Re: [PATCH net-next 1/3] net: macb: queue tie-off or disable during WOL suspend

["Karumanchi, Vineeth" <vineeth.karumanchi@amd.com>]

Hi Claudiu,


On 2/3/2024 9:08 PM, claudiu beznea wrote:
<...>
>> +		for (q = 0, queue = bp->queues; q < bp->num_queues;
>> +		     ++q, ++queue) {
>> +			/* Disable RX queues */
> Operation in this for loop could be moved in the the above IRQ disable
> loop. Have you tried it? are there any issues with it?

OK. I haven't tried it. I will try and update.

> 
>> +			if (bp->caps & MACB_CAPS_QUEUE_DISABLE) {
>> +				queue_writel(queue, RBQP, GEM_RBQP_DISABLE);
>> +			} else {
>> +				/* Tie off RX queues */
>> +				queue_writel(queue, RBQP,
>> +					     lower_32_bits(bp->rx_ring_tieoff_dma));
> I think this should be guarded by:
> #ifdef CONFIG_ARCH_DMA_ADDR_T_64BIT
>> +				queue_writel(queue, RBQPH,
>> +					     upper_32_bits(bp->rx_ring_tieoff_dma));
>> +			}
>> +		}
>> +		/* Enable Receive engine */
>> +		ctrlmask = macb_readl(bp, NCR);
> Is this needed?

Yes, not needed, we can use earlier value directly.

> 
>> +		ctrlmask |= MACB_BIT(RE);
<...>
-- 
🙏 vineeth