Bug 215527 - Random freeze in bfq_put_idle_entity / bfq_exit_icq_bfqq
Summary: Random freeze in bfq_put_idle_entity / bfq_exit_icq_bfqq
Status: RESOLVED CODE_FIX
Alias: None
Product: IO/Storage
Classification: Unclassified
Component: Block Layer (show other bugs)
Hardware: All Linux
: P1 normal
Assignee: Paolo Valente
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2022-01-24 13:22 UTC by Bernd Buschinski
Modified: 2022-02-18 13:54 UTC (History)
1 user (show)

See Also:
Kernel Version: 5.15.15
Tree: Mainline
Regression: No


Attachments

Description Bernd Buschinski 2022-01-24 13:22:42 UTC
With kernel 5.15.15 (possible all 5.15*, don't know when it really started) I get random freezes and a system log like:


Jan 24 13:10:03 localhost kernel: PGD 0 P4D 0 
Jan 24 13:10:03 localhost kernel: Oops: 0002 [#1] SMP NOPTI
Jan 24 13:10:03 localhost kernel: CPU: 3 PID: 2028 Comm: kworker/3:1H Not tainted 5.15.15-gentoo #1
Jan 24 13:10:03 localhost kernel: Hardware name: Dell Inc. Latitude 7410/0M5G57, BIOS 1.11.0 12/12/2021
Jan 24 13:10:03 localhost kernel: Workqueue: kblockd blk_mq_run_work_fn
Jan 24 13:10:03 localhost kernel: RIP: 0010:bfq_idle_extract+0x50/0xa0
Jan 24 13:10:03 localhost kernel: Code: 65 48 c7 45 30 00 00 00 00 48 8d 73 08 48 89 ef e8 85 39 08 00 4d 85 e4 74 35 49 8b 84 24 38 01 00 00 49 8b 94 24 30 01 00 00 <48> 89 42 08 48 89 10 48 b8 00 01 00 00 00 00 ad de 49 89 84 24 30
Jan 24 13:10:03 localhost kernel: RSP: 0018:ffffc9000134fce0 EFLAGS: 00010082
Jan 24 13:10:03 localhost kernel: RAX: 0000000000000000 RBX: ffff8881dd007158 RCX: 0000000000000000
Jan 24 13:10:03 localhost kernel: RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffff8881df2d8098
Jan 24 13:10:03 localhost kernel: RBP: ffff8881df2d8098 R08: ffff8881dd007160 R09: ffff8881f8155280
Jan 24 13:10:03 localhost kernel: R10: ffff8881f8155280 R11: fefefefefefefeff R12: ffff8881df2d8010
Jan 24 13:10:03 localhost kernel: R13: ffff888109736760 R14: ffff888109736760 R15: 000000000000001e
Jan 24 13:10:03 localhost kernel: FS:  0000000000000000(0000) GS:ffff88848d6c0000(0000) knlGS:0000000000000000
Jan 24 13:10:03 localhost kernel: CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
Jan 24 13:10:03 localhost kernel: CR2: 0000000000000008 CR3: 00000001174be001 CR4: 00000000003706e0
Jan 24 13:10:03 localhost kernel: DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
Jan 24 13:10:03 localhost kernel: DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Jan 24 13:10:03 localhost kernel: Call Trace:
Jan 24 13:10:03 localhost kernel:  <TASK>
Jan 24 13:10:03 localhost kernel:  bfq_put_idle_entity+0xd/0x50
Jan 24 13:10:03 localhost kernel:  bfq_bfqq_served+0xaa/0x1b0
Jan 24 13:10:03 localhost kernel:  bfq_dispatch_request+0x2c9/0x1270
Jan 24 13:10:03 localhost kernel:  ? mod_delayed_work_on+0x42/0x60
Jan 24 13:10:03 localhost kernel:  __blk_mq_do_dispatch_sched+0x1c8/0x2a0
Jan 24 13:10:03 localhost kernel:  __blk_mq_sched_dispatch_requests+0xcd/0x130
Jan 24 13:10:03 localhost kernel:  blk_mq_sched_dispatch_requests+0x2a/0x50
Jan 24 13:10:03 localhost kernel:  __blk_mq_run_hw_queue+0x28/0x60
Jan 24 13:10:03 localhost kernel:  process_one_work+0x1d1/0x370
Jan 24 13:10:03 localhost kernel:  worker_thread+0x48/0x3c0
Jan 24 13:10:03 localhost kernel:  ? rescuer_thread+0x350/0x350
Jan 24 13:10:03 localhost kernel:  kthread+0x11f/0x140
Jan 24 13:10:03 localhost kernel:  ? set_kthread_struct+0x40/0x40
Jan 24 13:10:03 localhost kernel:  ret_from_fork+0x1f/0x30
Jan 24 13:10:03 localhost kernel:  </TASK>
Jan 24 13:10:03 localhost kernel: Modules linked in: xt_MASQUERADE xt_addrtype iptable_nat br_netfilter snd_hda_codec_hdmi snd_ctl_led snd_hda_codec_realtek snd_hda_codec_generic snd_hda_intel snd_intel_dspcfg snd_hda_codec snd_hwdep snd_hda_core snd_pcm snd_timer snd efivarfs
Jan 24 13:10:03 localhost kernel: CR2: 0000000000000008
Jan 24 13:10:03 localhost kernel: ---[ end trace 15e78f85422205d7 ]---
Jan 24 13:10:03 localhost kernel: RIP: 0010:bfq_idle_extract+0x50/0xa0
Jan 24 13:10:03 localhost kernel: Code: 65 48 c7 45 30 00 00 00 00 48 8d 73 08 48 89 ef e8 85 39 08 00 4d 85 e4 74 35 49 8b 84 24 38 01 00 00 49 8b 94 24 30 01 00 00 <48> 89 42 08 48 89 10 48 b8 00 01 00 00 00 00 ad de 49 89 84 24 30
Jan 24 13:10:03 localhost kernel: RSP: 0018:ffffc9000134fce0 EFLAGS: 00010082
Jan 24 13:10:03 localhost kernel: RAX: 0000000000000000 RBX: ffff8881dd007158 RCX: 0000000000000000
Jan 24 13:10:03 localhost kernel: RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffff8881df2d8098
Jan 24 13:10:03 localhost kernel: RBP: ffff8881df2d8098 R08: ffff8881dd007160 R09: ffff8881f8155280
Jan 24 13:10:03 localhost kernel: R10: ffff8881f8155280 R11: fefefefefefefeff R12: ffff8881df2d8010
Jan 24 13:10:03 localhost kernel: R13: ffff888109736760 R14: ffff888109736760 R15: 000000000000001e
Jan 24 13:10:03 localhost kernel: FS:  0000000000000000(0000) GS:ffff88848d6c0000(0000) knlGS:0000000000000000
Jan 24 13:10:03 localhost kernel: CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
Jan 24 13:10:03 localhost kernel: CR2: 0000000000000008 CR3: 00000001174be001 CR4: 00000000003706e0
Jan 24 13:10:03 localhost kernel: DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
Jan 24 13:10:03 localhost kernel: DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Jan 24 13:10:26 localhost kernel:       (detected by 4, t=21002 jiffies, g=361417, q=1457)
Jan 24 13:10:26 localhost kernel: Sending NMI from CPU 4 to CPUs 6:
Jan 24 13:10:26 localhost kernel: NMI backtrace for cpu 6
Jan 24 13:10:26 localhost kernel: CPU: 6 PID: 10426 Comm: systemd-udevd Tainted: G      D           5.15.15-gentoo #1
Jan 24 13:10:26 localhost kernel: Hardware name: Dell Inc. Latitude 7410/0M5G57, BIOS 1.11.0 12/12/2021
Jan 24 13:10:26 localhost kernel: RIP: 0010:queued_spin_lock_slowpath+0x3d/0x190
Jan 24 13:10:26 localhost kernel: Code: 0f ba 2a 08 0f 92 c1 8b 02 0f b6 c9 c1 e1 08 30 e4 09 c8 a9 00 01 ff ff 0f 85 eb 00 00 00 85 c0 74 0e 8b 02 84 c0 74 08 f3 90 <8b> 02 84 c0 75 f8 b8 01 00 00 00 66 89 02 c3 8b 37 b8 00 02 00 00
Jan 24 13:10:26 localhost kernel: RSP: 0018:ffffc90002c17e08 EFLAGS: 00000002
Jan 24 13:10:26 localhost kernel: RAX: 0000000000000101 RBX: ffff8881078fe0e8 RCX: 0000000000000000
Jan 24 13:10:26 localhost kernel: RDX: ffff888125898c20 RSI: 0000000000000000 RDI: ffff888125898c20
Jan 24 13:10:26 localhost kernel: RBP: ffff88810eea78e0 R08: ffff8881f3801d80 R09: ffffc90002c17df0
Jan 24 13:10:26 localhost kernel: R10: 0000000000000000 R11: ffff88848d7a9c70 R12: 0000000000000086
Jan 24 13:10:26 localhost kernel: R13: ffff888125898800 R14: 0000000000000000 R15: ffff888125898c20
Jan 24 13:10:26 localhost kernel: FS:  0000000000000000(0000) GS:ffff88848d780000(0000) knlGS:0000000000000000
Jan 24 13:10:26 localhost kernel: CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
Jan 24 13:10:26 localhost kernel: CR2: 00005637e53ed010 CR3: 0000000006012006 CR4: 00000000003706e0
Jan 24 13:10:26 localhost kernel: DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
Jan 24 13:10:26 localhost kernel: DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Jan 24 13:10:26 localhost kernel: Call Trace:
Jan 24 13:10:26 localhost kernel:  <TASK>
Jan 24 13:10:26 localhost kernel:  _raw_spin_lock_irqsave+0x20/0x30
Jan 24 13:10:26 localhost kernel:  bfq_exit_icq_bfqq+0x6a/0x220
Jan 24 13:10:26 localhost kernel:  bfq_exit_icq+0x56/0x70
Jan 24 13:10:26 localhost kernel:  put_io_context_active+0x45/0x70
Jan 24 13:10:26 localhost kernel:  do_exit+0x687/0x990
Jan 24 13:10:26 localhost kernel:  do_group_exit+0x2e/0x90
Jan 24 13:10:26 localhost kernel:  __x64_sys_exit_group+0xf/0x10
Jan 24 13:10:26 localhost kernel:  do_syscall_64+0x38/0x90
Jan 24 13:10:26 localhost kernel:  entry_SYSCALL_64_after_hwframe+0x44/0xae
Jan 24 13:10:26 localhost kernel: RIP: 0033:0x7efdc46568d9
Jan 24 13:10:26 localhost kernel: Code: Unable to access opcode bytes at RIP 0x7efdc46568af.
Jan 24 13:10:26 localhost kernel: RSP: 002b:00007ffc39fd9c28 EFLAGS: 00000206 ORIG_RAX: 00000000000000e7
Jan 24 13:10:26 localhost kernel: RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007efdc46568d9
Jan 24 13:10:26 localhost kernel: RDX: 000000000000003c RSI: 00000000000000e7 RDI: 0000000000000000
Jan 24 13:10:26 localhost kernel: RBP: 0000000000000000 R08: fffffffffffffee0 R09: 00005637e5349ea0
Jan 24 13:10:26 localhost kernel: R10: 26ca40ab28001f9e R11: 0000000000000206 R12: 00005637e5410210
Jan 24 13:10:26 localhost kernel: R13: 00005637e53f6610 R14: 00005637e5450cc0 R15: 00005637e5349ea0
Jan 24 13:10:26 localhost kernel:  </TASK>


Not sure what more information I could report, as this is "pretty" random.
Please ask for more info, I will try to provide it as best as I can :)
Comment 1 The Linux kernel's regression tracker (Thorsten Leemhuis) 2022-01-26 07:44:23 UTC
(In reply to Bernd Buschinski from comment #0)
> With kernel 5.15.15 (possible all 5.15*, don't know when it really started)
> I get random freezes and a system log like:

But it used to work fine in older versions (like 5.14?)? Or has it always been there?
Comment 2 Bernd Buschinski 2022-01-26 08:12:33 UTC
(In reply to The Linux kernel's regression tracker (Thorsten Leemhuis) from comment #1)
> (In reply to Bernd Buschinski from comment #0)
> > With kernel 5.15.15 (possible all 5.15*, don't know when it really started)
> > I get random freezes and a system log like:
> 
> But it used to work fine in older versions (like 5.14?)? Or has it always
> been there?

As far as I can remember, it used to work fine with 5.14*.
BUT as this happens more or less randomly, it might be possible that I just didn't notice/trigger it.
Comment 3 The Linux kernel's regression tracker (Thorsten Leemhuis) 2022-01-26 08:45:17 UTC
(In reply to Bernd Buschinski from comment #2)

> As far as I can remember, it used to work fine with 5.14*.

thx, I'll track it as a regression then. Might not be one, but "random freezes" is something that shouldn't happen, so it might be worth tracking anyway.

> BUT as this happens more or less randomly, it might be possible that I just
> didn't notice/trigger it.

It's up to the maintainers to handle this, but two quick notes from my side:

 * a bisection (https://www.kernel.org/doc/html/latest/admin-guide/bug-bisect.html ) would be helpful, but I guess it won't be possible here?

 * you might want to give 5.17-rc1 a try, maybe you are lucky and this already was fixed in between
Comment 4 Bernd Buschinski 2022-01-26 15:52:15 UTC
(In reply to The Linux kernel's regression tracker (Thorsten Leemhuis) from comment #3)
> (In reply to Bernd Buschinski from comment #2)
> 
> > As far as I can remember, it used to work fine with 5.14*.
> 
> thx, I'll track it as a regression then. Might not be one, but "random
> freezes" is something that shouldn't happen, so it might be worth tracking
> anyway.
> 
> > BUT as this happens more or less randomly, it might be possible that I just
> > didn't notice/trigger it.
> 
> It's up to the maintainers to handle this, but two quick notes from my side:
> 
>  * a bisection
> (https://www.kernel.org/doc/html/latest/admin-guide/bug-bisect.html ) would
> be helpful, but I guess it won't be possible here?
> 
>  * you might want to give 5.17-rc1 a try, maybe you are lucky and this
> already was fixed in between

I will try 5.16.3 "soon" and if this does not help 5.17-rc*.
*if* I find a kind of reliable reproducer, I will try to bisect.

The only hint I got is: it seems to happen more frequently when building (big) docker images, but this is a 1 out of 10 situations. But sometimes didn't even happen after 20 retries, and it takes a long time.
Comment 5 Bernd Buschinski 2022-01-31 06:05:52 UTC
For the record: I tried to use 5.16.3 and 5.17-rc2, but I am blocked by another regression (https://bugzilla.kernel.org/show_bug.cgi?id=215550), meaning I can not really daily drive them, so I can't test if this issue is fixed in newer versions.
Comment 6 The Linux kernel's regression tracker (Thorsten Leemhuis) 2022-02-08 11:54:06 UTC
Paolo, what the status here?
Comment 7 The Linux kernel's regression tracker (Thorsten Leemhuis) 2022-02-18 11:28:39 UTC
Nothing happened for a while? What up here? Bernd, is the problem still happening and did you try latest mainline? Paolo, did you have a chance to look into this?
Comment 8 Bernd Buschinski 2022-02-18 13:54:39 UTC
I am currently using 5.16.8 (for a few days now), the issue did not appear again.
So I assume it was fixed in between.

If it ever reappears, or if I find a reproducer, I would reopen this ticket.

Note You need to log in before you can comment on or make changes to this bug.