On Wed, Oct 30, 2019 at 09:32:08AM +0100, Olivier Matz wrote:
> On Tue, Oct 29, 2019 at 06:20:47PM +0100, Olivier Matz wrote:
> > On Tue, Oct 29, 2019 at 01:31:22PM +0300, Andrew Rybchenko wrote:
> > > On 10/28/19 5:01 PM, Olivier Matz wrote:
> > > > In rte_mempool_populate_default(), we determine the page size,
> > > > which is needed for calc_size and allocation of memory.
> > > >
> > > > Move this in a function and export it, it will be used in next
> > > > commit.
> > > >
> > > > Signed-off-by: Olivier Matz <olivier.m...@6wind.com>
> > >
> > > One question below:
> > > Reviewed-by: Andrew Rybchenko <arybche...@solarflare.com>
> > >
> > > [snip]
> > >
> > > > diff --git a/lib/librte_mempool/rte_mempool_version.map
> > > > b/lib/librte_mempool/rte_mempool_version.map
> > > > index 17cbca460..4eff2767d 100644
> > > > --- a/lib/librte_mempool/rte_mempool_version.map
> > > > +++ b/lib/librte_mempool/rte_mempool_version.map
> > > > @@ -56,5 +56,6 @@ DPDK_18.05 {
> > > > EXPERIMENTAL {
> > > > global:
> > > > + rte_mempool_get_page_size;
> > > > rte_mempool_ops_get_info;
> > > > };
> > >
> > > Should internal function be here?
> > >
> >
> > Good question. Let me ask a friend ;)
>
> I was influenced by a warning saying "rte_mempool_get_page_size is
> flagged as experimental but is not listed in version map", but actually
> it should not be flagged as experimental. I'll remove both.
>
> My friend also suggested me to add it in a private header, which is a
> good idea, but I think it should be in another patch because there are
> already several functions in this case.
Finally, I had to keep it in the API, because the octeontx2 driver will
need it. I also kept the @internal tag in the comment, because I think
this function should only be used in mempool drivers.
