From: Dimitri Daskalakis <dimitri.daskalakis1@gmail.com>
To: Jakub Kicinski <kuba@kernel.org>
Cc: "David S . Miller" <davem@davemloft.net>,
Alexander Duyck <alexanderduyck@fb.com>,
Eric Dumazet <edumazet@google.com>,
Paolo Abeni <pabeni@redhat.com>,
Andrew Lunn <andrew+netdev@lunn.ch>,
Simon Horman <horms@kernel.org>,
Mike Marciniszyn <mike.marciniszyn@gmail.com>,
Lee Trager <lee@trager.us>,
Jacob Keller <jacob.e.keller@intel.com>,
Mohsin Bashir <mohsin.bashr@gmail.com>,
Mina Almasry <almasrymina@google.com>,
netdev@vger.kernel.org, kernel-team@meta.com
Subject: Re: [PATCH net 0/2] Fix page fragment handling when PAGE_SIZE > 4K
Date: Sun, 29 Mar 2026 08:10:49 -0700 [thread overview]
Message-ID: <fc702208-daee-49dd-9b67-9cb123b50fd8@gmail.com> (raw)
In-Reply-To: <20260327202958.53d26d46@kernel.org>
On 3/27/26 8:29 PM, Jakub Kicinski wrote:
> On Tue, 24 Mar 2026 12:51:21 -0700 Dimitri Daskalakis wrote:
>> FBNIC operates on fixed size descriptors (4K). When the OS supports pages
>> larger than 4K, we fragment the page across multiple descriptors.
>>
>> While performance testing, I found several issues with our page fragment
>> handling, resulting in low throughput and potential RX stalls.
> Google AI reviewer points out that we will also be toast if someone
> configures the ring size to just 16 descriptors. Can you check what
> is the min size that doesn't wedge w/ 64kB pages and let's just set
> the min to that instead of the current 16?
Yeah, this looks legit. Bumping the min to 32 should work, but I'll
test/verify before submitting the fix.
next prev parent reply other threads:[~2026-03-29 15:10 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-03-24 19:51 [PATCH net 0/2] Fix page fragment handling when PAGE_SIZE > 4K Dimitri Daskalakis
2026-03-24 19:51 ` [PATCH net 1/2] eth: fbnic: Account for page fragments when updating BDQ tail Dimitri Daskalakis
2026-03-24 19:51 ` [PATCH net 2/2] eth: fbnic: Fix debugfs output for BDQ's with page frags Dimitri Daskalakis
2026-03-28 3:29 ` [PATCH net 0/2] Fix page fragment handling when PAGE_SIZE > 4K Jakub Kicinski
2026-03-29 15:10 ` Dimitri Daskalakis [this message]
2026-03-28 3:40 ` patchwork-bot+netdevbpf
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=fc702208-daee-49dd-9b67-9cb123b50fd8@gmail.com \
--to=dimitri.daskalakis1@gmail.com \
--cc=alexanderduyck@fb.com \
--cc=almasrymina@google.com \
--cc=andrew+netdev@lunn.ch \
--cc=davem@davemloft.net \
--cc=edumazet@google.com \
--cc=horms@kernel.org \
--cc=jacob.e.keller@intel.com \
--cc=kernel-team@meta.com \
--cc=kuba@kernel.org \
--cc=lee@trager.us \
--cc=mike.marciniszyn@gmail.com \
--cc=mohsin.bashr@gmail.com \
--cc=netdev@vger.kernel.org \
--cc=pabeni@redhat.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.