[EXT] Re: [PATCH 1/1] net: stmmac: add support for platform specific reset

[Shenwei Wang]

> -----Original Message-----
> From: Simon Horman <simon.horman at corigine.com>
> Sent: Wednesday, March 22, 2023 11:33 AM
> To: Shenwei Wang <shenwei.wang at nxp.com>
> Cc: 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>;
> Shawn Guo <shawnguo at kernel.org>; dl-linux-imx <linux-imx at nxp.com>;
> Giuseppe Cavallaro <peppe.cavallaro at st.com>; Alexandre Torgue
> <alexandre.torgue at foss.st.com>; Jose Abreu <joabreu at synopsys.com>; Sascha
> Hauer <s.hauer at pengutronix.de>; Pengutronix Kernel Team
> <kernel at pengutronix.de>; Fabio Estevam <festevam at gmail.com>; Wong Vee
> Khee <veekhee at apple.com>; Jon Hunter <jonathanh at nvidia.com>;
> Mohammad Athari Bin Ismail <mohammad.athari.ismail at intel.com>; Andrey
> Konovalov <andrey.konovalov at linaro.org>; Revanth Kumar Uppala
> <ruppala at nvidia.com>; Tan Tee Min <tee.min.tan at linux.intel.com>;
> netdev at vger.kernel.org; linux-stm32 at st-md-mailman.stormreply.com; linux-
> arm-kernel at lists.infradead.org; imx at lists.linux.dev
> Subject: [EXT] Re: [PATCH 1/1] net: stmmac: add support for platform specific
> reset
> 
> Caution: EXT Email
> 
> On Tue, Mar 21, 2023 at 02:09:20PM -0500, Shenwei Wang wrote:
> > This patch adds support for platform-specific reset logic in the
> > stmmac driver. Some SoCs require a different reset mechanism than the
> > standard dwmac IP reset. To support these platforms, a new function
> > pointer 'fix_soc_reset' is added to the plat_stmmacenet_data structure.
> > The stmmac_reset macro in hwif.h is modified to call the 'fix_soc_reset'
> > function if it exists. This enables the driver to use the
> > platform-specific reset logic when necessary.
> >
> > Signed-off-by: Shenwei Wang <shenwei.wang at nxp.com>
> > ---
> >  drivers/net/ethernet/stmicro/stmmac/hwif.h | 10 +++++++++-
> >  include/linux/stmmac.h                     |  1 +
> >  2 files changed, 10 insertions(+), 1 deletion(-)
> >
> > diff --git a/drivers/net/ethernet/stmicro/stmmac/hwif.h
> > b/drivers/net/ethernet/stmicro/stmmac/hwif.h
> > index 16a7421715cb..e24ce870690e 100644
> > --- a/drivers/net/ethernet/stmicro/stmmac/hwif.h
> > +++ b/drivers/net/ethernet/stmicro/stmmac/hwif.h
> > @@ -215,7 +215,15 @@ struct stmmac_dma_ops {  };
> >
> >  #define stmmac_reset(__priv, __args...) \
> > -     stmmac_do_callback(__priv, dma, reset, __args)
> > +({ \
> > +     int __result = -EINVAL; \
> > +     if ((__priv) && (__priv)->plat && (__priv)->plat->fix_soc_reset) { \
> > +             __result = (__priv)->plat->fix_soc_reset((__priv)->plat, ##__args); \
> > +     } else { \
> > +             __result = stmmac_do_callback(__priv, dma, reset, __args); \
> > +     } \
> > +     __result; \
> > +})
> 
> Hi Shenwei Wang,
> 
> I am wondering if any consideration was given to an approach that has a bit
> better type safety.
> 

The patch aims to add the missing feature to the original source file in a consistent coding style. 
I am also okay to use the function prototype if everyone agrees.

Thanks,
Shenwei

> Something like this (*compile tested only!*):
> 
> static inline int stmmac_reset(struct stmmac_priv *priv, void __iomem *ioaddr) {
>        struct plat_stmmacenet_data *plat = priv ? priv->plat : NULL;
> 
>        if (plat && plat->fix_soc_reset)
>                return plat->fix_soc_reset(plat, ioaddr);
> 
>        return stmmac_do_callback(priv, dma, reset, ioaddr); }
> 
> In which case the first parameter of the fix_soc_reset field of struct
> plat_stmmacenet_data can become struct plat_stmmacenet_data *.
> 
> diff --git a/include/linux/stmmac.h b/include/linux/stmmac.h index
> 9044477fad61..b26ade7e4be8 100644
> --- a/include/linux/stmmac.h
> +++ b/include/linux/stmmac.h
> @@ -223,6 +223,8 @@ struct plat_stmmacenet_data {
>         struct stmmac_rxq_cfg rx_queues_cfg[MTL_MAX_RX_QUEUES];
>         struct stmmac_txq_cfg tx_queues_cfg[MTL_MAX_TX_QUEUES];
>         void (*fix_mac_speed)(void *priv, unsigned int speed);
> +       int (*fix_soc_reset)(struct plat_stmmacenet_data *,
> +                            void __iomem *ioaddr);
>         int (*serdes_powerup)(struct net_device *ndev, void *priv);
>         void (*serdes_powerdown)(struct net_device *ndev, void *priv);
>         void (*speed_mode_2500)(struct net_device *ndev, void *priv);
> 
> I do see that the approach you have is
> in keeping with the existing structure of stmmac_do_callback().
> But I guess my question there is: why is that model used?
> And could there be a plan to move away from that model?