* [net-next 01/15] net/mlx5e: Fix spelling mistake "supoported" -> "supported"
2022-02-17 7:56 [pull request][net-next 00/15] mlx5 updates 2022-02-16 Saeed Mahameed
@ 2022-02-17 7:56 ` Saeed Mahameed
2022-02-17 11:10 ` patchwork-bot+netdevbpf
2022-02-17 7:56 ` [net-next 02/15] net/mlx5e: Add support for using xdp->data_meta Saeed Mahameed
` (13 subsequent siblings)
14 siblings, 1 reply; 17+ messages in thread
From: Saeed Mahameed @ 2022-02-17 7:56 UTC (permalink / raw)
To: David S. Miller, Jakub Kicinski
Cc: netdev, Colin Ian King, Roi Dayan, Saeed Mahameed
From: Colin Ian King <colin.i.king@gmail.com>
There is a spelling mistake in a NL_SET_ERR_MSG_MOD error
message. Fix it.
Fixes: 3b49a7edec1d ("net/mlx5e: TC, Reject rules with multiple CT actions")
Signed-off-by: Colin Ian King <colin.i.king@gmail.com>
Reviewed-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/ct.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/ct.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/ct.c
index 85f0cb88127f..9fb1a9a8bc02 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/ct.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/ct.c
@@ -21,7 +21,7 @@ tc_act_can_offload_ct(struct mlx5e_tc_act_parse_state *parse_state,
}
if (parse_state->ct && !clear_action) {
- NL_SET_ERR_MSG_MOD(extack, "Multiple CT actions are not supoported");
+ NL_SET_ERR_MSG_MOD(extack, "Multiple CT actions are not supported");
return false;
}
--
2.34.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* Re: [net-next 01/15] net/mlx5e: Fix spelling mistake "supoported" -> "supported"
2022-02-17 7:56 ` [net-next 01/15] net/mlx5e: Fix spelling mistake "supoported" -> "supported" Saeed Mahameed
@ 2022-02-17 11:10 ` patchwork-bot+netdevbpf
0 siblings, 0 replies; 17+ messages in thread
From: patchwork-bot+netdevbpf @ 2022-02-17 11:10 UTC (permalink / raw)
To: Saeed Mahameed; +Cc: davem, kuba, netdev, colin.i.king, roid, saeedm
Hello:
This series was applied to netdev/net-next.git (master)
by Saeed Mahameed <saeedm@nvidia.com>:
On Wed, 16 Feb 2022 23:56:18 -0800 you wrote:
> From: Colin Ian King <colin.i.king@gmail.com>
>
> There is a spelling mistake in a NL_SET_ERR_MSG_MOD error
> message. Fix it.
>
> Fixes: 3b49a7edec1d ("net/mlx5e: TC, Reject rules with multiple CT actions")
> Signed-off-by: Colin Ian King <colin.i.king@gmail.com>
> Reviewed-by: Roi Dayan <roid@nvidia.com>
> Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
>
> [...]
Here is the summary with links:
- [net-next,01/15] net/mlx5e: Fix spelling mistake "supoported" -> "supported"
https://git.kernel.org/netdev/net-next/c/9625bf39bd41
- [net-next,02/15] net/mlx5e: Add support for using xdp->data_meta
https://git.kernel.org/netdev/net-next/c/c1e80bf4ad3e
- [net-next,03/15] net/mlx5e: Generalize packet merge error message
https://git.kernel.org/netdev/net-next/c/b98d2d722f81
- [net-next,04/15] net/mlx5e: Default to Striding RQ when not conflicting with CQE compression
https://git.kernel.org/netdev/net-next/c/1d5024f88dad
- [net-next,05/15] net/mlx5e: RX, Restrict bulk size for small Striding RQs
https://git.kernel.org/netdev/net-next/c/4b5fba4a3ac7
- [net-next,06/15] net/mlx5e: E-Switch, Add PTP counters for uplink representor
https://git.kernel.org/netdev/net-next/c/7c5f940d264e
- [net-next,07/15] net/mlx5e: E-Switch, Add support for tx_port_ts in switchdev mode
https://git.kernel.org/netdev/net-next/c/bfbdd77ac52f
- [net-next,08/15] net/mlx5e: TC, Move flow hashtable to be per rep
https://git.kernel.org/netdev/net-next/c/d1a3138f7913
- [net-next,09/15] net/mlx5e: Pass actions param to actions_match_supported()
https://git.kernel.org/netdev/net-next/c/0610f8dc0309
- [net-next,10/15] net/mlx5e: Add post act offload/unoffload API
https://git.kernel.org/netdev/net-next/c/314e1105831b
- [net-next,11/15] net/mlx5e: Create new flow attr for multi table actions
https://git.kernel.org/netdev/net-next/c/8300f225268b
- [net-next,12/15] net/mlx5e: Use multi table support for CT and sample actions
https://git.kernel.org/netdev/net-next/c/a81283263bb0
- [net-next,13/15] net/mlx5e: TC, Clean redundant counter flag from tc action parsers
https://git.kernel.org/netdev/net-next/c/2a829fe25d28
- [net-next,14/15] net/mlx5e: TC, Make post_act parse CT and sample actions
https://git.kernel.org/netdev/net-next/c/7843bd604081
- [net-next,15/15] net/mlx5e: TC, Allow sample action with CT
https://git.kernel.org/netdev/net-next/c/b070e70381ee
You are awesome, thank you!
--
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html
^ permalink raw reply [flat|nested] 17+ messages in thread
* [net-next 02/15] net/mlx5e: Add support for using xdp->data_meta
2022-02-17 7:56 [pull request][net-next 00/15] mlx5 updates 2022-02-16 Saeed Mahameed
2022-02-17 7:56 ` [net-next 01/15] net/mlx5e: Fix spelling mistake "supoported" -> "supported" Saeed Mahameed
@ 2022-02-17 7:56 ` Saeed Mahameed
2022-02-17 7:56 ` [net-next 03/15] net/mlx5e: Generalize packet merge error message Saeed Mahameed
` (12 subsequent siblings)
14 siblings, 0 replies; 17+ messages in thread
From: Saeed Mahameed @ 2022-02-17 7:56 UTC (permalink / raw)
To: David S. Miller, Jakub Kicinski; +Cc: netdev, Alex Liu, Saeed Mahameed
From: Alex Liu <liualex@fb.com>
Add support for using xdp->data_meta for cross-program communication
Pass "true" to the last argument of xdp_prepare_buff().
After SKB is built, call skb_metadata_set() if metadata was pushed.
Signed-off-by: Alex Liu <liualex@fb.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en_rx.c | 17 ++++++++++++-----
1 file changed, 12 insertions(+), 5 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
index 91fdf957cd7c..3fe4f06f3e71 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
@@ -1489,7 +1489,7 @@ static inline void mlx5e_complete_rx_cqe(struct mlx5e_rq *rq,
static inline
struct sk_buff *mlx5e_build_linear_skb(struct mlx5e_rq *rq, void *va,
u32 frag_size, u16 headroom,
- u32 cqe_bcnt)
+ u32 cqe_bcnt, u32 metasize)
{
struct sk_buff *skb = build_skb(va, frag_size);
@@ -1501,6 +1501,9 @@ struct sk_buff *mlx5e_build_linear_skb(struct mlx5e_rq *rq, void *va,
skb_reserve(skb, headroom);
skb_put(skb, cqe_bcnt);
+ if (metasize)
+ skb_metadata_set(skb, metasize);
+
return skb;
}
@@ -1508,7 +1511,7 @@ static void mlx5e_fill_xdp_buff(struct mlx5e_rq *rq, void *va, u16 headroom,
u32 len, struct xdp_buff *xdp)
{
xdp_init_buff(xdp, rq->buff.frame0_sz, &rq->xdp_rxq);
- xdp_prepare_buff(xdp, va, headroom, len, false);
+ xdp_prepare_buff(xdp, va, headroom, len, true);
}
static struct sk_buff *
@@ -1521,6 +1524,7 @@ mlx5e_skb_from_cqe_linear(struct mlx5e_rq *rq, struct mlx5_cqe64 *cqe,
struct sk_buff *skb;
void *va, *data;
u32 frag_size;
+ u32 metasize;
va = page_address(di->page) + wi->offset;
data = va + rx_headroom;
@@ -1537,7 +1541,8 @@ mlx5e_skb_from_cqe_linear(struct mlx5e_rq *rq, struct mlx5_cqe64 *cqe,
rx_headroom = xdp.data - xdp.data_hard_start;
frag_size = MLX5_SKB_FRAG_SZ(rx_headroom + cqe_bcnt);
- skb = mlx5e_build_linear_skb(rq, va, frag_size, rx_headroom, cqe_bcnt);
+ metasize = xdp.data - xdp.data_meta;
+ skb = mlx5e_build_linear_skb(rq, va, frag_size, rx_headroom, cqe_bcnt, metasize);
if (unlikely(!skb))
return NULL;
@@ -1836,6 +1841,7 @@ mlx5e_skb_from_cqe_mpwrq_linear(struct mlx5e_rq *rq, struct mlx5e_mpw_info *wi,
struct sk_buff *skb;
void *va, *data;
u32 frag_size;
+ u32 metasize;
/* Check packet size. Note LRO doesn't use linear SKB */
if (unlikely(cqe_bcnt > rq->hw_mtu)) {
@@ -1861,7 +1867,8 @@ mlx5e_skb_from_cqe_mpwrq_linear(struct mlx5e_rq *rq, struct mlx5e_mpw_info *wi,
rx_headroom = xdp.data - xdp.data_hard_start;
frag_size = MLX5_SKB_FRAG_SZ(rx_headroom + cqe_bcnt32);
- skb = mlx5e_build_linear_skb(rq, va, frag_size, rx_headroom, cqe_bcnt32);
+ metasize = xdp.data - xdp.data_meta;
+ skb = mlx5e_build_linear_skb(rq, va, frag_size, rx_headroom, cqe_bcnt32, metasize);
if (unlikely(!skb))
return NULL;
@@ -1892,7 +1899,7 @@ mlx5e_skb_from_cqe_shampo(struct mlx5e_rq *rq, struct mlx5e_mpw_info *wi,
dma_sync_single_range_for_cpu(rq->pdev, head->addr, 0, frag_size, DMA_FROM_DEVICE);
prefetchw(hdr);
prefetch(data);
- skb = mlx5e_build_linear_skb(rq, hdr, frag_size, rx_headroom, head_size);
+ skb = mlx5e_build_linear_skb(rq, hdr, frag_size, rx_headroom, head_size, 0);
if (unlikely(!skb))
return NULL;
--
2.34.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* [net-next 03/15] net/mlx5e: Generalize packet merge error message
2022-02-17 7:56 [pull request][net-next 00/15] mlx5 updates 2022-02-16 Saeed Mahameed
2022-02-17 7:56 ` [net-next 01/15] net/mlx5e: Fix spelling mistake "supoported" -> "supported" Saeed Mahameed
2022-02-17 7:56 ` [net-next 02/15] net/mlx5e: Add support for using xdp->data_meta Saeed Mahameed
@ 2022-02-17 7:56 ` Saeed Mahameed
2022-02-17 7:56 ` [net-next 04/15] net/mlx5e: Default to Striding RQ when not conflicting with CQE compression Saeed Mahameed
` (11 subsequent siblings)
14 siblings, 0 replies; 17+ messages in thread
From: Saeed Mahameed @ 2022-02-17 7:56 UTC (permalink / raw)
To: David S. Miller, Jakub Kicinski
Cc: netdev, Tariq Toukan, Khalid Manaa, Gal Pressman, Saeed Mahameed
From: Tariq Toukan <tariqt@nvidia.com>
Update the old error message for LRO state modify with the new
general name.
Signed-off-by: Tariq Toukan <tariqt@nvidia.com>
Reviewed-by: Khalid Manaa <khalidm@nvidia.com>
Reviewed-by: Gal Pressman <gal@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
| 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
--git a/drivers/net/ethernet/mellanox/mlx5/core/en/rss.c b/drivers/net/ethernet/mellanox/mlx5/core/en/rss.c
index c1cdd8c2e37a..7f93426b88b3 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/rss.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/rss.c
@@ -442,7 +442,7 @@ int mlx5e_rss_packet_merge_set_param(struct mlx5e_rss *rss,
goto inner_tir;
err = mlx5e_tir_modify(tir, builder);
if (err) {
- mlx5e_rss_warn(rss->mdev, "Failed to update LRO state of indirect TIR %#x for traffic type %d: err = %d\n",
+ mlx5e_rss_warn(rss->mdev, "Failed to update packet merge state of indirect TIR %#x for traffic type %d: err = %d\n",
mlx5e_tir_get_tirn(tir), tt, err);
if (!final_err)
final_err = err;
@@ -457,7 +457,7 @@ int mlx5e_rss_packet_merge_set_param(struct mlx5e_rss *rss,
continue;
err = mlx5e_tir_modify(tir, builder);
if (err) {
- mlx5e_rss_warn(rss->mdev, "Failed to update LRO state of inner indirect TIR %#x for traffic type %d: err = %d\n",
+ mlx5e_rss_warn(rss->mdev, "Failed to update packet merge state of inner indirect TIR %#x for traffic type %d: err = %d\n",
mlx5e_tir_get_tirn(tir), tt, err);
if (!final_err)
final_err = err;
--
2.34.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* [net-next 04/15] net/mlx5e: Default to Striding RQ when not conflicting with CQE compression
2022-02-17 7:56 [pull request][net-next 00/15] mlx5 updates 2022-02-16 Saeed Mahameed
` (2 preceding siblings ...)
2022-02-17 7:56 ` [net-next 03/15] net/mlx5e: Generalize packet merge error message Saeed Mahameed
@ 2022-02-17 7:56 ` Saeed Mahameed
2022-02-17 7:56 ` [net-next 05/15] net/mlx5e: RX, Restrict bulk size for small Striding RQs Saeed Mahameed
` (10 subsequent siblings)
14 siblings, 0 replies; 17+ messages in thread
From: Saeed Mahameed @ 2022-02-17 7:56 UTC (permalink / raw)
To: David S. Miller, Jakub Kicinski
Cc: netdev, Tariq Toukan, Gal Pressman, Saeed Mahameed
From: Tariq Toukan <tariqt@nvidia.com>
CQE compression is turned on by default on slow pci systems to help
reduce the load on pci.
In this case, Striding RQ was turned off as CQEs of packets that span
several strides were not compressed, significantly reducing the compression
effectiveness.
This issue does not exist when using the newer mini_cqe format "stride_index".
Hence, allow defaulting to Striding RQ in this case.
Signed-off-by: Tariq Toukan <tariqt@nvidia.com>
Reviewed-by: Gal Pressman <gal@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en/params.c | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/params.c b/drivers/net/ethernet/mellanox/mlx5/core/en/params.c
index d41936d65483..2d9707c024b0 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/params.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/params.c
@@ -359,12 +359,13 @@ void mlx5e_build_rq_params(struct mlx5_core_dev *mdev,
{
/* Prefer Striding RQ, unless any of the following holds:
* - Striding RQ configuration is not possible/supported.
- * - Slow PCI heuristic.
+ * - CQE compression is ON, and stride_index mini_cqe layout is not supported.
* - Legacy RQ would use linear SKB while Striding RQ would use non-linear.
*
* No XSK params: checking the availability of striding RQ in general.
*/
- if (!slow_pci_heuristic(mdev) &&
+ if ((!MLX5E_GET_PFLAG(params, MLX5E_PFLAG_RX_CQE_COMPRESS) ||
+ MLX5_CAP_GEN(mdev, mini_cqe_resp_stride_index)) &&
mlx5e_striding_rq_possible(mdev, params) &&
(mlx5e_rx_mpwqe_is_linear_skb(mdev, params, NULL) ||
!mlx5e_rx_is_linear_skb(params, NULL)))
--
2.34.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* [net-next 05/15] net/mlx5e: RX, Restrict bulk size for small Striding RQs
2022-02-17 7:56 [pull request][net-next 00/15] mlx5 updates 2022-02-16 Saeed Mahameed
` (3 preceding siblings ...)
2022-02-17 7:56 ` [net-next 04/15] net/mlx5e: Default to Striding RQ when not conflicting with CQE compression Saeed Mahameed
@ 2022-02-17 7:56 ` Saeed Mahameed
2022-02-17 7:56 ` [net-next 06/15] net/mlx5e: E-Switch, Add PTP counters for uplink representor Saeed Mahameed
` (9 subsequent siblings)
14 siblings, 0 replies; 17+ messages in thread
From: Saeed Mahameed @ 2022-02-17 7:56 UTC (permalink / raw)
To: David S. Miller, Jakub Kicinski; +Cc: netdev, Tariq Toukan, Saeed Mahameed
From: Tariq Toukan <tariqt@nvidia.com>
In RQs of type multi-packet WQE (Striding RQ), each WQE is relatively
large (typically 256KB) but their number is relatively small (8 in
default).
Re-mapping the descriptors' buffers before re-posting them is done via
UMR (User-Mode Memory Registration) operations.
On the one hand, posting UMR WQEs in bulks reduces communication overhead
with the HW and better utilizes its processing units.
On the other hand, delaying the WQE repost operations for a small RQ
(say, of 4 WQEs) might drastically hit its performance, causing packet
drops due to no receive buffer, for high or bursty incoming packets rate.
Here we restrict the bulk size for too small RQs. Effectively, with the current
constants, RQ of size 4 (minimum allowed) would have no bulking, while larger
RQs will continue working with bulks of 2.
Signed-off-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en.h | 1 +
drivers/net/ethernet/mellanox/mlx5/core/en/params.c | 6 ++++++
drivers/net/ethernet/mellanox/mlx5/core/en/params.h | 1 +
drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 1 +
drivers/net/ethernet/mellanox/mlx5/core/en_rx.c | 3 +--
5 files changed, 10 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h
index e29ac77e9bec..2704c7537481 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h
@@ -714,6 +714,7 @@ struct mlx5e_rq {
u8 umr_in_progress;
u8 umr_last_bulk;
u8 umr_completed;
+ u8 min_wqe_bulk;
struct mlx5e_shampo_hd *shampo;
} mpwqe;
};
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/params.c b/drivers/net/ethernet/mellanox/mlx5/core/en/params.c
index 2d9707c024b0..0bd8698f7226 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/params.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/params.c
@@ -178,6 +178,12 @@ u8 mlx5e_mpwqe_get_log_num_strides(struct mlx5_core_dev *mdev,
mlx5e_mpwqe_get_log_stride_size(mdev, params, xsk);
}
+u8 mlx5e_mpwqe_get_min_wqe_bulk(unsigned int wq_sz)
+{
+#define UMR_WQE_BULK (2)
+ return min_t(unsigned int, UMR_WQE_BULK, wq_sz / 2 - 1);
+}
+
u16 mlx5e_get_rq_headroom(struct mlx5_core_dev *mdev,
struct mlx5e_params *params,
struct mlx5e_xsk_param *xsk)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/params.h b/drivers/net/ethernet/mellanox/mlx5/core/en/params.h
index 433e6967692d..47a368112e31 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/params.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/params.h
@@ -129,6 +129,7 @@ u8 mlx5e_mpwqe_get_log_stride_size(struct mlx5_core_dev *mdev,
u8 mlx5e_mpwqe_get_log_num_strides(struct mlx5_core_dev *mdev,
struct mlx5e_params *params,
struct mlx5e_xsk_param *xsk);
+u8 mlx5e_mpwqe_get_min_wqe_bulk(unsigned int wq_sz);
u16 mlx5e_get_rq_headroom(struct mlx5_core_dev *mdev,
struct mlx5e_params *params,
struct mlx5e_xsk_param *xsk);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index b157c7aac4ca..b2ed2f6d4a92 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -595,6 +595,7 @@ static int mlx5e_alloc_rq(struct mlx5e_params *params,
rq->mpwqe.log_stride_sz = mlx5e_mpwqe_get_log_stride_size(mdev, params, xsk);
rq->mpwqe.num_strides =
BIT(mlx5e_mpwqe_get_log_num_strides(mdev, params, xsk));
+ rq->mpwqe.min_wqe_bulk = mlx5e_mpwqe_get_min_wqe_bulk(wq_sz);
rq->buff.frame0_sz = (1 << rq->mpwqe.log_stride_sz);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
index 3fe4f06f3e71..4cb7c7135b6a 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
@@ -960,8 +960,7 @@ INDIRECT_CALLABLE_SCOPE bool mlx5e_post_rx_mpwqes(struct mlx5e_rq *rq)
if (unlikely(rq->mpwqe.umr_in_progress > rq->mpwqe.umr_last_bulk))
rq->stats->congst_umr++;
-#define UMR_WQE_BULK (2)
- if (likely(missing < UMR_WQE_BULK))
+ if (likely(missing < rq->mpwqe.min_wqe_bulk))
return false;
if (rq->page_pool)
--
2.34.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* [net-next 06/15] net/mlx5e: E-Switch, Add PTP counters for uplink representor
2022-02-17 7:56 [pull request][net-next 00/15] mlx5 updates 2022-02-16 Saeed Mahameed
` (4 preceding siblings ...)
2022-02-17 7:56 ` [net-next 05/15] net/mlx5e: RX, Restrict bulk size for small Striding RQs Saeed Mahameed
@ 2022-02-17 7:56 ` Saeed Mahameed
2022-02-17 7:56 ` [net-next 07/15] net/mlx5e: E-Switch, Add support for tx_port_ts in switchdev mode Saeed Mahameed
` (8 subsequent siblings)
14 siblings, 0 replies; 17+ messages in thread
From: Saeed Mahameed @ 2022-02-17 7:56 UTC (permalink / raw)
To: David S. Miller, Jakub Kicinski
Cc: netdev, Aya Levin, Tariq Toukan, Maor Dickman, Saeed Mahameed
From: Aya Levin <ayal@nvidia.com>
There is a configuration where the uplink interface is the synchronizer.
Add PTP counters for this interface for monitoring.
Signed-off-by: Aya Levin <ayal@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Reviewed-by: Maor Dickman <maord@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en_rep.c | 1 +
drivers/net/ethernet/mellanox/mlx5/core/en_stats.c | 2 +-
drivers/net/ethernet/mellanox/mlx5/core/en_stats.h | 1 +
3 files changed, 3 insertions(+), 1 deletion(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
index a3f257623af4..4459e5585075 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
@@ -1094,6 +1094,7 @@ static mlx5e_stats_grp_t mlx5e_ul_rep_stats_grps[] = {
&MLX5E_STATS_GRP(ipsec_sw),
&MLX5E_STATS_GRP(ipsec_hw),
#endif
+ &MLX5E_STATS_GRP(ptp),
};
static unsigned int mlx5e_ul_rep_stats_grps_num(struct mlx5e_priv *priv)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c
index 26e326fe503c..3e5d8c788026 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c
@@ -2348,7 +2348,7 @@ MLX5E_DEFINE_STATS_GRP(channels, 0);
MLX5E_DEFINE_STATS_GRP(per_port_buff_congest, 0);
MLX5E_DEFINE_STATS_GRP(eth_ext, 0);
static MLX5E_DEFINE_STATS_GRP(tls, 0);
-static MLX5E_DEFINE_STATS_GRP(ptp, 0);
+MLX5E_DEFINE_STATS_GRP(ptp, 0);
static MLX5E_DEFINE_STATS_GRP(qos, 0);
/* The stats groups order is opposite to the update_stats() order calls */
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h
index 2c1ed5b81be6..14eaf923e7b3 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h
@@ -459,5 +459,6 @@ extern MLX5E_DECLARE_STATS_GRP(channels);
extern MLX5E_DECLARE_STATS_GRP(per_port_buff_congest);
extern MLX5E_DECLARE_STATS_GRP(ipsec_hw);
extern MLX5E_DECLARE_STATS_GRP(ipsec_sw);
+extern MLX5E_DECLARE_STATS_GRP(ptp);
#endif /* __MLX5_EN_STATS_H__ */
--
2.34.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* [net-next 07/15] net/mlx5e: E-Switch, Add support for tx_port_ts in switchdev mode
2022-02-17 7:56 [pull request][net-next 00/15] mlx5 updates 2022-02-16 Saeed Mahameed
` (5 preceding siblings ...)
2022-02-17 7:56 ` [net-next 06/15] net/mlx5e: E-Switch, Add PTP counters for uplink representor Saeed Mahameed
@ 2022-02-17 7:56 ` Saeed Mahameed
2022-02-17 7:56 ` [net-next 08/15] net/mlx5e: TC, Move flow hashtable to be per rep Saeed Mahameed
` (7 subsequent siblings)
14 siblings, 0 replies; 17+ messages in thread
From: Saeed Mahameed @ 2022-02-17 7:56 UTC (permalink / raw)
To: David S. Miller, Jakub Kicinski
Cc: netdev, Aya Levin, Tariq Toukan, Maor Dickman, Saeed Mahameed
From: Aya Levin <ayal@nvidia.com>
When turning on tx_port_ts (private flag) a PTP-SQ is created. Consider
this queue when adding rules matching SQs to VPORTs. Otherwise the
traffic on this queue won't reach the wire.
Signed-off-by: Aya Levin <ayal@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Reviewed-by: Maor Dickman <maord@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
.../net/ethernet/mellanox/mlx5/core/en_rep.c | 18 +++++++++++++++---
1 file changed, 15 insertions(+), 3 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
index 4459e5585075..6070b8a13818 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
@@ -55,6 +55,7 @@
#include "diag/en_rep_tracepoint.h"
#include "en_accel/ipsec.h"
#include "en/tc/int_port.h"
+#include "en/ptp.h"
#define MLX5E_REP_PARAMS_DEF_LOG_SQ_SIZE \
max(0x7, MLX5E_PARAMS_MINIMUM_LOG_SQ_SIZE)
@@ -401,13 +402,18 @@ int mlx5e_add_sqs_fwd_rules(struct mlx5e_priv *priv)
struct mlx5_eswitch *esw = priv->mdev->priv.eswitch;
struct mlx5e_rep_priv *rpriv = priv->ppriv;
struct mlx5_eswitch_rep *rep = rpriv->rep;
+ int n, tc, nch, num_sqs = 0;
struct mlx5e_channel *c;
- int n, tc, num_sqs = 0;
int err = -ENOMEM;
+ bool ptp_sq;
u32 *sqs;
- sqs = kcalloc(priv->channels.num * mlx5e_get_dcb_num_tc(&priv->channels.params),
- sizeof(*sqs), GFP_KERNEL);
+ ptp_sq = !!(priv->channels.ptp &&
+ MLX5E_GET_PFLAG(&priv->channels.params, MLX5E_PFLAG_TX_PORT_TS));
+ nch = priv->channels.num + ptp_sq;
+
+ sqs = kcalloc(nch * mlx5e_get_dcb_num_tc(&priv->channels.params), sizeof(*sqs),
+ GFP_KERNEL);
if (!sqs)
goto out;
@@ -416,6 +422,12 @@ int mlx5e_add_sqs_fwd_rules(struct mlx5e_priv *priv)
for (tc = 0; tc < c->num_tc; tc++)
sqs[num_sqs++] = c->sq[tc].sqn;
}
+ if (ptp_sq) {
+ struct mlx5e_ptp *ptp_ch = priv->channels.ptp;
+
+ for (tc = 0; tc < ptp_ch->num_tc; tc++)
+ sqs[num_sqs++] = ptp_ch->ptpsq[tc].txqsq.sqn;
+ }
err = mlx5e_sqs2vport_start(esw, rep, sqs, num_sqs);
kfree(sqs);
--
2.34.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* [net-next 08/15] net/mlx5e: TC, Move flow hashtable to be per rep
2022-02-17 7:56 [pull request][net-next 00/15] mlx5 updates 2022-02-16 Saeed Mahameed
` (6 preceding siblings ...)
2022-02-17 7:56 ` [net-next 07/15] net/mlx5e: E-Switch, Add support for tx_port_ts in switchdev mode Saeed Mahameed
@ 2022-02-17 7:56 ` Saeed Mahameed
2022-02-17 7:56 ` [net-next 09/15] net/mlx5e: Pass actions param to actions_match_supported() Saeed Mahameed
` (6 subsequent siblings)
14 siblings, 0 replies; 17+ messages in thread
From: Saeed Mahameed @ 2022-02-17 7:56 UTC (permalink / raw)
To: David S. Miller, Jakub Kicinski
Cc: netdev, Paul Blakey, Roi Dayan, Saeed Mahameed
From: Paul Blakey <paulb@nvidia.com>
To allow shared tc block offload between two or more reps of the
same eswitch, move the tc flow hashtable to be per rep, instead
of per eswitch.
Signed-off-by: Paul Blakey <paulb@nvidia.com>
Reviewed-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
.../ethernet/mellanox/mlx5/core/en/rep/tc.c | 4 +-
.../net/ethernet/mellanox/mlx5/core/en_rep.c | 12 ++++-
.../net/ethernet/mellanox/mlx5/core/en_rep.h | 6 +--
.../net/ethernet/mellanox/mlx5/core/en_tc.c | 44 ++++++++++---------
.../net/ethernet/mellanox/mlx5/core/en_tc.h | 9 +++-
5 files changed, 43 insertions(+), 32 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/rep/tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en/rep/tc.c
index 0991345c4ae5..86fa0bdbee36 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/rep/tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/rep/tc.c
@@ -263,14 +263,14 @@ int mlx5e_rep_tc_init(struct mlx5e_rep_priv *rpriv)
INIT_LIST_HEAD(&uplink_priv->unready_flows);
/* init shared tc flow table */
- err = mlx5e_tc_esw_init(&uplink_priv->tc_ht);
+ err = mlx5e_tc_esw_init(uplink_priv);
return err;
}
void mlx5e_rep_tc_cleanup(struct mlx5e_rep_priv *rpriv)
{
/* delete shared tc flow table */
- mlx5e_tc_esw_cleanup(&rpriv->uplink_priv.tc_ht);
+ mlx5e_tc_esw_cleanup(&rpriv->uplink_priv);
mutex_destroy(&rpriv->uplink_priv.unready_flows_lock);
}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
index 6070b8a13818..6b7e7ea6ded2 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
@@ -942,15 +942,21 @@ static int mlx5e_init_rep_tx(struct mlx5e_priv *priv)
return err;
}
+ err = mlx5e_tc_ht_init(&rpriv->tc_ht);
+ if (err)
+ goto err_ht_init;
+
if (rpriv->rep->vport == MLX5_VPORT_UPLINK) {
err = mlx5e_init_uplink_rep_tx(rpriv);
if (err)
- goto destroy_tises;
+ goto err_init_tx;
}
return 0;
-destroy_tises:
+err_init_tx:
+ mlx5e_tc_ht_cleanup(&rpriv->tc_ht);
+err_ht_init:
mlx5e_destroy_tises(priv);
return err;
}
@@ -970,6 +976,8 @@ static void mlx5e_cleanup_rep_tx(struct mlx5e_priv *priv)
if (rpriv->rep->vport == MLX5_VPORT_UPLINK)
mlx5e_cleanup_uplink_rep_tx(rpriv);
+
+ mlx5e_tc_ht_cleanup(&rpriv->tc_ht);
}
static void mlx5e_rep_enable(struct mlx5e_priv *priv)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h
index b01dacb6f527..0b619c7846d0 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h
@@ -64,11 +64,6 @@ struct mlx5e_tc_tun_encap;
struct mlx5e_post_act;
struct mlx5_rep_uplink_priv {
- /* Filters DB - instantiated by the uplink representor and shared by
- * the uplink's VFs
- */
- struct rhashtable tc_ht;
-
/* indirect block callbacks are invoked on bind/unbind events
* on registered higher level devices (e.g. tunnel devices)
*
@@ -113,6 +108,7 @@ struct mlx5e_rep_priv {
struct list_head vport_sqs_list;
struct mlx5_rep_uplink_priv uplink_priv; /* valid for uplink rep */
struct rtnl_link_stats64 prev_vf_vport_stats;
+ struct rhashtable tc_ht;
};
static inline
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index 099d4ce16049..342ab0688f13 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -3609,12 +3609,11 @@ static const struct rhashtable_params tc_ht_params = {
static struct rhashtable *get_tc_ht(struct mlx5e_priv *priv,
unsigned long flags)
{
- struct mlx5_eswitch *esw = priv->mdev->priv.eswitch;
- struct mlx5e_rep_priv *uplink_rpriv;
+ struct mlx5e_rep_priv *rpriv;
if (flags & MLX5_TC_FLAG(ESW_OFFLOAD)) {
- uplink_rpriv = mlx5_eswitch_get_uplink_priv(esw, REP_ETH);
- return &uplink_rpriv->uplink_priv.tc_ht;
+ rpriv = priv->ppriv;
+ return &rpriv->tc_ht;
} else /* NIC offload */
return &priv->fs.tc.ht;
}
@@ -4447,10 +4446,27 @@ void mlx5e_tc_nic_cleanup(struct mlx5e_priv *priv)
mlx5_chains_destroy(tc->chains);
}
-int mlx5e_tc_esw_init(struct rhashtable *tc_ht)
+int mlx5e_tc_ht_init(struct rhashtable *tc_ht)
+{
+ int err;
+
+ err = rhashtable_init(tc_ht, &tc_ht_params);
+ if (err)
+ return err;
+
+ lockdep_set_class(&tc_ht->mutex, &tc_ht_lock_key);
+
+ return 0;
+}
+
+void mlx5e_tc_ht_cleanup(struct rhashtable *tc_ht)
+{
+ rhashtable_free_and_destroy(tc_ht, _mlx5e_tc_del_flow, NULL);
+}
+
+int mlx5e_tc_esw_init(struct mlx5_rep_uplink_priv *uplink_priv)
{
const size_t sz_enc_opts = sizeof(struct tunnel_match_enc_opts);
- struct mlx5_rep_uplink_priv *uplink_priv;
struct mlx5e_rep_priv *rpriv;
struct mapping_ctx *mapping;
struct mlx5_eswitch *esw;
@@ -4458,7 +4474,6 @@ int mlx5e_tc_esw_init(struct rhashtable *tc_ht)
u64 mapping_id;
int err = 0;
- uplink_priv = container_of(tc_ht, struct mlx5_rep_uplink_priv, tc_ht);
rpriv = container_of(uplink_priv, struct mlx5e_rep_priv, uplink_priv);
priv = netdev_priv(rpriv->netdev);
esw = priv->mdev->priv.eswitch;
@@ -4498,12 +4513,6 @@ int mlx5e_tc_esw_init(struct rhashtable *tc_ht)
}
uplink_priv->tunnel_enc_opts_mapping = mapping;
- err = rhashtable_init(tc_ht, &tc_ht_params);
- if (err)
- goto err_ht_init;
-
- lockdep_set_class(&tc_ht->mutex, &tc_ht_lock_key);
-
uplink_priv->encap = mlx5e_tc_tun_init(priv);
if (IS_ERR(uplink_priv->encap)) {
err = PTR_ERR(uplink_priv->encap);
@@ -4513,8 +4522,6 @@ int mlx5e_tc_esw_init(struct rhashtable *tc_ht)
return 0;
err_register_fib_notifier:
- rhashtable_destroy(tc_ht);
-err_ht_init:
mapping_destroy(uplink_priv->tunnel_enc_opts_mapping);
err_enc_opts_mapping:
mapping_destroy(uplink_priv->tunnel_mapping);
@@ -4528,13 +4535,8 @@ int mlx5e_tc_esw_init(struct rhashtable *tc_ht)
return err;
}
-void mlx5e_tc_esw_cleanup(struct rhashtable *tc_ht)
+void mlx5e_tc_esw_cleanup(struct mlx5_rep_uplink_priv *uplink_priv)
{
- struct mlx5_rep_uplink_priv *uplink_priv;
-
- uplink_priv = container_of(tc_ht, struct mlx5_rep_uplink_priv, tc_ht);
-
- rhashtable_free_and_destroy(tc_ht, _mlx5e_tc_del_flow, NULL);
mlx5e_tc_tun_cleanup(uplink_priv->encap);
mapping_destroy(uplink_priv->tunnel_enc_opts_mapping);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.h b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.h
index c6221728b767..533c897bd517 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.h
@@ -167,8 +167,11 @@ enum {
#define MLX5_TC_FLAG(flag) BIT(MLX5E_TC_FLAG_##flag##_BIT)
-int mlx5e_tc_esw_init(struct rhashtable *tc_ht);
-void mlx5e_tc_esw_cleanup(struct rhashtable *tc_ht);
+int mlx5e_tc_esw_init(struct mlx5_rep_uplink_priv *uplink_priv);
+void mlx5e_tc_esw_cleanup(struct mlx5_rep_uplink_priv *uplink_priv);
+
+int mlx5e_tc_ht_init(struct rhashtable *tc_ht);
+void mlx5e_tc_ht_cleanup(struct rhashtable *tc_ht);
int mlx5e_configure_flower(struct net_device *dev, struct mlx5e_priv *priv,
struct flow_cls_offload *f, unsigned long flags);
@@ -304,6 +307,8 @@ int mlx5e_set_fwd_to_int_port_actions(struct mlx5e_priv *priv,
#else /* CONFIG_MLX5_CLS_ACT */
static inline int mlx5e_tc_nic_init(struct mlx5e_priv *priv) { return 0; }
static inline void mlx5e_tc_nic_cleanup(struct mlx5e_priv *priv) {}
+static inline int mlx5e_tc_ht_init(struct rhashtable *tc_ht) { return 0; }
+static inline void mlx5e_tc_ht_cleanup(struct rhashtable *tc_ht) {}
static inline int
mlx5e_setup_tc_block_cb(enum tc_setup_type type, void *type_data, void *cb_priv)
{ return -EOPNOTSUPP; }
--
2.34.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* [net-next 09/15] net/mlx5e: Pass actions param to actions_match_supported()
2022-02-17 7:56 [pull request][net-next 00/15] mlx5 updates 2022-02-16 Saeed Mahameed
` (7 preceding siblings ...)
2022-02-17 7:56 ` [net-next 08/15] net/mlx5e: TC, Move flow hashtable to be per rep Saeed Mahameed
@ 2022-02-17 7:56 ` Saeed Mahameed
2022-02-17 7:56 ` [net-next 10/15] net/mlx5e: Add post act offload/unoffload API Saeed Mahameed
` (5 subsequent siblings)
14 siblings, 0 replies; 17+ messages in thread
From: Saeed Mahameed @ 2022-02-17 7:56 UTC (permalink / raw)
To: David S. Miller, Jakub Kicinski; +Cc: netdev, Roi Dayan, Saeed Mahameed
From: Roi Dayan <roid@nvidia.com>
Currently the mlx5_flow object contains a single mlx5_attr instance.
However, multi table actions (e.g. CT) instantiate multiple attr instances.
Currently action_match_supported() reads the actions flag from the
flow's attribute instance. Modify the function to receive the action
flags as a parameter which is set by the calling function and
pass the aggregated actions to actions_match_supported().
Signed-off-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
.../net/ethernet/mellanox/mlx5/core/en/tc/act/act.h | 1 +
drivers/net/ethernet/mellanox/mlx5/core/en_tc.c | 10 +++++++---
2 files changed, 8 insertions(+), 3 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/act.h b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/act.h
index bfbc91c116a5..fc7c06688b51 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/act.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/act.h
@@ -16,6 +16,7 @@ struct mlx5e_tc_act_parse_state {
unsigned int num_actions;
struct mlx5e_tc_flow *flow;
struct netlink_ext_ack *extack;
+ u32 actions;
bool ct;
bool encap;
bool decap;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index 342ab0688f13..a709b2e9f3f2 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -3257,11 +3257,11 @@ actions_match_supported_fdb(struct mlx5e_priv *priv,
static bool
actions_match_supported(struct mlx5e_priv *priv,
struct flow_action *flow_action,
+ u32 actions,
struct mlx5e_tc_flow_parse_attr *parse_attr,
struct mlx5e_tc_flow *flow,
struct netlink_ext_ack *extack)
{
- u32 actions = flow->attr->action;
bool ct_flow, ct_clear;
ct_clear = flow->attr->ct_attr.ct_action & TCA_CT_ACT_CLEAR;
@@ -3344,6 +3344,8 @@ parse_tc_actions(struct mlx5e_tc_act_parse_state *parse_state,
err = tc_act->parse_action(parse_state, act, priv, attr);
if (err)
return err;
+
+ parse_state->actions |= attr->action;
}
flow_action_for_each(i, act, flow_action) {
@@ -3445,7 +3447,8 @@ parse_tc_nic_actions(struct mlx5e_priv *priv,
if (err)
return err;
- if (!actions_match_supported(priv, flow_action, parse_attr, flow, extack))
+ if (!actions_match_supported(priv, flow_action, parse_state->actions,
+ parse_attr, flow, extack))
return -EOPNOTSUPP;
return 0;
@@ -3574,7 +3577,8 @@ parse_tc_fdb_actions(struct mlx5e_priv *priv,
if (err)
return err;
- if (!actions_match_supported(priv, flow_action, parse_attr, flow, extack))
+ if (!actions_match_supported(priv, flow_action, parse_state->actions,
+ parse_attr, flow, extack))
return -EOPNOTSUPP;
return 0;
--
2.34.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* [net-next 10/15] net/mlx5e: Add post act offload/unoffload API
2022-02-17 7:56 [pull request][net-next 00/15] mlx5 updates 2022-02-16 Saeed Mahameed
` (8 preceding siblings ...)
2022-02-17 7:56 ` [net-next 09/15] net/mlx5e: Pass actions param to actions_match_supported() Saeed Mahameed
@ 2022-02-17 7:56 ` Saeed Mahameed
2022-02-17 7:56 ` [net-next 11/15] net/mlx5e: Create new flow attr for multi table actions Saeed Mahameed
` (4 subsequent siblings)
14 siblings, 0 replies; 17+ messages in thread
From: Saeed Mahameed @ 2022-02-17 7:56 UTC (permalink / raw)
To: David S. Miller, Jakub Kicinski; +Cc: netdev, Roi Dayan, Saeed Mahameed
From: Roi Dayan <roid@nvidia.com>
Introduce mlx5e_tc_post_act_offload() and mlx5e_tc_post_act_unoffload()
to be able to unoffload and reoffload existing post action rules handles.
For example in neigh update events, the driver removes and readds rules in
hardware.
Signed-off-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
.../mellanox/mlx5/core/en/tc/post_act.c | 67 +++++++++++++------
.../mellanox/mlx5/core/en/tc/post_act.h | 8 +++
2 files changed, 54 insertions(+), 21 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/post_act.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/post_act.c
index 9e0e229cf164..ce7ba1951e25 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/post_act.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/post_act.c
@@ -75,21 +75,47 @@ mlx5e_tc_post_act_destroy(struct mlx5e_post_act *post_act)
kfree(post_act);
}
+int
+mlx5e_tc_post_act_offload(struct mlx5e_post_act *post_act,
+ struct mlx5e_post_act_handle *handle)
+{
+ struct mlx5_flow_spec *spec;
+ int err;
+
+ spec = kvzalloc(sizeof(*spec), GFP_KERNEL);
+ if (!spec)
+ return -ENOMEM;
+
+ /* Post action rule matches on fte_id and executes original rule's tc rule action */
+ mlx5e_tc_match_to_reg_match(spec, FTEID_TO_REG, handle->id, MLX5_POST_ACTION_MASK);
+
+ handle->rule = mlx5_tc_rule_insert(post_act->priv, spec, handle->attr);
+ if (IS_ERR(handle->rule)) {
+ err = PTR_ERR(handle->rule);
+ netdev_warn(post_act->priv->netdev, "Failed to add post action rule");
+ goto err_rule;
+ }
+
+ kvfree(spec);
+ return 0;
+
+err_rule:
+ kvfree(spec);
+ return err;
+}
+
struct mlx5e_post_act_handle *
mlx5e_tc_post_act_add(struct mlx5e_post_act *post_act, struct mlx5_flow_attr *attr)
{
u32 attr_sz = ns_to_attr_sz(post_act->ns_type);
- struct mlx5e_post_act_handle *handle = NULL;
- struct mlx5_flow_attr *post_attr = NULL;
- struct mlx5_flow_spec *spec = NULL;
+ struct mlx5e_post_act_handle *handle;
+ struct mlx5_flow_attr *post_attr;
int err;
handle = kzalloc(sizeof(*handle), GFP_KERNEL);
- spec = kvzalloc(sizeof(*spec), GFP_KERNEL);
post_attr = mlx5_alloc_flow_attr(post_act->ns_type);
- if (!handle || !spec || !post_attr) {
+ if (!handle || !post_attr) {
kfree(post_attr);
- kvfree(spec);
kfree(handle);
return ERR_PTR(-ENOMEM);
}
@@ -113,36 +139,35 @@ mlx5e_tc_post_act_add(struct mlx5e_post_act *post_act, struct mlx5_flow_attr *at
if (err)
goto err_xarray;
- /* Post action rule matches on fte_id and executes original rule's
- * tc rule action
- */
- mlx5e_tc_match_to_reg_match(spec, FTEID_TO_REG,
- handle->id, MLX5_POST_ACTION_MASK);
-
- handle->rule = mlx5_tc_rule_insert(post_act->priv, spec, post_attr);
- if (IS_ERR(handle->rule)) {
- err = PTR_ERR(handle->rule);
- netdev_warn(post_act->priv->netdev, "Failed to add post action rule");
- goto err_rule;
- }
handle->attr = post_attr;
+ err = mlx5e_tc_post_act_offload(post_act, handle);
+ if (err)
+ goto err_rule;
+
- kvfree(spec);
return handle;
err_rule:
xa_erase(&post_act->ids, handle->id);
err_xarray:
kfree(post_attr);
- kvfree(spec);
kfree(handle);
return ERR_PTR(err);
}
void
-mlx5e_tc_post_act_del(struct mlx5e_post_act *post_act, struct mlx5e_post_act_handle *handle)
+mlx5e_tc_post_act_unoffload(struct mlx5e_post_act *post_act,
+ struct mlx5e_post_act_handle *handle)
{
mlx5_tc_rule_delete(post_act->priv, handle->rule, handle->attr);
+ handle->rule = NULL;
+}
+
+void
+mlx5e_tc_post_act_del(struct mlx5e_post_act *post_act, struct mlx5e_post_act_handle *handle)
+{
+ if (!IS_ERR_OR_NULL(handle->rule))
+ mlx5e_tc_post_act_unoffload(post_act, handle);
xa_erase(&post_act->ids, handle->id);
kfree(handle->attr);
kfree(handle);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/post_act.h b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/post_act.h
index b530ec1981a5..f476774c0b75 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/post_act.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/post_act.h
@@ -24,6 +24,14 @@ mlx5e_tc_post_act_add(struct mlx5e_post_act *post_act, struct mlx5_flow_attr *at
void
mlx5e_tc_post_act_del(struct mlx5e_post_act *post_act, struct mlx5e_post_act_handle *handle);
+int
+mlx5e_tc_post_act_offload(struct mlx5e_post_act *post_act,
+ struct mlx5e_post_act_handle *handle);
+
+void
+mlx5e_tc_post_act_unoffload(struct mlx5e_post_act *post_act,
+ struct mlx5e_post_act_handle *handle);
+
struct mlx5_flow_table *
mlx5e_tc_post_act_get_ft(struct mlx5e_post_act *post_act);
--
2.34.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* [net-next 11/15] net/mlx5e: Create new flow attr for multi table actions
2022-02-17 7:56 [pull request][net-next 00/15] mlx5 updates 2022-02-16 Saeed Mahameed
` (9 preceding siblings ...)
2022-02-17 7:56 ` [net-next 10/15] net/mlx5e: Add post act offload/unoffload API Saeed Mahameed
@ 2022-02-17 7:56 ` Saeed Mahameed
2022-02-17 7:56 ` [net-next 12/15] net/mlx5e: Use multi table support for CT and sample actions Saeed Mahameed
` (3 subsequent siblings)
14 siblings, 0 replies; 17+ messages in thread
From: Saeed Mahameed @ 2022-02-17 7:56 UTC (permalink / raw)
To: David S. Miller, Jakub Kicinski; +Cc: netdev, Roi Dayan, Saeed Mahameed
From: Roi Dayan <roid@nvidia.com>
Some TC actions use post actions for their implementation.
For example CT and sample actions.
Create a new flow attr after each multi table action and
create a post action rule for it.
First flow attr being offloaded normally and linked to the next
attr (post action rule) with setting an id on reg_c.
Post action rules match the id on reg_c and continue to the next one.
The flow counter is allocated on the last rule.
Signed-off-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
.../mellanox/mlx5/core/en/tc/act/act.c | 73 ++++
.../mellanox/mlx5/core/en/tc/act/act.h | 24 ++
.../mellanox/mlx5/core/en/tc/post_act.c | 6 -
.../mellanox/mlx5/core/en/tc/sample.c | 3 +
.../ethernet/mellanox/mlx5/core/en/tc_ct.c | 3 +
.../ethernet/mellanox/mlx5/core/en/tc_priv.h | 7 +
.../mellanox/mlx5/core/en/tc_tun_encap.c | 48 ++-
.../net/ethernet/mellanox/mlx5/core/en_tc.c | 383 +++++++++++++++---
.../net/ethernet/mellanox/mlx5/core/en_tc.h | 3 +-
9 files changed, 483 insertions(+), 67 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/act.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/act.c
index e600924e30ea..cb8f7593a00c 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/act.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/act.c
@@ -2,6 +2,7 @@
// Copyright (c) 2021, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
#include "act.h"
+#include "en/tc/post_act.h"
#include "en/tc_priv.h"
#include "mlx5_core.h"
@@ -101,3 +102,75 @@ mlx5e_tc_act_init_parse_state(struct mlx5e_tc_act_parse_state *parse_state,
parse_state->num_actions = flow_action->num_entries;
parse_state->extack = extack;
}
+
+void
+mlx5e_tc_act_reorder_flow_actions(struct flow_action *flow_action,
+ struct mlx5e_tc_flow_action *flow_action_reorder)
+{
+ struct flow_action_entry *act;
+ int i, j = 0;
+
+ flow_action_for_each(i, act, flow_action) {
+ /* Add CT action to be first. */
+ if (act->id == FLOW_ACTION_CT)
+ flow_action_reorder->entries[j++] = act;
+ }
+
+ flow_action_for_each(i, act, flow_action) {
+ if (act->id == FLOW_ACTION_CT)
+ continue;
+ flow_action_reorder->entries[j++] = act;
+ }
+}
+
+int
+mlx5e_tc_act_post_parse(struct mlx5e_tc_act_parse_state *parse_state,
+ struct flow_action *flow_action,
+ struct mlx5_flow_attr *attr,
+ enum mlx5_flow_namespace_type ns_type)
+{
+ struct flow_action_entry *act;
+ struct mlx5e_tc_act *tc_act;
+ struct mlx5e_priv *priv;
+ int err = 0, i;
+
+ priv = parse_state->flow->priv;
+
+ flow_action_for_each(i, act, flow_action) {
+ tc_act = mlx5e_tc_act_get(act->id, ns_type);
+ if (!tc_act || !tc_act->post_parse ||
+ !tc_act->can_offload(parse_state, act, i, attr))
+ continue;
+
+ err = tc_act->post_parse(parse_state, priv, attr);
+ if (err)
+ goto out;
+ }
+
+out:
+ return err;
+}
+
+int
+mlx5e_tc_act_set_next_post_act(struct mlx5e_tc_flow *flow,
+ struct mlx5_flow_attr *attr,
+ struct mlx5_flow_attr *next_attr)
+{
+ struct mlx5_core_dev *mdev = flow->priv->mdev;
+ struct mlx5e_tc_mod_hdr_acts *mod_acts;
+ int err;
+
+ mod_acts = &attr->parse_attr->mod_hdr_acts;
+
+ /* Set handle on current post act rule to next post act rule. */
+ err = mlx5e_tc_post_act_set_handle(mdev, next_attr->post_act_handle, mod_acts);
+ if (err) {
+ mlx5_core_warn(mdev, "Failed setting post action handle");
+ return err;
+ }
+
+ attr->action |= MLX5_FLOW_CONTEXT_ACTION_FWD_DEST |
+ MLX5_FLOW_CONTEXT_ACTION_MOD_HDR;
+
+ return 0;
+}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/act.h b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/act.h
index fc7c06688b51..519fa1056d9f 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/act.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/act.h
@@ -42,6 +42,15 @@ struct mlx5e_tc_act {
int (*post_parse)(struct mlx5e_tc_act_parse_state *parse_state,
struct mlx5e_priv *priv,
struct mlx5_flow_attr *attr);
+
+ bool (*is_multi_table_act)(struct mlx5e_priv *priv,
+ const struct flow_action_entry *act,
+ struct mlx5_flow_attr *attr);
+};
+
+struct mlx5e_tc_flow_action {
+ unsigned int num_entries;
+ struct flow_action_entry **entries;
};
extern struct mlx5e_tc_act mlx5e_tc_act_drop;
@@ -74,4 +83,19 @@ mlx5e_tc_act_init_parse_state(struct mlx5e_tc_act_parse_state *parse_state,
struct flow_action *flow_action,
struct netlink_ext_ack *extack);
+void
+mlx5e_tc_act_reorder_flow_actions(struct flow_action *flow_action,
+ struct mlx5e_tc_flow_action *flow_action_reorder);
+
+int
+mlx5e_tc_act_post_parse(struct mlx5e_tc_act_parse_state *parse_state,
+ struct flow_action *flow_action,
+ struct mlx5_flow_attr *attr,
+ enum mlx5_flow_namespace_type ns_type);
+
+int
+mlx5e_tc_act_set_next_post_act(struct mlx5e_tc_flow *flow,
+ struct mlx5_flow_attr *attr,
+ struct mlx5_flow_attr *next_attr);
+
#endif /* __MLX5_EN_TC_ACT_H__ */
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/post_act.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/post_act.c
index ce7ba1951e25..27854ac844a0 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/post_act.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/post_act.c
@@ -140,15 +140,9 @@ mlx5e_tc_post_act_add(struct mlx5e_post_act *post_act, struct mlx5_flow_attr *at
goto err_xarray;
handle->attr = post_attr;
- err = mlx5e_tc_post_act_offload(post_act, handle);
- if (err)
- goto err_rule;
-
return handle;
-err_rule:
- xa_erase(&post_act->ids, handle->id);
err_xarray:
kfree(post_attr);
kfree(handle);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/sample.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/sample.c
index 32230e677029..90b86c4aaf51 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/sample.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/sample.c
@@ -533,6 +533,9 @@ mlx5e_tc_sample_offload(struct mlx5e_tc_psample *tc_psample,
err = PTR_ERR(post_act_handle);
goto err_post_act;
}
+ err = mlx5e_tc_post_act_offload(tc_psample->post_act, post_act_handle);
+ if (err)
+ goto err_post_rule;
sample_flow->post_act_handle = post_act_handle;
} else {
err = add_post_rule(esw, sample_flow, spec, attr, &default_tbl_id);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c
index 0f4d3b9dd979..228ebb19fcab 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c
@@ -1823,6 +1823,9 @@ __mlx5_tc_ct_flow_offload(struct mlx5_tc_ct_priv *ct_priv,
ct_dbg("Failed to allocate post action handle");
goto err_post_act_handle;
}
+ err = mlx5e_tc_post_act_offload(ct_priv->post_act, handle);
+ if (err)
+ goto err_alloc_pre;
ct_flow->post_act_handle = handle;
/* Base flow attributes of both rules on original rule attribute */
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_priv.h b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_priv.h
index 9ffba584b982..f76624699a8d 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_priv.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_priv.h
@@ -109,6 +109,7 @@ struct mlx5e_tc_flow {
struct completion init_done;
struct completion del_hw_done;
struct mlx5_flow_attr *attr;
+ struct list_head attrs;
};
struct mlx5_flow_handle *
@@ -129,6 +130,12 @@ mlx5e_tc_offload_fdb_rules(struct mlx5_eswitch *esw,
struct mlx5_flow_spec *spec,
struct mlx5_flow_attr *attr);
+struct mlx5_flow_attr *
+mlx5e_tc_get_encap_attr(struct mlx5e_tc_flow *flow);
+
+void mlx5e_tc_unoffload_flow_post_acts(struct mlx5e_tc_flow *flow);
+int mlx5e_tc_offload_flow_post_acts(struct mlx5e_tc_flow *flow);
+
bool mlx5e_is_eswitch_flow(struct mlx5e_tc_flow *flow);
bool mlx5e_is_ft_flow(struct mlx5e_tc_flow *flow);
bool mlx5e_is_offloaded_flow(struct mlx5e_tc_flow *flow);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun_encap.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun_encap.c
index 1f8d339ff0c3..171bc6b36aa4 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun_encap.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun_encap.c
@@ -173,19 +173,29 @@ void mlx5e_tc_encap_flows_add(struct mlx5e_priv *priv,
list_for_each_entry(flow, flow_list, tmp_list) {
if (!mlx5e_is_offloaded_flow(flow) || !flow_flag_test(flow, SLOW))
continue;
- attr = flow->attr;
- esw_attr = attr->esw_attr;
- spec = &attr->parse_attr->spec;
+ spec = &flow->attr->parse_attr->spec;
+
+ attr = mlx5e_tc_get_encap_attr(flow);
+ esw_attr = attr->esw_attr;
esw_attr->dests[flow->tmp_entry_index].pkt_reformat = e->pkt_reformat;
esw_attr->dests[flow->tmp_entry_index].flags |= MLX5_ESW_DEST_ENCAP_VALID;
/* Do not offload flows with unresolved neighbors */
if (!mlx5e_tc_flow_all_encaps_valid(esw_attr))
continue;
+
+ err = mlx5e_tc_offload_flow_post_acts(flow);
+ if (err) {
+ mlx5_core_warn(priv->mdev, "Failed to update flow post acts, %d\n",
+ err);
+ continue;
+ }
+
/* update from slow path rule to encap rule */
- rule = mlx5e_tc_offload_fdb_rules(esw, flow, spec, attr);
+ rule = mlx5e_tc_offload_fdb_rules(esw, flow, spec, flow->attr);
if (IS_ERR(rule)) {
+ mlx5e_tc_unoffload_flow_post_acts(flow);
err = PTR_ERR(rule);
mlx5_core_warn(priv->mdev, "Failed to update cached encapsulation flow, %d\n",
err);
@@ -214,12 +224,13 @@ void mlx5e_tc_encap_flows_del(struct mlx5e_priv *priv,
list_for_each_entry(flow, flow_list, tmp_list) {
if (!mlx5e_is_offloaded_flow(flow) || flow_flag_test(flow, SLOW))
continue;
- attr = flow->attr;
- esw_attr = attr->esw_attr;
- spec = &attr->parse_attr->spec;
+ spec = &flow->attr->parse_attr->spec;
/* update from encap rule to slow path rule */
rule = mlx5e_tc_offload_to_slow_path(esw, flow, spec);
+
+ attr = mlx5e_tc_get_encap_attr(flow);
+ esw_attr = attr->esw_attr;
/* mark the flow's encap dest as non-valid */
esw_attr->dests[flow->tmp_entry_index].flags &= ~MLX5_ESW_DEST_ENCAP_VALID;
@@ -230,7 +241,8 @@ void mlx5e_tc_encap_flows_del(struct mlx5e_priv *priv,
continue;
}
- mlx5e_tc_unoffload_fdb_rules(esw, flow, attr);
+ mlx5e_tc_unoffload_fdb_rules(esw, flow, flow->attr);
+ mlx5e_tc_unoffload_flow_post_acts(flow);
flow->rule[0] = rule;
/* was unset when fast path rule removed */
flow_flag_set(flow, OFFLOADED);
@@ -495,6 +507,9 @@ void mlx5e_detach_encap(struct mlx5e_priv *priv,
struct mlx5e_encap_entry *e = flow->encaps[out_index].e;
struct mlx5_eswitch *esw = priv->mdev->priv.eswitch;
+ if (!mlx5e_is_eswitch_flow(flow))
+ return;
+
if (attr->esw_attr->dests[out_index].flags &
MLX5_ESW_DEST_CHAIN_WITH_SRC_PORT_CHANGE)
mlx5e_detach_encap_route(priv, flow, out_index);
@@ -1360,17 +1375,19 @@ static void mlx5e_reoffload_encap(struct mlx5e_priv *priv,
list_for_each_entry(flow, encap_flows, tmp_list) {
struct mlx5e_tc_flow_parse_attr *parse_attr;
- struct mlx5_flow_attr *attr = flow->attr;
struct mlx5_esw_flow_attr *esw_attr;
struct mlx5_flow_handle *rule;
+ struct mlx5_flow_attr *attr;
struct mlx5_flow_spec *spec;
if (flow_flag_test(flow, FAILED))
continue;
+ spec = &flow->attr->parse_attr->spec;
+
+ attr = mlx5e_tc_get_encap_attr(flow);
esw_attr = attr->esw_attr;
parse_attr = attr->parse_attr;
- spec = &parse_attr->spec;
err = mlx5e_update_vf_tunnel(esw, esw_attr, &parse_attr->mod_hdr_acts,
e->out_dev, e->route_dev_ifindex,
@@ -1392,9 +1409,18 @@ static void mlx5e_reoffload_encap(struct mlx5e_priv *priv,
esw_attr->dests[flow->tmp_entry_index].flags |= MLX5_ESW_DEST_ENCAP_VALID;
if (!mlx5e_tc_flow_all_encaps_valid(esw_attr))
goto offload_to_slow_path;
+
+ err = mlx5e_tc_offload_flow_post_acts(flow);
+ if (err) {
+ mlx5_core_warn(priv->mdev, "Failed to update flow post acts, %d\n",
+ err);
+ goto offload_to_slow_path;
+ }
+
/* update from slow path rule to encap rule */
- rule = mlx5e_tc_offload_fdb_rules(esw, flow, spec, attr);
+ rule = mlx5e_tc_offload_fdb_rules(esw, flow, spec, flow->attr);
if (IS_ERR(rule)) {
+ mlx5e_tc_unoffload_flow_post_acts(flow);
err = PTR_ERR(rule);
mlx5_core_warn(priv->mdev, "Failed to update cached encapsulation flow, %d\n",
err);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index a709b2e9f3f2..b9d6a2e8b240 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -115,6 +115,7 @@ struct mlx5e_tc_attr_to_reg_mapping mlx5e_tc_attr_to_reg_mappings[] = {
static struct lock_class_key tc_ht_lock_key;
static void mlx5e_put_flow_tunnel_id(struct mlx5e_tc_flow *flow);
+static void free_flow_post_acts(struct mlx5e_tc_flow *flow);
void
mlx5e_tc_match_to_reg_match(struct mlx5_flow_spec *spec,
@@ -273,6 +274,23 @@ get_sample_priv(struct mlx5e_priv *priv)
return NULL;
}
+static struct mlx5e_post_act *
+get_post_action(struct mlx5e_priv *priv)
+{
+ struct mlx5_eswitch *esw = priv->mdev->priv.eswitch;
+ struct mlx5_rep_uplink_priv *uplink_priv;
+ struct mlx5e_rep_priv *uplink_rpriv;
+
+ if (is_mdev_switchdev_mode(priv->mdev)) {
+ uplink_rpriv = mlx5_eswitch_get_uplink_priv(esw, REP_ETH);
+ uplink_priv = &uplink_rpriv->uplink_priv;
+
+ return uplink_priv->post_act;
+ }
+
+ return priv->fs.tc.post_act;
+}
+
struct mlx5_flow_handle *
mlx5_tc_rule_insert(struct mlx5e_priv *priv,
struct mlx5_flow_spec *spec,
@@ -1193,6 +1211,8 @@ static void mlx5e_tc_del_nic_flow(struct mlx5e_priv *priv,
if (flow_flag_test(flow, HAIRPIN))
mlx5e_hairpin_flow_del(priv, flow);
+ free_flow_post_acts(flow);
+
kvfree(attr->parse_attr);
kfree(flow->attr);
}
@@ -1425,6 +1445,9 @@ set_encap_dests(struct mlx5e_priv *priv,
int out_index;
int err = 0;
+ if (!mlx5e_is_eswitch_flow(flow))
+ return 0;
+
parse_attr = attr->parse_attr;
esw_attr = attr->esw_attr;
*vf_tun = false;
@@ -1480,6 +1503,9 @@ clean_encap_dests(struct mlx5e_priv *priv,
struct mlx5_esw_flow_attr *esw_attr;
int out_index;
+ if (!mlx5e_is_eswitch_flow(flow))
+ return;
+
esw_attr = attr->esw_attr;
*vf_tun = false;
@@ -1627,7 +1653,7 @@ mlx5e_tc_add_fdb_flow(struct mlx5e_priv *priv,
* (1) there's no error
* (2) there's an encap action and we don't have valid neigh
*/
- if (!encap_valid)
+ if (!encap_valid || flow_flag_test(flow, SLOW))
flow->rule[0] = mlx5e_tc_offload_to_slow_path(esw, flow, &parse_attr->spec);
else
flow->rule[0] = mlx5e_tc_offload_fdb_rules(esw, flow, &parse_attr->spec, attr);
@@ -1712,6 +1738,8 @@ static void mlx5e_tc_del_fdb_flow(struct mlx5e_priv *priv,
if (flow_flag_test(flow, L3_TO_L2_DECAP))
mlx5e_detach_decap(priv, flow);
+ free_flow_post_acts(flow);
+
kvfree(attr->esw_attr->rx_tun_attr);
kvfree(attr->parse_attr);
kfree(flow->attr);
@@ -1719,7 +1747,10 @@ static void mlx5e_tc_del_fdb_flow(struct mlx5e_priv *priv,
struct mlx5_fc *mlx5e_tc_get_counter(struct mlx5e_tc_flow *flow)
{
- return flow->attr->counter;
+ struct mlx5_flow_attr *attr;
+
+ attr = list_first_entry(&flow->attrs, struct mlx5_flow_attr, list);
+ return attr->counter;
}
/* Iterate over tmp_list of flows attached to flow_list head. */
@@ -3316,52 +3347,6 @@ bool mlx5e_same_hw_devs(struct mlx5e_priv *priv, struct mlx5e_priv *peer_priv)
return (fsystem_guid == psystem_guid);
}
-static int
-parse_tc_actions(struct mlx5e_tc_act_parse_state *parse_state,
- struct flow_action *flow_action)
-{
- struct netlink_ext_ack *extack = parse_state->extack;
- struct mlx5e_tc_flow *flow = parse_state->flow;
- struct mlx5_flow_attr *attr = flow->attr;
- enum mlx5_flow_namespace_type ns_type;
- struct mlx5e_priv *priv = flow->priv;
- const struct flow_action_entry *act;
- struct mlx5e_tc_act *tc_act;
- int err, i;
-
- ns_type = mlx5e_get_flow_namespace(flow);
-
- flow_action_for_each(i, act, flow_action) {
- tc_act = mlx5e_tc_act_get(act->id, ns_type);
- if (!tc_act) {
- NL_SET_ERR_MSG_MOD(extack, "Not implemented offload action");
- return -EOPNOTSUPP;
- }
-
- if (!tc_act->can_offload(parse_state, act, i, attr))
- return -EOPNOTSUPP;
-
- err = tc_act->parse_action(parse_state, act, priv, attr);
- if (err)
- return err;
-
- parse_state->actions |= attr->action;
- }
-
- flow_action_for_each(i, act, flow_action) {
- tc_act = mlx5e_tc_act_get(act->id, ns_type);
- if (!tc_act || !tc_act->post_parse ||
- !tc_act->can_offload(parse_state, act, i, attr))
- continue;
-
- err = tc_act->post_parse(parse_state, priv, attr);
- if (err)
- return err;
- }
-
- return 0;
-}
-
static int
actions_prepare_mod_hdr_actions(struct mlx5e_priv *priv,
struct mlx5e_tc_flow *flow,
@@ -3400,6 +3385,300 @@ actions_prepare_mod_hdr_actions(struct mlx5e_priv *priv,
return 0;
}
+static struct mlx5_flow_attr*
+mlx5e_clone_flow_attr_for_post_act(struct mlx5_flow_attr *attr,
+ enum mlx5_flow_namespace_type ns_type)
+{
+ struct mlx5e_tc_flow_parse_attr *parse_attr;
+ u32 attr_sz = ns_to_attr_sz(ns_type);
+ struct mlx5_flow_attr *attr2;
+
+ attr2 = mlx5_alloc_flow_attr(ns_type);
+ parse_attr = kvzalloc(sizeof(*parse_attr), GFP_KERNEL);
+ if (!attr2 || !parse_attr) {
+ kvfree(parse_attr);
+ kfree(attr2);
+ return attr2;
+ }
+
+ memcpy(attr2, attr, attr_sz);
+ INIT_LIST_HEAD(&attr2->list);
+ parse_attr->filter_dev = attr->parse_attr->filter_dev;
+ attr2->action = 0;
+ attr2->flags = 0;
+ attr2->parse_attr = parse_attr;
+ return attr2;
+}
+
+static struct mlx5_core_dev *
+get_flow_counter_dev(struct mlx5e_tc_flow *flow)
+{
+ return mlx5e_is_eswitch_flow(flow) ? flow->attr->esw_attr->counter_dev : flow->priv->mdev;
+}
+
+struct mlx5_flow_attr *
+mlx5e_tc_get_encap_attr(struct mlx5e_tc_flow *flow)
+{
+ struct mlx5_esw_flow_attr *esw_attr;
+ struct mlx5_flow_attr *attr;
+ int i;
+
+ list_for_each_entry(attr, &flow->attrs, list) {
+ esw_attr = attr->esw_attr;
+ for (i = 0; i < MLX5_MAX_FLOW_FWD_VPORTS; i++) {
+ if (esw_attr->dests[i].flags & MLX5_ESW_DEST_ENCAP)
+ return attr;
+ }
+ }
+
+ return NULL;
+}
+
+void
+mlx5e_tc_unoffload_flow_post_acts(struct mlx5e_tc_flow *flow)
+{
+ struct mlx5e_post_act *post_act = get_post_action(flow->priv);
+ struct mlx5_flow_attr *attr;
+
+ list_for_each_entry(attr, &flow->attrs, list) {
+ if (list_is_last(&attr->list, &flow->attrs))
+ break;
+
+ mlx5e_tc_post_act_unoffload(post_act, attr->post_act_handle);
+ }
+}
+
+static void
+free_flow_post_acts(struct mlx5e_tc_flow *flow)
+{
+ struct mlx5_core_dev *counter_dev = get_flow_counter_dev(flow);
+ struct mlx5e_post_act *post_act = get_post_action(flow->priv);
+ struct mlx5_flow_attr *attr, *tmp;
+ bool vf_tun;
+
+ list_for_each_entry_safe(attr, tmp, &flow->attrs, list) {
+ if (list_is_last(&attr->list, &flow->attrs))
+ break;
+
+ if (attr->post_act_handle)
+ mlx5e_tc_post_act_del(post_act, attr->post_act_handle);
+
+ clean_encap_dests(flow->priv, flow, attr, &vf_tun);
+
+ if (attr->action & MLX5_FLOW_CONTEXT_ACTION_COUNT)
+ mlx5_fc_destroy(counter_dev, attr->counter);
+
+ if (attr->action & MLX5_FLOW_CONTEXT_ACTION_MOD_HDR) {
+ mlx5e_mod_hdr_dealloc(&attr->parse_attr->mod_hdr_acts);
+ if (attr->modify_hdr)
+ mlx5_modify_header_dealloc(flow->priv->mdev, attr->modify_hdr);
+ }
+
+ list_del(&attr->list);
+ kvfree(attr->parse_attr);
+ kfree(attr);
+ }
+}
+
+int
+mlx5e_tc_offload_flow_post_acts(struct mlx5e_tc_flow *flow)
+{
+ struct mlx5e_post_act *post_act = get_post_action(flow->priv);
+ struct mlx5_flow_attr *attr;
+ int err = 0;
+
+ list_for_each_entry(attr, &flow->attrs, list) {
+ if (list_is_last(&attr->list, &flow->attrs))
+ break;
+
+ err = mlx5e_tc_post_act_offload(post_act, attr->post_act_handle);
+ if (err)
+ break;
+ }
+
+ return err;
+}
+
+/* TC filter rule HW translation:
+ *
+ * +---------------------+
+ * + ft prio (tc chain) +
+ * + original match +
+ * +---------------------+
+ * |
+ * | if multi table action
+ * |
+ * v
+ * +---------------------+
+ * + post act ft |<----.
+ * + match fte id | | split on multi table action
+ * + do actions |-----'
+ * +---------------------+
+ * |
+ * |
+ * v
+ * Do rest of the actions after last multi table action.
+ */
+static int
+alloc_flow_post_acts(struct mlx5e_tc_flow *flow, struct netlink_ext_ack *extack)
+{
+ struct mlx5e_post_act *post_act = get_post_action(flow->priv);
+ struct mlx5_flow_attr *attr, *next_attr = NULL;
+ struct mlx5e_post_act_handle *handle;
+ bool vf_tun, encap_valid = true;
+ int err;
+
+ /* This is going in reverse order as needed.
+ * The first entry is the last attribute.
+ */
+ list_for_each_entry(attr, &flow->attrs, list) {
+ if (!next_attr) {
+ /* Set counter action on last post act rule. */
+ attr->action |= MLX5_FLOW_CONTEXT_ACTION_COUNT;
+ } else {
+ attr->action &= ~MLX5_FLOW_CONTEXT_ACTION_COUNT;
+ err = mlx5e_tc_act_set_next_post_act(flow, attr, next_attr);
+ if (err)
+ goto out_free;
+ }
+
+ /* Don't add post_act rule for first attr (last in the list).
+ * It's being handled by the caller.
+ */
+ if (list_is_last(&attr->list, &flow->attrs))
+ break;
+
+ err = set_encap_dests(flow->priv, flow, attr, extack, &encap_valid, &vf_tun);
+ if (err)
+ goto out_free;
+
+ if (!encap_valid)
+ flow_flag_set(flow, SLOW);
+
+ err = actions_prepare_mod_hdr_actions(flow->priv, flow, attr, extack);
+ if (err)
+ goto out_free;
+
+ if (attr->action & MLX5_FLOW_CONTEXT_ACTION_MOD_HDR) {
+ err = mlx5e_tc_add_flow_mod_hdr(flow->priv, flow, attr);
+ if (err)
+ goto out_free;
+ }
+
+ if (attr->action & MLX5_FLOW_CONTEXT_ACTION_COUNT) {
+ err = alloc_flow_attr_counter(get_flow_counter_dev(flow), attr);
+ if (err)
+ goto out_free;
+ }
+
+ handle = mlx5e_tc_post_act_add(post_act, attr);
+ if (IS_ERR(handle)) {
+ err = PTR_ERR(handle);
+ goto out_free;
+ }
+
+ attr->post_act_handle = handle;
+ next_attr = attr;
+ }
+
+ if (flow_flag_test(flow, SLOW))
+ goto out;
+
+ err = mlx5e_tc_offload_flow_post_acts(flow);
+ if (err)
+ goto out_free;
+
+out:
+ return 0;
+
+out_free:
+ free_flow_post_acts(flow);
+ return err;
+}
+
+static int
+parse_tc_actions(struct mlx5e_tc_act_parse_state *parse_state,
+ struct flow_action *flow_action)
+{
+ struct netlink_ext_ack *extack = parse_state->extack;
+ struct mlx5e_tc_flow_action flow_action_reorder;
+ struct mlx5e_tc_flow *flow = parse_state->flow;
+ struct mlx5_flow_attr *attr = flow->attr;
+ enum mlx5_flow_namespace_type ns_type;
+ struct mlx5e_priv *priv = flow->priv;
+ struct flow_action_entry *act, **_act;
+ struct mlx5e_tc_act *tc_act;
+ int err, i;
+
+ flow_action_reorder.num_entries = flow_action->num_entries;
+ flow_action_reorder.entries = kcalloc(flow_action->num_entries,
+ sizeof(flow_action), GFP_KERNEL);
+ if (!flow_action_reorder.entries)
+ return -ENOMEM;
+
+ mlx5e_tc_act_reorder_flow_actions(flow_action, &flow_action_reorder);
+
+ ns_type = mlx5e_get_flow_namespace(flow);
+ list_add(&attr->list, &flow->attrs);
+
+ flow_action_for_each(i, _act, &flow_action_reorder) {
+ act = *_act;
+ tc_act = mlx5e_tc_act_get(act->id, ns_type);
+ if (!tc_act) {
+ NL_SET_ERR_MSG_MOD(extack, "Not implemented offload action");
+ err = -EOPNOTSUPP;
+ goto out_free;
+ }
+
+ if (!tc_act->can_offload(parse_state, act, i, attr)) {
+ err = -EOPNOTSUPP;
+ goto out_free;
+ }
+
+ err = tc_act->parse_action(parse_state, act, priv, attr);
+ if (err)
+ goto out_free;
+
+ parse_state->actions |= attr->action;
+
+ /* Split attr for multi table act if not the last act. */
+ if (tc_act->is_multi_table_act &&
+ tc_act->is_multi_table_act(priv, act, attr) &&
+ i < flow_action_reorder.num_entries - 1) {
+ err = mlx5e_tc_act_post_parse(parse_state, flow_action, attr, ns_type);
+ if (err)
+ goto out_free;
+
+ attr = mlx5e_clone_flow_attr_for_post_act(flow->attr, ns_type);
+ if (!attr) {
+ err = -ENOMEM;
+ goto out_free;
+ }
+
+ list_add(&attr->list, &flow->attrs);
+ }
+ }
+
+ kfree(flow_action_reorder.entries);
+
+ err = mlx5e_tc_act_post_parse(parse_state, flow_action, attr, ns_type);
+ if (err)
+ goto out_free_post_acts;
+
+ err = alloc_flow_post_acts(flow, extack);
+ if (err)
+ goto out_free_post_acts;
+
+ return 0;
+
+out_free:
+ kfree(flow_action_reorder.entries);
+out_free_post_acts:
+ free_flow_post_acts(flow);
+
+ return err;
+}
+
static int
flow_action_supported(struct flow_action *flow_action,
struct netlink_ext_ack *extack)
@@ -3652,7 +3931,12 @@ mlx5_alloc_flow_attr(enum mlx5_flow_namespace_type type)
sizeof(struct mlx5_nic_flow_attr);
struct mlx5_flow_attr *attr;
- return kzalloc(sizeof(*attr) + ex_attr_size, GFP_KERNEL);
+ attr = kzalloc(sizeof(*attr) + ex_attr_size, GFP_KERNEL);
+ if (!attr)
+ return attr;
+
+ INIT_LIST_HEAD(&attr->list);
+ return attr;
}
static int
@@ -3686,6 +3970,7 @@ mlx5e_alloc_flow(struct mlx5e_priv *priv, int attr_size,
INIT_LIST_HEAD(&flow->encaps[out_index].list);
INIT_LIST_HEAD(&flow->hairpin);
INIT_LIST_HEAD(&flow->l3_to_l2_reformat);
+ INIT_LIST_HEAD(&flow->attrs);
refcount_set(&flow->refcnt, 1);
init_completion(&flow->init_done);
init_completion(&flow->del_hw_done);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.h b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.h
index 533c897bd517..a80b00946f1b 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.h
@@ -53,7 +53,6 @@
ESW_FLOW_ATTR_SZ :\
NIC_FLOW_ATTR_SZ)
-
int mlx5e_tc_num_filters(struct mlx5e_priv *priv, unsigned long flags);
struct mlx5e_tc_update_priv {
@@ -84,6 +83,8 @@ struct mlx5_flow_attr {
u8 tun_ip_version;
int tunnel_id; /* mapped tunnel id */
u32 flags;
+ struct list_head list;
+ struct mlx5e_post_act_handle *post_act_handle;
union {
struct mlx5_esw_flow_attr esw_attr[0];
struct mlx5_nic_flow_attr nic_attr[0];
--
2.34.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* [net-next 12/15] net/mlx5e: Use multi table support for CT and sample actions
2022-02-17 7:56 [pull request][net-next 00/15] mlx5 updates 2022-02-16 Saeed Mahameed
` (10 preceding siblings ...)
2022-02-17 7:56 ` [net-next 11/15] net/mlx5e: Create new flow attr for multi table actions Saeed Mahameed
@ 2022-02-17 7:56 ` Saeed Mahameed
2022-02-17 7:56 ` [net-next 13/15] net/mlx5e: TC, Clean redundant counter flag from tc action parsers Saeed Mahameed
` (2 subsequent siblings)
14 siblings, 0 replies; 17+ messages in thread
From: Saeed Mahameed @ 2022-02-17 7:56 UTC (permalink / raw)
To: David S. Miller, Jakub Kicinski; +Cc: netdev, Roi Dayan, Saeed Mahameed
From: Roi Dayan <roid@nvidia.com>
CT and sample actions use post actions for their implementation.
Flag those actions as multi table actions so the post act infrastructure
will handle the post actions allocation.
Signed-off-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
.../mellanox/mlx5/core/en/tc/act/ct.c | 12 +++
.../mellanox/mlx5/core/en/tc/act/sample.c | 21 ++++++
.../mellanox/mlx5/core/en/tc/act/sample.h | 14 ++++
.../mellanox/mlx5/core/en/tc/post_act.c | 3 +-
.../mellanox/mlx5/core/en/tc/sample.c | 74 +++++--------------
.../ethernet/mellanox/mlx5/core/en/tc_ct.c | 50 ++++---------
6 files changed, 80 insertions(+), 94 deletions(-)
create mode 100644 drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/sample.h
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/ct.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/ct.c
index 9fb1a9a8bc02..1492d3e49c59 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/ct.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/ct.c
@@ -56,8 +56,20 @@ tc_act_parse_ct(struct mlx5e_tc_act_parse_state *parse_state,
return 0;
}
+static bool
+tc_act_is_multi_table_act_ct(struct mlx5e_priv *priv,
+ const struct flow_action_entry *act,
+ struct mlx5_flow_attr *attr)
+{
+ if (act->ct.action & TCA_CT_ACT_CLEAR)
+ return false;
+
+ return true;
+}
+
struct mlx5e_tc_act mlx5e_tc_act_ct = {
.can_offload = tc_act_can_offload_ct,
.parse_action = tc_act_parse_ct,
+ .is_multi_table_act = tc_act_is_multi_table_act_ct,
};
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/sample.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/sample.c
index 539fea13ce9f..f39538e3d027 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/sample.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/sample.c
@@ -4,6 +4,7 @@
#include <net/psample.h>
#include "act.h"
#include "en/tc_priv.h"
+#include "en/tc/act/sample.h"
static bool
tc_act_can_offload_sample(struct mlx5e_tc_act_parse_state *parse_state,
@@ -42,7 +43,27 @@ tc_act_parse_sample(struct mlx5e_tc_act_parse_state *parse_state,
return 0;
}
+bool
+mlx5e_tc_act_sample_is_multi_table(struct mlx5_core_dev *mdev,
+ struct mlx5_flow_attr *attr)
+{
+ if (MLX5_CAP_GEN(mdev, reg_c_preserve) ||
+ attr->action & MLX5_FLOW_CONTEXT_ACTION_DECAP)
+ return true;
+
+ return false;
+}
+
+static bool
+tc_act_is_multi_table_act_sample(struct mlx5e_priv *priv,
+ const struct flow_action_entry *act,
+ struct mlx5_flow_attr *attr)
+{
+ return mlx5e_tc_act_sample_is_multi_table(priv->mdev, attr);
+}
+
struct mlx5e_tc_act mlx5e_tc_act_sample = {
.can_offload = tc_act_can_offload_sample,
.parse_action = tc_act_parse_sample,
+ .is_multi_table_act = tc_act_is_multi_table_act_sample,
};
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/sample.h b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/sample.h
new file mode 100644
index 000000000000..3efb3a15c5d2
--- /dev/null
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/sample.h
@@ -0,0 +1,14 @@
+/* SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB */
+/* Copyright (c) 2021, NVIDIA CORPORATION & AFFILIATES. All rights reserved. */
+
+#ifndef __MLX5_EN_TC_ACT_SAMPLE_H__
+#define __MLX5_EN_TC_ACT_SAMPLE_H__
+
+#include <net/flow_offload.h>
+#include "en/tc_priv.h"
+
+bool
+mlx5e_tc_act_sample_is_multi_table(struct mlx5_core_dev *mdev,
+ struct mlx5_flow_attr *attr);
+
+#endif /* __MLX5_EN_TC_ACT_SAMPLE_H__ */
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/post_act.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/post_act.c
index 27854ac844a0..32abc91adf23 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/post_act.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/post_act.c
@@ -126,8 +126,7 @@ mlx5e_tc_post_act_add(struct mlx5e_post_act *post_act, struct mlx5_flow_attr *at
post_attr->ft = post_act->ft;
post_attr->inner_match_level = MLX5_MATCH_NONE;
post_attr->outer_match_level = MLX5_MATCH_NONE;
- post_attr->action &= ~(MLX5_FLOW_CONTEXT_ACTION_DECAP);
- post_attr->flags &= ~MLX5_ATTR_FLAG_SAMPLE;
+ post_attr->action &= ~MLX5_FLOW_CONTEXT_ACTION_DECAP;
handle->ns_type = post_act->ns_type;
/* Splits were handled before post action */
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/sample.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/sample.c
index 90b86c4aaf51..fd4504518578 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/sample.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/sample.c
@@ -5,6 +5,7 @@
#include <net/psample.h>
#include "en/mapping.h"
#include "en/tc/post_act.h"
+#include "en/tc/act/sample.h"
#include "en/mod_hdr.h"
#include "sample.h"
#include "eswitch.h"
@@ -46,14 +47,12 @@ struct mlx5e_sample_flow {
struct mlx5_flow_handle *pre_rule;
struct mlx5_flow_attr *post_attr;
struct mlx5_flow_handle *post_rule;
- struct mlx5e_post_act_handle *post_act_handle;
};
struct mlx5e_sample_restore {
struct hlist_node hlist;
struct mlx5_modify_hdr *modify_hdr;
struct mlx5_flow_handle *rule;
- struct mlx5e_post_act_handle *post_act_handle;
u32 obj_id;
int count;
};
@@ -231,69 +230,46 @@ sampler_put(struct mlx5e_tc_psample *tc_psample, struct mlx5e_sampler *sampler)
*/
static struct mlx5_modify_hdr *
sample_modify_hdr_get(struct mlx5_core_dev *mdev, u32 obj_id,
- struct mlx5e_post_act_handle *handle)
+ struct mlx5e_tc_mod_hdr_acts *mod_acts)
{
- struct mlx5e_tc_mod_hdr_acts mod_acts = {};
struct mlx5_modify_hdr *modify_hdr;
int err;
- err = mlx5e_tc_match_to_reg_set(mdev, &mod_acts, MLX5_FLOW_NAMESPACE_FDB,
+ err = mlx5e_tc_match_to_reg_set(mdev, mod_acts, MLX5_FLOW_NAMESPACE_FDB,
CHAIN_TO_REG, obj_id);
if (err)
goto err_set_regc0;
- if (handle) {
- err = mlx5e_tc_post_act_set_handle(mdev, handle, &mod_acts);
- if (err)
- goto err_post_act;
- }
-
modify_hdr = mlx5_modify_header_alloc(mdev, MLX5_FLOW_NAMESPACE_FDB,
- mod_acts.num_actions,
- mod_acts.actions);
+ mod_acts->num_actions,
+ mod_acts->actions);
if (IS_ERR(modify_hdr)) {
err = PTR_ERR(modify_hdr);
goto err_modify_hdr;
}
- mlx5e_mod_hdr_dealloc(&mod_acts);
+ mlx5e_mod_hdr_dealloc(mod_acts);
return modify_hdr;
err_modify_hdr:
-err_post_act:
- mlx5e_mod_hdr_dealloc(&mod_acts);
+ mlx5e_mod_hdr_dealloc(mod_acts);
err_set_regc0:
return ERR_PTR(err);
}
-static u32
-restore_hash(u32 obj_id, struct mlx5e_post_act_handle *post_act_handle)
-{
- return jhash_2words(obj_id, hash32_ptr(post_act_handle), 0);
-}
-
-static bool
-restore_equal(struct mlx5e_sample_restore *restore, u32 obj_id,
- struct mlx5e_post_act_handle *post_act_handle)
-{
- return restore->obj_id == obj_id && restore->post_act_handle == post_act_handle;
-}
-
static struct mlx5e_sample_restore *
sample_restore_get(struct mlx5e_tc_psample *tc_psample, u32 obj_id,
- struct mlx5e_post_act_handle *post_act_handle)
+ struct mlx5e_tc_mod_hdr_acts *mod_acts)
{
struct mlx5_eswitch *esw = tc_psample->esw;
struct mlx5_core_dev *mdev = esw->dev;
struct mlx5e_sample_restore *restore;
struct mlx5_modify_hdr *modify_hdr;
- u32 hash_key;
int err;
mutex_lock(&tc_psample->restore_lock);
- hash_key = restore_hash(obj_id, post_act_handle);
- hash_for_each_possible(tc_psample->restore_hashtbl, restore, hlist, hash_key)
- if (restore_equal(restore, obj_id, post_act_handle))
+ hash_for_each_possible(tc_psample->restore_hashtbl, restore, hlist, obj_id)
+ if (restore->obj_id == obj_id)
goto add_ref;
restore = kzalloc(sizeof(*restore), GFP_KERNEL);
@@ -302,9 +278,8 @@ sample_restore_get(struct mlx5e_tc_psample *tc_psample, u32 obj_id,
goto err_alloc;
}
restore->obj_id = obj_id;
- restore->post_act_handle = post_act_handle;
- modify_hdr = sample_modify_hdr_get(mdev, obj_id, post_act_handle);
+ modify_hdr = sample_modify_hdr_get(mdev, obj_id, mod_acts);
if (IS_ERR(modify_hdr)) {
err = PTR_ERR(modify_hdr);
goto err_modify_hdr;
@@ -317,7 +292,7 @@ sample_restore_get(struct mlx5e_tc_psample *tc_psample, u32 obj_id,
goto err_restore;
}
- hash_add(tc_psample->restore_hashtbl, &restore->hlist, hash_key);
+ hash_add(tc_psample->restore_hashtbl, &restore->hlist, obj_id);
add_ref:
restore->count++;
mutex_unlock(&tc_psample->restore_lock);
@@ -494,10 +469,10 @@ mlx5e_tc_sample_offload(struct mlx5e_tc_psample *tc_psample,
struct mlx5_flow_spec *spec,
struct mlx5_flow_attr *attr)
{
- struct mlx5e_post_act_handle *post_act_handle = NULL;
struct mlx5_esw_flow_attr *esw_attr = attr->esw_attr;
struct mlx5_esw_flow_attr *pre_esw_attr;
struct mlx5_mapped_obj restore_obj = {};
+ struct mlx5e_tc_mod_hdr_acts *mod_acts;
struct mlx5e_sample_flow *sample_flow;
struct mlx5e_sample_attr *sample_attr;
struct mlx5_flow_attr *pre_attr;
@@ -522,21 +497,11 @@ mlx5e_tc_sample_offload(struct mlx5e_tc_psample *tc_psample,
* original flow table.
*/
esw = tc_psample->esw;
- if (MLX5_CAP_GEN(esw->dev, reg_c_preserve) ||
- attr->action & MLX5_FLOW_CONTEXT_ACTION_DECAP) {
+ if (mlx5e_tc_act_sample_is_multi_table(esw->dev, attr)) {
struct mlx5_flow_table *ft;
ft = mlx5e_tc_post_act_get_ft(tc_psample->post_act);
default_tbl_id = ft->id;
- post_act_handle = mlx5e_tc_post_act_add(tc_psample->post_act, attr);
- if (IS_ERR(post_act_handle)) {
- err = PTR_ERR(post_act_handle);
- goto err_post_act;
- }
- err = mlx5e_tc_post_act_offload(tc_psample->post_act, post_act_handle);
- if (err)
- goto err_post_rule;
- sample_flow->post_act_handle = post_act_handle;
} else {
err = add_post_rule(esw, sample_flow, spec, attr, &default_tbl_id);
if (err)
@@ -563,7 +528,8 @@ mlx5e_tc_sample_offload(struct mlx5e_tc_psample *tc_psample,
sample_attr->restore_obj_id = obj_id;
/* Create sample restore context. */
- sample_flow->restore = sample_restore_get(tc_psample, obj_id, post_act_handle);
+ mod_acts = &attr->parse_attr->mod_hdr_acts;
+ sample_flow->restore = sample_restore_get(tc_psample, obj_id, mod_acts);
if (IS_ERR(sample_flow->restore)) {
err = PTR_ERR(sample_flow->restore);
goto err_sample_restore;
@@ -589,6 +555,7 @@ mlx5e_tc_sample_offload(struct mlx5e_tc_psample *tc_psample,
pre_attr->outer_match_level = attr->outer_match_level;
pre_attr->chain = attr->chain;
pre_attr->prio = attr->prio;
+ pre_attr->ft = attr->ft;
pre_attr->sample_attr = *sample_attr;
pre_esw_attr = pre_attr->esw_attr;
pre_esw_attr->in_mdev = esw_attr->in_mdev;
@@ -614,9 +581,6 @@ mlx5e_tc_sample_offload(struct mlx5e_tc_psample *tc_psample,
if (sample_flow->post_rule)
del_post_rule(esw, sample_flow, attr);
err_post_rule:
- if (post_act_handle)
- mlx5e_tc_post_act_del(tc_psample->post_act, post_act_handle);
-err_post_act:
kfree(sample_flow);
return ERR_PTR(err);
}
@@ -642,9 +606,7 @@ mlx5e_tc_sample_unoffload(struct mlx5e_tc_psample *tc_psample,
sample_restore_put(tc_psample, sample_flow->restore);
mapping_remove(esw->offloads.reg_c0_obj_pool, attr->sample_attr.restore_obj_id);
sampler_put(tc_psample, sample_flow->sampler);
- if (sample_flow->post_act_handle)
- mlx5e_tc_post_act_del(tc_psample->post_act, sample_flow->post_act_handle);
- else
+ if (sample_flow->post_rule)
del_post_rule(esw, sample_flow, attr);
kfree(sample_flow->pre_attr);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c
index 228ebb19fcab..875e77af0ae6 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c
@@ -18,6 +18,7 @@
#include "lib/fs_chains.h"
#include "en/tc_ct.h"
+#include "en/tc_priv.h"
#include "en/mod_hdr.h"
#include "en/mapping.h"
#include "en/tc/post_act.h"
@@ -68,7 +69,6 @@ struct mlx5_tc_ct_priv {
struct mlx5_ct_flow {
struct mlx5_flow_attr *pre_ct_attr;
struct mlx5_flow_handle *pre_ct_rule;
- struct mlx5e_post_act_handle *post_act_handle;
struct mlx5_ct_ft *ft;
u32 chain_mapping;
};
@@ -1756,7 +1756,7 @@ mlx5_tc_ct_del_ft_cb(struct mlx5_tc_ct_priv *ct_priv, struct mlx5_ct_ft *ft)
/* We translate the tc filter with CT action to the following HW model:
*
* +---------------------+
- * + ft prio (tc chain) +
+ * + ft prio (tc chain) +
* + original match +
* +---------------------+
* | set chain miss mapping
@@ -1766,7 +1766,7 @@ mlx5_tc_ct_del_ft_cb(struct mlx5_tc_ct_priv *ct_priv, struct mlx5_ct_ft *ft)
* v
* +---------------------+
* + pre_ct/pre_ct_nat + if matches +-------------------------+
- * + zone+nat match +---------------->+ post_act (see below) +
+ * + zone+nat match +---------------->+ post_act (see below) +
* +---------------------+ set zone +-------------------------+
* | set zone
* v
@@ -1781,7 +1781,7 @@ mlx5_tc_ct_del_ft_cb(struct mlx5_tc_ct_priv *ct_priv, struct mlx5_ct_ft *ft)
* | do nat (if needed)
* v
* +--------------+
- * + post_act + original filter actions
+ * + post_act + original filter actions
* + fte_id match +------------------------>
* +--------------+
*/
@@ -1792,9 +1792,8 @@ __mlx5_tc_ct_flow_offload(struct mlx5_tc_ct_priv *ct_priv,
{
bool nat = attr->ct_attr.ct_action & TCA_CT_ACT_NAT;
struct mlx5e_priv *priv = netdev_priv(ct_priv->netdev);
- struct mlx5e_tc_mod_hdr_acts pre_mod_acts = {};
+ struct mlx5e_tc_mod_hdr_acts *pre_mod_acts;
u32 attr_sz = ns_to_attr_sz(ct_priv->ns_type);
- struct mlx5e_post_act_handle *handle;
struct mlx5_flow_attr *pre_ct_attr;
struct mlx5_modify_hdr *mod_hdr;
struct mlx5_ct_flow *ct_flow;
@@ -1817,17 +1816,6 @@ __mlx5_tc_ct_flow_offload(struct mlx5_tc_ct_priv *ct_priv,
}
ct_flow->ft = ft;
- handle = mlx5e_tc_post_act_add(ct_priv->post_act, attr);
- if (IS_ERR(handle)) {
- err = PTR_ERR(handle);
- ct_dbg("Failed to allocate post action handle");
- goto err_post_act_handle;
- }
- err = mlx5e_tc_post_act_offload(ct_priv->post_act, handle);
- if (err)
- goto err_alloc_pre;
- ct_flow->post_act_handle = handle;
-
/* Base flow attributes of both rules on original rule attribute */
ct_flow->pre_ct_attr = mlx5_alloc_flow_attr(ct_priv->ns_type);
if (!ct_flow->pre_ct_attr) {
@@ -1837,6 +1825,7 @@ __mlx5_tc_ct_flow_offload(struct mlx5_tc_ct_priv *ct_priv,
pre_ct_attr = ct_flow->pre_ct_attr;
memcpy(pre_ct_attr, attr, attr_sz);
+ pre_mod_acts = &pre_ct_attr->parse_attr->mod_hdr_acts;
/* Modify the original rule's action to fwd and modify, leave decap */
pre_ct_attr->action = attr->action & MLX5_FLOW_CONTEXT_ACTION_DECAP;
@@ -1855,25 +1844,19 @@ __mlx5_tc_ct_flow_offload(struct mlx5_tc_ct_priv *ct_priv,
}
ct_flow->chain_mapping = chain_mapping;
- err = mlx5e_tc_match_to_reg_set(priv->mdev, &pre_mod_acts, ct_priv->ns_type,
+ err = mlx5e_tc_match_to_reg_set(priv->mdev, pre_mod_acts, ct_priv->ns_type,
CHAIN_TO_REG, chain_mapping);
if (err) {
ct_dbg("Failed to set chain register mapping");
goto err_mapping;
}
- err = mlx5e_tc_post_act_set_handle(priv->mdev, handle, &pre_mod_acts);
- if (err) {
- ct_dbg("Failed to set post action handle");
- goto err_mapping;
- }
-
/* If original flow is decap, we do it before going into ct table
* so add a rewrite for the tunnel match_id.
*/
if ((pre_ct_attr->action & MLX5_FLOW_CONTEXT_ACTION_DECAP) &&
attr->chain == 0) {
- err = mlx5e_tc_match_to_reg_set(priv->mdev, &pre_mod_acts,
+ err = mlx5e_tc_match_to_reg_set(priv->mdev, pre_mod_acts,
ct_priv->ns_type,
TUNNEL_TO_REG,
attr->tunnel_id);
@@ -1884,8 +1867,8 @@ __mlx5_tc_ct_flow_offload(struct mlx5_tc_ct_priv *ct_priv,
}
mod_hdr = mlx5_modify_header_alloc(priv->mdev, ct_priv->ns_type,
- pre_mod_acts.num_actions,
- pre_mod_acts.actions);
+ pre_mod_acts->num_actions,
+ pre_mod_acts->actions);
if (IS_ERR(mod_hdr)) {
err = PTR_ERR(mod_hdr);
ct_dbg("Failed to create pre ct mod hdr");
@@ -1905,20 +1888,18 @@ __mlx5_tc_ct_flow_offload(struct mlx5_tc_ct_priv *ct_priv,
}
attr->ct_attr.ct_flow = ct_flow;
- mlx5e_mod_hdr_dealloc(&pre_mod_acts);
+ mlx5e_mod_hdr_dealloc(pre_mod_acts);
return ct_flow->pre_ct_rule;
err_insert_orig:
mlx5_modify_header_dealloc(priv->mdev, pre_ct_attr->modify_hdr);
err_mapping:
- mlx5e_mod_hdr_dealloc(&pre_mod_acts);
+ mlx5e_mod_hdr_dealloc(pre_mod_acts);
mlx5_chains_put_chain_mapping(ct_priv->chains, ct_flow->chain_mapping);
err_get_chain:
kfree(ct_flow->pre_ct_attr);
err_alloc_pre:
- mlx5e_tc_post_act_del(ct_priv->post_act, handle);
-err_post_act_handle:
mlx5_tc_ct_del_ft_cb(ct_priv, ft);
err_ft:
kfree(ct_flow);
@@ -1955,11 +1936,8 @@ __mlx5_tc_ct_delete_flow(struct mlx5_tc_ct_priv *ct_priv,
mlx5_tc_rule_delete(priv, ct_flow->pre_ct_rule, pre_ct_attr);
mlx5_modify_header_dealloc(priv->mdev, pre_ct_attr->modify_hdr);
- if (ct_flow->post_act_handle) {
- mlx5_chains_put_chain_mapping(ct_priv->chains, ct_flow->chain_mapping);
- mlx5e_tc_post_act_del(ct_priv->post_act, ct_flow->post_act_handle);
- mlx5_tc_ct_del_ft_cb(ct_priv, ct_flow->ft);
- }
+ mlx5_chains_put_chain_mapping(ct_priv->chains, ct_flow->chain_mapping);
+ mlx5_tc_ct_del_ft_cb(ct_priv, ct_flow->ft);
kfree(ct_flow->pre_ct_attr);
kfree(ct_flow);
--
2.34.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* [net-next 13/15] net/mlx5e: TC, Clean redundant counter flag from tc action parsers
2022-02-17 7:56 [pull request][net-next 00/15] mlx5 updates 2022-02-16 Saeed Mahameed
` (11 preceding siblings ...)
2022-02-17 7:56 ` [net-next 12/15] net/mlx5e: Use multi table support for CT and sample actions Saeed Mahameed
@ 2022-02-17 7:56 ` Saeed Mahameed
2022-02-17 7:56 ` [net-next 14/15] net/mlx5e: TC, Make post_act parse CT and sample actions Saeed Mahameed
2022-02-17 7:56 ` [net-next 15/15] net/mlx5e: TC, Allow sample action with CT Saeed Mahameed
14 siblings, 0 replies; 17+ messages in thread
From: Saeed Mahameed @ 2022-02-17 7:56 UTC (permalink / raw)
To: David S. Miller, Jakub Kicinski; +Cc: netdev, Roi Dayan, Saeed Mahameed
From: Roi Dayan <roid@nvidia.com>
When tc actions being parsed only the last flow attr created needs the
counter flag and the previous flags being reset.
Clean the flag from the tc action parsers.
Signed-off-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/accept.c | 3 +--
drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/drop.c | 3 +--
drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/goto.c | 3 +--
drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/mirred.c | 3 +--
drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/mirred_nic.c | 3 +--
.../ethernet/mellanox/mlx5/core/en/tc/act/redirect_ingress.c | 3 +--
drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/trap.c | 3 +--
drivers/net/ethernet/mellanox/mlx5/core/en_tc.c | 1 -
8 files changed, 7 insertions(+), 15 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/accept.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/accept.c
index 2b53738938a9..21aab96357b5 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/accept.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/accept.c
@@ -19,8 +19,7 @@ tc_act_parse_accept(struct mlx5e_tc_act_parse_state *parse_state,
struct mlx5e_priv *priv,
struct mlx5_flow_attr *attr)
{
- attr->action |= MLX5_FLOW_CONTEXT_ACTION_FWD_DEST |
- MLX5_FLOW_CONTEXT_ACTION_COUNT;
+ attr->action |= MLX5_FLOW_CONTEXT_ACTION_FWD_DEST;
attr->flags |= MLX5_ATTR_FLAG_ACCEPT;
return 0;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/drop.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/drop.c
index 3d5f23636a02..dd025a95c439 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/drop.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/drop.c
@@ -19,8 +19,7 @@ tc_act_parse_drop(struct mlx5e_tc_act_parse_state *parse_state,
struct mlx5e_priv *priv,
struct mlx5_flow_attr *attr)
{
- attr->action |= MLX5_FLOW_CONTEXT_ACTION_DROP |
- MLX5_FLOW_CONTEXT_ACTION_COUNT;
+ attr->action |= MLX5_FLOW_CONTEXT_ACTION_DROP;
return 0;
}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/goto.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/goto.c
index fb1be822ad25..4726bcb46eec 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/goto.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/goto.c
@@ -76,8 +76,7 @@ tc_act_parse_goto(struct mlx5e_tc_act_parse_state *parse_state,
struct mlx5e_priv *priv,
struct mlx5_flow_attr *attr)
{
- attr->action |= MLX5_FLOW_CONTEXT_ACTION_FWD_DEST |
- MLX5_FLOW_CONTEXT_ACTION_COUNT;
+ attr->action |= MLX5_FLOW_CONTEXT_ACTION_FWD_DEST;
attr->dest_chain = act->chain_index;
return 0;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/mirred.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/mirred.c
index 99fb98b3e71b..d08abec93704 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/mirred.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/mirred.c
@@ -296,8 +296,7 @@ tc_act_parse_mirred(struct mlx5e_tc_act_parse_state *parse_state,
if (err)
return err;
- attr->action |= MLX5_FLOW_CONTEXT_ACTION_FWD_DEST |
- MLX5_FLOW_CONTEXT_ACTION_COUNT;
+ attr->action |= MLX5_FLOW_CONTEXT_ACTION_FWD_DEST;
return 0;
}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/mirred_nic.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/mirred_nic.c
index 16681cf6e93e..90b4c1b34776 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/mirred_nic.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/mirred_nic.c
@@ -40,8 +40,7 @@ tc_act_parse_mirred_nic(struct mlx5e_tc_act_parse_state *parse_state,
{
attr->parse_attr->mirred_ifindex[0] = act->dev->ifindex;
flow_flag_set(parse_state->flow, HAIRPIN);
- attr->action |= MLX5_FLOW_CONTEXT_ACTION_FWD_DEST |
- MLX5_FLOW_CONTEXT_ACTION_COUNT;
+ attr->action |= MLX5_FLOW_CONTEXT_ACTION_FWD_DEST;
return 0;
}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/redirect_ingress.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/redirect_ingress.c
index 9dd244147385..ad09a8a5f36e 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/redirect_ingress.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/redirect_ingress.c
@@ -58,8 +58,7 @@ tc_act_parse_redirect_ingress(struct mlx5e_tc_act_parse_state *parse_state,
struct net_device *out_dev = act->dev;
int err;
- attr->action |= MLX5_FLOW_CONTEXT_ACTION_FWD_DEST |
- MLX5_FLOW_CONTEXT_ACTION_COUNT;
+ attr->action |= MLX5_FLOW_CONTEXT_ACTION_FWD_DEST;
err = mlx5e_set_fwd_to_int_port_actions(priv, attr, out_dev->ifindex,
MLX5E_TC_INT_PORT_INGRESS,
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/trap.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/trap.c
index 9ea293fdc434..a7d9eab19e4a 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/trap.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/trap.c
@@ -26,8 +26,7 @@ tc_act_parse_trap(struct mlx5e_tc_act_parse_state *parse_state,
struct mlx5e_priv *priv,
struct mlx5_flow_attr *attr)
{
- attr->action |= MLX5_FLOW_CONTEXT_ACTION_FWD_DEST |
- MLX5_FLOW_CONTEXT_ACTION_COUNT;
+ attr->action |= MLX5_FLOW_CONTEXT_ACTION_FWD_DEST;
attr->flags |= MLX5_ATTR_FLAG_SLOW_PATH;
return 0;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index b9d6a2e8b240..76a015dfc5fc 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -3536,7 +3536,6 @@ alloc_flow_post_acts(struct mlx5e_tc_flow *flow, struct netlink_ext_ack *extack)
/* Set counter action on last post act rule. */
attr->action |= MLX5_FLOW_CONTEXT_ACTION_COUNT;
} else {
- attr->action &= ~MLX5_FLOW_CONTEXT_ACTION_COUNT;
err = mlx5e_tc_act_set_next_post_act(flow, attr, next_attr);
if (err)
goto out_free;
--
2.34.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* [net-next 14/15] net/mlx5e: TC, Make post_act parse CT and sample actions
2022-02-17 7:56 [pull request][net-next 00/15] mlx5 updates 2022-02-16 Saeed Mahameed
` (12 preceding siblings ...)
2022-02-17 7:56 ` [net-next 13/15] net/mlx5e: TC, Clean redundant counter flag from tc action parsers Saeed Mahameed
@ 2022-02-17 7:56 ` Saeed Mahameed
2022-02-17 7:56 ` [net-next 15/15] net/mlx5e: TC, Allow sample action with CT Saeed Mahameed
14 siblings, 0 replies; 17+ messages in thread
From: Saeed Mahameed @ 2022-02-17 7:56 UTC (permalink / raw)
To: David S. Miller, Jakub Kicinski; +Cc: netdev, Roi Dayan, Saeed Mahameed
From: Roi Dayan <roid@nvidia.com>
Before this commit post_act can be used for normal rules
and didn't handle special cases like CT and sample.
With this commit post_act rule can also handle the special cases
when needed.
Signed-off-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en/tc/post_act.c | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/post_act.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/post_act.c
index 32abc91adf23..dea137dd744b 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/post_act.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/post_act.c
@@ -1,6 +1,7 @@
// SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB
// Copyright (c) 2021, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+#include "en/tc_priv.h"
#include "en_tc.h"
#include "post_act.h"
#include "mlx5_core.h"
@@ -89,7 +90,7 @@ mlx5e_tc_post_act_offload(struct mlx5e_post_act *post_act,
/* Post action rule matches on fte_id and executes original rule's tc rule action */
mlx5e_tc_match_to_reg_match(spec, FTEID_TO_REG, handle->id, MLX5_POST_ACTION_MASK);
- handle->rule = mlx5_tc_rule_insert(post_act->priv, spec, handle->attr);
+ handle->rule = mlx5e_tc_rule_offload(post_act->priv, spec, handle->attr);
if (IS_ERR(handle->rule)) {
err = PTR_ERR(handle->rule);
netdev_warn(post_act->priv->netdev, "Failed to add post action rule");
@@ -152,7 +153,7 @@ void
mlx5e_tc_post_act_unoffload(struct mlx5e_post_act *post_act,
struct mlx5e_post_act_handle *handle)
{
- mlx5_tc_rule_delete(post_act->priv, handle->rule, handle->attr);
+ mlx5e_tc_rule_unoffload(post_act->priv, handle->rule, handle->attr);
handle->rule = NULL;
}
--
2.34.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* [net-next 15/15] net/mlx5e: TC, Allow sample action with CT
2022-02-17 7:56 [pull request][net-next 00/15] mlx5 updates 2022-02-16 Saeed Mahameed
` (13 preceding siblings ...)
2022-02-17 7:56 ` [net-next 14/15] net/mlx5e: TC, Make post_act parse CT and sample actions Saeed Mahameed
@ 2022-02-17 7:56 ` Saeed Mahameed
14 siblings, 0 replies; 17+ messages in thread
From: Saeed Mahameed @ 2022-02-17 7:56 UTC (permalink / raw)
To: David S. Miller, Jakub Kicinski; +Cc: netdev, Roi Dayan, Saeed Mahameed
From: Roi Dayan <roid@nvidia.com>
Allow sample+CT actions but still block sample+CT NAT
as it is not supported.
Signed-off-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/ct.c | 6 ------
.../net/ethernet/mellanox/mlx5/core/en/tc/act/sample.c | 8 +++++---
2 files changed, 5 insertions(+), 9 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/ct.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/ct.c
index 1492d3e49c59..7368f95f2310 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/ct.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/ct.c
@@ -14,12 +14,6 @@ tc_act_can_offload_ct(struct mlx5e_tc_act_parse_state *parse_state,
bool clear_action = act->ct.action & TCA_CT_ACT_CLEAR;
struct netlink_ext_ack *extack = parse_state->extack;
- if (flow_flag_test(parse_state->flow, SAMPLE)) {
- NL_SET_ERR_MSG_MOD(extack,
- "Sample action with connection tracking is not supported");
- return false;
- }
-
if (parse_state->ct && !clear_action) {
NL_SET_ERR_MSG_MOD(extack, "Multiple CT actions are not supported");
return false;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/sample.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/sample.c
index f39538e3d027..2c0196431302 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/sample.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/sample.c
@@ -13,10 +13,12 @@ tc_act_can_offload_sample(struct mlx5e_tc_act_parse_state *parse_state,
struct mlx5_flow_attr *attr)
{
struct netlink_ext_ack *extack = parse_state->extack;
+ bool ct_nat;
- if (flow_flag_test(parse_state->flow, CT)) {
- NL_SET_ERR_MSG_MOD(extack,
- "Sample action with connection tracking is not supported");
+ ct_nat = attr->ct_attr.ct_action & TCA_CT_ACT_NAT;
+
+ if (flow_flag_test(parse_state->flow, CT) && ct_nat) {
+ NL_SET_ERR_MSG_MOD(extack, "Sample action with CT NAT is not supported");
return false;
}
--
2.34.1
^ permalink raw reply related [flat|nested] 17+ messages in thread