[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Wed, 29 Mar 2023 06:54:05 +0200
From: Felix Fietkau <nbd@....name>
To: Jakub Kicinski <kuba@...nel.org>
Cc: netdev@...r.kernel.org, Frank Wunderlich <frank-w@...lic-files.de>,
Daniel Golle <daniel@...rotopia.org>
Subject: Re: [PATCH net] net: ethernet: mtk_eth_soc: fix tx throughput
regression with direct 1G links
On 29.03.23 04:01, Jakub Kicinski wrote:
> On Fri, 24 Mar 2023 15:04:04 +0100 Felix Fietkau wrote:
>> Using the QDMA tx scheduler to throttle tx to line speed works fine for
>> switch ports, but apparently caused a regression on non-switch ports.
>>
>> Based on a number of tests, it seems that this throttling can be safely
>> dropped without re-introducing the issues on switch ports that the
>> tx scheduling changes resolved.
>>
>> Link: https://lore.kernel.org/netdev/trinity-92c3826f-c2c8-40af-8339-bc6d0d3ffea4-1678213958520@3c-app-gmx-bs16/
>> Fixes: f63959c7eec3 ("net: ethernet: mtk_eth_soc: implement multi-queue support for per-port queues")
>> Reported-by: Frank Wunderlich <frank-w@...lic-files.de>
>> Reported-by: Daniel Golle <daniel@...rotopia.org>
>> Tested-by: Daniel Golle <daniel@...rotopia.org>
>> Signed-off-by: Felix Fietkau <nbd@....name>
>
> My reading of the discussion was that this patch is good, even if it
> doesn't fix all the models, but it's marked as Changes Requested in PW.
> Could you confirm that we should apply this one as is, just to be sure?
Yes, please apply this one. I will send another patch to take care of
MT7623 soon.
- Felix
Powered by blists - more mailing lists