On Mon, Jun 16, 2025 at 11:49:40AM +0100, Bruce Richardson wrote:
> Coverity (correctly) identified an issue[1] where, after the recent
> rework[2], the internal flag, used by argparse to track what arguments
> were previously encountered or not, was out of range for the type and no
> longer having
Definers are HW objects that are used for matching, rte items are
translated to definers, each definer holds the fields and bit-masks
used for HW flow matching.
This patch recognize IP-in-IP tunnel type in definer layer, including
the following combinations:
- IPv4 over IPv4
- IPv4 over IPv6
-
IP-in-IP tunnel type includes the following combination:
- IPv4 over IPv4
- IPv4 over IPv6
- IPv6 over IPv4
- IPv6 over IPv6
It's okay to have MLX5_FLOW_LAYER_IPIP flag when validate IPv6 item.
And it's okay to have outer L3 layer as either IPv4 or IPv6 when
setting the IP-in-IP tunnel item fl
This patchset support IPIP Tunnel in NT2HWS and HWS, including the
following {IPv4 x IPv6} combinations:
- IPv4 over IPv4
- IPv4 over IPv6
- IPv6 over IPv4
- IPv6 over IPv6
This feature cover both basic IPIP and IPIP over VXLAN Tunnel types.
Junfeng Guo (2):
net/mlx5/hws: recognize IPIP in
On Fri, Jun 20, 2025 at 4:07 AM Pravin Pathak wrote:
>
> 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
>
> S
On Fri, Jun 20, 2025 at 10:57 AM Rahul Bhansali wrote:
>
> Adds null pointer check before executing CPT instruction.
>
> Signed-off-by: Rahul Bhansali
Add Fixes tag and start the commit with common/cnxk: fix ...
On Thu, Jun 19, 2025 at 09:10:33AM +0200, David Marchand wrote:
> 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 ("tail
On Thu, Jun 19, 2025 at 09:10:32AM +0200, David Marchand wrote:
> 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
On Thu, Jun 19, 2025 at 09:10:28AM +0200, David Marchand wrote:
> 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_te
Install only architecture specific headers to avoid header pollution.
Signed-off-by: Tomasz Duszynski
---
lib/pmu/meson.build | 10 +-
1 file changed, 5 insertions(+), 5 deletions(-)
diff --git a/lib/pmu/meson.build b/lib/pmu/meson.build
index bcb6d10f1a..53013d0ed6 100644
--- a/lib/pmu
RTE_LIB_PMU is defined by build system automatically if DPDK is build
for Linux. Otherwise it's missing.
That said all cases where conditional compilation is required
can be handled without introducing other macros.
Signed-off-by: Tomasz Duszynski
---
lib/pmu/rte_pmu.h | 3 ---
1 file changed,
Make sure that out-of-bound access does not happen by saving one byte in
buffer for NUL terminator.
Fixes: a8926a65ad1d ("pmu: support Arm")
Fixes: 960c43184c4d ("pmu: introduce library for reading PMU events")
Cc: tduszyn...@marvell.com
Signed-off-by: Tomasz Duszynski
---
lib/pmu/pmu.c |
DPDK test suite provides much broader architecture coverage than
what can be tested locally so enable the test to help identify
potential issues.
Signed-off-by: Tomasz Duszynski
---
app/test/test_pmu.c | 44 +---
1 file changed, 41 insertions(+), 3 deletio
Add a dummy rte_pmu_read() definition for chkincs when
ALLOW_EXPERIMENTAL_API is not defined to suppress warnings from
use of experimental APIs in tracepoints.
Fixes: 960c43184c4d ("pmu: introduce library for reading PMU events")
Signed-off-by: Tomasz Duszynski
---
lib/pmu/rte_pmu.h | 4
1
In order to profile app, one needs to store significant amount of samples
somewhere for an analysis later on.
Since trace library supports storing data in a CTF format,
lets take advantage of that and add a dedicated PMU tracepoint.
Signed-off-by: Tomasz Duszynski
---
MAINTAINERS
On 6/19/25 10:10 AM, David Marchand wrote:
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
Acked-by: Andrew Rybchenko
On Thu, Jun 19, 2025 at 09:10:31AM +0200, David Marchand wrote:
> 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:
When sending a burst of output packets on a stopped transmit queue,
the packets are written to a memory mapped address.
On queue start the packets are processed and transmitted by the NIC.
Signed-off-by: Oleksandr Kolomeiets
---
drivers/net/ntnic/ntnic_ethdev.c | 4
1 file changed, 4 insert
Check if kernel returns 0 interrupt vectors and try another interrupt
type in that case. Failing to check the vector count can select an
interrupt type that's unusable.
Signed-off-by: Pekka Riikonen
---
drivers/bus/pci/linux/pci_vfio.c | 4
1 file changed, 4 insertions(+)
diff --git a/dr
Check if kernel returns 0 interrupt vectors and try another interrupt
type in that case. Failing to check the vector count can select an
interrupt type that's unusable.
Signed-off-by: Pekka Riikonen
---
drivers/bus/pci/linux/pci_vfio.c | 4
1 file changed, 4 insertions(+)
diff --git a/dr
On Fri, Jun 20, 2025 at 02:30:23PM +0100, Anatoly Burakov wrote:
> This issue was reported by static analysis. It is a false positive,
> because both `rte_socket_count` and `rte_socket_id_by_idx` only report
> information about physical sockets, and these specific calls are made
> during EAL initia
This issue was reported by static analysis. It is a false positive,
because both `rte_socket_count` and `rte_socket_id_by_idx` only report
information about physical sockets, and these specific calls are made
during EAL initialization, so no other sockets (i.e. external) could have
been available,
Following RSA encrypt op, same plaintext buffer is used as output
buffer for decrypt op, hence comparing plaintext buffer against
same buffer pointer in crypto op always succeed irrespective of
whether decrypt op succeeds or not. This patch fixes this issue
with a local buffer for crypto op.
Fixes
For i40e kernel drivers which support either vlan(v1) or vlan(v2)
VIRTCHNL OP,it will set strip on when setting filter on. But dpdk
side will not change strip flag. To be consistent with dpdk side,
explicitly disable strip again.
Bugzilla ID:1725
Signed-off-by: Amiya Ranjan Mohakud
---
drivers/
https://bugs.dpdk.org/show_bug.cgi?id=1729
Bug ID: 1729
Summary: RSS offload types are undocumented
Product: DPDK
Version: 25.03
Hardware: All
OS: All
Status: UNCONFIRMED
Severity: normal
Priorit
Replace static per-architecture ops with a callback-based ops
structure. This makes architecture-specific PMU implementations more
modular and easier to extend.
Signed-off-by: Tomasz Duszynski
---
lib/pmu/pmu.c | 17 +
lib/pmu/pmu_arm64.c | 19 +--
lib/p
This series does some cleanup and refactoring around the rc1 code like:
trimming unused headers,
switching to callbacks for per-arch handling, and adding trace support. It also
re-enables existing
base test to help catch reported issues on some architectures.
v3:
- do not export __rte_pmu_trace_
If private exponent is available, it should be included within
RSA session as per RFC 8017 (A.1.2). OpenSSL 1.1.1 implementation
rely on this private exponent, to implicitly reject invalid cipher.
Hence, check if it is available for session and include it.
Fixes: 3e9d6bd447fb ("crypto/openssl: add
On Thu, Jun 19, 2025 at 09:10:30AM +0200, David Marchand wrote:
> 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
> ou
This patchset includes:
* feature start/stop queues on HW layer.
* feature deferred start for queues.
* Improvement for memory mappings when IOMMU is unoptimized.
* Improvement for logging
Oleksandr Kolomeiets (4):
net/ntnic: implement start/stop for Rx/Tx queue
Handle rx_deferred_start and tx_deferred_start flags
during configuration done in rx_queue_setup and tx_queue_setup,
so that marked queues do not start with dev_start.
Signed-off-by: Oleksandr Kolomeiets
---
drivers/net/ntnic/dbsconfig/ntnic_dbsconfig.c | 46 +++
drivers/net/ntni
The following functions exported by the driver were stubs
which merely changed the status flags:
* rx_queue_start
* rx_queue_stop
* tx_queue_start
* tx_queue_stop
Proper implementation was added to control queues's state.
Signed-off-by: Oleksandr Kolomeiets
---
drivers/n
Perform unmapping in a default container, which is used by queues.
Handle multiple mappings when IOMMU is unoptimized.
Signed-off-by: Oleksandr Kolomeiets
---
drivers/net/ntnic/include/ntos_drv.h | 1 +
drivers/net/ntnic/ntnic_ethdev.c | 26 --
drivers/net/ntnic/ntni
Don't the Intel i40e and e1000 drivers support RSS on IP addresses [1]?
It looks like they do, but don't report it.
For i40e:
Accepted: 0x2288 FRAG_IPV4 NONFRAG_IPV4_OTHER FRAG_IPV6 NONFRAG_IPV6_OTHER
Unsupported: 0x8104 IPV4 IPV6 IPV6_EX
For e1000:
Accepted: 0x8104 IPV4 IPV6 IPV6_EX
Unsupported:
On Fri, Jun 20, 2025 at 12:24:24PM +0200, Morten Brørup wrote:
> Don't the Intel i40e and e1000 drivers support RSS on IP addresses [1]?
> It looks like they do, but don't report it.
>
> For i40e:
> Accepted: 0x2288 FRAG_IPV4 NONFRAG_IPV4_OTHER FRAG_IPV6 NONFRAG_IPV6_OTHER
> Unsupported: 0x8104 IP
As per RFC 8017 (A.1.2), RSA private key should be represented
with both exponent and CRT forms of data for inter operability.
Fixes: a379799cbaaf ("test/crypto: add RSA key type CRT")
Cc: sta...@dpdk.org
Signed-off-by: Gowrishankar Muthukrishnan
---
app/test/test_cryptodev_asym.c |
IBM - Power Systems Testing
DPDK v25.07-rc1-8-gdbeb8ae8d0
* Basic PF on Mellanox: No issue found
* Performance: not tested.
* OS:- RHEL 9.5 kernel: 5.14.0-503.11.1.el9_5.ppc64le
with gcc version 11.5.0 20240719 (Red Hat 11.5.0-2)
Systems tested:
- LPARs on IBM Power10 CHRP IBM,9105-22
For i40e kernel drivers which support either vlan(v1) or vlan(v2)
VIRTCHNL OP,it will set strip on when setting filter on. But dpdk
side will not change strip flag. To be consistent with dpdk side,
explicitly disable strip again.
Bugzilla ID:1725
Cc: sta...@dpdk.org
Signed-off-by: Amiya Ranja
38 matches
Mail list logo
