kswapd0: page allocation failure. order:2, mode:0x4020问题怎么解决

[mh_memory]

kswapd0: page allocation failure. order:2, mode:0x4020
Call Trace:
[<80263238>] dump_stack+0x8/0x34
[<8005915c>] __alloc_pages_nodemask+0x4dc/0x54c
[<80059238>] __get_free_pages+0x18/0x4c
[<8018b494>] __alloc_skb+0x7c/0x150
[<8014e7e4>] rx_submit+0x3c/0x228
[<8014f494>] usbnet_bh+0x1d8/0x2dc
[<80029970>] tasklet_action+0x88/0xdc
[<8002a0d0>] __do_softirq+0xb0/0x148
[<8002a1b0>] do_softirq+0x48/0x6c
[<80006fec>] ret_from_irq+0x0/0x4
[<801027e8>] __down_read_trylock+0x44/0x50
[<8005fbd8>] shrink_slab+0x48/0x1dc
[<800600e0>] kswapd+0x374/0x504
[<8003813c>] kthread+0x7c/0x88
[<8000917c>] kernel_thread_helper+0x10/0x18

Mem-Info:
Normal per-cpu:
CPU 0: hi: 18, btch: 3 usd: 16
Active_anon:704 active_file:1536 inactive_anon:790
inactive_file:7047 unevictable:0 dirty:0 writeback:3 unstable:0
free:1075 slab:1417 mapped:1017 pagetables:111 bounce:0
Normal free:4300kB min:4096kB low:5120kB high:6144kB active_anon:2816kB inactive_anon:3160kB active_file:6144kB inactive_file:28188kB unevictable:0kB present:65024kB pages_scanned:0 all_unreclaimable? no
lowmem_reserve[]: 0 0
Normal: 299*4kB 340*8kB 0*16kB 0*32kB 2*64kB 0*128kB 1*256kB 0*512kB 0*1024kB 0*2048kB 0*4096kB = 4300kB
8740 total pagecache pages
16384 pages RAM
936 pages reserved
8827 pages shared
7970 pages non-shared
eth2: kevent 2 may have been dropped

[mh_memory]

引用 6 楼 morris88 的回复:

swapoff -a

内核配置里面CONFIG_SWAP是关掉的、 然后不止是kswap0，有时还打印： sleep: page allocation failure. order:2, mode:0x4020 events/0: page allocation failure. order:2, mode:0x4020 这样的信息

[mh_memory]

然后用top命令查看的话，buff和cached还各占了10+M，内存64M、

[morris88]

__alloc_skb 分配内存失败

[mh_memory]

为什么怎么说？然后怎么处理？

[morris88]

kswapd0: page allocation failure. order:2, mode:0x4020 Call Trace: [<8005915c>] __alloc_pages_nodemask+0x4dc/0x54c [<80059238>] __get_free_pages+0x18/0x4c [<8018b494>] __alloc_skb+0x7c/0x150 增加内存吧

[mh_memory]

你的意思是，根据__alloc_skb这条语句得出分配内存失败。 然后而且是内存不足导致分配失败的么？

[mh_memory]

page allocation failure这种问题怎么解决？

[mh_memory]

增加/proc/sys/vm/min_free_kbytes值还是没效果，问题照样发生、、

[morris88]

是不是服务器负载过重？

[mh_memory]

嵌入式产品，流量比较大的情况下测试一小会就出现这种问题了、、

[morris88]

关掉swap吧，一般嵌入式不需要这个功能

[morris88]

swapoff -a

[zhouruheng2012]

shell@android:/ $ shell@android:/ $ [ 244.931462] Thread-111: page allocation failure: order:6, mode:0xd1 [ 244.933215] [<c05ae2a8>] (unwind_backtrace+0x0/0xf0) from [<c062f8c4>] (warn_alloc_failed+0xe4/0x108) [ 244.948788] [<c062f8c4>] (warn_alloc_failed+0xe4/0x108) from [<c062fedc>] (__alloc_pages_nodemask+0x5f4/0x6d4) [ 244.960932] [<c062fedc>] (__alloc_pages_nodemask+0x5f4/0x6d4) from [<c05aee18>] (__dma_alloc+0xb0/0x2d0) [ 244.970512] [<c05aee18>] (__dma_alloc+0xb0/0x2d0) from [<c05af0b0>] (dma_alloc_coherent+0x54/0x60) [ 244.980830] [<c05af0b0>] (dma_alloc_coherent+0x54/0x60) from [<bf06adb0>] (_mali_osk_mem_allocioregion+0x14/0x18 [mali]) [ 244.992075] [<bf06adb0>] (_mali_osk_mem_allocioregion+0x14/0x18 [mali]) from [<bf071dd0>] (os_allocator_allocate_page_table_block+0x88/0x10c [mali]) [ 245.008033] [<bf071dd0>] (os_allocator_allocate_page_table_block+0x88/0x10c [mali]) from [<bf0714d4>] (mali_allocation_engine_allocate_page_tables+0x2) [ 245.024760] [<bf0714d4>] (mali_allocation_engine_allocate_page_tables+0x20/0x58 [mali]) from [<bf0704fc>] (mali_mmu_get_table_page+0x150/0x250 [mali]) [ 245.039043] [<bf0704fc>] (mali_mmu_get_table_page+0x150/0x250 [mali]) from [<bf06fde0>] (mali_mmu_pagedir_map+0x5c/0xfc [mali]) [ 245.052842] [<bf06fde0>] (mali_mmu_pagedir_map+0x5c/0xfc [mali]) from [<bf071620>] (mali_allocation_engine_allocate_memory+0x24/0x1a8 [mali]) [ 245.066213] [<bf071620>] (mali_allocation_engine_allocate_memory+0x24/0x1a8 [mali]) from [<bf070754>] (_mali_ukk_mem_mmap+0x9c/0xe4 [mali]) [ 245.079946] [<bf070754>] (_mali_ukk_mem_mmap+0x9c/0xe4 [mali]) from [<bf06d34c>] (mali_mmap+0xa4/0xcc [mali]) [ 245.090338] [<bf06d34c>] (mali_mmap+0xa4/0xcc [mali]) from [<c064b898>] (mmap_region+0x2b8/0x4f4) [ 245.101502] [<c064b898>] (mmap_region+0x2b8/0x4f4) from [<c064be68>] (sys_mmap_pgoff+0x8c/0xb8) [ 245.110305] [<c064be68>] (sys_mmap_pgoff+0x8c/0xb8) from [<c05a99c0>] (ret_fast_syscall+0x0/0x30) [ 245.120605] Mem-info: [ 245.123645] Normal per-cpu: [ 245.126471] CPU 0: hi: 186, btch: 31 usd: 95 [ 245.131340] active_anon:38634 inactive_anon:2604 isolated_anon:0 [ 245.131348] active_file:10649 inactive_file:10686 isolated_file:0 [ 245.131356] unevictable:367 dirty:0 writeback:0 unstable:0 [ 245.131364] free:795 slab_reclaimable:947 slab_unreclaimable:1756 [ 245.131372] mapped:17030 shmem:2617 pagetables:2355 bounce:0 [ 245.163752] Normal free:3180kB min:2584kB low:3228kB high:3876kB active_anon:154536kB inactive_anon:10416kB active_file:42596kB inactive_file:42744kB o [ 245.210811] lowmem_reserve[]: 0 0 0 [ 245.214440] Normal: 533*4kB 73*8kB 23*16kB 3*32kB 0*64kB 0*128kB 0*256kB 0*512kB 0*1024kB 0*2048kB 0*4096kB = 3180kB [ 245.226131] 24284 total pagecache pages [ 245.231198] 0 pages in swap cache [ 245.234544] Swap cache stats: add 0, delete 0, find 0/0 [ 245.239834] Free swap = 0kB [ 245.242771] Total swap = 0kB [ 245.266275] 131072 pages of RAM [ 245.269784] 1842 free pages [ 245.272638] 31115 reserved pages [ 245.275962] 2358 slab pages [ 245.279016] 245018 pages shared [ 245.282203] 0 pages swap cached 我也遇到这种问题，系统能跑起来，楼主怎么解决的？