On 1/18/23 18:44, Rongwei Liu wrote:
When a DPDK application must be upgraded,
the traffic downtime should be shortened as much as possible.
During the migration time, the old application may stay alive
while the new application is starting and being configured.
In order to optimize the switch t
On 2/1/23 01:55, Thomas Monjalon wrote:
31/01/2023 19:14, Jerin Jacob:
On Wed, Jan 18, 2023 at 9:15 PM Rongwei Liu wrote:
When a DPDK application must be upgraded,
the traffic downtime should be shortened as much as possible.
During the migration time, the old application may stay alive
while
The LaunchTime defines the scheduling time of the packet from
the packet buffer to the MAC. The launchtime of each packet is
specified as an offset applied to the BaseT registers while BaseT
is automatically incremented each cycle.
This patch supports Tx timestamp based packet pacing by leveraging
Add definitions for packet pacing(launch time offloading) related
registers.
Signed-off-by: Simei Su
---
drivers/net/igc/base/igc_defines.h | 9 +
drivers/net/igc/base/igc_regs.h| 8
2 files changed, 17 insertions(+)
diff --git a/drivers/net/igc/base/igc_defines.h
b/driver
[PATCH v2 1/2] expose packet pacing registers
[PATCH v2 2/2] enable launch time offloading
v2:
* Refine title and commit log.
* Add release notes.
* Rename variable name.
Simei Su (2):
net/igc/base: expose packet pacing registers
net/igc: enable launch time offloading
doc/guides/rel_notes/r
Definers are a limited resource in the system per GVMI, to
avoid failure we try to improve bt checking if it is possible
to reuse the definers in some cases. Added a cache on the context
for this purpose.
Signed-off-by: Alex Vesker
---
drivers/net/mlx5/hws/mlx5dr_context.c | 12 ++-
drivers/net
To have the same name convention for future caches, use
cache and cache item naming.
Signed-off-by: Alex Vesker
---
drivers/net/mlx5/hws/mlx5dr_pat_arg.c | 22 +++---
drivers/net/mlx5/hws/mlx5dr_pat_arg.h | 6 +++---
2 files changed, 14 insertions(+), 14 deletions(-)
diff --git
Add support for dumping range and hash definers objects.
Hash definer is a per matcher object describing the fields
used for hashing. Range definer is per match template object
describing the fields used for range matching.
Both are optional based on the given match templates.
Signed-off-by: Alex
FW WQE and HW WQE are done in a similar way but not to
jeopardize the performance rule creation is done over
the new FW rule creation function. The deletion function
is shared between both flows.
Signed-off-by: Alex Vesker
---
drivers/net/mlx5/hws/mlx5dr_rule.c | 180 +++-
Calculate and create an additional definer used for range
check during matcher creation. In such case two definers will
be created one for specific matching and a range definer.
Since range HW GTA WQE doesn't support the needed range
support rule insertion rule insertion is done using the FW
GTA WQ
Hash definers allow performing hashing over a subset of
the fields which are used for matching. This allows combining
match templates which were considered invalid until now.
During matcher creation mlx5dr code will process the match
templates and check if such hash definer is needed based
on the d
Support range matching over selected items and range
is not supported over all the items. The range match
is done using:
item->last.field - maximum value
item->mask.field - bitmask
item->spec.field - minimum value
When items are processed if item last and mask fields are
non zero range matching wi
Until now definer creation and deletion get and put
functions were used, the get function would calculate the
definer field copy (fc), header layout (hl) and definer
layout internally without taking into account other match
templates used over the same matcher.
This logic had to be split to allow s
This check can be later on reused for other places, it
will look better in a function
Signed-off-by: Alex Vesker
---
drivers/net/mlx5/hws/mlx5dr_matcher.c | 43 +--
1 file changed, 27 insertions(+), 16 deletions(-)
diff --git a/drivers/net/mlx5/hws/mlx5dr_matcher.c
b/dr
FW WQE supports complex rules, constructed from 2 STEs,
for example:
Hash(DefinerA)
SteMatch(DefinerB)
SteRange(DefinerC)
DefinerA is a subset of DefinerB
This complex rule is written using a single FW command which
has a single WQE control, STE match data0 and STE
Send STE WQE function wraps the send WQE command to support WQE
build and FDB abstraction. Sending using FW is different from
sending from HW since FW returns the completion immediately which
requires us to retry on failure and prepare the completion as
part of the send process.
Signed-off-by: Ale
Rename rtc params create for new format.
Signed-off-by: Alex Vesker
---
drivers/common/mlx5/mlx5_prm.h| 16 ++--
drivers/net/mlx5/hws/mlx5dr_cmd.c | 13 +++--
drivers/net/mlx5/hws/mlx5dr_cmd.h | 11 +++
drivers/net/mlx5/hws/mlx5dr_matcher.c | 19 ++
Read the capabilities required to determine support for GENERATE_WQE.
Signed-off-by: Alex Vesker
---
drivers/common/mlx5/mlx5_prm.h| 6 --
drivers/net/mlx5/hws/mlx5dr_cmd.c | 12
drivers/net/mlx5/hws/mlx5dr_cmd.h | 3 +++
3 files changed, 19 insertions(+), 2 deletions(-)
The generate WQE command is used as an interface to writing GTA
WQEs with fields that are not supported in current HW, for example
extended match definer.
Signed-off-by: Alex Vesker
---
drivers/common/mlx5/mlx5_prm.h| 27 +-
drivers/net/mlx5/hws/mlx5dr_cmd.c | 47
Until now we supported asynchronous drain, triggering the queue
to start the drain, now we added support for synchronous which
assures all the work was processed on the queue.
This is useful when working over a FW command and HW queue in parallel
sending arguments over the HW queue and match over
ConnectX and BlueField devices which support HWS are capable of performing
advanced match, range-match and hashing operations on packet headers. This
patch-set introduces support for (1) range matching - allows combining
range and exact match over provided item fields using spec, last and mask.
A u
The action and match templates were stored on the matcher in
a fixed size array to reduce cache misses and reuse template
calculations. This approuch introduced two issues:
-limitation of fixed array
-definer is bindind to match template and cannot be used with
union definer since the layout is fi
On 2/1/23 09:10, Ivan Malov wrote:
Hello everyone,
Since making automatic, or implicit, offload decisions does
not belong in testpmd responsibility domain, it should be
safer to avoid calling the "negotiate metadata delivery"
API with some default selection unless the user asks to
do so explicit
Hello everyone,
Since making automatic, or implicit, offload decisions does
not belong in testpmd responsibility domain, it should be
safer to avoid calling the "negotiate metadata delivery"
API with some default selection unless the user asks to
do so explicitly, via internal CLI or app options.
Trying to allocate memory on the first detected numa node has less
chance to find some memory actually available rather than on the main
lcore numa node (especially when the DPDK application is started only
on one numa node).
Fixes: 705356f0811f ("eal: simplify control thread creation")
Fixes: bb0
> -Original Message-
> From: Thomas Monjalon
> Sent: 2023年2月1日 0:06
> To: You, KaisenX
> Cc: dev@dpdk.org; sta...@dpdk.org; Zhou, YidingX
> ; david.march...@redhat.com; Matz, Olivier
> ; ferruh.yi...@amd.com; You, KaisenX
> ; zhou...@loongson.cn; Burakov, Anatoly
> ; sta...@dpdk.org
>
From: Mike Baucom
The Rx queue stats are being reported incorrectly due to a workaround
for ring counters sometimes being reported as zero. When a queue is
stopped and started, the ring counter is reset to zero. The workaround
interprets the zero as incorrect, and sets the reported ring count a
From: Venkat Duvvuru
By clearing mbuf->hash.fdir.id in the data path, the driver is
corrupting the RSS hash value populated in the mbuf as they
are defined as a union.
This patch fixes the problem by removing the code that clears
mbuf->hash.fdir.id.
Fixes: 17b6c8386d73 ("net/bnxt: fix mark hand
From: Kalesh AP
The Tx queue stats are being reported incorrectly due to a workaround
for ring counters sometimes being reported as zero. When a queue is
stopped and started, the ring counter is reset to zero. The workaround
interprets the zero as incorrect, and sets the reported ring count as
From: Kalesh AP
This patchset contains bnxt PMD bug fixes. Please apply.
Kalesh AP (1):
net/bnxt: fix Tx queue stats after queue stop and start
Mike Baucom (1):
net/bnxt: fix Rx queue stats after queue stop and start
Venkat Duvvuru (1):
net/bnxt: do not corrupt RSS hash value in the mbuf
On Friday, January 27, 2023 11:37 PM, Ferruh Yigit wrote:
> On 1/18/2023 6:00 AM, Jiawen Wu wrote:
> > In some external applications, developers may fill in wrong
> > packet_type in rte_mbuf for transmission. It will result in Tx ring
> > hang when Tx checksum offload is on. So change it to parse f
On Friday, January 27, 2023 11:36 PM, Ferruh Yigit wrote:
> On 1/18/2023 6:00 AM, Jiawen Wu wrote:
> > When round up buffer size to 1K, to configure the register, hardware
> > will receive packets exceeding the buffer size in LRO mode. It will
> > cause a segment fault in the receive function.
> >
That's all right. Thanks very much for your attention~
> -邮件原件-
> 发件人: Konstantin Ananyev
> 发送时间: Wednesday, February 1, 2023 9:11 AM
> 收件人: Feifei Wang
> 抄送: dev@dpdk.org; nd
> 主题: Re: 回复: [PATCH v3 0/3] Direct re-arming of buffers on receive side
>
> Hi Feifei,
>
> > +ping konstanti
Maybe many PMDs do not support oerrors statistics, which cause this
problem isn't found.
LGTM
Acked-by: Huisong Li
在 2023/1/31 19:56, Ferruh Yigit 写道:
There is an inconsistency at displaying Tx dropped value for per port
forwarding stats and accumulated forwarding stats.
While displaying per
Hi Feifei,
+ping konstantin，
Would you please give some comments for this patch series?
Thanks very much.
Sure, will have a look in next few days.
Apologies for the delay.
> -Original Message-
> From: Thomas Monjalon
> Sent: Tuesday, January 31, 2023 4:42 PM
> To: Honnappa Nagarahalli
> Cc: dev@dpdk.org; bruce.richard...@intel.com; m...@smartsharesystems.com;
> Tyler Retzlaff ; david.march...@redhat.com;
> jer...@marvell.com; konstantin.anan...@huawei.com
31/01/2023 17:17, Jerin Jacob:
> On Fri, Jan 27, 2023 at 8:31 PM Thomas Monjalon wrote:
> >
> > 27/01/2023 11:42, Nithin Kumar Dabilpuram:
> > > From: Thomas Monjalon
> > > > 27/01/2023 06:02, Nithin Kumar Dabilpuram:
> > > > > From: Thomas Monjalon
> > > > > > Ferruh is proposing to have a comm
31/01/2023 19:14, Jerin Jacob:
> On Wed, Jan 18, 2023 at 9:15 PM Rongwei Liu wrote:
> >
> > When a DPDK application must be upgraded,
> > the traffic downtime should be shortened as much as possible.
> > During the migration time, the old application may stay alive
> > while the new application is
31/01/2023 19:26, Tyler Retzlaff:
> On Tue, Jan 31, 2023 at 01:23:34PM +0100, Morten Brørup wrote:
> > > From: David Marchand [mailto:david.march...@redhat.com]
> > > Sent: Tuesday, 31 January 2023 12.15
> > >
> > > On Wed, Jan 18, 2023 at 9:31 AM Morten Brørup
> > > wrote:
> > > >
> > > > +To: T
Honnappa, please could you give your view on the future of atomics in DPDK?
12/01/2023 22:26, Tyler Retzlaff:
> Introduce atomics abstraction that permits optional use of standard C11
> atomics when meson is provided the new enable_stdatomics=true option.
>
> Signed-off-by: Tyler Retzlaff
> ---
On 1/31/2023 6:46 PM, Jerin Jacob wrote:
> On Wed, Feb 1, 2023 at 12:09 AM Ferruh Yigit wrote:
>>
>> On 1/30/2023 4:01 PM, Ankur Dwivedi wrote:
>>
>> <...>
>>
> diff --git a/lib/ethdev/meson.build b/lib/ethdev/meson.build index
> 39250b5da1..f5c0865023 100644
> --- a/lib/ethdev/meson.b
hi folks,
just raising this to the top of mailboxes to call for reviewers. it's
going to be a lot of work by the time it's done so your feedback is
gretly appreciated.
thanks!
On Thu, Jan 12, 2023 at 01:26:24PM -0800, Tyler Retzlaff wrote:
> Introduce an abstraction for compiler specific atomics
top-post
this is a re-submission of v5 of the series due to a mistake i made
in the first submission that caused it not to be correctly submitted
to patchwork.
apologies for the noise.
On Tue, Jan 31, 2023 at 12:30:14PM -0800, Tyler Retzlaff wrote:
> Announce deprecation of rte_ctrl_thread_c
Notify deprecation of rte_ctrl_thread_create API, it will be removed as
it exposes platform-specific thread details.
Signed-off-by: Tyler Retzlaff
Acked-by: Morten Brørup
Reviewed-by: Mattias Rönnblom
---
doc/guides/rel_notes/deprecation.rst | 5 +
1 file changed, 5 insertions(+)
diff --g
Add missing thread index referencing rte_thread.h under the basic
topic.
Signed-off-by: Tyler Retzlaff
---
doc/api/doxy-api-index.md | 1 +
1 file changed, 1 insertion(+)
diff --git a/doc/api/doxy-api-index.md b/doc/api/doxy-api-index.md
index de488c7..e8ed51a 100644
--- a/doc/api/doxy-api-inde
Add rte_control_thread_create API as a replacement for
rte_ctrl_thread_create to allow deprecation of the use of platform
specific types in DPDK public API.
Duplicate the rte_ctrl_thread_create test adapted to use
rte_control_thread create to keep both APIs under test until
rte_ctrl_thread_create
Announce deprecation of rte_ctrl_thread_create API which is the final
remaining stable API exposing pthread_t.
Provide an equivalent replacement API rte_control_thread_create that
uses the new rte_thread_t type.
Provide a unit test for the new rte_control_thread_create API.
Add missing doxygen i
From: Tyler Retzlaff
Add rte_control_thread_create API as a replacement for
rte_ctrl_thread_create to allow deprecation of the use of platform
specific types in DPDK public API.
Duplicate the rte_ctrl_thread_create test adapted to use
rte_control_thread create to keep both APIs under test until
From: Tyler Retzlaff
Add missing thread index referencing rte_thread.h under the basic
topic.
Signed-off-by: Tyler Retzlaff
---
doc/api/doxy-api-index.md | 1 +
1 file changed, 1 insertion(+)
diff --git a/doc/api/doxy-api-index.md b/doc/api/doxy-api-index.md
index de488c7..e8ed51a 100644
---
Announce deprecation of rte_ctrl_thread_create API which is the final
remaining stable API exposing pthread_t.
Provide an equivalent replacement API rte_control_thread_create that
uses the new rte_thread_t type.
Provide a unit test for the new rte_control_thread_create API.
Add missing doxygen i
From: Tyler Retzlaff
Notify deprecation of rte_ctrl_thread_create API, it will be removed as
it exposes platform-specific thread details.
Signed-off-by: Tyler Retzlaff
Acked-by: Morten Brørup
Reviewed-by: Mattias Rönnblom
---
doc/guides/rel_notes/deprecation.rst | 5 +
1 file changed, 5
On Wed, Feb 1, 2023 at 12:09 AM Ferruh Yigit wrote:
>
> On 1/30/2023 4:01 PM, Ankur Dwivedi wrote:
>
> <...>
>
> >>> diff --git a/lib/ethdev/meson.build b/lib/ethdev/meson.build index
> >>> 39250b5da1..f5c0865023 100644
> >>> --- a/lib/ethdev/meson.build
> >>> +++ b/lib/ethdev/meson.build
> >>> @@
On 1/30/2023 4:01 PM, Ankur Dwivedi wrote:
<...>
>>> diff --git a/lib/ethdev/meson.build b/lib/ethdev/meson.build index
>>> 39250b5da1..f5c0865023 100644
>>> --- a/lib/ethdev/meson.build
>>> +++ b/lib/ethdev/meson.build
>>> @@ -24,6 +24,7 @@ headers = files(
>>> 'rte_ethdev.h',
>>>
On Tue, Jan 31, 2023 at 01:23:34PM +0100, Morten Brørup wrote:
> > From: David Marchand [mailto:david.march...@redhat.com]
> > Sent: Tuesday, 31 January 2023 12.15
> >
> > On Wed, Jan 18, 2023 at 9:31 AM Morten Brørup
> > wrote:
> > >
> > > +To: Thomas & David, you probably have some opinions on
On Wed, Jan 18, 2023 at 9:15 PM Rongwei Liu wrote:
>
> When a DPDK application must be upgraded,
> the traffic downtime should be shortened as much as possible.
> During the migration time, the old application may stay alive
> while the new application is starting and being configured.
>
> In orde
On Tue, Jan 31, 2023 at 11:20 PM Thomas Monjalon wrote:
>
> 31/01/2023 15:45, Ori Kam:
> > From: Jerin Jacob
> > > On Tue, Jan 31, 2023 at 2:31 PM Rongwei Liu wrote:
> > > > From: Jerin Jacob
> > > > > On Tue, Jan 31, 2023 at 8:23 AM Rongwei Liu
> > > > > > From: Jerin Jacob
> > > > > > > On
31/01/2023 15:45, Ori Kam:
> From: Jerin Jacob
> > On Tue, Jan 31, 2023 at 2:31 PM Rongwei Liu wrote:
> > > From: Jerin Jacob
> > > > On Tue, Jan 31, 2023 at 8:23 AM Rongwei Liu
> > > > > From: Jerin Jacob
> > > > > > On Mon, Jan 30, 2023 at 8:17 AM Rongwei Liu
> > > > > > > From: Jerin Jacob
30/01/2023 18:00, Jiawei Wang:
> --- a/devtools/libabigail.abignore
> +++ b/devtools/libabigail.abignore
> @@ -20,6 +20,11 @@
> [suppress_file]
> soname_regexp = ^librte_.*mlx.*glue\.
>
> +; Ignore fields inserted in middle padding of rte_eth_txconf
> +[suppress_type]
> +name =
> -Original Message-
> From: David Marchand
> Sent: Thursday, January 26, 2023 9:30 AM
> To: Van Haaren, Harry
> Cc: dev@dpdk.org; dpdk...@iol.unh.edu; c...@dpdk.org;
> honnappa.nagaraha...@arm.com; mattias.ronnblom
> ; tho...@monjalon.net; Morten Brørup
> ; Tyler Retzlaff ;
> Aaron Conol
The AESNI_MB PMD supports 8 and 16 byte tag lengths for ZUC256 when
intel-ipsec-mb version is 1.3 or newer.
The conditional check to enable these tag lengths had the incorrect
operator, which enabled 8 and 16 byte tag lengths for versions below 1.2,
which is not supported.
Fixes: 7babda4316f9 ("cr
On 1/19/23 19:46, David Marchand wrote:
Now that all locks in this library are annotated, we can enable the
check.
Signed-off-by: David Marchand
---
lib/vhost/meson.build | 2 ++
1 file changed, 2 insertions(+)
diff --git a/lib/vhost/meson.build b/lib/vhost/meson.build
index bc7272053b..
On Tue, Jan 31, 2023 at 10:53:17AM +0100, David Marchand wrote:
> On Mon, Jan 30, 2023 at 9:35 PM Tyler Retzlaff
> wrote:
> > thank you both, i'll rebase the rte_ctrl_thread_create patch series once
> > this is merged to get it moving forward again.
> >
>
> We are missing a reference to rte_threa
On 1/19/23 19:46, David Marchand wrote:
Access to vdpa_device_list must be protected with vdpa_device_list_lock
spinlock.
Signed-off-by: David Marchand
---
Changes since RFC v3:
- rebased,
---
lib/vhost/vdpa.c | 19 ++-
1 file changed, 10 insertions(+), 9 deletions(-)
On 12/15/2022 1:41 AM, Chengwen Feng wrote:
> This patch declare the hns3 PMD supports rule keep capability, and also
> add "--disable-flow-flush" parameter to test for such capability.
>
> Chengwen Feng (2):
> net/hns3: declare support rule keep capability
> app/testpmd: add disable-flow-flus
On 1/19/23 19:46, David Marchand wrote:
The starting point for this is __vhost_iova_to_vva() which requires the
lock to be taken. Annotate all the code leading to a call to it.
vdpa and vhost_crypto code are annotated but they end up not taking
a IOTLB lock and have been marked with a FIXME a
On 1/19/23 19:46, David Marchand wrote:
clang does not support conditionally held locks when statically analysing
taken locks with thread safety checks.
Always take iotlb locks regardless of VIRTIO_F_IOMMU_PLATFORM feature.
Signed-off-by: David Marchand
---
lib/vhost/vhost.c | 8 +++-
On 1/19/23 19:46, David Marchand wrote:
vq->async is initialised and must be accessed under vq->access_lock.
Signed-off-by: David Marchand
---
Changes since RFC v3:
- rebased,
- fixed annotations vq->access_lock -> &vq->access_lock,
- reworked free_vq,
---
lib/vhost/vhost.c | 4
On 1/19/23 19:46, David Marchand wrote:
vhost_user_lock/unlock_all_queue_pairs must be waived since clang
annotations can't express taking a runtime number of locks.
vhost_queue_stats_update() requirement can be expressed with a required
tag.
Signed-off-by: David Marchand
---
Changes since
On 1/19/23 19:46, David Marchand wrote:
Be a bit more strict when a programmatic error is detected wrt to the
with regards to*
access_lock not being taken.
Mark the new helper with __rte_assert_exclusive_lock so that clang
understands where locks are expected to be taken.
Signed-off-by: D
Few minor comments inline. Otherwise,
Reviewed-by: Honnappa Nagarahalli
> -Original Message-
> From: Jie Hai
> Sent: Monday, January 30, 2023 8:29 PM
> To: Honnappa Nagarahalli ;
> konstantin.v.anan...@yandex.ru; dev@dpdk.org
> Cc: liudongdo...@huawei.com; haij...@huawei.com
> Subject:
Few minor nits. Otherwise,
Reviewed-by: Honnappa Nagarahalli
> -Original Message-
> From: Jie Hai
> Sent: Monday, January 30, 2023 8:29 PM
> To: Honnappa Nagarahalli ;
> konstantin.v.anan...@yandex.ru; dev@dpdk.org
> Cc: liudongdo...@huawei.com; haij...@huawei.com
> Subject: [PATCH v3 2
On 1/19/23 19:46, David Marchand wrote:
Dedicate send_vhost_slave_message() helper to the case when no reply is
needed.
Add a send_vhost_slave_message_process_reply() helper for the opposite.
This new helper merges both send_vhost_slave_message() and the code
previously in process_slave_messa
On Tue, 31 Jan 2023 10:40:08 +0100
"Robin Jarry" wrote:
> Hi Stephen,
>
> Stephen Hemminger, Jan 26, 2023 at 18:03:
> > Alternatively, you could should check for lcore out of range.
> >
> > Simplified as:
> > struct lcore_telemetry_info info = { .d = d };
> > char *endptr; // init not r
On 1/19/23 19:46, David Marchand wrote:
clang offers some thread safety checks, statically verifying that locks
are taken and released in the code.
To use those checks, the full code leading to taking or releasing locks
must be annotated with some attributes.
Wrap those attributes into our ow
On Fri, Jan 27, 2023 at 8:31 PM Thomas Monjalon wrote:
>
> 27/01/2023 11:42, Nithin Kumar Dabilpuram:
> > From: Thomas Monjalon
> > > 27/01/2023 06:02, Nithin Kumar Dabilpuram:
> > > > From: Thomas Monjalon
> > > > > Ferruh is proposing to have a command "port config ..."
> > > > > to configure
31/01/2023 16:05, Kaisen You:
> lib/eal/common/eal_common_thread.c | 1 +
> lib/eal/common/malloc_heap.c | 4
> 2 files changed, 5 insertions(+)
Why the title refers to net/iavf?
On 1/17/23 23:36, Nicolas Chautru wrote:
The documentation for the Intel vRAN Boost PMD
is being updated in term of filename and content
to match with new branding and product name.
Signed-off-by: Nicolas Chautru
---
.../bbdevs/features/{acc200.ini => vrb1.ini} | 2 +-
doc/guides/bbdevs
On 1/17/23 23:36, Nicolas Chautru wrote:
Rename remaining variable, comments and enums names
when generic usage is applicable. VRB1 is specific to
current HW support and VRB is more generic.
Signed-off-by: Nicolas Chautru
---
drivers/baseband/acc/acc_common.h | 12 +-
drivers
On 1/17/23 23:36, Nicolas Chautru wrote:
Include vrb_ prefix for all generic functions for VRB PMD.
Signed-off-by: Nicolas Chautru
---
drivers/baseband/acc/rte_vrb_pmd.c | 248 ++---
drivers/baseband/acc/vrb_pmd.h | 2 +-
2 files changed, 124 insertions(+),
On 1/17/23 23:36, Nicolas Chautru wrote:
Adding prefix to register names specific to VRB1 (ACC200)
to avoid future enum collision.
Signed-off-by: Nicolas Chautru
---
drivers/baseband/acc/acc200_pf_enum.h | 174 +-
drivers/baseband/acc/acc200_vf_enum.h | 128 +++
Hi Jerin and Rongwei,
> -Original Message-
> From: Jerin Jacob
> Sent: Tuesday, 31 January 2023 16:37
>
> On Tue, Jan 31, 2023 at 2:31 PM Rongwei Liu wrote:
> >
> > Hi Jerin:
> >
> > BR
> > Rongwei
> >
> > > -Original Message-
> > > From: Jerin Jacob
> > > Sent: Tuesday, Januar
On 1/17/23 23:36, Nicolas Chautru wrote:
Make the device structure more generic so that to
have personnalization of the VRB PMD to support
multiple variants.
Signed-off-by: Nicolas Chautru
---
drivers/baseband/acc/acc200_pf_enum.h | 8 +-
drivers/baseband/acc/acc_common.h | 6 ++
On Tue, Jan 31, 2023 at 2:31 PM Rongwei Liu wrote:
>
> Hi Jerin:
>
> BR
> Rongwei
>
> > -Original Message-
> > From: Jerin Jacob
> > Sent: Tuesday, January 31, 2023 16:46
> > To: Rongwei Liu
> > Cc: dev@dpdk.org; Matan Azrad ; Slava Ovsiienko
> > ; Ori Kam ; NBU-Contact-
> > Thomas Monja
Hi Jiawei,
> -Original Message-
> From: Jiawei(Jonny) Wang
> Sent: Monday, 30 January 2023 19:01
>
> For the multiple hardware ports connect to a single DPDK port (mhpsdp),
> currently, there is no information to indicate the packet belongs to
> which hardware port.
>
> This patch intro
On 1/17/23 23:36, Nicolas Chautru wrote:
Renaming files as part of refactor to have generic
intel vRAN Boost PMD (vrb) whose acc200 is only one variant.
Signed-off-by: Nicolas Chautru
---
drivers/baseband/acc/meson.build | 2 +-
drivers/baseband/acc/rte_acc100_pmd.
Hi Rongwei,
> -Original Message-
> From: Rongwei Liu
> Sent: Wednesday, 18 January 2023 17:45
>
> When a DPDK application must be upgraded,
> the traffic downtime should be shortened as much as possible.
> During the migration time, the old application may stay alive
> while the new appl
> -Original Message-
> From: Jerin Jacob
> Sent: 27 January 2023 15:58
> To: Thomas Monjalon
> Cc: Shivah Shankar Shankar Narayan Rao ;
> Srikanth Yalavarthi ; dev@dpdk.org; Jerin Jacob
> Kollanukkaran ; Anup Prabhu
> ; ferruh.yi...@amd.com;
> bruce.richard...@intel.com; david.march...@re
From: Huisong Li
Currently, the interface for verifying duplicate RSS rules has
some problems:
1) If the value of 'func' in configuring RSS rule is default
value, this rule is mistakenly considered as a duplicate rule.
2) If key length is zero or 'key' is NULL in configuring RSS rule
this r
From: Huisong Li
The RSS configuration will be miscleared when driver receives a RSS rule
which has more one RSS action.
Fixes: c37ca66f2b27 ("net/hns3: support RSS")
Cc: sta...@dpdk.org
Signed-off-by: Huisong Li
Signed-off-by: Dongdong Liu
---
drivers/net/hns3/hns3_flow.c | 4 +---
1 file c
From: Huisong Li
The hw::rss_info::conf::func was set to the macro RTE_ETH_HASH_FUNCTION_MAX
and hw::rss_info::conf::queue was set to NULL when all rules are flushed,
which indicates no flow rules is issued. See
commit eb158fc756a5 ("net/hns3: fix config when creating RSS rule after
flush").
Actu
From: Huisong Li
The types of the rule will no longer be used when user flush all rules or
destroy a rule. But user would receive some RSS types warnings, like,
"modified RSS types based on hardware support, requested:0x137f83fffc
configured:0x3ffc".
Fixes: ec674cb742e5 ("net/hns3: fix flushing
From: Huisong Li
When the type in 'struct rte_flow_action' is RTE_FLOW_ACTION_TYPE_RSS, the
'conf' pointer references the 'struct rte_flow_action_rss' instead of the
'struct hns3_rss_conf' in driver. But driver uses 'struct hns3_rss_conf' to
convert this 'conf' pointer to get RSS action configura
From: Huisong Li
All rules from user are saved in RSS filter list, so use RSS
filter list to check duplicated rule.
Fixes: c37ca66f2b27 ("net/hns3: support RSS")
Cc: sta...@dpdk.org
Signed-off-by: Huisong Li
Signed-off-by: Dongdong Liu
---
drivers/net/hns3/hns3_flow.c | 35 ++
From: Huisong Li
The hns3_rss_set_algo_key() is used to set RSS hash algorithm and key to
hardware. The maximum times of command sent to firmware depend on the
length of key. However, now this times is fixed, which isn't good for
key expansion. In addition, hash algorithm comes from rss_info::has
From: Huisong Li
Extract common function to query device specifications.
Fixes: 9c740336f024 ("net/hns3: get device specifications from firmware")
Cc: sta...@dpdk.org
Signed-off-by: Huisong Li
Signed-off-by: Dongdong Liu
---
drivers/net/hns3/hns3_common.c| 75
From: Huisong Li
For better compatibility, the RSS key size of PF and VF are obtained from
firmware. However, many places still used the old macro HNS3_RSS_KEY_SIZE
as the key size.
Fixes: 9c740336f024 ("net/hns3: get device specifications from firmware")
Cc: sta...@dpdk.org
Signed-off-by: Huis
This patchset is to do some bugfixes for hns3 rss.
v2->v3:
- Fix segmentation fault when key_len is 40 and key is NULL for
[PATCH V2 09/10].
v1->v2:
- Fix missing comparison of types and level when verifying duplicate
rules for [PATCH 10/10].
Huisong Li (10):
net/hns3: fix error log about
From: Huisong Li
The error log about indirection table size during initialization phase
of PF and VF is unreasonable when the indirection table size obtained
from firmware or PF function should be zero. In addition, VF driver
should use error level to print this log.
Fixes: 0fce2c46dc16 ("net/hn
On 1/17/23 17:50, Hernan Vargas wrote:
To make the test compatible with devices that do not support early
termination, the iteration count assert can be removed.
Signed-off-by: Hernan Vargas
---
app/test-bbdev/test_bbdev_perf.c | 6 +-
1 file changed, 1 insertion(+), 5 deletions(-)
d
On 1/17/23 17:50, Hernan Vargas wrote:
Assert also if the opaque date is invalid.
data*
Signed-off-by: Hernan Vargas
---
app/test-bbdev/test_bbdev_perf.c | 8 +++-
1 file changed, 3 insertions(+), 5 deletions(-)
diff --git a/app/test-bbdev/test_bbdev_perf.c b/app/test-bbdev/test_
1 - 100 of 167 matches
Mail list logo
