linux-kernel - Re: [PATCH v2] hung_task: Dump blocker task if it is not hung

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: <d63b51f2-7c63-49ff-8dfe-f92a60db7c19@linux.dev>
Date: Thu, 31 Jul 2025 12:45:17 +0800
From: Lance Yang <lance.yang@...ux.dev>
To: Sergey Senozhatsky <senozhatsky@...omium.org>
Cc: "Masami Hiramatsu (Google)" <mhiramat@...nel.org>,
 Andrew Morton <akpm@...ux-foundation.org>,
 Peter Zijlstra <peterz@...radead.org>, Ingo Molnar <mingo@...hat.com>,
 Will Deacon <will@...nel.org>, Boqun Feng <boqun.feng@...il.com>,
 Waiman Long <longman@...hat.com>, Joel Granados <joel.granados@...nel.org>,
 Anna Schumaker <anna.schumaker@...cle.com>, Lance Yang
 <ioworker0@...il.com>, Kent Overstreet <kent.overstreet@...ux.dev>,
 Yongliang Gao <leonylgao@...cent.com>, Steven Rostedt <rostedt@...dmis.org>,
 Tomasz Figa <tfiga@...omium.org>, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2] hung_task: Dump blocker task if it is not hung



On 2025/7/31 11:37, Sergey Senozhatsky wrote:
> On (25/07/31 11:25), Lance Yang wrote:
>>>> Wrote a simple 3 tasks circular lock test (same that I had in real
>>>> life).  The output looks good:
>>>
>>> Nice work! The log is now much clearer ;)
>>
>> Forgot to add:
>>
>> This circular lock test is great. It would be very valuable to have
>> it in samples/hung_task ;)
>>
>> Would you be willing to submit it as a patch?
> 
> Well, that particular test is a deadlock, not sure how useful a test
> that deadlocks is going to be.  We probably need some alternative if
> we want to test backtrace de-dup.

Yep, you're right. A permanent deadlock is a no-go for a reusable test.

Maybe we need a test like a blocker chain (T1->T2->T3), where T3 holds
the lock and just sleeps for a while. Once it wakes up and releases the
lock, everything can finish cleanly.

Also, I might try to cook up a sample for that, time permitting ;)