dpdk.git
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>
6 years agonet/cxgbe: remove LSB interrupt support
Rahul Lakkireddy [Sat, 27 May 2017 03:46:28 +0000 (09:16 +0530)]
net/cxgbe: remove LSB interrupt support

Remove RTE_PCI_DRV_INTR_LSC from driver flags, since LSC interrupts
are currently not supported.

Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
6 years agonet/cxgbe: fix port statistics
Rahul Lakkireddy [Sat, 27 May 2017 03:46:27 +0000 (09:16 +0530)]
net/cxgbe: fix port statistics

Do not count pause frames as part of normal TX/RX Frame/Byte counts.
Also, switch to using software counters rather than hardware for RX
stats.

Fixes: 856505d303f4 ("cxgbe: add port statistics")
Cc: stable@dpdk.org
Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
6 years agonet/cxgbe: add compressed error vector
Rahul Lakkireddy [Sat, 27 May 2017 03:46:26 +0000 (09:16 +0530)]
net/cxgbe: add compressed error vector

Add support for compressed error vector available in cpl_rx_pkt for
Chelsio T6.

Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
6 years agonet/cxgbe: update Rx path for Chelsio T6
Rahul Lakkireddy [Sat, 27 May 2017 03:46:25 +0000 (09:16 +0530)]
net/cxgbe: update Rx path for Chelsio T6

Update RX path to reflect Chelsio T6 register value changes.
Update ingress pack boundary value based on maximum payload size
that can be accommodated by underlying PCI.  Update ingress pad
boundary value based on smallest memory controller bus width
possible.  Enforce alignment for free list pointer start address.

Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
6 years agonet/cxgbe: update RXQ channel mapping for Chelsio T6
Rahul Lakkireddy [Sat, 27 May 2017 03:46:24 +0000 (09:16 +0530)]
net/cxgbe: update RXQ channel mapping for Chelsio T6

Update MPS buffer groups and TP ingress channel congestion map to
correct RXQ allocation for Chelsio T6.

Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
6 years agonet/cxgbe: update Tx path for Chelsio T6
Rahul Lakkireddy [Sat, 27 May 2017 03:46:23 +0000 (09:16 +0530)]
net/cxgbe: update Tx path for Chelsio T6

Update TX path to reflect Chelsio T6 changes.

Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
6 years agonet/cxgbe: update hardware info prints
Rahul Lakkireddy [Sat, 27 May 2017 03:46:22 +0000 (09:16 +0530)]
net/cxgbe: update hardware info prints

Update informational prints pertaining to underlying hardware.
Add bootstrap firmware version and expansion ROM version prints.
Move the prints to a single function.

Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
6 years agonet/cxgbe/base: add forward error correction support
Rahul Lakkireddy [Sat, 27 May 2017 03:46:21 +0000 (09:16 +0530)]
net/cxgbe/base: add forward error correction support

Add forward error correction (FEC) support and update link
configuration.

Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
6 years agonet/cxgbe: update link speeds and port modules
Rahul Lakkireddy [Sat, 27 May 2017 03:46:20 +0000 (09:16 +0530)]
net/cxgbe: update link speeds and port modules

Add 25G and 100G link speeds and update supported port modules.

Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
6 years agonet/cxgbe: grab available ports after firmware reset
Rahul Lakkireddy [Sat, 27 May 2017 03:46:19 +0000 (09:16 +0530)]
net/cxgbe: grab available ports after firmware reset

Move code to get the available ports from the firmware after it had
been reset.  This ensures that driver uses the latest info on available
ports after firmware reset.

Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
6 years agonet/cxgbe/base: update flash part information
Rahul Lakkireddy [Sat, 27 May 2017 03:46:18 +0000 (09:16 +0530)]
net/cxgbe/base: update flash part information

Update flash part identification and decoding to accommodate new
flash parts.

Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
6 years agonet/cxgbe/base: update register dump
Rahul Lakkireddy [Sat, 27 May 2017 03:46:17 +0000 (09:16 +0530)]
net/cxgbe/base: update register dump

Add Chelsio T6 registers to register dump.

Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
6 years agonet/cxgbe: add support to run Chelsio T6 cards
Rahul Lakkireddy [Sat, 27 May 2017 03:46:16 +0000 (09:16 +0530)]
net/cxgbe: add support to run Chelsio T6 cards

Add code to detect and run T6 devices.  Update PCI ID Device table
with Chelsio T6 device ids and update documentation.

Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
6 years agonet/avp: remove redundant assignment
Ferruh Yigit [Thu, 25 May 2017 17:52:45 +0000 (18:52 +0100)]
net/avp: remove redundant assignment

dev_info->driver_name is assigned by rte_eth_dev_info_get()
and existing value overwritten, removing assignment.

Fixes: 1a85922369c4 ("net/avp: add device configuration")

Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
Acked-by: Allain Legacy <allain.legacy@windriver.com>
6 years agonet/tap: add Rx/Tx checksum offload support
Pascal Mazon [Mon, 22 May 2017 11:20:43 +0000 (13:20 +0200)]
net/tap: add Rx/Tx checksum offload support

This patch adds basic offloading support, widely expected in a PMD.

Verify IPv4 and UDP/TCP checksums upon packet reception, and set
ol_flags accordingly.

On Tx, set IPv4 and UDP/TCP checksums when required, considering
ol_flags.

Signed-off-by: Pascal Mazon <pascal.mazon@6wind.com>
6 years agonet/tap: fix some flow collision
Pascal Mazon [Mon, 22 May 2017 11:19:53 +0000 (13:19 +0200)]
net/tap: fix some flow collision

The following two flow rules (testpmd syntax) should not collide:
flow create 0 priority 1 ingress pattern eth / ipv4 / end actions drop / end
flow create 0 priority 1 ingress pattern eth / ipv6 / end actions drop / end

But the eth_type in the associated TC rule was set to either "ip" or
"ipv6".  For TC, they could thus not have the same priority.

Use ETH_P_ALL only in the TC message to make sure those rules can
coexist.

Fixes: de96fe68ae95 ("net/tap: add basic flow API patterns and actions")
Cc: stable@dpdk.org
Signed-off-by: Pascal Mazon <pascal.mazon@6wind.com>
6 years agobus/fslmc: reduce the debug log messages
Hemant Agrawal [Fri, 26 May 2017 06:51:25 +0000 (12:21 +0530)]
bus/fslmc: reduce the debug log messages

Reducing the amount of logging and making the err statement
as a single line statements.

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
6 years agonet/dpaa2: add support for firmware version get
Hemant Agrawal [Fri, 26 May 2017 06:51:24 +0000 (12:21 +0530)]
net/dpaa2: add support for firmware version get

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
6 years agobus/fslmc: add support to detect SoC version
Hemant Agrawal [Fri, 26 May 2017 06:51:23 +0000 (12:21 +0530)]
bus/fslmc: add support to detect SoC version

This is required to configure the stashing destination
based on the cluster size.

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
6 years agonet/dpaa2: configure jumbo frames
Hemant Agrawal [Fri, 26 May 2017 06:51:22 +0000 (12:21 +0530)]
net/dpaa2: configure jumbo frames

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
6 years agonet/dpaa2: add support for flow control
Hemant Agrawal [Fri, 26 May 2017 06:51:21 +0000 (12:21 +0530)]
net/dpaa2: add support for flow control

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
6 years agonet/dpaa2: add link status config support
Hemant Agrawal [Fri, 26 May 2017 06:51:20 +0000 (12:21 +0530)]
net/dpaa2: add link status config support

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
6 years agonet/dpaa2: add support for VLAN strip
Hemant Agrawal [Fri, 26 May 2017 06:51:19 +0000 (12:21 +0530)]
net/dpaa2: add support for VLAN strip

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
6 years agonet/dpaa2: add support for VLAN filter and offload
Hemant Agrawal [Fri, 26 May 2017 06:51:18 +0000 (12:21 +0530)]
net/dpaa2: add support for VLAN filter and offload

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
6 years agonet/dpaa2: add support for multicast promiscuous mode
Hemant Agrawal [Fri, 26 May 2017 06:51:17 +0000 (12:21 +0530)]
net/dpaa2: add support for multicast promiscuous mode

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
6 years agonet/dpaa2: add support for MAC address filtering
Hemant Agrawal [Fri, 26 May 2017 06:51:16 +0000 (12:21 +0530)]
net/dpaa2: add support for MAC address filtering

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
6 years agobus/fslmc: update TAILQ usages in dpaa2 objects
Hemant Agrawal [Fri, 26 May 2017 06:51:15 +0000 (12:21 +0530)]
bus/fslmc: update TAILQ usages in dpaa2 objects

This patch updates the usages of malloc and TAILQ in
dpbp and dpio objects.

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
6 years agonet/dpaa2: add support for tail drop on queue
Hemant Agrawal [Fri, 26 May 2017 06:51:14 +0000 (12:21 +0530)]
net/dpaa2: add support for tail drop on queue

This will help in limiting the size of queues and avoid
them growing practically infinite.

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
6 years agonet/dpaa2: add support for congestion notification
Hemant Agrawal [Fri, 26 May 2017 06:51:13 +0000 (12:21 +0530)]
net/dpaa2: add support for congestion notification

In case of HW egress FQ is congested, skip further
transmission of frames.

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
6 years agobus/fslmc: export qbman results in map file
Hemant Agrawal [Fri, 26 May 2017 06:51:12 +0000 (12:21 +0530)]
bus/fslmc: export qbman results in map file

In order to check the congestion state of Egress FQs,
this qbman function will be used by the DPAA2 Ethernet PMD.

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
6 years agonet/dpaa2: support parallel Rx
Hemant Agrawal [Fri, 26 May 2017 06:51:11 +0000 (12:21 +0530)]
net/dpaa2: support parallel Rx

Typically when the PMD issues a Rx command to DPAA2 hardware,
the HW writes the available descriptors into the given memory.
The RX function then processes the frames and prepare them as
mbufs.

This patch adds support to issue another pull request to hardware
in another memory location, before we start processing the output
of the first request. This help in controlling the CPU cycles
wasted during the wait for the hardware to write the descriptors.

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
6 years agobus/fslmc: support for parallel Rx DQ requests
Hemant Agrawal [Fri, 26 May 2017 06:51:10 +0000 (12:21 +0530)]
bus/fslmc: support for parallel Rx DQ requests

DPAA2 hardware support the option to configure
multiple memories for Rx recv buffer (DPAA2-DQRR).
Each Rx request executing is called as 'DQ' request.
This patch adds routines to get information w.r.t each DQ request.

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
6 years agonet/dpaa2: improve the error handling in dev init
Hemant Agrawal [Fri, 26 May 2017 06:51:09 +0000 (12:21 +0530)]
net/dpaa2: improve the error handling in dev init

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
6 years agonet/dpaa2: stop using software annotation
Hemant Agrawal [Fri, 26 May 2017 06:51:08 +0000 (12:21 +0530)]
net/dpaa2: stop using software annotation

The DPAA2 driver is not using the DPAA2 frame descriptor
software annotation area. This patch reduces the
PTA length to zero and adjust the RX Buffer Layout
accordingly.

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
6 years agonet/dpaa2: remove port level buffer layout definition
Hemant Agrawal [Fri, 26 May 2017 06:51:07 +0000 (12:21 +0530)]
net/dpaa2: remove port level buffer layout definition

This code in device init is not required, as the buffer
layout setting is also being programmed during queue configuration.

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
6 years agomk: change dpaa2 machine flag to cortex-a72
Hemant Agrawal [Fri, 26 May 2017 06:51:06 +0000 (12:21 +0530)]
mk: change dpaa2 machine flag to cortex-a72

DPAA2 devices now support cortex-a72. They no longer support a57.
Also fp and simd is no more required to be stated explicitly for
standard a72 core.

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
6 years agonet/sfc: advertise that device is detachable
Andrew Rybchenko [Thu, 25 May 2017 12:03:18 +0000 (13:03 +0100)]
net/sfc: advertise that device is detachable

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
6 years agonet/liquidio: set device speed capability info
Shijith Thotton [Thu, 25 May 2017 09:24:18 +0000 (14:54 +0530)]
net/liquidio: set device speed capability info

Add subsystem device id of CN23xx cards and expose speed capability of
devices. Update liquidio feature file to mark support.

Signed-off-by: Shijith Thotton <shijith.thotton@caviumnetworks.com>
6 years agonet/igb: fix add/delete of flex filters
Markus Theil [Mon, 22 May 2017 10:17:50 +0000 (12:17 +0200)]
net/igb: fix add/delete of flex filters

Before this patch, flex_filter->index was always zero when it was read
and used after rte_zmalloc. The corresponding code was therefore moved
into the add and delete parts of the if/else statement.

Fixes: 231d43909a31 ("igb: migrate flex filter to new API")
Cc: stable@dpdk.org
Signed-off-by: Markus Theil <markus.theil@tu-ilmenau.de>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
6 years agonet/igb: support flex filters on i210/i211
Markus Theil [Mon, 22 May 2017 10:17:49 +0000 (12:17 +0200)]
net/igb: support flex filters on i210/i211

i210/i211 also support flex filters. Tested on i210.

Signed-off-by: Markus Theil <markus.theil@tu-ilmenau.de>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
6 years agonet/mlx5: fix completion buffer size
Shahaf Shuler [Thu, 18 May 2017 10:44:45 +0000 (13:44 +0300)]
net/mlx5: fix completion buffer size

Completion buffer size was computed wrongly, causing
completion polling to wraparound too early and miss entries.

Fixing it by using Direct Verbs to query the CQ info.

Fixes: 6218063b39a6 ("net/mlx5: refactor Rx data path")
Fixes: 1d88ba171942 ("net/mlx5: refactor Tx data path")
Cc: stable@dpdk.org
Signed-off-by: Shahaf Shuler <shahafs@mellanox.com>
Acked-by: Yongseok Koh <yskoh@mellanox.com>
Acked-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>
6 years agonet/af_packet: fix packet bytes counting
Chas Williams [Wed, 17 May 2017 18:03:35 +0000 (14:03 -0400)]
net/af_packet: fix packet bytes counting

On error, we also need to zero the bytes transmitted.

Fixes: f4ff17a465d2 ("net/af_packet: guard against buffer overruns in Tx path")
Cc: stable@dpdk.org
Signed-off-by: Chas Williams <ciwillia@brocade.com>
6 years agonet/af_packet: handle possible null pointer
Chas Williams [Wed, 17 May 2017 18:03:33 +0000 (14:03 -0400)]
net/af_packet: handle possible null pointer

Fixes: 1b93c2aa81b4 ("net/af_packet: add interface name to internals")
Cc: stable@dpdk.org
Signed-off-by: Chas Williams <ciwillia@brocade.com>
6 years agonet/sfc: support multi-process
Andrew Rybchenko [Thu, 18 May 2017 14:00:04 +0000 (15:00 +0100)]
net/sfc: support multi-process

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Andy Moreton <amoreton@solarflare.com>
6 years agonet/sfc: use locally stored data for logging
Andrew Rybchenko [Thu, 18 May 2017 14:00:03 +0000 (15:00 +0100)]
net/sfc: use locally stored data for logging

Required to be able to use logging in the secondary process
where Ethernet device pointer stored in sfc_adapter is invalid.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Andy Moreton <amoreton@solarflare.com>
6 years agonet/sfc: carefully cleanup on init failure and shutdown
Andrew Rybchenko [Thu, 18 May 2017 14:00:02 +0000 (15:00 +0100)]
net/sfc: carefully cleanup on init failure and shutdown

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Andy Moreton <amoreton@solarflare.com>
6 years agonet/enic: flow API documentation
John Daley [Wed, 17 May 2017 22:38:11 +0000 (15:38 -0700)]
net/enic: flow API documentation

Update enic NIC guide, release notes and add flow API to the
supported features list.

Signed-off-by: John Daley <johndale@cisco.com>
Reviewed-by: Nelson Escobar <neescoba@cisco.com>
6 years agonet/enic: flow API debug
John Daley [Wed, 17 May 2017 22:38:10 +0000 (15:38 -0700)]
net/enic: flow API debug

Added a debug function to print enic filters and actions when
rte_validate_flow is called. Compiled in CONFIG_RTE_LIBRTE_ENIC_DEBUG_FLOW
is enabled and log level is INFO.

Signed-off-by: John Daley <johndale@cisco.com>
Reviewed-by: Nelson Escobar <neescoba@cisco.com>
6 years agonet/enic: flow API for Legacy NICs
John Daley [Wed, 17 May 2017 22:38:09 +0000 (15:38 -0700)]
net/enic: flow API for Legacy NICs

5-tuple exact Flow support for 1200 series adapters. This allows:
Attributes: ingress
Items: ipv4, ipv6, udp, tcp (must exactly match src/dst IP
       addresses and ports and all must be specified).
Actions: queue and void
Selectors: 'is'

Signed-off-by: John Daley <johndale@cisco.com>
Reviewed-by: Nelson Escobar <neescoba@cisco.com>
6 years agonet/enic: flow API for NICs with advanced filters disabled
John Daley [Wed, 17 May 2017 22:38:08 +0000 (15:38 -0700)]
net/enic: flow API for NICs with advanced filters disabled

Flow support for 1300 series adapters with the 'Advanced Filter'
mode disabled via the UCS management interface. This allows:
Attributes: ingress
Items: Outer eth, ipv4, ipv6, udp, sctp, tcp, vxlan. Inner eth, ipv4,
       ipv6, udp, tcp.
Actions: queue and void
Selectors: 'is', 'spec' and 'mask'. 'last' is not supported

With advanced filters disabled, an IPv4 or IPv6 item must be specified
in the pattern.

Signed-off-by: John Daley <johndale@cisco.com>
Reviewed-by: Nelson Escobar <neescoba@cisco.com>
6 years agonet/enic: flow API mark and flag support
John Daley [Wed, 17 May 2017 22:38:07 +0000 (15:38 -0700)]
net/enic: flow API mark and flag support

For VICs with filter tagging, support the MARK and FLAG actions
by setting appropriate mbuf ol_flags if there is a filter match.

Signed-off-by: John Daley <johndale@cisco.com>
Reviewed-by: Nelson Escobar <neescoba@cisco.com>
6 years agonet/enic: flow API for NICs with advanced filters enabled
John Daley [Wed, 17 May 2017 22:38:06 +0000 (15:38 -0700)]
net/enic: flow API for NICs with advanced filters enabled

Flow support for 1300 series adapters with the 'Advanced Filter'
mode enabled via the UCS management interface. This enables:
Attributes: ingress
Items: Outer eth, ipv4, ipv6, udp, sctp, tcp, vxlan. Inner eth, ipv4,
       ipv6, udp, tcp.
Actions: queue, and void
Selectors: 'is', 'spec' and 'mask'. 'last' is not supported

Signed-off-by: John Daley <johndale@cisco.com>
Reviewed-by: Nelson Escobar <neescoba@cisco.com>
6 years agonet/enic: flow API skeleton
John Daley [Wed, 17 May 2017 22:38:05 +0000 (15:38 -0700)]
net/enic: flow API skeleton

Stub callbacks for the generic flow API and a new FLOW debug define.

Signed-off-by: John Daley <johndale@cisco.com>
Reviewed-by: Nelson Escobar <neescoba@cisco.com>
6 years agonet/enic/base: bring NIC interface functions up to date
John Daley [Wed, 17 May 2017 22:38:04 +0000 (15:38 -0700)]
net/enic/base: bring NIC interface functions up to date

Update the base functions for the Cisco VIC. These files are mostly
common with other VIC drivers so are left alone is as much as possible.
Includes in a new filter/action interface which is needed for Generic
Flow API PMD support. Update FDIR code to use the new interface.

Signed-off-by: John Daley <johndale@cisco.com>
Reviewed-by: Nelson Escobar <neescoba@cisco.com>
6 years agonet/i40e/base: fix Tx error stats on VF
Wenzhuo Lu [Thu, 18 May 2017 03:29:04 +0000 (11:29 +0800)]
net/i40e/base: fix Tx error stats on VF

Unfortunately the datasheet has a mistake. The
address of the TX error counter is wrong.

Fixes: 8db9e2a1b232 ("i40e: base driver")
CC: stable@dpdk.org
Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
6 years agonet/igb: support MAC filters for i210 and i211 chips
Markus Theil [Wed, 17 May 2017 18:06:28 +0000 (20:06 +0200)]
net/igb: support MAC filters for i210 and i211 chips

i210 and i211 also support unicast MAC filters.
The patch was tested on i210 based hw, for i211
support was looked up in the specs.

Signed-off-by: Markus Theil <markus.theil@tu-ilmenau.de>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
6 years agonet/i40e: improve FDIR programming times
Michael Lilja [Wed, 17 May 2017 14:57:57 +0000 (16:57 +0200)]
net/i40e: improve FDIR programming times

Previously, the FDIR programming time is +11ms on i40e.
This patch will result in an average programming time of
22usec with a max of 60usec .

Signed-off-by: Michael Lilja <ml@napatech.com>
Acked-by: Beilei Xing <beilei.xing@intel.com>
6 years agonet/ark: fix return value of null not checked
John Miller [Tue, 16 May 2017 16:14:18 +0000 (12:14 -0400)]
net/ark: fix return value of null not checked

Coverity issue: 144517
Fixes: 1131cbf0fb2b ("net/ark: stub PMD for Atomic Rules Arkville")
Cc: stable@dpdk.org
Signed-off-by: John Miller <john.miller@atomicrules.com>
6 years agonet/ark: fix null pointer dereference
John Miller [Tue, 16 May 2017 16:14:17 +0000 (12:14 -0400)]
net/ark: fix null pointer dereference

Coverity issue: 144520
Fixes: 727b3fe292bc ("net/ark: integrate PMD")
Cc: stable@dpdk.org
Signed-off-by: John Miller <john.miller@atomicrules.com>
6 years agonet/ark: fix return code not checked
John Miller [Tue, 16 May 2017 16:14:16 +0000 (12:14 -0400)]
net/ark: fix return code not checked

Coverity issue: 144514
Fixes: 727b3fe292bc ("net/ark: integrate PMD")
Cc: stable@dpdk.org
Signed-off-by: John Miller <john.miller@atomicrules.com>
6 years agonet/ark: fix buffer not null terminated
John Miller [Tue, 16 May 2017 16:14:15 +0000 (12:14 -0400)]
net/ark: fix buffer not null terminated

Coverity issue: 144512
Coverity issue: 144513
Fixes: 9c7188a68d7b ("net/ark: provide API for hardware modules pktchkr and pktgen")
Cc: stable@dpdk.org
Signed-off-by: John Miller <john.miller@atomicrules.com>
6 years agodrivers/net: add generic ethdev macro to get PCI device
Ferruh Yigit [Mon, 15 May 2017 10:24:03 +0000 (11:24 +0100)]
drivers/net: add generic ethdev macro to get PCI device

Instead of many PMD define their own macro, define a generic one in
ethdev and use that in PMDs.

Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
Acked-by: Allain Legacy <allain.legacy@windriver.com>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
Acked-by: Andrew Rybchenko <arybchenko@solarflare.com>
6 years agodrivers/net: remove unnecessary macro for unused variables
Ferruh Yigit [Fri, 12 May 2017 10:33:03 +0000 (11:33 +0100)]
drivers/net: remove unnecessary macro for unused variables

remove __rte_unused instances that are not required.

Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
Acked-by: Allain Legacy <allain.legacy@windriver.com>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
6 years agonet/tap: do not set remote MAC if not necessary
Pascal Mazon [Fri, 12 May 2017 13:01:40 +0000 (15:01 +0200)]
net/tap: do not set remote MAC if not necessary

Check for the current MAC address on both the remote and the tap
netdevices before setting a new value.

While there, remove wrong empty lines and ensure tap_ioctl() return
value is negative, just like what is done throughout this code.

Fixes: 2bc06869cd94 ("net/tap: add remote netdevice traffic capture")

Signed-off-by: Pascal Mazon <pascal.mazon@6wind.com>
6 years agonet/tap: create netdevice during probing
Pascal Mazon [Fri, 12 May 2017 13:01:39 +0000 (15:01 +0200)]
net/tap: create netdevice during probing

This has three main benefits:
 - tun_alloc is now generic again for any queue,
 - mtu no longer needs to be handled in tap_setup_queue(),
 - an actual netdevice is created as soon as the device is probed.

On top of it, code in eth_dev_tap_create() has been reworked to have a
more logical behavior; initialization can now fail if a remote is
requested but cannot be set up.

Fixes: 2bc06869cd94 ("net/tap: add remote netdevice traffic capture")

Signed-off-by: Pascal Mazon <pascal.mazon@6wind.com>
6 years agonet/tap: drop unnecessary nested block
Pascal Mazon [Fri, 12 May 2017 13:01:38 +0000 (15:01 +0200)]
net/tap: drop unnecessary nested block

This is cosmetic; the code is functionally equivalent.

Signed-off-by: Pascal Mazon <pascal.mazon@6wind.com>
6 years agonet/tap: remove unnecessary functions
Pascal Mazon [Fri, 12 May 2017 13:01:37 +0000 (15:01 +0200)]
net/tap: remove unnecessary functions

These functions are only two lines each and are used only once.

Signed-off-by: Pascal Mazon <pascal.mazon@6wind.com>
6 years agonet/tap: add debug messages
Pascal Mazon [Fri, 12 May 2017 13:01:36 +0000 (15:01 +0200)]
net/tap: add debug messages

Print a detailed debug message inside tap_ioctl() directly. The caller
now only needs to check for return value.

Signed-off-by: Pascal Mazon <pascal.mazon@6wind.com>
6 years agonet/tap: add support for fixed MAC addresses
Pascal Mazon [Wed, 12 Apr 2017 07:30:21 +0000 (09:30 +0200)]
net/tap: add support for fixed MAC addresses

Support for a fixed MAC address for testing with the last octet
incrementing by one for each interface defined with the new 'mac=fixed'
string on the --vdev option. The default option is still to randomize
the MAC address for each tap interface.

Signed-off-by: Keith Wiles <keith.wiles@intel.com>
Signed-off-by: Pascal Mazon <pascal.mazon@6wind.com>
Acked-by: Keith Wiles <keith.wiles@intel.com>
6 years agonet/i40e: add return value checks
Herakliusz Lipiec [Tue, 9 May 2017 14:37:34 +0000 (15:37 +0100)]
net/i40e: add return value checks

Coverity issue: 1379362
Coverity issue: 1379365
Fixes: 71d35259ff67 ("i40e: tear down flow director")
Cc: stable@dpdk.org
Signed-off-by: Herakliusz Lipiec <herakliusz.lipiec@intel.com>
Acked-by: Helin Zhang <helin.zhang@intel.com>
6 years agonet/null: do not touch mbuf next or nb segs on Rx
Ferruh Yigit [Thu, 4 May 2017 15:43:58 +0000 (16:43 +0100)]
net/null: do not touch mbuf next or nb segs on Rx

mbuf next and nb_segs fields already have the default values when get
from mempool, no need to update them in PMD.

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

Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
Reviewed-by: Olivier Matz <olivier.matz@6wind.com>
6 years agonet/ixgbe: do not touch mbuf initialized fields
Ferruh Yigit [Thu, 4 May 2017 15:59:23 +0000 (16:59 +0100)]
net/ixgbe: do not touch mbuf initialized fields

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

Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
6 years agonet/thunderx: add compile-time checks
Jerin Jacob [Mon, 1 May 2017 06:40:18 +0000 (12:10 +0530)]
net/thunderx: add compile-time checks

The thunderx PMD is sensitive to the layout of the mbuf on
the RX path. Add in some compile-time checks to make sure the mbuf layout
assumptions are valid, and to provide hints to anyone changing the mbuf
where things may need to be updated.

Signed-off-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>
6 years agonet/fm10k: add compile-time checks to vector driver
Bruce Richardson [Fri, 28 Apr 2017 16:21:15 +0000 (17:21 +0100)]
net/fm10k: add compile-time checks to vector driver

The vector PMD is very sensitive to the layout of the mbuf, especially on
the RX path. Add in some compile-time checks to make sure the mbuf layout
assumptions are valid, and to provide hints to anyone changing the mbuf
where things may need to be updated.

Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
6 years agonet/i40e: add compile-time checks to vector driver
Bruce Richardson [Fri, 28 Apr 2017 16:21:14 +0000 (17:21 +0100)]
net/i40e: add compile-time checks to vector driver

The vector PMD is very sensitive to the layout of the mbuf, especially on
the RX path. Add in some compile-time checks to make sure the mbuf layout
assumptions are valid, and to provide hints to anyone changing the mbuf
where things may need to be updated.

Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
6 years agonet/ixgbe: enable ixgbe vector PMD for i686
Bruce Richardson [Fri, 28 Apr 2017 16:21:13 +0000 (17:21 +0100)]
net/ixgbe: enable ixgbe vector PMD for i686

The ixgbe vector PMD was disabled for i686 unnecessarily. Enable it.

Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
6 years agonet/ixgbe: add compile-time checks to vector driver
Bruce Richardson [Fri, 28 Apr 2017 16:21:12 +0000 (17:21 +0100)]
net/ixgbe: add compile-time checks to vector driver

The vector PMD is very sensitive to the layout of the mbuf, especially on
the RX path. Add in some compile-time checks to make sure the mbuf layout
assumptions are valid, and to provide hints to anyone changing the mbuf
where things may need to be updated.

Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
6 years agoethdev: keep port id unchanged if obtaining by name failed
Ilya Maximets [Thu, 18 May 2017 12:19:40 +0000 (15:19 +0300)]
ethdev: keep port id unchanged if obtaining by name failed

Currently, 'rte_eth_dev_get_port_by_name' changes transmitted
'port_id' unconditionally. This is undocumented and misleading
behaviour as user may expect unchanged value in case of error.

Otherwise, there is no sense having both return value and
a pointer in the function.

Fixes: 9c5b8d8b9feb ("ethdev: clean port id retrieval when attaching")

Signed-off-by: Ilya Maximets <i.maximets@samsung.com>
6 years agoapp/testpmd: add TCP flags in flow rules
Beilei Xing [Thu, 18 May 2017 09:06:12 +0000 (17:06 +0800)]
app/testpmd: add TCP flags in flow rules

This commit adds TCP flags support in flow API as
some drivers have SYN filter.

Signed-off-by: Beilei Xing <beilei.xing@intel.com>
Tested-by: Wei Zhao <wei.zhao1@intel.com>
Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>