* [dpdk-dev] [PATCH 0/6] DPDK PMD for new QLogic FastLinQ QL4xxxx 25G/40G CNAs
@ 2016-02-20 15:40 Harish Patil
2016-02-20 15:40 ` [dpdk-dev] [PATCH 1/6] qede: add maintainers Harish Patil
` (6 more replies)
0 siblings, 7 replies; 25+ messages in thread
From: Harish Patil @ 2016-02-20 15:40 UTC (permalink / raw)
To: dev
This patch set introduces DPDK poll mode driver for new QLogic FastLinQ QL4xxxx
25G/40G capable family of CNAs as well as their SR-IOV Virtual Functions (VF).
The overall PMD driver design includes a common module called ecore that deals
with the low level HW and a upper layer portion that provides the glue logic.
Specifically, the ecore module contains all of the common logic,
e.g. initialization, cleanup, infrastructure for interrupt handling, link
management, slowpath etc. as well as protocol agnostic features and supplying
an abstraction layer for other modules.
The higher layer implements DPDK exported APIs/driver entry points by
interfacing with the common module for configuration/status and also the
fastpath routines.
Included in the patch set is the supporting documentation and maintainers.
Please apply.
Thanks,
Harish Patil (6):
qede: add maintainers
qede: add documentation
qede: add QLogic PCI ids
qede: add driver common module
qede: add driver
qede: enable PMD build
MAINTAINERS | 7 +
config/common_bsdapp | 15 +
config/common_linuxapp | 16 +
doc/guides/nics/index.rst | 1 +
doc/guides/nics/qede.rst | 344 +
drivers/net/Makefile | 1 +
drivers/net/qede/LICENSE.qede_pmd | 28 +
drivers/net/qede/Makefile | 95 +
drivers/net/qede/ecore/bcm_osal.c | 130 +
drivers/net/qede/ecore/bcm_osal.h | 408 +
drivers/net/qede/ecore/common_hsi.h | 714 ++
drivers/net/qede/ecore/ecore.h | 785 ++
drivers/net/qede/ecore/ecore_attn_values.h | 13287 ++++++++++++++++++++++
drivers/net/qede/ecore/ecore_chain.h | 724 ++
drivers/net/qede/ecore/ecore_cxt.c | 2164 ++++
drivers/net/qede/ecore/ecore_cxt.h | 173 +
drivers/net/qede/ecore/ecore_cxt_api.h | 79 +
drivers/net/qede/ecore/ecore_dcbx.c | 950 ++
drivers/net/qede/ecore/ecore_dcbx.h | 55 +
drivers/net/qede/ecore/ecore_dcbx_api.h | 166 +
drivers/net/qede/ecore/ecore_dev.c | 3907 +++++++
drivers/net/qede/ecore/ecore_dev_api.h | 497 +
drivers/net/qede/ecore/ecore_gtt_reg_addr.h | 42 +
drivers/net/qede/ecore/ecore_gtt_values.h | 33 +
drivers/net/qede/ecore/ecore_hsi_common.h | 1912 ++++
drivers/net/qede/ecore/ecore_hsi_eth.h | 1912 ++++
drivers/net/qede/ecore/ecore_hsi_tools.h | 1081 ++
drivers/net/qede/ecore/ecore_hw.c | 1000 ++
drivers/net/qede/ecore/ecore_hw.h | 273 +
drivers/net/qede/ecore/ecore_hw_defs.h | 49 +
drivers/net/qede/ecore/ecore_init_fw_funcs.c | 1275 +++
drivers/net/qede/ecore/ecore_init_fw_funcs.h | 263 +
drivers/net/qede/ecore/ecore_init_ops.c | 610 +
drivers/net/qede/ecore/ecore_init_ops.h | 103 +
drivers/net/qede/ecore/ecore_int.c | 2234 ++++
drivers/net/qede/ecore/ecore_int.h | 234 +
drivers/net/qede/ecore/ecore_int_api.h | 277 +
drivers/net/qede/ecore/ecore_iov_api.h | 931 ++
drivers/net/qede/ecore/ecore_iro.h | 168 +
drivers/net/qede/ecore/ecore_iro_values.h | 59 +
drivers/net/qede/ecore/ecore_l2.c | 1801 +++
drivers/net/qede/ecore/ecore_l2.h | 151 +
drivers/net/qede/ecore/ecore_l2_api.h | 401 +
drivers/net/qede/ecore/ecore_mcp.c | 1952 ++++
drivers/net/qede/ecore/ecore_mcp.h | 304 +
drivers/net/qede/ecore/ecore_mcp_api.h | 629 +
drivers/net/qede/ecore/ecore_proto_if.h | 88 +
drivers/net/qede/ecore/ecore_rt_defs.h | 449 +
drivers/net/qede/ecore/ecore_sp_api.h | 42 +
drivers/net/qede/ecore/ecore_sp_commands.c | 531 +
drivers/net/qede/ecore/ecore_sp_commands.h | 137 +
drivers/net/qede/ecore/ecore_spq.c | 989 ++
drivers/net/qede/ecore/ecore_spq.h | 302 +
drivers/net/qede/ecore/ecore_sriov.c | 3422 ++++++
drivers/net/qede/ecore/ecore_sriov.h | 390 +
drivers/net/qede/ecore/ecore_status.h | 30 +
drivers/net/qede/ecore/ecore_utils.h | 31 +
drivers/net/qede/ecore/ecore_vf.c | 1319 +++
drivers/net/qede/ecore/ecore_vf.h | 415 +
drivers/net/qede/ecore/ecore_vf_api.h | 185 +
drivers/net/qede/ecore/ecore_vfpf_if.h | 588 +
drivers/net/qede/ecore/eth_common.h | 526 +
drivers/net/qede/ecore/mcp_public.h | 1243 ++
drivers/net/qede/ecore/nvm_cfg.h | 935 ++
drivers/net/qede/ecore/reg_addr.h | 1112 ++
drivers/net/qede/qede_eth_if.c | 461 +
drivers/net/qede/qede_eth_if.h | 180 +
drivers/net/qede/qede_ethdev.c | 815 ++
drivers/net/qede/qede_ethdev.h | 138 +
drivers/net/qede/qede_if.h | 167 +
drivers/net/qede/qede_logs.h | 93 +
drivers/net/qede/qede_main.c | 587 +
drivers/net/qede/qede_rxtx.c | 1346 +++
drivers/net/qede/qede_rxtx.h | 183 +
drivers/net/qede/rte_pmd_qede_version.map | 4 +
lib/librte_eal/common/include/rte_pci_dev_ids.h | 44 +
mk/rte.app.mk | 2 +
scripts/test-build.sh | 1 +
78 files changed, 58995 insertions(+)
create mode 100644 doc/guides/nics/qede.rst
create mode 100644 drivers/net/qede/LICENSE.qede_pmd
create mode 100644 drivers/net/qede/Makefile
create mode 100644 drivers/net/qede/ecore/bcm_osal.c
create mode 100644 drivers/net/qede/ecore/bcm_osal.h
create mode 100644 drivers/net/qede/ecore/common_hsi.h
create mode 100644 drivers/net/qede/ecore/ecore.h
create mode 100644 drivers/net/qede/ecore/ecore_attn_values.h
create mode 100644 drivers/net/qede/ecore/ecore_chain.h
create mode 100644 drivers/net/qede/ecore/ecore_cxt.c
create mode 100644 drivers/net/qede/ecore/ecore_cxt.h
create mode 100644 drivers/net/qede/ecore/ecore_cxt_api.h
create mode 100644 drivers/net/qede/ecore/ecore_dcbx.c
create mode 100644 drivers/net/qede/ecore/ecore_dcbx.h
create mode 100644 drivers/net/qede/ecore/ecore_dcbx_api.h
create mode 100644 drivers/net/qede/ecore/ecore_dev.c
create mode 100644 drivers/net/qede/ecore/ecore_dev_api.h
create mode 100644 drivers/net/qede/ecore/ecore_gtt_reg_addr.h
create mode 100644 drivers/net/qede/ecore/ecore_gtt_values.h
create mode 100644 drivers/net/qede/ecore/ecore_hsi_common.h
create mode 100644 drivers/net/qede/ecore/ecore_hsi_eth.h
create mode 100644 drivers/net/qede/ecore/ecore_hsi_tools.h
create mode 100644 drivers/net/qede/ecore/ecore_hw.c
create mode 100644 drivers/net/qede/ecore/ecore_hw.h
create mode 100644 drivers/net/qede/ecore/ecore_hw_defs.h
create mode 100644 drivers/net/qede/ecore/ecore_init_fw_funcs.c
create mode 100644 drivers/net/qede/ecore/ecore_init_fw_funcs.h
create mode 100644 drivers/net/qede/ecore/ecore_init_ops.c
create mode 100644 drivers/net/qede/ecore/ecore_init_ops.h
create mode 100644 drivers/net/qede/ecore/ecore_int.c
create mode 100644 drivers/net/qede/ecore/ecore_int.h
create mode 100644 drivers/net/qede/ecore/ecore_int_api.h
create mode 100644 drivers/net/qede/ecore/ecore_iov_api.h
create mode 100644 drivers/net/qede/ecore/ecore_iro.h
create mode 100644 drivers/net/qede/ecore/ecore_iro_values.h
create mode 100644 drivers/net/qede/ecore/ecore_l2.c
create mode 100644 drivers/net/qede/ecore/ecore_l2.h
create mode 100644 drivers/net/qede/ecore/ecore_l2_api.h
create mode 100644 drivers/net/qede/ecore/ecore_mcp.c
create mode 100644 drivers/net/qede/ecore/ecore_mcp.h
create mode 100644 drivers/net/qede/ecore/ecore_mcp_api.h
create mode 100644 drivers/net/qede/ecore/ecore_proto_if.h
create mode 100644 drivers/net/qede/ecore/ecore_rt_defs.h
create mode 100644 drivers/net/qede/ecore/ecore_sp_api.h
create mode 100644 drivers/net/qede/ecore/ecore_sp_commands.c
create mode 100644 drivers/net/qede/ecore/ecore_sp_commands.h
create mode 100644 drivers/net/qede/ecore/ecore_spq.c
create mode 100644 drivers/net/qede/ecore/ecore_spq.h
create mode 100644 drivers/net/qede/ecore/ecore_sriov.c
create mode 100644 drivers/net/qede/ecore/ecore_sriov.h
create mode 100644 drivers/net/qede/ecore/ecore_status.h
create mode 100644 drivers/net/qede/ecore/ecore_utils.h
create mode 100644 drivers/net/qede/ecore/ecore_vf.c
create mode 100644 drivers/net/qede/ecore/ecore_vf.h
create mode 100644 drivers/net/qede/ecore/ecore_vf_api.h
create mode 100644 drivers/net/qede/ecore/ecore_vfpf_if.h
create mode 100644 drivers/net/qede/ecore/eth_common.h
create mode 100644 drivers/net/qede/ecore/mcp_public.h
create mode 100644 drivers/net/qede/ecore/nvm_cfg.h
create mode 100644 drivers/net/qede/ecore/reg_addr.h
create mode 100644 drivers/net/qede/qede_eth_if.c
create mode 100644 drivers/net/qede/qede_eth_if.h
create mode 100644 drivers/net/qede/qede_ethdev.c
create mode 100644 drivers/net/qede/qede_ethdev.h
create mode 100644 drivers/net/qede/qede_if.h
create mode 100644 drivers/net/qede/qede_logs.h
create mode 100644 drivers/net/qede/qede_main.c
create mode 100644 drivers/net/qede/qede_rxtx.c
create mode 100644 drivers/net/qede/qede_rxtx.h
create mode 100644 drivers/net/qede/rte_pmd_qede_version.map
--
1.8.3.1
^ permalink raw reply [flat|nested] 25+ messages in thread
* [dpdk-dev] [PATCH 1/6] qede: add maintainers
2016-02-20 15:40 [dpdk-dev] [PATCH 0/6] DPDK PMD for new QLogic FastLinQ QL4xxxx 25G/40G CNAs Harish Patil
@ 2016-02-20 15:40 ` Harish Patil
2016-02-20 15:40 ` [dpdk-dev] [PATCH 2/6] qede: add documentation Harish Patil
` (5 subsequent siblings)
6 siblings, 0 replies; 25+ messages in thread
From: Harish Patil @ 2016-02-20 15:40 UTC (permalink / raw)
To: dev; +Cc: Sony Chacko
Signed-off-by: Harish Patil <harish.patil@qlogic.com>
Signed-off-by: Rasesh Mody <rasesh.mody@qlogic.com>
Signed-off-by: Sony Chacko <sony.chacko@qlogic.com>
---
MAINTAINERS | 7 +++++++
1 file changed, 7 insertions(+)
diff --git a/MAINTAINERS b/MAINTAINERS
index 628bc05..1b27467 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -352,6 +352,13 @@ Null PMD
M: Tetsuya Mukawa <mukawa@igel.co.jp>
F: drivers/net/null/
+QLogic qede PMD
+M: Harish Patil <harish.patil@qlogic.com>
+M: Rasesh Mody <rasesh.mody@qlogic.com>
+M: Sony Chacko <sony.chacko@qlogic.com>
+F: drivers/net/qede/
+F: doc/guides/nics/qede.rst
+
Intel AES-NI Multi-Buffer
M: Declan Doherty <declan.doherty@intel.com>
F: drivers/crypto/aesni_mb/
--
1.8.3.1
^ permalink raw reply [flat|nested] 25+ messages in thread
* [dpdk-dev] [PATCH 2/6] qede: add documentation
2016-02-20 15:40 [dpdk-dev] [PATCH 0/6] DPDK PMD for new QLogic FastLinQ QL4xxxx 25G/40G CNAs Harish Patil
2016-02-20 15:40 ` [dpdk-dev] [PATCH 1/6] qede: add maintainers Harish Patil
@ 2016-02-20 15:40 ` Harish Patil
2016-02-22 16:52 ` Mcnamara, John
2016-02-22 17:38 ` Mcnamara, John
2016-02-20 15:40 ` [dpdk-dev] [PATCH 3/6] qede: add QLogic PCI ids Harish Patil
` (4 subsequent siblings)
6 siblings, 2 replies; 25+ messages in thread
From: Harish Patil @ 2016-02-20 15:40 UTC (permalink / raw)
To: dev; +Cc: Sony Chacko
Signed-off-by: Harish Patil <harish.patil@qlogic.com>
Signed-off-by: Rasesh Mody <rasesh.mody@qlogic.com>
Signed-off-by: Sony Chacko <sony.chacko@qlogic.com>
---
doc/guides/nics/index.rst | 1 +
doc/guides/nics/qede.rst | 344 ++++++++++++++++++++++++++++++++++++++++++++++
2 files changed, 345 insertions(+)
create mode 100644 doc/guides/nics/qede.rst
diff --git a/doc/guides/nics/index.rst b/doc/guides/nics/index.rst
index 8618114..f092f75 100644
--- a/doc/guides/nics/index.rst
+++ b/doc/guides/nics/index.rst
@@ -50,6 +50,7 @@ Network Interface Controller Drivers
virtio
vmxnet3
pcap_ring
+ qede
**Figures**
diff --git a/doc/guides/nics/qede.rst b/doc/guides/nics/qede.rst
new file mode 100644
index 0000000..de4fecd
--- /dev/null
+++ b/doc/guides/nics/qede.rst
@@ -0,0 +1,344 @@
+.. BSD LICENSE
+ Copyright (c) 2016 QLogic Corporation
+ All rights reserved.
+
+ Redistribution and use in source and binary forms, with or without
+ modification, are permitted provided that the following conditions
+ are met:
+
+ * Redistributions of source code must retain the above copyright
+ notice, this list of conditions and the following disclaimer.
+ * Redistributions in binary form must reproduce the above copyright
+ notice, this list of conditions and the following disclaimer in
+ the documentation and/or other materials provided with the
+ distribution.
+ * Neither the name of QLogic Corporation nor the names of its
+ contributors may be used to endorse or promote products derived
+ from this software without specific prior written permission.
+
+ THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+
+QEDE Poll Mode Driver
+======================
+
+The QEDE poll mode driver library (**librte_pmd_qede**) implements support
+for **QLogic FastLinQ QL4xxxx 25G/40G CNA** family of adapters as well
+as their virtual functions (VF) in SR-IOV context. It is supported on
+several standard Linux distros like RHEL7.x and SLES12.x OS.
+It is compile-tested under FreeBSD OS.
+
+More information can be found at `QLogic Corporation's Official Website
+<http://www.qlogic.com>`_.
+
+Supported Features
+------------------
+
+- Unicast/multicast/VLAN filtering
+- Promiscuous mode
+- Allmulti mode
+- Port hardware statistics
+- Jumbo frames (using single buffer)
+- VLAN tag insertion/removal
+- Stateless checksum offloads (IPv4/TCP/UDP)
+- Multiple Rx/Tx queues (queue-pairs)
+- RSS (with default table/key)
+- TSS
+- Multiple MAC address
+- Default pause flow control
+- SR-IOV VF
+
+Non-supported Features
+----------------------
+
+- Scatter-Gather Rx/Tx frames
+- User configurable RETA table/key
+- Unequal number of Rx/Tx queues
+- MTU change (dynamic)
+- SR-IOV PF
+- Tunneling offloads
+
+Supported QLogic NICs
+---------------------
+
+- QLogic FastLinQ QL4xxxx 25G/40G CNAs
+
+Prerequisites
+-------------
+
+- Requires storm firmware version **8.7.x.** and management
+ firmware version **8.7.x and higher**. Storm firmware may be available
+ under /usr/lib/firmware/qed/ in certain newer Linux
+ distributions (e.g. qed_init_values_zipped-8.7.8.0.bin).
+
+- If the required firmware files are not available then visit
+ `QLogic Driver Download Center <http://driverdownloads.qlogic.com>`
+
+- This driver relies on external zlib library (-lz) for uncompressing
+ the firmware file.
+
+Config File Options
+~~~~~~~~~~~~~~~~~~~
+
+The following options can be modified in the ``.config`` file. Please note that
+enabling debugging options may affect system performance.
+
+- ``CONFIG_RTE_LIBRTE_QEDE_PMD`` (default **y**)
+
+ Toggle compilation of QEDE PMD driver.
+
+- ``CONFIG_RTE_LIBRTE_QEDE_DEBUG_INFO`` (default **n**)
+
+ Toggle display of generic debugging messages.
+
+- ``CONFIG_RTE_LIBRTE_QEDE_DEBUG_ECORE`` (default **n**)
+
+ Toggle display of ecore related messages.
+
+- ``CONFIG_RTE_LIBRTE_QEDE_DEBUG_TX`` (default **n**)
+
+ Toggle display of transmit fast path run-time messages.
+
+- ``CONFIG_RTE_LIBRTE_QEDE_DEBUG_RX`` (default **n**)
+
+ Toggle display of receive fast path run-time messages.
+
+- ``CONFIG_RTE_LIBRTE_QEDE_RX_COAL_US`` (default **24**)
+
+ Change Rx interrupt coalescing timer (in us).
+
+- ``CONFIG_RTE_LIBRTE_QEDE_TX_COAL_US`` (default **48**)
+
+ Change Tx interrupt coalescing timer (in us).
+
+- ``CONFIG_RTE_LIBRTE_QEDE_TX_SWITCHING`` (default **y**)
+
+ Toggle Tx switching
+
+- ``CONFIG_RTE_LIBRTE_QEDE_FW`` (default **n**)
+
+ Path of firmware file (overrides default location)
+
+Driver Compilation
+~~~~~~~~~~~~~~~~~~
+
+To compile QEDE PMD for Linux x86_64 gcc target, run the following "make"
+command::
+
+ cd <DPDK-source-directory>
+ make config T=x86_64-native-linuxapp-gcc install
+
+To compile QEDE PMD for Linux x86_64 clang target, run the following "make"
+command::
+
+ cd <DPDK-source-directory>
+ make config T=x86_64-native-linuxapp-clang install
+
+To compile QEDE PMD for FreeBSD x86_64 clang target, run the following "gmake"
+command::
+
+ cd <DPDK-source-directory>
+ gmake config T=x86_64-native-bsdapp-clang install
+
+To compile QEDE PMD for FreeBSD x86_64 gcc target, run the following "gmake"
+command::
+
+ cd <DPDK-source-directory>
+ gmake config T=x86_64-native-bsdapp-gcc install -Wl,-rpath=/usr/local/lib/gcc48 CC=gcc48
+
+
+Sample Application Notes
+~~~~~~~~~~~~~~~~~~~~~~~~
+
+This section demonstrates how to launch ``testpmd`` with QLogic 579xx
+devices managed by ``librte_pmd_qede`` in Linux operating system.
+
+#. Request huge pages:
+
+ .. code-block:: console
+
+ echo 1024 > /sys/kernel/mm/hugepages/hugepages-2048kB/nr_hugepages/nr_hugepages
+
+#. Load ``igb_uio`` or ``vfio-pci`` driver:
+
+ .. code-block:: console
+
+ insmod ./x86_64-native-linuxapp-gcc/kmod/igb_uio.ko
+
+ or
+
+ .. code-block:: console
+
+ modprobe vfio-pci
+
+#. Bind the QLogic 579xx adapters to ``igb_uio`` or ``vfio-pci`` loaded in the
+ previous step::
+
+ ./tools/dpdk_nic_bind.py --bind igb_uio 0000:84:00.0 0000:84:00.1 0000:84:00.2 0000:84:00.3
+
+ or
+
+ Setup VFIO permissions for regular users and then bind to ``vfio-pci``:
+
+ .. code-block:: console
+
+ sudo chmod a+x /dev/vfio
+
+ sudo chmod 0666 /dev/vfio/*
+
+ ./tools/dpdk_nic_bind.py --bind vfio-pci 0000:84:00.0 0000:84:00.1 0000:84:00.2 0000:84:00.3
+
+#. Start ``testpmd`` with basic parameters:
+
+ .. code-block:: console
+
+ testpmd -c 0xf8000 -n 4 -- -i --nb-cores=4 --portmask=0xf --rxd=4096 --txd=4096 --txfreet=4068 --rxq=2 --txq=2 --rss-ip --rss-udp
+
+ [...]
+
+ EAL: PCI device 0000:84:00.0 on NUMA socket 1
+ EAL: probe driver: 1077:1634 rte_qede_pmd
+ EAL: Not managed by a supported kernel driver, skipped
+ EAL: PCI device 0000:84:00.1 on NUMA socket 1
+ EAL: probe driver: 1077:1634 rte_qede_pmd
+ EAL: Not managed by a supported kernel driver, skipped
+ EAL: PCI device 0000:88:00.0 on NUMA socket 1
+ EAL: probe driver: 1077:1656 rte_qede_pmd
+ EAL: PCI memory mapped at 0x7f738b200000
+ EAL: PCI memory mapped at 0x7f738b280000
+ EAL: PCI memory mapped at 0x7f738b300000
+ [QEDE PMD: (88:00.0:dpdk-port-0)]qed_load_firmware_data: Loading the firmware file /lib/firmware/qed/qed_init_values_zipped.bin...
+ [QEDE PMD: (88:00.0:dpdk-port-0)]qede_print_adapter_info:Chip details - BB1
+ [QEDE PMD: (88:00.0:dpdk-port-0)]qede_print_adapter_info:Driver version:QEDE PMD 8.7.9.0_1.0.0
+ [QEDE PMD: (88:00.0:dpdk-port-0)]qede_print_adapter_info:Firmware version:8.7.7.0
+ [QEDE PMD: (88:00.0:dpdk-port-0)]qede_print_adapter_info:Management firmware version:8.7.8.0
+ [QEDE PMD: (88:00.0:dpdk-port-0)]qede_common_dev_init:macaddr 00:0e:1e:d2:08:e8
+ EAL: PCI device 0000:88:00.1 on NUMA socket 1
+ EAL: probe driver: 1077:1656 rte_qede_pmd
+ EAL: PCI memory mapped at 0x7f738b310000
+ EAL: PCI memory mapped at 0x7f738b390000
+ EAL: PCI memory mapped at 0x7f738b410000
+ [QEDE PMD: (88:00.1:dpdk-port-1)]qede_common_dev_init:macaddr 00:0e:1e:d2:08:e9
+ EAL: PCI device 0000:88:00.2 on NUMA socket 1
+ EAL: probe driver: 1077:1656 rte_qede_pmd
+ EAL: PCI memory mapped at 0x7f738b420000
+ EAL: PCI memory mapped at 0x7f738b4a0000
+ EAL: PCI memory mapped at 0x7f738b520000
+
+ [QEDE PMD: (88:00.2:dpdk-port-2)]qede_common_dev_init:macaddr 00:0e:1e:d2:08:ea
+ EAL: PCI device 0000:88:00.3 on NUMA socket 1
+ EAL: probe driver: 1077:1656 rte_qede_pmd
+ EAL: PCI memory mapped at 0x7f738b530000
+ EAL: PCI memory mapped at 0x7f738b5b0000
+ EAL: PCI memory mapped at 0x7f738b630000
+
+ [QEDE PMD: (88:00.3:dpdk-port-3)]qede_common_dev_init:macaddr 00:0e:1e:d2:08:eb
+ Interactive-mode selected
+ Configuring Port 0 (socket 0)
+ [QEDE PMD: (88:00.0:dpdk-port-0)]qede_tx_queue_setup:txq 0 num_desc 4096 tx_free_thresh 4068 socket 0
+ [QEDE PMD: (88:00.0:dpdk-port-0)]qede_tx_queue_setup:txq 1 num_desc 4096 tx_free_thresh 4068 socket 0
+ [QEDE PMD: (88:00.0:dpdk-port-0)]qede_rx_queue_setup:rxq 0 num_desc 4096 rx_buf_size=2148 socket 0
+ [QEDE PMD: (88:00.0:dpdk-port-0)]qede_rx_queue_setup:rxq 1 num_desc 4096 rx_buf_size=2148 socket 0
+ [QEDE PMD: (88:00.0:dpdk-port-0)]qede_dev_start:port 0
+ [QEDE PMD: (88:00.0:dpdk-port-0)]qede_config_rss:RSS flag is set
+ [QEDE PMD: (88:00.0:dpdk-port-0)]qede_dev_start:link status: down
+ Port 0: 00:0E:1E:D2:08:E8
+ Configuring Port 1 (socket 0)
+ [QEDE PMD: (88:00.1:dpdk-port-1)]qede_tx_queue_setup:txq 0 num_desc 4096 tx_free_thresh 4068 socket 0
+ [QEDE PMD: (88:00.1:dpdk-port-1)]qede_tx_queue_setup:txq 1 num_desc 4096 tx_free_thresh 4068 socket 0
+ [QEDE PMD: (88:00.1:dpdk-port-1)]qede_rx_queue_setup:rxq 0 num_desc 4096 rx_buf_size=2148 socket 0
+ [QEDE PMD: (88:00.1:dpdk-port-1)]qede_rx_queue_setup:rxq 1 num_desc 4096 rx_buf_size=2148 socket 0
+ [QEDE PMD: (88:00.1:dpdk-port-1)]qede_dev_start:port 1
+ ...
+ Checking link statuses...
+ Port 0 Link Up - speed 25000 Mbps - full-duplex
+ Port 1 Link Up - speed 25000 Mbps - full-duplex
+ Port 2 Link Up - speed 25000 Mbps - full-duplex
+ Port 3 Link Up - speed 25000 Mbps - full-duplex
+ Done
+ testpmd>
+
+
+SR-IOV: Prerequisites and Sample Application Notes
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+This section provides instructions to configure SR-IOV with Linux OS.
+
+**Note**: QEDE PMD will be used to bind to SR-IOV VF device and
+ Linux native kernel driver (QEDE) will function as SR-IOV PF driver.
+
+#. Verify SR-IOV and ARI capabilities are enabled on the adapter using ``lspci``:
+
+ .. code-block:: console
+
+ lspci -s <slot> -vvv
+
+ Example output:
+
+ .. code-block:: console
+
+ [...]
+ Capabilities: [1b8 v1] Alternative Routing-ID Interpretation (ARI)
+ [...]
+ Capabilities: [1c0 v1] Single Root I/O Virtualization (SR-IOV)
+ [...]
+ Kernel driver in use: igb_uio
+
+#. Load the kernel module:
+
+ .. code-block:: console
+
+ modprobe qede
+
+ Example output:
+
+ .. code-block:: console
+
+ systemd-udevd[4848]: renamed network interface eth0 to ens5f0
+ systemd-udevd[4848]: renamed network interface eth1 to ens5f1
+
+#. Bring up the PF ports:
+
+ .. code-block:: console
+
+ ifconfig ens5f0 up
+ ifconfig ens5f1 up
+
+#. Create VF device(s):
+
+ Echo the number of VFs to be created into "sriov_numvfs" sysfs entry
+ of the parent PF.
+
+ Example output:
+
+ .. code-block:: console
+
+ echo 2 > /sys/devices/pci0000:00/0000:00:03.0/0000:81:00.0/sriov_numvfs
+
+
+#. Assign VF MAC address:
+
+ Assign MAC address to the VF using iproute2 utility. The syntax is:
+ ip link set <PF iface> vf <VF id> mac <macaddr>
+
+ Example output:
+
+ .. code-block:: console
+
+ ip link set ens5f0 vf 0 mac 52:54:00:2f:9d:e8
+
+
+#. PCI Passthrough:
+
+ The VF devices may be passed through to the guest VM using virt-manager or
+ virsh. QEDE PMD should be used to bind the VF devices in the guest VM
+ using the instructions outlined in the Application notes below.
--
1.8.3.1
^ permalink raw reply [flat|nested] 25+ messages in thread
* [dpdk-dev] [PATCH 3/6] qede: add QLogic PCI ids
2016-02-20 15:40 [dpdk-dev] [PATCH 0/6] DPDK PMD for new QLogic FastLinQ QL4xxxx 25G/40G CNAs Harish Patil
2016-02-20 15:40 ` [dpdk-dev] [PATCH 1/6] qede: add maintainers Harish Patil
2016-02-20 15:40 ` [dpdk-dev] [PATCH 2/6] qede: add documentation Harish Patil
@ 2016-02-20 15:40 ` Harish Patil
2016-02-21 1:17 ` Stephen Hemminger
2016-02-20 15:40 ` [dpdk-dev] [PATCH 5/6] qede: add driver Harish Patil
` (3 subsequent siblings)
6 siblings, 1 reply; 25+ messages in thread
From: Harish Patil @ 2016-02-20 15:40 UTC (permalink / raw)
To: dev; +Cc: Sony Chacko
Signed-off-by: Harish Patil <harish.patil@qlogic.com>
Signed-off-by: Rasesh Mody <rasesh.mody@qlogic.com>
Signed-off-by: Sony Chacko <sony.chacko@qlogic.com>
---
lib/librte_eal/common/include/rte_pci_dev_ids.h | 44 +++++++++++++++++++++++++
1 file changed, 44 insertions(+)
diff --git a/lib/librte_eal/common/include/rte_pci_dev_ids.h b/lib/librte_eal/common/include/rte_pci_dev_ids.h
index d088191..0c1a3fe 100644
--- a/lib/librte_eal/common/include/rte_pci_dev_ids.h
+++ b/lib/librte_eal/common/include/rte_pci_dev_ids.h
@@ -152,6 +152,14 @@
#define RTE_PCI_DEV_ID_DECL_BNX2XVF(vend, dev)
#endif
+#ifndef RTE_PCI_DEV_ID_DECL_QEDE
+#define RTE_PCI_DEV_ID_DECL_QEDE(vend, dev)
+#endif
+
+#ifndef RTE_PCI_DEV_ID_DECL_QEDEVF
+#define RTE_PCI_DEV_ID_DECL_QEDEVF(vend, dev)
+#endif
+
#ifndef PCI_VENDOR_ID_INTEL
/** Vendor ID used by Intel devices */
#define PCI_VENDOR_ID_INTEL 0x8086
@@ -177,6 +185,11 @@
#define PCI_VENDOR_ID_BROADCOM 0x14E4
#endif
+#ifndef PCI_VENDOR_ID_QLOGIC
+/** Vendor ID used by QLogic devices */
+#define PCI_VENDOR_ID_QLOGIC 0x1077
+#endif
+
/******************** Physical EM devices from e1000_hw.h ********************/
#define E1000_DEV_ID_82542 0x1000
@@ -651,6 +664,35 @@ RTE_PCI_DEV_ID_DECL_BNX2X(PCI_VENDOR_ID_BROADCOM, BNX2X_DEV_ID_57811_MF)
RTE_PCI_DEV_ID_DECL_BNX2X(PCI_VENDOR_ID_BROADCOM, BNX2X_DEV_ID_57840_MF)
#endif
+/****************** QLogic 25G/40G devices ******************/
+#define CHIP_NUM_57980E 0x1634
+#define CHIP_NUM_57980S 0x1629
+#define CHIP_NUM_VF 0x1630
+#define CHIP_NUM_57980S_40 0x1634
+#define CHIP_NUM_57980S_VF 0x1637 /* will change to 1664 */
+#define CHIP_NUM_57980S_25 0x1656
+#define CHIP_NUM_57980S_IOV 0x1664
+#define CHIP_NUM_AH 0x8070
+
+#ifndef PCI_DEVICE_ID_NX2_57980E
+#define PCI_DEVICE_ID_NX2_57980E CHIP_NUM_57980E
+#define PCI_DEVICE_ID_NX2_57980S CHIP_NUM_57980S
+#define PCI_DEVICE_ID_NX2_VF CHIP_NUM_VF
+#define PCI_DEVICE_ID_57980S_40 CHIP_NUM_57980S_40
+#define PCI_DEVICE_ID_57980S_VF CHIP_NUM_57980S_VF
+#define PCI_DEVICE_ID_57980S_25 CHIP_NUM_57980S_25
+#define PCI_DEVICE_ID_57980S_IOV CHIP_NUM_57980S_IOV
+#define PCI_DEVICE_ID_AH CHIP_NUM_AH
+#endif
+
+RTE_PCI_DEV_ID_DECL_QEDE(PCI_VENDOR_ID_QLOGIC, PCI_DEVICE_ID_NX2_57980E)
+RTE_PCI_DEV_ID_DECL_QEDE(PCI_VENDOR_ID_QLOGIC, PCI_DEVICE_ID_NX2_57980S)
+RTE_PCI_DEV_ID_DECL_QEDE(PCI_VENDOR_ID_QLOGIC, PCI_DEVICE_ID_57980S_40)
+RTE_PCI_DEV_ID_DECL_QEDE(PCI_VENDOR_ID_QLOGIC, PCI_DEVICE_ID_57980S_25)
+RTE_PCI_DEV_ID_DECL_QEDEVF(PCI_VENDOR_ID_QLOGIC, PCI_DEVICE_ID_NX2_VF)
+RTE_PCI_DEV_ID_DECL_QEDEVF(PCI_VENDOR_ID_QLOGIC, PCI_DEVICE_ID_57980S_VF)
+RTE_PCI_DEV_ID_DECL_QEDEVF(PCI_VENDOR_ID_QLOGIC, PCI_DEVICE_ID_57980S_IOV)
+
/*
* Undef all RTE_PCI_DEV_ID_DECL_* here.
*/
@@ -667,3 +709,5 @@ RTE_PCI_DEV_ID_DECL_BNX2X(PCI_VENDOR_ID_BROADCOM, BNX2X_DEV_ID_57840_MF)
#undef RTE_PCI_DEV_ID_DECL_VMXNET3
#undef RTE_PCI_DEV_ID_DECL_FM10K
#undef RTE_PCI_DEV_ID_DECL_FM10KVF
+#undef RTE_PCI_DEV_ID_DECL_QEDE
+#undef RTE_PCI_DEV_ID_DECL_QEDEVF
--
1.8.3.1
^ permalink raw reply [flat|nested] 25+ messages in thread
* [dpdk-dev] [PATCH 5/6] qede: add driver
2016-02-20 15:40 [dpdk-dev] [PATCH 0/6] DPDK PMD for new QLogic FastLinQ QL4xxxx 25G/40G CNAs Harish Patil
` (2 preceding siblings ...)
2016-02-20 15:40 ` [dpdk-dev] [PATCH 3/6] qede: add QLogic PCI ids Harish Patil
@ 2016-02-20 15:40 ` Harish Patil
2016-02-21 1:26 ` Stephen Hemminger
2016-02-20 15:40 ` [dpdk-dev] [PATCH 6/6] qede: enable PMD build Harish Patil
` (2 subsequent siblings)
6 siblings, 1 reply; 25+ messages in thread
From: Harish Patil @ 2016-02-20 15:40 UTC (permalink / raw)
To: dev; +Cc: Sony Chacko
Signed-off-by: Harish Patil <harish.patil@qlogic.com>
Signed-off-by: Rasesh Mody <rasesh.mody@qlogic.com>
Signed-off-by: Sony Chacko <sony.chacko@qlogic.com>
---
drivers/net/qede/LICENSE.qede_pmd | 28 +
drivers/net/qede/Makefile | 95 ++
drivers/net/qede/qede_eth_if.c | 461 ++++++++++
drivers/net/qede/qede_eth_if.h | 180 ++++
drivers/net/qede/qede_ethdev.c | 815 +++++++++++++++++
drivers/net/qede/qede_ethdev.h | 138 +++
drivers/net/qede/qede_if.h | 167 ++++
drivers/net/qede/qede_logs.h | 93 ++
drivers/net/qede/qede_main.c | 587 +++++++++++++
drivers/net/qede/qede_rxtx.c | 1346 +++++++++++++++++++++++++++++
drivers/net/qede/qede_rxtx.h | 183 ++++
drivers/net/qede/rte_pmd_qede_version.map | 4 +
12 files changed, 4097 insertions(+)
create mode 100644 drivers/net/qede/LICENSE.qede_pmd
create mode 100644 drivers/net/qede/Makefile
create mode 100644 drivers/net/qede/qede_eth_if.c
create mode 100644 drivers/net/qede/qede_eth_if.h
create mode 100644 drivers/net/qede/qede_ethdev.c
create mode 100644 drivers/net/qede/qede_ethdev.h
create mode 100644 drivers/net/qede/qede_if.h
create mode 100644 drivers/net/qede/qede_logs.h
create mode 100644 drivers/net/qede/qede_main.c
create mode 100644 drivers/net/qede/qede_rxtx.c
create mode 100644 drivers/net/qede/qede_rxtx.h
create mode 100644 drivers/net/qede/rte_pmd_qede_version.map
diff --git a/drivers/net/qede/LICENSE.qede_pmd b/drivers/net/qede/LICENSE.qede_pmd
new file mode 100644
index 0000000..c7cbdcc
--- /dev/null
+++ b/drivers/net/qede/LICENSE.qede_pmd
@@ -0,0 +1,28 @@
+/*
+ * BSD LICENSE
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright
+ * notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ * notice, this list of conditions and the following disclaimer in the
+ * documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of QLogic Corporation nor the name of its contributors
+ * may be used to endorse or promote products derived from this software
+ * without specific prior written consent.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS'
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS
+ * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
+ * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
+ * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
+ * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
+ * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF
+ * THE POSSIBILITY OF SUCH DAMAGE.
+ */
diff --git a/drivers/net/qede/Makefile b/drivers/net/qede/Makefile
new file mode 100644
index 0000000..a62d80e
--- /dev/null
+++ b/drivers/net/qede/Makefile
@@ -0,0 +1,95 @@
+# Copyright (c) 2016 QLogic Corporation.
+# All rights reserved.
+# www.qlogic.com
+#
+# See LICENSE.qede_pmd for copyright and licensing details.
+
+include $(RTE_SDK)/mk/rte.vars.mk
+
+#
+# library name
+#
+LIB = librte_pmd_qede.a
+
+CFLAGS += -O3
+CFLAGS += $(WERROR_FLAGS)
+
+EXPORT_MAP := rte_pmd_qede_version.map
+
+LIBABIVER := 1
+
+#
+#OS
+#
+OS_TYPE := $(shell uname -s)
+
+#
+# CFLAGS
+#
+CFLAGS_ECORE_DRIVER = -Wno-unused-parameter
+CFLAGS_ECORE_DRIVER += -Wno-unused-value
+CFLAGS_ECORE_DRIVER += -Wno-sign-compare
+CFLAGS_ECORE_DRIVER += -Wno-missing-prototypes
+CFLAGS_ECORE_DRIVER += -Wno-cast-qual
+CFLAGS_ECORE_DRIVER += -Wno-unused-function
+CFLAGS_ECORE_DRIVER += -Wno-unused-variable
+CFLAGS_ECORE_DRIVER += -Wno-strict-aliasing
+CFLAGS_ECORE_DRIVER += -Wno-missing-prototypes
+CFLAGS_ECORE_DRIVER += -Wno-format-nonliteral
+ifeq ($(OS_TYPE),Linux)
+CFLAGS_ECORE_DRIVER += -Wno-shift-negative-value
+endif
+
+ifneq (,$(filter gcc gcc48,$(CC)))
+CFLAGS_ECORE_DRIVER += -Wno-unused-but-set-variable
+CFLAGS_ECORE_DRIVER += -Wno-missing-declarations
+CFLAGS_ECORE_DRIVER += -Wno-maybe-uninitialized
+CFLAGS_ECORE_DRIVER += -Wno-strict-prototypes
+else ifeq ($(CC), clang)
+CFLAGS_ECORE_DRIVER += -Wno-format-extra-args
+CFLAGS_ECORE_DRIVER += -Wno-visibility
+CFLAGS_ECORE_DRIVER += -Wno-empty-body
+CFLAGS_ECORE_DRIVER += -Wno-invalid-source-encoding
+CFLAGS_ECORE_DRIVER += -Wno-sometimes-uninitialized
+CFLAGS_ECORE_DRIVER += -Wno-pointer-bool-conversion
+else
+#icc flags
+endif
+
+#
+# Add extra flags for base ecore driver files
+# to disable warnings in them
+#
+#
+ECORE_DRIVER_OBJS=$(patsubst %.c,%.o,$(notdir $(wildcard $(SRCDIR)/ecore/*.c)))
+$(foreach obj, $(ECORE_DRIVER_OBJS), $(eval CFLAGS+=$(CFLAGS_ECORE_DRIVER)))
+
+#
+# all source are stored in SRCS-y
+#
+SRCS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += ecore/ecore_dev.c
+SRCS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += ecore/ecore_hw.c
+SRCS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += ecore/ecore_cxt.c
+SRCS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += ecore/ecore_l2.c
+SRCS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += ecore/ecore_sp_commands.c
+SRCS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += ecore/ecore_init_fw_funcs.c
+SRCS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += ecore/ecore_spq.c
+SRCS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += ecore/ecore_init_ops.c
+SRCS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += ecore/ecore_mcp.c
+SRCS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += ecore/ecore_int.c
+SRCS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += ecore/ecore_dcbx.c
+SRCS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += ecore/bcm_osal.c
+SRCS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += ecore/ecore_sriov.c
+SRCS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += ecore/ecore_vf.c
+SRCS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += qede_ethdev.c
+SRCS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += qede_eth_if.c
+SRCS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += qede_main.c
+SRCS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += qede_rxtx.c
+
+# dependent libs:
+DEPDIRS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += lib/librte_eal lib/librte_ether
+DEPDIRS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += lib/librte_mempool lib/librte_mbuf
+DEPDIRS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += lib/librte_net lib/librte_malloc
+
+
+include $(RTE_SDK)/mk/rte.lib.mk
diff --git a/drivers/net/qede/qede_eth_if.c b/drivers/net/qede/qede_eth_if.c
new file mode 100644
index 0000000..7c48c8b
--- /dev/null
+++ b/drivers/net/qede/qede_eth_if.c
@@ -0,0 +1,461 @@
+/*
+ * Copyright (c) 2016 QLogic Corporation.
+ * All rights reserved.
+ * www.qlogic.com
+ *
+ * See LICENSE.qede_pmd for copyright and licensing details.
+ */
+
+#include "qede_ethdev.h"
+
+static int
+qed_start_vport(struct ecore_dev *edev, struct qed_start_vport_params *p_params)
+{
+ int rc, i;
+
+ for_each_hwfn(edev, i) {
+ struct ecore_hwfn *p_hwfn = &edev->hwfns[i];
+ u8 tx_switching = 0;
+ struct ecore_sp_vport_start_params start = { 0 };
+
+ start.tpa_mode = p_params->gro_enable ? ECORE_TPA_MODE_GRO :
+ ECORE_TPA_MODE_NONE;
+ start.remove_inner_vlan = p_params->remove_inner_vlan;
+ start.tx_switching = tx_switching;
+ start.only_untagged = false; /* untagged only */
+ start.drop_ttl0 = p_params->drop_ttl0;
+ start.concrete_fid = p_hwfn->hw_info.concrete_fid;
+ start.opaque_fid = p_hwfn->hw_info.opaque_fid;
+ start.concrete_fid = p_hwfn->hw_info.concrete_fid;
+ start.handle_ptp_pkts = p_params->handle_ptp_pkts;
+ start.vport_id = p_params->vport_id;
+ start.max_buffers_per_cqe = 16; /* TODO-is this right */
+ start.mtu = p_params->mtu;
+
+ rc = ecore_sp_vport_start(p_hwfn, &start);
+ if (rc) {
+ DP_ERR(edev, "Failed to start VPORT\n");
+ return rc;
+ }
+
+ ecore_hw_start_fastpath(p_hwfn);
+
+ DP_VERBOSE(edev, ECORE_MSG_SPQ,
+ "Started V-PORT %d with MTU %d\n",
+ p_params->vport_id, p_params->mtu);
+ }
+
+ ecore_reset_vport_stats(edev);
+
+ return 0;
+}
+
+static int qed_stop_vport(struct ecore_dev *edev, uint8_t vport_id)
+{
+ int rc, i;
+
+ for_each_hwfn(edev, i) {
+ struct ecore_hwfn *p_hwfn = &edev->hwfns[i];
+ rc = ecore_sp_vport_stop(p_hwfn,
+ p_hwfn->hw_info.opaque_fid, vport_id);
+
+ if (rc) {
+ DP_ERR(edev, "Failed to stop VPORT\n");
+ return rc;
+ }
+ }
+
+ return 0;
+}
+
+static int
+qed_update_vport(struct ecore_dev *edev, struct qed_update_vport_params *params)
+{
+ struct ecore_sp_vport_update_params sp_params;
+ struct ecore_rss_params sp_rss_params;
+ int rc, i;
+
+ memset(&sp_params, 0, sizeof(sp_params));
+ memset(&sp_rss_params, 0, sizeof(sp_rss_params));
+
+ /* Translate protocol params into sp params */
+ sp_params.vport_id = params->vport_id;
+ sp_params.update_vport_active_rx_flg = params->update_vport_active_flg;
+ sp_params.update_vport_active_tx_flg = params->update_vport_active_flg;
+ sp_params.vport_active_rx_flg = params->vport_active_flg;
+ sp_params.vport_active_tx_flg = params->vport_active_flg;
+ sp_params.update_inner_vlan_removal_flg =
+ params->update_inner_vlan_removal_flg;
+ sp_params.inner_vlan_removal_flg = params->inner_vlan_removal_flg;
+ sp_params.update_tx_switching_flg = params->update_tx_switching_flg;
+ sp_params.tx_switching_flg = params->tx_switching_flg;
+ sp_params.accept_any_vlan = params->accept_any_vlan;
+ sp_params.update_accept_any_vlan_flg =
+ params->update_accept_any_vlan_flg;
+
+ /* RSS - is a bit tricky, since upper-layer isn't familiar with hwfns.
+ * We need to re-fix the rss values per engine for CMT.
+ */
+
+ if (edev->num_hwfns > 1 && params->update_rss_flg) {
+ struct qed_update_vport_rss_params *rss = ¶ms->rss_params;
+ int k, max = 0;
+
+ /* Find largest entry, since it's possible RSS needs to
+ * be disabled [in case only 1 queue per-hwfn]
+ */
+ for (k = 0; k < ECORE_RSS_IND_TABLE_SIZE; k++)
+ max = (max > rss->rss_ind_table[k]) ?
+ max : rss->rss_ind_table[k];
+
+ /* Either fix RSS values or disable RSS */
+ if (edev->num_hwfns < max + 1) {
+ int divisor = (max + edev->num_hwfns - 1) /
+ edev->num_hwfns;
+
+ DP_VERBOSE(edev, ECORE_MSG_SPQ,
+ "CMT - fixing RSS values (modulo %02x)\n",
+ divisor);
+
+ for (k = 0; k < ECORE_RSS_IND_TABLE_SIZE; k++)
+ rss->rss_ind_table[k] =
+ rss->rss_ind_table[k] % divisor;
+ } else {
+ DP_VERBOSE(edev, ECORE_MSG_SPQ,
+ "CMT - 1 queue per-hwfn; Disabling RSS\n");
+ params->update_rss_flg = 0;
+ }
+ }
+
+ /* Now, update the RSS configuration for actual configuration */
+ if (params->update_rss_flg) {
+ sp_rss_params.update_rss_config = 1;
+ sp_rss_params.rss_enable = 1;
+ sp_rss_params.update_rss_capabilities = 1;
+ sp_rss_params.update_rss_ind_table = 1;
+ sp_rss_params.update_rss_key = 1;
+ sp_rss_params.rss_caps = ECORE_RSS_IPV4 | ECORE_RSS_IPV6 |
+ ECORE_RSS_IPV4_TCP | ECORE_RSS_IPV6_TCP;
+ sp_rss_params.rss_table_size_log = 7; /* 2^7 = 128 */
+ rte_memcpy(sp_rss_params.rss_ind_table,
+ params->rss_params.rss_ind_table,
+ ECORE_RSS_IND_TABLE_SIZE * sizeof(uint16_t));
+ rte_memcpy(sp_rss_params.rss_key, params->rss_params.rss_key,
+ ECORE_RSS_KEY_SIZE * sizeof(uint32_t));
+ }
+ sp_params.rss_params = &sp_rss_params;
+
+ for_each_hwfn(edev, i) {
+ struct ecore_hwfn *p_hwfn = &edev->hwfns[i];
+
+ sp_params.opaque_fid = p_hwfn->hw_info.opaque_fid;
+ rc = ecore_sp_vport_update(p_hwfn, &sp_params,
+ ECORE_SPQ_MODE_EBLOCK, NULL);
+ if (rc) {
+ DP_ERR(edev, "Failed to update VPORT\n");
+ return rc;
+ }
+
+ DP_VERBOSE(edev, ECORE_MSG_SPQ,
+ "Updated V-PORT %d: active_flag %d [update %d]\n",
+ params->vport_id, params->vport_active_flg,
+ params->update_vport_active_flg);
+ }
+
+ return 0;
+}
+
+static int
+qed_start_rxq(struct ecore_dev *edev,
+ uint8_t rss_id, uint8_t rx_queue_id,
+ uint8_t vport_id, uint16_t sb,
+ uint8_t sb_index, uint16_t bd_max_bytes,
+ dma_addr_t bd_chain_phys_addr,
+ dma_addr_t cqe_pbl_addr,
+ uint16_t cqe_pbl_size, void OSAL_IOMEM**pp_prod)
+{
+ struct ecore_hwfn *p_hwfn;
+ int rc, hwfn_index;
+
+ hwfn_index = rss_id % edev->num_hwfns;
+ p_hwfn = &edev->hwfns[hwfn_index];
+
+ rc = ecore_sp_eth_rx_queue_start(p_hwfn,
+ p_hwfn->hw_info.opaque_fid,
+ rx_queue_id / edev->num_hwfns,
+ vport_id,
+ vport_id,
+ sb,
+ sb_index,
+ bd_max_bytes,
+ bd_chain_phys_addr,
+ cqe_pbl_addr, cqe_pbl_size, pp_prod);
+
+ if (rc) {
+ DP_ERR(edev, "Failed to start RXQ#%d\n", rx_queue_id);
+ return rc;
+ }
+
+ DP_VERBOSE(edev, ECORE_MSG_SPQ,
+ "Started RX-Q %d [rss %d] on V-PORT %d and SB %d\n",
+ rx_queue_id, rss_id, vport_id, sb);
+
+ return 0;
+}
+
+static int
+qed_stop_rxq(struct ecore_dev *edev, struct qed_stop_rxq_params *params)
+{
+ int rc, hwfn_index;
+ struct ecore_hwfn *p_hwfn;
+
+ hwfn_index = params->rss_id % edev->num_hwfns;
+ p_hwfn = &edev->hwfns[hwfn_index];
+
+ rc = ecore_sp_eth_rx_queue_stop(p_hwfn,
+ params->rx_queue_id / edev->num_hwfns,
+ params->eq_completion_only, false);
+ if (rc) {
+ DP_ERR(edev, "Failed to stop RXQ#%d\n", params->rx_queue_id);
+ return rc;
+ }
+
+ return 0;
+}
+
+static int
+qed_start_txq(struct ecore_dev *edev,
+ uint8_t rss_id, uint16_t tx_queue_id,
+ uint8_t vport_id, uint16_t sb,
+ uint8_t sb_index,
+ dma_addr_t pbl_addr,
+ uint16_t pbl_size, void OSAL_IOMEM**pp_doorbell)
+{
+ struct ecore_hwfn *p_hwfn;
+ int rc, hwfn_index;
+
+ hwfn_index = rss_id % edev->num_hwfns;
+ p_hwfn = &edev->hwfns[hwfn_index];
+
+ rc = ecore_sp_eth_tx_queue_start(p_hwfn,
+ p_hwfn->hw_info.opaque_fid,
+ tx_queue_id / edev->num_hwfns,
+ vport_id,
+ vport_id,
+ sb,
+ sb_index,
+ pbl_addr, pbl_size, pp_doorbell);
+
+ if (rc) {
+ DP_ERR(edev, "Failed to start TXQ#%d\n", tx_queue_id);
+ return rc;
+ }
+
+ DP_VERBOSE(edev, ECORE_MSG_SPQ,
+ "Started TX-Q %d [rss %d] on V-PORT %d and SB %d\n",
+ tx_queue_id, rss_id, vport_id, sb);
+
+ return 0;
+}
+
+static int
+qed_stop_txq(struct ecore_dev *edev, struct qed_stop_txq_params *params)
+{
+ struct ecore_hwfn *p_hwfn;
+ int rc, hwfn_index;
+
+ hwfn_index = params->rss_id % edev->num_hwfns;
+ p_hwfn = &edev->hwfns[hwfn_index];
+
+ rc = ecore_sp_eth_tx_queue_stop(p_hwfn,
+ params->tx_queue_id / edev->num_hwfns);
+ if (rc) {
+ DP_ERR(edev, "Failed to stop TXQ#%d\n", params->tx_queue_id);
+ return rc;
+ }
+
+ return 0;
+}
+
+static int
+qed_fp_cqe_completion(struct ecore_dev *edev,
+ uint8_t rss_id, struct eth_slow_path_rx_cqe *cqe)
+{
+
+ return ecore_eth_cqe_completion(&edev->hwfns[rss_id % edev->num_hwfns],
+ cqe);
+}
+
+static int qed_fastpath_stop(struct ecore_dev *edev)
+{
+ ecore_hw_stop_fastpath(edev);
+
+ return 0;
+}
+
+static void
+qed_get_vport_stats(struct ecore_dev *edev, struct ecore_eth_stats *stats)
+{
+ ecore_get_vport_stats(edev, stats);
+}
+
+static int
+qed_configure_filter_ucast(struct ecore_dev *edev,
+ struct qed_filter_ucast_params *params)
+{
+ struct ecore_filter_ucast ucast;
+
+ if (!params->vlan_valid && !params->mac_valid) {
+ DP_NOTICE(edev, true,
+ "Tried configuring a unicast filter,"
+ "but both MAC and VLAN are not set\n");
+ return -EINVAL;
+ }
+
+ memset(&ucast, 0, sizeof(ucast));
+ switch (params->type) {
+ case QED_FILTER_XCAST_TYPE_ADD:
+ ucast.opcode = ECORE_FILTER_ADD;
+ break;
+ case QED_FILTER_XCAST_TYPE_DEL:
+ ucast.opcode = ECORE_FILTER_REMOVE;
+ break;
+ case QED_FILTER_XCAST_TYPE_REPLACE:
+ ucast.opcode = ECORE_FILTER_REPLACE;
+ break;
+ default:
+ DP_NOTICE(edev, true, "Unknown unicast filter type %d\n",
+ params->type);
+ }
+
+ if (params->vlan_valid && params->mac_valid) {
+ ucast.type = ECORE_FILTER_MAC_VLAN;
+ ether_addr_copy((struct ether_addr *)¶ms->mac,
+ (struct ether_addr *)&ucast.mac);
+ ucast.vlan = params->vlan;
+ } else if (params->mac_valid) {
+ ucast.type = ECORE_FILTER_MAC;
+ ether_addr_copy((struct ether_addr *)¶ms->mac,
+ (struct ether_addr *)&ucast.mac);
+ } else {
+ ucast.type = ECORE_FILTER_VLAN;
+ ucast.vlan = params->vlan;
+ }
+
+ ucast.is_rx_filter = true;
+ ucast.is_tx_filter = true;
+
+ return ecore_filter_ucast_cmd(edev, &ucast, ECORE_SPQ_MODE_CB, NULL);
+}
+
+static int
+qed_configure_filter_mcast(struct ecore_dev *edev,
+ struct qed_filter_mcast_params *params)
+{
+ struct ecore_filter_mcast mcast;
+ int i;
+
+ memset(&mcast, 0, sizeof(mcast));
+ switch (params->type) {
+ case QED_FILTER_XCAST_TYPE_ADD:
+ mcast.opcode = ECORE_FILTER_ADD;
+ break;
+ case QED_FILTER_XCAST_TYPE_DEL:
+ mcast.opcode = ECORE_FILTER_REMOVE;
+ break;
+ default:
+ DP_NOTICE(edev, true, "Unknown multicast filter type %d\n",
+ params->type);
+ }
+
+ mcast.num_mc_addrs = params->num;
+ for (i = 0; i < mcast.num_mc_addrs; i++)
+ ether_addr_copy((struct ether_addr *)¶ms->mac[i],
+ (struct ether_addr *)&mcast.mac[i]);
+
+ return ecore_filter_mcast_cmd(edev, &mcast, ECORE_SPQ_MODE_CB, NULL);
+}
+
+int
+qed_configure_filter_rx_mode(struct ecore_dev *edev,
+ enum qed_filter_rx_mode_type type)
+{
+ struct ecore_filter_accept_flags accept_flags;
+
+ memset(&accept_flags, 0, sizeof(accept_flags));
+
+ accept_flags.update_rx_mode_config = 1;
+ accept_flags.update_tx_mode_config = 1;
+ accept_flags.rx_accept_filter = ECORE_ACCEPT_UCAST_MATCHED |
+ ECORE_ACCEPT_MCAST_MATCHED;
+ ECORE_ACCEPT_BCAST;
+ accept_flags.tx_accept_filter = ECORE_ACCEPT_UCAST_MATCHED |
+ ECORE_ACCEPT_MCAST_MATCHED | ECORE_ACCEPT_BCAST;
+
+ if (type == QED_FILTER_RX_MODE_TYPE_PROMISC)
+ accept_flags.rx_accept_filter |= ECORE_ACCEPT_UCAST_UNMATCHED;
+ else if (type == QED_FILTER_RX_MODE_TYPE_MULTI_PROMISC)
+ accept_flags.rx_accept_filter |= ECORE_ACCEPT_MCAST_UNMATCHED;
+ else if (type == (QED_FILTER_RX_MODE_TYPE_MULTI_PROMISC |
+ QED_FILTER_RX_MODE_TYPE_PROMISC))
+ accept_flags.rx_accept_filter |= ECORE_ACCEPT_UCAST_UNMATCHED |
+ ECORE_ACCEPT_MCAST_UNMATCHED;
+
+ return ecore_filter_accept_cmd(edev, 0, accept_flags, false, false,
+ ECORE_SPQ_MODE_CB, NULL);
+}
+
+static int
+qed_configure_filter(struct ecore_dev *edev, struct qed_filter_params *params)
+{
+ switch (params->type) {
+ case QED_FILTER_TYPE_UCAST:
+ return qed_configure_filter_ucast(edev, ¶ms->filter.ucast);
+ case QED_FILTER_TYPE_MCAST:
+ return qed_configure_filter_mcast(edev, ¶ms->filter.mcast);
+ case QED_FILTER_TYPE_RX_MODE:
+ return qed_configure_filter_rx_mode(edev,
+ params->filter.
+ accept_flags);
+ default:
+ DP_NOTICE(edev, true, "Unknown filter type %d\n",
+ (int)params->type);
+ return -EINVAL;
+ }
+}
+
+static const struct qed_eth_ops qed_eth_ops_pass = {
+ INIT_STRUCT_FIELD(common, &qed_common_ops_pass),
+ INIT_STRUCT_FIELD(fill_dev_info, &qed_fill_eth_dev_info),
+ INIT_STRUCT_FIELD(vport_start, &qed_start_vport),
+ INIT_STRUCT_FIELD(vport_stop, &qed_stop_vport),
+ INIT_STRUCT_FIELD(vport_update, &qed_update_vport),
+ INIT_STRUCT_FIELD(q_rx_start, &qed_start_rxq),
+ INIT_STRUCT_FIELD(q_tx_start, &qed_start_txq),
+ INIT_STRUCT_FIELD(q_rx_stop, &qed_stop_rxq),
+ INIT_STRUCT_FIELD(q_tx_stop, &qed_stop_txq),
+ INIT_STRUCT_FIELD(eth_cqe_completion, &qed_fp_cqe_completion),
+ INIT_STRUCT_FIELD(fastpath_stop, &qed_fastpath_stop),
+ INIT_STRUCT_FIELD(get_vport_stats, &qed_get_vport_stats),
+ INIT_STRUCT_FIELD(filter_config, &qed_configure_filter),
+};
+
+uint32_t qed_get_protocol_version(enum qed_protocol protocol)
+{
+ switch (protocol) {
+ case QED_PROTOCOL_ETH:
+ return QED_ETH_INTERFACE_VERSION;
+ case QED_PROTOCOL_ISCSI:
+ return QED_ISCSI_INTERFACE_VERSION;
+ default:
+ return 0;
+ }
+}
+
+const struct qed_eth_ops *qed_get_eth_ops(uint32_t version)
+{
+ if (version != QED_ETH_INTERFACE_VERSION)
+ return NULL;
+
+ return &qed_eth_ops_pass;
+}
diff --git a/drivers/net/qede/qede_eth_if.h b/drivers/net/qede/qede_eth_if.h
new file mode 100644
index 0000000..2fe5832
--- /dev/null
+++ b/drivers/net/qede/qede_eth_if.h
@@ -0,0 +1,180 @@
+/*
+ * Copyright (c) 2016 QLogic Corporation.
+ * All rights reserved.
+ * www.qlogic.com
+ *
+ * See LICENSE.qede_pmd for copyright and licensing details.
+ */
+
+#ifndef _QEDE_ETH_IF_H
+#define _QEDE_ETH_IF_H
+
+#include "qede_if.h"
+
+/*forward decl */
+struct eth_slow_path_rx_cqe;
+
+#define INIT_STRUCT_FIELD(field, value) .field = value
+
+#define QED_ETH_INTERFACE_VERSION 609
+#define QED_ISCSI_INTERFACE_VERSION 609
+#define QED_ROCE_INTERFACE_VERSION 609
+
+enum qed_filter_rx_mode_type {
+ QED_FILTER_RX_MODE_TYPE_REGULAR,
+ QED_FILTER_RX_MODE_TYPE_MULTI_PROMISC,
+ QED_FILTER_RX_MODE_TYPE_PROMISC,
+};
+
+enum qed_filter_xcast_params_type {
+ QED_FILTER_XCAST_TYPE_ADD,
+ QED_FILTER_XCAST_TYPE_DEL,
+ QED_FILTER_XCAST_TYPE_REPLACE,
+};
+
+enum qed_filter_type {
+ QED_FILTER_TYPE_UCAST,
+ QED_FILTER_TYPE_MCAST,
+ QED_FILTER_TYPE_RX_MODE,
+ QED_MAX_FILTER_TYPES,
+};
+
+struct qed_dev_eth_info {
+ struct qed_dev_info common;
+
+ uint8_t num_queues;
+ uint8_t num_tc;
+
+ struct ether_addr port_mac;
+ uint8_t num_vlan_filters;
+};
+
+struct qed_update_vport_rss_params {
+ uint16_t rss_ind_table[128];
+ uint32_t rss_key[10];
+};
+
+struct qed_stop_rxq_params {
+ uint8_t rss_id;
+ uint8_t rx_queue_id;
+ uint8_t vport_id;
+ bool eq_completion_only;
+};
+
+struct qed_update_vport_params {
+ uint8_t vport_id;
+ uint8_t update_vport_active_flg;
+ uint8_t vport_active_flg;
+ uint8_t update_inner_vlan_removal_flg;
+ uint8_t inner_vlan_removal_flg;
+ uint8_t update_tx_switching_flg;
+ uint8_t tx_switching_flg;
+ uint8_t update_accept_any_vlan_flg;
+ uint8_t accept_any_vlan;
+ uint8_t update_rss_flg;
+ struct qed_update_vport_rss_params rss_params;
+};
+
+struct qed_start_vport_params {
+ bool remove_inner_vlan;
+ bool handle_ptp_pkts;
+ bool gro_enable;
+ bool drop_ttl0;
+ uint8_t vport_id;
+ uint16_t mtu;
+ bool clear_stats;
+};
+
+struct qed_stop_txq_params {
+ uint8_t rss_id;
+ uint8_t tx_queue_id;
+};
+
+struct qed_filter_ucast_params {
+ enum qed_filter_xcast_params_type type;
+ uint8_t vlan_valid;
+ uint16_t vlan;
+ uint8_t mac_valid;
+ unsigned char mac[ETHER_ADDR_LEN];
+} __attribute__ ((__packed__));
+
+struct qed_filter_mcast_params {
+ enum qed_filter_xcast_params_type type;
+ uint8_t num;
+ unsigned char mac[64][ETHER_ADDR_LEN];
+};
+
+union qed_filter_type_params {
+ enum qed_filter_rx_mode_type accept_flags;
+ struct qed_filter_ucast_params ucast;
+ struct qed_filter_mcast_params mcast;
+};
+
+struct qed_filter_params {
+ enum qed_filter_type type;
+ union qed_filter_type_params filter;
+};
+
+struct qed_eth_ops {
+ const struct qed_common_ops *common;
+
+ int (*fill_dev_info) (struct ecore_dev *edev,
+ struct qed_dev_eth_info *info);
+
+ int (*vport_start) (struct ecore_dev *edev,
+ struct qed_start_vport_params *params);
+
+ int (*vport_stop) (struct ecore_dev *edev, uint8_t vport_id);
+
+ int (*vport_update) (struct ecore_dev *edev,
+ struct qed_update_vport_params *params);
+
+ int (*q_rx_start) (struct ecore_dev *cdev,
+ uint8_t rss_id, uint8_t rx_queue_id,
+ uint8_t vport_id, uint16_t sb,
+ uint8_t sb_index, uint16_t bd_max_bytes,
+ dma_addr_t bd_chain_phys_addr,
+ dma_addr_t cqe_pbl_addr,
+ uint16_t cqe_pbl_size, void OSAL_IOMEM**pp_prod);
+
+ int (*q_rx_stop) (struct ecore_dev *edev,
+ struct qed_stop_rxq_params *params);
+
+ int (*q_tx_start) (struct ecore_dev *edev,
+ uint8_t rss_id, uint16_t tx_queue_id,
+ uint8_t vport_id, uint16_t sb,
+ uint8_t sb_index,
+ dma_addr_t pbl_addr,
+ uint16_t pbl_size, void OSAL_IOMEM**pp_doorbell);
+
+ int (*q_tx_stop) (struct ecore_dev *edev,
+ struct qed_stop_txq_params *params);
+
+ int (*eth_cqe_completion) (struct ecore_dev *edev,
+ uint8_t rss_id,
+ struct eth_slow_path_rx_cqe *cqe);
+
+ int (*fastpath_stop) (struct ecore_dev *edev);
+
+ void (*get_vport_stats) (struct ecore_dev *edev,
+ struct ecore_eth_stats *stats);
+
+ int (*filter_config) (struct ecore_dev *edev,
+ struct qed_filter_params *params);
+};
+
+/* externs */
+
+extern const struct qed_common_ops qed_common_ops_pass;
+
+extern int qed_fill_eth_dev_info(struct ecore_dev *edev,
+ struct qed_dev_eth_info *info);
+
+const struct qed_eth_ops *qed_get_eth_ops(uint32_t version);
+
+void qed_put_eth_ops(void);
+
+int qed_configure_filter_rx_mode(struct ecore_dev *edev,
+ enum qed_filter_rx_mode_type type);
+
+#endif /* _QEDE_ETH_IF_H */
diff --git a/drivers/net/qede/qede_ethdev.c b/drivers/net/qede/qede_ethdev.c
new file mode 100644
index 0000000..d991ca5
--- /dev/null
+++ b/drivers/net/qede/qede_ethdev.c
@@ -0,0 +1,815 @@
+/*
+ * Copyright (c) 2016 QLogic Corporation.
+ * All rights reserved.
+ * www.qlogic.com
+ *
+ * See LICENSE.qede_pmd for copyright and licensing details.
+ */
+
+#include "qede_ethdev.h"
+
+/* Globals */
+static const struct qed_eth_ops *qed_ops;
+static const char *drivername = "qede pmd";
+
+static void qede_interrupt_action(struct ecore_hwfn *p_hwfn)
+{
+ ecore_int_sp_dpc((osal_int_ptr_t) (p_hwfn));
+}
+
+static void
+qede_interrupt_handler(__rte_unused struct rte_intr_handle *handle, void *param)
+{
+ struct rte_eth_dev *eth_dev = (struct rte_eth_dev *)param;
+ struct qede_dev *qdev = eth_dev->data->dev_private;
+ struct ecore_dev *edev = &qdev->edev;
+
+ qede_interrupt_action(ECORE_LEADING_HWFN(edev));
+ if (rte_intr_enable(&(eth_dev->pci_dev->intr_handle)))
+ DP_ERR(edev, "rte_intr_enable failed\n");
+}
+
+static void
+qede_alloc_etherdev(struct qede_dev *qdev, struct qed_dev_eth_info *info)
+{
+ rte_memcpy(&qdev->dev_info, info, sizeof(*info));
+ qdev->num_tc = qdev->dev_info.num_tc;
+ qdev->ops = qed_ops;
+}
+
+static void qede_print_adapter_info(struct qede_dev *qdev)
+{
+ struct ecore_dev *edev = &qdev->edev;
+ struct qed_dev_info *info = &qdev->dev_info.common;
+ char ver_str[QED_DRV_VER_STR_SIZE] = { 0 };
+
+ RTE_LOG(INFO, PMD,
+ " Chip details : %s%d\n",
+ ECORE_IS_BB(edev) ? "BB" : "AH",
+ CHIP_REV_IS_A0(edev) ? 0 : 1);
+
+ sprintf(ver_str, "%s %s_%d.%d.%d", QEDE_PMD_VER_PREFIX,
+ edev->ver_str, QEDE_PMD_VERSION_MAJOR,
+ QEDE_PMD_VERSION_MINOR, QEDE_PMD_VERSION_PATCH);
+ strcpy(qdev->drv_ver, ver_str);
+ RTE_LOG(INFO, PMD, " Driver version : %s\n", ver_str);
+
+ ver_str[0] = '\0';
+ sprintf(ver_str, "%d.%d.%d.%d", info->fw_major, info->fw_minor,
+ info->fw_rev, info->fw_eng);
+ RTE_LOG(INFO, PMD, " Firmware version : %s\n", ver_str);
+
+ ver_str[0] = '\0';
+ sprintf(ver_str, "%d.%d.%d.%d",
+ (info->mfw_rev >> 24) & 0xff,
+ (info->mfw_rev >> 16) & 0xff,
+ (info->mfw_rev >> 8) & 0xff, (info->mfw_rev) & 0xff);
+ RTE_LOG(INFO, PMD, " Management firmware version : %s\n", ver_str);
+
+ RTE_LOG(INFO, PMD, " Firmware file : %s\n", QEDE_FW_FILE_NAME);
+}
+
+static int
+qede_set_ucast_rx_mac(struct qede_dev *qdev,
+ enum qed_filter_xcast_params_type opcode,
+ uint8_t mac[ETHER_ADDR_LEN])
+{
+ struct ecore_dev *edev = &qdev->edev;
+ struct qed_filter_params filter_cmd;
+
+ memset(&filter_cmd, 0, sizeof(filter_cmd));
+ filter_cmd.type = QED_FILTER_TYPE_UCAST;
+ filter_cmd.filter.ucast.type = opcode;
+ filter_cmd.filter.ucast.mac_valid = 1;
+ rte_memcpy(&filter_cmd.filter.ucast.mac[0], &mac[0], ETHER_ADDR_LEN);
+ return qdev->ops->filter_config(edev, &filter_cmd);
+}
+
+static void
+qede_mac_addr_add(struct rte_eth_dev *eth_dev, struct ether_addr *mac_addr,
+ __rte_unused uint32_t index, __rte_unused uint32_t pool)
+{
+ struct qede_dev *qdev = eth_dev->data->dev_private;
+ struct ecore_dev *edev = &qdev->edev;
+ int rc;
+
+ PMD_INIT_FUNC_TRACE(edev);
+
+ DP_NOTICE(edev, false, "%s\n", __func__);
+
+ /* Skip adding macaddr if promiscuous mode is set */
+ if (rte_eth_promiscuous_get(eth_dev->data->port_id) == 1) {
+ DP_NOTICE(edev, false, "Port is in promiscuous mode\n");
+ return;
+ }
+
+ /* Add MAC filters according to the unicast secondary macs */
+ rc = qede_set_ucast_rx_mac(qdev, QED_FILTER_XCAST_TYPE_ADD,
+ mac_addr->addr_bytes);
+ if (rc)
+ DP_ERR(edev, "Unable to add filter\n");
+}
+
+static void
+qede_mac_addr_remove(__rte_unused struct rte_eth_dev *eth_dev,
+ __rte_unused uint32_t index)
+{
+ struct qede_dev *qdev = eth_dev->data->dev_private;
+ struct ecore_dev *edev = &qdev->edev;
+
+ /* TBD: Not implemented currently because DPDK does not provide
+ * macaddr and instead just passes the index. So pmd needs to
+ * maintain index mapping to macaddr.
+ */
+ DP_NOTICE(edev, false, "%s: Unsupported operation\n", __func__);
+}
+
+static void qede_config_accept_any_vlan(struct qede_dev *qdev, bool action)
+{
+ struct ecore_dev *edev = &qdev->edev;
+ struct qed_update_vport_params params;
+ int rc;
+
+ /* Proceed only if action actually needs to be performed */
+ if (qdev->accept_any_vlan == action)
+ return;
+
+ memset(¶ms, 0, sizeof(params));
+
+ params.vport_id = 0;
+ params.accept_any_vlan = action;
+ params.update_accept_any_vlan_flg = 1;
+
+ rc = qdev->ops->vport_update(edev, ¶ms);
+ if (rc) {
+ DP_ERR(edev, "Failed to %s accept-any-vlan\n",
+ action ? "enable" : "disable");
+ } else {
+ DP_INFO(edev, "%s accept-any-vlan\n",
+ action ? "enabled" : "disabled");
+ qdev->accept_any_vlan = action;
+ }
+}
+
+void qede_config_rx_mode(struct rte_eth_dev *eth_dev)
+{
+ struct qede_dev *qdev = eth_dev->data->dev_private;
+ struct ecore_dev *edev = &qdev->edev;
+ /* TODO: - QED_FILTER_TYPE_UCAST */
+ enum qed_filter_rx_mode_type accept_flags =
+ QED_FILTER_RX_MODE_TYPE_REGULAR;
+ struct qed_filter_params rx_mode;
+ int rc;
+
+ /* Configure the struct for the Rx mode */
+ memset(&rx_mode, 0, sizeof(struct qed_filter_params));
+ rx_mode.type = QED_FILTER_TYPE_RX_MODE;
+
+ rc = qede_set_ucast_rx_mac(qdev, QED_FILTER_XCAST_TYPE_REPLACE,
+ eth_dev->data->mac_addrs[0].addr_bytes);
+ if (rte_eth_promiscuous_get(eth_dev->data->port_id) == 1) {
+ accept_flags = QED_FILTER_RX_MODE_TYPE_PROMISC;
+ } else {
+ rc = qede_set_ucast_rx_mac(qdev, QED_FILTER_XCAST_TYPE_ADD,
+ eth_dev->data->
+ mac_addrs[0].addr_bytes);
+ if (rc) {
+ DP_ERR(edev, "Unable to add filter\n");
+ return;
+ }
+ }
+
+ /* take care of VLAN mode */
+ if (rte_eth_promiscuous_get(eth_dev->data->port_id) == 1)
+ qede_config_accept_any_vlan(qdev, true);
+ else if (!qdev->non_configured_vlans) {
+ /* If we dont have non-configured VLANs and promisc
+ * is not set, then check if we need to disable
+ * accept_any_vlan mode.
+ * Because in this case, accept_any_vlan mode is set
+ * as part of IFF_RPOMISC flag handling.
+ */
+ qede_config_accept_any_vlan(qdev, false);
+ }
+ rx_mode.filter.accept_flags = accept_flags;
+ (void)qdev->ops->filter_config(edev, &rx_mode);
+}
+
+static int qede_dev_configure(struct rte_eth_dev *eth_dev)
+{
+ struct qede_dev *qdev = eth_dev->data->dev_private;
+ struct ecore_dev *edev = &qdev->edev;
+ int rc = 0;
+
+ PMD_INIT_FUNC_TRACE(edev);
+
+ if (eth_dev->data->nb_rx_queues != eth_dev->data->nb_tx_queues) {
+ DP_NOTICE(edev, false,
+ "unequal number of rx/tx queues"
+ "is not supported RX=%u TX=%u\n",
+ eth_dev->data->nb_rx_queues,
+ eth_dev->data->nb_tx_queues);
+ return -EINVAL;
+ }
+
+ qdev->num_rss = eth_dev->data->nb_rx_queues;
+
+ /* Initial state */
+ qdev->state = QEDE_CLOSE;
+
+ if (!eth_dev->data->dev_conf.rxmode.hw_strip_crc)
+ DP_INFO(edev, "crc stripping is always enabled in hw\n");
+
+ if (eth_dev->data->dev_conf.rxmode.enable_scatter == 1) {
+ DP_ERR(edev, "RX scatter packets is not supported\n");
+ return -EINVAL;
+ }
+
+ DP_INFO(edev, "Allocated %d RSS queues on %d TC/s\n",
+ QEDE_RSS_CNT(qdev), qdev->num_tc);
+
+ DP_INFO(edev, "my_id %u rel_pf_id %u abs_pf_id %u"
+ " port %u first_on_engine %d\n",
+ edev->hwfns[0].my_id,
+ edev->hwfns[0].rel_pf_id,
+ edev->hwfns[0].abs_pf_id,
+ edev->hwfns[0].port_id, edev->hwfns[0].first_on_engine);
+
+ return 0;
+}
+
+/* Info about HW descriptor ring limitations */
+static const struct rte_eth_desc_lim qede_rx_desc_lim = {
+ .nb_max = NUM_RX_BDS_MAX,
+ .nb_min = 128,
+ .nb_align = 128 /* lowest common multiple */
+};
+
+static const struct rte_eth_desc_lim qede_tx_desc_lim = {
+ .nb_max = NUM_TX_BDS_MAX,
+ .nb_min = 256,
+ .nb_align = 256
+};
+
+static void
+qede_dev_info_get(struct rte_eth_dev *eth_dev,
+ struct rte_eth_dev_info *dev_info)
+{
+ struct qede_dev *qdev = eth_dev->data->dev_private;
+ struct ecore_dev *edev = &qdev->edev;
+
+ PMD_INIT_FUNC_TRACE(edev);
+
+ dev_info->min_rx_bufsize = (uint32_t) (ETHER_MIN_MTU +
+ QEDE_ETH_OVERHEAD);
+ dev_info->max_rx_pktlen = (uint32_t) ETH_TX_MAX_NON_LSO_PKT_LEN;
+ dev_info->rx_desc_lim = qede_rx_desc_lim;
+ dev_info->tx_desc_lim = qede_tx_desc_lim;
+ dev_info->max_rx_queues = (uint16_t) QEDE_MAX_RSS_CNT(qdev);
+ dev_info->max_tx_queues = (uint16_t) QEDE_MAX_TSS_CNT(qdev);
+ dev_info->max_mac_addrs = (uint32_t) (RESC_NUM(&edev->hwfns[0],
+ ECORE_MAC));
+ dev_info->max_vfs = (uint16_t) NUM_OF_VFS(&qdev->edev);
+ dev_info->driver_name = qdev->drv_ver;
+ dev_info->reta_size = ETH_RSS_RETA_SIZE_128;
+ dev_info->flow_type_rss_offloads = (uint64_t) QEDE_RSS_OFFLOAD_ALL;
+ dev_info->default_txconf = (struct rte_eth_txconf) {
+ .txq_flags = QEDE_TXQ_FLAGS,};
+ dev_info->rx_offload_capa = (DEV_RX_OFFLOAD_VLAN_STRIP |
+ DEV_RX_OFFLOAD_IPV4_CKSUM |
+ DEV_RX_OFFLOAD_UDP_CKSUM |
+ DEV_RX_OFFLOAD_TCP_CKSUM);
+ dev_info->tx_offload_capa = (DEV_TX_OFFLOAD_VLAN_INSERT |
+ DEV_TX_OFFLOAD_IPV4_CKSUM |
+ DEV_TX_OFFLOAD_UDP_CKSUM |
+ DEV_TX_OFFLOAD_TCP_CKSUM);
+}
+
+/* return 0 means link status changed, -1 means not changed */
+static int
+qede_link_update(struct rte_eth_dev *eth_dev, __rte_unused int wait_to_complete)
+{
+ struct qede_dev *qdev = eth_dev->data->dev_private;
+ struct ecore_dev *edev = &qdev->edev;
+ uint16_t speed, link_duplex;
+ /* Possible to use ethtool also */
+ struct qed_link_output link;
+ struct rte_eth_link *old = ð_dev->data->dev_link;
+
+ memset(&link, 0, sizeof(struct qed_link_output));
+ qdev->ops->common->get_link(edev, &link);
+ if (old->link_status == link.link_up)
+ return -1;
+
+ /* Speed */
+ switch (link.speed) {
+ case 40000:
+ speed = ETH_LINK_SPEED_40G;
+ break;
+ case 25000:
+ speed = 25000; /* Hardcode: don't have ETH_LINK_SPEED_25G */
+ break;
+ default:
+ DP_ERR(edev, "Unknown link speed %u\n", link.speed);
+ speed = 0;
+ }
+
+ eth_dev->data->dev_link.link_speed = speed;
+
+ /* Duplex/Simplex */
+ switch (link.duplex) {
+ case QEDE_DUPLEX_HALF:
+ link_duplex = ETH_LINK_HALF_DUPLEX;
+ break;
+ case QEDE_DUPLEX_FULL:
+ link_duplex = ETH_LINK_FULL_DUPLEX;
+ break;
+ case QEDE_DUPLEX_UNKNOWN:
+ default:
+ link_duplex = -1;
+ }
+
+ eth_dev->data->dev_link.link_duplex = link_duplex;
+ eth_dev->data->dev_link.link_status = link.link_up;
+
+ /* Link state changed */
+ return 0;
+}
+
+static void
+qede_rx_mode_setting(struct rte_eth_dev *eth_dev,
+ enum qed_filter_rx_mode_type accept_flags)
+{
+ struct qede_dev *qdev = eth_dev->data->dev_private;
+ struct ecore_dev *edev = &qdev->edev;
+ struct qed_filter_params rx_mode;
+
+ DP_INFO(edev, "%s mode %u\n", __func__, accept_flags);
+
+ memset(&rx_mode, 0, sizeof(struct qed_filter_params));
+ rx_mode.type = QED_FILTER_TYPE_RX_MODE;
+ rx_mode.filter.accept_flags = accept_flags;
+ qdev->ops->filter_config(edev, &rx_mode);
+}
+
+static void qede_promiscuous_enable(struct rte_eth_dev *eth_dev)
+{
+ struct qede_dev *qdev = eth_dev->data->dev_private;
+ struct ecore_dev *edev = &qdev->edev;
+
+ PMD_INIT_FUNC_TRACE(edev);
+
+ enum qed_filter_rx_mode_type type = QED_FILTER_RX_MODE_TYPE_PROMISC;
+
+ if (rte_eth_allmulticast_get(eth_dev->data->port_id) == 1)
+ type |= QED_FILTER_RX_MODE_TYPE_MULTI_PROMISC;
+
+ qede_rx_mode_setting(eth_dev, type);
+}
+
+static void qede_promiscuous_disable(struct rte_eth_dev *eth_dev)
+{
+ struct qede_dev *qdev = eth_dev->data->dev_private;
+ struct ecore_dev *edev = &qdev->edev;
+
+ PMD_INIT_FUNC_TRACE(edev);
+
+ if (rte_eth_allmulticast_get(eth_dev->data->port_id) == 1)
+ qede_rx_mode_setting(eth_dev,
+ QED_FILTER_RX_MODE_TYPE_MULTI_PROMISC);
+ else
+ qede_rx_mode_setting(eth_dev, QED_FILTER_RX_MODE_TYPE_REGULAR);
+}
+
+static void qede_dev_close(struct rte_eth_dev *eth_dev)
+{
+ struct qede_dev *qdev = eth_dev->data->dev_private;
+ struct ecore_dev *edev = &qdev->edev;
+
+ PMD_INIT_FUNC_TRACE(edev);
+
+ /* dev_stop() shall cleanup fp resources in hw but without releasing
+ * dma memories and sw structures so that dev_start() can be called
+ * by the app without reconfiguration. However, in dev_close() we
+ * can release all the resources and device can be brought up newly
+ */
+ if (qdev->state != QEDE_STOP)
+ qede_dev_stop(eth_dev);
+ else
+ DP_INFO(edev, "Device is already stopped\n");
+
+ qede_free_mem_load(qdev);
+
+ qede_free_fp_arrays(qdev);
+
+ qede_dev_set_link_state(eth_dev, false);
+
+ qdev->ops->common->slowpath_stop(edev);
+
+ qdev->ops->common->remove(edev);
+
+ rte_intr_disable(&(eth_dev->pci_dev->intr_handle));
+
+ rte_intr_callback_unregister(&(eth_dev->pci_dev->intr_handle),
+ qede_interrupt_handler, (void *)eth_dev);
+
+ qdev->state = QEDE_CLOSE;
+}
+
+static void
+qede_get_stats(struct rte_eth_dev *eth_dev, struct rte_eth_stats *eth_stats)
+{
+ struct qede_dev *qdev = eth_dev->data->dev_private;
+ struct ecore_dev *edev = &qdev->edev;
+ struct ecore_eth_stats stats;
+
+ qdev->ops->get_vport_stats(edev, &stats);
+
+ /* RX Stats */
+ eth_stats->ipackets = stats.rx_ucast_pkts +
+ stats.rx_mcast_pkts + stats.rx_bcast_pkts;
+
+ eth_stats->ibytes = stats.rx_ucast_bytes +
+ stats.rx_mcast_bytes + stats.rx_bcast_bytes;
+
+ eth_stats->imcasts = stats.rx_mcast_pkts;
+
+ eth_stats->ierrors = stats.rx_crc_errors +
+ stats.rx_align_errors +
+ stats.rx_carrier_errors +
+ stats.rx_oversize_packets +
+ stats.rx_jabbers + stats.rx_undersize_packets;
+
+ eth_stats->rx_nombuf = stats.no_buff_discards;
+
+ eth_stats->imissed = stats.mftag_filter_discards +
+ stats.mac_filter_discards +
+ stats.no_buff_discards + stats.brb_truncates + stats.brb_discards;
+
+ /* TX stats */
+ eth_stats->opackets = stats.tx_ucast_pkts +
+ stats.tx_mcast_pkts + stats.tx_bcast_pkts;
+
+ eth_stats->obytes = stats.tx_ucast_bytes +
+ stats.tx_mcast_bytes + stats.tx_bcast_bytes;
+
+ eth_stats->oerrors = stats.tx_err_drop_pkts;
+
+ DP_INFO(edev,
+ "no_buff_discards=%" PRIu64 ""
+ " mac_filter_discards=%" PRIu64 ""
+ " brb_truncates=%" PRIu64 ""
+ " brb_discards=%" PRIu64 "\n",
+ stats.no_buff_discards,
+ stats.mac_filter_discards,
+ stats.brb_truncates, stats.brb_discards);
+}
+
+int qede_dev_set_link_state(struct rte_eth_dev *eth_dev, bool link_up)
+{
+ struct qede_dev *qdev = QEDE_INIT_QDEV(eth_dev);
+ struct ecore_dev *edev = QEDE_INIT_EDEV(qdev);
+ struct qed_link_params link_params;
+ int rc;
+
+ DP_INFO(edev, "setting link state %d\n", link_up);
+ memset(&link_params, 0, sizeof(link_params));
+ link_params.link_up = link_up;
+ rc = qdev->ops->common->set_link(edev, &link_params);
+ if (rc != ECORE_SUCCESS)
+ DP_ERR(edev, "Unable to set link state %d\n", link_up);
+
+ return rc;
+}
+
+static int qede_dev_set_link_up(struct rte_eth_dev *eth_dev)
+{
+ return qede_dev_set_link_state(eth_dev, true);
+}
+
+static int qede_dev_set_link_down(struct rte_eth_dev *eth_dev)
+{
+ return qede_dev_set_link_state(eth_dev, false);
+}
+
+static void qede_reset_stats(struct rte_eth_dev *eth_dev)
+{
+ struct qede_dev *qdev = eth_dev->data->dev_private;
+ struct ecore_dev *edev = &qdev->edev;
+
+ ecore_reset_vport_stats(edev);
+}
+
+static void qede_allmulticast_enable(struct rte_eth_dev *eth_dev)
+{
+ enum qed_filter_rx_mode_type type =
+ QED_FILTER_RX_MODE_TYPE_MULTI_PROMISC;
+
+ if (rte_eth_promiscuous_get(eth_dev->data->port_id) == 1)
+ type |= QED_FILTER_RX_MODE_TYPE_PROMISC;
+
+ qede_rx_mode_setting(eth_dev, type);
+}
+
+static void qede_allmulticast_disable(struct rte_eth_dev *eth_dev)
+{
+ if (rte_eth_promiscuous_get(eth_dev->data->port_id) == 1)
+ qede_rx_mode_setting(eth_dev, QED_FILTER_RX_MODE_TYPE_PROMISC);
+ else
+ qede_rx_mode_setting(eth_dev, QED_FILTER_RX_MODE_TYPE_REGULAR);
+}
+
+static struct eth_dev_ops qede_eth_dev_ops = {
+ .dev_configure = qede_dev_configure,
+ .dev_infos_get = qede_dev_info_get,
+ .rx_queue_setup = qede_rx_queue_setup,
+ .rx_queue_release = qede_rx_queue_release,
+ .tx_queue_setup = qede_tx_queue_setup,
+ .tx_queue_release = qede_tx_queue_release,
+ .dev_start = qede_dev_start,
+ .dev_set_link_up = qede_dev_set_link_up,
+ .dev_set_link_down = qede_dev_set_link_down,
+ .link_update = qede_link_update,
+ .promiscuous_enable = qede_promiscuous_enable,
+ .promiscuous_disable = qede_promiscuous_disable,
+ .allmulticast_enable = qede_allmulticast_enable,
+ .allmulticast_disable = qede_allmulticast_disable,
+ .dev_stop = qede_dev_stop,
+ .dev_close = qede_dev_close,
+ .stats_get = qede_get_stats,
+ .stats_reset = qede_reset_stats,
+ .mac_addr_add = qede_mac_addr_add,
+ .mac_addr_remove = qede_mac_addr_remove,
+};
+
+static void qede_update_pf_params(struct ecore_dev *edev)
+{
+ struct ecore_pf_params pf_params;
+ /* 16 rx + 16 tx */
+ memset(&pf_params, 0, sizeof(struct ecore_pf_params));
+ pf_params.eth_pf_params.num_cons = 32;
+ qed_ops->common->update_pf_params(edev, &pf_params);
+}
+
+static int qede_common_dev_init(struct rte_eth_dev *eth_dev, bool is_vf)
+{
+ struct rte_pci_device *pci_dev;
+ struct rte_pci_addr pci_addr;
+ struct qede_dev *adapter;
+ struct ecore_dev *edev;
+ struct qed_dev_eth_info dev_info;
+ struct qed_slowpath_params params;
+ uint32_t qed_ver;
+ static bool do_once = true;
+ uint8_t bulletin_change;
+ uint8_t vf_mac[ETHER_ADDR_LEN];
+ uint8_t is_mac_forced;
+ bool is_mac_exist;
+ /* Fix up ecore debug level */
+ uint32_t dp_module = ~0 & ~ECORE_MSG_HW;
+ uint8_t dp_level = ECORE_LEVEL_VERBOSE;
+ int rc;
+
+ /* Extract key data structures */
+ adapter = eth_dev->data->dev_private;
+ edev = &adapter->edev;
+ pci_addr = eth_dev->pci_dev->addr;
+
+ PMD_INIT_FUNC_TRACE(edev);
+
+ snprintf(edev->name, NAME_SIZE, PCI_SHORT_PRI_FMT ":dpdk-port-%u",
+ pci_addr.bus, pci_addr.devid, pci_addr.function,
+ eth_dev->data->port_id);
+
+ eth_dev->rx_pkt_burst = qede_recv_pkts;
+ eth_dev->tx_pkt_burst = qede_xmit_pkts;
+
+ if (rte_eal_process_type() != RTE_PROC_PRIMARY) {
+ DP_NOTICE(edev, false,
+ "Skipping device init from secondary process\n");
+ return 0;
+ }
+
+ pci_dev = eth_dev->pci_dev;
+
+ rte_eth_copy_pci_info(eth_dev, pci_dev);
+
+ qed_ver = qed_get_protocol_version(QED_PROTOCOL_ETH);
+
+ if (qed_ver != QEDE_ETH_INTERFACE_VERSION) {
+ DP_ERR(edev, "Version mismatch [%08x != %08x]\n",
+ qed_ver, QEDE_ETH_INTERFACE_VERSION);
+ return -EINVAL;
+ }
+
+ qed_ops = qed_get_eth_ops(QEDE_ETH_INTERFACE_VERSION);
+ if (!qed_ops) {
+ DP_ERR(edev, "Failed to get qed_eth_ops_pass\n");
+ return -EINVAL;
+ }
+
+ DP_INFO(edev, "Starting qede probe\n");
+
+ rc = qed_ops->common->probe(edev, pci_dev, QED_PROTOCOL_ETH,
+ dp_module, dp_level, is_vf);
+
+ if (rc != 0) {
+ DP_ERR(edev, "qede probe failed rc 0x%x\n", rc);
+ return -ENODEV;
+ }
+
+ qede_update_pf_params(edev);
+
+ rte_intr_callback_register(&(eth_dev->pci_dev->intr_handle),
+ qede_interrupt_handler, (void *)eth_dev);
+
+ if (rte_intr_enable(&(eth_dev->pci_dev->intr_handle))) {
+ DP_ERR(edev, "rte_intr_enable() failed\n");
+ return -ENODEV;
+ }
+
+ /* Start the Slowpath-process */
+ memset(¶ms, 0, sizeof(struct qed_slowpath_params));
+ params.int_mode = ECORE_INT_MODE_MSIX;
+ params.drv_major = QEDE_MAJOR_VERSION;
+ params.drv_minor = QEDE_MINOR_VERSION;
+ params.drv_rev = QEDE_REVISION_VERSION;
+ params.drv_eng = QEDE_ENGINEERING_VERSION;
+ strncpy((char *)params.name, "qede LAN", QED_DRV_VER_STR_SIZE);
+
+ rc = qed_ops->common->slowpath_start(edev, ¶ms);
+ if (rc) {
+ DP_ERR(edev, "Cannot start slowpath rc=0x%x\n", rc);
+ return -ENODEV;
+ }
+
+ rc = qed_ops->fill_dev_info(edev, &dev_info);
+ if (rc) {
+ DP_ERR(edev, "Cannot get device_info rc=0x%x\n", rc);
+ qed_ops->common->slowpath_stop(edev);
+ qed_ops->common->remove(edev);
+ return -ENODEV;
+ }
+
+ qede_alloc_etherdev(adapter, &dev_info);
+
+ adapter->ops->common->set_id(edev, edev->name, DRV_MODULE_VERSION);
+
+ /* Allocate memory for storing primary macaddr */
+ eth_dev->data->mac_addrs = rte_zmalloc(edev->name, ETHER_ADDR_LEN,
+ RTE_CACHE_LINE_SIZE);
+
+ if (eth_dev->data->mac_addrs == NULL) {
+ DP_ERR(edev, "Failed to allocate MAC address\n");
+ qed_ops->common->slowpath_stop(edev);
+ qed_ops->common->remove(edev);
+ return -ENOMEM;
+ }
+
+ if (!is_vf)
+ ether_addr_copy((struct ether_addr *)edev->hwfns[0].
+ hw_info.hw_mac_addr,
+ ð_dev->data->mac_addrs[0]);
+ else {
+ ecore_vf_read_bulletin(&edev->hwfns[0], &bulletin_change);
+ if (bulletin_change) {
+ is_mac_exist =
+ ecore_vf_bulletin_get_forced_mac(&edev->hwfns[0],
+ vf_mac,
+ &is_mac_forced);
+ if (is_mac_exist && is_mac_forced) {
+ DP_INFO(edev, "VF macaddr received from PF\n");
+ ether_addr_copy((struct ether_addr *)&vf_mac,
+ ð_dev->data->mac_addrs[0]);
+ } else {
+ DP_NOTICE(edev, false,
+ "No VF macaddr assigned\n");
+ }
+ }
+ }
+
+ eth_dev->dev_ops = &qede_eth_dev_ops;
+
+ if (do_once) {
+ qede_print_adapter_info(adapter);
+ do_once = false;
+ }
+
+ DP_NOTICE(edev, false, "macaddr %02x:%02x:%02x:%02x:%02x:%02x\n",
+ eth_dev->data->mac_addrs[0].addr_bytes[0],
+ eth_dev->data->mac_addrs[0].addr_bytes[1],
+ eth_dev->data->mac_addrs[0].addr_bytes[2],
+ eth_dev->data->mac_addrs[0].addr_bytes[3],
+ eth_dev->data->mac_addrs[0].addr_bytes[4],
+ eth_dev->data->mac_addrs[0].addr_bytes[5]);
+
+ return rc;
+}
+
+static int qedevf_eth_dev_init(struct rte_eth_dev *eth_dev)
+{
+ return qede_common_dev_init(eth_dev, 1);
+}
+
+static int qede_eth_dev_init(struct rte_eth_dev *eth_dev)
+{
+ return qede_common_dev_init(eth_dev, 0);
+}
+
+static int qede_dev_common_uninit(struct rte_eth_dev *eth_dev)
+{
+ /* only uninitialize in the primary process */
+ if (rte_eal_process_type() != RTE_PROC_PRIMARY)
+ return 0;
+
+ /* safe to close dev here */
+ qede_dev_close(eth_dev);
+
+ eth_dev->dev_ops = NULL;
+ eth_dev->rx_pkt_burst = NULL;
+ eth_dev->tx_pkt_burst = NULL;
+
+ if (eth_dev->data->mac_addrs)
+ rte_free(eth_dev->data->mac_addrs);
+
+ eth_dev->data->mac_addrs = NULL;
+
+ return 0;
+}
+
+static int qede_eth_dev_uninit(struct rte_eth_dev *eth_dev)
+{
+ return qede_dev_common_uninit(eth_dev);
+}
+
+static int qedevf_eth_dev_uninit(struct rte_eth_dev *eth_dev)
+{
+ return qede_dev_common_uninit(eth_dev);
+}
+
+static struct rte_pci_id pci_id_qedevf_map[] = {
+#define RTE_PCI_DEV_ID_DECL_QEDEVF(vend, dev) {RTE_PCI_DEVICE(vend, dev)},
+#include "rte_pci_dev_ids.h"
+ {.vendor_id = 0,}
+};
+
+static struct rte_pci_id pci_id_qede_map[] = {
+#define RTE_PCI_DEV_ID_DECL_QEDE(vend, dev) {RTE_PCI_DEVICE(vend, dev)},
+#include "rte_pci_dev_ids.h"
+ {.vendor_id = 0,}
+};
+
+static struct eth_driver rte_qedevf_pmd = {
+ .pci_drv = {
+ .name = "rte_qedevf_pmd",
+ .id_table = pci_id_qedevf_map,
+ .drv_flags =
+ RTE_PCI_DRV_NEED_MAPPING | RTE_PCI_DRV_INTR_LSC,
+ },
+ .eth_dev_init = qedevf_eth_dev_init,
+ .eth_dev_uninit = qedevf_eth_dev_uninit,
+ .dev_private_size = sizeof(struct qede_dev),
+};
+
+static struct eth_driver rte_qede_pmd = {
+ .pci_drv = {
+ .name = "rte_qede_pmd",
+ .id_table = pci_id_qede_map,
+ .drv_flags =
+ RTE_PCI_DRV_NEED_MAPPING | RTE_PCI_DRV_INTR_LSC,
+ },
+ .eth_dev_init = qede_eth_dev_init,
+ .eth_dev_uninit = qede_eth_dev_uninit,
+ .dev_private_size = sizeof(struct qede_dev),
+};
+
+static int
+rte_qedevf_pmd_init(const char *name __rte_unused,
+ const char *params __rte_unused)
+{
+ rte_eth_driver_register(&rte_qedevf_pmd);
+
+ return 0;
+}
+
+static int
+rte_qede_pmd_init(const char *name __rte_unused,
+ const char *params __rte_unused)
+{
+ rte_eth_driver_register(&rte_qede_pmd);
+
+ return 0;
+}
+
+static struct rte_driver rte_qedevf_driver = {
+ .type = PMD_PDEV,
+ .init = rte_qede_pmd_init
+};
+
+static struct rte_driver rte_qede_driver = {
+ .type = PMD_PDEV,
+ .init = rte_qedevf_pmd_init
+};
+
+PMD_REGISTER_DRIVER(rte_qede_driver);
+PMD_REGISTER_DRIVER(rte_qedevf_driver);
diff --git a/drivers/net/qede/qede_ethdev.h b/drivers/net/qede/qede_ethdev.h
new file mode 100644
index 0000000..7cddb5e
--- /dev/null
+++ b/drivers/net/qede/qede_ethdev.h
@@ -0,0 +1,138 @@
+/*
+ * Copyright (c) 2016 QLogic Corporation.
+ * All rights reserved.
+ * www.qlogic.com
+ *
+ * See LICENSE.qede_pmd for copyright and licensing details.
+ */
+
+
+#ifndef _QEDE_ETHDEV_H_
+#define _QEDE_ETHDEV_H_
+
+#include <rte_ether.h>
+#include <rte_ethdev.h>
+#include <rte_dev.h>
+
+/* ecore includes */
+#include "ecore/bcm_osal.h"
+#include "ecore/ecore.h"
+#include "ecore/ecore_dev_api.h"
+#include "ecore/ecore_l2_api.h"
+#include "ecore/ecore_vf_api.h"
+#include "ecore/ecore_hsi_common.h"
+#include "ecore/ecore_int_api.h"
+#include "ecore/ecore_chain.h"
+#include "ecore/ecore_status.h"
+#include "ecore/ecore_hsi_eth.h"
+#include "ecore/ecore_dev_api.h"
+#include "ecore/ecore_iov_api.h"
+
+#include "qede_logs.h"
+#include "qede_if.h"
+#include "qede_eth_if.h"
+
+#include "qede_rxtx.h"
+
+#define qede_stringify1(x...) #x
+#define qede_stringify(x...) qede_stringify1(x)
+
+/* Driver versions */
+#define QEDE_PMD_VER_PREFIX "QEDE PMD"
+#define QEDE_PMD_VERSION_MAJOR 1
+#define QEDE_PMD_VERSION_MINOR 0
+#define QEDE_PMD_VERSION_PATCH 0
+
+#define QEDE_MAJOR_VERSION 8
+#define QEDE_MINOR_VERSION 7
+#define QEDE_REVISION_VERSION 9
+#define QEDE_ENGINEERING_VERSION 0
+
+#define DRV_MODULE_VERSION qede_stringify(QEDE_MAJOR_VERSION) "." \
+ qede_stringify(QEDE_MINOR_VERSION) "." \
+ qede_stringify(QEDE_REVISION_VERSION) "." \
+ qede_stringify(QEDE_ENGINEERING_VERSION)
+
+#define QEDE_ETH_INTERFACE_VERSION 609
+
+#define QEDE_RSS_INDIR_INITED (1 << 0)
+#define QEDE_RSS_KEY_INITED (1 << 1)
+#define QEDE_RSS_CAPS_INITED (1 << 2)
+
+#define QEDE_MAX_RSS_CNT(edev) ((edev)->dev_info.num_queues)
+#define QEDE_MAX_TSS_CNT(edev) ((edev)->dev_info.num_queues * \
+ (edev)->dev_info.num_tc)
+
+#define QEDE_RSS_CNT(edev) ((edev)->num_rss)
+#define QEDE_TSS_CNT(edev) ((edev)->num_rss * (edev)->num_tc)
+
+#define QEDE_DUPLEX_FULL 1
+#define QEDE_DUPLEX_HALF 2
+#define QEDE_DUPLEX_UNKNOWN 0xff
+
+#define QEDE_INIT_QDEV(eth_dev) (eth_dev->data->dev_private)
+
+#define QEDE_INIT_EDEV(adapter) (&((struct qede_dev *)adapter)->edev)
+
+#define QEDE_INIT(eth_dev) { \
+ struct qede_dev *qdev = eth_dev->data->dev_private; \
+ struct ecore_dev *edev = &qdev->edev; \
+}
+
+extern const char *QEDE_FW_FILE_NAME;
+
+/* Port/function states */
+enum dev_state {
+ QEDE_START,
+ QEDE_STOP,
+ QEDE_CLOSE
+};
+
+struct qed_int_param {
+ uint32_t int_mode;
+ uint8_t num_vectors;
+ uint8_t min_msix_cnt;
+};
+
+struct qed_int_params {
+ struct qed_int_param in;
+ struct qed_int_param out;
+ bool fp_initialized;
+};
+
+/*
+ * Structure to store private data for each port.
+ */
+struct qede_dev {
+ struct ecore_dev edev;
+ uint8_t protocol;
+ const struct qed_eth_ops *ops;
+ struct qed_dev_eth_info dev_info;
+ struct ecore_sb_info *sb_array;
+ struct qede_fastpath *fp_array;
+ uint16_t num_rss;
+ uint8_t num_tc;
+ uint16_t mtu;
+ uint32_t rss_params_inited;
+ struct qed_update_vport_rss_params rss_params;
+ uint32_t flags;
+ bool gro_disable;
+ struct qede_rx_queue **rx_queues;
+ struct qede_tx_queue **tx_queues;
+ enum dev_state state;
+
+ /* Vlans */
+ osal_list_t vlan_list;
+ uint16_t configured_vlans;
+ uint16_t non_configured_vlans;
+ bool accept_any_vlan;
+ uint16_t vxlan_dst_port;
+
+ bool handle_hw_err;
+ char drv_ver[QED_DRV_VER_STR_SIZE];
+};
+
+int qede_dev_set_link_state(struct rte_eth_dev *eth_dev, bool link_up);
+void qede_config_rx_mode(struct rte_eth_dev *eth_dev);
+
+#endif /* _QEDE_ETHDEV_H_ */
diff --git a/drivers/net/qede/qede_if.h b/drivers/net/qede/qede_if.h
new file mode 100644
index 0000000..72969eb
--- /dev/null
+++ b/drivers/net/qede/qede_if.h
@@ -0,0 +1,167 @@
+/*
+ * Copyright (c) 2016 QLogic Corporation.
+ * All rights reserved.
+ * www.qlogic.com
+ *
+ * See LICENSE.qede_pmd for copyright and licensing details.
+ */
+
+#ifndef _QEDE_IF_H
+#define _QEDE_IF_H
+
+#include "qede_ethdev.h"
+
+/* forward */
+struct ecore_dev;
+struct qed_sb_info;
+struct qed_pf_params;
+enum ecore_int_mode;
+
+struct qed_dev_info {
+ uint8_t num_hwfns;
+ uint8_t hw_mac[ETHER_ADDR_LEN];
+ bool is_mf_default;
+
+ /* FW version */
+ uint16_t fw_major;
+ uint16_t fw_minor;
+ uint16_t fw_rev;
+ uint16_t fw_eng;
+
+ /* MFW version */
+ uint32_t mfw_rev;
+
+ bool roce_supported;
+
+ uint32_t flash_size;
+ uint8_t mf_mode;
+ bool tx_switching;
+ /* To be added... */
+};
+
+enum qed_sb_type {
+ QED_SB_TYPE_L2_QUEUE,
+ QED_SB_TYPE_STORAGE,
+ QED_SB_TYPE_CNQ,
+};
+
+enum qed_protocol {
+ QED_PROTOCOL_ETH,
+ QED_PROTOCOL_ISCSI,
+};
+
+struct qed_link_params {
+ bool link_up;
+
+#define QED_LINK_OVERRIDE_SPEED_AUTONEG (1 << 0)
+#define QED_LINK_OVERRIDE_SPEED_ADV_SPEEDS (1 << 1)
+#define QED_LINK_OVERRIDE_SPEED_FORCED_SPEED (1 << 2)
+#define QED_LINK_OVERRIDE_PAUSE_CONFIG (1 << 3)
+ uint32_t override_flags;
+ bool autoneg;
+ uint32_t adv_speeds;
+ uint32_t forced_speed;
+#define QED_LINK_PAUSE_AUTONEG_ENABLE (1 << 0)
+#define QED_LINK_PAUSE_RX_ENABLE (1 << 1)
+#define QED_LINK_PAUSE_TX_ENABLE (1 << 2)
+ uint32_t pause_config;
+};
+
+struct qed_link_output {
+ bool link_up;
+ uint32_t supported_caps; /* In SUPPORTED defs */
+ uint32_t advertised_caps; /* In ADVERTISED defs */
+ uint32_t lp_caps; /* In ADVERTISED defs */
+ uint32_t speed; /* In Mb/s */
+ uint8_t duplex; /* In DUPLEX defs */
+ uint8_t port; /* In PORT defs */
+ bool autoneg;
+ uint32_t pause_config;
+};
+
+#define QED_DRV_VER_STR_SIZE 80
+struct qed_slowpath_params {
+ uint32_t int_mode;
+ uint8_t drv_major;
+ uint8_t drv_minor;
+ uint8_t drv_rev;
+ uint8_t drv_eng;
+ uint8_t name[QED_DRV_VER_STR_SIZE];
+};
+
+#define ILT_PAGE_SIZE_TCFC 0x8000 /* 32KB */
+
+struct qed_common_cb_ops {
+ void (*link_update) (void *dev, struct qed_link_output *link);
+};
+
+struct qed_selftest_ops {
+/**
+ * @brief registers - Perform register tests
+ *
+ * @param edev
+ *
+ * @return 0 on success, error otherwise.
+ */
+ int (*registers) (struct ecore_dev *edev);
+};
+
+struct qed_common_ops {
+ int (*probe) (struct ecore_dev *edev,
+ struct rte_pci_device *pci_dev,
+ enum qed_protocol protocol,
+ uint32_t dp_module, uint8_t dp_level, bool is_vf);
+ void (*set_id) (struct ecore_dev *edev,
+ char name[], const char ver_str[]);
+ enum _ecore_status_t (*chain_alloc) (struct ecore_dev *edev,
+ enum ecore_chain_use_mode
+ intended_use,
+ enum ecore_chain_mode mode,
+ enum ecore_chain_cnt_type cnt_type,
+ uint32_t num_elems,
+ osal_size_t elem_size,
+ struct ecore_chain *p_chain);
+
+ void (*chain_free) (struct ecore_dev *edev,
+ struct ecore_chain *p_chain);
+
+ void (*get_link) (struct ecore_dev *edev,
+ struct qed_link_output *if_link);
+ int (*set_link) (struct ecore_dev *edev,
+ struct qed_link_params *params);
+
+ int (*drain) (struct ecore_dev *edev);
+
+ void (*remove) (struct ecore_dev *edev);
+
+ int (*slowpath_stop) (struct ecore_dev *edev);
+
+ void (*update_pf_params) (struct ecore_dev *edev,
+ struct ecore_pf_params *params);
+
+ int (*slowpath_start) (struct ecore_dev *edev,
+ struct qed_slowpath_params *params);
+
+ int (*set_fp_int) (struct ecore_dev *edev, uint16_t cnt);
+
+ uint32_t(*sb_init) (struct ecore_dev *edev,
+ struct ecore_sb_info *sb_info,
+ void *sb_virt_addr,
+ dma_addr_t sb_phy_addr,
+ uint16_t sb_id, enum qed_sb_type type);
+
+ bool(*can_link_change) (struct ecore_dev *edev);
+ void (*update_msglvl) (struct ecore_dev *edev,
+ uint32_t dp_module, uint8_t dp_level);
+};
+
+/**
+ * @brief qed_get_protocol_version
+ *
+ * @param protocol
+ *
+ * @return version supported by qed for given protocol driver
+ */
+uint32_t qed_get_protocol_version(enum qed_protocol protocol);
+
+#endif /* _QEDE_IF_H */
diff --git a/drivers/net/qede/qede_logs.h b/drivers/net/qede/qede_logs.h
new file mode 100644
index 0000000..ee51004
--- /dev/null
+++ b/drivers/net/qede/qede_logs.h
@@ -0,0 +1,93 @@
+/*
+ * Copyright (c) 2016 QLogic Corporation.
+ * All rights reserved.
+ * www.qlogic.com
+ *
+ * See LICENSE.qede_pmd for copyright and licensing details.
+ */
+
+#ifndef _QEDE_LOGS_H_
+#define _QEDE_LOGS_H_
+
+#define DP_ERR(p_dev, fmt, ...) \
+ rte_log(RTE_LOG_ERR, RTE_LOGTYPE_PMD, \
+ "[%s:%d(%s)]" fmt, \
+ __func__, __LINE__, \
+ (p_dev)->name ? (p_dev)->name : "", \
+ ##__VA_ARGS__);
+
+#define DP_NOTICE(p_dev, is_assert, fmt, ...) \
+do { \
+ rte_log(RTE_LOG_NOTICE, RTE_LOGTYPE_PMD,\
+ "[QEDE PMD: (%s)]%s:" fmt, \
+ (p_dev)->name ? (p_dev)->name : "", \
+ __func__, \
+ ##__VA_ARGS__); \
+ OSAL_ASSERT(!is_assert); \
+} while (0)
+
+#ifdef RTE_LIBRTE_QEDE_DEBUG_INFO
+
+#define DP_INFO(p_dev, fmt, ...) \
+ rte_log(RTE_LOG_INFO, RTE_LOGTYPE_PMD, \
+ "[%s:%d(%s)]" fmt, \
+ __func__, __LINE__, \
+ (p_dev)->name ? (p_dev)->name : "", \
+ ##__VA_ARGS__)
+#else
+#define DP_INFO(p_dev, fmt, ...) do { } while (0)
+
+#endif
+
+#ifdef RTE_LIBRTE_QEDE_DEBUG_ECORE
+#define DP_VERBOSE(p_dev, module, fmt, ...) \
+do { \
+ if ((p_dev)->dp_module & module) \
+ rte_log(RTE_LOG_DEBUG, RTE_LOGTYPE_PMD, \
+ "[%s:%d(%s)]" fmt, \
+ __func__, __LINE__, \
+ (p_dev)->name ? (p_dev)->name : "", \
+ ##__VA_ARGS__); \
+} while (0)
+#else
+#define DP_VERBOSE(p_dev, fmt, ...) do { } while (0)
+#endif
+
+#define PMD_INIT_LOG(level, edev, fmt, args...) \
+ rte_log(RTE_LOG_ ## level, RTE_LOGTYPE_PMD, \
+ "[qede_pmd: %s] %s() " fmt "\n", \
+ (edev)->name, __func__, ##args)
+
+#ifdef RTE_LIBRTE_QEDE_DEBUG_INIT
+#define PMD_INIT_FUNC_TRACE(edev) PMD_INIT_LOG(DEBUG, edev, " >>")
+#else
+#define PMD_INIT_FUNC_TRACE(edev) do { } while (0)
+#endif
+
+#ifdef RTE_LIBRTE_QEDE_DEBUG_TX
+#define PMD_TX_LOG(level, q, fmt, args...) \
+ RTE_LOG(level, PMD, "%s(): port=%u queue=%u " fmt "\n", \
+ __func__, q->port_id, q->queue_id, ## args)
+#else
+#define PMD_TX_LOG(level, fmt, args...) do { } while (0)
+#endif
+
+#ifdef RTE_LIBRTE_QEDE_DEBUG_RX
+#define PMD_RX_LOG(level, q, fmt, args...) \
+ RTE_LOG(level, PMD, "%s(): port=%u queue=%u " fmt "\n", \
+ __func__, q->port_id, q->queue_id, ## args);
+#else
+#define PMD_RX_LOG(level, q, fmt, args...) do { } while (0)
+#endif
+
+#ifdef RTE_LIBRTE_QEDE_DEBUG_DRIVER
+#define PMD_DRV_LOG_RAW(level, fmt, args...) \
+ RTE_LOG(level, PMD, "%s(): " fmt, __func__, ## args)
+#else
+#define PMD_DRV_LOG_RAW(level, fmt, args...) do { } while (0)
+#endif
+
+#define PMD_DRV_LOG(level, fmt, args...) \
+ PMD_DRV_LOG_RAW(level, fmt "\n", ## args)
+
+#endif /* _QEDE_LOGS_H_ */
diff --git a/drivers/net/qede/qede_main.c b/drivers/net/qede/qede_main.c
new file mode 100644
index 0000000..9f3a236
--- /dev/null
+++ b/drivers/net/qede/qede_main.c
@@ -0,0 +1,587 @@
+/*
+ * Copyright (c) 2016 QLogic Corporation.
+ * All rights reserved.
+ * www.qlogic.com
+ *
+ * See LICENSE.qede_pmd for copyright and licensing details.
+ */
+
+#include <sys/stat.h>
+#include <fcntl.h>
+#include <unistd.h>
+#include <zlib.h>
+
+#include "qede_ethdev.h"
+
+
+static uint8_t npar_tx_switching = 1;
+
+#define CONFIG_QED_BINARY_FW
+
+#ifdef RTE_LIBRTE_QEDE_TX_SWITCHING
+static uint8_t tx_switching = 1;
+#else
+static uint8_t tx_switching;
+#endif
+
+#ifndef RTE_LIBRTE_QEDE_FW
+const char *QEDE_FW_FILE_NAME =
+ "/lib/firmware/qed/qed_init_values_zipped-8.7.7.0.bin";
+#else
+const char *QEDE_FW_FILE_NAME = RTE_LIBRTE_QEDE_FW;
+#endif
+
+static void
+qed_update_pf_params(struct ecore_dev *edev, struct ecore_pf_params *params)
+{
+ int i;
+
+ for (i = 0; i < edev->num_hwfns; i++) {
+ struct ecore_hwfn *p_hwfn = &edev->hwfns[i];
+ p_hwfn->pf_params = *params;
+ }
+}
+
+static void qed_init_pci(struct ecore_dev *edev, struct rte_pci_device *pci_dev)
+{
+ edev->regview = pci_dev->mem_resource[0].addr;
+ edev->doorbells = pci_dev->mem_resource[2].addr;
+}
+
+static int
+qed_probe(struct ecore_dev *edev, struct rte_pci_device *pci_dev,
+ enum qed_protocol protocol, uint32_t dp_module,
+ uint8_t dp_level, bool is_vf)
+{
+ struct qede_dev *qdev = (struct qede_dev *)edev;
+ int rc;
+
+ ecore_init_struct(edev);
+ qdev->protocol = protocol;
+ if (is_vf) {
+ edev->b_is_vf = true;
+ edev->sriov_info.b_hw_channel = true;
+ }
+ ecore_init_dp(edev, dp_module, dp_level, NULL);
+ qed_init_pci(edev, pci_dev);
+ rc = ecore_hw_prepare(edev, ECORE_PCI_DEFAULT);
+ if (rc) {
+ DP_ERR(edev, "hw prepare failed\n");
+ return rc;
+ }
+
+ return rc;
+}
+
+static int qed_nic_setup(struct ecore_dev *edev)
+{
+ int rc, i;
+
+ /* Save the PF params - @@@TBD this will work only for L2 */
+ for (i = 0; i < edev->num_hwfns; i++) {
+ struct ecore_hwfn *p_hwfn = &edev->hwfns[i];
+ p_hwfn->using_ll2 = false;
+ }
+
+ rc = ecore_resc_alloc(edev);
+ if (rc)
+ return rc;
+
+ DP_INFO(edev, "Allocated qed resources\n");
+ ecore_resc_setup(edev);
+
+ return rc;
+}
+
+static int qed_alloc_stream_mem(struct ecore_dev *edev)
+{
+ int i;
+
+ for_each_hwfn(edev, i) {
+ struct ecore_hwfn *p_hwfn = &edev->hwfns[i];
+
+ p_hwfn->stream = OSAL_ZALLOC(p_hwfn->p_dev, GFP_KERNEL,
+ sizeof(*p_hwfn->stream));
+ if (!p_hwfn->stream)
+ return -ENOMEM;
+ }
+
+ return 0;
+}
+
+static void qed_free_stream_mem(struct ecore_dev *edev)
+{
+ int i;
+
+ for_each_hwfn(edev, i) {
+ struct ecore_hwfn *p_hwfn = &edev->hwfns[i];
+
+ if (!p_hwfn->stream)
+ return;
+
+ OSAL_FREE(p_hwfn->p_dev, p_hwfn->stream);
+ }
+}
+
+static int qed_load_firmware_data(struct ecore_dev *edev)
+{
+ int fd;
+ struct stat st;
+
+ fd = open(QEDE_FW_FILE_NAME, O_RDONLY);
+ if (fd < 0) {
+ DP_NOTICE(edev, false, "Can't open firmware file\n");
+ return -ENOENT;
+ }
+
+ if (fstat(fd, &st) < 0) {
+ DP_NOTICE(edev, false, "Can't stat firmware file\n");
+ return -1;
+ }
+
+ edev->firmware = rte_zmalloc("qede_fw", st.st_size,
+ RTE_CACHE_LINE_SIZE);
+ if (!edev->firmware) {
+ DP_NOTICE(edev, false, "Can't allocate memory for firmware\n");
+ close(fd);
+ return -ENOMEM;
+ }
+
+ if (read(fd, edev->firmware, st.st_size) != st.st_size) {
+ DP_NOTICE(edev, false, "Can't read firmware data\n");
+ close(fd);
+ return -1;
+ }
+
+ edev->fw_len = st.st_size;
+ if (edev->fw_len < 104) {
+ DP_NOTICE(edev, false, "Invalid fw size: %" PRIu64"\n",
+ edev->fw_len);
+ return -EINVAL;
+ }
+
+ return 0;
+}
+
+static int qed_slowpath_start(struct ecore_dev *edev,
+ struct qed_slowpath_params *params)
+{
+ bool allow_npar_tx_switching;
+ const uint8_t *data = NULL;
+ struct ecore_hwfn *hwfn;
+ struct ecore_mcp_drv_version drv_version;
+ struct qede_dev *qdev = (struct qede_dev *)edev;
+ int rc;
+#ifdef QED_ENC_SUPPORTED
+ struct ecore_tunn_start_params tunn_info;
+#endif
+
+#ifdef CONFIG_QED_BINARY_FW
+ if (IS_PF(edev)) {
+ rc = qed_load_firmware_data(edev);
+ if (rc) {
+ DP_NOTICE(edev, true,
+ "Failed to find fw file %s\n",
+ QEDE_FW_FILE_NAME);
+ goto err;
+ }
+ }
+#endif
+
+ rc = qed_nic_setup(edev);
+ if (rc)
+ goto err;
+
+ /* set int_coalescing_mode */
+ edev->int_coalescing_mode = ECORE_COAL_MODE_ENABLE;
+
+ /* Should go with CONFIG_QED_BINARY_FW */
+ if (IS_PF(edev)) {
+ /* Allocate stream for unzipping */
+ rc = qed_alloc_stream_mem(edev);
+ if (rc) {
+ DP_NOTICE(edev, true,
+ "Failed to allocate stream memory\n");
+ goto err2;
+ }
+ }
+
+ /* Start the slowpath */
+#ifdef CONFIG_QED_BINARY_FW
+ if (IS_PF(edev))
+ data = edev->firmware;
+#endif
+ allow_npar_tx_switching = npar_tx_switching ? true : false;
+
+#ifdef QED_ENC_SUPPORTED
+ memset(&tunn_info, 0, sizeof(tunn_info));
+ tunn_info.tunn_mode |= 1 << QED_MODE_VXLAN_TUNN |
+ 1 << QED_MODE_L2GRE_TUNN |
+ 1 << QED_MODE_IPGRE_TUNN |
+ 1 << QED_MODE_L2GENEVE_TUNN | 1 << QED_MODE_IPGENEVE_TUNN;
+ tunn_info.tunn_clss_vxlan = QED_TUNN_CLSS_MAC_VLAN;
+ tunn_info.tunn_clss_l2gre = QED_TUNN_CLSS_MAC_VLAN;
+ tunn_info.tunn_clss_ipgre = QED_TUNN_CLSS_MAC_VLAN;
+ rc = ecore_hw_init(edev, &tunn_info, true, ECORE_INT_MODE_MSIX,
+ allow_npar_tx_switching, data);
+#else
+ rc = ecore_hw_init(edev, NULL, true, ECORE_INT_MODE_MSIX,
+ allow_npar_tx_switching, data);
+#endif
+ if (rc) {
+ DP_ERR(edev, "ecore_hw_init failed\n");
+ goto err2;
+ }
+
+ DP_INFO(edev, "HW inited and function started\n");
+
+#ifdef CONFIG_ECORE_LL2
+ /* Allocate LL2 interface if needed */
+ if (edev->hwfns[0].using_ll2) {
+ rc = qed_ll2_alloc_if(edev);
+ if (rc)
+ goto err3;
+ }
+#endif
+
+ if (IS_PF(edev)) {
+ hwfn = ECORE_LEADING_HWFN(edev);
+ drv_version.version = (params->drv_major << 24) |
+ (params->drv_minor << 16) |
+ (params->drv_rev << 8) | (params->drv_eng);
+ /* TBD: strlcpy() */
+ strncpy((char *)drv_version.name, (const char *)params->name,
+ MCP_DRV_VER_STR_SIZE - 4);
+ rc = ecore_mcp_send_drv_version(hwfn, hwfn->p_main_ptt,
+ &drv_version);
+ if (rc) {
+ DP_NOTICE(edev, true,
+ "Failed sending drv version command\n");
+ return rc;
+ }
+ }
+
+ ecore_reset_vport_stats(edev);
+
+ return 0;
+
+ ecore_hw_stop(edev);
+err2:
+ ecore_resc_free(edev);
+err:
+#ifdef CONFIG_QED_BINARY_FW
+ if (IS_PF(edev)) {
+ if (edev->firmware)
+ rte_free(edev->firmware);
+ edev->firmware = NULL;
+ }
+#endif
+ return rc;
+}
+
+static int
+qed_fill_dev_info(struct ecore_dev *edev, struct qed_dev_info *dev_info)
+{
+ struct ecore_ptt *ptt = NULL;
+
+ memset(dev_info, 0, sizeof(struct qed_dev_info));
+ dev_info->num_hwfns = edev->num_hwfns;
+ dev_info->is_mf_default = IS_MF_DEFAULT(&edev->hwfns[0]);
+ dev_info->roce_supported = (edev->hwfns[0].hw_info.personality
+ == ECORE_PCI_ETH_ROCE);
+ rte_memcpy(&dev_info->hw_mac, &edev->hwfns[0].hw_info.hw_mac_addr,
+ ETHER_ADDR_LEN);
+
+ if (IS_PF(edev)) {
+ dev_info->fw_major = FW_MAJOR_VERSION;
+ dev_info->fw_minor = FW_MINOR_VERSION;
+ dev_info->fw_rev = FW_REVISION_VERSION;
+ dev_info->fw_eng = FW_ENGINEERING_VERSION;
+ dev_info->mf_mode = edev->mf_mode;
+ dev_info->tx_switching = tx_switching ? true : false;
+ } else {
+ ecore_vf_get_fw_version(&edev->hwfns[0], &dev_info->fw_major,
+ &dev_info->fw_minor, &dev_info->fw_rev,
+ &dev_info->fw_eng);
+ }
+
+ if (IS_PF(edev)) {
+ ptt = ecore_ptt_acquire(ECORE_LEADING_HWFN(edev));
+ if (ptt) {
+ ecore_mcp_get_mfw_ver(edev, ptt,
+ &dev_info->mfw_rev, NULL);
+
+ ecore_mcp_get_flash_size(ECORE_LEADING_HWFN(edev), ptt,
+ &dev_info->flash_size);
+
+ /* Workaround to allow PHY-read commands for
+ * B0 bringup.
+ */
+ if (ECORE_IS_BB_B0(edev))
+ dev_info->flash_size = 0xffffffff;
+
+ ecore_ptt_release(ECORE_LEADING_HWFN(edev), ptt);
+ }
+ } else {
+ ecore_mcp_get_mfw_ver(edev, ptt, &dev_info->mfw_rev, NULL);
+ }
+
+ return 0;
+}
+
+int
+qed_fill_eth_dev_info(struct ecore_dev *edev, struct qed_dev_eth_info *info)
+{
+ struct qede_dev *qdev = (struct qede_dev *)edev;
+ int i;
+
+ memset(info, 0, sizeof(*info));
+
+ info->num_tc = 1 /* @@@TBD aelior MULTI_COS */;
+
+ if (IS_PF(edev)) {
+ info->num_queues = 0;
+ for_each_hwfn(edev, i)
+ info->num_queues +=
+ FEAT_NUM(&edev->hwfns[i], ECORE_PF_L2_QUE);
+
+ info->num_vlan_filters = RESC_NUM(&edev->hwfns[0], ECORE_VLAN);
+
+ rte_memcpy(&info->port_mac, &edev->hwfns[0].hw_info.hw_mac_addr,
+ ETHER_ADDR_LEN);
+ } else {
+ ecore_vf_get_num_rxqs(&edev->hwfns[0], &info->num_queues);
+
+ ecore_vf_get_num_vlan_filters(&edev->hwfns[0],
+ &info->num_vlan_filters);
+
+ ecore_vf_get_port_mac(&edev->hwfns[0],
+ (uint8_t *) &info->port_mac);
+ }
+
+ qed_fill_dev_info(edev, &info->common);
+
+ if (IS_VF(edev))
+ memset(&info->common.hw_mac, 0, ETHER_ADDR_LEN);
+
+ return 0;
+}
+
+static void
+qed_set_id(struct ecore_dev *edev, char name[NAME_SIZE],
+ const char ver_str[VER_SIZE])
+{
+ int i;
+
+ rte_memcpy(edev->name, name, NAME_SIZE);
+ for_each_hwfn(edev, i) {
+ snprintf(edev->hwfns[i].name, NAME_SIZE, "%s-%d", name, i);
+ }
+ rte_memcpy(edev->ver_str, ver_str, VER_SIZE);
+ edev->drv_type = DRV_ID_DRV_TYPE_LINUX;
+}
+
+static uint32_t
+qed_sb_init(struct ecore_dev *edev, struct ecore_sb_info *sb_info,
+ void *sb_virt_addr, dma_addr_t sb_phy_addr,
+ uint16_t sb_id, enum qed_sb_type type)
+{
+ struct ecore_hwfn *p_hwfn;
+ int hwfn_index;
+ uint16_t rel_sb_id;
+ uint8_t n_hwfns;
+ uint32_t rc;
+
+ /* RoCE uses single engine and CMT uses two engines. When using both
+ * we force only a single engine. Storage uses only engine 0 too.
+ */
+ if (type == QED_SB_TYPE_L2_QUEUE)
+ n_hwfns = edev->num_hwfns;
+ else
+ n_hwfns = 1;
+
+ hwfn_index = sb_id % n_hwfns;
+ p_hwfn = &edev->hwfns[hwfn_index];
+ rel_sb_id = sb_id / n_hwfns;
+
+ DP_INFO(edev, "hwfn [%d] <--[init]-- SB %04x [0x%04x upper]\n",
+ hwfn_index, rel_sb_id, sb_id);
+
+ rc = ecore_int_sb_init(p_hwfn, p_hwfn->p_main_ptt, sb_info,
+ sb_virt_addr, sb_phy_addr, rel_sb_id);
+
+ return rc;
+}
+
+static void qed_fill_link(struct ecore_hwfn *hwfn,
+ struct qed_link_output *if_link)
+{
+ struct ecore_mcp_link_params params;
+ struct ecore_mcp_link_state link;
+ struct ecore_mcp_link_capabilities link_caps;
+ uint32_t media_type;
+ uint8_t change = 0;
+
+ memset(if_link, 0, sizeof(*if_link));
+
+ /* Prepare source inputs */
+ if (IS_PF(hwfn->p_dev)) {
+ rte_memcpy(¶ms, ecore_mcp_get_link_params(hwfn),
+ sizeof(params));
+ rte_memcpy(&link, ecore_mcp_get_link_state(hwfn), sizeof(link));
+ rte_memcpy(&link_caps, ecore_mcp_get_link_capabilities(hwfn),
+ sizeof(link_caps));
+ } else {
+ ecore_vf_read_bulletin(hwfn, &change);
+ ecore_vf_get_link_params(hwfn, ¶ms);
+ ecore_vf_get_link_state(hwfn, &link);
+ ecore_vf_get_link_caps(hwfn, &link_caps);
+ }
+
+ /* Set the link parameters to pass to protocol driver */
+ if (link.link_up)
+ if_link->link_up = true;
+
+ if (link.link_up)
+ if_link->speed = link.speed;
+
+ if_link->duplex = QEDE_DUPLEX_FULL;
+}
+
+static void
+qed_get_current_link(struct ecore_dev *edev, struct qed_link_output *if_link)
+{
+ qed_fill_link(&edev->hwfns[0], if_link);
+
+#ifdef CONFIG_QED_SRIOV
+ for_each_hwfn(cdev, i)
+ qed_inform_vf_link_state(&cdev->hwfns[i]);
+#endif
+
+}
+
+static int qed_set_link(struct ecore_dev *edev, struct qed_link_params *params)
+{
+ struct ecore_hwfn *hwfn;
+ struct ecore_ptt *ptt;
+ int rc;
+
+ if (IS_VF(edev))
+ return 0;
+
+ /* The link should be set only once per PF */
+ hwfn = &edev->hwfns[0];
+
+ ptt = ecore_ptt_acquire(hwfn);
+ if (!ptt)
+ return -EBUSY;
+
+ rc = ecore_mcp_set_link(hwfn, ptt, params->link_up);
+
+ ecore_ptt_release(hwfn, ptt);
+
+ return rc;
+}
+
+static int qed_drain(struct ecore_dev *edev)
+{
+ struct ecore_hwfn *hwfn;
+ struct ecore_ptt *ptt;
+ int i, rc;
+
+ if (IS_VF(edev))
+ return 0;
+
+ for_each_hwfn(edev, i) {
+ hwfn = &edev->hwfns[i];
+ ptt = ecore_ptt_acquire(hwfn);
+ if (!ptt) {
+ DP_NOTICE(hwfn, true, "Failed to drain NIG; No PTT\n");
+ return -EBUSY;
+ }
+ rc = ecore_mcp_drain(hwfn, ptt);
+ if (rc)
+ return rc;
+ ecore_ptt_release(hwfn, ptt);
+ }
+
+ return 0;
+}
+
+static int qed_nic_stop(struct ecore_dev *edev)
+{
+ int i, rc;
+
+ rc = ecore_hw_stop(edev);
+ for (i = 0; i < edev->num_hwfns; i++) {
+ struct ecore_hwfn *p_hwfn = &edev->hwfns[i];
+
+ if (p_hwfn->b_sp_dpc_enabled)
+ p_hwfn->b_sp_dpc_enabled = false;
+ }
+ return rc;
+}
+
+static int qed_nic_reset(struct ecore_dev *edev)
+{
+ int rc;
+
+ rc = ecore_hw_reset(edev);
+ if (rc)
+ return rc;
+
+ ecore_resc_free(edev);
+
+ return 0;
+}
+
+static int qed_slowpath_stop(struct ecore_dev *edev)
+{
+#ifdef CONFIG_QED_SRIOV
+ int i;
+#endif
+
+ if (!edev)
+ return -ENODEV;
+/* Should be under CONFIG_ECORE_LL2 */
+#ifdef CONFIG_ECORE_LL2
+ qed_ll2_dealloc_if(edev);
+#endif
+
+ if (IS_PF(edev)) {
+ qed_free_stream_mem(edev);
+
+#ifdef CONFIG_QED_SRIOV
+ if (IS_QED_ETH_IF(edev))
+ qed_sriov_disable(edev, true);
+#endif
+ qed_nic_stop(edev);
+ }
+
+ qed_nic_reset(edev);
+
+ return 0;
+}
+
+static void qed_remove(struct ecore_dev *edev)
+{
+ if (!edev)
+ return;
+
+ ecore_hw_remove(edev);
+}
+
+const struct qed_common_ops qed_common_ops_pass = {
+ INIT_STRUCT_FIELD(probe, &qed_probe),
+ INIT_STRUCT_FIELD(update_pf_params, &qed_update_pf_params),
+ INIT_STRUCT_FIELD(slowpath_start, &qed_slowpath_start),
+ INIT_STRUCT_FIELD(set_id, &qed_set_id),
+ INIT_STRUCT_FIELD(chain_alloc, &ecore_chain_alloc),
+ INIT_STRUCT_FIELD(chain_free, &ecore_chain_free),
+ INIT_STRUCT_FIELD(sb_init, &qed_sb_init),
+ INIT_STRUCT_FIELD(get_link, &qed_get_current_link),
+ INIT_STRUCT_FIELD(set_link, &qed_set_link),
+ INIT_STRUCT_FIELD(drain, &qed_drain),
+ INIT_STRUCT_FIELD(slowpath_stop, &qed_slowpath_stop),
+ INIT_STRUCT_FIELD(remove, &qed_remove),
+};
diff --git a/drivers/net/qede/qede_rxtx.c b/drivers/net/qede/qede_rxtx.c
new file mode 100644
index 0000000..cb85331
--- /dev/null
+++ b/drivers/net/qede/qede_rxtx.c
@@ -0,0 +1,1346 @@
+/*
+ * Copyright (c) 2016 QLogic Corporation.
+ * All rights reserved.
+ * www.qlogic.com
+ *
+ * See LICENSE.qede_pmd for copyright and licensing details.
+ */
+
+#include "qede_rxtx.h"
+
+static bool gro_disable = 1; /* mod_param */
+
+static inline struct
+rte_mbuf *qede_rxmbuf_alloc(struct rte_mempool *mp)
+{
+ struct rte_mbuf *m;
+
+ m = __rte_mbuf_raw_alloc(mp);
+ __rte_mbuf_sanity_check(m, 0);
+
+ return m;
+}
+
+static inline int qede_alloc_rx_buffer(struct qede_rx_queue *rxq)
+{
+ struct rte_mbuf *new_mb = NULL;
+ struct eth_rx_bd *rx_bd;
+ dma_addr_t mapping;
+ uint16_t idx = rxq->sw_rx_prod & NUM_RX_BDS(rxq);
+
+ new_mb = qede_rxmbuf_alloc(rxq->mb_pool);
+ if (unlikely(!new_mb)) {
+ PMD_RX_LOG(ERR, rxq,
+ "Failed to allocate rx buffer"
+ "sw_rx_prod %u sw_rx_cons %u mp entries %u free %u",
+ idx, rxq->sw_rx_cons & NUM_RX_BDS(rxq),
+ rte_mempool_count(rxq->mb_pool),
+ rte_mempool_free_count(rxq->mb_pool));
+ return -ENOMEM;
+ }
+ rxq->sw_rx_ring[idx].mbuf = new_mb;
+ rxq->sw_rx_ring[idx].page_offset = 0;
+ mapping = new_mb->buf_physaddr;
+ /* Advance PROD and get BD pointer */
+ rx_bd = (struct eth_rx_bd *)ecore_chain_produce(&rxq->rx_bd_ring);
+ rx_bd->addr.hi = rte_cpu_to_le_32(U64_HI(mapping));
+ rx_bd->addr.lo = rte_cpu_to_le_32(U64_LO(mapping));
+ rxq->sw_rx_prod++;
+ return 0;
+}
+
+static void qede_rx_queue_release_mbufs(struct qede_rx_queue *rxq)
+{
+ uint16_t i;
+
+ if (rxq->sw_rx_ring != NULL) {
+ for (i = 0; i < rxq->nb_rx_desc; i++) {
+ if (NULL != rxq->sw_rx_ring[i].mbuf) {
+ rte_pktmbuf_free(rxq->sw_rx_ring[i].mbuf);
+ rxq->sw_rx_ring[i].mbuf = NULL;
+ }
+ }
+ }
+}
+
+void qede_rx_queue_release(void *rx_queue)
+{
+ struct qede_rx_queue *rxq = rx_queue;
+
+ if (rxq != NULL) {
+ qede_rx_queue_release_mbufs(rxq);
+ rte_free(rxq->sw_rx_ring);
+ rxq->sw_rx_ring = NULL;
+ rte_free(rxq);
+ rx_queue = NULL;
+ }
+}
+
+static uint16_t qede_set_rx_buf_size(struct rte_mempool *mp, uint16_t len)
+{
+ uint16_t data_size;
+ uint16_t buf_size;
+
+ data_size = rte_pktmbuf_data_room_size(mp) - RTE_PKTMBUF_HEADROOM;
+ buf_size = RTE_MAX(len, data_size);
+ return buf_size + QEDE_ETH_OVERHEAD;
+}
+
+int
+qede_rx_queue_setup(struct rte_eth_dev *dev, uint16_t queue_idx,
+ uint16_t nb_desc, unsigned int socket_id,
+ const struct rte_eth_rxconf *rx_conf,
+ struct rte_mempool *mp)
+{
+ struct qede_dev *qdev = dev->data->dev_private;
+ struct ecore_dev *edev = &qdev->edev;
+ struct rte_eth_dev_data *eth_data = dev->data;
+ struct qede_rx_queue *rxq;
+ uint16_t pkt_len = (uint16_t) dev->data->dev_conf.rxmode.max_rx_pkt_len;
+ size_t size;
+ int rc;
+ int i;
+
+ PMD_INIT_FUNC_TRACE(edev);
+
+ /* Note: Ring size/align is controlled by struct rte_eth_desc_lim */
+ if (!rte_is_power_of_2(nb_desc)) {
+ DP_NOTICE(edev, false, "Ring size %u is not power of 2\n",
+ nb_desc);
+ return -EINVAL;
+ }
+
+ /* Free memory prior to re-allocation if needed... */
+ if (dev->data->rx_queues[queue_idx] != NULL) {
+ qede_rx_queue_release(dev->data->rx_queues[queue_idx]);
+ dev->data->rx_queues[queue_idx] = NULL;
+ }
+
+ /* First allocate the rx queue data structure */
+ rxq = rte_zmalloc_socket("qede_rx_queue", sizeof(struct qede_rx_queue),
+ RTE_CACHE_LINE_SIZE, socket_id);
+
+ if (!rxq) {
+ DP_NOTICE(edev, false,
+ "Unable to allocate memory for rxq on socket %u",
+ socket_id);
+ return -ENOMEM;
+ }
+
+ rxq->qdev = qdev;
+ rxq->mb_pool = mp;
+ rxq->nb_rx_desc = nb_desc;
+ rxq->queue_id = queue_idx;
+ rxq->port_id = dev->data->port_id;
+
+ rxq->rx_buf_size = qede_set_rx_buf_size(mp, pkt_len);
+ if (pkt_len > ETHER_MAX_LEN) {
+ dev->data->dev_conf.rxmode.jumbo_frame = 1;
+ DP_NOTICE(edev, false, "jumbo frame enabled\n");
+ } else {
+ dev->data->dev_conf.rxmode.jumbo_frame = 0;
+ }
+
+ qdev->mtu = rxq->rx_buf_size;
+ DP_INFO(edev, "rx_buf_size=%u\n", qdev->mtu);
+
+ /* Allocate the parallel driver ring for Rx buffers */
+ size = sizeof(*rxq->sw_rx_ring) * rxq->nb_rx_desc;
+ rxq->sw_rx_ring = rte_zmalloc_socket("sw_rx_ring", size,
+ RTE_CACHE_LINE_SIZE, socket_id);
+ if (!rxq->sw_rx_ring) {
+ DP_NOTICE(edev, false,
+ "Unable to alloc memory for sw_rx_ring on socket %u\n",
+ socket_id);
+ rte_free(rxq);
+ rxq = NULL;
+ return -ENOMEM;
+ }
+
+ /* Allocate FW Rx ring */
+ rc = qdev->ops->common->chain_alloc(edev,
+ ECORE_CHAIN_USE_TO_CONSUME_PRODUCE,
+ ECORE_CHAIN_MODE_NEXT_PTR,
+ ECORE_CHAIN_CNT_TYPE_U16,
+ rxq->nb_rx_desc,
+ sizeof(struct eth_rx_bd),
+ &rxq->rx_bd_ring);
+
+ if (rc != ECORE_SUCCESS) {
+ DP_NOTICE(edev, false,
+ "Unable to alloc memory for rxbd ring on socket %u\n",
+ socket_id);
+ rte_free(rxq->sw_rx_ring);
+ rxq->sw_rx_ring = NULL;
+ rte_free(rxq);
+ rxq = NULL;
+ }
+
+ /* Allocate FW completion ring */
+ rc = qdev->ops->common->chain_alloc(edev,
+ ECORE_CHAIN_USE_TO_CONSUME,
+ ECORE_CHAIN_MODE_PBL,
+ ECORE_CHAIN_CNT_TYPE_U16,
+ rxq->nb_rx_desc,
+ sizeof(union eth_rx_cqe),
+ &rxq->rx_comp_ring);
+
+ if (rc != ECORE_SUCCESS) {
+ DP_NOTICE(edev, false,
+ "Unable to alloc memory for cqe ring on socket %u\n",
+ socket_id);
+ /* TBD: Freeing RX BD ring */
+ rte_free(rxq->sw_rx_ring);
+ rxq->sw_rx_ring = NULL;
+ rte_free(rxq);
+ }
+
+ /* Allocate buffers for the Rx ring */
+ for (i = 0; i < rxq->nb_rx_desc; i++) {
+ rc = qede_alloc_rx_buffer(rxq);
+ if (rc) {
+ DP_NOTICE(edev, false,
+ "RX buffer allocation failed at idx=%d\n", i);
+ goto err4;
+ }
+ }
+
+ dev->data->rx_queues[queue_idx] = rxq;
+ if (!qdev->rx_queues)
+ qdev->rx_queues = (struct qede_rx_queue **)dev->data->rx_queues;
+
+ DP_NOTICE(edev, false, "rxq %d num_desc %u rx_buf_size=%u socket %u\n",
+ queue_idx, nb_desc, qdev->mtu, socket_id);
+
+ return 0;
+err4:
+ qede_rx_queue_release(rxq);
+ return -ENOMEM;
+}
+
+static void qede_tx_queue_release_mbufs(struct qede_tx_queue *txq)
+{
+ unsigned i;
+
+ PMD_TX_LOG(DEBUG, txq, "releasing %u mbufs\n", txq->nb_tx_desc);
+
+ if (txq->sw_tx_ring != NULL) {
+ for (i = 0; i < txq->nb_tx_desc; i++) {
+ if (txq->sw_tx_ring[i].mbuf != NULL) {
+ rte_pktmbuf_free(txq->sw_tx_ring[i].mbuf);
+ txq->sw_tx_ring[i].mbuf = NULL;
+ }
+ }
+ }
+}
+
+void qede_tx_queue_release(void *tx_queue)
+{
+ struct qede_tx_queue *txq = tx_queue;
+
+ if (txq != NULL) {
+ qede_tx_queue_release_mbufs(txq);
+ if (txq->sw_tx_ring) {
+ rte_free(txq->sw_tx_ring);
+ txq->sw_tx_ring = NULL;
+ }
+ rte_free(txq);
+ }
+ tx_queue = NULL;
+}
+
+int
+qede_tx_queue_setup(struct rte_eth_dev *dev,
+ uint16_t queue_idx,
+ uint16_t nb_desc,
+ unsigned int socket_id,
+ const struct rte_eth_txconf *tx_conf)
+{
+ struct qede_dev *qdev = dev->data->dev_private;
+ struct ecore_dev *edev = &qdev->edev;
+ struct qede_tx_queue *txq;
+ int rc;
+
+ PMD_INIT_FUNC_TRACE(edev);
+
+ if (!rte_is_power_of_2(nb_desc)) {
+ DP_NOTICE(edev, false, "Ring size %u is not power of 2\n",
+ nb_desc);
+ return -EINVAL;
+ }
+
+ /* Free memory prior to re-allocation if needed... */
+ if (dev->data->tx_queues[queue_idx] != NULL) {
+ qede_tx_queue_release(dev->data->tx_queues[queue_idx]);
+ dev->data->tx_queues[queue_idx] = NULL;
+ }
+
+ txq = rte_zmalloc_socket("qede_tx_queue", sizeof(struct qede_tx_queue),
+ RTE_CACHE_LINE_SIZE, socket_id);
+
+ if (txq == NULL) {
+ DP_ERR(edev,
+ "Unable to allocate memory for txq on socket %u",
+ socket_id);
+ return -ENOMEM;
+ }
+
+ txq->nb_tx_desc = nb_desc;
+ txq->qdev = qdev;
+ txq->port_id = dev->data->port_id;
+
+ rc = qdev->ops->common->chain_alloc(edev,
+ ECORE_CHAIN_USE_TO_CONSUME_PRODUCE,
+ ECORE_CHAIN_MODE_PBL,
+ ECORE_CHAIN_CNT_TYPE_U16,
+ txq->nb_tx_desc,
+ sizeof(union eth_tx_bd_types),
+ &txq->tx_pbl);
+ if (rc != ECORE_SUCCESS) {
+ DP_ERR(edev,
+ "Unable to allocate memory for txbd ring on socket %u",
+ socket_id);
+ qede_tx_queue_release(txq);
+ return -ENOMEM;
+ }
+
+ /* Allocate software ring */
+ txq->sw_tx_ring = rte_zmalloc_socket("txq->sw_tx_ring",
+ (sizeof(struct qede_tx_entry) *
+ txq->nb_tx_desc),
+ RTE_CACHE_LINE_SIZE, socket_id);
+
+ if (!txq->sw_tx_ring) {
+ DP_ERR(edev,
+ "Unable to allocate memory for txbd ring on socket %u",
+ socket_id);
+ qede_tx_queue_release(txq);
+ return -ENOMEM;
+ }
+
+ txq->queue_id = queue_idx;
+
+ txq->nb_tx_avail = txq->nb_tx_desc;
+
+ txq->tx_free_thresh =
+ tx_conf->tx_free_thresh ? tx_conf->tx_free_thresh :
+ (txq->nb_tx_desc - QEDE_DEFAULT_TX_FREE_THRESH);
+
+ dev->data->tx_queues[queue_idx] = txq;
+ if (!qdev->tx_queues)
+ qdev->tx_queues = (struct qede_tx_queue **)dev->data->tx_queues;
+
+ txq->txq_counter = 0;
+
+ DP_NOTICE(edev, false,
+ "txq %u num_desc %u tx_free_thresh %u socket %u\n",
+ queue_idx, nb_desc, txq->tx_free_thresh, socket_id);
+
+ return 0;
+}
+
+/* This function inits fp content and resets the SB, RXQ and TXQ arrays */
+static void qede_init_fp(struct qede_dev *qdev)
+{
+ struct qede_fastpath *fp;
+ int rss_id, txq_index, tc;
+
+ memset((void *)qdev->fp_array, 0, (QEDE_RSS_CNT(qdev) *
+ sizeof(*qdev->fp_array)));
+ memset((void *)qdev->sb_array, 0, (QEDE_RSS_CNT(qdev) *
+ sizeof(*qdev->sb_array)));
+ for_each_rss(rss_id) {
+ fp = &qdev->fp_array[rss_id];
+
+ fp->qdev = qdev;
+ fp->rss_id = rss_id;
+
+ /* Point rxq to generic rte queues that was created
+ * as part of queue creation.
+ */
+ fp->rxq = qdev->rx_queues[rss_id];
+ fp->sb_info = &qdev->sb_array[rss_id];
+
+ for (tc = 0; tc < qdev->num_tc; tc++) {
+ txq_index = tc * QEDE_RSS_CNT(qdev) + rss_id;
+ fp->txqs[tc] = qdev->tx_queues[txq_index];
+ fp->txqs[tc]->queue_id = txq_index;
+ /* Updating it to main structure */
+ snprintf(fp->name, sizeof(fp->name), "%s-fp-%d",
+ "qdev", rss_id);
+ }
+ }
+
+ qdev->gro_disable = gro_disable;
+}
+
+void qede_free_fp_arrays(struct qede_dev *qdev)
+{
+ /* It asseumes qede_free_mem_load() is called before */
+ if (qdev->fp_array != NULL) {
+ rte_free(qdev->fp_array);
+ qdev->fp_array = NULL;
+ }
+
+ if (qdev->sb_array != NULL) {
+ rte_free(qdev->sb_array);
+ qdev->sb_array = NULL;
+ }
+}
+
+int qede_alloc_fp_array(struct qede_dev *qdev)
+{
+ struct qede_fastpath *fp;
+ struct ecore_dev *edev = &qdev->edev;
+ int i;
+
+ qdev->fp_array = rte_calloc("fp", QEDE_RSS_CNT(qdev),
+ sizeof(*qdev->fp_array),
+ RTE_CACHE_LINE_SIZE);
+
+ if (!qdev->fp_array) {
+ DP_NOTICE(edev, true, "fp array allocation failed\n");
+ return -ENOMEM;
+ }
+
+ qdev->sb_array = rte_calloc("sb", QEDE_RSS_CNT(qdev),
+ sizeof(*qdev->sb_array),
+ RTE_CACHE_LINE_SIZE);
+
+ if (!qdev->sb_array) {
+ DP_NOTICE(edev, true, "sb array allocation failed\n");
+ rte_free(qdev->fp_array);
+ return -ENOMEM;
+ }
+
+ return 0;
+}
+
+/* This function allocates fast-path status block memory */
+static int
+qede_alloc_mem_sb(struct qede_dev *qdev, struct ecore_sb_info *sb_info,
+ uint16_t sb_id)
+{
+ struct ecore_dev *edev = &qdev->edev;
+ struct status_block *sb_virt;
+ dma_addr_t sb_phys;
+ int rc;
+
+ sb_virt = OSAL_DMA_ALLOC_COHERENT(edev, &sb_phys, sizeof(*sb_virt));
+
+ if (!sb_virt) {
+ DP_ERR(edev, "Status block allocation failed\n");
+ return -ENOMEM;
+ }
+
+ rc = qdev->ops->common->sb_init(edev, sb_info,
+ sb_virt, sb_phys, sb_id,
+ QED_SB_TYPE_L2_QUEUE);
+ if (rc) {
+ DP_ERR(edev, "Status block initialization failed\n");
+ /* TBD: No dma_free_coherent possible */
+ return rc;
+ }
+
+ return 0;
+}
+
+static int qede_alloc_mem_fp(struct qede_dev *qdev, struct qede_fastpath *fp)
+{
+ return qede_alloc_mem_sb(qdev, fp->sb_info, fp->rss_id);
+}
+
+static void qede_shrink_txq(struct qede_dev *qdev, uint16_t num_rss)
+{
+ /* @@@TBD - this should also re-set the qed interrupts */
+}
+
+/* This function allocates all qede memory at NIC load. */
+static int qede_alloc_mem_load(struct qede_dev *qdev)
+{
+ int rc = 0, rss_id;
+ struct ecore_dev *edev = &qdev->edev;
+
+ for (rss_id = 0; rss_id < QEDE_RSS_CNT(qdev); rss_id++) {
+ struct qede_fastpath *fp = &qdev->fp_array[rss_id];
+
+ rc = qede_alloc_mem_fp(qdev, fp);
+ if (rc)
+ break;
+ }
+
+ if (rss_id != QEDE_RSS_CNT(qdev)) {
+ /* Failed allocating memory for all the queues */
+ if (!rss_id) {
+ DP_ERR(edev,
+ "Failed to alloc memory for leading queue\n");
+ rc = -ENOMEM;
+ } else {
+ DP_NOTICE(edev, false,
+ "Failed to allocate memory for all of"
+ "RSS queues\n"
+ "Desired: %d queues, allocated: %d queues\n",
+ QEDE_RSS_CNT(qdev), rss_id);
+ qede_shrink_txq(qdev, rss_id);
+ }
+ qdev->num_rss = rss_id;
+ }
+
+ return 0;
+}
+
+static inline void
+qede_update_rx_prod(struct qede_dev *edev, struct qede_rx_queue *rxq)
+{
+ uint16_t bd_prod = ecore_chain_get_prod_idx(&rxq->rx_bd_ring);
+ uint16_t cqe_prod = ecore_chain_get_prod_idx(&rxq->rx_comp_ring);
+ struct eth_rx_prod_data rx_prods = { 0 };
+
+ /* Update producers */
+ rx_prods.bd_prod = rte_cpu_to_le_16(bd_prod);
+ rx_prods.cqe_prod = rte_cpu_to_le_16(cqe_prod);
+
+ /* Make sure that the BD and SGE data is updated before updating the
+ * producers since FW might read the BD/SGE right after the producer
+ * is updated.
+ */
+ rte_wmb();
+
+ internal_ram_wr(rxq->hw_rxq_prod_addr, sizeof(rx_prods),
+ (uint32_t *) &rx_prods);
+
+ /* mmiowb is needed to synchronize doorbell writes from more than one
+ * processor. It guarantees that the write arrives to the device before
+ * the napi lock is released and another qede_poll is called (possibly
+ * on another CPU). Without this barrier, the next doorbell can bypass
+ * this doorbell. This is applicable to IA64/Altix systems.
+ */
+ rte_wmb();
+
+ PMD_RX_LOG(DEBUG, rxq, "bd_prod %u cqe_prod %u\n", bd_prod, cqe_prod);
+}
+
+static inline uint32_t
+qede_rxfh_indir_default(uint32_t index, uint32_t n_rx_rings)
+{
+ return index % n_rx_rings;
+}
+
+static void qede_prandom_bytes(uint32_t *buff, size_t bytes)
+{
+ unsigned i;
+
+ srand((unsigned int)time(NULL));
+
+ for (i = 0; i < ECORE_RSS_KEY_SIZE; i++)
+ buff[i] = rand();
+}
+
+static int
+qede_config_rss(struct rte_eth_dev *eth_dev,
+ struct qed_update_vport_rss_params *rss_params)
+{
+ enum rte_eth_rx_mq_mode mode = eth_dev->data->dev_conf.rxmode.mq_mode;
+ struct rte_eth_rss_conf rss_conf =
+ eth_dev->data->dev_conf.rx_adv_conf.rss_conf;
+ struct qede_dev *qdev = eth_dev->data->dev_private;
+ struct ecore_dev *edev = &qdev->edev;
+ unsigned i;
+
+ PMD_INIT_FUNC_TRACE(edev);
+
+ /* Check if RSS conditions are met */
+
+ if (!(mode & ETH_MQ_RX_RSS)) {
+ DP_INFO(edev, "RSS flag is not set\n");
+ return -EINVAL;
+ } else {
+ DP_INFO(edev, "RSS flag is set\n");
+ }
+
+ if (rss_conf.rss_hf == 0) {
+ DP_NOTICE(edev, false, "No RSS hash function to apply\n");
+ return -EINVAL;
+ }
+
+ if (QEDE_RSS_CNT(qdev) == 1) {
+ DP_NOTICE(edev, false, "RSS is not enabled with one queue\n");
+ return -EINVAL;
+ }
+
+ memset(rss_params, 0, sizeof(*rss_params));
+
+ for (i = 0; i < 128; i++)
+ rss_params->rss_ind_table[i] = qede_rxfh_indir_default(i,
+ QEDE_RSS_CNT(qdev));
+
+ /* key and protocols */
+ if (rss_conf.rss_key == NULL) {
+ qede_prandom_bytes(rss_params->rss_key,
+ sizeof(rss_params->rss_key));
+ } else {
+ /* Fill given by user */
+ DP_NOTICE(edev, false,
+ "User provided rss key is not supported\n");
+ return -EINVAL;
+ }
+
+ DP_INFO(edev, "RSS check passes\n");
+
+ return 0;
+}
+
+static int qede_start_queues(struct rte_eth_dev *eth_dev, bool clear_stats)
+{
+ struct qede_dev *qdev = eth_dev->data->dev_private;
+ struct ecore_dev *edev = &qdev->edev;
+ struct qed_update_vport_rss_params *rss_params = &qdev->rss_params;
+ struct qed_dev_info *qed_info = &qdev->dev_info.common;
+ struct qed_update_vport_params vport_update_params;
+ struct qed_start_vport_params start = { 0 };
+ int vlan_removal_en = 1;
+ int rc, tc, i;
+
+ if (!qdev->num_rss) {
+ DP_ERR(edev,
+ "Cannot update V-VPORT as active as"
+ "there are no Rx queues\n");
+ return -EINVAL;
+ }
+
+ start.remove_inner_vlan = vlan_removal_en;
+ start.gro_enable = !qdev->gro_disable;
+ start.mtu = qdev->mtu;
+ start.vport_id = 0;
+ start.drop_ttl0 = true;
+ start.clear_stats = clear_stats;
+
+ rc = qdev->ops->vport_start(edev, &start);
+ if (rc) {
+ DP_ERR(edev, "Start V-PORT failed %d\n", rc);
+ return rc;
+ }
+
+ DP_INFO(edev,
+ "Start vport ramrod passed, vport_id = %d,"
+ " MTU = %d, vlan_removal_en = %d\n",
+ start.vport_id, qdev->mtu + 0xe, vlan_removal_en);
+
+ for_each_rss(i) {
+ struct qede_fastpath *fp = &qdev->fp_array[i];
+ dma_addr_t p_phys_table;
+ uint16_t page_cnt;
+
+ p_phys_table = ecore_chain_get_pbl_phys(&fp->rxq->rx_comp_ring);
+ page_cnt = ecore_chain_get_page_cnt(&fp->rxq->rx_comp_ring);
+
+ ecore_sb_ack(fp->sb_info, IGU_INT_DISABLE, 0); /* @DPDK */
+
+ rc = qdev->ops->q_rx_start(edev, i, i, 0,
+ fp->sb_info->igu_sb_id,
+ RX_PI,
+ fp->rxq->rx_buf_size,
+ fp->rxq->rx_bd_ring.p_phys_addr,
+ p_phys_table,
+ page_cnt,
+ &fp->rxq->hw_rxq_prod_addr);
+ if (rc) {
+ DP_ERR(edev, "Start RXQ #%d failed %d\n", i, rc);
+ return rc;
+ }
+
+ fp->rxq->hw_cons_ptr = &fp->sb_info->sb_virt->pi_array[RX_PI];
+
+ qede_update_rx_prod(qdev, fp->rxq);
+
+ for (tc = 0; tc < qdev->num_tc; tc++) {
+ struct qede_tx_queue *txq = fp->txqs[tc];
+ int txq_index = tc * QEDE_RSS_CNT(qdev) + i;
+
+ p_phys_table = ecore_chain_get_pbl_phys(&txq->tx_pbl);
+ page_cnt = ecore_chain_get_page_cnt(&txq->tx_pbl);
+ rc = qdev->ops->q_tx_start(edev, i, txq_index,
+ 0,
+ fp->sb_info->igu_sb_id,
+ TX_PI(tc),
+ p_phys_table, page_cnt,
+ &txq->doorbell_addr);
+ if (rc) {
+ DP_ERR(edev, "Start txq %u failed %d\n",
+ txq_index, rc);
+ return rc;
+ }
+
+ txq->hw_cons_ptr =
+ &fp->sb_info->sb_virt->pi_array[TX_PI(tc)];
+ SET_FIELD(txq->tx_db.data.params,
+ ETH_DB_DATA_DEST, DB_DEST_XCM);
+ SET_FIELD(txq->tx_db.data.params, ETH_DB_DATA_AGG_CMD,
+ DB_AGG_CMD_SET);
+ SET_FIELD(txq->tx_db.data.params,
+ ETH_DB_DATA_AGG_VAL_SEL,
+ DQ_XCM_ETH_TX_BD_PROD_CMD);
+
+ txq->tx_db.data.agg_flags = DQ_XCM_ETH_DQ_CF_CMD;
+ }
+ }
+
+ /* Prepare and send the vport enable */
+ memset(&vport_update_params, 0, sizeof(vport_update_params));
+ vport_update_params.vport_id = start.vport_id;
+ vport_update_params.update_vport_active_flg = 1;
+ vport_update_params.vport_active_flg = 1;
+
+ /* @DPDK */
+ if (qed_info->mf_mode == MF_NPAR && qed_info->tx_switching) {
+ /* TBD: Check SRIOV enabled for VF */
+ vport_update_params.update_tx_switching_flg = 1;
+ vport_update_params.tx_switching_flg = 1;
+ }
+
+ if (!qede_config_rss(eth_dev, rss_params))
+ vport_update_params.update_rss_flg = 1;
+
+ DP_INFO(edev, "Updating RSS flag to %d\n",
+ vport_update_params.update_rss_flg);
+
+ rte_memcpy(&vport_update_params.rss_params, rss_params,
+ sizeof(*rss_params));
+
+ rc = qdev->ops->vport_update(edev, &vport_update_params);
+ if (rc) {
+ DP_ERR(edev, "Update V-PORT failed %d\n", rc);
+ return rc;
+ }
+
+ return 0;
+}
+
+#ifdef ENC_SUPPORTED
+static bool qede_tunn_exist(uint16_t flag)
+{
+ return !!((PARSING_AND_ERR_FLAGS_TUNNELEXIST_MASK <<
+ PARSING_AND_ERR_FLAGS_TUNNELEXIST_SHIFT) & flag);
+}
+
+static inline uint8_t qede_check_tunn_csum(uint16_t flag)
+{
+ uint8_t tcsum = 0;
+ uint16_t csum_flag = 0;
+
+ if ((PARSING_AND_ERR_FLAGS_TUNNELL4CHKSMWASCALCULATED_MASK <<
+ PARSING_AND_ERR_FLAGS_TUNNELL4CHKSMWASCALCULATED_SHIFT) & flag)
+ csum_flag |= PARSING_AND_ERR_FLAGS_TUNNELL4CHKSMERROR_MASK <<
+ PARSING_AND_ERR_FLAGS_TUNNELL4CHKSMERROR_SHIFT;
+
+ if ((PARSING_AND_ERR_FLAGS_L4CHKSMWASCALCULATED_MASK <<
+ PARSING_AND_ERR_FLAGS_L4CHKSMWASCALCULATED_SHIFT) & flag) {
+ csum_flag |= PARSING_AND_ERR_FLAGS_L4CHKSMERROR_MASK <<
+ PARSING_AND_ERR_FLAGS_L4CHKSMERROR_SHIFT;
+ tcsum = QEDE_TUNN_CSUM_UNNECESSARY;
+ }
+
+ csum_flag |= PARSING_AND_ERR_FLAGS_TUNNELIPHDRERROR_MASK <<
+ PARSING_AND_ERR_FLAGS_TUNNELIPHDRERROR_SHIFT |
+ PARSING_AND_ERR_FLAGS_IPHDRERROR_MASK <<
+ PARSING_AND_ERR_FLAGS_IPHDRERROR_SHIFT;
+
+ if (csum_flag & flag)
+ return QEDE_CSUM_ERROR;
+
+ return QEDE_CSUM_UNNECESSARY | tcsum;
+}
+#else
+static inline uint8_t qede_tunn_exist(uint16_t flag)
+{
+ return 0;
+}
+
+static inline uint8_t qede_check_tunn_csum(uint16_t flag)
+{
+ return 0;
+}
+#endif
+
+static inline uint8_t qede_check_notunn_csum(uint16_t flag)
+{
+ uint8_t csum = 0;
+ uint16_t csum_flag = 0;
+
+ if ((PARSING_AND_ERR_FLAGS_L4CHKSMWASCALCULATED_MASK <<
+ PARSING_AND_ERR_FLAGS_L4CHKSMWASCALCULATED_SHIFT) & flag) {
+ csum_flag |= PARSING_AND_ERR_FLAGS_L4CHKSMERROR_MASK <<
+ PARSING_AND_ERR_FLAGS_L4CHKSMERROR_SHIFT;
+ csum = QEDE_CSUM_UNNECESSARY;
+ }
+
+ csum_flag |= PARSING_AND_ERR_FLAGS_IPHDRERROR_MASK <<
+ PARSING_AND_ERR_FLAGS_IPHDRERROR_SHIFT;
+
+ if (csum_flag & flag)
+ return QEDE_CSUM_ERROR;
+
+ return csum;
+}
+
+static inline uint8_t qede_check_csum(uint16_t flag)
+{
+ if (likely(!qede_tunn_exist(flag)))
+ return qede_check_notunn_csum(flag);
+ else
+ return qede_check_tunn_csum(flag);
+}
+
+static inline void qede_rx_bd_ring_consume(struct qede_rx_queue *rxq)
+{
+ ecore_chain_consume(&rxq->rx_bd_ring);
+ rxq->sw_rx_cons++;
+}
+
+static inline void
+qede_reuse_page(struct qede_dev *qdev,
+ struct qede_rx_queue *rxq, struct qede_rx_entry *curr_cons)
+{
+ struct eth_rx_bd *rx_bd_prod = ecore_chain_produce(&rxq->rx_bd_ring);
+ uint16_t idx = rxq->sw_rx_cons & NUM_RX_BDS(rxq);
+ struct qede_rx_entry *curr_prod;
+ dma_addr_t new_mapping;
+
+ curr_prod = &rxq->sw_rx_ring[idx];
+ *curr_prod = *curr_cons;
+
+ new_mapping = curr_prod->mbuf->buf_physaddr + curr_prod->page_offset;
+
+ rx_bd_prod->addr.hi = rte_cpu_to_le_32(U64_HI(new_mapping));
+ rx_bd_prod->addr.lo = rte_cpu_to_le_32(U64_LO(new_mapping));
+
+ rxq->sw_rx_prod++;
+}
+
+static inline void
+qede_recycle_rx_bd_ring(struct qede_rx_queue *rxq,
+ struct qede_dev *qdev, uint8_t count)
+{
+ struct qede_rx_entry *curr_cons;
+
+ for (; count > 0; count--) {
+ curr_cons = &rxq->sw_rx_ring[rxq->sw_rx_cons & NUM_RX_BDS(rxq)];
+ qede_reuse_page(qdev, rxq, curr_cons);
+ qede_rx_bd_ring_consume(rxq);
+ }
+}
+
+uint16_t
+qede_recv_pkts(void *p_rxq, struct rte_mbuf **rx_pkts, uint16_t nb_pkts)
+{
+ struct qede_rx_queue *rxq = p_rxq;
+ struct qede_dev *qdev = rxq->qdev;
+ struct ecore_dev *edev = &qdev->edev;
+ struct qede_fastpath *fp = &qdev->fp_array[rxq->queue_id];
+ uint16_t hw_comp_cons, sw_comp_cons, sw_rx_index;
+ uint16_t rx_pkt = 0;
+ union eth_rx_cqe *cqe;
+ struct eth_fast_path_rx_reg_cqe *fp_cqe;
+ register struct rte_mbuf *rx_mb = NULL;
+ enum eth_rx_cqe_type cqe_type;
+ uint16_t len, pad;
+ uint16_t preload_idx;
+ uint8_t csum_flag;
+ uint16_t parse_flag;
+
+ hw_comp_cons = rte_le_to_cpu_16(*rxq->hw_cons_ptr);
+ sw_comp_cons = ecore_chain_get_cons_idx(&rxq->rx_comp_ring);
+
+ rte_rmb();
+
+ if (hw_comp_cons == sw_comp_cons)
+ return 0;
+
+ while (sw_comp_cons != hw_comp_cons) {
+ /* Get the CQE from the completion ring */
+ cqe =
+ (union eth_rx_cqe *)ecore_chain_consume(&rxq->rx_comp_ring);
+ cqe_type = cqe->fast_path_regular.type;
+
+ if (unlikely(cqe_type == ETH_RX_CQE_TYPE_SLOW_PATH)) {
+ PMD_RX_LOG(DEBUG, rxq, "Got a slowath CQE\n");
+
+ qdev->ops->eth_cqe_completion(edev, fp->rss_id,
+ (struct eth_slow_path_rx_cqe *) cqe);
+ goto next_cqe;
+ }
+
+ /* Get the data from the SW ring */
+ sw_rx_index = rxq->sw_rx_cons & NUM_RX_BDS(rxq);
+ rx_mb = rxq->sw_rx_ring[sw_rx_index].mbuf;
+ assert(rx_mb != NULL);
+
+ /* non GRO */
+ fp_cqe = &cqe->fast_path_regular;
+
+ len = rte_le_to_cpu_16(fp_cqe->len_on_first_bd);
+ pad = fp_cqe->placement_offset;
+#if 0
+ PMD_RX_LOG(DEBUG, rxq,
+ "CQE type = 0x%x, flags = 0x%x, vlan = 0x%x"
+ " len = %u, parsing_flags = %d\n",
+ cqe_type, fp_cqe->bitfields,
+ rte_le_to_cpu_16(fp_cqe->vlan_tag),
+ len, rte_le_to_cpu_16(fp_cqe->pars_flags.flags));
+#endif
+
+ /* If this is an error packet then drop it */
+ parse_flag =
+ rte_le_to_cpu_16(cqe->fast_path_regular.pars_flags.flags);
+ csum_flag = qede_check_csum(parse_flag);
+ if (unlikely(csum_flag == QEDE_CSUM_ERROR)) {
+ PMD_RX_LOG(ERR, rxq,
+ "CQE in CONS = %u has error, flags = 0x%x"
+ "dropping incoming packet\n",
+ sw_comp_cons, parse_flag);
+ rxq->rx_hw_errors++;
+ qede_recycle_rx_bd_ring(rxq, qdev, fp_cqe->bd_num);
+ goto next_cqe;
+ }
+
+ if (unlikely(qede_alloc_rx_buffer(rxq) != 0)) {
+ PMD_RX_LOG(ERR, rxq,
+ "New buffer allocation failed,"
+ "dropping incoming packet\n");
+ qede_recycle_rx_bd_ring(rxq, qdev, fp_cqe->bd_num);
+ rte_eth_devices[rxq->port_id].
+ data->rx_mbuf_alloc_failed++;
+ rxq->rx_alloc_errors++;
+ break;
+ }
+
+ qede_rx_bd_ring_consume(rxq);
+
+ /* Prefetch next mbuf while processing current one. */
+ preload_idx = rxq->sw_rx_cons & NUM_RX_BDS(rxq);
+ rte_prefetch0(rxq->sw_rx_ring[preload_idx].mbuf);
+
+ if (fp_cqe->bd_num != 1)
+ PMD_RX_LOG(DEBUG, rxq,
+ "Jumbo-over-BD packet not supported\n");
+
+ rx_mb->buf_len = len + pad;
+ rx_mb->data_off = pad;
+ rx_mb->nb_segs = 1;
+ rx_mb->data_len = len;
+ rx_mb->pkt_len = len;
+ rx_mb->port = rxq->port_id;
+ rte_prefetch1(rte_pktmbuf_mtod(rx_mb, void *));
+
+ if (CQE_HAS_VLAN(fp_cqe->pars_flags.flags)) {
+ rx_mb->vlan_tci = rte_le_to_cpu_16(fp_cqe->vlan_tag);
+ rx_mb->ol_flags |= PKT_RX_VLAN_PKT;
+ }
+
+ rx_pkts[rx_pkt] = rx_mb;
+ rx_pkt++;
+next_cqe:
+ ecore_chain_recycle_consumed(&rxq->rx_comp_ring);
+ sw_comp_cons = ecore_chain_get_cons_idx(&rxq->rx_comp_ring);
+ if (rx_pkt == nb_pkts) {
+ PMD_RX_LOG(DEBUG, rxq,
+ "Budget reached nb_pkts=%u received=%u\n",
+ rx_pkt, nb_pkts);
+ break;
+ }
+ }
+
+ qede_update_rx_prod(qdev, rxq);
+
+ PMD_RX_LOG(DEBUG, rxq, "rx_pkts=%u core=%d\n", rx_pkt, rte_lcore_id());
+
+ return rx_pkt;
+}
+
+static inline int
+qede_free_tx_pkt(struct ecore_dev *edev, struct qede_tx_queue *txq)
+{
+ uint16_t idx = TX_CONS(txq);
+ struct eth_tx_bd *tx_data_bd;
+ struct rte_mbuf *mbuf = txq->sw_tx_ring[idx].mbuf;
+
+ if (unlikely(!mbuf)) {
+ PMD_TX_LOG(ERR, txq,
+ "null mbuf nb_tx_desc %u nb_tx_avail %u"
+ "sw_tx_cons %u sw_tx_prod %u\n",
+ txq->nb_tx_desc, txq->nb_tx_avail, idx,
+ TX_PROD(txq));
+ return -1;
+ }
+
+ /* Free now */
+ rte_pktmbuf_free_seg(mbuf);
+ txq->sw_tx_ring[idx].mbuf = NULL;
+ ecore_chain_consume(&txq->tx_pbl);
+ txq->nb_tx_avail++;
+
+ return 0;
+}
+
+static inline uint16_t
+qede_process_tx_compl(struct ecore_dev *edev, struct qede_tx_queue *txq)
+{
+ uint16_t tx_compl = 0;
+ uint16_t hw_bd_cons;
+ int rc;
+
+ hw_bd_cons = rte_le_to_cpu_16(*txq->hw_cons_ptr);
+ rte_compiler_barrier();
+
+ while (hw_bd_cons != ecore_chain_get_cons_idx(&txq->tx_pbl)) {
+ rc = qede_free_tx_pkt(edev, txq);
+ if (rc) {
+ DP_NOTICE(edev, true,
+ "hw_bd_cons = %d, chain_cons=%d\n",
+ hw_bd_cons,
+ ecore_chain_get_cons_idx(&txq->tx_pbl));
+ break;
+ }
+ txq->sw_tx_cons++; /* Making TXD available */
+ tx_compl++;
+ }
+
+ PMD_TX_LOG(DEBUG, txq, "Tx compl %u sw_tx_cons %u avail %u\n",
+ tx_compl, txq->sw_tx_cons, txq->nb_tx_avail);
+ return tx_compl;
+}
+
+uint16_t
+qede_xmit_pkts(void *p_txq, struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
+{
+ struct qede_tx_queue *txq = p_txq;
+ struct qede_dev *qdev = txq->qdev;
+ struct ecore_dev *edev = &qdev->edev;
+ struct qede_fastpath *fp = &qdev->fp_array[txq->queue_id];
+ struct eth_tx_1st_bd *first_bd;
+ uint16_t nb_tx_pkts;
+ uint16_t nb_pkt_sent = 0;
+ uint16_t bd_prod;
+ uint16_t idx;
+ uint16_t tx_count;
+
+ if (unlikely(txq->nb_tx_avail < txq->tx_free_thresh)) {
+ PMD_TX_LOG(DEBUG, txq, "send=%u avail=%u free_thresh=%u\n",
+ nb_pkts, txq->nb_tx_avail, txq->tx_free_thresh);
+ (void)qede_process_tx_compl(edev, txq);
+ }
+
+ nb_tx_pkts = RTE_MIN(nb_pkts, (txq->nb_tx_avail / MAX_NUM_TX_BDS));
+ if (unlikely(nb_tx_pkts == 0)) {
+ PMD_TX_LOG(DEBUG, txq, "Out of BDs nb_pkts=%u avail=%u\n",
+ nb_pkts, txq->nb_tx_avail);
+ return 0;
+ }
+
+ tx_count = nb_tx_pkts;
+ while (nb_tx_pkts--) {
+ /* Fill the entry in the SW ring and the BDs in the FW ring */
+ idx = TX_PROD(txq);
+ struct rte_mbuf *mbuf = *tx_pkts++;
+ txq->sw_tx_ring[idx].mbuf = mbuf;
+ first_bd = (struct eth_tx_1st_bd *)
+ ecore_chain_produce(&txq->tx_pbl);
+ first_bd->data.bd_flags.bitfields =
+ 1 << ETH_TX_1ST_BD_FLAGS_START_BD_SHIFT;
+ /* Map mbug linear data for DMA and set in the first BD */
+ QEDE_BD_SET_ADDR_LEN(first_bd, RTE_MBUF_DATA_DMA_ADDR(mbuf),
+ mbuf->data_len);
+
+ /* Descriptor based VLAN insertion */
+ if (mbuf->ol_flags & (PKT_TX_VLAN_PKT | PKT_TX_QINQ_PKT)) {
+ first_bd->data.vlan = rte_cpu_to_le_16(mbuf->vlan_tci);
+ first_bd->data.bd_flags.bitfields |=
+ 1 << ETH_TX_1ST_BD_FLAGS_VLAN_INSERTION_SHIFT;
+ }
+
+ /* Offload the IP checksum in the hardware */
+ if (mbuf->ol_flags & PKT_TX_IP_CKSUM) {
+ first_bd->data.bd_flags.bitfields |=
+ 1 << ETH_TX_1ST_BD_FLAGS_IP_CSUM_SHIFT;
+ }
+
+ /* L4 checksum offload (tcp or udp) */
+ if (mbuf->ol_flags & (PKT_TX_TCP_CKSUM | PKT_TX_UDP_CKSUM)) {
+ first_bd->data.bd_flags.bitfields |=
+ 1 << ETH_TX_1ST_BD_FLAGS_L4_CSUM_SHIFT;
+ /* IPv6 + extn. -> later */
+ }
+ first_bd->data.nbds = MAX_NUM_TX_BDS;
+ txq->sw_tx_prod++;
+ rte_prefetch0(txq->sw_tx_ring[TX_PROD(txq)].mbuf);
+ txq->nb_tx_avail--;
+ bd_prod =
+ rte_cpu_to_le_16(ecore_chain_get_prod_idx(&txq->tx_pbl));
+ nb_pkt_sent++;
+ }
+
+ /* Write value of prod idx into bd_prod */
+ txq->tx_db.data.bd_prod = bd_prod;
+ rte_wmb();
+ rte_compiler_barrier();
+ DIRECT_REG_WR(edev, txq->doorbell_addr, txq->tx_db.raw);
+ rte_wmb();
+
+ /* Check again for Tx completions if enabled */
+#ifdef RTE_LIBRTE_QEDE_TX_COMP_END
+ (void)qede_process_tx_compl(edev, txq);
+#endif
+
+ PMD_TX_LOG(DEBUG, txq, "to_send=%u can_send=%u sent=%u core=%d\n",
+ nb_pkts, tx_count, nb_pkt_sent, rte_lcore_id());
+
+ return nb_pkt_sent;
+}
+
+int qede_dev_start(struct rte_eth_dev *eth_dev)
+{
+ struct qede_dev *qdev = eth_dev->data->dev_private;
+ struct ecore_dev *edev = &qdev->edev;
+ struct qed_link_output link_output;
+ int rc;
+
+ DP_NOTICE(edev, false, "port %u\n", eth_dev->data->port_id);
+
+ PMD_INIT_FUNC_TRACE(edev);
+
+ if (qdev->state == QEDE_START) {
+ DP_INFO(edev, "device already started\n");
+ return 0;
+ }
+
+ if (qdev->state == QEDE_CLOSE) {
+ rc = qede_alloc_fp_array(qdev);
+ qede_init_fp(qdev);
+ rc = qede_alloc_mem_load(qdev);
+ DP_INFO(edev, "Allocated %d RSS queues on %d TC/s\n",
+ QEDE_RSS_CNT(qdev), qdev->num_tc);
+ } else if (qdev->state == QEDE_STOP) {
+ DP_INFO(edev, "restarting port %u\n", eth_dev->data->port_id);
+ } else {
+ DP_INFO(edev, "unknown state port %u\n",
+ eth_dev->data->port_id);
+ return -EINVAL;
+ }
+
+ rc = qede_start_queues(eth_dev, true);
+
+ if (rc) {
+ DP_ERR(edev, "Failed to start queues\n");
+ /* TBD: free */
+ return rc;
+ }
+
+ DP_INFO(edev, "Start VPORT, RXQ and TXQ succeeded\n");
+
+ qede_dev_set_link_state(eth_dev, true);
+
+ /* Query whether link is already-up */
+ memset(&link_output, 0, sizeof(link_output));
+ qdev->ops->common->get_link(edev, &link_output);
+ DP_NOTICE(edev, false, "link status: %s\n",
+ link_output.link_up ? "up" : "down");
+
+ qdev->state = QEDE_START;
+
+ qede_config_rx_mode(eth_dev);
+
+ DP_INFO(edev, "dev_state is QEDE_START\n");
+
+ return 0;
+}
+
+static int qede_drain_txq(struct qede_dev *qdev,
+ struct qede_tx_queue *txq, bool allow_drain)
+{
+ struct ecore_dev *edev = &qdev->edev;
+ int rc, cnt = 1000;
+
+ while (txq->sw_tx_cons != txq->sw_tx_prod) {
+ qede_process_tx_compl(edev, txq);
+ if (!cnt) {
+ if (allow_drain) {
+ DP_NOTICE(edev, true,
+ "Tx queue[%u] is stuck,"
+ "requesting MCP to drain\n",
+ txq->queue_id);
+ rc = qdev->ops->common->drain(edev);
+ if (rc)
+ return rc;
+ return qede_drain_txq(qdev, txq, false);
+ } else {
+ DP_NOTICE(edev, true,
+ "Timeout waiting for tx queue[%d]:"
+ "PROD=%d, CONS=%d\n",
+ txq->queue_id, txq->sw_tx_prod,
+ txq->sw_tx_cons);
+ return -ENODEV;
+ }
+ }
+ cnt--;
+ DELAY(1000);
+ rte_compiler_barrier();
+ }
+
+ /* FW finished processing, wait for HW to transmit all tx packets */
+ DELAY(2000);
+
+ return 0;
+}
+
+static int qede_stop_queues(struct qede_dev *qdev)
+{
+ struct qed_update_vport_params vport_update_params;
+ struct ecore_dev *edev = &qdev->edev;
+ int rc, tc, i;
+
+ /* Disable the vport */
+ memset(&vport_update_params, 0, sizeof(vport_update_params));
+ vport_update_params.vport_id = 0;
+ vport_update_params.update_vport_active_flg = 1;
+ vport_update_params.vport_active_flg = 0;
+ vport_update_params.update_rss_flg = 0;
+
+ DP_INFO(edev, "vport_update\n");
+
+ rc = qdev->ops->vport_update(edev, &vport_update_params);
+ if (rc) {
+ DP_ERR(edev, "Failed to update vport\n");
+ return rc;
+ }
+
+ DP_INFO(edev, "Flushing tx queues\n");
+
+ /* Flush Tx queues. If needed, request drain from MCP */
+ for_each_rss(i) {
+ struct qede_fastpath *fp = &qdev->fp_array[i];
+ for (tc = 0; tc < qdev->num_tc; tc++) {
+ struct qede_tx_queue *txq = fp->txqs[tc];
+ rc = qede_drain_txq(qdev, txq, true);
+ if (rc)
+ return rc;
+ }
+ }
+
+ /* Stop all Queues in reverse order */
+ for (i = QEDE_RSS_CNT(qdev) - 1; i >= 0; i--) {
+ struct qed_stop_rxq_params rx_params;
+
+ /* Stop the Tx Queue(s) */
+ for (tc = 0; tc < qdev->num_tc; tc++) {
+ struct qed_stop_txq_params tx_params;
+
+ tx_params.rss_id = i;
+ tx_params.tx_queue_id = tc * QEDE_RSS_CNT(qdev) + i;
+
+ DP_INFO(edev, "Stopping tx queues\n");
+ rc = qdev->ops->q_tx_stop(edev, &tx_params);
+ if (rc) {
+ DP_ERR(edev, "Failed to stop TXQ #%d\n",
+ tx_params.tx_queue_id);
+ return rc;
+ }
+ }
+
+ /* Stop the Rx Queue */
+ memset(&rx_params, 0, sizeof(rx_params));
+ rx_params.rss_id = i;
+ rx_params.rx_queue_id = i;
+ rx_params.eq_completion_only = 1;
+
+ DP_INFO(edev, "Stopping rx queues\n");
+
+ rc = qdev->ops->q_rx_stop(edev, &rx_params);
+ if (rc) {
+ DP_ERR(edev, "Failed to stop RXQ #%d\n", i);
+ return rc;
+ }
+ }
+
+ DP_INFO(edev, "Stopping vports\n");
+
+ /* Stop the vport */
+ rc = qdev->ops->vport_stop(edev, 0);
+ if (rc)
+ DP_ERR(edev, "Failed to stop VPORT\n");
+
+ return rc;
+}
+
+void qede_reset_fp_rings(struct qede_dev *qdev)
+{
+ uint16_t rss_id;
+ uint8_t tc;
+
+ for_each_rss(rss_id) {
+ DP_INFO(&qdev->edev, "reset fp chain for rss %u\n", rss_id);
+ struct qede_fastpath *fp = &qdev->fp_array[rss_id];
+ ecore_chain_reset(&fp->rxq->rx_bd_ring);
+ ecore_chain_reset(&fp->rxq->rx_comp_ring);
+ for (tc = 0; tc < qdev->num_tc; tc++) {
+ struct qede_tx_queue *txq = fp->txqs[tc];
+ ecore_chain_reset(&txq->tx_pbl);
+ }
+ }
+}
+
+/* This function frees all memory of a single fp */
+static void qede_free_mem_fp(struct qede_dev *qdev, struct qede_fastpath *fp)
+{
+ uint8_t tc;
+
+ qede_rx_queue_release(fp->rxq);
+ for (tc = 0; tc < qdev->num_tc; tc++)
+ qede_tx_queue_release(fp->txqs[tc]);
+}
+
+void qede_free_mem_load(struct qede_dev *qdev)
+{
+ uint8_t rss_id;
+
+ for_each_rss(rss_id) {
+ struct qede_fastpath *fp = &qdev->fp_array[rss_id];
+ qede_free_mem_fp(qdev, fp);
+ }
+ /* qdev->num_rss = 0; */
+}
+
+/*
+ * Stop an Ethernet device. The device can be restarted with a call to
+ * rte_eth_dev_start().
+ * Do not change link state and do not release sw structures.
+ */
+void qede_dev_stop(struct rte_eth_dev *eth_dev)
+{
+ struct qede_dev *qdev = eth_dev->data->dev_private;
+ struct ecore_dev *edev = &qdev->edev;
+ int rc;
+
+ DP_NOTICE(edev, false, "port %u\n", eth_dev->data->port_id);
+
+ PMD_INIT_FUNC_TRACE(edev);
+
+ if (qdev->state != QEDE_START) {
+ DP_INFO(edev, "device not yet started\n");
+ return;
+ }
+
+ rc = qede_stop_queues(qdev);
+
+ if (rc)
+ DP_ERR(edev, "Didn't succeed to close queues\n");
+
+ DP_INFO(edev, "Stopped queues\n");
+
+ qdev->ops->fastpath_stop(edev);
+
+ qede_reset_fp_rings(qdev);
+
+ qdev->state = QEDE_STOP;
+
+ DP_INFO(edev, "dev_state is QEDE_STOP\n");
+}
diff --git a/drivers/net/qede/qede_rxtx.h b/drivers/net/qede/qede_rxtx.h
new file mode 100644
index 0000000..1ce1b81
--- /dev/null
+++ b/drivers/net/qede/qede_rxtx.h
@@ -0,0 +1,183 @@
+/*
+ * Copyright (c) 2016 QLogic Corporation.
+ * All rights reserved.
+ * www.qlogic.com
+ *
+ * See LICENSE.qede_pmd for copyright and licensing details.
+ */
+
+
+#ifndef _QEDE_RXTX_H_
+#define _QEDE_RXTX_H_
+
+#include "qede_ethdev.h"
+
+/* Ring Descriptors */
+#define RX_RING_SIZE_POW 16 /* 64K */
+#define RX_RING_SIZE (1ULL << RX_RING_SIZE_POW)
+#define NUM_RX_BDS_MAX (RX_RING_SIZE - 1)
+#define NUM_RX_BDS_MIN 128
+#define NUM_RX_BDS_DEF NUM_RX_BDS_MAX
+#define NUM_RX_BDS(q) (q->nb_rx_desc - 1)
+
+#define TX_RING_SIZE_POW 16 /* 64K */
+#define TX_RING_SIZE (1ULL << TX_RING_SIZE_POW)
+#define NUM_TX_BDS_MAX (TX_RING_SIZE - 1)
+#define NUM_TX_BDS_MIN 128
+#define NUM_TX_BDS_DEF NUM_TX_BDS_MAX
+#define NUM_TX_BDS(q) (q->nb_tx_desc - 1)
+
+#define TX_CONS(txq) (txq->sw_tx_cons & NUM_TX_BDS(txq))
+#define TX_PROD(txq) (txq->sw_tx_prod & NUM_TX_BDS(txq))
+
+/* Number of TX BDs per packet used currently */
+#define MAX_NUM_TX_BDS 1
+
+#define QEDE_DEFAULT_TX_FREE_THRESH 32
+
+#define QEDE_CSUM_ERROR (1 << 0)
+#define QEDE_CSUM_UNNECESSARY (1 << 1)
+#define QEDE_TUNN_CSUM_UNNECESSARY (1 << 2)
+
+#define RTE_MBUF_DATA_DMA_ADDR(mb) \
+ ((uint64_t) ((mb)->buf_physaddr + (mb)->data_off))
+
+#define QEDE_BD_SET_ADDR_LEN(bd, maddr, len) \
+ do { \
+ (bd)->addr.hi = rte_cpu_to_le_32(U64_HI(maddr)); \
+ (bd)->addr.lo = rte_cpu_to_le_32(U64_LO(maddr)); \
+ (bd)->nbytes = rte_cpu_to_le_16(len); \
+ } while (0)
+
+#define CQE_HAS_VLAN(flags) \
+ ((flags) & (PARSING_AND_ERR_FLAGS_TAG8021QEXIST_MASK \
+ << PARSING_AND_ERR_FLAGS_TAG8021QEXIST_SHIFT))
+
+#define QEDE_IP_HEADER_ALIGNMENT_PADDING 2
+
+/* Max supported alignment is 256 (8 shift)
+ * minimal alignment shift 6 is optimal for 57xxx HW performance
+ */
+#define QEDE_L1_CACHE_SHIFT 6
+#define QEDE_RX_ALIGN_SHIFT (RTE_MAX(6, RTE_MIN(8, QEDE_L1_CACHE_SHIFT)))
+#define QEDE_FW_RX_ALIGN_END (1UL << QEDE_RX_ALIGN_SHIFT)
+
+/* L2 header size + 2*VLANs (8 bytes) + LLC SNAP (8 bytes) */
+#define QEDE_ETH_OVERHEAD (ETHER_HDR_LEN + ETHER_CRC_LEN + \
+ 8 + 8 + QEDE_IP_HEADER_ALIGNMENT_PADDING + \
+ QEDE_FW_RX_ALIGN_END)
+
+/* TBD: Excluding IPV6 */
+#define QEDE_RSS_OFFLOAD_ALL (ETH_RSS_IPV4 | ETH_RSS_NONFRAG_IPV4_TCP | \
+ ETH_RSS_NONFRAG_IPV4_UDP)
+
+#define QEDE_TXQ_FLAGS ((uint32_t)ETH_TXQ_FLAGS_NOMULTSEGS)
+
+#define MAX_NUM_TC 8
+
+#define for_each_rss(i) for (i = 0; i < qdev->num_rss; i++)
+
+/*
+ * RX BD descriptor ring
+ */
+struct qede_rx_entry {
+ struct rte_mbuf *mbuf;
+ uint32_t page_offset;
+ /* allows expansion .. */
+};
+
+/*
+ * Structure associated with each RX queue.
+ */
+struct qede_rx_queue {
+ struct rte_mempool *mb_pool;
+ struct ecore_chain rx_bd_ring;
+ struct ecore_chain rx_comp_ring;
+ uint16_t *hw_cons_ptr;
+ void OSAL_IOMEM *hw_rxq_prod_addr;
+ struct qede_rx_entry *sw_rx_ring;
+ uint16_t sw_rx_cons;
+ uint16_t sw_rx_prod;
+ uint16_t nb_rx_desc;
+ uint16_t queue_id;
+ uint16_t port_id;
+ uint16_t rx_buf_size;
+ uint64_t rx_hw_errors;
+ uint64_t rx_alloc_errors;
+ struct qede_dev *qdev;
+};
+
+/*
+ * TX BD descriptor ring
+ */
+struct qede_tx_entry {
+ struct rte_mbuf *mbuf;
+ uint8_t flags;
+};
+
+union db_prod {
+ struct eth_db_data data;
+ uint32_t raw;
+};
+
+struct qede_tx_queue {
+ struct ecore_chain tx_pbl;
+ struct qede_tx_entry *sw_tx_ring;
+ uint16_t nb_tx_desc;
+ uint16_t nb_tx_avail;
+ uint16_t tx_free_thresh;
+ uint16_t queue_id;
+ uint16_t *hw_cons_ptr;
+ uint16_t sw_tx_cons;
+ uint16_t sw_tx_prod;
+ void OSAL_IOMEM *doorbell_addr;
+ volatile union db_prod tx_db;
+ uint16_t port_id;
+ uint64_t txq_counter;
+ struct qede_dev *qdev;
+};
+
+struct qede_fastpath {
+ struct qede_dev *qdev;
+ uint8_t rss_id;
+ struct ecore_sb_info *sb_info;
+ struct qede_rx_queue *rxq;
+ struct qede_tx_queue *txqs[MAX_NUM_TC];
+ char name[80];
+};
+
+/*
+ * RX/TX function prototypes
+ */
+int qede_rx_queue_setup(struct rte_eth_dev *dev, uint16_t queue_idx,
+ uint16_t nb_desc, unsigned int socket_id,
+ const struct rte_eth_rxconf *rx_conf,
+ struct rte_mempool *mp);
+
+int qede_tx_queue_setup(struct rte_eth_dev *dev,
+ uint16_t queue_idx,
+ uint16_t nb_desc,
+ unsigned int socket_id,
+ const struct rte_eth_txconf *tx_conf);
+
+void qede_rx_queue_release(void *rx_queue);
+
+void qede_tx_queue_release(void *tx_queue);
+
+int qede_dev_start(struct rte_eth_dev *eth_dev);
+
+void qede_dev_stop(struct rte_eth_dev *eth_dev);
+
+void qede_reset_fp_rings(struct qede_dev *qdev);
+
+void qede_free_fp_arrays(struct qede_dev *qdev);
+
+void qede_free_mem_load(struct qede_dev *qdev);
+
+uint16_t qede_xmit_pkts(void *p_txq, struct rte_mbuf **tx_pkts,
+ uint16_t nb_pkts);
+
+uint16_t qede_recv_pkts(void *p_rxq, struct rte_mbuf **rx_pkts,
+ uint16_t nb_pkts);
+
+#endif /* _QEDE_RXTX_H_ */
diff --git a/drivers/net/qede/rte_pmd_qede_version.map b/drivers/net/qede/rte_pmd_qede_version.map
new file mode 100644
index 0000000..5151684
--- /dev/null
+++ b/drivers/net/qede/rte_pmd_qede_version.map
@@ -0,0 +1,4 @@
+DPDK_2.2 {
+
+ local: *;
+};
--
1.8.3.1
^ permalink raw reply [flat|nested] 25+ messages in thread
* [dpdk-dev] [PATCH 6/6] qede: enable PMD build
2016-02-20 15:40 [dpdk-dev] [PATCH 0/6] DPDK PMD for new QLogic FastLinQ QL4xxxx 25G/40G CNAs Harish Patil
` (3 preceding siblings ...)
2016-02-20 15:40 ` [dpdk-dev] [PATCH 5/6] qede: add driver Harish Patil
@ 2016-02-20 15:40 ` Harish Patil
2016-02-20 20:49 ` [dpdk-dev] [PATCH 0/6] DPDK PMD for new QLogic FastLinQ QL4xxxx 25G/40G CNAs Thomas Monjalon
2016-03-08 14:01 ` Bruce Richardson
6 siblings, 0 replies; 25+ messages in thread
From: Harish Patil @ 2016-02-20 15:40 UTC (permalink / raw)
To: dev; +Cc: Sony Chacko
Signed-off-by: Harish Patil <harish.patil@qlogic.com>
Signed-off-by: Rasesh Mody <rasesh.mody@qlogic.com>
Signed-off-by: Sony Chacko <sony.chacko@qlogic.com>
---
config/common_bsdapp | 15 +++++++++++++++
config/common_linuxapp | 16 ++++++++++++++++
drivers/net/Makefile | 1 +
mk/rte.app.mk | 2 ++
scripts/test-build.sh | 1 +
5 files changed, 35 insertions(+)
diff --git a/config/common_bsdapp b/config/common_bsdapp
index 696382c..b1a6f38 100644
--- a/config/common_bsdapp
+++ b/config/common_bsdapp
@@ -295,6 +295,21 @@ CONFIG_RTE_LIBRTE_BOND_DEBUG_ALB=n
CONFIG_RTE_LIBRTE_BOND_DEBUG_ALB_L1=n
#
+# QLogic 25G/40G PMD
+#
+CONFIG_RTE_LIBRTE_QEDE_PMD=y
+CONFIG_RTE_LIBRTE_QEDE_DEBUG_INIT=n
+CONFIG_RTE_LIBRTE_QEDE_DEBUG_INFO=n
+CONFIG_RTE_LIBRTE_QEDE_DEBUG_ECORE=n
+CONFIG_RTE_LIBRTE_QEDE_DEBUG_TX=n
+CONFIG_RTE_LIBRTE_QEDE_DEBUG_RX=n
+CONFIG_RTE_LIBRTE_QEDE_RX_COAL_US=24
+CONFIG_RTE_LIBRTE_QEDE_TX_COAL_US=48
+CONFIG_RTE_LIBRTE_QEDE_TX_SWITCHING=y
+#Provides path/name of the firmware file
+CONFIG_RTE_LIBRTE_QEDE_FW=n
+
+#
# Compile null PMD
#
CONFIG_RTE_LIBRTE_PMD_NULL=y
diff --git a/config/common_linuxapp b/config/common_linuxapp
index f1638db..52e966c 100644
--- a/config/common_linuxapp
+++ b/config/common_linuxapp
@@ -308,6 +308,22 @@ CONFIG_RTE_LIBRTE_PMD_AF_PACKET=y
#
CONFIG_RTE_LIBRTE_PMD_XENVIRT=n
+
+#
+# QLogic 25G/40G PMD
+#
+CONFIG_RTE_LIBRTE_QEDE_PMD=y
+CONFIG_RTE_LIBRTE_QEDE_DEBUG_INIT=n
+CONFIG_RTE_LIBRTE_QEDE_DEBUG_INFO=n
+CONFIG_RTE_LIBRTE_QEDE_DEBUG_ECORE=n
+CONFIG_RTE_LIBRTE_QEDE_DEBUG_TX=n
+CONFIG_RTE_LIBRTE_QEDE_DEBUG_RX=n
+CONFIG_RTE_LIBRTE_QEDE_RX_COAL_US=24
+CONFIG_RTE_LIBRTE_QEDE_TX_COAL_US=48
+CONFIG_RTE_LIBRTE_QEDE_TX_SWITCHING=y
+#Provides path/name of the firmware file
+CONFIG_RTE_LIBRTE_QEDE_FW=n
+
#
# Compile null PMD
#
diff --git a/drivers/net/Makefile b/drivers/net/Makefile
index 6e4497e..fd10734 100644
--- a/drivers/net/Makefile
+++ b/drivers/net/Makefile
@@ -51,6 +51,7 @@ DIRS-$(CONFIG_RTE_LIBRTE_PMD_SZEDATA2) += szedata2
DIRS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += virtio
DIRS-$(CONFIG_RTE_LIBRTE_VMXNET3_PMD) += vmxnet3
DIRS-$(CONFIG_RTE_LIBRTE_PMD_XENVIRT) += xenvirt
+DIRS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += qede
include $(RTE_SDK)/mk/rte.sharelib.mk
include $(RTE_SDK)/mk/rte.subdir.mk
diff --git a/mk/rte.app.mk b/mk/rte.app.mk
index 8ecab41..bebb06e 100644
--- a/mk/rte.app.mk
+++ b/mk/rte.app.mk
@@ -109,6 +109,7 @@ _LDLIBS-$(CONFIG_RTE_LIBRTE_MLX5_PMD) += -libverbs
_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_SZEDATA2) += -lsze2
_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_XENVIRT) += -lxenstore
_LDLIBS-$(CONFIG_RTE_LIBRTE_MPIPE_PMD) += -lgxio
+_LDLIBS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += -lz
# QAT PMD has a dependency on libcrypto (from openssl) for calculating HMAC precomputes
_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_QAT) += -lcrypto
endif # CONFIG_RTE_BUILD_COMBINE_LIBS or not CONFIG_RTE_BUILD_SHARED_LIBS
@@ -152,6 +153,7 @@ _LDLIBS-$(CONFIG_RTE_LIBRTE_MPIPE_PMD) += -lrte_pmd_mpipe
_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_RING) += -lrte_pmd_ring
_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_PCAP) += -lrte_pmd_pcap
_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_AF_PACKET) += -lrte_pmd_af_packet
+_LDLIBS-$(CONFIG_RTE_LIBRTE_QEDE_PMD) += -lrte_pmd_qede
_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_NULL) += -lrte_pmd_null
_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_QAT) += -lrte_pmd_qat
diff --git a/scripts/test-build.sh b/scripts/test-build.sh
index 6d28c5d..dfdc190 100755
--- a/scripts/test-build.sh
+++ b/scripts/test-build.sh
@@ -118,6 +118,7 @@ config () # <directory> <target> <options>
test "$DPDK_DEP_ZLIB" != y || \
sed -ri 's,(BNX2X_PMD=)n,\1y,' $1/.config
sed -ri 's,(NFP_PMD=)n,\1y,' $1/.config
+ sed -ri 's,(QEDE_PMD=)n,\1y,' $1/.config
test "$DPDK_DEP_PCAP" != y || \
sed -ri 's,(PCAP=)n,\1y,' $1/.config
test -z "$AESNI_MULTI_BUFFER_LIB_PATH" || \
--
1.8.3.1
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [dpdk-dev] [PATCH 0/6] DPDK PMD for new QLogic FastLinQ QL4xxxx 25G/40G CNAs
2016-02-20 15:40 [dpdk-dev] [PATCH 0/6] DPDK PMD for new QLogic FastLinQ QL4xxxx 25G/40G CNAs Harish Patil
` (4 preceding siblings ...)
2016-02-20 15:40 ` [dpdk-dev] [PATCH 6/6] qede: enable PMD build Harish Patil
@ 2016-02-20 20:49 ` Thomas Monjalon
2016-02-22 16:47 ` Harish Patil
2016-03-08 14:01 ` Bruce Richardson
6 siblings, 1 reply; 25+ messages in thread
From: Thomas Monjalon @ 2016-02-20 20:49 UTC (permalink / raw)
To: Harish Patil; +Cc: dev
2016-02-20 07:40, Harish Patil:
> This patch set introduces DPDK poll mode driver for new QLogic FastLinQ QL4xxxx
> 25G/40G capable family of CNAs as well as their SR-IOV Virtual Functions (VF).
>
> The overall PMD driver design includes a common module called ecore that deals
> with the low level HW and a upper layer portion that provides the glue logic.
>
> Specifically, the ecore module contains all of the common logic,
> e.g. initialization, cleanup, infrastructure for interrupt handling, link
> management, slowpath etc. as well as protocol agnostic features and supplying
> an abstraction layer for other modules.
>
> The higher layer implements DPDK exported APIs/driver entry points by
> interfacing with the common module for configuration/status and also the
> fastpath routines.
A new driver is always a good news :)
Please could you share some performance numbers?
What is the status about the integration of this driver in other environments?
The layer named ecore seems to be what is named a base driver in other
DPDK drivers. Maybe you could consider renaming the directory to base/ for
consistency?
About the format of the patches, I think it's better to split driver imports
in several pieces to ease reviewing and later reference for specific bugs.
What do you think of introducing basic features one by one?
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [dpdk-dev] [PATCH 3/6] qede: add QLogic PCI ids
2016-02-20 15:40 ` [dpdk-dev] [PATCH 3/6] qede: add QLogic PCI ids Harish Patil
@ 2016-02-21 1:17 ` Stephen Hemminger
2016-02-22 23:23 ` Harish Patil
0 siblings, 1 reply; 25+ messages in thread
From: Stephen Hemminger @ 2016-02-21 1:17 UTC (permalink / raw)
To: Harish Patil; +Cc: dev, Sony Chacko
On Sat, 20 Feb 2016 07:40:28 -0800
Harish Patil <harish.patil@qlogic.com> wrote:
> diff --git a/lib/librte_eal/common/include/rte_pci_dev_ids.h b/lib/librte_eal/common/include/rte_pci_dev_ids.h
> index d088191..0c1a3fe 100644
> --- a/lib/librte_eal/common/include/rte_pci_dev_ids.h
> +++ b/lib/librte_eal/common/include/rte_pci_dev_ids.h
> @@ -152,6 +152,14 @@
> #define RTE_PCI_DEV_ID_DECL_BNX2XVF(vend, dev)
> #endif
>
> +#ifndef RTE_PCI_DEV_ID_DECL_QEDE
> +#define RTE_PCI_DEV_ID_DECL_QEDE(vend, dev)
> +#endif
> +
> +#ifndef RTE_PCI_DEV_ID_DECL_QEDEVF
> +#define RTE_PCI_DEV_ID_DECL_QEDEVF(vend, dev)
> +#endif
> +
> #ifndef PCI_VENDOR_ID_INTEL
> /** Vendor ID used by Intel devices */
> #define PCI_VENDOR_ID_INTEL 0x8086
> @@ -177,6 +185,11 @@
> #define PCI_VENDOR_ID_BROADCOM 0x14E4
> #endif
>
> +#ifndef PCI_VENDOR_ID_QLOGIC
> +/** Vendor ID used by QLogic devices */
> +#define PCI_VENDOR_ID_QLOGIC 0x1077
> +#endif
> +
The goal is get away from having every PCI id for every device
stuffed into rte_pci_dev_ids.h. So it would be good not to add
more now.
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [dpdk-dev] [PATCH 5/6] qede: add driver
2016-02-20 15:40 ` [dpdk-dev] [PATCH 5/6] qede: add driver Harish Patil
@ 2016-02-21 1:26 ` Stephen Hemminger
2016-02-23 2:28 ` Harish Patil
0 siblings, 1 reply; 25+ messages in thread
From: Stephen Hemminger @ 2016-02-21 1:26 UTC (permalink / raw)
To: Harish Patil; +Cc: dev, Sony Chacko
On Sat, 20 Feb 2016 07:40:30 -0800
Harish Patil <harish.patil@qlogic.com> wrote:
> Signed-off-by: Harish Patil <harish.patil@qlogic.com>
> Signed-off-by: Rasesh Mody <rasesh.mody@qlogic.com>
> Signed-off-by: Sony Chacko <sony.chacko@qlogic.com>
> ---
> drivers/net/qede/LICENSE.qede_pmd | 28 +
> drivers/net/qede/Makefile | 95 ++
> drivers/net/qede/qede_eth_if.c | 461 ++++++++++
> drivers/net/qede/qede_eth_if.h | 180 ++++
> drivers/net/qede/qede_ethdev.c | 815 +++++++++++++++++
> drivers/net/qede/qede_ethdev.h | 138 +++
> drivers/net/qede/qede_if.h | 167 ++++
> drivers/net/qede/qede_logs.h | 93 ++
> drivers/net/qede/qede_main.c | 587 +++++++++++++
> drivers/net/qede/qede_rxtx.c | 1346 +++++++++++++++++++++++++++++
> drivers/net/qede/qede_rxtx.h | 183 ++++
> drivers/net/qede/rte_pmd_qede_version.map | 4 +
> 12 files changed, 4097 insertions(+)
> create mode 100644 drivers/net/qede/LICENSE.qede_pmd
> create mode 100644 drivers/net/qede/Makefile
> create mode 100644 drivers/net/qede/qede_eth_if.c
> create mode 100644 drivers/net/qede/qede_eth_if.h
> create mode 100644 drivers/net/qede/qede_ethdev.c
> create mode 100644 drivers/net/qede/qede_ethdev.h
> create mode 100644 drivers/net/qede/qede_if.h
> create mode 100644 drivers/net/qede/qede_logs.h
> create mode 100644 drivers/net/qede/qede_main.c
> create mode 100644 drivers/net/qede/qede_rxtx.c
> create mode 100644 drivers/net/qede/qede_rxtx.h
> create mode 100644 drivers/net/qede/rte_pmd_qede_version.map
DPDK doesn't follow all the kernel style rules, but lots of them.
The biggest catch was the ; in the LOG macros.
Minor stuff about spacing around casts etc.
Running checkpatch with --ignore PREFER_KERNEL_TYPES,LINE_SPACING,PARENTHESIS_ALIGNMENT,BIT_MACRO,NETWORKING_BLOCK_COMMENT_STYLE,SPLIT_STRING,COMPARISON_TO_NULL,CONCATENATED_STRING
Produces:
CHECK:SPACING: spaces preferred around that '*' (ctx:VxO)
#410: FILE: drivers/net/qede/qede_eth_if.c:175:
+ uint16_t cqe_pbl_size, void OSAL_IOMEM**pp_prod)
^
CHECK:SPACING: spaces preferred around that '*' (ctx:VxO)
#467: FILE: drivers/net/qede/qede_eth_if.c:232:
+ uint16_t pbl_size, void OSAL_IOMEM**pp_doorbell)
^
CHECK:BRACES: Blank lines aren't necessary after an open brace '{'
#519: FILE: drivers/net/qede/qede_eth_if.c:284:
+{
+
WARNING:SPACING: Unnecessary space before function pointer arguments
#823: FILE: drivers/net/qede/qede_eth_if.h:121:
+ int (*fill_dev_info) (struct ecore_dev *edev,
WARNING:SPACING: Unnecessary space before function pointer arguments
#826: FILE: drivers/net/qede/qede_eth_if.h:124:
+ int (*vport_start) (struct ecore_dev *edev,
WARNING:SPACING: Unnecessary space before function pointer arguments
#829: FILE: drivers/net/qede/qede_eth_if.h:127:
+ int (*vport_stop) (struct ecore_dev *edev, uint8_t vport_id);
WARNING:SPACING: Unnecessary space before function pointer arguments
#831: FILE: drivers/net/qede/qede_eth_if.h:129:
+ int (*vport_update) (struct ecore_dev *edev,
WARNING:SPACING: Unnecessary space before function pointer arguments
#834: FILE: drivers/net/qede/qede_eth_if.h:132:
+ int (*q_rx_start) (struct ecore_dev *cdev,
CHECK:SPACING: spaces preferred around that '*' (ctx:VxO)
#840: FILE: drivers/net/qede/qede_eth_if.h:138:
+ uint16_t cqe_pbl_size, void OSAL_IOMEM**pp_prod);
^
WARNING:SPACING: Unnecessary space before function pointer arguments
#842: FILE: drivers/net/qede/qede_eth_if.h:140:
+ int (*q_rx_stop) (struct ecore_dev *edev,
WARNING:SPACING: Unnecessary space before function pointer arguments
#845: FILE: drivers/net/qede/qede_eth_if.h:143:
+ int (*q_tx_start) (struct ecore_dev *edev,
CHECK:SPACING: spaces preferred around that '*' (ctx:VxO)
#850: FILE: drivers/net/qede/qede_eth_if.h:148:
+ uint16_t pbl_size, void OSAL_IOMEM**pp_doorbell);
^
WARNING:SPACING: Unnecessary space before function pointer arguments
#852: FILE: drivers/net/qede/qede_eth_if.h:150:
+ int (*q_tx_stop) (struct ecore_dev *edev,
WARNING:SPACING: Unnecessary space before function pointer arguments
#855: FILE: drivers/net/qede/qede_eth_if.h:153:
+ int (*eth_cqe_completion) (struct ecore_dev *edev,
WARNING:SPACING: Unnecessary space before function pointer arguments
#859: FILE: drivers/net/qede/qede_eth_if.h:157:
+ int (*fastpath_stop) (struct ecore_dev *edev);
WARNING:SPACING: Unnecessary space before function pointer arguments
#861: FILE: drivers/net/qede/qede_eth_if.h:159:
+ void (*get_vport_stats) (struct ecore_dev *edev,
WARNING:SPACING: Unnecessary space before function pointer arguments
#864: FILE: drivers/net/qede/qede_eth_if.h:162:
+ int (*filter_config) (struct ecore_dev *edev,
CHECK:AVOID_EXTERNS: extern prototypes should be avoided in .h files
#872: FILE: drivers/net/qede/qede_eth_if.h:170:
+extern int qed_fill_eth_dev_info(struct ecore_dev *edev,
CHECK:SPACING: No space is necessary after a cast
#905: FILE: drivers/net/qede/qede_ethdev.c:17:
+ ecore_int_sp_dpc((osal_int_ptr_t) (p_hwfn));
CHECK:UNNECESSARY_PARENTHESES: Unnecessary parentheses around eth_dev->pci_dev->intr_handle
#916: FILE: drivers/net/qede/qede_ethdev.c:28:
+ if (rte_intr_enable(&(eth_dev->pci_dev->intr_handle)))
CHECK:BRACES: braces {} should be used on all arms of this statement
#1071: FILE: drivers/net/qede/qede_ethdev.c:183:
+ if (rte_eth_promiscuous_get(eth_dev->data->port_id) == 1)
[...]
+ else if (!qdev->non_configured_vlans) {
[...]
WARNING:MISSING_SPACE: break quoted strings at a space character
#1097: FILE: drivers/net/qede/qede_ethdev.c:209:
+ "unequal number of rx/tx queues"
+ "is not supported RX=%u TX=%u\n",
CHECK:SPACING: No space is necessary after a cast
#1151: FILE: drivers/net/qede/qede_ethdev.c:263:
+ dev_info->min_rx_bufsize = (uint32_t) (ETHER_MIN_MTU +
CHECK:SPACING: No space is necessary after a cast
#1153: FILE: drivers/net/qede/qede_ethdev.c:265:
+ dev_info->max_rx_pktlen = (uint32_t) ETH_TX_MAX_NON_LSO_PKT_LEN;
CHECK:SPACING: No space is necessary after a cast
#1156: FILE: drivers/net/qede/qede_ethdev.c:268:
+ dev_info->max_rx_queues = (uint16_t) QEDE_MAX_RSS_CNT(qdev);
CHECK:SPACING: No space is necessary after a cast
#1157: FILE: drivers/net/qede/qede_ethdev.c:269:
+ dev_info->max_tx_queues = (uint16_t) QEDE_MAX_TSS_CNT(qdev);
CHECK:SPACING: No space is necessary after a cast
#1158: FILE: drivers/net/qede/qede_ethdev.c:270:
+ dev_info->max_mac_addrs = (uint32_t) (RESC_NUM(&edev->hwfns[0],
CHECK:SPACING: No space is necessary after a cast
#1160: FILE: drivers/net/qede/qede_ethdev.c:272:
+ dev_info->max_vfs = (uint16_t) NUM_OF_VFS(&qdev->edev);
CHECK:SPACING: No space is necessary after a cast
#1163: FILE: drivers/net/qede/qede_ethdev.c:275:
+ dev_info->flow_type_rss_offloads = (uint64_t) QEDE_RSS_OFFLOAD_ALL;
CHECK:UNNECESSARY_PARENTHESES: Unnecessary parentheses around eth_dev->pci_dev->intr_handle
#1299: FILE: drivers/net/qede/qede_ethdev.c:411:
+ rte_intr_disable(&(eth_dev->pci_dev->intr_handle));
CHECK:UNNECESSARY_PARENTHESES: Unnecessary parentheses around eth_dev->pci_dev->intr_handle
#1301: FILE: drivers/net/qede/qede_ethdev.c:413:
+ rte_intr_callback_unregister(&(eth_dev->pci_dev->intr_handle),
CHECK:UNNECESSARY_PARENTHESES: Unnecessary parentheses around eth_dev->pci_dev->intr_handle
#1511: FILE: drivers/net/qede/qede_ethdev.c:623:
+ rte_intr_callback_register(&(eth_dev->pci_dev->intr_handle),
CHECK:UNNECESSARY_PARENTHESES: Unnecessary parentheses around eth_dev->pci_dev->intr_handle
#1514: FILE: drivers/net/qede/qede_ethdev.c:626:
+ if (rte_intr_enable(&(eth_dev->pci_dev->intr_handle))) {
WARNING:SPACING: Unnecessary space before function pointer arguments
#1948: FILE: drivers/net/qede/qede_if.h:95:
+ void (*link_update) (void *dev, struct qed_link_output *link);
WARNING:SPACING: Unnecessary space before function pointer arguments
#1959: FILE: drivers/net/qede/qede_if.h:106:
+ int (*registers) (struct ecore_dev *edev);
WARNING:SPACING: Unnecessary space before function pointer arguments
#1963: FILE: drivers/net/qede/qede_if.h:110:
+ int (*probe) (struct ecore_dev *edev,
WARNING:SPACING: Unnecessary space before function pointer arguments
#1967: FILE: drivers/net/qede/qede_if.h:114:
+ void (*set_id) (struct ecore_dev *edev,
WARNING:SPACING: Unnecessary space before function pointer arguments
#1969: FILE: drivers/net/qede/qede_if.h:116:
+ enum _ecore_status_t (*chain_alloc) (struct ecore_dev *edev,
WARNING:SPACING: Unnecessary space before function pointer arguments
#1978: FILE: drivers/net/qede/qede_if.h:125:
+ void (*chain_free) (struct ecore_dev *edev,
WARNING:SPACING: Unnecessary space before function pointer arguments
#1981: FILE: drivers/net/qede/qede_if.h:128:
+ void (*get_link) (struct ecore_dev *edev,
WARNING:SPACING: Unnecessary space before function pointer arguments
#1983: FILE: drivers/net/qede/qede_if.h:130:
+ int (*set_link) (struct ecore_dev *edev,
WARNING:SPACING: Unnecessary space before function pointer arguments
#1986: FILE: drivers/net/qede/qede_if.h:133:
+ int (*drain) (struct ecore_dev *edev);
WARNING:SPACING: Unnecessary space before function pointer arguments
#1988: FILE: drivers/net/qede/qede_if.h:135:
+ void (*remove) (struct ecore_dev *edev);
WARNING:SPACING: Unnecessary space before function pointer arguments
#1990: FILE: drivers/net/qede/qede_if.h:137:
+ int (*slowpath_stop) (struct ecore_dev *edev);
WARNING:SPACING: Unnecessary space before function pointer arguments
#1992: FILE: drivers/net/qede/qede_if.h:139:
+ void (*update_pf_params) (struct ecore_dev *edev,
WARNING:SPACING: Unnecessary space before function pointer arguments
#1995: FILE: drivers/net/qede/qede_if.h:142:
+ int (*slowpath_start) (struct ecore_dev *edev,
WARNING:SPACING: Unnecessary space before function pointer arguments
#1998: FILE: drivers/net/qede/qede_if.h:145:
+ int (*set_fp_int) (struct ecore_dev *edev, uint16_t cnt);
WARNING:SPACING: missing space after return type
#2000: FILE: drivers/net/qede/qede_if.h:147:
+ uint32_t(*sb_init) (struct ecore_dev *edev,
WARNING:SPACING: Unnecessary space before function pointer arguments
#2000: FILE: drivers/net/qede/qede_if.h:147:
+ uint32_t(*sb_init) (struct ecore_dev *edev,
WARNING:SPACING: missing space after return type
#2006: FILE: drivers/net/qede/qede_if.h:153:
+ bool(*can_link_change) (struct ecore_dev *edev);
WARNING:SPACING: Unnecessary space before function pointer arguments
#2006: FILE: drivers/net/qede/qede_if.h:153:
+ bool(*can_link_change) (struct ecore_dev *edev);
WARNING:SPACING: Unnecessary space before function pointer arguments
#2007: FILE: drivers/net/qede/qede_if.h:154:
+ void (*update_msglvl) (struct ecore_dev *edev,
WARNING:TRAILING_SEMICOLON: macros should not use a trailing semicolon
#2038: FILE: drivers/net/qede/qede_logs.h:12:
+#define DP_ERR(p_dev, fmt, ...) \
+ rte_log(RTE_LOG_ERR, RTE_LOGTYPE_PMD, \
+ "[%s:%d(%s)]" fmt, \
+ __func__, __LINE__, \
+ (p_dev)->name ? (p_dev)->name : "", \
+ ##__VA_ARGS__);
WARNING:TRAILING_SEMICOLON: macros should not use a trailing semicolon
#2102: FILE: drivers/net/qede/qede_logs.h:76:
+#define PMD_RX_LOG(level, q, fmt, args...) \
+ RTE_LOG(level, PMD, "%s(): port=%u queue=%u " fmt "\n", \
+ __func__, q->port_id, q->queue_id, ## args);
WARNING:SUSPECT_CODE_INDENT: suspect code indent for conditional statements (16, 20)
#2469: FILE: drivers/net/qede/qede_main.c:344:
+ for_each_hwfn(edev, i)
+ info->num_queues +=
CHECK:SPACING: No space is necessary after a cast
#2484: FILE: drivers/net/qede/qede_main.c:359:
+ (uint8_t *) &info->port_mac);
WARNING:SUSPECT_CODE_INDENT: suspect code indent for conditional statements (8, 12)
#2582: FILE: drivers/net/qede/qede_main.c:457:
+ for_each_hwfn(cdev, i)
+ qed_inform_vf_link_state(&cdev->hwfns[i]);
CHECK:BRACES: Blank lines aren't necessary before a close brace '}'
#2586: FILE: drivers/net/qede/qede_main.c:461:
+
+}
WARNING:MISSING_SPACE: break quoted strings at a space character
#2753: FILE: drivers/net/qede/qede_rxtx.c:35:
+ "Failed to allocate rx buffer"
+ "sw_rx_prod %u sw_rx_cons %u mp entries %u free %u",
WARNING:CONSTANT_COMPARISON: Comparisons should place the constant on the right side of the test
#2776: FILE: drivers/net/qede/qede_rxtx.c:58:
+ if (NULL != rxq->sw_rx_ring[i].mbuf) {
CHECK:SPACING: No space is necessary after a cast
#2817: FILE: drivers/net/qede/qede_rxtx.c:99:
+ uint16_t pkt_len = (uint16_t) dev->data->dev_conf.rxmode.max_rx_pkt_len;
WARNING:MISSING_SPACE: break quoted strings at a space character
#3199: FILE: drivers/net/qede/qede_rxtx.c:481:
+ "Failed to allocate memory for all of"
+ "RSS queues\n"
CHECK:SPACING: No space is necessary after a cast
#3228: FILE: drivers/net/qede/qede_rxtx.c:510:
+ (uint32_t *) &rx_prods);
WARNING:UNNECESSARY_ELSE: else is not generally useful after a break or return
#3275: FILE: drivers/net/qede/qede_rxtx.c:557:
+ return -EINVAL;
+ } else {
WARNING:MISSING_SPACE: break quoted strings at a space character
#3325: FILE: drivers/net/qede/qede_rxtx.c:607:
+ "Cannot update V-VPORT as active as"
+ "there are no Rx queues\n");
CHECK:SPACING: No space is necessary after a cast
#3587: FILE: drivers/net/qede/qede_rxtx.c:869:
+ (struct eth_slow_path_rx_cqe *) cqe);
CHECK:REDUNDANT_CODE: if this code is redundant consider removing it
#3601: FILE: drivers/net/qede/qede_rxtx.c:883:
+#if 0
WARNING:MISSING_SPACE: break quoted strings at a space character
#3617: FILE: drivers/net/qede/qede_rxtx.c:899:
+ "CQE in CONS = %u has error, flags = 0x%x"
+ "dropping incoming packet\n",
WARNING:MISSING_SPACE: break quoted strings at a space character
#3688: FILE: drivers/net/qede/qede_rxtx.c:970:
+ "null mbuf nb_tx_desc %u nb_tx_avail %u"
+ "sw_tx_cons %u sw_tx_prod %u\n",
WARNING:UNNECESSARY_ELSE: else is not generally useful after a break or return
#3893: FILE: drivers/net/qede/qede_rxtx.c:1175:
+ return qede_drain_txq(qdev, txq, false);
+ } else {
CHECK:SPACING: No space is necessary after a cast
#4113: FILE: drivers/net/qede/qede_rxtx.h:43:
+ ((uint64_t) ((mb)->buf_physaddr + (mb)->data_off))
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [dpdk-dev] [PATCH 0/6] DPDK PMD for new QLogic FastLinQ QL4xxxx 25G/40G CNAs
2016-02-20 20:49 ` [dpdk-dev] [PATCH 0/6] DPDK PMD for new QLogic FastLinQ QL4xxxx 25G/40G CNAs Thomas Monjalon
@ 2016-02-22 16:47 ` Harish Patil
2016-02-22 17:35 ` Thomas Monjalon
0 siblings, 1 reply; 25+ messages in thread
From: Harish Patil @ 2016-02-22 16:47 UTC (permalink / raw)
To: Thomas Monjalon; +Cc: dev
>
>2016-02-20 07:40, Harish Patil:
>> This patch set introduces DPDK poll mode driver for new QLogic FastLinQ
>>QL4xxxx
>> 25G/40G capable family of CNAs as well as their SR-IOV Virtual
>>Functions (VF).
>>
>> The overall PMD driver design includes a common module called ecore
>>that deals
>> with the low level HW and a upper layer portion that provides the glue
>>logic.
>>
>> Specifically, the ecore module contains all of the common logic,
>> e.g. initialization, cleanup, infrastructure for interrupt handling,
>>link
>> management, slowpath etc. as well as protocol agnostic features and
>>supplying
>> an abstraction layer for other modules.
>>
>> The higher layer implements DPDK exported APIs/driver entry points by
>> interfacing with the common module for configuration/status and also the
>> fastpath routines.
>
>A new driver is always a good news :)
Thank you!
>Please could you share some performance numbers?
We have measured ~68 MPPS @ 64B with zero drop for the 4x25G adapter
running bi-di RFC traffic.
>
>What is the status about the integration of this driver in other
>environments?
Not very sure what you mean here. If you are asking about linux, then
qede/qed kernel drivers are part of kernel.org.
If its about BSD then we have compile tested the current poll mode driver
under FreeBSD (64 bit).
>
>The layer named ecore seems to be what is named a base driver in other
>DPDK drivers. Maybe you could consider renaming the directory to base/ for
>consistency?
Sure
>
>About the format of the patches, I think it's better to split driver
>imports
>in several pieces to ease reviewing and later reference for specific bugs.
>What do you think of introducing basic features one by one?
>
I don’t know how easy its going to be to split the driver on a feature by
feature basis.
Some of them could be coupled/dependent on other sections of the L2 code.
But yes, we shall think thru’ this and get back to you.
By the way, the patch 4/6 is still held up (not posted) due to size
restrictions.
Could you please allow it to be posted?
Thanks,
Harish
________________________________
This message and any attached documents contain information from the sending company or its parent company(s), subsidiaries, divisions or branch offices that may be confidential. If you are not the intended recipient, you may not read, copy, distribute, or use this information. If you have received this transmission in error, please notify the sender immediately by reply e-mail and then delete this message.
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [dpdk-dev] [PATCH 2/6] qede: add documentation
2016-02-20 15:40 ` [dpdk-dev] [PATCH 2/6] qede: add documentation Harish Patil
@ 2016-02-22 16:52 ` Mcnamara, John
2016-02-24 7:17 ` Harish Patil
2016-02-22 17:38 ` Mcnamara, John
1 sibling, 1 reply; 25+ messages in thread
From: Mcnamara, John @ 2016-02-22 16:52 UTC (permalink / raw)
To: Harish Patil, dev; +Cc: Sony Chacko
> -----Original Message-----
> From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Harish Patil
> Sent: Saturday, February 20, 2016 3:40 PM
> To: dev@dpdk.org
> Cc: Sony Chacko <sony.chacko@qlogic.com>
> Subject: [dpdk-dev] [PATCH 2/6] qede: add documentation
>
> Signed-off-by: Harish Patil <harish.patil@qlogic.com>
> Signed-off-by: Rasesh Mody <rasesh.mody@qlogic.com>
> Signed-off-by: Sony Chacko <sony.chacko@qlogic.com>
> ---
> doc/guides/nics/index.rst | 1 +
> doc/guides/nics/qede.rst | 344
Hi,
Thanks for the docs. The overall format and content seem good. A few
comments below.
> +#. Bind the QLogic 579xx adapters to ``igb_uio`` or ``vfio-pci`` loaded
> in the
> + previous step::
> +
> + ./tools/dpdk_nic_bind.py --bind igb_uio 0000:84:00.0 0000:84:00.1
> + 0000:84:00.2 0000:84:00.3
Fixed width lines should be constrained to 80 characters or else they will
go off the page in the PDF docs. The usual workaround is to use a command-
line continuation (or text wrap). For example:
./tools/dpdk_nic_bind.py --bind igb_uio \
0000:84:00.0 0000:84:00.1 \
0000:84:00.2 0000:84:00.3
Or similar. This also applies to the command-lines in other sections.
> +
> +#. Start ``testpmd`` with basic parameters:
> +
> + .. code-block:: console
> +
> + testpmd -c 0xf8000 -n 4 -- -i --nb-cores=4 --portmask=0xf
> + --rxd=4096 --txd=4096 --txfreet=4068 --rxq=2 --txq=2 --rss-ip
> + --rss-udp
Same comment about using continuations.
> +
> + [...]
> +
> + EAL: PCI device 0000:84:00.0 on NUMA socket 1
> + EAL: probe driver: 1077:1634 rte_qede_pmd
Align the text in this section to the same level of indentation.
> + EAL: Not managed by a supported kernel driver, skipped
> + EAL: PCI device 0000:84:00.1 on NUMA socket 1
> + EAL: probe driver: 1077:1634 rte_qede_pmd
> + EAL: Not managed by a supported kernel driver, skipped
> + EAL: PCI device 0000:88:00.0 on NUMA socket 1
> + EAL: probe driver: 1077:1656 rte_qede_pmd
> + EAL: PCI memory mapped at 0x7f738b200000
> + EAL: PCI memory mapped at 0x7f738b280000
> + EAL: PCI memory mapped at 0x7f738b300000
> + [QEDE PMD: (88:00.0:dpdk-port-0)]qed_load_firmware_data: Loading the
> firmware file /lib/firmware/qed/qed_init_values_zipped.bin...
Again, try wrap the code/console section at 80 chars in some way that still
maintains the meaning. Maybe something like the following (with a note to
say that the text has been wrapped for clarity):
[QEDE PMD: (88:00.0:dpdk-port-0)]qed_load_firmware_data:
Loading the firmware file
/lib/firmware/qed/qed_init_values_zipped.bin...
[QEDE PMD: (88:00.0:dpdk-port-0)]
qede_print_adapter_info:Chip details - BB1
[QEDE PMD: (88:00.0:dpdk-port-0)]
qede_print_adapter_info:Driver version:QEDE PMD 8.7.9.0_1.0.0
[QEDE PMD: (88:00.0:dpdk-port-0)]
qede_print_adapter_info:Firmware version:8.7.7.0
You can test the PDF output as follows:
make -j doc-guides-pdf
pdf_veiwer build/doc/pdf/guides/nics.pdf
John
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [dpdk-dev] [PATCH 0/6] DPDK PMD for new QLogic FastLinQ QL4xxxx 25G/40G CNAs
2016-02-22 16:47 ` Harish Patil
@ 2016-02-22 17:35 ` Thomas Monjalon
2016-02-23 18:13 ` Harish Patil
0 siblings, 1 reply; 25+ messages in thread
From: Thomas Monjalon @ 2016-02-22 17:35 UTC (permalink / raw)
To: Harish Patil; +Cc: dev
2016-02-22 16:47, Harish Patil:
> >Please could you share some performance numbers?
>
> We have measured ~68 MPPS @ 64B with zero drop for the 4x25G adapter
> running bi-di RFC traffic.
How many queues/cores to achieve this performance?
> >What is the status about the integration of this driver in other
> >environments?
>
> Not very sure what you mean here. If you are asking about linux, then
> qede/qed kernel drivers are part of kernel.org.
> If its about BSD then we have compile tested the current poll mode driver
> under FreeBSD (64 bit).
OK, so the base driver is shared only between Linux and DPDK?
Nothing in BSD kernels?
> By the way, the patch 4/6 is still held up (not posted) due to size
> restrictions.
> Could you please allow it to be posted?
It was posted but curiously out of the thread:
http://dpdk.org/ml/archives/dev/2016-February/033612.html
> ________________________________
>
> This message and any attached documents contain information from the sending company or its parent company(s), subsidiaries, divisions or branch offices that may be confidential. If you are not the intended recipient, you may not read, copy, distribute, or use this information. If you have received this transmission in error, please notify the sender immediately by reply e-mail and then delete this message.
Please avoid this message --^
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [dpdk-dev] [PATCH 2/6] qede: add documentation
2016-02-20 15:40 ` [dpdk-dev] [PATCH 2/6] qede: add documentation Harish Patil
2016-02-22 16:52 ` Mcnamara, John
@ 2016-02-22 17:38 ` Mcnamara, John
1 sibling, 0 replies; 25+ messages in thread
From: Mcnamara, John @ 2016-02-22 17:38 UTC (permalink / raw)
To: Harish Patil, dev; +Cc: Sony Chacko
> +
> + Assign MAC address to the VF using iproute2 utility. The syntax is:
> + ip link set <PF iface> vf <VF id> mac <macaddr>
Also, make this a code/console section by adding :: to the previous line
and indenting 3-4 spaces.
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [dpdk-dev] [PATCH 3/6] qede: add QLogic PCI ids
2016-02-21 1:17 ` Stephen Hemminger
@ 2016-02-22 23:23 ` Harish Patil
0 siblings, 0 replies; 25+ messages in thread
From: Harish Patil @ 2016-02-22 23:23 UTC (permalink / raw)
To: Stephen Hemminger; +Cc: dev, Sony Chacko
>
>On Sat, 20 Feb 2016 07:40:28 -0800
>Harish Patil <harish.patil@qlogic.com> wrote:
>
>> diff --git a/lib/librte_eal/common/include/rte_pci_dev_ids.h
>>b/lib/librte_eal/common/include/rte_pci_dev_ids.h
>> index d088191..0c1a3fe 100644
>> --- a/lib/librte_eal/common/include/rte_pci_dev_ids.h
>> +++ b/lib/librte_eal/common/include/rte_pci_dev_ids.h
>> @@ -152,6 +152,14 @@
>> #define RTE_PCI_DEV_ID_DECL_BNX2XVF(vend, dev)
>> #endif
>>
>> +#ifndef RTE_PCI_DEV_ID_DECL_QEDE
>> +#define RTE_PCI_DEV_ID_DECL_QEDE(vend, dev)
>> +#endif
>> +
>> +#ifndef RTE_PCI_DEV_ID_DECL_QEDEVF
>> +#define RTE_PCI_DEV_ID_DECL_QEDEVF(vend, dev)
>> +#endif
>> +
>> #ifndef PCI_VENDOR_ID_INTEL
>> /** Vendor ID used by Intel devices */
>> #define PCI_VENDOR_ID_INTEL 0x8086
>> @@ -177,6 +185,11 @@
>> #define PCI_VENDOR_ID_BROADCOM 0x14E4
>> #endif
>>
>> +#ifndef PCI_VENDOR_ID_QLOGIC
>> +/** Vendor ID used by QLogic devices */
>> +#define PCI_VENDOR_ID_QLOGIC 0x1077
>> +#endif
>> +
>
>The goal is get away from having every PCI id for every device
>stuffed into rte_pci_dev_ids.h. So it would be good not to add
>more now.
>
Okay, I shall make the PCI ids local to this driver.
Thanks,
Harish
________________________________
This message and any attached documents contain information from the sending company or its parent company(s), subsidiaries, divisions or branch offices that may be confidential. If you are not the intended recipient, you may not read, copy, distribute, or use this information. If you have received this transmission in error, please notify the sender immediately by reply e-mail and then delete this message.
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [dpdk-dev] [PATCH 5/6] qede: add driver
2016-02-21 1:26 ` Stephen Hemminger
@ 2016-02-23 2:28 ` Harish Patil
2016-02-23 5:30 ` Stephen Hemminger
2016-02-23 5:33 ` Stephen Hemminger
0 siblings, 2 replies; 25+ messages in thread
From: Harish Patil @ 2016-02-23 2:28 UTC (permalink / raw)
To: Stephen Hemminger; +Cc: dev, Sony Chacko
>
>On Sat, 20 Feb 2016 07:40:30 -0800
>Harish Patil <harish.patil@qlogic.com> wrote:
>
>> Signed-off-by: Harish Patil <harish.patil@qlogic.com>
>> Signed-off-by: Rasesh Mody <rasesh.mody@qlogic.com>
>> Signed-off-by: Sony Chacko <sony.chacko@qlogic.com>
>> ---
>> drivers/net/qede/LICENSE.qede_pmd | 28 +
>> drivers/net/qede/Makefile | 95 ++
>> drivers/net/qede/qede_eth_if.c | 461 ++++++++++
>> drivers/net/qede/qede_eth_if.h | 180 ++++
>> drivers/net/qede/qede_ethdev.c | 815 +++++++++++++++++
>> drivers/net/qede/qede_ethdev.h | 138 +++
>> drivers/net/qede/qede_if.h | 167 ++++
>> drivers/net/qede/qede_logs.h | 93 ++
>> drivers/net/qede/qede_main.c | 587 +++++++++++++
>> drivers/net/qede/qede_rxtx.c | 1346
>>+++++++++++++++++++++++++++++
>> drivers/net/qede/qede_rxtx.h | 183 ++++
>> drivers/net/qede/rte_pmd_qede_version.map | 4 +
>> 12 files changed, 4097 insertions(+)
>> create mode 100644 drivers/net/qede/LICENSE.qede_pmd
>> create mode 100644 drivers/net/qede/Makefile
>> create mode 100644 drivers/net/qede/qede_eth_if.c
>> create mode 100644 drivers/net/qede/qede_eth_if.h
>> create mode 100644 drivers/net/qede/qede_ethdev.c
>> create mode 100644 drivers/net/qede/qede_ethdev.h
>> create mode 100644 drivers/net/qede/qede_if.h
>> create mode 100644 drivers/net/qede/qede_logs.h
>> create mode 100644 drivers/net/qede/qede_main.c
>> create mode 100644 drivers/net/qede/qede_rxtx.c
>> create mode 100644 drivers/net/qede/qede_rxtx.h
>> create mode 100644 drivers/net/qede/rte_pmd_qede_version.map
>
>DPDK doesn't follow all the kernel style rules, but lots of them.
>The biggest catch was the ; in the LOG macros.
>Minor stuff about spacing around casts etc.
>
>
>Running checkpatch with --ignore
>PREFER_KERNEL_TYPES,LINE_SPACING,PARENTHESIS_ALIGNMENT,BIT_MACRO,NETWORKIN
>G_BLOCK_COMMENT_STYLE,SPLIT_STRING,COMPARISON_TO_NULL,CONCATENATED_STRING
>
>Produces:
>
>
>CHECK:SPACING: spaces preferred around that '*' (ctx:VxO)
>#410: FILE: drivers/net/qede/qede_eth_if.c:175:
>+ uint16_t cqe_pbl_size, void OSAL_IOMEM**pp_prod)
> ^
>
>CHECK:SPACING: spaces preferred around that '*' (ctx:VxO)
>#467: FILE: drivers/net/qede/qede_eth_if.c:232:
>+ uint16_t pbl_size, void OSAL_IOMEM**pp_doorbell)
> ^
>
>CHECK:BRACES: Blank lines aren't necessary after an open brace '{'
>#519: FILE: drivers/net/qede/qede_eth_if.c:284:
>+{
>+
>
>WARNING:SPACING: Unnecessary space before function pointer arguments
>#823: FILE: drivers/net/qede/qede_eth_if.h:121:
>+ int (*fill_dev_info) (struct ecore_dev *edev,
>
>WARNING:SPACING: Unnecessary space before function pointer arguments
>#826: FILE: drivers/net/qede/qede_eth_if.h:124:
>+ int (*vport_start) (struct ecore_dev *edev,
>
>WARNING:SPACING: Unnecessary space before function pointer arguments
>#829: FILE: drivers/net/qede/qede_eth_if.h:127:
>+ int (*vport_stop) (struct ecore_dev *edev, uint8_t vport_id);
>
>WARNING:SPACING: Unnecessary space before function pointer arguments
>#831: FILE: drivers/net/qede/qede_eth_if.h:129:
>+ int (*vport_update) (struct ecore_dev *edev,
>
>WARNING:SPACING: Unnecessary space before function pointer arguments
>#834: FILE: drivers/net/qede/qede_eth_if.h:132:
>+ int (*q_rx_start) (struct ecore_dev *cdev,
>
>CHECK:SPACING: spaces preferred around that '*' (ctx:VxO)
>#840: FILE: drivers/net/qede/qede_eth_if.h:138:
>+ uint16_t cqe_pbl_size, void OSAL_IOMEM**pp_prod);
> ^
>
>WARNING:SPACING: Unnecessary space before function pointer arguments
>#842: FILE: drivers/net/qede/qede_eth_if.h:140:
>+ int (*q_rx_stop) (struct ecore_dev *edev,
>
>WARNING:SPACING: Unnecessary space before function pointer arguments
>#845: FILE: drivers/net/qede/qede_eth_if.h:143:
>+ int (*q_tx_start) (struct ecore_dev *edev,
>
>CHECK:SPACING: spaces preferred around that '*' (ctx:VxO)
>#850: FILE: drivers/net/qede/qede_eth_if.h:148:
>+ uint16_t pbl_size, void OSAL_IOMEM**pp_doorbell);
> ^
>
>WARNING:SPACING: Unnecessary space before function pointer arguments
>#852: FILE: drivers/net/qede/qede_eth_if.h:150:
>+ int (*q_tx_stop) (struct ecore_dev *edev,
>
>WARNING:SPACING: Unnecessary space before function pointer arguments
>#855: FILE: drivers/net/qede/qede_eth_if.h:153:
>+ int (*eth_cqe_completion) (struct ecore_dev *edev,
>
>WARNING:SPACING: Unnecessary space before function pointer arguments
>#859: FILE: drivers/net/qede/qede_eth_if.h:157:
>+ int (*fastpath_stop) (struct ecore_dev *edev);
>
>WARNING:SPACING: Unnecessary space before function pointer arguments
>#861: FILE: drivers/net/qede/qede_eth_if.h:159:
>+ void (*get_vport_stats) (struct ecore_dev *edev,
>
>WARNING:SPACING: Unnecessary space before function pointer arguments
>#864: FILE: drivers/net/qede/qede_eth_if.h:162:
>+ int (*filter_config) (struct ecore_dev *edev,
>
>CHECK:AVOID_EXTERNS: extern prototypes should be avoided in .h files
>#872: FILE: drivers/net/qede/qede_eth_if.h:170:
>+extern int qed_fill_eth_dev_info(struct ecore_dev *edev,
>
>CHECK:SPACING: No space is necessary after a cast
>#905: FILE: drivers/net/qede/qede_ethdev.c:17:
>+ ecore_int_sp_dpc((osal_int_ptr_t) (p_hwfn));
>
>CHECK:UNNECESSARY_PARENTHESES: Unnecessary parentheses around
>eth_dev->pci_dev->intr_handle
>#916: FILE: drivers/net/qede/qede_ethdev.c:28:
>+ if (rte_intr_enable(&(eth_dev->pci_dev->intr_handle)))
>
>CHECK:BRACES: braces {} should be used on all arms of this statement
>#1071: FILE: drivers/net/qede/qede_ethdev.c:183:
>+ if (rte_eth_promiscuous_get(eth_dev->data->port_id) == 1)
>[...]
>+ else if (!qdev->non_configured_vlans) {
>[...]
>
>WARNING:MISSING_SPACE: break quoted strings at a space character
>#1097: FILE: drivers/net/qede/qede_ethdev.c:209:
>+ "unequal number of rx/tx queues"
>+ "is not supported RX=%u TX=%u\n",
>
>CHECK:SPACING: No space is necessary after a cast
>#1151: FILE: drivers/net/qede/qede_ethdev.c:263:
>+ dev_info->min_rx_bufsize = (uint32_t) (ETHER_MIN_MTU +
>
>CHECK:SPACING: No space is necessary after a cast
>#1153: FILE: drivers/net/qede/qede_ethdev.c:265:
>+ dev_info->max_rx_pktlen = (uint32_t) ETH_TX_MAX_NON_LSO_PKT_LEN;
>
>CHECK:SPACING: No space is necessary after a cast
>#1156: FILE: drivers/net/qede/qede_ethdev.c:268:
>+ dev_info->max_rx_queues = (uint16_t) QEDE_MAX_RSS_CNT(qdev);
>
>CHECK:SPACING: No space is necessary after a cast
>#1157: FILE: drivers/net/qede/qede_ethdev.c:269:
>+ dev_info->max_tx_queues = (uint16_t) QEDE_MAX_TSS_CNT(qdev);
>
>CHECK:SPACING: No space is necessary after a cast
>#1158: FILE: drivers/net/qede/qede_ethdev.c:270:
>+ dev_info->max_mac_addrs = (uint32_t) (RESC_NUM(&edev->hwfns[0],
>
>CHECK:SPACING: No space is necessary after a cast
>#1160: FILE: drivers/net/qede/qede_ethdev.c:272:
>+ dev_info->max_vfs = (uint16_t) NUM_OF_VFS(&qdev->edev);
>
>CHECK:SPACING: No space is necessary after a cast
>#1163: FILE: drivers/net/qede/qede_ethdev.c:275:
>+ dev_info->flow_type_rss_offloads = (uint64_t) QEDE_RSS_OFFLOAD_ALL;
>
>CHECK:UNNECESSARY_PARENTHESES: Unnecessary parentheses around
>eth_dev->pci_dev->intr_handle
>#1299: FILE: drivers/net/qede/qede_ethdev.c:411:
>+ rte_intr_disable(&(eth_dev->pci_dev->intr_handle));
>
>CHECK:UNNECESSARY_PARENTHESES: Unnecessary parentheses around
>eth_dev->pci_dev->intr_handle
>#1301: FILE: drivers/net/qede/qede_ethdev.c:413:
>+ rte_intr_callback_unregister(&(eth_dev->pci_dev->intr_handle),
>
>CHECK:UNNECESSARY_PARENTHESES: Unnecessary parentheses around
>eth_dev->pci_dev->intr_handle
>#1511: FILE: drivers/net/qede/qede_ethdev.c:623:
>+ rte_intr_callback_register(&(eth_dev->pci_dev->intr_handle),
>
>CHECK:UNNECESSARY_PARENTHESES: Unnecessary parentheses around
>eth_dev->pci_dev->intr_handle
>#1514: FILE: drivers/net/qede/qede_ethdev.c:626:
>+ if (rte_intr_enable(&(eth_dev->pci_dev->intr_handle))) {
>
>WARNING:SPACING: Unnecessary space before function pointer arguments
>#1948: FILE: drivers/net/qede/qede_if.h:95:
>+ void (*link_update) (void *dev, struct qed_link_output *link);
>
>WARNING:SPACING: Unnecessary space before function pointer arguments
>#1959: FILE: drivers/net/qede/qede_if.h:106:
>+ int (*registers) (struct ecore_dev *edev);
>
>WARNING:SPACING: Unnecessary space before function pointer arguments
>#1963: FILE: drivers/net/qede/qede_if.h:110:
>+ int (*probe) (struct ecore_dev *edev,
>
>WARNING:SPACING: Unnecessary space before function pointer arguments
>#1967: FILE: drivers/net/qede/qede_if.h:114:
>+ void (*set_id) (struct ecore_dev *edev,
>
>WARNING:SPACING: Unnecessary space before function pointer arguments
>#1969: FILE: drivers/net/qede/qede_if.h:116:
>+ enum _ecore_status_t (*chain_alloc) (struct ecore_dev *edev,
>
>WARNING:SPACING: Unnecessary space before function pointer arguments
>#1978: FILE: drivers/net/qede/qede_if.h:125:
>+ void (*chain_free) (struct ecore_dev *edev,
>
>WARNING:SPACING: Unnecessary space before function pointer arguments
>#1981: FILE: drivers/net/qede/qede_if.h:128:
>+ void (*get_link) (struct ecore_dev *edev,
>
>WARNING:SPACING: Unnecessary space before function pointer arguments
>#1983: FILE: drivers/net/qede/qede_if.h:130:
>+ int (*set_link) (struct ecore_dev *edev,
>
>WARNING:SPACING: Unnecessary space before function pointer arguments
>#1986: FILE: drivers/net/qede/qede_if.h:133:
>+ int (*drain) (struct ecore_dev *edev);
>
>WARNING:SPACING: Unnecessary space before function pointer arguments
>#1988: FILE: drivers/net/qede/qede_if.h:135:
>+ void (*remove) (struct ecore_dev *edev);
>
>WARNING:SPACING: Unnecessary space before function pointer arguments
>#1990: FILE: drivers/net/qede/qede_if.h:137:
>+ int (*slowpath_stop) (struct ecore_dev *edev);
>
>WARNING:SPACING: Unnecessary space before function pointer arguments
>#1992: FILE: drivers/net/qede/qede_if.h:139:
>+ void (*update_pf_params) (struct ecore_dev *edev,
>
>WARNING:SPACING: Unnecessary space before function pointer arguments
>#1995: FILE: drivers/net/qede/qede_if.h:142:
>+ int (*slowpath_start) (struct ecore_dev *edev,
>
>WARNING:SPACING: Unnecessary space before function pointer arguments
>#1998: FILE: drivers/net/qede/qede_if.h:145:
>+ int (*set_fp_int) (struct ecore_dev *edev, uint16_t cnt);
>
>WARNING:SPACING: missing space after return type
>#2000: FILE: drivers/net/qede/qede_if.h:147:
>+ uint32_t(*sb_init) (struct ecore_dev *edev,
>
>WARNING:SPACING: Unnecessary space before function pointer arguments
>#2000: FILE: drivers/net/qede/qede_if.h:147:
>+ uint32_t(*sb_init) (struct ecore_dev *edev,
>
>WARNING:SPACING: missing space after return type
>#2006: FILE: drivers/net/qede/qede_if.h:153:
>+ bool(*can_link_change) (struct ecore_dev *edev);
>
>WARNING:SPACING: Unnecessary space before function pointer arguments
>#2006: FILE: drivers/net/qede/qede_if.h:153:
>+ bool(*can_link_change) (struct ecore_dev *edev);
>
>WARNING:SPACING: Unnecessary space before function pointer arguments
>#2007: FILE: drivers/net/qede/qede_if.h:154:
>+ void (*update_msglvl) (struct ecore_dev *edev,
>
>WARNING:TRAILING_SEMICOLON: macros should not use a trailing semicolon
>#2038: FILE: drivers/net/qede/qede_logs.h:12:
>+#define DP_ERR(p_dev, fmt, ...) \
>+ rte_log(RTE_LOG_ERR, RTE_LOGTYPE_PMD, \
>+ "[%s:%d(%s)]" fmt, \
>+ __func__, __LINE__, \
>+ (p_dev)->name ? (p_dev)->name : "", \
>+ ##__VA_ARGS__);
>
>WARNING:TRAILING_SEMICOLON: macros should not use a trailing semicolon
>#2102: FILE: drivers/net/qede/qede_logs.h:76:
>+#define PMD_RX_LOG(level, q, fmt, args...) \
>+ RTE_LOG(level, PMD, "%s(): port=%u queue=%u " fmt "\n", \
>+ __func__, q->port_id, q->queue_id, ## args);
>
>WARNING:SUSPECT_CODE_INDENT: suspect code indent for conditional
>statements (16, 20)
>#2469: FILE: drivers/net/qede/qede_main.c:344:
>+ for_each_hwfn(edev, i)
>+ info->num_queues +=
>
>CHECK:SPACING: No space is necessary after a cast
>#2484: FILE: drivers/net/qede/qede_main.c:359:
>+ (uint8_t *) &info->port_mac);
>
>WARNING:SUSPECT_CODE_INDENT: suspect code indent for conditional
>statements (8, 12)
>#2582: FILE: drivers/net/qede/qede_main.c:457:
>+ for_each_hwfn(cdev, i)
>+ qed_inform_vf_link_state(&cdev->hwfns[i]);
>
>CHECK:BRACES: Blank lines aren't necessary before a close brace '}'
>#2586: FILE: drivers/net/qede/qede_main.c:461:
>+
>+}
>
>WARNING:MISSING_SPACE: break quoted strings at a space character
>#2753: FILE: drivers/net/qede/qede_rxtx.c:35:
>+ "Failed to allocate rx buffer"
>+ "sw_rx_prod %u sw_rx_cons %u mp entries %u free %u",
>
>WARNING:CONSTANT_COMPARISON: Comparisons should place the constant on the
>right side of the test
>#2776: FILE: drivers/net/qede/qede_rxtx.c:58:
>+ if (NULL != rxq->sw_rx_ring[i].mbuf) {
>
>CHECK:SPACING: No space is necessary after a cast
>#2817: FILE: drivers/net/qede/qede_rxtx.c:99:
>+ uint16_t pkt_len = (uint16_t) dev->data->dev_conf.rxmode.max_rx_pkt_len;
>
>WARNING:MISSING_SPACE: break quoted strings at a space character
>#3199: FILE: drivers/net/qede/qede_rxtx.c:481:
>+ "Failed to allocate memory for all of"
>+ "RSS queues\n"
>
>CHECK:SPACING: No space is necessary after a cast
>#3228: FILE: drivers/net/qede/qede_rxtx.c:510:
>+ (uint32_t *) &rx_prods);
>
>WARNING:UNNECESSARY_ELSE: else is not generally useful after a break or
>return
>#3275: FILE: drivers/net/qede/qede_rxtx.c:557:
>+ return -EINVAL;
>+ } else {
>
>WARNING:MISSING_SPACE: break quoted strings at a space character
>#3325: FILE: drivers/net/qede/qede_rxtx.c:607:
>+ "Cannot update V-VPORT as active as"
>+ "there are no Rx queues\n");
>
>CHECK:SPACING: No space is necessary after a cast
>#3587: FILE: drivers/net/qede/qede_rxtx.c:869:
>+ (struct eth_slow_path_rx_cqe *) cqe);
>
>CHECK:REDUNDANT_CODE: if this code is redundant consider removing it
>#3601: FILE: drivers/net/qede/qede_rxtx.c:883:
>+#if 0
>
>WARNING:MISSING_SPACE: break quoted strings at a space character
>#3617: FILE: drivers/net/qede/qede_rxtx.c:899:
>+ "CQE in CONS = %u has error, flags = 0x%x"
>+ "dropping incoming packet\n",
>
>WARNING:MISSING_SPACE: break quoted strings at a space character
>#3688: FILE: drivers/net/qede/qede_rxtx.c:970:
>+ "null mbuf nb_tx_desc %u nb_tx_avail %u"
>+ "sw_tx_cons %u sw_tx_prod %u\n",
>
>WARNING:UNNECESSARY_ELSE: else is not generally useful after a break or
>return
>#3893: FILE: drivers/net/qede/qede_rxtx.c:1175:
>+ return qede_drain_txq(qdev, txq, false);
>+ } else {
>
>CHECK:SPACING: No space is necessary after a cast
>#4113: FILE: drivers/net/qede/qede_rxtx.h:43:
>+ ((uint64_t) ((mb)->buf_physaddr + (mb)->data_off))
>
>
All of the checkpatch warnings had been fixed (except one which cannot be
fixed) using the checkpatch script available under DPDK scripts/
directory. The linux checkpatch version is 0.32.
[root@dut4019 dpdk]# ./scripts/checkpatches.sh patches-send/*.patch
### patches-send/0004-qede-add-driver-common-module.patch
WARNING:CAMELCASE: Avoid CamelCase: <inflateEnd>
#251: FILE: drivers/net/qede/ecore/bcm_osal.c:120:
+ inflateEnd(p_hwfn->stream);
total: 0 errors, 1 warnings, 54467 lines checked
Wouldn’t that suffice?
________________________________
This message and any attached documents contain information from the sending company or its parent company(s), subsidiaries, divisions or branch offices that may be confidential. If you are not the intended recipient, you may not read, copy, distribute, or use this information. If you have received this transmission in error, please notify the sender immediately by reply e-mail and then delete this message.
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [dpdk-dev] [PATCH 5/6] qede: add driver
2016-02-23 2:28 ` Harish Patil
@ 2016-02-23 5:30 ` Stephen Hemminger
2016-02-23 5:33 ` Stephen Hemminger
1 sibling, 0 replies; 25+ messages in thread
From: Stephen Hemminger @ 2016-02-23 5:30 UTC (permalink / raw)
To: Harish Patil; +Cc: dev, Sony Chacko
On Tue, 23 Feb 2016 02:28:25 +0000
Harish Patil <harish.patil@qlogic.com> wrote:
> All of the checkpatch warnings had been fixed (except one which cannot be
> fixed) using the checkpatch script available under DPDK scripts/
> directory. The linux checkpatch version is 0.32.
>
> [root@dut4019 dpdk]# ./scripts/checkpatches.sh patches-send/*.patch
>
> ### patches-send/0004-qede-add-driver-common-module.patch
>
> WARNING:CAMELCASE: Avoid CamelCase: <inflateEnd>
> #251: FILE: drivers/net/qede/ecore/bcm_osal.c:120:
> + inflateEnd(p_hwfn->stream);
>
> total: 0 errors, 1 warnings, 54467 lines checked
>
> Wouldn’t that suffice?
The problem is that the DPDK version is shutting off warnings about stuff
that is really incorrect and should be fixed.
Look at the warnings about spaces in strings, I don't know why the DPDK wrapper
disables that.
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [dpdk-dev] [PATCH 5/6] qede: add driver
2016-02-23 2:28 ` Harish Patil
2016-02-23 5:30 ` Stephen Hemminger
@ 2016-02-23 5:33 ` Stephen Hemminger
2016-02-23 19:04 ` Harish Patil
1 sibling, 1 reply; 25+ messages in thread
From: Stephen Hemminger @ 2016-02-23 5:33 UTC (permalink / raw)
To: Harish Patil; +Cc: dev, Sony Chacko
On Tue, 23 Feb 2016 02:28:25 +0000
Harish Patil <harish.patil@qlogic.com> wrote:
> All of the checkpatch warnings had been fixed (except one which cannot be
> fixed) using the checkpatch script available under DPDK scripts/
> directory. The linux checkpatch version is 0.32.
>
> [root@dut4019 dpdk]# ./scripts/checkpatches.sh patches-send/*.patch
>
> ### patches-send/0004-qede-add-driver-common-module.patch
>
> WARNING:CAMELCASE: Avoid CamelCase: <inflateEnd>
> #251: FILE: drivers/net/qede/ecore/bcm_osal.c:120:
> + inflateEnd(p_hwfn->stream);
>
> total: 0 errors, 1 warnings, 54467 lines checked
>
> Wouldn’t that suffice?
Maybe your version of checkpatch from kernel is older than more recent
version I am using (from Linux 4.4).
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [dpdk-dev] [PATCH 0/6] DPDK PMD for new QLogic FastLinQ QL4xxxx 25G/40G CNAs
2016-02-22 17:35 ` Thomas Monjalon
@ 2016-02-23 18:13 ` Harish Patil
0 siblings, 0 replies; 25+ messages in thread
From: Harish Patil @ 2016-02-23 18:13 UTC (permalink / raw)
To: Thomas Monjalon; +Cc: dev
>
>2016-02-22 16:47, Harish Patil:
>> >Please could you share some performance numbers?
>>
>> We have measured ~68 MPPS @ 64B with zero drop for the 4x25G adapter
>> running bi-di RFC traffic.
>
>How many queues/cores to achieve this performance?
Using single queue/core per port. Eg: Running L2FWD application in the
4x25G configuration shall use 4 cores with single queue pair per port.
>
>> >What is the status about the integration of this driver in other
>> >environments?
>>
>> Not very sure what you mean here. If you are asking about linux, then
>> qede/qed kernel drivers are part of kernel.org.
>> If its about BSD then we have compile tested the current poll mode
>>driver
>> under FreeBSD (64 bit).
>
>OK, so the base driver is shared only between Linux and DPDK?
>Nothing in BSD kernels?
BSD driver also uses the common base code.
>
>> By the way, the patch 4/6 is still held up (not posted) due to size
>> restrictions.
>> Could you please allow it to be posted?
>
>It was posted but curiously out of the thread:
> http://dpdk.org/ml/archives/dev/2016-February/033612.html
Ok thanks
>
>> ________________________________
>>
>> This message and any attached documents contain information from the
>>sending company or its parent company(s), subsidiaries, divisions or
>>branch offices that may be confidential. If you are not the intended
>>recipient, you may not read, copy, distribute, or use this information.
>>If you have received this transmission in error, please notify the
>>sender immediately by reply e-mail and then delete this message.
>
>Please avoid this message --^
>
I have got it fixed at the exchange level. Hopefully should be fixed with
this email onwards.
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [dpdk-dev] [PATCH 5/6] qede: add driver
2016-02-23 5:33 ` Stephen Hemminger
@ 2016-02-23 19:04 ` Harish Patil
2016-02-23 19:06 ` Harish Patil
0 siblings, 1 reply; 25+ messages in thread
From: Harish Patil @ 2016-02-23 19:04 UTC (permalink / raw)
To: Stephen Hemminger; +Cc: dev
>
>On Tue, 23 Feb 2016 02:28:25 +0000
>Harish Patil <harish.patil@qlogic.com> wrote:
>
>> All of the checkpatch warnings had been fixed (except one which cannot
>>be
>> fixed) using the checkpatch script available under DPDK scripts/
>> directory. The linux checkpatch version is 0.32.
>>
>> [root@dut4019 dpdk]# ./scripts/checkpatches.sh patches-send/*.patch
>>
>> ### patches-send/0004-qede-add-driver-common-module.patch
>>
>> WARNING:CAMELCASE: Avoid CamelCase: <inflateEnd>
>> #251: FILE: drivers/net/qede/ecore/bcm_osal.c:120:
>> + inflateEnd(p_hwfn->stream);
>>
>> total: 0 errors, 1 warnings, 54467 lines checked
>>
>> Wouldn’t that suffice?
>
>Maybe your version of checkpatch from kernel is older than more recent
>version I am using (from Linux 4.4).
>
I had used checkpatch version 0.32 as mentioned above.
The latest checkpatch version in net-next also shows 0.32.
my $V = '0.32’;
I see similar checkpatch related conversations going on in other threads
in the distribution.
Please advise.
+Thomas
Thanks,
Harish
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [dpdk-dev] [PATCH 5/6] qede: add driver
2016-02-23 19:04 ` Harish Patil
@ 2016-02-23 19:06 ` Harish Patil
0 siblings, 0 replies; 25+ messages in thread
From: Harish Patil @ 2016-02-23 19:06 UTC (permalink / raw)
To: Stephen Hemminger, Thomas Monjalon; +Cc: dev
>>
>>
>>
>>On Tue, 23 Feb 2016 02:28:25 +0000
>>Harish Patil <harish.patil@qlogic.com> wrote:
>>
>>> All of the checkpatch warnings had been fixed (except one which cannot
>>>be
>>> fixed) using the checkpatch script available under DPDK scripts/
>>> directory. The linux checkpatch version is 0.32.
>>>
>>> [root@dut4019 dpdk]# ./scripts/checkpatches.sh patches-send/*.patch
>>>
>>> ### patches-send/0004-qede-add-driver-common-module.patch
>>>
>>> WARNING:CAMELCASE: Avoid CamelCase: <inflateEnd>
>>> #251: FILE: drivers/net/qede/ecore/bcm_osal.c:120:
>>> + inflateEnd(p_hwfn->stream);
>>>
>>> total: 0 errors, 1 warnings, 54467 lines checked
>>>
>>> Wouldn’t that suffice?
>>
>>Maybe your version of checkpatch from kernel is older than more recent
>>version I am using (from Linux 4.4).
>>
>I had used checkpatch version 0.32 as mentioned above.
>The latest checkpatch version in net-next also shows 0.32.
>my $V = '0.32’;
>
>I see similar checkpatch related conversations going on in other threads
>in the distribution.
>Please advise.
>+Thomas
>
>Thanks,
>Harish
>
>
>
+Thomas
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [dpdk-dev] [PATCH 2/6] qede: add documentation
2016-02-22 16:52 ` Mcnamara, John
@ 2016-02-24 7:17 ` Harish Patil
2016-02-24 9:26 ` Mcnamara, John
0 siblings, 1 reply; 25+ messages in thread
From: Harish Patil @ 2016-02-24 7:17 UTC (permalink / raw)
To: Mcnamara, John, dev
>>
>>-----Original Message-----
>> From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Harish Patil
>> Sent: Saturday, February 20, 2016 3:40 PM
>> To: dev@dpdk.org
>> Cc: Sony Chacko <sony.chacko@qlogic.com>
>> Subject: [dpdk-dev] [PATCH 2/6] qede: add documentation
>>
>> Signed-off-by: Harish Patil <harish.patil@qlogic.com>
>> Signed-off-by: Rasesh Mody <rasesh.mody@qlogic.com>
>> Signed-off-by: Sony Chacko <sony.chacko@qlogic.com>
>> ---
>> doc/guides/nics/index.rst | 1 +
>> doc/guides/nics/qede.rst | 344
>
>Hi,
>
>Thanks for the docs. The overall format and content seem good. A few
>comments below.
>
>
>> +#. Bind the QLogic 579xx adapters to ``igb_uio`` or ``vfio-pci`` loaded
>> in the
>> + previous step::
>> +
>> + ./tools/dpdk_nic_bind.py --bind igb_uio 0000:84:00.0 0000:84:00.1
>> + 0000:84:00.2 0000:84:00.3
>
>Fixed width lines should be constrained to 80 characters or else they will
>go off the page in the PDF docs. The usual workaround is to use a command-
>line continuation (or text wrap). For example:
>
>
> ./tools/dpdk_nic_bind.py --bind igb_uio \
> 0000:84:00.0 0000:84:00.1 \
> 0000:84:00.2 0000:84:00.3
>
>Or similar. This also applies to the command-lines in other sections.
>> +
>> +#. Start ``testpmd`` with basic parameters:
>> +
>> + .. code-block:: console
>> +
>> + testpmd -c 0xf8000 -n 4 -- -i --nb-cores=4 --portmask=0xf
>> + --rxd=4096 --txd=4096 --txfreet=4068 --rxq=2 --txq=2 --rss-ip
>> + --rss-udp
>
>Same comment about using continuations.
>
>
>> +
>> + [...]
>> +
>> + EAL: PCI device 0000:84:00.0 on NUMA socket 1
>> + EAL: probe driver: 1077:1634 rte_qede_pmd
>
>Align the text in this section to the same level of indentation.
>
>
>> + EAL: Not managed by a supported kernel driver, skipped
>> + EAL: PCI device 0000:84:00.1 on NUMA socket 1
>> + EAL: probe driver: 1077:1634 rte_qede_pmd
>> + EAL: Not managed by a supported kernel driver, skipped
>> + EAL: PCI device 0000:88:00.0 on NUMA socket 1
>> + EAL: probe driver: 1077:1656 rte_qede_pmd
>> + EAL: PCI memory mapped at 0x7f738b200000
>> + EAL: PCI memory mapped at 0x7f738b280000
>> + EAL: PCI memory mapped at 0x7f738b300000
>> + [QEDE PMD: (88:00.0:dpdk-port-0)]qed_load_firmware_data: Loading
>>the
>> firmware file /lib/firmware/qed/qed_init_values_zipped.bin...
>
>Again, try wrap the code/console section at 80 chars in some way that
>still
>maintains the meaning. Maybe something like the following (with a note to
>say that the text has been wrapped for clarity):
>
> [QEDE PMD: (88:00.0:dpdk-port-0)]qed_load_firmware_data:
> Loading the firmware file
> /lib/firmware/qed/qed_init_values_zipped.bin...
> [QEDE PMD: (88:00.0:dpdk-port-0)]
> qede_print_adapter_info:Chip details - BB1
> [QEDE PMD: (88:00.0:dpdk-port-0)]
> qede_print_adapter_info:Driver version:QEDE PMD
>8.7.9.0_1.0.0
> [QEDE PMD: (88:00.0:dpdk-port-0)]
> qede_print_adapter_info:Firmware version:8.7.7.0
>
>
>You can test the PDF output as follows:
>
> make -j doc-guides-pdf
> pdf_veiwer build/doc/pdf/guides/nics.pdf
>
>John
>
These are the actual output from the console that was pasted, that’s why
they are more than 80 chars.
Sure, will take care of that as you mentioned.
Thanks,
Harish
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [dpdk-dev] [PATCH 2/6] qede: add documentation
2016-02-24 7:17 ` Harish Patil
@ 2016-02-24 9:26 ` Mcnamara, John
0 siblings, 0 replies; 25+ messages in thread
From: Mcnamara, John @ 2016-02-24 9:26 UTC (permalink / raw)
To: Harish Patil, dev
> -----Original Message-----
> From: Harish Patil [mailto:harish.patil@qlogic.com]
> Sent: Wednesday, February 24, 2016 7:18 AM
> To: Mcnamara, John <john.mcnamara@intel.com>; dev@dpdk.org
> Subject: Re: [dpdk-dev] [PATCH 2/6] qede: add documentation
>
> >>
> >
> >Again, try wrap the code/console section at 80 chars in some way that
> >still maintains the meaning. Maybe something like the following (with a
> >note to say that the text has been wrapped for clarity):
> >
> > ...
>
> These are the actual output from the console that was pasted, that’s why
> they are more than 80 chars.
> Sure, will take care of that as you mentioned.
Hi,
I understand that the text is actual output but if you generate the pdf
documentation you will see that the fixed width text goes off the page.
So some sort of compromise is required and wrapping the text is what
is usually done in the DPDK docs.
Since it may be confusing to the end user if they see different output
in the docs and on the console it is probably worth adding a line in the
preceding paragraph to say something like "(text wrapped for clarity)".
John.
--
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [dpdk-dev] [PATCH 0/6] DPDK PMD for new QLogic FastLinQ QL4xxxx 25G/40G CNAs
2016-02-20 15:40 [dpdk-dev] [PATCH 0/6] DPDK PMD for new QLogic FastLinQ QL4xxxx 25G/40G CNAs Harish Patil
` (5 preceding siblings ...)
2016-02-20 20:49 ` [dpdk-dev] [PATCH 0/6] DPDK PMD for new QLogic FastLinQ QL4xxxx 25G/40G CNAs Thomas Monjalon
@ 2016-03-08 14:01 ` Bruce Richardson
2016-03-08 14:24 ` Harish Patil
6 siblings, 1 reply; 25+ messages in thread
From: Bruce Richardson @ 2016-03-08 14:01 UTC (permalink / raw)
To: Harish Patil; +Cc: dev
On Sat, Feb 20, 2016 at 07:40:25AM -0800, Harish Patil wrote:
> This patch set introduces DPDK poll mode driver for new QLogic FastLinQ QL4xxxx
> 25G/40G capable family of CNAs as well as their SR-IOV Virtual Functions (VF).
>
> The overall PMD driver design includes a common module called ecore that deals
> with the low level HW and a upper layer portion that provides the glue logic.
>
> Specifically, the ecore module contains all of the common logic,
> e.g. initialization, cleanup, infrastructure for interrupt handling, link
> management, slowpath etc. as well as protocol agnostic features and supplying
> an abstraction layer for other modules.
>
> The higher layer implements DPDK exported APIs/driver entry points by
> interfacing with the common module for configuration/status and also the
> fastpath routines.
>
> Included in the patch set is the supporting documentation and maintainers.
>
> Please apply.
>
> Thanks,
>
> Harish Patil (6):
> qede: add maintainers
> qede: add documentation
> qede: add QLogic PCI ids
> qede: add driver common module
> qede: add driver
> qede: enable PMD build
Hi Harish,
there are quite a few comments to be addressed on this patchset. Are there plans
for a V2 in time for the code freeze deadline later this week?
/Bruce
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [dpdk-dev] [PATCH 0/6] DPDK PMD for new QLogic FastLinQ QL4xxxx 25G/40G CNAs
2016-03-08 14:01 ` Bruce Richardson
@ 2016-03-08 14:24 ` Harish Patil
2016-03-10 17:22 ` Harish Patil
0 siblings, 1 reply; 25+ messages in thread
From: Harish Patil @ 2016-03-08 14:24 UTC (permalink / raw)
To: Bruce Richardson; +Cc: dev
>
>On Sat, Feb 20, 2016 at 07:40:25AM -0800, Harish Patil wrote:
>> This patch set introduces DPDK poll mode driver for new QLogic FastLinQ
>>QL4xxxx
>> 25G/40G capable family of CNAs as well as their SR-IOV Virtual
>>Functions (VF).
>>
>> The overall PMD driver design includes a common module called ecore
>>that deals
>> with the low level HW and a upper layer portion that provides the glue
>>logic.
>>
>> Specifically, the ecore module contains all of the common logic,
>> e.g. initialization, cleanup, infrastructure for interrupt handling,
>>link
>> management, slowpath etc. as well as protocol agnostic features and
>>supplying
>> an abstraction layer for other modules.
>>
>> The higher layer implements DPDK exported APIs/driver entry points by
>> interfacing with the common module for configuration/status and also the
>> fastpath routines.
>>
>> Included in the patch set is the supporting documentation and
>>maintainers.
>>
>> Please apply.
>>
>> Thanks,
>>
>> Harish Patil (6):
>> qede: add maintainers
>> qede: add documentation
>> qede: add QLogic PCI ids
>> qede: add driver common module
>> qede: add driver
>> qede: enable PMD build
>Hi Harish,
>
>there are quite a few comments to be addressed on this patchset. Are
>there plans
>for a V2 in time for the code freeze deadline later this week?
>
> /Bruce
>
>
Hi Bruce,
Yes we are working on V2 series with the comments addressed. Will send out
the patches soon.
Thanks,
Harish
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [dpdk-dev] [PATCH 0/6] DPDK PMD for new QLogic FastLinQ QL4xxxx 25G/40G CNAs
2016-03-08 14:24 ` Harish Patil
@ 2016-03-10 17:22 ` Harish Patil
0 siblings, 0 replies; 25+ messages in thread
From: Harish Patil @ 2016-03-10 17:22 UTC (permalink / raw)
To: Bruce Richardson; +Cc: dev
>>
>>
>>
>>On Sat, Feb 20, 2016 at 07:40:25AM -0800, Harish Patil wrote:
>>> This patch set introduces DPDK poll mode driver for new QLogic FastLinQ
>>>QL4xxxx
>>> 25G/40G capable family of CNAs as well as their SR-IOV Virtual
>>>Functions (VF).
>>>
>>> The overall PMD driver design includes a common module called ecore
>>>that deals
>>> with the low level HW and a upper layer portion that provides the glue
>>>logic.
>>>
>>> Specifically, the ecore module contains all of the common logic,
>>> e.g. initialization, cleanup, infrastructure for interrupt handling,
>>>link
>>> management, slowpath etc. as well as protocol agnostic features and
>>>supplying
>>> an abstraction layer for other modules.
>>>
>>> The higher layer implements DPDK exported APIs/driver entry points by
>>> interfacing with the common module for configuration/status and also
>>>the
>>> fastpath routines.
>>>
>>> Included in the patch set is the supporting documentation and
>>>maintainers.
>>>
>>> Please apply.
>>>
>>> Thanks,
>>>
>>> Harish Patil (6):
>>> qede: add maintainers
>>> qede: add documentation
>>> qede: add QLogic PCI ids
>>> qede: add driver common module
>>> qede: add driver
>>> qede: enable PMD build
>>Hi Harish,
>>
>>there are quite a few comments to be addressed on this patchset. Are
>>there plans
>>for a V2 in time for the code freeze deadline later this week?
>>
>> /Bruce
>>
>>
>
>Hi Bruce,
>Yes we are working on V2 series with the comments addressed. Will send out
>the patches soon.
>Thanks,
>Harish
>
>
Hi Bruce
FYI - We have submitted v2 patch series after incorporating all review
comments.
Thanks,
Harish
^ permalink raw reply [flat|nested] 25+ messages in thread
end of thread, other threads:[~2016-03-10 17:22 UTC | newest]
Thread overview: 25+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2016-02-20 15:40 [dpdk-dev] [PATCH 0/6] DPDK PMD for new QLogic FastLinQ QL4xxxx 25G/40G CNAs Harish Patil
2016-02-20 15:40 ` [dpdk-dev] [PATCH 1/6] qede: add maintainers Harish Patil
2016-02-20 15:40 ` [dpdk-dev] [PATCH 2/6] qede: add documentation Harish Patil
2016-02-22 16:52 ` Mcnamara, John
2016-02-24 7:17 ` Harish Patil
2016-02-24 9:26 ` Mcnamara, John
2016-02-22 17:38 ` Mcnamara, John
2016-02-20 15:40 ` [dpdk-dev] [PATCH 3/6] qede: add QLogic PCI ids Harish Patil
2016-02-21 1:17 ` Stephen Hemminger
2016-02-22 23:23 ` Harish Patil
2016-02-20 15:40 ` [dpdk-dev] [PATCH 5/6] qede: add driver Harish Patil
2016-02-21 1:26 ` Stephen Hemminger
2016-02-23 2:28 ` Harish Patil
2016-02-23 5:30 ` Stephen Hemminger
2016-02-23 5:33 ` Stephen Hemminger
2016-02-23 19:04 ` Harish Patil
2016-02-23 19:06 ` Harish Patil
2016-02-20 15:40 ` [dpdk-dev] [PATCH 6/6] qede: enable PMD build Harish Patil
2016-02-20 20:49 ` [dpdk-dev] [PATCH 0/6] DPDK PMD for new QLogic FastLinQ QL4xxxx 25G/40G CNAs Thomas Monjalon
2016-02-22 16:47 ` Harish Patil
2016-02-22 17:35 ` Thomas Monjalon
2016-02-23 18:13 ` Harish Patil
2016-03-08 14:01 ` Bruce Richardson
2016-03-08 14:24 ` Harish Patil
2016-03-10 17:22 ` Harish Patil
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).