net/mlx4: Fix limiting number of IRQ's instead of RSS queues

This fix a performance bug introduced by commit 90b1ebe "mlx4: set
maximal number of default RSS queues", which limits the numbers of IRQs
opened by core module.
The limit should be on the number of queues in the indirection table -
rx_rings, and not on the number of IRQ's. Also, limiting on mlx4_core
initialization instead of in mlx4_en, prevented using "ethtool -L" to
utilize all the CPU's, when performance mode is prefered, since limiting
this number to 8 reduces overall packet rate by 15%-50% in multiple TCP
streams applications.

For example, after running ethtool -L <ethx> rx 16

          Packet rate
Before the fix  897799
After the fix   1142070

Results were obtained using netperf:

S=200 ; ( for i in $(seq 1 $S) ; do ( \
  netperf -H 11.7.13.55 -t TCP_RR -l 30 &) ; \
  wait ; done | grep "1        1" | awk '{SUM+=$6} END {print SUM}' )

CC: Yuval Mintz <yuvalmin@broadcom.com>
Signed-off-by: Ido Shamay <idos@mellanox.com>
Signed-off-by: Amir Vadai <amirv@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This commit is contained in:
Ido Shamay 2014-02-21 12:39:18 +02:00 committed by David S. Miller
parent 0251248232
commit bb2146bc88
2 changed files with 5 additions and 3 deletions

View File

@ -322,6 +322,7 @@ void mlx4_en_set_num_rx_rings(struct mlx4_en_dev *mdev)
{
int i;
int num_of_eqs;
int num_rx_rings;
struct mlx4_dev *dev = mdev->dev;
mlx4_foreach_port(i, dev, MLX4_PORT_TYPE_ETH) {
@ -335,8 +336,10 @@ void mlx4_en_set_num_rx_rings(struct mlx4_en_dev *mdev)
dev->caps.comp_pool/
dev->caps.num_ports) - 1;
num_rx_rings = min_t(int, num_of_eqs,
netif_get_num_default_rss_queues());
mdev->profile.prof[i].rx_ring_num =
rounddown_pow_of_two(num_of_eqs);
rounddown_pow_of_two(num_rx_rings);
}
}

View File

@ -41,7 +41,6 @@
#include <linux/slab.h>
#include <linux/io-mapping.h>
#include <linux/delay.h>
#include <linux/netdevice.h>
#include <linux/kmod.h>
#include <linux/mlx4/device.h>
@ -1974,7 +1973,7 @@ static void mlx4_enable_msi_x(struct mlx4_dev *dev)
struct mlx4_priv *priv = mlx4_priv(dev);
struct msix_entry *entries;
int nreq = min_t(int, dev->caps.num_ports *
min_t(int, netif_get_num_default_rss_queues() + 1,
min_t(int, num_online_cpus() + 1,
MAX_MSIX_P_PORT) + MSIX_LEGACY_SZ, MAX_MSIX);
int i;