From: Peter Zijlstra <peterz@infradead.org>
To: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Thomas Gleixner <tglx@linutronix.de>,
Ingo Molnar <mingo@kernel.org>,
Martin Schwidefsky <schwidefsky@de.ibm.com>,
linux-kernel@vger.kernel.org, linux-s390@vger.kernel.org,
Stefan Liebler <stli@linux.ibm.com>
Subject: Re: WARN_ON_ONCE(!new_owner) within wake_futex_pi() triggered
Date: Mon, 28 Jan 2019 14:44:10 +0100 [thread overview]
Message-ID: <20190128134410.GA28485@hirez.programming.kicks-ass.net> (raw)
In-Reply-To: <20181129112321.GB3449@osiris>
On Thu, Nov 29, 2018 at 12:23:21PM +0100, Heiko Carstens wrote:
> And indeed, if I run only this test case in an endless loop and do
> some parallel work (like kernel compile) it currently seems to be
> possible to reproduce the warning:
>
> while true; do time ./testrun.sh nptl/tst-robustpi8 --direct ; done
>
> within the build directory of glibc (2.28).
Right; so that reproduces for me.
After staring at all that for a while; trying to remember how it all
worked (or supposed to work rather), I became suspiscous of commit:
56222b212e8e ("futex: Drop hb->lock before enqueueing on the rtmutex")
And indeed, when I revert that; the above reproducer no longer works (as
in, it no longer triggers in minutes and has -- so far -- held up for an
hour+ or so).
That patch in particular allows futex_unlock_pi() to 'start' early:
futex_lock_pi() futex_unlock_pi()
lock hb
queue
lock wait_lock
unlock hb
lock hb
futex_top_waiter
get_pi_state
lock wait_lock
rt_mutex_proxy_start // fail
unlock wait_lock
// acquired wait_lock
wake_futex_pi()
rt_mutex_next_owner() // whoops, no waiter
WARN
lock hb
unqueue_me_pi
So reverting that patch should cure things, because then there is no hb
lock break between queue/unqueue and futex_unlock_pi() cannot observe
this half-arsed state.
Now obviously reverting that makes RT unhappy; let me see what the
options are.
(concurrently tglx generated a trace that corroborates)
next prev parent reply other threads:[~2019-01-28 13:44 UTC|newest]
Thread overview: 47+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-11-27 8:11 WARN_ON_ONCE(!new_owner) within wake_futex_pi() triggered Heiko Carstens
2018-11-28 14:32 ` Thomas Gleixner
2018-11-29 11:23 ` Heiko Carstens
2019-01-21 12:21 ` Heiko Carstens
2019-01-21 13:12 ` Thomas Gleixner
2019-01-22 21:14 ` Thomas Gleixner
2019-01-23 9:24 ` Heiko Carstens
2019-01-23 12:33 ` Thomas Gleixner
2019-01-23 12:40 ` Heiko Carstens
2019-01-28 13:44 ` Peter Zijlstra [this message]
2019-01-28 13:58 ` Peter Zijlstra
2019-01-28 15:53 ` Thomas Gleixner
2019-01-29 8:49 ` Peter Zijlstra
2019-01-29 22:15 ` [PATCH] futex: Handle early deadlock return correctly Thomas Gleixner
2019-01-30 12:01 ` Thomas Gleixner
2019-02-08 12:05 ` [tip:locking/urgent] " tip-bot for Thomas Gleixner
2019-01-29 9:01 ` WARN_ON_ONCE(!new_owner) within wake_futex_pi() triggered Heiko Carstens
2019-01-29 9:33 ` Peter Zijlstra
2019-01-29 9:45 ` Thomas Gleixner
2019-01-29 10:24 ` Heiko Carstens
2019-01-29 10:35 ` Peter Zijlstra
2019-01-29 13:03 ` Thomas Gleixner
2019-01-29 13:23 ` Heiko Carstens
[not found] ` <20190129151058.GG26906@osiris>
2019-01-29 17:16 ` Sebastian Sewior
2019-01-29 21:45 ` Thomas Gleixner
[not found] ` <20190130094913.GC5299@osiris>
2019-01-30 12:15 ` Thomas Gleixner
[not found] ` <20190130125955.GD5299@osiris>
2019-01-30 13:24 ` Sebastian Sewior
2019-01-30 13:29 ` Thomas Gleixner
2019-01-30 14:33 ` Thomas Gleixner
2019-01-30 17:56 ` Thomas Gleixner
2019-01-30 21:07 ` Sebastian Sewior
2019-01-30 23:13 ` WARN_ON_ONCE(!new_owner) within wake_futex_pi() triggerede Thomas Gleixner
2019-01-30 23:35 ` Paul E. McKenney
2019-01-30 23:55 ` Thomas Gleixner
2019-01-31 0:27 ` Thomas Gleixner
2019-01-31 1:45 ` Paul E. McKenney
2019-01-31 16:52 ` Heiko Carstens
2019-01-31 17:06 ` Sebastian Sewior
2019-01-31 20:42 ` Heiko Carstens
2019-02-01 16:12 ` Heiko Carstens
2019-02-01 21:59 ` Thomas Gleixner
[not found] ` <20190202091043.GA3381@osiris>
2019-02-02 10:14 ` Thomas Gleixner
2019-02-02 11:20 ` Heiko Carstens
2019-02-03 16:30 ` Thomas Gleixner
2019-02-04 11:40 ` Heiko Carstens
2019-01-31 1:44 ` Paul E. McKenney
2019-01-30 13:25 ` WARN_ON_ONCE(!new_owner) within wake_futex_pi() triggered Thomas Gleixner
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20190128134410.GA28485@hirez.programming.kicks-ass.net \
--to=peterz@infradead.org \
--cc=heiko.carstens@de.ibm.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-s390@vger.kernel.org \
--cc=mingo@kernel.org \
--cc=schwidefsky@de.ibm.com \
--cc=stli@linux.ibm.com \
--cc=tglx@linutronix.de \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).