I tried running the debian 4.7.5 kernel on our parisc debian buildd servers, but when it reaches some load I usually see after some time this kernel crash: [ 2142.421978] ------------[ cut here ]------------ [ 2142.422194] kernel BUG at /build/linux/linux-4.7.5/mm/vmscan.c:1395! [ 2142.422714] CPU: 0 PID: 47 Comm: kswapd1 Not tainted 4.7.0-1-parisc64-smp #1 Debian 4.7.5-1 [ 2142.423047] task: 00000000bb2761b0 ti: 00000000bb27c000 task.ti: 00000000bb27c000 [ 2142.431931] [ 2142.432001] YZrvWESTHLNXBCVMcbcbcbcbOGFRQPDI [ 2142.442117] PSW: 00001000000001001111111100001110 Not tainted [ 2142.442329] r00-03 000000ff0804ff0e 0000000040cef520 000000004041ad6c 00000000bb27c530 [ 2142.451906] r04-07 0000000040cd1520 00000000bf2ac010 00000000bb27c5f0 0000000000000001 [ 2142.461626] r08-11 00000000bf2ac000 0000000000000010 0000000000000020 0000000000000001 [ 2142.493204] r12-15 0000000000000000 0000000000000001 00000000bb27c3f8 00000000bb27c410 [ 2142.634877] r16-19 00000000bb27c254 0000000041a5f260 00000000bb27c280 0000000000000000 [ 2142.635189] r20-23 0000000000000001 0000000000000000 00000000bb27c254 00000000bb27c410 [ 2142.778062] r24-27 00000000bb27c620 000000004041aa38 000000004041ad6c 0000000040cd1520 [ 2142.922203] r28-31 ffffffffffffffea 00000000bb27c690 00000000bb27c620 0000000000040048 [ 2143.065747] sr00-03 000000000325d000 000000000325d000 0000000000000000 000000000325d000 [ 2143.209974] sr04-07 0000000000000000 0000000000000000 0000000000000000 0000000000000000 [ 2143.210245] [ 2143.353754] IASQ: 0000000000000000 0000000000000000 IAOQ: 000000004041ae58 000000004041ae5c [ 2143.354040] IIR: 03ffe01f ISR: 0000000010340000 IOR: 000001ec9f27c6f0 [ 2143.497716] CPU: 0 CR30: 00000000bb27c000 CR31: ffffff7fffffffff [ 2143.642443] ORIG_R28: 00000000bb27c830 [ 2143.642625] IAOQ[0]: isolate_lru_pages.isra.20+0x1d8/0x2c0 [ 2143.786212] IAOQ[1]: isolate_lru_pages.isra.20+0x1dc/0x2c0 [ 2143.786408] RP(r2): isolate_lru_pages.isra.20+0xec/0x2c0 [ 2143.930262] Backtrace: [ 2143.930371] [<000000004041dddc>] shrink_active_list+0x184/0x720 [ 2144.073396] [<0000000040420938>] kswapd+0x6d8/0x1008 [ 2144.073600] [<00000000402c9864>] __wake_up_locked+0x84/0xa8 [ 2144.217657] [<000000004028fa84>] kthread+0x1ec/0x248 [ 2144.217848] [<0000000040205020>] end_fault_vector+0x20/0xc0 [ 2144.360930] [<000000004021e494>] execute_on_irq_stack+0x54/0x108 [ 2144.361148] [<0000000040300a98>] rcu_irq_exit+0x40/0xa0 [ 2144.505159] [<000000004025d1d8>] irq_exit+0x108/0x120 [ 2144.505345] [<000000004021faa4>] do_cpu_irq_mask+0x2bc/0x4a8 [ 2144.648511] [<0000000040206074>] intr_return+0x0/0xc [ 2144.648700] [<00000000402d9e50>] cpuacct_charge+0x50/0xe8 [ 2144.793737] [<00000000402b4fa0>] check_preempt_wakeup+0x260/0x2d0 [ 2144.793962] [<00000000402a3390>] check_preempt_curr+0x120/0x128 [ 2144.936497] [<00000000402b47a4>] update_curr+0x18c/0x2d8 [ 2145.080725] [<00000000402d9e50>] cpuacct_charge+0x50/0xe8 [ 2145.080922] [<00000000402b47a4>] update_curr+0x18c/0x2d8 [ 2145.225161] [ 2145.225240] CPU: 0 PID: 47 Comm: kswapd1 Not tainted 4.7.0-1-parisc64-smp #1 Debian 4.7.5-1 [ 2145.229139] Backtrace: [ 2145.229139] [<000000004021c738>] show_stack+0x68/0x80 [ 2145.229139] [<00000000406e937c>] dump_stack+0xec/0x168 [ 2145.229139] [<000000004021ca04>] die_if_kernel+0x25c/0x430 [ 2145.229139] [<000000004021dd90>] handle_interruption+0xc98/0xcc0 [ 2145.229139] [ 2145.657495] ---[ end trace bf38fbc7135d5765 ]--- Once I could capture this warning as well (which crashed the machine later on): [ 559.048134] list_add corruption. next->prev should be prev (00000000ba709020), but was 0000000042c98860. (next=0000000043604ee0). [ 559.048999] ------------[ cut here ]------------ [ 559.049227] WARNING: CPU: 2 PID: 17367 at /build/linux/linux-4.7.5/lib/list_debug.c:29 __list_add+0x11c/0x128 [ 559.058096] Modules linked in: autofs4 nfsd auth_rpcgss oid_registry nfs_acl nfs lockd grace fscache sunrpc sg ext4 ecb crc16 jbd2 crc32c_generic mbcache sr_mod cdrom sd_mod sym53c8xx scsi_transport_spi scsi_mod tulip [ 559.077491] CPU: 2 PID: 17367 Comm: cp Not tainted 4.7.0-1-parisc64-smp #2 Debian 4.7.5-1 [ 559.087139] task: 00000000bb06e370 ti: 00000000b8cb0000 task.ti: 00000000b8cb0000 [ 559.120020] [ 559.120084] YZrvWESTHLNXBCVMcbcbcbcbOGFRQPDI [ 559.120254] PSW: 00001000000001001111100000001110 Not tainted [ 559.260535] r00-03 000000ff0804f80e 0000000040d42520 000000004071a118 00000000b8cb0d80 [ 559.404336] r04-07 0000000040cd1520 00000000420b2e60 00000000ba709020 0000000043604ee0 [ 559.404647] r08-11 0000000000000002 0000000000000001 0000000000000000 0000000040c4d940 [ 559.547814] r12-15 0000000000000000 00000000420b2e40 000000000800000f 0000000000000000 [ 559.691424] r16-19 0000000000000000 0000000000000000 00000000b8cb0650 000000000800000e [ 559.835467] r20-23 0000000000000001 000000000800000e 0000000000000174 0000000000000174 [ 559.979173] r24-27 00000000b8cb1370 00000000406fdaa8 000000004070be24 0000000040cd1520 [ 559.979457] r28-31 0000000000000075 00000000b8cb13e0 00000000b8cb0e20 0000000000000075 [ 560.121995] sr00-03 0000000003e8c800 0000000000000000 0000000000000000 0000000003e8c800 [ 560.265947] sr04-07 0000000000000000 0000000000000000 0000000000000000 0000000000000000 [ 560.410434] [ 560.410503] IASQ: 0000000000000000 0000000000000000 IAOQ: 000000004071a11c 000000004071a120 [ 560.554676] IIR: 03ffe01f ISR: 0000000010240000 IOR: 00000003d47518d0 [ 560.554917] CPU: 2 CR30: 00000000b8cb0000 CR31: fffffffffff3ffff [ 560.697554] ORIG_R28: 0000000040cf4520 [ 560.697701] IAOQ[0]: __list_add+0x11c/0x128 [ 560.842128] IAOQ[1]: __list_add+0x120/0x128 [ 560.842285] RP(r2): __list_add+0x118/0x128 [ 560.986271] Backtrace: [ 560.986388] [<000000004040fee0>] __pagevec_lru_add_fn+0x1c8/0x2c8 [ 561.130001] [<000000004041076c>] pagevec_lru_move_fn+0x1c4/0x288 [ 561.130221] [<0000000040410aa8>] __lru_cache_add+0x118/0x158 [ 561.274461] [<00000000404120ac>] lru_cache_add+0x64/0x80 [ 561.274664] [<00000000403ef490>] add_to_page_cache_lru+0x148/0x250 [ 561.417955] [<00000000403f114c>] pagecache_get_page+0x2a4/0x600 [ 561.418174] [<00000000403f2a40>] grab_cache_page_write_begin+0x90/0xe0 [ 561.564306] [<0000000028984080>] ext4_da_write_begin+0x130/0x7e0 [ext4] [ 561.705674] [<00000000403f2e28>] generic_perform_write+0x158/0x370 [ 561.705905] [<00000000403f4ac8>] __generic_file_write_iter+0x260/0x2d8 [ 561.849151] [<000000002896fdf4>] ext4_file_write_iter+0x10c/0x548 [ext4] [ 561.992629] [<00000000404ba758>] new_sync_write+0x168/0x1d0 [ 561.992839] [<00000000404ba8a4>] __vfs_write+0xe4/0xf0 [ 562.136521] [<00000000404bc05c>] vfs_write+0xfc/0x278 [ 562.136710] [<00000000404be4f4>] SyS_write+0xcc/0x170 [ 562.279474] [<0000000040207024>] syscall_exit+0x0/0x14 [ 562.279655] [ 562.279721] ---[ end trace bce15645442c2913 ]--- I see there was one commit pushed into the stable tree: mm: workingset: fix crash in shadow node shrinker caused by replace_page_cache_page() http://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=22f2ac51b6d643666f4db093f13144f773ff3f3a Might this patch help to fix the issues as well? If not, maybe someone has an idea based on the traces above? I don't think it's triggered due to parisc specific changes... Thanks, Helge -- To unsubscribe from this list: send the line "unsubscribe linux-parisc" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
