> -Original Message-
> From: Juraj Linkeš
> Sent: Wednesday, October 25, 2023 8:57 PM
> To: tho...@monjalon.net; Honnappa Nagarahalli ;
> bruce.richard...@intel.com; Ruifeng Wang
> Cc: dev@dpdk.org; Juraj Linkeš
> Subject: [PATCH v3] config/arm: update aarch32 build with gcc13
>
> The a
Add MLX5 PMD support for indirect actions list.
Erez Shitrit (1):
net/mlx5/hws: allow destination into default miss FT
Gregory Etelson (4):
net/mlx5: reformat HWS code for HWS mirror action
net/mlx5: support HWS mirror action
net/mlx5: reformat HWS code for indirect list actions
net/mlx
From: Hamdan Igbaria
Add the ability to create forward table and FTE.
Signed-off-by: Hamdan Igbaria
Acked-by: Suanming Mou
---
drivers/common/mlx5/mlx5_prm.h| 4
drivers/net/mlx5/hws/mlx5dr_cmd.c | 13 +
drivers/net/mlx5/hws/mlx5dr_cmd.h | 19 +++
3 files
From: Hamdan Igbaria
Add mlx5dr_devx_obj struct to mlx5dr_action, so we could hold
the FT obj in dest table action.
Signed-off-by: Hamdan Igbaria
Acked-by: Suanming Mou
---
drivers/net/mlx5/hws/mlx5dr_action.c | 4
drivers/net/mlx5/hws/mlx5dr_action.h | 3 +++
drivers/net/mlx5/hws/mlx5dr
From: Erez Shitrit
In FDB it will direct the packet into the hypervisor vport.
That allows the user to mirror packets into the default-miss vport.
Signed-off-by: Erez Shitrit
Acked-by: Suanming Mou
---
drivers/net/mlx5/hws/mlx5dr_action.c | 11 +++
1 file changed, 11 insertions(+)
di
Reformat HWS code for HWS mirror action.
Signed-off-by: Gregory Etelson
Acked-by: Suanming Mou
---
drivers/net/mlx5/mlx5_flow_hw.c | 70 ++---
1 file changed, 39 insertions(+), 31 deletions(-)
diff --git a/drivers/net/mlx5/mlx5_flow_hw.c b/drivers/net/mlx5/mlx5_flow
From: Shun Hao
This patch supports mirroring by adding an dest_array action. The action
accecpts a list containing multiple destination actions, and can duplicate
packet and forward to each destination in the list.
Signed-off-by: Shun Hao
Acked-by: Alex Vesker
Acked-by: Suanming Mou
---
driv
From: Haifei Luo
In dest_array action, an optional reformat action can be applied to each
destination. This patch supports this by using the extended destination
entry.
Signed-off-by: Haifei Luo
Signed-off-by: Shun Hao
Acked-by: Suanming Mou
---
drivers/common/mlx5/mlx5_prm.h | 15
HWS mirror clones original packet to one or two destinations and
proceeds with the original packet path.
The mirror has no dedicated RTE flow action type.
Mirror object is referenced by INDIRECT_LIST action.
INDIRECT_LIST for a mirror built from actions list:
SAMPLE [/ SAMPLE] / / END
Mirro
Reformat HWS code for indirect list actions.
Signed-off-by: Gregory Etelson
Acked-by: Suanming Mou
---
drivers/net/mlx5/mlx5_flow.h| 4 +-
drivers/net/mlx5/mlx5_flow_hw.c | 235 ++--
2 files changed, 131 insertions(+), 108 deletions(-)
diff --git a/drivers/net
Support indirect list METER_MARK action.
Signed-off-by: Gregory Etelson
Acked-by: Suanming Mou
---
drivers/net/mlx5/mlx5_flow.c| 69 -
drivers/net/mlx5/mlx5_flow.h| 70 --
drivers/net/mlx5/mlx5_flow_hw.c | 430 +++-
3 files changed, 484 insertions(+
From: Hamdan Igbaria
Add support for creation of packet reformat object,
via the ALLOC_PACKET_REFORMAT_CONTEXT command.
Signed-off-by: Hamdan Igbaria
Acked-by: Suanming Mou
---
drivers/common/mlx5/mlx5_prm.h | 39 +
drivers/net/mlx5/hws/mlx5dr_cmd.c | 60 +
Testpmd always provides RAW_ENCAP flow action configuration with
encap buffer and the buffer size.
That implementation does not allow to create non-masked raw_encap
action in the template API actions template.
The patch adds the `size` parameter to testpmd `raw_encap` action
configuration.
Testpmd
> From: Stephen Hemminger [mailto:step...@networkplumber.org]
> Sent: Wednesday, 25 October 2023 23.33
>
> On Wed, 25 Oct 2023 19:54:06 +0200
> Morten Brørup wrote:
>
> > I agree with Thomas on this.
> >
> > If you want the log message, please degrade it to INFO or DEBUG level. It is
> only rele
Fix confusing indentations.
Fixes: 1e728b01120c ("net/iavf: rework Tx path")
Fixes: 6bc987ecb860 ("net/iavf: support IPsec inline crypto")
Cc: sta...@dpdk.org
Signed-off-by: David Marchand
---
drivers/net/iavf/iavf_rxtx.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/d
On Wed, Oct 25, 2023 at 11:13 AM Radu Nicolau wrote:
>
> Relax the check in the previous fix to allow packets
> with security offload flag set.
>
> Fixes: 3c715591ece0 ("net/iavf: fix checksum offloading")
> Cc: sta...@dpdk.org
>
> Signed-off-by: Radu Nicolau
Acked-by: David Marchand
--
Davi
Hi,
> -Original Message-
> From: Rongwei Liu
> Sent: Wednesday, October 25, 2023 12:50 PM
> To: dev@dpdk.org; Matan Azrad ; Slava Ovsiienko
> ; Ori Kam ; Suanming Mou
> ; NBU-Contact-Thomas Monjalon (EXTERNAL)
>
> Subject: [PATCH v7] net/mlx5: add test for live migration
>
> This patch
This patch series adds ECDH testsuite. It also enables ECDH
support in CNXK PMD.
Gowrishankar Muthukrishnan (4):
test/cryptodev: add ECDH tests
crypto/cnxk: use generic EC opcodes
crypto/cnxk: change order of ECFPM params
crypto/cnxk: add ECDH support
app/test/test_cryptodev_asym.c
Add ECDH tests.
Signed-off-by: Gowrishankar Muthukrishnan
---
app/test/test_cryptodev_asym.c | 729 +++
app/test/test_cryptodev_asym_util.h | 12 +
app/test/test_cryptodev_ecdh_test_vectors.h | 577 +++
app/test/test_cryptodev_ecdsa_test_vecto
Use generic EC opcodes for sign and verify ops in ECDSA and SM2
implementations.
Signed-off-by: Gowrishankar Muthukrishnan
---
drivers/common/cnxk/roc_ae.h | 14 +++---
drivers/crypto/cnxk/cnxk_ae.h | 50 +++
2 files changed, 38 insertions(+), 26 deletions(-)
Change order of ECFPM params to match changes in v2.0 microcode.
Signed-off-by: Gowrishankar Muthukrishnan
---
drivers/crypto/cnxk/cnxk_ae.h | 49 ++-
1 file changed, 20 insertions(+), 29 deletions(-)
diff --git a/drivers/crypto/cnxk/cnxk_ae.h b/drivers/crypto/cn
Add ECDH support in CNXK PMD.
Signed-off-by: Gowrishankar Muthukrishnan
---
doc/guides/cryptodevs/features/cn10k.ini | 1 +
doc/guides/cryptodevs/features/cn9k.ini | 1 +
drivers/common/cnxk/roc_ae.h | 7 +
drivers/crypto/cnxk/cn10k_cryptodev_ops.c | 12 ++
On Wed, Oct 18, 2023 at 11:32 AM David Marchand
wrote:
>
> As an external header, rte_thash_x86_gfni.h should be self sufficient.
>
> Fixes: 3d4e27fd7ff0 ("use abstracted bit count functions")
>
> Signed-off-by: David Marchand
Applied.
--
David Marchand
On Tue, Oct 24, 2023 at 2:23 AM Xu, Rosen wrote:
> > The driver header bus_ifpga_driver.h was not self sufficient.
> > Fix this by including dev_driver.h.
> >
> > Fixes: 925c074e377b ("bus/ifpga: make driver-only headers private")
> > Cc: sta...@dpdk.org
> >
> > Signed-off-by: David Marchand
> Ac
On Tue, Oct 17, 2023 at 3:15 PM Morten Brørup
wrote:
>
> > From: Tyler Retzlaff [mailto:roret...@linux.microsoft.com]
> > Sent: Tuesday, 17 October 2023 15.04
> >
> > Remove return from rte_free and eal_free_no_trace both functions are
> > declared to have a return type of void so they shouldn't
On Tue, Oct 17, 2023 at 4:06 PM Tyler Retzlaff
wrote:
>
> Most MSVC targets do not support inline asm and __MWAITX__ is a GCC
> implementation detail. Use a test of defined(RTE_TOOLCHAIN_MSVC) to
> avoid conditional compilation of inline asm and instead use intrinsics
> instead.
>
> Fixes: c7ed1ce
24/10/2023 14:57, lihuisong (C):
> 在 2023/10/17 22:26, Ferruh Yigit 写道:
> > On 10/17/2023 3:06 PM, Thomas Monjalon wrote:
> >> 11/10/2023 11:27, Jie Hai:
> >>> 55 files changed, 395 insertions(+), 106 deletions(-)
> >>
> >> Changing all drivers is suspicious.
> >> It shows that something is miss
The thread specific variable workspace indicated by "key_workspace"
should be freed explicitly when closing a device. For example, in
Linux, when exiting an application, the thread will not exit
explicitly and the thread resources will not be deconstructed.
The commit to solve this introduced a gl
The thread specific variable workspace indicated by "key_workspace"
should be freed explicitly when closing a device. For example, in
Linux, when exiting an application, the thread will not exit
explicitly and the thread resources will not be deconstructed.
The commit to solve this introduced a gl
release null mbuf results coredump.
This commit adding mbuf check before releasing.
Fixes: 12016895fcf3 ("net/iavf: fix buffer leak on Tx queue stop")
Signed-off-by: Kaiwen Deng
---
drivers/net/iavf/iavf_rxtx_vec_common.h | 7 +--
1 file changed, 5 insertions(+), 2 deletions(-)
diff --git
On 26-Oct-23 9:06 AM, David Marchand wrote:
Fix confusing indentations.
Fixes: 1e728b01120c ("net/iavf: rework Tx path")
Fixes: 6bc987ecb860 ("net/iavf: support IPsec inline crypto")
Cc: sta...@dpdk.org
Signed-off-by: David Marchand
---
drivers/net/iavf/iavf_rxtx.c | 4 ++--
1 file change
Since the representor port needs to access the resources of the
associated DCF when it is closed. Therefore, the correct close
port operation is to close all the representor ports first, and
then close the associated DCF port.
If the DCF port is closed before the representor port on pmd exit.
This
The default size of the lcores array in the lcore dma map
is MAX_WORKER_NB. However, when parsing configuration
parameters, MAX_LCORE_NB is used as a constraint.
Since MAX_LCORE_NB is greater than MAX_WORKER_NB, this
causes array access to go out of bounds when the value
of the `lcore_dma/lcore` co
https://bugs.dpdk.org/show_bug.cgi?id=1305
Bug ID: 1305
Summary: testpmd: early exit with tap driver in non interactive
mode
Product: DPDK
Version: unspecified
Hardware: All
OS: All
Status: UNCON
Sorry for late response... comment inline
> -Original Message-
> From: Thomas Monjalon
> Sent: Friday, October 13, 2023 9:17 PM
> To: Zhang, Qi Z
> Cc: Yang, Qiming ; Zeng, ZhichaoX
> ; dev@dpdk.org; or...@nvidia.com;
> david.march...@redhat.com; Mcnamara, John
>
> Subject: Re: [PATCH v
On 10/26/2023 3:50 AM, Chaoyong He wrote:
>>> On 10/24/2023 3:28 AM, Chaoyong He wrote:
This patch series aims to add the NFP vDPA PMD, we also grab the
common logic into the `drivers/common/nfp` directory.
---
v2:
* Grab more logic into the `drivers/common/nfp` direct
> On 10/26/2023 3:50 AM, Chaoyong He wrote:
> >>> On 10/24/2023 3:28 AM, Chaoyong He wrote:
> This patch series aims to add the NFP vDPA PMD, we also grab the
> common logic into the `drivers/common/nfp` directory.
>
> ---
> v2:
> * Grab more logic into the `drivers/co
> -Original Message-
> From: Ferruh Yigit
> Sent: Thursday, October 26, 2023 12:30 AM
> To: Brandes, Shai
> Cc: dev@dpdk.org; Beider, Ron ; Atrash, Wajeeh
> ; Bernstein, Amit
> Subject: RE: [EXTERNAL] [PATCH v2 1/5] net/ena: hal upgrade
>
> CAUTION: This email originated from outside
On 10/26/23 02:31, Tyler Retzlaff wrote:
Replace the use of gcc builtin __atomic_xxx intrinsics with
corresponding rte_atomic_xxx optional stdatomic API
Signed-off-by: Tyler Retzlaff
---
lib/vhost/vdpa.c| 3 ++-
lib/vhost/vhost.c | 42 ---
On 10/26/23 02:31, Tyler Retzlaff wrote:
Replace the use of gcc builtin __atomic_xxx intrinsics with corresponding
rte_atomic_xxx optional stdatomic API
Signed-off-by: Tyler Retzlaff
---
lib/bbdev/rte_bbdev.c | 6 +++---
lib/bbdev/rte_bbdev.h | 2 +-
2 files changed, 4 insertions(+), 4 d
This patch series is an implementation of revised ml/cnxk driver
to support models compiled with TVM compiler framework. TVM models
use a hybrid mode for execution, with regions of the model executing
on the ML accelerator and the rest executing on CPU cores.
This series of commits reorganizes the
Dropped support for device argument "poll_mem" for cnxk
ML driver. Support to use registers for polling is removed
and DDR addresses would be used for polling.
Signed-off-by: Srikanth Yalavarthi
---
doc/guides/mldevs/cnxk.rst | 16 -
drivers/ml/cnxk/cn10k_ml_dev.c | 36 +--
dri
Introduce generic cnxk device structure. This structure is
a top level device structure for the driver, which would
encapsulate the target / platform specific device structure.
Signed-off-by: Srikanth Yalavarthi
---
drivers/ml/cnxk/cn10k_ml_dev.c | 316 ++--
drivers/ml/cnxk/cn1
Added generic cnxk request structure. Moved common fields
from cn10k structures to cnxk structure. Moved job related
structures and enumerations to ops headers.
Signed-off-by: Srikanth Yalavarthi
---
drivers/ml/cnxk/cn10k_ml_dev.c | 72 +++
drivers/ml/cnxk/cn10k_ml_dev.h | 269 +
Introduce generic cnxk model and layer structure. These
structures would enable supporting models with multiple
layers. A model is a collection of multiple independent
layers with flow dependencies between the layers.
Signed-off-by: Srikanth Yalavarthi
---
drivers/ml/cnxk/cn10k_ml_dev.h | 9
Added cnxk wrapper function to handle ML device queue-pairs.
Signed-off-by: Srikanth Yalavarthi
---
drivers/ml/cnxk/cn10k_ml_ops.c | 135 +
drivers/ml/cnxk/cn10k_ml_ops.h | 7 +-
drivers/ml/cnxk/cnxk_ml_ops.c | 153 -
drivers/ml/cnxk
Implemented cnxk wrapper functions to start and stop
ML models. Wrapper functions would invoke the cn10k
model start and stop functions.
Signed-off-by: Srikanth Yalavarthi
---
drivers/ml/cnxk/cn10k_ml_ocm.c | 28 ++--
drivers/ml/cnxk/cn10k_ml_ocm.h | 12 +-
drivers/ml/cnxk/cn10k_ml_ops.c | 282
Introduced generic xstats structures and renamed cn10k
xstats enumerations with cnxk prefix.
Signed-off-by: Srikanth Yalavarthi
---
drivers/ml/cnxk/cn10k_ml_dev.h | 86 +---
drivers/ml/cnxk/cn10k_ml_model.h | 6 +-
drivers/ml/cnxk/cn10k_ml_ops.c | 169 ++---
Added cnxk wrapper function to update model params and
fetch model info.
Signed-off-by: Srikanth Yalavarthi
---
drivers/ml/cnxk/cn10k_ml_ops.c | 38 ++-
drivers/ml/cnxk/cn10k_ml_ops.h | 5 ++--
drivers/ml/cnxk/cnxk_ml_ops.c | 48 --
3 fil
Added cnxk wrapper for device dump and selftest debug
functions.
Signed-off-by: Srikanth Yalavarthi
---
drivers/ml/cnxk/cn10k_ml_model.c | 118 +
drivers/ml/cnxk/cn10k_ml_model.h | 1 +
drivers/ml/cnxk/cn10k_ml_ocm.c | 8 +-
drivers/ml/cnxk/cn10k_ml_ocm.h | 2 +-
dr
Renamed cn10k ML ops structure with cnxk prefix.
Signed-off-by: Srikanth Yalavarthi
---
drivers/ml/cnxk/cn10k_ml_dev.c | 2 +-
drivers/ml/cnxk/cn10k_ml_ops.c | 73 +-
drivers/ml/cnxk/cn10k_ml_ops.h | 34 +++-
drivers/ml/cnxk/cnxk_ml_ops.c | 36 ++
Added cnxk wrapper function to handle ML device stats
Signed-off-by: Srikanth Yalavarthi
---
drivers/ml/cnxk/cn10k_ml_ops.c | 32 --
drivers/ml/cnxk/cn10k_ml_ops.h | 2 --
drivers/ml/cnxk/cnxk_ml_ops.c | 36 --
3 files changed, 34 ins
Implement CNXK wrapper functions for dev_info_get,
dev_configure, dev_close, dev_start and dev_stop. The
wrapper functions allocate / release common resources
for the ML driver and invoke device specific functions.
Signed-off-by: Srikanth Yalavarthi
---
drivers/ml/cnxk/cn10k_ml_ops.c | 230 ++---
Added cnxk wrapper function to handle ML device and model
extended stats. Handling resources for the xstats is done
in the cnxk layer. Introduced internal xstats group.
Signed-off-by: Srikanth Yalavarthi
---
drivers/ml/cnxk/cn10k_ml_dev.h | 4 -
drivers/ml/cnxk/cn10k_ml_ops.c | 531 +++
Implemented cnxk wrapper functions to load and unload
ML models. Wrapper functions would invoke the cn10k
model load and unload functions.
Signed-off-by: Srikanth Yalavarthi
---
drivers/ml/cnxk/cn10k_ml_model.c | 244 -
drivers/ml/cnxk/cn10k_ml_model.h | 26 ++-
drivers/
Introduced model type, sub-type and layer type. Added
internal structures for TVM model objects.
Signed-off-by: Srikanth Yalavarthi
---
drivers/ml/cnxk/cn10k_ml_ocm.c | 3 ++
drivers/ml/cnxk/cn10k_ml_ops.c | 6 ++-
drivers/ml/cnxk/cnxk_ml_model.h | 66 +++-
dri
Added cnxk wrapper functions to quantize input data and
dequantize output data.
Signed-off-by: Srikanth Yalavarthi
---
drivers/ml/cnxk/cn10k_ml_ops.c | 164 -
drivers/ml/cnxk/cn10k_ml_ops.h | 7 --
drivers/ml/cnxk/cnxk_ml_io.c | 95 +++
driver
Enable support to parse model buffer to identify the
model type and model sub-type. Enabled basic checks
for Glow model type buffer.
Signed-off-by: Srikanth Yalavarthi
Signed-off-by: Anup Prabhu
---
drivers/ml/cnxk/cnxk_ml_model.c | 49
drivers/ml/cnxk/cnxk_ml_mode
Implemented cnxk layer fast-path functions and added support
for model specific fast-path functions. CNXK layer functions
would invoke model specific fast-path functions.
Added support for model specific poll handling functions and
updated internal inference sync function. Drop use of rte_ml_op
as
Enable unloading model using external tvmdp library. Updated
layer unload callback to support multiple layers.
Signed-off-by: Srikanth Yalavarthi
Signed-off-by: Anup Prabhu
---
drivers/ml/cnxk/cn10k_ml_ops.c | 8 +---
drivers/ml/cnxk/cnxk_ml_ops.c| 7 +--
drivers/ml/cnxk/mvtvm_m
Move error type structures to cnxk layer. cn10k layer to
handle fw and hw error sub-types only.
Signed-off-by: Srikanth Yalavarthi
---
drivers/ml/cnxk/cn10k_ml_dev.h | 41 ++-
drivers/ml/cnxk/cn10k_ml_ops.c | 93 +-
drivers/ml/cnxk/cnxk_ml_dev.c | 8
Enabled support to print TVM model layer info.
Signed-off-by: Srikanth Yalavarthi
---
drivers/ml/cnxk/cnxk_ml_model.c | 7 +++-
drivers/ml/cnxk/mvtvm_ml_model.c | 59
drivers/ml/cnxk/mvtvm_ml_model.h | 2 ++
drivers/ml/cnxk/mvtvm_ml_stubs.c | 8 +
drivers
Added support to configure and close TVMDP library based
on ML device configuration options.
Updated meson build to enable Jansson, TVM runtime, TVMDP
library as build dependencies.
Signed-off-by: Srikanth Yalavarthi
---
config/arm/arm64_cn10k_linux_gcc | 1 +
config/arm/arm64_cn9k_linux_gcc
Added support to parse TVM model objects from the model
archive buffer. Added support to check for all expected
objects and copy TVM model objects to internal buffers.
Signed-off-by: Srikanth Yalavarthi
Signed-off-by: Anup Prabhu
---
drivers/ml/cnxk/cnxk_ml_ops.c| 5 ++-
drivers/ml/cnxk/mv
Added model xstats entries to compute runtime latency.
Allocated internal resources for TVM model xstats.
Signed-off-by: Srikanth Yalavarthi
---
drivers/ml/cnxk/cn10k_ml_ops.c | 9 +++
drivers/ml/cnxk/cn10k_ml_ops.h | 2 +
drivers/ml/cnxk/cnxk_ml_ops.c| 131 ++
Added support to fetch TVM model layer information and
update internal structures based on the layer information
Set callback functions for layer load and unload and
enable model loading using TVMDP library. Added support
to fetch full metadata after model load.
Signed-off-by: Srikanth Yalavarthi
Enabled updating internal IO info structures for TVM model.
Compute static fields related to the model I/O.
Signed-off-by: Srikanth Yalavarthi
---
drivers/ml/cnxk/cnxk_ml_ops.c| 4 +
drivers/ml/cnxk/mvtvm_ml_model.c | 133 +++
drivers/ml/cnxk/mvtvm_ml_model.h |
Implemented generic ML malloc and free callbacks
Signed-off-by: Srikanth Yalavarthi
---
drivers/ml/cnxk/cn10k_ml_ops.c | 30 ++
drivers/ml/cnxk/cn10k_ml_ops.h | 3 +++
drivers/ml/cnxk/mvtvm_ml_ops.c | 2 ++
3 files changed, 35 insertions(+)
diff --git a/drivers/ml/
From: Anup Prabhu
Enabled check for OCM size requirement for multi-layer
TVM model. Compute OCM scratch and WB requirement for
all layers during the load stage.
Signed-off-by: Anup Prabhu
---
drivers/ml/cnxk/cnxk_ml_ops.c | 60 +++
1 file changed, 60 insertions(
From: Anup Prabhu
Enable fast-path ops support for TVM models. Models would
use TVMDP library function calls to execute inference
operations for Hybrid and LLVM model sub-types.
For TVM MRVL model subtypes that have a single MRVL layer,
the inference requests are directly enqueued to hardware
by
Added support to start and stop TVM models. TVM model
start would invoke layer start for all Glow layers part
of the model. TVM model stop would invoke layer stop
for all Glow layers part of the model.
Signed-off-by: Srikanth Yalavarthi
Signed-off-by: Anup Prabhu
---
drivers/ml/cnxk/cn10k_ml_op
From: Prince Takkar
Added support to update internal model info structure
for TVM models.
Signed-off-by: Prince Takkar
Signed-off-by: Srikanth Yalavarthi
---
drivers/ml/cnxk/mvtvm_ml_model.c | 66
drivers/ml/cnxk/mvtvm_ml_model.h | 2 +
drivers/ml/cnxk/mvtvm_
Implemented callback functions for IO allocation and free
for Glow layers.
Signed-off-by: Srikanth Yalavarthi
---
drivers/ml/cnxk/cn10k_ml_ops.c | 87 ++
drivers/ml/cnxk/cn10k_ml_ops.h | 3 ++
drivers/ml/cnxk/mvtvm_ml_ops.c | 2 +
3 files changed, 92 insertions(
From: Prince Takkar
Added support for quantize and dequantize callback
functions for TVM models.
Signed-off-by: Prince Takkar
---
drivers/ml/cnxk/mvtvm_ml_ops.c | 129 +
drivers/ml/cnxk/mvtvm_ml_ops.h | 4 +
2 files changed, 133 insertions(+)
diff --git a/dri
Enable support to create a mvtvm virtual device on
system's without a PCI based ML HW accelerator.
Signed-off-by: Srikanth Yalavarthi
---
doc/guides/mldevs/cnxk.rst | 50 +++-
drivers/ml/cnxk/cn10k_ml_dev.c | 8 ++
drivers/ml/cnxk/cn10k_ml_dev.h | 3 +
drivers/ml/cnxk/cnxk_ml_
From: Anup Prabhu
Enabled support for pre-quantized input and output in ML
test application.
Signed-off-by: Anup Prabhu
---
v2:
- Updated application help
v1:
- Initial changes
app/test-mldev/ml_options.c| 11 +-
app/test-mldev/ml_options.h| 28 ++
> -Original Message-
> From: Srikanth Yalavarthi
> Sent: 23 October 2023 19:32
> To: Bruce Richardson
> Cc: dmitry.kozl...@gmail.com; Aaron Conole ; Igor
> Russkikh ; David Marchand
> ; dev@dpdk.org; Shivah Shankar Shankar
> Narayan Rao ; Jerin Jacob Kollanukkaran
> ; sta...@dpdk.org; Sri
On 10/20/2023 1:06 PM, David Marchand wrote:
> In the process of releasing ethdev port in the primary process,
> secondary processes are asked to stop referencing such a port.
> Doing so, those processes can't predict whether the primary process
> will later invalidate/free the shared memory. So th
Add modexp tests for groups 5, 14, 15, 16, 17 and 18.
Signed-off-by: Gowrishankar Muthukrishnan
---
v3:
- Fixed test descriptions for clang requirement.
---
app/test/test_cryptodev_asym.c | 48 +
app/test/test_cryptodev_mod_test_vectors.h | 989 -
2 files change
25/10/2023 18:31, Thomas Monjalon:
> Real-time thread priority was been forbidden on Unix
> because of problems they can cause.
> Warnings and helpers are added to avoid deadlocks,
> so real-time can be allowed on all systems.
Unit test is failing:
DPDK:fast-tests / threads_autotest TIMEOUT 6
Real-time thread priority was been forbidden on Unix
because of problems they can cause.
Warnings and helpers are added to avoid deadlocks,
so real-time can be allowed on all systems.
Thomas Monjalon (2):
eal: add thread yield functions
eal/unix: allow creating thread with real-time priority
When running real-time threads, we may need to force scheduling
kernel threads or other real-time threads.
New functions are added to address these cases.
The yield functions should not have any interest for normal threads.
Note: other purposes may be addressed with rte_pause() or rte_delay_*().
When adding an API for creating threads,
the real-time priority has been forbidden on Unix.
There is a known issue with ring behaviour,
but it should not be completely forbidden.
Real-time thread can block some kernel threads on the same core,
making the system unstable.
That's why a pause is add
> From: Thomas Monjalon [mailto:tho...@monjalon.net]
> Sent: Thursday, 26 October 2023 15.37
>
> 25/10/2023 18:31, Thomas Monjalon:
> > Real-time thread priority was been forbidden on Unix
> > because of problems they can cause.
> > Warnings and helpers are added to avoid deadlocks,
> > so real-ti
On Thu, Oct 26, 2023 at 2:32 AM Tyler Retzlaff
wrote:
>
> Replace the use of gcc builtin __atomic_xxx intrinsics with
> corresponding rte_atomic_xxx optional stdatomic API.
>
> v3:
> * add missing atomic specification on head variable
> in struct rte_ring_headtail
> * adapt to use rte_atom
> From: Morten Brørup [mailto:m...@smartsharesystems.com]
> Sent: Thursday, 26 October 2023 15.45
>
> > From: Thomas Monjalon [mailto:tho...@monjalon.net]
> > Sent: Thursday, 26 October 2023 15.37
> >
> > 25/10/2023 18:31, Thomas Monjalon:
> > > Real-time thread priority was been forbidden on Unix
26/10/2023 15:57, Morten Brørup:
> > From: Morten Brørup [mailto:m...@smartsharesystems.com]
> > Sent: Thursday, 26 October 2023 15.45
> >
> > > From: Thomas Monjalon [mailto:tho...@monjalon.net]
> > > Sent: Thursday, 26 October 2023 15.37
> > >
> > > 25/10/2023 18:31, Thomas Monjalon:
> > > > Rea
> From: Thomas Monjalon [mailto:tho...@monjalon.net]
> Sent: Thursday, 26 October 2023 16.05
>
> 26/10/2023 15:57, Morten Brørup:
> > > From: Morten Brørup [mailto:m...@smartsharesystems.com]
> > > Sent: Thursday, 26 October 2023 15.45
> > >
> > > > From: Thomas Monjalon [mailto:tho...@monjalon.ne
On Thu, Oct 26, 2023 at 3:57 PM Morten Brørup
wrote:
>
> > From: Morten Brørup [mailto:m...@smartsharesystems.com]
> > Sent: Thursday, 26 October 2023 15.45
> >
> > > From: Thomas Monjalon [mailto:tho...@monjalon.net]
> > > Sent: Thursday, 26 October 2023 15.37
> > >
> > > 25/10/2023 18:31, Thoma
From: Shai Brandes
Hi,
This patchset contains alignment of the driver to the latest HAL version
which adds support for retrieving new metrics from the device and opens
a path to use additional device features that are not yet supported by
the driver.
The new driver features are mostly about add
From: Shai Brandes
ENA maintains a HAL that is shared by all supported host drivers.
Main features introduced to the HAL:
[1] Reworked the mechanism that queries the performance metrics
from the device.
[2] Added support for a new metric that allows monitoring the
available tracked connec
From: Shai Brandes
The driver publishes network performance metrics that the
application can use to troubleshoot performance issues,
monitor the workload, and benchmark applications to determine
whether they maximize the performance.
This patch adds support for the connection tracking allowance
From: Shai Brandes
RX overrun error occur when a packet arrives but there are
not enough free buffers in the RX ring to receive it.
The driver publishes the extended statistics with the RX
buffer overrun errors as reported by the device.
Signed-off-by: Shai Brandes
Reviewed-by: Amit Bernstein
From: Shai Brandes
ENA-express is powered by AWS scalable reliable datagram (SRD)
technology. SRD is a high performance network transport protocol
that uses dynamic routing to increase throughput and minimize
tail latency.
The driver expose the following ENA-express metrics via xstats:
* ena_srd
From: Shai Brandes
This release introduces:
* Upgraded ENA HAL.
* Support for connection tracking allowance utilization metric.
* Support for reporting rx overrun errors in xstats.
* Support for ENA-express metrics.
Signed-off-by: Shai Brandes
Reviewed-by: Amit Bernstein
---
drivers/net/ena/e
Real-time thread priority was been forbidden on Unix
because of problems they can cause.
Warnings and helpers are added to avoid deadlocks,
so real-time can be allowed on all systems.
Thomas Monjalon (2):
eal: add thread yield functions
eal/unix: allow creating thread with real-time priority
When running real-time threads, we may need to force scheduling
kernel threads or other real-time threads.
New functions are added to address these cases.
The yield functions should not have any interest for normal threads.
Note: other purposes may be addressed with rte_pause() or rte_delay_*().
When adding an API for creating threads,
the real-time priority has been forbidden on Unix.
There is a known issue with ring behaviour,
but it should not be completely forbidden.
Real-time thread can block some kernel threads on the same core,
making the system unstable.
That's why a sleep is add
26/10/2023 16:08, Morten Brørup:
> > From: Thomas Monjalon [mailto:tho...@monjalon.net]
> > Sent: Thursday, 26 October 2023 16.05
> >
> > 26/10/2023 15:57, Morten Brørup:
> > > > From: Morten Brørup [mailto:m...@smartsharesystems.com]
> > > > Sent: Thursday, 26 October 2023 15.45
> > > >
> > > > >
On 10/26/2023 7:42 AM, Chaoyong He wrote:
> This patch series aims to add the NFP vDPA PMD, we also grab the common
> logic into the `drivers/common/nfp` directory.
>
> ---
> v3:
> * Replace 'pthread_xxx' API with 'rte_thread_xxx' API.
> * Replace '__atomic_xxx' built-ins with 'rte_atomic_xxx'.
>
1 - 100 of 183 matches
Mail list logo