Add compilation tests for the use_cc_memcpy build option.
Signed-off-by: Mattias Rönnblom
---
.ci/linux-build.sh| 5 +
.github/workflows/build.yml | 7 +++
devtools/test-meson-builds.sh | 4 +++-
3 files changed, 15 insertions(+), 1 deletion(-)
diff --git a/.ci/linux-build
The octeon_ip driver relied on , but failed to provide a
direct include of this file.
Signed-off-by: Mattias Rönnblom
Acked-by: Stephen Hemminger
---
drivers/net/octeon_ep/otx_ep_ethdev.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/drivers/net/octeon_ep/otx_ep_ethdev.c
b/drivers/net/
This patch set make DPDK library, driver, and application code
optionally use the compiler/libc memcpy() when functions in
are invoked.
The new compiler memcpy-based rte_memcpy() implementations may be
enabled by means of a build-time option.
This patch set only make a difference on x86, PPC and
Provide build option to have functions in delegate to
the standard compiler/libc memcpy(), instead of using the various
custom DPDK, handcrafted, per-architecture rte_memcpy()
implementations.
A new meson build option 'use_cc_memcpy' is added. By default, the
traditional, custom DPDK rte_memcpy()
In build where use_cc_memcpy is set to true, the vhost user PMD
suffers a large performance drop on Intel P-cores for small packets,
at least when built by GCC and (to a much lesser extent) clang.
This patch addresses that issue by using a custom virtio
memcpy()-based packet copying routine.
Perf
The DLB2 PMD depended on being included as a side-effect
of being included.
In addition, DLB2 used rte_memcpy() but did not include ,
but rather depended on other include files to do so.
This patch addresses both of those issues.
Signed-off-by: Mattias Rönnblom
Acked-by: Bruce Richardson
---
The trie implementation of the fib library relied on , but
failed to provide a direct include of this file.
Signed-off-by: Mattias Rönnblom
Acked-by: Bruce Richardson
Acked-by: Stephen Hemminger
---
lib/fib/trie.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/lib/fib/trie.c b/lib/fib/tri
The distributor library relied on , but failed to provide
a direct include of this file.
Signed-off-by: Mattias Rönnblom
Acked-by: Bruce Richardson
---
lib/distributor/rte_distributor.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/lib/distributor/rte_distributor.c
b/lib/distributor/rte_
Have rte_bit_[test|set|clear|assign|flip]() and rte_bit_atomic_*()
handle volatile-marked pointers.
Signed-off-by: Mattias Rönnblom
Acked-by: Morten Brørup
Acked-by: Jack Bond-Preston
--
PATCH v3:
* Updated to reflect removed 'fun' parameter in __RTE_GEN_BIT_*()
(Jack Bond-Preston).
PATC
Add functionality to test and modify the value of individual bits in
32-bit or 64-bit words.
These functions have no implications on memory ordering, atomicity and
does not use volatile and thus does not prevent any compiler
optimizations.
Signed-off-by: Mattias Rönnblom
Acked-by: Morten Brørup
Extend bitops tests to cover the rte_bit_atomic_*() family of
functions.
Signed-off-by: Mattias Rönnblom
Acked-by: Morten Brørup
Acked-by: Tyler Retzlaff
Acked-by: Jack Bond-Preston
--
RFC v4:
* Add atomicity test for atomic bit flip.
RFC v3:
* Rename variable 'main' to make ICC happy.
--
Relax chkincs requirement of all DPDK header files having to contain
'extern "C"'.
Instructing a C++ toolchain to use C linkage is only necessarily if the
header file declares symbols (i.e., functions or global variables).
With this change, chkincs tries to find if any functions or references
to
Add atomic bit test/set/clear/assign/flip and
test-and-set/clear/assign/flip functions.
All atomic bit functions allow (and indeed, require) the caller to
specify a memory order.
Signed-off-by: Mattias Rönnblom
Acked-by: Morten Brørup
Acked-by: Tyler Retzlaff
Acked-by: Jack Bond-Preston
--
On Mon, Sep 9, 2024 at 1:04 PM Bruce Richardson
wrote:
>
> Commit 0f9ec0cbd2a9 ("net/iavf: fix VF reset when using DCF"),
> introduced a VF-reset adminq call into the reset sequence for iavf.
> However, that call was very early in the sequence before other adminq
> commands had been sent.
>
> To d
This patch set represent an attempt to improve and extend the RTE
bitops API, in particular for functions that operate on individual
bits.
All new functionality is exposed to the user as generic selection
macros, delegating the actual work to private (__-marked) static
inline functions. Public fun
Extend bitops tests to cover the
rte_bit_[test|set|clear|assign|flip]()
functions.
The tests are converted to use the test suite runner framework.
Signed-off-by: Mattias Rönnblom
Acked-by: Morten Brørup
Acked-by: Tyler Retzlaff
Acked-by: Jack Bond-Preston
--
RFC v6:
* Test rte_bit_*test()
Currently building with custom rdma-core installed in /opt/rdma-core
after setting PKG_CONFIG_PATH=/opt/rdma-core/lib64/pkgconfig/ results
in the below meson logs:
Run-time dependency libmana found: YES 1.0.54.0
Header "infiniband/manadv.h" has symbol "manadv_set_context_attr" : NO
Thus to fix thi
Hi Bruce, thanks for your reply.
> -Original Message-
> From: Richardson, Bruce
> Sent: Thursday, September 19, 2024 3:50 PM
> To: Zeng, ZhichaoX
> Cc: dev@dpdk.org
> Subject: Re: [PATCH v5] net/ice: support customized search path for DDP
> package
>
> On Thu, Sep 19, 2024 at 11:29:34AM
The "compiler.links()" function meson documentation [1] is a little
unclear, in a casual reading implies that the function was new in 0.60
meson release. In fact, it is only enhanced as described in that
release, but is present earlier.
As such, we can remove the version checks preceeding the calls
In order to work around some deprecated functions in meson, we need to
increase meson version. Increasing to 0.57 to also gain support for
other useful features for us in DPDK. Changes of interest to DPDK
include:
* Use get_external_property instead of get_cross_property
* Ability to use a VERSION
This patchset proposed increasing the minimum meson version to 0.57
and makes changes to update our build files appropriately for that
change: replacing deprecated functions, removing unnecessary version
checks and taking advantage of some new capabilities.
Why 0.57? No one particular reason; it's
Since minimum meson version is now 0.57 we can remove all version checks
for versions lower than that.
Signed-off-by: Bruce Richardson
---
config/meson.build | 2 +-
doc/api/meson.build | 2 +-
drivers/meson.build | 3 ---
lib/meson.build | 6 --
4 files changed, 2 insertions(+), 11 del
Rather than having to use run_command to shell out and read the VERSION
file for the DPDK version, we can use the support added directly to
meson in version 0.57.
Signed-off-by: Bruce Richardson
---
meson.build | 5 +
1 file changed, 1 insertion(+), 4 deletions(-)
diff --git a/meson.build b
The meson function "get_cross_property" is deprecated in meson 0.58 and
should be replaced by "get_external_property".
Signed-off-by: Bruce Richardson
---
config/arm/meson.build | 4 ++--
config/meson.build | 6 +++---
config/riscv/meson.build | 4 ++--
drivers/event/cn
Support EDDSA crypto algorithm in CNXK PMD.
Signed-off-by: Gowrishankar Muthukrishnan
---
doc/guides/cryptodevs/features/cn10k.ini | 1 +
drivers/common/cnxk/hw/cpt.h | 3 +-
drivers/common/cnxk/roc_ae.c | 52 +-
drivers/common/cnxk/roc_ae.h
Support EDDSA crypto algorithm in OpenSSL PMD.
Signed-off-by: Gowrishankar Muthukrishnan
---
drivers/crypto/openssl/openssl_pmd_private.h | 13 ++
drivers/crypto/openssl/rte_openssl_pmd.c | 223 +++
drivers/crypto/openssl/rte_openssl_pmd_ops.c | 131 +++
3 files chan
Add support for asymmetric EDDSA in cryptodev, as referenced in RFC:
https://datatracker.ietf.org/doc/html/rfc8032
Signed-off-by: Gowrishankar Muthukrishnan
---
doc/guides/cryptodevs/features/default.ini | 1 +
doc/guides/prog_guide/cryptodev_lib.rst| 2 +-
lib/cryptodev/rte_crypto_asym.h
Add test cases to validate EDDSA sign and verify ops,
as per RFC 8032.
Signed-off-by: Gowrishankar Muthukrishnan
---
v3:
- minor update in verify test.
---
app/test/test_cryptodev_asym.c | 345 +++-
app/test/test_cryptodev_ecdh_test_vectors.h | 94 +++-
app/test/test_cry
Add EDDSA support in fips_validation app.
Signed-off-by: Gowrishankar Muthukrishnan
---
examples/fips_validation/fips_validation.c| 2 +
examples/fips_validation/fips_validation.h| 23 ++
.../fips_validation/fips_validation_eddsa.c | 307 +
examples/fips_validation/m
Added support for EDDSA 25519 curve SIGN and VERIFY operations.
Signed-off-by: Gowrishankar Muthukrishnan
---
app/test-crypto-perf/cperf_ops.c | 52
app/test-crypto-perf/cperf_options.h | 2 +
app/test-crypto-perf/cperf_options_parsing.c | 9 +++-
app/t
> From: Bruce Richardson [mailto:bruce.richard...@intel.com]
> Sent: Friday, 20 September 2024 11.08
>
> On Fri, Sep 20, 2024 at 10:22:59AM +0200, Morten Brørup wrote:
> >Hi Bruce,
> >
> >
> >Meson version 1.3.2 emits a notice that meson.get_cross_property is
> >deprecated:
> >
> >
> >
This patch adds support for customizing firmware search path for
DDP package like the kernel behavior, it will read the search path
from "/sys/module/firmware_class/parameters/path", and try to load
DDP package.
Also, updates documentation for loading the DDP package in ice.rst.
Signed-off-by: Zh
Hi Bruce,
Meson version 1.3.2 emits a notice that meson.get_cross_property is deprecated:
NOTICE: Future-deprecated features used:
* 0.58.0: {'meson.get_cross_property'}
It seems to have been replaced by meson.get_external_property:
https://mesonbuild.com/Reference-manual_builtin_mes
On Fri, Sep 20, 2024 at 09:34:55AM +0100, Zeng, ZhichaoX wrote:
> Hi Bruce, thanks for your reply.
>
> > -Original Message-
> > From: Richardson, Bruce
> > Sent: Thursday, September 19, 2024 3:50 PM
> > To: Zeng, ZhichaoX
> > Cc: dev@dpdk.org
> > Subject: Re: [PATCH v5] net/ice: support
Some applications want to omit the trace feature.
Either to reduce the memory footprint, to reduce the exposed attack
surface, or for other reasons.
This patch adds an option in rte_config.h to include or omit trace in the
build. Trace is included by default.
Omitting trace works by omitting all
On Fri, Sep 20, 2024 at 10:22:59AM +0200, Morten Brørup wrote:
>Hi Bruce,
>
>
>Meson version 1.3.2 emits a notice that meson.get_cross_property is
>deprecated:
>
>
>NOTICE: Future-deprecated features used:
>
>* 0.58.0: {'meson.get_cross_property'}
>
>
>It seems to hav
On Fri, 20 Sep 2024 09:14:11 +0200
Stefan Laesser wrote:
> + *RTE_MBUF_DYNFIELD(mbuf, timestamp_dynfield_offset,
> + rte_mbuf_timestamp_t *) =
> + (uint64_t)ppd->tp_sec * 10 +
> ppd->tp_nsec;
Maybe the s
> Subject: [PATCH v3 2/6] crypto/openssl: support EDDSA
>
> Support EDDSA crypto algorithm in OpenSSL PMD.
>
> Signed-off-by: Gowrishankar Muthukrishnan
> ---
> drivers/crypto/openssl/openssl_pmd_private.h | 13 ++
> drivers/crypto/openssl/rte_openssl_pmd.c | 223 +++
> dri
Recheck-request: iol-marvell-Functional
putting in a ci lab retest for the failures on this series.
This patch introduces a way for backend to keep track
of the needed information to be able to reconnect without
frontend cooperation.
It will be used for VDUSE, which does not provide interface
for the backend to save and later recover local virtqueues
metadata needed to reconnect.
Vhost-user sup
> From: Stephen Hemminger [mailto:step...@networkplumber.org]
> Sent: Friday, 20 September 2024 16.30
>
> On Fri, 20 Sep 2024 09:14:11 +0200
> Stefan Laesser wrote:
>
> > + *RTE_MBUF_DYNFIELD(mbuf, timestamp_dynfield_offset,
> > + rte_mbuf_timestamp_t
Seeking feedback on the concept.
I have not yet benchmarked performance.
The mempool cache size is configurable, but it could hold 1.5 times the
configured size.
This was confusing for developers, and added complexity when configuring
mempools with caches.
This patch modifies the mempool cache to
Seeking feedback on the concept.
I have not yet benchmarked performance.
The mempool cache size is configurable, but it could hold 1.5 times the
configured size.
This was confusing for developers, and added complexity when configuring
mempools with caches.
This patch modifies the mempool cache to
Recheck-request: iol-marvell-Functional
putting in a ci lab retest for the failures on this series.
From: Jeremy Spewock
v2:
* rebase on next-dts
Jeremy Spewock (1):
dts: add send_packets to test suites and rework packet addressing
dts/framework/test_suite.py| 87 +++---
dts/framework/testbed_model/tg_node.py | 9 +++
2 files changed, 75 insertions(+), 21
From: Jeremy Spewock
Currently the only method provided in the test suite class for sending
packets sends a single packet and then captures the results. There is,
in some cases, a need to send multiple packets at once while not really
needing to capture any traffic received back. The method to do
On Wed, Sep 18, 2024 at 3:41 PM Dean Marx wrote:
>
> +
> +def tx_vlan_set(self, port: int, vlan: int, verify: bool = True) -> None:
One thing to note is that I think this method (unlike rx_vlan_set for
some reason) requires the ports to be stopped for it to work, so we
should probably decora
> Subject: [PATCH] net/mana: support rdma-core via pkg-config in meson
>
> Currently building with custom rdma-core installed in /opt/rdma-core after
> setting PKG_CONFIG_PATH=/opt/rdma-core/lib64/pkgconfig/ results in the below
> meson logs:
> Run-time dependency libmana found: YES 1.0.54.0 Heade
Seeking feedback on the concept.
I have not yet benchmarked performance.
The mempool cache size is configurable, but it could hold 1.5 times the
configured size.
This was confusing for developers, and added complexity when configuring
mempools with caches.
This patch modifies the mempool cache to
Add the packet timestamp from TPACKET_V2 to the mbuf
dynamic rx timestamp register if offload RTE_ETH_RX_OFFLOAD_TIMESTAMP
is enabled.
TPACKET_V2 provides the timestamp with nanosecond resolution
and UNIX origo, i.e. time since 1-JAN-1970 UTC.
Signed-off-by: Stefan Laesser
Acked-by: Morten Brøru
Hi Gregory,
Can you please resolve the compilation issues seen with this patch.
Hairpin offloads packet forwarding between ports.
Packet is expected on Rx port , Rx queue and is forwarded to Tx port
Tx queue .
Testpmd implements a static hairpin configuration scheme.
The new parameter allows
On 2024-09-20 19:13, Morten Brørup wrote:
Seeking feedback on the concept.
I have not yet benchmarked performance.
The mempool cache size is configurable, but it could hold 1.5 times the
configured size.
This was confusing for developers, and added complexity when configuring
mempools with cache
This patch enables VDUSE reconnection support making use of
the newly introduced reconnection mechanism in Vhost
library.
At DPDK VDUSE device creation time, there are two
possibilities:
1. The Kernel VDUSE device does not exist:
a. A reconnection file named after the VUDSE device name
is
This series adds support for VDUSE reconnection.
First patch introduces the reconnection file layout and
track the virtqueues available index updates in the
datapath and control queue.
Second patch adds VDUSE reconnect intialization and some
sanity checks to prevent incompatible reconnections.
C
On Thu, Sep 19, 2024 at 5:02 AM Juraj Linkeš wrote:
>
> > diff --git a/dts/framework/remote_session/testpmd_shell.py
> > b/dts/framework/remote_session/testpmd_shell.py
>
> > @@ -577,6 +577,497 @@ class TestPmdPortStats(TextParser):
> > tx_bps: int = field(metadata=TextParser.find_int(r"Tx-
On Thu, Sep 19, 2024 at 8:35 AM Juraj Linkeš wrote:
>
> > diff --git a/dts/framework/remote_session/testpmd_shell.py
> > b/dts/framework/remote_session/testpmd_shell.py
>
> > @@ -577,6 +577,497 @@ class TestPmdPortStats(TextParser):
> > tx_bps: int = field(metadata=TextParser.find_int(r"Tx-
56 matches
Mail list logo
