Hi Jie,
> -Original Message-
> From: Jie Wang
> Sent: Saturday, January 29, 2022 8:24 AM
> Subject: [PATCH v3 1/6] ethdev: add L2TPv2 RSS offload type
>
> This patch defines new RSS offload type for L2TPv2, which
> is required when users want to distribute packets based on
> the L2TPv2 s
> -Original Message-
> From: Jie Wang
> Sent: Saturday, January 29, 2022 8:25 AM
> To: dev@dpdk.org
> Cc: stevex.y...@intel.com; Ori Kam ;
> aman.deep.si...@intel.com;
> ferruh.yi...@intel.com; NBU-Contact-Thomas Monjalon (EXTERNAL)
> ;
> andrew.rybche...@oktetlabs.ru; jingjing...@int
> -Original Message-
> From: Jie Wang
> Sent: Saturday, January 29, 2022 8:25 AM
> Subject: [PATCH v3 3/6] app/testpmd: add 6 types of L2TPv2 message
>
> This patch adds L2TPv2 control message and 5 types of data message
> support for testpmd.
>
> The added L2TPv2 message types are li
I didn't want to remove the default parsing of tunnel as VxLan because I
thought it might be used,
Instead I moved it to the end, which makes it detect all supported tunnel
through udp_dst_port,
And only if no tunnel was matched it would default to VxLan.
That was the reason geneve weren't detect
Hi Andrew,
> On 1/20/22 19:26, Akhil Goyal wrote:
> > A new ethernet device op is added to give application control over
>
> ethernet -> Ethernet
Ok
>
> > the IP reassembly configuration. This operation is an optional
> > call from the application, default values are set by PMD and
> > exposed
Hi Andrew,
Thanks for the review.
> On 1/3/22 18:08, Akhil Goyal wrote:
> > IP Reassembly is a costly operation if it is done in software.
> > The operation becomes even more costlier if IP fragmants are encrypted.
> > However, if it is offloaded to HW, it can considerably save application
> > cyc
As discussed in the RFC[1] sent in 21.11, a new offload is
introduced in ethdev for IP reassembly.
This patchset add the IP reassembly RX offload.
Currently, the offload is tested along with inline IPsec processing.
It can also be updated as a standalone offload without IPsec, if there
are some ha
IP Reassembly is a costly operation if it is done in software.
The operation becomes even more costlier if IP fragments are encrypted.
However, if it is offloaded to HW, it can considerably save application
cycles.
Hence, a new offload RTE_ETH_RX_OFFLOAD_IP_REASSEMBLY is introduced in
ethdev for d
A new Ethernet device op is added to give application control over
the IP reassembly configuration. This operation is an optional
call from the application, default/max values are set by PMD and
exposed via rte_eth_dev_info.
Application should always first retrieve the capabilities from
rte_eth_dev
Hardware IP reassembly may be incomplete for multiple reasons like
reassembly timeout reached, duplicate fragments, etc.
To save application cycles to process these packets again, a new
mbuf dynflag is added to show that the mbuf received is not
reassembled properly.
Now if this dynflag is set, ap
A new option is added in IPsec to enable and attempt reassembly
of inbound packets.
Signed-off-by: Akhil Goyal
---
devtools/libabigail.abignore | 14 ++
lib/security/rte_security.h | 12 +++-
2 files changed, 25 insertions(+), 1 deletion(-)
diff --git a/devtools/libabigail.
When compiling with clang using -Wpedantic (or -Wgcc-compat) the use of
diagnose_if kicks up a warning:
.../include/rte_interrupts.h:623:1: error: 'diagnose_if' is a clang
extension [-Werror,-Wgcc-compat]
__rte_internal
^
.../include/rte_compat.h:36:16: note: expanded from macro '__rte_internal'
_
> > extension [-Werror,-Wgcc-compat]
> > __rte_internal
> > ^
>
> Which clang version is this?
>
Clang 10, 11, 12 and 13.
> Perhaps the allow internal API could use a different attribute that
> could work in both cases?
>
I've realised I've made a small error. It is not -Wall that includes this
>
> > +++ b/lib/eal/include/rte_compat.h
> > @@ -33,8 +33,11 @@ section(".text.internal")))
> > #elif !defined ALLOW_INTERNAL_API && __has_attribute(diagnose_if) /*
> > For clang */
>
> Why doesn't the __has_attribute take care of this?
> I would have thought that gcc would check the for the attri
Hi Yunjian,
That's interesting. Is it possible to elaborate the use case or
possibly provide the code snippet?
It is possible that it is a synchronization problem due to relaxed memory model
that Arm architecture uses. There could be a barrier missing in the code.
Thanks,
Honnap
From: Selwin Sebastian
Yellow Carp ethernet devices (V3xxx) use the existing PCI ID but
the window settings for the indirect PCS access have been
altered. Add the check for Yellow Carp Ethernet devices to
use the new register values.
Signed-off-by: Selwin Sebastian
---
drivers/net/axgbe/axgbe_
From: Selwin Sebastian
Adding support for Yellow Carp ethernet device
Selwin Sebastian (2):
net/axgbe: add support for Yellow Carp ethernet device
net/axgbe: disable the CDR wa for Yellow Carp devices
drivers/net/axgbe/axgbe_common.h | 2 ++
drivers/net/axgbe/axgbe_ethdev.c | 36 +
From: Selwin Sebastian
Yellow Carp ethernet devices (V3xxx) do not require
autonegotiation CDR workaround, hence disable the same.
Signed-off-by: Selwin Sebastian
---
drivers/net/axgbe/axgbe_ethdev.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/drivers/net/axgbe/axgbe_ethdev.c b/drive
[Public]
-Original Message-
From: Sebastian, Selwin
Sent: Monday, January 31, 2022 11:09 AM
To: dev@dpdk.org
Cc: Namburu, Chandu-babu ; ferruh.yi...@intel.com
Subject: [PATCH v1 1/2] net/axgbe: add support for Yellow Carp ethernet device
From: Selwin Sebastian
Yellow Carp ethernet d
19 matches
Mail list logo
