From: Ming Lei <ming.lei@redhat.com>
To: Keith Busch <kbusch@kernel.org>
Cc: bigeasy@linutronix.de, tglx@linutronix.de,
Sagi Grimberg <sagi@grimberg.me>,
linux-nvme@lists.infradead.org, hch@lst.de
Subject: Re: [PATCHv3 2/4] nvme/pci: Complete commands from primary handler
Date: Fri, 13 Dec 2019 06:55:52 +0800 [thread overview]
Message-ID: <20191212225552.GB24463@ming.t460p> (raw)
In-Reply-To: <20191212010240.GA29574@redsun51.ssa.fujisawa.hgst.com>
On Thu, Dec 12, 2019 at 10:02:40AM +0900, Keith Busch wrote:
> On Wed, Dec 11, 2019 at 04:40:47PM -0800, Sagi Grimberg wrote:
> > > Perhaps we can cycle the effective_affinity through the smp_affinity?
> >
> > Not sure I follow your thoughts.
>
> The only way the nvme driver's interrupt handler can saturate a
> cpu requires the smp_affinity have multiple online cpus set. The
As Sagi mentioned, there can be 24 drives, each drive can be 1:1
mapping, so each CPU may have to handle interrupts from 24 drives/queues,
then the CPU may become slower compared with the 24 hardware/queues.
> effective_affinity, however, is always a single cpu, which can be any
> of the online ones set in smp_affinity.
>
> If we can detect a cpu is spending too many cycles in nvme irq,
> perhaps we can re-spread the effective_affinity to different cpus. I
> don't currently know of a good way to do this, though.
That shouldn't have been done just for nvme irq given other storage
has similar issue.
I have tried toward that direction, and looks some cost has to be payed, such
as reading one time of local clock in interrupt or io path has to be done:
https://lore.kernel.org/lkml/20190827085344.30799-1-ming.lei@redhat.com/
It can be quite hard to figure out one efficient way to do that.
thanks,
Ming
_______________________________________________
linux-nvme mailing list
linux-nvme@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-nvme
next prev parent reply other threads:[~2019-12-12 22:56 UTC|newest]
Thread overview: 29+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-12-09 17:56 [PATCHv3 0/4] nvme pci interrupt handling improvements Keith Busch
2019-12-09 17:56 ` [PATCHv3 1/4] nvme/pci: Disable interrupts for threaded handler Keith Busch
2019-12-10 15:12 ` Daniel Wagner
2019-12-10 15:28 ` Sebastian Andrzej Siewior
2019-12-10 15:54 ` Keith Busch
2019-12-10 16:44 ` Daniel Wagner
2019-12-10 16:57 ` Keith Busch
2019-12-10 17:11 ` Daniel Wagner
2019-12-12 9:09 ` Christoph Hellwig
2019-12-12 15:53 ` Keith Busch
2019-12-09 17:56 ` [PATCHv3 2/4] nvme/pci: Complete commands from primary handler Keith Busch
2019-12-10 20:00 ` Sagi Grimberg
2019-12-10 20:25 ` Keith Busch
2019-12-10 21:14 ` Sagi Grimberg
2019-12-11 17:35 ` Keith Busch
2019-12-12 0:40 ` Sagi Grimberg
2019-12-12 1:02 ` Keith Busch
2019-12-12 22:55 ` Ming Lei [this message]
2019-12-12 23:30 ` Keith Busch
2019-12-13 0:52 ` Ming Lei
2019-12-12 9:14 ` Christoph Hellwig
2019-12-09 17:56 ` [PATCHv3 3/4] nvme/pci: Remove use_threaded_interrupts Keith Busch
2019-12-12 9:14 ` Christoph Hellwig
2019-12-12 15:45 ` Keith Busch
2019-12-18 7:29 ` Ming Lei
2019-12-18 15:50 ` Keith Busch
2019-12-19 1:10 ` Ming Lei
2019-12-09 17:56 ` [PATCHv3 4/4] nvme/pci: Poll threaded completions Keith Busch
2019-12-10 17:43 ` Daniel Wagner
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=20191212225552.GB24463@ming.t460p \
--to=ming.lei@redhat.com \
--cc=bigeasy@linutronix.de \
--cc=hch@lst.de \
--cc=kbusch@kernel.org \
--cc=linux-nvme@lists.infradead.org \
--cc=sagi@grimberg.me \
--cc=tglx@linutronix.de \
/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
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
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).