Joyce Kong [Wed, 17 Nov 2021 08:21:58 +0000 (08:21 +0000)]
app/testpmd: remove atomic operations for port status
The port_status changes do not need to be handled
atomically, as they are modified during initialization
or through the testpmd prompt instead of multiple
threads.
Signed-off-by: Joyce Kong <joyce.kong@arm.com> Reviewed-by: Ruifeng Wang <ruifeng.wang@arm.com> Reviewed-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com>
Michael Baum [Tue, 16 Nov 2021 14:36:35 +0000 (16:36 +0200)]
common/mlx5: fix redundant field in MR control structure
Inside the MR control structure there is a pointer to the common device.
This pointer enables access to the global cache as well as hardware
objects that may be required in case a new MR needs to be created.
The purpose of adding this pointer into the MR control structure was to
avoid its transfer as a parameter to all the functions of searching MR
in the caches.
However, adding it to this structure increased the Rx and Tx data-path
structures, all the fields that followed it were slightly moved away
which caused to a reduction in performance.
This patch removes the pointer from the structure. It can be accessed
through the "dev_gen_ptr" existing field using the "container_of"
operator.
Fixes: 334ed198ab4d ("common/mlx5: remove redundant parameter in MR search") Signed-off-by: Michael Baum <michaelba@nvidia.com> Acked-by: Matan Azrad <matan@nvidia.com>
Update public macros to have RTE_IP_FRAG_ prefix.
Update DPDK components to use new names.
Keep obsolete macro for compatibility reasons.
Renamed experimental function ``rte_frag_table_del_expired_entries``to
``rte_ip_frag_table_del_expired_entries`` to comply with other public
API naming convention.
Signed-off-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
This patch fixes various issues:
- replace _mm512_set_epi8 with _mm512_set_epi32 due to the lack
of support by some compilers (at least, gcc 8),
- check if AVX512F is supported along with GFNI, this is done if the code
is built on a platform that supports GFNI, but does not support AVX512,
- fix compilation problems on 32bit arch due to lack of support for
_mm_extract_epi64() by implementing XOR folding with
_mm_extract_epi32() on 32-bit arch,
Fixes: 4fd8c4cb0de1 ("hash: add new Toeplitz hash implementation") Signed-off-by: Vladimir Medvedkin <vladimir.medvedkin@intel.com> Acked-by: Lance Richardson <lance.richardson@broadcom.com> Acked-by: Kai Ji <kai.ji@intel.com>
Dmitry Kozlyuk [Fri, 12 Nov 2021 22:17:08 +0000 (01:17 +0300)]
doc: discourage using Meson 0.58 on Windows
Meson 0.58 and above cannot build DPDK on Windows with clang.
Recommend the latest known working version
and warn about the issue and the affected versions.
Suggested-by: Thomas Monjalon <thomas@monjalon.net> Signed-off-by: Dmitry Kozlyuk <dmitry.kozliuk@gmail.com>
Dmitry Kozlyuk [Fri, 12 Nov 2021 21:48:26 +0000 (00:48 +0300)]
config/x86: fix MinGW cross build with Meson 0.49
Cross build with MinGW was broken for the baseline meson 0.49.2.
Cause: in c_args = '-mno-avx512f' from config/x86/cross-mingw,
each character was treated as a separate compiler option:
meson.build:4:0: ERROR: Compiler x86_64-w64-mingw32-gcc can not compile programs.
With c_args = ['-mno-avx512f'] configuration passed, but build failed,
because Meson placed -mno-avx512f after -mavx512f in CFLAGS:
In file included from /usr/lib/gcc/x86_64-w64-mingw32/9.3-win32/include/immintrin.h:55,
from /usr/lib/gcc/x86_64-w64-mingw32/9.3-win32/include/x86intrin.h:32,
from ../dpdk/lib/net/net_crc_avx512.c:13:
/usr/lib/gcc/x86_64-w64-mingw32/9.3-win32/include/avx512fintrin.h:1650:1:
error: inlining failed in call to always_inline _mm512_ternarylogic_epi64:
target specific option mismatch
1650 | _mm512_ternarylogic_epi64 (__m512i __A, __m512i __B, __m512i __C,
| ^~~~~~~~~~~~~~~~~~~~~~~~~
../dpdk/lib/net/net_crc_avx512.c:59:9: note: called from here
59 | return _mm512_ternarylogic_epi64(tmp0, tmp1, data_block, 0x96);
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Additionally, -m[no-]avx512f flag is expected to be in machine_args
by all the checks in meson.build files.
Commit 419c6e9af69e ("net/i40e: fix build for Windows MinGW")
fixed the errors cause by MinGW using AVX512F on Windows.
The binutils AVX512F bug check is now portable,
so enable it for Windows to switch AVX512 support on and off
without any special logic for MinGW.
Fixes: 549bfc83168f ("config: disable AVX512 with MinGW") Signed-off-by: Dmitry Kozlyuk <dmitry.kozliuk@gmail.com> Acked-by: Bruce Richardson <bruce.richardson@intel.com>
Dmitry Kozlyuk [Fri, 12 Nov 2021 21:48:25 +0000 (00:48 +0300)]
buildtools: make AVX512 check portable
buildtools/binutils-avx512-check.sh was Unix-only
and could not be used in cross builds:
1) written in shell;
2) used the assembler binary that may be missing,
e.g. when building on Windows with LLVM;
3) located the assembler as ${AS:-as} and referenced objdump,
but those binaries may be overridden via --cross-file.
Rewrite the script in Python.
Use the C compiler for the check.
Locate objdump and the C compiler using Meson.
Signed-off-by: Dmitry Kozlyuk <dmitry.kozliuk@gmail.com> Acked-by: Bruce Richardson <bruce.richardson@intel.com>
Dmitry Kozlyuk [Fri, 12 Nov 2021 21:48:24 +0000 (00:48 +0300)]
config/x86: skip GNU binutils bug check for LLVM
AVX512 was disabled when GNU binutils were missing or had a known bug,
even if LLVM binutils were used for the build,
because binutils-avx512-check.sh was invoked regardless and failed.
In particular, this was the case for FreeBSD with clang (default).
Run the check only when GNU binutils are used.
Fixes: 68b1f1cda5b4 ("build: check AVX512 rather than binutils version") Cc: stable@dpdk.org Signed-off-by: Dmitry Kozlyuk <dmitry.kozliuk@gmail.com> Acked-by: Bruce Richardson <bruce.richardson@intel.com>
Elena Agostini [Tue, 16 Nov 2021 22:50:18 +0000 (22:50 +0000)]
gpu/cuda: introduce CUDA driver
This is the CUDA implementation of the gpudev library.
Functionalities implemented through CUDA Driver API are:
- Device probe and remove
- Manage device memory allocations
- Register/unregister external CPU memory in the device memory area
Signed-off-by: Elena Agostini <eagostini@nvidia.com>
David Marchand [Wed, 27 Oct 2021 14:04:57 +0000 (16:04 +0200)]
test: create a list of extra tests
We removed some tests from the "CI" (fast-tests) list because they
were not reliable enough or did not make sense as non regression tests.
Since we still build those tests code, leave an option for users to call
them.
This list can also serve as a point where to document why test X is not
suitable for the "CI" list.
Signed-off-by: David Marchand <david.marchand@redhat.com>
David Marchand [Wed, 27 Oct 2021 14:04:10 +0000 (16:04 +0200)]
test: remove unnecessary test commands
EAL and rwlock tests have been split into smaller unit tests now listed
in meson.
Nothing is referencing eal_flags_autotest and rwlock_autotest anymore,
since we dropped the python wrapper.
Gregory Etelson [Wed, 10 Nov 2021 16:57:59 +0000 (18:57 +0200)]
examples/multi_process: fix Rx packets distribution
MP servers distributes Rx packets between clients according to
round-robin scheme.
Current implementation always started packets distribution from
the first client. That procedure resulted in uniform distribution
in cases when Rx packets number was around clients number
multiplication. However, if RX burst repeatedly returned single
packet, round-robin scheme would not work because all packets
were assigned to the first client only.
The patch does not restart packets distribution from
the first client.
Packets distribution always continues to the next client.
These tests were using strings with hostname (dpdk.org) and this makes
test code do a DNS lookup. In container environment used for OpenSuse
build, DNS is unavailable. Replace dpdk.org with an IPv4 address
reserved for documentation (RFC5737) and use IPv6 in one example
(RFC3849). Actual addresses don't matter for this test which is
validating that code generated in classic BPF can be successfully
converted to eBPF.
Fixes: 2eccf6afbea9 ("bpf: add function to convert classic BPF to DPDK BPF") Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Luca Boccassi <bluca@debian.org>
Gagandeep Singh [Fri, 12 Nov 2021 07:53:48 +0000 (13:23 +0530)]
test/crypto: fix output buffer length
Input buffer length is getting appended with
padding length when the test case is encrypted
digest, but output buffer length is appended
with padding length for all the cases.
This patch fixes the output buffer length
by appending the padding length only when the
test case is of encrypted digest type.
Ciara Power [Thu, 11 Nov 2021 10:56:19 +0000 (10:56 +0000)]
test/crypto: remove unnecessary stats retrieval
The device stats are retrieved in the testcase teardown function,
but are not being used afterwards. Remove this unnecessary call.
The stats retrieval for the device is being tested already by a
dedicated stats testcase.
Fixes: 202d375c60bc ("app/test: add cryptodev unit and performance tests") Cc: stable@dpdk.org Signed-off-by: Ciara Power <ciara.power@intel.com> Acked-by: Anoob Joseph <anoobj@marvell.com>
Kai Ji [Tue, 9 Nov 2021 10:42:31 +0000 (10:42 +0000)]
test/crypto: fix data lengths
This patch fixes incorrect data lengths computation in cryptodev
unit test. Previously some data lengths were incorrectly set, which
was insensitive for crypto op unit tets but is critical for raw data
path API unit tests. The patch addressed the issue by setting the
correct data lengths for some tests.
Fixes: 681f540da52b ("cryptodev: do not use AAD in wireless algorithms") Fixes: e847fc512817 ("test/crypto: add encrypted digest case for AES-CTR-CMAC") Fixes: b1c1df46878d ("test/crypto: add ZUC test cases for auth-cipher") Cc: stable@dpdk.org Signed-off-by: Kai Ji <kai.ji@intel.com> Acked-by: Anoob Joseph <anoobj@marvell.com>
Anoob Joseph [Wed, 10 Nov 2021 13:04:08 +0000 (18:34 +0530)]
test/crypto: skip plain text compare for null cipher
NULL cipher is used for validating auth only cases. With NULL cipher,
validating plain text should not be done as the PMD is only expected
to update auth data.
Fixes: e847fc512817 ("test/crypto: add encrypted digest case for AES-CTR-CMAC") Cc: stable@dpdk.org Signed-off-by: Anoob Joseph <anoobj@marvell.com> Acked-by: Ciara Power <ciara.power@intel.com> Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Rashmi Shetty [Mon, 15 Nov 2021 16:34:44 +0000 (10:34 -0600)]
event/dlb2: fix delayed pop test in selftest
Number of events scheduled and available for dequeue
after token pop was set to dequeue_depth-1 instead of
dequeue_depth in test_delayed_pop. The expectation is
that all dequeue_depth number of events can be dequeued
once the last event is released.
Fixes: 07d55c418d47 ("event/dlb2: add delayed token pop logic") Cc: stable@dpdk.org Signed-off-by: Rashmi Shetty <rashmi.shetty@intel.com> Reviewed-by: Mike Ximing Chen <mike.ximing.chen@intel.com>
Mattias Rönnblom [Wed, 10 Nov 2021 11:32:10 +0000 (12:32 +0100)]
eventdev: negate maintenance capability flag
Replace RTE_EVENT_DEV_CAP_REQUIRES_MAINT, which signaled the need
for the application to call rte_event_maintain(), with
RTE_EVENT_DEV_CAP_MAINTENANCE_FREE, which does the opposite (i.e.,
signifies that the event device does not require maintenance).
This approach is more in line with how other eventdev hardware and/or
software limitations are handled in the Eventdev API.
Signed-off-by: Mattias Rönnblom <mattias.ronnblom@ericsson.com> Acked-by: Jerin Jacob <jerinj@marvell.com>
This patch fixes heap-use-after-free reported by ASan.
The application can use the queue_id as `-1` to delete all
the queues of the eth_device that are added to tx_adapter
instance.
In above case, the queue_del API is trying to use number of
queues from adapter level instead of eth_device queues.
When there are queues added from multiple eth devices,
it will result in heap-use-after-free as reported by ASAN.
This patch fixes the queue_del API to use correct number of
queues.
Bugzilla ID: 869 Fixes: a3bbf2e09756 ("eventdev: add eth Tx adapter implementation") 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> Tested-by: David Marchand <david.marchand@redhat.com>
Reported by clang 13.
This patch removes the hits variable from the cpu_load_collector function
within the performance thread example app as it is an unused but set
variable.
Conor Walsh [Mon, 15 Nov 2021 17:58:53 +0000 (17:58 +0000)]
event/sw: remove unused inflight events count
Reported by clang 13.
This patch removes the inflights variable from the sw_dump function
within the software section of the event driver as it is an unused but
set variable.
Bugzilla ID: 881 Fixes: c66baa68e453 ("event/sw: add dump function for easier debugging") Cc: stable@dpdk.org Reported-by: Liang Longfeng <longfengx.liang@intel.com> Signed-off-by: Conor Walsh <conor.walsh@intel.com> Acked-by: Harry van Haaren <harry.van.haaren@intel.com> Reviewed-by: David Marchand <david.marchand@redhat.com>
Conor Walsh [Mon, 15 Nov 2021 17:58:51 +0000 (17:58 +0000)]
bus/fslmc: remove unused device count
Reported by clang 13.
This patch removes the device_count variable from the rte_fslmc_scan
function within the fslmc bus driver as it is an unused but set variable.
Conor Walsh [Mon, 15 Nov 2021 17:58:50 +0000 (17:58 +0000)]
test/thash: fix build with clang 13
run_rss_calc() is used to compare the number of cycles spent computing a
hash value for different implementations.
clang 13 reports the hash variable as being unused, but run_rss_calc()
needs this variable as a placeholder for computing the hash value.
Bugzilla ID: 881 Fixes: 239fffe0402e ("test/thash: add performance tests for Toeplitz hash") Reported-by: Liang Longfeng <longfengx.liang@intel.com> Signed-off-by: Conor Walsh <conor.walsh@intel.com> Acked-by: Vladimir Medvedkin <vladimir.medvedkin@intel.com> Reviewed-by: David Marchand <david.marchand@redhat.com>
Conor Walsh [Mon, 15 Nov 2021 17:58:52 +0000 (17:58 +0000)]
common/qat: remove unused time counter
Reported by clang 13.
This patch removes the us variable from the qat_pf2vf_exch_msg function
within the PF to VF section of the QAT driver as it is an unused but set
variable.
Rongwei Liu [Thu, 11 Nov 2021 14:23:37 +0000 (16:23 +0200)]
app/flow-perf: support meter policy API
Add option "policy-mtr" to indicate if meter creation will include policy
or not. Meter creation will keep unchanged without it.
With "policy-mtr", the policy is introduced. API create_meter_policy
is to create a policy. API create_meter_rule will use it to create
a meter. The value of it is used to specify meter policy actions.
Signed-off-by: Haifei Luo <haifeil@nvidia.com> Signed-off-by: Jiawei Wang <jiaweiw@nvidia.com> Signed-off-by: Rongwei Liu <rongweil@nvidia.com> Acked-by: Wisam Jaddo <wisamm@nvidia.com>
Pavan Nikhilesh [Fri, 5 Nov 2021 08:38:51 +0000 (14:08 +0530)]
examples/l3fwd: use reserved addresses for EM mode
The l3fwd example should use the reserved IPv4/v6 reserved address
ranges defined in RFC5735, RFC5180 and RFC863 discard protocol for
the port number in the exact match mode of L3 forwarding.
Signed-off-by: Pavan Nikhilesh <pbhagavatula@marvell.com> Reviewed-by: David Marchand <david.marchand@redhat.com>
Pavan Nikhilesh [Fri, 5 Nov 2021 08:38:50 +0000 (14:08 +0530)]
examples/l3fwd: print device name when adding routes
Since the number of Ethernet ports have gone up, print the device name
(which for PCI devices is the BDF triplet), along with the routes.
This is also helpful for cases where allow listing order is not honored.
Pavan Nikhilesh [Fri, 5 Nov 2021 08:38:49 +0000 (14:08 +0530)]
examples/l3fwd: increase number of routes
Increase the number of routes from 8 to 16 that are statically added for
lpm and em mode as most of the SoCs support more than 8 interfaces.
The number of routes added is equal to the number of ethernet devices
ports enabled through port mask.
Ciara Loftus [Fri, 12 Nov 2021 10:30:02 +0000 (10:30 +0000)]
net/af_xdp: workaround custom program loading
Since v0.4.0, if the underlying kernel supports it, libbpf uses 'bpf
link' to manage the programs on the interfaces of the XDP sockets (xsks).
This is not compatible with the PMD's custom XDP program loading feature
which uses the netlink-based method for loading custom programs.
The conflict arises when libbpf searches for a custom program on the
interface using bpf link, but doesn't find one because the netlink
method was used. libbpf then proceeds to try to load the default program
on the interface, but fails due to the presence of the custom program.
To work around this, the PMD now uses the
XSK_LIBBPF_FLAGS__INHIBIT_PROG_LOAD flag which prevents libbpf from
attempting to search for or load a program. One repercussion is that
DPDK must now insert the xsk into the xsks_map as this was previously
handled by libbpf during the routines for program loading/probing.
Ideally, the PMD would use bpf link to load the custom program, however
at present there is no convenient and reliable way of detecting whether
the underlying kernel supports bpf link. Perhaps this may become
available in a future libbpf release, at which point we can switch the
PMD over to the new bpf link based method.
Ciara Loftus [Fri, 12 Nov 2021 10:30:01 +0000 (10:30 +0000)]
net/af_xdp: revert use BPF link for XDP programs
The commit ae70cc6e893b ("net/af_xdp: use BPF link for XDP programs")
caused compilation errors on kernels older than v5.8 due to absence of
the bpf_link_info struct and some definitions in the linux/bpf.h header.
Since relying on the reported kernel version is not a robust solution
and also since there doesn't appear to be a suitable definition in the
bpf header that the preprocessor could rely on to determine support for
bpf link, we will take a different approach to solving the issue that
the original patch attempted to solve. The next commit will address
this.
Fixes: ae70cc6e893b ("net/af_xdp: use BPF link for XDP programs") Signed-off-by: Ciara Loftus <ciara.loftus@intel.com>
Haiyue Wang [Mon, 15 Nov 2021 06:05:09 +0000 (14:05 +0800)]
app/testpmd: avoid process ID out of range
The 'proc_id' should be less than 'num_procs', if not, exit the testpmd
and show the error message.
Fixes: a550baf24af9 ("app/testpmd: support multi-process") Signed-off-by: Haiyue Wang <haiyue.wang@intel.com> Acked-by: Min Hu (Connor) <humin29@huawei.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
Apeksha Gupta [Mon, 15 Nov 2021 07:19:38 +0000 (12:49 +0530)]
net/enetfec: support queue configuration
This patch adds Rx/Tx queue configuration setup operations.
On packet reception the respective BD Ring status bit is set
which is then used for packet processing.
Apeksha Gupta [Mon, 15 Nov 2021 07:19:37 +0000 (12:49 +0530)]
net/enetfec: support UIO
Implemented the fec-uio driver in kernel. enetfec PMD uses
UIO interface to interact with "fec-uio" driver implemented in
kernel for PHY initialisation and for mapping the allocated memory
of register & BD from kernel to DPDK which gives access to
non-cacheable memory for BD.
Bing Zhao [Fri, 12 Nov 2021 14:38:42 +0000 (16:38 +0200)]
net/mlx5: fix delay drop bit set overflow
The attribute to record the global control of hairpin queues' delay
drop was defined as a bit-field with one bit, and the intention was
to reduce the memory overhead. In the meanwhile, the macro was
defined as an enumerated value 0x2.
No matter what value inputted via devarg, the lowest bit was always
zero and the higher bits would be ignored. For hairpin queues, the
delay drop attribute couldn't be enabled.
With the commit, the double logical negation is used to fix this.
Gregory Etelson [Thu, 11 Nov 2021 16:16:26 +0000 (18:16 +0200)]
net/mlx5: fix integrity conversion scheme
RTE flow integrity API provides top-level packet validations.
RTE integrity bits are not always translated one-to-one to
hardware integrity bits.
For example RTE l3_ok and l4_ok integrity bits require 2 hardware
integrity bits each.
The patch fixes RTE l3_ok and l4_ok bits translation to match
ConnectX-6 hardware.
There was a redundant check for the enabled E-Switch, this
resulted in device probing failure if the Tx scheduling was
requested and E-Switch was enabled.
Ruifeng Wang [Wed, 15 Sep 2021 08:33:39 +0000 (16:33 +0800)]
net/i40e: fix risk in descriptor read in scalar Rx
Rx descriptor is 16B/32B in size. If the DD bit is set, it indicates
that the rest of the descriptor words have valid values. Hence, the
word containing DD bit must be read first before reading the rest of
the descriptor words.
Since the entire descriptor is not read atomically, on relaxed memory
ordered systems like Aarch64, read of the word containing DD field
could be reordered after read of other words.
Read barrier is inserted between read of the word with DD field
and read of other words. The barrier ensures that the fetched data
is correct.
Testpmd single core test showed no performance drop on x86 or N1SDP.
On ThunderX2, 22% performance regression was observed.
Fixes: 7b0cf70135d1 ("net/i40e: support ARM platform") Cc: stable@dpdk.org Signed-off-by: Ruifeng Wang <ruifeng.wang@arm.com> Reviewed-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com>
The routine converting RTE flow modify field action into
field driver's presentation did not specify the field mask
correctly and this resulted into wrong conversion for
the actions with shifted fields.
Matan Azrad [Mon, 8 Nov 2021 12:22:04 +0000 (14:22 +0200)]
common/mlx5: fix build for zero-length headroom array
The structure of the striding RQ(MPRQ) buffer includes an array size
defined by the RTE_PKTMBUF_HEADROOM macro added in [1].
When RTE_PKTMBUF_HEADROOM is set to 0 in the compilation config file
the compilation with debug type failed:
"In file included from ../drivers/common/mlx5/mlx5_common.h:25,
from ../drivers/common/mlx5/linux/mlx5_nl.h:12,
from ../drivers/common/mlx5/linux/mlx5_nl.c:22:
../drivers/common/mlx5/mlx5_common_mr.h:96:10: error: ISO C forbids
zero-size array 'pad' [-Werror=pedantic]"
Actually, the array for the first stride headroom is not needed:
Each stride in the striding RQ buffer includes the headroom of the next
stride, so the headroom of the first stride should be allocated before
the starting point of the buffer posted to the HW(HW buffer).
The striding RQ buffer is used as an attached buffer to mbuf and have
shared information per stride.
The LRO support moved all the strides shared information to the top of
the buffer before the first stride headroom but didn't remove the old
memory of this headroom from the buffer.
Remove the old headroom memory from the striding RQ buffer.
Bing Zhao [Fri, 5 Nov 2021 06:10:57 +0000 (08:10 +0200)]
net/mlx5: fix RETA update without stopping device
The global redirection table is used to create the default flow
rules for the ingress traffic with the lowest priority. It is also
used to create the default RSS rule in the destination table when
there is a tunnel offload.
To update the RETA in-flight, there is no restriction in the ethdev
API. In the previous implementation of mlx5, a port restart was
needed to make the new configuration take effect.
The restart is heavy, e.g., all the queues will be released and
reallocated, users' rules will be flushed. Since the restart is
internal, there is a risk to crash the application when some change
in the ethdev is introduced but no workaround is done in mlx5 PMD.
The users' rules, including the default miss rule for tunnel
offload, should not be impacted by the RETA update. It is improper
to flush all rules when updating RETA.
With this patch, only the default rules will be flushed and
re-created with the new table configuration.
Jiawei Wang [Wed, 3 Nov 2021 13:07:59 +0000 (15:07 +0200)]
net/mlx5: fix tag ID conflict with sample action
For the flows containing sample action, the tag action was added
implicitly to store the unique flow index into metadata register in the
split prefix subflow, and then match on this index in the split suffix
subflow. The metadata register for flow index of sample split subflows
was also used to store application metadata TAG 0 item, this might cause
TAG 0 corruption in the flows with sample actions.
This patch uses the same metadata register C index as used for
ASO action since it's reserved and not used directly by the application,
and adds the checking in validation to make sure not to conflict
with ASO CT in the same flow.
Fixes: b4c0ddbfcc58 ("net/mlx5: split sample flow into two sub-flows") Cc: stable@dpdk.org Signed-off-by: Jiawei Wang <jiaweiw@nvidia.com> Acked-by: Viacheslav Ovsiienko <viacheslavo@nvidia.com>
Dmitry Kozlyuk [Tue, 9 Nov 2021 10:32:53 +0000 (12:32 +0200)]
common/mlx5: fix external memory pool registration
Registration of packet mempools with RTE_PKTMBUF_POOL_PINNED_EXT_MEM
was performed incorrectly: after population of such mempool chunks
only contain memory for rte_mbuf structures, while pointers to actual
external memory are not yet filled. MR LKeys could not be obtained
for external memory addresses of such mempools. Rx datapath assumes
all used mempools are registered and does not fallback to dynamic
MR creation in such case, so no packets could be received.
Skip registration of extmem pools on population because it is useless.
If used for Rx, they are registered at port start.
During registration, recognize such pools, inspect their mbufs
and recover the pages they reside in.
While MRs for these pages may already be created by rte_dev_dma_map(),
they are not reused to avoid synchronization on Rx datapath
in case these MRs are changed in the database.
Rongwei Liu [Tue, 2 Nov 2021 07:22:40 +0000 (09:22 +0200)]
net/mlx5: fix meter policy validation
When a user specifies meter policy like "g_actions queue / end
y_actions queue / r_action drop / end", validation logic missed
to set meter policy mode and it took a random value from the stack.
Define ALL policy modes for the mentioned cases.
Fixes: 4b7bf3ffb473 ("net/mlx5: support yellow in meter policy validation") Cc: stable@dpdk.org Signed-off-by: Rongwei Liu <rongweil@nvidia.com> Acked-by: Matan Azrad <matan@nvidia.com> Reviewed-by: Bing Zhao <bingz@nvidia.com>
git.droids-corp.org / dpdk.git / log