netdev - Re: [PATCH] gve: add missing NULL check for gve_alloc_pending

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: <b8d181e9-d818-4cf4-b470-a54b6df763a4@redhat.com>
Date: Tue, 3 Jun 2025 12:50:47 +0200
From: Paolo Abeni <pabeni@...hat.com>
To: ALOK TIWARI <alok.a.tiwari@...cle.com>, Bailey Forrest <bcf@...gle.com>,
 Eric Dumazet <edumazet@...gle.com>
Cc: Mina Almasry <almasrymina@...gle.com>, joshwash@...gle.com,
 willemb@...gle.com, pkaligineedi@...gle.com, kuba@...nel.org,
 jeroendb@...gle.com, hramamurthy@...gle.com, andrew+netdev@...n.ch,
 davem@...emloft.net, netdev@...r.kernel.org, linux-kernel@...r.kernel.org,
 darren.kenny@...cle.com
Subject: Re: [PATCH] gve: add missing NULL check for
 gve_alloc_pending_packet() in TX DQO

On 6/3/25 11:03 AM, ALOK TIWARI wrote:
> On 03-06-2025 00:54, Bailey Forrest wrote:
>> I think this patch isn't needed. gve_tx_add_skb_dqo() is only called
>> after checking gve_maybe_stop_tx_dqo(), which checks that
>> gve_alloc_pending_packet() will not return NULL.
> 
> Thank you for the clarification,
> 
> Even so, I felt it could be a bit misleading for developers and tools. 
> But if you believe the patch isn't required,I completely understand.
> In that case, I kindly request you to provide your NACK on the [PATCH 
> net v2] mail thread for formal tracking,
> so that other developers can also be aware of the reasoning and 
> understand the context.

IMHO it's indeed confusing that the same condition is checked in
gve_alloc_pending_packet() and ignored by gve_tx_add_skb_dqo().

Even gve_alloc_pending_packet() is only called after the
gve_maybe_stop_tx_dqo().

Either always ignore the NULL condition it in both places (possibly with
a comment) or always check it.

/P