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: <20231128130951.577af80b@kernel.org>
Date: Tue, 28 Nov 2023 13:09:51 -0800
From: Jakub Kicinski <kuba@...nel.org>
To: Igor Russkikh <irusskikh@...vell.com>
Cc: Linus Torvalds <torvalds@...ux-foundation.org>, Eric Dumazet
 <edumazet@...gle.com>, Paolo Abeni <pabeni@...hat.com>, Netdev
 <netdev@...r.kernel.org>
Subject: Re: [EXT] Aquantia ethernet driver suspend/resume issues

On Tue, 28 Nov 2023 20:18:49 +0100 Igor Russkikh wrote:
> > Another option you can consider is lowering the default ring size.
> > If I'm looking right you default to 4k descriptors for Tx.
> > Is it based on real life experience?  
> 
> Probably reducing default will help - but again not 100%.
> 
> I remember these numbers where chosen mainly to show up good 10Gbps
> line speed in tests, like iperf udp/tcp flood. But these of course
> artificial.
> 
> For sure "normal" user can survive even with lower digits.

For Rx under load larger rings are sometimes useful to avoid drops.
But your Tx rings are larger than Rx, which is a bit odd.

I was going to say that with BQL enabled you're very unlikely to ever
use much of the 4k Tx ring, anyway. But you don't have BQL support :S

My free advice is to recheck you really need these sizes and implement
BQL :)

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ