Re: [PATCH] net: macb: Apply RXUBR workaround only to versions with errata

[<Nicolas.Ferre@microchip.com>]

On 24/01/2019 at 14:38, Harini Katakam wrote:
> The interrupt handler contains a workaround for RX hang applicable
> to Zynq and AT91 only. Subsequent versions do not need this

AT91RM9200 only. It's not the case for other AT91 SoCs (reading errata 
list for them).

That being said I have to add a patch for making this perfectly clear in 
the comment just above the flag's test.

> workaround. This workaround unecessarily resets RX whenever RX used

Typo: unnecessarily

> bit read is observed, which can be often under heavy traffic. Hence
> introduce an CAPS mask and a check to enable this workaround.

Nack for this one, see below...

> Signed-off-by: Harini Katakam <harini.katakam@xilinx.com>
> ---
> Changes from RFC:
> - Use CAPS mask instead introducing and errata field.
> - Use check only on RX reset part; ISR should still be cleared.
> 
>   drivers/net/ethernet/cadence/macb.h      |  1 +
>   drivers/net/ethernet/cadence/macb_main.c | 16 ++++++++++------
>   2 files changed, 11 insertions(+), 6 deletions(-)
> 
> diff --git a/drivers/net/ethernet/cadence/macb.h b/drivers/net/ethernet/cadence/macb.h
> index 3d45f4c..2b412fa 100644
> --- a/drivers/net/ethernet/cadence/macb.h
> +++ b/drivers/net/ethernet/cadence/macb.h
> @@ -643,6 +643,7 @@
>   #define MACB_CAPS_JUMBO				0x00000020
>   #define MACB_CAPS_GEM_HAS_PTP			0x00000040
>   #define MACB_CAPS_BD_RD_PREFETCH		0x00000080
> +#define MACB_CAPS_NEEDS_RSTONUBR		0x00000100
>   #define MACB_CAPS_FIFO_MODE			0x10000000
>   #define MACB_CAPS_GIGABIT_MODE_AVAILABLE	0x20000000
>   #define MACB_CAPS_SG_DISABLED			0x40000000
> diff --git a/drivers/net/ethernet/cadence/macb_main.c b/drivers/net/ethernet/cadence/macb_main.c
> index 66cc792..0bda1cd 100644
> --- a/drivers/net/ethernet/cadence/macb_main.c
> +++ b/drivers/net/ethernet/cadence/macb_main.c
> @@ -1416,10 +1416,12 @@ static irqreturn_t macb_interrupt(int irq, void *dev_id)
>   		 * section 16.7.4 for details.
>   		 */
>   		if (status & MACB_BIT(RXUBR)) {
> -			ctrl = macb_readl(bp, NCR);
> -			macb_writel(bp, NCR, ctrl & ~MACB_BIT(RE));
> -			wmb();
> -			macb_writel(bp, NCR, ctrl | MACB_BIT(RE));
> +			if (bp->caps & MACB_CAPS_NEEDS_RSTONUBR) {
> +				ctrl = macb_readl(bp, NCR);
> +				macb_writel(bp, NCR, ctrl & ~MACB_BIT(RE));
> +				wmb();
> +				macb_writel(bp, NCR, ctrl | MACB_BIT(RE));
> +			}

As this interrupt routine is not doing anything else than just being 
called now, what about just removing the use of this status bit for 
platforms which don't need this RX reset?

>   			if (bp->caps & MACB_CAPS_ISR_CLEAR_ON_WRITE)
>   				queue_writel(queue, ISR, MACB_BIT(RXUBR));
> @@ -3864,7 +3866,8 @@ static int at91ether_init(struct platform_device *pdev)
>   }
>   
>   static const struct macb_config at91sam9260_config = {
> -	.caps = MACB_CAPS_USRIO_HAS_CLKEN | MACB_CAPS_USRIO_DEFAULT_IS_MII_GMII,
> +	.caps = MACB_CAPS_USRIO_HAS_CLKEN |
> +		MACB_CAPS_USRIO_DEFAULT_IS_MII_GMII | MACB_CAPS_NEEDS_RSTONUBR,
>   	.clk_init = macb_clk_init,
>   	.init = macb_init,

Actually it seems not the proper configuration structure to modify: You 
must change the emac_config instead.


>   };
> @@ -3928,7 +3931,8 @@ static const struct macb_config zynqmp_config = {
>   };
>   
>   static const struct macb_config zynq_config = {
> -	.caps = MACB_CAPS_GIGABIT_MODE_AVAILABLE | MACB_CAPS_NO_GIGABIT_HALF,
> +	.caps = MACB_CAPS_GIGABIT_MODE_AVAILABLE | MACB_CAPS_NO_GIGABIT_HALF |
> +		MACB_CAPS_NEEDS_RSTONUBR,
>   	.dma_burst_length = 16,
>   	.clk_init = macb_clk_init,
>   	.init = macb_init,


Best regards,
-- 
Nicolas Ferre