Re: [dpdk-dev] [RFC v2 11/14] net: add rte prefix to ip defines

[Olivier Matz <olivier.matz@6wind.com>]

On Mon, Apr 22, 2019 at 09:05:54AM -0700, Stephen Hemminger wrote:
> On Wed, 10 Apr 2019 10:32:15 +0200
> Olivier Matz <olivier.matz@6wind.com> wrote:
> 
> > diff --git a/app/test-acl/main.c b/app/test-acl/main.c
> > index b80179417..eb6294396 100644
> > --- a/app/test-acl/main.c
> > +++ b/app/test-acl/main.c
> > @@ -625,7 +625,7 @@ parse_ipv4_net(const char *in, uint32_t *addr, uint32_t *mask_len)
> >  	GET_CB_FIELD(in, d, 0, UINT8_MAX, '/');
> >  	GET_CB_FIELD(in, m, 0, sizeof(uint32_t) * CHAR_BIT, 0);
> >  
> > -	addr[0] = IPv4(a, b, c, d);
> > +	addr[0] = RTE_IPv4(a, b, c, d);
> 
> IPv4 is a DPDK only macro, could you leave it alone.

It is a public macro defined in rte_ip.h.
I think we should prefix it, to move it in the dpdk namespace. This would
avoid conflicts with an application defining its own IPv4() macro.