On Fri, Feb 10, 2023 at 7:54 PM David Marchand
<david.march...@redhat.com> wrote:
>
> On Fri, Feb 10, 2023 at 8:06 AM Jerin Jacob <jerinjac...@gmail.com> wrote:
> >
> > On Fri, Feb 10, 2023 at 12:30 PM Ankur Dwivedi <adwiv...@marvell.com> wrote:
> > >
> > > >On Thu, Feb 9, 2023 at 7:00 PM Ankur Dwivedi <adwiv...@marvell.com>
> > > >wrote:
> > > >>
> > > >> The file rte_mempool_trace.h contains tracepoints which are internal
> > > >> to the mempool library. This file is renamed to mempool_trace.h, and
> > > >> is made an internal header. The tracepoints in this file are removed
> > > >> from the experimental section in version.map file.
> > > >>
> > > >> Signed-off-by: Ankur Dwivedi <adwiv...@marvell.com>
> > > >
> > > >> @@ -47,22 +47,8 @@ EXPERIMENTAL {
> > > >> __rte_mempool_trace_generic_get;
> > > >> __rte_mempool_trace_get_bulk;
> > > >> __rte_mempool_trace_get_contig_blocks;
> > > >
> > > >I think, FP ones also can be removed.
> > >
> > > The FP symbols are used in header file rte_mempool.h. Removing the
> > > symbols will cause build
> > > failure in shared build.
> >
> > OK. Please update the below note documentation only for FP symbols then.
>
> I disagree.
>
> We may enhance this note, but it is not a matter of being FP / non-FP.
> A simple example is the "generic" non-FP tracepoints provided by EAL.
OK. Then what's your recommendation for the document update ?
Generic datatype exposed by EAL trace library or FP trace point
object. or something else?
I am trying to see what needs to be added/changed in documentation?
>
>
> --
> David Marchand
>
