dpdk.git
3 years agoapp/crypto-perf: fix spelling in output
Ciara Power [Mon, 18 Jan 2021 11:34:10 +0000 (11:34 +0000)]
app/crypto-perf: fix spelling in output

Fixes some spelling errors in app logs and help text.

Fixes: 7da018731c56 ("app/crypto-perf: add help option")
Fixes: f8be1786b1b8 ("app/crypto-perf: introduce performance test application")
Cc: stable@dpdk.org
Signed-off-by: Ciara Power <ciara.power@intel.com>
Acked-by: Declan Doherty <declan.doherty@intel.com>
3 years agocrypto/dpaa2_sec: support AES-CMAC integrity check
Hemant Agrawal [Thu, 14 Jan 2021 07:05:02 +0000 (12:35 +0530)]
crypto/dpaa2_sec: support AES-CMAC integrity check

This patch adds support for AES_CMAC integrity in non-security mode.
This patch modifies the camm flib to handles the AES CMAC
without conflicting the proto ALG operations. i.e. by creating
another ALG operation routine.

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Acked-by: Akhil Goyal <akhil.goyal@nxp.com>
3 years agocrypto/dpaa_sec: reduce log on queue closure
Hemant Agrawal [Thu, 14 Jan 2021 07:05:01 +0000 (12:35 +0530)]
crypto/dpaa_sec: reduce log on queue closure

If for some reason the queue is not close properly,
specially in test cases.
The QUEUE retire prints are flooding the screen.
They are not really required as WARNING.

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Acked-by: Akhil Goyal <akhil.goyal@nxp.com>
3 years agocommon/dpaax/caamflib: update ZUC-ZUC descriptor sharing
Akhil Goyal [Thu, 14 Jan 2021 07:05:00 +0000 (12:35 +0530)]
common/dpaax/caamflib: update ZUC-ZUC descriptor sharing

The descriptor sharing needed to be changed for ZUC+ZUC as we
were getting invalid CHA combination error due to sharing
being done on DECOs simultaneously.

Signed-off-by: Akhil Goyal <akhil.goyal@nxp.com>
3 years agotest/crypto: add AES-XCBC hash only test case
Hemant Agrawal [Thu, 14 Jan 2021 07:04:59 +0000 (12:34 +0530)]
test/crypto: add AES-XCBC hash only test case

This patch adds test case for AES-XCBC hash only for
Digest and Digest-verify

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Acked-by: Akhil Goyal <akhil.goyal@nxp.com>
3 years agocrypto/dpaa2_sec: support AES-XCBC-MAC
Akhil Goyal [Thu, 14 Jan 2021 07:04:58 +0000 (12:34 +0530)]
crypto/dpaa2_sec: support AES-XCBC-MAC

This patch add support for AES-XCBC-MAC for following cases
- AES-XCBC-MAC auth only
- AES-CBC/CTR + AES-XCBC-MAC (non-proto)
- AES-CBC/CTR + AES-XCBC-MAC (protocol offload)
- DES-CBC + AES-XCBC-MAC (non-proto)
- 3DES-CBC + AES-XCBC-MAC (non-proto)

Signed-off-by: Barry Cao <barry.cao@nxp.com>
Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Signed-off-by: Akhil Goyal <akhil.goyal@nxp.com>
3 years agocrypto/ccp: enable IOMMU
Amaranath Somalapuram [Fri, 25 Dec 2020 08:03:58 +0000 (13:33 +0530)]
crypto/ccp: enable IOMMU

CCP use vdev framework, and vdev framework don’t support IOMMU.
Adding custom IOMMU support for AMD CCP driver.

Signed-off-by: Amaranath Somalapuram <asomalap@amd.com>
3 years agotest/crypto: add case for enqueue/dequeue callbacks
Abhinandan Gujjar [Tue, 22 Dec 2020 14:42:43 +0000 (20:12 +0530)]
test/crypto: add case for enqueue/dequeue callbacks

This patch adds test cases for testing functionality of
enqueue and dequeue callback mechanism.

Signed-off-by: Abhinandan Gujjar <abhinandan.gujjar@intel.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
3 years agocryptodev: add enqueue and dequeue callbacks
Abhinandan Gujjar [Tue, 22 Dec 2020 14:42:42 +0000 (20:12 +0530)]
cryptodev: add enqueue and dequeue callbacks

This patch adds APIs to add/remove callback functions on crypto
enqueue/dequeue burst. The callback function will be called for
each burst of crypto ops received/sent on a given crypto device
queue pair.

Signed-off-by: Abhinandan Gujjar <abhinandan.gujjar@intel.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
Acked-by: Akhil Goyal <akhil.goyal@nxp.com>
3 years agocrypto/octeontx2: support AES-CBC SHA256-128-HMAC
Ankur Dwivedi [Sat, 19 Dec 2020 06:54:56 +0000 (12:24 +0530)]
crypto/octeontx2: support AES-CBC SHA256-128-HMAC

Support for aes-cbc sha256-128-hmac is added in lookaside protocol
mode. The functionality is verified using ipsec-secgw application.

Signed-off-by: Ankur Dwivedi <adwivedi@marvell.com>
3 years agocrypto/octeontx2: support AES-CBC SHA1-HMAC
Ankur Dwivedi [Sat, 19 Dec 2020 06:54:55 +0000 (12:24 +0530)]
crypto/octeontx2: support AES-CBC SHA1-HMAC

Support for aes-cbc sha1-hmac is added in lookaside protocol
mode. The functionality is verified using ipsec-secgw application.

Signed-off-by: Ankur Dwivedi <adwivedi@marvell.com>
3 years agocrypto/octeontx2: support CN98xx
Tejasree Kondoj [Thu, 31 Dec 2020 17:52:57 +0000 (23:22 +0530)]
crypto/octeontx2: support CN98xx

CN98xx SoC comes up with two CPT blocks wrt
CN96xx, CN93xx, to achieve higher performance.

Adding support to allocate all LFs of VF with even BDF from CPT0
and all LFs of VF with odd BDF from CPT1.
If LFs are not available in one block then they will be allocated
from alternate block.

Signed-off-by: Tejasree Kondoj <ktejasree@marvell.com>
Acked-by: Anoob Joseph <anoobj@marvell.com>
3 years agocrypto/octeontx2: support IPsec ESN and anti-replay
Tejasree Kondoj [Fri, 18 Dec 2020 14:14:10 +0000 (19:44 +0530)]
crypto/octeontx2: support IPsec ESN and anti-replay

Adding ESN and anti-replay support for lookaside IPsec.

Signed-off-by: Tejasree Kondoj <ktejasree@marvell.com>
Acked-by: Anoob Joseph <anoobj@marvell.com>
3 years agocrypto/octeontx2: make anti-replay routine generic
Tejasree Kondoj [Fri, 18 Dec 2020 14:14:09 +0000 (19:44 +0530)]
crypto/octeontx2: make anti-replay routine generic

Adding changes to make anti-replay routine common to both inline and
lookaside IPsec.

Signed-off-by: Tejasree Kondoj <ktejasree@marvell.com>
Acked-by: Anoob Joseph <anoobj@marvell.com>
3 years agocrypto/qat: fix access to uninitialized variable
Vladimir Medvedkin [Tue, 15 Dec 2020 18:23:51 +0000 (18:23 +0000)]
crypto/qat: fix access to uninitialized variable

QAT_LOG() has access to the uninitialized variable "name"

Fixes: 21792c443205 ("crypto/qat: add multi-process handling of driver ID")
Cc: stable@dpdk.org
Signed-off-by: Vladimir Medvedkin <vladimir.medvedkin@intel.com>
Acked-by: Arek Kusztal <arkadiuszx.kusztal@intel.com>
Acked-by: Declan Doherty <declan.doherty@intel.com>
3 years agoregex/mlx5: fix number of supported queues
Ori Kam [Thu, 7 Jan 2021 16:57:11 +0000 (18:57 +0200)]
regex/mlx5: fix number of supported queues

The RegEx engine has no limitation on number of queues.
This commits modifies the max supported queues reported to the application.

Fixes: fbc8c7003b93 ("regex/mlx5: add completion queue creation")
Cc: stable@dpdk.org
Signed-off-by: Ori Kam <orika@nvidia.com>
3 years agoregex/mlx5: support priority match
Ori Kam [Thu, 7 Jan 2021 16:57:10 +0000 (18:57 +0200)]
regex/mlx5: support priority match

The high priority match request flags means that the
RegEx engine should stop on the first match.

This commit add this flag check to the RegEx engine.

Signed-off-by: Ori Kam <orika@nvidia.com>
3 years agoregex/mlx5: fix support for group id
Ori Kam [Thu, 7 Jan 2021 16:57:09 +0000 (18:57 +0200)]
regex/mlx5: fix support for group id

In order to know which groups in the RegEx engine
should be used there is a need to check the req_flags.

This commit adds the missing check.

Fixes: 4d4e245ad637 ("regex/mlx5: support enqueue")
Cc: stable@dpdk.org
Signed-off-by: Ori Kam <orika@nvidia.com>
3 years agoregex/mlx5: support combined rule file
Ori Kam [Thu, 7 Jan 2021 16:57:08 +0000 (18:57 +0200)]
regex/mlx5: support combined rule file

The rof file holds programming instructions for
a given HW version.
In order to support future generation of HW it
was decided that the rof file will hold number
of rule configurations, and the driver will use
the one that matches the HW version.

In current code we force sync after each write block.
This has impact on performance.

The solution is to move the sync to the end of the
entire programming sequence.

Signed-off-by: Ori Kam <orika@nvidia.com>
3 years agoregex/mlx5: fix memory rule alignment
Ori Kam [Thu, 7 Jan 2021 16:57:07 +0000 (18:57 +0200)]
regex/mlx5: fix memory rule alignment

Due to Kernel requirement the memory allocated must be aligned to 2M.

Fixes: b34d816363b5 ("regex/mlx5: support rules import")
Cc: stable@dpdk.org
Signed-off-by: Ori Kam <orika@nvidia.com>
3 years agomaintainers: update for testpmd
Beilei Xing [Wed, 2 Dec 2020 06:33:22 +0000 (14:33 +0800)]
maintainers: update for testpmd

Replace Beilei Xing with Xiaoyun Li.

Signed-off-by: Beilei Xing <beilei.xing@intel.com>
Acked-by: Ferruh Yigit <ferruh.yigit@intel.com>
Acked-by: Xiaoyun Li <xiaoyun.li@intel.com>
3 years agopdump: cleanup logs and variables
Stephen Hemminger [Tue, 29 Dec 2020 20:09:01 +0000 (12:09 -0800)]
pdump: cleanup logs and variables

Checkpatch prefers 'unsigned int' over bare 'unsigned'.
Reword the error messages for brevity and clarity
so they don't have to be split across multiple lines.

Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
3 years agopdump: replace constant for device name size
Stephen Hemminger [Tue, 29 Dec 2020 20:09:00 +0000 (12:09 -0800)]
pdump: replace constant for device name size

The device string has an existing size in rte_dev.h
use that instead of defining our own.

Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
3 years agopdump: free mbuf in bulk
Stephen Hemminger [Tue, 29 Dec 2020 20:08:59 +0000 (12:08 -0800)]
pdump: free mbuf in bulk

Use rte_pktmbuf_free_bulk instead of loop when freeing
packets.

Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Acked-by: Morten Brørup <mb@smartsharesystems.com>
3 years agodoc: recommend GitHub Actions for CI
David Marchand [Wed, 13 Jan 2021 09:03:12 +0000 (10:03 +0100)]
doc: recommend GitHub Actions for CI

Update the contributing guidelines to describe GitHub Actions first and
add a warning about Travis usage.

Fixes: 87009585e293 ("ci: hook to GitHub Actions")

Signed-off-by: David Marchand <david.marchand@redhat.com>
Acked-by: Aaron Conole <aconole@redhat.com>
3 years agoexamples/l3fwd: remove limitation on Tx queue count
Harman Kalra [Tue, 12 Jan 2021 18:24:46 +0000 (23:54 +0530)]
examples/l3fwd: remove limitation on Tx queue count

In l3fwd no of transmit queues is calculated based on no of
lcores with which it is launched. Hence maximum no of tx
queues possible per port should depend on RTE_MAX_LCORE value.

Fixes: af75078fece3 ("first public release")
Cc: stable@dpdk.org
Signed-off-by: Harman Kalra <hkalra@marvell.com>
Reviewed-by: David Marchand <david.marchand@redhat.com>
3 years agofbarray: fix overlap check
Anatoly Burakov [Thu, 14 Jan 2021 15:02:45 +0000 (15:02 +0000)]
fbarray: fix overlap check

When we're attaching fbarrays in secondary processes, we check for
whether the intended memory address for the fbarray is already in use by
some other, local fbarray. However, the check for end-overlap (i.e. to
see if our memory area's end overlaps with some other fbarray) is
incorrectly counting end offset as part of the overlap. Fix the check.

Fixes: 5b61c62cfd76 ("fbarray: add internal tailq for mapped areas")
Cc: stable@dpdk.org
Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Zhihong Peng <zhihongx.peng@intel.com>
3 years agoeal/linux: improve no hugepages logging
Anatoly Burakov [Tue, 10 Nov 2020 15:41:36 +0000 (15:41 +0000)]
eal/linux: improve no hugepages logging

When no hugepages are found, we log a message about it, but we never
specify on which node. We also implicitly declare the page size based
on the directory name, but that's not very user friendly.

Fix both by changing the text of the message to note the NUMA node (if
applicable) and explicitly mention page size in kilobytes.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Reviewed-by: David Marchand <david.marchand@redhat.com>
3 years agotest: add ticketlock to fast tests
Stephen Hemminger [Thu, 14 Jan 2021 16:58:55 +0000 (08:58 -0800)]
test: add ticketlock to fast tests

The ticketlock test is fast and should be run all the time.

Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Acked-by: Aaron Conole <aconole@redhat.com>
3 years agotest: fix buffer overflow in Tx burst
Alvin Zhang [Wed, 25 Nov 2020 09:46:22 +0000 (17:46 +0800)]
test: fix buffer overflow in Tx burst

The Tx buffer may overflow when there is more than one port.

Fixes: 002ade70e933 ("app/test: measure cycles per packet in Rx/Tx")
Cc: stable@dpdk.org
Signed-off-by: Alvin Zhang <alvinx.zhang@intel.com>
Acked-by: Jeff Guo <jia.guo@intel.com>
Tested-by: Wei Ling <weix.ling@intel.com>
3 years agonet/ice: implement power management API
Liang Ma [Thu, 14 Jan 2021 14:46:12 +0000 (14:46 +0000)]
net/ice: implement power management API

Implement support for the power management API by implementing a
`get_monitor_addr` function that will return an address of an RX ring's
status bit.

Signed-off-by: Liang Ma <liang.j.ma@intel.com>
Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
3 years agonet/i40e: implement power management API
Liang Ma [Thu, 14 Jan 2021 14:46:11 +0000 (14:46 +0000)]
net/i40e: implement power management API

Implement support for the power management API by implementing a
`get_monitor_addr` function that will return an address of an RX ring's
status bit.

Signed-off-by: Liang Ma <liang.j.ma@intel.com>
Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
Acked-by: Jeff Guo <jia.guo@intel.com>
3 years agonet/ixgbe: implement power management API
Liang Ma [Thu, 14 Jan 2021 14:46:10 +0000 (14:46 +0000)]
net/ixgbe: implement power management API

Implement support for the power management API by implementing a
`get_monitor_addr` function that will return an address of an RX ring's
status bit.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Signed-off-by: Liang Ma <liang.j.ma@intel.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
3 years agoethdev: add simple power management API
Liang Ma [Thu, 14 Jan 2021 14:46:08 +0000 (14:46 +0000)]
ethdev: add simple power management API

Add a simple API to allow getting the monitor conditions for
power-optimized monitoring of the Rx queues from the PMD, as well as
release notes information.

Signed-off-by: Liang Ma <liang.j.ma@intel.com>
Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Acked-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
3 years agoeal: add monitor wakeup function
Anatoly Burakov [Thu, 14 Jan 2021 14:46:07 +0000 (14:46 +0000)]
eal: add monitor wakeup function

Now that we have everything in a C file, we can store the information
about our sleep, and have a native mechanism to wake up the sleeping
core. This mechanism would however only wake up a core that's sleeping
while monitoring - waking up from `rte_power_pause` won't work.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
3 years agoeal: remove sync version of power monitor
Anatoly Burakov [Thu, 14 Jan 2021 14:46:06 +0000 (14:46 +0000)]
eal: remove sync version of power monitor

Currently, the "sync" version of power monitor intrinsic is supposed to
be used for purposes of waking up a sleeping core. However, there are
better ways to achieve the same result, so remove the unneeded function.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
3 years agoeal: change power intrinsics API
Anatoly Burakov [Thu, 14 Jan 2021 14:46:05 +0000 (14:46 +0000)]
eal: change power intrinsics API

Instead of passing around pointers and integers, collect everything
into struct. This makes API design around these intrinsics much easier.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
3 years agoeal: avoid invalid power intrinsics API usage
Anatoly Burakov [Thu, 14 Jan 2021 14:46:04 +0000 (14:46 +0000)]
eal: avoid invalid power intrinsics API usage

Currently, the API documentation mandates that if the user wants to use
the power management intrinsics, they need to call the
`rte_cpu_get_intrinsics_support` API and check support for specific
intrinsics.

However, if the user does not do that, it is possible to get illegal
instruction error because we're using raw instruction opcodes, which may
or may not be supported at runtime.

Now that we have everything in a C file, we can check for support at
startup and prevent the user from possibly encountering illegal
instruction errors.

We also add return values to the API's as well, because why not.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
3 years agoeal: uninline power intrinsics
Anatoly Burakov [Thu, 14 Jan 2021 14:46:03 +0000 (14:46 +0000)]
eal: uninline power intrinsics

Currently, power intrinsics are inline functions. Make them part of the
ABI so that we can have various internal data associated with them
without exposing said data to the outside world.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
3 years agoconfig/arm: disable libnuma in cross builds
Juraj Linkeš [Fri, 15 Jan 2021 13:26:08 +0000 (14:26 +0100)]
config/arm: disable libnuma in cross builds

Some Arm SoCs are not NUMA systems. Add the capability to disable NUMA
for cross build and disable NUMA in Arm cross files.

Signed-off-by: Juraj Linkeš <juraj.linkes@pantheon.tech>
Reviewed-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>
Tested-by: Jerin Jacob <jerinj@marvell.com>
Tested-by: Pavan Nikhilesh <pbhagavatula@marvell.com>
Tested-by: Vimal Chungath <vcchunga@amazon.com>
Tested-by: Dharmik Thakkar <dharmik.thakkar@arm.com>
3 years agoconfig/arm: add core and NUMA counts to cross files
Juraj Linkeš [Fri, 15 Jan 2021 13:26:06 +0000 (14:26 +0100)]
config/arm: add core and NUMA counts to cross files

Add support for setting core count and numa nodes in cross files. The
values specified in cross files will override the default values.
Also add missing default values to Arm config.

Signed-off-by: Juraj Linkeš <juraj.linkes@pantheon.tech>
Reviewed-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>
Tested-by: Jerin Jacob <jerinj@marvell.com>
Tested-by: Pavan Nikhilesh <pbhagavatula@marvell.com>
Tested-by: Vimal Chungath <vcchunga@amazon.com>
Tested-by: Dharmik Thakkar <dharmik.thakkar@arm.com>
3 years agoconfig/arm: use native machine build arguments
Juraj Linkeš [Fri, 15 Jan 2021 13:26:05 +0000 (14:26 +0100)]
config/arm: use native machine build arguments

Letting the compiler decide is going to yield the best results for
native builds, so use native machine args usable for both GCC and Clang.

Signed-off-by: Juraj Linkeš <juraj.linkes@pantheon.tech>
Reviewed-by: Ruifeng Wang <ruifeng.wang@arm.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>
Tested-by: Jerin Jacob <jerinj@marvell.com>
Tested-by: Pavan Nikhilesh <pbhagavatula@marvell.com>
Tested-by: Vimal Chungath <vcchunga@amazon.com>
Tested-by: Dharmik Thakkar <dharmik.thakkar@arm.com>
3 years agoconfig/arm: isolate generic build
Juraj Linkeš [Fri, 15 Jan 2021 13:26:04 +0000 (14:26 +0100)]
config/arm: isolate generic build

Use generic configuration for the only build where it makes sense - the
generic build. For other builds, if we don't know either of implementer
ID or part number, the build is not supported.
Add part numbers to cross files where fallback to generic configuration
is assumed.

Signed-off-by: Juraj Linkeš <juraj.linkes@pantheon.tech>
Reviewed-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>
Tested-by: Jerin Jacob <jerinj@marvell.com>
Tested-by: Pavan Nikhilesh <pbhagavatula@marvell.com>
Tested-by: Vimal Chungath <vcchunga@amazon.com>
Tested-by: Dharmik Thakkar <dharmik.thakkar@arm.com>
3 years agoconfig/arm: organize config into dict
Juraj Linkeš [Fri, 15 Jan 2021 13:26:03 +0000 (14:26 +0100)]
config/arm: organize config into dict

Use dictionary lookup instead of checking for existing variables,
iterating over all elements in the list or checking lists for optional
configuration. Move variable contents into the dictionary for variables
that would be referenced only once.
Fallback to generic part number if the discovered part number is
unknown.

Signed-off-by: Juraj Linkeš <juraj.linkes@pantheon.tech>
Reviewed-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com>
Reviewed-by: Ruifeng Wang <ruifeng.wang@arm.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>
Tested-by: Jerin Jacob <jerinj@marvell.com>
Tested-by: Pavan Nikhilesh <pbhagavatula@marvell.com>
Tested-by: Vimal Chungath <vcchunga@amazon.com>
Tested-by: Dharmik Thakkar <dharmik.thakkar@arm.com>
3 years agoconfig/arm: simplify how flags are processed
Juraj Linkeš [Fri, 15 Jan 2021 13:26:02 +0000 (14:26 +0100)]
config/arm: simplify how flags are processed

Set flags in one loop. Append flags to a list and use the list in the
loop.

Signed-off-by: Juraj Linkeš <juraj.linkes@pantheon.tech>
Reviewed-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>
Tested-by: Jerin Jacob <jerinj@marvell.com>
Tested-by: Pavan Nikhilesh <pbhagavatula@marvell.com>
Tested-by: Vimal Chungath <vcchunga@amazon.com>
Tested-by: Dharmik Thakkar <dharmik.thakkar@arm.com>
3 years agoconfig/arm: reformat and move config and comments
Juraj Linkeš [Fri, 15 Jan 2021 13:26:01 +0000 (14:26 +0100)]
config/arm: reformat and move config and comments

Change formatting so that it's more consistent and readable, add/modify
comments/stdout messages, move configuration options to more appropriate
places and make the order consistent according to these rules:
1. First list generic configuration options, then list options that may
   be overwritten. List SoC-specific options last.
2. For SoC-specific options, list number of cores before the number of
   NUMA nodes, to make it consistent with config/meson.build.

Signed-off-by: Juraj Linkeš <juraj.linkes@pantheon.tech>
Reviewed-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>
Tested-by: Jerin Jacob <jerinj@marvell.com>
Tested-by: Pavan Nikhilesh <pbhagavatula@marvell.com>
Tested-by: Vimal Chungath <vcchunga@amazon.com>
Tested-by: Dharmik Thakkar <dharmik.thakkar@arm.com>
3 years agoconfig/arm: remove unused or superfluous variables
Juraj Linkeš [Fri, 15 Jan 2021 13:26:00 +0000 (14:26 +0100)]
config/arm: remove unused or superfluous variables

Remove variables that were either not used, referenced just once or not
needed.

Signed-off-by: Juraj Linkeš <juraj.linkes@pantheon.tech>
Reviewed-by: Ruifeng Wang <ruifeng.wang@arm.com>
Reviewed-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>
Tested-by: Jerin Jacob <jerinj@marvell.com>
Tested-by: Pavan Nikhilesh <pbhagavatula@marvell.com>
Tested-by: Vimal Chungath <vcchunga@amazon.com>
Tested-by: Dharmik Thakkar <dharmik.thakkar@arm.com>
3 years agoconfig/arm: rename variables
Juraj Linkeš [Fri, 15 Jan 2021 13:25:59 +0000 (14:25 +0100)]
config/arm: rename variables

Rename Arm build variables and values so that they better conform to Arm
specifications. Also rename generically sounding variable to names that
better capture what the variables hold.

Rename machine_args_generic to part_number_config_arm since the
variable contains more than just the generic machine args and is used
mainly as the fallback arm configuration.
Rename the default machine args to generic machine args to reflect that.
The rest of the variables are self-explanatory.

Signed-off-by: Juraj Linkeš <juraj.linkes@pantheon.tech>
Reviewed-by: Ruifeng Wang <ruifeng.wang@arm.com>
Reviewed-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>
Tested-by: Jerin Jacob <jerinj@marvell.com>
Tested-by: Pavan Nikhilesh <pbhagavatula@marvell.com>
Tested-by: Vimal Chungath <vcchunga@amazon.com>
Tested-by: Dharmik Thakkar <dharmik.thakkar@arm.com>
3 years agocfgfile: build on Windows
Narcisa Vasile [Tue, 5 Jan 2021 02:35:22 +0000 (18:35 -0800)]
cfgfile: build on Windows

The librte_cfgfile lib is functional on Windows.
Enable compilation of this lib for Windows.

Signed-off-by: Narcisa Vasile <navasile@linux.microsoft.com>
3 years agoraw/ioat: support secondary process
Kumar Amber [Fri, 8 Jan 2021 16:52:16 +0000 (22:22 +0530)]
raw/ioat: support secondary process

Add support for secondary processes in ioat devices. The update
allocates a memzone for a primary process or returns it in a
secondary process.

Signed-off-by: Kumar Amber <kumar.amber@intel.com>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
3 years agobitmap: support 128-byte cacheline in empty check
Nithin Dabilpuram [Tue, 5 Jan 2021 08:57:29 +0000 (14:27 +0530)]
bitmap: support 128-byte cacheline in empty check

Currently bitmap line not empty check API assumes cache line
of 64B and only checks 8 slabs. Since in 128B cacheline, we
have 16 slabs per cacheline, rte_bitmap_clear() will mark
complete line as empty as soon as 8 slabs are empty thereby
breaking bitmap scan functionality. Fix it by defining new
__rte_bitmap_line_not_empty() for 128B cacheline platform.

Signed-off-by: Nithin Dabilpuram <ndabilpuram@marvell.com>
Acked-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agotest/bitmap: test scan after half cacheline is cleared
Nithin Dabilpuram [Tue, 5 Jan 2021 08:57:28 +0000 (14:27 +0530)]
test/bitmap: test scan after half cacheline is cleared

Add a test case to test scan operation post clear of half
cacheline of slabs.

Also fix meson.build to include test_bitmap.c in the compilation.

Signed-off-by: Nithin Dabilpuram <ndabilpuram@marvell.com>
3 years agonet/netvsc: support VF device hot add/remove
Long Li [Mon, 21 Dec 2020 21:33:22 +0000 (13:33 -0800)]
net/netvsc: support VF device hot add/remove

When a VF device is present, netvsc can send or receive packets over the
VF device. The VF device driver communicates directly with the PCI device
via the PF from the host hypervisor. This is faster than exchanging data
with netvsp via vmbus, i.e. syntheic path.

In Azure and Hyper-v environments, VF device can be hot added or hot
removed at anytime while guest VM is running. This patch improves netvsc
to support VF device hot add/remove.

1. netvsc monitors all system hot add activities over the PCI bus. When it
detects a VF device is added to the system and is managed under this
netvsc device, it asks EAL to probe and start this VF device, then it
attaches and switches data path to the VF device.

2. After a VF device is attached to netvsc, netvsc monitors this device on
hot remove. When this VF device is hot removed, netvsc switches data path
to synthetic, stops this VF device and removes it from EAL.

3. If any failure happens during a VF device hot remove or add, the netvsc
falls back to synthetic path for all data traffic.

Signed-off-by: Long Li <longli@microsoft.com>
3 years agoeal/linux: allow multiple starts of event monitor
Long Li [Mon, 21 Dec 2020 21:32:36 +0000 (13:32 -0800)]
eal/linux: allow multiple starts of event monitor

In some cases, a device or infrastructure may want to enable hotplug
but application may also try and start hotplug as well. Therefore
change the monitor_started from a boolean into a reference count.

Signed-off-by: Long Li <longli@microsoft.com>
3 years agotest: fix terminal settings on exit
Stephen Hemminger [Thu, 14 Jan 2021 16:45:37 +0000 (08:45 -0800)]
test: fix terminal settings on exit

When running one test (via DPDK_TEST) the test program
would leave the terminal in raw mode.  This was because
it was setting up cmdline to do interactive input.

The fix is to use cmdline_new() for the interactive case.

This also fixes a memory leak because the test
runner was never calling cmdline_free().

Fixes: 9b848774a5dc ("test: use env variable to run tests")
Cc: stable@dpdk.org
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Tested-by: Harry van Haaren <harry.van.haaren@intel.com>
3 years agotest/rwlock: fix spelling and missing whitespace
Stephen Hemminger [Thu, 14 Jan 2021 16:48:37 +0000 (08:48 -0800)]
test/rwlock: fix spelling and missing whitespace

Trivial fix to for spelling errors and incorrect spacing.
No change to any built code.

Fixes: 7a61fc5d1b09 ("test/rwlock: add new test-cases")
Fixes: af75078fece3 ("first public release")
Cc: stable@dpdk.org
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
3 years agoeal/windows: fix C++ compatibility
Tyler Retzlaff [Fri, 15 Jan 2021 19:38:21 +0000 (11:38 -0800)]
eal/windows: fix C++ compatibility

Explicitly cast void * to type * so that EAL headers may be compiled
as C or C++.

Fixes: e8428a9d89f1 ("eal/windows: add some basic functions and macros")
Cc: stable@dpdk.org
Signed-off-by: Tyler Retzlaff <roretzla@linux.microsoft.com>
3 years agomaintainers: update for hns3
Lijun Ou [Sat, 16 Jan 2021 09:02:09 +0000 (17:02 +0800)]
maintainers: update for hns3

Because Wei Hu has changed to a new job and the
email address (xavier.huwei@huawei.com) has expired,
we remove him from the hns3 maintainer list.

All patches signed-off-by Wei Hu will be copied to Lijun Ou.

Signed-off-by: Lijun Ou <oulijun@huawei.com>
3 years agodevtools: reduce ABI checks and static binaries
Thomas Monjalon [Wed, 2 Dec 2020 16:47:03 +0000 (17:47 +0100)]
devtools: reduce ABI checks and static binaries

When testing compilation and checking ABI compatibility,
there is no real need of static binaries eating disks.

The static linkage of applications was already well tested,
though the static examples tested with meson were limited to "l3fwd" only.
The static build test with make is limited to "helloworld" example.

The ABI compatibility is checked on shared libraries,
and there is no need to test again on similar builds.
A new parameter is added to the function "build",
so the ABI check is enabled only for native gcc and clang shared builds,
32-bit, generic armv8 and ppc cross compilations.
In other words, it is disabled for some static builds and some Arm ones.

Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: David Marchand <david.marchand@redhat.com>
3 years agotest/mcslock: remove unneeded per lcore copy
Olivier Matz [Wed, 4 Nov 2020 17:04:25 +0000 (18:04 +0100)]
test/mcslock: remove unneeded per lcore copy

Each core already comes with its local storage for mcslock (in its
stack), therefore there is no need to define an additional per-lcore
mcslock.

Fixes: 32dcb9fd2a22 ("test/mcslock: add MCS queued lock unit test")

Signed-off-by: Olivier Matz <olivier.matz@6wind.com>
Reviewed-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com>
3 years agoservice: propagate init error in EAL
Olivier Matz [Wed, 13 Jan 2021 08:28:06 +0000 (09:28 +0100)]
service: propagate init error in EAL

Currently, when rte_service_init() fails at initialization, the
application always gets a ENOEXEC error code. For example, with testpmd,
this is displayed as:

  Cannot init EAL: Exec format error

This error code does not describe the real issue. Instead, use the error
code returned by the function.

Fixes: e39824500825 ("service: initialize with EAL")
Cc: stable@dpdk.org
Signed-off-by: Olivier Matz <olivier.matz@6wind.com>
Acked-by: Harry van Haaren <harry.van.haaren@intel.com>
3 years agoeal/windows: build reciprocal division functions
Tyler Retzlaff [Mon, 11 Jan 2021 23:16:36 +0000 (15:16 -0800)]
eal/windows: build reciprocal division functions

Build rte_reciprocal.c and export the following functions on windows:
  * rte_reciprocal_value
  * rte_reciprocal_value_u64

Signed-off-by: Tyler Retzlaff <roretzla@linux.microsoft.com>
Acked-by: Dmitry Kozlyuk <dmitry.kozliuk@gmail.com>
3 years agobus/pci: fix build with Windows SDK >= 10.0.20253
Tyler Retzlaff [Thu, 14 Jan 2021 21:22:35 +0000 (13:22 -0800)]
bus/pci: fix build with Windows SDK >= 10.0.20253

NetUIO device class and interface GUIDs are defined in system
headers starting from platform SDK v10.0.20253. Inspect SDK
version to avoid redefinition.

Pre-release SDKs do not promise compatibility and a narrow
subset of SDKs may still be subject to redefinition.

Fixes: c76ec01b4591 (bus/pci: support netuio on Windows)
Cc: stable@dpdk.org
Signed-off-by: Tyler Retzlaff <roretzla@linux.microsoft.com>
Acked-by: Dmitry Kozlyuk <dmitry.kozliuk@gmail.com>
Acked-by: Ranjit Menon <ranjit.menon@intel.com>
3 years agodoc: fix figure numbering in graph guide
Thomas Monjalon [Wed, 6 Jan 2021 09:19:43 +0000 (10:19 +0100)]
doc: fix figure numbering in graph guide

Some figures had a title inside the picture but not in RST file.
As a consequence, some versions of Sphinx are emitting a warning.

Warning, treated as error:
doc/guides/prog_guide/graph_lib.rst:64:
no number is assigned for figure: figure-anatomy-of-a-node

The titles are moved from SVG to RST,
except for graph_mem_layout.svg where in-picture title must be kept.

Fixes: 4dc6d8e63c16 ("doc: add graph library guide")
Cc: stable@dpdk.org
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Reviewed-by: Jerin Jacob <jerinj@marvell.com>
3 years agobus/dpaa: optimize device name parsing
Gaetan Rivet [Mon, 9 Nov 2020 13:37:55 +0000 (14:37 +0100)]
bus/dpaa: optimize device name parsing

Device name parsing is done on all buses during device iterations at
either EAL or ethdev levels.

When a bus implements device name parsing slowly, all iterations are
impacted. Efficient implementation is important.

The DPAA bus device name parsing has two issues: it allocates dynamic
memory and uses snprintf without a real need for it. Both can be
avoided, which improves the parsing performance.

The function is also simpler and shorter.

Signed-off-by: Gaetan Rivet <grive@u256.net>
3 years agoip_frag: remove padding length of fragment
Yicai Lu [Wed, 16 Dec 2020 13:36:30 +0000 (21:36 +0800)]
ip_frag: remove padding length of fragment

In some situations, we would get several ip fragments, which total
data length is less than min_ip_len(64) and padding with zeros.
We simulated intermediate fragments by modifying the MTU.
To illustrate the problem, we simplify the packet format and
ignore the impact of the packet header.In namespace2,
a packet whose data length is 1520 is sent.
When the packet passes tap2, the packet is divided into two
fragments: fragment A and B, similar to (1520 = 1510 + 10).
When the packet passes tap3, the larger fragment packet A is
divided into two fragments A1 and A2, similar to (1510 = 1500 + 10).
Finally, the bond interface receives three fragments:
A1, A2, and B (1520 = 1500 + 10 + 10).
One fragmented packet A2 is smaller than the minimum Ethernet
frame length, so it needs to be padded.

|---------------------------------------------------|
|                      HOST                         |
| |--------------|   |----------------------------| |
| |      ns2     |   |      |--------------|      | |
| |  |--------|  |   |  |--------|    |--------|  | |
| |  |  tap1  |  |   |  |  tap2  | ns1|  tap3  |  | |
| |  |mtu=1510|  |   |  |mtu=1510|    |mtu=1500|  | |
| |--|1.1.1.1 |--|   |--|1.1.1.2 |----|2.1.1.1 |--| |
|    |--------|         |--------|    |--------|    |
|         |                 |              |        |
|         |-----------------|              |        |
|                                          |        |
|                                      |--------|   |
|                                      |  bond  |   |
|--------------------------------------|mtu=1500|---|
                                       |--------|

When processing the preceding packets above,
DPDK would aggregate fragmented packets A2 and B.
And error packets are generated, which padding(zero)
is displayed in the middle of the packet.

A2 + B:
0000   fa 16 3e 9f fb 82 fa 47 b2 57 dc 20 08 00 45 00
0010   00 33 b4 66 00 ba 3f 01 c1 a5 01 01 01 01 02 01
0020   01 02 c0 c1 c2 c3 c4 c5 c6 c7 00 00 00 00 00 00
0030   00 00 00 00 00 00 00 00 00 00 00 00 c8 c9 ca cb
0040   cc cd ce cf d0 d1 d2 d3 d4 d5 d6 d7 d8 d9 da db
0050   dc dd de df e0 e1 e2 e3 e4 e5 e6

So, we would calculate the length of padding, and remove
the padding in pkt_len and data_len before aggregation.
And also we have the fix for both ipv4 and ipv6.

Fixes: 7f0983ee331c ("ip_frag: check fragment length of incoming packet")
Cc: stable@dpdk.org
Signed-off-by: Yicai Lu <luyicai@huawei.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
3 years agogso: support VXLAN UDP/IPv4
Yi Yang [Thu, 19 Nov 2020 06:43:31 +0000 (14:43 +0800)]
gso: support VXLAN UDP/IPv4

As most NICs do not support segmentation for VXLAN-encapsulated
UDP/IPv4 packets, this patch adds VXLAN UDP/IPv4 GSO support.

Signed-off-by: Yi Yang <yangyi01@inspur.com>
Acked-by: Jiayu Hu <jiayu.hu@intel.com>
3 years agodrivers/net: build i40e and mlx5 on Windows
Pallavi Kadam [Tue, 22 Dec 2020 00:45:11 +0000 (16:45 -0800)]
drivers/net: build i40e and mlx5 on Windows

Allows i40e and mlx5 PMDs to compile on Windows and disable other drivers.

Disable few i40e warnings with Clang such as comparison of integers of
different signs and macro redefinitions.

Signed-off-by: Pallavi Kadam <pallavi.kadam@intel.com>
Reviewed-by: Ranjit Menon <ranjit.menon@intel.com>
Acked-by: Tal Shnaiderman <talshn@nvidia.com>
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
3 years agoeal/windows: add random function
Pallavi Kadam [Tue, 22 Dec 2020 00:45:10 +0000 (16:45 -0800)]
eal/windows: add random function

The file rte_random.c is required to build i40e PMD on Windows.
Add rte_rand variable to export file.

Redefine _m_prefetchw for Clang toolchain due to following error
with respect to conflicting types:

FAILED: lib/76b5a35@@rte_eal@sta/librte_eal_common_rte_random.c.obj
clang @lib/76b5a35@@rte_eal@sta/librte_eal_common_rte_random.c.obj.rsp
In file included from ../lib/librte_eal/common/rte_random.c:13:
In file included from ..\lib/librte_eal/include\rte_eal.h:20:
In file included from ..\lib/librte_eal/include\rte_per_lcore.h:25:
In file included from ..\lib/librte_eal/windows/include\pthread.h:21:
In file included from ..\lib/librte_eal/windows/include\rte_windows.h:27:
In file included from C:\Program Files (x86)\Windows Kits\10\include\
10.0.18362.0\um\windows.h:171:
In file included from C:\Program Files (x86)\Windows Kits\10\include\
10.0.18362.0\shared\windef.h:24:
In file included from C:\Program Files (x86)\Windows Kits\10\include\
10.0.18362.0\shared\minwindef.h:182:
C:\Program Files (x86)\Windows Kits\10\include\10.0.18362.0\um\
winnt.h:3324:1: error: conflicting types for '_m_prefetchw'
_m_prefetchw (
^
C:\Program Files\LLVM\lib\clang\10.0.0\include\prfchwintrin.h:50:1:
note: previous definition is here
_m_prefetchw(void *__P)
^
1 error generated.

Signed-off-by: Pallavi Kadam <pallavi.kadam@intel.com>
Reviewed-by: Ranjit Menon <ranjit.menon@intel.com>
3 years agodoc: add Windows support for mlx5
Tal Shnaiderman [Sun, 3 Jan 2021 10:28:27 +0000 (12:28 +0200)]
doc: add Windows support for mlx5

Windows is supported by mlx5 PMD.
The mlx5 guide is updated with the needed information.

Signed-off-by: Tal Shnaiderman <talshn@nvidia.com>
3 years agocommon/mlx5: enable compilation on Windows
Ophir Munk [Tue, 12 Jan 2021 12:58:39 +0000 (14:58 +0200)]
common/mlx5: enable compilation on Windows

Enable mlx5 common driver on Windows with clang compilation.

Signed-off-by: Ophir Munk <ophirmu@nvidia.com>
Signed-off-by: Tal Shnaiderman <talshn@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
3 years agocommon/mlx5: fix pointer cast on Windows
Tal Shnaiderman [Thu, 7 Jan 2021 11:45:45 +0000 (13:45 +0200)]
common/mlx5: fix pointer cast on Windows

While compiling with clang 11 the callers of the
__mlx5_bit_off macro warns on the cast of pointers to
unsigned long which is a smaller int type in Windows.

warning: cast to smaller integer type 'unsigned long'
from 'u8 (*)[16]' [-Wpointer-to-int-cast]

To resolve it the type is changed to uintptr_t to be
compatible for both Linux and Windows.

Fixes: 865a0c15672c ("net/mlx5: add Direct Verbs prepare function")
Cc: stable@dpdk.org
Signed-off-by: Tal Shnaiderman <talshn@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
3 years agocommon/mlx5: remove doorbell allocation functions
Michael Baum [Wed, 6 Jan 2021 08:19:41 +0000 (08:19 +0000)]
common/mlx5: remove doorbell allocation functions

The mlx5_devx_dbr_page structure was used to allocate and release the
umem of the doorbells.
Since doorbell and buffer have used same umem, this structure is
useless.

Signed-off-by: Michael Baum <michaelba@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
3 years agonet/mlx5: move Rx RQ creation to common
Michael Baum [Wed, 6 Jan 2021 08:19:40 +0000 (08:19 +0000)]
net/mlx5: move Rx RQ creation to common

Using common function for Rx RQ creation.

Signed-off-by: Michael Baum <michaelba@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
3 years agocommon/mlx5: share DevX RQ creation
Michael Baum [Wed, 6 Jan 2021 08:19:39 +0000 (08:19 +0000)]
common/mlx5: share DevX RQ creation

The RQ object in DevX is used currently only in net driver, but it is
shared for future.

Add a structure that contains all the resources, and provide creation
and release functions for it.

Signed-off-by: Michael Baum <michaelba@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
3 years agonet/mlx5: move ASO SQ creation to common
Michael Baum [Wed, 6 Jan 2021 08:19:38 +0000 (08:19 +0000)]
net/mlx5: move ASO SQ creation to common

Using common function for ASO SQ creation.

Signed-off-by: Michael Baum <michaelba@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
3 years agonet/mlx5: move Tx SQ creation to common
Michael Baum [Wed, 6 Jan 2021 08:19:37 +0000 (08:19 +0000)]
net/mlx5: move Tx SQ creation to common

Using common function for Tx SQ creation.

Signed-off-by: Michael Baum <michaelba@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
3 years agonet/mlx5: move rearm and clock queue SQ creation to common
Michael Baum [Wed, 6 Jan 2021 08:19:36 +0000 (08:19 +0000)]
net/mlx5: move rearm and clock queue SQ creation to common

Using common function for DevX SQ creation for rearm and clock queue.

Signed-off-by: Michael Baum <michaelba@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
3 years agoregex/mlx5: move DevX SQ creation to common
Michael Baum [Wed, 6 Jan 2021 08:19:35 +0000 (08:19 +0000)]
regex/mlx5: move DevX SQ creation to common

Using common function for DevX SQ creation.

Signed-off-by: Michael Baum <michaelba@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
3 years agocommon/mlx5: share DevX SQ creation
Michael Baum [Wed, 6 Jan 2021 08:19:34 +0000 (08:19 +0000)]
common/mlx5: share DevX SQ creation

The SQ object in DevX is created in several places and in several
different drivers.
In all places almost all the details are the same, and in particular the
allocations of the required resources.

Add a structure that contains all the resources, and provide creation
and release functions for it.

Signed-off-by: Michael Baum <michaelba@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
3 years agocommon/mlx5: enhance page size configuration
Michael Baum [Wed, 6 Jan 2021 08:19:33 +0000 (08:19 +0000)]
common/mlx5: enhance page size configuration

The PRM calculates page size in 4K, so need to reduce the log_wq_pg_sz
attribute.

Signed-off-by: Michael Baum <michaelba@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
3 years agonet/mlx5: move Rx CQ creation to common
Michael Baum [Wed, 6 Jan 2021 08:19:32 +0000 (08:19 +0000)]
net/mlx5: move Rx CQ creation to common

Using common function for Rx CQ creation.

Signed-off-by: Michael Baum <michaelba@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
3 years agonet/mlx5: move Tx CQ creation to common
Michael Baum [Wed, 6 Jan 2021 08:19:31 +0000 (08:19 +0000)]
net/mlx5: move Tx CQ creation to common

Using common function for Tx CQ creation.

Signed-off-by: Michael Baum <michaelba@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
3 years agonet/mlx5: move ASO CQ creation to common
Michael Baum [Wed, 6 Jan 2021 08:19:30 +0000 (08:19 +0000)]
net/mlx5: move ASO CQ creation to common

Use common function for ASO CQ creation.

Signed-off-by: Michael Baum <michaelba@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
3 years agonet/mlx5: move rearm and clock queue CQ creation to common
Michael Baum [Wed, 6 Jan 2021 08:19:29 +0000 (08:19 +0000)]
net/mlx5: move rearm and clock queue CQ creation to common

Using common function for CQ creation at rearm queue and clock queue.

Signed-off-by: Michael Baum <michaelba@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
3 years agovdpa/mlx5: move DevX CQ creation to common
Michael Baum [Wed, 6 Jan 2021 08:19:28 +0000 (08:19 +0000)]
vdpa/mlx5: move DevX CQ creation to common

Using common function for DevX CQ creation.

Signed-off-by: Michael Baum <michaelba@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
3 years agoregex/mlx5: move DevX CQ creation to common
Michael Baum [Wed, 6 Jan 2021 08:19:27 +0000 (08:19 +0000)]
regex/mlx5: move DevX CQ creation to common

Using common function for DevX CQ creation.

Signed-off-by: Michael Baum <michaelba@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
3 years agocommon/mlx5: share DevX CQ creation
Michael Baum [Wed, 6 Jan 2021 08:19:26 +0000 (08:19 +0000)]
common/mlx5: share DevX CQ creation

The CQ object in DevX is created in several places and in several
different drivers.
In all places almost all the details are the same, and in particular the
allocations of the required resources.

Add a structure that contains all the resources, and provide creation
and release functions for it.

Signed-off-by: Michael Baum <michaelba@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
3 years agonet/mlx5: fix leak on ASO SQ creation failure
Michael Baum [Wed, 6 Jan 2021 08:19:25 +0000 (08:19 +0000)]
net/mlx5: fix leak on ASO SQ creation failure

In ASO SQ creation, the PMD allocates umem buffer for SQ.

When umem buffer allocation fails, the MR and CQ memory are not freed
what caused a memory leak.

Free it.

Fixes: f935ed4b645a ("net/mlx5: support flow hit action for aging")
Cc: stable@dpdk.org
Signed-off-by: Michael Baum <michaelba@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
3 years agonet/mlx5: remove CQE padding device argument
Michael Baum [Wed, 6 Jan 2021 08:19:24 +0000 (08:19 +0000)]
net/mlx5: remove CQE padding device argument

The data-path code doesn't take care on 'rxq_cqe_pad_en' and use padded
CQE for any case when the system cache-line size is 128B.

This makes the argument redundant.

Remove it.

Fixes: bc91e8db12cd ("net/mlx5: add 128B padding of Rx completion entry")
Cc: stable@dpdk.org
Signed-off-by: Michael Baum <michaelba@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
3 years agocommon/mlx5: fix completion queue entry size configuration
Michael Baum [Wed, 6 Jan 2021 08:19:23 +0000 (08:19 +0000)]
common/mlx5: fix completion queue entry size configuration

According to the current data-path implementation in the PMD the CQE
size must follow the cache-line size.
So, the configuration of the CQE size should be depended in
RTE_CACHE_LINE_SIZE.

Wrongly, part of the CQE creations didn't follow it exactly what caused
an incompatibility between HW and SW in the data-path when working in
128B cache-line size systems.

Adjust the rule for any CQE creation.
Remove the cqe_size attribute from the DevX CQ creation command and set
it inside the command translation according to the cache-line size.

Fixes: 79a7e409a2f6 ("common/mlx5: prepare support of packet pacing")
Fixes: 5cd0a83f413e ("common/mlx5: support more fields in DevX CQ create")
Cc: stable@dpdk.org
Signed-off-by: Michael Baum <michaelba@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
3 years agonet/mlx5: fix hairpin flow split decision
Dekel Peled [Sun, 10 Jan 2021 17:37:56 +0000 (19:37 +0200)]
net/mlx5: fix hairpin flow split decision

Previously, the identification of hairpin queue was done using
mlx5_rxq_get_type() function.
Recent patch replaced it with use of mlx5_rxq_get_hairpin_conf(),
and check of the return value conf != NULL.
The case of return value is NULL (queue is not hairpin) was not handled.
As result, non-hairpin flows were wrongly handled.
This patch adds the required check for return value is NULL.

Fixes: 509f8470de55 ("net/mlx5: do not split hairpin flow in explicit mode")
Cc: stable@dpdk.org
Signed-off-by: Dekel Peled <dekelp@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
3 years agonet/mlx5: split multi-thread flow handling per OS
Tal Shnaiderman [Thu, 7 Jan 2021 13:08:27 +0000 (15:08 +0200)]
net/mlx5: split multi-thread flow handling per OS

multi-threaded flows feature uses pthread function pthread_key_create
but for Windows the destruction option in the function is unimplemented.

To resolve it, Windows will implement destruction mechanism to cleanup
mlx5_flow_workspace object for each terminated thread.

Linux flow will keep the current behavior.

Signed-off-by: Tal Shnaiderman <talshn@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
Acked-by: Khoa To <khot@microsoft.com>
3 years agonet/octeontx2: support 24B custom L2 header parsing
Kiran Kumar K [Mon, 21 Dec 2020 07:45:18 +0000 (13:15 +0530)]
net/octeontx2: support 24B custom L2 header parsing

Adding support to parse 24B custom L2 header. Added devargs support to
configure the PKIND, and removed the restriction to support custom
headers on non SDP interface.

Signed-off-by: Kiran Kumar K <kirankumark@marvell.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>
3 years agonet/octeontx2: fix corruption in segments list
Sunil Kumar Kori [Mon, 21 Dec 2020 14:03:08 +0000 (19:33 +0530)]
net/octeontx2: fix corruption in segments list

On Tx, lastseg->next is not being reset to null for multi segmented
packet and same mbuf can be used on Rx which has a stale mbuf entry into
mbuf->next.

On Rx, application receives mbuf with mbuf->next uninitialized though
mbuf->nb_segs is correct. Application iterates over all segments using
mbuf->next ignoring mbuf->nb_segs which leads to undefined behavior.

So earlier assumption of just having right value in mbuf->nb_segs is
enough, is incorrect. Mbuf must contain valid and synced value in
nb_segs and next pointer.

Fixes: 364eb0e46683 ("net/octeontx2: avoid per packet barrier with multi segment")
Cc: stable@dpdk.org
Signed-off-by: Sunil Kumar Kori <skori@marvell.com>
Acked-by: Nithin Dabilpuram <ndabilpuram@marvell.com>
3 years agonet/mvneta: check allocation in Rx queue flush
Yunjian Wang [Mon, 7 Dec 2020 11:37:15 +0000 (19:37 +0800)]
net/mvneta: check allocation in Rx queue flush

The function rte_malloc() could return NULL, the return value
need to be checked.

Fixes: ce7ea764597e ("net/mvneta: support Rx/Tx")
Cc: stable@dpdk.org
Signed-off-by: Yunjian Wang <wangyunjian@huawei.com>
Acked-by: Liron Himi <lironh@marvell.com>
3 years agonet/bnxt: check chip reset in stop and close
Somnath Kotur [Thu, 24 Dec 2020 09:37:34 +0000 (15:07 +0530)]
net/bnxt: check chip reset in stop and close

While the error recovery thread is running, an application
can invoke dev_stop or dev_close_op thus triggering a race
and unwanted consequences if dev_close is invoked while the
recovery is not yet completed.
Fix by having another lock to synchronize between the 2 threads and
return EGAIN if adapter is in the middle of recovery when dev_stop
or dev_close ops are invoked

Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
3 years agonet/bnxt: fix error handling in device start
Somnath Kotur [Thu, 24 Dec 2020 09:37:33 +0000 (15:07 +0530)]
net/bnxt: fix error handling in device start

Call bnxt_dev_stop in error path of bnxt_dev_start_op() to keep
it simple and consistent

Fixes: c09f57b49c13 ("net/bnxt: add start/stop/link update operations")
Cc: stable@dpdk.org
Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
3 years agonet/bnxt: fix lock init and destroy
Somnath Kotur [Thu, 24 Dec 2020 09:37:32 +0000 (15:07 +0530)]
net/bnxt: fix lock init and destroy

Invoking init/uninit locks in init_resources and uninit_resources
would end up initializing and destroying locks on every port start
stop which is not desired.
Move the 2 routines to dev_init and dev_close respectively as
locks need to be initialized and destroyed only once during the
lifetime of the driver.

Fixes: 1cb3d39a48f7 ("net/bnxt: synchronize between flow related functions")
Cc: stable@dpdk.org
Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
3 years agonet/bnxt: add Rx logic for 58818 chips
Kalesh AP [Sun, 20 Dec 2020 05:24:30 +0000 (21:24 -0800)]
net/bnxt: add Rx logic for 58818 chips

1. On the new 58818 chips, the RX completion is largely the same except
   for the new completion opcode and the stripped VLAN format and
   checksum status. Added bnxt_parse_csum_v2(), bnxt_parse_pkt_type_v2()
   and bnxt_rx_vlan_v2() to support the new RX completion logic.
2. Disable vector mode RX/TX for 58818 chips for now.
3. The cfa_code format on 58818 chips is different than legacy chips.
   So skip cfa_code parsing logic on 58818 chips for now.

Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>