On Wed, Aug 10, 2022, at 9:32 PM, Qu Wenruo wrote: > On 2022/8/11 00:19, Chris Murphy wrote: >> Downstream bug - 5.19.0: dnf install hangs when system is under load >> https://bugzilla.redhat.com/show_bug.cgi?id=2117326 >> >> 5.19.0-65.fc37.x86_64 >> >> Setup >> btrfs raid10 on 8x plain partitions >> >> Command >> sudo dnf install pciutils >> >> Reproducible: >> About 1 in 3, correlates with the system being under heavy load, otherwise it's not happening >> >> Get stuck at >> Running scriptlet: sg3_utils-1.46-3.fc36.x86_64 2/2 >> >> ps aux status for dnf is D+, kill -9 does nothing, strace shows nothing. The hang last at least 10 minutes, didn't test beyond that. >> >> Full dmesg with sysrq+w is attached to the bug report. >> >> snippet >> >> [ 2268.057017] sysrq: Show Blocked State >> [ 2268.057866] task:kworker/u97:11 state:D stack: 0 pid: 340 ppid: 2 flags:0x00004000 >> [ 2268.058361] Workqueue: writeback wb_workfn (flush-btrfs-1) >> [ 2268.058825] Call Trace: >> [ 2268.059261] <TASK> >> [ 2268.059692] __schedule+0x335/0x1240 >> [ 2268.060145] ? __blk_mq_sched_dispatch_requests+0xe0/0x130 >> [ 2268.060611] schedule+0x4e/0xb0 >> [ 2268.061059] io_schedule+0x42/0x70 >> [ 2268.061473] blk_mq_get_tag+0x10c/0x290 > > All the hanging processes are waiting at blk_mq_get_tag(), thus I'm not > sure if it's really btrfs, or something in the block layer. > > Adding block layer guys into the thread. OK so it might just be the same problem I reported in this thread, which first appeared in the 5.12 merge window. The weird thing is, the 5.19 kernel is staying up for *days* unlike 5.12 through 5.18, except under heavy load I run dnf and then only dnf hangs. https://lore.kernel.org/linux-btrfs/ad78a32c-7790-4e21-be9f-81c5848a4953@xxxxxxxxxxxxxxxx/T/#t -- Chris Murphy
