#
March 20, 2025
Attendees
1. Patrick Robb
2. Paul Szczepanek
3. Luca Vizzarro
4. Cody Cheng
5. Aaron Conole
6. Matthew McGovern
7. Dean Marx
#
Minutes
=
On 2025-03-28 06:43, Mattias Rönnblom wrote:
Prefetch mbuf headers, resulting in ~10% throughput improvement when
the Ethernet RX and TX Adapters are hosted on the same core (likely
~2x in case a dedicated TX core is used).
Signed-off-by: Mattias Rönnblom
Tested-by: Peter Nilsson
What should
add support scatter multi segment received.
Signed-off-by: Wenbo Cao
---
doc/guides/nics/features/rnp.ini | 1 +
doc/guides/nics/rnp.rst | 2 +
drivers/net/rnp/rnp_ethdev.c | 2 +
drivers/net/rnp/rnp_rxtx.c | 130 ++-
drivers/net/rnp/rnp_rxtx.
Prefetch mbuf headers, resulting in ~10% throughput improvement when
the Ethernet RX and TX Adapters are hosted on the same core (likely
~2x in case a dedicated TX core is used).
Signed-off-by: Mattias Rönnblom
Tested-by: Peter Nilsson
---
lib/eventdev/rte_event_eth_tx_adapter.c | 20 ++
add support parse hw packet types result.
Signed-off-by: Wenbo Cao
---
doc/guides/nics/features/rnp.ini | 1 +
doc/guides/nics/rnp.rst | 2 ++
drivers/net/rnp/base/rnp_bdq_if.h | 4 +++
drivers/net/rnp/rnp_ethdev.c | 31 +
drivers/net/rnp/rnp_rxtx.c
This patch add support poll/irq link get mode.
Signed-off-by: Wenbo Cao
---
doc/guides/nics/features/rnp.ini | 2 +
doc/guides/nics/rnp.rst | 3 +
drivers/net/rnp/base/rnp_fw_cmd.c | 45
drivers/net/rnp/base/rnp_fw_cmd.h | 55 +
drivers/net/rnp/base/rnp_hw.h | 3 +
add set link_down/link_up implement
Signed-off-by: Wenbo Cao
---
doc/guides/nics/rnp.rst | 2 +
drivers/net/rnp/base/rnp_eth_regs.h | 3 +
drivers/net/rnp/base/rnp_fw_cmd.c | 22 +++
drivers/net/rnp/base/rnp_fw_cmd.h | 6 ++
drivers/net/rnp/base/rnp_mbx_fw.c | 33 +++
add basic support for device to start/stop function.
Signed-off-by: Wenbo Cao
---
doc/guides/nics/rnp.rst | 2 +
drivers/net/rnp/base/rnp_common.c | 22 +++
drivers/net/rnp/base/rnp_common.h | 1 +
drivers/net/rnp/base/rnp_dma_regs.h | 10 +
drivers/net/rnp/base/rnp_eth_reg
add support mac eth rx tx hw xstats.
Signed-off-by: Wenbo Cao
---
doc/guides/nics/features/rnp.ini| 1 +
doc/guides/nics/rnp.rst | 3 +
drivers/net/rnp/base/rnp_eth_regs.h | 4 +
drivers/net/rnp/base/rnp_mac_regs.h | 80 ++
drivers/net/rnp/rnp.h | 44
add mac filter for single/multiple port.
Signed-off-by: Wenbo Cao
---
doc/guides/nics/features/rnp.ini| 1 +
doc/guides/nics/rnp.rst | 1 +
drivers/net/rnp/base/rnp_crc32.c| 18 -
drivers/net/rnp/base/rnp_crc32.h| 1 +
drivers/net/rnp/base/rnp_eth_regs.h | 3
add mtu update limit for multiple port mode.
multiple mode just used the max-mtu of ports
to limit receive.
Signed-off-by: Wenbo Cao
---
doc/guides/nics/features/rnp.ini| 1 +
doc/guides/nics/rnp.rst | 3 +
drivers/net/rnp/base/rnp_eth_regs.h | 3 +
drivers/net/rnp/rnp.h
add support two method of MAC unicast promisc
mulcast promisc broadcast promisc mode
Signed-off-by: Wenbo Cao
---
doc/guides/nics/features/rnp.ini| 2 +
doc/guides/nics/rnp.rst | 11 +++
drivers/net/rnp/base/rnp_common.c | 5 ++
drivers/net/rnp/base/rnp_eth_regs.h | 15 ++
add platform method for get rx/tx burst function select
by upload func name.
Signed-off-by: Wenbo Cao
---
doc/guides/nics/rnp.rst | 2 ++
drivers/net/rnp/rnp_ethdev.c | 2 ++
drivers/net/rnp/rnp_rxtx.c | 58
drivers/net/rnp/rnp_rxtx.h | 6
4
add support get queue configure info for user debug
Signed-off-by: Wenbo Cao
---
doc/guides/nics/rnp.rst | 2 ++
drivers/net/rnp/rnp_ethdev.c | 2 ++
drivers/net/rnp/rnp_rxtx.c | 42
drivers/net/rnp/rnp_rxtx.h | 4
4 files changed, 50 inserti
add support rx vlan strip,filter,tx vlan/qinq insert.
Signed-off-by: Wenbo Cao
---
doc/guides/nics/features/rnp.ini| 2 +
doc/guides/nics/rnp.rst | 4 +
drivers/net/rnp/base/rnp_bdq_if.h | 2 +-
drivers/net/rnp/base/rnp_eth_regs.h | 5 +
drivers/net/rnp/base/rnp_hw.h
add support to update vid for vlan filter.
Signed-off-by: Wenbo Cao
---
doc/guides/nics/features/rnp.ini| 1 +
doc/guides/nics/rnp.rst | 3 +-
drivers/net/rnp/base/meson.build| 1 +
drivers/net/rnp/base/rnp_bitrev.h | 64
drivers/net/rnp/base/rnp_crc
add mac filter for single/multiple port.
Signed-off-by: Wenbo Cao
---
doc/guides/nics/features/rnp.ini| 1 +
doc/guides/nics/rnp.rst | 6 ++
drivers/net/rnp/base/rnp_eth_regs.h | 4 ++
drivers/net/rnp/base/rnp_hw.h | 3 +
drivers/net/rnp/base/rnp_mac.c | 91 +++
add support multiple segs mbuf send.
Signed-off-by: Wenbo Cao
---
drivers/net/rnp/rnp_ethdev.c | 2 +
drivers/net/rnp/rnp_rxtx.c | 126 ++-
drivers/net/rnp/rnp_rxtx.h | 3 +-
3 files changed, 128 insertions(+), 3 deletions(-)
diff --git a/drivers/net/rnp/
add only support simple recv pkts.
Signed-off-by: Wenbo Cao
---
drivers/net/rnp/rnp_ethdev.c | 7 ++
drivers/net/rnp/rnp_rxtx.c | 130 +++
drivers/net/rnp/rnp_rxtx.h | 5 ++
3 files changed, 142 insertions(+)
diff --git a/drivers/net/rnp/rnp_ethdev.c b/dr
support rx/tx queue stop/start,for rx queue stop
need to reset a queue,must stop all rx queue
during reset this queue.
Signed-off-by: Wenbo Cao
Reviewed-by: Stephen Hemminger
---
doc/guides/nics/features/rnp.ini | 1 +
doc/guides/nics/rnp.rst | 4 +
drivers/net/rnp/base/rnp_commo
add firmware communic method and basic device
init, uninit and close resource function.
Signed-off-by: Wenbo Cao
Reviewed-by: Ferruh Yigit
---
doc/guides/nics/rnp.rst | 10 +
drivers/net/rnp/base/meson.build| 4 +
drivers/net/rnp/base/rnp_common.c | 73 ++
drivers/net/
support tx/rx queue setup and release add hw bd
queue reset,sw queue reset.
Signed-off-by: Wenbo Cao
---
doc/guides/nics/rnp.rst | 3 +
drivers/net/rnp/base/meson.build| 1 +
drivers/net/rnp/base/rnp_bdq_if.c | 398 +++
drivers/net/rnp/base/rnp_bdq_if.h
add get device hardware capability function
Signed-off-by: Wenbo Cao
---
doc/guides/nics/features/rnp.ini | 1 +
doc/guides/nics/rnp.rst | 1 +
drivers/net/rnp/base/rnp_fw_cmd.c | 20 ++
drivers/net/rnp/base/rnp_fw_cmd.h | 80 +
drivers/net/rnp/base/rnp_m
add log function for trace or debug
Signed-off-by: Wenbo Cao
---
drivers/net/rnp/rnp_ethdev.c | 2 ++
drivers/net/rnp/rnp_logs.h | 36
2 files changed, 38 insertions(+)
create mode 100644 drivers/net/rnp/rnp_logs.h
diff --git a/drivers/net/rnp/rnp_ethdev
Add basic PCIe ethdev probe and remove.
Signed-off-by: Wenbo Cao
Reviewed-by: Stephen Hemminger
---
doc/guides/nics/rnp.rst | 10 ++
drivers/net/rnp/rnp.h| 16 +
drivers/net/rnp/rnp_ethdev.c | 66
3 files changed, 92 insertions(+)
c
For This patchset just to support the basic chip init work
and user can just found the eth_dev, but can't control more.
For Now just support 2*10g nic,the chip can support
2*10g,4*10g,4*1g,8*1g,8*10g.
The Feature rx side can support rx-cksum-offload,rss,vlan-filter
flow_clow,uncast_filter,mcast_fil
Implement the eram tables read and write operations
by dtb channel.
Signed-off-by: Bingbin Chen
---
drivers/net/zxdh/zxdh_np.c | 545 -
drivers/net/zxdh/zxdh_np.h | 15 +
2 files changed, 558 insertions(+), 2 deletions(-)
diff --git a/drivers/net/zxdh/zxdh_n
> From: Bruce Richardson [mailto:bruce.richard...@intel.com]
> Sent: Thursday, 27 March 2025 18.16
>
> On Wed, Feb 26, 2025 at 03:59:22PM +, Morten Brørup wrote:
> > The comparisons lcore_id < RTE_MAX_LCORE and lcore_id != LCORE_ID_ANY
> are
> > equivalent, but the latter compiles to fewer byt
On Thu, Mar 27, 2025 at 2:37 PM David Marchand
wrote:
>
> So far, each DPDK library (or driver) exposing symbols in an ABI had to
> maintain a version.map and use some macros for symbol versioning,
> specially crafted with the GNU linker in mind.
>
> This series proposes to rework the whole princi
On Thu, Mar 27, 2025 at 2:50 PM David Marchand
wrote:
>
> Annotate symbols with newly introduced export macros.
>
> For code not compiled via drivers/meson.build (think base drivers code),
> the exported symbols are added in some new .c listed in the sources so they
> get caught by drivers/meson.b
On Fri, 21 Mar 2025 15:31:46 +0800
Bingbin Chen wrote:
> v6:
> - Use byte stream to process bar channel return messages
> and modify some issues to ensure proper functionality.
>
> v5:
> - the VF read/write table function is fixed.
>
> v4:
> - change some tables to const.
> - use %u
On Wed, Mar 26, 2025 at 11:39:28AM +0100, David Marchand wrote:
> Caught in OBS for Fedora Rawhide on aarch64:
>
> [ 198s] In file included from ../lib/acl/acl_run_neon.h:7,
> [ 198s] from ../lib/acl/acl_run_neon.c:5:
> [ 198s] In function ‘alloc_completion’,
> [ 198s] inl
On Wed, 26 Feb 2025 15:41:47 +0800
Wenbo Cao wrote:
> For This patchset just to support the basic chip init work
> and user can just found the eth_dev, but can't control more.
> For Now just support 2*10g nic,the chip can support
> 2*10g,4*10g,4*1g,8*1g,8*10g.
> The Feature rx side can support rx
On Wed, Feb 26, 2025 at 03:59:22PM +, Morten Brørup wrote:
> The comparisons lcore_id < RTE_MAX_LCORE and lcore_id != LCORE_ID_ANY are
> equivalent, but the latter compiles to fewer bytes of code space.
> Similarly for lcore_id >= RTE_MAX_LCORE and lcore_id == LCORE_ID_ANY.
>
> The rte_mempool
On Thu, Mar 27, 2025 at 04:53:27AM -0500, tirthendu.sarkar wrote:
> Streamline code for AVX512 and SSE by removing separate files for them
> and adding runtime checks for selecting appropriate paths based on CPU
> capability. Also, update meson build file.
>
> Signed-off-by: Tirthendu Sarkar
> --
Remove obsolete, and generally not followed, advice in the coding
standards doc, around local variables.
Signed-off-by: Bruce Richardson
---
doc/guides/contributing/coding_style.rst | 4
1 file changed, 4 deletions(-)
diff --git a/doc/guides/contributing/coding_style.rst
b/doc/guides/cont
On Sat, Mar 22, 2025 at 12:59:11PM +0200, Gregory Etelson wrote:
> The patch converts include files with DPDK API to RUST and binds new
> RUST API files into raw module under dpdk crate.
>
> The RUST files and DPDK libraries build from C sources
> allow creation of DPDK application in RUST.
>
> R
> Hi Gregory,
>
> > Arm meson.build unconditionally relays on the `extra_features`
> > member of SoC configuration.
> >
> > SoC dictionary is populated for Arm CPUs with generic instruction set
> > only.
> > For Arm CPUs with native CPU instruction set the SoC dictionary is
> > empty.
> >
> > meso
On Thu, Mar 27, 2025 at 11:00:40AM +0200, Etelson, Gregory wrote:
> Hello Morten,
>
> Thank you for raising these questions !
>
> >
> > Do we want the DPDK project itself to support rust?
> > Or should parts of this be a DPDK hosted project, like grout?
>
> Rust packages management is different
Merge in additional fields used by the idpf driver and then convert it
over to using the common Tx queue structure
Signed-off-by: Shaiq Wani
---
drivers/net/intel/common/tx.h | 20 +++
drivers/net/intel/cpfl/cpfl_ethdev.c | 3 +-
drivers/net/intel/cpfl/cpfl_ethdev.h
On Thu, Mar 27, 2025 at 6:49 AM Eldho Paul Mathew wrote:
>
> Hi All,
>
> What is the importance of the 'ice' driver?
>
> Right now I am using 18.05.1. Does it have 'ice' driver support?
>
> We bought a new card: BCM57508 NetXtreme-E 10Gb/25Gb/40Gb/50Gb/100Gb/200Gb
> Ethernet
>
> To support this c
Display the version when a symbol was introduced.
This is needed for scripting the conversion from static to dynamically
generated version maps.
It is also useful when listing experimental symbols.
Signed-off-by: David Marchand
---
buildtools/map-list-symbol.sh | 2 +-
1 file changed, 1 insertio
Declaring inline helpers in version.map is unnecessary.
There is no exported symbol and this will probably be treated as an
error by MSVC linker.
eal_log_journal has no implementation and can be removed.
Signed-off-by: David Marchand
---
lib/eventdev/version.map | 12
lib/graph/ver
Switch the idpf driver to use the common Tx free function for
AVX2 and AVX512
Signed-off-by: Shaiq Wani
---
.../net/intel/idpf/idpf_common_rxtx_avx2.c| 68 +
.../net/intel/idpf/idpf_common_rxtx_avx512.c | 237 +-
2 files changed, 22 insertions(+), 283 deletions(-)
diff
Used the common Tx entry structure and common Tx mbuf ring replenish fn
in place of idpf-specific structure and function.
The vector driver code paths (AVX2, AVX512) use the smaller SW
ring structure.
Signed-off-by: Shaiq Wani
---
drivers/net/intel/idpf/idpf_common_rxtx.c | 26 --
Align the Tx queue struct field names in idpf and
cpfl driver with the common Tx queue struct.
Signed-off-by: Shaiq Wani
---
drivers/net/intel/cpfl/cpfl_rxtx.c| 40 -
drivers/net/intel/cpfl/cpfl_rxtx_vec_common.h | 4 +-
drivers/net/intel/idpf/idpf_common_rxtx.c | 72 +++
reworked the drivers to use the common functions
and structures from drivers/net/intel/common.
Shaiq Wani (4):
net/intel: align Tx queue struct field names
net/intel: use common Tx queue structure
net/intel: use common Tx entry structure
net/idpf: use common Tx free fn in idpf
drivers/ne
According to datasheet for 82599 and similar NICs, the Tx path must set
CC (Check Context) bit in the descriptor in cases when Tx switching is
enabled in the PF. In case of using VF functions, the fact that it is a
VF function implies that Tx switching is enabled, and thus the CC bit
must always be
When adding E610 support, some case statements for VF definitions were
missing, resulting in picking incorrect registers to write to in certain
cases. Fix it by adding necessary case statements for E610 VF.
Fixes: 316637762a5f ("net/ixgbe/base: enable E610 device")
Cc: sta...@dpdk.org
Signed-off-
In a couple of places, we are using explicit mac type comparisons to
trigger a VF-specific path, but we already have a function that does the
same thing, so use it.
Signed-off-by: Anatoly Burakov
---
drivers/net/intel/ixgbe/ixgbe_rxtx.c | 17 +
1 file changed, 5 insertions(+), 12
The definition of IXGBE_LE32_TO_CPUS macro is meant to modify the value
in place - similar to the le32_to_cpus() macro in kernel. Fixing the
definition allows us to remove some warning flags, and removes the need
for the uintptr_t typecasts.
Fixes: aa4fc14d2cee ("ixgbe: update base driver")
Cc: st
Now that base-code warning flags have been removed from a number of
drivers, we can simplify their build configuration by just including the
base code files in the regular list of driver files passed back. There
is no need to use pre-compiled objects.
Signed-off-by: Bruce Richardson
---
drivers/
Improve base code macros to reduce the number of issues with unused
variables in the code. Issues still remain with unused-but-set
variables, but completely unused variable warnings are eliminated.
Signed-off-by: Bruce Richardson
---
drivers/net/intel/ice/base/ice_osdep.h | 6 +++---
drivers/ne
Add a single-line fix to the base code, and then the remaining two
compiler warning disable flags can be removed from the driver base code
build file.
Fixes: 8db9e2a1b232 ("i40e: base driver")
Cc: sta...@dpdk.org
Signed-off-by: Bruce Richardson
---
drivers/net/intel/i40e/base/i40e_diag.c | 2 +
Fix warnings about unused values - parameters, variables, etc., and
remove the warning disable flags for them. Although modifying the
base-code files is not ideal, the changes required are minor, and only
affect two files from the imported base code.
Fixes: 8db9e2a1b232 ("i40e: base driver")
Cc: s
When building on FreeBSD, errors are reported in the base code by the
lock checker (-Wthread-safety). For example:
../drivers/net/intel/ixgbe/base/ixgbe_osdep.c:42:1: error: mutex 'lock->mutex'
is still held at the end of function [-Werror,-Wthread-safety-analysis]
42 | }
| ^
These erro
The fixes required to re-enable warnings in the fm10k base code are
trivial, so let's make the changes and get a clean compile without any
warning disable flags.
* provide definitions for the UNREFERENCED_PARAMETER macros
* fix the spelling of the work "fallthrough" in comments
* provide a definit
We can remove almost all of the "unused parameter" and "unused variable"
warnings by just improving the macro definitions in the osdep.h header.
Remaining two instances can be fixed by just one-line additions to the
code, so add those to give us a clean build with the warnings enabled.
Fixes: af75
Simplify the build of the various intel base code directories, by
clearing compiler warnings so the files can be compiled directly along
with the regular driver files.
v3: add fix for lock checker issues on FreeBSD
v2: added missing patch for fm10k
Bruce Richardson (9):
net/fm10k/base: fix comp
The meson.build file in the base folder was never used, as the base
files were just included in the main sources list for iavf.
Fixes: f1fdc9ddba5e ("drivers: move iavf common folder to iavf net")
Signed-off-by: Bruce Richardson
---
drivers/net/intel/iavf/base/meson.build | 10 --
1 fil
When adding E610 support, some case statements for VF definitions were
missing, resulting in picking incorrect registers to write to in certain
cases. Fix it by adding necessary case statements for E610 VF.
Fixes: 316637762a5f ("net/ixgbe/base: enable E610 device")
Cc: sta...@dpdk.org
Signed-off-
According to datasheet for 82599 and similar NICs, the Tx path must set
CC (Check Context) bit in the descriptor in cases when Tx switching is
enabled in the PF. In case of using VF functions, the fact that it is a
VF function implies that Tx switching is enabled, and thus the CC bit
must always be
In a couple of places, we are using explicit mac type comparisons to
trigger a VF-specific path, but we already have a function that does the
same thing, so use it.
Signed-off-by: Anatoly Burakov
---
drivers/net/intel/ixgbe/ixgbe_rxtx.c | 17 +
1 file changed, 5 insertions(+), 12
Declaring inline helpers in version.map is unnecessary.
There is no exported symbol and this will probably be treated as an
error by MSVC linker.
rte_dpaa2_dev_type is an enum token.
roc_se_ctx_swap has no implementation (leftover from a previous rework)
and can be removed.
Signed-off-by: David M
Hi All,
What is the importance of the 'ice' driver?
Right now I am using 18.05.1. Does it have 'ice' driver support?
We bought a new card: BCM57508 NetXtreme-E 10Gb/25Gb/40Gb/50Gb/100Gb/200Gb
Ethernet
To support this card using the 'ice' driver, is the DPDK 18.05.1 version
sufficient, or is the
Thanks Stephen.
I checked this module is not present in 20.11 dpdk version, but its in 23.11
But can you suggest if I need to port this back to 20.11, any necessary steps I
can follow or any dependencies with the new version etc,
Thanks
Rajesh
-Original Message-
From: Stephen Hemminger
For versioning symbols:
- MSVC uses pragmas on the symbol,
- GNU linker uses special asm directives,
To accommodate both GNU linker and MSVC linker, introduce new macros for
exporting and versioning symbols that will surround the whole function.
This has the advantage of hiding all the ugly detai
Switch to dynamically generated version maps.
As the map files get generated, tooling around checking, converting,
updating etc.. static version maps can be removed.
Signed-off-by: David Marchand
---
.github/workflows/build.yml | 1 -
MAINTAINERS
Rather than maintain a file in parallel of the code, symbols to be
exported can be marked with a token RTE_EXPORT_*SYMBOL.
>From those marks, the build framework generates map files only for
symbols actually compiled (which means that the WINDOWS_NO_EXPORT hack
becomes unnecessary).
The build fra
So far, each DPDK library (or driver) exposing symbols in an ABI had to
maintain a version.map and use some macros for symbol versioning,
specially crafted with the GNU linker in mind.
This series proposes to rework the whole principle, and instead rely on
marking the symbol exports in the source
> On Thu, Mar 27, 2025 at 12:10:12PM +, Konstantin Ananyev wrote:
> >
> >
> > > > >
> > > > > > -Original Message-
> > > > > > From: David Marchand
> > > > > > Sent: Thursday, March 27, 2025 10:37 AM
> > > > > > To: Bruce Richardson
> > > > > > Cc: dev@dpdk.org; bl...@debian.org; st
> From: Konstantin Ananyev [mailto:konstantin.anan...@huawei.com]
> Sent: Thursday, 27 March 2025 13.10
>
> > > >
> > > > > From: David Marchand
> > > > > Sent: Thursday, March 27, 2025 10:37 AM
> > > > >
> > > > > On Thu, Mar 27, 2025 at 9:55 AM Bruce Richardson
> > > > > wrote:
> > > > > >
> >
> > >
> > > > -Original Message-
> > > > From: David Marchand
> > > > Sent: Thursday, March 27, 2025 10:37 AM
> > > > To: Bruce Richardson
> > > > Cc: dev@dpdk.org; bl...@debian.org; sta...@dpdk.org; Konstantin
> > Ananyev ; David Christensen
> > > > ; Wathsala Vithanage
> >
> > > > Su
On Thu, Mar 27, 2025 at 12:10:12PM +, Konstantin Ananyev wrote:
>
>
> > > >
> > > > > -Original Message-
> > > > > From: David Marchand
> > > > > Sent: Thursday, March 27, 2025 10:37 AM
> > > > > To: Bruce Richardson
> > > > > Cc: dev@dpdk.org; bl...@debian.org; sta...@dpdk.org; Kon
https://bugs.dpdk.org/show_bug.cgi?id=1684
Bug ID: 1684
Summary: Colt Python holster
Product: DPDK
Version: 24.03
Hardware: All
OS: All
Status: UNCONFIRMED
Severity: normal
Priority: Normal
On 2025/3/13 23:37, Stephen Hemminger wrote:
Caution: This is an external email. Please be very careful when clicking links
or opening attachments. See http://nok.it/nsb for additional information.
On Thu, 13 Mar 2025 14:19:03 +0800
Yang Ming wrote:
This change ensures that the current pro
> From: Bruce Richardson [mailto:bruce.richard...@intel.com]
> Sent: Thursday, 27 March 2025 11.51
>
> On Thu, Mar 27, 2025 at 10:39:09AM +, Konstantin Ananyev wrote:
> >
> >
> > > -Original Message-
> > > From: David Marchand
> > > Sent: Thursday, March 27, 2025 10:37 AM
> > > To: Br
From: Pavan Nikhilesh
Since driver/common/cnxk has common code that is
used across multiple platforms, allow use of
GCC builtin atomic functions.
Signed-off-by: Pavan Nikhilesh
---
Rest of drivers i.e., drivers/*/cnxk have been migrated to use stdatomics in:
https://patchwork.dpdk.org/projec
From: Pavan Nikhilesh
Replace GCC inbuilt atomic functions with rte_atomic_xxx API.
Signed-off-by: Pavan Nikhilesh
---
drivers/crypto/cnxk/cn10k_cryptodev_ops.c | 36 ++-
drivers/crypto/cnxk/cn9k_cryptodev_ops.c | 20 +--
drivers/net/cnxk/cn10k_tx.h | 42
On 27/03/2025 07:55, Morten Brørup wrote:
> + Red Hat tech board members
>
>> From: Stephen Hemminger [mailto:step...@networkplumber.org]
>> Sent: Wednesday, 26 March 2025 20.21
>>
>> On Wed, 26 Mar 2025 19:06:58 +0100
>> Morten Brørup wrote:
>>
From: Bruce Richardson [mailto:bruce.richard..
Align the Tx queue struct field names in idpf and
cpfl driver with the common Tx queue struct.
Signed-off-by: Shaiq Wani
---
drivers/net/intel/cpfl/cpfl_rxtx.c| 40 -
drivers/net/intel/cpfl/cpfl_rxtx_vec_common.h | 4 +-
drivers/net/intel/idpf/idpf_common_rxtx.c | 72 +++
> From: Dengdui Huang [mailto:huangdeng...@huawei.com]
> Sent: Thursday, 27 March 2025 10.01
>
> The worker_id may come from user input.
> So it is necessary to verify it.
>
> Fixes: a95d70547c57 ("eal: factorize lcore main loop")
> Cc: sta...@dpdk.org
>
> Signed-off-by: Dengdui Huang
> ---
>
On Thu, Mar 27, 2025 at 10:39:09AM +, Konstantin Ananyev wrote:
>
>
> > -Original Message-
> > From: David Marchand
> > Sent: Thursday, March 27, 2025 10:37 AM
> > To: Bruce Richardson
> > Cc: dev@dpdk.org; bl...@debian.org; sta...@dpdk.org; Konstantin Ananyev
> > ; David Christens
Switch the idpf driver to use the common Tx free function for
AVX2 and AVX512.
Signed-off-by: Shaiq Wani
---
.../net/intel/idpf/idpf_common_rxtx_avx2.c| 10 +-
.../net/intel/idpf/idpf_common_rxtx_avx512.c | 237 +-
2 files changed, 22 insertions(+), 225 deletions(-)
diff -
Merge in additional fields used by the idpf driver and then convert it
over to using the common Tx queue structure.
Signed-off-by: Shaiq Wani
---
drivers/net/intel/common/tx.h | 20 +++
drivers/net/intel/cpfl/cpfl_ethdev.c | 3 +-
drivers/net/intel/cpfl/cpfl_ethdev.
reworked the drivers to use the common functions and structures
from drivers/net/intel/common.
Shaiq Wani (4):
net/intel: align Tx queue struct field names
net/intel: use common Tx queue structure
net/intel: use common Tx entry structure
net/idpf: use common Tx free fn in idpf
drivers/ne
> -Original Message-
> From: David Marchand
> Sent: Thursday, March 27, 2025 10:37 AM
> To: Bruce Richardson
> Cc: dev@dpdk.org; bl...@debian.org; sta...@dpdk.org; Konstantin Ananyev
> ; David Christensen
> ; Wathsala Vithanage
> Subject: Re: [PATCH] acl: fix build with GCC 15 on aarc
On Thu, Mar 27, 2025 at 9:55 AM Bruce Richardson
wrote:
>
> On Wed, Mar 26, 2025 at 11:39:28AM +0100, David Marchand wrote:
> > Caught in OBS for Fedora Rawhide on aarch64:
> >
> > [ 198s] In file included from ../lib/acl/acl_run_neon.h:7,
> > [ 198s] from ../lib/acl/acl_run_neo
>
> Caught in OBS for Fedora Rawhide on aarch64:
>
> [ 198s] In file included from ../lib/acl/acl_run_neon.h:7,
> [ 198s] from ../lib/acl/acl_run_neon.c:5:
> [ 198s] In function ‘alloc_completion’,
> [ 198s] inlined from ‘acl_start_next_trie’ at
> ../lib/acl/acl_r
After CPU isolation is configured, an infinite loop occurs when
dma-perf is executed using the default config file.
This patchset fix it.
Dengdui Huang (2):
eal: fix uncheck worker ID
app/dma-perf: fix infinite loop
app/test-dma-perf/main.c | 8 +++-
lib/eal/common/eal_common_
On 2025/3/17 21:56, Stephen Hemminger wrote:
Caution: This is an external email. Please be very careful when clicking links
or opening attachments. See http://nok.it/nsb for additional information.
On Fri, 14 Mar 2025 18:36:38 +0800
Yang Ming wrote:
The secordary process should not close s
When a core that is not used by the rte is specified in the config
for testing, the problem of infinite loop occurs. This patch fix it
by adding a check for lcore_id.
Fixes: 623dc9364dc6 ("app/dma-perf: introduce DMA performance test")
Cc: sta...@dpdk.org
Signed-off-by: Dengdui Huang
---
app/te
The direct passthrough interrupt mechanism, which allowed bypassing the
kernel, was obscure and is no longer supported. So this driver won't
work with latest SDK kernels. Additionally, the sysfs GPIO control
interface has been deprecated by Linux kernel itself.
That said, this change updates the P
When a core that is not used by the rte is specified in the config
for testing, the problem of infinite loop occurs. This patch fix it
by adding a check for lcore_id.
Fixes: 623dc9364dc6 ("app/dma-perf: introduce DMA performance test")
Cc: sta...@dpdk.org
Signed-off-by: Dengdui Huang
---
app/te
Hello Morten,
Thank you for raising these questions !
Do we want the DPDK project itself to support rust?
Or should parts of this be a DPDK hosted project, like grout?
Rust packages management is different.
Also DPDK Rust code will eventually provide a different API.
At this stage, DPDK host
The worker_id may come from user input.
So it is necessary to verify it.
Fixes: a95d70547c57 ("eal: factorize lcore main loop")
Cc: sta...@dpdk.org
Signed-off-by: Dengdui Huang
---
lib/eal/common/eal_common_launch.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/lib/eal/common/eal_commo
When a core that is not used by the rte is specified in the config
for testing, the problem of infinite loop occurs. This patch fix it
by adding a check for lcore_id.
Fixes: 623dc9364dc6 ("app/dma-perf: introduce DMA performance test")
Cc: sta...@dpdk.org
Signed-off-by: Dengdui Huang
---
app/te
On Wed, Mar 26, 2025 at 11:39:28AM +0100, David Marchand wrote:
> Caught in OBS for Fedora Rawhide on aarch64:
>
> [ 198s] In file included from ../lib/acl/acl_run_neon.h:7,
> [ 198s] from ../lib/acl/acl_run_neon.c:5:
> [ 198s] In function ‘alloc_completion’,
> [ 198s] inl
> From: Gregory Etelson [mailto:getel...@nvidia.com]
> Sent: Thursday, 6 March 2025 14.37
> Subject: [PATCH] rust: support DPDK API
>
> The patch converts include files with DPDK API to RUST and binds new
> RUST API files info dpdklib package.
>
> The RUST dpdklib files and DPDK libraries build f
Hello,
On Thu, Mar 27, 2025 at 9:02 AM Yang Ming wrote:
> On 2025/3/13 23:37, Stephen Hemminger wrote:
> > On Thu, 13 Mar 2025 14:19:03 +0800
> > Yang Ming wrote:
> >
> >> This change ensures that the current process is checked for
> >> being run with 'setarch' before verifying the value of
> >>
1 - 100 of 106 matches
Mail list logo
