[dpdk-dev] [PATCH v2 11/38] net/sfc: add port representors infrastructure

[Andrew Rybchenko]

From: Igor Romanov 

Provide minimal implementation for port representors that only can be
configured and can provide device information.

Signed-off-by: Igor Romanov 
Signed-off-by: Andrew Rybchenko 
Reviewed-by: Andy Moreton 
Reviewed-by: Ivan Malov 
---
 doc/guides/nics/sfc_efx.rst  |  13 +-
 drivers/net/sfc/meson.build  |   1 +
 drivers/net/sfc/sfc_ethdev.c | 153 +++-
 drivers/net/sfc/sfc_kvargs.c |   1 +
 drivers/net/sfc/sfc_kvargs.h |   2 +
 drivers/net/sfc/sfc_repr.c   | 458 +++
 drivers/net/sfc/sfc_repr.h   |  36 +++
 drivers/net/sfc/sfc_switch.h |   5 +
 8 files changed, 662 insertions(+), 7 deletions(-)
 create mode 100644 drivers/net/sfc/sfc_repr.c
 create mode 100644 drivers/net/sfc/sfc_repr.h

diff --git a/doc/guides/nics/sfc_efx.rst b/doc/guides/nics/sfc_efx.rst
index d66cb76dab..4719031508 100644
--- a/doc/guides/nics/sfc_efx.rst
+++ b/doc/guides/nics/sfc_efx.rst
@@ -74,6 +74,8 @@ SFC EFX PMD has support for:
 
 - SR-IOV PF
 
+- Port representors (see :ref: switch_representation)
+
 
 Non-supported Features
 --
@@ -382,7 +384,16 @@ boolean parameters value.
   software virtual switch (for example, Open vSwitch) makes the decision.
   Software virtual switch may install MAE rules to pass established traffic
   flows via hardware and offload software datapath as the result.
-  Default is legacy.
+  Default is legacy, unless representors are specified, in which case switchdev
+  is chosen.
+
+- ``representor`` parameter [list]
+
+  Instantiate port representor Ethernet devices for specified Virtual
+  Functions list.
+
+  It is a standard parameter whose format is described in
+  :ref:`ethernet_device_standard_device_arguments`.
 
 - ``rx_datapath`` [auto|efx|ef10|ef10_essb] (default **auto**)
 
diff --git a/drivers/net/sfc/meson.build b/drivers/net/sfc/meson.build
index 4fc2063f7a..98365e9e73 100644
--- a/drivers/net/sfc/meson.build
+++ b/drivers/net/sfc/meson.build
@@ -98,4 +98,5 @@ sources = files(
 'sfc_ef100_tx.c',
 'sfc_service.c',
 'sfc_repr_proxy.c',
+'sfc_repr.c',
 )
diff --git a/drivers/net/sfc/sfc_ethdev.c b/drivers/net/sfc/sfc_ethdev.c
index 1e9a31c937..e270b5cefd 100644
--- a/drivers/net/sfc/sfc_ethdev.c
+++ b/drivers/net/sfc/sfc_ethdev.c
@@ -28,6 +28,7 @@
 #include "sfc_flow.h"
 #include "sfc_dp.h"
 #include "sfc_dp_rx.h"
+#include "sfc_repr.h"
 #include "sfc_sw_stats.h"
 
 #define SFC_XSTAT_ID_INVALID_VAL  UINT64_MAX
@@ -1909,6 +1910,10 @@ static const struct eth_dev_ops sfc_eth_dev_ops = {
.pool_ops_supported = sfc_pool_ops_supported,
 };
 
+struct sfc_ethdev_init_data {
+   uint16_tnb_representors;
+};
+
 /**
  * Duplicate a string in potentially shared memory required for
  * multi-process support.
@@ -2190,7 +2195,7 @@ sfc_register_dp(void)
 }
 
 static int
-sfc_parse_switch_mode(struct sfc_adapter *sa)
+sfc_parse_switch_mode(struct sfc_adapter *sa, bool has_representors)
 {
const efx_nic_cfg_t *encp = efx_nic_cfg_get(sa->nic);
const char *switch_mode = NULL;
@@ -2205,7 +2210,8 @@ sfc_parse_switch_mode(struct sfc_adapter *sa)
 
if (switch_mode == NULL) {
sa->switchdev = encp->enc_mae_supported &&
-   !encp->enc_datapath_cap_evb;
+   (!encp->enc_datapath_cap_evb ||
+has_representors);
} else if (strcasecmp(switch_mode, SFC_KVARG_SWITCH_MODE_LEGACY) == 0) {
sa->switchdev = false;
} else if (strcasecmp(switch_mode,
@@ -2230,10 +2236,11 @@ sfc_parse_switch_mode(struct sfc_adapter *sa)
 }
 
 static int
-sfc_eth_dev_init(struct rte_eth_dev *dev)
+sfc_eth_dev_init(struct rte_eth_dev *dev, void *init_params)
 {
struct sfc_adapter_shared *sas = sfc_adapter_shared_by_eth_dev(dev);
struct rte_pci_device *pci_dev = RTE_ETH_DEV_TO_PCI(dev);
+   struct sfc_ethdev_init_data *init_data = init_params;
uint32_t logtype_main;
struct sfc_adapter *sa;
int rc;
@@ -2324,7 +2331,7 @@ sfc_eth_dev_init(struct rte_eth_dev *dev)
 * Selecting a default switch mode requires the NIC to be probed and
 * to have its capabilities filled in.
 */
-   rc = sfc_parse_switch_mode(sa);
+   rc = sfc_parse_switch_mode(sa, init_data->nb_representors > 0);
if (rc != 0)
goto fail_switch_mode;
 
@@ -2409,11 +2416,145 @@ static const struct rte_pci_id pci_id_sfc_efx_map[] = {
{ .vendor_id = 0 /* sentinel */ }
 };
 
+static int
+sfc_parse_rte_devargs(const char *args, struct rte_eth_devargs *devargs)
+{
+   struct rte_eth_devargs eth_da = { .nb_representor_ports = 0 };
+   int rc;
+
+   if (args != NULL) {
+   rc = rte_eth_devargs_parse(args, ð_da);
+   if (rc != 0) {
+   SFC_GENERIC_LOG(ERR,
+   "Failed to parse generic devargs '%s'",
+   

[dpdk-dev] [PATCH v2 13/38] common/sfc_efx/base: add API to get mport selector by ID

[Andrew Rybchenko]

From: Igor Romanov 

The conversion is required when mport ID is received via
mport allocation and mport selector is required for filter
creation.

Signed-off-by: Igor Romanov 
Signed-off-by: Andrew Rybchenko 
Reviewed-by: Andy Moreton 
Reviewed-by: Ivan Malov 
---
 drivers/common/sfc_efx/base/efx.h | 13 +
 drivers/common/sfc_efx/base/efx_mae.c | 17 +
 drivers/common/sfc_efx/version.map|  1 +
 3 files changed, 31 insertions(+)

diff --git a/drivers/common/sfc_efx/base/efx.h 
b/drivers/common/sfc_efx/base/efx.h
index 7f04b42bae..a59c2e47ef 100644
--- a/drivers/common/sfc_efx/base/efx.h
+++ b/drivers/common/sfc_efx/base/efx.h
@@ -4237,6 +4237,19 @@ efx_mae_mport_by_pcie_function(
__inuint32_t vf,
__out   efx_mport_sel_t *mportp);
 
+/*
+ * Get MPORT selector by an MPORT ID
+ *
+ * The resulting MPORT selector is opaque to the caller and can be
+ * passed as an argument to efx_mae_match_spec_mport_set()
+ * and efx_mae_action_set_populate_deliver().
+ */
+LIBEFX_API
+extern __checkReturn   efx_rc_t
+efx_mae_mport_by_id(
+   __inconst efx_mport_id_t *mport_idp,
+   __out   efx_mport_sel_t *mportp);
+
 /* Get MPORT ID by an MPORT selector */
 LIBEFX_API
 extern __checkReturn   efx_rc_t
diff --git a/drivers/common/sfc_efx/base/efx_mae.c 
b/drivers/common/sfc_efx/base/efx_mae.c
index b7afe8fdc8..f5d981f973 100644
--- a/drivers/common/sfc_efx/base/efx_mae.c
+++ b/drivers/common/sfc_efx/base/efx_mae.c
@@ -827,6 +827,23 @@ efx_mae_mport_id_by_selector(
return (rc);
 }
 
+   __checkReturn   efx_rc_t
+efx_mae_mport_by_id(
+   __inconst efx_mport_id_t *mport_idp,
+   __out   efx_mport_sel_t *mportp)
+{
+   efx_dword_t dword;
+
+   EFX_POPULATE_DWORD_2(dword,
+   MAE_MPORT_SELECTOR_TYPE, MAE_MPORT_SELECTOR_TYPE_MPORT_ID,
+   MAE_MPORT_SELECTOR_MPORT_ID, mport_idp->id);
+
+   memset(mportp, 0, sizeof (*mportp));
+   mportp->sel = __LE_TO_CPU_32(dword.ed_u32[0]);
+
+   return (0);
+}
+
__checkReturn   efx_rc_t
 efx_mae_match_spec_field_set(
__inefx_mae_match_spec_t *spec,
diff --git a/drivers/common/sfc_efx/version.map 
b/drivers/common/sfc_efx/version.map
index 611757ccde..8c5d813c19 100644
--- a/drivers/common/sfc_efx/version.map
+++ b/drivers/common/sfc_efx/version.map
@@ -126,6 +126,7 @@ INTERNAL {
efx_mae_match_specs_equal;
efx_mae_mport_by_pcie_function;
efx_mae_mport_by_phy_port;
+   efx_mae_mport_by_id;
efx_mae_mport_id_by_selector;
efx_mae_mport_invalid;
efx_mae_outer_rule_insert;
-- 
2.30.2

[dpdk-dev] [PATCH v2 14/38] common/sfc_efx/base: add mport alias MCDI wrappers

[Andrew Rybchenko]

From: Igor Romanov 

The APIs allow creation of mports for port representor
traffic filtering.

Signed-off-by: Igor Romanov 
Signed-off-by: Andrew Rybchenko 
Reviewed-by: Andy Moreton 
Reviewed-by: Ivan Malov 
---
 drivers/common/sfc_efx/base/efx.h | 13 
 drivers/common/sfc_efx/base/efx_mae.c | 90 +++
 drivers/common/sfc_efx/version.map|  2 +
 3 files changed, 105 insertions(+)

diff --git a/drivers/common/sfc_efx/base/efx.h 
b/drivers/common/sfc_efx/base/efx.h
index a59c2e47ef..0a178128ba 100644
--- a/drivers/common/sfc_efx/base/efx.h
+++ b/drivers/common/sfc_efx/base/efx.h
@@ -4599,6 +4599,19 @@ efx_mae_action_rule_remove(
__inefx_nic_t *enp,
__inconst efx_mae_rule_id_t *ar_idp);
 
+LIBEFX_API
+extern __checkReturn   efx_rc_t
+efx_mcdi_mport_alloc_alias(
+   __inefx_nic_t *enp,
+   __out   efx_mport_id_t *mportp,
+   __out_opt   uint32_t *labelp);
+
+LIBEFX_API
+extern __checkReturn   efx_rc_t
+efx_mae_mport_free(
+   __inefx_nic_t *enp,
+   __inconst efx_mport_id_t *mportp);
+
 #endif /* EFSYS_OPT_MAE */
 
 #if EFSYS_OPT_VIRTIO
diff --git a/drivers/common/sfc_efx/base/efx_mae.c 
b/drivers/common/sfc_efx/base/efx_mae.c
index f5d981f973..3f498fe189 100644
--- a/drivers/common/sfc_efx/base/efx_mae.c
+++ b/drivers/common/sfc_efx/base/efx_mae.c
@@ -3142,4 +3142,94 @@ efx_mae_action_rule_remove(
return (rc);
 }
 
+   __checkReturn   efx_rc_t
+efx_mcdi_mport_alloc_alias(
+   __inefx_nic_t *enp,
+   __out   efx_mport_id_t *mportp,
+   __out_opt   uint32_t *labelp)
+{
+   const efx_nic_cfg_t *encp = efx_nic_cfg_get(enp);
+   efx_mcdi_req_t req;
+   EFX_MCDI_DECLARE_BUF(payload,
+   MC_CMD_MAE_MPORT_ALLOC_ALIAS_IN_LEN,
+   MC_CMD_MAE_MPORT_ALLOC_ALIAS_OUT_LEN);
+   efx_rc_t rc;
+
+   if (encp->enc_mae_supported == B_FALSE) {
+   rc = ENOTSUP;
+   goto fail1;
+   }
+
+   req.emr_cmd = MC_CMD_MAE_MPORT_ALLOC;
+   req.emr_in_buf = payload;
+   req.emr_in_length = MC_CMD_MAE_MPORT_ALLOC_ALIAS_IN_LEN;
+   req.emr_out_buf = payload;
+   req.emr_out_length = MC_CMD_MAE_MPORT_ALLOC_ALIAS_OUT_LEN;
+
+   MCDI_IN_SET_DWORD(req, MAE_MPORT_ALLOC_IN_TYPE,
+ MC_CMD_MAE_MPORT_ALLOC_IN_MPORT_TYPE_ALIAS);
+   MCDI_IN_SET_DWORD(req, MAE_MPORT_ALLOC_ALIAS_IN_DELIVER_MPORT,
+ MAE_MPORT_SELECTOR_ASSIGNED);
+
+   efx_mcdi_execute(enp, &req);
+
+   if (req.emr_rc != 0) {
+   rc = req.emr_rc;
+   goto fail2;
+   }
+
+   mportp->id = MCDI_OUT_DWORD(req, MAE_MPORT_ALLOC_OUT_MPORT_ID);
+   if (labelp != NULL)
+   *labelp = MCDI_OUT_DWORD(req, MAE_MPORT_ALLOC_ALIAS_OUT_LABEL);
+
+   return (0);
+
+fail2:
+   EFSYS_PROBE(fail2);
+fail1:
+   EFSYS_PROBE1(fail1, efx_rc_t, rc);
+   return (rc);
+}
+
+   __checkReturn   efx_rc_t
+efx_mae_mport_free(
+   __inefx_nic_t *enp,
+   __inconst efx_mport_id_t *mportp)
+{
+   const efx_nic_cfg_t *encp = efx_nic_cfg_get(enp);
+   efx_mcdi_req_t req;
+   EFX_MCDI_DECLARE_BUF(payload,
+   MC_CMD_MAE_MPORT_FREE_IN_LEN,
+   MC_CMD_MAE_MPORT_FREE_OUT_LEN);
+   efx_rc_t rc;
+
+   if (encp->enc_mae_supported == B_FALSE) {
+   rc = ENOTSUP;
+   goto fail1;
+   }
+
+   req.emr_cmd = MC_CMD_MAE_MPORT_FREE;
+   req.emr_in_buf = payload;
+   req.emr_in_length = MC_CMD_MAE_MPORT_FREE_IN_LEN;
+   req.emr_out_buf = payload;
+   req.emr_out_length = MC_CMD_MAE_MPORT_FREE_OUT_LEN;
+
+   MCDI_IN_SET_DWORD(req, MAE_MPORT_FREE_IN_MPORT_ID, mportp->id);
+
+   efx_mcdi_execute(enp, &req);
+
+   if (req.emr_rc != 0) {
+   rc = req.emr_rc;
+   goto fail2;
+   }
+
+   return (0);
+
+fail2:
+   EFSYS_PROBE(fail2);
+fail1:
+   EFSYS_PROBE1(fail1, efx_rc_t, rc);
+   return (rc);
+}
+
 #endif /* EFSYS_OPT_MAE */
diff --git a/drivers/common/sfc_efx/version.map 
b/drivers/common/sfc_efx/version.map
index 8c5d813c19..3488367f68 100644
--- a/drivers/common/sfc_efx/version.map
+++ b/drivers/common/sfc_efx/version.map
@@ -127,6 +127,7 @@ INTERNAL {
efx_mae_mport_by_pcie_function;
efx_mae_mport_by_phy_port;
efx_mae_mport_by_id;
+   efx_mae_mport_free;
efx_mae_mport_id_by_selector;
efx_mae_mport_invalid;
efx_mae_outer_rule_insert;
@@ -136,6 +137,7 @@ INTERNAL {
efx_mcdi_get_proxy_handle;
efx_mcdi_get_timeout;
efx_mcdi_init;
+   efx_mcdi_mport_a

[dpdk-dev] [PATCH v2 15/38] net/sfc: add representor proxy port API

[Andrew Rybchenko]

From: Igor Romanov 

The API is required to create and destroy representor proxy
port assigned to representor.

Signed-off-by: Igor Romanov 
Signed-off-by: Andrew Rybchenko 
Reviewed-by: Andy Moreton 
Reviewed-by: Ivan Malov 
---
 drivers/net/sfc/sfc.c|  12 +
 drivers/net/sfc/sfc.h|   1 +
 drivers/net/sfc/sfc_ethdev.c |   2 +
 drivers/net/sfc/sfc_repr.c   |  20 ++
 drivers/net/sfc/sfc_repr_proxy.c | 320 ++-
 drivers/net/sfc/sfc_repr_proxy.h |  30 +++
 drivers/net/sfc/sfc_repr_proxy_api.h |  29 +++
 7 files changed, 412 insertions(+), 2 deletions(-)
 create mode 100644 drivers/net/sfc/sfc_repr_proxy_api.h

diff --git a/drivers/net/sfc/sfc.c b/drivers/net/sfc/sfc.c
index 152234cb61..f79f4d5ffc 100644
--- a/drivers/net/sfc/sfc.c
+++ b/drivers/net/sfc/sfc.c
@@ -1043,6 +1043,18 @@ sfc_attach(struct sfc_adapter *sa)
return rc;
 }
 
+void
+sfc_pre_detach(struct sfc_adapter *sa)
+{
+   sfc_log_init(sa, "entry");
+
+   SFC_ASSERT(!sfc_adapter_is_locked(sa));
+
+   sfc_repr_proxy_pre_detach(sa);
+
+   sfc_log_init(sa, "done");
+}
+
 void
 sfc_detach(struct sfc_adapter *sa)
 {
diff --git a/drivers/net/sfc/sfc.h b/drivers/net/sfc/sfc.h
index 628f32c13f..c3e92f3ab6 100644
--- a/drivers/net/sfc/sfc.h
+++ b/drivers/net/sfc/sfc.h
@@ -376,6 +376,7 @@ uint32_t sfc_register_logtype(const struct rte_pci_addr 
*pci_addr,
 int sfc_probe(struct sfc_adapter *sa);
 void sfc_unprobe(struct sfc_adapter *sa);
 int sfc_attach(struct sfc_adapter *sa);
+void sfc_pre_detach(struct sfc_adapter *sa);
 void sfc_detach(struct sfc_adapter *sa);
 int sfc_start(struct sfc_adapter *sa);
 void sfc_stop(struct sfc_adapter *sa);
diff --git a/drivers/net/sfc/sfc_ethdev.c b/drivers/net/sfc/sfc_ethdev.c
index e270b5cefd..efd5e6b1ab 100644
--- a/drivers/net/sfc/sfc_ethdev.c
+++ b/drivers/net/sfc/sfc_ethdev.c
@@ -345,6 +345,8 @@ sfc_dev_close(struct rte_eth_dev *dev)
return 0;
}
 
+   sfc_pre_detach(sa);
+
sfc_adapter_lock(sa);
switch (sa->state) {
case SFC_ETHDEV_STARTED:
diff --git a/drivers/net/sfc/sfc_repr.c b/drivers/net/sfc/sfc_repr.c
index 71eea0e209..ff5ea0d1ed 100644
--- a/drivers/net/sfc/sfc_repr.c
+++ b/drivers/net/sfc/sfc_repr.c
@@ -19,6 +19,7 @@
 #include "sfc_debug.h"
 #include "sfc_repr.h"
 #include "sfc_ethdev_state.h"
+#include "sfc_repr_proxy_api.h"
 #include "sfc_switch.h"
 
 /** Multi-process shared representor private data */
@@ -285,6 +286,7 @@ static int
 sfc_repr_dev_close(struct rte_eth_dev *dev)
 {
struct sfc_repr *sr = sfc_repr_by_eth_dev(dev);
+   struct sfc_repr_shared *srs = sfc_repr_shared_by_eth_dev(dev);
 
sfcr_info(sr, "entry");
 
@@ -306,6 +308,8 @@ sfc_repr_dev_close(struct rte_eth_dev *dev)
 * Rollback primary process sfc_repr_eth_dev_init() below.
 */
 
+   (void)sfc_repr_proxy_del_port(srs->pf_port_id, srs->repr_id);
+
dev->dev_ops = NULL;
 
sfc_repr_unlock(sr);
@@ -378,6 +382,18 @@ sfc_repr_eth_dev_init(struct rte_eth_dev *dev, void 
*init_params)
goto fail_mae_assign_switch_port;
}
 
+   ret = sfc_repr_proxy_add_port(repr_data->pf_port_id,
+ repr_data->repr_id,
+ dev->data->port_id,
+ &repr_data->mport_sel);
+   if (ret != 0) {
+   SFC_GENERIC_LOG(ERR, "%s() failed to add repr proxy port",
+   __func__);
+   SFC_ASSERT(ret > 0);
+   ret = -ret;
+   goto fail_create_port;
+   }
+
/*
 * Allocate process private data from heap, since it should not
 * be located in shared memory allocated using rte_malloc() API.
@@ -419,6 +435,10 @@ sfc_repr_eth_dev_init(struct rte_eth_dev *dev, void 
*init_params)
free(sr);
 
 fail_alloc_sr:
+   (void)sfc_repr_proxy_del_port(repr_data->pf_port_id,
+ repr_data->repr_id);
+
+fail_create_port:
 fail_mae_assign_switch_port:
SFC_GENERIC_LOG(ERR, "%s() failed: %s", __func__, rte_strerror(-ret));
return ret;
diff --git a/drivers/net/sfc/sfc_repr_proxy.c b/drivers/net/sfc/sfc_repr_proxy.c
index 6d3962304f..f64fa2efc7 100644
--- a/drivers/net/sfc/sfc_repr_proxy.c
+++ b/drivers/net/sfc/sfc_repr_proxy.c
@@ -13,17 +13,191 @@
 #include "sfc_log.h"
 #include "sfc_service.h"
 #include "sfc_repr_proxy.h"
+#include "sfc_repr_proxy_api.h"
 #include "sfc.h"
 
+/**
+ * Amount of time to wait for the representor proxy routine (which is
+ * running on a service core) to handle a request sent via mbox.
+ */
+#define SFC_REPR_PROXY_MBOX_POLL_TIMEOUT_MS1000
+
+static struct sfc_repr_proxy *
+sfc_repr_proxy_by_adapter(struct sfc_adapter *sa)
+{
+   return &sa->repr_proxy;
+}
+
+static struct sfc_adapter *
+sfc_get_adapter_by_pf_port_id(uint16_t pf_port_id)
+{
+   struct rte_eth_dev *dev

[dpdk-dev] [PATCH v2 16/38] net/sfc: implement representor queue setup and release

[Andrew Rybchenko]

From: Igor Romanov 

Implement queue creation and destruction both in port representors
and representor proxy.

Signed-off-by: Igor Romanov 
Signed-off-by: Andrew Rybchenko 
Reviewed-by: Andy Moreton 
Reviewed-by: Ivan Malov 
---
 drivers/net/sfc/sfc_repr.c   | 257 +++
 drivers/net/sfc/sfc_repr_proxy.c | 132 ++
 drivers/net/sfc/sfc_repr_proxy.h |  22 +++
 drivers/net/sfc/sfc_repr_proxy_api.h |  15 ++
 4 files changed, 426 insertions(+)

diff --git a/drivers/net/sfc/sfc_repr.c b/drivers/net/sfc/sfc_repr.c
index ff5ea0d1ed..ddd848466c 100644
--- a/drivers/net/sfc/sfc_repr.c
+++ b/drivers/net/sfc/sfc_repr.c
@@ -30,6 +30,17 @@ struct sfc_repr_shared {
uint16_tswitch_port_id;
 };
 
+struct sfc_repr_rxq {
+   /* Datapath members */
+   struct rte_ring *ring;
+};
+
+struct sfc_repr_txq {
+   /* Datapath members */
+   struct rte_ring *ring;
+   efx_mport_id_t  egress_mport;
+};
+
 /** Primary process representor private data */
 struct sfc_repr {
/**
@@ -50,6 +61,14 @@ struct sfc_repr {
SFC_GENERIC_LOG(ERR, __VA_ARGS__);  \
} while (0)
 
+#define sfcr_warn(sr, ...) \
+   do {\
+   const struct sfc_repr *_sr = (sr);  \
+   \
+   (void)_sr;  \
+   SFC_GENERIC_LOG(WARNING, __VA_ARGS__);  \
+   } while (0)
+
 #define sfcr_info(sr, ...) \
do {\
const struct sfc_repr *_sr = (sr);  \
@@ -269,6 +288,229 @@ sfc_repr_dev_infos_get(struct rte_eth_dev *dev,
return 0;
 }
 
+static int
+sfc_repr_ring_create(uint16_t pf_port_id, uint16_t repr_id,
+const char *type_name, uint16_t qid, uint16_t nb_desc,
+unsigned int socket_id, struct rte_ring **ring)
+{
+   char ring_name[RTE_RING_NAMESIZE];
+   int ret;
+
+   ret = snprintf(ring_name, sizeof(ring_name), "sfc_%u_repr_%u_%sq%u",
+  pf_port_id, repr_id, type_name, qid);
+   if (ret >= (int)sizeof(ring_name))
+   return -ENAMETOOLONG;
+
+   /*
+* Single producer/consumer rings are used since the API for Tx/Rx
+* packet burst for representors are guaranteed to be called from
+* a single thread, and the user of the other end (representor proxy)
+* is also single-threaded.
+*/
+   *ring = rte_ring_create(ring_name, nb_desc, socket_id,
+  RING_F_SP_ENQ | RING_F_SC_DEQ);
+   if (*ring == NULL)
+   return -rte_errno;
+
+   return 0;
+}
+
+static int
+sfc_repr_rx_qcheck_conf(struct sfc_repr *sr,
+   const struct rte_eth_rxconf *rx_conf)
+{
+   int ret = 0;
+
+   sfcr_info(sr, "entry");
+
+   if (rx_conf->rx_thresh.pthresh != 0 ||
+   rx_conf->rx_thresh.hthresh != 0 ||
+   rx_conf->rx_thresh.wthresh != 0) {
+   sfcr_warn(sr,
+   "RxQ prefetch/host/writeback thresholds are not 
supported");
+   }
+
+   if (rx_conf->rx_free_thresh != 0)
+   sfcr_warn(sr, "RxQ free threshold is not supported");
+
+   if (rx_conf->rx_drop_en == 0)
+   sfcr_warn(sr, "RxQ drop disable is not supported");
+
+   if (rx_conf->rx_deferred_start) {
+   sfcr_err(sr, "Deferred start is not supported");
+   ret = -EINVAL;
+   }
+
+   sfcr_info(sr, "done: %s", rte_strerror(-ret));
+
+   return ret;
+}
+
+static int
+sfc_repr_rx_queue_setup(struct rte_eth_dev *dev, uint16_t rx_queue_id,
+   uint16_t nb_rx_desc, unsigned int socket_id,
+   __rte_unused const struct rte_eth_rxconf *rx_conf,
+   struct rte_mempool *mb_pool)
+{
+   struct sfc_repr_shared *srs = sfc_repr_shared_by_eth_dev(dev);
+   struct sfc_repr *sr = sfc_repr_by_eth_dev(dev);
+   struct sfc_repr_rxq *rxq;
+   int ret;
+
+   sfcr_info(sr, "entry");
+
+   ret = sfc_repr_rx_qcheck_conf(sr, rx_conf);
+   if (ret != 0)
+   goto fail_check_conf;
+
+   ret = -ENOMEM;
+   rxq = rte_zmalloc_socket("sfc-repr-rxq", sizeof(*rxq),
+RTE_CACHE_LINE_SIZE, socket_id);
+   if (rxq == NULL) {
+   sfcr_err(sr, "%s() failed to alloc RxQ", __func__);
+   goto fail_rxq_alloc;
+   }
+
+   ret = sfc_repr_ring_create(srs->pf_port_id, srs->repr_id,
+  "rx", rx_queue_id, nb_rx_desc,
+  socket_id, &rxq->ring);
+   if (ret != 0) {
+   sfcr_err(sr, "%s()

[dpdk-dev] [PATCH v2 17/38] net/sfc: implement representor RxQ start/stop

[Andrew Rybchenko]

From: Igor Romanov 

Add extra libefx flags to Rx queue information initialization
function interface to be able to specify the ingress m-port
flag for a representor RxQ. Rx prefix of packets on that queue
will contain ingress m-port field required for packet forwarding
in representor proxy.

Signed-off-by: Igor Romanov 
Signed-off-by: Andrew Rybchenko 
Reviewed-by: Andy Moreton 
Reviewed-by: Ivan Malov 
---
 drivers/net/sfc/sfc_ev.h |   8 ++
 drivers/net/sfc/sfc_repr_proxy.c | 194 +++
 drivers/net/sfc/sfc_repr_proxy.h |   7 ++
 3 files changed, 209 insertions(+)

diff --git a/drivers/net/sfc/sfc_ev.h b/drivers/net/sfc/sfc_ev.h
index 590cfb1694..bcb7fbe466 100644
--- a/drivers/net/sfc/sfc_ev.h
+++ b/drivers/net/sfc/sfc_ev.h
@@ -110,6 +110,14 @@ sfc_counters_rxq_sw_index(const struct sfc_adapter_shared 
*sas)
return sas->counters_rxq_allocated ? 0 : SFC_SW_INDEX_INVALID;
 }
 
+static inline sfc_sw_index_t
+sfc_repr_rxq_sw_index(const struct sfc_adapter_shared *sas,
+ unsigned int repr_queue_id)
+{
+   return sfc_counters_rxq_sw_index(sas) + sfc_repr_nb_rxq(sas) +
+   repr_queue_id;
+}
+
 /*
  * Functions below define event queue to transmit/receive queue and vice
  * versa mapping.
diff --git a/drivers/net/sfc/sfc_repr_proxy.c b/drivers/net/sfc/sfc_repr_proxy.c
index 6a89cca40a..03b6421b04 100644
--- a/drivers/net/sfc/sfc_repr_proxy.c
+++ b/drivers/net/sfc/sfc_repr_proxy.c
@@ -15,6 +15,8 @@
 #include "sfc_repr_proxy.h"
 #include "sfc_repr_proxy_api.h"
 #include "sfc.h"
+#include "sfc_ev.h"
+#include "sfc_rx.h"
 
 /**
  * Amount of time to wait for the representor proxy routine (which is
@@ -136,6 +138,181 @@ sfc_repr_proxy_routine(void *arg)
return 0;
 }
 
+static int
+sfc_repr_proxy_rxq_attach(struct sfc_adapter *sa)
+{
+   struct sfc_adapter_shared * const sas = sfc_sa2shared(sa);
+   struct sfc_repr_proxy *rp = &sa->repr_proxy;
+   unsigned int i;
+
+   sfc_log_init(sa, "entry");
+
+   for (i = 0; i < sfc_repr_nb_rxq(sas); i++) {
+   sfc_sw_index_t sw_index = sfc_repr_rxq_sw_index(sas, i);
+
+   rp->dp_rxq[i].sw_index = sw_index;
+   }
+
+   sfc_log_init(sa, "done");
+
+   return 0;
+}
+
+static void
+sfc_repr_proxy_rxq_detach(struct sfc_adapter *sa)
+{
+   struct sfc_adapter_shared * const sas = sfc_sa2shared(sa);
+   struct sfc_repr_proxy *rp = &sa->repr_proxy;
+   unsigned int i;
+
+   sfc_log_init(sa, "entry");
+
+   for (i = 0; i < sfc_repr_nb_rxq(sas); i++)
+   rp->dp_rxq[i].sw_index = 0;
+
+   sfc_log_init(sa, "done");
+}
+
+static int
+sfc_repr_proxy_rxq_init(struct sfc_adapter *sa,
+   struct sfc_repr_proxy_dp_rxq *rxq)
+{
+   struct sfc_adapter_shared * const sas = sfc_sa2shared(sa);
+   uint16_t nb_rx_desc = SFC_REPR_PROXY_RX_DESC_COUNT;
+   struct sfc_rxq_info *rxq_info;
+   struct rte_eth_rxconf rxconf = {
+   .rx_free_thresh = SFC_REPR_PROXY_RXQ_REFILL_LEVEL,
+   .rx_drop_en = 1,
+   };
+   int rc;
+
+   sfc_log_init(sa, "entry");
+
+   rxq_info = &sas->rxq_info[rxq->sw_index];
+   if (rxq_info->state & SFC_RXQ_INITIALIZED) {
+   sfc_log_init(sa, "RxQ is already initialized - skip");
+   return 0;
+   }
+
+   nb_rx_desc = RTE_MIN(nb_rx_desc, sa->rxq_max_entries);
+   nb_rx_desc = RTE_MAX(nb_rx_desc, sa->rxq_min_entries);
+
+   rc = sfc_rx_qinit_info(sa, rxq->sw_index, EFX_RXQ_FLAG_INGRESS_MPORT);
+   if (rc != 0) {
+   sfc_err(sa, "failed to init representor proxy RxQ info");
+   goto fail_repr_rxq_init_info;
+   }
+
+   rc = sfc_rx_qinit(sa, rxq->sw_index, nb_rx_desc, sa->socket_id, &rxconf,
+ rxq->mp);
+   if (rc != 0) {
+   sfc_err(sa, "failed to init representor proxy RxQ");
+   goto fail_repr_rxq_init;
+   }
+
+   sfc_log_init(sa, "done");
+
+   return 0;
+
+fail_repr_rxq_init:
+fail_repr_rxq_init_info:
+   sfc_log_init(sa, "failed: %s", rte_strerror(rc));
+
+   return rc;
+}
+
+static void
+sfc_repr_proxy_rxq_fini(struct sfc_adapter *sa)
+{
+   struct sfc_adapter_shared * const sas = sfc_sa2shared(sa);
+   struct sfc_repr_proxy *rp = &sa->repr_proxy;
+   struct sfc_rxq_info *rxq_info;
+   unsigned int i;
+
+   sfc_log_init(sa, "entry");
+
+   if (!sfc_repr_available(sas)) {
+   sfc_log_init(sa, "representors not supported - skip");
+   return;
+   }
+
+   for (i = 0; i < sfc_repr_nb_rxq(sas); i++) {
+   struct sfc_repr_proxy_dp_rxq *rxq = &rp->dp_rxq[i];
+
+   rxq_info = &sas->rxq_info[rxq->sw_index];
+   if (rxq_info->state != SFC_RXQ_INITIALIZED) {
+   sfc_log_init(sa,
+   "representor RxQ %u is already finalized - 
skip",
+ 

[dpdk-dev] [PATCH v2 18/38] net/sfc: implement representor TxQ start/stop

[Andrew Rybchenko]

From: Igor Romanov 

Implement Tx queue start and stop in port representor proxy.

Signed-off-by: Igor Romanov 
Signed-off-by: Andrew Rybchenko 
Reviewed-by: Andy Moreton 
Reviewed-by: Ivan Malov 
---
 drivers/net/sfc/sfc_ev.h |   8 ++
 drivers/net/sfc/sfc_repr_proxy.c | 166 +++
 drivers/net/sfc/sfc_repr_proxy.h |  11 ++
 drivers/net/sfc/sfc_tx.c |  15 ++-
 drivers/net/sfc/sfc_tx.h |   1 +
 5 files changed, 199 insertions(+), 2 deletions(-)

diff --git a/drivers/net/sfc/sfc_ev.h b/drivers/net/sfc/sfc_ev.h
index bcb7fbe466..a4ababc2bc 100644
--- a/drivers/net/sfc/sfc_ev.h
+++ b/drivers/net/sfc/sfc_ev.h
@@ -118,6 +118,14 @@ sfc_repr_rxq_sw_index(const struct sfc_adapter_shared *sas,
repr_queue_id;
 }
 
+static inline sfc_sw_index_t
+sfc_repr_txq_sw_index(const struct sfc_adapter_shared *sas,
+ unsigned int repr_queue_id)
+{
+   /* Reserved TxQ for representors is the first reserved TxQ */
+   return sfc_repr_available(sas) ? repr_queue_id : SFC_SW_INDEX_INVALID;
+}
+
 /*
  * Functions below define event queue to transmit/receive queue and vice
  * versa mapping.
diff --git a/drivers/net/sfc/sfc_repr_proxy.c b/drivers/net/sfc/sfc_repr_proxy.c
index 03b6421b04..a5be8fa270 100644
--- a/drivers/net/sfc/sfc_repr_proxy.c
+++ b/drivers/net/sfc/sfc_repr_proxy.c
@@ -17,6 +17,7 @@
 #include "sfc.h"
 #include "sfc_ev.h"
 #include "sfc_rx.h"
+#include "sfc_tx.h"
 
 /**
  * Amount of time to wait for the representor proxy routine (which is
@@ -138,6 +139,155 @@ sfc_repr_proxy_routine(void *arg)
return 0;
 }
 
+static int
+sfc_repr_proxy_txq_attach(struct sfc_adapter *sa)
+{
+   struct sfc_adapter_shared * const sas = sfc_sa2shared(sa);
+   struct sfc_repr_proxy *rp = &sa->repr_proxy;
+   unsigned int i;
+
+   sfc_log_init(sa, "entry");
+
+   for (i = 0; i < sfc_repr_nb_txq(sas); i++) {
+   sfc_sw_index_t sw_index = sfc_repr_txq_sw_index(sas, i);
+
+   rp->dp_txq[i].sw_index = sw_index;
+   }
+
+   sfc_log_init(sa, "done");
+
+   return 0;
+}
+
+static void
+sfc_repr_proxy_txq_detach(struct sfc_adapter *sa)
+{
+   struct sfc_adapter_shared * const sas = sfc_sa2shared(sa);
+   struct sfc_repr_proxy *rp = &sa->repr_proxy;
+   unsigned int i;
+
+   sfc_log_init(sa, "entry");
+
+   for (i = 0; i < sfc_repr_nb_txq(sas); i++)
+   rp->dp_txq[i].sw_index = 0;
+
+   sfc_log_init(sa, "done");
+}
+
+int
+sfc_repr_proxy_txq_init(struct sfc_adapter *sa)
+{
+   struct sfc_adapter_shared * const sas = sfc_sa2shared(sa);
+   struct sfc_repr_proxy *rp = &sa->repr_proxy;
+   const struct rte_eth_txconf tx_conf = {
+   .tx_free_thresh = SFC_REPR_PROXY_TXQ_FREE_THRESH,
+   };
+   struct sfc_txq_info *txq_info;
+   unsigned int init_i;
+   unsigned int i;
+   int rc;
+
+   sfc_log_init(sa, "entry");
+
+   if (!sfc_repr_available(sas)) {
+   sfc_log_init(sa, "representors not supported - skip");
+   return 0;
+   }
+
+   for (init_i = 0; init_i < sfc_repr_nb_txq(sas); init_i++) {
+   struct sfc_repr_proxy_dp_txq *txq = &rp->dp_txq[init_i];
+
+   txq_info = &sfc_sa2shared(sa)->txq_info[txq->sw_index];
+   if (txq_info->state == SFC_TXQ_INITIALIZED) {
+   sfc_log_init(sa,
+   "representor proxy TxQ %u is already 
initialized - skip",
+   init_i);
+   continue;
+   }
+
+   sfc_tx_qinit_info(sa, txq->sw_index);
+
+   rc = sfc_tx_qinit(sa, txq->sw_index,
+ SFC_REPR_PROXY_TX_DESC_COUNT, sa->socket_id,
+ &tx_conf);
+
+   if (rc != 0) {
+   sfc_err(sa, "failed to init representor proxy TxQ %u",
+   init_i);
+   goto fail_init;
+   }
+   }
+
+   sfc_log_init(sa, "done");
+
+   return 0;
+
+fail_init:
+   for (i = 0; i < init_i; i++) {
+   struct sfc_repr_proxy_dp_txq *txq = &rp->dp_txq[i];
+
+   txq_info = &sfc_sa2shared(sa)->txq_info[txq->sw_index];
+   if (txq_info->state == SFC_TXQ_INITIALIZED)
+   sfc_tx_qfini(sa, txq->sw_index);
+   }
+   sfc_log_init(sa, "failed: %s", rte_strerror(rc));
+
+   return rc;
+}
+
+void
+sfc_repr_proxy_txq_fini(struct sfc_adapter *sa)
+{
+   struct sfc_adapter_shared * const sas = sfc_sa2shared(sa);
+   struct sfc_repr_proxy *rp = &sa->repr_proxy;
+   struct sfc_txq_info *txq_info;
+   unsigned int i;
+
+   sfc_log_init(sa, "entry");
+
+   if (!sfc_repr_available(sas)) {
+   sfc_log_init(sa, "representors not supported - skip");
+   return;
+   }
+
+   for (i = 0; i < sfc_repr_nb_txq(sas);

[dpdk-dev] [PATCH v2 20/38] net/sfc: implement port representor link update

[Andrew Rybchenko]

From: Igor Romanov 

Implement the callback by reporting link down if the representor
is not started, otherwise report link up with undefined link speed.

Link speed is undefined since representors can pass traffic to each
other even if the PF link is down.

Signed-off-by: Igor Romanov 
Signed-off-by: Andrew Rybchenko 
Reviewed-by: Andy Moreton 
Reviewed-by: Ivan Malov 
---
 drivers/net/sfc/sfc_repr.c | 19 +++
 1 file changed, 19 insertions(+)

diff --git a/drivers/net/sfc/sfc_repr.c b/drivers/net/sfc/sfc_repr.c
index f60106c196..9b70a3be76 100644
--- a/drivers/net/sfc/sfc_repr.c
+++ b/drivers/net/sfc/sfc_repr.c
@@ -463,6 +463,24 @@ sfc_repr_dev_infos_get(struct rte_eth_dev *dev,
return 0;
 }
 
+static int
+sfc_repr_dev_link_update(struct rte_eth_dev *dev,
+__rte_unused int wait_to_complete)
+{
+   struct sfc_repr *sr = sfc_repr_by_eth_dev(dev);
+   struct rte_eth_link link;
+
+   if (sr->state != SFC_ETHDEV_STARTED) {
+   sfc_port_link_mode_to_info(EFX_LINK_UNKNOWN, &link);
+   } else {
+   memset(&link, 0, sizeof(link));
+   link.link_status = ETH_LINK_UP;
+   link.link_speed = ETH_SPEED_NUM_UNKNOWN;
+   }
+
+   return rte_eth_linkstatus_set(dev, &link);
+}
+
 static int
 sfc_repr_ring_create(uint16_t pf_port_id, uint16_t repr_id,
 const char *type_name, uint16_t qid, uint16_t nb_desc,
@@ -760,6 +778,7 @@ static const struct eth_dev_ops sfc_repr_dev_ops = {
.dev_stop   = sfc_repr_dev_stop,
.dev_close  = sfc_repr_dev_close,
.dev_infos_get  = sfc_repr_dev_infos_get,
+   .link_update= sfc_repr_dev_link_update,
.rx_queue_setup = sfc_repr_rx_queue_setup,
.rx_queue_release   = sfc_repr_rx_queue_release,
.tx_queue_setup = sfc_repr_tx_queue_setup,
-- 
2.30.2

[dpdk-dev] [PATCH v2 19/38] net/sfc: implement port representor start and stop

[Andrew Rybchenko]

From: Igor Romanov 

Implement queue start and stop operation both in port
representors and representor proxy.

Signed-off-by: Igor Romanov 
Signed-off-by: Andrew Rybchenko 
Reviewed-by: Andy Moreton 
Reviewed-by: Ivan Malov 
---
 drivers/net/sfc/sfc_mae.h|   9 +-
 drivers/net/sfc/sfc_repr.c   | 181 +++
 drivers/net/sfc/sfc_repr_proxy.c | 453 ++-
 drivers/net/sfc/sfc_repr_proxy.h |  16 +
 drivers/net/sfc/sfc_repr_proxy_api.h |   3 +
 5 files changed, 644 insertions(+), 18 deletions(-)

diff --git a/drivers/net/sfc/sfc_mae.h b/drivers/net/sfc/sfc_mae.h
index 684f0daf7a..d835056aef 100644
--- a/drivers/net/sfc/sfc_mae.h
+++ b/drivers/net/sfc/sfc_mae.h
@@ -139,10 +139,17 @@ struct sfc_mae_counter_registry {
uint32_tservice_id;
 };
 
+/**
+ * MAE rules used to capture traffic generated by VFs and direct it to
+ * representors (one for each VF).
+ */
+#define SFC_MAE_NB_REPR_RULES_MAX  (64)
+
 /** Rules to forward traffic from PHY port to PF and from PF to PHY port */
 #define SFC_MAE_NB_SWITCHDEV_RULES (2)
 /** Maximum required internal MAE rules */
-#define SFC_MAE_NB_RULES_MAX   (SFC_MAE_NB_SWITCHDEV_RULES)
+#define SFC_MAE_NB_RULES_MAX   (SFC_MAE_NB_SWITCHDEV_RULES + \
+SFC_MAE_NB_REPR_RULES_MAX)
 
 struct sfc_mae_rule {
efx_mae_match_spec_t*spec;
diff --git a/drivers/net/sfc/sfc_repr.c b/drivers/net/sfc/sfc_repr.c
index ddd848466c..f60106c196 100644
--- a/drivers/net/sfc/sfc_repr.c
+++ b/drivers/net/sfc/sfc_repr.c
@@ -9,6 +9,7 @@
 
 #include 
 
+#include 
 #include 
 #include 
 #include 
@@ -21,6 +22,7 @@
 #include "sfc_ethdev_state.h"
 #include "sfc_repr_proxy_api.h"
 #include "sfc_switch.h"
+#include "sfc_dp_tx.h"
 
 /** Multi-process shared representor private data */
 struct sfc_repr_shared {
@@ -136,6 +138,179 @@ sfc_repr_lock_fini(__rte_unused struct sfc_repr *sr)
/* Just for symmetry of the API */
 }
 
+static void
+sfc_repr_rx_queue_stop(void *queue)
+{
+   struct sfc_repr_rxq *rxq = queue;
+
+   if (rxq == NULL)
+   return;
+
+   rte_ring_reset(rxq->ring);
+}
+
+static void
+sfc_repr_tx_queue_stop(void *queue)
+{
+   struct sfc_repr_txq *txq = queue;
+
+   if (txq == NULL)
+   return;
+
+   rte_ring_reset(txq->ring);
+}
+
+static int
+sfc_repr_start(struct rte_eth_dev *dev)
+{
+   struct sfc_repr *sr = sfc_repr_by_eth_dev(dev);
+   struct sfc_repr_shared *srs;
+   int ret;
+
+   sfcr_info(sr, "entry");
+
+   SFC_ASSERT(sfc_repr_lock_is_locked(sr));
+
+   switch (sr->state) {
+   case SFC_ETHDEV_CONFIGURED:
+   break;
+   case SFC_ETHDEV_STARTED:
+   sfcr_info(sr, "already started");
+   return 0;
+   default:
+   ret = -EINVAL;
+   goto fail_bad_state;
+   }
+
+   sr->state = SFC_ETHDEV_STARTING;
+
+   srs = sfc_repr_shared_by_eth_dev(dev);
+   ret = sfc_repr_proxy_start_repr(srs->pf_port_id, srs->repr_id);
+   if (ret != 0) {
+   SFC_ASSERT(ret > 0);
+   ret = -ret;
+   goto fail_start;
+   }
+
+   sr->state = SFC_ETHDEV_STARTED;
+
+   sfcr_info(sr, "done");
+
+   return 0;
+
+fail_start:
+   sr->state = SFC_ETHDEV_CONFIGURED;
+
+fail_bad_state:
+   sfcr_err(sr, "%s() failed: %s", __func__, rte_strerror(-ret));
+   return ret;
+}
+
+static int
+sfc_repr_dev_start(struct rte_eth_dev *dev)
+{
+   struct sfc_repr *sr = sfc_repr_by_eth_dev(dev);
+   int ret;
+
+   sfcr_info(sr, "entry");
+
+   sfc_repr_lock(sr);
+   ret = sfc_repr_start(dev);
+   sfc_repr_unlock(sr);
+
+   if (ret != 0)
+   goto fail_start;
+
+   sfcr_info(sr, "done");
+
+   return 0;
+
+fail_start:
+   sfcr_err(sr, "%s() failed: %s", __func__, rte_strerror(-ret));
+   return ret;
+}
+
+static int
+sfc_repr_stop(struct rte_eth_dev *dev)
+{
+   struct sfc_repr *sr = sfc_repr_by_eth_dev(dev);
+   struct sfc_repr_shared *srs;
+   unsigned int i;
+   int ret;
+
+   sfcr_info(sr, "entry");
+
+   SFC_ASSERT(sfc_repr_lock_is_locked(sr));
+
+   switch (sr->state) {
+   case SFC_ETHDEV_STARTED:
+   break;
+   case SFC_ETHDEV_CONFIGURED:
+   sfcr_info(sr, "already stopped");
+   return 0;
+   default:
+   sfcr_err(sr, "stop in unexpected state %u", sr->state);
+   SFC_ASSERT(B_FALSE);
+   ret = -EINVAL;
+   goto fail_bad_state;
+   }
+
+   srs = sfc_repr_shared_by_eth_dev(dev);
+   ret = sfc_repr_proxy_stop_repr(srs->pf_port_id, srs->repr_id);
+   if (ret != 0) {
+   SFC_ASSERT(ret > 0);
+   ret = -ret;
+   goto fail_stop;
+   }
+
+   for (i = 0; i < dev->data->nb_rx_queues; i++)
+   sfc_repr

[dpdk-dev] [PATCH v2 21/38] net/sfc: support multiple device probe

[Andrew Rybchenko]

From: Igor Romanov 

Support probing the device multiple times so that additional port
representors can be created with hotplug EAL API. To hotplug a
representor, the PF must be hotplugged with different representor
device argument.

Signed-off-by: Igor Romanov 
Signed-off-by: Andrew Rybchenko 
Reviewed-by: Andy Moreton 
Reviewed-by: Ivan Malov 
---
 drivers/net/sfc/sfc_ethdev.c | 55 
 drivers/net/sfc/sfc_repr.c   | 35 +--
 2 files changed, 57 insertions(+), 33 deletions(-)

diff --git a/drivers/net/sfc/sfc_ethdev.c b/drivers/net/sfc/sfc_ethdev.c
index efd5e6b1ab..f69bbde11a 100644
--- a/drivers/net/sfc/sfc_ethdev.c
+++ b/drivers/net/sfc/sfc_ethdev.c
@@ -2440,31 +2440,40 @@ sfc_parse_rte_devargs(const char *args, struct 
rte_eth_devargs *devargs)
 }
 
 static int
-sfc_eth_dev_create(struct rte_pci_device *pci_dev,
-  struct sfc_ethdev_init_data *init_data,
-  struct rte_eth_dev **devp)
+sfc_eth_dev_find_or_create(struct rte_pci_device *pci_dev,
+  struct sfc_ethdev_init_data *init_data,
+  struct rte_eth_dev **devp,
+  bool *dev_created)
 {
struct rte_eth_dev *dev;
+   bool created = false;
int rc;
 
-   rc = rte_eth_dev_create(&pci_dev->device, pci_dev->device.name,
-   sizeof(struct sfc_adapter_shared),
-   eth_dev_pci_specific_init, pci_dev,
-   sfc_eth_dev_init, init_data);
-   if (rc != 0) {
-   SFC_GENERIC_LOG(ERR, "Failed to create sfc ethdev '%s'",
-   pci_dev->device.name);
-   return rc;
-   }
-
dev = rte_eth_dev_allocated(pci_dev->device.name);
if (dev == NULL) {
-   SFC_GENERIC_LOG(ERR, "Failed to find allocated sfc ethdev '%s'",
+   rc = rte_eth_dev_create(&pci_dev->device, pci_dev->device.name,
+   sizeof(struct sfc_adapter_shared),
+   eth_dev_pci_specific_init, pci_dev,
+   sfc_eth_dev_init, init_data);
+   if (rc != 0) {
+   SFC_GENERIC_LOG(ERR, "Failed to create sfc ethdev '%s'",
+   pci_dev->device.name);
+   return rc;
+   }
+
+   created = true;
+
+   dev = rte_eth_dev_allocated(pci_dev->device.name);
+   if (dev == NULL) {
+   SFC_GENERIC_LOG(ERR,
+   "Failed to find allocated sfc ethdev '%s'",
pci_dev->device.name);
-   return -ENODEV;
+   return -ENODEV;
+   }
}
 
*devp = dev;
+   *dev_created = created;
 
return 0;
 }
@@ -2525,6 +2534,7 @@ static int sfc_eth_dev_pci_probe(struct rte_pci_driver 
*pci_drv __rte_unused,
struct sfc_ethdev_init_data init_data;
struct rte_eth_devargs eth_da;
struct rte_eth_dev *dev;
+   bool dev_created;
int rc;
 
if (pci_dev->device.devargs != NULL) {
@@ -2546,13 +2556,21 @@ static int sfc_eth_dev_pci_probe(struct rte_pci_driver 
*pci_drv __rte_unused,
return -ENOTSUP;
}
 
-   rc = sfc_eth_dev_create(pci_dev, &init_data, &dev);
+   /*
+* Driver supports RTE_PCI_DRV_PROBE_AGAIN. Hence create device only
+* if it does not already exist. Re-probing an existing device is
+* expected to allow additional representors to be configured.
+*/
+   rc = sfc_eth_dev_find_or_create(pci_dev, &init_data, &dev,
+   &dev_created);
if (rc != 0)
return rc;
 
rc = sfc_eth_dev_create_representors(dev, ð_da);
if (rc != 0) {
-   (void)rte_eth_dev_destroy(dev, sfc_eth_dev_uninit);
+   if (dev_created)
+   (void)rte_eth_dev_destroy(dev, sfc_eth_dev_uninit);
+
return rc;
}
 
@@ -2568,7 +2586,8 @@ static struct rte_pci_driver sfc_efx_pmd = {
.id_table = pci_id_sfc_efx_map,
.drv_flags =
RTE_PCI_DRV_INTR_LSC |
-   RTE_PCI_DRV_NEED_MAPPING,
+   RTE_PCI_DRV_NEED_MAPPING |
+   RTE_PCI_DRV_PROBE_AGAIN,
.probe = sfc_eth_dev_pci_probe,
.remove = sfc_eth_dev_pci_remove,
 };
diff --git a/drivers/net/sfc/sfc_repr.c b/drivers/net/sfc/sfc_repr.c
index 9b70a3be76..922f4da432 100644
--- a/drivers/net/sfc/sfc_repr.c
+++ b/drivers/net/sfc/sfc_repr.c
@@ -908,6 +908,7 @@ sfc_repr_create(struct rte_eth_dev *parent, uint16_t 
representor_id,
struct sfc_repr_init_data repr_data;
char name[RTE_ETH_NAME_MAX_LEN];
int ret;
+   struct rte_eth_dev *dev;
 
if (snprintf(name, sizeof(name), "

[dpdk-dev] [PATCH v2 22/38] net/sfc: implement representor Tx routine

[Andrew Rybchenko]

From: Igor Romanov 

Forward traffic that is transmitted from a port representor to the
corresponding virtual function using the dedicated TxQ.

Signed-off-by: Igor Romanov 
Signed-off-by: Andrew Rybchenko 
Reviewed-by: Andy Moreton 
Reviewed-by: Ivan Malov 
---
 drivers/net/sfc/sfc_repr.c   | 45 
 drivers/net/sfc/sfc_repr_proxy.c | 88 +++-
 drivers/net/sfc/sfc_repr_proxy.h |  8 +++
 3 files changed, 140 insertions(+), 1 deletion(-)

diff --git a/drivers/net/sfc/sfc_repr.c b/drivers/net/sfc/sfc_repr.c
index 922f4da432..3e74313b12 100644
--- a/drivers/net/sfc/sfc_repr.c
+++ b/drivers/net/sfc/sfc_repr.c
@@ -160,6 +160,49 @@ sfc_repr_tx_queue_stop(void *queue)
rte_ring_reset(txq->ring);
 }
 
+static uint16_t
+sfc_repr_tx_burst(void *tx_queue, struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
+{
+   struct sfc_repr_txq *txq = tx_queue;
+   unsigned int n_tx;
+   void **objs;
+   uint16_t i;
+
+   /*
+* mbuf is likely cache-hot. Set flag and egress m-port here instead of
+* doing that in representors proxy. Also, it should help to avoid
+* cache bounce. Moreover, potentially, it allows to use one
+* multi-producer single-consumer ring for all representors.
+*
+* The only potential problem is doing so many times if enqueue
+* fails and sender retries.
+*/
+   for (i = 0; i < nb_pkts; ++i) {
+   struct rte_mbuf *m = tx_pkts[i];
+
+   m->ol_flags |= sfc_dp_mport_override;
+   *RTE_MBUF_DYNFIELD(m, sfc_dp_mport_offset,
+  efx_mport_id_t *) = txq->egress_mport;
+   }
+
+   objs = (void *)&tx_pkts[0];
+   n_tx = rte_ring_sp_enqueue_burst(txq->ring, objs, nb_pkts, NULL);
+
+   /*
+* Remove m-port override flag from packets that were not enqueued
+* Setting the flag only for enqueued packets after the burst is
+* not possible since the ownership of enqueued packets is
+* transferred to representor proxy.
+*/
+   for (i = n_tx; i < nb_pkts; ++i) {
+   struct rte_mbuf *m = tx_pkts[i];
+
+   m->ol_flags &= ~sfc_dp_mport_override;
+   }
+
+   return n_tx;
+}
+
 static int
 sfc_repr_start(struct rte_eth_dev *dev)
 {
@@ -760,6 +803,7 @@ sfc_repr_dev_close(struct rte_eth_dev *dev)
 
(void)sfc_repr_proxy_del_port(srs->pf_port_id, srs->repr_id);
 
+   dev->tx_pkt_burst = NULL;
dev->dev_ops = NULL;
 
sfc_repr_unlock(sr);
@@ -880,6 +924,7 @@ sfc_repr_eth_dev_init(struct rte_eth_dev *dev, void 
*init_params)
goto fail_mac_addrs;
}
 
+   dev->tx_pkt_burst = sfc_repr_tx_burst;
dev->dev_ops = &sfc_repr_dev_ops;
 
sr->state = SFC_ETHDEV_INITIALIZED;
diff --git a/drivers/net/sfc/sfc_repr_proxy.c b/drivers/net/sfc/sfc_repr_proxy.c
index ea03d5afdd..d8934bab65 100644
--- a/drivers/net/sfc/sfc_repr_proxy.c
+++ b/drivers/net/sfc/sfc_repr_proxy.c
@@ -25,6 +25,12 @@
  */
 #define SFC_REPR_PROXY_MBOX_POLL_TIMEOUT_MS1000
 
+/**
+ * Amount of time to wait for the representor proxy routine (which is
+ * running on a service core) to terminate after service core is stopped.
+ */
+#define SFC_REPR_PROXY_ROUTINE_TERMINATE_TIMEOUT_MS1
+
 static struct sfc_repr_proxy *
 sfc_repr_proxy_by_adapter(struct sfc_adapter *sa)
 {
@@ -148,16 +154,71 @@ sfc_repr_proxy_mbox_handle(struct sfc_repr_proxy *rp)
__atomic_store_n(&mbox->ack, true, __ATOMIC_RELEASE);
 }
 
+static void
+sfc_repr_proxy_handle_tx(struct sfc_repr_proxy_dp_txq *rp_txq,
+struct sfc_repr_proxy_txq *repr_txq)
+{
+   /*
+* With multiple representor proxy queues configured it is
+* possible that not all of the corresponding representor
+* queues were created. Skip the queues that do not exist.
+*/
+   if (repr_txq->ring == NULL)
+   return;
+
+   if (rp_txq->available < RTE_DIM(rp_txq->tx_pkts)) {
+   rp_txq->available +=
+   rte_ring_sc_dequeue_burst(repr_txq->ring,
+   (void **)(&rp_txq->tx_pkts[rp_txq->available]),
+   RTE_DIM(rp_txq->tx_pkts) - rp_txq->available,
+   NULL);
+
+   if (rp_txq->available == rp_txq->transmitted)
+   return;
+   }
+
+   rp_txq->transmitted += rp_txq->pkt_burst(rp_txq->dp,
+   &rp_txq->tx_pkts[rp_txq->transmitted],
+   rp_txq->available - rp_txq->transmitted);
+
+   if (rp_txq->available == rp_txq->transmitted) {
+   rp_txq->available = 0;
+   rp_txq->transmitted = 0;
+   }
+}
+
 static int32_t
 sfc_repr_proxy_routine(void *arg)
 {
+   struct sfc_repr_proxy_port *port;
struct sfc_repr_proxy *rp = arg;
+   unsigned int i;
 
sfc_repr_proxy_

[dpdk-dev] [PATCH v2 23/38] net/sfc: use xword type for EF100 Rx prefix

[Andrew Rybchenko]

From: Igor Romanov 

Layout of the EF100 Rx prefix is defined in terms of a 32 bytes long
value type (xword). Replace oword with xword to avoid truncation.

Signed-off-by: Igor Romanov 
Signed-off-by: Andrew Rybchenko 
Reviewed-by: Andy Moreton 
Reviewed-by: Ivan Malov 
---
 drivers/net/sfc/sfc_ef100_rx.c | 18 +-
 1 file changed, 9 insertions(+), 9 deletions(-)

diff --git a/drivers/net/sfc/sfc_ef100_rx.c b/drivers/net/sfc/sfc_ef100_rx.c
index 1bf04f565a..6e58b8c243 100644
--- a/drivers/net/sfc/sfc_ef100_rx.c
+++ b/drivers/net/sfc/sfc_ef100_rx.c
@@ -379,7 +379,7 @@ static const efx_rx_prefix_layout_t 
sfc_ef100_rx_prefix_layout = {
 
 static bool
 sfc_ef100_rx_prefix_to_offloads(const struct sfc_ef100_rxq *rxq,
-   const efx_oword_t *rx_prefix,
+   const efx_xword_t *rx_prefix,
struct rte_mbuf *m)
 {
const efx_word_t *class;
@@ -399,19 +399,19 @@ sfc_ef100_rx_prefix_to_offloads(const struct 
sfc_ef100_rxq *rxq,
m->packet_type = sfc_ef100_rx_class_decode(*class, &ol_flags);
 
if ((rxq->flags & SFC_EF100_RXQ_RSS_HASH) &&
-   EFX_TEST_OWORD_BIT(rx_prefix[0],
+   EFX_TEST_XWORD_BIT(rx_prefix[0],
   ESF_GZ_RX_PREFIX_RSS_HASH_VALID_LBN)) {
ol_flags |= PKT_RX_RSS_HASH;
-   /* EFX_OWORD_FIELD converts little-endian to CPU */
-   m->hash.rss = EFX_OWORD_FIELD(rx_prefix[0],
+   /* EFX_XWORD_FIELD converts little-endian to CPU */
+   m->hash.rss = EFX_XWORD_FIELD(rx_prefix[0],
  ESF_GZ_RX_PREFIX_RSS_HASH);
}
 
if (rxq->flags & SFC_EF100_RXQ_USER_MARK) {
uint32_t user_mark;
 
-   /* EFX_OWORD_FIELD converts little-endian to CPU */
-   user_mark = EFX_OWORD_FIELD(rx_prefix[0],
+   /* EFX_XWORD_FIELD converts little-endian to CPU */
+   user_mark = EFX_XWORD_FIELD(rx_prefix[0],
ESF_GZ_RX_PREFIX_USER_MARK);
if (user_mark != SFC_EF100_USER_MARK_INVALID) {
ol_flags |= PKT_RX_FDIR | PKT_RX_FDIR_ID;
@@ -480,7 +480,7 @@ sfc_ef100_rx_process_ready_pkts(struct sfc_ef100_rxq *rxq,
while (rxq->ready_pkts > 0 && rx_pkts != rx_pkts_end) {
struct rte_mbuf *pkt;
struct rte_mbuf *lastseg;
-   const efx_oword_t *rx_prefix;
+   const efx_xword_t *rx_prefix;
uint16_t pkt_len;
uint16_t seg_len;
bool deliver;
@@ -495,9 +495,9 @@ sfc_ef100_rx_process_ready_pkts(struct sfc_ef100_rxq *rxq,
pkt->rearm_data[0] = rxq->rearm_data;
 
/* data_off already moved past Rx prefix */
-   rx_prefix = (const efx_oword_t *)sfc_ef100_rx_pkt_prefix(pkt);
+   rx_prefix = (const efx_xword_t *)sfc_ef100_rx_pkt_prefix(pkt);
 
-   pkt_len = EFX_OWORD_FIELD(rx_prefix[0],
+   pkt_len = EFX_XWORD_FIELD(rx_prefix[0],
  ESF_GZ_RX_PREFIX_LENGTH);
SFC_ASSERT(pkt_len > 0);
rte_pktmbuf_pkt_len(pkt) = pkt_len;
-- 
2.30.2

[dpdk-dev] [PATCH v2 24/38] net/sfc: handle ingress m-port in EF100 Rx prefix

[Andrew Rybchenko]

From: Igor Romanov 

Set ingress mport dynamic field in mbuf in EF100.

For a given PF, Rx queues of representor devices
function on top of the only Rx queue operated by
the PF representor proxy facility. This field is
a means to demultiplex traffic hitting the queue.

Signed-off-by: Igor Romanov 
Signed-off-by: Andrew Rybchenko 
Reviewed-by: Andy Moreton 
Reviewed-by: Ivan Malov 
---
 drivers/net/sfc/sfc_ef100_rx.c | 18 ++
 1 file changed, 18 insertions(+)

diff --git a/drivers/net/sfc/sfc_ef100_rx.c b/drivers/net/sfc/sfc_ef100_rx.c
index 6e58b8c243..378c0314ae 100644
--- a/drivers/net/sfc/sfc_ef100_rx.c
+++ b/drivers/net/sfc/sfc_ef100_rx.c
@@ -62,6 +62,7 @@ struct sfc_ef100_rxq {
 #define SFC_EF100_RXQ_RSS_HASH 0x10
 #define SFC_EF100_RXQ_USER_MARK0x20
 #define SFC_EF100_RXQ_FLAG_INTR_EN 0x40
+#define SFC_EF100_RXQ_INGRESS_MPORT0x80
unsigned intptr_mask;
unsigned intevq_phase_bit_shift;
unsigned intready_pkts;
@@ -370,6 +371,8 @@ static const efx_rx_prefix_layout_t 
sfc_ef100_rx_prefix_layout = {
SFC_EF100_RX_PREFIX_FIELD(LENGTH, B_FALSE),
SFC_EF100_RX_PREFIX_FIELD(RSS_HASH_VALID, B_FALSE),
SFC_EF100_RX_PREFIX_FIELD(CLASS, B_FALSE),
+   EFX_RX_PREFIX_FIELD(INGRESS_MPORT,
+   ESF_GZ_RX_PREFIX_INGRESS_MPORT, B_FALSE),
SFC_EF100_RX_PREFIX_FIELD(RSS_HASH, B_FALSE),
SFC_EF100_RX_PREFIX_FIELD(USER_MARK, B_FALSE),
 
@@ -419,6 +422,15 @@ sfc_ef100_rx_prefix_to_offloads(const struct sfc_ef100_rxq 
*rxq,
}
}
 
+   if (rxq->flags & SFC_EF100_RXQ_INGRESS_MPORT) {
+   ol_flags |= sfc_dp_mport_override;
+   *RTE_MBUF_DYNFIELD(m,
+   sfc_dp_mport_offset,
+   typeof(&((efx_mport_id_t *)0)->id)) =
+   EFX_XWORD_FIELD(rx_prefix[0],
+   ESF_GZ_RX_PREFIX_INGRESS_MPORT);
+   }
+
m->ol_flags = ol_flags;
return true;
 }
@@ -806,6 +818,12 @@ sfc_ef100_rx_qstart(struct sfc_dp_rxq *dp_rxq, unsigned 
int evq_read_ptr,
else
rxq->flags &= ~SFC_EF100_RXQ_USER_MARK;
 
+   if ((unsup_rx_prefix_fields &
+(1U << EFX_RX_PREFIX_FIELD_INGRESS_MPORT)) == 0)
+   rxq->flags |= SFC_EF100_RXQ_INGRESS_MPORT;
+   else
+   rxq->flags &= ~SFC_EF100_RXQ_INGRESS_MPORT;
+
rxq->prefix_size = pinfo->erpl_length;
rxq->rearm_data = sfc_ef100_mk_mbuf_rearm_data(rxq->dp.dpq.port_id,
   rxq->prefix_size);
-- 
2.30.2

[dpdk-dev] [PATCH v2 25/38] net/sfc: implement representor Rx routine

[Andrew Rybchenko]

From: Igor Romanov 

Implement traffic forwarding for representor and representor proxy
from virtual functions to representor Rx queues.

Signed-off-by: Igor Romanov 
Signed-off-by: Andrew Rybchenko 
Reviewed-by: Andy Moreton 
Reviewed-by: Ivan Malov 
---
 drivers/net/sfc/sfc_repr.c   |  12 +++
 drivers/net/sfc/sfc_repr_proxy.c | 134 +++
 drivers/net/sfc/sfc_repr_proxy.h |  11 +++
 3 files changed, 157 insertions(+)

diff --git a/drivers/net/sfc/sfc_repr.c b/drivers/net/sfc/sfc_repr.c
index 3e74313b12..067496c5f0 100644
--- a/drivers/net/sfc/sfc_repr.c
+++ b/drivers/net/sfc/sfc_repr.c
@@ -160,6 +160,16 @@ sfc_repr_tx_queue_stop(void *queue)
rte_ring_reset(txq->ring);
 }
 
+static uint16_t
+sfc_repr_rx_burst(void *rx_queue, struct rte_mbuf **rx_pkts, uint16_t nb_pkts)
+{
+   struct sfc_repr_rxq *rxq = rx_queue;
+   void **objs = (void *)&rx_pkts[0];
+
+   /* mbufs port is already filled correctly by representors proxy */
+   return rte_ring_sc_dequeue_burst(rxq->ring, objs, nb_pkts, NULL);
+}
+
 static uint16_t
 sfc_repr_tx_burst(void *tx_queue, struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
 {
@@ -803,6 +813,7 @@ sfc_repr_dev_close(struct rte_eth_dev *dev)
 
(void)sfc_repr_proxy_del_port(srs->pf_port_id, srs->repr_id);
 
+   dev->rx_pkt_burst = NULL;
dev->tx_pkt_burst = NULL;
dev->dev_ops = NULL;
 
@@ -924,6 +935,7 @@ sfc_repr_eth_dev_init(struct rte_eth_dev *dev, void 
*init_params)
goto fail_mac_addrs;
}
 
+   dev->rx_pkt_burst = sfc_repr_rx_burst;
dev->tx_pkt_burst = sfc_repr_tx_burst;
dev->dev_ops = &sfc_repr_dev_ops;
 
diff --git a/drivers/net/sfc/sfc_repr_proxy.c b/drivers/net/sfc/sfc_repr_proxy.c
index d8934bab65..535b07ea52 100644
--- a/drivers/net/sfc/sfc_repr_proxy.c
+++ b/drivers/net/sfc/sfc_repr_proxy.c
@@ -18,6 +18,7 @@
 #include "sfc_ev.h"
 #include "sfc_rx.h"
 #include "sfc_tx.h"
+#include "sfc_dp_rx.h"
 
 /**
  * Amount of time to wait for the representor proxy routine (which is
@@ -31,6 +32,8 @@
  */
 #define SFC_REPR_PROXY_ROUTINE_TERMINATE_TIMEOUT_MS1
 
+#define SFC_REPR_INVALID_ROUTE_PORT_ID  (UINT16_MAX)
+
 static struct sfc_repr_proxy *
 sfc_repr_proxy_by_adapter(struct sfc_adapter *sa)
 {
@@ -187,6 +190,113 @@ sfc_repr_proxy_handle_tx(struct sfc_repr_proxy_dp_txq 
*rp_txq,
}
 }
 
+static struct sfc_repr_proxy_port *
+sfc_repr_proxy_rx_route_mbuf(struct sfc_repr_proxy *rp, struct rte_mbuf *m)
+{
+   struct sfc_repr_proxy_port *port;
+   efx_mport_id_t mport_id;
+
+   mport_id.id = *RTE_MBUF_DYNFIELD(m, sfc_dp_mport_offset,
+typeof(&((efx_mport_id_t *)0)->id));
+
+   TAILQ_FOREACH(port, &rp->ports, entries) {
+   if (port->egress_mport.id == mport_id.id) {
+   m->port = port->rte_port_id;
+   m->ol_flags &= ~sfc_dp_mport_override;
+   return port;
+   }
+   }
+
+   return NULL;
+}
+
+/*
+ * Returns true if a packet is encountered which should be forwarded to a
+ * port which is different from the one that is currently routed.
+ */
+static bool
+sfc_repr_proxy_rx_route(struct sfc_repr_proxy *rp,
+   struct sfc_repr_proxy_dp_rxq *rp_rxq)
+{
+   unsigned int i;
+
+   for (i = rp_rxq->routed;
+i < rp_rxq->available && !rp_rxq->stop_route;
+i++, rp_rxq->routed++) {
+   struct sfc_repr_proxy_port *port;
+   struct rte_mbuf *m = rp_rxq->pkts[i];
+
+   port = sfc_repr_proxy_rx_route_mbuf(rp, m);
+   /* Cannot find destination representor */
+   if (port == NULL) {
+   /* Effectively drop the packet */
+   rp_rxq->forwarded++;
+   continue;
+   }
+
+   /* Currently routed packets are mapped to a different port */
+   if (port->repr_id != rp_rxq->route_port_id &&
+   rp_rxq->route_port_id != SFC_REPR_INVALID_ROUTE_PORT_ID)
+   return true;
+
+   rp_rxq->route_port_id = port->repr_id;
+   }
+
+   return false;
+}
+
+static void
+sfc_repr_proxy_rx_forward(struct sfc_repr_proxy *rp,
+ struct sfc_repr_proxy_dp_rxq *rp_rxq)
+{
+   struct sfc_repr_proxy_port *port;
+
+   if (rp_rxq->route_port_id != SFC_REPR_INVALID_ROUTE_PORT_ID) {
+   port = sfc_repr_proxy_find_port(rp, rp_rxq->route_port_id);
+
+   if (port != NULL && port->started) {
+   rp_rxq->forwarded +=
+   rte_ring_sp_enqueue_burst(port->rxq[0].ring,
+   (void **)(&rp_rxq->pkts[rp_rxq->forwarded]),
+   rp_rxq->routed - rp_rxq->forwarded, NULL);
+   } else {
+   /* Drop all routed packets if the port is not started

[dpdk-dev] [PATCH v2 26/38] net/sfc: add simple port representor statistics

[Andrew Rybchenko]

From: Igor Romanov 

Gather statistics of enqueued and dequeued packets in Rx and Tx burst
callbacks to report in stats_get callback.

Signed-off-by: Igor Romanov 
Signed-off-by: Andrew Rybchenko 
Reviewed-by: Andy Moreton 
Reviewed-by: Ivan Malov 
---
 drivers/net/sfc/sfc_repr.c | 60 --
 1 file changed, 58 insertions(+), 2 deletions(-)

diff --git a/drivers/net/sfc/sfc_repr.c b/drivers/net/sfc/sfc_repr.c
index 067496c5f0..87f10092c3 100644
--- a/drivers/net/sfc/sfc_repr.c
+++ b/drivers/net/sfc/sfc_repr.c
@@ -32,15 +32,21 @@ struct sfc_repr_shared {
uint16_tswitch_port_id;
 };
 
+struct sfc_repr_queue_stats {
+   union sfc_pkts_bytespackets_bytes;
+};
+
 struct sfc_repr_rxq {
/* Datapath members */
struct rte_ring *ring;
+   struct sfc_repr_queue_stats stats;
 };
 
 struct sfc_repr_txq {
/* Datapath members */
struct rte_ring *ring;
efx_mport_id_t  egress_mport;
+   struct sfc_repr_queue_stats stats;
 };
 
 /** Primary process representor private data */
@@ -165,15 +171,30 @@ sfc_repr_rx_burst(void *rx_queue, struct rte_mbuf 
**rx_pkts, uint16_t nb_pkts)
 {
struct sfc_repr_rxq *rxq = rx_queue;
void **objs = (void *)&rx_pkts[0];
+   unsigned int n_rx;
 
/* mbufs port is already filled correctly by representors proxy */
-   return rte_ring_sc_dequeue_burst(rxq->ring, objs, nb_pkts, NULL);
+   n_rx = rte_ring_sc_dequeue_burst(rxq->ring, objs, nb_pkts, NULL);
+
+   if (n_rx > 0) {
+   unsigned int n_bytes = 0;
+   unsigned int i = 0;
+
+   do {
+   n_bytes += rx_pkts[i]->pkt_len;
+   } while (++i < n_rx);
+
+   sfc_pkts_bytes_add(&rxq->stats.packets_bytes, n_rx, n_bytes);
+   }
+
+   return n_rx;
 }
 
 static uint16_t
 sfc_repr_tx_burst(void *tx_queue, struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
 {
struct sfc_repr_txq *txq = tx_queue;
+   unsigned int n_bytes = 0;
unsigned int n_tx;
void **objs;
uint16_t i;
@@ -193,6 +214,7 @@ sfc_repr_tx_burst(void *tx_queue, struct rte_mbuf 
**tx_pkts, uint16_t nb_pkts)
m->ol_flags |= sfc_dp_mport_override;
*RTE_MBUF_DYNFIELD(m, sfc_dp_mport_offset,
   efx_mport_id_t *) = txq->egress_mport;
+   n_bytes += tx_pkts[i]->pkt_len;
}
 
objs = (void *)&tx_pkts[0];
@@ -202,14 +224,18 @@ sfc_repr_tx_burst(void *tx_queue, struct rte_mbuf 
**tx_pkts, uint16_t nb_pkts)
 * Remove m-port override flag from packets that were not enqueued
 * Setting the flag only for enqueued packets after the burst is
 * not possible since the ownership of enqueued packets is
-* transferred to representor proxy.
+* transferred to representor proxy. The same logic applies to
+* counting the enqueued packets' bytes.
 */
for (i = n_tx; i < nb_pkts; ++i) {
struct rte_mbuf *m = tx_pkts[i];
 
m->ol_flags &= ~sfc_dp_mport_override;
+   n_bytes -= m->pkt_len;
}
 
+   sfc_pkts_bytes_add(&txq->stats.packets_bytes, n_tx, n_bytes);
+
return n_tx;
 }
 
@@ -827,6 +853,35 @@ sfc_repr_dev_close(struct rte_eth_dev *dev)
return 0;
 }
 
+static int
+sfc_repr_stats_get(struct rte_eth_dev *dev, struct rte_eth_stats *stats)
+{
+   union sfc_pkts_bytes queue_stats;
+   uint16_t i;
+
+   for (i = 0; i < dev->data->nb_rx_queues; i++) {
+   struct sfc_repr_rxq *rxq = dev->data->rx_queues[i];
+
+   sfc_pkts_bytes_get(&rxq->stats.packets_bytes,
+  &queue_stats);
+
+   stats->ipackets += queue_stats.pkts;
+   stats->ibytes += queue_stats.bytes;
+   }
+
+   for (i = 0; i < dev->data->nb_tx_queues; i++) {
+   struct sfc_repr_txq *txq = dev->data->tx_queues[i];
+
+   sfc_pkts_bytes_get(&txq->stats.packets_bytes,
+  &queue_stats);
+
+   stats->opackets += queue_stats.pkts;
+   stats->obytes += queue_stats.bytes;
+   }
+
+   return 0;
+}
+
 static const struct eth_dev_ops sfc_repr_dev_ops = {
.dev_configure  = sfc_repr_dev_configure,
.dev_start  = sfc_repr_dev_start,
@@ -834,6 +889,7 @@ static const struct eth_dev_ops sfc_repr_dev_ops = {
.dev_close  = sfc_repr_dev_close,
.dev_infos_get  = sfc_repr_dev_infos_get,
.link_update= sfc_repr_dev_link_update,
+   .stats_get  = sfc_repr_stats_get,
.rx_queue_setup = sfc_repr_rx_queue_setup,
.rx_queue_release   = sfc_repr_rx_queue_release,
.tx_queue

[dpdk-dev] [PATCH v2 27/38] net/sfc: free MAE lock once switch domain is assigned

[Andrew Rybchenko]

From: Viacheslav Galaktionov 

If for some reason the hardware switch ID initialization function fails,
MAE lock is still held after the function finishes. This patch fixes that.

Fixes: 1e7fbdf0ba19 ("net/sfc: support concept of switch domains/ports")
Cc: sta...@dpdk.org

Signed-off-by: Viacheslav Galaktionov 
Signed-off-by: Andrew Rybchenko 
Reviewed-by: Andy Moreton 
---
 drivers/net/sfc/sfc_switch.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/sfc/sfc_switch.c b/drivers/net/sfc/sfc_switch.c
index c37cdf4a61..80c884a599 100644
--- a/drivers/net/sfc/sfc_switch.c
+++ b/drivers/net/sfc/sfc_switch.c
@@ -214,9 +214,9 @@ sfc_mae_assign_switch_domain(struct sfc_adapter *sa,
 
 fail_mem_alloc:
sfc_hw_switch_id_fini(sa, hw_switch_id);
-   rte_spinlock_unlock(&sfc_mae_switch.lock);
 
 fail_hw_switch_id_init:
+   rte_spinlock_unlock(&sfc_mae_switch.lock);
return rc;
 }
 
-- 
2.30.2

[dpdk-dev] [PATCH v2 28/38] common/sfc_efx/base: add multi-host function M-port selector

[Andrew Rybchenko]

From: Viacheslav Galaktionov 

Provide helper function to compose multi-host aware PCIe
function M-port selector.

The firmware expects mport selectors to use different sets of values to
represent a PCIe interface in mport selectors and elsewhere. In order to
avoid having the user perform the conversion themselves, it is now done
automatically when a selector is constructed.

In addition, a type has been added to libefx for possible PCIe interfaces.
This is done to abstract different representations away from the users.

Allow to support matching traffic coming from an arbitrary PCIe
end-point of the NIC and redirect traffic to it.

Signed-off-by: Viacheslav Galaktionov 
Signed-off-by: Andrew Rybchenko 
Reviewed-by: Andy Moreton 
---
 drivers/common/sfc_efx/base/efx.h | 22 +++
 drivers/common/sfc_efx/base/efx_mae.c | 86 +++
 drivers/common/sfc_efx/version.map|  1 +
 3 files changed, 96 insertions(+), 13 deletions(-)

diff --git a/drivers/common/sfc_efx/base/efx.h 
b/drivers/common/sfc_efx/base/efx.h
index 0a178128ba..159e7957a3 100644
--- a/drivers/common/sfc_efx/base/efx.h
+++ b/drivers/common/sfc_efx/base/efx.h
@@ -82,6 +82,13 @@ efx_family(
 
 #if EFSYS_OPT_PCI
 
+/* PCIe interface numbers for multi-host configurations. */
+typedef enum efx_pcie_interface_e {
+   EFX_PCIE_INTERFACE_CALLER = 1000,
+   EFX_PCIE_INTERFACE_HOST_PRIMARY,
+   EFX_PCIE_INTERFACE_NIC_EMBEDDED,
+} efx_pcie_interface_t;
+
 typedef struct efx_pci_ops_s {
/*
 * Function for reading PCIe configuration space.
@@ -4237,6 +4244,21 @@ efx_mae_mport_by_pcie_function(
__inuint32_t vf,
__out   efx_mport_sel_t *mportp);
 
+/*
+ * Get MPORT selector of a multi-host PCIe function.
+ *
+ * The resulting MPORT selector is opaque to the caller and can be
+ * passed as an argument to efx_mae_match_spec_mport_set()
+ * and efx_mae_action_set_populate_deliver().
+ */
+LIBEFX_API
+extern __checkReturn   efx_rc_t
+efx_mae_mport_by_pcie_mh_function(
+   __inefx_pcie_interface_t intf,
+   __inuint32_t pf,
+   __inuint32_t vf,
+   __out   efx_mport_sel_t *mportp);
+
 /*
  * Get MPORT selector by an MPORT ID
  *
diff --git a/drivers/common/sfc_efx/base/efx_mae.c 
b/drivers/common/sfc_efx/base/efx_mae.c
index 3f498fe189..37cc48eafc 100644
--- a/drivers/common/sfc_efx/base/efx_mae.c
+++ b/drivers/common/sfc_efx/base/efx_mae.c
@@ -727,35 +727,95 @@ efx_mae_mport_by_pcie_function(
efx_dword_t dword;
efx_rc_t rc;
 
+   rc = efx_mae_mport_by_pcie_mh_function(EFX_PCIE_INTERFACE_CALLER,
+  pf, vf, mportp);
+   if (rc != 0)
+   goto fail1;
+
+   return (0);
+
+fail1:
+   EFSYS_PROBE1(fail1, efx_rc_t, rc);
+   return (rc);
+}
+
+static __checkReturn   efx_rc_t
+efx_mae_intf_to_selector(
+   __inefx_pcie_interface_t intf,
+   __out   uint32_t *selector_intfp)
+{
+   efx_rc_t rc;
+
+   switch (intf) {
+   case EFX_PCIE_INTERFACE_HOST_PRIMARY:
+   EFX_STATIC_ASSERT(MAE_MPORT_SELECTOR_HOST_PRIMARY <=
+   EFX_MASK32(MAE_MPORT_SELECTOR_FUNC_INTF_ID));
+   *selector_intfp = MAE_MPORT_SELECTOR_HOST_PRIMARY;
+   break;
+   case EFX_PCIE_INTERFACE_NIC_EMBEDDED:
+   EFX_STATIC_ASSERT(MAE_MPORT_SELECTOR_NIC_EMBEDDED <=
+   EFX_MASK32(MAE_MPORT_SELECTOR_FUNC_INTF_ID));
+   *selector_intfp = MAE_MPORT_SELECTOR_NIC_EMBEDDED;
+   break;
+   case EFX_PCIE_INTERFACE_CALLER:
+   EFX_STATIC_ASSERT(MAE_MPORT_SELECTOR_CALLER_INTF <=
+   EFX_MASK32(MAE_MPORT_SELECTOR_FUNC_INTF_ID));
+   *selector_intfp = MAE_MPORT_SELECTOR_CALLER_INTF;
+   break;
+   default:
+   rc = EINVAL;
+   goto fail1;
+   }
+
+   return (0);
+
+fail1:
+   EFSYS_PROBE1(fail1, efx_rc_t, rc);
+   return (rc);
+}
+
+   __checkReturn   efx_rc_t
+efx_mae_mport_by_pcie_mh_function(
+   __inefx_pcie_interface_t intf,
+   __inuint32_t pf,
+   __inuint32_t vf,
+   __out   efx_mport_sel_t *mportp)
+{
+   uint32_t selector_intf;
+   efx_dword_t dword;
+   efx_rc_t rc;
+
EFX_STATIC_ASSERT(EFX_PCI_VF_INVALID ==
MAE_MPORT_SELECTOR_FUNC_VF_ID_NULL);
 
-   if (pf > EFX_MASK32(MAE_MPORT_SELECTOR_FUNC_PF_ID)) {
-   rc = EINVAL;
+   rc = efx_mae_intf_to_selector(intf, &selector_intf);
+   if (rc != 0)
goto fail1;
+
+   if (pf > EFX_MASK32(MAE_MPORT_SELECTOR_FUNC_MH_PF_I

[dpdk-dev] [PATCH v2 29/38] common/sfc_efx/base: retrieve function interfaces for VNICs

[Andrew Rybchenko]

From: Viacheslav Galaktionov 

This information is required to be able to fully identify the function.
Add this information to the NIC configuration structure for easy access.

Signed-off-by: Viacheslav Galaktionov 
Signed-off-by: Andrew Rybchenko 
Reviewed-by: Andy Moreton 
---
 drivers/common/sfc_efx/base/ef10_impl.h |  3 +-
 drivers/common/sfc_efx/base/ef10_nic.c  |  4 +-
 drivers/common/sfc_efx/base/efx.h   |  1 +
 drivers/common/sfc_efx/base/efx_impl.h  |  6 +++
 drivers/common/sfc_efx/base/efx_mcdi.c  | 55 +++--
 5 files changed, 64 insertions(+), 5 deletions(-)

diff --git a/drivers/common/sfc_efx/base/ef10_impl.h 
b/drivers/common/sfc_efx/base/ef10_impl.h
index 7c8d51b7a5..d48f238479 100644
--- a/drivers/common/sfc_efx/base/ef10_impl.h
+++ b/drivers/common/sfc_efx/base/ef10_impl.h
@@ -1372,7 +1372,8 @@ extern__checkReturn   efx_rc_t
 efx_mcdi_get_function_info(
__inefx_nic_t *enp,
__out   uint32_t *pfp,
-   __out_opt   uint32_t *vfp);
+   __out_opt   uint32_t *vfp,
+   __out_opt   efx_pcie_interface_t *intfp);
 
 LIBEFX_INTERNAL
 extern __checkReturn   efx_rc_t
diff --git a/drivers/common/sfc_efx/base/ef10_nic.c 
b/drivers/common/sfc_efx/base/ef10_nic.c
index eda0ad3068..3cd9ff89d0 100644
--- a/drivers/common/sfc_efx/base/ef10_nic.c
+++ b/drivers/common/sfc_efx/base/ef10_nic.c
@@ -1847,6 +1847,7 @@ efx_mcdi_nic_board_cfg(
efx_nic_cfg_t *encp = &(enp->en_nic_cfg);
ef10_link_state_t els;
efx_port_t *epp = &(enp->en_port);
+   efx_pcie_interface_t intf;
uint32_t board_type = 0;
uint32_t base, nvec;
uint32_t port;
@@ -1875,11 +1876,12 @@ efx_mcdi_nic_board_cfg(
 *  - PCIe PF: pf = PF number, vf = 0x.
 *  - PCIe VF: pf = parent PF, vf = VF number.
 */
-   if ((rc = efx_mcdi_get_function_info(enp, &pf, &vf)) != 0)
+   if ((rc = efx_mcdi_get_function_info(enp, &pf, &vf, &intf)) != 0)
goto fail3;
 
encp->enc_pf = pf;
encp->enc_vf = vf;
+   encp->enc_intf = intf;
 
if ((rc = ef10_mcdi_get_pf_count(enp, &encp->enc_hw_pf_count)) != 0)
goto fail4;
diff --git a/drivers/common/sfc_efx/base/efx.h 
b/drivers/common/sfc_efx/base/efx.h
index 159e7957a3..996126217e 100644
--- a/drivers/common/sfc_efx/base/efx.h
+++ b/drivers/common/sfc_efx/base/efx.h
@@ -1511,6 +1511,7 @@ typedef struct efx_nic_cfg_s {
uint32_tenc_bist_mask;
 #endif /* EFSYS_OPT_BIST */
 #if EFSYS_OPT_RIVERHEAD || EFX_OPTS_EF10()
+   efx_pcie_interface_tenc_intf;
uint32_tenc_pf;
uint32_tenc_vf;
uint32_tenc_privilege_mask;
diff --git a/drivers/common/sfc_efx/base/efx_impl.h 
b/drivers/common/sfc_efx/base/efx_impl.h
index 992edbabe3..e0efbb8cdd 100644
--- a/drivers/common/sfc_efx/base/efx_impl.h
+++ b/drivers/common/sfc_efx/base/efx_impl.h
@@ -1529,6 +1529,12 @@ efx_mcdi_get_workarounds(
 
 #if EFSYS_OPT_RIVERHEAD || EFX_OPTS_EF10()
 
+LIBEFX_INTERNAL
+extern __checkReturn   efx_rc_t
+efx_mcdi_intf_from_pcie(
+   __inuint32_t pcie_intf,
+   __out   efx_pcie_interface_t *efx_intf);
+
 LIBEFX_INTERNAL
 extern __checkReturn   efx_rc_t
 efx_mcdi_init_evq(
diff --git a/drivers/common/sfc_efx/base/efx_mcdi.c 
b/drivers/common/sfc_efx/base/efx_mcdi.c
index b68fc0503d..69bf7ce70f 100644
--- a/drivers/common/sfc_efx/base/efx_mcdi.c
+++ b/drivers/common/sfc_efx/base/efx_mcdi.c
@@ -2130,6 +2130,36 @@ efx_mcdi_mac_stats_periodic(
 
 #if EFSYS_OPT_RIVERHEAD || EFX_OPTS_EF10()
 
+   __checkReturn   efx_rc_t
+efx_mcdi_intf_from_pcie(
+   __inuint32_t pcie_intf,
+   __out   efx_pcie_interface_t *efx_intf)
+{
+   efx_rc_t rc;
+
+   switch (pcie_intf) {
+   case PCIE_INTERFACE_CALLER:
+   *efx_intf = EFX_PCIE_INTERFACE_CALLER;
+   break;
+   case PCIE_INTERFACE_HOST_PRIMARY:
+   *efx_intf = EFX_PCIE_INTERFACE_HOST_PRIMARY;
+   break;
+   case PCIE_INTERFACE_NIC_EMBEDDED:
+   *efx_intf = EFX_PCIE_INTERFACE_NIC_EMBEDDED;
+   break;
+   default:
+   rc = EINVAL;
+   goto fail1;
+   }
+
+   return (0);
+
+fail1:
+   EFSYS_PROBE1(fail1, efx_rc_t, rc);
+
+   return (rc);
+}
+
 /*
  * This function returns the pf and vf number of a function.  If it is a pf the
  * vf number is 0x.  The vf number is the index of the vf on that
@@ -2140,18 +2170,21 @@ efx_mcdi_mac_stats_periodic(
 efx_mcdi_get_function_info(
__inefx_nic_t *enp,
__out   uint32_t *pfp,
-   __out_opt   uint32_t *vfp)
+   __out_opt   uint32_t *vfp,
+ 

[dpdk-dev] [PATCH v2 30/38] common/sfc_efx/base: add a means to read MAE mport journal

[Andrew Rybchenko]

From: Viacheslav Galaktionov 

This is required to provide the driver with the current state of mports.

Signed-off-by: Viacheslav Galaktionov 
Signed-off-by: Andrew Rybchenko 
Reviewed-by: Andy Moreton 
---
 drivers/common/sfc_efx/base/efx.h  |  56 +++
 drivers/common/sfc_efx/base/efx_mae.c  | 224 +
 drivers/common/sfc_efx/base/efx_mcdi.h |  54 ++
 drivers/common/sfc_efx/version.map |   1 +
 4 files changed, 335 insertions(+)

diff --git a/drivers/common/sfc_efx/base/efx.h 
b/drivers/common/sfc_efx/base/efx.h
index 996126217e..e77b297950 100644
--- a/drivers/common/sfc_efx/base/efx.h
+++ b/drivers/common/sfc_efx/base/efx.h
@@ -4205,6 +4205,42 @@ typedef struct efx_mport_id_s {
uint32_t id;
 } efx_mport_id_t;
 
+typedef enum efx_mport_type_e {
+   EFX_MPORT_TYPE_NET_PORT = 0,
+   EFX_MPORT_TYPE_ALIAS,
+   EFX_MPORT_TYPE_VNIC,
+} efx_mport_type_t;
+
+typedef enum efx_mport_vnic_client_type_e {
+   EFX_MPORT_VNIC_CLIENT_FUNCTION = 1,
+   EFX_MPORT_VNIC_CLIENT_PLUGIN,
+} efx_mport_vnic_client_type_t;
+
+typedef struct efx_mport_desc_s {
+   efx_mport_id_t  emd_id;
+   boolean_t   emd_can_receive_on;
+   boolean_t   emd_can_deliver_to;
+   boolean_t   emd_can_delete;
+   boolean_t   emd_zombie;
+   efx_mport_type_temd_type;
+   union {
+   struct {
+   uint32_tep_index;
+   } emd_net_port;
+   struct {
+   efx_mport_id_t  ea_target_mport_id;
+   } emd_alias;
+   struct {
+   efx_mport_vnic_client_type_tev_client_type;
+   efx_pcie_interface_tev_intf;
+   uint16_tev_pf;
+   uint16_tev_vf;
+   /* MCDI client handle for this VNIC. */
+   uint32_tev_handle;
+   } emd_vnic;
+   };
+} efx_mport_desc_t;
+
 #defineEFX_MPORT_NULL  (0U)
 
 /*
@@ -4635,6 +4671,26 @@ efx_mae_mport_free(
__inefx_nic_t *enp,
__inconst efx_mport_id_t *mportp);
 
+typedef __checkReturn  efx_rc_t
+(efx_mae_read_mport_journal_cb)(
+   __invoid *cb_datap,
+   __inefx_mport_desc_t *mportp,
+   __insize_t mport_len);
+
+/*
+ * Read mport descriptions from the MAE journal (which describes added and
+ * removed mports) and pass them to a user-supplied callback. The user gets
+ * only one chance to process the data it's given. Once the callback function
+ * finishes, that particular mport description will be gone.
+ * The journal will be fully repopulated on PCI reset (efx_nic_reset function).
+ */
+LIBEFX_API
+extern __checkReturn   efx_rc_t
+efx_mae_read_mport_journal(
+   __inefx_nic_t *enp,
+   __inefx_mae_read_mport_journal_cb *cbp,
+   __invoid *cb_datap);
+
 #endif /* EFSYS_OPT_MAE */
 
 #if EFSYS_OPT_VIRTIO
diff --git a/drivers/common/sfc_efx/base/efx_mae.c 
b/drivers/common/sfc_efx/base/efx_mae.c
index 37cc48eafc..110addd92d 100644
--- a/drivers/common/sfc_efx/base/efx_mae.c
+++ b/drivers/common/sfc_efx/base/efx_mae.c
@@ -3292,4 +3292,228 @@ efx_mae_mport_free(
return (rc);
 }
 
+static __checkReturn   efx_rc_t
+efx_mae_read_mport_journal_single(
+   __inuint8_t *entry_buf,
+   __out   efx_mport_desc_t *desc)
+{
+   uint32_t pcie_intf;
+   efx_rc_t rc;
+
+   memset(desc, 0, sizeof (*desc));
+
+   desc->emd_id.id = MCDI_STRUCT_DWORD(entry_buf,
+   MAE_MPORT_DESC_V2_MPORT_ID);
+
+   desc->emd_can_receive_on = MCDI_STRUCT_DWORD_FIELD(entry_buf,
+   MAE_MPORT_DESC_V2_FLAGS,
+   MAE_MPORT_DESC_V2_CAN_RECEIVE_ON);
+
+   desc->emd_can_deliver_to = MCDI_STRUCT_DWORD_FIELD(entry_buf,
+   MAE_MPORT_DESC_V2_FLAGS,
+   MAE_MPORT_DESC_V2_CAN_DELIVER_TO);
+
+   desc->emd_can_delete = MCDI_STRUCT_DWORD_FIELD(entry_buf,
+   MAE_MPORT_DESC_V2_FLAGS,
+   MAE_MPORT_DESC_V2_CAN_DELETE);
+
+   desc->emd_zombie = MCDI_STRUCT_DWORD_FIELD(entry_buf,
+   MAE_MPORT_DESC_V2_FLAGS,
+   MAE_MPORT_DESC_V2_IS_ZOMBIE);
+
+   desc->emd_type = MCDI_STRUCT_DWORD(entry_buf,
+   MAE_MPORT_DESC_V2_MPORT_TYPE);
+
+   /*
+* We can't check everything here. If some additional checks are
+* required, they should be performed by the callback function.
+*/
+   switch (desc->emd_type) {
+   case EFX_MPORT_TYPE_NET_PORT:
+   desc->emd_net_port.ep_index =
+

[dpdk-dev] [PATCH v2 31/38] common/sfc_efx/base: allow getting VNIC MCDI client handles

[Andrew Rybchenko]

From: Viacheslav Galaktionov 

Equality checks between VNICs should be done by comparing their client
handles. This means that clients should be able to retrieve client handles
for arbitrary functions and themselves.

Signed-off-by: Viacheslav Galaktionov 
Signed-off-by: Andrew Rybchenko 
Reviewed-by: Andy Moreton 
---
 drivers/common/sfc_efx/base/efx.h  | 15 ++
 drivers/common/sfc_efx/base/efx_mcdi.c | 73 ++
 drivers/common/sfc_efx/version.map |  2 +
 3 files changed, 90 insertions(+)

diff --git a/drivers/common/sfc_efx/base/efx.h 
b/drivers/common/sfc_efx/base/efx.h
index e77b297950..b61984a8e3 100644
--- a/drivers/common/sfc_efx/base/efx.h
+++ b/drivers/common/sfc_efx/base/efx.h
@@ -391,6 +391,21 @@ extern __checkReturn   boolean_t
 efx_mcdi_request_abort(
__inefx_nic_t *enp);
 
+LIBEFX_API
+extern __checkReturn   efx_rc_t
+efx_mcdi_get_client_handle(
+   __inefx_nic_t *enp,
+   __inefx_pcie_interface_t intf,
+   __inuint16_t pf,
+   __inuint16_t vf,
+   __out   uint32_t *handle);
+
+LIBEFX_API
+extern __checkReturn   efx_rc_t
+efx_mcdi_get_own_client_handle(
+   __inefx_nic_t *enp,
+   __out   uint32_t *handle);
+
 LIBEFX_API
 extern void
 efx_mcdi_fini(
diff --git a/drivers/common/sfc_efx/base/efx_mcdi.c 
b/drivers/common/sfc_efx/base/efx_mcdi.c
index 69bf7ce70f..cdf7181e0d 100644
--- a/drivers/common/sfc_efx/base/efx_mcdi.c
+++ b/drivers/common/sfc_efx/base/efx_mcdi.c
@@ -647,6 +647,79 @@ efx_mcdi_request_abort(
return (aborted);
 }
 
+   __checkReturn   efx_rc_t
+efx_mcdi_get_client_handle(
+   __inefx_nic_t *enp,
+   __inefx_pcie_interface_t intf,
+   __inuint16_t pf,
+   __inuint16_t vf,
+   __out   uint32_t *handle)
+{
+   efx_mcdi_req_t req;
+   EFX_MCDI_DECLARE_BUF(payload,
+   MC_CMD_GET_CLIENT_HANDLE_IN_LEN,
+   MC_CMD_GET_CLIENT_HANDLE_OUT_LEN);
+   efx_rc_t rc;
+
+   if (handle == NULL) {
+   rc = EINVAL;
+   goto fail1;
+   }
+
+   req.emr_cmd = MC_CMD_GET_CLIENT_HANDLE;
+   req.emr_in_buf = payload;
+   req.emr_in_length = MC_CMD_GET_CLIENT_HANDLE_IN_LEN;
+   req.emr_out_buf = payload;
+   req.emr_out_length = MC_CMD_GET_CLIENT_HANDLE_OUT_LEN;
+
+   MCDI_IN_SET_DWORD(req, GET_CLIENT_HANDLE_IN_TYPE,
+   MC_CMD_GET_CLIENT_HANDLE_IN_TYPE_FUNC);
+   MCDI_IN_SET_WORD(req, GET_CLIENT_HANDLE_IN_FUNC_PF, pf);
+   MCDI_IN_SET_WORD(req, GET_CLIENT_HANDLE_IN_FUNC_VF, vf);
+   MCDI_IN_SET_DWORD(req, GET_CLIENT_HANDLE_IN_FUNC_INTF, intf);
+
+   efx_mcdi_execute(enp, &req);
+
+   if (req.emr_rc != 0) {
+   rc = req.emr_rc;
+   goto fail2;
+   }
+
+   if (req.emr_out_length_used < MC_CMD_GET_CLIENT_HANDLE_OUT_LEN) {
+   rc = EMSGSIZE;
+   goto fail3;
+   }
+
+   *handle = MCDI_OUT_DWORD(req, GET_CLIENT_HANDLE_OUT_HANDLE);
+
+   return 0;
+fail3:
+   EFSYS_PROBE(fail3);
+fail2:
+   EFSYS_PROBE(fail2);
+fail1:
+   EFSYS_PROBE1(fail1, efx_rc_t, rc);
+   return (rc);
+}
+
+   __checkReturn   efx_rc_t
+efx_mcdi_get_own_client_handle(
+   __inefx_nic_t *enp,
+   __out   uint32_t *handle)
+{
+   efx_rc_t rc;
+
+   rc = efx_mcdi_get_client_handle(enp, PCIE_INTERFACE_CALLER,
+   PCIE_FUNCTION_PF_NULL, PCIE_FUNCTION_VF_NULL, handle);
+   if (rc != 0)
+   goto fail1;
+
+   return (0);
+fail1:
+   EFSYS_PROBE1(fail1, efx_rc_t, rc);
+   return (rc);
+}
+
void
 efx_mcdi_get_timeout(
__inefx_nic_t *enp,
diff --git a/drivers/common/sfc_efx/version.map 
b/drivers/common/sfc_efx/version.map
index 10216bb37d..346deb4b12 100644
--- a/drivers/common/sfc_efx/version.map
+++ b/drivers/common/sfc_efx/version.map
@@ -136,6 +136,8 @@ INTERNAL {
efx_mae_read_mport_journal;
 
efx_mcdi_fini;
+   efx_mcdi_get_client_handle;
+   efx_mcdi_get_own_client_handle;
efx_mcdi_get_proxy_handle;
efx_mcdi_get_timeout;
efx_mcdi_init;
-- 
2.30.2

[dpdk-dev] [PATCH v2 32/38] net/sfc: maintain controller to EFX interface mapping

[Andrew Rybchenko]

From: Viacheslav Galaktionov 

Newer hardware may have arbitrarily complex controller configurations,
and for this reason the mapping has been made dynamic: it is represented
with a dynamic array that is indexed by controller numbers and each
element contains an EFX interface number. Since the number of controllers
is expected to be small, this approach should not hurt the performance.

Signed-off-by: Viacheslav Galaktionov 
Signed-off-by: Andrew Rybchenko 
Reviewed-by: Andy Moreton 
---
 drivers/net/sfc/sfc_ethdev.c | 184 +++
 drivers/net/sfc/sfc_switch.c |  57 +++
 drivers/net/sfc/sfc_switch.h |   8 ++
 3 files changed, 249 insertions(+)

diff --git a/drivers/net/sfc/sfc_ethdev.c b/drivers/net/sfc/sfc_ethdev.c
index f69bbde11a..f93b9cc921 100644
--- a/drivers/net/sfc/sfc_ethdev.c
+++ b/drivers/net/sfc/sfc_ethdev.c
@@ -30,6 +30,7 @@
 #include "sfc_dp_rx.h"
 #include "sfc_repr.h"
 #include "sfc_sw_stats.h"
+#include "sfc_switch.h"
 
 #define SFC_XSTAT_ID_INVALID_VAL  UINT64_MAX
 #define SFC_XSTAT_ID_INVALID_NAME '\0'
@@ -1863,6 +1864,177 @@ sfc_rx_queue_intr_disable(struct rte_eth_dev *dev, 
uint16_t ethdev_qid)
return sap->dp_rx->intr_disable(rxq_info->dp);
 }
 
+struct sfc_mport_journal_ctx {
+   struct sfc_adapter  *sa;
+   uint16_tswitch_domain_id;
+   uint32_tmcdi_handle;
+   boolcontrollers_assigned;
+   efx_pcie_interface_t*controllers;
+   size_t  nb_controllers;
+};
+
+static int
+sfc_journal_ctx_add_controller(struct sfc_mport_journal_ctx *ctx,
+  efx_pcie_interface_t intf)
+{
+   efx_pcie_interface_t *new_controllers;
+   size_t i, target;
+   size_t new_size;
+
+   if (ctx->controllers == NULL) {
+   ctx->controllers = rte_malloc("sfc_controller_mapping",
+ sizeof(ctx->controllers[0]), 0);
+   if (ctx->controllers == NULL)
+   return ENOMEM;
+
+   ctx->controllers[0] = intf;
+   ctx->nb_controllers = 1;
+
+   return 0;
+   }
+
+   for (i = 0; i < ctx->nb_controllers; i++) {
+   if (ctx->controllers[i] == intf)
+   return 0;
+   if (ctx->controllers[i] > intf)
+   break;
+   }
+   target = i;
+
+   ctx->nb_controllers += 1;
+   new_size = ctx->nb_controllers * sizeof(ctx->controllers[0]);
+
+   new_controllers = rte_realloc(ctx->controllers, new_size, 0);
+   if (new_controllers == NULL) {
+   rte_free(ctx->controllers);
+   return ENOMEM;
+   }
+   ctx->controllers = new_controllers;
+
+   for (i = target + 1; i < ctx->nb_controllers; i++)
+   ctx->controllers[i] = ctx->controllers[i - 1];
+
+   ctx->controllers[target] = intf;
+
+   return 0;
+}
+
+static efx_rc_t
+sfc_process_mport_journal_entry(struct sfc_mport_journal_ctx *ctx,
+   efx_mport_desc_t *mport)
+{
+   efx_mport_sel_t ethdev_mport;
+   int rc;
+
+   sfc_dbg(ctx->sa,
+   "processing mport id %u (controller %u pf %u vf %u)",
+   mport->emd_id.id, mport->emd_vnic.ev_intf,
+   mport->emd_vnic.ev_pf, mport->emd_vnic.ev_vf);
+   efx_mae_mport_invalid(ðdev_mport);
+
+   if (!ctx->controllers_assigned) {
+   rc = sfc_journal_ctx_add_controller(ctx,
+   mport->emd_vnic.ev_intf);
+   if (rc != 0)
+   return rc;
+   }
+
+   return 0;
+}
+
+static efx_rc_t
+sfc_process_mport_journal_cb(void *data, efx_mport_desc_t *mport,
+size_t mport_len)
+{
+   struct sfc_mport_journal_ctx *ctx = data;
+
+   if (ctx == NULL || ctx->sa == NULL) {
+   sfc_err(ctx->sa, "received NULL context or SFC adapter");
+   return EINVAL;
+   }
+
+   if (mport_len != sizeof(*mport)) {
+   sfc_err(ctx->sa, "actual and expected mport buffer sizes 
differ");
+   return EINVAL;
+   }
+
+   SFC_ASSERT(sfc_adapter_is_locked(ctx->sa));
+
+   /*
+* If a zombie flag is set, it means the mport has been marked for
+* deletion and cannot be used for any new operations. The mport will
+* be destroyed completely once all references to it are released.
+*/
+   if (mport->emd_zombie) {
+   sfc_dbg(ctx->sa, "mport is a zombie, skipping");
+   return 0;
+   }
+   if (mport->emd_type != EFX_MPORT_TYPE_VNIC) {
+   sfc_dbg(ctx->sa, "mport is not a VNIC, skipping");
+   return 0;
+   }
+   if (mport->emd_vnic.ev_client_type != EFX_MPORT_VNIC_CLIENT_FUNCTION) {
+   sfc_dbg(ctx->sa, "mport is n

[dpdk-dev] [PATCH v2 33/38] net/sfc: store PCI address for represented entities

[Andrew Rybchenko]

From: Viacheslav Galaktionov 

This information will be useful when representor info API is implemented.

Signed-off-by: Viacheslav Galaktionov 
Signed-off-by: Andrew Rybchenko 
Reviewed-by: Andy Moreton 
---
 drivers/net/sfc/sfc_ethdev.c | 11 +--
 drivers/net/sfc/sfc_repr.c   | 20 +++-
 drivers/net/sfc/sfc_repr.h   | 10 +-
 drivers/net/sfc/sfc_switch.c | 14 ++
 drivers/net/sfc/sfc_switch.h | 11 +++
 5 files changed, 58 insertions(+), 8 deletions(-)

diff --git a/drivers/net/sfc/sfc_ethdev.c b/drivers/net/sfc/sfc_ethdev.c
index f93b9cc921..53008f477f 100644
--- a/drivers/net/sfc/sfc_ethdev.c
+++ b/drivers/net/sfc/sfc_ethdev.c
@@ -2688,6 +2688,7 @@ sfc_eth_dev_create_representors(struct rte_eth_dev *dev,
 
for (i = 0; i < eth_da->nb_representor_ports; ++i) {
const efx_nic_cfg_t *encp = efx_nic_cfg_get(sa->nic);
+   struct sfc_repr_entity_info entity;
efx_mport_sel_t mport_sel;
 
rc = efx_mae_mport_by_pcie_function(encp->enc_pf,
@@ -2700,8 +2701,14 @@ sfc_eth_dev_create_representors(struct rte_eth_dev *dev,
continue;
}
 
-   rc = sfc_repr_create(dev, eth_da->representor_ports[i],
-sa->mae.switch_domain_id, &mport_sel);
+   memset(&entity, 0, sizeof(entity));
+   entity.type = eth_da->type;
+   entity.intf = encp->enc_intf;
+   entity.pf = encp->enc_pf;
+   entity.vf = eth_da->representor_ports[i];
+
+   rc = sfc_repr_create(dev, &entity, sa->mae.switch_domain_id,
+&mport_sel);
if (rc != 0) {
sfc_err(sa, "cannot create representor %u: %s - ignore",
eth_da->representor_ports[i],
diff --git a/drivers/net/sfc/sfc_repr.c b/drivers/net/sfc/sfc_repr.c
index 87f10092c3..f87188ed7a 100644
--- a/drivers/net/sfc/sfc_repr.c
+++ b/drivers/net/sfc/sfc_repr.c
@@ -902,6 +902,9 @@ struct sfc_repr_init_data {
uint16_trepr_id;
uint16_tswitch_domain_id;
efx_mport_sel_t mport_sel;
+   efx_pcie_interface_tintf;
+   uint16_tpf;
+   uint16_tvf;
 };
 
 static int
@@ -939,6 +942,9 @@ sfc_repr_eth_dev_init(struct rte_eth_dev *dev, void 
*init_params)
switch_port_request.ethdev_mportp = ðdev_mport_sel;
switch_port_request.entity_mportp = &repr_data->mport_sel;
switch_port_request.ethdev_port_id = dev->data->port_id;
+   switch_port_request.port_data.repr.intf = repr_data->intf;
+   switch_port_request.port_data.repr.pf = repr_data->pf;
+   switch_port_request.port_data.repr.vf = repr_data->vf;
 
ret = sfc_repr_assign_mae_switch_port(repr_data->switch_domain_id,
  &switch_port_request,
@@ -1015,8 +1021,10 @@ sfc_repr_eth_dev_init(struct rte_eth_dev *dev, void 
*init_params)
 }
 
 int
-sfc_repr_create(struct rte_eth_dev *parent, uint16_t representor_id,
-   uint16_t switch_domain_id, const efx_mport_sel_t *mport_sel)
+sfc_repr_create(struct rte_eth_dev *parent,
+   struct sfc_repr_entity_info *entity,
+   uint16_t switch_domain_id,
+   const efx_mport_sel_t *mport_sel)
 {
struct sfc_repr_init_data repr_data;
char name[RTE_ETH_NAME_MAX_LEN];
@@ -1024,8 +1032,7 @@ sfc_repr_create(struct rte_eth_dev *parent, uint16_t 
representor_id,
struct rte_eth_dev *dev;
 
if (snprintf(name, sizeof(name), "net_%s_representor_%u",
-parent->device->name, representor_id) >=
-   (int)sizeof(name)) {
+parent->device->name, entity->vf) >= (int)sizeof(name)) {
SFC_GENERIC_LOG(ERR, "%s() failed name too long", __func__);
return -ENAMETOOLONG;
}
@@ -1034,9 +1041,12 @@ sfc_repr_create(struct rte_eth_dev *parent, uint16_t 
representor_id,
if (dev == NULL) {
memset(&repr_data, 0, sizeof(repr_data));
repr_data.pf_port_id = parent->data->port_id;
-   repr_data.repr_id = representor_id;
+   repr_data.repr_id = entity->vf;
repr_data.switch_domain_id = switch_domain_id;
repr_data.mport_sel = *mport_sel;
+   repr_data.intf = entity->intf;
+   repr_data.pf = entity->pf;
+   repr_data.vf = entity->vf;
 
ret = rte_eth_dev_create(parent->device, name,
 sizeof(struct sfc_repr_shared),
diff --git a/drivers/net/sfc/sfc_repr.h b/drivers/net/sfc/sfc_repr.h
index 1347206006..2093973761 100644
--- a/drivers/net/sfc/sfc_repr.h
+++ b/drivers/net/sfc/sfc_repr.h
@@ -26,7 +26,15 @@ extern "C" {
 /** Max count of the representor Tx queues */
 #define

[dpdk-dev] [PATCH v2 34/38] net/sfc: include controller and port in representor name

[Andrew Rybchenko]

From: Viacheslav Galaktionov 

Make representor names unique on multi-host configurations.

Signed-off-by: Viacheslav Galaktionov 
Signed-off-by: Andrew Rybchenko 
Reviewed-by: Andy Moreton 
---
 drivers/net/sfc/sfc_repr.c   | 28 ++--
 drivers/net/sfc/sfc_switch.c | 28 
 drivers/net/sfc/sfc_switch.h |  4 
 3 files changed, 58 insertions(+), 2 deletions(-)

diff --git a/drivers/net/sfc/sfc_repr.c b/drivers/net/sfc/sfc_repr.c
index f87188ed7a..b4ff4da60a 100644
--- a/drivers/net/sfc/sfc_repr.c
+++ b/drivers/net/sfc/sfc_repr.c
@@ -1028,11 +1028,35 @@ sfc_repr_create(struct rte_eth_dev *parent,
 {
struct sfc_repr_init_data repr_data;
char name[RTE_ETH_NAME_MAX_LEN];
+   int controller;
int ret;
+   int rc;
struct rte_eth_dev *dev;
 
-   if (snprintf(name, sizeof(name), "net_%s_representor_%u",
-parent->device->name, entity->vf) >= (int)sizeof(name)) {
+   controller = -1;
+   rc = sfc_mae_switch_domain_get_controller(switch_domain_id,
+ entity->intf, &controller);
+   if (rc != 0) {
+   SFC_GENERIC_LOG(ERR, "%s() failed to get DPDK controller for 
%d",
+   __func__, entity->intf);
+   return -rc;
+   }
+
+   switch (entity->type) {
+   case RTE_ETH_REPRESENTOR_VF:
+   ret = snprintf(name, sizeof(name), 
"net_%s_representor_c%upf%uvf%u",
+  parent->device->name, controller, entity->pf,
+  entity->vf);
+   break;
+   case RTE_ETH_REPRESENTOR_PF:
+   ret = snprintf(name, sizeof(name), "net_%s_representor_c%upf%u",
+  parent->device->name, controller, entity->pf);
+   break;
+   default:
+   return -ENOTSUP;
+   }
+
+   if (ret >= (int)sizeof(name)) {
SFC_GENERIC_LOG(ERR, "%s() failed name too long", __func__);
return -ENAMETOOLONG;
}
diff --git a/drivers/net/sfc/sfc_switch.c b/drivers/net/sfc/sfc_switch.c
index 7a0b332f33..225d07fa15 100644
--- a/drivers/net/sfc/sfc_switch.c
+++ b/drivers/net/sfc/sfc_switch.c
@@ -279,6 +279,34 @@ sfc_mae_switch_domain_map_controllers(uint16_t 
switch_domain_id,
return 0;
 }
 
+int
+sfc_mae_switch_domain_get_controller(uint16_t switch_domain_id,
+efx_pcie_interface_t intf,
+int *controller)
+{
+   const efx_pcie_interface_t *controllers;
+   size_t nb_controllers;
+   size_t i;
+   int rc;
+
+   rc = sfc_mae_switch_domain_controllers(switch_domain_id, &controllers,
+  &nb_controllers);
+   if (rc != 0)
+   return rc;
+
+   if (controllers == NULL)
+   return ENOENT;
+
+   for (i = 0; i < nb_controllers; i++) {
+   if (controllers[i] == intf) {
+   *controller = i;
+   return 0;
+   }
+   }
+
+   return ENOENT;
+}
+
 /* This function expects to be called only when the lock is held */
 static struct sfc_mae_switch_port *
 sfc_mae_find_switch_port_by_entity(const struct sfc_mae_switch_domain *domain,
diff --git a/drivers/net/sfc/sfc_switch.h b/drivers/net/sfc/sfc_switch.h
index a072507375..294baae9a2 100644
--- a/drivers/net/sfc/sfc_switch.h
+++ b/drivers/net/sfc/sfc_switch.h
@@ -63,6 +63,10 @@ int sfc_mae_switch_domain_map_controllers(uint16_t 
switch_domain_id,
  efx_pcie_interface_t *controllers,
  size_t nb_controllers);
 
+int sfc_mae_switch_domain_get_controller(uint16_t switch_domain_id,
+  efx_pcie_interface_t intf,
+  int *controller);
+
 int sfc_mae_assign_switch_port(uint16_t switch_domain_id,
   const struct sfc_mae_switch_port_request *req,
   uint16_t *switch_port_id);
-- 
2.30.2

[dpdk-dev] [PATCH v2 36/38] net/sfc: use switch port ID as representor ID

[Andrew Rybchenko]

From: Viacheslav Galaktionov 

Representor IDs must be unique for each representor. VFs, which are
currently used, are not unique as they may repeat in combination with
different PCI controllers and PFs. On the other hand, switch port IDs
are unique, so they are a better fit for this role.

Signed-off-by: Viacheslav Galaktionov 
Signed-off-by: Andrew Rybchenko 
Reviewed-by: Andy Moreton 
---
 drivers/net/sfc/sfc_repr.c | 8 +++-
 1 file changed, 3 insertions(+), 5 deletions(-)

diff --git a/drivers/net/sfc/sfc_repr.c b/drivers/net/sfc/sfc_repr.c
index b4ff4da60a..6ec83873ab 100644
--- a/drivers/net/sfc/sfc_repr.c
+++ b/drivers/net/sfc/sfc_repr.c
@@ -899,7 +899,6 @@ static const struct eth_dev_ops sfc_repr_dev_ops = {
 
 struct sfc_repr_init_data {
uint16_tpf_port_id;
-   uint16_trepr_id;
uint16_tswitch_domain_id;
efx_mport_sel_t mport_sel;
efx_pcie_interface_tintf;
@@ -957,7 +956,7 @@ sfc_repr_eth_dev_init(struct rte_eth_dev *dev, void 
*init_params)
}
 
ret = sfc_repr_proxy_add_port(repr_data->pf_port_id,
- repr_data->repr_id,
+ srs->switch_port_id,
  dev->data->port_id,
  &repr_data->mport_sel);
if (ret != 0) {
@@ -984,7 +983,7 @@ sfc_repr_eth_dev_init(struct rte_eth_dev *dev, void 
*init_params)
dev->process_private = sr;
 
srs->pf_port_id = repr_data->pf_port_id;
-   srs->repr_id = repr_data->repr_id;
+   srs->repr_id = srs->switch_port_id;
srs->switch_domain_id = repr_data->switch_domain_id;
 
dev->data->dev_flags |= RTE_ETH_DEV_REPRESENTOR;
@@ -1012,7 +1011,7 @@ sfc_repr_eth_dev_init(struct rte_eth_dev *dev, void 
*init_params)
 
 fail_alloc_sr:
(void)sfc_repr_proxy_del_port(repr_data->pf_port_id,
- repr_data->repr_id);
+ srs->switch_port_id);
 
 fail_create_port:
 fail_mae_assign_switch_port:
@@ -1065,7 +1064,6 @@ sfc_repr_create(struct rte_eth_dev *parent,
if (dev == NULL) {
memset(&repr_data, 0, sizeof(repr_data));
repr_data.pf_port_id = parent->data->port_id;
-   repr_data.repr_id = entity->vf;
repr_data.switch_domain_id = switch_domain_id;
repr_data.mport_sel = *mport_sel;
repr_data.intf = entity->intf;
-- 
2.30.2

[dpdk-dev] [PATCH v2 35/38] net/sfc: support new representor parameter syntax

[Andrew Rybchenko]

From: Viacheslav Galaktionov 

Allow the user to specify representor entities using the structured
parameter values.

Signed-off-by: Viacheslav Galaktionov 
Signed-off-by: Andrew Rybchenko 
Reviewed-by: Andy Moreton 
---
 drivers/net/sfc/sfc_ethdev.c | 181 ---
 drivers/net/sfc/sfc_switch.c |  24 +
 drivers/net/sfc/sfc_switch.h |   4 +
 3 files changed, 176 insertions(+), 33 deletions(-)

diff --git a/drivers/net/sfc/sfc_ethdev.c b/drivers/net/sfc/sfc_ethdev.c
index 53008f477f..69ab2a60b0 100644
--- a/drivers/net/sfc/sfc_ethdev.c
+++ b/drivers/net/sfc/sfc_ethdev.c
@@ -2650,18 +2650,143 @@ sfc_eth_dev_find_or_create(struct rte_pci_device 
*pci_dev,
return 0;
 }
 
+static int
+sfc_eth_dev_create_repr(struct sfc_adapter *sa,
+   efx_pcie_interface_t controller,
+   uint16_t port,
+   uint16_t repr_port,
+   enum rte_eth_representor_type type)
+{
+   struct sfc_repr_entity_info entity;
+   efx_mport_sel_t mport_sel;
+   int rc;
+
+   switch (type) {
+   case RTE_ETH_REPRESENTOR_NONE:
+   return 0;
+   case RTE_ETH_REPRESENTOR_VF:
+   case RTE_ETH_REPRESENTOR_PF:
+   break;
+   case RTE_ETH_REPRESENTOR_SF:
+   sfc_err(sa, "SF representors are not supported");
+   return ENOTSUP;
+   default:
+   sfc_err(sa, "unknown representor type: %d", type);
+   return ENOTSUP;
+   }
+
+   rc = efx_mae_mport_by_pcie_mh_function(controller,
+  port,
+  repr_port,
+  &mport_sel);
+   if (rc != 0) {
+   sfc_err(sa,
+   "failed to get m-port selector for controller %u port 
%u repr_port %u: %s",
+   controller, port, repr_port, rte_strerror(-rc));
+   return rc;
+   }
+
+   memset(&entity, 0, sizeof(entity));
+   entity.type = type;
+   entity.intf = controller;
+   entity.pf = port;
+   entity.vf = repr_port;
+
+   rc = sfc_repr_create(sa->eth_dev, &entity, sa->mae.switch_domain_id,
+&mport_sel);
+   if (rc != 0) {
+   sfc_err(sa,
+   "failed to create representor for controller %u port %u 
repr_port %u: %s",
+   controller, port, repr_port, rte_strerror(-rc));
+   return rc;
+   }
+
+   return 0;
+}
+
+static int
+sfc_eth_dev_create_repr_port(struct sfc_adapter *sa,
+const struct rte_eth_devargs *eth_da,
+efx_pcie_interface_t controller,
+uint16_t port)
+{
+   int first_error = 0;
+   uint16_t i;
+   int rc;
+
+   if (eth_da->type == RTE_ETH_REPRESENTOR_PF) {
+   return sfc_eth_dev_create_repr(sa, controller, port,
+  EFX_PCI_VF_INVALID,
+  eth_da->type);
+   }
+
+   for (i = 0; i < eth_da->nb_representor_ports; i++) {
+   rc = sfc_eth_dev_create_repr(sa, controller, port,
+eth_da->representor_ports[i],
+eth_da->type);
+   if (rc != 0 && first_error == 0)
+   first_error = rc;
+   }
+
+   return first_error;
+}
+
+static int
+sfc_eth_dev_create_repr_controller(struct sfc_adapter *sa,
+  const struct rte_eth_devargs *eth_da,
+  efx_pcie_interface_t controller)
+{
+   const efx_nic_cfg_t *encp;
+   int first_error = 0;
+   uint16_t default_port;
+   uint16_t i;
+   int rc;
+
+   if (eth_da->nb_ports == 0) {
+   encp = efx_nic_cfg_get(sa->nic);
+   default_port = encp->enc_intf == controller ? encp->enc_pf : 0;
+   return sfc_eth_dev_create_repr_port(sa, eth_da, controller,
+   default_port);
+   }
+
+   for (i = 0; i < eth_da->nb_ports; i++) {
+   rc = sfc_eth_dev_create_repr_port(sa, eth_da, controller,
+ eth_da->ports[i]);
+   if (rc != 0 && first_error == 0)
+   first_error = rc;
+   }
+
+   return first_error;
+}
+
 static int
 sfc_eth_dev_create_representors(struct rte_eth_dev *dev,
const struct rte_eth_devargs *eth_da)
 {
+   efx_pcie_interface_t intf;
+   const efx_nic_cfg_t *encp;
struct sfc_adapter *sa;
-   unsigned int i;
+   uint16_t switch_domain_id;
+   uint16_t i;
int rc;
 
-   if (eth_da->nb_representor_ports == 0)
-   return 0;
-
sa = sfc_adapter

[dpdk-dev] [PATCH v2 38/38] net/sfc: update comment about representor support

[Andrew Rybchenko]

From: Viacheslav Galaktionov 

The representor support has been implemented to some extent, and the fact
that ethdev mport is equivalent to entity mport is by design.

Fixes: 1fb65e4dae8 ("net/sfc: support flow action port ID in transfer rules")
Cc: sta...@dpdk.org

Signed-off-by: Viacheslav Galaktionov 
Signed-off-by: Andrew Rybchenko 
Reviewed-by: Andy Moreton 
---
 drivers/net/sfc/sfc_mae.c | 5 +
 1 file changed, 1 insertion(+), 4 deletions(-)

diff --git a/drivers/net/sfc/sfc_mae.c b/drivers/net/sfc/sfc_mae.c
index 7be77054ab..fa60c948ca 100644
--- a/drivers/net/sfc/sfc_mae.c
+++ b/drivers/net/sfc/sfc_mae.c
@@ -228,10 +228,7 @@ sfc_mae_attach(struct sfc_adapter *sa)
sfc_log_init(sa, "assign RTE switch port");
switch_port_request.type = SFC_MAE_SWITCH_PORT_INDEPENDENT;
switch_port_request.entity_mportp = &entity_mport;
-   /*
-* As of now, the driver does not support representors, so
-* RTE ethdev MPORT simply matches that of the entity.
-*/
+   /* RTE ethdev MPORT matches that of the entity for independent ports. */
switch_port_request.ethdev_mportp = &entity_mport;
switch_port_request.ethdev_port_id = sas->port_id;
rc = sfc_mae_assign_switch_port(mae->switch_domain_id,
-- 
2.30.2

[dpdk-dev] [PATCH v2 37/38] net/sfc: implement the representor info API

[Andrew Rybchenko]

From: Viacheslav Galaktionov 

Let the driver provide the user with information about available
representors by implementing the representor_info_get operation.

Due to the lack of any structure to representor IDs, every ID range
describes exactly one representor.

Signed-off-by: Viacheslav Galaktionov 
Signed-off-by: Andrew Rybchenko 
Reviewed-by: Andy Moreton 
---
 doc/guides/rel_notes/release_21_11.rst |   6 +
 drivers/net/sfc/sfc_ethdev.c   | 229 +
 drivers/net/sfc/sfc_switch.c   | 104 +--
 drivers/net/sfc/sfc_switch.h   |  24 +++
 4 files changed, 352 insertions(+), 11 deletions(-)

diff --git a/doc/guides/rel_notes/release_21_11.rst 
b/doc/guides/rel_notes/release_21_11.rst
index 89d4b33ef1..a89dcb2c63 100644
--- a/doc/guides/rel_notes/release_21_11.rst
+++ b/doc/guides/rel_notes/release_21_11.rst
@@ -101,6 +101,12 @@ New Features
   * Default VLAN strip behavior is changed. VLAN tag won't be stripped unless
 ``DEV_RX_OFFLOAD_VLAN_STRIP`` offload is enabled.
 
+* **Updated Solarflare network PMD.**
+
+  Updated the Solarflare ``sfc_efx`` driver with changes including:
+
+  * Added port representors support on SN1000 SmartNICs
+
 * **Updated Marvell cnxk crypto PMD.**
 
   * Added AES-CBC SHA1-HMAC support in lookaside protocol (IPsec) for CN10K.
diff --git a/drivers/net/sfc/sfc_ethdev.c b/drivers/net/sfc/sfc_ethdev.c
index 69ab2a60b0..54711d349f 100644
--- a/drivers/net/sfc/sfc_ethdev.c
+++ b/drivers/net/sfc/sfc_ethdev.c
@@ -1923,7 +1923,11 @@ static efx_rc_t
 sfc_process_mport_journal_entry(struct sfc_mport_journal_ctx *ctx,
efx_mport_desc_t *mport)
 {
+   struct sfc_mae_switch_port_request req;
+   efx_mport_sel_t entity_selector;
efx_mport_sel_t ethdev_mport;
+   uint16_t switch_port_id;
+   efx_rc_t efx_rc;
int rc;
 
sfc_dbg(ctx->sa,
@@ -1939,6 +1943,63 @@ sfc_process_mport_journal_entry(struct 
sfc_mport_journal_ctx *ctx,
return rc;
}
 
+   /* Build Mport selector */
+   efx_rc = efx_mae_mport_by_pcie_mh_function(mport->emd_vnic.ev_intf,
+   mport->emd_vnic.ev_pf,
+   mport->emd_vnic.ev_vf,
+   &entity_selector);
+   if (efx_rc != 0) {
+   sfc_err(ctx->sa, "failed to build entity mport selector for 
c%upf%uvf%u",
+   mport->emd_vnic.ev_intf,
+   mport->emd_vnic.ev_pf,
+   mport->emd_vnic.ev_vf);
+   return efx_rc;
+   }
+
+   rc = sfc_mae_switch_port_id_by_entity(ctx->switch_domain_id,
+ &entity_selector,
+ SFC_MAE_SWITCH_PORT_REPRESENTOR,
+ &switch_port_id);
+   switch (rc) {
+   case 0:
+   /* Already registered */
+   break;
+   case ENOENT:
+   /*
+* No representor has been created for this entity.
+* Create a dummy switch registry entry with an invalid ethdev
+* mport selector. When a corresponding representor is created,
+* this entry will be updated.
+*/
+   req.type = SFC_MAE_SWITCH_PORT_REPRESENTOR;
+   req.entity_mportp = &entity_selector;
+   req.ethdev_mportp = ðdev_mport;
+   req.ethdev_port_id = RTE_MAX_ETHPORTS;
+   req.port_data.repr.intf = mport->emd_vnic.ev_intf;
+   req.port_data.repr.pf = mport->emd_vnic.ev_pf;
+   req.port_data.repr.vf = mport->emd_vnic.ev_vf;
+
+   rc = sfc_mae_assign_switch_port(ctx->switch_domain_id,
+   &req, &switch_port_id);
+   if (rc != 0) {
+   sfc_err(ctx->sa,
+   "failed to assign MAE switch port for 
c%upf%uvf%u: %s",
+   mport->emd_vnic.ev_intf,
+   mport->emd_vnic.ev_pf,
+   mport->emd_vnic.ev_vf,
+   rte_strerror(rc));
+   return rc;
+   }
+   break;
+   default:
+   sfc_err(ctx->sa, "failed to find MAE switch port for 
c%upf%uvf%u: %s",
+   mport->emd_vnic.ev_intf,
+   mport->emd_vnic.ev_pf,
+   mport->emd_vnic.ev_vf,
+   rte_strerror(rc));
+   return rc;
+   }
+
return 0;
 }
 
@@ -2035,6 +2096,173 @@ sfc_process_mport_journal(struct sfc_adapter *sa)
return 0;
 }
 
+static void
+sfc_count_representors_cb(enum sfc_mae_switch_port_type type,
+ const efx_mport_sel_t *ethdev_mportp __rte_unused,
+

Re: [dpdk-dev] [PATCH v2 1/5] cryptodev: separate out internal structures

[Zhang, Roy Fan]

> -Original Message-
> From: Akhil Goyal 
> Sent: Monday, October 11, 2021 1:43 PM
> To: dev@dpdk.org
> Cc: tho...@monjalon.net; david.march...@redhat.com;
> hemant.agra...@nxp.com; ano...@marvell.com; De Lara Guarch, Pablo
> ; Trahe, Fiona ;
> Doherty, Declan ; ma...@nvidia.com;
> g.si...@nxp.com; Zhang, Roy Fan ;
> jianjay.z...@huawei.com; asoma...@amd.com; ruifeng.w...@arm.com;
> Ananyev, Konstantin ; Nicolau, Radu
> ; ajit.khapa...@broadcom.com;
> rnagadhee...@marvell.com; adwiv...@marvell.com; Power, Ciara
> ; Akhil Goyal 
> Subject: [PATCH v2 1/5] cryptodev: separate out internal structures
> 
> A new header file rte_cryptodev_core.h is added and all
> internal data structures which need not be exposed directly to
> application are moved to this file. These structures are mostly
> used by drivers, but they need to be in the public header file
> as they are accessed by datapath inline functions for
> performance reasons.
> 
> Signed-off-by: Akhil Goyal 
> ---

Acked-by: Fan Zhang 

Re: [dpdk-dev] [PATCH] mbuf: remove deprecated bad outer IPv4 checksum flag on Rx

[Andrew Rybchenko]

Hi Olivier,

On 9/30/21 12:51 AM, Olivier Matz wrote:
> Hi Andrew,
> 
> On Wed, Sep 29, 2021 at 10:37:34AM +0300, Andrew Rybchenko wrote:
>> Removed offload flag PKT_RX_EIP_CKSUM_BAD. PKT_RX_OUTER_IP_CKSUM_BAD
>> should be used as a replacement.
>>
>> Signed-off-by: Andrew Rybchenko 
> 
> FYI, I just submitted a patchset that removes all PKT_ flags and
> fixes the namespace. I will rebase it on top of your patch.

I'm sorry for inconvenience.

>> ---
>>  doc/guides/rel_notes/deprecation.rst   | 7 +--
>>  doc/guides/rel_notes/release_21_11.rst | 3 +++
>>  lib/mbuf/rte_mbuf_core.h   | 7 ---
>>  3 files changed, 4 insertions(+), 13 deletions(-)
>>
>> diff --git a/doc/guides/rel_notes/deprecation.rst 
>> b/doc/guides/rel_notes/deprecation.rst
>> index 59445a6f42..3175426dad 100644
>> --- a/doc/guides/rel_notes/deprecation.rst
>> +++ b/doc/guides/rel_notes/deprecation.rst
>> @@ -42,7 +42,7 @@ Deprecation Notices
>>  * mbuf: The mbuf offload flags ``PKT_*`` will be renamed as 
>> ``RTE_MBUF_F_*``.
>>A compatibility layer will be kept until DPDK 22.11, except for the flags
>>that are already deprecated (``PKT_RX_L4_CKSUM_BAD``, 
>> ``PKT_RX_IP_CKSUM_BAD``,
>> -  ``PKT_RX_EIP_CKSUM_BAD``, ``PKT_TX_QINQ_PKT``) which will be removed.
>> +  ``PKT_RX_OUTER_IP_CKSUM_BAD``, ``PKT_TX_QINQ_PKT``) which will be removed.
>>  
> 
> I think this one should not be modified.
> 
> You can add my ack to the v2.

Fixed in v2.

Thanks,
Andrew.

Re: [dpdk-dev] [PATCH v2 2/5] cryptodev: allocate max space for internal qp array

[Zhang, Roy Fan]

> -Original Message-
> From: Akhil Goyal 
> Sent: Monday, October 11, 2021 1:43 PM
> To: dev@dpdk.org
> Cc: tho...@monjalon.net; david.march...@redhat.com;
> hemant.agra...@nxp.com; ano...@marvell.com; De Lara Guarch, Pablo
> ; Trahe, Fiona ;
> Doherty, Declan ; ma...@nvidia.com;
> g.si...@nxp.com; Zhang, Roy Fan ;
> jianjay.z...@huawei.com; asoma...@amd.com; ruifeng.w...@arm.com;
> Ananyev, Konstantin ; Nicolau, Radu
> ; ajit.khapa...@broadcom.com;
> rnagadhee...@marvell.com; adwiv...@marvell.com; Power, Ciara
> ; Akhil Goyal 
> Subject: [PATCH v2 2/5] cryptodev: allocate max space for internal qp array
> 
> At queue_pair config stage, allocate memory for maximum
> number of queue pair pointers that a device can support.
> 
> This will allow fast path APIs(enqueue_burst/dequeue_burst) to
> refer pointer to internal QP data without checking for currently
> configured QPs.
> This is required to hide the rte_cryptodev and rte_cryptodev_data
> structure from user.
> 
> Signed-off-by: Akhil Goyal 
> ---
Acked-by: Fan Zhang 

Re: [dpdk-dev] [PATCH v2 4/5] cryptodev: update fast path APIs to use new flat array

[Zhang, Roy Fan]

Hi Akhil,

> -Original Message-
> From: Akhil Goyal 
> Sent: Monday, October 11, 2021 1:43 PM
> To: dev@dpdk.org
> Cc: tho...@monjalon.net; david.march...@redhat.com;
> hemant.agra...@nxp.com; ano...@marvell.com; De Lara Guarch, Pablo
> ; Trahe, Fiona ;
> Doherty, Declan ; ma...@nvidia.com;
> g.si...@nxp.com; Zhang, Roy Fan ;
> jianjay.z...@huawei.com; asoma...@amd.com; ruifeng.w...@arm.com;
> Ananyev, Konstantin ; Nicolau, Radu
> ; ajit.khapa...@broadcom.com;
> rnagadhee...@marvell.com; adwiv...@marvell.com; Power, Ciara
> ; Akhil Goyal 
> Subject: [PATCH v2 4/5] cryptodev: update fast path APIs to use new flat
> array
> 
> Rework fast-path cryptodev functions to use rte_crypto_fp_ops[].
> While it is an API/ABI breakage, this change is intended to be
> transparent for both users (no changes in user app is required) and
> PMD developers (no changes in PMD is required).
> 
> Signed-off-by: Akhil Goyal 
> ---
>  lib/cryptodev/rte_cryptodev.h | 27 +--
>  1 file changed, 17 insertions(+), 10 deletions(-)
> 
> diff --git a/lib/cryptodev/rte_cryptodev.h b/lib/cryptodev/rte_cryptodev.h
> index ce0dca72be..739ad529e5 100644
> --- a/lib/cryptodev/rte_cryptodev.h
> +++ b/lib/cryptodev/rte_cryptodev.h
> @@ -1832,13 +1832,18 @@ static inline uint16_t
>  rte_cryptodev_dequeue_burst(uint8_t dev_id, uint16_t qp_id,
>   struct rte_crypto_op **ops, uint16_t nb_ops)
>  {
> - struct rte_cryptodev *dev = &rte_cryptodevs[dev_id];
> + struct rte_crypto_fp_ops *fp_ops;

We may need to use const for fp_ops since we only call the function pointers in 
it.
 
> + void *qp;
> 
>   rte_cryptodev_trace_dequeue_burst(dev_id, qp_id, (void **)ops,
> nb_ops);
> - nb_ops = (*dev->dequeue_burst)
> - (dev->data->queue_pairs[qp_id], ops, nb_ops);
> +
> + fp_ops = &rte_crypto_fp_ops[dev_id];
> + qp = fp_ops->qp.data[qp_id];
> +
> + nb_ops = fp_ops->dequeue_burst(qp, ops, nb_ops);
> +
>  #ifdef RTE_CRYPTO_CALLBACKS
> - if (unlikely(dev->deq_cbs != NULL)) {
> + if (unlikely(fp_ops->qp.deq_cb != NULL)) {
>   struct rte_cryptodev_cb_rcu *list;
>   struct rte_cryptodev_cb *cb;
> 
> @@ -1848,7 +1853,7 @@ rte_cryptodev_dequeue_burst(uint8_t dev_id,
> uint16_t qp_id,
>* cb and cb->fn/cb->next, __ATOMIC_ACQUIRE memory
> order is
>* not required.
>*/
> - list = &dev->deq_cbs[qp_id];
> + list = (struct rte_cryptodev_cb_rcu *)&fp_ops-
> >qp.deq_cb[qp_id];
>   rte_rcu_qsbr_thread_online(list->qsbr, 0);
>   cb = __atomic_load_n(&list->next, __ATOMIC_RELAXED);
> 
> @@ -1899,10 +1904,13 @@ static inline uint16_t
>  rte_cryptodev_enqueue_burst(uint8_t dev_id, uint16_t qp_id,
>   struct rte_crypto_op **ops, uint16_t nb_ops)
>  {
> - struct rte_cryptodev *dev = &rte_cryptodevs[dev_id];
> + struct rte_crypto_fp_ops *fp_ops;

Same as above

> + void *qp;
> 
> + fp_ops = &rte_crypto_fp_ops[dev_id];
> + qp = fp_ops->qp.data[qp_id];
>  #ifdef RTE_CRYPTO_CALLBACKS
> - if (unlikely(dev->enq_cbs != NULL)) {
> + if (unlikely(fp_ops->qp.enq_cb != NULL)) {
>   struct rte_cryptodev_cb_rcu *list;
>   struct rte_cryptodev_cb *cb;
> 
> @@ -1912,7 +1920,7 @@ rte_cryptodev_enqueue_burst(uint8_t dev_id,
> uint16_t qp_id,
>* cb and cb->fn/cb->next, __ATOMIC_ACQUIRE memory
> order is
>* not required.
>*/
> - list = &dev->enq_cbs[qp_id];
> + list = (struct rte_cryptodev_cb_rcu *)&fp_ops-
> >qp.enq_cb[qp_id];
>   rte_rcu_qsbr_thread_online(list->qsbr, 0);
>   cb = __atomic_load_n(&list->next, __ATOMIC_RELAXED);
> 
> @@ -1927,8 +1935,7 @@ rte_cryptodev_enqueue_burst(uint8_t dev_id,
> uint16_t qp_id,
>  #endif
> 
>   rte_cryptodev_trace_enqueue_burst(dev_id, qp_id, (void **)ops,
> nb_ops);
> - return (*dev->enqueue_burst)(
> - dev->data->queue_pairs[qp_id], ops, nb_ops);
> + return fp_ops->enqueue_burst(qp, ops, nb_ops);
>  }
> 
> 
> --
> 2.25.1

Other than the minor comments above
Acked-by: Fan Zhang 

[dpdk-dev] [PATCH] test/service: fix race in attr check

[David Marchand]

The CI reported rare (and cryptic) failures like:

RTE>>service_autotest
 + --- +
 + Test Suite : service core test suite
 + --- +
 + TestCase [ 0] : unregister_all succeeded
 + TestCase [ 1] : service_name succeeded
 + TestCase [ 2] : service_get_by_name succeeded
Service dummy_service Summary
  dummy_service: stats 1calls 0 cycles 0avg: 0
Service dummy_service Summary
  dummy_service: stats 0calls 0 cycles 0avg: 0
 + TestCase [ 3] : service_dump succeeded
 + TestCase [ 4] : service_attr_get failed
 + TestCase [ 5] : service_lcore_attr_get succeeded
 + TestCase [ 6] : service_probe_capability succeeded
 + TestCase [ 7] : service_start_stop succeeded
 + TestCase [ 8] : service_lcore_add_del succeeded
 + TestCase [ 9] : service_lcore_start_stop succeeded
 + TestCase [10] : service_lcore_en_dis_able succeeded
 + TestCase [11] : service_mt_unsafe_poll succeeded
 + TestCase [12] : service_mt_safe_poll succeeded
perf test for MT Safe: 42.7 cycles per call
 + TestCase [13] : service_app_lcore_mt_safe succeeded
perf test for MT Unsafe: 73.3 cycles per call
 + TestCase [14] : service_app_lcore_mt_unsafe succeeded
 + TestCase [15] : service_may_be_active succeeded
 + TestCase [16] : service_active_two_cores succeeded
 + --- +
 + Test Suite Summary : service core test suite
 + --- +
 + Tests Total :   17
 + Tests Skipped :  0
 + Tests Executed :17
 + Tests Unsupported:   0
 + Tests Passed :  16
 + Tests Failed :   1
 + --- +
Test Failed
RTE>>
stderr:
EAL: Detected CPU lcores: 16
EAL: Detected NUMA nodes: 2
EAL: Detected static linkage of DPDK
EAL: Multi-process socket /var/run/dpdk/service_autotest/mp_socket
EAL: Selected IOVA mode 'PA'
EAL: No available 1048576 kB hugepages reported
EAL: VFIO support initialized
EAL: Device :03:00.0 is not NUMA-aware, defaulting socket to 0
APP: HPET is not enabled, using TSC as default timer
EAL: Test assert service_attr_get line 340 failed: attr_get() call didn't
 get call count (zero)

According to API, trying to stop a service lcore is not possible if this
lcore is the only one associated to a service.
Doing this will result in a -EBUSY return code from
rte_service_lcore_stop() which the service_attr_get subtest was not
checking.
This left the service lcore running, and a race existed with the main
lcore on checking the service attributes which triggered this CI
failure.

To fix this, dissociate the service lcore with current service.

Once fixed this first issue, a race still exists, because the
wait_slcore_inactive helper added in a previous fix was not
paired with a check that the service lcore _did_ stop.

Add missing check on rte_service_lcore_may_be_active.

Fixes: 4d55194d76a4 ("service: add attribute get function")
Fixes: 52bb6be259ff ("test/service: fix race condition on stopping lcore")
Cc: sta...@dpdk.org

Signed-off-by: David Marchand 
---
 app/test/test_service_cores.c | 8 +++-
 1 file changed, 7 insertions(+), 1 deletion(-)

diff --git a/app/test/test_service_cores.c b/app/test/test_service_cores.c
index ece104054e..6821dca0e6 100644
--- a/app/test/test_service_cores.c
+++ b/app/test/test_service_cores.c
@@ -318,10 +318,16 @@ service_attr_get(void)
TEST_ASSERT_EQUAL(1, cycles_gt_zero,
"attr_get() failed to get cycles (expected > zero)");
 
-   rte_service_lcore_stop(slcore_id);
+   TEST_ASSERT_EQUAL(0, rte_service_map_lcore_set(id, slcore_id, 0),
+   "Disabling valid service and core failed");
+   TEST_ASSERT_EQUAL(0, rte_service_lcore_stop(slcore_id),
+   "Failed to stop service lcore");
 
wait_slcore_inactive(slcore_id);
 
+   TEST_ASSERT_EQUAL(0, rte_service_lcore_may_be_active(slcore_id),
+ "Service lcore not stopped after waiting.");
+
TEST_ASSERT_EQUAL(0, rte_service_attr_get(id, attr_calls, &attr_value),
"Valid attr_get() call didn't return success");
TEST_ASSERT_EQUAL(1, (attr_value > 0),
-- 
2.23.0

Re: [dpdk-dev] [PATCH] test/service: fix race in attr check

[Aaron Conole]

David Marchand  writes:

> The CI reported rare (and cryptic) failures like:
>
> RTE>>service_autotest
>  + --- +
>  + Test Suite : service core test suite
>  + --- +
>  + TestCase [ 0] : unregister_all succeeded
>  + TestCase [ 1] : service_name succeeded
>  + TestCase [ 2] : service_get_by_name succeeded
> Service dummy_service Summary
>   dummy_service: stats 1  calls 0 cycles 0avg: 0
> Service dummy_service Summary
>   dummy_service: stats 0  calls 0 cycles 0avg: 0
>  + TestCase [ 3] : service_dump succeeded
>  + TestCase [ 4] : service_attr_get failed
>  + TestCase [ 5] : service_lcore_attr_get succeeded
>  + TestCase [ 6] : service_probe_capability succeeded
>  + TestCase [ 7] : service_start_stop succeeded
>  + TestCase [ 8] : service_lcore_add_del succeeded
>  + TestCase [ 9] : service_lcore_start_stop succeeded
>  + TestCase [10] : service_lcore_en_dis_able succeeded
>  + TestCase [11] : service_mt_unsafe_poll succeeded
>  + TestCase [12] : service_mt_safe_poll succeeded
> perf test for MT Safe: 42.7 cycles per call
>  + TestCase [13] : service_app_lcore_mt_safe succeeded
> perf test for MT Unsafe: 73.3 cycles per call
>  + TestCase [14] : service_app_lcore_mt_unsafe succeeded
>  + TestCase [15] : service_may_be_active succeeded
>  + TestCase [16] : service_active_two_cores succeeded
>  + --- +
>  + Test Suite Summary : service core test suite
>  + --- +
>  + Tests Total :   17
>  + Tests Skipped :  0
>  + Tests Executed :17
>  + Tests Unsupported:   0
>  + Tests Passed :  16
>  + Tests Failed :   1
>  + --- +
> Test Failed
> RTE>>
> stderr:
> EAL: Detected CPU lcores: 16
> EAL: Detected NUMA nodes: 2
> EAL: Detected static linkage of DPDK
> EAL: Multi-process socket /var/run/dpdk/service_autotest/mp_socket
> EAL: Selected IOVA mode 'PA'
> EAL: No available 1048576 kB hugepages reported
> EAL: VFIO support initialized
> EAL: Device :03:00.0 is not NUMA-aware, defaulting socket to 0
> APP: HPET is not enabled, using TSC as default timer
> EAL: Test assert service_attr_get line 340 failed: attr_get() call didn't
>  get call count (zero)
>
> According to API, trying to stop a service lcore is not possible if this
> lcore is the only one associated to a service.
> Doing this will result in a -EBUSY return code from
> rte_service_lcore_stop() which the service_attr_get subtest was not
> checking.
> This left the service lcore running, and a race existed with the main
> lcore on checking the service attributes which triggered this CI
> failure.
>
> To fix this, dissociate the service lcore with current service.
>
> Once fixed this first issue, a race still exists, because the
> wait_slcore_inactive helper added in a previous fix was not
> paired with a check that the service lcore _did_ stop.
>
> Add missing check on rte_service_lcore_may_be_active.
>
> Fixes: 4d55194d76a4 ("service: add attribute get function")
> Fixes: 52bb6be259ff ("test/service: fix race condition on stopping lcore")
> Cc: sta...@dpdk.org
>
> Signed-off-by: David Marchand 
> ---

Excellent catch.

Acked-by: Aaron Conole 

Re: [dpdk-dev] [PATCH v2 5/5] cryptodev: move device specific structures

[Zhang, Roy Fan]

Hi Akhil,

> -Original Message-
> From: Akhil Goyal 
> Sent: Monday, October 11, 2021 1:43 PM
> To: dev@dpdk.org
> Cc: tho...@monjalon.net; david.march...@redhat.com;
> hemant.agra...@nxp.com; ano...@marvell.com; De Lara Guarch, Pablo
> ; Trahe, Fiona ;
> Doherty, Declan ; ma...@nvidia.com;
> g.si...@nxp.com; Zhang, Roy Fan ;
> jianjay.z...@huawei.com; asoma...@amd.com; ruifeng.w...@arm.com;
> Ananyev, Konstantin ; Nicolau, Radu
> ; ajit.khapa...@broadcom.com;
> rnagadhee...@marvell.com; adwiv...@marvell.com; Power, Ciara
> ; Akhil Goyal 
> Subject: [PATCH v2 5/5] cryptodev: move device specific structures
> 
> The device specific structures - rte_cryptodev
> and rte_cryptodev_data are moved to cryptodev_pmd.h
> to hide it from the applications.
> 
> Signed-off-by: Akhil Goyal 
> ---
...
> diff --git a/lib/cryptodev/cryptodev_pmd.h
> b/lib/cryptodev/cryptodev_pmd.h
> index a71edbb991..0d1bef0e67 100644
> --- a/lib/cryptodev/cryptodev_pmd.h
> +++ b/lib/cryptodev/cryptodev_pmd.h
> @@ -52,6 +52,71 @@ struct rte_cryptodev_pmd_init_params {
>   unsigned int max_nb_queue_pairs;
>  };
> 
> +/**
> + * @internal
> + * The data part, with no function pointers, associated with each device.
> + *
> + * This structure is safe to place in shared memory to be common among
> + * different processes in a multi-process configuration.
> + */
> +struct rte_cryptodev_data {
> + /** Device ID for this instance */
> + uint8_t dev_id;
> + /** Socket ID where memory is allocated */
> + uint8_t socket_id;
> + /** Unique identifier name */
> + char name[RTE_CRYPTODEV_NAME_MAX_LEN];
> +
> + __extension__
> + /** Device state: STARTED(1)/STOPPED(0) */
> + uint8_t dev_started : 1;
> +
> + /** Session memory pool */
> + struct rte_mempool *session_pool;

Looks like we never used this pool ever - shall we take this chance to remove 
it?

> + /** Array of pointers to queue pairs. */
> + void **queue_pairs;
> + /** Number of device queue pairs. */
> + uint16_t nb_queue_pairs;

queue_pairs are likely the only item we use rte_cryptodev_data in the fast-path.
Also take this chance to move it to the beginning of the structure?

> +
> + /** PMD-specific private data */
> + void *dev_private;
> +} __rte_cache_aligned;
> +

Not really comments but some suggestions - could be done in the future as the
structure is moved to internal in this patchset.
Acked-by: Fan Zhang 

Re: [dpdk-dev] [PATCH v5 4/7] ethdev: copy fast-path API into separate structure

[Ananyev, Konstantin]

> > Copy public function pointers (rx_pkt_burst(), etc.) and related
> > pointers to internal data from rte_eth_dev structure into a
> > separate flat array. That array will remain in a public header.
> > The intention here is to make rte_eth_dev and related structures internal.
> > That should allow future possible changes to core eth_dev structures
> > to be transparent to the user and help to avoid ABI/API breakages.
> > The plan is to keep minimal part of data from rte_eth_dev public,
> > so we still can use inline functions for fast-path calls
> > (like rte_eth_rx_burst(), etc.) to avoid/minimize slowdown.
> > The whole idea beyond this new schema:
> > 1. PMDs keep to setup fast-path function pointers and related data
> >inside rte_eth_dev struct in the same way they did it before.
> > 2. Inside rte_eth_dev_start() and inside rte_eth_dev_probing_finish()
> >(for secondary process) we call eth_dev_fp_ops_setup, which
> >copies these function and data pointers into rte_eth_fp_ops[port_id].
> > 3. Inside rte_eth_dev_stop() and inside rte_eth_dev_release_port()
> >we call eth_dev_fp_ops_reset(), which resets rte_eth_fp_ops[port_id]
> >into some dummy values.
> > 4. fast-path ethdev API (rte_eth_rx_burst(), etc.) will use that new
> >flat array to call PMD specific functions.
> > That approach should allow us to make rte_eth_devices[] private
> > without introducing regression and help to avoid changes in drivers code.
> >
> > Signed-off-by: Konstantin Ananyev 
> > ---
> >  lib/ethdev/ethdev_private.c  | 52 ++
> >  lib/ethdev/ethdev_private.h  |  7 +
> >  lib/ethdev/rte_ethdev.c  | 27 ++
> >  lib/ethdev/rte_ethdev_core.h | 55 
> >  4 files changed, 141 insertions(+)
> >
> > diff --git a/lib/ethdev/ethdev_private.c b/lib/ethdev/ethdev_private.c
> > index 012cf73ca2..3eeda6e9f9 100644
> > --- a/lib/ethdev/ethdev_private.c
> > +++ b/lib/ethdev/ethdev_private.c
> > @@ -174,3 +174,55 @@ rte_eth_devargs_parse_representor_ports(char *str, 
> > void *data)
> > RTE_LOG(ERR, EAL, "wrong representor format: %s\n", str);
> > return str == NULL ? -1 : 0;
> >  }
> > +
> > +static uint16_t
> > +dummy_eth_rx_burst(__rte_unused void *rxq,
> > +   __rte_unused struct rte_mbuf **rx_pkts,
> > +   __rte_unused uint16_t nb_pkts)
> > +{
> > +   RTE_ETHDEV_LOG(ERR, "rx_pkt_burst for unconfigured port\n");
> > +   rte_errno = ENOTSUP;
> > +   return 0;
> > +}
> > +
> > +static uint16_t
> > +dummy_eth_tx_burst(__rte_unused void *txq,
> > +   __rte_unused struct rte_mbuf **tx_pkts,
> > +   __rte_unused uint16_t nb_pkts)
> > +{
> > +   RTE_ETHDEV_LOG(ERR, "tx_pkt_burst for unconfigured port\n");
> > +   rte_errno = ENOTSUP;
> > +   return 0;
> > +}
> > +
> > +void
> > +eth_dev_fp_ops_reset(struct rte_eth_fp_ops *fpo)
> 
> The port_id parameter is preferable, this will hide rte_eth_fp_ops as much as 
> possible.

Why do we need to hide it here?
rte_eth_fp_ops is a public structure, and it is a helper function that
just resets fields of this structure to some predefined dummy values.
Nice and simple, so I prefer to keep it like that. 

> 
> > +{
> > +   static void *dummy_data[RTE_MAX_QUEUES_PER_PORT];
> > +   static const struct rte_eth_fp_ops dummy_ops = {
> > +   .rx_pkt_burst = dummy_eth_rx_burst,
> > +   .tx_pkt_burst = dummy_eth_tx_burst,
> > +   .rxq = {.data = dummy_data, .clbk = dummy_data,},
> > +   .txq = {.data = dummy_data, .clbk = dummy_data,},
> > +   };
> > +
> > +   *fpo = dummy_ops;
> > +}
> > +
> > +void
> > +eth_dev_fp_ops_setup(struct rte_eth_fp_ops *fpo,
> > +   const struct rte_eth_dev *dev)
> 
> Because fp_ops and eth_dev is a one-to-one correspondence. It's better only 
> use
> port_id parameter.

Same as above:
All this internal helper function does - copies some fields from one structure 
to another.
Both structures are visible by ethdev layer.
No point to add extra assumptions and complexity here. 

> 
> > +{
> > +   fpo->rx_pkt_burst = dev->rx_pkt_burst;
> > +   fpo->tx_pkt_burst = dev->tx_pkt_burst;
> > +   fpo->tx_pkt_prepare = dev->tx_pkt_prepare;
> > +   fpo->rx_queue_count = dev->rx_queue_count;
> > +   fpo->rx_descriptor_status = dev->rx_descriptor_status;
> > +   fpo->tx_descriptor_status = dev->tx_descriptor_status;
> > +
> > +   fpo->rxq.data = dev->data->rx_queues;
> > +   fpo->rxq.clbk = (void **)(uintptr_t)dev->post_rx_burst_cbs;
> > +
> > +   fpo->txq.data = dev->data->tx_queues;
> > +   fpo->txq.clbk = (void **)(uintptr_t)dev->pre_tx_burst_cbs;
> > +}
> > diff --git a/lib/ethdev/ethdev_private.h b/lib/ethdev/ethdev_private.h
> > index 3724429577..5721be7bdc 100644
> > --- a/lib/ethdev/ethdev_private.h
> > +++ b/lib/ethdev/ethdev_private.h
> > @@ -26,4 +26,11 @@ eth_find_device(const struct rte_eth_dev *_start, 
> > rte_eth_cmp_t cmp,
> >  /* Parse devargs value for representor parameter. */
>

Re: [dpdk-dev] [RFC PATCH] ethdev: mtr: enhance input color table features

[Dumitrescu, Cristian]

Hi Jerin,

> -Original Message-
> From: jer...@marvell.com 
> Sent: Friday, August 20, 2021 9:24 AM
> To: Dumitrescu, Cristian ; Thomas Monjalon
> ; Yigit, Ferruh ; Andrew
> Rybchenko 
> Cc: dev@dpdk.org; arybche...@solarflare.com; l...@nvidia.com;
> ajit.khapa...@broadcom.com; Singh, Jasvinder
> ; ma...@nvidia.com;
> ndabilpu...@marvell.com; sk...@marvell.com; rkuduruma...@marvell.com;
> Jerin Jacob 
> Subject: [dpdk-dev] [RFC PATCH] ethdev: mtr: enhance input color table
> features
> 
> From: Jerin Jacob 
> 
> Currently, meter object supports only DSCP based on input color table,
> The patch enhance that to support VLAN based input color table,
> color table based on inner field for the tunnel use case,  and support
> for fallback color per meter if packet based on a different field.
> 
> All of the above features are exposed through capability and added
> additional capability to specify the implementation supports
> more than one input color table per ethdev port.
> 
> Signed-off-by: Jerin Jacob 
> ---
>  lib/ethdev/rte_mtr.h | 130
> ++-
>  1 file changed, 116 insertions(+), 14 deletions(-)
> 
> diff --git a/lib/ethdev/rte_mtr.h b/lib/ethdev/rte_mtr.h
> index dc246dd7af..311e8754de 100644
> --- a/lib/ethdev/rte_mtr.h
> +++ b/lib/ethdev/rte_mtr.h
> @@ -213,6 +213,16 @@ struct rte_mtr_meter_policy_params {
>   const struct rte_flow_action *actions[RTE_COLORS];
>  };
> 
> +/**
> + * Input color table
> + */
> +enum rte_mtr_input_color_tbl {
> + /** DSCP based input color table */
> + RTE_MTR_INPUT_COLOR_TBL_DSCP,
> + /** VLAN based input color table */
> + RTE_MTR_INPUT_COLOR_TBL_VLAN,
> +};
> +
>  /**
>   * Parameters for each traffic metering & policing object
>   *
> @@ -233,20 +243,44 @@ struct rte_mtr_params {
>*/
>   int use_prev_mtr_color;
> 
> - /** Meter input color. When non-NULL: it points to a pre-allocated
> and
> -  * pre-populated table with exactly 64 elements providing the input
> -  * color for each value of the IPv4/IPv6 Differentiated Services Code
> -  * Point (DSCP) input packet field. When NULL: it is equivalent to
> -  * setting this parameter to an all-green populated table (i.e. table
> -  * with all the 64 elements set to green color). The color blind mode
> -  * is configured by setting *use_prev_mtr_color* to 0 and
> *dscp_table*
> -  * to either NULL or to an all-green populated table. When
> -  * *use_prev_mtr_color* is non-zero value or when *dscp_table*
> contains
> -  * at least one yellow or red color element, then the color aware
> mode
> -  * is configured.
> -  */
> - enum rte_color *dscp_table;
> -
> + RTE_STD_C11
> + union {
> + /** Meter input color based on DSCP.
> +  * Valid when rte_mtr_input_color_tbl::tbl_selector is
> +  * set to RTE_MTR_INPUT_COLOR_TBL_DSCP.
> +  * When non-NULL: it points to a pre-allocated and pre-
> populated
> +  * table with exactly 64 elements providing the input color for
> +  * each value of the IPv4/IPv6 Differentiated Services Code
> +  * Point (DSCP) input packet field. When NULL:
> +  * it is equivalent to setting this parameter to an all-green
> +  * populated table (i.e. table with
> +  * all the 64 elements set to green color). The color blind
> mode
> +  * is configured by setting *use_prev_mtr_color* to 0 and
> +  * *dscp_table* to either NULL or to an all-green
> +  * populated table. When *use_prev_mtr_color* is non-zero
> value
> +  * or when *dscp_table* contains at least one yellow or
> +  * red color element, then the color aware mode is
> configured.
> +  * @see struct
> rte_mtr_capabilities::input_color_dscp_supported
> +  */
> + enum rte_color *dscp_table;
> + /** Meter input color based on VLAN.
> +  * Valid when rte_mtr_input_color_tbl::tbl_selector is
> +  * set to RTE_MTR_INPUT_COLOR_TBL_VLAN.
> +  * When non-NULL: it points to a pre-allocated and pre-
> populated
> +  * table with exactly 16 elements providing the input color for
> +  * each value of the DEI(1bit), PCP(3 bits) input packet field.
> +  * When NULL: it is equivalent to setting this parameter to an
> +  * all-green populated table (i.e. table with
> +  * all the 16 elements set to green color). The color blind
> mode
> +  * is configured by setting *use_prev_mtr_color* to 0 and
> +  * *vlan_table* to either NULL or to an all-green
> +  * populated table. When *use_prev_mtr_color* is non-zero
> value
> +  * or when *vlan_table* contains at least one yellow or
> +  * red color element, then the color aware mode is
> configured.
> +

Re: [dpdk-dev] [PATCH v5 4/7] ethdev: copy fast-path API into separate structure

[Ananyev, Konstantin]

> 
> Sorry to self-reply.
> 
> I think it's better the 'struct rte_eth_dev *dev' hold a pointer to the
> 'struct rte_eth_fp_ops', e.g.
> 
>   struct rte_eth_dev {
>   struct rte_eth_fp_ops *fp_ops;
>   ...  // other field
>   }
> 
> The eth framework set the pointer in the rte_eth_dev_pci_allocate(), and 
> driver fill
> corresponding callback:
>   dev->fp_ops->rx_pkt_burst = xxx_recv_pkts;
>   dev->fp_ops->tx_pkt_burst = xxx_xmit_pkts;
>   ...
> 
> In this way, the behavior of the primary and secondary processes can be 
> unified, which
> is basically the same as that of the original process.

I don't think it is a good thing to do, as it nullifies one of the main thing 
of this approach:
fp_ops[] points to actual rx/tx functions and data only when PMD is really 
ready to do rx/tx
(device is properly configured and started), in other cases it points to dummy 
stubs.
As another drawback it will mean code changes in aach and every driver.

Re: [dpdk-dev] [PATCH v2] ethdev: remove deprecated shared counter attribute

[Ori Kam]

Hi Andrew,

> -Original Message-
> From: Andrew Rybchenko 
> Subject: Re: [dpdk-dev] [PATCH v2] ethdev: remove deprecated shared counter 
> attribute
> 
> Hi Ori,
> 
> On 10/11/21 1:02 PM, Ori Kam wrote:
> > Hi Andrew,
> >
> > Sorry but I think I missed something.
> >
> >> -Original Message-
> >> From: dev  On Behalf Of Andrew Rybchenko
> >> Sent: Friday, October 8, 2021 1:26 PM
> >>
> >> Indirect actions should be used to do shared counters.
> >>
> >> Signed-off-by: Andrew Rybchenko 
> >> Acked-by: Thomas Monjalon 
> >> Acked-by: Ajit Khaparde 
> >> Acked-by: Somnath Kotur 
> >> Acked-by: Ori Kam 
> >> ---
> >> v2:
> >> - remove reserved field from count structure (review from Stephen)
> >> - apply mlx5 review notes from Matan
> >>
> >>  app/test-pmd/cmdline_flow.c|  10 --
> >>  doc/guides/prog_guide/rte_flow.rst |  19 +---
> >>  doc/guides/rel_notes/deprecation.rst   |   4 -
> >>  doc/guides/rel_notes/release_21_11.rst |   4 +
> >>  drivers/net/bnxt/tf_ulp/ulp_rte_parser.c   |   5 -
> >>  drivers/net/cnxk/cnxk_rte_flow.c   |   8 --
> >>  drivers/net/hns3/hns3_flow.c   |   3 +-
> >>  drivers/net/ice/ice_fdir_filter.c  |   4 +-
> >>  drivers/net/mlx5/mlx5.c|  11 --
> >>  drivers/net/mlx5/mlx5.h|   9 --
> >>  drivers/net/mlx5/mlx5_flow_dv.c| 118 ++---
> >>  drivers/net/mlx5/mlx5_flow_verbs.c |  22 +---
> >>  drivers/net/octeontx2/otx2_flow_parse.c|  10 --
> >>  drivers/net/sfc/sfc_mae.c  |   9 +-
> >>  drivers/net/softnic/rte_eth_softnic_flow.c |   7 --
> >>  lib/ethdev/rte_flow.h  |  16 +--
> >>  16 files changed, 22 insertions(+), 237 deletions(-)
> >>
> >
> > [Snip]
> >
> >> diff --git a/lib/ethdev/rte_flow.h b/lib/ethdev/rte_flow.h index
> >> 7b1ed7f110..9819c25d2f 100644
> >> --- a/lib/ethdev/rte_flow.h
> >> +++ b/lib/ethdev/rte_flow.h
> >> @@ -75,7 +75,7 @@ extern "C" {
> >>   * At least one direction must be specified.
> >>   *
> >>   * Specifying both directions at once for a given rule is not
> >> recommended
> >> - * but may be valid in a few cases (e.g. shared counter).
> >> + * but may be valid in a few cases.
> >>   */
> >>  struct rte_flow_attr {
> >>uint32_t group; /**< Priority group. */ @@ -2498,24 +2498,10 @@
> >> struct rte_flow_query_age {
> >>   * Counters can be retrieved and reset through ``rte_flow_query()``, see
> >>   * ``struct rte_flow_query_count``.
> >>   *
> >> - * @deprecated Shared attribute is deprecated, use generic
> >> - * RTE_FLOW_ACTION_TYPE_INDIRECT action.
> >> - *
> >> - * The shared flag indicates whether the counter is unique to the
> >> flow rule the
> >> - * action is specified with, or whether it is a shared counter.
> >> - *
> >> - * For a count action with the shared flag set, then then a global
> >> device
> >> - * namespace is assumed for the counter id, so that any matched flow
> >> rules using
> >> - * a count action with the same counter id on the same port will
> >> contribute to
> >> - * that counter.
> >> - *
> >>   * For ports within the same switch domain then the counter id namespace 
> >> extends
> >>   * to all ports within that switch domain.
> >
> > I don't think we need this anymore.
> 
> I agree. I'll remove it in v3 if required, but I hope it could be removed on 
> apply as well.
> 
> >
> >>   */
> >>  struct rte_flow_action_count {
> >> -  /** @deprecated Share counter ID with other flow rules. */
> >> -  uint32_t shared:1;
> >> -  uint32_t reserved:31; /**< Reserved, must be zero. */
> >>uint32_t id; /**< Counter ID. */
> >
> > Why do we need to keep the id field?
> 
> It is a very good question. I thought about it and preserved it for the 
> corner case of two COUNT actions in
> the same rule.
> If so, id is required to distinguish on query.
> I don't know if we really need it to have two basically duplicate counters in 
> the same rule. However, since
> order of actions matter, COUNT, VXLAN_ENCAP, COUNT should produce different 
> byte counters.
> 
Good thinking, I will not block this patch. 
Just please fix the comment, if it can be done in apply that will be good for me

> I suggest to continue discussion and gather more thought on it, but do not 
> block the patch, since strictly
> speaking it is a bit separate topic as noted above.
> 
Yes lets take it in one of the RTE_FLOW meetings or over mail.

> Thanks,
> Andrew.

Best,
Ori

Re: [dpdk-dev] [PATCH v2] app/testpmd: retain all original dev conf when config DCB

[Ferruh Yigit]

On 10/11/2021 10:12 AM, Min Hu (Connor) wrote:

From: Huisong Li 

When configuring DCB, testpmd retains the rx_mode/tx_mode configuration in
rte_port->dev_conf. But some configurations, such as the link_speed, were
not saved if they were set before configuring DCB.

Fixes: 1a572499beb6 ("app/testpmd: setup DCB forwarding based on traffic class")
Cc: sta...@dpdk.org

Signed-off-by: Huisong Li 
Signed-off-by: Min Hu (Connor) 
Acked-by: Xiaoyun Li 


Applied to dpdk-next-net/main, thanks.

Re: [dpdk-dev] [v1] ethdev: add telemetry endpoint for device info

[Bruce Richardson]

On Mon, Oct 11, 2021 at 03:40:58PM +0100, Ferruh Yigit wrote:
> On 9/29/2021 5:25 AM, Gowrishankar Muthukrishnan wrote:
> > Add telemetry endpoint /ethdev/info for device info.
> > 
> > Signed-off-by: Gowrishankar Muthukrishnan 
> > Change-Id: I3e6ee2bd1a80675473adf0bd884b194f98e28536
> > ---
> >   lib/ethdev/rte_ethdev.c | 92 +
> >   1 file changed, 92 insertions(+)
> > 
> > diff --git a/lib/ethdev/rte_ethdev.c b/lib/ethdev/rte_ethdev.c
> > index daf5ca9242..9b7bfa5f63 100644
> > --- a/lib/ethdev/rte_ethdev.c
> > +++ b/lib/ethdev/rte_ethdev.c
> > @@ -6242,6 +6242,96 @@ eth_dev_handle_port_link_status(const char *cmd 
> > __rte_unused,
> > return 0;
> >   }
> > +static int
> > +eth_dev_handle_port_info(const char *cmd __rte_unused,
> > +   const char *params,
> > +   struct rte_tel_data *d)
> > +{
> > +   struct rte_tel_data *rxq_state, *txq_state;
> > +   char mac_addr[RTE_ETHER_ADDR_LEN];
> > +   struct rte_eth_dev *eth_dev;
> > +   char *end_param;
> > +   int port_id, i;
> > +
> > +   if (params == NULL || strlen(params) == 0 || !isdigit(*params))
> > +   return -1;
> > +
> > +   port_id = strtoul(params, &end_param, 0);
> > +   if (*end_param != '\0')
> > +   RTE_ETHDEV_LOG(NOTICE,
> > +   "Extra parameters passed to ethdev telemetry command, 
> > ignoring");
> > +
> > +   if (!rte_eth_dev_is_valid_port(port_id))
> > +   return -EINVAL;
> > +
> > +   eth_dev = &rte_eth_devices[port_id];
> > +   if (!eth_dev)
> > +   return -EINVAL;
> > +
> > +   rxq_state = rte_tel_data_alloc();
> > +   if (!rxq_state)
> > +   return -ENOMEM;
> > +
> > +   txq_state = rte_tel_data_alloc();
> > +   if (!txq_state)
> > +   return -ENOMEM;
> > +
> > +   rte_tel_data_start_dict(d);
> > +   rte_tel_data_add_dict_string(d, "name", eth_dev->data->name);
> > +   rte_tel_data_add_dict_int(d, "state", eth_dev->state);
> > +   rte_tel_data_add_dict_int(d, "nb_rx_queues",
> > +   eth_dev->data->nb_rx_queues);
> > +   rte_tel_data_add_dict_int(d, "nb_tx_queues",
> > +   eth_dev->data->nb_tx_queues);
> > +   rte_tel_data_add_dict_int(d, "port_id", eth_dev->data->port_id);
> > +   rte_tel_data_add_dict_int(d, "mtu", eth_dev->data->mtu);
> > +   rte_tel_data_add_dict_int(d, "rx_mbuf_size_min",
> > +   eth_dev->data->min_rx_buf_size);
> > +   rte_tel_data_add_dict_int(d, "rx_mbuf_alloc_fail",
> > +   eth_dev->data->rx_mbuf_alloc_failed);
> > +   snprintf(mac_addr, RTE_ETHER_ADDR_LEN, "%02x:%02x:%02x:%02x:%02x:%02x",
> > +eth_dev->data->mac_addrs->addr_bytes[0],
> > +eth_dev->data->mac_addrs->addr_bytes[1],
> > +eth_dev->data->mac_addrs->addr_bytes[2],
> > +eth_dev->data->mac_addrs->addr_bytes[3],
> > +eth_dev->data->mac_addrs->addr_bytes[4],
> > +eth_dev->data->mac_addrs->addr_bytes[5]);
> > +   rte_tel_data_add_dict_string(d, "mac_addr", mac_addr);
> > +   rte_tel_data_add_dict_int(d, "promiscuous",
> > +   eth_dev->data->promiscuous);
> > +   rte_tel_data_add_dict_int(d, "scattered_rx",
> > +   eth_dev->data->scattered_rx);
> > +   rte_tel_data_add_dict_int(d, "all_multicast",
> > +   eth_dev->data->all_multicast);
> > +   rte_tel_data_add_dict_int(d, "dev_started", eth_dev->data->dev_started);
> > +   rte_tel_data_add_dict_int(d, "lro", eth_dev->data->lro);
> > +   rte_tel_data_add_dict_int(d, "dev_configured",
> > +   eth_dev->data->dev_configured);
> > +
> > +   rte_tel_data_start_array(rxq_state, RTE_TEL_INT_VAL);
> > +   for (i = 0; i < eth_dev->data->nb_rx_queues; i++)
> > +   rte_tel_data_add_array_int(rxq_state,
> > +   eth_dev->data->rx_queue_state[i]);
> > +
> > +   rte_tel_data_start_array(txq_state, RTE_TEL_INT_VAL);
> > +   for (i = 0; i < eth_dev->data->nb_tx_queues; i++)
> > +   rte_tel_data_add_array_int(txq_state,
> > +   eth_dev->data->tx_queue_state[i]);
> > +
> > +   rte_tel_data_add_dict_container(d, "rxq_state", rxq_state, 0);
> > +   rte_tel_data_add_dict_container(d, "txq_state", txq_state, 0);
> > +   rte_tel_data_add_dict_int(d, "numa_node", eth_dev->data->numa_node);
> > +   rte_tel_data_add_dict_int(d, "dev_flags", eth_dev->data->dev_flags);
> > +   rte_tel_data_add_dict_int(d, "rx_offloads",
> > +   eth_dev->data->dev_conf.rxmode.offloads);
> > +   rte_tel_data_add_dict_int(d, "tx_offloads",
> > +   eth_dev->data->dev_conf.txmode.offloads);
> > +   rte_tel_data_add_dict_int(d, "ethdev_rss_hf",
> > +   eth_dev->data->dev_conf.rx_adv_conf.rss_conf.rss_hf);
> > +
> > +   return 0;
> > +}
> > +
> >   int
> >   rte_eth_hairpin_queue_peer_update(uint16_t peer_port, uint16_t peer_queue,
> >   struct rte_hairp

Re: [dpdk-dev] [PATCH v25 2/6] dmadev: add control plane API support

[Bruce Richardson]

On Mon, Oct 11, 2021 at 03:33:44PM +0800, Chengwen Feng wrote:
> This patch add control plane API for dmadev.
> 
> Signed-off-by: Chengwen Feng 
> Acked-by: Bruce Richardson 
> Acked-by: Morten Brørup 
> Reviewed-by: Kevin Laatz 
> Reviewed-by: Conor Walsh 
> ---
>  doc/guides/prog_guide/dmadev.rst   |  38 ++
>  doc/guides/rel_notes/release_21_11.rst |   1 +
>  lib/dmadev/rte_dmadev.c| 360 +++
>  lib/dmadev/rte_dmadev.h| 464 +
>  lib/dmadev/rte_dmadev_pmd.h|  61 
>  lib/dmadev/version.map |   9 +
>  6 files changed, 933 insertions(+)
> 



> +/**
> + * A structure used to retrieve the information of a DMA device.
> + *
> + * @see rte_dma_info_get
> + */
> +struct rte_dma_info {
> + /** Device capabilities (RTE_DMA_CAPA_*). */
> + uint64_t dev_capa;
> + /** Maximum number of virtual DMA channels supported. */
> + uint16_t max_vchans;
> + /** Maximum allowed number of virtual DMA channel descriptors. */
> + uint16_t max_desc;
> + /** Minimum allowed number of virtual DMA channel descriptors. */
> + uint16_t min_desc;
> + /** Maximum number of source or destination scatter-gather entry
> +  * supported.
> +  * If the device does not support COPY_SG capability, this value can be
> +  * zero.
> +  * If the device supports COPY_SG capability, then rte_dma_copy_sg()
> +  * parameter nb_src/nb_dst should not exceed this value.
> +  */
> + uint16_t max_sges;
> + /** NUMA node connection, -1 if unknown. */
> + int16_t numa_node;
> + /** Number of virtual DMA channel configured. */
> + uint16_t nb_vchans;
> +};
> +
Since we really don't want apps to have to access the rte_dma_devices
array, I think we should add "const char *name" to the info struct. Apps may
find it useful for debugging or logging, and certainly the unit tests will
use it.

/Bruce

Re: [dpdk-dev] [v1] ethdev: add telemetry endpoint for device info

[Ferruh Yigit]

On 10/11/2021 4:40 PM, Bruce Richardson wrote:

On Mon, Oct 11, 2021 at 03:40:58PM +0100, Ferruh Yigit wrote:

On 9/29/2021 5:25 AM, Gowrishankar Muthukrishnan wrote:

Add telemetry endpoint /ethdev/info for device info.

Signed-off-by: Gowrishankar Muthukrishnan 
Change-Id: I3e6ee2bd1a80675473adf0bd884b194f98e28536
---
   lib/ethdev/rte_ethdev.c | 92 +
   1 file changed, 92 insertions(+)

diff --git a/lib/ethdev/rte_ethdev.c b/lib/ethdev/rte_ethdev.c
index daf5ca9242..9b7bfa5f63 100644
--- a/lib/ethdev/rte_ethdev.c
+++ b/lib/ethdev/rte_ethdev.c
@@ -6242,6 +6242,96 @@ eth_dev_handle_port_link_status(const char *cmd 
__rte_unused,
return 0;
   }
+static int
+eth_dev_handle_port_info(const char *cmd __rte_unused,
+   const char *params,
+   struct rte_tel_data *d)
+{
+   struct rte_tel_data *rxq_state, *txq_state;
+   char mac_addr[RTE_ETHER_ADDR_LEN];
+   struct rte_eth_dev *eth_dev;
+   char *end_param;
+   int port_id, i;
+
+   if (params == NULL || strlen(params) == 0 || !isdigit(*params))
+   return -1;
+
+   port_id = strtoul(params, &end_param, 0);
+   if (*end_param != '\0')
+   RTE_ETHDEV_LOG(NOTICE,
+   "Extra parameters passed to ethdev telemetry command, 
ignoring");
+
+   if (!rte_eth_dev_is_valid_port(port_id))
+   return -EINVAL;
+
+   eth_dev = &rte_eth_devices[port_id];
+   if (!eth_dev)
+   return -EINVAL;
+
+   rxq_state = rte_tel_data_alloc();
+   if (!rxq_state)
+   return -ENOMEM;
+
+   txq_state = rte_tel_data_alloc();
+   if (!txq_state)
+   return -ENOMEM;
+
+   rte_tel_data_start_dict(d);
+   rte_tel_data_add_dict_string(d, "name", eth_dev->data->name);
+   rte_tel_data_add_dict_int(d, "state", eth_dev->state);
+   rte_tel_data_add_dict_int(d, "nb_rx_queues",
+   eth_dev->data->nb_rx_queues);
+   rte_tel_data_add_dict_int(d, "nb_tx_queues",
+   eth_dev->data->nb_tx_queues);
+   rte_tel_data_add_dict_int(d, "port_id", eth_dev->data->port_id);
+   rte_tel_data_add_dict_int(d, "mtu", eth_dev->data->mtu);
+   rte_tel_data_add_dict_int(d, "rx_mbuf_size_min",
+   eth_dev->data->min_rx_buf_size);
+   rte_tel_data_add_dict_int(d, "rx_mbuf_alloc_fail",
+   eth_dev->data->rx_mbuf_alloc_failed);
+   snprintf(mac_addr, RTE_ETHER_ADDR_LEN, "%02x:%02x:%02x:%02x:%02x:%02x",
+eth_dev->data->mac_addrs->addr_bytes[0],
+eth_dev->data->mac_addrs->addr_bytes[1],
+eth_dev->data->mac_addrs->addr_bytes[2],
+eth_dev->data->mac_addrs->addr_bytes[3],
+eth_dev->data->mac_addrs->addr_bytes[4],
+eth_dev->data->mac_addrs->addr_bytes[5]);
+   rte_tel_data_add_dict_string(d, "mac_addr", mac_addr);
+   rte_tel_data_add_dict_int(d, "promiscuous",
+   eth_dev->data->promiscuous);
+   rte_tel_data_add_dict_int(d, "scattered_rx",
+   eth_dev->data->scattered_rx);
+   rte_tel_data_add_dict_int(d, "all_multicast",
+   eth_dev->data->all_multicast);
+   rte_tel_data_add_dict_int(d, "dev_started", eth_dev->data->dev_started);
+   rte_tel_data_add_dict_int(d, "lro", eth_dev->data->lro);
+   rte_tel_data_add_dict_int(d, "dev_configured",
+   eth_dev->data->dev_configured);
+
+   rte_tel_data_start_array(rxq_state, RTE_TEL_INT_VAL);
+   for (i = 0; i < eth_dev->data->nb_rx_queues; i++)
+   rte_tel_data_add_array_int(rxq_state,
+   eth_dev->data->rx_queue_state[i]);
+
+   rte_tel_data_start_array(txq_state, RTE_TEL_INT_VAL);
+   for (i = 0; i < eth_dev->data->nb_tx_queues; i++)
+   rte_tel_data_add_array_int(txq_state,
+   eth_dev->data->tx_queue_state[i]);
+
+   rte_tel_data_add_dict_container(d, "rxq_state", rxq_state, 0);
+   rte_tel_data_add_dict_container(d, "txq_state", txq_state, 0);
+   rte_tel_data_add_dict_int(d, "numa_node", eth_dev->data->numa_node);
+   rte_tel_data_add_dict_int(d, "dev_flags", eth_dev->data->dev_flags);
+   rte_tel_data_add_dict_int(d, "rx_offloads",
+   eth_dev->data->dev_conf.rxmode.offloads);
+   rte_tel_data_add_dict_int(d, "tx_offloads",
+   eth_dev->data->dev_conf.txmode.offloads);
+   rte_tel_data_add_dict_int(d, "ethdev_rss_hf",
+   eth_dev->data->dev_conf.rx_adv_conf.rss_conf.rss_hf);
+
+   return 0;
+}
+
   int
   rte_eth_hairpin_queue_peer_update(uint16_t peer_port, uint16_t peer_queue,
  struct rte_hairpin_peer_info *cur_info,
@@ -6323,4 +6413,6 @@ RTE_INIT(ethdev_init_telemetry)

Re: [dpdk-dev] [PATCH] net/hns3: remove similar macro function definitions

[Ferruh Yigit]

On 10/9/2021 8:48 AM, Min Hu (Connor) wrote:

From: Chengchang Tang 

For different capabilities, we declare different macro functions to
determine whether the capabilities are supported.

This patch declare a unified macro function to judge capabilities.

Signed-off-by: Chengchang Tang 
Signed-off-by: Min Hu (Connor) 


Applied to dpdk-next-net/main, thanks.

Re: [dpdk-dev] [PATCH v3 01/12] ethdev: add port representor item to flow API

[Ori Kam]

Hi Ivan,

> -Original Message-
> From: Ivan Malov 
> Sent: Sunday, October 10, 2021 5:39 PM
> Subject: [PATCH v3 01/12] ethdev: add port representor item to flow API
> 
> For use in "transfer" flows. Supposed to match traffic entering the embedded 
> switch from the given
> ethdev.
> 
> Must not be combined with direction attributes.
> 
> Signed-off-by: Ivan Malov 
> ---

Acked-by: Ori Kam 
Thanks,
Ori

Re: [dpdk-dev] [PATCH] net/ena: remove useless address check

[Ferruh Yigit]

On 10/11/2021 7:49 AM, Michał Krawczyk wrote:

pon., 4 paź 2021 o 19:27 Ferruh Yigit  napisał(a):


Reported by "gcc (GCC) 12.0.0 20211003 (experimental)":

./drivers/net/ena/ena_rss.c: In function ‘ena_rss_reta_query’:
./drivers/net/ena/ena_rss.c:140:66:
 error: the comparison will always evaluate as ‘false’ for the
 pointer operand in ‘reta_conf + 136’ must not be NULL
 [-Werror=address]
   140 |  (reta_size > RTE_RETA_GROUP_SIZE && ((reta_conf + 1) == NULL)))
   |   ^~

Fixing it by removing useless check.

Signed-off-by: Ferruh Yigit 

Acked-by: Michal Krawczyk 



Applied to dpdk-next-net/main, thanks.

[dpdk-dev] [PATCH v4 00/28] Support ingress policer

[skori]

From: Sunil Kumar Kori 

Series adds support for ingress meter for CN10K platform. Series is divided
in following two logical implementation:

 - common/cnxk: It implements RoC APIs over MBOX interfaces which are by 
network drivers.

common/cnxk: update policer MBOX APIs and HW definitions
common/cnxk: support RoC API to get level to index
common/cnxk: support RoC API to get profile count
common/cnxk: support RoC API to alloc bandwidth profiles
common/cnxk: support RoC API to free bandwidth profiles
common/cnxk: support RoC API to configure bandwidth profile
common/cnxk: support RoC API to toggle profile state
common/cnxk: support RoC API to dump bandwidth profile
common/cnxk: support RoC API to setup precolor table
common/cnxk: support RoC API to connect bandwidth profiles
common/cnxk: support RoC API to get stats to index
common/cnxk: support RoC API to read profile statistics
common/cnxk: support RoC API to reset profile stats
common/cnxk: support meter in action list

 - net/cnxk: It implenets ethdev ops for various metering operations.
For cn9k platform, ops are registerded as NULL because cn9k does
not support metering.

net/cnxk: support meter ops get API
net/cnxk: support ops to get meter capabilities
net/cnxk: support ops to create meter profile
net/cnxk: support ops to delete meter profile
net/cnxk: support ops to validate meter policy
net/cnxk: support ops to create meter policy
net/cnxk: support ops to delete meter policy
net/cnxk: support ops to create meter
net/cnxk: support ops to delete meter
net/cnxk: support ops to enable/disable meter
net/cnxk: support ops to update precolor DSCP table
net/cnxk: support ops to read/update meter stats
net/cnxk: support meter action to flow create
net/cnxk: support meter action to flow destroy

 doc/guides/nics/features/cnxk.ini  |1 +
 doc/guides/nics/features/cnxk_vf.ini   |1 +
 doc/guides/rel_notes/release_21_11.rst |1 +
 drivers/common/cnxk/hw/nix.h   |   62 +-
 drivers/common/cnxk/meson.build|1 +
 drivers/common/cnxk/roc_mbox.h |   34 +-
 drivers/common/cnxk/roc_nix.h  |  164 +++
 drivers/common/cnxk/roc_nix_bpf.c  | 1149 
 drivers/common/cnxk/roc_nix_priv.h |1 +
 drivers/common/cnxk/roc_npc.c  |7 +
 drivers/common/cnxk/roc_npc.h  |8 +-
 drivers/common/cnxk/roc_platform.h |1 +
 drivers/common/cnxk/roc_utils.c|3 +
 drivers/common/cnxk/version.map|   15 +
 drivers/net/cnxk/cn10k_rte_flow.c  |  193 +++-
 drivers/net/cnxk/cnxk_ethdev.c |   57 +
 drivers/net/cnxk/cnxk_ethdev.h |   92 ++
 drivers/net/cnxk/cnxk_ethdev_mtr.c | 1341 
 drivers/net/cnxk/cnxk_rte_flow.c   |4 +
 drivers/net/cnxk/meson.build   |1 +
 20 files changed, 3129 insertions(+), 7 deletions(-)
 create mode 100644 drivers/common/cnxk/roc_nix_bpf.c
 create mode 100644 drivers/net/cnxk/cnxk_ethdev_mtr.c

-- 
2.25.1

[dpdk-dev] [PATCH v4 01/28] common/cnxk: update policer MBOX APIs and HW definitions

[skori]

From: Sunil Kumar Kori 

To support ingress policer on CN10K, MBOX interfaces and HW
definitions are synced.

Signed-off-by: Sunil Kumar Kori 
Acked-by: Ray Kinsella 
---
v4:
 - Rebase support on dpdk-next-net-mrvl branch
 - Handled meter action during flow destroy
 - Handled meter cleanup during port shutdown
 
v3:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for tree hierarchy
 - Fix naming convention
v2:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for linear hierarchy
 - Review comments incorporated

 drivers/common/cnxk/hw/nix.h   | 13 ++---
 drivers/common/cnxk/roc_mbox.h | 34 +-
 2 files changed, 43 insertions(+), 4 deletions(-)

diff --git a/drivers/common/cnxk/hw/nix.h b/drivers/common/cnxk/hw/nix.h
index 6a0eb019ac..7685b7dc1b 100644
--- a/drivers/common/cnxk/hw/nix.h
+++ b/drivers/common/cnxk/hw/nix.h
@@ -692,9 +692,16 @@
 #define NIX_RX_BAND_PROF_ACTIONRESULT_DROP (0x1ull) /* [CN10K, .) */
 #define NIX_RX_BAND_PROF_ACTIONRESULT_RED  (0x2ull) /* [CN10K, .) */
 
-#define NIX_RX_BAND_PROF_LAYER_LEAF   (0x0ull) /* [CN10K, .) */
-#define NIX_RX_BAND_PROF_LAYER_MIDDLE (0x1ull) /* [CN10K, .) */
-#define NIX_RX_BAND_PROF_LAYER_TOP(0x2ull) /* [CN10K, .) */
+#define NIX_RX_BAND_PROF_LAYER_LEAF(0x0ull) /* [CN10K, .) */
+#define NIX_RX_BAND_PROF_LAYER_INVALID (0x1ull) /* [CN10K, .) */
+#define NIX_RX_BAND_PROF_LAYER_MIDDLE  (0x2ull) /* [CN10K, .) */
+#define NIX_RX_BAND_PROF_LAYER_TOP (0x3ull) /* [CN10K, .) */
+#define NIX_RX_BAND_PROF_LAYER_MAX (0x4ull) /* [CN10K, .) */
+
+#define NIX_RX_BAND_PROF_PC_MODE_VLAN (0x0ull) /* [CN10K, .) */
+#define NIX_RX_BAND_PROF_PC_MODE_DSCP (0x1ull) /* [CN10K, .) */
+#define NIX_RX_BAND_PROF_PC_MODE_GEN  (0x2ull) /* [CN10K, .) */
+#define NIX_RX_BAND_PROF_PC_MODE_RSVD (0x3ull) /* [CN10K, .) */
 
 #define NIX_RX_COLORRESULT_GREEN  (0x0ull) /* [CN10K, .) */
 #define NIX_RX_COLORRESULT_YELLOW (0x1ull) /* [CN10K, .) */
diff --git a/drivers/common/cnxk/roc_mbox.h b/drivers/common/cnxk/roc_mbox.h
index 75d1ff1ef3..bc40848450 100644
--- a/drivers/common/cnxk/roc_mbox.h
+++ b/drivers/common/cnxk/roc_mbox.h
@@ -234,7 +234,11 @@ struct mbox_msghdr {
  nix_inline_ipsec_lf_cfg, msg_rsp)\
M(NIX_CN10K_AQ_ENQ, 0x801b, nix_cn10k_aq_enq, nix_cn10k_aq_enq_req,\
  nix_cn10k_aq_enq_rsp)\
-   M(NIX_GET_HW_INFO, 0x801c, nix_get_hw_info, msg_req, nix_hw_info)
+   M(NIX_GET_HW_INFO, 0x801c, nix_get_hw_info, msg_req, nix_hw_info)  \
+   M(NIX_BANDPROF_ALLOC, 0x801d, nix_bandprof_alloc,  \
+ nix_bandprof_alloc_req, nix_bandprof_alloc_rsp)  \
+   M(NIX_BANDPROF_FREE, 0x801e, nix_bandprof_free, nix_bandprof_free_req, \
+ msg_rsp)
 
 /* Messages initiated by AF (range 0xC00 - 0xDFF) */
 #define MBOX_UP_CGX_MESSAGES   
\
@@ -772,6 +776,10 @@ struct nix_cn10k_aq_enq_req {
__io struct nix_rsse_s rss;
/* Valid when op == WRITE/INIT and ctype == NIX_AQ_CTYPE_MCE */
__io struct nix_rx_mce_s mce;
+   /* Valid when op == WRITE/INIT and
+* ctype == NIX_AQ_CTYPE_BAND_PROF
+*/
+   __io struct nix_band_prof_s prof;
};
/* Mask data when op == WRITE (1=write, 0=don't write) */
union {
@@ -785,6 +793,8 @@ struct nix_cn10k_aq_enq_req {
__io struct nix_rsse_s rss_mask;
/* Valid when op == WRITE and ctype == NIX_AQ_CTYPE_MCE */
__io struct nix_rx_mce_s mce_mask;
+   /* Valid when op == WRITE and ctype == NIX_AQ_CTYPE_BAND_PROF */
+   __io struct nix_band_prof_s prof_mask;
};
 };
 
@@ -796,6 +806,7 @@ struct nix_cn10k_aq_enq_rsp {
struct nix_cq_ctx_s cq;
struct nix_rsse_s rss;
struct nix_rx_mce_s mce;
+   struct nix_band_prof_s prof;
};
 };
 
@@ -1130,6 +1141,27 @@ struct nix_hw_info {
uint16_t __io rsvd[15];
 };
 
+struct nix_bandprof_alloc_req {
+   struct mbox_msghdr hdr;
+   /* Count of profiles needed per layer */
+   uint16_t __io prof_count[NIX_RX_BAND_PROF_LAYER_MAX];
+};
+
+struct nix_bandprof_alloc_rsp {
+   struct mbox_msghdr hdr;
+   uint16_t __io prof_count[NIX_RX_BAND_PROF_LAYER_MAX];
+
+#define BANDPROF_PER_PFFUNC 64
+   uint16_t __io prof_idx[NIX_RX_BAND_PROF_LAYER_MAX][BANDPROF_PER_PFFUNC];
+};
+
+struct nix_bandprof_free_req {
+   struct mbox_msghdr hdr;
+   uint8_t __io free_all;
+   uint16_t __io prof_count[NIX_RX_BAND_PROF_LAYER_MAX];
+   uint16_t __io prof_idx[NIX_RX_BAND_PROF_LAYER_MAX][BANDPROF_PER_PFFUNC];
+};
+
 /* SSO mailbox error codes
  * Range 501 - 600.
  */
-- 
2.25.1

[dpdk-dev] [PATCH v4 02/28] common/cnxk: support RoC API to get level to index

[skori]

From: Sunil Kumar Kori 

CN10K platform supports policer up to 3 level of hierarchy.
Implement RoC API to get corresponding index for given level.

Signed-off-by: Sunil Kumar Kori 
---
v4:
 - Rebase support on dpdk-next-net-mrvl branch
 - Handled meter action during flow destroy
 - Handled meter cleanup during port shutdown
 
v3:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for tree hierarchy
 - Fix naming convention
v2:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for linear hierarchy
 - Review comments incorporated

 drivers/common/cnxk/meson.build|  1 +
 drivers/common/cnxk/roc_nix.h  | 12 
 drivers/common/cnxk/roc_nix_bpf.c  | 22 ++
 drivers/common/cnxk/roc_nix_priv.h |  1 +
 drivers/common/cnxk/roc_utils.c|  3 +++
 drivers/common/cnxk/version.map|  1 +
 6 files changed, 40 insertions(+)
 create mode 100644 drivers/common/cnxk/roc_nix_bpf.c

diff --git a/drivers/common/cnxk/meson.build b/drivers/common/cnxk/meson.build
index 97db5f087b..32d3ad5087 100644
--- a/drivers/common/cnxk/meson.build
+++ b/drivers/common/cnxk/meson.build
@@ -25,6 +25,7 @@ sources = files(
 'roc_mbox.c',
 'roc_model.c',
 'roc_nix.c',
+'roc_nix_bpf.c',
 'roc_nix_debug.c',
 'roc_nix_fc.c',
 'roc_nix_irq.c',
diff --git a/drivers/common/cnxk/roc_nix.h b/drivers/common/cnxk/roc_nix.h
index b06895a565..bf4fd2f208 100644
--- a/drivers/common/cnxk/roc_nix.h
+++ b/drivers/common/cnxk/roc_nix.h
@@ -6,6 +6,8 @@
 #define _ROC_NIX_H_
 
 /* Constants */
+#define ROC_NIX_BPF_LEVEL_IDX_INVALID 0xFF
+
 enum roc_nix_rss_reta_sz {
ROC_NIX_RSS_RETA_SZ_64 = 64,
ROC_NIX_RSS_RETA_SZ_128 = 128,
@@ -29,6 +31,12 @@ enum roc_nix_vlan_type {
ROC_NIX_VLAN_TYPE_OUTER = 0x02,
 };
 
+enum roc_nix_bpf_level_flag {
+   ROC_NIX_BPF_LEVEL_F_LEAF = BIT(0),
+   ROC_NIX_BPF_LEVEL_F_MID = BIT(1),
+   ROC_NIX_BPF_LEVEL_F_TOP = BIT(2),
+};
+
 struct roc_nix_vlan_config {
uint32_t type;
union {
@@ -508,6 +516,10 @@ int __roc_api roc_nix_tm_prepare_rate_limited_tree(struct 
roc_nix *roc_nix);
 bool __roc_api roc_nix_tm_is_user_hierarchy_enabled(struct roc_nix *nix);
 int __roc_api roc_nix_tm_tree_type_get(struct roc_nix *nix);
 
+/* Ingress Policer API */
+uint8_t __roc_api
+roc_nix_bpf_level_to_idx(enum roc_nix_bpf_level_flag lvl_flag);
+
 /* MAC */
 int __roc_api roc_nix_mac_rxtx_start_stop(struct roc_nix *roc_nix, bool start);
 int __roc_api roc_nix_mac_link_event_start_stop(struct roc_nix *roc_nix,
diff --git a/drivers/common/cnxk/roc_nix_bpf.c 
b/drivers/common/cnxk/roc_nix_bpf.c
new file mode 100644
index 00..b588cc16e4
--- /dev/null
+++ b/drivers/common/cnxk/roc_nix_bpf.c
@@ -0,0 +1,22 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(C) 2021 Marvell.
+ */
+
+#include "roc_api.h"
+#include "roc_priv.h"
+
+uint8_t
+roc_nix_bpf_level_to_idx(enum roc_nix_bpf_level_flag level_f)
+{
+   uint8_t idx;
+
+   if (level_f & ROC_NIX_BPF_LEVEL_F_LEAF)
+   idx = 0;
+   else if (level_f & ROC_NIX_BPF_LEVEL_F_MID)
+   idx = 1;
+   else if (level_f & ROC_NIX_BPF_LEVEL_F_TOP)
+   idx = 2;
+   else
+   idx = ROC_NIX_BPF_LEVEL_IDX_INVALID;
+   return idx;
+}
diff --git a/drivers/common/cnxk/roc_nix_priv.h 
b/drivers/common/cnxk/roc_nix_priv.h
index 7653c5a328..02b1be852d 100644
--- a/drivers/common/cnxk/roc_nix_priv.h
+++ b/drivers/common/cnxk/roc_nix_priv.h
@@ -187,6 +187,7 @@ enum nix_err_status {
NIX_ERR_INVALID_RANGE,
NIX_ERR_INTERNAL,
NIX_ERR_OP_NOTSUP,
+   NIX_ERR_HW_NOTSUP,
NIX_ERR_QUEUE_INVALID_RANGE,
NIX_ERR_AQ_READ_FAILED,
NIX_ERR_AQ_WRITE_FAILED,
diff --git a/drivers/common/cnxk/roc_utils.c b/drivers/common/cnxk/roc_utils.c
index 751486f503..f1b5ef3b70 100644
--- a/drivers/common/cnxk/roc_utils.c
+++ b/drivers/common/cnxk/roc_utils.c
@@ -34,6 +34,9 @@ roc_error_msg_get(int errorcode)
case NIX_ERR_OP_NOTSUP:
err_msg = "Operation not supported";
break;
+   case NIX_ERR_HW_NOTSUP:
+   err_msg = "Hardware does not support";
+   break;
case NIX_ERR_QUEUE_INVALID_RANGE:
err_msg = "Invalid Queue range";
break;
diff --git a/drivers/common/cnxk/version.map b/drivers/common/cnxk/version.map
index ccf8ec157e..c84c229938 100644
--- a/drivers/common/cnxk/version.map
+++ b/drivers/common/cnxk/version.map
@@ -82,6 +82,7 @@ INTERNAL {
roc_model;
roc_se_auth_key_set;
roc_se_ciph_key_set;
+   roc_nix_bpf_level_to_idx;
roc_nix_cq_dump;
roc_nix_cq_fini;
roc_nix_cq_init;
-- 
2.25.1

[dpdk-dev] [PATCH v4 03/28] common/cnxk: support RoC API to get profile count

[skori]

From: Sunil Kumar Kori 

Implement interface to get available profile count for given
nixlf.

Signed-off-by: Sunil Kumar Kori 
---
v4:
 - Rebase support on dpdk-next-net-mrvl branch
 - Handled meter action during flow destroy
 - Handled meter cleanup during port shutdown
 
v3:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for tree hierarchy
 - Fix naming convention
v2:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for linear hierarchy
 - Review comments incorporated

 drivers/common/cnxk/roc_nix.h |  5 
 drivers/common/cnxk/roc_nix_bpf.c | 46 +++
 drivers/common/cnxk/version.map   |  1 +
 3 files changed, 52 insertions(+)

diff --git a/drivers/common/cnxk/roc_nix.h b/drivers/common/cnxk/roc_nix.h
index bf4fd2f208..4fef6d44e7 100644
--- a/drivers/common/cnxk/roc_nix.h
+++ b/drivers/common/cnxk/roc_nix.h
@@ -7,6 +7,7 @@
 
 /* Constants */
 #define ROC_NIX_BPF_LEVEL_IDX_INVALID 0xFF
+#define ROC_NIX_BPF_LEVEL_MAX3
 
 enum roc_nix_rss_reta_sz {
ROC_NIX_RSS_RETA_SZ_64 = 64,
@@ -517,6 +518,10 @@ bool __roc_api roc_nix_tm_is_user_hierarchy_enabled(struct 
roc_nix *nix);
 int __roc_api roc_nix_tm_tree_type_get(struct roc_nix *nix);
 
 /* Ingress Policer API */
+int __roc_api
+roc_nix_bpf_count_get(struct roc_nix *roc_nix, uint8_t lvl_mask,
+ uint16_t count[ROC_NIX_BPF_LEVEL_MAX] /* Out */);
+
 uint8_t __roc_api
 roc_nix_bpf_level_to_idx(enum roc_nix_bpf_level_flag lvl_flag);
 
diff --git a/drivers/common/cnxk/roc_nix_bpf.c 
b/drivers/common/cnxk/roc_nix_bpf.c
index b588cc16e4..af9dffa90c 100644
--- a/drivers/common/cnxk/roc_nix_bpf.c
+++ b/drivers/common/cnxk/roc_nix_bpf.c
@@ -5,6 +5,14 @@
 #include "roc_api.h"
 #include "roc_priv.h"
 
+#define NIX_MAX_BPF_COUNT_LEAF_LAYER 64
+#define NIX_MAX_BPF_COUNT_MID_LAYER  8
+#define NIX_MAX_BPF_COUNT_TOP_LAYER  1
+
+#define NIX_BPF_LEVEL_F_MASK   
\
+   (ROC_NIX_BPF_LEVEL_F_LEAF | ROC_NIX_BPF_LEVEL_F_MID |  \
+ROC_NIX_BPF_LEVEL_F_TOP)
+
 uint8_t
 roc_nix_bpf_level_to_idx(enum roc_nix_bpf_level_flag level_f)
 {
@@ -20,3 +28,41 @@ roc_nix_bpf_level_to_idx(enum roc_nix_bpf_level_flag level_f)
idx = ROC_NIX_BPF_LEVEL_IDX_INVALID;
return idx;
 }
+
+int
+roc_nix_bpf_count_get(struct roc_nix *roc_nix, uint8_t lvl_mask,
+ uint16_t count[ROC_NIX_BPF_LEVEL_MAX])
+{
+   uint8_t mask = lvl_mask & NIX_BPF_LEVEL_F_MASK;
+   uint8_t leaf_idx, mid_idx, top_idx;
+
+   PLT_SET_USED(roc_nix);
+
+   if (roc_model_is_cn9k())
+   return NIX_ERR_HW_NOTSUP;
+
+   if (!mask)
+   return NIX_ERR_PARAM;
+
+   /* Currently No MBOX interface is available to get number
+* of bandwidth profiles. So numbers per level are hard coded,
+* considering 3 RPM blocks and each block has 4 LMAC's.
+* So total 12 physical interfaces are in system. Each interface
+* supports following bandwidth profiles.
+*/
+
+   leaf_idx = roc_nix_bpf_level_to_idx(mask & ROC_NIX_BPF_LEVEL_F_LEAF);
+   mid_idx = roc_nix_bpf_level_to_idx(mask & ROC_NIX_BPF_LEVEL_F_MID);
+   top_idx = roc_nix_bpf_level_to_idx(mask & ROC_NIX_BPF_LEVEL_F_TOP);
+
+   if (leaf_idx != ROC_NIX_BPF_LEVEL_IDX_INVALID)
+   count[leaf_idx] = NIX_MAX_BPF_COUNT_LEAF_LAYER;
+
+   if (mid_idx != ROC_NIX_BPF_LEVEL_IDX_INVALID)
+   count[mid_idx] = NIX_MAX_BPF_COUNT_MID_LAYER;
+
+   if (top_idx != ROC_NIX_BPF_LEVEL_IDX_INVALID)
+   count[top_idx] = NIX_MAX_BPF_COUNT_TOP_LAYER;
+
+   return 0;
+}
diff --git a/drivers/common/cnxk/version.map b/drivers/common/cnxk/version.map
index c84c229938..6990c1f084 100644
--- a/drivers/common/cnxk/version.map
+++ b/drivers/common/cnxk/version.map
@@ -82,6 +82,7 @@ INTERNAL {
roc_model;
roc_se_auth_key_set;
roc_se_ciph_key_set;
+   roc_nix_bpf_count_get;
roc_nix_bpf_level_to_idx;
roc_nix_cq_dump;
roc_nix_cq_fini;
-- 
2.25.1

Re: [dpdk-dev] [PATCH v3 02/12] ethdev: add represented port item to flow API

[Ori Kam]

Hi Ivan,

> -Original Message-
> From: Ivan Malov 
> Sent: Sunday, October 10, 2021 5:39 PM
> To: dev@dpdk.org
> Cc: NBU-Contact-Thomas Monjalon ; Ori Kam 
> ; Xiaoyun Li
> ; Ferruh Yigit ; Andrew 
> Rybchenko
> 
> Subject: [PATCH v3 02/12] ethdev: add represented port item to flow API
> 
> For use in "transfer" flows. Supposed to match traffic entering the embedded 
> switch from the entity
> represented by the given ethdev.
> Such an entity can be a network (via a network port), a guest machine (via a 
> VF) or another ethdev in the
> same application.
> 
> Must not be combined with direction attributes.
> 
> Signed-off-by: Ivan Malov 
> ---

Acked-by: Ori Kam 
Thanks,
Ori

[dpdk-dev] [PATCH v4 04/28] common/cnxk: support RoC API to alloc bandwidth profiles

[skori]

From: Sunil Kumar Kori 

Implement RoC API to allocate HW resources i.e. bandwidth
profiles for policer processing on CN10K platform.

Signed-off-by: Sunil Kumar Kori 
---
v4:
 - Rebase support on dpdk-next-net-mrvl branch
 - Handled meter action during flow destroy
 - Handled meter cleanup during port shutdown
 
v3:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for tree hierarchy
 - Fix naming convention
v2:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for linear hierarchy
 - Review comments incorporated

 drivers/common/cnxk/roc_nix.h |  11 
 drivers/common/cnxk/roc_nix_bpf.c | 104 ++
 drivers/common/cnxk/version.map   |   1 +
 3 files changed, 116 insertions(+)

diff --git a/drivers/common/cnxk/roc_nix.h b/drivers/common/cnxk/roc_nix.h
index 4fef6d44e7..b192a2b217 100644
--- a/drivers/common/cnxk/roc_nix.h
+++ b/drivers/common/cnxk/roc_nix.h
@@ -6,6 +6,7 @@
 #define _ROC_NIX_H_
 
 /* Constants */
+#define ROC_NIX_BPF_PER_PFFUNC   64
 #define ROC_NIX_BPF_LEVEL_IDX_INVALID 0xFF
 #define ROC_NIX_BPF_LEVEL_MAX3
 
@@ -38,6 +39,12 @@ enum roc_nix_bpf_level_flag {
ROC_NIX_BPF_LEVEL_F_TOP = BIT(2),
 };
 
+struct roc_nix_bpf_objs {
+   uint16_t level;
+   uint16_t count;
+   uint16_t ids[ROC_NIX_BPF_PER_PFFUNC];
+};
+
 struct roc_nix_vlan_config {
uint32_t type;
union {
@@ -522,6 +529,10 @@ int __roc_api
 roc_nix_bpf_count_get(struct roc_nix *roc_nix, uint8_t lvl_mask,
  uint16_t count[ROC_NIX_BPF_LEVEL_MAX] /* Out */);
 
+int __roc_api roc_nix_bpf_alloc(struct roc_nix *roc_nix, uint8_t lvl_mask,
+   uint16_t per_lvl_cnt[ROC_NIX_BPF_LEVEL_MAX],
+   struct roc_nix_bpf_objs *profs /* Out */);
+
 uint8_t __roc_api
 roc_nix_bpf_level_to_idx(enum roc_nix_bpf_level_flag lvl_flag);
 
diff --git a/drivers/common/cnxk/roc_nix_bpf.c 
b/drivers/common/cnxk/roc_nix_bpf.c
index af9dffa90c..06394bda07 100644
--- a/drivers/common/cnxk/roc_nix_bpf.c
+++ b/drivers/common/cnxk/roc_nix_bpf.c
@@ -13,6 +13,19 @@
(ROC_NIX_BPF_LEVEL_F_LEAF | ROC_NIX_BPF_LEVEL_F_MID |  \
 ROC_NIX_BPF_LEVEL_F_TOP)
 
+static uint8_t sw_to_hw_lvl_map[] = {NIX_RX_BAND_PROF_LAYER_LEAF,
+NIX_RX_BAND_PROF_LAYER_MIDDLE,
+NIX_RX_BAND_PROF_LAYER_TOP};
+
+static inline struct mbox *
+get_mbox(struct roc_nix *roc_nix)
+{
+   struct nix *nix = roc_nix_to_nix_priv(roc_nix);
+   struct dev *dev = &nix->dev;
+
+   return dev->mbox;
+}
+
 uint8_t
 roc_nix_bpf_level_to_idx(enum roc_nix_bpf_level_flag level_f)
 {
@@ -66,3 +79,94 @@ roc_nix_bpf_count_get(struct roc_nix *roc_nix, uint8_t 
lvl_mask,
 
return 0;
 }
+
+int
+roc_nix_bpf_alloc(struct roc_nix *roc_nix, uint8_t lvl_mask,
+ uint16_t per_lvl_cnt[ROC_NIX_BPF_LEVEL_MAX],
+ struct roc_nix_bpf_objs *profs)
+{
+   uint8_t mask = lvl_mask & NIX_BPF_LEVEL_F_MASK;
+   struct mbox *mbox = get_mbox(roc_nix);
+   struct nix_bandprof_alloc_req *req;
+   struct nix_bandprof_alloc_rsp *rsp;
+   uint8_t leaf_idx, mid_idx, top_idx;
+   int rc = -ENOSPC, i;
+
+   if (roc_model_is_cn9k())
+   return NIX_ERR_HW_NOTSUP;
+
+   if (!mask)
+   return NIX_ERR_PARAM;
+
+   leaf_idx = roc_nix_bpf_level_to_idx(mask & ROC_NIX_BPF_LEVEL_F_LEAF);
+   mid_idx = roc_nix_bpf_level_to_idx(mask & ROC_NIX_BPF_LEVEL_F_MID);
+   top_idx = roc_nix_bpf_level_to_idx(mask & ROC_NIX_BPF_LEVEL_F_TOP);
+
+   if ((leaf_idx != ROC_NIX_BPF_LEVEL_IDX_INVALID) &&
+   (per_lvl_cnt[leaf_idx] > NIX_MAX_BPF_COUNT_LEAF_LAYER))
+   return NIX_ERR_INVALID_RANGE;
+
+   if ((mid_idx != ROC_NIX_BPF_LEVEL_IDX_INVALID) &&
+   (per_lvl_cnt[mid_idx] > NIX_MAX_BPF_COUNT_MID_LAYER))
+   return NIX_ERR_INVALID_RANGE;
+
+   if ((top_idx != ROC_NIX_BPF_LEVEL_IDX_INVALID) &&
+   (per_lvl_cnt[top_idx] > NIX_MAX_BPF_COUNT_TOP_LAYER))
+   return NIX_ERR_INVALID_RANGE;
+
+   req = mbox_alloc_msg_nix_bandprof_alloc(mbox);
+   if (req == NULL)
+   goto exit;
+
+   if (leaf_idx != ROC_NIX_BPF_LEVEL_IDX_INVALID) {
+   req->prof_count[sw_to_hw_lvl_map[leaf_idx]] =
+   per_lvl_cnt[leaf_idx];
+   }
+
+   if (mid_idx != ROC_NIX_BPF_LEVEL_IDX_INVALID) {
+   req->prof_count[sw_to_hw_lvl_map[mid_idx]] =
+   per_lvl_cnt[mid_idx];
+   }
+
+   if (top_idx != ROC_NIX_BPF_LEVEL_IDX_INVALID) {
+   req->prof_count[sw_to_hw_lvl_map[top_idx]] =
+   per_lvl_cnt[top_idx];
+   }
+
+   rc = mbox_process_msg(mbox, (void *)&rsp);
+   if (rc)
+   goto exit;
+
+   if (leaf_idx != ROC_NIX_BPF_LEVEL_IDX_INVALID) {
+   profs[leaf_idx].level = leaf_idx;
+

[dpdk-dev] [PATCH v4 05/28] common/cnxk: support RoC API to free bandwidth profiles

[skori]

From: Sunil Kumar Kori 

Implement RoC interface to free HW bandwidth profiles on
CN10K platform.

Signed-off-by: Sunil Kumar Kori 
---
v4:
 - Rebase support on dpdk-next-net-mrvl branch
 - Handled meter action during flow destroy
 - Handled meter cleanup during port shutdown
 
v3:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for tree hierarchy
 - Fix naming convention
v2:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for linear hierarchy
 - Review comments incorporated

 drivers/common/cnxk/roc_nix.h |  6 +
 drivers/common/cnxk/roc_nix_bpf.c | 40 +++
 drivers/common/cnxk/version.map   |  2 ++
 3 files changed, 48 insertions(+)

diff --git a/drivers/common/cnxk/roc_nix.h b/drivers/common/cnxk/roc_nix.h
index b192a2b217..a5642337b2 100644
--- a/drivers/common/cnxk/roc_nix.h
+++ b/drivers/common/cnxk/roc_nix.h
@@ -533,6 +533,12 @@ int __roc_api roc_nix_bpf_alloc(struct roc_nix *roc_nix, 
uint8_t lvl_mask,
uint16_t per_lvl_cnt[ROC_NIX_BPF_LEVEL_MAX],
struct roc_nix_bpf_objs *profs /* Out */);
 
+int __roc_api roc_nix_bpf_free(struct roc_nix *roc_nix,
+  struct roc_nix_bpf_objs *profs,
+  uint8_t num_prof);
+
+int __roc_api roc_nix_bpf_free_all(struct roc_nix *roc_nix);
+
 uint8_t __roc_api
 roc_nix_bpf_level_to_idx(enum roc_nix_bpf_level_flag lvl_flag);
 
diff --git a/drivers/common/cnxk/roc_nix_bpf.c 
b/drivers/common/cnxk/roc_nix_bpf.c
index 06394bda07..41d31bc6cd 100644
--- a/drivers/common/cnxk/roc_nix_bpf.c
+++ b/drivers/common/cnxk/roc_nix_bpf.c
@@ -170,3 +170,43 @@ roc_nix_bpf_alloc(struct roc_nix *roc_nix, uint8_t 
lvl_mask,
 exit:
return rc;
 }
+
+int
+roc_nix_bpf_free(struct roc_nix *roc_nix, struct roc_nix_bpf_objs *profs,
+uint8_t num_prof)
+{
+   struct mbox *mbox = get_mbox(roc_nix);
+   struct nix_bandprof_free_req *req;
+   uint8_t level;
+   int i, j;
+
+   if (num_prof >= NIX_RX_BAND_PROF_LAYER_MAX)
+   return NIX_ERR_INVALID_RANGE;
+
+   req = mbox_alloc_msg_nix_bandprof_free(mbox);
+   if (req == NULL)
+   return -ENOSPC;
+
+   for (i = 0; i < num_prof; i++) {
+   level = sw_to_hw_lvl_map[profs[i].level];
+   req->prof_count[level] = profs[i].count;
+   for (j = 0; j < profs[i].count; j++)
+   req->prof_idx[level][j] = profs[i].ids[j];
+   }
+
+   return mbox_process(mbox);
+}
+
+int
+roc_nix_bpf_free_all(struct roc_nix *roc_nix)
+{
+   struct mbox *mbox = get_mbox(roc_nix);
+   struct nix_bandprof_free_req *req;
+
+   req = mbox_alloc_msg_nix_bandprof_free(mbox);
+   if (req == NULL)
+   return -ENOSPC;
+
+   req->free_all = true;
+   return mbox_process(mbox);
+}
diff --git a/drivers/common/cnxk/version.map b/drivers/common/cnxk/version.map
index b8863a1155..3d474f8a96 100644
--- a/drivers/common/cnxk/version.map
+++ b/drivers/common/cnxk/version.map
@@ -84,6 +84,8 @@ INTERNAL {
roc_se_ciph_key_set;
roc_nix_bpf_alloc;
roc_nix_bpf_count_get;
+   roc_nix_bpf_free;
+   roc_nix_bpf_free_all;
roc_nix_bpf_level_to_idx;
roc_nix_cq_dump;
roc_nix_cq_fini;
-- 
2.25.1

[dpdk-dev] [PATCH v4 06/28] common/cnxk: support RoC API to configure bandwidth profile

[skori]

From: Sunil Kumar Kori 

Implement RoC API to configure HW bandwidth profile for
CN10K platform.

Signed-off-by: Sunil Kumar Kori 
---
v4:
 - Rebase support on dpdk-next-net-mrvl branch
 - Handled meter action during flow destroy
 - Handled meter cleanup during port shutdown
 
v3:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for tree hierarchy
 - Fix naming convention
v2:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for linear hierarchy
 - Review comments incorporated

 drivers/common/cnxk/hw/nix.h  |  49 ++
 drivers/common/cnxk/roc_nix.h |  63 
 drivers/common/cnxk/roc_nix_bpf.c | 239 ++
 drivers/common/cnxk/version.map   |   1 +
 4 files changed, 352 insertions(+)

diff --git a/drivers/common/cnxk/hw/nix.h b/drivers/common/cnxk/hw/nix.h
index 7685b7dc1b..b9e65c942f 100644
--- a/drivers/common/cnxk/hw/nix.h
+++ b/drivers/common/cnxk/hw/nix.h
@@ -2113,6 +2113,55 @@ struct nix_lso_format {
 #define NIX_RPM_MAX_HW_FRS  16380UL
 #define NIX_MIN_HW_FRS 60UL
 
+/** NIX policer rate limits */
+#define NIX_BPF_MAX_RATE_DIV_EXP  12
+#define NIX_BPF_MAX_RATE_EXPONENT 0xf
+#define NIX_BPF_MAX_RATE_MANTISSA 0xff
+
+#define NIX_BPF_RATE_CONST 200ULL
+
+/* NIX rate calculation in Bits/Sec
+ * PIR_ADD = ((256 + NIX_*_PIR[RATE_MANTISSA])
+ * << NIX_*_PIR[RATE_EXPONENT]) / 256
+ * PIR = (2E6 * PIR_ADD / (1 << NIX_*_PIR[RATE_DIVIDER_EXPONENT]))
+ *
+ * CIR_ADD = ((256 + NIX_*_CIR[RATE_MANTISSA])
+ * << NIX_*_CIR[RATE_EXPONENT]) / 256
+ * CIR = (2E6 * CIR_ADD / (CCLK_TICKS << NIX_*_CIR[RATE_DIVIDER_EXPONENT]))
+ */
+#define NIX_BPF_RATE(exponent, mantissa, div_exp)  
\
+   ((NIX_BPF_RATE_CONST * ((256 + (mantissa)) << (exponent))) /   \
+(((1ull << (div_exp)) * 256)))
+
+/* Meter rate limits in Bits/Sec */
+#define NIX_BPF_RATE_MIN NIX_BPF_RATE(0, 0, NIX_BPF_MAX_RATE_DIV_EXP)
+#define NIX_BPF_RATE_MAX   
\
+   NIX_BPF_RATE(NIX_BPF_MAX_RATE_EXPONENT, NIX_BPF_MAX_RATE_MANTISSA, 0)
+
+#define NIX_BPF_DEFAULT_ADJUST_MANTISSA 511
+#define NIX_BPF_DEFAULT_ADJUST_EXPONENT 0
+
+/** NIX burst limits */
+#define NIX_BPF_MAX_BURST_EXPONENT 0xf
+#define NIX_BPF_MAX_BURST_MANTISSA 0xff
+
+/* NIX burst calculation
+ * PIR_BURST = ((256 + NIX_*_PIR[BURST_MANTISSA])
+ * << (NIX_*_PIR[BURST_EXPONENT] + 1))
+ * / 256
+ *
+ * CIR_BURST = ((256 + NIX_*_CIR[BURST_MANTISSA])
+ * << (NIX_*_CIR[BURST_EXPONENT] + 1))
+ * / 256
+ */
+#define NIX_BPF_BURST(exponent, mantissa)  
\
+   (((256 + (mantissa)) << ((exponent) + 1)) / 256)
+
+/** Meter burst limits */
+#define NIX_BPF_BURST_MIN NIX_BPF_BURST(0, 0)
+#define NIX_BPF_BURST_MAX  
\
+   NIX_BPF_BURST(NIX_BPF_MAX_BURST_EXPONENT, NIX_BPF_MAX_BURST_MANTISSA)
+
 /* NIX rate limits */
 #define NIX_TM_MAX_RATE_DIV_EXP 12
 #define NIX_TM_MAX_RATE_EXPONENT 0xf
diff --git a/drivers/common/cnxk/roc_nix.h b/drivers/common/cnxk/roc_nix.h
index a5642337b2..c2cc823516 100644
--- a/drivers/common/cnxk/roc_nix.h
+++ b/drivers/common/cnxk/roc_nix.h
@@ -39,6 +39,65 @@ enum roc_nix_bpf_level_flag {
ROC_NIX_BPF_LEVEL_F_TOP = BIT(2),
 };
 
+enum roc_nix_bpf_color {
+   ROC_NIX_BPF_COLOR_GREEN,
+   ROC_NIX_BPF_COLOR_YELLOW,
+   ROC_NIX_BPF_COLOR_RED,
+   ROC_NIX_BPF_COLOR_MAX
+};
+
+enum roc_nix_bpf_algo {
+   ROC_NIX_BPF_ALGO_NONE,
+   ROC_NIX_BPF_ALGO_2698,
+   ROC_NIX_BPF_ALGO_4115,
+   ROC_NIX_BPF_ALGO_2697
+};
+
+enum roc_nix_bpf_lmode { ROC_NIX_BPF_LMODE_BYTE, ROC_NIX_BPF_LMODE_PACKET };
+
+enum roc_nix_bpf_action {
+   ROC_NIX_BPF_ACTION_PASS,
+   ROC_NIX_BPF_ACTION_DROP,
+   ROC_NIX_BPF_ACTION_RED
+};
+
+struct roc_nix_bpf_cfg {
+   enum roc_nix_bpf_algo alg;
+   enum roc_nix_bpf_lmode lmode;
+   enum roc_nix_bpf_color icolor;
+   enum roc_nix_bpf_pc_mode pc_mode;
+   bool tnl_ena;
+   union {
+   /* Valid when *alg* is set to ROC_NIX_BPF_ALGO_2697. */
+   struct {
+   uint64_t cir;
+   uint64_t cbs;
+   uint64_t ebs;
+   } algo2697;
+
+   /* Valid when *alg* is set to ROC_NIX_BPF_ALGO_2698. */
+   struct {
+   uint64_t cir;
+   uint64_t pir;
+   uint64_t cbs;
+   uint64_t pbs;
+   } algo2698;
+
+   /* Valid when *alg* is set to ROC_NIX_BPF_ALGO_4115. */
+   struct {
+   uint64_t cir;
+   uint64_t eir;
+   uint64_t cbs;
+   uint64_t ebs;
+   } algo4115;
+   };
+
+   enum roc_nix_bpf_action action[ROC

[dpdk-dev] [PATCH v4 07/28] common/cnxk: support RoC API to toggle profile state

[skori]

From: Sunil Kumar Kori 

Implement RoC API to enable or disable HW bandwidth profiles
on CN10K platform.

Signed-off-by: Sunil Kumar Kori 
---
v4:
 - Rebase support on dpdk-next-net-mrvl branch
 - Handled meter action during flow destroy
 - Handled meter cleanup during port shutdown
 
v3:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for tree hierarchy
 - Fix naming convention
v2:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for linear hierarchy
 - Review comments incorporated

 drivers/common/cnxk/roc_nix.h |  4 
 drivers/common/cnxk/roc_nix_bpf.c | 39 +++
 drivers/common/cnxk/version.map   |  1 +
 3 files changed, 44 insertions(+)

diff --git a/drivers/common/cnxk/roc_nix.h b/drivers/common/cnxk/roc_nix.h
index c2cc823516..5b3bf38a82 100644
--- a/drivers/common/cnxk/roc_nix.h
+++ b/drivers/common/cnxk/roc_nix.h
@@ -217,6 +217,7 @@ struct roc_nix_stats_queue {
 struct roc_nix_rq {
/* Input parameters */
uint16_t qid;
+   uint16_t bpf_id;
uint64_t aura_handle;
bool ipsech_ena;
uint16_t first_skip;
@@ -602,6 +603,9 @@ int __roc_api roc_nix_bpf_config(struct roc_nix *roc_nix, 
uint16_t id,
 enum roc_nix_bpf_level_flag lvl_flag,
 struct roc_nix_bpf_cfg *cfg);
 
+int __roc_api roc_nix_bpf_ena_dis(struct roc_nix *roc_nix, uint16_t id,
+ struct roc_nix_rq *rq, bool enable);
+
 uint8_t __roc_api
 roc_nix_bpf_level_to_idx(enum roc_nix_bpf_level_flag lvl_flag);
 
diff --git a/drivers/common/cnxk/roc_nix_bpf.c 
b/drivers/common/cnxk/roc_nix_bpf.c
index aa5829ee42..c61a697f1a 100644
--- a/drivers/common/cnxk/roc_nix_bpf.c
+++ b/drivers/common/cnxk/roc_nix_bpf.c
@@ -449,3 +449,42 @@ roc_nix_bpf_config(struct roc_nix *roc_nix, uint16_t id,
 
return mbox_process(mbox);
 }
+
+int
+roc_nix_bpf_ena_dis(struct roc_nix *roc_nix, uint16_t id, struct roc_nix_rq 
*rq,
+   bool enable)
+{
+   struct nix *nix = roc_nix_to_nix_priv(roc_nix);
+   struct mbox *mbox = get_mbox(roc_nix);
+   struct nix_cn10k_aq_enq_req *aq;
+   int rc;
+
+   if (roc_model_is_cn9k())
+   return NIX_ERR_HW_NOTSUP;
+
+   if (rq->qid >= nix->nb_rx_queues)
+   return NIX_ERR_QUEUE_INVALID_RANGE;
+
+   aq = mbox_alloc_msg_nix_cn10k_aq_enq(mbox);
+   if (aq == NULL)
+   return -ENOSPC;
+   aq->qidx = rq->qid;
+   aq->ctype = NIX_AQ_CTYPE_RQ;
+   aq->op = NIX_AQ_INSTOP_WRITE;
+
+   aq->rq.policer_ena = enable;
+   aq->rq_mask.policer_ena = ~(aq->rq_mask.policer_ena);
+   if (enable) {
+   aq->rq.band_prof_id = id;
+   aq->rq_mask.band_prof_id = ~(aq->rq_mask.band_prof_id);
+   }
+
+   rc = mbox_process(mbox);
+   if (rc)
+   goto exit;
+
+   rq->bpf_id = id;
+
+exit:
+   return rc;
+}
diff --git a/drivers/common/cnxk/version.map b/drivers/common/cnxk/version.map
index f34fd3b4a4..05b19beaa2 100644
--- a/drivers/common/cnxk/version.map
+++ b/drivers/common/cnxk/version.map
@@ -85,6 +85,7 @@ INTERNAL {
roc_nix_bpf_alloc;
roc_nix_bpf_config;
roc_nix_bpf_count_get;
+   roc_nix_bpf_ena_dis;
roc_nix_bpf_free;
roc_nix_bpf_free_all;
roc_nix_bpf_level_to_idx;
-- 
2.25.1

[dpdk-dev] [PATCH v4 08/28] common/cnxk: support RoC API to dump bandwidth profile

[skori]

From: Sunil Kumar Kori 

Implement RoC API to dump bandwidth profile on CN10K
platform.

Signed-off-by: Sunil Kumar Kori 
---
v4:
 - Rebase support on dpdk-next-net-mrvl branch
 - Handled meter action during flow destroy
 - Handled meter cleanup during port shutdown
 
v3:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for tree hierarchy
 - Fix naming convention
v2:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for linear hierarchy
 - Review comments incorporated

 drivers/common/cnxk/roc_nix.h  |  3 ++
 drivers/common/cnxk/roc_nix_bpf.c  | 86 ++
 drivers/common/cnxk/roc_platform.h |  1 +
 drivers/common/cnxk/version.map|  1 +
 4 files changed, 91 insertions(+)

diff --git a/drivers/common/cnxk/roc_nix.h b/drivers/common/cnxk/roc_nix.h
index 5b3bf38a82..d7e8dd94e3 100644
--- a/drivers/common/cnxk/roc_nix.h
+++ b/drivers/common/cnxk/roc_nix.h
@@ -606,6 +606,9 @@ int __roc_api roc_nix_bpf_config(struct roc_nix *roc_nix, 
uint16_t id,
 int __roc_api roc_nix_bpf_ena_dis(struct roc_nix *roc_nix, uint16_t id,
  struct roc_nix_rq *rq, bool enable);
 
+int __roc_api roc_nix_bpf_dump(struct roc_nix *roc_nix, uint16_t id,
+  enum roc_nix_bpf_level_flag lvl_flag);
+
 uint8_t __roc_api
 roc_nix_bpf_level_to_idx(enum roc_nix_bpf_level_flag lvl_flag);
 
diff --git a/drivers/common/cnxk/roc_nix_bpf.c 
b/drivers/common/cnxk/roc_nix_bpf.c
index c61a697f1a..8181753220 100644
--- a/drivers/common/cnxk/roc_nix_bpf.c
+++ b/drivers/common/cnxk/roc_nix_bpf.c
@@ -125,6 +125,60 @@ meter_burst_to_nix(uint64_t value, uint64_t *exponent_p, 
uint64_t *mantissa_p)
return NIX_BPF_BURST(exponent, mantissa);
 }
 
+static inline void
+nix_lf_bpf_dump(__io struct nix_band_prof_s *bpf)
+{
+   plt_dump("W0: cir_mantissa  \t\t\t%d\nW0: pebs_mantissa \t\t\t0x%03x",
+bpf->cir_mantissa, bpf->pebs_mantissa);
+   plt_dump("W0: peir_matissa \t\t\t\t%d\nW0: cbs_exponent \t\t\t%d",
+bpf->peir_mantissa, bpf->cbs_exponent);
+   plt_dump("W0: cir_exponent \t\t\t%d\nW0: pebs_exponent \t\t\t%d",
+bpf->cir_exponent, bpf->pebs_exponent);
+   plt_dump("W0: peir_exponent \t\t\t%d\n", bpf->peir_exponent);
+   plt_dump("W0: tnl_ena \t\t\t%d\n", bpf->tnl_ena);
+   plt_dump("W0: icolor \t\t\t%d\n", bpf->icolor);
+   plt_dump("W0: pc_mode \t\t\t%d\n", bpf->pc_mode);
+   plt_dump("W1: hl_en \t\t%d\nW1: band_prof_id \t\t%d", bpf->hl_en,
+bpf->band_prof_id);
+   plt_dump("W1: meter_algo \t\t%d\nW1: rc_action \t\t%d", bpf->meter_algo,
+bpf->rc_action);
+   plt_dump("W1: yc_action \t\t\t%d\nW1: gc_action \t\t\t%d",
+bpf->yc_action, bpf->gc_action);
+   plt_dump("W1: adjust_mantissa\t\t\t%d\nW1: adjust_exponent \t\t\t%d",
+bpf->adjust_mantissa, bpf->adjust_exponent);
+   plt_dump("W1: rdiv \t\t\t%d\n", bpf->rdiv);
+   plt_dump("W1: l_select \t\t%d\nW2: lmode \t\t%d", bpf->l_sellect,
+bpf->lmode);
+   plt_dump("W1: cbs_mantissa \t\t\t%d\n", bpf->cbs_mantissa);
+   plt_dump("W2: tsa \t\t\t0x%" PRIx64 "\n", (uint64_t)bpf->ts);
+   plt_dump("W3: c_accum \t\t%d\nW3: pe_accum \t\t%d", bpf->c_accum,
+bpf->pe_accum);
+   plt_dump("W4: green_pkt_pass \t\t\t0x%" PRIx64 "",
+(uint64_t)bpf->green_pkt_pass);
+   plt_dump("W5: yellow_pkt_pass \t\t\t0x%" PRIx64 "",
+(uint64_t)bpf->yellow_pkt_pass);
+   plt_dump("W6: red_pkt_pass \t\t\t0x%" PRIx64 "",
+(uint64_t)bpf->red_pkt_pass);
+   plt_dump("W7: green_octs_pass \t\t\t0x%" PRIx64 "",
+(uint64_t)bpf->green_octs_pass);
+   plt_dump("W8: yellow_octs_pass \t\t\t0x%" PRIx64 "",
+(uint64_t)bpf->yellow_octs_pass);
+   plt_dump("W9: red_octs_pass \t\t\t0x%" PRIx64 "",
+(uint64_t)bpf->red_octs_pass);
+   plt_dump("W10: green_pkt_drop \t\t\t0x%" PRIx64 "",
+(uint64_t)bpf->green_pkt_drop);
+   plt_dump("W11: yellow_pkt_drop \t\t\t0x%" PRIx64 "",
+(uint64_t)bpf->yellow_pkt_drop);
+   plt_dump("W12: red_pkt_drop \t\t\t0x%" PRIx64 "",
+(uint64_t)bpf->red_pkt_drop);
+   plt_dump("W13: green_octs_drop \t\t\t0x%" PRIx64 "",
+(uint64_t)bpf->green_octs_drop);
+   plt_dump("W14: yellow_octs_drop \t\t\t0x%" PRIx64 "",
+(uint64_t)bpf->yellow_octs_drop);
+   plt_dump("W15: red_octs_drop \t\t\t0x%" PRIx64 "",
+(uint64_t)bpf->red_octs_drop);
+}
+
 uint8_t
 roc_nix_bpf_level_to_idx(enum roc_nix_bpf_level_flag level_f)
 {
@@ -488,3 +542,35 @@ roc_nix_bpf_ena_dis(struct roc_nix *roc_nix, uint16_t id, 
struct roc_nix_rq *rq,
 exit:
return rc;
 }
+
+int
+roc_nix_bpf_dump(struct roc_nix *roc_nix, uint16_t id,
+enum roc_nix_bpf_level_flag lvl_flag)
+{
+   struct 

[dpdk-dev] [PATCH v4 09/28] common/cnxk: support RoC API to setup precolor table

[skori]

From: Sunil Kumar Kori 

For initial coloring of input packet, CN10K platform maintains
precolor table for VLAN, DSCP and Generic. Implement RoC
interface to setup pre color table.

Signed-off-by: Sunil Kumar Kori 
---
v4:
 - Rebase support on dpdk-next-net-mrvl branch
 - Handled meter action during flow destroy
 - Handled meter cleanup during port shutdown
 
v3:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for tree hierarchy
 - Fix naming convention
v2:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for linear hierarchy
 - Review comments incorporated

 drivers/common/cnxk/roc_nix.h |  20 
 drivers/common/cnxk/roc_nix_bpf.c | 193 ++
 drivers/common/cnxk/version.map   |   1 +
 3 files changed, 214 insertions(+)

diff --git a/drivers/common/cnxk/roc_nix.h b/drivers/common/cnxk/roc_nix.h
index d7e8dd94e3..d125941c5c 100644
--- a/drivers/common/cnxk/roc_nix.h
+++ b/drivers/common/cnxk/roc_nix.h
@@ -39,6 +39,15 @@ enum roc_nix_bpf_level_flag {
ROC_NIX_BPF_LEVEL_F_TOP = BIT(2),
 };
 
+enum roc_nix_bpf_pc_mode {
+   ROC_NIX_BPF_PC_MODE_VLAN_INNER,
+   ROC_NIX_BPF_PC_MODE_VLAN_OUTER,
+   ROC_NIX_BPF_PC_MODE_DSCP_INNER,
+   ROC_NIX_BPF_PC_MODE_DSCP_OUTER,
+   ROC_NIX_BPF_PC_MODE_GEN_INNER,
+   ROC_NIX_BPF_PC_MODE_GEN_OUTER
+};
+
 enum roc_nix_bpf_color {
ROC_NIX_BPF_COLOR_GREEN,
ROC_NIX_BPF_COLOR_YELLOW,
@@ -104,6 +113,13 @@ struct roc_nix_bpf_objs {
uint16_t ids[ROC_NIX_BPF_PER_PFFUNC];
 };
 
+struct roc_nix_bpf_precolor {
+#define ROC_NIX_BPF_PRE_COLOR_MAX 64
+   uint8_t count;
+   enum roc_nix_bpf_pc_mode mode;
+   enum roc_nix_bpf_color color[ROC_NIX_BPF_PRE_COLOR_MAX];
+};
+
 struct roc_nix_vlan_config {
uint32_t type;
union {
@@ -609,6 +625,10 @@ int __roc_api roc_nix_bpf_ena_dis(struct roc_nix *roc_nix, 
uint16_t id,
 int __roc_api roc_nix_bpf_dump(struct roc_nix *roc_nix, uint16_t id,
   enum roc_nix_bpf_level_flag lvl_flag);
 
+int __roc_api roc_nix_bpf_pre_color_tbl_setup(
+   struct roc_nix *roc_nix, uint16_t id,
+   enum roc_nix_bpf_level_flag lvl_flag, struct roc_nix_bpf_precolor *tbl);
+
 uint8_t __roc_api
 roc_nix_bpf_level_to_idx(enum roc_nix_bpf_level_flag lvl_flag);
 
diff --git a/drivers/common/cnxk/roc_nix_bpf.c 
b/drivers/common/cnxk/roc_nix_bpf.c
index 8181753220..8c84bbb656 100644
--- a/drivers/common/cnxk/roc_nix_bpf.c
+++ b/drivers/common/cnxk/roc_nix_bpf.c
@@ -9,6 +9,10 @@
 #define NIX_MAX_BPF_COUNT_MID_LAYER  8
 #define NIX_MAX_BPF_COUNT_TOP_LAYER  1
 
+#define NIX_BPF_PRECOLOR_GEN_TABLE_SIZE 16
+#define NIX_BPF_PRECOLOR_VLAN_TABLE_SIZE 16
+#define NIX_BPF_PRECOLOR_DSCP_TABLE_SIZE 64
+
 #define NIX_BPF_LEVEL_F_MASK   
\
(ROC_NIX_BPF_LEVEL_F_LEAF | ROC_NIX_BPF_LEVEL_F_MID |  \
 ROC_NIX_BPF_LEVEL_F_TOP)
@@ -179,6 +183,107 @@ nix_lf_bpf_dump(__io struct nix_band_prof_s *bpf)
 (uint64_t)bpf->red_octs_drop);
 }
 
+static inline void
+nix_precolor_conv_table_write(struct roc_nix *roc_nix, uint64_t val,
+ uint32_t off)
+{
+   struct nix *nix = roc_nix_to_nix_priv(roc_nix);
+   int64_t *addr;
+
+   addr = PLT_PTR_ADD(nix->base, off);
+   /* FIXME: Currently writing to this register throwing kernel dump.
+* plt_write64(val, addr);
+*/
+   PLT_SET_USED(val);
+   PLT_SET_USED(addr);
+}
+
+static uint8_t
+nix_precolor_vlan_table_update(struct roc_nix *roc_nix,
+  struct roc_nix_bpf_precolor *tbl)
+{
+   uint64_t val = 0, i;
+   uint8_t tn_ena;
+   uint32_t off;
+
+   for (i = 0; i < tbl->count; i++)
+   val |= (((uint64_t)tbl->color[i]) << (2 * i));
+
+   if (tbl->mode == ROC_NIX_BPF_PC_MODE_VLAN_INNER) {
+   off = NIX_LF_RX_VLAN1_COLOR_CONV;
+   tn_ena = true;
+   } else {
+   off = NIX_LF_RX_VLAN0_COLOR_CONV;
+   tn_ena = false;
+   }
+
+   nix_precolor_conv_table_write(roc_nix, val, off);
+   return tn_ena;
+}
+
+static uint8_t
+nix_precolor_inner_dscp_table_update(struct roc_nix *roc_nix,
+struct roc_nix_bpf_precolor *tbl)
+{
+   uint64_t val_lo = 0, val_hi = 0, i, j;
+
+   for (i = 0, j = 0; i < (tbl->count / 2); i++, j++)
+   val_lo |= (((uint64_t)tbl->color[i]) << (2 * j));
+
+   for (j = 0; i < tbl->count; i++, j++)
+   val_hi |= (((uint64_t)tbl->color[i]) << (2 * j));
+
+   nix_precolor_conv_table_write(roc_nix, val_lo,
+ NIX_LF_RX_IIP_COLOR_CONV_LO);
+   nix_precolor_conv_table_write(roc_nix, val_hi,
+ NIX_LF_RX_IIP_COLOR_CONV_HI);
+
+   return true;
+}
+
+static uint8_t
+nix_precolor_outer_dscp_table_update(struct roc_nix *roc_nix,
+

[dpdk-dev] [PATCH v4 10/28] common/cnxk: support RoC API to connect bandwidth profiles

[skori]

From: Sunil Kumar Kori 

To maintain chain of bandwidth profiles, they needs to be
connected. Implement RoC API to connect two bandwidth profiles
at different levels.

Signed-off-by: Sunil Kumar Kori 
---
v4:
 - Rebase support on dpdk-next-net-mrvl branch
 - Handled meter action during flow destroy
 - Handled meter cleanup during port shutdown
 
v3:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for tree hierarchy
 - Fix naming convention
v2:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for linear hierarchy
 - Review comments incorporated

 drivers/common/cnxk/roc_nix.h |  6 ++
 drivers/common/cnxk/roc_nix_bpf.c | 36 +++
 drivers/common/cnxk/version.map   |  1 +
 3 files changed, 43 insertions(+)

diff --git a/drivers/common/cnxk/roc_nix.h b/drivers/common/cnxk/roc_nix.h
index d125941c5c..be24d20462 100644
--- a/drivers/common/cnxk/roc_nix.h
+++ b/drivers/common/cnxk/roc_nix.h
@@ -7,6 +7,7 @@
 
 /* Constants */
 #define ROC_NIX_BPF_PER_PFFUNC   64
+#define ROC_NIX_BPF_ID_INVALID   0x
 #define ROC_NIX_BPF_LEVEL_IDX_INVALID 0xFF
 #define ROC_NIX_BPF_LEVEL_MAX3
 
@@ -629,6 +630,11 @@ int __roc_api roc_nix_bpf_pre_color_tbl_setup(
struct roc_nix *roc_nix, uint16_t id,
enum roc_nix_bpf_level_flag lvl_flag, struct roc_nix_bpf_precolor *tbl);
 
+/* Use ROC_NIX_BPF_ID_INVALID as dst_id to disconnect */
+int __roc_api roc_nix_bpf_connect(struct roc_nix *roc_nix,
+ enum roc_nix_bpf_level_flag lvl_flag,
+ uint16_t src_id, uint16_t dst_id);
+
 uint8_t __roc_api
 roc_nix_bpf_level_to_idx(enum roc_nix_bpf_level_flag lvl_flag);
 
diff --git a/drivers/common/cnxk/roc_nix_bpf.c 
b/drivers/common/cnxk/roc_nix_bpf.c
index 8c84bbb656..807723f7b7 100644
--- a/drivers/common/cnxk/roc_nix_bpf.c
+++ b/drivers/common/cnxk/roc_nix_bpf.c
@@ -767,3 +767,39 @@ roc_nix_bpf_pre_color_tbl_setup(struct roc_nix *roc_nix, 
uint16_t id,
 exit:
return rc;
 }
+
+int
+roc_nix_bpf_connect(struct roc_nix *roc_nix,
+   enum roc_nix_bpf_level_flag lvl_flag, uint16_t src_id,
+   uint16_t dst_id)
+{
+   struct mbox *mbox = get_mbox(roc_nix);
+   struct nix_cn10k_aq_enq_req *aq;
+   uint8_t level_idx;
+
+   if (roc_model_is_cn9k())
+   return NIX_ERR_HW_NOTSUP;
+
+   level_idx = roc_nix_bpf_level_to_idx(lvl_flag);
+   if (level_idx == ROC_NIX_BPF_LEVEL_IDX_INVALID)
+   return NIX_ERR_PARAM;
+
+   aq = mbox_alloc_msg_nix_cn10k_aq_enq(mbox);
+   if (aq == NULL)
+   return -ENOSPC;
+   aq->qidx = (sw_to_hw_lvl_map[level_idx] << 14) | src_id;
+   aq->ctype = NIX_AQ_CTYPE_BAND_PROF;
+   aq->op = NIX_AQ_INSTOP_WRITE;
+
+   if (dst_id == ROC_NIX_BPF_ID_INVALID) {
+   aq->prof.hl_en = false;
+   aq->prof_mask.hl_en = ~(aq->prof_mask.hl_en);
+   } else {
+   aq->prof.hl_en = true;
+   aq->prof.band_prof_id = dst_id;
+   aq->prof_mask.hl_en = ~(aq->prof_mask.hl_en);
+   aq->prof_mask.band_prof_id = ~(aq->prof_mask.band_prof_id);
+   }
+
+   return mbox_process(mbox);
+}
diff --git a/drivers/common/cnxk/version.map b/drivers/common/cnxk/version.map
index 836cbec4dc..dfd2cdc287 100644
--- a/drivers/common/cnxk/version.map
+++ b/drivers/common/cnxk/version.map
@@ -84,6 +84,7 @@ INTERNAL {
roc_se_ciph_key_set;
roc_nix_bpf_alloc;
roc_nix_bpf_config;
+   roc_nix_bpf_connect;
roc_nix_bpf_count_get;
roc_nix_bpf_dump;
roc_nix_bpf_ena_dis;
-- 
2.25.1

[dpdk-dev] [PATCH v4 11/28] common/cnxk: support RoC API to get stats to index

[skori]

From: Sunil Kumar Kori 

CN10K platform supports different stats for HW bandwidth profiles.
Implement RoC API to get index for given stats type.

Signed-off-by: Sunil Kumar Kori 
---
v4:
 - Rebase support on dpdk-next-net-mrvl branch
 - Handled meter action during flow destroy
 - Handled meter cleanup during port shutdown
 
v3:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for tree hierarchy
 - Fix naming convention
v2:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for linear hierarchy
 - Review comments incorporated

 drivers/common/cnxk/roc_nix.h | 18 
 drivers/common/cnxk/roc_nix_bpf.c | 34 +++
 drivers/common/cnxk/version.map   |  1 +
 3 files changed, 53 insertions(+)

diff --git a/drivers/common/cnxk/roc_nix.h b/drivers/common/cnxk/roc_nix.h
index be24d20462..c6cf42e8f4 100644
--- a/drivers/common/cnxk/roc_nix.h
+++ b/drivers/common/cnxk/roc_nix.h
@@ -10,6 +10,7 @@
 #define ROC_NIX_BPF_ID_INVALID   0x
 #define ROC_NIX_BPF_LEVEL_IDX_INVALID 0xFF
 #define ROC_NIX_BPF_LEVEL_MAX3
+#define ROC_NIX_BPF_STATS_MAX12
 
 enum roc_nix_rss_reta_sz {
ROC_NIX_RSS_RETA_SZ_64 = 64,
@@ -71,6 +72,21 @@ enum roc_nix_bpf_action {
ROC_NIX_BPF_ACTION_RED
 };
 
+enum roc_nix_bpf_stats {
+   ROC_NIX_BPF_GREEN_PKT_F_PASS = BIT_ULL(0),
+   ROC_NIX_BPF_GREEN_OCTS_F_PASS = BIT_ULL(1),
+   ROC_NIX_BPF_GREEN_PKT_F_DROP = BIT_ULL(2),
+   ROC_NIX_BPF_GREEN_OCTS_F_DROP = BIT_ULL(3),
+   ROC_NIX_BPF_YELLOW_PKT_F_PASS = BIT_ULL(4),
+   ROC_NIX_BPF_YELLOW_OCTS_F_PASS = BIT_ULL(5),
+   ROC_NIX_BPF_YELLOW_PKT_F_DROP = BIT_ULL(6),
+   ROC_NIX_BPF_YELLOW_OCTS_F_DROP = BIT_ULL(7),
+   ROC_NIX_BPF_RED_PKT_F_PASS = BIT_ULL(8),
+   ROC_NIX_BPF_RED_OCTS_F_PASS = BIT_ULL(9),
+   ROC_NIX_BPF_RED_PKT_F_DROP = BIT_ULL(10),
+   ROC_NIX_BPF_RED_OCTS_F_DROP = BIT_ULL(11),
+};
+
 struct roc_nix_bpf_cfg {
enum roc_nix_bpf_algo alg;
enum roc_nix_bpf_lmode lmode;
@@ -638,6 +654,8 @@ int __roc_api roc_nix_bpf_connect(struct roc_nix *roc_nix,
 uint8_t __roc_api
 roc_nix_bpf_level_to_idx(enum roc_nix_bpf_level_flag lvl_flag);
 
+uint8_t __roc_api roc_nix_bpf_stats_to_idx(enum roc_nix_bpf_stats lvl_flag);
+
 /* MAC */
 int __roc_api roc_nix_mac_rxtx_start_stop(struct roc_nix *roc_nix, bool start);
 int __roc_api roc_nix_mac_link_event_start_stop(struct roc_nix *roc_nix,
diff --git a/drivers/common/cnxk/roc_nix_bpf.c 
b/drivers/common/cnxk/roc_nix_bpf.c
index 807723f7b7..7c1ac9d728 100644
--- a/drivers/common/cnxk/roc_nix_bpf.c
+++ b/drivers/common/cnxk/roc_nix_bpf.c
@@ -300,6 +300,40 @@ roc_nix_bpf_level_to_idx(enum roc_nix_bpf_level_flag 
level_f)
return idx;
 }
 
+uint8_t
+roc_nix_bpf_stats_to_idx(enum roc_nix_bpf_stats level_f)
+{
+   uint8_t idx;
+
+   if (level_f & ROC_NIX_BPF_GREEN_PKT_F_PASS)
+   idx = 0;
+   else if (level_f & ROC_NIX_BPF_GREEN_OCTS_F_PASS)
+   idx = 1;
+   else if (level_f & ROC_NIX_BPF_GREEN_PKT_F_DROP)
+   idx = 2;
+   else if (level_f & ROC_NIX_BPF_GREEN_OCTS_F_DROP)
+   idx = 3;
+   else if (level_f & ROC_NIX_BPF_YELLOW_PKT_F_PASS)
+   idx = 4;
+   else if (level_f & ROC_NIX_BPF_YELLOW_OCTS_F_PASS)
+   idx = 5;
+   else if (level_f & ROC_NIX_BPF_YELLOW_PKT_F_DROP)
+   idx = 6;
+   else if (level_f & ROC_NIX_BPF_YELLOW_OCTS_F_DROP)
+   idx = 7;
+   else if (level_f & ROC_NIX_BPF_RED_PKT_F_PASS)
+   idx = 8;
+   else if (level_f & ROC_NIX_BPF_RED_OCTS_F_PASS)
+   idx = 9;
+   else if (level_f & ROC_NIX_BPF_RED_PKT_F_DROP)
+   idx = 10;
+   else if (level_f & ROC_NIX_BPF_RED_OCTS_F_DROP)
+   idx = 11;
+   else
+   idx = ROC_NIX_BPF_STATS_MAX;
+   return idx;
+}
+
 int
 roc_nix_bpf_count_get(struct roc_nix *roc_nix, uint8_t lvl_mask,
  uint16_t count[ROC_NIX_BPF_LEVEL_MAX])
diff --git a/drivers/common/cnxk/version.map b/drivers/common/cnxk/version.map
index dfd2cdc287..e9643f7469 100644
--- a/drivers/common/cnxk/version.map
+++ b/drivers/common/cnxk/version.map
@@ -92,6 +92,7 @@ INTERNAL {
roc_nix_bpf_free_all;
roc_nix_bpf_level_to_idx;
roc_nix_bpf_pre_color_tbl_setup;
+   roc_nix_bpf_stats_to_idx;
roc_nix_cq_dump;
roc_nix_cq_fini;
roc_nix_cq_init;
-- 
2.25.1

[dpdk-dev] [PATCH v4 12/28] common/cnxk: support RoC API to read profile statistics

[skori]

From: Sunil Kumar Kori 

CN10K platform provides statistics per bandwidth profile and
per nixlf. Implement RoC API to read stats for given bandwidth
profile.

Signed-off-by: Sunil Kumar Kori 
---
v4:
 - Rebase support on dpdk-next-net-mrvl branch
 - Handled meter action during flow destroy
 - Handled meter cleanup during port shutdown
 
v3:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for tree hierarchy
 - Fix naming convention
v2:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for linear hierarchy
 - Review comments incorporated

 drivers/common/cnxk/roc_nix.h |   9 ++
 drivers/common/cnxk/roc_nix_bpf.c | 197 ++
 drivers/common/cnxk/version.map   |   2 +
 3 files changed, 208 insertions(+)

diff --git a/drivers/common/cnxk/roc_nix.h b/drivers/common/cnxk/roc_nix.h
index c6cf42e8f4..0416778c11 100644
--- a/drivers/common/cnxk/roc_nix.h
+++ b/drivers/common/cnxk/roc_nix.h
@@ -651,6 +651,15 @@ int __roc_api roc_nix_bpf_connect(struct roc_nix *roc_nix,
  enum roc_nix_bpf_level_flag lvl_flag,
  uint16_t src_id, uint16_t dst_id);
 
+int __roc_api
+roc_nix_bpf_stats_read(struct roc_nix *roc_nix, uint16_t id, uint64_t mask,
+  enum roc_nix_bpf_level_flag lvl_flag,
+  uint64_t stats[ROC_NIX_BPF_STATS_MAX] /* Out */);
+
+int __roc_api
+roc_nix_bpf_lf_stats_read(struct roc_nix *roc_nix, uint64_t mask,
+ uint64_t stats[ROC_NIX_BPF_STATS_MAX] /* Out */);
+
 uint8_t __roc_api
 roc_nix_bpf_level_to_idx(enum roc_nix_bpf_level_flag lvl_flag);
 
diff --git a/drivers/common/cnxk/roc_nix_bpf.c 
b/drivers/common/cnxk/roc_nix_bpf.c
index 7c1ac9d728..a5e3575404 100644
--- a/drivers/common/cnxk/roc_nix_bpf.c
+++ b/drivers/common/cnxk/roc_nix_bpf.c
@@ -17,6 +17,9 @@
(ROC_NIX_BPF_LEVEL_F_LEAF | ROC_NIX_BPF_LEVEL_F_MID |  \
 ROC_NIX_BPF_LEVEL_F_TOP)
 
+#define NIX_RD_STATS(val)  plt_read64(nix->base + NIX_LF_RX_STATX(val))
+#define NIX_RST_STATS(val) plt_write64(0, nix->base + NIX_LF_RX_STATX(val))
+
 static uint8_t sw_to_hw_lvl_map[] = {NIX_RX_BAND_PROF_LAYER_LEAF,
 NIX_RX_BAND_PROF_LAYER_MIDDLE,
 NIX_RX_BAND_PROF_LAYER_TOP};
@@ -837,3 +840,197 @@ roc_nix_bpf_connect(struct roc_nix *roc_nix,
 
return mbox_process(mbox);
 }
+
+int
+roc_nix_bpf_stats_read(struct roc_nix *roc_nix, uint16_t id, uint64_t mask,
+  enum roc_nix_bpf_level_flag lvl_flag,
+  uint64_t stats[ROC_NIX_BPF_STATS_MAX])
+{
+   uint8_t yellow_pkt_pass, yellow_octs_pass, yellow_pkt_drop;
+   uint8_t green_octs_drop, yellow_octs_drop, red_octs_drop;
+   uint8_t green_pkt_pass, green_octs_pass, green_pkt_drop;
+   uint8_t red_pkt_pass, red_octs_pass, red_pkt_drop;
+   struct mbox *mbox = get_mbox(roc_nix);
+   struct nix_cn10k_aq_enq_rsp *rsp;
+   struct nix_cn10k_aq_enq_req *aq;
+   uint8_t level_idx;
+   int rc;
+
+   if (roc_model_is_cn9k())
+   return NIX_ERR_HW_NOTSUP;
+
+   level_idx = roc_nix_bpf_level_to_idx(lvl_flag);
+   if (level_idx == ROC_NIX_BPF_LEVEL_IDX_INVALID)
+   return NIX_ERR_PARAM;
+
+   aq = mbox_alloc_msg_nix_cn10k_aq_enq(mbox);
+   if (aq == NULL)
+   return -ENOSPC;
+   aq->qidx = (sw_to_hw_lvl_map[level_idx] << 14 | id);
+   aq->ctype = NIX_AQ_CTYPE_BAND_PROF;
+   aq->op = NIX_AQ_INSTOP_READ;
+   rc = mbox_process_msg(mbox, (void *)&rsp);
+   if (rc)
+   return rc;
+
+   green_pkt_pass =
+   roc_nix_bpf_stats_to_idx(mask & ROC_NIX_BPF_GREEN_PKT_F_PASS);
+   green_octs_pass =
+   roc_nix_bpf_stats_to_idx(mask & ROC_NIX_BPF_GREEN_OCTS_F_PASS);
+   green_pkt_drop =
+   roc_nix_bpf_stats_to_idx(mask & ROC_NIX_BPF_GREEN_PKT_F_DROP);
+   green_octs_drop =
+   roc_nix_bpf_stats_to_idx(mask & ROC_NIX_BPF_GREEN_OCTS_F_DROP);
+   yellow_pkt_pass =
+   roc_nix_bpf_stats_to_idx(mask & ROC_NIX_BPF_YELLOW_PKT_F_PASS);
+   yellow_octs_pass =
+   roc_nix_bpf_stats_to_idx(mask & ROC_NIX_BPF_YELLOW_OCTS_F_PASS);
+   yellow_pkt_drop =
+   roc_nix_bpf_stats_to_idx(mask & ROC_NIX_BPF_YELLOW_PKT_F_DROP);
+   yellow_octs_drop =
+   roc_nix_bpf_stats_to_idx(mask & ROC_NIX_BPF_YELLOW_OCTS_F_DROP);
+   red_pkt_pass =
+   roc_nix_bpf_stats_to_idx(mask & ROC_NIX_BPF_RED_PKT_F_PASS);
+   red_octs_pass =
+   roc_nix_bpf_stats_to_idx(mask & ROC_NIX_BPF_RED_OCTS_F_PASS);
+   red_pkt_drop =
+   roc_nix_bpf_stats_to_idx(mask & ROC_NIX_BPF_RED_PKT_F_DROP);
+   red_octs_drop =
+   roc_nix_bpf_stats_to_idx(mask & ROC_NIX_BPF_RED_OCTS_F_DROP);
+
+   if (green_pkt_pass != ROC_NIX_BPF_STATS_MAX)
+   stats[green_pk

[dpdk-dev] [PATCH v4 13/28] common/cnxk: support RoC API to reset profile stats

[skori]

From: Sunil Kumar Kori 

Implement RoC API to reset stats per bandwidth profile
or per nixlf.

Signed-off-by: Sunil Kumar Kori 
---
v4:
 - Rebase support on dpdk-next-net-mrvl branch
 - Handled meter action during flow destroy
 - Handled meter cleanup during port shutdown
 
v3:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for tree hierarchy
 - Fix naming convention
v2:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for linear hierarchy
 - Review comments incorporated

 drivers/common/cnxk/roc_nix.h |   7 ++
 drivers/common/cnxk/roc_nix_bpf.c | 113 ++
 drivers/common/cnxk/version.map   |   2 +
 3 files changed, 122 insertions(+)

diff --git a/drivers/common/cnxk/roc_nix.h b/drivers/common/cnxk/roc_nix.h
index 0416778c11..343bb2f8b7 100644
--- a/drivers/common/cnxk/roc_nix.h
+++ b/drivers/common/cnxk/roc_nix.h
@@ -656,10 +656,17 @@ roc_nix_bpf_stats_read(struct roc_nix *roc_nix, uint16_t 
id, uint64_t mask,
   enum roc_nix_bpf_level_flag lvl_flag,
   uint64_t stats[ROC_NIX_BPF_STATS_MAX] /* Out */);
 
+int __roc_api roc_nix_bpf_stats_reset(struct roc_nix *roc_nix, uint16_t id,
+ uint64_t mask,
+ enum roc_nix_bpf_level_flag lvl_flag);
+
 int __roc_api
 roc_nix_bpf_lf_stats_read(struct roc_nix *roc_nix, uint64_t mask,
  uint64_t stats[ROC_NIX_BPF_STATS_MAX] /* Out */);
 
+int __roc_api roc_nix_bpf_lf_stats_reset(struct roc_nix *roc_nix,
+uint64_t mask);
+
 uint8_t __roc_api
 roc_nix_bpf_level_to_idx(enum roc_nix_bpf_level_flag lvl_flag);
 
diff --git a/drivers/common/cnxk/roc_nix_bpf.c 
b/drivers/common/cnxk/roc_nix_bpf.c
index a5e3575404..df57958683 100644
--- a/drivers/common/cnxk/roc_nix_bpf.c
+++ b/drivers/common/cnxk/roc_nix_bpf.c
@@ -937,6 +937,86 @@ roc_nix_bpf_stats_read(struct roc_nix *roc_nix, uint16_t 
id, uint64_t mask,
return 0;
 }
 
+int
+roc_nix_bpf_stats_reset(struct roc_nix *roc_nix, uint16_t id, uint64_t mask,
+   enum roc_nix_bpf_level_flag lvl_flag)
+{
+   struct mbox *mbox = get_mbox(roc_nix);
+   struct nix_cn10k_aq_enq_req *aq;
+   uint8_t level_idx;
+
+   if (roc_model_is_cn9k())
+   return NIX_ERR_HW_NOTSUP;
+
+   level_idx = roc_nix_bpf_level_to_idx(lvl_flag);
+   if (level_idx == ROC_NIX_BPF_LEVEL_IDX_INVALID)
+   return NIX_ERR_PARAM;
+
+   aq = mbox_alloc_msg_nix_cn10k_aq_enq(mbox);
+   if (aq == NULL)
+   return -ENOSPC;
+   aq->qidx = (sw_to_hw_lvl_map[level_idx] << 14 | id);
+   aq->ctype = NIX_AQ_CTYPE_BAND_PROF;
+   aq->op = NIX_AQ_INSTOP_WRITE;
+
+   if (mask & ROC_NIX_BPF_GREEN_PKT_F_PASS) {
+   aq->prof.green_pkt_pass = 0;
+   aq->prof_mask.green_pkt_pass = ~(aq->prof_mask.green_pkt_pass);
+   }
+   if (mask & ROC_NIX_BPF_GREEN_OCTS_F_PASS) {
+   aq->prof.green_octs_pass = 0;
+   aq->prof_mask.green_octs_pass =
+   ~(aq->prof_mask.green_octs_pass);
+   }
+   if (mask & ROC_NIX_BPF_GREEN_PKT_F_DROP) {
+   aq->prof.green_pkt_drop = 0;
+   aq->prof_mask.green_pkt_drop = ~(aq->prof_mask.green_pkt_drop);
+   }
+   if (mask & ROC_NIX_BPF_GREEN_OCTS_F_DROP) {
+   aq->prof.green_octs_drop = 0;
+   aq->prof_mask.green_octs_drop =
+   ~(aq->prof_mask.green_octs_drop);
+   }
+   if (mask & ROC_NIX_BPF_YELLOW_PKT_F_PASS) {
+   aq->prof.yellow_pkt_pass = 0;
+   aq->prof_mask.yellow_pkt_pass =
+   ~(aq->prof_mask.yellow_pkt_pass);
+   }
+   if (mask & ROC_NIX_BPF_YELLOW_OCTS_F_PASS) {
+   aq->prof.yellow_octs_pass = 0;
+   aq->prof_mask.yellow_octs_pass =
+   ~(aq->prof_mask.yellow_octs_pass);
+   }
+   if (mask & ROC_NIX_BPF_YELLOW_PKT_F_DROP) {
+   aq->prof.yellow_pkt_drop = 0;
+   aq->prof_mask.yellow_pkt_drop =
+   ~(aq->prof_mask.yellow_pkt_drop);
+   }
+   if (mask & ROC_NIX_BPF_YELLOW_OCTS_F_DROP) {
+   aq->prof.yellow_octs_drop = 0;
+   aq->prof_mask.yellow_octs_drop =
+   ~(aq->prof_mask.yellow_octs_drop);
+   }
+   if (mask & ROC_NIX_BPF_RED_PKT_F_PASS) {
+   aq->prof.red_pkt_pass = 0;
+   aq->prof_mask.red_pkt_pass = ~(aq->prof_mask.red_pkt_pass);
+   }
+   if (mask & ROC_NIX_BPF_RED_OCTS_F_PASS) {
+   aq->prof.red_octs_pass = 0;
+   aq->prof_mask.red_octs_pass = ~(aq->prof_mask.red_octs_pass);
+   }
+   if (mask & ROC_NIX_BPF_RED_PKT_F_DROP) {
+   aq->prof.red_pkt_drop = 0;
+   aq->prof_mask.red_pkt_drop = ~(aq->prof_mask.red_pkt_drop);
+   }
+   if (mask & ROC_NI

[dpdk-dev] [PATCH v4 14/28] common/cnxk: support meter in action list

[skori]

From: Sunil Kumar Kori 

Meter action is added in supported action list.

Signed-off-by: Sunil Kumar Kori 
---
v4:
 - Rebase support on dpdk-next-net-mrvl branch
 - Handled meter action during flow destroy
 - Handled meter cleanup during port shutdown
 
v3:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for tree hierarchy
 - Fix naming convention
v2:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for linear hierarchy
 - Review comments incorporated

 drivers/common/cnxk/roc_npc.c | 7 +++
 drivers/common/cnxk/roc_npc.h | 8 +++-
 2 files changed, 14 insertions(+), 1 deletion(-)

diff --git a/drivers/common/cnxk/roc_npc.c b/drivers/common/cnxk/roc_npc.c
index b724ff9401..5a78d9652c 100644
--- a/drivers/common/cnxk/roc_npc.c
+++ b/drivers/common/cnxk/roc_npc.c
@@ -348,6 +348,7 @@ npc_parse_actions(struct roc_npc *roc_npc, const struct 
roc_npc_attr *attr,
 {
struct npc *npc = roc_npc_to_npc_priv(roc_npc);
const struct roc_npc_action_mark *act_mark;
+   const struct roc_npc_action_meter *act_mtr;
const struct roc_npc_action_queue *act_q;
const struct roc_npc_action_vf *vf_act;
bool vlan_insert_action = false;
@@ -451,6 +452,12 @@ npc_parse_actions(struct roc_npc *roc_npc, const struct 
roc_npc_attr *attr,
case ROC_NPC_ACTION_TYPE_VLAN_PCP_INSERT:
req_act |= ROC_NPC_ACTION_TYPE_VLAN_PCP_INSERT;
break;
+   case ROC_NPC_ACTION_TYPE_METER:
+   act_mtr = (const struct roc_npc_action_meter *)
+ actions->conf;
+   flow->mtr_id = act_mtr->mtr_id;
+   req_act |= ROC_NPC_ACTION_TYPE_METER;
+   break;
default:
errcode = NPC_ERR_ACTION_NOTSUP;
goto err_exit;
diff --git a/drivers/common/cnxk/roc_npc.h b/drivers/common/cnxk/roc_npc.h
index 65d4bd6edc..10d1ac82a4 100644
--- a/drivers/common/cnxk/roc_npc.h
+++ b/drivers/common/cnxk/roc_npc.h
@@ -58,7 +58,7 @@ struct roc_npc_flow_item_raw {
const uint8_t *pattern; /**< Byte string to look for. */
 };
 
-#define ROC_NPC_MAX_ACTION_COUNT 12
+#define ROC_NPC_MAX_ACTION_COUNT 19
 
 enum roc_npc_action_type {
ROC_NPC_ACTION_TYPE_END = (1 << 0),
@@ -77,6 +77,7 @@ enum roc_npc_action_type {
ROC_NPC_ACTION_TYPE_VLAN_INSERT = (1 << 13),
ROC_NPC_ACTION_TYPE_VLAN_ETHTYPE_INSERT = (1 << 14),
ROC_NPC_ACTION_TYPE_VLAN_PCP_INSERT = (1 << 15),
+   ROC_NPC_ACTION_TYPE_METER = (1 << 17),
 };
 
 struct roc_npc_action {
@@ -110,6 +111,10 @@ struct roc_npc_action_of_set_vlan_pcp {
uint8_t vlan_pcp; /**< VLAN priority. */
 };
 
+struct roc_npc_action_meter {
+   uint32_t mtr_id; /**< Meter id to be applied. > */
+};
+
 struct roc_npc_attr {
uint32_t priority;  /**< Rule priority level within group. */
uint32_t ingress : 1;   /**< Rule applies to ingress traffic. */
@@ -128,6 +133,7 @@ struct roc_npc_flow {
uint32_t mcam_id;
int32_t ctr_id;
uint32_t priority;
+   uint32_t mtr_id;
 #define ROC_NPC_MAX_MCAM_WIDTH_DWORDS 7
/* Contiguous match string */
uint64_t mcam_data[ROC_NPC_MAX_MCAM_WIDTH_DWORDS];
-- 
2.25.1

[dpdk-dev] [PATCH v4 15/28] net/cnxk: support meter ops get API

[skori]

From: Sunil Kumar Kori 

To enable support for ingress meter, supported operations
are exposed for CNXK platform.

Signed-off-by: Sunil Kumar Kori 
Signed-off-by: Rakesh Kudurumalla 
---
v4:
 - Rebase support on dpdk-next-net-mrvl branch
 - Handled meter action during flow destroy
 - Handled meter cleanup during port shutdown
 
v3:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for tree hierarchy
 - Fix naming convention
v2:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for linear hierarchy
 - Review comments incorporated

 drivers/net/cnxk/cnxk_ethdev.c |  1 +
 drivers/net/cnxk/cnxk_ethdev.h |  3 +++
 drivers/net/cnxk/cnxk_ethdev_mtr.c | 18 ++
 drivers/net/cnxk/meson.build   |  1 +
 4 files changed, 23 insertions(+)
 create mode 100644 drivers/net/cnxk/cnxk_ethdev_mtr.c

diff --git a/drivers/net/cnxk/cnxk_ethdev.c b/drivers/net/cnxk/cnxk_ethdev.c
index ec00e620eb..f694abd71d 100644
--- a/drivers/net/cnxk/cnxk_ethdev.c
+++ b/drivers/net/cnxk/cnxk_ethdev.c
@@ -1498,6 +1498,7 @@ struct eth_dev_ops cnxk_eth_dev_ops = {
.set_mc_addr_list = cnxk_nix_mc_addr_list_configure,
.set_queue_rate_limit = cnxk_nix_tm_set_queue_rate_limit,
.tm_ops_get = cnxk_nix_tm_ops_get,
+   .mtr_ops_get = cnxk_nix_mtr_ops_get,
 };
 
 static int
diff --git a/drivers/net/cnxk/cnxk_ethdev.h b/drivers/net/cnxk/cnxk_ethdev.h
index ff21b977b7..7a585b9dcf 100644
--- a/drivers/net/cnxk/cnxk_ethdev.h
+++ b/drivers/net/cnxk/cnxk_ethdev.h
@@ -433,6 +433,9 @@ int cnxk_nix_tm_ops_get(struct rte_eth_dev *eth_dev, void 
*ops);
 int cnxk_nix_tm_set_queue_rate_limit(struct rte_eth_dev *eth_dev,
 uint16_t queue_idx, uint16_t tx_rate);
 
+/* MTR */
+int cnxk_nix_mtr_ops_get(struct rte_eth_dev *dev, void *ops);
+
 /* RSS */
 uint32_t cnxk_rss_ethdev_to_nix(struct cnxk_eth_dev *dev, uint64_t ethdev_rss,
uint8_t rss_level);
diff --git a/drivers/net/cnxk/cnxk_ethdev_mtr.c 
b/drivers/net/cnxk/cnxk_ethdev_mtr.c
new file mode 100644
index 00..fdb493a4b9
--- /dev/null
+++ b/drivers/net/cnxk/cnxk_ethdev_mtr.c
@@ -0,0 +1,18 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(C) 2021 Marvell.
+ */
+
+#include "cnxk_ethdev.h"
+#include 
+
+const struct rte_mtr_ops nix_mtr_ops = {
+};
+
+int
+cnxk_nix_mtr_ops_get(struct rte_eth_dev *dev, void *ops)
+{
+   RTE_SET_USED(dev);
+
+   *(const void **)ops = &nix_mtr_ops;
+   return 0;
+}
diff --git a/drivers/net/cnxk/meson.build b/drivers/net/cnxk/meson.build
index d86188fed7..5dac078b2c 100644
--- a/drivers/net/cnxk/meson.build
+++ b/drivers/net/cnxk/meson.build
@@ -11,6 +11,7 @@ endif
 sources = files(
 'cnxk_ethdev.c',
 'cnxk_ethdev_devargs.c',
+'cnxk_ethdev_mtr.c',
 'cnxk_ethdev_ops.c',
 'cnxk_ethdev_sec.c',
 'cnxk_link.c',
-- 
2.25.1

[dpdk-dev] [PATCH v4 16/28] net/cnxk: support ops to get meter capabilities

[skori]

From: Sunil Kumar Kori 

Implement ethdev operation to get meter capabilities for
CNXK platform.

Signed-off-by: Sunil Kumar Kori 
Signed-off-by: Rakesh Kudurumalla 
---
v4:
 - Rebase support on dpdk-next-net-mrvl branch
 - Handled meter action during flow destroy
 - Handled meter cleanup during port shutdown
 
v3:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for tree hierarchy
 - Fix naming convention
v2:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for linear hierarchy
 - Review comments incorporated

 drivers/net/cnxk/cnxk_ethdev_mtr.c | 48 ++
 1 file changed, 48 insertions(+)

diff --git a/drivers/net/cnxk/cnxk_ethdev_mtr.c 
b/drivers/net/cnxk/cnxk_ethdev_mtr.c
index fdb493a4b9..12013f26c2 100644
--- a/drivers/net/cnxk/cnxk_ethdev_mtr.c
+++ b/drivers/net/cnxk/cnxk_ethdev_mtr.c
@@ -5,7 +5,55 @@
 #include "cnxk_ethdev.h"
 #include 
 
+#define NIX_MTR_COUNT_MAX  73 /* 64(leaf) + 8(mid) + 1(top) */
+#define NIX_MTR_COUNT_PER_FLOW 3  /* 1(leaf) + 1(mid) + 1(top) */
+
+static struct rte_mtr_capabilities mtr_capa = {
+   .n_max = NIX_MTR_COUNT_MAX,
+   .n_shared_max = NIX_MTR_COUNT_PER_FLOW,
+   /* .identical = , */
+   .shared_identical = true,
+   /* .shared_n_flows_per_mtr_max = ,*/
+   .chaining_n_mtrs_per_flow_max = NIX_MTR_COUNT_PER_FLOW,
+   .chaining_use_prev_mtr_color_supported = true,
+   .chaining_use_prev_mtr_color_enforced = true,
+   .meter_srtcm_rfc2697_n_max = NIX_MTR_COUNT_MAX,
+   .meter_trtcm_rfc2698_n_max = NIX_MTR_COUNT_MAX,
+   .meter_trtcm_rfc4115_n_max = NIX_MTR_COUNT_MAX,
+   .meter_rate_max = NIX_BPF_RATE_MAX / 8, /* Bytes per second */
+   .meter_policy_n_max = NIX_MTR_COUNT_MAX,
+   .color_aware_srtcm_rfc2697_supported = true,
+   .color_aware_trtcm_rfc2698_supported = true,
+   .color_aware_trtcm_rfc4115_supported = true,
+   .srtcm_rfc2697_byte_mode_supported = true,
+   .srtcm_rfc2697_packet_mode_supported = true,
+   .trtcm_rfc2698_byte_mode_supported = true,
+   .trtcm_rfc2698_packet_mode_supported = true,
+   .trtcm_rfc4115_byte_mode_supported = true,
+   .trtcm_rfc4115_packet_mode_supported = true,
+   .stats_mask = RTE_MTR_STATS_N_PKTS_GREEN | RTE_MTR_STATS_N_PKTS_YELLOW |
+ RTE_MTR_STATS_N_PKTS_RED | RTE_MTR_STATS_N_PKTS_DROPPED |
+ RTE_MTR_STATS_N_BYTES_GREEN |
+ RTE_MTR_STATS_N_BYTES_YELLOW | RTE_MTR_STATS_N_BYTES_RED |
+ RTE_MTR_STATS_N_BYTES_DROPPED};
+
+static int
+cnxk_nix_mtr_capabilities_get(struct rte_eth_dev *dev,
+  struct rte_mtr_capabilities *capa,
+  struct rte_mtr_error *error)
+{
+   RTE_SET_USED(dev);
+
+   if (!capa)
+   return -rte_mtr_error_set(error, EINVAL,
+ RTE_MTR_ERROR_TYPE_MTR_PARAMS, NULL,
+ "NULL input parameter");
+   *capa = mtr_capa;
+   return 0;
+}
+
 const struct rte_mtr_ops nix_mtr_ops = {
+   .capabilities_get = cnxk_nix_mtr_capabilities_get,
 };
 
 int
-- 
2.25.1

[dpdk-dev] [PATCH v4 17/28] net/cnxk: support ops to create meter profile

[skori]

From: Sunil Kumar Kori 

Implement API to add meter profile for CNXK platform.

Signed-off-by: Sunil Kumar Kori 
Signed-off-by: Rakesh Kudurumalla 
---
v4:
 - Rebase support on dpdk-next-net-mrvl branch
 - Handled meter action during flow destroy
 - Handled meter cleanup during port shutdown
 
v3:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for tree hierarchy
 - Fix naming convention
v2:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for linear hierarchy
 - Review comments incorporated

 drivers/net/cnxk/cnxk_ethdev.c |  14 +++
 drivers/net/cnxk/cnxk_ethdev.h |  13 +++
 drivers/net/cnxk/cnxk_ethdev_mtr.c | 138 +
 3 files changed, 165 insertions(+)

diff --git a/drivers/net/cnxk/cnxk_ethdev.c b/drivers/net/cnxk/cnxk_ethdev.c
index f694abd71d..d748d5da04 100644
--- a/drivers/net/cnxk/cnxk_ethdev.c
+++ b/drivers/net/cnxk/cnxk_ethdev.c
@@ -701,6 +701,14 @@ nix_free_queue_mem(struct cnxk_eth_dev *dev)
dev->sqs = NULL;
 }
 
+static int
+nix_ingress_policer_setup(struct cnxk_eth_dev *dev)
+{
+   TAILQ_INIT(&dev->mtr_profiles);
+
+   return 0;
+}
+
 static int
 nix_rss_default_setup(struct cnxk_eth_dev *dev)
 {
@@ -1113,6 +1121,12 @@ cnxk_nix_configure(struct rte_eth_dev *eth_dev)
goto free_nix_lf;
}
 
+   rc = nix_ingress_policer_setup(dev);
+   if (rc) {
+   plt_err("Failed to setup ingress policer rc=%d", rc);
+   goto free_nix_lf;
+   }
+
rc = roc_nix_tm_hierarchy_enable(nix, ROC_NIX_TM_DEFAULT, false);
if (rc) {
plt_err("Failed to enable default tm hierarchy, rc=%d", rc);
diff --git a/drivers/net/cnxk/cnxk_ethdev.h b/drivers/net/cnxk/cnxk_ethdev.h
index 7a585b9dcf..c076208d77 100644
--- a/drivers/net/cnxk/cnxk_ethdev.h
+++ b/drivers/net/cnxk/cnxk_ethdev.h
@@ -13,6 +13,7 @@
 #include 
 #include 
 #include 
+#include 
 #include 
 #include 
 #include 
@@ -158,6 +159,15 @@ struct cnxk_timesync_info {
uint64_t *tx_tstamp;
 } __plt_cache_aligned;
 
+struct cnxk_mtr_profile_node {
+   TAILQ_ENTRY(cnxk_mtr_profile_node) next;
+   struct rte_mtr_meter_profile profile; /**< Profile detail. */
+   uint32_t ref_cnt; /**< Use count. */
+   uint32_t id;  /**< Profile id. */
+};
+
+TAILQ_HEAD(cnxk_mtr_profiles, cnxk_mtr_profile_node);
+
 /* Security session private data */
 struct cnxk_eth_sec_sess {
/* List entry */
@@ -303,6 +313,9 @@ struct cnxk_eth_dev {
double clk_freq_mult;
uint64_t clk_delta;
 
+   /* Ingress policer */
+   struct cnxk_mtr_profiles mtr_profiles;
+
/* Rx burst for cleanup(Only Primary) */
eth_rx_burst_t rx_pkt_burst_no_offload;
 
diff --git a/drivers/net/cnxk/cnxk_ethdev_mtr.c 
b/drivers/net/cnxk/cnxk_ethdev_mtr.c
index 12013f26c2..bab03ebe95 100644
--- a/drivers/net/cnxk/cnxk_ethdev_mtr.c
+++ b/drivers/net/cnxk/cnxk_ethdev_mtr.c
@@ -37,6 +37,106 @@ static struct rte_mtr_capabilities mtr_capa = {
  RTE_MTR_STATS_N_BYTES_YELLOW | RTE_MTR_STATS_N_BYTES_RED |
  RTE_MTR_STATS_N_BYTES_DROPPED};
 
+static struct cnxk_mtr_profile_node *
+nix_mtr_profile_find(struct cnxk_eth_dev *dev, uint32_t profile_id)
+{
+   struct cnxk_mtr_profiles *fmps = &dev->mtr_profiles;
+   struct cnxk_mtr_profile_node *fmp;
+
+   TAILQ_FOREACH(fmp, fmps, next)
+   if (profile_id == fmp->id)
+   return fmp;
+
+   return NULL;
+}
+
+static int
+nix_mtr_profile_validate(struct cnxk_eth_dev *dev, uint32_t profile_id,
+struct rte_mtr_meter_profile *profile,
+struct rte_mtr_error *error)
+{
+   int rc = 0;
+
+   PLT_SET_USED(dev);
+
+   if (profile == NULL)
+   return -rte_mtr_error_set(error, EINVAL,
+ RTE_MTR_ERROR_TYPE_METER_PROFILE,
+ NULL, "Meter profile is null.");
+
+   if (profile_id == UINT32_MAX)
+   return -rte_mtr_error_set(error, EINVAL,
+ RTE_MTR_ERROR_TYPE_METER_PROFILE_ID,
+ NULL, "Meter profile id not valid.");
+
+   switch (profile->alg) {
+   case RTE_MTR_SRTCM_RFC2697:
+   if (profile->srtcm_rfc2697.cir > mtr_capa.meter_rate_max)
+   rc = -rte_mtr_error_set(error, EINVAL,
+   RTE_MTR_ERROR_TYPE_METER_PROFILE, NULL,
+   "CIR exceeds max meter rate");
+
+   if (profile->srtcm_rfc2697.cbs > NIX_BPF_BURST_MAX)
+   rc = -rte_mtr_error_set(error, EINVAL,
+   RTE_MTR_ERROR_TYPE_METER_PROFILE, NULL,
+   "CBS exceeds max meter burst size");
+
+   if (profile->srtcm_rfc2697.ebs > NIX_BPF_BURST_MAX)
+  

[dpdk-dev] [PATCH v4 18/28] net/cnxk: support ops to delete meter profile

[skori]

From: Sunil Kumar Kori 

Implement API to delete meter profile for CNXK platform.

Signed-off-by: Sunil Kumar Kori 
Signed-off-by: Rakesh Kudurumalla 
---
v4:
 - Rebase support on dpdk-next-net-mrvl branch
 - Handled meter action during flow destroy
 - Handled meter cleanup during port shutdown
 
v3:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for tree hierarchy
 - Fix naming convention
v2:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for linear hierarchy
 - Review comments incorporated

 drivers/net/cnxk/cnxk_ethdev_mtr.c | 30 ++
 1 file changed, 30 insertions(+)

diff --git a/drivers/net/cnxk/cnxk_ethdev_mtr.c 
b/drivers/net/cnxk/cnxk_ethdev_mtr.c
index bab03ebe95..f5e2c19480 100644
--- a/drivers/net/cnxk/cnxk_ethdev_mtr.c
+++ b/drivers/net/cnxk/cnxk_ethdev_mtr.c
@@ -189,9 +189,39 @@ cnxk_nix_mtr_profile_add(struct rte_eth_dev *eth_dev, 
uint32_t profile_id,
return 0;
 }
 
+static int
+cnxk_nix_mtr_profile_delete(struct rte_eth_dev *eth_dev, uint32_t profile_id,
+struct rte_mtr_error *error)
+{
+   struct cnxk_eth_dev *dev = cnxk_eth_pmd_priv(eth_dev);
+   struct cnxk_mtr_profile_node *fmp;
+
+   if (profile_id == UINT32_MAX)
+   return -rte_mtr_error_set(error, EINVAL,
+ RTE_MTR_ERROR_TYPE_METER_PROFILE_ID,
+ NULL, "Meter profile id not valid.");
+
+   fmp = nix_mtr_profile_find(dev, profile_id);
+   if (fmp == NULL)
+   return -rte_mtr_error_set(error, ENOENT,
+ RTE_MTR_ERROR_TYPE_METER_PROFILE_ID,
+ &profile_id,
+ "Meter profile is invalid.");
+
+   if (fmp->ref_cnt)
+   return -rte_mtr_error_set(error, EBUSY,
+ RTE_MTR_ERROR_TYPE_METER_PROFILE_ID,
+ NULL, "Meter profile is in use.");
+
+   TAILQ_REMOVE(&dev->mtr_profiles, fmp, next);
+   plt_free(fmp);
+   return 0;
+}
+
 const struct rte_mtr_ops nix_mtr_ops = {
.capabilities_get = cnxk_nix_mtr_capabilities_get,
.meter_profile_add = cnxk_nix_mtr_profile_add,
+   .meter_profile_delete = cnxk_nix_mtr_profile_delete,
 };
 
 int
-- 
2.25.1

[dpdk-dev] [PATCH v4 19/28] net/cnxk: support ops to validate meter policy

[skori]

From: Sunil Kumar Kori 

Implement API to validate meter policy for CNXK platform.

Signed-off-by: Sunil Kumar Kori 
Signed-off-by: Rakesh Kudurumalla 
---
v4:
 - Rebase support on dpdk-next-net-mrvl branch
 - Handled meter action during flow destroy
 - Handled meter cleanup during port shutdown
 
v3:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for tree hierarchy
 - Fix naming convention
v2:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for linear hierarchy
 - Review comments incorporated

 drivers/net/cnxk/cnxk_ethdev_mtr.c | 49 ++
 1 file changed, 49 insertions(+)

diff --git a/drivers/net/cnxk/cnxk_ethdev_mtr.c 
b/drivers/net/cnxk/cnxk_ethdev_mtr.c
index f5e2c19480..ec34327756 100644
--- a/drivers/net/cnxk/cnxk_ethdev_mtr.c
+++ b/drivers/net/cnxk/cnxk_ethdev_mtr.c
@@ -218,10 +218,59 @@ cnxk_nix_mtr_profile_delete(struct rte_eth_dev *eth_dev, 
uint32_t profile_id,
return 0;
 }
 
+static int
+cnxk_nix_mtr_policy_validate(struct rte_eth_dev *dev,
+ struct rte_mtr_meter_policy_params *policy,
+ struct rte_mtr_error *error)
+{
+   static const char *const action_color[] = {"Green", "Yellow", "Red"};
+   bool supported[RTE_COLORS] = {false, false, false};
+   const struct rte_flow_action *action;
+   char message[1024];
+   uint32_t i;
+
+   RTE_SET_USED(dev);
+
+   if (!policy)
+   return 0; /* Nothing to be validated */
+
+   for (i = 0; i < RTE_COLORS; i++) {
+   if (policy->actions[i]) {
+   for (action = policy->actions[i];
+action->type != RTE_FLOW_ACTION_TYPE_END;
+action++) {
+   if (action->type == RTE_FLOW_ACTION_TYPE_METER)
+   supported[i] = true;
+
+   if (action->type == RTE_FLOW_ACTION_TYPE_DROP)
+   supported[i] = true;
+
+   if (!supported[i]) {
+   sprintf(message,
+   "%s action is not valid",
+   action_color[i]);
+   return -rte_mtr_error_set(error,
+ ENOTSUP,
+ RTE_MTR_ERROR_TYPE_METER_POLICY, NULL,
+ message);
+   }
+   }
+   } else {
+   sprintf(message, "%s action is null", action_color[i]);
+   return -rte_mtr_error_set(error, EINVAL,
+   RTE_MTR_ERROR_TYPE_METER_POLICY, NULL,
+   message);
+   }
+   }
+
+   return 0;
+}
+
 const struct rte_mtr_ops nix_mtr_ops = {
.capabilities_get = cnxk_nix_mtr_capabilities_get,
.meter_profile_add = cnxk_nix_mtr_profile_add,
.meter_profile_delete = cnxk_nix_mtr_profile_delete,
+   .meter_policy_validate = cnxk_nix_mtr_policy_validate,
 };
 
 int
-- 
2.25.1

[dpdk-dev] [PATCH v4 20/28] net/cnxk: support ops to create meter policy

[skori]

From: Sunil Kumar Kori 

Implement API to add meter policy for CNXK platform.

Signed-off-by: Sunil Kumar Kori 
Signed-off-by: Rakesh Kudurumalla 
---
v4:
 - Rebase support on dpdk-next-net-mrvl branch
 - Handled meter action during flow destroy
 - Handled meter cleanup during port shutdown
 
v3:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for tree hierarchy
 - Fix naming convention
v2:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for linear hierarchy
 - Review comments incorporated

 drivers/net/cnxk/cnxk_ethdev.c |  1 +
 drivers/net/cnxk/cnxk_ethdev.h | 31 +++
 drivers/net/cnxk/cnxk_ethdev_mtr.c | 84 ++
 3 files changed, 116 insertions(+)

diff --git a/drivers/net/cnxk/cnxk_ethdev.c b/drivers/net/cnxk/cnxk_ethdev.c
index d748d5da04..6295fa0557 100644
--- a/drivers/net/cnxk/cnxk_ethdev.c
+++ b/drivers/net/cnxk/cnxk_ethdev.c
@@ -705,6 +705,7 @@ static int
 nix_ingress_policer_setup(struct cnxk_eth_dev *dev)
 {
TAILQ_INIT(&dev->mtr_profiles);
+   TAILQ_INIT(&dev->mtr_policy);
 
return 0;
 }
diff --git a/drivers/net/cnxk/cnxk_ethdev.h b/drivers/net/cnxk/cnxk_ethdev.h
index c076208d77..4cdce5dfd0 100644
--- a/drivers/net/cnxk/cnxk_ethdev.h
+++ b/drivers/net/cnxk/cnxk_ethdev.h
@@ -159,6 +159,35 @@ struct cnxk_timesync_info {
uint64_t *tx_tstamp;
 } __plt_cache_aligned;
 
+struct action_rss {
+   enum rte_eth_hash_function func;
+   uint32_t level;
+   uint64_t types;
+   uint32_t key_len;
+   uint32_t queue_num;
+   uint8_t *key;
+   uint16_t *queue;
+};
+
+struct policy_actions {
+   uint32_t action_fate;
+   union {
+   uint16_t queue;
+   uint32_t mtr_id;
+   struct action_rss *rss_desc;
+   };
+};
+
+struct cnxk_mtr_policy_node {
+   TAILQ_ENTRY(cnxk_mtr_policy_node) next;
+   /**< Pointer to the next flow meter structure. */
+   uint32_t id; /**< Policy id */
+   uint32_t mtr_id; /** Meter id */
+   struct rte_mtr_meter_policy_params policy;
+   struct policy_actions actions[RTE_COLORS];
+   uint32_t ref_cnt;
+};
+
 struct cnxk_mtr_profile_node {
TAILQ_ENTRY(cnxk_mtr_profile_node) next;
struct rte_mtr_meter_profile profile; /**< Profile detail. */
@@ -167,6 +196,7 @@ struct cnxk_mtr_profile_node {
 };
 
 TAILQ_HEAD(cnxk_mtr_profiles, cnxk_mtr_profile_node);
+TAILQ_HEAD(cnxk_mtr_policy, cnxk_mtr_policy_node);
 
 /* Security session private data */
 struct cnxk_eth_sec_sess {
@@ -315,6 +345,7 @@ struct cnxk_eth_dev {
 
/* Ingress policer */
struct cnxk_mtr_profiles mtr_profiles;
+   struct cnxk_mtr_policy mtr_policy;
 
/* Rx burst for cleanup(Only Primary) */
eth_rx_burst_t rx_pkt_burst_no_offload;
diff --git a/drivers/net/cnxk/cnxk_ethdev_mtr.c 
b/drivers/net/cnxk/cnxk_ethdev_mtr.c
index ec34327756..187a0b90b2 100644
--- a/drivers/net/cnxk/cnxk_ethdev_mtr.c
+++ b/drivers/net/cnxk/cnxk_ethdev_mtr.c
@@ -50,6 +50,18 @@ nix_mtr_profile_find(struct cnxk_eth_dev *dev, uint32_t 
profile_id)
return NULL;
 }
 
+static struct cnxk_mtr_policy_node *
+nix_mtr_policy_find(struct cnxk_eth_dev *dev, uint32_t meter_policy_id)
+{
+   struct cnxk_mtr_policy *fmps = &dev->mtr_policy;
+   struct cnxk_mtr_policy_node *fmp;
+
+   TAILQ_FOREACH(fmp, fmps, next)
+   if (meter_policy_id == fmp->id)
+   return fmp;
+   return NULL;
+}
+
 static int
 nix_mtr_profile_validate(struct cnxk_eth_dev *dev, uint32_t profile_id,
 struct rte_mtr_meter_profile *profile,
@@ -266,11 +278,83 @@ cnxk_nix_mtr_policy_validate(struct rte_eth_dev *dev,
return 0;
 }
 
+static void
+cnxk_fill_policy_actions(struct cnxk_mtr_policy_node *fmp,
+ struct rte_mtr_meter_policy_params *policy)
+
+{
+   const struct rte_flow_action_meter *mtr;
+   const struct rte_flow_action *action;
+   int i;
+
+   for (i = 0; i < RTE_COLORS; i++) {
+   if (policy->actions[i]) {
+   for (action = policy->actions[i];
+action->type != RTE_FLOW_ACTION_TYPE_END;
+action++) {
+   if (action->type ==
+   RTE_FLOW_ACTION_TYPE_METER) {
+   fmp->actions[i].action_fate =
+   action->type;
+   mtr = (const struct
+  rte_flow_action_meter *)
+ action->conf;
+   fmp->actions[i].mtr_id = mtr->mtr_id;
+   }
+
+   if (action->type == RTE_FLOW_ACTION_TYPE_DROP) {
+   fmp->actions[i].action_fate =
+

[dpdk-dev] [PATCH v4 21/28] net/cnxk: support ops to delete meter policy

[skori]

From: Sunil Kumar Kori 

Implement API to delete meter policy for CNXK platform.

Signed-off-by: Sunil Kumar Kori 
Signed-off-by: Rakesh Kudurumalla 
---
v4:
 - Rebase support on dpdk-next-net-mrvl branch
 - Handled meter action during flow destroy
 - Handled meter cleanup during port shutdown
 
v3:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for tree hierarchy
 - Fix naming convention
v2:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for linear hierarchy
 - Review comments incorporated

 drivers/net/cnxk/cnxk_ethdev_mtr.c | 26 ++
 1 file changed, 26 insertions(+)

diff --git a/drivers/net/cnxk/cnxk_ethdev_mtr.c 
b/drivers/net/cnxk/cnxk_ethdev_mtr.c
index 187a0b90b2..c3739749fc 100644
--- a/drivers/net/cnxk/cnxk_ethdev_mtr.c
+++ b/drivers/net/cnxk/cnxk_ethdev_mtr.c
@@ -349,12 +349,38 @@ cnxk_nix_mtr_policy_add(struct rte_eth_dev *eth_dev, 
uint32_t policy_id,
return rc;
 }
 
+static int
+cnxk_nix_mtr_policy_delete(struct rte_eth_dev *eth_dev, uint32_t policy_id,
+   struct rte_mtr_error *error)
+{
+   struct cnxk_eth_dev *dev = cnxk_eth_pmd_priv(eth_dev);
+   struct cnxk_mtr_policy_node *fmp;
+
+   fmp = nix_mtr_policy_find(dev, policy_id);
+   if (fmp == NULL) {
+   return -rte_mtr_error_set(error, ENOENT,
+ RTE_MTR_ERROR_TYPE_METER_POLICY_ID,
+ NULL, "No policy found");
+   }
+
+   if (fmp->ref_cnt)
+   return -rte_mtr_error_set(error, EBUSY,
+ RTE_MTR_ERROR_TYPE_METER_POLICY_ID,
+ NULL, "Meter policy is in use.");
+
+   TAILQ_REMOVE(&dev->mtr_policy, fmp, next);
+   plt_free(fmp);
+
+   return 0;
+}
+
 const struct rte_mtr_ops nix_mtr_ops = {
.capabilities_get = cnxk_nix_mtr_capabilities_get,
.meter_profile_add = cnxk_nix_mtr_profile_add,
.meter_profile_delete = cnxk_nix_mtr_profile_delete,
.meter_policy_validate = cnxk_nix_mtr_policy_validate,
.meter_policy_add = cnxk_nix_mtr_policy_add,
+   .meter_policy_delete = cnxk_nix_mtr_policy_delete,
 };
 
 int
-- 
2.25.1

[dpdk-dev] [PATCH v4 22/28] net/cnxk: support ops to create meter

[skori]

From: Sunil Kumar Kori 

Implement API to create meter instance for CNXK platform.

Signed-off-by: Sunil Kumar Kori 
Signed-off-by: Rakesh Kudurumalla 
---
v4:
 - Rebase support on dpdk-next-net-mrvl branch
 - Handled meter action during flow destroy
 - Handled meter cleanup during port shutdown
 
v3:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for tree hierarchy
 - Fix naming convention
v2:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for linear hierarchy
 - Review comments incorporated

 drivers/net/cnxk/cnxk_ethdev.c |  1 +
 drivers/net/cnxk/cnxk_ethdev.h | 25 
 drivers/net/cnxk/cnxk_ethdev_mtr.c | 93 ++
 3 files changed, 119 insertions(+)

diff --git a/drivers/net/cnxk/cnxk_ethdev.c b/drivers/net/cnxk/cnxk_ethdev.c
index 6295fa0557..2bb33d8f2d 100644
--- a/drivers/net/cnxk/cnxk_ethdev.c
+++ b/drivers/net/cnxk/cnxk_ethdev.c
@@ -706,6 +706,7 @@ nix_ingress_policer_setup(struct cnxk_eth_dev *dev)
 {
TAILQ_INIT(&dev->mtr_profiles);
TAILQ_INIT(&dev->mtr_policy);
+   TAILQ_INIT(&dev->mtr);
 
return 0;
 }
diff --git a/drivers/net/cnxk/cnxk_ethdev.h b/drivers/net/cnxk/cnxk_ethdev.h
index 4cdce5dfd0..10adeeec82 100644
--- a/drivers/net/cnxk/cnxk_ethdev.h
+++ b/drivers/net/cnxk/cnxk_ethdev.h
@@ -159,6 +159,28 @@ struct cnxk_timesync_info {
uint64_t *tx_tstamp;
 } __plt_cache_aligned;
 
+struct cnxk_meter_node {
+#define MAX_PRV_MTR_NODES 10
+   TAILQ_ENTRY(cnxk_meter_node) next;
+   /**< Pointer to the next flow meter structure. */
+   uint32_t id; /**< Usr mtr id. */
+   struct cnxk_mtr_profile_node *profile;
+   struct cnxk_mtr_policy_node *policy;
+   uint32_t bpf_id; /**< Hw mtr id. */
+   uint32_t rq_num;
+   uint32_t *rq_id;
+   uint16_t level;
+   uint32_t prev_id[MAX_PRV_MTR_NODES]; /**< Prev mtr id for chaining */
+   uint32_t prev_cnt;
+   uint32_t next_id; /**< Next mtr id for chaining */
+   bool is_prev;
+   bool is_next;
+   struct rte_mtr_params params;
+   struct roc_nix_bpf_objs profs;
+   bool is_used;
+   uint32_t ref_cnt;
+};
+
 struct action_rss {
enum rte_eth_hash_function func;
uint32_t level;
@@ -197,6 +219,7 @@ struct cnxk_mtr_profile_node {
 
 TAILQ_HEAD(cnxk_mtr_profiles, cnxk_mtr_profile_node);
 TAILQ_HEAD(cnxk_mtr_policy, cnxk_mtr_policy_node);
+TAILQ_HEAD(cnxk_mtr, cnxk_meter_node);
 
 /* Security session private data */
 struct cnxk_eth_sec_sess {
@@ -344,8 +367,10 @@ struct cnxk_eth_dev {
uint64_t clk_delta;
 
/* Ingress policer */
+   enum roc_nix_bpf_color precolor_tbl[ROC_NIX_BPF_PRE_COLOR_MAX];
struct cnxk_mtr_profiles mtr_profiles;
struct cnxk_mtr_policy mtr_policy;
+   struct cnxk_mtr mtr;
 
/* Rx burst for cleanup(Only Primary) */
eth_rx_burst_t rx_pkt_burst_no_offload;
diff --git a/drivers/net/cnxk/cnxk_ethdev_mtr.c 
b/drivers/net/cnxk/cnxk_ethdev_mtr.c
index c3739749fc..b258de0253 100644
--- a/drivers/net/cnxk/cnxk_ethdev_mtr.c
+++ b/drivers/net/cnxk/cnxk_ethdev_mtr.c
@@ -37,6 +37,18 @@ static struct rte_mtr_capabilities mtr_capa = {
  RTE_MTR_STATS_N_BYTES_YELLOW | RTE_MTR_STATS_N_BYTES_RED |
  RTE_MTR_STATS_N_BYTES_DROPPED};
 
+static struct cnxk_meter_node *
+nix_mtr_find(struct cnxk_eth_dev *dev, uint32_t meter_id)
+{
+   struct cnxk_mtr *fms = &dev->mtr;
+   struct cnxk_meter_node *fm;
+
+   TAILQ_FOREACH(fm, fms, next)
+   if (meter_id == fm->id)
+   return fm;
+   return NULL;
+}
+
 static struct cnxk_mtr_profile_node *
 nix_mtr_profile_find(struct cnxk_eth_dev *dev, uint32_t profile_id)
 {
@@ -374,6 +386,86 @@ cnxk_nix_mtr_policy_delete(struct rte_eth_dev *eth_dev, 
uint32_t policy_id,
return 0;
 }
 
+static int
+cnxk_nix_mtr_create(struct rte_eth_dev *eth_dev, uint32_t mtr_id,
+struct rte_mtr_params *params, int shared,
+struct rte_mtr_error *error)
+{
+   struct cnxk_eth_dev *dev = cnxk_eth_pmd_priv(eth_dev);
+   struct cnxk_mtr_profile_node *profile;
+   struct cnxk_mtr_policy_node *policy;
+   struct cnxk_mtr *fm = &dev->mtr;
+   struct cnxk_meter_node *mtr;
+   int i;
+
+   RTE_SET_USED(shared);
+
+   if (params == NULL)
+   return -rte_mtr_error_set(error, ENOENT,
+ RTE_MTR_ERROR_TYPE_MTR_PARAMS, NULL,
+ "Meter params are invalid.");
+
+   profile = nix_mtr_profile_find(dev, params->meter_profile_id);
+   if (profile == NULL)
+   return -rte_mtr_error_set(error, ENOENT,
+ RTE_MTR_ERROR_TYPE_METER_PROFILE_ID,
+ ¶ms->meter_profile_id,
+ "Meter profile is invalid.");
+
+   policy = nix_mtr_policy_find(dev, 

[dpdk-dev] [PATCH v4 23/28] net/cnxk: support ops to delete meter

[skori]

From: Sunil Kumar Kori 

Implement API to delete meter instance for CNXK platform.

Signed-off-by: Sunil Kumar Kori 
Signed-off-by: Rakesh Kudurumalla 
---
v4:
 - Rebase support on dpdk-next-net-mrvl branch
 - Handled meter action during flow destroy
 - Handled meter cleanup during port shutdown
 
v3:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for tree hierarchy
 - Fix naming convention
v2:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for linear hierarchy
 - Review comments incorporated

 drivers/net/cnxk/cnxk_ethdev_mtr.c | 90 ++
 1 file changed, 90 insertions(+)

diff --git a/drivers/net/cnxk/cnxk_ethdev_mtr.c 
b/drivers/net/cnxk/cnxk_ethdev_mtr.c
index b258de0253..28eba29f0a 100644
--- a/drivers/net/cnxk/cnxk_ethdev_mtr.c
+++ b/drivers/net/cnxk/cnxk_ethdev_mtr.c
@@ -8,6 +8,10 @@
 #define NIX_MTR_COUNT_MAX  73 /* 64(leaf) + 8(mid) + 1(top) */
 #define NIX_MTR_COUNT_PER_FLOW 3  /* 1(leaf) + 1(mid) + 1(top) */
 
+static const enum roc_nix_bpf_level_flag lvl_map[] = {ROC_NIX_BPF_LEVEL_F_LEAF,
+ ROC_NIX_BPF_LEVEL_F_MID,
+ ROC_NIX_BPF_LEVEL_F_TOP};
+
 static struct rte_mtr_capabilities mtr_capa = {
.n_max = NIX_MTR_COUNT_MAX,
.n_shared_max = NIX_MTR_COUNT_PER_FLOW,
@@ -466,6 +470,91 @@ cnxk_nix_mtr_create(struct rte_eth_dev *eth_dev, uint32_t 
mtr_id,
return 0;
 }
 
+static int
+cnxk_nix_mtr_destroy(struct rte_eth_dev *eth_dev, uint32_t mtr_id,
+ struct rte_mtr_error *error)
+{
+   struct cnxk_eth_dev *dev = cnxk_eth_pmd_priv(eth_dev);
+   struct roc_nix_bpf_objs profs = {0};
+   struct cnxk_mtr *fm = &dev->mtr;
+   struct roc_nix *nix = &dev->nix;
+   struct cnxk_meter_node *mtr;
+   struct cnxk_meter_node *mid_mtr;
+   struct cnxk_meter_node *top_mtr;
+   int rc = 0;
+
+   mtr = nix_mtr_find(dev, mtr_id);
+   if (mtr == NULL) {
+   return -rte_mtr_error_set(error, ENOENT,
+ RTE_MTR_ERROR_TYPE_MTR_ID, &mtr_id,
+ "Meter id is invalid.");
+   }
+
+   if (mtr->ref_cnt) {
+   return -rte_mtr_error_set(error, EADDRINUSE,
+ RTE_MTR_ERROR_TYPE_MTR_ID, &mtr_id,
+ "Meter id in use.");
+   }
+
+   switch (lvl_map[mtr->level]) {
+   case ROC_NIX_BPF_LEVEL_F_LEAF:
+   if (mtr->is_next) {
+   rc = roc_nix_bpf_connect(nix, ROC_NIX_BPF_LEVEL_F_LEAF,
+mtr->bpf_id,
+ROC_NIX_BPF_ID_INVALID);
+   }
+   break;
+   case ROC_NIX_BPF_LEVEL_F_MID:
+   while ((mtr->prev_cnt) + 1) {
+   mid_mtr =
+   nix_mtr_find(dev, mtr->prev_id[mtr->prev_cnt]);
+   rc = roc_nix_bpf_connect(nix, ROC_NIX_BPF_LEVEL_F_LEAF,
+mid_mtr->bpf_id,
+ROC_NIX_BPF_ID_INVALID);
+   mtr->prev_cnt--;
+   }
+   if (mtr->is_next) {
+   rc = roc_nix_bpf_connect(nix, ROC_NIX_BPF_LEVEL_F_MID,
+mtr->bpf_id,
+ROC_NIX_BPF_ID_INVALID);
+   }
+   break;
+   case ROC_NIX_BPF_LEVEL_F_TOP:
+   while (mtr->prev_cnt) {
+   top_mtr =
+   nix_mtr_find(dev, mtr->prev_id[mtr->prev_cnt]);
+   rc = roc_nix_bpf_connect(nix, ROC_NIX_BPF_LEVEL_F_MID,
+top_mtr->bpf_id,
+ROC_NIX_BPF_ID_INVALID);
+   mtr->prev_cnt--;
+   }
+   break;
+   default:
+   return -rte_mtr_error_set(error, EINVAL,
+ RTE_MTR_ERROR_TYPE_MTR_ID, NULL,
+ "Invalid meter level");
+   }
+
+   if (rc)
+   goto exit;
+
+   profs.level = mtr->level;
+   profs.count = 1;
+   profs.ids[0] = mtr->bpf_id;
+   rc = roc_nix_bpf_free(nix, &profs, 1);
+   if (rc)
+   goto exit;
+
+   mtr->policy->ref_cnt--;
+   mtr->profile->ref_cnt--;
+   TAILQ_REMOVE(fm, mtr, next);
+   plt_free(mtr->params.dscp_table);
+   plt_free(mtr);
+
+exit:
+   return rc;
+}
+
 const struct rte_mtr_ops nix_mtr_ops = {
.capabilities_get = cnxk_nix_mtr_capabilities_get,
.meter_profile_add = cnxk_nix_mtr_profile_add,
@@ -474,6 +563,7 @@ const struct rte_mtr_ops nix_mtr_ops = {
.meter_po

[dpdk-dev] [PATCH v4 24/28] net/cnxk: support ops to enable/disable meter

[skori]

From: Sunil Kumar Kori 

Implement API to enable or disable meter instance for
CNXK platform.

Signed-off-by: Sunil Kumar Kori 
Signed-off-by: Rakesh Kudurumalla 
---
v4:
 - Rebase support on dpdk-next-net-mrvl branch
 - Handled meter action during flow destroy
 - Handled meter cleanup during port shutdown
 
v3:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for tree hierarchy
 - Fix naming convention
v2:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for linear hierarchy
 - Review comments incorporated

 drivers/net/cnxk/cnxk_ethdev_mtr.c | 60 ++
 1 file changed, 60 insertions(+)

diff --git a/drivers/net/cnxk/cnxk_ethdev_mtr.c 
b/drivers/net/cnxk/cnxk_ethdev_mtr.c
index 28eba29f0a..b9359c185d 100644
--- a/drivers/net/cnxk/cnxk_ethdev_mtr.c
+++ b/drivers/net/cnxk/cnxk_ethdev_mtr.c
@@ -555,6 +555,64 @@ cnxk_nix_mtr_destroy(struct rte_eth_dev *eth_dev, uint32_t 
mtr_id,
return rc;
 }
 
+static int
+cnxk_nix_mtr_enable(struct rte_eth_dev *eth_dev, uint32_t mtr_id,
+struct rte_mtr_error *error)
+{
+   struct cnxk_eth_dev *dev = cnxk_eth_pmd_priv(eth_dev);
+   struct roc_nix *nix = &dev->nix;
+   struct cnxk_meter_node *mtr;
+   struct roc_nix_rq *rq;
+   uint32_t i;
+   int rc = 0;
+
+   mtr = nix_mtr_find(dev, mtr_id);
+   if (mtr == NULL) {
+   return -rte_mtr_error_set(error, ENOENT,
+ RTE_MTR_ERROR_TYPE_MTR_ID, NULL,
+ "Meter id is invalid.");
+   }
+
+   if (mtr->level != 0)
+   return 0;
+
+   for (i = 0; i < mtr->rq_num; i++) {
+   rq = &dev->rqs[mtr->rq_id[i]];
+   rc |= roc_nix_bpf_ena_dis(nix, mtr->bpf_id, rq, true);
+   }
+
+   return rc;
+}
+
+static int
+cnxk_nix_mtr_disable(struct rte_eth_dev *eth_dev, uint32_t mtr_id,
+ struct rte_mtr_error *error)
+{
+   struct cnxk_eth_dev *dev = cnxk_eth_pmd_priv(eth_dev);
+   struct roc_nix *nix = &dev->nix;
+   struct cnxk_meter_node *mtr;
+   struct roc_nix_rq *rq;
+   uint32_t i;
+   int rc = 0;
+
+   mtr = nix_mtr_find(dev, mtr_id);
+   if (mtr == NULL) {
+   return -rte_mtr_error_set(error, ENOENT,
+ RTE_MTR_ERROR_TYPE_MTR_ID, NULL,
+ "Meter id is invalid.");
+   }
+
+   if (mtr->level != 0)
+   return 0;
+
+   for (i = 0; i < mtr->rq_num; i++) {
+   rq = &dev->rqs[mtr->rq_id[i]];
+   rc |= roc_nix_bpf_ena_dis(nix, mtr->bpf_id, rq, false);
+   }
+
+   return rc;
+}
+
 const struct rte_mtr_ops nix_mtr_ops = {
.capabilities_get = cnxk_nix_mtr_capabilities_get,
.meter_profile_add = cnxk_nix_mtr_profile_add,
@@ -564,6 +622,8 @@ const struct rte_mtr_ops nix_mtr_ops = {
.meter_policy_delete = cnxk_nix_mtr_policy_delete,
.create = cnxk_nix_mtr_create,
.destroy = cnxk_nix_mtr_destroy,
+   .meter_enable = cnxk_nix_mtr_enable,
+   .meter_disable = cnxk_nix_mtr_disable,
 };
 
 int
-- 
2.25.1

[dpdk-dev] [PATCH v4 25/28] net/cnxk: support ops to update precolor DSCP table

[skori]

From: Sunil Kumar Kori 

Implement API to update DSCP table for pre-coloring for
incoming packet per nixlf for CNXK platform.

Signed-off-by: Sunil Kumar Kori 
Signed-off-by: Rakesh Kudurumalla 
---
v4:
 - Rebase support on dpdk-next-net-mrvl branch
 - Handled meter action during flow destroy
 - Handled meter cleanup during port shutdown
 
v3:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for tree hierarchy
 - Fix naming convention
v2:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for linear hierarchy
 - Review comments incorporated

 drivers/net/cnxk/cnxk_ethdev_mtr.c | 51 ++
 1 file changed, 51 insertions(+)

diff --git a/drivers/net/cnxk/cnxk_ethdev_mtr.c 
b/drivers/net/cnxk/cnxk_ethdev_mtr.c
index b9359c185d..be902ad47d 100644
--- a/drivers/net/cnxk/cnxk_ethdev_mtr.c
+++ b/drivers/net/cnxk/cnxk_ethdev_mtr.c
@@ -613,6 +613,56 @@ cnxk_nix_mtr_disable(struct rte_eth_dev *eth_dev, uint32_t 
mtr_id,
return rc;
 }
 
+static int
+cnxk_nix_mtr_dscp_table_update(struct rte_eth_dev *eth_dev, uint32_t mtr_id,
+   enum rte_color *dscp_table,
+   struct rte_mtr_error *error)
+{
+   enum roc_nix_bpf_color nix_dscp_tbl[ROC_NIX_BPF_PRE_COLOR_MAX];
+   enum roc_nix_bpf_color color_map[] = {ROC_NIX_BPF_COLOR_GREEN,
+ ROC_NIX_BPF_COLOR_YELLOW,
+ ROC_NIX_BPF_COLOR_RED};
+   struct cnxk_eth_dev *dev = cnxk_eth_pmd_priv(eth_dev);
+   struct roc_nix_bpf_precolor table;
+   struct roc_nix *nix = &dev->nix;
+   struct cnxk_meter_node *mtr;
+   int rc, i;
+
+   mtr = nix_mtr_find(dev, mtr_id);
+   if (mtr == NULL) {
+   return -rte_mtr_error_set(error, ENOENT,
+ RTE_MTR_ERROR_TYPE_MTR_ID, NULL,
+ "Meter object not found");
+   }
+
+   if (!dscp_table) {
+   for (i = 0; i < ROC_NIX_BPF_PRE_COLOR_MAX; i++)
+   nix_dscp_tbl[i] = ROC_NIX_BPF_COLOR_GREEN;
+   } else {
+   for (i = 0; i < ROC_NIX_BPF_PRE_COLOR_MAX; i++)
+   nix_dscp_tbl[i] = color_map[dscp_table[i]];
+   }
+
+   table.count = ROC_NIX_BPF_PRE_COLOR_MAX;
+   table.mode = ROC_NIX_BPF_PC_MODE_DSCP_OUTER;
+   for (i = 0; i < ROC_NIX_BPF_PRE_COLOR_MAX; i++)
+   table.color[i] = nix_dscp_tbl[i];
+
+   rc = roc_nix_bpf_pre_color_tbl_setup(nix, mtr->bpf_id,
+lvl_map[mtr->level], &table);
+   if (rc) {
+   rte_mtr_error_set(error, rc, RTE_MTR_ERROR_TYPE_UNSPECIFIED,
+ NULL, NULL);
+   goto exit;
+   }
+
+   for (i = 0; i < ROC_NIX_BPF_PRE_COLOR_MAX; i++)
+   dev->precolor_tbl[i] = nix_dscp_tbl[i];
+
+exit:
+   return rc;
+}
+
 const struct rte_mtr_ops nix_mtr_ops = {
.capabilities_get = cnxk_nix_mtr_capabilities_get,
.meter_profile_add = cnxk_nix_mtr_profile_add,
@@ -624,6 +674,7 @@ const struct rte_mtr_ops nix_mtr_ops = {
.destroy = cnxk_nix_mtr_destroy,
.meter_enable = cnxk_nix_mtr_enable,
.meter_disable = cnxk_nix_mtr_disable,
+   .meter_dscp_table_update = cnxk_nix_mtr_dscp_table_update,
 };
 
 int
-- 
2.25.1

[dpdk-dev] [PATCH v4 26/28] net/cnxk: support ops to read/update meter stats

[skori]

From: Sunil Kumar Kori 

Implement API to read and update stats corresponding to
given meter instance for CNXK platform.

Signed-off-by: Sunil Kumar Kori 
Signed-off-by: Rakesh Kudurumalla 
---
v4:
 - Rebase support on dpdk-next-net-mrvl branch
 - Handled meter action during flow destroy
 - Handled meter cleanup during port shutdown
 
v3:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for tree hierarchy
 - Fix naming convention
v2:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for linear hierarchy
 - Review comments incorporated

 drivers/net/cnxk/cnxk_ethdev_mtr.c | 141 +
 1 file changed, 141 insertions(+)

diff --git a/drivers/net/cnxk/cnxk_ethdev_mtr.c 
b/drivers/net/cnxk/cnxk_ethdev_mtr.c
index be902ad47d..08af3f628a 100644
--- a/drivers/net/cnxk/cnxk_ethdev_mtr.c
+++ b/drivers/net/cnxk/cnxk_ethdev_mtr.c
@@ -8,6 +8,21 @@
 #define NIX_MTR_COUNT_MAX  73 /* 64(leaf) + 8(mid) + 1(top) */
 #define NIX_MTR_COUNT_PER_FLOW 3  /* 1(leaf) + 1(mid) + 1(top) */
 
+#define NIX_BPF_STATS_MASK_ALL 
\
+   {  \
+   ROC_NIX_BPF_GREEN_PKT_F_PASS | ROC_NIX_BPF_GREEN_OCTS_F_PASS | \
+   ROC_NIX_BPF_GREEN_PKT_F_DROP | \
+   ROC_NIX_BPF_GREEN_OCTS_F_DROP |\
+   ROC_NIX_BPF_YELLOW_PKT_F_PASS |\
+   ROC_NIX_BPF_YELLOW_OCTS_F_PASS |   \
+   ROC_NIX_BPF_YELLOW_PKT_F_DROP |\
+   ROC_NIX_BPF_YELLOW_OCTS_F_DROP |   \
+   ROC_NIX_BPF_RED_PKT_F_PASS |   \
+   ROC_NIX_BPF_RED_OCTS_F_PASS |  \
+   ROC_NIX_BPF_RED_PKT_F_DROP |   \
+   ROC_NIX_BPF_RED_OCTS_F_DROP\
+   }
+
 static const enum roc_nix_bpf_level_flag lvl_map[] = {ROC_NIX_BPF_LEVEL_F_LEAF,
  ROC_NIX_BPF_LEVEL_F_MID,
  ROC_NIX_BPF_LEVEL_F_TOP};
@@ -663,6 +678,130 @@ cnxk_nix_mtr_dscp_table_update(struct rte_eth_dev 
*eth_dev, uint32_t mtr_id,
return rc;
 }
 
+static int
+cnxk_nix_mtr_stats_update(struct rte_eth_dev *eth_dev, uint32_t mtr_id,
+  uint64_t stats_mask, struct rte_mtr_error *error)
+{
+   struct cnxk_eth_dev *dev = cnxk_eth_pmd_priv(eth_dev);
+   struct cnxk_meter_node *mtr;
+
+   if (!stats_mask)
+   return -rte_mtr_error_set(error, EINVAL,
+ RTE_MTR_ERROR_TYPE_MTR_PARAMS, NULL,
+ "no bit is set to stats mask");
+
+   mtr = nix_mtr_find(dev, mtr_id);
+   if (mtr == NULL) {
+   return -rte_mtr_error_set(error, ENOENT,
+ RTE_MTR_ERROR_TYPE_MTR_ID, NULL,
+ "Meter object not found");
+   }
+
+   mtr->params.stats_mask = stats_mask;
+   return 0;
+}
+
+static int
+cnxk_nix_mtr_stats_read(struct rte_eth_dev *eth_dev, uint32_t mtr_id,
+struct rte_mtr_stats *stats, uint64_t *stats_mask,
+int clear, struct rte_mtr_error *error)
+{
+   uint8_t yellow_pkt_pass, yellow_octs_pass, yellow_pkt_drop;
+   uint8_t green_octs_drop, yellow_octs_drop, red_octs_drop;
+   uint8_t green_pkt_pass, green_octs_pass, green_pkt_drop;
+   struct cnxk_eth_dev *dev = cnxk_eth_pmd_priv(eth_dev);
+   uint8_t red_pkt_pass, red_octs_pass, red_pkt_drop;
+   uint64_t bpf_stats[ROC_NIX_BPF_STATS_MAX] = {0};
+   uint64_t mask = NIX_BPF_STATS_MASK_ALL;
+   struct roc_nix *nix = &dev->nix;
+   struct cnxk_meter_node *mtr;
+   int rc;
+
+   if (!stats)
+   return -rte_mtr_error_set(error, EINVAL,
+ RTE_MTR_ERROR_TYPE_MTR_PARAMS, NULL,
+ "stats pointer is NULL");
+
+   mtr = nix_mtr_find(dev, mtr_id);
+   if (mtr == NULL) {
+   return -rte_mtr_error_set(error, ENOENT,
+ RTE_MTR_ERROR_TYPE_MTR_ID, NULL,
+ "Meter object not found");
+   }
+
+   rc = roc_nix_bpf_stats_read(nix, mtr->bpf_id, mask, lvl_map[mtr->level],
+   bpf_stats);
+   if (rc) {
+   rte_mtr_error_set(error, rc, RTE_MTR_ERROR_TYPE_UNSPECIFIED,
+ NULL, NULL);
+   goto exit;
+   }
+
+   green_pkt_pass = roc_nix_bpf_stats_to_idx(ROC_NIX_BPF_GREEN_PKT_F_PASS);
+   green_octs_pass =
+

[dpdk-dev] [PATCH v4 27/28] net/cnxk: support meter action to flow create

[skori]

From: Sunil Kumar Kori 

Meters are configured per flow using rte_flow_create API.
Implement support for meter action applied on the flow.

Signed-off-by: Sunil Kumar Kori 
Signed-off-by: Rakesh Kudurumalla 
---
v4:
 - Rebase support on dpdk-next-net-mrvl branch
 - Handled meter action during flow destroy
 - Handled meter cleanup during port shutdown
 
v3:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for tree hierarchy
 - Fix naming convention
v2:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for linear hierarchy
 - Review comments incorporated

 doc/guides/nics/features/cnxk.ini  |   1 +
 doc/guides/nics/features/cnxk_vf.ini   |   1 +
 doc/guides/rel_notes/release_21_11.rst |   1 +
 drivers/net/cnxk/cn10k_rte_flow.c  | 171 -
 drivers/net/cnxk/cnxk_ethdev.h |  18 +
 drivers/net/cnxk/cnxk_ethdev_mtr.c | 506 +
 drivers/net/cnxk/cnxk_rte_flow.c   |   4 +
 7 files changed, 701 insertions(+), 1 deletion(-)

diff --git a/doc/guides/nics/features/cnxk.ini 
b/doc/guides/nics/features/cnxk.ini
index 1ced3ee903..f0586457ca 100644
--- a/doc/guides/nics/features/cnxk.ini
+++ b/doc/guides/nics/features/cnxk.ini
@@ -79,6 +79,7 @@ count= Y
 drop = Y
 flag = Y
 mark = Y
+meter= Y
 of_pop_vlan  = Y
 of_push_vlan = Y
 of_set_vlan_pcp  = Y
diff --git a/doc/guides/nics/features/cnxk_vf.ini 
b/doc/guides/nics/features/cnxk_vf.ini
index 139d9b9892..f3d9f23f17 100644
--- a/doc/guides/nics/features/cnxk_vf.ini
+++ b/doc/guides/nics/features/cnxk_vf.ini
@@ -71,6 +71,7 @@ count= Y
 drop = Y
 flag = Y
 mark = Y
+meter= Y
 of_pop_vlan  = Y
 of_push_vlan = Y
 of_set_vlan_pcp  = Y
diff --git a/doc/guides/rel_notes/release_21_11.rst 
b/doc/guides/rel_notes/release_21_11.rst
index 16c48d54a1..5ec80377ee 100644
--- a/doc/guides/rel_notes/release_21_11.rst
+++ b/doc/guides/rel_notes/release_21_11.rst
@@ -91,6 +91,7 @@ New Features
   * Added rte_tm support.
   * Added support for Inline IPsec for CN9K event mode and CN10K
 poll mode and event mode.
+  * Added support for ingress meter for CN10K platform.
 
 * **Updated af_packet ethdev driver.**
 
diff --git a/drivers/net/cnxk/cn10k_rte_flow.c 
b/drivers/net/cnxk/cn10k_rte_flow.c
index b04de6a7e6..00c0b172e4 100644
--- a/drivers/net/cnxk/cn10k_rte_flow.c
+++ b/drivers/net/cnxk/cn10k_rte_flow.c
@@ -6,6 +6,113 @@
 #include "cn10k_ethdev.h"
 #include "cn10k_rx.h"
 
+static int
+cn10k_mtr_connect(struct rte_eth_dev *eth_dev, uint32_t mtr_id)
+{
+   return nix_mtr_connect(eth_dev, mtr_id);
+}
+
+static int
+cn10k_mtr_configure(struct rte_eth_dev *eth_dev,
+   const struct rte_flow_action actions[])
+{
+   uint32_t mtr_id = 0x, prev_mtr_id = 0x, next_mtr_id = 0x;
+   const struct rte_flow_action_meter *mtr_conf;
+   const struct rte_flow_action_queue *q_conf;
+   const struct rte_flow_action_rss *rss_conf;
+   struct cnxk_mtr_policy_node *policy;
+   bool is_mtr_act = false;
+   int tree_level = 0;
+   int rc = -EINVAL, i;
+
+   for (i = 0; actions[i].type != RTE_FLOW_ACTION_TYPE_END; i++) {
+   if (actions[i].type == RTE_FLOW_ACTION_TYPE_METER) {
+   mtr_conf = (const struct rte_flow_action_meter
+   *)(actions->conf);
+   mtr_id = mtr_conf->mtr_id;
+   is_mtr_act = true;
+   }
+   if (actions[i].type == RTE_FLOW_ACTION_TYPE_QUEUE) {
+   q_conf = (const struct rte_flow_action_queue
+ *)(actions->conf);
+   if (is_mtr_act)
+   nix_mtr_rq_update(eth_dev, mtr_id, 1,
+ &q_conf->index);
+   }
+   if (actions[i].type == RTE_FLOW_ACTION_TYPE_RSS) {
+   rss_conf = (const struct rte_flow_action_rss
+   *)(actions->conf);
+   if (is_mtr_act)
+   nix_mtr_rq_update(eth_dev, mtr_id,
+ rss_conf->queue_num,
+ rss_conf->queue);
+   }
+   }
+
+   if (!is_mtr_act)
+   return rc;
+
+   prev_mtr_id = mtr_id;
+   next_mtr_id = mtr_id;
+   while (next_mtr_id != 0x) {
+   rc = nix_mtr_validate(eth_dev, next_mtr_id);
+   if (rc)
+   return rc;
+
+   rc = nix_mtr_policy_act_get(eth_dev, next_mtr_id, &policy);
+   if (rc)
+   return rc;
+
+   rc = nix_mtr_color_action_validate(eth_dev, mtr_id,
+

[dpdk-dev] [PATCH v4 28/28] net/cnxk: support meter action to flow destroy

[skori]

From: Sunil Kumar Kori 

Meters are configured per flow using rte_flow_create API.
Patch adds support for destroy operation for meter action
applied on the flow.

Signed-off-by: Sunil Kumar Kori 
Signed-off-by: Rakesh Kudurumalla 
---
v4:
 - Rebase support on dpdk-next-net-mrvl branch
 - Handled meter action during flow destroy
 - Handled meter cleanup during port shutdown
 
v3:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for tree hierarchy
 - Fix naming convention
v2:
 - Rebase support on latest DPDK
 - Handled multilevel chaining for linear hierarchy
 - Review comments incorporated

 drivers/net/cnxk/cn10k_rte_flow.c  | 22 +++-
 drivers/net/cnxk/cnxk_ethdev.c | 40 ++
 drivers/net/cnxk/cnxk_ethdev.h |  2 ++
 drivers/net/cnxk/cnxk_ethdev_mtr.c |  7 ++
 4 files changed, 70 insertions(+), 1 deletion(-)

diff --git a/drivers/net/cnxk/cn10k_rte_flow.c 
b/drivers/net/cnxk/cn10k_rte_flow.c
index 00c0b172e4..8c87452934 100644
--- a/drivers/net/cnxk/cn10k_rte_flow.c
+++ b/drivers/net/cnxk/cn10k_rte_flow.c
@@ -12,6 +12,14 @@ cn10k_mtr_connect(struct rte_eth_dev *eth_dev, uint32_t 
mtr_id)
return nix_mtr_connect(eth_dev, mtr_id);
 }
 
+static int
+cn10k_mtr_destroy(struct rte_eth_dev *eth_dev, uint32_t mtr_id)
+{
+   struct rte_mtr_error mtr_error;
+
+   return nix_mtr_destroy(eth_dev, mtr_id, &mtr_error);
+}
+
 static int
 cn10k_mtr_configure(struct rte_eth_dev *eth_dev,
const struct rte_flow_action actions[])
@@ -215,6 +223,8 @@ cn10k_flow_destroy(struct rte_eth_dev *eth_dev, struct 
rte_flow *rte_flow,
struct cnxk_eth_dev *dev = cnxk_eth_pmd_priv(eth_dev);
int mark_actions = 0, vtag_actions = 0;
struct roc_npc *npc = &dev->npc;
+   uint32_t mtr_id;
+   int rc;
 
mark_actions = roc_npc_mark_actions_get(npc);
if (mark_actions) {
@@ -237,5 +247,15 @@ cn10k_flow_destroy(struct rte_eth_dev *eth_dev, struct 
rte_flow *rte_flow,
}
}
 
-   return cnxk_flow_destroy(eth_dev, flow, error);
+   mtr_id = flow->mtr_id;
+   rc = cnxk_flow_destroy(eth_dev, flow, error);
+   if (!rc) {
+   rc = cn10k_mtr_destroy(eth_dev, mtr_id);
+   if (rc) {
+   rte_flow_error_set(error, ENXIO,
+   RTE_FLOW_ERROR_TYPE_UNSPECIFIED, NULL,
+   "Meter attached to this flow does not exist");
+   }
+   }
+   return rc;
 }
diff --git a/drivers/net/cnxk/cnxk_ethdev.c b/drivers/net/cnxk/cnxk_ethdev.c
index 2bb33d8f2d..69bd58ae45 100644
--- a/drivers/net/cnxk/cnxk_ethdev.c
+++ b/drivers/net/cnxk/cnxk_ethdev.c
@@ -141,6 +141,38 @@ nix_security_setup(struct cnxk_eth_dev *dev)
return rc;
 }
 
+static int
+nix_meter_fini(struct cnxk_eth_dev *dev)
+{
+   struct cnxk_meter_node *next_mtr = NULL;
+   struct roc_nix_bpf_objs profs = {0};
+   struct cnxk_meter_node *mtr = NULL;
+   struct cnxk_mtr *fms = &dev->mtr;
+   struct roc_nix *nix = &dev->nix;
+   struct roc_nix_rq *rq;
+   uint32_t i;
+   int rc;
+
+   RTE_TAILQ_FOREACH_SAFE(mtr, fms, next, next_mtr) {
+   for (i = 0; i < mtr->rq_num; i++) {
+   rq = &dev->rqs[mtr->rq_id[i]];
+   rc |= roc_nix_bpf_ena_dis(nix, mtr->bpf_id, rq, false);
+   }
+
+   profs.level = mtr->level;
+   profs.count = 1;
+   profs.ids[0] = mtr->bpf_id;
+   rc = roc_nix_bpf_free(nix, &profs, 1);
+
+   if (rc)
+   return rc;
+
+   TAILQ_REMOVE(fms, mtr, next);
+   plt_free(mtr);
+   }
+   return 0;
+}
+
 static int
 nix_security_release(struct cnxk_eth_dev *dev)
 {
@@ -1012,6 +1044,11 @@ cnxk_nix_configure(struct rte_eth_dev *eth_dev)
if (rc)
goto fail_configure;
 
+   /* Disable and free rte_meter entries */
+   rc = nix_meter_fini(dev);
+   if (rc)
+   goto fail_configure;
+
/* Cleanup security support */
rc = nix_security_release(dev);
if (rc)
@@ -1668,6 +1705,9 @@ cnxk_eth_dev_uninit(struct rte_eth_dev *eth_dev, bool 
reset)
 
roc_nix_npc_rx_ena_dis(nix, false);
 
+   /* Disable and free rte_meter entries */
+   nix_meter_fini(dev);
+
/* Disable and free rte_flow entries */
roc_npc_fini(&dev->npc);
 
diff --git a/drivers/net/cnxk/cnxk_ethdev.h b/drivers/net/cnxk/cnxk_ethdev.h
index 7e13956572..2d4b073985 100644
--- a/drivers/net/cnxk/cnxk_ethdev.h
+++ b/drivers/net/cnxk/cnxk_ethdev.h
@@ -589,6 +589,8 @@ int nix_mtr_level_update(struct rte_eth_dev *eth_dev, 
uint32_t id,
 uint32_t level);
 int nix_mtr_configure(struct rte_eth_dev *eth_dev, uint32_t id);
 int nix_mtr_connect(struct rte_eth_dev *eth_dev, uint32_

Re: [dpdk-dev] [EXT] [PATCH v3 8/8] examples/ipsec-secgw: add support for additional algorithms

[Nicolau, Radu]

Hi Akhil, thanks for the feedback, some comments below.

On 10/8/2021 8:07 PM, Akhil Goyal wrote:

Add support for AES-GMAC, AES_CTR, AES_XCBC_MAC,
AES_CCM, CHACHA20_POLY1305

Signed-off-by: Declan Doherty 
Signed-off-by: Radu Nicolau 
---
  examples/ipsec-secgw/ipsec.h |   3 +-
  examples/ipsec-secgw/sa.c| 133 ---
  2 files changed, 126 insertions(+), 10 deletions(-)


Documentation?
Release notes?


I will follow up with an updated patchset.



-
+#define MAX_KEY_SIZE   96

Max key length defined below is 64, then why 96?

I guess just to have some room, I will set it back to 64.



  /*
}

-   if (sa->aead_algo == RTE_CRYPTO_AEAD_AES_GCM) {
-   iv_length = 12;
+
+   if (sa->aead_algo == RTE_CRYPTO_AEAD_AES_GCM ||
+   sa->aead_algo == RTE_CRYPTO_AEAD_AES_CCM ||
+   sa->aead_algo ==
RTE_CRYPTO_AEAD_CHACHA20_POLY1305) {
+
+   if (ips->type ==
+
RTE_SECURITY_ACTION_TYPE_INLINE_CRYPTO) {
+   iv_length = 8;

How is IV length dependent on the action type?
It should be same for all modes for a particular algorithm. Right?


The issue here with inline crypto and AEAD algorithms is that we have 
the IV and the salt used for building the nonce but only the IV is 
included in the ESP header. And technically and according to the RFCs 
the value in the inline branch is the correct one, the other is actually 
the nonce buffer length, not the IV length. I'm not sure if this is the 
proper fix, probably it's not, but it's probably a change that will not 
break it for other crypto devices.

Re: [dpdk-dev] [PATCH v5 7/7] ethdev: hide eth dev related structures

[Ananyev, Konstantin]

> 
> On 10/7/21 2:27 PM, Konstantin Ananyev wrote:
> > Move rte_eth_dev, rte_eth_dev_data, rte_eth_rxtx_callback and related
> > data into private header (ethdev_driver.h).
> > Few minor changes to keep DPDK building after that.
> >
> > Signed-off-by: Konstantin Ananyev 
> 
> [snip]
> 
> I realize that many notes below correspond to just
> moved code, but I still think that ti would be nice
> to fix anyway.

Yes, that’s just cut and paste from rte_ethdev_core.h and ethdev_driver.h
It is probably a good idea to clean-up and might be re-layout rte_eth_dev and 
friends,
but I don't think it has to be part of this patch-set.
After all, if it will become internal structure, that work could be done for 
any DPDK release. 

Re: [dpdk-dev] [PATCH v5 5/7] ethdev: make fast-path functions to use new flat array

[Ananyev, Konstantin]

> 
> On 10/7/21 2:27 PM, Konstantin Ananyev wrote:
> > Rework fast-path ethdev functions to use rte_eth_fp_ops[].
> > While it is an API/ABI breakage, this change is intended to be
> > transparent for both users (no changes in user app is required) and
> > PMD developers (no changes in PMD is required).
> > One extra thing to note - RX/TX callback invocation will cause extra
> > function call with these changes. That might cause some insignificant
> > slowdown for code-path where RX/TX callbacks are heavily involved.
> 
> I'm sorry for nit picking here and below:
> 
> RX -> Rx, TX -> Tx everywhere above.
> 
> >
> > Signed-off-by: Konstantin Ananyev 
> > ---
> >  lib/ethdev/ethdev_private.c |  31 +
> >  lib/ethdev/rte_ethdev.h | 242 ++--
> >  lib/ethdev/version.map  |   3 +
> >  3 files changed, 208 insertions(+), 68 deletions(-)
> >
> > diff --git a/lib/ethdev/ethdev_private.c b/lib/ethdev/ethdev_private.c
> > index 3eeda6e9f9..1222c6f84e 100644
> > --- a/lib/ethdev/ethdev_private.c
> > +++ b/lib/ethdev/ethdev_private.c
> > @@ -226,3 +226,34 @@ eth_dev_fp_ops_setup(struct rte_eth_fp_ops *fpo,
> > fpo->txq.data = dev->data->tx_queues;
> > fpo->txq.clbk = (void **)(uintptr_t)dev->pre_tx_burst_cbs;
> >  }
> > +
> > +uint16_t
> > +rte_eth_call_rx_callbacks(uint16_t port_id, uint16_t queue_id,
> > +   struct rte_mbuf **rx_pkts, uint16_t nb_rx, uint16_t nb_pkts,
> > +   void *opaque)
> > +{
> > +   const struct rte_eth_rxtx_callback *cb = opaque;
> > +
> > +   while (cb != NULL) {
> > +   nb_rx = cb->fn.rx(port_id, queue_id, rx_pkts, nb_rx,
> > +   nb_pkts, cb->param);
> > +   cb = cb->next;
> > +   }
> > +
> > +   return nb_rx;
> > +}
> > +
> > +uint16_t
> > +rte_eth_call_tx_callbacks(uint16_t port_id, uint16_t queue_id,
> > +   struct rte_mbuf **tx_pkts, uint16_t nb_pkts, void *opaque)
> > +{
> > +   const struct rte_eth_rxtx_callback *cb = opaque;
> > +
> > +   while (cb != NULL) {
> > +   nb_pkts = cb->fn.tx(port_id, queue_id, tx_pkts, nb_pkts,
> > +   cb->param);
> > +   cb = cb->next;
> > +   }
> > +
> > +   return nb_pkts;
> > +}
> > diff --git a/lib/ethdev/rte_ethdev.h b/lib/ethdev/rte_ethdev.h
> > index cdd16d6e57..c0e1a40681 100644
> > --- a/lib/ethdev/rte_ethdev.h
> > +++ b/lib/ethdev/rte_ethdev.h
> > @@ -4904,6 +4904,33 @@ int rte_eth_representor_info_get(uint16_t port_id,
> >
> >  #include 
> >
> > +/**
> > + * @internal
> > + * Helper routine for eth driver rx_burst API.
> 
> rx -> Rx
> 
> > + * Should be called at exit from PMD's rte_eth_rx_bulk implementation.
> > + * Does necessary post-processing - invokes RX callbacks if any, etc.
> 
> RX -> Rx
> 
> > + *
> > + * @param port_id
> > + *  The port identifier of the Ethernet device.
> > + * @param queue_id
> > + *  The index of the receive queue from which to retrieve input packets.
> 
> Isn't:
> The index of the queue from which packets are received from?

I copied it from comments from rte_eth_rx_burst().
I suppose it is just two ways to say the same thing.

> 
> > + * @param rx_pkts
> > + *   The address of an array of pointers to *rte_mbuf* structures that
> > + *   have been retrieved from the device.
> > + * @param nb_pkts
> 
> Should be @param nb_rx

Ack, will fix. 

> 
> > + *   The number of packets that were retrieved from the device.
> > + * @param nb_pkts
> > + *   The number of elements in *rx_pkts* array.
> 
> @p should be used to refer to a paramter.

To be more precise you are talking about:
s/*rx_pkts*/@ rx_pkts/
?

> 
> The description does not help to understand why both nb_rx and
> nb_pkts are necessary. Isn't nb_pkts >= nb_rx and nb_rx
> sufficient?

Nope,  that's for callbacks call.
Will update the comment.
 
> > + * @param opaque
> > + *   Opaque pointer of RX queue callback related data.
> 
> RX -> Rx
> 
> > + *
> > + * @return
> > + *  The number of packets effectively supplied to the *rx_pkts* array.
> 
> @p should be used to refer to a parameter.
> 
> > + */
> > +uint16_t rte_eth_call_rx_callbacks(uint16_t port_id, uint16_t queue_id,
> > +   struct rte_mbuf **rx_pkts, uint16_t nb_rx, uint16_t nb_pkts,
> > +   void *opaque);
> > +
> >  /**
> >   *
> >   * Retrieve a burst of input packets from a receive queue of an Ethernet
> > @@ -4995,23 +5022,37 @@ static inline uint16_t
> >  rte_eth_rx_burst(uint16_t port_id, uint16_t queue_id,
> >  struct rte_mbuf **rx_pkts, const uint16_t nb_pkts)
> >  {
> > -   struct rte_eth_dev *dev = &rte_eth_devices[port_id];
> > uint16_t nb_rx;
> > +   struct rte_eth_fp_ops *p;
> 
> p is typically a very bad name in a funcion with
> many pointer variables etc. May be "fpo" as in previous
> patch?
> 
> > +   void *cb, *qd;
> 
> Please, avoid variable, expecially pointers, declaration in
> one line.

Here and in other places, I think local variable names and placement,
is just a matter of personal preference.

> 
> I'd suggest to us

Re: [dpdk-dev] [RFC PATCH 2/2] ethdev: add capability to keep indirect actions on restart

[Dmitry Kozlyuk]

> -Original Message-
> From: Andrew Rybchenko 
> Sent: 11 октября 2021 г. 16:58
> To: Dmitry Kozlyuk ; Ajit Khaparde
> 
> Cc: dpdk-dev ; Matan Azrad ; Ori Kam
> ; NBU-Contact-Thomas Monjalon
> ; Ferruh Yigit 
> Subject: Re: [dpdk-dev] [RFC PATCH 2/2] ethdev: add capability to keep 
> indirect
> actions on restart
> 
> External email: Use caution opening links or attachments
> 
> 
> On 10/7/21 11:16 AM, Dmitry Kozlyuk wrote:
> >> -Original Message-
> >> From: Ajit Khaparde 
> >> Sent: 6 октября 2021 г. 20:13
> >> To: Dmitry Kozlyuk 
> >> Cc: dpdk-dev ; Matan Azrad ; Ori Kam
> >> ; NBU-Contact-Thomas Monjalon
> >> ; Ferruh Yigit ; Andrew
> >> Rybchenko 
> >> Subject: Re: [dpdk-dev] [RFC PATCH 2/2] ethdev: add capability to
> >> keep indirect actions on restart
> >>
> >> On Wed, Sep 1, 2021 at 1:55 AM Dmitry Kozlyuk 
> wrote:
> >>>
> >>> rte_flow_action_handle_create() did not mention what happens with an
> >>> indirect action when a device is stopped, possibly reconfigured, and
> >>> started again. It is natural for some indirect actions to be
> >>> persistent, like counters and meters; keeping others just saves
> >>> application time and complexity. However, not all PMDs can support it.
> >>> It is proposed to add a device capability to indicate if indirect
> >>> actions are kept across the above sequence or implicitly destroyed.
> >>>
> >>> It may happen that in the future a PMD acquires support for a type
> >>> of indirect actions that it cannot keep across a restart. It is
> >>> undesirable to stop advertising the capability so that applications
> >>> that don't use actions of the problematic type can still take advantage 
> >>> of it.
> >>> This is why PMDs are allowed to keep only a subset of indirect
> >>> actions provided that the vendor mandatorily documents it.
> >> Sorry - I am seeing this late.
> >> This could become confusing.
> >> May be it is better for the PMDs to specify which actions are persistent.
> >> How about adding a bit for the possible actions of interest.
> >> And then PMDs can set bits for actions which can be persistent across
> >> stop, start and reconfigurations?
> >
> > This approach was considered, but there is a risk of quickly running out of
> capability bits. Each action would consume one bit plus as many bits as there 
> are
> special conditions for it in all the PMDs, because conditions are likely to 
> be PMD-
> specific. And the application will anyway need to consider specific 
> conditions to
> know which bit to test, so the meaning of the bits will be PMD-specific. On 
> the
> other hand, PMDs are not expected to exercise this loophole unless absolutely
> needed.
> >
> 
> May be we should separate at least transfer and non-transfer rules? Transfer
> rules are less configuration dependent.

Do you suggest splitting the bit from patch 1/2 in two?
Or did you mean indirect actions with only "transfer" bit set
and suggest splitting the bit from this patch in two?

Re: [dpdk-dev] [PATCH v2 0/5] cryptodev: hide internal structures

[Zhang, Roy Fan]

Hi Akhil,

The approach looks great but we may have to check if it works in multi-process
environment - since all enqueue/dequeue handlers are set by primary process
the secondary process may not recognize the fp_ops data.

We will run a quick test to see if it is true.

Regards,
Fan

> -Original Message-
> From: Akhil Goyal 
> Sent: Monday, October 11, 2021 1:43 PM
> To: dev@dpdk.org
> Cc: tho...@monjalon.net; david.march...@redhat.com;
> hemant.agra...@nxp.com; ano...@marvell.com; De Lara Guarch, Pablo
> ; Trahe, Fiona ;
> Doherty, Declan ; ma...@nvidia.com;
> g.si...@nxp.com; Zhang, Roy Fan ;
> jianjay.z...@huawei.com; asoma...@amd.com; ruifeng.w...@arm.com;
> Ananyev, Konstantin ; Nicolau, Radu
> ; ajit.khapa...@broadcom.com;
> rnagadhee...@marvell.com; adwiv...@marvell.com; Power, Ciara
> ; Akhil Goyal 
> Subject: [PATCH v2 0/5] cryptodev: hide internal structures
> 
> Structures rte_cryptodev and rte_cryptodev_data are not
> supposed to be directly used by the application. These
> are made public as they are used by inline datapath
> public APIs.
> This patchset, creates a new rte_cryptodev_core.h file
> which helps in defining a data structure to hold datapath
> APIs in a flat array based on the device identifier which
> is filled by the PMD.
> 
> Similar series for ethdev and eventdev are also floated on ML.
> https://patchwork.dpdk.org/project/dpdk/list/?series=19428
> https://patchwork.dpdk.org/project/dpdk/list/?series=19405
> 
> changes in v2: align with the latest versions of above series.
> 
> Akhil Goyal (5):
>   cryptodev: separate out internal structures
>   cryptodev: allocate max space for internal qp array
>   cryptodev: move inline APIs into separate structure
>   cryptodev: update fast path APIs to use new flat array
>   cryptodev: move device specific structures
> 
>  drivers/crypto/aesni_mb/rte_aesni_mb_pmd.c|   1 -
>  drivers/crypto/ccp/ccp_dev.h  |   2 +-
>  drivers/crypto/cnxk/cn10k_ipsec.c |   2 +-
>  drivers/crypto/cnxk/cn9k_ipsec.c  |   2 +-
>  .../crypto/cnxk/cnxk_cryptodev_capabilities.c |   2 +-
>  drivers/crypto/cnxk/cnxk_cryptodev_sec.c  |   2 +-
>  drivers/crypto/nitrox/nitrox_sym_reqmgr.c |   2 +-
>  drivers/crypto/octeontx/otx_cryptodev.c   |   1 -
>  .../crypto/octeontx/otx_cryptodev_hw_access.c |   2 +-
>  .../crypto/octeontx/otx_cryptodev_hw_access.h |   2 +-
>  drivers/crypto/octeontx/otx_cryptodev_ops.h   |   2 +-
>  .../crypto/octeontx2/otx2_cryptodev_mbox.c|   2 +-
>  drivers/crypto/scheduler/scheduler_failover.c |   2 +-
>  .../crypto/scheduler/scheduler_multicore.c|   2 +-
>  .../scheduler/scheduler_pkt_size_distr.c  |   2 +-
>  .../crypto/scheduler/scheduler_roundrobin.c   |   2 +-
>  drivers/event/cnxk/cnxk_eventdev.h|   2 +-
>  drivers/event/dpaa/dpaa_eventdev.c|   2 +-
>  drivers/event/dpaa2/dpaa2_eventdev.c  |   2 +-
>  drivers/event/octeontx/ssovf_evdev.c  |   2 +-
>  .../event/octeontx2/otx2_evdev_crypto_adptr.c |   2 +-
>  lib/cryptodev/cryptodev_pmd.c |  51 +++
>  lib/cryptodev/cryptodev_pmd.h |  82 +++-
>  lib/cryptodev/meson.build |   4 +-
>  lib/cryptodev/rte_cryptodev.c |  50 ++-
>  lib/cryptodev/rte_cryptodev.h | 367 +++---
>  lib/cryptodev/rte_cryptodev_core.h|  62 +++
>  lib/cryptodev/version.map |   7 +-
>  28 files changed, 398 insertions(+), 265 deletions(-)
>  create mode 100644 lib/cryptodev/rte_cryptodev_core.h
> 
> --
> 2.25.1

Re: [dpdk-dev] [PATCH v1] eventdev/rx-adapter: add telemetry callbacks

[Jerin Jacob]

On Thu, Oct 7, 2021 at 6:27 PM Ganapati Kundapura
 wrote:
>
> Added telemetry callbacks to get Rx adapter stats, reset stats and
> to get rx queue config information.

rx -> Rx

Change the subject to eventdev/rx_adapter

>
> Signed-off-by: Ganapati Kundapura 
>
> diff --git a/lib/eventdev/rte_event_eth_rx_adapter.c 
> b/lib/eventdev/rte_event_eth_rx_adapter.c
> index 9ac976c..fa7191c 100644
> --- a/lib/eventdev/rte_event_eth_rx_adapter.c
> +++ b/lib/eventdev/rte_event_eth_rx_adapter.c
> @@ -23,6 +23,7 @@
>  #include "eventdev_pmd.h"
>  #include "rte_eventdev_trace.h"
>  #include "rte_event_eth_rx_adapter.h"
> +#include 

Move this to the above block where all <...h> header files are grouped.


>
>  #define BATCH_SIZE 32
>  #define BLOCK_CNT_THRESHOLD10
> @@ -2852,6 +2853,7 @@ rte_event_eth_rx_adapter_stats_get(uint8_t id,
>struct rte_event_eth_rx_adapter_stats *stats)
>  {
> struct rte_event_eth_rx_adapter *rx_adapter;
> +   struct rte_eth_event_enqueue_buffer *buf;
> struct rte_event_eth_rx_adapter_stats dev_stats_sum = { 0 };
> struct rte_event_eth_rx_adapter_stats dev_stats;
> struct rte_eventdev *dev;
> @@ -2887,8 +2889,11 @@ rte_event_eth_rx_adapter_stats_get(uint8_t id,
> if (rx_adapter->service_inited)
> *stats = rx_adapter->stats;
>
> +   buf = &rx_adapter->event_enqueue_buffer;
> stats->rx_packets += dev_stats_sum.rx_packets;
> stats->rx_enq_count += dev_stats_sum.rx_enq_count;
> +   stats->rx_event_buf_count = buf->count;
> +   stats->rx_event_buf_size = buf->events_size;
>
> return 0;
>  }
> @@ -3052,3 +3057,146 @@ rte_event_eth_rx_adapter_queue_conf_get(uint8_t id,
>
> return 0;
>  }
> +
> +#define RXA_ADD_DICT(stats, s) rte_tel_data_add_dict_u64(d, #s, stats.s)
> +
> +static int
> +handle_rxa_stats(const char *cmd __rte_unused,
> +const char *params,
> +struct rte_tel_data *d)
> +{
> +   uint8_t rx_adapter_id;
> +   struct rte_event_eth_rx_adapter_stats rx_adptr_stats;
> +
> +   if (params == NULL || strlen(params) == 0 || !isdigit(*params))
> +   return -1;
> +
> +   /* Get Rx adapter ID from parameter string */
> +   rx_adapter_id = atoi(params);
> +   RTE_EVENT_ETH_RX_ADAPTER_ID_VALID_OR_ERR_RET(rx_adapter_id, -EINVAL);
> +
> +   /* Get Rx adapter stats */
> +   if (rte_event_eth_rx_adapter_stats_get(rx_adapter_id,
> +  &rx_adptr_stats)) {
> +   RTE_EDEV_LOG_ERR("Failed to get Rx adapter stats\n");
> +   return -1;
> +   }
> +
> +   rte_tel_data_start_dict(d);
> +   rte_tel_data_add_dict_u64(d, "rx_adapter_id", rx_adapter_id);
> +   RXA_ADD_DICT(rx_adptr_stats, rx_packets);
> +   RXA_ADD_DICT(rx_adptr_stats, rx_poll_count);
> +   RXA_ADD_DICT(rx_adptr_stats, rx_dropped);
> +   RXA_ADD_DICT(rx_adptr_stats, rx_enq_retry);
> +   RXA_ADD_DICT(rx_adptr_stats, rx_event_buf_count);
> +   RXA_ADD_DICT(rx_adptr_stats, rx_event_buf_size);
> +   RXA_ADD_DICT(rx_adptr_stats, rx_enq_count);
> +   RXA_ADD_DICT(rx_adptr_stats, rx_enq_start_ts);
> +   RXA_ADD_DICT(rx_adptr_stats, rx_enq_block_cycles);
> +   RXA_ADD_DICT(rx_adptr_stats, rx_enq_end_ts);
> +   RXA_ADD_DICT(rx_adptr_stats, rx_intr_packets);
> +
> +   return 0;
> +}
> +
> +static int
> +handle_rxa_stats_reset(const char *cmd __rte_unused,
> +  const char *params,
> +  struct rte_tel_data *d __rte_unused)
> +{
> +   uint8_t rx_adapter_id;
> +
> +   if (params == NULL || strlen(params) == 0 || ~isdigit(*params))
> +   return -1;
> +
> +   /* Get Rx adapter ID from parameter string */
> +   rx_adapter_id = atoi(params);
> +   RTE_EVENT_ETH_RX_ADAPTER_ID_VALID_OR_ERR_RET(rx_adapter_id, -EINVAL);
> +
> +   /* Reset Rx adapter stats */
> +   if (rte_event_eth_rx_adapter_stats_reset(rx_adapter_id)) {
> +   RTE_EDEV_LOG_ERR("Failed to reset Rx adapter stats\n");
> +   return -1;
> +   }
> +
> +   return 0;
> +}
> +
> +static int
> +handle_rxa_get_queue_conf(const char *cmd __rte_unused,
> + const char *params,
> + struct rte_tel_data *d)
> +{
> +   uint8_t rx_adapter_id;
> +   uint16_t rx_queue_id;
> +   int eth_dev_id;
> +   char *token, *l_params;
> +   struct rte_event_eth_rx_adapter_queue_conf queue_conf;
> +
> +   if (params == NULL || strlen(params) == 0 || !isdigit(*params))
> +   return -1;
> +
> +   /* Get Rx adapter ID from parameter string */
> +   l_params = strdup(params);
> +   token = strtok(l_params, ",");
> +   rx_adapter_id = strtoul(token, NULL, 10);
> +   RTE_EVENT_ETH_RX_ADAPTER_ID_VALID_OR_ERR_RET(rx_adapter_id, -EINVAL);
> +
> +   token = strtok(NULL, ",");
> +

Re: [dpdk-dev] [PATCH v5 2/7] ethdev: allocate max space for internal queue array

[Ananyev, Konstantin]

> > At queue configure stage always allocate space for maximum possible
> > number (RTE_MAX_QUEUES_PER_PORT) of queue pointers.
> > That will allow 'fast' inline functions (eth_rx_burst, etc.) to refer
> > pointer to internal queue data without extra checking of current number
> > of configured queues.
> > That would help in future to hide rte_eth_dev and related structures.
> > It means that from now on, each ethdev port will always consume:
> > ((2*sizeof(uintptr_t))* RTE_MAX_QUEUES_PER_PORT)
> > bytes of memory for its queue pointers.
> > With RTE_MAX_QUEUES_PER_PORT==1024 (default value) it is 16KB per port.
> >
> > Signed-off-by: Konstantin Ananyev 
> > ---
> >  lib/ethdev/rte_ethdev.c | 36 +---
> >  1 file changed, 9 insertions(+), 27 deletions(-)
> >
> > diff --git a/lib/ethdev/rte_ethdev.c b/lib/ethdev/rte_ethdev.c
> > index ed37f8871b..c8abda6dd7 100644
> > --- a/lib/ethdev/rte_ethdev.c
> > +++ b/lib/ethdev/rte_ethdev.c
> > @@ -897,7 +897,8 @@ eth_dev_rx_queue_config(struct rte_eth_dev *dev, 
> > uint16_t nb_queues)
> >
> > if (dev->data->rx_queues == NULL && nb_queues != 0) { /* first time 
> > configuration */
> > dev->data->rx_queues = rte_zmalloc("ethdev->rx_queues",
> > -   sizeof(dev->data->rx_queues[0]) * nb_queues,
> > +   sizeof(dev->data->rx_queues[0]) *
> > +   RTE_MAX_QUEUES_PER_PORT,
> > RTE_CACHE_LINE_SIZE);
> 
> Looking at it I have few questions:
> 1. Why is nb_queues == 0 case kept as an exception? Yes,
>strictly speaking it is not the problem of the patch,
>DPDK will still segfault (non-debug build) if I
>allocate Tx queues only but call rte_eth_rx_burst().

eth_dev_rx_queue_config(.., nb_queues=0) is used in few places to clean-up 
things.

>After reading the patch description I thought that
>we're trying to address it.

We do, though I can't see how we can address it in this patch.
Though it is a good idea - I think I can add extra check in 
eth_dev_fp_ops_setup()
or around and setup RX function pointers only when dev->data->rx_queues != NULL.
Same for TX.

> 2. Why do we need to allocate memory dynamically?
>Can we just make rx_queues an array of appropriate size?

Pavan already asked same question.
My answer to him:
Yep we can, and yes it will simplify this peace of code.
The main reason I decided no to do this change now -
it will change layout of the_eth_dev_data structure.
In this series I tried to mininize(/avoid) changes in rte_eth_dev and 
rte_eth_dev_data,
as much as possible to avoid any unforeseen performance and functional impacts.
If we'll manage to make rte_eth_dev and rte_eth_dev_data private we can in 
future
consider that one and other changes in rte_eth_dev and rte_eth_dev_data layouts
without worrying about ABI breakage

>May be wasting 512K unconditionally is too much.
> 3. If wasting 512K is too much, I'd consider to move
>allocation to eth_dev_get(). If

Don't understand where 512KB came from.
each ethdev port will always consume:
((2*sizeof(uintptr_t))* RTE_MAX_QUEUES_PER_PORT)
bytes of memory for its queue pointers.
With RTE_MAX_QUEUES_PER_PORT==1024 (default value) it is 16KB per port.
 
> > if (dev->data->rx_queues == NULL) {
> > dev->data->nb_rx_queues = 0;
> > @@ -908,21 +909,11 @@ eth_dev_rx_queue_config(struct rte_eth_dev *dev, 
> > uint16_t nb_queues)
> >
> > rxq = dev->data->rx_queues;
> >
> > -   for (i = nb_queues; i < old_nb_queues; i++)
> > +   for (i = nb_queues; i < old_nb_queues; i++) {
> > (*dev->dev_ops->rx_queue_release)(rxq[i]);
> > -   rxq = rte_realloc(rxq, sizeof(rxq[0]) * nb_queues,
> > -   RTE_CACHE_LINE_SIZE);
> > -   if (rxq == NULL)
> > -   return -(ENOMEM);
> > -   if (nb_queues > old_nb_queues) {
> > -   uint16_t new_qs = nb_queues - old_nb_queues;
> > -
> > -   memset(rxq + old_nb_queues, 0,
> > -   sizeof(rxq[0]) * new_qs);
> > +   rxq[i] = NULL;
> 
> It looks like the patch should be rebased on top of
> next-net main because of queue release patches.
> 
> [snip]

Re: [dpdk-dev] [dpdk-stable] [PATCH v2 2/2] net/i40e: fix risk in Rx descriptor read in scalar path

[Ferruh Yigit]

On 9/29/2021 4:29 PM, Honnappa Nagarahalli wrote:




On 9/15/2021 9:33 AM, Ruifeng Wang wrote:

Rx descriptor is 16B/32B in size. If the DD bit is set, it indicates
that the rest of the descriptor words have valid values. Hence, the
word containing DD bit must be read first before reading the rest of
the descriptor words.

Since the entire descriptor is not read atomically, on relaxed memory
ordered systems like Aarch64, read of the word containing DD field
could be reordered after read of other words.

Read barrier is inserted between read of the word with DD field and
read of other words. The barrier ensures that the fetched data is
correct.

Testpmd single core test showed no performance drop on x86 or N1SDP.
On ThunderX2, 22% performance regression was observed.



Is 22% performance drop value correct? That is a big drop, is it acceptable?

Agree, it is a big drop. Fixing it will require using the barrier less 
frequently. For ex: read 4 descriptors (4 words containing the DD bits) before 
using the barrier.



Is this performance drop valid for all Arm scalar datapath, or is it specific to
ThunderX2?

This is specific to ThunderX2. N1 CPU does not see any impact. A72 is not 
tested. Considering that the ThunderXx line of CPUs are not in further 
development, and it is scalar path, I would not suggest to make further changes 
to the code.

It would be good to test this on Kunpeng servers and get some feedback.


Hi Connor, Yisen, Lijun,

Can you please check this patch? I don't know if you are using i40e nic
on your platform but if you do can you please test it?

Overall this patch cause a big performance drop on Arm for i40e, I just
want to be sure this is not impacting any user negatively.






Fixes: 7b0cf70135d1 ("net/i40e: support ARM platform")
Cc: sta...@dpdk.org

Signed-off-by: Ruifeng Wang 
Reviewed-by: Honnappa Nagarahalli 

Re: [dpdk-dev] [PATCH] doc/windows_gsg: update section on driver installation

[Kadam, Pallavi]

On 10/9/2021 1:20 PM, Dmitry Kozlyuk wrote:

2021-10-09 12:38 (UTC-0700), Kadam, Pallavi:

On 10/8/2021 2:11 PM, Dmitry Kozlyuk wrote:

[...]
+virt2phys
+~
   
-1. From Device Manager, Action menu, select "Add legacy hardware".

-2. It will launch the "Add Hardware Wizard". Click "Next".
-3. Select second option "Install the hardware that I manually select
-   from a list (Advanced)".
-4. On the next screen, "Kernel bypass" will be shown as a device class.
-5. Select it, and click "Next".
-6. The previously installed drivers will now be installed for the
-   "Virtual to physical address translator" device.
+Access to physical addresses is provided by a kernel-mode driver, virt2phys.
+It is mandatory for allocating physically-contiguous memory which is required
+by hardware PMDs.

Should we add specific link to virt2phys README in this section as well?

https://git.dpdk.org/dpdk-kmods/tree/windows/README.rst

This link is given above and I consider it general information for all
drivers. However, reading it carefully, I think docs in dpdk-kmods could be
restructured as well:
* Keep windows/README.rst unchanged, except for Device Manager steps,
   which are actually specific to virt2phys.
* Create windows/virt2phys/README.rst with instructions above.
* Change windows/netuio/README.rst: remove the bulk of its content, as it
   repeats generic instructions, keep netuio-specific Device Manager part.
Then a specific link for virt2phys here would make sense
and there will be even less duplication in dpdk-kmods.
If no one objects, I'll do this in v2.
Thanks!

Sounds good. Thanks, Dmitry!

[dpdk-dev] [PATCH v5 0/5] ethdev: introduce shared Rx queue

[Xueming Li]

In current DPDK framework, all Rx queues is pre-loaded with mbufs for
incoming packets. When number of representors scale out in a switch
domain, the memory consumption became significant. Further more,
polling all ports leads to high cache miss, high latency and low
throughputs.

This patch introduces shared Rx queue. PF and representors with same
configuration in same switch domain could share Rx queue set by
specifying shared Rx queue offloading flag and sharing group.

All ports that Shared Rx queue actually shares One Rx queue and only
pre-load mbufs to one Rx queue, memory is saved.

Polling any queue using same shared Rx queue receives packets from all
member ports. Source port is identified by mbuf->port.

Multiple groups is supported by group ID. Port queue number in a shared
group should be identical. Queue index is 1:1 mapped in shared group.
An example of two share groups:
 Group0, 4 shared Rx queues per member port: PF, repr0, repr1
 Group1, 2 shared Rx queues per member port: repr2, repr3, ... repr127
 Poll first port for each group:
  core  portqueue
  0 0   0
  1 0   1
  2 0   2
  3 0   3
  4 2   0
  5 2   1

Shared Rx queue must be polled on single thread or core. If both PF0 and
representor0 joined same share group, can't poll pf0rxq0 on core1 and
rep0rxq0 on core2. Actually, polling one port within share group is
sufficient since polling any port in group will return packets for any
port in group.

There was some discussion to aggregate member ports in same group into a
dummy port, several ways to achieve it. Since it optional, need to collect
more feedback and requirement from user, make better decision later.

v1:
  - initial version
v2:
  - add testpmd patches
v3:
  - change common forwarding api to macro for performance, thanks Jerin.
  - save global variable accessed in forwarding to flowstream to minimize
cache miss
  - combined patches for each forwarding engine
  - support multiple groups in testpmd "--share-rxq" parameter
  - new api to aggregate shared rxq group
v4:
  - spelling fixes
  - remove shared-rxq support for all forwarding engines
  - add dedicate shared-rxq forwarding engine
v5:
 - fix grammars
 - remove aggregate api and leave it for later discussion
 - add release notes
 - add deployment example

Xueming Li (5):
  ethdev: introduce shared Rx queue
  app/testpmd: new parameter to enable shared Rx queue
  app/testpmd: dump port info for shared Rx queue
  app/testpmd: force shared Rx queue polled on same core
  app/testpmd: add forwarding engine for shared Rx queue

 app/test-pmd/config.c | 102 +++-
 app/test-pmd/meson.build  |   1 +
 app/test-pmd/parameters.c |  13 ++
 app/test-pmd/shared_rxq_fwd.c | 148 ++
 app/test-pmd/testpmd.c|  23 ++-
 app/test-pmd/testpmd.h|   5 +
 app/test-pmd/util.c   |   3 +
 doc/guides/nics/features.rst  |  11 ++
 doc/guides/nics/features/default.ini  |   1 +
 .../prog_guide/switch_representation.rst  |  10 ++
 doc/guides/rel_notes/release_21_11.rst|   4 +
 doc/guides/testpmd_app_ug/run_app.rst |   8 +
 doc/guides/testpmd_app_ug/testpmd_funcs.rst   |   5 +-
 lib/ethdev/rte_ethdev.c   |   1 +
 lib/ethdev/rte_ethdev.h   |   7 +
 15 files changed, 339 insertions(+), 3 deletions(-)
 create mode 100644 app/test-pmd/shared_rxq_fwd.c

-- 
2.33.0

[dpdk-dev] [PATCH v5 3/5] app/testpmd: dump port info for shared Rx queue

[Xueming Li]

In case of shared Rx queue, polling any member port returns mbufs for
all members. This patch dumps mbuf->port for each packet.

Signed-off-by: Xueming Li 
---
 app/test-pmd/util.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/app/test-pmd/util.c b/app/test-pmd/util.c
index 51506e49404..e98f136d5ed 100644
--- a/app/test-pmd/util.c
+++ b/app/test-pmd/util.c
@@ -100,6 +100,9 @@ dump_pkt_burst(uint16_t port_id, uint16_t queue, struct 
rte_mbuf *pkts[],
struct rte_flow_restore_info info = { 0, };
 
mb = pkts[i];
+   if (rxq_share > 0)
+   MKDUMPSTR(print_buf, buf_size, cur_len, "port %u, ",
+ mb->port);
eth_hdr = rte_pktmbuf_read(mb, 0, sizeof(_eth_hdr), &_eth_hdr);
eth_type = RTE_BE_TO_CPU_16(eth_hdr->ether_type);
packet_type = mb->packet_type;
-- 
2.33.0

[dpdk-dev] [PATCH v5 1/5] ethdev: introduce shared Rx queue

[Xueming Li]

In current DPDK framework, each Rx queue is pre-loaded with mbufs for
incoming packets. For some PMDs, when number of representors scale out
in a switch domain, the memory consumption became significant. Polling
all ports also leads to high cache miss, high latency and low
throughput.

This patch introduce shared Rx queue. Ports with same configuration in
a switch domain could share Rx queue set by specifying sharing group.
Polling any queue using same shared Rx queue receives packets from all
member ports. Source port is identified by mbuf->port.

Port queue number in a shared group should be identical. Queue index is
1:1 mapped in shared group.

Share Rx queue must be polled on single thread or core.

Multiple groups are supported by group ID.

Example grouping and polling model to reflect service priority:
 Group0, 2 shared Rx queues per port: PF, rep0, rep1
 Group1, 1 shared Rx queue per port: rep2, rep3, ... rep127
 Core0: poll PF queue0
 Core1: poll PF queue1
 Core2: poll rep2 queue0

Signed-off-by: Xueming Li 
Cc: Jerin Jacob 
---
Rx queue object could be used as shared Rx queue object, it's important
to clear all queue control callback api that using queue object:
  https://mails.dpdk.org/archives/dev/2021-July/215574.html
---
 doc/guides/nics/features.rst| 11 +++
 doc/guides/nics/features/default.ini|  1 +
 doc/guides/prog_guide/switch_representation.rst | 10 ++
 doc/guides/rel_notes/release_21_11.rst  |  4 
 lib/ethdev/rte_ethdev.c |  1 +
 lib/ethdev/rte_ethdev.h |  7 +++
 6 files changed, 34 insertions(+)

diff --git a/doc/guides/nics/features.rst b/doc/guides/nics/features.rst
index 4fce8cd1c97..69bc1d5719c 100644
--- a/doc/guides/nics/features.rst
+++ b/doc/guides/nics/features.rst
@@ -626,6 +626,17 @@ Supports inner packet L4 checksum.
   ``tx_offload_capa,tx_queue_offload_capa:DEV_TX_OFFLOAD_OUTER_UDP_CKSUM``.
 
 
+.. _nic_features_shared_rx_queue:
+
+Shared Rx queue
+---
+
+Supports shared Rx queue for ports in same switch domain.
+
+* **[uses] rte_eth_rxconf,rte_eth_rxmode**: 
``offloads:RTE_ETH_RX_OFFLOAD_SHARED_RXQ``.
+* **[provides] mbuf**: ``mbuf.port``.
+
+
 .. _nic_features_packet_type_parsing:
 
 Packet type parsing
diff --git a/doc/guides/nics/features/default.ini 
b/doc/guides/nics/features/default.ini
index 754184ddd4d..ebeb4c18512 100644
--- a/doc/guides/nics/features/default.ini
+++ b/doc/guides/nics/features/default.ini
@@ -19,6 +19,7 @@ Free Tx mbuf on demand =
 Queue start/stop =
 Runtime Rx queue setup =
 Runtime Tx queue setup =
+Shared Rx queue  =
 Burst mode info  =
 Power mgmt address monitor =
 MTU update   =
diff --git a/doc/guides/prog_guide/switch_representation.rst 
b/doc/guides/prog_guide/switch_representation.rst
index ff6aa91c806..47205f5f1cc 100644
--- a/doc/guides/prog_guide/switch_representation.rst
+++ b/doc/guides/prog_guide/switch_representation.rst
@@ -123,6 +123,16 @@ thought as a software "patch panel" front-end for 
applications.
 .. [1] `Ethernet switch device driver model (switchdev)
`_
 
+- For some PMDs, memory usage of representors is huge when number of
+  representor grows, mbufs are allocated for each descriptor of Rx queue.
+  Polling large number of ports brings more CPU load, cache miss and
+  latency. Shared Rx queue can be used to share Rx queue between PF and
+  representors among same switch. ``RTE_ETH_RX_OFFLOAD_SHARED_RXQ`` is
+  present in Rx offloading capability of device info. Setting the
+  offloading flag in device Rx mode or Rx queue configuration to enable
+  shared Rx queue. Polling any member port of the shared Rx queue can return
+  packets of all ports in the group, port ID is saved in ``mbuf.port``.
+
 Basic SR-IOV
 
 
diff --git a/doc/guides/rel_notes/release_21_11.rst 
b/doc/guides/rel_notes/release_21_11.rst
index c0a7f755189..aa6b5e2e9c5 100644
--- a/doc/guides/rel_notes/release_21_11.rst
+++ b/doc/guides/rel_notes/release_21_11.rst
@@ -134,6 +134,10 @@ New Features
   * Added tests to validate packets hard expiry.
   * Added tests to verify tunnel header verification in IPsec inbound.
 
+* **Added ethdev shared Rx queue support. **
+
+  * Added new Rx queue offloading capability flag.
+  * Added share group to Rx queue configuration.
 
 Removed Items
 -
diff --git a/lib/ethdev/rte_ethdev.c b/lib/ethdev/rte_ethdev.c
index fb69f6ea8d1..d78b50e1fa7 100644
--- a/lib/ethdev/rte_ethdev.c
+++ b/lib/ethdev/rte_ethdev.c
@@ -127,6 +127,7 @@ static const struct {
RTE_RX_OFFLOAD_BIT2STR(OUTER_UDP_CKSUM),
RTE_RX_OFFLOAD_BIT2STR(RSS_HASH),
RTE_ETH_RX_OFFLOAD_BIT2STR(BUFFER_SPLIT),
+   RTE_ETH_RX_OFFLOAD_BIT2STR(SHARED_RXQ),
 };
 
 #undef RTE_RX_OFFLOAD_BIT2STR
diff --git a/lib/ethdev/rte_ethdev.h b/lib/ethdev/rte_ethdev.h
index 39d2cd612cb..b94f7ba5a3f 100644
--

Re: [dpdk-dev] RHEL7 failures

[Thinh Tran]

Hi Linconln,

I have this patch set 100042  - build/ppc: enable build support for Power10
https://patches.dpdk.org/patch/100042

Thanks,
Thinh Tran



On 10/11/2021 9:42 AM, Lincoln Lavoie wrote:
Hi Thinh, The CI won't automatically go back and retest failures after a 
patch was applied.  This is a corner case, where a bad change was merged 
and affected all patches downstream of that merge.  We can retrigger 
specific patches, just ZjQcmQRYFpfptBannerStart

This Message Is From an External Sender
This message came from outside your organization.
ZjQcmQRYFpfptBannerEnd
Hi Thinh,

The CI won't automatically go back and retest failures after a patch was 
applied.  This is a corner case, where a bad change was merged and 
affected all patches downstream of that merge.  We can retrigger 
specific patches, just let us know the patch number or the URI of the 
patch(es).


Cheers,
Lincoln

On Mon, Oct 11, 2021 at 10:39 AM Thinh Tran > wrote:



Hi,
I'm wondering how or when the CI would re-pick up the patches those
were
failed due to this issue?

Regards,
Thinh Tran

On 10/6/2021 10:40 PM, Zhang, Qi Z wrote:
 > Hi Ferruh, David and Aman:
 >
 >       Sorry for late response due to PRC holiday.
 >       I have just tested the patch and there is no issue be
detected for functions
 >       Thank you so much for the help!
 >
 > Regards
 > Qi
 >
 >> -Original Message-
 >> From: Yigit, Ferruh mailto:ferruh.yi...@intel.com>>
 >> Sent: Wednesday, October 6, 2021 5:46 AM
 >> To: Lincoln Lavoie mailto:lylav...@iol.unh.edu>>; dev mailto:dev@dpdk.org>>; Yang, Qiming
 >> mailto:qiming.y...@intel.com>>; Zhang,
Qi Z mailto:qi.z.zh...@intel.com>>
 >> Cc: c...@dpdk.org ; Aaron Conole
mailto:acon...@redhat.com>>; dpdklab
 >> mailto:dpdk...@iol.unh.edu>>; Singh, Aman
Deep mailto:aman.deep.si...@intel.com>>;
 >> David Marchand mailto:david.march...@redhat.com>>
 >> Subject: Re: [dpdk-dev] RHEL7 failures
 >>
 >> On 10/5/2021 8:09 PM, Lincoln Lavoie wrote:
 >>> Hello Qiming and Qi,
 >>>
 >>> The CI is picking up failures when building on RHEL7, where
warnings
 >>> are being treated as errors.  This looks like something has been
 >>> merged into the mainline, as it's failing across all patches.
 >>>
 >>> Here is the specific failure:
 >>>
 >>> ./drivers/net/ice/base/ice_parser_rt.c: In function '_hv_bit_sel':
 >>> ../drivers/net/ice/base/ice_parser_rt.c:201:2: error: dereferencing
 >>> type-punned pointer will break strict-aliasing rules
 >>> [-Werror=strict-aliasing]
 >>>     d64 = *(u64 *)&b[0];
 >>>     ^
 >>> ../drivers/net/ice/base/ice_parser_rt.c: In function
'_reg_bit_sel':
 >>> ../drivers/net/ice/base/ice_parser_rt.c:458:2: error: dereferencing
 >>> type-punned pointer will break strict-aliasing rules
 >>> [-Werror=strict-aliasing]
 >>>     d32 = *(u32 *)&v[0];
 >>>     ^
 >>> cc1: all warnings being treated as errors
 >>>
 >>> You can download a full set of logs from here (for a failing run):
 >>> https://lab.dpdk.org/results/dashboard/patchsets/19162/

 >>>
 >>
 >> Issue was reported by David, Aman sent the fix [1] and it is
already merged by
 >> David [2], it should be fixed now, can you please double check?
 >>
 >> [1]
 >>
https://patches.dpdk.org/project/dpdk/patch/20211005115754.34117-1-aman

 >> .deep.si...@intel.com/ 
 >>
 >> [2]
 >>
https://git.dpdk.org/dpdk/commit/?id=16b809d144dc2df7f31695b5abc64a809

 >> 021b154



--
*Lincoln Lavoie*
Principal Engineer, Broadband Technologies
21 Madbury Rd., Ste. 100, Durham, NH 03824
lylav...@iol.unh.edu 
https://www.iol.unh.edu 
+1-603-674-2755 (m)

[dpdk-dev] [PATCH v5 4/5] app/testpmd: force shared Rx queue polled on same core

[Xueming Li]

Shared rxqs shares one set rx queue of groups zero. Shared Rx queue must
must be polled from one core.

Checks and stops forwarding if shared rxq being scheduled on multiple
cores.

Signed-off-by: Xueming Li 
---
 app/test-pmd/config.c  | 96 ++
 app/test-pmd/testpmd.c |  4 +-
 app/test-pmd/testpmd.h |  2 +
 3 files changed, 101 insertions(+), 1 deletion(-)

diff --git a/app/test-pmd/config.c b/app/test-pmd/config.c
index 6c7f9dee065..4ab500569ee 100644
--- a/app/test-pmd/config.c
+++ b/app/test-pmd/config.c
@@ -2885,6 +2885,102 @@ port_rss_hash_key_update(portid_t port_id, char 
rss_type[], uint8_t *hash_key,
}
 }
 
+/*
+ * Check whether a shared rxq scheduled on other lcores.
+ */
+static bool
+fwd_stream_on_other_lcores(uint16_t domain_id, portid_t src_port,
+  queueid_t src_rxq, lcoreid_t src_lc,
+  uint32_t shared_group)
+{
+   streamid_t sm_id;
+   streamid_t nb_fs_per_lcore;
+   lcoreid_t  nb_fc;
+   lcoreid_t  lc_id;
+   struct fwd_stream *fs;
+   struct rte_port *port;
+   struct rte_eth_rxconf *rxq_conf;
+
+   nb_fc = cur_fwd_config.nb_fwd_lcores;
+   for (lc_id = src_lc + 1; lc_id < nb_fc; lc_id++) {
+   sm_id = fwd_lcores[lc_id]->stream_idx;
+   nb_fs_per_lcore = fwd_lcores[lc_id]->stream_nb;
+   for (; sm_id < fwd_lcores[lc_id]->stream_idx + nb_fs_per_lcore;
+sm_id++) {
+   fs = fwd_streams[sm_id];
+   port = &ports[fs->rx_port];
+   rxq_conf = &port->rx_conf[fs->rx_queue];
+   if ((rxq_conf->offloads & RTE_ETH_RX_OFFLOAD_SHARED_RXQ)
+   == 0)
+   /* Not shared rxq. */
+   continue;
+   if (domain_id != port->dev_info.switch_info.domain_id)
+   continue;
+   if (fs->rx_queue != src_rxq)
+   continue;
+   if (rxq_conf->shared_group != shared_group)
+   continue;
+   printf("Shared Rx queue group %u can't be scheduled on 
different cores:\n",
+  shared_group);
+   printf("  lcore %hhu Port %hu queue %hu\n",
+  src_lc, src_port, src_rxq);
+   printf("  lcore %hhu Port %hu queue %hu\n",
+  lc_id, fs->rx_port, fs->rx_queue);
+   printf("  please use --nb-cores=%hu to limit forwarding 
cores\n",
+  nb_rxq);
+   return true;
+   }
+   }
+   return false;
+}
+
+/*
+ * Check shared rxq configuration.
+ *
+ * Shared group must not being scheduled on different core.
+ */
+bool
+pkt_fwd_shared_rxq_check(void)
+{
+   streamid_t sm_id;
+   streamid_t nb_fs_per_lcore;
+   lcoreid_t  nb_fc;
+   lcoreid_t  lc_id;
+   struct fwd_stream *fs;
+   uint16_t domain_id;
+   struct rte_port *port;
+   struct rte_eth_rxconf *rxq_conf;
+
+   nb_fc = cur_fwd_config.nb_fwd_lcores;
+   /*
+* Check streams on each core, make sure the same switch domain +
+* group + queue doesn't get scheduled on other cores.
+*/
+   for (lc_id = 0; lc_id < nb_fc; lc_id++) {
+   sm_id = fwd_lcores[lc_id]->stream_idx;
+   nb_fs_per_lcore = fwd_lcores[lc_id]->stream_nb;
+   for (; sm_id < fwd_lcores[lc_id]->stream_idx + nb_fs_per_lcore;
+sm_id++) {
+   fs = fwd_streams[sm_id];
+   /* Update lcore info stream being scheduled. */
+   fs->lcore = fwd_lcores[lc_id];
+   port = &ports[fs->rx_port];
+   rxq_conf = &port->rx_conf[fs->rx_queue];
+   if ((rxq_conf->offloads & RTE_ETH_RX_OFFLOAD_SHARED_RXQ)
+   == 0)
+   /* Not shared rxq. */
+   continue;
+   /* Check shared rxq not scheduled on remaining cores. */
+   domain_id = port->dev_info.switch_info.domain_id;
+   if (fwd_stream_on_other_lcores(domain_id, fs->rx_port,
+  fs->rx_queue, lc_id,
+  rxq_conf->shared_group))
+   return false;
+   }
+   }
+   return true;
+}
+
 /*
  * Setup forwarding configuration for each logical core.
  */
diff --git a/app/test-pmd/testpmd.c b/app/test-pmd/testpmd.c
index 417e92ade11..cab4b36b046 100644
--- a/app/test-pmd/testpmd.c
+++ b/app/test-pmd/testpmd.c
@@ -2241,10 +2241,12 @@ start_packet_forwarding(int with_tx_first)
 
  

[dpdk-dev] [PATCH v5 5/5] app/testpmd: add forwarding engine for shared Rx queue

[Xueming Li]

To support shared Rx queue, this patch introduces dedicate forwarding
engine. The engine groups received packets by mbuf->port into sub-group,
updates stream statistics and simply frees packets.

Signed-off-by: Xueming Li 
---
 app/test-pmd/meson.build|   1 +
 app/test-pmd/shared_rxq_fwd.c   | 148 
 app/test-pmd/testpmd.c  |   1 +
 app/test-pmd/testpmd.h  |   1 +
 doc/guides/testpmd_app_ug/run_app.rst   |   1 +
 doc/guides/testpmd_app_ug/testpmd_funcs.rst |   5 +-
 6 files changed, 156 insertions(+), 1 deletion(-)
 create mode 100644 app/test-pmd/shared_rxq_fwd.c

diff --git a/app/test-pmd/meson.build b/app/test-pmd/meson.build
index 98f3289bdfa..07042e45b12 100644
--- a/app/test-pmd/meson.build
+++ b/app/test-pmd/meson.build
@@ -21,6 +21,7 @@ sources = files(
 'noisy_vnf.c',
 'parameters.c',
 'rxonly.c',
+'shared_rxq_fwd.c',
 'testpmd.c',
 'txonly.c',
 'util.c',
diff --git a/app/test-pmd/shared_rxq_fwd.c b/app/test-pmd/shared_rxq_fwd.c
new file mode 100644
index 000..4e262b99bc7
--- /dev/null
+++ b/app/test-pmd/shared_rxq_fwd.c
@@ -0,0 +1,148 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright (c) 2021 NVIDIA Corporation & Affiliates
+ */
+
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+
+#include 
+#include 
+
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+
+#include "testpmd.h"
+
+/*
+ * Rx only sub-burst forwarding.
+ */
+static void
+forward_rx_only(uint16_t nb_rx, struct rte_mbuf **pkts_burst)
+{
+   rte_pktmbuf_free_bulk(pkts_burst, nb_rx);
+}
+
+/**
+ * Get packet source stream by source port and queue.
+ * All streams of same shared Rx queue locates on same core.
+ */
+static struct fwd_stream *
+forward_stream_get(struct fwd_stream *fs, uint16_t port)
+{
+   streamid_t sm_id;
+   struct fwd_lcore *fc;
+   struct fwd_stream **fsm;
+   streamid_t nb_fs;
+
+   fc = fs->lcore;
+   fsm = &fwd_streams[fc->stream_idx];
+   nb_fs = fc->stream_nb;
+   for (sm_id = 0; sm_id < nb_fs; sm_id++) {
+   if (fsm[sm_id]->rx_port == port &&
+   fsm[sm_id]->rx_queue == fs->rx_queue)
+   return fsm[sm_id];
+   }
+   return NULL;
+}
+
+/**
+ * Forward packet by source port and queue.
+ */
+static void
+forward_sub_burst(struct fwd_stream *src_fs, uint16_t port, uint16_t nb_rx,
+ struct rte_mbuf **pkts)
+{
+   struct fwd_stream *fs = forward_stream_get(src_fs, port);
+
+   if (fs != NULL) {
+   fs->rx_packets += nb_rx;
+   forward_rx_only(nb_rx, pkts);
+   } else {
+   /* Source stream not found, drop all packets. */
+   src_fs->fwd_dropped += nb_rx;
+   while (nb_rx > 0)
+   rte_pktmbuf_free(pkts[--nb_rx]);
+   }
+}
+
+/**
+ * Forward packets from shared Rx queue.
+ *
+ * Source port of packets are identified by mbuf->port.
+ */
+static void
+forward_shared_rxq(struct fwd_stream *fs, uint16_t nb_rx,
+  struct rte_mbuf **pkts_burst)
+{
+   uint16_t i, nb_sub_burst, port, last_port;
+
+   nb_sub_burst = 0;
+   last_port = pkts_burst[0]->port;
+   /* Locate sub-burst according to mbuf->port. */
+   for (i = 0; i < nb_rx - 1; ++i) {
+   rte_prefetch0(pkts_burst[i + 1]);
+   port = pkts_burst[i]->port;
+   if (i > 0 && last_port != port) {
+   /* Forward packets with same source port. */
+   forward_sub_burst(fs, last_port, nb_sub_burst,
+ &pkts_burst[i - nb_sub_burst]);
+   nb_sub_burst = 0;
+   last_port = port;
+   }
+   nb_sub_burst++;
+   }
+   /* Last sub-burst. */
+   nb_sub_burst++;
+   forward_sub_burst(fs, last_port, nb_sub_burst,
+ &pkts_burst[nb_rx - nb_sub_burst]);
+}
+
+static void
+shared_rxq_fwd(struct fwd_stream *fs)
+{
+   struct rte_mbuf *pkts_burst[nb_pkt_per_burst];
+   uint16_t nb_rx;
+   uint64_t start_tsc = 0;
+
+   get_start_cycles(&start_tsc);
+   nb_rx = rte_eth_rx_burst(fs->rx_port, fs->rx_queue, pkts_burst,
+nb_pkt_per_burst);
+   inc_rx_burst_stats(fs, nb_rx);
+   if (unlikely(nb_rx == 0))
+   return;
+   forward_shared_rxq(fs, nb_rx, pkts_burst);
+   get_end_cycles(fs, start_tsc);
+}
+
+struct fwd_engine shared_rxq_engine = {
+   .fwd_mode_name  = "shared_rxq",
+   .port_fwd_begin = NULL,
+   .port_fwd_end   = NULL,
+   .packet_fwd = shared_rxq_f

[dpdk-dev] [PATCH v5 2/5] app/testpmd: new parameter to enable shared Rx queue

[Xueming Li]

Adds "--rxq-share" parameter to enable shared rxq for each rxq.

Default shared rxq group 0 is used, Rx queues in same switch domain
shares same rxq according to queue index.

Shared Rx queue is enabled only if device support offloading flag
RTE_ETH_RX_OFFLOAD_SHARED_RXQ.

Signed-off-by: Xueming Li 
---
 app/test-pmd/config.c |  6 +-
 app/test-pmd/parameters.c | 13 +
 app/test-pmd/testpmd.c| 18 ++
 app/test-pmd/testpmd.h|  2 ++
 doc/guides/testpmd_app_ug/run_app.rst |  7 +++
 5 files changed, 45 insertions(+), 1 deletion(-)

diff --git a/app/test-pmd/config.c b/app/test-pmd/config.c
index 9c66329e96e..6c7f9dee065 100644
--- a/app/test-pmd/config.c
+++ b/app/test-pmd/config.c
@@ -2709,7 +2709,11 @@ rxtx_config_display(void)
printf("  RX threshold registers: pthresh=%d 
hthresh=%d "
" wthresh=%d\n",
pthresh_tmp, hthresh_tmp, wthresh_tmp);
-   printf("  RX Offloads=0x%"PRIx64"\n", offloads_tmp);
+   printf("  RX Offloads=0x%"PRIx64, offloads_tmp);
+   if (rx_conf->offloads & RTE_ETH_RX_OFFLOAD_SHARED_RXQ)
+   printf(" share group=%u",
+  rx_conf->shared_group);
+   printf("\n");
}
 
/* per tx queue config only for first queue to be less verbose 
*/
diff --git a/app/test-pmd/parameters.c b/app/test-pmd/parameters.c
index 3f94a82e321..30dae326310 100644
--- a/app/test-pmd/parameters.c
+++ b/app/test-pmd/parameters.c
@@ -167,6 +167,7 @@ usage(char* progname)
printf("  --tx-ip=src,dst: IP addresses in Tx-only mode\n");
printf("  --tx-udp=src[,dst]: UDP ports in Tx-only mode\n");
printf("  --eth-link-speed: force link speed.\n");
+   printf("  --rxq-share: number of ports per shared rxq groups, defaults 
to MAX(1 group)\n");
printf("  --disable-link-check: disable check on link status when "
   "starting/stopping ports.\n");
printf("  --disable-device-start: do not automatically start port\n");
@@ -607,6 +608,7 @@ launch_args_parse(int argc, char** argv)
{ "rxpkts", 1, 0, 0 },
{ "txpkts", 1, 0, 0 },
{ "txonly-multi-flow",  0, 0, 0 },
+   { "rxq-share",  2, 0, 0 },
{ "eth-link-speed", 1, 0, 0 },
{ "disable-link-check", 0, 0, 0 },
{ "disable-device-start",   0, 0, 0 },
@@ -1271,6 +1273,17 @@ launch_args_parse(int argc, char** argv)
}
if (!strcmp(lgopts[opt_idx].name, "txonly-multi-flow"))
txonly_multi_flow = 1;
+   if (!strcmp(lgopts[opt_idx].name, "rxq-share")) {
+   if (optarg == NULL) {
+   rxq_share = UINT32_MAX;
+   } else {
+   n = atoi(optarg);
+   if (n >= 0)
+   rxq_share = (uint32_t)n;
+   else
+   rte_exit(EXIT_FAILURE, 
"rxq-share must be >= 0\n");
+   }
+   }
if (!strcmp(lgopts[opt_idx].name, "no-flush-rx"))
no_flush_rx = 1;
if (!strcmp(lgopts[opt_idx].name, "eth-link-speed")) {
diff --git a/app/test-pmd/testpmd.c b/app/test-pmd/testpmd.c
index 97ae52e17ec..417e92ade11 100644
--- a/app/test-pmd/testpmd.c
+++ b/app/test-pmd/testpmd.c
@@ -498,6 +498,11 @@ uint8_t record_core_cycles;
  */
 uint8_t record_burst_stats;
 
+/*
+ * Number of ports per shared Rx queue group, 0 disable.
+ */
+uint32_t rxq_share;
+
 unsigned int num_sockets = 0;
 unsigned int socket_ids[RTE_MAX_NUMA_NODES];
 
@@ -1506,6 +1511,11 @@ init_config_port_offloads(portid_t pid, uint32_t 
socket_id)
port->dev_conf.txmode.offloads &=
~DEV_TX_OFFLOAD_MBUF_FAST_FREE;
 
+   if (rxq_share > 0 &&
+   (port->dev_info.rx_offload_capa & RTE_ETH_RX_OFFLOAD_SHARED_RXQ))
+   port->dev_conf.rxmode.offloads |=
+   RTE_ETH_RX_OFFLOAD_SHARED_RXQ;
+
/* Apply Rx offloads configuration */
for (i = 0; i < port->dev_info.max_rx_queues; i++)
port->rx_conf[i].offloads = port->dev_conf.rxmode.offloads;
@@ -3401,6 +3411,14 @@ rxtx_port_config(struct rte_port *port)
for (qid = 0; qid < nb_rxq; qid++) {
offloads = port->rx_conf[qid].offloads;
port->rx_conf[qid] = port->dev_info.default_

Re: [dpdk-dev] [PATCH 00/11] net/sfc: support per-queue stats on EF100

[Ferruh Yigit]

On 9/28/2021 12:29 PM, Andrew Rybchenko wrote:

Implement per-queue Rx and Tx statistics for EF100 in software.
Packets and bytes stats are collected by the driver.

Ivan Ilchenko (11):
   net/sfc: rename array of SW stats descriptions
   net/sfc: rename accumulative SW stats to total
   net/sfc: rename SW stats structures
   net/sfc: fix cleanup order of SW stats
   net/sfc: fix missing const of SW stats descriptions
   net/sfc: optimize getting number of SW stats
   net/sfc: prepare having no some SW stats on an adapter
   net/sfc: add toggle to disable total stat
   net/sfc: add support for SW stats groups
   net/sfc: collect per queue stats in EF100 Rx datapath
   net/sfc: collect per queue stats in EF100 Tx datapath



Series applied to dpdk-next-net/main, thanks.

Re: [dpdk-dev] [PATCH v5 4/7] ethdev: copy fast-path API into separate structure

[Ananyev, Konstantin]

> On 10/7/21 2:27 PM, Konstantin Ananyev wrote:
> > Copy public function pointers (rx_pkt_burst(), etc.) and related
> > pointers to internal data from rte_eth_dev structure into a
> > separate flat array. That array will remain in a public header.
> > The intention here is to make rte_eth_dev and related structures internal.
> > That should allow future possible changes to core eth_dev structures
> > to be transparent to the user and help to avoid ABI/API breakages.
> > The plan is to keep minimal part of data from rte_eth_dev public,
> > so we still can use inline functions for fast-path calls
> > (like rte_eth_rx_burst(), etc.) to avoid/minimize slowdown.
> > The whole idea beyond this new schema:
> > 1. PMDs keep to setup fast-path function pointers and related data
> >inside rte_eth_dev struct in the same way they did it before.
> > 2. Inside rte_eth_dev_start() and inside rte_eth_dev_probing_finish()
> >(for secondary process) we call eth_dev_fp_ops_setup, which
> >copies these function and data pointers into rte_eth_fp_ops[port_id].
> > 3. Inside rte_eth_dev_stop() and inside rte_eth_dev_release_port()
> >we call eth_dev_fp_ops_reset(), which resets rte_eth_fp_ops[port_id]
> >into some dummy values.
> > 4. fast-path ethdev API (rte_eth_rx_burst(), etc.) will use that new
> >flat array to call PMD specific functions.
> > That approach should allow us to make rte_eth_devices[] private
> > without introducing regression and help to avoid changes in drivers code.
> >
> > Signed-off-by: Konstantin Ananyev 
> 
> Overall LGTM, few nits below.
> 
> > ---
> >  lib/ethdev/ethdev_private.c  | 52 ++
> >  lib/ethdev/ethdev_private.h  |  7 +
> >  lib/ethdev/rte_ethdev.c  | 27 ++
> >  lib/ethdev/rte_ethdev_core.h | 55 
> >  4 files changed, 141 insertions(+)
> >
> > diff --git a/lib/ethdev/ethdev_private.c b/lib/ethdev/ethdev_private.c
> > index 012cf73ca2..3eeda6e9f9 100644
> > --- a/lib/ethdev/ethdev_private.c
> > +++ b/lib/ethdev/ethdev_private.c
> > @@ -174,3 +174,55 @@ rte_eth_devargs_parse_representor_ports(char *str, 
> > void *data)
> > RTE_LOG(ERR, EAL, "wrong representor format: %s\n", str);
> > return str == NULL ? -1 : 0;
> >  }
> > +
> > +static uint16_t
> > +dummy_eth_rx_burst(__rte_unused void *rxq,
> > +   __rte_unused struct rte_mbuf **rx_pkts,
> > +   __rte_unused uint16_t nb_pkts)
> > +{
> > +   RTE_ETHDEV_LOG(ERR, "rx_pkt_burst for unconfigured port\n");
> 
> May be "unconfigured" -> "stopped" ? Or "non-started" ?

Yes, it can be configured but not started.
So 'not started' seems like a better wording here.
Another option probably: 'not ready'.
What people think?

...

> 
> > +   rte_errno = ENOTSUP;
> > +   return 0;
> > +}
> > +
> > +struct rte_eth_fp_ops {
> > +
> > +   /**
> > +* Rx fast-path functions and related data.
> > +* 64-bit systems: occupies first 64B line
> > +*/
> 
> As I understand the above comment is for a group of below
> fields. If so, Doxygen annocation for member groups should
> be used.

Ok, and how to do it?

Re: [dpdk-dev] [EXT] Re: [PATCH v2 3/3] security: add reserved bitfields

[Akhil Goyal]

> 08/10/2021 22:45, Akhil Goyal:
> > In struct rte_security_ipsec_sa_options, for every new option
> > added, there is an ABI breakage, to avoid, a reserved_opts
> > bitfield is added to for the remaining bits available in the
> > structure.
> > Now for every new sa option, these reserved_opts can be reduced
> > and new option can be added.
> 
> How do you make sure this field is initialized to 0?
> 
Struct rte_security_ipsec_xform Is part of rte_security_capability as well
As a configuration structure in session create.
User, should ensure that if a device support that option(in capability), then
only these options will take into effect or else it will be don't care for the 
PMD.
The initial values of capabilities are set by PMD statically based on the 
features
that it support.
So if someone sets a bit in reserved_opts, it will work only if PMD support it
And sets the corresponding field in capabilities.
But yes, if a new field is added in future, and user sets the reserved_opts by 
mistake
And the PMD supports that feature as well, then that feature will be enabled.
This may or may not create issue depending on the feature which is enabled.

Should I add a note in the comments to clarify that reserved_opts should be set 
as 0
And future releases may change this without notice(But reserved in itself 
suggest that)?
Adding an explicit check in session_create does not make sense to me.
What do you suggest?

Regards,
Akhil

Re: [dpdk-dev] [PATCH v5 5/7] ethdev: make fast-path functions to use new flat array

[Andrew Rybchenko]

On 10/11/21 6:47 PM, Ananyev, Konstantin wrote:




On 10/7/21 2:27 PM, Konstantin Ananyev wrote:

Rework fast-path ethdev functions to use rte_eth_fp_ops[].
While it is an API/ABI breakage, this change is intended to be
transparent for both users (no changes in user app is required) and
PMD developers (no changes in PMD is required).
One extra thing to note - RX/TX callback invocation will cause extra
function call with these changes. That might cause some insignificant
slowdown for code-path where RX/TX callbacks are heavily involved.


I'm sorry for nit picking here and below:

RX -> Rx, TX -> Tx everywhere above.



Signed-off-by: Konstantin Ananyev 
---
  lib/ethdev/ethdev_private.c |  31 +
  lib/ethdev/rte_ethdev.h | 242 ++--
  lib/ethdev/version.map  |   3 +
  3 files changed, 208 insertions(+), 68 deletions(-)

diff --git a/lib/ethdev/ethdev_private.c b/lib/ethdev/ethdev_private.c
index 3eeda6e9f9..1222c6f84e 100644
--- a/lib/ethdev/ethdev_private.c
+++ b/lib/ethdev/ethdev_private.c
@@ -226,3 +226,34 @@ eth_dev_fp_ops_setup(struct rte_eth_fp_ops *fpo,
fpo->txq.data = dev->data->tx_queues;
fpo->txq.clbk = (void **)(uintptr_t)dev->pre_tx_burst_cbs;
  }
+
+uint16_t
+rte_eth_call_rx_callbacks(uint16_t port_id, uint16_t queue_id,
+   struct rte_mbuf **rx_pkts, uint16_t nb_rx, uint16_t nb_pkts,
+   void *opaque)
+{
+   const struct rte_eth_rxtx_callback *cb = opaque;
+
+   while (cb != NULL) {
+   nb_rx = cb->fn.rx(port_id, queue_id, rx_pkts, nb_rx,
+   nb_pkts, cb->param);
+   cb = cb->next;
+   }
+
+   return nb_rx;
+}
+
+uint16_t
+rte_eth_call_tx_callbacks(uint16_t port_id, uint16_t queue_id,
+   struct rte_mbuf **tx_pkts, uint16_t nb_pkts, void *opaque)
+{
+   const struct rte_eth_rxtx_callback *cb = opaque;
+
+   while (cb != NULL) {
+   nb_pkts = cb->fn.tx(port_id, queue_id, tx_pkts, nb_pkts,
+   cb->param);
+   cb = cb->next;
+   }
+
+   return nb_pkts;
+}
diff --git a/lib/ethdev/rte_ethdev.h b/lib/ethdev/rte_ethdev.h
index cdd16d6e57..c0e1a40681 100644
--- a/lib/ethdev/rte_ethdev.h
+++ b/lib/ethdev/rte_ethdev.h
@@ -4904,6 +4904,33 @@ int rte_eth_representor_info_get(uint16_t port_id,

  #include 

+/**
+ * @internal
+ * Helper routine for eth driver rx_burst API.


rx -> Rx


+ * Should be called at exit from PMD's rte_eth_rx_bulk implementation.
+ * Does necessary post-processing - invokes RX callbacks if any, etc.


RX -> Rx


+ *
+ * @param port_id
+ *  The port identifier of the Ethernet device.
+ * @param queue_id
+ *  The index of the receive queue from which to retrieve input packets.


Isn't:
The index of the queue from which packets are received from?


I copied it from comments from rte_eth_rx_burst().
I suppose it is just two ways to say the same thing.


May be it is just my problem that I don't understand the
initial description.






+ * @param rx_pkts
+ *   The address of an array of pointers to *rte_mbuf* structures that
+ *   have been retrieved from the device.
+ * @param nb_pkts


Should be @param nb_rx


Ack, will fix.




+ *   The number of packets that were retrieved from the device.
+ * @param nb_pkts
+ *   The number of elements in *rx_pkts* array.


@p should be used to refer to a paramter.


To be more precise you are talking about:
s/*rx_pkts*/@ rx_pkts/


s/"rx_pkts"/@p rx_pkts/


?



The description does not help to understand why both nb_rx and
nb_pkts are necessary. Isn't nb_pkts >= nb_rx and nb_rx
sufficient?


Nope,  that's for callbacks call.
Will update the comment.


Thanks.


+ * @param opaque
+ *   Opaque pointer of RX queue callback related data.


RX -> Rx


+ *
+ * @return
+ *  The number of packets effectively supplied to the *rx_pkts* array.


@p should be used to refer to a parameter.


+ */
+uint16_t rte_eth_call_rx_callbacks(uint16_t port_id, uint16_t queue_id,
+   struct rte_mbuf **rx_pkts, uint16_t nb_rx, uint16_t nb_pkts,
+   void *opaque);
+
  /**
   *
   * Retrieve a burst of input packets from a receive queue of an Ethernet
@@ -4995,23 +5022,37 @@ static inline uint16_t
  rte_eth_rx_burst(uint16_t port_id, uint16_t queue_id,
 struct rte_mbuf **rx_pkts, const uint16_t nb_pkts)
  {
-   struct rte_eth_dev *dev = &rte_eth_devices[port_id];
uint16_t nb_rx;
+   struct rte_eth_fp_ops *p;


p is typically a very bad name in a funcion with
many pointer variables etc. May be "fpo" as in previous
patch?


+   void *cb, *qd;


Please, avoid variable, expecially pointers, declaration in
one line.


Here and in other places, I think local variable names and placement,
is just a matter of personal preference.


Of course you can drop my notes as long as I'm alone.
I've started my comment from "Please" :)
May be I'm asking too much.

Also, I'm sorry, but I stricly against 'p' name since such
na

Re: [dpdk-dev] [PATCH v5 7/7] ethdev: hide eth dev related structures

[Andrew Rybchenko]

On 10/11/21 6:54 PM, Ananyev, Konstantin wrote:





On 10/7/21 2:27 PM, Konstantin Ananyev wrote:

Move rte_eth_dev, rte_eth_dev_data, rte_eth_rxtx_callback and related
data into private header (ethdev_driver.h).
Few minor changes to keep DPDK building after that.

Signed-off-by: Konstantin Ananyev 


[snip]

I realize that many notes below correspond to just
moved code, but I still think that ti would be nice
to fix anyway.


Yes, that’s just cut and paste from rte_ethdev_core.h and ethdev_driver.h
It is probably a good idea to clean-up and might be re-layout rte_eth_dev and 
friends,
but I don't think it has to be part of this patch-set.
After all, if it will become internal structure, that work could be done for 
any DPDK release.



OK, I can do cleanup as a follow up patch.

Re: [dpdk-dev] [PATCH v2 0/5] cryptodev: hide internal structures

[Ji, Kai]

Hi Akhil,

Just ran a quick mutli process test against the patch set, unfortunately it 
failed on the secondary process enqueue or dequeue.

USER1: Configuring vector 0, using session 0
USER1: Start enqueuing packets on dev 0 qp 0
USER1: Start dequeuing packets on dev 0 qp 0
USER1: Enqueuing - Dequeueing -Segmentation fault (core dumped)

It will happen on any PMD type.

Regards

Kai 

  
> -Original Message-
> From: dev  On Behalf Of Zhang, Roy Fan
> Sent: Monday, October 11, 2021 5:03 PM
> To: Akhil Goyal ; dev@dpdk.org
> Cc: tho...@monjalon.net; david.march...@redhat.com;
> hemant.agra...@nxp.com; ano...@marvell.com; De Lara Guarch, Pablo
> ; Trahe, Fiona ;
> Doherty, Declan ; ma...@nvidia.com;
> g.si...@nxp.com; jianjay.z...@huawei.com; asoma...@amd.com;
> ruifeng.w...@arm.com; Ananyev, Konstantin
> ; Nicolau, Radu ;
> ajit.khapa...@broadcom.com; rnagadhee...@marvell.com;
> adwiv...@marvell.com; Power, Ciara 
> Subject: Re: [dpdk-dev] [PATCH v2 0/5] cryptodev: hide internal structures
> 
> Hi Akhil,
> 
> The approach looks great but we may have to check if it works in multi-
> process environment - since all enqueue/dequeue handlers are set by
> primary process the secondary process may not recognize the fp_ops data.
> 
> We will run a quick test to see if it is true.
> 
> Regards,
> Fan
> 
> > -Original Message-
> > From: Akhil Goyal 
> > Sent: Monday, October 11, 2021 1:43 PM
> > To: dev@dpdk.org
> > Cc: tho...@monjalon.net; david.march...@redhat.com;
> > hemant.agra...@nxp.com; ano...@marvell.com; De Lara Guarch, Pablo
> > ; Trahe, Fiona
> > ; Doherty, Declan ;
> > ma...@nvidia.com; g.si...@nxp.com; Zhang, Roy Fan
> > ; jianjay.z...@huawei.com;
> asoma...@amd.com;
> > ruifeng.w...@arm.com; Ananyev, Konstantin
> > ; Nicolau, Radu
> > ; ajit.khapa...@broadcom.com;
> > rnagadhee...@marvell.com; adwiv...@marvell.com; Power, Ciara
> > ; Akhil Goyal 
> > Subject: [PATCH v2 0/5] cryptodev: hide internal structures
> >
> > Structures rte_cryptodev and rte_cryptodev_data are not supposed to be
> > directly used by the application. These are made public as they are
> > used by inline datapath public APIs.
> > This patchset, creates a new rte_cryptodev_core.h file which helps in
> > defining a data structure to hold datapath APIs in a flat array based
> > on the device identifier which is filled by the PMD.
> >
> > Similar series for ethdev and eventdev are also floated on ML.
> > https://patchwork.dpdk.org/project/dpdk/list/?series=19428
> > https://patchwork.dpdk.org/project/dpdk/list/?series=19405
> >
> > changes in v2: align with the latest versions of above series.
> >
> > Akhil Goyal (5):
> >   cryptodev: separate out internal structures
> >   cryptodev: allocate max space for internal qp array
> >   cryptodev: move inline APIs into separate structure
> >   cryptodev: update fast path APIs to use new flat array
> >   cryptodev: move device specific structures
> >
> >  drivers/crypto/aesni_mb/rte_aesni_mb_pmd.c|   1 -
> >  drivers/crypto/ccp/ccp_dev.h  |   2 +-
> >  drivers/crypto/cnxk/cn10k_ipsec.c |   2 +-
> >  drivers/crypto/cnxk/cn9k_ipsec.c  |   2 +-
> >  .../crypto/cnxk/cnxk_cryptodev_capabilities.c |   2 +-
> >  drivers/crypto/cnxk/cnxk_cryptodev_sec.c  |   2 +-
> >  drivers/crypto/nitrox/nitrox_sym_reqmgr.c |   2 +-
> >  drivers/crypto/octeontx/otx_cryptodev.c   |   1 -
> >  .../crypto/octeontx/otx_cryptodev_hw_access.c |   2 +-
> >  .../crypto/octeontx/otx_cryptodev_hw_access.h |   2 +-
> >  drivers/crypto/octeontx/otx_cryptodev_ops.h   |   2 +-
> >  .../crypto/octeontx2/otx2_cryptodev_mbox.c|   2 +-
> >  drivers/crypto/scheduler/scheduler_failover.c |   2 +-
> >  .../crypto/scheduler/scheduler_multicore.c|   2 +-
> >  .../scheduler/scheduler_pkt_size_distr.c  |   2 +-
> >  .../crypto/scheduler/scheduler_roundrobin.c   |   2 +-
> >  drivers/event/cnxk/cnxk_eventdev.h|   2 +-
> >  drivers/event/dpaa/dpaa_eventdev.c|   2 +-
> >  drivers/event/dpaa2/dpaa2_eventdev.c  |   2 +-
> >  drivers/event/octeontx/ssovf_evdev.c  |   2 +-
> >  .../event/octeontx2/otx2_evdev_crypto_adptr.c |   2 +-
> >  lib/cryptodev/cryptodev_pmd.c |  51 +++
> >  lib/cryptodev/cryptodev_pmd.h |  82 +++-
> >  lib/cryptodev/meson.build |   4 +-
> >  lib/cryptodev/rte_cryptodev.c |  50 ++-
> >  lib/cryptodev/rte_cryptodev.h | 367 +++---
> >  lib/cryptodev/rte_cryptodev_core.h|  62 +++
> >  lib/cryptodev/version.map |   7 +-
> >  28 files changed, 398 insertions(+), 265 deletions(-)  create mode
> > 100644 lib/cryptodev/rte_cryptodev_core.h
> >
> > --
> > 2.25.1

Re: [dpdk-dev] [PATCH v4 1/6] ethdev: introduce shared Rx queue

[Xueming(Steven) Li]

On Mon, 2021-10-11 at 13:47 +0300, Andrew Rybchenko wrote:
> On 9/30/21 5:55 PM, Xueming Li wrote:
> > In current DPDK framework, each RX queue is pre-loaded with mbufs for
> 
> RX -> Rx
> 
> > incoming packets. When number of representors scale out in a switch
> > domain, the memory consumption became significant. Most important,
> > polling all ports leads to high cache miss, high latency and low
> > throughput.
> 
> It should be highlighted that it is a problem of some PMDs.
> Not all.
> 
> > 
> > This patch introduces shared RX queue. Ports with same configuration in
> 
> "This patch introduces" -> "Introduce"
> 
> RX -> Rx
> 
> > a switch domain could share RX queue set by specifying sharing group.
> 
> RX -> Rx
> 
> > Polling any queue using same shared RX queue receives packets from all
> 
> RX -> Rx
> 
> > member ports. Source port is identified by mbuf->port.
> > 
> > Port queue number in a shared group should be identical. Queue index is
> > 1:1 mapped in shared group.
> > 
> > Share RX queue must be polled on single thread or core.
> 
> RX -> Rx
> 
> > 
> > Multiple groups is supported by group ID.
> 
> is -> are
> 
> > 
> > Signed-off-by: Xueming Li 
> > Cc: Jerin Jacob 
> 
> The patch should update release notes.
> 
> > ---
> > Rx queue object could be used as shared Rx queue object, it's important
> > to clear all queue control callback api that using queue object:
> >   https://mails.dpdk.org/archives/dev/2021-July/215574.html
> > ---
> >  doc/guides/nics/features.rst| 11 +++
> >  doc/guides/nics/features/default.ini|  1 +
> >  doc/guides/prog_guide/switch_representation.rst | 10 ++
> >  lib/ethdev/rte_ethdev.c |  1 +
> >  lib/ethdev/rte_ethdev.h |  7 +++
> >  5 files changed, 30 insertions(+)
> > 
> > diff --git a/doc/guides/nics/features.rst b/doc/guides/nics/features.rst
> > index 4fce8cd1c97..69bc1d5719c 100644
> > --- a/doc/guides/nics/features.rst
> > +++ b/doc/guides/nics/features.rst
> > @@ -626,6 +626,17 @@ Supports inner packet L4 checksum.
> >``tx_offload_capa,tx_queue_offload_capa:DEV_TX_OFFLOAD_OUTER_UDP_CKSUM``.
> >  
> >  
> > +.. _nic_features_shared_rx_queue:
> > +
> > +Shared Rx queue
> > +---
> > +
> > +Supports shared Rx queue for ports in same switch domain.
> > +
> > +* **[uses] rte_eth_rxconf,rte_eth_rxmode**: 
> > ``offloads:RTE_ETH_RX_OFFLOAD_SHARED_RXQ``.
> > +* **[provides] mbuf**: ``mbuf.port``.
> > +
> > +
> >  .. _nic_features_packet_type_parsing:
> >  
> >  Packet type parsing
> > diff --git a/doc/guides/nics/features/default.ini 
> > b/doc/guides/nics/features/default.ini
> > index 754184ddd4d..ebeb4c18512 100644
> > --- a/doc/guides/nics/features/default.ini
> > +++ b/doc/guides/nics/features/default.ini
> > @@ -19,6 +19,7 @@ Free Tx mbuf on demand =
> >  Queue start/stop =
> >  Runtime Rx queue setup =
> >  Runtime Tx queue setup =
> > +Shared Rx queue  =
> >  Burst mode info  =
> >  Power mgmt address monitor =
> >  MTU update   =
> > diff --git a/doc/guides/prog_guide/switch_representation.rst 
> > b/doc/guides/prog_guide/switch_representation.rst
> > index ff6aa91c806..bc7ce65fa3d 100644
> > --- a/doc/guides/prog_guide/switch_representation.rst
> > +++ b/doc/guides/prog_guide/switch_representation.rst
> > @@ -123,6 +123,16 @@ thought as a software "patch panel" front-end for 
> > applications.
> >  .. [1] `Ethernet switch device driver model (switchdev)
> > 
> > `_
> >  
> > +- Memory usage of representors is huge when number of representor grows,
> > +  because PMD always allocate mbuf for each descriptor of Rx queue.
> 
> It is a problem of some PMDs only. So, it must be rewritten to
> highlight it.
> 
> > +  Polling the large number of ports brings more CPU load, cache miss and
> > +  latency. Shared Rx queue can be used to share Rx queue between PF and
> > +  representors in same switch. ``RTE_ETH_RX_OFFLOAD_SHARED_RXQ`` is
> > +  present in Rx offloading capability of device info. Setting the
> > +  offloading flag in device Rx mode or Rx queue configuration to enable
> > +  shared Rx queue. Polling any member port of the shared Rx queue can 
> > return
> > +  packets of all ports in the group, port ID is saved in ``mbuf.port``.
> > +
> >  Basic SR-IOV
> >  
> >  
> > diff --git a/lib/ethdev/rte_ethdev.c b/lib/ethdev/rte_ethdev.c
> > index 61aa49efec6..73270c10492 100644
> > --- a/lib/ethdev/rte_ethdev.c
> > +++ b/lib/ethdev/rte_ethdev.c
> > @@ -127,6 +127,7 @@ static const struct {
> > RTE_RX_OFFLOAD_BIT2STR(OUTER_UDP_CKSUM),
> > RTE_RX_OFFLOAD_BIT2STR(RSS_HASH),
> > RTE_ETH_RX_OFFLOAD_BIT2STR(BUFFER_SPLIT),
> > +   RTE_ETH_RX_OFFLOAD_BIT2STR(SHARED_RXQ),
> >  };
> >  
> >  #undef RTE_RX_OFFLOAD_BIT2STR
> > diff --git a/lib/ethdev/rte_ethdev.h b/lib/ethdev/rte_ethdev.h
> > index afdc53b674c..d7ac625ee74 100644
> > 

Re: [dpdk-dev] [PATCH v5 2/7] ethdev: allocate max space for internal queue array

[Andrew Rybchenko]

On 10/11/21 7:25 PM, Ananyev, Konstantin wrote:




At queue configure stage always allocate space for maximum possible
number (RTE_MAX_QUEUES_PER_PORT) of queue pointers.
That will allow 'fast' inline functions (eth_rx_burst, etc.) to refer
pointer to internal queue data without extra checking of current number
of configured queues.
That would help in future to hide rte_eth_dev and related structures.
It means that from now on, each ethdev port will always consume:
((2*sizeof(uintptr_t))* RTE_MAX_QUEUES_PER_PORT)
bytes of memory for its queue pointers.
With RTE_MAX_QUEUES_PER_PORT==1024 (default value) it is 16KB per port.

Signed-off-by: Konstantin Ananyev 
---
  lib/ethdev/rte_ethdev.c | 36 +---
  1 file changed, 9 insertions(+), 27 deletions(-)

diff --git a/lib/ethdev/rte_ethdev.c b/lib/ethdev/rte_ethdev.c
index ed37f8871b..c8abda6dd7 100644
--- a/lib/ethdev/rte_ethdev.c
+++ b/lib/ethdev/rte_ethdev.c
@@ -897,7 +897,8 @@ eth_dev_rx_queue_config(struct rte_eth_dev *dev, uint16_t 
nb_queues)

if (dev->data->rx_queues == NULL && nb_queues != 0) { /* first time 
configuration */
dev->data->rx_queues = rte_zmalloc("ethdev->rx_queues",
-   sizeof(dev->data->rx_queues[0]) * nb_queues,
+   sizeof(dev->data->rx_queues[0]) *
+   RTE_MAX_QUEUES_PER_PORT,
RTE_CACHE_LINE_SIZE);


Looking at it I have few questions:
1. Why is nb_queues == 0 case kept as an exception? Yes,
strictly speaking it is not the problem of the patch,
DPDK will still segfault (non-debug build) if I
allocate Tx queues only but call rte_eth_rx_burst().


eth_dev_rx_queue_config(.., nb_queues=0) is used in few places to clean-up 
things.


No, as far as I know. For Tx only application (e.g. traffic generator)
it is 100% legal to configure with tx_queues=X, rx_queues=0.
The same is for Rx only application (e.g. packet capture).




After reading the patch description I thought that
we're trying to address it.


We do, though I can't see how we can address it in this patch.
Though it is a good idea - I think I can add extra check in 
eth_dev_fp_ops_setup()
or around and setup RX function pointers only when dev->data->rx_queues != NULL.
Same for TX.


You don't need to care about these pointers, if these arrays are
always allocated. See (3) below.




2. Why do we need to allocate memory dynamically?
Can we just make rx_queues an array of appropriate size?


Pavan already asked same question.
My answer to him:
Yep we can, and yes it will simplify this peace of code.
The main reason I decided no to do this change now -
it will change layout of the_eth_dev_data structure.
In this series I tried to mininize(/avoid) changes in rte_eth_dev and 
rte_eth_dev_data,
as much as possible to avoid any unforeseen performance and functional impacts.
If we'll manage to make rte_eth_dev and rte_eth_dev_data private we can in 
future
consider that one and other changes in rte_eth_dev and rte_eth_dev_data layouts
without worrying about ABI breakage


Thanks a lot. Makes sense.


May be wasting 512K unconditionally is too much.
3. If wasting 512K is too much, I'd consider to move
allocation to eth_dev_get(). If


Don't understand where 512KB came from.


32 port * 1024 queues * 2 types * 8 pointer size
if we allocate as in (2) above.


each ethdev port will always consume:
((2*sizeof(uintptr_t))* RTE_MAX_QUEUES_PER_PORT)
bytes of memory for its queue pointers.
With RTE_MAX_QUEUES_PER_PORT==1024 (default value) it is 16KB per port.


IMHO it will be a bit nicer if queue pointers arrays are allocated
on device get if size is fixed. It is just a suggestion. If you
disagree, feel free to drop it.


if (dev->data->rx_queues == NULL) {
dev->data->nb_rx_queues = 0;
@@ -908,21 +909,11 @@ eth_dev_rx_queue_config(struct rte_eth_dev *dev, uint16_t 
nb_queues)

rxq = dev->data->rx_queues;

-   for (i = nb_queues; i < old_nb_queues; i++)
+   for (i = nb_queues; i < old_nb_queues; i++) {
(*dev->dev_ops->rx_queue_release)(rxq[i]);
-   rxq = rte_realloc(rxq, sizeof(rxq[0]) * nb_queues,
-   RTE_CACHE_LINE_SIZE);
-   if (rxq == NULL)
-   return -(ENOMEM);
-   if (nb_queues > old_nb_queues) {
-   uint16_t new_qs = nb_queues - old_nb_queues;
-
-   memset(rxq + old_nb_queues, 0,
-   sizeof(rxq[0]) * new_qs);
+   rxq[i] = NULL;


It looks like the patch should be rebased on top of
next-net main because of queue release patches.

[snip]

Re: [dpdk-dev] [dpdk-ci] RHEL7 failures

[Brandon Lo]

Hi Thinh,

I have started a rerun of the patch
https://patches.dpdk.org/patch/100042 (patchset
https://lab.dpdk.org/results/dashboard/patchsets/19066/).
You should see the new results reported over email soon.

Thanks,
Brandon


On Mon, Oct 11, 2021 at 12:39 PM Thinh Tran  wrote:
>
> Hi Linconln,
>
> I have this patch set 100042  - build/ppc: enable build support for Power10
> https://patches.dpdk.org/patch/100042
>
> Thanks,
> Thinh Tran
>
>
>
> On 10/11/2021 9:42 AM, Lincoln Lavoie wrote:
> > Hi Thinh, The CI won't automatically go back and retest failures after a
> > patch was applied.  This is a corner case, where a bad change was merged
> > and affected all patches downstream of that merge.  We can retrigger
> > specific patches, just ZjQcmQRYFpfptBannerStart
> > This Message Is From an External Sender
> > This message came from outside your organization.
> > ZjQcmQRYFpfptBannerEnd
> > Hi Thinh,
> >
> > The CI won't automatically go back and retest failures after a patch was
> > applied.  This is a corner case, where a bad change was merged and
> > affected all patches downstream of that merge.  We can retrigger
> > specific patches, just let us know the patch number or the URI of the
> > patch(es).
> >
> > Cheers,
> > Lincoln
> >
> > On Mon, Oct 11, 2021 at 10:39 AM Thinh Tran  > > wrote:
> >
> >
> > Hi,
> > I'm wondering how or when the CI would re-pick up the patches those
> > were
> > failed due to this issue?
> >
> > Regards,
> > Thinh Tran
> >
> > On 10/6/2021 10:40 PM, Zhang, Qi Z wrote:
> >  > Hi Ferruh, David and Aman:
> >  >
> >  >   Sorry for late response due to PRC holiday.
> >  >   I have just tested the patch and there is no issue be
> > detected for functions
> >  >   Thank you so much for the help!
> >  >
> >  > Regards
> >  > Qi
> >  >
> >  >> -Original Message-
> >  >> From: Yigit, Ferruh  > >
> >  >> Sent: Wednesday, October 6, 2021 5:46 AM
> >  >> To: Lincoln Lavoie  > >; dev  > >; Yang, Qiming
> >  >> mailto:qiming.y...@intel.com>>; Zhang,
> > Qi Z mailto:qi.z.zh...@intel.com>>
> >  >> Cc: c...@dpdk.org ; Aaron Conole
> > mailto:acon...@redhat.com>>; dpdklab
> >  >> mailto:dpdk...@iol.unh.edu>>; Singh, Aman
> > Deep mailto:aman.deep.si...@intel.com>>;
> >  >> David Marchand  > >
> >  >> Subject: Re: [dpdk-dev] RHEL7 failures
> >  >>
> >  >> On 10/5/2021 8:09 PM, Lincoln Lavoie wrote:
> >  >>> Hello Qiming and Qi,
> >  >>>
> >  >>> The CI is picking up failures when building on RHEL7, where
> > warnings
> >  >>> are being treated as errors.  This looks like something has been
> >  >>> merged into the mainline, as it's failing across all patches.
> >  >>>
> >  >>> Here is the specific failure:
> >  >>>
> >  >>> ./drivers/net/ice/base/ice_parser_rt.c: In function '_hv_bit_sel':
> >  >>> ../drivers/net/ice/base/ice_parser_rt.c:201:2: error: dereferencing
> >  >>> type-punned pointer will break strict-aliasing rules
> >  >>> [-Werror=strict-aliasing]
> >  >>> d64 = *(u64 *)&b[0];
> >  >>> ^
> >  >>> ../drivers/net/ice/base/ice_parser_rt.c: In function
> > '_reg_bit_sel':
> >  >>> ../drivers/net/ice/base/ice_parser_rt.c:458:2: error: dereferencing
> >  >>> type-punned pointer will break strict-aliasing rules
> >  >>> [-Werror=strict-aliasing]
> >  >>> d32 = *(u32 *)&v[0];
> >  >>> ^
> >  >>> cc1: all warnings being treated as errors
> >  >>>
> >  >>> You can download a full set of logs from here (for a failing run):
> >  >>> https://lab.dpdk.org/results/dashboard/patchsets/19162/
> > 
> >  >>>
> >  >>
> >  >> Issue was reported by David, Aman sent the fix [1] and it is
> > already merged by
> >  >> David [2], it should be fixed now, can you please double check?
> >  >>
> >  >> [1]
> >  >>
> > https://patches.dpdk.org/project/dpdk/patch/20211005115754.34117-1-aman
> > 
> > 
> >  >> .deep.si...@intel.com/ 
> >  >>
> >  >> [2]
> >  >>
> > https://git.dpdk.org/dpdk/commit/?id=16b809d144dc2df7f31695b5abc64a809
> > 
> >  >> 021b154
> >
> >
> >
> > --
> > *Lincoln Lavoie*
> > Principal Engineer, Broadband Technologies
> > 21 Madbury Rd., Ste. 100, Durham, NH 03824
> > lylav...@iol.unh.edu 
> > https://www.iol.unh.edu 
> > +1-603-674-2755 (m)
> > 



-- 
Brandon Lo
UNH Inter