Re: [RFC v2 0/3] Introduce Topology NUMA grouping for lcores

2024-10-29 Thread Mattias Rönnblom
On 2024-10-30 05:40, Vipin Varghese wrote: One of the way to increase core density in a given physical package is to use easy to replicate tiles. These tiles are either core complexes or core complexes with IO (memory and PCIe). This results to possibility of having two types NUMA topology. - C

RE: [PATCH v01] net/af_packet: add rollover and defrag options

2024-10-29 Thread Gur Stavi
> > + if (strstr(pair->key, ETH_AF_PACKET_ROLLOVER) != NULL) { > > + rollover = atoi(pair->value); > > + if (rollover != 0 && rollover != 1) { > > + PMD_LOG(ERR, > > + "%s: invalid rollover val

[RFC v3 3/3] examples: add lcore topology API calls

2024-10-29 Thread Vipin Varghese
Enhance example code to allow topology based lcores API, while retaining default behaviour. - helloworld: allow lcoes to send hello to lcores in selected topology. - l2fwd: allow use of IO lcores topology. - skeleton: choose the lcore from IO topology which has more ports. v3 changes: - fix t

[RFC v3 2/3] test/lcore: enable tests for topology

2024-10-29 Thread Vipin Varghese
add functional test cases to validate topology supported lcore API. v3 changes: - fix test test-report/2024-October/817748.html Signed-off-by: Vipin Varghese --- app/test/test_lcores.c | 189 + 1 file changed, 189 insertions(+) diff --git a/app/test/tes

[RFC v3 1/3] eal/lcore: add topology based functions

2024-10-29 Thread Vipin Varghese
Introduce topology aware lcore mapping into lcore API. With higher core density, more and more cores are categorized into various chiplets based on IO (memory and PCIe) and Last Level Cache (mainly L3). Using hwloc library, the dpdk available lcores can be grouped into various groups nameley L1, L

[RFC v3 0/3] Introduce Topology NUMA grouping for lcores

2024-10-29 Thread Vipin Varghese
One of the way to increase core density in a given physical package is to use easy to replicate tiles. These tiles are either core complexes or core complexes with IO (memory and PCIe). This results to possibility of having two types NUMA topology. - CPU topology & IO topology - CPU+IO topology

[RFC v2 1/3] eal/lcore: add topology based functions

2024-10-29 Thread Vipin Varghese
Introduce topology aware lcore mapping into lcore API. With higher core density, more and more cores are categorized into various chiplets based on IO (memory and PCIe) and Last Level Cache (mainly L3). Using hwloc library, the dpdk available lcores can be grouped into various groups nameley L1,

[RFC v2 3/3] examples: add lcore topology API calls

2024-10-29 Thread Vipin Varghese
Enhance example code to allow topology based lcores API, while retaining default behaviour. - helloworld: allow lcoes to send hello to lcores in selected topology. - l2fwd: allow use of IO lcores topology. - skeleton: choose the lcore from IO topology which has more ports. Signed-off-by: Vipin

[RFC v2 2/3] test/lcore: enable tests for topology

2024-10-29 Thread Vipin Varghese
add functional test cases to validate topology supported lcore API. Signed-off-by: Vipin Varghese --- app/test/test_lcores.c | 189 + 1 file changed, 189 insertions(+) diff --git a/app/test/test_lcores.c b/app/test/test_lcores.c index bd5c0dd94b..3e57a3f2

[RFC v2 0/3] Introduce Topology NUMA grouping for lcores

2024-10-29 Thread Vipin Varghese
One of the way to increase core density in a given physical package is to use easy to replicate tiles. These tiles are either core complexes or core complexes with IO (memory and PCIe). This results to possibility of having two types NUMA topology. - CPU topology & IO topology - CPU+IO topology

Re: [PATCH RESEND v7 0/5] app/testpmd: support multiple process attach and detach port

2024-10-29 Thread lihuisong (C)
在 2024/10/30 6:12, Ferruh Yigit 写道: On 10/18/2024 3:48 AM, lihuisong (C) wrote: Hi Ferruh, Thanks for your considering again. please see reply inline. 在 2024/10/18 9:04, Ferruh Yigit 写道: On 10/8/2024 3:32 AM, lihuisong (C) wrote: Hi Thomas and Ferruh, We've discussed it on and off a few t

[PATCH] dpdk: remove redundant null check when parse kvargs

2024-10-29 Thread Chengwen Feng
As rte_kvargs_process() API already restrict only process key=value, so the callback will never exist 'value == NULL' case. This commit remove such above 'value == NULL' check. Plus, the net/tap driver should use rte_kvargs_process_opt() because it will treat only-key as normal case, so this commi

[PATCH v2] net/ice: fix incorrect reading of PHY timestamp

2024-10-29 Thread Soumyadeep Hore
In ICE PMD, previously the ready bitmap checking before reading PHY timestamp was not present. This caused incorrect Tx timestamping. The ready bitmap checking is enabled and PHY timestamp is read once the ready bitmap gives positive value. Fixes: 881169950d80 ("net/ice/base: implement initial PT

Re: [PATCH] net/vmxnet3: Fix vmxnet3 NULL pointer deref after configuration failure

2024-10-29 Thread Ferruh Yigit
On 10/26/2024 3:33 PM, Roger B Melton wrote: > Problem: > > If vxmnet3_dev_configure() fails, applications may call > vmxnet3_dev_close(). If the failure occurs before the vmxnet3 > hw->shared structure is allocated the close will lead to a segv. > > Root Cause: > > This crash is due to incorrec

Re: [PATCH 0/3] net/ena: ena release 2.11.0

2024-10-29 Thread Ferruh Yigit
On 10/23/2024 10:41 AM, shaib...@amazon.com wrote: > From: Shai Brandes > > Dear all, > Please find attached the patches for our upcoming release. > We welcome and appreciate any feedback you may have. > Thank you in advance. > Best regards, > Shai Brandes > > Shai Brandes (3): > net/ena: upgr

Re: [PATCH 3/3] net/ena: remove outdated devarg documentation

2024-10-29 Thread Ferruh Yigit
On 10/23/2024 10:41 AM, shaib...@amazon.com wrote: > From: Shai Brandes > > Removed obsolete devarg enforcing normal llq policy, > as its logic is now handled by the new `llq_policy` devarg. > > Signed-off-by: Shai Brandes > Fixes: d7918d19d25e ("net/ena: restructure LLQ policy user setting")

Re: [PATCH] net/hns3: fix inaccurate error log

2024-10-29 Thread Ferruh Yigit
On 10/23/2024 8:12 AM, Jie Hai wrote: > If register number obtained from firmware is different from the > register list driver hold, an error log occurs. The incorrect > use of `&&` makes this log inaccurate, change it to `&`. > > Signed-off-by: Jie Hai > Fixes: 99d3bd8b85d3 ("net/hns3: support

Re: [PATCH] testpmd: add hairpin-map parameter

2024-10-29 Thread Stephen Hemminger
On Thu, 28 Sep 2023 16:39:06 +0300 Gregory Etelson wrote: > +static __rte_always_inline > +char *parse_hairpin_map_entry(char *input, char **next) > +{ > + char *tail = strchr(input, ':'); > + > + if (!tail) > + return NULL; > + tail[0] = '\0'; > + *next = tail + 1; >

Re: [PATCH v4 00/86] Provide flow filter API and statistics

2024-10-29 Thread Ferruh Yigit
On 10/29/2024 4:41 PM, Serhii Iliushyk wrote: > The list of updates provided by the patchset: > - FW version > - Speed capabilities > - Link status (Link update only) > - Unicast MAC filter > - Multicast MAC filter > - Promiscuous mode (Enable only. The device always run promiscuous mod

Re: [PATCH v4 64/86] net/ntnic: update documentation

2024-10-29 Thread Ferruh Yigit
On 10/29/2024 4:42 PM, Serhii Iliushyk wrote: > From: Oleksandr Kolomeiets > > Update required documentation > > Signed-off-by: Oleksandr Kolomeiets > --- > doc/guides/nics/ntnic.rst | 30 ++ > doc/guides/rel_notes/release_24_11.rst | 2 ++ > 2 files chang

Re: [PATCH v4 01/86] net/ntnic: add API for configuration NT flow dev

2024-10-29 Thread Ferruh Yigit
On 10/29/2024 4:41 PM, Serhii Iliushyk wrote: > From: Danylo Vodopianov > > This API allows to enable of flow profile for NT SmartNIC > > Signed-off-by: Danylo Vodopianov > <...> > + for (i = 0; i < alloc_rx_queues; i++) { > +#ifdef SCATTER_GATHER > + eth_dev->rx_queue[i] = q

Re: [PATCH v4 05/86] net/ntnic: add minimal NT flow inline profile

2024-10-29 Thread Ferruh Yigit
On 10/29/2024 4:41 PM, Serhii Iliushyk wrote: > The flow profile implements a all flow related operations > Can you please give some more details about the profiles, and "inline profile" mentioned? > Signed-off-by: Serhii Iliushyk > --- > drivers/net/ntnic/include/flow_api.h | 15

Re: [PATCH v4 70/86] net/ntnic: add aging documentation

2024-10-29 Thread Ferruh Yigit
On 10/29/2024 4:42 PM, Serhii Iliushyk wrote: > From: Danylo Vodopianov > > ntnic.rst document was exntede with age feature specification. > ntnic.ini was extended with rte_flow action age support. > > Signed-off-by: Danylo Vodopianov > --- > doc/guides/nics/features/ntnic.ini | 1 + > do

Re: [PATCH v4 15/86] net/ntnic: add item IPv4

2024-10-29 Thread Ferruh Yigit
On 10/29/2024 4:41 PM, Serhii Iliushyk wrote: > Add possibility to use RTE_FLOW_ITEM_TYPE_IPV4 > > Signed-off-by: Serhii Iliushyk > --- > doc/guides/nics/features/ntnic.ini| 1 + > .../profile_inline/flow_api_profile_inline.c | 162 ++ > 2 files changed, 163 insert

Re: [PATCH 0/9] support multiple PF flower firmware

2024-10-29 Thread Ferruh Yigit
On 10/21/2024 7:32 AM, Chaoyong He wrote: > This patch series add support of the multiple PF flower firmware. > > Peng Zhang (9): > net/nfp: extract the function to allocate the PHY > net/nfp: extract the function to initialize the PF > net/nfp: extract the function to allocate PF > net/nf

Re: [PATCH v2 0/2] net/hns3: bugfix on register dump

2024-10-29 Thread Ferruh Yigit
On 10/22/2024 5:01 AM, Jie Hai wrote: > > Jie Hai (2): > net/hns3: fix variable type > net/hns3: fix pointer offset > Series applied to dpdk-next-net/main, thanks.

Re: [PATCH v5 1/1] testpmd: add hairpin-map parameter

2024-10-29 Thread Stephen Hemminger
On Tue, 29 Oct 2024 21:51:40 + Ferruh Yigit wrote: > On 9/22/2024 7:50 AM, Gregory Etelson wrote: > > Hairpin offloads packet forwarding between ports. > > Packet is expected on Rx port , Rx queue and is forwarded > > to Tx port Tx queue . > > > > Testpmd implements a static hairpin config

Re: [PATCH RESEND v7 0/5] app/testpmd: support multiple process attach and detach port

2024-10-29 Thread Ferruh Yigit
On 10/18/2024 3:48 AM, lihuisong (C) wrote: > Hi Ferruh, > > Thanks for your considering again. please see reply inline. > > 在 2024/10/18 9:04, Ferruh Yigit 写道: >> On 10/8/2024 3:32 AM, lihuisong (C) wrote: >>> Hi Thomas and Ferruh, >>> >>> We've discussed it on and off a few times, and we've rea

Re: [PATCH v5 1/1] testpmd: add hairpin-map parameter

2024-10-29 Thread Ferruh Yigit
On 9/22/2024 7:50 AM, Gregory Etelson wrote: > Hairpin offloads packet forwarding between ports. > Packet is expected on Rx port , Rx queue and is forwarded > to Tx port Tx queue . > > Testpmd implements a static hairpin configuration scheme. > > The new parameter allows explicit selection of R

Re: [PATCH v2 5/7] dts: add support for externally compiled DPDK

2024-10-29 Thread Luca Vizzarro
On 29/10/2024 15:48, Dean Marx wrote: I may have just misunderstood how the command line args are supposed to be used. I'm assuming based on your response that --config-file is only used for when the tarball/tree is located on the host, versus --tarball or --dpdk-tree being for when they're loc

Re: [PATCH v2 6/7] doc: update argument options for external DPDK build

2024-10-29 Thread Patrick Robb
On Tue, Oct 29, 2024 at 8:35 AM Luca Vizzarro wrote: > On 29/10/2024 01:23, Patrick Robb wrote: > > On Mon, Oct 21, 2024 at 9:50 AM Luca Vizzarro > > wrote: > > > > > > + > > + DPDK Build Options: > > + Arguments in this group (and subgroup) will be

[PATCH v3 1/1] net/bonding: make bonding functions stable

2024-10-29 Thread Lukas Sismis
Remove rte_experimental macros from the stable functions as they have been part of the stable API since 23.11. Signed-off-by: Lukas Sismis --- drivers/net/bonding/rte_eth_bond.h| 4 drivers/net/bonding/rte_eth_bond_8023ad.h | 1 - drivers/net/bonding/version.map | 15 +++

Re: [PATCH] net/bonding: make bonding functions stable

2024-10-29 Thread Lukáš Šišmiš
Sorry for multiple patch request versions, I am going to make a new one as per your suggestions. Thanks! Lukas On 29. 10. 24 18:36, Stephen Hemminger wrote: On Tue, 29 Oct 2024 16:41:13 +0100 Lukas Sismis wrote: diff --git a/drivers/net/bonding/version.map b/drivers/net/bonding/version.map

Re: [PATCH v2 5/7] dts: add support for externally compiled DPDK

2024-10-29 Thread Patrick Robb
On Tue, Oct 29, 2024 at 7:48 AM Luca Vizzarro wrote: > Hi Patrick, > > On 29/10/2024 01:19, Patrick Robb wrote: > > diff --git a/dts/conf.yaml b/dts/conf.yaml > > index 814744a1fc..8a65a481d6 100644 > > --- a/dts/conf.yaml > > +++ b/dts/conf.yaml > > @@ -5,12 +5,24 @@ > >

Re: [PATCH v2 5/7] dts: add support for externally compiled DPDK

2024-10-29 Thread Dean Marx
Also, I want to point out that even though the --config-file argument defaults to /dpdk/dts/conf.yaml, you still can't run ./main.py by itself.

Re: release candidate 24.11-rc1

2024-10-29 Thread Thinh Tran
IBM - Power Systems DPDK v24.11-rc1-6-g90cb8ff819 * Build CI on Fedora 38,39,40 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)

Re: [PATCH] mbuf: add transport mode ESP packet type

2024-10-29 Thread Ajit Khaparde
On Tue, Oct 29, 2024 at 7:48 AM Ori Kam wrote: > > Hi Ajit, Nithin and Olivier > > > -Original Message- > > From: Ajit Khaparde > > Sent: Friday, October 25, 2024 2:33 AM > > > > On Thu, Oct 24, 2024 at 12:30 PM Alexander Kozyrev > > wrote: > > > > > > >>And we definitely need RTE_PTYPE_

[PATCH v4 78/86] net/ntnic: migrate to the RTE spinlock

2024-10-29 Thread Serhii Iliushyk
Migarte form the pthread to rte_spinlock Signed-off-by: Serhii Iliushyk --- drivers/net/ntnic/include/flow_api.h | 6 +- drivers/net/ntnic/include/ntdrv_4ga.h | 3 +- .../net/ntnic/nthw/core/include/nthw_i2cm.h | 4 +- .../net/ntnic/nthw/core/include/nthw_rpf.h| 5 +-

RE: [PATCH v01] net/af_packet: add rollover and defrag options

2024-10-29 Thread Gur Stavi
> On Tue, 29 Oct 2024 15:48:05 +0200 > Gur Stavi wrote: > > > net_af_packet PMD multi "queue" support relies on Linux FANOUT > capability. > > Linux FANOUT is a SW based load balancer that is similar to HW RSS > which is > > more common for DPDK PMDs. Instead of multiple HW descriptor queues, AF

[PATCH v4 46/86] net/ntnic: add MAC Tx module registers

2024-10-29 Thread Serhii Iliushyk
From: Oleksandr Kolomeiets The Media Access Control Transmit module contains counters that keep track on transmitted packets. Signed-off-by: Oleksandr Kolomeiets --- .../supported/nthw_fpga_9563_055_049_.c | 38 ++- .../ntnic/nthw/supported/nthw_fpga_mod_defs.h | 3 +- .

[PATCH v4 51/86] net/ntnic: add Tx RPL module registers

2024-10-29 Thread Serhii Iliushyk
From: Oleksandr Kolomeiets The TX Replacer module can replace a range of bytes in a packet. The replacing data is stored in a table in the module and will often contain tunnel data. Signed-off-by: Oleksandr Kolomeiets --- .../supported/nthw_fpga_9563_055_049_.c | 41 ++-

Re: [PATCH] igc: fix invalid length and corrupted multi-segment mbufs

2024-10-29 Thread Bruce Richardson
On Mon, Oct 28, 2024 at 03:17:07PM +0100, Martin Weiser wrote: > > The issue only appeared with hardware-timestamping enabled > (RTE_ETH_RX_OFFLOAD_TIMESTAMP). > > The length of the prepended hardware timestamp was not subtracted from > the data length so that received packets were 16 bytes longe

Re: [PATCH 3/5] dts: use Pydantic in the configuration

2024-10-29 Thread Luca Vizzarro
Likewise, thank you Juraj for the comments. Most if not all suggestions have been applied in the new versions. On 17/09/2024 12:13, Juraj Linkeš wrote: - the config schema is no longer used for validation but kept as an    alternative format for the developer If it's not used, we should remov

[PATCH v6 4/5] net/ice: allowing stopping port to apply TM topology

2024-10-29 Thread Bruce Richardson
The rte_tm topology commit requires the port to be stopped on apply. Rather than just returning an error when the port is already started, we can stop the port, apply the topology to it and then restart it. Signed-off-by: Bruce Richardson Acked-by: Vladimir Medvedkin --- drivers/net/ice/ice_tm.

[PATCH v4 44/86] net/ntnic: add IFR module registers

2024-10-29 Thread Serhii Iliushyk
From: Oleksandr Kolomeiets The IP Fragmenter module can fragment outgoing packets based on a programmable MTU. Signed-off-by: Oleksandr Kolomeiets --- .../supported/nthw_fpga_9563_055_049_.c | 41 ++- 1 file changed, 40 insertions(+), 1 deletion(-) diff --git a/drivers/n

[PATCH v4 76/86] net/ntnic: flow update was added

2024-10-29 Thread Serhii Iliushyk
From: Danylo Vodopianov Flow action update was implemented. Signed-off-by: Danylo Vodopianov --- .../profile_inline/flow_api_profile_inline.c | 165 ++ 1 file changed, 165 insertions(+) diff --git a/drivers/net/ntnic/nthw/flow_api/profile_inline/flow_api_profile_inline.c b/

[PATCH v4 83/86] net/ntnic: add async template APIs implementation

2024-10-29 Thread Serhii Iliushyk
From: Danylo Vodopianov flow filter ops and inline API was exnteded with next APIs: 1. flow pattern template create 2. flow pattern template destroy 3. flow actions template create 4. flow actions template destroy 5. flow template table create 6. flow template table destroy Signed-off-by: Danylo

[PATCH v4 59/86] net/ntnic: add STA module

2024-10-29 Thread Serhii Iliushyk
From: Oleksandr Kolomeiets fpga map was extended with STA module support which enable statistics functionality. Signed-off-by: Oleksandr Kolomeiets --- .../supported/nthw_fpga_9563_055_049_.c | 92 ++- .../nthw/supported/nthw_fpga_mod_str_map.c| 1 + .../nthw/support

[PATCH V3 3/7] net/mlx5: add new devargs to control probe optimization

2024-10-29 Thread Minggang Li(Gavin)
From: Rongwei Liu Add a new devarg probe_opt_en to control probe optimization in PMD. By default, the value is 0 and no behavior changed. Signed-off-by: Rongwei Liu Acked-by: Viacheslav Ovsiienko --- doc/guides/nics/mlx5.rst| 7 +++ drivers/common/mlx5/linux/mlx5_nl.c

[PATCH v4 82/86] net/ntnic: add async flow create/delete API implementation

2024-10-29 Thread Serhii Iliushyk
From: Danylo Vodopianov Inline profile was extended with async flow create and delete features implementation. async & destroy API was added to the flow filter ops. Signed-off-by: Danylo Vodopianov --- drivers/net/ntnic/include/flow_api_engine.h | 36 +++ drivers/net/ntnic/nthw/flow_api/fl

[PATCH v4 47/86] net/ntnic: add RPP LR module registers

2024-10-29 Thread Serhii Iliushyk
From: Oleksandr Kolomeiets The RX Packet Process for Local Retransmit module can add bytes in the FPGA TX pipeline, which is needed when the packet increases in size. Note, this makes room for packet expansion, but the actual expansion is done by the modules. Signed-off-by: Oleksandr Kolomeiets

[PATCH v4 53/86] net/ntnic: enable RSS feature

2024-10-29 Thread Serhii Iliushyk
Enable receive side scaling Signed-off-by: Serhii Iliushyk --- v4 * Use RTE_MIN instead of the ternary operator. --- doc/guides/nics/features/ntnic.ini| 3 + drivers/net/ntnic/include/create_elements.h | 1 + drivers/net/ntnic/include/flow_api.h | 2 + drivers/net/ntnic

[PATCH v4 50/86] net/ntnic: add Tx INS module registers

2024-10-29 Thread Serhii Iliushyk
From: Oleksandr Kolomeiets The TX Inserter module injects zeros into an offset of a packet, effectively expanding the packet. Signed-off-by: Oleksandr Kolomeiets --- .../supported/nthw_fpga_9563_055_049_.c | 19 ++- .../ntnic/nthw/supported/nthw_fpga_mod_defs.h | 3 ++-

[PATCH v4 66/86] net/ntnic: add aging API to the inline profile

2024-10-29 Thread Serhii Iliushyk
From: Danylo Vodopianov Added implementation for flow get aging API. Module which operate with age queue was extended with get, count and size operations. Signed-off-by: Danylo Vodopianov --- drivers/net/ntnic/meson.build | 1 + .../flow_api/profile_inline/flm_age_queue.c |

[PATCH v4 45/86] net/ntnic: add MAC Rx module registers

2024-10-29 Thread Serhii Iliushyk
From: Oleksandr Kolomeiets The IP Fragmenter module can fragment outgoing packets based on a programmable MTU. Signed-off-by: Oleksandr Kolomeiets --- .../supported/nthw_fpga_9563_055_049_.c | 61 ++- .../ntnic/nthw/supported/nthw_fpga_mod_defs.h | 3 +- .../ntnic/nthw/s

[PATCH v4 79/86] net/ntnic: remove unnecessary type cast

2024-10-29 Thread Serhii Iliushyk
The dev_private has type void * and type casting is not necessary. Signed-off-by: Serhii Iliushyk --- .../net/ntnic/nthw/ntnic_meter/ntnic_meter.c | 18 +++ drivers/net/ntnic/ntnic_ethdev.c | 48 +-- drivers/net/ntnic/ntnic_filter/ntnic_filter.c | 16 +++ 3

[PATCH v4 86/86] net/ntnic: update documentation for set MTU

2024-10-29 Thread Serhii Iliushyk
Update ntnic.rst and release notes Signed-off-by: Serhii Iliushyk --- doc/guides/nics/ntnic.rst | 1 + doc/guides/rel_notes/release_24_11.rst | 1 + 2 files changed, 2 insertions(+) diff --git a/doc/guides/nics/ntnic.rst b/doc/guides/nics/ntnic.rst index 3794b4f216..bb927f9505 1006

[PATCH v4 75/86] net/ntnic: add flow action update

2024-10-29 Thread Serhii Iliushyk
From: Danylo Vodopianov flow_filter_ops was extended with flow action update API. Signed-off-by: Danylo Vodopianov --- drivers/net/ntnic/nthw/flow_api/flow_api.c | 16 .../profile_inline/flow_api_profile_inline.h | 5 + 2 files changed, 21 insertions(+) diff --

[PATCH v4 57/86] net/ntnic: added flm stat interface

2024-10-29 Thread Serhii Iliushyk
From: Danylo Vodopianov flm stat module interface was added. Signed-off-by: Danylo Vodopianov --- drivers/net/ntnic/include/flow_api.h | 2 ++ drivers/net/ntnic/include/flow_filter.h| 1 + drivers/net/ntnic/nthw/flow_api/flow_api.c | 11 +++ drivers/net/ntnic/ntnic_mod_reg.

[PATCH v4 74/86] net/ntnic: add action update

2024-10-29 Thread Serhii Iliushyk
From: Danylo Vodopianov rte_flow_ops was extended with action update feature. Signed-off-by: Danylo Vodopianov --- drivers/net/ntnic/ntnic_filter/ntnic_filter.c | 66 +++ drivers/net/ntnic/ntnic_mod_reg.h | 10 +++ 2 files changed, 76 insertions(+) diff --git a/dri

[PATCH v4 67/86] net/ntnic: add flow info and flow configure APIs

2024-10-29 Thread Serhii Iliushyk
From: Danylo Vodopianov Inline profile was extended with flow info and create APIS. Module which operate with age queue was extended with create and free operations. Signed-off-by: Danylo Vodopianov --- drivers/net/ntnic/include/flow_api.h | 3 + drivers/net/ntnic/include/flow_api_e

[PATCH v4 71/86] net/ntnic: add meter API

2024-10-29 Thread Serhii Iliushyk
From: Danylo Vodopianov Add meter API and implementation to the profile inline. management functions were extended with meter flow support. Signed-off-by: Danylo Vodopianov --- drivers/net/ntnic/include/flow_api.h | 1 + drivers/net/ntnic/include/flow_api_engine.h | 5 + .../fl

[PATCH v4 84/86] net/ntnic: update async flow API documentation

2024-10-29 Thread Serhii Iliushyk
Update ntnic.ini and release notes Signed-off-by: Serhii Iliushyk --- doc/guides/nics/ntnic.rst | 1 + doc/guides/rel_notes/release_24_11.rst | 1 + 2 files changed, 2 insertions(+) diff --git a/doc/guides/nics/ntnic.rst b/doc/guides/nics/ntnic.rst index e0dfbefacb..3794b4f216 1006

[PATCH v4 62/86] net/ntnic: added flow statistics

2024-10-29 Thread Serhii Iliushyk
From: Danylo Vodopianov xstats was extended with flow statistics support. Additional counters that shows learn, unlearn, lps, aps and other. Signed-off-by: Danylo Vodopianov --- .../net/ntnic/adapter/nt4ga_stat/nt4ga_stat.c | 40 drivers/net/ntnic/include/hw_mod_backend.h| 3 + dr

[PATCH v4 63/86] net/ntnic: add scrub registers

2024-10-29 Thread Serhii Iliushyk
From: Danylo Vodopianov Scrub fields were added to the fpga map file Remove duplicated macro Signed-off-by: Danylo Vodopianov --- .../supported/nthw_fpga_9563_055_049_.c | 17 - drivers/net/ntnic/ntnic_ethdev.c| 3 --- 2 files changed, 16 insertions(+),

[PATCH v4 73/86] net/ntnic: update meter documentation

2024-10-29 Thread Serhii Iliushyk
From: Danylo Vodopianov Update ntnic.ini ntnic.rst and release notes Signed-off-by: Danylo Vodopianov --- doc/guides/nics/features/ntnic.ini | 2 ++ doc/guides/nics/ntnic.rst | 1 + doc/guides/rel_notes/release_24_11.rst | 1 + 3 files changed, 4 insertions(+) diff --git a/do

[PATCH v4 55/86] net/ntnic: add rpf module

2024-10-29 Thread Serhii Iliushyk
From: Danylo Vodopianov The Receive Port FIFO module controls the small FPGA FIFO that packets are stored in before they enter the packet processor pipeline. Signed-off-by: Danylo Vodopianov --- .../net/ntnic/adapter/nt4ga_stat/nt4ga_stat.c | 25 +++- drivers/net/ntnic/include/ntnic_stat.h

[PATCH v4 54/86] net/ntnic: add statistics API

2024-10-29 Thread Serhii Iliushyk
From: Danylo Vodopianov Statistics init, setup, get, reset APIs and their implementation were added. Statistics fpga defines were added. Signed-off-by: Danylo Vodopianov --- drivers/net/ntnic/adapter/nt4ga_adapter.c | 29 +- .../net/ntnic/adapter/nt4ga_stat/nt4ga_stat.c | 192 +

[PATCH v4 65/86] net/ntnic: add flow aging API

2024-10-29 Thread Serhii Iliushyk
From: Danylo Vodopianov add flow aging API to the ops structure Signed-off-by: Danylo Vodopianov --- drivers/net/ntnic/nthw/flow_api/flow_api.c| 71 +++ drivers/net/ntnic/ntnic_filter/ntnic_filter.c | 88 +++ drivers/net/ntnic/ntnic_mod_reg.h | 21 ++

[PATCH v4 48/86] net/ntnic: add SLC LR module registers

2024-10-29 Thread Serhii Iliushyk
From: Oleksandr Kolomeiets The Slicer for Local Retransmit module can cut of the head a packet before the packet leaves the FPGA RX pipeline. This is used when the TX pipeline is configured to add a new head in the packet Signed-off-by: Oleksandr Kolomeiets --- .../supported/nthw_fpga_9563_055

[PATCH v4 72/86] net/ntnic: add meter module

2024-10-29 Thread Serhii Iliushyk
From: Danylo Vodopianov Meter module was added: 1. add/remove profile 2. create/destroy flow 3. add/remove meter policy 4. read/update stats eth_dev_ops struct was extended with ops above. Signed-off-by: Danylo Vodopianov --- drivers/net/ntnic/include/ntos_drv.h | 14 + drivers/net/

[PATCH v4 85/86] net/ntnic: add MTU configuration

2024-10-29 Thread Serhii Iliushyk
From: Oleksandr Kolomeiets Add supporting API rte_eth_dev_set_mtu Signed-off-by: Oleksandr Kolomeiets --- drivers/net/ntnic/include/flow_api_engine.h | 7 ++ drivers/net/ntnic/include/hw_mod_backend.h| 4 + .../ntnic/nthw/flow_api/hw_mod/hw_mod_tpe.c | 96 +++ .../pro

[PATCH v4 68/86] net/ntnic: add flow aging event

2024-10-29 Thread Serhii Iliushyk
From: Danylo Vodopianov Port thread was extended with new age event callback handler. LRN, INF, STA registers getter setter was added. Signed-off-by: Danylo Vodopianov --- drivers/net/ntnic/include/flow_api_engine.h | 13 ++ drivers/net/ntnic/include/hw_mod_backend.h| 11 ++ .../net/nt

[PATCH v4 49/86] net/ntnic: add Tx CPY module registers

2024-10-29 Thread Serhii Iliushyk
From: Oleksandr Kolomeiets The TX Copy module writes data to packet fields based on the lookup performed by the FLM module. This is used for NAT and can support other actions based on the RTE action MODIFY_FIELD. Signed-off-by: Oleksandr Kolomeiets --- .../supported/nthw_fpga_9563_055_049_

[PATCH v4 56/86] net/ntnic: add statistics poll

2024-10-29 Thread Serhii Iliushyk
From: Danylo Vodopianov Mechanism which poll statistics module and update values with dma module. Signed-off-by: Danylo Vodopianov --- .../net/ntnic/adapter/nt4ga_stat/nt4ga_stat.c | 343 ++ drivers/net/ntnic/include/ntdrv_4ga.h | 1 + drivers/net/ntnic/include/ntnic_

[PATCH v4 80/86] net/ntnic: add async create/destroy API declaration

2024-10-29 Thread Serhii Iliushyk
From: Danylo Vodopianov Fast path async create and destroy flow API implementation were added. Signed-off-by: Danylo Vodopianov --- drivers/net/ntnic/include/flow_api_engine.h | 8 ++ drivers/net/ntnic/ntnic_ethdev.c | 1 + drivers/net/ntnic/ntnic_filter/ntnic_filter.c | 105

[PATCH v4 77/86] net/ntnic: update documentation for flow actions update

2024-10-29 Thread Serhii Iliushyk
Update ntnic.rst and release notes Signed-off-by: Serhii Iliushyk --- doc/guides/nics/ntnic.rst | 1 + doc/guides/rel_notes/release_24_11.rst | 1 + 2 files changed, 2 insertions(+) diff --git a/doc/guides/nics/ntnic.rst b/doc/guides/nics/ntnic.rst index 4ae94b161c..e0dfbefacb 1006

[PATCH v4 81/86] net/ntnic: add async template API declaration

2024-10-29 Thread Serhii Iliushyk
From: Danylo Vodopianov rte_flow_ops was exnteded with next features support 1. flow pattern template create 2. flow pattern template destroy 3. flow actions template create 4. flow actions template destroy 5. flow template table create 6. flow template table destroy Signed-off-by: Danylo Vodopi

[PATCH v4 64/86] net/ntnic: update documentation

2024-10-29 Thread Serhii Iliushyk
From: Oleksandr Kolomeiets Update required documentation Signed-off-by: Oleksandr Kolomeiets --- doc/guides/nics/ntnic.rst | 30 ++ doc/guides/rel_notes/release_24_11.rst | 2 ++ 2 files changed, 32 insertions(+) diff --git a/doc/guides/nics/ntnic.rst b/d

[PATCH v4 70/86] net/ntnic: add aging documentation

2024-10-29 Thread Serhii Iliushyk
From: Danylo Vodopianov ntnic.rst document was exntede with age feature specification. ntnic.ini was extended with rte_flow action age support. Signed-off-by: Danylo Vodopianov --- doc/guides/nics/features/ntnic.ini | 1 + doc/guides/nics/ntnic.rst | 18 ++ do

[PATCH v4 58/86] net/ntnic: add tsm module

2024-10-29 Thread Serhii Iliushyk
From: Danylo Vodopianov tsm module which operate with timers in the physical nic was added. Necessary defines and implementation were added. The Time Stamp Module controls every aspect of packet timestamping, including time synchronization, time stamp format, PTP protocol, etc. Signed-off-by:

[PATCH v4 60/86] net/ntnic: add TSM module

2024-10-29 Thread Serhii Iliushyk
From: Oleksandr Kolomeiets fpga map was extended with tsm module support which enable statistics functionality. Signed-off-by: Oleksandr Kolomeiets --- doc/guides/nics/features/ntnic.ini| 1 + .../supported/nthw_fpga_9563_055_049_.c | 394 +- .../nthw/suppor

[PATCH v4 52/86] net/ntnic: update alignment for virt queue structs

2024-10-29 Thread Serhii Iliushyk
Update incorrect alignment Signed-off-by: Serhii Iliushyk --- v2 * Fix __rte_packed usage Original NT PMD driver use pragma pack(1) wich is similar with combination attributes packed and aligned In this case aligned(1) can be ignored in case of use attribute packed

[PATCH v4 61/86] net/ntnic: add xstats

2024-10-29 Thread Serhii Iliushyk
From: Danylo Vodopianov Extended statistics implementation and initialization were added. eth_dev_ops api was extended with new xstats apis. Signed-off-by: Danylo Vodopianov --- doc/guides/nics/features/ntnic.ini| 1 + drivers/net/ntnic/include/ntnic_stat.h| 36 + drive

[PATCH v4 69/86] net/ntnic: add termination thread

2024-10-29 Thread Serhii Iliushyk
From: Danylo Vodopianov Introduce clear_pdrv to unregister driver from global tracking. Modify drv_deinit to call clear_pdirv and ensure safe termination. Add flm sta and age event free. Signed-off-by: Danylo Vodopianov --- .../flow_api/profile_inline/flm_age_queue.c | 10 +++ .../flow_api

Re: [PATCH v1] net/ice: fix incorrect reading of PHY timestamp

2024-10-29 Thread Bruce Richardson
On Tue, Oct 29, 2024 at 04:07:45PM +0530, Singh, Aman Deep wrote: >On 25-10-2024 15:56, Soumyadeep Hore wrote: > > In ICE PMD, previously the ready bitmap checking before reading > PHY timestamp was not present. This caused incorrect Tx > timestamping. > > The ready bitmap checking is enabled

[PATCH v4 25/86] net/ntnic: add items gtp and actions raw encap/decap

2024-10-29 Thread Serhii Iliushyk
From: Danylo Vodopianov Add possibility to use * RTE_FLOW_ITEM_TYPE_GTP * RTE_FLOW_ITEM_TYPE_GTP_PSC * RTE_FLOW_ACTION_TYPE_RAW_ENCAP * RTE_FLOW_ACTION_TYPE_RAW_DECAP Signed-off-by: Danylo Vodopianov --- doc/guides/nics/features/ntnic.ini| 4 + drivers/net/ntnic/include/create_el

[PATCH v6 5/5] net/ice: provide parameter to limit scheduler layers

2024-10-29 Thread Bruce Richardson
In order to help with backward compatibility for applications, which may expect the ice driver tx scheduler (accessed via tm apis) to only have 3 layers, add in a devarg to allow the user to explicitly limit the number of scheduler layers visible to the application. Signed-off-by: Bruce Richardson

[PATCH v6 3/5] net/ice: enhance Tx scheduler hierarchy support

2024-10-29 Thread Bruce Richardson
Increase the flexibility of the Tx scheduler hierarchy support in the driver. If the HW/firmware allows it, allow creating up to 2k child nodes per scheduler node. Also expand the number of supported layers to the max available, rather than always just having 3 layers. One restriction on this chan

[PATCH v6 2/5] net/ice/base: make context alloc function non-static

2024-10-29 Thread Bruce Richardson
The function "ice_alloc_lan_q_ctx" will be needed by the driver code, so make it non-static. Signed-off-by: Bruce Richardson Acked-by: Vladimir Medvedkin --- drivers/net/ice/base/ice_sched.c | 2 +- drivers/net/ice/base/ice_sched.h | 3 +++ 2 files changed, 4 insertions(+), 1 deletion(-) diff

[PATCH v6 0/5] Improve rte_tm support in ICE driver

2024-10-29 Thread Bruce Richardson
This patchset expands the capabilities of the traffic management support in the ICE driver. It allows the driver to support different sizes of topologies, and support >256 queues and more than 3 hierarchy layers. --- v6: * remove char arithmetic in patch 1 * rework parameter checks in patch 3 to

[PATCH v6 1/5] net/ice: add option to download scheduler topology

2024-10-29 Thread Bruce Richardson
The DDP package file being loaded at init time may contain an alternative Tx Scheduler topology in it. Add driver option to load this topology at init time. Signed-off-by: Bruce Richardson --- doc/guides/nics/ice.rst| 15 +++ drivers/net/ice/base/ice_ddp.c | 20 ++

[PATCH v4 02/86] net/ntnic: add flow filter API

2024-10-29 Thread Serhii Iliushyk
Enable flow ops getter Signed-off-by: Serhii Iliushyk --- v2 * Change cast to void with __rte_unused --- drivers/net/ntnic/include/create_elements.h | 13 +++ .../ntnic/include/stream_binary_flow_api.h| 2 + drivers/net/ntnic/meson.build | 1 + drivers/net/ntnic/ntnic

Re: [PATCH v1 2/3] dts: rework testbed_model Port objects to contain unique identifiers

2024-10-29 Thread Dean Marx
On Wed, Aug 21, 2024 at 2:43 PM Nicholas Pratte wrote: > In order to leverage the usability of unique identifiers on ports, the > testbed_model Port object needs some refactoring/trimming of obsolete or > needless attributes. > > Bugzilla ID: 1478 > > Signed-off-by: Nicholas Pratte > Reviewed-b

Re: [PATCH v1 3/3] dts: rework test suite and dts runner to include test_run configs

2024-10-29 Thread Dean Marx
On Wed, Aug 21, 2024 at 2:43 PM Nicholas Pratte wrote: > The TestSuite object needs to access the running test run config in > order to identify the ports that need to be linked. To do this, > DTSRunner needs to be tweaked to accept test run configs to test suites, > and the TestSuite constructor

[PATCH v4 40/86] net/ntnic: sort FPGA registers alphanumerically

2024-10-29 Thread Serhii Iliushyk
From: Oleksandr Kolomeiets Beatification commit. It is required for pretty supporting different FPGA Signed-off-by: Oleksandr Kolomeiets --- .../supported/nthw_fpga_9563_055_049_.c | 364 +- 1 file changed, 182 insertions(+), 182 deletions(-) diff --git a/drivers/net/ntn

[PATCH v4 28/86] net/ntnic: add PDB module

2024-10-29 Thread Serhii Iliushyk
From: Danylo Vodopianov The Packet Description Builder module creates packet meta-data for example virtio-net headers. Signed-off-by: Danylo Vodopianov --- drivers/net/ntnic/include/hw_mod_backend.h| 3 + .../ntnic/nthw/flow_api/hw_mod/hw_mod_pdb.c | 144 ++ .../profile

[PATCH v4 08/86] net/ntnic: add create/destroy implementation for NT flows

2024-10-29 Thread Serhii Iliushyk
Implements flow create/destroy functions with minimal capabilities item any action port id Signed-off-by: Serhii Iliushyk --- doc/guides/nics/features/ntnic.ini| 6 + drivers/net/ntnic/include/flow_api.h | 3 + drivers/net/ntnic/include/flow_api_engine.h | 105 +++ ...

[PATCH v4 13/86] net/ntnic: add action drop

2024-10-29 Thread Serhii Iliushyk
Add possibility to use RTE_FLOW_ACTION_TYPE_DROP Signed-off-by: Serhii Iliushyk --- doc/guides/nics/features/ntnic.ini | 1 + .../profile_inline/flow_api_profile_inline.c | 12 2 files changed, 13 insertions(+) diff --git a/doc/guides/nics/features/ntnic.

Re: [PATCH v1 1/3] dts: rework port attributes in config module

2024-10-29 Thread Dean Marx
On Wed, Aug 21, 2024 at 2:43 PM Nicholas Pratte wrote: > The current design requires that a peer pci port is identified so that > test suites can create the correct port links. While this can work, it > also creates a lot of room for user error. Instead, devices should be > given a unique identif

  1   2   3   >