d/numam-dpdk
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
dc59260fb6
numam-dpdk/drivers/net/bnxt/bnxt_ethdev.c

6198 lines
152 KiB
C
Raw Normal View History

net/bnxt: convert to SPDX license tag Update the license header on bnxt files to be the standard BSD-3-Clause license used for the rest of DPDK, bring the files in compliance with the DPDK licensing policy. Signed-off-by: Scott Branden <scott.branden@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>
2018-04-02 22:34:32 +00:00
/* SPDX-License-Identifier: BSD-3-Clause
net/bnxt: update copyright year Update copyright year in various files. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
2021-01-27 16:09:43 +00:00
* Copyright(c) 2014-2021 Broadcom
net/bnxt: convert to SPDX license tag Update the license header on bnxt files to be the standard BSD-3-Clause license used for the rest of DPDK, bring the files in compliance with the DPDK licensing policy. Signed-off-by: Scott Branden <scott.branden@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>
2018-04-02 22:34:32 +00:00
* All rights reserved.
net/bnxt: add driver for Broadcom NetXtreme-C devices This patch adds the initial skeleton for bnxt driver along with the nic guide, and ties the driver into the build system. At this point, the driver simply fails init. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com> [Release Note Addition] Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2016-06-15 21:23:01 +00:00
*/
#include <inttypes.h>
#include <stdbool.h>
#include <rte_dev.h>
ethdev: make driver-only headers private The rte_ethdev_driver.h, rte_ethdev_vdev.h and rte_ethdev_pci.h files are for drivers only and should be a private to DPDK and not installed. Signed-off-by: Bruce Richardson <bruce.richardson@intel.com> Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com> Acked-by: Thomas Monjalon <thomas@monjalon.net> Acked-by: Steven Webster <steven.webster@windriver.com>
2021-01-29 16:48:19 +00:00
#include <ethdev_driver.h>
#include <ethdev_pci.h>
net/bnxt: add driver for Broadcom NetXtreme-C devices This patch adds the initial skeleton for bnxt driver along with the nic guide, and ties the driver into the build system. At this point, the driver simply fails init. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com> [Release Note Addition] Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2016-06-15 21:23:01 +00:00
#include <rte_malloc.h>
#include <rte_cycles.h>
net/bnxt: handle reset notify async event from FW When the FW upgrade is initiated the current instance of FW issues a HWRM_ASYNC_EVENT_CMPL_EVENT_ID_RESET_NOTIFY async notification to the driver. On receiving this notification, the PMD shall quiesce itself and poll on the HWRM_VER_GET FW command at regular intervals. Once the VER_GET command succeeds, the driver should go through the rediscovery process and re-initialize the device. Also register with FW for the reset notify async event. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:23 +00:00
#include <rte_alarm.h>
net/bnxt: support host memory based TruFlow This feature can be enabled by passing "-w 0000:0d:00.0,host-based-truflow=1” to the DPDK application. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-04-15 08:18:52 +00:00
#include <rte_kvargs.h>
net/bnxt: check max SIMD bitwidth When choosing a vector path to take, an extra condition must be satisfied to ensure the max SIMD bitwidth allows for the CPU enabled path. Signed-off-by: Ciara Power <ciara.power@intel.com> Acked-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-10-19 13:48:45 +00:00
#include <rte_vect.h>
net/bnxt: add driver for Broadcom NetXtreme-C devices This patch adds the initial skeleton for bnxt driver along with the nic guide, and ties the driver into the build system. At this point, the driver simply fails init. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com> [Release Note Addition] Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2016-06-15 21:23:01 +00:00
net/bnxt: add hardware resource manager init code Start adding support to use the HWRM API. Hardware Resource Manager or HWRM in short, is a set of API provided by the firmware running in the ASIC to manage the various resources. Initial commit just performs necessary HWRM queries for init, then fails as before. Now that struct bnxt is non-zero size, we can set dev_private_size correctly. The used HWRM calls so far: bnxt_hwrm_func_qcaps: This command returns capabilities of a function. bnxt_hwrm_ver_get: This function is called by a driver to determine the HWRM interface version supported by the HWRM firmware, the version of HWRM firmware implementation, the name of HWRM firmware, the versions of other embedded firmwares, and the names of other embedded firmwares, etc. Gets the firmware version and interface specifications. Returns an error if the firmware on the device is not supported by the driver and ensures the response space is large enough for the largest possible response. bnxt_hwrm_queue_qportcfg: This function is called by a driver to query queue configuration of a port. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:02 +00:00
#include "bnxt.h"
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
#include "bnxt_filter.h"
net/bnxt: add hardware resource manager init code Start adding support to use the HWRM API. Hardware Resource Manager or HWRM in short, is a set of API provided by the firmware running in the ASIC to manage the various resources. Initial commit just performs necessary HWRM queries for init, then fails as before. Now that struct bnxt is non-zero size, we can set dev_private_size correctly. The used HWRM calls so far: bnxt_hwrm_func_qcaps: This command returns capabilities of a function. bnxt_hwrm_ver_get: This function is called by a driver to determine the HWRM interface version supported by the HWRM firmware, the version of HWRM firmware implementation, the name of HWRM firmware, the versions of other embedded firmwares, and the names of other embedded firmwares, etc. Gets the firmware version and interface specifications. Returns an error if the firmware on the device is not supported by the driver and ensures the response space is large enough for the largest possible response. bnxt_hwrm_queue_qportcfg: This function is called by a driver to query queue configuration of a port. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:02 +00:00
#include "bnxt_hwrm.h"
net/bnxt: support async link notification This patch adds support to get Link notification asynchronously. The HW sends async notifications on default completion ring. The PMD processes these notifications and logs a message appropriately. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2016-10-11 21:47:50 +00:00
#include "bnxt_irq.h"
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
#include "bnxt_reps.h"
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
#include "bnxt_ring.h"
net/bnxt: add Rx queue create/destroy In this patch we are adding the bnxt_rx_queue_setup_op() and bnxt_rx_queue_release_op() functions. These will be tied to the rx_queue_setup and rx_queue_release dev_ops in a subsequent patch. In these functions we allocate/free memory for the RX queues. This still requires support to create a RX ring in the ASIC which will be completed in a future commit. Each Rx queue created via the rx_queue_setup dev_op will have an associated Rx ring in the hardware. The Rx logic in the hardware picks a Rx ring for each Rx frame received by the hardware depending on the properties like RSS, MAC and VLAN settings configured in the hardware. These packets in the end arrive on the Rx queue corresponding to the Rx ring in the hardware. We are also adding some functions like bnxt_mq_rx_configure() bnxt_free_rx_mbufs() and bnxt_free_rxq_stats() which will be used in subsequent patches. We are also adding hwrm_vnic_rss_cfg_* structures, which will be used in subsequent patches to enable RSS configuration. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:11 +00:00
#include "bnxt_rxq.h"
net/bnxt: add initial Rx code This patch adds initial implementation of rx_pkt_burst() function for Rx. bnxt_recv_pkts() is the top level function for doing Rx. This patch also adds code to allocate rings in the ASIC. For each Rx queue allocated in the PMD driver, a corresponding ring in hardware will be created. Every time a frame is received a Rx ring is selected based on the hardware configuration like RSS, MAC or VLAN, COS and such. The hardware uses a completion ring to indicate the availability of a packet. This patch also brings in functions like bnxt_init_one_rx_ring() bnxt_init_rx_ring_struct() which initializes various structures before a Rx can begin. bnxt_init_rxbds() initializes the Rx Buffer Descriptors while bnxt_alloc_rx_data() allocates a buffer in the host to receive the incoming packet. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:14 +00:00
#include "bnxt_rxr.h"
net/bnxt: add statistics Add the bnxt_stats_get_op() and bnxt_stats_reset_op() dev_ops to get and reset statistics. It also brings in the associated HWRM calls to handle the requests appropriately. We also have the bnxt_free_stats() function which will be used in the follow on patches to free the memory allocated by the driver for statistics. New HWRM calls: bnxt_hwrm_stat_clear: This command clears statistics of a context Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:12 +00:00
#include "bnxt_stats.h"
net/bnxt: add Tx queue create/destroy In this patch we are adding the bnxt_tx_queue_setup_op() and bnxt_tx_queue_release_op() functions. These will be tied to the tx_queue_setup and tx_queue_release dev_ops in a subsequent patch. In these functions we allocate/free memory for the TX queues. This still requires support to create a TX ring in the ASIC which will be completed in a future commit. Each Tx queue created via the tx_queue_setup dev_op will have an associated Tx ring in the hardware. A Tx request coming on the Tx queue gets sent to the corresponding Tx ring in the ASIC for subsequent transmission. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:10 +00:00
#include "bnxt_txq.h"
net/bnxt: add initial Tx code Initial implementation of tx_pkt_burst for transmit. bnxt_xmit_pkts() is the top level function that is called during Tx. bnxt_handle_tx_cp() is used to check and process the Tx completions generated for the Tx Buffer Descriptors sent by the hardware. This patch also adds code to allocate rings in the hardware. For each Tx queue allocated in the PMD driver, a corresponding ring in hardware will be created. Every time a Tx request is initiated via the bnxt_xmit_pkts() call, a Buffer Descriptor is created and is sent to the hardware via the associated Tx ring. On completing the Tx operation, the hardware will generates the status in the form of a completion. This completion is processed by the bnxt_handle_tx_cp() function. Functions like bnxt_init_tx_ring_struct() and bnxt_init_one_tx_ring() are used to initialize various members of the structure before starting Tx operations. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:13 +00:00
#include "bnxt_txr.h"
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
#include "bnxt_vnic.h"
#include "hsi_struct_def_dpdk.h"
net/bnxt: support get/set EEPROM Add support for get/set_eeprom, get_eeprom_length dev_ops. Defined the structures required to get/set the eeprom length/data in hsi_struct_defs hdr file along with implementation. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:44 +00:00
#include "bnxt_nvm_defs.h"
net/bnxt: get port and function info add helper functions to get port & function related information like parif, physical port id & vport id. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:53 +00:00
#include "bnxt_tf_common.h"
net/bnxt: add port default rules for ingress and egress ingress & egress port default rules are needed to send the packet from port_to_dpdk & dpdk_to_port respectively. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:28:34 +00:00
#include "ulp_flow_db.h"
net/bnxt: switch CFA code to dynamic mbuf field The CFA code from mark was stored in the deprecated mbuf field udata64. It is moved to a dynamic field in order to allow removal of udata64. Note: the new field has 32 bits, smaller than udata64. Signed-off-by: Thomas Monjalon <thomas@monjalon.net> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-10-25 22:41:16 +00:00
#include "rte_pmd_bnxt.h"
net/bnxt: add hardware resource manager init code Start adding support to use the HWRM API. Hardware Resource Manager or HWRM in short, is a set of API provided by the firmware running in the ASIC to manage the various resources. Initial commit just performs necessary HWRM queries for init, then fails as before. Now that struct bnxt is non-zero size, we can set dev_private_size correctly. The used HWRM calls so far: bnxt_hwrm_func_qcaps: This command returns capabilities of a function. bnxt_hwrm_ver_get: This function is called by a driver to determine the HWRM interface version supported by the HWRM firmware, the version of HWRM firmware implementation, the name of HWRM firmware, the versions of other embedded firmwares, and the names of other embedded firmwares, etc. Gets the firmware version and interface specifications. Returns an error if the firmware on the device is not supported by the driver and ensures the response space is large enough for the largest possible response. bnxt_hwrm_queue_qportcfg: This function is called by a driver to query queue configuration of a port. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:02 +00:00
net/bnxt: add driver for Broadcom NetXtreme-C devices This patch adds the initial skeleton for bnxt driver along with the nic guide, and ties the driver into the build system. At this point, the driver simply fails init. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com> [Release Note Addition] Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2016-06-15 21:23:01 +00:00
#define DRV_MODULE_NAME "bnxt"
static const char bnxt_version[] =
net/bnxt: do not double space version message The version message is double spaced in the log. Signed-off-by: Stephen Hemminger <sthemmin@microsoft.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-03-11 18:11:42 +00:00
"Broadcom NetXtreme driver " DRV_MODULE_NAME;
net/bnxt: add driver for Broadcom NetXtreme-C devices This patch adds the initial skeleton for bnxt driver along with the nic guide, and ties the driver into the build system. At this point, the driver simply fails init. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com> [Release Note Addition] Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2016-06-15 21:23:01 +00:00
net/bnxt: move macro definitions to header file Move device and vendor ids macro definition from bnxt_ethdev.c to bnxt.h. There is no functionality impact. Fixes: 882aa9c6e2e1 ("net/bnxt: move PCI device ids to the driver") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-10 09:44:32 +00:00
/*
* The set of PCI devices this driver supports
*/
drivers/net: make PCI device id struct const Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com> Acked-by: Shreyansh Jain <shreyansh.jain@nxp.com>
2017-01-26 14:07:44 +00:00
static const struct rte_pci_id bnxt_pci_id_map[] = {
net/bnxt: support Stratus VF device This patch adds support for Stratus VF devices. Other than adding the VF device ID, we also are adding support for short message format. The new short firmware message of size 16 bytes points to a location in host memory that contains the firmware message that the driver wants the adapter to process. This indirection requires the firmware to DMA the message into its own memory for processing. When the firmware receives a Short Command Format, it will DMA the firmware message from host memory into its internal memory and then processes the message as it would have if it received the message directly in the firmware communication channel. With BCM57454, the number of VFs increase to 1K and so the requirement for firmware communication channel memory on adapter becomes expensive. The short firmware message saves the amount of memory required to support 1K VFs on adapter. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-30 14:20:13 +00:00
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM,
net/bnxt: add device ID for Stratus VF Fixes: 1cd45aeb3270 ("net/bnxt: support Stratus VF device") Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
2018-04-17 01:11:26 +00:00
BROADCOM_DEV_ID_STRATUS_NIC_VF1) },
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM,
BROADCOM_DEV_ID_STRATUS_NIC_VF2) },
net/bnxt: add new device ids Add support for new device ids. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-04-26 19:20:02 +00:00
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_STRATUS_NIC) },
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_57414_VF) },
net/bnxt: move PCI device ids to the driver Moved defines since the driver had no such information. Used RTE_PCI_DEVICE in place of RTE_PCI_DEV_ID_DECL* stuff. Signed-off-by: David Marchand <david.marchand@6wind.com>
2016-07-11 14:40:44 +00:00
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_57304_VF) },
net/bnxt: support Broadcom StrataGX This patch adds support for the Broadcom StrataGX® BCM5871X series of Communications Processors. These ARM based processors target a broad range of networking applications including virtual CPE (vCPE) and NFV appliances, 10G service routers and gateways, control plane processing for Ethernet switches, and network attached storage (NAS). Other than adding the PCI Id for supporting the device, the patch also adds a memory barrier before the Tx doorbell and Completing ring doorbell is written to. Since ARM has a weakly ordered memory model this enforces a strict ordering of the descriptor writes before the doorbell writes happen. Signed-off-by: John Carney <john.carney@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2016-09-26 16:16:54 +00:00
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_NS2) },
net/bnxt: move PCI device ids to the driver Moved defines since the driver had no such information. Used RTE_PCI_DEVICE in place of RTE_PCI_DEV_ID_DECL* stuff. Signed-off-by: David Marchand <david.marchand@6wind.com>
2016-07-11 14:40:44 +00:00
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_57406_VF) },
net/bnxt: add new device IDs More PCI Device IDs for Cumulus, Cumulus+ and Whitney, Whitney+ SKUs. The NPAR model supported by firmware has been altered. It now allocates a unique Device ID for each NPAR partition for each device. In addition, ASIC's that are capable of supporting dual media have a unique DID depending whether they are configured in copper or fiber mode. This patch adds the necessary DIDs. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-09-26 16:18:30 +00:00
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_57407_MF) },
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_5741X_VF) },
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_5731X_VF) },
net/bnxt: support new PCI IDs Add support for PCI IDs which was removed as a part of commit 0e7bd0b2d6a5 ("net/bnxt: remove support for few devices") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-01-06 19:55:17 +00:00
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_57417_MF) },
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_57412) },
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_57414) },
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_57416_RJ45) },
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_57417_RJ45) },
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_57412_MF) },
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_57317_RJ45) },
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_57417_SFP) },
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_57416_SFP) },
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_57317_SFP) },
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_57414_MF) },
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_57416_MF) },
net/bnxt: introduce Broadcom Stingray Add support for Stingray PF devices ids: BCM58802 = 0xd802 BCM58804 = 0xd804 BCM58808 = 0x16f0 Add Broadcom Stingray machine support based on armv8. For product brief see: https://www.broadcom.com/products/ethernet-connectivity/controllers/bcm58800 Signed-off-by: Scott Branden <scott.branden@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-04-13 21:07:54 +00:00
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_58802) },
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_58804) },
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_58808) },
net/bnxt: support a StingRay VF id Add support for StingRay VF device 0xd800 Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-06-28 20:15:34 +00:00
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_58802_VF) },
net/bnxt: support thor controller This commit adds support to the bnxt PMD for devices based on the BCM57508 "thor" Ethernet controller. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
2019-06-02 17:42:44 +00:00
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_57508) },
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_57504) },
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_57502) },
net/bnxt: add new device id Added support for new thor device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-19 06:19:00 +00:00
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_57500_VF1) },
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_57500_VF2) },
net/bnxt: add PCI IDs for 57500 series NPAR devices Add PCI IDs for 57500 series NPAR devices Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-24 07:44:29 +00:00
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_57508_MF1) },
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_57504_MF1) },
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_57502_MF1) },
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_57508_MF2) },
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_57504_MF2) },
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_57502_MF2) },
net/bnxt: support 58818 chip family The new chip (Stingray 2) is part of the P5 chip family with a number of changes: 1. Implement the epoch doorbell bit for 58818 chip. With the new doorbell infrastructure and the unbounded index logic, now set the epoch doorbell bit to support proper doorbell operation on the new chip. Toggle epoch bit of all rings when it's wrapped to support doorbell overflow checking. 2. Get the legacy doorbell size from firmware. Legacy doorbell support has been removed in Stingray 2. So, the fast path doorbell pages start from the base of the BAR. Drivers need to use legacy_l2_db_space_size_kb field in the hwrm_func_qcfg_output response to get the legacy doorbell page offset from the BAR. 3. Set VALID doorbell bit on 58818 chip family. This class of chip has a valid doorbell bit added and it needs to be set. 4. Use "chip_num" returned by firmware. The "chip_num" field in the HWRM_VER_GET output returns the chip number. Use this value to identify chip category for 58818 chip family. 5. Added device ids for Stingray2 PF/VF devices. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-20 05:24:25 +00:00
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_58812) },
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_58814) },
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_58818) },
{ RTE_PCI_DEVICE(PCI_VENDOR_ID_BROADCOM, BROADCOM_DEV_ID_58818_VF) },
net/bnxt: move PCI device ids to the driver Moved defines since the driver had no such information. Used RTE_PCI_DEVICE in place of RTE_PCI_DEV_ID_DECL* stuff. Signed-off-by: David Marchand <david.marchand@6wind.com>
2016-07-11 14:40:44 +00:00
{ .vendor_id = 0, /* sentinel */ },
net/bnxt: add driver for Broadcom NetXtreme-C devices This patch adds the initial skeleton for bnxt driver along with the nic guide, and ties the driver into the build system. At this point, the driver simply fails init. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com> [Release Note Addition] Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2016-06-15 21:23:01 +00:00
};
net/bnxt: support host memory based TruFlow This feature can be enabled by passing "-w 0000:0d:00.0,host-based-truflow=1” to the DPDK application. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-04-15 08:18:52 +00:00
#define BNXT_DEVARG_TRUFLOW "host-based-truflow"
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
#define BNXT_DEVARG_FLOW_XSTAT "flow-xstat"
net/bnxt: set maximum flow count User could set max flow count by passing a devarg "-w 0000:0d:00.0,max_num_kflows=64" to a DPDK application; The value must be not less than 32K and be power-of-2; the default value is 32K. Signed-off-by: Shuanglin Wang <shuanglin.wang@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-06-12 12:50:12 +00:00
#define BNXT_DEVARG_MAX_NUM_KFLOWS "max-num-kflows"
net/bnxt: parse representors along with other devargs Representor dev-args need to be parsed during pci probe as they determine subsequent probe of VF representor ports as well. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:28:30 +00:00
#define BNXT_DEVARG_REPRESENTOR "representor"
net/bnxt: support representors on remote host domain In the Stingray use case, representors are conventionally run inside the SoC domain representing functions that are on the X86 domain. In order to support this mechanism of building representors for endpoints that are not in the same host domain, additional dev args have been in the PMD like so: rep-based-pf=<physical index> rep-is-pf=<VF=0 or PF=1> where `rep-based-pf` specifies the physical index of the base PF that the representor is derived off of. Since representor(s) can be created for endpoint PFs as well, rename struct bnxt_vf_representor to bnxt_representor and other such dev_ops and function names. devargs have also been extended to specify the exact CoS queue along with flow control enablement to be used for the conduit between the representor and the endpoint function. This is how a sample devargs would look with all the extended devargs -w 0000:06:02.0,host-based-truflow=1,representor=[1],rep-based-pf=8, rep-is-pf=1,rep-q-r2f=1,rep-fc-r2f=0,rep-q-f2r=1,rep-fc-f2r=1 Call CFA_PAIR_ALLOC only in case of Stingray instead of CFA_VFR_ALLOC. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:30 +00:00
#define BNXT_DEVARG_REP_BASED_PF "rep-based-pf"
#define BNXT_DEVARG_REP_IS_PF "rep-is-pf"
#define BNXT_DEVARG_REP_Q_R2F "rep-q-r2f"
#define BNXT_DEVARG_REP_Q_F2R "rep-q-f2r"
#define BNXT_DEVARG_REP_FC_R2F "rep-fc-r2f"
#define BNXT_DEVARG_REP_FC_F2R "rep-fc-f2r"
net/bnxt: initialize parent PF information Add support to query parent PF information (MAC address, function ID, port ID and default VNIC) from firmware. Current firmware returns zero for parent default vnic, a temporary Wh+-specific workaround is included until that can be fixed. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:51 +00:00
net/bnxt: support host memory based TruFlow This feature can be enabled by passing "-w 0000:0d:00.0,host-based-truflow=1” to the DPDK application. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-04-15 08:18:52 +00:00
static const char *const bnxt_dev_args[] = {
net/bnxt: parse representors along with other devargs Representor dev-args need to be parsed during pci probe as they determine subsequent probe of VF representor ports as well. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:28:30 +00:00
BNXT_DEVARG_REPRESENTOR,
net/bnxt: support host memory based TruFlow This feature can be enabled by passing "-w 0000:0d:00.0,host-based-truflow=1” to the DPDK application. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-04-15 08:18:52 +00:00
BNXT_DEVARG_TRUFLOW,
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
BNXT_DEVARG_FLOW_XSTAT,
net/bnxt: set maximum flow count User could set max flow count by passing a devarg "-w 0000:0d:00.0,max_num_kflows=64" to a DPDK application; The value must be not less than 32K and be power-of-2; the default value is 32K. Signed-off-by: Shuanglin Wang <shuanglin.wang@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-06-12 12:50:12 +00:00
BNXT_DEVARG_MAX_NUM_KFLOWS,
net/bnxt: support representors on remote host domain In the Stingray use case, representors are conventionally run inside the SoC domain representing functions that are on the X86 domain. In order to support this mechanism of building representors for endpoints that are not in the same host domain, additional dev args have been in the PMD like so: rep-based-pf=<physical index> rep-is-pf=<VF=0 or PF=1> where `rep-based-pf` specifies the physical index of the base PF that the representor is derived off of. Since representor(s) can be created for endpoint PFs as well, rename struct bnxt_vf_representor to bnxt_representor and other such dev_ops and function names. devargs have also been extended to specify the exact CoS queue along with flow control enablement to be used for the conduit between the representor and the endpoint function. This is how a sample devargs would look with all the extended devargs -w 0000:06:02.0,host-based-truflow=1,representor=[1],rep-based-pf=8, rep-is-pf=1,rep-q-r2f=1,rep-fc-r2f=0,rep-q-f2r=1,rep-fc-f2r=1 Call CFA_PAIR_ALLOC only in case of Stingray instead of CFA_VFR_ALLOC. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:30 +00:00
BNXT_DEVARG_REP_BASED_PF,
BNXT_DEVARG_REP_IS_PF,
BNXT_DEVARG_REP_Q_R2F,
BNXT_DEVARG_REP_Q_F2R,
BNXT_DEVARG_REP_FC_R2F,
BNXT_DEVARG_REP_FC_F2R,
net/bnxt: support host memory based TruFlow This feature can be enabled by passing "-w 0000:0d:00.0,host-based-truflow=1” to the DPDK application. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-04-15 08:18:52 +00:00
NULL
};
/*
* truflow == false to disable the feature
* truflow == true to enable the feature
*/
#define BNXT_DEVARG_TRUFLOW_INVALID(truflow) ((truflow) > 1)
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
/*
* flow_xstat == false to disable the feature
* flow_xstat == true to enable the feature
*/
#define BNXT_DEVARG_FLOW_XSTAT_INVALID(flow_xstat) ((flow_xstat) > 1)
net/bnxt: support representors on remote host domain In the Stingray use case, representors are conventionally run inside the SoC domain representing functions that are on the X86 domain. In order to support this mechanism of building representors for endpoints that are not in the same host domain, additional dev args have been in the PMD like so: rep-based-pf=<physical index> rep-is-pf=<VF=0 or PF=1> where `rep-based-pf` specifies the physical index of the base PF that the representor is derived off of. Since representor(s) can be created for endpoint PFs as well, rename struct bnxt_vf_representor to bnxt_representor and other such dev_ops and function names. devargs have also been extended to specify the exact CoS queue along with flow control enablement to be used for the conduit between the representor and the endpoint function. This is how a sample devargs would look with all the extended devargs -w 0000:06:02.0,host-based-truflow=1,representor=[1],rep-based-pf=8, rep-is-pf=1,rep-q-r2f=1,rep-fc-r2f=0,rep-q-f2r=1,rep-fc-f2r=1 Call CFA_PAIR_ALLOC only in case of Stingray instead of CFA_VFR_ALLOC. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:30 +00:00
/*
* rep_is_pf == false to indicate VF representor
* rep_is_pf == true to indicate PF representor
*/
#define BNXT_DEVARG_REP_IS_PF_INVALID(rep_is_pf) ((rep_is_pf) > 1)
/*
* rep_based_pf == Physical index of the PF
*/
#define BNXT_DEVARG_REP_BASED_PF_INVALID(rep_based_pf) ((rep_based_pf) > 15)
/*
* rep_q_r2f == Logical COS Queue index for the rep to endpoint direction
*/
#define BNXT_DEVARG_REP_Q_R2F_INVALID(rep_q_r2f) ((rep_q_r2f) > 3)
/*
* rep_q_f2r == Logical COS Queue index for the endpoint to rep direction
*/
#define BNXT_DEVARG_REP_Q_F2R_INVALID(rep_q_f2r) ((rep_q_f2r) > 3)
/*
* rep_fc_r2f == Flow control for the representor to endpoint direction
*/
#define BNXT_DEVARG_REP_FC_R2F_INVALID(rep_fc_r2f) ((rep_fc_r2f) > 1)
/*
* rep_fc_f2r == Flow control for the endpoint to representor direction
*/
#define BNXT_DEVARG_REP_FC_F2R_INVALID(rep_fc_f2r) ((rep_fc_f2r) > 1)
net/bnxt: switch CFA code to dynamic mbuf field The CFA code from mark was stored in the deprecated mbuf field udata64. It is moved to a dynamic field in order to allow removal of udata64. Note: the new field has 32 bits, smaller than udata64. Signed-off-by: Thomas Monjalon <thomas@monjalon.net> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-10-25 22:41:16 +00:00
int bnxt_cfa_code_dynfield_offset = -1;
net/bnxt: set maximum flow count User could set max flow count by passing a devarg "-w 0000:0d:00.0,max_num_kflows=64" to a DPDK application; The value must be not less than 32K and be power-of-2; the default value is 32K. Signed-off-by: Shuanglin Wang <shuanglin.wang@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-06-12 12:50:12 +00:00
/*
* max_num_kflows must be >= 32
* and must be a power-of-2 supported value
* return: 1 -> invalid
* 0 -> valid
*/
static int bnxt_devarg_max_num_kflow_invalid(uint16_t max_num_kflows)
{
if (max_num_kflows < 32 || !rte_is_power_of_2(max_num_kflows))
return 1;
return 0;
}
ethdev: allow returning error on VLAN offload ops Some devices may not support or fail setting VLAN offload configuration based on dynamic circumstances so the vlan_offload_set_t vector is modified to return an int so the caller can determine success or not. rte_eth_dev_set_vlan_offload is updated to return the value provided by the vector when called along with restoring the original offload configs on failure. Existing vlan_offload_set_t vectors are modified to return an int. Majority of cases return 0 but a few that actually can fail now return their failure codes. Finally, a vlan_offload_set_t vector is added to virtio to facilitate dynamically turning VLAN strip on or off. Signed-off-by: David Harton <dharton@cisco.com> Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
2017-09-01 02:36:28 +00:00
static int bnxt_vlan_offload_set_op(struct rte_eth_dev *dev, int mask);
net/bnxt: fix close operation We are not cleaning up all the memory and also not unregistering the driver during device close operation. This patch fixes the issue. Fixes: 893074951314 ("net/bnxt: free memory in close operation") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-06-28 20:15:31 +00:00
static int bnxt_dev_uninit(struct rte_eth_dev *eth_dev);
net/bnxt: handle reset notify async event from FW When the FW upgrade is initiated the current instance of FW issues a HWRM_ASYNC_EVENT_CMPL_EVENT_ID_RESET_NOTIFY async notification to the driver. On receiving this notification, the PMD shall quiesce itself and poll on the HWRM_VER_GET FW command at regular intervals. Once the VER_GET command succeeds, the driver should go through the rediscovery process and re-initialize the device. Also register with FW for the reset notify async event. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:23 +00:00
static int bnxt_init_resources(struct bnxt *bp, bool reconfig_dev);
static int bnxt_uninit_resources(struct bnxt *bp, bool reconfig_dev);
net/bnxt: support periodic FW health monitoring Periodically poll the FW heartbeat register and FW recovery counter registers to check the FW health. Polling frequency will be advertised by the FW in HWRM_ERROR_RECOVERY_QCFG response. Schedule the task upon receiving the async event from FW. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:29 +00:00
static void bnxt_cancel_fw_health_check(struct bnxt *bp);
net/bnxt: fix VLAN strip HWRM_VNIC_CFG command to configure VNIC dynamically with traffic running is not working. Driver has to free and recreate the VNIC and then reconfigure the VNIC filters. Fixes: 7fe5668d2ea3 ("net/bnxt: support VLAN filter and strip") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-01-16 13:04:50 +00:00
static int bnxt_restore_vlan_filters(struct bnxt *bp);
net/bnxt: fix recovery alarm race condition in port close There is a race condition when port is closed while error recovery is happening in driver. Fixed this by canceling the recovery threads during port close. Fixes: df6cd7c1f73a ("net/bnxt: handle reset notify async event from FW") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com>
2020-02-28 10:21:31 +00:00
static void bnxt_dev_recover(void *arg);
net/bnxt: log firmware status on early init failure Firmware reports any fatal error (either ASIC or Firmware related) via a new status register. This status register can provide more detailed information about the firmware errors, especially if error occurs before HWRM_VER_GET is issued. Attempt to map this register if it is present and check for firmware status when VER_GET command fails. Refactored the code to allocate the "bp->recovery_info" structure in bnxt_init_fw() instead of doing in bnxt_hwrm_error_recovery_qcfg(). Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-04-21 21:33:50 +00:00
static void bnxt_free_error_recovery_info(struct bnxt *bp);
net/bnxt: support port representor data path Added code to support Tx/Rx from a VF representor port. The VF-reps use the RX/TX rings of the Trusted VF/PF. For each VF-rep, the Trusted VF/PF driver issues a VFR_ALLOC FW cmd that returns "cfa_code" and "cfa_action" values. The FW sets up the filter tables in such a way that VF traffic by default (in absence of other rules) gets punted to the parent function i.e. either the Trusted VF or the PF. The cfa_code value in the RX-compl informs the driver of the source VF. For traffic being transmitted from the VF-rep, the TX BD is tagged with a cfa_action value that informs the HW to punt it to the corresponding VF. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:49 +00:00
static void bnxt_free_rep_info(struct bnxt *bp);
net/bnxt: support Tx loopback, set VF MAC and queues drop Add functions rte_pmd_bnxt_set_tx_loopback, rte_pmd_bnxt_set_all_queues_drop_en and rte_pmd_bnxt_set_vf_mac_addr to configure tx_loopback, queue_drop and VF MAC address setting in the hardware. It also adds the necessary functions to send the HWRM commands to the firmware. Signed-off-by: Steeven Li <steeven.li@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:16 +00:00
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
int is_bnxt_in_error(struct bnxt *bp)
{
if (bp->flags & BNXT_FLAG_FATAL_ERROR)
return -EIO;
if (bp->flags & BNXT_FLAG_FW_RESET)
return -EBUSY;
return 0;
}
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
/***********************/
/*
* High level utility functions
*/
net/bnxt: make locally used functions static bnxt_rss_ctxts() function is declared in header file bnxt.h, implemented in bnxt_ethdev.c, and called only in bnxt_ethdev.c. Also many functions are declared in header file bnxt_hwrm.h, implemented in bnxt_hwrm.c, and called only in bnxt_hwrm.c. This patch moves these function declarations into bnxt_ethdev.c and bnxt_hwrm.c, as static functions. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-02-06 16:33:13 +00:00
static uint16_t bnxt_rss_ctxts(const struct bnxt *bp)
net/bnxt: enable RSS for thor-based controllers Make changes needed to support rss for thor-based controllers. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-06-02 17:42:47 +00:00
{
net/bnxt: fix Rx rings in RSS redirection table This commit introduces a limit on the number of RX rings included in the RSS redirection table to a value no larger than the size supported by Thor as defined by BNXT_RSS_TBL_SIZE_THOR. Fixes: d819382543f3 ("net/bnxt: add RSS redirection table operations") Cc: stable@dpdk.org Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Samik Gupta <samik.gupta@broadcom.com>
2020-11-12 21:28:25 +00:00
unsigned int num_rss_rings = RTE_MIN(bp->rx_nr_rings,
net/bnxt: remove references to Thor Refactor code to remove references to Thor. Instead use P5 as in phase 5 of development cycle since it is applicable to boards other than Thor as well. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2020-12-01 19:15:23 +00:00
BNXT_RSS_TBL_SIZE_P5);
net/bnxt: fix Rx rings in RSS redirection table This commit introduces a limit on the number of RX rings included in the RSS redirection table to a value no larger than the size supported by Thor as defined by BNXT_RSS_TBL_SIZE_THOR. Fixes: d819382543f3 ("net/bnxt: add RSS redirection table operations") Cc: stable@dpdk.org Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Samik Gupta <samik.gupta@broadcom.com>
2020-11-12 21:28:25 +00:00
net/bnxt: remove references to Thor Refactor code to remove references to Thor. Instead use P5 as in phase 5 of development cycle since it is applicable to boards other than Thor as well. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2020-12-01 19:15:23 +00:00
if (!BNXT_CHIP_P5(bp))
net/bnxt: enable RSS for thor-based controllers Make changes needed to support rss for thor-based controllers. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-06-02 17:42:47 +00:00
return 1;
net/bnxt: fix Rx rings in RSS redirection table This commit introduces a limit on the number of RX rings included in the RSS redirection table to a value no larger than the size supported by Thor as defined by BNXT_RSS_TBL_SIZE_THOR. Fixes: d819382543f3 ("net/bnxt: add RSS redirection table operations") Cc: stable@dpdk.org Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Samik Gupta <samik.gupta@broadcom.com>
2020-11-12 21:28:25 +00:00
return RTE_ALIGN_MUL_CEIL(num_rss_rings,
net/bnxt: remove references to Thor Refactor code to remove references to Thor. Instead use P5 as in phase 5 of development cycle since it is applicable to boards other than Thor as well. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2020-12-01 19:15:23 +00:00
BNXT_RSS_ENTRIES_PER_CTX_P5) /
BNXT_RSS_ENTRIES_PER_CTX_P5;
net/bnxt: enable RSS for thor-based controllers Make changes needed to support rss for thor-based controllers. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-06-02 17:42:47 +00:00
}
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
uint16_t bnxt_rss_hash_tbl_size(const struct bnxt *bp)
net/bnxt: enable RSS for thor-based controllers Make changes needed to support rss for thor-based controllers. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-06-02 17:42:47 +00:00
{
net/bnxt: remove references to Thor Refactor code to remove references to Thor. Instead use P5 as in phase 5 of development cycle since it is applicable to boards other than Thor as well. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2020-12-01 19:15:23 +00:00
if (!BNXT_CHIP_P5(bp))
net/bnxt: enable RSS for thor-based controllers Make changes needed to support rss for thor-based controllers. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-06-02 17:42:47 +00:00
return HW_HASH_INDEX_SIZE;
net/bnxt: remove references to Thor Refactor code to remove references to Thor. Instead use P5 as in phase 5 of development cycle since it is applicable to boards other than Thor as well. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2020-12-01 19:15:23 +00:00
return bnxt_rss_ctxts(bp) * BNXT_RSS_ENTRIES_PER_CTX_P5;
net/bnxt: enable RSS for thor-based controllers Make changes needed to support rss for thor-based controllers. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-06-02 17:42:47 +00:00
}
net/bnxt: initialize parent PF information Add support to query parent PF information (MAC address, function ID, port ID and default VNIC) from firmware. Current firmware returns zero for parent default vnic, a temporary Wh+-specific workaround is included until that can be fixed. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:51 +00:00
static void bnxt_free_parent_info(struct bnxt *bp)
{
rte_free(bp->parent);
net/bnxt: fix resource cleanup Fix resource cleanup in port close. Once the pointers are freed, set them to NULL. Make sure access to the pointers is validated before use. Fixes: bb81e07323bb ("net/bnxt: support LED on/off") Fixes: 804e746c7b73 ("net/bnxt: add hardware resource manager init code") Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Fixes: 698aa7e95325 ("net/bnxt: add code to determine the Tx COS queue") Fixes: 322bd6e70272 ("net/bnxt: add port representor infrastructure") Fixes: 0bf5a0b5ebb8 ("net/bnxt: add a failure log") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-04-08 20:17:47 +00:00
bp->parent = NULL;
net/bnxt: initialize parent PF information Add support to query parent PF information (MAC address, function ID, port ID and default VNIC) from firmware. Current firmware returns zero for parent default vnic, a temporary Wh+-specific workaround is included until that can be fixed. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:51 +00:00
}
net/bnxt: fix allocation of PF info struct Fixes: 804e746c7b73 ("net/bnxt: add hardware resource manager init code") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 11:10:41 +00:00
static void bnxt_free_pf_info(struct bnxt *bp)
{
rte_free(bp->pf);
net/bnxt: fix resource cleanup Fix resource cleanup in port close. Once the pointers are freed, set them to NULL. Make sure access to the pointers is validated before use. Fixes: bb81e07323bb ("net/bnxt: support LED on/off") Fixes: 804e746c7b73 ("net/bnxt: add hardware resource manager init code") Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Fixes: 698aa7e95325 ("net/bnxt: add code to determine the Tx COS queue") Fixes: 322bd6e70272 ("net/bnxt: add port representor infrastructure") Fixes: 0bf5a0b5ebb8 ("net/bnxt: add a failure log") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-04-08 20:17:47 +00:00
bp->pf = NULL;
net/bnxt: fix allocation of PF info struct Fixes: 804e746c7b73 ("net/bnxt: add hardware resource manager init code") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 11:10:41 +00:00
}
net/bnxt: fix allocation of link info struct Dynamically alloc link info structure. Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 11:05:15 +00:00
static void bnxt_free_link_info(struct bnxt *bp)
{
rte_free(bp->link_info);
net/bnxt: fix resource cleanup Fix resource cleanup in port close. Once the pointers are freed, set them to NULL. Make sure access to the pointers is validated before use. Fixes: bb81e07323bb ("net/bnxt: support LED on/off") Fixes: 804e746c7b73 ("net/bnxt: add hardware resource manager init code") Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Fixes: 698aa7e95325 ("net/bnxt: add code to determine the Tx COS queue") Fixes: 322bd6e70272 ("net/bnxt: add port representor infrastructure") Fixes: 0bf5a0b5ebb8 ("net/bnxt: add a failure log") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-04-08 20:17:47 +00:00
bp->link_info = NULL;
net/bnxt: fix allocation of link info struct Dynamically alloc link info structure. Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 11:05:15 +00:00
}
net/bnxt: fix allocation of LED config info Dynamically allocate bnxt_led_cfg. This helps reduces memory footprint of struct bnxt. Fixes: bb81e07323bb ("net/bnxt: support LED on/off") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 07:01:51 +00:00
static void bnxt_free_leds_info(struct bnxt *bp)
{
net/bnxt: fix unnecessary memory allocation VFs are not privileged to issue HWRM_PORT_LED_QCFG/CFG. There is no need to allocate "bp->leds" memory. Fixes: 205b74295282 ("net/bnxt: fix allocation of LED config info") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-09 09:38:28 +00:00
if (BNXT_VF(bp))
return;
net/bnxt: fix allocation of LED config info Dynamically allocate bnxt_led_cfg. This helps reduces memory footprint of struct bnxt. Fixes: bb81e07323bb ("net/bnxt: support LED on/off") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 07:01:51 +00:00
rte_free(bp->leds);
bp->leds = NULL;
}
net/bnxt: fix allocation of flow stat related structs Consolidate flow stat related structs for performance improvement. The intention of this patch is to reduce the size struct bnxt which had grown because of recent changes and was impacting performance. Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 08:31:59 +00:00
static void bnxt_free_flow_stats_info(struct bnxt *bp)
{
rte_free(bp->flow_stat);
bp->flow_stat = NULL;
}
net/bnxt: fix allocation of COS queue info Fixes: 698aa7e95325 ("net/bnxt: add code to determine the Tx COS queue") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 18:07:32 +00:00
static void bnxt_free_cos_queues(struct bnxt *bp)
{
rte_free(bp->rx_cos_queue);
net/bnxt: fix resource cleanup Fix resource cleanup in port close. Once the pointers are freed, set them to NULL. Make sure access to the pointers is validated before use. Fixes: bb81e07323bb ("net/bnxt: support LED on/off") Fixes: 804e746c7b73 ("net/bnxt: add hardware resource manager init code") Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Fixes: 698aa7e95325 ("net/bnxt: add code to determine the Tx COS queue") Fixes: 322bd6e70272 ("net/bnxt: add port representor infrastructure") Fixes: 0bf5a0b5ebb8 ("net/bnxt: add a failure log") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-04-08 20:17:47 +00:00
bp->rx_cos_queue = NULL;
net/bnxt: fix allocation of COS queue info Fixes: 698aa7e95325 ("net/bnxt: add code to determine the Tx COS queue") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 18:07:32 +00:00
rte_free(bp->tx_cos_queue);
net/bnxt: fix resource cleanup Fix resource cleanup in port close. Once the pointers are freed, set them to NULL. Make sure access to the pointers is validated before use. Fixes: bb81e07323bb ("net/bnxt: support LED on/off") Fixes: 804e746c7b73 ("net/bnxt: add hardware resource manager init code") Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Fixes: 698aa7e95325 ("net/bnxt: add code to determine the Tx COS queue") Fixes: 322bd6e70272 ("net/bnxt: add port representor infrastructure") Fixes: 0bf5a0b5ebb8 ("net/bnxt: add a failure log") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-04-08 20:17:47 +00:00
bp->tx_cos_queue = NULL;
net/bnxt: fix allocation of COS queue info Fixes: 698aa7e95325 ("net/bnxt: add code to determine the Tx COS queue") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 18:07:32 +00:00
}
net/bnxt: handle reset notify async event from FW When the FW upgrade is initiated the current instance of FW issues a HWRM_ASYNC_EVENT_CMPL_EVENT_ID_RESET_NOTIFY async notification to the driver. On receiving this notification, the PMD shall quiesce itself and poll on the HWRM_VER_GET FW command at regular intervals. Once the VER_GET command succeeds, the driver should go through the rediscovery process and re-initialize the device. Also register with FW for the reset notify async event. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:23 +00:00
static void bnxt_free_mem(struct bnxt *bp, bool reconfig)
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
{
bnxt_free_filter_mem(bp);
bnxt_free_vnic_attributes(bp);
bnxt_free_vnic_mem(bp);
net/bnxt: handle reset notify async event from FW When the FW upgrade is initiated the current instance of FW issues a HWRM_ASYNC_EVENT_CMPL_EVENT_ID_RESET_NOTIFY async notification to the driver. On receiving this notification, the PMD shall quiesce itself and poll on the HWRM_VER_GET FW command at regular intervals. Once the VER_GET command succeeds, the driver should go through the rediscovery process and re-initialize the device. Also register with FW for the reset notify async event. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:23 +00:00
/* tx/rx rings are configured as part of *_queue_setup callbacks.
* If the number of rings change across fw update,
* we don't have much choice except to warn the user.
*/
if (!reconfig) {
bnxt_free_stats(bp);
bnxt_free_tx_rings(bp);
bnxt_free_rx_rings(bp);
}
net/bnxt: use dedicated CPR for async events This commit enables the creation of a dedicated completion ring for asynchronous event handling instead of handling these events on a receive completion ring. For the stingray platform and other platforms needing tighter control of resource utilization, we retain the ability to process async events on a receive completion ring. For Thor-based adapters, we use a dedicated NQ (notification queue) ring for async events (async events can't currently be received on a completion ring due to a firmware limitation). Rename "def_cp_ring" to "async_cp_ring" to better reflect its purpose (async event notifications) and to avoid confusion with VNIC default receive completion rings. Allow rxq 0 to be stopped when not being used for async events. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-24 16:49:32 +00:00
bnxt_free_async_cp_ring(bp);
net/bnxt: use common NQ ring Thor queue scaling is currently limited by the number of NQs that can be allocated. Fix by using a common NQ for all receive/transmit rings instead of allocating a separate NQ for each ring. Fixes: f8168ca0e690 ("net/bnxt: support thor controller") Cc: stable@dpdk.org Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
2019-10-04 03:48:59 +00:00
bnxt_free_rxtx_nq_ring(bp);
net/bnxt: fix memory leak During hot fw upgrade or error recovery, driver does cleanup and reallocation of resources. But driver reallocates memory for ring group info table without freeing the allocated memory during device init. Fix this memory leak by moving the freeing of ring group info table to bnxt_free_mem() in bnxt_uninit_resources(). Fixes: c09f57b49c13 ("net/bnxt: add start/stop/link update operations") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com>
2019-10-24 07:44:31 +00:00
rte_free(bp->grp_info);
bp->grp_info = NULL;
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
}
net/bnxt: initialize parent PF information Add support to query parent PF information (MAC address, function ID, port ID and default VNIC) from firmware. Current firmware returns zero for parent default vnic, a temporary Wh+-specific workaround is included until that can be fixed. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:51 +00:00
static int bnxt_alloc_parent_info(struct bnxt *bp)
{
bp->parent = rte_zmalloc("bnxt_parent_info",
sizeof(struct bnxt_parent_info), 0);
if (bp->parent == NULL)
return -ENOMEM;
return 0;
}
net/bnxt: fix allocation of PF info struct Fixes: 804e746c7b73 ("net/bnxt: add hardware resource manager init code") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 11:10:41 +00:00
static int bnxt_alloc_pf_info(struct bnxt *bp)
{
bp->pf = rte_zmalloc("bnxt_pf_info", sizeof(struct bnxt_pf_info), 0);
if (bp->pf == NULL)
return -ENOMEM;
return 0;
}
net/bnxt: fix allocation of link info struct Dynamically alloc link info structure. Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 11:05:15 +00:00
static int bnxt_alloc_link_info(struct bnxt *bp)
{
bp->link_info =
rte_zmalloc("bnxt_link_info", sizeof(struct bnxt_link_info), 0);
if (bp->link_info == NULL)
return -ENOMEM;
return 0;
}
net/bnxt: fix allocation of LED config info Dynamically allocate bnxt_led_cfg. This helps reduces memory footprint of struct bnxt. Fixes: bb81e07323bb ("net/bnxt: support LED on/off") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 07:01:51 +00:00
static int bnxt_alloc_leds_info(struct bnxt *bp)
{
net/bnxt: fix unnecessary memory allocation VFs are not privileged to issue HWRM_PORT_LED_QCFG/CFG. There is no need to allocate "bp->leds" memory. Fixes: 205b74295282 ("net/bnxt: fix allocation of LED config info") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-09 09:38:28 +00:00
if (BNXT_VF(bp))
return 0;
net/bnxt: fix allocation of LED config info Dynamically allocate bnxt_led_cfg. This helps reduces memory footprint of struct bnxt. Fixes: bb81e07323bb ("net/bnxt: support LED on/off") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 07:01:51 +00:00
bp->leds = rte_zmalloc("bnxt_leds",
BNXT_MAX_LED * sizeof(struct bnxt_led_info),
0);
if (bp->leds == NULL)
return -ENOMEM;
return 0;
}
net/bnxt: fix allocation of COS queue info Fixes: 698aa7e95325 ("net/bnxt: add code to determine the Tx COS queue") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 18:07:32 +00:00
static int bnxt_alloc_cos_queues(struct bnxt *bp)
{
bp->rx_cos_queue =
rte_zmalloc("bnxt_rx_cosq",
BNXT_COS_QUEUE_COUNT *
sizeof(struct bnxt_cos_queue_info),
0);
if (bp->rx_cos_queue == NULL)
return -ENOMEM;
bp->tx_cos_queue =
rte_zmalloc("bnxt_tx_cosq",
BNXT_COS_QUEUE_COUNT *
sizeof(struct bnxt_cos_queue_info),
0);
if (bp->tx_cos_queue == NULL)
return -ENOMEM;
return 0;
}
net/bnxt: fix allocation of flow stat related structs Consolidate flow stat related structs for performance improvement. The intention of this patch is to reduce the size struct bnxt which had grown because of recent changes and was impacting performance. Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 08:31:59 +00:00
static int bnxt_alloc_flow_stats_info(struct bnxt *bp)
{
bp->flow_stat = rte_zmalloc("bnxt_flow_xstat",
sizeof(struct bnxt_flow_stat_info), 0);
if (bp->flow_stat == NULL)
return -ENOMEM;
return 0;
}
net/bnxt: handle reset notify async event from FW When the FW upgrade is initiated the current instance of FW issues a HWRM_ASYNC_EVENT_CMPL_EVENT_ID_RESET_NOTIFY async notification to the driver. On receiving this notification, the PMD shall quiesce itself and poll on the HWRM_VER_GET FW command at regular intervals. Once the VER_GET command succeeds, the driver should go through the rediscovery process and re-initialize the device. Also register with FW for the reset notify async event. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:23 +00:00
static int bnxt_alloc_mem(struct bnxt *bp, bool reconfig)
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
{
int rc;
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
rc = bnxt_alloc_ring_grps(bp);
if (rc)
goto alloc_mem_err;
net/bnxt: use dedicated CPR for async events This commit enables the creation of a dedicated completion ring for asynchronous event handling instead of handling these events on a receive completion ring. For the stingray platform and other platforms needing tighter control of resource utilization, we retain the ability to process async events on a receive completion ring. For Thor-based adapters, we use a dedicated NQ (notification queue) ring for async events (async events can't currently be received on a completion ring due to a firmware limitation). Rename "def_cp_ring" to "async_cp_ring" to better reflect its purpose (async event notifications) and to avoid confusion with VNIC default receive completion rings. Allow rxq 0 to be stopped when not being used for async events. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-24 16:49:32 +00:00
rc = bnxt_alloc_async_ring_struct(bp);
if (rc)
goto alloc_mem_err;
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
rc = bnxt_alloc_vnic_mem(bp);
if (rc)
goto alloc_mem_err;
rc = bnxt_alloc_vnic_attributes(bp);
if (rc)
goto alloc_mem_err;
rc = bnxt_alloc_filter_mem(bp);
if (rc)
goto alloc_mem_err;
net/bnxt: use dedicated CPR for async events This commit enables the creation of a dedicated completion ring for asynchronous event handling instead of handling these events on a receive completion ring. For the stingray platform and other platforms needing tighter control of resource utilization, we retain the ability to process async events on a receive completion ring. For Thor-based adapters, we use a dedicated NQ (notification queue) ring for async events (async events can't currently be received on a completion ring due to a firmware limitation). Rename "def_cp_ring" to "async_cp_ring" to better reflect its purpose (async event notifications) and to avoid confusion with VNIC default receive completion rings. Allow rxq 0 to be stopped when not being used for async events. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-24 16:49:32 +00:00
rc = bnxt_alloc_async_cp_ring(bp);
if (rc)
goto alloc_mem_err;
net/bnxt: use common NQ ring Thor queue scaling is currently limited by the number of NQs that can be allocated. Fix by using a common NQ for all receive/transmit rings instead of allocating a separate NQ for each ring. Fixes: f8168ca0e690 ("net/bnxt: support thor controller") Cc: stable@dpdk.org Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
2019-10-04 03:48:59 +00:00
rc = bnxt_alloc_rxtx_nq_ring(bp);
if (rc)
goto alloc_mem_err;
net/bnxt: fix allocation of flow stat related structs Consolidate flow stat related structs for performance improvement. The intention of this patch is to reduce the size struct bnxt which had grown because of recent changes and was impacting performance. Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 08:31:59 +00:00
if (BNXT_FLOW_XSTATS_EN(bp)) {
rc = bnxt_alloc_flow_stats_info(bp);
if (rc)
goto alloc_mem_err;
}
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
return 0;
alloc_mem_err:
net/bnxt: handle reset notify async event from FW When the FW upgrade is initiated the current instance of FW issues a HWRM_ASYNC_EVENT_CMPL_EVENT_ID_RESET_NOTIFY async notification to the driver. On receiving this notification, the PMD shall quiesce itself and poll on the HWRM_VER_GET FW command at regular intervals. Once the VER_GET command succeeds, the driver should go through the rediscovery process and re-initialize the device. Also register with FW for the reset notify async event. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:23 +00:00
bnxt_free_mem(bp, reconfig);
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
return rc;
}
net/bnxt: fix VLAN strip HWRM_VNIC_CFG command to configure VNIC dynamically with traffic running is not working. Driver has to free and recreate the VNIC and then reconfigure the VNIC filters. Fixes: 7fe5668d2ea3 ("net/bnxt: support VLAN filter and strip") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-01-16 13:04:50 +00:00
static int bnxt_setup_one_vnic(struct bnxt *bp, uint16_t vnic_id)
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
{
net/bnxt: fix VLAN strip HWRM_VNIC_CFG command to configure VNIC dynamically with traffic running is not working. Driver has to free and recreate the VNIC and then reconfigure the VNIC filters. Fixes: 7fe5668d2ea3 ("net/bnxt: support VLAN filter and strip") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-01-16 13:04:50 +00:00
struct rte_eth_conf *dev_conf = &bp->eth_dev->data->dev_conf;
struct bnxt_vnic_info *vnic = &bp->vnic_info[vnic_id];
uint64_t rx_offloads = dev_conf->rxmode.offloads;
net/bnxt: fix queue start/stop operations Packets destined to the to-be-stopped queue should not be dropped (neither in HW nor in the driver), so re-program the RSS Table without this queue on stop and add it back to the table on start unless it is a Representor VF. Since 0th entry is used for default ring, use fw_grp_id + 1 to change the RSS table population logic by programming valid IDs instead of the default zeroth entry in case of an invalid fw_grp_id. Destroy and recreate the trio of Rx rings(compl, Rx, AG) every time in start so that HW is in sync with software. Fixes: 9b63c6fd70e3 ("net/bnxt: support Rx/Tx queue start/stop") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ray Jui <ray.jui@broadcom.com> Reviewed-by: Scott Branden <scott.branden@broadcom.com> Reviewed-by: Randy Schacher <stuart.schacher@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-06-28 20:15:35 +00:00
struct bnxt_rx_queue *rxq;
net/bnxt: fix VLAN strip HWRM_VNIC_CFG command to configure VNIC dynamically with traffic running is not working. Driver has to free and recreate the VNIC and then reconfigure the VNIC filters. Fixes: 7fe5668d2ea3 ("net/bnxt: support VLAN filter and strip") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-01-16 13:04:50 +00:00
unsigned int j;
int rc;
rc = bnxt_vnic_grp_alloc(bp, vnic);
if (rc)
goto err_out;
PMD_DRV_LOG(DEBUG, "vnic[%d] = %p vnic->fw_grp_ids = %p\n",
vnic_id, vnic, vnic->fw_grp_ids);
rc = bnxt_hwrm_vnic_alloc(bp, vnic);
if (rc)
goto err_out;
/* Alloc RSS context only if RSS mode is enabled */
if (dev_conf->rxmode.mq_mode & ETH_MQ_RX_RSS) {
int j, nr_ctxs = bnxt_rss_ctxts(bp);
net/bnxt: cap max Rx rings for Thor For Thor devices, RSS table can only accommodate 512 Rx queues. When RSS is enabled, Cap the max Rx rings to 512. For non-RSS case, the number will be limited by number of VNICs. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com>
2021-04-14 23:44:49 +00:00
/* RSS table size in Thor is 512.
* Cap max Rx rings to same value
*/
net/bnxt: remove references to Thor Refactor code to remove references to Thor. Instead use P5 as in phase 5 of development cycle since it is applicable to boards other than Thor as well. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2020-12-01 19:15:23 +00:00
if (bp->rx_nr_rings > BNXT_RSS_TBL_SIZE_P5) {
net/bnxt: fix Rx rings in RSS redirection table This commit introduces a limit on the number of RX rings included in the RSS redirection table to a value no larger than the size supported by Thor as defined by BNXT_RSS_TBL_SIZE_THOR. Fixes: d819382543f3 ("net/bnxt: add RSS redirection table operations") Cc: stable@dpdk.org Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Samik Gupta <samik.gupta@broadcom.com>
2020-11-12 21:28:25 +00:00
PMD_DRV_LOG(ERR, "RxQ cnt %d > reta_size %d\n",
net/bnxt: remove references to Thor Refactor code to remove references to Thor. Instead use P5 as in phase 5 of development cycle since it is applicable to boards other than Thor as well. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2020-12-01 19:15:23 +00:00
bp->rx_nr_rings, BNXT_RSS_TBL_SIZE_P5);
net/bnxt: cap max Rx rings for Thor For Thor devices, RSS table can only accommodate 512 Rx queues. When RSS is enabled, Cap the max Rx rings to 512. For non-RSS case, the number will be limited by number of VNICs. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com>
2021-04-14 23:44:49 +00:00
goto err_out;
net/bnxt: fix Rx rings in RSS redirection table This commit introduces a limit on the number of RX rings included in the RSS redirection table to a value no larger than the size supported by Thor as defined by BNXT_RSS_TBL_SIZE_THOR. Fixes: d819382543f3 ("net/bnxt: add RSS redirection table operations") Cc: stable@dpdk.org Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Samik Gupta <samik.gupta@broadcom.com>
2020-11-12 21:28:25 +00:00
}
net/bnxt: fix VLAN strip HWRM_VNIC_CFG command to configure VNIC dynamically with traffic running is not working. Driver has to free and recreate the VNIC and then reconfigure the VNIC filters. Fixes: 7fe5668d2ea3 ("net/bnxt: support VLAN filter and strip") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-01-16 13:04:50 +00:00
rc = 0;
for (j = 0; j < nr_ctxs; j++) {
rc = bnxt_hwrm_vnic_ctx_alloc(bp, vnic, j);
if (rc)
break;
}
if (rc) {
PMD_DRV_LOG(ERR,
"HWRM vnic %d ctx %d alloc failure rc: %x\n",
vnic_id, j, rc);
goto err_out;
}
vnic->num_lb_ctxts = nr_ctxs;
}
/*
* Firmware sets pf pair in default vnic cfg. If the VLAN strip
* setting is not available at this time, it will not be
* configured correctly in the CFA.
*/
if (rx_offloads & DEV_RX_OFFLOAD_VLAN_STRIP)
vnic->vlan_strip = true;
else
vnic->vlan_strip = false;
rc = bnxt_hwrm_vnic_cfg(bp, vnic);
if (rc)
goto err_out;
rc = bnxt_set_hwrm_vnic_filters(bp, vnic);
if (rc)
goto err_out;
for (j = 0; j < bp->rx_num_qs_per_vnic; j++) {
rxq = bp->eth_dev->data->rx_queues[j];
PMD_DRV_LOG(DEBUG,
"rxq[%d]->vnic=%p vnic->fw_grp_ids=%p\n",
j, rxq->vnic, rxq->vnic->fw_grp_ids);
if (BNXT_HAS_RING_GRPS(bp) && rxq->rx_deferred_start)
rxq->vnic->fw_grp_ids[j] = INVALID_HW_RING_ID;
net/bnxt: allow group ID 0 for RSS action Allow RSS action with group ID 0. The RSS match check will ensure if requested RSS action configuration parameters should be allowed as per the VNIC's RSS configuration or not and if it does match as it is by design in OVS-DPDK use case, the default vnic can be used to create the filter. As part of this ensure that rx_queue_cnt for the default vnic is setup during init as this field was being set only for non-zero vnics while handling RSS action for flow create. Check for out of bounds error while accessing the vnic array. Fixes: adc0f81c6552 ("net/bnxt: support RSS action") Cc: stable@dpdk.org Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-01-28 07:29:17 +00:00
else
vnic->rx_queue_cnt++;
net/bnxt: fix VLAN strip HWRM_VNIC_CFG command to configure VNIC dynamically with traffic running is not working. Driver has to free and recreate the VNIC and then reconfigure the VNIC filters. Fixes: 7fe5668d2ea3 ("net/bnxt: support VLAN filter and strip") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-01-16 13:04:50 +00:00
}
net/bnxt: allow group ID 0 for RSS action Allow RSS action with group ID 0. The RSS match check will ensure if requested RSS action configuration parameters should be allowed as per the VNIC's RSS configuration or not and if it does match as it is by design in OVS-DPDK use case, the default vnic can be used to create the filter. As part of this ensure that rx_queue_cnt for the default vnic is setup during init as this field was being set only for non-zero vnics while handling RSS action for flow create. Check for out of bounds error while accessing the vnic array. Fixes: adc0f81c6552 ("net/bnxt: support RSS action") Cc: stable@dpdk.org Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-01-28 07:29:17 +00:00
PMD_DRV_LOG(DEBUG, "vnic->rx_queue_cnt = %d\n", vnic->rx_queue_cnt);
net/bnxt: fix VLAN strip HWRM_VNIC_CFG command to configure VNIC dynamically with traffic running is not working. Driver has to free and recreate the VNIC and then reconfigure the VNIC filters. Fixes: 7fe5668d2ea3 ("net/bnxt: support VLAN filter and strip") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-01-16 13:04:50 +00:00
rc = bnxt_vnic_rss_configure(bp, vnic);
if (rc)
goto err_out;
bnxt_hwrm_vnic_plcmode_cfg(bp, vnic);
net/bnxt: fix configuring LRO While configuring LRO, driver should check the return value of bnxt_hwrm_vnic_tpa_cfg() HWRM command and return error when the FW command fails. Fixes: 0958d8b6435d ("net/bnxt: support LRO") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2021-04-02 03:25:51 +00:00
rc = bnxt_hwrm_vnic_tpa_cfg(bp, vnic,
(rx_offloads & DEV_RX_OFFLOAD_TCP_LRO) ?
true : false);
if (rc)
goto err_out;
net/bnxt: fix VLAN strip HWRM_VNIC_CFG command to configure VNIC dynamically with traffic running is not working. Driver has to free and recreate the VNIC and then reconfigure the VNIC filters. Fixes: 7fe5668d2ea3 ("net/bnxt: support VLAN filter and strip") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-01-16 13:04:50 +00:00
return 0;
err_out:
PMD_DRV_LOG(ERR, "HWRM vnic %d cfg failure rc: %x\n",
vnic_id, rc);
return rc;
}
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
static int bnxt_register_fc_ctx_mem(struct bnxt *bp)
{
int rc = 0;
net/bnxt: fix allocation of flow stat related structs Consolidate flow stat related structs for performance improvement. The intention of this patch is to reduce the size struct bnxt which had grown because of recent changes and was impacting performance. Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 08:31:59 +00:00
rc = bnxt_hwrm_ctx_rgtr(bp, bp->flow_stat->rx_fc_in_tbl.dma,
&bp->flow_stat->rx_fc_in_tbl.ctx_id);
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
if (rc)
return rc;
PMD_DRV_LOG(DEBUG,
"rx_fc_in_tbl.va = %p rx_fc_in_tbl.dma = %p"
" rx_fc_in_tbl.ctx_id = %d\n",
net/bnxt: fix allocation of flow stat related structs Consolidate flow stat related structs for performance improvement. The intention of this patch is to reduce the size struct bnxt which had grown because of recent changes and was impacting performance. Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 08:31:59 +00:00
bp->flow_stat->rx_fc_in_tbl.va,
(void *)((uintptr_t)bp->flow_stat->rx_fc_in_tbl.dma),
bp->flow_stat->rx_fc_in_tbl.ctx_id);
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
net/bnxt: fix allocation of flow stat related structs Consolidate flow stat related structs for performance improvement. The intention of this patch is to reduce the size struct bnxt which had grown because of recent changes and was impacting performance. Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 08:31:59 +00:00
rc = bnxt_hwrm_ctx_rgtr(bp, bp->flow_stat->rx_fc_out_tbl.dma,
&bp->flow_stat->rx_fc_out_tbl.ctx_id);
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
if (rc)
return rc;
PMD_DRV_LOG(DEBUG,
"rx_fc_out_tbl.va = %p rx_fc_out_tbl.dma = %p"
" rx_fc_out_tbl.ctx_id = %d\n",
net/bnxt: fix allocation of flow stat related structs Consolidate flow stat related structs for performance improvement. The intention of this patch is to reduce the size struct bnxt which had grown because of recent changes and was impacting performance. Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 08:31:59 +00:00
bp->flow_stat->rx_fc_out_tbl.va,
(void *)((uintptr_t)bp->flow_stat->rx_fc_out_tbl.dma),
bp->flow_stat->rx_fc_out_tbl.ctx_id);
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
net/bnxt: fix allocation of flow stat related structs Consolidate flow stat related structs for performance improvement. The intention of this patch is to reduce the size struct bnxt which had grown because of recent changes and was impacting performance. Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 08:31:59 +00:00
rc = bnxt_hwrm_ctx_rgtr(bp, bp->flow_stat->tx_fc_in_tbl.dma,
&bp->flow_stat->tx_fc_in_tbl.ctx_id);
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
if (rc)
return rc;
PMD_DRV_LOG(DEBUG,
"tx_fc_in_tbl.va = %p tx_fc_in_tbl.dma = %p"
" tx_fc_in_tbl.ctx_id = %d\n",
net/bnxt: fix allocation of flow stat related structs Consolidate flow stat related structs for performance improvement. The intention of this patch is to reduce the size struct bnxt which had grown because of recent changes and was impacting performance. Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 08:31:59 +00:00
bp->flow_stat->tx_fc_in_tbl.va,
(void *)((uintptr_t)bp->flow_stat->tx_fc_in_tbl.dma),
bp->flow_stat->tx_fc_in_tbl.ctx_id);
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
net/bnxt: fix allocation of flow stat related structs Consolidate flow stat related structs for performance improvement. The intention of this patch is to reduce the size struct bnxt which had grown because of recent changes and was impacting performance. Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 08:31:59 +00:00
rc = bnxt_hwrm_ctx_rgtr(bp, bp->flow_stat->tx_fc_out_tbl.dma,
&bp->flow_stat->tx_fc_out_tbl.ctx_id);
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
if (rc)
return rc;
PMD_DRV_LOG(DEBUG,
"tx_fc_out_tbl.va = %p tx_fc_out_tbl.dma = %p"
" tx_fc_out_tbl.ctx_id = %d\n",
net/bnxt: fix allocation of flow stat related structs Consolidate flow stat related structs for performance improvement. The intention of this patch is to reduce the size struct bnxt which had grown because of recent changes and was impacting performance. Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 08:31:59 +00:00
bp->flow_stat->tx_fc_out_tbl.va,
(void *)((uintptr_t)bp->flow_stat->tx_fc_out_tbl.dma),
bp->flow_stat->tx_fc_out_tbl.ctx_id);
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
net/bnxt: fix allocation of flow stat related structs Consolidate flow stat related structs for performance improvement. The intention of this patch is to reduce the size struct bnxt which had grown because of recent changes and was impacting performance. Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 08:31:59 +00:00
memset(bp->flow_stat->rx_fc_out_tbl.va,
0,
bp->flow_stat->rx_fc_out_tbl.size);
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
rc = bnxt_hwrm_cfa_counter_cfg(bp, BNXT_DIR_RX,
CFA_COUNTER_CFG_IN_COUNTER_TYPE_FC,
net/bnxt: fix allocation of flow stat related structs Consolidate flow stat related structs for performance improvement. The intention of this patch is to reduce the size struct bnxt which had grown because of recent changes and was impacting performance. Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 08:31:59 +00:00
bp->flow_stat->rx_fc_out_tbl.ctx_id,
bp->flow_stat->max_fc,
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
true);
if (rc)
return rc;
net/bnxt: fix allocation of flow stat related structs Consolidate flow stat related structs for performance improvement. The intention of this patch is to reduce the size struct bnxt which had grown because of recent changes and was impacting performance. Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 08:31:59 +00:00
memset(bp->flow_stat->tx_fc_out_tbl.va,
0,
bp->flow_stat->tx_fc_out_tbl.size);
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
rc = bnxt_hwrm_cfa_counter_cfg(bp, BNXT_DIR_TX,
CFA_COUNTER_CFG_IN_COUNTER_TYPE_FC,
net/bnxt: fix allocation of flow stat related structs Consolidate flow stat related structs for performance improvement. The intention of this patch is to reduce the size struct bnxt which had grown because of recent changes and was impacting performance. Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 08:31:59 +00:00
bp->flow_stat->tx_fc_out_tbl.ctx_id,
bp->flow_stat->max_fc,
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
true);
return rc;
}
static int bnxt_alloc_ctx_mem_buf(char *type, size_t size,
struct bnxt_ctx_mem_buf_info *ctx)
{
if (!ctx)
return -EINVAL;
ctx->va = rte_zmalloc(type, size, 0);
if (ctx->va == NULL)
return -ENOMEM;
rte_mem_lock_page(ctx->va);
ctx->size = size;
ctx->dma = rte_mem_virt2iova(ctx->va);
if (ctx->dma == RTE_BAD_IOVA)
return -ENOMEM;
return 0;
}
static int bnxt_init_fc_ctx_mem(struct bnxt *bp)
{
struct rte_pci_device *pdev = bp->pdev;
char type[RTE_MEMZONE_NAMESIZE];
uint16_t max_fc;
int rc = 0;
net/bnxt: fix allocation of flow stat related structs Consolidate flow stat related structs for performance improvement. The intention of this patch is to reduce the size struct bnxt which had grown because of recent changes and was impacting performance. Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 08:31:59 +00:00
max_fc = bp->flow_stat->max_fc;
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
sprintf(type, "bnxt_rx_fc_in_" PCI_PRI_FMT, pdev->addr.domain,
pdev->addr.bus, pdev->addr.devid, pdev->addr.function);
/* 4 bytes for each counter-id */
net/bnxt: fix allocation of flow stat related structs Consolidate flow stat related structs for performance improvement. The intention of this patch is to reduce the size struct bnxt which had grown because of recent changes and was impacting performance. Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 08:31:59 +00:00
rc = bnxt_alloc_ctx_mem_buf(type,
max_fc * 4,
&bp->flow_stat->rx_fc_in_tbl);
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
if (rc)
return rc;
sprintf(type, "bnxt_rx_fc_out_" PCI_PRI_FMT, pdev->addr.domain,
pdev->addr.bus, pdev->addr.devid, pdev->addr.function);
/* 16 bytes for each counter - 8 bytes pkt_count, 8 bytes byte_count */
net/bnxt: fix allocation of flow stat related structs Consolidate flow stat related structs for performance improvement. The intention of this patch is to reduce the size struct bnxt which had grown because of recent changes and was impacting performance. Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 08:31:59 +00:00
rc = bnxt_alloc_ctx_mem_buf(type,
max_fc * 16,
&bp->flow_stat->rx_fc_out_tbl);
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
if (rc)
return rc;
sprintf(type, "bnxt_tx_fc_in_" PCI_PRI_FMT, pdev->addr.domain,
pdev->addr.bus, pdev->addr.devid, pdev->addr.function);
/* 4 bytes for each counter-id */
net/bnxt: fix allocation of flow stat related structs Consolidate flow stat related structs for performance improvement. The intention of this patch is to reduce the size struct bnxt which had grown because of recent changes and was impacting performance. Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 08:31:59 +00:00
rc = bnxt_alloc_ctx_mem_buf(type,
max_fc * 4,
&bp->flow_stat->tx_fc_in_tbl);
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
if (rc)
return rc;
sprintf(type, "bnxt_tx_fc_out_" PCI_PRI_FMT, pdev->addr.domain,
pdev->addr.bus, pdev->addr.devid, pdev->addr.function);
/* 16 bytes for each counter - 8 bytes pkt_count, 8 bytes byte_count */
net/bnxt: fix allocation of flow stat related structs Consolidate flow stat related structs for performance improvement. The intention of this patch is to reduce the size struct bnxt which had grown because of recent changes and was impacting performance. Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 08:31:59 +00:00
rc = bnxt_alloc_ctx_mem_buf(type,
max_fc * 16,
&bp->flow_stat->tx_fc_out_tbl);
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
if (rc)
return rc;
rc = bnxt_register_fc_ctx_mem(bp);
return rc;
}
static int bnxt_init_ctx_mem(struct bnxt *bp)
{
int rc = 0;
if (!(bp->fw_cap & BNXT_FW_CAP_ADV_FLOW_COUNTERS) ||
net/bnxt: fix allocation of flow stat related structs Consolidate flow stat related structs for performance improvement. The intention of this patch is to reduce the size struct bnxt which had grown because of recent changes and was impacting performance. Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 08:31:59 +00:00
!(BNXT_PF(bp) || BNXT_VF_IS_TRUSTED(bp)) ||
!BNXT_FLOW_XSTATS_EN(bp))
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
return 0;
net/bnxt: fix allocation of flow stat related structs Consolidate flow stat related structs for performance improvement. The intention of this patch is to reduce the size struct bnxt which had grown because of recent changes and was impacting performance. Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 08:31:59 +00:00
rc = bnxt_hwrm_cfa_counter_qcaps(bp, &bp->flow_stat->max_fc);
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
if (rc)
return rc;
rc = bnxt_init_fc_ctx_mem(bp);
return rc;
}
net/bnxt: fix speed setting on certain adapters On BCM957508-N2100 adapters, FW will not allow any user other than BMC to shutdown the port. As a result, bnxt_get_hwrm_link_config() always returns link up. Because of this, driver will not update the new port configurations such as speed, autoneg during a port start. Fixed the condition to invoke bnxt_set_hwrm_link_config() in bnxt_init_chip(). Fixes: 7bc8e9a227cc ("net/bnxt: support async link notification") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-08-28 05:01:12 +00:00
static int bnxt_update_phy_setting(struct bnxt *bp)
net/bnxt: fix VLAN strip HWRM_VNIC_CFG command to configure VNIC dynamically with traffic running is not working. Driver has to free and recreate the VNIC and then reconfigure the VNIC filters. Fixes: 7fe5668d2ea3 ("net/bnxt: support VLAN filter and strip") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-01-16 13:04:50 +00:00
{
net/bnxt: support async link notification This patch adds support to get Link notification asynchronously. The HW sends async notifications on default completion ring. The PMD processes these notifications and logs a message appropriately. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2016-10-11 21:47:50 +00:00
struct rte_eth_link new;
net/bnxt: fix speed setting on certain adapters On BCM957508-N2100 adapters, FW will not allow any user other than BMC to shutdown the port. As a result, bnxt_get_hwrm_link_config() always returns link up. Because of this, driver will not update the new port configurations such as speed, autoneg during a port start. Fixed the condition to invoke bnxt_set_hwrm_link_config() in bnxt_init_chip(). Fixes: 7bc8e9a227cc ("net/bnxt: support async link notification") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-08-28 05:01:12 +00:00
int rc;
rc = bnxt_get_hwrm_link_config(bp, &new);
if (rc) {
PMD_DRV_LOG(ERR, "Failed to get link settings\n");
return rc;
}
/*
* On BCM957508-N2100 adapters, FW will not allow any user other
* than BMC to shutdown the port. bnxt_get_hwrm_link_config() call
* always returns link up. Force phy update always in that case.
*/
if (!new.link_status || IS_BNXT_DEV_957508_N2100(bp)) {
rc = bnxt_set_hwrm_link_config(bp, true);
if (rc) {
PMD_DRV_LOG(ERR, "Failed to update PHY settings\n");
return rc;
}
}
return rc;
}
net/bnxt: refactor init/uninit Move all the individual driver fields allocation routines to one routine - bnxt_drv_init(). This houses all such routines where memory needs to be allocated once during the driver's lifetime and does not need to be torn down during error recovery. Rename some function names in accordance with their functionality. bnxt_init_board() is doing nothing more than mapping the PCI bars, so rename it as such. Given that there is a bnxt_shutdown_nic that is called in dev_stop_op, rename it's counterpart - bnxt_init_chip() that is called in dev_start_op, to bnxt_start_nic. Also helps avoid confusion with some of the other bnxt_init_xxx routines. Rename bnxt_init_fw() to bnxt_get_config() as that is what that routine is doing mostly functionality wise. Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-01-14 02:19:14 +00:00
static int bnxt_start_nic(struct bnxt *bp)
net/bnxt: fix speed setting on certain adapters On BCM957508-N2100 adapters, FW will not allow any user other than BMC to shutdown the port. As a result, bnxt_get_hwrm_link_config() always returns link up. Because of this, driver will not update the new port configurations such as speed, autoneg during a port start. Fixed the condition to invoke bnxt_set_hwrm_link_config() in bnxt_init_chip(). Fixes: 7bc8e9a227cc ("net/bnxt: support async link notification") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-08-28 05:01:12 +00:00
{
net/bnxt: support enable/disable interrupt Implement Rx Queue interrupt enable/disable functions Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:45 +00:00
struct rte_pci_device *pci_dev = RTE_ETH_DEV_TO_PCI(bp->eth_dev);
struct rte_intr_handle *intr_handle = &pci_dev->intr_handle;
uint32_t intr_vector = 0;
uint32_t queue_id, base = BNXT_MISC_VEC_ID;
uint32_t vec = BNXT_MISC_VEC_ID;
net/bnxt: fix queue start/stop operations Packets destined to the to-be-stopped queue should not be dropped (neither in HW nor in the driver), so re-program the RSS Table without this queue on stop and add it back to the table on start unless it is a Representor VF. Since 0th entry is used for default ring, use fw_grp_id + 1 to change the RSS table population logic by programming valid IDs instead of the default zeroth entry in case of an invalid fw_grp_id. Destroy and recreate the trio of Rx rings(compl, Rx, AG) every time in start so that HW is in sync with software. Fixes: 9b63c6fd70e3 ("net/bnxt: support Rx/Tx queue start/stop") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ray Jui <ray.jui@broadcom.com> Reviewed-by: Scott Branden <scott.branden@broadcom.com> Reviewed-by: Randy Schacher <stuart.schacher@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-06-28 20:15:35 +00:00
unsigned int i, j;
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
int rc;
net: add rte prefix to ether defines Add 'RTE_' prefix to defines: - rename ETHER_ADDR_LEN as RTE_ETHER_ADDR_LEN. - rename ETHER_TYPE_LEN as RTE_ETHER_TYPE_LEN. - rename ETHER_CRC_LEN as RTE_ETHER_CRC_LEN. - rename ETHER_HDR_LEN as RTE_ETHER_HDR_LEN. - rename ETHER_MIN_LEN as RTE_ETHER_MIN_LEN. - rename ETHER_MAX_LEN as RTE_ETHER_MAX_LEN. - rename ETHER_MTU as RTE_ETHER_MTU. - rename ETHER_MAX_VLAN_FRAME_LEN as RTE_ETHER_MAX_VLAN_FRAME_LEN. - rename ETHER_MAX_VLAN_ID as RTE_ETHER_MAX_VLAN_ID. - rename ETHER_MAX_JUMBO_FRAME_LEN as RTE_ETHER_MAX_JUMBO_FRAME_LEN. - rename ETHER_MIN_MTU as RTE_ETHER_MIN_MTU. - rename ETHER_LOCAL_ADMIN_ADDR as RTE_ETHER_LOCAL_ADMIN_ADDR. - rename ETHER_GROUP_ADDR as RTE_ETHER_GROUP_ADDR. - rename ETHER_TYPE_IPv4 as RTE_ETHER_TYPE_IPv4. - rename ETHER_TYPE_IPv6 as RTE_ETHER_TYPE_IPv6. - rename ETHER_TYPE_ARP as RTE_ETHER_TYPE_ARP. - rename ETHER_TYPE_VLAN as RTE_ETHER_TYPE_VLAN. - rename ETHER_TYPE_RARP as RTE_ETHER_TYPE_RARP. - rename ETHER_TYPE_QINQ as RTE_ETHER_TYPE_QINQ. - rename ETHER_TYPE_ETAG as RTE_ETHER_TYPE_ETAG. - rename ETHER_TYPE_1588 as RTE_ETHER_TYPE_1588. - rename ETHER_TYPE_SLOW as RTE_ETHER_TYPE_SLOW. - rename ETHER_TYPE_TEB as RTE_ETHER_TYPE_TEB. - rename ETHER_TYPE_LLDP as RTE_ETHER_TYPE_LLDP. - rename ETHER_TYPE_MPLS as RTE_ETHER_TYPE_MPLS. - rename ETHER_TYPE_MPLSM as RTE_ETHER_TYPE_MPLSM. - rename ETHER_VXLAN_HLEN as RTE_ETHER_VXLAN_HLEN. - rename ETHER_ADDR_FMT_SIZE as RTE_ETHER_ADDR_FMT_SIZE. - rename VXLAN_GPE_TYPE_IPV4 as RTE_VXLAN_GPE_TYPE_IPV4. - rename VXLAN_GPE_TYPE_IPV6 as RTE_VXLAN_GPE_TYPE_IPV6. - rename VXLAN_GPE_TYPE_ETH as RTE_VXLAN_GPE_TYPE_ETH. - rename VXLAN_GPE_TYPE_NSH as RTE_VXLAN_GPE_TYPE_NSH. - rename VXLAN_GPE_TYPE_MPLS as RTE_VXLAN_GPE_TYPE_MPLS. - rename VXLAN_GPE_TYPE_GBP as RTE_VXLAN_GPE_TYPE_GBP. - rename VXLAN_GPE_TYPE_VBNG as RTE_VXLAN_GPE_TYPE_VBNG. - rename ETHER_VXLAN_GPE_HLEN as RTE_ETHER_VXLAN_GPE_HLEN. Do not update the command line library to avoid adding a dependency to librte_net. Signed-off-by: Olivier Matz <olivier.matz@6wind.com> Reviewed-by: Stephen Hemminger <stephen@networkplumber.org> Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2019-05-21 16:13:05 +00:00
if (bp->eth_dev->data->mtu > RTE_ETHER_MTU) {
net/bnxt: switch to the new offload API Update bnxt PMD to new ethdev offloads API. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-04-07 17:40:55 +00:00
bp->eth_dev->data->dev_conf.rxmode.offloads |=
DEV_RX_OFFLOAD_JUMBO_FRAME;
net/bnxt: support set MTU This patch adds support to modify MTU using the set_mtu dev_op. To support frames > 2k, the PMD creates an aggregator ring. When a frame greater than 2k is received, it is fragmented and the resulting fragments are DMA'ed to the aggregator ring. Now the driver can support jumbo frames upto 9500 bytes. Signed-off-by: Steeven Li <steeven.li@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:09 +00:00
bp->flags |= BNXT_FLAG_JUMBO;
} else {
net/bnxt: switch to the new offload API Update bnxt PMD to new ethdev offloads API. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-04-07 17:40:55 +00:00
bp->eth_dev->data->dev_conf.rxmode.offloads &=
~DEV_RX_OFFLOAD_JUMBO_FRAME;
net/bnxt: support set MTU This patch adds support to modify MTU using the set_mtu dev_op. To support frames > 2k, the PMD creates an aggregator ring. When a frame greater than 2k is received, it is fragmented and the resulting fragments are DMA'ed to the aggregator ring. Now the driver can support jumbo frames upto 9500 bytes. Signed-off-by: Steeven Li <steeven.li@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:09 +00:00
bp->flags &= ~BNXT_FLAG_JUMBO;
}
net/bnxt: support thor controller This commit adds support to the bnxt PMD for devices based on the BCM57508 "thor" Ethernet controller. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
2019-06-02 17:42:44 +00:00
/* THOR does not support ring groups.
* But we will use the array to save RSS context IDs.
*/
net/bnxt: remove references to Thor Refactor code to remove references to Thor. Instead use P5 as in phase 5 of development cycle since it is applicable to boards other than Thor as well. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2020-12-01 19:15:23 +00:00
if (BNXT_CHIP_P5(bp))
bp->max_ring_grps = BNXT_MAX_RSS_CTXTS_P5;
net/bnxt: support thor controller This commit adds support to the bnxt PMD for devices based on the BCM57508 "thor" Ethernet controller. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
2019-06-02 17:42:44 +00:00
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
rc = bnxt_alloc_all_hwrm_stat_ctxs(bp);
if (rc) {
net/bnxt: use dynamic log type This patch implements driver specific log type doing away with usage of RTE_LOG() for logging. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:55 +00:00
PMD_DRV_LOG(ERR, "HWRM stat ctx alloc failure rc: %x\n", rc);
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
goto err_out;
}
rc = bnxt_alloc_hwrm_rings(bp);
if (rc) {
net/bnxt: use dynamic log type This patch implements driver specific log type doing away with usage of RTE_LOG() for logging. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:55 +00:00
PMD_DRV_LOG(ERR, "HWRM ring alloc failure rc: %x\n", rc);
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
goto err_out;
}
rc = bnxt_alloc_all_hwrm_ring_grps(bp);
if (rc) {
net/bnxt: use dynamic log type This patch implements driver specific log type doing away with usage of RTE_LOG() for logging. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:55 +00:00
PMD_DRV_LOG(ERR, "HWRM ring grp alloc failure: %x\n", rc);
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
goto err_out;
}
net/bnxt: support CoS classification Class of Service (CoS) is a way to manage multiple types of traffic over a network to offer different types of services to applications. CoS classification (priority to cosqueue) is determined by the user and configured through the PF driver. DPDK driver queries this configuration and maps the cos queue ids to different VNICs. This patch adds this support. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-04 03:48:58 +00:00
if (!(bp->vnic_cap_flags & BNXT_VNIC_CAP_COS_CLASSIFY))
goto skip_cosq_cfg;
for (j = 0, i = 0; i < BNXT_COS_QUEUE_COUNT; i++) {
if (bp->rx_cos_queue[i].id != 0xff) {
struct bnxt_vnic_info *vnic = &bp->vnic_info[j++];
if (!vnic) {
PMD_DRV_LOG(ERR,
"Num pools more than FW profile\n");
rc = -EINVAL;
goto err_out;
}
vnic->cos_queue_id = bp->rx_cos_queue[i].id;
bp->rx_cosq_cnt++;
}
}
skip_cosq_cfg:
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
rc = bnxt_mq_rx_configure(bp);
if (rc) {
net/bnxt: use dynamic log type This patch implements driver specific log type doing away with usage of RTE_LOG() for logging. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:55 +00:00
PMD_DRV_LOG(ERR, "MQ mode configure failure rc: %x\n", rc);
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
goto err_out;
}
net/bnxt: modify VNIC accounting Modify VNIC accounting when enabling RFS on newer chips. Unlike legacy chips, newer chips don't need additional VNIC resources for ntuple filter. Fix the code accordingly so that we don't reserve and allocate additional VNICs on newer chips. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-20 05:24:27 +00:00
/* default vnic 0 */
rc = bnxt_setup_one_vnic(bp, 0);
if (rc)
goto err_out;
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
/* VNIC configuration */
net/bnxt: modify VNIC accounting Modify VNIC accounting when enabling RFS on newer chips. Unlike legacy chips, newer chips don't need additional VNIC resources for ntuple filter. Fix the code accordingly so that we don't reserve and allocate additional VNICs on newer chips. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-20 05:24:27 +00:00
if (BNXT_RFS_NEEDS_VNIC(bp)) {
for (i = 1; i < bp->nr_vnics; i++) {
rc = bnxt_setup_one_vnic(bp, i);
if (rc)
goto err_out;
}
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
}
net/bnxt: fix VLAN strip HWRM_VNIC_CFG command to configure VNIC dynamically with traffic running is not working. Driver has to free and recreate the VNIC and then reconfigure the VNIC filters. Fixes: 7fe5668d2ea3 ("net/bnxt: support VLAN filter and strip") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-01-16 13:04:50 +00:00
net/bnxt: support set VF QOS and MAC anti spoof This patch adds support to 1) enable VF MAC anti spoof. 2) QOS configuration for specified VF. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:17 +00:00
rc = bnxt_hwrm_cfa_l2_set_rx_mask(bp, &bp->vnic_info[0], 0, NULL);
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
if (rc) {
net/bnxt: use dynamic log type This patch implements driver specific log type doing away with usage of RTE_LOG() for logging. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:55 +00:00
PMD_DRV_LOG(ERR,
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
"HWRM cfa l2 rx mask failure rc: %x\n", rc);
goto err_out;
}
net/bnxt: support enable/disable interrupt Implement Rx Queue interrupt enable/disable functions Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:45 +00:00
/* check and configure queue intr-vector mapping */
if ((rte_intr_cap_multiple(intr_handle) ||
!RTE_ETH_DEV_SRIOV(bp->eth_dev).active) &&
bp->eth_dev->data->dev_conf.intr_conf.rxq != 0) {
intr_vector = bp->eth_dev->data->nb_rx_queues;
net/bnxt: use dynamic log type This patch implements driver specific log type doing away with usage of RTE_LOG() for logging. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:55 +00:00
PMD_DRV_LOG(DEBUG, "intr_vector = %d\n", intr_vector);
net/bnxt: support enable/disable interrupt Implement Rx Queue interrupt enable/disable functions Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:45 +00:00
if (intr_vector > bp->rx_cp_nr_rings) {
net/bnxt: use dynamic log type This patch implements driver specific log type doing away with usage of RTE_LOG() for logging. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:55 +00:00
PMD_DRV_LOG(ERR, "At most %d intr queues supported",
net/bnxt: support enable/disable interrupt Implement Rx Queue interrupt enable/disable functions Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:45 +00:00
bp->rx_cp_nr_rings);
return -ENOTSUP;
}
net/bnxt: fix error handling in port start 1. during port start, if bnxt_init_chip() return error bnxt_dev_start_op() invokes bnxt_shutdown_nic() which in turn calls bnxt_free_all_hwrm_resources() to free up resources. Hence remove the bnxt_free_all_hwrm_resources() from bnxt_init_chip() failure path. 2. fix to check the return value of rte_intr_enable() as this call can fail. 3. set bp->dev_stopped to 0 only when port start succeeds. 4. handle failure cases in bnxt_init_chip() routine to do proper cleanup and return correct error value. Fixes: b7778e8a1c00 ("net/bnxt: refactor to properly allocate resources for PF/VF") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-07-18 03:35:55 +00:00
rc = rte_intr_efd_enable(intr_handle, intr_vector);
if (rc)
return rc;
net/bnxt: support enable/disable interrupt Implement Rx Queue interrupt enable/disable functions Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:45 +00:00
}
if (rte_intr_dp_is_en(intr_handle) && !intr_handle->intr_vec) {
intr_handle->intr_vec =
rte_zmalloc("intr_vec",
bp->eth_dev->data->nb_rx_queues *
sizeof(int), 0);
if (intr_handle->intr_vec == NULL) {
net/bnxt: use dynamic log type This patch implements driver specific log type doing away with usage of RTE_LOG() for logging. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:55 +00:00
PMD_DRV_LOG(ERR, "Failed to allocate %d rx_queues"
net/bnxt: support enable/disable interrupt Implement Rx Queue interrupt enable/disable functions Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:45 +00:00
" intr_vec", bp->eth_dev->data->nb_rx_queues);
net/bnxt: fix error handling in port start 1. during port start, if bnxt_init_chip() return error bnxt_dev_start_op() invokes bnxt_shutdown_nic() which in turn calls bnxt_free_all_hwrm_resources() to free up resources. Hence remove the bnxt_free_all_hwrm_resources() from bnxt_init_chip() failure path. 2. fix to check the return value of rte_intr_enable() as this call can fail. 3. set bp->dev_stopped to 0 only when port start succeeds. 4. handle failure cases in bnxt_init_chip() routine to do proper cleanup and return correct error value. Fixes: b7778e8a1c00 ("net/bnxt: refactor to properly allocate resources for PF/VF") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-07-18 03:35:55 +00:00
rc = -ENOMEM;
net/bnxt: fix double free in port start failure During port start when bnxt_start_nic() fails, it tries to free "intr_handle->intr_vec" but the variable is not set to NULL after that. If port start fails, driver invokes bnxt_dev_stop() which will lead to a double free of "intr_handle->intr_vec". Fix it by removing the call to free "intr_handle->intr_vec" in the bnxt_start_nic() failure path as it is anyway doing in bnxt_dev_stop(). Fixes: 9d276b439aaf ("net/bnxt: fix error handling in device start") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-04-01 02:53:34 +00:00
goto err_out;
net/bnxt: support enable/disable interrupt Implement Rx Queue interrupt enable/disable functions Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:45 +00:00
}
net/bnxt: use dynamic log type This patch implements driver specific log type doing away with usage of RTE_LOG() for logging. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:55 +00:00
PMD_DRV_LOG(DEBUG, "intr_handle->intr_vec = %p "
net/bnxt: support enable/disable interrupt Implement Rx Queue interrupt enable/disable functions Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:45 +00:00
"intr_handle->nb_efd = %d intr_handle->max_intr = %d\n",
net/bnxt: use dynamic log type This patch implements driver specific log type doing away with usage of RTE_LOG() for logging. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:55 +00:00
intr_handle->intr_vec, intr_handle->nb_efd,
net/bnxt: support enable/disable interrupt Implement Rx Queue interrupt enable/disable functions Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:45 +00:00
intr_handle->max_intr);
net/bnxt: fix interrupt vector initialization Initialize the vector array when it is valid, thereby preventing a case were it may be accessed when the array is unallocated Fixes: 1fe427fd08ee ("net/bnxt: support enable/disable interrupt") Cc: stable@dpdk.org Signed-off-by: Naresh Kumar PBS <nareshkumar.pbs@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com>
2019-06-08 19:22:07 +00:00
for (queue_id = 0; queue_id < bp->eth_dev->data->nb_rx_queues;
queue_id++) {
net/bnxt: fix Rx interrupt vector The receive interrupt vector should be offset by the constant RTE_INTR_VEC_RXTX_OFFSET; otherwise setting up some queue interrupts will fail. Fixes: 1fe427fd08ee ("net/bnxt: support enable/disable interrupt") Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Tested-by: Rahul Gupta <rahul.gupta@broadcom.com>
2019-07-19 06:19:02 +00:00
intr_handle->intr_vec[queue_id] =
vec + BNXT_RX_VEC_START;
net/bnxt: fix interrupt vector initialization Initialize the vector array when it is valid, thereby preventing a case were it may be accessed when the array is unallocated Fixes: 1fe427fd08ee ("net/bnxt: support enable/disable interrupt") Cc: stable@dpdk.org Signed-off-by: Naresh Kumar PBS <nareshkumar.pbs@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com>
2019-06-08 19:22:07 +00:00
if (vec < base + intr_handle->nb_efd - 1)
vec++;
}
net/bnxt: support enable/disable interrupt Implement Rx Queue interrupt enable/disable functions Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:45 +00:00
}
/* enable uio/vfio intr/eventfd mapping */
net/bnxt: fix error handling in port start 1. during port start, if bnxt_init_chip() return error bnxt_dev_start_op() invokes bnxt_shutdown_nic() which in turn calls bnxt_free_all_hwrm_resources() to free up resources. Hence remove the bnxt_free_all_hwrm_resources() from bnxt_init_chip() failure path. 2. fix to check the return value of rte_intr_enable() as this call can fail. 3. set bp->dev_stopped to 0 only when port start succeeds. 4. handle failure cases in bnxt_init_chip() routine to do proper cleanup and return correct error value. Fixes: b7778e8a1c00 ("net/bnxt: refactor to properly allocate resources for PF/VF") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-07-18 03:35:55 +00:00
rc = rte_intr_enable(intr_handle);
net/bnxt: fix probe in FreeBSD In FreeBSD environment, nic_uio driver does not support interrupts and rte_intr_callback_register() will fail to register interrupts which in turn causes bnxt driver probe failure. Fixed driver to ignore interrupt callback failures in FreeBSD. Also fixed to not use a dedicated completion ring for async events from FW and process these events on RXQ0 in FreeBSD. Fixes: 6de4c538b393 ("net/bnxt: fix error handling in port start") Fixes: 43f78b380f89 ("net/bnxt: retry IRQ callback deregistration") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com>
2020-01-16 13:04:47 +00:00
#ifndef RTE_EXEC_ENV_FREEBSD
/* In FreeBSD OS, nic_uio driver does not support interrupts */
net/bnxt: fix error handling in port start 1. during port start, if bnxt_init_chip() return error bnxt_dev_start_op() invokes bnxt_shutdown_nic() which in turn calls bnxt_free_all_hwrm_resources() to free up resources. Hence remove the bnxt_free_all_hwrm_resources() from bnxt_init_chip() failure path. 2. fix to check the return value of rte_intr_enable() as this call can fail. 3. set bp->dev_stopped to 0 only when port start succeeds. 4. handle failure cases in bnxt_init_chip() routine to do proper cleanup and return correct error value. Fixes: b7778e8a1c00 ("net/bnxt: refactor to properly allocate resources for PF/VF") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-07-18 03:35:55 +00:00
if (rc)
net/bnxt: fix double free in port start failure During port start when bnxt_start_nic() fails, it tries to free "intr_handle->intr_vec" but the variable is not set to NULL after that. If port start fails, driver invokes bnxt_dev_stop() which will lead to a double free of "intr_handle->intr_vec". Fix it by removing the call to free "intr_handle->intr_vec" in the bnxt_start_nic() failure path as it is anyway doing in bnxt_dev_stop(). Fixes: 9d276b439aaf ("net/bnxt: fix error handling in device start") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-04-01 02:53:34 +00:00
goto err_out;
net/bnxt: fix probe in FreeBSD In FreeBSD environment, nic_uio driver does not support interrupts and rte_intr_callback_register() will fail to register interrupts which in turn causes bnxt driver probe failure. Fixed driver to ignore interrupt callback failures in FreeBSD. Also fixed to not use a dedicated completion ring for async events from FW and process these events on RXQ0 in FreeBSD. Fixes: 6de4c538b393 ("net/bnxt: fix error handling in port start") Fixes: 43f78b380f89 ("net/bnxt: retry IRQ callback deregistration") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com>
2020-01-16 13:04:47 +00:00
#endif
net/bnxt: support enable/disable interrupt Implement Rx Queue interrupt enable/disable functions Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:45 +00:00
net/bnxt: fix speed setting on certain adapters On BCM957508-N2100 adapters, FW will not allow any user other than BMC to shutdown the port. As a result, bnxt_get_hwrm_link_config() always returns link up. Because of this, driver will not update the new port configurations such as speed, autoneg during a port start. Fixed the condition to invoke bnxt_set_hwrm_link_config() in bnxt_init_chip(). Fixes: 7bc8e9a227cc ("net/bnxt: support async link notification") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-08-28 05:01:12 +00:00
rc = bnxt_update_phy_setting(bp);
if (rc)
net/bnxt: fix double free in port start failure During port start when bnxt_start_nic() fails, it tries to free "intr_handle->intr_vec" but the variable is not set to NULL after that. If port start fails, driver invokes bnxt_dev_stop() which will lead to a double free of "intr_handle->intr_vec". Fix it by removing the call to free "intr_handle->intr_vec" in the bnxt_start_nic() failure path as it is anyway doing in bnxt_dev_stop(). Fixes: 9d276b439aaf ("net/bnxt: fix error handling in device start") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-04-01 02:53:34 +00:00
goto err_out;
net/bnxt: support async link notification This patch adds support to get Link notification asynchronously. The HW sends async notifications on default completion ring. The PMD processes these notifications and logs a message appropriately. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2016-10-11 21:47:50 +00:00
net/bnxt: support flow mark action Add support for RTE_FLOW_ACTION_TYPE_MARK. Use the flow_id provided by FW during flow creation to lookup the mark id provided by the application. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com>
2019-12-21 02:38:49 +00:00
bp->mark_table = rte_zmalloc("bnxt_mark_table", BNXT_MARK_TABLE_SZ, 0);
if (!bp->mark_table)
PMD_DRV_LOG(ERR, "Allocation of mark table failed\n");
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
return 0;
err_out:
net/bnxt: return proper error code If the FW fails bnxt_hwrm_func_reset() with an error status, instead of returning -1, return a more standard value of -EIO. Similarly sometimes the status returned by certain FW commands may not be generic. Return a more standard value of -EIO in that case as well. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-08 20:24:29 +00:00
/* Some of the error status returned by FW may not be from errno.h */
if (rc > 0)
rc = -EIO;
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
return rc;
}
static int bnxt_shutdown_nic(struct bnxt *bp)
{
bnxt_free_all_hwrm_resources(bp);
bnxt_free_all_filters(bp);
bnxt_free_all_vnics(bp);
return 0;
}
net/bnxt: get device infos Gets device info from the bp structure filled in the init() function. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:04 +00:00
/*
* Device configuration and status function
*/
net/bnxt: fix setting link speed bnxt PMD uses the macro BNXT_SUPPORTED_SPEEDS to validate the user requested speed. But this has all the speed values supported by the PMD and is not chip specific. The check against this macro returns success when the user tries set the speed to 100G on a port even if the chip does not support 100G speed. Fixed it to use bnxt_get_speed_capabilities() to check the supported speeds by the chip. Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-05-22 17:42:08 +00:00
uint32_t bnxt_get_speed_capabilities(struct bnxt *bp)
net/bnxt: support speed capabilities query Added information about supported speeds for the port in the "dev_infos_get". As other PMDs are returning the speed capabilities, apps may expect this behavior from bnxt PMD. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-04-21 21:33:45 +00:00
{
net/bnxt: fix resource cleanup Fix resource cleanup in port close. Once the pointers are freed, set them to NULL. Make sure access to the pointers is validated before use. Fixes: bb81e07323bb ("net/bnxt: support LED on/off") Fixes: 804e746c7b73 ("net/bnxt: add hardware resource manager init code") Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Fixes: 698aa7e95325 ("net/bnxt: add code to determine the Tx COS queue") Fixes: 322bd6e70272 ("net/bnxt: add port representor infrastructure") Fixes: 0bf5a0b5ebb8 ("net/bnxt: add a failure log") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-04-08 20:17:47 +00:00
uint32_t link_speed = 0;
net/bnxt: support speed capabilities query Added information about supported speeds for the port in the "dev_infos_get". As other PMDs are returning the speed capabilities, apps may expect this behavior from bnxt PMD. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-04-21 21:33:45 +00:00
uint32_t speed_capa = 0;
net/bnxt: fix resource cleanup Fix resource cleanup in port close. Once the pointers are freed, set them to NULL. Make sure access to the pointers is validated before use. Fixes: bb81e07323bb ("net/bnxt: support LED on/off") Fixes: 804e746c7b73 ("net/bnxt: add hardware resource manager init code") Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Fixes: 698aa7e95325 ("net/bnxt: add code to determine the Tx COS queue") Fixes: 322bd6e70272 ("net/bnxt: add port representor infrastructure") Fixes: 0bf5a0b5ebb8 ("net/bnxt: add a failure log") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-04-08 20:17:47 +00:00
if (bp->link_info == NULL)
return 0;
link_speed = bp->link_info->support_speeds;
net/bnxt: fix PAM4 link negotiation In some instances link was not coming up if PAM4 signaling is enabled. Added check to disable autoneg if FW indicates auto speeds are zero. Use default auto speeds if PAM4 auto speeds is not set. Added a fix for forced link setting. Fixes: c23f9ded0391 ("net/bnxt: support 200G PAM4 link") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-11-02 02:43:16 +00:00
/* If PAM4 is configured, use PAM4 supported speed */
if (link_speed == 0 && bp->link_info->support_pam4_speeds > 0)
link_speed = bp->link_info->support_pam4_speeds;
net/bnxt: support speed capabilities query Added information about supported speeds for the port in the "dev_infos_get". As other PMDs are returning the speed capabilities, apps may expect this behavior from bnxt PMD. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-04-21 21:33:45 +00:00
if (link_speed & HWRM_PORT_PHY_QCFG_OUTPUT_LINK_SPEED_100MB)
speed_capa |= ETH_LINK_SPEED_100M;
if (link_speed & HWRM_PORT_PHY_QCFG_OUTPUT_SUPPORT_SPEEDS_100MBHD)
speed_capa |= ETH_LINK_SPEED_100M_HD;
if (link_speed & HWRM_PORT_PHY_QCFG_OUTPUT_SUPPORT_SPEEDS_1GB)
speed_capa |= ETH_LINK_SPEED_1G;
if (link_speed & HWRM_PORT_PHY_QCFG_OUTPUT_SUPPORT_SPEEDS_2_5GB)
speed_capa |= ETH_LINK_SPEED_2_5G;
if (link_speed & HWRM_PORT_PHY_QCFG_OUTPUT_SUPPORT_SPEEDS_10GB)
speed_capa |= ETH_LINK_SPEED_10G;
if (link_speed & HWRM_PORT_PHY_QCFG_OUTPUT_SUPPORT_SPEEDS_20GB)
speed_capa |= ETH_LINK_SPEED_20G;
if (link_speed & HWRM_PORT_PHY_QCFG_OUTPUT_SUPPORT_SPEEDS_25GB)
speed_capa |= ETH_LINK_SPEED_25G;
if (link_speed & HWRM_PORT_PHY_QCFG_OUTPUT_SUPPORT_SPEEDS_40GB)
speed_capa |= ETH_LINK_SPEED_40G;
if (link_speed & HWRM_PORT_PHY_QCFG_OUTPUT_SUPPORT_SPEEDS_50GB)
speed_capa |= ETH_LINK_SPEED_50G;
if (link_speed & HWRM_PORT_PHY_QCFG_OUTPUT_SUPPORT_SPEEDS_100GB)
speed_capa |= ETH_LINK_SPEED_100G;
net/bnxt: support 200G PAM4 link Thor based NICs can support PAM4 as wells as NRZ link negotiation. With this patch we are adding support for 200G link speeds based on PAM4 signaling. While PAM4 can negotiate speeds for 50G and 100G as well, the PMD will use NRZ signalling for these speeds. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-08-20 03:51:02 +00:00
if (link_speed & HWRM_PORT_PHY_QCFG_OUTPUT_SUPPORT_PAM4_SPEEDS_50G)
speed_capa |= ETH_LINK_SPEED_50G;
if (link_speed & HWRM_PORT_PHY_QCFG_OUTPUT_SUPPORT_PAM4_SPEEDS_100G)
speed_capa |= ETH_LINK_SPEED_100G;
if (link_speed & HWRM_PORT_PHY_QCFG_OUTPUT_SUPPORT_PAM4_SPEEDS_200G)
net/bnxt: support 200G link speed When the driver is loaded on a 200G NIC, the port speed is not displayed correctly. Parse the 200G speed before displaying it. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-05-13 04:51:25 +00:00
speed_capa |= ETH_LINK_SPEED_200G;
net/bnxt: support speed capabilities query Added information about supported speeds for the port in the "dev_infos_get". As other PMDs are returning the speed capabilities, apps may expect this behavior from bnxt PMD. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-04-21 21:33:45 +00:00
net/bnxt: fix allocation of link info struct Dynamically alloc link info structure. Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 11:05:15 +00:00
if (bp->link_info->auto_mode ==
HWRM_PORT_PHY_QCFG_OUTPUT_AUTO_MODE_NONE)
net/bnxt: support speed capabilities query Added information about supported speeds for the port in the "dev_infos_get". As other PMDs are returning the speed capabilities, apps may expect this behavior from bnxt PMD. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-04-21 21:33:45 +00:00
speed_capa |= ETH_LINK_SPEED_FIXED;
return speed_capa;
}
ethdev: change device info get callback to return int Change eth_dev_infos_get_t return value from void to int. Make eth_dev_infos_get_t implementations across all drivers to return negative errno values if case of error conditions. Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2019-09-12 16:42:28 +00:00
static int bnxt_dev_info_get_op(struct rte_eth_dev *eth_dev,
struct rte_eth_dev_info *dev_info)
net/bnxt: get device infos Gets device info from the bp structure filled in the init() function. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:04 +00:00
{
net/bnxt: fix crash in secondary process The secondary process should not modify device state when init is called. The pci device information pointed to by dev_private pointer is not correct in secondary process. Fixes: b7778e8a1c00 ("net/bnxt: refactor to properly allocate resources for PF/VF") Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 17:17:43 +00:00
struct rte_pci_device *pdev = RTE_DEV_TO_PCI(eth_dev->device);
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = eth_dev->data->dev_private;
net/bnxt: get device infos Gets device info from the bp structure filled in the init() function. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:04 +00:00
uint16_t max_vnics, i, j, vpool, vrxq;
net/bnxt: handle multi queue mode properly We are currently not handling multi queue RX/RSS modes correctly. If RSS is not requested, create one VNIC per RXQ. Fixes: 6133f207970c ("net/bnxt: add Rx queue create/destroy") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:26 +00:00
unsigned int max_rx_rings;
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
int rc;
rc = is_bnxt_in_error(bp);
if (rc)
return rc;
net/bnxt: get device infos Gets device info from the bp structure filled in the init() function. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:04 +00:00
/* MAC Specifics */
net/bnxt: fix number of MAC addresses for VMDq We were hardcoding the max MAC addresses to 32, while the HW can support more than that. This was restricting the number of VMDQ pools that we could support. Use the value obtained from FW instead. Fixes: 804e746c7b73 ("net/bnxt: add hardware resource manager init code") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-10-05 15:06:44 +00:00
dev_info->max_mac_addrs = bp->max_l2_ctx;
net/bnxt: get device infos Gets device info from the bp structure filled in the init() function. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:04 +00:00
dev_info->max_hash_mac_addrs = 0;
/* PF/VF specifics */
net/bnxt: refactor to properly allocate resources for PF/VF 1) Move the function reset to bnxt_dev_init. On the same lines, setup, enable and request interrupt to init path. Memory allocation is also being done in the init path. 2) After a function reset, configure the VFs. Distribute resources evenly between all functions (PF and VF) for now. In the future, this should be controllable. 3) The bnxt_vf_info and bnxt_pf_info had lot of duplication. Move the common items to struct bnxt. And only unique items specific to PF remain in the struct bnxt_pf_info. 4) Program the firmware to allow certain commands sent by a VF. Disallowing these will prevent clean VF driver cleanup. 5) Since PF/VF need to allocate resources from a pool in the hardware, use func_qcaps and func_qcfg to appropriately query the capabilities and available resources. 6) If a PF is being initialized and no VFs are allocated, explicitly call func_cfg to allocate the resources. 7) Once resources are requested from the firmware, update local copy of resource count in struct bnxt only after sending the func_qcfg to make sure the allocation request in the firmware went through. The changes in this patch will be used by the subsequent patches to allow proper initialization of PF/VF instance. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:06:59 +00:00
if (BNXT_PF(bp))
net/bnxt: fix crash in secondary process The secondary process should not modify device state when init is called. The pci device information pointed to by dev_private pointer is not correct in secondary process. Fixes: b7778e8a1c00 ("net/bnxt: refactor to properly allocate resources for PF/VF") Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 17:17:43 +00:00
dev_info->max_vfs = pdev->max_vfs;
net/bnxt: fix max rings computation The current max_rings computation does not take into account the case when max_nq_rings is <= num_async_cpr. This results in a wrong value like 0, when max_nq_rings is 1. Fix this by subtracting num_async_cpr only when max_cp_rings > num_async_cpr. Apart from this, the entire logic is currently spread across a few macros, making it hard to read and debug this code. Move this code into an inline function. max_msix is not used in the max_rings calculation. Apparently the max_msix field returned in HWRM_RESC_QCAPS is only valid for Thor and newer chips. On Wh+ it will be equal to min_compl_rings. Also, when a function reset is performed on an application quit, FW will not reset the VF resource pool as per design. This can lead to a strange condition wherein the max_msix field on Wh+ keeps changing on each application re-load thereby throwing throwing off the max_rings computation. Fixes: f03e66cb64ce ("net/bnxt: limit queue count for NS3/Stingray devices") Cc: stable@dpdk.org Signed-off-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-08-28 11:13:22 +00:00
max_rx_rings = bnxt_max_rings(bp);
net/bnxt: handle multi queue mode properly We are currently not handling multi queue RX/RSS modes correctly. If RSS is not requested, create one VNIC per RXQ. Fixes: 6133f207970c ("net/bnxt: add Rx queue create/destroy") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:26 +00:00
/* For the sake of symmetry, max_rx_queues = max_tx_queues */
dev_info->max_rx_queues = max_rx_rings;
dev_info->max_tx_queues = max_rx_rings;
net/bnxt: enable RSS for thor-based controllers Make changes needed to support rss for thor-based controllers. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-06-02 17:42:47 +00:00
dev_info->reta_size = bnxt_rss_hash_tbl_size(bp);
net/bnxt: set the hash key size We were not setting the dev_info.hash_key_size. Setting it now. Fixes: 0a6d2a720078 ("net/bnxt: get device infos") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:34 +00:00
dev_info->hash_key_size = 40;
net/bnxt: refactor to properly allocate resources for PF/VF 1) Move the function reset to bnxt_dev_init. On the same lines, setup, enable and request interrupt to init path. Memory allocation is also being done in the init path. 2) After a function reset, configure the VFs. Distribute resources evenly between all functions (PF and VF) for now. In the future, this should be controllable. 3) The bnxt_vf_info and bnxt_pf_info had lot of duplication. Move the common items to struct bnxt. And only unique items specific to PF remain in the struct bnxt_pf_info. 4) Program the firmware to allow certain commands sent by a VF. Disallowing these will prevent clean VF driver cleanup. 5) Since PF/VF need to allocate resources from a pool in the hardware, use func_qcaps and func_qcfg to appropriately query the capabilities and available resources. 6) If a PF is being initialized and no VFs are allocated, explicitly call func_cfg to allocate the resources. 7) Once resources are requested from the firmware, update local copy of resource count in struct bnxt only after sending the func_qcfg to make sure the allocation request in the firmware went through. The changes in this patch will be used by the subsequent patches to allow proper initialization of PF/VF instance. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:06:59 +00:00
max_vnics = bp->max_vnics;
net/bnxt: get device infos Gets device info from the bp structure filled in the init() function. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:04 +00:00
net/bnxt: fix forwarding with higher mbuf size Maximum frame length supported by hw is 9600 bytes. When user launches testpmd with --mbuf-size=9729, we are posting buffer descriptors of size 9601 to the ring. This was causing packets getting discarded. Fixes: daef48efe5e5 ("net/bnxt: support set MTU") Cc: stable@dpdk.org Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com>
2019-10-02 17:17:31 +00:00
/* MTU specifics */
dev_info->min_mtu = RTE_ETHER_MIN_MTU;
dev_info->max_mtu = BNXT_MAX_MTU;
net/bnxt: get device infos Gets device info from the bp structure filled in the init() function. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:04 +00:00
/* Fast path specifics */
dev_info->min_rx_bufsize = 1;
net/bnxt: fix forwarding with higher mbuf size Maximum frame length supported by hw is 9600 bytes. When user launches testpmd with --mbuf-size=9729, we are posting buffer descriptors of size 9601 to the ring. This was causing packets getting discarded. Fixes: daef48efe5e5 ("net/bnxt: support set MTU") Cc: stable@dpdk.org Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com>
2019-10-02 17:17:31 +00:00
dev_info->max_rx_pktlen = BNXT_MAX_PKT_LEN;
net/bnxt: switch to the new offload API Update bnxt PMD to new ethdev offloads API. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-04-07 17:40:55 +00:00
dev_info->rx_offload_capa = BNXT_DEV_RX_OFFLOAD_SUPPORT;
if (bp->flags & BNXT_FLAG_PTP_SUPPORTED)
dev_info->rx_offload_capa |= DEV_RX_OFFLOAD_TIMESTAMP;
net/bnxt: support fast mbuf free Add support for DEV_TX_OFFLOAD_MBUF_FAST_FREE to bnxt vector mode transmit. This offload may be enabled only when multi-segment transmit is not needed, all transmitted mbufs for a given queue will be allocated from the same pool, and all transmitted mbufs will have a reference count of 1. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-10-06 17:37:52 +00:00
dev_info->tx_queue_offload_capa = DEV_TX_OFFLOAD_MBUF_FAST_FREE;
dev_info->tx_offload_capa = BNXT_DEV_TX_OFFLOAD_SUPPORT |
dev_info->tx_queue_offload_capa;
net/bnxt: switch to the new offload API Update bnxt PMD to new ethdev offloads API. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-04-07 17:40:55 +00:00
dev_info->flow_type_rss_offloads = BNXT_ETH_RSS_SUPPORT;
net/bnxt: get device infos Gets device info from the bp structure filled in the init() function. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:04 +00:00
net/bnxt: support speed capabilities query Added information about supported speeds for the port in the "dev_infos_get". As other PMDs are returning the speed capabilities, apps may expect this behavior from bnxt PMD. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-04-21 21:33:45 +00:00
dev_info->speed_capa = bnxt_get_speed_capabilities(bp);
net/bnxt: get device infos Gets device info from the bp structure filled in the init() function. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:04 +00:00
/* *INDENT-OFF* */
dev_info->default_rxconf = (struct rte_eth_rxconf) {
.rx_thresh = {
.pthresh = 8,
.hthresh = 8,
.wthresh = 0,
},
.rx_free_thresh = 32,
net/bnxt: fix drop enable in get Rx queue info Return correct value for rx_drop_en. Add per-queue field to track rx_drop_en configuration. Fixes: 2fc201884be8 ("net/bnxt: support rxq/txq get information") Cc: stable@dpdk.org Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 17:30:34 +00:00
.rx_drop_en = BNXT_DEFAULT_RX_DROP_EN,
net/bnxt: get device infos Gets device info from the bp structure filled in the init() function. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:04 +00:00
};
dev_info->default_txconf = (struct rte_eth_txconf) {
.tx_thresh = {
.pthresh = 32,
.hthresh = 0,
.wthresh = 0,
},
.tx_free_thresh = 32,
.tx_rs_thresh = 32,
};
net/bnxt: support async link notification This patch adds support to get Link notification asynchronously. The HW sends async notifications on default completion ring. The PMD processes these notifications and logs a message appropriately. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2016-10-11 21:47:50 +00:00
eth_dev->data->dev_conf.intr_conf.lsc = 1;
net/bnxt: support enable/disable interrupt Implement Rx Queue interrupt enable/disable functions Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:45 +00:00
eth_dev->data->dev_conf.intr_conf.rxq = 1;
net/bnxt: set descriptor rings limits Set MIN and MAX descriptor count for TX and RX rings. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-06-28 20:15:30 +00:00
dev_info->rx_desc_lim.nb_min = BNXT_MIN_RING_DESC;
dev_info->rx_desc_lim.nb_max = BNXT_MAX_RX_RING_DESC;
dev_info->tx_desc_lim.nb_min = BNXT_MIN_RING_DESC;
dev_info->tx_desc_lim.nb_max = BNXT_MAX_TX_RING_DESC;
net/bnxt: support enable/disable interrupt Implement Rx Queue interrupt enable/disable functions Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:45 +00:00
net/bnxt: provide switch info if VFR are configured Some applications need switch_info of the device to be returned as a part of eth_dev_info_get(). The offload logic in such applications could use this info. Pass this info to the when VF representors are configured. Fixes: 322bd6e70272 ("net/bnxt: add port representor infrastructure") Cc: stable@dpdk.org Signed-off-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-11 01:55:59 +00:00
if (BNXT_PF(bp) || BNXT_VF_IS_TRUSTED(bp)) {
dev_info->switch_info.name = eth_dev->device->name;
dev_info->switch_info.domain_id = bp->switch_domain_id;
dev_info->switch_info.port_id =
BNXT_PF(bp) ? BNXT_SWITCH_PORT_ID_PF :
BNXT_SWITCH_PORT_ID_TRUSTED_VF;
}
net/bnxt: get device infos Gets device info from the bp structure filled in the init() function. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:04 +00:00
/* *INDENT-ON* */
/*
* TODO: default_rxconf, default_txconf, rx_desc_lim, and tx_desc_lim
* need further investigation.
*/
/* VMDq resources */
vpool = 64; /* ETH_64_POOLS */
vrxq = 128; /* ETH_VMDQ_DCB_NUM_QUEUES */
for (i = 0; i < 4; vpool >>= 1, i++) {
if (max_vnics > vpool) {
for (j = 0; j < 5; vrxq >>= 1, j++) {
if (dev_info->max_rx_queues > vrxq) {
if (vpool > vrxq)
vpool = vrxq;
goto found;
}
}
/* Not enough resources to support VMDq */
break;
}
}
/* Not enough resources to support VMDq */
vpool = 0;
vrxq = 0;
found:
dev_info->max_vmdq_pools = vpool;
dev_info->vmdq_queue_num = vrxq;
dev_info->vmdq_pool_base = 0;
dev_info->vmdq_queue_base = 0;
ethdev: change device info get callback to return int Change eth_dev_infos_get_t return value from void to int. Make eth_dev_infos_get_t implementations across all drivers to return negative errno values if case of error conditions. Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2019-09-12 16:42:28 +00:00
return 0;
net/bnxt: get device infos Gets device info from the bp structure filled in the init() function. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:04 +00:00
}
net/bnxt: add device configure operation The dev_configure_op function calls bnxt_set_hwrm_link_config() to setup the PHY. This calls the new bnxt_parse_eth_link_*() functions to translate from the DPDK macro values to those used by HWRM calls, then calls bnxt_hwrm_port_phy_cfg() to issue the HWRM call. New HWRM calls: bnxt_hwrm_port_phy_cfg: This command configures the PHY device for the port. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:05 +00:00
/* Configure the device based on the configuration provided */
static int bnxt_dev_configure_op(struct rte_eth_dev *eth_dev)
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = eth_dev->data->dev_private;
net/bnxt: switch to the new offload API Update bnxt PMD to new ethdev offloads API. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-04-07 17:40:55 +00:00
uint64_t rx_offloads = eth_dev->data->dev_conf.rxmode.offloads;
net/bnxt: check VF resources if resource manager is enabled If HWRM resource manager is enabled, check VF resources before proceeding. Make sure there are enough resources allocated and return an error in case of insufficient error. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-06-28 20:15:47 +00:00
int rc;
net/bnxt: switch to the new offload API Update bnxt PMD to new ethdev offloads API. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-04-07 17:40:55 +00:00
net/bnxt: add device configure operation The dev_configure_op function calls bnxt_set_hwrm_link_config() to setup the PHY. This calls the new bnxt_parse_eth_link_*() functions to translate from the DPDK macro values to those used by HWRM calls, then calls bnxt_hwrm_port_phy_cfg() to issue the HWRM call. New HWRM calls: bnxt_hwrm_port_phy_cfg: This command configures the PHY device for the port. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:05 +00:00
bp->rx_queues = (void *)eth_dev->data->rx_queues;
bp->tx_queues = (void *)eth_dev->data->tx_queues;
net/bnxt: fix VF resource allocation If the resource requirement of a VF is more than what is allotted to it by the FW, a VF can request to reallocate some of the resources. This is possible only when the NIC is running the Resource Manager aware CHiMP FW. If necessary, calculate Tx and Rx ring count using the new RM API. Otherwise use the Tx and Rx ring count as-is. Update the cp ring count based on the Tx and Rx ring count. Fixes: b7778e8a1c00 ("net/bnxt: refactor to properly allocate resources for PF/VF") Signed-off-by: Jay Ding <jay.ding@broadcom.com> Signed-off-by: Scott Branden <scott.branden@broadcom.com> Reviewed-by: Ray Jui <ray.jui@broadcom.com> Reviewed-by: Randy Schacher <stuart.schacher@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-05-22 18:13:45 +00:00
bp->tx_nr_rings = eth_dev->data->nb_tx_queues;
bp->rx_nr_rings = eth_dev->data->nb_rx_queues;
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
rc = is_bnxt_in_error(bp);
if (rc)
return rc;
net/bnxt: fix VF resource allocation If the resource requirement of a VF is more than what is allotted to it by the FW, a VF can request to reallocate some of the resources. This is possible only when the NIC is running the Resource Manager aware CHiMP FW. If necessary, calculate Tx and Rx ring count using the new RM API. Otherwise use the Tx and Rx ring count as-is. Update the cp ring count based on the Tx and Rx ring count. Fixes: b7778e8a1c00 ("net/bnxt: refactor to properly allocate resources for PF/VF") Signed-off-by: Jay Ding <jay.ding@broadcom.com> Signed-off-by: Scott Branden <scott.branden@broadcom.com> Reviewed-by: Ray Jui <ray.jui@broadcom.com> Reviewed-by: Randy Schacher <stuart.schacher@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-05-22 18:13:45 +00:00
if (BNXT_VF(bp) && (bp->flags & BNXT_FLAG_NEW_RM)) {
net/bnxt: check VF resources if resource manager is enabled If HWRM resource manager is enabled, check VF resources before proceeding. Make sure there are enough resources allocated and return an error in case of insufficient error. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-06-28 20:15:47 +00:00
rc = bnxt_hwrm_check_vf_rings(bp);
if (rc) {
PMD_DRV_LOG(ERR, "HWRM insufficient resources\n");
return -ENOSPC;
}
net/bnxt: fix VF resource allocation If the resource requirement of a VF is more than what is allotted to it by the FW, a VF can request to reallocate some of the resources. This is possible only when the NIC is running the Resource Manager aware CHiMP FW. If necessary, calculate Tx and Rx ring count using the new RM API. Otherwise use the Tx and Rx ring count as-is. Update the cp ring count based on the Tx and Rx ring count. Fixes: b7778e8a1c00 ("net/bnxt: refactor to properly allocate resources for PF/VF") Signed-off-by: Jay Ding <jay.ding@broadcom.com> Signed-off-by: Scott Branden <scott.branden@broadcom.com> Reviewed-by: Ray Jui <ray.jui@broadcom.com> Reviewed-by: Randy Schacher <stuart.schacher@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-05-22 18:13:45 +00:00
net/bnxt: free default completion ring before VF config If the VF resources need to be reconfigured using the bnxt_hwrm_func_reserve_vf_resc, make sure that the default completion ring is freed first before the resources are reserved. Reallocate the ring once the VF resources are configured. Fixes: 7bc8e9a227cc ("net/bnxt: support async link notification") Cc: stable@dpdk.org Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-10 01:41:43 +00:00
/* If a resource has already been allocated - in this case
* it is the async completion ring, free it. Reallocate it after
* resource reservation. This will ensure the resource counts
* are calculated correctly.
*/
net/bnxt: fix race between interrupt handler and dev config Fix a possible that while interrupt handler is in the middle of processing, dev_configure_op() might come in and try to destroy/realloc the default completion ring on which interrupt completions arrive. Synchronize the two so they don't race each other Fixes: 7bc8e9a227cc ("net/bnxt: support async link notification") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-11 04:44:16 +00:00
pthread_mutex_lock(&bp->def_cp_lock);
net/bnxt: free default completion ring before VF config If the VF resources need to be reconfigured using the bnxt_hwrm_func_reserve_vf_resc, make sure that the default completion ring is freed first before the resources are reserved. Reallocate the ring once the VF resources are configured. Fixes: 7bc8e9a227cc ("net/bnxt: support async link notification") Cc: stable@dpdk.org Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-10 01:41:43 +00:00
if (!BNXT_HAS_NQ(bp) && bp->async_cp_ring) {
bnxt_disable_int(bp);
bnxt_free_cp_ring(bp, bp->async_cp_ring);
}
net/bnxt: check VF resources if resource manager is enabled If HWRM resource manager is enabled, check VF resources before proceeding. Make sure there are enough resources allocated and return an error in case of insufficient error. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-06-28 20:15:47 +00:00
rc = bnxt_hwrm_func_reserve_vf_resc(bp, false);
net/bnxt: fix VF resource allocation If the resource requirement of a VF is more than what is allotted to it by the FW, a VF can request to reallocate some of the resources. This is possible only when the NIC is running the Resource Manager aware CHiMP FW. If necessary, calculate Tx and Rx ring count using the new RM API. Otherwise use the Tx and Rx ring count as-is. Update the cp ring count based on the Tx and Rx ring count. Fixes: b7778e8a1c00 ("net/bnxt: refactor to properly allocate resources for PF/VF") Signed-off-by: Jay Ding <jay.ding@broadcom.com> Signed-off-by: Scott Branden <scott.branden@broadcom.com> Reviewed-by: Ray Jui <ray.jui@broadcom.com> Reviewed-by: Randy Schacher <stuart.schacher@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-05-22 18:13:45 +00:00
if (rc) {
PMD_DRV_LOG(ERR, "HWRM resource alloc fail:%x\n", rc);
net/bnxt: fix race between interrupt handler and dev config Fix a possible that while interrupt handler is in the middle of processing, dev_configure_op() might come in and try to destroy/realloc the default completion ring on which interrupt completions arrive. Synchronize the two so they don't race each other Fixes: 7bc8e9a227cc ("net/bnxt: support async link notification") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-11 04:44:16 +00:00
pthread_mutex_unlock(&bp->def_cp_lock);
net/bnxt: fix VF resource allocation If the resource requirement of a VF is more than what is allotted to it by the FW, a VF can request to reallocate some of the resources. This is possible only when the NIC is running the Resource Manager aware CHiMP FW. If necessary, calculate Tx and Rx ring count using the new RM API. Otherwise use the Tx and Rx ring count as-is. Update the cp ring count based on the Tx and Rx ring count. Fixes: b7778e8a1c00 ("net/bnxt: refactor to properly allocate resources for PF/VF") Signed-off-by: Jay Ding <jay.ding@broadcom.com> Signed-off-by: Scott Branden <scott.branden@broadcom.com> Reviewed-by: Ray Jui <ray.jui@broadcom.com> Reviewed-by: Randy Schacher <stuart.schacher@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-05-22 18:13:45 +00:00
return -ENOSPC;
}
net/bnxt: free default completion ring before VF config If the VF resources need to be reconfigured using the bnxt_hwrm_func_reserve_vf_resc, make sure that the default completion ring is freed first before the resources are reserved. Reallocate the ring once the VF resources are configured. Fixes: 7bc8e9a227cc ("net/bnxt: support async link notification") Cc: stable@dpdk.org Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-10 01:41:43 +00:00
if (!BNXT_HAS_NQ(bp) && bp->async_cp_ring) {
rc = bnxt_alloc_async_cp_ring(bp);
net/bnxt: fix race between interrupt handler and dev config Fix a possible that while interrupt handler is in the middle of processing, dev_configure_op() might come in and try to destroy/realloc the default completion ring on which interrupt completions arrive. Synchronize the two so they don't race each other Fixes: 7bc8e9a227cc ("net/bnxt: support async link notification") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-11 04:44:16 +00:00
if (rc) {
pthread_mutex_unlock(&bp->def_cp_lock);
net/bnxt: free default completion ring before VF config If the VF resources need to be reconfigured using the bnxt_hwrm_func_reserve_vf_resc, make sure that the default completion ring is freed first before the resources are reserved. Reallocate the ring once the VF resources are configured. Fixes: 7bc8e9a227cc ("net/bnxt: support async link notification") Cc: stable@dpdk.org Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-10 01:41:43 +00:00
return rc;
net/bnxt: fix race between interrupt handler and dev config Fix a possible that while interrupt handler is in the middle of processing, dev_configure_op() might come in and try to destroy/realloc the default completion ring on which interrupt completions arrive. Synchronize the two so they don't race each other Fixes: 7bc8e9a227cc ("net/bnxt: support async link notification") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-11 04:44:16 +00:00
}
net/bnxt: free default completion ring before VF config If the VF resources need to be reconfigured using the bnxt_hwrm_func_reserve_vf_resc, make sure that the default completion ring is freed first before the resources are reserved. Reallocate the ring once the VF resources are configured. Fixes: 7bc8e9a227cc ("net/bnxt: support async link notification") Cc: stable@dpdk.org Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-10 01:41:43 +00:00
bnxt_enable_int(bp);
}
net/bnxt: fix race between interrupt handler and dev config Fix a possible that while interrupt handler is in the middle of processing, dev_configure_op() might come in and try to destroy/realloc the default completion ring on which interrupt completions arrive. Synchronize the two so they don't race each other Fixes: 7bc8e9a227cc ("net/bnxt: support async link notification") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-11 04:44:16 +00:00
pthread_mutex_unlock(&bp->def_cp_lock);
net/bnxt: fix VF resource allocation If the resource requirement of a VF is more than what is allotted to it by the FW, a VF can request to reallocate some of the resources. This is possible only when the NIC is running the Resource Manager aware CHiMP FW. If necessary, calculate Tx and Rx ring count using the new RM API. Otherwise use the Tx and Rx ring count as-is. Update the cp ring count based on the Tx and Rx ring count. Fixes: b7778e8a1c00 ("net/bnxt: refactor to properly allocate resources for PF/VF") Signed-off-by: Jay Ding <jay.ding@broadcom.com> Signed-off-by: Scott Branden <scott.branden@broadcom.com> Reviewed-by: Ray Jui <ray.jui@broadcom.com> Reviewed-by: Randy Schacher <stuart.schacher@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-05-22 18:13:45 +00:00
}
net/bnxt: add device configure operation The dev_configure_op function calls bnxt_set_hwrm_link_config() to setup the PHY. This calls the new bnxt_parse_eth_link_*() functions to translate from the DPDK macro values to those used by HWRM calls, then calls bnxt_hwrm_port_phy_cfg() to issue the HWRM call. New HWRM calls: bnxt_hwrm_port_phy_cfg: This command configures the PHY device for the port. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:05 +00:00
/* Inherit new configurations */
net/bnxt: check on-chip resources Check for availability of on-chip resources - like Queue count, number stat context, number of ring groups before inheriting and initializing as per application requirements. Also check before creating a Tx or Rx queue make sure there are enough resources to complete the request. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-08 20:24:37 +00:00
if (eth_dev->data->nb_rx_queues > bp->max_rx_rings ||
eth_dev->data->nb_tx_queues > bp->max_tx_rings ||
net/bnxt: use dedicated CPR for async events This commit enables the creation of a dedicated completion ring for asynchronous event handling instead of handling these events on a receive completion ring. For the stingray platform and other platforms needing tighter control of resource utilization, we retain the ability to process async events on a receive completion ring. For Thor-based adapters, we use a dedicated NQ (notification queue) ring for async events (async events can't currently be received on a completion ring due to a firmware limitation). Rename "def_cp_ring" to "async_cp_ring" to better reflect its purpose (async event notifications) and to avoid confusion with VNIC default receive completion rings. Allow rxq 0 to be stopped when not being used for async events. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-24 16:49:32 +00:00
eth_dev->data->nb_rx_queues + eth_dev->data->nb_tx_queues
+ BNXT_NUM_ASYNC_CPR(bp) > bp->max_cp_rings ||
net/bnxt: check on-chip resources Check for availability of on-chip resources - like Queue count, number stat context, number of ring groups before inheriting and initializing as per application requirements. Also check before creating a Tx or Rx queue make sure there are enough resources to complete the request. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-08 20:24:37 +00:00
eth_dev->data->nb_rx_queues + eth_dev->data->nb_tx_queues >
net/bnxt: support thor controller This commit adds support to the bnxt PMD for devices based on the BCM57508 "thor" Ethernet controller. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
2019-06-02 17:42:44 +00:00
bp->max_stat_ctx)
goto resource_error;
if (BNXT_HAS_RING_GRPS(bp) &&
(uint32_t)(eth_dev->data->nb_rx_queues) > bp->max_ring_grps)
goto resource_error;
if (!(eth_dev->data->dev_conf.rxmode.mq_mode & ETH_MQ_RX_RSS) &&
bp->max_vnics < eth_dev->data->nb_rx_queues)
goto resource_error;
net/bnxt: check on-chip resources Check for availability of on-chip resources - like Queue count, number stat context, number of ring groups before inheriting and initializing as per application requirements. Also check before creating a Tx or Rx queue make sure there are enough resources to complete the request. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-08 20:24:37 +00:00
net/bnxt: add device configure operation The dev_configure_op function calls bnxt_set_hwrm_link_config() to setup the PHY. This calls the new bnxt_parse_eth_link_*() functions to translate from the DPDK macro values to those used by HWRM calls, then calls bnxt_hwrm_port_phy_cfg() to issue the HWRM call. New HWRM calls: bnxt_hwrm_port_phy_cfg: This command configures the PHY device for the port. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:05 +00:00
bp->rx_cp_nr_rings = bp->rx_nr_rings;
bp->tx_cp_nr_rings = bp->tx_nr_rings;
drivers/net: fix RSS hash offload flag if no RSS By default RSS hash delivery (offload) is bound to RSS mode and it is incorrect to advertise it as enabled if Rx multi-queue mode has no RSS. Fixes: 8b945a7f7dcb ("drivers/net: update Rx RSS hash offload capabilities") Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2019-11-14 16:40:50 +00:00
if (eth_dev->data->dev_conf.rxmode.mq_mode & ETH_MQ_RX_RSS_FLAG)
rx_offloads |= DEV_RX_OFFLOAD_RSS_HASH;
drivers/net: update Rx RSS hash offload capabilities Add DEV_RX_OFFLOAD_RSS_HASH flag for all PMDs that support RSS hash delivery. Signed-off-by: Pavan Nikhilesh <pbhagavatula@marvell.com> Reviewed-by: Andrew Rybchenko <arybchenko@solarflare.com> Reviewed-by: Hemant Agrawal <hemant.agrawal@nxp.com> Acked-by: Jerin Jacob <jerinj@marvell.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2019-11-11 13:19:08 +00:00
eth_dev->data->dev_conf.rxmode.offloads = rx_offloads;
net/bnxt: set MTU in dev config for jumbo packets MTU setting does not take effect after rte_eth_dev_configure is called with jumbo enable unless it is configured using the set_mtu dev_op. Fixes: daef48efe5e5 ("net/bnxt: support set MTU") Cc: stable@dpdk.org Signed-off-by: Qingmin Liu <qingmin.liu@broadcom.com> Signed-off-by: Scott Branden <scott.branden@broadcom.com> Reviewed-by: Jay Ding <jay.ding@broadcom.com> Reviewed-by: Randy Schacher <stuart.schacher@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-05-01 01:06:10 +00:00
if (rx_offloads & DEV_RX_OFFLOAD_JUMBO_FRAME) {
net/bnxt: add device configure operation The dev_configure_op function calls bnxt_set_hwrm_link_config() to setup the PHY. This calls the new bnxt_parse_eth_link_*() functions to translate from the DPDK macro values to those used by HWRM calls, then calls bnxt_hwrm_port_phy_cfg() to issue the HWRM call. New HWRM calls: bnxt_hwrm_port_phy_cfg: This command configures the PHY device for the port. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:05 +00:00
eth_dev->data->mtu =
net: add rte prefix to ether defines Add 'RTE_' prefix to defines: - rename ETHER_ADDR_LEN as RTE_ETHER_ADDR_LEN. - rename ETHER_TYPE_LEN as RTE_ETHER_TYPE_LEN. - rename ETHER_CRC_LEN as RTE_ETHER_CRC_LEN. - rename ETHER_HDR_LEN as RTE_ETHER_HDR_LEN. - rename ETHER_MIN_LEN as RTE_ETHER_MIN_LEN. - rename ETHER_MAX_LEN as RTE_ETHER_MAX_LEN. - rename ETHER_MTU as RTE_ETHER_MTU. - rename ETHER_MAX_VLAN_FRAME_LEN as RTE_ETHER_MAX_VLAN_FRAME_LEN. - rename ETHER_MAX_VLAN_ID as RTE_ETHER_MAX_VLAN_ID. - rename ETHER_MAX_JUMBO_FRAME_LEN as RTE_ETHER_MAX_JUMBO_FRAME_LEN. - rename ETHER_MIN_MTU as RTE_ETHER_MIN_MTU. - rename ETHER_LOCAL_ADMIN_ADDR as RTE_ETHER_LOCAL_ADMIN_ADDR. - rename ETHER_GROUP_ADDR as RTE_ETHER_GROUP_ADDR. - rename ETHER_TYPE_IPv4 as RTE_ETHER_TYPE_IPv4. - rename ETHER_TYPE_IPv6 as RTE_ETHER_TYPE_IPv6. - rename ETHER_TYPE_ARP as RTE_ETHER_TYPE_ARP. - rename ETHER_TYPE_VLAN as RTE_ETHER_TYPE_VLAN. - rename ETHER_TYPE_RARP as RTE_ETHER_TYPE_RARP. - rename ETHER_TYPE_QINQ as RTE_ETHER_TYPE_QINQ. - rename ETHER_TYPE_ETAG as RTE_ETHER_TYPE_ETAG. - rename ETHER_TYPE_1588 as RTE_ETHER_TYPE_1588. - rename ETHER_TYPE_SLOW as RTE_ETHER_TYPE_SLOW. - rename ETHER_TYPE_TEB as RTE_ETHER_TYPE_TEB. - rename ETHER_TYPE_LLDP as RTE_ETHER_TYPE_LLDP. - rename ETHER_TYPE_MPLS as RTE_ETHER_TYPE_MPLS. - rename ETHER_TYPE_MPLSM as RTE_ETHER_TYPE_MPLSM. - rename ETHER_VXLAN_HLEN as RTE_ETHER_VXLAN_HLEN. - rename ETHER_ADDR_FMT_SIZE as RTE_ETHER_ADDR_FMT_SIZE. - rename VXLAN_GPE_TYPE_IPV4 as RTE_VXLAN_GPE_TYPE_IPV4. - rename VXLAN_GPE_TYPE_IPV6 as RTE_VXLAN_GPE_TYPE_IPV6. - rename VXLAN_GPE_TYPE_ETH as RTE_VXLAN_GPE_TYPE_ETH. - rename VXLAN_GPE_TYPE_NSH as RTE_VXLAN_GPE_TYPE_NSH. - rename VXLAN_GPE_TYPE_MPLS as RTE_VXLAN_GPE_TYPE_MPLS. - rename VXLAN_GPE_TYPE_GBP as RTE_VXLAN_GPE_TYPE_GBP. - rename VXLAN_GPE_TYPE_VBNG as RTE_VXLAN_GPE_TYPE_VBNG. - rename ETHER_VXLAN_GPE_HLEN as RTE_ETHER_VXLAN_GPE_HLEN. Do not update the command line library to avoid adding a dependency to librte_net. Signed-off-by: Olivier Matz <olivier.matz@6wind.com> Reviewed-by: Stephen Hemminger <stephen@networkplumber.org> Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2019-05-21 16:13:05 +00:00
eth_dev->data->dev_conf.rxmode.max_rx_pkt_len -
RTE_ETHER_HDR_LEN - RTE_ETHER_CRC_LEN - VLAN_TAG_SIZE *
BNXT_NUM_VLANS;
net/bnxt: set MTU in dev config for jumbo packets MTU setting does not take effect after rte_eth_dev_configure is called with jumbo enable unless it is configured using the set_mtu dev_op. Fixes: daef48efe5e5 ("net/bnxt: support set MTU") Cc: stable@dpdk.org Signed-off-by: Qingmin Liu <qingmin.liu@broadcom.com> Signed-off-by: Scott Branden <scott.branden@broadcom.com> Reviewed-by: Jay Ding <jay.ding@broadcom.com> Reviewed-by: Randy Schacher <stuart.schacher@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-05-01 01:06:10 +00:00
bnxt_mtu_set_op(eth_dev, eth_dev->data->mtu);
}
net/bnxt: support async link notification This patch adds support to get Link notification asynchronously. The HW sends async notifications on default completion ring. The PMD processes these notifications and logs a message appropriately. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2016-10-11 21:47:50 +00:00
return 0;
net/bnxt: support thor controller This commit adds support to the bnxt PMD for devices based on the BCM57508 "thor" Ethernet controller. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
2019-06-02 17:42:44 +00:00
resource_error:
PMD_DRV_LOG(ERR,
"Insufficient resources to support requested config\n");
PMD_DRV_LOG(ERR,
"Num Queues Requested: Tx %d, Rx %d\n",
eth_dev->data->nb_tx_queues,
eth_dev->data->nb_rx_queues);
PMD_DRV_LOG(ERR,
"MAX: TxQ %d, RxQ %d, CQ %d Stat %d, Grp %d, Vnic %d\n",
bp->max_tx_rings, bp->max_rx_rings, bp->max_cp_rings,
bp->max_stat_ctx, bp->max_ring_grps, bp->max_vnics);
return -ENOSPC;
net/bnxt: support async link notification This patch adds support to get Link notification asynchronously. The HW sends async notifications on default completion ring. The PMD processes these notifications and logs a message appropriately. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2016-10-11 21:47:50 +00:00
}
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
void bnxt_print_link_info(struct rte_eth_dev *eth_dev)
net/bnxt: support async link notification This patch adds support to get Link notification asynchronously. The HW sends async notifications on default completion ring. The PMD processes these notifications and logs a message appropriately. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2016-10-11 21:47:50 +00:00
{
struct rte_eth_link *link = &eth_dev->data->dev_link;
if (link->link_status)
net/bnxt: use dynamic log type This patch implements driver specific log type doing away with usage of RTE_LOG() for logging. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:55 +00:00
PMD_DRV_LOG(INFO, "Port %d Link Up - speed %u Mbps - %s\n",
ethdev: increase port id range Extend port_id definition from uint8_t to uint16_t in lib and drivers data structures, specifically rte_eth_dev_data. Modify the APIs, drivers and app using port_id at the same time. Fix some checkpatch issues from the original code and remove some unnecessary cast operations. release_17_11 and deprecation docs have been updated in this patch. Signed-off-by: Zhiyong Yang <zhiyong.yang@intel.com> Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2017-09-29 07:17:24 +00:00
eth_dev->data->port_id,
net/bnxt: support async link notification This patch adds support to get Link notification asynchronously. The HW sends async notifications on default completion ring. The PMD processes these notifications and logs a message appropriately. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2016-10-11 21:47:50 +00:00
(uint32_t)link->link_speed,
(link->link_duplex == ETH_LINK_FULL_DUPLEX) ?
("full-duplex") : ("half-duplex\n"));
else
net/bnxt: use dynamic log type This patch implements driver specific log type doing away with usage of RTE_LOG() for logging. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:55 +00:00
PMD_DRV_LOG(INFO, "Port %d Link Down\n",
ethdev: increase port id range Extend port_id definition from uint8_t to uint16_t in lib and drivers data structures, specifically rte_eth_dev_data. Modify the APIs, drivers and app using port_id at the same time. Fix some checkpatch issues from the original code and remove some unnecessary cast operations. release_17_11 and deprecation docs have been updated in this patch. Signed-off-by: Zhiyong Yang <zhiyong.yang@intel.com> Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2017-09-29 07:17:24 +00:00
eth_dev->data->port_id);
net/bnxt: support async link notification This patch adds support to get Link notification asynchronously. The HW sends async notifications on default completion ring. The PMD processes these notifications and logs a message appropriately. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2016-10-11 21:47:50 +00:00
}
net/bnxt: compute and store scattered Rx status In preparation for a bnxt vector-mode driver, compute and store scattered_rx status for the device when started. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
2019-05-29 21:02:22 +00:00
/*
* Determine whether the current configuration requires support for scattered
* receive; return 1 if scattered receive is required and 0 if not.
*/
static int bnxt_scattered_rx(struct rte_eth_dev *eth_dev)
{
uint16_t buf_size;
int i;
net/bnxt: advertise scatter Rx offload capability Scattered receive is supported but not included in receive offload capabilities. Fix by adding it and including in scattered receive calculation. Fixes: 9c1507d96ab8 ("net/bnxt: switch to the new offload API") Cc: stable@dpdk.org Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-04 03:49:02 +00:00
if (eth_dev->data->dev_conf.rxmode.offloads & DEV_RX_OFFLOAD_SCATTER)
return 1;
net/bnxt: fix Rx completion ring size calculation The size of the receive completion ring should be recalculated when MTU is increased to a size that requires scattered receive or when LRO is enabled. Move logic for this calculation from the ring configuration path to the device start path. - Made size calculation dependent only on scattered_rx status. - Moved calculation of scattered_rx up in the initialization sequence. - Made LRO offload status part of scattered_rx calculation. When the completion ring size is too small, completion overflows can occur causing the ring to be disabled in hardware. Fixes: 04067844a3e9 ("net/bnxt: reduce CQ queue size without aggregation ring") Cc: stable@dpdk.org Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-01-29 18:07:09 +00:00
if (eth_dev->data->dev_conf.rxmode.offloads & DEV_RX_OFFLOAD_TCP_LRO)
return 1;
net/bnxt: compute and store scattered Rx status In preparation for a bnxt vector-mode driver, compute and store scattered_rx status for the device when started. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
2019-05-29 21:02:22 +00:00
for (i = 0; i < eth_dev->data->nb_rx_queues; i++) {
struct bnxt_rx_queue *rxq = eth_dev->data->rx_queues[i];
buf_size = (uint16_t)(rte_pktmbuf_data_room_size(rxq->mb_pool) -
RTE_PKTMBUF_HEADROOM);
if (eth_dev->data->dev_conf.rxmode.max_rx_pkt_len > buf_size)
return 1;
}
return 0;
}
net/bnxt: implement SSE vector mode Introduce SSE vector mode support for the bnxt pmd. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-05-29 21:02:23 +00:00
static eth_rx_burst_t
net/bnxt: support flow mark action Add support for RTE_FLOW_ACTION_TYPE_MARK. Use the flow_id provided by FW during flow creation to lookup the mark id provided by the application. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com>
2019-12-21 02:38:49 +00:00
bnxt_receive_function(struct rte_eth_dev *eth_dev)
net/bnxt: implement SSE vector mode Introduce SSE vector mode support for the bnxt pmd. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-05-29 21:02:23 +00:00
{
net/bnxt: support flow mark action Add support for RTE_FLOW_ACTION_TYPE_MARK. Use the flow_id provided by FW during flow creation to lookup the mark id provided by the application. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com>
2019-12-21 02:38:49 +00:00
struct bnxt *bp = eth_dev->data->dev_private;
net/bnxt: add Rx logic for 58818 chips 1. On the new 58818 chips, the RX completion is largely the same except for the new completion opcode and the stripped VLAN format and checksum status. Added bnxt_parse_csum_v2(), bnxt_parse_pkt_type_v2() and bnxt_rx_vlan_v2() to support the new RX completion logic. 2. Disable vector mode RX/TX for 58818 chips for now. 3. The cfa_code format on 58818 chips is different than legacy chips. So skip cfa_code parsing logic on 58818 chips for now. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-20 05:24:30 +00:00
/* Disable vector mode RX for Stingray2 for now */
if (BNXT_CHIP_SR2(bp)) {
bp->flags &= ~BNXT_FLAG_RX_VECTOR_PKT_MODE;
return bnxt_recv_pkts;
}
net/bnxt: support NEON Add bnxt vector PMD support using NEON SIMD instructions. Also update the 20.08 release notes with this information. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Ruifeng Wang <ruifeng.wang@arm.com>
2020-07-07 22:22:25 +00:00
#if defined(RTE_ARCH_X86) || defined(RTE_ARCH_ARM64)
net/bnxt: support PTP for Thor On Thor, direct access to PTP registers (via GRC) is not supported. Driver must use HWRM to access the timestamp information. Vectorized Rx/Tx cannot be enabled if RTE_LIBRTE_IEEE1588=y. Remove the PTP flags handling code from the vector Rx path. Add support to read tx timestamp value and the time from the timesync clock. On Thor, Rx timestamps are provided directly in the Rx completion records to the driver. Only 32 bits of the timestamp is present in the completion. Driver needs to read the current 48 bit free running timer using the HWRM_PORT_TS_QUERY command and combine the upper 16 bits from the HWRM response with the lower 32 bits in the Rx completion to produce the 48 bit timestamp for the Rx packet. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 01:23:35 +00:00
#ifndef RTE_LIBRTE_IEEE1588
net/bnxt: implement SSE vector mode Introduce SSE vector mode support for the bnxt pmd. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-05-29 21:02:23 +00:00
/*
* Vector mode receive can be enabled only if scatter rx is not
* in use and rx offloads are limited to VLAN stripping and
* CRC stripping.
*/
if (!eth_dev->data->scattered_rx &&
!(eth_dev->data->dev_conf.rxmode.offloads &
~(DEV_RX_OFFLOAD_VLAN_STRIP |
DEV_RX_OFFLOAD_KEEP_CRC |
DEV_RX_OFFLOAD_JUMBO_FRAME |
DEV_RX_OFFLOAD_IPV4_CKSUM |
DEV_RX_OFFLOAD_UDP_CKSUM |
DEV_RX_OFFLOAD_TCP_CKSUM |
DEV_RX_OFFLOAD_OUTER_IPV4_CKSUM |
net/bnxt: fix outer UDP checksum Rx offload capability Rx outer UDP checksum offload has been supported for some time, but this has not been advertised in offload capability flags. Fix this, and allow vector mode receive to be enabled when DEV_RX_OFFLOAD_OUTER_UDP_CKSUM is requested. Fixes: 04a681426d49 ("net/bnxt: fix L4 checksum error indication in Rx") Cc: stable@dpdk.org Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-17 19:24:31 +00:00
DEV_RX_OFFLOAD_OUTER_UDP_CKSUM |
net/bnxt: fix vector Rx selection Take DEV_RX_OFFLOAD_RSS_HASH into account when deciding whether vector mode receive can be enabled. Fixes: 8b945a7f7dcb ("drivers/net: update Rx RSS hash offload capabilities") Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-11-19 22:12:15 +00:00
DEV_RX_OFFLOAD_RSS_HASH |
net/bnxt: disable vector mode on host based TruFlow If bp->truflow is not set then don't enable vector mode. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-04-15 08:19:09 +00:00
DEV_RX_OFFLOAD_VLAN_FILTER)) &&
net/bnxt: check max SIMD bitwidth When choosing a vector path to take, an extra condition must be satisfied to ensure the max SIMD bitwidth allows for the CPU enabled path. Signed-off-by: Ciara Power <ciara.power@intel.com> Acked-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-10-19 13:48:45 +00:00
!BNXT_TRUFLOW_EN(bp) && BNXT_NUM_ASYNC_CPR(bp) &&
rte_vect_get_max_simd_bitwidth() >= RTE_VECT_SIMD_128) {
net/bnxt: implement SSE vector mode Introduce SSE vector mode support for the bnxt pmd. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-05-29 21:02:23 +00:00
PMD_DRV_LOG(INFO, "Using vector mode receive for port %d\n",
eth_dev->data->port_id);
net/bnxt: support flow mark action Add support for RTE_FLOW_ACTION_TYPE_MARK. Use the flow_id provided by FW during flow creation to lookup the mark id provided by the application. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com>
2019-12-21 02:38:49 +00:00
bp->flags |= BNXT_FLAG_RX_VECTOR_PKT_MODE;
net/bnxt: implement SSE vector mode Introduce SSE vector mode support for the bnxt pmd. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-05-29 21:02:23 +00:00
return bnxt_recv_pkts_vec;
}
PMD_DRV_LOG(INFO, "Vector mode receive disabled for port %d\n",
eth_dev->data->port_id);
PMD_DRV_LOG(INFO,
"Port %d scatter: %d rx offload: %" PRIX64 "\n",
eth_dev->data->port_id,
eth_dev->data->scattered_rx,
eth_dev->data->dev_conf.rxmode.offloads);
net/bnxt: support PTP for Thor On Thor, direct access to PTP registers (via GRC) is not supported. Driver must use HWRM to access the timestamp information. Vectorized Rx/Tx cannot be enabled if RTE_LIBRTE_IEEE1588=y. Remove the PTP flags handling code from the vector Rx path. Add support to read tx timestamp value and the time from the timesync clock. On Thor, Rx timestamps are provided directly in the Rx completion records to the driver. Only 32 bits of the timestamp is present in the completion. Driver needs to read the current 48 bit free running timer using the HWRM_PORT_TS_QUERY command and combine the upper 16 bits from the HWRM response with the lower 32 bits in the Rx completion to produce the 48 bit timestamp for the Rx packet. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 01:23:35 +00:00
#endif
net/bnxt: implement SSE vector mode Introduce SSE vector mode support for the bnxt pmd. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-05-29 21:02:23 +00:00
#endif
net/bnxt: support flow mark action Add support for RTE_FLOW_ACTION_TYPE_MARK. Use the flow_id provided by FW during flow creation to lookup the mark id provided by the application. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com>
2019-12-21 02:38:49 +00:00
bp->flags &= ~BNXT_FLAG_RX_VECTOR_PKT_MODE;
net/bnxt: implement SSE vector mode Introduce SSE vector mode support for the bnxt pmd. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-05-29 21:02:23 +00:00
return bnxt_recv_pkts;
}
static eth_tx_burst_t
net/bnxt: add Rx logic for 58818 chips 1. On the new 58818 chips, the RX completion is largely the same except for the new completion opcode and the stripped VLAN format and checksum status. Added bnxt_parse_csum_v2(), bnxt_parse_pkt_type_v2() and bnxt_rx_vlan_v2() to support the new RX completion logic. 2. Disable vector mode RX/TX for 58818 chips for now. 3. The cfa_code format on 58818 chips is different than legacy chips. So skip cfa_code parsing logic on 58818 chips for now. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-20 05:24:30 +00:00
bnxt_transmit_function(struct rte_eth_dev *eth_dev)
net/bnxt: implement SSE vector mode Introduce SSE vector mode support for the bnxt pmd. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-05-29 21:02:23 +00:00
{
net/bnxt: add Rx logic for 58818 chips 1. On the new 58818 chips, the RX completion is largely the same except for the new completion opcode and the stripped VLAN format and checksum status. Added bnxt_parse_csum_v2(), bnxt_parse_pkt_type_v2() and bnxt_rx_vlan_v2() to support the new RX completion logic. 2. Disable vector mode RX/TX for 58818 chips for now. 3. The cfa_code format on 58818 chips is different than legacy chips. So skip cfa_code parsing logic on 58818 chips for now. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-20 05:24:30 +00:00
struct bnxt *bp = eth_dev->data->dev_private;
/* Disable vector mode TX for Stingray2 for now */
if (BNXT_CHIP_SR2(bp))
return bnxt_xmit_pkts;
net/bnxt: support NEON Add bnxt vector PMD support using NEON SIMD instructions. Also update the 20.08 release notes with this information. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Ruifeng Wang <ruifeng.wang@arm.com>
2020-07-07 22:22:25 +00:00
#if defined(RTE_ARCH_X86) || defined(RTE_ARCH_ARM64)
net/bnxt: support PTP for Thor On Thor, direct access to PTP registers (via GRC) is not supported. Driver must use HWRM to access the timestamp information. Vectorized Rx/Tx cannot be enabled if RTE_LIBRTE_IEEE1588=y. Remove the PTP flags handling code from the vector Rx path. Add support to read tx timestamp value and the time from the timesync clock. On Thor, Rx timestamps are provided directly in the Rx completion records to the driver. Only 32 bits of the timestamp is present in the completion. Driver needs to read the current 48 bit free running timer using the HWRM_PORT_TS_QUERY command and combine the upper 16 bits from the HWRM response with the lower 32 bits in the Rx completion to produce the 48 bit timestamp for the Rx packet. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 01:23:35 +00:00
#ifndef RTE_LIBRTE_IEEE1588
net/bnxt: support fast mbuf free Add support for DEV_TX_OFFLOAD_MBUF_FAST_FREE to bnxt vector mode transmit. This offload may be enabled only when multi-segment transmit is not needed, all transmitted mbufs for a given queue will be allocated from the same pool, and all transmitted mbufs will have a reference count of 1. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-10-06 17:37:52 +00:00
uint64_t offloads = eth_dev->data->dev_conf.txmode.offloads;
net/bnxt: disable Tx vector mode on TruFlow The vector mode in the tx handler is disabled when truflow is enabled since truflow now requires bd action record support. Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:28:22 +00:00
net/bnxt: implement SSE vector mode Introduce SSE vector mode support for the bnxt pmd. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-05-29 21:02:23 +00:00
/*
net/bnxt: disable vector mode Tx with VLAN offload The vector mode transmit path does not currently support VLAN tag insertion, so we need to disable vector transmit when transmit VLAN insertion offload is enabled. Fixes: bc4a000f2f53 ("net/bnxt: implement SSE vector mode") Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-19 06:19:04 +00:00
* Vector mode transmit can be enabled only if not using scatter rx
* or tx offloads.
net/bnxt: implement SSE vector mode Introduce SSE vector mode support for the bnxt pmd. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-05-29 21:02:23 +00:00
*/
if (!eth_dev->data->scattered_rx &&
net/bnxt: support fast mbuf free Add support for DEV_TX_OFFLOAD_MBUF_FAST_FREE to bnxt vector mode transmit. This offload may be enabled only when multi-segment transmit is not needed, all transmitted mbufs for a given queue will be allocated from the same pool, and all transmitted mbufs will have a reference count of 1. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-10-06 17:37:52 +00:00
!(offloads & ~DEV_TX_OFFLOAD_MBUF_FAST_FREE) &&
net/bnxt: check max SIMD bitwidth When choosing a vector path to take, an extra condition must be satisfied to ensure the max SIMD bitwidth allows for the CPU enabled path. Signed-off-by: Ciara Power <ciara.power@intel.com> Acked-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-10-19 13:48:45 +00:00
!BNXT_TRUFLOW_EN(bp) &&
rte_vect_get_max_simd_bitwidth() >= RTE_VECT_SIMD_128) {
net/bnxt: implement SSE vector mode Introduce SSE vector mode support for the bnxt pmd. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-05-29 21:02:23 +00:00
PMD_DRV_LOG(INFO, "Using vector mode transmit for port %d\n",
eth_dev->data->port_id);
return bnxt_xmit_pkts_vec;
}
PMD_DRV_LOG(INFO, "Vector mode transmit disabled for port %d\n",
eth_dev->data->port_id);
PMD_DRV_LOG(INFO,
"Port %d scatter: %d tx offload: %" PRIX64 "\n",
eth_dev->data->port_id,
eth_dev->data->scattered_rx,
net/bnxt: support fast mbuf free Add support for DEV_TX_OFFLOAD_MBUF_FAST_FREE to bnxt vector mode transmit. This offload may be enabled only when multi-segment transmit is not needed, all transmitted mbufs for a given queue will be allocated from the same pool, and all transmitted mbufs will have a reference count of 1. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-10-06 17:37:52 +00:00
offloads);
net/bnxt: support PTP for Thor On Thor, direct access to PTP registers (via GRC) is not supported. Driver must use HWRM to access the timestamp information. Vectorized Rx/Tx cannot be enabled if RTE_LIBRTE_IEEE1588=y. Remove the PTP flags handling code from the vector Rx path. Add support to read tx timestamp value and the time from the timesync clock. On Thor, Rx timestamps are provided directly in the Rx completion records to the driver. Only 32 bits of the timestamp is present in the completion. Driver needs to read the current 48 bit free running timer using the HWRM_PORT_TS_QUERY command and combine the upper 16 bits from the HWRM response with the lower 32 bits in the Rx completion to produce the 48 bit timestamp for the Rx packet. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 01:23:35 +00:00
#endif
net/bnxt: implement SSE vector mode Introduce SSE vector mode support for the bnxt pmd. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-05-29 21:02:23 +00:00
#endif
return bnxt_xmit_pkts;
}
net/bnxt: inform firmware about IF state changes Use latest firmware API to inform firmware about IF state changes. Firmware has the option to clean up resources during IF down and to require the driver to reserve resources again during IF up. During port start, HWRM_FUNC_DRV_IF_CHANGE command response flags indicates that firmware has reset. Add logic to re-probe the firmware and re-setup resources. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 01:23:24 +00:00
static int bnxt_handle_if_change_status(struct bnxt *bp)
{
int rc;
/* Since fw has undergone a reset and lost all contexts,
* set fatal flag to not issue hwrm during cleanup
*/
bp->flags |= BNXT_FLAG_FATAL_ERROR;
bnxt_uninit_resources(bp, true);
/* clear fatal flag so that re-init happens */
bp->flags &= ~BNXT_FLAG_FATAL_ERROR;
rc = bnxt_init_resources(bp, true);
bp->flags &= ~BNXT_FLAG_IF_CHANGE_HOT_FW_RESET_DONE;
return rc;
}
net/bnxt: add link state operations Adds dev_ops to set link Up or Down as appropriate. Uses the bnxt_set_hwrm_link_config() API added previously. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:35 +00:00
static int bnxt_dev_set_link_up_op(struct rte_eth_dev *eth_dev)
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = eth_dev->data->dev_private;
net/bnxt: fix link handling and configuration Remove a case where we were sending a deprecated field to the FW. There is no need to send auto_link_speed to the FW. Also set the auto_mode correctly depending on the setting requested. Fixes: 7bc8e9a227cc ("net/bnxt: support async link notification") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-11-09 17:46:28 +00:00
int rc = 0;
net/bnxt: fix link state operations VFs does not have the privilege to change link configuration. But the driver silently returns success to these ethdev callbacks without actually issuing the HWRM command to bring the link up/down. Fixes: 5c206086feaa ("net/bnxt: add link state operations") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-03-18 09:35:22 +00:00
if (!BNXT_SINGLE_PF(bp))
return -ENOTSUP;
net/bnxt: fix allocation of link info struct Dynamically alloc link info structure. Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 11:05:15 +00:00
if (!bp->link_info->link_up)
net/bnxt: fix link handling and configuration Remove a case where we were sending a deprecated field to the FW. There is no need to send auto_link_speed to the FW. Also set the auto_mode correctly depending on the setting requested. Fixes: 7bc8e9a227cc ("net/bnxt: support async link notification") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-11-09 17:46:28 +00:00
rc = bnxt_set_hwrm_link_config(bp, true);
if (!rc)
eth_dev->data->dev_link.link_status = 1;
net/bnxt: add link state operations Adds dev_ops to set link Up or Down as appropriate. Uses the bnxt_set_hwrm_link_config() API added previously. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:35 +00:00
net/bnxt: fix link handling and configuration Remove a case where we were sending a deprecated field to the FW. There is no need to send auto_link_speed to the FW. Also set the auto_mode correctly depending on the setting requested. Fixes: 7bc8e9a227cc ("net/bnxt: support async link notification") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-11-09 17:46:28 +00:00
bnxt_print_link_info(eth_dev);
net/bnxt: return error if setting link up fails Currently bnxt driver does not return error in case setting link up fails. Fixes: 5c206086feaa ("net/bnxt: add link state operations") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-10 01:41:44 +00:00
return rc;
net/bnxt: add link state operations Adds dev_ops to set link Up or Down as appropriate. Uses the bnxt_set_hwrm_link_config() API added previously. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:35 +00:00
}
static int bnxt_dev_set_link_down_op(struct rte_eth_dev *eth_dev)
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = eth_dev->data->dev_private;
net/bnxt: add link state operations Adds dev_ops to set link Up or Down as appropriate. Uses the bnxt_set_hwrm_link_config() API added previously. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:35 +00:00
net/bnxt: fix link state operations VFs does not have the privilege to change link configuration. But the driver silently returns success to these ethdev callbacks without actually issuing the HWRM command to bring the link up/down. Fixes: 5c206086feaa ("net/bnxt: add link state operations") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-03-18 09:35:22 +00:00
if (!BNXT_SINGLE_PF(bp))
return -ENOTSUP;
net/bnxt: add link state operations Adds dev_ops to set link Up or Down as appropriate. Uses the bnxt_set_hwrm_link_config() API added previously. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:35 +00:00
eth_dev->data->dev_link.link_status = 0;
bnxt_set_hwrm_link_config(bp, false);
net/bnxt: fix allocation of link info struct Dynamically alloc link info structure. Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 11:05:15 +00:00
bp->link_info->link_up = 0;
net/bnxt: fix link handling and configuration Remove a case where we were sending a deprecated field to the FW. There is no need to send auto_link_speed to the FW. Also set the auto_mode correctly depending on the setting requested. Fixes: 7bc8e9a227cc ("net/bnxt: support async link notification") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-11-09 17:46:28 +00:00
net/bnxt: add link state operations Adds dev_ops to set link Up or Down as appropriate. Uses the bnxt_set_hwrm_link_config() API added previously. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:35 +00:00
return 0;
}
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
static void bnxt_free_switch_domain(struct bnxt *bp)
{
net/bnxt: add a failure log Check and log an error message if switch domain free API fails Coverity issue: 362757 Fixes: 322bd6e70272 ("net/bnxt: add port representor infrastructure") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
2020-10-30 07:03:03 +00:00
int rc = 0;
net/bnxt: fix resource cleanup Fix resource cleanup in port close. Once the pointers are freed, set them to NULL. Make sure access to the pointers is validated before use. Fixes: bb81e07323bb ("net/bnxt: support LED on/off") Fixes: 804e746c7b73 ("net/bnxt: add hardware resource manager init code") Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Fixes: 698aa7e95325 ("net/bnxt: add code to determine the Tx COS queue") Fixes: 322bd6e70272 ("net/bnxt: add port representor infrastructure") Fixes: 0bf5a0b5ebb8 ("net/bnxt: add a failure log") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-04-08 20:17:47 +00:00
if (!(BNXT_PF(bp) || BNXT_VF_IS_TRUSTED(bp)))
return;
rc = rte_eth_switch_domain_free(bp->switch_domain_id);
if (rc)
PMD_DRV_LOG(ERR, "free switch domain:%d fail: %d\n",
bp->switch_domain_id, rc);
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
}
net/bnxt: fix PTP support for Thor On Thor, Rx timestamp is present in the Rx completion record. Only 32 bits of the timestamp is present in the completion. The driver needs to periodically poll the current 48 bit free running timer using the HWRM_PORT_TS_QUERY command. It can combine the upper 16 bits from the HWRM response with the lower 32 bits in the Rx completion to produce the 48 bit timestamp for the Rx packet. This patch adds an alarm thread to periodically poll the current 48 bit free running timer using the HWRM_PORT_TS_QUERY command. This avoids issuing the hwrm command from the Rx handler. This patch also handles the timer roll over condition. Fixes: 6cbd89f9f3d8 ("net/bnxt: support PTP for Thor") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-02-24 15:55:53 +00:00
static void bnxt_ptp_get_current_time(void *arg)
{
struct bnxt *bp = arg;
struct bnxt_ptp_cfg *ptp = bp->ptp_cfg;
int rc;
rc = is_bnxt_in_error(bp);
if (rc)
return;
if (!ptp)
return;
bnxt_hwrm_port_ts_query(bp, BNXT_PTP_FLAGS_CURRENT_TIME,
&ptp->current_time);
rc = rte_eal_alarm_set(US_PER_S, bnxt_ptp_get_current_time, (void *)bp);
if (rc != 0) {
PMD_DRV_LOG(ERR, "Failed to re-schedule PTP alarm\n");
bp->flags2 &= ~BNXT_FLAGS2_PTP_ALARM_SCHEDULED;
}
}
static int bnxt_schedule_ptp_alarm(struct bnxt *bp)
{
struct bnxt_ptp_cfg *ptp = bp->ptp_cfg;
int rc;
if (bp->flags2 & BNXT_FLAGS2_PTP_ALARM_SCHEDULED)
return 0;
bnxt_hwrm_port_ts_query(bp, BNXT_PTP_FLAGS_CURRENT_TIME,
&ptp->current_time);
rc = rte_eal_alarm_set(US_PER_S, bnxt_ptp_get_current_time, (void *)bp);
return rc;
}
static void bnxt_cancel_ptp_alarm(struct bnxt *bp)
{
if (bp->flags2 & BNXT_FLAGS2_PTP_ALARM_SCHEDULED) {
rte_eal_alarm_cancel(bnxt_ptp_get_current_time, (void *)bp);
bp->flags2 &= ~BNXT_FLAGS2_PTP_ALARM_SCHEDULED;
}
}
static void bnxt_ptp_stop(struct bnxt *bp)
{
bnxt_cancel_ptp_alarm(bp);
bp->flags2 &= ~BNXT_FLAGS2_PTP_TIMESYNC_ENABLED;
}
static int bnxt_ptp_start(struct bnxt *bp)
{
int rc;
rc = bnxt_schedule_ptp_alarm(bp);
if (rc != 0) {
PMD_DRV_LOG(ERR, "Failed to schedule PTP alarm\n");
} else {
bp->flags2 |= BNXT_FLAGS2_PTP_TIMESYNC_ENABLED;
bp->flags2 |= BNXT_FLAGS2_PTP_ALARM_SCHEDULED;
}
return rc;
}
net/bnxt: check chip reset in stop and close While the error recovery thread is running, an application can invoke dev_stop or dev_close_op thus triggering a race and unwanted consequences if dev_close is invoked while the recovery is not yet completed. Fix by having another lock to synchronize between the 2 threads and return EGAIN if adapter is in the middle of recovery when dev_stop or dev_close ops are invoked Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-24 09:37:34 +00:00
static int bnxt_dev_stop(struct rte_eth_dev *eth_dev)
net/bnxt: fix crash when closing This patch fixes segfault encountered during dev_uninit/close routine. KNI sample app can be used to reproduce the issue. Fixes: c09f57b49c13 ("net/bnxt: add start/stop/link update operations") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2016-09-29 17:03:44 +00:00
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = eth_dev->data->dev_private;
net/bnxt: fix enabling/disabling interrupts 1. Disable interrupts in dev_stop_op() 2. Enable interrupts in dev_start_op() 3. Clean queue intr-vector mapping in dev_stop_op() and thus fix a possible memory leak. Fixes: c09f57b49c13 ("net/bnxt: add start/stop/link update operations") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-07-17 10:41:35 +00:00
struct rte_pci_device *pci_dev = RTE_ETH_DEV_TO_PCI(eth_dev);
struct rte_intr_handle *intr_handle = &pci_dev->intr_handle;
net/bnxt: fix link update 1. When port is stopped, we can forcibly set the link status for the device to down. 2. VFs and MH PFs do not have the privilege to bring the link down. As a result driver prints "Link Up" when port is stopped. 3. When driver receives link status/speed/config async event from fw, driver invokes bnxt_link_update() with exp_link_status as ETH_LINK_UP This is not logically correct as the async event could be for Link up or link down or for speed change. Fixes: 074cacb9907a ("net/bnxt: fix link during port toggle") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-10-06 16:01:56 +00:00
struct rte_eth_link link;
ethdev: change stop operation callback to return int Change eth_dev_stop_t return value from void to int. Make eth_dev_stop_t implementations across all drivers to return negative errno values if case of error conditions. Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2020-10-15 13:30:45 +00:00
int ret;
net/bnxt: fix enabling/disabling interrupts 1. Disable interrupts in dev_stop_op() 2. Enable interrupts in dev_start_op() 3. Clean queue intr-vector mapping in dev_stop_op() and thus fix a possible memory leak. Fixes: c09f57b49c13 ("net/bnxt: add start/stop/link update operations") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-07-17 10:41:35 +00:00
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
eth_dev->data->dev_started = 0;
net/bnxt: reduce CQ queue size without aggregation ring Don't allocate extra completion queue entries for aggregation ring when aggregation ring will not be used. Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
2020-09-09 15:52:59 +00:00
eth_dev->data->scattered_rx = 0;
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
/* Prevent crashes when queues are still in use */
eth_dev->rx_pkt_burst = &bnxt_dummy_recv_pkts;
eth_dev->tx_pkt_burst = &bnxt_dummy_xmit_pkts;
net/bnxt: fix enabling/disabling interrupts 1. Disable interrupts in dev_stop_op() 2. Enable interrupts in dev_start_op() 3. Clean queue intr-vector mapping in dev_stop_op() and thus fix a possible memory leak. Fixes: c09f57b49c13 ("net/bnxt: add start/stop/link update operations") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-07-17 10:41:35 +00:00
bnxt_disable_int(bp);
/* disable uio/vfio intr/eventfd mapping */
rte_intr_disable(intr_handle);
net/bnxt: fix crash when closing This patch fixes segfault encountered during dev_uninit/close routine. KNI sample app can be used to reproduce the issue. Fixes: c09f57b49c13 ("net/bnxt: add start/stop/link update operations") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2016-09-29 17:03:44 +00:00
net/bnxt: refactor VFR port clean up When parent VF or PF ports are cleaned up, the child VF representor ports also need to be cleaned up. If not cleaned up, then deleting the parent VF shall result in not cleaning up the hardware rules and updating the firmware of VFR removal. The issue can occur even when application is exited without deleting VFR ports. Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Shahaji Bhosle <sbhosle@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-11 01:55:48 +00:00
/* Stop the child representors for this device */
ethdev: change stop operation callback to return int Change eth_dev_stop_t return value from void to int. Make eth_dev_stop_t implementations across all drivers to return negative errno values if case of error conditions. Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2020-10-15 13:30:45 +00:00
ret = bnxt_rep_stop_all(bp);
if (ret != 0)
return ret;
net/bnxt: refactor VFR port clean up When parent VF or PF ports are cleaned up, the child VF representor ports also need to be cleaned up. If not cleaned up, then deleting the parent VF shall result in not cleaning up the hardware rules and updating the firmware of VFR removal. The issue can occur even when application is exited without deleting VFR ports. Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Shahaji Bhosle <sbhosle@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-11 01:55:48 +00:00
net/bnxt: fix port stop process and cleanup resources The port deinitialization now cleans up all the resources properly. If all the ports are stopped then ULP context is freed. Added fix to update the correct tfp pointer in the ULP context with the changes to support multi control channels. Fixes: 70e64b27af5b ("net/bnxt: support ULP session manager cleanup") Cc: stable@dpdk.org Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Shahaji Bhosle <sbhosle@broadcom.com>
2020-09-11 01:55:39 +00:00
/* delete the bnxt ULP port details */
bnxt_ulp_port_deinit(bp);
net/bnxt: fix port default rule create/destroy Currently, the flow_ids of port_to_app/app_to_port & tx_cfa_action for the first port are getting over-written by the second port because these fields are stored in the ulp context which is common across the ports. This patch fixes the problem by having per port structure to store these fields. Fixes: 9f702636d7ba ("net/bnxt: add port default rules for ingress and egress") Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com>
2020-07-23 11:56:33 +00:00
net/bnxt: support periodic FW health monitoring Periodically poll the FW heartbeat register and FW recovery counter registers to check the FW health. Polling frequency will be advertised by the FW in HWRM_ERROR_RECOVERY_QCFG response. Schedule the task upon receiving the async event from FW. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:29 +00:00
bnxt_cancel_fw_health_check(bp);
net/bnxt: fix PTP support for Thor On Thor, Rx timestamp is present in the Rx completion record. Only 32 bits of the timestamp is present in the completion. The driver needs to periodically poll the current 48 bit free running timer using the HWRM_PORT_TS_QUERY command. It can combine the upper 16 bits from the HWRM response with the lower 32 bits in the Rx completion to produce the 48 bit timestamp for the Rx packet. This patch adds an alarm thread to periodically poll the current 48 bit free running timer using the HWRM_PORT_TS_QUERY command. This avoids issuing the hwrm command from the Rx handler. This patch also handles the timer roll over condition. Fixes: 6cbd89f9f3d8 ("net/bnxt: support PTP for Thor") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-02-24 15:55:53 +00:00
if (BNXT_P5_PTP_TIMESYNC_ENABLED(bp))
bnxt_cancel_ptp_alarm(bp);
net/bnxt: fix link status during device recovery Driver should not send the phy_cfg request to bring link down during reset recovery. If the driver sends the phy_cfg request in recovery process, then FW needs to re-establish the link which in turn increases the recovery time based on PHY type and link partners. Fixes: df6cd7c1f73a ("net/bnxt: handle reset notify async event from FW") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 05:34:16 +00:00
/* Do not bring link down during reset recovery */
net/bnxt: fix link update 1. When port is stopped, we can forcibly set the link status for the device to down. 2. VFs and MH PFs do not have the privilege to bring the link down. As a result driver prints "Link Up" when port is stopped. 3. When driver receives link status/speed/config async event from fw, driver invokes bnxt_link_update() with exp_link_status as ETH_LINK_UP This is not logically correct as the async event could be for Link up or link down or for speed change. Fixes: 074cacb9907a ("net/bnxt: fix link during port toggle") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-10-06 16:01:56 +00:00
if (!is_bnxt_in_error(bp)) {
net/bnxt: fix link status during device recovery Driver should not send the phy_cfg request to bring link down during reset recovery. If the driver sends the phy_cfg request in recovery process, then FW needs to re-establish the link which in turn increases the recovery time based on PHY type and link partners. Fixes: df6cd7c1f73a ("net/bnxt: handle reset notify async event from FW") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 05:34:16 +00:00
bnxt_dev_set_link_down_op(eth_dev);
net/bnxt: fix link update 1. When port is stopped, we can forcibly set the link status for the device to down. 2. VFs and MH PFs do not have the privilege to bring the link down. As a result driver prints "Link Up" when port is stopped. 3. When driver receives link status/speed/config async event from fw, driver invokes bnxt_link_update() with exp_link_status as ETH_LINK_UP This is not logically correct as the async event could be for Link up or link down or for speed change. Fixes: 074cacb9907a ("net/bnxt: fix link during port toggle") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-10-06 16:01:56 +00:00
/* Wait for link to be reset */
if (BNXT_SINGLE_PF(bp))
rte_delay_ms(500);
/* clear the recorded link status */
memset(&link, 0, sizeof(link));
rte_eth_linkstatus_set(eth_dev, &link);
}
net/bnxt: fix enabling/disabling interrupts 1. Disable interrupts in dev_stop_op() 2. Enable interrupts in dev_start_op() 3. Clean queue intr-vector mapping in dev_stop_op() and thus fix a possible memory leak. Fixes: c09f57b49c13 ("net/bnxt: add start/stop/link update operations") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-07-17 10:41:35 +00:00
/* Clean queue intr-vector mapping */
rte_intr_efd_disable(intr_handle);
if (intr_handle->intr_vec != NULL) {
rte_free(intr_handle->intr_vec);
intr_handle->intr_vec = NULL;
}
net/bnxt: support xstats get/reset This patch adds support to get and reset xstats dev_ops dev_ops added: xstats_get, xstats_get_name, xstats_reset HWRM commands added: hwrm_port_qstats, hwrm_port_clr_stats Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:04 +00:00
bnxt_hwrm_port_clr_stats(bp);
net/bnxt: fix Rx mbuf and agg ring leak in dev stop In the start/stop_op operation, mbufs allocated for rings were not freed 1) add bnxt_free_tx_mbuf/bnxt_free_rx_mbuf in bnxt_dev_stop_op to free MBUF before freeing the rings. 2) MBUF allocation and free routines were not in sync. Allocation uses the ring->ring_size including any rounded up and multiple factors while the free routine uses the requested queue size. Fixes: c09f57b49c13 ("net/bnxt: add start/stop/link update operations") Cc: stable@dpdk.org Signed-off-by: Jay Ding <jay.ding@broadcom.com> Signed-off-by: Scott Branden <scott.branden@broadcom.com> Reviewed-by: Ray Jui <ray.jui@broadcom.com> Reviewed-by: Randy Schacher <stuart.schacher@broadcom.com> Signed-off-by: Xiaoxin Peng <xiaoxin.peng@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-05-01 01:06:14 +00:00
bnxt_free_tx_mbufs(bp);
bnxt_free_rx_mbufs(bp);
net/bnxt: fix async link handling and update When updating the link because of an async link notification there is no need to set wait_for_completion. At this point the link related information should be available without need to poll. Use rte_eth_linkstatus_set instead of memcpy to ensure atomicity while updating the link status. We force the physical link down as a part of device stop. But we are not waiting there enough and handling the async notification before exiting. It just sits in the default CQ till we do a device start. Fix it by calling the CQ handler in device stop. Fixes: 7bc8e9a227cc ("net/bnxt: support async link notification") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com>
2019-10-02 17:17:34 +00:00
/* Process any remaining notifications in default completion queue */
bnxt_int_handler(eth_dev);
net/bnxt: fix crash when closing This patch fixes segfault encountered during dev_uninit/close routine. KNI sample app can be used to reproduce the issue. Fixes: c09f57b49c13 ("net/bnxt: add start/stop/link update operations") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2016-09-29 17:03:44 +00:00
bnxt_shutdown_nic(bp);
net/bnxt: use true/false for bool types Fixes: 0b533591238f ("net/bnxt: inform firmware about IF state changes") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-04-21 21:33:47 +00:00
bnxt_hwrm_if_change(bp, false);
net/bnxt: fix a memory leak in port stop The memory for mark table is allocated during port start. But the allocated memory is freed only during port close or driver unload which in turn causes a memory leakage on each port start/stop. Fixed it by moving the memory free to port stop. Fixes: 94eb699bc82e ("net/bnxt: support flow mark action") Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-01-16 13:04:52 +00:00
rte_free(bp->mark_table);
bp->mark_table = NULL;
net/bnxt: support flow mark action Add support for RTE_FLOW_ACTION_TYPE_MARK. Use the flow_id provided by FW during flow creation to lookup the mark id provided by the application. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com>
2019-12-21 02:38:49 +00:00
bp->flags &= ~BNXT_FLAG_RX_VECTOR_PKT_MODE;
net/bnxt: fix VNIC allocation on port toggle bnxt_init_chip called on port start keeps incrementing rx_cosq_cnt. Hence more vnics are allocated with less number of rings on each port stop/start operation. Eventually vnic allocation fails due to incorrect ring group as no ring gets allocated to the vnic. Fixes: 84d49664b5b2 ("net/bnxt: support CoS classification") Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-11-04 10:02:38 +00:00
bp->rx_cosq_cnt = 0;
net/bnxt: fix allocation of flow stat related structs Consolidate flow stat related structs for performance improvement. The intention of this patch is to reduce the size struct bnxt which had grown because of recent changes and was impacting performance. Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 08:31:59 +00:00
/* All filters are deleted on a port stop. */
if (BNXT_FLOW_XSTATS_EN(bp))
bp->flow_stat->flow_count = 0;
ethdev: change stop operation callback to return int Change eth_dev_stop_t return value from void to int. Make eth_dev_stop_t implementations across all drivers to return negative errno values if case of error conditions. Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2020-10-15 13:30:45 +00:00
return 0;
net/bnxt: fix crash when closing This patch fixes segfault encountered during dev_uninit/close routine. KNI sample app can be used to reproduce the issue. Fixes: c09f57b49c13 ("net/bnxt: add start/stop/link update operations") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2016-09-29 17:03:44 +00:00
}
net/bnxt: check chip reset in stop and close While the error recovery thread is running, an application can invoke dev_stop or dev_close_op thus triggering a race and unwanted consequences if dev_close is invoked while the recovery is not yet completed. Fix by having another lock to synchronize between the 2 threads and return EGAIN if adapter is in the middle of recovery when dev_stop or dev_close ops are invoked Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-24 09:37:34 +00:00
/* Unload the driver, release resources */
static int bnxt_dev_stop_op(struct rte_eth_dev *eth_dev)
{
struct bnxt *bp = eth_dev->data->dev_private;
pthread_mutex_lock(&bp->err_recovery_lock);
if (bp->flags & BNXT_FLAG_FW_RESET) {
PMD_DRV_LOG(ERR,
"Adapter recovering from error..Please retry\n");
net/bnxt: fix lock handling in stop and close err_recovery_lock needs to be released before returning in stop and close_op if FW_RESET flag is set. Fixes: 6f5f3b99821e ("net/bnxt: check chip reset in stop and close") Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-01-14 02:27:16 +00:00
pthread_mutex_unlock(&bp->err_recovery_lock);
net/bnxt: check chip reset in stop and close While the error recovery thread is running, an application can invoke dev_stop or dev_close_op thus triggering a race and unwanted consequences if dev_close is invoked while the recovery is not yet completed. Fix by having another lock to synchronize between the 2 threads and return EGAIN if adapter is in the middle of recovery when dev_stop or dev_close ops are invoked Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-24 09:37:34 +00:00
return -EAGAIN;
}
pthread_mutex_unlock(&bp->err_recovery_lock);
return bnxt_dev_stop(eth_dev);
}
net/bnxt: fix error handling in device start Call bnxt_dev_stop in error path of bnxt_dev_start_op() to keep it simple and consistent Fixes: c09f57b49c13 ("net/bnxt: add start/stop/link update operations") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-24 09:37:33 +00:00
static int bnxt_dev_start_op(struct rte_eth_dev *eth_dev)
{
struct bnxt *bp = eth_dev->data->dev_private;
uint64_t rx_offloads = eth_dev->data->dev_conf.rxmode.offloads;
int vlan_mask = 0;
int rc, retry_cnt = BNXT_IF_CHANGE_RETRY_COUNT;
if (!eth_dev->data->nb_tx_queues || !eth_dev->data->nb_rx_queues) {
PMD_DRV_LOG(ERR, "Queues are not configured yet!\n");
return -EINVAL;
}
if (bp->rx_cp_nr_rings > RTE_ETHDEV_QUEUE_STAT_CNTRS)
PMD_DRV_LOG(ERR,
"RxQ cnt %d > RTE_ETHDEV_QUEUE_STAT_CNTRS %d\n",
bp->rx_cp_nr_rings, RTE_ETHDEV_QUEUE_STAT_CNTRS);
do {
rc = bnxt_hwrm_if_change(bp, true);
if (rc == 0 || rc != -EAGAIN)
break;
rte_delay_ms(BNXT_IF_CHANGE_RETRY_INTERVAL);
} while (retry_cnt--);
if (rc)
return rc;
if (bp->flags & BNXT_FLAG_IF_CHANGE_HOT_FW_RESET_DONE) {
rc = bnxt_handle_if_change_status(bp);
if (rc)
return rc;
}
bnxt_enable_int(bp);
net/bnxt: fix Rx completion ring size calculation The size of the receive completion ring should be recalculated when MTU is increased to a size that requires scattered receive or when LRO is enabled. Move logic for this calculation from the ring configuration path to the device start path. - Made size calculation dependent only on scattered_rx status. - Moved calculation of scattered_rx up in the initialization sequence. - Made LRO offload status part of scattered_rx calculation. When the completion ring size is too small, completion overflows can occur causing the ring to be disabled in hardware. Fixes: 04067844a3e9 ("net/bnxt: reduce CQ queue size without aggregation ring") Cc: stable@dpdk.org Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-01-29 18:07:09 +00:00
eth_dev->data->scattered_rx = bnxt_scattered_rx(eth_dev);
net/bnxt: refactor init/uninit Move all the individual driver fields allocation routines to one routine - bnxt_drv_init(). This houses all such routines where memory needs to be allocated once during the driver's lifetime and does not need to be torn down during error recovery. Rename some function names in accordance with their functionality. bnxt_init_board() is doing nothing more than mapping the PCI bars, so rename it as such. Given that there is a bnxt_shutdown_nic that is called in dev_stop_op, rename it's counterpart - bnxt_init_chip() that is called in dev_start_op, to bnxt_start_nic. Also helps avoid confusion with some of the other bnxt_init_xxx routines. Rename bnxt_init_fw() to bnxt_get_config() as that is what that routine is doing mostly functionality wise. Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-01-14 02:19:14 +00:00
rc = bnxt_start_nic(bp);
net/bnxt: fix error handling in device start Call bnxt_dev_stop in error path of bnxt_dev_start_op() to keep it simple and consistent Fixes: c09f57b49c13 ("net/bnxt: add start/stop/link update operations") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-24 09:37:33 +00:00
if (rc)
goto error;
eth_dev->data->dev_started = 1;
bnxt_link_update_op(eth_dev, 1);
if (rx_offloads & DEV_RX_OFFLOAD_VLAN_FILTER)
vlan_mask |= ETH_VLAN_FILTER_MASK;
if (rx_offloads & DEV_RX_OFFLOAD_VLAN_STRIP)
vlan_mask |= ETH_VLAN_STRIP_MASK;
rc = bnxt_vlan_offload_set_op(eth_dev, vlan_mask);
if (rc)
goto error;
/* Initialize bnxt ULP port details */
rc = bnxt_ulp_port_init(bp);
if (rc)
goto error;
eth_dev->rx_pkt_burst = bnxt_receive_function(eth_dev);
eth_dev->tx_pkt_burst = bnxt_transmit_function(eth_dev);
bnxt_schedule_fw_health_check(bp);
net/bnxt: fix PTP support for Thor On Thor, Rx timestamp is present in the Rx completion record. Only 32 bits of the timestamp is present in the completion. The driver needs to periodically poll the current 48 bit free running timer using the HWRM_PORT_TS_QUERY command. It can combine the upper 16 bits from the HWRM response with the lower 32 bits in the Rx completion to produce the 48 bit timestamp for the Rx packet. This patch adds an alarm thread to periodically poll the current 48 bit free running timer using the HWRM_PORT_TS_QUERY command. This avoids issuing the hwrm command from the Rx handler. This patch also handles the timer roll over condition. Fixes: 6cbd89f9f3d8 ("net/bnxt: support PTP for Thor") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-02-24 15:55:53 +00:00
if (BNXT_P5_PTP_TIMESYNC_ENABLED(bp))
bnxt_schedule_ptp_alarm(bp);
net/bnxt: fix error handling in device start Call bnxt_dev_stop in error path of bnxt_dev_start_op() to keep it simple and consistent Fixes: c09f57b49c13 ("net/bnxt: add start/stop/link update operations") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-24 09:37:33 +00:00
return 0;
error:
net/bnxt: check chip reset in stop and close While the error recovery thread is running, an application can invoke dev_stop or dev_close_op thus triggering a race and unwanted consequences if dev_close is invoked while the recovery is not yet completed. Fix by having another lock to synchronize between the 2 threads and return EGAIN if adapter is in the middle of recovery when dev_stop or dev_close ops are invoked Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-24 09:37:34 +00:00
bnxt_dev_stop(eth_dev);
net/bnxt: fix error handling in device start Call bnxt_dev_stop in error path of bnxt_dev_start_op() to keep it simple and consistent Fixes: c09f57b49c13 ("net/bnxt: add start/stop/link update operations") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-24 09:37:33 +00:00
return rc;
}
net/bnxt: fix lock init and destroy Invoking init/uninit locks in init_resources and uninit_resources would end up initializing and destroying locks on every port start stop which is not desired. Move the 2 routines to dev_init and dev_close respectively as locks need to be initialized and destroyed only once during the lifetime of the driver. Fixes: 1cb3d39a48f7 ("net/bnxt: synchronize between flow related functions") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-24 09:37:32 +00:00
static void
bnxt_uninit_locks(struct bnxt *bp)
{
pthread_mutex_destroy(&bp->flow_lock);
pthread_mutex_destroy(&bp->def_cp_lock);
pthread_mutex_destroy(&bp->health_check_lock);
net/bnxt: check chip reset in stop and close While the error recovery thread is running, an application can invoke dev_stop or dev_close_op thus triggering a race and unwanted consequences if dev_close is invoked while the recovery is not yet completed. Fix by having another lock to synchronize between the 2 threads and return EGAIN if adapter is in the middle of recovery when dev_stop or dev_close ops are invoked Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-24 09:37:34 +00:00
pthread_mutex_destroy(&bp->err_recovery_lock);
net/bnxt: fix lock init and destroy Invoking init/uninit locks in init_resources and uninit_resources would end up initializing and destroying locks on every port start stop which is not desired. Move the 2 routines to dev_init and dev_close respectively as locks need to be initialized and destroyed only once during the lifetime of the driver. Fixes: 1cb3d39a48f7 ("net/bnxt: synchronize between flow related functions") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-24 09:37:32 +00:00
if (bp->rep_info) {
pthread_mutex_destroy(&bp->rep_info->vfr_lock);
pthread_mutex_destroy(&bp->rep_info->vfr_start_lock);
}
}
net/bnxt: refactor init/uninit Move all the individual driver fields allocation routines to one routine - bnxt_drv_init(). This houses all such routines where memory needs to be allocated once during the driver's lifetime and does not need to be torn down during error recovery. Rename some function names in accordance with their functionality. bnxt_init_board() is doing nothing more than mapping the PCI bars, so rename it as such. Given that there is a bnxt_shutdown_nic that is called in dev_stop_op, rename it's counterpart - bnxt_init_chip() that is called in dev_start_op, to bnxt_start_nic. Also helps avoid confusion with some of the other bnxt_init_xxx routines. Rename bnxt_init_fw() to bnxt_get_config() as that is what that routine is doing mostly functionality wise. Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-01-14 02:19:14 +00:00
static void bnxt_drv_uninit(struct bnxt *bp)
{
bnxt_free_leds_info(bp);
bnxt_free_cos_queues(bp);
bnxt_free_link_info(bp);
bnxt_free_parent_info(bp);
bnxt_uninit_locks(bp);
rte_memzone_free((const struct rte_memzone *)bp->tx_mem_zone);
bp->tx_mem_zone = NULL;
rte_memzone_free((const struct rte_memzone *)bp->rx_mem_zone);
bp->rx_mem_zone = NULL;
net/bnxt: fix VF info allocation 1. Renamed bnxt_hwrm_alloc_vf_info()/bnxt_hwrm_free_vf_info to bnxt_alloc_vf_info()/bnxt_free_vf_info as it does not issue any HWRM command to fw. 2. Fix missing unlock when memory allocation fails. Fixes: b7778e8a1c00 ("net/bnxt: refactor to properly allocate resources for PF/VF") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-03-04 09:07:28 +00:00
bnxt_free_vf_info(bp);
net/bnxt: fix resource cleanup Fix resource cleanup in port close. Once the pointers are freed, set them to NULL. Make sure access to the pointers is validated before use. Fixes: bb81e07323bb ("net/bnxt: support LED on/off") Fixes: 804e746c7b73 ("net/bnxt: add hardware resource manager init code") Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Fixes: 698aa7e95325 ("net/bnxt: add code to determine the Tx COS queue") Fixes: 322bd6e70272 ("net/bnxt: add port representor infrastructure") Fixes: 0bf5a0b5ebb8 ("net/bnxt: add a failure log") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-04-08 20:17:47 +00:00
bnxt_free_pf_info(bp);
net/bnxt: refactor init/uninit Move all the individual driver fields allocation routines to one routine - bnxt_drv_init(). This houses all such routines where memory needs to be allocated once during the driver's lifetime and does not need to be torn down during error recovery. Rename some function names in accordance with their functionality. bnxt_init_board() is doing nothing more than mapping the PCI bars, so rename it as such. Given that there is a bnxt_shutdown_nic that is called in dev_stop_op, rename it's counterpart - bnxt_init_chip() that is called in dev_start_op, to bnxt_start_nic. Also helps avoid confusion with some of the other bnxt_init_xxx routines. Rename bnxt_init_fw() to bnxt_get_config() as that is what that routine is doing mostly functionality wise. Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-01-14 02:19:14 +00:00
rte_free(bp->grp_info);
bp->grp_info = NULL;
}
ethdev: allow drivers to return error on close The device operation .dev_close was returning void. This driver interface is changed to return an int. Note that the API rte_eth_dev_close() is still returning void, although a deprecation notice is pending to change it as well. Signed-off-by: Thomas Monjalon <thomas@monjalon.net> Reviewed-by: Rosen Xu <rosen.xu@intel.com> Reviewed-by: Sachin Saxena <sachin.saxena@oss.nxp.com> Reviewed-by: Liron Himi <lironh@marvell.com> Reviewed-by: Haiyue Wang <haiyue.wang@intel.com> Acked-by: Jeff Guo <jia.guo@intel.com> Reviewed-by: Andrew Rybchenko <arybchenko@solarflare.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2020-09-28 23:14:10 +00:00
static int bnxt_dev_close_op(struct rte_eth_dev *eth_dev)
net/bnxt: free memory in close operation This patch adds code to free all resources except the one corresponding to HWRM, which are required to notify the HWRM that the driver is unloaded (these are freed in uninit()). Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:33 +00:00
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = eth_dev->data->dev_private;
ethdev: change stop operation callback to return int Change eth_dev_stop_t return value from void to int. Make eth_dev_stop_t implementations across all drivers to return negative errno values if case of error conditions. Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2020-10-15 13:30:45 +00:00
int ret = 0;
net/bnxt: free memory in close operation This patch adds code to free all resources except the one corresponding to HWRM, which are required to notify the HWRM that the driver is unloaded (these are freed in uninit()). Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:33 +00:00
drivers/net: check process type in close operation The secondary processes are not allowed to release shared resources. Only process-private resources should be freed in a secondary process. Most of the time, there is no process-private resource, so the close operation is just forbidden in a secondary process. After adding proper check in the port close functions, some redundant checks in the device remove functions are dropped. Signed-off-by: Thomas Monjalon <thomas@monjalon.net> Reviewed-by: Rosen Xu <rosen.xu@intel.com> Reviewed-by: Sachin Saxena <sachin.saxena@oss.nxp.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Liron Himi <lironh@marvell.com> Reviewed-by: Haiyue Wang <haiyue.wang@intel.com> Acked-by: Jeff Guo <jia.guo@intel.com> Reviewed-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2020-09-28 23:14:31 +00:00
if (rte_eal_process_type() != RTE_PROC_PRIMARY)
return 0;
net/bnxt: check chip reset in stop and close While the error recovery thread is running, an application can invoke dev_stop or dev_close_op thus triggering a race and unwanted consequences if dev_close is invoked while the recovery is not yet completed. Fix by having another lock to synchronize between the 2 threads and return EGAIN if adapter is in the middle of recovery when dev_stop or dev_close ops are invoked Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-24 09:37:34 +00:00
pthread_mutex_lock(&bp->err_recovery_lock);
if (bp->flags & BNXT_FLAG_FW_RESET) {
PMD_DRV_LOG(ERR,
"Adapter recovering from error...Please retry\n");
net/bnxt: fix lock handling in stop and close err_recovery_lock needs to be released before returning in stop and close_op if FW_RESET flag is set. Fixes: 6f5f3b99821e ("net/bnxt: check chip reset in stop and close") Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-01-14 02:27:16 +00:00
pthread_mutex_unlock(&bp->err_recovery_lock);
net/bnxt: check chip reset in stop and close While the error recovery thread is running, an application can invoke dev_stop or dev_close_op thus triggering a race and unwanted consequences if dev_close is invoked while the recovery is not yet completed. Fix by having another lock to synchronize between the 2 threads and return EGAIN if adapter is in the middle of recovery when dev_stop or dev_close ops are invoked Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-24 09:37:34 +00:00
return -EAGAIN;
}
pthread_mutex_unlock(&bp->err_recovery_lock);
net/bnxt: fix recovery alarm race condition in port close There is a race condition when port is closed while error recovery is happening in driver. Fixed this by canceling the recovery threads during port close. Fixes: df6cd7c1f73a ("net/bnxt: handle reset notify async event from FW") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com>
2020-02-28 10:21:31 +00:00
/* cancel the recovery handler before remove dev */
rte_eal_alarm_cancel(bnxt_dev_reset_and_resume, (void *)bp);
rte_eal_alarm_cancel(bnxt_dev_recover, (void *)bp);
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
bnxt_cancel_fc_thread(bp);
net/bnxt: fix recovery alarm race condition in port close There is a race condition when port is closed while error recovery is happening in driver. Fixed this by canceling the recovery threads during port close. Fixes: df6cd7c1f73a ("net/bnxt: handle reset notify async event from FW") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com>
2020-02-28 10:21:31 +00:00
net/bnxt: remove a redundant variable Use "dev->data->dev_started" state, instead of local "dev_stopped" to check whether port has been started or not. Fixes: 316e412299fd ("net/bnxt: fix crash when closing") Cc: stable@dpdk.org Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
2020-01-28 07:29:21 +00:00
if (eth_dev->data->dev_started)
net/bnxt: check chip reset in stop and close While the error recovery thread is running, an application can invoke dev_stop or dev_close_op thus triggering a race and unwanted consequences if dev_close is invoked while the recovery is not yet completed. Fix by having another lock to synchronize between the 2 threads and return EGAIN if adapter is in the middle of recovery when dev_stop or dev_close ops are invoked Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-24 09:37:34 +00:00
ret = bnxt_dev_stop(eth_dev);
net/bnxt: fix crash when closing This patch fixes segfault encountered during dev_uninit/close routine. KNI sample app can be used to reproduce the issue. Fixes: c09f57b49c13 ("net/bnxt: add start/stop/link update operations") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2016-09-29 17:03:44 +00:00
net/bnxt: release port upon close Set RTE_ETH_DEV_CLOSE_REMOVE upon probe so all the private resources for the port can be freed by rte_eth_dev_close(). Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-01-16 13:04:54 +00:00
bnxt_uninit_resources(bp, false);
net/bnxt: fix close operation We are not cleaning up all the memory and also not unregistering the driver during device close operation. This patch fixes the issue. Fixes: 893074951314 ("net/bnxt: free memory in close operation") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-06-28 20:15:31 +00:00
net/bnxt: refactor init/uninit Move all the individual driver fields allocation routines to one routine - bnxt_drv_init(). This houses all such routines where memory needs to be allocated once during the driver's lifetime and does not need to be torn down during error recovery. Rename some function names in accordance with their functionality. bnxt_init_board() is doing nothing more than mapping the PCI bars, so rename it as such. Given that there is a bnxt_shutdown_nic that is called in dev_stop_op, rename it's counterpart - bnxt_init_chip() that is called in dev_start_op, to bnxt_start_nic. Also helps avoid confusion with some of the other bnxt_init_xxx routines. Rename bnxt_init_fw() to bnxt_get_config() as that is what that routine is doing mostly functionality wise. Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-01-14 02:19:14 +00:00
bnxt_drv_uninit(bp);
ethdev: allow drivers to return error on close The device operation .dev_close was returning void. This driver interface is changed to return an int. Note that the API rte_eth_dev_close() is still returning void, although a deprecation notice is pending to change it as well. Signed-off-by: Thomas Monjalon <thomas@monjalon.net> Reviewed-by: Rosen Xu <rosen.xu@intel.com> Reviewed-by: Sachin Saxena <sachin.saxena@oss.nxp.com> Reviewed-by: Liron Himi <lironh@marvell.com> Reviewed-by: Haiyue Wang <haiyue.wang@intel.com> Acked-by: Jeff Guo <jia.guo@intel.com> Reviewed-by: Andrew Rybchenko <arybchenko@solarflare.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2020-09-28 23:14:10 +00:00
ethdev: change stop operation callback to return int Change eth_dev_stop_t return value from void to int. Make eth_dev_stop_t implementations across all drivers to return negative errno values if case of error conditions. Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2020-10-15 13:30:45 +00:00
return ret;
net/bnxt: free memory in close operation This patch adds code to free all resources except the one corresponding to HWRM, which are required to notify the HWRM that the driver is unloaded (these are freed in uninit()). Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:33 +00:00
}
net/bnxt: add MAC address This patch adds dev_ops to Add/Remove MAC addresses. These use the bnxt_hwrm_set_filter() defined previously Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:34 +00:00
static void bnxt_mac_addr_remove_op(struct rte_eth_dev *eth_dev,
uint32_t index)
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = eth_dev->data->dev_private;
net/bnxt: add MAC address This patch adds dev_ops to Add/Remove MAC addresses. These use the bnxt_hwrm_set_filter() defined previously Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:34 +00:00
uint64_t pool_mask = eth_dev->data->mac_pool_sel[index];
struct bnxt_vnic_info *vnic;
struct bnxt_filter_info *filter, *temp_filter;
net/bnxt: fix a bit shift operation We are left shifting more bits than we should be doing. This patch fixes that. Coverity issue: 127546 Fixes: 778b759ba10e ("net/bnxt: add MAC address") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-10-24 21:19:45 +00:00
uint32_t i;
net/bnxt: add MAC address This patch adds dev_ops to Add/Remove MAC addresses. These use the bnxt_hwrm_set_filter() defined previously Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:34 +00:00
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
if (is_bnxt_in_error(bp))
return;
net/bnxt: add MAC address This patch adds dev_ops to Add/Remove MAC addresses. These use the bnxt_hwrm_set_filter() defined previously Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:34 +00:00
/*
* Loop through all VNICs from the specified filter flow pools to
* remove the corresponding MAC addr filter
*/
net/bnxt: get rid of ff pools and use VNIC info array There was no direct association between the rxq's VNIC and the vnic_info[]. Explicitly associate the two in bnxt_mq_rx_configure(). Fixes: 0a256e4a548b ("net/bnxt: fix Rx ring count limitation") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
2018-09-29 01:59:52 +00:00
for (i = 0; i < bp->nr_vnics; i++) {
net/bnxt: fix bit shift size Some(?) compilers will treat the unmarked constant 1 as a 32-bit integer, but the shift operation is in a loop that could run up to 63 times -- undefined behavior! Coverity issue: 127546 Fixes: 778b759ba10e ("net/bnxt: add MAC address") Signed-off-by: John W. Linville <linville@tuxdriver.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2016-09-29 17:39:36 +00:00
if (!(pool_mask & (1ULL << i)))
net/bnxt: add MAC address This patch adds dev_ops to Add/Remove MAC addresses. These use the bnxt_hwrm_set_filter() defined previously Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:34 +00:00
continue;
net/bnxt: get rid of ff pools and use VNIC info array There was no direct association between the rxq's VNIC and the vnic_info[]. Explicitly associate the two in bnxt_mq_rx_configure(). Fixes: 0a256e4a548b ("net/bnxt: fix Rx ring count limitation") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
2018-09-29 01:59:52 +00:00
vnic = &bp->vnic_info[i];
filter = STAILQ_FIRST(&vnic->filter);
while (filter) {
temp_filter = STAILQ_NEXT(filter, next);
if (filter->mac_index == index) {
STAILQ_REMOVE(&vnic->filter, filter,
bnxt_filter_info, next);
bnxt_hwrm_clear_l2_filter(bp, filter);
net/bnxt: fix redundant MAC address check filter->mac_index is used to check, if a same mac is already programmed. Hence, filter->dflt member is not needed which is also used for mac addr redundancy check. This patch fixes it by moving mac_index based redundant check from bnxt_mac_addr_add_op to bnxt_add_mac_filter Fixes: 6118503d8071 ("net/bnxt: fix VLAN filtering") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-11-04 10:02:37 +00:00
bnxt_free_filter(bp, filter);
net/bnxt: add MAC address This patch adds dev_ops to Add/Remove MAC addresses. These use the bnxt_hwrm_set_filter() defined previously Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:34 +00:00
}
net/bnxt: get rid of ff pools and use VNIC info array There was no direct association between the rxq's VNIC and the vnic_info[]. Explicitly associate the two in bnxt_mq_rx_configure(). Fixes: 0a256e4a548b ("net/bnxt: fix Rx ring count limitation") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
2018-09-29 01:59:52 +00:00
filter = temp_filter;
net/bnxt: add MAC address This patch adds dev_ops to Add/Remove MAC addresses. These use the bnxt_hwrm_set_filter() defined previously Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:34 +00:00
}
}
}
net/bnxt: fix VLAN filtering Currently, when hw-vlan-filter is enabled on testpmd, driver is receiving all vlan packets. Instead, it should only receive untagged packets and vlan packets for which the VLAN filter is programmed. This is because, the default rule to match on MAC is not getting deleted, when hw-vlan-filter is ON. This patch fixes the problem, by deleting the default MAC rule and programming a new rule to receive only untagged packets, when hw-vlan-filter is enabled & another rule for each vlan, as and when that vlan is configured on that port. Fixes: 246c5cc5f05e ("net/bnxt: use correct flags during VLAN configuration") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 23:25:59 +00:00
static int bnxt_add_mac_filter(struct bnxt *bp, struct bnxt_vnic_info *vnic,
net/bnxt: fix redundant MAC address check filter->mac_index is used to check, if a same mac is already programmed. Hence, filter->dflt member is not needed which is also used for mac addr redundancy check. This patch fixes it by moving mac_index based redundant check from bnxt_mac_addr_add_op to bnxt_add_mac_filter Fixes: 6118503d8071 ("net/bnxt: fix VLAN filtering") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-11-04 10:02:37 +00:00
struct rte_ether_addr *mac_addr, uint32_t index,
uint32_t pool)
net/bnxt: fix VLAN filtering Currently, when hw-vlan-filter is enabled on testpmd, driver is receiving all vlan packets. Instead, it should only receive untagged packets and vlan packets for which the VLAN filter is programmed. This is because, the default rule to match on MAC is not getting deleted, when hw-vlan-filter is ON. This patch fixes the problem, by deleting the default MAC rule and programming a new rule to receive only untagged packets, when hw-vlan-filter is enabled & another rule for each vlan, as and when that vlan is configured on that port. Fixes: 246c5cc5f05e ("net/bnxt: use correct flags during VLAN configuration") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 23:25:59 +00:00
{
struct bnxt_filter_info *filter;
int rc = 0;
net/bnxt: fix redundant MAC address check filter->mac_index is used to check, if a same mac is already programmed. Hence, filter->dflt member is not needed which is also used for mac addr redundancy check. This patch fixes it by moving mac_index based redundant check from bnxt_mac_addr_add_op to bnxt_add_mac_filter Fixes: 6118503d8071 ("net/bnxt: fix VLAN filtering") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-11-04 10:02:37 +00:00
/* Attach requested MAC address to the new l2_filter */
STAILQ_FOREACH(filter, &vnic->filter, next) {
if (filter->mac_index == index) {
net/bnxt: fix log message level When an existing mac_addr is tried to get programmed again, a message is displayed that the mac_addr already exists. However the message is of type ERR. This patch changes the message to type DEBUG Fixes: 938a87db4324 ("net/bnxt: fix redundant MAC address check") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-11-13 08:29:44 +00:00
PMD_DRV_LOG(DEBUG,
net/bnxt: fix redundant MAC address check filter->mac_index is used to check, if a same mac is already programmed. Hence, filter->dflt member is not needed which is also used for mac addr redundancy check. This patch fixes it by moving mac_index based redundant check from bnxt_mac_addr_add_op to bnxt_add_mac_filter Fixes: 6118503d8071 ("net/bnxt: fix VLAN filtering") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-11-04 10:02:37 +00:00
"MAC addr already existed for pool %d\n",
pool);
return 0;
}
}
net/bnxt: fix VLAN filtering Currently, when hw-vlan-filter is enabled on testpmd, driver is receiving all vlan packets. Instead, it should only receive untagged packets and vlan packets for which the VLAN filter is programmed. This is because, the default rule to match on MAC is not getting deleted, when hw-vlan-filter is ON. This patch fixes the problem, by deleting the default MAC rule and programming a new rule to receive only untagged packets, when hw-vlan-filter is enabled & another rule for each vlan, as and when that vlan is configured on that port. Fixes: 246c5cc5f05e ("net/bnxt: use correct flags during VLAN configuration") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 23:25:59 +00:00
filter = bnxt_alloc_filter(bp);
if (!filter) {
PMD_DRV_LOG(ERR, "L2 filter alloc failed\n");
return -ENODEV;
}
/* bnxt_alloc_filter copies default MAC to filter->l2_addr. So,
* if the MAC that's been programmed now is a different one, then,
* copy that addr to filter->l2_addr
*/
if (mac_addr)
memcpy(filter->l2_addr, mac_addr, RTE_ETHER_ADDR_LEN);
filter->flags |= HWRM_CFA_L2_FILTER_ALLOC_INPUT_FLAGS_OUTERMOST;
rc = bnxt_hwrm_set_l2_filter(bp, vnic->fw_vnic_id, filter);
if (!rc) {
net/bnxt: fix redundant MAC address check filter->mac_index is used to check, if a same mac is already programmed. Hence, filter->dflt member is not needed which is also used for mac addr redundancy check. This patch fixes it by moving mac_index based redundant check from bnxt_mac_addr_add_op to bnxt_add_mac_filter Fixes: 6118503d8071 ("net/bnxt: fix VLAN filtering") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-11-04 10:02:37 +00:00
filter->mac_index = index;
if (filter->mac_index == 0)
net/bnxt: fix VLAN filtering Currently, when hw-vlan-filter is enabled on testpmd, driver is receiving all vlan packets. Instead, it should only receive untagged packets and vlan packets for which the VLAN filter is programmed. This is because, the default rule to match on MAC is not getting deleted, when hw-vlan-filter is ON. This patch fixes the problem, by deleting the default MAC rule and programming a new rule to receive only untagged packets, when hw-vlan-filter is enabled & another rule for each vlan, as and when that vlan is configured on that port. Fixes: 246c5cc5f05e ("net/bnxt: use correct flags during VLAN configuration") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 23:25:59 +00:00
STAILQ_INSERT_HEAD(&vnic->filter, filter, next);
net/bnxt: fix redundant MAC address check filter->mac_index is used to check, if a same mac is already programmed. Hence, filter->dflt member is not needed which is also used for mac addr redundancy check. This patch fixes it by moving mac_index based redundant check from bnxt_mac_addr_add_op to bnxt_add_mac_filter Fixes: 6118503d8071 ("net/bnxt: fix VLAN filtering") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-11-04 10:02:37 +00:00
else
net/bnxt: fix VLAN filtering Currently, when hw-vlan-filter is enabled on testpmd, driver is receiving all vlan packets. Instead, it should only receive untagged packets and vlan packets for which the VLAN filter is programmed. This is because, the default rule to match on MAC is not getting deleted, when hw-vlan-filter is ON. This patch fixes the problem, by deleting the default MAC rule and programming a new rule to receive only untagged packets, when hw-vlan-filter is enabled & another rule for each vlan, as and when that vlan is configured on that port. Fixes: 246c5cc5f05e ("net/bnxt: use correct flags during VLAN configuration") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 23:25:59 +00:00
STAILQ_INSERT_TAIL(&vnic->filter, filter, next);
} else {
bnxt_free_filter(bp, filter);
}
return rc;
}
ethdev: fix adding invalid MAC address Some customers find adding MAC addr to VF sometimes can fail, but it is still stored in dev->data->mac_addrs[ ]. So this can lead to some errors that assumes the non-zero entry in dev->data->mac_addrs[ ] is valid. Following acknowledgements are from specific NIC PMD maintainer for their managing part. This patch changes the ethdev internal API, it should not be backported to a stable/LTS release so far. Fixes: af75078fece3 ("first public release") Signed-off-by: Wei Dai <wei.dai@intel.com> Acked-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com> Acked-by: Yuanhan Liu <yuanhan.liu@linux.intel.com> Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
2017-05-05 00:40:00 +00:00
static int bnxt_mac_addr_add_op(struct rte_eth_dev *eth_dev,
net: add rte prefix to ether structures Add 'rte_' prefix to structures: - rename struct ether_addr as struct rte_ether_addr. - rename struct ether_hdr as struct rte_ether_hdr. - rename struct vlan_hdr as struct rte_vlan_hdr. - rename struct vxlan_hdr as struct rte_vxlan_hdr. - rename struct vxlan_gpe_hdr as struct rte_vxlan_gpe_hdr. Do not update the command line library to avoid adding a dependency to librte_net. Signed-off-by: Olivier Matz <olivier.matz@6wind.com> Reviewed-by: Stephen Hemminger <stephen@networkplumber.org> Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2019-05-21 16:13:03 +00:00
struct rte_ether_addr *mac_addr,
ethdev: fix adding invalid MAC address Some customers find adding MAC addr to VF sometimes can fail, but it is still stored in dev->data->mac_addrs[ ]. So this can lead to some errors that assumes the non-zero entry in dev->data->mac_addrs[ ] is valid. Following acknowledgements are from specific NIC PMD maintainer for their managing part. This patch changes the ethdev internal API, it should not be backported to a stable/LTS release so far. Fixes: af75078fece3 ("first public release") Signed-off-by: Wei Dai <wei.dai@intel.com> Acked-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com> Acked-by: Yuanhan Liu <yuanhan.liu@linux.intel.com> Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
2017-05-05 00:40:00 +00:00
uint32_t index, uint32_t pool)
net/bnxt: add MAC address This patch adds dev_ops to Add/Remove MAC addresses. These use the bnxt_hwrm_set_filter() defined previously Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:34 +00:00
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = eth_dev->data->dev_private;
net/bnxt: get rid of ff pools and use VNIC info array There was no direct association between the rxq's VNIC and the vnic_info[]. Explicitly associate the two in bnxt_mq_rx_configure(). Fixes: 0a256e4a548b ("net/bnxt: fix Rx ring count limitation") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
2018-09-29 01:59:52 +00:00
struct bnxt_vnic_info *vnic = &bp->vnic_info[pool];
net/bnxt: fix adding MAC address HWRM command to add MAC address can fail. Driver should check the return value of HWRM command and do the house keeping properly. Fixes: 778b759ba10e ("net/bnxt: add MAC address") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-18 03:35:57 +00:00
int rc = 0;
net/bnxt: add MAC address This patch adds dev_ops to Add/Remove MAC addresses. These use the bnxt_hwrm_set_filter() defined previously Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:34 +00:00
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
rc = is_bnxt_in_error(bp);
if (rc)
return rc;
net/bnxt: fix boolean operator usage Use boolean AND operator instead of bitwise operator. Coverity issue: 323488 Fixes: b42c15c83e88 ("net/bnxt: support trusted VF") Cc: stable@dpdk.org Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-10-22 18:45:10 +00:00
if (BNXT_VF(bp) && !BNXT_VF_IS_TRUSTED(bp)) {
net/bnxt: use dynamic log type This patch implements driver specific log type doing away with usage of RTE_LOG() for logging. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:55 +00:00
PMD_DRV_LOG(ERR, "Cannot add MAC address to a VF interface\n");
ethdev: fix adding invalid MAC address Some customers find adding MAC addr to VF sometimes can fail, but it is still stored in dev->data->mac_addrs[ ]. So this can lead to some errors that assumes the non-zero entry in dev->data->mac_addrs[ ] is valid. Following acknowledgements are from specific NIC PMD maintainer for their managing part. This patch changes the ethdev internal API, it should not be backported to a stable/LTS release so far. Fixes: af75078fece3 ("first public release") Signed-off-by: Wei Dai <wei.dai@intel.com> Acked-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com> Acked-by: Yuanhan Liu <yuanhan.liu@linux.intel.com> Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
2017-05-05 00:40:00 +00:00
return -ENOTSUP;
net/bnxt: support VF Add support to the bnxt PMD to load on a PCI VF. 1) VF cannot change parameters like - speed, autoneg and pause 2) If the VF MAC address shows up as all 0's it has to be provisioned by the PF in the hypervisor. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-09-26 16:18:46 +00:00
}
net/bnxt: add MAC address This patch adds dev_ops to Add/Remove MAC addresses. These use the bnxt_hwrm_set_filter() defined previously Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:34 +00:00
if (!vnic) {
net/bnxt: use dynamic log type This patch implements driver specific log type doing away with usage of RTE_LOG() for logging. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:55 +00:00
PMD_DRV_LOG(ERR, "VNIC not found for pool %d!\n", pool);
ethdev: fix adding invalid MAC address Some customers find adding MAC addr to VF sometimes can fail, but it is still stored in dev->data->mac_addrs[ ]. So this can lead to some errors that assumes the non-zero entry in dev->data->mac_addrs[ ] is valid. Following acknowledgements are from specific NIC PMD maintainer for their managing part. This patch changes the ethdev internal API, it should not be backported to a stable/LTS release so far. Fixes: af75078fece3 ("first public release") Signed-off-by: Wei Dai <wei.dai@intel.com> Acked-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com> Acked-by: Yuanhan Liu <yuanhan.liu@linux.intel.com> Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
2017-05-05 00:40:00 +00:00
return -EINVAL;
net/bnxt: add MAC address This patch adds dev_ops to Add/Remove MAC addresses. These use the bnxt_hwrm_set_filter() defined previously Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:34 +00:00
}
net/bnxt: fix adding MAC address HWRM command to add MAC address can fail. Driver should check the return value of HWRM command and do the house keeping properly. Fixes: 778b759ba10e ("net/bnxt: add MAC address") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-18 03:35:57 +00:00
net/bnxt: fix MAC address setting when port is stopped Driver destroys the VNIC when the port is brought down. Port HW filter setting such as default MAC address and unicast MAC filters will be applied when port is started. Fixed to return success silently for these callbacks when port is stopped. Fixes: 39b88344e364 ("net/bnxt: fix enable/disable VLAN filtering") Cc: stable@dpdk.org Reported-by: Stephen Hemminger <stephen@networkplumber.org> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-02-27 08:18:51 +00:00
/* Filter settings will get applied when port is started */
if (!eth_dev->data->dev_started)
return 0;
net/bnxt: fix redundant MAC address check filter->mac_index is used to check, if a same mac is already programmed. Hence, filter->dflt member is not needed which is also used for mac addr redundancy check. This patch fixes it by moving mac_index based redundant check from bnxt_mac_addr_add_op to bnxt_add_mac_filter Fixes: 6118503d8071 ("net/bnxt: fix VLAN filtering") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-11-04 10:02:37 +00:00
rc = bnxt_add_mac_filter(bp, vnic, mac_addr, index, pool);
net/bnxt: fix adding MAC address HWRM command to add MAC address can fail. Driver should check the return value of HWRM command and do the house keeping properly. Fixes: 778b759ba10e ("net/bnxt: add MAC address") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-18 03:35:57 +00:00
return rc;
net/bnxt: add MAC address This patch adds dev_ops to Add/Remove MAC addresses. These use the bnxt_hwrm_set_filter() defined previously Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:34 +00:00
}
net/bnxt: fix link update 1. When port is stopped, we can forcibly set the link status for the device to down. 2. VFs and MH PFs do not have the privilege to bring the link down. As a result driver prints "Link Up" when port is stopped. 3. When driver receives link status/speed/config async event from fw, driver invokes bnxt_link_update() with exp_link_status as ETH_LINK_UP This is not logically correct as the async event could be for Link up or link down or for speed change. Fixes: 074cacb9907a ("net/bnxt: fix link during port toggle") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-10-06 16:01:56 +00:00
int bnxt_link_update_op(struct rte_eth_dev *eth_dev, int wait_to_complete)
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
{
int rc = 0;
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = eth_dev->data->dev_private;
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
struct rte_eth_link new;
net/bnxt: fix link update 1. When port is stopped, we can forcibly set the link status for the device to down. 2. VFs and MH PFs do not have the privilege to bring the link down. As a result driver prints "Link Up" when port is stopped. 3. When driver receives link status/speed/config async event from fw, driver invokes bnxt_link_update() with exp_link_status as ETH_LINK_UP This is not logically correct as the async event could be for Link up or link down or for speed change. Fixes: 074cacb9907a ("net/bnxt: fix link during port toggle") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-10-06 16:01:56 +00:00
int cnt = wait_to_complete ? BNXT_MAX_LINK_WAIT_CNT :
BNXT_MIN_LINK_WAIT_CNT;
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
rc = is_bnxt_in_error(bp);
if (rc)
return rc;
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
memset(&new, 0, sizeof(new));
net/bnxt: fix resource cleanup Fix resource cleanup in port close. Once the pointers are freed, set them to NULL. Make sure access to the pointers is validated before use. Fixes: bb81e07323bb ("net/bnxt: support LED on/off") Fixes: 804e746c7b73 ("net/bnxt: add hardware resource manager init code") Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Fixes: 698aa7e95325 ("net/bnxt: add code to determine the Tx COS queue") Fixes: 322bd6e70272 ("net/bnxt: add port representor infrastructure") Fixes: 0bf5a0b5ebb8 ("net/bnxt: add a failure log") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-04-08 20:17:47 +00:00
if (bp->link_info == NULL)
goto out;
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
do {
/* Retrieve link info from hardware */
rc = bnxt_get_hwrm_link_config(bp, &new);
if (rc) {
new.link_speed = ETH_LINK_SPEED_100M;
new.link_duplex = ETH_LINK_FULL_DUPLEX;
net/bnxt: use dynamic log type This patch implements driver specific log type doing away with usage of RTE_LOG() for logging. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:55 +00:00
PMD_DRV_LOG(ERR,
net/bnxt: handle VF/PF initialization appropriately 1) For a VF, query the firmware to determine if a MAC address is already configured. If not configure a random default MAC address. 2) Do not initialize the default completion ring in bnxt_alloc_hwrm_rings(). 3) While registering for async events with the firmware, use func_vf_cfg for a VF and use func_cfg for a PF. 4) Query the VNIC plcmode config using the bnxt_hwrm_vnic_plcmodes_qcfg before a VNIC is updated. Reconfigure the VNIC with the plcmode configuration queried earlier. Not doing this could overwrite the plcmodes in some cases. 5) Reorg the bnxt_handle_fwd_req to properly handle the forwarded requests. The previous code did not handle it completely. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:00 +00:00
"Failed to retrieve link rc = 0x%x!\n", rc);
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
goto out;
}
net/bnxt: fix link update 1. When port is stopped, we can forcibly set the link status for the device to down. 2. VFs and MH PFs do not have the privilege to bring the link down. As a result driver prints "Link Up" when port is stopped. 3. When driver receives link status/speed/config async event from fw, driver invokes bnxt_link_update() with exp_link_status as ETH_LINK_UP This is not logically correct as the async event could be for Link up or link down or for speed change. Fixes: 074cacb9907a ("net/bnxt: fix link during port toggle") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-10-06 16:01:56 +00:00
if (!wait_to_complete || new.link_status)
net/bnxt: support async link notification This patch adds support to get Link notification asynchronously. The HW sends async notifications on default completion ring. The PMD processes these notifications and logs a message appropriately. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2016-10-11 21:47:50 +00:00
break;
net/bnxt: fix unconditional wait in link update There is an unconditional delay in link update op. Fixed it to wait only if wait for request completion is set. Fixes: 7bc8e9a227cc ("net/bnxt: support async link notification") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com>
2019-07-18 03:35:58 +00:00
rte_delay_ms(BNXT_LINK_WAIT_INTERVAL);
} while (cnt--);
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
net/bnxt: fix link update 1. When port is stopped, we can forcibly set the link status for the device to down. 2. VFs and MH PFs do not have the privilege to bring the link down. As a result driver prints "Link Up" when port is stopped. 3. When driver receives link status/speed/config async event from fw, driver invokes bnxt_link_update() with exp_link_status as ETH_LINK_UP This is not logically correct as the async event could be for Link up or link down or for speed change. Fixes: 074cacb9907a ("net/bnxt: fix link during port toggle") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-10-06 16:01:56 +00:00
/* Only single function PF can bring phy down.
* When port is stopped, report link down for VF/MH/NPAR functions.
*/
if (!BNXT_SINGLE_PF(bp) && !eth_dev->data->dev_started)
memset(&new, 0, sizeof(new));
net/bnxt: support async link notification This patch adds support to get Link notification asynchronously. The HW sends async notifications on default completion ring. The PMD processes these notifications and logs a message appropriately. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2016-10-11 21:47:50 +00:00
out:
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
/* Timed out or success */
net/bnxt: support async link notification This patch adds support to get Link notification asynchronously. The HW sends async notifications on default completion ring. The PMD processes these notifications and logs a message appropriately. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2016-10-11 21:47:50 +00:00
if (new.link_status != eth_dev->data->dev_link.link_status ||
net/bnxt: update indentation and coding style Indentation and coding style fixes. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2020-12-05 00:15:05 +00:00
new.link_speed != eth_dev->data->dev_link.link_speed) {
net/bnxt: fix async link handling and update When updating the link because of an async link notification there is no need to set wait_for_completion. At this point the link related information should be available without need to poll. Use rte_eth_linkstatus_set instead of memcpy to ensure atomicity while updating the link status. We force the physical link down as a part of device stop. But we are not waiting there enough and handling the async notification before exiting. It just sits in the default CQ till we do a device start. Fix it by calling the CQ handler in device stop. Fixes: 7bc8e9a227cc ("net/bnxt: support async link notification") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com>
2019-10-02 17:17:34 +00:00
rte_eth_linkstatus_set(eth_dev, &new);
net/bnxt: add support for LSC interrupt event Add support to bnxt driver to register RTE_ETH_EVENT_INTR_LSC event and monitor physical link status. Signed-off-by: Qingmin Liu <qingmin.liu@broadcom.com> Signed-off-by: Scott Branden <scott.branden@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Randy Schacher <stuart.schacher@broadcom.com>
2018-05-01 01:06:05 +00:00
ethdev: remove underscore prefix from internal API '_rte_eth_dev_callback_process()' & '_rte_eth_dev_reset()' internal APIs has unconventional underscore ('_') prefix. Although this is not documented most probably this is to mark them as internal. Since we have '__rte_internal' flag to mark this, removing '_' from API names. For '_rte_eth_dev_reset()', there is already a public API named 'rte_eth_dev_reset()', so renaming '_rte_eth_dev_reset()' to 'rte_eth_dev_internal_reset'. Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com> Acked-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: David Marchand <david.marchand@redhat.com> Acked-by: Sachin Saxena <sachin.saxena@nxp.com>
2020-09-09 13:01:48 +00:00
rte_eth_dev_callback_process(eth_dev,
RTE_ETH_EVENT_INTR_LSC,
NULL);
net/bnxt: add support for LSC interrupt event Add support to bnxt driver to register RTE_ETH_EVENT_INTR_LSC event and monitor physical link status. Signed-off-by: Qingmin Liu <qingmin.liu@broadcom.com> Signed-off-by: Scott Branden <scott.branden@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Randy Schacher <stuart.schacher@broadcom.com>
2018-05-01 01:06:05 +00:00
net/bnxt: support async link notification This patch adds support to get Link notification asynchronously. The HW sends async notifications on default completion ring. The PMD processes these notifications and logs a message appropriately. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2016-10-11 21:47:50 +00:00
bnxt_print_link_info(eth_dev);
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
}
net/bnxt: support async link notification This patch adds support to get Link notification asynchronously. The HW sends async notifications on default completion ring. The PMD processes these notifications and logs a message appropriately. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2016-10-11 21:47:50 +00:00
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
return rc;
}
ethdev: change promiscuous callbacks to return status Enabling/disabling of promiscuous mode is not always successful and it should be taken into account to be able to handle it properly. When correct return status is unclear from driver code, -EAGAIN is used. Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Matan Azrad <matan@mellanox.com> Acked-by: Hyong Youb Kim <hyonkim@cisco.com>
2019-09-14 11:37:24 +00:00
static int bnxt_promiscuous_enable_op(struct rte_eth_dev *eth_dev)
net/bnxt: add promiscuous mode This patch adds the promiscuous mode enable and disable dev_ops. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added in an earlier commit. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:31 +00:00
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = eth_dev->data->dev_private;
net/bnxt: add promiscuous mode This patch adds the promiscuous mode enable and disable dev_ops. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added in an earlier commit. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:31 +00:00
struct bnxt_vnic_info *vnic;
ethdev: change promiscuous callbacks to return status Enabling/disabling of promiscuous mode is not always successful and it should be taken into account to be able to handle it properly. When correct return status is unclear from driver code, -EAGAIN is used. Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Matan Azrad <matan@mellanox.com> Acked-by: Hyong Youb Kim <hyonkim@cisco.com>
2019-09-14 11:37:24 +00:00
uint32_t old_flags;
int rc;
net/bnxt: add promiscuous mode This patch adds the promiscuous mode enable and disable dev_ops. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added in an earlier commit. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:31 +00:00
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
rc = is_bnxt_in_error(bp);
if (rc)
return rc;
net/bnxt: handle HW filter setting when port is stopped Driver destroy the VNIC when the port is brought down. Port HW filter setting such as promiscuos, allmulti and VLAN filtering will be applied when port is started. Fixed to return success silently for these callbacks when port is stopped. Also fixed to clear "bp->dev_stopped" before invoking bnxt_vlan_offload_set_op() in bnxt_dev_start_op(). Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
2020-01-16 13:04:51 +00:00
/* Filter settings will get applied when port is started */
net/bnxt: remove a redundant variable Use "dev->data->dev_started" state, instead of local "dev_stopped" to check whether port has been started or not. Fixes: 316e412299fd ("net/bnxt: fix crash when closing") Cc: stable@dpdk.org Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
2020-01-28 07:29:21 +00:00
if (!eth_dev->data->dev_started)
net/bnxt: handle HW filter setting when port is stopped Driver destroy the VNIC when the port is brought down. Port HW filter setting such as promiscuos, allmulti and VLAN filtering will be applied when port is started. Fixed to return success silently for these callbacks when port is stopped. Also fixed to clear "bp->dev_stopped" before invoking bnxt_vlan_offload_set_op() in bnxt_dev_start_op(). Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
2020-01-16 13:04:51 +00:00
return 0;
net/bnxt: add promiscuous mode This patch adds the promiscuous mode enable and disable dev_ops. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added in an earlier commit. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:31 +00:00
if (bp->vnic_info == NULL)
ethdev: change promiscuous callbacks to return status Enabling/disabling of promiscuous mode is not always successful and it should be taken into account to be able to handle it properly. When correct return status is unclear from driver code, -EAGAIN is used. Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Matan Azrad <matan@mellanox.com> Acked-by: Hyong Youb Kim <hyonkim@cisco.com>
2019-09-14 11:37:24 +00:00
return 0;
net/bnxt: add promiscuous mode This patch adds the promiscuous mode enable and disable dev_ops. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added in an earlier commit. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:31 +00:00
net/bnxt: use macro for getting default VNIC Minor code cleanup. BNXT_GET_DEFAULT_VNIC macro returns the default VNIC of the function and fixed code to use the macro. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 17:17:45 +00:00
vnic = BNXT_GET_DEFAULT_VNIC(bp);
net/bnxt: add promiscuous mode This patch adds the promiscuous mode enable and disable dev_ops. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added in an earlier commit. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:31 +00:00
ethdev: change promiscuous callbacks to return status Enabling/disabling of promiscuous mode is not always successful and it should be taken into account to be able to handle it properly. When correct return status is unclear from driver code, -EAGAIN is used. Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Matan Azrad <matan@mellanox.com> Acked-by: Hyong Youb Kim <hyonkim@cisco.com>
2019-09-14 11:37:24 +00:00
old_flags = vnic->flags;
net/bnxt: add promiscuous mode This patch adds the promiscuous mode enable and disable dev_ops. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added in an earlier commit. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:31 +00:00
vnic->flags |= BNXT_VNIC_INFO_PROMISC;
ethdev: change promiscuous callbacks to return status Enabling/disabling of promiscuous mode is not always successful and it should be taken into account to be able to handle it properly. When correct return status is unclear from driver code, -EAGAIN is used. Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Matan Azrad <matan@mellanox.com> Acked-by: Hyong Youb Kim <hyonkim@cisco.com>
2019-09-14 11:37:24 +00:00
rc = bnxt_hwrm_cfa_l2_set_rx_mask(bp, vnic, 0, NULL);
if (rc != 0)
vnic->flags = old_flags;
return rc;
net/bnxt: add promiscuous mode This patch adds the promiscuous mode enable and disable dev_ops. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added in an earlier commit. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:31 +00:00
}
ethdev: change promiscuous callbacks to return status Enabling/disabling of promiscuous mode is not always successful and it should be taken into account to be able to handle it properly. When correct return status is unclear from driver code, -EAGAIN is used. Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Matan Azrad <matan@mellanox.com> Acked-by: Hyong Youb Kim <hyonkim@cisco.com>
2019-09-14 11:37:24 +00:00
static int bnxt_promiscuous_disable_op(struct rte_eth_dev *eth_dev)
net/bnxt: add promiscuous mode This patch adds the promiscuous mode enable and disable dev_ops. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added in an earlier commit. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:31 +00:00
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = eth_dev->data->dev_private;
net/bnxt: add promiscuous mode This patch adds the promiscuous mode enable and disable dev_ops. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added in an earlier commit. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:31 +00:00
struct bnxt_vnic_info *vnic;
ethdev: change promiscuous callbacks to return status Enabling/disabling of promiscuous mode is not always successful and it should be taken into account to be able to handle it properly. When correct return status is unclear from driver code, -EAGAIN is used. Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Matan Azrad <matan@mellanox.com> Acked-by: Hyong Youb Kim <hyonkim@cisco.com>
2019-09-14 11:37:24 +00:00
uint32_t old_flags;
int rc;
net/bnxt: add promiscuous mode This patch adds the promiscuous mode enable and disable dev_ops. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added in an earlier commit. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:31 +00:00
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
rc = is_bnxt_in_error(bp);
if (rc)
return rc;
net/bnxt: handle HW filter setting when port is stopped Driver destroy the VNIC when the port is brought down. Port HW filter setting such as promiscuos, allmulti and VLAN filtering will be applied when port is started. Fixed to return success silently for these callbacks when port is stopped. Also fixed to clear "bp->dev_stopped" before invoking bnxt_vlan_offload_set_op() in bnxt_dev_start_op(). Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
2020-01-16 13:04:51 +00:00
/* Filter settings will get applied when port is started */
net/bnxt: remove a redundant variable Use "dev->data->dev_started" state, instead of local "dev_stopped" to check whether port has been started or not. Fixes: 316e412299fd ("net/bnxt: fix crash when closing") Cc: stable@dpdk.org Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
2020-01-28 07:29:21 +00:00
if (!eth_dev->data->dev_started)
net/bnxt: handle HW filter setting when port is stopped Driver destroy the VNIC when the port is brought down. Port HW filter setting such as promiscuos, allmulti and VLAN filtering will be applied when port is started. Fixed to return success silently for these callbacks when port is stopped. Also fixed to clear "bp->dev_stopped" before invoking bnxt_vlan_offload_set_op() in bnxt_dev_start_op(). Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
2020-01-16 13:04:51 +00:00
return 0;
net/bnxt: add promiscuous mode This patch adds the promiscuous mode enable and disable dev_ops. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added in an earlier commit. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:31 +00:00
if (bp->vnic_info == NULL)
ethdev: change promiscuous callbacks to return status Enabling/disabling of promiscuous mode is not always successful and it should be taken into account to be able to handle it properly. When correct return status is unclear from driver code, -EAGAIN is used. Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Matan Azrad <matan@mellanox.com> Acked-by: Hyong Youb Kim <hyonkim@cisco.com>
2019-09-14 11:37:24 +00:00
return 0;
net/bnxt: add promiscuous mode This patch adds the promiscuous mode enable and disable dev_ops. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added in an earlier commit. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:31 +00:00
net/bnxt: use macro for getting default VNIC Minor code cleanup. BNXT_GET_DEFAULT_VNIC macro returns the default VNIC of the function and fixed code to use the macro. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 17:17:45 +00:00
vnic = BNXT_GET_DEFAULT_VNIC(bp);
net/bnxt: add promiscuous mode This patch adds the promiscuous mode enable and disable dev_ops. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added in an earlier commit. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:31 +00:00
ethdev: change promiscuous callbacks to return status Enabling/disabling of promiscuous mode is not always successful and it should be taken into account to be able to handle it properly. When correct return status is unclear from driver code, -EAGAIN is used. Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Matan Azrad <matan@mellanox.com> Acked-by: Hyong Youb Kim <hyonkim@cisco.com>
2019-09-14 11:37:24 +00:00
old_flags = vnic->flags;
net/bnxt: add promiscuous mode This patch adds the promiscuous mode enable and disable dev_ops. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added in an earlier commit. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:31 +00:00
vnic->flags &= ~BNXT_VNIC_INFO_PROMISC;
ethdev: change promiscuous callbacks to return status Enabling/disabling of promiscuous mode is not always successful and it should be taken into account to be able to handle it properly. When correct return status is unclear from driver code, -EAGAIN is used. Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Matan Azrad <matan@mellanox.com> Acked-by: Hyong Youb Kim <hyonkim@cisco.com>
2019-09-14 11:37:24 +00:00
rc = bnxt_hwrm_cfa_l2_set_rx_mask(bp, vnic, 0, NULL);
if (rc != 0)
vnic->flags = old_flags;
return rc;
net/bnxt: add promiscuous mode This patch adds the promiscuous mode enable and disable dev_ops. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added in an earlier commit. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:31 +00:00
}
ethdev: change allmulticast callbacks to return status Enabling/disabling of allmulticast mode is not always successful and it should be taken into account to be able to handle it properly. When correct return status is unclear from driver code, -EAGAIN is used. Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Hyong Youb Kim <hyonkim@cisco.com>
2019-09-24 12:56:10 +00:00
static int bnxt_allmulticast_enable_op(struct rte_eth_dev *eth_dev)
net/bnxt: add multicast This patch adds dev_ops to enable/disable multicast traffic. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added earlier. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:32 +00:00
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = eth_dev->data->dev_private;
net/bnxt: add multicast This patch adds dev_ops to enable/disable multicast traffic. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added earlier. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:32 +00:00
struct bnxt_vnic_info *vnic;
ethdev: change allmulticast callbacks to return status Enabling/disabling of allmulticast mode is not always successful and it should be taken into account to be able to handle it properly. When correct return status is unclear from driver code, -EAGAIN is used. Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Hyong Youb Kim <hyonkim@cisco.com>
2019-09-24 12:56:10 +00:00
uint32_t old_flags;
int rc;
net/bnxt: add multicast This patch adds dev_ops to enable/disable multicast traffic. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added earlier. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:32 +00:00
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
rc = is_bnxt_in_error(bp);
if (rc)
return rc;
net/bnxt: handle HW filter setting when port is stopped Driver destroy the VNIC when the port is brought down. Port HW filter setting such as promiscuos, allmulti and VLAN filtering will be applied when port is started. Fixed to return success silently for these callbacks when port is stopped. Also fixed to clear "bp->dev_stopped" before invoking bnxt_vlan_offload_set_op() in bnxt_dev_start_op(). Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
2020-01-16 13:04:51 +00:00
/* Filter settings will get applied when port is started */
net/bnxt: remove a redundant variable Use "dev->data->dev_started" state, instead of local "dev_stopped" to check whether port has been started or not. Fixes: 316e412299fd ("net/bnxt: fix crash when closing") Cc: stable@dpdk.org Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
2020-01-28 07:29:21 +00:00
if (!eth_dev->data->dev_started)
net/bnxt: handle HW filter setting when port is stopped Driver destroy the VNIC when the port is brought down. Port HW filter setting such as promiscuos, allmulti and VLAN filtering will be applied when port is started. Fixed to return success silently for these callbacks when port is stopped. Also fixed to clear "bp->dev_stopped" before invoking bnxt_vlan_offload_set_op() in bnxt_dev_start_op(). Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
2020-01-16 13:04:51 +00:00
return 0;
net/bnxt: add multicast This patch adds dev_ops to enable/disable multicast traffic. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added earlier. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:32 +00:00
if (bp->vnic_info == NULL)
ethdev: change allmulticast callbacks to return status Enabling/disabling of allmulticast mode is not always successful and it should be taken into account to be able to handle it properly. When correct return status is unclear from driver code, -EAGAIN is used. Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Hyong Youb Kim <hyonkim@cisco.com>
2019-09-24 12:56:10 +00:00
return 0;
net/bnxt: add multicast This patch adds dev_ops to enable/disable multicast traffic. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added earlier. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:32 +00:00
net/bnxt: use macro for getting default VNIC Minor code cleanup. BNXT_GET_DEFAULT_VNIC macro returns the default VNIC of the function and fixed code to use the macro. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 17:17:45 +00:00
vnic = BNXT_GET_DEFAULT_VNIC(bp);
net/bnxt: add multicast This patch adds dev_ops to enable/disable multicast traffic. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added earlier. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:32 +00:00
ethdev: change allmulticast callbacks to return status Enabling/disabling of allmulticast mode is not always successful and it should be taken into account to be able to handle it properly. When correct return status is unclear from driver code, -EAGAIN is used. Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Hyong Youb Kim <hyonkim@cisco.com>
2019-09-24 12:56:10 +00:00
old_flags = vnic->flags;
net/bnxt: add multicast This patch adds dev_ops to enable/disable multicast traffic. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added earlier. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:32 +00:00
vnic->flags |= BNXT_VNIC_INFO_ALLMULTI;
ethdev: change allmulticast callbacks to return status Enabling/disabling of allmulticast mode is not always successful and it should be taken into account to be able to handle it properly. When correct return status is unclear from driver code, -EAGAIN is used. Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Hyong Youb Kim <hyonkim@cisco.com>
2019-09-24 12:56:10 +00:00
rc = bnxt_hwrm_cfa_l2_set_rx_mask(bp, vnic, 0, NULL);
if (rc != 0)
vnic->flags = old_flags;
return rc;
net/bnxt: add multicast This patch adds dev_ops to enable/disable multicast traffic. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added earlier. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:32 +00:00
}
ethdev: change allmulticast callbacks to return status Enabling/disabling of allmulticast mode is not always successful and it should be taken into account to be able to handle it properly. When correct return status is unclear from driver code, -EAGAIN is used. Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Hyong Youb Kim <hyonkim@cisco.com>
2019-09-24 12:56:10 +00:00
static int bnxt_allmulticast_disable_op(struct rte_eth_dev *eth_dev)
net/bnxt: add multicast This patch adds dev_ops to enable/disable multicast traffic. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added earlier. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:32 +00:00
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = eth_dev->data->dev_private;
net/bnxt: add multicast This patch adds dev_ops to enable/disable multicast traffic. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added earlier. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:32 +00:00
struct bnxt_vnic_info *vnic;
ethdev: change allmulticast callbacks to return status Enabling/disabling of allmulticast mode is not always successful and it should be taken into account to be able to handle it properly. When correct return status is unclear from driver code, -EAGAIN is used. Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Hyong Youb Kim <hyonkim@cisco.com>
2019-09-24 12:56:10 +00:00
uint32_t old_flags;
int rc;
net/bnxt: add multicast This patch adds dev_ops to enable/disable multicast traffic. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added earlier. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:32 +00:00
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
rc = is_bnxt_in_error(bp);
if (rc)
return rc;
net/bnxt: handle HW filter setting when port is stopped Driver destroy the VNIC when the port is brought down. Port HW filter setting such as promiscuos, allmulti and VLAN filtering will be applied when port is started. Fixed to return success silently for these callbacks when port is stopped. Also fixed to clear "bp->dev_stopped" before invoking bnxt_vlan_offload_set_op() in bnxt_dev_start_op(). Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
2020-01-16 13:04:51 +00:00
/* Filter settings will get applied when port is started */
net/bnxt: remove a redundant variable Use "dev->data->dev_started" state, instead of local "dev_stopped" to check whether port has been started or not. Fixes: 316e412299fd ("net/bnxt: fix crash when closing") Cc: stable@dpdk.org Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
2020-01-28 07:29:21 +00:00
if (!eth_dev->data->dev_started)
net/bnxt: handle HW filter setting when port is stopped Driver destroy the VNIC when the port is brought down. Port HW filter setting such as promiscuos, allmulti and VLAN filtering will be applied when port is started. Fixed to return success silently for these callbacks when port is stopped. Also fixed to clear "bp->dev_stopped" before invoking bnxt_vlan_offload_set_op() in bnxt_dev_start_op(). Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
2020-01-16 13:04:51 +00:00
return 0;
net/bnxt: add multicast This patch adds dev_ops to enable/disable multicast traffic. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added earlier. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:32 +00:00
if (bp->vnic_info == NULL)
ethdev: change allmulticast callbacks to return status Enabling/disabling of allmulticast mode is not always successful and it should be taken into account to be able to handle it properly. When correct return status is unclear from driver code, -EAGAIN is used. Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Hyong Youb Kim <hyonkim@cisco.com>
2019-09-24 12:56:10 +00:00
return 0;
net/bnxt: add multicast This patch adds dev_ops to enable/disable multicast traffic. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added earlier. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:32 +00:00
net/bnxt: use macro for getting default VNIC Minor code cleanup. BNXT_GET_DEFAULT_VNIC macro returns the default VNIC of the function and fixed code to use the macro. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 17:17:45 +00:00
vnic = BNXT_GET_DEFAULT_VNIC(bp);
net/bnxt: add multicast This patch adds dev_ops to enable/disable multicast traffic. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added earlier. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:32 +00:00
ethdev: change allmulticast callbacks to return status Enabling/disabling of allmulticast mode is not always successful and it should be taken into account to be able to handle it properly. When correct return status is unclear from driver code, -EAGAIN is used. Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Hyong Youb Kim <hyonkim@cisco.com>
2019-09-24 12:56:10 +00:00
old_flags = vnic->flags;
net/bnxt: add multicast This patch adds dev_ops to enable/disable multicast traffic. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added earlier. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:32 +00:00
vnic->flags &= ~BNXT_VNIC_INFO_ALLMULTI;
ethdev: change allmulticast callbacks to return status Enabling/disabling of allmulticast mode is not always successful and it should be taken into account to be able to handle it properly. When correct return status is unclear from driver code, -EAGAIN is used. Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Hyong Youb Kim <hyonkim@cisco.com>
2019-09-24 12:56:10 +00:00
rc = bnxt_hwrm_cfa_l2_set_rx_mask(bp, vnic, 0, NULL);
if (rc != 0)
vnic->flags = old_flags;
return rc;
net/bnxt: add multicast This patch adds dev_ops to enable/disable multicast traffic. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added earlier. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:32 +00:00
}
net/bnxt: fix RSS RETA indirection table ops We are trying to update the indirection table for all the VNICs. We should update the table only for the default vnic0. Fix the reta update function to only update table entries that are selected by the update mask. Translate queue number to firmware group ID when updating an entry. Fix reta query op to only return table entries as identfied by the provided mask. Translate firmware group IDs to queue numbers. Removed extraneous code from bnxt_reta_query_op(). Fixes: d819382543f3 ("net/bnxt: add RSS redirection table operations") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Rahul Gupta <rahul.gupta@broadcom.com>
2019-05-29 21:02:24 +00:00
/* Return bnxt_rx_queue pointer corresponding to a given rxq. */
static struct bnxt_rx_queue *bnxt_qid_to_rxq(struct bnxt *bp, uint16_t qid)
{
if (qid >= bp->rx_nr_rings)
return NULL;
return bp->eth_dev->data->rx_queues[qid];
}
/* Return rxq corresponding to a given rss table ring/group ID. */
static uint16_t bnxt_rss_to_qid(struct bnxt *bp, uint16_t fwr)
{
net/bnxt: enable RSS for thor-based controllers Make changes needed to support rss for thor-based controllers. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-06-02 17:42:47 +00:00
struct bnxt_rx_queue *rxq;
net/bnxt: fix RSS RETA indirection table ops We are trying to update the indirection table for all the VNICs. We should update the table only for the default vnic0. Fix the reta update function to only update table entries that are selected by the update mask. Translate queue number to firmware group ID when updating an entry. Fix reta query op to only return table entries as identfied by the provided mask. Translate firmware group IDs to queue numbers. Removed extraneous code from bnxt_reta_query_op(). Fixes: d819382543f3 ("net/bnxt: add RSS redirection table operations") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Rahul Gupta <rahul.gupta@broadcom.com>
2019-05-29 21:02:24 +00:00
unsigned int i;
net/bnxt: enable RSS for thor-based controllers Make changes needed to support rss for thor-based controllers. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-06-02 17:42:47 +00:00
if (!BNXT_HAS_RING_GRPS(bp)) {
for (i = 0; i < bp->rx_nr_rings; i++) {
rxq = bp->eth_dev->data->rx_queues[i];
if (rxq->rx_ring->rx_ring_struct->fw_ring_id == fwr)
return rxq->index;
}
} else {
for (i = 0; i < bp->rx_nr_rings; i++) {
if (bp->grp_info[i].fw_grp_id == fwr)
return i;
}
net/bnxt: fix RSS RETA indirection table ops We are trying to update the indirection table for all the VNICs. We should update the table only for the default vnic0. Fix the reta update function to only update table entries that are selected by the update mask. Translate queue number to firmware group ID when updating an entry. Fix reta query op to only return table entries as identfied by the provided mask. Translate firmware group IDs to queue numbers. Removed extraneous code from bnxt_reta_query_op(). Fixes: d819382543f3 ("net/bnxt: add RSS redirection table operations") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Rahul Gupta <rahul.gupta@broadcom.com>
2019-05-29 21:02:24 +00:00
}
return INVALID_HW_RING_ID;
}
net/bnxt: add RSS redirection table operations Add code to Update/query reta dev_ops Uses the bnxt_hwrm_vnic_rss_cfg() HWRM API added earlier. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:36 +00:00
static int bnxt_reta_update_op(struct rte_eth_dev *eth_dev,
struct rte_eth_rss_reta_entry64 *reta_conf,
uint16_t reta_size)
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = eth_dev->data->dev_private;
net/bnxt: add RSS redirection table operations Add code to Update/query reta dev_ops Uses the bnxt_hwrm_vnic_rss_cfg() HWRM API added earlier. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:36 +00:00
struct rte_eth_conf *dev_conf = &bp->eth_dev->data->dev_conf;
net/bnxt: use macro for getting default VNIC BNXT_GET_DEFAULT_VNIC macro returns the default VNIC of the function and fixed all occurrences of code to use the macro. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-12-02 04:18:09 +00:00
struct bnxt_vnic_info *vnic = BNXT_GET_DEFAULT_VNIC(bp);
net/bnxt: enable RSS for thor-based controllers Make changes needed to support rss for thor-based controllers. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-06-02 17:42:47 +00:00
uint16_t tbl_size = bnxt_rss_hash_tbl_size(bp);
net/bnxt: fix RSS RETA indirection table ops We are trying to update the indirection table for all the VNICs. We should update the table only for the default vnic0. Fix the reta update function to only update table entries that are selected by the update mask. Translate queue number to firmware group ID when updating an entry. Fix reta query op to only return table entries as identfied by the provided mask. Translate firmware group IDs to queue numbers. Removed extraneous code from bnxt_reta_query_op(). Fixes: d819382543f3 ("net/bnxt: add RSS redirection table operations") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Rahul Gupta <rahul.gupta@broadcom.com>
2019-05-29 21:02:24 +00:00
uint16_t idx, sft;
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
int i, rc;
rc = is_bnxt_in_error(bp);
if (rc)
return rc;
net/bnxt: add RSS redirection table operations Add code to Update/query reta dev_ops Uses the bnxt_hwrm_vnic_rss_cfg() HWRM API added earlier. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:36 +00:00
net/bnxt: fix RSS RETA indirection table ops We are trying to update the indirection table for all the VNICs. We should update the table only for the default vnic0. Fix the reta update function to only update table entries that are selected by the update mask. Translate queue number to firmware group ID when updating an entry. Fix reta query op to only return table entries as identfied by the provided mask. Translate firmware group IDs to queue numbers. Removed extraneous code from bnxt_reta_query_op(). Fixes: d819382543f3 ("net/bnxt: add RSS redirection table operations") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Rahul Gupta <rahul.gupta@broadcom.com>
2019-05-29 21:02:24 +00:00
if (!vnic->rss_table)
return -EINVAL;
net/bnxt: add RSS redirection table operations Add code to Update/query reta dev_ops Uses the bnxt_hwrm_vnic_rss_cfg() HWRM API added earlier. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:36 +00:00
if (!(dev_conf->rxmode.mq_mode & ETH_MQ_RX_RSS_FLAG))
return -EINVAL;
net/bnxt: fix RSS RETA indirection table ops We are trying to update the indirection table for all the VNICs. We should update the table only for the default vnic0. Fix the reta update function to only update table entries that are selected by the update mask. Translate queue number to firmware group ID when updating an entry. Fix reta query op to only return table entries as identfied by the provided mask. Translate firmware group IDs to queue numbers. Removed extraneous code from bnxt_reta_query_op(). Fixes: d819382543f3 ("net/bnxt: add RSS redirection table operations") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Rahul Gupta <rahul.gupta@broadcom.com>
2019-05-29 21:02:24 +00:00
if (reta_size != tbl_size) {
net/bnxt: use dynamic log type This patch implements driver specific log type doing away with usage of RTE_LOG() for logging. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:55 +00:00
PMD_DRV_LOG(ERR, "The configured hash table lookup size "
net/bnxt: add RSS redirection table operations Add code to Update/query reta dev_ops Uses the bnxt_hwrm_vnic_rss_cfg() HWRM API added earlier. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:36 +00:00
"(%d) must equal the size supported by the hardware "
net/bnxt: fix RSS RETA indirection table ops We are trying to update the indirection table for all the VNICs. We should update the table only for the default vnic0. Fix the reta update function to only update table entries that are selected by the update mask. Translate queue number to firmware group ID when updating an entry. Fix reta query op to only return table entries as identfied by the provided mask. Translate firmware group IDs to queue numbers. Removed extraneous code from bnxt_reta_query_op(). Fixes: d819382543f3 ("net/bnxt: add RSS redirection table operations") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Rahul Gupta <rahul.gupta@broadcom.com>
2019-05-29 21:02:24 +00:00
"(%d)\n", reta_size, tbl_size);
net/bnxt: add RSS redirection table operations Add code to Update/query reta dev_ops Uses the bnxt_hwrm_vnic_rss_cfg() HWRM API added earlier. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:36 +00:00
return -EINVAL;
}
net/bnxt: fix RSS RETA indirection table ops We are trying to update the indirection table for all the VNICs. We should update the table only for the default vnic0. Fix the reta update function to only update table entries that are selected by the update mask. Translate queue number to firmware group ID when updating an entry. Fix reta query op to only return table entries as identfied by the provided mask. Translate firmware group IDs to queue numbers. Removed extraneous code from bnxt_reta_query_op(). Fixes: d819382543f3 ("net/bnxt: add RSS redirection table operations") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Rahul Gupta <rahul.gupta@broadcom.com>
2019-05-29 21:02:24 +00:00
for (i = 0; i < reta_size; i++) {
struct bnxt_rx_queue *rxq;
idx = i / RTE_RETA_GROUP_SIZE;
sft = i % RTE_RETA_GROUP_SIZE;
if (!(reta_conf[idx].mask & (1ULL << sft)))
continue;
rxq = bnxt_qid_to_rxq(bp, reta_conf[idx].reta[sft]);
if (!rxq) {
PMD_DRV_LOG(ERR, "Invalid ring in reta_conf.\n");
return -EINVAL;
}
net/bnxt: remove references to Thor Refactor code to remove references to Thor. Instead use P5 as in phase 5 of development cycle since it is applicable to boards other than Thor as well. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2020-12-01 19:15:23 +00:00
if (BNXT_CHIP_P5(bp)) {
net/bnxt: enable RSS for thor-based controllers Make changes needed to support rss for thor-based controllers. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-06-02 17:42:47 +00:00
vnic->rss_table[i * 2] =
rxq->rx_ring->rx_ring_struct->fw_ring_id;
vnic->rss_table[i * 2 + 1] =
rxq->cp_ring->cp_ring_struct->fw_ring_id;
} else {
vnic->rss_table[i] =
vnic->fw_grp_ids[reta_conf[idx].reta[sft]];
}
net/bnxt: add RSS redirection table operations Add code to Update/query reta dev_ops Uses the bnxt_hwrm_vnic_rss_cfg() HWRM API added earlier. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:36 +00:00
}
net/bnxt: fix RSS RETA indirection table ops We are trying to update the indirection table for all the VNICs. We should update the table only for the default vnic0. Fix the reta update function to only update table entries that are selected by the update mask. Translate queue number to firmware group ID when updating an entry. Fix reta query op to only return table entries as identfied by the provided mask. Translate firmware group IDs to queue numbers. Removed extraneous code from bnxt_reta_query_op(). Fixes: d819382543f3 ("net/bnxt: add RSS redirection table operations") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Rahul Gupta <rahul.gupta@broadcom.com>
2019-05-29 21:02:24 +00:00
net/bnxt: propagate FW command failure to application In bnxt_reta_update_op() and bnxt_rss_hash_update_op(), driver does not propagate the error back to the application when the fw command fails. Fixes: 378ab645bb0b ("net/bnxt: fix RSS RETA indirection table ops") Fixes: fcc0aa1edc10 ("net/bnxt: add RSS hash configuration") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-11-12 10:47:54 +00:00
rc = bnxt_hwrm_vnic_rss_cfg(bp, vnic);
return rc;
net/bnxt: add RSS redirection table operations Add code to Update/query reta dev_ops Uses the bnxt_hwrm_vnic_rss_cfg() HWRM API added earlier. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:36 +00:00
}
static int bnxt_reta_query_op(struct rte_eth_dev *eth_dev,
struct rte_eth_rss_reta_entry64 *reta_conf,
uint16_t reta_size)
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = eth_dev->data->dev_private;
net/bnxt: use macro for getting default VNIC BNXT_GET_DEFAULT_VNIC macro returns the default VNIC of the function and fixed all occurrences of code to use the macro. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-12-02 04:18:09 +00:00
struct bnxt_vnic_info *vnic = BNXT_GET_DEFAULT_VNIC(bp);
net/bnxt: enable RSS for thor-based controllers Make changes needed to support rss for thor-based controllers. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-06-02 17:42:47 +00:00
uint16_t tbl_size = bnxt_rss_hash_tbl_size(bp);
net/bnxt: fix RSS RETA indirection table ops We are trying to update the indirection table for all the VNICs. We should update the table only for the default vnic0. Fix the reta update function to only update table entries that are selected by the update mask. Translate queue number to firmware group ID when updating an entry. Fix reta query op to only return table entries as identfied by the provided mask. Translate firmware group IDs to queue numbers. Removed extraneous code from bnxt_reta_query_op(). Fixes: d819382543f3 ("net/bnxt: add RSS redirection table operations") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Rahul Gupta <rahul.gupta@broadcom.com>
2019-05-29 21:02:24 +00:00
uint16_t idx, sft, i;
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
int rc;
rc = is_bnxt_in_error(bp);
if (rc)
return rc;
net/bnxt: add RSS redirection table operations Add code to Update/query reta dev_ops Uses the bnxt_hwrm_vnic_rss_cfg() HWRM API added earlier. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:36 +00:00
/* Retrieve from the default VNIC */
if (!vnic)
return -EINVAL;
if (!vnic->rss_table)
return -EINVAL;
net/bnxt: fix RSS RETA indirection table ops We are trying to update the indirection table for all the VNICs. We should update the table only for the default vnic0. Fix the reta update function to only update table entries that are selected by the update mask. Translate queue number to firmware group ID when updating an entry. Fix reta query op to only return table entries as identfied by the provided mask. Translate firmware group IDs to queue numbers. Removed extraneous code from bnxt_reta_query_op(). Fixes: d819382543f3 ("net/bnxt: add RSS redirection table operations") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Rahul Gupta <rahul.gupta@broadcom.com>
2019-05-29 21:02:24 +00:00
if (reta_size != tbl_size) {
net/bnxt: use dynamic log type This patch implements driver specific log type doing away with usage of RTE_LOG() for logging. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:55 +00:00
PMD_DRV_LOG(ERR, "The configured hash table lookup size "
net/bnxt: add RSS redirection table operations Add code to Update/query reta dev_ops Uses the bnxt_hwrm_vnic_rss_cfg() HWRM API added earlier. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:36 +00:00
"(%d) must equal the size supported by the hardware "
net/bnxt: fix RSS RETA indirection table ops We are trying to update the indirection table for all the VNICs. We should update the table only for the default vnic0. Fix the reta update function to only update table entries that are selected by the update mask. Translate queue number to firmware group ID when updating an entry. Fix reta query op to only return table entries as identfied by the provided mask. Translate firmware group IDs to queue numbers. Removed extraneous code from bnxt_reta_query_op(). Fixes: d819382543f3 ("net/bnxt: add RSS redirection table operations") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Rahul Gupta <rahul.gupta@broadcom.com>
2019-05-29 21:02:24 +00:00
"(%d)\n", reta_size, tbl_size);
net/bnxt: add RSS redirection table operations Add code to Update/query reta dev_ops Uses the bnxt_hwrm_vnic_rss_cfg() HWRM API added earlier. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:36 +00:00
return -EINVAL;
}
net/bnxt: fix RSS RETA indirection table ops We are trying to update the indirection table for all the VNICs. We should update the table only for the default vnic0. Fix the reta update function to only update table entries that are selected by the update mask. Translate queue number to firmware group ID when updating an entry. Fix reta query op to only return table entries as identfied by the provided mask. Translate firmware group IDs to queue numbers. Removed extraneous code from bnxt_reta_query_op(). Fixes: d819382543f3 ("net/bnxt: add RSS redirection table operations") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Rahul Gupta <rahul.gupta@broadcom.com>
2019-05-29 21:02:24 +00:00
for (idx = 0, i = 0; i < reta_size; i++) {
idx = i / RTE_RETA_GROUP_SIZE;
sft = i % RTE_RETA_GROUP_SIZE;
if (reta_conf[idx].mask & (1ULL << sft)) {
uint16_t qid;
net/bnxt: remove references to Thor Refactor code to remove references to Thor. Instead use P5 as in phase 5 of development cycle since it is applicable to boards other than Thor as well. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2020-12-01 19:15:23 +00:00
if (BNXT_CHIP_P5(bp))
net/bnxt: enable RSS for thor-based controllers Make changes needed to support rss for thor-based controllers. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-06-02 17:42:47 +00:00
qid = bnxt_rss_to_qid(bp,
vnic->rss_table[i * 2]);
else
qid = bnxt_rss_to_qid(bp, vnic->rss_table[i]);
net/bnxt: fix RSS RETA indirection table ops We are trying to update the indirection table for all the VNICs. We should update the table only for the default vnic0. Fix the reta update function to only update table entries that are selected by the update mask. Translate queue number to firmware group ID when updating an entry. Fix reta query op to only return table entries as identfied by the provided mask. Translate firmware group IDs to queue numbers. Removed extraneous code from bnxt_reta_query_op(). Fixes: d819382543f3 ("net/bnxt: add RSS redirection table operations") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Rahul Gupta <rahul.gupta@broadcom.com>
2019-05-29 21:02:24 +00:00
if (qid == INVALID_HW_RING_ID) {
PMD_DRV_LOG(ERR, "Inv. entry in rss table.\n");
return -EINVAL;
}
reta_conf[idx].reta[sft] = qid;
}
net/bnxt: support async link notification This patch adds support to get Link notification asynchronously. The HW sends async notifications on default completion ring. The PMD processes these notifications and logs a message appropriately. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2016-10-11 21:47:50 +00:00
}
net/bnxt: add RSS redirection table operations Add code to Update/query reta dev_ops Uses the bnxt_hwrm_vnic_rss_cfg() HWRM API added earlier. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:36 +00:00
return 0;
}
net/bnxt: add RSS hash configuration Add rss_hash_update and rss_hash_conf_get dev_ops Uses the bnxt_hwrm_vnic_rss_cfg() HWRM API added in the previous patch. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:37 +00:00
static int bnxt_rss_hash_update_op(struct rte_eth_dev *eth_dev,
struct rte_eth_rss_conf *rss_conf)
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = eth_dev->data->dev_private;
net/bnxt: add RSS hash configuration Add rss_hash_update and rss_hash_conf_get dev_ops Uses the bnxt_hwrm_vnic_rss_cfg() HWRM API added in the previous patch. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:37 +00:00
struct rte_eth_conf *dev_conf = &bp->eth_dev->data->dev_conf;
struct bnxt_vnic_info *vnic;
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
int rc;
rc = is_bnxt_in_error(bp);
if (rc)
return rc;
net/bnxt: add RSS hash configuration Add rss_hash_update and rss_hash_conf_get dev_ops Uses the bnxt_hwrm_vnic_rss_cfg() HWRM API added in the previous patch. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:37 +00:00
/*
* If RSS enablement were different than dev_configure,
* then return -EINVAL
*/
if (dev_conf->rxmode.mq_mode & ETH_MQ_RX_RSS_FLAG) {
if (!rss_conf->rss_hf)
net/bnxt: use dynamic log type This patch implements driver specific log type doing away with usage of RTE_LOG() for logging. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:55 +00:00
PMD_DRV_LOG(ERR, "Hash type NONE\n");
net/bnxt: add RSS hash configuration Add rss_hash_update and rss_hash_conf_get dev_ops Uses the bnxt_hwrm_vnic_rss_cfg() HWRM API added in the previous patch. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:37 +00:00
} else {
if (rss_conf->rss_hf & BNXT_ETH_RSS_SUPPORT)
return -EINVAL;
}
net/bnxt: fix config RSS update We are not configuring the RSS settings updated by rss_hash_update(). Fixing it. Fixes: fcc0aa1edc10 ("net/bnxt: add RSS hash configuration") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:33 +00:00
bp->flags |= BNXT_FLAG_UPDATE_HASH;
net/bnxt: fix using RSS config struct There is no need to maintain local copy. This helps reduce the size of the bnxt structure. Fixes: fcc0aa1edc10 ("net/bnxt: add RSS hash configuration") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
2020-05-15 11:25:11 +00:00
memcpy(&eth_dev->data->dev_conf.rx_adv_conf.rss_conf,
rss_conf,
sizeof(*rss_conf));
net/bnxt: fix config RSS update We are not configuring the RSS settings updated by rss_hash_update(). Fixing it. Fixes: fcc0aa1edc10 ("net/bnxt: add RSS hash configuration") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:33 +00:00
net/bnxt: allow dynamic creation of VNIC Refactor code to allow dynamic creation of VNIC for RSS or Queue Action during flow create. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Rahul Gupta <rahul.gupta@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
2019-10-02 23:25:43 +00:00
/* Update the default RSS VNIC(s) */
net/bnxt: use macro for getting default VNIC BNXT_GET_DEFAULT_VNIC macro returns the default VNIC of the function and fixed all occurrences of code to use the macro. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-12-02 04:18:09 +00:00
vnic = BNXT_GET_DEFAULT_VNIC(bp);
net/bnxt: allow dynamic creation of VNIC Refactor code to allow dynamic creation of VNIC for RSS or Queue Action during flow create. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Rahul Gupta <rahul.gupta@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
2019-10-02 23:25:43 +00:00
vnic->hash_type = bnxt_rte_to_hwrm_hash_types(rss_conf->rss_hf);
net/bnxt: support RSS hash selection Add support to select RSS hash based on innermost or outermost headers. If an application is started without any specific settings the default mode configured by FW or HW shall be used. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-18 23:33:41 +00:00
vnic->hash_mode =
bnxt_rte_to_hwrm_hash_level(bp, rss_conf->rss_hf,
ETH_RSS_LEVEL(rss_conf->rss_hf));
net/bnxt: get rid of ff pools and use VNIC info array There was no direct association between the rxq's VNIC and the vnic_info[]. Explicitly associate the two in bnxt_mq_rx_configure(). Fixes: 0a256e4a548b ("net/bnxt: fix Rx ring count limitation") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
2018-09-29 01:59:52 +00:00
net/bnxt: allow dynamic creation of VNIC Refactor code to allow dynamic creation of VNIC for RSS or Queue Action during flow create. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Rahul Gupta <rahul.gupta@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
2019-10-02 23:25:43 +00:00
/*
net/bnxt: validate RSS hash key length In bnxt_rss_hash_update_op, driver is proceeding with bnxt_hwrm_vnic_rss_cfg even though RSS hashkey length is invalid. This patch fixes the problem by returning -EINVAL when RSS hashkey length is invalid. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Rahul Gupta <rahul.gupta@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
2019-10-02 23:25:55 +00:00
* If hashkey is not specified, use the previously configured
* hashkey
net/bnxt: allow dynamic creation of VNIC Refactor code to allow dynamic creation of VNIC for RSS or Queue Action during flow create. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Rahul Gupta <rahul.gupta@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
2019-10-02 23:25:43 +00:00
*/
net/bnxt: validate RSS hash key length In bnxt_rss_hash_update_op, driver is proceeding with bnxt_hwrm_vnic_rss_cfg even though RSS hashkey length is invalid. This patch fixes the problem by returning -EINVAL when RSS hashkey length is invalid. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Rahul Gupta <rahul.gupta@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
2019-10-02 23:25:55 +00:00
if (!rss_conf->rss_key)
goto rss_config;
net/bnxt: get rid of ff pools and use VNIC info array There was no direct association between the rxq's VNIC and the vnic_info[]. Explicitly associate the two in bnxt_mq_rx_configure(). Fixes: 0a256e4a548b ("net/bnxt: fix Rx ring count limitation") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
2018-09-29 01:59:52 +00:00
net/bnxt: validate RSS hash key length In bnxt_rss_hash_update_op, driver is proceeding with bnxt_hwrm_vnic_rss_cfg even though RSS hashkey length is invalid. This patch fixes the problem by returning -EINVAL when RSS hashkey length is invalid. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Rahul Gupta <rahul.gupta@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
2019-10-02 23:25:55 +00:00
if (rss_conf->rss_key_len != HW_HASH_KEY_SIZE) {
PMD_DRV_LOG(ERR,
"Invalid hashkey length, should be 16 bytes\n");
return -EINVAL;
}
memcpy(vnic->rss_hash_key, rss_conf->rss_key, rss_conf->rss_key_len);
rss_config:
net/bnxt: propagate FW command failure to application In bnxt_reta_update_op() and bnxt_rss_hash_update_op(), driver does not propagate the error back to the application when the fw command fails. Fixes: 378ab645bb0b ("net/bnxt: fix RSS RETA indirection table ops") Fixes: fcc0aa1edc10 ("net/bnxt: add RSS hash configuration") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-11-12 10:47:54 +00:00
rc = bnxt_hwrm_vnic_rss_cfg(bp, vnic);
return rc;
net/bnxt: add RSS hash configuration Add rss_hash_update and rss_hash_conf_get dev_ops Uses the bnxt_hwrm_vnic_rss_cfg() HWRM API added in the previous patch. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:37 +00:00
}
static int bnxt_rss_hash_conf_get_op(struct rte_eth_dev *eth_dev,
struct rte_eth_rss_conf *rss_conf)
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = eth_dev->data->dev_private;
net/bnxt: use macro for getting default VNIC BNXT_GET_DEFAULT_VNIC macro returns the default VNIC of the function and fixed all occurrences of code to use the macro. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-12-02 04:18:09 +00:00
struct bnxt_vnic_info *vnic = BNXT_GET_DEFAULT_VNIC(bp);
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
int len, rc;
net/bnxt: add RSS hash configuration Add rss_hash_update and rss_hash_conf_get dev_ops Uses the bnxt_hwrm_vnic_rss_cfg() HWRM API added in the previous patch. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:37 +00:00
uint32_t hash_types;
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
rc = is_bnxt_in_error(bp);
if (rc)
return rc;
net/bnxt: add RSS hash configuration Add rss_hash_update and rss_hash_conf_get dev_ops Uses the bnxt_hwrm_vnic_rss_cfg() HWRM API added in the previous patch. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:37 +00:00
/* RSS configuration is the same for all VNICs */
if (vnic && vnic->rss_hash_key) {
if (rss_conf->rss_key) {
len = rss_conf->rss_key_len <= HW_HASH_KEY_SIZE ?
rss_conf->rss_key_len : HW_HASH_KEY_SIZE;
memcpy(rss_conf->rss_key, vnic->rss_hash_key, len);
}
hash_types = vnic->hash_type;
rss_conf->rss_hf = 0;
if (hash_types & HWRM_VNIC_RSS_CFG_INPUT_HASH_TYPE_IPV4) {
rss_conf->rss_hf |= ETH_RSS_IPV4;
hash_types &= ~HWRM_VNIC_RSS_CFG_INPUT_HASH_TYPE_IPV4;
}
if (hash_types & HWRM_VNIC_RSS_CFG_INPUT_HASH_TYPE_TCP_IPV4) {
rss_conf->rss_hf |= ETH_RSS_NONFRAG_IPV4_TCP;
hash_types &=
~HWRM_VNIC_RSS_CFG_INPUT_HASH_TYPE_TCP_IPV4;
}
if (hash_types & HWRM_VNIC_RSS_CFG_INPUT_HASH_TYPE_UDP_IPV4) {
rss_conf->rss_hf |= ETH_RSS_NONFRAG_IPV4_UDP;
hash_types &=
~HWRM_VNIC_RSS_CFG_INPUT_HASH_TYPE_UDP_IPV4;
}
if (hash_types & HWRM_VNIC_RSS_CFG_INPUT_HASH_TYPE_IPV6) {
rss_conf->rss_hf |= ETH_RSS_IPV6;
hash_types &= ~HWRM_VNIC_RSS_CFG_INPUT_HASH_TYPE_IPV6;
}
if (hash_types & HWRM_VNIC_RSS_CFG_INPUT_HASH_TYPE_TCP_IPV6) {
rss_conf->rss_hf |= ETH_RSS_NONFRAG_IPV6_TCP;
hash_types &=
~HWRM_VNIC_RSS_CFG_INPUT_HASH_TYPE_TCP_IPV6;
}
if (hash_types & HWRM_VNIC_RSS_CFG_INPUT_HASH_TYPE_UDP_IPV6) {
rss_conf->rss_hf |= ETH_RSS_NONFRAG_IPV6_UDP;
hash_types &=
~HWRM_VNIC_RSS_CFG_INPUT_HASH_TYPE_UDP_IPV6;
}
net/bnxt: support RSS hash selection Add support to select RSS hash based on innermost or outermost headers. If an application is started without any specific settings the default mode configured by FW or HW shall be used. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-18 23:33:41 +00:00
rss_conf->rss_hf |=
bnxt_hwrm_to_rte_rss_level(bp, vnic->hash_mode);
net/bnxt: add RSS hash configuration Add rss_hash_update and rss_hash_conf_get dev_ops Uses the bnxt_hwrm_vnic_rss_cfg() HWRM API added in the previous patch. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:37 +00:00
if (hash_types) {
net/bnxt: use dynamic log type This patch implements driver specific log type doing away with usage of RTE_LOG() for logging. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:55 +00:00
PMD_DRV_LOG(ERR,
fix various typos found by Lintian Cc: stable@dpdk.org Signed-off-by: Luca Boccassi <bluca@debian.org>
2020-02-29 16:37:06 +00:00
"Unknown RSS config from firmware (%08x), RSS disabled",
net/bnxt: add RSS hash configuration Add rss_hash_update and rss_hash_conf_get dev_ops Uses the bnxt_hwrm_vnic_rss_cfg() HWRM API added in the previous patch. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:37 +00:00
vnic->hash_type);
return -ENOTSUP;
}
} else {
rss_conf->rss_hf = 0;
}
return 0;
}
net/bnxt: add flow control operations Add flow_ctrl_get and flow_ctrl_set dev_ops. Uses the bnxt_set_hwrm_link_config() HWRM API added in earlier patch. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:38 +00:00
static int bnxt_flow_ctrl_get_op(struct rte_eth_dev *dev,
drivers/net: remove unnecessary macro for unused variables remove __rte_unused instances that are not required. Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com> Acked-by: Allain Legacy <allain.legacy@windriver.com> Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
2017-05-12 10:33:03 +00:00
struct rte_eth_fc_conf *fc_conf)
net/bnxt: add flow control operations Add flow_ctrl_get and flow_ctrl_set dev_ops. Uses the bnxt_set_hwrm_link_config() HWRM API added in earlier patch. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:38 +00:00
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = dev->data->dev_private;
net/bnxt: add flow control operations Add flow_ctrl_get and flow_ctrl_set dev_ops. Uses the bnxt_set_hwrm_link_config() HWRM API added in earlier patch. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:38 +00:00
struct rte_eth_link link_info;
int rc;
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
rc = is_bnxt_in_error(bp);
if (rc)
return rc;
net/bnxt: add flow control operations Add flow_ctrl_get and flow_ctrl_set dev_ops. Uses the bnxt_set_hwrm_link_config() HWRM API added in earlier patch. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:38 +00:00
rc = bnxt_get_hwrm_link_config(bp, &link_info);
if (rc)
return rc;
memset(fc_conf, 0, sizeof(*fc_conf));
net/bnxt: fix allocation of link info struct Dynamically alloc link info structure. Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 11:05:15 +00:00
if (bp->link_info->auto_pause)
net/bnxt: add flow control operations Add flow_ctrl_get and flow_ctrl_set dev_ops. Uses the bnxt_set_hwrm_link_config() HWRM API added in earlier patch. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:38 +00:00
fc_conf->autoneg = 1;
net/bnxt: fix allocation of link info struct Dynamically alloc link info structure. Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 11:05:15 +00:00
switch (bp->link_info->pause) {
net/bnxt: add flow control operations Add flow_ctrl_get and flow_ctrl_set dev_ops. Uses the bnxt_set_hwrm_link_config() HWRM API added in earlier patch. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:38 +00:00
case 0:
fc_conf->mode = RTE_FC_NONE;
break;
case HWRM_PORT_PHY_QCFG_OUTPUT_PAUSE_TX:
fc_conf->mode = RTE_FC_TX_PAUSE;
break;
case HWRM_PORT_PHY_QCFG_OUTPUT_PAUSE_RX:
fc_conf->mode = RTE_FC_RX_PAUSE;
break;
case (HWRM_PORT_PHY_QCFG_OUTPUT_PAUSE_TX |
HWRM_PORT_PHY_QCFG_OUTPUT_PAUSE_RX):
fc_conf->mode = RTE_FC_FULL;
break;
}
return 0;
}
static int bnxt_flow_ctrl_set_op(struct rte_eth_dev *dev,
struct rte_eth_fc_conf *fc_conf)
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = dev->data->dev_private;
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
int rc;
rc = is_bnxt_in_error(bp);
if (rc)
return rc;
net/bnxt: add flow control operations Add flow_ctrl_get and flow_ctrl_set dev_ops. Uses the bnxt_set_hwrm_link_config() HWRM API added in earlier patch. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:38 +00:00
net/bnxt: fix single PF per port check The check BNXT_SINGLE_PF(bp) returns false for a VF. So there is no extra check needed for VF along with BNXT_SINGLE_PF(bp). Also make error messages more explicit. Fixes: ff947c6ce15f ("net/bnxt: add check for multi host PF per port") Fixes: f86febfb46da ("net/bnxt: support VF") Fixes: 3e12fdb78e82 ("net/bnxt: support VLAN pvid") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-04-29 05:53:00 +00:00
if (!BNXT_SINGLE_PF(bp)) {
PMD_DRV_LOG(ERR,
"Flow Control Settings cannot be modified on VF or on shared PF\n");
net/bnxt: support VF Add support to the bnxt PMD to load on a PCI VF. 1) VF cannot change parameters like - speed, autoneg and pause 2) If the VF MAC address shows up as all 0's it has to be provisioned by the PF in the hypervisor. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-09-26 16:18:46 +00:00
return -ENOTSUP;
}
net/bnxt: support NIC Partitioning Adding code to enable support for NIC Partitioning or NPAR 1.0 As a part of NPAR, we don't allow port settings like speed or flow control to be changed. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-09-26 16:18:16 +00:00
net/bnxt: add flow control operations Add flow_ctrl_get and flow_ctrl_set dev_ops. Uses the bnxt_set_hwrm_link_config() HWRM API added in earlier patch. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:38 +00:00
switch (fc_conf->mode) {
case RTE_FC_NONE:
net/bnxt: fix allocation of link info struct Dynamically alloc link info structure. Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 11:05:15 +00:00
bp->link_info->auto_pause = 0;
bp->link_info->force_pause = 0;
net/bnxt: add flow control operations Add flow_ctrl_get and flow_ctrl_set dev_ops. Uses the bnxt_set_hwrm_link_config() HWRM API added in earlier patch. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:38 +00:00
break;
case RTE_FC_RX_PAUSE:
if (fc_conf->autoneg) {
net/bnxt: fix allocation of link info struct Dynamically alloc link info structure. Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 11:05:15 +00:00
bp->link_info->auto_pause =
net/bnxt: add flow control operations Add flow_ctrl_get and flow_ctrl_set dev_ops. Uses the bnxt_set_hwrm_link_config() HWRM API added in earlier patch. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:38 +00:00
HWRM_PORT_PHY_CFG_INPUT_AUTO_PAUSE_RX;
net/bnxt: fix allocation of link info struct Dynamically alloc link info structure. Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 11:05:15 +00:00
bp->link_info->force_pause = 0;
net/bnxt: add flow control operations Add flow_ctrl_get and flow_ctrl_set dev_ops. Uses the bnxt_set_hwrm_link_config() HWRM API added in earlier patch. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:38 +00:00
} else {
net/bnxt: fix allocation of link info struct Dynamically alloc link info structure. Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 11:05:15 +00:00
bp->link_info->auto_pause = 0;
bp->link_info->force_pause =
net/bnxt: add flow control operations Add flow_ctrl_get and flow_ctrl_set dev_ops. Uses the bnxt_set_hwrm_link_config() HWRM API added in earlier patch. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:38 +00:00
HWRM_PORT_PHY_CFG_INPUT_FORCE_PAUSE_RX;
}
break;
case RTE_FC_TX_PAUSE:
if (fc_conf->autoneg) {
net/bnxt: fix allocation of link info struct Dynamically alloc link info structure. Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 11:05:15 +00:00
bp->link_info->auto_pause =
net/bnxt: add flow control operations Add flow_ctrl_get and flow_ctrl_set dev_ops. Uses the bnxt_set_hwrm_link_config() HWRM API added in earlier patch. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:38 +00:00
HWRM_PORT_PHY_CFG_INPUT_AUTO_PAUSE_TX;
net/bnxt: fix allocation of link info struct Dynamically alloc link info structure. Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 11:05:15 +00:00
bp->link_info->force_pause = 0;
net/bnxt: add flow control operations Add flow_ctrl_get and flow_ctrl_set dev_ops. Uses the bnxt_set_hwrm_link_config() HWRM API added in earlier patch. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:38 +00:00
} else {
net/bnxt: fix allocation of link info struct Dynamically alloc link info structure. Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 11:05:15 +00:00
bp->link_info->auto_pause = 0;
bp->link_info->force_pause =
net/bnxt: add flow control operations Add flow_ctrl_get and flow_ctrl_set dev_ops. Uses the bnxt_set_hwrm_link_config() HWRM API added in earlier patch. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:38 +00:00
HWRM_PORT_PHY_CFG_INPUT_FORCE_PAUSE_TX;
}
break;
case RTE_FC_FULL:
if (fc_conf->autoneg) {
net/bnxt: fix allocation of link info struct Dynamically alloc link info structure. Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 11:05:15 +00:00
bp->link_info->auto_pause =
net/bnxt: add flow control operations Add flow_ctrl_get and flow_ctrl_set dev_ops. Uses the bnxt_set_hwrm_link_config() HWRM API added in earlier patch. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:38 +00:00
HWRM_PORT_PHY_CFG_INPUT_AUTO_PAUSE_TX |
HWRM_PORT_PHY_CFG_INPUT_AUTO_PAUSE_RX;
net/bnxt: fix allocation of link info struct Dynamically alloc link info structure. Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 11:05:15 +00:00
bp->link_info->force_pause = 0;
net/bnxt: add flow control operations Add flow_ctrl_get and flow_ctrl_set dev_ops. Uses the bnxt_set_hwrm_link_config() HWRM API added in earlier patch. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:38 +00:00
} else {
net/bnxt: fix allocation of link info struct Dynamically alloc link info structure. Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 11:05:15 +00:00
bp->link_info->auto_pause = 0;
bp->link_info->force_pause =
net/bnxt: add flow control operations Add flow_ctrl_get and flow_ctrl_set dev_ops. Uses the bnxt_set_hwrm_link_config() HWRM API added in earlier patch. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:38 +00:00
HWRM_PORT_PHY_CFG_INPUT_FORCE_PAUSE_TX |
HWRM_PORT_PHY_CFG_INPUT_FORCE_PAUSE_RX;
}
break;
}
return bnxt_set_hwrm_link_config(bp, true);
}
net/bnxt: support tunneling Add support for udp_tunnel_port_add/del dev_ops to configure a UDP port for VXLAN and Geneve Tunnel protocols. The HWRM supports only one global destination port for a tunnel type, use a reference counter to keep track of its usage. Cache the configured VXLAN/Geneve ports and use that value to check if the right UDP port is being freed up. Skip calling bnxt_hwrm_tunnel_dst_port_alloc if the same UDP port is being programmed. Skip calling bnxt_hwrm_tunnel_dst_port_free if no UDP port has been configured. Also update tx offload capabilities Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:03 +00:00
/* Add UDP tunneling port */
static int
bnxt_udp_tunnel_port_add_op(struct rte_eth_dev *eth_dev,
struct rte_eth_udp_tunnel *udp_tunnel)
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = eth_dev->data->dev_private;
net/bnxt: support tunneling Add support for udp_tunnel_port_add/del dev_ops to configure a UDP port for VXLAN and Geneve Tunnel protocols. The HWRM supports only one global destination port for a tunnel type, use a reference counter to keep track of its usage. Cache the configured VXLAN/Geneve ports and use that value to check if the right UDP port is being freed up. Skip calling bnxt_hwrm_tunnel_dst_port_alloc if the same UDP port is being programmed. Skip calling bnxt_hwrm_tunnel_dst_port_free if no UDP port has been configured. Also update tx offload capabilities Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:03 +00:00
uint16_t tunnel_type = 0;
int rc = 0;
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
rc = is_bnxt_in_error(bp);
if (rc)
return rc;
net/bnxt: support tunneling Add support for udp_tunnel_port_add/del dev_ops to configure a UDP port for VXLAN and Geneve Tunnel protocols. The HWRM supports only one global destination port for a tunnel type, use a reference counter to keep track of its usage. Cache the configured VXLAN/Geneve ports and use that value to check if the right UDP port is being freed up. Skip calling bnxt_hwrm_tunnel_dst_port_alloc if the same UDP port is being programmed. Skip calling bnxt_hwrm_tunnel_dst_port_free if no UDP port has been configured. Also update tx offload capabilities Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:03 +00:00
switch (udp_tunnel->prot_type) {
case RTE_TUNNEL_TYPE_VXLAN:
if (bp->vxlan_port_cnt) {
net/bnxt: use dynamic log type This patch implements driver specific log type doing away with usage of RTE_LOG() for logging. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:55 +00:00
PMD_DRV_LOG(ERR, "Tunnel Port %d already programmed\n",
net/bnxt: support tunneling Add support for udp_tunnel_port_add/del dev_ops to configure a UDP port for VXLAN and Geneve Tunnel protocols. The HWRM supports only one global destination port for a tunnel type, use a reference counter to keep track of its usage. Cache the configured VXLAN/Geneve ports and use that value to check if the right UDP port is being freed up. Skip calling bnxt_hwrm_tunnel_dst_port_alloc if the same UDP port is being programmed. Skip calling bnxt_hwrm_tunnel_dst_port_free if no UDP port has been configured. Also update tx offload capabilities Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:03 +00:00
udp_tunnel->udp_port);
if (bp->vxlan_port != udp_tunnel->udp_port) {
net/bnxt: use dynamic log type This patch implements driver specific log type doing away with usage of RTE_LOG() for logging. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:55 +00:00
PMD_DRV_LOG(ERR, "Only one port allowed\n");
net/bnxt: support tunneling Add support for udp_tunnel_port_add/del dev_ops to configure a UDP port for VXLAN and Geneve Tunnel protocols. The HWRM supports only one global destination port for a tunnel type, use a reference counter to keep track of its usage. Cache the configured VXLAN/Geneve ports and use that value to check if the right UDP port is being freed up. Skip calling bnxt_hwrm_tunnel_dst_port_alloc if the same UDP port is being programmed. Skip calling bnxt_hwrm_tunnel_dst_port_free if no UDP port has been configured. Also update tx offload capabilities Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:03 +00:00
return -ENOSPC;
}
bp->vxlan_port_cnt++;
return 0;
}
tunnel_type =
HWRM_TUNNEL_DST_PORT_ALLOC_INPUT_TUNNEL_TYPE_VXLAN;
bp->vxlan_port_cnt++;
break;
case RTE_TUNNEL_TYPE_GENEVE:
if (bp->geneve_port_cnt) {
net/bnxt: use dynamic log type This patch implements driver specific log type doing away with usage of RTE_LOG() for logging. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:55 +00:00
PMD_DRV_LOG(ERR, "Tunnel Port %d already programmed\n",
net/bnxt: support tunneling Add support for udp_tunnel_port_add/del dev_ops to configure a UDP port for VXLAN and Geneve Tunnel protocols. The HWRM supports only one global destination port for a tunnel type, use a reference counter to keep track of its usage. Cache the configured VXLAN/Geneve ports and use that value to check if the right UDP port is being freed up. Skip calling bnxt_hwrm_tunnel_dst_port_alloc if the same UDP port is being programmed. Skip calling bnxt_hwrm_tunnel_dst_port_free if no UDP port has been configured. Also update tx offload capabilities Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:03 +00:00
udp_tunnel->udp_port);
if (bp->geneve_port != udp_tunnel->udp_port) {
net/bnxt: use dynamic log type This patch implements driver specific log type doing away with usage of RTE_LOG() for logging. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:55 +00:00
PMD_DRV_LOG(ERR, "Only one port allowed\n");
net/bnxt: support tunneling Add support for udp_tunnel_port_add/del dev_ops to configure a UDP port for VXLAN and Geneve Tunnel protocols. The HWRM supports only one global destination port for a tunnel type, use a reference counter to keep track of its usage. Cache the configured VXLAN/Geneve ports and use that value to check if the right UDP port is being freed up. Skip calling bnxt_hwrm_tunnel_dst_port_alloc if the same UDP port is being programmed. Skip calling bnxt_hwrm_tunnel_dst_port_free if no UDP port has been configured. Also update tx offload capabilities Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:03 +00:00
return -ENOSPC;
}
bp->geneve_port_cnt++;
return 0;
}
tunnel_type =
HWRM_TUNNEL_DST_PORT_ALLOC_INPUT_TUNNEL_TYPE_GENEVE;
bp->geneve_port_cnt++;
break;
default:
net/bnxt: use dynamic log type This patch implements driver specific log type doing away with usage of RTE_LOG() for logging. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:55 +00:00
PMD_DRV_LOG(ERR, "Tunnel type is not supported\n");
net/bnxt: support tunneling Add support for udp_tunnel_port_add/del dev_ops to configure a UDP port for VXLAN and Geneve Tunnel protocols. The HWRM supports only one global destination port for a tunnel type, use a reference counter to keep track of its usage. Cache the configured VXLAN/Geneve ports and use that value to check if the right UDP port is being freed up. Skip calling bnxt_hwrm_tunnel_dst_port_alloc if the same UDP port is being programmed. Skip calling bnxt_hwrm_tunnel_dst_port_free if no UDP port has been configured. Also update tx offload capabilities Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:03 +00:00
return -ENOTSUP;
}
rc = bnxt_hwrm_tunnel_dst_port_alloc(bp, udp_tunnel->udp_port,
tunnel_type);
return rc;
}
static int
bnxt_udp_tunnel_port_del_op(struct rte_eth_dev *eth_dev,
struct rte_eth_udp_tunnel *udp_tunnel)
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = eth_dev->data->dev_private;
net/bnxt: support tunneling Add support for udp_tunnel_port_add/del dev_ops to configure a UDP port for VXLAN and Geneve Tunnel protocols. The HWRM supports only one global destination port for a tunnel type, use a reference counter to keep track of its usage. Cache the configured VXLAN/Geneve ports and use that value to check if the right UDP port is being freed up. Skip calling bnxt_hwrm_tunnel_dst_port_alloc if the same UDP port is being programmed. Skip calling bnxt_hwrm_tunnel_dst_port_free if no UDP port has been configured. Also update tx offload capabilities Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:03 +00:00
uint16_t tunnel_type = 0;
uint16_t port = 0;
int rc = 0;
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
rc = is_bnxt_in_error(bp);
if (rc)
return rc;
net/bnxt: support tunneling Add support for udp_tunnel_port_add/del dev_ops to configure a UDP port for VXLAN and Geneve Tunnel protocols. The HWRM supports only one global destination port for a tunnel type, use a reference counter to keep track of its usage. Cache the configured VXLAN/Geneve ports and use that value to check if the right UDP port is being freed up. Skip calling bnxt_hwrm_tunnel_dst_port_alloc if the same UDP port is being programmed. Skip calling bnxt_hwrm_tunnel_dst_port_free if no UDP port has been configured. Also update tx offload capabilities Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:03 +00:00
switch (udp_tunnel->prot_type) {
case RTE_TUNNEL_TYPE_VXLAN:
if (!bp->vxlan_port_cnt) {
net/bnxt: use dynamic log type This patch implements driver specific log type doing away with usage of RTE_LOG() for logging. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:55 +00:00
PMD_DRV_LOG(ERR, "No Tunnel port configured yet\n");
net/bnxt: support tunneling Add support for udp_tunnel_port_add/del dev_ops to configure a UDP port for VXLAN and Geneve Tunnel protocols. The HWRM supports only one global destination port for a tunnel type, use a reference counter to keep track of its usage. Cache the configured VXLAN/Geneve ports and use that value to check if the right UDP port is being freed up. Skip calling bnxt_hwrm_tunnel_dst_port_alloc if the same UDP port is being programmed. Skip calling bnxt_hwrm_tunnel_dst_port_free if no UDP port has been configured. Also update tx offload capabilities Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:03 +00:00
return -EINVAL;
}
if (bp->vxlan_port != udp_tunnel->udp_port) {
net/bnxt: use dynamic log type This patch implements driver specific log type doing away with usage of RTE_LOG() for logging. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:55 +00:00
PMD_DRV_LOG(ERR, "Req Port: %d. Configured port: %d\n",
net/bnxt: support tunneling Add support for udp_tunnel_port_add/del dev_ops to configure a UDP port for VXLAN and Geneve Tunnel protocols. The HWRM supports only one global destination port for a tunnel type, use a reference counter to keep track of its usage. Cache the configured VXLAN/Geneve ports and use that value to check if the right UDP port is being freed up. Skip calling bnxt_hwrm_tunnel_dst_port_alloc if the same UDP port is being programmed. Skip calling bnxt_hwrm_tunnel_dst_port_free if no UDP port has been configured. Also update tx offload capabilities Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:03 +00:00
udp_tunnel->udp_port, bp->vxlan_port);
return -EINVAL;
}
if (--bp->vxlan_port_cnt)
return 0;
tunnel_type =
HWRM_TUNNEL_DST_PORT_FREE_INPUT_TUNNEL_TYPE_VXLAN;
port = bp->vxlan_fw_dst_port_id;
break;
case RTE_TUNNEL_TYPE_GENEVE:
if (!bp->geneve_port_cnt) {
net/bnxt: use dynamic log type This patch implements driver specific log type doing away with usage of RTE_LOG() for logging. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:55 +00:00
PMD_DRV_LOG(ERR, "No Tunnel port configured yet\n");
net/bnxt: support tunneling Add support for udp_tunnel_port_add/del dev_ops to configure a UDP port for VXLAN and Geneve Tunnel protocols. The HWRM supports only one global destination port for a tunnel type, use a reference counter to keep track of its usage. Cache the configured VXLAN/Geneve ports and use that value to check if the right UDP port is being freed up. Skip calling bnxt_hwrm_tunnel_dst_port_alloc if the same UDP port is being programmed. Skip calling bnxt_hwrm_tunnel_dst_port_free if no UDP port has been configured. Also update tx offload capabilities Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:03 +00:00
return -EINVAL;
}
if (bp->geneve_port != udp_tunnel->udp_port) {
net/bnxt: use dynamic log type This patch implements driver specific log type doing away with usage of RTE_LOG() for logging. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:55 +00:00
PMD_DRV_LOG(ERR, "Req Port: %d. Configured port: %d\n",
net/bnxt: support tunneling Add support for udp_tunnel_port_add/del dev_ops to configure a UDP port for VXLAN and Geneve Tunnel protocols. The HWRM supports only one global destination port for a tunnel type, use a reference counter to keep track of its usage. Cache the configured VXLAN/Geneve ports and use that value to check if the right UDP port is being freed up. Skip calling bnxt_hwrm_tunnel_dst_port_alloc if the same UDP port is being programmed. Skip calling bnxt_hwrm_tunnel_dst_port_free if no UDP port has been configured. Also update tx offload capabilities Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:03 +00:00
udp_tunnel->udp_port, bp->geneve_port);
return -EINVAL;
}
if (--bp->geneve_port_cnt)
return 0;
tunnel_type =
HWRM_TUNNEL_DST_PORT_FREE_INPUT_TUNNEL_TYPE_GENEVE;
port = bp->geneve_fw_dst_port_id;
break;
default:
net/bnxt: use dynamic log type This patch implements driver specific log type doing away with usage of RTE_LOG() for logging. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:55 +00:00
PMD_DRV_LOG(ERR, "Tunnel type is not supported\n");
net/bnxt: support tunneling Add support for udp_tunnel_port_add/del dev_ops to configure a UDP port for VXLAN and Geneve Tunnel protocols. The HWRM supports only one global destination port for a tunnel type, use a reference counter to keep track of its usage. Cache the configured VXLAN/Geneve ports and use that value to check if the right UDP port is being freed up. Skip calling bnxt_hwrm_tunnel_dst_port_alloc if the same UDP port is being programmed. Skip calling bnxt_hwrm_tunnel_dst_port_free if no UDP port has been configured. Also update tx offload capabilities Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:03 +00:00
return -ENOTSUP;
}
rc = bnxt_hwrm_tunnel_dst_port_free(bp, port, tunnel_type);
return rc;
}
net/bnxt: support VLAN filter and strip This patch adds VLAN strip and offload callbacks. To add a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists: if VLAN matches vlan_id VLAN filter already exists, just skip and continue else add a new MAC+VLAN filter else Remove the old MAC only filter Add a new MAC+VLAN filter To remove a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists && VLAN matches vlan_id remove the MAC+VLAN filter add a new MAC only filter else VLAN filter doesn't exist, just skip and continue Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:05 +00:00
static int bnxt_del_vlan_filter(struct bnxt *bp, uint16_t vlan_id)
{
net/bnxt: fix MAC/VLAN filter allocation We were adding the VLAN filters to all the VNICs of the function. Also, we were adding these VLANs to all the existing MAC only filters. This was resulting in fewer VLANs getting added. By default we should allocate MAC+VLAN filter only to the default VNIC of the function using the default mac address. Similar logic was followed in the VLAN deletion code. This patch fixes it. Use inner VLAN fields instead of outer VLAN during filter deletion to be in sync with VLAN addition code. Fixes: 246c5cc5f05e ("net/bnxt: use correct flags during VLAN configuration") Cc: stable@dpdk.org Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-18 03:36:14 +00:00
struct bnxt_filter_info *filter;
net/bnxt: support VLAN filter and strip This patch adds VLAN strip and offload callbacks. To add a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists: if VLAN matches vlan_id VLAN filter already exists, just skip and continue else add a new MAC+VLAN filter else Remove the old MAC only filter Add a new MAC+VLAN filter To remove a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists && VLAN matches vlan_id remove the MAC+VLAN filter add a new MAC only filter else VLAN filter doesn't exist, just skip and continue Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:05 +00:00
struct bnxt_vnic_info *vnic;
int rc = 0;
net/bnxt: fix MAC/VLAN filter allocation We were adding the VLAN filters to all the VNICs of the function. Also, we were adding these VLANs to all the existing MAC only filters. This was resulting in fewer VLANs getting added. By default we should allocate MAC+VLAN filter only to the default VNIC of the function using the default mac address. Similar logic was followed in the VLAN deletion code. This patch fixes it. Use inner VLAN fields instead of outer VLAN during filter deletion to be in sync with VLAN addition code. Fixes: 246c5cc5f05e ("net/bnxt: use correct flags during VLAN configuration") Cc: stable@dpdk.org Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-18 03:36:14 +00:00
uint32_t chk = HWRM_CFA_L2_FILTER_ALLOC_INPUT_ENABLES_L2_IVLAN;
net/bnxt: get rid of ff pools and use VNIC info array There was no direct association between the rxq's VNIC and the vnic_info[]. Explicitly associate the two in bnxt_mq_rx_configure(). Fixes: 0a256e4a548b ("net/bnxt: fix Rx ring count limitation") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
2018-09-29 01:59:52 +00:00
net/bnxt: fix MAC/VLAN filter allocation We were adding the VLAN filters to all the VNICs of the function. Also, we were adding these VLANs to all the existing MAC only filters. This was resulting in fewer VLANs getting added. By default we should allocate MAC+VLAN filter only to the default VNIC of the function using the default mac address. Similar logic was followed in the VLAN deletion code. This patch fixes it. Use inner VLAN fields instead of outer VLAN during filter deletion to be in sync with VLAN addition code. Fixes: 246c5cc5f05e ("net/bnxt: use correct flags during VLAN configuration") Cc: stable@dpdk.org Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-18 03:36:14 +00:00
vnic = BNXT_GET_DEFAULT_VNIC(bp);
filter = STAILQ_FIRST(&vnic->filter);
while (filter) {
/* Search for this matching MAC+VLAN filter */
net/bnxt: fix redundant MAC address check filter->mac_index is used to check, if a same mac is already programmed. Hence, filter->dflt member is not needed which is also used for mac addr redundancy check. This patch fixes it by moving mac_index based redundant check from bnxt_mac_addr_add_op to bnxt_add_mac_filter Fixes: 6118503d8071 ("net/bnxt: fix VLAN filtering") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-11-04 10:02:37 +00:00
if (bnxt_vlan_filter_exists(bp, filter, chk, vlan_id)) {
net/bnxt: fix MAC/VLAN filter allocation We were adding the VLAN filters to all the VNICs of the function. Also, we were adding these VLANs to all the existing MAC only filters. This was resulting in fewer VLANs getting added. By default we should allocate MAC+VLAN filter only to the default VNIC of the function using the default mac address. Similar logic was followed in the VLAN deletion code. This patch fixes it. Use inner VLAN fields instead of outer VLAN during filter deletion to be in sync with VLAN addition code. Fixes: 246c5cc5f05e ("net/bnxt: use correct flags during VLAN configuration") Cc: stable@dpdk.org Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-18 03:36:14 +00:00
/* Delete the filter */
rc = bnxt_hwrm_clear_l2_filter(bp, filter);
if (rc)
return rc;
STAILQ_REMOVE(&vnic->filter, filter,
bnxt_filter_info, next);
net/bnxt: fix redundant MAC address check filter->mac_index is used to check, if a same mac is already programmed. Hence, filter->dflt member is not needed which is also used for mac addr redundancy check. This patch fixes it by moving mac_index based redundant check from bnxt_mac_addr_add_op to bnxt_add_mac_filter Fixes: 6118503d8071 ("net/bnxt: fix VLAN filtering") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-11-04 10:02:37 +00:00
bnxt_free_filter(bp, filter);
net/bnxt: fix MAC/VLAN filter allocation We were adding the VLAN filters to all the VNICs of the function. Also, we were adding these VLANs to all the existing MAC only filters. This was resulting in fewer VLANs getting added. By default we should allocate MAC+VLAN filter only to the default VNIC of the function using the default mac address. Similar logic was followed in the VLAN deletion code. This patch fixes it. Use inner VLAN fields instead of outer VLAN during filter deletion to be in sync with VLAN addition code. Fixes: 246c5cc5f05e ("net/bnxt: use correct flags during VLAN configuration") Cc: stable@dpdk.org Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-18 03:36:14 +00:00
PMD_DRV_LOG(INFO,
net/bnxt: fix redundant MAC address check filter->mac_index is used to check, if a same mac is already programmed. Hence, filter->dflt member is not needed which is also used for mac addr redundancy check. This patch fixes it by moving mac_index based redundant check from bnxt_mac_addr_add_op to bnxt_add_mac_filter Fixes: 6118503d8071 ("net/bnxt: fix VLAN filtering") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-11-04 10:02:37 +00:00
"Deleted vlan filter for %d\n",
net/bnxt: fix MAC/VLAN filter allocation We were adding the VLAN filters to all the VNICs of the function. Also, we were adding these VLANs to all the existing MAC only filters. This was resulting in fewer VLANs getting added. By default we should allocate MAC+VLAN filter only to the default VNIC of the function using the default mac address. Similar logic was followed in the VLAN deletion code. This patch fixes it. Use inner VLAN fields instead of outer VLAN during filter deletion to be in sync with VLAN addition code. Fixes: 246c5cc5f05e ("net/bnxt: use correct flags during VLAN configuration") Cc: stable@dpdk.org Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-18 03:36:14 +00:00
vlan_id);
net/bnxt: fix redundant MAC address check filter->mac_index is used to check, if a same mac is already programmed. Hence, filter->dflt member is not needed which is also used for mac addr redundancy check. This patch fixes it by moving mac_index based redundant check from bnxt_mac_addr_add_op to bnxt_add_mac_filter Fixes: 6118503d8071 ("net/bnxt: fix VLAN filtering") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-11-04 10:02:37 +00:00
return 0;
net/bnxt: support VLAN filter and strip This patch adds VLAN strip and offload callbacks. To add a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists: if VLAN matches vlan_id VLAN filter already exists, just skip and continue else add a new MAC+VLAN filter else Remove the old MAC only filter Add a new MAC+VLAN filter To remove a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists && VLAN matches vlan_id remove the MAC+VLAN filter add a new MAC only filter else VLAN filter doesn't exist, just skip and continue Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:05 +00:00
}
net/bnxt: fix MAC/VLAN filter allocation We were adding the VLAN filters to all the VNICs of the function. Also, we were adding these VLANs to all the existing MAC only filters. This was resulting in fewer VLANs getting added. By default we should allocate MAC+VLAN filter only to the default VNIC of the function using the default mac address. Similar logic was followed in the VLAN deletion code. This patch fixes it. Use inner VLAN fields instead of outer VLAN during filter deletion to be in sync with VLAN addition code. Fixes: 246c5cc5f05e ("net/bnxt: use correct flags during VLAN configuration") Cc: stable@dpdk.org Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-18 03:36:14 +00:00
filter = STAILQ_NEXT(filter, next);
net/bnxt: support VLAN filter and strip This patch adds VLAN strip and offload callbacks. To add a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists: if VLAN matches vlan_id VLAN filter already exists, just skip and continue else add a new MAC+VLAN filter else Remove the old MAC only filter Add a new MAC+VLAN filter To remove a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists && VLAN matches vlan_id remove the MAC+VLAN filter add a new MAC only filter else VLAN filter doesn't exist, just skip and continue Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:05 +00:00
}
net/bnxt: fix MAC/VLAN filter allocation We were adding the VLAN filters to all the VNICs of the function. Also, we were adding these VLANs to all the existing MAC only filters. This was resulting in fewer VLANs getting added. By default we should allocate MAC+VLAN filter only to the default VNIC of the function using the default mac address. Similar logic was followed in the VLAN deletion code. This patch fixes it. Use inner VLAN fields instead of outer VLAN during filter deletion to be in sync with VLAN addition code. Fixes: 246c5cc5f05e ("net/bnxt: use correct flags during VLAN configuration") Cc: stable@dpdk.org Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-18 03:36:14 +00:00
return -ENOENT;
net/bnxt: support VLAN filter and strip This patch adds VLAN strip and offload callbacks. To add a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists: if VLAN matches vlan_id VLAN filter already exists, just skip and continue else add a new MAC+VLAN filter else Remove the old MAC only filter Add a new MAC+VLAN filter To remove a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists && VLAN matches vlan_id remove the MAC+VLAN filter add a new MAC only filter else VLAN filter doesn't exist, just skip and continue Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:05 +00:00
}
static int bnxt_add_vlan_filter(struct bnxt *bp, uint16_t vlan_id)
{
net/bnxt: fix MAC/VLAN filter allocation We were adding the VLAN filters to all the VNICs of the function. Also, we were adding these VLANs to all the existing MAC only filters. This was resulting in fewer VLANs getting added. By default we should allocate MAC+VLAN filter only to the default VNIC of the function using the default mac address. Similar logic was followed in the VLAN deletion code. This patch fixes it. Use inner VLAN fields instead of outer VLAN during filter deletion to be in sync with VLAN addition code. Fixes: 246c5cc5f05e ("net/bnxt: use correct flags during VLAN configuration") Cc: stable@dpdk.org Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-18 03:36:14 +00:00
struct bnxt_filter_info *filter;
net/bnxt: support VLAN filter and strip This patch adds VLAN strip and offload callbacks. To add a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists: if VLAN matches vlan_id VLAN filter already exists, just skip and continue else add a new MAC+VLAN filter else Remove the old MAC only filter Add a new MAC+VLAN filter To remove a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists && VLAN matches vlan_id remove the MAC+VLAN filter add a new MAC only filter else VLAN filter doesn't exist, just skip and continue Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:05 +00:00
struct bnxt_vnic_info *vnic;
int rc = 0;
net/bnxt: use correct flags during VLAN configuration Setting of VLAN filter cmd was being done with incorrect flag value. We need to use inner vlan fields instead of outer vlan. Fixes: 7fe5668d2ea3 ("net/bnxt: support VLAN filter and strip") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-06-28 20:15:49 +00:00
uint32_t en = HWRM_CFA_L2_FILTER_ALLOC_INPUT_ENABLES_L2_IVLAN |
HWRM_CFA_L2_FILTER_ALLOC_INPUT_ENABLES_L2_IVLAN_MASK;
uint32_t chk = HWRM_CFA_L2_FILTER_ALLOC_INPUT_ENABLES_L2_IVLAN;
net/bnxt: support VLAN filter and strip This patch adds VLAN strip and offload callbacks. To add a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists: if VLAN matches vlan_id VLAN filter already exists, just skip and continue else add a new MAC+VLAN filter else Remove the old MAC only filter Add a new MAC+VLAN filter To remove a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists && VLAN matches vlan_id remove the MAC+VLAN filter add a new MAC only filter else VLAN filter doesn't exist, just skip and continue Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:05 +00:00
net/bnxt: fix MAC/VLAN filter allocation We were adding the VLAN filters to all the VNICs of the function. Also, we were adding these VLANs to all the existing MAC only filters. This was resulting in fewer VLANs getting added. By default we should allocate MAC+VLAN filter only to the default VNIC of the function using the default mac address. Similar logic was followed in the VLAN deletion code. This patch fixes it. Use inner VLAN fields instead of outer VLAN during filter deletion to be in sync with VLAN addition code. Fixes: 246c5cc5f05e ("net/bnxt: use correct flags during VLAN configuration") Cc: stable@dpdk.org Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-18 03:36:14 +00:00
/* Implementation notes on the use of VNIC in this command:
*
* By default, these filters belong to default vnic for the function.
* Once these filters are set up, only destination VNIC can be modified.
* If the destination VNIC is not specified in this command,
* then the HWRM shall only create an l2 context id.
*/
net/bnxt: get rid of ff pools and use VNIC info array There was no direct association between the rxq's VNIC and the vnic_info[]. Explicitly associate the two in bnxt_mq_rx_configure(). Fixes: 0a256e4a548b ("net/bnxt: fix Rx ring count limitation") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
2018-09-29 01:59:52 +00:00
net/bnxt: fix MAC/VLAN filter allocation We were adding the VLAN filters to all the VNICs of the function. Also, we were adding these VLANs to all the existing MAC only filters. This was resulting in fewer VLANs getting added. By default we should allocate MAC+VLAN filter only to the default VNIC of the function using the default mac address. Similar logic was followed in the VLAN deletion code. This patch fixes it. Use inner VLAN fields instead of outer VLAN during filter deletion to be in sync with VLAN addition code. Fixes: 246c5cc5f05e ("net/bnxt: use correct flags during VLAN configuration") Cc: stable@dpdk.org Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-18 03:36:14 +00:00
vnic = BNXT_GET_DEFAULT_VNIC(bp);
filter = STAILQ_FIRST(&vnic->filter);
/* Check if the VLAN has already been added */
while (filter) {
net/bnxt: fix redundant MAC address check filter->mac_index is used to check, if a same mac is already programmed. Hence, filter->dflt member is not needed which is also used for mac addr redundancy check. This patch fixes it by moving mac_index based redundant check from bnxt_mac_addr_add_op to bnxt_add_mac_filter Fixes: 6118503d8071 ("net/bnxt: fix VLAN filtering") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-11-04 10:02:37 +00:00
if (bnxt_vlan_filter_exists(bp, filter, chk, vlan_id))
net/bnxt: fix MAC/VLAN filter allocation We were adding the VLAN filters to all the VNICs of the function. Also, we were adding these VLANs to all the existing MAC only filters. This was resulting in fewer VLANs getting added. By default we should allocate MAC+VLAN filter only to the default VNIC of the function using the default mac address. Similar logic was followed in the VLAN deletion code. This patch fixes it. Use inner VLAN fields instead of outer VLAN during filter deletion to be in sync with VLAN addition code. Fixes: 246c5cc5f05e ("net/bnxt: use correct flags during VLAN configuration") Cc: stable@dpdk.org Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-18 03:36:14 +00:00
return -EEXIST;
filter = STAILQ_NEXT(filter, next);
}
/* No match found. Alloc a fresh filter and issue the L2_FILTER_ALLOC
* command to create MAC+VLAN filter with the right flags, enables set.
*/
filter = bnxt_alloc_filter(bp);
if (!filter) {
PMD_DRV_LOG(ERR,
"MAC/VLAN filter alloc failed\n");
return -ENOMEM;
net/bnxt: support VLAN filter and strip This patch adds VLAN strip and offload callbacks. To add a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists: if VLAN matches vlan_id VLAN filter already exists, just skip and continue else add a new MAC+VLAN filter else Remove the old MAC only filter Add a new MAC+VLAN filter To remove a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists && VLAN matches vlan_id remove the MAC+VLAN filter add a new MAC only filter else VLAN filter doesn't exist, just skip and continue Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:05 +00:00
}
net/bnxt: fix MAC/VLAN filter allocation We were adding the VLAN filters to all the VNICs of the function. Also, we were adding these VLANs to all the existing MAC only filters. This was resulting in fewer VLANs getting added. By default we should allocate MAC+VLAN filter only to the default VNIC of the function using the default mac address. Similar logic was followed in the VLAN deletion code. This patch fixes it. Use inner VLAN fields instead of outer VLAN during filter deletion to be in sync with VLAN addition code. Fixes: 246c5cc5f05e ("net/bnxt: use correct flags during VLAN configuration") Cc: stable@dpdk.org Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-18 03:36:14 +00:00
/* MAC + VLAN ID filter */
net/bnxt: fix VLAN filtering Currently, when hw-vlan-filter is enabled on testpmd, driver is receiving all vlan packets. Instead, it should only receive untagged packets and vlan packets for which the VLAN filter is programmed. This is because, the default rule to match on MAC is not getting deleted, when hw-vlan-filter is ON. This patch fixes the problem, by deleting the default MAC rule and programming a new rule to receive only untagged packets, when hw-vlan-filter is enabled & another rule for each vlan, as and when that vlan is configured on that port. Fixes: 246c5cc5f05e ("net/bnxt: use correct flags during VLAN configuration") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 23:25:59 +00:00
/* If l2_ivlan == 0 and l2_ivlan_mask != 0, only
* untagged packets are received
*
* If l2_ivlan != 0 and l2_ivlan_mask != 0, untagged
* packets and only the programmed vlan's packets are received
*/
net/bnxt: fix MAC/VLAN filter allocation We were adding the VLAN filters to all the VNICs of the function. Also, we were adding these VLANs to all the existing MAC only filters. This was resulting in fewer VLANs getting added. By default we should allocate MAC+VLAN filter only to the default VNIC of the function using the default mac address. Similar logic was followed in the VLAN deletion code. This patch fixes it. Use inner VLAN fields instead of outer VLAN during filter deletion to be in sync with VLAN addition code. Fixes: 246c5cc5f05e ("net/bnxt: use correct flags during VLAN configuration") Cc: stable@dpdk.org Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-18 03:36:14 +00:00
filter->l2_ivlan = vlan_id;
filter->l2_ivlan_mask = 0x0FFF;
filter->enables |= en;
net/bnxt: support creating SMAC and inner DMAC filters We are currently creating only outer DMAC filters. Create SMAC and inner DMAC filters using HWRM_CFA_L2_FILTER_ALLOC. For this the HWRM_CFA_L2_FILTER_ALLOC has already been updated. Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Rahul Gupta <rahul.gupta@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 23:25:45 +00:00
filter->flags |= HWRM_CFA_L2_FILTER_ALLOC_INPUT_FLAGS_OUTERMOST;
net/bnxt: fix VLAN filtering Currently, when hw-vlan-filter is enabled on testpmd, driver is receiving all vlan packets. Instead, it should only receive untagged packets and vlan packets for which the VLAN filter is programmed. This is because, the default rule to match on MAC is not getting deleted, when hw-vlan-filter is ON. This patch fixes the problem, by deleting the default MAC rule and programming a new rule to receive only untagged packets, when hw-vlan-filter is enabled & another rule for each vlan, as and when that vlan is configured on that port. Fixes: 246c5cc5f05e ("net/bnxt: use correct flags during VLAN configuration") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 23:25:59 +00:00
net/bnxt: fix MAC/VLAN filter allocation We were adding the VLAN filters to all the VNICs of the function. Also, we were adding these VLANs to all the existing MAC only filters. This was resulting in fewer VLANs getting added. By default we should allocate MAC+VLAN filter only to the default VNIC of the function using the default mac address. Similar logic was followed in the VLAN deletion code. This patch fixes it. Use inner VLAN fields instead of outer VLAN during filter deletion to be in sync with VLAN addition code. Fixes: 246c5cc5f05e ("net/bnxt: use correct flags during VLAN configuration") Cc: stable@dpdk.org Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-18 03:36:14 +00:00
rc = bnxt_hwrm_set_l2_filter(bp, vnic->fw_vnic_id, filter);
if (rc) {
/* Free the newly allocated filter as we were
* not able to create the filter in hardware.
*/
net/bnxt: fix redundant MAC address check filter->mac_index is used to check, if a same mac is already programmed. Hence, filter->dflt member is not needed which is also used for mac addr redundancy check. This patch fixes it by moving mac_index based redundant check from bnxt_mac_addr_add_op to bnxt_add_mac_filter Fixes: 6118503d8071 ("net/bnxt: fix VLAN filtering") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-11-04 10:02:37 +00:00
bnxt_free_filter(bp, filter);
net/bnxt: fix MAC/VLAN filter allocation We were adding the VLAN filters to all the VNICs of the function. Also, we were adding these VLANs to all the existing MAC only filters. This was resulting in fewer VLANs getting added. By default we should allocate MAC+VLAN filter only to the default VNIC of the function using the default mac address. Similar logic was followed in the VLAN deletion code. This patch fixes it. Use inner VLAN fields instead of outer VLAN during filter deletion to be in sync with VLAN addition code. Fixes: 246c5cc5f05e ("net/bnxt: use correct flags during VLAN configuration") Cc: stable@dpdk.org Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-18 03:36:14 +00:00
return rc;
}
net/bnxt: fix redundant MAC address check filter->mac_index is used to check, if a same mac is already programmed. Hence, filter->dflt member is not needed which is also used for mac addr redundancy check. This patch fixes it by moving mac_index based redundant check from bnxt_mac_addr_add_op to bnxt_add_mac_filter Fixes: 6118503d8071 ("net/bnxt: fix VLAN filtering") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-11-04 10:02:37 +00:00
filter->mac_index = 0;
/* Add this new filter to the list */
if (vlan_id == 0)
STAILQ_INSERT_HEAD(&vnic->filter, filter, next);
else
STAILQ_INSERT_TAIL(&vnic->filter, filter, next);
net/bnxt: fix MAC/VLAN filter allocation We were adding the VLAN filters to all the VNICs of the function. Also, we were adding these VLANs to all the existing MAC only filters. This was resulting in fewer VLANs getting added. By default we should allocate MAC+VLAN filter only to the default VNIC of the function using the default mac address. Similar logic was followed in the VLAN deletion code. This patch fixes it. Use inner VLAN fields instead of outer VLAN during filter deletion to be in sync with VLAN addition code. Fixes: 246c5cc5f05e ("net/bnxt: use correct flags during VLAN configuration") Cc: stable@dpdk.org Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-18 03:36:14 +00:00
PMD_DRV_LOG(INFO,
"Added Vlan filter for %d\n", vlan_id);
net/bnxt: support VLAN filter and strip This patch adds VLAN strip and offload callbacks. To add a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists: if VLAN matches vlan_id VLAN filter already exists, just skip and continue else add a new MAC+VLAN filter else Remove the old MAC only filter Add a new MAC+VLAN filter To remove a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists && VLAN matches vlan_id remove the MAC+VLAN filter add a new MAC only filter else VLAN filter doesn't exist, just skip and continue Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:05 +00:00
return rc;
}
static int bnxt_vlan_filter_set_op(struct rte_eth_dev *eth_dev,
net/bnxt: get rid of ff pools and use VNIC info array There was no direct association between the rxq's VNIC and the vnic_info[]. Explicitly associate the two in bnxt_mq_rx_configure(). Fixes: 0a256e4a548b ("net/bnxt: fix Rx ring count limitation") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
2018-09-29 01:59:52 +00:00
uint16_t vlan_id, int on)
net/bnxt: support VLAN filter and strip This patch adds VLAN strip and offload callbacks. To add a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists: if VLAN matches vlan_id VLAN filter already exists, just skip and continue else add a new MAC+VLAN filter else Remove the old MAC only filter Add a new MAC+VLAN filter To remove a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists && VLAN matches vlan_id remove the MAC+VLAN filter add a new MAC only filter else VLAN filter doesn't exist, just skip and continue Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:05 +00:00
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = eth_dev->data->dev_private;
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
int rc;
rc = is_bnxt_in_error(bp);
if (rc)
return rc;
net/bnxt: support VLAN filter and strip This patch adds VLAN strip and offload callbacks. To add a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists: if VLAN matches vlan_id VLAN filter already exists, just skip and continue else add a new MAC+VLAN filter else Remove the old MAC only filter Add a new MAC+VLAN filter To remove a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists && VLAN matches vlan_id remove the MAC+VLAN filter add a new MAC only filter else VLAN filter doesn't exist, just skip and continue Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:05 +00:00
net/bnxt: fix VLAN add when port is stopped Driver destroys the vnic when the port is brought down. When user tries to add a vlan when port is stopped, driver issues HWRM command to FW with invalid vnic_id and it fails. Fixed to return an error while setting vlan when port is not started. Fixes: b4e190d55c25 ("net/bnxt: fix MAC/VLAN filter allocation") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-04-21 21:33:49 +00:00
if (!eth_dev->data->dev_started) {
PMD_DRV_LOG(ERR, "port must be started before setting vlan\n");
return -EINVAL;
}
net/bnxt: support VLAN filter and strip This patch adds VLAN strip and offload callbacks. To add a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists: if VLAN matches vlan_id VLAN filter already exists, just skip and continue else add a new MAC+VLAN filter else Remove the old MAC only filter Add a new MAC+VLAN filter To remove a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists && VLAN matches vlan_id remove the MAC+VLAN filter add a new MAC only filter else VLAN filter doesn't exist, just skip and continue Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:05 +00:00
/* These operations apply to ALL existing MAC/VLAN filters */
if (on)
return bnxt_add_vlan_filter(bp, vlan_id);
else
return bnxt_del_vlan_filter(bp, vlan_id);
}
net/bnxt: fix VLAN filtering Currently, when hw-vlan-filter is enabled on testpmd, driver is receiving all vlan packets. Instead, it should only receive untagged packets and vlan packets for which the VLAN filter is programmed. This is because, the default rule to match on MAC is not getting deleted, when hw-vlan-filter is ON. This patch fixes the problem, by deleting the default MAC rule and programming a new rule to receive only untagged packets, when hw-vlan-filter is enabled & another rule for each vlan, as and when that vlan is configured on that port. Fixes: 246c5cc5f05e ("net/bnxt: use correct flags during VLAN configuration") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 23:25:59 +00:00
static int bnxt_del_dflt_mac_filter(struct bnxt *bp,
struct bnxt_vnic_info *vnic)
{
struct bnxt_filter_info *filter;
int rc;
filter = STAILQ_FIRST(&vnic->filter);
while (filter) {
net/bnxt: fix redundant MAC address check filter->mac_index is used to check, if a same mac is already programmed. Hence, filter->dflt member is not needed which is also used for mac addr redundancy check. This patch fixes it by moving mac_index based redundant check from bnxt_mac_addr_add_op to bnxt_add_mac_filter Fixes: 6118503d8071 ("net/bnxt: fix VLAN filtering") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-11-04 10:02:37 +00:00
if (filter->mac_index == 0 &&
net/bnxt: fix VLAN filtering Currently, when hw-vlan-filter is enabled on testpmd, driver is receiving all vlan packets. Instead, it should only receive untagged packets and vlan packets for which the VLAN filter is programmed. This is because, the default rule to match on MAC is not getting deleted, when hw-vlan-filter is ON. This patch fixes the problem, by deleting the default MAC rule and programming a new rule to receive only untagged packets, when hw-vlan-filter is enabled & another rule for each vlan, as and when that vlan is configured on that port. Fixes: 246c5cc5f05e ("net/bnxt: use correct flags during VLAN configuration") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 23:25:59 +00:00
!memcmp(filter->l2_addr, bp->mac_addr,
RTE_ETHER_ADDR_LEN)) {
rc = bnxt_hwrm_clear_l2_filter(bp, filter);
net/bnxt: fix redundant MAC address check filter->mac_index is used to check, if a same mac is already programmed. Hence, filter->dflt member is not needed which is also used for mac addr redundancy check. This patch fixes it by moving mac_index based redundant check from bnxt_mac_addr_add_op to bnxt_add_mac_filter Fixes: 6118503d8071 ("net/bnxt: fix VLAN filtering") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-11-04 10:02:37 +00:00
if (!rc) {
STAILQ_REMOVE(&vnic->filter, filter,
bnxt_filter_info, next);
bnxt_free_filter(bp, filter);
}
return rc;
net/bnxt: fix VLAN filtering Currently, when hw-vlan-filter is enabled on testpmd, driver is receiving all vlan packets. Instead, it should only receive untagged packets and vlan packets for which the VLAN filter is programmed. This is because, the default rule to match on MAC is not getting deleted, when hw-vlan-filter is ON. This patch fixes the problem, by deleting the default MAC rule and programming a new rule to receive only untagged packets, when hw-vlan-filter is enabled & another rule for each vlan, as and when that vlan is configured on that port. Fixes: 246c5cc5f05e ("net/bnxt: use correct flags during VLAN configuration") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 23:25:59 +00:00
}
filter = STAILQ_NEXT(filter, next);
}
return 0;
}
ethdev: allow returning error on VLAN offload ops Some devices may not support or fail setting VLAN offload configuration based on dynamic circumstances so the vlan_offload_set_t vector is modified to return an int so the caller can determine success or not. rte_eth_dev_set_vlan_offload is updated to return the value provided by the vector when called along with restoring the original offload configs on failure. Existing vlan_offload_set_t vectors are modified to return an int. Majority of cases return 0 but a few that actually can fail now return their failure codes. Finally, a vlan_offload_set_t vector is added to virtio to facilitate dynamically turning VLAN strip on or off. Signed-off-by: David Harton <dharton@cisco.com> Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
2017-09-01 02:36:28 +00:00
static int
net/bnxt: fix enable/disable VLAN filtering There is no condition check for the user requested operation for VLAN filtering. As a result, VLAN filtering is getting disabled when the user enables/disables VLAN stripping on same port. The function bnxt_hwrm_clear_l2_filter() didn't actually free L2 filter in HW if the reference count of filter is zero. Fixed it by incrementing the reference count of filter in bnxt_alloc_filter() routine. Because of the recent changes in bnxt_hwrm_clear_l2_filter(), change was needed in the routine bnxt_set_default_mac_addr_op() to destroy and re-create the default filter when the user changes the default MAC of the port. Fixes: 5c1171c97216 ("net/bnxt: refactor filter/flow") Fixes: 6118503d8071 ("net/bnxt: fix VLAN filtering") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-01-16 13:04:49 +00:00
bnxt_config_vlan_hw_filter(struct bnxt *bp, uint64_t rx_offloads)
net/bnxt: support VLAN filter and strip This patch adds VLAN strip and offload callbacks. To add a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists: if VLAN matches vlan_id VLAN filter already exists, just skip and continue else add a new MAC+VLAN filter else Remove the old MAC only filter Add a new MAC+VLAN filter To remove a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists && VLAN matches vlan_id remove the MAC+VLAN filter add a new MAC only filter else VLAN filter doesn't exist, just skip and continue Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:05 +00:00
{
net/bnxt: fix VLAN filtering Currently, when hw-vlan-filter is enabled on testpmd, driver is receiving all vlan packets. Instead, it should only receive untagged packets and vlan packets for which the VLAN filter is programmed. This is because, the default rule to match on MAC is not getting deleted, when hw-vlan-filter is ON. This patch fixes the problem, by deleting the default MAC rule and programming a new rule to receive only untagged packets, when hw-vlan-filter is enabled & another rule for each vlan, as and when that vlan is configured on that port. Fixes: 246c5cc5f05e ("net/bnxt: use correct flags during VLAN configuration") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 23:25:59 +00:00
struct bnxt_vnic_info *vnic;
net/bnxt: support VLAN filter and strip This patch adds VLAN strip and offload callbacks. To add a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists: if VLAN matches vlan_id VLAN filter already exists, just skip and continue else add a new MAC+VLAN filter else Remove the old MAC only filter Add a new MAC+VLAN filter To remove a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists && VLAN matches vlan_id remove the MAC+VLAN filter add a new MAC only filter else VLAN filter doesn't exist, just skip and continue Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:05 +00:00
unsigned int i;
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
int rc;
net/bnxt: fix VLAN filtering Currently, when hw-vlan-filter is enabled on testpmd, driver is receiving all vlan packets. Instead, it should only receive untagged packets and vlan packets for which the VLAN filter is programmed. This is because, the default rule to match on MAC is not getting deleted, when hw-vlan-filter is ON. This patch fixes the problem, by deleting the default MAC rule and programming a new rule to receive only untagged packets, when hw-vlan-filter is enabled & another rule for each vlan, as and when that vlan is configured on that port. Fixes: 246c5cc5f05e ("net/bnxt: use correct flags during VLAN configuration") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 23:25:59 +00:00
vnic = BNXT_GET_DEFAULT_VNIC(bp);
if (!(rx_offloads & DEV_RX_OFFLOAD_VLAN_FILTER)) {
/* Remove any VLAN filters programmed */
net/bnxt: fix redundant MAC address check filter->mac_index is used to check, if a same mac is already programmed. Hence, filter->dflt member is not needed which is also used for mac addr redundancy check. This patch fixes it by moving mac_index based redundant check from bnxt_mac_addr_add_op to bnxt_add_mac_filter Fixes: 6118503d8071 ("net/bnxt: fix VLAN filtering") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-11-04 10:02:37 +00:00
for (i = 0; i < RTE_ETHER_MAX_VLAN_ID; i++)
net/bnxt: fix VLAN filtering Currently, when hw-vlan-filter is enabled on testpmd, driver is receiving all vlan packets. Instead, it should only receive untagged packets and vlan packets for which the VLAN filter is programmed. This is because, the default rule to match on MAC is not getting deleted, when hw-vlan-filter is ON. This patch fixes the problem, by deleting the default MAC rule and programming a new rule to receive only untagged packets, when hw-vlan-filter is enabled & another rule for each vlan, as and when that vlan is configured on that port. Fixes: 246c5cc5f05e ("net/bnxt: use correct flags during VLAN configuration") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 23:25:59 +00:00
bnxt_del_vlan_filter(bp, i);
net/bnxt: fix redundant MAC address check filter->mac_index is used to check, if a same mac is already programmed. Hence, filter->dflt member is not needed which is also used for mac addr redundancy check. This patch fixes it by moving mac_index based redundant check from bnxt_mac_addr_add_op to bnxt_add_mac_filter Fixes: 6118503d8071 ("net/bnxt: fix VLAN filtering") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-11-04 10:02:37 +00:00
rc = bnxt_add_mac_filter(bp, vnic, NULL, 0, 0);
net/bnxt: fix VLAN filtering Currently, when hw-vlan-filter is enabled on testpmd, driver is receiving all vlan packets. Instead, it should only receive untagged packets and vlan packets for which the VLAN filter is programmed. This is because, the default rule to match on MAC is not getting deleted, when hw-vlan-filter is ON. This patch fixes the problem, by deleting the default MAC rule and programming a new rule to receive only untagged packets, when hw-vlan-filter is enabled & another rule for each vlan, as and when that vlan is configured on that port. Fixes: 246c5cc5f05e ("net/bnxt: use correct flags during VLAN configuration") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 23:25:59 +00:00
if (rc)
return rc;
} else {
/* Default filter will allow packets that match the
* dest mac. So, it has to be deleted, otherwise, we
* will endup receiving vlan packets for which the
* filter is not programmed, when hw-vlan-filter
* configuration is ON
*/
bnxt_del_dflt_mac_filter(bp, vnic);
/* This filter will allow only untagged packets */
bnxt_add_vlan_filter(bp, 0);
net/bnxt: support VLAN filter and strip This patch adds VLAN strip and offload callbacks. To add a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists: if VLAN matches vlan_id VLAN filter already exists, just skip and continue else add a new MAC+VLAN filter else Remove the old MAC only filter Add a new MAC+VLAN filter To remove a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists && VLAN matches vlan_id remove the MAC+VLAN filter add a new MAC only filter else VLAN filter doesn't exist, just skip and continue Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:05 +00:00
}
net/bnxt: fix VLAN filtering Currently, when hw-vlan-filter is enabled on testpmd, driver is receiving all vlan packets. Instead, it should only receive untagged packets and vlan packets for which the VLAN filter is programmed. This is because, the default rule to match on MAC is not getting deleted, when hw-vlan-filter is ON. This patch fixes the problem, by deleting the default MAC rule and programming a new rule to receive only untagged packets, when hw-vlan-filter is enabled & another rule for each vlan, as and when that vlan is configured on that port. Fixes: 246c5cc5f05e ("net/bnxt: use correct flags during VLAN configuration") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 23:25:59 +00:00
PMD_DRV_LOG(DEBUG, "VLAN Filtering: %d\n",
!!(rx_offloads & DEV_RX_OFFLOAD_VLAN_FILTER));
net/bnxt: support VLAN filter and strip This patch adds VLAN strip and offload callbacks. To add a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists: if VLAN matches vlan_id VLAN filter already exists, just skip and continue else add a new MAC+VLAN filter else Remove the old MAC only filter Add a new MAC+VLAN filter To remove a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists && VLAN matches vlan_id remove the MAC+VLAN filter add a new MAC only filter else VLAN filter doesn't exist, just skip and continue Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:05 +00:00
net/bnxt: fix enable/disable VLAN filtering There is no condition check for the user requested operation for VLAN filtering. As a result, VLAN filtering is getting disabled when the user enables/disables VLAN stripping on same port. The function bnxt_hwrm_clear_l2_filter() didn't actually free L2 filter in HW if the reference count of filter is zero. Fixed it by incrementing the reference count of filter in bnxt_alloc_filter() routine. Because of the recent changes in bnxt_hwrm_clear_l2_filter(), change was needed in the routine bnxt_set_default_mac_addr_op() to destroy and re-create the default filter when the user changes the default MAC of the port. Fixes: 5c1171c97216 ("net/bnxt: refactor filter/flow") Fixes: 6118503d8071 ("net/bnxt: fix VLAN filtering") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-01-16 13:04:49 +00:00
return 0;
}
net/bnxt: fix VLAN strip HWRM_VNIC_CFG command to configure VNIC dynamically with traffic running is not working. Driver has to free and recreate the VNIC and then reconfigure the VNIC filters. Fixes: 7fe5668d2ea3 ("net/bnxt: support VLAN filter and strip") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-01-16 13:04:50 +00:00
static int bnxt_free_one_vnic(struct bnxt *bp, uint16_t vnic_id)
{
struct bnxt_vnic_info *vnic = &bp->vnic_info[vnic_id];
unsigned int i;
int rc;
/* Destroy vnic filters and vnic */
if (bp->eth_dev->data->dev_conf.rxmode.offloads &
DEV_RX_OFFLOAD_VLAN_FILTER) {
for (i = 0; i < RTE_ETHER_MAX_VLAN_ID; i++)
bnxt_del_vlan_filter(bp, i);
}
bnxt_del_dflt_mac_filter(bp, vnic);
net/bnxt: fix RSS context cleanup The PMD is allocating an extra RSS context with each port start. But it is freeing only one RSS context during port stop. So at some point we run out of RSS contexts when we do multiple port stop/start sequences. bnxt_hwrm_vnic_ctx_alloc() is called by bnxt_setup_one_vnic(), but bnxt_hwrm_vnic_ctx_free() is not called in the corresponding bnxt_free_one_vnic(). Fix this by calling bnxt_hwrm_vnic_ctx_free() in bnxt_free_one_vnic(). Fixes: 7fe5668d2ea3 ("net/bnxt: support VLAN filter and strip") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-03-16 05:42:40 +00:00
rc = bnxt_hwrm_vnic_ctx_free(bp, vnic);
if (rc)
return rc;
net/bnxt: fix VLAN strip HWRM_VNIC_CFG command to configure VNIC dynamically with traffic running is not working. Driver has to free and recreate the VNIC and then reconfigure the VNIC filters. Fixes: 7fe5668d2ea3 ("net/bnxt: support VLAN filter and strip") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-01-16 13:04:50 +00:00
rc = bnxt_hwrm_vnic_free(bp, vnic);
if (rc)
return rc;
rte_free(vnic->fw_grp_ids);
vnic->fw_grp_ids = NULL;
net/bnxt: fix VNIC Rx queue count on VNIC free bnxt_free_one_vnic and bnxt_setup_one_vnic are called on configuring port vlan stripping. bnxt_setup_one_vnic keeps incrementing the vnic rx_queue_cnt. Fix to reset vnic rx_queue_cnt in bnxt_free_one_vnic. Fixes: cfadfee41ed1 ("net/bnxt: fix VLAN strip") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-04-23 15:02:24 +00:00
vnic->rx_queue_cnt = 0;
net/bnxt: fix VLAN strip HWRM_VNIC_CFG command to configure VNIC dynamically with traffic running is not working. Driver has to free and recreate the VNIC and then reconfigure the VNIC filters. Fixes: 7fe5668d2ea3 ("net/bnxt: support VLAN filter and strip") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-01-16 13:04:50 +00:00
return 0;
}
static int
bnxt_config_vlan_hw_stripping(struct bnxt *bp, uint64_t rx_offloads)
{
struct bnxt_vnic_info *vnic = BNXT_GET_DEFAULT_VNIC(bp);
int rc;
/* Destroy, recreate and reconfigure the default vnic */
rc = bnxt_free_one_vnic(bp, 0);
if (rc)
return rc;
/* default vnic 0 */
rc = bnxt_setup_one_vnic(bp, 0);
if (rc)
return rc;
if (bp->eth_dev->data->dev_conf.rxmode.offloads &
DEV_RX_OFFLOAD_VLAN_FILTER) {
rc = bnxt_add_vlan_filter(bp, 0);
net/bnxt: fix return code handling in VLAN config return value stored in "ret" but it has been overwritten before use. Coverity issue: 353621 Fixes: 7fe5668d2ea3 ("net/bnxt: support VLAN filter and strip") Fixes: df6cd7c1f73a ("net/bnxt: handle reset notify async event from FW") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-01-28 07:29:23 +00:00
if (rc)
return rc;
rc = bnxt_restore_vlan_filters(bp);
if (rc)
return rc;
net/bnxt: fix VLAN strip HWRM_VNIC_CFG command to configure VNIC dynamically with traffic running is not working. Driver has to free and recreate the VNIC and then reconfigure the VNIC filters. Fixes: 7fe5668d2ea3 ("net/bnxt: support VLAN filter and strip") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-01-16 13:04:50 +00:00
} else {
rc = bnxt_add_mac_filter(bp, vnic, NULL, 0, 0);
net/bnxt: fix return code handling in VLAN config return value stored in "ret" but it has been overwritten before use. Coverity issue: 353621 Fixes: 7fe5668d2ea3 ("net/bnxt: support VLAN filter and strip") Fixes: df6cd7c1f73a ("net/bnxt: handle reset notify async event from FW") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-01-28 07:29:23 +00:00
if (rc)
return rc;
net/bnxt: fix VLAN strip HWRM_VNIC_CFG command to configure VNIC dynamically with traffic running is not working. Driver has to free and recreate the VNIC and then reconfigure the VNIC filters. Fixes: 7fe5668d2ea3 ("net/bnxt: support VLAN filter and strip") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-01-16 13:04:50 +00:00
}
rc = bnxt_hwrm_cfa_l2_set_rx_mask(bp, vnic, 0, NULL);
if (rc)
return rc;
PMD_DRV_LOG(DEBUG, "VLAN Strip Offload: %d\n",
!!(rx_offloads & DEV_RX_OFFLOAD_VLAN_STRIP));
return rc;
}
net/bnxt: fix enable/disable VLAN filtering There is no condition check for the user requested operation for VLAN filtering. As a result, VLAN filtering is getting disabled when the user enables/disables VLAN stripping on same port. The function bnxt_hwrm_clear_l2_filter() didn't actually free L2 filter in HW if the reference count of filter is zero. Fixed it by incrementing the reference count of filter in bnxt_alloc_filter() routine. Because of the recent changes in bnxt_hwrm_clear_l2_filter(), change was needed in the routine bnxt_set_default_mac_addr_op() to destroy and re-create the default filter when the user changes the default MAC of the port. Fixes: 5c1171c97216 ("net/bnxt: refactor filter/flow") Fixes: 6118503d8071 ("net/bnxt: fix VLAN filtering") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-01-16 13:04:49 +00:00
static int
bnxt_vlan_offload_set_op(struct rte_eth_dev *dev, int mask)
{
uint64_t rx_offloads = dev->data->dev_conf.rxmode.offloads;
struct bnxt *bp = dev->data->dev_private;
int rc;
rc = is_bnxt_in_error(bp);
if (rc)
return rc;
net/bnxt: handle HW filter setting when port is stopped Driver destroy the VNIC when the port is brought down. Port HW filter setting such as promiscuos, allmulti and VLAN filtering will be applied when port is started. Fixed to return success silently for these callbacks when port is stopped. Also fixed to clear "bp->dev_stopped" before invoking bnxt_vlan_offload_set_op() in bnxt_dev_start_op(). Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
2020-01-16 13:04:51 +00:00
/* Filter settings will get applied when port is started */
net/bnxt: remove a redundant variable Use "dev->data->dev_started" state, instead of local "dev_stopped" to check whether port has been started or not. Fixes: 316e412299fd ("net/bnxt: fix crash when closing") Cc: stable@dpdk.org Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
2020-01-28 07:29:21 +00:00
if (!dev->data->dev_started)
net/bnxt: handle HW filter setting when port is stopped Driver destroy the VNIC when the port is brought down. Port HW filter setting such as promiscuos, allmulti and VLAN filtering will be applied when port is started. Fixed to return success silently for these callbacks when port is stopped. Also fixed to clear "bp->dev_stopped" before invoking bnxt_vlan_offload_set_op() in bnxt_dev_start_op(). Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
2020-01-16 13:04:51 +00:00
return 0;
net/bnxt: fix enable/disable VLAN filtering There is no condition check for the user requested operation for VLAN filtering. As a result, VLAN filtering is getting disabled when the user enables/disables VLAN stripping on same port. The function bnxt_hwrm_clear_l2_filter() didn't actually free L2 filter in HW if the reference count of filter is zero. Fixed it by incrementing the reference count of filter in bnxt_alloc_filter() routine. Because of the recent changes in bnxt_hwrm_clear_l2_filter(), change was needed in the routine bnxt_set_default_mac_addr_op() to destroy and re-create the default filter when the user changes the default MAC of the port. Fixes: 5c1171c97216 ("net/bnxt: refactor filter/flow") Fixes: 6118503d8071 ("net/bnxt: fix VLAN filtering") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-01-16 13:04:49 +00:00
if (mask & ETH_VLAN_FILTER_MASK) {
/* Enable or disable VLAN filtering */
rc = bnxt_config_vlan_hw_filter(bp, rx_offloads);
if (rc)
return rc;
}
net/bnxt: support VLAN filter and strip This patch adds VLAN strip and offload callbacks. To add a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists: if VLAN matches vlan_id VLAN filter already exists, just skip and continue else add a new MAC+VLAN filter else Remove the old MAC only filter Add a new MAC+VLAN filter To remove a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists && VLAN matches vlan_id remove the MAC+VLAN filter add a new MAC only filter else VLAN filter doesn't exist, just skip and continue Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:05 +00:00
if (mask & ETH_VLAN_STRIP_MASK) {
/* Enable or disable VLAN stripping */
net/bnxt: fix VLAN strip HWRM_VNIC_CFG command to configure VNIC dynamically with traffic running is not working. Driver has to free and recreate the VNIC and then reconfigure the VNIC filters. Fixes: 7fe5668d2ea3 ("net/bnxt: support VLAN filter and strip") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-01-16 13:04:50 +00:00
rc = bnxt_config_vlan_hw_stripping(bp, rx_offloads);
if (rc)
return rc;
net/bnxt: support VLAN filter and strip This patch adds VLAN strip and offload callbacks. To add a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists: if VLAN matches vlan_id VLAN filter already exists, just skip and continue else add a new MAC+VLAN filter else Remove the old MAC only filter Add a new MAC+VLAN filter To remove a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists && VLAN matches vlan_id remove the MAC+VLAN filter add a new MAC only filter else VLAN filter doesn't exist, just skip and continue Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:05 +00:00
}
net/bnxt: support for QinQ insertion and stripping Driver will accelerate only outer/S-VLAN insertion by turning on the appropriate bits in the Tx Buffer Descriptor when the packet arrives for transmission. The TPID to be used for this S-VLAN is conveyed by the vlan_tpid_set dev_op which will terminate in the driver. In the Rx path, driver will continue providing the stripped vlan tag in the mbuf's vlan tci field. This would be the outermost vlan tag in a double-tagged packet or the vlan tag for a single vlan tagged pkt. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 17:17:44 +00:00
if (mask & ETH_VLAN_EXTEND_MASK) {
if (rx_offloads & DEV_RX_OFFLOAD_VLAN_EXTEND)
PMD_DRV_LOG(DEBUG, "Extend VLAN supported\n");
else
PMD_DRV_LOG(INFO, "Extend VLAN unsupported\n");
}
return 0;
}
static int
bnxt_vlan_tpid_set_op(struct rte_eth_dev *dev, enum rte_vlan_type vlan_type,
uint16_t tpid)
{
struct bnxt *bp = dev->data->dev_private;
int qinq = dev->data->dev_conf.rxmode.offloads &
DEV_RX_OFFLOAD_VLAN_EXTEND;
if (vlan_type != ETH_VLAN_TYPE_INNER &&
vlan_type != ETH_VLAN_TYPE_OUTER) {
PMD_DRV_LOG(ERR,
"Unsupported vlan type.");
return -EINVAL;
}
if (!qinq) {
PMD_DRV_LOG(ERR,
"QinQ not enabled. Needs to be ON as we can "
"accelerate only outer vlan\n");
return -EINVAL;
}
if (vlan_type == ETH_VLAN_TYPE_OUTER) {
switch (tpid) {
case RTE_ETHER_TYPE_QINQ:
bp->outer_tpid_bd =
TX_BD_LONG_CFA_META_VLAN_TPID_TPID88A8;
break;
case RTE_ETHER_TYPE_VLAN:
bp->outer_tpid_bd =
TX_BD_LONG_CFA_META_VLAN_TPID_TPID8100;
break;
net: use named constants for deprecated QinQ TPIDs Add named constants for deprecated QinQ TPIDs. Update drivers which have already been using existing TPID named constants from librte_net to use the new named constants rather than magic numbers. Signed-off-by: Ivan Malov <ivan.malov@oktetlabs.ru> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2020-06-19 09:21:15 +00:00
case RTE_ETHER_TYPE_QINQ1:
net/bnxt: support for QinQ insertion and stripping Driver will accelerate only outer/S-VLAN insertion by turning on the appropriate bits in the Tx Buffer Descriptor when the packet arrives for transmission. The TPID to be used for this S-VLAN is conveyed by the vlan_tpid_set dev_op which will terminate in the driver. In the Rx path, driver will continue providing the stripped vlan tag in the mbuf's vlan tci field. This would be the outermost vlan tag in a double-tagged packet or the vlan tag for a single vlan tagged pkt. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 17:17:44 +00:00
bp->outer_tpid_bd =
TX_BD_LONG_CFA_META_VLAN_TPID_TPID9100;
break;
net: use named constants for deprecated QinQ TPIDs Add named constants for deprecated QinQ TPIDs. Update drivers which have already been using existing TPID named constants from librte_net to use the new named constants rather than magic numbers. Signed-off-by: Ivan Malov <ivan.malov@oktetlabs.ru> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2020-06-19 09:21:15 +00:00
case RTE_ETHER_TYPE_QINQ2:
net/bnxt: support for QinQ insertion and stripping Driver will accelerate only outer/S-VLAN insertion by turning on the appropriate bits in the Tx Buffer Descriptor when the packet arrives for transmission. The TPID to be used for this S-VLAN is conveyed by the vlan_tpid_set dev_op which will terminate in the driver. In the Rx path, driver will continue providing the stripped vlan tag in the mbuf's vlan tci field. This would be the outermost vlan tag in a double-tagged packet or the vlan tag for a single vlan tagged pkt. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 17:17:44 +00:00
bp->outer_tpid_bd =
TX_BD_LONG_CFA_META_VLAN_TPID_TPID9200;
break;
net: use named constants for deprecated QinQ TPIDs Add named constants for deprecated QinQ TPIDs. Update drivers which have already been using existing TPID named constants from librte_net to use the new named constants rather than magic numbers. Signed-off-by: Ivan Malov <ivan.malov@oktetlabs.ru> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2020-06-19 09:21:15 +00:00
case RTE_ETHER_TYPE_QINQ3:
net/bnxt: support for QinQ insertion and stripping Driver will accelerate only outer/S-VLAN insertion by turning on the appropriate bits in the Tx Buffer Descriptor when the packet arrives for transmission. The TPID to be used for this S-VLAN is conveyed by the vlan_tpid_set dev_op which will terminate in the driver. In the Rx path, driver will continue providing the stripped vlan tag in the mbuf's vlan tci field. This would be the outermost vlan tag in a double-tagged packet or the vlan tag for a single vlan tagged pkt. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 17:17:44 +00:00
bp->outer_tpid_bd =
TX_BD_LONG_CFA_META_VLAN_TPID_TPID9300;
break;
default:
PMD_DRV_LOG(ERR, "Invalid TPID: %x\n", tpid);
return -EINVAL;
}
bp->outer_tpid_bd |= tpid;
PMD_DRV_LOG(INFO, "outer_tpid_bd = %x\n", bp->outer_tpid_bd);
} else if (vlan_type == ETH_VLAN_TYPE_INNER) {
PMD_DRV_LOG(ERR,
"Can accelerate only outer vlan in QinQ\n");
return -EINVAL;
}
ethdev: allow returning error on VLAN offload ops Some devices may not support or fail setting VLAN offload configuration based on dynamic circumstances so the vlan_offload_set_t vector is modified to return an int so the caller can determine success or not. rte_eth_dev_set_vlan_offload is updated to return the value provided by the vector when called along with restoring the original offload configs on failure. Existing vlan_offload_set_t vectors are modified to return an int. Majority of cases return 0 but a few that actually can fail now return their failure codes. Finally, a vlan_offload_set_t vector is added to virtio to facilitate dynamically turning VLAN strip on or off. Signed-off-by: David Harton <dharton@cisco.com> Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
2017-09-01 02:36:28 +00:00
return 0;
net/bnxt: support VLAN filter and strip This patch adds VLAN strip and offload callbacks. To add a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists: if VLAN matches vlan_id VLAN filter already exists, just skip and continue else add a new MAC+VLAN filter else Remove the old MAC only filter Add a new MAC+VLAN filter To remove a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists && VLAN matches vlan_id remove the MAC+VLAN filter add a new MAC only filter else VLAN filter doesn't exist, just skip and continue Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:05 +00:00
}
ethdev: return diagnostic when setting MAC address Change the prototype and the behavior of dev_ops->eth_mac_addr_set(): a return code is added to notify the caller (librte_ether) if an error occurred in the PMD. The new default MAC address is now copied in dev->data->mac_addrs[0] only if the operation is successful. The patch also updates all the PMDs accordingly. Signed-off-by: Olivier Matz <olivier.matz@6wind.com> Signed-off-by: Ivan Malov <ivan.malov@oktetlabs.ru> Acked-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com> Acked-by: Shreyansh Jain <shreyansh.jain@nxp.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com> Acked-by: Thomas Monjalon <thomas@monjalon.net> Acked-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>
2018-04-11 16:32:51 +00:00
static int
net: add rte prefix to ether structures Add 'rte_' prefix to structures: - rename struct ether_addr as struct rte_ether_addr. - rename struct ether_hdr as struct rte_ether_hdr. - rename struct vlan_hdr as struct rte_vlan_hdr. - rename struct vxlan_hdr as struct rte_vxlan_hdr. - rename struct vxlan_gpe_hdr as struct rte_vxlan_gpe_hdr. Do not update the command line library to avoid adding a dependency to librte_net. Signed-off-by: Olivier Matz <olivier.matz@6wind.com> Reviewed-by: Stephen Hemminger <stephen@networkplumber.org> Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2019-05-21 16:13:03 +00:00
bnxt_set_default_mac_addr_op(struct rte_eth_dev *dev,
net/bnxt: support for QinQ insertion and stripping Driver will accelerate only outer/S-VLAN insertion by turning on the appropriate bits in the Tx Buffer Descriptor when the packet arrives for transmission. The TPID to be used for this S-VLAN is conveyed by the vlan_tpid_set dev_op which will terminate in the driver. In the Rx path, driver will continue providing the stripped vlan tag in the mbuf's vlan tci field. This would be the outermost vlan tag in a double-tagged packet or the vlan tag for a single vlan tagged pkt. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 17:17:44 +00:00
struct rte_ether_addr *addr)
net/bnxt: support multicast filter and set MAC addr This patch adds support for set_mc_addr_list and mac_addr_set dev_ops Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:06 +00:00
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = dev->data->dev_private;
net/bnxt: support multicast filter and set MAC addr This patch adds support for set_mc_addr_list and mac_addr_set dev_ops Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:06 +00:00
/* Default Filter is tied to VNIC 0 */
net/bnxt: use macro for getting default VNIC BNXT_GET_DEFAULT_VNIC macro returns the default VNIC of the function and fixed all occurrences of code to use the macro. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-12-02 04:18:09 +00:00
struct bnxt_vnic_info *vnic = BNXT_GET_DEFAULT_VNIC(bp);
net/bnxt: support multicast filter and set MAC addr This patch adds support for set_mc_addr_list and mac_addr_set dev_ops Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:06 +00:00
int rc;
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
rc = is_bnxt_in_error(bp);
if (rc)
return rc;
net/bnxt: support trusted VF In the current model, VFs are not trusted. So it is not allowed to send many HWRM commands. Newer firmware has added support to allow VF to be trusted. Now the VF queries if it is a trusted entity and based on that it can send HWRM commands to the firmware. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-09-29 02:00:00 +00:00
if (BNXT_VF(bp) && !BNXT_VF_IS_TRUSTED(bp))
ethdev: return diagnostic when setting MAC address Change the prototype and the behavior of dev_ops->eth_mac_addr_set(): a return code is added to notify the caller (librte_ether) if an error occurred in the PMD. The new default MAC address is now copied in dev->data->mac_addrs[0] only if the operation is successful. The patch also updates all the PMDs accordingly. Signed-off-by: Olivier Matz <olivier.matz@6wind.com> Signed-off-by: Ivan Malov <ivan.malov@oktetlabs.ru> Acked-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com> Acked-by: Shreyansh Jain <shreyansh.jain@nxp.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com> Acked-by: Thomas Monjalon <thomas@monjalon.net> Acked-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>
2018-04-11 16:32:51 +00:00
return -EPERM;
net/bnxt: support multicast filter and set MAC addr This patch adds support for set_mc_addr_list and mac_addr_set dev_ops Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:06 +00:00
net/bnxt: fix setting primary MAC address 1. Default filter is tied to VNIC 0 at index 0. After finding the filter with mac_index 0 and set the new MAC address, looping through remaining filters is unnecessary. 2. Added a check for NULL MAC address. 3. bnxt_hwrm_set_l2_filter() clears the existing filter configuration first before applying new filter settings. Hence there is no need to invoke bnxt_hwrm_clear_l2_filter() explicitly in bnxt_set_default_mac_addr_op(). Fixes: d69851df12b2 ("net/bnxt: support multicast filter and set MAC addr") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-18 03:35:59 +00:00
if (rte_is_zero_ether_addr(addr))
return -EINVAL;
net/bnxt: support multicast filter and set MAC addr This patch adds support for set_mc_addr_list and mac_addr_set dev_ops Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:06 +00:00
net/bnxt: fix MAC address setting when port is stopped Driver destroys the VNIC when the port is brought down. Port HW filter setting such as default MAC address and unicast MAC filters will be applied when port is started. Fixed to return success silently for these callbacks when port is stopped. Fixes: 39b88344e364 ("net/bnxt: fix enable/disable VLAN filtering") Cc: stable@dpdk.org Reported-by: Stephen Hemminger <stephen@networkplumber.org> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-02-27 08:18:51 +00:00
/* Filter settings will get applied when port is started */
if (!dev->data->dev_started)
return 0;
net/bnxt: fix enable/disable VLAN filtering There is no condition check for the user requested operation for VLAN filtering. As a result, VLAN filtering is getting disabled when the user enables/disables VLAN stripping on same port. The function bnxt_hwrm_clear_l2_filter() didn't actually free L2 filter in HW if the reference count of filter is zero. Fixed it by incrementing the reference count of filter in bnxt_alloc_filter() routine. Because of the recent changes in bnxt_hwrm_clear_l2_filter(), change was needed in the routine bnxt_set_default_mac_addr_op() to destroy and re-create the default filter when the user changes the default MAC of the port. Fixes: 5c1171c97216 ("net/bnxt: refactor filter/flow") Fixes: 6118503d8071 ("net/bnxt: fix VLAN filtering") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-01-16 13:04:49 +00:00
/* Check if the requested MAC is already added */
if (memcmp(addr, bp->mac_addr, RTE_ETHER_ADDR_LEN) == 0)
return 0;
net/bnxt: fix setting primary MAC address 1. Default filter is tied to VNIC 0 at index 0. After finding the filter with mac_index 0 and set the new MAC address, looping through remaining filters is unnecessary. 2. Added a check for NULL MAC address. 3. bnxt_hwrm_set_l2_filter() clears the existing filter configuration first before applying new filter settings. Hence there is no need to invoke bnxt_hwrm_clear_l2_filter() explicitly in bnxt_set_default_mac_addr_op(). Fixes: d69851df12b2 ("net/bnxt: support multicast filter and set MAC addr") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-18 03:35:59 +00:00
net/bnxt: fix enable/disable VLAN filtering There is no condition check for the user requested operation for VLAN filtering. As a result, VLAN filtering is getting disabled when the user enables/disables VLAN stripping on same port. The function bnxt_hwrm_clear_l2_filter() didn't actually free L2 filter in HW if the reference count of filter is zero. Fixed it by incrementing the reference count of filter in bnxt_alloc_filter() routine. Because of the recent changes in bnxt_hwrm_clear_l2_filter(), change was needed in the routine bnxt_set_default_mac_addr_op() to destroy and re-create the default filter when the user changes the default MAC of the port. Fixes: 5c1171c97216 ("net/bnxt: refactor filter/flow") Fixes: 6118503d8071 ("net/bnxt: fix VLAN filtering") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-01-16 13:04:49 +00:00
/* Destroy filter and re-create it */
bnxt_del_dflt_mac_filter(bp, vnic);
net/bnxt: fix setting primary MAC address 1. Default filter is tied to VNIC 0 at index 0. After finding the filter with mac_index 0 and set the new MAC address, looping through remaining filters is unnecessary. 2. Added a check for NULL MAC address. 3. bnxt_hwrm_set_l2_filter() clears the existing filter configuration first before applying new filter settings. Hence there is no need to invoke bnxt_hwrm_clear_l2_filter() explicitly in bnxt_set_default_mac_addr_op(). Fixes: d69851df12b2 ("net/bnxt: support multicast filter and set MAC addr") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-18 03:35:59 +00:00
net/bnxt: fix enable/disable VLAN filtering There is no condition check for the user requested operation for VLAN filtering. As a result, VLAN filtering is getting disabled when the user enables/disables VLAN stripping on same port. The function bnxt_hwrm_clear_l2_filter() didn't actually free L2 filter in HW if the reference count of filter is zero. Fixed it by incrementing the reference count of filter in bnxt_alloc_filter() routine. Because of the recent changes in bnxt_hwrm_clear_l2_filter(), change was needed in the routine bnxt_set_default_mac_addr_op() to destroy and re-create the default filter when the user changes the default MAC of the port. Fixes: 5c1171c97216 ("net/bnxt: refactor filter/flow") Fixes: 6118503d8071 ("net/bnxt: fix VLAN filtering") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-01-16 13:04:49 +00:00
memcpy(bp->mac_addr, addr, RTE_ETHER_ADDR_LEN);
if (dev->data->dev_conf.rxmode.offloads & DEV_RX_OFFLOAD_VLAN_FILTER) {
/* This filter will allow only untagged packets */
rc = bnxt_add_vlan_filter(bp, 0);
} else {
rc = bnxt_add_mac_filter(bp, vnic, addr, 0, 0);
net/bnxt: support multicast filter and set MAC addr This patch adds support for set_mc_addr_list and mac_addr_set dev_ops Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:06 +00:00
}
ethdev: return diagnostic when setting MAC address Change the prototype and the behavior of dev_ops->eth_mac_addr_set(): a return code is added to notify the caller (librte_ether) if an error occurred in the PMD. The new default MAC address is now copied in dev->data->mac_addrs[0] only if the operation is successful. The patch also updates all the PMDs accordingly. Signed-off-by: Olivier Matz <olivier.matz@6wind.com> Signed-off-by: Ivan Malov <ivan.malov@oktetlabs.ru> Acked-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com> Acked-by: Shreyansh Jain <shreyansh.jain@nxp.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com> Acked-by: Thomas Monjalon <thomas@monjalon.net> Acked-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>
2018-04-11 16:32:51 +00:00
net/bnxt: fix enable/disable VLAN filtering There is no condition check for the user requested operation for VLAN filtering. As a result, VLAN filtering is getting disabled when the user enables/disables VLAN stripping on same port. The function bnxt_hwrm_clear_l2_filter() didn't actually free L2 filter in HW if the reference count of filter is zero. Fixed it by incrementing the reference count of filter in bnxt_alloc_filter() routine. Because of the recent changes in bnxt_hwrm_clear_l2_filter(), change was needed in the routine bnxt_set_default_mac_addr_op() to destroy and re-create the default filter when the user changes the default MAC of the port. Fixes: 5c1171c97216 ("net/bnxt: refactor filter/flow") Fixes: 6118503d8071 ("net/bnxt: fix VLAN filtering") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-01-16 13:04:49 +00:00
PMD_DRV_LOG(DEBUG, "Set MAC addr\n");
return rc;
net/bnxt: support multicast filter and set MAC addr This patch adds support for set_mc_addr_list and mac_addr_set dev_ops Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:06 +00:00
}
static int
bnxt_dev_set_mc_addr_list_op(struct rte_eth_dev *eth_dev,
net: add rte prefix to ether structures Add 'rte_' prefix to structures: - rename struct ether_addr as struct rte_ether_addr. - rename struct ether_hdr as struct rte_ether_hdr. - rename struct vlan_hdr as struct rte_vlan_hdr. - rename struct vxlan_hdr as struct rte_vxlan_hdr. - rename struct vxlan_gpe_hdr as struct rte_vxlan_gpe_hdr. Do not update the command line library to avoid adding a dependency to librte_net. Signed-off-by: Olivier Matz <olivier.matz@6wind.com> Reviewed-by: Stephen Hemminger <stephen@networkplumber.org> Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2019-05-21 16:13:03 +00:00
struct rte_ether_addr *mc_addr_set,
net/bnxt: support multicast filter and set MAC addr This patch adds support for set_mc_addr_list and mac_addr_set dev_ops Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:06 +00:00
uint32_t nb_mc_addr)
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = eth_dev->data->dev_private;
net/bnxt: support multicast filter and set MAC addr This patch adds support for set_mc_addr_list and mac_addr_set dev_ops Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:06 +00:00
char *mc_addr_list = (char *)mc_addr_set;
struct bnxt_vnic_info *vnic;
uint32_t off = 0, i = 0;
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
int rc;
rc = is_bnxt_in_error(bp);
if (rc)
return rc;
net/bnxt: support multicast filter and set MAC addr This patch adds support for set_mc_addr_list and mac_addr_set dev_ops Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:06 +00:00
net/bnxt: use macro for getting default VNIC Minor code cleanup. BNXT_GET_DEFAULT_VNIC macro returns the default VNIC of the function and fixed code to use the macro. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 17:17:45 +00:00
vnic = BNXT_GET_DEFAULT_VNIC(bp);
net/bnxt: support multicast filter and set MAC addr This patch adds support for set_mc_addr_list and mac_addr_set dev_ops Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:06 +00:00
if (nb_mc_addr > BNXT_MAX_MC_ADDRS) {
vnic->flags |= BNXT_VNIC_INFO_ALLMULTI;
goto allmulti;
}
/* TODO Check for Duplicate mcast addresses */
vnic->flags &= ~BNXT_VNIC_INFO_ALLMULTI;
for (i = 0; i < nb_mc_addr; i++) {
net: add rte prefix to ether defines Add 'RTE_' prefix to defines: - rename ETHER_ADDR_LEN as RTE_ETHER_ADDR_LEN. - rename ETHER_TYPE_LEN as RTE_ETHER_TYPE_LEN. - rename ETHER_CRC_LEN as RTE_ETHER_CRC_LEN. - rename ETHER_HDR_LEN as RTE_ETHER_HDR_LEN. - rename ETHER_MIN_LEN as RTE_ETHER_MIN_LEN. - rename ETHER_MAX_LEN as RTE_ETHER_MAX_LEN. - rename ETHER_MTU as RTE_ETHER_MTU. - rename ETHER_MAX_VLAN_FRAME_LEN as RTE_ETHER_MAX_VLAN_FRAME_LEN. - rename ETHER_MAX_VLAN_ID as RTE_ETHER_MAX_VLAN_ID. - rename ETHER_MAX_JUMBO_FRAME_LEN as RTE_ETHER_MAX_JUMBO_FRAME_LEN. - rename ETHER_MIN_MTU as RTE_ETHER_MIN_MTU. - rename ETHER_LOCAL_ADMIN_ADDR as RTE_ETHER_LOCAL_ADMIN_ADDR. - rename ETHER_GROUP_ADDR as RTE_ETHER_GROUP_ADDR. - rename ETHER_TYPE_IPv4 as RTE_ETHER_TYPE_IPv4. - rename ETHER_TYPE_IPv6 as RTE_ETHER_TYPE_IPv6. - rename ETHER_TYPE_ARP as RTE_ETHER_TYPE_ARP. - rename ETHER_TYPE_VLAN as RTE_ETHER_TYPE_VLAN. - rename ETHER_TYPE_RARP as RTE_ETHER_TYPE_RARP. - rename ETHER_TYPE_QINQ as RTE_ETHER_TYPE_QINQ. - rename ETHER_TYPE_ETAG as RTE_ETHER_TYPE_ETAG. - rename ETHER_TYPE_1588 as RTE_ETHER_TYPE_1588. - rename ETHER_TYPE_SLOW as RTE_ETHER_TYPE_SLOW. - rename ETHER_TYPE_TEB as RTE_ETHER_TYPE_TEB. - rename ETHER_TYPE_LLDP as RTE_ETHER_TYPE_LLDP. - rename ETHER_TYPE_MPLS as RTE_ETHER_TYPE_MPLS. - rename ETHER_TYPE_MPLSM as RTE_ETHER_TYPE_MPLSM. - rename ETHER_VXLAN_HLEN as RTE_ETHER_VXLAN_HLEN. - rename ETHER_ADDR_FMT_SIZE as RTE_ETHER_ADDR_FMT_SIZE. - rename VXLAN_GPE_TYPE_IPV4 as RTE_VXLAN_GPE_TYPE_IPV4. - rename VXLAN_GPE_TYPE_IPV6 as RTE_VXLAN_GPE_TYPE_IPV6. - rename VXLAN_GPE_TYPE_ETH as RTE_VXLAN_GPE_TYPE_ETH. - rename VXLAN_GPE_TYPE_NSH as RTE_VXLAN_GPE_TYPE_NSH. - rename VXLAN_GPE_TYPE_MPLS as RTE_VXLAN_GPE_TYPE_MPLS. - rename VXLAN_GPE_TYPE_GBP as RTE_VXLAN_GPE_TYPE_GBP. - rename VXLAN_GPE_TYPE_VBNG as RTE_VXLAN_GPE_TYPE_VBNG. - rename ETHER_VXLAN_GPE_HLEN as RTE_ETHER_VXLAN_GPE_HLEN. Do not update the command line library to avoid adding a dependency to librte_net. Signed-off-by: Olivier Matz <olivier.matz@6wind.com> Reviewed-by: Stephen Hemminger <stephen@networkplumber.org> Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2019-05-21 16:13:05 +00:00
memcpy(vnic->mc_list + off, &mc_addr_list[i],
RTE_ETHER_ADDR_LEN);
off += RTE_ETHER_ADDR_LEN;
net/bnxt: support multicast filter and set MAC addr This patch adds support for set_mc_addr_list and mac_addr_set dev_ops Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:06 +00:00
}
vnic->mc_addr_cnt = i;
net/bnxt: fix multicast filter programming Fixed multicast filter programming and allmulti programming. Fixed to skip programming multicast macs if the user requests allmulti mode. Also removed a comment in bnxt_hwrm_cfa_l2_set_rx_mask() which is no longer valid now. Fixes: d69851df12b2 ("net/bnxt: support multicast filter and set MAC addr") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 23:26:00 +00:00
if (vnic->mc_addr_cnt)
vnic->flags |= BNXT_VNIC_INFO_MCAST;
else
vnic->flags &= ~BNXT_VNIC_INFO_MCAST;
net/bnxt: support multicast filter and set MAC addr This patch adds support for set_mc_addr_list and mac_addr_set dev_ops Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:06 +00:00
allmulti:
net/bnxt: support set VF QOS and MAC anti spoof This patch adds support to 1) enable VF MAC anti spoof. 2) QOS configuration for specified VF. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:17 +00:00
return bnxt_hwrm_cfa_l2_set_rx_mask(bp, vnic, 0, NULL);
net/bnxt: support multicast filter and set MAC addr This patch adds support for set_mc_addr_list and mac_addr_set dev_ops Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:06 +00:00
}
net/bnxt: support get FW version This patch adds support for fw_version_get dev_op Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:08 +00:00
static int
bnxt_fw_version_get(struct rte_eth_dev *dev, char *fw_version, size_t fw_size)
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = dev->data->dev_private;
net/bnxt: support get FW version This patch adds support for fw_version_get dev_op Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:08 +00:00
uint8_t fw_major = (bp->fw_ver >> 24) & 0xff;
uint8_t fw_minor = (bp->fw_ver >> 16) & 0xff;
uint8_t fw_updt = (bp->fw_ver >> 8) & 0xff;
net/bnxt: fix FW version query bnxt_fw_version_get is not indicating the complete FW version. Fix it to indicate complete version string. Fixes: e2652b0a20a0 ("net/bnxt: support get FW version") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-05-07 05:43:28 +00:00
uint8_t fw_rsvd = bp->fw_ver & 0xff;
net/bnxt: support get FW version This patch adds support for fw_version_get dev_op Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:08 +00:00
int ret;
net/bnxt: fix FW version query bnxt_fw_version_get is not indicating the complete FW version. Fix it to indicate complete version string. Fixes: e2652b0a20a0 ("net/bnxt: support get FW version") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-05-07 05:43:28 +00:00
ret = snprintf(fw_version, fw_size, "%d.%d.%d.%d",
fw_major, fw_minor, fw_updt, fw_rsvd);
drivers/net: fix FW version query Fixes a few different things: * Remove 'fw_version' NULL checks, it is allowed if the 'fw_size' is zero, 'fw_version' being NULL but 'fw_size' not zero condition checked in ethdev layer * Be sure required buffer size is returned if provided one is not big enough, instead of returning success (0) * Document in doxygen comment the '-EINVAL' is a valid return type * Take into account that 'snprintf' can return negative value * Cast length to 'size_t' to compare it with 'fw_size' Fixes: bb42aa9ffe4e ("net/atlantic: configure device start/stop") Fixes: ff70acdf4299 ("net/axgbe: support reading FW version") Fixes: e2652b0a20a0 ("net/bnxt: support get FW version") Fixes: cf0fab1d2ca5 ("net/dpaa: support firmware version get API") Fixes: 748eccb97cdc ("net/dpaa2: add support for firmware version get") Fixes: b883c0644a24 ("net/e1000: add firmware version get") Fixes: 293430677e9c ("net/enic: add handler to return firmware version") Fixes: 1f5ca0b460cd ("net/hns3: support some device operations") Fixes: bd5b86732bc7 ("net/hns3: modify format for firmware version") Fixes: ed0dfdd0e976 ("net/i40e: add firmware version get") Fixes: e31cb9a36298 ("net/ice: support FW version getting") Fixes: 4f09bc55ac3d ("net/igc: implement device base operations") Fixes: eec10fb0ce6b ("net/ionic: support FW version") Fixes: 8b0b56574269 ("net/ixgbe: add firmware version get") Fixes: 4d9f5b8adc02 ("net/octeontx2: add FW version get operation") Fixes: f97b56f9f12e ("net/qede: support FW version query") Fixes: 83fef46a22b2 ("net/sfc: add callback to retrieve FW version") Fixes: bc84ac0fadef ("net/txgbe: support getting FW version") Fixes: 21913471202f ("ethdev: add firmware version get") Cc: stable@dpdk.org Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Acked-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru> Acked-by: Haiyue Wang <haiyue.wang@intel.com> Acked-by: Rasesh Mody <rmody@marvell.com> Acked-by: Jiawen Wu <jiawenwu@trustnetic.com> Acked-by: Beilei Xing <beilei.xing@intel.com> Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>
2021-04-21 16:20:57 +00:00
if (ret < 0)
return -EINVAL;
net/bnxt: support get FW version This patch adds support for fw_version_get dev_op Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:08 +00:00
ret += 1; /* add the size of '\0' */
drivers/net: fix FW version query Fixes a few different things: * Remove 'fw_version' NULL checks, it is allowed if the 'fw_size' is zero, 'fw_version' being NULL but 'fw_size' not zero condition checked in ethdev layer * Be sure required buffer size is returned if provided one is not big enough, instead of returning success (0) * Document in doxygen comment the '-EINVAL' is a valid return type * Take into account that 'snprintf' can return negative value * Cast length to 'size_t' to compare it with 'fw_size' Fixes: bb42aa9ffe4e ("net/atlantic: configure device start/stop") Fixes: ff70acdf4299 ("net/axgbe: support reading FW version") Fixes: e2652b0a20a0 ("net/bnxt: support get FW version") Fixes: cf0fab1d2ca5 ("net/dpaa: support firmware version get API") Fixes: 748eccb97cdc ("net/dpaa2: add support for firmware version get") Fixes: b883c0644a24 ("net/e1000: add firmware version get") Fixes: 293430677e9c ("net/enic: add handler to return firmware version") Fixes: 1f5ca0b460cd ("net/hns3: support some device operations") Fixes: bd5b86732bc7 ("net/hns3: modify format for firmware version") Fixes: ed0dfdd0e976 ("net/i40e: add firmware version get") Fixes: e31cb9a36298 ("net/ice: support FW version getting") Fixes: 4f09bc55ac3d ("net/igc: implement device base operations") Fixes: eec10fb0ce6b ("net/ionic: support FW version") Fixes: 8b0b56574269 ("net/ixgbe: add firmware version get") Fixes: 4d9f5b8adc02 ("net/octeontx2: add FW version get operation") Fixes: f97b56f9f12e ("net/qede: support FW version query") Fixes: 83fef46a22b2 ("net/sfc: add callback to retrieve FW version") Fixes: bc84ac0fadef ("net/txgbe: support getting FW version") Fixes: 21913471202f ("ethdev: add firmware version get") Cc: stable@dpdk.org Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Acked-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru> Acked-by: Haiyue Wang <haiyue.wang@intel.com> Acked-by: Rasesh Mody <rmody@marvell.com> Acked-by: Jiawen Wu <jiawenwu@trustnetic.com> Acked-by: Beilei Xing <beilei.xing@intel.com> Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>
2021-04-21 16:20:57 +00:00
if (fw_size < (size_t)ret)
net/bnxt: support get FW version This patch adds support for fw_version_get dev_op Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:08 +00:00
return ret;
else
return 0;
}
net/bnxt: support rxq/txq get information Add support for txq_info_get and rxq_info_get dev_ops Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:11 +00:00
static void
bnxt_rxq_info_get_op(struct rte_eth_dev *dev, uint16_t queue_id,
struct rte_eth_rxq_info *qinfo)
{
net/bnxt: add checks for firmware reset Driver should fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Added missing checks for fw reset in few routines. Fixes: be14720def9c ("net/bnxt: support FW reset") Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-11-13 08:29:40 +00:00
struct bnxt *bp = dev->data->dev_private;
net/bnxt: support rxq/txq get information Add support for txq_info_get and rxq_info_get dev_ops Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:11 +00:00
struct bnxt_rx_queue *rxq;
net/bnxt: add checks for firmware reset Driver should fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Added missing checks for fw reset in few routines. Fixes: be14720def9c ("net/bnxt: support FW reset") Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-11-13 08:29:40 +00:00
if (is_bnxt_in_error(bp))
return;
net/bnxt: support rxq/txq get information Add support for txq_info_get and rxq_info_get dev_ops Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:11 +00:00
rxq = dev->data->rx_queues[queue_id];
qinfo->mp = rxq->mb_pool;
qinfo->scattered_rx = dev->data->scattered_rx;
qinfo->nb_desc = rxq->nb_rx_desc;
qinfo->conf.rx_free_thresh = rxq->rx_free_thresh;
net/bnxt: fix drop enable in get Rx queue info Return correct value for rx_drop_en. Add per-queue field to track rx_drop_en configuration. Fixes: 2fc201884be8 ("net/bnxt: support rxq/txq get information") Cc: stable@dpdk.org Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 17:30:34 +00:00
qinfo->conf.rx_drop_en = rxq->drop_en;
net/bnxt: fix Rx queue start/stop Driver should not change "deferred_start" state of the rx queues. It should get the state in queue_setup_op() and use that value. Since the deferred start state was being used in the packet receive functions to determine whether a stopped rx ring should be polled, introduced a per-rxq flag to track queue stopped/started state. Fixes: 9b63c6fd70e3 ("net/bnxt: support Rx/Tx queue start/stop") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 17:17:32 +00:00
qinfo->conf.rx_deferred_start = rxq->rx_deferred_start;
net/bnxt: fix queue get info Return current offloads in rxq_info_get()/txq_info_get(). Fixes: 2fc201884be8 ("net/bnxt: support rxq/txq get information") Cc: stable@dpdk.org Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 17:30:35 +00:00
qinfo->conf.offloads = dev->data->dev_conf.rxmode.offloads;
net/bnxt: support rxq/txq get information Add support for txq_info_get and rxq_info_get dev_ops Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:11 +00:00
}
static void
bnxt_txq_info_get_op(struct rte_eth_dev *dev, uint16_t queue_id,
struct rte_eth_txq_info *qinfo)
{
net/bnxt: add checks for firmware reset Driver should fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Added missing checks for fw reset in few routines. Fixes: be14720def9c ("net/bnxt: support FW reset") Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-11-13 08:29:40 +00:00
struct bnxt *bp = dev->data->dev_private;
net/bnxt: support rxq/txq get information Add support for txq_info_get and rxq_info_get dev_ops Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:11 +00:00
struct bnxt_tx_queue *txq;
net/bnxt: add checks for firmware reset Driver should fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Added missing checks for fw reset in few routines. Fixes: be14720def9c ("net/bnxt: support FW reset") Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-11-13 08:29:40 +00:00
if (is_bnxt_in_error(bp))
return;
net/bnxt: support rxq/txq get information Add support for txq_info_get and rxq_info_get dev_ops Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:11 +00:00
txq = dev->data->tx_queues[queue_id];
qinfo->nb_desc = txq->nb_tx_desc;
qinfo->conf.tx_thresh.pthresh = txq->pthresh;
qinfo->conf.tx_thresh.hthresh = txq->hthresh;
qinfo->conf.tx_thresh.wthresh = txq->wthresh;
qinfo->conf.tx_free_thresh = txq->tx_free_thresh;
qinfo->conf.tx_rs_thresh = 0;
qinfo->conf.tx_deferred_start = txq->tx_deferred_start;
net/bnxt: support fast mbuf free Add support for DEV_TX_OFFLOAD_MBUF_FAST_FREE to bnxt vector mode transmit. This offload may be enabled only when multi-segment transmit is not needed, all transmitted mbufs for a given queue will be allocated from the same pool, and all transmitted mbufs will have a reference count of 1. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-10-06 17:37:52 +00:00
qinfo->conf.offloads = txq->offloads;
net/bnxt: support rxq/txq get information Add support for txq_info_get and rxq_info_get dev_ops Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:11 +00:00
}
net/bnxt: fix getting burst mode for Arm Transmit and receive burst mode get operations incorrectly return "Vector SSE" on ARM64 platforms, change to return "Vector Neon" instead. Fixes: 398358341419 ("net/bnxt: support NEON") Cc: stable@dpdk.org Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
2020-09-09 15:52:54 +00:00
static const struct {
eth_rx_burst_t pkt_burst;
const char *info;
} bnxt_rx_burst_info[] = {
{bnxt_recv_pkts, "Scalar"},
#if defined(RTE_ARCH_X86)
{bnxt_recv_pkts_vec, "Vector SSE"},
#elif defined(RTE_ARCH_ARM64)
{bnxt_recv_pkts_vec, "Vector Neon"},
#endif
};
net/bnxt: support Rx/Tx burst mode info Retrieve burst mode options according to the selected Rx/Tx burst function name. Update 20.08 release notes with this information. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-07 22:22:24 +00:00
static int
bnxt_rx_burst_mode_get(struct rte_eth_dev *dev, __rte_unused uint16_t queue_id,
struct rte_eth_burst_mode *mode)
{
eth_rx_burst_t pkt_burst = dev->rx_pkt_burst;
net/bnxt: fix getting burst mode for Arm Transmit and receive burst mode get operations incorrectly return "Vector SSE" on ARM64 platforms, change to return "Vector Neon" instead. Fixes: 398358341419 ("net/bnxt: support NEON") Cc: stable@dpdk.org Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
2020-09-09 15:52:54 +00:00
size_t i;
net/bnxt: support Rx/Tx burst mode info Retrieve burst mode options according to the selected Rx/Tx burst function name. Update 20.08 release notes with this information. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-07 22:22:24 +00:00
net/bnxt: fix getting burst mode for Arm Transmit and receive burst mode get operations incorrectly return "Vector SSE" on ARM64 platforms, change to return "Vector Neon" instead. Fixes: 398358341419 ("net/bnxt: support NEON") Cc: stable@dpdk.org Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
2020-09-09 15:52:54 +00:00
for (i = 0; i < RTE_DIM(bnxt_rx_burst_info); i++) {
if (pkt_burst == bnxt_rx_burst_info[i].pkt_burst) {
snprintf(mode->info, sizeof(mode->info), "%s",
bnxt_rx_burst_info[i].info);
return 0;
}
net/bnxt: support Rx/Tx burst mode info Retrieve burst mode options according to the selected Rx/Tx burst function name. Update 20.08 release notes with this information. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-07 22:22:24 +00:00
}
return -EINVAL;
}
net/bnxt: fix getting burst mode for Arm Transmit and receive burst mode get operations incorrectly return "Vector SSE" on ARM64 platforms, change to return "Vector Neon" instead. Fixes: 398358341419 ("net/bnxt: support NEON") Cc: stable@dpdk.org Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
2020-09-09 15:52:54 +00:00
static const struct {
eth_tx_burst_t pkt_burst;
const char *info;
} bnxt_tx_burst_info[] = {
{bnxt_xmit_pkts, "Scalar"},
#if defined(RTE_ARCH_X86)
{bnxt_xmit_pkts_vec, "Vector SSE"},
#elif defined(RTE_ARCH_ARM64)
{bnxt_xmit_pkts_vec, "Vector Neon"},
#endif
};
net/bnxt: support Rx/Tx burst mode info Retrieve burst mode options according to the selected Rx/Tx burst function name. Update 20.08 release notes with this information. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-07 22:22:24 +00:00
static int
bnxt_tx_burst_mode_get(struct rte_eth_dev *dev, __rte_unused uint16_t queue_id,
struct rte_eth_burst_mode *mode)
{
eth_tx_burst_t pkt_burst = dev->tx_pkt_burst;
net/bnxt: fix getting burst mode for Arm Transmit and receive burst mode get operations incorrectly return "Vector SSE" on ARM64 platforms, change to return "Vector Neon" instead. Fixes: 398358341419 ("net/bnxt: support NEON") Cc: stable@dpdk.org Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
2020-09-09 15:52:54 +00:00
size_t i;
net/bnxt: support Rx/Tx burst mode info Retrieve burst mode options according to the selected Rx/Tx burst function name. Update 20.08 release notes with this information. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-07 22:22:24 +00:00
net/bnxt: fix getting burst mode for Arm Transmit and receive burst mode get operations incorrectly return "Vector SSE" on ARM64 platforms, change to return "Vector Neon" instead. Fixes: 398358341419 ("net/bnxt: support NEON") Cc: stable@dpdk.org Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
2020-09-09 15:52:54 +00:00
for (i = 0; i < RTE_DIM(bnxt_tx_burst_info); i++) {
if (pkt_burst == bnxt_tx_burst_info[i].pkt_burst) {
snprintf(mode->info, sizeof(mode->info), "%s",
bnxt_tx_burst_info[i].info);
return 0;
}
net/bnxt: support Rx/Tx burst mode info Retrieve burst mode options according to the selected Rx/Tx burst function name. Update 20.08 release notes with this information. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-07 22:22:24 +00:00
}
return -EINVAL;
}
net/bnxt: fix crash after removing and adding slaves On removing the slave interface, slave_remove in bonding module calls _rte_eth_dev_reset which in turn frees both Tx and Rx queues. 1. segfault is seen after removing/adding the slave interface and starting bond interface. In this below path, when mtu is set for the slave interface, queues are not created yet and driver reference to queue[0] causes the segfault. slave_configure: rte_eth_dev_set_mtu rte_eth_dev_configure rte_eth_rx_queue_setup 2. segfault is seen on starting the port after removing from bond device. This is a testpmd bug where in, on starting the port, testpmd is supposed to recreate the queues before starting the port. Fixed these by adding check for queues created before accessing them. Fixes: daef48efe5e5 ("net/bnxt: support set MTU") Cc: stable@dpdk.org Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-24 07:44:27 +00:00
int bnxt_mtu_set_op(struct rte_eth_dev *eth_dev, uint16_t new_mtu)
net/bnxt: support set MTU This patch adds support to modify MTU using the set_mtu dev_op. To support frames > 2k, the PMD creates an aggregator ring. When a frame greater than 2k is received, it is fragmented and the resulting fragments are DMA'ed to the aggregator ring. Now the driver can support jumbo frames upto 9500 bytes. Signed-off-by: Steeven Li <steeven.li@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:09 +00:00
{
struct bnxt *bp = eth_dev->data->dev_private;
net/bnxt: implement SSE vector mode Introduce SSE vector mode support for the bnxt pmd. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-05-29 21:02:23 +00:00
uint32_t new_pkt_size;
net/bnxt: support set MTU This patch adds support to modify MTU using the set_mtu dev_op. To support frames > 2k, the PMD creates an aggregator ring. When a frame greater than 2k is received, it is fragmented and the resulting fragments are DMA'ed to the aggregator ring. Now the driver can support jumbo frames upto 9500 bytes. Signed-off-by: Steeven Li <steeven.li@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:09 +00:00
uint32_t rc = 0;
uint32_t i;
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
rc = is_bnxt_in_error(bp);
if (rc)
return rc;
net/bnxt: fix crash after removing and adding slaves On removing the slave interface, slave_remove in bonding module calls _rte_eth_dev_reset which in turn frees both Tx and Rx queues. 1. segfault is seen after removing/adding the slave interface and starting bond interface. In this below path, when mtu is set for the slave interface, queues are not created yet and driver reference to queue[0] causes the segfault. slave_configure: rte_eth_dev_set_mtu rte_eth_dev_configure rte_eth_rx_queue_setup 2. segfault is seen on starting the port after removing from bond device. This is a testpmd bug where in, on starting the port, testpmd is supposed to recreate the queues before starting the port. Fixed these by adding check for queues created before accessing them. Fixes: daef48efe5e5 ("net/bnxt: support set MTU") Cc: stable@dpdk.org Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-24 07:44:27 +00:00
/* Exit if receive queues are not configured yet */
if (!eth_dev->data->nb_rx_queues)
return rc;
net/bnxt: implement SSE vector mode Introduce SSE vector mode support for the bnxt pmd. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-05-29 21:02:23 +00:00
new_pkt_size = new_mtu + RTE_ETHER_HDR_LEN + RTE_ETHER_CRC_LEN +
VLAN_TAG_SIZE * BNXT_NUM_VLANS;
/*
net/bnxt: reduce CQ queue size without aggregation ring Don't allocate extra completion queue entries for aggregation ring when aggregation ring will not be used. Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
2020-09-09 15:52:59 +00:00
* Disallow any MTU change that would require scattered receive support
* if it is not already enabled.
net/bnxt: implement SSE vector mode Introduce SSE vector mode support for the bnxt pmd. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-05-29 21:02:23 +00:00
*/
if (eth_dev->data->dev_started &&
net/bnxt: reduce CQ queue size without aggregation ring Don't allocate extra completion queue entries for aggregation ring when aggregation ring will not be used. Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
2020-09-09 15:52:59 +00:00
!eth_dev->data->scattered_rx &&
net/bnxt: implement SSE vector mode Introduce SSE vector mode support for the bnxt pmd. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-05-29 21:02:23 +00:00
(new_pkt_size >
eth_dev->data->min_rx_buf_size - RTE_PKTMBUF_HEADROOM)) {
PMD_DRV_LOG(ERR,
"MTU change would require scattered rx support. ");
PMD_DRV_LOG(ERR, "Stop port before changing MTU.\n");
return -EINVAL;
}
net: add rte prefix to ether defines Add 'RTE_' prefix to defines: - rename ETHER_ADDR_LEN as RTE_ETHER_ADDR_LEN. - rename ETHER_TYPE_LEN as RTE_ETHER_TYPE_LEN. - rename ETHER_CRC_LEN as RTE_ETHER_CRC_LEN. - rename ETHER_HDR_LEN as RTE_ETHER_HDR_LEN. - rename ETHER_MIN_LEN as RTE_ETHER_MIN_LEN. - rename ETHER_MAX_LEN as RTE_ETHER_MAX_LEN. - rename ETHER_MTU as RTE_ETHER_MTU. - rename ETHER_MAX_VLAN_FRAME_LEN as RTE_ETHER_MAX_VLAN_FRAME_LEN. - rename ETHER_MAX_VLAN_ID as RTE_ETHER_MAX_VLAN_ID. - rename ETHER_MAX_JUMBO_FRAME_LEN as RTE_ETHER_MAX_JUMBO_FRAME_LEN. - rename ETHER_MIN_MTU as RTE_ETHER_MIN_MTU. - rename ETHER_LOCAL_ADMIN_ADDR as RTE_ETHER_LOCAL_ADMIN_ADDR. - rename ETHER_GROUP_ADDR as RTE_ETHER_GROUP_ADDR. - rename ETHER_TYPE_IPv4 as RTE_ETHER_TYPE_IPv4. - rename ETHER_TYPE_IPv6 as RTE_ETHER_TYPE_IPv6. - rename ETHER_TYPE_ARP as RTE_ETHER_TYPE_ARP. - rename ETHER_TYPE_VLAN as RTE_ETHER_TYPE_VLAN. - rename ETHER_TYPE_RARP as RTE_ETHER_TYPE_RARP. - rename ETHER_TYPE_QINQ as RTE_ETHER_TYPE_QINQ. - rename ETHER_TYPE_ETAG as RTE_ETHER_TYPE_ETAG. - rename ETHER_TYPE_1588 as RTE_ETHER_TYPE_1588. - rename ETHER_TYPE_SLOW as RTE_ETHER_TYPE_SLOW. - rename ETHER_TYPE_TEB as RTE_ETHER_TYPE_TEB. - rename ETHER_TYPE_LLDP as RTE_ETHER_TYPE_LLDP. - rename ETHER_TYPE_MPLS as RTE_ETHER_TYPE_MPLS. - rename ETHER_TYPE_MPLSM as RTE_ETHER_TYPE_MPLSM. - rename ETHER_VXLAN_HLEN as RTE_ETHER_VXLAN_HLEN. - rename ETHER_ADDR_FMT_SIZE as RTE_ETHER_ADDR_FMT_SIZE. - rename VXLAN_GPE_TYPE_IPV4 as RTE_VXLAN_GPE_TYPE_IPV4. - rename VXLAN_GPE_TYPE_IPV6 as RTE_VXLAN_GPE_TYPE_IPV6. - rename VXLAN_GPE_TYPE_ETH as RTE_VXLAN_GPE_TYPE_ETH. - rename VXLAN_GPE_TYPE_NSH as RTE_VXLAN_GPE_TYPE_NSH. - rename VXLAN_GPE_TYPE_MPLS as RTE_VXLAN_GPE_TYPE_MPLS. - rename VXLAN_GPE_TYPE_GBP as RTE_VXLAN_GPE_TYPE_GBP. - rename VXLAN_GPE_TYPE_VBNG as RTE_VXLAN_GPE_TYPE_VBNG. - rename ETHER_VXLAN_GPE_HLEN as RTE_ETHER_VXLAN_GPE_HLEN. Do not update the command line library to avoid adding a dependency to librte_net. Signed-off-by: Olivier Matz <olivier.matz@6wind.com> Reviewed-by: Stephen Hemminger <stephen@networkplumber.org> Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2019-05-21 16:13:05 +00:00
if (new_mtu > RTE_ETHER_MTU) {
net/bnxt: support set MTU This patch adds support to modify MTU using the set_mtu dev_op. To support frames > 2k, the PMD creates an aggregator ring. When a frame greater than 2k is received, it is fragmented and the resulting fragments are DMA'ed to the aggregator ring. Now the driver can support jumbo frames upto 9500 bytes. Signed-off-by: Steeven Li <steeven.li@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:09 +00:00
bp->flags |= BNXT_FLAG_JUMBO;
net/bnxt: switch to the new offload API Update bnxt PMD to new ethdev offloads API. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-04-07 17:40:55 +00:00
bp->eth_dev->data->dev_conf.rxmode.offloads |=
DEV_RX_OFFLOAD_JUMBO_FRAME;
net/bnxt: support set MTU This patch adds support to modify MTU using the set_mtu dev_op. To support frames > 2k, the PMD creates an aggregator ring. When a frame greater than 2k is received, it is fragmented and the resulting fragments are DMA'ed to the aggregator ring. Now the driver can support jumbo frames upto 9500 bytes. Signed-off-by: Steeven Li <steeven.li@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:09 +00:00
} else {
net/bnxt: switch to the new offload API Update bnxt PMD to new ethdev offloads API. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-04-07 17:40:55 +00:00
bp->eth_dev->data->dev_conf.rxmode.offloads &=
~DEV_RX_OFFLOAD_JUMBO_FRAME;
net/bnxt: support set MTU This patch adds support to modify MTU using the set_mtu dev_op. To support frames > 2k, the PMD creates an aggregator ring. When a frame greater than 2k is received, it is fragmented and the resulting fragments are DMA'ed to the aggregator ring. Now the driver can support jumbo frames upto 9500 bytes. Signed-off-by: Steeven Li <steeven.li@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:09 +00:00
bp->flags &= ~BNXT_FLAG_JUMBO;
}
net/bnxt: fix crash after removing and adding slaves On removing the slave interface, slave_remove in bonding module calls _rte_eth_dev_reset which in turn frees both Tx and Rx queues. 1. segfault is seen after removing/adding the slave interface and starting bond interface. In this below path, when mtu is set for the slave interface, queues are not created yet and driver reference to queue[0] causes the segfault. slave_configure: rte_eth_dev_set_mtu rte_eth_dev_configure rte_eth_rx_queue_setup 2. segfault is seen on starting the port after removing from bond device. This is a testpmd bug where in, on starting the port, testpmd is supposed to recreate the queues before starting the port. Fixed these by adding check for queues created before accessing them. Fixes: daef48efe5e5 ("net/bnxt: support set MTU") Cc: stable@dpdk.org Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-24 07:44:27 +00:00
/* Is there a change in mtu setting? */
if (eth_dev->data->dev_conf.rxmode.max_rx_pkt_len == new_pkt_size)
return rc;
net/bnxt: support set MTU This patch adds support to modify MTU using the set_mtu dev_op. To support frames > 2k, the PMD creates an aggregator ring. When a frame greater than 2k is received, it is fragmented and the resulting fragments are DMA'ed to the aggregator ring. Now the driver can support jumbo frames upto 9500 bytes. Signed-off-by: Steeven Li <steeven.li@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:09 +00:00
for (i = 0; i < bp->nr_vnics; i++) {
struct bnxt_vnic_info *vnic = &bp->vnic_info[i];
net/bnxt: fix set MTU There is no need to update bnxt_hwrm_vnic_plcmode_cfg if new MTU is not greater than the max data the mbuf can accommodate. Fixes: daef48efe5e5 ("net/bnxt: support set MTU") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-06-28 20:15:43 +00:00
uint16_t size = 0;
net/bnxt: support set MTU This patch adds support to modify MTU using the set_mtu dev_op. To support frames > 2k, the PMD creates an aggregator ring. When a frame greater than 2k is received, it is fragmented and the resulting fragments are DMA'ed to the aggregator ring. Now the driver can support jumbo frames upto 9500 bytes. Signed-off-by: Steeven Li <steeven.li@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:09 +00:00
net/bnxt: fix ping with MTU change Driver was setting maximum receive unit differently in bnxt_mtu_set_op and bnxt_hwrm_vnic_alloc. Moreover firmware adds the 4 bytes for FCS. Fixed it by setting correct maximum receive unit value. Fixes: daef48efe5e5 ("net/bnxt: support set MTU") Cc: stable@dpdk.org Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-24 07:44:17 +00:00
vnic->mru = BNXT_VNIC_MRU(new_mtu);
net/bnxt: support set MTU This patch adds support to modify MTU using the set_mtu dev_op. To support frames > 2k, the PMD creates an aggregator ring. When a frame greater than 2k is received, it is fragmented and the resulting fragments are DMA'ed to the aggregator ring. Now the driver can support jumbo frames upto 9500 bytes. Signed-off-by: Steeven Li <steeven.li@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:09 +00:00
rc = bnxt_hwrm_vnic_cfg(bp, vnic);
if (rc)
break;
net/bnxt: fix set MTU There is no need to update bnxt_hwrm_vnic_plcmode_cfg if new MTU is not greater than the max data the mbuf can accommodate. Fixes: daef48efe5e5 ("net/bnxt: support set MTU") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-06-28 20:15:43 +00:00
size = rte_pktmbuf_data_room_size(bp->rx_queues[0]->mb_pool);
size -= RTE_PKTMBUF_HEADROOM;
if (size < new_mtu) {
rc = bnxt_hwrm_vnic_plcmode_cfg(bp, vnic);
if (rc)
return rc;
}
net/bnxt: support set MTU This patch adds support to modify MTU using the set_mtu dev_op. To support frames > 2k, the PMD creates an aggregator ring. When a frame greater than 2k is received, it is fragmented and the resulting fragments are DMA'ed to the aggregator ring. Now the driver can support jumbo frames upto 9500 bytes. Signed-off-by: Steeven Li <steeven.li@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:09 +00:00
}
net/bnxt: fix crash after removing and adding slaves On removing the slave interface, slave_remove in bonding module calls _rte_eth_dev_reset which in turn frees both Tx and Rx queues. 1. segfault is seen after removing/adding the slave interface and starting bond interface. In this below path, when mtu is set for the slave interface, queues are not created yet and driver reference to queue[0] causes the segfault. slave_configure: rte_eth_dev_set_mtu rte_eth_dev_configure rte_eth_rx_queue_setup 2. segfault is seen on starting the port after removing from bond device. This is a testpmd bug where in, on starting the port, testpmd is supposed to recreate the queues before starting the port. Fixed these by adding check for queues created before accessing them. Fixes: daef48efe5e5 ("net/bnxt: support set MTU") Cc: stable@dpdk.org Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-24 07:44:27 +00:00
if (!rc)
eth_dev->data->dev_conf.rxmode.max_rx_pkt_len = new_pkt_size;
net/bnxt: fix forwarding with higher mbuf size Maximum frame length supported by hw is 9600 bytes. When user launches testpmd with --mbuf-size=9729, we are posting buffer descriptors of size 9601 to the ring. This was causing packets getting discarded. Fixes: daef48efe5e5 ("net/bnxt: support set MTU") Cc: stable@dpdk.org Signed-off-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com>
2019-10-02 17:17:31 +00:00
PMD_DRV_LOG(INFO, "New MTU is %d\n", new_mtu);
net/bnxt: support set MTU This patch adds support to modify MTU using the set_mtu dev_op. To support frames > 2k, the PMD creates an aggregator ring. When a frame greater than 2k is received, it is fragmented and the resulting fragments are DMA'ed to the aggregator ring. Now the driver can support jumbo frames upto 9500 bytes. Signed-off-by: Steeven Li <steeven.li@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:09 +00:00
return rc;
}
net/bnxt: support VLAN pvid This patch adds code to support vlan_pvid_set dev_op Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:12 +00:00
static int
bnxt_vlan_pvid_set_op(struct rte_eth_dev *dev, uint16_t pvid, int on)
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = dev->data->dev_private;
net/bnxt: support VLAN pvid This patch adds code to support vlan_pvid_set dev_op Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:12 +00:00
uint16_t vlan = bp->vlan;
int rc;
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
rc = is_bnxt_in_error(bp);
if (rc)
return rc;
net/bnxt: fix single PF per port check The check BNXT_SINGLE_PF(bp) returns false for a VF. So there is no extra check needed for VF along with BNXT_SINGLE_PF(bp). Also make error messages more explicit. Fixes: ff947c6ce15f ("net/bnxt: add check for multi host PF per port") Fixes: f86febfb46da ("net/bnxt: support VF") Fixes: 3e12fdb78e82 ("net/bnxt: support VLAN pvid") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-04-29 05:53:00 +00:00
if (!BNXT_SINGLE_PF(bp)) {
PMD_DRV_LOG(ERR, "PVID cannot be modified on VF or on shared PF\n");
net/bnxt: support VLAN pvid This patch adds code to support vlan_pvid_set dev_op Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:12 +00:00
return -ENOTSUP;
}
bp->vlan = on ? pvid : 0;
rc = bnxt_hwrm_set_default_vlan(bp, 0, 0);
if (rc)
bp->vlan = vlan;
return rc;
}
net/bnxt: support LED on/off This patch adds support for dev_led_on/off dev_ops HWRM calls added: bnxt_hwrm_port_led_qcaps() bnxt_hwrm_port_led_cfg() Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:15 +00:00
static int
bnxt_dev_led_on_op(struct rte_eth_dev *dev)
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = dev->data->dev_private;
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
int rc;
rc = is_bnxt_in_error(bp);
if (rc)
return rc;
net/bnxt: support LED on/off This patch adds support for dev_led_on/off dev_ops HWRM calls added: bnxt_hwrm_port_led_qcaps() bnxt_hwrm_port_led_cfg() Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:15 +00:00
return bnxt_hwrm_port_led_cfg(bp, true);
}
static int
bnxt_dev_led_off_op(struct rte_eth_dev *dev)
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = dev->data->dev_private;
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
int rc;
rc = is_bnxt_in_error(bp);
if (rc)
return rc;
net/bnxt: support LED on/off This patch adds support for dev_led_on/off dev_ops HWRM calls added: bnxt_hwrm_port_led_qcaps() bnxt_hwrm_port_led_cfg() Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:15 +00:00
return bnxt_hwrm_port_led_cfg(bp, false);
}
net/bnxt: support Rx queue count add support for rx_queue_count dev_op Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:35 +00:00
static uint32_t
bnxt_rx_queue_count_op(struct rte_eth_dev *dev, uint16_t rx_queue_id)
{
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
struct bnxt *bp = (struct bnxt *)dev->data->dev_private;
net/bnxt: support Rx queue count add support for rx_queue_count dev_op Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:35 +00:00
struct bnxt_cp_ring_info *cpr;
net/bnxt: fix Rx queue count bnxt_rx_queue_count_op() incorrectly returns the number of filled but unprocessed completion queue entries instead of the number of filled but unprocessed received packet completions. Fix by properly accounting for the number of completion ring entries used by the various received packet completion types. Fixes: 34c0ba839bae ("net/bnxt: fix Rx queue count") Cc: stable@dpdk.org Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-02-18 18:19:20 +00:00
uint32_t desc = 0, raw_cons;
net/bnxt: support Rx queue count add support for rx_queue_count dev_op Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:35 +00:00
struct bnxt_rx_queue *rxq;
struct rx_pkt_cmpl *rxcmp;
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
int rc;
rc = is_bnxt_in_error(bp);
if (rc)
return rc;
net/bnxt: support Rx queue count add support for rx_queue_count dev_op Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:35 +00:00
rxq = dev->data->rx_queues[rx_queue_id];
cpr = rxq->cp_ring;
net/bnxt: fix Rx queue count Fix Computing of number of used descriptors in an Rx queue. Fixes: 1b7ceba3e375 ("net/bnxt: support Rx queue count") Cc: stable@dpdk.org Signed-off-by: Rahul Gupta <rahul.gupta@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 17:17:42 +00:00
raw_cons = cpr->cp_raw_cons;
net/bnxt: support Rx queue count add support for rx_queue_count dev_op Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:35 +00:00
net/bnxt: fix Rx queue count Fix Computing of number of used descriptors in an Rx queue. Fixes: 1b7ceba3e375 ("net/bnxt: support Rx queue count") Cc: stable@dpdk.org Signed-off-by: Rahul Gupta <rahul.gupta@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 17:17:42 +00:00
while (1) {
net/bnxt: fix Rx queue count bnxt_rx_queue_count_op() incorrectly returns the number of filled but unprocessed completion queue entries instead of the number of filled but unprocessed received packet completions. Fix by properly accounting for the number of completion ring entries used by the various received packet completion types. Fixes: 34c0ba839bae ("net/bnxt: fix Rx queue count") Cc: stable@dpdk.org Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-02-18 18:19:20 +00:00
uint32_t agg_cnt, cons, cmpl_type;
net/bnxt: support Rx queue count add support for rx_queue_count dev_op Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:35 +00:00
cons = RING_CMP(cpr->cp_ring_struct, raw_cons);
rxcmp = (struct rx_pkt_cmpl *)&cpr->cp_desc_ring[cons];
net/bnxt: fix Rx queue count bnxt_rx_queue_count_op() incorrectly returns the number of filled but unprocessed completion queue entries instead of the number of filled but unprocessed received packet completions. Fix by properly accounting for the number of completion ring entries used by the various received packet completion types. Fixes: 34c0ba839bae ("net/bnxt: fix Rx queue count") Cc: stable@dpdk.org Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-02-18 18:19:20 +00:00
if (!CMP_VALID(rxcmp, raw_cons, cpr->cp_ring_struct))
net/bnxt: fix Rx queue count Fix Computing of number of used descriptors in an Rx queue. Fixes: 1b7ceba3e375 ("net/bnxt: support Rx queue count") Cc: stable@dpdk.org Signed-off-by: Rahul Gupta <rahul.gupta@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 17:17:42 +00:00
break;
net/bnxt: fix Rx queue count bnxt_rx_queue_count_op() incorrectly returns the number of filled but unprocessed completion queue entries instead of the number of filled but unprocessed received packet completions. Fix by properly accounting for the number of completion ring entries used by the various received packet completion types. Fixes: 34c0ba839bae ("net/bnxt: fix Rx queue count") Cc: stable@dpdk.org Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-02-18 18:19:20 +00:00
cmpl_type = CMP_TYPE(rxcmp);
switch (cmpl_type) {
case CMPL_BASE_TYPE_RX_L2:
case CMPL_BASE_TYPE_RX_L2_V2:
agg_cnt = BNXT_RX_L2_AGG_BUFS(rxcmp);
raw_cons = raw_cons + CMP_LEN(cmpl_type) + agg_cnt;
desc++;
break;
case CMPL_BASE_TYPE_RX_TPA_END:
if (BNXT_CHIP_P5(rxq->bp)) {
struct rx_tpa_v2_end_cmpl_hi *p5_tpa_end;
p5_tpa_end = (void *)rxcmp;
agg_cnt = BNXT_TPA_END_AGG_BUFS_TH(p5_tpa_end);
} else {
struct rx_tpa_end_cmpl *tpa_end;
tpa_end = (void *)rxcmp;
agg_cnt = BNXT_TPA_END_AGG_BUFS(tpa_end);
}
raw_cons = raw_cons + CMP_LEN(cmpl_type) + agg_cnt;
net/bnxt: fix Rx queue count Fix Computing of number of used descriptors in an Rx queue. Fixes: 1b7ceba3e375 ("net/bnxt: support Rx queue count") Cc: stable@dpdk.org Signed-off-by: Rahul Gupta <rahul.gupta@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 17:17:42 +00:00
desc++;
net/bnxt: fix Rx queue count bnxt_rx_queue_count_op() incorrectly returns the number of filled but unprocessed completion queue entries instead of the number of filled but unprocessed received packet completions. Fix by properly accounting for the number of completion ring entries used by the various received packet completion types. Fixes: 34c0ba839bae ("net/bnxt: fix Rx queue count") Cc: stable@dpdk.org Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-02-18 18:19:20 +00:00
break;
default:
raw_cons += CMP_LEN(cmpl_type);
net/bnxt: support Rx queue count add support for rx_queue_count dev_op Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:35 +00:00
}
}
return desc;
}
net/bnxt: support Rx descriptor status Add support for rx_descriptor_status dev_op Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:36 +00:00
static int
bnxt_rx_descriptor_status_op(void *rx_queue, uint16_t offset)
{
net/bnxt: fix Rx descriptor status Fix a number of issues in the bnxt receive descriptor status function, including: - Provide status of receive descriptor instead of completion descriptor. - Remove invalid comparison of raw ring index with masked ring index. - Correct misinterpretation of offset parameter as ring index. - Correct misuse of completion ring index for mbuf ring (the two rings have different sizes). Fixes: 0fe613bb87b2 ("net/bnxt: support Rx descriptor status") Cc: stable@dpdk.org Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Andy Gospodarek <gospo@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-03-02 14:29:27 +00:00
struct bnxt_rx_queue *rxq = rx_queue;
net/bnxt: support Rx descriptor status Add support for rx_descriptor_status dev_op Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:36 +00:00
struct bnxt_cp_ring_info *cpr;
net/bnxt: fix Rx descriptor status Fix a number of issues in the bnxt receive descriptor status function, including: - Provide status of receive descriptor instead of completion descriptor. - Remove invalid comparison of raw ring index with masked ring index. - Correct misinterpretation of offset parameter as ring index. - Correct misuse of completion ring index for mbuf ring (the two rings have different sizes). Fixes: 0fe613bb87b2 ("net/bnxt: support Rx descriptor status") Cc: stable@dpdk.org Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Andy Gospodarek <gospo@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-03-02 14:29:27 +00:00
struct bnxt_rx_ring_info *rxr;
uint32_t desc, raw_cons;
struct bnxt *bp = rxq->bp;
net/bnxt: support Rx descriptor status Add support for rx_descriptor_status dev_op Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:36 +00:00
struct rx_pkt_cmpl *rxcmp;
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
int rc;
net/bnxt: support Rx descriptor status Add support for rx_descriptor_status dev_op Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:36 +00:00
net/bnxt: fix Rx descriptor status Fix a number of issues in the bnxt receive descriptor status function, including: - Provide status of receive descriptor instead of completion descriptor. - Remove invalid comparison of raw ring index with masked ring index. - Correct misinterpretation of offset parameter as ring index. - Correct misuse of completion ring index for mbuf ring (the two rings have different sizes). Fixes: 0fe613bb87b2 ("net/bnxt: support Rx descriptor status") Cc: stable@dpdk.org Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Andy Gospodarek <gospo@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-03-02 14:29:27 +00:00
rc = is_bnxt_in_error(bp);
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
if (rc)
return rc;
net/bnxt: support Rx descriptor status Add support for rx_descriptor_status dev_op Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:36 +00:00
if (offset >= rxq->nb_rx_desc)
return -EINVAL;
net/bnxt: fix Rx descriptor status Fix a number of issues in the bnxt receive descriptor status function, including: - Provide status of receive descriptor instead of completion descriptor. - Remove invalid comparison of raw ring index with masked ring index. - Correct misinterpretation of offset parameter as ring index. - Correct misuse of completion ring index for mbuf ring (the two rings have different sizes). Fixes: 0fe613bb87b2 ("net/bnxt: support Rx descriptor status") Cc: stable@dpdk.org Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Andy Gospodarek <gospo@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-03-02 14:29:27 +00:00
rxr = rxq->rx_ring;
cpr = rxq->cp_ring;
net/bnxt: support Rx descriptor status Add support for rx_descriptor_status dev_op Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:36 +00:00
net/bnxt: fix Rx descriptor status Fix a number of issues in the bnxt receive descriptor status function, including: - Provide status of receive descriptor instead of completion descriptor. - Remove invalid comparison of raw ring index with masked ring index. - Correct misinterpretation of offset parameter as ring index. - Correct misuse of completion ring index for mbuf ring (the two rings have different sizes). Fixes: 0fe613bb87b2 ("net/bnxt: support Rx descriptor status") Cc: stable@dpdk.org Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Andy Gospodarek <gospo@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-03-02 14:29:27 +00:00
/*
* For the vector receive case, the completion at the requested
* offset can be indexed directly.
*/
#if defined(RTE_ARCH_X86) || defined(RTE_ARCH_ARM64)
if (bp->flags & BNXT_FLAG_RX_VECTOR_PKT_MODE) {
struct rx_pkt_cmpl *rxcmp;
uint32_t cons;
/* Check status of completion descriptor. */
raw_cons = cpr->cp_raw_cons +
offset * CMP_LEN(CMPL_BASE_TYPE_RX_L2);
cons = RING_CMP(cpr->cp_ring_struct, raw_cons);
rxcmp = (struct rx_pkt_cmpl *)&cpr->cp_desc_ring[cons];
if (CMP_VALID(rxcmp, raw_cons, cpr->cp_ring_struct))
net/bnxt: support Rx descriptor status Add support for rx_descriptor_status dev_op Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:36 +00:00
return RTE_ETH_RX_DESC_DONE;
net/bnxt: fix Rx descriptor status Fix a number of issues in the bnxt receive descriptor status function, including: - Provide status of receive descriptor instead of completion descriptor. - Remove invalid comparison of raw ring index with masked ring index. - Correct misinterpretation of offset parameter as ring index. - Correct misuse of completion ring index for mbuf ring (the two rings have different sizes). Fixes: 0fe613bb87b2 ("net/bnxt: support Rx descriptor status") Cc: stable@dpdk.org Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Andy Gospodarek <gospo@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-03-02 14:29:27 +00:00
/* Check whether rx desc has an mbuf attached. */
cons = RING_CMP(rxr->rx_ring_struct, raw_cons / 2);
if (cons >= rxq->rxrearm_start &&
cons < rxq->rxrearm_start + rxq->rxrearm_nb) {
return RTE_ETH_RX_DESC_UNAVAIL;
}
return RTE_ETH_RX_DESC_AVAIL;
net/bnxt: support Rx descriptor status Add support for rx_descriptor_status dev_op Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:36 +00:00
}
net/bnxt: fix Rx descriptor status Fix a number of issues in the bnxt receive descriptor status function, including: - Provide status of receive descriptor instead of completion descriptor. - Remove invalid comparison of raw ring index with masked ring index. - Correct misinterpretation of offset parameter as ring index. - Correct misuse of completion ring index for mbuf ring (the two rings have different sizes). Fixes: 0fe613bb87b2 ("net/bnxt: support Rx descriptor status") Cc: stable@dpdk.org Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Andy Gospodarek <gospo@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-03-02 14:29:27 +00:00
#endif
net/bnxt: support Rx descriptor status Add support for rx_descriptor_status dev_op Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:36 +00:00
net/bnxt: fix Rx descriptor status Fix a number of issues in the bnxt receive descriptor status function, including: - Provide status of receive descriptor instead of completion descriptor. - Remove invalid comparison of raw ring index with masked ring index. - Correct misinterpretation of offset parameter as ring index. - Correct misuse of completion ring index for mbuf ring (the two rings have different sizes). Fixes: 0fe613bb87b2 ("net/bnxt: support Rx descriptor status") Cc: stable@dpdk.org Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Andy Gospodarek <gospo@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-03-02 14:29:27 +00:00
/*
* For the non-vector receive case, scan the completion ring to
* locate the completion descriptor for the requested offset.
*/
raw_cons = cpr->cp_raw_cons;
desc = 0;
while (1) {
uint32_t agg_cnt, cons, cmpl_type;
cons = RING_CMP(cpr->cp_ring_struct, raw_cons);
rxcmp = (struct rx_pkt_cmpl *)&cpr->cp_desc_ring[cons];
if (!CMP_VALID(rxcmp, raw_cons, cpr->cp_ring_struct))
break;
cmpl_type = CMP_TYPE(rxcmp);
switch (cmpl_type) {
case CMPL_BASE_TYPE_RX_L2:
case CMPL_BASE_TYPE_RX_L2_V2:
if (desc == offset) {
cons = rxcmp->opaque;
if (rxr->rx_buf_ring[cons])
return RTE_ETH_RX_DESC_DONE;
else
return RTE_ETH_RX_DESC_UNAVAIL;
}
agg_cnt = BNXT_RX_L2_AGG_BUFS(rxcmp);
raw_cons = raw_cons + CMP_LEN(cmpl_type) + agg_cnt;
desc++;
break;
case CMPL_BASE_TYPE_RX_TPA_END:
if (desc == offset)
return RTE_ETH_RX_DESC_DONE;
if (BNXT_CHIP_P5(rxq->bp)) {
struct rx_tpa_v2_end_cmpl_hi *p5_tpa_end;
p5_tpa_end = (void *)rxcmp;
agg_cnt = BNXT_TPA_END_AGG_BUFS_TH(p5_tpa_end);
} else {
struct rx_tpa_end_cmpl *tpa_end;
tpa_end = (void *)rxcmp;
agg_cnt = BNXT_TPA_END_AGG_BUFS(tpa_end);
}
raw_cons = raw_cons + CMP_LEN(cmpl_type) + agg_cnt;
desc++;
break;
default:
raw_cons += CMP_LEN(cmpl_type);
}
}
net/bnxt: support Rx descriptor status Add support for rx_descriptor_status dev_op Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:36 +00:00
return RTE_ETH_RX_DESC_AVAIL;
}
net/bnxt: support Tx descriptor status Add support for tx_descriptor_status dev_op Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:37 +00:00
static int
bnxt_tx_descriptor_status_op(void *tx_queue, uint16_t offset)
{
struct bnxt_tx_queue *txq = (struct bnxt_tx_queue *)tx_queue;
struct bnxt_tx_ring_info *txr;
struct bnxt_cp_ring_info *cpr;
net/bnxt: optimize Tx completion handling Avoid copying mbuf pointers to separate array for bulk mbuf free when handling transmit completions for vector mode transmit. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-03-06 15:19:11 +00:00
struct rte_mbuf **tx_buf;
net/bnxt: support Tx descriptor status Add support for tx_descriptor_status dev_op Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:37 +00:00
struct tx_pkt_cmpl *txcmp;
uint32_t cons, cp_cons;
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
int rc;
net/bnxt: support Tx descriptor status Add support for tx_descriptor_status dev_op Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:37 +00:00
if (!txq)
return -EINVAL;
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
rc = is_bnxt_in_error(txq->bp);
if (rc)
return rc;
net/bnxt: support Tx descriptor status Add support for tx_descriptor_status dev_op Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:37 +00:00
cpr = txq->cp_ring;
txr = txq->tx_ring;
if (offset >= txq->nb_tx_desc)
return -EINVAL;
cons = RING_CMP(cpr->cp_ring_struct, offset);
txcmp = (struct tx_pkt_cmpl *)&cpr->cp_desc_ring[cons];
cp_cons = cpr->cp_raw_cons;
if (cons > cp_cons) {
if (CMPL_VALID(txcmp, cpr->valid))
return RTE_ETH_TX_DESC_UNAVAIL;
} else {
if (CMPL_VALID(txcmp, !cpr->valid))
return RTE_ETH_TX_DESC_UNAVAIL;
}
tx_buf = &txr->tx_buf_ring[cons];
net/bnxt: optimize Tx completion handling Avoid copying mbuf pointers to separate array for bulk mbuf free when handling transmit completions for vector mode transmit. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-03-06 15:19:11 +00:00
if (*tx_buf == NULL)
net/bnxt: support Tx descriptor status Add support for tx_descriptor_status dev_op Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:37 +00:00
return RTE_ETH_TX_DESC_DONE;
return RTE_ETH_TX_DESC_FULL;
}
net/bnxt: enable filter ctrl ops for port representor Inorder to offload flows on the vfrep device, it must be populated with rte_flow_ops. This patch enables the same. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-06 08:25:01 +00:00
int
ethdev: replace callback getting filter operations Since rte_flow is the only API for filtering operations, the legacy driver interface filter_ctrl was too much complicated for the simple task of getting the struct rte_flow_ops. The filter type RTE_ETH_FILTER_GENERIC and the filter operarion RTE_ETH_FILTER_GET are removed. The new driver callback flow_ops_get replaces filter_ctrl. Signed-off-by: Thomas Monjalon <thomas@monjalon.net> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Acked-by: Haiyue Wang <haiyue.wang@intel.com> Acked-by: Rosen Xu <rosen.xu@intel.com> Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com> Reviewed-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2021-03-21 09:00:00 +00:00
bnxt_flow_ops_get_op(struct rte_eth_dev *dev,
const struct rte_flow_ops **ops)
net/bnxt: support flow filter ops This patch adds support for flow validate/create/destroy/flush, ethertype add/del ops Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:39 +00:00
{
net/bnxt: register TruFlow flow API ops Register bnxt_ulp_rte_flow_ops when host based TRUFLOW is enabled. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-04-15 08:19:08 +00:00
struct bnxt *bp = dev->data->dev_private;
net/bnxt: support flow filter ops This patch adds support for flow validate/create/destroy/flush, ethertype add/del ops Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:39 +00:00
int ret = 0;
net/bnxt: fix port stop process and cleanup resources The port deinitialization now cleans up all the resources properly. If all the ports are stopped then ULP context is freed. Added fix to update the correct tfp pointer in the ULP context with the changes to support multi control channels. Fixes: 70e64b27af5b ("net/bnxt: support ULP session manager cleanup") Cc: stable@dpdk.org Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Shahaji Bhosle <sbhosle@broadcom.com>
2020-09-11 01:55:39 +00:00
if (!bp)
return -EIO;
net/bnxt: enable filter ctrl ops for port representor Inorder to offload flows on the vfrep device, it must be populated with rte_flow_ops. This patch enables the same. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-06 08:25:01 +00:00
if (BNXT_ETH_DEV_IS_REPRESENTOR(dev)) {
net/bnxt: support representors on remote host domain In the Stingray use case, representors are conventionally run inside the SoC domain representing functions that are on the X86 domain. In order to support this mechanism of building representors for endpoints that are not in the same host domain, additional dev args have been in the PMD like so: rep-based-pf=<physical index> rep-is-pf=<VF=0 or PF=1> where `rep-based-pf` specifies the physical index of the base PF that the representor is derived off of. Since representor(s) can be created for endpoint PFs as well, rename struct bnxt_vf_representor to bnxt_representor and other such dev_ops and function names. devargs have also been extended to specify the exact CoS queue along with flow control enablement to be used for the conduit between the representor and the endpoint function. This is how a sample devargs would look with all the extended devargs -w 0000:06:02.0,host-based-truflow=1,representor=[1],rep-based-pf=8, rep-is-pf=1,rep-q-r2f=1,rep-fc-r2f=0,rep-q-f2r=1,rep-fc-f2r=1 Call CFA_PAIR_ALLOC only in case of Stingray instead of CFA_VFR_ALLOC. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:30 +00:00
struct bnxt_representor *vfr = dev->data->dev_private;
net/bnxt: enable filter ctrl ops for port representor Inorder to offload flows on the vfrep device, it must be populated with rte_flow_ops. This patch enables the same. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-06 08:25:01 +00:00
bp = vfr->parent_dev->data->dev_private;
net/bnxt: fix function id used in flow flush The function id being used in the flush is incorrect, fixed the flush of the flows. Fixes: 74bcfc062489 ("net/bnxt: add session and function flow flush") Cc: stable@dpdk.org Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Reviewed-by: Shahaji Bhosle <sbhosle@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-11 01:55:47 +00:00
/* parent is deleted while children are still valid */
net/bnxt: refactor VFR port clean up When parent VF or PF ports are cleaned up, the child VF representor ports also need to be cleaned up. If not cleaned up, then deleting the parent VF shall result in not cleaning up the hardware rules and updating the firmware of VFR removal. The issue can occur even when application is exited without deleting VFR ports. Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Shahaji Bhosle <sbhosle@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-11 01:55:48 +00:00
if (!bp) {
ethdev: replace callback getting filter operations Since rte_flow is the only API for filtering operations, the legacy driver interface filter_ctrl was too much complicated for the simple task of getting the struct rte_flow_ops. The filter type RTE_ETH_FILTER_GENERIC and the filter operarion RTE_ETH_FILTER_GET are removed. The new driver callback flow_ops_get replaces filter_ctrl. Signed-off-by: Thomas Monjalon <thomas@monjalon.net> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Acked-by: Haiyue Wang <haiyue.wang@intel.com> Acked-by: Rosen Xu <rosen.xu@intel.com> Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com> Reviewed-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2021-03-21 09:00:00 +00:00
PMD_DRV_LOG(DEBUG, "BNXT Port:%d VFR Error\n",
dev->data->port_id);
net/bnxt: fix port stop process and cleanup resources The port deinitialization now cleans up all the resources properly. If all the ports are stopped then ULP context is freed. Added fix to update the correct tfp pointer in the ULP context with the changes to support multi control channels. Fixes: 70e64b27af5b ("net/bnxt: support ULP session manager cleanup") Cc: stable@dpdk.org Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Shahaji Bhosle <sbhosle@broadcom.com>
2020-09-11 01:55:39 +00:00
return -EIO;
net/bnxt: refactor VFR port clean up When parent VF or PF ports are cleaned up, the child VF representor ports also need to be cleaned up. If not cleaned up, then deleting the parent VF shall result in not cleaning up the hardware rules and updating the firmware of VFR removal. The issue can occur even when application is exited without deleting VFR ports. Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Shahaji Bhosle <sbhosle@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-11 01:55:48 +00:00
}
net/bnxt: enable filter ctrl ops for port representor Inorder to offload flows on the vfrep device, it must be populated with rte_flow_ops. This patch enables the same. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-06 08:25:01 +00:00
}
ret = is_bnxt_in_error(bp);
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
if (ret)
return ret;
ethdev: replace callback getting filter operations Since rte_flow is the only API for filtering operations, the legacy driver interface filter_ctrl was too much complicated for the simple task of getting the struct rte_flow_ops. The filter type RTE_ETH_FILTER_GENERIC and the filter operarion RTE_ETH_FILTER_GET are removed. The new driver callback flow_ops_get replaces filter_ctrl. Signed-off-by: Thomas Monjalon <thomas@monjalon.net> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Acked-by: Haiyue Wang <haiyue.wang@intel.com> Acked-by: Rosen Xu <rosen.xu@intel.com> Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com> Reviewed-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2021-03-21 09:00:00 +00:00
/* PMD supports thread-safe flow operations. rte_flow API
* functions can avoid mutex for multi-thread safety.
*/
dev->data->dev_flags |= RTE_ETH_DEV_FLOW_OPS_THREAD_SAFE;
net/bnxt: set thread safe flow ops flag PMD supports thread-safe flow operations. Set the RTE_ETH_DEV_FLOW_OPS_THREAD_SAFE dev_flag to indicate this info to the application. rte_flow API functions can avoid using its own mutex for safe multi-thread flow handling. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-10-26 03:56:15 +00:00
ethdev: replace callback getting filter operations Since rte_flow is the only API for filtering operations, the legacy driver interface filter_ctrl was too much complicated for the simple task of getting the struct rte_flow_ops. The filter type RTE_ETH_FILTER_GENERIC and the filter operarion RTE_ETH_FILTER_GET are removed. The new driver callback flow_ops_get replaces filter_ctrl. Signed-off-by: Thomas Monjalon <thomas@monjalon.net> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Acked-by: Haiyue Wang <haiyue.wang@intel.com> Acked-by: Rosen Xu <rosen.xu@intel.com> Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com> Reviewed-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2021-03-21 09:00:00 +00:00
if (BNXT_TRUFLOW_EN(bp))
*ops = &bnxt_ulp_rte_flow_ops;
else
*ops = &bnxt_flow_ops;
net/bnxt: set thread safe flow ops flag PMD supports thread-safe flow operations. Set the RTE_ETH_DEV_FLOW_OPS_THREAD_SAFE dev_flag to indicate this info to the application. rte_flow API functions can avoid using its own mutex for safe multi-thread flow handling. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-10-26 03:56:15 +00:00
net/bnxt: support flow filter ops This patch adds support for flow validate/create/destroy/flush, ethertype add/del ops Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:39 +00:00
return ret;
}
net/bnxt: support getting ptypes This patch adds support for dev_supported_ptypes_get Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:43 +00:00
static const uint32_t *
bnxt_dev_supported_ptypes_get_op(struct rte_eth_dev *dev)
{
static const uint32_t ptypes[] = {
RTE_PTYPE_L2_ETHER_VLAN,
RTE_PTYPE_L3_IPV4_EXT_UNKNOWN,
RTE_PTYPE_L3_IPV6_EXT_UNKNOWN,
RTE_PTYPE_L4_ICMP,
RTE_PTYPE_L4_TCP,
RTE_PTYPE_L4_UDP,
RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN,
RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN,
RTE_PTYPE_INNER_L4_ICMP,
RTE_PTYPE_INNER_L4_TCP,
RTE_PTYPE_INNER_L4_UDP,
RTE_PTYPE_UNKNOWN
};
net/bnxt: implement SSE vector mode Introduce SSE vector mode support for the bnxt pmd. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-05-29 21:02:23 +00:00
if (!dev->rx_pkt_burst)
return NULL;
return ptypes;
net/bnxt: support getting ptypes This patch adds support for dev_supported_ptypes_get Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:43 +00:00
}
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
static int bnxt_map_regs(struct bnxt *bp, uint32_t *reg_arr, int count,
int reg_win)
{
uint32_t reg_base = *reg_arr & 0xfffff000;
uint32_t win_off;
int i;
for (i = 0; i < count; i++) {
if ((reg_arr[i] & 0xfffff000) != reg_base)
return -ERANGE;
}
win_off = BNXT_GRCPF_REG_WINDOW_BASE_OUT + (reg_win - 1) * 4;
net/bnxt: fix big endian build The result of the endian conversion was not being used incorrectly. Fixes: b11cceb83a34 ("net/bnxt: support timesync") Cc: stable@dpdk.org Reported-by: Thomas Monjalon <thomas@monjalon.net> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-04-10 09:16:22 +00:00
rte_write32(reg_base, (uint8_t *)bp->bar0 + win_off);
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
return 0;
}
static int bnxt_map_ptp_regs(struct bnxt *bp)
{
struct bnxt_ptp_cfg *ptp = bp->ptp_cfg;
uint32_t *reg_arr;
int rc, i;
reg_arr = ptp->rx_regs;
rc = bnxt_map_regs(bp, reg_arr, BNXT_PTP_RX_REGS, 5);
if (rc)
return rc;
reg_arr = ptp->tx_regs;
rc = bnxt_map_regs(bp, reg_arr, BNXT_PTP_TX_REGS, 6);
if (rc)
return rc;
for (i = 0; i < BNXT_PTP_RX_REGS; i++)
ptp->rx_mapped_regs[i] = 0x5000 + (ptp->rx_regs[i] & 0xfff);
for (i = 0; i < BNXT_PTP_TX_REGS; i++)
ptp->tx_mapped_regs[i] = 0x6000 + (ptp->tx_regs[i] & 0xfff);
return 0;
}
static void bnxt_unmap_ptp_regs(struct bnxt *bp)
{
net/bnxt: fix big endian build The result of the endian conversion was not being used incorrectly. Fixes: b11cceb83a34 ("net/bnxt: support timesync") Cc: stable@dpdk.org Reported-by: Thomas Monjalon <thomas@monjalon.net> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-04-10 09:16:22 +00:00
rte_write32(0, (uint8_t *)bp->bar0 +
BNXT_GRCPF_REG_WINDOW_BASE_OUT + 16);
rte_write32(0, (uint8_t *)bp->bar0 +
BNXT_GRCPF_REG_WINDOW_BASE_OUT + 20);
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
}
static uint64_t bnxt_cc_read(struct bnxt *bp)
{
uint64_t ns;
ns = rte_le_to_cpu_32(rte_read32((uint8_t *)bp->bar0 +
BNXT_GRCPF_REG_SYNC_TIME));
ns |= (uint64_t)(rte_le_to_cpu_32(rte_read32((uint8_t *)bp->bar0 +
BNXT_GRCPF_REG_SYNC_TIME + 4))) << 32;
return ns;
}
static int bnxt_get_tx_ts(struct bnxt *bp, uint64_t *ts)
{
struct bnxt_ptp_cfg *ptp = bp->ptp_cfg;
uint32_t fifo;
fifo = rte_le_to_cpu_32(rte_read32((uint8_t *)bp->bar0 +
ptp->tx_mapped_regs[BNXT_PTP_TX_FIFO]));
if (fifo & BNXT_PTP_TX_FIFO_EMPTY)
return -EAGAIN;
fifo = rte_le_to_cpu_32(rte_read32((uint8_t *)bp->bar0 +
ptp->tx_mapped_regs[BNXT_PTP_TX_FIFO]));
*ts = rte_le_to_cpu_32(rte_read32((uint8_t *)bp->bar0 +
ptp->tx_mapped_regs[BNXT_PTP_TX_TS_L]));
*ts |= (uint64_t)rte_le_to_cpu_32(rte_read32((uint8_t *)bp->bar0 +
ptp->tx_mapped_regs[BNXT_PTP_TX_TS_H])) << 32;
net/bnxt: fix Tx timestamp init Fix to read the sequence ID register to get Tx timestamp. Reading the sequence ID register is necessary for the HW FIFO to advance and thereby get the correct value of the timestamp on Tx side. This patch fixes that. Fixes: b11cceb83a34 ("net/bnxt: support timesync") Cc: stable@dpdk.org Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
2021-03-16 05:40:48 +00:00
rte_read32((uint8_t *)bp->bar0 + ptp->tx_mapped_regs[BNXT_PTP_TX_SEQ]);
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
return 0;
}
net/bnxt: fix Rx timestamp when FIFO pending bit is set Fix to clear the Rx FIFO while reading the timestamp. If the Rx FIFO has pending bit set, keep reading to clear it and return the last valid timestamp instead of unconditionally returning an error. Fixes: b11cceb83a34 ("net/bnxt: support timesync") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-04-26 06:07:55 +00:00
static int bnxt_clr_rx_ts(struct bnxt *bp, uint64_t *last_ts)
{
struct bnxt_ptp_cfg *ptp = bp->ptp_cfg;
struct bnxt_pf_info *pf = bp->pf;
uint16_t port_id;
int i = 0;
uint32_t fifo;
if (!ptp || (bp->flags & BNXT_FLAG_CHIP_P5))
return -EINVAL;
port_id = pf->port_id;
fifo = rte_le_to_cpu_32(rte_read32((uint8_t *)bp->bar0 +
ptp->rx_mapped_regs[BNXT_PTP_RX_FIFO]));
while ((fifo & BNXT_PTP_RX_FIFO_PENDING) && (i < BNXT_PTP_RX_PND_CNT)) {
rte_write32(1 << port_id, (uint8_t *)bp->bar0 +
ptp->rx_mapped_regs[BNXT_PTP_RX_FIFO_ADV]);
fifo = rte_le_to_cpu_32(rte_read32((uint8_t *)bp->bar0 +
ptp->rx_mapped_regs[BNXT_PTP_RX_FIFO]));
*last_ts = rte_le_to_cpu_32(rte_read32((uint8_t *)bp->bar0 +
ptp->rx_mapped_regs[BNXT_PTP_RX_TS_L]));
*last_ts |= (uint64_t)rte_le_to_cpu_32(rte_read32((uint8_t *)bp->bar0 +
ptp->rx_mapped_regs[BNXT_PTP_RX_TS_H])) << 32;
i++;
}
if (i >= BNXT_PTP_RX_PND_CNT)
return -EBUSY;
return 0;
}
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
static int bnxt_get_rx_ts(struct bnxt *bp, uint64_t *ts)
{
struct bnxt_ptp_cfg *ptp = bp->ptp_cfg;
net/bnxt: fix allocation of PF info struct Fixes: 804e746c7b73 ("net/bnxt: add hardware resource manager init code") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 11:10:41 +00:00
struct bnxt_pf_info *pf = bp->pf;
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
uint16_t port_id;
uint32_t fifo;
fifo = rte_le_to_cpu_32(rte_read32((uint8_t *)bp->bar0 +
ptp->rx_mapped_regs[BNXT_PTP_RX_FIFO]));
if (!(fifo & BNXT_PTP_RX_FIFO_PENDING))
return -EAGAIN;
port_id = pf->port_id;
net/bnxt: fix big endian build The result of the endian conversion was not being used incorrectly. Fixes: b11cceb83a34 ("net/bnxt: support timesync") Cc: stable@dpdk.org Reported-by: Thomas Monjalon <thomas@monjalon.net> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-04-10 09:16:22 +00:00
rte_write32(1 << port_id, (uint8_t *)bp->bar0 +
ptp->rx_mapped_regs[BNXT_PTP_RX_FIFO_ADV]);
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
fifo = rte_le_to_cpu_32(rte_read32((uint8_t *)bp->bar0 +
ptp->rx_mapped_regs[BNXT_PTP_RX_FIFO]));
net/bnxt: fix Rx timestamp when FIFO pending bit is set Fix to clear the Rx FIFO while reading the timestamp. If the Rx FIFO has pending bit set, keep reading to clear it and return the last valid timestamp instead of unconditionally returning an error. Fixes: b11cceb83a34 ("net/bnxt: support timesync") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-04-26 06:07:55 +00:00
if (fifo & BNXT_PTP_RX_FIFO_PENDING)
return bnxt_clr_rx_ts(bp, ts);
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
*ts = rte_le_to_cpu_32(rte_read32((uint8_t *)bp->bar0 +
ptp->rx_mapped_regs[BNXT_PTP_RX_TS_L]));
*ts |= (uint64_t)rte_le_to_cpu_32(rte_read32((uint8_t *)bp->bar0 +
ptp->rx_mapped_regs[BNXT_PTP_RX_TS_H])) << 32;
return 0;
}
static int
bnxt_timesync_write_time(struct rte_eth_dev *dev, const struct timespec *ts)
{
uint64_t ns;
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = dev->data->dev_private;
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
struct bnxt_ptp_cfg *ptp = bp->ptp_cfg;
if (!ptp)
net/bnxt: fix timesync when PTP is not supported Fixed to return error when PTP support is not supported on the port. Also, removed an unnecessary check inside bnxt_get_rx_ts(). Fixes: b11cceb83a34 ("net/bnxt: support timesync") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2021-03-18 09:35:23 +00:00
return -ENOTSUP;
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
ns = rte_timespec_to_ns(ts);
/* Set the timecounters to a new value. */
ptp->tc.nsec = ns;
net/bnxt: fix Rx and Tx timestamps timesync adjust and write_time APIs needed to account for the Rx and Tx timestamp counters as well. Fix it since it was not done earlier. Fixes: b11cceb83a34 ("net/bnxt: support timesync") Cc: stable@dpdk.org Reviewed-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
2021-03-12 05:21:09 +00:00
ptp->tx_tstamp_tc.nsec = ns;
ptp->rx_tstamp_tc.nsec = ns;
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
return 0;
}
static int
bnxt_timesync_read_time(struct rte_eth_dev *dev, struct timespec *ts)
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = dev->data->dev_private;
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
struct bnxt_ptp_cfg *ptp = bp->ptp_cfg;
net/bnxt: support PTP for Thor On Thor, direct access to PTP registers (via GRC) is not supported. Driver must use HWRM to access the timestamp information. Vectorized Rx/Tx cannot be enabled if RTE_LIBRTE_IEEE1588=y. Remove the PTP flags handling code from the vector Rx path. Add support to read tx timestamp value and the time from the timesync clock. On Thor, Rx timestamps are provided directly in the Rx completion records to the driver. Only 32 bits of the timestamp is present in the completion. Driver needs to read the current 48 bit free running timer using the HWRM_PORT_TS_QUERY command and combine the upper 16 bits from the HWRM response with the lower 32 bits in the Rx completion to produce the 48 bit timestamp for the Rx packet. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 01:23:35 +00:00
uint64_t ns, systime_cycles = 0;
int rc = 0;
net/bnxt: support getting ptypes This patch adds support for dev_supported_ptypes_get Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:43 +00:00
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
if (!ptp)
net/bnxt: fix timesync when PTP is not supported Fixed to return error when PTP support is not supported on the port. Also, removed an unnecessary check inside bnxt_get_rx_ts(). Fixes: b11cceb83a34 ("net/bnxt: support timesync") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2021-03-18 09:35:23 +00:00
return -ENOTSUP;
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
net/bnxt: remove references to Thor Refactor code to remove references to Thor. Instead use P5 as in phase 5 of development cycle since it is applicable to boards other than Thor as well. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2020-12-01 19:15:23 +00:00
if (BNXT_CHIP_P5(bp))
net/bnxt: support PTP for Thor On Thor, direct access to PTP registers (via GRC) is not supported. Driver must use HWRM to access the timestamp information. Vectorized Rx/Tx cannot be enabled if RTE_LIBRTE_IEEE1588=y. Remove the PTP flags handling code from the vector Rx path. Add support to read tx timestamp value and the time from the timesync clock. On Thor, Rx timestamps are provided directly in the Rx completion records to the driver. Only 32 bits of the timestamp is present in the completion. Driver needs to read the current 48 bit free running timer using the HWRM_PORT_TS_QUERY command and combine the upper 16 bits from the HWRM response with the lower 32 bits in the Rx completion to produce the 48 bit timestamp for the Rx packet. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 01:23:35 +00:00
rc = bnxt_hwrm_port_ts_query(bp, BNXT_PTP_FLAGS_CURRENT_TIME,
&systime_cycles);
else
systime_cycles = bnxt_cc_read(bp);
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
ns = rte_timecounter_update(&ptp->tc, systime_cycles);
*ts = rte_ns_to_timespec(ns);
net/bnxt: support PTP for Thor On Thor, direct access to PTP registers (via GRC) is not supported. Driver must use HWRM to access the timestamp information. Vectorized Rx/Tx cannot be enabled if RTE_LIBRTE_IEEE1588=y. Remove the PTP flags handling code from the vector Rx path. Add support to read tx timestamp value and the time from the timesync clock. On Thor, Rx timestamps are provided directly in the Rx completion records to the driver. Only 32 bits of the timestamp is present in the completion. Driver needs to read the current 48 bit free running timer using the HWRM_PORT_TS_QUERY command and combine the upper 16 bits from the HWRM response with the lower 32 bits in the Rx completion to produce the 48 bit timestamp for the Rx packet. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 01:23:35 +00:00
return rc;
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
}
static int
bnxt_timesync_enable(struct rte_eth_dev *dev)
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = dev->data->dev_private;
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
struct bnxt_ptp_cfg *ptp = bp->ptp_cfg;
uint32_t shift = 0;
net/bnxt: support PTP for Thor On Thor, direct access to PTP registers (via GRC) is not supported. Driver must use HWRM to access the timestamp information. Vectorized Rx/Tx cannot be enabled if RTE_LIBRTE_IEEE1588=y. Remove the PTP flags handling code from the vector Rx path. Add support to read tx timestamp value and the time from the timesync clock. On Thor, Rx timestamps are provided directly in the Rx completion records to the driver. Only 32 bits of the timestamp is present in the completion. Driver needs to read the current 48 bit free running timer using the HWRM_PORT_TS_QUERY command and combine the upper 16 bits from the HWRM response with the lower 32 bits in the Rx completion to produce the 48 bit timestamp for the Rx packet. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 01:23:35 +00:00
int rc;
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
if (!ptp)
net/bnxt: fix timesync when PTP is not supported Fixed to return error when PTP support is not supported on the port. Also, removed an unnecessary check inside bnxt_get_rx_ts(). Fixes: b11cceb83a34 ("net/bnxt: support timesync") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2021-03-18 09:35:23 +00:00
return -ENOTSUP;
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
ptp->rx_filter = 1;
ptp->tx_tstamp_en = 1;
ptp->rxctl = BNXT_PTP_MSG_EVENTS;
net/bnxt: support PTP for Thor On Thor, direct access to PTP registers (via GRC) is not supported. Driver must use HWRM to access the timestamp information. Vectorized Rx/Tx cannot be enabled if RTE_LIBRTE_IEEE1588=y. Remove the PTP flags handling code from the vector Rx path. Add support to read tx timestamp value and the time from the timesync clock. On Thor, Rx timestamps are provided directly in the Rx completion records to the driver. Only 32 bits of the timestamp is present in the completion. Driver needs to read the current 48 bit free running timer using the HWRM_PORT_TS_QUERY command and combine the upper 16 bits from the HWRM response with the lower 32 bits in the Rx completion to produce the 48 bit timestamp for the Rx packet. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 01:23:35 +00:00
rc = bnxt_hwrm_ptp_cfg(bp);
if (rc)
return rc;
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
memset(&ptp->tc, 0, sizeof(struct rte_timecounter));
memset(&ptp->rx_tstamp_tc, 0, sizeof(struct rte_timecounter));
memset(&ptp->tx_tstamp_tc, 0, sizeof(struct rte_timecounter));
ptp->tc.cc_mask = BNXT_CYCLECOUNTER_MASK;
ptp->tc.cc_shift = shift;
ptp->tc.nsec_mask = (1ULL << shift) - 1;
ptp->rx_tstamp_tc.cc_mask = BNXT_CYCLECOUNTER_MASK;
ptp->rx_tstamp_tc.cc_shift = shift;
ptp->rx_tstamp_tc.nsec_mask = (1ULL << shift) - 1;
ptp->tx_tstamp_tc.cc_mask = BNXT_CYCLECOUNTER_MASK;
ptp->tx_tstamp_tc.cc_shift = shift;
ptp->tx_tstamp_tc.nsec_mask = (1ULL << shift) - 1;
net/bnxt: remove references to Thor Refactor code to remove references to Thor. Instead use P5 as in phase 5 of development cycle since it is applicable to boards other than Thor as well. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2020-12-01 19:15:23 +00:00
if (!BNXT_CHIP_P5(bp))
net/bnxt: support PTP for Thor On Thor, direct access to PTP registers (via GRC) is not supported. Driver must use HWRM to access the timestamp information. Vectorized Rx/Tx cannot be enabled if RTE_LIBRTE_IEEE1588=y. Remove the PTP flags handling code from the vector Rx path. Add support to read tx timestamp value and the time from the timesync clock. On Thor, Rx timestamps are provided directly in the Rx completion records to the driver. Only 32 bits of the timestamp is present in the completion. Driver needs to read the current 48 bit free running timer using the HWRM_PORT_TS_QUERY command and combine the upper 16 bits from the HWRM response with the lower 32 bits in the Rx completion to produce the 48 bit timestamp for the Rx packet. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 01:23:35 +00:00
bnxt_map_ptp_regs(bp);
net/bnxt: fix PTP support for Thor On Thor, Rx timestamp is present in the Rx completion record. Only 32 bits of the timestamp is present in the completion. The driver needs to periodically poll the current 48 bit free running timer using the HWRM_PORT_TS_QUERY command. It can combine the upper 16 bits from the HWRM response with the lower 32 bits in the Rx completion to produce the 48 bit timestamp for the Rx packet. This patch adds an alarm thread to periodically poll the current 48 bit free running timer using the HWRM_PORT_TS_QUERY command. This avoids issuing the hwrm command from the Rx handler. This patch also handles the timer roll over condition. Fixes: 6cbd89f9f3d8 ("net/bnxt: support PTP for Thor") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-02-24 15:55:53 +00:00
else
rc = bnxt_ptp_start(bp);
net/bnxt: support PTP for Thor On Thor, direct access to PTP registers (via GRC) is not supported. Driver must use HWRM to access the timestamp information. Vectorized Rx/Tx cannot be enabled if RTE_LIBRTE_IEEE1588=y. Remove the PTP flags handling code from the vector Rx path. Add support to read tx timestamp value and the time from the timesync clock. On Thor, Rx timestamps are provided directly in the Rx completion records to the driver. Only 32 bits of the timestamp is present in the completion. Driver needs to read the current 48 bit free running timer using the HWRM_PORT_TS_QUERY command and combine the upper 16 bits from the HWRM response with the lower 32 bits in the Rx completion to produce the 48 bit timestamp for the Rx packet. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 01:23:35 +00:00
net/bnxt: fix PTP support for Thor On Thor, Rx timestamp is present in the Rx completion record. Only 32 bits of the timestamp is present in the completion. The driver needs to periodically poll the current 48 bit free running timer using the HWRM_PORT_TS_QUERY command. It can combine the upper 16 bits from the HWRM response with the lower 32 bits in the Rx completion to produce the 48 bit timestamp for the Rx packet. This patch adds an alarm thread to periodically poll the current 48 bit free running timer using the HWRM_PORT_TS_QUERY command. This avoids issuing the hwrm command from the Rx handler. This patch also handles the timer roll over condition. Fixes: 6cbd89f9f3d8 ("net/bnxt: support PTP for Thor") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-02-24 15:55:53 +00:00
return rc;
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
}
static int
bnxt_timesync_disable(struct rte_eth_dev *dev)
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = dev->data->dev_private;
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
struct bnxt_ptp_cfg *ptp = bp->ptp_cfg;
if (!ptp)
net/bnxt: fix timesync when PTP is not supported Fixed to return error when PTP support is not supported on the port. Also, removed an unnecessary check inside bnxt_get_rx_ts(). Fixes: b11cceb83a34 ("net/bnxt: support timesync") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2021-03-18 09:35:23 +00:00
return -ENOTSUP;
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
ptp->rx_filter = 0;
ptp->tx_tstamp_en = 0;
ptp->rxctl = 0;
bnxt_hwrm_ptp_cfg(bp);
net/bnxt: remove references to Thor Refactor code to remove references to Thor. Instead use P5 as in phase 5 of development cycle since it is applicable to boards other than Thor as well. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2020-12-01 19:15:23 +00:00
if (!BNXT_CHIP_P5(bp))
net/bnxt: support PTP for Thor On Thor, direct access to PTP registers (via GRC) is not supported. Driver must use HWRM to access the timestamp information. Vectorized Rx/Tx cannot be enabled if RTE_LIBRTE_IEEE1588=y. Remove the PTP flags handling code from the vector Rx path. Add support to read tx timestamp value and the time from the timesync clock. On Thor, Rx timestamps are provided directly in the Rx completion records to the driver. Only 32 bits of the timestamp is present in the completion. Driver needs to read the current 48 bit free running timer using the HWRM_PORT_TS_QUERY command and combine the upper 16 bits from the HWRM response with the lower 32 bits in the Rx completion to produce the 48 bit timestamp for the Rx packet. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 01:23:35 +00:00
bnxt_unmap_ptp_regs(bp);
net/bnxt: fix PTP support for Thor On Thor, Rx timestamp is present in the Rx completion record. Only 32 bits of the timestamp is present in the completion. The driver needs to periodically poll the current 48 bit free running timer using the HWRM_PORT_TS_QUERY command. It can combine the upper 16 bits from the HWRM response with the lower 32 bits in the Rx completion to produce the 48 bit timestamp for the Rx packet. This patch adds an alarm thread to periodically poll the current 48 bit free running timer using the HWRM_PORT_TS_QUERY command. This avoids issuing the hwrm command from the Rx handler. This patch also handles the timer roll over condition. Fixes: 6cbd89f9f3d8 ("net/bnxt: support PTP for Thor") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-02-24 15:55:53 +00:00
else
bnxt_ptp_stop(bp);
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
return 0;
}
static int
bnxt_timesync_read_rx_timestamp(struct rte_eth_dev *dev,
struct timespec *timestamp,
uint32_t flags __rte_unused)
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = dev->data->dev_private;
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
struct bnxt_ptp_cfg *ptp = bp->ptp_cfg;
uint64_t rx_tstamp_cycles = 0;
uint64_t ns;
if (!ptp)
net/bnxt: fix timesync when PTP is not supported Fixed to return error when PTP support is not supported on the port. Also, removed an unnecessary check inside bnxt_get_rx_ts(). Fixes: b11cceb83a34 ("net/bnxt: support timesync") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2021-03-18 09:35:23 +00:00
return -ENOTSUP;
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
net/bnxt: remove references to Thor Refactor code to remove references to Thor. Instead use P5 as in phase 5 of development cycle since it is applicable to boards other than Thor as well. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2020-12-01 19:15:23 +00:00
if (BNXT_CHIP_P5(bp))
net/bnxt: support PTP for Thor On Thor, direct access to PTP registers (via GRC) is not supported. Driver must use HWRM to access the timestamp information. Vectorized Rx/Tx cannot be enabled if RTE_LIBRTE_IEEE1588=y. Remove the PTP flags handling code from the vector Rx path. Add support to read tx timestamp value and the time from the timesync clock. On Thor, Rx timestamps are provided directly in the Rx completion records to the driver. Only 32 bits of the timestamp is present in the completion. Driver needs to read the current 48 bit free running timer using the HWRM_PORT_TS_QUERY command and combine the upper 16 bits from the HWRM response with the lower 32 bits in the Rx completion to produce the 48 bit timestamp for the Rx packet. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 01:23:35 +00:00
rx_tstamp_cycles = ptp->rx_timestamp;
else
bnxt_get_rx_ts(bp, &rx_tstamp_cycles);
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
ns = rte_timecounter_update(&ptp->rx_tstamp_tc, rx_tstamp_cycles);
*timestamp = rte_ns_to_timespec(ns);
return 0;
}
static int
bnxt_timesync_read_tx_timestamp(struct rte_eth_dev *dev,
struct timespec *timestamp)
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = dev->data->dev_private;
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
struct bnxt_ptp_cfg *ptp = bp->ptp_cfg;
uint64_t tx_tstamp_cycles = 0;
uint64_t ns;
net/bnxt: support PTP for Thor On Thor, direct access to PTP registers (via GRC) is not supported. Driver must use HWRM to access the timestamp information. Vectorized Rx/Tx cannot be enabled if RTE_LIBRTE_IEEE1588=y. Remove the PTP flags handling code from the vector Rx path. Add support to read tx timestamp value and the time from the timesync clock. On Thor, Rx timestamps are provided directly in the Rx completion records to the driver. Only 32 bits of the timestamp is present in the completion. Driver needs to read the current 48 bit free running timer using the HWRM_PORT_TS_QUERY command and combine the upper 16 bits from the HWRM response with the lower 32 bits in the Rx completion to produce the 48 bit timestamp for the Rx packet. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 01:23:35 +00:00
int rc = 0;
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
if (!ptp)
net/bnxt: fix timesync when PTP is not supported Fixed to return error when PTP support is not supported on the port. Also, removed an unnecessary check inside bnxt_get_rx_ts(). Fixes: b11cceb83a34 ("net/bnxt: support timesync") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2021-03-18 09:35:23 +00:00
return -ENOTSUP;
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
net/bnxt: remove references to Thor Refactor code to remove references to Thor. Instead use P5 as in phase 5 of development cycle since it is applicable to boards other than Thor as well. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2020-12-01 19:15:23 +00:00
if (BNXT_CHIP_P5(bp))
net/bnxt: support PTP for Thor On Thor, direct access to PTP registers (via GRC) is not supported. Driver must use HWRM to access the timestamp information. Vectorized Rx/Tx cannot be enabled if RTE_LIBRTE_IEEE1588=y. Remove the PTP flags handling code from the vector Rx path. Add support to read tx timestamp value and the time from the timesync clock. On Thor, Rx timestamps are provided directly in the Rx completion records to the driver. Only 32 bits of the timestamp is present in the completion. Driver needs to read the current 48 bit free running timer using the HWRM_PORT_TS_QUERY command and combine the upper 16 bits from the HWRM response with the lower 32 bits in the Rx completion to produce the 48 bit timestamp for the Rx packet. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 01:23:35 +00:00
rc = bnxt_hwrm_port_ts_query(bp, BNXT_PTP_FLAGS_PATH_TX,
&tx_tstamp_cycles);
else
rc = bnxt_get_tx_ts(bp, &tx_tstamp_cycles);
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
ns = rte_timecounter_update(&ptp->tx_tstamp_tc, tx_tstamp_cycles);
*timestamp = rte_ns_to_timespec(ns);
net/bnxt: support PTP for Thor On Thor, direct access to PTP registers (via GRC) is not supported. Driver must use HWRM to access the timestamp information. Vectorized Rx/Tx cannot be enabled if RTE_LIBRTE_IEEE1588=y. Remove the PTP flags handling code from the vector Rx path. Add support to read tx timestamp value and the time from the timesync clock. On Thor, Rx timestamps are provided directly in the Rx completion records to the driver. Only 32 bits of the timestamp is present in the completion. Driver needs to read the current 48 bit free running timer using the HWRM_PORT_TS_QUERY command and combine the upper 16 bits from the HWRM response with the lower 32 bits in the Rx completion to produce the 48 bit timestamp for the Rx packet. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 01:23:35 +00:00
return rc;
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
}
static int
bnxt_timesync_adjust_time(struct rte_eth_dev *dev, int64_t delta)
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = dev->data->dev_private;
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
struct bnxt_ptp_cfg *ptp = bp->ptp_cfg;
if (!ptp)
net/bnxt: fix timesync when PTP is not supported Fixed to return error when PTP support is not supported on the port. Also, removed an unnecessary check inside bnxt_get_rx_ts(). Fixes: b11cceb83a34 ("net/bnxt: support timesync") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2021-03-18 09:35:23 +00:00
return -ENOTSUP;
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
ptp->tc.nsec += delta;
net/bnxt: fix Rx and Tx timestamps timesync adjust and write_time APIs needed to account for the Rx and Tx timestamp counters as well. Fix it since it was not done earlier. Fixes: b11cceb83a34 ("net/bnxt: support timesync") Cc: stable@dpdk.org Reviewed-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
2021-03-12 05:21:09 +00:00
ptp->tx_tstamp_tc.nsec += delta;
ptp->rx_tstamp_tc.nsec += delta;
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
return 0;
}
net/bnxt: support get/set EEPROM Add support for get/set_eeprom, get_eeprom_length dev_ops. Defined the structures required to get/set the eeprom length/data in hsi_struct_defs hdr file along with implementation. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:44 +00:00
static int
bnxt_get_eeprom_length_op(struct rte_eth_dev *dev)
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = dev->data->dev_private;
net/bnxt: support get/set EEPROM Add support for get/set_eeprom, get_eeprom_length dev_ops. Defined the structures required to get/set the eeprom length/data in hsi_struct_defs hdr file along with implementation. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:44 +00:00
int rc;
uint32_t dir_entries;
uint32_t entry_length;
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
rc = is_bnxt_in_error(bp);
if (rc)
return rc;
net/bnxt: use macro for PCI log format Fixes: 19e6af01bb36 ("net/bnxt: support get/set EEPROM") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-01-16 13:04:53 +00:00
PMD_DRV_LOG(INFO, PCI_PRI_FMT "\n",
bp->pdev->addr.domain, bp->pdev->addr.bus,
bp->pdev->addr.devid, bp->pdev->addr.function);
net/bnxt: support get/set EEPROM Add support for get/set_eeprom, get_eeprom_length dev_ops. Defined the structures required to get/set the eeprom length/data in hsi_struct_defs hdr file along with implementation. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:44 +00:00
rc = bnxt_hwrm_nvm_get_dir_info(bp, &dir_entries, &entry_length);
if (rc != 0)
return rc;
return dir_entries * entry_length;
}
static int
bnxt_get_eeprom_op(struct rte_eth_dev *dev,
struct rte_dev_eeprom_info *in_eeprom)
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = dev->data->dev_private;
net/bnxt: support get/set EEPROM Add support for get/set_eeprom, get_eeprom_length dev_ops. Defined the structures required to get/set the eeprom length/data in hsi_struct_defs hdr file along with implementation. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:44 +00:00
uint32_t index;
uint32_t offset;
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
int rc;
rc = is_bnxt_in_error(bp);
if (rc)
return rc;
net/bnxt: support get/set EEPROM Add support for get/set_eeprom, get_eeprom_length dev_ops. Defined the structures required to get/set the eeprom length/data in hsi_struct_defs hdr file along with implementation. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:44 +00:00
net/bnxt: use macro for PCI log format Fixes: 19e6af01bb36 ("net/bnxt: support get/set EEPROM") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-01-16 13:04:53 +00:00
PMD_DRV_LOG(INFO, PCI_PRI_FMT " in_eeprom->offset = %d len = %d\n",
bp->pdev->addr.domain, bp->pdev->addr.bus,
bp->pdev->addr.devid, bp->pdev->addr.function,
in_eeprom->offset, in_eeprom->length);
net/bnxt: support get/set EEPROM Add support for get/set_eeprom, get_eeprom_length dev_ops. Defined the structures required to get/set the eeprom length/data in hsi_struct_defs hdr file along with implementation. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:44 +00:00
if (in_eeprom->offset == 0) /* special offset value to get directory */
return bnxt_get_nvram_directory(bp, in_eeprom->length,
in_eeprom->data);
index = in_eeprom->offset >> 24;
offset = in_eeprom->offset & 0xffffff;
if (index != 0)
return bnxt_hwrm_get_nvram_item(bp, index - 1, offset,
in_eeprom->length, in_eeprom->data);
return 0;
}
static bool bnxt_dir_type_is_ape_bin_format(uint16_t dir_type)
{
switch (dir_type) {
case BNX_DIR_TYPE_CHIMP_PATCH:
case BNX_DIR_TYPE_BOOTCODE:
case BNX_DIR_TYPE_BOOTCODE_2:
case BNX_DIR_TYPE_APE_FW:
case BNX_DIR_TYPE_APE_PATCH:
case BNX_DIR_TYPE_KONG_FW:
case BNX_DIR_TYPE_KONG_PATCH:
case BNX_DIR_TYPE_BONO_FW:
case BNX_DIR_TYPE_BONO_PATCH:
net/bnxt: add fallthrough comment The FALLTHROUGH comment for cascading switch cases was missing in a few places. This patch adds that. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-05-22 18:13:41 +00:00
/* FALLTHROUGH */
net/bnxt: support get/set EEPROM Add support for get/set_eeprom, get_eeprom_length dev_ops. Defined the structures required to get/set the eeprom length/data in hsi_struct_defs hdr file along with implementation. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:44 +00:00
return true;
}
return false;
}
static bool bnxt_dir_type_is_other_exec_format(uint16_t dir_type)
{
switch (dir_type) {
case BNX_DIR_TYPE_AVS:
case BNX_DIR_TYPE_EXP_ROM_MBA:
case BNX_DIR_TYPE_PCIE:
case BNX_DIR_TYPE_TSCF_UCODE:
case BNX_DIR_TYPE_EXT_PHY:
case BNX_DIR_TYPE_CCM:
case BNX_DIR_TYPE_ISCSI_BOOT:
case BNX_DIR_TYPE_ISCSI_BOOT_IPV6:
case BNX_DIR_TYPE_ISCSI_BOOT_IPV4N6:
net/bnxt: add fallthrough comment The FALLTHROUGH comment for cascading switch cases was missing in a few places. This patch adds that. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-05-22 18:13:41 +00:00
/* FALLTHROUGH */
net/bnxt: support get/set EEPROM Add support for get/set_eeprom, get_eeprom_length dev_ops. Defined the structures required to get/set the eeprom length/data in hsi_struct_defs hdr file along with implementation. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:44 +00:00
return true;
}
return false;
}
static bool bnxt_dir_type_is_executable(uint16_t dir_type)
{
return bnxt_dir_type_is_ape_bin_format(dir_type) ||
bnxt_dir_type_is_other_exec_format(dir_type);
}
static int
bnxt_set_eeprom_op(struct rte_eth_dev *dev,
struct rte_dev_eeprom_info *in_eeprom)
{
net/bnxt: remove unnecessary cast The device private pointer (dev_private) is of type void * therefore no cast is necessary in C. Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2019-05-29 19:14:53 +00:00
struct bnxt *bp = dev->data->dev_private;
net/bnxt: support get/set EEPROM Add support for get/set_eeprom, get_eeprom_length dev_ops. Defined the structures required to get/set the eeprom length/data in hsi_struct_defs hdr file along with implementation. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:44 +00:00
uint8_t index, dir_op;
uint16_t type, ext, ordinal, attr;
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
int rc;
rc = is_bnxt_in_error(bp);
if (rc)
return rc;
net/bnxt: support get/set EEPROM Add support for get/set_eeprom, get_eeprom_length dev_ops. Defined the structures required to get/set the eeprom length/data in hsi_struct_defs hdr file along with implementation. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:44 +00:00
net/bnxt: use macro for PCI log format Fixes: 19e6af01bb36 ("net/bnxt: support get/set EEPROM") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-01-16 13:04:53 +00:00
PMD_DRV_LOG(INFO, PCI_PRI_FMT " in_eeprom->offset = %d len = %d\n",
bp->pdev->addr.domain, bp->pdev->addr.bus,
bp->pdev->addr.devid, bp->pdev->addr.function,
in_eeprom->offset, in_eeprom->length);
net/bnxt: support get/set EEPROM Add support for get/set_eeprom, get_eeprom_length dev_ops. Defined the structures required to get/set the eeprom length/data in hsi_struct_defs hdr file along with implementation. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:44 +00:00
if (!BNXT_PF(bp)) {
net/bnxt: use dynamic log type This patch implements driver specific log type doing away with usage of RTE_LOG() for logging. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:55 +00:00
PMD_DRV_LOG(ERR, "NVM write not supported from a VF\n");
net/bnxt: support get/set EEPROM Add support for get/set_eeprom, get_eeprom_length dev_ops. Defined the structures required to get/set the eeprom length/data in hsi_struct_defs hdr file along with implementation. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:44 +00:00
return -EINVAL;
}
type = in_eeprom->magic >> 16;
if (type == 0xffff) { /* special value for directory operations */
index = in_eeprom->magic & 0xff;
dir_op = in_eeprom->magic >> 8;
if (index == 0)
return -EINVAL;
switch (dir_op) {
case 0x0e: /* erase */
if (in_eeprom->offset != ~in_eeprom->magic)
return -EINVAL;
return bnxt_hwrm_erase_nvram_directory(bp, index - 1);
default:
return -EINVAL;
}
}
/* Create or re-write an NVM item: */
if (bnxt_dir_type_is_executable(type) == true)
return -EOPNOTSUPP;
ext = in_eeprom->magic & 0xffff;
ordinal = in_eeprom->offset >> 16;
attr = in_eeprom->offset & 0xffff;
return bnxt_hwrm_flash_nvram(bp, type, ordinal, ext, attr,
in_eeprom->data, in_eeprom->length);
}
net/bnxt: add driver for Broadcom NetXtreme-C devices This patch adds the initial skeleton for bnxt driver along with the nic guide, and ties the driver into the build system. At this point, the driver simply fails init. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com> [Release Note Addition] Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2016-06-15 21:23:01 +00:00
/*
* Initialization
*/
net/bnxt: make ethdev functions struct const Reported-by: Stephen Hemminger <stephen@networkplumber.org> Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
2017-01-12 22:19:37 +00:00
static const struct eth_dev_ops bnxt_dev_ops = {
net/bnxt: get device infos Gets device info from the bp structure filled in the init() function. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:04 +00:00
.dev_infos_get = bnxt_dev_info_get_op,
net/bnxt: add hardware resource manager init code Start adding support to use the HWRM API. Hardware Resource Manager or HWRM in short, is a set of API provided by the firmware running in the ASIC to manage the various resources. Initial commit just performs necessary HWRM queries for init, then fails as before. Now that struct bnxt is non-zero size, we can set dev_private_size correctly. The used HWRM calls so far: bnxt_hwrm_func_qcaps: This command returns capabilities of a function. bnxt_hwrm_ver_get: This function is called by a driver to determine the HWRM interface version supported by the HWRM firmware, the version of HWRM firmware implementation, the name of HWRM firmware, the versions of other embedded firmwares, and the names of other embedded firmwares, etc. Gets the firmware version and interface specifications. Returns an error if the firmware on the device is not supported by the driver and ensures the response space is large enough for the largest possible response. bnxt_hwrm_queue_qportcfg: This function is called by a driver to query queue configuration of a port. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:02 +00:00
.dev_close = bnxt_dev_close_op,
net/bnxt: add device configure operation The dev_configure_op function calls bnxt_set_hwrm_link_config() to setup the PHY. This calls the new bnxt_parse_eth_link_*() functions to translate from the DPDK macro values to those used by HWRM calls, then calls bnxt_hwrm_port_phy_cfg() to issue the HWRM call. New HWRM calls: bnxt_hwrm_port_phy_cfg: This command configures the PHY device for the port. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:05 +00:00
.dev_configure = bnxt_dev_configure_op,
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
.dev_start = bnxt_dev_start_op,
.dev_stop = bnxt_dev_stop_op,
net/bnxt: add link state operations Adds dev_ops to set link Up or Down as appropriate. Uses the bnxt_set_hwrm_link_config() API added previously. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:35 +00:00
.dev_set_link_up = bnxt_dev_set_link_up_op,
.dev_set_link_down = bnxt_dev_set_link_down_op,
net/bnxt: add statistics Add the bnxt_stats_get_op() and bnxt_stats_reset_op() dev_ops to get and reset statistics. It also brings in the associated HWRM calls to handle the requests appropriately. We also have the bnxt_free_stats() function which will be used in the follow on patches to free the memory allocated by the driver for statistics. New HWRM calls: bnxt_hwrm_stat_clear: This command clears statistics of a context Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:12 +00:00
.stats_get = bnxt_stats_get_op,
.stats_reset = bnxt_stats_reset_op,
net/bnxt: add Rx queue create/destroy In this patch we are adding the bnxt_rx_queue_setup_op() and bnxt_rx_queue_release_op() functions. These will be tied to the rx_queue_setup and rx_queue_release dev_ops in a subsequent patch. In these functions we allocate/free memory for the RX queues. This still requires support to create a RX ring in the ASIC which will be completed in a future commit. Each Rx queue created via the rx_queue_setup dev_op will have an associated Rx ring in the hardware. The Rx logic in the hardware picks a Rx ring for each Rx frame received by the hardware depending on the properties like RSS, MAC and VLAN settings configured in the hardware. These packets in the end arrive on the Rx queue corresponding to the Rx ring in the hardware. We are also adding some functions like bnxt_mq_rx_configure() bnxt_free_rx_mbufs() and bnxt_free_rxq_stats() which will be used in subsequent patches. We are also adding hwrm_vnic_rss_cfg_* structures, which will be used in subsequent patches to enable RSS configuration. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:11 +00:00
.rx_queue_setup = bnxt_rx_queue_setup_op,
.rx_queue_release = bnxt_rx_queue_release_op,
net/bnxt: add Tx queue create/destroy In this patch we are adding the bnxt_tx_queue_setup_op() and bnxt_tx_queue_release_op() functions. These will be tied to the tx_queue_setup and tx_queue_release dev_ops in a subsequent patch. In these functions we allocate/free memory for the TX queues. This still requires support to create a TX ring in the ASIC which will be completed in a future commit. Each Tx queue created via the tx_queue_setup dev_op will have an associated Tx ring in the hardware. A Tx request coming on the Tx queue gets sent to the corresponding Tx ring in the ASIC for subsequent transmission. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:10 +00:00
.tx_queue_setup = bnxt_tx_queue_setup_op,
.tx_queue_release = bnxt_tx_queue_release_op,
net/bnxt: support enable/disable interrupt Implement Rx Queue interrupt enable/disable functions Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:45 +00:00
.rx_queue_intr_enable = bnxt_rx_queue_intr_enable_op,
.rx_queue_intr_disable = bnxt_rx_queue_intr_disable_op,
net/bnxt: add RSS redirection table operations Add code to Update/query reta dev_ops Uses the bnxt_hwrm_vnic_rss_cfg() HWRM API added earlier. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:36 +00:00
.reta_update = bnxt_reta_update_op,
.reta_query = bnxt_reta_query_op,
net/bnxt: add RSS hash configuration Add rss_hash_update and rss_hash_conf_get dev_ops Uses the bnxt_hwrm_vnic_rss_cfg() HWRM API added in the previous patch. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:37 +00:00
.rss_hash_update = bnxt_rss_hash_update_op,
.rss_hash_conf_get = bnxt_rss_hash_conf_get_op,
net/bnxt: add start/stop/link update operations This patch adds code to add the start, stop and link update dev_ops. Also adds wrapper functions like bnxt_init_chip(), bnxt_init_nic(), bnxt_alloc_mem(), bnxt_free_mem() The BNXT driver will now minimally pass traffic with testpmd. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:30 +00:00
.link_update = bnxt_link_update_op,
net/bnxt: add promiscuous mode This patch adds the promiscuous mode enable and disable dev_ops. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added in an earlier commit. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:31 +00:00
.promiscuous_enable = bnxt_promiscuous_enable_op,
.promiscuous_disable = bnxt_promiscuous_disable_op,
net/bnxt: add multicast This patch adds dev_ops to enable/disable multicast traffic. Uses the bnxt_hwrm_cfa_l2_set_rx_mask() API added earlier. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:32 +00:00
.allmulticast_enable = bnxt_allmulticast_enable_op,
.allmulticast_disable = bnxt_allmulticast_disable_op,
net/bnxt: add MAC address This patch adds dev_ops to Add/Remove MAC addresses. These use the bnxt_hwrm_set_filter() defined previously Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:34 +00:00
.mac_addr_add = bnxt_mac_addr_add_op,
.mac_addr_remove = bnxt_mac_addr_remove_op,
net/bnxt: add flow control operations Add flow_ctrl_get and flow_ctrl_set dev_ops. Uses the bnxt_set_hwrm_link_config() HWRM API added in earlier patch. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:38 +00:00
.flow_ctrl_get = bnxt_flow_ctrl_get_op,
.flow_ctrl_set = bnxt_flow_ctrl_set_op,
net/bnxt: support tunneling Add support for udp_tunnel_port_add/del dev_ops to configure a UDP port for VXLAN and Geneve Tunnel protocols. The HWRM supports only one global destination port for a tunnel type, use a reference counter to keep track of its usage. Cache the configured VXLAN/Geneve ports and use that value to check if the right UDP port is being freed up. Skip calling bnxt_hwrm_tunnel_dst_port_alloc if the same UDP port is being programmed. Skip calling bnxt_hwrm_tunnel_dst_port_free if no UDP port has been configured. Also update tx offload capabilities Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:03 +00:00
.udp_tunnel_port_add = bnxt_udp_tunnel_port_add_op,
.udp_tunnel_port_del = bnxt_udp_tunnel_port_del_op,
net/bnxt: support VLAN filter and strip This patch adds VLAN strip and offload callbacks. To add a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists: if VLAN matches vlan_id VLAN filter already exists, just skip and continue else add a new MAC+VLAN filter else Remove the old MAC only filter Add a new MAC+VLAN filter To remove a VLAN filter: For each VNIC and each associated filter(s) if VLAN exists && VLAN matches vlan_id remove the MAC+VLAN filter add a new MAC only filter else VLAN filter doesn't exist, just skip and continue Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:05 +00:00
.vlan_filter_set = bnxt_vlan_filter_set_op,
.vlan_offload_set = bnxt_vlan_offload_set_op,
net/bnxt: support for QinQ insertion and stripping Driver will accelerate only outer/S-VLAN insertion by turning on the appropriate bits in the Tx Buffer Descriptor when the packet arrives for transmission. The TPID to be used for this S-VLAN is conveyed by the vlan_tpid_set dev_op which will terminate in the driver. In the Rx path, driver will continue providing the stripped vlan tag in the mbuf's vlan tci field. This would be the outermost vlan tag in a double-tagged packet or the vlan tag for a single vlan tagged pkt. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 17:17:44 +00:00
.vlan_tpid_set = bnxt_vlan_tpid_set_op,
net/bnxt: support VLAN pvid This patch adds code to support vlan_pvid_set dev_op Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:12 +00:00
.vlan_pvid_set = bnxt_vlan_pvid_set_op,
net/bnxt: support set MTU This patch adds support to modify MTU using the set_mtu dev_op. To support frames > 2k, the PMD creates an aggregator ring. When a frame greater than 2k is received, it is fragmented and the resulting fragments are DMA'ed to the aggregator ring. Now the driver can support jumbo frames upto 9500 bytes. Signed-off-by: Steeven Li <steeven.li@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:09 +00:00
.mtu_set = bnxt_mtu_set_op,
net/bnxt: support multicast filter and set MAC addr This patch adds support for set_mc_addr_list and mac_addr_set dev_ops Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:06 +00:00
.mac_addr_set = bnxt_set_default_mac_addr_op,
net/bnxt: support xstats get/reset This patch adds support to get and reset xstats dev_ops dev_ops added: xstats_get, xstats_get_name, xstats_reset HWRM commands added: hwrm_port_qstats, hwrm_port_clr_stats Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:04 +00:00
.xstats_get = bnxt_dev_xstats_get_op,
.xstats_get_names = bnxt_dev_xstats_get_names_op,
.xstats_reset = bnxt_dev_xstats_reset_op,
net/bnxt: support get FW version This patch adds support for fw_version_get dev_op Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:08 +00:00
.fw_version_get = bnxt_fw_version_get,
net/bnxt: support multicast filter and set MAC addr This patch adds support for set_mc_addr_list and mac_addr_set dev_ops Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:06 +00:00
.set_mc_addr_list = bnxt_dev_set_mc_addr_list_op,
net/bnxt: support rxq/txq get information Add support for txq_info_get and rxq_info_get dev_ops Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:11 +00:00
.rxq_info_get = bnxt_rxq_info_get_op,
.txq_info_get = bnxt_txq_info_get_op,
net/bnxt: support Rx/Tx burst mode info Retrieve burst mode options according to the selected Rx/Tx burst function name. Update 20.08 release notes with this information. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-07 22:22:24 +00:00
.rx_burst_mode_get = bnxt_rx_burst_mode_get,
.tx_burst_mode_get = bnxt_tx_burst_mode_get,
net/bnxt: support LED on/off This patch adds support for dev_led_on/off dev_ops HWRM calls added: bnxt_hwrm_port_led_qcaps() bnxt_hwrm_port_led_cfg() Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:15 +00:00
.dev_led_on = bnxt_dev_led_on_op,
.dev_led_off = bnxt_dev_led_off_op,
net/bnxt: support Rx/Tx queue start/stop Currently this is implemented entirely in the PMD as there is no explicit support in the HW. Re-program the RSS Table without this queue on stop and add it back to the table on start. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:58 +00:00
.rx_queue_start = bnxt_rx_queue_start,
.rx_queue_stop = bnxt_rx_queue_stop,
.tx_queue_start = bnxt_tx_queue_start,
.tx_queue_stop = bnxt_tx_queue_stop,
ethdev: replace callback getting filter operations Since rte_flow is the only API for filtering operations, the legacy driver interface filter_ctrl was too much complicated for the simple task of getting the struct rte_flow_ops. The filter type RTE_ETH_FILTER_GENERIC and the filter operarion RTE_ETH_FILTER_GET are removed. The new driver callback flow_ops_get replaces filter_ctrl. Signed-off-by: Thomas Monjalon <thomas@monjalon.net> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Acked-by: Haiyue Wang <haiyue.wang@intel.com> Acked-by: Rosen Xu <rosen.xu@intel.com> Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com> Reviewed-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2021-03-21 09:00:00 +00:00
.flow_ops_get = bnxt_flow_ops_get_op,
net/bnxt: support getting ptypes This patch adds support for dev_supported_ptypes_get Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:43 +00:00
.dev_supported_ptypes_get = bnxt_dev_supported_ptypes_get_op,
net/bnxt: support get/set EEPROM Add support for get/set_eeprom, get_eeprom_length dev_ops. Defined the structures required to get/set the eeprom length/data in hsi_struct_defs hdr file along with implementation. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:44 +00:00
.get_eeprom_length = bnxt_get_eeprom_length_op,
.get_eeprom = bnxt_get_eeprom_op,
.set_eeprom = bnxt_set_eeprom_op,
net/bnxt: support timesync Implemented the 'time_sync' related APIs for supporting the PTP protocol. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-12-05 07:26:56 +00:00
.timesync_enable = bnxt_timesync_enable,
.timesync_disable = bnxt_timesync_disable,
.timesync_read_time = bnxt_timesync_read_time,
.timesync_write_time = bnxt_timesync_write_time,
.timesync_adjust_time = bnxt_timesync_adjust_time,
.timesync_read_rx_timestamp = bnxt_timesync_read_rx_timestamp,
.timesync_read_tx_timestamp = bnxt_timesync_read_tx_timestamp,
net/bnxt: add hardware resource manager init code Start adding support to use the HWRM API. Hardware Resource Manager or HWRM in short, is a set of API provided by the firmware running in the ASIC to manage the various resources. Initial commit just performs necessary HWRM queries for init, then fails as before. Now that struct bnxt is non-zero size, we can set dev_private_size correctly. The used HWRM calls so far: bnxt_hwrm_func_qcaps: This command returns capabilities of a function. bnxt_hwrm_ver_get: This function is called by a driver to determine the HWRM interface version supported by the HWRM firmware, the version of HWRM firmware implementation, the name of HWRM firmware, the versions of other embedded firmwares, and the names of other embedded firmwares, etc. Gets the firmware version and interface specifications. Returns an error if the firmware on the device is not supported by the driver and ensures the response space is large enough for the largest possible response. bnxt_hwrm_queue_qportcfg: This function is called by a driver to query queue configuration of a port. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:02 +00:00
};
net/bnxt: support FW reset Added code to perform FW_RESET. When the driver detects error in FW, it has to initiate the recovery by resetting the cores. FW advertise the method to do a core reset, reset register offsets and values to perform reset in response of HWRM_ERROR_RECOVERY_QCFG command. There are 2 ways to recover from the error. 1. Master function issues core resets to recover from error. 2. Master function detects chimp dead condition and notify the Kong processor about the chimp dead case through FW_RESET HWRM command. Kong Processor send an RESET_NOTIFY async event with REASON_CODE_FW_EXCEPTION_FATAL to all the PF’s/VF’s that chimp is dead and it is going to reset the chimp. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 01:23:30 +00:00
static uint32_t bnxt_map_reset_regs(struct bnxt *bp, uint32_t reg)
{
uint32_t offset;
/* Only pre-map the reset GRC registers using window 3 */
rte_write32(reg & 0xfffff000, (uint8_t *)bp->bar0 +
BNXT_GRCPF_REG_WINDOW_BASE_OUT + 8);
offset = BNXT_GRCP_WINDOW_3_BASE + (reg & 0xffc);
return offset;
}
net/bnxt: map status registers for FW health monitoring HWRM_ERROR_RECOVERY_QCFG command returns the FW status registers offset for periodic firmware health check monitoring. Map them to GRC window 2. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:27 +00:00
int bnxt_map_fw_health_status_regs(struct bnxt *bp)
{
struct bnxt_error_recovery_info *info = bp->recovery_info;
uint32_t reg_base = 0xffffffff;
int i;
/* Only pre-map the monitoring GRC registers using window 2 */
for (i = 0; i < BNXT_FW_STATUS_REG_CNT; i++) {
uint32_t reg = info->status_regs[i];
if (BNXT_FW_STATUS_REG_TYPE(reg) != BNXT_FW_STATUS_REG_TYPE_GRC)
continue;
if (reg_base == 0xffffffff)
reg_base = reg & 0xfffff000;
if ((reg & 0xfffff000) != reg_base)
return -ERANGE;
/* Use mask 0xffc as the Lower 2 bits indicates
* address space location
*/
info->mapped_status_regs[i] = BNXT_GRCP_WINDOW_2_BASE +
(reg & 0xffc);
}
if (reg_base == 0xffffffff)
return 0;
rte_write32(reg_base, (uint8_t *)bp->bar0 +
BNXT_GRCPF_REG_WINDOW_BASE_OUT + 4);
return 0;
}
net/bnxt: support FW reset Added code to perform FW_RESET. When the driver detects error in FW, it has to initiate the recovery by resetting the cores. FW advertise the method to do a core reset, reset register offsets and values to perform reset in response of HWRM_ERROR_RECOVERY_QCFG command. There are 2 ways to recover from the error. 1. Master function issues core resets to recover from error. 2. Master function detects chimp dead condition and notify the Kong processor about the chimp dead case through FW_RESET HWRM command. Kong Processor send an RESET_NOTIFY async event with REASON_CODE_FW_EXCEPTION_FATAL to all the PF’s/VF’s that chimp is dead and it is going to reset the chimp. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 01:23:30 +00:00
static void bnxt_write_fw_reset_reg(struct bnxt *bp, uint32_t index)
{
struct bnxt_error_recovery_info *info = bp->recovery_info;
uint32_t delay = info->delay_after_reset[index];
uint32_t val = info->reset_reg_val[index];
uint32_t reg = info->reset_reg[index];
uint32_t type, offset;
net/bnxt: fix PCI write check CID 363716 (#1 of 1): Unchecked return value (CHECKED_RETURN) check_return: Calling rte_pci_write_config without checking return value (as is done elsewhere 46 out of 49 times). Coverity issue: 363716 Fixes: be14720def9c ("net/bnxt: support FW reset") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2021-03-16 05:41:25 +00:00
int ret;
net/bnxt: support FW reset Added code to perform FW_RESET. When the driver detects error in FW, it has to initiate the recovery by resetting the cores. FW advertise the method to do a core reset, reset register offsets and values to perform reset in response of HWRM_ERROR_RECOVERY_QCFG command. There are 2 ways to recover from the error. 1. Master function issues core resets to recover from error. 2. Master function detects chimp dead condition and notify the Kong processor about the chimp dead case through FW_RESET HWRM command. Kong Processor send an RESET_NOTIFY async event with REASON_CODE_FW_EXCEPTION_FATAL to all the PF’s/VF’s that chimp is dead and it is going to reset the chimp. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 01:23:30 +00:00
type = BNXT_FW_STATUS_REG_TYPE(reg);
offset = BNXT_FW_STATUS_REG_OFF(reg);
switch (type) {
case BNXT_FW_STATUS_REG_TYPE_CFG:
net/bnxt: fix PCI write check CID 363716 (#1 of 1): Unchecked return value (CHECKED_RETURN) check_return: Calling rte_pci_write_config without checking return value (as is done elsewhere 46 out of 49 times). Coverity issue: 363716 Fixes: be14720def9c ("net/bnxt: support FW reset") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2021-03-16 05:41:25 +00:00
ret = rte_pci_write_config(bp->pdev, &val, sizeof(val), offset);
if (ret < 0) {
PMD_DRV_LOG(ERR, "Failed to write %#x at PCI offset %#x",
val, offset);
return;
}
net/bnxt: support FW reset Added code to perform FW_RESET. When the driver detects error in FW, it has to initiate the recovery by resetting the cores. FW advertise the method to do a core reset, reset register offsets and values to perform reset in response of HWRM_ERROR_RECOVERY_QCFG command. There are 2 ways to recover from the error. 1. Master function issues core resets to recover from error. 2. Master function detects chimp dead condition and notify the Kong processor about the chimp dead case through FW_RESET HWRM command. Kong Processor send an RESET_NOTIFY async event with REASON_CODE_FW_EXCEPTION_FATAL to all the PF’s/VF’s that chimp is dead and it is going to reset the chimp. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 01:23:30 +00:00
break;
case BNXT_FW_STATUS_REG_TYPE_GRC:
offset = bnxt_map_reset_regs(bp, offset);
rte_write32(val, (uint8_t *)bp->bar0 + offset);
break;
case BNXT_FW_STATUS_REG_TYPE_BAR0:
rte_write32(val, (uint8_t *)bp->bar0 + offset);
break;
}
/* wait on a specific interval of time until core reset is complete */
if (delay)
rte_delay_ms(delay);
}
net/bnxt: handle reset notify async event from FW When the FW upgrade is initiated the current instance of FW issues a HWRM_ASYNC_EVENT_CMPL_EVENT_ID_RESET_NOTIFY async notification to the driver. On receiving this notification, the PMD shall quiesce itself and poll on the HWRM_VER_GET FW command at regular intervals. Once the VER_GET command succeeds, the driver should go through the rediscovery process and re-initialize the device. Also register with FW for the reset notify async event. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:23 +00:00
static void bnxt_dev_cleanup(struct bnxt *bp)
{
net/bnxt: fix link status during device recovery Driver should not send the phy_cfg request to bring link down during reset recovery. If the driver sends the phy_cfg request in recovery process, then FW needs to re-establish the link which in turn increases the recovery time based on PHY type and link partners. Fixes: df6cd7c1f73a ("net/bnxt: handle reset notify async event from FW") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 05:34:16 +00:00
bp->eth_dev->data->dev_link.link_status = 0;
net/bnxt: fix allocation of link info struct Dynamically alloc link info structure. Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 11:05:15 +00:00
bp->link_info->link_up = 0;
net/bnxt: remove a redundant variable Use "dev->data->dev_started" state, instead of local "dev_stopped" to check whether port has been started or not. Fixes: 316e412299fd ("net/bnxt: fix crash when closing") Cc: stable@dpdk.org Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
2020-01-28 07:29:21 +00:00
if (bp->eth_dev->data->dev_started)
net/bnxt: check chip reset in stop and close While the error recovery thread is running, an application can invoke dev_stop or dev_close_op thus triggering a race and unwanted consequences if dev_close is invoked while the recovery is not yet completed. Fix by having another lock to synchronize between the 2 threads and return EGAIN if adapter is in the middle of recovery when dev_stop or dev_close ops are invoked Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-24 09:37:34 +00:00
bnxt_dev_stop(bp->eth_dev);
net/bnxt: handle reset notify async event from FW When the FW upgrade is initiated the current instance of FW issues a HWRM_ASYNC_EVENT_CMPL_EVENT_ID_RESET_NOTIFY async notification to the driver. On receiving this notification, the PMD shall quiesce itself and poll on the HWRM_VER_GET FW command at regular intervals. Once the VER_GET command succeeds, the driver should go through the rediscovery process and re-initialize the device. Also register with FW for the reset notify async event. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:23 +00:00
bnxt_uninit_resources(bp, true);
}
net/bnxt: fix firmware fatal error handling During some fatal firmware error conditions, the PCI config space register 0x2e which normally contains the subsystem ID will become 0xffff. This register will revert back to the normal value after the chip has completed core reset. If we detect this condition, we can poll this config register immediately for the value to revert. Because we use config read cycles to poll this register, there is no possibility of Master Abort if we happen to read it during core reset. This speeds up recovery significantly as we don't have to wait for the conservative min_time before polling to see if the firmware has come out of reset. As soon as this register changes value we can proceed to re-initialize the device. Fixes: df6cd7c1f73a ("net/bnxt: handle reset notify async event from FW") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-02-24 15:55:51 +00:00
static int
bnxt_check_fw_reset_done(struct bnxt *bp)
{
int timeout = bp->fw_reset_max_msecs;
uint16_t val = 0;
int rc;
do {
rc = rte_pci_read_config(bp->pdev, &val, sizeof(val), PCI_SUBSYSTEM_ID_OFFSET);
if (rc < 0) {
PMD_DRV_LOG(ERR, "Failed to read PCI offset 0x%x", PCI_SUBSYSTEM_ID_OFFSET);
return rc;
}
if (val != 0xffff)
break;
rte_delay_ms(1);
} while (timeout--);
if (val == 0xffff) {
PMD_DRV_LOG(ERR, "Firmware reset aborted, PCI config space invalid\n");
return -1;
}
return 0;
}
net/bnxt: restore VLAN filters during reset recovery Older Firmware could have state information such as MAC Filters, VLAN settings etc configured by user. But new Firmware is unaware of this state information and as a result driver should restore these settings during reset recovery. This patch restores the user configured vlan settings prior to hot FW upgrade or FW error. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-12-10 15:05:25 +00:00
static int bnxt_restore_vlan_filters(struct bnxt *bp)
{
struct rte_eth_dev *dev = bp->eth_dev;
struct rte_vlan_filter_conf *vfc;
int vidx, vbit, rc;
uint16_t vlan_id;
for (vlan_id = 1; vlan_id <= RTE_ETHER_MAX_VLAN_ID; vlan_id++) {
vfc = &dev->data->vlan_filter_conf;
vidx = vlan_id / 64;
vbit = vlan_id % 64;
/* Each bit corresponds to a VLAN id */
if (vfc->ids[vidx] & (UINT64_C(1) << vbit)) {
rc = bnxt_add_vlan_filter(bp, vlan_id);
if (rc)
return rc;
}
}
return 0;
}
net/bnxt: restore MAC filters during reset recovery Older Firmware could have state information such as MAC Filters, VLAN settings etc configured by user. But new Firmware is unaware of this state information and as a result driver should restore these settings during reset recovery. This patch restores the user configured mac addresses prior to hot FW upgrade or FW error. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-12-10 15:05:24 +00:00
static int bnxt_restore_mac_filters(struct bnxt *bp)
{
struct rte_eth_dev *dev = bp->eth_dev;
struct rte_eth_dev_info dev_info;
struct rte_ether_addr *addr;
uint64_t pool_mask;
uint32_t pool = 0;
net/bnxt: fix mismatched type comparison in MAC restore dev_info.max_mac_addrs is of type uint32_t. But the counter i is of type uint16_t. This mismatch may cause the loop condition may always be true. Change the loop counter variable to uint32_t. Fixes: b02f1573cd07 ("net/bnxt: restore MAC filters during reset recovery") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
2021-04-30 20:14:10 +00:00
uint32_t i;
net/bnxt: restore MAC filters during reset recovery Older Firmware could have state information such as MAC Filters, VLAN settings etc configured by user. But new Firmware is unaware of this state information and as a result driver should restore these settings during reset recovery. This patch restores the user configured mac addresses prior to hot FW upgrade or FW error. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-12-10 15:05:24 +00:00
int rc;
net/bnxt: fix logical AND in if condition The if condition in bnxt_restore_mac_filters needs to check for the result of logical AND. But it was not doing it resulting in an incorrect check. Fixes: b02f1573cd07 ("net/bnxt: restore MAC filters during reset recovery") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-24 00:04:37 +00:00
if (BNXT_VF(bp) && !BNXT_VF_IS_TRUSTED(bp))
net/bnxt: restore MAC filters during reset recovery Older Firmware could have state information such as MAC Filters, VLAN settings etc configured by user. But new Firmware is unaware of this state information and as a result driver should restore these settings during reset recovery. This patch restores the user configured mac addresses prior to hot FW upgrade or FW error. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-12-10 15:05:24 +00:00
return 0;
rc = bnxt_dev_info_get_op(dev, &dev_info);
if (rc)
return rc;
/* replay MAC address configuration */
for (i = 1; i < dev_info.max_mac_addrs; i++) {
addr = &dev->data->mac_addrs[i];
/* skip zero address */
if (rte_is_zero_ether_addr(addr))
continue;
pool = 0;
pool_mask = dev->data->mac_pool_sel[i];
do {
if (pool_mask & 1ULL) {
rc = bnxt_mac_addr_add_op(dev, addr, i, pool);
if (rc)
return rc;
}
pool_mask >>= 1;
pool++;
} while (pool_mask);
}
return 0;
}
net/bnxt: handle reset notify async event from FW When the FW upgrade is initiated the current instance of FW issues a HWRM_ASYNC_EVENT_CMPL_EVENT_ID_RESET_NOTIFY async notification to the driver. On receiving this notification, the PMD shall quiesce itself and poll on the HWRM_VER_GET FW command at regular intervals. Once the VER_GET command succeeds, the driver should go through the rediscovery process and re-initialize the device. Also register with FW for the reset notify async event. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:23 +00:00
static int bnxt_restore_filters(struct bnxt *bp)
{
struct rte_eth_dev *dev = bp->eth_dev;
int ret = 0;
net/bnxt: fix return code handling in VLAN config return value stored in "ret" but it has been overwritten before use. Coverity issue: 353621 Fixes: 7fe5668d2ea3 ("net/bnxt: support VLAN filter and strip") Fixes: df6cd7c1f73a ("net/bnxt: handle reset notify async event from FW") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-01-28 07:29:23 +00:00
if (dev->data->all_multicast) {
net/bnxt: handle reset notify async event from FW When the FW upgrade is initiated the current instance of FW issues a HWRM_ASYNC_EVENT_CMPL_EVENT_ID_RESET_NOTIFY async notification to the driver. On receiving this notification, the PMD shall quiesce itself and poll on the HWRM_VER_GET FW command at regular intervals. Once the VER_GET command succeeds, the driver should go through the rediscovery process and re-initialize the device. Also register with FW for the reset notify async event. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:23 +00:00
ret = bnxt_allmulticast_enable_op(dev);
net/bnxt: fix return code handling in VLAN config return value stored in "ret" but it has been overwritten before use. Coverity issue: 353621 Fixes: 7fe5668d2ea3 ("net/bnxt: support VLAN filter and strip") Fixes: df6cd7c1f73a ("net/bnxt: handle reset notify async event from FW") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-01-28 07:29:23 +00:00
if (ret)
return ret;
}
if (dev->data->promiscuous) {
net/bnxt: handle reset notify async event from FW When the FW upgrade is initiated the current instance of FW issues a HWRM_ASYNC_EVENT_CMPL_EVENT_ID_RESET_NOTIFY async notification to the driver. On receiving this notification, the PMD shall quiesce itself and poll on the HWRM_VER_GET FW command at regular intervals. Once the VER_GET command succeeds, the driver should go through the rediscovery process and re-initialize the device. Also register with FW for the reset notify async event. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:23 +00:00
ret = bnxt_promiscuous_enable_op(dev);
net/bnxt: fix return code handling in VLAN config return value stored in "ret" but it has been overwritten before use. Coverity issue: 353621 Fixes: 7fe5668d2ea3 ("net/bnxt: support VLAN filter and strip") Fixes: df6cd7c1f73a ("net/bnxt: handle reset notify async event from FW") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-01-28 07:29:23 +00:00
if (ret)
return ret;
}
net/bnxt: handle reset notify async event from FW When the FW upgrade is initiated the current instance of FW issues a HWRM_ASYNC_EVENT_CMPL_EVENT_ID_RESET_NOTIFY async notification to the driver. On receiving this notification, the PMD shall quiesce itself and poll on the HWRM_VER_GET FW command at regular intervals. Once the VER_GET command succeeds, the driver should go through the rediscovery process and re-initialize the device. Also register with FW for the reset notify async event. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:23 +00:00
net/bnxt: restore MAC filters during reset recovery Older Firmware could have state information such as MAC Filters, VLAN settings etc configured by user. But new Firmware is unaware of this state information and as a result driver should restore these settings during reset recovery. This patch restores the user configured mac addresses prior to hot FW upgrade or FW error. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-12-10 15:05:24 +00:00
ret = bnxt_restore_mac_filters(bp);
net/bnxt: restore VLAN filters during reset recovery Older Firmware could have state information such as MAC Filters, VLAN settings etc configured by user. But new Firmware is unaware of this state information and as a result driver should restore these settings during reset recovery. This patch restores the user configured vlan settings prior to hot FW upgrade or FW error. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-12-10 15:05:25 +00:00
if (ret)
return ret;
ret = bnxt_restore_vlan_filters(bp);
net/bnxt: handle reset notify async event from FW When the FW upgrade is initiated the current instance of FW issues a HWRM_ASYNC_EVENT_CMPL_EVENT_ID_RESET_NOTIFY async notification to the driver. On receiving this notification, the PMD shall quiesce itself and poll on the HWRM_VER_GET FW command at regular intervals. Once the VER_GET command succeeds, the driver should go through the rediscovery process and re-initialize the device. Also register with FW for the reset notify async event. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:23 +00:00
/* TODO restore other filters as well */
return ret;
}
net/bnxt: fix FW readiness check during recovery Moved fw readiness check to a new routine bnxt_check_fw_ready(). During error recovery, driver needs to wait for fw readiness. For that, it uses bnxt_hwrm_ver_get() function now and that function does parsing of the VER_GET response as well. Added a new lightweight function bnxt_hwrm_poll_ver_get() for polling the firmware readiness which issues VER_GET and checks for success without processing the command response. Fixes: df6cd7c1f73a ("net/bnxt: handle reset notify async event from FW") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2021-02-24 15:55:52 +00:00
static int bnxt_check_fw_ready(struct bnxt *bp)
{
int timeout = bp->fw_reset_max_msecs;
int rc = 0;
do {
rc = bnxt_hwrm_poll_ver_get(bp);
if (rc == 0)
break;
rte_delay_ms(BNXT_FW_READY_WAIT_INTERVAL);
timeout -= BNXT_FW_READY_WAIT_INTERVAL;
} while (rc && timeout > 0);
if (rc)
PMD_DRV_LOG(ERR, "FW is not Ready after reset\n");
return rc;
}
net/bnxt: handle reset notify async event from FW When the FW upgrade is initiated the current instance of FW issues a HWRM_ASYNC_EVENT_CMPL_EVENT_ID_RESET_NOTIFY async notification to the driver. On receiving this notification, the PMD shall quiesce itself and poll on the HWRM_VER_GET FW command at regular intervals. Once the VER_GET command succeeds, the driver should go through the rediscovery process and re-initialize the device. Also register with FW for the reset notify async event. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:23 +00:00
static void bnxt_dev_recover(void *arg)
{
struct bnxt *bp = arg;
int rc = 0;
net/bnxt: check chip reset in stop and close While the error recovery thread is running, an application can invoke dev_stop or dev_close_op thus triggering a race and unwanted consequences if dev_close is invoked while the recovery is not yet completed. Fix by having another lock to synchronize between the 2 threads and return EGAIN if adapter is in the middle of recovery when dev_stop or dev_close ops are invoked Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-24 09:37:34 +00:00
pthread_mutex_lock(&bp->err_recovery_lock);
net/bnxt: fix firmware fatal error handling During some fatal firmware error conditions, the PCI config space register 0x2e which normally contains the subsystem ID will become 0xffff. This register will revert back to the normal value after the chip has completed core reset. If we detect this condition, we can poll this config register immediately for the value to revert. Because we use config read cycles to poll this register, there is no possibility of Master Abort if we happen to read it during core reset. This speeds up recovery significantly as we don't have to wait for the conservative min_time before polling to see if the firmware has come out of reset. As soon as this register changes value we can proceed to re-initialize the device. Fixes: df6cd7c1f73a ("net/bnxt: handle reset notify async event from FW") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-02-24 15:55:51 +00:00
if (!bp->fw_reset_min_msecs) {
rc = bnxt_check_fw_reset_done(bp);
if (rc)
goto err;
}
net/bnxt: handle fatal event from FW When firmware hit some unrecoverable error conditions, firmware initiate the recovery by sending an async event EVENT_CMPL_EVENT_ID_RESET_NOTIFY with data1 set to RESET_NOTIFY_EVENT_DATA1_REASON_CODE_FW_EXCEPTION_FATAL to all host drivers and will reset the chip. The recovery procedure is same sequence as the one for hot FW upgrade. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 01:23:25 +00:00
/* Clear Error flag so that device re-init should happen */
bp->flags &= ~BNXT_FLAG_FATAL_ERROR;
net/bnxt: fix FW readiness check during recovery Moved fw readiness check to a new routine bnxt_check_fw_ready(). During error recovery, driver needs to wait for fw readiness. For that, it uses bnxt_hwrm_ver_get() function now and that function does parsing of the VER_GET response as well. Added a new lightweight function bnxt_hwrm_poll_ver_get() for polling the firmware readiness which issues VER_GET and checks for success without processing the command response. Fixes: df6cd7c1f73a ("net/bnxt: handle reset notify async event from FW") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2021-02-24 15:55:52 +00:00
rc = bnxt_check_fw_ready(bp);
if (rc)
net/bnxt: handle reset notify async event from FW When the FW upgrade is initiated the current instance of FW issues a HWRM_ASYNC_EVENT_CMPL_EVENT_ID_RESET_NOTIFY async notification to the driver. On receiving this notification, the PMD shall quiesce itself and poll on the HWRM_VER_GET FW command at regular intervals. Once the VER_GET command succeeds, the driver should go through the rediscovery process and re-initialize the device. Also register with FW for the reset notify async event. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:23 +00:00
goto err;
rc = bnxt_init_resources(bp, true);
if (rc) {
PMD_DRV_LOG(ERR,
"Failed to initialize resources after reset\n");
goto err;
}
/* clear reset flag as the device is initialized now */
bp->flags &= ~BNXT_FLAG_FW_RESET;
rc = bnxt_dev_start_op(bp->eth_dev);
if (rc) {
PMD_DRV_LOG(ERR, "Failed to start port after reset\n");
net/bnxt: fix port stop on error recovery failure During live FW upgrade or error recovery, if restoring the filter settings fail after port start, driver invokes bnxt_uninit_resources() only. Fix it to invoke bnxt_dev_stop_op() first before calling bnxt_uninit_resources(). Fixes: df6cd7c1f73a ("net/bnxt: handle reset notify async event from FW") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Rahul Gupta <rahul.gupta@broadcom.com>
2020-02-06 16:33:12 +00:00
goto err_start;
net/bnxt: handle reset notify async event from FW When the FW upgrade is initiated the current instance of FW issues a HWRM_ASYNC_EVENT_CMPL_EVENT_ID_RESET_NOTIFY async notification to the driver. On receiving this notification, the PMD shall quiesce itself and poll on the HWRM_VER_GET FW command at regular intervals. Once the VER_GET command succeeds, the driver should go through the rediscovery process and re-initialize the device. Also register with FW for the reset notify async event. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:23 +00:00
}
rc = bnxt_restore_filters(bp);
if (rc)
net/bnxt: fix port stop on error recovery failure During live FW upgrade or error recovery, if restoring the filter settings fail after port start, driver invokes bnxt_uninit_resources() only. Fix it to invoke bnxt_dev_stop_op() first before calling bnxt_uninit_resources(). Fixes: df6cd7c1f73a ("net/bnxt: handle reset notify async event from FW") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Rahul Gupta <rahul.gupta@broadcom.com>
2020-02-06 16:33:12 +00:00
goto err_start;
net/bnxt: handle reset notify async event from FW When the FW upgrade is initiated the current instance of FW issues a HWRM_ASYNC_EVENT_CMPL_EVENT_ID_RESET_NOTIFY async notification to the driver. On receiving this notification, the PMD shall quiesce itself and poll on the HWRM_VER_GET FW command at regular intervals. Once the VER_GET command succeeds, the driver should go through the rediscovery process and re-initialize the device. Also register with FW for the reset notify async event. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:23 +00:00
PMD_DRV_LOG(INFO, "Recovered from FW reset\n");
net/bnxt: check chip reset in stop and close While the error recovery thread is running, an application can invoke dev_stop or dev_close_op thus triggering a race and unwanted consequences if dev_close is invoked while the recovery is not yet completed. Fix by having another lock to synchronize between the 2 threads and return EGAIN if adapter is in the middle of recovery when dev_stop or dev_close ops are invoked Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-24 09:37:34 +00:00
pthread_mutex_unlock(&bp->err_recovery_lock);
net/bnxt: handle reset notify async event from FW When the FW upgrade is initiated the current instance of FW issues a HWRM_ASYNC_EVENT_CMPL_EVENT_ID_RESET_NOTIFY async notification to the driver. On receiving this notification, the PMD shall quiesce itself and poll on the HWRM_VER_GET FW command at regular intervals. Once the VER_GET command succeeds, the driver should go through the rediscovery process and re-initialize the device. Also register with FW for the reset notify async event. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:23 +00:00
return;
net/bnxt: fix port stop on error recovery failure During live FW upgrade or error recovery, if restoring the filter settings fail after port start, driver invokes bnxt_uninit_resources() only. Fix it to invoke bnxt_dev_stop_op() first before calling bnxt_uninit_resources(). Fixes: df6cd7c1f73a ("net/bnxt: handle reset notify async event from FW") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Rahul Gupta <rahul.gupta@broadcom.com>
2020-02-06 16:33:12 +00:00
err_start:
net/bnxt: check chip reset in stop and close While the error recovery thread is running, an application can invoke dev_stop or dev_close_op thus triggering a race and unwanted consequences if dev_close is invoked while the recovery is not yet completed. Fix by having another lock to synchronize between the 2 threads and return EGAIN if adapter is in the middle of recovery when dev_stop or dev_close ops are invoked Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-24 09:37:34 +00:00
bnxt_dev_stop(bp->eth_dev);
net/bnxt: handle reset notify async event from FW When the FW upgrade is initiated the current instance of FW issues a HWRM_ASYNC_EVENT_CMPL_EVENT_ID_RESET_NOTIFY async notification to the driver. On receiving this notification, the PMD shall quiesce itself and poll on the HWRM_VER_GET FW command at regular intervals. Once the VER_GET command succeeds, the driver should go through the rediscovery process and re-initialize the device. Also register with FW for the reset notify async event. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:23 +00:00
err:
bp->flags |= BNXT_FLAG_FATAL_ERROR;
bnxt_uninit_resources(bp, false);
net/bnxt: check chip reset in stop and close While the error recovery thread is running, an application can invoke dev_stop or dev_close_op thus triggering a race and unwanted consequences if dev_close is invoked while the recovery is not yet completed. Fix by having another lock to synchronize between the 2 threads and return EGAIN if adapter is in the middle of recovery when dev_stop or dev_close ops are invoked Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-24 09:37:34 +00:00
pthread_mutex_unlock(&bp->err_recovery_lock);
net/bnxt: handle reset notify async event from FW When the FW upgrade is initiated the current instance of FW issues a HWRM_ASYNC_EVENT_CMPL_EVENT_ID_RESET_NOTIFY async notification to the driver. On receiving this notification, the PMD shall quiesce itself and poll on the HWRM_VER_GET FW command at regular intervals. Once the VER_GET command succeeds, the driver should go through the rediscovery process and re-initialize the device. Also register with FW for the reset notify async event. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:23 +00:00
PMD_DRV_LOG(ERR, "Failed to recover from FW reset\n");
}
void bnxt_dev_reset_and_resume(void *arg)
{
struct bnxt *bp = arg;
net/bnxt: fix firmware fatal error handling During some fatal firmware error conditions, the PCI config space register 0x2e which normally contains the subsystem ID will become 0xffff. This register will revert back to the normal value after the chip has completed core reset. If we detect this condition, we can poll this config register immediately for the value to revert. Because we use config read cycles to poll this register, there is no possibility of Master Abort if we happen to read it during core reset. This speeds up recovery significantly as we don't have to wait for the conservative min_time before polling to see if the firmware has come out of reset. As soon as this register changes value we can proceed to re-initialize the device. Fixes: df6cd7c1f73a ("net/bnxt: handle reset notify async event from FW") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-02-24 15:55:51 +00:00
uint32_t us = US_PER_MS * bp->fw_reset_min_msecs;
uint16_t val = 0;
net/bnxt: handle reset notify async event from FW When the FW upgrade is initiated the current instance of FW issues a HWRM_ASYNC_EVENT_CMPL_EVENT_ID_RESET_NOTIFY async notification to the driver. On receiving this notification, the PMD shall quiesce itself and poll on the HWRM_VER_GET FW command at regular intervals. Once the VER_GET command succeeds, the driver should go through the rediscovery process and re-initialize the device. Also register with FW for the reset notify async event. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:23 +00:00
int rc;
bnxt_dev_cleanup(bp);
net/bnxt: support hot firmware upgrade for Stingray FW sets HWRM_FUNC_QCAPS_OUTPUT_FLAGS_ERR_RECOVER_RELOAD in HWRM_FUNC_QCAPS command, if device requires to invoke fastboot FW during FW reset. Driver has to poll for shutdown bit in fw_status register: 1. in case of hot fw upgrade, this bit will be set after all function drivers unregistered with fw. 2. in case of fw initiated error recovery, this bit will be set after fw has collected the core dump Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Rahul Gupta <rahul.gupta@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 01:23:31 +00:00
bnxt_wait_for_device_shutdown(bp);
net/bnxt: fix firmware fatal error handling During some fatal firmware error conditions, the PCI config space register 0x2e which normally contains the subsystem ID will become 0xffff. This register will revert back to the normal value after the chip has completed core reset. If we detect this condition, we can poll this config register immediately for the value to revert. Because we use config read cycles to poll this register, there is no possibility of Master Abort if we happen to read it during core reset. This speeds up recovery significantly as we don't have to wait for the conservative min_time before polling to see if the firmware has come out of reset. As soon as this register changes value we can proceed to re-initialize the device. Fixes: df6cd7c1f73a ("net/bnxt: handle reset notify async event from FW") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-02-24 15:55:51 +00:00
/* During some fatal firmware error conditions, the PCI config space
* register 0x2e which normally contains the subsystem ID will become
* 0xffff. This register will revert back to the normal value after
* the chip has completed core reset. If we detect this condition,
* we can poll this config register immediately for the value to revert.
*/
if (bp->flags & BNXT_FLAG_FATAL_ERROR) {
rc = rte_pci_read_config(bp->pdev, &val, sizeof(val), PCI_SUBSYSTEM_ID_OFFSET);
if (rc < 0) {
PMD_DRV_LOG(ERR, "Failed to read PCI offset 0x%x", PCI_SUBSYSTEM_ID_OFFSET);
return;
}
if (val == 0xffff) {
bp->fw_reset_min_msecs = 0;
us = 1;
}
}
rc = rte_eal_alarm_set(us, bnxt_dev_recover, (void *)bp);
net/bnxt: handle reset notify async event from FW When the FW upgrade is initiated the current instance of FW issues a HWRM_ASYNC_EVENT_CMPL_EVENT_ID_RESET_NOTIFY async notification to the driver. On receiving this notification, the PMD shall quiesce itself and poll on the HWRM_VER_GET FW command at regular intervals. Once the VER_GET command succeeds, the driver should go through the rediscovery process and re-initialize the device. Also register with FW for the reset notify async event. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:23 +00:00
if (rc)
PMD_DRV_LOG(ERR, "Error setting recovery alarm");
}
net/bnxt: support periodic FW health monitoring Periodically poll the FW heartbeat register and FW recovery counter registers to check the FW health. Polling frequency will be advertised by the FW in HWRM_ERROR_RECOVERY_QCFG response. Schedule the task upon receiving the async event from FW. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:29 +00:00
uint32_t bnxt_read_fw_status_reg(struct bnxt *bp, uint32_t index)
{
struct bnxt_error_recovery_info *info = bp->recovery_info;
uint32_t reg = info->status_regs[index];
uint32_t type, offset, val = 0;
net/bnxt: check PCI config read Return value where return value of rte_pci_read_config was not checked. Fix it. Coverity issue: 349919 Fixes: 9d0cbaecc91a ("net/bnxt: support periodic FW health monitoring") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2021-04-30 20:14:11 +00:00
int ret = 0;
net/bnxt: support periodic FW health monitoring Periodically poll the FW heartbeat register and FW recovery counter registers to check the FW health. Polling frequency will be advertised by the FW in HWRM_ERROR_RECOVERY_QCFG response. Schedule the task upon receiving the async event from FW. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:29 +00:00
type = BNXT_FW_STATUS_REG_TYPE(reg);
offset = BNXT_FW_STATUS_REG_OFF(reg);
switch (type) {
case BNXT_FW_STATUS_REG_TYPE_CFG:
net/bnxt: check PCI config read Return value where return value of rte_pci_read_config was not checked. Fix it. Coverity issue: 349919 Fixes: 9d0cbaecc91a ("net/bnxt: support periodic FW health monitoring") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2021-04-30 20:14:11 +00:00
ret = rte_pci_read_config(bp->pdev, &val, sizeof(val), offset);
if (ret < 0)
PMD_DRV_LOG(ERR, "Failed to read PCI offset %#x",
offset);
net/bnxt: support periodic FW health monitoring Periodically poll the FW heartbeat register and FW recovery counter registers to check the FW health. Polling frequency will be advertised by the FW in HWRM_ERROR_RECOVERY_QCFG response. Schedule the task upon receiving the async event from FW. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:29 +00:00
break;
case BNXT_FW_STATUS_REG_TYPE_GRC:
offset = info->mapped_status_regs[index];
/* FALLTHROUGH */
case BNXT_FW_STATUS_REG_TYPE_BAR0:
val = rte_le_to_cpu_32(rte_read32((uint8_t *)bp->bar0 +
offset));
break;
}
return val;
}
net/bnxt: support FW reset Added code to perform FW_RESET. When the driver detects error in FW, it has to initiate the recovery by resetting the cores. FW advertise the method to do a core reset, reset register offsets and values to perform reset in response of HWRM_ERROR_RECOVERY_QCFG command. There are 2 ways to recover from the error. 1. Master function issues core resets to recover from error. 2. Master function detects chimp dead condition and notify the Kong processor about the chimp dead case through FW_RESET HWRM command. Kong Processor send an RESET_NOTIFY async event with REASON_CODE_FW_EXCEPTION_FATAL to all the PF’s/VF’s that chimp is dead and it is going to reset the chimp. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 01:23:30 +00:00
static int bnxt_fw_reset_all(struct bnxt *bp)
{
struct bnxt_error_recovery_info *info = bp->recovery_info;
uint32_t i;
int rc = 0;
if (info->flags & BNXT_FLAG_ERROR_RECOVERY_HOST) {
/* Reset through master function driver */
for (i = 0; i < info->reg_array_cnt; i++)
bnxt_write_fw_reset_reg(bp, i);
/* Wait for time specified by FW after triggering reset */
rte_delay_ms(info->master_func_wait_period_after_reset);
} else if (info->flags & BNXT_FLAG_ERROR_RECOVERY_CO_CPU) {
/* Reset with the help of Kong processor */
rc = bnxt_hwrm_fw_reset(bp);
if (rc)
PMD_DRV_LOG(ERR, "Failed to reset FW\n");
}
return rc;
}
static void bnxt_fw_reset_cb(void *arg)
{
struct bnxt *bp = arg;
struct bnxt_error_recovery_info *info = bp->recovery_info;
int rc = 0;
/* Only Master function can do FW reset */
if (bnxt_is_master_func(bp) &&
bnxt_is_recovery_enabled(bp)) {
rc = bnxt_fw_reset_all(bp);
if (rc) {
PMD_DRV_LOG(ERR, "Adapter recovery failed\n");
return;
}
}
/* if recovery method is ERROR_RECOVERY_CO_CPU, KONG will send
* EXCEPTION_FATAL_ASYNC event to all the functions
* (including MASTER FUNC). After receiving this Async, all the active
* drivers should treat this case as FW initiated recovery
*/
if (info->flags & BNXT_FLAG_ERROR_RECOVERY_HOST) {
bp->fw_reset_min_msecs = BNXT_MIN_FW_READY_TIMEOUT;
bp->fw_reset_max_msecs = BNXT_MAX_FW_RESET_TIMEOUT;
/* To recover from error */
rte_eal_alarm_set(US_PER_MS, bnxt_dev_reset_and_resume,
(void *)bp);
}
}
net/bnxt: support periodic FW health monitoring Periodically poll the FW heartbeat register and FW recovery counter registers to check the FW health. Polling frequency will be advertised by the FW in HWRM_ERROR_RECOVERY_QCFG response. Schedule the task upon receiving the async event from FW. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:29 +00:00
/* Driver should poll FW heartbeat, reset_counter with the frequency
* advertised by FW in HWRM_ERROR_RECOVERY_QCFG.
* When the driver detects heartbeat stop or change in reset_counter,
* it has to trigger a reset to recover from the error condition.
* A “master PF” is the function who will have the privilege to
* initiate the chimp reset. The master PF will be elected by the
* firmware and will be notified through async message.
*/
static void bnxt_check_fw_health(void *arg)
{
struct bnxt *bp = arg;
struct bnxt_error_recovery_info *info = bp->recovery_info;
net/bnxt: support FW reset Added code to perform FW_RESET. When the driver detects error in FW, it has to initiate the recovery by resetting the cores. FW advertise the method to do a core reset, reset register offsets and values to perform reset in response of HWRM_ERROR_RECOVERY_QCFG command. There are 2 ways to recover from the error. 1. Master function issues core resets to recover from error. 2. Master function detects chimp dead condition and notify the Kong processor about the chimp dead case through FW_RESET HWRM command. Kong Processor send an RESET_NOTIFY async event with REASON_CODE_FW_EXCEPTION_FATAL to all the PF’s/VF’s that chimp is dead and it is going to reset the chimp. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 01:23:30 +00:00
uint32_t val = 0, wait_msec;
net/bnxt: support periodic FW health monitoring Periodically poll the FW heartbeat register and FW recovery counter registers to check the FW health. Polling frequency will be advertised by the FW in HWRM_ERROR_RECOVERY_QCFG response. Schedule the task upon receiving the async event from FW. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:29 +00:00
if (!info || !bnxt_is_recovery_enabled(bp) ||
is_bnxt_in_error(bp))
return;
val = bnxt_read_fw_status_reg(bp, BNXT_FW_HEARTBEAT_CNT_REG);
if (val == info->last_heart_beat)
goto reset;
info->last_heart_beat = val;
val = bnxt_read_fw_status_reg(bp, BNXT_FW_RECOVERY_CNT_REG);
if (val != info->last_reset_counter)
goto reset;
info->last_reset_counter = val;
rte_eal_alarm_set(US_PER_MS * info->driver_polling_freq,
bnxt_check_fw_health, (void *)bp);
return;
reset:
/* Stop DMA to/from device */
bp->flags |= BNXT_FLAG_FATAL_ERROR;
bp->flags |= BNXT_FLAG_FW_RESET;
net/bnxt: prevent device access in error state Driver should prevent any DMA with the device when it detects an error. When firmware is in fatal state, stop tx/rx by assigning them to dummy functions. Fixes: be14720def9c ("net/bnxt: support FW reset") Fixes: 9d0cbaecc91a ("net/bnxt: support periodic FW health monitoring") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Andy Gospodarek <gospo@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-05-03 05:21:50 +00:00
bnxt_stop_rxtx(bp);
net/bnxt: support periodic FW health monitoring Periodically poll the FW heartbeat register and FW recovery counter registers to check the FW health. Polling frequency will be advertised by the FW in HWRM_ERROR_RECOVERY_QCFG response. Schedule the task upon receiving the async event from FW. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:29 +00:00
PMD_DRV_LOG(ERR, "Detected FW dead condition\n");
net/bnxt: support FW reset Added code to perform FW_RESET. When the driver detects error in FW, it has to initiate the recovery by resetting the cores. FW advertise the method to do a core reset, reset register offsets and values to perform reset in response of HWRM_ERROR_RECOVERY_QCFG command. There are 2 ways to recover from the error. 1. Master function issues core resets to recover from error. 2. Master function detects chimp dead condition and notify the Kong processor about the chimp dead case through FW_RESET HWRM command. Kong Processor send an RESET_NOTIFY async event with REASON_CODE_FW_EXCEPTION_FATAL to all the PF’s/VF’s that chimp is dead and it is going to reset the chimp. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 01:23:30 +00:00
if (bnxt_is_master_func(bp))
wait_msec = info->master_func_wait_period;
else
wait_msec = info->normal_func_wait_period;
rte_eal_alarm_set(US_PER_MS * wait_msec,
bnxt_fw_reset_cb, (void *)bp);
net/bnxt: support periodic FW health monitoring Periodically poll the FW heartbeat register and FW recovery counter registers to check the FW health. Polling frequency will be advertised by the FW in HWRM_ERROR_RECOVERY_QCFG response. Schedule the task upon receiving the async event from FW. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:29 +00:00
}
void bnxt_schedule_fw_health_check(struct bnxt *bp)
{
uint32_t polling_freq;
net/bnxt: add separate mutex for FW health check def_cp_lock was added to sync race between dev_configure and int_handler. It should not be used to synchronize scheduling of FW health check between dev_start and async event handler as well, use a separate mutex for the same. Fixes: a73b8e939f10 ("net/bnxt: fix race between start and interrupt handler") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-11 01:56:03 +00:00
pthread_mutex_lock(&bp->health_check_lock);
net/bnxt: support periodic FW health monitoring Periodically poll the FW heartbeat register and FW recovery counter registers to check the FW health. Polling frequency will be advertised by the FW in HWRM_ERROR_RECOVERY_QCFG response. Schedule the task upon receiving the async event from FW. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:29 +00:00
if (!bnxt_is_recovery_enabled(bp))
net/bnxt: add separate mutex for FW health check def_cp_lock was added to sync race between dev_configure and int_handler. It should not be used to synchronize scheduling of FW health check between dev_start and async event handler as well, use a separate mutex for the same. Fixes: a73b8e939f10 ("net/bnxt: fix race between start and interrupt handler") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-11 01:56:03 +00:00
goto done;
net/bnxt: support periodic FW health monitoring Periodically poll the FW heartbeat register and FW recovery counter registers to check the FW health. Polling frequency will be advertised by the FW in HWRM_ERROR_RECOVERY_QCFG response. Schedule the task upon receiving the async event from FW. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:29 +00:00
if (bp->flags & BNXT_FLAG_FW_HEALTH_CHECK_SCHEDULED)
net/bnxt: add separate mutex for FW health check def_cp_lock was added to sync race between dev_configure and int_handler. It should not be used to synchronize scheduling of FW health check between dev_start and async event handler as well, use a separate mutex for the same. Fixes: a73b8e939f10 ("net/bnxt: fix race between start and interrupt handler") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-11 01:56:03 +00:00
goto done;
net/bnxt: support periodic FW health monitoring Periodically poll the FW heartbeat register and FW recovery counter registers to check the FW health. Polling frequency will be advertised by the FW in HWRM_ERROR_RECOVERY_QCFG response. Schedule the task upon receiving the async event from FW. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:29 +00:00
polling_freq = bp->recovery_info->driver_polling_freq;
rte_eal_alarm_set(US_PER_MS * polling_freq,
bnxt_check_fw_health, (void *)bp);
bp->flags |= BNXT_FLAG_FW_HEALTH_CHECK_SCHEDULED;
net/bnxt: add separate mutex for FW health check def_cp_lock was added to sync race between dev_configure and int_handler. It should not be used to synchronize scheduling of FW health check between dev_start and async event handler as well, use a separate mutex for the same. Fixes: a73b8e939f10 ("net/bnxt: fix race between start and interrupt handler") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-11 01:56:03 +00:00
done:
pthread_mutex_unlock(&bp->health_check_lock);
net/bnxt: support periodic FW health monitoring Periodically poll the FW heartbeat register and FW recovery counter registers to check the FW health. Polling frequency will be advertised by the FW in HWRM_ERROR_RECOVERY_QCFG response. Schedule the task upon receiving the async event from FW. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:29 +00:00
}
static void bnxt_cancel_fw_health_check(struct bnxt *bp)
{
rte_eal_alarm_cancel(bnxt_check_fw_health, (void *)bp);
bp->flags &= ~BNXT_FLAG_FW_HEALTH_CHECK_SCHEDULED;
}
net/bnxt: refactor chip type selection Use switch-case statement to replace if-else statement. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-12-02 04:18:07 +00:00
static bool bnxt_vf_pciid(uint16_t device_id)
net/bnxt: add hardware resource manager init code Start adding support to use the HWRM API. Hardware Resource Manager or HWRM in short, is a set of API provided by the firmware running in the ASIC to manage the various resources. Initial commit just performs necessary HWRM queries for init, then fails as before. Now that struct bnxt is non-zero size, we can set dev_private_size correctly. The used HWRM calls so far: bnxt_hwrm_func_qcaps: This command returns capabilities of a function. bnxt_hwrm_ver_get: This function is called by a driver to determine the HWRM interface version supported by the HWRM firmware, the version of HWRM firmware implementation, the name of HWRM firmware, the versions of other embedded firmwares, and the names of other embedded firmwares, etc. Gets the firmware version and interface specifications. Returns an error if the firmware on the device is not supported by the driver and ensures the response space is large enough for the largest possible response. bnxt_hwrm_queue_qportcfg: This function is called by a driver to query queue configuration of a port. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:02 +00:00
{
net/bnxt: refactor chip type selection Use switch-case statement to replace if-else statement. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-12-02 04:18:07 +00:00
switch (device_id) {
case BROADCOM_DEV_ID_57304_VF:
case BROADCOM_DEV_ID_57406_VF:
case BROADCOM_DEV_ID_5731X_VF:
case BROADCOM_DEV_ID_5741X_VF:
case BROADCOM_DEV_ID_57414_VF:
case BROADCOM_DEV_ID_STRATUS_NIC_VF1:
case BROADCOM_DEV_ID_STRATUS_NIC_VF2:
case BROADCOM_DEV_ID_58802_VF:
case BROADCOM_DEV_ID_57500_VF1:
case BROADCOM_DEV_ID_57500_VF2:
net/bnxt: support 58818 chip family The new chip (Stingray 2) is part of the P5 chip family with a number of changes: 1. Implement the epoch doorbell bit for 58818 chip. With the new doorbell infrastructure and the unbounded index logic, now set the epoch doorbell bit to support proper doorbell operation on the new chip. Toggle epoch bit of all rings when it's wrapped to support doorbell overflow checking. 2. Get the legacy doorbell size from firmware. Legacy doorbell support has been removed in Stingray 2. So, the fast path doorbell pages start from the base of the BAR. Drivers need to use legacy_l2_db_space_size_kb field in the hwrm_func_qcfg_output response to get the legacy doorbell page offset from the BAR. 3. Set VALID doorbell bit on 58818 chip family. This class of chip has a valid doorbell bit added and it needs to be set. 4. Use "chip_num" returned by firmware. The "chip_num" field in the HWRM_VER_GET output returns the chip number. Use this value to identify chip category for 58818 chip family. 5. Added device ids for Stingray2 PF/VF devices. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-20 05:24:25 +00:00
case BROADCOM_DEV_ID_58818_VF:
net/bnxt: refactor chip type selection Use switch-case statement to replace if-else statement. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-12-02 04:18:07 +00:00
/* FALLTHROUGH */
net/bnxt: add hardware resource manager init code Start adding support to use the HWRM API. Hardware Resource Manager or HWRM in short, is a set of API provided by the firmware running in the ASIC to manage the various resources. Initial commit just performs necessary HWRM queries for init, then fails as before. Now that struct bnxt is non-zero size, we can set dev_private_size correctly. The used HWRM calls so far: bnxt_hwrm_func_qcaps: This command returns capabilities of a function. bnxt_hwrm_ver_get: This function is called by a driver to determine the HWRM interface version supported by the HWRM firmware, the version of HWRM firmware implementation, the name of HWRM firmware, the versions of other embedded firmwares, and the names of other embedded firmwares, etc. Gets the firmware version and interface specifications. Returns an error if the firmware on the device is not supported by the driver and ensures the response space is large enough for the largest possible response. bnxt_hwrm_queue_qportcfg: This function is called by a driver to query queue configuration of a port. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:02 +00:00
return true;
net/bnxt: refactor chip type selection Use switch-case statement to replace if-else statement. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-12-02 04:18:07 +00:00
default:
return false;
}
net/bnxt: add hardware resource manager init code Start adding support to use the HWRM API. Hardware Resource Manager or HWRM in short, is a set of API provided by the firmware running in the ASIC to manage the various resources. Initial commit just performs necessary HWRM queries for init, then fails as before. Now that struct bnxt is non-zero size, we can set dev_private_size correctly. The used HWRM calls so far: bnxt_hwrm_func_qcaps: This command returns capabilities of a function. bnxt_hwrm_ver_get: This function is called by a driver to determine the HWRM interface version supported by the HWRM firmware, the version of HWRM firmware implementation, the name of HWRM firmware, the versions of other embedded firmwares, and the names of other embedded firmwares, etc. Gets the firmware version and interface specifications. Returns an error if the firmware on the device is not supported by the driver and ensures the response space is large enough for the largest possible response. bnxt_hwrm_queue_qportcfg: This function is called by a driver to query queue configuration of a port. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:02 +00:00
}
net/bnxt: remove references to Thor Refactor code to remove references to Thor. Instead use P5 as in phase 5 of development cycle since it is applicable to boards other than Thor as well. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2020-12-01 19:15:23 +00:00
/* Phase 5 device */
static bool bnxt_p5_device(uint16_t device_id)
net/bnxt: add wrapper function for Thor device check This patch adds a wrapper function bnxt_thor_device() which checks for the pci device is Thor family. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com>
2019-10-24 07:44:30 +00:00
{
net/bnxt: refactor chip type selection Use switch-case statement to replace if-else statement. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-12-02 04:18:07 +00:00
switch (device_id) {
case BROADCOM_DEV_ID_57508:
case BROADCOM_DEV_ID_57504:
case BROADCOM_DEV_ID_57502:
case BROADCOM_DEV_ID_57508_MF1:
case BROADCOM_DEV_ID_57504_MF1:
case BROADCOM_DEV_ID_57502_MF1:
case BROADCOM_DEV_ID_57508_MF2:
case BROADCOM_DEV_ID_57504_MF2:
case BROADCOM_DEV_ID_57502_MF2:
case BROADCOM_DEV_ID_57500_VF1:
case BROADCOM_DEV_ID_57500_VF2:
net/bnxt: support 58818 chip family The new chip (Stingray 2) is part of the P5 chip family with a number of changes: 1. Implement the epoch doorbell bit for 58818 chip. With the new doorbell infrastructure and the unbounded index logic, now set the epoch doorbell bit to support proper doorbell operation on the new chip. Toggle epoch bit of all rings when it's wrapped to support doorbell overflow checking. 2. Get the legacy doorbell size from firmware. Legacy doorbell support has been removed in Stingray 2. So, the fast path doorbell pages start from the base of the BAR. Drivers need to use legacy_l2_db_space_size_kb field in the hwrm_func_qcfg_output response to get the legacy doorbell page offset from the BAR. 3. Set VALID doorbell bit on 58818 chip family. This class of chip has a valid doorbell bit added and it needs to be set. 4. Use "chip_num" returned by firmware. The "chip_num" field in the HWRM_VER_GET output returns the chip number. Use this value to identify chip category for 58818 chip family. 5. Added device ids for Stingray2 PF/VF devices. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-20 05:24:25 +00:00
case BROADCOM_DEV_ID_58812:
case BROADCOM_DEV_ID_58814:
case BROADCOM_DEV_ID_58818:
case BROADCOM_DEV_ID_58818_VF:
net/bnxt: refactor chip type selection Use switch-case statement to replace if-else statement. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-12-02 04:18:07 +00:00
/* FALLTHROUGH */
net/bnxt: add wrapper function for Thor device check This patch adds a wrapper function bnxt_thor_device() which checks for the pci device is Thor family. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com>
2019-10-24 07:44:30 +00:00
return true;
net/bnxt: refactor chip type selection Use switch-case statement to replace if-else statement. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-12-02 04:18:07 +00:00
default:
return false;
}
net/bnxt: add wrapper function for Thor device check This patch adds a wrapper function bnxt_thor_device() which checks for the pci device is Thor family. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com>
2019-10-24 07:44:30 +00:00
}
net/bnxt: set ring coalesce parameters for Stratus NIC Set ring coalesce parameters for Stratus NIC. Other skews don't necessarily need this. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-06-28 20:15:32 +00:00
bool bnxt_stratus_device(struct bnxt *bp)
{
net/bnxt: refactor chip type selection Use switch-case statement to replace if-else statement. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-12-02 04:18:07 +00:00
uint16_t device_id = bp->pdev->id.device_id;
net/bnxt: set ring coalesce parameters for Stratus NIC Set ring coalesce parameters for Stratus NIC. Other skews don't necessarily need this. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-06-28 20:15:32 +00:00
net/bnxt: refactor chip type selection Use switch-case statement to replace if-else statement. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-12-02 04:18:07 +00:00
switch (device_id) {
case BROADCOM_DEV_ID_STRATUS_NIC:
case BROADCOM_DEV_ID_STRATUS_NIC_VF1:
case BROADCOM_DEV_ID_STRATUS_NIC_VF2:
/* FALLTHROUGH */
net/bnxt: set ring coalesce parameters for Stratus NIC Set ring coalesce parameters for Stratus NIC. Other skews don't necessarily need this. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-06-28 20:15:32 +00:00
return true;
net/bnxt: refactor chip type selection Use switch-case statement to replace if-else statement. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-12-02 04:18:07 +00:00
default:
return false;
}
net/bnxt: set ring coalesce parameters for Stratus NIC Set ring coalesce parameters for Stratus NIC. Other skews don't necessarily need this. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-06-28 20:15:32 +00:00
}
net/bnxt: refactor init/uninit Move all the individual driver fields allocation routines to one routine - bnxt_drv_init(). This houses all such routines where memory needs to be allocated once during the driver's lifetime and does not need to be torn down during error recovery. Rename some function names in accordance with their functionality. bnxt_init_board() is doing nothing more than mapping the PCI bars, so rename it as such. Given that there is a bnxt_shutdown_nic that is called in dev_stop_op, rename it's counterpart - bnxt_init_chip() that is called in dev_start_op, to bnxt_start_nic. Also helps avoid confusion with some of the other bnxt_init_xxx routines. Rename bnxt_init_fw() to bnxt_get_config() as that is what that routine is doing mostly functionality wise. Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-01-14 02:19:14 +00:00
static int bnxt_map_pci_bars(struct rte_eth_dev *eth_dev)
net/bnxt: add hardware resource manager init code Start adding support to use the HWRM API. Hardware Resource Manager or HWRM in short, is a set of API provided by the firmware running in the ASIC to manage the various resources. Initial commit just performs necessary HWRM queries for init, then fails as before. Now that struct bnxt is non-zero size, we can set dev_private_size correctly. The used HWRM calls so far: bnxt_hwrm_func_qcaps: This command returns capabilities of a function. bnxt_hwrm_ver_get: This function is called by a driver to determine the HWRM interface version supported by the HWRM firmware, the version of HWRM firmware implementation, the name of HWRM firmware, the versions of other embedded firmwares, and the names of other embedded firmwares, etc. Gets the firmware version and interface specifications. Returns an error if the firmware on the device is not supported by the driver and ensures the response space is large enough for the largest possible response. bnxt_hwrm_queue_qportcfg: This function is called by a driver to query queue configuration of a port. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:02 +00:00
{
drivers/net: add generic ethdev macro to get PCI device Instead of many PMD define their own macro, define a generic one in ethdev and use that in PMDs. Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com> Acked-by: Allain Legacy <allain.legacy@windriver.com> Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com> Acked-by: Andrew Rybchenko <arybchenko@solarflare.com>
2017-05-15 10:24:03 +00:00
struct rte_pci_device *pci_dev = RTE_ETH_DEV_TO_PCI(eth_dev);
net/bnxt: fix return values to standard error codes Fixed the return values of few routines to return standard error code. Also fixed few error logs to more meaningful one. Fixes: 804e746c7b73 ("net/bnxt: add hardware resource manager init code") Fixes: e3d8f1e6a665 ("net/bnxt: cache address of doorbell to subsequent access") Fixes: 19e6af01bb36 ("net/bnxt: support get/set EEPROM") Fixes: b7435d660a8c ("net/bnxt: add ntuple filtering support") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-07-18 03:36:06 +00:00
struct bnxt *bp = eth_dev->data->dev_private;
net/bnxt: add hardware resource manager init code Start adding support to use the HWRM API. Hardware Resource Manager or HWRM in short, is a set of API provided by the firmware running in the ASIC to manage the various resources. Initial commit just performs necessary HWRM queries for init, then fails as before. Now that struct bnxt is non-zero size, we can set dev_private_size correctly. The used HWRM calls so far: bnxt_hwrm_func_qcaps: This command returns capabilities of a function. bnxt_hwrm_ver_get: This function is called by a driver to determine the HWRM interface version supported by the HWRM firmware, the version of HWRM firmware implementation, the name of HWRM firmware, the versions of other embedded firmwares, and the names of other embedded firmwares, etc. Gets the firmware version and interface specifications. Returns an error if the firmware on the device is not supported by the driver and ensures the response space is large enough for the largest possible response. bnxt_hwrm_queue_qportcfg: This function is called by a driver to query queue configuration of a port. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:02 +00:00
/* enable device (incl. PCI PM wakeup), and bus-mastering */
net/bnxt: fix return values to standard error codes Fixed the return values of few routines to return standard error code. Also fixed few error logs to more meaningful one. Fixes: 804e746c7b73 ("net/bnxt: add hardware resource manager init code") Fixes: e3d8f1e6a665 ("net/bnxt: cache address of doorbell to subsequent access") Fixes: 19e6af01bb36 ("net/bnxt: support get/set EEPROM") Fixes: b7435d660a8c ("net/bnxt: add ntuple filtering support") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-07-18 03:36:06 +00:00
bp->bar0 = (void *)pci_dev->mem_resource[0].addr;
bp->doorbell_base = (void *)pci_dev->mem_resource[2].addr;
if (!bp->bar0 || !bp->doorbell_base) {
PMD_DRV_LOG(ERR, "Unable to access Hardware\n");
return -ENODEV;
net/bnxt: add hardware resource manager init code Start adding support to use the HWRM API. Hardware Resource Manager or HWRM in short, is a set of API provided by the firmware running in the ASIC to manage the various resources. Initial commit just performs necessary HWRM queries for init, then fails as before. Now that struct bnxt is non-zero size, we can set dev_private_size correctly. The used HWRM calls so far: bnxt_hwrm_func_qcaps: This command returns capabilities of a function. bnxt_hwrm_ver_get: This function is called by a driver to determine the HWRM interface version supported by the HWRM firmware, the version of HWRM firmware implementation, the name of HWRM firmware, the versions of other embedded firmwares, and the names of other embedded firmwares, etc. Gets the firmware version and interface specifications. Returns an error if the firmware on the device is not supported by the driver and ensures the response space is large enough for the largest possible response. bnxt_hwrm_queue_qportcfg: This function is called by a driver to query queue configuration of a port. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:02 +00:00
}
bp->eth_dev = eth_dev;
net/bnxt: localize mapping of ethdev to PCI device Use existing information about pci and interrupt handle to minimize the number of places that assume eth_dev contains pci_device information. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Jan Blunck <jblunck@infradead.org>
2016-12-23 15:57:59 +00:00
bp->pdev = pci_dev;
net/bnxt: add hardware resource manager init code Start adding support to use the HWRM API. Hardware Resource Manager or HWRM in short, is a set of API provided by the firmware running in the ASIC to manage the various resources. Initial commit just performs necessary HWRM queries for init, then fails as before. Now that struct bnxt is non-zero size, we can set dev_private_size correctly. The used HWRM calls so far: bnxt_hwrm_func_qcaps: This command returns capabilities of a function. bnxt_hwrm_ver_get: This function is called by a driver to determine the HWRM interface version supported by the HWRM firmware, the version of HWRM firmware implementation, the name of HWRM firmware, the versions of other embedded firmwares, and the names of other embedded firmwares, etc. Gets the firmware version and interface specifications. Returns an error if the firmware on the device is not supported by the driver and ensures the response space is large enough for the largest possible response. bnxt_hwrm_queue_qportcfg: This function is called by a driver to query queue configuration of a port. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:02 +00:00
return 0;
}
net/bnxt: remove unnecessary macro for unused variables Remove "__rte_unused" instances that are not required. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Rahul Gupta <rahul.gupta@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-12-10 15:05:27 +00:00
static int bnxt_alloc_ctx_mem_blk(struct bnxt *bp,
net/bnxt: support thor controller This commit adds support to the bnxt PMD for devices based on the BCM57508 "thor" Ethernet controller. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
2019-06-02 17:42:44 +00:00
struct bnxt_ctx_pg_info *ctx_pg,
uint32_t mem_size,
const char *suffix,
uint16_t idx)
{
struct bnxt_ring_mem_info *rmem = &ctx_pg->ring_mem;
const struct rte_memzone *mz = NULL;
char mz_name[RTE_MEMZONE_NAMESIZE];
rte_iova_t mz_phys_addr;
uint64_t valid_bits = 0;
uint32_t sz;
int i;
if (!mem_size)
return 0;
rmem->nr_pages = RTE_ALIGN_MUL_CEIL(mem_size, BNXT_PAGE_SIZE) /
BNXT_PAGE_SIZE;
rmem->page_size = BNXT_PAGE_SIZE;
rmem->pg_arr = ctx_pg->ctx_pg_arr;
rmem->dma_arr = ctx_pg->ctx_dma_arr;
rmem->flags = BNXT_RMEM_VALID_PTE_FLAG;
valid_bits = PTU_PTE_VALID;
if (rmem->nr_pages > 1) {
net/bnxt: fix context memory allocation There is a bug in context memory allocation because of which it results in reusing the context memory allocated for the first port while allocating memory for next ports. Fix it by passing the port id in the name field while allocating context memory. Fixes: f8168ca0e690 ("net/bnxt: support thor controller") Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-07-29 21:28:05 +00:00
snprintf(mz_name, RTE_MEMZONE_NAMESIZE,
"bnxt_ctx_pg_tbl%s_%x_%d",
suffix, idx, bp->eth_dev->data->port_id);
net/bnxt: support thor controller This commit adds support to the bnxt PMD for devices based on the BCM57508 "thor" Ethernet controller. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
2019-06-02 17:42:44 +00:00
mz_name[RTE_MEMZONE_NAMESIZE - 1] = 0;
mz = rte_memzone_lookup(mz_name);
if (!mz) {
mz = rte_memzone_reserve_aligned(mz_name,
rmem->nr_pages * 8,
SOCKET_ID_ANY,
RTE_MEMZONE_2MB |
RTE_MEMZONE_SIZE_HINT_ONLY |
RTE_MEMZONE_IOVA_CONTIG,
BNXT_PAGE_SIZE);
if (mz == NULL)
return -ENOMEM;
}
memset(mz->addr, 0, mz->len);
mz_phys_addr = mz->iova;
rmem->pg_tbl = mz->addr;
rmem->pg_tbl_map = mz_phys_addr;
rmem->pg_tbl_mz = mz;
}
net/bnxt: fix context memory allocation There is a bug in context memory allocation because of which it results in reusing the context memory allocated for the first port while allocating memory for next ports. Fix it by passing the port id in the name field while allocating context memory. Fixes: f8168ca0e690 ("net/bnxt: support thor controller") Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-07-29 21:28:05 +00:00
snprintf(mz_name, RTE_MEMZONE_NAMESIZE, "bnxt_ctx_%s_%x_%d",
suffix, idx, bp->eth_dev->data->port_id);
net/bnxt: support thor controller This commit adds support to the bnxt PMD for devices based on the BCM57508 "thor" Ethernet controller. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
2019-06-02 17:42:44 +00:00
mz = rte_memzone_lookup(mz_name);
if (!mz) {
mz = rte_memzone_reserve_aligned(mz_name,
mem_size,
SOCKET_ID_ANY,
RTE_MEMZONE_1GB |
RTE_MEMZONE_SIZE_HINT_ONLY |
RTE_MEMZONE_IOVA_CONTIG,
BNXT_PAGE_SIZE);
if (mz == NULL)
return -ENOMEM;
}
memset(mz->addr, 0, mz->len);
mz_phys_addr = mz->iova;
for (sz = 0, i = 0; sz < mem_size; sz += BNXT_PAGE_SIZE, i++) {
rmem->pg_arr[i] = ((char *)mz->addr) + sz;
rmem->dma_arr[i] = mz_phys_addr + sz;
if (rmem->nr_pages > 1) {
if (i == rmem->nr_pages - 2 &&
(rmem->flags & BNXT_RMEM_RING_PTE_FLAG))
valid_bits |= PTU_PTE_NEXT_TO_LAST;
else if (i == rmem->nr_pages - 1 &&
(rmem->flags & BNXT_RMEM_RING_PTE_FLAG))
valid_bits |= PTU_PTE_LAST;
rmem->pg_tbl[i] = rte_cpu_to_le_64(rmem->dma_arr[i] |
valid_bits);
}
}
rmem->mz = mz;
if (rmem->vmem_size)
rmem->vmem = (void **)mz->addr;
rmem->dma_arr[0] = mz_phys_addr;
return 0;
}
static void bnxt_free_ctx_mem(struct bnxt *bp)
{
int i;
if (!bp->ctx || !(bp->ctx->flags & BNXT_CTX_FLAG_INITED))
return;
bp->ctx->flags &= ~BNXT_CTX_FLAG_INITED;
rte_memzone_free(bp->ctx->qp_mem.ring_mem.mz);
rte_memzone_free(bp->ctx->srq_mem.ring_mem.mz);
rte_memzone_free(bp->ctx->cq_mem.ring_mem.mz);
rte_memzone_free(bp->ctx->vnic_mem.ring_mem.mz);
rte_memzone_free(bp->ctx->stat_mem.ring_mem.mz);
rte_memzone_free(bp->ctx->qp_mem.ring_mem.pg_tbl_mz);
rte_memzone_free(bp->ctx->srq_mem.ring_mem.pg_tbl_mz);
rte_memzone_free(bp->ctx->cq_mem.ring_mem.pg_tbl_mz);
rte_memzone_free(bp->ctx->vnic_mem.ring_mem.pg_tbl_mz);
rte_memzone_free(bp->ctx->stat_mem.ring_mem.pg_tbl_mz);
net/bnxt: fix number of TQM ring Newer firmware advertises the number of TQM rings to allocate context memory for. Use the firmware specified value and fall back to the old value derived from "bp->max_q" if it is not available. Fixes: f8168ca0e690 ("net/bnxt: support thor controller") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-05-06 06:27:09 +00:00
for (i = 0; i < bp->ctx->tqm_fp_rings_count + 1; i++) {
net/bnxt: support thor controller This commit adds support to the bnxt PMD for devices based on the BCM57508 "thor" Ethernet controller. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
2019-06-02 17:42:44 +00:00
if (bp->ctx->tqm_mem[i])
rte_memzone_free(bp->ctx->tqm_mem[i]->ring_mem.mz);
}
rte_free(bp->ctx);
bp->ctx = NULL;
}
net/bnxt: fix build with some compilers The driver was defining its own version of roundup which was conflicting with another version defined elsewhere. Change the local definition of roundup to avoid compilation errors. Fixes: f8168ca0e690 ("net/bnxt: support thor controller") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-06-14 04:20:57 +00:00
#define bnxt_roundup(x, y) ((((x) + ((y) - 1)) / (y)) * (y))
net/bnxt: support thor controller This commit adds support to the bnxt PMD for devices based on the BCM57508 "thor" Ethernet controller. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
2019-06-02 17:42:44 +00:00
#define min_t(type, x, y) ({ \
type __min1 = (x); \
type __min2 = (y); \
__min1 < __min2 ? __min1 : __min2; })
#define max_t(type, x, y) ({ \
type __max1 = (x); \
type __max2 = (y); \
__max1 > __max2 ? __max1 : __max2; })
#define clamp_t(type, _x, min, max) min_t(type, max_t(type, _x, min), max)
int bnxt_alloc_ctx_mem(struct bnxt *bp)
{
struct bnxt_ctx_pg_info *ctx_pg;
struct bnxt_ctx_mem_info *ctx;
uint32_t mem_size, ena, entries;
net/bnxt: fix TQM ring context memory size The current formulas to calculate the TQM slow path and fast path ring context memory sizes are not quite correct. TQM slow path entry is array index 0 of ctx->tqm_mem[]. The other array entries are for fast path. Fix these sizes according to firmware spec. for 57500 and newer chips. Fixes: cc5e26b8ef98 ("net/bnxt: increase TQM entry allocation") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-05-06 06:27:10 +00:00
uint32_t entries_sp, min;
net/bnxt: support thor controller This commit adds support to the bnxt PMD for devices based on the BCM57508 "thor" Ethernet controller. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
2019-06-02 17:42:44 +00:00
int i, rc;
rc = bnxt_hwrm_func_backing_store_qcaps(bp);
if (rc) {
PMD_DRV_LOG(ERR, "Query context mem capability failed\n");
return rc;
}
ctx = bp->ctx;
if (!ctx || (ctx->flags & BNXT_CTX_FLAG_INITED))
return 0;
ctx_pg = &ctx->qp_mem;
ctx_pg->entries = ctx->qp_min_qp1_entries + ctx->qp_max_l2_entries;
net/bnxt: modify context memory allocation Newer devices like SR2 may have chip backing store and do not require host backed memory allocation. In these cases, HWRM_FUNC_BACKING_STORE_QCAPS will return a zero entry size to indicate contexts for which the host should not allocate backing store. Selectively allocate context memory based on device capabilities and only enable backing store for the appropriate contexts Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-20 05:24:29 +00:00
if (ctx->qp_entry_size) {
mem_size = ctx->qp_entry_size * ctx_pg->entries;
rc = bnxt_alloc_ctx_mem_blk(bp, ctx_pg, mem_size, "qp_mem", 0);
if (rc)
return rc;
}
net/bnxt: support thor controller This commit adds support to the bnxt PMD for devices based on the BCM57508 "thor" Ethernet controller. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
2019-06-02 17:42:44 +00:00
ctx_pg = &ctx->srq_mem;
ctx_pg->entries = ctx->srq_max_l2_entries;
net/bnxt: modify context memory allocation Newer devices like SR2 may have chip backing store and do not require host backed memory allocation. In these cases, HWRM_FUNC_BACKING_STORE_QCAPS will return a zero entry size to indicate contexts for which the host should not allocate backing store. Selectively allocate context memory based on device capabilities and only enable backing store for the appropriate contexts Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-20 05:24:29 +00:00
if (ctx->srq_entry_size) {
mem_size = ctx->srq_entry_size * ctx_pg->entries;
rc = bnxt_alloc_ctx_mem_blk(bp, ctx_pg, mem_size, "srq_mem", 0);
if (rc)
return rc;
}
net/bnxt: support thor controller This commit adds support to the bnxt PMD for devices based on the BCM57508 "thor" Ethernet controller. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
2019-06-02 17:42:44 +00:00
ctx_pg = &ctx->cq_mem;
ctx_pg->entries = ctx->cq_max_l2_entries;
net/bnxt: modify context memory allocation Newer devices like SR2 may have chip backing store and do not require host backed memory allocation. In these cases, HWRM_FUNC_BACKING_STORE_QCAPS will return a zero entry size to indicate contexts for which the host should not allocate backing store. Selectively allocate context memory based on device capabilities and only enable backing store for the appropriate contexts Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-20 05:24:29 +00:00
if (ctx->cq_entry_size) {
mem_size = ctx->cq_entry_size * ctx_pg->entries;
rc = bnxt_alloc_ctx_mem_blk(bp, ctx_pg, mem_size, "cq_mem", 0);
if (rc)
return rc;
}
net/bnxt: support thor controller This commit adds support to the bnxt PMD for devices based on the BCM57508 "thor" Ethernet controller. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
2019-06-02 17:42:44 +00:00
ctx_pg = &ctx->vnic_mem;
ctx_pg->entries = ctx->vnic_max_vnic_entries +
ctx->vnic_max_ring_table_entries;
net/bnxt: modify context memory allocation Newer devices like SR2 may have chip backing store and do not require host backed memory allocation. In these cases, HWRM_FUNC_BACKING_STORE_QCAPS will return a zero entry size to indicate contexts for which the host should not allocate backing store. Selectively allocate context memory based on device capabilities and only enable backing store for the appropriate contexts Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-20 05:24:29 +00:00
if (ctx->vnic_entry_size) {
mem_size = ctx->vnic_entry_size * ctx_pg->entries;
rc = bnxt_alloc_ctx_mem_blk(bp, ctx_pg, mem_size, "vnic_mem", 0);
if (rc)
return rc;
}
net/bnxt: support thor controller This commit adds support to the bnxt PMD for devices based on the BCM57508 "thor" Ethernet controller. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
2019-06-02 17:42:44 +00:00
ctx_pg = &ctx->stat_mem;
ctx_pg->entries = ctx->stat_max_entries;
net/bnxt: modify context memory allocation Newer devices like SR2 may have chip backing store and do not require host backed memory allocation. In these cases, HWRM_FUNC_BACKING_STORE_QCAPS will return a zero entry size to indicate contexts for which the host should not allocate backing store. Selectively allocate context memory based on device capabilities and only enable backing store for the appropriate contexts Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-20 05:24:29 +00:00
if (ctx->stat_entry_size) {
mem_size = ctx->stat_entry_size * ctx_pg->entries;
rc = bnxt_alloc_ctx_mem_blk(bp, ctx_pg, mem_size, "stat_mem", 0);
if (rc)
return rc;
}
net/bnxt: support thor controller This commit adds support to the bnxt PMD for devices based on the BCM57508 "thor" Ethernet controller. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
2019-06-02 17:42:44 +00:00
net/bnxt: fix TQM ring context memory size The current formulas to calculate the TQM slow path and fast path ring context memory sizes are not quite correct. TQM slow path entry is array index 0 of ctx->tqm_mem[]. The other array entries are for fast path. Fix these sizes according to firmware spec. for 57500 and newer chips. Fixes: cc5e26b8ef98 ("net/bnxt: increase TQM entry allocation") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-05-06 06:27:10 +00:00
min = ctx->tqm_min_entries_per_ring;
entries_sp = ctx->qp_max_l2_entries +
ctx->vnic_max_vnic_entries +
2 * ctx->qp_min_qp1_entries + min;
entries_sp = bnxt_roundup(entries_sp, ctx->tqm_entries_multiple);
entries = ctx->qp_max_l2_entries + ctx->qp_min_qp1_entries;
net/bnxt: fix build with some compilers The driver was defining its own version of roundup which was conflicting with another version defined elsewhere. Change the local definition of roundup to avoid compilation errors. Fixes: f8168ca0e690 ("net/bnxt: support thor controller") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-06-14 04:20:57 +00:00
entries = bnxt_roundup(entries, ctx->tqm_entries_multiple);
net/bnxt: fix TQM ring context memory size The current formulas to calculate the TQM slow path and fast path ring context memory sizes are not quite correct. TQM slow path entry is array index 0 of ctx->tqm_mem[]. The other array entries are for fast path. Fix these sizes according to firmware spec. for 57500 and newer chips. Fixes: cc5e26b8ef98 ("net/bnxt: increase TQM entry allocation") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-05-06 06:27:10 +00:00
entries = clamp_t(uint32_t, entries, min,
net/bnxt: support thor controller This commit adds support to the bnxt PMD for devices based on the BCM57508 "thor" Ethernet controller. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
2019-06-02 17:42:44 +00:00
ctx->tqm_max_entries_per_ring);
net/bnxt: fix number of TQM ring Newer firmware advertises the number of TQM rings to allocate context memory for. Use the firmware specified value and fall back to the old value derived from "bp->max_q" if it is not available. Fixes: f8168ca0e690 ("net/bnxt: support thor controller") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-05-06 06:27:09 +00:00
for (i = 0, ena = 0; i < ctx->tqm_fp_rings_count + 1; i++) {
net/bnxt: update to new version of backing store Update HWRM headers to version 1.10.2.15 which updates the backing store API for additional TQM rings. Add support for 9th TQM ring using latest firmware interface. Also make sure that we set only necessary bits in the enables field in backing store request. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2021-02-24 15:55:48 +00:00
/* i=0 is for TQM_SP. i=1 to i=8 applies to RING0 to RING7.
* i > 8 is other ext rings.
*/
net/bnxt: support thor controller This commit adds support to the bnxt PMD for devices based on the BCM57508 "thor" Ethernet controller. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
2019-06-02 17:42:44 +00:00
ctx_pg = ctx->tqm_mem[i];
net/bnxt: fix TQM ring context memory size The current formulas to calculate the TQM slow path and fast path ring context memory sizes are not quite correct. TQM slow path entry is array index 0 of ctx->tqm_mem[]. The other array entries are for fast path. Fix these sizes according to firmware spec. for 57500 and newer chips. Fixes: cc5e26b8ef98 ("net/bnxt: increase TQM entry allocation") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-05-06 06:27:10 +00:00
ctx_pg->entries = i ? entries : entries_sp;
net/bnxt: modify context memory allocation Newer devices like SR2 may have chip backing store and do not require host backed memory allocation. In these cases, HWRM_FUNC_BACKING_STORE_QCAPS will return a zero entry size to indicate contexts for which the host should not allocate backing store. Selectively allocate context memory based on device capabilities and only enable backing store for the appropriate contexts Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-20 05:24:29 +00:00
if (ctx->tqm_entry_size) {
mem_size = ctx->tqm_entry_size * ctx_pg->entries;
net/bnxt: update to new version of backing store Update HWRM headers to version 1.10.2.15 which updates the backing store API for additional TQM rings. Add support for 9th TQM ring using latest firmware interface. Also make sure that we set only necessary bits in the enables field in backing store request. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2021-02-24 15:55:48 +00:00
rc = bnxt_alloc_ctx_mem_blk(bp, ctx_pg, mem_size,
"tqm_mem", i);
net/bnxt: modify context memory allocation Newer devices like SR2 may have chip backing store and do not require host backed memory allocation. In these cases, HWRM_FUNC_BACKING_STORE_QCAPS will return a zero entry size to indicate contexts for which the host should not allocate backing store. Selectively allocate context memory based on device capabilities and only enable backing store for the appropriate contexts Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-20 05:24:29 +00:00
if (rc)
return rc;
}
net/bnxt: update to new version of backing store Update HWRM headers to version 1.10.2.15 which updates the backing store API for additional TQM rings. Add support for 9th TQM ring using latest firmware interface. Also make sure that we set only necessary bits in the enables field in backing store request. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2021-02-24 15:55:48 +00:00
if (i < BNXT_MAX_TQM_LEGACY_RINGS)
ena |= HWRM_FUNC_BACKING_STORE_CFG_INPUT_ENABLES_TQM_SP << i;
else
ena |= HWRM_FUNC_BACKING_STORE_CFG_INPUT_ENABLES_TQM_RING8;
net/bnxt: support thor controller This commit adds support to the bnxt PMD for devices based on the BCM57508 "thor" Ethernet controller. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
2019-06-02 17:42:44 +00:00
}
ena |= FUNC_BACKING_STORE_CFG_INPUT_DFLT_ENABLES;
rc = bnxt_hwrm_func_backing_store_cfg(bp, ena);
if (rc)
PMD_DRV_LOG(ERR,
"Failed to configure context mem: rc = %d\n", rc);
else
ctx->flags |= BNXT_CTX_FLAG_INITED;
net/bnxt: fix return values to standard error codes Fixed the return values of few routines to return standard error code. Also fixed few error logs to more meaningful one. Fixes: 804e746c7b73 ("net/bnxt: add hardware resource manager init code") Fixes: e3d8f1e6a665 ("net/bnxt: cache address of doorbell to subsequent access") Fixes: 19e6af01bb36 ("net/bnxt: support get/set EEPROM") Fixes: b7435d660a8c ("net/bnxt: add ntuple filtering support") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-07-18 03:36:06 +00:00
return rc;
net/bnxt: support thor controller This commit adds support to the bnxt PMD for devices based on the BCM57508 "thor" Ethernet controller. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
2019-06-02 17:42:44 +00:00
}
drivers/net: do not use ethdev driver Signed-off-by: Jan Blunck <jblunck@infradead.org> Signed-off-by: Gaetan Rivet <gaetan.rivet@6wind.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2017-04-11 15:44:24 +00:00
net/bnxt: fix extended port counter statistics 1. refactor stats allocation code to new routine 2. check for extended statistics support depends on "hwrm_spec_code" which is set in bnxt_hwrm_ver_get called later. Hence we were never querying extended port stats as flags field was not updated. Fixed this by moving the stats allocation after the call to bnxt_hwrm_ver_get. 3. we were incorrectly passing the host address used for port statistics to PORT_QSTATS_EXT command. Fixed this by passing the correct extended stats address. Fixes: f55e12f33416 ("net/bnxt: support extended port counters") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-17 10:41:25 +00:00
static int bnxt_alloc_stats_mem(struct bnxt *bp)
{
struct rte_pci_device *pci_dev = bp->pdev;
char mz_name[RTE_MEMZONE_NAMESIZE];
const struct rte_memzone *mz = NULL;
uint32_t total_alloc_len;
rte_iova_t mz_phys_addr;
if (pci_dev->id.device_id == BROADCOM_DEV_ID_NS2)
return 0;
snprintf(mz_name, RTE_MEMZONE_NAMESIZE,
"bnxt_" PCI_PRI_FMT "-%s", pci_dev->addr.domain,
pci_dev->addr.bus, pci_dev->addr.devid,
pci_dev->addr.function, "rx_port_stats");
mz_name[RTE_MEMZONE_NAMESIZE - 1] = 0;
mz = rte_memzone_lookup(mz_name);
total_alloc_len =
RTE_CACHE_LINE_ROUNDUP(sizeof(struct rx_port_stats) +
sizeof(struct rx_port_stats_ext) + 512);
if (!mz) {
mz = rte_memzone_reserve(mz_name, total_alloc_len,
SOCKET_ID_ANY,
RTE_MEMZONE_2MB |
RTE_MEMZONE_SIZE_HINT_ONLY |
RTE_MEMZONE_IOVA_CONTIG);
if (mz == NULL)
return -ENOMEM;
}
memset(mz->addr, 0, mz->len);
mz_phys_addr = mz->iova;
bp->rx_mem_zone = (const void *)mz;
bp->hw_rx_port_stats = mz->addr;
bp->hw_rx_port_stats_map = mz_phys_addr;
snprintf(mz_name, RTE_MEMZONE_NAMESIZE,
"bnxt_" PCI_PRI_FMT "-%s", pci_dev->addr.domain,
pci_dev->addr.bus, pci_dev->addr.devid,
pci_dev->addr.function, "tx_port_stats");
mz_name[RTE_MEMZONE_NAMESIZE - 1] = 0;
mz = rte_memzone_lookup(mz_name);
total_alloc_len =
RTE_CACHE_LINE_ROUNDUP(sizeof(struct tx_port_stats) +
sizeof(struct tx_port_stats_ext) + 512);
if (!mz) {
mz = rte_memzone_reserve(mz_name,
total_alloc_len,
SOCKET_ID_ANY,
RTE_MEMZONE_2MB |
RTE_MEMZONE_SIZE_HINT_ONLY |
RTE_MEMZONE_IOVA_CONTIG);
if (mz == NULL)
return -ENOMEM;
}
memset(mz->addr, 0, mz->len);
mz_phys_addr = mz->iova;
bp->tx_mem_zone = (const void *)mz;
bp->hw_tx_port_stats = mz->addr;
bp->hw_tx_port_stats_map = mz_phys_addr;
bp->flags |= BNXT_FLAG_PORT_STATS;
/* Display extended statistics if FW supports it */
if (bp->hwrm_spec_code < HWRM_SPEC_CODE_1_8_4 ||
bp->hwrm_spec_code == HWRM_SPEC_CODE_1_9_0 ||
!(bp->flags & BNXT_FLAG_EXT_STATS_SUPPORTED))
return 0;
bp->hw_rx_port_stats_ext = (void *)
((uint8_t *)bp->hw_rx_port_stats +
sizeof(struct rx_port_stats));
bp->hw_rx_port_stats_ext_map = bp->hw_rx_port_stats_map +
sizeof(struct rx_port_stats);
bp->flags |= BNXT_FLAG_EXT_RX_PORT_STATS;
if (bp->hwrm_spec_code < HWRM_SPEC_CODE_1_9_2 ||
bp->flags & BNXT_FLAG_EXT_STATS_SUPPORTED) {
bp->hw_tx_port_stats_ext = (void *)
((uint8_t *)bp->hw_tx_port_stats +
sizeof(struct tx_port_stats));
bp->hw_tx_port_stats_ext_map =
bp->hw_tx_port_stats_map +
sizeof(struct tx_port_stats);
bp->flags |= BNXT_FLAG_EXT_TX_PORT_STATS;
}
return 0;
}
net/bnxt: fix VF probe when MAC address is zero VF driver should not fail probe if the host PF driver has not assigned any MAC address for the VF. It should generate a random MAC address and configure the MAC and then continue probing the device. Fixes: be160484a48d ("net/bnxt: check if MAC address is all zeros") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-17 10:41:27 +00:00
static int bnxt_setup_mac_addr(struct rte_eth_dev *eth_dev)
{
struct bnxt *bp = eth_dev->data->dev_private;
int rc = 0;
eth_dev->data->mac_addrs = rte_zmalloc("bnxt_mac_addr_tbl",
RTE_ETHER_ADDR_LEN *
bp->max_l2_ctx,
0);
if (eth_dev->data->mac_addrs == NULL) {
PMD_DRV_LOG(ERR, "Failed to alloc MAC addr tbl\n");
return -ENOMEM;
}
net/bnxt: fix storing MAC address twice We are maintaining MAC address twice which is unnecessary. This patch removes the copy which helps reduce size of bnxt structure. Fixes: b7778e8a1c00 ("net/bnxt: refactor to properly allocate resources for PF/VF") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-05-15 05:04:36 +00:00
if (!BNXT_HAS_DFLT_MAC_SET(bp)) {
net/bnxt: fix VF probe when MAC address is zero VF driver should not fail probe if the host PF driver has not assigned any MAC address for the VF. It should generate a random MAC address and configure the MAC and then continue probing the device. Fixes: be160484a48d ("net/bnxt: check if MAC address is all zeros") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-17 10:41:27 +00:00
if (BNXT_PF(bp))
return -EINVAL;
/* Generate a random MAC address, if none was assigned by PF */
PMD_DRV_LOG(INFO, "VF MAC address not assigned by Host PF\n");
bnxt_eth_hw_addr_random(bp->mac_addr);
PMD_DRV_LOG(INFO,
"Assign random MAC:%02X:%02X:%02X:%02X:%02X:%02X\n",
bp->mac_addr[0], bp->mac_addr[1], bp->mac_addr[2],
bp->mac_addr[3], bp->mac_addr[4], bp->mac_addr[5]);
rc = bnxt_hwrm_set_mac(bp);
net/bnxt: fix storing MAC address twice We are maintaining MAC address twice which is unnecessary. This patch removes the copy which helps reduce size of bnxt structure. Fixes: b7778e8a1c00 ("net/bnxt: refactor to properly allocate resources for PF/VF") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-05-15 05:04:36 +00:00
if (rc)
return rc;
net/bnxt: fix VF probe when MAC address is zero VF driver should not fail probe if the host PF driver has not assigned any MAC address for the VF. It should generate a random MAC address and configure the MAC and then continue probing the device. Fixes: be160484a48d ("net/bnxt: check if MAC address is all zeros") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-17 10:41:27 +00:00
}
/* Copy the permanent MAC from the FUNC_QCAPS response */
memcpy(&eth_dev->data->mac_addrs[0], bp->mac_addr, RTE_ETHER_ADDR_LEN);
return rc;
}
net/bnxt: handle reset notify async event from FW When the FW upgrade is initiated the current instance of FW issues a HWRM_ASYNC_EVENT_CMPL_EVENT_ID_RESET_NOTIFY async notification to the driver. On receiving this notification, the PMD shall quiesce itself and poll on the HWRM_VER_GET FW command at regular intervals. Once the VER_GET command succeeds, the driver should go through the rediscovery process and re-initialize the device. Also register with FW for the reset notify async event. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:23 +00:00
static int bnxt_restore_dflt_mac(struct bnxt *bp)
{
int rc = 0;
/* MAC is already configured in FW */
net/bnxt: fix storing MAC address twice We are maintaining MAC address twice which is unnecessary. This patch removes the copy which helps reduce size of bnxt structure. Fixes: b7778e8a1c00 ("net/bnxt: refactor to properly allocate resources for PF/VF") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-05-15 05:04:36 +00:00
if (BNXT_HAS_DFLT_MAC_SET(bp))
net/bnxt: handle reset notify async event from FW When the FW upgrade is initiated the current instance of FW issues a HWRM_ASYNC_EVENT_CMPL_EVENT_ID_RESET_NOTIFY async notification to the driver. On receiving this notification, the PMD shall quiesce itself and poll on the HWRM_VER_GET FW command at regular intervals. Once the VER_GET command succeeds, the driver should go through the rediscovery process and re-initialize the device. Also register with FW for the reset notify async event. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:23 +00:00
return 0;
/* Restore the old MAC configured */
rc = bnxt_hwrm_set_mac(bp);
if (rc)
PMD_DRV_LOG(ERR, "Failed to restore MAC address\n");
return rc;
}
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
static void bnxt_config_vf_req_fwd(struct bnxt *bp)
{
if (!BNXT_PF(bp))
return;
net/bnxt: fix PF support in SR-IOV mode 1. Implement HWRM_FUNC_VF_RESOURCE_CFG command and use it to reserve resources for VFs when NEW RM is enabled. 2. Invoke PF’s FUNC_CFG before configuring VFs resources. 3. Don’t consider max_rx_em_flows in max_l2_ctx calculation when VFs are configured. 4. Issue HWRM_FUNC_QCFG instead of HWRM_FUNC_QCAPS to find out the actual allocated resources for VF. 5. Don’t add random mac to the VF. 6. Handle completion type CMPL_BASE_TYPE_HWRM_FWD_REQ instead of CMPL_BASE_TYPE_HWRM_FWD_RESP. 7. Don't enable HWRM_FUNC_DRV_RGTR_INPUT_FLAGS_FWD_NONE_MODE when the list of HWRM commands that needs to be forwarded to the PF is specified in HWRM_FUNC_DRV_RGTR. 8. Update the HWRM commands list that can be forwarded to the PF. Fixes: b7778e8a1c00 ("net/bnxt: refactor to properly allocate resources for PF/VF") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-10-09 11:11:21 +00:00
memset(bp->pf->vf_req_fwd, 0, sizeof(bp->pf->vf_req_fwd));
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
net/bnxt: fix PF support in SR-IOV mode 1. Implement HWRM_FUNC_VF_RESOURCE_CFG command and use it to reserve resources for VFs when NEW RM is enabled. 2. Invoke PF’s FUNC_CFG before configuring VFs resources. 3. Don’t consider max_rx_em_flows in max_l2_ctx calculation when VFs are configured. 4. Issue HWRM_FUNC_QCFG instead of HWRM_FUNC_QCAPS to find out the actual allocated resources for VF. 5. Don’t add random mac to the VF. 6. Handle completion type CMPL_BASE_TYPE_HWRM_FWD_REQ instead of CMPL_BASE_TYPE_HWRM_FWD_RESP. 7. Don't enable HWRM_FUNC_DRV_RGTR_INPUT_FLAGS_FWD_NONE_MODE when the list of HWRM commands that needs to be forwarded to the PF is specified in HWRM_FUNC_DRV_RGTR. 8. Update the HWRM commands list that can be forwarded to the PF. Fixes: b7778e8a1c00 ("net/bnxt: refactor to properly allocate resources for PF/VF") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-10-09 11:11:21 +00:00
if (!(bp->fw_cap & BNXT_FW_CAP_LINK_ADMIN))
BNXT_HWRM_CMD_TO_FORWARD(HWRM_PORT_PHY_QCFG);
BNXT_HWRM_CMD_TO_FORWARD(HWRM_FUNC_CFG);
BNXT_HWRM_CMD_TO_FORWARD(HWRM_FUNC_VF_CFG);
BNXT_HWRM_CMD_TO_FORWARD(HWRM_CFA_L2_FILTER_ALLOC);
BNXT_HWRM_CMD_TO_FORWARD(HWRM_OEM_CMD);
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
}
net/bnxt: fetch SVIF information from firmware SVIF (source virtual interface) is used to represent a physical port, physical function, or a virtual function. SVIF is compared during L2 context and exact match lookups in TX direction. SVIF is masked for port information during L2 context and exact match lookup in RX direction. Hence, driver needs this SVIF information to program L2 context and Exact match tables. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-04-15 08:18:50 +00:00
uint16_t
net/bnxt: enhance port DB 1. Add "enum bnxt_ulp_intf_type” as the second parameter for the port & func helper functions 2. Return vfrep related port & func information in the helper functions 3. Allocate phy_port_list dynamically based on port count 4. Introduce ulp_func_id_tbl array for book keeping func related information indexed by func_id Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:28:28 +00:00
bnxt_get_svif(uint16_t port_id, bool func_svif,
enum bnxt_ulp_intf_type type)
net/bnxt: fetch SVIF information from firmware SVIF (source virtual interface) is used to represent a physical port, physical function, or a virtual function. SVIF is compared during L2 context and exact match lookups in TX direction. SVIF is masked for port information during L2 context and exact match lookup in RX direction. Hence, driver needs this SVIF information to program L2 context and Exact match tables. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-04-15 08:18:50 +00:00
{
struct rte_eth_dev *eth_dev;
struct bnxt *bp;
eth_dev = &rte_eth_devices[port_id];
net/bnxt: enhance port DB 1. Add "enum bnxt_ulp_intf_type” as the second parameter for the port & func helper functions 2. Return vfrep related port & func information in the helper functions 3. Allocate phy_port_list dynamically based on port count 4. Introduce ulp_func_id_tbl array for book keeping func related information indexed by func_id Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:28:28 +00:00
if (BNXT_ETH_DEV_IS_REPRESENTOR(eth_dev)) {
net/bnxt: support representors on remote host domain In the Stingray use case, representors are conventionally run inside the SoC domain representing functions that are on the X86 domain. In order to support this mechanism of building representors for endpoints that are not in the same host domain, additional dev args have been in the PMD like so: rep-based-pf=<physical index> rep-is-pf=<VF=0 or PF=1> where `rep-based-pf` specifies the physical index of the base PF that the representor is derived off of. Since representor(s) can be created for endpoint PFs as well, rename struct bnxt_vf_representor to bnxt_representor and other such dev_ops and function names. devargs have also been extended to specify the exact CoS queue along with flow control enablement to be used for the conduit between the representor and the endpoint function. This is how a sample devargs would look with all the extended devargs -w 0000:06:02.0,host-based-truflow=1,representor=[1],rep-based-pf=8, rep-is-pf=1,rep-q-r2f=1,rep-fc-r2f=0,rep-q-f2r=1,rep-fc-f2r=1 Call CFA_PAIR_ALLOC only in case of Stingray instead of CFA_VFR_ALLOC. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:30 +00:00
struct bnxt_representor *vfr = eth_dev->data->dev_private;
net/bnxt: enhance port DB 1. Add "enum bnxt_ulp_intf_type” as the second parameter for the port & func helper functions 2. Return vfrep related port & func information in the helper functions 3. Allocate phy_port_list dynamically based on port count 4. Introduce ulp_func_id_tbl array for book keeping func related information indexed by func_id Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:28:28 +00:00
if (!vfr)
return 0;
if (type == BNXT_ULP_INTF_TYPE_VF_REP)
return vfr->svif;
eth_dev = vfr->parent_dev;
}
net/bnxt: fetch SVIF information from firmware SVIF (source virtual interface) is used to represent a physical port, physical function, or a virtual function. SVIF is compared during L2 context and exact match lookups in TX direction. SVIF is masked for port information during L2 context and exact match lookup in RX direction. Hence, driver needs this SVIF information to program L2 context and Exact match tables. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-04-15 08:18:50 +00:00
bp = eth_dev->data->dev_private;
return func_svif ? bp->func_svif : bp->port_svif;
}
net/bnxt: fetch VNIC info VNIC is needed for the driver to program the action record for rx flows. VNIC determines what receive rings to use to place the received packets. This patch introduces a routine that will convert a given dpdk port to VNIC. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-04-15 08:18:51 +00:00
uint16_t
net/bnxt: enhance port DB 1. Add "enum bnxt_ulp_intf_type” as the second parameter for the port & func helper functions 2. Return vfrep related port & func information in the helper functions 3. Allocate phy_port_list dynamically based on port count 4. Introduce ulp_func_id_tbl array for book keeping func related information indexed by func_id Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:28:28 +00:00
bnxt_get_vnic_id(uint16_t port, enum bnxt_ulp_intf_type type)
net/bnxt: fetch VNIC info VNIC is needed for the driver to program the action record for rx flows. VNIC determines what receive rings to use to place the received packets. This patch introduces a routine that will convert a given dpdk port to VNIC. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-04-15 08:18:51 +00:00
{
struct rte_eth_dev *eth_dev;
struct bnxt_vnic_info *vnic;
struct bnxt *bp;
eth_dev = &rte_eth_devices[port];
net/bnxt: enhance port DB 1. Add "enum bnxt_ulp_intf_type” as the second parameter for the port & func helper functions 2. Return vfrep related port & func information in the helper functions 3. Allocate phy_port_list dynamically based on port count 4. Introduce ulp_func_id_tbl array for book keeping func related information indexed by func_id Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:28:28 +00:00
if (BNXT_ETH_DEV_IS_REPRESENTOR(eth_dev)) {
net/bnxt: support representors on remote host domain In the Stingray use case, representors are conventionally run inside the SoC domain representing functions that are on the X86 domain. In order to support this mechanism of building representors for endpoints that are not in the same host domain, additional dev args have been in the PMD like so: rep-based-pf=<physical index> rep-is-pf=<VF=0 or PF=1> where `rep-based-pf` specifies the physical index of the base PF that the representor is derived off of. Since representor(s) can be created for endpoint PFs as well, rename struct bnxt_vf_representor to bnxt_representor and other such dev_ops and function names. devargs have also been extended to specify the exact CoS queue along with flow control enablement to be used for the conduit between the representor and the endpoint function. This is how a sample devargs would look with all the extended devargs -w 0000:06:02.0,host-based-truflow=1,representor=[1],rep-based-pf=8, rep-is-pf=1,rep-q-r2f=1,rep-fc-r2f=0,rep-q-f2r=1,rep-fc-f2r=1 Call CFA_PAIR_ALLOC only in case of Stingray instead of CFA_VFR_ALLOC. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:30 +00:00
struct bnxt_representor *vfr = eth_dev->data->dev_private;
net/bnxt: enhance port DB 1. Add "enum bnxt_ulp_intf_type” as the second parameter for the port & func helper functions 2. Return vfrep related port & func information in the helper functions 3. Allocate phy_port_list dynamically based on port count 4. Introduce ulp_func_id_tbl array for book keeping func related information indexed by func_id Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:28:28 +00:00
if (!vfr)
return 0;
if (type == BNXT_ULP_INTF_TYPE_VF_REP)
return vfr->dflt_vnic_id;
eth_dev = vfr->parent_dev;
}
net/bnxt: fetch VNIC info VNIC is needed for the driver to program the action record for rx flows. VNIC determines what receive rings to use to place the received packets. This patch introduces a routine that will convert a given dpdk port to VNIC. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-04-15 08:18:51 +00:00
bp = eth_dev->data->dev_private;
vnic = BNXT_GET_DEFAULT_VNIC(bp);
return vnic->fw_vnic_id;
}
net/bnxt: add session and function flow flush The ulp flow flush has been extended to support session flow flush and function flow flush. The session flow flush is called when there the device is sole owner of the session and it deletes all the flows associated with that session. The function flow flush is called if the device function is not the sole owner of the session, it deletes all the flows that are associated with that device function. Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
2020-04-15 14:49:13 +00:00
uint16_t
net/bnxt: enhance port DB 1. Add "enum bnxt_ulp_intf_type” as the second parameter for the port & func helper functions 2. Return vfrep related port & func information in the helper functions 3. Allocate phy_port_list dynamically based on port count 4. Introduce ulp_func_id_tbl array for book keeping func related information indexed by func_id Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:28:28 +00:00
bnxt_get_fw_func_id(uint16_t port, enum bnxt_ulp_intf_type type)
net/bnxt: add session and function flow flush The ulp flow flush has been extended to support session flow flush and function flow flush. The session flow flush is called when there the device is sole owner of the session and it deletes all the flows associated with that session. The function flow flush is called if the device function is not the sole owner of the session, it deletes all the flows that are associated with that device function. Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
2020-04-15 14:49:13 +00:00
{
struct rte_eth_dev *eth_dev;
struct bnxt *bp;
eth_dev = &rte_eth_devices[port];
net/bnxt: enhance port DB 1. Add "enum bnxt_ulp_intf_type” as the second parameter for the port & func helper functions 2. Return vfrep related port & func information in the helper functions 3. Allocate phy_port_list dynamically based on port count 4. Introduce ulp_func_id_tbl array for book keeping func related information indexed by func_id Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:28:28 +00:00
if (BNXT_ETH_DEV_IS_REPRESENTOR(eth_dev)) {
net/bnxt: support representors on remote host domain In the Stingray use case, representors are conventionally run inside the SoC domain representing functions that are on the X86 domain. In order to support this mechanism of building representors for endpoints that are not in the same host domain, additional dev args have been in the PMD like so: rep-based-pf=<physical index> rep-is-pf=<VF=0 or PF=1> where `rep-based-pf` specifies the physical index of the base PF that the representor is derived off of. Since representor(s) can be created for endpoint PFs as well, rename struct bnxt_vf_representor to bnxt_representor and other such dev_ops and function names. devargs have also been extended to specify the exact CoS queue along with flow control enablement to be used for the conduit between the representor and the endpoint function. This is how a sample devargs would look with all the extended devargs -w 0000:06:02.0,host-based-truflow=1,representor=[1],rep-based-pf=8, rep-is-pf=1,rep-q-r2f=1,rep-fc-r2f=0,rep-q-f2r=1,rep-fc-f2r=1 Call CFA_PAIR_ALLOC only in case of Stingray instead of CFA_VFR_ALLOC. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:30 +00:00
struct bnxt_representor *vfr = eth_dev->data->dev_private;
net/bnxt: enhance port DB 1. Add "enum bnxt_ulp_intf_type” as the second parameter for the port & func helper functions 2. Return vfrep related port & func information in the helper functions 3. Allocate phy_port_list dynamically based on port count 4. Introduce ulp_func_id_tbl array for book keeping func related information indexed by func_id Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:28:28 +00:00
if (!vfr)
return 0;
if (type == BNXT_ULP_INTF_TYPE_VF_REP)
return vfr->fw_fid;
eth_dev = vfr->parent_dev;
}
net/bnxt: add session and function flow flush The ulp flow flush has been extended to support session flow flush and function flow flush. The session flow flush is called when there the device is sole owner of the session and it deletes all the flows associated with that session. The function flow flush is called if the device function is not the sole owner of the session, it deletes all the flows that are associated with that device function. Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
2020-04-15 14:49:13 +00:00
bp = eth_dev->data->dev_private;
return bp->fw_fid;
}
net/bnxt: get port and function info add helper functions to get port & function related information like parif, physical port id & vport id. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:53 +00:00
enum bnxt_ulp_intf_type
bnxt_get_interface_type(uint16_t port)
{
struct rte_eth_dev *eth_dev;
struct bnxt *bp;
eth_dev = &rte_eth_devices[port];
if (BNXT_ETH_DEV_IS_REPRESENTOR(eth_dev))
return BNXT_ULP_INTF_TYPE_VF_REP;
bp = eth_dev->data->dev_private;
net/bnxt: enhance port DB 1. Add "enum bnxt_ulp_intf_type” as the second parameter for the port & func helper functions 2. Return vfrep related port & func information in the helper functions 3. Allocate phy_port_list dynamically based on port count 4. Introduce ulp_func_id_tbl array for book keeping func related information indexed by func_id Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:28:28 +00:00
if (BNXT_PF(bp))
return BNXT_ULP_INTF_TYPE_PF;
else if (BNXT_VF_IS_TRUSTED(bp))
return BNXT_ULP_INTF_TYPE_TRUSTED_VF;
else if (BNXT_VF(bp))
return BNXT_ULP_INTF_TYPE_VF;
return BNXT_ULP_INTF_TYPE_INVALID;
net/bnxt: get port and function info add helper functions to get port & function related information like parif, physical port id & vport id. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:53 +00:00
}
uint16_t
bnxt_get_phy_port_id(uint16_t port_id)
{
net/bnxt: support representors on remote host domain In the Stingray use case, representors are conventionally run inside the SoC domain representing functions that are on the X86 domain. In order to support this mechanism of building representors for endpoints that are not in the same host domain, additional dev args have been in the PMD like so: rep-based-pf=<physical index> rep-is-pf=<VF=0 or PF=1> where `rep-based-pf` specifies the physical index of the base PF that the representor is derived off of. Since representor(s) can be created for endpoint PFs as well, rename struct bnxt_vf_representor to bnxt_representor and other such dev_ops and function names. devargs have also been extended to specify the exact CoS queue along with flow control enablement to be used for the conduit between the representor and the endpoint function. This is how a sample devargs would look with all the extended devargs -w 0000:06:02.0,host-based-truflow=1,representor=[1],rep-based-pf=8, rep-is-pf=1,rep-q-r2f=1,rep-fc-r2f=0,rep-q-f2r=1,rep-fc-f2r=1 Call CFA_PAIR_ALLOC only in case of Stingray instead of CFA_VFR_ALLOC. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:30 +00:00
struct bnxt_representor *vfr;
net/bnxt: get port and function info add helper functions to get port & function related information like parif, physical port id & vport id. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:53 +00:00
struct rte_eth_dev *eth_dev;
struct bnxt *bp;
eth_dev = &rte_eth_devices[port_id];
if (BNXT_ETH_DEV_IS_REPRESENTOR(eth_dev)) {
vfr = eth_dev->data->dev_private;
net/bnxt: enhance port DB 1. Add "enum bnxt_ulp_intf_type” as the second parameter for the port & func helper functions 2. Return vfrep related port & func information in the helper functions 3. Allocate phy_port_list dynamically based on port count 4. Introduce ulp_func_id_tbl array for book keeping func related information indexed by func_id Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:28:28 +00:00
if (!vfr)
return 0;
net/bnxt: get port and function info add helper functions to get port & function related information like parif, physical port id & vport id. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:53 +00:00
eth_dev = vfr->parent_dev;
}
bp = eth_dev->data->dev_private;
return BNXT_PF(bp) ? bp->pf->port_id : bp->parent->port_id;
}
uint16_t
net/bnxt: enhance port DB 1. Add "enum bnxt_ulp_intf_type” as the second parameter for the port & func helper functions 2. Return vfrep related port & func information in the helper functions 3. Allocate phy_port_list dynamically based on port count 4. Introduce ulp_func_id_tbl array for book keeping func related information indexed by func_id Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:28:28 +00:00
bnxt_get_parif(uint16_t port_id, enum bnxt_ulp_intf_type type)
net/bnxt: get port and function info add helper functions to get port & function related information like parif, physical port id & vport id. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:53 +00:00
{
struct rte_eth_dev *eth_dev;
struct bnxt *bp;
eth_dev = &rte_eth_devices[port_id];
if (BNXT_ETH_DEV_IS_REPRESENTOR(eth_dev)) {
net/bnxt: support representors on remote host domain In the Stingray use case, representors are conventionally run inside the SoC domain representing functions that are on the X86 domain. In order to support this mechanism of building representors for endpoints that are not in the same host domain, additional dev args have been in the PMD like so: rep-based-pf=<physical index> rep-is-pf=<VF=0 or PF=1> where `rep-based-pf` specifies the physical index of the base PF that the representor is derived off of. Since representor(s) can be created for endpoint PFs as well, rename struct bnxt_vf_representor to bnxt_representor and other such dev_ops and function names. devargs have also been extended to specify the exact CoS queue along with flow control enablement to be used for the conduit between the representor and the endpoint function. This is how a sample devargs would look with all the extended devargs -w 0000:06:02.0,host-based-truflow=1,representor=[1],rep-based-pf=8, rep-is-pf=1,rep-q-r2f=1,rep-fc-r2f=0,rep-q-f2r=1,rep-fc-f2r=1 Call CFA_PAIR_ALLOC only in case of Stingray instead of CFA_VFR_ALLOC. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:30 +00:00
struct bnxt_representor *vfr = eth_dev->data->dev_private;
net/bnxt: enhance port DB 1. Add "enum bnxt_ulp_intf_type” as the second parameter for the port & func helper functions 2. Return vfrep related port & func information in the helper functions 3. Allocate phy_port_list dynamically based on port count 4. Introduce ulp_func_id_tbl array for book keeping func related information indexed by func_id Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:28:28 +00:00
if (!vfr)
return 0;
if (type == BNXT_ULP_INTF_TYPE_VF_REP)
return vfr->fw_fid - 1;
net/bnxt: get port and function info add helper functions to get port & function related information like parif, physical port id & vport id. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:53 +00:00
eth_dev = vfr->parent_dev;
}
bp = eth_dev->data->dev_private;
return BNXT_PF(bp) ? bp->fw_fid - 1 : bp->parent->fid - 1;
}
uint16_t
bnxt_get_vport(uint16_t port_id)
{
return (1 << bnxt_get_phy_port_id(port_id));
}
net/bnxt: log firmware status on early init failure Firmware reports any fatal error (either ASIC or Firmware related) via a new status register. This status register can provide more detailed information about the firmware errors, especially if error occurs before HWRM_VER_GET is issued. Attempt to map this register if it is present and check for firmware status when VER_GET command fails. Refactored the code to allocate the "bp->recovery_info" structure in bnxt_init_fw() instead of doing in bnxt_hwrm_error_recovery_qcfg(). Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-04-21 21:33:50 +00:00
static void bnxt_alloc_error_recovery_info(struct bnxt *bp)
{
struct bnxt_error_recovery_info *info = bp->recovery_info;
if (info) {
if (!(bp->fw_cap & BNXT_FW_CAP_HCOMM_FW_STATUS))
memset(info, 0, sizeof(*info));
return;
}
if (!(bp->fw_cap & BNXT_FW_CAP_ERROR_RECOVERY))
return;
info = rte_zmalloc("bnxt_hwrm_error_recovery_qcfg",
sizeof(*info), 0);
if (!info)
bp->fw_cap &= ~BNXT_FW_CAP_ERROR_RECOVERY;
bp->recovery_info = info;
}
static void bnxt_check_fw_status(struct bnxt *bp)
{
uint32_t fw_status;
if (!(bp->recovery_info &&
(bp->fw_cap & BNXT_FW_CAP_HCOMM_FW_STATUS)))
return;
fw_status = bnxt_read_fw_status_reg(bp, BNXT_FW_STATUS_REG);
if (fw_status != BNXT_FW_STATUS_HEALTHY)
PMD_DRV_LOG(ERR, "Firmware not responding, status: %#x\n",
fw_status);
}
static int bnxt_map_hcomm_fw_status_reg(struct bnxt *bp)
{
struct bnxt_error_recovery_info *info = bp->recovery_info;
uint32_t status_loc;
uint32_t sig_ver;
rte_write32(HCOMM_STATUS_STRUCT_LOC, (uint8_t *)bp->bar0 +
BNXT_GRCPF_REG_WINDOW_BASE_OUT + 4);
sig_ver = rte_le_to_cpu_32(rte_read32((uint8_t *)bp->bar0 +
BNXT_GRCP_WINDOW_2_BASE +
offsetof(struct hcomm_status,
sig_ver)));
/* If the signature is absent, then FW does not support this feature */
if ((sig_ver & HCOMM_STATUS_SIGNATURE_MASK) !=
HCOMM_STATUS_SIGNATURE_VAL)
return 0;
if (!info) {
info = rte_zmalloc("bnxt_hwrm_error_recovery_qcfg",
sizeof(*info), 0);
if (!info)
return -ENOMEM;
bp->recovery_info = info;
} else {
memset(info, 0, sizeof(*info));
}
status_loc = rte_le_to_cpu_32(rte_read32((uint8_t *)bp->bar0 +
BNXT_GRCP_WINDOW_2_BASE +
offsetof(struct hcomm_status,
fw_status_loc)));
/* Only pre-map the FW health status GRC register */
if (BNXT_FW_STATUS_REG_TYPE(status_loc) != BNXT_FW_STATUS_REG_TYPE_GRC)
return 0;
info->status_regs[BNXT_FW_STATUS_REG] = status_loc;
info->mapped_status_regs[BNXT_FW_STATUS_REG] =
BNXT_GRCP_WINDOW_2_BASE + (status_loc & BNXT_GRCP_OFFSET_MASK);
rte_write32((status_loc & BNXT_GRCP_BASE_MASK), (uint8_t *)bp->bar0 +
BNXT_GRCPF_REG_WINDOW_BASE_OUT + 4);
bp->fw_cap |= BNXT_FW_CAP_HCOMM_FW_STATUS;
return 0;
}
net/bnxt: refactor init/uninit Move all the individual driver fields allocation routines to one routine - bnxt_drv_init(). This houses all such routines where memory needs to be allocated once during the driver's lifetime and does not need to be torn down during error recovery. Rename some function names in accordance with their functionality. bnxt_init_board() is doing nothing more than mapping the PCI bars, so rename it as such. Given that there is a bnxt_shutdown_nic that is called in dev_stop_op, rename it's counterpart - bnxt_init_chip() that is called in dev_start_op, to bnxt_start_nic. Also helps avoid confusion with some of the other bnxt_init_xxx routines. Rename bnxt_init_fw() to bnxt_get_config() as that is what that routine is doing mostly functionality wise. Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-01-14 02:19:14 +00:00
/* This function gets the FW version along with the
* capabilities(MAX and current) of the function, vnic,
* error recovery, phy and other chip related info
*/
static int bnxt_get_config(struct bnxt *bp)
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
{
uint16_t mtu;
int rc = 0;
net/bnxt: add a field for FW capabilities The purpose of this patch is used to add a new field "fw_cap" in bnxt structure. The new "fw_cap" field store all the capabilities reported by the FW. This information will help us to determine the device capabilities when registering feature support with FW. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-12-10 15:05:29 +00:00
bp->fw_cap = 0;
net/bnxt: log firmware status on early init failure Firmware reports any fatal error (either ASIC or Firmware related) via a new status register. This status register can provide more detailed information about the firmware errors, especially if error occurs before HWRM_VER_GET is issued. Attempt to map this register if it is present and check for firmware status when VER_GET command fails. Refactored the code to allocate the "bp->recovery_info" structure in bnxt_init_fw() instead of doing in bnxt_hwrm_error_recovery_qcfg(). Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-04-21 21:33:50 +00:00
rc = bnxt_map_hcomm_fw_status_reg(bp);
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
if (rc)
return rc;
net/bnxt: log firmware status on early init failure Firmware reports any fatal error (either ASIC or Firmware related) via a new status register. This status register can provide more detailed information about the firmware errors, especially if error occurs before HWRM_VER_GET is issued. Attempt to map this register if it is present and check for firmware status when VER_GET command fails. Refactored the code to allocate the "bp->recovery_info" structure in bnxt_init_fw() instead of doing in bnxt_hwrm_error_recovery_qcfg(). Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-04-21 21:33:50 +00:00
rc = bnxt_hwrm_ver_get(bp, DFLT_HWRM_CMD_TIMEOUT);
if (rc) {
bnxt_check_fw_status(bp);
return rc;
}
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
rc = bnxt_hwrm_func_reset(bp);
if (rc)
return -EIO;
net/bnxt: support CoS classification Class of Service (CoS) is a way to manage multiple types of traffic over a network to offer different types of services to applications. CoS classification (priority to cosqueue) is determined by the user and configured through the PF driver. DPDK driver queries this configuration and maps the cos queue ids to different VNICs. This patch adds this support. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-04 03:48:58 +00:00
rc = bnxt_hwrm_vnic_qcaps(bp);
net/bnxt: support creating SMAC and inner DMAC filters We are currently creating only outer DMAC filters. Create SMAC and inner DMAC filters using HWRM_CFA_L2_FILTER_ALLOC. For this the HWRM_CFA_L2_FILTER_ALLOC has already been updated. Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Rahul Gupta <rahul.gupta@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 23:25:45 +00:00
if (rc)
return rc;
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
rc = bnxt_hwrm_queue_qportcfg(bp);
if (rc)
return rc;
net/bnxt: support CoS classification Class of Service (CoS) is a way to manage multiple types of traffic over a network to offer different types of services to applications. CoS classification (priority to cosqueue) is determined by the user and configured through the PF driver. DPDK driver queries this configuration and maps the cos queue ids to different VNICs. This patch adds this support. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-04 03:48:58 +00:00
/* Get the MAX capabilities for this function.
* This function also allocates context memory for TQM rings and
* informs the firmware about this allocated backing store memory.
*/
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
rc = bnxt_hwrm_func_qcaps(bp);
if (rc)
return rc;
net/bnxt: support CoS classification Class of Service (CoS) is a way to manage multiple types of traffic over a network to offer different types of services to applications. CoS classification (priority to cosqueue) is determined by the user and configured through the PF driver. DPDK driver queries this configuration and maps the cos queue ids to different VNICs. This patch adds this support. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-04 03:48:58 +00:00
rc = bnxt_hwrm_func_qcfg(bp, &mtu);
net/bnxt: support LRO on Thor adapters Add support for LRO for adapters based on Thor (BCM57508). Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-04 03:48:57 +00:00
if (rc)
return rc;
net/bnxt: modify VNIC accounting Modify VNIC accounting when enabling RFS on newer chips. Unlike legacy chips, newer chips don't need additional VNIC resources for ntuple filter. Fix the code accordingly so that we don't reserve and allocate additional VNICs on newer chips. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-20 05:24:27 +00:00
rc = bnxt_hwrm_cfa_adv_flow_mgmt_qcaps(bp);
if (rc)
return rc;
net/bnxt: fetch SVIF information from firmware SVIF (source virtual interface) is used to represent a physical port, physical function, or a virtual function. SVIF is compared during L2 context and exact match lookups in TX direction. SVIF is masked for port information during L2 context and exact match lookup in RX direction. Hence, driver needs this SVIF information to program L2 context and Exact match tables. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-04-15 08:18:50 +00:00
bnxt_hwrm_port_mac_qcfg(bp);
net/bnxt: initialize parent PF information Add support to query parent PF information (MAC address, function ID, port ID and default VNIC) from firmware. Current firmware returns zero for parent default vnic, a temporary Wh+-specific workaround is included until that can be fixed. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:51 +00:00
bnxt_hwrm_parent_pf_qcfg(bp);
net/bnxt: support HWRM port PHY qcaps Issue HWRM_PORT_PHY_QCAPS to the firmware to get the physical port count of the device. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:54 +00:00
bnxt_hwrm_port_phy_qcaps(bp);
net/bnxt: log firmware status on early init failure Firmware reports any fatal error (either ASIC or Firmware related) via a new status register. This status register can provide more detailed information about the firmware errors, especially if error occurs before HWRM_VER_GET is issued. Attempt to map this register if it is present and check for firmware status when VER_GET command fails. Refactored the code to allocate the "bp->recovery_info" structure in bnxt_init_fw() instead of doing in bnxt_hwrm_error_recovery_qcfg(). Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-04-21 21:33:50 +00:00
bnxt_alloc_error_recovery_info(bp);
net/bnxt: query firmware error recovery capabilities In Driver initiated error recovery process, driver has to know about the registers offset and values to initiate FW reset. The HWRM command HWRM_ERROR_RECOVERY_QCFG is used to obtain all the registers and values required to initiate FW reset. This command response includes FW heart_beat register, health status register, Error counter register, register offsets and values to do chip reset if firmware crashes and becomes unresponsive. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:26 +00:00
/* Get the adapter error recovery support info */
rc = bnxt_hwrm_error_recovery_qcfg(bp);
if (rc)
net/bnxt: add a field for FW capabilities The purpose of this patch is used to add a new field "fw_cap" in bnxt structure. The new "fw_cap" field store all the capabilities reported by the FW. This information will help us to determine the device capabilities when registering feature support with FW. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-12-10 15:05:29 +00:00
bp->fw_cap &= ~BNXT_FW_CAP_ERROR_RECOVERY;
net/bnxt: query firmware error recovery capabilities In Driver initiated error recovery process, driver has to know about the registers offset and values to initiate FW reset. The HWRM command HWRM_ERROR_RECOVERY_QCFG is used to obtain all the registers and values required to initiate FW reset. This command response includes FW heart_beat register, health status register, Error counter register, register offsets and values to do chip reset if firmware crashes and becomes unresponsive. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:26 +00:00
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
bnxt_hwrm_port_led_qcaps(bp);
return 0;
}
net/bnxt: synchronize between flow related functions Currently, there are four flow related functions, namely bnxt_flow_create, bnxt_flow_destroy, bnxt_flow_validate, bnxt_flow_flush. All these functions are not multi-thread safe. This patch fixes it by synchronizing these functions with a lock. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
2019-10-02 23:25:57 +00:00
static int
bnxt_init_locks(struct bnxt *bp)
{
int err;
err = pthread_mutex_init(&bp->flow_lock, NULL);
net/bnxt: fix race between interrupt handler and dev config Fix a possible that while interrupt handler is in the middle of processing, dev_configure_op() might come in and try to destroy/realloc the default completion ring on which interrupt completions arrive. Synchronize the two so they don't race each other Fixes: 7bc8e9a227cc ("net/bnxt: support async link notification") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-11 04:44:16 +00:00
if (err) {
net/bnxt: synchronize between flow related functions Currently, there are four flow related functions, namely bnxt_flow_create, bnxt_flow_destroy, bnxt_flow_validate, bnxt_flow_flush. All these functions are not multi-thread safe. This patch fixes it by synchronizing these functions with a lock. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
2019-10-02 23:25:57 +00:00
PMD_DRV_LOG(ERR, "Unable to initialize flow_lock\n");
net/bnxt: fix race between interrupt handler and dev config Fix a possible that while interrupt handler is in the middle of processing, dev_configure_op() might come in and try to destroy/realloc the default completion ring on which interrupt completions arrive. Synchronize the two so they don't race each other Fixes: 7bc8e9a227cc ("net/bnxt: support async link notification") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-11 04:44:16 +00:00
return err;
}
err = pthread_mutex_init(&bp->def_cp_lock, NULL);
net/bnxt: fix cleanup on mutex init failure In case mutex init fails during initialization, start cleanup and fail the initialization process. Fixes: a73b8e939f10 ("net/bnxt: fix race between start and interrupt handler") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2020-12-01 20:52:53 +00:00
if (err) {
net/bnxt: fix race between interrupt handler and dev config Fix a possible that while interrupt handler is in the middle of processing, dev_configure_op() might come in and try to destroy/realloc the default completion ring on which interrupt completions arrive. Synchronize the two so they don't race each other Fixes: 7bc8e9a227cc ("net/bnxt: support async link notification") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-11 04:44:16 +00:00
PMD_DRV_LOG(ERR, "Unable to initialize def_cp_lock\n");
net/bnxt: fix cleanup on mutex init failure In case mutex init fails during initialization, start cleanup and fail the initialization process. Fixes: a73b8e939f10 ("net/bnxt: fix race between start and interrupt handler") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2020-12-01 20:52:53 +00:00
return err;
}
net/bnxt: add separate mutex for FW health check def_cp_lock was added to sync race between dev_configure and int_handler. It should not be used to synchronize scheduling of FW health check between dev_start and async event handler as well, use a separate mutex for the same. Fixes: a73b8e939f10 ("net/bnxt: fix race between start and interrupt handler") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-11 01:56:03 +00:00
err = pthread_mutex_init(&bp->health_check_lock, NULL);
net/bnxt: check chip reset in stop and close While the error recovery thread is running, an application can invoke dev_stop or dev_close_op thus triggering a race and unwanted consequences if dev_close is invoked while the recovery is not yet completed. Fix by having another lock to synchronize between the 2 threads and return EGAIN if adapter is in the middle of recovery when dev_stop or dev_close ops are invoked Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-24 09:37:34 +00:00
if (err) {
net/bnxt: add separate mutex for FW health check def_cp_lock was added to sync race between dev_configure and int_handler. It should not be used to synchronize scheduling of FW health check between dev_start and async event handler as well, use a separate mutex for the same. Fixes: a73b8e939f10 ("net/bnxt: fix race between start and interrupt handler") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-11 01:56:03 +00:00
PMD_DRV_LOG(ERR, "Unable to initialize health_check_lock\n");
net/bnxt: check chip reset in stop and close While the error recovery thread is running, an application can invoke dev_stop or dev_close_op thus triggering a race and unwanted consequences if dev_close is invoked while the recovery is not yet completed. Fix by having another lock to synchronize between the 2 threads and return EGAIN if adapter is in the middle of recovery when dev_stop or dev_close ops are invoked Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-24 09:37:34 +00:00
return err;
}
err = pthread_mutex_init(&bp->err_recovery_lock, NULL);
if (err)
PMD_DRV_LOG(ERR, "Unable to initialize err_recovery_lock\n");
net/bnxt: synchronize between flow related functions Currently, there are four flow related functions, namely bnxt_flow_create, bnxt_flow_destroy, bnxt_flow_validate, bnxt_flow_flush. All these functions are not multi-thread safe. This patch fixes it by synchronizing these functions with a lock. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
2019-10-02 23:25:57 +00:00
return err;
}
net/bnxt: handle reset notify async event from FW When the FW upgrade is initiated the current instance of FW issues a HWRM_ASYNC_EVENT_CMPL_EVENT_ID_RESET_NOTIFY async notification to the driver. On receiving this notification, the PMD shall quiesce itself and poll on the HWRM_VER_GET FW command at regular intervals. Once the VER_GET command succeeds, the driver should go through the rediscovery process and re-initialize the device. Also register with FW for the reset notify async event. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:23 +00:00
static int bnxt_init_resources(struct bnxt *bp, bool reconfig_dev)
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
{
net/bnxt: support port representor data path Added code to support Tx/Rx from a VF representor port. The VF-reps use the RX/TX rings of the Trusted VF/PF. For each VF-rep, the Trusted VF/PF driver issues a VFR_ALLOC FW cmd that returns "cfa_code" and "cfa_action" values. The FW sets up the filter tables in such a way that VF traffic by default (in absence of other rules) gets punted to the parent function i.e. either the Trusted VF or the PF. The cfa_code value in the RX-compl informs the driver of the source VF. For traffic being transmitted from the VF-rep, the TX BD is tagged with a cfa_action value that informs the HW to punt it to the corresponding VF. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:49 +00:00
int rc = 0;
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
net/bnxt: refactor init/uninit Move all the individual driver fields allocation routines to one routine - bnxt_drv_init(). This houses all such routines where memory needs to be allocated once during the driver's lifetime and does not need to be torn down during error recovery. Rename some function names in accordance with their functionality. bnxt_init_board() is doing nothing more than mapping the PCI bars, so rename it as such. Given that there is a bnxt_shutdown_nic that is called in dev_stop_op, rename it's counterpart - bnxt_init_chip() that is called in dev_start_op, to bnxt_start_nic. Also helps avoid confusion with some of the other bnxt_init_xxx routines. Rename bnxt_init_fw() to bnxt_get_config() as that is what that routine is doing mostly functionality wise. Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-01-14 02:19:14 +00:00
rc = bnxt_get_config(bp);
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
if (rc)
return rc;
net/bnxt: handle reset notify async event from FW When the FW upgrade is initiated the current instance of FW issues a HWRM_ASYNC_EVENT_CMPL_EVENT_ID_RESET_NOTIFY async notification to the driver. On receiving this notification, the PMD shall quiesce itself and poll on the HWRM_VER_GET FW command at regular intervals. Once the VER_GET command succeeds, the driver should go through the rediscovery process and re-initialize the device. Also register with FW for the reset notify async event. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:23 +00:00
if (!reconfig_dev) {
rc = bnxt_setup_mac_addr(bp->eth_dev);
if (rc)
return rc;
} else {
rc = bnxt_restore_dflt_mac(bp);
if (rc)
return rc;
}
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
bnxt_config_vf_req_fwd(bp);
rc = bnxt_hwrm_func_driver_register(bp);
if (rc) {
PMD_DRV_LOG(ERR, "Failed to register driver");
return -EBUSY;
}
if (BNXT_PF(bp)) {
if (bp->pdev->max_vfs) {
rc = bnxt_hwrm_allocate_vfs(bp, bp->pdev->max_vfs);
if (rc) {
PMD_DRV_LOG(ERR, "Failed to allocate VFs\n");
return rc;
}
} else {
rc = bnxt_hwrm_allocate_pf_only(bp);
if (rc) {
PMD_DRV_LOG(ERR,
"Failed to allocate PF resources");
return rc;
}
}
}
net/bnxt: handle reset notify async event from FW When the FW upgrade is initiated the current instance of FW issues a HWRM_ASYNC_EVENT_CMPL_EVENT_ID_RESET_NOTIFY async notification to the driver. On receiving this notification, the PMD shall quiesce itself and poll on the HWRM_VER_GET FW command at regular intervals. Once the VER_GET command succeeds, the driver should go through the rediscovery process and re-initialize the device. Also register with FW for the reset notify async event. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:23 +00:00
rc = bnxt_alloc_mem(bp, reconfig_dev);
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
if (rc)
return rc;
rc = bnxt_setup_int(bp);
if (rc)
return rc;
rc = bnxt_request_int(bp);
if (rc)
return rc;
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
rc = bnxt_init_ctx_mem(bp);
if (rc) {
PMD_DRV_LOG(ERR, "Failed to init adv_flow_counters\n");
return rc;
}
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
return 0;
}
net/bnxt: support host memory based TruFlow This feature can be enabled by passing "-w 0000:0d:00.0,host-based-truflow=1” to the DPDK application. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-04-15 08:18:52 +00:00
static int
bnxt_parse_devarg_truflow(__rte_unused const char *key,
const char *value, void *opaque_arg)
{
struct bnxt *bp = opaque_arg;
unsigned long truflow;
char *end = NULL;
if (!value || !opaque_arg) {
PMD_DRV_LOG(ERR,
"Invalid parameter passed to truflow devargs.\n");
return -EINVAL;
}
truflow = strtoul(value, &end, 10);
if (end == NULL || *end != '\0' ||
(truflow == ULONG_MAX && errno == ERANGE)) {
PMD_DRV_LOG(ERR,
"Invalid parameter passed to truflow devargs.\n");
return -EINVAL;
}
if (BNXT_DEVARG_TRUFLOW_INVALID(truflow)) {
PMD_DRV_LOG(ERR,
"Invalid value passed to truflow devargs.\n");
return -EINVAL;
}
net/bnxt: fix TruFlow devarg handling Set the TRUFLOW Enable bit in bp->flags only if the value passed in devargs was 1. Otherwise set it to 0. Fixes: 313ac35ac701 ("net/bnxt: support ULP session manager init") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:28 +00:00
if (truflow) {
bp->flags |= BNXT_FLAG_TRUFLOW_EN;
net/bnxt: support host memory based TruFlow This feature can be enabled by passing "-w 0000:0d:00.0,host-based-truflow=1” to the DPDK application. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-04-15 08:18:52 +00:00
PMD_DRV_LOG(INFO, "Host-based truflow feature enabled.\n");
net/bnxt: fix TruFlow devarg handling Set the TRUFLOW Enable bit in bp->flags only if the value passed in devargs was 1. Otherwise set it to 0. Fixes: 313ac35ac701 ("net/bnxt: support ULP session manager init") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:28 +00:00
} else {
bp->flags &= ~BNXT_FLAG_TRUFLOW_EN;
PMD_DRV_LOG(INFO, "Host-based truflow feature disabled.\n");
}
net/bnxt: support host memory based TruFlow This feature can be enabled by passing "-w 0000:0d:00.0,host-based-truflow=1” to the DPDK application. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-04-15 08:18:52 +00:00
return 0;
}
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
static int
bnxt_parse_devarg_flow_xstat(__rte_unused const char *key,
const char *value, void *opaque_arg)
{
struct bnxt *bp = opaque_arg;
unsigned long flow_xstat;
char *end = NULL;
if (!value || !opaque_arg) {
PMD_DRV_LOG(ERR,
"Invalid parameter passed to flow_xstat devarg.\n");
return -EINVAL;
}
flow_xstat = strtoul(value, &end, 10);
if (end == NULL || *end != '\0' ||
(flow_xstat == ULONG_MAX && errno == ERANGE)) {
PMD_DRV_LOG(ERR,
"Invalid parameter passed to flow_xstat devarg.\n");
return -EINVAL;
}
if (BNXT_DEVARG_FLOW_XSTAT_INVALID(flow_xstat)) {
PMD_DRV_LOG(ERR,
"Invalid value passed to flow_xstat devarg.\n");
return -EINVAL;
}
net/bnxt: fix allocation of flow stat related structs Consolidate flow stat related structs for performance improvement. The intention of this patch is to reduce the size struct bnxt which had grown because of recent changes and was impacting performance. Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 08:31:59 +00:00
bp->flags |= BNXT_FLAG_FLOW_XSTATS_EN;
if (BNXT_FLOW_XSTATS_EN(bp))
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
PMD_DRV_LOG(INFO, "flow_xstat feature enabled.\n");
return 0;
}
net/bnxt: set maximum flow count User could set max flow count by passing a devarg "-w 0000:0d:00.0,max_num_kflows=64" to a DPDK application; The value must be not less than 32K and be power-of-2; the default value is 32K. Signed-off-by: Shuanglin Wang <shuanglin.wang@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-06-12 12:50:12 +00:00
static int
bnxt_parse_devarg_max_num_kflows(__rte_unused const char *key,
const char *value, void *opaque_arg)
{
struct bnxt *bp = opaque_arg;
unsigned long max_num_kflows;
char *end = NULL;
if (!value || !opaque_arg) {
PMD_DRV_LOG(ERR,
"Invalid parameter passed to max_num_kflows devarg.\n");
return -EINVAL;
}
max_num_kflows = strtoul(value, &end, 10);
if (end == NULL || *end != '\0' ||
(max_num_kflows == ULONG_MAX && errno == ERANGE)) {
PMD_DRV_LOG(ERR,
"Invalid parameter passed to max_num_kflows devarg.\n");
return -EINVAL;
}
if (bnxt_devarg_max_num_kflow_invalid(max_num_kflows)) {
PMD_DRV_LOG(ERR,
"Invalid value passed to max_num_kflows devarg.\n");
return -EINVAL;
}
bp->max_num_kflows = max_num_kflows;
if (bp->max_num_kflows)
PMD_DRV_LOG(INFO, "max_num_kflows set as %ldK.\n",
max_num_kflows);
return 0;
}
net/bnxt: support representors on remote host domain In the Stingray use case, representors are conventionally run inside the SoC domain representing functions that are on the X86 domain. In order to support this mechanism of building representors for endpoints that are not in the same host domain, additional dev args have been in the PMD like so: rep-based-pf=<physical index> rep-is-pf=<VF=0 or PF=1> where `rep-based-pf` specifies the physical index of the base PF that the representor is derived off of. Since representor(s) can be created for endpoint PFs as well, rename struct bnxt_vf_representor to bnxt_representor and other such dev_ops and function names. devargs have also been extended to specify the exact CoS queue along with flow control enablement to be used for the conduit between the representor and the endpoint function. This is how a sample devargs would look with all the extended devargs -w 0000:06:02.0,host-based-truflow=1,representor=[1],rep-based-pf=8, rep-is-pf=1,rep-q-r2f=1,rep-fc-r2f=0,rep-q-f2r=1,rep-fc-f2r=1 Call CFA_PAIR_ALLOC only in case of Stingray instead of CFA_VFR_ALLOC. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:30 +00:00
static int
bnxt_parse_devarg_rep_is_pf(__rte_unused const char *key,
const char *value, void *opaque_arg)
{
struct bnxt_representor *vfr_bp = opaque_arg;
unsigned long rep_is_pf;
char *end = NULL;
if (!value || !opaque_arg) {
PMD_DRV_LOG(ERR,
"Invalid parameter passed to rep_is_pf devargs.\n");
return -EINVAL;
}
rep_is_pf = strtoul(value, &end, 10);
if (end == NULL || *end != '\0' ||
(rep_is_pf == ULONG_MAX && errno == ERANGE)) {
PMD_DRV_LOG(ERR,
"Invalid parameter passed to rep_is_pf devargs.\n");
return -EINVAL;
}
if (BNXT_DEVARG_REP_IS_PF_INVALID(rep_is_pf)) {
PMD_DRV_LOG(ERR,
"Invalid value passed to rep_is_pf devargs.\n");
return -EINVAL;
}
vfr_bp->flags |= rep_is_pf;
if (BNXT_REP_PF(vfr_bp))
PMD_DRV_LOG(INFO, "PF representor\n");
else
PMD_DRV_LOG(INFO, "VF representor\n");
return 0;
}
static int
bnxt_parse_devarg_rep_based_pf(__rte_unused const char *key,
const char *value, void *opaque_arg)
{
struct bnxt_representor *vfr_bp = opaque_arg;
unsigned long rep_based_pf;
char *end = NULL;
if (!value || !opaque_arg) {
PMD_DRV_LOG(ERR,
"Invalid parameter passed to rep_based_pf "
"devargs.\n");
return -EINVAL;
}
rep_based_pf = strtoul(value, &end, 10);
if (end == NULL || *end != '\0' ||
(rep_based_pf == ULONG_MAX && errno == ERANGE)) {
PMD_DRV_LOG(ERR,
"Invalid parameter passed to rep_based_pf "
"devargs.\n");
return -EINVAL;
}
if (BNXT_DEVARG_REP_BASED_PF_INVALID(rep_based_pf)) {
PMD_DRV_LOG(ERR,
"Invalid value passed to rep_based_pf devargs.\n");
return -EINVAL;
}
vfr_bp->rep_based_pf = rep_based_pf;
net/bnxt: modify HWRM command to create reps Use cfa pair alloc for configuring reps. Instead of cfa_vfr_alloc for Wh+ and cfa_pair_alloc for Stingray, converge to cfa_pair_alloc/free for both devices. Set the command request structure bits accordingly. As part of this, remove the old cfa_vfr_alloc cmd definitions as FW has deprecated support for those commands. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Shahaji Bhosle <sbhosle@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-10-26 03:56:08 +00:00
vfr_bp->flags |= BNXT_REP_BASED_PF_VALID;
net/bnxt: support representors on remote host domain In the Stingray use case, representors are conventionally run inside the SoC domain representing functions that are on the X86 domain. In order to support this mechanism of building representors for endpoints that are not in the same host domain, additional dev args have been in the PMD like so: rep-based-pf=<physical index> rep-is-pf=<VF=0 or PF=1> where `rep-based-pf` specifies the physical index of the base PF that the representor is derived off of. Since representor(s) can be created for endpoint PFs as well, rename struct bnxt_vf_representor to bnxt_representor and other such dev_ops and function names. devargs have also been extended to specify the exact CoS queue along with flow control enablement to be used for the conduit between the representor and the endpoint function. This is how a sample devargs would look with all the extended devargs -w 0000:06:02.0,host-based-truflow=1,representor=[1],rep-based-pf=8, rep-is-pf=1,rep-q-r2f=1,rep-fc-r2f=0,rep-q-f2r=1,rep-fc-f2r=1 Call CFA_PAIR_ALLOC only in case of Stingray instead of CFA_VFR_ALLOC. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:30 +00:00
PMD_DRV_LOG(INFO, "rep-based-pf = %d\n", vfr_bp->rep_based_pf);
return 0;
}
static int
bnxt_parse_devarg_rep_q_r2f(__rte_unused const char *key,
const char *value, void *opaque_arg)
{
struct bnxt_representor *vfr_bp = opaque_arg;
unsigned long rep_q_r2f;
char *end = NULL;
if (!value || !opaque_arg) {
PMD_DRV_LOG(ERR,
"Invalid parameter passed to rep_q_r2f "
"devargs.\n");
return -EINVAL;
}
rep_q_r2f = strtoul(value, &end, 10);
if (end == NULL || *end != '\0' ||
(rep_q_r2f == ULONG_MAX && errno == ERANGE)) {
PMD_DRV_LOG(ERR,
"Invalid parameter passed to rep_q_r2f "
"devargs.\n");
return -EINVAL;
}
if (BNXT_DEVARG_REP_Q_R2F_INVALID(rep_q_r2f)) {
PMD_DRV_LOG(ERR,
"Invalid value passed to rep_q_r2f devargs.\n");
return -EINVAL;
}
vfr_bp->rep_q_r2f = rep_q_r2f;
vfr_bp->flags |= BNXT_REP_Q_R2F_VALID;
PMD_DRV_LOG(INFO, "rep-q-r2f = %d\n", vfr_bp->rep_q_r2f);
return 0;
}
static int
bnxt_parse_devarg_rep_q_f2r(__rte_unused const char *key,
const char *value, void *opaque_arg)
{
struct bnxt_representor *vfr_bp = opaque_arg;
unsigned long rep_q_f2r;
char *end = NULL;
if (!value || !opaque_arg) {
PMD_DRV_LOG(ERR,
"Invalid parameter passed to rep_q_f2r "
"devargs.\n");
return -EINVAL;
}
rep_q_f2r = strtoul(value, &end, 10);
if (end == NULL || *end != '\0' ||
(rep_q_f2r == ULONG_MAX && errno == ERANGE)) {
PMD_DRV_LOG(ERR,
"Invalid parameter passed to rep_q_f2r "
"devargs.\n");
return -EINVAL;
}
if (BNXT_DEVARG_REP_Q_F2R_INVALID(rep_q_f2r)) {
PMD_DRV_LOG(ERR,
"Invalid value passed to rep_q_f2r devargs.\n");
return -EINVAL;
}
vfr_bp->rep_q_f2r = rep_q_f2r;
vfr_bp->flags |= BNXT_REP_Q_F2R_VALID;
PMD_DRV_LOG(INFO, "rep-q-f2r = %d\n", vfr_bp->rep_q_f2r);
return 0;
}
static int
bnxt_parse_devarg_rep_fc_r2f(__rte_unused const char *key,
const char *value, void *opaque_arg)
{
struct bnxt_representor *vfr_bp = opaque_arg;
unsigned long rep_fc_r2f;
char *end = NULL;
if (!value || !opaque_arg) {
PMD_DRV_LOG(ERR,
"Invalid parameter passed to rep_fc_r2f "
"devargs.\n");
return -EINVAL;
}
rep_fc_r2f = strtoul(value, &end, 10);
if (end == NULL || *end != '\0' ||
(rep_fc_r2f == ULONG_MAX && errno == ERANGE)) {
PMD_DRV_LOG(ERR,
"Invalid parameter passed to rep_fc_r2f "
"devargs.\n");
return -EINVAL;
}
if (BNXT_DEVARG_REP_FC_R2F_INVALID(rep_fc_r2f)) {
PMD_DRV_LOG(ERR,
"Invalid value passed to rep_fc_r2f devargs.\n");
return -EINVAL;
}
vfr_bp->flags |= BNXT_REP_FC_R2F_VALID;
vfr_bp->rep_fc_r2f = rep_fc_r2f;
PMD_DRV_LOG(INFO, "rep-fc-r2f = %lu\n", rep_fc_r2f);
return 0;
}
static int
bnxt_parse_devarg_rep_fc_f2r(__rte_unused const char *key,
const char *value, void *opaque_arg)
{
struct bnxt_representor *vfr_bp = opaque_arg;
unsigned long rep_fc_f2r;
char *end = NULL;
if (!value || !opaque_arg) {
PMD_DRV_LOG(ERR,
"Invalid parameter passed to rep_fc_f2r "
"devargs.\n");
return -EINVAL;
}
rep_fc_f2r = strtoul(value, &end, 10);
if (end == NULL || *end != '\0' ||
(rep_fc_f2r == ULONG_MAX && errno == ERANGE)) {
PMD_DRV_LOG(ERR,
"Invalid parameter passed to rep_fc_f2r "
"devargs.\n");
return -EINVAL;
}
if (BNXT_DEVARG_REP_FC_F2R_INVALID(rep_fc_f2r)) {
PMD_DRV_LOG(ERR,
"Invalid value passed to rep_fc_f2r devargs.\n");
return -EINVAL;
}
vfr_bp->flags |= BNXT_REP_FC_F2R_VALID;
vfr_bp->rep_fc_f2r = rep_fc_f2r;
PMD_DRV_LOG(INFO, "rep-fc-f2r = %lu\n", rep_fc_f2r);
return 0;
}
net/bnxt: check kvargs parsing Check return value of rte_kvargs_process() Coverity issue: 357765 Fixes: ba404aacc5cf ("net/bnxt: set maximum flow count") Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Fixes: 7b0940653720 ("net/bnxt: support host memory based TruFlow") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com>
2021-03-17 01:08:40 +00:00
static int
net/bnxt: support host memory based TruFlow This feature can be enabled by passing "-w 0000:0d:00.0,host-based-truflow=1” to the DPDK application. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-04-15 08:18:52 +00:00
bnxt_parse_dev_args(struct bnxt *bp, struct rte_devargs *devargs)
{
struct rte_kvargs *kvlist;
net/bnxt: check kvargs parsing Check return value of rte_kvargs_process() Coverity issue: 357765 Fixes: ba404aacc5cf ("net/bnxt: set maximum flow count") Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Fixes: 7b0940653720 ("net/bnxt: support host memory based TruFlow") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com>
2021-03-17 01:08:40 +00:00
int ret;
net/bnxt: support host memory based TruFlow This feature can be enabled by passing "-w 0000:0d:00.0,host-based-truflow=1” to the DPDK application. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-04-15 08:18:52 +00:00
if (devargs == NULL)
net/bnxt: check kvargs parsing Check return value of rte_kvargs_process() Coverity issue: 357765 Fixes: ba404aacc5cf ("net/bnxt: set maximum flow count") Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Fixes: 7b0940653720 ("net/bnxt: support host memory based TruFlow") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com>
2021-03-17 01:08:40 +00:00
return 0;
net/bnxt: support host memory based TruFlow This feature can be enabled by passing "-w 0000:0d:00.0,host-based-truflow=1” to the DPDK application. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-04-15 08:18:52 +00:00
kvlist = rte_kvargs_parse(devargs->args, bnxt_dev_args);
if (kvlist == NULL)
net/bnxt: check kvargs parsing Check return value of rte_kvargs_process() Coverity issue: 357765 Fixes: ba404aacc5cf ("net/bnxt: set maximum flow count") Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Fixes: 7b0940653720 ("net/bnxt: support host memory based TruFlow") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com>
2021-03-17 01:08:40 +00:00
return -EINVAL;
net/bnxt: support host memory based TruFlow This feature can be enabled by passing "-w 0000:0d:00.0,host-based-truflow=1” to the DPDK application. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-04-15 08:18:52 +00:00
/*
* Handler for "truflow" devarg.
eal: replace blacklist/whitelist options Replace -w / --pci-whitelist with -a / --allow options and --pci-blacklist with --block. The -b short option remains unchanged. Allow the old options for now, but print a nag warning since old options are deprecated. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Luca Boccassi <bluca@debian.org> Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
2020-11-10 22:55:40 +00:00
* Invoked as for ex: "-a 0000:00:0d.0,host-based-truflow=1"
net/bnxt: support host memory based TruFlow This feature can be enabled by passing "-w 0000:0d:00.0,host-based-truflow=1” to the DPDK application. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-04-15 08:18:52 +00:00
*/
net/bnxt: check kvargs parsing Check return value of rte_kvargs_process() Coverity issue: 357765 Fixes: ba404aacc5cf ("net/bnxt: set maximum flow count") Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Fixes: 7b0940653720 ("net/bnxt: support host memory based TruFlow") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com>
2021-03-17 01:08:40 +00:00
ret = rte_kvargs_process(kvlist, BNXT_DEVARG_TRUFLOW,
bnxt_parse_devarg_truflow, bp);
if (ret)
goto err;
net/bnxt: support host memory based TruFlow This feature can be enabled by passing "-w 0000:0d:00.0,host-based-truflow=1” to the DPDK application. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-04-15 08:18:52 +00:00
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
/*
* Handler for "flow_xstat" devarg.
eal: replace blacklist/whitelist options Replace -w / --pci-whitelist with -a / --allow options and --pci-blacklist with --block. The -b short option remains unchanged. Allow the old options for now, but print a nag warning since old options are deprecated. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Luca Boccassi <bluca@debian.org> Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
2020-11-10 22:55:40 +00:00
* Invoked as for ex: "-a 0000:00:0d.0,flow_xstat=1"
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
*/
net/bnxt: check kvargs parsing Check return value of rte_kvargs_process() Coverity issue: 357765 Fixes: ba404aacc5cf ("net/bnxt: set maximum flow count") Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Fixes: 7b0940653720 ("net/bnxt: support host memory based TruFlow") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com>
2021-03-17 01:08:40 +00:00
ret = rte_kvargs_process(kvlist, BNXT_DEVARG_FLOW_XSTAT,
bnxt_parse_devarg_flow_xstat, bp);
if (ret)
goto err;
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
net/bnxt: set maximum flow count User could set max flow count by passing a devarg "-w 0000:0d:00.0,max_num_kflows=64" to a DPDK application; The value must be not less than 32K and be power-of-2; the default value is 32K. Signed-off-by: Shuanglin Wang <shuanglin.wang@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-06-12 12:50:12 +00:00
/*
* Handler for "max_num_kflows" devarg.
eal: replace blacklist/whitelist options Replace -w / --pci-whitelist with -a / --allow options and --pci-blacklist with --block. The -b short option remains unchanged. Allow the old options for now, but print a nag warning since old options are deprecated. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Luca Boccassi <bluca@debian.org> Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
2020-11-10 22:55:40 +00:00
* Invoked as for ex: "-a 000:00:0d.0,max_num_kflows=32"
net/bnxt: set maximum flow count User could set max flow count by passing a devarg "-w 0000:0d:00.0,max_num_kflows=64" to a DPDK application; The value must be not less than 32K and be power-of-2; the default value is 32K. Signed-off-by: Shuanglin Wang <shuanglin.wang@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-06-12 12:50:12 +00:00
*/
net/bnxt: check kvargs parsing Check return value of rte_kvargs_process() Coverity issue: 357765 Fixes: ba404aacc5cf ("net/bnxt: set maximum flow count") Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Fixes: 7b0940653720 ("net/bnxt: support host memory based TruFlow") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com>
2021-03-17 01:08:40 +00:00
ret = rte_kvargs_process(kvlist, BNXT_DEVARG_MAX_NUM_KFLOWS,
bnxt_parse_devarg_max_num_kflows, bp);
if (ret)
goto err;
net/bnxt: set maximum flow count User could set max flow count by passing a devarg "-w 0000:0d:00.0,max_num_kflows=64" to a DPDK application; The value must be not less than 32K and be power-of-2; the default value is 32K. Signed-off-by: Shuanglin Wang <shuanglin.wang@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-06-12 12:50:12 +00:00
net/bnxt: check kvargs parsing Check return value of rte_kvargs_process() Coverity issue: 357765 Fixes: ba404aacc5cf ("net/bnxt: set maximum flow count") Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Fixes: 7b0940653720 ("net/bnxt: support host memory based TruFlow") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com>
2021-03-17 01:08:40 +00:00
err:
net/bnxt: support host memory based TruFlow This feature can be enabled by passing "-w 0000:0d:00.0,host-based-truflow=1” to the DPDK application. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-04-15 08:18:52 +00:00
rte_kvargs_free(kvlist);
net/bnxt: check kvargs parsing Check return value of rte_kvargs_process() Coverity issue: 357765 Fixes: ba404aacc5cf ("net/bnxt: set maximum flow count") Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Fixes: 7b0940653720 ("net/bnxt: support host memory based TruFlow") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com>
2021-03-17 01:08:40 +00:00
return ret;
net/bnxt: support host memory based TruFlow This feature can be enabled by passing "-w 0000:0d:00.0,host-based-truflow=1” to the DPDK application. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-04-15 08:18:52 +00:00
}
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
static int bnxt_alloc_switch_domain(struct bnxt *bp)
{
int rc = 0;
if (BNXT_PF(bp) || BNXT_VF_IS_TRUSTED(bp)) {
rc = rte_eth_switch_domain_alloc(&bp->switch_domain_id);
if (rc)
PMD_DRV_LOG(ERR,
"Failed to alloc switch domain: %d\n", rc);
else
PMD_DRV_LOG(INFO,
"Switch domain allocated %d\n",
bp->switch_domain_id);
}
return rc;
}
net/bnxt: refactor init/uninit Move all the individual driver fields allocation routines to one routine - bnxt_drv_init(). This houses all such routines where memory needs to be allocated once during the driver's lifetime and does not need to be torn down during error recovery. Rename some function names in accordance with their functionality. bnxt_init_board() is doing nothing more than mapping the PCI bars, so rename it as such. Given that there is a bnxt_shutdown_nic that is called in dev_stop_op, rename it's counterpart - bnxt_init_chip() that is called in dev_start_op, to bnxt_start_nic. Also helps avoid confusion with some of the other bnxt_init_xxx routines. Rename bnxt_init_fw() to bnxt_get_config() as that is what that routine is doing mostly functionality wise. Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-01-14 02:19:14 +00:00
/* Allocate and initialize various fields in bnxt struct that
* need to be allocated/destroyed only once in the lifetime of the driver
*/
static int bnxt_drv_init(struct rte_eth_dev *eth_dev)
net/bnxt: add driver for Broadcom NetXtreme-C devices This patch adds the initial skeleton for bnxt driver along with the nic guide, and ties the driver into the build system. At this point, the driver simply fails init. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com> [Release Note Addition] Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2016-06-15 21:23:01 +00:00
{
drivers/net: add generic ethdev macro to get PCI device Instead of many PMD define their own macro, define a generic one in ethdev and use that in PMDs. Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com> Acked-by: Allain Legacy <allain.legacy@windriver.com> Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com> Acked-by: Andrew Rybchenko <arybchenko@solarflare.com>
2017-05-15 10:24:03 +00:00
struct rte_pci_device *pci_dev = RTE_ETH_DEV_TO_PCI(eth_dev);
net/bnxt: refactor init/uninit Move all the individual driver fields allocation routines to one routine - bnxt_drv_init(). This houses all such routines where memory needs to be allocated once during the driver's lifetime and does not need to be torn down during error recovery. Rename some function names in accordance with their functionality. bnxt_init_board() is doing nothing more than mapping the PCI bars, so rename it as such. Given that there is a bnxt_shutdown_nic that is called in dev_stop_op, rename it's counterpart - bnxt_init_chip() that is called in dev_start_op, to bnxt_start_nic. Also helps avoid confusion with some of the other bnxt_init_xxx routines. Rename bnxt_init_fw() to bnxt_get_config() as that is what that routine is doing mostly functionality wise. Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-01-14 02:19:14 +00:00
struct bnxt *bp = eth_dev->data->dev_private;
int rc = 0;
net/bnxt: support host memory based TruFlow This feature can be enabled by passing "-w 0000:0d:00.0,host-based-truflow=1” to the DPDK application. Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-04-15 08:18:52 +00:00
net/bnxt: support flow mark action Add support for RTE_FLOW_ACTION_TYPE_MARK. Use the flow_id provided by FW during flow creation to lookup the mark id provided by the application. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com>
2019-12-21 02:38:49 +00:00
bp->flags &= ~BNXT_FLAG_RX_VECTOR_PKT_MODE;
net/bnxt: fix crash in secondary process The secondary process should not modify device state when init is called. The pci device information pointed to by dev_private pointer is not correct in secondary process. Fixes: b7778e8a1c00 ("net/bnxt: refactor to properly allocate resources for PF/VF") Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 17:17:43 +00:00
net/bnxt: localize mapping of ethdev to PCI device Use existing information about pci and interrupt handle to minimize the number of places that assume eth_dev contains pci_device information. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Jan Blunck <jblunck@infradead.org>
2016-12-23 15:57:59 +00:00
if (bnxt_vf_pciid(pci_dev->id.device_id))
net/bnxt: add hardware resource manager init code Start adding support to use the HWRM API. Hardware Resource Manager or HWRM in short, is a set of API provided by the firmware running in the ASIC to manage the various resources. Initial commit just performs necessary HWRM queries for init, then fails as before. Now that struct bnxt is non-zero size, we can set dev_private_size correctly. The used HWRM calls so far: bnxt_hwrm_func_qcaps: This command returns capabilities of a function. bnxt_hwrm_ver_get: This function is called by a driver to determine the HWRM interface version supported by the HWRM firmware, the version of HWRM firmware implementation, the name of HWRM firmware, the versions of other embedded firmwares, and the names of other embedded firmwares, etc. Gets the firmware version and interface specifications. Returns an error if the firmware on the device is not supported by the driver and ensures the response space is large enough for the largest possible response. bnxt_hwrm_queue_qportcfg: This function is called by a driver to query queue configuration of a port. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:02 +00:00
bp->flags |= BNXT_FLAG_VF;
net/bnxt: remove references to Thor Refactor code to remove references to Thor. Instead use P5 as in phase 5 of development cycle since it is applicable to boards other than Thor as well. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Acked-by: Lance Richardson <lance.richardson@broadcom.com>
2020-12-01 19:15:23 +00:00
if (bnxt_p5_device(pci_dev->id.device_id))
bp->flags |= BNXT_FLAG_CHIP_P5;
net/bnxt: support thor controller This commit adds support to the bnxt PMD for devices based on the BCM57508 "thor" Ethernet controller. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
2019-06-02 17:42:44 +00:00
net/bnxt: use dedicated CPR for async events This commit enables the creation of a dedicated completion ring for asynchronous event handling instead of handling these events on a receive completion ring. For the stingray platform and other platforms needing tighter control of resource utilization, we retain the ability to process async events on a receive completion ring. For Thor-based adapters, we use a dedicated NQ (notification queue) ring for async events (async events can't currently be received on a completion ring due to a firmware limitation). Rename "def_cp_ring" to "async_cp_ring" to better reflect its purpose (async event notifications) and to avoid confusion with VNIC default receive completion rings. Allow rxq 0 to be stopped when not being used for async events. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-24 16:49:32 +00:00
if (pci_dev->id.device_id == BROADCOM_DEV_ID_58802 ||
pci_dev->id.device_id == BROADCOM_DEV_ID_58804 ||
pci_dev->id.device_id == BROADCOM_DEV_ID_58808 ||
pci_dev->id.device_id == BROADCOM_DEV_ID_58802_VF)
bp->flags |= BNXT_FLAG_STINGRAY;
net/bnxt: switch CFA code to dynamic mbuf field The CFA code from mark was stored in the deprecated mbuf field udata64. It is moved to a dynamic field in order to allow removal of udata64. Note: the new field has 32 bits, smaller than udata64. Signed-off-by: Thomas Monjalon <thomas@monjalon.net> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-10-25 22:41:16 +00:00
if (BNXT_TRUFLOW_EN(bp)) {
/* extra mbuf field is required to store CFA code from mark */
static const struct rte_mbuf_dynfield bnxt_cfa_code_dynfield_desc = {
.name = RTE_PMD_BNXT_CFA_CODE_DYNFIELD_NAME,
.size = sizeof(bnxt_cfa_code_dynfield_t),
.align = __alignof__(bnxt_cfa_code_dynfield_t),
};
bnxt_cfa_code_dynfield_offset =
rte_mbuf_dynfield_register(&bnxt_cfa_code_dynfield_desc);
if (bnxt_cfa_code_dynfield_offset < 0) {
PMD_DRV_LOG(ERR,
"Failed to register mbuf field for TruFlow mark\n");
return -rte_errno;
}
}
net/bnxt: refactor init/uninit Move all the individual driver fields allocation routines to one routine - bnxt_drv_init(). This houses all such routines where memory needs to be allocated once during the driver's lifetime and does not need to be torn down during error recovery. Rename some function names in accordance with their functionality. bnxt_init_board() is doing nothing more than mapping the PCI bars, so rename it as such. Given that there is a bnxt_shutdown_nic that is called in dev_stop_op, rename it's counterpart - bnxt_init_chip() that is called in dev_start_op, to bnxt_start_nic. Also helps avoid confusion with some of the other bnxt_init_xxx routines. Rename bnxt_init_fw() to bnxt_get_config() as that is what that routine is doing mostly functionality wise. Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-01-14 02:19:14 +00:00
rc = bnxt_map_pci_bars(eth_dev);
net/bnxt: add hardware resource manager init code Start adding support to use the HWRM API. Hardware Resource Manager or HWRM in short, is a set of API provided by the firmware running in the ASIC to manage the various resources. Initial commit just performs necessary HWRM queries for init, then fails as before. Now that struct bnxt is non-zero size, we can set dev_private_size correctly. The used HWRM calls so far: bnxt_hwrm_func_qcaps: This command returns capabilities of a function. bnxt_hwrm_ver_get: This function is called by a driver to determine the HWRM interface version supported by the HWRM firmware, the version of HWRM firmware implementation, the name of HWRM firmware, the versions of other embedded firmwares, and the names of other embedded firmwares, etc. Gets the firmware version and interface specifications. Returns an error if the firmware on the device is not supported by the driver and ensures the response space is large enough for the largest possible response. bnxt_hwrm_queue_qportcfg: This function is called by a driver to query queue configuration of a port. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:02 +00:00
if (rc) {
net/bnxt: use dynamic log type This patch implements driver specific log type doing away with usage of RTE_LOG() for logging. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:55 +00:00
PMD_DRV_LOG(ERR,
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
"Failed to initialize board rc: %x\n", rc);
return rc;
net/bnxt: add hardware resource manager init code Start adding support to use the HWRM API. Hardware Resource Manager or HWRM in short, is a set of API provided by the firmware running in the ASIC to manage the various resources. Initial commit just performs necessary HWRM queries for init, then fails as before. Now that struct bnxt is non-zero size, we can set dev_private_size correctly. The used HWRM calls so far: bnxt_hwrm_func_qcaps: This command returns capabilities of a function. bnxt_hwrm_ver_get: This function is called by a driver to determine the HWRM interface version supported by the HWRM firmware, the version of HWRM firmware implementation, the name of HWRM firmware, the versions of other embedded firmwares, and the names of other embedded firmwares, etc. Gets the firmware version and interface specifications. Returns an error if the firmware on the device is not supported by the driver and ensures the response space is large enough for the largest possible response. bnxt_hwrm_queue_qportcfg: This function is called by a driver to query queue configuration of a port. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:02 +00:00
}
net/bnxt: fix allocation of PF info struct Fixes: 804e746c7b73 ("net/bnxt: add hardware resource manager init code") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 11:10:41 +00:00
rc = bnxt_alloc_pf_info(bp);
if (rc)
net/bnxt: refactor init/uninit Move all the individual driver fields allocation routines to one routine - bnxt_drv_init(). This houses all such routines where memory needs to be allocated once during the driver's lifetime and does not need to be torn down during error recovery. Rename some function names in accordance with their functionality. bnxt_init_board() is doing nothing more than mapping the PCI bars, so rename it as such. Given that there is a bnxt_shutdown_nic that is called in dev_stop_op, rename it's counterpart - bnxt_init_chip() that is called in dev_start_op, to bnxt_start_nic. Also helps avoid confusion with some of the other bnxt_init_xxx routines. Rename bnxt_init_fw() to bnxt_get_config() as that is what that routine is doing mostly functionality wise. Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-01-14 02:19:14 +00:00
return rc;
net/bnxt: fix allocation of PF info struct Fixes: 804e746c7b73 ("net/bnxt: add hardware resource manager init code") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 11:10:41 +00:00
net/bnxt: fix allocation of link info struct Dynamically alloc link info structure. Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 11:05:15 +00:00
rc = bnxt_alloc_link_info(bp);
if (rc)
net/bnxt: refactor init/uninit Move all the individual driver fields allocation routines to one routine - bnxt_drv_init(). This houses all such routines where memory needs to be allocated once during the driver's lifetime and does not need to be torn down during error recovery. Rename some function names in accordance with their functionality. bnxt_init_board() is doing nothing more than mapping the PCI bars, so rename it as such. Given that there is a bnxt_shutdown_nic that is called in dev_stop_op, rename it's counterpart - bnxt_init_chip() that is called in dev_start_op, to bnxt_start_nic. Also helps avoid confusion with some of the other bnxt_init_xxx routines. Rename bnxt_init_fw() to bnxt_get_config() as that is what that routine is doing mostly functionality wise. Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-01-14 02:19:14 +00:00
return rc;
net/bnxt: initialize parent PF information Add support to query parent PF information (MAC address, function ID, port ID and default VNIC) from firmware. Current firmware returns zero for parent default vnic, a temporary Wh+-specific workaround is included until that can be fixed. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:51 +00:00
rc = bnxt_alloc_parent_info(bp);
if (rc)
net/bnxt: refactor init/uninit Move all the individual driver fields allocation routines to one routine - bnxt_drv_init(). This houses all such routines where memory needs to be allocated once during the driver's lifetime and does not need to be torn down during error recovery. Rename some function names in accordance with their functionality. bnxt_init_board() is doing nothing more than mapping the PCI bars, so rename it as such. Given that there is a bnxt_shutdown_nic that is called in dev_stop_op, rename it's counterpart - bnxt_init_chip() that is called in dev_start_op, to bnxt_start_nic. Also helps avoid confusion with some of the other bnxt_init_xxx routines. Rename bnxt_init_fw() to bnxt_get_config() as that is what that routine is doing mostly functionality wise. Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-01-14 02:19:14 +00:00
return rc;
net/bnxt: fix allocation of link info struct Dynamically alloc link info structure. Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 11:05:15 +00:00
net/bnxt: add hardware resource manager init code Start adding support to use the HWRM API. Hardware Resource Manager or HWRM in short, is a set of API provided by the firmware running in the ASIC to manage the various resources. Initial commit just performs necessary HWRM queries for init, then fails as before. Now that struct bnxt is non-zero size, we can set dev_private_size correctly. The used HWRM calls so far: bnxt_hwrm_func_qcaps: This command returns capabilities of a function. bnxt_hwrm_ver_get: This function is called by a driver to determine the HWRM interface version supported by the HWRM firmware, the version of HWRM firmware implementation, the name of HWRM firmware, the versions of other embedded firmwares, and the names of other embedded firmwares, etc. Gets the firmware version and interface specifications. Returns an error if the firmware on the device is not supported by the driver and ensures the response space is large enough for the largest possible response. bnxt_hwrm_queue_qportcfg: This function is called by a driver to query queue configuration of a port. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:02 +00:00
rc = bnxt_alloc_hwrm_resources(bp);
if (rc) {
net/bnxt: use dynamic log type This patch implements driver specific log type doing away with usage of RTE_LOG() for logging. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:55 +00:00
PMD_DRV_LOG(ERR,
net/bnxt: fix memory allocation for command response Driver re-allocates memory for the command response buffer when the installed firmware version is newer (and has a larger max response length) than the version of HWRM that was used to build the PMD. This change helps to avoid the re-allocation by allocating the memory for the command response buffer with PAGE_SIZE. Coverity issue: 366256, 366204, 366180 Fixes: b7778e8a1c00 ("net/bnxt: refactor to properly allocate resources for PF/VF") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com>
2021-03-20 06:49:17 +00:00
"Failed to allocate response buffer rc: %x\n", rc);
net/bnxt: refactor init/uninit Move all the individual driver fields allocation routines to one routine - bnxt_drv_init(). This houses all such routines where memory needs to be allocated once during the driver's lifetime and does not need to be torn down during error recovery. Rename some function names in accordance with their functionality. bnxt_init_board() is doing nothing more than mapping the PCI bars, so rename it as such. Given that there is a bnxt_shutdown_nic that is called in dev_stop_op, rename it's counterpart - bnxt_init_chip() that is called in dev_start_op, to bnxt_start_nic. Also helps avoid confusion with some of the other bnxt_init_xxx routines. Rename bnxt_init_fw() to bnxt_get_config() as that is what that routine is doing mostly functionality wise. Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-01-14 02:19:14 +00:00
return rc;
net/bnxt: add hardware resource manager init code Start adding support to use the HWRM API. Hardware Resource Manager or HWRM in short, is a set of API provided by the firmware running in the ASIC to manage the various resources. Initial commit just performs necessary HWRM queries for init, then fails as before. Now that struct bnxt is non-zero size, we can set dev_private_size correctly. The used HWRM calls so far: bnxt_hwrm_func_qcaps: This command returns capabilities of a function. bnxt_hwrm_ver_get: This function is called by a driver to determine the HWRM interface version supported by the HWRM firmware, the version of HWRM firmware implementation, the name of HWRM firmware, the versions of other embedded firmwares, and the names of other embedded firmwares, etc. Gets the firmware version and interface specifications. Returns an error if the firmware on the device is not supported by the driver and ensures the response space is large enough for the largest possible response. bnxt_hwrm_queue_qportcfg: This function is called by a driver to query queue configuration of a port. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:02 +00:00
}
net/bnxt: fix allocation of LED config info Dynamically allocate bnxt_led_cfg. This helps reduces memory footprint of struct bnxt. Fixes: bb81e07323bb ("net/bnxt: support LED on/off") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 07:01:51 +00:00
rc = bnxt_alloc_leds_info(bp);
if (rc)
net/bnxt: refactor init/uninit Move all the individual driver fields allocation routines to one routine - bnxt_drv_init(). This houses all such routines where memory needs to be allocated once during the driver's lifetime and does not need to be torn down during error recovery. Rename some function names in accordance with their functionality. bnxt_init_board() is doing nothing more than mapping the PCI bars, so rename it as such. Given that there is a bnxt_shutdown_nic that is called in dev_stop_op, rename it's counterpart - bnxt_init_chip() that is called in dev_start_op, to bnxt_start_nic. Also helps avoid confusion with some of the other bnxt_init_xxx routines. Rename bnxt_init_fw() to bnxt_get_config() as that is what that routine is doing mostly functionality wise. Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-01-14 02:19:14 +00:00
return rc;
net/bnxt: fix allocation of LED config info Dynamically allocate bnxt_led_cfg. This helps reduces memory footprint of struct bnxt. Fixes: bb81e07323bb ("net/bnxt: support LED on/off") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 07:01:51 +00:00
net/bnxt: fix allocation of COS queue info Fixes: 698aa7e95325 ("net/bnxt: add code to determine the Tx COS queue") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 18:07:32 +00:00
rc = bnxt_alloc_cos_queues(bp);
if (rc)
net/bnxt: refactor init/uninit Move all the individual driver fields allocation routines to one routine - bnxt_drv_init(). This houses all such routines where memory needs to be allocated once during the driver's lifetime and does not need to be torn down during error recovery. Rename some function names in accordance with their functionality. bnxt_init_board() is doing nothing more than mapping the PCI bars, so rename it as such. Given that there is a bnxt_shutdown_nic that is called in dev_stop_op, rename it's counterpart - bnxt_init_chip() that is called in dev_start_op, to bnxt_start_nic. Also helps avoid confusion with some of the other bnxt_init_xxx routines. Rename bnxt_init_fw() to bnxt_get_config() as that is what that routine is doing mostly functionality wise. Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-01-14 02:19:14 +00:00
return rc;
net/bnxt: fix allocation of COS queue info Fixes: 698aa7e95325 ("net/bnxt: add code to determine the Tx COS queue") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 18:07:32 +00:00
net/bnxt: fix lock init and destroy Invoking init/uninit locks in init_resources and uninit_resources would end up initializing and destroying locks on every port start stop which is not desired. Move the 2 routines to dev_init and dev_close respectively as locks need to be initialized and destroyed only once during the lifetime of the driver. Fixes: 1cb3d39a48f7 ("net/bnxt: synchronize between flow related functions") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-24 09:37:32 +00:00
rc = bnxt_init_locks(bp);
net/bnxt: refactor init/uninit Move all the individual driver fields allocation routines to one routine - bnxt_drv_init(). This houses all such routines where memory needs to be allocated once during the driver's lifetime and does not need to be torn down during error recovery. Rename some function names in accordance with their functionality. bnxt_init_board() is doing nothing more than mapping the PCI bars, so rename it as such. Given that there is a bnxt_shutdown_nic that is called in dev_stop_op, rename it's counterpart - bnxt_init_chip() that is called in dev_start_op, to bnxt_start_nic. Also helps avoid confusion with some of the other bnxt_init_xxx routines. Rename bnxt_init_fw() to bnxt_get_config() as that is what that routine is doing mostly functionality wise. Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-01-14 02:19:14 +00:00
if (rc)
return rc;
rc = bnxt_alloc_switch_domain(bp);
if (rc)
return rc;
return rc;
}
static int
bnxt_dev_init(struct rte_eth_dev *eth_dev, void *params __rte_unused)
{
struct rte_pci_device *pci_dev = RTE_ETH_DEV_TO_PCI(eth_dev);
static int version_printed;
struct bnxt *bp;
int rc;
if (version_printed++ == 0)
PMD_DRV_LOG(INFO, "%s\n", bnxt_version);
eth_dev->dev_ops = &bnxt_dev_ops;
eth_dev->rx_queue_count = bnxt_rx_queue_count_op;
eth_dev->rx_descriptor_status = bnxt_rx_descriptor_status_op;
eth_dev->tx_descriptor_status = bnxt_tx_descriptor_status_op;
eth_dev->rx_pkt_burst = &bnxt_recv_pkts;
eth_dev->tx_pkt_burst = &bnxt_xmit_pkts;
/*
* For secondary processes, we don't initialise any further
* as primary has already done this work.
*/
if (rte_eal_process_type() != RTE_PROC_PRIMARY)
return 0;
rte_eth_copy_pci_info(eth_dev, pci_dev);
eth_dev->data->dev_flags |= RTE_ETH_DEV_AUTOFILL_QUEUE_XSTATS;
bp = eth_dev->data->dev_private;
/* Parse dev arguments passed on when starting the DPDK application. */
net/bnxt: check kvargs parsing Check return value of rte_kvargs_process() Coverity issue: 357765 Fixes: ba404aacc5cf ("net/bnxt: set maximum flow count") Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Fixes: 7b0940653720 ("net/bnxt: support host memory based TruFlow") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Lance Richardson <lance.richardson@broadcom.com>
2021-03-17 01:08:40 +00:00
rc = bnxt_parse_dev_args(bp, pci_dev->device.devargs);
if (rc)
goto error_free;
net/bnxt: refactor init/uninit Move all the individual driver fields allocation routines to one routine - bnxt_drv_init(). This houses all such routines where memory needs to be allocated once during the driver's lifetime and does not need to be torn down during error recovery. Rename some function names in accordance with their functionality. bnxt_init_board() is doing nothing more than mapping the PCI bars, so rename it as such. Given that there is a bnxt_shutdown_nic that is called in dev_stop_op, rename it's counterpart - bnxt_init_chip() that is called in dev_start_op, to bnxt_start_nic. Also helps avoid confusion with some of the other bnxt_init_xxx routines. Rename bnxt_init_fw() to bnxt_get_config() as that is what that routine is doing mostly functionality wise. Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-01-14 02:19:14 +00:00
rc = bnxt_drv_init(eth_dev);
net/bnxt: fix lock init and destroy Invoking init/uninit locks in init_resources and uninit_resources would end up initializing and destroying locks on every port start stop which is not desired. Move the 2 routines to dev_init and dev_close respectively as locks need to be initialized and destroyed only once during the lifetime of the driver. Fixes: 1cb3d39a48f7 ("net/bnxt: synchronize between flow related functions") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-12-24 09:37:32 +00:00
if (rc)
goto error_free;
net/bnxt: handle reset notify async event from FW When the FW upgrade is initiated the current instance of FW issues a HWRM_ASYNC_EVENT_CMPL_EVENT_ID_RESET_NOTIFY async notification to the driver. On receiving this notification, the PMD shall quiesce itself and poll on the HWRM_VER_GET FW command at regular intervals. Once the VER_GET command succeeds, the driver should go through the rediscovery process and re-initialize the device. Also register with FW for the reset notify async event. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:23 +00:00
rc = bnxt_init_resources(bp, false);
net/bnxt: add hardware resource manager init code Start adding support to use the HWRM API. Hardware Resource Manager or HWRM in short, is a set of API provided by the firmware running in the ASIC to manage the various resources. Initial commit just performs necessary HWRM queries for init, then fails as before. Now that struct bnxt is non-zero size, we can set dev_private_size correctly. The used HWRM calls so far: bnxt_hwrm_func_qcaps: This command returns capabilities of a function. bnxt_hwrm_ver_get: This function is called by a driver to determine the HWRM interface version supported by the HWRM firmware, the version of HWRM firmware implementation, the name of HWRM firmware, the versions of other embedded firmwares, and the names of other embedded firmwares, etc. Gets the firmware version and interface specifications. Returns an error if the firmware on the device is not supported by the driver and ensures the response space is large enough for the largest possible response. bnxt_hwrm_queue_qportcfg: This function is called by a driver to query queue configuration of a port. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:02 +00:00
if (rc)
goto error_free;
net/bnxt: reset function earlier in initialization Move function reset to beginnng of initialization sequence. Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-06-02 17:42:40 +00:00
net/bnxt: fix extended port counter statistics 1. refactor stats allocation code to new routine 2. check for extended statistics support depends on "hwrm_spec_code" which is set in bnxt_hwrm_ver_get called later. Hence we were never querying extended port stats as flags field was not updated. Fixed this by moving the stats allocation after the call to bnxt_hwrm_ver_get. 3. we were incorrectly passing the host address used for port statistics to PORT_QSTATS_EXT command. Fixed this by passing the correct extended stats address. Fixes: f55e12f33416 ("net/bnxt: support extended port counters") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-07-17 10:41:25 +00:00
rc = bnxt_alloc_stats_mem(bp);
if (rc)
goto error_free;
net/bnxt: use dynamic log type This patch implements driver specific log type doing away with usage of RTE_LOG() for logging. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-01-26 17:31:55 +00:00
PMD_DRV_LOG(INFO,
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
DRV_MODULE_NAME "found at mem %" PRIX64 ", node addr %pM\n",
pci_dev->mem_resource[0].phys_addr,
pci_dev->mem_resource[0].addr);
net/bnxt: refactor to properly allocate resources for PF/VF 1) Move the function reset to bnxt_dev_init. On the same lines, setup, enable and request interrupt to init path. Memory allocation is also being done in the init path. 2) After a function reset, configure the VFs. Distribute resources evenly between all functions (PF and VF) for now. In the future, this should be controllable. 3) The bnxt_vf_info and bnxt_pf_info had lot of duplication. Move the common items to struct bnxt. And only unique items specific to PF remain in the struct bnxt_pf_info. 4) Program the firmware to allow certain commands sent by a VF. Disallowing these will prevent clean VF driver cleanup. 5) Since PF/VF need to allocate resources from a pool in the hardware, use func_qcaps and func_qcfg to appropriately query the capabilities and available resources. 6) If a PF is being initialized and no VFs are allocated, explicitly call func_cfg to allocate the resources. 7) Once resources are requested from the firmware, update local copy of resource count in struct bnxt only after sending the func_qcfg to make sure the allocation request in the firmware went through. The changes in this patch will be used by the subsequent patches to allow proper initialization of PF/VF instance. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:06:59 +00:00
net/bnxt: add driver register/unregister Move init() cleanup into uninit() function Fix .dev_private_size New HWRM calls: bnxt_hwrm_func_driver_register: This command is used by the function driver to register its information with the HWRM. bnxt_hwrm_func_driver_unregister: This command is used by the function driver to unregister with the HWRM. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:03 +00:00
return 0;
net/bnxt: add driver for Broadcom NetXtreme-C devices This patch adds the initial skeleton for bnxt driver along with the nic guide, and ties the driver into the build system. At this point, the driver simply fails init. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com> [Release Note Addition] Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2016-06-15 21:23:01 +00:00
net/bnxt: add hardware resource manager init code Start adding support to use the HWRM API. Hardware Resource Manager or HWRM in short, is a set of API provided by the firmware running in the ASIC to manage the various resources. Initial commit just performs necessary HWRM queries for init, then fails as before. Now that struct bnxt is non-zero size, we can set dev_private_size correctly. The used HWRM calls so far: bnxt_hwrm_func_qcaps: This command returns capabilities of a function. bnxt_hwrm_ver_get: This function is called by a driver to determine the HWRM interface version supported by the HWRM firmware, the version of HWRM firmware implementation, the name of HWRM firmware, the versions of other embedded firmwares, and the names of other embedded firmwares, etc. Gets the firmware version and interface specifications. Returns an error if the firmware on the device is not supported by the driver and ensures the response space is large enough for the largest possible response. bnxt_hwrm_queue_qportcfg: This function is called by a driver to query queue configuration of a port. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com>
2016-06-15 21:23:02 +00:00
error_free:
drivers/net: do not use ethdev driver Signed-off-by: Jan Blunck <jblunck@infradead.org> Signed-off-by: Gaetan Rivet <gaetan.rivet@6wind.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2017-04-11 15:44:24 +00:00
bnxt_dev_uninit(eth_dev);
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
return rc;
}
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
static void bnxt_free_ctx_mem_buf(struct bnxt_ctx_mem_buf_info *ctx)
{
if (!ctx)
return;
if (ctx->va)
rte_free(ctx->va);
ctx->va = NULL;
ctx->dma = RTE_BAD_IOVA;
ctx->ctx_id = BNXT_CTX_VAL_INVAL;
}
static void bnxt_unregister_fc_ctx_mem(struct bnxt *bp)
{
bnxt_hwrm_cfa_counter_cfg(bp, BNXT_DIR_RX,
CFA_COUNTER_CFG_IN_COUNTER_TYPE_FC,
net/bnxt: fix allocation of flow stat related structs Consolidate flow stat related structs for performance improvement. The intention of this patch is to reduce the size struct bnxt which had grown because of recent changes and was impacting performance. Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 08:31:59 +00:00
bp->flow_stat->rx_fc_out_tbl.ctx_id,
bp->flow_stat->max_fc,
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
false);
bnxt_hwrm_cfa_counter_cfg(bp, BNXT_DIR_TX,
CFA_COUNTER_CFG_IN_COUNTER_TYPE_FC,
net/bnxt: fix allocation of flow stat related structs Consolidate flow stat related structs for performance improvement. The intention of this patch is to reduce the size struct bnxt which had grown because of recent changes and was impacting performance. Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 08:31:59 +00:00
bp->flow_stat->tx_fc_out_tbl.ctx_id,
bp->flow_stat->max_fc,
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
false);
net/bnxt: fix allocation of flow stat related structs Consolidate flow stat related structs for performance improvement. The intention of this patch is to reduce the size struct bnxt which had grown because of recent changes and was impacting performance. Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 08:31:59 +00:00
if (bp->flow_stat->rx_fc_in_tbl.ctx_id != BNXT_CTX_VAL_INVAL)
bnxt_hwrm_ctx_unrgtr(bp, bp->flow_stat->rx_fc_in_tbl.ctx_id);
bp->flow_stat->rx_fc_in_tbl.ctx_id = BNXT_CTX_VAL_INVAL;
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
net/bnxt: fix allocation of flow stat related structs Consolidate flow stat related structs for performance improvement. The intention of this patch is to reduce the size struct bnxt which had grown because of recent changes and was impacting performance. Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 08:31:59 +00:00
if (bp->flow_stat->rx_fc_out_tbl.ctx_id != BNXT_CTX_VAL_INVAL)
bnxt_hwrm_ctx_unrgtr(bp, bp->flow_stat->rx_fc_out_tbl.ctx_id);
bp->flow_stat->rx_fc_out_tbl.ctx_id = BNXT_CTX_VAL_INVAL;
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
net/bnxt: fix allocation of flow stat related structs Consolidate flow stat related structs for performance improvement. The intention of this patch is to reduce the size struct bnxt which had grown because of recent changes and was impacting performance. Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 08:31:59 +00:00
if (bp->flow_stat->tx_fc_in_tbl.ctx_id != BNXT_CTX_VAL_INVAL)
bnxt_hwrm_ctx_unrgtr(bp, bp->flow_stat->tx_fc_in_tbl.ctx_id);
bp->flow_stat->tx_fc_in_tbl.ctx_id = BNXT_CTX_VAL_INVAL;
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
net/bnxt: fix allocation of flow stat related structs Consolidate flow stat related structs for performance improvement. The intention of this patch is to reduce the size struct bnxt which had grown because of recent changes and was impacting performance. Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 08:31:59 +00:00
if (bp->flow_stat->tx_fc_out_tbl.ctx_id != BNXT_CTX_VAL_INVAL)
bnxt_hwrm_ctx_unrgtr(bp, bp->flow_stat->tx_fc_out_tbl.ctx_id);
bp->flow_stat->tx_fc_out_tbl.ctx_id = BNXT_CTX_VAL_INVAL;
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
}
static void bnxt_uninit_fc_ctx_mem(struct bnxt *bp)
{
bnxt_unregister_fc_ctx_mem(bp);
net/bnxt: fix allocation of flow stat related structs Consolidate flow stat related structs for performance improvement. The intention of this patch is to reduce the size struct bnxt which had grown because of recent changes and was impacting performance. Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 08:31:59 +00:00
bnxt_free_ctx_mem_buf(&bp->flow_stat->rx_fc_in_tbl);
bnxt_free_ctx_mem_buf(&bp->flow_stat->rx_fc_out_tbl);
bnxt_free_ctx_mem_buf(&bp->flow_stat->tx_fc_in_tbl);
bnxt_free_ctx_mem_buf(&bp->flow_stat->tx_fc_out_tbl);
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
}
static void bnxt_uninit_ctx_mem(struct bnxt *bp)
{
net/bnxt: fix allocation of flow stat related structs Consolidate flow stat related structs for performance improvement. The intention of this patch is to reduce the size struct bnxt which had grown because of recent changes and was impacting performance. Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2020-05-15 08:31:59 +00:00
if (BNXT_FLOW_XSTATS_EN(bp))
bnxt_uninit_fc_ctx_mem(bp);
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
}
net/bnxt: log firmware status on early init failure Firmware reports any fatal error (either ASIC or Firmware related) via a new status register. This status register can provide more detailed information about the firmware errors, especially if error occurs before HWRM_VER_GET is issued. Attempt to map this register if it is present and check for firmware status when VER_GET command fails. Refactored the code to allocate the "bp->recovery_info" structure in bnxt_init_fw() instead of doing in bnxt_hwrm_error_recovery_qcfg(). Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-04-21 21:33:50 +00:00
static void
bnxt_free_error_recovery_info(struct bnxt *bp)
{
rte_free(bp->recovery_info);
bp->recovery_info = NULL;
bp->fw_cap &= ~BNXT_FW_CAP_ERROR_RECOVERY;
}
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
static int
net/bnxt: handle reset notify async event from FW When the FW upgrade is initiated the current instance of FW issues a HWRM_ASYNC_EVENT_CMPL_EVENT_ID_RESET_NOTIFY async notification to the driver. On receiving this notification, the PMD shall quiesce itself and poll on the HWRM_VER_GET FW command at regular intervals. Once the VER_GET command succeeds, the driver should go through the rediscovery process and re-initialize the device. Also register with FW for the reset notify async event. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:23 +00:00
bnxt_uninit_resources(struct bnxt *bp, bool reconfig_dev)
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
{
int rc;
bnxt_free_int(bp);
net/bnxt: handle reset notify async event from FW When the FW upgrade is initiated the current instance of FW issues a HWRM_ASYNC_EVENT_CMPL_EVENT_ID_RESET_NOTIFY async notification to the driver. On receiving this notification, the PMD shall quiesce itself and poll on the HWRM_VER_GET FW command at regular intervals. Once the VER_GET command succeeds, the driver should go through the rediscovery process and re-initialize the device. Also register with FW for the reset notify async event. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:23 +00:00
bnxt_free_mem(bp, reconfig_dev);
net/bnxt: fix PF support in SR-IOV mode 1. Implement HWRM_FUNC_VF_RESOURCE_CFG command and use it to reserve resources for VFs when NEW RM is enabled. 2. Invoke PF’s FUNC_CFG before configuring VFs resources. 3. Don’t consider max_rx_em_flows in max_l2_ctx calculation when VFs are configured. 4. Issue HWRM_FUNC_QCFG instead of HWRM_FUNC_QCAPS to find out the actual allocated resources for VF. 5. Don’t add random mac to the VF. 6. Handle completion type CMPL_BASE_TYPE_HWRM_FWD_REQ instead of CMPL_BASE_TYPE_HWRM_FWD_RESP. 7. Don't enable HWRM_FUNC_DRV_RGTR_INPUT_FLAGS_FWD_NONE_MODE when the list of HWRM commands that needs to be forwarded to the PF is specified in HWRM_FUNC_DRV_RGTR. 8. Update the HWRM commands list that can be forwarded to the PF. Fixes: b7778e8a1c00 ("net/bnxt: refactor to properly allocate resources for PF/VF") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-10-09 11:11:21 +00:00
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
bnxt_hwrm_func_buf_unrgtr(bp);
net/bnxt: fix resource cleanup Fix resource cleanup in port close. Once the pointers are freed, set them to NULL. Make sure access to the pointers is validated before use. Fixes: bb81e07323bb ("net/bnxt: support LED on/off") Fixes: 804e746c7b73 ("net/bnxt: add hardware resource manager init code") Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Fixes: 698aa7e95325 ("net/bnxt: add code to determine the Tx COS queue") Fixes: 322bd6e70272 ("net/bnxt: add port representor infrastructure") Fixes: 0bf5a0b5ebb8 ("net/bnxt: add a failure log") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-04-08 20:17:47 +00:00
if (bp->pf != NULL) {
rte_free(bp->pf->vf_req_buf);
bp->pf->vf_req_buf = NULL;
}
net/bnxt: fix PF support in SR-IOV mode 1. Implement HWRM_FUNC_VF_RESOURCE_CFG command and use it to reserve resources for VFs when NEW RM is enabled. 2. Invoke PF’s FUNC_CFG before configuring VFs resources. 3. Don’t consider max_rx_em_flows in max_l2_ctx calculation when VFs are configured. 4. Issue HWRM_FUNC_QCFG instead of HWRM_FUNC_QCAPS to find out the actual allocated resources for VF. 5. Don’t add random mac to the VF. 6. Handle completion type CMPL_BASE_TYPE_HWRM_FWD_REQ instead of CMPL_BASE_TYPE_HWRM_FWD_RESP. 7. Don't enable HWRM_FUNC_DRV_RGTR_INPUT_FLAGS_FWD_NONE_MODE when the list of HWRM commands that needs to be forwarded to the PF is specified in HWRM_FUNC_DRV_RGTR. 8. Update the HWRM commands list that can be forwarded to the PF. Fixes: b7778e8a1c00 ("net/bnxt: refactor to properly allocate resources for PF/VF") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-10-09 11:11:21 +00:00
net/bnxt: prevent device access when device is in reset Refactor init and uninit functions so that the driver can fail the eth_dev_ops callbacks and accessing Tx and Rx queues when device is in reset or in error state. Transmit and receive queues are freed during reset cleanup and reallocated during recovery. So we block all data path handling in this state. The eth_dev dev_started field is updated depending on the status of the device. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Santoshkumar Karanappa Rastapur <santosh.rastapur@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:22 +00:00
rc = bnxt_hwrm_func_driver_unregister(bp, 0);
bp->flags &= ~BNXT_FLAG_REGISTERED;
bnxt_free_ctx_mem(bp);
net/bnxt: query firmware error recovery capabilities In Driver initiated error recovery process, driver has to know about the registers offset and values to initiate FW reset. The HWRM command HWRM_ERROR_RECOVERY_QCFG is used to obtain all the registers and values required to initiate FW reset. This command response includes FW heart_beat register, health status register, Error counter register, register offsets and values to do chip reset if firmware crashes and becomes unresponsive. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:26 +00:00
if (!reconfig_dev) {
net/bnxt: handle reset notify async event from FW When the FW upgrade is initiated the current instance of FW issues a HWRM_ASYNC_EVENT_CMPL_EVENT_ID_RESET_NOTIFY async notification to the driver. On receiving this notification, the PMD shall quiesce itself and poll on the HWRM_VER_GET FW command at regular intervals. Once the VER_GET command succeeds, the driver should go through the rediscovery process and re-initialize the device. Also register with FW for the reset notify async event. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:23 +00:00
bnxt_free_hwrm_resources(bp);
net/bnxt: log firmware status on early init failure Firmware reports any fatal error (either ASIC or Firmware related) via a new status register. This status register can provide more detailed information about the firmware errors, especially if error occurs before HWRM_VER_GET is issued. Attempt to map this register if it is present and check for firmware status when VER_GET command fails. Refactored the code to allocate the "bp->recovery_info" structure in bnxt_init_fw() instead of doing in bnxt_hwrm_error_recovery_qcfg(). Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-04-21 21:33:50 +00:00
bnxt_free_error_recovery_info(bp);
net/bnxt: query firmware error recovery capabilities In Driver initiated error recovery process, driver has to know about the registers offset and values to initiate FW reset. The HWRM command HWRM_ERROR_RECOVERY_QCFG is used to obtain all the registers and values required to initiate FW reset. This command response includes FW heart_beat register, health status register, Error counter register, register offsets and values to do chip reset if firmware crashes and becomes unresponsive. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-10-02 01:23:26 +00:00
}
net/bnxt: add flow stats in extended stats This patch allows to display flow stats in extended stats. To do this, DMA-able memory is registered with the FW during device initialization. Then the driver uses an alarm thread to query the per flow stats using the HWRM_CFA_COUNTER_QSTATS HWRM command at regular intervals and stores it locally which will be displayed when the application queries the xstats. The DMA-able memory is unregistered during driver cleanup. This functionality can be enabled using the flow-xstat devarg and will be disabled by default. The intention behind this is to allow stats to be displayed for all the flows in one shot instead of querying one at a time. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
2020-04-16 04:27:01 +00:00
bnxt_uninit_ctx_mem(bp);
net/bnxt: fix crash during close We are freeing flow_stats a little early. This results in a segfault when the driver accesses the members during cleanup. Move the call to bnxt_free_flow_stats_info() to prevent this. Fixes: 02a95625fe9c ("net/bnxt: add flow stats in extended stats") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-05-22 21:27:31 +00:00
bnxt_free_flow_stats_info(bp);
net/bnxt: fix resource cleanup Fix resource cleanup in port close. Once the pointers are freed, set them to NULL. Make sure access to the pointers is validated before use. Fixes: bb81e07323bb ("net/bnxt: support LED on/off") Fixes: 804e746c7b73 ("net/bnxt: add hardware resource manager init code") Fixes: 1d0704f4d793 ("net/bnxt: add device configure operation") Fixes: 698aa7e95325 ("net/bnxt: add code to determine the Tx COS queue") Fixes: 322bd6e70272 ("net/bnxt: add port representor infrastructure") Fixes: 0bf5a0b5ebb8 ("net/bnxt: add a failure log") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2021-04-08 20:17:47 +00:00
if (bp->rep_info != NULL)
bnxt_free_switch_domain(bp);
net/bnxt: support port representor data path Added code to support Tx/Rx from a VF representor port. The VF-reps use the RX/TX rings of the Trusted VF/PF. For each VF-rep, the Trusted VF/PF driver issues a VFR_ALLOC FW cmd that returns "cfa_code" and "cfa_action" values. The FW sets up the filter tables in such a way that VF traffic by default (in absence of other rules) gets punted to the parent function i.e. either the Trusted VF or the PF. The cfa_code value in the RX-compl informs the driver of the source VF. For traffic being transmitted from the VF-rep, the TX BD is tagged with a cfa_action value that informs the HW to punt it to the corresponding VF. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:49 +00:00
bnxt_free_rep_info(bp);
net/bnxt: support PTP for Thor On Thor, direct access to PTP registers (via GRC) is not supported. Driver must use HWRM to access the timestamp information. Vectorized Rx/Tx cannot be enabled if RTE_LIBRTE_IEEE1588=y. Remove the PTP flags handling code from the vector Rx path. Add support to read tx timestamp value and the time from the timesync clock. On Thor, Rx timestamps are provided directly in the Rx completion records to the driver. Only 32 bits of the timestamp is present in the completion. Driver needs to read the current 48 bit free running timer using the HWRM_PORT_TS_QUERY command and combine the upper 16 bits from the HWRM response with the lower 32 bits in the Rx completion to produce the 48 bit timestamp for the Rx packet. Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2019-10-02 01:23:35 +00:00
rte_free(bp->ptp_cfg);
bp->ptp_cfg = NULL;
net/bnxt: add driver for Broadcom NetXtreme-C devices This patch adds the initial skeleton for bnxt driver along with the nic guide, and ties the driver into the build system. At this point, the driver simply fails init. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com> [Release Note Addition] Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2016-06-15 21:23:01 +00:00
return rc;
}
static int
net/bnxt: fix close operation We are not cleaning up all the memory and also not unregistering the driver during device close operation. This patch fixes the issue. Fixes: 893074951314 ("net/bnxt: free memory in close operation") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-06-28 20:15:31 +00:00
bnxt_dev_uninit(struct rte_eth_dev *eth_dev)
{
net/bnxt: support xstats get by id This patch adds support for xstats_get_by_id/xstats_get_names_by_id. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-09-28 21:43:32 +00:00
if (rte_eal_process_type() != RTE_PROC_PRIMARY)
return -EPERM;
net/bnxt: fix close operation We are not cleaning up all the memory and also not unregistering the driver during device close operation. This patch fixes the issue. Fixes: 893074951314 ("net/bnxt: free memory in close operation") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2018-06-28 20:15:31 +00:00
PMD_DRV_LOG(DEBUG, "Calling Device uninit\n");
net/bnxt: fix device init error path 1. bnxt_dev_init() invokes bnxt_dev_uninit() on failure. So there is no need to do individual function cleanups in failure path. 2. rearrange the check for primary process to remove an unwanted goto. 3. fix to invoke bnxt_hwrm_func_buf_unrgtr() in bnxt_dev_uninit() when it is needed. Fixes: b7778e8a1c00 ("net/bnxt: refactor to properly allocate resources for PF/VF") Cc: stable@dpdk.org Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
2019-07-18 03:36:00 +00:00
net/bnxt: release port upon close Set RTE_ETH_DEV_CLOSE_REMOVE upon probe so all the private resources for the port can be freed by rte_eth_dev_close(). Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-01-16 13:04:54 +00:00
if (eth_dev->state != RTE_ETH_DEV_UNUSED)
net/bnxt: fix crash when closing This patch fixes segfault encountered during dev_uninit/close routine. KNI sample app can be used to reproduce the issue. Fixes: c09f57b49c13 ("net/bnxt: add start/stop/link update operations") Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2016-09-29 17:03:44 +00:00
bnxt_dev_close_op(eth_dev);
net/bnxt: support hotplug This patch adds support for port hotplug framework. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2016-09-26 16:18:52 +00:00
net/bnxt: release port upon close Set RTE_ETH_DEV_CLOSE_REMOVE upon probe so all the private resources for the port can be freed by rte_eth_dev_close(). Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-01-16 13:04:54 +00:00
return 0;
net/bnxt: add driver for Broadcom NetXtreme-C devices This patch adds the initial skeleton for bnxt driver along with the nic guide, and ties the driver into the build system. At this point, the driver simply fails init. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com> [Release Note Addition] Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2016-06-15 21:23:01 +00:00
}
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
static int bnxt_pci_remove_dev_with_reps(struct rte_eth_dev *eth_dev)
{
struct bnxt *bp = eth_dev->data->dev_private;
struct rte_eth_dev *vf_rep_eth_dev;
int ret = 0, i;
if (!bp)
return -EINVAL;
for (i = 0; i < bp->num_reps; i++) {
vf_rep_eth_dev = bp->rep_info[i].vfr_eth_dev;
if (!vf_rep_eth_dev)
continue;
net/bnxt: refactor VFR port clean up When parent VF or PF ports are cleaned up, the child VF representor ports also need to be cleaned up. If not cleaned up, then deleting the parent VF shall result in not cleaning up the hardware rules and updating the firmware of VFR removal. The issue can occur even when application is exited without deleting VFR ports. Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Shahaji Bhosle <sbhosle@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-11 01:55:48 +00:00
PMD_DRV_LOG(DEBUG, "BNXT Port:%d VFR pci remove\n",
vf_rep_eth_dev->data->port_id);
net/bnxt: support representors on remote host domain In the Stingray use case, representors are conventionally run inside the SoC domain representing functions that are on the X86 domain. In order to support this mechanism of building representors for endpoints that are not in the same host domain, additional dev args have been in the PMD like so: rep-based-pf=<physical index> rep-is-pf=<VF=0 or PF=1> where `rep-based-pf` specifies the physical index of the base PF that the representor is derived off of. Since representor(s) can be created for endpoint PFs as well, rename struct bnxt_vf_representor to bnxt_representor and other such dev_ops and function names. devargs have also been extended to specify the exact CoS queue along with flow control enablement to be used for the conduit between the representor and the endpoint function. This is how a sample devargs would look with all the extended devargs -w 0000:06:02.0,host-based-truflow=1,representor=[1],rep-based-pf=8, rep-is-pf=1,rep-q-r2f=1,rep-fc-r2f=0,rep-q-f2r=1,rep-fc-f2r=1 Call CFA_PAIR_ALLOC only in case of Stingray instead of CFA_VFR_ALLOC. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:30 +00:00
rte_eth_dev_destroy(vf_rep_eth_dev, bnxt_representor_uninit);
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
}
net/bnxt: refactor VFR port clean up When parent VF or PF ports are cleaned up, the child VF representor ports also need to be cleaned up. If not cleaned up, then deleting the parent VF shall result in not cleaning up the hardware rules and updating the firmware of VFR removal. The issue can occur even when application is exited without deleting VFR ports. Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Shahaji Bhosle <sbhosle@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-11 01:55:48 +00:00
PMD_DRV_LOG(DEBUG, "BNXT Port:%d pci remove\n",
eth_dev->data->port_id);
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
ret = rte_eth_dev_destroy(eth_dev, bnxt_dev_uninit);
return ret;
}
net/bnxt: support port representor data path Added code to support Tx/Rx from a VF representor port. The VF-reps use the RX/TX rings of the Trusted VF/PF. For each VF-rep, the Trusted VF/PF driver issues a VFR_ALLOC FW cmd that returns "cfa_code" and "cfa_action" values. The FW sets up the filter tables in such a way that VF traffic by default (in absence of other rules) gets punted to the parent function i.e. either the Trusted VF or the PF. The cfa_code value in the RX-compl informs the driver of the source VF. For traffic being transmitted from the VF-rep, the TX BD is tagged with a cfa_action value that informs the HW to punt it to the corresponding VF. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:49 +00:00
static void bnxt_free_rep_info(struct bnxt *bp)
drivers/net: do not use ethdev driver Signed-off-by: Jan Blunck <jblunck@infradead.org> Signed-off-by: Gaetan Rivet <gaetan.rivet@6wind.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2017-04-11 15:44:24 +00:00
{
net/bnxt: support port representor data path Added code to support Tx/Rx from a VF representor port. The VF-reps use the RX/TX rings of the Trusted VF/PF. For each VF-rep, the Trusted VF/PF driver issues a VFR_ALLOC FW cmd that returns "cfa_code" and "cfa_action" values. The FW sets up the filter tables in such a way that VF traffic by default (in absence of other rules) gets punted to the parent function i.e. either the Trusted VF or the PF. The cfa_code value in the RX-compl informs the driver of the source VF. For traffic being transmitted from the VF-rep, the TX BD is tagged with a cfa_action value that informs the HW to punt it to the corresponding VF. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:49 +00:00
rte_free(bp->rep_info);
bp->rep_info = NULL;
rte_free(bp->cfa_code_map);
bp->cfa_code_map = NULL;
}
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
net/bnxt: support port representor data path Added code to support Tx/Rx from a VF representor port. The VF-reps use the RX/TX rings of the Trusted VF/PF. For each VF-rep, the Trusted VF/PF driver issues a VFR_ALLOC FW cmd that returns "cfa_code" and "cfa_action" values. The FW sets up the filter tables in such a way that VF traffic by default (in absence of other rules) gets punted to the parent function i.e. either the Trusted VF or the PF. The cfa_code value in the RX-compl informs the driver of the source VF. For traffic being transmitted from the VF-rep, the TX BD is tagged with a cfa_action value that informs the HW to punt it to the corresponding VF. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:49 +00:00
static int bnxt_init_rep_info(struct bnxt *bp)
{
int i = 0, rc;
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
net/bnxt: support port representor data path Added code to support Tx/Rx from a VF representor port. The VF-reps use the RX/TX rings of the Trusted VF/PF. For each VF-rep, the Trusted VF/PF driver issues a VFR_ALLOC FW cmd that returns "cfa_code" and "cfa_action" values. The FW sets up the filter tables in such a way that VF traffic by default (in absence of other rules) gets punted to the parent function i.e. either the Trusted VF or the PF. The cfa_code value in the RX-compl informs the driver of the source VF. For traffic being transmitted from the VF-rep, the TX BD is tagged with a cfa_action value that informs the HW to punt it to the corresponding VF. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:49 +00:00
if (bp->rep_info)
return 0;
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
net/bnxt: support port representor data path Added code to support Tx/Rx from a VF representor port. The VF-reps use the RX/TX rings of the Trusted VF/PF. For each VF-rep, the Trusted VF/PF driver issues a VFR_ALLOC FW cmd that returns "cfa_code" and "cfa_action" values. The FW sets up the filter tables in such a way that VF traffic by default (in absence of other rules) gets punted to the parent function i.e. either the Trusted VF or the PF. The cfa_code value in the RX-compl informs the driver of the source VF. For traffic being transmitted from the VF-rep, the TX BD is tagged with a cfa_action value that informs the HW to punt it to the corresponding VF. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:49 +00:00
bp->rep_info = rte_zmalloc("bnxt_rep_info",
sizeof(bp->rep_info[0]) * BNXT_MAX_VF_REPS,
0);
if (!bp->rep_info) {
PMD_DRV_LOG(ERR, "Failed to alloc memory for rep info\n");
return -ENOMEM;
}
bp->cfa_code_map = rte_zmalloc("bnxt_cfa_code_map",
sizeof(*bp->cfa_code_map) *
BNXT_MAX_CFA_CODE, 0);
if (!bp->cfa_code_map) {
PMD_DRV_LOG(ERR, "Failed to alloc memory for cfa_code_map\n");
bnxt_free_rep_info(bp);
return -ENOMEM;
}
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
net/bnxt: support port representor data path Added code to support Tx/Rx from a VF representor port. The VF-reps use the RX/TX rings of the Trusted VF/PF. For each VF-rep, the Trusted VF/PF driver issues a VFR_ALLOC FW cmd that returns "cfa_code" and "cfa_action" values. The FW sets up the filter tables in such a way that VF traffic by default (in absence of other rules) gets punted to the parent function i.e. either the Trusted VF or the PF. The cfa_code value in the RX-compl informs the driver of the source VF. For traffic being transmitted from the VF-rep, the TX BD is tagged with a cfa_action value that informs the HW to punt it to the corresponding VF. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:49 +00:00
for (i = 0; i < BNXT_MAX_CFA_CODE; i++)
bp->cfa_code_map[i] = BNXT_VF_IDX_INVALID;
rc = pthread_mutex_init(&bp->rep_info->vfr_lock, NULL);
if (rc) {
PMD_DRV_LOG(ERR, "Unable to initialize vfr_lock\n");
bnxt_free_rep_info(bp);
return rc;
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
}
net/bnxt: fix VF representor port add Fix VF representor port add when it's endpoint interface is down. While adding vf representor port to a bridge, vnic & svif information of vf representors endpoint(VF) would be needed to program default flow rules. However, if the endpoint interface is down when vf representor port is added, firmware will return invalid vnic & svif information. This patch fixes the problem by registering to DEFAULT_VNIC_CHANGE async event and once the async event is received, use the endpoint information(VF's fid) to fetch it's vnic & svif information and program the default flow rules. Fixes: 322bd6e70272 ("net/bnxt: add port representor infrastructure") Cc: stable@dpdk.org Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-29 14:04:59 +00:00
rc = pthread_mutex_init(&bp->rep_info->vfr_start_lock, NULL);
if (rc) {
PMD_DRV_LOG(ERR, "Unable to initialize vfr_start_lock\n");
bnxt_free_rep_info(bp);
return rc;
}
net/bnxt: support port representor data path Added code to support Tx/Rx from a VF representor port. The VF-reps use the RX/TX rings of the Trusted VF/PF. For each VF-rep, the Trusted VF/PF driver issues a VFR_ALLOC FW cmd that returns "cfa_code" and "cfa_action" values. The FW sets up the filter tables in such a way that VF traffic by default (in absence of other rules) gets punted to the parent function i.e. either the Trusted VF or the PF. The cfa_code value in the RX-compl informs the driver of the source VF. For traffic being transmitted from the VF-rep, the TX BD is tagged with a cfa_action value that informs the HW to punt it to the corresponding VF. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:49 +00:00
return rc;
}
static int bnxt_rep_port_probe(struct rte_pci_device *pci_dev,
net/bnxt: fix pass by reference Pass 'eth_da' pointer instead of pass by value to bnxt_rep_port_probe() Coverity issue: 360841 Fixes: 322bd6e70272 ("net/bnxt: add port representor infrastructure") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
2020-10-30 07:03:04 +00:00
struct rte_eth_devargs *eth_da,
net/bnxt: support representors on remote host domain In the Stingray use case, representors are conventionally run inside the SoC domain representing functions that are on the X86 domain. In order to support this mechanism of building representors for endpoints that are not in the same host domain, additional dev args have been in the PMD like so: rep-based-pf=<physical index> rep-is-pf=<VF=0 or PF=1> where `rep-based-pf` specifies the physical index of the base PF that the representor is derived off of. Since representor(s) can be created for endpoint PFs as well, rename struct bnxt_vf_representor to bnxt_representor and other such dev_ops and function names. devargs have also been extended to specify the exact CoS queue along with flow control enablement to be used for the conduit between the representor and the endpoint function. This is how a sample devargs would look with all the extended devargs -w 0000:06:02.0,host-based-truflow=1,representor=[1],rep-based-pf=8, rep-is-pf=1,rep-q-r2f=1,rep-fc-r2f=0,rep-q-f2r=1,rep-fc-f2r=1 Call CFA_PAIR_ALLOC only in case of Stingray instead of CFA_VFR_ALLOC. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:30 +00:00
struct rte_eth_dev *backing_eth_dev,
const char *dev_args)
net/bnxt: support port representor data path Added code to support Tx/Rx from a VF representor port. The VF-reps use the RX/TX rings of the Trusted VF/PF. For each VF-rep, the Trusted VF/PF driver issues a VFR_ALLOC FW cmd that returns "cfa_code" and "cfa_action" values. The FW sets up the filter tables in such a way that VF traffic by default (in absence of other rules) gets punted to the parent function i.e. either the Trusted VF or the PF. The cfa_code value in the RX-compl informs the driver of the source VF. For traffic being transmitted from the VF-rep, the TX BD is tagged with a cfa_action value that informs the HW to punt it to the corresponding VF. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:49 +00:00
{
struct rte_eth_dev *vf_rep_eth_dev;
char name[RTE_ETH_NAME_MAX_LEN];
struct bnxt *backing_bp;
uint16_t num_rep;
int i, ret = 0;
net/bnxt: fix resource leak Fix a potential resource leak in case of errors during dev args parsing during device probe. Fixes: 6dc83230b43b ("net/bnxt: support port representor data path") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
2020-10-20 23:24:28 +00:00
struct rte_kvargs *kvlist = NULL;
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
ethdev: introduce representor type To support more representor type, this patch introduces representor type enum. The enum is subject to be extended to support new representor in patches upcoming. For each devarg structure, only one type supported. Signed-off-by: Xueming Li <xuemingl@nvidia.com> Acked-by: Viacheslav Ovsiienko <viacheslavo@nvidia.com> Acked-by: Thomas Monjalon <thomas@monjalon.net> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Acked-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru> Acked-by: Hyong Youb Kim <hyonkim@cisco.com>
2021-03-11 13:13:24 +00:00
if (eth_da->type == RTE_ETH_REPRESENTOR_NONE)
return 0;
if (eth_da->type != RTE_ETH_REPRESENTOR_VF) {
PMD_DRV_LOG(ERR, "unsupported representor type %d\n",
eth_da->type);
return -ENOTSUP;
}
net/bnxt: fix pass by reference Pass 'eth_da' pointer instead of pass by value to bnxt_rep_port_probe() Coverity issue: 360841 Fixes: 322bd6e70272 ("net/bnxt: add port representor infrastructure") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
2020-10-30 07:03:04 +00:00
num_rep = eth_da->nb_representor_ports;
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
if (num_rep > BNXT_MAX_VF_REPS) {
PMD_DRV_LOG(ERR, "nb_representor_ports = %d > %d MAX VF REPS\n",
net/bnxt: support port representor data path Added code to support Tx/Rx from a VF representor port. The VF-reps use the RX/TX rings of the Trusted VF/PF. For each VF-rep, the Trusted VF/PF driver issues a VFR_ALLOC FW cmd that returns "cfa_code" and "cfa_action" values. The FW sets up the filter tables in such a way that VF traffic by default (in absence of other rules) gets punted to the parent function i.e. either the Trusted VF or the PF. The cfa_code value in the RX-compl informs the driver of the source VF. For traffic being transmitted from the VF-rep, the TX BD is tagged with a cfa_action value that informs the HW to punt it to the corresponding VF. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:49 +00:00
num_rep, BNXT_MAX_VF_REPS);
return -EINVAL;
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
}
net/bnxt: fix VFR cleanup during init failure If VF-rep port add fails for some reason, code was rolling back all ports added so far. With some applications, there is no need to do that. Just log failure message for the VF rep port add and continue. Also include RTE_MAX_ETH_PORTS value in the bounds check as one port will be taken by the uplink port anyway Fixes: 6dc83230b43b ("net/bnxt: support port representor data path") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-11 01:55:50 +00:00
if (num_rep >= RTE_MAX_ETHPORTS) {
net/bnxt: support port representor data path Added code to support Tx/Rx from a VF representor port. The VF-reps use the RX/TX rings of the Trusted VF/PF. For each VF-rep, the Trusted VF/PF driver issues a VFR_ALLOC FW cmd that returns "cfa_code" and "cfa_action" values. The FW sets up the filter tables in such a way that VF traffic by default (in absence of other rules) gets punted to the parent function i.e. either the Trusted VF or the PF. The cfa_code value in the RX-compl informs the driver of the source VF. For traffic being transmitted from the VF-rep, the TX BD is tagged with a cfa_action value that informs the HW to punt it to the corresponding VF. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:49 +00:00
PMD_DRV_LOG(ERR,
"nb_representor_ports = %d > %d MAX ETHPORTS\n",
num_rep, RTE_MAX_ETHPORTS);
return -EINVAL;
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
}
net/bnxt: support port representor data path Added code to support Tx/Rx from a VF representor port. The VF-reps use the RX/TX rings of the Trusted VF/PF. For each VF-rep, the Trusted VF/PF driver issues a VFR_ALLOC FW cmd that returns "cfa_code" and "cfa_action" values. The FW sets up the filter tables in such a way that VF traffic by default (in absence of other rules) gets punted to the parent function i.e. either the Trusted VF or the PF. The cfa_code value in the RX-compl informs the driver of the source VF. For traffic being transmitted from the VF-rep, the TX BD is tagged with a cfa_action value that informs the HW to punt it to the corresponding VF. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:49 +00:00
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
backing_bp = backing_eth_dev->data->dev_private;
if (!(BNXT_PF(backing_bp) || BNXT_VF_IS_TRUSTED(backing_bp))) {
PMD_DRV_LOG(ERR,
"Not a PF or trusted VF. No Representor support\n");
/* Returning an error is not an option.
* Applications are not handling this correctly
*/
net/bnxt: support port representor data path Added code to support Tx/Rx from a VF representor port. The VF-reps use the RX/TX rings of the Trusted VF/PF. For each VF-rep, the Trusted VF/PF driver issues a VFR_ALLOC FW cmd that returns "cfa_code" and "cfa_action" values. The FW sets up the filter tables in such a way that VF traffic by default (in absence of other rules) gets punted to the parent function i.e. either the Trusted VF or the PF. The cfa_code value in the RX-compl informs the driver of the source VF. For traffic being transmitted from the VF-rep, the TX BD is tagged with a cfa_action value that informs the HW to punt it to the corresponding VF. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:49 +00:00
return 0;
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
}
net/bnxt: support port representor data path Added code to support Tx/Rx from a VF representor port. The VF-reps use the RX/TX rings of the Trusted VF/PF. For each VF-rep, the Trusted VF/PF driver issues a VFR_ALLOC FW cmd that returns "cfa_code" and "cfa_action" values. The FW sets up the filter tables in such a way that VF traffic by default (in absence of other rules) gets punted to the parent function i.e. either the Trusted VF or the PF. The cfa_code value in the RX-compl informs the driver of the source VF. For traffic being transmitted from the VF-rep, the TX BD is tagged with a cfa_action value that informs the HW to punt it to the corresponding VF. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:49 +00:00
if (bnxt_init_rep_info(backing_bp))
return 0;
for (i = 0; i < num_rep; i++) {
net/bnxt: support representors on remote host domain In the Stingray use case, representors are conventionally run inside the SoC domain representing functions that are on the X86 domain. In order to support this mechanism of building representors for endpoints that are not in the same host domain, additional dev args have been in the PMD like so: rep-based-pf=<physical index> rep-is-pf=<VF=0 or PF=1> where `rep-based-pf` specifies the physical index of the base PF that the representor is derived off of. Since representor(s) can be created for endpoint PFs as well, rename struct bnxt_vf_representor to bnxt_representor and other such dev_ops and function names. devargs have also been extended to specify the exact CoS queue along with flow control enablement to be used for the conduit between the representor and the endpoint function. This is how a sample devargs would look with all the extended devargs -w 0000:06:02.0,host-based-truflow=1,representor=[1],rep-based-pf=8, rep-is-pf=1,rep-q-r2f=1,rep-fc-r2f=0,rep-q-f2r=1,rep-fc-f2r=1 Call CFA_PAIR_ALLOC only in case of Stingray instead of CFA_VFR_ALLOC. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:30 +00:00
struct bnxt_representor representor = {
net/bnxt: fix pass by reference Pass 'eth_da' pointer instead of pass by value to bnxt_rep_port_probe() Coverity issue: 360841 Fixes: 322bd6e70272 ("net/bnxt: add port representor infrastructure") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
2020-10-30 07:03:04 +00:00
.vf_id = eth_da->representor_ports[i],
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
.switch_domain_id = backing_bp->switch_domain_id,
net/bnxt: support port representor data path Added code to support Tx/Rx from a VF representor port. The VF-reps use the RX/TX rings of the Trusted VF/PF. For each VF-rep, the Trusted VF/PF driver issues a VFR_ALLOC FW cmd that returns "cfa_code" and "cfa_action" values. The FW sets up the filter tables in such a way that VF traffic by default (in absence of other rules) gets punted to the parent function i.e. either the Trusted VF or the PF. The cfa_code value in the RX-compl informs the driver of the source VF. For traffic being transmitted from the VF-rep, the TX BD is tagged with a cfa_action value that informs the HW to punt it to the corresponding VF. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:49 +00:00
.parent_dev = backing_eth_dev
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
};
if (representor.vf_id >= BNXT_MAX_VF_REPS) {
PMD_DRV_LOG(ERR, "VF-Rep id %d >= %d MAX VF ID\n",
representor.vf_id, BNXT_MAX_VF_REPS);
continue;
}
/* representor port net_bdf_port */
snprintf(name, sizeof(name), "net_%s_representor_%d",
net/bnxt: fix pass by reference Pass 'eth_da' pointer instead of pass by value to bnxt_rep_port_probe() Coverity issue: 360841 Fixes: 322bd6e70272 ("net/bnxt: add port representor infrastructure") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
2020-10-30 07:03:04 +00:00
pci_dev->device.name, eth_da->representor_ports[i]);
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
net/bnxt: support representors on remote host domain In the Stingray use case, representors are conventionally run inside the SoC domain representing functions that are on the X86 domain. In order to support this mechanism of building representors for endpoints that are not in the same host domain, additional dev args have been in the PMD like so: rep-based-pf=<physical index> rep-is-pf=<VF=0 or PF=1> where `rep-based-pf` specifies the physical index of the base PF that the representor is derived off of. Since representor(s) can be created for endpoint PFs as well, rename struct bnxt_vf_representor to bnxt_representor and other such dev_ops and function names. devargs have also been extended to specify the exact CoS queue along with flow control enablement to be used for the conduit between the representor and the endpoint function. This is how a sample devargs would look with all the extended devargs -w 0000:06:02.0,host-based-truflow=1,representor=[1],rep-based-pf=8, rep-is-pf=1,rep-q-r2f=1,rep-fc-r2f=0,rep-q-f2r=1,rep-fc-f2r=1 Call CFA_PAIR_ALLOC only in case of Stingray instead of CFA_VFR_ALLOC. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:30 +00:00
kvlist = rte_kvargs_parse(dev_args, bnxt_dev_args);
if (kvlist) {
/*
* Handler for "rep_is_pf" devarg.
eal: replace blacklist/whitelist options Replace -w / --pci-whitelist with -a / --allow options and --pci-blacklist with --block. The -b short option remains unchanged. Allow the old options for now, but print a nag warning since old options are deprecated. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Luca Boccassi <bluca@debian.org> Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
2020-11-10 22:55:40 +00:00
* Invoked as for ex: "-a 000:00:0d.0,
net/bnxt: support representors on remote host domain In the Stingray use case, representors are conventionally run inside the SoC domain representing functions that are on the X86 domain. In order to support this mechanism of building representors for endpoints that are not in the same host domain, additional dev args have been in the PMD like so: rep-based-pf=<physical index> rep-is-pf=<VF=0 or PF=1> where `rep-based-pf` specifies the physical index of the base PF that the representor is derived off of. Since representor(s) can be created for endpoint PFs as well, rename struct bnxt_vf_representor to bnxt_representor and other such dev_ops and function names. devargs have also been extended to specify the exact CoS queue along with flow control enablement to be used for the conduit between the representor and the endpoint function. This is how a sample devargs would look with all the extended devargs -w 0000:06:02.0,host-based-truflow=1,representor=[1],rep-based-pf=8, rep-is-pf=1,rep-q-r2f=1,rep-fc-r2f=0,rep-q-f2r=1,rep-fc-f2r=1 Call CFA_PAIR_ALLOC only in case of Stingray instead of CFA_VFR_ALLOC. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:30 +00:00
* rep-based-pf=<pf index> rep-is-pf=<VF=0 or PF=1>"
*/
net/bnxt: fix resource leak Fix a potential resource leak in case of errors during dev args parsing during device probe. Fixes: 6dc83230b43b ("net/bnxt: support port representor data path") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
2020-10-20 23:24:28 +00:00
ret = rte_kvargs_process(kvlist, BNXT_DEVARG_REP_IS_PF,
bnxt_parse_devarg_rep_is_pf,
(void *)&representor);
if (ret) {
ret = -EINVAL;
goto err;
}
net/bnxt: support representors on remote host domain In the Stingray use case, representors are conventionally run inside the SoC domain representing functions that are on the X86 domain. In order to support this mechanism of building representors for endpoints that are not in the same host domain, additional dev args have been in the PMD like so: rep-based-pf=<physical index> rep-is-pf=<VF=0 or PF=1> where `rep-based-pf` specifies the physical index of the base PF that the representor is derived off of. Since representor(s) can be created for endpoint PFs as well, rename struct bnxt_vf_representor to bnxt_representor and other such dev_ops and function names. devargs have also been extended to specify the exact CoS queue along with flow control enablement to be used for the conduit between the representor and the endpoint function. This is how a sample devargs would look with all the extended devargs -w 0000:06:02.0,host-based-truflow=1,representor=[1],rep-based-pf=8, rep-is-pf=1,rep-q-r2f=1,rep-fc-r2f=0,rep-q-f2r=1,rep-fc-f2r=1 Call CFA_PAIR_ALLOC only in case of Stingray instead of CFA_VFR_ALLOC. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:30 +00:00
/*
* Handler for "rep_based_pf" devarg.
eal: replace blacklist/whitelist options Replace -w / --pci-whitelist with -a / --allow options and --pci-blacklist with --block. The -b short option remains unchanged. Allow the old options for now, but print a nag warning since old options are deprecated. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Luca Boccassi <bluca@debian.org> Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
2020-11-10 22:55:40 +00:00
* Invoked as for ex: "-a 000:00:0d.0,
net/bnxt: support representors on remote host domain In the Stingray use case, representors are conventionally run inside the SoC domain representing functions that are on the X86 domain. In order to support this mechanism of building representors for endpoints that are not in the same host domain, additional dev args have been in the PMD like so: rep-based-pf=<physical index> rep-is-pf=<VF=0 or PF=1> where `rep-based-pf` specifies the physical index of the base PF that the representor is derived off of. Since representor(s) can be created for endpoint PFs as well, rename struct bnxt_vf_representor to bnxt_representor and other such dev_ops and function names. devargs have also been extended to specify the exact CoS queue along with flow control enablement to be used for the conduit between the representor and the endpoint function. This is how a sample devargs would look with all the extended devargs -w 0000:06:02.0,host-based-truflow=1,representor=[1],rep-based-pf=8, rep-is-pf=1,rep-q-r2f=1,rep-fc-r2f=0,rep-q-f2r=1,rep-fc-f2r=1 Call CFA_PAIR_ALLOC only in case of Stingray instead of CFA_VFR_ALLOC. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:30 +00:00
* rep-based-pf=<pf index> rep-is-pf=<VF=0 or PF=1>"
*/
net/bnxt: fix resource leak Fix a potential resource leak in case of errors during dev args parsing during device probe. Fixes: 6dc83230b43b ("net/bnxt: support port representor data path") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
2020-10-20 23:24:28 +00:00
ret = rte_kvargs_process(kvlist,
BNXT_DEVARG_REP_BASED_PF,
bnxt_parse_devarg_rep_based_pf,
(void *)&representor);
if (ret) {
ret = -EINVAL;
goto err;
}
net/bnxt: support representors on remote host domain In the Stingray use case, representors are conventionally run inside the SoC domain representing functions that are on the X86 domain. In order to support this mechanism of building representors for endpoints that are not in the same host domain, additional dev args have been in the PMD like so: rep-based-pf=<physical index> rep-is-pf=<VF=0 or PF=1> where `rep-based-pf` specifies the physical index of the base PF that the representor is derived off of. Since representor(s) can be created for endpoint PFs as well, rename struct bnxt_vf_representor to bnxt_representor and other such dev_ops and function names. devargs have also been extended to specify the exact CoS queue along with flow control enablement to be used for the conduit between the representor and the endpoint function. This is how a sample devargs would look with all the extended devargs -w 0000:06:02.0,host-based-truflow=1,representor=[1],rep-based-pf=8, rep-is-pf=1,rep-q-r2f=1,rep-fc-r2f=0,rep-q-f2r=1,rep-fc-f2r=1 Call CFA_PAIR_ALLOC only in case of Stingray instead of CFA_VFR_ALLOC. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:30 +00:00
/*
* Handler for "rep_based_pf" devarg.
eal: replace blacklist/whitelist options Replace -w / --pci-whitelist with -a / --allow options and --pci-blacklist with --block. The -b short option remains unchanged. Allow the old options for now, but print a nag warning since old options are deprecated. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Luca Boccassi <bluca@debian.org> Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
2020-11-10 22:55:40 +00:00
* Invoked as for ex: "-a 000:00:0d.0,
net/bnxt: support representors on remote host domain In the Stingray use case, representors are conventionally run inside the SoC domain representing functions that are on the X86 domain. In order to support this mechanism of building representors for endpoints that are not in the same host domain, additional dev args have been in the PMD like so: rep-based-pf=<physical index> rep-is-pf=<VF=0 or PF=1> where `rep-based-pf` specifies the physical index of the base PF that the representor is derived off of. Since representor(s) can be created for endpoint PFs as well, rename struct bnxt_vf_representor to bnxt_representor and other such dev_ops and function names. devargs have also been extended to specify the exact CoS queue along with flow control enablement to be used for the conduit between the representor and the endpoint function. This is how a sample devargs would look with all the extended devargs -w 0000:06:02.0,host-based-truflow=1,representor=[1],rep-based-pf=8, rep-is-pf=1,rep-q-r2f=1,rep-fc-r2f=0,rep-q-f2r=1,rep-fc-f2r=1 Call CFA_PAIR_ALLOC only in case of Stingray instead of CFA_VFR_ALLOC. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:30 +00:00
* rep-based-pf=<pf index> rep-is-pf=<VF=0 or PF=1>"
*/
net/bnxt: fix resource leak Fix a potential resource leak in case of errors during dev args parsing during device probe. Fixes: 6dc83230b43b ("net/bnxt: support port representor data path") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
2020-10-20 23:24:28 +00:00
ret = rte_kvargs_process(kvlist, BNXT_DEVARG_REP_Q_R2F,
bnxt_parse_devarg_rep_q_r2f,
(void *)&representor);
if (ret) {
ret = -EINVAL;
goto err;
}
net/bnxt: support representors on remote host domain In the Stingray use case, representors are conventionally run inside the SoC domain representing functions that are on the X86 domain. In order to support this mechanism of building representors for endpoints that are not in the same host domain, additional dev args have been in the PMD like so: rep-based-pf=<physical index> rep-is-pf=<VF=0 or PF=1> where `rep-based-pf` specifies the physical index of the base PF that the representor is derived off of. Since representor(s) can be created for endpoint PFs as well, rename struct bnxt_vf_representor to bnxt_representor and other such dev_ops and function names. devargs have also been extended to specify the exact CoS queue along with flow control enablement to be used for the conduit between the representor and the endpoint function. This is how a sample devargs would look with all the extended devargs -w 0000:06:02.0,host-based-truflow=1,representor=[1],rep-based-pf=8, rep-is-pf=1,rep-q-r2f=1,rep-fc-r2f=0,rep-q-f2r=1,rep-fc-f2r=1 Call CFA_PAIR_ALLOC only in case of Stingray instead of CFA_VFR_ALLOC. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:30 +00:00
/*
* Handler for "rep_based_pf" devarg.
eal: replace blacklist/whitelist options Replace -w / --pci-whitelist with -a / --allow options and --pci-blacklist with --block. The -b short option remains unchanged. Allow the old options for now, but print a nag warning since old options are deprecated. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Luca Boccassi <bluca@debian.org> Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
2020-11-10 22:55:40 +00:00
* Invoked as for ex: "-a 000:00:0d.0,
net/bnxt: support representors on remote host domain In the Stingray use case, representors are conventionally run inside the SoC domain representing functions that are on the X86 domain. In order to support this mechanism of building representors for endpoints that are not in the same host domain, additional dev args have been in the PMD like so: rep-based-pf=<physical index> rep-is-pf=<VF=0 or PF=1> where `rep-based-pf` specifies the physical index of the base PF that the representor is derived off of. Since representor(s) can be created for endpoint PFs as well, rename struct bnxt_vf_representor to bnxt_representor and other such dev_ops and function names. devargs have also been extended to specify the exact CoS queue along with flow control enablement to be used for the conduit between the representor and the endpoint function. This is how a sample devargs would look with all the extended devargs -w 0000:06:02.0,host-based-truflow=1,representor=[1],rep-based-pf=8, rep-is-pf=1,rep-q-r2f=1,rep-fc-r2f=0,rep-q-f2r=1,rep-fc-f2r=1 Call CFA_PAIR_ALLOC only in case of Stingray instead of CFA_VFR_ALLOC. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:30 +00:00
* rep-based-pf=<pf index> rep-is-pf=<VF=0 or PF=1>"
*/
net/bnxt: fix resource leak Fix a potential resource leak in case of errors during dev args parsing during device probe. Fixes: 6dc83230b43b ("net/bnxt: support port representor data path") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
2020-10-20 23:24:28 +00:00
ret = rte_kvargs_process(kvlist, BNXT_DEVARG_REP_Q_F2R,
bnxt_parse_devarg_rep_q_f2r,
(void *)&representor);
if (ret) {
ret = -EINVAL;
goto err;
}
net/bnxt: support representors on remote host domain In the Stingray use case, representors are conventionally run inside the SoC domain representing functions that are on the X86 domain. In order to support this mechanism of building representors for endpoints that are not in the same host domain, additional dev args have been in the PMD like so: rep-based-pf=<physical index> rep-is-pf=<VF=0 or PF=1> where `rep-based-pf` specifies the physical index of the base PF that the representor is derived off of. Since representor(s) can be created for endpoint PFs as well, rename struct bnxt_vf_representor to bnxt_representor and other such dev_ops and function names. devargs have also been extended to specify the exact CoS queue along with flow control enablement to be used for the conduit between the representor and the endpoint function. This is how a sample devargs would look with all the extended devargs -w 0000:06:02.0,host-based-truflow=1,representor=[1],rep-based-pf=8, rep-is-pf=1,rep-q-r2f=1,rep-fc-r2f=0,rep-q-f2r=1,rep-fc-f2r=1 Call CFA_PAIR_ALLOC only in case of Stingray instead of CFA_VFR_ALLOC. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:30 +00:00
/*
* Handler for "rep_based_pf" devarg.
eal: replace blacklist/whitelist options Replace -w / --pci-whitelist with -a / --allow options and --pci-blacklist with --block. The -b short option remains unchanged. Allow the old options for now, but print a nag warning since old options are deprecated. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Luca Boccassi <bluca@debian.org> Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
2020-11-10 22:55:40 +00:00
* Invoked as for ex: "-a 000:00:0d.0,
net/bnxt: support representors on remote host domain In the Stingray use case, representors are conventionally run inside the SoC domain representing functions that are on the X86 domain. In order to support this mechanism of building representors for endpoints that are not in the same host domain, additional dev args have been in the PMD like so: rep-based-pf=<physical index> rep-is-pf=<VF=0 or PF=1> where `rep-based-pf` specifies the physical index of the base PF that the representor is derived off of. Since representor(s) can be created for endpoint PFs as well, rename struct bnxt_vf_representor to bnxt_representor and other such dev_ops and function names. devargs have also been extended to specify the exact CoS queue along with flow control enablement to be used for the conduit between the representor and the endpoint function. This is how a sample devargs would look with all the extended devargs -w 0000:06:02.0,host-based-truflow=1,representor=[1],rep-based-pf=8, rep-is-pf=1,rep-q-r2f=1,rep-fc-r2f=0,rep-q-f2r=1,rep-fc-f2r=1 Call CFA_PAIR_ALLOC only in case of Stingray instead of CFA_VFR_ALLOC. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:30 +00:00
* rep-based-pf=<pf index> rep-is-pf=<VF=0 or PF=1>"
*/
net/bnxt: fix resource leak Fix a potential resource leak in case of errors during dev args parsing during device probe. Fixes: 6dc83230b43b ("net/bnxt: support port representor data path") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
2020-10-20 23:24:28 +00:00
ret = rte_kvargs_process(kvlist, BNXT_DEVARG_REP_FC_R2F,
bnxt_parse_devarg_rep_fc_r2f,
(void *)&representor);
if (ret) {
ret = -EINVAL;
goto err;
}
net/bnxt: support representors on remote host domain In the Stingray use case, representors are conventionally run inside the SoC domain representing functions that are on the X86 domain. In order to support this mechanism of building representors for endpoints that are not in the same host domain, additional dev args have been in the PMD like so: rep-based-pf=<physical index> rep-is-pf=<VF=0 or PF=1> where `rep-based-pf` specifies the physical index of the base PF that the representor is derived off of. Since representor(s) can be created for endpoint PFs as well, rename struct bnxt_vf_representor to bnxt_representor and other such dev_ops and function names. devargs have also been extended to specify the exact CoS queue along with flow control enablement to be used for the conduit between the representor and the endpoint function. This is how a sample devargs would look with all the extended devargs -w 0000:06:02.0,host-based-truflow=1,representor=[1],rep-based-pf=8, rep-is-pf=1,rep-q-r2f=1,rep-fc-r2f=0,rep-q-f2r=1,rep-fc-f2r=1 Call CFA_PAIR_ALLOC only in case of Stingray instead of CFA_VFR_ALLOC. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:30 +00:00
/*
* Handler for "rep_based_pf" devarg.
eal: replace blacklist/whitelist options Replace -w / --pci-whitelist with -a / --allow options and --pci-blacklist with --block. The -b short option remains unchanged. Allow the old options for now, but print a nag warning since old options are deprecated. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Luca Boccassi <bluca@debian.org> Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
2020-11-10 22:55:40 +00:00
* Invoked as for ex: "-a 000:00:0d.0,
net/bnxt: support representors on remote host domain In the Stingray use case, representors are conventionally run inside the SoC domain representing functions that are on the X86 domain. In order to support this mechanism of building representors for endpoints that are not in the same host domain, additional dev args have been in the PMD like so: rep-based-pf=<physical index> rep-is-pf=<VF=0 or PF=1> where `rep-based-pf` specifies the physical index of the base PF that the representor is derived off of. Since representor(s) can be created for endpoint PFs as well, rename struct bnxt_vf_representor to bnxt_representor and other such dev_ops and function names. devargs have also been extended to specify the exact CoS queue along with flow control enablement to be used for the conduit between the representor and the endpoint function. This is how a sample devargs would look with all the extended devargs -w 0000:06:02.0,host-based-truflow=1,representor=[1],rep-based-pf=8, rep-is-pf=1,rep-q-r2f=1,rep-fc-r2f=0,rep-q-f2r=1,rep-fc-f2r=1 Call CFA_PAIR_ALLOC only in case of Stingray instead of CFA_VFR_ALLOC. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:30 +00:00
* rep-based-pf=<pf index> rep-is-pf=<VF=0 or PF=1>"
*/
net/bnxt: fix resource leak Fix a potential resource leak in case of errors during dev args parsing during device probe. Fixes: 6dc83230b43b ("net/bnxt: support port representor data path") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
2020-10-20 23:24:28 +00:00
ret = rte_kvargs_process(kvlist, BNXT_DEVARG_REP_FC_F2R,
bnxt_parse_devarg_rep_fc_f2r,
(void *)&representor);
if (ret) {
ret = -EINVAL;
goto err;
}
net/bnxt: support representors on remote host domain In the Stingray use case, representors are conventionally run inside the SoC domain representing functions that are on the X86 domain. In order to support this mechanism of building representors for endpoints that are not in the same host domain, additional dev args have been in the PMD like so: rep-based-pf=<physical index> rep-is-pf=<VF=0 or PF=1> where `rep-based-pf` specifies the physical index of the base PF that the representor is derived off of. Since representor(s) can be created for endpoint PFs as well, rename struct bnxt_vf_representor to bnxt_representor and other such dev_ops and function names. devargs have also been extended to specify the exact CoS queue along with flow control enablement to be used for the conduit between the representor and the endpoint function. This is how a sample devargs would look with all the extended devargs -w 0000:06:02.0,host-based-truflow=1,representor=[1],rep-based-pf=8, rep-is-pf=1,rep-q-r2f=1,rep-fc-r2f=0,rep-q-f2r=1,rep-fc-f2r=1 Call CFA_PAIR_ALLOC only in case of Stingray instead of CFA_VFR_ALLOC. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:30 +00:00
}
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
ret = rte_eth_dev_create(&pci_dev->device, name,
net/bnxt: support representors on remote host domain In the Stingray use case, representors are conventionally run inside the SoC domain representing functions that are on the X86 domain. In order to support this mechanism of building representors for endpoints that are not in the same host domain, additional dev args have been in the PMD like so: rep-based-pf=<physical index> rep-is-pf=<VF=0 or PF=1> where `rep-based-pf` specifies the physical index of the base PF that the representor is derived off of. Since representor(s) can be created for endpoint PFs as well, rename struct bnxt_vf_representor to bnxt_representor and other such dev_ops and function names. devargs have also been extended to specify the exact CoS queue along with flow control enablement to be used for the conduit between the representor and the endpoint function. This is how a sample devargs would look with all the extended devargs -w 0000:06:02.0,host-based-truflow=1,representor=[1],rep-based-pf=8, rep-is-pf=1,rep-q-r2f=1,rep-fc-r2f=0,rep-q-f2r=1,rep-fc-f2r=1 Call CFA_PAIR_ALLOC only in case of Stingray instead of CFA_VFR_ALLOC. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:30 +00:00
sizeof(struct bnxt_representor),
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
NULL, NULL,
net/bnxt: support representors on remote host domain In the Stingray use case, representors are conventionally run inside the SoC domain representing functions that are on the X86 domain. In order to support this mechanism of building representors for endpoints that are not in the same host domain, additional dev args have been in the PMD like so: rep-based-pf=<physical index> rep-is-pf=<VF=0 or PF=1> where `rep-based-pf` specifies the physical index of the base PF that the representor is derived off of. Since representor(s) can be created for endpoint PFs as well, rename struct bnxt_vf_representor to bnxt_representor and other such dev_ops and function names. devargs have also been extended to specify the exact CoS queue along with flow control enablement to be used for the conduit between the representor and the endpoint function. This is how a sample devargs would look with all the extended devargs -w 0000:06:02.0,host-based-truflow=1,representor=[1],rep-based-pf=8, rep-is-pf=1,rep-q-r2f=1,rep-fc-r2f=0,rep-q-f2r=1,rep-fc-f2r=1 Call CFA_PAIR_ALLOC only in case of Stingray instead of CFA_VFR_ALLOC. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:30 +00:00
bnxt_representor_init,
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
&representor);
net/bnxt: fix VFR cleanup during init failure If VF-rep port add fails for some reason, code was rolling back all ports added so far. With some applications, there is no need to do that. Just log failure message for the VF rep port add and continue. Also include RTE_MAX_ETH_PORTS value in the bounds check as one port will be taken by the uplink port anyway Fixes: 6dc83230b43b ("net/bnxt: support port representor data path") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-11 01:55:50 +00:00
if (ret) {
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
PMD_DRV_LOG(ERR, "failed to create bnxt vf "
"representor %s.", name);
net/bnxt: fix VFR cleanup during init failure If VF-rep port add fails for some reason, code was rolling back all ports added so far. With some applications, there is no need to do that. Just log failure message for the VF rep port add and continue. Also include RTE_MAX_ETH_PORTS value in the bounds check as one port will be taken by the uplink port anyway Fixes: 6dc83230b43b ("net/bnxt: support port representor data path") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-11 01:55:50 +00:00
goto err;
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
}
net/bnxt: fix VFR cleanup during init failure If VF-rep port add fails for some reason, code was rolling back all ports added so far. With some applications, there is no need to do that. Just log failure message for the VF rep port add and continue. Also include RTE_MAX_ETH_PORTS value in the bounds check as one port will be taken by the uplink port anyway Fixes: 6dc83230b43b ("net/bnxt: support port representor data path") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-11 01:55:50 +00:00
vf_rep_eth_dev = rte_eth_dev_allocated(name);
if (!vf_rep_eth_dev) {
PMD_DRV_LOG(ERR, "Failed to find the eth_dev"
" for VF-Rep: %s.", name);
ret = -ENODEV;
goto err;
}
PMD_DRV_LOG(DEBUG, "BNXT Port:%d VFR pci probe\n",
net/bnxt: support representors on remote host domain In the Stingray use case, representors are conventionally run inside the SoC domain representing functions that are on the X86 domain. In order to support this mechanism of building representors for endpoints that are not in the same host domain, additional dev args have been in the PMD like so: rep-based-pf=<physical index> rep-is-pf=<VF=0 or PF=1> where `rep-based-pf` specifies the physical index of the base PF that the representor is derived off of. Since representor(s) can be created for endpoint PFs as well, rename struct bnxt_vf_representor to bnxt_representor and other such dev_ops and function names. devargs have also been extended to specify the exact CoS queue along with flow control enablement to be used for the conduit between the representor and the endpoint function. This is how a sample devargs would look with all the extended devargs -w 0000:06:02.0,host-based-truflow=1,representor=[1],rep-based-pf=8, rep-is-pf=1,rep-q-r2f=1,rep-fc-r2f=0,rep-q-f2r=1,rep-fc-f2r=1 Call CFA_PAIR_ALLOC only in case of Stingray instead of CFA_VFR_ALLOC. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:30 +00:00
backing_eth_dev->data->port_id);
net/bnxt: fix VFR cleanup during init failure If VF-rep port add fails for some reason, code was rolling back all ports added so far. With some applications, there is no need to do that. Just log failure message for the VF rep port add and continue. Also include RTE_MAX_ETH_PORTS value in the bounds check as one port will be taken by the uplink port anyway Fixes: 6dc83230b43b ("net/bnxt: support port representor data path") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-11 01:55:50 +00:00
backing_bp->rep_info[representor.vf_id].vfr_eth_dev =
vf_rep_eth_dev;
backing_bp->num_reps++;
net/bnxt: support representors on remote host domain In the Stingray use case, representors are conventionally run inside the SoC domain representing functions that are on the X86 domain. In order to support this mechanism of building representors for endpoints that are not in the same host domain, additional dev args have been in the PMD like so: rep-based-pf=<physical index> rep-is-pf=<VF=0 or PF=1> where `rep-based-pf` specifies the physical index of the base PF that the representor is derived off of. Since representor(s) can be created for endpoint PFs as well, rename struct bnxt_vf_representor to bnxt_representor and other such dev_ops and function names. devargs have also been extended to specify the exact CoS queue along with flow control enablement to be used for the conduit between the representor and the endpoint function. This is how a sample devargs would look with all the extended devargs -w 0000:06:02.0,host-based-truflow=1,representor=[1],rep-based-pf=8, rep-is-pf=1,rep-q-r2f=1,rep-fc-r2f=0,rep-q-f2r=1,rep-fc-f2r=1 Call CFA_PAIR_ALLOC only in case of Stingray instead of CFA_VFR_ALLOC. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:30 +00:00
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
}
net/bnxt: fix resource leak Fix a potential resource leak in case of errors during dev args parsing during device probe. Fixes: 6dc83230b43b ("net/bnxt: support port representor data path") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
2020-10-20 23:24:28 +00:00
rte_kvargs_free(kvlist);
net/bnxt: fix VFR cleanup during init failure If VF-rep port add fails for some reason, code was rolling back all ports added so far. With some applications, there is no need to do that. Just log failure message for the VF rep port add and continue. Also include RTE_MAX_ETH_PORTS value in the bounds check as one port will be taken by the uplink port anyway Fixes: 6dc83230b43b ("net/bnxt: support port representor data path") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-11 01:55:50 +00:00
return 0;
err:
/* If num_rep > 1, then rollback already created
* ports, since we'll be failing the probe anyway
*/
if (num_rep > 1)
bnxt_pci_remove_dev_with_reps(backing_eth_dev);
net/bnxt: fix resource leak Fix a potential resource leak in case of errors during dev args parsing during device probe. Fixes: 6dc83230b43b ("net/bnxt: support port representor data path") Cc: stable@dpdk.org Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
2020-10-20 23:24:28 +00:00
rte_errno = -ret;
rte_kvargs_free(kvlist);
net/bnxt: fix VFR cleanup during init failure If VF-rep port add fails for some reason, code was rolling back all ports added so far. With some applications, there is no need to do that. Just log failure message for the VF rep port add and continue. Also include RTE_MAX_ETH_PORTS value in the bounds check as one port will be taken by the uplink port anyway Fixes: 6dc83230b43b ("net/bnxt: support port representor data path") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-11 01:55:50 +00:00
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
return ret;
drivers/net: do not use ethdev driver Signed-off-by: Jan Blunck <jblunck@infradead.org> Signed-off-by: Gaetan Rivet <gaetan.rivet@6wind.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2017-04-11 15:44:24 +00:00
}
net/bnxt: support port representor data path Added code to support Tx/Rx from a VF representor port. The VF-reps use the RX/TX rings of the Trusted VF/PF. For each VF-rep, the Trusted VF/PF driver issues a VFR_ALLOC FW cmd that returns "cfa_code" and "cfa_action" values. The FW sets up the filter tables in such a way that VF traffic by default (in absence of other rules) gets punted to the parent function i.e. either the Trusted VF or the PF. The cfa_code value in the RX-compl informs the driver of the source VF. For traffic being transmitted from the VF-rep, the TX BD is tagged with a cfa_action value that informs the HW to punt it to the corresponding VF. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:49 +00:00
static int bnxt_pci_probe(struct rte_pci_driver *pci_drv __rte_unused,
struct rte_pci_device *pci_dev)
{
struct rte_eth_devargs eth_da = { .nb_representor_ports = 0 };
struct rte_eth_dev *backing_eth_dev;
uint16_t num_rep;
int ret = 0;
if (pci_dev->device.devargs) {
ret = rte_eth_devargs_parse(pci_dev->device.devargs->args,
&eth_da);
if (ret)
return ret;
}
num_rep = eth_da.nb_representor_ports;
PMD_DRV_LOG(DEBUG, "nb_representor_ports = %d\n",
num_rep);
/* We could come here after first level of probe is already invoked
* as part of an application bringup(OVS-DPDK vswitchd), so first check
* for already allocated eth_dev for the backing device (PF/Trusted VF)
*/
backing_eth_dev = rte_eth_dev_allocated(pci_dev->device.name);
if (backing_eth_dev == NULL) {
ret = rte_eth_dev_create(&pci_dev->device, pci_dev->device.name,
sizeof(struct bnxt),
eth_dev_pci_specific_init, pci_dev,
bnxt_dev_init, NULL);
if (ret || !num_rep)
return ret;
backing_eth_dev = rte_eth_dev_allocated(pci_dev->device.name);
}
net/bnxt: refactor VFR port clean up When parent VF or PF ports are cleaned up, the child VF representor ports also need to be cleaned up. If not cleaned up, then deleting the parent VF shall result in not cleaning up the hardware rules and updating the firmware of VFR removal. The issue can occur even when application is exited without deleting VFR ports. Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Shahaji Bhosle <sbhosle@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-11 01:55:48 +00:00
PMD_DRV_LOG(DEBUG, "BNXT Port:%d pci probe\n",
backing_eth_dev->data->port_id);
net/bnxt: check representors devargs before probe Check for num_rep before invoking rep port probe. num_rep should be !=0 if representor devargs provided. Fixes: 6dc83230b43b ("net/bnxt: support port representor data path") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
2020-09-25 10:40:44 +00:00
if (!num_rep)
return ret;
net/bnxt: support port representor data path Added code to support Tx/Rx from a VF representor port. The VF-reps use the RX/TX rings of the Trusted VF/PF. For each VF-rep, the Trusted VF/PF driver issues a VFR_ALLOC FW cmd that returns "cfa_code" and "cfa_action" values. The FW sets up the filter tables in such a way that VF traffic by default (in absence of other rules) gets punted to the parent function i.e. either the Trusted VF or the PF. The cfa_code value in the RX-compl informs the driver of the source VF. For traffic being transmitted from the VF-rep, the TX BD is tagged with a cfa_action value that informs the HW to punt it to the corresponding VF. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:49 +00:00
/* probe representor ports now */
net/bnxt: fix pass by reference Pass 'eth_da' pointer instead of pass by value to bnxt_rep_port_probe() Coverity issue: 360841 Fixes: 322bd6e70272 ("net/bnxt: add port representor infrastructure") Cc: stable@dpdk.org Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
2020-10-30 07:03:04 +00:00
ret = bnxt_rep_port_probe(pci_dev, &eth_da, backing_eth_dev,
net/bnxt: support representors on remote host domain In the Stingray use case, representors are conventionally run inside the SoC domain representing functions that are on the X86 domain. In order to support this mechanism of building representors for endpoints that are not in the same host domain, additional dev args have been in the PMD like so: rep-based-pf=<physical index> rep-is-pf=<VF=0 or PF=1> where `rep-based-pf` specifies the physical index of the base PF that the representor is derived off of. Since representor(s) can be created for endpoint PFs as well, rename struct bnxt_vf_representor to bnxt_representor and other such dev_ops and function names. devargs have also been extended to specify the exact CoS queue along with flow control enablement to be used for the conduit between the representor and the endpoint function. This is how a sample devargs would look with all the extended devargs -w 0000:06:02.0,host-based-truflow=1,representor=[1],rep-based-pf=8, rep-is-pf=1,rep-q-r2f=1,rep-fc-r2f=0,rep-q-f2r=1,rep-fc-f2r=1 Call CFA_PAIR_ALLOC only in case of Stingray instead of CFA_VFR_ALLOC. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:30 +00:00
pci_dev->device.devargs->args);
net/bnxt: support port representor data path Added code to support Tx/Rx from a VF representor port. The VF-reps use the RX/TX rings of the Trusted VF/PF. For each VF-rep, the Trusted VF/PF driver issues a VFR_ALLOC FW cmd that returns "cfa_code" and "cfa_action" values. The FW sets up the filter tables in such a way that VF traffic by default (in absence of other rules) gets punted to the parent function i.e. either the Trusted VF or the PF. The cfa_code value in the RX-compl informs the driver of the source VF. For traffic being transmitted from the VF-rep, the TX BD is tagged with a cfa_action value that informs the HW to punt it to the corresponding VF. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:49 +00:00
return ret;
}
drivers/net: do not use ethdev driver Signed-off-by: Jan Blunck <jblunck@infradead.org> Signed-off-by: Gaetan Rivet <gaetan.rivet@6wind.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2017-04-11 15:44:24 +00:00
static int bnxt_pci_remove(struct rte_pci_device *pci_dev)
{
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
struct rte_eth_dev *eth_dev;
eth_dev = rte_eth_dev_allocated(pci_dev->device.name);
if (!eth_dev)
return 0; /* Invoked typically only by OVS-DPDK, by the
* time it comes here the eth_dev is already
* deleted by rte_eth_dev_close(), so returning
* +ve value will at least help in proper cleanup
*/
net/bnxt: refactor VFR port clean up When parent VF or PF ports are cleaned up, the child VF representor ports also need to be cleaned up. If not cleaned up, then deleting the parent VF shall result in not cleaning up the hardware rules and updating the firmware of VFR removal. The issue can occur even when application is exited without deleting VFR ports. Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Shahaji Bhosle <sbhosle@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-11 01:55:48 +00:00
PMD_DRV_LOG(DEBUG, "BNXT Port:%d pci remove\n", eth_dev->data->port_id);
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
if (rte_eal_process_type() == RTE_PROC_PRIMARY) {
if (eth_dev->data->dev_flags & RTE_ETH_DEV_REPRESENTOR)
return rte_eth_dev_destroy(eth_dev,
net/bnxt: support representors on remote host domain In the Stingray use case, representors are conventionally run inside the SoC domain representing functions that are on the X86 domain. In order to support this mechanism of building representors for endpoints that are not in the same host domain, additional dev args have been in the PMD like so: rep-based-pf=<physical index> rep-is-pf=<VF=0 or PF=1> where `rep-based-pf` specifies the physical index of the base PF that the representor is derived off of. Since representor(s) can be created for endpoint PFs as well, rename struct bnxt_vf_representor to bnxt_representor and other such dev_ops and function names. devargs have also been extended to specify the exact CoS queue along with flow control enablement to be used for the conduit between the representor and the endpoint function. This is how a sample devargs would look with all the extended devargs -w 0000:06:02.0,host-based-truflow=1,representor=[1],rep-based-pf=8, rep-is-pf=1,rep-q-r2f=1,rep-fc-r2f=0,rep-q-f2r=1,rep-fc-f2r=1 Call CFA_PAIR_ALLOC only in case of Stingray instead of CFA_VFR_ALLOC. Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-09-22 07:06:30 +00:00
bnxt_representor_uninit);
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
else
return rte_eth_dev_destroy(eth_dev,
bnxt_dev_uninit);
} else {
drivers/net: enable device detach on secondary process With the enabling for hotplug on multi-process, rte_eth_dev_pci_generic_remove can be used to detach the device from a secondary process also. But we need to take care of the uninit callback parameter to make sure it handles the secondary case correctly. Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
2018-10-16 00:16:31 +00:00
return rte_eth_dev_pci_generic_remove(pci_dev, NULL);
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
}
drivers/net: do not use ethdev driver Signed-off-by: Jan Blunck <jblunck@infradead.org> Signed-off-by: Gaetan Rivet <gaetan.rivet@6wind.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2017-04-11 15:44:24 +00:00
}
static struct rte_pci_driver bnxt_rte_pmd = {
.id_table = bnxt_pci_id_map,
net/bnxt: add port representor infrastructure Defines data structures and code to init/uninit VF representors during pci_probe and pci_remove respectively. Most of the dev_ops for the VF representor are just stubs for now and will be will be filled out in next patch. To create a representor using testpmd: testpmd -c 0xff -wB:D.F,representor=1 -- -i testpmd -c 0xff -w05:02.0,representor=[1] -- -i To create a representor using ovs-dpdk: 1. First add the trusted VF port to a bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0 2. Add the representor port to the bridge ovs-vsctl add-port ovsbr0 vf_rep1 -- set Interface vf_rep1 type=dpdk options:dpdk-devargs=0000:06:02.0,representor=1 Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com> Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2020-07-02 23:27:48 +00:00
.drv_flags = RTE_PCI_DRV_NEED_MAPPING | RTE_PCI_DRV_INTR_LSC |
RTE_PCI_DRV_PROBE_AGAIN, /* Needed in case of VF-REPs
* and OVS-DPDK
*/
drivers/net: do not use ethdev driver Signed-off-by: Jan Blunck <jblunck@infradead.org> Signed-off-by: Gaetan Rivet <gaetan.rivet@6wind.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2017-04-11 15:44:24 +00:00
.probe = bnxt_pci_probe,
.remove = bnxt_pci_remove,
net/bnxt: add driver for Broadcom NetXtreme-C devices This patch adds the initial skeleton for bnxt driver along with the nic guide, and ties the driver into the build system. At this point, the driver simply fails init. Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com> Reviewed-by: David Christensen <david.christensen@broadcom.com> [Release Note Addition] Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
2016-06-15 21:23:01 +00:00
};
net/bnxt: support Tx loopback, set VF MAC and queues drop Add functions rte_pmd_bnxt_set_tx_loopback, rte_pmd_bnxt_set_all_queues_drop_en and rte_pmd_bnxt_set_vf_mac_addr to configure tx_loopback, queue_drop and VF MAC address setting in the hardware. It also adds the necessary functions to send the HWRM commands to the firmware. Signed-off-by: Steeven Li <steeven.li@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:16 +00:00
static bool
is_device_supported(struct rte_eth_dev *dev, struct rte_pci_driver *drv)
{
ethdev: remove driver name from device private data rte_driver->name has the driver name and all physical and virtual devices has access to it. Previously it was not possible for virtual ethernet devices to access rte_driver->name field (because eth_dev used to keep only pci_dev), and it was required to save driver name in the device private struct. After re-works on bus and vdev, it is possible for all bus types to access rte_driver. It is able to remove the driver name from ethdev device private data and use eth_dev->device->driver->name. Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com> Acked-by: Shreyansh Jain <shreyansh.jain@nxp.com> Acked-by: Jan Blunck <jblunck@infradead.org>
2017-06-12 15:25:12 +00:00
if (strcmp(dev->device->driver->name, drv->driver.name))
net/bnxt: support Tx loopback, set VF MAC and queues drop Add functions rte_pmd_bnxt_set_tx_loopback, rte_pmd_bnxt_set_all_queues_drop_en and rte_pmd_bnxt_set_vf_mac_addr to configure tx_loopback, queue_drop and VF MAC address setting in the hardware. It also adds the necessary functions to send the HWRM commands to the firmware. Signed-off-by: Steeven Li <steeven.li@broadcom.com> Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
2017-06-01 17:07:16 +00:00
return false;
return true;
}
bool is_bnxt_supported(struct rte_eth_dev *dev)
{
return is_device_supported(dev, &bnxt_rte_pmd);
}
log: register with standardized names Let's try to enforce the convention where most drivers use a pmd. logtype with their class reflected in it, and libraries use a lib. logtype. Introduce two new macros: - RTE_LOG_REGISTER_DEFAULT can be used when a single logtype is used in a component. It is associated to the default name provided by the build system, - RTE_LOG_REGISTER_SUFFIX can be used when multiple logtypes are used, and then the passed name is appended to the default name, RTE_LOG_REGISTER is left untouched for existing external users and for components that do not comply with the convention. There is a new Meson variable log_prefix to adapt the default name for baseband (pmd.bb.), bus (no pmd.) and mempool (no pmd.) classes. Note: achieved with below commands + reverted change on net/bonding + edits on crypto/virtio, compress/mlx5, regex/mlx5 $ git grep -l RTE_LOG_REGISTER drivers/ | while read file; do pattern=${file##drivers/}; class=${pattern%%/*}; pattern=${pattern#$class/}; drv=${pattern%%/*}; case "$class" in baseband) pattern=pmd.bb.$drv;; bus) pattern=bus.$drv;; mempool) pattern=mempool.$drv;; *) pattern=pmd.$class.$drv;; esac sed -i -e 's/RTE_LOG_REGISTER(\(.*\), '$pattern',/RTE_LOG_REGISTER_DEFAULT(\1,/' $file; sed -i -e 's/RTE_LOG_REGISTER(\(.*\), '$pattern'\.\(.*\),/RTE_LOG_REGISTER_SUFFIX(\1, \2,/' $file; done $ git grep -l RTE_LOG_REGISTER lib/ | while read file; do pattern=${file##lib/}; pattern=lib.${pattern%%/*}; sed -i -e 's/RTE_LOG_REGISTER(\(.*\), '$pattern',/RTE_LOG_REGISTER_DEFAULT(\1,/' $file; sed -i -e 's/RTE_LOG_REGISTER(\(.*\), '$pattern'\.\(.*\),/RTE_LOG_REGISTER_SUFFIX(\1, \2,/' $file; done Signed-off-by: David Marchand <david.marchand@redhat.com> Signed-off-by: Thomas Monjalon <thomas@monjalon.net> Acked-by: Bruce Richardson <bruce.richardson@intel.com>
2021-04-26 12:51:08 +00:00
RTE_LOG_REGISTER_SUFFIX(bnxt_logtype_driver, driver, NOTICE);
drivers/net: do not use ethdev driver Signed-off-by: Jan Blunck <jblunck@infradead.org> Signed-off-by: Gaetan Rivet <gaetan.rivet@6wind.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org>
2017-04-11 15:44:24 +00:00
RTE_PMD_REGISTER_PCI(net_bnxt, bnxt_rte_pmd);
drivers: rename register macro prefix All macros related to driver registeration renamed from DRIVER_* to RTE_PMD_* This includes: DRIVER_REGISTER_PCI -> RTE_PMD_REGISTER_PCI DRIVER_REGISTER_PCI_TABLE -> RTE_PMD_REGISTER_PCI_TABLE DRIVER_REGISTER_VDEV -> RTE_PMD_REGISTER_VDEV DRIVER_REGISTER_PARAM_STRING -> RTE_PMD_REGISTER_PARAM_STRING DRIVER_EXPORT_* -> RTE_PMD_EXPORT_* Fix PMDINFOGEN tool to look for matches of RTE_PMD_REGISTER_*. Signed-off-by: Shreyansh Jain <shreyansh.jain@nxp.com> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
2016-10-10 05:43:15 +00:00
RTE_PMD_REGISTER_PCI_TABLE(net_bnxt, bnxt_pci_id_map);
drivers/net: fix vfio kmod dependency vfio is the kernel framework used by the vfio-pci kernel driver. DPDK drivers do not rely solely on vfio, but rather on vfio-pci to gain access to pci resources. Fixes: 0880c40113ef ("drivers: advertise kmod dependencies in pmdinfo") Cc: stable@dpdk.org Signed-off-by: David Marchand <david.marchand@6wind.com> Acked-by: Olivier Matz <olivier.matz@6wind.com>
2017-05-20 13:12:37 +00:00
RTE_PMD_REGISTER_KMOD_DEP(net_bnxt, "* igb_uio | uio_pci_generic | vfio-pci");
Reference in New Issue Copy Permalink