Re: [lwip-devel] What is xchar() macro?

[Joel Cunningham]

On 11/01/2017 10:07 AM, Ambroz Bizjak wrote:
> Also, C99 documents isprint() etc as functions not macros, so "#ifndef
> isprint" seems broken in that respect.

In the GCC ARM Embedded toolchain, they are defined as macros for plain 
C (does a check for #ifndef __cplusplus), otherwise there are function 
declarations.

> On Wed, Nov 1, 2017 at 4:05 PM, Ambroz Bizjak <address@hidden> wrote:
> > My guess is that xchar is completely nonstandard and the code works
> > only because <ctype.h> is not included leading to the custom
> > definitions within "#ifndef isprint" to be used. Might be interesting
> > to see what happens if one adds #include <ctype.h> at the top of
> > ip6_addr.c

ctype.h was already implicitly included through existing headers in my 
port, that's how come the #ifndef isprint block was not compiled in and 
thus I didn't get xchar. I agree with you that other ports which have 
successfully compiled the ip6_addr.c must not have had ctype.h included 
or isprint and friends were not macros.

I think a simple fix would be moving the xchar macro out of that block 
and we could guard with #ifndef xchar for whatever bizarre platform 
provides it
> > On Wed, Nov 1, 2017 at 1:05 PM, goldsimon <address@hidden> wrote:
> > > Ambroz Bizjak wrote:
> > > > Clearly xchar(i) returns the uppercase hexadecimal char for the given
> > > > integer in range [0,15], or in other words "0123456789ABCDEF"[i].
> > > Of course. I rather read Joel's question like "where is this function standardized 
> > > to keep the lwip code portable when using it".
> > >
> > > Simon
> > >
> > > _______________________________________________
> > > lwip-devel mailing list
> > > address@hidden
> > > https://lists.nongnu.org/mailman/listinfo/lwip-devel
> _______________________________________________
> lwip-devel mailing list
> address@hidden
> https://lists.nongnu.org/mailman/listinfo/lwip-devel
Joel