Add support for ODM DMA device in devbind.
Signed-off-by: Anoob Joseph
Signed-off-by: Gowrishankar Muthukrishnan
Signed-off-by: Vidya Sagar Velumuri
---
usertools/dpdk-devbind.py | 5 -
1 file changed, 4 insertions(+), 1 deletion(-)
diff --git a/usertools/dpdk-devbind.py b/usertools/dpdk-
Anoob Joseph, Apr 17, 2024 at 09:11:
Add support for ODM DMA device in devbind.
Signed-off-by: Anoob Joseph
Signed-off-by: Gowrishankar Muthukrishnan
Signed-off-by: Vidya Sagar Velumuri
Acked-by: Robin Jarry
The "rte_flow_item_geneve_opt" structure describes the GENEVE TLV option
header according to RFC 8926 [1]:
struct rte_flow_item_geneve_opt {
rte_be16_t option_class;
uint8_t option_type;
uint8_t option_len;
uint32_t *data;
};
The "option_len" field is used for two
The "rte_flow_conv()" function, enables, among other things, to copy
item list.
For GENEVE option item, the function copies it without considering deep
copy. It copies the "data" pointer without copying the pointed values.
This patch adds deep copy for after regular copy.
Fixes: 2b4c72b4d10d ("e
The "rte_flow_item_geneve_opt" structure has field for option length.
This field has 2 different usages which might limit each other:
1. field to matching with value/mask.
2. descriptor for data array size.
This patch adds a new field "data_array_size" into
"rte_flow_item_geneve_opt" structure
Add Odyssey ODM DMA device. This PMD abstracts ODM hardware unit on
Odyssey SoC which can perform mem to mem copies.
The hardware unit can support upto 32 queues (vchan) and 16 VFs. It
supports 'fill' operation with specific values. It also supports
SG mode of operation with upto 4 src pointers an
Add framework for Odyssey ODM DMA device.
Signed-off-by: Anoob Joseph
Signed-off-by: Gowrishankar Muthukrishnan
Signed-off-by: Vidya Sagar Velumuri
---
MAINTAINERS | 6 +++
drivers/dma/meson.build | 1 +
drivers/dma/odm/meson.build | 14 +++
drivers/dma/odm/odm.h
Add ODM registers and structures. Add mailbox structs as well.
Signed-off-by: Anoob Joseph
Signed-off-by: Gowrishankar Muthukrishnan
Signed-off-by: Vidya Sagar Velumuri
---
drivers/dma/odm/odm.h | 116 +
drivers/dma/odm/odm_priv.h | 49
From: Gowrishankar Muthukrishnan
Add ODM device init and fini.
Signed-off-by: Anoob Joseph
Signed-off-by: Gowrishankar Muthukrishnan
Signed-off-by: Vidya Sagar Velumuri
---
drivers/dma/odm/meson.build | 2 +-
drivers/dma/odm/odm.c| 97
drivers/d
From: Gowrishankar Muthukrishnan
Add DMA dev stats.
Signed-off-by: Anoob Joseph
Signed-off-by: Gowrishankar Muthukrishnan
Signed-off-by: Vidya Sagar Velumuri
---
drivers/dma/odm/odm_dmadev.c | 63 ++--
1 file changed, 61 insertions(+), 2 deletions(-)
diff --g
From: Gowrishankar Muthukrishnan
Add DMA device control ops.
Signed-off-by: Anoob Joseph
Signed-off-by: Gowrishankar Muthukrishnan
Signed-off-by: Vidya Sagar Velumuri
---
drivers/dma/odm/odm.c| 144 ++-
drivers/dma/odm/odm.h| 58 ++
From: Vidya Sagar Velumuri
Add ODM copy and copy SG ops.
Signed-off-by: Anoob Joseph
Signed-off-by: Gowrishankar Muthukrishnan
Signed-off-by: Vidya Sagar Velumuri
---
drivers/dma/odm/odm_dmadev.c | 236 +++
1 file changed, 236 insertions(+)
diff --git a/drive
From: Vidya Sagar Velumuri
Add all remaining ops such as fill, burst_capacity etc. Also update the
documentation.
Signed-off-by: Anoob Joseph
Signed-off-by: Gowrishankar Muthukrishnan
Signed-off-by: Vidya Sagar Velumuri
---
MAINTAINERS | 1 +
doc/guides/dmadevs/index.rst |
On 4/16/2024 6:32 PM, Stephen Hemminger wrote:
> On Tue, 16 Apr 2024 17:29:25 +0100
> Ferruh Yigit wrote:
>
>> On 4/16/2024 4:19 PM, Stephen Hemminger wrote:
>>> Use RTE_DIM instead of computing directly with sizeof.
>>> Patch automatically generated via cocci/rte_dim.cocci.
>>> Code in base/ sub
From: Pavan Nikhilesh
Re-organize event DMA ops structure to allow holding
source and destination pointers without the need for
additional memory, the mempool allocating memory for
rte_event_dma_adapter_ops can size the structure to
accommodate all the needed source and destination
pointers.
Add
From: Pavan Nikhilesh
Use DMA ops to store metadata, remove use of completion pool.
Signed-off-by: Pavan Nikhilesh
Acked-by: Vamsi Attunuru
---
drivers/dma/cnxk/cnxk_dmadev.c | 53 ++-
drivers/dma/cnxk/cnxk_dmadev.h | 24 +--
drivers/dma/cnxk/cnxk_dmadev_fp
This patch introduces a new flag RTE_VHOST_USER_ASYNC_CONNECT,
which in combination with the flag RTE_VHOST_USER_CLIENT makes
rte_vhost_driver_start connect asynchronously to the vhost server.
Signed-off-by: Daniil Ushkov
---
lib/vhost/rte_vhost.h | 1 +
lib/vhost/socket.c| 28 +
This patch introduces a new flag RTE_VHOST_USER_ASYNC_CONNECT,
which in combination with the flag RTE_VHOST_USER_CLIENT makes
rte_vhost_driver_start connect asynchronously to the vhost server.
Signed-off-by: Ушков Даниил
---
Fixes:
1. Fix warning about the name in signature.
2. Fix warning about
On 2024/4/9 10:50, Stephen Hemminger wrote:
> On Tue, 9 Apr 2024 10:06:01 +0800
> huangdengdui wrote:
>
>> On 2024/4/8 16:45, Ferruh Yigit wrote:
>>> On 4/8/2024 6:52 AM, huangdengdui wrote:
On 2024/4/6 0:44, Stephen Hemminger wrote:
> On Sat, 30 Mar 2024 15:44:09 +0800
The version variable is not initialized. Therefore, if the -V option
is not specified, the value of $version is obtained from the context,
which may cause the version map parsing failure.
Fixes: 6edec7f202ac ("devtools: list symbols by version")
Cc: sta...@dpdk.org
Signed-off-by: Dengdui Huang
-
On 8/18/2022 6:43 AM, Ke Zhang wrote:
> Increase the loop count so that there is a longer
> threshold time for the LACP handshake process.
>
> Fixes: 5e41ab250dfa ("app/test: unit tests for bonding mode 4")
> Cc: sta...@dpdk.org
>
> Signed-off-by: Ke Zhang
>
Patch seems to fix unit test where L
> -Original Message-
> From: Kundapura, Ganapati
> Sent: Tuesday, April 16, 2024 9:12 AM
> To: dev@dpdk.org
> Cc: Gujjar, Abhinandan S ; Power, Ciara
> ; gak...@marvell.com; fanzhang@gmail.com
> Subject: [PATCH v1] crypto: fix build issues on crypto callbacks macro
> undefined
>
>
> -Original Message-
> From: luca.bocca...@gmail.com
> Sent: Thursday, April 4, 2024 2:11 AM
> To: sta...@dpdk.org
> Cc: dev@dpdk.org; Abhishek Marathe ;
> Ali Alnubani ; benjamin.wal...@intel.com; David
> Christensen ; Hemant Agrawal
> ; Ian Stokes ; Jerin Jacob
> ; John McNamara ; Ju-
>
On Tue, Apr 16, 2024 at 2:12 PM Luca Vizzarro wrote:
>
> Thank you for your review Juraj!
>
> On 16/04/2024 09:40, Juraj Linkeš wrote:
>
> >> When sending a command using an instance of InteractiveShell the output
> >> is meant to filter out the leading shell prompt. The filtering logic is
> >> pr
On 4/16/2024 4:44 PM, Ferruh Yigit wrote:
> On 6/6/2022 3:34 PM, Gaoxiang Liu wrote:
>> When the primary port of bond slaves with bond mode 4 linked down,
>> the system id of the other slave ports channged.
>> It may cause some switches to renegotiate,
>> and the process takes a few seconds. It is
On Tue, Apr 16, 2024 at 2:15 PM Luca Vizzarro wrote:
>
> On 16/04/2024 09:48, Juraj Linkeš wrote:
> > Oh, the first commit message was confusing. It said leading prompt
> > which I understood to be the first prompt (the one with the command).
> > I see that this commit actually addresses what I th
On Tue, Apr 16, 2024 at 2:24 PM Luca Vizzarro wrote:
>
> On 16/04/2024 10:03, Juraj Linkeš wrote:
> >> +@dataclass
> >> +class TestPmdPort(TextParser):
> >
> > This and the classes above are missing docstrings.
>
> Ack.
>
> >> @@ -225,6 +664,38 @@ def set_forward_mode(self, mode:
> >> TestPmdForw
On 11/3/2023 12:08 AM, Ferruh Yigit wrote:
> On 11/2/2023 10:17 PM, Thomas Monjalon wrote:
>> 02/11/2023 22:21, Ajit Khaparde:
>>> On Thu, Nov 2, 2023 at 2:13 PM Ferruh Yigit wrote:
On 6/25/2023 4:45 PM, Thomas Monjalon wrote:
> 23/06/2023 16:00, Ferruh Yigit:
>> On 2/3/2023 1:28 PM,
In the previous implementation, the maximum reclamation size was set
to RTE_HASH_RCU_DQ_RECLAIM_MAX and it was not configurable. This patch
uses the configuration argument to set the maximum reclamation size.
Fixes: 769b2de7fb52 ("hash: implement RCU resources reclamation")
Cc: dharmik.thak...@arm
Don't forget to run the ../devtools/dts-check-format.sh script, it
found one issue:
framework/testbed_model/os_session.py:351: [D] D417 Missing argument
descriptions in the docstring [pydocstyle]
On Tue, Apr 16, 2024 at 8:18 PM Nicholas Pratte wrote:
>
> The previous implementation configures and
On 17/04/2024 14:06, Juraj Linkeš wrote:
Actually, filtering out the last prompt is fine. I got this mixed up
with processing the last prompt. We must process the last prompt, but
we don't need to print it (and there's not much reason to do so).
Yeah, my confusion was about displaying the last
On 17/04/2024 14:22, Juraj Linkeš wrote:
I agree that it looks much better. I gave it a first attempt to come up
with a regular expression that is not too complicated and is able to
match blocks individually. I've noticed that blocks start with:
\n* Infos for port X
but
On 17/04/2024 15:25, Luca Vizzarro wrote:
On 17/04/2024 14:22, Juraj Linkeš wrote:
I'll
experiment with some look ahead constructs. The easiest solution is to
match everything that is not * ([^*]+) but can we be certain that there
won't be any asterisk in the actual information?
We can't. But
This patchset adds SVE support for the signature comparison in the cuckoo
hash lookup and improves the existing NEON implementation. These
optimizations required changes to the data format and signature of the
relevant functions to support dense hitmasks (no padding) and having the
primary and seco
Upon a successful comparison, NEON sets all the bits in the lane to 1
We can skip shifting by simply masking with specific masks.
Signed-off-by: Yoan Picchi
Reviewed-by: Ruifeng Wang
Reviewed-by: Nathan Brown
---
lib/hash/arch/arm/compare_signatures.h | 24 +++-
1 file chan
Current hitmask includes padding due to Intel's SIMD
implementation detail. This patch allows non Intel SIMD
implementations to benefit from a dense hitmask.
In addition, the new dense hitmask interweave the primary
and secondary matches which allow a better cache usage and
enable future improvemen
This patch adds unit test for rte_hash_lookup_bulk().
It also update the test_full_bucket test to the current number of entries
in a hash bucket.
Signed-off-by: Yoan Picchi
Signed-off-by: Harjot Singh
Reviewed-by: Ruifeng Wang
Reviewed-by: Nathan Brown
---
app/test/test_hash.c | 99 ++
- Implemented SVE code for comparing signatures in bulk lookup.
- New SVE code is ~5% slower than optimized NEON for N2 processor for
128b vectors.
Signed-off-by: Yoan Picchi
Signed-off-by: Harjot Singh
Reviewed-by: Nathan Brown
Reviewed-by: Ruifeng Wang
---
lib/hash/arch/arm/compare_signatur
VMware was acquired by Broadcom, and vmware e-mail addresses will
be invalid in the future.
Add broadcom.com addresses for those still in the company.
Place broadcom.com addresses first, as they are now the active ones.
Retain the old addresses from git history behind them.
Signed-off-by: Jochen
Latencystats uses variable length array and floating point when they
are not necessary to acheive the same result. While testing also noticed
that the code was computing wrong values on my test system, and that
include files were missing.
Stephen Hemminger (5):
latencystats: replace use of VLA
The temporary array latencystats is not needed if the algorithm
is converted into one pass.
Signed-off-by: Stephen Hemminger
---
lib/latencystats/rte_latencystats.c | 31 +++--
1 file changed, 16 insertions(+), 15 deletions(-)
diff --git a/lib/latencystats/rte_latencysta
The timer_hz is not always an integral number of nanoseconds.
For examples, cycles per nanoseconds on my test system is 2.8.
Fix by using floating point where needed.
Signed-off-by: Stephen Hemminger
---
lib/latencystats/rte_latencystats.c | 12 +---
1 file changed, 5 insertions(+), 7 de
The cycle counts do not need to be stored as floating point.
Instead keep track of latency in cycles, and convert to
nanoseconds when read.
Change Exponential Weighted Moving Average weight from .2 to .25
to avoid use of floating point for that.
Signed-off-by: Stephen Hemminger
---
lib/latencys
All messages that because of an error should be at log level
NOTICE or above. Do not break log messages across lines.
Signed-off-by: Stephen Hemminger
---
lib/latencystats/rte_latencystats.c | 30 ++---
1 file changed, 15 insertions(+), 15 deletions(-)
diff --git a/lib/l
Include what is used here.
Signed-off-by: Stephen Hemminger
---
lib/latencystats/rte_latencystats.c | 85 ++---
1 file changed, 52 insertions(+), 33 deletions(-)
diff --git a/lib/latencystats/rte_latencystats.c
b/lib/latencystats/rte_latencystats.c
index 62038a9f5d..31e
On Wed, Apr 17, 2024 at 10:07:27AM -0700, Stephen Hemminger wrote:
> Include what is used here.
>
There are more changes in this patch than just includes cleanup.
> Signed-off-by: Stephen Hemminger
> ---
> lib/latencystats/rte_latencystats.c | 85 ++---
> 1 file changed,
> From: Stephen Hemminger [mailto:step...@networkplumber.org]
> Sent: Wednesday, 17 April 2024 19.07
>
> The temporary array latencystats is not needed if the algorithm
> is converted into one pass.
>
> Signed-off-by: Stephen Hemminger
> ---
> lib/latencystats/rte_latencystats.c | 31 ++
MSVC does not support VLAs, replace VLAs with standard C arrays
or alloca(). alloca() is available for all toolchain/platform
combinations officially supported by DPDK.
Signed-off-by: Tyler Retzlaff
---
lib/hash/rte_cuckoo_hash.c | 4 ++--
lib/hash/rte_thash.c | 2 +-
2 files changed, 3 in
Include alloca.h for Linux and malloc.h for Windows to get declaration
of alloca().
Signed-off-by: Tyler Retzlaff
---
lib/eal/linux/include/rte_os.h | 1 +
lib/eal/windows/include/rte_os.h | 1 +
2 files changed, 2 insertions(+)
diff --git a/lib/eal/linux/include/rte_os.h b/lib/eal/linux/incl
MSVC does not support VLAs, replace VLAs with standard C arrays
or alloca(). alloca() is available for all toolchain/platform
combinations officially supported by DPDK.
Signed-off-by: Tyler Retzlaff
---
lib/gro/rte_gro.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/lib
As per guidance technical board meeting 2024/04/17. This series
removes the use of VLAs from code built for Windows for all 3
toolchains. If there are additional opportunities to convert VLAs
to regular C arrays please provide the details for incorporation
into the series.
MSVC does not support VL
MSVC does not support VLAs, replace VLAs with standard C arrays
or alloca(). alloca() is available for all toolchain/platform
combinations officially supported by DPDK.
Signed-off-by: Tyler Retzlaff
---
lib/ethdev/rte_ethdev.c | 9 +
1 file changed, 5 insertions(+), 4 deletions(-)
diff
MSVC does not support VLAs, replace VLAs with standard C arrays
or alloca(). alloca() is available for all toolchain/platform
combinations officially supported by DPDK.
Signed-off-by: Tyler Retzlaff
---
drivers/net/ice/ice_rxtx.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git
MSVC does not support VLAs, replace VLAs with standard C arrays
or alloca(). alloca() is available for all toolchain/platform
combinations officially supported by DPDK.
Signed-off-by: Tyler Retzlaff
---
lib/rcu/rte_rcu_qsbr.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --gi
MSVC does not support VLAs, replace VLAs with standard C arrays
or alloca(). alloca() is available for all toolchain/platform
combinations officially supported by DPDK.
Signed-off-by: Tyler Retzlaff
---
drivers/net/i40e/i40e_testpmd.c | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
d
MSVC does not support VLAs, replace VLAs with standard C arrays
or alloca(). alloca() is available for all toolchain/platform
combinations officially supported by DPDK.
Signed-off-by: Tyler Retzlaff
---
app/test-pmd/cmdline.c| 2 +-
app/test-pmd/cmdline_flow.c | 9 -
app/test
MSVC does not support VLAs, replace VLAs with standard C arrays
or alloca(). alloca() is available for all toolchain/platform
combinations officially supported by DPDK.
Signed-off-by: Tyler Retzlaff
---
lib/latencystats/rte_latencystats.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
di
MSVC does not support VLAs, replace VLAs with standard C arrays
or alloca(). alloca() is available for all toolchain/platform
combinations officially supported by DPDK.
Signed-off-by: Tyler Retzlaff
---
lib/lpm/rte_lpm.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/lib/lpm
MSVC does not support VLAs, replace VLAs with standard C arrays
or alloca(). alloca() is available for all toolchain/platform
combinations officially supported by DPDK.
Signed-off-by: Tyler Retzlaff
---
drivers/net/ixgbe/ixgbe_ethdev.c | 5 +++--
drivers/net/ixgbe/ixgbe_rxtx_vec_common.
MSVC does not support VLAs, replace VLAs with standard C arrays
or alloca(). alloca() is available for all toolchain/platform
combinations officially supported by DPDK.
Signed-off-by: Tyler Retzlaff
---
drivers/net/mlx5/mlx5.c | 5 ++---
drivers/net/mlx5/mlx5_flow.c | 6 +++---
2 files chan
MSVC does not support VLAs, replace VLAs with standard C arrays
or alloca(). alloca() is available for all toolchain/platform
combinations officially supported by DPDK.
Signed-off-by: Tyler Retzlaff
---
drivers/common/idpf/idpf_common_rxtx.c| 2 +-
drivers/common/idpf/idpf_common_rxtx_av
MSVC does not support VLAs, replace VLAs with standard C arrays
or alloca(). alloca() is available for all toolchain/platform
combinations officially supported by DPDK.
Signed-off-by: Tyler Retzlaff
---
app/test/test.c | 2 +-
app/test/test_cmdline_string.c| 2 +-
MSVC does not support VLAs, replace VLAs with standard C arrays
or alloca(). alloca() is available for all toolchain/platform
combinations officially supported by DPDK.
Signed-off-by: Tyler Retzlaff
---
drivers/common/mlx5/mlx5_common.h| 4 ++--
drivers/common/mlx5/mlx5_devx_cmds.c | 7 +++--
MSVC does not support optional C11 VLAs. When building for Windows
enable -Wvla so that mingw and clang also fail if a VLA is used.
Signed-off-by: Tyler Retzlaff
---
config/meson.build | 4
1 file changed, 4 insertions(+)
diff --git a/config/meson.build b/config/meson.build
index 8c8b019..
> This introduced additional check in Rx path, not sure what is the performance
> impact.
>
> I can see Long already acked the v3, I just want to double check.
> If Tx queue number > Rx queue number is not a common usecase, perhaps it can
> be an option to forbid it instead of getting performanc
On Wed, Apr 17, 2024 at 10:07:23AM -0700, Stephen Hemminger wrote:
> The temporary array latencystats is not needed if the algorithm
> is converted into one pass.
>
> Signed-off-by: Stephen Hemminger
> ---
if this series is merged first, my series here
https://patchwork.dpdk.org/project/dpdk/pat
On Wed, Apr 17, 2024 at 10:07:24AM -0700, Stephen Hemminger wrote:
> The timer_hz is not always an integral number of nanoseconds.
> For examples, cycles per nanoseconds on my test system is 2.8.
> Fix by using floating point where needed.
>
> Signed-off-by: Stephen Hemminger
> ---
> lib/latency
On Wed, Apr 17, 2024 at 10:07:25AM -0700, Stephen Hemminger wrote:
> The cycle counts do not need to be stored as floating point.
> Instead keep track of latency in cycles, and convert to
> nanoseconds when read.
>
> Change Exponential Weighted Moving Average weight from .2 to .25
> to avoid use o
On Wed, Apr 17, 2024 at 10:07:26AM -0700, Stephen Hemminger wrote:
> All messages that because of an error should be at log level
> NOTICE or above. Do not break log messages across lines.
>
> Signed-off-by: Stephen Hemminger
> ---
> lib/latencystats/rte_latencystats.c | 30 ++---
Vmxnet3 uses capability registers to advertize the supported
capabilities of UPT device. It uses DCR0_REG command register
for this purpose. However, the register command enum misses
one command which technically is not used by dpdk vmxnet3
driver yet. This can cause issue for commands added later.
From: maoyingming
Fixes: 2542ad53d8 (doc: add description of the offload API)
Fixes: ae2c2cb606 (net/cnxk: avoid command copy from Tx queue)
Reviewed-by: lishuo02
Signed-off-by: maoyingming
---
doc/guides/prog_guide/mbuf_lib.rst | 8
drivers/net/cnxk/cn10k_tx.h| 2 +-
drivers
On Wed, Apr 17, 2024 at 4:17 PM Luca Vizzarro wrote:
>
> On 17/04/2024 14:06, Juraj Linkeš wrote:
> > Actually, filtering out the last prompt is fine. I got this mixed up
> > with processing the last prompt. We must process the last prompt, but
> > we don't need to print it (and there's not much r
Vmxnet3 uses capability registers to advertise the supported
capabilities of UPT device. It uses DCR0_REG command register
for this purpose. However, the register command enum misses
one command which technically is not used by dpdk vmxnet3
driver yet. This can cause issue for commands added later.
On Wed, Apr 17, 2024 at 5:29 PM Luca Vizzarro wrote:
>
> On 17/04/2024 15:25, Luca Vizzarro wrote:
> > On 17/04/2024 14:22, Juraj Linkeš wrote:
> >>> I'll
> >>> experiment with some look ahead constructs. The easiest solution is to
> >>> match everything that is not * ([^*]+) but can we be certain
> - int32_t positions[num_keys];
> + int32_t *positions = alloca(sizeof(int32_t) * num_keys);
A general comment, using this patch as an example...
I wonder if adding const has any effect on the compilers' optimizers:
int32_t * const positions = alloca(sizeof(int32_t) * num_keys);
Perhap
> MSVC does not support optional C11 VLAs. When building for Windows
> enable -Wvla so that mingw and clang also fail if a VLA is used.
Minor detail, doesn't affect my Ack for the series...
Applications built for Windows with mingw and clang might use VLAs in the
application itself.
Perhaps we
> From: Tyler Retzlaff [mailto:roret...@linux.microsoft.com]
> Sent: Thursday, 18 April 2024 01.42
>
> As per guidance technical board meeting 2024/04/17. This series
> removes the use of VLAs from code built for Windows for all 3
> toolchains. If there are additional opportunities to convert VLAs
77 matches
Mail list logo
