To get better performance, using async mode to replace sync mode
However, case UADK_CHAIN_CIPHER_AUTH and UADK_CHAIN_AUTH_CIPHER
still use sync mode for the first operation and async mode for
the second operation since the dependence.
Also RTE_CRYPTO_AUTH_OP_VERIFY will hold the generated auth in
To get better performance, using async mode to replace sync mode
Signed-off-by: Zhangfei Gao
---
doc/guides/compressdevs/uadk.rst | 15 +++
drivers/compress/uadk/meson.build | 4 +-
drivers/compress/uadk/uadk_compress_pmd.c | 94 ++-
.../compress/uad
Realize async mode to replace sync mode for better performance
v3:
uadk 2.9 is released
use different label name to fix build warning:
"duplicate label table_uadk_versions"
v2:
Add version check, requires uadk >= 2.9
Add version table describing DPDK and external UADK library
ver
added RTE_PMD_DLB2_ prefix to dlb2 token pop mode enmus
to avoid name comflict. These enums are passed to public
API rte_pmd_dlb2_set_token_pop_mode().
Fixes: c667583d82f4 ("event/dlb2: add token pop API")
Cc: sta...@dpdk.org
Signed-off-by: Pravin Pathak
---
drivers/event/dlb2/dlb2.c |
Adds null pointer check before executing CPT instruction.
Signed-off-by: Rahul Bhansali
---
drivers/common/cnxk/roc_nix_inl.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/drivers/common/cnxk/roc_nix_inl.c
b/drivers/common/cnxk/roc_nix_inl.c
index bee8e25c7c..c7637dd
Add lock on security session update, write and read to
prevent corruption.
Signed-off-by: Rahul Bhansali
---
drivers/net/cnxk/cn10k_ethdev_sec.c | 65 +++--
drivers/net/cnxk/cnxk_ethdev_sec.c | 60 --
2 files changed, 109 insertions(+), 16 deletio
Hi, Maxime
I understand your point. However we coould have a situation like this, could
resize twice:
VHOST_CONFIG: (/usr/local/var/run/stdvio4) read message VHOST_USER_SET_MEM_TABLE
VHOST_CONFIG: (/usr/local/var/run/stdvio4) guest memory region size: 0x4000
VHOST_CONFIG: (/usr/local/var/run
The VTNET_SQ_CQ_QUEUE_IDX is orphaned, drop it.
Signed-off-by: Hengqi Chen
---
drivers/net/virtio/virtqueue.h | 1 -
1 file changed, 1 deletion(-)
diff --git a/drivers/net/virtio/virtqueue.h b/drivers/net/virtio/virtqueue.h
index 3b468bd4f4..1f0e6ae77e 100644
--- a/drivers/net/virtio/virtqueue.
Yes the windows failure was unrelated to your patch - sorry about the
noise.
Looks like Paul has applied this commit to next-dts.
On Mon, Jun 16, 2025 at 4:23 AM Clemens Famulla-Conrad <
cfamullacon...@suse.com> wrote:
> > Thanks. This needs to pass the linter, see my suggestions below.
>
> Than
> From: Stephen Hemminger [mailto:step...@networkplumber.org]
> Sent: Monday, 16 June 2025 17.30
>
> On Mon, 16 Jun 2025 10:29:05 +0300
> Shani Peretz wrote:
>
> > This feature is designed to monitor the lifecycle of mempool objects
> > as they move between the application and the PMD.
> >
> > I
- Used global node mbuf in ip[4|6]_lookup/rewrite nodes
- Redefine node_mbuf_priv1() to rte_node_mbuf_overload_fields_get()
Signed-off-by: Nitin Saxena
---
lib/node/ip4_lookup.c | 14 ---
lib/node/ip4_lookup_fib.c | 17 ++---
lib/node/ip4_rewrite.c
This patch defines rte_node specific dynamic field structure
(rte_node_mbuf_dynfield_t)
rte_node_mbuf_dynfield_t structure holds two types of fields
- Persistent data fields which are preserved across graph walk.
Currently size of persistent data fields is zero.
- Overloadable data fields which
Currently each rte_node registers separate mbuf dynamic fields for their
own purpose. This leads to wastage of mbuf space as once mbuf get passed
a particular node, the registered dynamic field(by that node) is no
longer used.
This patch series adds a global/common mbuf dynamic field which is
reus
Hi Thomas,
Thanks,
Nitin
On Wed, Jun 11, 2025 at 5:15 PM Thomas Monjalon wrote:
>
> 28/04/2025 12:37, Nitin Saxena:
> > Currently each rte_node registers separate mbuf dynamic fields for their
> > own purpose. This leads to wastage of mbuf space as once mbuf get passed
> > a particular node, the
> -Original Message-
> From: Richardson, Bruce
> Sent: Wednesday, June 18, 2025 9:41 PM
> To: Ye, MingjinX
> Cc: dev@dpdk.org; sta...@dpdk.org; Burakov, Anatoly
>
> Subject: Re: [PATCH v2 2/2] net/ice: fix updating DCF RSS hash
>
> On Wed, Jun 18, 2025 at 07:08:44AM +, Mingjin Ye
On Wed, 18 Jun 2025 at 05:13, Akhil Goyal wrote:
>
> > Realize async mode to replace sync mode for better performance
> >
> > v2:
> > Add version check, requires uadk >= 2.9
>
> This version DO NOT exist as of now.
Sorry, Akhil
The tag 2.9 was just released today
And will send v3 with different
Enable Tx QINQ offload if the VF reports support for inserting both an
outer and inner VLAN tag. The VF capabilities report the locations for
placing each of the tags - either L2TAG1 in the tx descriptor or L2TAG2
in the context descriptor. Use this information to configure the
descriptors correctl
Re-enable support for VLAN insertion and implement support for QinQ
insertion on the iavf avx-512 transmit path.
Signed-off-by: Ciara Loftus
---
drivers/net/intel/iavf/iavf_rxtx.h| 6 +-
drivers/net/intel/iavf/iavf_rxtx_vec_avx2.c | 24 +++
drivers/net/intel/iavf/iavf_rxtx_vec
Prior to this change, it was possible the scalar path would be selected
even if the avx-512 path was available. This was due to the ordering of
the logic in the iavf_set_tx_function function. Support for all three
vector paths (sse, avx2 and avx-512) was first established and then in
that order, th
Nice idea,
Some comments:
1. The first link is broken for me.
2. I like the idea that there is a separate GitHub repository. If I
understand correctly the new repository is a clone of the main one,
and only used by the bot is that correct?
3. I think main usage is also for the reviewer
On Thu, Jun 19, 2025 at 9:34 AM Pravin Pathak wrote:
>
> From: Tirthendu Sarkar
>
> dlb2_eventdev_info_get() that implements rte_event_dev_info_get() should
> return the maximum available credits as supported by HW.
>
> Set maximum credits before device probing by checking HW version.
>
> Fixes:
>Sent: Thursday, June 12, 2025 05:43
>To: Jerin Jacob; tho...@monjalon.net; david.march...@redhat.com; Pavan
>Nikhilesh Bhagavatula
>Cc: dev@dpdk.org
>Subject: [PATCH] test/event: fix event vector adapter timeouts
>
>From: Pavan Nikhilesh
>
>Update event vector adapter test to depend on service
>
> Subject: [PATCH] net/af_xdp: add rx/tx queue support for af_xdp
>
> Driver like GVE do support rx/tx queue configuration
> rather than combined queue. Enable AF_XDP vdev to use
> rx/tx queue configuration instead of combined queue
> configuration if available.
>
> Signed-off-by: Shivaji Kant
>
Hi Kai Ji,
Debugged it using openssl source code used in these distros and the failure
found due to below
Implicit rejection added in PKCS padding.
https://github.com/openssl/openssl/pull/13817/commits/3b99dcb1a6df6853581d3035c6d9a23a831716bb
As part of the above fix, rsa_ossl_private_decrypt()
On Thu, Jun 19, 2025 at 12:45:00PM +0530, Shaiq Wani wrote:
> v1 - fixed the series name.
> v2 - Addressed review comments,fixed idpf kernel driver version.
>
>
>
> Shaiq Wani (2):
> doc: update recommended matching list for cpfl PMD
> doc: update recommended matching list for idpf PMD
>
>
Updated the recommended MEV-ts release versions corresponding to
DPDK releases in the idpf driver documentation.
Signed-off-by: Shaiq Wani
---
doc/guides/nics/idpf.rst | 4
1 file changed, 4 insertions(+)
diff --git a/doc/guides/nics/idpf.rst b/doc/guides/nics/idpf.rst
index b55de3abd3..b9
v1 - fixed the series name.
v2 - Addressed review comments,fixed idpf kernel driver version.
Shaiq Wani (2):
doc: update recommended matching list for cpfl PMD
doc: update recommended matching list for idpf PMD
doc/guides/nics/cpfl.rst | 2 ++
doc/guides/nics/idpf.rst | 4
2 files cha
Rdma-core exposes E-Switch Manager vport metadata
for matching in flow rules.
Use that metadata when available.
Signed-off-by: Rongwei Liu
Acked-by: Dariusz Sosnowski
---
drivers/net/mlx5/mlx5_flow_dv.c | 18 +++---
1 file changed, 15 insertions(+), 3 deletions(-)
diff --git a/driv
Updated the recommended MEV-ts release version corresponding
to DPDK release in the cpfl driver documentation.
Signed-off-by: Shaiq Wani
---
doc/guides/nics/cpfl.rst | 2 ++
1 file changed, 2 insertions(+)
diff --git a/doc/guides/nics/cpfl.rst b/doc/guides/nics/cpfl.rst
index 3bfab2136d..aa5bed
Caught by UBSan:
../lib/eal/x86/include/rte_atomic_64.h:206:21: runtime error:
member access within misaligned address 0x7ffd9c67f228 for
type 'const rte_int128_t', which requires 16 byte alignment
0x7ffd9c67f228: note: pointer points here
00 00 00 00 c0 5d 3e 00 01 00 00
On Wed, Jun 18, 2025 at 07:08:44AM +, Mingjin Ye wrote:
> The driver does nothing if the rss key is invalid during the rss
> hash update.
>
> Extract the update rss key code from ice_dcf_dev_rss_hash_update to
> ice_dcf_set_rss_key and make it consistent with the pf behaviour.
>
> Fixes: c223
On Tue, Jun 17, 2025 at 10:36:10PM +, Soumyadeep Hore wrote:
> Added TxPP support for E830 adapters.
> ---
> v2:
> - Fixed Compilation warnings
> ---
> v3:
> - Fixed further Compilation warnings (uinitialised one)
> ---
> v4:
> - Addressed Bruce's comments
> ---
> v5:
> - Addressed Bruce's and
Currently, there are no ways to retrieve the queue size
set by vhost backend. Implement the {rxq,txq}_info_get
callbacks so that DPDK applications can use the queue
size to setup mempool properly.
Signed-off-by: Hengqi Chen
---
drivers/net/virtio/virtio_ethdev.c | 24
1
If mlx5 malloc with specified socket failed,
malloc function would return an error.
Cross NUMA support means that when there is no
memory on the local NUMA, use other available
NUMA memory for port init and start.
To support cross NUMA,
added a flag to enable NUMA fallback to any NUMA.
Fallback w
This test was doing nothing as it could not find the telemetry client
script following the test suite rework.
Caught while looking at UNH unit test logs:
/root/workspace/Generic-Unit-Test-DPDK/dpdk/app/test/suites/test_telemetry.sh:
18: /root/workspace/Generic-Unit-Test-DPDK/dpdk/app/usertools/dp
When troubleshooting unit test failures and repeating jobs in GHA,
the absence of ccache makes the whole process way slower.
Signed-off-by: David Marchand
---
.github/workflows/build.yml | 6 ++
1 file changed, 6 insertions(+)
diff --git a/.github/workflows/build.yml b/.github/workflows/bui
pl->map is a uint32_t.
Caught by UBSan:
../lib/cmdline/cmdline_parse_portlist.c:27:17: runtime error:
left shift of 1 by 31 places cannot be represented in type 'int'
SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior
../lib/cmdline/cmdline_parse_portlist.c:27:17 in
Fixes: a
In a minimal build, the mempool/stack driver is disabled.
Separate the code specific to this external driver and rename unrelated
variables.
Signed-off-by: David Marchand
---
app/test/test_mempool.c | 32
1 file changed, 20 insertions(+), 12 deletions(-)
diff --
Doing arithmetics with the NULL pointer is undefined.
Caught by UBSan:
../lib/cmdline/cmdline_parse_portlist.c:40:19: runtime error:
applying non-zero offset 1 to null pointer
SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior
../lib/cmdline/cmdline_parse_portlist.c:40:19 in
For '.' and '..' directories (or any short file name),
a out of bound issue occurs.
Caught by UBSan:
EAL: Detected shared linkage of DPDK
../lib/eal/common/eal_common_options.c:420:15: runtime error: index -2
out of bounds for type 'char[256]'
#0 0x7f867eedf206 in eal_plugindir_init
This series fixes a number of issues reported by UBSan and adds a simple
job in GHA to avoid introducing undefined behavior in the core
components.
There is way more work/fixes to do if we want to run with a full set of
components, but baby steps first.
--
David Marchand
David Marchand (10):
Enable UBSan in GHA.
There are still a lot of issues so only run unit tests for a "mini"
target.
Building with debugoptimized forces -O2 and consumes too much memory
with UBSan, prefer plain build (iow -O0) even though this hides a number
of build issues.
Signed-off-by: David Marchand
---
.ci/l
Doing arithmetics with the NULL pointer is undefined.
Caught by UBSan:
../app/test/test_tailq.c:111:9: runtime error:
member access within null pointer of type 'struct rte_tailq_head'
Fixes: f6b4f6c9c123 ("tailq: use a single cast macro")
Cc: sta...@dpdk.org
Signed-off-by: David Marchan
Caught by UBSan:
../lib/hash/rte_thash.c:421:8: runtime error: load of misaligned address
0x0001816c2da3 for type 'uint32_t' (aka 'unsigned int'),
which requires 4 byte alignment
Fixes: 28ebff11c2dc ("hash: add predictable RSS")
Cc: sta...@dpdk.org
Signed-off-by: David Marchand
44 matches
Mail list logo
