netdev - Re: [PATCH RFC net-next] net: ag71xx: disable GRO by default

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: <CAKxU2N8CPgfODgsRufcfTg6MjnXCTgqGXv5cTqZjrx7D9ACURw@mail.gmail.com>
Date: Sat, 17 Aug 2024 14:34:26 -0700
From: Rosen Penev <rosenp@...il.com>
To: Andrew Lunn <andrew@...n.ch>
Cc: netdev@...r.kernel.org, davem@...emloft.net, edumazet@...gle.com, 
	kuba@...nel.org, pabeni@...hat.com, linux-kernel@...r.kernel.org, 
	o.rempel@...gutronix.de, nbd@....name
Subject: Re: [PATCH RFC net-next] net: ag71xx: disable GRO by default

On Sat, Aug 17, 2024 at 11:39 AM Andrew Lunn <andrew@...n.ch> wrote:
>
> On Fri, Aug 16, 2024 at 03:47:33PM -0700, Rosen Penev wrote:
> > ```
> > Currently this is handled in userspace with ethtool. Not sure if this
> > should be done in the kernel or if this is even the proper place for it.
> > ```
>
> Comments like this should be placed under the ---. If the patch is
> merged, anything in the commit message under the --- is then
> discarded.
Well, I don't mean for the patch to be merged. I'm mostly trying to
get feedback on it. From what I see in the tree, it's not common to
disable NETIF_F_SOFT_FEATURES.
>
> > ag71xx is usually paired with qca8k or ar9331, both DSA drivers.
>
> Can it be used without a switch? It looks like this option will
> disable offloads which are useful when there is no switch.
Most of the time it is used with a switch. Even in cases where only
one port is available it still goes through the switch (QCA9531 single
port device using AR9331 DSA).

Some older devices might not go through the switch. I'm not sure.
>
>         Andrew