[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAFki+L=RaSZXASAaAxBf=RJqXWju+pkSj3ftMkmoqCLPfg0v=g@mail.gmail.com>
Date: Mon, 17 May 2021 17:13:18 -0400
From: Nitesh Lal <nilal@...hat.com>
To: Thomas Gleixner <tglx@...utronix.de>
Cc: Robin Murphy <robin.murphy@....com>,
Jesse Brandeburg <jesse.brandeburg@...el.com>,
"frederic@...nel.org" <frederic@...nel.org>,
"juri.lelli@...hat.com" <juri.lelli@...hat.com>,
Marcelo Tosatti <mtosatti@...hat.com>,
Ingo Molnar <mingo@...nel.org>, linux-kernel@...r.kernel.org,
intel-wired-lan@...ts.osuosl.org, jbrandeb@...nel.org,
Alex Belits <abelits@...vell.com>,
"linux-api@...r.kernel.org" <linux-api@...r.kernel.org>,
"bhelgaas@...gle.com" <bhelgaas@...gle.com>,
"linux-pci@...r.kernel.org" <linux-pci@...r.kernel.org>,
"rostedt@...dmis.org" <rostedt@...dmis.org>,
"peterz@...radead.org" <peterz@...radead.org>,
"davem@...emloft.net" <davem@...emloft.net>,
"akpm@...ux-foundation.org" <akpm@...ux-foundation.org>,
"sfr@...b.auug.org.au" <sfr@...b.auug.org.au>,
"stephen@...workplumber.org" <stephen@...workplumber.org>,
"rppt@...ux.vnet.ibm.com" <rppt@...ux.vnet.ibm.com>,
"jinyuqi@...wei.com" <jinyuqi@...wei.com>,
"zhangshaokun@...ilicon.com" <zhangshaokun@...ilicon.com>,
netdev@...r.kernel.org, chris.friesen@...driver.com,
Marc Zyngier <maz@...nel.org>
Subject: Re: [PATCH tip:irq/core v1] genirq: remove auto-set of the mask when
setting the hint
On Mon, May 17, 2021 at 3:47 PM Thomas Gleixner <tglx@...utronix.de> wrote:
>
> Nitesh,
>
> On Mon, May 17 2021 at 14:21, Nitesh Lal wrote:
> > On Mon, May 17, 2021 at 1:26 PM Robin Murphy <robin.murphy@....com> wrote:
> >
> > We can use irq_set_affinity() to set the hint mask as well, however, maybe
> > there is a specific reason behind separating those two in the
> > first place (maybe not?).
>
> Yes, because kernel side settings might overwrite the hint.
>
> > But even in this case, we have to either modify the PMU drivers' IRQs
> > affinity from the userspace or we will have to make changes in the existing
> > request_irq code path.
>
> Adjusting them from user space does not work for various reasons,
> especially CPU hotplug.
>
> > I am not sure about the latter because we already have the required controls
> > to adjust the device IRQ mask (by using default_smp_affinity or by modifying
> > them manually).
>
> default_smp_affinity does not help at all and there is nothing a module
> can modify manually.
Right, it will not help a module.
>
> I'll send out a patch series which cleans that up soon.
Ack, thanks.
>
> Thanks,
>
> tglx
>
>
--
Nitesh
Powered by blists - more mailing lists