On Thu, Apr 05, 2001 at 10:57:01PM -0500, Craig A. Berry wrote:
> At 07:07 PM 4/5/2001 -0500, Jarkko Hietaniemi wrote:
> >[ 9560] By: jhi on 2001/04/05 17:47:01
> > Log: Introduce d_u32align / U32_REQUIRES_ALIGNMENT,
>
>
> With Compaq C V6.2-003 on OpenVMS Alpha V7.2-1 I get:
>
> CC/DECC
>/Include=[]/Standard=Relaxed_ANSI/Prefix=All/Obj=.obj/NoList/NOANSI_ALIAS/Define=(U32_ALIGNMENT_REQUIRED,"VERSION=""2.13""","XS_VERSION=""2.13""")/Include=([---])/Optimize
> MD5.c
>
> #define U32_ALIGNMENT_REQUIRED /**/
> ....................................^
> %CC-W-MACROREDEF, The redefinition of the macro "U32_ALIGNMENT_REQUIRED" conflicts
>with a current definition because the replacement lists differ. The redefinition is
>now in effect.
> at line number 3336 in file D0:[CRAIG.PERL]CONFIG.H;1
>
> The problem appears to be that Digest::MD5 defines the macro on the command
> line but we also now have it in config.h. I'm not sure whether to chop it
> out of the Makefile.PL for the extension or remove it from config, and I
The Makefile.PL would spot be the first to go, I'd say. I'll take that
code away from Makefile.PL since the U32_ALIGNMENT_REQUIRED is really
supposed to come from config.
> don't know of a way to wrap an "#ifndef" around something when we generate
> config.h. One could also argue that this is a compiler bug because after
> stripping comments and trailing whitespace the command-line definition and
> the config definition are identical.
Other than that, how's the snapshot looking in VMS?
--
$jhi++; # http://www.iki.fi/jhi/
# There is this special biologist word we use for 'stable'.
# It is 'dead'. -- Jack Cohen
