netdev - Re: [PATCH net 0/4][pull request] igb: fix igb_msix_other() handling for PREEMPT

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20250219164158.qYsR4V25@linutronix.de>
Date: Wed, 19 Feb 2025 17:41:58 +0100
From: Sebastian Andrzej Siewior <bigeasy@...utronix.de>
To: Steven Rostedt <rostedt@...dmis.org>
Cc: Wander Lairson Costa <wander@...hat.com>,
	Tony Nguyen <anthony.l.nguyen@...el.com>, davem@...emloft.net,
	kuba@...nel.org, pabeni@...hat.com, edumazet@...gle.com,
	andrew+netdev@...n.ch, netdev@...r.kernel.org, clrkwllms@...nel.org,
	jgarzik@...hat.com, yuma@...hat.com, linux-rt-devel@...ts.linux.dev
Subject: Re: [PATCH net 0/4][pull request] igb: fix igb_msix_other() handling
 for PREEMPT_RT

On 2025-02-19 10:29:16 [-0500], Steven Rostedt wrote:
> On Tue, 18 Feb 2025 11:50:55 -0300
> Wander Lairson Costa <wander@...hat.com> wrote:
> 
> >      kworker/0:0-8       [000] b..13  2121.730643: e1000_write_posted_mbx <-e1000_rar_set_vf
> >      kworker/0:0-8       [000] D.Zf2  2121.730645: igbvf_reset_L14: (igbvf_reset+0x62/0x120 [igbvf])
> >      kworker/0:0-8       [000] .N...  2121.730649: igbvf_reset_L16: (igbvf_reset+0x7b/0x120 [igbvf])
> >   irq/63-ens14f0-1112    [000] b..12  2121.730652: igb_msix_other <-irq_thread_fn
> 
> The preempt count is hex, so 'f' means 15. But that that latency field looks corrupted.

It is high but it kind of makes sense. We cap it at 15 so it might be
higher. But then we would have nesting but why? What confuses me a bit
is the Z because this would indicate NMI.
Also the entry after is a N and nothing else. I would expect a
sched_switch right after unlock so there should be no further entry from
kworker which must run at RT priority because it is boosted by
irq/63-ens14f0-1112.

> Thanks,
> 
> -- Steve

Sebastian