All of lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH 00/13] cxgbe: add CXGBE VF PMD and updates
@ 2018-03-10 22:48 Rahul Lakkireddy
  2018-03-10 22:48 ` [PATCH 01/13] cxgbe: add skeleton VF driver Rahul Lakkireddy
                   ` (14 more replies)
  0 siblings, 15 replies; 26+ messages in thread
From: Rahul Lakkireddy @ 2018-03-10 22:48 UTC (permalink / raw)
  To: dev; +Cc: kumaras, nirranjan, indranil

Patches 1 - 9 add support for cxgbe VF driver.  Patches 10 - 12 fix
bugs and convert license in cxgbe files to SPDX License Tag.  Patch
13 adds compile time option to keep outer vlan tag in Q-in-Q packet.

Patch 1 adds minimal cxgbe VF driver.

Patch 2 adds firmware mailbox support for VF.

Patch 3 adds base functions to enable VF ports in subsequent patches.

Patch 4 adds cxgbe VF probe to initialize VF devices.

Patch 5 initializes SGE and queues for VF.

Patch 6 enables RSS for VF.

Patch 7 updates TX and RX path for VF.

Patch 8 adds support for VF port statistics.

Patch 9 adds support to set mac address.

Patch 10 fixes bug where the other ports under same PF are not closed
properly.

Patch 11 exports RSS hash functions in device info and adds check
to prevent configuring unsupported hash functions.

Patch 12 converts all cxgbe files to use SPDX license tag.

Patch 13 adds compile time option to keep outer vlan tag in Q-in-Q
packet.

Thanks,
Rahul

Kumar Sanghvi (10):
  cxgbe: add skeleton VF driver
  cxgbe: add VF firmware mailbox support
  cxgbe: add base for enabling VF ports
  cxgbe: add probe to initialize VF devices
  cxgbe: initialize SGE and queues for VF
  cxgbe: enable RSS for VF
  cxgbe: update TX and RX path for VF
  cxgbe: add VF port statistics
  cxgbe: add support to set mac address
  cxgbe: add option to keep outer VLAN tag in Q-in-Q

Rahul Lakkireddy (3):
  cxgbe: fix check to close other ports properly
  cxgbe: export supported RSS hash functions
  cxgbe: convert to SPDX license tags

 config/common_base                      |   1 +
 doc/guides/nics/cxgbe.rst               |  36 +-
 doc/guides/nics/features/cxgbevf.ini    |  28 +
 doc/guides/rel_notes/release_18_05.rst  |   5 +
 drivers/net/cxgbe/Makefile              |  36 +-
 drivers/net/cxgbe/base/adapter.h        |  42 +-
 drivers/net/cxgbe/base/common.h         | 111 ++--
 drivers/net/cxgbe/base/t4_chip_type.h   |  34 +-
 drivers/net/cxgbe/base/t4_hw.c          | 117 +++--
 drivers/net/cxgbe/base/t4_hw.h          |  34 +-
 drivers/net/cxgbe/base/t4_msg.h         |  34 +-
 drivers/net/cxgbe/base/t4_pci_id_tbl.h  |  34 +-
 drivers/net/cxgbe/base/t4_regs.h        | 116 +++--
 drivers/net/cxgbe/base/t4_regs_values.h |  34 +-
 drivers/net/cxgbe/base/t4fw_interface.h | 229 +++++++--
 drivers/net/cxgbe/base/t4vf_hw.c        | 874 ++++++++++++++++++++++++++++++++
 drivers/net/cxgbe/base/t4vf_hw.h        |  15 +
 drivers/net/cxgbe/cxgbe.h               |  44 +-
 drivers/net/cxgbe/cxgbe_compat.h        |  34 +-
 drivers/net/cxgbe/cxgbe_ethdev.c        | 143 +++---
 drivers/net/cxgbe/cxgbe_main.c          | 118 +++--
 drivers/net/cxgbe/cxgbe_pfvf.h          |  43 ++
 drivers/net/cxgbe/cxgbevf_ethdev.c      | 197 +++++++
 drivers/net/cxgbe/cxgbevf_main.c        | 306 +++++++++++
 drivers/net/cxgbe/sge.c                 | 415 ++++++++++++---
 25 files changed, 2417 insertions(+), 663 deletions(-)
 create mode 100644 doc/guides/nics/features/cxgbevf.ini
 create mode 100644 drivers/net/cxgbe/base/t4vf_hw.c
 create mode 100644 drivers/net/cxgbe/base/t4vf_hw.h
 create mode 100644 drivers/net/cxgbe/cxgbe_pfvf.h
 create mode 100644 drivers/net/cxgbe/cxgbevf_ethdev.c
 create mode 100644 drivers/net/cxgbe/cxgbevf_main.c

-- 
2.14.1

^ permalink raw reply	[flat|nested] 26+ messages in thread

* [PATCH 01/13] cxgbe: add skeleton VF driver
  2018-03-10 22:48 [PATCH 00/13] cxgbe: add CXGBE VF PMD and updates Rahul Lakkireddy
@ 2018-03-10 22:48 ` Rahul Lakkireddy
  2018-03-10 22:48 ` [PATCH 02/13] cxgbe: add VF firmware mailbox support Rahul Lakkireddy
                   ` (13 subsequent siblings)
  14 siblings, 0 replies; 26+ messages in thread
From: Rahul Lakkireddy @ 2018-03-10 22:48 UTC (permalink / raw)
  To: dev; +Cc: kumaras, nirranjan, indranil

From: Kumar Sanghvi <kumaras@chelsio.com>

Add minimal VF driver.  Declare functions common to both PF and VF
functionality in separate header file and import the header file.

Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
---
 doc/guides/nics/features/cxgbevf.ini   | 28 +++++++++++
 doc/guides/rel_notes/release_18_05.rst |  5 ++
 drivers/net/cxgbe/Makefile             |  1 +
 drivers/net/cxgbe/cxgbe_ethdev.c       | 70 +++++++++++---------------
 drivers/net/cxgbe/cxgbe_pfvf.h         | 38 ++++++++++++++
 drivers/net/cxgbe/cxgbevf_ethdev.c     | 91 ++++++++++++++++++++++++++++++++++
 6 files changed, 193 insertions(+), 40 deletions(-)
 create mode 100644 doc/guides/nics/features/cxgbevf.ini
 create mode 100644 drivers/net/cxgbe/cxgbe_pfvf.h
 create mode 100644 drivers/net/cxgbe/cxgbevf_ethdev.c

diff --git a/doc/guides/nics/features/cxgbevf.ini b/doc/guides/nics/features/cxgbevf.ini
new file mode 100644
index 000000000..7706106d1
--- /dev/null
+++ b/doc/guides/nics/features/cxgbevf.ini
@@ -0,0 +1,28 @@
+;
+; Supported features of the 'cxgbevf' network poll mode driver.
+;
+; Refer to default.ini for the full list of available PMD features.
+;
+[Features]
+Speed capabilities   = Y
+Link status          = Y
+Queue start/stop     = Y
+MTU update           = Y
+Jumbo frame          = Y
+Scattered Rx         = Y
+TSO                  = Y
+Promiscuous mode     = Y
+Allmulticast mode    = Y
+RSS hash             = Y
+CRC offload          = Y
+VLAN offload         = Y
+L3 checksum offload  = Y
+L4 checksum offload  = Y
+Packet type parsing  = Y
+Basic stats          = Y
+Stats per queue      = Y
+Multiprocess aware   = Y
+Linux UIO            = Y
+Linux VFIO           = Y
+x86-32               = Y
+x86-64               = Y
diff --git a/doc/guides/rel_notes/release_18_05.rst b/doc/guides/rel_notes/release_18_05.rst
index f2525bbd9..93ee69249 100644
--- a/doc/guides/rel_notes/release_18_05.rst
+++ b/doc/guides/rel_notes/release_18_05.rst
@@ -45,6 +45,11 @@ New Features
 
   Support to update RSS hash and key has been added to CXGBE PMD.
 
+* **Added CXGBE VF PMD.**
+
+  CXGBE VF Poll Mode Driver has been added to run DPDK over Chelsio
+  T5/T6 NIC VF instances.
+
 * **Updated Solarflare network PMD.**
 
   Updated the sfc_efx driver including the following changes:
diff --git a/drivers/net/cxgbe/Makefile b/drivers/net/cxgbe/Makefile
index 65df14253..397977e02 100644
--- a/drivers/net/cxgbe/Makefile
+++ b/drivers/net/cxgbe/Makefile
@@ -80,6 +80,7 @@ VPATH += $(SRCDIR)/base
 # all source are stored in SRCS-y
 #
 SRCS-$(CONFIG_RTE_LIBRTE_CXGBE_PMD) += cxgbe_ethdev.c
+SRCS-$(CONFIG_RTE_LIBRTE_CXGBE_PMD) += cxgbevf_ethdev.c
 SRCS-$(CONFIG_RTE_LIBRTE_CXGBE_PMD) += cxgbe_main.c
 SRCS-$(CONFIG_RTE_LIBRTE_CXGBE_PMD) += sge.c
 SRCS-$(CONFIG_RTE_LIBRTE_CXGBE_PMD) += t4_hw.c
diff --git a/drivers/net/cxgbe/cxgbe_ethdev.c b/drivers/net/cxgbe/cxgbe_ethdev.c
index 781d75cc2..7f523265d 100644
--- a/drivers/net/cxgbe/cxgbe_ethdev.c
+++ b/drivers/net/cxgbe/cxgbe_ethdev.c
@@ -63,6 +63,7 @@
 #include <rte_dev.h>
 
 #include "cxgbe.h"
+#include "cxgbe_pfvf.h"
 
 /*
  * Macros needed to support the PCI Device ID Table ...
@@ -135,8 +136,8 @@ static uint16_t cxgbe_recv_pkts(void *rx_queue, struct rte_mbuf **rx_pkts,
 	return work_done;
 }
 
-static void cxgbe_dev_info_get(struct rte_eth_dev *eth_dev,
-			       struct rte_eth_dev_info *device_info)
+void cxgbe_dev_info_get(struct rte_eth_dev *eth_dev,
+			struct rte_eth_dev_info *device_info)
 {
 	struct port_info *pi = (struct port_info *)(eth_dev->data->dev_private);
 	struct adapter *adapter = pi->adapter;
@@ -178,7 +179,7 @@ static void cxgbe_dev_info_get(struct rte_eth_dev *eth_dev,
 	cxgbe_get_speed_caps(pi, &device_info->speed_capa);
 }
 
-static void cxgbe_dev_promiscuous_enable(struct rte_eth_dev *eth_dev)
+void cxgbe_dev_promiscuous_enable(struct rte_eth_dev *eth_dev)
 {
 	struct port_info *pi = (struct port_info *)(eth_dev->data->dev_private);
 	struct adapter *adapter = pi->adapter;
@@ -187,7 +188,7 @@ static void cxgbe_dev_promiscuous_enable(struct rte_eth_dev *eth_dev)
 		      1, -1, 1, -1, false);
 }
 
-static void cxgbe_dev_promiscuous_disable(struct rte_eth_dev *eth_dev)
+void cxgbe_dev_promiscuous_disable(struct rte_eth_dev *eth_dev)
 {
 	struct port_info *pi = (struct port_info *)(eth_dev->data->dev_private);
 	struct adapter *adapter = pi->adapter;
@@ -196,7 +197,7 @@ static void cxgbe_dev_promiscuous_disable(struct rte_eth_dev *eth_dev)
 		      0, -1, 1, -1, false);
 }
 
-static void cxgbe_dev_allmulticast_enable(struct rte_eth_dev *eth_dev)
+void cxgbe_dev_allmulticast_enable(struct rte_eth_dev *eth_dev)
 {
 	struct port_info *pi = (struct port_info *)(eth_dev->data->dev_private);
 	struct adapter *adapter = pi->adapter;
@@ -207,7 +208,7 @@ static void cxgbe_dev_allmulticast_enable(struct rte_eth_dev *eth_dev)
 		      -1, 1, 1, -1, false);
 }
 
-static void cxgbe_dev_allmulticast_disable(struct rte_eth_dev *eth_dev)
+void cxgbe_dev_allmulticast_disable(struct rte_eth_dev *eth_dev)
 {
 	struct port_info *pi = (struct port_info *)(eth_dev->data->dev_private);
 	struct adapter *adapter = pi->adapter;
@@ -218,8 +219,8 @@ static void cxgbe_dev_allmulticast_disable(struct rte_eth_dev *eth_dev)
 		      -1, 0, 1, -1, false);
 }
 
-static int cxgbe_dev_link_update(struct rte_eth_dev *eth_dev,
-				 __rte_unused int wait_to_complete)
+int cxgbe_dev_link_update(struct rte_eth_dev *eth_dev,
+			  __rte_unused int wait_to_complete)
 {
 	struct port_info *pi = (struct port_info *)(eth_dev->data->dev_private);
 	struct adapter *adapter = pi->adapter;
@@ -239,7 +240,7 @@ static int cxgbe_dev_link_update(struct rte_eth_dev *eth_dev,
 	return 0;
 }
 
-static int cxgbe_dev_mtu_set(struct rte_eth_dev *eth_dev, uint16_t mtu)
+int cxgbe_dev_mtu_set(struct rte_eth_dev *eth_dev, uint16_t mtu)
 {
 	struct port_info *pi = (struct port_info *)(eth_dev->data->dev_private);
 	struct adapter *adapter = pi->adapter;
@@ -267,17 +268,10 @@ static int cxgbe_dev_mtu_set(struct rte_eth_dev *eth_dev, uint16_t mtu)
 	return err;
 }
 
-static int cxgbe_dev_tx_queue_start(struct rte_eth_dev *eth_dev,
-				    uint16_t tx_queue_id);
-static int cxgbe_dev_rx_queue_start(struct rte_eth_dev *eth_dev,
-				    uint16_t tx_queue_id);
-static void cxgbe_dev_tx_queue_release(void *q);
-static void cxgbe_dev_rx_queue_release(void *q);
-
 /*
  * Stop device.
  */
-static void cxgbe_dev_close(struct rte_eth_dev *eth_dev)
+void cxgbe_dev_close(struct rte_eth_dev *eth_dev)
 {
 	struct port_info *pi = (struct port_info *)(eth_dev->data->dev_private);
 	struct adapter *adapter = pi->adapter;
@@ -316,7 +310,7 @@ static void cxgbe_dev_close(struct rte_eth_dev *eth_dev)
 /* Start the device.
  * It returns 0 on success.
  */
-static int cxgbe_dev_start(struct rte_eth_dev *eth_dev)
+int cxgbe_dev_start(struct rte_eth_dev *eth_dev)
 {
 	struct port_info *pi = (struct port_info *)(eth_dev->data->dev_private);
 	struct adapter *adapter = pi->adapter;
@@ -368,7 +362,7 @@ static int cxgbe_dev_start(struct rte_eth_dev *eth_dev)
 /*
  * Stop device: disable rx and tx functions to allow for reconfiguring.
  */
-static void cxgbe_dev_stop(struct rte_eth_dev *eth_dev)
+void cxgbe_dev_stop(struct rte_eth_dev *eth_dev)
 {
 	struct port_info *pi = (struct port_info *)(eth_dev->data->dev_private);
 	struct adapter *adapter = pi->adapter;
@@ -387,7 +381,7 @@ static void cxgbe_dev_stop(struct rte_eth_dev *eth_dev)
 	t4_sge_eth_clear_queues(pi);
 }
 
-static int cxgbe_dev_configure(struct rte_eth_dev *eth_dev)
+int cxgbe_dev_configure(struct rte_eth_dev *eth_dev)
 {
 	struct port_info *pi = (struct port_info *)(eth_dev->data->dev_private);
 	struct adapter *adapter = pi->adapter;
@@ -409,8 +403,7 @@ static int cxgbe_dev_configure(struct rte_eth_dev *eth_dev)
 	return 0;
 }
 
-static int cxgbe_dev_tx_queue_start(struct rte_eth_dev *eth_dev,
-				    uint16_t tx_queue_id)
+int cxgbe_dev_tx_queue_start(struct rte_eth_dev *eth_dev, uint16_t tx_queue_id)
 {
 	int ret;
 	struct sge_eth_txq *txq = (struct sge_eth_txq *)
@@ -425,8 +418,7 @@ static int cxgbe_dev_tx_queue_start(struct rte_eth_dev *eth_dev,
 	return ret;
 }
 
-static int cxgbe_dev_tx_queue_stop(struct rte_eth_dev *eth_dev,
-				   uint16_t tx_queue_id)
+int cxgbe_dev_tx_queue_stop(struct rte_eth_dev *eth_dev, uint16_t tx_queue_id)
 {
 	int ret;
 	struct sge_eth_txq *txq = (struct sge_eth_txq *)
@@ -441,10 +433,10 @@ static int cxgbe_dev_tx_queue_stop(struct rte_eth_dev *eth_dev,
 	return ret;
 }
 
-static int cxgbe_dev_tx_queue_setup(struct rte_eth_dev *eth_dev,
-				    uint16_t queue_idx,	uint16_t nb_desc,
-				    unsigned int socket_id,
-				    const struct rte_eth_txconf *tx_conf)
+int cxgbe_dev_tx_queue_setup(struct rte_eth_dev *eth_dev,
+			     uint16_t queue_idx, uint16_t nb_desc,
+			     unsigned int socket_id,
+			     const struct rte_eth_txconf *tx_conf)
 {
 	struct port_info *pi = (struct port_info *)(eth_dev->data->dev_private);
 	struct adapter *adapter = pi->adapter;
@@ -495,7 +487,7 @@ static int cxgbe_dev_tx_queue_setup(struct rte_eth_dev *eth_dev,
 	return err;
 }
 
-static void cxgbe_dev_tx_queue_release(void *q)
+void cxgbe_dev_tx_queue_release(void *q)
 {
 	struct sge_eth_txq *txq = (struct sge_eth_txq *)q;
 
@@ -511,8 +503,7 @@ static void cxgbe_dev_tx_queue_release(void *q)
 	}
 }
 
-static int cxgbe_dev_rx_queue_start(struct rte_eth_dev *eth_dev,
-				    uint16_t rx_queue_id)
+int cxgbe_dev_rx_queue_start(struct rte_eth_dev *eth_dev, uint16_t rx_queue_id)
 {
 	int ret;
 	struct port_info *pi = (struct port_info *)(eth_dev->data->dev_private);
@@ -531,8 +522,7 @@ static int cxgbe_dev_rx_queue_start(struct rte_eth_dev *eth_dev,
 	return ret;
 }
 
-static int cxgbe_dev_rx_queue_stop(struct rte_eth_dev *eth_dev,
-				   uint16_t rx_queue_id)
+int cxgbe_dev_rx_queue_stop(struct rte_eth_dev *eth_dev, uint16_t rx_queue_id)
 {
 	int ret;
 	struct port_info *pi = (struct port_info *)(eth_dev->data->dev_private);
@@ -550,11 +540,11 @@ static int cxgbe_dev_rx_queue_stop(struct rte_eth_dev *eth_dev,
 	return ret;
 }
 
-static int cxgbe_dev_rx_queue_setup(struct rte_eth_dev *eth_dev,
-				    uint16_t queue_idx,	uint16_t nb_desc,
-				    unsigned int socket_id,
-				    const struct rte_eth_rxconf *rx_conf,
-				    struct rte_mempool *mp)
+int cxgbe_dev_rx_queue_setup(struct rte_eth_dev *eth_dev,
+			     uint16_t queue_idx, uint16_t nb_desc,
+			     unsigned int socket_id,
+			     const struct rte_eth_rxconf *rx_conf,
+			     struct rte_mempool *mp)
 {
 	struct port_info *pi = (struct port_info *)(eth_dev->data->dev_private);
 	struct adapter *adapter = pi->adapter;
@@ -628,7 +618,7 @@ static int cxgbe_dev_rx_queue_setup(struct rte_eth_dev *eth_dev,
 	return err;
 }
 
-static void cxgbe_dev_rx_queue_release(void *q)
+void cxgbe_dev_rx_queue_release(void *q)
 {
 	struct sge_eth_rxq *rxq = (struct sge_eth_rxq *)q;
 	struct sge_rspq *rq = &rxq->rspq;
@@ -774,7 +764,7 @@ static int cxgbe_flow_ctrl_set(struct rte_eth_dev *eth_dev,
 			     &pi->link_cfg);
 }
 
-static const uint32_t *
+const uint32_t *
 cxgbe_dev_supported_ptypes_get(struct rte_eth_dev *eth_dev)
 {
 	static const uint32_t ptypes[] = {
diff --git a/drivers/net/cxgbe/cxgbe_pfvf.h b/drivers/net/cxgbe/cxgbe_pfvf.h
new file mode 100644
index 000000000..e3d8533ca
--- /dev/null
+++ b/drivers/net/cxgbe/cxgbe_pfvf.h
@@ -0,0 +1,38 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2018 Chelsio Communications.
+ * All rights reserved.
+ */
+
+#ifndef _CXGBE_PFVF_H_
+#define _CXGBE_PFVF_H_
+
+void cxgbe_dev_rx_queue_release(void *q);
+void cxgbe_dev_tx_queue_release(void *q);
+void cxgbe_dev_stop(struct rte_eth_dev *eth_dev);
+void cxgbe_dev_close(struct rte_eth_dev *eth_dev);
+void cxgbe_dev_info_get(struct rte_eth_dev *eth_dev,
+			struct rte_eth_dev_info *device_info);
+void cxgbe_dev_promiscuous_enable(struct rte_eth_dev *eth_dev);
+void cxgbe_dev_promiscuous_disable(struct rte_eth_dev *eth_dev);
+void cxgbe_dev_allmulticast_enable(struct rte_eth_dev *eth_dev);
+void cxgbe_dev_allmulticast_disable(struct rte_eth_dev *eth_dev);
+int cxgbe_dev_configure(struct rte_eth_dev *eth_dev);
+int cxgbe_dev_tx_queue_setup(struct rte_eth_dev *eth_dev, uint16_t queue_idx,
+			     uint16_t nb_desc, unsigned int socket_id,
+			     const struct rte_eth_txconf *tx_conf);
+int cxgbe_dev_rx_queue_setup(struct rte_eth_dev *eth_dev, uint16_t queue_idx,
+			     uint16_t nb_desc, unsigned int socket_id,
+			     const struct rte_eth_rxconf *rx_conf,
+			     struct rte_mempool *mp);
+int cxgbe_dev_tx_queue_start(struct rte_eth_dev *eth_dev,
+			     uint16_t tx_queue_id);
+int cxgbe_dev_rx_queue_start(struct rte_eth_dev *eth_dev,
+			     uint16_t tx_queue_id);
+int cxgbe_dev_tx_queue_stop(struct rte_eth_dev *eth_dev, uint16_t tx_queue_id);
+int cxgbe_dev_rx_queue_stop(struct rte_eth_dev *eth_dev, uint16_t rx_queue_id);
+int cxgbe_dev_mtu_set(struct rte_eth_dev *eth_dev, uint16_t mtu);
+int cxgbe_dev_start(struct rte_eth_dev *eth_dev);
+int cxgbe_dev_link_update(struct rte_eth_dev *eth_dev,
+			  int wait_to_complete);
+const uint32_t *cxgbe_dev_supported_ptypes_get(struct rte_eth_dev *eth_dev);
+#endif /* _CXGBE_PFVF_H_ */
diff --git a/drivers/net/cxgbe/cxgbevf_ethdev.c b/drivers/net/cxgbe/cxgbevf_ethdev.c
new file mode 100644
index 000000000..7d5b8c07f
--- /dev/null
+++ b/drivers/net/cxgbe/cxgbevf_ethdev.c
@@ -0,0 +1,91 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2018 Chelsio Communications.
+ * All rights reserved.
+ */
+
+#include <rte_ethdev_driver.h>
+#include <rte_ethdev_pci.h>
+
+#include "cxgbe.h"
+#include "cxgbe_pfvf.h"
+
+/*
+ * Macros needed to support the PCI Device ID Table ...
+ */
+#define CH_PCI_DEVICE_ID_TABLE_DEFINE_BEGIN \
+	static const struct rte_pci_id cxgb4vf_pci_tbl[] = {
+#define CH_PCI_DEVICE_ID_FUNCTION 0x8
+
+#define PCI_VENDOR_ID_CHELSIO 0x1425
+
+#define CH_PCI_ID_TABLE_ENTRY(devid) \
+		{ RTE_PCI_DEVICE(PCI_VENDOR_ID_CHELSIO, (devid)) }
+
+#define CH_PCI_DEVICE_ID_TABLE_DEFINE_END \
+		{ .vendor_id = 0, } \
+	}
+
+/*
+ *... and the PCI ID Table itself ...
+ */
+#include "t4_pci_id_tbl.h"
+
+static const struct eth_dev_ops cxgbevf_eth_dev_ops = {
+	.dev_start              = cxgbe_dev_start,
+	.dev_stop               = cxgbe_dev_stop,
+	.dev_close              = cxgbe_dev_close,
+	.promiscuous_enable     = cxgbe_dev_promiscuous_enable,
+	.promiscuous_disable    = cxgbe_dev_promiscuous_disable,
+	.allmulticast_enable    = cxgbe_dev_allmulticast_enable,
+	.allmulticast_disable   = cxgbe_dev_allmulticast_disable,
+	.dev_configure          = cxgbe_dev_configure,
+	.dev_infos_get          = cxgbe_dev_info_get,
+	.dev_supported_ptypes_get = cxgbe_dev_supported_ptypes_get,
+	.link_update            = cxgbe_dev_link_update,
+	.mtu_set                = cxgbe_dev_mtu_set,
+	.tx_queue_setup         = cxgbe_dev_tx_queue_setup,
+	.tx_queue_start         = cxgbe_dev_tx_queue_start,
+	.tx_queue_stop          = cxgbe_dev_tx_queue_stop,
+	.tx_queue_release       = cxgbe_dev_tx_queue_release,
+	.rx_queue_setup         = cxgbe_dev_rx_queue_setup,
+	.rx_queue_start         = cxgbe_dev_rx_queue_start,
+	.rx_queue_stop          = cxgbe_dev_rx_queue_stop,
+	.rx_queue_release       = cxgbe_dev_rx_queue_release,
+};
+
+/*
+ * Initialize driver
+ * It returns 0 on success.
+ */
+static int eth_cxgbevf_dev_init(struct rte_eth_dev *eth_dev)
+{
+	CXGBE_FUNC_TRACE();
+
+	eth_dev->dev_ops = &cxgbevf_eth_dev_ops;
+
+	/* XXX: Do probe */
+	return -EIO;
+}
+
+static int eth_cxgbevf_pci_probe(struct rte_pci_driver *pci_drv __rte_unused,
+				 struct rte_pci_device *pci_dev)
+{
+	return rte_eth_dev_pci_generic_probe(pci_dev, sizeof(struct port_info),
+					     eth_cxgbevf_dev_init);
+}
+
+static int eth_cxgbevf_pci_remove(struct rte_pci_device *pci_dev)
+{
+	return rte_eth_dev_pci_generic_remove(pci_dev, NULL);
+}
+
+static struct rte_pci_driver rte_cxgbevf_pmd = {
+	.id_table = cxgb4vf_pci_tbl,
+	.drv_flags = RTE_PCI_DRV_NEED_MAPPING,
+	.probe = eth_cxgbevf_pci_probe,
+	.remove = eth_cxgbevf_pci_remove,
+};
+
+RTE_PMD_REGISTER_PCI(net_cxgbevf, rte_cxgbevf_pmd);
+RTE_PMD_REGISTER_PCI_TABLE(net_cxgbevf, cxgb4vf_pci_tbl);
+RTE_PMD_REGISTER_KMOD_DEP(net_cxgbevf, "* igb_uio | vfio-pci");
-- 
2.14.1

^ permalink raw reply related	[flat|nested] 26+ messages in thread

* [PATCH 02/13] cxgbe: add VF firmware mailbox support
  2018-03-10 22:48 [PATCH 00/13] cxgbe: add CXGBE VF PMD and updates Rahul Lakkireddy
  2018-03-10 22:48 ` [PATCH 01/13] cxgbe: add skeleton VF driver Rahul Lakkireddy
@ 2018-03-10 22:48 ` Rahul Lakkireddy
  2018-03-10 22:48 ` [PATCH 03/13] cxgbe: add base for enabling VF ports Rahul Lakkireddy
                   ` (12 subsequent siblings)
  14 siblings, 0 replies; 26+ messages in thread
From: Rahul Lakkireddy @ 2018-03-10 22:48 UTC (permalink / raw)
  To: dev; +Cc: kumaras, nirranjan, indranil

From: Kumar Sanghvi <kumaras@chelsio.com>

Add firmware mailbox communication support for VF.  Add is_pf4()
to check if driver is attached to PF4.  Use is_pf4() to determine
whether to use PF or VF mailbox communication.

Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
---
 drivers/net/cxgbe/Makefile       |   1 +
 drivers/net/cxgbe/base/common.h  |  21 ++++
 drivers/net/cxgbe/base/t4_hw.c   |  72 +++++++++----
 drivers/net/cxgbe/base/t4_regs.h |   4 +
 drivers/net/cxgbe/base/t4vf_hw.c | 224 +++++++++++++++++++++++++++++++++++++++
 drivers/net/cxgbe/base/t4vf_hw.h |  14 +++
 6 files changed, 318 insertions(+), 18 deletions(-)
 create mode 100644 drivers/net/cxgbe/base/t4vf_hw.c
 create mode 100644 drivers/net/cxgbe/base/t4vf_hw.h

diff --git a/drivers/net/cxgbe/Makefile b/drivers/net/cxgbe/Makefile
index 397977e02..66c445ea3 100644
--- a/drivers/net/cxgbe/Makefile
+++ b/drivers/net/cxgbe/Makefile
@@ -84,5 +84,6 @@ SRCS-$(CONFIG_RTE_LIBRTE_CXGBE_PMD) += cxgbevf_ethdev.c
 SRCS-$(CONFIG_RTE_LIBRTE_CXGBE_PMD) += cxgbe_main.c
 SRCS-$(CONFIG_RTE_LIBRTE_CXGBE_PMD) += sge.c
 SRCS-$(CONFIG_RTE_LIBRTE_CXGBE_PMD) += t4_hw.c
+SRCS-$(CONFIG_RTE_LIBRTE_CXGBE_PMD) += t4vf_hw.c
 
 include $(RTE_SDK)/mk/rte.lib.mk
diff --git a/drivers/net/cxgbe/base/common.h b/drivers/net/cxgbe/base/common.h
index 365e9e692..b67cf0b12 100644
--- a/drivers/net/cxgbe/base/common.h
+++ b/drivers/net/cxgbe/base/common.h
@@ -36,6 +36,7 @@
 
 #include "cxgbe_compat.h"
 #include "t4_hw.h"
+#include "t4vf_hw.h"
 #include "t4_chip_type.h"
 #include "t4fw_interface.h"
 
@@ -290,6 +291,11 @@ static inline int t4_wait_op_done(struct adapter *adapter, int reg, u32 mask,
 				   delay, NULL);
 }
 
+static inline int is_pf4(struct adapter *adap)
+{
+	return adap->pf == 4;
+}
+
 #define for_each_port(adapter, iter) \
 	for (iter = 0; iter < (adapter)->params.nports; ++iter)
 
@@ -400,6 +406,21 @@ static inline int t4_wr_mbox_ns(struct adapter *adap, int mbox, const void *cmd,
 	return t4_wr_mbox_meat(adap, mbox, cmd, size, rpl, false);
 }
 
+int t4vf_wr_mbox_core(struct adapter *, const void *, int, void *, bool);
+
+static inline int t4vf_wr_mbox(struct adapter *adapter, const void *cmd,
+			       int size, void *rpl)
+{
+	return t4vf_wr_mbox_core(adapter, cmd, size, rpl, true);
+}
+
+static inline int t4vf_wr_mbox_ns(struct adapter *adapter, const void *cmd,
+				  int size, void *rpl)
+{
+	return t4vf_wr_mbox_core(adapter, cmd, size, rpl, false);
+}
+
+
 void t4_read_indirect(struct adapter *adap, unsigned int addr_reg,
 		      unsigned int data_reg, u32 *vals, unsigned int nregs,
 		      unsigned int start_idx);
diff --git a/drivers/net/cxgbe/base/t4_hw.c b/drivers/net/cxgbe/base/t4_hw.c
index c66e2a6f7..8e2b8a7d5 100644
--- a/drivers/net/cxgbe/base/t4_hw.c
+++ b/drivers/net/cxgbe/base/t4_hw.c
@@ -2342,7 +2342,11 @@ int t4_config_rss_range(struct adapter *adapter, int mbox, unsigned int viid,
 		 * Send this portion of the RRS table update to the firmware;
 		 * bail out on any errors.
 		 */
-		ret = t4_wr_mbox(adapter, mbox, &cmd, sizeof(cmd), NULL);
+		if (is_pf4(adapter))
+			ret = t4_wr_mbox(adapter, mbox, &cmd, sizeof(cmd),
+					 NULL);
+		else
+			ret = t4vf_wr_mbox(adapter, &cmd, sizeof(cmd), NULL);
 		if (ret)
 			return ret;
 	}
@@ -2372,7 +2376,10 @@ int t4_config_vi_rss(struct adapter *adapter, int mbox, unsigned int viid,
 	c.retval_len16 = cpu_to_be32(FW_LEN16(c));
 	c.u.basicvirtual.defaultq_to_udpen = cpu_to_be32(flags |
 			V_FW_RSS_VI_CONFIG_CMD_DEFAULTQ(defq));
-	return t4_wr_mbox(adapter, mbox, &c, sizeof(c), NULL);
+	if (is_pf4(adapter))
+		return t4_wr_mbox(adapter, mbox, &c, sizeof(c), NULL);
+	else
+		return t4vf_wr_mbox(adapter, &c, sizeof(c), NULL);
 }
 
 /**
@@ -4082,12 +4089,17 @@ int t4_free_vi(struct adapter *adap, unsigned int mbox, unsigned int pf,
 
 	memset(&c, 0, sizeof(c));
 	c.op_to_vfn = cpu_to_be32(V_FW_CMD_OP(FW_VI_CMD) | F_FW_CMD_REQUEST |
-				  F_FW_CMD_EXEC | V_FW_VI_CMD_PFN(pf) |
-				  V_FW_VI_CMD_VFN(vf));
+				  F_FW_CMD_EXEC);
+	if (is_pf4(adap))
+		c.op_to_vfn |= cpu_to_be32(V_FW_VI_CMD_PFN(pf) |
+					   V_FW_VI_CMD_VFN(vf));
 	c.alloc_to_len16 = cpu_to_be32(F_FW_VI_CMD_FREE | FW_LEN16(c));
 	c.type_to_viid = cpu_to_be16(V_FW_VI_CMD_VIID(viid));
 
-	return t4_wr_mbox(adap, mbox, &c, sizeof(c), &c);
+	if (is_pf4(adap))
+		return t4_wr_mbox(adap, mbox, &c, sizeof(c), &c);
+	else
+		return t4vf_wr_mbox(adap, &c, sizeof(c), NULL);
 }
 
 /**
@@ -4133,7 +4145,11 @@ int t4_set_rxmode(struct adapter *adap, unsigned int mbox, unsigned int viid,
 			    V_FW_VI_RXMODE_CMD_ALLMULTIEN(all_multi) |
 			    V_FW_VI_RXMODE_CMD_BROADCASTEN(bcast) |
 			    V_FW_VI_RXMODE_CMD_VLANEXEN(vlanex));
-	return t4_wr_mbox_meat(adap, mbox, &c, sizeof(c), NULL, sleep_ok);
+	if (is_pf4(adap))
+		return t4_wr_mbox_meat(adap, mbox, &c, sizeof(c), NULL,
+				       sleep_ok);
+	else
+		return t4vf_wr_mbox(adap, &c, sizeof(c), NULL);
 }
 
 /**
@@ -4180,7 +4196,10 @@ int t4_change_mac(struct adapter *adap, unsigned int mbox, unsigned int viid,
 				      V_FW_VI_MAC_CMD_IDX(idx));
 	memcpy(p->macaddr, addr, sizeof(p->macaddr));
 
-	ret = t4_wr_mbox(adap, mbox, &c, sizeof(c), &c);
+	if (is_pf4(adap))
+		ret = t4_wr_mbox(adap, mbox, &c, sizeof(c), &c);
+	else
+		ret = t4vf_wr_mbox(adap, &c, sizeof(c), &c);
 	if (ret == 0) {
 		ret = G_FW_VI_MAC_CMD_IDX(be16_to_cpu(p->valid_to_idx));
 		if (ret >= max_mac_addr)
@@ -4214,7 +4233,10 @@ int t4_enable_vi_params(struct adapter *adap, unsigned int mbox,
 				     V_FW_VI_ENABLE_CMD_EEN(tx_en) |
 				     V_FW_VI_ENABLE_CMD_DCB_INFO(dcb_en) |
 				     FW_LEN16(c));
-	return t4_wr_mbox_ns(adap, mbox, &c, sizeof(c), NULL);
+	if (is_pf4(adap))
+		return t4_wr_mbox_ns(adap, mbox, &c, sizeof(c), NULL);
+	else
+		return t4vf_wr_mbox_ns(adap, &c, sizeof(c), NULL);
 }
 
 /**
@@ -4255,15 +4277,20 @@ int t4_iq_start_stop(struct adapter *adap, unsigned int mbox, bool start,
 
 	memset(&c, 0, sizeof(c));
 	c.op_to_vfn = cpu_to_be32(V_FW_CMD_OP(FW_IQ_CMD) | F_FW_CMD_REQUEST |
-				  F_FW_CMD_EXEC | V_FW_IQ_CMD_PFN(pf) |
-				  V_FW_IQ_CMD_VFN(vf));
+				  F_FW_CMD_EXEC);
 	c.alloc_to_len16 = cpu_to_be32(V_FW_IQ_CMD_IQSTART(start) |
 				       V_FW_IQ_CMD_IQSTOP(!start) |
 				       FW_LEN16(c));
 	c.iqid = cpu_to_be16(iqid);
 	c.fl0id = cpu_to_be16(fl0id);
 	c.fl1id = cpu_to_be16(fl1id);
-	return t4_wr_mbox(adap, mbox, &c, sizeof(c), NULL);
+	if (is_pf4(adap)) {
+		c.op_to_vfn |= cpu_to_be32(V_FW_IQ_CMD_PFN(pf) |
+					   V_FW_IQ_CMD_VFN(vf));
+		return t4_wr_mbox(adap, mbox, &c, sizeof(c), NULL);
+	} else {
+		return t4vf_wr_mbox(adap, &c, sizeof(c), NULL);
+	}
 }
 
 /**
@@ -4287,14 +4314,19 @@ int t4_iq_free(struct adapter *adap, unsigned int mbox, unsigned int pf,
 
 	memset(&c, 0, sizeof(c));
 	c.op_to_vfn = cpu_to_be32(V_FW_CMD_OP(FW_IQ_CMD) | F_FW_CMD_REQUEST |
-				  F_FW_CMD_EXEC | V_FW_IQ_CMD_PFN(pf) |
-				  V_FW_IQ_CMD_VFN(vf));
+				  F_FW_CMD_EXEC);
+	if (is_pf4(adap))
+		c.op_to_vfn |= cpu_to_be32(V_FW_IQ_CMD_PFN(pf) |
+					   V_FW_IQ_CMD_VFN(vf));
 	c.alloc_to_len16 = cpu_to_be32(F_FW_IQ_CMD_FREE | FW_LEN16(c));
 	c.type_to_iqandstindex = cpu_to_be32(V_FW_IQ_CMD_TYPE(iqtype));
 	c.iqid = cpu_to_be16(iqid);
 	c.fl0id = cpu_to_be16(fl0id);
 	c.fl1id = cpu_to_be16(fl1id);
-	return t4_wr_mbox(adap, mbox, &c, sizeof(c), NULL);
+	if (is_pf4(adap))
+		return t4_wr_mbox(adap, mbox, &c, sizeof(c), NULL);
+	else
+		return t4vf_wr_mbox(adap, &c, sizeof(c), NULL);
 }
 
 /**
@@ -4314,12 +4346,16 @@ int t4_eth_eq_free(struct adapter *adap, unsigned int mbox, unsigned int pf,
 
 	memset(&c, 0, sizeof(c));
 	c.op_to_vfn = cpu_to_be32(V_FW_CMD_OP(FW_EQ_ETH_CMD) |
-				  F_FW_CMD_REQUEST | F_FW_CMD_EXEC |
-				  V_FW_EQ_ETH_CMD_PFN(pf) |
-				  V_FW_EQ_ETH_CMD_VFN(vf));
+				  F_FW_CMD_REQUEST | F_FW_CMD_EXEC);
+	if (is_pf4(adap))
+		c.op_to_vfn |= cpu_to_be32(V_FW_IQ_CMD_PFN(pf) |
+					   V_FW_IQ_CMD_VFN(vf));
 	c.alloc_to_len16 = cpu_to_be32(F_FW_EQ_ETH_CMD_FREE | FW_LEN16(c));
 	c.eqid_pkd = cpu_to_be32(V_FW_EQ_ETH_CMD_EQID(eqid));
-	return t4_wr_mbox(adap, mbox, &c, sizeof(c), NULL);
+	if (is_pf4(adap))
+		return t4_wr_mbox(adap, mbox, &c, sizeof(c), NULL);
+	else
+		return t4vf_wr_mbox(adap, &c, sizeof(c), NULL);
 }
 
 /**
diff --git a/drivers/net/cxgbe/base/t4_regs.h b/drivers/net/cxgbe/base/t4_regs.h
index 0f0bca910..657d8a7d7 100644
--- a/drivers/net/cxgbe/base/t4_regs.h
+++ b/drivers/net/cxgbe/base/t4_regs.h
@@ -443,6 +443,8 @@
 /* registers for module CIM */
 #define CIM_BASE_ADDR 0x7b00
 
+#define A_CIM_VF_EXT_MAILBOX_CTRL 0x0
+
 #define A_CIM_PF_MAILBOX_DATA 0x240
 #define A_CIM_PF_MAILBOX_CTRL 0x280
 
@@ -462,6 +464,8 @@
 #define V_UPCRST(x) ((x) << S_UPCRST)
 #define F_UPCRST    V_UPCRST(1U)
 
+#define NUM_CIM_PF_MAILBOX_DATA_INSTANCES 16
+
 /* registers for module TP */
 #define A_TP_OUT_CONFIG 0x7d04
 
diff --git a/drivers/net/cxgbe/base/t4vf_hw.c b/drivers/net/cxgbe/base/t4vf_hw.c
new file mode 100644
index 000000000..909659ce7
--- /dev/null
+++ b/drivers/net/cxgbe/base/t4vf_hw.c
@@ -0,0 +1,224 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2018 Chelsio Communications.
+ * All rights reserved.
+ */
+
+#include <rte_ethdev_driver.h>
+#include <rte_ether.h>
+
+#include "common.h"
+#include "t4_regs.h"
+
+/*
+ * Get the reply to a mailbox command and store it in @rpl in big-endian order.
+ */
+static void get_mbox_rpl(struct adapter *adap, __be64 *rpl, int nflit,
+			 u32 mbox_addr)
+{
+	for ( ; nflit; nflit--, mbox_addr += 8)
+		*rpl++ = htobe64(t4_read_reg64(adap, mbox_addr));
+}
+
+/**
+ * t4vf_wr_mbox_core - send a command to FW through the mailbox
+ * @adapter: the adapter
+ * @cmd: the command to write
+ * @size: command length in bytes
+ * @rpl: where to optionally store the reply
+ * @sleep_ok: if true we may sleep while awaiting command completion
+ *
+ * Sends the given command to FW through the mailbox and waits for the
+ * FW to execute the command.  If @rpl is not %NULL it is used to store
+ * the FW's reply to the command.  The command and its optional reply
+ * are of the same length.  FW can take up to 500 ms to respond.
+ * @sleep_ok determines whether we may sleep while awaiting the response.
+ * If sleeping is allowed we use progressive backoff otherwise we spin.
+ *
+ * The return value is 0 on success or a negative errno on failure.  A
+ * failure can happen either because we are not able to execute the
+ * command or FW executes it but signals an error.  In the latter case
+ * the return value is the error code indicated by FW (negated).
+ */
+int t4vf_wr_mbox_core(struct adapter *adapter,
+		      const void __attribute__((__may_alias__)) *cmd,
+		      int size, void *rpl, bool sleep_ok)
+{
+	/*
+	 * We delay in small increments at first in an effort to maintain
+	 * responsiveness for simple, fast executing commands but then back
+	 * off to larger delays to a maximum retry delay.
+	 */
+	static const int delay[] = {
+		1, 1, 3, 5, 10, 10, 20, 50, 100
+	};
+
+
+	u32 mbox_ctl = T4VF_CIM_BASE_ADDR + A_CIM_VF_EXT_MAILBOX_CTRL;
+	__be64 cmd_rpl[MBOX_LEN / 8];
+	struct mbox_entry entry;
+	unsigned int delay_idx;
+	u32 v, mbox_data;
+	const __be64 *p;
+	int i, ret;
+	int ms;
+
+	/* In T6, mailbox size is changed to 128 bytes to avoid
+	 * invalidating the entire prefetch buffer.
+	 */
+	if (CHELSIO_CHIP_VERSION(adapter->params.chip) <= CHELSIO_T5)
+		mbox_data = T4VF_MBDATA_BASE_ADDR;
+	else
+		mbox_data = T6VF_MBDATA_BASE_ADDR;
+
+	/*
+	 * Commands must be multiples of 16 bytes in length and may not be
+	 * larger than the size of the Mailbox Data register array.
+	 */
+	if ((size % 16) != 0 ||
+			size > NUM_CIM_VF_MAILBOX_DATA_INSTANCES * 4)
+		return -EINVAL;
+
+	/*
+	 * Queue ourselves onto the mailbox access list.  When our entry is at
+	 * the front of the list, we have rights to access the mailbox.  So we
+	 * wait [for a while] till we're at the front [or bail out with an
+	 * EBUSY] ...
+	 */
+	t4_os_atomic_add_tail(&entry, &adapter->mbox_list, &adapter->mbox_lock);
+
+	delay_idx = 0;
+	ms = delay[0];
+
+	for (i = 0; ; i += ms) {
+		/*
+		 * If we've waited too long, return a busy indication.  This
+		 * really ought to be based on our initial position in the
+		 * mailbox access list but this is a start.  We very rarely
+		 * contend on access to the mailbox ...
+		 */
+		if (i > (2 * FW_CMD_MAX_TIMEOUT)) {
+			t4_os_atomic_list_del(&entry, &adapter->mbox_list,
+					      &adapter->mbox_lock);
+			ret = -EBUSY;
+			return ret;
+		}
+
+		/*
+		 * If we're at the head, break out and start the mailbox
+		 * protocol.
+		 */
+		if (t4_os_list_first_entry(&adapter->mbox_list) == &entry)
+			break;
+
+		/*
+		 * Delay for a bit before checking again ...
+		 */
+		if (sleep_ok) {
+			ms = delay[delay_idx];  /* last element may repeat */
+			if (delay_idx < ARRAY_SIZE(delay) - 1)
+				delay_idx++;
+			msleep(ms);
+		} else {
+			rte_delay_ms(ms);
+		}
+	}
+
+	/*
+	 * Loop trying to get ownership of the mailbox.  Return an error
+	 * if we can't gain ownership.
+	 */
+	v = G_MBOWNER(t4_read_reg(adapter, mbox_ctl));
+	for (i = 0; v == X_MBOWNER_NONE && i < 3; i++)
+		v = G_MBOWNER(t4_read_reg(adapter, mbox_ctl));
+
+	if (v != X_MBOWNER_PL) {
+		t4_os_atomic_list_del(&entry, &adapter->mbox_list,
+				      &adapter->mbox_lock);
+		ret = (v == X_MBOWNER_FW) ? -EBUSY : -ETIMEDOUT;
+		return ret;
+	}
+
+	/*
+	 * Write the command array into the Mailbox Data register array and
+	 * transfer ownership of the mailbox to the firmware.
+	 */
+	for (i = 0, p = cmd; i < size; i += 8)
+		t4_write_reg64(adapter, mbox_data + i, be64_to_cpu(*p++));
+
+	t4_read_reg(adapter, mbox_data);          /* flush write */
+	t4_write_reg(adapter, mbox_ctl,
+			F_MBMSGVALID | V_MBOWNER(X_MBOWNER_FW));
+	t4_read_reg(adapter, mbox_ctl);          /* flush write */
+	delay_idx = 0;
+	ms = delay[0];
+
+	/*
+	 * Spin waiting for firmware to acknowledge processing our command.
+	 */
+	for (i = 0; i < FW_CMD_MAX_TIMEOUT; i++) {
+		if (sleep_ok) {
+			ms = delay[delay_idx];  /* last element may repeat */
+			if (delay_idx < ARRAY_SIZE(delay) - 1)
+				delay_idx++;
+			msleep(ms);
+		} else {
+			rte_delay_ms(ms);
+		}
+
+		/*
+		 * If we're the owner, see if this is the reply we wanted.
+		 */
+		v = t4_read_reg(adapter, mbox_ctl);
+		if (G_MBOWNER(v) == X_MBOWNER_PL) {
+			/*
+			 * If the Message Valid bit isn't on, revoke ownership
+			 * of the mailbox and continue waiting for our reply.
+			 */
+			if ((v & F_MBMSGVALID) == 0) {
+				t4_write_reg(adapter, mbox_ctl,
+					     V_MBOWNER(X_MBOWNER_NONE));
+				continue;
+			}
+
+			/*
+			 * We now have our reply.  Extract the command return
+			 * value, copy the reply back to our caller's buffer
+			 * (if specified) and revoke ownership of the mailbox.
+			 * We return the (negated) firmware command return
+			 * code (this depends on FW_SUCCESS == 0).  (Again we
+			 * avoid clogging the log with FW_VI_STATS_CMD
+			 * reply results.)
+			 */
+
+			/*
+			 * Retrieve the command reply and release the mailbox.
+			 */
+			get_mbox_rpl(adapter, cmd_rpl, size / 8, mbox_data);
+			t4_write_reg(adapter, mbox_ctl,
+				     V_MBOWNER(X_MBOWNER_NONE));
+			t4_os_atomic_list_del(&entry, &adapter->mbox_list,
+					      &adapter->mbox_lock);
+
+			/* return value in high-order host-endian word */
+			v = be64_to_cpu(cmd_rpl[0]);
+
+			if (rpl) {
+				/* request bit in high-order BE word */
+				WARN_ON((be32_to_cpu(*(const u32 *)cmd)
+					 & F_FW_CMD_REQUEST) == 0);
+				memcpy(rpl, cmd_rpl, size);
+			}
+			return -((int)G_FW_CMD_RETVAL(v));
+		}
+	}
+
+	/*
+	 * We timed out.  Return the error ...
+	 */
+	dev_err(adapter, "command %#x timed out\n",
+		*(const u8 *)cmd);
+	dev_err(adapter, "    Control = %#x\n", t4_read_reg(adapter, mbox_ctl));
+	t4_os_atomic_list_del(&entry, &adapter->mbox_list, &adapter->mbox_lock);
+	ret = -ETIMEDOUT;
+	return ret;
+}
diff --git a/drivers/net/cxgbe/base/t4vf_hw.h b/drivers/net/cxgbe/base/t4vf_hw.h
new file mode 100644
index 000000000..d16b452ab
--- /dev/null
+++ b/drivers/net/cxgbe/base/t4vf_hw.h
@@ -0,0 +1,14 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2018 Chelsio Communications.
+ * All rights reserved.
+ */
+
+#ifndef __T4VF_HW_H
+#define __T4VF_HW_H
+
+#define T4VF_CIM_BASE_ADDR	0x0300
+#define T4VF_MBDATA_BASE_ADDR	0x0240
+#define T6VF_MBDATA_BASE_ADDR	0x0280
+
+#define NUM_CIM_VF_MAILBOX_DATA_INSTANCES NUM_CIM_PF_MAILBOX_DATA_INSTANCES
+#endif /* __T4VF_HW_H */
-- 
2.14.1

^ permalink raw reply related	[flat|nested] 26+ messages in thread

* [PATCH 03/13] cxgbe: add base for enabling VF ports
  2018-03-10 22:48 [PATCH 00/13] cxgbe: add CXGBE VF PMD and updates Rahul Lakkireddy
  2018-03-10 22:48 ` [PATCH 01/13] cxgbe: add skeleton VF driver Rahul Lakkireddy
  2018-03-10 22:48 ` [PATCH 02/13] cxgbe: add VF firmware mailbox support Rahul Lakkireddy
@ 2018-03-10 22:48 ` Rahul Lakkireddy
  2018-03-10 22:48 ` [PATCH 04/13] cxgbe: add probe to initialize VF devices Rahul Lakkireddy
                   ` (11 subsequent siblings)
  14 siblings, 0 replies; 26+ messages in thread
From: Rahul Lakkireddy @ 2018-03-10 22:48 UTC (permalink / raw)
  To: dev; +Cc: kumaras, nirranjan, indranil

From: Kumar Sanghvi <kumaras@chelsio.com>

Add base to enable VF ports in subsequent patches.

Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
---
 drivers/net/cxgbe/base/common.h         |  31 +++
 drivers/net/cxgbe/base/t4_hw.c          |   9 +-
 drivers/net/cxgbe/base/t4_regs.h        |   2 +
 drivers/net/cxgbe/base/t4fw_interface.h |  73 ++++++
 drivers/net/cxgbe/base/t4vf_hw.c        | 434 ++++++++++++++++++++++++++++++++
 drivers/net/cxgbe/base/t4vf_hw.h        |   1 +
 6 files changed, 544 insertions(+), 6 deletions(-)

diff --git a/drivers/net/cxgbe/base/common.h b/drivers/net/cxgbe/base/common.h
index b67cf0b12..3307827c8 100644
--- a/drivers/net/cxgbe/base/common.h
+++ b/drivers/net/cxgbe/base/common.h
@@ -210,12 +210,29 @@ struct arch_specific_params {
 	u16 mps_tcam_size;
 };
 
+/*
+ * Maximum resources provisioned for a PCI VF.
+ */
+struct vf_resources {
+	unsigned int nvi;		/* N virtual interfaces */
+	unsigned int neq;		/* N egress Qs */
+	unsigned int nethctrl;		/* N egress ETH or CTRL Qs */
+	unsigned int niqflint;		/* N ingress Qs/w free list(s) & intr */
+	unsigned int niq;		/* N ingress Qs */
+	unsigned int tc;		/* PCI-E traffic class */
+	unsigned int pmask;		/* port access rights mask */
+	unsigned int nexactf;		/* N exact MPS filters */
+	unsigned int r_caps;		/* read capabilities */
+	unsigned int wx_caps;		/* write/execute capabilities */
+};
+
 struct adapter_params {
 	struct sge_params sge;
 	struct tp_params  tp;
 	struct vpd_params vpd;
 	struct pci_params pci;
 	struct devlog_params devlog;
+	struct vf_resources vfres;
 	enum pcie_memwin drv_memwin;
 
 	unsigned int sf_size;             /* serial flash size in bytes */
@@ -312,9 +329,11 @@ int t4_fw_hello(struct adapter *adap, unsigned int mbox, unsigned int evt_mbox,
 		enum dev_master master, enum dev_state *state);
 int t4_fw_bye(struct adapter *adap, unsigned int mbox);
 int t4_fw_reset(struct adapter *adap, unsigned int mbox, int reset);
+int t4vf_fw_reset(struct adapter *adap);
 int t4_fw_halt(struct adapter *adap, unsigned int mbox, int reset);
 int t4_fw_restart(struct adapter *adap, unsigned int mbox, int reset);
 int t4_fl_pkt_align(struct adapter *adap);
+int t4vf_get_vfres(struct adapter *adap);
 int t4_fixup_host_params_compat(struct adapter *adap, unsigned int page_size,
 				unsigned int cache_line_size,
 				enum chip_type chip_compat);
@@ -324,6 +343,12 @@ int t4_fw_initialize(struct adapter *adap, unsigned int mbox);
 int t4_query_params(struct adapter *adap, unsigned int mbox, unsigned int pf,
 		    unsigned int vf, unsigned int nparams, const u32 *params,
 		    u32 *val);
+int t4vf_query_params(struct adapter *adap, unsigned int nparams,
+		      const u32 *params, u32 *vals);
+int t4vf_get_dev_params(struct adapter *adap);
+int t4vf_get_vpd_params(struct adapter *adap);
+int t4vf_set_params(struct adapter *adapter, unsigned int nparams,
+		    const u32 *params, const u32 *vals);
 int t4_set_params_timeout(struct adapter *adap, unsigned int mbox,
 			  unsigned int pf, unsigned int vf,
 			  unsigned int nparams, const u32 *params,
@@ -440,13 +465,17 @@ void t4_get_port_stats_offset(struct adapter *adap, int idx,
 			      struct port_stats *stats,
 			      struct port_stats *offset);
 void t4_clr_port_stats(struct adapter *adap, int idx);
+void init_link_config(struct link_config *lc, fw_port_cap32_t pcaps,
+		      fw_port_cap32_t acaps);
 void t4_reset_link_config(struct adapter *adap, int idx);
 int t4_get_version_info(struct adapter *adapter);
 void t4_dump_version_info(struct adapter *adapter);
 int t4_get_flash_params(struct adapter *adapter);
 int t4_get_chip_type(struct adapter *adap, int ver);
 int t4_prep_adapter(struct adapter *adapter);
+int t4vf_prep_adapter(struct adapter *adapter);
 int t4_port_init(struct adapter *adap, int mbox, int pf, int vf);
+int t4vf_port_init(struct adapter *adap);
 int t4_init_rss_mode(struct adapter *adap, int mbox);
 int t4_config_rss_range(struct adapter *adapter, int mbox, unsigned int viid,
 			int start, int n, const u16 *rspq, unsigned int nrspq);
@@ -469,8 +498,10 @@ int t4_init_tp_params(struct adapter *adap);
 int t4_filter_field_shift(const struct adapter *adap, unsigned int filter_sel);
 int t4_handle_fw_rpl(struct adapter *adap, const __be64 *rpl);
 unsigned int t4_get_regs_len(struct adapter *adap);
+unsigned int t4vf_get_pf_from_vf(struct adapter *adap);
 void t4_get_regs(struct adapter *adap, void *buf, size_t buf_size);
 int t4_seeprom_read(struct adapter *adapter, u32 addr, u32 *data);
 int t4_seeprom_write(struct adapter *adapter, u32 addr, u32 data);
 int t4_seeprom_wp(struct adapter *adapter, int enable);
+fw_port_cap32_t fwcaps16_to_caps32(fw_port_cap16_t caps16);
 #endif /* __CHELSIO_COMMON_H */
diff --git a/drivers/net/cxgbe/base/t4_hw.c b/drivers/net/cxgbe/base/t4_hw.c
index 8e2b8a7d5..436a9953d 100644
--- a/drivers/net/cxgbe/base/t4_hw.c
+++ b/drivers/net/cxgbe/base/t4_hw.c
@@ -55,9 +55,6 @@
 #include "t4_regs_values.h"
 #include "t4fw_interface.h"
 
-static void init_link_config(struct link_config *lc, unsigned int pcaps,
-			     unsigned int acaps);
-
 /**
  * t4_read_mtu_tbl - returns the values in the HW path MTU table
  * @adap: the adapter
@@ -2804,7 +2801,7 @@ void t4_dump_version_info(struct adapter *adapter)
  *
  *     Returns the equivalent 32-bit Port Capabilities value.
  */
-static fw_port_cap32_t fwcaps16_to_caps32(fw_port_cap16_t caps16)
+fw_port_cap32_t fwcaps16_to_caps32(fw_port_cap16_t caps16)
 {
 	fw_port_cap32_t caps32 = 0;
 
@@ -4585,8 +4582,8 @@ void t4_reset_link_config(struct adapter *adap, int idx)
  * Initializes the SW state maintained for each link, including the link's
  * capabilities and default speed/flow-control/autonegotiation settings.
  */
-static void init_link_config(struct link_config *lc, fw_port_cap32_t pcaps,
-			     fw_port_cap32_t acaps)
+void init_link_config(struct link_config *lc, fw_port_cap32_t pcaps,
+		      fw_port_cap32_t acaps)
 {
 	lc->pcaps = pcaps;
 	lc->requested_speed = 0;
diff --git a/drivers/net/cxgbe/base/t4_regs.h b/drivers/net/cxgbe/base/t4_regs.h
index 657d8a7d7..f553d146d 100644
--- a/drivers/net/cxgbe/base/t4_regs.h
+++ b/drivers/net/cxgbe/base/t4_regs.h
@@ -852,6 +852,7 @@
 #define F_PFCIM    V_PFCIM(1U)
 
 #define A_PL_WHOAMI 0x19400
+#define A_PL_VF_WHOAMI 0x0
 
 #define A_PL_RST 0x19428
 
@@ -866,6 +867,7 @@
 #define F_PIORSTMODE    V_PIORSTMODE(1U)
 
 #define A_PL_REV 0x1943c
+#define A_PL_VF_REV 0x4
 
 #define S_REV    0
 #define M_REV    0xfU
diff --git a/drivers/net/cxgbe/base/t4fw_interface.h b/drivers/net/cxgbe/base/t4fw_interface.h
index d71c5a4af..b40bfb960 100644
--- a/drivers/net/cxgbe/base/t4fw_interface.h
+++ b/drivers/net/cxgbe/base/t4fw_interface.h
@@ -178,6 +178,7 @@ enum fw_cmd_opcodes {
 	FW_INITIALIZE_CMD              = 0x06,
 	FW_CAPS_CONFIG_CMD             = 0x07,
 	FW_PARAMS_CMD                  = 0x08,
+	FW_PFVF_CMD		       = 0x09,
 	FW_IQ_CMD                      = 0x10,
 	FW_EQ_ETH_CMD                  = 0x12,
 	FW_VI_CMD                      = 0x14,
@@ -190,6 +191,10 @@ enum fw_cmd_opcodes {
 	FW_DEBUG_CMD                   = 0x81,
 };
 
+enum fw_cmd_cap {
+	FW_CMD_CAP_PORT		= 0x04,
+};
+
 /*
  * Generic command header flit0
  */
@@ -484,6 +489,8 @@ enum fw_params_mnem {
 enum fw_params_param_dev {
 	FW_PARAMS_PARAM_DEV_CCLK	= 0x00, /* chip core clock in khz */
 	FW_PARAMS_PARAM_DEV_PORTVEC	= 0x01, /* the port vector */
+	FW_PARAMS_PARAM_DEV_FWREV	= 0x0B, /* fw version */
+	FW_PARAMS_PARAM_DEV_TPREV	= 0x0C, /* tp version */
 	FW_PARAMS_PARAM_DEV_ULPTX_MEMWRITE_DSGL = 0x17,
 };
 
@@ -533,6 +540,10 @@ enum fw_params_param_dmaq {
 #define G_FW_PARAMS_PARAM_YZ(x) \
 	(((x) >> S_FW_PARAMS_PARAM_YZ) & M_FW_PARAMS_PARAM_YZ)
 
+#define S_FW_PARAMS_PARAM_XYZ		0
+#define M_FW_PARAMS_PARAM_XYZ		0xffffff
+#define V_FW_PARAMS_PARAM_XYZ(x)	((x) << S_FW_PARAMS_PARAM_XYZ)
+
 struct fw_params_cmd {
 	__be32 op_to_vfn;
 	__be32 retval_len16;
@@ -554,6 +565,68 @@ struct fw_params_cmd {
 #define G_FW_PARAMS_CMD_VFN(x)	\
 	(((x) >> S_FW_PARAMS_CMD_VFN) & M_FW_PARAMS_CMD_VFN)
 
+struct fw_pfvf_cmd {
+	__be32 op_to_vfn;
+	__be32 retval_len16;
+	__be32 niqflint_niq;
+	__be32 type_to_neq;
+	__be32 tc_to_nexactf;
+	__be32 r_caps_to_nethctrl;
+	__be16 nricq;
+	__be16 nriqp;
+	__be32 r4;
+};
+
+#define S_FW_PFVF_CMD_NIQFLINT          20
+#define M_FW_PFVF_CMD_NIQFLINT          0xfff
+#define G_FW_PFVF_CMD_NIQFLINT(x)       \
+	(((x) >> S_FW_PFVF_CMD_NIQFLINT) & M_FW_PFVF_CMD_NIQFLINT)
+
+#define S_FW_PFVF_CMD_NIQ               0
+#define M_FW_PFVF_CMD_NIQ               0xfffff
+#define G_FW_PFVF_CMD_NIQ(x)            \
+	(((x) >> S_FW_PFVF_CMD_NIQ) & M_FW_PFVF_CMD_NIQ)
+
+#define S_FW_PFVF_CMD_PMASK             20
+#define M_FW_PFVF_CMD_PMASK             0xf
+#define G_FW_PFVF_CMD_PMASK(x)          \
+	(((x) >> S_FW_PFVF_CMD_PMASK) & M_FW_PFVF_CMD_PMASK)
+
+#define S_FW_PFVF_CMD_NEQ               0
+#define M_FW_PFVF_CMD_NEQ               0xfffff
+#define G_FW_PFVF_CMD_NEQ(x)            \
+	(((x) >> S_FW_PFVF_CMD_NEQ) & M_FW_PFVF_CMD_NEQ)
+
+#define S_FW_PFVF_CMD_TC                24
+#define M_FW_PFVF_CMD_TC                0xff
+#define G_FW_PFVF_CMD_TC(x)             \
+	(((x) >> S_FW_PFVF_CMD_TC) & M_FW_PFVF_CMD_TC)
+
+#define S_FW_PFVF_CMD_NVI               16
+#define M_FW_PFVF_CMD_NVI               0xff
+#define G_FW_PFVF_CMD_NVI(x)            \
+	(((x) >> S_FW_PFVF_CMD_NVI) & M_FW_PFVF_CMD_NVI)
+
+#define S_FW_PFVF_CMD_NEXACTF           0
+#define M_FW_PFVF_CMD_NEXACTF           0xffff
+#define G_FW_PFVF_CMD_NEXACTF(x)        \
+	(((x) >> S_FW_PFVF_CMD_NEXACTF) & M_FW_PFVF_CMD_NEXACTF)
+
+#define S_FW_PFVF_CMD_R_CAPS            24
+#define M_FW_PFVF_CMD_R_CAPS            0xff
+#define G_FW_PFVF_CMD_R_CAPS(x)         \
+	(((x) >> S_FW_PFVF_CMD_R_CAPS) & M_FW_PFVF_CMD_R_CAPS)
+
+#define S_FW_PFVF_CMD_WX_CAPS           16
+#define M_FW_PFVF_CMD_WX_CAPS           0xff
+#define G_FW_PFVF_CMD_WX_CAPS(x)        \
+	(((x) >> S_FW_PFVF_CMD_WX_CAPS) & M_FW_PFVF_CMD_WX_CAPS)
+
+#define S_FW_PFVF_CMD_NETHCTRL          0
+#define M_FW_PFVF_CMD_NETHCTRL          0xffff
+#define G_FW_PFVF_CMD_NETHCTRL(x)       \
+	(((x) >> S_FW_PFVF_CMD_NETHCTRL) & M_FW_PFVF_CMD_NETHCTRL)
+
 /*
  * ingress queue type; the first 1K ingress queues can have associated 0,
  * 1 or 2 free lists and an interrupt, all other ingress queues lack these
diff --git a/drivers/net/cxgbe/base/t4vf_hw.c b/drivers/net/cxgbe/base/t4vf_hw.c
index 909659ce7..6f222c4a1 100644
--- a/drivers/net/cxgbe/base/t4vf_hw.c
+++ b/drivers/net/cxgbe/base/t4vf_hw.c
@@ -9,6 +9,34 @@
 #include "common.h"
 #include "t4_regs.h"
 
+/**
+ * t4vf_wait_dev_ready - wait till to reads of registers work
+ *
+ * Wait for the device to become ready (signified by our "who am I" register
+ * returning a value other than all 1's).  Return an error if it doesn't
+ * become ready ...
+ */
+static int t4vf_wait_dev_ready(struct adapter *adapter)
+{
+	const u32 whoami = T4VF_PL_BASE_ADDR + A_PL_VF_WHOAMI;
+	const u32 notready1 = 0xffffffff;
+	const u32 notready2 = 0xeeeeeeee;
+	u32 val;
+
+	val = t4_read_reg(adapter, whoami);
+	if (val != notready1 && val != notready2)
+		return 0;
+
+	msleep(500);
+	val = t4_read_reg(adapter, whoami);
+	if (val != notready1 && val != notready2)
+		return 0;
+
+	dev_err(adapter, "Device didn't become ready for access, whoami = %#x\n",
+		val);
+	return -EIO;
+}
+
 /*
  * Get the reply to a mailbox command and store it in @rpl in big-endian order.
  */
@@ -222,3 +250,409 @@ int t4vf_wr_mbox_core(struct adapter *adapter,
 	ret = -ETIMEDOUT;
 	return ret;
 }
+
+/**
+ * t4vf_fw_reset - issue a reset to FW
+ * @adapter: the adapter
+ *
+ * Issues a reset command to FW.  For a Physical Function this would
+ * result in the Firmware resetting all of its state.  For a Virtual
+ * Function this just resets the state associated with the VF.
+ */
+int t4vf_fw_reset(struct adapter *adapter)
+{
+	struct fw_reset_cmd cmd;
+
+	memset(&cmd, 0, sizeof(cmd));
+	cmd.op_to_write = cpu_to_be32(V_FW_CMD_OP(FW_RESET_CMD) |
+				      F_FW_CMD_WRITE);
+	cmd.retval_len16 = cpu_to_be32(V_FW_CMD_LEN16(FW_LEN16(cmd)));
+	return t4vf_wr_mbox(adapter, &cmd, sizeof(cmd), NULL);
+}
+
+/**
+ * t4vf_prep_adapter - prepare SW and HW for operation
+ * @adapter: the adapter
+ *
+ * Initialize adapter SW state for the various HW modules, set initial
+ * values for some adapter tunables, take PHYs out of reset, and
+ * initialize the MDIO interface.
+ */
+int t4vf_prep_adapter(struct adapter *adapter)
+{
+	u32 pl_vf_rev;
+	int ret, ver;
+
+	ret = t4vf_wait_dev_ready(adapter);
+	if (ret < 0)
+		return ret;
+
+	/*
+	 * Default port and clock for debugging in case we can't reach
+	 * firmware.
+	 */
+	adapter->params.nports = 1;
+	adapter->params.vfres.pmask = 1;
+	adapter->params.vpd.cclk = 50000;
+
+	pl_vf_rev = G_REV(t4_read_reg(adapter, A_PL_VF_REV));
+	adapter->params.pci.device_id = adapter->pdev->id.device_id;
+	adapter->params.pci.vendor_id = adapter->pdev->id.vendor_id;
+
+	/*
+	 * WE DON'T NEED adapter->params.chip CODE ONCE PL_REV CONTAINS
+	 * ADAPTER (VERSION << 4 | REVISION)
+	 */
+	ver = CHELSIO_PCI_ID_VER(adapter->params.pci.device_id);
+	adapter->params.chip = 0;
+	switch (ver) {
+	case CHELSIO_T5:
+		adapter->params.chip |= CHELSIO_CHIP_CODE(CHELSIO_T5,
+							  pl_vf_rev);
+		adapter->params.arch.sge_fl_db = F_DBPRIO | F_DBTYPE;
+		adapter->params.arch.mps_tcam_size =
+			NUM_MPS_T5_CLS_SRAM_L_INSTANCES;
+		break;
+	case CHELSIO_T6:
+		adapter->params.chip |= CHELSIO_CHIP_CODE(CHELSIO_T6,
+							  pl_vf_rev);
+		adapter->params.arch.sge_fl_db = 0;
+		adapter->params.arch.mps_tcam_size =
+			NUM_MPS_T5_CLS_SRAM_L_INSTANCES;
+		break;
+	default:
+		dev_err(adapter, "%s: Device %d is not supported\n",
+			__func__, adapter->params.pci.device_id);
+		return -EINVAL;
+	}
+	return 0;
+}
+
+/**
+ * t4vf_query_params - query FW or device parameters
+ * @adapter: the adapter
+ * @nparams: the number of parameters
+ * @params: the parameter names
+ * @vals: the parameter values
+ *
+ * Reads the values of firmware or device parameters.  Up to 7 parameters
+ * can be queried at once.
+ */
+int t4vf_query_params(struct adapter *adapter, unsigned int nparams,
+		      const u32 *params, u32 *vals)
+{
+	struct fw_params_cmd cmd, rpl;
+	struct fw_params_param *p;
+	unsigned int i;
+	size_t len16;
+	int ret;
+
+	if (nparams > 7)
+		return -EINVAL;
+
+	memset(&cmd, 0, sizeof(cmd));
+	cmd.op_to_vfn = cpu_to_be32(V_FW_CMD_OP(FW_PARAMS_CMD) |
+				    F_FW_CMD_REQUEST |
+				    F_FW_CMD_READ);
+	len16 = DIV_ROUND_UP(offsetof(struct fw_params_cmd,
+			     param[nparams]), 16);
+	cmd.retval_len16 = cpu_to_be32(V_FW_CMD_LEN16(len16));
+	for (i = 0, p = &cmd.param[0]; i < nparams; i++, p++)
+		p->mnem = cpu_to_be32(*params++);
+	ret = t4vf_wr_mbox(adapter, &cmd, sizeof(cmd), &rpl);
+	if (ret == 0)
+		for (i = 0, p = &rpl.param[0]; i < nparams; i++, p++)
+			*vals++ = be32_to_cpu(p->val);
+	return ret;
+}
+
+/**
+ * t4vf_get_vpd_params - retrieve device VPD paremeters
+ * @adapter: the adapter
+ *
+ * Retrives various device Vital Product Data parameters.  The parameters
+ * are stored in @adapter->params.vpd.
+ */
+int t4vf_get_vpd_params(struct adapter *adapter)
+{
+	struct vpd_params *vpd_params = &adapter->params.vpd;
+	u32 params[7], vals[7];
+	int v;
+
+	params[0] = (V_FW_PARAMS_MNEM(FW_PARAMS_MNEM_DEV) |
+		     V_FW_PARAMS_PARAM_X(FW_PARAMS_PARAM_DEV_CCLK));
+	v = t4vf_query_params(adapter, 1, params, vals);
+	if (v != FW_SUCCESS)
+		return v;
+	vpd_params->cclk = vals[0];
+	dev_debug(adapter, "%s: vpd_params->cclk = %u\n",
+		  __func__, vpd_params->cclk);
+	return 0;
+}
+
+/**
+ * t4vf_get_dev_params - retrieve device paremeters
+ * @adapter: the adapter
+ *
+ * Retrives fw and tp version.
+ */
+int t4vf_get_dev_params(struct adapter *adapter)
+{
+	u32 params[7], vals[7];
+	int v;
+
+	params[0] = (V_FW_PARAMS_MNEM(FW_PARAMS_MNEM_DEV) |
+		     V_FW_PARAMS_PARAM_X(FW_PARAMS_PARAM_DEV_FWREV));
+	params[1] = (V_FW_PARAMS_MNEM(FW_PARAMS_MNEM_DEV) |
+		     V_FW_PARAMS_PARAM_X(FW_PARAMS_PARAM_DEV_TPREV));
+	v = t4vf_query_params(adapter, 2, params, vals);
+	if (v != FW_SUCCESS)
+		return v;
+	adapter->params.fw_vers = vals[0];
+	adapter->params.tp_vers = vals[1];
+
+	dev_info(adapter, "Firmware version: %u.%u.%u.%u\n",
+		 G_FW_HDR_FW_VER_MAJOR(adapter->params.fw_vers),
+		 G_FW_HDR_FW_VER_MINOR(adapter->params.fw_vers),
+		 G_FW_HDR_FW_VER_MICRO(adapter->params.fw_vers),
+		 G_FW_HDR_FW_VER_BUILD(adapter->params.fw_vers));
+
+	dev_info(adapter, "TP Microcode version: %u.%u.%u.%u\n",
+		 G_FW_HDR_FW_VER_MAJOR(adapter->params.tp_vers),
+		 G_FW_HDR_FW_VER_MINOR(adapter->params.tp_vers),
+		 G_FW_HDR_FW_VER_MICRO(adapter->params.tp_vers),
+		 G_FW_HDR_FW_VER_BUILD(adapter->params.tp_vers));
+	return 0;
+}
+
+/**
+ * t4vf_set_params - sets FW or device parameters
+ * @adapter: the adapter
+ * @nparams: the number of parameters
+ * @params: the parameter names
+ * @vals: the parameter values
+ *
+ * Sets the values of firmware or device parameters.  Up to 7 parameters
+ * can be specified at once.
+ */
+int t4vf_set_params(struct adapter *adapter, unsigned int nparams,
+		    const u32 *params, const u32 *vals)
+{
+	struct fw_params_param *p;
+	struct fw_params_cmd cmd;
+	unsigned int i;
+	size_t len16;
+
+	if (nparams > 7)
+		return -EINVAL;
+
+	memset(&cmd, 0, sizeof(cmd));
+	cmd.op_to_vfn = cpu_to_be32(V_FW_CMD_OP(FW_PARAMS_CMD) |
+				    F_FW_CMD_REQUEST |
+				    F_FW_CMD_WRITE);
+	len16 = DIV_ROUND_UP(offsetof(struct fw_params_cmd,
+			     param[nparams]), 16);
+	cmd.retval_len16 = cpu_to_be32(V_FW_CMD_LEN16(len16));
+	for (i = 0, p = &cmd.param[0]; i < nparams; i++, p++) {
+		p->mnem = cpu_to_be32(*params++);
+		p->val = cpu_to_be32(*vals++);
+	}
+	return t4vf_wr_mbox(adapter, &cmd, sizeof(cmd), NULL);
+}
+
+unsigned int t4vf_get_pf_from_vf(struct adapter *adapter)
+{
+	u32 whoami;
+
+	whoami = t4_read_reg(adapter, T4VF_PL_BASE_ADDR + A_PL_VF_WHOAMI);
+	return (CHELSIO_CHIP_VERSION(adapter->params.chip) <= CHELSIO_T5 ?
+			G_SOURCEPF(whoami) : G_T6_SOURCEPF(whoami));
+}
+
+/**
+ * t4vf_get_vfres - retrieve VF resource limits
+ * @adapter: the adapter
+ *
+ * Retrieves configured resource limits and capabilities for a virtual
+ * function.  The results are stored in @adapter->vfres.
+ */
+int t4vf_get_vfres(struct adapter *adapter)
+{
+	struct vf_resources *vfres = &adapter->params.vfres;
+	struct fw_pfvf_cmd cmd, rpl;
+	u32 word;
+	int v;
+
+	/*
+	 * Execute PFVF Read command to get VF resource limits; bail out early
+	 * with error on command failure.
+	 */
+	memset(&cmd, 0, sizeof(cmd));
+	cmd.op_to_vfn = cpu_to_be32(V_FW_CMD_OP(FW_PFVF_CMD) |
+				    F_FW_CMD_REQUEST |
+				    F_FW_CMD_READ);
+	cmd.retval_len16 = cpu_to_be32(FW_LEN16(cmd));
+	v = t4vf_wr_mbox(adapter, &cmd, sizeof(cmd), &rpl);
+	if (v != FW_SUCCESS)
+		return v;
+
+	/*
+	 * Extract VF resource limits and return success.
+	 */
+	word = be32_to_cpu(rpl.niqflint_niq);
+	vfres->niqflint = G_FW_PFVF_CMD_NIQFLINT(word);
+	vfres->niq = G_FW_PFVF_CMD_NIQ(word);
+
+	word = be32_to_cpu(rpl.type_to_neq);
+	vfres->neq = G_FW_PFVF_CMD_NEQ(word);
+	vfres->pmask = G_FW_PFVF_CMD_PMASK(word);
+
+	word = be32_to_cpu(rpl.tc_to_nexactf);
+	vfres->tc = G_FW_PFVF_CMD_TC(word);
+	vfres->nvi = G_FW_PFVF_CMD_NVI(word);
+	vfres->nexactf = G_FW_PFVF_CMD_NEXACTF(word);
+
+	word = be32_to_cpu(rpl.r_caps_to_nethctrl);
+	vfres->r_caps = G_FW_PFVF_CMD_R_CAPS(word);
+	vfres->wx_caps = G_FW_PFVF_CMD_WX_CAPS(word);
+	vfres->nethctrl = G_FW_PFVF_CMD_NETHCTRL(word);
+	return 0;
+}
+
+static int t4vf_alloc_vi(struct adapter *adapter, int port_id)
+{
+	struct fw_vi_cmd cmd, rpl;
+	int v;
+
+	/*
+	 * Execute a VI command to allocate Virtual Interface and return its
+	 * VIID.
+	 */
+	memset(&cmd, 0, sizeof(cmd));
+	cmd.op_to_vfn = cpu_to_be32(V_FW_CMD_OP(FW_VI_CMD) |
+				    F_FW_CMD_REQUEST |
+				    F_FW_CMD_WRITE |
+				    F_FW_CMD_EXEC);
+	cmd.alloc_to_len16 = cpu_to_be32(FW_LEN16(cmd) |
+					 F_FW_VI_CMD_ALLOC);
+	cmd.portid_pkd = V_FW_VI_CMD_PORTID(port_id);
+	v = t4vf_wr_mbox(adapter, &cmd, sizeof(cmd), &rpl);
+	if (v != FW_SUCCESS)
+		return v;
+	return G_FW_VI_CMD_VIID(be16_to_cpu(rpl.type_to_viid));
+}
+
+int t4vf_port_init(struct adapter *adapter)
+{
+	unsigned int fw_caps = adapter->params.fw_caps_support;
+	struct fw_port_cmd port_cmd, port_rpl;
+	struct fw_vi_cmd vi_cmd, vi_rpl;
+	fw_port_cap32_t pcaps, acaps;
+	enum fw_port_type port_type;
+	int mdio_addr;
+	int ret, i;
+
+	for_each_port(adapter, i) {
+		struct port_info *p = adap2pinfo(adapter, i);
+
+		/*
+		 * If we haven't yet determined if we're talking to Firmware
+		 * which knows the new 32-bit Port Caps, it's time to find
+		 * out now.  This will also tell new Firmware to send us Port
+		 * Status Updates using the new 32-bit Port Capabilities
+		 * version of the Port Information message.
+		 */
+		if (fw_caps == FW_CAPS_UNKNOWN) {
+			u32 param, val;
+
+			param = (V_FW_PARAMS_MNEM(FW_PARAMS_MNEM_PFVF) |
+				 V_FW_PARAMS_PARAM_X
+					 (FW_PARAMS_PARAM_PFVF_PORT_CAPS32));
+			val = 1;
+			ret = t4vf_set_params(adapter, 1, &param, &val);
+			fw_caps = (ret == 0 ? FW_CAPS32 : FW_CAPS16);
+			adapter->params.fw_caps_support = fw_caps;
+		}
+
+		ret = t4vf_alloc_vi(adapter, p->port_id);
+		if (ret < 0) {
+			dev_err(&pdev->dev, "cannot allocate VI for port %d:"
+				" err=%d\n", p->port_id, ret);
+			return ret;
+		}
+		p->viid = ret;
+
+		/*
+		 * Execute a VI Read command to get our Virtual Interface
+		 * information like MAC address, etc.
+		 */
+		memset(&vi_cmd, 0, sizeof(vi_cmd));
+		vi_cmd.op_to_vfn = cpu_to_be32(V_FW_CMD_OP(FW_VI_CMD) |
+					       F_FW_CMD_REQUEST |
+					       F_FW_CMD_READ);
+		vi_cmd.alloc_to_len16 = cpu_to_be32(FW_LEN16(vi_cmd));
+		vi_cmd.type_to_viid = cpu_to_be16(V_FW_VI_CMD_VIID(p->viid));
+		ret = t4vf_wr_mbox(adapter, &vi_cmd, sizeof(vi_cmd), &vi_rpl);
+		if (ret != FW_SUCCESS)
+			return ret;
+
+		p->rss_size = G_FW_VI_CMD_RSSSIZE
+				(be16_to_cpu(vi_rpl.norss_rsssize));
+		t4_os_set_hw_addr(adapter, i, vi_rpl.mac);
+
+		/*
+		 * If we don't have read access to our port information, we're
+		 * done now.  Else, execute a PORT Read command to get it ...
+		 */
+		if (!(adapter->params.vfres.r_caps & FW_CMD_CAP_PORT))
+			return 0;
+
+		memset(&port_cmd, 0, sizeof(port_cmd));
+		port_cmd.op_to_portid = cpu_to_be32
+				(V_FW_CMD_OP(FW_PORT_CMD) | F_FW_CMD_REQUEST |
+				F_FW_CMD_READ |
+				V_FW_PORT_CMD_PORTID(p->port_id));
+		port_cmd.action_to_len16 = cpu_to_be32
+				(V_FW_PORT_CMD_ACTION(fw_caps == FW_CAPS16 ?
+					FW_PORT_ACTION_GET_PORT_INFO :
+					FW_PORT_ACTION_GET_PORT_INFO32) |
+					FW_LEN16(port_cmd));
+		ret = t4vf_wr_mbox(adapter, &port_cmd, sizeof(port_cmd),
+				   &port_rpl);
+		if (ret != FW_SUCCESS)
+			return ret;
+
+		/*
+		 * Extract the various fields from the Port Information message.
+		 */
+		if (fw_caps == FW_CAPS16) {
+			u32 lstatus = be32_to_cpu
+					(port_rpl.u.info.lstatus_to_modtype);
+
+			port_type = G_FW_PORT_CMD_PTYPE(lstatus);
+			mdio_addr = ((lstatus & F_FW_PORT_CMD_MDIOCAP) ?
+				      (int)G_FW_PORT_CMD_MDIOADDR(lstatus) :
+				      -1);
+			pcaps = fwcaps16_to_caps32
+					(be16_to_cpu(port_rpl.u.info.pcap));
+			acaps = fwcaps16_to_caps32
+					(be16_to_cpu(port_rpl.u.info.acap));
+		} else {
+			u32 lstatus32 = be32_to_cpu
+				(port_rpl.u.info32.lstatus32_to_cbllen32);
+
+			port_type = G_FW_PORT_CMD_PORTTYPE32(lstatus32);
+			mdio_addr = ((lstatus32 & F_FW_PORT_CMD_MDIOCAP32) ?
+				      (int)G_FW_PORT_CMD_MDIOADDR32(lstatus32) :
+				      -1);
+			pcaps = be32_to_cpu(port_rpl.u.info32.pcaps32);
+			acaps = be32_to_cpu(port_rpl.u.info32.acaps32);
+		}
+
+		p->port_type = port_type;
+		p->mdio_addr = mdio_addr;
+		p->mod_type = FW_PORT_MOD_TYPE_NA;
+		init_link_config(&p->link_cfg, pcaps, acaps);
+	}
+	return 0;
+}
diff --git a/drivers/net/cxgbe/base/t4vf_hw.h b/drivers/net/cxgbe/base/t4vf_hw.h
index d16b452ab..55e436e74 100644
--- a/drivers/net/cxgbe/base/t4vf_hw.h
+++ b/drivers/net/cxgbe/base/t4vf_hw.h
@@ -6,6 +6,7 @@
 #ifndef __T4VF_HW_H
 #define __T4VF_HW_H
 
+#define T4VF_PL_BASE_ADDR	0x0200
 #define T4VF_CIM_BASE_ADDR	0x0300
 #define T4VF_MBDATA_BASE_ADDR	0x0240
 #define T6VF_MBDATA_BASE_ADDR	0x0280
-- 
2.14.1

^ permalink raw reply related	[flat|nested] 26+ messages in thread

* [PATCH 04/13] cxgbe: add probe to initialize VF devices
  2018-03-10 22:48 [PATCH 00/13] cxgbe: add CXGBE VF PMD and updates Rahul Lakkireddy
                   ` (2 preceding siblings ...)
  2018-03-10 22:48 ` [PATCH 03/13] cxgbe: add base for enabling VF ports Rahul Lakkireddy
@ 2018-03-10 22:48 ` Rahul Lakkireddy
  2018-03-10 22:48 ` [PATCH 05/13] cxgbe: initialize SGE and queues for VF Rahul Lakkireddy
                   ` (10 subsequent siblings)
  14 siblings, 0 replies; 26+ messages in thread
From: Rahul Lakkireddy @ 2018-03-10 22:48 UTC (permalink / raw)
  To: dev; +Cc: kumaras, nirranjan, indranil

From: Kumar Sanghvi <kumaras@chelsio.com>

Add probe to initialize VF devices.  Separate init/de-init paths
for PF and VF.  Do firmware state initialization wrt VF and retrieve
various operational parameters by querying firmware.  Finally configure
and initialize ports.

Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
---
 drivers/net/cxgbe/Makefile         |   1 +
 drivers/net/cxgbe/base/adapter.h   |   1 +
 drivers/net/cxgbe/base/t4_hw.c     |   2 +-
 drivers/net/cxgbe/cxgbe.h          |   3 +
 drivers/net/cxgbe/cxgbe_main.c     |  17 ++-
 drivers/net/cxgbe/cxgbevf_ethdev.c |  65 ++++++++-
 drivers/net/cxgbe/cxgbevf_main.c   | 280 +++++++++++++++++++++++++++++++++++++
 7 files changed, 359 insertions(+), 10 deletions(-)
 create mode 100644 drivers/net/cxgbe/cxgbevf_main.c

diff --git a/drivers/net/cxgbe/Makefile b/drivers/net/cxgbe/Makefile
index 66c445ea3..fe177ff90 100644
--- a/drivers/net/cxgbe/Makefile
+++ b/drivers/net/cxgbe/Makefile
@@ -82,6 +82,7 @@ VPATH += $(SRCDIR)/base
 SRCS-$(CONFIG_RTE_LIBRTE_CXGBE_PMD) += cxgbe_ethdev.c
 SRCS-$(CONFIG_RTE_LIBRTE_CXGBE_PMD) += cxgbevf_ethdev.c
 SRCS-$(CONFIG_RTE_LIBRTE_CXGBE_PMD) += cxgbe_main.c
+SRCS-$(CONFIG_RTE_LIBRTE_CXGBE_PMD) += cxgbevf_main.c
 SRCS-$(CONFIG_RTE_LIBRTE_CXGBE_PMD) += sge.c
 SRCS-$(CONFIG_RTE_LIBRTE_CXGBE_PMD) += t4_hw.c
 SRCS-$(CONFIG_RTE_LIBRTE_CXGBE_PMD) += t4vf_hw.c
diff --git a/drivers/net/cxgbe/base/adapter.h b/drivers/net/cxgbe/base/adapter.h
index 169402836..6b2fc8b34 100644
--- a/drivers/net/cxgbe/base/adapter.h
+++ b/drivers/net/cxgbe/base/adapter.h
@@ -68,6 +68,7 @@ struct port_info {
 	u8     port_type;               /* firmware port type */
 	u8     mod_type;                /* firmware module type */
 	u8     port_id;                 /* physical port ID */
+	u8     pidx;			/* port index for this PF */
 	u8     tx_chan;                 /* associated channel */
 
 	u8     n_rx_qsets;              /* # of rx qsets */
diff --git a/drivers/net/cxgbe/base/t4_hw.c b/drivers/net/cxgbe/base/t4_hw.c
index 436a9953d..b3938943c 100644
--- a/drivers/net/cxgbe/base/t4_hw.c
+++ b/drivers/net/cxgbe/base/t4_hw.c
@@ -4488,7 +4488,7 @@ static void t4_handle_get_port_info(struct port_info *pi, const __be64 *rpl)
 		lc->auto_fec = fec;
 		pi->port_type = port_type;
 		pi->mod_type = mod_type;
-		t4_os_portmod_changed(adapter, pi->port_id);
+		t4_os_portmod_changed(adapter, pi->pidx);
 	}
 	if (link_ok != lc->link_ok || speed != lc->speed ||
 	    fc != lc->fc || fec != lc->fec) { /* something changed */
diff --git a/drivers/net/cxgbe/cxgbe.h b/drivers/net/cxgbe/cxgbe.h
index 489e09dda..040aca423 100644
--- a/drivers/net/cxgbe/cxgbe.h
+++ b/drivers/net/cxgbe/cxgbe.h
@@ -49,6 +49,7 @@
 #define CXGBE_DEFAULT_RSS_KEY_LEN     40 /* 320-bits */
 
 int cxgbe_probe(struct adapter *adapter);
+int cxgbevf_probe(struct adapter *adapter);
 void cxgbe_get_speed_caps(struct port_info *pi, u32 *speed_caps);
 int cxgbe_up(struct adapter *adap);
 int cxgbe_down(struct port_info *pi);
@@ -63,5 +64,7 @@ void cfg_queues(struct rte_eth_dev *eth_dev);
 int cfg_queue_count(struct rte_eth_dev *eth_dev);
 int setup_rss(struct port_info *pi);
 void cxgbe_enable_rx_queues(struct port_info *pi);
+void print_port_info(struct adapter *adap);
+void print_adapter_info(struct adapter *adap);
 
 #endif /* _CXGBE_H_ */
diff --git a/drivers/net/cxgbe/cxgbe_main.c b/drivers/net/cxgbe/cxgbe_main.c
index 5792032e7..8d78ad427 100644
--- a/drivers/net/cxgbe/cxgbe_main.c
+++ b/drivers/net/cxgbe/cxgbe_main.c
@@ -370,7 +370,7 @@ static int init_rss(struct adapter *adap)
 /**
  * Dump basic information about the adapter.
  */
-static void print_adapter_info(struct adapter *adap)
+void print_adapter_info(struct adapter *adap)
 {
 	/**
 	 * Hardware/Firmware/etc. Version/Revision IDs.
@@ -378,7 +378,7 @@ static void print_adapter_info(struct adapter *adap)
 	t4_dump_version_info(adap);
 }
 
-static void print_port_info(struct adapter *adap)
+void print_port_info(struct adapter *adap)
 {
 	int i;
 	char buf[80];
@@ -917,7 +917,7 @@ int link_start(struct port_info *pi)
 			ret = 0;
 		}
 	}
-	if (ret == 0)
+	if (ret == 0 && is_pf4(adapter))
 		ret = t4_link_l1cfg(adapter, adapter->mbox, pi->tx_chan,
 				    &pi->link_cfg);
 	if (ret == 0) {
@@ -1196,7 +1196,8 @@ int cxgbe_up(struct adapter *adap)
 {
 	enable_rx(adap, &adap->sge.fw_evtq);
 	t4_sge_tx_monitor_start(adap);
-	t4_intr_enable(adap);
+	if (is_pf4(adap))
+		t4_intr_enable(adap);
 	adap->flags |= FULL_INIT_DONE;
 
 	/* TODO: deadman watchdog ?? */
@@ -1217,7 +1218,7 @@ int cxgbe_down(struct port_info *pi)
 		return err;
 	}
 
-	t4_reset_link_config(adapter, pi->port_id);
+	t4_reset_link_config(adapter, pi->pidx);
 	return 0;
 }
 
@@ -1230,7 +1231,8 @@ void cxgbe_close(struct adapter *adapter)
 	int i;
 
 	if (adapter->flags & FULL_INIT_DONE) {
-		t4_intr_disable(adapter);
+		if (is_pf4(adapter))
+			t4_intr_disable(adapter);
 		t4_sge_tx_monitor_stop(adapter);
 		t4_free_sge_resources(adapter);
 		for_each_port(adapter, i) {
@@ -1248,7 +1250,7 @@ void cxgbe_close(struct adapter *adapter)
 		adapter->flags &= ~FULL_INIT_DONE;
 	}
 
-	if (adapter->flags & FW_OK)
+	if (is_pf4(adapter) && (adapter->flags & FW_OK))
 		t4_fw_bye(adapter, adapter->mbox);
 }
 
@@ -1355,6 +1357,7 @@ int cxgbe_probe(struct adapter *adapter)
 		pi->adapter = adapter;
 		pi->xact_addr_filt = -1;
 		pi->port_id = i;
+		pi->pidx = i;
 
 		pi->eth_dev->device = &adapter->pdev->device;
 		pi->eth_dev->dev_ops = adapter->eth_dev->dev_ops;
diff --git a/drivers/net/cxgbe/cxgbevf_ethdev.c b/drivers/net/cxgbe/cxgbevf_ethdev.c
index 7d5b8c07f..3b1deac52 100644
--- a/drivers/net/cxgbe/cxgbevf_ethdev.c
+++ b/drivers/net/cxgbe/cxgbevf_ethdev.c
@@ -59,12 +59,73 @@ static const struct eth_dev_ops cxgbevf_eth_dev_ops = {
  */
 static int eth_cxgbevf_dev_init(struct rte_eth_dev *eth_dev)
 {
+	struct port_info *pi = (struct port_info *)(eth_dev->data->dev_private);
+	struct rte_pci_device *pci_dev;
+	char name[RTE_ETH_NAME_MAX_LEN];
+	struct adapter *adapter = NULL;
+	int err = 0;
+
 	CXGBE_FUNC_TRACE();
 
 	eth_dev->dev_ops = &cxgbevf_eth_dev_ops;
+	eth_dev->rx_pkt_burst = NULL;
+	eth_dev->tx_pkt_burst = NULL;
+	pci_dev = RTE_ETH_DEV_TO_PCI(eth_dev);
+
+	/* for secondary processes, we attach to ethdevs allocated by primary
+	 * and do minimal initialization.
+	 */
+	if (rte_eal_process_type() != RTE_PROC_PRIMARY) {
+		int i;
+
+		for (i = 1; i < MAX_NPORTS; i++) {
+			struct rte_eth_dev *rest_eth_dev;
+			char namei[RTE_ETH_NAME_MAX_LEN];
+
+			snprintf(namei, sizeof(namei), "%s_%d",
+				 pci_dev->device.name, i);
+			rest_eth_dev = rte_eth_dev_attach_secondary(namei);
+			if (rest_eth_dev) {
+				rest_eth_dev->device = &pci_dev->device;
+				rest_eth_dev->dev_ops =
+					eth_dev->dev_ops;
+				rest_eth_dev->rx_pkt_burst =
+					eth_dev->rx_pkt_burst;
+				rest_eth_dev->tx_pkt_burst =
+					eth_dev->tx_pkt_burst;
+			}
+		}
+		return 0;
+	}
+
+	snprintf(name, sizeof(name), "cxgbevfadapter%d",
+		 eth_dev->data->port_id);
+	adapter = rte_zmalloc(name, sizeof(*adapter), 0);
+	if (!adapter)
+		return -1;
+
+	adapter->use_unpacked_mode = 1;
+	adapter->regs = (void *)pci_dev->mem_resource[0].addr;
+	if (!adapter->regs) {
+		dev_err(adapter, "%s: cannot map device registers\n", __func__);
+		err = -ENOMEM;
+		goto out_free_adapter;
+	}
+	adapter->pdev = pci_dev;
+	adapter->eth_dev = eth_dev;
+	pi->adapter = adapter;
+	err = cxgbevf_probe(adapter);
+	if (err) {
+		dev_err(adapter, "%s: cxgbevf probe failed with err %d\n",
+			__func__, err);
+		goto out_free_adapter;
+	}
+
+	return 0;
 
-	/* XXX: Do probe */
-	return -EIO;
+out_free_adapter:
+	rte_free(adapter);
+	return err;
 }
 
 static int eth_cxgbevf_pci_probe(struct rte_pci_driver *pci_drv __rte_unused,
diff --git a/drivers/net/cxgbe/cxgbevf_main.c b/drivers/net/cxgbe/cxgbevf_main.c
new file mode 100644
index 000000000..0624267c7
--- /dev/null
+++ b/drivers/net/cxgbe/cxgbevf_main.c
@@ -0,0 +1,280 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2018 Chelsio Communications.
+ * All rights reserved.
+ */
+
+#include <rte_ethdev_driver.h>
+#include <rte_ethdev_pci.h>
+#include <rte_malloc.h>
+
+#include "common.h"
+#include "t4_regs.h"
+#include "t4_msg.h"
+#include "cxgbe.h"
+
+/*
+ * Figure out how many Ports and Queue Sets we can support.  This depends on
+ * knowing our Virtual Function Resources and may be called a second time if
+ * we fall back from MSI-X to MSI Interrupt Mode.
+ */
+static void size_nports_qsets(struct adapter *adapter)
+{
+	struct vf_resources *vfres = &adapter->params.vfres;
+	unsigned int ethqsets, pmask_nports;
+
+	/*
+	 * The number of "ports" which we support is equal to the number of
+	 * Virtual Interfaces with which we've been provisioned.
+	 */
+	adapter->params.nports = vfres->nvi;
+	if (adapter->params.nports > MAX_NPORTS) {
+		dev_warn(adapter->pdev_dev, "only using %d of %d maximum"
+			 " allowed virtual interfaces\n", MAX_NPORTS,
+			 adapter->params.nports);
+		adapter->params.nports = MAX_NPORTS;
+	}
+
+	/*
+	 * We may have been provisioned with more VIs than the number of
+	 * ports we're allowed to access (our Port Access Rights Mask).
+	 * This is obviously a configuration conflict but we don't want to
+	 * do anything silly just because of that.
+	 */
+	pmask_nports = hweight32(adapter->params.vfres.pmask);
+	if (pmask_nports < adapter->params.nports) {
+		dev_warn(adapter->pdev_dev, "only using %d of %d provissioned"
+			 " virtual interfaces; limited by Port Access Rights"
+			 " mask %#x\n", pmask_nports, adapter->params.nports,
+			 adapter->params.vfres.pmask);
+		adapter->params.nports = pmask_nports;
+	}
+
+	/*
+	 * We need to reserve an Ingress Queue for the Asynchronous Firmware
+	 * Event Queue.
+	 *
+	 * For each Queue Set, we'll need the ability to allocate two Egress
+	 * Contexts -- one for the Ingress Queue Free List and one for the TX
+	 * Ethernet Queue.
+	 */
+	ethqsets = vfres->niqflint - 1;
+	if (vfres->nethctrl != ethqsets)
+		ethqsets = min(vfres->nethctrl, ethqsets);
+	if (vfres->neq < ethqsets * 2)
+		ethqsets = vfres->neq / 2;
+	if (ethqsets > MAX_ETH_QSETS)
+		ethqsets = MAX_ETH_QSETS;
+	adapter->sge.max_ethqsets = ethqsets;
+
+	if (adapter->sge.max_ethqsets < adapter->params.nports) {
+		dev_warn(adapter->pdev_dev, "only using %d of %d available"
+			 " virtual interfaces (too few Queue Sets)\n",
+			 adapter->sge.max_ethqsets, adapter->params.nports);
+		adapter->params.nports = adapter->sge.max_ethqsets;
+	}
+}
+
+static int adap_init0vf(struct adapter *adapter)
+{
+	u32 param, val = 0;
+	int err;
+
+	err = t4vf_fw_reset(adapter);
+	if (err < 0) {
+		dev_err(adapter->pdev_dev, "FW reset failed: err=%d\n", err);
+		return err;
+	}
+
+	/*
+	 * Grab basic operational parameters.  These will predominantly have
+	 * been set up by the Physical Function Driver or will be hard coded
+	 * into the adapter.  We just have to live with them ...  Note that
+	 * we _must_ get our VPD parameters before our SGE parameters because
+	 * we need to know the adapter's core clock from the VPD in order to
+	 * properly decode the SGE Timer Values.
+	 */
+	err = t4vf_get_dev_params(adapter);
+	if (err) {
+		dev_err(adapter->pdev_dev, "unable to retrieve adapter"
+			" device parameters: err=%d\n", err);
+		return err;
+	}
+
+	err = t4vf_get_vpd_params(adapter);
+	if (err) {
+		dev_err(adapter->pdev_dev, "unable to retrieve adapter"
+			" VPD parameters: err=%d\n", err);
+		return err;
+	}
+
+	adapter->pf = t4vf_get_pf_from_vf(adapter);
+
+	/* If we're running on newer firmware, let it know that we're
+	 * prepared to deal with encapsulated CPL messages.  Older
+	 * firmware won't understand this and we'll just get
+	 * unencapsulated messages ...
+	 */
+	param = V_FW_PARAMS_MNEM(FW_PARAMS_MNEM_PFVF) |
+		V_FW_PARAMS_PARAM_X(FW_PARAMS_PARAM_PFVF_CPLFW4MSG_ENCAP);
+	val = 1;
+	t4vf_set_params(adapter, 1, &param, &val);
+
+	/*
+	 * Grab our Virtual Interface resource allocation, extract the
+	 * features that we're interested in and do a bit of sanity testing on
+	 * what we discover.
+	 */
+	err = t4vf_get_vfres(adapter);
+	if (err) {
+		dev_err(adapter->pdev_dev, "unable to get virtual interface"
+			" resources: err=%d\n", err);
+		return err;
+	}
+
+	/*
+	 * Check for various parameter sanity issues.
+	 */
+	if (adapter->params.vfres.pmask == 0) {
+		dev_err(adapter->pdev_dev, "no port access configured\n"
+			"usable!\n");
+		return -EINVAL;
+	}
+	if (adapter->params.vfres.nvi == 0) {
+		dev_err(adapter->pdev_dev, "no virtual interfaces configured/"
+			"usable!\n");
+		return -EINVAL;
+	}
+
+	/*
+	 * Initialize nports and max_ethqsets now that we have our Virtual
+	 * Function Resources.
+	 */
+	size_nports_qsets(adapter);
+	adapter->flags |= FW_OK;
+	return 0;
+}
+
+int cxgbevf_probe(struct adapter *adapter)
+{
+	struct port_info *pi;
+	unsigned int pmask;
+	int err = 0;
+	int i;
+
+	t4_os_lock_init(&adapter->mbox_lock);
+	TAILQ_INIT(&adapter->mbox_list);
+	err = t4vf_prep_adapter(adapter);
+	if (err)
+		return err;
+
+	if (!is_t4(adapter->params.chip)) {
+		adapter->bar2 = (void *)adapter->pdev->mem_resource[2].addr;
+		if (!adapter->bar2) {
+			dev_err(adapter, "cannot map device bar2 region\n");
+			err = -ENOMEM;
+			return err;
+		}
+	}
+
+	err = adap_init0vf(adapter);
+	if (err) {
+		dev_err(adapter, "%s: Adapter initialization failed, error %d\n",
+				__func__, err);
+		goto out_free;
+	}
+
+	pmask = adapter->params.vfres.pmask;
+	for_each_port(adapter, i) {
+		const unsigned int numa_node = rte_socket_id();
+		char name[RTE_ETH_NAME_MAX_LEN];
+		struct rte_eth_dev *eth_dev;
+		int port_id;
+
+		if (pmask == 0)
+			break;
+		port_id = ffs(pmask) - 1;
+		pmask &= ~(1 << port_id);
+
+		snprintf(name, sizeof(name), "%s_%d",
+			 adapter->pdev->device.name, i);
+
+		if (i == 0) {
+			/* First port is already allocated by DPDK */
+			eth_dev = adapter->eth_dev;
+			goto allocate_mac;
+		}
+
+		/*
+		 * now do all data allocation - for eth_dev structure,
+		 * and internal (private) data for the remaining ports
+		 */
+
+		/* reserve an ethdev entry */
+		eth_dev = rte_eth_dev_allocate(name);
+		if (!eth_dev) {
+			err = -ENOMEM;
+			goto out_free;
+		}
+		eth_dev->data->dev_private =
+			rte_zmalloc_socket(name, sizeof(struct port_info),
+					   RTE_CACHE_LINE_SIZE, numa_node);
+		if (!eth_dev->data->dev_private)
+			goto out_free;
+
+allocate_mac:
+		pi = (struct port_info *)eth_dev->data->dev_private;
+		adapter->port[i] = pi;
+		pi->eth_dev = eth_dev;
+		pi->adapter = adapter;
+		pi->xact_addr_filt = -1;
+		pi->port_id = port_id;
+		pi->pidx = i;
+
+		pi->eth_dev->device = &adapter->pdev->device;
+		pi->eth_dev->dev_ops = adapter->eth_dev->dev_ops;
+		pi->eth_dev->tx_pkt_burst = adapter->eth_dev->tx_pkt_burst;
+		pi->eth_dev->rx_pkt_burst = adapter->eth_dev->rx_pkt_burst;
+
+		rte_eth_copy_pci_info(pi->eth_dev, adapter->pdev);
+		pi->eth_dev->data->mac_addrs = rte_zmalloc(name,
+							   ETHER_ADDR_LEN, 0);
+		if (!pi->eth_dev->data->mac_addrs) {
+			dev_err(adapter, "%s: Mem allocation failed for storing mac addr, aborting\n",
+				__func__);
+			err = -ENOMEM;
+			goto out_free;
+		}
+	}
+
+	if (adapter->flags & FW_OK) {
+		err = t4vf_port_init(adapter);
+		if (err) {
+			dev_err(adapter, "%s: t4_port_init failed with err %d\n",
+				__func__, err);
+			goto out_free;
+		}
+	}
+
+	cfg_queues(adapter->eth_dev);
+	print_adapter_info(adapter);
+	print_port_info(adapter);
+
+	return 0;
+
+out_free:
+	for_each_port(adapter, i) {
+		pi = adap2pinfo(adapter, i);
+		if (pi->viid != 0)
+			t4_free_vi(adapter, adapter->mbox, adapter->pf,
+				   0, pi->viid);
+		/* Skip first port since it'll be de-allocated by DPDK */
+		if (i == 0)
+			continue;
+		if (pi->eth_dev) {
+			if (pi->eth_dev->data->dev_private)
+				rte_free(pi->eth_dev->data->dev_private);
+			rte_eth_dev_release_port(pi->eth_dev);
+		}
+	}
+	return -err;
+}
-- 
2.14.1

^ permalink raw reply related	[flat|nested] 26+ messages in thread

* [PATCH 05/13] cxgbe: initialize SGE and queues for VF
  2018-03-10 22:48 [PATCH 00/13] cxgbe: add CXGBE VF PMD and updates Rahul Lakkireddy
                   ` (3 preceding siblings ...)
  2018-03-10 22:48 ` [PATCH 04/13] cxgbe: add probe to initialize VF devices Rahul Lakkireddy
@ 2018-03-10 22:48 ` Rahul Lakkireddy
  2018-03-10 22:48 ` [PATCH 06/13] cxgbe: enable RSS " Rahul Lakkireddy
                   ` (9 subsequent siblings)
  14 siblings, 0 replies; 26+ messages in thread
From: Rahul Lakkireddy @ 2018-03-10 22:48 UTC (permalink / raw)
  To: dev; +Cc: kumaras, nirranjan, indranil

From: Kumar Sanghvi <kumaras@chelsio.com>

Query firmware and initialize SGE parameters and enable queue
allocation for VF.  Calculate pcie channel and queue congestion
management for VF.

Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
---
 drivers/net/cxgbe/base/adapter.h        |   2 +
 drivers/net/cxgbe/base/common.h         |   1 +
 drivers/net/cxgbe/base/t4_regs.h        |  12 ++
 drivers/net/cxgbe/base/t4fw_interface.h |   6 +
 drivers/net/cxgbe/base/t4vf_hw.c        |  40 ++++++
 drivers/net/cxgbe/cxgbe_ethdev.c        |  13 +-
 drivers/net/cxgbe/cxgbe_main.c          |   3 +-
 drivers/net/cxgbe/cxgbevf_main.c        |   5 +
 drivers/net/cxgbe/sge.c                 | 239 +++++++++++++++++++++++++++++---
 9 files changed, 296 insertions(+), 25 deletions(-)

diff --git a/drivers/net/cxgbe/base/adapter.h b/drivers/net/cxgbe/base/adapter.h
index 6b2fc8b34..95752d1b4 100644
--- a/drivers/net/cxgbe/base/adapter.h
+++ b/drivers/net/cxgbe/base/adapter.h
@@ -249,6 +249,7 @@ struct sge_txq {
 	unsigned int equeidx;	   /* last sent credit request */
 	unsigned int last_pidx;	   /* last pidx recorded by tx monitor */
 	unsigned int last_coal_idx;/* last coal-idx recorded by tx monitor */
+	unsigned int abs_id;
 
 	int db_disabled;            /* doorbell state */
 	unsigned short db_pidx;     /* doorbell producer index */
@@ -719,6 +720,7 @@ int t4_eth_xmit(struct sge_eth_txq *txq, struct rte_mbuf *mbuf,
 int t4_ethrx_handler(struct sge_rspq *q, const __be64 *rsp,
 		     const struct pkt_gl *gl);
 int t4_sge_init(struct adapter *adap);
+int t4vf_sge_init(struct adapter *adap);
 int t4_sge_alloc_eth_txq(struct adapter *adap, struct sge_eth_txq *txq,
 			 struct rte_eth_dev *eth_dev, uint16_t queue_id,
 			 unsigned int iqid, int socket_id);
diff --git a/drivers/net/cxgbe/base/common.h b/drivers/net/cxgbe/base/common.h
index 3307827c8..d74903dfd 100644
--- a/drivers/net/cxgbe/base/common.h
+++ b/drivers/net/cxgbe/base/common.h
@@ -333,6 +333,7 @@ int t4vf_fw_reset(struct adapter *adap);
 int t4_fw_halt(struct adapter *adap, unsigned int mbox, int reset);
 int t4_fw_restart(struct adapter *adap, unsigned int mbox, int reset);
 int t4_fl_pkt_align(struct adapter *adap);
+int t4vf_fl_pkt_align(struct adapter *adap, u32 sge_control, u32 sge_control2);
 int t4vf_get_vfres(struct adapter *adap);
 int t4_fixup_host_params_compat(struct adapter *adap, unsigned int page_size,
 				unsigned int cache_line_size,
diff --git a/drivers/net/cxgbe/base/t4_regs.h b/drivers/net/cxgbe/base/t4_regs.h
index f553d146d..43d6a0c73 100644
--- a/drivers/net/cxgbe/base/t4_regs.h
+++ b/drivers/net/cxgbe/base/t4_regs.h
@@ -77,6 +77,7 @@
 #define SGE_BASE_ADDR 0x1000
 
 #define A_SGE_PF_KDOORBELL 0x0
+#define A_SGE_VF_KDOORBELL 0x0
 
 #define S_QID    15
 #define M_QID    0x1ffffU
@@ -103,6 +104,9 @@
 
 #define A_SGE_PF_GTS 0x4
 
+#define T4VF_SGE_BASE_ADDR 0x0000
+#define A_SGE_VF_GTS 0x4
+
 #define S_INGRESSQID    16
 #define M_INGRESSQID    0xffffU
 #define V_INGRESSQID(x) ((x) << S_INGRESSQID)
@@ -191,6 +195,8 @@
 #define V_QUEUESPERPAGEPF0(x) ((x) << S_QUEUESPERPAGEPF0)
 #define G_QUEUESPERPAGEPF0(x) (((x) >> S_QUEUESPERPAGEPF0) & M_QUEUESPERPAGEPF0)
 
+#define A_SGE_EGRESS_QUEUES_PER_PAGE_VF 0x1014
+
 #define S_ERR_CPL_EXCEED_IQE_SIZE    22
 #define V_ERR_CPL_EXCEED_IQE_SIZE(x) ((x) << S_ERR_CPL_EXCEED_IQE_SIZE)
 #define F_ERR_CPL_EXCEED_IQE_SIZE    V_ERR_CPL_EXCEED_IQE_SIZE(1U)
@@ -280,6 +286,11 @@
 
 #define A_SGE_CONM_CTRL 0x1094
 
+#define S_T6_EGRTHRESHOLDPACKING    16
+#define M_T6_EGRTHRESHOLDPACKING    0xffU
+#define G_T6_EGRTHRESHOLDPACKING(x) (((x) >> S_T6_EGRTHRESHOLDPACKING) & \
+				     M_T6_EGRTHRESHOLDPACKING)
+
 #define S_EGRTHRESHOLD    8
 #define M_EGRTHRESHOLD    0x3fU
 #define V_EGRTHRESHOLD(x) ((x) << S_EGRTHRESHOLD)
@@ -370,6 +381,7 @@
 #define G_STATSOURCE_T5(x) (((x) >> S_STATSOURCE_T5) & M_STATSOURCE_T5)
 
 #define A_SGE_INGRESS_QUEUES_PER_PAGE_PF 0x10f4
+#define A_SGE_INGRESS_QUEUES_PER_PAGE_VF 0x10f8
 
 #define A_SGE_CONTROL2 0x1124
 
diff --git a/drivers/net/cxgbe/base/t4fw_interface.h b/drivers/net/cxgbe/base/t4fw_interface.h
index b40bfb960..6c5c97270 100644
--- a/drivers/net/cxgbe/base/t4fw_interface.h
+++ b/drivers/net/cxgbe/base/t4fw_interface.h
@@ -480,6 +480,7 @@ struct fw_caps_config_cmd {
 enum fw_params_mnem {
 	FW_PARAMS_MNEM_DEV		= 1,	/* device params */
 	FW_PARAMS_MNEM_PFVF		= 2,	/* function params */
+	FW_PARAMS_MNEM_REG		= 3,	/* limited register access */
 	FW_PARAMS_MNEM_DMAQ		= 4,	/* dma queue params */
 };
 
@@ -887,6 +888,11 @@ struct fw_eq_eth_cmd {
 #define G_FW_EQ_ETH_CMD_EQID(x)	\
 	(((x) >> S_FW_EQ_ETH_CMD_EQID) & M_FW_EQ_ETH_CMD_EQID)
 
+#define S_FW_EQ_ETH_CMD_PHYSEQID        0
+#define M_FW_EQ_ETH_CMD_PHYSEQID        0xfffff
+#define G_FW_EQ_ETH_CMD_PHYSEQID(x)     \
+	(((x) >> S_FW_EQ_ETH_CMD_PHYSEQID) & M_FW_EQ_ETH_CMD_PHYSEQID)
+
 #define S_FW_EQ_ETH_CMD_FETCHRO		22
 #define M_FW_EQ_ETH_CMD_FETCHRO		0x1
 #define V_FW_EQ_ETH_CMD_FETCHRO(x)	((x) << S_FW_EQ_ETH_CMD_FETCHRO)
diff --git a/drivers/net/cxgbe/base/t4vf_hw.c b/drivers/net/cxgbe/base/t4vf_hw.c
index 6f222c4a1..8e48588b0 100644
--- a/drivers/net/cxgbe/base/t4vf_hw.c
+++ b/drivers/net/cxgbe/base/t4vf_hw.c
@@ -460,6 +460,46 @@ int t4vf_set_params(struct adapter *adapter, unsigned int nparams,
 	return t4vf_wr_mbox(adapter, &cmd, sizeof(cmd), NULL);
 }
 
+/**
+ * t4vf_fl_pkt_align - return the fl packet alignment
+ * @adapter: the adapter
+ *
+ * T4 has a single field to specify the packing and padding boundary.
+ * T5 onwards has separate fields for this and hence the alignment for
+ * next packet offset is maximum of these two.
+ */
+int t4vf_fl_pkt_align(struct adapter *adapter, u32 sge_control,
+		      u32 sge_control2)
+{
+	unsigned int ingpadboundary, ingpackboundary, fl_align, ingpad_shift;
+
+	/* T4 uses a single control field to specify both the PCIe Padding and
+	 * Packing Boundary.  T5 introduced the ability to specify these
+	 * separately.  The actual Ingress Packet Data alignment boundary
+	 * within Packed Buffer Mode is the maximum of these two
+	 * specifications.
+	 */
+	if (CHELSIO_CHIP_VERSION(adapter->params.chip) <= CHELSIO_T5)
+		ingpad_shift = X_INGPADBOUNDARY_SHIFT;
+	else
+		ingpad_shift = X_T6_INGPADBOUNDARY_SHIFT;
+
+	ingpadboundary = 1 << (G_INGPADBOUNDARY(sge_control) + ingpad_shift);
+
+	fl_align = ingpadboundary;
+	if (!is_t4(adapter->params.chip)) {
+		ingpackboundary = G_INGPACKBOUNDARY(sge_control2);
+		if (ingpackboundary == X_INGPACKBOUNDARY_16B)
+			ingpackboundary = 16;
+		else
+			ingpackboundary = 1 << (ingpackboundary +
+					X_INGPACKBOUNDARY_SHIFT);
+
+		fl_align = max(ingpadboundary, ingpackboundary);
+	}
+	return fl_align;
+}
+
 unsigned int t4vf_get_pf_from_vf(struct adapter *adapter)
 {
 	u32 whoami;
diff --git a/drivers/net/cxgbe/cxgbe_ethdev.c b/drivers/net/cxgbe/cxgbe_ethdev.c
index 7f523265d..16031f38d 100644
--- a/drivers/net/cxgbe/cxgbe_ethdev.c
+++ b/drivers/net/cxgbe/cxgbe_ethdev.c
@@ -481,9 +481,8 @@ int cxgbe_dev_tx_queue_setup(struct rte_eth_dev *eth_dev,
 	err = t4_sge_alloc_eth_txq(adapter, txq, eth_dev, queue_idx,
 				   s->fw_evtq.cntxt_id, socket_id);
 
-	dev_debug(adapter, "%s: txq->q.cntxt_id= %d err = %d\n",
-		  __func__, txq->q.cntxt_id, err);
-
+	dev_debug(adapter, "%s: txq->q.cntxt_id= %u txq->q.abs_id= %u err = %d\n",
+		  __func__, txq->q.cntxt_id, txq->q.abs_id, err);
 	return err;
 }
 
@@ -610,11 +609,13 @@ int cxgbe_dev_rx_queue_setup(struct rte_eth_dev *eth_dev,
 
 	err = t4_sge_alloc_rxq(adapter, &rxq->rspq, false, eth_dev, msi_idx,
 			       &rxq->fl, t4_ethrx_handler,
-			       t4_get_tp_ch_map(adapter, pi->tx_chan), mp,
+			       is_pf4(adapter) ?
+			       t4_get_tp_ch_map(adapter, pi->tx_chan) : 0, mp,
 			       queue_idx, socket_id);
 
-	dev_debug(adapter, "%s: err = %d; port_id = %d; cntxt_id = %u\n",
-		  __func__, err, pi->port_id, rxq->rspq.cntxt_id);
+	dev_debug(adapter, "%s: err = %d; port_id = %d; cntxt_id = %u; abs_id = %u\n",
+		  __func__, err, pi->port_id, rxq->rspq.cntxt_id,
+		  rxq->rspq.abs_id);
 	return err;
 }
 
diff --git a/drivers/net/cxgbe/cxgbe_main.c b/drivers/net/cxgbe/cxgbe_main.c
index 8d78ad427..d9e772717 100644
--- a/drivers/net/cxgbe/cxgbe_main.c
+++ b/drivers/net/cxgbe/cxgbe_main.c
@@ -1064,7 +1064,8 @@ int setup_rss(struct port_info *pi)
 static void enable_rx(struct adapter *adap, struct sge_rspq *q)
 {
 	/* 0-increment GTS to start the timer and enable interrupts */
-	t4_write_reg(adap, MYPF_REG(A_SGE_PF_GTS),
+	t4_write_reg(adap, is_pf4(adap) ? MYPF_REG(A_SGE_PF_GTS) :
+					  T4VF_SGE_BASE_ADDR + A_SGE_VF_GTS,
 		     V_SEINTARM(q->intr_params) |
 		     V_INGRESSQID(q->cntxt_id));
 }
diff --git a/drivers/net/cxgbe/cxgbevf_main.c b/drivers/net/cxgbe/cxgbevf_main.c
index 0624267c7..f4d0f4de5 100644
--- a/drivers/net/cxgbe/cxgbevf_main.c
+++ b/drivers/net/cxgbe/cxgbevf_main.c
@@ -108,6 +108,11 @@ static int adap_init0vf(struct adapter *adapter)
 	}
 
 	adapter->pf = t4vf_get_pf_from_vf(adapter);
+	err = t4vf_sge_init(adapter);
+	if (err) {
+		dev_err(adapter->pdev_dev, "error in sge init\n");
+		return err;
+	}
 
 	/* If we're running on newer firmware, let it know that we're
 	 * prepared to deal with encapsulated CPL messages.  Older
diff --git a/drivers/net/cxgbe/sge.c b/drivers/net/cxgbe/sge.c
index 6ff8bc46b..aba1a49f3 100644
--- a/drivers/net/cxgbe/sge.c
+++ b/drivers/net/cxgbe/sge.c
@@ -1689,6 +1689,7 @@ int t4_sge_alloc_rxq(struct adapter *adap, struct sge_rspq *iq, bool fwevtq,
 	char z_name[RTE_MEMZONE_NAMESIZE];
 	char z_name_sw[RTE_MEMZONE_NAMESIZE];
 	unsigned int nb_refill;
+	u8 pciechan;
 
 	/* Size needs to be multiple of 16, including status entry. */
 	iq->size = cxgbe_roundup(iq->size, 16);
@@ -1706,8 +1707,19 @@ int t4_sge_alloc_rxq(struct adapter *adap, struct sge_rspq *iq, bool fwevtq,
 
 	memset(&c, 0, sizeof(c));
 	c.op_to_vfn = htonl(V_FW_CMD_OP(FW_IQ_CMD) | F_FW_CMD_REQUEST |
-			    F_FW_CMD_WRITE | F_FW_CMD_EXEC |
-			    V_FW_IQ_CMD_PFN(adap->pf) | V_FW_IQ_CMD_VFN(0));
+			    F_FW_CMD_WRITE | F_FW_CMD_EXEC);
+
+	if (is_pf4(adap)) {
+		pciechan = cong > 0 ? cxgbe_ffs(cong) - 1 : pi->tx_chan;
+		c.op_to_vfn |= htonl(V_FW_IQ_CMD_PFN(adap->pf) |
+				     V_FW_IQ_CMD_VFN(0));
+		if (cong >= 0)
+			c.iqns_to_fl0congen = htonl(F_FW_IQ_CMD_IQFLINTCONGEN |
+						    F_FW_IQ_CMD_IQRO);
+	} else {
+		pciechan = pi->port_id;
+	}
+
 	c.alloc_to_len16 = htonl(F_FW_IQ_CMD_ALLOC | F_FW_IQ_CMD_IQSTART |
 				 (sizeof(c) / 16));
 	c.type_to_iqandstindex =
@@ -1719,16 +1731,12 @@ int t4_sge_alloc_rxq(struct adapter *adap, struct sge_rspq *iq, bool fwevtq,
 		      V_FW_IQ_CMD_IQANDSTINDEX(intr_idx >= 0 ? intr_idx :
 							       -intr_idx - 1));
 	c.iqdroprss_to_iqesize =
-		htons(V_FW_IQ_CMD_IQPCIECH(cong > 0 ? cxgbe_ffs(cong) - 1 :
-						      pi->tx_chan) |
+		htons(V_FW_IQ_CMD_IQPCIECH(pciechan) |
 		      F_FW_IQ_CMD_IQGTSMODE |
 		      V_FW_IQ_CMD_IQINTCNTTHRESH(iq->pktcnt_idx) |
 		      V_FW_IQ_CMD_IQESIZE(ilog2(iq->iqe_len) - 4));
 	c.iqsize = htons(iq->size);
 	c.iqaddr = cpu_to_be64(iq->phys_addr);
-	if (cong >= 0)
-		c.iqns_to_fl0congen = htonl(F_FW_IQ_CMD_IQFLINTCONGEN |
-					    F_FW_IQ_CMD_IQRO);
 
 	if (fl) {
 		struct sge_eth_rxq *rxq = container_of(fl, struct sge_eth_rxq,
@@ -1768,7 +1776,7 @@ int t4_sge_alloc_rxq(struct adapter *adap, struct sge_rspq *iq, bool fwevtq,
 			       0 : F_FW_IQ_CMD_FL0PACKEN) |
 			      F_FW_IQ_CMD_FL0FETCHRO | F_FW_IQ_CMD_FL0DATARO |
 			      F_FW_IQ_CMD_FL0PADEN);
-		if (cong >= 0)
+		if (is_pf4(adap) && cong >= 0)
 			c.iqns_to_fl0congen |=
 				htonl(V_FW_IQ_CMD_FL0CNGCHMAP(cong) |
 				      F_FW_IQ_CMD_FL0CONGCIF |
@@ -1789,7 +1797,10 @@ int t4_sge_alloc_rxq(struct adapter *adap, struct sge_rspq *iq, bool fwevtq,
 		c.fl0addr = cpu_to_be64(fl->addr);
 	}
 
-	ret = t4_wr_mbox(adap, adap->mbox, &c, sizeof(c), &c);
+	if (is_pf4(adap))
+		ret = t4_wr_mbox(adap, adap->mbox, &c, sizeof(c), &c);
+	else
+		ret = t4vf_wr_mbox(adap, &c, sizeof(c), &c);
 	if (ret)
 		goto err;
 
@@ -1806,7 +1817,7 @@ int t4_sge_alloc_rxq(struct adapter *adap, struct sge_rspq *iq, bool fwevtq,
 	iq->stat = (void *)&iq->desc[iq->size * 8];
 	iq->eth_dev = eth_dev;
 	iq->handler = hnd;
-	iq->port_id = pi->port_id;
+	iq->port_id = pi->pidx;
 	iq->mb_pool = mp;
 
 	/* set offset to -1 to distinguish ingress queues without FL */
@@ -1846,7 +1857,7 @@ int t4_sge_alloc_rxq(struct adapter *adap, struct sge_rspq *iq, bool fwevtq,
 	 * a lot easier to fix in one place ...  For now we do something very
 	 * simple (and hopefully less wrong).
 	 */
-	if (!is_t4(adap->params.chip) && cong >= 0) {
+	if (is_pf4(adap) && !is_t4(adap->params.chip) && cong >= 0) {
 		u32 param, val;
 		int i;
 
@@ -1893,9 +1904,11 @@ int t4_sge_alloc_rxq(struct adapter *adap, struct sge_rspq *iq, bool fwevtq,
 	return ret;
 }
 
-static void init_txq(struct adapter *adap, struct sge_txq *q, unsigned int id)
+static void init_txq(struct adapter *adap, struct sge_txq *q, unsigned int id,
+		     unsigned int abs_id)
 {
 	q->cntxt_id = id;
+	q->abs_id = abs_id;
 	q->bar2_addr = bar2_address(adap, q->cntxt_id, T4_BAR2_QTYPE_EGRESS,
 				    &q->bar2_qid);
 	q->cidx = 0;
@@ -1943,6 +1956,7 @@ int t4_sge_alloc_eth_txq(struct adapter *adap, struct sge_eth_txq *txq,
 	struct port_info *pi = (struct port_info *)(eth_dev->data->dev_private);
 	char z_name[RTE_MEMZONE_NAMESIZE];
 	char z_name_sw[RTE_MEMZONE_NAMESIZE];
+	u8 pciechan;
 
 	/* Add status entries */
 	nentries = txq->q.size + s->stat_len / sizeof(struct tx_desc);
@@ -1961,16 +1975,22 @@ int t4_sge_alloc_eth_txq(struct adapter *adap, struct sge_eth_txq *txq,
 
 	memset(&c, 0, sizeof(c));
 	c.op_to_vfn = htonl(V_FW_CMD_OP(FW_EQ_ETH_CMD) | F_FW_CMD_REQUEST |
-			    F_FW_CMD_WRITE | F_FW_CMD_EXEC |
-			    V_FW_EQ_ETH_CMD_PFN(adap->pf) |
-			    V_FW_EQ_ETH_CMD_VFN(0));
+			    F_FW_CMD_WRITE | F_FW_CMD_EXEC);
+	if (is_pf4(adap)) {
+		pciechan = pi->tx_chan;
+		c.op_to_vfn |= htonl(V_FW_EQ_ETH_CMD_PFN(adap->pf) |
+				     V_FW_EQ_ETH_CMD_VFN(0));
+	} else {
+		pciechan = pi->port_id;
+	}
+
 	c.alloc_to_len16 = htonl(F_FW_EQ_ETH_CMD_ALLOC |
 				 F_FW_EQ_ETH_CMD_EQSTART | (sizeof(c) / 16));
 	c.autoequiqe_to_viid = htonl(F_FW_EQ_ETH_CMD_AUTOEQUEQE |
 				     V_FW_EQ_ETH_CMD_VIID(pi->viid));
 	c.fetchszm_to_iqid =
 		htonl(V_FW_EQ_ETH_CMD_HOSTFCMODE(X_HOSTFCMODE_NONE) |
-		      V_FW_EQ_ETH_CMD_PCIECHN(pi->tx_chan) |
+		      V_FW_EQ_ETH_CMD_PCIECHN(pciechan) |
 		      F_FW_EQ_ETH_CMD_FETCHRO | V_FW_EQ_ETH_CMD_IQID(iqid));
 	c.dcaen_to_eqsize =
 		htonl(V_FW_EQ_ETH_CMD_FBMIN(X_FETCHBURSTMIN_64B) |
@@ -1978,7 +1998,10 @@ int t4_sge_alloc_eth_txq(struct adapter *adap, struct sge_eth_txq *txq,
 		      V_FW_EQ_ETH_CMD_EQSIZE(nentries));
 	c.eqaddr = rte_cpu_to_be_64(txq->q.phys_addr);
 
-	ret = t4_wr_mbox(adap, adap->mbox, &c, sizeof(c), &c);
+	if (is_pf4(adap))
+		ret = t4_wr_mbox(adap, adap->mbox, &c, sizeof(c), &c);
+	else
+		ret = t4vf_wr_mbox(adap, &c, sizeof(c), &c);
 	if (ret) {
 		rte_free(txq->q.sdesc);
 		txq->q.sdesc = NULL;
@@ -1986,7 +2009,8 @@ int t4_sge_alloc_eth_txq(struct adapter *adap, struct sge_eth_txq *txq,
 		return ret;
 	}
 
-	init_txq(adap, &txq->q, G_FW_EQ_ETH_CMD_EQID(ntohl(c.eqid_pkd)));
+	init_txq(adap, &txq->q, G_FW_EQ_ETH_CMD_EQID(ntohl(c.eqid_pkd)),
+		 G_FW_EQ_ETH_CMD_PHYSEQID(ntohl(c.physeqid_pkd)));
 	txq->stats.tso = 0;
 	txq->stats.pkts = 0;
 	txq->stats.tx_cso = 0;
@@ -2281,3 +2305,182 @@ int t4_sge_init(struct adapter *adap)
 
 	return 0;
 }
+
+int t4vf_sge_init(struct adapter *adap)
+{
+	struct sge_params *sge_params = &adap->params.sge;
+	u32 sge_ingress_queues_per_page;
+	u32 sge_egress_queues_per_page;
+	u32 sge_control, sge_control2;
+	u32 fl_small_pg, fl_large_pg;
+	u32 sge_ingress_rx_threshold;
+	u32 sge_timer_value_0_and_1;
+	u32 sge_timer_value_2_and_3;
+	u32 sge_timer_value_4_and_5;
+	u32 sge_congestion_control;
+	struct sge *s = &adap->sge;
+	unsigned int s_hps, s_qpp;
+	u32 sge_host_page_size;
+	u32 params[7], vals[7];
+	int v;
+
+	/* query basic params from fw */
+	params[0] = (V_FW_PARAMS_MNEM(FW_PARAMS_MNEM_REG) |
+		     V_FW_PARAMS_PARAM_XYZ(A_SGE_CONTROL));
+	params[1] = (V_FW_PARAMS_MNEM(FW_PARAMS_MNEM_REG) |
+		     V_FW_PARAMS_PARAM_XYZ(A_SGE_HOST_PAGE_SIZE));
+	params[2] = (V_FW_PARAMS_MNEM(FW_PARAMS_MNEM_REG) |
+		     V_FW_PARAMS_PARAM_XYZ(A_SGE_FL_BUFFER_SIZE0));
+	params[3] = (V_FW_PARAMS_MNEM(FW_PARAMS_MNEM_REG) |
+		     V_FW_PARAMS_PARAM_XYZ(A_SGE_FL_BUFFER_SIZE1));
+	params[4] = (V_FW_PARAMS_MNEM(FW_PARAMS_MNEM_REG) |
+		     V_FW_PARAMS_PARAM_XYZ(A_SGE_TIMER_VALUE_0_AND_1));
+	params[5] = (V_FW_PARAMS_MNEM(FW_PARAMS_MNEM_REG) |
+		     V_FW_PARAMS_PARAM_XYZ(A_SGE_TIMER_VALUE_2_AND_3));
+	params[6] = (V_FW_PARAMS_MNEM(FW_PARAMS_MNEM_REG) |
+		     V_FW_PARAMS_PARAM_XYZ(A_SGE_TIMER_VALUE_4_AND_5));
+	v = t4vf_query_params(adap, 7, params, vals);
+	if (v != FW_SUCCESS)
+		return v;
+
+	sge_control = vals[0];
+	sge_host_page_size = vals[1];
+	fl_small_pg = vals[2];
+	fl_large_pg = vals[3];
+	sge_timer_value_0_and_1 = vals[4];
+	sge_timer_value_2_and_3 = vals[5];
+	sge_timer_value_4_and_5 = vals[6];
+
+	/*
+	 * Start by vetting the basic SGE parameters which have been set up by
+	 * the Physical Function Driver.
+	 */
+
+	/* We only bother using the Large Page logic if the Large Page Buffer
+	 * is larger than our Page Size Buffer.
+	 */
+	if (fl_large_pg <= fl_small_pg)
+		fl_large_pg = 0;
+
+	/* The Page Size Buffer must be exactly equal to our Page Size and the
+	 * Large Page Size Buffer should be 0 (per above) or a power of 2.
+	 */
+	if (fl_small_pg != CXGBE_PAGE_SIZE ||
+	    (fl_large_pg & (fl_large_pg - 1)) != 0) {
+		dev_err(adapter->pdev_dev, "bad SGE FL buffer sizes [%d, %d]\n",
+			fl_small_pg, fl_large_pg);
+		return -EINVAL;
+	}
+
+	if ((sge_control & F_RXPKTCPLMODE) !=
+	    V_RXPKTCPLMODE(X_RXPKTCPLMODE_SPLIT)) {
+		dev_err(adapter->pdev_dev, "bad SGE CPL MODE\n");
+		return -EINVAL;
+	}
+
+
+	/* Grab ingress packing boundary from SGE_CONTROL2 for */
+	params[0] = (V_FW_PARAMS_MNEM(FW_PARAMS_MNEM_REG) |
+		     V_FW_PARAMS_PARAM_XYZ(A_SGE_CONTROL2));
+	v = t4vf_query_params(adap, 1, params, vals);
+	if (v != FW_SUCCESS) {
+		dev_err(adapter, "Unable to get SGE Control2; "
+			"probably old firmware.\n");
+		return v;
+	}
+	sge_control2 = vals[0];
+
+	params[0] = (V_FW_PARAMS_MNEM(FW_PARAMS_MNEM_REG) |
+		     V_FW_PARAMS_PARAM_XYZ(A_SGE_INGRESS_RX_THRESHOLD));
+	params[1] = (V_FW_PARAMS_MNEM(FW_PARAMS_MNEM_REG) |
+		     V_FW_PARAMS_PARAM_XYZ(A_SGE_CONM_CTRL));
+	v = t4vf_query_params(adap, 2, params, vals);
+	if (v != FW_SUCCESS)
+		return v;
+	sge_ingress_rx_threshold = vals[0];
+	sge_congestion_control = vals[1];
+	params[0] = (V_FW_PARAMS_MNEM(FW_PARAMS_MNEM_REG) |
+		     V_FW_PARAMS_PARAM_XYZ(A_SGE_EGRESS_QUEUES_PER_PAGE_VF));
+	params[1] = (V_FW_PARAMS_MNEM(FW_PARAMS_MNEM_REG) |
+		     V_FW_PARAMS_PARAM_XYZ(A_SGE_INGRESS_QUEUES_PER_PAGE_VF));
+	v = t4vf_query_params(adap, 2, params, vals);
+	if (v != FW_SUCCESS) {
+		dev_warn(adap, "Unable to get VF SGE Queues/Page; "
+			 "probably old firmware.\n");
+		return v;
+	}
+	sge_egress_queues_per_page = vals[0];
+	sge_ingress_queues_per_page = vals[1];
+
+	/*
+	 * We need the Queues/Page for our VF.  This is based on the
+	 * PF from which we're instantiated and is indexed in the
+	 * register we just read.
+	 */
+	s_hps = (S_HOSTPAGESIZEPF0 +
+		 (S_HOSTPAGESIZEPF1 - S_HOSTPAGESIZEPF0) * adap->pf);
+	sge_params->hps =
+		((sge_host_page_size >> s_hps) & M_HOSTPAGESIZEPF0);
+
+	s_qpp = (S_QUEUESPERPAGEPF0 +
+		 (S_QUEUESPERPAGEPF1 - S_QUEUESPERPAGEPF0) * adap->pf);
+	sge_params->eq_qpp =
+		((sge_egress_queues_per_page >> s_qpp)
+		 & M_QUEUESPERPAGEPF0);
+	sge_params->iq_qpp =
+		((sge_ingress_queues_per_page >> s_qpp)
+		 & M_QUEUESPERPAGEPF0);
+
+	/*
+	 * Now translate the queried parameters into our internal forms.
+	 */
+	if (fl_large_pg)
+		s->fl_pg_order = ilog2(fl_large_pg) - PAGE_SHIFT;
+	s->stat_len = ((sge_control & F_EGRSTATUSPAGESIZE)
+			? 128 : 64);
+	s->pktshift = G_PKTSHIFT(sge_control);
+	s->fl_align = t4vf_fl_pkt_align(adap, sge_control, sge_control2);
+
+	/*
+	 * A FL with <= fl_starve_thres buffers is starving and a periodic
+	 * timer will attempt to refill it.  This needs to be larger than the
+	 * SGE's Egress Congestion Threshold.  If it isn't, then we can get
+	 * stuck waiting for new packets while the SGE is waiting for us to
+	 * give it more Free List entries.  (Note that the SGE's Egress
+	 * Congestion Threshold is in units of 2 Free List pointers.)
+	 */
+	switch (CHELSIO_CHIP_VERSION(adap->params.chip)) {
+	case CHELSIO_T5:
+		s->fl_starve_thres =
+			G_EGRTHRESHOLDPACKING(sge_congestion_control);
+		break;
+	case CHELSIO_T6:
+	default:
+		s->fl_starve_thres =
+			G_T6_EGRTHRESHOLDPACKING(sge_congestion_control);
+		break;
+	}
+	s->fl_starve_thres = s->fl_starve_thres * 2 + 1;
+
+	/*
+	 * Save RX interrupt holdoff timer values and counter
+	 * threshold values from the SGE parameters.
+	 */
+	s->timer_val[0] = core_ticks_to_us(adap,
+			G_TIMERVALUE0(sge_timer_value_0_and_1));
+	s->timer_val[1] = core_ticks_to_us(adap,
+			G_TIMERVALUE1(sge_timer_value_0_and_1));
+	s->timer_val[2] = core_ticks_to_us(adap,
+			G_TIMERVALUE2(sge_timer_value_2_and_3));
+	s->timer_val[3] = core_ticks_to_us(adap,
+			G_TIMERVALUE3(sge_timer_value_2_and_3));
+	s->timer_val[4] = core_ticks_to_us(adap,
+			G_TIMERVALUE4(sge_timer_value_4_and_5));
+	s->timer_val[5] = core_ticks_to_us(adap,
+			G_TIMERVALUE5(sge_timer_value_4_and_5));
+	s->counter_val[0] = G_THRESHOLD_0(sge_ingress_rx_threshold);
+	s->counter_val[1] = G_THRESHOLD_1(sge_ingress_rx_threshold);
+	s->counter_val[2] = G_THRESHOLD_2(sge_ingress_rx_threshold);
+	s->counter_val[3] = G_THRESHOLD_3(sge_ingress_rx_threshold);
+	return 0;
+}
-- 
2.14.1

^ permalink raw reply related	[flat|nested] 26+ messages in thread

* [PATCH 06/13] cxgbe: enable RSS for VF
  2018-03-10 22:48 [PATCH 00/13] cxgbe: add CXGBE VF PMD and updates Rahul Lakkireddy
                   ` (4 preceding siblings ...)
  2018-03-10 22:48 ` [PATCH 05/13] cxgbe: initialize SGE and queues for VF Rahul Lakkireddy
@ 2018-03-10 22:48 ` Rahul Lakkireddy
  2018-03-10 22:48 ` [PATCH 07/13] cxgbe: update TX and RX path " Rahul Lakkireddy
                   ` (8 subsequent siblings)
  14 siblings, 0 replies; 26+ messages in thread
From: Rahul Lakkireddy @ 2018-03-10 22:48 UTC (permalink / raw)
  To: dev; +Cc: kumaras, nirranjan, indranil

From: Kumar Sanghvi <kumaras@chelsio.com>

Query firmware and configure RSS for VF.

Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
---
 drivers/net/cxgbe/base/common.h         | 22 ++++++++++
 drivers/net/cxgbe/base/t4fw_interface.h | 78 +++++++++++++++++++++++++++++++++
 drivers/net/cxgbe/base/t4vf_hw.c        | 71 ++++++++++++++++++++++++++++++
 drivers/net/cxgbe/cxgbe.h               |  1 +
 drivers/net/cxgbe/cxgbe_main.c          | 13 +++---
 drivers/net/cxgbe/cxgbevf_main.c        | 16 +++++++
 6 files changed, 196 insertions(+), 5 deletions(-)

diff --git a/drivers/net/cxgbe/base/common.h b/drivers/net/cxgbe/base/common.h
index d74903dfd..567e55ec2 100644
--- a/drivers/net/cxgbe/base/common.h
+++ b/drivers/net/cxgbe/base/common.h
@@ -210,6 +210,26 @@ struct arch_specific_params {
 	u16 mps_tcam_size;
 };
 
+/*
+ * Global Receive Side Scaling (RSS) parameters in host-native format.
+ */
+struct rss_params {
+	unsigned int mode;			/* RSS mode */
+	union {
+		struct {
+			uint synmapen:1;	/* SYN Map Enable */
+			uint syn4tupenipv6:1;	/* en 4-tuple IPv6 SYNs hash */
+			uint syn2tupenipv6:1;	/* en 2-tuple IPv6 SYNs hash */
+			uint syn4tupenipv4:1;	/* en 4-tuple IPv4 SYNs hash */
+			uint syn2tupenipv4:1;	/* en 2-tuple IPv4 SYNs hash */
+			uint ofdmapen:1;	/* Offload Map Enable */
+			uint tnlmapen:1;	/* Tunnel Map Enable */
+			uint tnlalllookup:1;	/* Tunnel All Lookup */
+			uint hashtoeplitz:1;	/* use Toeplitz hash */
+		} basicvirtual;
+	} u;
+};
+
 /*
  * Maximum resources provisioned for a PCI VF.
  */
@@ -232,6 +252,7 @@ struct adapter_params {
 	struct vpd_params vpd;
 	struct pci_params pci;
 	struct devlog_params devlog;
+	struct rss_params rss;
 	struct vf_resources vfres;
 	enum pcie_memwin drv_memwin;
 
@@ -348,6 +369,7 @@ int t4vf_query_params(struct adapter *adap, unsigned int nparams,
 		      const u32 *params, u32 *vals);
 int t4vf_get_dev_params(struct adapter *adap);
 int t4vf_get_vpd_params(struct adapter *adap);
+int t4vf_get_rss_glb_config(struct adapter *adap);
 int t4vf_set_params(struct adapter *adapter, unsigned int nparams,
 		    const u32 *params, const u32 *vals);
 int t4_set_params_timeout(struct adapter *adap, unsigned int mbox,
diff --git a/drivers/net/cxgbe/base/t4fw_interface.h b/drivers/net/cxgbe/base/t4fw_interface.h
index 6c5c97270..3a89814ff 100644
--- a/drivers/net/cxgbe/base/t4fw_interface.h
+++ b/drivers/net/cxgbe/base/t4fw_interface.h
@@ -187,6 +187,7 @@ enum fw_cmd_opcodes {
 	FW_VI_ENABLE_CMD               = 0x17,
 	FW_PORT_CMD                    = 0x1b,
 	FW_RSS_IND_TBL_CMD             = 0x20,
+	FW_RSS_GLB_CONFIG_CMD	       = 0x22,
 	FW_RSS_VI_CONFIG_CMD           = 0x23,
 	FW_DEBUG_CMD                   = 0x81,
 };
@@ -1779,6 +1780,83 @@ struct fw_rss_ind_tbl_cmd {
 #define G_FW_RSS_IND_TBL_CMD_IQ2(x)	\
 	(((x) >> S_FW_RSS_IND_TBL_CMD_IQ2) & M_FW_RSS_IND_TBL_CMD_IQ2)
 
+struct fw_rss_glb_config_cmd {
+	__be32 op_to_write;
+	__be32 retval_len16;
+	union fw_rss_glb_config {
+		struct fw_rss_glb_config_manual {
+			__be32 mode_pkd;
+			__be32 r3;
+			__be64 r4;
+			__be64 r5;
+		} manual;
+		struct fw_rss_glb_config_basicvirtual {
+			__be32 mode_keymode;
+			__be32 synmapen_to_hashtoeplitz;
+			__be64 r8;
+			__be64 r9;
+		} basicvirtual;
+	} u;
+};
+
+#define S_FW_RSS_GLB_CONFIG_CMD_MODE    28
+#define M_FW_RSS_GLB_CONFIG_CMD_MODE    0xf
+#define G_FW_RSS_GLB_CONFIG_CMD_MODE(x) \
+	(((x) >> S_FW_RSS_GLB_CONFIG_CMD_MODE) & M_FW_RSS_GLB_CONFIG_CMD_MODE)
+
+#define FW_RSS_GLB_CONFIG_CMD_MODE_BASICVIRTUAL 1
+
+#define S_FW_RSS_GLB_CONFIG_CMD_SYNMAPEN 8
+#define V_FW_RSS_GLB_CONFIG_CMD_SYNMAPEN(x) \
+	((x) << S_FW_RSS_GLB_CONFIG_CMD_SYNMAPEN)
+#define F_FW_RSS_GLB_CONFIG_CMD_SYNMAPEN V_FW_RSS_GLB_CONFIG_CMD_SYNMAPEN(1U)
+
+#define S_FW_RSS_GLB_CONFIG_CMD_SYN4TUPENIPV6 7
+#define V_FW_RSS_GLB_CONFIG_CMD_SYN4TUPENIPV6(x) \
+	((x) << S_FW_RSS_GLB_CONFIG_CMD_SYN4TUPENIPV6)
+#define F_FW_RSS_GLB_CONFIG_CMD_SYN4TUPENIPV6 \
+	V_FW_RSS_GLB_CONFIG_CMD_SYN4TUPENIPV6(1U)
+
+#define S_FW_RSS_GLB_CONFIG_CMD_SYN2TUPENIPV6 6
+#define V_FW_RSS_GLB_CONFIG_CMD_SYN2TUPENIPV6(x) \
+	((x) << S_FW_RSS_GLB_CONFIG_CMD_SYN2TUPENIPV6)
+#define F_FW_RSS_GLB_CONFIG_CMD_SYN2TUPENIPV6 \
+	V_FW_RSS_GLB_CONFIG_CMD_SYN2TUPENIPV6(1U)
+
+#define S_FW_RSS_GLB_CONFIG_CMD_SYN4TUPENIPV4 5
+#define V_FW_RSS_GLB_CONFIG_CMD_SYN4TUPENIPV4(x) \
+	((x) << S_FW_RSS_GLB_CONFIG_CMD_SYN4TUPENIPV4)
+#define F_FW_RSS_GLB_CONFIG_CMD_SYN4TUPENIPV4 \
+	V_FW_RSS_GLB_CONFIG_CMD_SYN4TUPENIPV4(1U)
+
+#define S_FW_RSS_GLB_CONFIG_CMD_SYN2TUPENIPV4 4
+#define V_FW_RSS_GLB_CONFIG_CMD_SYN2TUPENIPV4(x) \
+	((x) << S_FW_RSS_GLB_CONFIG_CMD_SYN2TUPENIPV4)
+#define F_FW_RSS_GLB_CONFIG_CMD_SYN2TUPENIPV4 \
+	V_FW_RSS_GLB_CONFIG_CMD_SYN2TUPENIPV4(1U)
+
+#define S_FW_RSS_GLB_CONFIG_CMD_OFDMAPEN 3
+#define V_FW_RSS_GLB_CONFIG_CMD_OFDMAPEN(x) \
+	((x) << S_FW_RSS_GLB_CONFIG_CMD_OFDMAPEN)
+#define F_FW_RSS_GLB_CONFIG_CMD_OFDMAPEN V_FW_RSS_GLB_CONFIG_CMD_OFDMAPEN(1U)
+
+#define S_FW_RSS_GLB_CONFIG_CMD_TNLMAPEN 2
+#define V_FW_RSS_GLB_CONFIG_CMD_TNLMAPEN(x) \
+	((x) << S_FW_RSS_GLB_CONFIG_CMD_TNLMAPEN)
+#define F_FW_RSS_GLB_CONFIG_CMD_TNLMAPEN V_FW_RSS_GLB_CONFIG_CMD_TNLMAPEN(1U)
+
+#define S_FW_RSS_GLB_CONFIG_CMD_TNLALLLKP 1
+#define V_FW_RSS_GLB_CONFIG_CMD_TNLALLLKP(x) \
+	((x) << S_FW_RSS_GLB_CONFIG_CMD_TNLALLLKP)
+#define F_FW_RSS_GLB_CONFIG_CMD_TNLALLLKP \
+	V_FW_RSS_GLB_CONFIG_CMD_TNLALLLKP(1U)
+
+#define S_FW_RSS_GLB_CONFIG_CMD_HASHTOEPLITZ 0
+#define V_FW_RSS_GLB_CONFIG_CMD_HASHTOEPLITZ(x) \
+	((x) << S_FW_RSS_GLB_CONFIG_CMD_HASHTOEPLITZ)
+#define F_FW_RSS_GLB_CONFIG_CMD_HASHTOEPLITZ \
+	V_FW_RSS_GLB_CONFIG_CMD_HASHTOEPLITZ(1U)
+
 struct fw_rss_vi_config_cmd {
 	__be32 op_to_viid;
 	__be32 retval_len16;
diff --git a/drivers/net/cxgbe/base/t4vf_hw.c b/drivers/net/cxgbe/base/t4vf_hw.c
index 8e48588b0..f0408757d 100644
--- a/drivers/net/cxgbe/base/t4vf_hw.c
+++ b/drivers/net/cxgbe/base/t4vf_hw.c
@@ -509,6 +509,77 @@ unsigned int t4vf_get_pf_from_vf(struct adapter *adapter)
 			G_SOURCEPF(whoami) : G_T6_SOURCEPF(whoami));
 }
 
+/**
+ * t4vf_get_rss_glb_config - retrieve adapter RSS Global Configuration
+ * @adapter: the adapter
+ *
+ * Retrieves global RSS mode and parameters with which we have to live
+ * and stores them in the @adapter's RSS parameters.
+ */
+int t4vf_get_rss_glb_config(struct adapter *adapter)
+{
+	struct rss_params *rss = &adapter->params.rss;
+	struct fw_rss_glb_config_cmd cmd, rpl;
+	int v;
+
+	/*
+	 * Execute an RSS Global Configuration read command to retrieve
+	 * our RSS configuration.
+	 */
+	memset(&cmd, 0, sizeof(cmd));
+	cmd.op_to_write = cpu_to_be32(V_FW_CMD_OP(FW_RSS_GLB_CONFIG_CMD) |
+				      F_FW_CMD_REQUEST |
+				      F_FW_CMD_READ);
+	cmd.retval_len16 = cpu_to_be32(FW_LEN16(cmd));
+	v = t4vf_wr_mbox(adapter, &cmd, sizeof(cmd), &rpl);
+	if (v != FW_SUCCESS)
+		return v;
+
+	/*
+	 * Translate the big-endian RSS Global Configuration into our
+	 * cpu-endian format based on the RSS mode.  We also do first level
+	 * filtering at this point to weed out modes which don't support
+	 * VF Drivers ...
+	 */
+	rss->mode = G_FW_RSS_GLB_CONFIG_CMD_MODE
+			(be32_to_cpu(rpl.u.manual.mode_pkd));
+	switch (rss->mode) {
+	case FW_RSS_GLB_CONFIG_CMD_MODE_BASICVIRTUAL: {
+		u32 word = be32_to_cpu
+				(rpl.u.basicvirtual.synmapen_to_hashtoeplitz);
+
+		rss->u.basicvirtual.synmapen =
+			((word & F_FW_RSS_GLB_CONFIG_CMD_SYNMAPEN) != 0);
+		rss->u.basicvirtual.syn4tupenipv6 =
+			((word & F_FW_RSS_GLB_CONFIG_CMD_SYN4TUPENIPV6) != 0);
+		rss->u.basicvirtual.syn2tupenipv6 =
+			((word & F_FW_RSS_GLB_CONFIG_CMD_SYN2TUPENIPV6) != 0);
+		rss->u.basicvirtual.syn4tupenipv4 =
+			((word & F_FW_RSS_GLB_CONFIG_CMD_SYN4TUPENIPV4) != 0);
+		rss->u.basicvirtual.syn2tupenipv4 =
+			((word & F_FW_RSS_GLB_CONFIG_CMD_SYN2TUPENIPV4) != 0);
+		rss->u.basicvirtual.ofdmapen =
+			((word & F_FW_RSS_GLB_CONFIG_CMD_OFDMAPEN) != 0);
+		rss->u.basicvirtual.tnlmapen =
+			((word & F_FW_RSS_GLB_CONFIG_CMD_TNLMAPEN) != 0);
+		rss->u.basicvirtual.tnlalllookup =
+			((word  & F_FW_RSS_GLB_CONFIG_CMD_TNLALLLKP) != 0);
+		rss->u.basicvirtual.hashtoeplitz =
+			((word & F_FW_RSS_GLB_CONFIG_CMD_HASHTOEPLITZ) != 0);
+
+		/* we need at least Tunnel Map Enable to be set */
+		if (!rss->u.basicvirtual.tnlmapen)
+			return -EINVAL;
+		break;
+	}
+
+	default:
+		/* all unknown/unsupported RSS modes result in an error */
+		return -EINVAL;
+	}
+	return 0;
+}
+
 /**
  * t4vf_get_vfres - retrieve VF resource limits
  * @adapter: the adapter
diff --git a/drivers/net/cxgbe/cxgbe.h b/drivers/net/cxgbe/cxgbe.h
index 040aca423..748037d83 100644
--- a/drivers/net/cxgbe/cxgbe.h
+++ b/drivers/net/cxgbe/cxgbe.h
@@ -62,6 +62,7 @@ void init_rspq(struct adapter *adap, struct sge_rspq *q, unsigned int us,
 int setup_sge_fwevtq(struct adapter *adapter);
 void cfg_queues(struct rte_eth_dev *eth_dev);
 int cfg_queue_count(struct rte_eth_dev *eth_dev);
+int init_rss(struct adapter *adap);
 int setup_rss(struct port_info *pi);
 void cxgbe_enable_rx_queues(struct port_info *pi);
 void print_port_info(struct adapter *adap);
diff --git a/drivers/net/cxgbe/cxgbe_main.c b/drivers/net/cxgbe/cxgbe_main.c
index d9e772717..882bcb7ef 100644
--- a/drivers/net/cxgbe/cxgbe_main.c
+++ b/drivers/net/cxgbe/cxgbe_main.c
@@ -346,14 +346,17 @@ static void setup_memwin(struct adapter *adap)
 					MEMWIN_NIC));
 }
 
-static int init_rss(struct adapter *adap)
+int init_rss(struct adapter *adap)
 {
 	unsigned int i;
-	int err;
 
-	err = t4_init_rss_mode(adap, adap->mbox);
-	if (err)
-		return err;
+	if (is_pf4(adap)) {
+		int err;
+
+		err = t4_init_rss_mode(adap, adap->mbox);
+		if (err)
+			return err;
+	}
 
 	for_each_port(adap, i) {
 		struct port_info *pi = adap2pinfo(adap, i);
diff --git a/drivers/net/cxgbe/cxgbevf_main.c b/drivers/net/cxgbe/cxgbevf_main.c
index f4d0f4de5..f3b83ecb9 100644
--- a/drivers/net/cxgbe/cxgbevf_main.c
+++ b/drivers/net/cxgbe/cxgbevf_main.c
@@ -114,6 +114,19 @@ static int adap_init0vf(struct adapter *adapter)
 		return err;
 	}
 
+	err = t4vf_get_rss_glb_config(adapter);
+	if (err) {
+		dev_err(adapter->pdev_dev, "unable to retrieve adapter"
+			" RSS parameters: err=%d\n", err);
+		return err;
+	}
+	if (adapter->params.rss.mode !=
+	    FW_RSS_GLB_CONFIG_CMD_MODE_BASICVIRTUAL) {
+		dev_err(adapter->pdev_dev, "unable to operate with global RSS"
+			" mode %d\n", adapter->params.rss.mode);
+		return -EINVAL;
+	}
+
 	/* If we're running on newer firmware, let it know that we're
 	 * prepared to deal with encapsulated CPL messages.  Older
 	 * firmware won't understand this and we'll just get
@@ -264,6 +277,9 @@ int cxgbevf_probe(struct adapter *adapter)
 	print_adapter_info(adapter);
 	print_port_info(adapter);
 
+	err = init_rss(adapter);
+	if (err)
+		goto out_free;
 	return 0;
 
 out_free:
-- 
2.14.1

^ permalink raw reply related	[flat|nested] 26+ messages in thread

* [PATCH 07/13] cxgbe: update TX and RX path for VF
  2018-03-10 22:48 [PATCH 00/13] cxgbe: add CXGBE VF PMD and updates Rahul Lakkireddy
                   ` (5 preceding siblings ...)
  2018-03-10 22:48 ` [PATCH 06/13] cxgbe: enable RSS " Rahul Lakkireddy
@ 2018-03-10 22:48 ` Rahul Lakkireddy
  2018-03-10 22:48 ` [PATCH 08/13] cxgbe: add VF port statistics Rahul Lakkireddy
                   ` (7 subsequent siblings)
  14 siblings, 0 replies; 26+ messages in thread
From: Rahul Lakkireddy @ 2018-03-10 22:48 UTC (permalink / raw)
  To: dev; +Cc: kumaras, nirranjan, indranil

From: Kumar Sanghvi <kumaras@chelsio.com>

On TX path, add fw_eth_tx_pkt_vm_wr to transmit packets over VF.
Use is_pf4() to correctly calculate the work request size and
offsets within the work request.  On RX path, use pktshift to adjust
data offset within the mbuf.

Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
---
 drivers/net/cxgbe/base/adapter.h        |   5 ++
 drivers/net/cxgbe/base/t4fw_interface.h |  25 ++++++
 drivers/net/cxgbe/cxgbe_ethdev.c        |   8 +-
 drivers/net/cxgbe/cxgbe_pfvf.h          |   4 +
 drivers/net/cxgbe/cxgbevf_ethdev.c      |   4 +-
 drivers/net/cxgbe/sge.c                 | 139 +++++++++++++++++++++++++-------
 6 files changed, 149 insertions(+), 36 deletions(-)

diff --git a/drivers/net/cxgbe/base/adapter.h b/drivers/net/cxgbe/base/adapter.h
index 95752d1b4..fcea4055a 100644
--- a/drivers/net/cxgbe/base/adapter.h
+++ b/drivers/net/cxgbe/base/adapter.h
@@ -198,6 +198,7 @@ struct sge_eth_rxq {                /* a SW Ethernet Rx queue */
  * scenario where a packet needs 32 bytes.
  */
 #define ETH_COALESCE_PKT_NUM 15
+#define ETH_COALESCE_VF_PKT_NUM 7
 #define ETH_COALESCE_PKT_PER_DESC 2
 
 struct tx_eth_coal_desc {
@@ -227,6 +228,10 @@ struct eth_coalesce {
 	unsigned int len;
 	unsigned int flits;
 	unsigned int max;
+	__u8 ethmacdst[ETHER_ADDR_LEN];
+	__u8 ethmacsrc[ETHER_ADDR_LEN];
+	__be16 ethtype;
+	__be16 vlantci;
 };
 
 struct sge_txq {
diff --git a/drivers/net/cxgbe/base/t4fw_interface.h b/drivers/net/cxgbe/base/t4fw_interface.h
index 3a89814ff..274f00b95 100644
--- a/drivers/net/cxgbe/base/t4fw_interface.h
+++ b/drivers/net/cxgbe/base/t4fw_interface.h
@@ -84,6 +84,8 @@ enum fw_memtype {
 enum fw_wr_opcodes {
 	FW_ETH_TX_PKT_WR	= 0x08,
 	FW_ETH_TX_PKTS_WR	= 0x09,
+	FW_ETH_TX_PKT_VM_WR	= 0x11,
+	FW_ETH_TX_PKTS_VM_WR	= 0x12,
 	FW_ETH_TX_PKTS2_WR      = 0x78,
 };
 
@@ -146,6 +148,29 @@ struct fw_eth_tx_pkts_wr {
 	__u8   type;
 };
 
+struct fw_eth_tx_pkt_vm_wr {
+	__be32 op_immdlen;
+	__be32 equiq_to_len16;
+	__be32 r3[2];
+	__u8   ethmacdst[6];
+	__u8   ethmacsrc[6];
+	__be16 ethtype;
+	__be16 vlantci;
+};
+
+struct fw_eth_tx_pkts_vm_wr {
+	__be32 op_pkd;
+	__be32 equiq_to_len16;
+	__be32 r3;
+	__be16 plen;
+	__u8   npkt;
+	__u8   r4;
+	__u8   ethmacdst[6];
+	__u8   ethmacsrc[6];
+	__be16 ethtype;
+	__be16 vlantci;
+};
+
 /******************************************************************************
  *  C O M M A N D s
  *********************/
diff --git a/drivers/net/cxgbe/cxgbe_ethdev.c b/drivers/net/cxgbe/cxgbe_ethdev.c
index 16031f38d..e84facd33 100644
--- a/drivers/net/cxgbe/cxgbe_ethdev.c
+++ b/drivers/net/cxgbe/cxgbe_ethdev.c
@@ -86,8 +86,8 @@
  */
 #include "t4_pci_id_tbl.h"
 
-static uint16_t cxgbe_xmit_pkts(void *tx_queue, struct rte_mbuf **tx_pkts,
-				uint16_t nb_pkts)
+uint16_t cxgbe_xmit_pkts(void *tx_queue, struct rte_mbuf **tx_pkts,
+			 uint16_t nb_pkts)
 {
 	struct sge_eth_txq *txq = (struct sge_eth_txq *)tx_queue;
 	uint16_t pkts_sent, pkts_remain;
@@ -120,8 +120,8 @@ static uint16_t cxgbe_xmit_pkts(void *tx_queue, struct rte_mbuf **tx_pkts,
 	return total_sent;
 }
 
-static uint16_t cxgbe_recv_pkts(void *rx_queue, struct rte_mbuf **rx_pkts,
-				uint16_t nb_pkts)
+uint16_t cxgbe_recv_pkts(void *rx_queue, struct rte_mbuf **rx_pkts,
+			 uint16_t nb_pkts)
 {
 	struct sge_eth_rxq *rxq = (struct sge_eth_rxq *)rx_queue;
 	unsigned int work_done;
diff --git a/drivers/net/cxgbe/cxgbe_pfvf.h b/drivers/net/cxgbe/cxgbe_pfvf.h
index e3d8533ca..19bfd6d92 100644
--- a/drivers/net/cxgbe/cxgbe_pfvf.h
+++ b/drivers/net/cxgbe/cxgbe_pfvf.h
@@ -34,5 +34,9 @@ int cxgbe_dev_mtu_set(struct rte_eth_dev *eth_dev, uint16_t mtu);
 int cxgbe_dev_start(struct rte_eth_dev *eth_dev);
 int cxgbe_dev_link_update(struct rte_eth_dev *eth_dev,
 			  int wait_to_complete);
+uint16_t cxgbe_xmit_pkts(void *tx_queue, struct rte_mbuf **tx_pkts,
+			 uint16_t nb_pkts);
+uint16_t cxgbe_recv_pkts(void *rx_queue, struct rte_mbuf **rx_pkts,
+			 uint16_t nb_pkts);
 const uint32_t *cxgbe_dev_supported_ptypes_get(struct rte_eth_dev *eth_dev);
 #endif /* _CXGBE_PFVF_H_ */
diff --git a/drivers/net/cxgbe/cxgbevf_ethdev.c b/drivers/net/cxgbe/cxgbevf_ethdev.c
index 3b1deac52..a96630341 100644
--- a/drivers/net/cxgbe/cxgbevf_ethdev.c
+++ b/drivers/net/cxgbe/cxgbevf_ethdev.c
@@ -68,8 +68,8 @@ static int eth_cxgbevf_dev_init(struct rte_eth_dev *eth_dev)
 	CXGBE_FUNC_TRACE();
 
 	eth_dev->dev_ops = &cxgbevf_eth_dev_ops;
-	eth_dev->rx_pkt_burst = NULL;
-	eth_dev->tx_pkt_burst = NULL;
+	eth_dev->rx_pkt_burst = &cxgbe_recv_pkts;
+	eth_dev->tx_pkt_burst = &cxgbe_xmit_pkts;
 	pci_dev = RTE_ETH_DEV_TO_PCI(eth_dev);
 
 	/* for secondary processes, we attach to ethdevs allocated by primary
diff --git a/drivers/net/cxgbe/sge.c b/drivers/net/cxgbe/sge.c
index aba1a49f3..54e13fb9a 100644
--- a/drivers/net/cxgbe/sge.c
+++ b/drivers/net/cxgbe/sge.c
@@ -337,7 +337,11 @@ static inline void ring_fl_db(struct adapter *adap, struct sge_fl *q)
 		 * mechanism.
 		 */
 		if (unlikely(!q->bar2_addr)) {
-			t4_write_reg_relaxed(adap, MYPF_REG(A_SGE_PF_KDOORBELL),
+			u32 reg = is_pf4(adap) ? MYPF_REG(A_SGE_PF_KDOORBELL) :
+						 T4VF_SGE_BASE_ADDR +
+						 A_SGE_VF_KDOORBELL;
+
+			t4_write_reg_relaxed(adap, reg,
 					     val | V_QID(q->cntxt_id));
 		} else {
 			writel_relaxed(val | V_QID(q->bar2_qid),
@@ -570,12 +574,16 @@ static inline int is_eth_imm(const struct rte_mbuf *m)
 /**
  * calc_tx_flits - calculate the number of flits for a packet Tx WR
  * @m: the packet
+ * @adap: adapter structure pointer
  *
  * Returns the number of flits needed for a Tx WR for the given Ethernet
  * packet, including the needed WR and CPL headers.
  */
-static inline unsigned int calc_tx_flits(const struct rte_mbuf *m)
+static inline unsigned int calc_tx_flits(const struct rte_mbuf *m,
+					 struct adapter *adap)
 {
+	size_t wr_size = is_pf4(adap) ? sizeof(struct fw_eth_tx_pkt_wr) :
+					sizeof(struct fw_eth_tx_pkt_vm_wr);
 	unsigned int flits;
 	int hdrlen;
 
@@ -600,11 +608,10 @@ static inline unsigned int calc_tx_flits(const struct rte_mbuf *m)
 	 */
 	flits = sgl_len(m->nb_segs);
 	if (m->tso_segsz)
-		flits += (sizeof(struct fw_eth_tx_pkt_wr) +
-			  sizeof(struct cpl_tx_pkt_lso_core) +
+		flits += (wr_size + sizeof(struct cpl_tx_pkt_lso_core) +
 			  sizeof(struct cpl_tx_pkt_core)) / sizeof(__be64);
 	else
-		flits += (sizeof(struct fw_eth_tx_pkt_wr) +
+		flits += (wr_size +
 			  sizeof(struct cpl_tx_pkt_core)) / sizeof(__be64);
 	return flits;
 }
@@ -848,14 +855,20 @@ static void tx_timer_cb(void *data)
 static inline void ship_tx_pkt_coalesce_wr(struct adapter *adap,
 					   struct sge_eth_txq *txq)
 {
-	u32 wr_mid;
-	struct sge_txq *q = &txq->q;
+	struct fw_eth_tx_pkts_vm_wr *vmwr;
+	const size_t fw_hdr_copy_len = (sizeof(vmwr->ethmacdst) +
+					sizeof(vmwr->ethmacsrc) +
+					sizeof(vmwr->ethtype) +
+					sizeof(vmwr->vlantci));
 	struct fw_eth_tx_pkts_wr *wr;
+	struct sge_txq *q = &txq->q;
 	unsigned int ndesc;
+	u32 wr_mid;
 
 	/* fill the pkts WR header */
 	wr = (void *)&q->desc[q->pidx];
 	wr->op_pkd = htonl(V_FW_WR_OP(FW_ETH_TX_PKTS2_WR));
+	vmwr = (void *)&q->desc[q->pidx];
 
 	wr_mid = V_FW_WR_LEN16(DIV_ROUND_UP(q->coalesce.flits, 2));
 	ndesc = flits_to_desc(q->coalesce.flits);
@@ -863,12 +876,18 @@ static inline void ship_tx_pkt_coalesce_wr(struct adapter *adap,
 	wr->plen = cpu_to_be16(q->coalesce.len);
 	wr->npkt = q->coalesce.idx;
 	wr->r3 = 0;
-	wr->type = q->coalesce.type;
+	if (is_pf4(adap)) {
+		wr->op_pkd = htonl(V_FW_WR_OP(FW_ETH_TX_PKTS2_WR));
+		wr->type = q->coalesce.type;
+	} else {
+		wr->op_pkd = htonl(V_FW_WR_OP(FW_ETH_TX_PKTS_VM_WR));
+		vmwr->r4 = 0;
+		memcpy((void *)vmwr->ethmacdst, (void *)q->coalesce.ethmacdst,
+		       fw_hdr_copy_len);
+	}
 
 	/* zero out coalesce structure members */
-	q->coalesce.idx = 0;
-	q->coalesce.flits = 0;
-	q->coalesce.len = 0;
+	memset((void *)&q->coalesce, 0, sizeof(struct eth_coalesce));
 
 	txq_advance(q, ndesc);
 	txq->stats.coal_wr++;
@@ -896,13 +915,27 @@ static inline int should_tx_packet_coalesce(struct sge_eth_txq *txq,
 					    unsigned int *nflits,
 					    struct adapter *adap)
 {
+	struct fw_eth_tx_pkts_vm_wr *wr;
+	const size_t fw_hdr_copy_len = (sizeof(wr->ethmacdst) +
+					sizeof(wr->ethmacsrc) +
+					sizeof(wr->ethtype) +
+					sizeof(wr->vlantci));
 	struct sge_txq *q = &txq->q;
 	unsigned int flits, ndesc;
 	unsigned char type = 0;
-	int credits;
+	int credits, wr_size;
 
 	/* use coal WR type 1 when no frags are present */
 	type = (mbuf->nb_segs == 1) ? 1 : 0;
+	if (!is_pf4(adap)) {
+		if (!type)
+			return 0;
+
+		if (q->coalesce.idx && memcmp((void *)q->coalesce.ethmacdst,
+					      rte_pktmbuf_mtod(mbuf, void *),
+					      fw_hdr_copy_len))
+			ship_tx_pkt_coalesce_wr(adap, txq);
+	}
 
 	if (unlikely(type != q->coalesce.type && q->coalesce.idx))
 		ship_tx_pkt_coalesce_wr(adap, txq);
@@ -948,16 +981,21 @@ static inline int should_tx_packet_coalesce(struct sge_eth_txq *txq,
 
 new:
 	/* start a new pkts WR, the WR header is not filled below */
-	flits += sizeof(struct fw_eth_tx_pkts_wr) / sizeof(__be64);
+	wr_size = is_pf4(adap) ? sizeof(struct fw_eth_tx_pkts_wr) :
+				 sizeof(struct fw_eth_tx_pkts_vm_wr);
+	flits += wr_size / sizeof(__be64);
 	ndesc = flits_to_desc(q->coalesce.flits + flits);
 	credits = txq_avail(q) - ndesc;
 
 	if (unlikely(credits < 0 || wraps_around(q, ndesc)))
 		return 0;
-	q->coalesce.flits += 2;
+	q->coalesce.flits += wr_size / sizeof(__be64);
 	q->coalesce.type = type;
 	q->coalesce.ptr = (unsigned char *)&q->desc[q->pidx] +
-			   2 * sizeof(__be64);
+			   q->coalesce.flits * sizeof(__be64);
+	if (!is_pf4(adap))
+		memcpy((void *)q->coalesce.ethmacdst,
+		       rte_pktmbuf_mtod(mbuf, void *), fw_hdr_copy_len);
 	return 1;
 }
 
@@ -987,6 +1025,8 @@ static inline int tx_do_packet_coalesce(struct sge_eth_txq *txq,
 	struct cpl_tx_pkt_core *cpl;
 	struct tx_sw_desc *sd;
 	unsigned int idx = q->coalesce.idx, len = mbuf->pkt_len;
+	unsigned int max_coal_pkt_num = is_pf4(adap) ? ETH_COALESCE_PKT_NUM :
+						       ETH_COALESCE_VF_PKT_NUM;
 
 #ifdef RTE_LIBRTE_CXGBE_TPUT
 	RTE_SET_USED(nb_pkts);
@@ -1030,9 +1070,12 @@ static inline int tx_do_packet_coalesce(struct sge_eth_txq *txq,
 		cntrl |= F_TXPKT_VLAN_VLD | V_TXPKT_VLAN(mbuf->vlan_tci);
 	}
 
-	cpl->ctrl0 = htonl(V_TXPKT_OPCODE(CPL_TX_PKT_XT) |
-			   V_TXPKT_INTF(pi->tx_chan) |
-			   V_TXPKT_PF(adap->pf));
+	cpl->ctrl0 = htonl(V_TXPKT_OPCODE(CPL_TX_PKT_XT));
+	if (is_pf4(adap))
+		cpl->ctrl0 |= htonl(V_TXPKT_INTF(pi->tx_chan) |
+				    V_TXPKT_PF(adap->pf));
+	else
+		cpl->ctrl0 |= htonl(V_TXPKT_INTF(pi->port_id));
 	cpl->pack = htons(0);
 	cpl->len = htons(len);
 	cpl->ctrl1 = cpu_to_be64(cntrl);
@@ -1061,7 +1104,7 @@ static inline int tx_do_packet_coalesce(struct sge_eth_txq *txq,
 	sd->coalesce.idx = (idx & 1) + 1;
 
 	/* send the coaelsced work request if max reached */
-	if (++q->coalesce.idx == ETH_COALESCE_PKT_NUM
+	if (++q->coalesce.idx == max_coal_pkt_num
 #ifndef RTE_LIBRTE_CXGBE_TPUT
 	    || q->coalesce.idx >= nb_pkts
 #endif
@@ -1085,6 +1128,7 @@ int t4_eth_xmit(struct sge_eth_txq *txq, struct rte_mbuf *mbuf,
 	struct adapter *adap;
 	struct rte_mbuf *m = mbuf;
 	struct fw_eth_tx_pkt_wr *wr;
+	struct fw_eth_tx_pkt_vm_wr *vmwr;
 	struct cpl_tx_pkt_core *cpl;
 	struct tx_sw_desc *d;
 	dma_addr_t addr[m->nb_segs];
@@ -1141,7 +1185,7 @@ int t4_eth_xmit(struct sge_eth_txq *txq, struct rte_mbuf *mbuf,
 	if (txq->q.coalesce.idx)
 		ship_tx_pkt_coalesce_wr(adap, txq);
 
-	flits = calc_tx_flits(m);
+	flits = calc_tx_flits(m, adap);
 	ndesc = flits_to_desc(flits);
 	credits = txq_avail(&txq->q) - ndesc;
 
@@ -1163,31 +1207,55 @@ int t4_eth_xmit(struct sge_eth_txq *txq, struct rte_mbuf *mbuf,
 	}
 
 	wr = (void *)&txq->q.desc[txq->q.pidx];
+	vmwr = (void *)&txq->q.desc[txq->q.pidx];
 	wr->equiq_to_len16 = htonl(wr_mid);
-	wr->r3 = rte_cpu_to_be_64(0);
-	end = (u64 *)wr + flits;
+	if (is_pf4(adap)) {
+		wr->r3 = rte_cpu_to_be_64(0);
+		end = (u64 *)wr + flits;
+	} else {
+		const size_t fw_hdr_copy_len = (sizeof(vmwr->ethmacdst) +
+						sizeof(vmwr->ethmacsrc) +
+						sizeof(vmwr->ethtype) +
+						sizeof(vmwr->vlantci));
+
+		vmwr->r3[0] = rte_cpu_to_be_32(0);
+		vmwr->r3[1] = rte_cpu_to_be_32(0);
+		memcpy((void *)vmwr->ethmacdst, rte_pktmbuf_mtod(m, void *),
+		       fw_hdr_copy_len);
+		end = (u64 *)vmwr + flits;
+	}
 
 	len = 0;
 	len += sizeof(*cpl);
 
 	/* Coalescing skipped and we send through normal path */
 	if (!(m->ol_flags & PKT_TX_TCP_SEG)) {
-		wr->op_immdlen = htonl(V_FW_WR_OP(FW_ETH_TX_PKT_WR) |
+		wr->op_immdlen = htonl(V_FW_WR_OP(is_pf4(adap) ?
+						  FW_ETH_TX_PKT_WR :
+						  FW_ETH_TX_PKT_VM_WR) |
 				       V_FW_WR_IMMDLEN(len));
-		cpl = (void *)(wr + 1);
+		if (is_pf4(adap))
+			cpl = (void *)(wr + 1);
+		else
+			cpl = (void *)(vmwr + 1);
 		if (m->ol_flags & PKT_TX_IP_CKSUM) {
 			cntrl = hwcsum(adap->params.chip, m) |
 				F_TXPKT_IPCSUM_DIS;
 			txq->stats.tx_cso++;
 		}
 	} else {
-		lso = (void *)(wr + 1);
+		if (is_pf4(adap))
+			lso = (void *)(wr + 1);
+		else
+			lso = (void *)(vmwr + 1);
 		v6 = (m->ol_flags & PKT_TX_IPV6) != 0;
 		l3hdr_len = m->l3_len;
 		l4hdr_len = m->l4_len;
 		eth_xtra_len = m->l2_len - ETHER_HDR_LEN;
 		len += sizeof(*lso);
-		wr->op_immdlen = htonl(V_FW_WR_OP(FW_ETH_TX_PKT_WR) |
+		wr->op_immdlen = htonl(V_FW_WR_OP(is_pf4(adap) ?
+						  FW_ETH_TX_PKT_WR :
+						  FW_ETH_TX_PKT_VM_WR) |
 				       V_FW_WR_IMMDLEN(len));
 		lso->lso_ctrl = htonl(V_LSO_OPCODE(CPL_TX_PKT_LSO) |
 				      F_LSO_FIRST_SLICE | F_LSO_LAST_SLICE |
@@ -1221,9 +1289,14 @@ int t4_eth_xmit(struct sge_eth_txq *txq, struct rte_mbuf *mbuf,
 		cntrl |= F_TXPKT_VLAN_VLD | V_TXPKT_VLAN(m->vlan_tci);
 	}
 
-	cpl->ctrl0 = htonl(V_TXPKT_OPCODE(CPL_TX_PKT_XT) |
-			   V_TXPKT_INTF(pi->tx_chan) |
-			   V_TXPKT_PF(adap->pf));
+	cpl->ctrl0 = htonl(V_TXPKT_OPCODE(CPL_TX_PKT_XT));
+	if (is_pf4(adap))
+		cpl->ctrl0 |= htonl(V_TXPKT_INTF(pi->tx_chan) |
+				    V_TXPKT_PF(adap->pf));
+	else
+		cpl->ctrl0 |= htonl(V_TXPKT_INTF(pi->port_id) |
+				    V_TXPKT_PF(0));
+
 	cpl->pack = htons(0);
 	cpl->len = htons(m->pkt_len);
 	cpl->ctrl1 = cpu_to_be64(cntrl);
@@ -1468,6 +1541,7 @@ static int process_responses(struct sge_rspq *q, int budget,
 		rsp_type = G_RSPD_TYPE(rc->u.type_gen);
 
 		if (likely(rsp_type == X_RSPD_TYPE_FLBUF)) {
+			struct sge *s = &q->adapter->sge;
 			unsigned int stat_pidx;
 			int stat_pidx_diff;
 
@@ -1554,6 +1628,7 @@ static int process_responses(struct sge_rspq *q, int budget,
 					pkt->vlan_tci = ntohs(cpl->vlan);
 				}
 
+				rte_pktmbuf_adj(pkt, s->pktshift);
 				rxq->stats.pkts++;
 				rxq->stats.rx_bytes += pkt->pkt_len;
 				rx_pkts[budget - budget_left] = pkt;
@@ -1612,7 +1687,11 @@ int cxgbe_poll(struct sge_rspq *q, struct rte_mbuf **rx_pkts,
 		val = V_CIDXINC(cidx_inc) | V_SEINTARM(params);
 
 		if (unlikely(!q->bar2_addr)) {
-			t4_write_reg(q->adapter, MYPF_REG(A_SGE_PF_GTS),
+			u32 reg = is_pf4(q->adapter) ? MYPF_REG(A_SGE_PF_GTS) :
+						       T4VF_SGE_BASE_ADDR +
+						       A_SGE_VF_GTS;
+
+			t4_write_reg(q->adapter, reg,
 				     val | V_INGRESSQID((u32)q->cntxt_id));
 		} else {
 			writel(val | V_INGRESSQID(q->bar2_qid),
-- 
2.14.1

^ permalink raw reply related	[flat|nested] 26+ messages in thread

* [PATCH 08/13] cxgbe: add VF port statistics
  2018-03-10 22:48 [PATCH 00/13] cxgbe: add CXGBE VF PMD and updates Rahul Lakkireddy
                   ` (6 preceding siblings ...)
  2018-03-10 22:48 ` [PATCH 07/13] cxgbe: update TX and RX path " Rahul Lakkireddy
@ 2018-03-10 22:48 ` Rahul Lakkireddy
  2018-03-10 22:48 ` [PATCH 09/13] cxgbe: add support to set mac address Rahul Lakkireddy
                   ` (6 subsequent siblings)
  14 siblings, 0 replies; 26+ messages in thread
From: Rahul Lakkireddy @ 2018-03-10 22:48 UTC (permalink / raw)
  To: dev; +Cc: kumaras, nirranjan, indranil

From: Kumar Sanghvi <kumaras@chelsio.com>

Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
---
 drivers/net/cxgbe/base/common.h         |   2 +
 drivers/net/cxgbe/base/t4_regs.h        |  10 +++
 drivers/net/cxgbe/base/t4fw_interface.h |  13 ++++
 drivers/net/cxgbe/base/t4vf_hw.c        | 105 ++++++++++++++++++++++++++++++++
 drivers/net/cxgbe/cxgbe.h               |   1 +
 drivers/net/cxgbe/cxgbevf_ethdev.c      |  44 +++++++++++++
 drivers/net/cxgbe/cxgbevf_main.c        |   5 ++
 7 files changed, 180 insertions(+)

diff --git a/drivers/net/cxgbe/base/common.h b/drivers/net/cxgbe/base/common.h
index 567e55ec2..fcb6474e8 100644
--- a/drivers/net/cxgbe/base/common.h
+++ b/drivers/net/cxgbe/base/common.h
@@ -484,6 +484,8 @@ unsigned int t4_get_mps_bg_map(struct adapter *adapter, unsigned int pidx);
 unsigned int t4_get_tp_ch_map(struct adapter *adapter, unsigned int pidx);
 const char *t4_get_port_type_description(enum fw_port_type port_type);
 void t4_get_port_stats(struct adapter *adap, int idx, struct port_stats *p);
+void t4vf_get_port_stats(struct adapter *adapter, int pidx,
+			 struct port_stats *p);
 void t4_get_port_stats_offset(struct adapter *adap, int idx,
 			      struct port_stats *stats,
 			      struct port_stats *offset);
diff --git a/drivers/net/cxgbe/base/t4_regs.h b/drivers/net/cxgbe/base/t4_regs.h
index 43d6a0c73..ae33aa33e 100644
--- a/drivers/net/cxgbe/base/t4_regs.h
+++ b/drivers/net/cxgbe/base/t4_regs.h
@@ -601,6 +601,7 @@
 
 /* registers for module MPS */
 #define MPS_BASE_ADDR 0x9000
+#define T4VF_MPS_BASE_ADDR 0x0100
 
 #define S_REPLICATE    11
 #define V_REPLICATE(x) ((x) << S_REPLICATE)
@@ -807,6 +808,15 @@
 #define A_MPS_STAT_RX_BG_3_LB_TRUNC_FRAME_L 0x96b8
 #define A_MPS_STAT_RX_BG_3_LB_TRUNC_FRAME_H 0x96bc
 
+#define A_MPS_VF_STAT_TX_VF_BCAST_FRAMES_L 0x88
+#define A_MPS_VF_STAT_TX_VF_MCAST_FRAMES_L 0x98
+#define A_MPS_VF_STAT_TX_VF_UCAST_FRAMES_L 0xa8
+#define A_MPS_VF_STAT_TX_VF_DROP_FRAMES_L 0xb0
+#define A_MPS_VF_STAT_RX_VF_BCAST_FRAMES_L 0xd0
+#define A_MPS_VF_STAT_RX_VF_MCAST_FRAMES_L 0xe0
+#define A_MPS_VF_STAT_RX_VF_UCAST_FRAMES_L 0xf0
+#define A_MPS_VF_STAT_RX_VF_ERR_FRAMES_L 0xf8
+
 /* registers for module ULP_RX */
 #define ULP_RX_BASE_ADDR 0x19150
 
diff --git a/drivers/net/cxgbe/base/t4fw_interface.h b/drivers/net/cxgbe/base/t4fw_interface.h
index 274f00b95..68b633e3f 100644
--- a/drivers/net/cxgbe/base/t4fw_interface.h
+++ b/drivers/net/cxgbe/base/t4fw_interface.h
@@ -210,6 +210,7 @@ enum fw_cmd_opcodes {
 	FW_VI_MAC_CMD                  = 0x15,
 	FW_VI_RXMODE_CMD               = 0x16,
 	FW_VI_ENABLE_CMD               = 0x17,
+	FW_VI_STATS_CMD		       = 0x1a,
 	FW_PORT_CMD                    = 0x1b,
 	FW_RSS_IND_TBL_CMD             = 0x20,
 	FW_RSS_GLB_CONFIG_CMD	       = 0x22,
@@ -1183,6 +1184,9 @@ struct fw_vi_enable_cmd {
 	(((x) >> S_FW_VI_ENABLE_CMD_DCB_INFO) & M_FW_VI_ENABLE_CMD_DCB_INFO)
 #define F_FW_VI_ENABLE_CMD_DCB_INFO	V_FW_VI_ENABLE_CMD_DCB_INFO(1U)
 
+/* VI VF stats offset definitions */
+#define VI_VF_NUM_STATS 16
+
 /* VI PF stats offset definitions */
 #define VI_PF_NUM_STATS	17
 enum fw_vi_stats_pf_index {
@@ -1260,6 +1264,15 @@ struct fw_vi_stats_cmd {
 	} u;
 };
 
+#define S_FW_VI_STATS_CMD_VIID		0
+#define V_FW_VI_STATS_CMD_VIID(x)	((x) << S_FW_VI_STATS_CMD_VIID)
+
+#define S_FW_VI_STATS_CMD_NSTATS	12
+#define V_FW_VI_STATS_CMD_NSTATS(x)	((x) << S_FW_VI_STATS_CMD_NSTATS)
+
+#define S_FW_VI_STATS_CMD_IX		0
+#define V_FW_VI_STATS_CMD_IX(x)		((x) << S_FW_VI_STATS_CMD_IX)
+
 /* old 16-bit port capabilities bitmap */
 enum fw_port_cap {
 	FW_PORT_CAP_SPEED_100M		= 0x0001,
diff --git a/drivers/net/cxgbe/base/t4vf_hw.c b/drivers/net/cxgbe/base/t4vf_hw.c
index f0408757d..9fd0b8791 100644
--- a/drivers/net/cxgbe/base/t4vf_hw.c
+++ b/drivers/net/cxgbe/base/t4vf_hw.c
@@ -630,6 +630,111 @@ int t4vf_get_vfres(struct adapter *adapter)
 	return 0;
 }
 
+/**
+ * t4vf_get_port_stats_fw - collect "port" statistics via Firmware
+ * @adapter: the adapter
+ * @pidx: the port index
+ * @s: the stats structure to fill
+ *
+ * Collect statistics for the "port"'s Virtual Interface via Firmware
+ * commands.
+ */
+static int t4vf_get_port_stats_fw(struct adapter *adapter, int pidx,
+				  struct port_stats *p)
+{
+	struct port_info *pi = adap2pinfo(adapter, pidx);
+	unsigned int rem = VI_VF_NUM_STATS;
+	struct fw_vi_stats_vf fwstats;
+	__be64 *fwsp = (__be64 *)&fwstats;
+
+	/*
+	 * Grab the Virtual Interface statistics a chunk at a time via mailbox
+	 * commands.  We could use a Work Request and get all of them at once
+	 * but that's an asynchronous interface which is awkward to use.
+	 */
+	while (rem) {
+		unsigned int ix = VI_VF_NUM_STATS - rem;
+		unsigned int nstats = min(6U, rem);
+		struct fw_vi_stats_cmd cmd, rpl;
+		size_t len = (offsetof(struct fw_vi_stats_cmd, u) +
+			      sizeof(struct fw_vi_stats_ctl));
+		size_t len16 = DIV_ROUND_UP(len, 16);
+		int ret;
+
+		memset(&cmd, 0, sizeof(cmd));
+		cmd.op_to_viid = cpu_to_be32(V_FW_CMD_OP(FW_VI_STATS_CMD) |
+					     V_FW_VI_STATS_CMD_VIID(pi->viid) |
+					     F_FW_CMD_REQUEST |
+					     F_FW_CMD_READ);
+		cmd.retval_len16 = cpu_to_be32(V_FW_CMD_LEN16(len16));
+		cmd.u.ctl.nstats_ix =
+			cpu_to_be16(V_FW_VI_STATS_CMD_IX(ix) |
+				    V_FW_VI_STATS_CMD_NSTATS(nstats));
+		ret = t4vf_wr_mbox_ns(adapter, &cmd, len, &rpl);
+		if (ret != FW_SUCCESS)
+			return ret;
+
+		memcpy(fwsp, &rpl.u.ctl.stat0, sizeof(__be64) * nstats);
+
+		rem -= nstats;
+		fwsp += nstats;
+	}
+
+	/*
+	 * Translate firmware statistics into host native statistics.
+	 */
+	p->tx_bcast_frames = be64_to_cpu(fwstats.tx_bcast_frames);
+	p->tx_mcast_frames = be64_to_cpu(fwstats.tx_mcast_frames);
+	p->tx_ucast_frames = be64_to_cpu(fwstats.tx_ucast_frames);
+	p->tx_drop = be64_to_cpu(fwstats.tx_drop_frames);
+
+	p->rx_bcast_frames = be64_to_cpu(fwstats.rx_bcast_frames);
+	p->rx_mcast_frames = be64_to_cpu(fwstats.rx_mcast_frames);
+	p->rx_ucast_frames = be64_to_cpu(fwstats.rx_ucast_frames);
+	p->rx_len_err = be64_to_cpu(fwstats.rx_err_frames);
+
+	return 0;
+}
+
+/**
+ *      t4vf_get_port_stats - collect "port" statistics
+ *      @adapter: the adapter
+ *      @pidx: the port index
+ *      @s: the stats structure to fill
+ *
+ *      Collect statistics for the "port"'s Virtual Interface.
+ */
+void t4vf_get_port_stats(struct adapter *adapter, int pidx,
+			 struct port_stats *p)
+{
+	/*
+	 * If this is not the first Virtual Interface for our Virtual
+	 * Function, we need to use Firmware commands to retrieve its
+	 * MPS statistics.
+	 */
+	if (pidx != 0)
+		t4vf_get_port_stats_fw(adapter, pidx, p);
+
+	/*
+	 * But for the first VI, we can grab its statistics via the MPS
+	 * register mapped into the VF register space.
+	 */
+#define GET_STAT(name) \
+	t4_read_reg64(adapter, \
+			T4VF_MPS_BASE_ADDR + A_MPS_VF_STAT_##name##_L)
+	p->tx_bcast_frames = GET_STAT(TX_VF_BCAST_FRAMES);
+	p->tx_mcast_frames = GET_STAT(TX_VF_MCAST_FRAMES);
+	p->tx_ucast_frames = GET_STAT(TX_VF_UCAST_FRAMES);
+	p->tx_drop = GET_STAT(TX_VF_DROP_FRAMES);
+
+	p->rx_bcast_frames = GET_STAT(RX_VF_BCAST_FRAMES);
+	p->rx_mcast_frames = GET_STAT(RX_VF_MCAST_FRAMES);
+	p->rx_ucast_frames = GET_STAT(RX_VF_UCAST_FRAMES);
+
+	p->rx_len_err = GET_STAT(RX_VF_ERR_FRAMES);
+#undef GET_STAT
+}
+
 static int t4vf_alloc_vi(struct adapter *adapter, int port_id)
 {
 	struct fw_vi_cmd cmd, rpl;
diff --git a/drivers/net/cxgbe/cxgbe.h b/drivers/net/cxgbe/cxgbe.h
index 748037d83..da8bdd03c 100644
--- a/drivers/net/cxgbe/cxgbe.h
+++ b/drivers/net/cxgbe/cxgbe.h
@@ -55,6 +55,7 @@ int cxgbe_up(struct adapter *adap);
 int cxgbe_down(struct port_info *pi);
 void cxgbe_close(struct adapter *adapter);
 void cxgbe_stats_get(struct port_info *pi, struct port_stats *stats);
+void cxgbevf_stats_get(struct port_info *pi, struct port_stats *stats);
 void cxgbe_stats_reset(struct port_info *pi);
 int link_start(struct port_info *pi);
 void init_rspq(struct adapter *adap, struct sge_rspq *q, unsigned int us,
diff --git a/drivers/net/cxgbe/cxgbevf_ethdev.c b/drivers/net/cxgbe/cxgbevf_ethdev.c
index a96630341..27308c71f 100644
--- a/drivers/net/cxgbe/cxgbevf_ethdev.c
+++ b/drivers/net/cxgbe/cxgbevf_ethdev.c
@@ -30,6 +30,49 @@
  */
 #include "t4_pci_id_tbl.h"
 
+/*
+ * Get port statistics.
+ */
+static int cxgbevf_dev_stats_get(struct rte_eth_dev *eth_dev,
+				 struct rte_eth_stats *eth_stats)
+{
+	struct port_info *pi = (struct port_info *)(eth_dev->data->dev_private);
+	struct adapter *adapter = pi->adapter;
+	struct sge *s = &adapter->sge;
+	struct port_stats ps;
+	unsigned int i;
+
+	cxgbevf_stats_get(pi, &ps);
+
+	/* RX Stats */
+	eth_stats->ierrors  = ps.rx_len_err;
+
+	/* TX Stats */
+	eth_stats->opackets = ps.tx_bcast_frames + ps.tx_mcast_frames +
+			      ps.tx_ucast_frames;
+	eth_stats->oerrors  = ps.tx_drop;
+
+	for (i = 0; i < pi->n_rx_qsets; i++) {
+		struct sge_eth_rxq *rxq =
+			&s->ethrxq[pi->first_qset + i];
+
+		eth_stats->q_ipackets[i] = rxq->stats.pkts;
+		eth_stats->q_ibytes[i] = rxq->stats.rx_bytes;
+		eth_stats->ipackets += eth_stats->q_ipackets[i];
+		eth_stats->ibytes += eth_stats->q_ibytes[i];
+	}
+
+	for (i = 0; i < pi->n_tx_qsets; i++) {
+		struct sge_eth_txq *txq =
+			&s->ethtxq[pi->first_qset + i];
+
+		eth_stats->q_opackets[i] = txq->stats.pkts;
+		eth_stats->q_obytes[i] = txq->stats.tx_bytes;
+		eth_stats->q_errors[i] = txq->stats.mapping_err;
+	}
+	return 0;
+}
+
 static const struct eth_dev_ops cxgbevf_eth_dev_ops = {
 	.dev_start              = cxgbe_dev_start,
 	.dev_stop               = cxgbe_dev_stop,
@@ -51,6 +94,7 @@ static const struct eth_dev_ops cxgbevf_eth_dev_ops = {
 	.rx_queue_start         = cxgbe_dev_rx_queue_start,
 	.rx_queue_stop          = cxgbe_dev_rx_queue_stop,
 	.rx_queue_release       = cxgbe_dev_rx_queue_release,
+	.stats_get		= cxgbevf_dev_stats_get,
 };
 
 /*
diff --git a/drivers/net/cxgbe/cxgbevf_main.c b/drivers/net/cxgbe/cxgbevf_main.c
index f3b83ecb9..6c81fd12e 100644
--- a/drivers/net/cxgbe/cxgbevf_main.c
+++ b/drivers/net/cxgbe/cxgbevf_main.c
@@ -74,6 +74,11 @@ static void size_nports_qsets(struct adapter *adapter)
 	}
 }
 
+void cxgbevf_stats_get(struct port_info *pi, struct port_stats *stats)
+{
+	t4vf_get_port_stats(pi->adapter, pi->pidx, stats);
+}
+
 static int adap_init0vf(struct adapter *adapter)
 {
 	u32 param, val = 0;
-- 
2.14.1

^ permalink raw reply related	[flat|nested] 26+ messages in thread

* [PATCH 09/13] cxgbe: add support to set mac address
  2018-03-10 22:48 [PATCH 00/13] cxgbe: add CXGBE VF PMD and updates Rahul Lakkireddy
                   ` (7 preceding siblings ...)
  2018-03-10 22:48 ` [PATCH 08/13] cxgbe: add VF port statistics Rahul Lakkireddy
@ 2018-03-10 22:48 ` Rahul Lakkireddy
  2018-03-10 22:48 ` [PATCH 10/13] cxgbe: fix check to close other ports properly Rahul Lakkireddy
                   ` (5 subsequent siblings)
  14 siblings, 0 replies; 26+ messages in thread
From: Rahul Lakkireddy @ 2018-03-10 22:48 UTC (permalink / raw)
  To: dev; +Cc: kumaras, nirranjan, indranil

From: Kumar Sanghvi <kumaras@chelsio.com>

Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
---
 drivers/net/cxgbe/cxgbe_ethdev.c   | 17 +++++++++++++++++
 drivers/net/cxgbe/cxgbe_pfvf.h     |  1 +
 drivers/net/cxgbe/cxgbevf_ethdev.c |  1 +
 3 files changed, 19 insertions(+)

diff --git a/drivers/net/cxgbe/cxgbe_ethdev.c b/drivers/net/cxgbe/cxgbe_ethdev.c
index e84facd33..fdea65ba6 100644
--- a/drivers/net/cxgbe/cxgbe_ethdev.c
+++ b/drivers/net/cxgbe/cxgbe_ethdev.c
@@ -1030,6 +1030,22 @@ static int cxgbe_get_regs(struct rte_eth_dev *eth_dev,
 	return 0;
 }
 
+void cxgbe_mac_addr_set(struct rte_eth_dev *dev, struct ether_addr *addr)
+{
+	struct port_info *pi = (struct port_info *)(dev->data->dev_private);
+	struct adapter *adapter = pi->adapter;
+	int ret;
+
+	ret = t4_change_mac(adapter, adapter->mbox, pi->viid,
+			    pi->xact_addr_filt, (u8 *)addr, true, true);
+	if (ret < 0) {
+		dev_err(adapter, "failed to set mac addr; err = %d\n",
+			ret);
+		return;
+	}
+	pi->xact_addr_filt = ret;
+}
+
 static const struct eth_dev_ops cxgbe_eth_dev_ops = {
 	.dev_start		= cxgbe_dev_start,
 	.dev_stop		= cxgbe_dev_stop,
@@ -1061,6 +1077,7 @@ static const struct eth_dev_ops cxgbe_eth_dev_ops = {
 	.get_reg		= cxgbe_get_regs,
 	.rss_hash_update	= cxgbe_dev_rss_hash_update,
 	.rss_hash_conf_get	= cxgbe_dev_rss_hash_conf_get,
+	.mac_addr_set		= cxgbe_mac_addr_set,
 };
 
 /*
diff --git a/drivers/net/cxgbe/cxgbe_pfvf.h b/drivers/net/cxgbe/cxgbe_pfvf.h
index 19bfd6d92..0c1c17071 100644
--- a/drivers/net/cxgbe/cxgbe_pfvf.h
+++ b/drivers/net/cxgbe/cxgbe_pfvf.h
@@ -16,6 +16,7 @@ void cxgbe_dev_promiscuous_enable(struct rte_eth_dev *eth_dev);
 void cxgbe_dev_promiscuous_disable(struct rte_eth_dev *eth_dev);
 void cxgbe_dev_allmulticast_enable(struct rte_eth_dev *eth_dev);
 void cxgbe_dev_allmulticast_disable(struct rte_eth_dev *eth_dev);
+void cxgbe_mac_addr_set(struct rte_eth_dev *dev, struct ether_addr *addr);
 int cxgbe_dev_configure(struct rte_eth_dev *eth_dev);
 int cxgbe_dev_tx_queue_setup(struct rte_eth_dev *eth_dev, uint16_t queue_idx,
 			     uint16_t nb_desc, unsigned int socket_id,
diff --git a/drivers/net/cxgbe/cxgbevf_ethdev.c b/drivers/net/cxgbe/cxgbevf_ethdev.c
index 27308c71f..4885b9748 100644
--- a/drivers/net/cxgbe/cxgbevf_ethdev.c
+++ b/drivers/net/cxgbe/cxgbevf_ethdev.c
@@ -95,6 +95,7 @@ static const struct eth_dev_ops cxgbevf_eth_dev_ops = {
 	.rx_queue_stop          = cxgbe_dev_rx_queue_stop,
 	.rx_queue_release       = cxgbe_dev_rx_queue_release,
 	.stats_get		= cxgbevf_dev_stats_get,
+	.mac_addr_set		= cxgbe_mac_addr_set,
 };
 
 /*
-- 
2.14.1

^ permalink raw reply related	[flat|nested] 26+ messages in thread

* [PATCH 10/13] cxgbe: fix check to close other ports properly
  2018-03-10 22:48 [PATCH 00/13] cxgbe: add CXGBE VF PMD and updates Rahul Lakkireddy
                   ` (8 preceding siblings ...)
  2018-03-10 22:48 ` [PATCH 09/13] cxgbe: add support to set mac address Rahul Lakkireddy
@ 2018-03-10 22:48 ` Rahul Lakkireddy
  2018-03-28 17:24   ` Ferruh Yigit
  2018-03-10 22:48 ` [PATCH 11/13] cxgbe: export supported RSS hash functions Rahul Lakkireddy
                   ` (4 subsequent siblings)
  14 siblings, 1 reply; 26+ messages in thread
From: Rahul Lakkireddy @ 2018-03-10 22:48 UTC (permalink / raw)
  To: dev; +Cc: kumaras, nirranjan, indranil

Fixes: 084b35314ff2 ("net/cxgbe: rework ethdev device allocation")
Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
---
 drivers/net/cxgbe/cxgbe_main.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/cxgbe/cxgbe_main.c b/drivers/net/cxgbe/cxgbe_main.c
index 882bcb7ef..b3b2b006a 100644
--- a/drivers/net/cxgbe/cxgbe_main.c
+++ b/drivers/net/cxgbe/cxgbe_main.c
@@ -1246,7 +1246,7 @@ void cxgbe_close(struct adapter *adapter)
 					   adapter->pf, 0, pi->viid);
 			rte_free(pi->eth_dev->data->mac_addrs);
 			/* Skip first port since it'll be freed by DPDK stack */
-			if (!i) {
+			if (i) {
 				rte_free(pi->eth_dev->data->dev_private);
 				rte_eth_dev_release_port(pi->eth_dev);
 			}
-- 
2.14.1

^ permalink raw reply related	[flat|nested] 26+ messages in thread

* [PATCH 11/13] cxgbe: export supported RSS hash functions
  2018-03-10 22:48 [PATCH 00/13] cxgbe: add CXGBE VF PMD and updates Rahul Lakkireddy
                   ` (9 preceding siblings ...)
  2018-03-10 22:48 ` [PATCH 10/13] cxgbe: fix check to close other ports properly Rahul Lakkireddy
@ 2018-03-10 22:48 ` Rahul Lakkireddy
  2018-03-28 17:25   ` Ferruh Yigit
  2018-03-10 22:48 ` [PATCH 12/13] cxgbe: convert to SPDX license tags Rahul Lakkireddy
                   ` (3 subsequent siblings)
  14 siblings, 1 reply; 26+ messages in thread
From: Rahul Lakkireddy @ 2018-03-10 22:48 UTC (permalink / raw)
  To: dev; +Cc: kumaras, nirranjan, indranil

Export supported RSS hash functions in device info.  Also add check to
prevent configuring unsupported RSS hash functions.

Fixes: 58c5a23c1c4f ("net/cxgbe: support updating RSS hash configuration and key")
Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
---
 drivers/net/cxgbe/cxgbe.h        | 5 +++++
 drivers/net/cxgbe/cxgbe_ethdev.c | 1 +
 drivers/net/cxgbe/cxgbe_main.c   | 6 +++++-
 3 files changed, 11 insertions(+), 1 deletion(-)

diff --git a/drivers/net/cxgbe/cxgbe.h b/drivers/net/cxgbe/cxgbe.h
index da8bdd03c..9438239bc 100644
--- a/drivers/net/cxgbe/cxgbe.h
+++ b/drivers/net/cxgbe/cxgbe.h
@@ -47,6 +47,11 @@
 #define CXGBE_MAX_RX_PKTLEN (9000 + ETHER_HDR_LEN + ETHER_CRC_LEN) /* max pkt */
 
 #define CXGBE_DEFAULT_RSS_KEY_LEN     40 /* 320-bits */
+#define CXGBE_RSS_HF_ALL (ETH_RSS_IPV4 | ETH_RSS_IPV6 | \
+			  ETH_RSS_NONFRAG_IPV4_TCP | \
+			  ETH_RSS_NONFRAG_IPV4_UDP | \
+			  ETH_RSS_NONFRAG_IPV6_TCP | \
+			  ETH_RSS_NONFRAG_IPV6_UDP)
 
 int cxgbe_probe(struct adapter *adapter);
 int cxgbevf_probe(struct adapter *adapter);
diff --git a/drivers/net/cxgbe/cxgbe_ethdev.c b/drivers/net/cxgbe/cxgbe_ethdev.c
index fdea65ba6..2610d62ef 100644
--- a/drivers/net/cxgbe/cxgbe_ethdev.c
+++ b/drivers/net/cxgbe/cxgbe_ethdev.c
@@ -173,6 +173,7 @@ void cxgbe_dev_info_get(struct rte_eth_dev *eth_dev,
 
 	device_info->reta_size = pi->rss_size;
 	device_info->hash_key_size = CXGBE_DEFAULT_RSS_KEY_LEN;
+	device_info->flow_type_rss_offloads = CXGBE_RSS_HF_ALL;
 
 	device_info->rx_desc_lim = cxgbe_desc_lim;
 	device_info->tx_desc_lim = cxgbe_desc_lim;
diff --git a/drivers/net/cxgbe/cxgbe_main.c b/drivers/net/cxgbe/cxgbe_main.c
index b3b2b006a..138e1c159 100644
--- a/drivers/net/cxgbe/cxgbe_main.c
+++ b/drivers/net/cxgbe/cxgbe_main.c
@@ -365,7 +365,7 @@ int init_rss(struct adapter *adap)
 		if (!pi->rss)
 			return -ENOMEM;
 
-		pi->rss_hf = ETH_RSS_TCP | ETH_RSS_UDP;
+		pi->rss_hf = CXGBE_RSS_HF_ALL;
 	}
 	return 0;
 }
@@ -957,6 +957,10 @@ int cxgbe_write_rss_conf(const struct port_info *pi, uint64_t rss_hf)
 		return -EINVAL;
 	}
 
+	/* Don't allow unsupported hash functions */
+	if (rss_hf & ~CXGBE_RSS_HF_ALL)
+		return -EINVAL;
+
 	if (rss_hf & ETH_RSS_IPV4)
 		flags |= F_FW_RSS_VI_CONFIG_CMD_IP4TWOTUPEN;
 
-- 
2.14.1

^ permalink raw reply related	[flat|nested] 26+ messages in thread

* [PATCH 12/13] cxgbe: convert to SPDX license tags
  2018-03-10 22:48 [PATCH 00/13] cxgbe: add CXGBE VF PMD and updates Rahul Lakkireddy
                   ` (10 preceding siblings ...)
  2018-03-10 22:48 ` [PATCH 11/13] cxgbe: export supported RSS hash functions Rahul Lakkireddy
@ 2018-03-10 22:48 ` Rahul Lakkireddy
  2018-03-10 22:48 ` [PATCH 13/13] cxgbe: add option to keep outer VLAN tag in Q-in-Q Rahul Lakkireddy
                   ` (2 subsequent siblings)
  14 siblings, 0 replies; 26+ messages in thread
From: Rahul Lakkireddy @ 2018-03-10 22:48 UTC (permalink / raw)
  To: dev; +Cc: kumaras, nirranjan, indranil

Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
---
 doc/guides/nics/cxgbe.rst               | 32 +++----------------------------
 drivers/net/cxgbe/Makefile              | 33 +++-----------------------------
 drivers/net/cxgbe/base/adapter.h        | 34 +++------------------------------
 drivers/net/cxgbe/base/common.h         | 34 +++------------------------------
 drivers/net/cxgbe/base/t4_chip_type.h   | 34 +++------------------------------
 drivers/net/cxgbe/base/t4_hw.c          | 34 +++------------------------------
 drivers/net/cxgbe/base/t4_hw.h          | 34 +++------------------------------
 drivers/net/cxgbe/base/t4_msg.h         | 34 +++------------------------------
 drivers/net/cxgbe/base/t4_pci_id_tbl.h  | 34 +++------------------------------
 drivers/net/cxgbe/base/t4_regs.h        | 34 +++------------------------------
 drivers/net/cxgbe/base/t4_regs_values.h | 34 +++------------------------------
 drivers/net/cxgbe/base/t4fw_interface.h | 34 +++------------------------------
 drivers/net/cxgbe/cxgbe.h               | 34 +++------------------------------
 drivers/net/cxgbe/cxgbe_compat.h        | 34 +++------------------------------
 drivers/net/cxgbe/cxgbe_ethdev.c        | 34 +++------------------------------
 drivers/net/cxgbe/cxgbe_main.c          | 34 +++------------------------------
 drivers/net/cxgbe/sge.c                 | 34 +++------------------------------
 17 files changed, 51 insertions(+), 524 deletions(-)

diff --git a/doc/guides/nics/cxgbe.rst b/doc/guides/nics/cxgbe.rst
index c4afe8678..f2e209901 100644
--- a/doc/guides/nics/cxgbe.rst
+++ b/doc/guides/nics/cxgbe.rst
@@ -1,32 +1,6 @@
-..  BSD LICENSE
-    Copyright 2015-2017 Chelsio Communications.
-    All rights reserved.
-
-    Redistribution and use in source and binary forms, with or without
-    modification, are permitted provided that the following conditions
-    are met:
-
-    * Redistributions of source code must retain the above copyright
-    notice, this list of conditions and the following disclaimer.
-    * Redistributions in binary form must reproduce the above copyright
-    notice, this list of conditions and the following disclaimer in
-    the documentation and/or other materials provided with the
-    distribution.
-    * Neither the name of Chelsio Communications nor the names of its
-    contributors may be used to endorse or promote products derived
-    from this software without specific prior written permission.
-
-    THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
-    "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
-    LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
-    A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
-    OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
-    SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
-    LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
-    DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
-    THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
-    (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
-    OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+.. SPDX-License-Identifier: BSD-3-Clause
+   Copyright(c) 2014-2018 Chelsio Communications.
+   All rights reserved.
 
 CXGBE Poll Mode Driver
 ======================
diff --git a/drivers/net/cxgbe/Makefile b/drivers/net/cxgbe/Makefile
index fe177ff90..8fba1a540 100644
--- a/drivers/net/cxgbe/Makefile
+++ b/drivers/net/cxgbe/Makefile
@@ -1,33 +1,6 @@
-#   BSD LICENSE
-#
-#   Copyright(c) 2014-2015 Chelsio Communications.
-#   All rights reserved.
-#
-#   Redistribution and use in source and binary forms, with or without
-#   modification, are permitted provided that the following conditions
-#   are met:
-#
-#     * Redistributions of source code must retain the above copyright
-#       notice, this list of conditions and the following disclaimer.
-#     * Redistributions in binary form must reproduce the above copyright
-#       notice, this list of conditions and the following disclaimer in
-#       the documentation and/or other materials provided with the
-#       distribution.
-#     * Neither the name of Chelsio Communications nor the names of its
-#       contributors may be used to endorse or promote products derived
-#       from this software without specific prior written permission.
-#
-#   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
-#   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
-#   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
-#   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
-#   OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
-#   SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
-#   LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
-#   DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
-#   THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
-#   (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
-#   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+# SPDX-License-Identifier: BSD-3-Clause
+# Copyright(c) 2014-2018 Chelsio Communications.
+# All rights reserved.
 
 include $(RTE_SDK)/mk/rte.vars.mk
 
diff --git a/drivers/net/cxgbe/base/adapter.h b/drivers/net/cxgbe/base/adapter.h
index fcea4055a..55cb2e91c 100644
--- a/drivers/net/cxgbe/base/adapter.h
+++ b/drivers/net/cxgbe/base/adapter.h
@@ -1,34 +1,6 @@
-/*-
- *   BSD LICENSE
- *
- *   Copyright(c) 2014-2017 Chelsio Communications.
- *   All rights reserved.
- *
- *   Redistribution and use in source and binary forms, with or without
- *   modification, are permitted provided that the following conditions
- *   are met:
- *
- *     * Redistributions of source code must retain the above copyright
- *       notice, this list of conditions and the following disclaimer.
- *     * Redistributions in binary form must reproduce the above copyright
- *       notice, this list of conditions and the following disclaimer in
- *       the documentation and/or other materials provided with the
- *       distribution.
- *     * Neither the name of Chelsio Communications nor the names of its
- *       contributors may be used to endorse or promote products derived
- *       from this software without specific prior written permission.
- *
- *   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
- *   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
- *   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
- *   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
- *   OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
- *   SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
- *   LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
- *   DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
- *   THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
- *   (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
- *   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2014-2018 Chelsio Communications.
+ * All rights reserved.
  */
 
 /* This file should not be included directly.  Include common.h instead. */
diff --git a/drivers/net/cxgbe/base/common.h b/drivers/net/cxgbe/base/common.h
index fcb6474e8..155a30288 100644
--- a/drivers/net/cxgbe/base/common.h
+++ b/drivers/net/cxgbe/base/common.h
@@ -1,34 +1,6 @@
-/*-
- *   BSD LICENSE
- *
- *   Copyright(c) 2014-2017 Chelsio Communications.
- *   All rights reserved.
- *
- *   Redistribution and use in source and binary forms, with or without
- *   modification, are permitted provided that the following conditions
- *   are met:
- *
- *     * Redistributions of source code must retain the above copyright
- *       notice, this list of conditions and the following disclaimer.
- *     * Redistributions in binary form must reproduce the above copyright
- *       notice, this list of conditions and the following disclaimer in
- *       the documentation and/or other materials provided with the
- *       distribution.
- *     * Neither the name of Chelsio Communications nor the names of its
- *       contributors may be used to endorse or promote products derived
- *       from this software without specific prior written permission.
- *
- *   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
- *   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
- *   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
- *   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
- *   OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
- *   SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
- *   LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
- *   DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
- *   THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
- *   (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
- *   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2014-2018 Chelsio Communications.
+ * All rights reserved.
  */
 
 #ifndef __CHELSIO_COMMON_H
diff --git a/drivers/net/cxgbe/base/t4_chip_type.h b/drivers/net/cxgbe/base/t4_chip_type.h
index cd7a9282d..c0c5d0b2c 100644
--- a/drivers/net/cxgbe/base/t4_chip_type.h
+++ b/drivers/net/cxgbe/base/t4_chip_type.h
@@ -1,34 +1,6 @@
-/*-
- *   BSD LICENSE
- *
- *   Copyright(c) 2014-2017 Chelsio Communications.
- *   All rights reserved.
- *
- *   Redistribution and use in source and binary forms, with or without
- *   modification, are permitted provided that the following conditions
- *   are met:
- *
- *     * Redistributions of source code must retain the above copyright
- *       notice, this list of conditions and the following disclaimer.
- *     * Redistributions in binary form must reproduce the above copyright
- *       notice, this list of conditions and the following disclaimer in
- *       the documentation and/or other materials provided with the
- *       distribution.
- *     * Neither the name of Chelsio Communications nor the names of its
- *       contributors may be used to endorse or promote products derived
- *       from this software without specific prior written permission.
- *
- *   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
- *   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
- *   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
- *   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
- *   OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
- *   SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
- *   LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
- *   DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
- *   THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
- *   (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
- *   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2014-2018 Chelsio Communications.
+ * All rights reserved.
  */
 
 #ifndef __T4_CHIP_TYPE_H__
diff --git a/drivers/net/cxgbe/base/t4_hw.c b/drivers/net/cxgbe/base/t4_hw.c
index b3938943c..e5ef73b67 100644
--- a/drivers/net/cxgbe/base/t4_hw.c
+++ b/drivers/net/cxgbe/base/t4_hw.c
@@ -1,34 +1,6 @@
-/*-
- *   BSD LICENSE
- *
- *   Copyright(c) 2014-2017 Chelsio Communications.
- *   All rights reserved.
- *
- *   Redistribution and use in source and binary forms, with or without
- *   modification, are permitted provided that the following conditions
- *   are met:
- *
- *     * Redistributions of source code must retain the above copyright
- *       notice, this list of conditions and the following disclaimer.
- *     * Redistributions in binary form must reproduce the above copyright
- *       notice, this list of conditions and the following disclaimer in
- *       the documentation and/or other materials provided with the
- *       distribution.
- *     * Neither the name of Chelsio Communications nor the names of its
- *       contributors may be used to endorse or promote products derived
- *       from this software without specific prior written permission.
- *
- *   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
- *   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
- *   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
- *   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
- *   OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
- *   SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
- *   LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
- *   DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
- *   THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
- *   (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
- *   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2014-2018 Chelsio Communications.
+ * All rights reserved.
  */
 
 #include <netinet/in.h>
diff --git a/drivers/net/cxgbe/base/t4_hw.h b/drivers/net/cxgbe/base/t4_hw.h
index 074988419..ac12afc04 100644
--- a/drivers/net/cxgbe/base/t4_hw.h
+++ b/drivers/net/cxgbe/base/t4_hw.h
@@ -1,34 +1,6 @@
-/*-
- *   BSD LICENSE
- *
- *   Copyright(c) 2014-2017 Chelsio Communications.
- *   All rights reserved.
- *
- *   Redistribution and use in source and binary forms, with or without
- *   modification, are permitted provided that the following conditions
- *   are met:
- *
- *     * Redistributions of source code must retain the above copyright
- *       notice, this list of conditions and the following disclaimer.
- *     * Redistributions in binary form must reproduce the above copyright
- *       notice, this list of conditions and the following disclaimer in
- *       the documentation and/or other materials provided with the
- *       distribution.
- *     * Neither the name of Chelsio Communications nor the names of its
- *       contributors may be used to endorse or promote products derived
- *       from this software without specific prior written permission.
- *
- *   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
- *   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
- *   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
- *   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
- *   OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
- *   SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
- *   LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
- *   DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
- *   THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
- *   (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
- *   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2014-2018 Chelsio Communications.
+ * All rights reserved.
  */
 
 #ifndef __T4_HW_H
diff --git a/drivers/net/cxgbe/base/t4_msg.h b/drivers/net/cxgbe/base/t4_msg.h
index 6acd749ad..74b4fc193 100644
--- a/drivers/net/cxgbe/base/t4_msg.h
+++ b/drivers/net/cxgbe/base/t4_msg.h
@@ -1,34 +1,6 @@
-/*-
- *   BSD LICENSE
- *
- *   Copyright(c) 2014-2017 Chelsio Communications.
- *   All rights reserved.
- *
- *   Redistribution and use in source and binary forms, with or without
- *   modification, are permitted provided that the following conditions
- *   are met:
- *
- *     * Redistributions of source code must retain the above copyright
- *       notice, this list of conditions and the following disclaimer.
- *     * Redistributions in binary form must reproduce the above copyright
- *       notice, this list of conditions and the following disclaimer in
- *       the documentation and/or other materials provided with the
- *       distribution.
- *     * Neither the name of Chelsio Communications nor the names of its
- *       contributors may be used to endorse or promote products derived
- *       from this software without specific prior written permission.
- *
- *   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
- *   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
- *   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
- *   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
- *   OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
- *   SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
- *   LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
- *   DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
- *   THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
- *   (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
- *   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2014-2018 Chelsio Communications.
+ * All rights reserved.
  */
 
 #ifndef T4_MSG_H
diff --git a/drivers/net/cxgbe/base/t4_pci_id_tbl.h b/drivers/net/cxgbe/base/t4_pci_id_tbl.h
index 1230e7383..5f5cbe048 100644
--- a/drivers/net/cxgbe/base/t4_pci_id_tbl.h
+++ b/drivers/net/cxgbe/base/t4_pci_id_tbl.h
@@ -1,34 +1,6 @@
-/*-
- *   BSD LICENSE
- *
- *   Copyright(c) 2014-2017 Chelsio Communications.
- *   All rights reserved.
- *
- *   Redistribution and use in source and binary forms, with or without
- *   modification, are permitted provided that the following conditions
- *   are met:
- *
- *     * Redistributions of source code must retain the above copyright
- *       notice, this list of conditions and the following disclaimer.
- *     * Redistributions in binary form must reproduce the above copyright
- *       notice, this list of conditions and the following disclaimer in
- *       the documentation and/or other materials provided with the
- *       distribution.
- *     * Neither the name of Chelsio Communications nor the names of its
- *       contributors may be used to endorse or promote products derived
- *       from this software without specific prior written permission.
- *
- *   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
- *   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
- *   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
- *   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
- *   OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
- *   SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
- *   LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
- *   DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
- *   THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
- *   (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
- *   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2014-2018 Chelsio Communications.
+ * All rights reserved.
  */
 
 #ifndef __T4_PCI_ID_TBL_H__
diff --git a/drivers/net/cxgbe/base/t4_regs.h b/drivers/net/cxgbe/base/t4_regs.h
index ae33aa33e..28ff21927 100644
--- a/drivers/net/cxgbe/base/t4_regs.h
+++ b/drivers/net/cxgbe/base/t4_regs.h
@@ -1,34 +1,6 @@
-/*-
- *   BSD LICENSE
- *
- *   Copyright(c) 2014-2017 Chelsio Communications.
- *   All rights reserved.
- *
- *   Redistribution and use in source and binary forms, with or without
- *   modification, are permitted provided that the following conditions
- *   are met:
- *
- *     * Redistributions of source code must retain the above copyright
- *       notice, this list of conditions and the following disclaimer.
- *     * Redistributions in binary form must reproduce the above copyright
- *       notice, this list of conditions and the following disclaimer in
- *       the documentation and/or other materials provided with the
- *       distribution.
- *     * Neither the name of Chelsio Communications nor the names of its
- *       contributors may be used to endorse or promote products derived
- *       from this software without specific prior written permission.
- *
- *   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
- *   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
- *   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
- *   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
- *   OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
- *   SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
- *   LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
- *   DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
- *   THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
- *   (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
- *   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2014-2018 Chelsio Communications.
+ * All rights reserved.
  */
 
 #define MYPF_BASE 0x1b000
diff --git a/drivers/net/cxgbe/base/t4_regs_values.h b/drivers/net/cxgbe/base/t4_regs_values.h
index 9085ff6d1..a9414d202 100644
--- a/drivers/net/cxgbe/base/t4_regs_values.h
+++ b/drivers/net/cxgbe/base/t4_regs_values.h
@@ -1,34 +1,6 @@
-/*-
- *   BSD LICENSE
- *
- *   Copyright(c) 2014-2017 Chelsio Communications.
- *   All rights reserved.
- *
- *   Redistribution and use in source and binary forms, with or without
- *   modification, are permitted provided that the following conditions
- *   are met:
- *
- *     * Redistributions of source code must retain the above copyright
- *       notice, this list of conditions and the following disclaimer.
- *     * Redistributions in binary form must reproduce the above copyright
- *       notice, this list of conditions and the following disclaimer in
- *       the documentation and/or other materials provided with the
- *       distribution.
- *     * Neither the name of Chelsio Communications nor the names of its
- *       contributors may be used to endorse or promote products derived
- *       from this software without specific prior written permission.
- *
- *   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
- *   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
- *   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
- *   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
- *   OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
- *   SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
- *   LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
- *   DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
- *   THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
- *   (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
- *   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2014-2018 Chelsio Communications.
+ * All rights reserved.
  */
 
 #ifndef __T4_REGS_VALUES_H__
diff --git a/drivers/net/cxgbe/base/t4fw_interface.h b/drivers/net/cxgbe/base/t4fw_interface.h
index 68b633e3f..852e8f3c7 100644
--- a/drivers/net/cxgbe/base/t4fw_interface.h
+++ b/drivers/net/cxgbe/base/t4fw_interface.h
@@ -1,34 +1,6 @@
-/*-
- *   BSD LICENSE
- *
- *   Copyright(c) 2014-2017 Chelsio Communications.
- *   All rights reserved.
- *
- *   Redistribution and use in source and binary forms, with or without
- *   modification, are permitted provided that the following conditions
- *   are met:
- *
- *     * Redistributions of source code must retain the above copyright
- *       notice, this list of conditions and the following disclaimer.
- *     * Redistributions in binary form must reproduce the above copyright
- *       notice, this list of conditions and the following disclaimer in
- *       the documentation and/or other materials provided with the
- *       distribution.
- *     * Neither the name of Chelsio Communications nor the names of its
- *       contributors may be used to endorse or promote products derived
- *       from this software without specific prior written permission.
- *
- *   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
- *   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
- *   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
- *   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
- *   OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
- *   SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
- *   LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
- *   DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
- *   THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
- *   (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
- *   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2014-2018 Chelsio Communications.
+ * All rights reserved.
  */
 
 #ifndef _T4FW_INTERFACE_H_
diff --git a/drivers/net/cxgbe/cxgbe.h b/drivers/net/cxgbe/cxgbe.h
index 9438239bc..b1a4e08df 100644
--- a/drivers/net/cxgbe/cxgbe.h
+++ b/drivers/net/cxgbe/cxgbe.h
@@ -1,34 +1,6 @@
-/*-
- *   BSD LICENSE
- *
- *   Copyright(c) 2014-2017 Chelsio Communications.
- *   All rights reserved.
- *
- *   Redistribution and use in source and binary forms, with or without
- *   modification, are permitted provided that the following conditions
- *   are met:
- *
- *     * Redistributions of source code must retain the above copyright
- *       notice, this list of conditions and the following disclaimer.
- *     * Redistributions in binary form must reproduce the above copyright
- *       notice, this list of conditions and the following disclaimer in
- *       the documentation and/or other materials provided with the
- *       distribution.
- *     * Neither the name of Chelsio Communications nor the names of its
- *       contributors may be used to endorse or promote products derived
- *       from this software without specific prior written permission.
- *
- *   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
- *   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
- *   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
- *   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
- *   OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
- *   SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
- *   LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
- *   DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
- *   THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
- *   (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
- *   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2014-2018 Chelsio Communications.
+ * All rights reserved.
  */
 
 #ifndef _CXGBE_H_
diff --git a/drivers/net/cxgbe/cxgbe_compat.h b/drivers/net/cxgbe/cxgbe_compat.h
index 03bba9fe5..779bcf165 100644
--- a/drivers/net/cxgbe/cxgbe_compat.h
+++ b/drivers/net/cxgbe/cxgbe_compat.h
@@ -1,34 +1,6 @@
-/*-
- *   BSD LICENSE
- *
- *   Copyright(c) 2014-2017 Chelsio Communications.
- *   All rights reserved.
- *
- *   Redistribution and use in source and binary forms, with or without
- *   modification, are permitted provided that the following conditions
- *   are met:
- *
- *     * Redistributions of source code must retain the above copyright
- *       notice, this list of conditions and the following disclaimer.
- *     * Redistributions in binary form must reproduce the above copyright
- *       notice, this list of conditions and the following disclaimer in
- *       the documentation and/or other materials provided with the
- *       distribution.
- *     * Neither the name of Chelsio Communications nor the names of its
- *       contributors may be used to endorse or promote products derived
- *       from this software without specific prior written permission.
- *
- *   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
- *   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
- *   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
- *   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
- *   OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
- *   SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
- *   LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
- *   DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
- *   THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
- *   (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
- *   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2014-2018 Chelsio Communications.
+ * All rights reserved.
  */
 
 #ifndef _CXGBE_COMPAT_H_
diff --git a/drivers/net/cxgbe/cxgbe_ethdev.c b/drivers/net/cxgbe/cxgbe_ethdev.c
index 2610d62ef..feae01d6a 100644
--- a/drivers/net/cxgbe/cxgbe_ethdev.c
+++ b/drivers/net/cxgbe/cxgbe_ethdev.c
@@ -1,34 +1,6 @@
-/*-
- *   BSD LICENSE
- *
- *   Copyright(c) 2014-2017 Chelsio Communications.
- *   All rights reserved.
- *
- *   Redistribution and use in source and binary forms, with or without
- *   modification, are permitted provided that the following conditions
- *   are met:
- *
- *     * Redistributions of source code must retain the above copyright
- *       notice, this list of conditions and the following disclaimer.
- *     * Redistributions in binary form must reproduce the above copyright
- *       notice, this list of conditions and the following disclaimer in
- *       the documentation and/or other materials provided with the
- *       distribution.
- *     * Neither the name of Chelsio Communications nor the names of its
- *       contributors may be used to endorse or promote products derived
- *       from this software without specific prior written permission.
- *
- *   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
- *   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
- *   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
- *   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
- *   OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
- *   SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
- *   LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
- *   DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
- *   THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
- *   (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
- *   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2014-2018 Chelsio Communications.
+ * All rights reserved.
  */
 
 #include <sys/queue.h>
diff --git a/drivers/net/cxgbe/cxgbe_main.c b/drivers/net/cxgbe/cxgbe_main.c
index 138e1c159..01a80ace8 100644
--- a/drivers/net/cxgbe/cxgbe_main.c
+++ b/drivers/net/cxgbe/cxgbe_main.c
@@ -1,34 +1,6 @@
-/*-
- *   BSD LICENSE
- *
- *   Copyright(c) 2014-2017 Chelsio Communications.
- *   All rights reserved.
- *
- *   Redistribution and use in source and binary forms, with or without
- *   modification, are permitted provided that the following conditions
- *   are met:
- *
- *     * Redistributions of source code must retain the above copyright
- *       notice, this list of conditions and the following disclaimer.
- *     * Redistributions in binary form must reproduce the above copyright
- *       notice, this list of conditions and the following disclaimer in
- *       the documentation and/or other materials provided with the
- *       distribution.
- *     * Neither the name of Chelsio Communications nor the names of its
- *       contributors may be used to endorse or promote products derived
- *       from this software without specific prior written permission.
- *
- *   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
- *   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
- *   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
- *   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
- *   OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
- *   SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
- *   LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
- *   DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
- *   THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
- *   (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
- *   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2014-2018 Chelsio Communications.
+ * All rights reserved.
  */
 
 #include <sys/queue.h>
diff --git a/drivers/net/cxgbe/sge.c b/drivers/net/cxgbe/sge.c
index 54e13fb9a..83e26d0c6 100644
--- a/drivers/net/cxgbe/sge.c
+++ b/drivers/net/cxgbe/sge.c
@@ -1,34 +1,6 @@
-/*-
- *   BSD LICENSE
- *
- *   Copyright(c) 2014-2015 Chelsio Communications.
- *   All rights reserved.
- *
- *   Redistribution and use in source and binary forms, with or without
- *   modification, are permitted provided that the following conditions
- *   are met:
- *
- *     * Redistributions of source code must retain the above copyright
- *       notice, this list of conditions and the following disclaimer.
- *     * Redistributions in binary form must reproduce the above copyright
- *       notice, this list of conditions and the following disclaimer in
- *       the documentation and/or other materials provided with the
- *       distribution.
- *     * Neither the name of Chelsio Communications nor the names of its
- *       contributors may be used to endorse or promote products derived
- *       from this software without specific prior written permission.
- *
- *   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
- *   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
- *   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
- *   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
- *   OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
- *   SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
- *   LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
- *   DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
- *   THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
- *   (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
- *   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2014-2018 Chelsio Communications.
+ * All rights reserved.
  */
 
 #include <sys/queue.h>
-- 
2.14.1

^ permalink raw reply related	[flat|nested] 26+ messages in thread

* [PATCH 13/13] cxgbe: add option to keep outer VLAN tag in Q-in-Q
  2018-03-10 22:48 [PATCH 00/13] cxgbe: add CXGBE VF PMD and updates Rahul Lakkireddy
                   ` (11 preceding siblings ...)
  2018-03-10 22:48 ` [PATCH 12/13] cxgbe: convert to SPDX license tags Rahul Lakkireddy
@ 2018-03-10 22:48 ` Rahul Lakkireddy
  2018-03-26 20:51 ` [PATCH 00/13] cxgbe: add CXGBE VF PMD and updates Ferruh Yigit
  2018-03-28 17:25 ` Ferruh Yigit
  14 siblings, 0 replies; 26+ messages in thread
From: Rahul Lakkireddy @ 2018-03-10 22:48 UTC (permalink / raw)
  To: dev; +Cc: kumaras, nirranjan, indranil

From: Kumar Sanghvi <kumaras@chelsio.com>

Add compile time option to keep outer VLAN tag in Q-in-Q packets.

Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
---
 config/common_base               |  1 +
 doc/guides/nics/cxgbe.rst        |  4 +++
 drivers/net/cxgbe/base/t4_regs.h | 54 ++++++++++++++++++++++++++++++++++++++++
 drivers/net/cxgbe/cxgbe_main.c   | 43 ++++++++++++++++++++++++++++++++
 drivers/net/cxgbe/sge.c          |  3 ++-
 5 files changed, 104 insertions(+), 1 deletion(-)

diff --git a/config/common_base b/config/common_base
index ad03cf433..325e9847c 100644
--- a/config/common_base
+++ b/config/common_base
@@ -167,6 +167,7 @@ CONFIG_RTE_LIBRTE_CXGBE_DEBUG_MBOX=n
 CONFIG_RTE_LIBRTE_CXGBE_DEBUG_TX=n
 CONFIG_RTE_LIBRTE_CXGBE_DEBUG_RX=n
 CONFIG_RTE_LIBRTE_CXGBE_TPUT=y
+CONFIG_RTE_LIBRTE_CXGBE_KEEP_OVLAN=n
 
 # NXP DPAA Bus
 CONFIG_RTE_LIBRTE_DPAA_BUS=n
diff --git a/doc/guides/nics/cxgbe.rst b/doc/guides/nics/cxgbe.rst
index f2e209901..a84547f6f 100644
--- a/doc/guides/nics/cxgbe.rst
+++ b/doc/guides/nics/cxgbe.rst
@@ -110,6 +110,10 @@ enabling debugging options may affect system performance.
 
   Toggle behaviour to prefer Throughput or Latency.
 
+- ``CONFIG_RTE_LIBRTE_CXGBE_KEEP_OVLAN`` (default **n**)
+
+  Toggle behaviour to keep/remove Outer VLAN in Q-in-Q.
+
 .. _driver-compilation:
 
 Driver compilation and testing
diff --git a/drivers/net/cxgbe/base/t4_regs.h b/drivers/net/cxgbe/base/t4_regs.h
index 28ff21927..c0d6ddcac 100644
--- a/drivers/net/cxgbe/base/t4_regs.h
+++ b/drivers/net/cxgbe/base/t4_regs.h
@@ -571,6 +571,9 @@
 #define V_CSUM_HAS_PSEUDO_HDR(x) ((x) << S_CSUM_HAS_PSEUDO_HDR)
 #define F_CSUM_HAS_PSEUDO_HDR    V_CSUM_HAS_PSEUDO_HDR(1U)
 
+#define S_RM_OVLAN	9
+#define V_RM_OVLAN(x)	((x) << S_RM_OVLAN)
+
 /* registers for module MPS */
 #define MPS_BASE_ADDR 0x9000
 #define T4VF_MPS_BASE_ADDR 0x0100
@@ -789,6 +792,57 @@
 #define A_MPS_VF_STAT_RX_VF_UCAST_FRAMES_L 0xf0
 #define A_MPS_VF_STAT_RX_VF_ERR_FRAMES_L 0xf8
 
+#define A_MPS_PORT0_RX_IVLAN	0x3011c
+
+#define S_IVLAN_ETYPE		0
+#define M_IVLAN_ETYPE		0xffffU
+#define V_IVLAN_ETYPE(x)	((x) << S_IVLAN_ETYPE)
+
+#define MPS_PORT_RX_IVLAN_STRIDE	0x4000
+#define MPS_PORT_RX_IVLAN(idx)		\
+	(A_MPS_PORT0_RX_IVLAN + (idx) * MPS_PORT_RX_IVLAN_STRIDE)
+
+#define A_MPS_PORT0_RX_OVLAN0	0x30120
+
+#define S_OVLAN_MASK	16
+#define M_OVLAN_MASK	0xffffU
+#define V_OVLAN_MASK(x)	((x) << S_OVLAN_MASK)
+
+#define S_OVLAN_ETYPE		0
+#define M_OVLAN_ETYPE		0xffffU
+#define V_OVLAN_ETYPE(x)	((x) << S_OVLAN_ETYPE)
+
+#define MPS_PORT_RX_OVLAN_STRIDE	0x4000
+#define MPS_PORT_RX_OVLAN_BASE(idx)	\
+(A_MPS_PORT0_RX_OVLAN0 + (idx) * MPS_PORT_RX_OVLAN_STRIDE)
+#define MPS_PORT_RX_OVLAN_REG(idx, reg)	(MPS_PORT_RX_OVLAN_BASE(idx) + (reg))
+
+#define A_RX_OVLAN0	0x0
+#define A_RX_OVLAN1	0x4
+#define A_RX_OVLAN2	0x8
+
+#define A_MPS_PORT0_RX_CTL	0x30100
+
+#define S_OVLAN_EN0	0
+#define V_OVLAN_EN0(x)	((x) << S_OVLAN_EN0)
+#define F_OVLAN_EN0	V_OVLAN_EN0(1)
+
+#define S_OVLAN_EN1	1
+#define V_OVLAN_EN1(x)	((x) << S_OVLAN_EN1)
+#define F_OVLAN_EN1	V_OVLAN_EN1(1)
+
+#define S_OVLAN_EN2	2
+#define V_OVLAN_EN2(x)	((x) << S_OVLAN_EN2)
+#define F_OVLAN_EN2	V_OVLAN_EN2(1)
+
+#define S_IVLAN_EN	4
+#define V_IVLAN_EN(x)	((x) << S_IVLAN_EN)
+#define F_IVLAN_EN	V_IVLAN_EN(1)
+
+#define MPS_PORT_RX_CTL_STRIDE	0x4000
+#define MPS_PORT_RX_CTL(idx)	\
+	(A_MPS_PORT0_RX_CTL + (idx) * MPS_PORT_RX_CTL_STRIDE)
+
 /* registers for module ULP_RX */
 #define ULP_RX_BASE_ADDR 0x19150
 
diff --git a/drivers/net/cxgbe/cxgbe_main.c b/drivers/net/cxgbe/cxgbe_main.c
index 01a80ace8..7cbdb2aa0 100644
--- a/drivers/net/cxgbe/cxgbe_main.c
+++ b/drivers/net/cxgbe/cxgbe_main.c
@@ -392,6 +392,48 @@ void print_port_info(struct adapter *adap)
 	}
 }
 
+static void configure_vlan_types(struct adapter *adapter)
+{
+	int i;
+
+	for_each_port(adapter, i) {
+		/* OVLAN Type 0x88a8 */
+		t4_set_reg_field(adapter, MPS_PORT_RX_OVLAN_REG(i, A_RX_OVLAN0),
+				 V_OVLAN_MASK(M_OVLAN_MASK) |
+				 V_OVLAN_ETYPE(M_OVLAN_ETYPE),
+				 V_OVLAN_MASK(M_OVLAN_MASK) |
+				 V_OVLAN_ETYPE(0x88a8));
+		/* OVLAN Type 0x9100 */
+		t4_set_reg_field(adapter, MPS_PORT_RX_OVLAN_REG(i, A_RX_OVLAN1),
+				 V_OVLAN_MASK(M_OVLAN_MASK) |
+				 V_OVLAN_ETYPE(M_OVLAN_ETYPE),
+				 V_OVLAN_MASK(M_OVLAN_MASK) |
+				 V_OVLAN_ETYPE(0x9100));
+		/* OVLAN Type 0x8100 */
+		t4_set_reg_field(adapter, MPS_PORT_RX_OVLAN_REG(i, A_RX_OVLAN2),
+				 V_OVLAN_MASK(M_OVLAN_MASK) |
+				 V_OVLAN_ETYPE(M_OVLAN_ETYPE),
+				 V_OVLAN_MASK(M_OVLAN_MASK) |
+				 V_OVLAN_ETYPE(0x8100));
+
+		/* IVLAN 0X8100 */
+		t4_set_reg_field(adapter, MPS_PORT_RX_IVLAN(i),
+				 V_IVLAN_ETYPE(M_IVLAN_ETYPE),
+				 V_IVLAN_ETYPE(0x8100));
+
+		t4_set_reg_field(adapter, MPS_PORT_RX_CTL(i),
+				 F_OVLAN_EN0 | F_OVLAN_EN1 |
+				 F_OVLAN_EN2 | F_IVLAN_EN,
+				 F_OVLAN_EN0 | F_OVLAN_EN1 |
+				 F_OVLAN_EN2 | F_IVLAN_EN);
+	}
+
+#ifdef RTE_LIBRTE_CXGBE_KEEP_OVLAN
+	t4_tp_wr_bits_indirect(adapter, A_TP_INGRESS_CONFIG,
+			       V_RM_OVLAN(1), V_RM_OVLAN(0));
+#endif
+}
+
 static void configure_pcie_ext_tag(struct adapter *adapter)
 {
 	u16 v;
@@ -808,6 +850,7 @@ static int adap_init0(struct adapter *adap)
 	t4_init_sge_params(adap);
 	t4_init_tp_params(adap);
 	configure_pcie_ext_tag(adap);
+	configure_vlan_types(adap);
 
 	adap->params.drv_memwin = MEMWIN_NIC;
 	adap->flags |= FW_OK;
diff --git a/drivers/net/cxgbe/sge.c b/drivers/net/cxgbe/sge.c
index 83e26d0c6..0d866354e 100644
--- a/drivers/net/cxgbe/sge.c
+++ b/drivers/net/cxgbe/sge.c
@@ -1596,7 +1596,8 @@ static int process_responses(struct sge_rspq *q, int budget,
 				}
 
 				if (cpl->vlan_ex) {
-					pkt->ol_flags |= PKT_RX_VLAN;
+					pkt->ol_flags |= PKT_RX_VLAN |
+							 PKT_RX_VLAN_STRIPPED;
 					pkt->vlan_tci = ntohs(cpl->vlan);
 				}
 
-- 
2.14.1

^ permalink raw reply related	[flat|nested] 26+ messages in thread

* Re: [PATCH 00/13] cxgbe: add CXGBE VF PMD and updates
  2018-03-10 22:48 [PATCH 00/13] cxgbe: add CXGBE VF PMD and updates Rahul Lakkireddy
                   ` (12 preceding siblings ...)
  2018-03-10 22:48 ` [PATCH 13/13] cxgbe: add option to keep outer VLAN tag in Q-in-Q Rahul Lakkireddy
@ 2018-03-26 20:51 ` Ferruh Yigit
  2018-03-27  7:01   ` Rahul Lakkireddy
  2018-03-28 17:25 ` Ferruh Yigit
  14 siblings, 1 reply; 26+ messages in thread
From: Ferruh Yigit @ 2018-03-26 20:51 UTC (permalink / raw)
  To: Rahul Lakkireddy, dev; +Cc: kumaras, nirranjan, indranil

On 3/10/2018 10:48 PM, Rahul Lakkireddy wrote:
> Patches 1 - 9 add support for cxgbe VF driver.  Patches 10 - 12 fix
> bugs and convert license in cxgbe files to SPDX License Tag.  Patch
> 13 adds compile time option to keep outer vlan tag in Q-in-Q packet.
> 
> Patch 1 adds minimal cxgbe VF driver.

Can you please update driver documentation with new supported device?

> 
> Patch 2 adds firmware mailbox support for VF.
> 
> Patch 3 adds base functions to enable VF ports in subsequent patches.
> 
> Patch 4 adds cxgbe VF probe to initialize VF devices.
> 
> Patch 5 initializes SGE and queues for VF.
> 
> Patch 6 enables RSS for VF.
> 
> Patch 7 updates TX and RX path for VF.
> 
> Patch 8 adds support for VF port statistics.
> 
> Patch 9 adds support to set mac address.
> 
> Patch 10 fixes bug where the other ports under same PF are not closed
> properly.
> 
> Patch 11 exports RSS hash functions in device info and adds check
> to prevent configuring unsupported hash functions.
> 
> Patch 12 converts all cxgbe files to use SPDX license tag.
> 
> Patch 13 adds compile time option to keep outer vlan tag in Q-in-Q
> packet.

We are trying to reduce the config options, is it possible to provide this
functionality with a runtime option (devargs) ?

Or there is already an offload option DEV_RX_OFFLOAD_QINQ_STRIP, I guess this is
different (is it?), if so does it make sense to have another offload option to
cover your case?


> 
> Thanks,
> Rahul
> 
> Kumar Sanghvi (10):
>   cxgbe: add skeleton VF driver
>   cxgbe: add VF firmware mailbox support
>   cxgbe: add base for enabling VF ports
>   cxgbe: add probe to initialize VF devices
>   cxgbe: initialize SGE and queues for VF
>   cxgbe: enable RSS for VF
>   cxgbe: update TX and RX path for VF
>   cxgbe: add VF port statistics
>   cxgbe: add support to set mac address
>   cxgbe: add option to keep outer VLAN tag in Q-in-Q
> 
> Rahul Lakkireddy (3):
>   cxgbe: fix check to close other ports properly
>   cxgbe: export supported RSS hash functions
>   cxgbe: convert to SPDX license tags

<...>

^ permalink raw reply	[flat|nested] 26+ messages in thread

* Re: [PATCH 00/13] cxgbe: add CXGBE VF PMD and updates
  2018-03-26 20:51 ` [PATCH 00/13] cxgbe: add CXGBE VF PMD and updates Ferruh Yigit
@ 2018-03-27  7:01   ` Rahul Lakkireddy
  2018-03-27 17:26     ` Ferruh Yigit
  2018-03-28  4:49     ` Shahaf Shuler
  0 siblings, 2 replies; 26+ messages in thread
From: Rahul Lakkireddy @ 2018-03-27  7:01 UTC (permalink / raw)
  To: Ferruh Yigit; +Cc: dev, Kumar A S, Nirranjan Kirubaharan, Indranil Choudhury

On Tuesday, March 03/27/18, 2018 at 02:21:56 +0530, Ferruh Yigit wrote:
> On 3/10/2018 10:48 PM, Rahul Lakkireddy wrote:
> > Patches 1 - 9 add support for cxgbe VF driver.  Patches 10 - 12 fix
> > bugs and convert license in cxgbe files to SPDX License Tag.  Patch
> > 13 adds compile time option to keep outer vlan tag in Q-in-Q packet.
> > 
> > Patch 1 adds minimal cxgbe VF driver.
> 
> Can you please update driver documentation with new supported device?
> 

There is no addition of new devices. The existing supported devices will
be supported for VF PMD as well.

> > 
> > Patch 2 adds firmware mailbox support for VF.
> > 
> > Patch 3 adds base functions to enable VF ports in subsequent patches.
> > 
> > Patch 4 adds cxgbe VF probe to initialize VF devices.
> > 
> > Patch 5 initializes SGE and queues for VF.
> > 
> > Patch 6 enables RSS for VF.
> > 
> > Patch 7 updates TX and RX path for VF.
> > 
> > Patch 8 adds support for VF port statistics.
> > 
> > Patch 9 adds support to set mac address.
> > 
> > Patch 10 fixes bug where the other ports under same PF are not closed
> > properly.
> > 
> > Patch 11 exports RSS hash functions in device info and adds check
> > to prevent configuring unsupported hash functions.
> > 
> > Patch 12 converts all cxgbe files to use SPDX license tag.
> > 
> > Patch 13 adds compile time option to keep outer vlan tag in Q-in-Q
> > packet.
> 
> We are trying to reduce the config options, is it possible to provide this
> functionality with a runtime option (devargs) ?
> 

Thank you for pointing to this. It seems like a good option.

> Or there is already an offload option DEV_RX_OFFLOAD_QINQ_STRIP, I guess this is
> different (is it?), if so does it make sense to have another offload option to
> cover your case?
> 
> 

Yes, this is different. Here, its about stripping or preserving
Outer-VLAN tag from double-vlan in Rx. We have few customers who need
this for their use-case. So, adding another offload option would also
help. Let us know which is preferred - either one, devargs OR another
offload, seems fine.

Thanks,
Rahul

^ permalink raw reply	[flat|nested] 26+ messages in thread

* Re: [PATCH 00/13] cxgbe: add CXGBE VF PMD and updates
  2018-03-27  7:01   ` Rahul Lakkireddy
@ 2018-03-27 17:26     ` Ferruh Yigit
  2018-03-28  7:44       ` Rahul Lakkireddy
  2018-03-28  4:49     ` Shahaf Shuler
  1 sibling, 1 reply; 26+ messages in thread
From: Ferruh Yigit @ 2018-03-27 17:26 UTC (permalink / raw)
  To: Rahul Lakkireddy
  Cc: dev, Kumar A S, Nirranjan Kirubaharan, Indranil Choudhury, Shahaf Shuler

On 3/27/2018 8:01 AM, Rahul Lakkireddy wrote:
> On Tuesday, March 03/27/18, 2018 at 02:21:56 +0530, Ferruh Yigit wrote:
>> On 3/10/2018 10:48 PM, Rahul Lakkireddy wrote:
>>> Patches 1 - 9 add support for cxgbe VF driver.  Patches 10 - 12 fix
>>> bugs and convert license in cxgbe files to SPDX License Tag.  Patch
>>> 13 adds compile time option to keep outer vlan tag in Q-in-Q packet.
>>>
>>> Patch 1 adds minimal cxgbe VF driver.
>>
>> Can you please update driver documentation with new supported device?
>>
> 
> There is no addition of new devices. The existing supported devices will
> be supported for VF PMD as well.

Sorry I didn't get this part, new PMD is introduced (VF driver) but no new
device support added? If so why new driver is added?

> 
>>>
>>> Patch 2 adds firmware mailbox support for VF.
>>>
>>> Patch 3 adds base functions to enable VF ports in subsequent patches.
>>>
>>> Patch 4 adds cxgbe VF probe to initialize VF devices.
>>>
>>> Patch 5 initializes SGE and queues for VF.
>>>
>>> Patch 6 enables RSS for VF.
>>>
>>> Patch 7 updates TX and RX path for VF.
>>>
>>> Patch 8 adds support for VF port statistics.
>>>
>>> Patch 9 adds support to set mac address.
>>>
>>> Patch 10 fixes bug where the other ports under same PF are not closed
>>> properly.
>>>
>>> Patch 11 exports RSS hash functions in device info and adds check
>>> to prevent configuring unsupported hash functions.
>>>
>>> Patch 12 converts all cxgbe files to use SPDX license tag.
>>>
>>> Patch 13 adds compile time option to keep outer vlan tag in Q-in-Q
>>> packet.
>>
>> We are trying to reduce the config options, is it possible to provide this
>> functionality with a runtime option (devargs) ?
>>
> 
> Thank you for pointing to this. It seems like a good option.
> 
>> Or there is already an offload option DEV_RX_OFFLOAD_QINQ_STRIP, I guess this is
>> different (is it?), if so does it make sense to have another offload option to
>> cover your case?
>>
>>
> 
> Yes, this is different. Here, its about stripping or preserving
> Outer-VLAN tag from double-vlan in Rx. We have few customers who need
> this for their use-case. So, adding another offload option would also
> help. Let us know which is preferred - either one, devargs OR another
> offload, seems fine.

As far as I can see this is done by hw configuration, so I would think this is a
kind of offload and tend to add it as another offload flag. cc'ed Shahaf for
comment.

Out of curiosity, if it is OK, what is the use case of keeping outer-VLAN tag
but remove inner-VLAN?

> 
> Thanks,
> Rahul
> 

^ permalink raw reply	[flat|nested] 26+ messages in thread

* Re: [PATCH 00/13] cxgbe: add CXGBE VF PMD and updates
  2018-03-27  7:01   ` Rahul Lakkireddy
  2018-03-27 17:26     ` Ferruh Yigit
@ 2018-03-28  4:49     ` Shahaf Shuler
  2018-03-28  7:39       ` Rahul Lakkireddy
  1 sibling, 1 reply; 26+ messages in thread
From: Shahaf Shuler @ 2018-03-28  4:49 UTC (permalink / raw)
  To: Rahul Lakkireddy, Ferruh Yigit
  Cc: dev, Kumar A S, Nirranjan Kirubaharan, Indranil Choudhury

Tuesday, March 27, 2018 10:02 AM, Rahul Lakkireddy:

[...]

> > >
> > > Patch 13 adds compile time option to keep outer vlan tag in Q-in-Q
> > > packet.
> >
> > We are trying to reduce the config options, is it possible to provide
> > this functionality with a runtime option (devargs) ?
> >
> 
> Thank you for pointing to this. It seems like a good option.
> 
> > Or there is already an offload option DEV_RX_OFFLOAD_QINQ_STRIP, I
> > guess this is different (is it?), if so does it make sense to have
> > another offload option to cover your case?
> >
> >
> 
> Yes, this is different. Here, its about stripping or preserving Outer-VLAN tag
> from double-vlan in Rx. We have few customers who need this for their use-
> case. So, adding another offload option would also help. Let us know which is
> preferred - either one, devargs OR another offload, seems fine.

What exactly is the offload you try to expose? To strip the inner VLAN while keeping the outer one? 

> 
> Thanks,
> Rahul

^ permalink raw reply	[flat|nested] 26+ messages in thread

* Re: [PATCH 00/13] cxgbe: add CXGBE VF PMD and updates
  2018-03-28  4:49     ` Shahaf Shuler
@ 2018-03-28  7:39       ` Rahul Lakkireddy
  2018-03-28  8:30         ` Shahaf Shuler
  0 siblings, 1 reply; 26+ messages in thread
From: Rahul Lakkireddy @ 2018-03-28  7:39 UTC (permalink / raw)
  To: Shahaf Shuler
  Cc: Ferruh Yigit, dev, Kumar A S, Nirranjan Kirubaharan, Indranil Choudhury

On Wednesday, March 03/28/18, 2018 at 10:19:17 +0530, Shahaf Shuler wrote:
> Tuesday, March 27, 2018 10:02 AM, Rahul Lakkireddy:
> 
> [...]
> 
> > > >
> > > > Patch 13 adds compile time option to keep outer vlan tag in Q-in-Q
> > > > packet.
> > >
> > > We are trying to reduce the config options, is it possible to provide
> > > this functionality with a runtime option (devargs) ?
> > >
> > 
> > Thank you for pointing to this. It seems like a good option.
> > 
> > > Or there is already an offload option DEV_RX_OFFLOAD_QINQ_STRIP, I
> > > guess this is different (is it?), if so does it make sense to have
> > > another offload option to cover your case?
> > >
> > >
> > 
> > Yes, this is different. Here, its about stripping or preserving Outer-VLAN tag
> > from double-vlan in Rx. We have few customers who need this for their use-
> > case. So, adding another offload option would also help. Let us know which is
> > preferred - either one, devargs OR another offload, seems fine.
> 
> What exactly is the offload you try to expose? To strip the inner VLAN while keeping the outer one? 
> 

Sorry about that, I should have provided more details on this earlier.
Chelsio hardware supports below for double-vlans:
1) Dropping Outer VLAN in hw; inner vlan is stripped and available as
packet metadata.
2) Preserving Outer VLAN so, Outer VLAN is stripped and available as
packet metadata. Inner VLAN is available as part of packet itself.

So, the configurability here is with respect to Outer VLAN and
subsequently what is available as part of packet metadata (outer or
inner vlan).  Inner VLAN gets retained in both cases.  However, based
on how Outer VLAN is configured, Inner VLAN will be available either
as packet metadata or as part of packet itself.

We are not very sure if DEV_RX_OFFLOAD_QINQ_STRIP could be used here.
If it holds the same meaning then we will use the same.

Thanks,
Rahul

^ permalink raw reply	[flat|nested] 26+ messages in thread

* Re: [PATCH 00/13] cxgbe: add CXGBE VF PMD and updates
  2018-03-27 17:26     ` Ferruh Yigit
@ 2018-03-28  7:44       ` Rahul Lakkireddy
  0 siblings, 0 replies; 26+ messages in thread
From: Rahul Lakkireddy @ 2018-03-28  7:44 UTC (permalink / raw)
  To: Ferruh Yigit
  Cc: dev, Kumar A S, Nirranjan Kirubaharan, Indranil Choudhury, Shahaf Shuler

On Tuesday, March 03/27/18, 2018 at 22:56:44 +0530, Ferruh Yigit wrote:
> On 3/27/2018 8:01 AM, Rahul Lakkireddy wrote:
> > On Tuesday, March 03/27/18, 2018 at 02:21:56 +0530, Ferruh Yigit wrote:
> >> On 3/10/2018 10:48 PM, Rahul Lakkireddy wrote:
> >>> Patches 1 - 9 add support for cxgbe VF driver.  Patches 10 - 12 fix
> >>> bugs and convert license in cxgbe files to SPDX License Tag.  Patch
> >>> 13 adds compile time option to keep outer vlan tag in Q-in-Q packet.
> >>>
> >>> Patch 1 adds minimal cxgbe VF driver.
> >>
> >> Can you please update driver documentation with new supported device?
> >>
> > 
> > There is no addition of new devices. The existing supported devices will
> > be supported for VF PMD as well.
> 
> Sorry I didn't get this part, new PMD is introduced (VF driver) but no new
> device support added? If so why new driver is added?
> 

Sorry for the confusion. By mistake, I replied in relation to existing
cxgbe PMD.  New device ids are already handled by Patch 1/13 for VF.
Will get the docs updated.

> > 
> >>>
> >>> Patch 2 adds firmware mailbox support for VF.
> >>>
> >>> Patch 3 adds base functions to enable VF ports in subsequent patches.
> >>>
> >>> Patch 4 adds cxgbe VF probe to initialize VF devices.
> >>>
> >>> Patch 5 initializes SGE and queues for VF.
> >>>
> >>> Patch 6 enables RSS for VF.
> >>>
> >>> Patch 7 updates TX and RX path for VF.
> >>>
> >>> Patch 8 adds support for VF port statistics.
> >>>
> >>> Patch 9 adds support to set mac address.
> >>>
> >>> Patch 10 fixes bug where the other ports under same PF are not closed
> >>> properly.
> >>>
> >>> Patch 11 exports RSS hash functions in device info and adds check
> >>> to prevent configuring unsupported hash functions.
> >>>
> >>> Patch 12 converts all cxgbe files to use SPDX license tag.
> >>>
> >>> Patch 13 adds compile time option to keep outer vlan tag in Q-in-Q
> >>> packet.
> >>
> >> We are trying to reduce the config options, is it possible to provide this
> >> functionality with a runtime option (devargs) ?
> >>
> > 
> > Thank you for pointing to this. It seems like a good option.
> > 
> >> Or there is already an offload option DEV_RX_OFFLOAD_QINQ_STRIP, I guess this is
> >> different (is it?), if so does it make sense to have another offload option to
> >> cover your case?
> >>
> >>
> > 
> > Yes, this is different. Here, its about stripping or preserving
> > Outer-VLAN tag from double-vlan in Rx. We have few customers who need
> > this for their use-case. So, adding another offload option would also
> > help. Let us know which is preferred - either one, devargs OR another
> > offload, seems fine.
> 
> As far as I can see this is done by hw configuration, so I would think this is a
> kind of offload and tend to add it as another offload flag. cc'ed Shahaf for
> comment.
> 

Replied to Shahaf in earlier mail with more details on how Chelsio HW
handles double-vlans.

> Out of curiosity, if it is OK, what is the use case of keeping outer-VLAN tag
> but remove inner-VLAN?
> 
> > 
> > Thanks,
> > Rahul
> > 
> 

^ permalink raw reply	[flat|nested] 26+ messages in thread

* Re: [PATCH 00/13] cxgbe: add CXGBE VF PMD and updates
  2018-03-28  7:39       ` Rahul Lakkireddy
@ 2018-03-28  8:30         ` Shahaf Shuler
  2018-03-28  9:29           ` Rahul Lakkireddy
  0 siblings, 1 reply; 26+ messages in thread
From: Shahaf Shuler @ 2018-03-28  8:30 UTC (permalink / raw)
  To: Rahul Lakkireddy
  Cc: Ferruh Yigit, dev, Kumar A S, Nirranjan Kirubaharan, Indranil Choudhury

Wednesday, March 28, 2018 10:40 AM, Rahul Lakkireddy:
> On Wednesday, March 03/28/18, 2018 at 10:19:17 +0530, Shahaf Shuler
> >
> > What exactly is the offload you try to expose? To strip the inner VLAN while
> keeping the outer one?
> >
> 
> Sorry about that, I should have provided more details on this earlier.
> Chelsio hardware supports below for double-vlans:
> 1) Dropping Outer VLAN in hw; inner vlan is stripped and available as packet
> metadata.

So this is basically the DEV_RX_OFFLOAD_QINQ_STRIP with limitation from your device you cannot return the outer vlan in the packet metadata

> 2) Preserving Outer VLAN so, Outer VLAN is stripped and available as packet
> metadata. Inner VLAN is available as part of packet itself.

And this one is the regular DEV_RX_OFFLOAD_VLAN_STRIP

> 
> So, the configurability here is with respect to Outer VLAN and subsequently
> what is available as part of packet metadata (outer or inner vlan).  Inner
> VLAN gets retained in both cases.  However, based on how Outer VLAN is
> configured, Inner VLAN will be available either as packet metadata or as part
> of packet itself.
> 
> We are not very sure if DEV_RX_OFFLOAD_QINQ_STRIP could be used here.
> If it holds the same meaning then we will use the same.

Yeah , we are missing good documentation of what each offload means. 
> 
> Thanks,
> Rahul

^ permalink raw reply	[flat|nested] 26+ messages in thread

* Re: [PATCH 00/13] cxgbe: add CXGBE VF PMD and updates
  2018-03-28  8:30         ` Shahaf Shuler
@ 2018-03-28  9:29           ` Rahul Lakkireddy
  2018-03-28 14:06             ` Ferruh Yigit
  0 siblings, 1 reply; 26+ messages in thread
From: Rahul Lakkireddy @ 2018-03-28  9:29 UTC (permalink / raw)
  To: Shahaf Shuler
  Cc: Ferruh Yigit, dev, Kumar A S, Nirranjan Kirubaharan, Indranil Choudhury

On Wednesday, March 03/28/18, 2018 at 14:00:58 +0530, Shahaf Shuler wrote:
> Wednesday, March 28, 2018 10:40 AM, Rahul Lakkireddy:
> > On Wednesday, March 03/28/18, 2018 at 10:19:17 +0530, Shahaf Shuler
> > >
> > > What exactly is the offload you try to expose? To strip the inner VLAN while
> > keeping the outer one?
> > >
> > 
> > Sorry about that, I should have provided more details on this earlier.
> > Chelsio hardware supports below for double-vlans:
> > 1) Dropping Outer VLAN in hw; inner vlan is stripped and available as packet
> > metadata.
> 
> So this is basically the DEV_RX_OFFLOAD_QINQ_STRIP with limitation from your device you cannot return the outer vlan in the packet metadata
> 

Since this is specific to Chelsio devices, is it better handled by
devargs approach instead? We will respin this particular patch
separately.
 
Hi Ferruh,
Should we re-post the entire series minus this particular patch?

> > 2) Preserving Outer VLAN so, Outer VLAN is stripped and available as packet
> > metadata. Inner VLAN is available as part of packet itself.
> 
> And this one is the regular DEV_RX_OFFLOAD_VLAN_STRIP
> 
> > 
> > So, the configurability here is with respect to Outer VLAN and subsequently
> > what is available as part of packet metadata (outer or inner vlan).  Inner
> > VLAN gets retained in both cases.  However, based on how Outer VLAN is
> > configured, Inner VLAN will be available either as packet metadata or as part
> > of packet itself.
> > 
> > We are not very sure if DEV_RX_OFFLOAD_QINQ_STRIP could be used here.
> > If it holds the same meaning then we will use the same.
> 
> Yeah , we are missing good documentation of what each offload means. 
> > 
> > Thanks,
> > Rahul

^ permalink raw reply	[flat|nested] 26+ messages in thread

* Re: [PATCH 00/13] cxgbe: add CXGBE VF PMD and updates
  2018-03-28  9:29           ` Rahul Lakkireddy
@ 2018-03-28 14:06             ` Ferruh Yigit
  0 siblings, 0 replies; 26+ messages in thread
From: Ferruh Yigit @ 2018-03-28 14:06 UTC (permalink / raw)
  To: Rahul Lakkireddy, Shahaf Shuler
  Cc: dev, Kumar A S, Nirranjan Kirubaharan, Indranil Choudhury

On 3/28/2018 10:29 AM, Rahul Lakkireddy wrote:
> On Wednesday, March 03/28/18, 2018 at 14:00:58 +0530, Shahaf Shuler wrote:
>> Wednesday, March 28, 2018 10:40 AM, Rahul Lakkireddy:
>>> On Wednesday, March 03/28/18, 2018 at 10:19:17 +0530, Shahaf Shuler
>>>>
>>>> What exactly is the offload you try to expose? To strip the inner VLAN while
>>> keeping the outer one?
>>>>
>>>
>>> Sorry about that, I should have provided more details on this earlier.
>>> Chelsio hardware supports below for double-vlans:
>>> 1) Dropping Outer VLAN in hw; inner vlan is stripped and available as packet
>>> metadata.
>>
>> So this is basically the DEV_RX_OFFLOAD_QINQ_STRIP with limitation from your device you cannot return the outer vlan in the packet metadata
>>
> 
> Since this is specific to Chelsio devices, is it better handled by
> devargs approach instead? We will respin this particular patch
> separately.
>  
> Hi Ferruh,
> Should we re-post the entire series minus this particular patch?

No new version required.
I will not merge patch 13/13.
Also expecting an incremental documentation patch to update device information.

> 
>>> 2) Preserving Outer VLAN so, Outer VLAN is stripped and available as packet
>>> metadata. Inner VLAN is available as part of packet itself.
>>
>> And this one is the regular DEV_RX_OFFLOAD_VLAN_STRIP
>>
>>>
>>> So, the configurability here is with respect to Outer VLAN and subsequently
>>> what is available as part of packet metadata (outer or inner vlan).  Inner
>>> VLAN gets retained in both cases.  However, based on how Outer VLAN is
>>> configured, Inner VLAN will be available either as packet metadata or as part
>>> of packet itself.
>>>
>>> We are not very sure if DEV_RX_OFFLOAD_QINQ_STRIP could be used here.
>>> If it holds the same meaning then we will use the same.
>>
>> Yeah , we are missing good documentation of what each offload means. 
>>>
>>> Thanks,
>>> Rahul

^ permalink raw reply	[flat|nested] 26+ messages in thread

* Re: [PATCH 10/13] cxgbe: fix check to close other ports properly
  2018-03-10 22:48 ` [PATCH 10/13] cxgbe: fix check to close other ports properly Rahul Lakkireddy
@ 2018-03-28 17:24   ` Ferruh Yigit
  0 siblings, 0 replies; 26+ messages in thread
From: Ferruh Yigit @ 2018-03-28 17:24 UTC (permalink / raw)
  To: Rahul Lakkireddy, dev; +Cc: kumaras, nirranjan, indranil

On 3/10/2018 10:48 PM, Rahul Lakkireddy wrote:
> Fixes: 084b35314ff2 ("net/cxgbe: rework ethdev device allocation")
> Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
> Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>

Squashed into relevant commit in next-net, thanks.

^ permalink raw reply	[flat|nested] 26+ messages in thread

* Re: [PATCH 11/13] cxgbe: export supported RSS hash functions
  2018-03-10 22:48 ` [PATCH 11/13] cxgbe: export supported RSS hash functions Rahul Lakkireddy
@ 2018-03-28 17:25   ` Ferruh Yigit
  0 siblings, 0 replies; 26+ messages in thread
From: Ferruh Yigit @ 2018-03-28 17:25 UTC (permalink / raw)
  To: Rahul Lakkireddy, dev; +Cc: kumaras, nirranjan, indranil

On 3/10/2018 10:48 PM, Rahul Lakkireddy wrote:
> Export supported RSS hash functions in device info.  Also add check to
> prevent configuring unsupported RSS hash functions.
> 
> Fixes: 58c5a23c1c4f ("net/cxgbe: support updating RSS hash configuration and key")
> Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
> Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>

Squashed into relevant commit in next-net, thanks.

^ permalink raw reply	[flat|nested] 26+ messages in thread

* Re: [PATCH 00/13] cxgbe: add CXGBE VF PMD and updates
  2018-03-10 22:48 [PATCH 00/13] cxgbe: add CXGBE VF PMD and updates Rahul Lakkireddy
                   ` (13 preceding siblings ...)
  2018-03-26 20:51 ` [PATCH 00/13] cxgbe: add CXGBE VF PMD and updates Ferruh Yigit
@ 2018-03-28 17:25 ` Ferruh Yigit
  14 siblings, 0 replies; 26+ messages in thread
From: Ferruh Yigit @ 2018-03-28 17:25 UTC (permalink / raw)
  To: Rahul Lakkireddy, dev; +Cc: kumaras, nirranjan, indranil

On 3/10/2018 10:48 PM, Rahul Lakkireddy wrote:
> Patches 1 - 9 add support for cxgbe VF driver.  Patches 10 - 12 fix
> bugs and convert license in cxgbe files to SPDX License Tag.  Patch
> 13 adds compile time option to keep outer vlan tag in Q-in-Q packet.
> 
> Patch 1 adds minimal cxgbe VF driver.
> 
> Patch 2 adds firmware mailbox support for VF.
> 
> Patch 3 adds base functions to enable VF ports in subsequent patches.
> 
> Patch 4 adds cxgbe VF probe to initialize VF devices.
> 
> Patch 5 initializes SGE and queues for VF.
> 
> Patch 6 enables RSS for VF.
> 
> Patch 7 updates TX and RX path for VF.
> 
> Patch 8 adds support for VF port statistics.
> 
> Patch 9 adds support to set mac address.
> 
> Patch 10 fixes bug where the other ports under same PF are not closed
> properly.
> 
> Patch 11 exports RSS hash functions in device info and adds check
> to prevent configuring unsupported hash functions.
> 
> Patch 12 converts all cxgbe files to use SPDX license tag.
> 
> Patch 13 adds compile time option to keep outer vlan tag in Q-in-Q
> packet.
> 
> Thanks,
> Rahul
> 
> Kumar Sanghvi (10):
>   cxgbe: add skeleton VF driver
>   cxgbe: add VF firmware mailbox support
>   cxgbe: add base for enabling VF ports
>   cxgbe: add probe to initialize VF devices
>   cxgbe: initialize SGE and queues for VF
>   cxgbe: enable RSS for VF
>   cxgbe: update TX and RX path for VF
>   cxgbe: add VF port statistics
>   cxgbe: add support to set mac address
>   cxgbe: add option to keep outer VLAN tag in Q-in-Q
> 
> Rahul Lakkireddy (3):
>   cxgbe: fix check to close other ports properly
>   cxgbe: export supported RSS hash functions
>   cxgbe: convert to SPDX license tags

Series applied to dpdk-next-net/master with exceptions, thanks.

Exceptions:
Patch 10/13 squashed into next-net
Patch 11/13 squashed into next-net
Patch 13/13 dropped from set and not merged

Also expecting an incremental documentation patch to update device information.

^ permalink raw reply	[flat|nested] 26+ messages in thread

end of thread, other threads:[~2018-03-28 17:26 UTC | newest]

Thread overview: 26+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2018-03-10 22:48 [PATCH 00/13] cxgbe: add CXGBE VF PMD and updates Rahul Lakkireddy
2018-03-10 22:48 ` [PATCH 01/13] cxgbe: add skeleton VF driver Rahul Lakkireddy
2018-03-10 22:48 ` [PATCH 02/13] cxgbe: add VF firmware mailbox support Rahul Lakkireddy
2018-03-10 22:48 ` [PATCH 03/13] cxgbe: add base for enabling VF ports Rahul Lakkireddy
2018-03-10 22:48 ` [PATCH 04/13] cxgbe: add probe to initialize VF devices Rahul Lakkireddy
2018-03-10 22:48 ` [PATCH 05/13] cxgbe: initialize SGE and queues for VF Rahul Lakkireddy
2018-03-10 22:48 ` [PATCH 06/13] cxgbe: enable RSS " Rahul Lakkireddy
2018-03-10 22:48 ` [PATCH 07/13] cxgbe: update TX and RX path " Rahul Lakkireddy
2018-03-10 22:48 ` [PATCH 08/13] cxgbe: add VF port statistics Rahul Lakkireddy
2018-03-10 22:48 ` [PATCH 09/13] cxgbe: add support to set mac address Rahul Lakkireddy
2018-03-10 22:48 ` [PATCH 10/13] cxgbe: fix check to close other ports properly Rahul Lakkireddy
2018-03-28 17:24   ` Ferruh Yigit
2018-03-10 22:48 ` [PATCH 11/13] cxgbe: export supported RSS hash functions Rahul Lakkireddy
2018-03-28 17:25   ` Ferruh Yigit
2018-03-10 22:48 ` [PATCH 12/13] cxgbe: convert to SPDX license tags Rahul Lakkireddy
2018-03-10 22:48 ` [PATCH 13/13] cxgbe: add option to keep outer VLAN tag in Q-in-Q Rahul Lakkireddy
2018-03-26 20:51 ` [PATCH 00/13] cxgbe: add CXGBE VF PMD and updates Ferruh Yigit
2018-03-27  7:01   ` Rahul Lakkireddy
2018-03-27 17:26     ` Ferruh Yigit
2018-03-28  7:44       ` Rahul Lakkireddy
2018-03-28  4:49     ` Shahaf Shuler
2018-03-28  7:39       ` Rahul Lakkireddy
2018-03-28  8:30         ` Shahaf Shuler
2018-03-28  9:29           ` Rahul Lakkireddy
2018-03-28 14:06             ` Ferruh Yigit
2018-03-28 17:25 ` Ferruh Yigit

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.