Re: [PATCH 0/5] Fix GBETH TX scheduling policy for RZ/{V2H, V2N, G3E, T2H, N2H}

[Geert Uytterhoeven]

Hi Ovidiu,

On Wed, 22 Jul 2026 at 10:54, Ovidiu Panait
<ovidiu.panait.rb@xxxxxxxxxxx> wrote:
> This series switches the GBETH TX scheduling policy to
> Weighted-Round-Robin (WRR) on RZ/V2H, RZ/V2N, RZ/G3E, RZ/T2H and RZ/N2H
> SoCs.
>
> Since the GBETH nodes don't currently specify a TX scheduling policy,
> stmmac falls back to Strict Priority. In that configuration the highest
> priority queue gets all the traffic and starves the others under load.
>
> Under sustained UDP TX load with multiple data streams, this starvation
> triggers adapter resets due to TX queue timeouts:
>
>   iperf3 -c <ip> -i0 -t60 --bind-dev end0 -u -b0 -P4
>   end0: NETDEV WATCHDOG: CPU: 1: transmit queue 1 timed out 5228 ms
>   end0: Reset adapter.
>
> Investigation showed that only the highest priority queue advances while
> the others stall for more than 5 seconds, causing the netdev watchdog
> reset.

FTR. fast-tracking to fixes for v7.3.

Gr{oetje,eeting}s,

                        Geert

-- 
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@xxxxxxxxxxxxxx

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds