mlx4: improve Rx performance with better prefetching
authorNelio Laranjeiro <nelio.laranjeiro@6wind.com>
Fri, 30 Oct 2015 19:00:07 +0000 (20:00 +0100)
committerThomas Monjalon <thomas.monjalon@6wind.com>
Fri, 30 Oct 2015 23:21:58 +0000 (00:21 +0100)
commita21bc2fabcce51daa2e5261995c03b839abc0fa3
tree01c3f1b0368b57ca60c4f537006d032048eaa58b
parenta7e11a0ce9013065e82481dc8e3b106c3b7c4b20
mlx4: improve Rx performance with better prefetching

Prefetching initial bytes of mbuf structures earlier and in two cache lines
instead of one improves performance of mlx4_rx_burst(), which accesses the
mbuf->next field not present in the first 128 bytes.

Signed-off-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>
drivers/net/mlx4/mlx4.c