[PATCH 2/2] devtools/dump-cpu-flags: add tool to update CPU flags table

This patchset allows users to specify the CPU for which the generated code should be optimized for by passing the CPU name. MSVC does not provide this functionality natively, so logic was added. This additional logic relies on a table which stores instruction set availability (like AXV512F) for di

[PATCH 2/6] net/netvsc: introduce get_vmbus_device to get the vmbus device

From: Long Li Introduce a function get the vmbus device from hn_data. For secondary process, the vmbus device is in eth_dev's private region. Signed-off-by: Long Li --- drivers/net/netvsc/hn_nvs.c | 15 +++ drivers/net/netvsc/hn_nvs.h | 2 ++ 2 files changed, 17 insertions(+) dif

[PATCH 1/2] config: allow AVX512 instructions to be used with MSVC

Up to now MSVC has being used with the default mode, which uses SSE2 instructions for scalar floating-point and vector calculations. https://learn.microsoft.com/en-us/cpp/build/reference/arch-x64?view=msvc-170 This patch allows users to specify the CPU for which the generated code should be optimi

[PATCH 0/2] allow AVX512 instructions to be used with MSVC

Up to now MSVC has being used with the default mode, which uses SSE2 instructions for scalar floating-point and vector calculations. https://learn.microsoft.com/en-us/cpp/build/reference/arch-x64?view=msvc-170 This patchset allows users to specify the CPU for which the generated code should be opt

[PATCH 0/6] Support VMBUS channels without monitoring enabled

From: Long Li Hyperv may expose VMBUS channels without monitoring enabled. In this case, it programs almost all the data traffic to VF. This patchset enabled vmbus/netvsc to use channels without monitoring enabled. Long Li (6): net/netvsc: introduce private data for storing vmbus device for

[PATCH 4/6] bus/vmbus: support channels without monitoring enabled

From: Long Li Hyperv host may offer channels without monitor enabled. The max monitor ID it supports is 128. Over those channels without monitor enabled, Hyperv does not send or receive large amount of data traffic and almost all the data traffic is going over the VF. Change the code to not fail

[PATCH 5/6] bus/vmbus: add rte_vmbus_device to all functions accessing vmbus

From: Long Li The secondary process can access its vmbus device through device private region. Add and pass it on all call chains leading to vmbus code. Signed-off-by: Long Li --- drivers/bus/vmbus/linux/vmbus_uio.c | 2 +- drivers/bus/vmbus/private.h | 2 +- drivers/bus/vmbus/rte_bu

[PATCH 6/6] bus/vmbus: set event for channel without monitoring support

From: Long Li For vmbus channels without monitoring support, use kernel UIO interface to indicate packet through interrupt page and UIO file handle. Signed-off-by: Long Li --- drivers/bus/vmbus/vmbus_channel.c | 21 ++--- 1 file changed, 18 insertions(+), 3 deletions(-) diff -

[PATCH 1/6] net/netvsc: introduce private data for storing vmbus device for secondary process

From: Long Li To prepare for supporting to set hyperv event from secondary process when the channel has monitoring disable, introduce a private data region for storing the vmbus device. The secondary process will get access to its vmbus device in case it needs to signal the host. Signed-off-by:

[PATCH 3/6] bus/vmbus: store UIO fd for secondary process

From: Long Li Secondary process will get access to vmbus device and this UIO fd for signaling hyperv host on channels without monitoring support. Signed-off-by: Long Li --- drivers/bus/vmbus/vmbus_common_uio.c | 9 ++--- 1 file changed, 6 insertions(+), 3 deletions(-) diff --git a/drivers

Re: [PATCH] app/test: fix DMA API tests in IOVA as PA mode

Reviewed-by: Chengwen Feng On 2025/2/27 21:27, Bruce Richardson wrote: > When running without IOMMU for address translation, i.e. IOVAs are > physical rather than virtual addresses, we need to translate the > pointers to IOVAs for the completion API tests. > > Fixes: 9942ebb9c698 ("test/dma: add

MPLS Push action

Hi, When MPLS push action is used, using testpmd action of_push_mpls, it accepts the ethertype only. How do I specify the MPLS label and other fields for the MPLS header to be added ? In the doc , in section 7.2.3

Re: release candidate 25.03-rc1

IBM - Power Systems DPDK v25.03-rc1 * Build CI on Fedora 38,39,40 and 41 container images for ppc64le * Basic PF on Mellanox: No issue found * Performance: not tested. * OS: - RHEL 9.4 kernel: 5.14.0-427.40.1.el9_4.ppc64le with gcc version 11.4.1 20231218 (Red Hat 11.4.1-3) (GCC)

[PATCH] doc: update ionic driver guides

Update broken links and add a link for the DSC3-400 product brief. Cc: sta...@dpdk.org Signed-off-by: Andrew Boyer --- doc/guides/cryptodevs/ionic.rst | 14 +- doc/guides/nics/ionic.rst | 18 +- 2 files changed, 22 insertions(+), 10 deletions(-) diff --git a/do

[v2] app/crypto-perf: add RSA support

From: Akhil Goyal Add RSA support in crypto-perf application. Signed-off-by: Akhil Goyal --- v2: - fixed RSA padding type. --- app/test-crypto-perf/cperf_ops.c | 69 app/test-crypto-perf/cperf_options.h | 4 + app/test-crypto-perf/cperf_options_parsing.c | 38

[v8 3/5] examples/vhost_crypto: fix user callbacks

In order to handle new vhost user connection, use new_connection and destroy_connection callbacks. Fixes: f5188211c721 ("examples/vhost_crypto: add sample application") Cc: sta...@dpdk.org Signed-off-by: Gowrishankar Muthukrishnan Acked-by: Akhil Goyal --- examples/vhost_crypto/main.c | 4 ++--

[v8 2/5] vhost: update vhost_user crypto session parameters

As per requirements on vhost_user spec, session id should be located at the end of session parameter. Update VhostUserCryptoSessionParam structure to support newer QEMU versions (v9). Due to additional parameters added in QEMU, received payload from QEMU would be larger than existing payload. Henc

[v8 4/5] vhost: support asymmetric RSA crypto ops

Support asymmetric RSA crypto operations in vhost-user. Signed-off-by: Gowrishankar Muthukrishnan Acked-by: Akhil Goyal --- doc/guides/rel_notes/release_25_03.rst | 3 + lib/vhost/vhost_crypto.c | 486 +++-- lib/vhost/virtio_crypto.h | 67

[v8 5/5] examples/vhost_crypto: support asymmetric crypto

Support asymmetric crypto operations. Signed-off-by: Gowrishankar Muthukrishnan Acked-by: Akhil Goyal --- doc/guides/sample_app_ug/vhost_crypto.rst | 5 +++ examples/vhost_crypto/main.c | 50 +-- 2 files changed, 43 insertions(+), 12 deletions(-) diff --git a/

[v8 1/5] vhost: skip crypto op fetch before vring init

Until virtio avail ring is initialized (by VHOST_USER_SET_VRING_ADDR), worker thread should not try to fetch crypto op, which would lead to memory fault. Fixes: 939066d96563 ("vhost/crypto: add public function implementation") Cc: sta...@dpdk.org Signed-off-by: Gowrishankar Muthukrishnan Acked-b

[v8 0/5] vhost: add RSA support

This patch series supports asymmetric RSA in vhost crypto library. It also includes changes to improve vhost crypto library: * support newer QEMU versions. * fix broken vhost_crypto example application. * stabilize crypto fastpath operations. Gowrishankar Muthukrishnan (5): vhost: skip crypto

RE: [EXTERNAL] Re: [v6 1/5] vhost: skip crypto op fetch before vring init

> > > > Ha, and also you should be able to remove: > > __rte_no_thread_safety_analysis /* FIXME: requires iotlb_lock? */ in > > vhost_crypto_process_one_req() once implemented. > > > Removing it would break compilation for thread safety flag. http://mails.dpdk.org/archives/test-report/2025-Februar

[v7 5/5] examples/vhost_crypto: support asymmetric crypto

Support asymmetric crypto operations. Signed-off-by: Gowrishankar Muthukrishnan Acked-by: Akhil Goyal --- doc/guides/sample_app_ug/vhost_crypto.rst | 5 +++ examples/vhost_crypto/main.c | 50 +-- 2 files changed, 43 insertions(+), 12 deletions(-) diff --git a/

[v7 4/5] vhost: support asymmetric RSA crypto ops

Support asymmetric RSA crypto operations in vhost-user. Signed-off-by: Gowrishankar Muthukrishnan Acked-by: Akhil Goyal --- doc/guides/rel_notes/release_25_03.rst | 3 + lib/vhost/vhost_crypto.c | 486 +++-- lib/vhost/virtio_crypto.h | 67

[v7 3/5] examples/vhost_crypto: fix user callbacks

In order to handle new vhost user connection, use new_connection and destroy_connection callbacks. Fixes: f5188211c721 ("examples/vhost_crypto: add sample application") Cc: sta...@dpdk.org Signed-off-by: Gowrishankar Muthukrishnan Acked-by: Akhil Goyal --- examples/vhost_crypto/main.c | 4 ++--

[v7 2/5] vhost: update vhost_user crypto session parameters

As per requirements on vhost_user spec, session id should be located at the end of session parameter. Update VhostUserCryptoSessionParam structure to support newer QEMU versions (v9). Due to additional parameters added in QEMU, received payload from QEMU would be larger than existing payload. Henc

[v7 1/5] vhost: skip crypto op fetch before vring init

Until virtio avail ring is initialized (by VHOST_USER_SET_VRING_ADDR), worker thread should not try to fetch crypto op, which would lead to memory fault. Fixes: 939066d96563 ("vhost/crypto: add public function implementation") Cc: sta...@dpdk.org Signed-off-by: Gowrishankar Muthukrishnan Acked-b

[v7 0/5] vhost: add RSA support

This patch series supports asymmetric RSA in vhost crypto library. It also includes changes to improve vhost crypto library: * support newer QEMU versions. * fix broken vhost_crypto example application. * stabilize crypto fastpath operations. Gowrishankar Muthukrishnan (5): vhost: skip crypto

[PATCH] app/test: fix DMA API tests in IOVA as PA mode

When running without IOMMU for address translation, i.e. IOVAs are physical rather than virtual addresses, we need to translate the pointers to IOVAs for the completion API tests. Fixes: 9942ebb9c698 ("test/dma: add dmadev API test") Cc: fengcheng...@huawei.com Cc: sta...@dpdk.org Signed-off-by:

RE: [EXTERNAL] Re: [v6 1/5] vhost: skip crypto op fetch before vring init

Hi Maxime, > > > > You should only unlock at the end of the function, otherwise there is > > not much protection. > > Ha, and also you should be able to remove: > __rte_no_thread_safety_analysis /* FIXME: requires iotlb_lock? */ in > vhost_crypto_process_one_req() once implemented. > Ack. Thank

[PATCH] net/mlx5: fix GTP flags matching

Support gtp flags in non template on top of HWS. Currently, only extension flag was supported, Added support to all bits under v_pt_rsv_flags. Fixes: c55c2bf35333 ("net/mlx5/hws: add definer layer") Cc: sta...@dpdk.org Signed-off-by: Maayan Kashani Acked-by: Dariusz Sosnowski --- drivers/net/m

[PATCH 2/2] net/mlx5: fix error info in actions construct

In some cases in debug it misses the error info. Fix to update the error structure. Fixes: 654ebd8cb7a3 ("net/mlx5: support flow table resizing") Cc: sta...@dpdk.org Signed-off-by: Maayan Kashani Acked-by: Dariusz Sosnowski --- drivers/net/mlx5/mlx5_flow_hw.c | 7 --- 1 file changed, 4 ins

[PATCH 1/2] net/mlx5: fix non template set VLAN VID

Support set vlan vid in non template on top of HWS. Update relevant return errors in the relevant functions to avoid crash. Mask the vlan vid action in non template mode such that the action template create will use the vid value. Fixes: 00a0a6b80674 ("net/mlx5: support indirect actions in non-tem

[PATCH] net/mlx5: fix crash when using represented port w/o port ID

For non template API on top of HWS, when trying to use represented-port item w/o setting the ethdev_port_id, it crashes. Added default values to match the case for SWS. Default port is now eswitch manager id. Fixes: c55c2bf35333 ("net/mlx5/hws: add definer layer") Cc: sta...@dpdk.org Signed-off-

[PATCH] doc: add limitation under unified FDB

Add limitation for unified FDB group 0 table creation. Signed-off-by: Maayan Kashani Acked-by: Bing Zhao --- doc/guides/nics/mlx5.rst | 6 ++ 1 file changed, 6 insertions(+) diff --git a/doc/guides/nics/mlx5.rst b/doc/guides/nics/mlx5.rst index 190c58b54d6..efa4c708d51 100644 --- a/doc/gui

[PATCH] net/mlx5: fix LACP packets handling in isolated mode

In HWS(dv_flow_en=2) mode, in the routine that enables traffic, isolated mode check was before LACP default rules creation as opposed to legacy mode handling. As a result, the LACP default rules were not created in HWS mode and LACP packets arrived to testpmd. The fix was to create the LACP defau

[PATCH] net/mlx5: fix assert failure on hairpin queue release

Assert was triggered because of ctrl_ref mismatch on hairpin queue. Fixed the mismatch. Fixes: 09c2555 ("net/mlx5: support shared Rx queue") Cc: sta...@dpdk.org Signed-off-by: Maayan Kashani Acked-by: Dariusz Sosnowski --- drivers/net/mlx5/mlx5.h | 1 + drivers/net/mlx5/mlx5_flow.c | 4

RE: [EXTERNAL] Re: [PATCH v2] net/virtio: add virtio hash report feature

> -Original Message- > From: Maxime Coquelin > Sent: Tuesday, February 25, 2025 9:39 PM > To: Shiva Shankar Kommula ; dev@dpdk.org; > chen...@nvidia.com > Cc: david.march...@redhat.com; Jerin Jacob ; Nithin > Kumar Dabilpuram ; Srujana Challa > > Subject: [EXTERNAL] Re: [PATCH v2] net/vir

Re: [v6 1/5] vhost: skip crypto op fetch before vring init

On 2/27/25 10:15 AM, Maxime Coquelin wrote: Hi Gowri, Thanks for the change, but I think there is an issue with the locking, more below: On 2/26/25 7:43 PM, Gowrishankar Muthukrishnan wrote: Until virtio avail ring is initialized (by VHOST_USER_SET_VRING_ADDR), worker thread should not try

Re: [PATCH] mempool: micro optimizations

On Thu, Feb 27, 2025 at 10:14:27AM +0100, Morten Brørup wrote: > > From: Bruce Richardson [mailto:bruce.richard...@intel.com] > > Sent: Wednesday, 26 February 2025 17.53 > > > > On Wed, Feb 26, 2025 at 03:59:22PM +, Morten Brørup wrote: > > > The comparisons lcore_id < RTE_MAX_LCORE and lcore_

Re: [v6 1/5] vhost: skip crypto op fetch before vring init

Hi Gowri, Thanks for the change, but I think there is an issue with the locking, more below: On 2/26/25 7:43 PM, Gowrishankar Muthukrishnan wrote: Until virtio avail ring is initialized (by VHOST_USER_SET_VRING_ADDR), worker thread should not try to fetch crypto op, which would lead to memory f

RE: [PATCH] mempool: micro optimizations

> From: Bruce Richardson [mailto:bruce.richard...@intel.com] > Sent: Wednesday, 26 February 2025 17.53 > > On Wed, Feb 26, 2025 at 03:59:22PM +, Morten Brørup wrote: > > The comparisons lcore_id < RTE_MAX_LCORE and lcore_id != LCORE_ID_ANY > are > > equivalent, but the latter compiles to fewer

RE: [PATCH v3 2/3] eal: only use numbers as align parameters for MSVC

> > After the instruction set updates for MSVC the error below popped up: > > ..\lib\eal\x86\include\rte_vect.h(82): error C2059: syntax error: '(' > > The issue is that MSVC does not allow __rte_aligned(RTE_X86_ZMM_SIZE). > It only accepts numbers that are power of 2. So, even though > RTE_X