[RFC 4/4] eal/malloc: remove type argument from internal malloc routines

[Tyler Retzlaff]

On Fri, Apr 26, 2024 at 03:52:29PM -0700, Stephen Hemminger wrote:
> On Fri, 26 Apr 2024 09:16:27 -0700
> Tyler Retzlaff <roretzla at linux.microsoft.com> wrote:
> 
> > > 
> > > diff --git a/lib/eal/common/eal_common_memzone.c b/lib/eal/common/eal_common_memzone.c
> > > index 32e6b78f87..2d9b6aa3e3 100644
> > > --- a/lib/eal/common/eal_common_memzone.c
> > > +++ b/lib/eal/common/eal_common_memzone.c
> > > @@ -191,14 +191,12 @@ memzone_reserve_aligned_thread_unsafe(const char *name, size_t len,
> > >  	if (len == 0 && bound == 0) {
> > >  		/* no size constraints were placed, so use malloc elem len */
> > >  		requested_len = 0;
> > > -		mz_addr = malloc_heap_alloc_biggest(NULL, socket_id, flags,
> > > -				align, contig);
> > > +		mz_addr = malloc_heap_alloc_biggest(socket_id, flags, align, contig);  
> > 
> > i may have missed if this was discussed already. for the public api i
> > understand for now we need to keep the unused parameter in the function
> > signatures but for internal api/functions i would prefer the parameter
> > be removed entirely.
> > 
> > also somewhat related side-note i don't think msvc has a way of marking
> > function parameters unused as is done with __rte_unused. currently i
> > expand the macro empty and suppress the warning globally which is not
> > great.
> 
> I dropped the parameter from all the internal routines. Are you suggesting having
> an different name/version for use internally?

oh, it looks like i can't read a diff.

ignore me! thanks!