dpdk.git
6 years agopdump: remove unnecessary header include
Reshma Pattan [Mon, 12 Jun 2017 09:46:11 +0000 (10:46 +0100)]
pdump: remove unnecessary header include

Missed to remove unnecessary header file rte_pci.h.
Removed it now.

Fixes: bb900072ffaa ("pdump: revert PCI device name conversion")

Signed-off-by: Reshma Pattan <reshma.pattan@intel.com>
6 years agoethdev: add isolated mode to flow API
Adrien Mazarguil [Wed, 14 Jun 2017 14:48:51 +0000 (16:48 +0200)]
ethdev: add isolated mode to flow API

Isolated mode can be requested by applications on individual ports to avoid
ingress traffic outside of the flow rules they define.

Besides making ingress more deterministic, it allows PMDs to safely reuse
resources otherwise assigned to handle the remaining traffic, such as
global RSS configuration settings, VLAN filters, MAC address entries,
legacy filter API rules and so on in order to expand the set of possible
flow rule types.

To minimize code complexity, PMDs implementing this mode may provide
partial (or even no) support for flow rules when not enabled (e.g. no
priorities, no RSS action). Applications written to use the flow API are
therefore encouraged to enable it.

Once effective, leaving isolated mode may not be possible depending on PMD
implementation.

Signed-off-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>
Acked-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>
Acked-by: Andrew Rybchenko <arybchenko@solarflare.com>
6 years agohash: fix icc build
Ferruh Yigit [Tue, 13 Jun 2017 16:42:12 +0000 (17:42 +0100)]
hash: fix icc build

build error with icc version 17.0.4 (gcc version 7.0.0 compatibility):

In file included from .../dpdk/lib/librte_hash/rte_fbk_hash.h(59),
                 from .../dpdk/lib/librte_hash/rte_fbk_hash.c(54):
.../dpdk/x86_64-native-linuxapp-icc/include/rte_hash_crc.h(480):
 error #1292: unknown attribute "fallthrough"
                __attribute__ ((fallthrough));
                                ^

In file included from .../dpdk/lib/librte_hash/rte_fbk_hash.h(59),
                 from .../dpdk/lib/librte_hash/rte_fbk_hash.c(54):
.../dpdk/x86_64-native-linuxapp-icc/include/rte_hash_crc.h(486):
 error #1292: unknown attribute "fallthrough"
                __attribute__ ((fallthrough));
                                ^
This code patch hit when gcc > 7 installed and ICC doesn't recognize
fallthrough attribute.

Fixed by disabling code when compiled with ICC.

Fixes: 3dfb9facb055 ("lib: add switch fall-through comments")

Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
6 years agonet/mlx5: fix build with gcc 7.1
Ferruh Yigit [Tue, 13 Jun 2017 16:42:11 +0000 (17:42 +0100)]
net/mlx5: fix build with gcc 7.1

build error:
.../dpdk/drivers/net/mlx5/mlx5_fdir.c:
  In function ‘fdir_filter_to_flow_desc’:
.../dpdk/drivers/net/mlx5/mlx5_fdir.c:146:18:
 error: this statement may fall through [-Werror=implicit-fallthrough=]
   desc->dst_port = fdir_filter->input.flow.udp4_flow.dst_port;
   ~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
.../dpdk/drivers/net/mlx5/mlx5_fdir.c:147:2: note: here
  case RTE_ETH_FLOW_NONFRAG_IPV4_OTHER:
  ^~~~

Fixed by adding fallthrough comment to the code.

Fixes: 76f5c99e6840 ("mlx5: support flow director")
Cc: stable@dpdk.org
Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>
6 years agonet/enic: fix build with gcc 7.1
Ferruh Yigit [Tue, 13 Jun 2017 16:42:10 +0000 (17:42 +0100)]
net/enic: fix build with gcc 7.1

build error:

.../dpdk/drivers/net/enic/base/vnic_dev.c:
  In function ‘vnic_dev_get_mac_addr’:
.../dpdk/drivers/net/enic/base/vnic_dev.c:470:12:
  error: ‘a0’ is used uninitialized in this function
  [-Werror=uninitialized]
  args[0] = *a0;
            ^~~
...dpdk/drivers/net/enic/base/vnic_dev.c:
  In function ‘vnic_dev_classifier’:
...dpdk/drivers/net/enic/base/vnic_dev.c:471:12:
  error: ‘a1’ may be used uninitialized in this function
  [-Werror=maybe-uninitialized]
  args[1] = *a1;
            ^~~
Fixed by providing initial values.

Fixes: 9913fbb91df0 ("enic/base: common code")
Cc: stable@dpdk.org
Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
6 years agonet/i40e: fix memset size
Ferruh Yigit [Tue, 13 Jun 2017 16:42:09 +0000 (17:42 +0100)]
net/i40e: fix memset size

This causes build error with gcc 7.1.1 :

...dpdk/drivers/net/i40e/i40e_flow.c:2357:2:
error: ‘memset’ used with length equal to number of elements without
       multiplication by element size [-Werror=memset-elt-size]
  memset(off_arr, 0, I40E_MAX_FLXPLD_FIED);
  ^~~~~~

...dpdk/drivers/net/i40e/i40e_flow.c:2358:2:
error: ‘memset’ used with length equal to number of elements without
       multiplication by element size [-Werror=memset-elt-size]
  memset(len_arr, 0, I40E_MAX_FLXPLD_FIED);
  ^~~~~~

Fixed by providing correct size to memset.

Fixes: 6ced3dd72f5f ("net/i40e: support flexible payload parsing for FDIR")

Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
Acked-by: Beilei Xing <beilei.xing@intel.com>
6 years agokni: fix build with gcc 7.1
Ferruh Yigit [Tue, 13 Jun 2017 16:42:08 +0000 (17:42 +0100)]
kni: fix build with gcc 7.1

build error:
.../dpdk/build/build/lib/librte_eal/linuxapp/kni/igb_main.c:
  In function ‘igb_kni_probe’:
.../dpdk/build/build/lib/librte_eal/linuxapp/kni/igb_main.c:2483:30:
  error: ‘%d’ directive output may be truncated writing between 1 and 5
  bytes into a region of size between 0 and 11
  [-Werror=format-truncation=]
        "%d.%d, 0x%08x, %d.%d.%d",
                              ^~
.../dpdk/build/build/lib/librte_eal/linuxapp/kni/igb_main.c:2483:8:
  note: directive argument in the range [0, 65535]
        "%d.%d, 0x%08x, %d.%d.%d",
        ^~~~~~~~~~~~~~~~~~~~~~~~~
.../dpdk/build/build/lib/librte_eal/linuxapp/kni/igb_main.c:2481:4:
  note: ‘snprintf’ output between 23 and 43 bytes into a destination of
  size 32
    snprintf(adapter->fw_version,
    ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~
        sizeof(adapter->fw_version),
        ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
        "%d.%d, 0x%08x, %d.%d.%d",
        ~~~~~~~~~~~~~~~~~~~~~~~~~~
        fw.eep_major, fw.eep_minor, fw.etrack_id,
        ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
        fw.or_major, fw.or_build, fw.or_patch);
        ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Fixed by increasing buffer size to 43 as suggested in compiler log.

Fixes: b9ee370557f1 ("kni: update kernel driver ethtool baseline")
Cc: stable@dpdk.org
Reported-by: Nirmoy Das <ndas@suse.de>
Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
Reviewed-by: Markos Chandras <mchandras@suse.de>
6 years agonet/thunderx: manage PCI device mapping for SQS VFs
Jerin Jacob [Fri, 9 Jun 2017 10:27:46 +0000 (15:57 +0530)]
net/thunderx: manage PCI device mapping for SQS VFs

Since the commit e84ad157b7bc ("pci: unmap resources if probe fails"),
EAL unmaps the PCI device if ethdev probe returns positive or
negative value.

nicvf thunderx PMD needs special treatment for Secondary queue set(SQS)
PCIe VF devices, where, it expects to not unmap or free the memory
without registering the ethdev subsystem.

Enable the same behavior by using RTE_PCI_DRV_KEEP_MAPPED_RES
PCI driver flag.

Signed-off-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
6 years agoeal/pci: introduce a PCI driver flag
Jerin Jacob [Fri, 9 Jun 2017 10:27:45 +0000 (15:57 +0530)]
eal/pci: introduce a PCI driver flag

Some ethdev devices like nicvf thunderx PMD need special treatment for
Secondary queue set(SQS) PCIe VF devices, where, it expects to not unmap
or free the memory without registering the ethdev subsystem.

Introducing a new RTE_PCI_DRV_KEEP_MAPPED_RES
PCI driver flag to request PCI subsystem to not unmap the mapped PCI
resources(PCI BAR address) if unsupported device detected.

Suggested-by: Ferruh Yigit <ferruh.yigit@intel.com>
Signed-off-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
6 years agodrivers/net: document missing speed capabilities feature
Ferruh Yigit [Mon, 15 May 2017 12:30:46 +0000 (13:30 +0100)]
drivers/net: document missing speed capabilities feature

Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
Acked-by: John McNamara <john.mcnamara@intel.com>
6 years agoethdev: remove driver name from device private data
Ferruh Yigit [Mon, 12 Jun 2017 15:25:12 +0000 (16:25 +0100)]
ethdev: remove driver name from device private data

rte_driver->name has the driver name and all physical and virtual
devices has access to it.

Previously it was not possible for virtual ethernet devices to access
rte_driver->name field (because eth_dev used to keep only pci_dev),
and it was required to save driver name in the device private struct.

After re-works on bus and vdev, it is possible for all bus types to
access rte_driver.

It is able to remove the driver name from ethdev device private data and
use eth_dev->device->driver->name.

Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
Acked-by: Shreyansh Jain <shreyansh.jain@nxp.com>
Acked-by: Jan Blunck <jblunck@infradead.org>
6 years agonet/ring: use EAL APIs in PMD specific API
Ferruh Yigit [Mon, 12 Jun 2017 15:25:11 +0000 (16:25 +0100)]
net/ring: use EAL APIs in PMD specific API

When ring PMD created via PMD specific API instead of EAL abstraction
it is missing the virtual device creation done by EAL vdev.

And this makes eth_dev unusable exact same as other PMDs used, because
of some missing fields, like rte_device->name.

Now API calls EAL APIs to create ring PMDs.

Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
6 years agonet/ring: set ethernet device field
Ferruh Yigit [Mon, 12 Jun 2017 15:25:10 +0000 (16:25 +0100)]
net/ring: set ethernet device field

The eth_dev->device link was missing for ring PMD, adding it.

This is to generalize rte_device access from eth_dev.

Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
6 years agonet/szedata2: add more supported firmwares
Matej Vido [Mon, 12 Jun 2017 12:03:22 +0000 (14:03 +0200)]
net/szedata2: add more supported firmwares

Add IBUF and OBUF offsets definitions for new firmwares.

Signed-off-by: Matej Vido <vido@cesnet.cz>
6 years agonet/szedata2: move ibuf and obuf to specific header
Matej Vido [Mon, 12 Jun 2017 12:03:21 +0000 (14:03 +0200)]
net/szedata2: move ibuf and obuf to specific header

Signed-off-by: Matej Vido <vido@cesnet.cz>
6 years agonet/szedata2: refactor ibuf and obuf address definition
Matej Vido [Mon, 12 Jun 2017 12:03:20 +0000 (14:03 +0200)]
net/szedata2: refactor ibuf and obuf address definition

This is to prepare for firmwares with multiple ibufs and obufs.
Ibufs and obufs are the modules in FPGA firmware implementing
the Ethernet port.
There is one ibuf+obuf per Ethernet port.
The cards and firmwares allow one physical port to be one Ethernet
port or split into more Ethernet ports, e.g. one 100GE physical
port can be one Ethernet port of 100GE or split into ten Ethernet
ports of 10GE.
All DMA queues in the device are shared between all Ethernet ports.
Offsets of ibufs and obufs are defined in array.
Functions which operate on ibufs and obufs iterate over this array.

Signed-off-by: Matej Vido <vido@cesnet.cz>
6 years agonet/szedata2: refactor ibuf and obuf read and write
Matej Vido [Mon, 12 Jun 2017 12:03:19 +0000 (14:03 +0200)]
net/szedata2: refactor ibuf and obuf read and write

Remove unused read and write functions.
Use rte_read*, rte_write* functions to access ibuf and obuf
address space.

Signed-off-by: Matej Vido <vido@cesnet.cz>
6 years agonet/szedata2: refactor ibuf and obuf names
Matej Vido [Mon, 12 Jun 2017 12:03:18 +0000 (14:03 +0200)]
net/szedata2: refactor ibuf and obuf names

Prefix "cgmii" is removed because it is too specific.
There are different ibuf/obuf modules in different firmwares
but the address space definition is the same.
This patch makes the name general.

Signed-off-by: Matej Vido <vido@cesnet.cz>
6 years agonet/igb: flush all the filter
Wei Zhao [Mon, 12 Jun 2017 06:48:28 +0000 (14:48 +0800)]
net/igb: flush all the filter

This patch adds a function to flush all the fliter list
and filter rule on a port.

Signed-off-by: Wei Zhao <wei.zhao1@intel.com>
6 years agonet/igb: destroy consistent filter
Wei Zhao [Mon, 12 Jun 2017 06:48:27 +0000 (14:48 +0800)]
net/igb: destroy consistent filter

This patch adds a function to destroy the flow fliter.

Signed-off-by: Wei Zhao <wei.zhao1@intel.com>
6 years agonet/igb: create consistent filter
Wei Zhao [Mon, 12 Jun 2017 06:48:26 +0000 (14:48 +0800)]
net/igb: create consistent filter

This patch adds a function to create the flow directory filter.

Signed-off-by: Wei Zhao <wei.zhao1@intel.com>
6 years agonet/igb: parse flow API flex filter
Wei Zhao [Mon, 12 Jun 2017 06:48:25 +0000 (14:48 +0800)]
net/igb: parse flow API flex filter

check if the rule is a flex byte rule, and get the flex info.

Signed-off-by: Wei Zhao <wei.zhao1@intel.com>
6 years agonet/igb: parse flow API TCP SYN filter
Wei Zhao [Mon, 12 Jun 2017 06:48:24 +0000 (14:48 +0800)]
net/igb: parse flow API TCP SYN filter

check if the rule is a TCP SYN rule, and get the SYN info.

Signed-off-by: Wei Zhao <wei.zhao1@intel.com>
6 years agonet/igb: parse flow API ethertype filter
Wei Zhao [Mon, 12 Jun 2017 06:48:23 +0000 (14:48 +0800)]
net/igb: parse flow API ethertype filter

check if the rule is a ethertype rule, and get the ethertype info.

Signed-off-by: Wei Zhao <wei.zhao1@intel.com>
6 years agonet/igb: parse flow API n-tuple filter
Wei Zhao [Mon, 12 Jun 2017 06:48:22 +0000 (14:48 +0800)]
net/igb: parse flow API n-tuple filter

Add rule validate function and check if the rule is a
n-tuple rule, and get the n-tuple info.

Signed-off-by: Wei Zhao <wei.zhao1@intel.com>
6 years agonet/igb: restore flex type filter
Wei Zhao [Mon, 12 Jun 2017 06:48:21 +0000 (14:48 +0800)]
net/igb: restore flex type filter

Add support for restoring flex type filter in SW.

Signed-off-by: Wei Zhao <wei.zhao1@intel.com>
6 years agonet/igb: restore ether type filter
Wei Zhao [Mon, 12 Jun 2017 06:48:20 +0000 (14:48 +0800)]
net/igb: restore ether type filter

Add support for restoring ether type filter in SW.

Signed-off-by: Wei Zhao <wei.zhao1@intel.com>
6 years agonet/igb: restore n-tuple filter
Wei Zhao [Mon, 12 Jun 2017 06:48:19 +0000 (14:48 +0800)]
net/igb: restore n-tuple filter

Add support for restoring n-tuple filter in SW.

Signed-off-by: Wei Zhao <wei.zhao1@intel.com>
6 years agonet/igb: store and restore TCP SYN filter
Wei Zhao [Mon, 12 Jun 2017 06:48:18 +0000 (14:48 +0800)]
net/igb: store and restore TCP SYN filter

Add support for storing and restoring TCP SYN filter in SW.

Signed-off-by: Wei Zhao <wei.zhao1@intel.com>
6 years agonet/liquidio: do not touch mbuf initialized fields
Shijith Thotton [Thu, 8 Jun 2017 11:22:51 +0000 (16:52 +0530)]
net/liquidio: do not touch mbuf initialized fields

Avoid re-initializing of mbuf fields which are set while in pool.
Replaced lio_recv_buffer_alloc with rte_pktmbuf_alloc.

See commit 8f094a9ac5d7 ("mbuf: set mbuf fields while in pool").

Signed-off-by: Shijith Thotton <shijith.thotton@caviumnetworks.com>
6 years agonet/bnxt: fix reporting of link status
Ajit Khaparde [Fri, 9 Jun 2017 04:24:48 +0000 (23:24 -0500)]
net/bnxt: fix reporting of link status

This patch fixes incorrect reporting of link status

1) When link is down, set speed to zero. Otherwise a wrong non-zero
   speed will be displayed.

2) DAC cables can detect there is a signal, but it necessarily does not
   mean link is up. Code previously treated this as link up.

Fixes: 7bc8e9a227cc ("net/bnxt: support async link notification")
Cc: stable@dpdk.org
Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com>
Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
6 years agonet/bnxt: update HWRM defines
Ajit Khaparde [Fri, 9 Jun 2017 04:24:47 +0000 (23:24 -0500)]
net/bnxt: update HWRM defines

Some HWRM defines are missing from hsi_struct_def_dpdk.h
This patch adds them.

Also remove duplicate HWRM_RING_GRP_ALLOC entry.

Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
6 years agonet/bnxt: move PMD specific functions
Ajit Khaparde [Fri, 9 Jun 2017 04:24:46 +0000 (23:24 -0500)]
net/bnxt: move PMD specific functions

Move PMD specific functions in the appropriate rte_pmd_bnxt.c file

Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
6 years agonet/i40e: support ether pattern for FDIR
Beilei Xing [Fri, 9 Jun 2017 08:21:23 +0000 (16:21 +0800)]
net/i40e: support ether pattern for FDIR

Previously, i40e PMD will select ethertype filter
parser when adding ether pattern rules. In fact,
FDIR also supports ether pattern.
This patch adds ether pattern support for FDIR.

Signed-off-by: Beilei Xing <beilei.xing@intel.com>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
6 years agonet/i40e: update supported patterns for FDIR
Beilei Xing [Fri, 9 Jun 2017 08:21:22 +0000 (16:21 +0800)]
net/i40e: update supported patterns for FDIR

This patch updates supported patterns for flow
director filters.

Signed-off-by: Beilei Xing <beilei.xing@intel.com>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
6 years agonet/i40e: support input set selection for FDIR
Beilei Xing [Fri, 9 Jun 2017 08:21:21 +0000 (16:21 +0800)]
net/i40e: support input set selection for FDIR

This patch supports input set selection for flow
director filter.

Signed-off-by: Beilei Xing <beilei.xing@intel.com>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
6 years agonet/i40e: support flexible payload parsing for FDIR
Beilei Xing [Fri, 9 Jun 2017 08:21:20 +0000 (16:21 +0800)]
net/i40e: support flexible payload parsing for FDIR

This patch adds flexible payload parsing support for
flow director filter.

Signed-off-by: Beilei Xing <beilei.xing@intel.com>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
6 years agonet/i40e: add NVGRE flow parsing
Beilei Xing [Wed, 7 Jun 2017 06:53:59 +0000 (14:53 +0800)]
net/i40e: add NVGRE flow parsing

This patch adds NVGRE flow parsing function to support NVGRE
classification.

Signed-off-by: Beilei Xing <beilei.xing@intel.com>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
6 years agonet/i40e: refactor VXLAN flow parsing function
Beilei Xing [Wed, 7 Jun 2017 06:53:58 +0000 (14:53 +0800)]
net/i40e: refactor VXLAN flow parsing function

The current vxlan parsing function is not easy to read when parsing
filter type, this patch optimizes the function and makes it more
readable.

Signed-off-by: Beilei Xing <beilei.xing@intel.com>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
6 years agonet/liquidio: fix MTU calculation from port configuration
Shijith Thotton [Tue, 6 Jun 2017 11:04:34 +0000 (16:34 +0530)]
net/liquidio: fix MTU calculation from port configuration

max_rx_pkt_len member of port RX configuration indicates max frame
length. Ethernet header and CRC length should be subtracted from it to
find MTU.

Fixes: 605164c8e79d ("net/liquidio: add API to validate VF MTU")
Cc: stable@dpdk.org
Signed-off-by: Shijith Thotton <shijith.thotton@caviumnetworks.com>
6 years agonet/mlx4: support user space Rx interrupt event
Moti Haimovsky [Tue, 6 Jun 2017 14:48:29 +0000 (17:48 +0300)]
net/mlx4: support user space Rx interrupt event

Implement rxq interrupt callbacks

Signed-off-by: Moti Haimovsky <motih@mellanox.com>
Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>
6 years agonet/qede: refactor Tx routine
Harish Patil [Wed, 7 Jun 2017 07:42:22 +0000 (00:42 -0700)]
net/qede: refactor Tx routine

Refactor TX routine such that TX BD updates can all be grouped together.
Based on the TX offloads requested the TX bitfields are calculated in
a temporary variable and TX BDs are updated at the end. This will minimize
the if checks also. This change is done to easily accommodate newer TX
offload operations in the future.

Signed-off-by: Harish Patil <harish.patil@cavium.com>
6 years agonet/qede: fix VXLAN tunnel Tx offload flag setting
Harish Patil [Wed, 7 Jun 2017 07:42:21 +0000 (00:42 -0700)]
net/qede: fix VXLAN tunnel Tx offload flag setting

This patch fixes missing PKT_TX_TUNNEL_VXLAN Tx offload flag from the
supported Tx offloads and an incorrect tunnel TX BD bit setting.

Fixes: 3d4bb4411683 ("net/qede: add fastpath support for VXLAN tunneling")
Cc: stable@dpdk.org
Signed-off-by: Harish Patil <harish.patil@cavium.com>
6 years agonet/qede/base: upgrade the FW to 8.20.0.0
Rasesh Mody [Wed, 7 Jun 2017 07:42:20 +0000 (00:42 -0700)]
net/qede/base: upgrade the FW to 8.20.0.0

This patch adds changes to upgrade to 8.20.0.0 FW.

Signed-off-by: Rasesh Mody <rasesh.mody@cavium.com>
6 years agonet/qede: refactoring multi-queue implementation
Harish Patil [Wed, 7 Jun 2017 07:42:19 +0000 (00:42 -0700)]
net/qede: refactoring multi-queue implementation

This patch does the following refactoring and cleanup:

- As part of multi-queue support a struct member called 'type' was added
  in struct qede_fastpath in order to identify whether a queue is RX or
  TX and take actions based on that. This was unnecessary in the first
  place since pointers to RX and TX queues are already available in
  rte_eth_dev->data. So all usage of fp->type is removed.

- Remove remaining additional layer of internal callbacks for RX/TX
  queues and fastpath related operations from the qed_eth_ops_pass.
  With this change the files qede_eth_if.[c,h] are no longer needed.

- Add new per-queue start/stop APIs instead of clubbing it all together.

- Remove multiple TXQs references (num_tc and fp->txqs) since CoS is not
  supported.

- Enable sharing of the status block for each queue pair.

- Remove enum qede_dev_state and instead make use of existing port
  states RTE_ETH_QUEUE_STATE_STOPPED/RTE_ETH_QUEUE_STATE_STARTED.

- Move qede_dev_start() and qede_dev_stop() to qede_ethdev.c from
  qede_rxtc.c.

Signed-off-by: Harish Patil <harish.patil@cavium.com>
6 years agonet/qede: refactoring vport handling code
Harish Patil [Wed, 7 Jun 2017 07:42:18 +0000 (00:42 -0700)]
net/qede: refactoring vport handling code

The refactoring is mainly for two reasons:

- To remove an additional layer of internal callbacks for all vport
  related operations from the struct qed_eth_ops_pass. Instead, we
  can invoke base APIs directly.

- Splitting a single large vport-update configuration into multiple and
  independent vport-update operations. Each configuration would touch
  only the required config bits that needs an update.

Signed-off-by: Harish Patil <harish.patil@cavium.com>
6 years agodoc: update release notes for bnxt PMD
Ajit Khaparde [Thu, 1 Jun 2017 17:07:23 +0000 (12:07 -0500)]
doc: update release notes for bnxt PMD

Update release doc briefly describing updates to bnxt PMD.

Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
6 years agonet/bnxt: support to set VF rxmode
Ajit Khaparde [Thu, 1 Jun 2017 17:07:22 +0000 (12:07 -0500)]
net/bnxt: support to set VF rxmode

This patch adds support to configure the VF L2 Rx settings.
The per VF setting is maintained in bnxt_child_vf_info.l2_rx_mask

Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
6 years agonet/bnxt: configure a default VF VLAN
Ajit Khaparde [Thu, 1 Jun 2017 17:07:21 +0000 (12:07 -0500)]
net/bnxt: configure a default VF VLAN

This patch adds code to insert a default VF VLAN.
Also track the current default VLAN per vnic for the VF.
When setting the default VLAN, avoid setting it to the current value.

Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com>
Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
6 years agonet/bnxt: support to add a VF MAC address
Ajit Khaparde [Thu, 1 Jun 2017 17:07:20 +0000 (12:07 -0500)]
net/bnxt: support to add a VF MAC address

This patch adds support to allocate a filter and program
it in the hardware for every MAC address added to the specified
function.

Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com>
Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
6 years agonet/bnxt: determine the Rx status of VF
Ajit Khaparde [Thu, 1 Jun 2017 17:07:19 +0000 (12:07 -0500)]
net/bnxt: determine the Rx status of VF

This patch adds code to determine the Rx status of a VF.
It adds the rte_pmd_bnxt_get_vf_rx_status call, which calculates
the VNIC count of the function to get the Rx status.

Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com>
Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
6 years agonet/bnxt: support get and clear VF specific stats
Ajit Khaparde [Thu, 1 Jun 2017 17:07:18 +0000 (12:07 -0500)]
net/bnxt: support get and clear VF specific stats

This patch adds code to get and clear VF stats.

It also adds the necessary HWRM structures to send the command
to the firmware.

Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com>
Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
6 years agonet/bnxt: support set VF QOS and MAC anti spoof
Ajit Khaparde [Thu, 1 Jun 2017 17:07:17 +0000 (12:07 -0500)]
net/bnxt: support set VF QOS and MAC anti spoof

This patch adds support to
1) enable VF MAC anti spoof.
2) QOS configuration for specified VF.

Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
6 years agonet/bnxt: support Tx loopback, set VF MAC and queues drop
Ajit Khaparde [Thu, 1 Jun 2017 17:07:16 +0000 (12:07 -0500)]
net/bnxt: support Tx loopback, set VF MAC and queues drop

Add functions rte_pmd_bnxt_set_tx_loopback,
rte_pmd_bnxt_set_all_queues_drop_en and
rte_pmd_bnxt_set_vf_mac_addr to configure tx_loopback,
queue_drop and VF MAC address setting in the hardware.
It also adds the necessary functions to send the HWRM commands
to the firmware.

Signed-off-by: Steeven Li <steeven.li@broadcom.com>
Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
6 years agonet/bnxt: support LED on/off
Ajit Khaparde [Thu, 1 Jun 2017 17:07:15 +0000 (12:07 -0500)]
net/bnxt: support LED on/off

This patch adds support for dev_led_on/off dev_ops

HWRM calls added:
bnxt_hwrm_port_led_qcaps()
bnxt_hwrm_port_led_cfg()

Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
6 years agodoc: add LED support to default features file
Ajit Khaparde [Thu, 1 Jun 2017 17:07:14 +0000 (12:07 -0500)]
doc: add LED support to default features file

Update default.ini to indicate LED on/off support

Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
6 years agonet/bnxt: refactor the query stats
Ajit Khaparde [Thu, 1 Jun 2017 17:07:13 +0000 (12:07 -0500)]
net/bnxt: refactor the query stats

1) Use hwrm_stat_ctx_query command to query statistics
   Using hwrm_stat_ctx_query command will allow polling
   the statistics from hardware instead of using the current push
   model from the hardware which does a DMA of the stats to the host
   at fixed intervals.

2) Use the rx_mbuf_alloc_fail to track mbuf alloc failures.

3) We were wrongly incrementing hwrm_cmd_seq in bnxt_hwrm_stat_clear
   and bnxt_hwrm_stat_ctx_alloc functions.  This patch fixes that.

Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com>
Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
6 years agonet/bnxt: support VLAN pvid
Ajit Khaparde [Thu, 1 Jun 2017 17:07:12 +0000 (12:07 -0500)]
net/bnxt: support VLAN pvid

This patch adds code to support vlan_pvid_set dev_op

Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
6 years agonet/bnxt: support rxq/txq get information
Ajit Khaparde [Thu, 1 Jun 2017 17:07:11 +0000 (12:07 -0500)]
net/bnxt: support rxq/txq get information

Add support for txq_info_get and rxq_info_get dev_ops

Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
6 years agonet/bnxt: support LRO
Ajit Khaparde [Thu, 1 Jun 2017 17:07:10 +0000 (12:07 -0500)]
net/bnxt: support LRO

This patch adds support to enable and disable LRO
To support this feature, the driver creates an aggregator ring.
When the hardware starts doing LRO, it sends a tpa_start completion.
When the driver receives a tpa_end completion, it indicates that the
LRO chaining is complete.

Signed-off-by: Steeven Li <steeven.li@broadcom.com>
Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
6 years agonet/bnxt: support set MTU
Ajit Khaparde [Thu, 1 Jun 2017 17:07:09 +0000 (12:07 -0500)]
net/bnxt: support set MTU

This patch adds support to modify MTU using the set_mtu dev_op.
To support frames > 2k, the PMD creates an aggregator ring.
When a frame greater than 2k is received, it is fragmented
and the resulting fragments are DMA'ed to the aggregator ring.
Now the driver can support jumbo frames upto 9500 bytes.

Signed-off-by: Steeven Li <steeven.li@broadcom.com>
Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
6 years agonet/bnxt: support get FW version
Ajit Khaparde [Thu, 1 Jun 2017 17:07:08 +0000 (12:07 -0500)]
net/bnxt: support get FW version

This patch adds support for fw_version_get dev_op

Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
6 years agodoc: document bnxt allmulticast mode support
Ajit Khaparde [Thu, 1 Jun 2017 17:07:07 +0000 (12:07 -0500)]
doc: document bnxt allmulticast mode support

This patch update bnxt.int to document Allmulticast mode
which is already supported by the PMD.

Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
6 years agonet/bnxt: support multicast filter and set MAC addr
Ajit Khaparde [Thu, 1 Jun 2017 17:07:06 +0000 (12:07 -0500)]
net/bnxt: support multicast filter and set MAC addr

This patch adds support for set_mc_addr_list and
mac_addr_set dev_ops

Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
6 years agonet/bnxt: support VLAN filter and strip
Ajit Khaparde [Thu, 1 Jun 2017 17:07:05 +0000 (12:07 -0500)]
net/bnxt: support VLAN filter and strip

This patch adds VLAN strip and offload callbacks.
To add a VLAN filter:
    For each VNIC and each associated filter(s)
        if VLAN exists:
            if VLAN matches vlan_id
                VLAN filter already exists, just skip and continue
            else
                add a new MAC+VLAN filter
        else
            Remove the old MAC only filter
            Add a new MAC+VLAN filter

To remove a VLAN filter:
    For each VNIC and each associated filter(s)
        if VLAN exists && VLAN matches vlan_id
            remove the MAC+VLAN filter
            add a new MAC only filter
        else
            VLAN filter doesn't exist, just skip and continue

Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
6 years agonet/bnxt: support xstats get/reset
Ajit Khaparde [Thu, 1 Jun 2017 17:07:04 +0000 (12:07 -0500)]
net/bnxt: support xstats get/reset

This patch adds support to get and reset xstats dev_ops

dev_ops added:
xstats_get, xstats_get_name, xstats_reset

HWRM commands added:
hwrm_port_qstats, hwrm_port_clr_stats

Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
6 years agonet/bnxt: support tunneling
Ajit Khaparde [Thu, 1 Jun 2017 17:07:03 +0000 (12:07 -0500)]
net/bnxt: support tunneling

Add support for udp_tunnel_port_add/del dev_ops to configure a UDP port
for VXLAN and Geneve Tunnel protocols.

The HWRM supports only one global destination port for a tunnel type,
use a reference counter to keep track of its usage.
Cache the configured VXLAN/Geneve ports and use that value to check
if the right UDP port is being freed up.
Skip calling bnxt_hwrm_tunnel_dst_port_alloc if the same UDP port is
being programmed.
Skip calling bnxt_hwrm_tunnel_dst_port_free if no UDP port has been
configured.

Also update tx offload capabilities

Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
6 years agonet/bnxt: add additional HWRM debug info
Ajit Khaparde [Thu, 1 Jun 2017 17:07:02 +0000 (12:07 -0500)]
net/bnxt: add additional HWRM debug info

Add the cmd_err and opaque_0 and opaque_1 fields to HWRM error
messages.  These allow better debugging of some classes of HWRM
errors.

Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com>
Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
6 years agonet/bnxt: support lack of huge pages
Ajit Khaparde [Thu, 1 Jun 2017 17:07:01 +0000 (12:07 -0500)]
net/bnxt: support lack of huge pages

rte_malloc_virt2phy() does not return a physical address if huge pages
aren't in use.  Further, rte_memzone->phys_addr is not a physical address.

Use rte_mem_virt2phy() and manually lock pages to support lack of
huge pages.

Also check the return value of rte_mem_virt2phy()

Verify the function returns an address. Otherwise return an error and
log a message.

Signed-off-by: Stephen Hurd <stephen.hurd@broadcom.com>
Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
6 years agonet/bnxt: handle VF/PF initialization appropriately
Ajit Khaparde [Thu, 1 Jun 2017 17:07:00 +0000 (12:07 -0500)]
net/bnxt: handle VF/PF initialization appropriately

1) For a VF, query the firmware to determine if a MAC address is
   already configured. If not configure a random default MAC address.

2) Do not initialize the default completion ring in
   bnxt_alloc_hwrm_rings().

3) While registering for async events with the firmware,
   use func_vf_cfg for a VF and use func_cfg for a PF.

4) Query the VNIC plcmode config using the bnxt_hwrm_vnic_plcmodes_qcfg
   before a VNIC is updated. Reconfigure the VNIC with the plcmode
   configuration queried earlier. Not doing this could overwrite
   the plcmodes in some cases.

5) Reorg the bnxt_handle_fwd_req to properly handle the forwarded
   requests. The previous code did not handle it completely.

Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
6 years agonet/bnxt: refactor to properly allocate resources for PF/VF
Ajit Khaparde [Thu, 1 Jun 2017 17:06:59 +0000 (12:06 -0500)]
net/bnxt: refactor to properly allocate resources for PF/VF

1) Move the function reset to bnxt_dev_init.
   On the same lines, setup, enable and request interrupt to init path.
   Memory allocation is also being done in the init path.

2) After a function reset, configure the VFs.  Distribute resources
   evenly between all functions (PF and VF) for now. In the future, this
   should be controllable.

3) The bnxt_vf_info and bnxt_pf_info had lot of duplication. Move the
   common items to struct bnxt. And only unique items specific to PF
   remain in the struct bnxt_pf_info.

4) Program the firmware to allow certain commands sent by a VF.
   Disallowing these will prevent clean VF driver cleanup.

5) Since PF/VF need to allocate resources from a pool in the hardware,
   use func_qcaps and func_qcfg to appropriately query the capabilities
   and available resources.

6) If a PF is being initialized and no VFs are allocated, explicitly
   call func_cfg to allocate the resources.

7) Once resources are requested from the firmware, update local copy
   of resource count in struct bnxt only after sending the func_qcfg to
   make sure the allocation request in the firmware went through.

The changes in this patch will be used by the subsequent patches
to allow proper initialization of PF/VF instance.

Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
6 years agonet/bnxt: update to HWRM version 1.7.7
Ajit Khaparde [Thu, 1 Jun 2017 17:06:58 +0000 (12:06 -0500)]
net/bnxt: update to HWRM version 1.7.7

This patch updates the Broadcom bnxt PMD to version 1.7.7
Most of the changes in the patch are in the hsi_struct_def_dpdk.h - an
autogenerated file. The changes in the *.c files are because of changes
in the macro names.

Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
6 years agonet/i40e: fix VF statistics
Qi Zhang [Mon, 5 Jun 2017 21:14:20 +0000 (17:14 -0400)]
net/i40e: fix VF statistics

CRC bytes should be excluded, so rx/tx bytes of VF stats is aligned
with PF stats.

Fixes: 9aace75fc82e ("i40e: fix statistics")
Cc: stable@dpdk.org
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
6 years agonet/e1000: add support 2-tuple filter on i210/i211
Wei Zhao [Mon, 5 Jun 2017 05:40:57 +0000 (13:40 +0800)]
net/e1000: add support 2-tuple filter on i210/i211

Add support of i210 and i211 type nic in 2-tuple filter.

Signed-off-by: Wei Zhao <wei.zhao1@intel.com>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
6 years agonet/i40e: exclude internal packet's byte count
Qi Zhang [Thu, 1 Jun 2017 17:55:45 +0000 (13:55 -0400)]
net/i40e: exclude internal packet's byte count

Tx/Rx byte counts of internal managed packet should be
excluded from the total rx/tx bytes.

Fixes: 9aace75fc82e ("i40e: fix statistics")
Cc: stable@dpdk.org
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Beilei Xing <beilei.xing@intel.com>
6 years agonet/ixgbe: enable flex bytes for generic flow API
Qi Zhang [Thu, 1 Jun 2017 17:36:45 +0000 (13:36 -0400)]
net/ixgbe: enable flex bytes for generic flow API

Add fdir flex byte support for rte_flow APIs.

Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
6 years agonet/ixgbe: fix fdir mask not be reset
Qi Zhang [Thu, 1 Jun 2017 17:36:44 +0000 (13:36 -0400)]
net/ixgbe: fix fdir mask not be reset

When the last fdir flow be destroyed, the flag "mask_added"
should be reset, so the remain mask info will not take effect
when a new flow be added.

Fixes: a14de8b498d1 ("net/ixgbe: destroy consistent filter")
Cc: stable@dpdk.org
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
6 years agonet/ixgbe: remove redundant code
Qi Zhang [Thu, 1 Jun 2017 17:36:43 +0000 (13:36 -0400)]
net/ixgbe: remove redundant code

Remove redundant code.
item->type != RTE_FLOW_ITEM_TYPE_END already cover
item->type == RTE_FLOW_ITEM_TYPE_VLAN.

Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
6 years agonet/i40e: update actions for FDIR
Beilei Xing [Wed, 31 May 2017 23:47:30 +0000 (07:47 +0800)]
net/i40e: update actions for FDIR

This commit adds support of FLAG action and PASSTHRU
action for flow director.

Signed-off-by: Beilei Xing <beilei.xing@intel.com>
Acked-by: Jingjing Wu <jingjing.wu@intel.com>
6 years agoethdev: move bypass functions to ixgbe PMD
Radu Nicolau [Wed, 31 May 2017 11:10:25 +0000 (12:10 +0100)]
ethdev: move bypass functions to ixgbe PMD

Move all bypass functions to ixgbe pmd and remove function
pointers from the eth_dev_ops struct.

Signed-off-by: Radu Nicolau <radu.nicolau@intel.com>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
6 years agonet/igb: fix checksum valid flags
Wei Zhao [Wed, 31 May 2017 03:03:12 +0000 (11:03 +0800)]
net/igb: fix checksum valid flags

To enable L4 checksum offload for SCTP packets, E1000_RXCSUM_CRCOFL
should be set.

Otherwise, for SCTP packets even L4 checksum not calculated, it will
return PKT_RX_L4_CKSUM_GOOD.

Fixes: af75078fece3 ("first public release")

Signed-off-by: Wei Zhao <wei.zhao1@intel.com>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
6 years agonet/mlx5: fix redundant free of Tx buffer
Yongseok Koh [Wed, 31 May 2017 17:48:45 +0000 (10:48 -0700)]
net/mlx5: fix redundant free of Tx buffer

SW completion ring of Tx (txq->elts) stores individual mbufs even if a
multi-segmented packet is sent. rte_pktmbuf_free_seg() must be used when
cleaning up the completion ring. Otherwise, chained mbufs are redundantly
freed and finally it would cause a crash.

Fixes: 1d88ba171942 ("net/mlx5: refactor Tx data path")
CC: stable@dpdk.org
Signed-off-by: Yongseok Koh <yskoh@mellanox.com>
Acked-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>
6 years agonet/mlx5: fix exception handling
Yongseok Koh [Tue, 30 May 2017 01:02:59 +0000 (18:02 -0700)]
net/mlx5: fix exception handling

A sanity check is required in priv_fdir_disable(). If resizing Rx queue
fails, this can cause a crash by referencing a NULL pointer.

Fixes: 76f5c99e6840 ("mlx5: support flow director")
Fixes: 0cdddf4d0626 ("net/mlx5: split Rx queue structure")
Cc: stable@dpdk.org
Signed-off-by: Yongseok Koh <yskoh@mellanox.com>
Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>
6 years agonet/sfc: rely on one desc is one mbuf in simple EF10 Tx reap
Andrew Rybchenko [Fri, 26 May 2017 13:50:44 +0000 (14:50 +0100)]
net/sfc: rely on one desc is one mbuf in simple EF10 Tx reap

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Andy Moreton <amoreton@solarflare.com>
Reviewed-by: David Riddoch <driddoch@solarflare.com>
6 years agonet/sfc: check added but not completed descs on EF10 Tx reap
Andrew Rybchenko [Fri, 26 May 2017 13:50:43 +0000 (14:50 +0100)]
net/sfc: check added but not completed descs on EF10 Tx reap

There is not point to check other Tx descriptors.
It is important if Tx datapath does not reset Tx descriptor
mbuf pointer to NULL on completion (EF10 simple Tx will do).

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Andy Moreton <amoreton@solarflare.com>
Reviewed-by: David Riddoch <driddoch@solarflare.com>
6 years agonet/ring: fix adding MAC addresses
Chas Williams [Fri, 26 May 2017 13:18:05 +0000 (09:18 -0400)]
net/ring: fix adding MAC addresses

When .mac_addr_add() was changed to allow a return code, ring was changed
to return -ENOTSUP.  This changes the behavior of the rte_ring driver in
a way that does not maintain backward compatibility.  Additionally, if
the intent is to return -ENOTSUP, you could simply not define the stubs.

Fixes: 6d01e580ac5d ("ethdev: fix adding invalid MAC address")
Cc: stable@dpdk.org
Signed-off-by: Chas Williams <ciwillia@brocade.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
6 years agonet/mlx5: remove already initialised mbuf fields
Nélio Laranjeiro [Mon, 29 May 2017 09:45:09 +0000 (11:45 +0200)]
net/mlx5: remove already initialised mbuf fields

Since commit 8f094a9ac5d7 ("mbuf: set mbuf fields while in pool"), some
fields are already initialised and do not need to be modified by the PMD
anymore.

Signed-off-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>
Acked-by: Shahaf Shuler <shahafs@mellanox.com>
Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>
6 years agonet/mlx5: fix flow application order on stop/start
Nélio Laranjeiro [Mon, 29 May 2017 09:40:58 +0000 (11:40 +0200)]
net/mlx5: fix flow application order on stop/start

Flow rules must be applied in the same order as they have been created and
thus destroyed in the reverse order.

Fixes: 2097d0d1e2cc ("net/mlx5: support basic flow items and actions")
Cc: stable@dpdk.org
Signed-off-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>
Acked-by: Yongseok Koh <yskoh@mellanox.com>
Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>
6 years agonet/ixgbe/base: update shared code version to 2017.05.16
Wei Dai [Sat, 27 May 2017 08:33:18 +0000 (16:33 +0800)]
net/ixgbe/base: update shared code version to 2017.05.16

* Remove PHY access for some 1G ports
* Disable X550EM-x 1GBASE-T led switch support

Signed-off-by: Wei Dai <wei.dai@intel.com>
6 years agonet/ixgbe/base: disable X550EM-x 1GBASE-T led switch support
Wei Dai [Sat, 27 May 2017 08:33:17 +0000 (16:33 +0800)]
net/ixgbe/base: disable X550EM-x 1GBASE-T led switch support

This patch disables X550EM_X 1Gbase-t led_[on|off] support since
the LEDs are wired to the PHY and the driver can not access the
PHY. led_[on|off]

Support is disabled by setting the function pointer to NULL.
init_led_link_act is also set to NULL.

Signed-off-by: Wei Dai <wei.dai@intel.com>
6 years agonet/ixgbe/base: remove PHY access for some 1G ports
Wei Dai [Sat, 27 May 2017 08:33:16 +0000 (16:33 +0800)]
net/ixgbe/base: remove PHY access for some 1G ports

This patch removes some some 1GBASE-T PHY access since the FW
configures the PHY. SW shall not configure or initialize link.
Accessing the PHY would require the use of MDI clause 22 which
should be avoided in high layer driver code.

Signed-off-by: Wei Dai <wei.dai@intel.com>
6 years agonet/sfc: handle already flushed Tx queue gracefully
Andrew Rybchenko [Sat, 27 May 2017 07:55:34 +0000 (08:55 +0100)]
net/sfc: handle already flushed Tx queue gracefully

Tx queue may be already flushed because of previous Tx error or
MC reboot.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
6 years agonet/sfc: add Tx queue flush failed flag for sanity
Andrew Rybchenko [Sat, 27 May 2017 07:55:33 +0000 (08:55 +0100)]
net/sfc: add Tx queue flush failed flag for sanity

Avoid usage of flushing state when Tx queue flush init failed.

Fixes: fed9aeb46c19 ("net/sfc: implement transmit path start / stop")
Cc: stable@dpdk.org
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
6 years agonet/sfc: handle already flushed Rx queue gracefully
Andrew Rybchenko [Sat, 27 May 2017 07:55:32 +0000 (08:55 +0100)]
net/sfc: handle already flushed Rx queue gracefully

Rx queue may be already flushed because of previous Rx error or
MC reboot.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
6 years agonet/sfc/base: let caller know that queue is already flushed
Andy Moreton [Sat, 27 May 2017 07:55:31 +0000 (08:55 +0100)]
net/sfc/base: let caller know that queue is already flushed

Tx/Rx queue may be already flushed due to Tx/Rx error on the queue or
MC reboot. Caller needs to know that the queue is already flushed to
avoid waiting for flush done event.

Signed-off-by: Andy Moreton <amoreton@solarflare.com>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
6 years agonet/sfc/base: fix error code usage in common code
Andy Moreton [Sat, 27 May 2017 07:55:30 +0000 (08:55 +0100)]
net/sfc/base: fix error code usage in common code

MCDI results returned in req.emr_rc have already been translated
from MC_CMD_ERR_* to errno names, so using an MC_CMD_ERR_* value
is incorrect.

Fixes: e7cd430c864f ("net/sfc/base: import SFN7xxx family support")
Cc: stable@dpdk.org
Signed-off-by: Andy Moreton <amoreton@solarflare.com>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
6 years agonet/cxgbe: configure PCIe extended tags
Rahul Lakkireddy [Sat, 27 May 2017 03:48:00 +0000 (09:18 +0530)]
net/cxgbe: configure PCIe extended tags

Add support to configure minimum and maximum PCIe extended tag.

Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
6 years agonet/cxgbe: remove rmb bottleneck in Rx path
Rahul Lakkireddy [Sat, 27 May 2017 03:47:59 +0000 (09:17 +0530)]
net/cxgbe: remove rmb bottleneck in Rx path

rmb before determining rsp_type is a bottleneck.
Once we determine rsp-type is FL, we can directly go ahead and read
packets based on q->stat->pidx and budget_left.

This removes bottleneck of rmb once per every RX packet.
Now, rmb exists once per RX batch.

Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
6 years agonet/cxgbe: fix rxq default params for ports under same PF
Rahul Lakkireddy [Sat, 27 May 2017 03:47:58 +0000 (09:17 +0530)]
net/cxgbe: fix rxq default params for ports under same PF

Enabling rx queues with default interrupt parameters doesn't happen
for other ports under same PF due to FULL_INIT_DONE flag being set
by the first port.

Fix is to to allow each port to enable its own rx queues with default
parameters.

Fixes: 0462d115441d ("cxgbe: add device related operations")

Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
6 years agonet/cxgbe: improve latency for slow traffic
Rahul Lakkireddy [Sat, 27 May 2017 03:47:57 +0000 (09:17 +0530)]
net/cxgbe: improve latency for slow traffic

TX coalescing waits for ETH_COALESCE_PKT_NUM packets to be coalesced
across bursts before transmitting them.  For slow traffic, such as
100 PPS, this approach increases latency since packets are received
one at a time and tx coalescing has to wait for ETH_COALESCE_PKT
number of packets to arrive before transmitting.

To fix this:

- Update rx path to use status page instead and only receive packets
  when either the ingress interrupt timer threshold (5 us) or
  the ingress interrupt packet count threshold (32 packets) fires.
  (i.e. whichever happens first).

- If number of packets coalesced is <= number of packets sent
  by tx burst function, stop coalescing and transmit these packets
  immediately.

Also added compile time option to favor throughput over latency by
default.

Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>