Re: [PATCH v7 00/15] XSC PMD for Yunsilicon NICs

On 2025/2/15 7:27, Stephen Hemminger wrote: > On Tue, 28 Jan 2025 22:47:22 +0800 > "Renyong Wan" wrote: > >> This xsc PMD (**librte_net_xsc**) provides poll mode driver for >> Yunsilicon metaScale serials NICs. >> >> Features: >> - >> - MTU update >> - TSO >> - RSS hash >> - RSS key update

Re: [PATCH v12 00/28] [v12]drivers/net Add Support mucse N10 Pmd Driver

I ran this driver through the evaluation copy of PVS studio and it reported many warnings you should address. Not all of them are valid, the tool seems to have a lot of false positives. MESSAGES LocationCodeMessage Fails/Info rnp_mac.c (258) V610 Un

PVS studio warnings in zxdh

Now that zxdh is main branch, ran the PVS studio analyzer over it and reported many new warnings. MESSAGES LocationCodeMessage Fails/Info zxdh_common.c (223) V1032 The pointer 'recv_buf' is cast to a more strictly aligned pointer type. zxdh_ethde

Re: [PATCH v7 00/15] XSC PMD for Yunsilicon NICs

On Tue, 28 Jan 2025 22:47:22 +0800 "Renyong Wan" wrote: > This xsc PMD (**librte_net_xsc**) provides poll mode driver for > Yunsilicon metaScale serials NICs. > > Features: > - > - MTU update > - TSO > - RSS hash > - RSS key update > - RSS reta update > - L3 checksum offload > - L4 chec

Re: [PATCH v2 7/7] dts: remove node distinction

Right off the bat, I like this organization a lot more. Definitely a lot easier to navigate the code. Reviewed-by: Nicholas Pratte On Wed, Feb 12, 2025 at 11:46 AM Luca Vizzarro wrote: > > Remove the distinction between SUT and TG nodes for configuration > purposes. As DPDK and the traffic gene

Re: [PATCH v2 6/7] dts: revamp runtime internals

This definitely cleans things up a lot! Reviewed-by: Nicholas Pratte > @@ -217,9 +218,7 @@ def get_supported_capabilities( > ) > if cls.capabilities_to_check: > capabilities_to_check_map = cls._get_decorated_capabilities_map() > -with TestPmdShell( > -

[PATCH v2 5/6] net/enetfec: remove unneeded deferred start check

Already checked in ethdev now. Signed-off-by: Stephen Hemminger --- drivers/net/enetfec/enet_ethdev.c | 16 ++-- 1 file changed, 2 insertions(+), 14 deletions(-) diff --git a/drivers/net/enetfec/enet_ethdev.c b/drivers/net/enetfec/enet_ethdev.c index 0d27b63953..7c2926409e 100644 -

[PATCH v2 6/6] net/virtio: remove unneeded deferred start check

Already checked in ethdev now. Signed-off-by: Stephen Hemminger --- drivers/net/virtio/virtio_rxtx.c | 10 -- 1 file changed, 10 deletions(-) diff --git a/drivers/net/virtio/virtio_rxtx.c b/drivers/net/virtio/virtio_rxtx.c index b67f063b31..6ae3a6eb12 100644 --- a/drivers/net/virtio/vir

[PATCH v2 4/6] net/dpaa2: remove unneeded deferred start check

Already checked in ethdev now. Signed-off-by: Stephen Hemminger --- drivers/net/dpaa2/dpaa2_ethdev.c | 14 -- 1 file changed, 14 deletions(-) diff --git a/drivers/net/dpaa2/dpaa2_ethdev.c b/drivers/net/dpaa2/dpaa2_ethdev.c index a9bce854c3..8624bba0ce 100644 --- a/drivers/net/dpaa2/

[PATCH v2 3/6] net/dpaa: remove unnecessary deferred start check

This check is now done in ethdev layer. Signed-off-by: Stephen Hemminger --- drivers/net/dpaa/dpaa_ethdev.c | 10 -- 1 file changed, 10 deletions(-) diff --git a/drivers/net/dpaa/dpaa_ethdev.c b/drivers/net/dpaa/dpaa_ethdev.c index e8d34e5898..ce48f0 100644 --- a/drivers/net/dpaa/dp

[PATCH v2 1/6] ethdev: check that device supports deferred start

The check for supporting deferred start should be handled at the ethdev level for all devices. Signed-off-by: Stephen Hemminger Acked-by: Andrew Rybchenko --- lib/ethdev/rte_ethdev.c | 14 ++ 1 file changed, 14 insertions(+) diff --git a/lib/ethdev/rte_ethdev.c b/lib/ethdev/rte_eth

[PATCH v2 0/6] fix the handling of deferred start in ethdev

There is a inconsistency and lack of documentation around the deferred start configuration flag. The queue deferred start only makes sense if the driver has a way to start the queue later via ethdev ops. Therefore move the check into the ethdev queue configuration handling. Fix the driver documenta

[PATCH v2 6/7] vhost: replace open coded TAILQ_FOREACH_SAFE

Proper macro is now in EAL rte_queue.h use it instead. Signed-off-by: Stephen Hemminger --- lib/vhost/socket.c | 11 ++- 1 file changed, 2 insertions(+), 9 deletions(-) diff --git a/lib/vhost/socket.c b/lib/vhost/socket.c index 4c6c551c3d..dc9c9da9d1 100644 --- a/lib/vhost/socket.c +++

[PATCH v2 5/7] net/iavf: replace local version of TAILQ_FOREACH_SAFE

Now in EAL as rte_queue.h Signed-off-by: Stephen Hemminger --- drivers/net/intel/iavf/iavf_vchnl.c | 8 +--- 1 file changed, 1 insertion(+), 7 deletions(-) diff --git a/drivers/net/intel/iavf/iavf_vchnl.c b/drivers/net/intel/iavf/iavf_vchnl.c index 6feca8435e..0b7c35250c 100644 --- a/drive

[PATCH v2 4/7] net/bnxt: fix use after free

The filter cleanup loop was using STAILQ_FOREACH and rte_free and would dereference the filter after free. Found by build with -Dbsanitize=address,undefined Fixes: e8fe0e067b68 ("net/bnxt: fix allocation of PF info struct") Cc: ajit.khapa...@broadcom.com Cc: sta...@dpdk.org Signed-off-by: Stephe

[PATCH v2 7/7] raw/ifpga: use EAL version of TAILQ_FOREACH_SAFE

Prefer the EAL version over local version of macro. Signed-off-by: Stephen Hemminger Reviewed-by: Rosen Xu --- drivers/raw/ifpga/base/opae_osdep.h | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/raw/ifpga/base/opae_osdep.h b/drivers/raw/ifpga/base/opae_osdep.h index e35a21c80e..b48

[PATCH v2 0/7] Introduce FOREACH_SAFE macros

This series adds common macros for safe iteration over lists. It is a subset copy of the macros from FreeBSD that are missing from the Linux header sys/queue.h Chose this over several other options: - let each driver define their own as needed. One Intel driver got it wrong, others will as w

[PATCH v2 3/7] bus/fslmc: fix use after free

The cleanup loop would deference the dpio_dev after freeing. Use TAILQ_FOREACH_SAFE to fix that. Found by building with sanitizer undefined flag. Fixes: e55d0494ab98 ("bus/fslmc: support secondary process") Cc: shreyansh.j...@nxp.com Cc: sta...@dpdk.org Signed-off-by: Stephen Hemminger --- drive

[PATCH v2 2/7] net/qede: fix use after free

The loop cleaning up flowdir resources was using SLIST_FOREACH but the inner loop would call rte_free. Found by building with address sanitizer undefined check. Also remove needless initialization, and null check. Fixes: f5765f66f9bb ("net/qede: refactor flow director into generic aRFS") Cc: shah

[PATCH v2 1/7] eal: add queue macro extensions from FreeBSD

The Linux version of sys/queue.h is frozen at an older version and is missing the _SAFE macro variants. Several drivers started introducing the own workarounds for this. Should be handled in EAL. Signed-off-by: Stephen Hemminger --- lib/eal/include/meson.build | 3 +- lib/eal/include/rte_queue

Re: [PATCH v3 1/2] net: add ptype parse for tunnel packets

On Fri, 14 Feb 2025 09:56:37 +0800 Jie Hai wrote: > Add packet types parse for vxlan/vxlan-gpe/gtp/geneve packets. > > Signed-off-by: Jie Hai > --- Not sure about this. The original purpose of ptype was to support hardware offload information. Does any driver do this ptype detection in tunnel

RE: [EXTERNAL] [RFC PATCH v6 1/2] eventdev: add atomic queue to test-eventdev app

> On Tue, 2025-02-04 at 16:11 +, Pavan Nikhilesh Bhagavatula wrote: > > > Add an atomic queue test to the test-eventdev app, which is based on the > > > order queue test that exclusively uses atomic queues. > > > > > > This makes it compatible with event devices such as the > > > distributed so

[PATCH] net/mlx5: fix flow null ops seg fault

flow null driver ops may get selected and try to call .list_create causing a seg fault. Full details and back traces in bugzilla. Add a null ops implementation for .list_create. Bugzilla ID: 1661 Fixes: e38776c36c8a ("net/mlx5: introduce HWS for non-template flow API") Cc: sta...@dpdk.org Signe

Re: [EXTERNAL] [RFC PATCH v6 1/2] eventdev: add atomic queue to test-eventdev app

On Tue, 2025-02-04 at 16:11 +, Pavan Nikhilesh Bhagavatula wrote: > > Add an atomic queue test to the test-eventdev app, which is based on the > > order queue test that exclusively uses atomic queues. > > > > This makes it compatible with event devices such as the > > distributed software ev

[DPDK/ethdev Bug 1661] mlx5: rte_flow_create can cause a seg fault

https://bugs.dpdk.org/show_bug.cgi?id=1661 Bug ID: 1661 Summary: mlx5: rte_flow_create can cause a seg fault Product: DPDK Version: 24.11 Hardware: All OS: All Status: UNCONFIRMED Severity: normal

Re: [PATCH v7 0/7] latencystats: cleanup

On Thu, Feb 13, 2025 at 01:20:29PM -0800, Stephen Hemminger wrote: > Latencystats uses unnecessary floating point and is inaccurate if the number > of cycles per ns is not an integral value. Cleanup the code and add > more tests as well. > > v7 - rebase and drop already merged VLA patch > > Steph

RE: [PATCH v7 2/7] latencystats: do not use floating point

> The cycle counts do not need to be stored as floating point. > Instead keep track of latency in cycles, and convert to > nanoseconds when read. > > Use scaled math with exponential Weighted Moving Average weight > of .25 to avoid use of floating point for that. > > The average latency took to

Re: [PATCH v2] examples/l3fwd: add option to set Tx burst size

Acked-by: Chengwen Feng On 2024/12/4 10:06, Jie Hai wrote: > The application send packets only when the buffer is full, or the > buffer is empty and the packets to be sent extends TX_PKT_BURST. > The change of MAX_PKT_BURST make TX buffer size and TX_PKT_BURST > increase, while the default cache