Satheesh Paul [Tue, 25 Jan 2022 04:17:59 +0000 (09:47 +0530)]
common/cnxk: fix base rule merge
Egress rules do not have a base steering rule and the
current base steering rule is applicable only for ingress
rules. Hence, when creating a flow rule, merge base steering
rule only for ingress rules.
Fixes: f9af90807466 ("common/cnxk: add mcam utility API") Cc: stable@dpdk.org Signed-off-by: Satheesh Paul <psatheesh@marvell.com> Reviewed-by: Kiran Kumar K <kirankumark@marvell.com>
Satheesh Paul [Tue, 25 Jan 2022 04:15:26 +0000 (09:45 +0530)]
common/cnxk: fix flow deletion
When a flow is deleted, the counters are deleted without
being cleared first. This results in counter values being
retained and shown in flows created later. Hence, counter
is cleared before being deleted.
Updated driver to clear and delete counters upon flow deletion.
Fixes: a07f7ced436d ("common/cnxk: add NPC init and fini") Cc: stable@dpdk.org Signed-off-by: Satheesh Paul <psatheesh@marvell.com> Reviewed-by: Jerin Jacob <jerinj@marvell.com>
Satheesh Paul [Tue, 25 Jan 2022 04:12:48 +0000 (09:42 +0530)]
common/cnxk: fix null access in IPv6 flow parsing
Added null check in IPv6 flow pattern parsing.
Fixes: 474e275b1bc6 ("common/cnxk: support extensions attributes in IPv6 item") Signed-off-by: Satheesh Paul <psatheesh@marvell.com> Reviewed-by: Jerin Jacob <jerinj@marvell.com>
Min Hu (Connor) [Thu, 17 Feb 2022 11:36:55 +0000 (19:36 +0800)]
app/testpmd: check starting port is not in bonding
In bond, start or stop slave port should be operated by bonding port.
This patch add port_is_bonding_slave in start_port function.
Fixes: 0e545d3047fe ("app/testpmd: check stopping port is not in bonding") Cc: stable@dpdk.org Signed-off-by: Min Hu (Connor) <humin29@huawei.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
Junjie Wan [Tue, 15 Feb 2022 10:59:40 +0000 (18:59 +0800)]
net/bonding: fix slaves initializing on MTU setting
If a initial process for the bonding device is like:
rte_eth_dev_configure
rte_eth_dev_set_mtu
queue setup and start, etc.
Pass the vdev args to application, and init bonding device only.
-a 0000:af:00.0 --vdev="net_bonding0,mode=2,slave=0000:af:00.0"
It will fail and complain for the slave device
"Port 0 must be configured before MTU set"
Test can be reproduced with ovs.
Fixes: b26bee10ee37 ("ethdev: forbid MTU set before device configure") Cc: stable@dpdk.org Signed-off-by: Junjie Wan <wanjunjie@bytedance.com> Tested-by: Ferruh Yigit <ferruh.yigit@intel.com> Acked-by: Min Hu (Connor) <humin29@huawei.com>
Rahul Bhansali [Wed, 19 Jan 2022 10:52:03 +0000 (16:22 +0530)]
net/cnxk: fix mbuf data length
If multi-segment is enabled and single segment/packet
is received, then mbuf data_len is not updated in
cn9k_nix_cqe_to_mbuf() function.
Also, in case of timestamp is enabled, mbuf data_len
and pkt_len will be updated for all packets including
multi segmented packets.
Fixes: 7c6bee34064f ("net/cnxk: enable PTP processing in vector Rx") Cc: stable@dpdk.org Signed-off-by: Rahul Bhansali <rbhansali@marvell.com> Acked-by: Jerin Jacob <jerinj@marvell.com>
David Marchand [Wed, 16 Feb 2022 14:13:16 +0000 (15:13 +0100)]
ethdev: fix MAC address in telemetry device info
The right size for a human readable MAC is RTE_ETHER_ADDR_FMT_SIZE.
While at it, the net library provides a helper for MAC address
formatting. Prefer it.
John Miller [Tue, 15 Feb 2022 22:19:22 +0000 (16:19 -0600)]
net/ark: support arbitrary mbuf size
Added arbitrary mbuf size per queue capability.
Updated ARK_UDM_CONST3 value to reflect the version number
read from the HW that is required to support this change.
Signed-off-by: John Miller <john.miller@atomicrules.com>
John Miller [Tue, 15 Feb 2022 22:19:20 +0000 (16:19 -0600)]
net/ark: add device capabilities record
Add a device capabilities record for supported features.
Certain variants require that PCIe read-requests be correctly
throttled. This is called "rqpacing" in Arkville, and has to do
with credit and flow control on certain Arkville implementations.
Signed-off-by: John Miller <john.miller@atomicrules.com>
Martin Spinler [Tue, 15 Feb 2022 12:55:42 +0000 (13:55 +0100)]
net/nfb: fix multicast/promiscuous mode switching
In the firmware, the promisc mode overrides the multicast mode.
So when the promisc mode is turned off, driver must check if the
multicast mode was active before and conditionally reactivate it.
Fixes: 6435f9a0ac22 ("net/nfb: add new netcope driver") Cc: stable@dpdk.org Signed-off-by: Martin Spinler <spinler@cesnet.cz>
Joyce Kong [Fri, 17 Dec 2021 05:36:00 +0000 (05:36 +0000)]
net/i40e: add flow mark capability to NEON Rx
This commit adds a flow director support to i40e NEON
vector RX path.
I40e can have 16 and 32 byte descriptors, and the Flow
Director ID data and indication-bit are in different
locations for each size descriptor. The support is
implemented in two separate functions as they require
vastly different operations.
The 16B descriptor re-purposes the "filter-status" u32
field to indicate FDIR ID when the FLM bit is set. No
extra loads are required, however we do have to store
to mbuf->fdir.hi, which is not stored to in the RX path
before this patch.
The 32B descriptor requires loading the 2nd 16 bytes of
each descriptor, to get the FLEXBH_STAT and FD Filter ID
from qword3. The resulting data must also be stored to
mbuf->fdir.hi, same as the 16B code path.
Signed-off-by: Joyce Kong <joyce.kong@arm.com> Reviewed-by: Ruifeng Wang <ruifeng.wang@arm.com>
Pablo de Lara [Wed, 23 Feb 2022 18:22:45 +0000 (18:22 +0000)]
doc: support IPsec Multi-buffer lib v1.2
Updated AESNI MB and AESNI GCM, KASUMI, ZUC and SNOW3G PMD documentation
guides with information about the latest Intel IPSec Multi-buffer
library supported.
Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Pablo de Lara [Wed, 23 Feb 2022 16:01:16 +0000 (16:01 +0000)]
crypto/ipsec_mb: fix length and offset settings
KASUMI, SNOW3G and ZUC require lengths and offsets to
be set in bits or bytes depending on the algorithm.
There were some algorithms that were mixing these two,
so this commit is fixing this issue.
Fixes: ae8e085c608d ("crypto/aesni_mb: support KASUMI F8/F9") Fixes: 6c42e0cf4d12 ("crypto/aesni_mb: support SNOW3G-UEA2/UIA2") Fixes: fd8df85487c4 ("crypto/aesni_mb: support ZUC-EEA3/EIA3") Fixes: 8c835018de84 ("crypto/ipsec_mb: support ZUC-256 for aesni_mb") Cc: stable@dpdk.org Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Pablo de Lara [Wed, 23 Feb 2022 16:01:15 +0000 (16:01 +0000)]
crypto/ipsec_mb: fix ZUC operation overwrite
ZUC PMD batches crypto operations depending on their type
(encryption + tag generation, tag verification + decryption, etc),
to allow parallelization.
The array used to store the pointers to these operations was
always the same array provided by dequeue_burst() function,
and it was looping around the same positions (from 0 to ZUC_MAX_BURST - 1).
A new internal array is used to avoid overwriting the pointers
of the array provided by dequeue_burst() function.
Fixes: cf7685d68f00 ("crypto/zuc: add driver for ZUC library") Cc: stable@dpdk.org Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Pablo de Lara [Wed, 23 Feb 2022 16:01:14 +0000 (16:01 +0000)]
crypto/ipsec_mb: fix ZUC authentication verify
ZUC authentication is done over multiple buffers at a time.
When authentication verification is done, multiple scratch buffers
are using to generate the tags that will be compared afterwards.
However, the same scratch buffer was used always, instead of having
different ones for each crypto operation.
Fixes: 0b133c36ad7d ("crypto/zuc: support IPsec Multi-buffer lib v0.54") Cc: stable@dpdk.org Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Pablo de Lara [Wed, 23 Feb 2022 16:01:13 +0000 (16:01 +0000)]
crypto/ipsec_mb: check missing operation types
When processing crypto operations in ZUC PMD,
there were two operation types that were set at session level,
but not checked when the operations are enqueued and processed,
leaving the buffers untouched silently.
Fixes: cde8df1bda9d ("crypto/ipsec_mb: move zuc PMD") Cc: stable@dpdk.org Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Brian Dooley [Mon, 21 Feb 2022 18:06:58 +0000 (18:06 +0000)]
examples/l2fwd-crypto: fix port mask overflow
Coverity flags an issue with 32-bit value. If max ethports value is
configured with a value larger than 32 there will be an issue.
Coverity issue: 375863 Unintentional integer overflow
Add support to enable per port packet pool and also override
vector pool size from command line args. This is useful
on some HW to tune performance based on usecase.
Update error prints in data path to RTE_LOG_DP().
Error prints in fast path are not good for performance
as they slow down the application when few bad packets are
received.
Arek Kusztal [Mon, 21 Feb 2022 10:48:31 +0000 (10:48 +0000)]
crypto/qat: refactor asymmetric session
This patch refactors asymmetric session in Intel
QuickAssist Technology PMD and fixes some issues
with xform. Code will be now bit more scalable,
and easier readable.
Arek Kusztal [Mon, 21 Feb 2022 10:48:27 +0000 (10:48 +0000)]
crypto/qat: refactor asymmetric crypto functions
This commit refactors asummetric crypto functions
in Intel QuickAssist Technology PMD.
Functions right now are shorter and far easier readable,
plus it facilitates addition of new algorithms.
Kai Ji [Wed, 23 Feb 2022 00:50:05 +0000 (08:50 +0800)]
crypto/qat: unify raw data path functions
This patch unifies QAT's raw dp api implementations
to the same enqueue/dequeue methods used in crypto operations.
The specific functions for different QAT generation are updated
respectively. The qat_sym_hw_dp.c is removed as no longer required.
Signed-off-by: Kai Ji <kai.ji@intel.com> Acked-by: Fan Zhang <roy.fan.zhang@intel.com>
Kai Ji [Wed, 23 Feb 2022 00:50:04 +0000 (08:50 +0800)]
crypto/qat: rework burst data path
This patch enable the op_build_request function in
qat_enqueue_op_burst, and the qat_dequeue_process_response
function in qat_dequeue_op_burst.
The op_build_request invoked in crypto build request op is based
on crypto operations setup'd during session init.
Signed-off-by: Kai Ji <kai.ji@intel.com> Acked-by: Fan Zhang <roy.fan.zhang@intel.com>
Kai Ji [Wed, 23 Feb 2022 00:50:03 +0000 (08:50 +0800)]
crypto/qat: unify asymmetric functions
This patch removes qat_asym_pmd.c and integrates all the
functions into qat_asym.c. The unified/integrated asym crypto
pmd functions should make them easier to maintain.
Signed-off-by: Kai Ji <kai.ji@intel.com> Acked-by: Fan Zhang <roy.fan.zhang@intel.com>
Kai Ji [Wed, 23 Feb 2022 00:50:02 +0000 (08:50 +0800)]
crypto/qat: unify symmetric functions
This patch removes qat_sym_pmd.c and integrates all the functions into
qat_sym.c. The unified/integrated qat sym crypto pmd functions should
make them easier to maintain.
Signed-off-by: Kai Ji <kai.ji@intel.com> Acked-by: Fan Zhang <roy.fan.zhang@intel.com>
Kai Ji [Wed, 23 Feb 2022 00:50:01 +0000 (08:50 +0800)]
crypto/qat: rework asymmetric op build operation
This patch reworks the asymmetric crypto data path
implementation in QAT driver. The changes include asymmetric
crypto data path separation for QAT hardware generations, and
code optimisation of the device capabilities declaration.
Signed-off-by: Kai Ji <kai.ji@intel.com> Acked-by: Fan Zhang <roy.fan.zhang@intel.com>
Kai Ji [Wed, 23 Feb 2022 00:50:00 +0000 (08:50 +0800)]
crypto/qat: rework session functions
This patch introduces a set of set_session methods to QAT
generations. In addition, the reuse of QAT session between
generations is prohibit as the support of min_qat_dev_gen_id'
is removed.
Signed-off-by: Kai Ji <kai.ji@intel.com> Acked-by: Fan Zhang <roy.fan.zhang@intel.com>
Kai Ji [Wed, 23 Feb 2022 00:49:59 +0000 (08:49 +0800)]
crypto/qat: support symmetric build op request
This patch adds common inline functions for QAT symmetric
crypto driver to process crypto op, and the implementation of
build op request function for QAT generation 1.
Signed-off-by: Kai Ji <kai.ji@intel.com> Acked-by: Fan Zhang <roy.fan.zhang@intel.com>
Kai Ji [Wed, 23 Feb 2022 00:49:58 +0000 (08:49 +0800)]
common/qat: define build request and dequeue operations
This patch introduce build request and dequeue op function
pointers to the qat queue pair implementation. The function
pointers are assigned during qat session generation based on input
crypto operation request.
Signed-off-by: Kai Ji <kai.ji@intel.com> Acked-by: Fan Zhang <roy.fan.zhang@intel.com>
When creating the asymmetric session mempool, the maximum private
session size of all devices is used when creating the mempool
object size.
The return value for ``rte_cryptodev_asym_get_private_session_size``
is unsigned int, whereas the variable was uint8_t, leading to a
possible overflow issue.
To fix this, the variable for private session size is now changed to
unsigned int to match the function return type.
Fixes: 1f1e4b7cbaad ("cryptodev: use single mempool for asymmetric session") Reported-by: Arek Kusztal <arkadiuszx.kusztal@intel.com> Signed-off-by: Ciara Power <ciara.power@intel.com> Acked-by: Fan Zhang <roy.fan.zhang@intel.com> Acked-by: Arek Kusztal <arkadiuszx.kusztal@intel.com>
Ciara Power [Tue, 15 Feb 2022 11:07:58 +0000 (11:07 +0000)]
crypto/ipsec_mb: remove useless check
The qp is never NULL when it reaches the qp_setup_cleanup error handling
code. This is due to checks earlier in the function that return an error
if qp is NULL.
Shijith Thotton [Thu, 24 Feb 2022 04:46:07 +0000 (10:16 +0530)]
app/eventdev: add crypto producer mode
In crypto producer mode, producer core enqueues cryptodev with software
generated crypto ops and worker core dequeues crypto completion events
from the eventdev. Event crypto metadata used for above processing is
pre-populated in each crypto session.
Parameter --prod_type_cryptodev can be used to enable crypto producer
mode. Parameter --crypto_adptr_mode can be set to select the crypto
adapter mode, 0 for OP_NEW and 1 for OP_FORWARD.
This mode can be used to measure the performance of crypto adapter.
eca_cryptodev_cdev_flush() is internal function and called with
valid range of cdevs.
crypto_cdev_info structure is allocated at adapter creation time
and retrieved from the adapter for a valid cdevs which cannot be NULL
and hence no need for NULL check.
Fixes: 2ae84b39ae7b ("eventdev/crypto: store operations in circular buffer") Signed-off-by: Ganapati Kundapura <ganapati.kundapura@intel.com> Acked-by: Abhinandan Gujjar <abhinandan.gujjar@intel.com>
Timothy McDaniel [Wed, 16 Feb 2022 19:38:27 +0000 (13:38 -0600)]
event/dlb2: poll HW CQ inflights before mapping queue
When attempting to link a port and queue immediately after unlinking,
the CQ inflights may not all be processed. Poll the h/w register for
outstanding inflights instead of reading once, in case the inflights
are still being processed. Also return EBUSY if the inflight
processing is not completed in a suitable amount of time.
Fixes: 1857f1922ce2 ("event/dlb2: use new implementation of resource file") Cc: stable@dpdk.org Signed-off-by: Timothy McDaniel <timothy.mcdaniel@intel.com>
Remove the memcpy usage in queue config get function for
`event` variable which is 8 byte size and use direct copy.
Also provide vector information and event buffer size in the
queue config info.
Fixes: da781e6488 ("eventdev/eth_rx: support Rx queue config get") Cc: stable@dpdk.org Signed-off-by: Naga Harish K S V <s.v.naga.harish.k@intel.com> Acked-by: Jay Jayatheerthan <jay.jayatheerthan@intel.com>
Megha Ajmera [Wed, 23 Feb 2022 17:36:30 +0000 (17:36 +0000)]
examples/qos_sched: fix core mask overflow
Masking of core mask was incorrect. Instead of using 1U for shifting, it
should be using 1LU as the result is assigned to uint64.
CID 375859: Potentially overflowing expression "1U << app_main_core" with
type "unsigned int" (32 bits, unsigned) is evaluated using 32-bit
arithmetic, and then used in a context that expects an expression of
type "uint64_t" (64 bits, unsigned).
Bruce Richardson [Wed, 16 Feb 2022 16:06:09 +0000 (16:06 +0000)]
doc: improve configuration examples in idxd guide
The documentation on how to configure device instances using
accel-config can be improved by a number of changes:
* For initial example, when only configuring one queue, omit
configuration of a second engine, which is unused later.
* Add the "max-batch-size" setting to the options being configured for
each queue
* Add a final, more complete example, showing configuration of multiple
queues on a device.
Cc: stable@dpdk.org Signed-off-by: Bruce Richardson <bruce.richardson@intel.com> Reviewed-by: Kevin Laatz <kevin.laatz@intel.com>
Bruce Richardson [Wed, 16 Feb 2022 16:06:08 +0000 (16:06 +0000)]
dma/idxd: configure maximum batch size to high value
When configuring an Intel DSA instance using the utility script
dpdk_idxd_cfg.py, explicitly set the max supported batch size value to a
high value, to ensure large bursts are supported if so desired. The
default in the linux kernel is now just 32 [1], which may not be
sufficient for all DPDK apps.
Bruce Richardson [Wed, 16 Feb 2022 16:06:07 +0000 (16:06 +0000)]
test/dma: fix missing checks for device capacity
For some DMA HW devices, e.g. those using the idxd driver, the maximum
burst size is configurable, which can lead to test failures if the value
is set too small. Add explicit check for this to give reasonable error
messages for devices which need their config adjusted.
Fixes: 1b86a66a30c2 ("test/dma: add more comprehensive copy tests") Fixes: 8fa5d2683940 ("test/dma: add burst capacity test") Cc: stable@dpdk.org Signed-off-by: Bruce Richardson <bruce.richardson@intel.com> Reviewed-by: Conor Walsh <conor.walsh@intel.com> Reviewed-by: Kevin Laatz <kevin.laatz@intel.com>
Chengwen Feng [Thu, 17 Feb 2022 02:59:11 +0000 (10:59 +0800)]
dma/hisilicon: use common PCI device naming
For DMA device 0000:7d:0.0, the original generated dmadev name starts
with the "7d:0.0", which is not expected.
This patch uses rte_pci_device_name API to generates the dmadev name.
Chengwen Feng [Thu, 17 Feb 2022 02:59:07 +0000 (10:59 +0800)]
dma/hisilicon: support Kunpeng 930
The Kunpeng930 DMA devices have the same PCI device id with Kunpeng920,
but with different PCI revision and register layout. This patch
introduces the basic initialization for Kunpeng930 DMA devices.
Bruce Richardson [Mon, 14 Feb 2022 15:30:38 +0000 (15:30 +0000)]
build: make ring mempool driver mandatory
The default mempool driver is one based on the rte_ring, and as such it
needs to be present to have just about any app (which doesn't override
the mempool) run. Given this state of affairs it is probably best to add
this default mempool driver to the always-enabled list to ensure we get
a runnable build in all cases. This means that, for example, to run some
NIC tests with testpmd in a minimal build, in most cases, all the user
should need to do is specify "-Denable_drivers=net/<nic_drv>" for the
build.
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com> Acked-by: Morten Brørup <mb@smartsharesystems.com>
Bruce Richardson [Tue, 15 Feb 2022 17:30:29 +0000 (17:30 +0000)]
buildtools/chkincs: check missing C++ guards
Simply compiling a C header with a C++ compiler is not enough to flag
missing 'extern "C"' guards. To catch missing guards, we can just use a
simple grep for the 'extern "C"' part, and error out if any files have a
miss.
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com> Acked-by: Tyler Retzlaff <roretzla@linux.microsoft.com>
Bruce Richardson [Tue, 15 Feb 2022 17:30:28 +0000 (17:30 +0000)]
buildtools/chkincs: check SDK headers for C++ compatibility
With a one-line change to the lib meson.build file we can add the SDK
headers to the list of files to be checked using the chkincs binary.
Unfortunately, many of those SDK header depend upon headers in the PCI
and vdev bus drivers, so we need to update chkincs build to ensure those
dependencies are added. We also need to allow internal APIs to be
present in these SDK headers.
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com> Acked-by: Tyler Retzlaff <roretzla@linux.microsoft.com>
Brian Dooley [Wed, 16 Feb 2022 15:14:56 +0000 (15:14 +0000)]
cryptodev: add missing C++ guards
Some public header files were missing 'extern "C"' C++ guards,
and couldn't be used by C++ applications. Add the missing guards.
Fixes: 7a3357205755 ("lib: remove C++ include guard from private headers") Cc: stable@dpdk.org Signed-off-by: Brian Dooley <brian.dooley@intel.com> Acked-by: Bruce Richardson <bruce.richardson@intel.com> Acked-by: Tyler Retzlaff <roretzla@linux.microsoft.com>
Brian Dooley [Wed, 16 Feb 2022 15:14:55 +0000 (15:14 +0000)]
bpf: add missing C++ guards
Some public header files were missing 'extern "C"' C++ guards,
and couldn't be used by C++ applications. Add the missing guards.
Fixes: 7a3357205755 ("lib: remove C++ include guard from private headers") Cc: stable@dpdk.org Signed-off-by: Brian Dooley <brian.dooley@intel.com> Acked-by: Bruce Richardson <bruce.richardson@intel.com> Acked-by: Tyler Retzlaff <roretzla@linux.microsoft.com>
git.droids-corp.org / dpdk.git / log