Hi DPDK,
I am a pmd driver developer from Realtek NIC department,
when I was porting r8168pmd already verified on x86 to an ARM64 SoC Unisoc:
UIS8650
I found that after NIC Rx init (in general, Rx ring and buffers should have
been prepared for NIC to DMA read),
the NIC status reg showed RDU (Rx
Hello DPDK Community!
This is a friendly reminder to please register to join us September 6-8 for
the 2022 DPDK Userspace event! We are meeting face-to-face in Arcachon,
France, and will also provide a virtual option for those unable to travel.
Please check out the agenda to see the fantastic lin
The pcap file will be synchronized to the disk when stopping the device.
It takes a long time if the file is large that would cause the
'detach sync request' timeout when the device is closed under multi-process
scenario.
This commit fixes the issue by performing synchronization in Tx path
Fixes:
> From: Amit Prakash Shukla [mailto:amitpraka...@marvell.com]
> Sent: Wednesday, 24 August 2022 16.04
>
> GCC 12 raises the following warning:
>
> meson --werror --buildtype=debugoptimized
> --cross-file config/x86/cross-mingw -Dexamples=helloworld build
> ninja -C build
>
> In function 'i
> From: Kevin Laatz [mailto:kevin.la...@intel.com]
> Sent: Wednesday, 24 August 2022 18.25
>
> Currently, there is no way to measure lcore busyness in a passive way,
> without any modifications to the application. This patchset adds a new
> EAL
> API that will be able to passively track core busyn
Add new protocol ID IPV6_NETX_PROTO support for FDIR and RSS,
and always include proto or next header for l3 only RSS case.
The feature depends on the specific ICE DDP packet.
Signed-off-by: Steve Yang
---
drivers/net/ice/base/ice_flow.c | 6 +-
drivers/net/ice/base/ice_flow.h
> Subject: [PATCH] security: remove get_userdata function pointer
>
> Removes get_userdata function pointer as it is being unused
> and make fast accessing method that uses dynamic field to get
> userdata as default for optimizing security path.
>
> Signed-off-by: Srujana Challa
Acked-by: Akhil
> Subject: [PATCH] examples/ipsec-secgw: compilation fix for GCC-12
>
> Typecasting uint32_t array to uint16_t and accessing it as max array
> size(at time of declaration of uint32_t array) causes gcc-12 to
> throw an error.
>
> GCC 12 raises the following warning:
>
> In function 'send_multi_pk
> Subject: [PATCH 1/2] examples/ipsec-secgw: use Tx cksum offload conditionally
>
> Use Tx checksum offload only when all the ports have it enabled as
> the qconf for a particular lcore stores ipv4_offloads for all the
> Tx ports and each lcore can Tx to any port.
>
> Fixes: 03128be4cd4d ("exam
> Subject: [PATCH 2/2] examples/ipsec-secgw: fix issue with IP hdr manipulation
Title should be
examples/ipsec-secgw: fix IP header manipulation
Will fix while applying.
>
> Fix issue with ip header pointer computation to pick the right offset.
>
> Fixes: 6eb3ba03995c ("examples/ipsec-secgw:
> Subject: [PATCH] test/crypto: add AES-CCM test vectors
>
> From: Archana Muniganti
>
> Added ESP tunnel mode known vectors for AES-CCM along with
> combined mode support.
>
> Signed-off-by: Archana Muniganti
Acked-by: Akhil Goyal
Hello Stephen,
We see the same issue with stable 21.11 version of dpdk.
Regards,
Michael
From: Stephen Hemminger
Sent: Wednesday, August 24, 2022 5:28 PM
To: Michael Lekar
Cc: dev
Subject: Re: dpdk-22.07
Please choose a stable version 22.11 soon, or 21.11. bugs in 22.07 will not be
fixed
Hello,
On Fri, Aug 5, 2022 at 1:19 AM Harris, James R wrote:
> Can we keep rte_pci_register(), or a new variation of it that keeps the
> rte_pci_driver structure hidden? Hiding rte_pci_register() would mean
> SPDK can no longer work with a packaged DPDK. Or the DPDK packages
> would need to set
On 8/25/2022 8:20 AM, Yiding Zhou wrote:
The pcap file will be synchronized to the disk when stopping the device.
It takes a long time if the file is large that would cause the
'detach sync request' timeout when the device is closed under multi-process
scenario.
This commit fixes the issue by pe
On 7/21/2022 9:05 AM, David Marchand wrote:
Introduce a new command and remove the last part of specific port init
from testpmd.
Signed-off-by: David Marchand
Since there is no objection, I will proceed with this set, it was
intended for previous release anyway.
In this release we will have
On 25/08/2022 08:47, Morten Brørup wrote:
From: Kevin Laatz [mailto:kevin.la...@intel.com]
Sent: Wednesday, 24 August 2022 18.25
Currently, there is no way to measure lcore busyness in a passive way,
without any modifications to the application. This patchset adds a new
EAL
API that will be able
> -Original Message-
> From: Ferruh Yigit
> Sent: Thursday, August 25, 2022 6:09 PM
> To: Zhou, YidingX ; dev@dpdk.org
> Cc: Zhang, Qi Z ; sta...@dpdk.org
> Subject: Re: [PATCH] net/pcap: reduce time for stopping device
>
> On 8/25/2022 8:20 AM, Yiding Zhou wrote:
> > The pcap file will
On 7/21/2022 9:05 AM, David Marchand wrote:
Since the switch to meson, ixgbe bypass commands were ineffective as the
RTE_LIBRTE_IXGBE_BYPASS build flag was not set, even though the
net/ixgbe driver had this feature compiled in.
Fixes: 16ade738fd0d ("app/testpmd: build with meson")
Cc: sta...@dpd
On 8/25/2022 12:17 PM, Zhou, YidingX wrote:
-Original Message-
From: Ferruh Yigit
Sent: Thursday, August 25, 2022 6:09 PM
To: Zhou, YidingX ; dev@dpdk.org
Cc: Zhang, Qi Z ; sta...@dpdk.org
Subject: Re: [PATCH] net/pcap: reduce time for stopping device
On 8/25/2022 8:20 AM, Yiding Zho
Hi David,
Thanks for your kind reply.
On Wed, Aug 24, 2022 at 22:09, David Marchand wrote:
On Mon, Aug 22, 2022 at 9:42 AM Min Zhou wrote:
This patch fixes the following build failure seen on CentOS 8
with gcc 12.1 because of uninitialized struct variable:
[..]
../lib/vhost/virtio_net.c:
Release status meeting minutes 2022-08-25
=
Agenda:
* Release Dates
* Subtrees
* Roadmaps
* LTS
* Defects
* Opens
Participants:
* ARM
* Canonical [No]
* Debian/Microsoft [No]
* Intel
* Marvell
* Nvidia
* Red Hat
* Xilinx/AMD
Release Dates
-
T
Hi,
> -Original Message-
> From: Morten Brørup
> Sent: Thursday, August 18, 2022 4:17 PM
> To: Jiang, Cheng1 ; Stephen Hemminger
>
> Cc: maxime.coque...@redhat.com; Xia, Chenbo ;
> dev@dpdk.org; Hu, Jiayu ; Ding, Xuan
> ; Ma, WenwuX ; Wang,
> YuanX ; Yang, YvonneX
> Subject: RE: [RFC 1/
Hi Nicolas,
On 7/7/22 01:28, Nicolas Chautru wrote:
Updating the enum for rte_bbdev_op_type
to allow to keep ABI compatible for enum insertion
while adding padded maximum value for array need.
Removing RTE_BBDEV_OP_TYPE_COUNT and instead exposing
RTE_BBDEV_OP_TYPE_PADDED_MAX.
Signed-off-by: Nic
This is only a first pass.
On Wed, Aug 24, 2022 at 10:31 AM Min Zhou wrote:
> diff --git a/MAINTAINERS b/MAINTAINERS
> index 32ffdd1a61..f00b82b5ce 100644
> --- a/MAINTAINERS
> +++ b/MAINTAINERS
> @@ -311,6 +311,12 @@ F: config/riscv/
> F: doc/guides/linux_gsg/cross_build_dpdk_for_riscv.rst
> F
On Thu, Aug 25, 2022 at 3:56 PM David Marchand
wrote:
>
> This is only a first pass.
Additionnally, please run ./devtools/check-meson.py.
I see tabulations where we expect spaces.
--
David Marchand
On 7/6/22 23:16, Chautru, Nicolas wrote:
+};
+
/** Device statistics. */
struct rte_bbdev_stats {
uint64_t enqueued_count; /**< Count of all operations enqueued */
@@ -285,12 +300,14 @@ struct rte_bbdev_driver_info {
/** Set if device supports per-queue interrupts */
Hello,
On Thu, Aug 25, 2022 at 2:37 PM zhoumin wrote:
> >
> > I had seen a similar warning during 22.07 when cross compiling but did
> > not investigate much.
> > The patch that I had written at the time was:
> >
> > diff --git a/lib/vhost/virtio_net.c b/lib/vhost/virtio_net.c
> > index 35fa4670f
> > Added changes to enable CMAN (RED or PIE) at init from profile configuration
> > file.
> >
> > By default CMAN code is enable but not in use, when there is no RED or PIE
> > profile configured.
> >
> > Signed-off-by: Marcin Danilewicz
>
> Acked-by: Jasvinder Singh
Applied, thanks.
On 7/7/22 01:28, Nicolas Chautru wrote:
Added device status information, so that the PMD can
expose information related to the underlying accelerator device status.
Minor order change in structure to fit into padding hole.
Signed-off-by: Nicolas Chautru
---
drivers/baseband/acc100/rte_acc1
Currently, the intel-ipsec-mb library only supports SGL for
GCM and ChaCha20-Poly1305 algorithms through the JOB API.
To add SGL support for other algorithms, a workaround approach is
added in the AESNI_MB PMD. SGL feature flags can now be added to
the PMD.
This patchset also includes a fix for S
The segment size for some tests was too small to hold the auth digest.
This caused issues when using op->sym->auth.digest.data for comparisons
in AESNI_MB PMD after a subsequent patch enables SGL.
For example, if segment size is 2, and digest size is 4, then 4 bytes
are read from op->sym->auth.dig
Currently, for a sessionless op, the session created is reset before
being put back into the mempool. This causes issues as the object isn't
correctly released into the mempool.
Fixes: c68d7aa354f6 ("crypto/aesni_mb: use architecture independent macros")
Fixes: b3bbd9e5f265 ("cryptodev: support de
The intel-ipsec-mb library supports SGL for GCM and ChaChaPoly
algorithms using the JOB API.
This support was added to AESNI_MB PMD previously, but the SGL feature
flags could not be added due to no SGL support for other algorithms.
This patch adds a workaround SGL approach for other algorithms
us
More tests are added to test variations of OOP SGL for snow3g.
This includes LB_IN_SGL_OUT and SGL_IN_LB_OUT.
Signed-off-by: Ciara Power
---
app/test/test_cryptodev.c | 48 +++
1 file changed, 39 insertions(+), 9 deletions(-)
diff --git a/app/test/test_crypto
The current blockcipher test function only has support for two types of
SGL test, INPLACE or OOP_SGL_IN_LB_OUT. These types are hardcoded into
the function, with the number of segments always set to 3.
To ensure all SGL types are tested, blockcipher test vectors now have
fields to specify SGL type
Hello,
I cannot find many details of the SoC on the internet. Does it
use coherent IO? Depending on that, different barriers might be needed. Other
than this, I would not think it needs anything special.
If you could send an RFC to the DPDK mailing list, I am happy to review and
Hi Markus,
On 24/08/2022 20:28, Markus Theil wrote:
If DPDK applications should be used with a minimal set of privileges,
using the msr kernel module on linux should not be necessary.
Since at least kernel 4.4 the rdmsr call to obtain the last non-turbo
boost frequency can be left out, if the s
On 7/7/22 01:28, Nicolas Chautru wrote:
Adding more options in the API to expose the number
of queues exposed and related priority.
Signed-off-by: Nicolas Chautru
---
lib/bbdev/rte_bbdev.h | 4
1 file changed, 4 insertions(+)
diff --git a/lib/bbdev/rte_bbdev.h b/lib/bbdev/rte_bbdev.
Currently, there is no way to measure lcore polling busyness in a passive
way, without any modifications to the application. This patchset adds a new
EAL API that will be able to passively track core polling busyness. As part
of the set, new telemetry endpoints are added to read the generate metric
From: Anatoly Burakov
Currently, there is no way to measure lcore poll busyness in a passive way,
without any modifications to the application. This patch adds a new EAL API
that will be able to passively track core polling busyness.
The poll busyness is calculated by relying on the fact that mo
From: Anatoly Burakov
Expose per-lcore cpuset information to telemetry.
Signed-off-by: Anatoly Burakov
---
lib/eal/common/eal_common_lcore_telemetry.c | 47 +
1 file changed, 47 insertions(+)
diff --git a/lib/eal/common/eal_common_lcore_telemetry.c
b/lib/eal/common/eal_co
Add a new section to the howto guides for using the new lcore poll
busyness telemetry endpoints and describe general usage.
Signed-off-by: Kevin Laatz
---
v3: update naming to poll busyness
---
doc/guides/howto/lcore_busyness.rst | 79 +
1 file changed, 79 insertions
v2:
* Extend and tidy up the macro unit test (Morten).
* Remove unneeded includes from rte_common.h (Morten, Bruce).
Dmitry Kozlyuk (4):
eal: fix pointer arithmetic with an expression argument
eal: deduplicate roundup code
eal: uninline rte_str_to_size
eal: remove unneeded includes
RTE_PTR_SUB(ptr, x) and RTE_PTR_ALIGN_FLOOR() worked incorrectly
if "ptr" was an expression:
uint32_t arr[3];
RTE_PTR_SUB(arr + 1, sizeof(arr[0]));
// expected: (uint32_t *)((uintptr_t)(arr + 1) - 4) == arr
// actual: (uint32_t *)((uintptr_t) arr + 1 - 4) != arr
RTE_PTR_AL
RTE_CACHE_LINE_ROUNDUP() implementation repeated RTE_ALIGN_MUL_CEIL().
In other places RTE_CACHE_LINE_SIZE is assumed to be a power-of-2,
so define RTE_CACHE_LINE_ROUNDUP() using RTE_ALIGN_CEIL().
Signed-off-by: Dmitry Kozlyuk
Reviewed-by: Morten Brørup
Acked-by: Bruce Richardson
Acked-by: Chen
There is no reason for rte_str_to_size() to be inline.
Move the implementation out of .
Export it as a stable ABI because it always has been public.
Signed-off-by: Dmitry Kozlyuk
Acked-by: Morten Brørup
Acked-by: Bruce Richardson
Acked-by: Chengwen Feng
---
lib/eal/common/eal_common_string_fn
Do not include , , and from ,
because they are not used by this file directly.
Include the needed headers directly from the files that need them.
Signed-off-by: Dmitry Kozlyuk
---
Done as a separate commit to keep EAL change clean.
app/test-bbdev/test_bbdev_vector.c | 1 +
app/te
On 7/20/2022 2:28 AM, humin (Q) wrote:
-邮件原件-
发件人: wangyunjian
发送时间: 2022年7月18日 21:09
收件人: dev@dpdk.org
抄送: ch...@att.com; humin (Q) ; Huangshaozhang
; jilei (F) ; wangyunjian
; sta...@dpdk.org
主题: [dpdk-dev] [PATCH] net/bonding: fix array overflow in Rx burst
In bond_ethdev_rx_burst
On 7/20/2022 7:39 AM, humin (Q) wrote:
-邮件原件-
发件人: wangyunjian
发送时间: 2022年7月13日 19:11
收件人: dev@dpdk.org
抄送: ch...@att.com; humin (Q) ; Huangshaozhang
; wangyunjian ; sta...@dpdk.org
主题: [dpdk-dev] [PATCH] net/bonding: fix double get slave link status
When link status polling mode is u
If DPDK applications should be used with a minimal set of privileges,
using the msr kernel module on linux should not be necessary.
Since at least kernel 4.4 the rdmsr call to obtain the last non-turbo
boost frequency can be left out, if the sysfs interface is used.
Also RHEL 7 with recent kernel
Thanks Maxime,
> -Original Message-
> From: Maxime Coquelin
> Sent: Thursday, August 25, 2022 7:19 AM
> To: Chautru, Nicolas ; dev@dpdk.org;
> tho...@monjalon.net; gak...@marvell.com; hemant.agra...@nxp.com
> Cc: t...@redhat.com; m...@ashroe.eu; Richardson, Bruce
> ; david.march...@redhat
v6: added one comment in commit 2/7 suggested by Maxime.
v5: update base on review from Tom Rix. Number of typos reported and resolved,
removed the commit related to rw_lock for now, added a commit for
code clean up from review, resolved one rebase issue between 2 commits, used
size of array for s
Added device status information, so that the PMD can
expose information related to the underlying accelerator device status.
Minor order change in structure to fit into padding hole.
Signed-off-by: Nicolas Chautru
---
drivers/baseband/acc100/rte_acc100_pmd.c | 1 +
drivers/baseband/fp
Updating the enum for rte_bbdev_op_type
to allow to keep ABI compatible for enum insertion
while adding padded maximum value for array need.
Removing RTE_BBDEV_OP_TYPE_COUNT and instead exposing
RTE_BBDEV_OP_TYPE_PADDED_MAX.
Signed-off-by: Nicolas Chautru
Acked-by: Maxime Coquelin
---
app/test-
Adding more options in the API to expose the number
of queues exposed and related priority.
Signed-off-by: Nicolas Chautru
Acked-by: Maxime Coquelin
---
lib/bbdev/rte_bbdev.h | 4
1 file changed, 4 insertions(+)
diff --git a/lib/bbdev/rte_bbdev.h b/lib/bbdev/rte_bbdev.h
index 9b1ffa4..ac9
Extension of bbdev operation to support FFT based operations.
Signed-off-by: Nicolas Chautru
Acked-by: Hemant Agrawal
---
doc/guides/prog_guide/bbdev.rst | 130 +++
lib/bbdev/rte_bbdev.c | 10 ++-
lib/bbdev/rte_bbdev.h | 76 +
This allows to expose more information with regards to any
queue related failure and warning which cannot be supported
in existing API.
Signed-off-by: Nicolas Chautru
---
app/test-bbdev/test_bbdev_perf.c | 2 ++
lib/bbdev/rte_bbdev.c| 19 +++
lib/bbdev/rte_bbdev.h
Add support in existing bbdev PMDs for the explicit number of queue
and priority for each operation type configured on the device.
Signed-off-by: Nicolas Chautru
---
drivers/baseband/acc100/rte_acc100_pmd.c | 29 +-
drivers/baseband/fpga_5gnr_fec/rte_fpga_5gnr_fec.c
Code clean up due to if-check not required
Signed-off-by: Nicolas Chautru
---
lib/bbdev/rte_bbdev_op.h | 15 ---
1 file changed, 4 insertions(+), 11 deletions(-)
diff --git a/lib/bbdev/rte_bbdev_op.h b/lib/bbdev/rte_bbdev_op.h
index afa1a71..386eed8 100644
--- a/lib/bbdev/rte_bbdev_
v3: Fix build (CI).
v2:
* Extend and tidy up the macro unit test (Morten).
* Remove unneeded includes from rte_common.h (Morten, Bruce).
Dmitry Kozlyuk (4):
eal: fix pointer arithmetic with an expression argument
eal: deduplicate roundup code
eal: uninline rte_str_to_size
eal: rem
RTE_PTR_SUB(ptr, x) and RTE_PTR_ALIGN_FLOOR() worked incorrectly
if "ptr" was an expression:
uint32_t arr[3];
RTE_PTR_SUB(arr + 1, sizeof(arr[0]));
// expected: (uint32_t *)((uintptr_t)(arr + 1) - 4) == arr
// actual: (uint32_t *)((uintptr_t) arr + 1 - 4) != arr
RTE_PTR_AL
RTE_CACHE_LINE_ROUNDUP() implementation repeated RTE_ALIGN_MUL_CEIL().
In other places RTE_CACHE_LINE_SIZE is assumed to be a power-of-2,
so define RTE_CACHE_LINE_ROUNDUP() using RTE_ALIGN_CEIL().
Signed-off-by: Dmitry Kozlyuk
Reviewed-by: Morten Brørup
Acked-by: Bruce Richardson
Acked-by: Chen
There is no reason for rte_str_to_size() to be inline.
Move the implementation out of .
Export it as a stable ABI because it always has been public.
Signed-off-by: Dmitry Kozlyuk
Acked-by: Morten Brørup
Acked-by: Bruce Richardson
Acked-by: Chengwen Feng
---
lib/eal/common/eal_common_string_fn
Do not include , , and from ,
because they are not used by this file directly.
Include the needed headers directly from the files that need them.
Signed-off-by: Dmitry Kozlyuk
---
app/test-bbdev/test_bbdev_vector.c | 1 +
app/test-crypto-perf/cperf_test_vector_parsing.c | 1 +
app
The ice has the feature to extract protocol fields into flex descriptor
by programming per queue. However, the dynamic field for proto_ext are
allocated by PMD, it is the responsibility of application to reserved
the field, before start DPDK.
Application with parse the offset and proto_ext name to
Hi Suanming,
The issues you've mentioned will be fixed in v2, thanks!
> -Original Message-
> From: Suanming Mou
> Sent: Thursday, August 11, 2022 10:33 AM
> To: Rong, Leyi ; Wang, Yipeng1
> ; zaoxing...@gmail.com; Gobriel, Sameh
>
> Cc: dev@dpdk.org
> Subject: RE: [PATCH 2/2] test/membe
Tested-by: Jin Ling
-Original Message-
From: Kevin Liu
Sent: 2022年8月26日 18:15
To: dev@dpdk.org
Cc: Yang, Qiming ; Zhang, Qi Z ;
Yang, SteveX ; Liu, KevinX
Subject: [PATCH v4] net/ice: refactor proto_ext to remove global variable
The ice has the feature to extract protocol fields into
> -Original Message-
> From: Ferruh Yigit
> Sent: Tuesday, August 16, 2022 4:59 PM
> To: Rong, Leyi ; Wang, Yipeng1
> ; zaoxing...@gmail.com; Gobriel, Sameh
>
> Cc: dev@dpdk.org
> Subject: Re: [PATCH 1/2] member: implement NitroSketch mode
>
> On 8/10/2022 8:45 AM, Leyi Rong wrote:
> >
Offloading small packets to DMA degrades throughput 10%~20%,
and this is because DMA offloading is not free and DMA is not
good at processing small packets. In addition, control plane
packets are usually small, and assign those packets to DMA will
significantly increase latency, which may cause tim
From: James Hershaw
Prepend `0x` to the NFP HWINFO header value that is printed to improve
the readability of the printed statement.
Signed-off-by: James Hershaw
Reviewed-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/nfpcore/nfp_hwinfo.c | 2 +-
1 file changed, 1 insertio
From: James Hershaw
Previously the nfp driver as supported NFP chips with the Netronome PCIe
ID. This patch extends the PMD to also support NFP chips with the
Corigine PCIe vendor ID (0x1da8), which at this point are assumed to be
otherwise identical from a software perspective.
Signed-off-by: J
This patch series do some upgrade of the log system of nfp PMD:
Use DPDK debug macro to control the nfp Rx/Tx log.
Add the support of nfp cpp log macro.
Long Wu (2):
net/nfp: add support for nfp cpp log print
net/nfp: use dpdk debug macro to control nfp Rx/Tx log print
drivers/net/nfp/nfp_co
From: Long Wu
Nfp cpp log print is not available before. Add support for cpp
print log and set its initial priority to "NOTICE".
Signed-off-by: Long Wu
Reviewed-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/nfp_common.c | 1 +
drivers/net/nfp/nfp_logs.h | 5 -
2 fil
From: Long Wu
Nfp log print was controlled by nfp's own macro before. This
commit changes to use dpdk debug rx/tx macro to control it.
Signed-off-by: Long Wu
Reviewed-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/nfp_logs.h | 4 ++--
1 file changed, 2 insertions(+), 2 del
74 matches
Mail list logo
