172kB low:212kB high:256kB active:0kB inactive:0kB
present:10604kB pages_scanned:0 all_unreclaimable? yes
lowmem_reserve[]: 0 481 481 481
DMA32 free:2984kB min:8016kB low:10020kB high:12024kB active:160936kB
inactive:6700kB present:492776kB pages_scanned:565 all_unreclaimable? no
lowmem_reserve[]: 0 0 0 0
Normal free:0kB min:0kB low:0kB high:0kB active:0kB inactive:0kB present:0kB
pages_scanned:0 all_unreclaimable? no
lowmem_reserve[]: 0 0 0 0
HighMem free:0kB min:128kB low:128kB high:128kB active:0kB inactive:0kB
present:0kB pages_scanned:0 all_unreclaimable? no
lowmem_reserve[]: 0 0 0 0
DMA: 0*4kB 1*8kB 0*16kB 0*32kB 1*64kB 1*128kB 1*256kB 1*512kB 1*1024kB 0*2048kB
0*4096kB = 1992kB
DMA32: 0*4kB 1*8kB 2*16kB 0*32kB 0*64kB 1*128kB 1*256kB 1*512kB 0*1024kB
1*2048kB 0*4096kB = 2984kB
Normal: empty
HighMem: empty
Swap cache: add 38, delete 28, find 0/0, race 0+0
Free swap  = 995868kB
Total swap = 996020kB
Free swap:       995868kB
128997 pages of RAM
3326 reserved pages
5576 pages shared
10 pages swap cached
Comment 1 Natalie Protasevich 2008-03-24 13:27:37 UTC
This bug needs to be refreshed. Pawel can you please retest with newest kernel?
Comment 2 Pawel Staszewski 2008-03-24 15:16:03 UTC
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
BGP paol # w
 00:10:39 up 1 day, 23:02,  1 user,  load average: 0.01, 0.02, 0.00
USER     TTY        LOGIN@   IDLE   JCPU   PCPU WHAT
paol     pts/0     00:10    0.00s  0.00s  0.02s sshd: paol [priv]
BGP paol # uname -a
Linux BGP 2.6.25-rc6-git6 #1 SMP Sun Mar 23 01:06:42 CET 2008 x86_64 Intel(R) 

With the newest kernel
i have no problem.
only this: Fix inflate_threshold_root. Now=15 size=11 bits
Comment 3 Natalie Protasevich 2008-03-24 15:44:34 UTC
Thanks for the update.
Hmm, this comes from resize() in TCP code. CC-int to Stephen, maybe this bug needs to be re-assigned now. 
Comment 4 Stephen Hemminger 2008-03-31 08:14:30 UTC
Are you using Jumbo frames? I suspect the problem is that e1000
driver has to round up allocations, and when using Jumbo frames this can
mean a 9K frame requires 16K contiguous allocation.  On a busy system
that will fail. Newer kernels generate less fragmented space, but it
is really a hardware problem.
Comment 5 Stephen Hemminger 2008-03-31 08:18:21 UTC
The message 'inflate_threshold_root' comes from the fib_trie routing
code. How many routes do you have and what is the distribution?

It would be useful to see the output of /proc/net/fib_triestat and 
/proc/net/fib_trie
Comment 6 Pawel Staszewski 2008-03-31 11:16:21 UTC
Created attachment 15540 [details]
cat /proc/net/fib_trie
Comment 7 Pawel Staszewski 2008-03-31 11:39:22 UTC
this is bgp router with installed quagga as bgp daemon.
cat /proc/net/fib_triestat
Basic info: size of leaf: 40 bytes, size of tnode: 40 bytes.
Local:
        Aver depth:     5.12
        Max depth:      7
        Leaves:         31
        Internal nodes: 28
          1: 26  2: 2
        Pointers: 60
Null ptrs: 2
Total size: 3  kB
Main:
        Aver depth:     2.26
        Max depth:      6
        Leaves:         235924
        Internal nodes: 57854
          1: 31632  2: 11422  3: 8475  4: 3755  5: 1676  6: 893  18: 1
        Pointers: 609760
Null ptrs: 315983
Total size: 16240  kB

Comment 8 Pawel Staszewski 2008-03-31 11:40:46 UTC
and MTU is 1500 no jumbo frames.
Comment 9 Alan 2008-12-04 06:22:14 UTC
Is this still a problem ?
Comment 10 Pawel Staszewski 2009-05-26 12:25:56 UTC
Kernel 2.6.29.3


Server #1
Sometimes i have the same info like before:
e1000: eth0 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: None
e1000: eth2 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: None
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits

On three different machines.

cat /proc/net/fib_triestat
Basic info: size of leaf: 20 bytes, size of tnode: 36 bytes.
Main:
        Aver depth:     2.27
        Max depth:      7
        Leaves:         268917
        Prefixes:       282352
        Internal nodes: 65708
          1: 36354  2: 12657  3: 9215  4: 4399  5: 2011  6: 1071  18: 1
        Pointers: 662480
Null ptrs: 327856
Total size: 17871  kB

Counters:
---------
gets = 2195484652
backtracks = 636049593
semantic match passed = 2195493159
semantic match miss = 212348885
null node hit= 1353421394
skipped node resize = 0

Local:
        Aver depth:     3.58
        Max depth:      4
        Leaves:         12
        Prefixes:       13
        Internal nodes: 9
          1: 7  2: 2
        Pointers: 22
Null ptrs: 2
Total size: 1  kB

Counters:
---------
gets = 2195935298
backtracks = 1041035352
semantic match passed = 418139
semantic match miss = 0
null node hit= 290544
skipped node resize = 0

Server #2
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits
Fix inflate_threshold_root. Now=15 size=11 bits



 cat /proc/net/fib_triestat
Basic info: size of leaf: 20 bytes, size of tnode: 36 bytes.
Local:
        Aver depth:     3.75
        Max depth:      5
        Leaves:         12
        Prefixes:       13
        Internal nodes: 10
          1: 9  2: 1
        Pointers: 22
Null ptrs: 1
Total size: 2  kB

Counters:
---------
gets = 4070990500
backtracks = 993990300
semantic match passed = 311572
semantic match miss = 0
null node hit= 1556177
skipped node resize = 0

Main:
        Aver depth:     2.26
        Max depth:      7
        Leaves:         268918
        Prefixes:       282353
        Internal nodes: 65142
          1: 34900  2: 12879  3: 9559  4: 4591  5: 2106  6: 1106  18: 1
        Pointers: 671564
Null ptrs: 337505
Total size: 17887  kB

Counters:
---------
gets = 4070765628
backtracks = 41807032
semantic match passed = 4068421081
semantic match miss = 114232
null node hit= 2025898189
skipped node resize = 0

On this servers is working bgpd (quagga 0.99.12)
Comment 11 Jarek Poplawski 2009-07-11 06:52:05 UTC
These problems were tracked on the netdev@ list under the subject: "rib_trie / Fix inflate_threshold_root" (25 Jun 2009).
Fixes in the mainline (2.6.31 tree) start with: "ipv4: Fix fib_trie rebalancing"
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=e0f7cb8c8cc6cccce28d2ce39ad8c60d23c3799f
plus next parts.

Note You need to log in before you can comment on or make changes to this bug.