netdev - Re: [PATCH net-next 0/3] gve: Improve RX buffer length management

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <CAJcM6BECrKh2updwNk9c-4oDPqhgku_2KVzTimDASit1c1JWvg@mail.gmail.com>
Date: Fri, 24 Oct 2025 10:31:01 -0700
From: Ankit Garg <nktgrg@...gle.com>
To: Mina Almasry <almasrymina@...gle.com>
Cc: Jakub Kicinski <kuba@...nel.org>, Joshua Washington <joshwash@...gle.com>, netdev@...r.kernel.org
Subject: Re: [PATCH net-next 0/3] gve: Improve RX buffer length management

On Thu, Oct 23, 2025 at 7:41 PM Mina Almasry <almasrymina@...gle.com> wrote:
>
> On Thu, Oct 23, 2025 at 5:12 PM Jakub Kicinski <kuba@...nel.org> wrote:
> >
> > On Wed, 22 Oct 2025 11:22:22 -0700 Joshua Washington wrote:
> > > This patch series improves the management of the RX buffer length for
> > > the DQO queue format in the gve driver. The goal is to make RX buffer
> > > length config more explicit, easy to change, and performant by default.
> >
> > I suppose this was done in prep for later buffers for ZC?
> > It's less urgent given the change of plans in
> > https://lore.kernel.org/all/20251016184031.66c92962@kernel.org/
> > but still relevant AFAICT. Just wanted to confirm.. Mina?
>
> This is very likely unrelated to the large ZC buffers and instead
> related to another effort we're preparing for. Not really 'urgent' but
> as always a bit pressing :D
>
> --
> Thanks,
> Mina

This is being done in preparation for enabling HW-GRO by default later
on. Our device only coalesces packets till 19 buffers are filled and
with 2k buffer, biggest HW coalesced packet will be limited to 38k bytes.

To enable 64k sized packet, 4K buffers are needed.