[PATCH 3/3] event/dlb2: optimize credit allocations

This commit implements the changes required for using suggested port type hint feature. Each port uses different credit quanta based on port type specified using port configuration flags. Each port has separate quanta defined in dlb2_priv.h Producer and consumer ports will need larger quanta value

[PATCH 2/3] event/dlb2: add fence bypass option for producer ports

If producer thread is only acting as a bridge between NIC and DLB, then performance can be greatly improved by bypassing the fence instrucntion. DLB enqueue API calls memory fence once per enqueue burst. If prodcuer thread is just reading from NIC and sending to DLB without updating the read buffe

[PATCH 1/3] event/dlb2: add producer port probing optimization

For best performance, applications running on certain cores should use the DLB device locally available on the same tile along with other resources. To allocate optimal resources, probing is done for each producer port (PP) for a given CPU and the best performing ports are allocated to producers. T

[PATCH 0/3] DLB2 Performance Optimizations

This patchset contains performance optimizations for the DLB2 eventdev PMD. The port probing patch is dependent on patch 115303 "eal: make eal_parse_coremask external", since the application can pass in a coremask to aid in discovering the best performing port/core combinations. The fence bypass

[PATCH] eal: make eal_parse_coremask external

DLB2 has a need to parse a user supplied coremask as part of an optimization that associates optimal core/resource pairs. Therefore eal_parse_coremask has been renamed to rte_eal_parse_coremask and exported. Signed-off-by: Timothy McDaniel --- lib/eal/common/eal_common_options.c | 6 +++--- lib

[PATCH V2 6/6] pipeline: configure the hash function for learner tables

Make the hash function configurable for the learner pipeline tables. Signed-off-by: Cristian Dumitrescu Signed-off-by: Kamalakannan R. --- lib/pipeline/rte_swx_pipeline.c | 78 +--- lib/pipeline/rte_swx_pipeline.h | 6 ++ lib/pipeline/rte_swx_pipeline_inte

[PATCH V2 5/6] table: configure the hash function for learner tables

Make the hash function configurable. The internal hash function that was not configurable, mask-based and limited to 64 bytes is removed. Signed-off-by: Cristian Dumitrescu Signed-off-by: Kamalakannan R. --- lib/table/rte_swx_table_learner.c | 220 -- lib/table/rte_s

[PATCH V2 4/6] pipeline: configure the hash function for regular tables

Make the hash function configurable for the regular pipeline tables. Signed-off-by: Cristian Dumitrescu Signed-off-by: Kamalakannan R. --- lib/pipeline/rte_swx_ctl.c | 1 + lib/pipeline/rte_swx_ctl.h | 3 ++ lib/pipeline/rte_swx_pipeline.c | 56 +++

[PATCH V2 3/6] table: configure the hash function for regular tables

Make the hash function configurable. The internal hash function that was not configurable, mask-based and limited to 64 bytes is removed. Signed-off-by: Cristian Dumitrescu Signed-off-by: Kamalakannan R. --- lib/table/rte_swx_table.h| 8 ++ lib/table/rte_swx_table_em.c | 266 ++---

[PATCH V2 2/6] table: add key comparison functions

Add key comparison functions to be used by the exact match and the learner table types as part of the performance critical lookup operation. Since the key size is fixed, it is possible to select a specialized memory copy function as opposed to using the variable size version, resulting in a perform

[PATCH V2 1/6] table: add hash function prototype

Add hash function prototype to be used by the exact match and the learner table types. The hash function is not mask-based, so the table key fields have to be contiguous in memory. Signed-off-by: Cristian Dumitrescu Signed-off-by: Kamalakannan R. --- lib/table/meson.build | 1 + lib/ta

[PATCH V2 0/6] pipeline: make the hash function configurable per table

The exact match and learner tables use a hash function for the lookup operation. This patch set makes the hash function configurable and removes some limitations. The hash function previously used by these table types had the following limitations: a) Not configurable: An internally hardcoded vers

[PATCH v2 37/37] baseband/acc100: reset pointer after rte_free

Set local pointer to NULL after rte_free. This needs to be set explicitly since logic may check for null pointers. Fixes: 060e7672930 ("baseband/acc100: add queue configuration") Cc: sta...@dpdk.org Signed-off-by: Hernan Vargas --- drivers/baseband/acc100/rte_acc100_pmd.c | 3 +++ 1 file change

[PATCH v2 36/37] baseband/acc100: make HARQ layout memory 4GB

HARQ layout memory should be 4GB instead of 2GB. Fixes: 4cf90079797 ("baseband/acc100: add HW register definitions") Cc: sta...@dpdk.org Signed-off-by: Hernan Vargas --- drivers/baseband/acc100/acc100_pmd.h | 6 -- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/baseba

[PATCH v2 35/37] baseband/acc100: add protection for NULL HARQ input

It is possible to cause an invalid HW operation in case the user provides the BBDEV API and HARQ operation with input enabled and zero input. Adding protection for that case. Signed-off-by: Hernan Vargas --- drivers/baseband/acc100/rte_acc100_pmd.c | 8 1 file changed, 8 insertions(+)

[PATCH v2 34/37] baseband/acc100: update meson file sdk dependency

Update meson files with FlexRAN SDK dependency. Signed-off-by: Hernan Vargas --- drivers/baseband/acc100/meson.build | 21 + 1 file changed, 21 insertions(+) diff --git a/drivers/baseband/acc100/meson.build b/drivers/baseband/acc100/meson.build index 9a1a3b8b07..3b934a25ca

[PATCH v2 33/37] baseband/acc100: set device min alignment to 1

Historical mistakes, there should be no 64B alignment requirement for the buffer being processed. Any 1B alignment is sufficient. Fixes: 9200ffa5cd5 ("baseband/acc100: add info get function") Cc: sta...@dpdk.org Signed-off-by: Hernan Vargas --- drivers/baseband/acc100/rte_acc100_pmd.c | 2 +- 1

[PATCH v2 32/37] baseband/acc100: fix debug print for LDPC FCW

Print full size of FCW LDPC structure on debug messages. This is just a cosmetic fix, no need to fix on previous code base. Signed-off-by: Hernan Vargas --- drivers/baseband/acc100/rte_acc100_pmd.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/baseband/acc100/rte_ac

[PATCH v2 31/37] baseband/acc100: fix clearing PF IR outside handler

Clearing of PF info ring outside of handler may cause interrupt to be missed. A condition in the ACC100 PMD implementation may cause an interrupt functional handler call to be missed due to related bit being cleared when checking PF info ring status. Fixes: 06531464151 ("baseband/acc100: support i

[PATCH v2 30/37] baseband/acc100: reduce input length for CRC24B

Input length should be reduced only for CRC24B. Signed-off-by: Hernan Vargas --- drivers/baseband/acc100/rte_acc100_pmd.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/drivers/baseband/acc100/rte_acc100_pmd.c b/drivers/baseband/acc100/rte_acc100_pmd.c index 4ce4c9d218..7

[PATCH v2 29/37] baseband/acc100: update device info

Remove unused capabilities, use dummy operation as start count for number of queues. Signed-off-by: Hernan Vargas --- drivers/baseband/acc100/rte_acc100_pmd.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/drivers/baseband/acc100/rte_acc100_pmd.c b/drivers/baseband/acc

[PATCH v2 28/37] baseband/acc100: make desc optimization optional

Add ACC100_DESC_OPTIMIZATION flag to enable muxing of encode operations with common FCW. Signed-off-by: Hernan Vargas --- drivers/baseband/acc100/rte_acc100_pmd.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/baseband/acc100/rte_acc100_pmd.c b/drivers/baseband/ac

[PATCH v2 27/37] baseband/acc100: store FCW from first CB descriptor

Store the descriptor from the first code block from a transport block. Copy the LDPC FCW from the first descriptor into the rest of the CBs in that TB. Signed-off-by: Hernan Vargas --- drivers/baseband/acc100/rte_acc100_pmd.c | 4 1 file changed, 4 insertions(+) diff --git a/drivers/baseba

[PATCH v2 26/37] baseband/acc100: allocate ring/queue mem when NULL

Allocate info ring, tail pointers and HARQ layout memory for a device only if it hasn't already been allocated. Signed-off-by: Hernan Vargas --- drivers/baseband/acc100/rte_acc100_pmd.c | 9 ++--- 1 file changed, 6 insertions(+), 3 deletions(-) diff --git a/drivers/baseband/acc100/rte_acc10

[PATCH v2 25/37] baseband/acc100: update log messages

Add extra values for some log messages. No functional impact. Signed-off-by: Hernan Vargas --- drivers/baseband/acc100/rte_acc100_pmd.c | 14 +- 1 file changed, 9 insertions(+), 5 deletions(-) diff --git a/drivers/baseband/acc100/rte_acc100_pmd.c b/drivers/baseband/acc100/rte_acc10

[PATCH v2 24/37] baseband/acc100: rename ldpc encode function arg

Rename total_enqueued_cbs to total_enqueued_descs in the enqueue_ldpc_enc_n_op_cb function. No functional impact. Signed-off-by: Hernan Vargas --- drivers/baseband/acc100/rte_acc100_pmd.c | 10 ++ 1 file changed, 6 insertions(+), 4 deletions(-) diff --git a/drivers/baseband/acc100/rte_a

[PATCH v2 23/37] baseband/acc100: update uplink CB input length

Use the FCW E parameter for rate matching as the code block input length. Signed-off-by: Hernan Vargas --- drivers/baseband/acc100/rte_acc100_pmd.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/baseband/acc100/rte_acc100_pmd.c b/drivers/baseband/acc100/rte_acc100_p

[PATCH v2 22/37] baseband/acc100: enforce additional check on FCW

Enforce additional check on Frame Control Word validity and add stronger alignment for decompression mode. Signed-off-by: Hernan Vargas --- drivers/baseband/acc100/rte_acc100_pmd.c | 40 ++-- 1 file changed, 37 insertions(+), 3 deletions(-) diff --git a/drivers/baseband/acc1

[PATCH v2 21/37] baseband/acc100: check for unlikely operation vals

Add unlikely checks for NULL operation values. Signed-off-by: Hernan Vargas --- drivers/baseband/acc100/rte_acc100_pmd.c | 10 ++ 1 file changed, 10 insertions(+) diff --git a/drivers/baseband/acc100/rte_acc100_pmd.c b/drivers/baseband/acc100/rte_acc100_pmd.c index 70a29f92a1..ea850e2d

[PATCH v2 20/37] baseband/acc100: check turbo dec/enc input

Add NULL check for the turbo decoder and encoder input length. Signed-off-by: Hernan Vargas --- drivers/baseband/acc100/rte_acc100_pmd.c | 23 ++- 1 file changed, 18 insertions(+), 5 deletions(-) diff --git a/drivers/baseband/acc100/rte_acc100_pmd.c b/drivers/baseband/acc10

[PATCH v2 19/37] baseband/acc100: add queue stop operation

Implement new feature queue stop operation. Signed-off-by: Hernan Vargas --- drivers/baseband/acc100/rte_acc100_pmd.c | 58 1 file changed, 58 insertions(+) diff --git a/drivers/baseband/acc100/rte_acc100_pmd.c b/drivers/baseband/acc100/rte_acc100_pmd.c index 460233a49

[PATCH v2 18/37] baseband/acc100: implement configurable queue depth

Implement new feature to make queue depth configurable based on decode or encode mode. Signed-off-by: Hernan Vargas --- drivers/baseband/acc100/rte_acc100_pmd.c | 12 +--- 1 file changed, 9 insertions(+), 3 deletions(-) diff --git a/drivers/baseband/acc100/rte_acc100_pmd.c b/drivers/ba

[PATCH v2 17/37] baseband/acc100: configure PMON control registers

Adding back feature from ACC101. Expose the device status and add protection for corner cases. Enable performance monitor control registers. Signed-off-by: Hernan Vargas --- drivers/baseband/acc100/acc100_pf_enum.h | 52 +--- drivers/baseband/acc100/acc100_pmd.h | 9 +++-

[PATCH v2 16/37] baseband/acc100: add ring companion address

Store the virtual address of companion ring as part of queue information. Use this address to calculate the op address. Signed-off-by: Hernan Vargas --- drivers/baseband/acc100/acc100_pmd.h | 12 ++ drivers/baseband/acc100/rte_acc100_pmd.c | 143 ++- 2 files changed, 100

[PATCH v2 15/37] baseband/acc100: add workaround for deRM corner cases

Add function to asses if de-ratematch pre-processing should be run in SW for corner cases. Signed-off-by: Hernan Vargas --- drivers/baseband/acc100/acc100_pmd.h | 13 +++ drivers/baseband/acc100/rte_acc100_pmd.c | 103 ++- 2 files changed, 114 insertions(+), 2 deletions(

[PATCH v2 14/37] baseband/acc100: update validate LDPC enc/dec

Update validate functions to check for valid LDPC parameters to avoid any HW issues. Adding protection for null corner case and for HARQ inbound size out of range. HARQ input size from application may be invalid and causing HW issue. Add checks to ensure that if HARQ is invalid, set to some valid s

[PATCH v2 13/37] baseband/acc10x: limit cases for HARQ pruning

Add flag ACC101_HARQ_PRUNING_OPTIMIZATION to limit cases when HARQ pruning is valid. Signed-off-by: Hernan Vargas --- drivers/baseband/acc100/rte_acc100_pmd.c | 52 +++- 1 file changed, 41 insertions(+), 11 deletions(-) diff --git a/drivers/baseband/acc100/rte_acc100_pmd.c

[PATCH v2 12/37] baseband/acc100: add LDPC transport block support

Add LDPC enqueue functions to handle transport blocks. Signed-off-by: Hernan Vargas --- drivers/baseband/acc100/rte_acc100_pmd.c | 197 ++- 1 file changed, 195 insertions(+), 2 deletions(-) diff --git a/drivers/baseband/acc100/rte_acc100_pmd.c b/drivers/baseband/acc100/rte_

[PATCH v2 11/37] baseband/acc100: separate validation functions from debug

Use new flag RTE_LIBRTE_BBDEV_SKIP_VALIDATE enable/disable validation functions. The validation API will be enabled by default. Signed-off-by: Hernan Vargas --- drivers/baseband/acc100/rte_acc100_pmd.c | 37 1 file changed, 19 insertions(+), 18 deletions(-) diff --git a

[PATCH v2 10/37] baseband/acc100: avoid mux for small inbound frames

Update check_mux to avoid multiplexing small inbound frames. Preventing to multiplex code blocks when K < 512B per specs. Signed-off-by: Hernan Vargas --- drivers/baseband/acc100/acc100_pmd.h | 1 + drivers/baseband/acc100/rte_acc100_pmd.c | 16 +++- 2 files changed, 12 insertio

[PATCH v2 09/37] baseband/acc100: add HARQ index helper function

Refactor code to use the HARQ index helper function and make harq_idx uint32. Signed-off-by: Hernan Vargas --- drivers/baseband/acc100/rte_acc100_pmd.c | 32 +++- 1 file changed, 14 insertions(+), 18 deletions(-) diff --git a/drivers/baseband/acc100/rte_acc100_pmd.c b/drive

[PATCH v2 08/37] baseband/acc100: add scatter-gather support

Add flag to support scatter-gather for the mbuf Signed-off-by: Hernan Vargas --- drivers/baseband/acc100/rte_acc100_pmd.c | 45 1 file changed, 31 insertions(+), 14 deletions(-) diff --git a/drivers/baseband/acc100/rte_acc100_pmd.c b/drivers/baseband/acc100/rte_acc100_

[PATCH v2 07/37] baseband/acc100: add LDPC encoder padding function

LDPC Encoder input may need to be padded to avoid small beat for ACC100. Padding 5GDL input buffer length (BLEN) to avoid case (BLEN % 64) <= 8. Adding protection for corner case to avoid for 5GDL occurrence of last beat within the ACC100 fabric with <= 8B which might trigger a fabric corner case h

[PATCH v2 06/37] baseband/acc100: add default e value for FCW

Refactor frame control word LDPC encoder fill function to take a default e value as a parameter. No functional impact. Signed-off-by: Hernan Vargas --- drivers/baseband/acc100/rte_acc100_pmd.c | 8 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/baseband/acc100/rte

[PATCH v2 05/37] baseband/acc100: memory leak fix

Move check for undefined device before allocating queue data structure. Coverity issue: 375803, 375813, 375819, 375827, 375831 Fixes: 060e7672930 ("baseband/acc100: add queue configuration") Cc: sta...@dpdk.org Signed-off-by: Hernan Vargas --- drivers/baseband/acc100/rte_acc100_pmd.c | 8 --

[PATCH v2 04/37] baseband/acc100: free SW ring mem for reconfiguration

Free base address of unaligned memory for SW rings to manage the missed corner case when there is a reconfiguration. Fixes: 060e7672930 ("baseband/acc100: add queue configuration") Cc: sta...@dpdk.org Signed-off-by: Hernan Vargas --- drivers/baseband/acc100/rte_acc100_pmd.c | 6 +- 1 file c

[PATCH v2 03/37] baseband/acc100: add function to check AQ availability

It is possible for some corner case to run more batch enqueue than supported. A protection is required to avoid that corner case. Enhance all ACC100 enqueue operations with check to see if there is room in the atomic queue for enqueueing batches into the queue manager Check room in AQ for the enque

[PATCH v2 02/37] baseband/acc100: update ring availability calculation

Refactor of the queue availability computation to prevent the application to dequeue more than what may have been enqueued. Signed-off-by: Hernan Vargas --- drivers/baseband/acc100/rte_acc100_pmd.c | 39 1 file changed, 27 insertions(+), 12 deletions(-) diff --git a/dri

[PATCH v2 01/37] baseband/acc100: add enqueue status

Add enqueue status as part of rte_bbdev_queue_data. This is a new feature to update queue status and indicate the reason why a previous enqueue may or may not have consumed all requested operations. Signed-off-by: Hernan Vargas --- drivers/baseband/acc100/rte_acc100_pmd.c | 85 ++

[PATCH v2 00/37] baseband/acc100: changes for 22.11

Upstreaming ACC100 changes for 22.11. This patch series is dependant on series: https://patches.dpdk.org/project/dpdk/patch/1657150110-69957 Hernan Vargas (37): baseband/acc100: add enqueue status baseband/acc100: update ring availability calculation baseband/acc100: add function to check AQ

Re: [dpdk-kmods v2] windows/netuio: fix BAR parsing

On 8/11/2022 3:17 PM, Pallavi Kadam wrote: Current code was always checking the 'prev_bar & PCI_TYPE_64BIT' though only the first BAR slot of a 64-bit BAR contains flags. Also for certain PCIe devices, BAR values were not continuous. This patch fixes this incorrectness and maps the BAR addresses

[RFC v3] non-temporal memcpy

This RFC proposes a set of functions optimized for non-temporal memory copy. At this stage, I am asking for acceptance of the concept and API. Feedback on the x86 implementation is also welcome. Applications sometimes copy data to another memory location, which is only used much later. In this ca

Re: [PATCH v2 1/3] net: rename octeon ep PMD

On Tue, Aug 16, 2022 at 11:05 AM Veerasenareddy Burru wrote: > > > > > -Original Message- > > From: Sathesh Edara > > Sent: Wednesday, July 27, 2022 2:22 AM > > To: Satananda Burla ; Jerin Jacob Kollanukkaran > > ; Sathesh B Edara ; Thomas > > Monjalon ; Radha Chintakuntla > > ; Veerasena

[PATCH v1] examples/fips_validation: add parsing for AES GMAC

Added functionality to parse algorithm for AES GMAC test Signed-off-by: Brian Dooley --- examples/fips_validation/fips_validation.c | 2 ++ examples/fips_validation/fips_validation.h | 1 + examples/fips_validation/main.c| 6 ++ 3 files changed, 9 insertions(+) diff --git a/exam

[Bug 1042] [dpdk-22.07](ABI) unit_tests_eal/link_bonding_rssconf: link_bonding_rssconf_autotest test failed

https://bugs.dpdk.org/show_bug.cgi?id=1042 lingwei (weix.l...@intel.com) changed: What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution|---

[PATCH v1] examples/fips_validation: add parsing for AES CTR

Added functionality to parse algorithm for AES CTR test Signed-off-by: Brian Dooley --- examples/fips_validation/fips_validation.c | 2 ++ examples/fips_validation/fips_validation.h | 2 ++ examples/fips_validation/fips_validation_aes.c | 5 + examples/fips_validation/main.c

RE: [Bug 1069] Testpmd parameter --enable-lro fails on Mellanox ConnectX-5

Hello, can you please specify the OFED and FW version you use? Regards, Asaf Penso >-Original Message- >From: bugzi...@dpdk.org >Sent: Friday, August 19, 2022 10:50 AM >To: dev@dpdk.org >Subject: [Bug 1069] Testpmd parameter --enable-lro fails on Mellanox >ConnectX-5 > >https://bugs.dpdk

[Bug 1001] [meson test] Debug-tests/dump_* all meson test time out because commands are not registered to command list

https://bugs.dpdk.org/show_bug.cgi?id=1001 liweiyuan (weiyuanx...@intel.com) changed: What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution|-

[Bug 1069] Testpmd parameter --enable-lro fails on Mellanox ConnectX-5

https://bugs.dpdk.org/show_bug.cgi?id=1069 Bug ID: 1069 Summary: Testpmd parameter --enable-lro fails on Mellanox ConnectX-5 Product: DPDK Version: 22.03 Hardware: All OS: Linux Status: UNCONFIRM

RE: [PATCH 0/1] Add security perf application

Hi Akhil, Please see inline. Thanks, Anoob > > Hi Anoob, > > Subject: [PATCH 0/1] Add security perf application > > > > Add performance application to test security session create & destroy > > rates supported by the security enabled cryptodev PMD. The application > > would create specified num

Re: [PATCH] event/cnxk: move crypto adapter ops to respective files

On Wed, Jul 27, 2022 at 12:46 PM Shijith Thotton wrote: > > Moved the common crypto adapter ops to file specific to eventdev > adapters. > > Signed-off-by: Shijith Thotton Applied to dpdk-next-net-eventdev/for-main. Thanks > --- > drivers/event/cnxk/cnxk_eventdev.c | 121 --