Re: [PATCH v2 net 2/2] net: stmmac: dwmac-imx: pause the TXC clock in fixed-link

[Russell King (Oracle)]

On Thu, Jul 27, 2023 at 01:36:45PM -0500, Andrew Halaney wrote:
> > diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac-imx.c b/drivers/net/ethernet/stmicro/stmmac/dwmac-imx.c
> > index 53ee5a42c071..e7819960128e 100644
> > --- a/drivers/net/ethernet/stmicro/stmmac/dwmac-imx.c
> > +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac-imx.c
> > @@ -40,6 +40,9 @@
> >  #define DMA_BUS_MODE			0x00001000
> >  #define DMA_BUS_MODE_SFT_RESET		(0x1 << 0)
> >  #define RMII_RESET_SPEED		(0x3 << 14)
> > +#define MII_RESET_SPEED			(0x2 << 14)
> > +#define RGMII_RESET_SPEED		(0x0 << 14)
> > +#define CTRL_SPEED_MASK			(0x3 << 14)
> 
> GENMASK() would be cleaner, as well as BIT() usage, but I do see
> the driver currently does shifts.. so /me shrugs

BIT() is only useful for single-bit items, not for use with bitfields,
and their use with bitfields just makes the whole thing perverse.

#define CTRL_SPEED_MASK		GENMASK(15, 14)
#define CTRL_SPEED_RGMII_RESET	0
#define CTRL_SPEED_MII_RESET	2
#define CTRL_SPEED_RMII_RESET	3

and then its use:

	FIELD_PREP(CTRL_SPEED_MASK, CTRL_SPEED_RGMII_RESET)
or
	FIELD_PREP(CTRL_SPEED_MASK, CTRL_SPEED_MII_RESET)
or
	FIELD_PREP(CTRL_SPEED_MASK, CTRL_SPEED_RMII_RESET)

alternatively:

        if (iface == MX93_GPR_ENET_QOS_INTF_SEL_RMII)
                speed = CTRL_SPEED_RMII_RESET;
        else (iface == MX93_GPR_ENET_QOS_INTF_SEL_MII)
                speed = CTRL_SPEED_MII_RESET;
	else
		speed = CTRL_SPEED_RGMII_RESET;

	old_ctrl = readl(dwmac->base_addr + MAC_CTRL_REG);
	ctrl = old_ctrl & ~CTRL_SPEED_MASK;
	ctrl |= FIELD_PREP(CTRL_SPEED_MASK, speed);
	writel(ctrl, dwmac->base_addr + MAC_CTRL_REG);

> I don't have any documentation for the registers here, and as you can
> see I'm an amateur with respect to memory ordering based on my prior
> comment.
> 
> But you:
> 
>     1. Read intf_reg_off into variable iface
>     2. Write the RESET_SPEED for the appropriate mode to MAC_CTRL_REG
>     3. wmb() to ensure that write goes through

I wonder about whether that wmb() is required. If the mapping is
device-like rather than memory-like, the write should be committed
before the read that regmap_update_bits() does according to the ARM
memory model. Maybe a bit of information about where this barrier
has come from would be good, and maybe getting it reviewed by the
arm64 barrier specialist, Will Deacon. :)

wmb() is normally required to be paired with a rmb(), but we're not
talking about system memory here, so I also wonder whether wmb() is
the correct barrier to use.

Adding Will.

-- 
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTP is here! 80Mbps down 10Mbps up. Decent connectivity at last!