Consult the official release version of dpdk-kmod

Hi, We are honored to use dpdk and dpdk-kmod, two outstanding open source softwares. However, according to the company's open source software usage standards, only officially released versions can be introduced as open source software. As far as I know, there is no official release of dpdk-kmod

Re: [RFC 00/27] Add VDUSE support to Vhost library

Hi Maxime, On Fri, Mar 31, 2023 at 11:43 PM Maxime Coquelin wrote: > > This series introduces a new type of backend, VDUSE, > to the Vhost library. > > VDUSE stands for vDPA device in Userspace, it enables > implementing a Virtio device in userspace and have it > attached to the Kernel vDPA bus.

Hardware timestamps

I have tried to get hw timestamps of packets on Intel X550T(10G) and AQC107(10G). But RTE_ETH_RX_OFFLOAD_TIMESTAMP = 0 in dev_info.rx_offload_capa. Tell me is it possible to get hw timestamps on these NICs with DPDK? Which 10G NICs support this option?

RE: Testpmd/l3fwd port shutdown failure on Arm Altra systems

Hi Qiming, It’s not an issue in VPP. It’s the XL710 NIC link down issue in DPDK testpmd. The ethernet links btw two XL710 NICs occasionally go down in FD.io VPP lab. If possible, could we have a talk on this issue this afternoon? Thanks. From: Yang, Qiming Sent: Tuesday, April 4, 2023 9:47 AM To

Re: [Bug 1209] vmxnet3 interface not receiving any packets with dpdk-22.11.1

Hi Jochen, Could you please help here? Regards Amiya On Fri, Mar 31, 2023 at 1:10 PM David Marchand wrote: > Hello Jochen, > > On Fri, Mar 31, 2023 at 9:07 AM wrote: > > > > Bug ID 1209 > > Summary vmxnet3 interface not receiving any packets with dpdk-22.11.1 > > Product DPDK > > Version 22.1

Clarify FEC capabilities

Hi, We are currently working on implementing the Forward Error Correction (FEC) feature in our driver, but we have encountered some difficulties in understanding the interpretation of the semantics of 'fec_capa', a bitmask of allowed FEC modes. Specifically, we are unclear about the meaning o

[PATCH v2 3/3] net/iavf: support Rx timestamp offload on SSE

This patch enables Rx timestamp offload on SSE data path. Enable timestamp offload with the command '--enable-rx-timestamp', pay attention that getting Rx timestamp offload will drop the performance. Signed-off-by: Zhichao Zeng --- v2: fix compile warning and timestamp error --- drivers/net/ia

[PATCH v2 2/3] net/iavf: support Rx timestamp offload on AVX2

This patch enables Rx timestamp offload on AVX2 data path. Enable timestamp offload with the command '--enable-rx-timestamp', pay attention that getting Rx timestamp offload will drop the performance. Signed-off-by: Zhichao Zeng --- v2: fix compile warning --- drivers/net/iavf/iavf_rxtx_vec_av

[PATCH v2 1/3] net/iavf: support Rx timestamp offload on AVX512

This patch enables Rx timestamp offload on AVX512 data path. Enable timestamp offload with the command '--enable-rx-timestamp', pay attention that getting Rx timestamp offload will drop the performance. Signed-off-by: Wenjun Wu Signed-off-by: Zhichao Zeng --- v2: fix compile warning --- drive

[PATCH v2 0/3] Enable iavf Rx Timestamp offload on vector path

Enable timestamp offload with the command '--enable-rx-timestamp', pay attention that getting Rx timestamp offload will drop the performance. --- v2: fix compile warning and SSE path Zhichao Zeng (3): net/iavf: support Rx timestamp offload on AVX512 net/iavf: support Rx timestamp offload on A

[Bug 1214] BCM57508 bnxt_hwrm_send_message Error(timeout) sending msg

https://bugs.dpdk.org/show_bug.cgi?id=1214 Bug ID: 1214 Summary: BCM57508 bnxt_hwrm_send_message Error(timeout) sending msg Product: DPDK Version: 21.08 Hardware: x86 OS: Linux Status: UNCONFIRME

[PATCH] dmadev: add tracepoints

Add tracepoints at important APIs for tracing support. Signed-off-by: Chengwen Feng --- lib/dmadev/meson.build | 2 +- lib/dmadev/rte_dmadev.c | 39 ++-- lib/dmadev/rte_dmadev.h | 56 --- lib/dmadev/rte_dmadev_trace.h| 133 +

Re: [PATCH] app/testpmd: txonly multiflow port change support

On Tue, 11 Apr 2023 13:11:07 -0700 Joshua Washington wrote: > Google cloud routes traffic using IP addesses without the support of MAC > addresses, so changing source IP address for txonly-multi-flow can have > negative performance implications for net/gve when using testpmd. This > change adds a

Re: [PATCH 3/9] eal: use barrier intrinsics when compiling with msvc

On Tue, Apr 11, 2023 at 10:10:48AM +0100, Bruce Richardson wrote: > On Mon, Apr 10, 2023 at 01:58:48PM -0700, Tyler Retzlaff wrote: > > On Mon, Apr 10, 2023 at 09:02:00PM +0100, Konstantin Ananyev wrote: > > > 06/04/2023 01:07, Tyler Retzlaff пишет: > > > >On Wed, Apr 05, 2023 at 10:57:02AM +,

[PATCH v4 14/14] eal: always define MSVC as little endian

The MSVC compiler does not target big endian platforms so define little endian always. Signed-off-by: Tyler Retzlaff --- lib/eal/include/generic/rte_byteorder.h | 2 ++ 1 file changed, 2 insertions(+) diff --git a/lib/eal/include/generic/rte_byteorder.h b/lib/eal/include/generic/rte_byteorder.

[PATCH v4 13/14] telemetry: avoid expanding versioned symbol macros on MSVC

Windows does not support versioned symbols. Fortunately Windows also doesn't have an exported stable ABI. Export rte_tel_data_add_array_int -> rte_tel_data_add_array_int_24 and rte_tel_data_add_dict_int -> rte_tel_data_add_dict_int_v24 functions. Windows does have a way to achieve similar version

[PATCH v4 11/14] eal: expand most macros to empty when using MSVC

For now expand a lot of common rte macros empty. The catch here is we need to test that most of the macros do what they should but at the same time they are blocking work needed to bootstrap of the unit tests. Later we will return and provide (where possible) expansions that work correctly for msv

[PATCH v4 08/14] eal: typedef cpu flag enum as int

Forward declaration of a enum is a non-standard extension and is not supported by MSVC. Use an int instead. Abstract the use of the int/enum rte_cpu_flag_t in function parameter lists by re-typdefing the enum rte_cpu_flag_t to the rte_cpu_flag_t identifier. Remove the use of __extension__ on func

[PATCH v4 12/14] eal: exclude exposure of rte atomic APIs for MSVC builds

It's discouraged to use rte_atomics APIs instead standard APIs should be used from C11. Since MSVC is a new toolchain/platform combination block visibility of the rte_atomic APIs from day 1. Signed-off-by: Tyler Retzlaff --- lib/eal/include/generic/rte_atomic.h | 7 +++ lib/eal/x86/include/r

[PATCH v4 10/14] eal: hide typedefs based on GCC vector extensions

When compiling with MSVC don't expose typedefs based on GCC vector extensions. Signed-off-by: Tyler Retzlaff --- lib/eal/include/generic/rte_vect.h | 4 1 file changed, 4 insertions(+) diff --git a/lib/eal/include/generic/rte_vect.h b/lib/eal/include/generic/rte_vect.h index 3fec2bf..7775

[PATCH v4 07/14] eal: use byte swap intrinsics

Inline assembly is not supported for MSVC x64 instead expand use _byteswap_u{short,long} intrinsics instead. Signed-off-by: Tyler Retzlaff --- lib/eal/x86/include/rte_byteorder.h | 27 +++ 1 file changed, 27 insertions(+) diff --git a/lib/eal/x86/include/rte_byteorder.h

[PATCH v4 09/14] eal: hide GCC extension based alignment markers

When compiling with MSVC don't expose typedefs used as alignment markers. Signed-off-by: Tyler Retzlaff --- lib/eal/include/rte_common.h | 4 1 file changed, 4 insertions(+) diff --git a/lib/eal/include/rte_common.h b/lib/eal/include/rte_common.h index 15765b4..2f464e3 100644 --- a/lib/eal

[PATCH v4 06/14] eal: use prefetch intrinsics

Inline assembly is not supported for MSVC x64 instead use _mm_prefetch and _mm_cldemote intrinsics. Signed-off-by: Tyler Retzlaff --- lib/eal/x86/include/rte_prefetch.h | 29 + 1 file changed, 29 insertions(+) diff --git a/lib/eal/x86/include/rte_prefetch.h b/lib/ea

[PATCH v4 05/14] eal: use umonitor umwait and tpause intrinsics

Inline assembly is not supported for MSVC x64 instead use _umonitor, _umwait and _tpause intrinsics. Signed-off-by: Tyler Retzlaff --- lib/eal/x86/rte_power_intrinsics.c | 12 1 file changed, 12 insertions(+) diff --git a/lib/eal/x86/rte_power_intrinsics.c b/lib/eal/x86/rte_power_

[PATCH v4 04/14] eal: use cpuid and cpuidex intrinsics

Inline assembly is not supported for MSVC x64 instead use __cpuid and __cpuidex intrinsics. Signed-off-by: Tyler Retzlaff --- lib/eal/x86/rte_cpuflags.c | 4 lib/eal/x86/rte_cpuid.h | 7 +++ lib/eal/x86/rte_cycles.c | 36 lib/eal/x86/rte

[PATCH v4 03/14] eal: use barrier intrinsics

Inline assembly is not supported for MSVC x64 instead expand rte_compiler_barrier as _ReadWriteBarrier and for rte_smp_mb _m_mfence intrinsics. Signed-off-by: Tyler Retzlaff --- lib/eal/include/generic/rte_atomic.h | 4 lib/eal/x86/include/rte_atomic.h | 5 - 2 files changed, 8 inse

[PATCH v4 01/14] eal: use rdtsc intrinsic

Inline assembly is not supported for MSVC x64. Convert code to use __rdtsc intrinsic. Signed-off-by: Tyler Retzlaff --- lib/eal/x86/include/rte_cycles.h | 14 -- 1 file changed, 8 insertions(+), 6 deletions(-) diff --git a/lib/eal/x86/include/rte_cycles.h b/lib/eal/x86/include/rte_c

[PATCH v4 00/14] msvc integration changes

In accordance with draft plan http://mails.dpdk.org/archives/web/2023-February/002023.html introduces conditionally compiled code to enable building with MSVC that _does not_ require C99/C11 meaning it can be integrated now. This series covers minimal changes for item #2 in draft plan for EAL depe

[PATCH v4 02/14] eal: use rtm and xtest intrinsics

Inline assembly is not supported for MSVC x64. Convert code to use _xend, _xabort and _xtest intrinsics. Signed-off-by: Tyler Retzlaff --- config/x86/meson.build| 6 ++ lib/eal/x86/include/rte_rtm.h | 18 +- 2 files changed, 11 insertions(+), 13 deletions(-) diff --

Re: [PATCH v3 11/11] telemetry: avoid expanding versioned symbol macros on msvc

On Tue, Apr 11, 2023 at 11:24:07AM +0100, Bruce Richardson wrote: > On Wed, Apr 05, 2023 at 05:45:19PM -0700, Tyler Retzlaff wrote: > > Windows does not support versioned symbols. Fortunately Windows also > > doesn't have an exported stable ABI. > > > > Export rte_tel_data_add_array_int -> rte_tel

[PATCH v3] app/testpmd: txonly multiflow port change support

Google cloud routes traffic using IP addresses without the support of MAC addresses, so changing source IP address for txonly-multi-flow can have negative performance implications for net/gve when using testpmd. This change adds a new flag --txonly-alter-port, which allows the alteration of source

[PATCH v2] app/testpmd: txonly multiflow port change support

Google cloud routes traffic using IP addresses without the support of MAC addresses, so changing source IP address for txonly-multi-flow can have negative performance implications for net/gve when using testpmd. This change adds a new flag --txonly-alter-port, which allows the alteration of source

[PATCH] app/testpmd: txonly multiflow port change support

Google cloud routes traffic using IP addesses without the support of MAC addresses, so changing source IP address for txonly-multi-flow can have negative performance implications for net/gve when using testpmd. This change adds a new flag --txonly-alter-port, which allows the alteration of source p

RE: [PATCH v2] vfio: do not coalesce DMA mappings

[AMD Official Use Only - General] > -Original Message- > From: Thomas Monjalon > Sent: Tuesday, April 11, 2023 8:43 PM > To: Burakov, Anatoly ; David Marchand > ; Gupta, Nipun > Cc: dev@dpdk.org; Yigit, Ferruh ; Agarwal, Nikhil > > Subject: Re: [PATCH v2] vfio: do not coalesce DMA map

[Bug 1212] [dpdk-20.11.8]pvp_qemu_multi_paths_port_restart:perf_pvp_qemu_vector_rx_mac: performance drop about 23.5% when send small packets

https://bugs.dpdk.org/show_bug.cgi?id=1212 Luca Boccassi (luca.bocca...@gmail.com) changed: What|Removed |Added Resolution|--- |FIXED Statu

Re: [PATCH v2] vfio: do not coalesce DMA mappings

04/04/2023 18:32, Nipun Gupta: > On 4/4/2023 8:43 PM, Burakov, Anatoly wrote: > > The patch probably shouldn't include the mailmap changes :) > > Sure, will send a separate patch for it. No please, we squash mailmap changes with 1st patch.

Re: [PATCH v1 1/2] dts: fabric requirements

04/04/2023 13:51, Juraj Linkeš: > On Mon, Apr 3, 2023 at 5:18 PM Thomas Monjalon wrote: > > > 03/04/2023 16:56, Juraj Linkeš: > > > On Mon, Apr 3, 2023 at 2:33 PM Thomas Monjalon > > wrote: > > > > > > > 03/04/2023 13:46, Juraj Linkeš: > > > > > Replace pexpect with Fabric. > > > > > > > > You s

RE: [PATCH] doc: fix invalid auth algo in cryptoperf app

> -Original Message- > From: Akhil Goyal > Sent: Tuesday 11 April 2023 14:05 > To: dev@dpdk.org > Cc: ano...@marvell.com; Power, Ciara ; > sta...@dpdk.org; Akhil Goyal > Subject: [PATCH] doc: fix invalid auth algo in cryptoperf app > > 3des-cbc is not an authentication algorithm, henc

[PATCH] doc: fix invalid auth algo in cryptoperf app

3des-cbc is not an authentication algorithm, hence need to be removed. Fixes: c6baca7adc94 ("doc: describe new performance test application") Cc: sta...@dpdk.org Signed-off-by: Akhil Goyal --- doc/guides/tools/cryptoperf.rst | 1 - 1 file changed, 1 deletion(-) diff --git a/doc/guides/tools/cr

Re: [PATCH v3 11/11] telemetry: avoid expanding versioned symbol macros on msvc

On Wed, Apr 05, 2023 at 05:45:19PM -0700, Tyler Retzlaff wrote: > Windows does not support versioned symbols. Fortunately Windows also > doesn't have an exported stable ABI. > > Export rte_tel_data_add_array_int -> rte_tel_data_add_array_int_24 > and rte_tel_data_add_dict_int -> rte_tel_data_add_d

[PATCH 3/3] test/security: add unittest for inline ingress oop

Add unittest for inline ingress out-of-place processing. Signed-off-by: Nithin Dabilpuram --- app/test/test_cryptodev_security_ipsec.c | 8 +++ app/test/test_cryptodev_security_ipsec.h | 1 + app/test/test_security_inline_proto.c| 85 3 files changed, 94 insertions

[PATCH 2/3] net/cnxk: support inline ingress out of place session

Add support for inline ingress session with out-of-place support. Signed-off-by: Nithin Dabilpuram --- Depends-on: series-27660 ("common/cnxk: allocate dynamic BPIDs) drivers/event/cnxk/cn10k_worker.h | 28 - drivers/net/cnxk/cn10k_ethdev.c | 13 +- drivers/net/cnxk/cn10k_ethdev_se

[PATCH 1/3] security: introduce out of place support for inline ingress

Similar to out of place(OOP) processing support that exists for Lookaside crypto/security sessions, Inline ingress security sessions may also need out of place processing in usecases where original encrypted packet needs to be retained for post processing. So for NIC's which have such a kind of HW

Re: 20.11.7 patches review and test

On Tue, 11 Apr 2023 at 08:22, Morten Brørup wrote: > > > From: Morten Brørup [mailto:m...@smartsharesystems.com] > > Sent: Monday, 28 November 2022 13.08 > > > > > From: Luca Boccassi [mailto:luca.bocca...@gmail.com] > > > Sent: Monday, 28 November 2022 12.55 > > > > > > On Mon, 28 Nov 2022 at 11:

[PATCH 21/21] common/cnxk: support of 1:n pool:aura per NIX LF

From: Rahul Bhansali This will add the support of 1:n pool:aura per NIX LF when inl_cpt_channel devargs is set to inline device, otherwise it will create 1:1 pool:aura for CN103/CN106B0 SOCs. With 1:N, global pool will be created with Aura 0, and per NIX individual aura will be created and mappe

[PATCH 20/21] net/cnxk: add receive error mask

From: Rakesh Kudurumalla receive errors related to MACSEC and USXGMI are masked for cn10kb_b0 and cn10kb Signed-off-by: Rakesh Kudurumalla --- drivers/net/cnxk/cnxk_ethdev.c | 5 + 1 file changed, 5 insertions(+) diff --git a/drivers/net/cnxk/cnxk_ethdev.c b/drivers/net/cnxk/cnxk_ethdev.c

[PATCH 19/21] common/cnxk: access valid pass value

From: Hanumanth Pothula There is a possibility of accessing an invalid pass value on rvu device look up failure, as the return value is dropped. Hence pass through the return value of rvu device look, to make sure valid pass value is accessed. Signed-off-by: Hanumanth Pothula --- drivers/comm

[PATCH 17/21] common/cnxk: add more comments to mbox code

From: Harman Kalra Adding more comments to the mbox routines to understand the flow well. Signed-off-by: Harman Kalra --- drivers/common/cnxk/roc_dev.c | 20 +--- drivers/common/cnxk/roc_mbox.c | 5 + 2 files changed, 22 insertions(+), 3 deletions(-) diff --git a/drivers

[PATCH 18/21] common/cnxk: add CN105xxN B0 model

From: Harman Kalra Adding support for CN105xxN B0 pass Signed-off-by: Harman Kalra --- drivers/common/cnxk/roc_model.c | 1 + drivers/common/cnxk/roc_model.h | 9 - 2 files changed, 9 insertions(+), 1 deletion(-) diff --git a/drivers/common/cnxk/roc_model.c b/drivers/common/cnxk/roc_m

[PATCH 16/21] common/cnxk: sync between mbox up and down messages

From: Harman Kalra An issue is observed where if PF is with DPDK and VF as kernel netdev does not responds to link events. It was due to recent design change in kernel where sender checks whether previous interrupt is received before triggering current interrupt by waiting for mailbox data regist

[PATCH 15/21] common/cnxk: update errata info

Update errata info based on CN10KA B0 and CN10KB A0. Also remove duplicate model check roc_model_is_cn103xx() Signed-off-by: Nithin Dabilpuram --- drivers/common/cnxk/roc_errata.h | 20 drivers/common/cnxk/roc_features.h | 2 +- drivers/common/cnxk/roc_model.h| 6 ---

[PATCH 14/21] common/cnxk: avoid STALL with dual rate on CNF95N

From: Satha Rao Due to errata RED_ALGO STALL with dual shaper rate will hangs on platforms CNF95N and CNF95O. Set READ_ALGO to DISCARD with dual shaper rate on CNF95N and CNF95O. Signed-off-by: Satha Rao --- drivers/common/cnxk/roc_nix_tm_utils.c | 3 ++- 1 file changed, 2 insertions(+), 1 del

[PATCH 13/21] common/cnxk: set relchan in TL4 config for each SDP queue

From: Veerasenareddy Burru set distinct relchan in each TL4 queue connected to SDP. currently rechan in TL4 SDP config is getting set to 0 for all SDP-NIX queues. Each TL4 queues for SDP need to be configured with distinct channel for SDP to provide per channel backpressure to NIX. Signed-off-b

[PATCH 12/21] common/cnxk: fix parameter in NIX dump

From: Gowrishankar Muthukrishnan Fix parameter passed to nix_dump to what expected in format specifier. Fixes: d2f168dfa5de ("common/cnxk: support 10K B0 for inline IPsec") Signed-off-by: Gowrishankar Muthukrishnan --- drivers/common/cnxk/roc_nix_debug.c | 4 ++-- 1 file changed, 2 insertions(

[PATCH 11/21] common/cnxk: fix null pointer dereference

From: Gowrishankar Muthukrishnan Fix null pointer dereferences reported by klocwork. Fixes: 4398c4092f3d ("common/cnxk: dump inline device RQ context") Fixes: 79dc6f324e82 ("common/cnxk: add inline function for statistics") Signed-off-by: Gowrishankar Muthukrishnan --- drivers/common/cnxk/roc

[PATCH 10/21] common/cnxk: add receive error mask

From: Rakesh Kudurumalla Adding support to configure receive error mask for 106B0 Signed-off-by: Rakesh Kudurumalla --- drivers/common/cnxk/roc_features.h | 6 ++ drivers/common/cnxk/roc_nix.h | 16 2 files changed, 22 insertions(+) diff --git a/drivers/common/cnxk/

[PATCH 09/21] common/cnxk: fetch eng caps for inl outb inst format

Fetch engine caps and use it along with model check to determine inline outbound instruction format with NIX Tx offset or address. Signed-off-by: Nithin Dabilpuram --- drivers/common/cnxk/roc_cpt.h | 3 + drivers/common/cnxk/roc_nix_inl.c | 101 drivers/com

[PATCH 08/21] common/cnxk: update age drop statistics

From: Satha Rao Update age drop statistics. Added telemetry statistics for age drops. Signed-off-by: Satha Rao --- drivers/common/cnxk/cnxk_telemetry_nix.c | 4 drivers/common/cnxk/hw/nix.h | 2 ++ drivers/common/cnxk/roc_features.h | 6 ++ drivers/common/cnxk/roc_ni

[PATCH 07/21] common/cnxk: make aura flow control config more predictable

Restrict shared BPID config only when force BP is enabled and make aura flow control config more predictable by not disabling it if there is a collision but ignore new config and log the same. Also remove BPID setup from Rx adapter as it is now evaluated and configured every time ethdev is stopped

[PATCH 06/21] common/cnxk: add percent drop threshold to pool

From: Sunil Kumar Kori Currently hard coded drop threshold(95%) is configured to aura/pool as a threshold for drop limit. Patch adds a input parameter to RoC API so that user passed percentage value can be configured. Signed-off-by: Sunil Kumar Kori --- drivers/common/cnxk/roc_nix.h

[PATCH 05/21] common/cnxk: dump SW SSO work count as xstat

Dump SW SSO work count as xstat. Signed-off-by: Nithin Dabilpuram --- drivers/common/cnxk/roc_nix_inl_dev_irq.c | 1 + drivers/common/cnxk/roc_nix_inl_priv.h| 1 + drivers/common/cnxk/roc_nix_stats.c | 17 +++-- drivers/common/cnxk/roc_nix_xstats.h | 4 4 files

[PATCH 04/21] common/cnxk: reduce sqes per sqb by one

From: Satha Rao Each SQB reserves last SQE to store pointer to next SQB. So each SQB will holds either 31 or 63 based on send descriptors selected. This patch also consider sqb_slack to maintain threshold buffers to sync between HW and SW. Threshold will be maximum of 30% of queue size or sqb_sl

[PATCH 03/21] common/cnxk: skip flow ctrl set on non-existent meta aura

Skip setting flow control on local meta aura if it is not yet created. Also in flow control mode set, do get to confirm if it is in same state to avoid a set and unnecessary mbox failures. Signed-off-by: Nithin Dabilpuram --- .mailmap | 1 + drivers/common/cnxk/roc_dev.

[PATCH 02/21] common/cnxk: add pool BPID to RQ while using common pool

From: Shijith Thotton When RQs of two different traffic classes are using the same mempool, BPIDs could differ between the RQs and BPID of only one RQ can be configured per pool. In such cases, a new BPID is configured on both RQs and pool or pool back-pressure is disabled. CN103xx and CN106xx B

[PATCH 01/21] common/cnxk: allocate dynamic BPIDs

From: Satha Rao New mail box to allocate/free dynamic BPIDs based on NIX type. Added to new mail box APIs to get/set RX channel config with new BPIDs. Signed-off-by: Satha Rao --- Depends-on: series-27659 ("add hwpools and support exchanging mbufs between pools") drivers/common/cnxk/roc_cpt

Re: [PATCH 3/9] eal: use barrier intrinsics when compiling with msvc

On Mon, Apr 10, 2023 at 01:58:48PM -0700, Tyler Retzlaff wrote: > On Mon, Apr 10, 2023 at 09:02:00PM +0100, Konstantin Ananyev wrote: > > 06/04/2023 01:07, Tyler Retzlaff пишет: > > >On Wed, Apr 05, 2023 at 10:57:02AM +, Konstantin Ananyev wrote: > > >> > > >>>Inline assembly is not support

Re: [PATCH v3 4/5] telemetry: rename local variables

On Fri, Apr 07, 2023 at 12:50:06PM -0700, Tyler Retzlaff wrote: > On Wed, Apr 05, 2023 at 05:03:25PM +0100, Bruce Richardson wrote: > > In the newly separated out function, rename "tmp" to "buf" to have more > > meaningful variable names. > > > > Signed-off-by: Bruce Richardson > > > > --- > Ack

Re: [PATCH v2 1/5] telemetry: fix autotest failures on Alpine

On Fri, Apr 07, 2023 at 12:21:16PM -0700, Tyler Retzlaff wrote: > On Wed, Apr 05, 2023 at 04:44:10PM +0100, Bruce Richardson wrote: > > On Alpine linux, the telemetry_data_autotest was failing for the > > test where we had dictionaries embedded in other dictionaries up > > to three levels deep. Ind

[PATCH 5/5] mempool/cnxk: add support for exchanging mbufs between pools

Add the following cnxk mempool PMD APIs to facilitate exchanging mbufs between pools. * rte_pmd_cnxk_mempool_is_hwpool() - Allows user to check whether a pool is hwpool or not. * rte_pmd_cnxk_mempool_range_check_disable() - Disables range checking on any rte_mempool. * rte_pmd_cnxk_mempool

[PATCH 4/5] mempool/cnxk: add hwpool ops

Add hwpool ops which can used to create a rte_mempool that attaches to another rte_mempool. The hwpool will not have its own buffers and will have a dummy populate callback. Only an NPA aura will be allocated for this rte_mempool. The buffers will be allocate from the NPA pool of the attached rte_m

[PATCH 3/5] mempool/cnxk: add NPA aura range get/set APIs

Current APIs to set range on auras modifies both the aura range limits in software and pool range limits in NPA hardware. Newly added ROC APIs allow to set/get aura range limits in software alone without modifying hardware. The existing aura range set functionality has been moved as a pool range

[PATCH 2/5] common/cnxk: add NPA aura create/destroy ROC APIs

Add ROC APIs which allows to create NPA auras independently and attach it to an existing NPA pool. Also add API to destroy NPA auras independently. Signed-off-by: Ashwin Sekhar T K --- drivers/common/cnxk/roc_npa.c | 219 drivers/common/cnxk/roc_npa.h | 4 +

[PATCH 1/5] mempool/cnxk: use pool config to pass flags

Use lower bits of pool_config to pass flags specific to cnxk mempool PMD ops. Signed-off-by: Ashwin Sekhar T K --- drivers/mempool/cnxk/cnxk_mempool.h | 24 drivers/mempool/cnxk/cnxk_mempool_ops.c | 17 ++--- drivers/net/cnxk/cnxk_ethdev_sec.c | 25 +

[PATCH 0/5] add hwpools and support exchanging mbufs between pools

This patch series enables the creation of special types of pool in cnxk mempool PMD called as hwpools. A hwpool will not have any buffers of its own but instead will only have a hardware AURA associated with it. This hwpool will be attached to another normal rte_mempool and the hwpool will use the

RE: 20.11.7 patches review and test

> From: Morten Brørup [mailto:m...@smartsharesystems.com] > Sent: Monday, 28 November 2022 13.08 > > > From: Luca Boccassi [mailto:luca.bocca...@gmail.com] > > Sent: Monday, 28 November 2022 12.55 > > > > On Mon, 28 Nov 2022 at 11:35, Morten Brørup > > wrote: > > > > > > +CC Andrew Rybchenko as c

[PATCH] mempool/cnxk: avoid indefinite wait when counting batch alloc pointers

Avoid waiting indefinitely when counting batch alloc pointers by adding a wait timeout. Signed-off-by: Ashwin Sekhar T K --- drivers/common/cnxk/roc_npa.h| 15 +-- drivers/mempool/cnxk/cn10k_mempool_ops.c | 3 ++- 2 files changed, 11 insertions(+), 7 deletions(-) diff -

[PATCH] common/cnxk: update ROC erratas

Update the models where errata IPBUNPA-37480 is applicable. Signed-off-by: Ashwin Sekhar T K --- drivers/common/cnxk/roc_errata.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/common/cnxk/roc_errata.h b/drivers/common/cnxk/roc_errata.h index 2d15e639b7..1333bde629

[PATCH 1/2] net/cnxk: optimize flow control hysteresis

From: Pavan Nikhilesh Set flow control hysteresis to ignore immediate sequence of decrement-increment to avoid unnecessary LLC traffic. Enable hysteresis when SQ length is more than 512 as it lower queue lengths will require faster updates. Signed-off-by: Pavan Nikhilesh Signed-off-by: Satha Ra

[PATCH 2/2] event/cnxk: restrict stashing to multi-core

From: Pavan Nikhilesh Enable WQE stashing by default only when multiple event ports are enabled. Signed-off-by: Pavan Nikhilesh --- drivers/event/cnxk/cn10k_eventdev.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/event/cnxk/cn10k_eventdev.c b/drivers/event/cnxk/