re strings, which will
cause find_device not to find any device.
This fix canonicalize the user provided string before it is being
passed to find_device.
Signed-off-by: Shani Peretz
---
app/test-pmd/meson.build | 2 +-
app/test-pmd/testpmd.c | 30 ++
2 files c
cmp_dev_name to perform a smarter comparison
instead of a direct string match to find the device.
Shani Peretz (1):
app/testpmd: canonicalize short PCI name format
app/test-pmd/meson.build | 2 +-
app/test-pmd/testpmd.c | 30 ++
2 files changed, 31 insertions(+), 1
the device.
Shani Peretz (1):
app/testpmd: canonicalize short PCI name format
app/test-pmd/testpmd.c | 31 ++-
1 file changed, 30 insertions(+), 1 deletion(-)
--
2.34.1
are stings, which will
cause find_device not to find any device.
This fix canonicalize the user provided string before it is being
passed to find_device.
Signed-off-by: Shani Peretz
---
app/test-pmd/testpmd.c | 31 ++-
1 file changed, 30 insertions(+), 1 deletion(-)
di
> -Original Message-
> From: Morten Brørup
> Sent: Thursday, 19 June 2025 15:57
> To: Stephen Hemminger ; Shani Peretz
>
> Cc: dev@dpdk.org
> Subject: RE: [RFC PATCH 0/5] Introduce mempool object new debug
> capabilities
>
> External email: Use cautio
> -Original Message-
> From: Stephen Hemminger
> Sent: Monday, 16 June 2025 18:30
> To: Shani Peretz
> Cc: dev@dpdk.org
> Subject: Re: [RFC PATCH 0/5] Introduce mempool object new debug
> capabilities
>
> External email: Use caution opening links or attachm
decreasing the maximum inline data size to fit within
acceptable limits.
Fixes: 0c2f7837c673 ("net/mlx5: mitigate Tx queue parameter adjustment")
Signed-off-by: Shani Peretz
Acked-by: Viacheslav Ovsiienko
---
drivers/net/mlx5/mlx5_txq.c | 2 ++
1 file changed, 2 insertions(+)
di
Granularity rule matching snippet
Signed-off-by: Shani Peretz
---
examples/flow_filtering/jump_flow.c | 40 ++
examples/flow_filtering/jump_flow.h | 13
examples/flow_filtering/meson.build | 4 +
.../snippets/snippet_match_nsh.c | 74
This snippet demonstrates port affinity matching in bonded ports,
ensuring packets are sent back through their ingress port.
The NR_QUEUES constant was moved to common.h to support mapping
four TX queues (0-3) to two port affinities (0,1→1 and 2,3→2).
Signed-off-by: Shani Peretz
---
examples
specific flow groups,
with three new snippets demonstrating this feature.
Shani Peretz (2):
examples/flow_filtering: add match port affinity snippet
examples/flow_filtering: add jump flow
examples/flow_filtering/common.h | 2 +
examples/flow_filtering/flow_skeleton.c | 2
heir meson file
to be built when RTE_MEMPOOL_DEBUG_OBJECTS_HISTORY is enabled.
Signed-off-by: Shani Peretz
---
config/meson.build | 1 +
drivers/meson.build | 7 +++
drivers/net/af_packet/meson.build | 1 +
drivers/net/af_xdp/meson.build | 1 +
drive
dumps the mempool object history to console or to a file.
The dump will contain:
- Operation history for each mempool object
- Summary and statistics about all mempool objects
testpmd> dump_mempool_objects_history
testpmd> dump_mempool_objects_history
Signed-off-by: Shani Peretz
---
ap
-off-by: Shani Peretz
---
.../dpdk-mempool_object_history_parser.py | 129 ++
1 file changed, 129 insertions(+)
create mode 100755 usertools/dpdk-mempool_object_history_parser.py
diff --git a/usertools/dpdk-mempool_object_history_parser.py
b/usertools/dpdk
record operations on mempool objects when it is allocated
and released inside the mlx5 PMD.
Signed-off-by: Shani Peretz
---
drivers/net/mlx5/mlx5_rx.c | 9 +
drivers/net/mlx5/mlx5_rx.h | 2 ++
drivers/net/mlx5/mlx5_rxq.c | 9 +++--
drivers/net/mlx5/mlx5_rxtx_vec.c
mempool object header that records
the most recent operations.
Each operation is represented by a 4-bit value, and for each mempool
object the history field stores 16 steps.
Signed-off-by: Shani Peretz
---
lib/ethdev/rte_ethdev.h | 14 +
lib/mempool/rte_mempool.c | 111
ython script that can parse, analyze the data and
present it in an human readable format.
5. Added compilation flag to enable the feature.
Shani Peretz (5):
mempool: record mempool objects operations history
drivers: add mempool history compilation flag
net/mlx5: mark an operation in me
Signed-off-by: Shani Peretz
---
doc/guides/sample_app_ug/flow_filtering.rst | 19 +++-
examples/flow_filtering/common.h | 12 ++-
examples/flow_filtering/flow_skeleton.c | 28 ++---
examples/flow_filtering/main.c| 31 --
examples/flow_filtering/meson.build
> -Original Message-
> From: Shani Peretz
> Sent: Wednesday, 26 March 2025 9:14
> To: Akhil Goyal ; dev@dpdk.org
> Cc: Suanming Mou ; sta...@dpdk.org; Brian
> Dooley ; Pablo de Lara
>
> Subject: RE: [EXTERNAL] [PATCH] app/crypto-perf: fix aad offset alignment
it in devargs.name but then we will still need another function
that converts the short name to its canonical form.
> -Original Message-
> From: Stephen Hemminger
> Sent: Wednesday, 30 April 2025 19:00
> To: Shani Peretz
> Cc: dev@dpdk.org; Tyler Retzlaff ; Parav Pandit
&g
/document/d/1LmMlJ31P1G77K0TGkBfXWz0DEj6zdprP0bG5p_wu77w/edit?usp=sharing
> -Original Message-
> From: Stephen Hemminger
> Sent: Monday, 17 March 2025 16:11
> To: Shani Peretz
> Cc: dev@dpdk.org; Tyler Retzlaff ; Parav Pandit
> ; Xueming Li ; Nipun Gupta
> ; Nikhil Agarw
> -Original Message-
> From: Tanzeel Ahmed
> Sent: Saturday, 29 March 2025 22:54
> To: Ori Kam
> Cc: dev@dpdk.org; Tanzeel Ahmed
> Subject: [PATCH] examples/flow_filtering: fix make clean
>
> External email: Use caution opening links or attachments
>
>
> make clean is unable to dele
> -Original Message-
> From: Akhil Goyal
> Sent: Monday, 17 March 2025 12:23
> To: Shani Peretz ; dev@dpdk.org
> Cc: Suanming Mou ; sta...@dpdk.org; Brian
> Dooley ; Pablo de Lara
>
> Subject: RE: [EXTERNAL] [PATCH] app/crypto-perf: fix aad offset alignment
The bitmap goal is to prevent double allocations and deallocations in
per core cache mode.
This validation occurs only in debug mode, ensuring it doesn't
impact performance.
Signed-off-by: Shani Peretz
Acked-by: Bing Zhao
---
drivers/net/mlx5/mlx5_utils.c
Enhancing ipool debug capabilities by introducing new ipool log
component.
Also adding various logs in different verbosities for ipool operations.
Signed-off-by: Shani Peretz
Acked-by: Bing Zhao
---
drivers/net/mlx5/mlx5_utils.c | 50 ++-
drivers/net/mlx5
Enhanced ipool debugging: Added new log component and verbosity levels for
operations.
Introduced a bitmap in debug mode to track allocations/deallocations,
preventing doubles in per-core cache mode.
Shani Peretz (2):
net/mlx5: add ipool debug capabilities
net/mlx5: added a bitmap that
.@dpdk.org
Signed-off-by: Shani Peretz
---
app/test-crypto-perf/cperf_ops.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/app/test-crypto-perf/cperf_ops.c b/app/test-crypto-perf/cperf_ops.c
index 6d5f510220..f9be51e17f 100644
--- a/app/test-crypto-perf/cperf_ops.c
+++ b/app/t
> -Original Message-
> From: Thomas Monjalon
> Sent: Wednesday, 19 February 2025 17:20
> To: Shani Peretz
> Cc: dev@dpdk.org; ferruh.yi...@amd.com; Ori Kam ;
> Stephen Hemminger
> Subject: Re: [PATCH] examples/flow_filtering: fix destination IP mask
>
>
> -Original Message-
> From: Stephen Hemminger
> Sent: Thursday, 20 February 2025 20:33
> To: Shani Peretz
> Cc: dev@dpdk.org; Tyler Retzlaff ; Parav Pandit
> ; Xueming Li ; Nipun Gupta
> ; Nikhil Agarwal ; Hemant
> Agrawal ; Sachin Saxena
> ; Rosen Xu ; Chen
, void *addr, int addr_size, int *out_size)
So I now use it in rte_strscpy.
In addition, should I replace the call to rte_strscpy with strlcpy?
> -Original Message-
> From: Stephen Hemminger
> Sent: Tuesday, 11 February 2025 20:05
> To: Bruce Richardson
> Cc: Shani Peretz
Added tests to PCI and vdev bus that compare and find devices
in various forms of names under test_devargs.
The tests check the functionality of the rte_cmp_dev_name function by
calling it with different variations of names, like short and full,
full and full, etc.
Signed-off-by: Shani Peretz
gt;name,
so when a new PCI device is registering the name displayed in the device
list will be the parsed version.
Fixes: 23eaa9059ec2 ("bus/pci: use given name as generic name")
Cc: sta...@dpdk.org
Signed-off-by: Shani Peretz
---
drivers/bus/pci/pci_common.c | 14 ++
1 file ch
Change find_device and rte_cmp_dev_name to accept a struct
with pointer and size instead of a void* pointer.
Signed-off-by: Shani Peretz
---
app/test/test_devargs.c | 23 +++
app/test/test_vdev.c | 23 ++-
drivers/bus
al: add hotplug add/remove device")
Signed-off-by: Shani Peretz
---
app/test/test_devargs.c | 2 +-
app/test/test_vdev.c | 10 ++
drivers/bus/auxiliary/auxiliary_common.c | 17 +++---
drivers/bus/cdx/cdx.c| 13 +---
drivers
Change find_device and rte_cmp_dev_name to accept a struct
with pointer and size instead of a void* pointer.
Signed-off-by: Shani Peretz
---
app/test/test_devargs.c | 23 +++
app/test/test_vdev.c | 23 ++-
drivers/bus
Added tests to PCI and vdev bus that compare and find devices
in various forms of names under test_devargs.
The tests check the functionality of the rte_cmp_dev_name function by
calling it with different variations of names, like short and full,
full and full, etc.
Signed-off-by: Shani Peretz
ed address.
This will allow consistent comparisons between different representations
of same devices.
In addition, fixed vdev test to use the rte_cmp_dev_name function
instead of the custom one.
Fixes: a3ee360f4440 ("eal: add hotplug add/remove device")
Signed-off-by: Shani Peretz
---
app/tes
gt;name,
so when a new PCI device is registering the name displayed in the device
list will be the parsed version.
Fixes: 23eaa9059ec2 ("bus/pci: use given name as generic name")
Cc: sta...@dpdk.org
Signed-off-by: Shani Peretz
---
drivers/bus/pci/pci_common.c | 14 ++
1 file ch
Change find_device and rte_cmp_dev_name to accept a struct
with pointer and size instead of a void* pointer.
Signed-off-by: Shani Peretz
---
app/test/test_devargs.c | 23 +++
app/test/test_vdev.c | 23 ++-
drivers/bus
ed address.
This will allow consistent comparisons between different representations
of same devices.
In addition, fixed vdev test to use the rte_cmp_dev_name function
instead of the custom one.
Fixes: a3ee360f4440 ("eal: add hotplug add/remove device")
Signed-off-by: Shani Peretz
---
app/tes
Added tests to PCI and vdev bus that compare and find devices
in various forms of names under test_devargs.
The tests check the functionality of the rte_cmp_dev_name function by
calling it with different variations of names, like short and full,
full and full, etc.
Signed-off-by: Shani Peretz
gt;name,
so when a new PCI device is registering the name displayed in the device
list will be the parsed version.
Fixes: 23eaa9059ec2 ("bus/pci: use given name as generic name")
Cc: sta...@dpdk.org
Signed-off-by: Shani Peretz
---
drivers/bus/pci/pci_common.c | 14 ++
1 file ch
This series fixes two issues:
1. saving PCI name as the parsed one instead of user-provided string when
registering a new device
2. fixes rte_cmp_dev_name function that currently compares the names by
compering 2 strings
Also added test to check the functionality of rte_cmp_dev_name.
Shani
> -Original Message-
> From: Stephen Hemminger
> Sent: Wednesday, 5 February 2025 20:46
> To: Shani Peretz
> Cc: dev@dpdk.org; NBU-Contact-Thomas Monjalon (EXTERNAL)
> ; Tyler Retzlaff ; Parav
> Pandit ; Xueming Li ; Nipun Gupta
> ; Nikhil Agarwal ; Hemant
&
> -Original Message-
> From: Stephen Hemminger
> Sent: Wednesday, 5 February 2025 18:42
> To: Shani Peretz
> Cc: dev@dpdk.org; NBU-Contact-Thomas Monjalon (EXTERNAL)
> ; Tyler Retzlaff ; Parav
> Pandit ; Xueming Li ; Nipun Gupta
> ; Nikhil Agarwal ; Hemant
&
> -Original Message-
> From: Stephen Hemminger
> Sent: Wednesday, 29 January 2025 18:25
> To: Shani Peretz
> Cc: dev@dpdk.org; NBU-Contact-Thomas Monjalon (EXTERNAL)
> ; Tyler Retzlaff ; Parav
> Pandit ; Xueming Li ; Nipun Gupta
> ; Nikhil Agarwal ; Hemant
&
test_devargs.
Fixes: a3ee360f4440 ("eal: add hotplug add/remove device")
Signed-off-by: Shani Peretz
---
app/test/test_devargs.c | 123 ++-
app/test/test_vdev.c | 10 +-
drivers/bus/auxiliary/auxiliary_common.c | 17 +++-
irpin queues
- hairpin_out_of_buffer_rxq* (queue-level)
counts drops of a specific queue
Both the port-level and per-queue counters can be enabled,
disabled, and queried.
Signed-off-by: Shani Peretz
---
drivers/common/mlx5/mlx5_devx_cmds.c| 9 +-
drivers/common/mlx5/mlx5_devx_cmds.h
Added the following testpmd subcommands:
toggle the counter on and off
> port (port_id) enable|disable
query the state of counters:
> set xstats-hide-disabled on|off
Signed-off-by: Shani Peretz
---
app/test-pmd/cmdline.c | 125 +++
app/te
those provided by
eth_basic_stats_get()) will be considered unsupported and
will not be toggleable.
Also all xstats will be considered unsupported for dynamic enable/disable,
and each PMD will be able to override this in its implementation.
Signed-off-by: Shani Peretz
---
lib/ethdev
lizes these
APIs, introducing a new xstat counter designed to track hairpin packet drops.
Shani Peretz (3):
ethdev: add new API to enable/disable xstat by counters ID
app/testpmd: added testpmd commands to enable/disable xstat
net/mlx5: add port and queue level counters of hairpin drops
app/tes
t;examples/flow_filtering: introduce use cases snippets")
Signed-off-by: Shani Peretz
---
examples/flow_filtering/snippets/snippet_match_ipv4.c | 10 +-
examples/flow_filtering/snippets/snippet_match_ipv4.h | 1 -
2 files changed, 5 insertions(+), 6 deletions(-)
diff --git a/examples
age-
> From: Stephen Hemminger
> Sent: Tuesday, 24 December 2024 19:19
> To: Shani Peretz
> Cc: dev@dpdk.org; Aman Singh ; Dariusz
> Sosnowski ; Slava Ovsiienko
> ; Bing Zhao ; Ori Kam
> ; Suanming Mou ; Matan Azrad
> ; NBU-Contact-Thomas Monjalon (EXTERNAL)
> ; Fer
ueues they want statistics from.
If the counters are exceeded, the user would have the option to disable the
counter, thereby freeing up the resource for reuse elsewhere.
> -Original Message-
> From: Stephen Hemminger
> Sent: Sunday, 22 December 2024 18:46
> To: Shani Peretz
&g
ounter -
counts drops from all the hairpin queues
- Hairpin_out_of_buffer_rxq* - Queue-level counter -
counts drops of a specific queue
Both the port-level and per-queue counters can be
enabled, disabled, and queried.
Signed-off-by: Shani Peretz
---
app/te
that time (b69408ae453).
Before asan misalignment fix (average kflows/sec) -
Insertion - 4461.269, Deletion - 7799.9992
After:
Insertion - 4579.0642 , Deletion - 7913.0034
Fixes: 9a4c36880704 ("common/mlx5: optimize cache list object memory")
Cc: suanmi...@nvidia.com
Signed-off-by: Shani Peretz
corresponding feature.
Signed-off-by: Shani Peretz
---
doc/guides/prog_guide/ethdev/flow_offload.rst | 2 +
doc/guides/sample_app_ug/flow_filtering.rst | 295 +++---
examples/flow_filtering/Makefile | 19 +-
examples/flow_filtering/common.h | 29 ++
examples
corresponding feature.
Signed-off-by: Shani Peretz
---
doc/guides/prog_guide/rte_flow.rst| 2 +
doc/guides/sample_app_ug/flow_filtering.rst | 295 +++---
examples/flow_filtering/Makefile | 19 +-
examples/flow_filtering/common.h | 29 ++
examples
have their packet drops
tracked by these counters.
This approach provides flexibility for users to monitor packet drops
at either a port-level or a queue-level granularity.
Signed-off-by: Shani Peretz
---
drivers/common/mlx5/mlx5_devx_cmds.c | 8 +-
drivers/common/mlx5/mlx5_devx_cmds.h | 2
corresponding feature.
Signed-off-by: Shani Peretz
---
doc/guides/prog_guide/ethdev/flow_offload.rst | 2 +
doc/guides/sample_app_ug/flow_filtering.rst | 295 +++---
examples/flow_filtering/common.h | 29 ++
examples/flow_filtering/flow_blocks.c | 112 ---
examples
/remove device")
Cc: jblu...@infradead.org
Signed-off-by: Shani Peretz
Acked-by: Dariusz Sosnowski
Acked-by: Thomas Monjalon
---
app/test/test_vdev.c| 10 ++
drivers/bus/pci/pci_common.c| 11 +++
lib/eal/common/eal_common_dev.c | 11 ++-
lib/ea
/remove device")
Cc: jblu...@infradead.org
Signed-off-by: Shani Peretz
Acked-by: Dariusz Sosnowski
Acked-by: Thomas Monjalon
---
app/test/test_vdev.c| 10 ++
drivers/bus/pci/pci_common.c| 11 +++
lib/eal/common/eal_common_dev.c | 11 ++-
lib/ea
/remove device")
Cc: jblu...@infradead.org
Signed-off-by: Shani Peretz
Acked-by: Dariusz Sosnowski
Acked-by: Thomas Monjalon
---
app/test/test_vdev.c| 10 ++
drivers/bus/pci/pci_common.c| 10 ++
lib/eal/common/eal_common_dev.c | 11 ++-
lib/ea
tion phase.
The `hairpin_out_of_buffer` counter will be created only if there is
at least one hairpin Rx queue present on the device.
Signed-off-by: Shani Peretz
Acked-by: Dariusz Sosnowski
---
doc/guides/nics/mlx5.rst| 3 ++
doc/guides/rel_notes/release_24_07.rst | 1 +
driver
Exposing ethtool dev_internal_queue_oob counter which counts
the number of hairpin dropped packets in xstats.
Signed-off-by: Shani Peretz
Acked-by: Dariusz Sosnowski
---
doc/guides/rel_notes/release_24_07.rst | 1 +
drivers/net/mlx5/linux/mlx5_ethdev_os.c | 4
2 files changed, 5
Added global hairpin out of buffer counter (from ethtool)
Also added port-level hairpin out of buffer counter.
Shani Peretz (2):
net/mlx5: add global hairpin out of buffer counter
net/mlx5: add hairpin out of buffer counter
doc/guides/nics/mlx5.rst| 3 ++
doc/guides
.
The fix ensures that the function appropriately handles
the argument before finishing.
Fixes: 72a3dec7126f ("ethdev: add indirect flow list action")
Signed-off-by: Shani Peretz
---
app/test-pmd/cmdline_flow.c | 46 -
1 file changed, 25 inserti
66 matches
Mail list logo
