Commit 5e040d4b authored by Edward Cree's avatar Edward Cree Committed by David S. Miller

sfc: don't score irq moderation points for GRO

We already scored points when handling the RX event, no-one else does this,
 and looking at the history it appears this was originally meant to only
 score on merges, not on GRO_NORMAL.  Moreover, it gets in the way of
 changing GRO to not immediately pass GRO_NORMAL skbs to the stack.
Performance testing with four TCP streams received on a single CPU (where
 throughput was line rate of 9.4Gbps in all tests) showed a 13.7% reduction
 in RX CPU usage (n=6, p=0.03).
Signed-off-by: default avatarEdward Cree <ecree@solarflare.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
parent 5e6d9fc7
...@@ -412,7 +412,6 @@ efx_rx_packet_gro(struct efx_channel *channel, struct efx_rx_buffer *rx_buf, ...@@ -412,7 +412,6 @@ efx_rx_packet_gro(struct efx_channel *channel, struct efx_rx_buffer *rx_buf,
unsigned int n_frags, u8 *eh) unsigned int n_frags, u8 *eh)
{ {
struct napi_struct *napi = &channel->napi_str; struct napi_struct *napi = &channel->napi_str;
gro_result_t gro_result;
struct efx_nic *efx = channel->efx; struct efx_nic *efx = channel->efx;
struct sk_buff *skb; struct sk_buff *skb;
...@@ -449,9 +448,7 @@ efx_rx_packet_gro(struct efx_channel *channel, struct efx_rx_buffer *rx_buf, ...@@ -449,9 +448,7 @@ efx_rx_packet_gro(struct efx_channel *channel, struct efx_rx_buffer *rx_buf,
skb_record_rx_queue(skb, channel->rx_queue.core_index); skb_record_rx_queue(skb, channel->rx_queue.core_index);
gro_result = napi_gro_frags(napi); napi_gro_frags(napi);
if (gro_result != GRO_DROP)
channel->irq_mod_score += 2;
} }
/* Allocate and construct an SKB around page fragments */ /* Allocate and construct an SKB around page fragments */
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment