[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <fe234fbbec7f30e34bfc44231b8e061d4ca8dbf4.camel@redhat.com>
Date: Fri, 16 Feb 2024 16:33:05 -0600
From: Crystal Wood <crwood@...hat.com>
To: Thomas Gleixner <tglx@...utronix.de>
Cc: linux-kernel@...r.kernel.org, Brent Roswell <browsell@...hat.com>
Subject: Re: [PATCH] genirq: Wake IRQ threads immediately when changing
affinity
On Mon, 2024-01-22 at 17:53 -0600, Crystal Wood wrote:
> This avoids a situation where the IRQs are being migrated away from
> an isolated cpu, but the migration does not happen until some arbitrary
> time in the future, causing an isolation disruption.
>
> Note that this is of the most benefit on systems where the IRQ affinity
> itself does not need to be deferred to the IRQ handler, but even where
> that's not the case, the total dirsuption will be less.
>
> Signed-off-by: Crystal Wood <crwood@...hat.com>
Hi, do you have any thoughts on this patch?
..well, apart from the "dirsuption" typo I just noticed :-P
Thanks,
Crystal
Powered by blists - more mailing lists