Hi Maxime,
> -Original Message-
> From: Maxime Coquelin
> Sent: Tuesday, June 6, 2023 4:18 PM
> To: dev@dpdk.org; Xia, Chenbo ;
> david.march...@redhat.com; m...@redhat.com; f...@redhat.com;
> jasow...@redhat.com; Liang, Cunming ; Xie, Yongji
> ; echau...@redhat.com; epere...@redhat.com;
> -Original Message-
> From: Bill Zhou
> Sent: Tuesday, June 6, 2023 2:07 PM
> To: Ori Kam ; Alex Vesker ; Slava
> Ovsiienko ; NBU-Contact-Thomas Monjalon
> (EXTERNAL) ; Matan Azrad ;
> Suanming Mou
> Cc: dev@dpdk.org; Raslan Darawsheh
> Subject: [PATCH v1 2/2] net/mlx5/hws: add suppor
> -Original Message-
> From: Nipun Gupta
> Sent: Tuesday, June 6, 2023 12:57 PM
> To: Harman Kalra ; dev@dpdk.org;
> tho...@monjalon.net; david.march...@redhat.com;
> anatoly.bura...@intel.com; step...@networkplumber.org
> Cc: ferruh.yi...@amd.com; harpreet.an...@amd.com;
> nikhil.agar..
This change adds support for plugging and unplugging
CDX devices on AMD CDX bus. Also, CDX dev iterator support
has been added for the CDX bus.
Signed-off-by: Nipun Gupta
Acked-by: Ferruh Yigit
---
drivers/bus/cdx/bus_cdx_driver.h | 1 +
drivers/bus/cdx/cdx.c| 122
MSI's are exposed to the devices using VFIO (vfio-cdx). This
patch uses the same to add support for MSI for the devices on
the cdx bus.
Signed-off-by: Nipun Gupta
Acked-by: Ferruh Yigit
---
drivers/bus/cdx/bus_cdx_driver.h | 25 +
drivers/bus/cdx/cdx.c| 11 ++
drivers/bus/cdx/
AMD CDX bus can use VFIO interface for mapping and unmapping
of DMA addresses in the IOMMU. This change adds the callback
support for map and unmap APIs as well as fetching the IOMMU
class.
Signed-off-by: Nipun Gupta
Acked-by: Ferruh Yigit
---
drivers/bus/cdx/cdx.c | 30
AMD CDX bus supports multiple type of devices, which can be
exposed to user-space via vfio-cdx.
vfio-cdx provides the MMIO IO_MEMORY regions as well as the
DMA interface for the device (IOMMU).
This support aims to enable the DPDK to support the cdx
devices in user-space using VFIO interface.
Si
The CDX devices are memory mapped on system bus for embedded CPUs.
It uses sysfs interface and the vfio-cdx driver to discover
and initialize the CDX devices.
The CDX bus and VFIO support is available at Xilinx/AMD open source tree:
https://github.com/Xilinx/linux-xlnx (drivers/cdx/ and drivers/v
> -Original Message-
> From: Deng, KaiwenX
> Sent: Wednesday, June 7, 2023 10:04 AM
> To: dev@dpdk.org
> Cc: sta...@dpdk.org; Yang, Qiming ; Zhou, YidingX
> ; Deng, KaiwenX ; Chas
> Williams ; Min Hu (Connor) ; Wu,
> Jingjing ; Xing, Beilei ; Zhang,
> Qi Z ; Mike Pattrick ; Doherty,
> D
Update graph documentation to introduce new multicore dispatch model.
Signed-off-by: Haiyue Wang
Signed-off-by: Cunming Liang
Signed-off-by: Zhirun Yan
---
doc/guides/prog_guide/graph_lib.rst | 68 +++--
1 file changed, 64 insertions(+), 4 deletions(-)
diff --git a/doc
Add functional test for mcore dispatch model including graph clone,
graph model set/get, node worker affinity, graph worker binding/unbinding.
Signed-off-by: Haiyue Wang
Signed-off-by: Cunming Liang
Signed-off-by: Zhirun Yan
---
app/test/test_graph.c | 130 +
Add new parameter "model" to choose mcore dispatch or rtc model.
And in dispatch model, the node will affinity to worker core successively.
Note:
RTE_GRAPH_MODEL_SELECT is set to GRAPH_MODEL_RTC by default. Must set
model the same as RTE_GRAPH_MODEL_SELECT If set it as rtc or mcore
dispatch explic
Add stats for cross-core dispatching scheduler if stats collection is
enabled.
Signed-off-by: Haiyue Wang
Signed-off-by: Cunming Liang
Signed-off-by: Zhirun Yan
---
lib/graph/graph_debug.c| 6 ++
lib/graph/graph_stats.c| 76 +++---
lib/g
This patch enables to chose new scheduler model. Must define
RTE_GRAPH_MODEL_SELECT before including rte_graph_worker.h
to enable specific model choosing.
Signed-off-by: Haiyue Wang
Signed-off-by: Cunming Liang
Signed-off-by: Zhirun Yan
---
lib/graph/rte_graph_worker.h | 12
1 fil
This patch introduces the task scheduler mechanism to enable dispatching
tasks to another worker cores. Currently, there is only a local work
queue for one graph to walk. We introduce a scheduler worker queue in
each worker core for dispatching tasks. It will perform the walk on
scheduler work queu
This patch enables to create and destroy scheduling workqueue into
common graph operations.
Signed-off-by: Haiyue Wang
Signed-off-by: Cunming Liang
Signed-off-by: Zhirun Yan
---
lib/graph/graph.c | 10 ++
1 file changed, 10 insertions(+)
diff --git a/lib/graph/graph.c b/lib/graph/grap
This patch introduces key functions to allow a worker thread to
enable enqueue and move streams of objects to the next nodes over
different cores for mcore dispatch model.
Signed-off-by: Haiyue Wang
Signed-off-by: Cunming Liang
Signed-off-by: Zhirun Yan
---
lib/graph/graph.c
Add graph_mcore_dispatch_wq_node to hold graph scheduling workqueue
node.
Signed-off-by: Haiyue Wang
Signed-off-by: Cunming Liang
Signed-off-by: Zhirun Yan
---
lib/graph/graph.c | 2 ++
lib/graph/graph_populate.c | 1 +
lib/graph/graph_private.h | 12
This patch adds graph API for supporting to clone the graph object for
a specified worker core. The new graph will also clone all nodes.
Signed-off-by: Haiyue Wang
Signed-off-by: Cunming Liang
Signed-off-by: Zhirun Yan
---
lib/graph/graph.c | 89 +++
Move clone_name() into graph_private.h as a common function for both node
and graph to naming a new cloned object.
Signed-off-by: Haiyue Wang
Signed-off-by: Cunming Liang
Signed-off-by: Zhirun Yan
---
lib/graph/graph_private.h | 41 +++
lib/graph/node.c
Add lcore_id for graph to hold affinity core id where graph would run on.
Add bind/unbind API to set/unset graph affinity attribute. lcore_id will
be set as MAX by default, it means not enable this attribute.
Signed-off-by: Haiyue Wang
Signed-off-by: Cunming Liang
Signed-off-by: Zhirun Yan
---
Add lcore_id for node to hold affinity core id and impl
rte_graph_model_mcore_dispatch_lcore_affinity_set to set node affinity
with specific lcore.
Signed-off-by: Haiyue Wang
Signed-off-by: Cunming Liang
Signed-off-by: Zhirun Yan
---
lib/graph/graph_private.h | 2 +
lib/graph
Node process is a single and reusable block, move the code into an inline
function.
Signed-off-by: Haiyue Wang
Signed-off-by: Cunming Liang
Signed-off-by: Zhirun Yan
---
lib/graph/rte_graph_model_rtc.h | 20 ++---
lib/graph/rte_graph_worker_common.h | 33 +++
Add new get/set APIs to configure graph worker model which is used to
determine which model will be chosen.
Signed-off-by: Haiyue Wang
Signed-off-by: Cunming Liang
Signed-off-by: Zhirun Yan
---
lib/graph/meson.build | 1 +
lib/graph/rte_graph_worker.c| 21 ++
lib
To support multiple graph worker model, split graph into common
and default. Naming the current walk function as rte_graph_model_rtc
cause the default model is RTC(Run-to-completion).
Signed-off-by: Haiyue Wang
Signed-off-by: Cunming Liang
Signed-off-by: Zhirun Yan
---
lib/graph/graph_pcap.c
Rename rte_graph_work.h to rte_graph_work_common.h for supporting
multiple graph worker model.
Signed-off-by: Haiyue Wang
Signed-off-by: Cunming Liang
Signed-off-by: Zhirun Yan
---
MAINTAINERS | 3 ++-
lib/graph/graph_pcap.c
V9:
Fix CI build issues for doc building(move TAILQ next pointer out of union) in
patch 09,10.
Fix graph model check in rte_graph_worker_model_set() in patch 04.
Fix typo in doc.
V8:
No performance dorp for original l3fwd-graph and graph_perf_autotest.
Update graph model set/get functions and ad
> Acked-by: Arek Kusztal
>
> Some things will need to be changed before the next release (few of them I
> described below), but I will ack it. Especially that other algorithms have
> similar
> issues.
>
Sure. I also thought any improvements can collectively addressed as it is
another EC.
> >
> On 6/6/2023 2:23 AM, Chaoyong He wrote:
> > A gentle ping on this series.
...
> >>> Long Wu (2):
> >>> net/bonding: add independent LACP sending function
> >>> app/testpmd: add support for bonding port's LACP negotiation
> >>>
> >>> app/test-pmd/config.c | 19
> >
When iavf send query-stats command in eal-intr-thread through
virtual channel, there will be no response received from
iavf_dev_virtchnl_handler for this command during block and wait.
Because iavf_dev_virtchnl_handler is also registered in eal-intr-thread.
When vf device is bonded as BONDING_MODE
From: Long Wu
In order to support inclusive naming, some of the macro in DPDK will
need to be renamed. Do this through deprecation process now for 23.07.
Signed-off-by: Long Wu
Reviewed-by: Chaoyong He
---
Depends-on: patch-127084 ("doc: announce inclusive naming function name")
---
doc/guide
Refactor the logic of parse metadata to include all the supported
metadata type in one function.
If a packet received carried metadata type 'PORT_ID', then it should
be dispatched to the corresponding ring of the representor port,
else it should be put into the array of the caller function.
This
The only different logic between the NFDk xmit function of PF representor
and PF is if the logic about metadata exists. Refactor the NFDk xmit
function and use a parameter to distinguish them.
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/flower/nfp_flower.c | 15
The only different logic between the NFD3 xmit function of PF representor
port and PF is if the logic about metadata exists. So squash the NFD3
xmit function and use a parameter to distinguish them.
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/flower/nfp_flower.c
The dispatch layer aims to choose the corresponding Tx function
within NFD3 version and NFDk version for reprsentor port.
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/flower/nfp_flower.c | 29 +++--
drivers/net/nfp/flower/nfp_flower.h | 3
Add the xmit function of NFDk for ctrl VNIC.
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/flower/nfp_flower_ctrl.c | 138 ++-
drivers/net/nfp/nfdk/nfp_nfdk.h | 2 +
drivers/net/nfp/nfdk/nfp_nfdk_dp.c | 2 +-
3 files changed,
The dispatch layer aims to choose the corresponding Tx function
within NFD3 version and NFDk version for ctrl VNIC.
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/flower/nfp_flower.c | 1 +
drivers/net/nfp/flower/nfp_flower.h | 3 +++
drivers/net/nfp/fl
NFD3 and NFDk have the same format metadata for the RX direction,
while they are different for the TX direction.
Add the insert metadata logic of NFDk.
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/flower/nfp_flower.c | 8 +++-
drivers/net/nfp/nfdk/nfp_nfdk.
Extract the insert metadata logic of NFD3 into a single function to
simplify the logics.
Add the data structure aims to hold all the different function pointers
which related to different NFD version.
Also add a register function to initialize the function pointers.
Signed-off-by: Chaoyong He
R
Add a check function for the NFD version, this will make the logics
which using this version free from validaty check, thus simplify
the driver.
Signed-off-by: Chaoyong He
Reviewed-by: Niklas Söderlund
---
drivers/net/nfp/flower/nfp_flower.c | 3 +++
drivers/net/nfp/nfp_common.c| 37 ++
The symbol '_abi_flower_extra_features' is in IMEM for NFP4000,
but in EMU_CACHE for NFP3800 because which does not have IMEM.
The original logic can't read symbol from EMU_CACHE, so the probe
process will fail when we try to offload flows use NFP3800.
Modify the related data structure and logics
This patch series add the support of rte_flow for card which using
flower firmware with NFDk:
1. Fix the underlying API logic to support this feature.
2. Import a unified check and dispatch layer for different versions of NFD.
3. Refactor the meta data logic to adapt this dispatch layer.
4. Refacto
> -Original Message-
> From: Zeng, ZhichaoX
> Sent: Thursday, April 13, 2023 1:35 PM
> To: dev@dpdk.org
> Cc: Zhang, Qi Z ; Xu, Ke1 ; Zeng,
> ZhichaoX ; Yang, Qiming
> Subject: [PATCH 3/4] net/ice: enable UDP fragmentation offload
>
> This commit enables transmit segmentation offload
> -Original Message-
> From: Xu, Ke1
> Sent: Wednesday, May 24, 2023 1:21 PM
> To: Zeng, ZhichaoX ; dev@dpdk.org
> Cc: Zhang, Qi Z ; Wu, Jingjing
> ; Xing, Beilei ; Xu, Ke1
>
> Subject: RE: [PATCH 4/4] net/iavf: enable UDP fragmentation offload
>
> > From: Zeng, ZhichaoX
> > Sent: Th
> -Original Message-
> From: Qiming Yang
> Sent: Friday, May 26, 2023 4:23 PM
> To: dev@dpdk.org
> Cc: Yang, Qiming
> Subject: [PATCH] net/igc: support device I225-LMVP
>
> Add support for I225-LMVP device.
>
> Signed-off-by: Qiming Yang
Acked-by: Qi Zhang
Applied to dpdk-next-ne
> -Original Message-
> From: Qiming Yang
> Sent: Tuesday, May 30, 2023 5:54 PM
> To: dev@dpdk.org
> Cc: Yang, Qiming
> Subject: [PATCH] net/e1000: support device I219
>
> support device I219 LM22, V22, LM23 and V23.
>
> Signed-off-by: Qiming Yang
Acked-by: Qi Zhang
Applied to dpd
On 6/6/2023 9:34 AM, Konstantin Ananyev wrote:
>
>
>>
>> [...]
Probably I am missing something, but why it is not possible to do something
>>> like that:
rte_eth_recycle_mbufs(rx_port_id=X, rx_queue_id=Y, tx_port_id=N,
tx_queue_id=M, ...);
rte_eth_recycle_mbufs(rx_po
On 6/6/2023 10:12 PM, Mike Pattrick wrote:
> Previously the noisy neighbour vnf simulation would only operate in io
> mode, forwarding packets as is. However, this limited the usefulness of
> noisy neighbour simulation.
>
> This feature has now been expanded to supporting mac, macswap, and
> 5tswa
Replace the use of rte_atomic.h types and functions, instead use GCC
supplied C++11 memory model builtins.
Signed-off-by: Tyler Retzlaff
Acked-by: Morten Brørup
Acked-by: Bruce Richardson
---
drivers/net/ring/rte_eth_ring.c | 20 ++--
1 file changed, 10 insertions(+), 10 deleti
Replace the use of rte_atomic.h types and functions, instead use GCC
supplied C++11 memory model builtins.
Signed-off-by: Tyler Retzlaff
Acked-by: Morten Brørup
---
drivers/net/null/rte_eth_null.c | 28 ++--
1 file changed, 18 insertions(+), 10 deletions(-)
diff --git a
Replace the use of rte_atomic.h types and functions, instead use GCC
supplied C++11 memory model builtins.
Signed-off-by: Tyler Retzlaff
Acked-by: Morten Brørup
Acked-by: Bruce Richardson
Acked-by: Kevin Laatz
---
drivers/dma/idxd/idxd_internal.h | 3 +--
drivers/dma/idxd/idxd_pci.c | 1
Replace the use of rte_atomic.h types and functions, instead use GCC
supplied C++11 memory model builtins.
Signed-off-by: Tyler Retzlaff
Acked-by: Morten Brørup
---
drivers/net/ixgbe/ixgbe_bypass.c | 1 -
drivers/net/ixgbe/ixgbe_ethdev.c | 18 --
drivers/net/ixgbe/ixgbe_ethdev.
Replace the use of rte_atomic.h types and functions, instead use GCC
supplied C++11 memory model builtins.
This series covers the libraries and drivers that are built on Windows.
The code has be converted to use the __atomic builtins but there are
additional during conversion I notice that there
Replace the use of rte_atomic.h types and functions, instead use GCC
supplied C++11 memory model builtins.
Signed-off-by: Tyler Retzlaff
Acked-by: Morten Brørup
---
lib/stack/rte_stack_lf_generic.h | 16 +---
1 file changed, 9 insertions(+), 7 deletions(-)
diff --git a/lib/stack/rt
Replace the use of rte_atomic.h types and functions, instead use GCC
supplied C++11 memory model builtins.
Signed-off-by: Tyler Retzlaff
Acked-by: Morten Brørup
---
drivers/net/ice/ice_dcf.c| 1 -
drivers/net/ice/ice_dcf_ethdev.c | 1 -
drivers/net/ice/ice_ethdev.c | 12 --
On 6/6/2023 10:37 PM, Ferruh Yigit wrote:
> On 6/6/2023 10:12 PM, Mike Pattrick wrote:
>> Previously the noisy neighbour vnf simulation would only operate in io
>> mode, forwarding packets as is. However, this limited the usefulness of
>> noisy neighbour simulation.
>>
>> This feature has now been
On 6/6/2023 10:12 PM, Mike Pattrick wrote:
> Previously the noisy neighbour vnf simulation would only operate in io
> mode, forwarding packets as is. However, this limited the usefulness of
> noisy neighbour simulation.
>
> This feature has now been expanded to supporting mac, macswap, and
> 5tswa
Previously the noisy neighbour vnf simulation would only operate in io
mode, forwarding packets as is. However, this limited the usefulness of
noisy neighbour simulation.
This feature has now been expanded to supporting mac, macswap, and
5tswap modes. To facilitate adding this support, some new he
Hi David,
> -Original Message-
> From: David Marchand
> Sent: Tuesday, June 6, 2023 2:21 AM
> To: Chautru, Nicolas
> Cc: Maxime Coquelin ; Stephen Hemminger
> ; dev@dpdk.org; Rix, Tom
> ; hemant.agra...@nxp.com; Vargas, Hernan
>
> Subject: Re: [PATCH v1 1/1] doc: announce change in bbd
Acked-by: Arek Kusztal
Some things will need to be changed before the next release (few of them I
described below), but I will ack it. Especially that other algorithms have
similar issues.
> -Original Message-
> From: Gowrishankar Muthukrishnan
> Sent: Sunday, June 4, 2023 11:43 AM
>
On 5/8/2023 2:21 AM, zhoumin wrote:
> Kindly ping.
>
Hi Min,
Sorry for the delay.
> Should we solve this problem by modifying PMDs?
>
> Is it suitable to maintain the RSS configurations for NIC by testpmd in
> order to set DCB?
>
>
> On Thur, Apr 27, 2023 at 4:33PM, zhoumin wrote:
>> Kindly
On 6/6/2023 4:19 PM, Olivier Matz wrote:
> Hi Zeng,
>
> On Thu, Apr 13, 2023 at 01:34:00PM +0800, Zhichao Zeng wrote:
>> This commit calculates the correct pseudo header for the
>> UDP fragmentation offload by adding UDP_SEG flag.
>>
>> Signed-off-by: Zhichao Zeng
>
> Sorry for the late feedback
On 6/6/2023 2:23 AM, Chaoyong He wrote:
> A gentle ping on this series.
>
>> -Original Message-
>> From: Chaoyong He
>> Sent: Friday, May 12, 2023 9:51 AM
>> To: dev@dpdk.org
>> Cc: oss-drivers ; Niklas Soderlund
>>
>> Subject: RE: [PATCH v3 0/2] enhance bonding PMD to support the LACP
>>
On Tue, 6 Jun 2023 16:50:53 +0100
Ferruh Yigit wrote:
> On 6/6/2023 4:39 PM, Stephen Hemminger wrote:
> > On Tue, 6 Jun 2023 20:11:26 +0800
> > Dongdong Liu wrote:
> >
> >> Deprecation notice to add "func" field to ``rte_eth_rss_conf``
> >> structure for RSS hash algorithm.
> >>
> >> Signed-o
On 5/27/2023 3:11 AM, Huisong Li wrote:
> This patchset fix some bugs and support attaching and detaching port
> in primary and secondary.
>
Hi Huisong,
As commented on v4, I have some concerns on this set.
The set does multiple ethdev/testpmd change, but the main target of the
patch is not des
On 2/28/2023 2:21 AM, lihuisong (C) wrote:
>
> 在 2023/2/16 0:09, Ferruh Yigit 写道:
>> On 1/12/2023 2:44 AM, lihuisong (C) wrote:
>>> 在 2023/1/11 20:51, Ferruh Yigit 写道:
On 12/6/2022 9:26 AM, Huisong Li wrote:
> The driver assignment was moved back at the end of the device probing
> bec
On Thu, May 18, 2023 at 6:52 PM wrote:
>
> From: Jerin Jacob
>
> Based on techboard meeting[1] action item, defining the process for a
> new library approval in principle.
>
> [1]
> https://mails.dpdk.org/archives/dev/2023-January/260035.html
>
> Signed-off-by: Jerin Jacob
> ---
> RFC..v1:
> - F
On 6/6/2023 4:39 PM, Stephen Hemminger wrote:
> On Tue, 6 Jun 2023 20:11:26 +0800
> Dongdong Liu wrote:
>
>> Deprecation notice to add "func" field to ``rte_eth_rss_conf``
>> structure for RSS hash algorithm.
>>
>> Signed-off-by: Dongdong Liu
>> ---
>
> New fields do not require deprecation not
The DPDK can be compiled to be run in IOVA VA mode with
'enable_iova_as_pa=false' meson option. It disables support
of the physical address field in the mbufs and application
can not be run in 'IOVA as PA mode'.
The patch adds warning if some bus requires the 'IOVA as PA'
mode in runtime and the '
On 5/25/2023 9:39 PM, Stephen Hemminger wrote:
> The various reserved fields added to ethdev could not be
> safely used for future extensions because they were never
> checked on input. Therefore ABI would be broken if these
> fields were added in a future DPDK release.
>
> Fixes: 436b3a6b6e62 ("e
On Tue, Jun 06, 2023 at 06:15:11PM +0300, Viacheslav Ovsiienko wrote:
> The DPDK can be compiled to be run in IOVA VA mode with
> 'enable_iova_as_pa=false' meson option. If there is no
> explicit EAL --iova-mode parameter specified in the command
> line the rte_eal_init() tried to deduce VA or PA m
Hi Zeng,
On Thu, Apr 13, 2023 at 01:34:00PM +0800, Zhichao Zeng wrote:
> This commit calculates the correct pseudo header for the
> UDP fragmentation offload by adding UDP_SEG flag.
>
> Signed-off-by: Zhichao Zeng
Sorry for the late feedback.
The patch looks good to me, thanks!
I don't think w
The DPDK can be compiled to be run in IOVA VA mode with
'enable_iova_as_pa=false' meson option. If there is no
explicit EAL --iova-mode parameter specified in the command
line the rte_eal_init() tried to deduce VA or PA mode without
taking into account the above mentioned compile time option,
resul
The patch adds GRO support for TCP/ipv6 packets. This does not
include the support for vxlan, udp ipv6 packets.
Signed-off-by: Kumara Parameshwaran
---
v1:
* Changes to support GRO for TCP/ipv6 packets. This does not include
vxlan changes.
* The GRO is performed only fo
Add functional test for mcore dispatch model including graph clone,
graph model set/get, node worker affinity, graph worker binding/unbinding.
Signed-off-by: Haiyue Wang
Signed-off-by: Cunming Liang
Signed-off-by: Zhirun Yan
---
app/test/test_graph.c | 130 +
Update graph documentation to introduce new multicore dispatch model.
Signed-off-by: Haiyue Wang
Signed-off-by: Cunming Liang
Signed-off-by: Zhirun Yan
---
doc/guides/prog_guide/graph_lib.rst | 68 +++--
1 file changed, 64 insertions(+), 4 deletions(-)
diff --git a/doc
Add new parameter "model" to choose mcore dispatch or rtc model.
And in dispatch model, the node will affinity to worker core successively.
Note:
RTE_GRAPH_MODEL_SELECT is set to GRAPH_MODEL_RTC by default. Must set
model the same as RTE_GRAPH_MODEL_SELECT If set it as rtc or mcore
dispatch explic
Add stats for cross-core dispatching scheduler if stats collection is
enabled.
Signed-off-by: Haiyue Wang
Signed-off-by: Cunming Liang
Signed-off-by: Zhirun Yan
---
lib/graph/graph_debug.c| 6 ++
lib/graph/graph_stats.c| 76 +++---
lib/g
This patch introduces the task scheduler mechanism to enable dispatching
tasks to another worker cores. Currently, there is only a local work
queue for one graph to walk. We introduce a scheduler worker queue in
each worker core for dispatching tasks. It will perform the walk on
scheduler work queu
This patch enables to chose new scheduler model. Must define
RTE_GRAPH_MODEL_SELECT before including rte_graph_worker.h
to enable specific model choosing.
Signed-off-by: Haiyue Wang
Signed-off-by: Cunming Liang
Signed-off-by: Zhirun Yan
---
lib/graph/rte_graph_worker.h | 12
1 fil
This patch enables to create and destroy scheduling workqueue into
common graph operations.
Signed-off-by: Haiyue Wang
Signed-off-by: Cunming Liang
Signed-off-by: Zhirun Yan
---
lib/graph/graph.c | 10 ++
1 file changed, 10 insertions(+)
diff --git a/lib/graph/graph.c b/lib/graph/grap
This patch introduces key functions to allow a worker thread to
enable enqueue and move streams of objects to the next nodes over
different cores for mcore dispatch model.
Signed-off-by: Haiyue Wang
Signed-off-by: Cunming Liang
Signed-off-by: Zhirun Yan
---
lib/graph/graph.c
Add graph_mcore_dispatch_wq_node to hold graph scheduling workqueue
node.
Signed-off-by: Haiyue Wang
Signed-off-by: Cunming Liang
Signed-off-by: Zhirun Yan
---
lib/graph/graph.c | 2 ++
lib/graph/graph_populate.c | 1 +
lib/graph/graph_private.h | 12
This patch adds graph API for supporting to clone the graph object for
a specified worker core. The new graph will also clone all nodes.
Signed-off-by: Haiyue Wang
Signed-off-by: Cunming Liang
Signed-off-by: Zhirun Yan
---
lib/graph/graph.c | 89 +++
Move clone_name() into graph_private.h as a common function for both node
and graph to naming a new cloned object.
Signed-off-by: Haiyue Wang
Signed-off-by: Cunming Liang
Signed-off-by: Zhirun Yan
---
lib/graph/graph_private.h | 41 +++
lib/graph/node.c
Add lcore_id for graph to hold affinity core id where graph would run on.
Add bind/unbind API to set/unset graph affinity attribute. lcore_id will
be set as MAX by default, it means not enable this attribute.
Signed-off-by: Haiyue Wang
Signed-off-by: Cunming Liang
Signed-off-by: Zhirun Yan
---
Add lcore_id for node to hold affinity core id and impl
rte_graph_model_mcore_dispatch_lcore_affinity_set to set node affinity
with specific lcore.
Signed-off-by: Haiyue Wang
Signed-off-by: Cunming Liang
Signed-off-by: Zhirun Yan
---
lib/graph/graph_private.h | 2 +
lib/graph
Add new get/set APIs to configure graph worker model which is used to
determine which model will be chosen.
Signed-off-by: Haiyue Wang
Signed-off-by: Cunming Liang
Signed-off-by: Zhirun Yan
---
lib/graph/meson.build | 1 +
lib/graph/rte_graph_worker.c| 21 ++
lib
To support multiple graph worker model, split graph into common
and default. Naming the current walk function as rte_graph_model_rtc
cause the default model is RTC(Run-to-completion).
Signed-off-by: Haiyue Wang
Signed-off-by: Cunming Liang
Signed-off-by: Zhirun Yan
---
lib/graph/graph_pcap.c
Node process is a single and reusable block, move the code into an inline
function.
Signed-off-by: Haiyue Wang
Signed-off-by: Cunming Liang
Signed-off-by: Zhirun Yan
---
lib/graph/rte_graph_model_rtc.h | 20 ++---
lib/graph/rte_graph_worker_common.h | 33 +++
Rename rte_graph_work.h to rte_graph_work_common.h for supporting
multiple graph worker model.
Signed-off-by: Haiyue Wang
Signed-off-by: Cunming Liang
Signed-off-by: Zhirun Yan
---
MAINTAINERS | 3 ++-
lib/graph/graph_pcap.c
V8:
No performance dorp for original l3fwd-graph and graph_perf_autotest.
Update graph model set/get functions and add graph_model_is_valid() in patch 04.
Update doc for new scheme usage(choose model in runtime or compile time).
Update dispatch schedule struct into union.
Change enum rte_graph_wor
On 5/29/2023 3:26 AM, Jie Hai wrote:
>
> Each stream has a read-only "disabled" field that control if this
> stream should be used to forward. This field depends on states
> of Rx/Tx queues, please see
> commit 3c4426db54fc ("app/testpmd: do not poll stopped queues").
>
> Currently, the testpmd
31/05/2023 08:02, Xia, Chenbo:
> > -Original Message-
> > From: Thomas Monjalon
> > Sent: Wednesday, May 31, 2023 12:02 AM
> > To: dev@dpdk.org
> > Cc: sta...@dpdk.org; Gaetan Rivet ; David Marchand
> >
> > Subject: [PATCH] pci: fix comment referencing renamed function
> >
> > When renam
17/03/2023 16:24, Xueming Li:
> The bus cleanup callback is used to sunset all devices on bus
> gracefully. This patch supports the callback by unplug all
> devices on auxiliary bus.
>
> Signed-off-by: Xueming Li
Applied, thanks.
Hi Maxime,
Kind reminder to review this series when you have a chance:
https://patchwork.dpdk.org/project/dpdk/list/?series=28189
Thanks,
Hernan
> -Original Message-
> From: Vargas, Hernan
> Sent: Thursday, May 25, 2023 1:28 PM
> To: dev@dpdk.org; maxime.coque...@redhat.com; gak...@marv
On 6/2/2023 7:29 PM, Mike Pattrick wrote:
> Previously the noisy neighbour vnf simulation would only operate in io
> mode, forwarding packets as is. However, this limited the usefulness of
> noisy neighbour simulation.
>
> This feature has now been expanded to supporting mac, macswap, and
> 5tswap
21/03/2023 20:31, lon...@linuxonhyperv.com:
> From: Long Li
>
> Implement VMBUS cleanup callback from eal_cleanup().
>
> Signed-off-by: Long Li
> ---
> drivers/bus/vmbus/vmbus_common.c | 29 +
> 1 file changed, 29 insertions(+)
>
> diff --git a/drivers/bus/vmbus/vm
On 6/1/2023 5:23 PM, Kevin Traynor wrote:
> On 26/05/2023 18:32, Mike Pattrick wrote:
>> On Fri, May 26, 2023 at 11:34 AM Kevin Traynor
>> wrote:
>>>
>>> On 17/04/2023 19:55, Mike Pattrick wrote:
Previously the noisy neighbour vnf simulation would only operate in io
mode, forwarding pack
01/06/2023 16:42, David Marchand:
> On Fri, May 26, 2023 at 11:03 AM Kaisen You wrote:
> >
> > When a DPDK application is started on only one numa node, memory is
> > allocated for only one socket. When interrupt threads use memory,
> > memory may not be found on the socket where the interrupt thr
1 - 100 of 186 matches
Mail list logo
