Re: [PATCH net-next v3 1/4] net: stmmac: Switch to zero-copy in non-XDP RX path

[Furong Xu <0x1207@gmail.com>]

On Sun, 26 Jan 2025 13:35:45 +0200, Ido Schimmel wrote:

> On Sun, Jan 26, 2025 at 06:37:14PM +0800, Furong Xu wrote:
> > The "SPH feature" splits header into buf->page (non-zero offset) and
> > splits payload into buf->sec_page (zero offset).
> > 
> > For buf->page, pp_params.max_len should be the size of L3/L4 header,
> > and with a offset of NET_SKB_PAD.
> > 
> > For buf->sec_page, pp_params.max_len should be dma_conf->dma_buf_sz,
> > and with a offset of 0.
> > 
> > This is always true:
> > sizeof(L3/L4 header) + NET_SKB_PAD < dma_conf->dma_buf_sz + 0  
> 
> Thanks, understood, but are there situations where the device is
> unable to split a packet? For example, a large L2 packet. I am trying
> to understand if there are situations where the device will write
> more than "dma_conf->dma_buf_sz - NET_SKB_PAD" to the head buffer.

Nice catch!
When receiving a large L2/non-IP packet, more than "dma_conf->dma_buf_sz
- NET_SKB_PAD" will be written.

So we should:
pp_params.max_len = dma_conf->dma_buf_sz + stmmac_rx_offset(priv);

Thanks a lot, Ido
Have a nice weekend :)