RE: [RFA] Use "-idirafter /usr/include/w32api" in cygwin specs

["Billinghurst, David (CRTS)" <David dot Billinghurst at riotinto dot com>]

The first problem is fixed in CVS of 2001-03-12 01:00 UTC, but the failure
of stage2/xgcc to compile f/top.c remains.  If I copy stage2/f/top.o and
continue the bootstrap then all is OK.

This second failure seems to be caused by this change, which was made approx
2001-03-11 00:12 UTC, as the gcc-3.0 branch just prior (2001-03-11 00:10:00
UTC) bootstraps.  I have had a look at the diffs, and they seem OK to my
untrained eye.

 

> -----Original Message-----
> From:	'Christopher Faylor' [SMTP:cgf@redhat.com]
> Sent:	Monday, March 12, 2001 3:38 PM
> To:	Billinghurst, David (CRTS)
> Cc:	'gcc-patches@gcc.gnu.org'
> Subject:	Re: [RFA] Use "-idirafter /usr/include/w32api" in cygwin
> specs
> 
> On Mon, Mar 12, 2001 at 02:59:16AM -0000, Billinghurst, David (CRTS)
> wrote:
> >With this change the 3.0 branch does not bootstrap for me, while it did
> so
> >around 12 hours previously.  There is one trivial problem with this
> patch,
> >and a second perhaps unrelated one that prevents a bootstrap. 
> >
> >System is PII - NT4 SP5 - 20010222 cygwin1.dll snapshot.
> >
> >Configure and build in usual way
> >${srcdir}/configure
> >make bootstrap
> >
> >Firstly, the stage1 compiler generates warnings
> >
> >	<command line>: warning: "cpu" re-asserted
> >	<command line>: warning: "machine" re-asserted
> >
> >This is due to two occurrences of %{cpp_cpu} in the *cpp section of the
> >specs.  See below for some more details
> 
> Are you sure you're up-to-date with regard to CVS?  I did check in a
> typo which added an additional cpp_cpu from an erroneous cut-and-paste
> from the cygwin release but I rectified it within an hour afterwards.
> 
> cgf