> From: Wang, Haiyue [mailto:haiyue.w...@intel.com]
> Sent: Wednesday, 13 April 2022 04.47
> To: Daly, Jeff; dev@dpdk.org
> Cc: sta...@dpdk.org; Stephen Douthit; Yang, Qiming
>
> > From: Jeff Daly
> > Sent: Wednesday, April 13, 2022 01:42
> > To: dev@dpdk.org
> > Cc: sta...@dpdk.org; Stephen Dout
> From: Stephen Hemminger [mailto:step...@networkplumber.org]
> Sent: Wednesday, 13 April 2022 04.10
>
> A couple of small micro optimizations on the RIB code.
>
> Stephen Hemminger (2):
> rib: mark error checks with unlikely
> rib6: mark error tests with unlikely
>
> lib/rib/rte_rib.c | 2
Ping.
On 3/22/22 11:28 PM, Nithin Dabilpuram wrote:
Enable Tx IPv4 checksum offload only when Tx inline crypto, lookaside
crypto/protocol or cpu crypto is needed.
For Tx Inline protocol offload, checksum computation
is implicitly taken care by HW.
Signed-off-by: Nithin Dabilpuram
---
example
the kni alloc queue is filled with physical addresses of mbufs
for kernel consumption. Any unused mbufs in the alloc queue are
freed during shutdown sequence in rte_kni_release.
In the current existing implementation, for freeing one entry of
alloc queue all the objects of the mempool are traverse
the kni alloc queue is filled with physical addresses of mbufs for kernel
consumption. Any unused mbufs in the alloc queue are freed during
shutdown sequence in rte_kni_release.
In the current existing implementation, for freeing one entry of alloc queue
all the objects of the mempool are traverse
Hi Kevin,
The dpdk 21.11.1-rc1 test result from Red Hat looks good.
We tested below 17 scenarios and all got PASS on RHEL8:
- Guest with device assignment(PF) throughput testing(1G hugepage size):
PASS
- Guest with device assignment(PF) throughput testing(2M hugepage size)
: PASS
Hi everyone,
I have resubmitted the patch.Thanks.
https://patchwork.dpdk.org/project/dpdk/patch/1649649325-1942-1-git-send-email-chcch...@163.com/
Huichao,Cai
> -Original Message-
> From: Jeff Daly
> Sent: Wednesday, April 13, 2022 01:42
> To: dev@dpdk.org
> Cc: sta...@dpdk.org; Stephen Douthit ; Wang, Haiyue
>
> Subject: [PATCH v6 2/2] net/ixgbe: Fix SFP detection and linking on hotplug
>
> Currently the ixgbe driver does not ID any SFP exce
Also mark some conditional functions as const.
Signed-off-by: Stephen Hemminger
---
lib/rib/rte_rib6.c | 25 -
1 file changed, 12 insertions(+), 13 deletions(-)
diff --git a/lib/rib/rte_rib6.c b/lib/rib/rte_rib6.c
index 042ac1f090bf..650bf1b8f681 100644
--- a/lib/rib/rte
Also mark some conditional functions as const.
Signed-off-by: Stephen Hemminger
---
lib/rib/rte_rib.c | 26 +-
1 file changed, 13 insertions(+), 13 deletions(-)
diff --git a/lib/rib/rte_rib.c b/lib/rib/rte_rib.c
index cd9e823068d2..2a3de5065a31 100644
--- a/lib/rib/rte_r
A couple of small micro optimizations on the RIB code.
Stephen Hemminger (2):
rib: mark error checks with unlikely
rib6: mark error tests with unlikely
lib/rib/rte_rib.c | 26 +-
lib/rib/rte_rib6.c | 25 -
2 files changed, 25 insertions(+), 26
> -Original Message-
> From: Jeff Daly
> Sent: Wednesday, April 13, 2022 01:42
> To: dev@dpdk.org
> Cc: sta...@dpdk.org; Wang, Haiyue ; Xiaolong Ye
> ; Xiao
> Zhang ; Zhao1, Wei ; Lunyuan Cui
>
> Subject: [PATCH v6 1/2] net/ixgbe: Limit SDP3 check of TX_DISABLE to
> appropriate devices
On Wed, Nov 10, 2021 at 05:33:38PM -0800, Narcisa Ana Maria Vasile wrote:
> From: Narcisa Vasile
>
> Use a portable, type-safe representation for the thread identifier.
> Add functions for comparing thread ids and obtaining the thread id
> for the current thread.
>
> Signed-off-by: Narcisa Vasil
Added new macros for pointer operations, bitwise operations,
spinlocks and 32 bit read and write.
Signed-off-by: Srikanth Yalavarthi
---
drivers/common/cnxk/roc_bits.h | 12
drivers/common/cnxk/roc_platform.h | 28 +++-
2 files changed, 31 insertions(+),
Currently the ixgbe driver does not ID any SFP except for the first one
plugged in. This can lead to no-link, or incorrect speed conditions.
For example:
* If link is initially established with a 1G SFP, and later a 1G/10G
multispeed part is later installed, then the MAC link setup functions are
1ca05831b9b added a check that SDP3 (used as a TX_DISABLE output to the
SFP cage on these cards) is not asserted to avoid incorrectly reporting
link up when the SFP's laser is turned off.
ff8162cb957 limited this workaround to fiber ports
This patch:
* Adds devarg 'fiber_sdp3_no_tx_disable' not a
v6:
* Fixed minor coding style issue in patch 1/2
Currently the ixgbe driver does not ID any SFP except for the first one
plugged in. This can lead to no-link, or incorrect speed conditions.
For example:
* If link is initially established with a 1G SFP, and later a 1G/10G
multispeed part is later installed, then the MAC link setup functions are
1ca05831b9b added a check that SDP3 (used as a TX_DISABLE output to the
SFP cage on these cards) is not asserted to avoid incorrectly reporting
link up when the SFP's laser is turned off.
ff8162cb957 limited this workaround to fiber ports
This patch:
* Adds devarg 'fiber_sdp3_no_tx_disable' not a
v5:
* Restore missing patch pieces preventing interrupts on Linux from
working correctly.
Hi, Tyler
On 4/12/2022 3:43 AM, Tyler Retzlaff wrote:
Add function to translate Windows error codes to errno-style error
codes. The possible return values are chosen so that we have as
much semantical compatibility between platforms as possible.
Signed-off-by: Narcisa Vasile
Signed-off-by: Tyl
> A sequence lock (seqlock) is synchronization primitive which allows
> for data-race free, low-overhead, high-frequency reads, especially for
> data structures shared across many cores and which are updated
> relatively infrequently.
>
> A seqlock permits multiple parallel readers. The variant
Hi Jacob,
> -Original Message-
> From: Jerin Jacob
> Sent: Thursday, April 7, 2022 9:27 PM
> To: Wu, WenxuanX
> Cc: Thomas Monjalon ; Andrew Rybchenko
> ; Li, Xiaoyun ;
> Singh, Aman Deep ; Zhang, Yuying
> ; Zhang, Qi Z ; dpdk-dev
> ; Stephen Hemminger ;
> Morten Brørup ; Viacheslav Ovsi
Hi Andrew,
> -Original Message-
> From: Andrew Rybchenko
> Sent: Thursday, April 7, 2022 6:48 PM
> To: Wu, WenxuanX ; tho...@monjalon.net; Li,
> Xiaoyun ; Singh, Aman Deep
> ; Zhang, Yuying ;
> Zhang, Qi Z
> Cc: dev@dpdk.org; step...@networkplumber.org;
> m...@smartsharesystems.com; viac
On Thu, Apr 07, 2022 at 03:14:33PM +0200, David Marchand wrote:
> Hello Tyler,
>
> On Wed, Mar 30, 2022 at 11:00 AM Tyler Retzlaff
> wrote:
> >
> > add missing code to affinitize main_lcore from lcore configuration.
>
> Nit: Add*
>
> >
> > Signed-off-by: Tyler Retzlaff
> > ---
> > lib/eal/win
Was there a consensus on if this should be re-drafted as a bug and fixes or
left as a backportable feature/improvement? I am good either way, just
wanted to clarify if I had an additional action at this time.
Cheers
On Wed, Apr 6, 2022 at 3:33 AM Olivier Matz wrote:
> Hi,
>
> On Tue, Apr 05, 2
On Mon, 11 Apr 2022 03:06:46 +
"Zhang, Ke1X" wrote:
> This is a coding waring as below:
>
> _coding style issues_
>
> WARNING:TYPO_SPELLING: 'slave' may be misspelled - perhaps 'secondary'?
> #64:
> when creating a bonding device, if the slave device's rss key length
>
> WARNING:TYPO_SP
Multi-Packet RQ can further save PCIe bandwidth by posting a single large
buffer for multiple packets.
Instead of posting a buffer per a packet, one large buffer is posted
to receive multiple packets on the buffer.
Add support for multi-packet RQ on Windows.
The feature is disabled by default and
The following series enables support of 3 hardware offloads on Windows which
improve PMD throughput.
RX throughput improvements:
**Multi-packet RQ.
**CQE compression.
TX throughput improvement:
**Multi packet send.
Tal Shnaiderman (3):
net/mlx5: support multi-packet RQ on Windows
net/mlx5:
CQE Compression reduces PCI overhead by coalescing and compressing
multiple CQEs into a single merged CQE.
Add supported for the CQE compression feature on Windows.
feature is enabled by default unless not supported by the HW
or if the rxq_cqe_comp_en PMD argument is explicitly disabled.
Signed-o
Add support for enhanced multi-packet write on Windows.
Enhanced multi-packet write allows the Tx burst function to pack up
multiple packets in a single descriptor session to save PCI bandwidth
and improve performance.
The feature can be controlled by the txq_mpw_en PMD argument:
txq_mpw_en=1 -
On Fri, Apr 08, 2022 at 03:15:03PM +0100, Kevin Laatz wrote:
> DMA devices are created during PCI probe in EAL init. In order to
> perform cleanup for those devices, they need to be stopped and closed.
> This patch adds the necessary cleanup to ensure clean exit.
>
> Signed-off-by: Kevin Laatz
>
For those still using the "igb_uio" kernel module, there may be
occasions where the module is installed in /lib/modules/$VERSION folder
rather than being insmod'ed from the build directory. To support this, a
number of changes are needed to the makefile:
* Change the "clean" command to a generic w
The github actions API is currently undergoing some kind of outage - and
has been for the last 12-14 hours (from what I can see). This
particular issue seems to impact how we can retrieve results from
various runs. I am working towards a workaround that will let us move
past this.
GitHub is awar
Establish unit test for testing thread api. Initial unit tests
for rte_thread_{get,set}_affinity_by_id().
Signed-off-by: Narcisa Vasile
Signed-off-by: Tyler Retzlaff
---
app/test/meson.build| 2 ++
app/test/test_threads.c | 89 +
2 files chan
Implement functions for getting/setting thread affinity.
Threads can be pinned to specific cores by setting their
affinity attribute.
note: rte_convert_cpuset_to_affinity has a limitation that all cpus of
the set belong to the same processor group.
Signed-off-by: Narcisa Vasile
Signed-off-by: Ty
Provide a portable type-safe thread identifier.
Provide rte_thread_self for obtaining current thread identifier.
Signed-off-by: Narcisa Vasile
Signed-off-by: Tyler Retzlaff
---
lib/eal/include/rte_thread.h | 22 ++
lib/eal/unix/rte_thread.c| 11 +++
lib/eal/versi
Add function to translate Windows error codes to errno-style error
codes. The possible return values are chosen so that we have as
much semantical compatibility between platforms as possible.
Signed-off-by: Narcisa Vasile
Signed-off-by: Tyler Retzlaff
---
lib/eal/windows/rte_thread.c | 49 +
this series provides basic dependencies for additional eal thread api
additions. series includes
* basic platform error number conversion.
* function to get current thread identifier.
* functions to get and set affinity with platform agnostic thread
identifier.
* minimal unit test of get and set
From: zhichao zeng
Promote the MDIO APIs to be stable.
Signed-off-by: zhichao zeng
---
drivers/net/ixgbe/rte_pmd_ixgbe.h | 5 -
drivers/net/ixgbe/version.map | 10 +-
2 files changed, 5 insertions(+), 10 deletions(-)
diff --git a/drivers/net/ixgbe/rte_pmd_ixgbe.h
b/drivers/n
For memif non-zero-copy mode, there is a branch to compare
the mbuf and memif buffer size during memory copying. Add
a fast memory copy path by removing this branch with mbuf
and memif buffer size defined at compile time. The removal
of the branch leads to performance uplift.
When mbuf >= memif bu
> -Original Message-
> From: Kevin Traynor
> Sent: Friday, April 1, 2022 6:22 PM
> To: sta...@dpdk.org
> Cc: dev@dpdk.org; Abhishek Marathe ;
> Ali Alnubani ; Walker, Benjamin
> ; David Christensen
> ; hariprasad.govindhara...@intel.com; Hemant
> Agrawal ; Stokes, Ian ;
> Jerin Jacob ; Mcn
> -Original Message-
> From: Richardson, Bruce
> Sent: Tuesday, April 12, 2022 16:39
> To: Zeng, ZhichaoX
> Cc: Wang, Haiyue ; dev@dpdk.org
> Subject: Re: [DPDK] drivers/net/ixgbe: remove expired symbols
>
> On Tue, Apr 12, 2022 at 02:27:06PM +, zhichaox.z...@intel.com wrote:
> > Fro
On Tue, Apr 12, 2022 at 9:23 AM Wang, Haiyue wrote:
>
> > -Original Message-
> > From: Zeng, ZhichaoX
> > Sent: Tuesday, April 12, 2022 22:27
> > To: Wang, Haiyue
> > Cc: dev@dpdk.org; Zeng, ZhichaoX
> > Subject: [DPDK] drivers/net/ixgbe: remove expired symbols
> >
> > From: zhichao zen
On Tue, Apr 12, 2022 at 02:27:06PM +, zhichaox.z...@intel.com wrote:
> From: zhichao zeng
>
> Remove expired experimental symbols.
>
> Signed-off-by: zhichao zeng
> ---
> drivers/net/ixgbe/rte_pmd_ixgbe.h | 20 +---
> drivers/net/ixgbe/version.map | 10 +-
> 2 f
From: zhichao zeng
Promote the MDIO APIs to be stable.
Signed-off-by: zhichao zeng
---
drivers/net/ixgbe/rte_pmd_ixgbe.h | 20 +---
drivers/net/ixgbe/version.map | 10 +-
2 files changed, 10 insertions(+), 20 deletions(-)
diff --git a/drivers/net/ixgbe/rte_pmd_ixgb
Hi,
> -Original Message-
> From: Dmitry Kozlyuk
> Sent: Thursday, March 31, 2022 5:39 PM
> To: dev@dpdk.org
> Cc: sta...@dpdk.org; Slava Ovsiienko ; Matan
> Azrad
> Subject: [PATCH] net/mlx5: fix TxQ failure when inlining is impossible
>
> When txq_inline_max is too large and an mbuf is
Hi,
> -Original Message-
> From: Dmitry Kozlyuk
> Sent: Thursday, March 31, 2022 5:33 PM
> To: dev@dpdk.org
> Cc: sta...@dpdk.org; Matan Azrad ; Slava Ovsiienko
>
> Subject: [PATCH] common/mlx5: fix MR range calculation
>
> MR end for a mempool chunk may be calculated incorrectly.
> For
Hi,
> -Original Message-
> From: Alexander Kozyrev
> Sent: Saturday, March 12, 2022 1:08 AM
> To: dev@dpdk.org
> Cc: Raslan Darawsheh ; Slava Ovsiienko
> ; Matan Azrad
> Subject: [PATCH v3] net/mlx5: handle MPRQ incompatibility with external
> buffers
>
> Multi-Packet Rx queue uses PMD-
> -Original Message-
> From: Zeng, ZhichaoX
> Sent: Tuesday, April 12, 2022 22:27
> To: Wang, Haiyue
> Cc: dev@dpdk.org; Zeng, ZhichaoX
> Subject: [DPDK] drivers/net/ixgbe: remove expired symbols
>
> From: zhichao zeng
>
> Remove expired experimental symbols.
The git commit can be en
50 matches
Mail list logo
