netdev - Re: [PATCH 1/2] stmmac: introduce flag to dynamically disable TX offload for rockchip devices

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: <1554489537.1199.0@gmail.com>
Date:   Fri, 05 Apr 2019 21:38:57 +0300
From:   "Leonidas P. Papadakos" <papadakospan@...il.com>
To:     Robin Murphy <robin.murphy@....com>
Cc:     Jose Abreu <jose.abreu@...opsys.com>,
        Philipp Tomsich <philipp.tomsich@...obroma-systems.com>,
        Heiko Stübner <heiko@...ech.de>,
        Christoph Müllner 
        <christoph.muellner@...obroma-systems.com>,
        Maxime Coquelin <mcoquelin.stm32@...il.com>,
        Alexandre Torgue <alexandre.torgue@...com>,
        netdev@...r.kernel.org, LKML <linux-kernel@...r.kernel.org>,
        Linux ARM <linux-arm-kernel@...ts.infradead.org>,
        Klaus Goger <klaus.goger@...obroma-systems.com>
Subject: Re: [PATCH 1/2] stmmac: introduce flag to dynamically disable TX
 offload for rockchip devices

> What's your overall setup? So far with my RK3328 box I've found that 
> with an IPv4 link and an MTU of 2000 I have to dial PBL right down to 
> 1 or 2 (with no-pbl-x8 as well) to prevent Tx stalling, and with the 
> MTU bumped up to 3000 it just won't play at any PBL setting...
> 
> Robin.

I'm using a Libre Computer Renegade with RK3328 SoC(maybe it's a board 
specific thing?)
I'm using txpbl as from what I can gather, the rx has no issues, no 
need to change it.

When it comes to rx,
running the server (iperf3 -s) on the Renegade and iperf3 -c 
<renegade_ip> on my laptop,
I get about 940 Mbps. It's always been like that, satisfactory.

When it comes to tx (the one all the fuss is about)
I eventually find that just with snps,txpbl = <0x21>, and no other pbl 
thing set I get about 905 Mbps on average sent to my laptop, now 
running the iperf3 server.

I must say I haven't touched the MTUs but I never needed to as the 
issue presented itself in normal usage. Simple iperf tests will fail 
after same kilobytes if no tweaks are applied.