net: stmmac: fix up RX flow hash indirection table when setting channels
Subsystems (4)
PATCH 1/1 net: stmmac: fix up RX flow hash indirection table when setting channels
stmmac_reinit_queues() fails to fix up the RX hash. Even if the number of channels gets restricted, the output of `ethtool -x' indicates that all RX queues are used:
$ ethtool -l enp0s29f2
Channel parameters for enp0s29f2:
Pre-set maximums:
RX: 8
TX: 8
Other: n/a
Combined: n/a
Current hardware settings:
RX: 8
TX: 8
Other: n/a
Combined: n/a
$ ethtool -x enp0s29f2
RX flow hash indirection table for enp0s29f2 with 8 RX ring(s):
0: 0 1 2 3 4 5 6 7
8: 0 1 2 3 4 5 6 7
[...]
$ ethtool -L enp0s29f2 rx 3
$ ethtool -x enp0s29f2
RX flow hash indirection table for enp0s29f2 with 3 RX ring(s):
0: 0 1 2 3 4 5 6 7
8: 0 1 2 3 4 5 6 7
[...]Fix this by setting the indirection table according to the number of specified queues. The result is now as expected:
$ ethtool -L enp0s29f2 rx 3
$ ethtool -x enp0s29f2
RX flow hash indirection table for enp0s29f2 with 3 RX ring(s):
0: 0 1 2 0 1 2 0 1
8: 2 0 1 2 0 1 2 0
[...]Tested on Intel Elkhart Lake.
--- drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/net/ethernet/stmicro/stmmac/stmmac_main.c +3 -1
@@ -6948,7 +6948,7 @@ static void stmmac_napi_del(struct net_device *dev)int stmmac_reinit_queues (struct net_device *dev,u32 rx_cnt,u32 tx_cnt) {struct stmmac_priv *priv =netdev_priv (dev); -int ret =0 ; +int ret =0 , i;if (netif_running (dev))stmmac_release (dev);
@@ -6957,6 +6957,8 @@ int stmmac_reinit_queues(struct net_device *dev, u32 rx_cnt, u32 tx_cnt) priv->plat->rx_queues_to_use = rx_cnt; priv->plat->tx_queues_to_use = tx_cnt; +for (i =0 ; i <ARRAY_SIZE (priv->rss.table); i++) + priv->rss.table[i] =ethtool_rxfh_indir_default (i, rx_cnt);stmmac_napi_add (dev);
Signature · 1 line
2.39.2
Jakub Kicinski
maintainerdrivers/net/ethernet/stmicro/stmmac/stmmac_main.c int stmmac_reinit_queues(struct net_device *dev, u32 rx_cnt, u32 tx_cnt) · line 6961
Quoted 4 lines · On Fri, 31 Mar 2023 11:23:47 +0200 Corinna Vinschen wrote:
drivers/net/ethernet/stmicro/stmmac/stmmac_main.c +2
-0priv->plat->rx_queues_to_use = rx_cnt; priv->plat->tx_queues_to_use = tx_cnt; +for (i =0 ; i <ARRAY_SIZE (priv->rss.table); i++) + priv->rss.table[i] =ethtool_rxfh_indir_default (i, rx_cnt);
You need to check if (!netif_is_rxfh_configured())
if user set the config to not RSS to all queues we shouldn't reset.
PATCH v2 1/1 net: stmmac: fix up RX flow hash indirection table when setting channels
stmmac_reinit_queues() fails to fix up the RX hash. Even if the number of channels gets restricted, the output of `ethtool -x' indicates that all RX queues are used:
$ ethtool -l enp0s29f2
Channel parameters for enp0s29f2:
Pre-set maximums:
RX: 8
TX: 8
Other: n/a
Combined: n/a
Current hardware settings:
RX: 8
TX: 8
Other: n/a
Combined: n/a
$ ethtool -x enp0s29f2
RX flow hash indirection table for enp0s29f2 with 8 RX ring(s):
0: 0 1 2 3 4 5 6 7
8: 0 1 2 3 4 5 6 7
[...]
$ ethtool -L enp0s29f2 rx 3
$ ethtool -x enp0s29f2
RX flow hash indirection table for enp0s29f2 with 3 RX ring(s):
0: 0 1 2 3 4 5 6 7
8: 0 1 2 3 4 5 6 7
[...]Fix this by setting the indirection table according to the number of specified queues. The result is now as expected:
$ ethtool -L enp0s29f2 rx 3
$ ethtool -x enp0s29f2
RX flow hash indirection table for enp0s29f2 with 3 RX ring(s):
0: 0 1 2 0 1 2 0 1
8: 2 0 1 2 0 1 2 0
[...]Tested on Intel Elkhart Lake.
--- drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-)
drivers/net/ethernet/stmicro/stmmac/stmmac_main.c +5 -1
@@ -6948,7 +6948,7 @@ static void stmmac_napi_del(struct net_device *dev)int stmmac_reinit_queues (struct net_device *dev,u32 rx_cnt,u32 tx_cnt) {struct stmmac_priv *priv =netdev_priv (dev); -int ret =0 ; +int ret =0 , i;if (netif_running (dev))stmmac_release (dev);
@@ -6957,6 +6957,10 @@ int stmmac_reinit_queues(struct net_device *dev, u32 rx_cnt, u32 tx_cnt) priv->plat->rx_queues_to_use = rx_cnt; priv->plat->tx_queues_to_use = tx_cnt; +if (!netif_is_rxfh_configured (dev)) +for (i =0 ; i <ARRAY_SIZE (priv->rss.table); i++) + priv->rss.table[i] =ethtool_rxfh_indir_default (i, + rx_cnt);stmmac_napi_add (dev);
Signature · 1 line
2.39.2
This patch was applied to netdev/net.git (main) by Paolo Abeni <[email protected]>:
Hello:
This patch was applied to netdev/net.git (main)
by Paolo Abeni <[email protected]>:
Quoted 28 lines · On Mon, 3 Apr 2023 14:11:20 +0200 you wrote:
stmmac_reinit_queues()fails to fix up the RX hash. Even if the number of channels gets restricted, the output of `ethtool -x' indicates that all RX queues are used:$ ethtool -l enp0s29f2 Channel parameters for enp0s29f2: Pre-set maximums: RX: 8 TX: 8 Other: n/a Combined: n/a Current hardware settings: RX: 8 TX: 8 Other: n/a Combined: n/a $ ethtool -x enp0s29f2 RX flow hash indirection table for enp0s29f2 with 8 RX ring(s): 0: 0 1 2 3 4 5 6 7 8: 0 1 2 3 4 5 6 7 [...] $ ethtool -L enp0s29f2 rx 3 $ ethtool -x enp0s29f2 RX flow hash indirection table for enp0s29f2 with 3 RX ring(s): 0: 0 1 2 3 4 5 6 7 8: 0 1 2 3 4 5 6 7 [...]
Quoted 1 line
[...]
Here is the summary with links:
- [v2,net] net: stmmac: fix up RX flow hash indirection table when setting channels
https://git.kernel.org/netdev/net/c/218c597325f4
You are awesome, thank you!
Signature · 2 lines
Deet-doot-dot, I am a bot. https://korg.docs.kernel.org/patchwork/pwbot.html