> -Original Message-
> From: Feifei Wang
> Sent: Wednesday, June 30, 2021 2:41 PM
> To: Xing, Beilei
> Cc: dev@dpdk.org; n...@arm.com; Feifei Wang ;
> Ruifeng Wang
> Subject: [PATCH v3 1/2] net/i40e: improve performance for scalar Tx
>
> For i40e scalar Tx path, if implement FAST_FRE
Add 21.08 release note for ETS-based Tx QoS
Signed-off-by: Ting Xu
---
doc/guides/rel_notes/release_21_08.rst | 7 +++
1 file changed, 7 insertions(+)
diff --git a/doc/guides/rel_notes/release_21_08.rst
b/doc/guides/rel_notes/release_21_08.rst
index a6ecfdf3ce..35aa76a270 100644
--- a/doc/
Add check in the Tx packet preparation function, to guarantee that the
packet with specific user priority is distributed to the correct Tx
queue according to the configured Tx queue TC mapping.
Signed-off-by: Ting Xu
---
drivers/net/iavf/iavf.h | 10 +
drivers/net/iavf/iavf_rxtx.c |
This patch added the support for VF to config the ETS-based Tx QoS,
including querying current QoS configuration from PF and config queue TC
mapping. PF QoS is configured in advance and the queried info is
provided to the user for future usage. VF queues are mapped to different
TCs in PF through vi
This patch supports the ETS-based QoS configuration. It enables the DCF
to configure bandwidth limits for each VF VSI of different TCs. A
hierarchy scheduler tree is built with port, TC and VSI nodes.
Signed-off-by: Qiming Yang
Signed-off-by: Ting Xu
---
drivers/net/ice/ice_dcf.c| 9 +
When link status changes, DCF will receive virtchnl PF event message.
Add support to handle this event, change link status and update link
info.
Signed-off-by: Ting Xu
---
drivers/net/ice/ice_dcf.h| 6
drivers/net/ice/ice_dcf_ethdev.c | 54 ++--
drivers/
In the adminq command query port ETS function, the root node teid is
needed. However, for DCF, the root node is not initialized, which will
cause error when we refer to the variable. In this patch, we will check
whether the root node is available or not first.
Signed-off-by: Ting Xu
---
drivers/
This patch adds new virtchnl opcodes and structures for QoS
configuration, which includes:
1. VIRTCHNL_VF_OFFLOAD_TC, to negotiate the capability supporting QoS
configuration. If VF and PF both have this flag, then the ETS-based QoS
offload function is supported.
2. VIRTCHNL_OP_DCF_CONFIG_BW, DCF i
This patch enables the ETS-based Tx QoS for IAVF. Kernel tool is used to
configure ETS first. DCF is used to set bandwidth limit for VFs of each
TC. IAVF is supported to query QoS capability and set queue TC mapping.
Traffic Management API is utilized to configure the QoS hierarchy
scheduler tree.
On Mon, May 10, 2021 at 11:42 AM Feifei Wang wrote:
>
> For "order_launch_lcores" function, wmb after that the main lcore
> updates the variable "t->err", which represents the end of the test
> signal, is unnecessary. Because after the main lcore updates this
> siginal variable, it will jump out o
> -Original Message-
> From: dev On Behalf Of Xueming(Steven) Li
> Sent: Tuesday, June 15, 2021 6:56 PM
> To: Christian Ehrhardt
> Cc: dev ; Luca Boccassi ; NBU-Contact-Thomas
> Monjalon ; dpdk stable
> ; Yuanhan Liu
> Subject: Re: [dpdk-dev] [PATCH 2/2] devtools: auto detect branch t
For i40e vector Tx path, if tx_offload is set as FAST_FREE_MBUF mode,
no mbuf fast free operations are executed. To fix this, add mbuf fast
free mode for vector Tx path.
Furthermore, for i40e vector Tx path, if implement FAST_FREE_MBUF mode,
it means per-queue all mbufs come from the same mempool
For i40e scalar Tx path, if implement FAST_FREE_MBUF mode, it means
per-queue all mbufs come from the same mempool and have refcnt = 1.
Thus we can use bulk free of the buffers when mbuf fast free mode is
enabled.
Following are the test results with this patch:
MRR L3FWD Test:
two ports & bi-dir
For i40e Tx path, use bulk free of the buffers when mbuf fast free
mode is enabled. This can efficiently improve the performance.
v2:
1. fix bug when tx_rs_thres > RTE_I40E_TX_MAX_FREE_BUF_SZ (Beilei)
v3:
1. change variable name for more readable (Beilei)
Feifei Wang (2):
net/i40e: improve per
Please mark [PATCH v2], and record changes under the ---
Ray K
> -Original Message-
> From: Mohsin Kazmi
> Sent: Tuesday 29 June 2021 18:07
> To: dev@dpdk.org
> Cc: Kinsella, Ray ; Mohsin Kazmi
> ; sta...@dpdk.org; Olivier Matz
> ; Ananyev, Konstantin
> ; Thomas Monjalon
> ; Tomasz Kulas
Current fix scan scripts scanned specified range in HEAD branch.
When users ran it in an earlier branch, few patches were scanned
due to the fixes in the range are newer and not merged to HEAD
branch.
This patch introduces optional argument, default to HEAD
if not specified. Checks the specified
When scanning fixes from current(HEAD) branch, local tags were included
and reported as version. For example:
$ git tag --contains --merged
20.11_backport_202010506 // user tag
v20.11
v20.11.1
This patch matches DPDK officail version pattern in search, selects
the most early tag. Offici
Hello Ferruh,
On Tue, Jun 29, 2021 at 3:46 PM Ferruh Yigit wrote:
>
> Add 'RTE_ETH' namespace to all enums & macros in a backward compatible
> way. The macros for backward compatibility can be removed in next LTS.
>
> Signed-off-by: Ferruh Yigit
- I did not do a full check but I noticed that ET
From: Michael Baum
> The issue can cause illegal physical address access while a huge-page A is
> released and huge-page B is allocated on the same virtual address.
> The old MR can be matched using the virtual address of huge-page B but the
> HW will access the physical address of huge-page A w
Hi,
It looks good to me. I verified on arm platforms and didn't find any problem
currently.
BR,
Richael
> -Original Message-
> From: David Hunt
> Sent: Wednesday, June 23, 2021 8:04 PM
> To: dev@dpdk.org
> Cc: david.h...@intel.com; anatoly.bura...@intel.com; Richael Zhuang
> ; step...@n
> -邮件原件-
> 发件人: Xing, Beilei
> 发送时间: 2021年6月30日 11:43
> 收件人: Feifei Wang
> 抄送: dev@dpdk.org; nd ; Ruifeng Wang
>
> 主题: RE: [PATCH v2 1/2] net/i40e: improve performance for scalar Tx
>
>
>
> > -Original Message-
> > From: Feifei Wang
> > Sent: Wednesday, June 30, 2021 10:04 A
The following changes since commit 51266ce05ab1caa442c54754063d147be61485ec:
net/cxgbe: add MAC matchall to track promisc traffic (2021-06-29 12:37:48
+0300)
are available in the Git repository at:
http://dpdk.org/git/next/dpdk-next-net-mrvl
for you to fetch changes up to 55e702407d76fe6a0
> -Original Message-
> From: Maxime Coquelin
> Sent: Wednesday, June 30, 2021 12:12 AM
> To: dev@dpdk.org; Xia, Chenbo ;
> david.march...@redhat.com
> Cc: Maxime Coquelin
> Subject: [PATCH v7 7/7] vhost: convert inflight data to DPDK allocation API
>
> Inflight metadata are allocated usi
> -Original Message-
> From: Maxime Coquelin
> Sent: Wednesday, June 30, 2021 12:12 AM
> To: dev@dpdk.org; Xia, Chenbo ;
> david.march...@redhat.com
> Cc: Maxime Coquelin ; sta...@dpdk.org
> Subject: [PATCH v7 4/7] vhost: fix NUMA reallocation with multiqueue
>
> Since the Vhost-user devi
> -Original Message-
> From: Maxime Coquelin
> Sent: Wednesday, June 30, 2021 12:11 AM
> To: dev@dpdk.org; Xia, Chenbo ;
> david.march...@redhat.com
> Cc: Maxime Coquelin
> Subject: [PATCH v7 3/7] vhost: fix missing cache logging NUMA realloc
>
> When the guest allocates virtqueues on a
> -Original Message-
> From: Wang, Haiyue
> Sent: Wednesday, June 30, 2021 12:41 AM
> To: dev@dpdk.org
> Cc: david.march...@redhat.com; Wang, Haiyue ;
> Yang, Qiming ; Zhang, Qi Z
> Subject: [PATCH v2] net/ice: enable to set HW debug mask
>
> The HW debug mask is always zero, so user
> -Original Message-
> From: Feifei Wang
> Sent: Wednesday, June 30, 2021 10:04 AM
> To: Xing, Beilei
> Cc: dev@dpdk.org; n...@arm.com; Feifei Wang ;
> Ruifeng Wang
> Subject: [PATCH v2 1/2] net/i40e: improve performance for scalar Tx
>
> For i40e scalar Tx path, if implement FAST_FR
> -Original Message-
> From: Thomas Monjalon
> Sent: Tuesday, June 29, 2021 6:40 PM
> To: Honnappa Nagarahalli
> Cc: dev@dpdk.org; David Marchand ;
> Ruifeng Wang ; Richael Zhuang
> ; nd
> Subject: Re: [dpdk-dev] Arm roadmap for 21.08
>
> 03/06/2021 06:41, Honnappa Nagarahalli:
> >
Hi, all
one question about 'rxm->hash.rss' and 'mb->hash.fdir'.
In Rx recv packets function,
'rxm->hash.rss' will report rss hash result from Rx desc.
'rxm->hash.fdir' will report filter identifier from Rx desc.
But function implementation differs from som
Currently, we store metadata in BE and need to change
it to CPU endian when attaching to mbuf.
For dv_xmeta_en = 2, no bit shift.
For dv_xmeta_en = 1, 16 bits shift.
For the non-vector mode, endian change should be first,
then bit shift.
For the vector mode, the behavior is right. In this update
For i40e vector Tx path, if tx_offload is set as FAST_FREE_MBUF mode,
no mbuf fast free operations are executed. To fix this, add mbuf fast
free mode for vector Tx path.
Furthermore, for i40e vector Tx path, if implement FAST_FREE_MBUF mode,
it means per-queue all mbufs come from the same mempool
For i40e scalar Tx path, if implement FAST_FREE_MBUF mode, it means
per-queue all mbufs come from the same mempool and have refcnt = 1.
Thus we can use bulk free of the buffers when mbuf fast free mode is
enabled.
Following are the test results with this patch:
MRR L3FWD Test:
two ports & bi-dir
For i40e Tx path, use bulk free of the buffers when mbuf fast free
mode is enabled. This can efficiently improve the performance.
v2:
1. fix bug when tx_rs_thres > RTE_I40E_TX_MAX_FREE_BUF_SZ (Beilei)
Feifei Wang (2):
net/i40e: improve performance for scalar Tx
net/i40e: improve performance f
On Fri, Jun 25, 2021 at 11:36:21AM +0300, Dmitry Kozlyuk wrote:
> 2021-06-24 17:27 (UTC-0700), Narcisa Ana Maria Vasile:
> > From: Narcisa Vasile
> >
> > Fix count_cpu() to ensure it iterates through all the CPUs in a set.
> > count_cpu() iterates through the CPUs in the set 's' and counts the
>
在 2021/5/12 9:05, Min Hu (Connor) 写道:
在 2021/5/12 0:14, Ferruh Yigit 写道:
On 5/7/2021 10:08 AM, Min Hu (Connor) wrote:
From: Chengwen Feng
This patch adds link status event pre-conditions.
Signed-off-by: Chengwen Feng
Signed-off-by: Min Hu (Connor)
---
doc/guides/nics/hns3.rst | 10 +
> -Original Message-
> From: Chen, LingliX
> Sent: Thursday, June 24, 2021 5:35 PM
> To: Xing, Beilei ; Zhang, Qi Z
> Cc: dev@dpdk.org; Xing, Beilei ; sta...@dpdk.org
> Subject: RE: [dpdk-dev] [PATCH v2] net/i40e: fix FDIR input set conflict issue
>
>
> > -Original Message-
>
From: Chengwen Feng
This patch adds link status event pre-conditions.
Signed-off-by: Chengwen Feng
Signed-off-by: Min Hu (Connor)
---
v2:
* not modified, only waiting for 5.13 kernel released.
---
doc/guides/nics/hns3.rst | 10 ++
1 file changed, 10 insertions(+)
diff --git a/doc/gui
From: Narcisa Vasile
rte_cpuset_t describes a set of CPUs by using an array of masks
named '_bits'. Each element in the '_bits' array represents
a bit mask, with each bit corresponding to a CPU.
The maximum number of CPUs is given by 'CPU_SETSIZE'.
The number of bit masks is computed using '_NUM_
> -Original Message-
> From: Liu, Lingyu
> Sent: Wednesday, June 30, 2021 3:46 PM
> To: dev@dpdk.org; Zhang, Qi Z ; Xing, Beilei
> ; Wu, Jingjing
> Cc: Guo, Junfeng ; Liu, Lingyu
> Subject: [PATCH v1] doc: update GTPoGRE FDIR/RSS feature in iavf driver
>
> GTPoGRE inner most layer FD
>
> For Arm platforms, reading descs can get re-ordered, then the status of DD
> bits will be discontinuous, so add the logic to only process continuous descs
> by checking DD bits.
>
> Fixes: 4861cde46116 ("i40e: new poll mode driver")
> Cc: sta...@dpdk.org
>
> Signed-off-by: Joyce Kong
> Rev
GTPoGRE inner most layer FDIR and RSS will be supported
in iavf driver.
Signed-off-by: Lingyu Liu
---
doc/guides/rel_notes/release_21_08.rst | 4
1 file changed, 4 insertions(+)
diff --git a/doc/guides/rel_notes/release_21_08.rst
b/doc/guides/rel_notes/release_21_08.rst
index a6ecfdf3ce..
On Fri, Jun 25, 2021 at 3:34 PM Ajit Khaparde
wrote:
>
> This patchset adds runtime queue setup support.
>
> Ajit Khaparde (2):
> net/bnxt: add support for runtime queue setup
> net/bnxt: fix ring alloc and free logic
Patchset applied to dpdk-next-net-brcm, for-next-net branch.
Thanks
>
> dr
Resolve name collisions with Windows types
Signed-off-by: Jie Zhou
Acked-by: Tal Shnaiderman
Acked-by: Dmitry Kozlyuk
---
app/test-pmd/cmdline_flow.c | 512 ++--
1 file changed, 256 insertions(+), 256 deletions(-)
diff --git a/app/test-pmd/cmdline_flow.c b/app/
Function print_fdir_mask and print_fdir_flex_payload are only called
when either i40e or ixgbe presents. Extend existing #if defined to
include these two functions, to remove "unused function" compilation
warning.
Signed-off-by: Jie Zhou
Acked-by: Tal Shnaiderman
Acked-by: Dmitry Kozlyuk
---
a
- Make printf format OS independent
- Replace htons with RTE_BE16
- Replace POSIX specific inet_aton with OS independent inet_pton
- Replace sleep with rte_delay_us_sleep
- Repalce random with rte_rand
- #ifndef mman related code for now
- Fix header inclusion
- Include rte_os_shim.h in tes
- Disable unsupported Apps on Windows
- Enable building of testpmd on Windows
Signed-off-by: Jie Zhou
Acked-by: Tal Shnaiderman
Acked-by: Dmitry Kozlyuk
---
app/meson.build| 4
app/pdump/meson.build | 6 ++
app/proc-info/meson.build | 6 ++
Passing an uint32_t pointer to an enum pointer parameter causes
pointer-sign warning on Windows (converts between pointers to
integer types with different sign), since enum is implicitly
converted to int on Windows.
And the current enum pointer parameter of that function is actually
misleading and
Add clock_gettime on Windows in rte_os_shim.h
Signed-off-by: Jie Zhou
Acked-by: Dmitry Kozlyuk
---
lib/eal/windows/include/rte_os_shim.h | 33 +++
1 file changed, 33 insertions(+)
diff --git a/lib/eal/windows/include/rte_os_shim.h
b/lib/eal/windows/include/rte_os_shim.
Add device event stubs in eal_dev.c for Windows
Signed-off-by: Jie Zhou
Acked-by: Tal Shnaiderman
Acked-by: Dmitry Kozlyuk
---
lib/eal/windows/eal_dev.c | 33 +
lib/eal/windows/meson.build | 1 +
2 files changed, 34 insertions(+)
create mode 100644 lib/eal/w
Add required macros by testpmd on Windows in rte_os_shim.h
Signed-off-by: Jie Zhou
Acked-by: Tal Shnaiderman
Acked-by: Dmitry Kozlyuk
---
lib/eal/windows/include/rte_os_shim.h | 9 +
1 file changed, 9 insertions(+)
diff --git a/lib/eal/windows/include/rte_os_shim.h
b/lib/eal/windows/
This patchset is to enable testpmd on windows. It mainly includes:
- Enable building libraries on Windows that testpmd depends on
- Add necessary macros required by testpmd on Windows in rte_os_shim.h
- Add device event stubs for Windows
- Resolve name collisions with Windows types
- Add clock_gett
Enable building libraries that testpmd depends on for Windows
Signed-off-by: Jie Zhou
Acked-by: Dmitry Kozlyuk
---
lib/meson.build | 7 +++
1 file changed, 7 insertions(+)
diff --git a/lib/meson.build b/lib/meson.build
index 4a64756a68..1673ca4323 100644
--- a/lib/meson.build
+++ b/lib/mes
> Use m only after it was checked not to be NULL.
>
> Fixes: 202d375c60bc1 ("app/test: add cryptodev unit and performance
> tests")
> Cc: sta...@dpdk.org
> Cc: Declan Doherty
>
Applied to dpdk-next-crypto
Thanks.
> Add myself as Crypto API, QAT, SW PMDs based on ipsec-mb,
> NULL PMD, and crypto perf test maintainer. Also remove
> Declan, Deepak, and John from the maintainers of these
> areas.
>
> Signed-off-by: Fan Zhang
> Acked-by: Declan Doherty
> Acked-by: Fiona Trahe
> Acked-by: John Griffin
> ---
Function print_fdir_mask and print_fdir_flex_payload are only called
when either i40e or ixgbe presents. Extend existing #if defined to
include these two functions, to remove "unused function" compilation
warning.
Signed-off-by: Jie Zhou
Acked-by: Tal Shnaiderman
Acked-by: Dmitry Kozlyuk
---
a
From: Jie Zhou
- Disable unsupported Apps on Windows
- Enable building of testpmd on Windows
Signed-off-by: Jie Zhou
Acked-by: Tal Shnaiderman
Acked-by: Dmitry Kozlyuk
---
app/meson.build| 4
app/pdump/meson.build | 6 ++
app/proc-info/meson.build
- Make printf format OS independent
- Replace htons with RTE_BE16
- Replace POSIX specific inet_aton with OS independent inet_pton
- Replace sleep with rte_delay_us_sleep
- Repalce random with rte_rand
- #ifndef mman related code for now
- Fix header inclusion
- Include rte_os_shim.h in tes
Resolve name collisions with Windows types
Signed-off-by: Jie Zhou
Acked-by: Tal Shnaiderman
Acked-by: Dmitry Kozlyuk
---
app/test-pmd/cmdline_flow.c | 512 ++--
1 file changed, 256 insertions(+), 256 deletions(-)
diff --git a/app/test-pmd/cmdline_flow.c b/app/
Passing an uint32_t pointer to an enum pointer parameter causes
pointer-sign warning on Windows (converts between pointers to
integer types with different sign), since enum is implicitly
converted to int on Windows.
And the current enum pointer parameter of that function is actually
misleading and
Add clock_gettime on Windows in rte_os_shim.h
Signed-off-by: Jie Zhou
Acked-by: Dmitry Kozlyuk
---
lib/eal/windows/include/rte_os_shim.h | 33 +++
1 file changed, 33 insertions(+)
diff --git a/lib/eal/windows/include/rte_os_shim.h
b/lib/eal/windows/include/rte_os_shim.
Add device event stubs in eal_dev.c for Windows
Signed-off-by: Jie Zhou
Acked-by: Tal Shnaiderman
Acked-by: Dmitry Kozlyuk
---
lib/eal/windows/eal_dev.c | 33 +
lib/eal/windows/meson.build | 1 +
2 files changed, 34 insertions(+)
create mode 100644 lib/eal/w
This patchset is to enable testpmd on windows. It mainly includes:
- Enable building libraries on Windows that testpmd depends on
- Add necessary macros required by testpmd on Windows in rte_os_shim.h
- Add device event stubs for Windows
- Resolve name collisions with Windows types
- Add clock_gett
Add required macros by testpmd on Windows in rte_os_shim.h
Signed-off-by: Jie Zhou
Acked-by: Tal Shnaiderman
Acked-by: Dmitry Kozlyuk
---
lib/eal/windows/include/rte_os_shim.h | 9 +
1 file changed, 9 insertions(+)
diff --git a/lib/eal/windows/include/rte_os_shim.h
b/lib/eal/windows/
Enable building libraries that testpmd depends on for Windows
Signed-off-by: Jie Zhou
Acked-by: Dmitry Kozlyuk
---
lib/meson.build | 7 +++
1 file changed, 7 insertions(+)
diff --git a/lib/meson.build b/lib/meson.build
index 4a64756a68..1673ca4323 100644
--- a/lib/meson.build
+++ b/lib/mes
> This patch enable out-of-place auth-cipher operations where
> digest should be encrypted among with the rest of raw data.
> It also adds support for partially encrypted digest when using
> auth-cipher operations.
>
> Fixes: 7c87e2d7b359 ("crypto/snow3g: use IPsec library")
> Cc: pablo.de.lara.gu
> Add in missing rte_mbuf size in mempool allocation for out-of-place op.
>
> Fixes: bf9d6702eca9 ("app/crypto-perf: use single mempool")
> Cc: pablo.de.lara.gua...@intel.com
>
> Signed-off-by: Kai Ji
> Acked-by: Fan Zhang
> Acked-by: Pablo de Lara
> ---
> v2:
> -- commit message update
> In Arm platform, when "RTE_ARCH_ARM64_MEMCPY" is set as true, gcc will
> report variable uninitialized warning:
>
> ../drivers/crypto/qat/qat_sym_session.c: In function
> ‘partial_hash_compute’:
> ../lib/eal/include/generic/rte_byteorder.h:241:24: warning:
> ‘’ may be used uninitialized in this
>
> When used with crypto adapter, metabuf pool would be shared across
> cores. Account for the same in pool size.
>
> Signed-off-by: Anoob Joseph
> ---
Series
Acked-by: Akhil Goyal
Applied to dpdk-next-crypto
>
> Add rte_security (lookaside protocol - IPsec) support in crypto_cn10k.
>
> IPsec operations can be offloaded to CPT's SE and IE engines, which
> can process IPsec protcol operations including atomic sequence number
> increment (for outbound operations) and anti replay window check (for
> inbo
On 29/06/2021 17:28, Tyler Retzlaff wrote:
> On Tue, Jun 29, 2021 at 05:00:31PM +0100, Ray Kinsella wrote:
>> Clarifying the ABI policy on the promotion of experimental APIS to stable.
>> We have a fair number of APIs that have been experimental for more than
>> 2 years. This policy ammendment i
On Mon, Jun 28, 2021 at 07:29:11AM -0700, Tyler Retzlaff wrote:
> On Mon, Jun 28, 2021 at 01:55:02PM +0300, Andrew Rybchenko wrote:
> > On 6/24/21 1:34 AM, Jie Zhou wrote:
> > >Replace parse_fec_mode misleading return type name mode with fec_capa
> > >
> > >Fixes: b19da32e3151 ("app/testpmd: add FE
On Mon, Jun 28, 2021 at 07:10:01AM -0700, Tyler Retzlaff wrote:
> On Mon, Jun 28, 2021 at 01:35:09PM +0300, Andrew Rybchenko wrote:
> > On 6/28/21 1:01 PM, Andrew Rybchenko wrote:
> > >On 6/24/21 1:34 AM, Jie Zhou wrote:
> > >>Enable building libraries that testpmd depends on for Windows
> > >>
> >
On Mon, May 24, 2021 at 2:30 PM Thierry Herbelot
wrote:
>
> mz is known to be NULL: do not use it to print a memzone name.
>
> Fixes: 242e18c056890 ("net/octeontx_ep: add Rx queue setup and release")
> Cc: sta...@dpdk.org
> Cc: Nalla Pradeep
> Cc: Radha Mohan Chintakuntla
> Cc: Veerasenareddy Bu
On Tue, 29 Jun 2021 12:55:05 +0200
Juraj Linkeš wrote:
> diff --git a/buildtools/get-numa-count.py b/buildtools/get-numa-count.py
> new file mode 100644
> index 00..3b67564fd4
> --- /dev/null
> +++ b/buildtools/get-numa-count.py
> @@ -0,0 +1,24 @@
> +#!/usr/bin/env python3
> +# SPDX-Licen
Preparation the headers for the hardware offload
misses the outer ipv4 checksum offload.
It results in bad checksum computed by hardware NIC.
This patch fixes the issue by setting the outer ipv4
checksum field to 0.
Fixes: 4fb7e803eb1a ("ethdev: add Tx preparation")
Cc: sta...@dpdk.org
Signed-of
>
> 18/06/2021 01:26, Honnappa Nagarahalli:
> > > On Tue, Jun 15, 2021 at 09:54:51PM -0500, Joyce Kong wrote:
> > > > Since C11 memory model is adopted in DPDK now[1], use GCC's atomic
> > > > builtins in test cases.
> > >
> > > as previously discussed these atomics are not "C11" they are direct
The HW debug mask is always zero, so user can't enable the related debug
function like ICE_DBG_XXX etc, add the devarg 'hw_debug_mask' to set the
debug mask log output at runtime.
Signed-off-by: Haiyue Wang
---
v2: Add the doc in ice.rst
---
doc/guides/nics/ice.rst | 8
drivers/ne
On Thu, Jun 24, 2021 at 02:14:16PM +0200, David Marchand wrote:
> On Thu, Jun 24, 2021 at 12:31 PM Kinsella, Ray wrote:
> >
> > Hi Anatoly & Thomas,
> >
> > The following eal experimental symbols are present in both v21.05 and
> > v19.11 release. These symbols should be considered for promotion t
On Tue, Jun 29, 2021 at 05:00:31PM +0100, Ray Kinsella wrote:
> Clarifying the ABI policy on the promotion of experimental APIS to stable.
> We have a fair number of APIs that have been experimental for more than
> 2 years. This policy ammendment indicates that these APIs should be
> promoted or re
On Tue, Jun 29, 2021 at 1:14 PM wrote:
>
> From: Pavan Nikhilesh
>
> Add multi-segment Rx vector routine, form the primary mbufs using
> vector path switch to scalar path when extracting segments.
>
> Signed-off-by: Pavan Nikhilesh
> Series-acked-by: Nithin Dabilpuram
Series applied to dpdk-
This patch saves the NUMA node the virtqueue is allocated
on at init time, in order to allocate all other data on the
same node.
While most of the data are allocated before numa_realloc()
is called and so the data will be reallocated properly, some
data like the log cache are most likely allocated
Inflight metadata are allocated using glibc's calloc.
This patch converts them to rte_zmalloc_socket to take
care of the NUMA affinity.
Signed-off-by: Maxime Coquelin
---
lib/vhost/vhost.c | 4 +--
lib/vhost/vhost_user.c | 71 ++
2 files changed, 60
This patch improves the numa_realloc() function by making use
of rte_realloc_socket(), which takes care of the memory copy
and freeing of the old data.
Suggested-by: David Marchand
Signed-off-by: Maxime Coquelin
Reviewed-by: Chenbo Xia
---
lib/vhost/vhost_user.c | 186 ++---
Since the Vhost-user device initialization has been reworked,
enabling the application to start using the device as soon as
the first queue pair is ready, NUMA reallocation no more
happened on queue pairs other than the first one since
numa_realloc() was returning early if the device was running.
When the guest allocates virtqueues on a different NUMA node
than the one the Vhost metadata are allocated, both the Vhost
device struct and the virtqueues struct are reallocated.
However, reallocating the log cache on the new NUMA node was
not done. This patch fixes this by reallocating it if it
When the guest allocates virtqueues on a different NUMA node
than the one the Vhost metadata are allocated, both the Vhost
device struct and the virtqueues struct are reallocated.
However, reallocating the guest pages table was missing, which
likely causes at least one cross-NUMA accesses for ever
When the guest allocates virtqueues on a different NUMA node
than the one the Vhost metadata are allocated, both the Vhost
device struct and the virtqueues struct are reallocated.
However, reallocating the Vhost memory table was missing, which
likely causes at least one cross-NUMA accesses for eve
This patch series first fixes missing reallocations of some
Virtqueue and device metadata.
Then, it improves the numa_realloc function by using
rte_realloc_socket API that takes care of the memcpy &
freeing. The VQs NUMA IDs are also saved in the VQ metadata
and used for every allocations so that
Clarifying the ABI policy on the promotion of experimental APIS to stable.
We have a fair number of APIs that have been experimental for more than
2 years. This policy ammendment indicates that these APIs should be
promoted or removed, or should at least form a conservation between the
maintainer a
Currently, l3fwd-power enforces the limitation of having one queue per
lcore. This is no longer necessary, so remove the limitation.
Signed-off-by: Anatoly Burakov
---
examples/l3fwd-power/main.c | 6 --
1 file changed, 6 deletions(-)
diff --git a/examples/l3fwd-power/main.c b/examples/l3fw
Currently, there is a hard limitation on the PMD power management
support that only allows it to support a single queue per lcore. This is
not ideal as most DPDK use cases will poll multiple queues per core.
The PMD power management mechanism relies on ethdev Rx callbacks, so it
is very difficult
Use the new multi-monitor intrinsic to allow monitoring multiple ethdev
Rx queues while entering the energy efficient power state. The multi
version will be used unconditionally if supported, and the UMWAIT one
will only be used when multi-monitor is not supported by the hardware.
Signed-off-by: A
Currently, we expect that only one callback can be active at any given
moment, for a particular queue configuration, which is relatively easy
to implement in a thread-safe way. However, we're about to add support
for multiple queues per lcore, which will greatly increase the
possibility of various
Previously, the semantics of power monitor were such that we were
checking current value against the expected value, and if they matched,
then the sleep was aborted. This is somewhat inflexible, because it only
allowed us to check for a specific value in a specific way.
This commit replaces the co
Use RTM and WAITPKG instructions to perform a wait-for-writes similar to
what UMWAIT does, but without the limitation of having to listen for
just one event. This works because the optimized power state used by the
TPAUSE instruction will cause a wake up on RTM transaction abort, so if
we add the a
This patchset introduces several changes related to PMD power management:
- Changed monitoring intrinsics to use callbacks as a comparison function, based
on previous patchset [1] but incorporating feedback [2] - this hopefully will
make it possible to add support for .get_monitor_addr in virt
Implement support for .get_monitor_addr in AF_XDP driver.
Signed-off-by: Anatoly Burakov
---
Notes:
v2:
- Rewrite using the callback mechanism
drivers/net/af_xdp/rte_eth_af_xdp.c | 34 +
1 file changed, 34 insertions(+)
diff --git a/drivers/net/af_xdp/rte_e
> -Original Message-
> From: Ji, Kai
> Sent: Tuesday, June 29, 2021 4:19 PM
> To: dev@dpdk.org
> Cc: Zhang, Roy Fan ; Ji, Kai ; De
> Lara Guarch, Pablo
> Subject: [dpdk-dev v2] crypto/aesni_gcm: fix performance issue
>
> This patch fixes the aesni_gcm performance issue on systems with AV
This patch fixes the aesni_gcm performance issue on systems with AVX512
CPU flag presented but with VAES CPU flag missing, such as Skylake.
Fixes: 81fe96a0cece ("crypto/aesni_gcm: use architecture independent API")
Cc: pablo.de.lara.gua...@intel.com
Signed-off-by: Kai Ji
---
v2: Checkpatch fix
10/06/2021 16:32, Shy Shyman:
> Below is NVIDIA Mellanox's roadmap for DPDK21.08, on which we are currently
> working:
[...]
Added in online roadmap:
https://git.dpdk.org/tools/dpdk-web/commit/?id=08ac0a75618a909
1 - 100 of 182 matches
Mail list logo
