Re: 24.11.3 patches review and test

On 15/08/2025 18:47, Ali Alnubani wrote: > > On 8/13/25 11:25 PM, Ali Alnubani wrote: >> On 7/24/25 7:15 PM, Kevin Traynor wrote: >> >>> Hi all, >>> >>> Here is a list of patches targeted for stable release 24.11.3. >>> >>> The planned date for the final release is 12th August 2025. >>> >>> Please

Re: 24.11.3 patches review and test

On 8/13/25 11:25 PM, Ali Alnubani wrote: On 7/24/25 7:15 PM, Kevin Traynor wrote: Hi all, Here is a list of patches targeted for stable release 24.11.3. The planned date for the final release is 12th August 2025. Please help with testing and validation of your use cases and report any issu

[PATCH v3 7/7] config: enable comma warnings

Clang (3.9 or later) has a -Wcomma that emits warnings for questionable uses of the comma operator. Don't enable it for drivers since many drivers still use comma operator unnecessarily. Signed-off-by: Stephen Hemminger Acked-by: Konstantin Ananyev Acked-by: Chengwen Feng --- config/meson.bui

[PATCH v3 5/7] app/testpmd: replace comma operator with bracket

Use while with brackets in favor of comma operator. In flow print replace with temporary variable. Signed-off-by: Stephen Hemminger Acked-by: Chengwen Feng --- app/test-pmd/cmdline_flow.c | 6 -- app/test-pmd/config.c | 10 ++ 2 files changed, 10 insertions(+), 6 deletions(-

[PATCH v3 6/7] examples: remove unnecessary use of comma operator

Use of comma as statement separator is discouraged and reported as warning by clang with -Wcomma Signed-off-by: Stephen Hemminger --- examples/l2fwd-event/l2fwd_event_generic.c | 2 +- examples/l3fwd/l3fwd_event_generic.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git

[PATCH v3 1/7] eal: replace unnecessary comma operator

Use of comma as statement separator is discouraged and reported as warning by clang with -Wcomma Signed-off-by: Stephen Hemminger Acked-by: Konstantin Ananyev Acked-by: Chengwen Feng --- lib/eal/common/rte_malloc.c | 4 ++-- lib/eal/linux/eal_alarm.c | 6 -- 2 files changed, 6 insertions

[PATCH v3 3/7] graph: replace unnecessary comma operator

Use of comma as statement separator is discouraged and reported as warning by clang with -Wcomma Signed-off-by: Stephen Hemminger Acked-by: Konstantin Ananyev Acked-by: Chengwen Feng --- lib/graph/graph_stats.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/lib/graph/graph

[PATCH v3 4/7] test: replace unnecessary comma operators

Use of comma as statement separator is discouraged and reported as warning by clang with -Wcomma Signed-off-by: Stephen Hemminger Acked-by: Konstantin Ananyev Acked-by: Chengwen Feng --- app/test/test_compressdev.c | 4 ++-- app/test/test_link_bonding_rssconf.c | 2 +- app/test/test_t

[PATCH v3 2/7] ipsec: replace unnecessary comma operator

Use of comma as statement separator is discouraged and reported as warning by clang with -Wcomma Signed-off-by: Stephen Hemminger Acked-by: Konstantin Ananyev Acked-by: Chengwen Feng --- lib/ipsec/misc.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/lib/ipsec/misc.h b/l

[PATCH v3 0/7] avoid unnecessary use of comma operator

Clang has ability to detect suspicious use of comma operator where a semicolon would have the same effect. Easy enough to fix in DPDK code. V3 - fix examples as well Stephen Hemminger (7): eal: replace unnecessary comma operator ipsec: replace unnecessary comma operator graph: replace unnec

Re: [PATCH v2 0/6] Enable warnings about use of comma operator

On Fri, 15 Aug 2025 08:40:58 +0200 Morten Brørup wrote: > > From: Stephen Hemminger [mailto:step...@networkplumber.org] > > Sent: Thursday, 14 August 2025 23.29 > > > > This series was motivated by recent thread on netdev > > about build with -Wcomma in clang. This warning catches code > > that

Re: [PATCH] bpf: remove dependency on vla

On Fri, 15 Aug 2025 14:26:07 + Konstantin Ananyev wrote: > > The code for ethdev callbacks was using variable length arrays > > which is a feature not supported on MSVC and later C standards. > > Replace with alloca(). > > I am not a big fun of such mechanical replacement of vla with alloc

Re: [PATCH v5 2/2] net/zxdh: add support flow director ops

On Fri, 15 Aug 2025 09:33:17 +0800 fengchengwen wrote: > Hi Stephen, > > On 8/15/2025 12:59 AM, Stephen Hemminger wrote: > > On Thu, 14 Aug 2025 10:52:42 +0800 > > Bingbin Chen wrote: > > > >> + } else { > >> + entry = calloc(1, sizeof(struct rte_flow)); > >> +

Re: [PATCH v11 09/14] ethdev: add port mirroring feature

Sending a CI testing retest for this series because of a suspected false failure on the vlan testsuite. On Fri, Aug 8, 2025 at 12:59 PM Stephen Hemminger < step...@networkplumber.org> wrote: > This adds new feature port mirroring to the ethdev layer. > And standalone tests for those features. > >

Re: [PATCH v5 0/2] add support flow director ops

Sending a CI testing retest for the v5 of this series because of a suspected false failure on the packet capture testsuite. On Wed, Aug 13, 2025 at 11:03 PM Bingbin Chen wrote: > v5: > - modify some error level printing information, > and other issues from Maintainer's. > > V4: > - repla

Re: [PATCH 0/3] fix build with MinGW 13

Sending a CI testing retest for this series because of a suspected false failure on the dynamic config testsuite. On Wed, Aug 13, 2025 at 11:28 AM Thomas Monjalon wrote: > After an upgrade to MinGW version 13, some compilation errors appear > in PCI, mlx5 and bbdev code. > This series fixes them

Re: [PATCH] net/mlx5/hws: fix TIR action support in FDB

Sending a CI testing retest for this series because of a suspected false failure on the packet capture testsuite. On Thu, Aug 14, 2025 at 9:22 AM Dariusz Sosnowski wrote: > TIR action was not added as an allowed action in FDB domain. > This prevented the usage of RSS flow action in transfer flow

RE: [PATCH v3] mbuf: de-inline sanity checking a reinitialized mbuf

> Sanity checking a reinitialized mbuf (a.k.a. raw mbuf) has been refactored > to follow the same design pattern as sanity checking a normal mbuf, and > now depends on RTE_LIBRTE_MBUF_DEBUG instead of RTE_ENABLE_ASSERT. > > The details of the changes are as follows: > > Non-inlined functions rte_

RE: [PATCH] bpf: remove dependency on vla

> The code for ethdev callbacks was using variable length arrays > which is a feature not supported on MSVC and later C standards. > Replace with alloca(). I am not a big fun of such mechanical replacement of vla with alloca() Specially in that particular case, we can have internal function tha

RE: [PATCH v2 6/6] config: enable comma warnings

> Clang (3.9 or later) has a -Wcomma that emits warnings for questionable > uses of the comma operator. > > Don't enable it for drivers since many drivers still use > comma operator unnecessarily. > > Signed-off-by: Stephen Hemminger > --- > config/meson.build | 1 + > drivers/meson.build |

RE: [PATCH v2 4/6] test: replace unnecessary comma operators

> Use of comma as statement separator is discouraged and > reported as warning by clang with -Wcomma > > Signed-off-by: Stephen Hemminger > --- > app/test/test_compressdev.c | 4 ++-- > app/test/test_link_bonding_rssconf.c | 2 +- > app/test/test_thash.c| 2 +- > 3 fil

RE: [PATCH v2 3/6] graph: replace unnecessary comma operator

> Use of comma as statement separator is discouraged and > reported as warning by clang with -Wcomma > > Signed-off-by: Stephen Hemminger > --- > lib/graph/graph_stats.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/lib/graph/graph_stats.c b/lib/graph/graph_stats.c

RE: [PATCH v2 1/6] eal: replace unnecessary comma operator

> Use of comma as statement separator is discouraged and > reported as warning by clang with -Wcomma > > Signed-off-by: Stephen Hemminger > --- > lib/eal/common/rte_malloc.c | 4 ++-- > lib/eal/linux/eal_alarm.c | 6 -- > 2 files changed, 6 insertions(+), 4 deletions(-) > > diff --git

RE: [PATCH v2 2/6] ipsec: replace unnecessary comma operator

> -Original Message- > From: Stephen Hemminger > Sent: Thursday, August 14, 2025 10:29 PM > To: dev@dpdk.org > Cc: Stephen Hemminger ; Konstantin Ananyev > ; Vladimir > Medvedkin > Subject: [PATCH v2 2/6] ipsec: replace unnecessary comma operator > > Use of comma as statement separat

[PATCH v2] dts: enable port binding on the TG

Currently, driver binding for all nodes relies on a devbind_script_path attribute in the node main sessions. This attribute is being set on the SUT node, but not the TG node. Consequently, if the TG node ports are not bound to the correct driver before DTS execution, the TG ports will fail to bind

Re: Warnings in headers when building libpcap with dpdk and clang

>> Hello, >> >> I notice many warnings. To reproduce: >> >> On Debian bookworm. >> >> sudo apt install dpdk-dev >> >> git clone https://git.tcpdump.org/libpcap >> >> # build with clang >> >> $ ./autogen.sh >> >> $ ./configure CC=clang-16 --with-dpdk >> >> $ make >> >> Warning lines: >> /usr/include

[PATCH V2] net/mlx5: add support for flows targeting multicast MAC addresses

Rules for multicast MAC addresses are intended to filter multicast traffic and are managed through multicast MAC add/remove APIs. In mlx5_dev_spawn function, devices (PF, VFs, and SFs) retrieve the netdev-configured MAC addresses via netlink and store them in the PMD device data, which includes mul

[PATCH] app/testpmd: add contrack CT state inspect commands

Add command in testpmd user guide to inspect conntract CT states. The conntract possible CT states are SYN_RECV, ESTABLISHED, FIN_WAIT, CLOSE_WAIT, LAST_ACK and TIME_WAIT. Fixes: 4d07cbefe3ba0 ("app/testpmd: add commands for conntrack") Cc: sta...@dpdk.org Signed-off-by: Khadem Ullah <14pwcse1..

[PATCH] app/testpmd: add contrack state inspect commands commands

Add command in testpmd user guide to inspect conntract CT states. The conntract possible CT states are SYN_RECV, ESTABLISHED, FIN_WAIT, CLOSE_WAIT, LAST_ACK and TIME_WAIT. Fixes: 4d07cbefe3ba0 ("app/testpmd: add commands for conntrack") Cc: sta...@dpdk.org Signed-off-by: Khadem Ullah <14pwcse1..

Re: [PATCH] dts: enable port binding on the TG

Hi Andrew, thank you for the very quick turnaround! I've just noticed that this is not precisely what I had in mind, but in fairness this approach is perfectly good. On 14/08/2025 21:38, Andrew Bailey wrote: Currently, driver binding for all nodes relies on a devbind_script_path attribute in

Re: 回复：[PATCH v4 12/16] net/nbl: add nbl device rxtx queue setup and release ops

Hi Dimon, On Fri, 15 Aug 2025, Dimon wrote: Hi Ivan, Thank you for your review. Regarding your following review comments. > +int nbl_rx_queue_setup(struct rte_eth_dev *eth_dev, u16 queue_idx, > +         u16 nb_desc, unsigned int socket_id, > +         const struct rte_eth_rxconf *conf, struc

[PATCH v6 0/2] add support flow director ops

v6: - fix rte_free compilation errors. v5: - modify some error level printing information, and other issues from Maintainer's. V4: - replace rte_malloc with calloc in the flow_dev_dump function. V3: - fix several checkpatch warnings. V2: - resolve code style and gcc compilation is