Bruce,
With this patch [1], it is noted that the ring producer and consumer data
should not be on adjacent cache lines, for performance reasons.
[1]:
https://git.dpdk.org/dpdk/commit/lib/librte_ring/rte_ring.h?id=d9f0d3a1ffd4b66e75485cc8b63b9aedfbdfe8b0
(It's obvious that they cannot share the
Hi Ali,
Thanks for testing and report!
Regards,
Xueming
> -Original Message-
> From: Ali Alnubani
> Sent: 8/25/2023 0:02
> To: Xueming(Steven) Li ; sta...@dpdk.org
> Cc: xuemi...@nvdia.com; dev@dpdk.org; Abhishek Marathe
> ; benjamin.wal...@intel.com; David
> Christensen ; Hemant Agrawa
> -Original Message-
> From: Zhang, Yuying
> Sent: Saturday, August 12, 2023 3:55 PM
> To: dev@dpdk.org; Xing, Beilei ; Zhang, Qi Z
> ; Wu, Jingjing
> Cc: Zhang, Yuying
> Subject: [PATCH v1 3/5] net/cpfl: add cpfl control queue message handle
>
> Add cpfl driver control queue message
> -Original Message-
> From: Zhang, Yuying
> Sent: Saturday, August 12, 2023 3:55 PM
> To: dev@dpdk.org; Xing, Beilei ; Zhang, Qi Z
> ; Wu, Jingjing
> Cc: Zhang, Yuying
> Subject: [PATCH v1 2/5] common/idpf/base: refine idpf ctlq message structure
>
> Add cfg data in idpf_ctlq_msg.
On Fri, Aug 25, 2023 at 9:27 AM Ashwin Sekhar T K wrote:
>
> As done in FLR, similar to CQ disable, in process of
> aura disable we need to explicitly clear BP_ENA in order to
> deassert backpressure if it was earlier asserted.
>
> Fixes: f765f5611240 ("common/cnxk: add NPA pool HW operations")
>
As done in FLR, similar to CQ disable, in process of
aura disable we need to explicitly clear BP_ENA in order to
deassert backpressure if it was earlier asserted.
Fixes: f765f5611240 ("common/cnxk: add NPA pool HW operations")
Signed-off-by: Nithin Dabilpuram
Signed-off-by: Ashwin Sekhar T K
--
> -Original Message-
> From: Zhang, Yuying
> Sent: Saturday, August 12, 2023 3:55 PM
> To: dev@dpdk.org; Xing, Beilei ; Zhang, Qi Z
> ; Wu, Jingjing
> Cc: Zhang, Yuying
> Subject: [PATCH v1 1/5] net/cpfl: setup rte flow skeleton
>
> Setup the rte_flow backend skeleton. Introduce the
As done in FLR, similar to CQ disable, in process of
aura disable we need to explicitly clear BP_ENA in order to
deassert backpressure if it was earlier asserted.
Fixes: f765f5611240 ("common/cnxk: add NPA pool HW operations")
Signed-off-by: Nithin Dabilpuram
Signed-off-by: Ashwin Sekhar T K
--
> -Original Message-
> From: Yang, Qiming
> Sent: Monday, August 21, 2023 1:30 PM
> To: dev@dpdk.org
> Cc: Zhang, Qi Z ; Yang, Qiming
>
> Subject: [PATCH] doc: update 23.07 Intel NICs matching list
>
> Updated i40e and ice match list table.
>
> Signed-off-by: Qiming Yang
Applied to
> -Original Message-
> From: Xueming(Steven) Li
> Sent: Thursday, August 17, 2023 9:14 AM
> To: sta...@dpdk.org
> Cc: xuemi...@nvdia.com; dev@dpdk.org; Abhishek Marathe
> ; Ali Alnubani ;
> benjamin.wal...@intel.com; David Christensen ;
> Hemant Agrawal ; Ian Stokes
> ; Jerin Jacob ; John
On Thu, Aug 24, 2023 at 04:18:06PM +0200, David Marchand wrote:
> On Thu, Aug 24, 2023 at 2:47 PM Morten Brørup
> wrote:
> > > From: David Marchand [mailto:david.march...@redhat.com]
> > > However, I am a bit puzzled why the prefetch change makes the compiler
> > > consider this loop differently.
Recheck-request: iol-sample-apps-testing
This is a test of the retesting system, please ignore this email.
On Thu, Aug 3, 2023 at 3:50 AM David Marchand
wrote:
> Rather than rely on Linux headers to find some PCI(e) standard constants
> or reinvent the same PCI capability helper, this series co
Hi David, you can ignore the failures UNH reported for this patch - we made
a mistake reconfiguring jenkins for the refactored unit test suite. As
discussed we are not running the arm64 unit tests until one of the
solutions we talked about is in place.
On Wed, Aug 23, 2023 at 4:33 AM Zhang, Qi Z
> From: David Marchand [mailto:david.march...@redhat.com]
> Sent: Thursday, 24 August 2023 16.18
>
> On Thu, Aug 24, 2023 at 2:47 PM Morten Brørup
> wrote:
> > > From: David Marchand [mailto:david.march...@redhat.com]
> > > However, I am a bit puzzled why the prefetch change makes the compiler
>
On Thu, Aug 24, 2023 at 2:47 PM Morten Brørup
wrote:
> > From: David Marchand [mailto:david.march...@redhat.com]
> > However, I am a bit puzzled why the prefetch change makes the compiler
> > consider this loop differently.
> > We have the same constructs everywhere in this library and x86_64
> >
Hi all,
Here is a new stable release:
https://fast.dpdk.org/rel/dpdk-21.11.5.tar.xz
The git tree is at:
https://dpdk.org/browse/dpdk-stable/?h=21.11
This LTS release contains ~160 fixes from main branch
up to DPDK 23.07.
Thanks to the authors who helped with backports and to the
> From: David Marchand [mailto:david.march...@redhat.com]
> Sent: Thursday, 24 August 2023 14.06
>
> On Fri, Aug 11, 2023 at 9:21 PM Tyler Retzlaff
> wrote:
> >
> > Inline assembly is not supported for MSVC x64 instead use _mm_prefetch
> > and _mm_cldemote intrinsics.
> >
> > Signed-off-by: Tyler
On Thu, Aug 24, 2023 at 2:06 PM David Marchand
wrote:
> I had a look at the vhost library and I think the compiler thinks size may be
> 0.
> Changing the loop on size with a do { } while (size > 0); resolves the
> warning.
> I can post a change for this, as we hit a similar issue in the past
> a
On Wed, Aug 23, 2023 at 12:15 PM Ankur Dwivedi wrote:
>
> Adds sa to port mapping in roc mcs. This helps to propagate the tx and
> rx pn threshold events to the correct ethernet device.
>
> Signed-off-by: Ankur Dwivedi
> ---
> drivers/common/cnxk/roc_mcs.c | 22 +++---
>
On Tue, Aug 22, 2023 at 11:00 PM Ashwin Sekhar T K wrote:
>
> As done in FLR, similar to CQ disable, in process of
> aura disable we needs explicitly clear BP_ENA in order to
> deassert backpressure if it was earlier asserted.
>
> Signed-off-by: Nithin Dabilpuram
> Signed-off-by: Ashwin Sekhar T
On Fri, Aug 11, 2023 at 9:21 PM Tyler Retzlaff
wrote:
>
> Inline assembly is not supported for MSVC x64 instead use _mm_prefetch
> and _mm_cldemote intrinsics.
>
> Signed-off-by: Tyler Retzlaff
> Acked-by: Bruce Richardson
> Acked-by: Morten Brørup
> Acked-by: Konstantin Ananyev
> ---
> lib/e
> From: Mattias Rönnblom [mailto:hof...@lysator.liu.se]
> Sent: Thursday, 24 August 2023 12.53
>
> On 2023-08-24 10:05, Morten Brørup wrote:
> >> From: Honnappa Nagarahalli [mailto:honnappa.nagaraha...@arm.com]
> >> Sent: Tuesday, 22 August 2023 07.47
> >>
> >>> From: Morten Brørup
> >>> Sent: Mo
On 2023-08-22 14:32, Jerin Jacob wrote:
On Tue, Aug 22, 2023 at 2:12 PM Mattias Rönnblom wrote:
On 2023-08-18 08:09, Jerin Jacob wrote:
On Fri, Jun 16, 2023 at 1:17 PM Mattias Rönnblom
wrote:
Hi Mattais,
Finally, got some time to review this series, and thanks for excellent
documentatio
Sync the logic from kernel driver, adjust the definition of structure,
and extend the usage of nfp BAR from 8 to 24.
This will greatly enhance the scalability of nfp PMD.
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/nfp_ethdev.c | 8 +-
drivers/net/n
Sync the logic from kernel driver, use the new entry function from the
PCIe module instead of the cppcore module.
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/nfp_ethdev.c | 9 +-
drivers/net/nfp/nfpcore/nfp6000_pcie.c | 197 ++---
Sync the logic from kernel driver and remove the unneeded header
file include statements.
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/nfpcore/nfp6000_pcie.c | 211 +
drivers/net/nfp/nfpcore/nfp_cpp.h | 9 ++
2 files changed, 150 in
Modify the comment and standard the coding style.
Move the definition of data structure into the implement file to make
the API small.
Also sync the logic from kernel driver and remove the unneeded header
file include statements.
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
driv
Add a new header file for the PCIe module.
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/meson.build | 2 +-
.../nfpcore/{nfp_cpp_pcie_ops.c => nfp6000_pcie.c} | 2 ++
drivers/net/nfp/nfpcore/nfp6000_pcie.h | 13
Add the nfp_dev module to simplify the logic.
The data structure we import can gather and replace several macro
which scatter everywhere. Which also make the logic of choose value
for different devices unnecessary and so eliminate such logic.
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderl
Rename the data fields to sync with the logic in kernel driver.
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/flower/nfp_flower.c | 6 +++---
drivers/net/nfp/nfp_common.h| 4 ++--
drivers/net/nfp/nfp_ethdev.c| 16
3 files changed
Add a header file to holds the API declarations of this module.
Also sync the logic from kernel driver and remove the unneeded header
file include statements.
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/nfpcore/nfp_cpp.h | 16 --
drivers/net/nfp/nfpcore/nf
Move the definition of data structure into the implement file.
Also sync the logic from kernel driver and remove the unneeded header
file include statements.
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/nfp_ethdev.c | 2 +-
drivers/net/nfp/nfpcore/nfp
Split out the macro which are not easily find a DPDK substitute into
a new header file.
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/nfpcore/nfp_nsp.c | 1 +
drivers/net/nfp/nfpcore/nfp_nsp.h | 23 --
drivers/net/nfp/nfpcore/nfp_nsp_eth.c
Move the function declaration into the header file, and sync some
logic from kernel driver.
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/nfpcore/nfp6000/nfp6000.h | 20
drivers/net/nfp/nfpcore/nfp_cpp_pcie_ops.c | 1 +
drivers/net/nfp/nfpco
Sync the logic from kernel driver and remove the unneeded header
file include statements.
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/nfpcore/nfp_resource.c | 21 -
1 file changed, 8 insertions(+), 13 deletions(-)
diff --git a/drivers/net/nf
Add several read/write APIs and remove the unneeded header file
include statements.
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/nfpcore/nfp_rtsym.c | 305 +---
drivers/net/nfp/nfpcore/nfp_rtsym.h | 53 ++---
2 files changed, 299 insertio
Remove the unneeded header file include statements.
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/nfpcore/nfp_mip.c | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git a/drivers/net/nfp/nfpcore/nfp_mip.c
b/drivers/net/nfp/nfpcore/nfp_mip.c
inde
Move the definition of data structure and macro into the implement file.
Also sync the logic from kernel driver and remove the unneeded header file
include statements.
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/nfpcore/nfp_mip.h | 4 +-
drivers/net/nfp/nfpco
Rename some parameter and variable to make the logic easier to
understand.
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/nfpcore/nfp_cpp.h | 10 ++---
drivers/net/nfp/nfpcore/nfp_cpp_pcie_ops.c | 42 -
drivers/net/nfp/nfpcore/nfp_cppcore.c
Move the definition of data structure and macro into the implement file.
Also remove the unneeded header file include statements.
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/nfpcore/nfp_hwinfo.c | 84 +---
drivers/net/nfp/nfpcore/nfp_hwin
Unify the guide line of header file, we choose '__FOO_BAR_H__' style.
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/nfpcore/nfp_cpp.h | 2 +-
drivers/net/nfp/nfpcore/nfp_hwinfo.h | 2 +-
drivers/net/nfp/nfpcore/nfp_mip.h | 2 +-
drivers/net/nfp/nfpcore
Use space character to align instead of TAB character.
There should one blank line to split the block of logic, no more no less.
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/nfpcore/nfp6000/nfp6000.h | 4 +-
drivers/net/nfp/nfpcore/nfp_cpp.h | 18 +--
Follow the DPDK coding style, use the kdoc comment style.
Also add some comment to help understand logic.
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/nfpcore/nfp_cpp.h | 504 -
drivers/net/nfp/nfpcore/nfp_cpp_pcie_ops.c | 39 +-
dri
Revise the logic, using the DPDK memory management API to replace the
malloc()/free().
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/flower/nfp_flower.c| 2 +-
drivers/net/nfp/nfp_ethdev.c | 8
drivers/net/nfp/nfpcore/nfp_cppcore.c | 10 +
Add log statement to the important control logic, which means
more strict check of the return value of function call.
Also remove some verbose info log statement.
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/nfpcore/nfp_cpp_pcie_ops.c | 37 ++
There should only declare one local variable in each line, and the local
variable should be arranged from short to long in the function.
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/nfpcore/nfp_cpp_pcie_ops.c | 58 --
drivers/net/nfp/nfpcore/n
Remove the unneeded logic.
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/nfpcore/nfp_cpp.h | 107
drivers/net/nfp/nfpcore/nfp_cpp_pcie_ops.c | 12 -
drivers/net/nfp/nfpcore/nfp_cppcore.c | 162 --
drivers/net/nfp/nfpcore/nfp_mutex.c
Unify the type of integer variable to the DPDK prefer style.
Also change the return type of 'nfp_eth_speed2rate()' to the
corresponding enum, which make it more readable.
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/nfpcore/nfp_cpp.h | 53 ++-
Each parameter of function should occupy one line, and indent two TAB
character.
All the statement which span multi line should indent two TAB character.
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/nfpcore/nfp_cpp.h | 80 +-
drivers/net/nfp/nfp
To compliance with the coding standard, make the pointer variable
explicitly comparing to 'NULL' and the integer variable explicitly
comparing to '0'.
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/nfpcore/nfp_cpp_pcie_ops.c | 12 ++--
drivers/net/nfp/nfpco
This patch series aims to:
- Make the coding style satisfy with DPDK.
- Sync the logic with kernel driver.
- Make the sub-module more modular.
- Extend the nfp configure BAR from 8 to 24.
Chaoyong He (27):
net/nfp: explicitly compare to null and 0
net/nfp: unify the indent coding style
net/n
On 24/08/2023 07:07, Xu, HailinX wrote:
-Original Message-
From: Kevin Traynor
Sent: Thursday, August 3, 2023 12:31 AM
To: sta...@dpdk.org
Cc: dev@dpdk.org; Abhishek Marathe ; Ali
Alnubani ; Walker, Benjamin
; David Christensen ;
Hemant Agrawal ; Stokes, Ian
; Jerin Jacob ; Mcnamara, Joh
Release status meeting minutes 2023-08-17
=
Agenda:
* Release Dates
* Subtrees
* Roadmaps
* LTS
* Defects
* Opens
Participants:
* AMD
* Intel
* Marvell
* Red Hat
Release Dates
-
The following are the proposed working dates for 23.11:
* V1:
On 2023-08-24 10:05, Morten Brørup wrote:
From: Honnappa Nagarahalli [mailto:honnappa.nagaraha...@arm.com]
Sent: Tuesday, 22 August 2023 07.47
From: Morten Brørup
Sent: Monday, August 21, 2023 2:37 AM
From: Honnappa Nagarahalli [mailto:honnappa.nagaraha...@arm.com]
Sent: Monday, 21 August 20
Remove device from the list before doing actual cleanup to avoid use
after free.
Bugzilla ID: 1276
Fixes: 17c839f74da3 ("bus: add platform bus")
Signed-off-by: Tomasz Duszynski
---
drivers/bus/platform/platform.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/bus/pl
As we made our move to C11, GCC requirements has been bumped.
Wrappers for old versions of GCC can be dropped.
Signed-off-by: David Marchand
Acked-by: Bruce Richardson
---
Changes since v1:
- removed empty #else #endif block in x86 rte_byteorder.h,
- updated lib/table/rte_lru_x86.h (thanks Bruce
> -Original Message-
> From: Qiao, Wenjing
> Sent: Friday, August 11, 2023 5:31 PM
> To: Zhang, Yuying ; Xing, Beilei
>
> Cc: dev@dpdk.org; Liu, Mingxia ; Qiao, Wenjing
>
> Subject: [PATCH 3/4] net/cpfl: introduce CPF common library
>
> Add common library support for CPFL rte_flow to
> -Original Message-
> From: Qiao, Wenjing
> Sent: Friday, August 11, 2023 5:31 PM
> To: Zhang, Yuying ; Xing, Beilei
>
> Cc: dev@dpdk.org; Liu, Mingxia ; Qiao, Wenjing
> ; Zhang, Qi Z
> Subject: [PATCH 4/4] net/cpfl: setup ctrl path
>
> Setup the control vport and control queue for
On 8/23/2023 5:03 PM, Tyler Retzlaff wrote:
> On Wed, Aug 23, 2023 at 10:19:39AM +0100, Ferruh Yigit wrote:
>> On 8/22/2023 11:30 PM, Konstantin Ananyev wrote:
>>> 18/08/2023 14:48, Bruce Richardson пишет:
On Fri, Aug 18, 2023 at 02:25:14PM +0100, Ferruh Yigit wrote:
> On 8/17/2023 3:18 PM
On Thu, Aug 24, 2023 at 10:30:34AM +0200, David Marchand wrote:
> As we made our move to C11, GCC requirements has been bumped.
> Wrappers for old versions of GCC can be dropped.
>
> Signed-off-by: David Marchand
> ---
> lib/eal/arm/include/rte_byteorder.h | 19 ---
> lib/eal
Hello Oliver,
thanks, your response helped a lot, I managed to find the root cause of the
instability which is on our side.
It was due to other internal developments.
I'll still add an error check on the enqueue ops to catch eventual problems
earlier, if that suits you.
Best regards,
Julien
On
From: Julien Hascoet
As the ring is big enough by construction, the ring
enqueue ops cannot fail; therefore, we check and panic
on it as soon as it is detected.
Signed-off-by: Julien Hascoet
---
app/test/test_mbuf.c | 13 ++---
1 file changed, 10 insertions(+), 3 deletions(-)
diff --g
Hello Tyler,
On Thu, Aug 24, 2023 at 10:30 AM David Marchand
wrote:
> diff --git a/lib/eal/include/rte_debug.h b/lib/eal/include/rte_debug.h
> index 2c4b94a7c9..74593cd4d4 100644
> --- a/lib/eal/include/rte_debug.h
> +++ b/lib/eal/include/rte_debug.h
> @@ -60,11 +60,7 @@ void rte_dump_stack(void)
As we made our move to C11, GCC requirements has been bumped.
Wrappers for old versions of GCC can be dropped.
Signed-off-by: David Marchand
---
lib/eal/arm/include/rte_byteorder.h | 19 ---
lib/eal/arm/include/rte_vect.h | 12
lib/eal/include/generic/rt
On passing NULL as a last parameter to xxx_create_ext_with_params(),
rxa_config_params_validate() uses default values and dereferences
NULL pointer leading to segmentation fault.
Fixed by returning after using default values without dereferencing
NULL pointer.
Fixes: 8be6c94d6d90 ("eventdev/eth_r
> From: Honnappa Nagarahalli [mailto:honnappa.nagaraha...@arm.com]
> Sent: Tuesday, 22 August 2023 07.47
>
> > From: Morten Brørup
> > Sent: Monday, August 21, 2023 2:37 AM
> >
> > > From: Honnappa Nagarahalli [mailto:honnappa.nagaraha...@arm.com]
> > > Sent: Monday, 21 August 2023 08.04
> > >
>
> +/**
> + * Perform security processing of packets and do Rx inject after processing.
> + *
> + * Rx inject would behave similarly to ethdev loopback but with the
> additional
> + * security processing. In case of ethdev loopback, application would be
> + * submitting packets to ethdev Tx queues
Add recycle mbufs engine for testpmd. This engine forward pkts with
I/O forward mode. But enable mbufs recycle feature to recycle used
txq mbufs for rxq mbuf ring, which can bypass mempool path and save
CPU cycles.
Suggested-by: Jerin Jacob
Signed-off-by: Feifei Wang
Reviewed-by: Ruifeng Wang
-
Define specific function implementation for ixgbe driver.
Currently, recycle buffer mode can support 128bit
vector path. And can be enabled both in fast free and
no fast free mode.
Suggested-by: Honnappa Nagarahalli
Signed-off-by: Feifei Wang
Reviewed-by: Ruifeng Wang
Reviewed-by: Honnappa Naga
Define specific function implementation for i40e driver.
Currently, mbufs recycle mode can support 128bit
vector path and avx2 path. And can be enabled both in
fast free and no fast free mode.
Suggested-by: Honnappa Nagarahalli
Signed-off-by: Feifei Wang
Reviewed-by: Ruifeng Wang
Reviewed-by: H
Add 'rte_eth_recycle_rx_queue_info_get' and 'rte_eth_recycle_mbufs'
APIs to recycle used mbufs from a transmit queue of an Ethernet device,
and move these mbufs into a mbuf ring for a receive queue of an Ethernet
device. This can bypass mempool 'put/get' operations hence saving CPU
cycles.
For eac
Currently, the transmit side frees the buffers into the lcore cache and
the receive side allocates buffers from the lcore cache. The transmit
side typically frees 32 buffers resulting in 32*8=256B of stores to
lcore cache. The receive side allocates 32 buffers and stores them in
the receive side so
> -Original Message-
> From: Qiao, Wenjing
> Sent: Friday, August 11, 2023 6:00 PM
> To: Zhang, Yuying ; Xing, Beilei
>
> Cc: dev@dpdk.org; Liu, Mingxia ; Qiao, Wenjing
>
> Subject: [PATCH v2 2/4] net/cpfl: add flow json parser
>
> A JSON file will be used to direct DPDK CPF PMD to
>
Get rid of all the complicated code which copies data on the stack:
- allocate a new segment from the same mempool than the original mbuf,
- copy headers data in this segment,
- chain the new segment in place of headers of the original mbuf,
- use existing helpers for computing IP and TCP/UDP check
Checking that one of RTE_MBUF_F_TX_IPV4 or RTE_MBUF_F_TX_IP_CKSUM is
present is not compliant with the offloading API which specifies that IP
checksum requires RTE_MBUF_F_TX_IP_CKSUM.
On the other hand, RTE_MBUF_F_TX_IP_CKSUM is invalid for IPv6 packets,
so we can simply check for RTE_MBUF_F_TX_IP_
The L4 checksum offloading API does not require l4_len to be set.
Make the driver discover the L4 headers size by itself.
Fixes: 6546e76056e3 ("net/tap: calculate checksums of multi segs packets")
Cc: sta...@dpdk.org
Signed-off-by: David Marchand
Tested-by: Ales Musil
---
.mailmap
76 matches
Mail list logo
