netdev - Re: [PATCH net] ice: don't stop netdev tx queues when setting up XSK socket

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] [day] [month] [year] [list]

Message-ID: <ZR1oF7he5WaToeR1@boxer>
Date: Wed, 4 Oct 2023 15:26:47 +0200
From: Maciej Fijalkowski <maciej.fijalkowski@...el.com>
To: Jakub Kicinski <kuba@...nel.org>
CC: Tony Nguyen <anthony.l.nguyen@...el.com>, <davem@...emloft.net>,
	<pabeni@...hat.com>, <edumazet@...gle.com>, <netdev@...r.kernel.org>, "Kamil
 Maziarz" <kamil.maziarz@...el.com>, <magnus.karlsson@...el.com>,
	<ast@...nel.org>, <daniel@...earbox.net>, <hawk@...nel.org>,
	<john.fastabend@...il.com>, <bpf@...r.kernel.org>, Chandan Kumar Rout
	<chandanx.rout@...el.com>
Subject: Re: [PATCH net] ice: don't stop netdev tx queues when setting up XSK
 socket

On Tue, Oct 03, 2023 at 03:49:20PM -0700, Jakub Kicinski wrote:
> On Mon, 25 Sep 2023 10:19:57 -0700 Tony Nguyen wrote:
> > Avoid stopping netdev tx queues during XSK setup by removing
> > netif_tx_stop_queue() and netif_tx_start_queue().
> > These changes prevent unnecessary stopping and starting of netdev
> > transmit queues during the setup of XDP socket. Without this change,
> > after stopping the XDP traffic flow tracker and then stopping
> > the XDP prog - NETDEV WATCHDOG transmit queue timed out appears.

what is xdp traffic flow tracker? what do you mean by stopping xdp prog -
or you removing it while keeping xsk pool present on interface? or are you
just downing interface?

can you provide steps to reproduce it?

> 
> I think we need more info about what happens here.
> 
> Maybe ice_qp_ena() fails before it gets to the start?
> If we don't understand what happens, exactly, we may be papering
> over other bugs.

+1

> -- 
> pw-bot: cr
>