p71924506
/
linux_old1
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

net/mlx4_en: avoid one cache line miss to ring doorbell 

This patch caches doorbell address directly in struct mlx4_en_tx_ring.

This removes the need to bring in cpu caches whole struct mlx4_uar
in fast path.

Note that mlx4_uar is not guaranteed to be on a local node,
because mlx4_bf_alloc() uses a single free list (priv->bf_list)
regardless of its node parameter.

This kind of change does matter in presence of light/moderate traffic.
In high stress, this read-only line would be kept hot in caches.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Tariq Toukan <tariqt@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This commit is contained in:
Eric Dumazet 2021-09-30 17:52:49 -07:00 committed by David S. Miller
parent 0693b27644
commit 9ac936276f
2 changed files with 3 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
drivers/net/ethernet/mellanox/mlx4/en_tx.c
View File

@ -130,6 +130,7 @@ int mlx4_en_create_tx_ring(struct mlx4_en_priv *priv,
ring->bf_enabled = !!(priv->pflags & ring->bf_enabled = !!(priv->pflags &
MLX4_EN_PRIV_FLAGS_BLUEFLAME); MLX4_EN_PRIV_FLAGS_BLUEFLAME);
} }
ring->doorbell_address = ring->bf.uar->map + MLX4_SEND_DOORBELL;
ring->hwtstamp_tx_type = priv->hwtstamp_config.tx_type; ring->hwtstamp_tx_type = priv->hwtstamp_config.tx_type;
ring->queue_index = queue_index; ring->queue_index = queue_index;
@ -753,8 +754,7 @@ void mlx4_en_xmit_doorbell(struct mlx4_en_tx_ring *ring)
#else #else
iowrite32be( iowrite32be(
#endif #endif
(__force u32)ring->doorbell_qpn, (__force u32)ring->doorbell_qpn, ring->doorbell_address);
ring->bf.uar->map + MLX4_SEND_DOORBELL);
} }
static void mlx4_en_tx_write_desc(struct mlx4_en_tx_ring *ring, static void mlx4_en_tx_write_desc(struct mlx4_en_tx_ring *ring,

1
drivers/net/ethernet/mellanox/mlx4/mlx4_en.h
View File

@ -283,6 +283,7 @@ struct mlx4_en_tx_ring {
struct mlx4_bf bf; struct mlx4_bf bf;
/* Following part should be mostly read */ /* Following part should be mostly read */
void __iomem *doorbell_address;
__be32 doorbell_qpn; __be32 doorbell_qpn;
__be32 mr_key; __be32 mr_key;
u32 size; /* number of TXBBs */ u32 size; /* number of TXBBs */