On Thu, Mar 22, 2007 at 07:47:35PM -0700, Guy Harris wrote:
>
> On Mar 22, 2007, at 7:10 PM, Stephen Fisher wrote:
> > ber.c:70: warning: ISO C90 does not support the `ll' printf length
> > modifier
>
> ...and if I compile
>
> long long int
> fnord(long long int foo)
> {
> return foo + 17;
> }
>
> with "gcc -O2 -Wall -W -pedantic -c fnord.c", I get
>
> fnord.c:1: warning: ISO C90 does not support ???long long???
> fnord.c:2: warning: ISO C90 does not support ???long long???
>
> (GCC 4.x).
>
> I think -pedantic is a hopeless cause here; the only way to get it to
> keep quiet would be to get rid of FT_INT64/FT_UINT64 and everything
> else that uses 64-bit integral data types, or to support them only in
> 64-bit environments (which means no support on most of the desktop
> machines on which Wireshark is run; someday, they'll all be 64-bit,
> and maybe even be running in 64-bit environments, but that's probably
> a few years off).
The alternative is to compile with -std=c99 for the long long case, but
do we really want to do that? I agree that -pedantic is a bit too much
to tackle, at least for now. Let's work on warnings with the usual
flags for now.
Steve