On Thu, Jul 18, 2024 at 8:28 PM Bailey Forrest bcf@google.com wrote:
On Thu, Jul 18, 2024 at 4:07 PM Willem de Bruijn willemdebruijn.kernel@gmail.com wrote:
This however loops skb->len / gso_size. While the above modulo operation skips many segments that span a frag. Not sure if the more intuitive approach could be as performant.
Yes, the original intention of the code was to loop over nr_frags, instead of (skb->len / gso_size).
But perhaps that's premature optimization if it makes the code significantly harder to follow.
Thanks. I don't mean to ask for a wholesale rewrite if not needed.
But perhaps the logic can be explained in the commit in a way that it is more immediately obvious.
Praveen suggested that. I'll respond to his reply in more detail.