RE: [PATCH] net/mlx5: zero encap UDP csum for IPv4 too

Hi, > -Original Message- > From: Bing Zhao > Sent: Monday, November 13, 2023 3:30 PM > To: Matan Azrad ; Slava Ovsiienko > ; Raslan Darawsheh ; Suanming > Mou ; Ori Kam > Cc: dev@dpdk.org; Eli Britstein ; sta...@dpdk.org > Subject: [PATCH] net/mlx5: zero encap UDP csum for IPv4 too > >

[PATCH v2] crypto/ipsec_mb: do not dequeue ops after flush

Previously it was possible to increment `processed_jobs` to a value greater than requested `nb_ops`, because after flushing at most `nb_ops` jobs the while loop continued, so `processed_jobs` could still be incremented and it was possible for this variable to be greater than `nb_ops`. If `ops` prov

RE: [PATCH v2] crypto/ipsec_mb: do not dequeue ops after flush

> -Original Message- > From: Karas, Krzysztof > Sent: Monday, November 13, 2023 8:02 AM > To: Ji, Kai ; De Lara Guarch, Pablo > ; Cornu, Marcel D > ; Power, Ciara > Cc: dev@dpdk.org; Karas, Krzysztof ; > sta...@dpdk.org > Subject: [PATCH v2] crypto/ipsec_mb: do not dequeue ops after fl

RE: [PATCH v1] doc: support IPsec Multi-buffer lib v1.5

> Hi Venkat, > > > Subject: [PATCH v1] doc: support IPsec Multi-buffer lib v1.5 > > > > Updated AESNI MB and AESNI GCM, KASUMI, ZUC, SNOW3G and > > CHACHA20_POLY1305 PMD documentation guides with information about > > the latest Intel IPsec Multi-buffer library supported. > > > > Signed-off-by: Si

RE: [EXT] Re: [PATCH v3 3/3] test/dma: add SG copy tests

Hi Fengchengwen, > > On 2023/11/3 23:38, Gowrishankar Muthukrishnan wrote: > > Add scatter-gather copy tests. > > > > Signed-off-by: Vidya Sagar Velumuri > > Signed-off-by: Gowrishankar Muthukrishnan > > --- > > app/test/test_dmadev.c | 132 +- > > app/test/test

RE: [EXT] Re: [PATCH v3 1/3] test/dma: use unit test framework

Hi Fengchengwen, Thanks for your valuable comments, I have addressed all of them in v4 (sending shortly). Please review the same. > > > > For the dmadev, which have API test, memory-to-memory test, memory-to- > device test, currently packed into two test-suites, But I think its better use > sub

Re: [PATCH v2] vhost: make experimental API's stable

On Fri, Nov 10, 2023 at 9:52 AM Maxime Coquelin wrote: > > From: Stephen Hemminger > > Transistion API's added from 2008 to 2022 to stable, except > Vhost-async's which are likely to require changes to be adopted. > > Signed-off-by: Stephen Hemminger > Signed-off-by: Maxime Coquelin I have one

Re: [PATCH v2] vhost: make experimental API's stable

On 11/13/23 10:52, David Marchand wrote: On Fri, Nov 10, 2023 at 9:52 AM Maxime Coquelin wrote: From: Stephen Hemminger Transistion API's added from 2008 to 2022 to stable, except Vhost-async's which are likely to require changes to be adopted. Signed-off-by: Stephen Hemminger Signed-of

Re: [PATCH v7 24/25] eventdev: make many API's stable

On Tue, Oct 24, 2023 at 12:42 PM Stephen Hemminger wrote: > > All API's in eventdev up to 22.11 release should be made stable. > > Signed-off-by: Stephen Hemminger Acked-by: Jerin Jacob

Re: [PATCH v7 23/25] node: remove some of the experimental tags

On Sun, Nov 12, 2023 at 8:29 PM David Marchand wrote: > > On Tue, Oct 24, 2023 at 4:57 AM Stephen Hemminger > wrote: > > > > The node library has been around long enough that experimental > > tag should be removed. The IPv6 support was added in 23.03 > > but that is still enough time that it can

RE: [PATCH] net/mlx5: fix MPRQ stride size check

> -Original Message- > From: Alexander Kozyrev > Sent: Thursday, October 12, 2023 7:35 PM > To: dev@dpdk.org > Cc: sta...@dpdk.org; Suanming Mou ; Slava > Ovsiienko > Subject: [PATCH] net/mlx5: fix MPRQ stride size check > > We should only check that MPRQ stride size is bigger than the m

[PATCH] net/mlx5: allow jump to current group

In current implementation, jump action is allowed only if target group is different than the current flow group. Change jump action validation such that it allows jump to current group. Signed-off-by: Michael Baum Acked-by: Matan Azrad --- drivers/net/mlx5/mlx5_flow_dv.c | 7 --- 1 file ch

[PATCH 00/21] replace strtok with strtok_r

Multiple threads calling the same function may cause condition race issues, which often leads to abnormal behavior and can cause more serious vulnerabilities such as abnormal termination, denial of service, and compromised data integrity. The strtok() is non-reentrant, it is better to replace it w

[PATCH 01/21] app/graph: replace strtok with strtok_r

Multiple threads calling the same function may cause condition race issues, which often leads to abnormal behavior and can cause more serious vulnerabilities such as abnormal termination, denial of service, and compromised data integrity. The strtok() is non-reentrant, it is better to replace it w

[PATCH 03/21] app/test-compress-perf: replace strtok with strtok_r

Multiple threads calling the same function may cause condition race issues, which often leads to abnormal behavior and can cause more serious vulnerabilities such as abnormal termination, denial of service, and compromised data integrity. The strtok() is non-reentrant, it is better to replace it w

[PATCH 02/21] app/test-bbdev: replace strtok with strtok_r

Multiple threads calling the same function may cause condition race issues, which often leads to abnormal behavior and can cause more serious vulnerabilities such as abnormal termination, denial of service, and compromised data integrity. The strtok() is non-reentrant, it is better to replace it w

[PATCH 04/21] app/test-crypto-perf: replace strtok with strtok_r

Multiple threads calling the same function may cause condition race issues, which often leads to abnormal behavior and can cause more serious vulnerabilities such as abnormal termination, denial of service, and compromised data integrity. The strtok() is non-reentrant, it is better to replace it w

[PATCH 05/21] app/test-dma-perf: replace strtok with strtok_r

Multiple threads calling the same function may cause condition race issues, which often leads to abnormal behavior and can cause more serious vulnerabilities such as abnormal termination, denial of service, and compromised data integrity. The strtok() is non-reentrant, it is better to replace it w

[PATCH 06/21] app/test-fib: replace strtok with strtok_r

Multiple threads calling the same function may cause condition race issues, which often leads to abnormal behavior and can cause more serious vulnerabilities such as abnormal termination, denial of service, and compromised data integrity. The strtok() is non-reentrant, it is better to replace it w

[PATCH 07/21] app/dpdk-test-flow-perf: replace strtok with strtok_r

Multiple threads calling the same function may cause condition race issues, which often leads to abnormal behavior and can cause more serious vulnerabilities such as abnormal termination, denial of service, and compromised data integrity. The strtok() is non-reentrant, it is better to replace it w

[PATCH 08/21] app/test-mldev: replace strtok with strtok_r

Multiple threads calling the same function may cause condition race issues, which often leads to abnormal behavior and can cause more serious vulnerabilities such as abnormal termination, denial of service, and compromised data integrity. The strtok() is non-reentrant, it is better to replace it w

[PATCH 09/21] lib/dmadev: replace strtok with strtok_r

Multiple threads calling the same function may cause condition race issues, which often leads to abnormal behavior and can cause more serious vulnerabilities such as abnormal termination, denial of service, and compromised data integrity. The strtok() is non-reentrant, it is better to replace it w

[PATCH 10/21] lib/eal: replace strtok with strtok_r

Multiple threads calling the same function may cause condition race issues, which often leads to abnormal behavior and can cause more serious vulnerabilities such as abnormal termination, denial of service, and compromised data integrity. The strtok() is non-reentrant, it is better to replace it w

[PATCH 11/21] lib/ethdev: replace strtok with strtok_r

Multiple threads calling the same function may cause condition race issues, which often leads to abnormal behavior and can cause more serious vulnerabilities such as abnormal termination, denial of service, and compromised data integrity. The strtok() is non-reentrant, it is better to replace it w

[PATCH 13/21] lib/telemetry: replace strtok with strtok_r

Multiple threads calling the same function may cause condition race issues, which often leads to abnormal behavior and can cause more serious vulnerabilities such as abnormal termination, denial of service, and compromised data integrity. The strtok() is non-reentrant, it is better to replace it w

[PATCH 12/21] lib/eventdev: replace strtok with strtok_r

Multiple threads calling the same function may cause condition race issues, which often leads to abnormal behavior and can cause more serious vulnerabilities such as abnormal termination, denial of service, and compromised data integrity. The strtok() is non-reentrant, it is better to replace it w

[PATCH 14/21] lib/telemetry: replace strtok with strtok_r

Multiple threads calling the same function may cause condition race issues, which often leads to abnormal behavior and can cause more serious vulnerabilities such as abnormal termination, denial of service, and compromised data integrity. The strtok() is non-reentrant, it is better to replace it w

[PATCH 16/21] common/cnxk: replace strtok with strtok_r

Multiple threads calling the same function may cause condition race issues, which often leads to abnormal behavior and can cause more serious vulnerabilities such as abnormal termination, denial of service, and compromised data integrity. The strtok() is non-reentrant, it is better to replace it w

[PATCH 15/21] bus/fslmc: replace strtok with strtok_r

Multiple threads calling the same function may cause condition race issues, which often leads to abnormal behavior and can cause more serious vulnerabilities such as abnormal termination, denial of service, and compromised data integrity. The strtok() is non-reentrant, it is better to replace it w

RE: [v2] doc: support flow matching on representor ID

> -Original Message- > From: Sean Zhang (Networking SW) > Sent: Tuesday, November 1, 2022 5:45 PM > To: NBU-Contact-Thomas Monjalon (EXTERNAL) ; > Raslan Darawsheh ; Matan Azrad > ; Slava Ovsiienko > Cc: dev@dpdk.org > Subject: [v2] doc: support flow matching on representor ID > > Add no

[PATCH 17/21] event/cnxk: replace strtok with strtok_r

Multiple threads calling the same function may cause condition race issues, which often leads to abnormal behavior and can cause more serious vulnerabilities such as abnormal termination, denial of service, and compromised data integrity. The strtok() is non-reentrant, it is better to replace it w

[PATCH 18/21] net/ark: replace strtok with strtok_r

Multiple threads calling the same function may cause condition race issues, which often leads to abnormal behavior and can cause more serious vulnerabilities such as abnormal termination, denial of service, and compromised data integrity. The strtok() is non-reentrant, it is better to replace it w

[PATCH 19/21] raw/cnxk_gpio: replace strtok with strtok_r

Multiple threads calling the same function may cause condition race issues, which often leads to abnormal behavior and can cause more serious vulnerabilities such as abnormal termination, denial of service, and compromised data integrity. The strtok() is non-reentrant, it is better to replace it w

[PATCH 20/21] examples/l2fwd-crypto: replace strtok with strtok_r

Multiple threads calling the same function may cause condition race issues, which often leads to abnormal behavior and can cause more serious vulnerabilities such as abnormal termination, denial of service, and compromised data integrity. The strtok() is non-reentrant, it is better to replace it w

[PATCH 21/21] examples/vhost: replace strtok with strtok_r

Multiple threads calling the same function may cause condition race issues, which often leads to abnormal behavior and can cause more serious vulnerabilities such as abnormal termination, denial of service, and compromised data integrity. The strtok() is non-reentrant, it is better to replace it w

[PATCH v6] app/test: secondary process passes allow parameters

In EAL related test cases, the allow parameters are not passed to the secondary process, resulting in unexpected NICs being loaded. This patch fixes this issue by appending the allow parameters to the secondary process. Fixes: af75078fece3 ("first public release") Cc: sta...@dpdk.org Signed-off-

[PATCH] net/iavf: fix error devargs parsing

This patch fixes an unknown parsing result without devarg "no-poll-on-link-down". Fixes: 5b3124a0a6ef ("net/iavf: support no polling when link down") Cc: sta...@dpdk.org Signed-off-by: Mingjin Ye --- drivers/net/iavf/iavf_ethdev.c | 7 +-- 1 file changed, 1 insertion(+), 6 deletions(-) dif

Re: [PATCH 00/21] replace strtok with strtok_r

13/11/2023 11:45, Jie Hai: > Multiple threads calling the same function may cause condition > race issues, which often leads to abnormal behavior and can cause > more serious vulnerabilities such as abnormal termination, denial > of service, and compromised data integrity. > > The strtok() is non-

Re: [PATCH v6] app/test: secondary process passes allow parameters

On Mon, Nov 13, 2023 at 10:42:22AM +, Mingjin Ye wrote: > In EAL related test cases, the allow parameters are not passed to > the secondary process, resulting in unexpected NICs being loaded. > > This patch fixes this issue by appending the allow parameters to > the secondary process. > > Fix

RE: [PATCH v2] crypto/ipsec_mb: do not dequeue ops after flush

> -Original Message- > From: Karas, Krzysztof > Sent: Monday, November 13, 2023 8:02 AM > To: Ji, Kai ; De Lara Guarch, Pablo > ; Cornu, Marcel D > ; Power, Ciara > Cc: dev@dpdk.org; Karas, Krzysztof ; > sta...@dpdk.org > Subject: [PATCH v2] crypto/ipsec_mb: do not dequeue ops after flu

Re: [PATCH 00/21] replace strtok with strtok_r

Hi Jie, Good fix There are two minor I think need to modify: 1. The [PATCH 13/21] lib/telemetry should be lib/security 2. All commits should add Cc because it's potential bug. The other LGTM, with above fixed Series-acked-by: Chengwen Feng Thanks Chengwen On 2023/11/13 18:45, Jie Hai wrote: >

RE: [PATCH v2] crypto/ipsec_mb: do not dequeue ops after flush

> -Original Message- > From: Cornu, Marcel D > Sent: Monday, November 13, 2023 11:13 AM > To: Karas, Krzysztof ; Ji, Kai ; > De > Lara Guarch, Pablo ; Power, Ciara > > Cc: dev@dpdk.org; sta...@dpdk.org > Subject: RE: [PATCH v2] crypto/ipsec_mb: do not dequeue ops after flush > > > >

Re:Re: [PATCH v3 1/2] examples/l3fwd: relax RSS requirement with option

At 2023-11-13 07:33:49, "Stephen Hemminger" wrote: >On Sun, 12 Nov 2023 07:16:10 + >Trevor Tao wrote: > >> Now the port Rx mq_mode had been set to RTE_ETH_MQ_RX_RSS >> by default, but some hw and/or virtual interface does not >> support the RSS and offload mode presupposed, e.g., some >> virt

Re: BUILD bug hidden in SFC driver.

Hi Stephen, On Sat, 11 Nov 2023, Stephen Hemminger wrote: While examining the use of VLA in DPDK, ran into a bug in sfc driver. If DPDK is built with -Wvla, then the RTE_BUILD_BUG_ON() macro won't work as written. Experimenting with a better more portable version of that macro as: #def

RE: Re: [PATCH v3 1/2] examples/l3fwd: relax RSS requirement with option

From: taozj888 [mailto:taozj...@163.com] Sent: Monday, 13 November 2023 12.47 At 2023-11-13 07:33:49, "Stephen Hemminger" wrote: >On Sun, 12 Nov 2023 07:16:10 + >Trevor Tao wrote: > >> Now the port Rx mq_mode had been set to RTE_ETH_MQ_RX_RSS >> by default, but some hw and/or virtual in

[PATCH] net/mlx5: fix the LACP redirection in Rx domain

When the "lacp_by_user" is not set from the application in bond mode, the LACP traffic should be handled by the kernel driver by default. This commit adds the missing support in the template API when "dv_flow_en=2". The behavior will be the same as that in the DV mode with "dv_flow_en=1". The LACP

[PATCH v4 0/3] test/dma: add vchan reconfig and SG tests

This patch series reworks DMA tests to follow unit test framework followed by new vchan reconfig and SG tests. v4: - Suggestions from Fengchenwen addressed. Gowrishankar Muthukrishnan (3): test/dma: use unit test framework test/dma: test multiple vchan test/dma: add SG copy tests app/tes

[PATCH v4 1/3] test/dma: use unit test framework

Use unit test framework to execute DMA tests. Signed-off-by: Gowrishankar Muthukrishnan --- app/test/test_dmadev.c | 235 + app/test/test_dmadev_api.c | 95 +-- 2 files changed, 193 insertions(+), 137 deletions(-) diff --git a/app/test/test_d

[PATCH v4 2/3] test/dma: test multiple vchan

Support API with multiple vchan test. Signed-off-by: Gowrishankar Muthukrishnan --- app/test/test_dmadev_api.c | 63 +- 1 file changed, 55 insertions(+), 8 deletions(-) diff --git a/app/test/test_dmadev_api.c b/app/test/test_dmadev_api.c index 73d4db825a..2b8

[PATCH v4 3/3] test/dma: add SG copy tests

Add scatter-gather copy tests. Signed-off-by: Vidya Sagar Velumuri Signed-off-by: Gowrishankar Muthukrishnan --- app/test/test_dmadev.c | 139 ++- app/test/test_dmadev_api.c | 164 ++--- 2 files changed, 289 insertions(+), 14 delet

[PATCH v2] crypto/openssl: fix asym memory leaks

Numerous memory leaks were detected by ASAN in the OpenSSL PMD asymmetric code path. These are now fixed to free all variables allocated by OpenSSL functions such as BN_bin2bn and OSSL_PARAM_BLD_new. Some need to exist until the op is processed, for example the BIGNUMs associated with DSA. The po

Re: [PATCH v1] doc: support IPsec Multi-buffer lib v1.5

13/11/2023 10:13, Akhil Goyal: > > Hi Venkat, > > > > > Subject: [PATCH v1] doc: support IPsec Multi-buffer lib v1.5 > > > > > > Updated AESNI MB and AESNI GCM, KASUMI, ZUC, SNOW3G and > > > CHACHA20_POLY1305 PMD documentation guides with information about > > > the latest Intel IPsec Multi-buffer

RE: [PATCH] net/mlx5: fix the LACP redirection in Rx domain

Hi, One small comment, v2 is needed. > -Original Message- > From: Bing Zhao > Sent: Monday, November 13, 2023 8:42 PM > To: Matan Azrad ; Slava Ovsiienko > ; Raslan Darawsheh ; Suanming > Mou ; Ori Kam > Cc: dev@dpdk.org; Dariusz Sosnowski ; > sta...@dpdk.org > Subject: [PATCH] net/mlx5

[PATCH v3] crypto/openssl: fix asym memory leaks

Numerous memory leaks were detected by ASAN in the OpenSSL PMD asymmetric code path. These are now fixed to free all variables allocated by OpenSSL functions such as BN_bin2bn and OSSL_PARAM_BLD_new. Some need to exist until the op is processed, for example the BIGNUMs associated with DSA. The po

RE: Re: [PATCH v3 1/2] examples/l3fwd: relax RSS requirement with option

>> Now the port Rx mq_mode had been set to RTE_ETH_MQ_RX_RSS >> by default, but some hw and/or virtual interface does not >> support the RSS and offload mode presupposed, e.g., some >> virtio interfaces in the cloud don't support >> RSS and the error msg may like: >> >> virtio_dev_configure():

Re: [virtio] Invitation to comment on Virtual I/O Device (VIRTIO) Version 1.3 - ends December 8th

It seems that the original mail did not make it to some of the mailing lists it was intended to go to, possibly due to the html part of the original mail. Therefore, I'm trying again, this time with plain text only. Apologies in advance for any duplicates. On Wed, Nov 08 2023, Paul Knight wrote:

RE: [PATCH] net/mlx5: fix the LACP redirection in Rx domain

Hi Suanming, My answers are inline. > -Original Message- > From: Suanming Mou > Sent: Monday, November 13, 2023 9:14 PM > To: Bing Zhao ; Matan Azrad ; Slava > Ovsiienko ; Raslan Darawsheh > ; Ori Kam > Cc: dev@dpdk.org; Dariusz Sosnowski ; > sta...@dpdk.org > Subject: RE: [PATCH] net/m

RE: [PATCH] net/mlx5: fix the LACP redirection in Rx domain

Hi, > -Original Message- > From: Bing Zhao > Sent: Monday, November 13, 2023 9:51 PM > To: Suanming Mou ; Matan Azrad > ; Slava Ovsiienko ; Raslan > Darawsheh ; Ori Kam > Cc: dev@dpdk.org; Dariusz Sosnowski ; > sta...@dpdk.org > Subject: RE: [PATCH] net/mlx5: fix the LACP redirection in

[PATCH] maintainers: Adding new maintainer to Nvidia mlx5

Adding Dariusz Sosnowski as NVIDIA mlx5 maintainer. Signed-off-by: Maayan Kashani --- MAINTAINERS | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/MAINTAINERS b/MAINTAINERS index 80e071d13e..c9fccc97b6 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -867,10 +867,11 @@ F: doc/g

[PATCH] app/test: don't count skipped tests as executed

The logic around skipped tests is a little confusing in the unit test runner. * Any explicitly disabled tests are counted as skipped but not executed. * Any tests that return TEST_SKIPPED are counted as both skipped and executed, using the same statistics counters. This makes the stats very st

Re: [PATCH v1 1/3] net/axgbe: packet size doesn't exceed the configured MTU

On 11/11/2023 4:00 PM, Jesna K E wrote: > Signed-off-by: Jesna K E > Hi Jesna, Description is missing, making it hard to understand the problem and motivation of the change. > --- > drivers/net/axgbe/axgbe_ethdev.c | 6 -- > drivers/net/axgbe/axgbe_rxtx.c | 20 ++-- > 2

[Bug 1313] Build errors with clang 17.0.3 and ASan

https://bugs.dpdk.org/show_bug.cgi?id=1313 Bug ID: 1313 Summary: Build errors with clang 17.0.3 and ASan Product: DPDK Version: 23.11 Hardware: All OS: All Status: UNCONFIRMED Severity: normal Pr

RE: [PATCH] maintainers: Adding new maintainer to Nvidia mlx5

> -Original Message- > From: Maayan Kashani > Sent: Monday, November 13, 2023 4:55 PM > To: dev@dpdk.org > Cc: Maayan Kashani ; Matan Azrad > ; Raslan Darawsheh ; NBU- > Contact-Thomas Monjalon (EXTERNAL) > Subject: [PATCH] maintainers: Adding new maintainer to Nvidia mlx5 > > Adding Dar

Re: [PATCH v1 2/3] net/axgbe: correct API call when offload enabled

On 11/11/2023 4:00 PM, Jesna K E wrote: > Fixes: 9963b5131af8 ("net/axgbe: support multi-process") > Can you please add more description? > Signed-off-by: Jesna K E > --- > drivers/net/axgbe/axgbe_ethdev.c | 10 ++ > 1 file changed, 6 insertions(+), 4 deletions(-) > > diff --git a/dr

Re: [PATCH] graph: mark API's as stable

On Tue, Aug 15, 2023 at 3:46 AM Yan, Zhirun wrote: > > On Thu, Aug 10, 2023 at 11:35 PM Stephen Hemminger > > wrote: > > > > > > The graph library has been marked experimental since initial release > > > in 2020. Time to take the training wheels off. > > > > > > Signed-off-by: Stephen Hemminger

Re: [PATCH v7 00/25] Promote 22.11 and older API's to stable

On Tue, Oct 24, 2023 at 4:54 AM Stephen Hemminger wrote: > > Since 23.11 is an LTS release it is time to remove the experimental > bandaid off many API's. There are about 890 API's marked with experimental > on current main branch. This addresses the easy to remove ones and > gets it down to about

RE: [PATCH v7 1/2] bus/pci: support PASID control

>+Ok, that is indeed an issue. >+I found some patches exposing this capability with vfio-pci but I am not sure >what is the latest work on the topic. That's right, some kernels may expose the capability some not. We realized it when Bruce reported my earlier patch with the api rte_pci_find_ext_

Re: [PATCH v1 1/1] ml/cnxk: fix fast-path synchronization

On Mon, Nov 13, 2023 at 11:01 AM Srikanth Yalavarthi wrote: > > Segfaults are reported with TVM/LLVM models as NULL value > is read by dequeue thread for the op handle. This is due > synchronization issue between enqueue and dequeue threads. > > This fix ensures the op handle written to internal r

[PATCH v4 0/2] example/l3fwd: relax l3fwd rx RSS/Offload if needed

This series try to relax l3fwd rx RSS/Offload mode requirement if they are not supported by underlying hw or virtual devices, there is an option named relax_rx_mode added to enable this option. Trevor Tao (2): examples/l3fwd: relax RSS requirement with option examples/l3fwd: relax RX Offload w

[PATCH v4 1/2] examples/l3fwd: relax RSS requirement with option

Now the port Rx mq_mode had been set to RTE_ETH_MQ_RX_RSS by default, but some hw and/or virtual interface does not support the RSS and offload mode presupposed, e.g., some virtio interfaces in the cloud don't support RSS and the error msg may like: virtio_dev_configure(): RSS support requested bu

[PATCH v4 2/2] examples/l3fwd: relax RX Offload with option

Now the port Rx offload mode is set to RTE_ETH_RX_OFFLOAD_CHECKSUM by default, but some hw and/or virtual interface does not support the offload mode presupposed, e.g., some virtio interfaces in the cloud may only partly support RTE_ETH_RX_OFFLOAD_UDP_CKSUM/ RTE_ETH_RX_OFFLOAD_TCP_CKSUM, but not RT

Re:RE: Re: [PATCH v3 1/2] examples/l3fwd: relax RSS requirement with option

Hi all: At 2023-11-13 21:19:14, "Konstantin Ananyev" wrote: > > >>> Now the port Rx mq_mode had been set to RTE_ETH_MQ_RX_RSS >>> by default, but some hw and/or virtual interface does not >>> support the RSS and offload mode presupposed, e.g., some >>> virtio interfaces in the cloud don't suppor

[PATCH v5 0/2] example/l3fwd: relax l3fwd rx RSS/Offload if needed

This series try to relax l3fwd rx RSS/Offload mode requirement if they are not supported by underlying hw or virtual devices, there is an option named relax_rx_mode added to enable this option. Trevor Tao (2): examples/l3fwd: relax RSS requirement with option examples/l3fwd: relax RX Offload w

[PATCH v5 2/2] examples/l3fwd: relax RX Offload with option

Now the port Rx offload mode is set to RTE_ETH_RX_OFFLOAD_CHECKSUM by default, but some hw and/or virtual interface does not support the offload mode presupposed, e.g., some virtio interfaces in the cloud may only partly support RTE_ETH_RX_OFFLOAD_UDP_CKSUM/ RTE_ETH_RX_OFFLOAD_TCP_CKSUM, but not RT

[PATCH v5 1/2] examples/l3fwd: relax RSS requirement with option

Now the port Rx mq_mode had been set to RTE_ETH_MQ_RX_RSS by default, but some hw and/or virtual interface does not support the RSS and offload mode presupposed, e.g., some virtio interfaces in the cloud don't support RSS and the error msg may like: virtio_dev_configure(): RSS support requested bu

[PATCH v6 0/5] dumpcap and pcapng fixes

It fixes issues related to timestamping. The design choices are to maximize performance in the primary process; and do all the time adjustment in the secondary (dumpcap) since the dumpcap needs to system calls anyway to write the result. This patches set changes where the adjustment is calculated

[PATCH v6 1/5] pdump: fix setting rte_errno on mp error

The response from MP server sets err_value to negative on error. The convention for rte_errno is to use a positive value on error. This makes errors like duplicate registration show up with the correct error value. Fixes: 660098d61f57 ("pdump: use generic multi-process channel") Signed-off-by: Ste

[PATCH v6 2/5] dumpcap: allow multiple invocations

If dumpcap is run twice with each instance pointing a different interface, it would fail because of overlap in ring a pool names. Fix by putting process id in the name. It is still not allowed to do multiple invocations on the same interface because only one callback is allowed and only one copy o

[PATCH v6 3/5] pcapng: modify timestamp calculation

The computation of timestamp is best done in the part of pcapng library that is in secondary process. The secondary process is already doing a bunch of system calls which makes it not performance sensitive. This does change the rte_pcapng_copy() and rte_pcapng_write_stats() experimental API's. Sim

[PATCH v6 4/5] pcapng: avoid using alloca()

The function alloca() like VLA's has problems if the caller passes a large value. Instead use a fixed size buffer (2K) which will be more than sufficient for the info related blocks in the file. Add bounds checks as well. Signed-off-by: Stephen Hemminger Acked-by: Morten Brørup --- lib/pcapng/r

[PATCH v6 5/5] test: cleanups to pcapng test

Overhaul of the pcapng test: - promote it to be a fast test so it gets regularly run. - create null device and use i. - use UDP discard packets that are valid so that for debugging the resulting pcapng file can be looked at with wireshark. - do basic checks on resulting pcap file that l

[PATCH v1] crypto/qat: skip SM3 HMAC on QAT when no SM3 slice exists

Add check to skip SM3 HMAC on QAT when no SM3 slice exists Fixes: 171c655 ("crypto/qat: add SM3 HMAC to GEN4 devices") Cc: arkadiuszx.kusz...@intel.com Cc: sta...@dpdk.org Signed-off-by: Sivaramakrishnan Venkat --- drivers/crypto/qat/dev/qat_crypto_pmd_gen3.c | 4 +++- 1 file changed, 3 inserti

Re: BUILD bug hidden in SFC driver.

On Sat, Nov 11, 2023 at 08:56:34AM -0800, Stephen Hemminger wrote: > While examining the use of VLA in DPDK, ran into a bug in sfc driver. > > If DPDK is built with -Wvla, then the RTE_BUILD_BUG_ON() macro won't work > as written. Experimenting with a better more portable version of that macro > a

Re: [RFC] eal: use _Static_assert() for RTE_BUILD_BUG_ON

On Sat, Nov 11, 2023 at 09:21:53AM -0800, Stephen Hemminger wrote: > The method of doing sizeof a bad array element was an interesting > hack but it has a couple of problems. First, it won't work if > VLA checking is enabled. It doesn't enforce that the expression > is constant. > > Replace that w

Re: [RFC] eal: use _Static_assert() for RTE_BUILD_BUG_ON

On Sat, Nov 11, 2023 at 06:52:26PM +0100, Morten Brørup wrote: > > From: Stephen Hemminger [mailto:step...@networkplumber.org] > > Sent: Saturday, 11 November 2023 18.22 > > > > The method of doing sizeof a bad array element was an interesting > > hack but it has a couple of problems. First, it wo

Re: BUILD bug hidden in SFC driver.

On Mon, 13 Nov 2023 16:04:51 +0400 (+04) Ivan Malov wrote: > diff --git a/drivers/net/sfc/sfc_ef100_tx.c b/drivers/net/sfc/sfc_ef100_tx.c > index 1b6374775f..01f37c2616 100644 > --- a/drivers/net/sfc/sfc_ef100_tx.c > +++ b/drivers/net/sfc/sfc_ef100_tx.c > @@ -563,7 +563,7 @@ sfc_ef100_tx_pkt_desc

Re: [PATCH v4 1/2] examples/l3fwd: relax RSS requirement with option

On Mon, 13 Nov 2023 16:02:54 + Trevor Tao wrote: > Now the port Rx mq_mode had been set to RTE_ETH_MQ_RX_RSS > by default, but some hw and/or virtual interface does not > support the RSS and offload mode presupposed, e.g., some > virtio interfaces in the cloud don't support > RSS and the erro

Re: [PATCH v1 2/3] net/axgbe: correct API call when offload enabled

On 11/13/2023 3:23 PM, Ferruh Yigit wrote: > On 11/11/2023 4:00 PM, Jesna K E wrote: >> Fixes: 9963b5131af8 ("net/axgbe: support multi-process") >> > > Can you please add more description? > > >> Signed-off-by: Jesna K E >> --- >> drivers/net/axgbe/axgbe_ethdev.c | 10 ++ >> 1 file cha

[PATCH v2 1/3] event/opdl: fix non-constant compile time assertion

RTE_BUILD_BUG_ON() was being used with a non-constant value. The inline function rte_is_power_of_2() is not constant since inline expansion happens later in the compile process. Replace it with macro which will be constant. Fixes: 4236ce9bf5bf ("event/opdl: add OPDL ring infrastructure library") C

[PATCH v2 0/3] use static_assertion for build errors

This series fixes a couple places where expressions that could not be evaluated as constant early in compiler passes were used. And then converts RTE_BUILD_BUG_ON() with static_assert. Stephen Hemminger (3): event/opdl: fix non-constant compile time assertion net/sfc: fix non-constant expressi

[PATCH v2 2/3] net/sfc: fix non-constant expression inr RTE_BUILD_BUG_ON()

The macro RTE_MIN has some hidden assignments to provide type safety which means the statement can not be fully evaluted in first pass of compiler. Replace RTE_MIN() with equivalent macro. This will cause errors from checkpatch about multiple evaluations of same expression in macro but it is ok in

[PATCH v2 3/3] eal: replace out of bounds VLA with static_assert

Both Gcc, clang and MSVC have better way to do compile time assertions rather than using out of bounds array access. The old method would fail if -Wvla is enabled because compiler can't determine size in that code. Also, the use of new _Static_assert will catch broken code that is passing non-cons

Re: [PATCH 00/21] replace strtok with strtok_r

On Mon, Nov 13, 2023 at 06:45:29PM +0800, Jie Hai wrote: > Multiple threads calling the same function may cause condition > race issues, which often leads to abnormal behavior and can cause > more serious vulnerabilities such as abnormal termination, denial > of service, and compromised data integr

Re: [PATCH v2 3/3] eal: replace out of bounds VLA with static_assert

On Mon, Nov 13, 2023 at 09:06:05AM -0800, Stephen Hemminger wrote: > Both Gcc, clang and MSVC have better way to do compile time > assertions rather than using out of bounds array access. > The old method would fail if -Wvla is enabled because compiler > can't determine size in that code. Also, th

Re: [PATCH v2 1/3] event/opdl: fix non-constant compile time assertion

On Mon, Nov 13, 2023 at 09:06:03AM -0800, Stephen Hemminger wrote: > RTE_BUILD_BUG_ON() was being used with a non-constant value. > The inline function rte_is_power_of_2() is not constant since > inline expansion happens later in the compile process. > Replace it with macro which will be constant.

[PATCH] net/mlx5: remove non constant size from type cast

Placing a non-constant size in the subscript [size] of a type cast is causing unnecessary generation of a VLA. Remove size and just provide [] indicating the type is an array of unspecified size. Signed-off-by: Tyler Retzlaff --- drivers/net/mlx5/mlx5_rxq.c | 5 ++--- drivers/net/mlx5/mlx5_txq.c

[PATCH v1] bus/pci: revise support PASID control

This commit revises PASID control function to accept PASID offset to pasid *structure* instead of taking exact register for controlling the feature. PASID control function was introduced in earlier commit. Pls see commit 5a6878335b81 ("event/dlb2: disable PASID") and commit 60ea19609aec ("bus/pci:

RE: [PATCH v7 1/2] bus/pci: support PASID control

I have pushed another patch addressing David's comments. https://patches.dpdk.org/project/dpdk/patch/20231113172759.3529518-1-abdullah.sevin...@intel.com/

RE: [PATCH v1] crypto/qat: skip SM3 HMAC on QAT when no SM3 slice exists

> -Original Message- > From: Sivaramakrishnan Venkat > Sent: Monday, November 13, 2023 4:23 PM > To: Ji, Kai > Cc: dev@dpdk.org; sta...@dpdk.org; gak...@marvell.com; Sivaramakrishnan, > VenkatX ; Kusztal, ArkadiuszX > > Subject: [PATCH v1] crypto/qat: skip SM3 HMAC on QAT when no SM3

Re: [PATCH v4 1/2] examples/l3fwd: relax RSS requirement with option

On Mon, 13 Nov 2023 08:34:47 -0800 Stephen Hemminger wrote: > On Mon, 13 Nov 2023 16:02:54 + > Trevor Tao wrote: > > > Now the port Rx mq_mode had been set to RTE_ETH_MQ_RX_RSS > > by default, but some hw and/or virtual interface does not > > support the RSS and offload mode presupposed, e.

  1   2   >