Re: [net PATCH] net: stmmac: replace priv->speed with the portTransmitRate from the tc-cbs parameters

From: Vladimir Oltean
Date: Fri Jun 07 2024 - 09:47:36 EST


On Fri, Jun 07, 2024 at 06:33:27PM +0800, Xiaolei Wang wrote:
> Since the given offload->sendslope only applies to the
> current link speed, and userspace may reprogram it when
> the link speed changes, don't even bother tracking the
> port's link speed, and deduce the port transmit rate
> from idleslope - sentslope instead.
>
> Signed-off-by: Xiaolei Wang <xiaolei.wang@xxxxxxxxxxxxx>

Patches to the "net" tree usually need a Fixes: tag pointing to the
first commit introducing an issue. They also need an explanation of the
problem being addressed and how it can negatively affect an user.

Still on the process topic, please increment the patch version from the
previous submissions, and post a change log under the "---" sign below,
as well as links on patchwork or lore to previous versions.

> ---
> drivers/net/ethernet/stmicro/stmmac/stmmac_tc.c | 16 ++++++++--------
> 1 file changed, 8 insertions(+), 8 deletions(-)
>
> diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_tc.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_tc.c
> index 222540b55480..48500864017b 100644
> --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_tc.c
> +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_tc.c
> @@ -348,6 +348,7 @@ static int tc_setup_cbs(struct stmmac_priv *priv,
> u32 mode_to_use;
> u64 value;
> int ret;
> + s64 port_transmit_rate_kbps;
>
> /* Queue 0 is not AVB capable */
> if (queue <= 0 || queue >= tx_queues_count)
> @@ -355,27 +356,24 @@ static int tc_setup_cbs(struct stmmac_priv *priv,
> if (!priv->dma_cap.av)
> return -EOPNOTSUPP;
>
> + port_transmit_rate_kbps = qopt->idleslope - qopt->sendslope;
> +
> /* Port Transmit Rate and Speed Divider */
> - switch (priv->speed) {
> + switch (div_s64(port_transmit_rate_kbps, 1000)) {
> case SPEED_10000:
> ptr = 32;
> - speed_div = 10000000;
> break;
> case SPEED_5000:
> ptr = 32;
> - speed_div = 5000000;
> break;
> case SPEED_2500:
> ptr = 8;
> - speed_div = 2500000;
> break;
> case SPEED_1000:
> ptr = 8;
> - speed_div = 1000000;
> break;
> case SPEED_100:
> ptr = 4;
> - speed_div = 100000;
> break;
> default:
> return -EOPNOTSUPP;

This can be further compressed after the elimination of speed_div:

switch (div_s64(port_transmit_rate_kbps, 1000)) {
case SPEED_10000:
case SPEED_5000:
ptr = 32;
case SPEED_2500:
case SPEED_1000:
ptr = 8;
break;
case SPEED_100:
ptr = 4;
break;
default:
return -EOPNOTSUPP;
}

> @@ -397,11 +395,13 @@ static int tc_setup_cbs(struct stmmac_priv *priv,
> priv->plat->tx_queues_cfg[queue].mode_to_use = MTL_QUEUE_DCB;
> }
>
> + port_transmit_rate_kbps = qopt->idleslope - qopt->sendslope;
> +

You don't need to calculate it twice in the same function.

> /* Final adjustments for HW */
> - value = div_s64(qopt->idleslope * 1024ll * ptr, speed_div);
> + value = div_s64(qopt->idleslope * 1024ll * ptr, port_transmit_rate_kbps);
> priv->plat->tx_queues_cfg[queue].idle_slope = value & GENMASK(31, 0);
>
> - value = div_s64(-qopt->sendslope * 1024ll * ptr, speed_div);
> + value = div_s64(-qopt->sendslope * 1024ll * ptr, port_transmit_rate_kbps);
> priv->plat->tx_queues_cfg[queue].send_slope = value & GENMASK(31, 0);
>
> value = qopt->hicredit * 1024ll * 8;
> --
> 2.25.1
>