From: Sunil Kumar Kori
Patch implements priority flow control support for CNXK platforms.
Signed-off-by: Sunil Kumar Kori
---
v2:
- fix application restart issue.
drivers/net/cnxk/cnxk_ethdev.c | 27 +
drivers/net/cnxk/cnxk_ethdev.h | 16 +++
drivers/net/cnxk/cnxk_ethdev_ops.c
From: Sunil Kumar Kori
CNXK platforms support priority flow control(802.1qbb) to pause
respective traffic per class on that link.
Patch adds RoC interface to configure priority flow control on MAC
block i.e. CGX on cn9k and RPM on cn10k.
Signed-off-by: Sunil Kumar Kori
---
v2:
- fix RoC API n
Hi Ori,
> -Original Message-
> From: Ori Kam
> Sent: Tuesday, January 11, 2022 3:24 PM
> To: Sean Zhang (Networking SW) ; Matan Azrad
> ; NBU-Contact-Thomas Monjalon (EXTERNAL)
> ; Ferruh Yigit ; Andrew
> Rybchenko
> Cc: dev@dpdk.org
> Subject: RE: [RFC 1/3] ethdev: support GRE optional
From: Elena Agostini
GPU CUDA implementation of the new gpudev functions
to expose GPU memory to the CPU.
Today GDRCopy library is required to pin and DMA map
the GPU memory through the BAR1 of the GPU and expose
it to the CPU.
Goal here is to hide technical details GDRCopy library
and expose t
From: Elena Agostini
GPU CUDA implementation of the new gpudev functions
to expose GPU memory to the CPU.
Today GDRCopy library is required to pin and DMA map
the GPU memory through the BAR1 of the GPU and expose
it to the CPU.
Goal here is to hide technical details GDRCopy library
and expose t
Move crypto ops to circular buffer to retain crypto
ops when cryptodev/eventdev are temporarily full
---
v3:
* update eca_ops_buffer_flush() to flush out all the crypto
ops out of circular buffer.
* remove freeing of failed crypto ops from eca_ops_enqueue_burst()
and add to cirular buffer for
update rte_event_crypto_adapter_caps_get() to return
SW_CAP if PMD callback is not registered.
Signed-off-by: Ganapati Kundapura
diff --git a/lib/eventdev/rte_eventdev.c b/lib/eventdev/rte_eventdev.c
index 79b9ea3..6988bf1 100644
--- a/lib/eventdev/rte_eventdev.c
+++ b/lib/eventdev/rte_eventdev.
On 12/19/2021 8:09 AM, Asaf Sinai wrote:
Hi,
How to build 'igb_uio.ko', without the need to build all DPDK code?
'igb_uio' is in its own repo for a while:
https://git.dpdk.org/dpdk-kmods/
And build instructions for it:
https://patches.dpdk.org/project/dpdk/patch/617638b7bdd97cf232289909fcd74
06/01/2022 17:46, Thomas Monjalon:
> 29/11/2021 17:08, Josh Soref:
> > -* external phy devices registred through kerenl apis
> > +* external phy devices registred through kernel apis
>
> another spelling to fix:
> "registred"
> and while at it, "apis" should be "APIs"
>
> -
When enqueueing/dequeueing to/from the ring we try to optimize by manual
loop unrolling. The check for this optimization looks like:
if (likely(idx + n < size)) {
where 'idx' points to the first usable element (empty slot for enqueue,
data for dequeue). The correct comparison here shoul
> From: wangyunjian
[...]
> > From: Dmitry Kozlyuk [mailto:dkozl...@nvidia.com]
[...]
> > Thanks for attaching all the details.
> > Can you please reproduce it with --log-level=pmd.common.mlx5:debug and
> > send the logs?
> >
> > > For example, if the environment is configured with 10GB hugepages
When using Python 3.10, this warning appears:
DeprecationWarning: The distutils package is deprecated
and slated for removal in Python 3.12.
Use setuptools or check PEP 632 for potential alternatives
The PEP 632 recommends replacing "distutils.version" with "packaging".
Bugzilla ID: 914
Cc:
Thank you Morten, Olivier and Konstantin for taking look at it.
I've just sent another version, with updates in commit message suggested
by Olivier.
With regards
Andrzej Ostruszka
On 11/9/2021 9:42 AM, psathe...@marvell.com wrote:
From: Satheesh Paul
This patch adds support to configure channel mask which will
be used by rte flow when adding flow rules on SDP interfaces.
Hi Satheesh,
+ Ori & Andrew.
What 'SDP' stands for?
And can this new devarg be provided with flo
> From: Andrzej Ostruszka [mailto:a...@semihalf.com]
> Sent: Tuesday, 11 January 2022 12.38
>
> When enqueueing/dequeueing to/from the ring we try to optimize by
> manual
> loop unrolling. The check for this optimization looks like:
>
> if (likely(idx + n < size)) {
>
> where 'idx' points
On Mon, Jan 10, 2022 at 2:13 PM Rasesh Mody wrote:
>
> > From: Anatoly Burakov
> > Sent: Tuesday, November 30, 2021 10:29 PM
> >
> > External Email
> >
> > --
> > Expression "a && 1" is equivalent to just "a", so fix the accident
On 11/30/2021 5:45 AM, Ashwin Sekhar T K wrote:
Before issuing the batch alloc, we clear the first word of
cache lines so that NPA can update the status. Make sure that
this line clear is flushed before the batch alloc is issued.
Signed-off-by: Ashwin Sekhar T K
---
drivers/common/cnxk/roc_io
> From: Min Hu (Connor) [mailto:humi...@huawei.com]
> Sent: Tuesday, 11 January 2022 12.55
>
> Added the ethdev dump API which provides functions for query private
> info
> from device. There exists many private properties in different PMD
> drivers,
> such as adapter state, Rx/Tx func algorithm i
On 1/11/2022 12:08 PM, Ferruh Yigit wrote:
On 11/30/2021 5:45 AM, Ashwin Sekhar T K wrote:
Before issuing the batch alloc, we clear the first word of
cache lines so that NPA can update the status. Make sure that
this line clear is flushed before the batch alloc is issued.
Signed-off-by: Ashwin
CAS is compare and swap. CASL is compare and swap with release semantics.
But on CNXK platform, the functionality of CAS* instructions is completely
different when it is done to specific addresses. These APIs are meant for use
for such special cases. These cannot be made ARM generic.
Ashwin Sek
On 11/30/2021 6:41 AM, Rakesh Kudurumalla wrote:
Patch updates configured meter bpf is in rq context
during meter creation
RQ is receive queue, right? Can you please use the long version for
clarification?
Signed-off-by: Rakesh Kudurumalla
---
drivers/net/cnxk/cn10k_rte_flow.c | 9
Please use --cc-cmd devtools/get-maintainer.sh so all maintainers are Cc'ed.
11/01/2022 12:54, Min Hu (Connor):
> --- a/doc/guides/rel_notes/release_22_03.rst
> +++ b/doc/guides/rel_notes/release_22_03.rst
> @@ -55,6 +55,12 @@ New Features
> Also, make sure to start the actual text at the ma
This patch fixes the stack buffer overflow error reported
from AddressSanitizer.
Function send_packetsx4() tries to access out of bound data
from rte_mbuf and fill it into TX buffer even in the case
where no pending packets (len = 0).
Performance impact:- No
ASAN error report:-
==819==ERROR: Addre
On Tue, Jan 11, 2022 at 5:14 PM Thomas Monjalon wrote:
>
> When using Python 3.10, this warning appears:
> DeprecationWarning: The distutils package is deprecated
> and slated for removal in Python 3.12.
> Use setuptools or check PEP 632 for potential alternatives
>
> The PEP 632 recommends
11/01/2022 14:11, Jerin Jacob:
> On Tue, Jan 11, 2022 at 5:14 PM Thomas Monjalon wrote:
> >
> > When using Python 3.10, this warning appears:
> > DeprecationWarning: The distutils package is deprecated
> > and slated for removal in Python 3.12.
> > Use setuptools or check PEP 632 for potenti
Hello
15/02/2021 13:00, Burakov, Anatoly:
> Thanks for your review, Thomas, i'll submit a new version addressing the
> feedback!
It seems this patch was forgotten.
Please could you update?
08/10/2021 04:06, Menon, Ranjit:
> On 10/2/2021 9:18 AM, Dmitry Kozlyuk wrote:
> > Fixes for building with Windows Development Kit 22000.1
> > and some build configuration cleanup.
> >
> > Dmitry Kozlyuk (4):
> >windows: specify signature hash algorithm
> >windows/virt2phys: support only x8
Collection of fixes for idxd driver, including one small enhancement to the
unit tests to help catch future errors too.
V2:
* Changed from single patch for one issue (now patch 1) to multiple
patches to cover other issues discovered.
Bruce Richardson (4):
dma/idxd: fix burst capacity calculat
When the maximum burst size supported by HW is less than the available
ring space, incorrect capacity was returned when there was already some
jobs queued up for submission. This was because the capacity calculation
failed to subtract the number of already-enqueued jobs from the max
burst size. Aft
Recent kernel changes[1][2] mean that we cannot guarantee that the paths
in sysfs used for creating/binding a DSA or workqueue instance will be
as given in the utility script, since they are now "compatibility-mode
only". Update script to support both new paths and compatibility ones.
[1]
https:/
The burst capacity calculation code assumes that the write and read
(i.e. ids_returned) values both wrap at the ring-size, but the read
value instead wraps as UINT16_MAX. Therefore, instead of just adding
ring-size to the write value in case the read is greater, we need to
just always mask the resu
To ensure we catch any bugs in calculation due to wrap-around of the id
values, increase the number of iterations of the burst_capacity test.
Signed-off-by: Bruce Richardson
---
app/test/test_dmadev.c | 7 ---
1 file changed, 4 insertions(+), 3 deletions(-)
diff --git a/app/test/test_dmadev
Tested-by: Sunil Pai G
Thanks and regards,
Sunil
On 1/11/2022 12:26 PM, Ashwin Sekhar Thalakalath Kottilveetil wrote:
CAS is compare and swap. CASL is compare and swap with release semantics.
What does 'release semantics' mean? What is functional difference in both?
But on CNXK platform, the functionality of CAS* instructions is completely
On Tue, Jan 11, 2022 at 01:00:25PM +0100, Morten Brørup wrote:
[...]
> Also this version of the patch:
>
> Reviewed-by: Morten Brørup
My apologies Morten, don't know how I missed your tag.
With regards
Andrzej Ostruszka
> From: Andrzej Ostruszka [mailto:a...@semihalf.com]
> Sent: Tuesday, 11 January 2022 14.46
>
> On Tue, Jan 11, 2022 at 01:00:25PM +0100, Morten Brørup wrote:
> [...]
> > Also this version of the patch:
> >
> > Reviewed-by: Morten Brørup
>
> My apologies Morten, don't know how I missed your tag.
12/10/2021 02:42, Dmitry Kozlyuk:
> v3:
> * Fix Release build (Ranjit).
> * Drop PnpLockdown=1 patch as it is now in dependency series.
> v2:
> * Following ofline review by DmitryM:
> - Add comment explaining tracking approach for validation team.
> - Replace deprecated allo
On 11/01/2022 10:51, Thomas Monjalon wrote:
06/01/2022 17:46, Thomas Monjalon:
29/11/2021 17:08, Josh Soref:
-* external phy devices registred through kerenl apis
+* external phy devices registred through kernel apis
another spelling to fix:
"registred"
and while at it
19/10/2021 21:01, William Tu:
> The patch adds three Intel device IDs,
> I350 (0x1521), 82574L (0x10D3), and 82599 (0x10ED).
>
> Signed-off-by: William Tu
> Acked-by: Dmitry Kozlyuk
> Acked-by: Pallavi Kadam
Applied, thanks.
11/01/2022 15:00, Kevin Traynor:
> On 11/01/2022 10:51, Thomas Monjalon wrote:
> > 06/01/2022 17:46, Thomas Monjalon:
> >> 29/11/2021 17:08, Josh Soref:
> > Applied, thanks.
> >
> > The git log became:
> >
> > fix spelling in comments and strings
> >
> > The tool comes from https:
Hi,
Please find reply inline.
Thanks,
Satheesh.
-Original Message-
From: Ferruh Yigit
Sent: 11 January 2022 05:26 PM
To: Satheesh Paul ; Nithin Kumar Dabilpuram
; Kiran Kumar Kokkilagadda ;
Sunil Kumar Kori ; Satha Koteswara Rao Kottidi
Cc: dev@dpdk.org; Ori Kam ; Andrew Rybchenko
2021-10-12 04:11 (UTC+0300), Dmitry Kozlyuk:
> [...]
This patchset is not supposed to be merged for now.
It should have been marked as RFC, sorry for the misleading title.
Offline discussions revealed a conflict between
edge-triggered interrupts in DPDK contract
and level-triggered IO completions
16/12/2021 13:03, Ferruh Yigit:
> Linux is using '-Wimplicit-fallthrough=5' compiler option, which doesn't
> take any fall through comments into account but only uses compiler
> 'fallthrough' attribute to document fall through action is intended.
>
> "falls through" comment was used in the code wh
https://bugs.dpdk.org/show_bug.cgi?id=914
Thomas Monjalon (tho...@monjalon.net) changed:
What|Removed |Added
Status|CONFIRMED |RESOLVED
Resolutio
On 05/01/2022 10:32, Maxime Gouin wrote:
reported by code analysis tool C++test (version 10.4):
/build/dpdk-20.11/drivers/net/nfp/nfp_net.c
546 Conditions "(hw->ctrl &NFP_NET_CFG_CTRL_ENABLE) &&
!! (hw->cap &NFP_NET_CFG_C" is always evaluated to false
547 Condition "! (hw->cap &N
On 05/01/2022 10:32, Maxime Gouin wrote:
reported by code analysis tool C++test (version 10.4):
/build/dpdk-20.11/drivers/net/nfp/nfpcore/nfp_target.h
375 Condition "island < 1" is always evaluated to false
415 Condition "island < 1" is always evaluated to false
547 Condition "target < 0"
On 05/01/2022 10:26, Maxime Gouin wrote:
reported by code analysis tool C++test (version 10.4):
/build/dpdk-20.11/drivers/bus/ifpga/ifpga_bus.c
67Condition "afu_dev" is always evaluated to true
81Condition "afu_dev" is always evaluated to true
The "for" loop already checks that afu_de
17/12/2021 10:20, Anoob Joseph:
> Print more info when command timeout happens. Print software and
> hardware queue information.
>
> Signed-off-by: Anoob Joseph
> Signed-off-by: Tejasree Kondoj
> ---
> +void
> +cnxk_cpt_dump_on_err(struct cnxk_cpt_qp *qp)
> +{
> + struct pending_queue *pend_
> >> Add Yuying Zhang as a co-maintainer.
> >>
> >> Signed-off-by: Yuying Zhang
> >
> > Acked-by: Xiaoyun Li
> >
>
> Acked-by: Ferruh Yigit
>
> > Thanks for the volunteer.
>
> +1
Applied, thanks for volunteering,
we need more maintainers for testpmd.
I think there is room for one more, wit
> 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
> cycles.
>
> Hence, a new offload RTE_ETH_RX_OFFLOAD_IP_REASSEMBLY is introduced
On 23/12/2021 15:25, Vladimir Medvedkin wrote:
This patch fixes the division by 0, which occurs if the number of routes is
less than 10.
Can be triggered by passing -n argument with value < 10:
9 causing a divide by zero - another example of inflation :-)
./dpdk-test-fib -- -n 9
...
Floatin
> 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 values are set by PMD and
> exposed via rte_eth_dev_info.
> Application should always first retreive the capabilities from
>
17/12/2021 14:15, Ferruh Yigit:
> Remove sample application which is not clear if it is still relevant.
>
> Signed-off-by: Ferruh Yigit
> ---
>
> Please comment if there is an reason to keep the sample application.
This is not an exercise, we are going to remove this example!
Last chance to say
On 21/12/2021 19:57, Robert Sanford wrote:
This patchset makes the following changes to net/bonding:
- Clean up minor errors in spelling, whitespace, C++ wrappers, and
comments.
- Replace directly overwriting of slave port's rte_eth_conf by copying
it, but only updating it via rte_eth_dev_c
14/12/2021 08:56, Gaoxiang Liu:
> Because dynamic fields are registered by the DPDK application,
> so it is up to the application to decide whether to copy the value of
> dynamic fields.
> So delete dynamic fields copy in __rte_pktmbuf_copy_hdr.
> It's more flexible for the DPDK application,
> and
On 11/01/2022 13:41, Bruce Richardson wrote:
To ensure we catch any bugs in calculation due to wrap-around of the id
values, increase the number of iterations of the burst_capacity test.
Signed-off-by: Bruce Richardson
---
app/test/test_dmadev.c | 7 ---
1 file changed, 4 insertions(+),
13/12/2021 17:48, Kevin Traynor:
> Regular LTS releases have previously aligned to DPDK main branch
> releases so that fixes being backported have already gone through
> DPDK main branch release validation.
>
> Now that DPDK main branch has moved to 3 releases per year, the LTS
> releases should c
On 11/01/2022 13:41, Bruce Richardson wrote:
Recent kernel changes[1][2] mean that we cannot guarantee that the paths
in sysfs used for creating/binding a DSA or workqueue instance will be
as given in the utility script, since they are now "compatibility-mode
only". Update script to support both
On 11/01/2022 13:41, Bruce Richardson wrote:
The burst capacity calculation code assumes that the write and read
(i.e. ids_returned) values both wrap at the ring-size, but the read
value instead wraps as UINT16_MAX. Therefore, instead of just adding
ring-size to the write value in case the read i
07/12/2021 12:51, Kevin Traynor:
> On 07/12/2021 11:01, Ferruh Yigit wrote:
> > Not related with this set specifically, a more general question about
> > updating
> > old release notes.
> > For me release notes should be frozen with the release and shouldn't be
> > updated
> > at all afterwards,
> 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 ol_flag (RTE_MBUF_F_RX_IPREASSEMBLY_INCOMPLETE) is added to
> show that the mbuf received is not r
>
> > 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 ol_flag (RTE_MBUF_F_RX_IPREASSEMBLY_INCOMPLETE) is added to
> > show that the mbuf receiv
> > diff --git a/lib/ethdev/rte_ethdev.c b/lib/ethdev/rte_ethdev.c
> > index d9a03f12f9..ecc6c1fe37 100644
> > --- a/lib/ethdev/rte_ethdev.c
> > +++ b/lib/ethdev/rte_ethdev.c
> > @@ -6473,6 +6473,36 @@ rte_eth_rx_metadata_negotiate(uint16_t port_id,
> uint64_t *features)
> >(*de
On Tue, Nov 30, 2021 at 6:42 AM Lance Richardson
wrote:
>
> When the xstats_names parameter to rte_eth_xstats_get_names()
> is non-NULL and the size parameter is less than the required
> number of entries, the driver must return the required size
> without modifying (and over-running) the caller's
On Tue, Jan 4, 2022 at 12:19 AM Kalesh A P
wrote:
>
> From: Kalesh AP
>
> This series contains bnxt PMD bug fixes. Please apply.
Patchset applied to dpdk-next-net-brcm. Thanks
>
> Ajit Khaparde (4):
> net/bnxt: fix ring teardown
> net/bnxt: fix PAM4 mask setting
> net/bnxt: fix pointer acc
> -Original Message-
> From: Ferruh Yigit
> Sent: Tuesday, January 11, 2022 9:46 PM
> To: Ashwin Sekhar Thalakalath Kottilveetil ;
> dev@dpdk.org; Honnappa Nagarahalli ;
> Ruifeng Wang
> Cc: Nithin Kumar Dabilpuram ;
> jer...@marvell.com; Sunil Kumar Kori ; Satha
> Koteswara Rao Kottidi ;
Series-
Acked-by: Hemant Agrawal
On 1/3/2022 3:31 PM, nipun.gu...@nxp.com wrote:
From: Nipun Gupta
This series adds few features and important fixes on DPAA,
PFE and ENETC devices.
Features added:
- level 2 support for shaping on DPAA2
- loopback configuration for DPNI devices on DPAA2
-
> -Original Message-
> From: Ruifeng Wang
> Sent: Wednesday, January 12, 2022 11:01 AM
> To: Ferruh Yigit ; Ashwin Sekhar Thalakalath
> Kottilveetil ; dev@dpdk.org; Honnappa Nagarahalli
>
> Cc: Nithin Kumar Dabilpuram ;
> jer...@marvell.com; Sunil Kumar Kori ; Satha
> Koteswara Rao Kottid
dpdk is fairly expensive to build in GitHub.
It's helpful to abandon old builds as soon as there's a new
build waiting instead of wasting resources on the previous
round.
Josh Soref (1):
ci: restrict concurrency
.github/workflows/build.yml | 3 +++
1 file changed, 3 insertions(+)
--
2.32.0
Signed-off-by: Josh Soref
---
.github/workflows/build.yml | 3 +++
1 file changed, 3 insertions(+)
diff --git a/.github/workflows/build.yml b/.github/workflows/build.yml
index 6cf997d6..a171d430 100644
--- a/.github/workflows/build.yml
+++ b/.github/workflows/build.yml
@@ -12,6 +12,9 @@ defaults
> From: Min Hu (Connor) [mailto:humi...@huawei.com]
> Sent: Wednesday, 12 January 2022 03.40
>
> Added the ethdev dump API which provides functions for query private
> info
> from device. There exists many private properties in different PMD
> drivers,
> such as adapter state, Rx/Tx func algorithm
Users can create the desired number of RxQ and TxQ in DPDK. For
example, if the number of RxQ = 2 and the number of TxQ = 5,
a total of 8 file descriptors will be created for a tap device,
including RxQ, TxQ, and one for keepalive. The RxQ and TxQ
with the same ID are paired by dup(2).
In this sce
This is a followup to
https://github.com/DPDK/dpdk/commit/7be78d027918dbc846e502780faf94d5acdf5f75
The engine that identifies items runs here:
https://github.com/jsoref/dpdk/actions/workflows/spelling.yml
I generally use Google Sheets to select automatic corrections for words.
I'll sometimes
Secondary process support had been disabled for the AF_XDP PMD
because there was no logic in place to share the AF_XDP socket
file descriptors between the processes. This commit introduces
this logic using the IPC APIs.
Since AF_XDP rings are single-producer single-consumer, rx/tx
in the secondary
74 matches
Mail list logo
