mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Joe Damato <jdamato@fastly.com>
To: Shailend Chand <shailend@google.com>
Cc: netdev@vger.kernel.org, Jeroen de Borst <jeroendb@google.com>,
	Praveen Kaligineedi <pkaligineedi@google.com>,
	"David S. Miller" <davem@davemloft.net>,
	Eric Dumazet <edumazet@google.com>,
	Jakub Kicinski <kuba@kernel.org>, Paolo Abeni <pabeni@redhat.com>,
	Willem de Bruijn <willemb@google.com>,
	Harshitha Ramamurthy <hramamurthy@google.com>,
	Ziwei Xiao <ziweixiao@google.com>,
	open list <linux-kernel@vger.kernel.org>
Subject: Re: [RFC net-next 3/6] gve: Use napi_affinity_no_change
Date: Tue, 13 Aug 2024 22:44:45 +0100	[thread overview]
Message-ID: <ZrvTzX8CfyM40c8I@LQ3V64L9R2.home> (raw)
In-Reply-To: <CANLc=av2PUaXQ5KVPQGppOdD5neHUtUgioqOO4fA=+Qb594Z4w@mail.gmail.com>

On Tue, Aug 13, 2024 at 11:55:31AM -0700, Shailend Chand wrote:
> On Mon, Aug 12, 2024 at 7:57 AM Joe Damato <jdamato@fastly.com> wrote:
> >
> > Use napi_affinity_no_change instead of gve's internal implementation,
> > simplifying and centralizing the logic.
> >
> > Signed-off-by: Joe Damato <jdamato@fastly.com>
> > ---
> >  drivers/net/ethernet/google/gve/gve_main.c | 14 +-------------
> >  1 file changed, 1 insertion(+), 13 deletions(-)
> >
> > diff --git a/drivers/net/ethernet/google/gve/gve_main.c b/drivers/net/ethernet/google/gve/gve_main.c
> > index 661566db68c8..ad5e85b8c6a5 100644
> > --- a/drivers/net/ethernet/google/gve/gve_main.c
> > +++ b/drivers/net/ethernet/google/gve/gve_main.c
> > @@ -298,18 +298,6 @@ static irqreturn_t gve_intr_dqo(int irq, void *arg)
> >         return IRQ_HANDLED;
> >  }
> >
> > -static int gve_is_napi_on_home_cpu(struct gve_priv *priv, u32 irq)
> > -{
> > -       int cpu_curr = smp_processor_id();
> > -       const struct cpumask *aff_mask;
> > -
> > -       aff_mask = irq_get_effective_affinity_mask(irq);
> > -       if (unlikely(!aff_mask))
> > -               return 1;
> > -
> > -       return cpumask_test_cpu(cpu_curr, aff_mask);
> > -}
> > -
> >  int gve_napi_poll(struct napi_struct *napi, int budget)
> >  {
> >         struct gve_notify_block *block;
> > @@ -383,7 +371,7 @@ int gve_napi_poll_dqo(struct napi_struct *napi, int budget)
> >                 /* Reschedule by returning budget only if already on the correct
> >                  * cpu.
> >                  */
> > -               if (likely(gve_is_napi_on_home_cpu(priv, block->irq)))
> > +               if (likely(napi_affinity_no_change(block->irq)))
> 
> Nice to centralize this code! Evolving this to cache the affinity mask
> like the other drivers would probably also require a means to update
> the cache when the affinity changes.

Thanks for taking a look.

The gve driver already calls irq_get_effective_affinity_mask in the
hot path, so I'm planning on submitting a rfcv2 which will do this:

-               if (likely(gve_is_napi_on_home_cpu(priv, block->irq)))
+               const struct cpumask *aff_mask =
+                       irq_get_effective_affinity_mask(block->irq);
+
+               if (likely(napi_affinity_no_change(aff_mask)))
                        return budget;

with a change like that there'd be no behavioral change to gve since
it didn't cache before and still won't be caching after this change.

I think a change can be made to gve in a separate patch set to
support caching the affinity mask and does not need to be included
with this change.

What do you think?

  reply	other threads:[~2024-08-13 21:44 UTC|newest]

Thread overview: 27+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-08-12 14:56 [RFC net-next 0/6] Cleanup IRQ affinity checks in several drivers Joe Damato
2024-08-12 14:56 ` [RFC net-next 1/6] netdevice: Add napi_affinity_no_change Joe Damato
2024-08-12 20:23   ` Stanislav Fomichev
2024-08-12 21:08     ` Joe Damato
2024-08-12 22:36       ` Stanislav Fomichev
2024-08-13  9:11         ` Joe Damato
2024-08-13 10:03           ` Joe Damato
2024-08-13 13:05           ` Simon Horman
2024-08-12 14:56 ` [RFC net-next 2/6] mlx5: Use napi_affinity_no_change Joe Damato
2024-08-12 14:56 ` [RFC net-next 3/6] gve: " Joe Damato
2024-08-13 18:55   ` Shailend Chand
2024-08-13 21:44     ` Joe Damato [this message]
2024-08-13 21:50       ` Shailend Chand
2024-08-12 14:56 ` [RFC net-next 4/6] i40e: " Joe Damato
2024-08-12 14:56 ` [RFC net-next 5/6] iavf: " Joe Damato
2024-08-12 14:56 ` [RFC net-next 6/6] mlx4: " Joe Damato
2024-08-14  0:17 ` [RFC net-next 0/6] Cleanup IRQ affinity checks in several drivers Jakub Kicinski
2024-08-14  7:14   ` Joe Damato
2024-08-14 12:12     ` Joe Damato
2024-08-14 15:09       ` Jakub Kicinski
2024-08-14 15:19         ` Joe Damato
2024-08-14 16:03           ` Shay Drori
2024-08-14 18:01             ` Joe Damato
2024-08-15  0:20               ` Jakub Kicinski
2024-08-15 10:22                 ` Joe Damato
2024-08-20  6:40                   ` Shay Drori
2024-08-20  8:33                     ` Joe Damato

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=ZrvTzX8CfyM40c8I@LQ3V64L9R2.home \
    --to=jdamato@fastly.com \
    --cc=davem@davemloft.net \
    --cc=edumazet@google.com \
    --cc=hramamurthy@google.com \
    --cc=jeroendb@google.com \
    --cc=kuba@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=netdev@vger.kernel.org \
    --cc=pabeni@redhat.com \
    --cc=pkaligineedi@google.com \
    --cc=shailend@google.com \
    --cc=willemb@google.com \
    --cc=ziweixiao@google.com \
    /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

all inboxes | Powered by JetHome®