z_wr_int can be interrupted when CONFIG_PREEMPT_NONE=y is set #746
Comments
|
Here is another failure that occurred: [ 5103.739075] z_wr_int/11: page allocation failure: order:0, mode:0x4020 |
|
This is a failed atomic allocations in the network layer (__alloc_skb) likely because we've consumed to much of the reserve. It will result in a packet drop which is harmless since it will trigger a retransmit but does show we're still pushing the VM pretty hard. |
|
I only observe these failed allocations in the context of a z_wr_int thread. I have yet to observe them in any other situation, which is strange. I suspect that the use of PF_MEMALLOC had an unforeseen side effect. |
|
This issue is resolved by pull request #883. |
) Bumps [sysinfo](https://github.com/GuillaumeGomez/sysinfo) from 0.28.1 to 0.28.2. - [Release notes](https://github.com/GuillaumeGomez/sysinfo/releases) - [Changelog](https://github.com/GuillaumeGomez/sysinfo/blob/master/CHANGELOG.md) - [Commits](https://github.com/GuillaumeGomez/sysinfo/commits) --- updated-dependencies: - dependency-name: sysinfo dependency-type: direct:production update-type: version-update:semver-patch ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
When stress testing the patches in #726, the following appeared in dmesg:
[ 1773.498133] z_wr_int/11: page allocation failure: order:0, mode:0x4020
[ 1773.498140] Pid: 3846, comm: z_wr_int/11 Tainted: P O 3.2.17 #1
[ 1773.498143] Call Trace:
[ 1773.498146] [] ? warn_alloc_failed+0xfc/0x150
[ 1773.498162] [] ? __alloc_pages_nodemask+0x580/0x7f0
[ 1773.498169] [] ? new_slab+0x209/0x220
[ 1773.498174] [] ? __slab_alloc.clone.64+0x204/0x290
[ 1773.498178] [] ? __netdev_alloc_skb+0x17/0x40
[ 1773.498184] [] ? __wake_up_common+0x50/0x80
[ 1773.498188] [] ? __kmalloc_track_caller+0x10b/0x160
[ 1773.498191] [] ? __alloc_skb+0x6e/0x170
[ 1773.498195] [] ? __netdev_alloc_skb+0x17/0x40
[ 1773.498208] [] ? e1000_alloc_rx_buffers+0x1fa/0x2e0 [e1000e]
[ 1773.498216] [] ? e1000_clean_rx_irq+0x28a/0x420 [e1000e]
[ 1773.498223] [] ? e1000_clean+0xb0/0x2a0 [e1000e]
[ 1773.498228] [] ? net_rx_action+0x79/0x110
[ 1773.498233] [] ? __do_softirq+0x90/0x110
[ 1773.498239] [] ? call_softirq+0x1c/0x30
[ 1773.498243] [] ? do_softirq+0x4d/0x80
[ 1773.498247] [] ? irq_exit+0x96/0xb0
[ 1773.498250] [] ? do_IRQ+0x5c/0xd0
[ 1773.498254] [] ? common_interrupt+0x6b/0x6b
[ 1773.498256] [] ? __mutex_lock_slowpath+0x1da/0x2a0
[ 1773.498282] [] ? dsl_dir_transfer_space+0x67/0x100 [zfs]
[ 1773.498286] [] ? mutex_lock+0x9/0x20
[ 1773.498297] [] ? arc_released+0x28/0x80 [zfs]
[ 1773.498309] [] ? dbuf_rele_and_unlock+0x46e/0x680 [zfs]
[ 1773.498319] [] ? arc_space_return+0x17c8/0x2de0 [zfs]
[ 1773.498331] [] ? zio_worst_error+0x290/0x1040 [zfs]
[ 1773.498338] [] ? kmem_free_debug+0x41/0x220 [spl]
[ 1773.498348] [] ? zio_worst_error+0x730/0x1040 [zfs]
[ 1773.498354] [] ? kmem_free_debug+0x41/0x220 [spl]
[ 1773.498369] [] ? vdev_queue_io_done+0x638/0x3760 [zfs]
[ 1773.498379] [] ? zio_worst_error+0x730/0x1040 [zfs]
[ 1773.498390] [] ? zio_execute+0x95/0x2a0 [zfs]
[ 1773.498404] [] ? vdev_queue_offset_compare+0x30b/0x5c0 [zfs]
[ 1773.498408] [] ? __mutex_lock_slowpath+0x1e2/0x2a0
[ 1773.498423] [] ? vdev_queue_io_done+0xa7/0x3760 [zfs]
[ 1773.498434] [] ? zio_buf_free+0x418/0xa90 [zfs]
[ 1773.498444] [] ? zio_execute+0x95/0x2a0 [zfs]
[ 1773.498450] [] ? __taskq_dispatch+0x7e0/0xb30 [spl]
[ 1773.498454] [] ? __schedule+0x29e/0x740
[ 1773.498460] [] ? try_to_wake_up+0x270/0x270
[ 1773.498465] [] ? __taskq_dispatch+0x580/0xb30 [spl]
[ 1773.498471] [] ? __taskq_dispatch+0x580/0xb30 [spl]
[ 1773.498477] [] ? kthread+0x96/0xa0
[ 1773.498481] [] ? kernel_thread_helper+0x4/0x10
[ 1773.498486] [] ? kthread_worker_fn+0x180/0x180
[ 1773.498490] [] ? gs_change+0xb/0xb
[ 1773.498492] Mem-Info:
[ 1773.498495] DMA per-cpu:
[ 1773.498498] CPU 0: hi: 0, btch: 1 usd: 0
[ 1773.498500] CPU 1: hi: 0, btch: 1 usd: 0
[ 1773.498503] CPU 2: hi: 0, btch: 1 usd: 0
[ 1773.498505] CPU 3: hi: 0, btch: 1 usd: 0
[ 1773.498508] CPU 4: hi: 0, btch: 1 usd: 0
[ 1773.498510] CPU 5: hi: 0, btch: 1 usd: 0
[ 1773.498512] DMA32 per-cpu:
[ 1773.498515] CPU 0: hi: 186, btch: 31 usd: 13
[ 1773.498517] CPU 1: hi: 186, btch: 31 usd: 24
[ 1773.498520] CPU 2: hi: 186, btch: 31 usd: 19
[ 1773.498522] CPU 3: hi: 186, btch: 31 usd: 14
[ 1773.498525] CPU 4: hi: 186, btch: 31 usd: 18
[ 1773.498527] CPU 5: hi: 186, btch: 31 usd: 0
[ 1773.498529] Normal per-cpu:
[ 1773.498531] CPU 0: hi: 186, btch: 31 usd: 0
[ 1773.498533] CPU 1: hi: 186, btch: 31 usd: 0
[ 1773.498535] CPU 2: hi: 186, btch: 31 usd: 0
[ 1773.498538] CPU 3: hi: 186, btch: 31 usd: 4
[ 1773.498540] CPU 4: hi: 186, btch: 31 usd: 2
[ 1773.498542] CPU 5: hi: 186, btch: 31 usd: 3
[ 1773.498548] active_anon:1831687 inactive_anon:213435 isolated_anon:0
[ 1773.498549] active_file:113 inactive_file:914 isolated_file:0
[ 1773.498551] unevictable:547 dirty:0 writeback:0 unstable:0
[ 1773.498552] free:14153 slab_reclaimable:17162 slab_unreclaimable:59568
[ 1773.498553] mapped:910 shmem:2 pagetables:5903 bounce:0
[ 1773.498560] DMA free:15896kB min:64kB low:80kB high:96kB active_anon:0kB inactive_anon:0kB active_file:0kB inactive_file:0kB unevictable:0kB isolated(anon):0kB isolated(file):0kB present:15640kB mlocked:0kB dirty:0kB writeback:0kB mapped:0kB shmem:0kB slab_reclaimable:0kB slab_unreclaimable:0kB kernel_stack:0kB pagetables:0kB unstable:0kB bounce:0kB writeback_tmp:0kB pages_scanned:0 all_unreclaimable? yes
[ 1773.498567] lowmem_reserve[]: 0 3118 15214 15214
[ 1773.498576] DMA32 free:40564kB min:13836kB low:17292kB high:20752kB active_anon:557436kB inactive_anon:156204kB active_file:0kB inactive_file:0kB unevictable:0kB isolated(anon):0kB isolated(file):0kB present:3193664kB mlocked:0kB dirty:0kB writeback:0kB mapped:4kB shmem:0kB slab_reclaimable:28556kB slab_unreclaimable:93192kB kernel_stack:144kB pagetables:3212kB unstable:0kB bounce:0kB writeback_tmp:0kB pages_scanned:2 all_unreclaimable? no
[ 1773.498583] lowmem_reserve[]: 0 0 12096 12096
[ 1773.498592] Normal free:152kB min:53676kB low:67092kB high:80512kB active_anon:6769312kB inactive_anon:697536kB active_file:452kB inactive_file:3656kB unevictable:2188kB isolated(anon):0kB isolated(file):0kB present:12386304kB mlocked:2188kB dirty:0kB writeback:0kB mapped:3636kB shmem:8kB slab_reclaimable:40092kB slab_unreclaimable:145080kB kernel_stack:2712kB pagetables:20400kB unstable:0kB bounce:0kB writeback_tmp:0kB pages_scanned:42 all_unreclaimable? no
[ 1773.498599] lowmem_reserve[]: 0 0 0 0
[ 1773.498602] DMA: 0_4kB 1_8kB 1_16kB 0_32kB 2_64kB 1_128kB 1_256kB 0_512kB 1_1024kB 1_2048kB 3_4096kB = 15896kB
[ 1773.498611] DMA32: 413_4kB 267_8kB 151_16kB 73_32kB 32_64kB 8_128kB 13_256kB 18_512kB 11_1024kB 0_2048kB 1_4096kB = 39516kB
[ 1773.498620] Normal: 0_4kB 0_8kB 0_16kB 0_32kB 0_64kB 0_128kB 0_256kB 0_512kB 0_1024kB 0_2048kB 0*4096kB = 0kB
[ 1773.498629] 23224 total pagecache pages
[ 1773.498631] 21711 pages in swap cache
[ 1773.498633] Swap cache stats: add 406690, delete 384979, find 25632/30547
[ 1773.498635] Free swap = 15335732kB
[ 1773.498637] Total swap = 16777212kB
[ 1773.562577] 3964912 pages RAM
[ 1773.562577] 102929 pages reserved
[ 1773.562577] 29333 pages shared
[ 1773.562577] 3318044 pages non-shared
[ 1773.562577] SLUB: Unable to allocate memory on node -1 (gfp=0x20)
[ 1773.562577] cache: kmalloc-2048, object size: 2048, buffer size: 2048, default order: 3, min order: 0
[ 1773.562577] node 0: slabs: 131, objs: 2096, free: 0
This is Linux 3.2.17 with CONFIG_PREEMPT_NONE=y set. I checked /proc/config.gz to be certain that preemption should be disabled. This should not be able to occur. There also appears to be a bug in e1000, as there is no reason why its page allocation should fail, but at least it gave us a backtrace so we can investigate why interrupts are on.
The text was updated successfully, but these errors were encountered: