[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <5f5fbd7.1073c.16c0446ea63.Coremail.luferry@163.com>
Date: Thu, 18 Jul 2019 16:50:52 +0800 (CST)
From: luferry <luferry@....com>
To: "Thomas Gleixner" <tglx@...utronix.de>
Cc: "Peter Zijlstra (Intel)" <peterz@...radead.org>,
"Rik van Riel" <riel@...riel.com>,
"Greg Kroah-Hartman" <gregkh@...uxfoundation.org>,
"Josh Poimboeuf" <jpoimboe@...hat.com>,
linux-kernel@...r.kernel.org
Subject: Re:Re: [PATCH v2] smp: avoid generic_exec_single cause system
lockup
At 2019-07-18 16:07:58, "Thomas Gleixner" <tglx@...utronix.de> wrote:
>On Thu, 18 Jul 2019, luferry@....com wrote:
>
>> From: luferry <luferry@....com>
>>
>> The race can reproduced by sending wait enabled IPI in softirq/irq env
>
>Which code path is doing that?
>
>Thanks,
>
> tglx
Thanks for your kindly reply.
I checked kernel and found no code path can run into this.
Actually , i encounter with this problem by my own code.
I need to do some specific urgent work periodicity and these
work may run for quite a while. So i can't disable irq during these work
which stops me from using hrtimer to do this. So i did add an extra
sofitrq action which may invoke smp_call.
Powered by blists - more mailing lists