[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20171128083114.GA1540@linutronix.de>
Date: Tue, 28 Nov 2017 09:31:14 +0100
From: Sebastian Andrzej Siewior <bigeasy@...utronix.de>
To: Steven Rostedt <rostedt@...dmis.org>
Cc: Clark Williams <williams@...hat.com>,
LKML <linux-kernel@...r.kernel.org>,
Peter Zijlstra <peterz@...radead.org>,
Thomas Gleixner <tglx@...utronix.de>,
Ingo Molnar <mingo@...nel.org>,
Daniel Bristot de Oliveira <bristot@...hat.com>,
John Kacur <jkacur@...hat.com>, Scott Wood <swood@...hat.com>,
Mike Galbraith <efault@....de>
Subject: Re: [PATCH RT] fix IPI balancing for 4.14-rt
On 2017-11-27 17:25:21 [-0500], Steven Rostedt wrote:
> On Thu, 23 Nov 2017 18:25:39 +0100
> Sebastian Andrzej Siewior <bigeasy@...utronix.de> wrote:
>
>
> > I cherry-picked commit 4bdced5c9a29 ("sched/rt: Simplify the IPI based
> > RT balancing logic") and while refreshing the queue I noticed that the
> > irq_work struct moved and added the fix below into the original patch
> > where the IRQ_WORK_HARD_IRQ flag was added.
>
> Perhaps you should keep it as a separate patch. 4bdced5c9a29 is now in
> mainline. The next time you port to mainline, you may drop this fix,
> and cause the scheduling IPI to run in threaded context (which would be
> bad).
>
> Having it as a separate patch, would remind us that it would need to be
> added to mainline in the future.
That patch in v4.14 rt-devel RT git tree has it as one commit. The patch
in RT queue has the IRQ_WORK_HARD_IRQ added in the patch where that flag
was introduced. For reference:
https://git.kernel.org/pub/scm/linux/kernel/git/rt/linux-rt-devel.git/tree/patches/irqwork-push_most_work_into_softirq_context.patch?h=linux-4.14.y-rt-patches&id=657d8cd9f93891840fb1cd1666a8e590d19e72ba#n144
> Feel free to add to that that patch:
>
> Signed-off-by: Steven Rostedt (VMware) <rostedt@...dmis.org>
>
> -- Steve
Sebastian
Powered by blists - more mailing lists