On 06/26/2017 09:53 PM, Fengguang Wu wrote: > Greetings, > > Sorry the report is a bit late, this bug looks still active: > First I've heard about it! > [ 36.037364] sector 2, nr/cnr 0/2 > [ 36.037367] bio ffff88001af84e80, biotail ffff88001af84e80, len 0 > [ 36.037367] bio ffff88001af84e80, biotail ffff88001af84e80, len 0 > [ 36.037424] ------------[ cut here ]------------ > [ 36.037424] ------------[ cut here ]------------ > [ 36.037428] kernel BUG at block/blk-mq.c:415! > [ 36.037428] kernel BUG at block/blk-mq.c:415! > [ 36.037433] invalid opcode: 0000 [#1] PREEMPT SMP > [ 36.037433] invalid opcode: 0000 [#1] PREEMPT SMP > [ 36.037434] Modules linked in: > [ 36.037434] Modules linked in: Do you have the full dmesg of this? I looked at the one attached to the bug report, but that's a completely different bug (NULL deref in drm). Are you sure it's still the above blk-mq BUG() we're hitting? -- Jens Axboe