Linux-NVME Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Christoph Hellwig <hch@lst.de>
To: Sagi Grimberg <sagi@grimberg.me>
Cc: Hannes Reinecke <hare@kernel.org>, Christoph Hellwig <hch@lst.de>,
	Keith Busch <kbusch@kernel.org>,
	linux-nvme@lists.infradead.org
Subject: Re: [PATCH 2/4] nvme-tcp: align I/O cpu with blk-mq mapping
Date: Thu, 4 Jul 2024 07:37:25 +0200	[thread overview]
Message-ID: <20240704053725.GB19735@lst.de> (raw)
In-Reply-To: <b6702586-12c4-4390-9085-2cc6d9fbb462@grimberg.me>

On Wed, Jul 03, 2024 at 05:19:39PM +0300, Sagi Grimberg wrote:
>
>
> On 03/07/2024 16:50, Hannes Reinecke wrote:
>> When 'wq_unbound' is selected we should select the
>> the first CPU from a given blk-mq hctx mapping to queue
>> the tcp workqueue item. With this we can instruct the
>> workqueue code to keep the I/O affinity and avoid
>> a performance penalty.
>
> wq_unbound is designed to keep io_cpu to be UNBOUND, my recollection
> was the the person introducing it was trying to make the io_cpu always be
> on a specific NUMA node, or a subset of cpus within a numa node. So he uses
> that and tinkers with wq cpumask via sysfs.

Honestly, I think we really should kill this unbound module paramter
before it causes more harm.


  parent reply	other threads:[~2024-07-04  5:37 UTC|newest]

Thread overview: 37+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-07-03 13:50 [PATCH 0/4] nvme-tcp: improve scalability Hannes Reinecke
2024-07-03 13:50 ` [PATCH 1/4] nvme-tcp: per-controller I/O workqueues Hannes Reinecke
2024-07-03 14:11   ` Sagi Grimberg
2024-07-03 14:46     ` Hannes Reinecke
2024-07-03 15:16       ` Sagi Grimberg
2024-07-03 17:07         ` Tejun Heo
2024-07-03 19:14           ` Sagi Grimberg
2024-07-03 19:17             ` Tejun Heo
2024-07-03 19:41               ` Sagi Grimberg
2024-07-04  7:36               ` Hannes Reinecke
2024-07-05  7:10                 ` Christoph Hellwig
2024-07-05  8:11                   ` Hannes Reinecke
2024-07-05  8:16                     ` Jens Axboe
2024-07-04  5:36   ` Christoph Hellwig
2024-07-03 13:50 ` [PATCH 2/4] nvme-tcp: align I/O cpu with blk-mq mapping Hannes Reinecke
2024-07-03 14:19   ` Sagi Grimberg
2024-07-03 14:53     ` Hannes Reinecke
2024-07-03 15:03       ` Sagi Grimberg
2024-07-03 15:40         ` Hannes Reinecke
2024-07-03 19:38           ` Sagi Grimberg
2024-07-03 19:47             ` Sagi Grimberg
2024-07-04  6:43             ` Hannes Reinecke
2024-07-04  9:07               ` Sagi Grimberg
2024-07-04 14:03                 ` Hannes Reinecke
2024-07-04  5:37     ` Christoph Hellwig [this message]
2024-07-04  9:13       ` Sagi Grimberg
2024-07-03 13:50 ` [PATCH 3/4] workqueue: introduce helper workqueue_unbound_affinity_scope() Hannes Reinecke
2024-07-03 17:31   ` Tejun Heo
2024-07-04  6:04     ` Hannes Reinecke
2024-07-03 13:50 ` [PATCH 4/4] nvme-tcp: switch to 'cpu' affinity scope for unbound workqueues Hannes Reinecke
2024-07-03 14:22   ` Sagi Grimberg
2024-07-03 15:01     ` Hannes Reinecke
2024-07-03 15:09       ` Sagi Grimberg
2024-07-03 15:50         ` Hannes Reinecke
2024-07-04  9:11           ` Sagi Grimberg
2024-07-04 15:54             ` Hannes Reinecke
2024-07-05 11:48               ` Sagi Grimberg

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=20240704053725.GB19735@lst.de \
    --to=hch@lst.de \
    --cc=hare@kernel.org \
    --cc=kbusch@kernel.org \
    --cc=linux-nvme@lists.infradead.org \
    --cc=sagi@grimberg.me \
    /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