linux-kernel - Re: [RT PATCH 1/2] timers: wakeup all timer waiters

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: <e014d4ad-316f-3afe-fcac-923cd2f9a7e6@linutronix.de>
Date:	Thu, 14 Jul 2016 19:14:58 +0200
From:	Sebastian Andrzej Siewior <bigeasy@...utronix.de>
To:	Steven Rostedt <rostedt@...dmis.org>
Cc:	linux-rt-users@...r.kernel.org, linux-kernel@...r.kernel.org,
	tglx@...utronix.de
Subject: Re: [RT PATCH 1/2] timers: wakeup all timer waiters

On 07/14/2016 06:13 PM, Steven Rostedt wrote:
>>  
>> -# define wakeup_timer_waiters(b)	wake_up(&(b)->wait_for_running_timer)
>> +# define wakeup_timer_waiters(b)	wake_up_all(&(b)->wait_for_running_timer)
> 
> OK, I just received this patch (way after patch 2)
> 
> I'm assuming that patch two was done such that you don't do a
> "wake_up_all" under a spinlock.

No. I pulled in new timer code in and had to redo this part of RT.

While doing so I noticed that we drop the base lock during timer
invocations and so it could be possible that we have two invocations
of del_timer_sync() on a timer on the same "base" (one after the
other). This is patch #1.

After that I saw that we do the wake up under the base lock but there
is no reason for it. So here is patch #2.

Patch #1 is something that could happen in theory and I did not run in
any problem.

Sebastian