* [PATCH V2 mlx5-next 00/11] Mellanox, mlx5-next updates 2019-04-25
@ 2019-04-29 18:13 Saeed Mahameed
2019-04-29 18:14 ` [PATCH V2 mlx5-next 01/11] net/mlx5: E-Switch: Introduce prio tag mode Saeed Mahameed
` (11 more replies)
0 siblings, 12 replies; 17+ messages in thread
From: Saeed Mahameed @ 2019-04-29 18:13 UTC (permalink / raw)
To: Saeed Mahameed, Leon Romanovsky; +Cc: Jason Gunthorpe, netdev, linux-rdma
Hi,
This series provides misc low level updates to mlx5 core driver, to be
shared between rdma and net-next trees.
1) From Aya: Enable general events on all physical link types and
restrict general event handling of subtype DELAY_DROP_TIMEOUT in mlx5 rdma
driver to ethernet links only as it was intended.
2) From Eli: Introduce low level bits for prio tag mode
3) From Maor: Low level steering updates to support RDMA RX flow
steering and enables RoCE loopback traffic when switchdev is enabled.
4) From Vu and Parav: Two small mlx5 core cleanups
5) From Yevgeny add HW definitions of geneve offloads
In case of no objections this series will be applied to mlx5-next
branch.
v2:
- Address comments from Leon.
- Static checker fixes.
Thanks,
Saeed.
---
Aya Levin (2):
IB/mlx5: Restrict 'DELAY_DROP_TIMEOUT' subtype to Ethernet interfaces
net/mlx5: Enable general events on all interfaces
Eli Britstein (1):
net/mlx5: E-Switch: Introduce prio tag mode
Maor Gottlieb (4):
net/mlx5: Pass flow steering objects to fs_cmd
net/mlx5: Add support in RDMA RX steering
net/mlx5: Add new miss flow table action
net/mlx5: Eswitch, enable RoCE loopback traffic
Parav Pandit (1):
net/mlx5: Get rid of storing copy of device name
Vu Pham (1):
net/mlx5: Separate and generalize dma device from pci device
Yevgeny Kliteynik (2):
net/mlx5: Geneve, Add basic Geneve encap/decap flow table capabilities
net/mlx5: Geneve, Add flow table capabilities for Geneve decap with
TLV options
drivers/infiniband/hw/mlx5/main.c | 16 +-
.../net/ethernet/mellanox/mlx5/core/Makefile | 2 +-
.../net/ethernet/mellanox/mlx5/core/alloc.c | 19 +-
drivers/net/ethernet/mellanox/mlx5/core/cmd.c | 9 +-
.../mlx5/core/diag/fw_tracer_tracepoint.h | 5 +-
.../net/ethernet/mellanox/mlx5/core/en_main.c | 14 +-
.../net/ethernet/mellanox/mlx5/core/en_rep.c | 2 +-
.../net/ethernet/mellanox/mlx5/core/en_tc.c | 5 +-
drivers/net/ethernet/mellanox/mlx5/core/eq.c | 3 +-
.../net/ethernet/mellanox/mlx5/core/eswitch.h | 8 +-
.../mellanox/mlx5/core/eswitch_offloads.c | 4 +
.../ethernet/mellanox/mlx5/core/fpga/ipsec.c | 86 +++++----
.../net/ethernet/mellanox/mlx5/core/fs_cmd.c | 113 ++++++-----
.../net/ethernet/mellanox/mlx5/core/fs_cmd.h | 33 ++--
.../net/ethernet/mellanox/mlx5/core/fs_core.c | 73 ++++---
.../net/ethernet/mellanox/mlx5/core/fs_core.h | 5 +-
.../net/ethernet/mellanox/mlx5/core/health.c | 2 +-
.../net/ethernet/mellanox/mlx5/core/main.c | 15 +-
.../ethernet/mellanox/mlx5/core/mlx5_core.h | 65 ++++---
.../ethernet/mellanox/mlx5/core/pagealloc.c | 20 +-
.../net/ethernet/mellanox/mlx5/core/rdma.c | 182 ++++++++++++++++++
.../net/ethernet/mellanox/mlx5/core/rdma.h | 20 ++
include/linux/mlx5/device.h | 10 +-
include/linux/mlx5/driver.h | 10 +-
include/linux/mlx5/fs.h | 1 +
include/linux/mlx5/mlx5_ifc.h | 82 ++++++--
26 files changed, 566 insertions(+), 238 deletions(-)
create mode 100644 drivers/net/ethernet/mellanox/mlx5/core/rdma.c
create mode 100644 drivers/net/ethernet/mellanox/mlx5/core/rdma.h
--
2.20.1
^ permalink raw reply [flat|nested] 17+ messages in thread
* [PATCH V2 mlx5-next 01/11] net/mlx5: E-Switch: Introduce prio tag mode
2019-04-29 18:13 [PATCH V2 mlx5-next 00/11] Mellanox, mlx5-next updates 2019-04-25 Saeed Mahameed
@ 2019-04-29 18:14 ` Saeed Mahameed
2019-04-29 18:14 ` [PATCH V2 mlx5-next 02/11] net/mlx5: Get rid of storing copy of device name Saeed Mahameed
` (10 subsequent siblings)
11 siblings, 0 replies; 17+ messages in thread
From: Saeed Mahameed @ 2019-04-29 18:14 UTC (permalink / raw)
To: Saeed Mahameed, Leon Romanovsky
Cc: Jason Gunthorpe, netdev, linux-rdma, Eli Britstein, Oz Shlomo
From: Eli Britstein <elibr@mellanox.com>
Current ConnectX HW is unable to perform VLAN pop in TX path and VLAN
push on RX path. To workaround that limitation untagged packets will be
tagged with VLAN ID 0x000 (priority tag) and pop/push actions will be
replaced by VLAN re-write actions (which are supported by the HW).
Introduce prio tag mode as a pre-step to controlling the workaround
behavior.
Signed-off-by: Eli Britstein <elibr@mellanox.com>
Reviewed-by: Oz Shlomo <ozsh@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
include/linux/mlx5/mlx5_ifc.h | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/include/linux/mlx5/mlx5_ifc.h b/include/linux/mlx5/mlx5_ifc.h
index 4b37519bd6a5..eeedf3f53ed3 100644
--- a/include/linux/mlx5/mlx5_ifc.h
+++ b/include/linux/mlx5/mlx5_ifc.h
@@ -951,7 +951,9 @@ struct mlx5_ifc_cmd_hca_cap_bits {
u8 log_max_srq_sz[0x8];
u8 log_max_qp_sz[0x8];
- u8 reserved_at_90[0xb];
+ u8 reserved_at_90[0x8];
+ u8 prio_tag_required[0x1];
+ u8 reserved_at_99[0x2];
u8 log_max_qp[0x5];
u8 reserved_at_a0[0xb];
--
2.20.1
^ permalink raw reply related [flat|nested] 17+ messages in thread
* [PATCH V2 mlx5-next 02/11] net/mlx5: Get rid of storing copy of device name
2019-04-29 18:13 [PATCH V2 mlx5-next 00/11] Mellanox, mlx5-next updates 2019-04-25 Saeed Mahameed
2019-04-29 18:14 ` [PATCH V2 mlx5-next 01/11] net/mlx5: E-Switch: Introduce prio tag mode Saeed Mahameed
@ 2019-04-29 18:14 ` Saeed Mahameed
2019-04-29 18:14 ` [PATCH V2 mlx5-next 03/11] net/mlx5: Separate and generalize dma device from pci device Saeed Mahameed
` (9 subsequent siblings)
11 siblings, 0 replies; 17+ messages in thread
From: Saeed Mahameed @ 2019-04-29 18:14 UTC (permalink / raw)
To: Saeed Mahameed, Leon Romanovsky
Cc: Jason Gunthorpe, netdev, linux-rdma, Parav Pandit
From: Parav Pandit <parav@mellanox.com>
Currently mlx5 core stores copy of the PCI device name in a
mlx5_priv structure and uses pr_warn, pr_err helpers.
Get rid of the copy of this name; instead store the parent device
pointer that contains name as well as dma specific parameters.
This also allows to use kernel's well defined dev_warn, dev_err, dev_dbg
device specific print routines.
This is also a preparation patch to access non PCI parent device in
future.
Signed-off-by: Parav Pandit <parav@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx5/core/cmd.c | 2 +-
.../mlx5/core/diag/fw_tracer_tracepoint.h | 5 +-
.../net/ethernet/mellanox/mlx5/core/en_tc.c | 5 +-
.../net/ethernet/mellanox/mlx5/core/eswitch.h | 8 +--
.../net/ethernet/mellanox/mlx5/core/health.c | 2 +-
.../net/ethernet/mellanox/mlx5/core/main.c | 15 ++---
.../ethernet/mellanox/mlx5/core/mlx5_core.h | 65 ++++++++++---------
.../ethernet/mellanox/mlx5/core/pagealloc.c | 5 +-
include/linux/mlx5/driver.h | 3 +-
9 files changed, 57 insertions(+), 53 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
index 0a2ffe794a54..779fa1e9ee4e 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
@@ -1347,7 +1347,7 @@ static void set_wqname(struct mlx5_core_dev *dev)
struct mlx5_cmd *cmd = &dev->cmd;
snprintf(cmd->wq_name, sizeof(cmd->wq_name), "mlx5_cmd_%s",
- dev->priv.name);
+ dev_name(dev->device));
}
static void clean_debug_files(struct mlx5_core_dev *dev)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/diag/fw_tracer_tracepoint.h b/drivers/net/ethernet/mellanox/mlx5/core/diag/fw_tracer_tracepoint.h
index 7b5901d42994..3038be575923 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/diag/fw_tracer_tracepoint.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/diag/fw_tracer_tracepoint.h
@@ -47,7 +47,7 @@ TRACE_EVENT(mlx5_fw,
TP_ARGS(tracer, trace_timestamp, lost, event_id, msg),
TP_STRUCT__entry(
- __string(dev_name, tracer->dev->priv.name)
+ __string(dev_name, dev_name(tracer->dev->device))
__field(u64, trace_timestamp)
__field(bool, lost)
__field(u8, event_id)
@@ -55,7 +55,8 @@ TRACE_EVENT(mlx5_fw,
),
TP_fast_assign(
- __assign_str(dev_name, tracer->dev->priv.name);
+ __assign_str(dev_name,
+ dev_name(tracer->dev->device));
__entry->trace_timestamp = trace_timestamp;
__entry->lost = lost;
__entry->event_id = event_id;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index b4967a0ff8c7..a40e801cf713 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -663,7 +663,8 @@ static int mlx5e_hairpin_flow_add(struct mlx5e_priv *priv,
}
netdev_dbg(priv->netdev, "add hairpin: tirn %x rqn %x peer %s sqn %x prio %d (log) data %d packets %d\n",
- hp->tirn, hp->pair->rqn[0], hp->pair->peer_mdev->priv.name,
+ hp->tirn, hp->pair->rqn[0],
+ dev_name(hp->pair->peer_mdev->device),
hp->pair->sqn[0], match_prio, params.log_data_size, params.log_num_packets);
hpe->hp = hp;
@@ -700,7 +701,7 @@ static void mlx5e_hairpin_flow_del(struct mlx5e_priv *priv,
hpe = list_entry(next, struct mlx5e_hairpin_entry, flows);
netdev_dbg(priv->netdev, "del hairpin: peer %s\n",
- hpe->hp->pair->peer_mdev->priv.name);
+ dev_name(hpe->hp->pair->peer_mdev->device));
mlx5e_hairpin_destroy(hpe->hp);
hash_del(&hpe->hairpin_hlist);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h
index 3f3cd32ae60a..c1d6bb90cf04 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h
@@ -376,11 +376,11 @@ bool mlx5_esw_multipath_prereq(struct mlx5_core_dev *dev0,
#define MLX5_DEBUG_ESWITCH_MASK BIT(3)
-#define esw_info(dev, format, ...) \
- pr_info("(%s): E-Switch: " format, (dev)->priv.name, ##__VA_ARGS__)
+#define esw_info(__dev, format, ...) \
+ dev_info((__dev)->device, "E-Switch: " format, ##__VA_ARGS__)
-#define esw_warn(dev, format, ...) \
- pr_warn("(%s): E-Switch: " format, (dev)->priv.name, ##__VA_ARGS__)
+#define esw_warn(__dev, format, ...) \
+ dev_warn((__dev)->device, "E-Switch: " format, ##__VA_ARGS__)
#define esw_debug(dev, format, ...) \
mlx5_core_dbg_mask(dev, MLX5_DEBUG_ESWITCH_MASK, format, ##__VA_ARGS__)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/health.c b/drivers/net/ethernet/mellanox/mlx5/core/health.c
index 3b98fcdd7d0e..a2656f4008d9 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/health.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/health.c
@@ -380,7 +380,7 @@ int mlx5_health_init(struct mlx5_core_dev *dev)
return -ENOMEM;
strcpy(name, "mlx5_health");
- strcat(name, dev->priv.name);
+ strcat(name, dev_name(dev->device));
health->wq = create_singlethread_workqueue(name);
kfree(name);
if (!health->wq)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/main.c b/drivers/net/ethernet/mellanox/mlx5/core/main.c
index b200a29d1420..bedf809737b5 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/main.c
@@ -741,7 +741,6 @@ static int mlx5_pci_init(struct mlx5_core_dev *dev, struct pci_dev *pdev,
struct mlx5_priv *priv = &dev->priv;
int err = 0;
- dev->pdev = pdev;
priv->pci_dev_data = id->driver_data;
pci_set_drvdata(dev->pdev, dev);
@@ -1242,14 +1241,11 @@ static const struct devlink_ops mlx5_devlink_ops = {
#endif
};
-static int mlx5_mdev_init(struct mlx5_core_dev *dev, int profile_idx, const char *name)
+static int mlx5_mdev_init(struct mlx5_core_dev *dev, int profile_idx)
{
struct mlx5_priv *priv = &dev->priv;
int err;
- strncpy(priv->name, name, MLX5_MAX_NAME_LEN);
- priv->name[MLX5_MAX_NAME_LEN - 1] = 0;
-
dev->profile = &profile[profile_idx];
INIT_LIST_HEAD(&priv->ctx_list);
@@ -1267,9 +1263,10 @@ static int mlx5_mdev_init(struct mlx5_core_dev *dev, int profile_idx, const char
INIT_LIST_HEAD(&priv->pgdir_list);
spin_lock_init(&priv->mkey_lock);
- priv->dbg_root = debugfs_create_dir(name, mlx5_debugfs_root);
+ priv->dbg_root = debugfs_create_dir(dev_name(dev->device),
+ mlx5_debugfs_root);
if (!priv->dbg_root) {
- pr_err("mlx5_core: %s error, Cannot create debugfs dir, aborting\n", name);
+ dev_err(dev->device, "mlx5_core: error, Cannot create debugfs dir, aborting\n");
return -ENOMEM;
}
@@ -1312,8 +1309,10 @@ static int init_one(struct pci_dev *pdev, const struct pci_device_id *id)
}
dev = devlink_priv(devlink);
+ dev->device = &pdev->dev;
+ dev->pdev = pdev;
- err = mlx5_mdev_init(dev, prof_sel, dev_name(&pdev->dev));
+ err = mlx5_mdev_init(dev, prof_sel);
if (err)
goto mdev_init_err;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h b/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h
index d66f4f082ef7..0e111b70c178 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h
@@ -41,6 +41,7 @@
#include <linux/ptp_clock_kernel.h>
#include <linux/mlx5/cq.h>
#include <linux/mlx5/fs.h>
+#include <linux/mlx5/driver.h>
#define DRIVER_NAME "mlx5_core"
#define DRIVER_VERSION "5.0-0"
@@ -48,53 +49,57 @@
extern uint mlx5_core_debug_mask;
#define mlx5_core_dbg(__dev, format, ...) \
- pr_debug("%s:%s:%d:(pid %d): " format, (__dev)->priv.name, \
+ dev_dbg((__dev)->device, "%s:%d:(pid %d): " format, \
__func__, __LINE__, current->pid, \
##__VA_ARGS__)
-#define mlx5_core_dbg_once(__dev, format, ...) \
- pr_debug_once("%s:%s:%d:(pid %d): " format, (__dev)->priv.name, \
- __func__, __LINE__, current->pid, \
+#define mlx5_core_dbg_once(__dev, format, ...) \
+ dev_dbg_once((__dev)->device, \
+ "%s:%d:(pid %d): " format, \
+ __func__, __LINE__, current->pid, \
##__VA_ARGS__)
-#define mlx5_core_dbg_mask(__dev, mask, format, ...) \
-do { \
- if ((mask) & mlx5_core_debug_mask) \
- mlx5_core_dbg(__dev, format, ##__VA_ARGS__); \
+#define mlx5_core_dbg_mask(__dev, mask, format, ...) \
+do { \
+ if ((mask) & mlx5_core_debug_mask) \
+ mlx5_core_dbg(__dev, format, ##__VA_ARGS__); \
} while (0)
-#define mlx5_core_err(__dev, format, ...) \
- pr_err("%s:%s:%d:(pid %d): " format, (__dev)->priv.name, \
- __func__, __LINE__, current->pid, \
+#define mlx5_core_err(__dev, format, ...) \
+ dev_err((__dev)->device, "%s:%d:(pid %d): " format, \
+ __func__, __LINE__, current->pid, \
##__VA_ARGS__)
-#define mlx5_core_err_rl(__dev, format, ...) \
- pr_err_ratelimited("%s:%s:%d:(pid %d): " format, (__dev)->priv.name, \
- __func__, __LINE__, current->pid, \
- ##__VA_ARGS__)
+#define mlx5_core_err_rl(__dev, format, ...) \
+ dev_err_ratelimited((__dev)->device, \
+ "%s:%d:(pid %d): " format, \
+ __func__, __LINE__, current->pid, \
+ ##__VA_ARGS__)
-#define mlx5_core_warn(__dev, format, ...) \
- pr_warn("%s:%s:%d:(pid %d): " format, (__dev)->priv.name, \
- __func__, __LINE__, current->pid, \
- ##__VA_ARGS__)
+#define mlx5_core_warn(__dev, format, ...) \
+ dev_warn((__dev)->device, "%s:%d:(pid %d): " format, \
+ __func__, __LINE__, current->pid, \
+ ##__VA_ARGS__)
#define mlx5_core_warn_once(__dev, format, ...) \
- pr_warn_once("%s:%s:%d:(pid %d): " format, (__dev)->priv.name, \
+ dev_warn_once((__dev)->device, "%s:%d:(pid %d): " format, \
__func__, __LINE__, current->pid, \
##__VA_ARGS__)
-#define mlx5_core_warn_rl(__dev, format, ...) \
- pr_warn_ratelimited("%s:%s:%d:(pid %d): " format, (__dev)->priv.name, \
- __func__, __LINE__, current->pid, \
- ##__VA_ARGS__)
+#define mlx5_core_warn_rl(__dev, format, ...) \
+ dev_warn_ratelimited((__dev)->device, \
+ "%s:%d:(pid %d): " format, \
+ __func__, __LINE__, current->pid, \
+ ##__VA_ARGS__)
-#define mlx5_core_info(__dev, format, ...) \
- pr_info("%s " format, (__dev)->priv.name, ##__VA_ARGS__)
+#define mlx5_core_info(__dev, format, ...) \
+ dev_info((__dev)->device, format, ##__VA_ARGS__)
-#define mlx5_core_info_rl(__dev, format, ...) \
- pr_info_ratelimited("%s:%s:%d:(pid %d): " format, (__dev)->priv.name, \
- __func__, __LINE__, current->pid, \
- ##__VA_ARGS__)
+#define mlx5_core_info_rl(__dev, format, ...) \
+ dev_info_ratelimited((__dev)->device, \
+ "%s:%d:(pid %d): " format, \
+ __func__, __LINE__, current->pid, \
+ ##__VA_ARGS__)
enum {
MLX5_CMD_DATA, /* print command payload only */
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/pagealloc.c b/drivers/net/ethernet/mellanox/mlx5/core/pagealloc.c
index 41025387ff2c..d87d42f4f6dd 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/pagealloc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/pagealloc.c
@@ -600,8 +600,7 @@ int mlx5_wait_for_pages(struct mlx5_core_dev *dev, int *pages)
return 0;
}
- mlx5_core_dbg(dev, "Waiting for %d pages from %s\n", prev_pages,
- dev->priv.name);
+ mlx5_core_dbg(dev, "Waiting for %d pages\n", prev_pages);
while (*pages) {
if (time_after(jiffies, end)) {
mlx5_core_warn(dev, "aborting while there are %d pending pages\n", *pages);
@@ -614,6 +613,6 @@ int mlx5_wait_for_pages(struct mlx5_core_dev *dev, int *pages)
msleep(50);
}
- mlx5_core_dbg(dev, "All pages received from %s\n", dev->priv.name);
+ mlx5_core_dbg(dev, "All pages received\n");
return 0;
}
diff --git a/include/linux/mlx5/driver.h b/include/linux/mlx5/driver.h
index 6c43191c0186..582a9680b182 100644
--- a/include/linux/mlx5/driver.h
+++ b/include/linux/mlx5/driver.h
@@ -56,7 +56,6 @@
enum {
MLX5_BOARD_ID_LEN = 64,
- MLX5_MAX_NAME_LEN = 16,
};
enum {
@@ -514,7 +513,6 @@ struct mlx5_rl_table {
};
struct mlx5_priv {
- char name[MLX5_MAX_NAME_LEN];
struct mlx5_eq_table *eq_table;
/* pages stuff */
@@ -641,6 +639,7 @@ struct mlx5_fw_tracer;
struct mlx5_vxlan;
struct mlx5_core_dev {
+ struct device *device;
struct pci_dev *pdev;
/* sync pci state */
struct mutex pci_status_mutex;
--
2.20.1
^ permalink raw reply related [flat|nested] 17+ messages in thread
* [PATCH V2 mlx5-next 03/11] net/mlx5: Separate and generalize dma device from pci device
2019-04-29 18:13 [PATCH V2 mlx5-next 00/11] Mellanox, mlx5-next updates 2019-04-25 Saeed Mahameed
2019-04-29 18:14 ` [PATCH V2 mlx5-next 01/11] net/mlx5: E-Switch: Introduce prio tag mode Saeed Mahameed
2019-04-29 18:14 ` [PATCH V2 mlx5-next 02/11] net/mlx5: Get rid of storing copy of device name Saeed Mahameed
@ 2019-04-29 18:14 ` Saeed Mahameed
2019-04-29 18:14 ` [PATCH V2 mlx5-next 04/11] IB/mlx5: Restrict 'DELAY_DROP_TIMEOUT' subtype to Ethernet interfaces Saeed Mahameed
` (8 subsequent siblings)
11 siblings, 0 replies; 17+ messages in thread
From: Saeed Mahameed @ 2019-04-29 18:14 UTC (permalink / raw)
To: Saeed Mahameed, Leon Romanovsky
Cc: Jason Gunthorpe, netdev, linux-rdma, Vu Pham, Parav Pandit
From: Vu Pham <vuhuong@mellanox.com>
The mlx5 Sub-Function (SF) sub device will be introduced in
subsequent patches. It will be created as mediated device and
belong to mdev bus. It is necessary to treat dma operations on
PF, VF and SF in uniform way, hence reduce the dependency on
pdev pci dev struct and work directly out of newly introduced
'struct device' from previous patch.
This patch does not change any functionality.
Signed-off-by: Vu Pham <vuhuong@mellanox.com>
Reviewed-by: Parav Pandit <parav@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
drivers/infiniband/hw/mlx5/main.c | 10 ++++++----
.../net/ethernet/mellanox/mlx5/core/alloc.c | 19 ++++++++++---------
drivers/net/ethernet/mellanox/mlx5/core/cmd.c | 7 +++----
.../net/ethernet/mellanox/mlx5/core/en_main.c | 14 +++++++-------
.../net/ethernet/mellanox/mlx5/core/en_rep.c | 2 +-
.../ethernet/mellanox/mlx5/core/pagealloc.c | 15 +++++++--------
6 files changed, 34 insertions(+), 33 deletions(-)
diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index 3c76ee24a17e..44d00cc5ffe2 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -181,7 +181,7 @@ static int mlx5_netdev_event(struct notifier_block *this,
ibdev->rep->vport);
if (rep_ndev == ndev)
roce->netdev = ndev;
- } else if (ndev->dev.parent == &mdev->pdev->dev) {
+ } else if (ndev->dev.parent == mdev->device) {
roce->netdev = ndev;
}
write_unlock(&roce->netdev_lock);
@@ -5666,7 +5666,8 @@ static int mlx5_ib_init_multiport_master(struct mlx5_ib_dev *dev)
}
if (bound) {
- dev_dbg(&mpi->mdev->pdev->dev, "removing port from unaffiliated list.\n");
+ dev_dbg(mpi->mdev->device,
+ "removing port from unaffiliated list.\n");
mlx5_ib_dbg(dev, "port %d bound\n", i + 1);
list_del(&mpi->list);
break;
@@ -5865,7 +5866,7 @@ int mlx5_ib_stage_init_init(struct mlx5_ib_dev *dev)
dev->ib_dev.local_dma_lkey = 0 /* not supported for now */;
dev->ib_dev.phys_port_cnt = dev->num_ports;
dev->ib_dev.num_comp_vectors = mlx5_comp_vectors_count(mdev);
- dev->ib_dev.dev.parent = &mdev->pdev->dev;
+ dev->ib_dev.dev.parent = mdev->device;
mutex_init(&dev->cap_mask_mutex);
INIT_LIST_HEAD(&dev->qp_list);
@@ -6554,7 +6555,8 @@ static void *mlx5_ib_add_slave_port(struct mlx5_core_dev *mdev)
if (!bound) {
list_add_tail(&mpi->list, &mlx5_ib_unaffiliated_port_list);
- dev_dbg(&mdev->pdev->dev, "no suitable IB device found to bind to, added to unaffiliated list.\n");
+ dev_dbg(mdev->device,
+ "no suitable IB device found to bind to, added to unaffiliated list.\n");
}
mutex_unlock(&mlx5_ib_multiport_mutex);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/alloc.c b/drivers/net/ethernet/mellanox/mlx5/core/alloc.c
index 9008e17126db..549f962cd86e 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/alloc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/alloc.c
@@ -57,15 +57,16 @@ static void *mlx5_dma_zalloc_coherent_node(struct mlx5_core_dev *dev,
int node)
{
struct mlx5_priv *priv = &dev->priv;
+ struct device *device = dev->device;
int original_node;
void *cpu_handle;
mutex_lock(&priv->alloc_mutex);
- original_node = dev_to_node(&dev->pdev->dev);
- set_dev_node(&dev->pdev->dev, node);
- cpu_handle = dma_alloc_coherent(&dev->pdev->dev, size, dma_handle,
+ original_node = dev_to_node(device);
+ set_dev_node(device, node);
+ cpu_handle = dma_alloc_coherent(device, size, dma_handle,
GFP_KERNEL);
- set_dev_node(&dev->pdev->dev, original_node);
+ set_dev_node(device, original_node);
mutex_unlock(&priv->alloc_mutex);
return cpu_handle;
}
@@ -110,7 +111,7 @@ EXPORT_SYMBOL(mlx5_buf_alloc);
void mlx5_buf_free(struct mlx5_core_dev *dev, struct mlx5_frag_buf *buf)
{
- dma_free_coherent(&dev->pdev->dev, buf->size, buf->frags->buf,
+ dma_free_coherent(dev->device, buf->size, buf->frags->buf,
buf->frags->map);
kfree(buf->frags);
@@ -139,7 +140,7 @@ int mlx5_frag_buf_alloc_node(struct mlx5_core_dev *dev, int size,
if (!frag->buf)
goto err_free_buf;
if (frag->map & ((1 << buf->page_shift) - 1)) {
- dma_free_coherent(&dev->pdev->dev, frag_sz,
+ dma_free_coherent(dev->device, frag_sz,
buf->frags[i].buf, buf->frags[i].map);
mlx5_core_warn(dev, "unexpected map alignment: %pad, page_shift=%d\n",
&frag->map, buf->page_shift);
@@ -152,7 +153,7 @@ int mlx5_frag_buf_alloc_node(struct mlx5_core_dev *dev, int size,
err_free_buf:
while (i--)
- dma_free_coherent(&dev->pdev->dev, PAGE_SIZE, buf->frags[i].buf,
+ dma_free_coherent(dev->device, PAGE_SIZE, buf->frags[i].buf,
buf->frags[i].map);
kfree(buf->frags);
err_out:
@@ -168,7 +169,7 @@ void mlx5_frag_buf_free(struct mlx5_core_dev *dev, struct mlx5_frag_buf *buf)
for (i = 0; i < buf->npages; i++) {
int frag_sz = min_t(int, size, PAGE_SIZE);
- dma_free_coherent(&dev->pdev->dev, frag_sz, buf->frags[i].buf,
+ dma_free_coherent(dev->device, frag_sz, buf->frags[i].buf,
buf->frags[i].map);
size -= frag_sz;
}
@@ -274,7 +275,7 @@ void mlx5_db_free(struct mlx5_core_dev *dev, struct mlx5_db *db)
__set_bit(db->index, db->u.pgdir->bitmap);
if (bitmap_full(db->u.pgdir->bitmap, db_per_page)) {
- dma_free_coherent(&(dev->pdev->dev), PAGE_SIZE,
+ dma_free_coherent(dev->device, PAGE_SIZE,
db->u.pgdir->db_page, db->u.pgdir->db_dma);
list_del(&db->u.pgdir->list);
bitmap_free(db->u.pgdir->bitmap);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
index 779fa1e9ee4e..746c8cc95e48 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
@@ -1852,7 +1852,7 @@ static void create_msg_cache(struct mlx5_core_dev *dev)
static int alloc_cmd_page(struct mlx5_core_dev *dev, struct mlx5_cmd *cmd)
{
- struct device *ddev = &dev->pdev->dev;
+ struct device *ddev = dev->device;
cmd->cmd_alloc_buf = dma_alloc_coherent(ddev, MLX5_ADAPTER_PAGE_SIZE,
&cmd->alloc_dma, GFP_KERNEL);
@@ -1883,7 +1883,7 @@ static int alloc_cmd_page(struct mlx5_core_dev *dev, struct mlx5_cmd *cmd)
static void free_cmd_page(struct mlx5_core_dev *dev, struct mlx5_cmd *cmd)
{
- struct device *ddev = &dev->pdev->dev;
+ struct device *ddev = dev->device;
dma_free_coherent(ddev, cmd->alloc_size, cmd->cmd_alloc_buf,
cmd->alloc_dma);
@@ -1908,8 +1908,7 @@ int mlx5_cmd_init(struct mlx5_core_dev *dev)
return -EINVAL;
}
- cmd->pool = dma_pool_create("mlx5_cmd", &dev->pdev->dev, size, align,
- 0);
+ cmd->pool = dma_pool_create("mlx5_cmd", dev->device, size, align, 0);
if (!cmd->pool)
return -ENOMEM;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index b5fdbd3190d9..ae0f154d16db 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -1891,7 +1891,7 @@ static int mlx5e_open_channel(struct mlx5e_priv *priv, int ix,
c->tstamp = &priv->tstamp;
c->ix = ix;
c->cpu = cpu;
- c->pdev = &priv->mdev->pdev->dev;
+ c->pdev = priv->mdev->device;
c->netdev = priv->netdev;
c->mkey_be = cpu_to_be32(priv->mdev->mlx5e_res.mkey.key);
c->num_tc = params->num_tc;
@@ -2137,7 +2137,7 @@ static void mlx5e_build_rq_param(struct mlx5e_priv *priv,
MLX5_SET(rqc, rqc, vsd, params->vlan_strip_disable);
MLX5_SET(rqc, rqc, scatter_fcs, params->scatter_fcs_en);
- param->wq.buf_numa_node = dev_to_node(&mdev->pdev->dev);
+ param->wq.buf_numa_node = dev_to_node(mdev->device);
}
static void mlx5e_build_drop_rq_param(struct mlx5e_priv *priv,
@@ -2152,7 +2152,7 @@ static void mlx5e_build_drop_rq_param(struct mlx5e_priv *priv,
mlx5e_get_rqwq_log_stride(MLX5_WQ_TYPE_CYCLIC, 1));
MLX5_SET(rqc, rqc, counter_set_id, priv->drop_rq_q_counter);
- param->wq.buf_numa_node = dev_to_node(&mdev->pdev->dev);
+ param->wq.buf_numa_node = dev_to_node(mdev->device);
}
static void mlx5e_build_sq_param_common(struct mlx5e_priv *priv,
@@ -2164,7 +2164,7 @@ static void mlx5e_build_sq_param_common(struct mlx5e_priv *priv,
MLX5_SET(wq, wq, log_wq_stride, ilog2(MLX5_SEND_WQE_BB));
MLX5_SET(wq, wq, pd, priv->mdev->mlx5e_res.pdn);
- param->wq.buf_numa_node = dev_to_node(&priv->mdev->pdev->dev);
+ param->wq.buf_numa_node = dev_to_node(priv->mdev->device);
}
static void mlx5e_build_sq_param(struct mlx5e_priv *priv,
@@ -3046,8 +3046,8 @@ static int mlx5e_alloc_drop_cq(struct mlx5_core_dev *mdev,
struct mlx5e_cq *cq,
struct mlx5e_cq_param *param)
{
- param->wq.buf_numa_node = dev_to_node(&mdev->pdev->dev);
- param->wq.db_numa_node = dev_to_node(&mdev->pdev->dev);
+ param->wq.buf_numa_node = dev_to_node(mdev->device);
+ param->wq.db_numa_node = dev_to_node(mdev->device);
return mlx5e_alloc_cq_common(mdev, param, cq);
}
@@ -4639,7 +4639,7 @@ static void mlx5e_build_nic_netdev(struct net_device *netdev)
bool fcs_supported;
bool fcs_enabled;
- SET_NETDEV_DEV(netdev, &mdev->pdev->dev);
+ SET_NETDEV_DEV(netdev, mdev->device);
netdev->netdev_ops = &mlx5e_netdev_ops;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
index a66b6ed80b30..fc21403f140e 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
@@ -1389,7 +1389,7 @@ static void mlx5e_build_rep_netdev(struct net_device *netdev)
struct mlx5_core_dev *mdev = priv->mdev;
if (rep->vport == MLX5_VPORT_UPLINK) {
- SET_NETDEV_DEV(netdev, &priv->mdev->pdev->dev);
+ SET_NETDEV_DEV(netdev, mdev->device);
netdev->netdev_ops = &mlx5e_netdev_ops_uplink_rep;
/* we want a persistent mac for the uplink rep */
mlx5_query_nic_vport_mac_address(mdev, 0, netdev->dev_addr);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/pagealloc.c b/drivers/net/ethernet/mellanox/mlx5/core/pagealloc.c
index d87d42f4f6dd..91bd258ecf1b 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/pagealloc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/pagealloc.c
@@ -200,7 +200,7 @@ static void free_4k(struct mlx5_core_dev *dev, u64 addr)
rb_erase(&fwp->rb_node, &dev->priv.page_root);
if (fwp->free_count != 1)
list_del(&fwp->list);
- dma_unmap_page(&dev->pdev->dev, addr & MLX5_U64_4K_PAGE_MASK,
+ dma_unmap_page(dev->device, addr & MLX5_U64_4K_PAGE_MASK,
PAGE_SIZE, DMA_BIDIRECTIONAL);
__free_page(fwp->page);
kfree(fwp);
@@ -211,11 +211,12 @@ static void free_4k(struct mlx5_core_dev *dev, u64 addr)
static int alloc_system_page(struct mlx5_core_dev *dev, u16 func_id)
{
+ struct device *device = dev->device;
+ int nid = dev_to_node(device);
struct page *page;
u64 zero_addr = 1;
u64 addr;
int err;
- int nid = dev_to_node(&dev->pdev->dev);
page = alloc_pages_node(nid, GFP_HIGHUSER, 0);
if (!page) {
@@ -223,9 +224,8 @@ static int alloc_system_page(struct mlx5_core_dev *dev, u16 func_id)
return -ENOMEM;
}
map:
- addr = dma_map_page(&dev->pdev->dev, page, 0,
- PAGE_SIZE, DMA_BIDIRECTIONAL);
- if (dma_mapping_error(&dev->pdev->dev, addr)) {
+ addr = dma_map_page(device, page, 0, PAGE_SIZE, DMA_BIDIRECTIONAL);
+ if (dma_mapping_error(device, addr)) {
mlx5_core_warn(dev, "failed dma mapping page\n");
err = -ENOMEM;
goto err_mapping;
@@ -240,8 +240,7 @@ static int alloc_system_page(struct mlx5_core_dev *dev, u16 func_id)
err = insert_page(dev, addr, page, func_id);
if (err) {
mlx5_core_err(dev, "failed to track allocated page\n");
- dma_unmap_page(&dev->pdev->dev, addr, PAGE_SIZE,
- DMA_BIDIRECTIONAL);
+ dma_unmap_page(device, addr, PAGE_SIZE, DMA_BIDIRECTIONAL);
}
err_mapping:
@@ -249,7 +248,7 @@ static int alloc_system_page(struct mlx5_core_dev *dev, u16 func_id)
__free_page(page);
if (zero_addr == 0)
- dma_unmap_page(&dev->pdev->dev, zero_addr, PAGE_SIZE,
+ dma_unmap_page(device, zero_addr, PAGE_SIZE,
DMA_BIDIRECTIONAL);
return err;
--
2.20.1
^ permalink raw reply related [flat|nested] 17+ messages in thread
* [PATCH V2 mlx5-next 04/11] IB/mlx5: Restrict 'DELAY_DROP_TIMEOUT' subtype to Ethernet interfaces
2019-04-29 18:13 [PATCH V2 mlx5-next 00/11] Mellanox, mlx5-next updates 2019-04-25 Saeed Mahameed
` (2 preceding siblings ...)
2019-04-29 18:14 ` [PATCH V2 mlx5-next 03/11] net/mlx5: Separate and generalize dma device from pci device Saeed Mahameed
@ 2019-04-29 18:14 ` Saeed Mahameed
2019-04-29 18:14 ` [PATCH V2 mlx5-next 05/11] net/mlx5: Enable general events on all interfaces Saeed Mahameed
` (7 subsequent siblings)
11 siblings, 0 replies; 17+ messages in thread
From: Saeed Mahameed @ 2019-04-29 18:14 UTC (permalink / raw)
To: Saeed Mahameed, Leon Romanovsky
Cc: Jason Gunthorpe, netdev, linux-rdma, Aya Levin
From: Aya Levin <ayal@mellanox.com>
Subtype 'DELAY_DROP_TIMEOUT' (under 'GENERAL' event) is restricted to
Ethernet interfaces. This patch doesn't change functionality or breaks
current flow. In the downstream patch, non Ethernet (like IB) interfaces
will receive 'GENERAL' event.
Fixes: 5d3c537f9070 ("net/mlx5: Handle event of power detection in the PCIE slot")
Signed-off-by: Aya Levin <ayal@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
drivers/infiniband/hw/mlx5/main.c | 6 +++++-
1 file changed, 5 insertions(+), 1 deletion(-)
diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index 44d00cc5ffe2..fae6a6a1fbea 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -4347,9 +4347,13 @@ static void delay_drop_handler(struct work_struct *work)
static void handle_general_event(struct mlx5_ib_dev *ibdev, struct mlx5_eqe *eqe,
struct ib_event *ibev)
{
+ u8 port = (eqe->data.port.port >> 4) & 0xf;
+
switch (eqe->sub_type) {
case MLX5_GENERAL_SUBTYPE_DELAY_DROP_TIMEOUT:
- schedule_work(&ibdev->delay_drop.delay_drop_work);
+ if (mlx5_ib_port_link_layer(&ibdev->ib_dev, port) ==
+ IB_LINK_LAYER_ETHERNET)
+ schedule_work(&ibdev->delay_drop.delay_drop_work);
break;
default: /* do nothing */
return;
--
2.20.1
^ permalink raw reply related [flat|nested] 17+ messages in thread
* [PATCH V2 mlx5-next 05/11] net/mlx5: Enable general events on all interfaces
2019-04-29 18:13 [PATCH V2 mlx5-next 00/11] Mellanox, mlx5-next updates 2019-04-25 Saeed Mahameed
` (3 preceding siblings ...)
2019-04-29 18:14 ` [PATCH V2 mlx5-next 04/11] IB/mlx5: Restrict 'DELAY_DROP_TIMEOUT' subtype to Ethernet interfaces Saeed Mahameed
@ 2019-04-29 18:14 ` Saeed Mahameed
2019-04-29 18:14 ` [PATCH V2 mlx5-next 06/11] net/mlx5: Pass flow steering objects to fs_cmd Saeed Mahameed
` (6 subsequent siblings)
11 siblings, 0 replies; 17+ messages in thread
From: Saeed Mahameed @ 2019-04-29 18:14 UTC (permalink / raw)
To: Saeed Mahameed, Leon Romanovsky
Cc: Jason Gunthorpe, netdev, linux-rdma, Aya Levin
From: Aya Levin <ayal@mellanox.com>
Open events of type 'GENERAL' to all types of interfaces. Prior to this
patch, 'GENERAL' events were captured only by Ethernet interfaces. Other
interface types (non-Ethernet) were excluded and couldn't receive
'GENERAL' events.
Fixes: 5d3c537f9070 ("net/mlx5: Handle event of power detection in the PCIE slot")
Signed-off-by: Aya Levin <ayal@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx5/core/eq.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eq.c b/drivers/net/ethernet/mellanox/mlx5/core/eq.c
index bb6e5b5d9681..3b617b5e1d9d 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eq.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eq.c
@@ -504,8 +504,7 @@ static u64 gather_async_events_mask(struct mlx5_core_dev *dev)
if (MLX5_VPORT_MANAGER(dev))
async_event_mask |= (1ull << MLX5_EVENT_TYPE_NIC_VPORT_CHANGE);
- if (MLX5_CAP_GEN(dev, port_type) == MLX5_CAP_PORT_TYPE_ETH &&
- MLX5_CAP_GEN(dev, general_notification_event))
+ if (MLX5_CAP_GEN(dev, general_notification_event))
async_event_mask |= (1ull << MLX5_EVENT_TYPE_GENERAL_EVENT);
if (MLX5_CAP_GEN(dev, port_module_event))
--
2.20.1
^ permalink raw reply related [flat|nested] 17+ messages in thread
* [PATCH V2 mlx5-next 06/11] net/mlx5: Pass flow steering objects to fs_cmd
2019-04-29 18:13 [PATCH V2 mlx5-next 00/11] Mellanox, mlx5-next updates 2019-04-25 Saeed Mahameed
` (4 preceding siblings ...)
2019-04-29 18:14 ` [PATCH V2 mlx5-next 05/11] net/mlx5: Enable general events on all interfaces Saeed Mahameed
@ 2019-04-29 18:14 ` Saeed Mahameed
2019-04-29 18:14 ` [PATCH V2 mlx5-next 07/11] net/mlx5: Add support in RDMA RX steering Saeed Mahameed
` (5 subsequent siblings)
11 siblings, 0 replies; 17+ messages in thread
From: Saeed Mahameed @ 2019-04-29 18:14 UTC (permalink / raw)
To: Saeed Mahameed, Leon Romanovsky
Cc: Jason Gunthorpe, netdev, linux-rdma, Maor Gottlieb
From: Maor Gottlieb <maorg@mellanox.com>
Pass the flow steering objects instead of their attributes
to fs_cmd in order to decrease number of arguments and in
addition it will be used to update object fields.
Pass the flow steering root namespace instead of the device
so will have context to the namespace in the fs_cmd layer.
Signed-off-by: Maor Gottlieb <maorg@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
.../ethernet/mellanox/mlx5/core/fpga/ipsec.c | 86 ++++++++--------
.../net/ethernet/mellanox/mlx5/core/fs_cmd.c | 99 ++++++++++---------
.../net/ethernet/mellanox/mlx5/core/fs_cmd.h | 33 +++----
.../net/ethernet/mellanox/mlx5/core/fs_core.c | 42 ++++----
4 files changed, 128 insertions(+), 132 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fpga/ipsec.c b/drivers/net/ethernet/mellanox/mlx5/core/fpga/ipsec.c
index 5a22c5874f3b..52c47d3dd5a5 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fpga/ipsec.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fpga/ipsec.c
@@ -989,32 +989,33 @@ static enum fs_flow_table_type egress_to_fs_ft(bool egress)
return egress ? FS_FT_NIC_TX : FS_FT_NIC_RX;
}
-static int fpga_ipsec_fs_create_flow_group(struct mlx5_core_dev *dev,
+static int fpga_ipsec_fs_create_flow_group(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
u32 *in,
- unsigned int *group_id,
+ struct mlx5_flow_group *fg,
bool is_egress)
{
- int (*create_flow_group)(struct mlx5_core_dev *dev,
+ int (*create_flow_group)(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft, u32 *in,
- unsigned int *group_id) =
+ struct mlx5_flow_group *fg) =
mlx5_fs_cmd_get_default(egress_to_fs_ft(is_egress))->create_flow_group;
char *misc_params_c = MLX5_ADDR_OF(create_flow_group_in, in,
match_criteria.misc_parameters);
+ struct mlx5_core_dev *dev = ns->dev;
u32 saved_outer_esp_spi_mask;
u8 match_criteria_enable;
int ret;
if (MLX5_CAP_FLOWTABLE(dev,
flow_table_properties_nic_receive.ft_field_support.outer_esp_spi))
- return create_flow_group(dev, ft, in, group_id);
+ return create_flow_group(ns, ft, in, fg);
match_criteria_enable =
MLX5_GET(create_flow_group_in, in, match_criteria_enable);
saved_outer_esp_spi_mask =
MLX5_GET(fte_match_set_misc, misc_params_c, outer_esp_spi);
if (!match_criteria_enable || !saved_outer_esp_spi_mask)
- return create_flow_group(dev, ft, in, group_id);
+ return create_flow_group(ns, ft, in, fg);
MLX5_SET(fte_match_set_misc, misc_params_c, outer_esp_spi, 0);
@@ -1023,7 +1024,7 @@ static int fpga_ipsec_fs_create_flow_group(struct mlx5_core_dev *dev,
MLX5_SET(create_flow_group_in, in, match_criteria_enable,
match_criteria_enable & ~MLX5_MATCH_MISC_PARAMETERS);
- ret = create_flow_group(dev, ft, in, group_id);
+ ret = create_flow_group(ns, ft, in, fg);
MLX5_SET(fte_match_set_misc, misc_params_c, outer_esp_spi, saved_outer_esp_spi_mask);
MLX5_SET(create_flow_group_in, in, match_criteria_enable, match_criteria_enable);
@@ -1031,17 +1032,18 @@ static int fpga_ipsec_fs_create_flow_group(struct mlx5_core_dev *dev,
return ret;
}
-static int fpga_ipsec_fs_create_fte(struct mlx5_core_dev *dev,
+static int fpga_ipsec_fs_create_fte(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
struct mlx5_flow_group *fg,
struct fs_fte *fte,
bool is_egress)
{
- int (*create_fte)(struct mlx5_core_dev *dev,
+ int (*create_fte)(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
struct mlx5_flow_group *fg,
struct fs_fte *fte) =
mlx5_fs_cmd_get_default(egress_to_fs_ft(is_egress))->create_fte;
+ struct mlx5_core_dev *dev = ns->dev;
struct mlx5_fpga_device *fdev = dev->fpga;
struct mlx5_fpga_ipsec *fipsec = fdev->ipsec;
struct mlx5_fpga_ipsec_rule *rule;
@@ -1053,7 +1055,7 @@ static int fpga_ipsec_fs_create_fte(struct mlx5_core_dev *dev,
!(fte->action.action &
(MLX5_FLOW_CONTEXT_ACTION_ENCRYPT |
MLX5_FLOW_CONTEXT_ACTION_DECRYPT)))
- return create_fte(dev, ft, fg, fte);
+ return create_fte(ns, ft, fg, fte);
rule = kzalloc(sizeof(*rule), GFP_KERNEL);
if (!rule)
@@ -1070,7 +1072,7 @@ static int fpga_ipsec_fs_create_fte(struct mlx5_core_dev *dev,
WARN_ON(rule_insert(fipsec, rule));
modify_spec_mailbox(dev, fte, &mbox_mod);
- ret = create_fte(dev, ft, fg, fte);
+ ret = create_fte(ns, ft, fg, fte);
restore_spec_mailbox(fte, &mbox_mod);
if (ret) {
_rule_delete(fipsec, rule);
@@ -1081,19 +1083,20 @@ static int fpga_ipsec_fs_create_fte(struct mlx5_core_dev *dev,
return ret;
}
-static int fpga_ipsec_fs_update_fte(struct mlx5_core_dev *dev,
+static int fpga_ipsec_fs_update_fte(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
- unsigned int group_id,
+ struct mlx5_flow_group *fg,
int modify_mask,
struct fs_fte *fte,
bool is_egress)
{
- int (*update_fte)(struct mlx5_core_dev *dev,
+ int (*update_fte)(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
- unsigned int group_id,
+ struct mlx5_flow_group *fg,
int modify_mask,
struct fs_fte *fte) =
mlx5_fs_cmd_get_default(egress_to_fs_ft(is_egress))->update_fte;
+ struct mlx5_core_dev *dev = ns->dev;
bool is_esp = fte->action.esp_id;
struct mailbox_mod mbox_mod;
int ret;
@@ -1102,24 +1105,25 @@ static int fpga_ipsec_fs_update_fte(struct mlx5_core_dev *dev,
!(fte->action.action &
(MLX5_FLOW_CONTEXT_ACTION_ENCRYPT |
MLX5_FLOW_CONTEXT_ACTION_DECRYPT)))
- return update_fte(dev, ft, group_id, modify_mask, fte);
+ return update_fte(ns, ft, fg, modify_mask, fte);
modify_spec_mailbox(dev, fte, &mbox_mod);
- ret = update_fte(dev, ft, group_id, modify_mask, fte);
+ ret = update_fte(ns, ft, fg, modify_mask, fte);
restore_spec_mailbox(fte, &mbox_mod);
return ret;
}
-static int fpga_ipsec_fs_delete_fte(struct mlx5_core_dev *dev,
+static int fpga_ipsec_fs_delete_fte(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
struct fs_fte *fte,
bool is_egress)
{
- int (*delete_fte)(struct mlx5_core_dev *dev,
+ int (*delete_fte)(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
struct fs_fte *fte) =
mlx5_fs_cmd_get_default(egress_to_fs_ft(is_egress))->delete_fte;
+ struct mlx5_core_dev *dev = ns->dev;
struct mlx5_fpga_device *fdev = dev->fpga;
struct mlx5_fpga_ipsec *fipsec = fdev->ipsec;
struct mlx5_fpga_ipsec_rule *rule;
@@ -1131,7 +1135,7 @@ static int fpga_ipsec_fs_delete_fte(struct mlx5_core_dev *dev,
!(fte->action.action &
(MLX5_FLOW_CONTEXT_ACTION_ENCRYPT |
MLX5_FLOW_CONTEXT_ACTION_DECRYPT)))
- return delete_fte(dev, ft, fte);
+ return delete_fte(ns, ft, fte);
rule = rule_search(fipsec, fte);
if (!rule)
@@ -1141,84 +1145,84 @@ static int fpga_ipsec_fs_delete_fte(struct mlx5_core_dev *dev,
rule_delete(fipsec, rule);
modify_spec_mailbox(dev, fte, &mbox_mod);
- ret = delete_fte(dev, ft, fte);
+ ret = delete_fte(ns, ft, fte);
restore_spec_mailbox(fte, &mbox_mod);
return ret;
}
static int
-mlx5_fpga_ipsec_fs_create_flow_group_egress(struct mlx5_core_dev *dev,
+mlx5_fpga_ipsec_fs_create_flow_group_egress(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
u32 *in,
- unsigned int *group_id)
+ struct mlx5_flow_group *fg)
{
- return fpga_ipsec_fs_create_flow_group(dev, ft, in, group_id, true);
+ return fpga_ipsec_fs_create_flow_group(ns, ft, in, fg, true);
}
static int
-mlx5_fpga_ipsec_fs_create_fte_egress(struct mlx5_core_dev *dev,
+mlx5_fpga_ipsec_fs_create_fte_egress(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
struct mlx5_flow_group *fg,
struct fs_fte *fte)
{
- return fpga_ipsec_fs_create_fte(dev, ft, fg, fte, true);
+ return fpga_ipsec_fs_create_fte(ns, ft, fg, fte, true);
}
static int
-mlx5_fpga_ipsec_fs_update_fte_egress(struct mlx5_core_dev *dev,
+mlx5_fpga_ipsec_fs_update_fte_egress(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
- unsigned int group_id,
+ struct mlx5_flow_group *fg,
int modify_mask,
struct fs_fte *fte)
{
- return fpga_ipsec_fs_update_fte(dev, ft, group_id, modify_mask, fte,
+ return fpga_ipsec_fs_update_fte(ns, ft, fg, modify_mask, fte,
true);
}
static int
-mlx5_fpga_ipsec_fs_delete_fte_egress(struct mlx5_core_dev *dev,
+mlx5_fpga_ipsec_fs_delete_fte_egress(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
struct fs_fte *fte)
{
- return fpga_ipsec_fs_delete_fte(dev, ft, fte, true);
+ return fpga_ipsec_fs_delete_fte(ns, ft, fte, true);
}
static int
-mlx5_fpga_ipsec_fs_create_flow_group_ingress(struct mlx5_core_dev *dev,
+mlx5_fpga_ipsec_fs_create_flow_group_ingress(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
u32 *in,
- unsigned int *group_id)
+ struct mlx5_flow_group *fg)
{
- return fpga_ipsec_fs_create_flow_group(dev, ft, in, group_id, false);
+ return fpga_ipsec_fs_create_flow_group(ns, ft, in, fg, false);
}
static int
-mlx5_fpga_ipsec_fs_create_fte_ingress(struct mlx5_core_dev *dev,
+mlx5_fpga_ipsec_fs_create_fte_ingress(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
struct mlx5_flow_group *fg,
struct fs_fte *fte)
{
- return fpga_ipsec_fs_create_fte(dev, ft, fg, fte, false);
+ return fpga_ipsec_fs_create_fte(ns, ft, fg, fte, false);
}
static int
-mlx5_fpga_ipsec_fs_update_fte_ingress(struct mlx5_core_dev *dev,
+mlx5_fpga_ipsec_fs_update_fte_ingress(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
- unsigned int group_id,
+ struct mlx5_flow_group *fg,
int modify_mask,
struct fs_fte *fte)
{
- return fpga_ipsec_fs_update_fte(dev, ft, group_id, modify_mask, fte,
+ return fpga_ipsec_fs_update_fte(ns, ft, fg, modify_mask, fte,
false);
}
static int
-mlx5_fpga_ipsec_fs_delete_fte_ingress(struct mlx5_core_dev *dev,
+mlx5_fpga_ipsec_fs_delete_fte_ingress(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
struct fs_fte *fte)
{
- return fpga_ipsec_fs_delete_fte(dev, ft, fte, false);
+ return fpga_ipsec_fs_delete_fte(ns, ft, fte, false);
}
static struct mlx5_flow_cmds fpga_ipsec_ingress;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c
index c44ccb67c4a3..e89555da485e 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c
@@ -39,7 +39,7 @@
#include "mlx5_core.h"
#include "eswitch.h"
-static int mlx5_cmd_stub_update_root_ft(struct mlx5_core_dev *dev,
+static int mlx5_cmd_stub_update_root_ft(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
u32 underlay_qpn,
bool disconnect)
@@ -47,47 +47,43 @@ static int mlx5_cmd_stub_update_root_ft(struct mlx5_core_dev *dev,
return 0;
}
-static int mlx5_cmd_stub_create_flow_table(struct mlx5_core_dev *dev,
- u16 vport,
- enum fs_flow_table_op_mod op_mod,
- enum fs_flow_table_type type,
- unsigned int level,
+static int mlx5_cmd_stub_create_flow_table(struct mlx5_flow_root_namespace *ns,
+ struct mlx5_flow_table *ft,
unsigned int log_size,
- struct mlx5_flow_table *next_ft,
- unsigned int *table_id, u32 flags)
+ struct mlx5_flow_table *next_ft)
{
return 0;
}
-static int mlx5_cmd_stub_destroy_flow_table(struct mlx5_core_dev *dev,
+static int mlx5_cmd_stub_destroy_flow_table(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft)
{
return 0;
}
-static int mlx5_cmd_stub_modify_flow_table(struct mlx5_core_dev *dev,
+static int mlx5_cmd_stub_modify_flow_table(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
struct mlx5_flow_table *next_ft)
{
return 0;
}
-static int mlx5_cmd_stub_create_flow_group(struct mlx5_core_dev *dev,
+static int mlx5_cmd_stub_create_flow_group(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
u32 *in,
- unsigned int *group_id)
+ struct mlx5_flow_group *fg)
{
return 0;
}
-static int mlx5_cmd_stub_destroy_flow_group(struct mlx5_core_dev *dev,
+static int mlx5_cmd_stub_destroy_flow_group(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
- unsigned int group_id)
+ struct mlx5_flow_group *fg)
{
return 0;
}
-static int mlx5_cmd_stub_create_fte(struct mlx5_core_dev *dev,
+static int mlx5_cmd_stub_create_fte(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
struct mlx5_flow_group *group,
struct fs_fte *fte)
@@ -95,28 +91,29 @@ static int mlx5_cmd_stub_create_fte(struct mlx5_core_dev *dev,
return 0;
}
-static int mlx5_cmd_stub_update_fte(struct mlx5_core_dev *dev,
+static int mlx5_cmd_stub_update_fte(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
- unsigned int group_id,
+ struct mlx5_flow_group *group,
int modify_mask,
struct fs_fte *fte)
{
return -EOPNOTSUPP;
}
-static int mlx5_cmd_stub_delete_fte(struct mlx5_core_dev *dev,
+static int mlx5_cmd_stub_delete_fte(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
struct fs_fte *fte)
{
return 0;
}
-static int mlx5_cmd_update_root_ft(struct mlx5_core_dev *dev,
+static int mlx5_cmd_update_root_ft(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft, u32 underlay_qpn,
bool disconnect)
{
u32 in[MLX5_ST_SZ_DW(set_flow_table_root_in)] = {0};
u32 out[MLX5_ST_SZ_DW(set_flow_table_root_out)] = {0};
+ struct mlx5_core_dev *dev = ns->dev;
if ((MLX5_CAP_GEN(dev, port_type) == MLX5_CAP_PORT_TYPE_IB) &&
underlay_qpn == 0)
@@ -143,29 +140,26 @@ static int mlx5_cmd_update_root_ft(struct mlx5_core_dev *dev,
return mlx5_cmd_exec(dev, in, sizeof(in), out, sizeof(out));
}
-static int mlx5_cmd_create_flow_table(struct mlx5_core_dev *dev,
- u16 vport,
- enum fs_flow_table_op_mod op_mod,
- enum fs_flow_table_type type,
- unsigned int level,
+static int mlx5_cmd_create_flow_table(struct mlx5_flow_root_namespace *ns,
+ struct mlx5_flow_table *ft,
unsigned int log_size,
- struct mlx5_flow_table *next_ft,
- unsigned int *table_id, u32 flags)
+ struct mlx5_flow_table *next_ft)
{
- int en_encap = !!(flags & MLX5_FLOW_TABLE_TUNNEL_EN_REFORMAT);
- int en_decap = !!(flags & MLX5_FLOW_TABLE_TUNNEL_EN_DECAP);
+ int en_encap = !!(ft->flags & MLX5_FLOW_TABLE_TUNNEL_EN_REFORMAT);
+ int en_decap = !!(ft->flags & MLX5_FLOW_TABLE_TUNNEL_EN_DECAP);
u32 out[MLX5_ST_SZ_DW(create_flow_table_out)] = {0};
u32 in[MLX5_ST_SZ_DW(create_flow_table_in)] = {0};
+ struct mlx5_core_dev *dev = ns->dev;
int err;
MLX5_SET(create_flow_table_in, in, opcode,
MLX5_CMD_OP_CREATE_FLOW_TABLE);
- MLX5_SET(create_flow_table_in, in, table_type, type);
- MLX5_SET(create_flow_table_in, in, flow_table_context.level, level);
+ MLX5_SET(create_flow_table_in, in, table_type, ft->type);
+ MLX5_SET(create_flow_table_in, in, flow_table_context.level, ft->level);
MLX5_SET(create_flow_table_in, in, flow_table_context.log_size, log_size);
- if (vport) {
- MLX5_SET(create_flow_table_in, in, vport_number, vport);
+ if (ft->vport) {
+ MLX5_SET(create_flow_table_in, in, vport_number, ft->vport);
MLX5_SET(create_flow_table_in, in, other_vport, 1);
}
@@ -174,7 +168,7 @@ static int mlx5_cmd_create_flow_table(struct mlx5_core_dev *dev,
MLX5_SET(create_flow_table_in, in, flow_table_context.reformat_en,
en_encap);
- switch (op_mod) {
+ switch (ft->op_mod) {
case FS_FT_OP_MOD_NORMAL:
if (next_ft) {
MLX5_SET(create_flow_table_in, in,
@@ -195,16 +189,17 @@ static int mlx5_cmd_create_flow_table(struct mlx5_core_dev *dev,
err = mlx5_cmd_exec(dev, in, sizeof(in), out, sizeof(out));
if (!err)
- *table_id = MLX5_GET(create_flow_table_out, out,
- table_id);
+ ft->id = MLX5_GET(create_flow_table_out, out,
+ table_id);
return err;
}
-static int mlx5_cmd_destroy_flow_table(struct mlx5_core_dev *dev,
+static int mlx5_cmd_destroy_flow_table(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft)
{
u32 in[MLX5_ST_SZ_DW(destroy_flow_table_in)] = {0};
u32 out[MLX5_ST_SZ_DW(destroy_flow_table_out)] = {0};
+ struct mlx5_core_dev *dev = ns->dev;
MLX5_SET(destroy_flow_table_in, in, opcode,
MLX5_CMD_OP_DESTROY_FLOW_TABLE);
@@ -218,12 +213,13 @@ static int mlx5_cmd_destroy_flow_table(struct mlx5_core_dev *dev,
return mlx5_cmd_exec(dev, in, sizeof(in), out, sizeof(out));
}
-static int mlx5_cmd_modify_flow_table(struct mlx5_core_dev *dev,
+static int mlx5_cmd_modify_flow_table(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
struct mlx5_flow_table *next_ft)
{
u32 in[MLX5_ST_SZ_DW(modify_flow_table_in)] = {0};
u32 out[MLX5_ST_SZ_DW(modify_flow_table_out)] = {0};
+ struct mlx5_core_dev *dev = ns->dev;
MLX5_SET(modify_flow_table_in, in, opcode,
MLX5_CMD_OP_MODIFY_FLOW_TABLE);
@@ -263,13 +259,14 @@ static int mlx5_cmd_modify_flow_table(struct mlx5_core_dev *dev,
return mlx5_cmd_exec(dev, in, sizeof(in), out, sizeof(out));
}
-static int mlx5_cmd_create_flow_group(struct mlx5_core_dev *dev,
+static int mlx5_cmd_create_flow_group(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
u32 *in,
- unsigned int *group_id)
+ struct mlx5_flow_group *fg)
{
u32 out[MLX5_ST_SZ_DW(create_flow_group_out)] = {0};
int inlen = MLX5_ST_SZ_BYTES(create_flow_group_in);
+ struct mlx5_core_dev *dev = ns->dev;
int err;
MLX5_SET(create_flow_group_in, in, opcode,
@@ -283,23 +280,24 @@ static int mlx5_cmd_create_flow_group(struct mlx5_core_dev *dev,
err = mlx5_cmd_exec(dev, in, inlen, out, sizeof(out));
if (!err)
- *group_id = MLX5_GET(create_flow_group_out, out,
- group_id);
+ fg->id = MLX5_GET(create_flow_group_out, out,
+ group_id);
return err;
}
-static int mlx5_cmd_destroy_flow_group(struct mlx5_core_dev *dev,
+static int mlx5_cmd_destroy_flow_group(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
- unsigned int group_id)
+ struct mlx5_flow_group *fg)
{
u32 out[MLX5_ST_SZ_DW(destroy_flow_group_out)] = {0};
u32 in[MLX5_ST_SZ_DW(destroy_flow_group_in)] = {0};
+ struct mlx5_core_dev *dev = ns->dev;
MLX5_SET(destroy_flow_group_in, in, opcode,
MLX5_CMD_OP_DESTROY_FLOW_GROUP);
MLX5_SET(destroy_flow_group_in, in, table_type, ft->type);
MLX5_SET(destroy_flow_group_in, in, table_id, ft->id);
- MLX5_SET(destroy_flow_group_in, in, group_id, group_id);
+ MLX5_SET(destroy_flow_group_in, in, group_id, fg->id);
if (ft->vport) {
MLX5_SET(destroy_flow_group_in, in, vport_number, ft->vport);
MLX5_SET(destroy_flow_group_in, in, other_vport, 1);
@@ -505,23 +503,25 @@ static int mlx5_cmd_set_fte(struct mlx5_core_dev *dev,
return err;
}
-static int mlx5_cmd_create_fte(struct mlx5_core_dev *dev,
+static int mlx5_cmd_create_fte(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
struct mlx5_flow_group *group,
struct fs_fte *fte)
{
+ struct mlx5_core_dev *dev = ns->dev;
unsigned int group_id = group->id;
return mlx5_cmd_set_fte(dev, 0, 0, ft, group_id, fte);
}
-static int mlx5_cmd_update_fte(struct mlx5_core_dev *dev,
+static int mlx5_cmd_update_fte(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
- unsigned int group_id,
+ struct mlx5_flow_group *fg,
int modify_mask,
struct fs_fte *fte)
{
int opmod;
+ struct mlx5_core_dev *dev = ns->dev;
int atomic_mod_cap = MLX5_CAP_FLOWTABLE(dev,
flow_table_properties_nic_receive.
flow_modify_en);
@@ -529,15 +529,16 @@ static int mlx5_cmd_update_fte(struct mlx5_core_dev *dev,
return -EOPNOTSUPP;
opmod = 1;
- return mlx5_cmd_set_fte(dev, opmod, modify_mask, ft, group_id, fte);
+ return mlx5_cmd_set_fte(dev, opmod, modify_mask, ft, fg->id, fte);
}
-static int mlx5_cmd_delete_fte(struct mlx5_core_dev *dev,
+static int mlx5_cmd_delete_fte(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
struct fs_fte *fte)
{
u32 out[MLX5_ST_SZ_DW(delete_fte_out)] = {0};
u32 in[MLX5_ST_SZ_DW(delete_fte_in)] = {0};
+ struct mlx5_core_dev *dev = ns->dev;
MLX5_SET(delete_fte_in, in, opcode, MLX5_CMD_OP_DELETE_FLOW_TABLE_ENTRY);
MLX5_SET(delete_fte_in, in, table_type, ft->type);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.h b/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.h
index 6228ba7bfa1a..e340f9af2f5a 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.h
@@ -36,45 +36,42 @@
#include "fs_core.h"
struct mlx5_flow_cmds {
- int (*create_flow_table)(struct mlx5_core_dev *dev,
- u16 vport,
- enum fs_flow_table_op_mod op_mod,
- enum fs_flow_table_type type,
- unsigned int level, unsigned int log_size,
- struct mlx5_flow_table *next_ft,
- unsigned int *table_id, u32 flags);
- int (*destroy_flow_table)(struct mlx5_core_dev *dev,
+ int (*create_flow_table)(struct mlx5_flow_root_namespace *ns,
+ struct mlx5_flow_table *ft,
+ unsigned int log_size,
+ struct mlx5_flow_table *next_ft);
+ int (*destroy_flow_table)(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft);
- int (*modify_flow_table)(struct mlx5_core_dev *dev,
+ int (*modify_flow_table)(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
struct mlx5_flow_table *next_ft);
- int (*create_flow_group)(struct mlx5_core_dev *dev,
+ int (*create_flow_group)(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
u32 *in,
- unsigned int *group_id);
+ struct mlx5_flow_group *fg);
- int (*destroy_flow_group)(struct mlx5_core_dev *dev,
+ int (*destroy_flow_group)(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
- unsigned int group_id);
+ struct mlx5_flow_group *fg);
- int (*create_fte)(struct mlx5_core_dev *dev,
+ int (*create_fte)(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
struct mlx5_flow_group *fg,
struct fs_fte *fte);
- int (*update_fte)(struct mlx5_core_dev *dev,
+ int (*update_fte)(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
- unsigned int group_id,
+ struct mlx5_flow_group *fg,
int modify_mask,
struct fs_fte *fte);
- int (*delete_fte)(struct mlx5_core_dev *dev,
+ int (*delete_fte)(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
struct fs_fte *fte);
- int (*update_root_ft)(struct mlx5_core_dev *dev,
+ int (*update_root_ft)(struct mlx5_flow_root_namespace *ns,
struct mlx5_flow_table *ft,
u32 underlay_qpn,
bool disconnect);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
index 9fcef7e3b86d..6024df821734 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
@@ -403,7 +403,7 @@ static void del_hw_flow_table(struct fs_node *node)
trace_mlx5_fs_del_ft(ft);
if (node->active) {
- err = root->cmds->destroy_flow_table(dev, ft);
+ err = root->cmds->destroy_flow_table(root, ft);
if (err)
mlx5_core_warn(dev, "flow steering can't destroy ft\n");
}
@@ -435,7 +435,7 @@ static void modify_fte(struct fs_fte *fte)
dev = get_dev(&fte->node);
root = find_root(&ft->node);
- err = root->cmds->update_fte(dev, ft, fg->id, fte->modify_mask, fte);
+ err = root->cmds->update_fte(root, ft, fg, fte->modify_mask, fte);
if (err)
mlx5_core_warn(dev,
"%s can't del rule fg id=%d fte_index=%d\n",
@@ -492,7 +492,7 @@ static void del_hw_fte(struct fs_node *node)
dev = get_dev(&ft->node);
root = find_root(&ft->node);
if (node->active) {
- err = root->cmds->delete_fte(dev, ft, fte);
+ err = root->cmds->delete_fte(root, ft, fte);
if (err)
mlx5_core_warn(dev,
"flow steering can't delete fte in index %d of flow group id %d\n",
@@ -532,7 +532,7 @@ static void del_hw_flow_group(struct fs_node *node)
trace_mlx5_fs_del_fg(fg);
root = find_root(&ft->node);
- if (fg->node.active && root->cmds->destroy_flow_group(dev, ft, fg->id))
+ if (fg->node.active && root->cmds->destroy_flow_group(root, ft, fg))
mlx5_core_warn(dev, "flow steering can't destroy fg %d of ft %d\n",
fg->id, ft->id);
}
@@ -783,7 +783,7 @@ static int connect_fts_in_prio(struct mlx5_core_dev *dev,
fs_for_each_ft(iter, prio) {
i++;
- err = root->cmds->modify_flow_table(dev, iter, ft);
+ err = root->cmds->modify_flow_table(root, iter, ft);
if (err) {
mlx5_core_warn(dev, "Failed to modify flow table %d\n",
iter->id);
@@ -831,11 +831,11 @@ static int update_root_ft_create(struct mlx5_flow_table *ft, struct fs_prio
if (list_empty(&root->underlay_qpns)) {
/* Don't set any QPN (zero) in case QPN list is empty */
qpn = 0;
- err = root->cmds->update_root_ft(root->dev, ft, qpn, false);
+ err = root->cmds->update_root_ft(root, ft, qpn, false);
} else {
list_for_each_entry(uqp, &root->underlay_qpns, list) {
qpn = uqp->qpn;
- err = root->cmds->update_root_ft(root->dev, ft,
+ err = root->cmds->update_root_ft(root, ft,
qpn, false);
if (err)
break;
@@ -871,7 +871,7 @@ static int _mlx5_modify_rule_destination(struct mlx5_flow_rule *rule,
memcpy(&rule->dest_attr, dest, sizeof(*dest));
root = find_root(&ft->node);
- err = root->cmds->update_fte(get_dev(&ft->node), ft, fg->id,
+ err = root->cmds->update_fte(root, ft, fg,
modify_mask, fte);
up_write_ref_node(&fte->node, false);
@@ -1013,9 +1013,7 @@ static struct mlx5_flow_table *__mlx5_create_flow_table(struct mlx5_flow_namespa
tree_init_node(&ft->node, del_hw_flow_table, del_sw_flow_table);
log_table_sz = ft->max_fte ? ilog2(ft->max_fte) : 0;
next_ft = find_next_chained_ft(fs_prio);
- err = root->cmds->create_flow_table(root->dev, ft->vport, ft->op_mod,
- ft->type, ft->level, log_table_sz,
- next_ft, &ft->id, ft->flags);
+ err = root->cmds->create_flow_table(root, ft, log_table_sz, next_ft);
if (err)
goto free_ft;
@@ -1032,7 +1030,7 @@ static struct mlx5_flow_table *__mlx5_create_flow_table(struct mlx5_flow_namespa
trace_mlx5_fs_add_ft(ft);
return ft;
destroy_ft:
- root->cmds->destroy_flow_table(root->dev, ft);
+ root->cmds->destroy_flow_table(root, ft);
free_ft:
kfree(ft);
unlock_root:
@@ -1114,7 +1112,6 @@ struct mlx5_flow_group *mlx5_create_flow_group(struct mlx5_flow_table *ft,
start_flow_index);
int end_index = MLX5_GET(create_flow_group_in, fg_in,
end_flow_index);
- struct mlx5_core_dev *dev = get_dev(&ft->node);
struct mlx5_flow_group *fg;
int err;
@@ -1129,7 +1126,7 @@ struct mlx5_flow_group *mlx5_create_flow_group(struct mlx5_flow_table *ft,
if (IS_ERR(fg))
return fg;
- err = root->cmds->create_flow_group(dev, ft, fg_in, &fg->id);
+ err = root->cmds->create_flow_group(root, ft, fg_in, fg);
if (err) {
tree_put_node(&fg->node, false);
return ERR_PTR(err);
@@ -1269,11 +1266,9 @@ add_rule_fte(struct fs_fte *fte,
fs_get_obj(ft, fg->node.parent);
root = find_root(&fg->node);
if (!(fte->status & FS_FTE_STATUS_EXISTING))
- err = root->cmds->create_fte(get_dev(&ft->node),
- ft, fg, fte);
+ err = root->cmds->create_fte(root, ft, fg, fte);
else
- err = root->cmds->update_fte(get_dev(&ft->node), ft, fg->id,
- modify_mask, fte);
+ err = root->cmds->update_fte(root, ft, fg, modify_mask, fte);
if (err)
goto free_handle;
@@ -1339,7 +1334,6 @@ static int create_auto_flow_group(struct mlx5_flow_table *ft,
struct mlx5_flow_group *fg)
{
struct mlx5_flow_root_namespace *root = find_root(&ft->node);
- struct mlx5_core_dev *dev = get_dev(&ft->node);
int inlen = MLX5_ST_SZ_BYTES(create_flow_group_in);
void *match_criteria_addr;
u8 src_esw_owner_mask_on;
@@ -1369,7 +1363,7 @@ static int create_auto_flow_group(struct mlx5_flow_table *ft,
memcpy(match_criteria_addr, fg->mask.match_criteria,
sizeof(fg->mask.match_criteria));
- err = root->cmds->create_flow_group(dev, ft, in, &fg->id);
+ err = root->cmds->create_flow_group(root, ft, in, fg);
if (!err) {
fg->node.active = true;
trace_mlx5_fs_add_fg(fg);
@@ -1941,12 +1935,12 @@ static int update_root_ft_destroy(struct mlx5_flow_table *ft)
if (list_empty(&root->underlay_qpns)) {
/* Don't set any QPN (zero) in case QPN list is empty */
qpn = 0;
- err = root->cmds->update_root_ft(root->dev, new_root_ft,
+ err = root->cmds->update_root_ft(root, new_root_ft,
qpn, false);
} else {
list_for_each_entry(uqp, &root->underlay_qpns, list) {
qpn = uqp->qpn;
- err = root->cmds->update_root_ft(root->dev,
+ err = root->cmds->update_root_ft(root,
new_root_ft, qpn,
false);
if (err)
@@ -2762,7 +2756,7 @@ int mlx5_fs_add_rx_underlay_qpn(struct mlx5_core_dev *dev, u32 underlay_qpn)
goto update_ft_fail;
}
- err = root->cmds->update_root_ft(dev, root->root_ft, underlay_qpn,
+ err = root->cmds->update_root_ft(root, root->root_ft, underlay_qpn,
false);
if (err) {
mlx5_core_warn(dev, "Failed adding underlay QPN (%u) to root FT err(%d)\n",
@@ -2806,7 +2800,7 @@ int mlx5_fs_remove_rx_underlay_qpn(struct mlx5_core_dev *dev, u32 underlay_qpn)
goto out;
}
- err = root->cmds->update_root_ft(dev, root->root_ft, underlay_qpn,
+ err = root->cmds->update_root_ft(root, root->root_ft, underlay_qpn,
true);
if (err)
mlx5_core_warn(dev, "Failed removing underlay QPN (%u) from root FT err(%d)\n",
--
2.20.1
^ permalink raw reply related [flat|nested] 17+ messages in thread
* [PATCH V2 mlx5-next 07/11] net/mlx5: Add support in RDMA RX steering
2019-04-29 18:13 [PATCH V2 mlx5-next 00/11] Mellanox, mlx5-next updates 2019-04-25 Saeed Mahameed
` (5 preceding siblings ...)
2019-04-29 18:14 ` [PATCH V2 mlx5-next 06/11] net/mlx5: Pass flow steering objects to fs_cmd Saeed Mahameed
@ 2019-04-29 18:14 ` Saeed Mahameed
2019-04-29 18:14 ` [PATCH V2 mlx5-next 08/11] net/mlx5: Add new miss flow table action Saeed Mahameed
` (4 subsequent siblings)
11 siblings, 0 replies; 17+ messages in thread
From: Saeed Mahameed @ 2019-04-29 18:14 UTC (permalink / raw)
To: Saeed Mahameed, Leon Romanovsky
Cc: Jason Gunthorpe, netdev, linux-rdma, Maor Gottlieb, Mark Bloch
From: Maor Gottlieb <maorg@mellanox.com>
Add new flow steering namespace - MLX5_FLOW_NAMESPACE_RDMA_RX.
Flow steering rules in this namespace are used to filter
RDMA traffic.
Signed-off-by: Maor Gottlieb <maorg@mellanox.com>
Reviewed-by: Mark Bloch <markb@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
.../net/ethernet/mellanox/mlx5/core/fs_cmd.c | 1 +
.../net/ethernet/mellanox/mlx5/core/fs_core.c | 27 +++++++++++++++++++
.../net/ethernet/mellanox/mlx5/core/fs_core.h | 2 ++
include/linux/mlx5/device.h | 6 +++++
include/linux/mlx5/fs.h | 1 +
include/linux/mlx5/mlx5_ifc.h | 2 +-
6 files changed, 38 insertions(+), 1 deletion(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c
index e89555da485e..629c5ab1c0d1 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c
@@ -854,6 +854,7 @@ const struct mlx5_flow_cmds *mlx5_fs_cmd_get_default(enum fs_flow_table_type typ
case FS_FT_SNIFFER_RX:
case FS_FT_SNIFFER_TX:
case FS_FT_NIC_TX:
+ case FS_FT_RDMA_RX:
return mlx5_fs_cmd_get_fw_cmds();
default:
return mlx5_fs_cmd_get_stub_cmds();
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
index 6024df821734..3c2302a2b9d4 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
@@ -2054,6 +2054,10 @@ struct mlx5_flow_namespace *mlx5_get_flow_namespace(struct mlx5_core_dev *dev,
if (steering->sniffer_tx_root_ns)
return &steering->sniffer_tx_root_ns->ns;
return NULL;
+ case MLX5_FLOW_NAMESPACE_RDMA_RX:
+ if (steering->rdma_rx_root_ns)
+ return &steering->rdma_rx_root_ns->ns;
+ return NULL;
default:
break;
}
@@ -2450,6 +2454,7 @@ void mlx5_cleanup_fs(struct mlx5_core_dev *dev)
steering->fdb_sub_ns = NULL;
cleanup_root_ns(steering->sniffer_rx_root_ns);
cleanup_root_ns(steering->sniffer_tx_root_ns);
+ cleanup_root_ns(steering->rdma_rx_root_ns);
cleanup_root_ns(steering->egress_root_ns);
mlx5_cleanup_fc_stats(dev);
kmem_cache_destroy(steering->ftes_cache);
@@ -2491,6 +2496,22 @@ static int init_sniffer_rx_root_ns(struct mlx5_flow_steering *steering)
return 0;
}
+static int init_rdma_rx_root_ns(struct mlx5_flow_steering *steering)
+{
+ struct fs_prio *prio;
+
+ steering->rdma_rx_root_ns = create_root_ns(steering, FS_FT_RDMA_RX);
+ if (!steering->rdma_rx_root_ns)
+ return -ENOMEM;
+
+ /* Create single prio */
+ prio = fs_create_prio(&steering->rdma_rx_root_ns->ns, 0, 1);
+ if (IS_ERR(prio)) {
+ cleanup_root_ns(steering->rdma_rx_root_ns);
+ return PTR_ERR(prio);
+ }
+ return 0;
+}
static int init_fdb_root_ns(struct mlx5_flow_steering *steering)
{
struct mlx5_flow_namespace *ns;
@@ -2727,6 +2748,12 @@ int mlx5_init_fs(struct mlx5_core_dev *dev)
goto err;
}
+ if (MLX5_CAP_FLOWTABLE_RDMA_RX(dev, ft_support)) {
+ err = init_rdma_rx_root_ns(steering);
+ if (err)
+ goto err;
+ }
+
if (MLX5_IPSEC_DEV(dev) || MLX5_CAP_FLOWTABLE_NIC_TX(dev, ft_support)) {
err = init_egress_root_ns(steering);
if (err)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h
index 87de0e4d9124..e43c6f6d46a7 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h
@@ -67,6 +67,7 @@ enum fs_flow_table_type {
FS_FT_FDB = 0X4,
FS_FT_SNIFFER_RX = 0X5,
FS_FT_SNIFFER_TX = 0X6,
+ FS_FT_RDMA_RX = 0X7,
FS_FT_MAX_TYPE = FS_FT_SNIFFER_TX,
};
@@ -90,6 +91,7 @@ struct mlx5_flow_steering {
struct mlx5_flow_root_namespace **esw_ingress_root_ns;
struct mlx5_flow_root_namespace *sniffer_tx_root_ns;
struct mlx5_flow_root_namespace *sniffer_rx_root_ns;
+ struct mlx5_flow_root_namespace *rdma_rx_root_ns;
struct mlx5_flow_root_namespace *egress_root_ns;
};
diff --git a/include/linux/mlx5/device.h b/include/linux/mlx5/device.h
index f93a5598b942..28ebb6c93542 100644
--- a/include/linux/mlx5/device.h
+++ b/include/linux/mlx5/device.h
@@ -1166,6 +1166,12 @@ enum mlx5_qcam_feature_groups {
#define MLX5_CAP_FLOWTABLE_SNIFFER_TX_MAX(mdev, cap) \
MLX5_CAP_FLOWTABLE_MAX(mdev, flow_table_properties_nic_transmit_sniffer.cap)
+#define MLX5_CAP_FLOWTABLE_RDMA_RX(mdev, cap) \
+ MLX5_CAP_FLOWTABLE(mdev, flow_table_properties_nic_receive_rdma.cap)
+
+#define MLX5_CAP_FLOWTABLE_RDMA_RX_MAX(mdev, cap) \
+ MLX5_CAP_FLOWTABLE_MAX(mdev, flow_table_properties_nic_receive_rdma.cap)
+
#define MLX5_CAP_ESW_FLOWTABLE(mdev, cap) \
MLX5_GET(flow_table_eswitch_cap, \
mdev->caps.hca_cur[MLX5_CAP_ESWITCH_FLOW_TABLE], cap)
diff --git a/include/linux/mlx5/fs.h b/include/linux/mlx5/fs.h
index fd91df3a4e09..e690ba0f965c 100644
--- a/include/linux/mlx5/fs.h
+++ b/include/linux/mlx5/fs.h
@@ -73,6 +73,7 @@ enum mlx5_flow_namespace_type {
MLX5_FLOW_NAMESPACE_SNIFFER_RX,
MLX5_FLOW_NAMESPACE_SNIFFER_TX,
MLX5_FLOW_NAMESPACE_EGRESS,
+ MLX5_FLOW_NAMESPACE_RDMA_RX,
};
enum {
diff --git a/include/linux/mlx5/mlx5_ifc.h b/include/linux/mlx5/mlx5_ifc.h
index eeedf3f53ed3..89e7194b3d97 100644
--- a/include/linux/mlx5/mlx5_ifc.h
+++ b/include/linux/mlx5/mlx5_ifc.h
@@ -598,7 +598,7 @@ struct mlx5_ifc_flow_table_nic_cap_bits {
struct mlx5_ifc_flow_table_prop_layout_bits flow_table_properties_nic_receive;
- u8 reserved_at_400[0x200];
+ struct mlx5_ifc_flow_table_prop_layout_bits flow_table_properties_nic_receive_rdma;
struct mlx5_ifc_flow_table_prop_layout_bits flow_table_properties_nic_receive_sniffer;
--
2.20.1
^ permalink raw reply related [flat|nested] 17+ messages in thread
* [PATCH V2 mlx5-next 08/11] net/mlx5: Add new miss flow table action
2019-04-29 18:13 [PATCH V2 mlx5-next 00/11] Mellanox, mlx5-next updates 2019-04-25 Saeed Mahameed
` (6 preceding siblings ...)
2019-04-29 18:14 ` [PATCH V2 mlx5-next 07/11] net/mlx5: Add support in RDMA RX steering Saeed Mahameed
@ 2019-04-29 18:14 ` Saeed Mahameed
2019-04-29 18:14 ` [PATCH V2 mlx5-next 09/11] net/mlx5: Eswitch, enable RoCE loopback traffic Saeed Mahameed
` (3 subsequent siblings)
11 siblings, 0 replies; 17+ messages in thread
From: Saeed Mahameed @ 2019-04-29 18:14 UTC (permalink / raw)
To: Saeed Mahameed, Leon Romanovsky
Cc: Jason Gunthorpe, netdev, linux-rdma, Maor Gottlieb, Mark Bloch
From: Maor Gottlieb <maorg@mellanox.com>
Flow table supports three types of miss action:
1. Default miss action - go to default miss table according to table.
2. Go to specific table.
3. Switch domain - go to the root table of an alternative steering
table domain.
New table miss action was added - switch_domain.
The next domain for RDMA_RX namespace is the NIC RX domain.
Signed-off-by: Maor Gottlieb <maorg@mellanox.com>
Reviewed-by: Mark Bloch <markb@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c | 13 ++++++++++---
drivers/net/ethernet/mellanox/mlx5/core/fs_core.c | 6 +++++-
drivers/net/ethernet/mellanox/mlx5/core/fs_core.h | 1 +
include/linux/mlx5/mlx5_ifc.h | 10 +++++++++-
4 files changed, 25 insertions(+), 5 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c
index 629c5ab1c0d1..013b1ca4a791 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c
@@ -172,9 +172,14 @@ static int mlx5_cmd_create_flow_table(struct mlx5_flow_root_namespace *ns,
case FS_FT_OP_MOD_NORMAL:
if (next_ft) {
MLX5_SET(create_flow_table_in, in,
- flow_table_context.table_miss_action, 1);
+ flow_table_context.table_miss_action,
+ MLX5_FLOW_TABLE_MISS_ACTION_FWD);
MLX5_SET(create_flow_table_in, in,
flow_table_context.table_miss_id, next_ft->id);
+ } else {
+ MLX5_SET(create_flow_table_in, in,
+ flow_table_context.table_miss_action,
+ ns->def_miss_action);
}
break;
@@ -246,13 +251,15 @@ static int mlx5_cmd_modify_flow_table(struct mlx5_flow_root_namespace *ns,
MLX5_MODIFY_FLOW_TABLE_MISS_TABLE_ID);
if (next_ft) {
MLX5_SET(modify_flow_table_in, in,
- flow_table_context.table_miss_action, 1);
+ flow_table_context.table_miss_action,
+ MLX5_FLOW_TABLE_MISS_ACTION_FWD);
MLX5_SET(modify_flow_table_in, in,
flow_table_context.table_miss_id,
next_ft->id);
} else {
MLX5_SET(modify_flow_table_in, in,
- flow_table_context.table_miss_action, 0);
+ flow_table_context.table_miss_action,
+ ns->def_miss_action);
}
}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
index 3c2302a2b9d4..fb5b61727ee7 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
@@ -2504,6 +2504,9 @@ static int init_rdma_rx_root_ns(struct mlx5_flow_steering *steering)
if (!steering->rdma_rx_root_ns)
return -ENOMEM;
+ steering->rdma_rx_root_ns->def_miss_action =
+ MLX5_FLOW_TABLE_MISS_ACTION_SWITCH_DOMAIN;
+
/* Create single prio */
prio = fs_create_prio(&steering->rdma_rx_root_ns->ns, 0, 1);
if (IS_ERR(prio)) {
@@ -2748,7 +2751,8 @@ int mlx5_init_fs(struct mlx5_core_dev *dev)
goto err;
}
- if (MLX5_CAP_FLOWTABLE_RDMA_RX(dev, ft_support)) {
+ if (MLX5_CAP_FLOWTABLE_RDMA_RX(dev, ft_support) &&
+ MLX5_CAP_FLOWTABLE_RDMA_RX(dev, table_miss_action_domain)) {
err = init_rdma_rx_root_ns(steering);
if (err)
goto err;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h
index e43c6f6d46a7..0c6c5fef4548 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h
@@ -218,6 +218,7 @@ struct mlx5_flow_root_namespace {
struct mutex chain_lock;
struct list_head underlay_qpns;
const struct mlx5_flow_cmds *cmds;
+ enum mlx5_flow_table_miss_action def_miss_action;
};
int mlx5_init_fc_stats(struct mlx5_core_dev *dev);
diff --git a/include/linux/mlx5/mlx5_ifc.h b/include/linux/mlx5/mlx5_ifc.h
index 89e7194b3d97..7d9264b282d1 100644
--- a/include/linux/mlx5/mlx5_ifc.h
+++ b/include/linux/mlx5/mlx5_ifc.h
@@ -370,7 +370,9 @@ struct mlx5_ifc_flow_table_prop_layout_bits {
u8 reformat_l3_tunnel_to_l2[0x1];
u8 reformat_l2_to_l3_tunnel[0x1];
u8 reformat_and_modify_action[0x1];
- u8 reserved_at_15[0xb];
+ u8 reserved_at_15[0x2];
+ u8 table_miss_action_domain[0x1];
+ u8 reserved_at_18[0x8];
u8 reserved_at_20[0x2];
u8 log_max_ft_size[0x6];
u8 log_max_modify_header_context[0x8];
@@ -1284,6 +1286,12 @@ enum mlx5_flow_destination_type {
MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE_NUM = 0x101,
};
+enum mlx5_flow_table_miss_action {
+ MLX5_FLOW_TABLE_MISS_ACTION_DEF,
+ MLX5_FLOW_TABLE_MISS_ACTION_FWD,
+ MLX5_FLOW_TABLE_MISS_ACTION_SWITCH_DOMAIN,
+};
+
struct mlx5_ifc_dest_format_struct_bits {
u8 destination_type[0x8];
u8 destination_id[0x18];
--
2.20.1
^ permalink raw reply related [flat|nested] 17+ messages in thread
* [PATCH V2 mlx5-next 09/11] net/mlx5: Eswitch, enable RoCE loopback traffic
2019-04-29 18:13 [PATCH V2 mlx5-next 00/11] Mellanox, mlx5-next updates 2019-04-25 Saeed Mahameed
` (7 preceding siblings ...)
2019-04-29 18:14 ` [PATCH V2 mlx5-next 08/11] net/mlx5: Add new miss flow table action Saeed Mahameed
@ 2019-04-29 18:14 ` Saeed Mahameed
2019-04-29 18:41 ` Leon Romanovsky
2019-04-29 18:14 ` [PATCH V2 mlx5-next 10/11] net/mlx5: Geneve, Add basic Geneve encap/decap flow table capabilities Saeed Mahameed
` (2 subsequent siblings)
11 siblings, 1 reply; 17+ messages in thread
From: Saeed Mahameed @ 2019-04-29 18:14 UTC (permalink / raw)
To: Saeed Mahameed, Leon Romanovsky
Cc: Jason Gunthorpe, netdev, linux-rdma, Maor Gottlieb, Mark Bloch
From: Maor Gottlieb <maorg@mellanox.com>
When in switchdev mode, we would like to treat loopback RoCE
traffic (on eswitch manager) as RDMA and not as regular
Ethernet traffic
In order to enable it we add flow steering rule that forward RoCE
loopback traffic to the HW RoCE filter (by adding allow rule).
In addition we add RoCE address in GID index 0, which will be
set in the RoCE loopback packet.
Signed-off-by: Maor Gottlieb <maorg@mellanox.com>
Reviewed-by: Mark Bloch <markb@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
.../net/ethernet/mellanox/mlx5/core/Makefile | 2 +-
.../mellanox/mlx5/core/eswitch_offloads.c | 4 +
.../net/ethernet/mellanox/mlx5/core/rdma.c | 182 ++++++++++++++++++
.../net/ethernet/mellanox/mlx5/core/rdma.h | 20 ++
include/linux/mlx5/driver.h | 7 +
5 files changed, 214 insertions(+), 1 deletion(-)
create mode 100644 drivers/net/ethernet/mellanox/mlx5/core/rdma.c
create mode 100644 drivers/net/ethernet/mellanox/mlx5/core/rdma.h
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/Makefile b/drivers/net/ethernet/mellanox/mlx5/core/Makefile
index 1a16f6d73cbc..5f0be9b36a04 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/Makefile
+++ b/drivers/net/ethernet/mellanox/mlx5/core/Makefile
@@ -35,7 +35,7 @@ mlx5_core-$(CONFIG_MLX5_ESWITCH) += en_rep.o en_tc.o en/tc_tun.o lib/port_tu
#
# Core extra
#
-mlx5_core-$(CONFIG_MLX5_ESWITCH) += eswitch.o eswitch_offloads.o ecpf.o
+mlx5_core-$(CONFIG_MLX5_ESWITCH) += eswitch.o eswitch_offloads.o ecpf.o rdma.o
mlx5_core-$(CONFIG_MLX5_MPFS) += lib/mpfs.o
mlx5_core-$(CONFIG_VXLAN) += lib/vxlan.o
mlx5_core-$(CONFIG_PTP_1588_CLOCK) += lib/clock.o
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
index 6c8a17ca236e..9a1598bca4d1 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
@@ -37,6 +37,7 @@
#include <linux/mlx5/fs.h>
#include "mlx5_core.h"
#include "eswitch.h"
+#include "rdma.h"
#include "en.h"
#include "fs_core.h"
#include "lib/devcom.h"
@@ -1713,6 +1714,8 @@ int esw_offloads_init(struct mlx5_eswitch *esw, int vf_nvports,
esw->host_info.num_vfs = vf_nvports;
}
+ mlx5_rdma_enable_roce(esw->dev);
+
return 0;
err_reps:
@@ -1751,6 +1754,7 @@ void esw_offloads_cleanup(struct mlx5_eswitch *esw)
num_vfs = esw->dev->priv.sriov.num_vfs;
}
+ mlx5_rdma_disable_roce(esw->dev);
esw_offloads_devcom_cleanup(esw);
esw_offloads_unload_all_reps(esw, num_vfs);
esw_offloads_steering_cleanup(esw);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/rdma.c b/drivers/net/ethernet/mellanox/mlx5/core/rdma.c
new file mode 100644
index 000000000000..86f77456f873
--- /dev/null
+++ b/drivers/net/ethernet/mellanox/mlx5/core/rdma.c
@@ -0,0 +1,182 @@
+// SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB
+/* Copyright (c) 2019 Mellanox Technologies */
+
+#include <linux/mlx5/vport.h>
+#include <rdma/ib_verbs.h>
+#include <net/addrconf.h>
+
+#include "lib/mlx5.h"
+#include "eswitch.h"
+#include "fs_core.h"
+#include "rdma.h"
+
+static void mlx5_rdma_disable_roce_steering(struct mlx5_core_dev *dev)
+{
+ struct mlx5_core_roce *roce = &dev->priv.roce;
+
+ if (!roce->ft)
+ return;
+
+ mlx5_del_flow_rules(roce->allow_rule);
+ mlx5_destroy_flow_group(roce->fg);
+ mlx5_destroy_flow_table(roce->ft);
+}
+
+static int mlx5_rdma_enable_roce_steering(struct mlx5_core_dev *dev)
+{
+ int inlen = MLX5_ST_SZ_BYTES(create_flow_group_in);
+ struct mlx5_core_roce *roce = &dev->priv.roce;
+ struct mlx5_flow_handle *flow_rule = NULL;
+ struct mlx5_flow_table_attr ft_attr = {};
+ struct mlx5_flow_namespace *ns = NULL;
+ struct mlx5_flow_act flow_act = {};
+ struct mlx5_flow_spec *spec;
+ struct mlx5_flow_table *ft;
+ struct mlx5_flow_group *fg;
+ void *match_criteria;
+ u32 *flow_group_in;
+ void *misc;
+ int err;
+
+ if (!(MLX5_CAP_FLOWTABLE_RDMA_RX(dev, ft_support) &&
+ MLX5_CAP_FLOWTABLE_RDMA_RX(dev, table_miss_action_domain)))
+ return -EOPNOTSUPP;
+
+ flow_group_in = kvzalloc(inlen, GFP_KERNEL);
+ if (!flow_group_in)
+ return -ENOMEM;
+ spec = kvzalloc(sizeof(*spec), GFP_KERNEL);
+ if (!spec) {
+ kvfree(flow_group_in);
+ return -ENOMEM;
+ }
+
+ ns = mlx5_get_flow_namespace(dev, MLX5_FLOW_NAMESPACE_RDMA_RX);
+ if (!ns) {
+ mlx5_core_err(dev, "Failed to get RDMA RX namespace");
+ err = -EOPNOTSUPP;
+ goto free;
+ }
+
+ ft_attr.max_fte = 1;
+ ft = mlx5_create_flow_table(ns, &ft_attr);
+ if (IS_ERR(ft)) {
+ mlx5_core_err(dev, "Failed to create RDMA RX flow table");
+ err = PTR_ERR(ft);
+ goto free;
+ }
+
+ MLX5_SET(create_flow_group_in, flow_group_in, match_criteria_enable,
+ MLX5_MATCH_MISC_PARAMETERS);
+ match_criteria = MLX5_ADDR_OF(create_flow_group_in, flow_group_in,
+ match_criteria);
+ MLX5_SET_TO_ONES(fte_match_param, match_criteria,
+ misc_parameters.source_port);
+
+ fg = mlx5_create_flow_group(ft, flow_group_in);
+ if (IS_ERR(fg)) {
+ err = PTR_ERR(fg);
+ mlx5_core_err(dev, "Failed to create RDMA RX flow group err(%d)\n", err);
+ goto destroy_flow_table;
+ }
+
+ spec->match_criteria_enable = MLX5_MATCH_MISC_PARAMETERS;
+ misc = MLX5_ADDR_OF(fte_match_param, spec->match_value,
+ misc_parameters);
+ MLX5_SET(fte_match_set_misc, misc, source_port,
+ dev->priv.eswitch->manager_vport);
+ misc = MLX5_ADDR_OF(fte_match_param, spec->match_criteria,
+ misc_parameters);
+ MLX5_SET_TO_ONES(fte_match_set_misc, misc, source_port);
+
+ flow_act.action = MLX5_FLOW_CONTEXT_ACTION_ALLOW;
+ flow_rule = mlx5_add_flow_rules(ft, spec, &flow_act, NULL, 0);
+ if (IS_ERR(flow_rule)) {
+ err = PTR_ERR(flow_rule);
+ mlx5_core_err(dev, "Failed to add RoCE allow rule, err=%d\n",
+ err);
+ goto destroy_flow_group;
+ }
+
+ kvfree(spec);
+ kvfree(flow_group_in);
+ roce->ft = ft;
+ roce->fg = fg;
+ roce->allow_rule = flow_rule;
+
+ return 0;
+
+destroy_flow_table:
+ mlx5_destroy_flow_table(ft);
+destroy_flow_group:
+ mlx5_destroy_flow_group(fg);
+free:
+ kvfree(spec);
+ kvfree(flow_group_in);
+ return err;
+}
+
+static void mlx5_rdma_del_roce_addr(struct mlx5_core_dev *dev)
+{
+ mlx5_core_roce_gid_set(dev, 0, 0, 0,
+ NULL, NULL, false, 0, 0);
+}
+
+static void mlx5_rdma_make_default_gid(struct mlx5_core_dev *dev, union ib_gid *gid)
+{
+ u8 hw_id[ETH_ALEN];
+
+ mlx5_query_nic_vport_mac_address(dev, 0, hw_id);
+ gid->global.subnet_prefix = cpu_to_be64(0xfe80000000000000LL);
+ addrconf_addr_eui48(&gid->raw[8], hw_id);
+}
+
+static int mlx5_rdma_add_roce_addr(struct mlx5_core_dev *dev)
+{
+ union ib_gid gid;
+ u8 mac[ETH_ALEN];
+
+ mlx5_rdma_make_default_gid(dev, &gid);
+ return mlx5_core_roce_gid_set(dev, 0,
+ MLX5_ROCE_VERSION_1,
+ 0, gid.raw, mac,
+ false, 0, 1);
+}
+
+void mlx5_rdma_disable_roce(struct mlx5_core_dev *dev)
+{
+ mlx5_rdma_disable_roce_steering(dev);
+ mlx5_rdma_del_roce_addr(dev);
+ mlx5_nic_vport_disable_roce(dev);
+}
+
+void mlx5_rdma_enable_roce(struct mlx5_core_dev *dev)
+{
+ int err;
+
+ err = mlx5_nic_vport_enable_roce(dev);
+ if (err) {
+ mlx5_core_err(dev, "Failed to enable RoCE: %d\n", err);
+ return;
+ }
+
+ err = mlx5_rdma_add_roce_addr(dev);
+ if (err) {
+ mlx5_core_err(dev, "Failed to add RoCE address: %d\n", err);
+ goto disable_roce;
+ }
+
+ err = mlx5_rdma_enable_roce_steering(dev);
+ if (err) {
+ mlx5_core_err(dev, "Failed to enable RoCE steering: %d\n", err);
+ goto del_roce_addr;
+ }
+
+ return;
+
+del_roce_addr:
+ mlx5_rdma_del_roce_addr(dev);
+disable_roce:
+ mlx5_nic_vport_disable_roce(dev);
+ return;
+}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/rdma.h b/drivers/net/ethernet/mellanox/mlx5/core/rdma.h
new file mode 100644
index 000000000000..750cff2a71a4
--- /dev/null
+++ b/drivers/net/ethernet/mellanox/mlx5/core/rdma.h
@@ -0,0 +1,20 @@
+/* SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB */
+/* Copyright (c) 2019 Mellanox Technologies. */
+
+#ifndef __MLX5_RDMA_H__
+#define __MLX5_RDMA_H__
+
+#include "mlx5_core.h"
+
+#ifdef CONFIG_MLX5_ESWITCH
+
+void mlx5_rdma_enable_roce(struct mlx5_core_dev *dev);
+void mlx5_rdma_disable_roce(struct mlx5_core_dev *dev);
+
+#else /* CONFIG_MLX5_ESWITCH */
+
+static inline void mlx5_rdma_enable_roce(struct mlx5_core_dev *dev) {}
+static inline void mlx5_rdma_disable_roce(struct mlx5_core_dev *dev) {}
+
+#endif /* CONFIG_MLX5_ESWITCH */
+#endif /* __MLX5_RDMA_H__ */
diff --git a/include/linux/mlx5/driver.h b/include/linux/mlx5/driver.h
index 582a9680b182..7fa95270dd59 100644
--- a/include/linux/mlx5/driver.h
+++ b/include/linux/mlx5/driver.h
@@ -512,6 +512,12 @@ struct mlx5_rl_table {
struct mlx5_rl_entry *rl_entry;
};
+struct mlx5_core_roce {
+ struct mlx5_flow_table *ft;
+ struct mlx5_flow_group *fg;
+ struct mlx5_flow_handle *allow_rule;
+};
+
struct mlx5_priv {
struct mlx5_eq_table *eq_table;
@@ -565,6 +571,7 @@ struct mlx5_priv {
struct mlx5_lag *lag;
struct mlx5_devcom *devcom;
unsigned long pci_dev_data;
+ struct mlx5_core_roce roce;
struct mlx5_fc_stats fc_stats;
struct mlx5_rl_table rl_table;
--
2.20.1
^ permalink raw reply related [flat|nested] 17+ messages in thread
* [PATCH V2 mlx5-next 10/11] net/mlx5: Geneve, Add basic Geneve encap/decap flow table capabilities
2019-04-29 18:13 [PATCH V2 mlx5-next 00/11] Mellanox, mlx5-next updates 2019-04-25 Saeed Mahameed
` (8 preceding siblings ...)
2019-04-29 18:14 ` [PATCH V2 mlx5-next 09/11] net/mlx5: Eswitch, enable RoCE loopback traffic Saeed Mahameed
@ 2019-04-29 18:14 ` Saeed Mahameed
2019-04-29 18:14 ` [PATCH V2 mlx5-next 11/11] net/mlx5: Geneve, Add flow table capabilities for Geneve decap with TLV options Saeed Mahameed
2019-04-29 23:56 ` [PATCH V2 mlx5-next 00/11] Mellanox, mlx5-next updates 2019-04-25 Saeed Mahameed
11 siblings, 0 replies; 17+ messages in thread
From: Saeed Mahameed @ 2019-04-29 18:14 UTC (permalink / raw)
To: Saeed Mahameed, Leon Romanovsky
Cc: Jason Gunthorpe, netdev, linux-rdma, Yevgeny Kliteynik, Oz Shlomo
From: Yevgeny Kliteynik <kliteyn@mellanox.com>
Introduce support for Geneve flow specification and allow
the creation of rules that are matching on basic Geneve
protocol fields: VNI, OAM bit, protocol type, options length.
Reviewed-by: Oz Shlomo <ozsh@mellanox.com>
Signed-off-by: Yevgeny Kliteynik <kliteyn@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
include/linux/mlx5/mlx5_ifc.h | 16 +++++++++++++---
1 file changed, 13 insertions(+), 3 deletions(-)
diff --git a/include/linux/mlx5/mlx5_ifc.h b/include/linux/mlx5/mlx5_ifc.h
index 7d9264b282d1..268ac126b3bb 100644
--- a/include/linux/mlx5/mlx5_ifc.h
+++ b/include/linux/mlx5/mlx5_ifc.h
@@ -307,7 +307,11 @@ struct mlx5_ifc_flow_table_fields_supported_bits {
u8 outer_gre_protocol[0x1];
u8 outer_gre_key[0x1];
u8 outer_vxlan_vni[0x1];
- u8 reserved_at_1a[0x5];
+ u8 outer_geneve_vni[0x1];
+ u8 outer_geneve_oam[0x1];
+ u8 outer_geneve_protocol_type[0x1];
+ u8 outer_geneve_opt_len[0x1];
+ u8 reserved_at_1e[0x1];
u8 source_eswitch_port[0x1];
u8 inner_dmac[0x1];
@@ -480,7 +484,9 @@ struct mlx5_ifc_fte_match_set_misc_bits {
u8 vxlan_vni[0x18];
u8 reserved_at_b8[0x8];
- u8 reserved_at_c0[0x20];
+ u8 geneve_vni[0x18];
+ u8 reserved_at_d8[0x7];
+ u8 geneve_oam[0x1];
u8 reserved_at_e0[0xc];
u8 outer_ipv6_flow_label[0x14];
@@ -488,7 +494,11 @@ struct mlx5_ifc_fte_match_set_misc_bits {
u8 reserved_at_100[0xc];
u8 inner_ipv6_flow_label[0x14];
- u8 reserved_at_120[0x28];
+ u8 reserved_at_120[0xa];
+ u8 geneve_opt_len[0x6];
+ u8 geneve_protocol_type[0x10];
+
+ u8 reserved_at_140[0x8];
u8 bth_dst_qp[0x18];
u8 reserved_at_160[0x20];
u8 outer_esp_spi[0x20];
--
2.20.1
^ permalink raw reply related [flat|nested] 17+ messages in thread
* [PATCH V2 mlx5-next 11/11] net/mlx5: Geneve, Add flow table capabilities for Geneve decap with TLV options
2019-04-29 18:13 [PATCH V2 mlx5-next 00/11] Mellanox, mlx5-next updates 2019-04-25 Saeed Mahameed
` (9 preceding siblings ...)
2019-04-29 18:14 ` [PATCH V2 mlx5-next 10/11] net/mlx5: Geneve, Add basic Geneve encap/decap flow table capabilities Saeed Mahameed
@ 2019-04-29 18:14 ` Saeed Mahameed
2019-04-29 23:56 ` [PATCH V2 mlx5-next 00/11] Mellanox, mlx5-next updates 2019-04-25 Saeed Mahameed
11 siblings, 0 replies; 17+ messages in thread
From: Saeed Mahameed @ 2019-04-29 18:14 UTC (permalink / raw)
To: Saeed Mahameed, Leon Romanovsky
Cc: Jason Gunthorpe, netdev, linux-rdma, Yevgeny Kliteynik, Oz Shlomo
From: Yevgeny Kliteynik <kliteyn@mellanox.com>
Introduce specification for Geneve decap flow with encapsulation options
and allow creation of rules that are matching on Geneve TLV options.
Reviewed-by: Oz Shlomo <ozsh@mellanox.com>
Signed-off-by: Yevgeny Kliteynik <kliteyn@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
.../net/ethernet/mellanox/mlx5/core/fs_core.h | 2 +-
include/linux/mlx5/device.h | 4 +-
include/linux/mlx5/mlx5_ifc.h | 50 +++++++++++++++++--
3 files changed, 49 insertions(+), 7 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h
index 0c6c5fef4548..a08c3d09a50f 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h
@@ -152,7 +152,7 @@ struct mlx5_ft_underlay_qp {
u32 qpn;
};
-#define MLX5_FTE_MATCH_PARAM_RESERVED reserved_at_800
+#define MLX5_FTE_MATCH_PARAM_RESERVED reserved_at_a00
/* Calculate the fte_match_param length and without the reserved length.
* Make sure the reserved field is the last.
*/
diff --git a/include/linux/mlx5/device.h b/include/linux/mlx5/device.h
index 28ebb6c93542..4ab801040e98 100644
--- a/include/linux/mlx5/device.h
+++ b/include/linux/mlx5/device.h
@@ -1001,7 +1001,8 @@ enum {
MLX5_MATCH_OUTER_HEADERS = 1 << 0,
MLX5_MATCH_MISC_PARAMETERS = 1 << 1,
MLX5_MATCH_INNER_HEADERS = 1 << 2,
-
+ MLX5_MATCH_MISC_PARAMETERS_2 = 1 << 3,
+ MLX5_MATCH_MISC_PARAMETERS_3 = 1 << 4,
};
enum {
@@ -1045,6 +1046,7 @@ enum mlx5_mpls_supported_fields {
};
enum mlx5_flex_parser_protos {
+ MLX5_FLEX_PROTO_GENEVE = 1 << 3,
MLX5_FLEX_PROTO_CW_MPLS_GRE = 1 << 4,
MLX5_FLEX_PROTO_CW_MPLS_UDP = 1 << 5,
};
diff --git a/include/linux/mlx5/mlx5_ifc.h b/include/linux/mlx5/mlx5_ifc.h
index 268ac126b3bb..6a7fc18a9fe3 100644
--- a/include/linux/mlx5/mlx5_ifc.h
+++ b/include/linux/mlx5/mlx5_ifc.h
@@ -86,6 +86,11 @@ enum {
enum {
MLX5_GENERAL_OBJ_TYPES_CAP_SW_ICM = (1ULL << MLX5_OBJ_TYPE_SW_ICM),
+ MLX5_GENERAL_OBJ_TYPES_CAP_GENEVE_TLV_OPT = (1ULL << 11),
+};
+
+enum {
+ MLX5_OBJ_TYPE_GENEVE_TLV_OPT = 0x000b,
};
enum {
@@ -339,7 +344,8 @@ struct mlx5_ifc_flow_table_fields_supported_bits {
u8 inner_tcp_flags[0x1];
u8 reserved_at_37[0x9];
- u8 reserved_at_40[0x5];
+ u8 geneve_tlv_option_0_data[0x1];
+ u8 reserved_at_41[0x4];
u8 outer_first_mpls_over_udp[0x4];
u8 outer_first_mpls_over_gre[0x4];
u8 inner_first_mpls[0x4];
@@ -528,6 +534,12 @@ struct mlx5_ifc_fte_match_set_misc2_bits {
u8 reserved_at_1a0[0x60];
};
+struct mlx5_ifc_fte_match_set_misc3_bits {
+ u8 reserved_at_0[0x120];
+ u8 geneve_tlv_option_0_data[0x20];
+ u8 reserved_at_140[0xc0];
+};
+
struct mlx5_ifc_cmd_pas_bits {
u8 pa_h[0x20];
@@ -1247,9 +1259,13 @@ struct mlx5_ifc_cmd_hca_cap_bits {
u8 num_of_uars_per_page[0x20];
u8 flex_parser_protocols[0x20];
- u8 reserved_at_560[0x20];
- u8 reserved_at_580[0x3c];
+ u8 max_geneve_tlv_options[0x8];
+ u8 reserved_at_568[0x3];
+ u8 max_geneve_tlv_option_data_len[0x5];
+ u8 reserved_at_570[0x10];
+
+ u8 reserved_at_580[0x1c];
u8 mini_cqe_resp_stride_index[0x1];
u8 cqe_128_always[0x1];
u8 cqe_compression_128[0x1];
@@ -1283,7 +1299,9 @@ struct mlx5_ifc_cmd_hca_cap_bits {
u8 uctx_cap[0x20];
- u8 reserved_at_6c0[0x140];
+ u8 reserved_at_6c0[0x4];
+ u8 flex_parser_id_geneve_tlv_option_0[0x4];
+ u8 reserved_at_6c8[0x138];
};
enum mlx5_flow_destination_type {
@@ -1341,7 +1359,9 @@ struct mlx5_ifc_fte_match_param_bits {
struct mlx5_ifc_fte_match_set_misc2_bits misc_parameters_2;
- u8 reserved_at_800[0x800];
+ struct mlx5_ifc_fte_match_set_misc3_bits misc_parameters_3;
+
+ u8 reserved_at_a00[0x600];
};
enum {
@@ -4850,6 +4870,7 @@ enum {
MLX5_QUERY_FLOW_GROUP_OUT_MATCH_CRITERIA_ENABLE_MISC_PARAMETERS = 0x1,
MLX5_QUERY_FLOW_GROUP_OUT_MATCH_CRITERIA_ENABLE_INNER_HEADERS = 0x2,
MLX5_QUERY_FLOW_GROUP_IN_MATCH_CRITERIA_ENABLE_MISC_PARAMETERS_2 = 0x3,
+ MLX5_QUERY_FLOW_GROUP_IN_MATCH_CRITERIA_ENABLE_MISC_PARAMETERS_3 = 0x4,
};
struct mlx5_ifc_query_flow_group_out_bits {
@@ -9545,6 +9566,20 @@ struct mlx5_ifc_sw_icm_bits {
u8 sw_icm_start_addr[0x40];
u8 reserved_at_c0[0x140];
+};
+
+struct mlx5_ifc_geneve_tlv_option_bits {
+ u8 modify_field_select[0x40];
+
+ u8 reserved_at_40[0x18];
+ u8 geneve_option_fte_index[0x8];
+
+ u8 option_class[0x10];
+ u8 option_type[0x8];
+ u8 reserved_at_78[0x3];
+ u8 option_data_length[0x5];
+
+ u8 reserved_at_80[0x180];
};
struct mlx5_ifc_create_umem_in_bits {
@@ -9589,6 +9624,11 @@ struct mlx5_ifc_create_sw_icm_in_bits {
struct mlx5_ifc_sw_icm_bits sw_icm;
};
+struct mlx5_ifc_create_geneve_tlv_option_in_bits {
+ struct mlx5_ifc_general_obj_in_cmd_hdr_bits hdr;
+ struct mlx5_ifc_geneve_tlv_option_bits geneve_tlv_opt;
+};
+
struct mlx5_ifc_mtrc_string_db_param_bits {
u8 string_db_base_address[0x20];
--
2.20.1
^ permalink raw reply related [flat|nested] 17+ messages in thread
* Re: [PATCH V2 mlx5-next 09/11] net/mlx5: Eswitch, enable RoCE loopback traffic
2019-04-29 18:14 ` [PATCH V2 mlx5-next 09/11] net/mlx5: Eswitch, enable RoCE loopback traffic Saeed Mahameed
@ 2019-04-29 18:41 ` Leon Romanovsky
2019-04-29 18:45 ` Parav Pandit
0 siblings, 1 reply; 17+ messages in thread
From: Leon Romanovsky @ 2019-04-29 18:41 UTC (permalink / raw)
To: Saeed Mahameed
Cc: Jason Gunthorpe, netdev, linux-rdma, Maor Gottlieb, Mark Bloch
On Mon, Apr 29, 2019 at 06:14:16PM +0000, Saeed Mahameed wrote:
> From: Maor Gottlieb <maorg@mellanox.com>
>
> When in switchdev mode, we would like to treat loopback RoCE
> traffic (on eswitch manager) as RDMA and not as regular
> Ethernet traffic
> In order to enable it we add flow steering rule that forward RoCE
> loopback traffic to the HW RoCE filter (by adding allow rule).
> In addition we add RoCE address in GID index 0, which will be
> set in the RoCE loopback packet.
>
> Signed-off-by: Maor Gottlieb <maorg@mellanox.com>
> Reviewed-by: Mark Bloch <markb@mellanox.com>
> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
> ---
> .../net/ethernet/mellanox/mlx5/core/Makefile | 2 +-
> .../mellanox/mlx5/core/eswitch_offloads.c | 4 +
> .../net/ethernet/mellanox/mlx5/core/rdma.c | 182 ++++++++++++++++++
> .../net/ethernet/mellanox/mlx5/core/rdma.h | 20 ++
> include/linux/mlx5/driver.h | 7 +
> 5 files changed, 214 insertions(+), 1 deletion(-)
> create mode 100644 drivers/net/ethernet/mellanox/mlx5/core/rdma.c
> create mode 100644 drivers/net/ethernet/mellanox/mlx5/core/rdma.h
>
Thanks,
Acked-by: Leon Romanovsky <leonro@mellanox.com>
^ permalink raw reply [flat|nested] 17+ messages in thread
* RE: [PATCH V2 mlx5-next 09/11] net/mlx5: Eswitch, enable RoCE loopback traffic
2019-04-29 18:41 ` Leon Romanovsky
@ 2019-04-29 18:45 ` Parav Pandit
2019-04-29 22:30 ` Mark Bloch
0 siblings, 1 reply; 17+ messages in thread
From: Parav Pandit @ 2019-04-29 18:45 UTC (permalink / raw)
To: Leon Romanovsky, Saeed Mahameed
Cc: Jason Gunthorpe, netdev, linux-rdma, Maor Gottlieb, Mark Bloch
> -----Original Message-----
> From: netdev-owner@vger.kernel.org <netdev-owner@vger.kernel.org> On
> Behalf Of Leon Romanovsky
> Sent: Monday, April 29, 2019 1:41 PM
> To: Saeed Mahameed <saeedm@mellanox.com>
> Cc: Jason Gunthorpe <jgg@mellanox.com>; netdev@vger.kernel.org; linux-
> rdma@vger.kernel.org; Maor Gottlieb <maorg@mellanox.com>; Mark Bloch
> <markb@mellanox.com>
> Subject: Re: [PATCH V2 mlx5-next 09/11] net/mlx5: Eswitch, enable RoCE
> loopback traffic
>
> On Mon, Apr 29, 2019 at 06:14:16PM +0000, Saeed Mahameed wrote:
> > From: Maor Gottlieb <maorg@mellanox.com>
> >
> > When in switchdev mode, we would like to treat loopback RoCE traffic
> > (on eswitch manager) as RDMA and not as regular Ethernet traffic In
> > order to enable it we add flow steering rule that forward RoCE
> > loopback traffic to the HW RoCE filter (by adding allow rule).
> > In addition we add RoCE address in GID index 0, which will be set in
> > the RoCE loopback packet.
> >
I likely don't understand nor I reviewed the patches.
Part that I don't understand is GID index 0 for RoCE.
RoCE traffic runs over all the GID entries and for all practical purposes from non_zero index.
How will it work?
It is better if you explain it in the commit log, why its done this way, 'what' part is already present the patch.
> > Signed-off-by: Maor Gottlieb <maorg@mellanox.com>
> > Reviewed-by: Mark Bloch <markb@mellanox.com>
> > Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
> > ---
> > .../net/ethernet/mellanox/mlx5/core/Makefile | 2 +-
> > .../mellanox/mlx5/core/eswitch_offloads.c | 4 +
> > .../net/ethernet/mellanox/mlx5/core/rdma.c | 182 ++++++++++++++++++
> > .../net/ethernet/mellanox/mlx5/core/rdma.h | 20 ++
> > include/linux/mlx5/driver.h | 7 +
> > 5 files changed, 214 insertions(+), 1 deletion(-) create mode 100644
> > drivers/net/ethernet/mellanox/mlx5/core/rdma.c
> > create mode 100644 drivers/net/ethernet/mellanox/mlx5/core/rdma.h
> >
>
> Thanks,
> Acked-by: Leon Romanovsky <leonro@mellanox.com>
^ permalink raw reply [flat|nested] 17+ messages in thread
* Re: [PATCH V2 mlx5-next 09/11] net/mlx5: Eswitch, enable RoCE loopback traffic
2019-04-29 18:45 ` Parav Pandit
@ 2019-04-29 22:30 ` Mark Bloch
2019-04-29 22:34 ` Parav Pandit
0 siblings, 1 reply; 17+ messages in thread
From: Mark Bloch @ 2019-04-29 22:30 UTC (permalink / raw)
To: Parav Pandit, Leon Romanovsky, Saeed Mahameed
Cc: Jason Gunthorpe, netdev, linux-rdma, Maor Gottlieb
On 4/29/19 11:45 AM, Parav Pandit wrote:
>
>
>> -----Original Message-----
>> From: netdev-owner@vger.kernel.org <netdev-owner@vger.kernel.org> On
>> Behalf Of Leon Romanovsky
>> Sent: Monday, April 29, 2019 1:41 PM
>> To: Saeed Mahameed <saeedm@mellanox.com>
>> Cc: Jason Gunthorpe <jgg@mellanox.com>; netdev@vger.kernel.org; linux-
>> rdma@vger.kernel.org; Maor Gottlieb <maorg@mellanox.com>; Mark Bloch
>> <markb@mellanox.com>
>> Subject: Re: [PATCH V2 mlx5-next 09/11] net/mlx5: Eswitch, enable RoCE
>> loopback traffic
>>
>> On Mon, Apr 29, 2019 at 06:14:16PM +0000, Saeed Mahameed wrote:
>>> From: Maor Gottlieb <maorg@mellanox.com>
>>>
>>> When in switchdev mode, we would like to treat loopback RoCE traffic
>>> (on eswitch manager) as RDMA and not as regular Ethernet traffic In
>>> order to enable it we add flow steering rule that forward RoCE
>>> loopback traffic to the HW RoCE filter (by adding allow rule).
>>> In addition we add RoCE address in GID index 0, which will be set in
>>> the RoCE loopback packet.
>>>
> I likely don't understand nor I reviewed the patches.
> Part that I don't understand is GID index 0 for RoCE.
> RoCE traffic runs over all the GID entries and for all practical purposes from non_zero index.
> How will it work?
Currently in switchdev mode we only support RAW Ethernet QP and no RoCE capabilities are reported to ib_core.
Which means no GIDs are inserted to the HW's GID table and RoCE isn't enabled on the vport.
However, there are cases where an internal RC QP might be needed, and for that we need a GID.
The patches from Maor make sure a GID entry at index 0 is valid (and we do that only in switchdev mode),
and with steering we make sure such traffic is only used for loopback purposes.
Mark
> It is better if you explain it in the commit log, why its done this way, 'what' part is already present the patch.
>
>
>>> Signed-off-by: Maor Gottlieb <maorg@mellanox.com>
>>> Reviewed-by: Mark Bloch <markb@mellanox.com>
>>> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
>>> ---
>>> .../net/ethernet/mellanox/mlx5/core/Makefile | 2 +-
>>> .../mellanox/mlx5/core/eswitch_offloads.c | 4 +
>>> .../net/ethernet/mellanox/mlx5/core/rdma.c | 182 ++++++++++++++++++
>>> .../net/ethernet/mellanox/mlx5/core/rdma.h | 20 ++
>>> include/linux/mlx5/driver.h | 7 +
>>> 5 files changed, 214 insertions(+), 1 deletion(-) create mode 100644
>>> drivers/net/ethernet/mellanox/mlx5/core/rdma.c
>>> create mode 100644 drivers/net/ethernet/mellanox/mlx5/core/rdma.h
>>>
>>
>> Thanks,
>> Acked-by: Leon Romanovsky <leonro@mellanox.com>
^ permalink raw reply [flat|nested] 17+ messages in thread
* RE: [PATCH V2 mlx5-next 09/11] net/mlx5: Eswitch, enable RoCE loopback traffic
2019-04-29 22:30 ` Mark Bloch
@ 2019-04-29 22:34 ` Parav Pandit
0 siblings, 0 replies; 17+ messages in thread
From: Parav Pandit @ 2019-04-29 22:34 UTC (permalink / raw)
To: Mark Bloch, Leon Romanovsky, Saeed Mahameed
Cc: Jason Gunthorpe, netdev, linux-rdma, Maor Gottlieb
Hi Mark,
> -----Original Message-----
> From: Mark Bloch
> Sent: Monday, April 29, 2019 5:31 PM
> To: Parav Pandit <parav@mellanox.com>; Leon Romanovsky
> <leonro@mellanox.com>; Saeed Mahameed <saeedm@mellanox.com>
> Cc: Jason Gunthorpe <jgg@mellanox.com>; netdev@vger.kernel.org; linux-
> rdma@vger.kernel.org; Maor Gottlieb <maorg@mellanox.com>
> Subject: Re: [PATCH V2 mlx5-next 09/11] net/mlx5: Eswitch, enable RoCE
> loopback traffic
>
>
>
> On 4/29/19 11:45 AM, Parav Pandit wrote:
> >
> >
> >> -----Original Message-----
> >> From: netdev-owner@vger.kernel.org <netdev-owner@vger.kernel.org>
> On
> >> Behalf Of Leon Romanovsky
> >> Sent: Monday, April 29, 2019 1:41 PM
> >> To: Saeed Mahameed <saeedm@mellanox.com>
> >> Cc: Jason Gunthorpe <jgg@mellanox.com>; netdev@vger.kernel.org;
> >> linux- rdma@vger.kernel.org; Maor Gottlieb <maorg@mellanox.com>;
> Mark
> >> Bloch <markb@mellanox.com>
> >> Subject: Re: [PATCH V2 mlx5-next 09/11] net/mlx5: Eswitch, enable
> >> RoCE loopback traffic
> >>
> >> On Mon, Apr 29, 2019 at 06:14:16PM +0000, Saeed Mahameed wrote:
> >>> From: Maor Gottlieb <maorg@mellanox.com>
> >>>
> >>> When in switchdev mode, we would like to treat loopback RoCE traffic
> >>> (on eswitch manager) as RDMA and not as regular Ethernet traffic In
> >>> order to enable it we add flow steering rule that forward RoCE
> >>> loopback traffic to the HW RoCE filter (by adding allow rule).
> >>> In addition we add RoCE address in GID index 0, which will be set in
> >>> the RoCE loopback packet.
> >>>
> > I likely don't understand nor I reviewed the patches.
> > Part that I don't understand is GID index 0 for RoCE.
> > RoCE traffic runs over all the GID entries and for all practical purposes from
> non_zero index.
> > How will it work?
>
> Currently in switchdev mode we only support RAW Ethernet QP and no RoCE
> capabilities are reported to ib_core.
> Which means no GIDs are inserted to the HW's GID table and RoCE isn't
> enabled on the vport.
>
> However, there are cases where an internal RC QP might be needed, and for
> that we need a GID.
> The patches from Maor make sure a GID entry at index 0 is valid (and we do
> that only in switchdev mode), and with steering we make sure such traffic is
> only used for loopback purposes.
Got it. Thanks for the detailed description. Its clear to me now.
^ permalink raw reply [flat|nested] 17+ messages in thread
* Re: [PATCH V2 mlx5-next 00/11] Mellanox, mlx5-next updates 2019-04-25
2019-04-29 18:13 [PATCH V2 mlx5-next 00/11] Mellanox, mlx5-next updates 2019-04-25 Saeed Mahameed
` (10 preceding siblings ...)
2019-04-29 18:14 ` [PATCH V2 mlx5-next 11/11] net/mlx5: Geneve, Add flow table capabilities for Geneve decap with TLV options Saeed Mahameed
@ 2019-04-29 23:56 ` Saeed Mahameed
11 siblings, 0 replies; 17+ messages in thread
From: Saeed Mahameed @ 2019-04-29 23:56 UTC (permalink / raw)
To: Leon Romanovsky; +Cc: Jason Gunthorpe, netdev, linux-rdma
On Mon, 2019-04-29 at 18:13 +0000, Saeed Mahameed wrote:
> Hi,
>
> This series provides misc low level updates to mlx5 core driver, to
> be
> shared between rdma and net-next trees.
>
> 1) From Aya: Enable general events on all physical link types and
> restrict general event handling of subtype DELAY_DROP_TIMEOUT in mlx5
> rdma
> driver to ethernet links only as it was intended.
>
> 2) From Eli: Introduce low level bits for prio tag mode
>
> 3) From Maor: Low level steering updates to support RDMA RX flow
> steering and enables RoCE loopback traffic when switchdev is enabled.
>
> 4) From Vu and Parav: Two small mlx5 core cleanups
>
> 5) From Yevgeny add HW definitions of geneve offloads
>
> In case of no objections this series will be applied to mlx5-next
> branch.
>
> v2:
> - Address comments from Leon.
> - Static checker fixes.
>
> Thanks,
> Saeed.
>
>
Applied to mlx5-next,
Thanks !
^ permalink raw reply [flat|nested] 17+ messages in thread
end of thread, other threads:[~2019-04-29 23:56 UTC | newest]
Thread overview: 17+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2019-04-29 18:13 [PATCH V2 mlx5-next 00/11] Mellanox, mlx5-next updates 2019-04-25 Saeed Mahameed
2019-04-29 18:14 ` [PATCH V2 mlx5-next 01/11] net/mlx5: E-Switch: Introduce prio tag mode Saeed Mahameed
2019-04-29 18:14 ` [PATCH V2 mlx5-next 02/11] net/mlx5: Get rid of storing copy of device name Saeed Mahameed
2019-04-29 18:14 ` [PATCH V2 mlx5-next 03/11] net/mlx5: Separate and generalize dma device from pci device Saeed Mahameed
2019-04-29 18:14 ` [PATCH V2 mlx5-next 04/11] IB/mlx5: Restrict 'DELAY_DROP_TIMEOUT' subtype to Ethernet interfaces Saeed Mahameed
2019-04-29 18:14 ` [PATCH V2 mlx5-next 05/11] net/mlx5: Enable general events on all interfaces Saeed Mahameed
2019-04-29 18:14 ` [PATCH V2 mlx5-next 06/11] net/mlx5: Pass flow steering objects to fs_cmd Saeed Mahameed
2019-04-29 18:14 ` [PATCH V2 mlx5-next 07/11] net/mlx5: Add support in RDMA RX steering Saeed Mahameed
2019-04-29 18:14 ` [PATCH V2 mlx5-next 08/11] net/mlx5: Add new miss flow table action Saeed Mahameed
2019-04-29 18:14 ` [PATCH V2 mlx5-next 09/11] net/mlx5: Eswitch, enable RoCE loopback traffic Saeed Mahameed
2019-04-29 18:41 ` Leon Romanovsky
2019-04-29 18:45 ` Parav Pandit
2019-04-29 22:30 ` Mark Bloch
2019-04-29 22:34 ` Parav Pandit
2019-04-29 18:14 ` [PATCH V2 mlx5-next 10/11] net/mlx5: Geneve, Add basic Geneve encap/decap flow table capabilities Saeed Mahameed
2019-04-29 18:14 ` [PATCH V2 mlx5-next 11/11] net/mlx5: Geneve, Add flow table capabilities for Geneve decap with TLV options Saeed Mahameed
2019-04-29 23:56 ` [PATCH V2 mlx5-next 00/11] Mellanox, mlx5-next updates 2019-04-25 Saeed Mahameed
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).