Skip to contentSkip to conversation

vxge: prefetch RxD descriptors

PATCH In review Benjamin LaHaise 1 participant 1 message Aug 4, 2009

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH 4/6 vxge: prefetch RxD descriptors

This patch prefetches RxD descriptors which helps to lower the latency of a cache miss in vxge_hw_ring_rxd_next_completed. This lowers the % of CPU time used by vxge_hw_ring_rxd_next_completed() where the descriptor is accessed in profiling netperf on a P4 Xeon from 1.5% to 1.0%.

---

drivers/net/vxge/vxge-main.c renamed from drivers/net/vxge/vxge-main.c.orig +1 -0
@@ -445,6 +431,7 @@ vxge_rx_1b_compl(struct __vxge_hw_ring *
 	vxge_hw_ring_replenish(ringh, 0);
 
 	do {
+		prefetch((char *)dtr + L1_CACHE_BYTES);
 		rx_priv = vxge_hw_ring_rxd_private_get(dtr);
 		skb = rx_priv->skb;
 		data_size = rx_priv->data_size;
drivers/net/vxge/vxge-traffic.c +1 -0
index 7be0ae1..9a6b10d 100644
@@ -731,6 +731,7 @@ vxge_hw_channel_dtr_try_complete(struct __vxge_hw_channel *channel, void **dtrh)
 	vxge_assert(channel->compl_index < channel->length);
 
 	*dtrh =	channel->work_arr[channel->compl_index];
+	prefetch(*dtrh);
 }
 
 /*