This series provides knobs which will enable users to
minimize memory consumption of mlx5 Functions (PF/VF/SF).
mlx5 exposes two new generic devlink params for EQ size
configuration and uses devlink generic param max_macs.
Patches summary:
- Patch-1 Introduce log_max_current_uc_list_wr_supported bit
- Patches-2-3 Provides I/O EQ size param which enables to save
up to 128KB.
- Patches-4-5 Provides event EQ size param which enables to save
up to 512KB.
- Patch-6 Clarify max_macs param.
- Patch-7 Provides max_macs param which enables to save up to 70KB
In total, this series can save up to 700KB per Function.
---
changelog:
v2->v3:
- change type of EQ size param to u32 per Jiri suggestion.
- separate ifc changes to new patch
v1->v2:
- convert io_eq_size and event_eq_size from devlink_resources to
generic devlink_params
Shay Drory (7):
net/mlx5: Introduce log_max_current_uc_list_wr_supported bit
devlink: Add new "io_eq_size" generic device param
net/mlx5: Let user configure io_eq_size param
devlink: Add new "event_eq_size" generic device param
net/mlx5: Let user configure event_eq_size param
devlink: Clarifies max_macs generic devlink param
net/mlx5: Let user configure max_macs generic param
.../networking/devlink/devlink-params.rst | 12 ++-
Documentation/networking/devlink/mlx5.rst | 10 +++
.../net/ethernet/mellanox/mlx5/core/devlink.c | 88 +++++++++++++++++++
drivers/net/ethernet/mellanox/mlx5/core/eq.c | 34 ++++++-
.../net/ethernet/mellanox/mlx5/core/main.c | 21 +++++
include/linux/mlx5/mlx5_ifc.h | 2 +-
include/net/devlink.h | 8 ++
net/core/devlink.c | 10 +++
8 files changed, 180 insertions(+), 5 deletions(-)
--
2.21.3
Downstream patch will use this bit in order to know whether the device
supports changing of max_uc_list.
Signed-off-by: Shay Drory <[email protected]>
Reviewed-by: Moshe Shemesh <[email protected]>
---
include/linux/mlx5/mlx5_ifc.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/include/linux/mlx5/mlx5_ifc.h b/include/linux/mlx5/mlx5_ifc.h
index fbaab440a484..e9db12aae8f9 100644
--- a/include/linux/mlx5/mlx5_ifc.h
+++ b/include/linux/mlx5/mlx5_ifc.h
@@ -1621,7 +1621,7 @@ struct mlx5_ifc_cmd_hca_cap_bits {
u8 ext_stride_num_range[0x1];
u8 roce_rw_supported[0x1];
- u8 reserved_at_3a2[0x1];
+ u8 log_max_current_uc_list_wr_supported[0x1];
u8 log_max_stride_sz_rq[0x5];
u8 reserved_at_3a8[0x3];
u8 log_min_stride_sz_rq[0x5];
--
2.21.3
Add new device generic parameter to determine the size of the
I/O completion EQs.
For example, to reduce I/O EQ size to 64, execute:
$ devlink dev param set pci/0000:06:00.0 \
name io_eq_size value 64 cmode driverinit
$ devlink dev reload pci/0000:06:00.0
Signed-off-by: Shay Drory <[email protected]>
Reviewed-by: Moshe Shemesh <[email protected]>
---
Documentation/networking/devlink/devlink-params.rst | 3 +++
include/net/devlink.h | 4 ++++
net/core/devlink.c | 5 +++++
3 files changed, 12 insertions(+)
diff --git a/Documentation/networking/devlink/devlink-params.rst b/Documentation/networking/devlink/devlink-params.rst
index b7dfe693a332..cd9342305a13 100644
--- a/Documentation/networking/devlink/devlink-params.rst
+++ b/Documentation/networking/devlink/devlink-params.rst
@@ -129,3 +129,6 @@ own name.
will NACK any attempt of other host to reset the device. This parameter
is useful for setups where a device is shared by different hosts, such
as multi-host setup.
+ * - ``io_eq_size``
+ - u16
+ - Control the size of I/O completion EQs.
diff --git a/include/net/devlink.h b/include/net/devlink.h
index 3276a29f2b81..b5f4acd0e0cd 100644
--- a/include/net/devlink.h
+++ b/include/net/devlink.h
@@ -459,6 +459,7 @@ enum devlink_param_generic_id {
DEVLINK_PARAM_GENERIC_ID_ENABLE_RDMA,
DEVLINK_PARAM_GENERIC_ID_ENABLE_VNET,
DEVLINK_PARAM_GENERIC_ID_ENABLE_IWARP,
+ DEVLINK_PARAM_GENERIC_ID_IO_EQ_SIZE,
/* add new param generic ids above here*/
__DEVLINK_PARAM_GENERIC_ID_MAX,
@@ -511,6 +512,9 @@ enum devlink_param_generic_id {
#define DEVLINK_PARAM_GENERIC_ENABLE_IWARP_NAME "enable_iwarp"
#define DEVLINK_PARAM_GENERIC_ENABLE_IWARP_TYPE DEVLINK_PARAM_TYPE_BOOL
+#define DEVLINK_PARAM_GENERIC_IO_EQ_SIZE_NAME "io_eq_size"
+#define DEVLINK_PARAM_GENERIC_IO_EQ_SIZE_TYPE DEVLINK_PARAM_TYPE_U32
+
#define DEVLINK_PARAM_GENERIC(_id, _cmodes, _get, _set, _validate) \
{ \
.id = DEVLINK_PARAM_GENERIC_ID_##_id, \
diff --git a/net/core/devlink.c b/net/core/devlink.c
index db3b52110cf2..0d4e63d11585 100644
--- a/net/core/devlink.c
+++ b/net/core/devlink.c
@@ -4466,6 +4466,11 @@ static const struct devlink_param devlink_param_generic[] = {
.name = DEVLINK_PARAM_GENERIC_ENABLE_IWARP_NAME,
.type = DEVLINK_PARAM_GENERIC_ENABLE_IWARP_TYPE,
},
+ {
+ .id = DEVLINK_PARAM_GENERIC_ID_IO_EQ_SIZE,
+ .name = DEVLINK_PARAM_GENERIC_IO_EQ_SIZE_NAME,
+ .type = DEVLINK_PARAM_GENERIC_IO_EQ_SIZE_TYPE,
+ },
};
static int devlink_param_generic_verify(const struct devlink_param *param)
--
2.21.3
Currently, each I/O EQ is taking 128KB of memory. This size
is not needed in all use cases, and is critical with large scale.
Hence, allow user to configure the size of I/O EQs.
For example, to reduce I/O EQ size to 64, execute:
$ devlink dev param set pci/0000:00:0b.0 name io_eq_size value 64 \
cmode driverinit
$ devlink dev reload pci/0000:00:0b.0
Signed-off-by: Shay Drory <[email protected]>
Reviewed-by: Moshe Shemesh <[email protected]>
---
Documentation/networking/devlink/mlx5.rst | 4 ++++
.../net/ethernet/mellanox/mlx5/core/devlink.c | 14 ++++++++++++++
drivers/net/ethernet/mellanox/mlx5/core/eq.c | 18 +++++++++++++++++-
3 files changed, 35 insertions(+), 1 deletion(-)
diff --git a/Documentation/networking/devlink/mlx5.rst b/Documentation/networking/devlink/mlx5.rst
index 4e4b97f7971a..291e7f63af73 100644
--- a/Documentation/networking/devlink/mlx5.rst
+++ b/Documentation/networking/devlink/mlx5.rst
@@ -14,8 +14,12 @@ Parameters
* - Name
- Mode
+ - Validation
* - ``enable_roce``
- driverinit
+ * - ``io_eq_size``
+ - driverinit
+ - The range is between 64 and 4096.
The ``mlx5`` driver also implements the following driver-specific
parameters.
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/devlink.c b/drivers/net/ethernet/mellanox/mlx5/core/devlink.c
index 1c98652b244a..177c6e9159f8 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/devlink.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/devlink.c
@@ -546,6 +546,13 @@ static int mlx5_devlink_enable_remote_dev_reset_get(struct devlink *devlink, u32
return 0;
}
+static int mlx5_devlink_eq_depth_validate(struct devlink *devlink, u32 id,
+ union devlink_param_value val,
+ struct netlink_ext_ack *extack)
+{
+ return (val.vu16 >= 64 && val.vu16 <= 4096) ? 0 : -EINVAL;
+}
+
static const struct devlink_param mlx5_devlink_params[] = {
DEVLINK_PARAM_DRIVER(MLX5_DEVLINK_PARAM_ID_FLOW_STEERING_MODE,
"flow_steering_mode", DEVLINK_PARAM_TYPE_STRING,
@@ -570,6 +577,8 @@ static const struct devlink_param mlx5_devlink_params[] = {
DEVLINK_PARAM_GENERIC(ENABLE_REMOTE_DEV_RESET, BIT(DEVLINK_PARAM_CMODE_RUNTIME),
mlx5_devlink_enable_remote_dev_reset_get,
mlx5_devlink_enable_remote_dev_reset_set, NULL),
+ DEVLINK_PARAM_GENERIC(IO_EQ_SIZE, BIT(DEVLINK_PARAM_CMODE_DRIVERINIT),
+ NULL, NULL, mlx5_devlink_eq_depth_validate),
};
static void mlx5_devlink_set_params_init_values(struct devlink *devlink)
@@ -608,6 +617,11 @@ static void mlx5_devlink_set_params_init_values(struct devlink *devlink)
value);
}
#endif
+
+ value.vu32 = MLX5_COMP_EQ_SIZE;
+ devlink_param_driverinit_value_set(devlink,
+ DEVLINK_PARAM_GENERIC_ID_IO_EQ_SIZE,
+ value);
}
static const struct devlink_param enable_eth_param =
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eq.c b/drivers/net/ethernet/mellanox/mlx5/core/eq.c
index 792e0d6aa861..7686d7c9c824 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eq.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eq.c
@@ -19,6 +19,7 @@
#include "lib/clock.h"
#include "diag/fw_tracer.h"
#include "mlx5_irq.h"
+#include "devlink.h"
enum {
MLX5_EQE_OWNER_INIT_VAL = 0x1,
@@ -796,6 +797,21 @@ static void destroy_comp_eqs(struct mlx5_core_dev *dev)
}
}
+static u16 comp_eq_depth_devlink_param_get(struct mlx5_core_dev *dev)
+{
+ struct devlink *devlink = priv_to_devlink(dev);
+ union devlink_param_value val;
+ int err;
+
+ err = devlink_param_driverinit_value_get(devlink,
+ DEVLINK_PARAM_GENERIC_ID_IO_EQ_SIZE,
+ &val);
+ if (!err)
+ return val.vu32;
+ mlx5_core_dbg(dev, "Failed to get param. using default. err = %d\n", err);
+ return MLX5_COMP_EQ_SIZE;
+}
+
static int create_comp_eqs(struct mlx5_core_dev *dev)
{
struct mlx5_eq_table *table = dev->priv.eq_table;
@@ -807,7 +823,7 @@ static int create_comp_eqs(struct mlx5_core_dev *dev)
INIT_LIST_HEAD(&table->comp_eqs_list);
ncomp_eqs = table->num_comp_eqs;
- nent = MLX5_COMP_EQ_SIZE;
+ nent = comp_eq_depth_devlink_param_get(dev);
for (i = 0; i < ncomp_eqs; i++) {
struct mlx5_eq_param param = {};
int vecidx = i;
--
2.21.3
Add new device generic parameter to determine the size of the
asynchronous control events EQ.
For example, to reduce event EQ size to 64, execute:
$ devlink dev param set pci/0000:06:00.0 \
name event_eq_size value 64 cmode driverinit
$ devlink dev reload pci/0000:06:00.0
Signed-off-by: Shay Drory <[email protected]>
Reviewed-by: Moshe Shemesh <[email protected]>
---
Documentation/networking/devlink/devlink-params.rst | 3 +++
include/net/devlink.h | 4 ++++
net/core/devlink.c | 5 +++++
3 files changed, 12 insertions(+)
diff --git a/Documentation/networking/devlink/devlink-params.rst b/Documentation/networking/devlink/devlink-params.rst
index cd9342305a13..0eddee6e66f3 100644
--- a/Documentation/networking/devlink/devlink-params.rst
+++ b/Documentation/networking/devlink/devlink-params.rst
@@ -132,3 +132,6 @@ own name.
* - ``io_eq_size``
- u16
- Control the size of I/O completion EQs.
+ * - ``event_eq_size``
+ - u16
+ - Control the size of asynchronous control events EQ.
diff --git a/include/net/devlink.h b/include/net/devlink.h
index b5f4acd0e0cd..8d5349d2fb68 100644
--- a/include/net/devlink.h
+++ b/include/net/devlink.h
@@ -460,6 +460,7 @@ enum devlink_param_generic_id {
DEVLINK_PARAM_GENERIC_ID_ENABLE_VNET,
DEVLINK_PARAM_GENERIC_ID_ENABLE_IWARP,
DEVLINK_PARAM_GENERIC_ID_IO_EQ_SIZE,
+ DEVLINK_PARAM_GENERIC_ID_EVENT_EQ_SIZE,
/* add new param generic ids above here*/
__DEVLINK_PARAM_GENERIC_ID_MAX,
@@ -515,6 +516,9 @@ enum devlink_param_generic_id {
#define DEVLINK_PARAM_GENERIC_IO_EQ_SIZE_NAME "io_eq_size"
#define DEVLINK_PARAM_GENERIC_IO_EQ_SIZE_TYPE DEVLINK_PARAM_TYPE_U32
+#define DEVLINK_PARAM_GENERIC_EVENT_EQ_SIZE_NAME "event_eq_size"
+#define DEVLINK_PARAM_GENERIC_EVENT_EQ_SIZE_TYPE DEVLINK_PARAM_TYPE_U32
+
#define DEVLINK_PARAM_GENERIC(_id, _cmodes, _get, _set, _validate) \
{ \
.id = DEVLINK_PARAM_GENERIC_ID_##_id, \
diff --git a/net/core/devlink.c b/net/core/devlink.c
index 0d4e63d11585..d9f3c994e704 100644
--- a/net/core/devlink.c
+++ b/net/core/devlink.c
@@ -4471,6 +4471,11 @@ static const struct devlink_param devlink_param_generic[] = {
.name = DEVLINK_PARAM_GENERIC_IO_EQ_SIZE_NAME,
.type = DEVLINK_PARAM_GENERIC_IO_EQ_SIZE_TYPE,
},
+ {
+ .id = DEVLINK_PARAM_GENERIC_ID_EVENT_EQ_SIZE,
+ .name = DEVLINK_PARAM_GENERIC_EVENT_EQ_SIZE_NAME,
+ .type = DEVLINK_PARAM_GENERIC_EVENT_EQ_SIZE_TYPE,
+ },
};
static int devlink_param_generic_verify(const struct devlink_param *param)
--
2.21.3
Event EQ is an EQ which received the notification of almost all the
events generated by the NIC.
Currently, each event EQ is taking 512KB of memory. This size is not
needed in most use cases, and is critical with large scale. Hence,
allow user to configure the size of the event EQ.
For example to reduce event EQ size to 64, execute::
$ devlink dev param set pci/0000:00:0b.0 name event_eq_size value 64 \
cmode driverinit
$ devlink dev reload pci/0000:00:0b.0
Signed-off-by: Shay Drory <[email protected]>
Reviewed-by: Moshe Shemesh <[email protected]>
---
Documentation/networking/devlink/mlx5.rst | 3 +++
.../net/ethernet/mellanox/mlx5/core/devlink.c | 7 +++++++
drivers/net/ethernet/mellanox/mlx5/core/eq.c | 16 +++++++++++++++-
3 files changed, 25 insertions(+), 1 deletion(-)
diff --git a/Documentation/networking/devlink/mlx5.rst b/Documentation/networking/devlink/mlx5.rst
index 291e7f63af73..38089f0aefcf 100644
--- a/Documentation/networking/devlink/mlx5.rst
+++ b/Documentation/networking/devlink/mlx5.rst
@@ -20,6 +20,9 @@ Parameters
* - ``io_eq_size``
- driverinit
- The range is between 64 and 4096.
+ * - ``event_eq_size``
+ - driverinit
+ - The range is between 64 and 4096.
The ``mlx5`` driver also implements the following driver-specific
parameters.
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/devlink.c b/drivers/net/ethernet/mellanox/mlx5/core/devlink.c
index 177c6e9159f8..37b7600c5545 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/devlink.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/devlink.c
@@ -579,6 +579,8 @@ static const struct devlink_param mlx5_devlink_params[] = {
mlx5_devlink_enable_remote_dev_reset_set, NULL),
DEVLINK_PARAM_GENERIC(IO_EQ_SIZE, BIT(DEVLINK_PARAM_CMODE_DRIVERINIT),
NULL, NULL, mlx5_devlink_eq_depth_validate),
+ DEVLINK_PARAM_GENERIC(EVENT_EQ_SIZE, BIT(DEVLINK_PARAM_CMODE_DRIVERINIT),
+ NULL, NULL, mlx5_devlink_eq_depth_validate),
};
static void mlx5_devlink_set_params_init_values(struct devlink *devlink)
@@ -622,6 +624,11 @@ static void mlx5_devlink_set_params_init_values(struct devlink *devlink)
devlink_param_driverinit_value_set(devlink,
DEVLINK_PARAM_GENERIC_ID_IO_EQ_SIZE,
value);
+
+ value.vu32 = MLX5_NUM_ASYNC_EQE;
+ devlink_param_driverinit_value_set(devlink,
+ DEVLINK_PARAM_GENERIC_ID_EVENT_EQ_SIZE,
+ value);
}
static const struct devlink_param enable_eth_param =
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eq.c b/drivers/net/ethernet/mellanox/mlx5/core/eq.c
index 7686d7c9c824..b695aad71ee1 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eq.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eq.c
@@ -623,6 +623,20 @@ static void cleanup_async_eq(struct mlx5_core_dev *dev,
name, err);
}
+static u16 async_eq_depth_devlink_param_get(struct mlx5_core_dev *dev)
+{
+ struct devlink *devlink = priv_to_devlink(dev);
+ union devlink_param_value val;
+ int err;
+
+ err = devlink_param_driverinit_value_get(devlink,
+ DEVLINK_PARAM_GENERIC_ID_EVENT_EQ_SIZE,
+ &val);
+ if (!err)
+ return val.vu32;
+ mlx5_core_dbg(dev, "Failed to get param. using default. err = %d\n", err);
+ return MLX5_NUM_ASYNC_EQE;
+}
static int create_async_eqs(struct mlx5_core_dev *dev)
{
struct mlx5_eq_table *table = dev->priv.eq_table;
@@ -647,7 +661,7 @@ static int create_async_eqs(struct mlx5_core_dev *dev)
param = (struct mlx5_eq_param) {
.irq_index = MLX5_IRQ_EQ_CTRL,
- .nent = MLX5_NUM_ASYNC_EQE,
+ .nent = async_eq_depth_devlink_param_get(dev),
};
gather_async_events_mask(dev, param.mask);
--
2.21.3
The generic param max_macs documentation isn't clear.
Replace it with a more descriptive documentation
Signed-off-by: Shay Drory <[email protected]>
Reviewed-by: Jiri Pirko <[email protected]>
---
Documentation/networking/devlink/devlink-params.rst | 6 ++++--
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/Documentation/networking/devlink/devlink-params.rst b/Documentation/networking/devlink/devlink-params.rst
index 0eddee6e66f3..2cbdce4e6a1f 100644
--- a/Documentation/networking/devlink/devlink-params.rst
+++ b/Documentation/networking/devlink/devlink-params.rst
@@ -118,8 +118,10 @@ own name.
errors.
* - ``max_macs``
- u32
- - Specifies the maximum number of MAC addresses per ethernet port of
- this device.
+ - Typically macvlan, vlan net devices mac are also programmed in their
+ parent netdevice's Function rx filter. This parameter limit the
+ maximum number of unicast mac address filters to receive traffic from
+ per ethernet port of this device.
* - ``region_snapshot_enable``
- Boolean
- Enable capture of ``devlink-region`` snapshots.
--
2.21.3
Currently, max_macs is taking 70Kbytes of memory per function. This
size is not needed in all use cases, and is critical with large scale.
Hence, allow user to configure the number of max_macs.
For example, to reduce the number of max_macs to 1, execute::
$ devlink dev param set pci/0000:00:0b.0 name max_macs value 1 \
cmode driverinit
$ devlink dev reload pci/0000:00:0b.0
Signed-off-by: Shay Drory <[email protected]>
Reviewed-by: Moshe Shemesh <[email protected]>
Reviewed-by: Parav Pandit <[email protected]>
---
Documentation/networking/devlink/mlx5.rst | 3 +
.../net/ethernet/mellanox/mlx5/core/devlink.c | 67 +++++++++++++++++++
.../net/ethernet/mellanox/mlx5/core/main.c | 21 ++++++
3 files changed, 91 insertions(+)
diff --git a/Documentation/networking/devlink/mlx5.rst b/Documentation/networking/devlink/mlx5.rst
index 38089f0aefcf..38e94ed65936 100644
--- a/Documentation/networking/devlink/mlx5.rst
+++ b/Documentation/networking/devlink/mlx5.rst
@@ -23,6 +23,9 @@ Parameters
* - ``event_eq_size``
- driverinit
- The range is between 64 and 4096.
+ * - ``max_macs``
+ - driverinit
+ - The range is between 1 and 2^31. Only power of 2 values are supported.
The ``mlx5`` driver also implements the following driver-specific
parameters.
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/devlink.c b/drivers/net/ethernet/mellanox/mlx5/core/devlink.c
index 37b7600c5545..d1093bb2d436 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/devlink.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/devlink.c
@@ -773,6 +773,66 @@ static void mlx5_devlink_auxdev_params_unregister(struct devlink *devlink)
mlx5_devlink_eth_param_unregister(devlink);
}
+static int mlx5_devlink_max_uc_list_validate(struct devlink *devlink, u32 id,
+ union devlink_param_value val,
+ struct netlink_ext_ack *extack)
+{
+ struct mlx5_core_dev *dev = devlink_priv(devlink);
+
+ if (val.vu32 == 0) {
+ NL_SET_ERR_MSG_MOD(extack, "max_macs value must be greater than 0");
+ return -EINVAL;
+ }
+
+ if (!is_power_of_2(val.vu32)) {
+ NL_SET_ERR_MSG_MOD(extack, "Only power of 2 values are supported for max_macs");
+ return -EINVAL;
+ }
+
+ if (ilog2(val.vu32) >
+ MLX5_CAP_GEN_MAX(dev, log_max_current_uc_list)) {
+ NL_SET_ERR_MSG_MOD(extack, "max_macs value is out of the supported range");
+ return -EINVAL;
+ }
+
+ return 0;
+}
+
+static const struct devlink_param max_uc_list_param =
+ DEVLINK_PARAM_GENERIC(MAX_MACS, BIT(DEVLINK_PARAM_CMODE_DRIVERINIT),
+ NULL, NULL, mlx5_devlink_max_uc_list_validate);
+
+static int mlx5_devlink_max_uc_list_param_register(struct devlink *devlink)
+{
+ struct mlx5_core_dev *dev = devlink_priv(devlink);
+ union devlink_param_value value;
+ int err;
+
+ if (!MLX5_CAP_GEN_MAX(dev, log_max_current_uc_list_wr_supported))
+ return 0;
+
+ err = devlink_param_register(devlink, &max_uc_list_param);
+ if (err)
+ return err;
+
+ value.vu32 = 1 << MLX5_CAP_GEN(dev, log_max_current_uc_list);
+ devlink_param_driverinit_value_set(devlink,
+ DEVLINK_PARAM_GENERIC_ID_MAX_MACS,
+ value);
+ return 0;
+}
+
+static void
+mlx5_devlink_max_uc_list_param_unregister(struct devlink *devlink)
+{
+ struct mlx5_core_dev *dev = devlink_priv(devlink);
+
+ if (!MLX5_CAP_GEN_MAX(dev, log_max_current_uc_list_wr_supported))
+ return;
+
+ devlink_param_unregister(devlink, &max_uc_list_param);
+}
+
#define MLX5_TRAP_DROP(_id, _group_id) \
DEVLINK_TRAP_GENERIC(DROP, DROP, _id, \
DEVLINK_TRAP_GROUP_GENERIC_ID_##_group_id, \
@@ -832,6 +892,10 @@ int mlx5_devlink_register(struct devlink *devlink)
if (err)
goto auxdev_reg_err;
+ err = mlx5_devlink_max_uc_list_param_register(devlink);
+ if (err)
+ goto max_uc_list_err;
+
err = mlx5_devlink_traps_register(devlink);
if (err)
goto traps_reg_err;
@@ -842,6 +906,8 @@ int mlx5_devlink_register(struct devlink *devlink)
return 0;
traps_reg_err:
+ mlx5_devlink_max_uc_list_param_unregister(devlink);
+max_uc_list_err:
mlx5_devlink_auxdev_params_unregister(devlink);
auxdev_reg_err:
devlink_params_unregister(devlink, mlx5_devlink_params,
@@ -852,6 +918,7 @@ int mlx5_devlink_register(struct devlink *devlink)
void mlx5_devlink_unregister(struct devlink *devlink)
{
mlx5_devlink_traps_unregister(devlink);
+ mlx5_devlink_max_uc_list_param_unregister(devlink);
mlx5_devlink_auxdev_params_unregister(devlink);
devlink_params_unregister(devlink, mlx5_devlink_params,
ARRAY_SIZE(mlx5_devlink_params));
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/main.c b/drivers/net/ethernet/mellanox/mlx5/core/main.c
index d97c9e86d7b3..b1a82226623c 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/main.c
@@ -484,10 +484,26 @@ static int handle_hca_cap_odp(struct mlx5_core_dev *dev, void *set_ctx)
return set_caps(dev, set_ctx, MLX5_SET_HCA_CAP_OP_MOD_ODP);
}
+static int max_uc_list_get_devlink_param(struct mlx5_core_dev *dev)
+{
+ struct devlink *devlink = priv_to_devlink(dev);
+ union devlink_param_value val;
+ int err;
+
+ err = devlink_param_driverinit_value_get(devlink,
+ DEVLINK_PARAM_GENERIC_ID_MAX_MACS,
+ &val);
+ if (!err)
+ return val.vu32;
+ mlx5_core_dbg(dev, "Failed to get param. err = %d\n", err);
+ return err;
+}
+
static int handle_hca_cap(struct mlx5_core_dev *dev, void *set_ctx)
{
struct mlx5_profile *prof = &dev->profile;
void *set_hca_cap;
+ int max_uc_list;
int err;
err = mlx5_core_get_caps(dev, MLX5_CAP_GENERAL);
@@ -561,6 +577,11 @@ static int handle_hca_cap(struct mlx5_core_dev *dev, void *set_ctx)
if (MLX5_CAP_GEN(dev, roce_rw_supported))
MLX5_SET(cmd_hca_cap, set_hca_cap, roce, mlx5_is_roce_init_enabled(dev));
+ max_uc_list = max_uc_list_get_devlink_param(dev);
+ if (max_uc_list > 0)
+ MLX5_SET(cmd_hca_cap, set_hca_cap, log_max_current_uc_list,
+ ilog2(max_uc_list));
+
return set_caps(dev, set_ctx, MLX5_SET_HCA_CAP_OP_MOD_GENERAL_DEVICE);
}
--
2.21.3
Wed, Dec 08, 2021 at 03:17:17PM CET, [email protected] wrote:
>Add new device generic parameter to determine the size of the
>I/O completion EQs.
>
>For example, to reduce I/O EQ size to 64, execute:
>$ devlink dev param set pci/0000:06:00.0 \
> name io_eq_size value 64 cmode driverinit
>$ devlink dev reload pci/0000:06:00.0
>
>Signed-off-by: Shay Drory <[email protected]>
>Reviewed-by: Moshe Shemesh <[email protected]>
>---
> Documentation/networking/devlink/devlink-params.rst | 3 +++
> include/net/devlink.h | 4 ++++
> net/core/devlink.c | 5 +++++
> 3 files changed, 12 insertions(+)
>
>diff --git a/Documentation/networking/devlink/devlink-params.rst b/Documentation/networking/devlink/devlink-params.rst
>index b7dfe693a332..cd9342305a13 100644
>--- a/Documentation/networking/devlink/devlink-params.rst
>+++ b/Documentation/networking/devlink/devlink-params.rst
>@@ -129,3 +129,6 @@ own name.
> will NACK any attempt of other host to reset the device. This parameter
> is useful for setups where a device is shared by different hosts, such
> as multi-host setup.
>+ * - ``io_eq_size``
>+ - u16
You missed this.
>+ - Control the size of I/O completion EQs.
>diff --git a/include/net/devlink.h b/include/net/devlink.h
>index 3276a29f2b81..b5f4acd0e0cd 100644
>--- a/include/net/devlink.h
>+++ b/include/net/devlink.h
>@@ -459,6 +459,7 @@ enum devlink_param_generic_id {
> DEVLINK_PARAM_GENERIC_ID_ENABLE_RDMA,
> DEVLINK_PARAM_GENERIC_ID_ENABLE_VNET,
> DEVLINK_PARAM_GENERIC_ID_ENABLE_IWARP,
>+ DEVLINK_PARAM_GENERIC_ID_IO_EQ_SIZE,
>
> /* add new param generic ids above here*/
> __DEVLINK_PARAM_GENERIC_ID_MAX,
>@@ -511,6 +512,9 @@ enum devlink_param_generic_id {
> #define DEVLINK_PARAM_GENERIC_ENABLE_IWARP_NAME "enable_iwarp"
> #define DEVLINK_PARAM_GENERIC_ENABLE_IWARP_TYPE DEVLINK_PARAM_TYPE_BOOL
>
>+#define DEVLINK_PARAM_GENERIC_IO_EQ_SIZE_NAME "io_eq_size"
>+#define DEVLINK_PARAM_GENERIC_IO_EQ_SIZE_TYPE DEVLINK_PARAM_TYPE_U32
>+
> #define DEVLINK_PARAM_GENERIC(_id, _cmodes, _get, _set, _validate) \
> { \
> .id = DEVLINK_PARAM_GENERIC_ID_##_id, \
>diff --git a/net/core/devlink.c b/net/core/devlink.c
>index db3b52110cf2..0d4e63d11585 100644
>--- a/net/core/devlink.c
>+++ b/net/core/devlink.c
>@@ -4466,6 +4466,11 @@ static const struct devlink_param devlink_param_generic[] = {
> .name = DEVLINK_PARAM_GENERIC_ENABLE_IWARP_NAME,
> .type = DEVLINK_PARAM_GENERIC_ENABLE_IWARP_TYPE,
> },
>+ {
>+ .id = DEVLINK_PARAM_GENERIC_ID_IO_EQ_SIZE,
>+ .name = DEVLINK_PARAM_GENERIC_IO_EQ_SIZE_NAME,
>+ .type = DEVLINK_PARAM_GENERIC_IO_EQ_SIZE_TYPE,
>+ },
> };
>
> static int devlink_param_generic_verify(const struct devlink_param *param)
>--
>2.21.3
>
Wed, Dec 08, 2021 at 03:17:19PM CET, [email protected] wrote:
>Add new device generic parameter to determine the size of the
>asynchronous control events EQ.
>
>For example, to reduce event EQ size to 64, execute:
>$ devlink dev param set pci/0000:06:00.0 \
> name event_eq_size value 64 cmode driverinit
>$ devlink dev reload pci/0000:06:00.0
>
>Signed-off-by: Shay Drory <[email protected]>
>Reviewed-by: Moshe Shemesh <[email protected]>
>---
> Documentation/networking/devlink/devlink-params.rst | 3 +++
> include/net/devlink.h | 4 ++++
> net/core/devlink.c | 5 +++++
> 3 files changed, 12 insertions(+)
>
>diff --git a/Documentation/networking/devlink/devlink-params.rst b/Documentation/networking/devlink/devlink-params.rst
>index cd9342305a13..0eddee6e66f3 100644
>--- a/Documentation/networking/devlink/devlink-params.rst
>+++ b/Documentation/networking/devlink/devlink-params.rst
>@@ -132,3 +132,6 @@ own name.
> * - ``io_eq_size``
> - u16
> - Control the size of I/O completion EQs.
>+ * - ``event_eq_size``
>+ - u16
And this.
>+ - Control the size of asynchronous control events EQ.
>diff --git a/include/net/devlink.h b/include/net/devlink.h
>index b5f4acd0e0cd..8d5349d2fb68 100644
>--- a/include/net/devlink.h
>+++ b/include/net/devlink.h
>@@ -460,6 +460,7 @@ enum devlink_param_generic_id {
> DEVLINK_PARAM_GENERIC_ID_ENABLE_VNET,
> DEVLINK_PARAM_GENERIC_ID_ENABLE_IWARP,
> DEVLINK_PARAM_GENERIC_ID_IO_EQ_SIZE,
>+ DEVLINK_PARAM_GENERIC_ID_EVENT_EQ_SIZE,
>
> /* add new param generic ids above here*/
> __DEVLINK_PARAM_GENERIC_ID_MAX,
>@@ -515,6 +516,9 @@ enum devlink_param_generic_id {
> #define DEVLINK_PARAM_GENERIC_IO_EQ_SIZE_NAME "io_eq_size"
> #define DEVLINK_PARAM_GENERIC_IO_EQ_SIZE_TYPE DEVLINK_PARAM_TYPE_U32
>
>+#define DEVLINK_PARAM_GENERIC_EVENT_EQ_SIZE_NAME "event_eq_size"
>+#define DEVLINK_PARAM_GENERIC_EVENT_EQ_SIZE_TYPE DEVLINK_PARAM_TYPE_U32
>+
> #define DEVLINK_PARAM_GENERIC(_id, _cmodes, _get, _set, _validate) \
> { \
> .id = DEVLINK_PARAM_GENERIC_ID_##_id, \
>diff --git a/net/core/devlink.c b/net/core/devlink.c
>index 0d4e63d11585..d9f3c994e704 100644
>--- a/net/core/devlink.c
>+++ b/net/core/devlink.c
>@@ -4471,6 +4471,11 @@ static const struct devlink_param devlink_param_generic[] = {
> .name = DEVLINK_PARAM_GENERIC_IO_EQ_SIZE_NAME,
> .type = DEVLINK_PARAM_GENERIC_IO_EQ_SIZE_TYPE,
> },
>+ {
>+ .id = DEVLINK_PARAM_GENERIC_ID_EVENT_EQ_SIZE,
>+ .name = DEVLINK_PARAM_GENERIC_EVENT_EQ_SIZE_NAME,
>+ .type = DEVLINK_PARAM_GENERIC_EVENT_EQ_SIZE_TYPE,
>+ },
> };
>
> static int devlink_param_generic_verify(const struct devlink_param *param)
>--
>2.21.3
>
On Wed, 2021-12-08 at 16:17 +0200, Shay Drory wrote:
> This series provides knobs which will enable users to
> minimize memory consumption of mlx5 Functions (PF/VF/SF).
> mlx5 exposes two new generic devlink params for EQ size
> configuration and uses devlink generic param max_macs.
>
> Patches summary:
> - Patch-1 Introduce log_max_current_uc_list_wr_supported bit
> - Patches-2-3 Provides I/O EQ size param which enables to save
> up to 128KB.
> - Patches-4-5 Provides event EQ size param which enables to save
> up to 512KB.
> - Patch-6 Clarify max_macs param.
> - Patch-7 Provides max_macs param which enables to save up to 70KB
>
> In total, this series can save up to 700KB per Function.
>
> ---
> changelog:
> v2->v3:
> - change type of EQ size param to u32 per Jiri suggestion.
> - separate ifc changes to new patch
> v1->v2:
> - convert io_eq_size and event_eq_size from devlink_resources to
> generic devlink_params
Jakub are ok with this version ?
I would like to take it to my trees.
On Mon, 13 Dec 2021 23:06:26 +0000 Saeed Mahameed wrote:
> On Wed, 2021-12-08 at 16:17 +0200, Shay Drory wrote:
> > This series provides knobs which will enable users to
> > minimize memory consumption of mlx5 Functions (PF/VF/SF).
> > mlx5 exposes two new generic devlink params for EQ size
> > configuration and uses devlink generic param max_macs.
> >
> > Patches summary:
> > - Patch-1 Introduce log_max_current_uc_list_wr_supported bit
> > - Patches-2-3 Provides I/O EQ size param which enables to save
> > up to 128KB.
> > - Patches-4-5 Provides event EQ size param which enables to save
> > up to 512KB.
> > - Patch-6 Clarify max_macs param.
> > - Patch-7 Provides max_macs param which enables to save up to 70KB
> >
> > In total, this series can save up to 700KB per Function.
>
> Jakub are ok with this version ?
> I would like to take it to my trees.
Yup, go for it. Sorry I didn't ack - I was going to apply it today but
Dave reminded me that you probably want to take it thru your tree.
On Mon, 2021-12-13 at 17:01 -0800, Jakub Kicinski wrote:
> On Mon, 13 Dec 2021 23:06:26 +0000 Saeed Mahameed wrote:
> > On Wed, 2021-12-08 at 16:17 +0200, Shay Drory wrote:
> > > This series provides knobs which will enable users to
> > > minimize memory consumption of mlx5 Functions (PF/VF/SF).
> > > mlx5 exposes two new generic devlink params for EQ size
> > > configuration and uses devlink generic param max_macs.
> > >
> > > Patches summary:
> > > - Patch-1 Introduce log_max_current_uc_list_wr_supported bit
> > > - Patches-2-3 Provides I/O EQ size param which enables to save
> > > up to 128KB.
> > > - Patches-4-5 Provides event EQ size param which enables to save
> > > up to 512KB.
> > > - Patch-6 Clarify max_macs param.
> > > - Patch-7 Provides max_macs param which enables to save up to
> > > 70KB
> > >
> > > In total, this series can save up to 700KB per Function.
> >
> > Jakub are ok with this version ?
> > I would like to take it to my trees.
>
> Yup, go for it. Sorry I didn't ack - I was going to apply it today
> but
> Dave reminded me that you probably want to take it thru your tree.
Thanks ! applied. Will be sent in future PRs.
First patch needs to go to mlx5-next shared branch with rdma.
So this will be sent in two separate PRs.