[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <c3b56d02-e415-b6e9-2c22-9c3d341e07e9@iogearbox.net>
Date: Sat, 27 Feb 2021 01:09:26 +0100
From: Daniel Borkmann <daniel@...earbox.net>
To: Lorenzo Bianconi <lorenzo@...nel.org>, bpf@...r.kernel.org
Cc: netdev@...r.kernel.org, davem@...emloft.net, kuba@...nel.org,
ast@...nel.org, lorenzo.bianconi@...hat.com, brouer@...hat.com,
toke@...hat.com, freysteinn.alfredsson@....se
Subject: Re: [PATCH bpf-next] bpf: devmap: move drop error path to devmap for
XDP_REDIRECT
On 2/17/21 2:56 PM, Lorenzo Bianconi wrote:
> We want to change the current ndo_xdp_xmit drop semantics because
> it will allow us to implement better queue overflow handling.
> This is working towards the larger goal of a XDP TX queue-hook.
> Move XDP_REDIRECT error path handling from each XDP ethernet driver to
> devmap code. According to the new APIs, the driver running the
> ndo_xdp_xmit pointer, will break tx loop whenever the hw reports a tx
> error and it will just return to devmap caller the number of successfully
> transmitted frames. It will be devmap responsability to free dropped frames.
> Move each XDP ndo_xdp_xmit capable driver to the new APIs:
> - veth
> - virtio-net
> - mvneta
> - mvpp2
> - socionext
> - amazon ena
> - bnxt
> - freescale (dpaa2, dpaa)
> - xen-frontend
> - qede
> - ice
> - igb
> - ixgbe
> - i40e
> - mlx5
> - ti (cpsw, cpsw-new)
> - tun
> - sfc
I presume for a number of these drivers the refactoring changes were just compile-
tested due to lack of HW, right? If so, please also Cc related driver maintainers
aside from the few of us, so they have a chance to review & ACK the patch if it looks
good to them. I presume Ed saw it by accident, but for others it might easily get
lost in the daily mail flood.
> More details about the new ndo_xdp_xmit design can be found here [0].
>
> [0] https://github.com/xdp-project/xdp-project/blob/master/areas/core/redesign01_ndo_xdp_xmit.org
I'd probably move this below the "---" if it's not essential to the commit itself or
rather take relevant parts out and move it into the commit desc so it doesn't get
lost for future ref given things could likely reschuffle inside the repo in the future,
just a nit.
> Signed-off-by: Lorenzo Bianconi <lorenzo@...nel.org>
Powered by blists - more mailing lists