https://bugs.exim.org/show_bug.cgi?id=2152
Michael Haubenwallner <michael.haubenwall...@ssi-schaefer.com> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |michael.haubenwallner@ssi-s
| |chaefer.com
--- Comment #13 from Michael Haubenwallner
<michael.haubenwall...@ssi-schaefer.com> ---
(In reply to Reini Urban from comment #12)
> (In reply to Philip Hazel from comment #11)
> > (In reply to Zoltan Herczeg from comment #10)
> > > > +/* some cmake's define it still */
> > > > +#if defined(__CYGWIN__) && defined(WIN32)
> > > > #define WIN32
> > > > #endif
> > >
> > > What was the conclusion of this change?
> >
> > I applied the patch, changing "defined(WIN32)" into "!defined(WIN32)".
>
> I'll check why it was needed in the first place for Cygwin. But only next
> week. Right now I'm on a hot Greek island without access to my Windows
> machine.
Back from the island already?
These 3 lines were introduced in r849:
https://vcs.pcre.org/pcre2/code/trunk/src/pcre2grep.c?r1=843&r2=849
Something really feels wrong here:
For the earlier #define WIN32, the condition "not defined __CYGWIN__" was
explicitly added, so why should there be need to define WIN32 later with
Cygwin?
Probably the idea was to #undef WIN32 - as in:
#if defined(__CYGWIN__) && defined(WIN32)
-#define WIN32
+#undef WIN32
#endif
Then the cmake comment would make some sense actually!
--
You are receiving this mail because:
You are on the CC list for the bug.
--
## List details at https://lists.exim.org/mailman/listinfo/pcre-dev
