> -----Original Message----- > From: Jerin Jacob <jerinjac...@gmail.com> > Sent: Wednesday, June 7, 2023 8:46 PM > To: Yan, Zhirun <zhirun....@intel.com> > Cc: dev@dpdk.org; jer...@marvell.com; kirankum...@marvell.com; > ndabilpu...@marvell.com; step...@networkplumber.org; > pbhagavat...@marvell.com; Liang, Cunming <cunming.li...@intel.com>; Wang, > Haiyue <haiyue.w...@intel.com>; mattias.ronnblom > <mattias.ronnb...@ericsson.com> > Subject: Re: [PATCH v9 17/17] doc: update multicore dispatch model in graph > guides > > On Wed, Jun 7, 2023 at 9:30 AM Zhirun Yan <zhirun....@intel.com> wrote: > > > > Update graph documentation to introduce new multicore dispatch model. > > Please squash this to relevant implementation patches, No need for separate > patch. This is the contribution guideline followed. > Got it. I will squash this to the patch of walk model choosing in patch 13. > > > > Signed-off-by: Haiyue Wang <haiyue.w...@intel.com> > > Signed-off-by: Cunming Liang <cunming.li...@intel.com> > > Signed-off-by: Zhirun Yan <zhirun....@intel.com> > > --- > > doc/guides/prog_guide/graph_lib.rst | 68 > > +++++++++++++++++++++++++++-- > > 1 file changed, 64 insertions(+), 4 deletions(-) > > > > diff --git a/doc/guides/prog_guide/graph_lib.rst > > b/doc/guides/prog_guide/graph_lib.rst > > index 1cfdc86433..8c2c2816ed 100644 > > --- a/doc/guides/prog_guide/graph_lib.rst > > +++ b/doc/guides/prog_guide/graph_lib.rst > > @@ -189,14 +189,74 @@ In the above example, A graph object will be > > created with ethdev Rx node of port 0 and queue 0, all ipv4* nodes in > > the system, and ethdev tx node of all ports. > > > > -Multicore graph processing > > -~~~~~~~~~~~~~~~~~~~~~~~~~~ > > -In the current graph library implementation, specifically, > > -``rte_graph_walk()`` and ``rte_node_enqueue*`` fast path API > > functions > > +Graph models chossing > > Graph models > > > +~~~~~~~~~~~~~~~~~~~~~ > > +Currently, there are 2 different walking models. Use macro > > +RTE_GRAPH_MODEL_SELECT to set the model in compile time. Also offers > > +the ability to choose models in runtime > > by not defining RTE_GRAPH_MODEL_SELECT. > > > +For application, must #define RTE_GRAPH_MODEL_SELECT before including > > +rte_graph_worker.h > > Some rewording as suggestion. > > Graph models > ~~~~~~~~~~~~ > There are two different kinds of graph walking models. User can select the > model using ``rte_graph_worker_model_set()`` API. If the application decides > to > use only one model, the fast path check can be avoided by defining the model > with RTE_GRAPH_MODEL_SELECT. For example: > > .. code-block:: console > > #define RTE_GRAPH_MODEL_SELECT RTE_GRAPH_MODEL_RTC #include > "rte_graph_worker.h" > >
Thanks for your refine. I will update in next version. > > > + > > +In l3fwd-graph, set RTE_GRAPH_MODEL_SELECT as the model explicitly > > +for performance-sensitive use case. > > +Or set the macro as GRAPH_MODEL_MCORE_RUNTIME_SELECT. And parse > > +``"--model=NAME"`` in cmdline and use > > +``rte_graph_worker_model_set()`` to set the walking model in runtime. > > Please move to doc/guides/sample_app_ug/l3_forward_graph.rst > Yes, will move it and squash this part into example patch. > > + > > +RTC (Run-To-Completion) > > +^^^^^^^^^^^^^^^^^^^^^^^ > > +This is the default graph walking model. Specifically, > > +``rte_graph_walk_rtc()`` and ``rte_node_enqueue*`` fast path API > > +functions > > are designed to work on single-core to have better performance. > > The fast path API works on graph object, So the multi-core graph > > processing strategy would be to create graph object PER WORKER. > > > > +Example: > > + > > +Graph: node-0 -> node-1 -> node-2 @Core0. > > + > > +.. code-block:: diff > > + > > + + - - - - - - - - - - - - - - - - - - - - - + > > + ' Core #0 ' > > + ' ' > > + ' +--------+ +---------+ +--------+ ' > > + ' | Node-0 | --> | Node-1 | --> | Node-2 | ' > > + ' +--------+ +---------+ +--------+ ' > > + ' ' > > + + - - - - - - - - - - - - - - - - - - - - - + > > + > > +Dispatch model > > +^^^^^^^^^^^^^^ > > +The dispatch model enables a cross-core dispatching mechanism which > > +employs a scheduling work-queue to dispatch streams to other worker > > +cores which being associated with the destination node. > > + > > +Use ``rte_graph_model_mcore_dispatch_lcore_affinity_set()`` to set > > +lcore affinity with the node. > > +Each worker core will have a graph repetition. Use > > +``rte_graph_clone()`` to clone graph for each worker and > > +use``rte_graph_model_mcore_dispatch_core_bind()`` to bind graph with the > worker core. > > + > > +Example: > > + > > +Graph topo: node-0 -> Core1; node-1 -> node-2; node-2 -> node-3. > > +Config graph: node-0 @Core0; node-1/3 @Core1; node-2 @Core2. > > + > > +.. code-block:: diff > > + > > + + - - - - - -+ +- - - - - - - - - - - - - + + - - - - - -+ > > + ' Core #0 ' ' Core #1 ' ' Core #2 ' > > + ' ' ' ' ' ' > > + ' +--------+ ' ' +--------+ +--------+ ' ' +--------+ ' > > + ' | Node-0 | - - - ->| Node-1 | | Node-3 |<- - - - | Node-2 | ' > > + ' +--------+ ' ' +--------+ +--------+ ' ' +--------+ ' > > + ' ' ' | ' ' ^ ' > > + + - - - - - -+ +- - -|- - - - - - - - - - + + - - -|- - -+ > > + | | > > + + - - - - - - - - - - - - - - - - + > > + > > + > > In fast path > > ~~~~~~~~~~~~ > > Typical fast-path code looks like below, where the application > > -- > > 2.37.2 > >
