[PATCH net-next 1/1] net: stmmac: Remove redundant checking for rx_coalesce_usecs

[Gan, Yi Fang]

Hi Andrew,

Thank you for your feedback. I will submit V2 to update the commit message.

The value of rx-usecs will not be affected when the tx-usecs is set. When the command "ethtool -C eth24 tx-usecs 42" is applied, the value of rx-usecs is remaining the same as previously.

Best Regards,
Gan Yi Fang

> -----Original Message-----
> From: Andrew Lunn <andrew at lunn.ch>
> Sent: Tuesday, May 23, 2023 8:48 PM
> To: Gan, Yi Fang <yi.fang.gan at intel.com>
> Cc: Giuseppe Cavallaro <peppe.cavallaro at st.com>; Alexandre Torgue
> <alexandre.torgue at st.com>; Jose Abreu <joabreu at synopsys.com>; David S .
> Miller <davem at davemloft.net>; Eric Dumazet <edumazet at google.com>; Jakub
> Kicinski <kuba at kernel.org>; Paolo Abeni <pabeni at redhat.com>; Maxime
> Coquelin <mcoquelin.stm32 at gmail.com>; netdev at vger.kernel.org; linux-
> stm32 at st-md-mailman.stormreply.com; linux-arm-kernel at lists.infradead.org;
> linux-kernel at vger.kernel.org; Looi, Hong Aun <hong.aun.looi at intel.com>; Sit,
> Michael Wei Hong <michael.wei.hong.sit at intel.com>
> Subject: Re: [PATCH net-next 1/1] net: stmmac: Remove redundant checking for
> rx_coalesce_usecs
> 
> On Tue, May 23, 2023 at 02:19:52AM -0400, Gan Yi Fang wrote:
> > The datatype of rx_coalesce_usecs is u32, always larger or equal to zero.
> > Previous checking does not include value 0, this patch removes the
> > checking to handle the value 0.
> >
> > Signed-off-by: Gan Yi Fang <yi.fang.gan at intel.com>
> > ---
> >  drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c
> > b/drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c
> > index 35c8dd92d369..6ed0e683b5e0 100644
> > --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c
> > +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c
> > @@ -917,7 +917,7 @@ static int __stmmac_set_coalesce(struct net_device
> *dev,
> >  	else if (queue >= max_cnt)
> >  		return -EINVAL;
> >
> > -	if (priv->use_riwt && (ec->rx_coalesce_usecs > 0)) {
> > +	if (priv->use_riwt) {
> >  		rx_riwt = stmmac_usec2riwt(ec->rx_coalesce_usecs, priv);
> >
> >  		if ((rx_riwt > MAX_DMA_RIWT) || (rx_riwt <
> MIN_DMA_RIWT))
> 
> This appears to be a user visible ABI change. For the current code, a value of
> zero here is ignored, and 0 is returned. With this change, 0 will result in rx_riwt
> being calculated as 0, which is less than MIN_DMA_RIWT, so you get -EINVAL
> returned.
> 
> I don't know this uAPI too well. What values are passed to this function for:
> 
> ethtool -C eth24 tx-usecs 42
> 
> where you only want to change transmit coalesce? Is rx_usecs 0?
> 
> At minimum you need to explain in the commit message: "This change in
> behaviour making the value of 0 cause an error is not a problem because...."
> 
>     Andrew
> 
> ---
> pw-bot: cr