linux-pci.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Thomas Gleixner <tglx@linutronix.de>
To: Keith Busch <kbusch@kernel.org>,
	Chris Friesen <chris.friesen@windriver.com>
Cc: Bjorn Helgaas <helgaas@kernel.org>,
	linux-pci@vger.kernel.org, Christoph Hellwig <hch@lst.de>,
	Nitesh Narayan Lal <nitesh@redhat.com>
Subject: Re: PCI, isolcpus, and irq affinity
Date: Mon, 12 Oct 2020 21:44:49 +0200	[thread overview]
Message-ID: <87tuuzp6tq.fsf@nanos.tec.linutronix.de> (raw)
In-Reply-To: <20201012190726.GB1032142@dhcp-10-100-145-180.wdl.wdc.com>

On Mon, Oct 12 2020 at 12:07, Keith Busch wrote:

> On Mon, Oct 12, 2020 at 12:58:41PM -0600, Chris Friesen wrote:
>> On 10/12/2020 11:50 AM, Thomas Gleixner wrote:
>> > On Mon, Oct 12 2020 at 11:58, Bjorn Helgaas wrote:
>> > > On Mon, Oct 12, 2020 at 09:49:37AM -0600, Chris Friesen wrote:
>> > > > I've got a linux system running the RT kernel with threaded irqs.  On
>> > > > startup we affine the various irq threads to the housekeeping CPUs, but I
>> > > > recently hit a scenario where after some days of uptime we ended up with a
>> > > > number of NVME irq threads affined to application cores instead (not good
>> > > > when we're trying to run low-latency applications).
>> > 
>> > These threads and the associated interupt vectors are completely
>> > harmless and fully idle as long as there is nothing on those isolated
>> > CPUs which does disk I/O.
>> 
>> Some of the irq threads are affined (by the kernel presumably) to multiple
>> CPUs (nvme1q2 and nvme0q2 were both affined 0x38000038, a couple of other
>> queues were affined 0x1c00001c0).
>
> That means you have more CPUs than your controller has queues. When that
> happens, some sharing of the queue resources among CPUs is required.
>  
>> In this case could disk I/O submitted by one of those CPUs end up
>> interrupting another one?
>
> If you dispatch IO from any CPU in the mask, then the completion side
> wakes the thread to run on one of the CPUs in the affinity mask.

Pre 4.17, yes.

From 4.17 onwards the irq thread is following the effective affinity of
the hardware interrupt which is a single CPU target.

Since 5.6 the effective affinity is steered to a housekeeping CPU if the
cpumask of a queue spawns multiple CPUs.

Thanks,

        tglx

  reply	other threads:[~2020-10-12 19:44 UTC|newest]

Thread overview: 14+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-10-12 15:49 Chris Friesen
2020-10-12 16:58 ` Bjorn Helgaas
2020-10-12 17:39   ` Sean V Kelley
2020-10-12 19:18     ` Chris Friesen
2020-10-12 17:42   ` Nitesh Narayan Lal
2020-10-12 17:50   ` Thomas Gleixner
2020-10-12 18:58     ` Chris Friesen
2020-10-12 19:07       ` Keith Busch
2020-10-12 19:44         ` Thomas Gleixner [this message]
2020-10-15 18:47         ` Chris Friesen
2020-10-15 19:02           ` Keith Busch
2020-10-12 19:31       ` Thomas Gleixner
2020-10-12 20:24         ` David Woodhouse
2020-10-12 22:25           ` Thomas Gleixner

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=87tuuzp6tq.fsf@nanos.tec.linutronix.de \
    --to=tglx@linutronix.de \
    --cc=chris.friesen@windriver.com \
    --cc=hch@lst.de \
    --cc=helgaas@kernel.org \
    --cc=kbusch@kernel.org \
    --cc=linux-pci@vger.kernel.org \
    --cc=nitesh@redhat.com \
    --subject='Re: PCI, isolcpus, and irq affinity' \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).