Re: [Ada] Bootstrapping mainline GNAT fails

[Gordon Sadler <gbsadler1 at lcisp dot com>]

On Sun, Mar 17, 2002 at 11:14:52AM +0100, Florian Weimer wrote:
> Laurent Guerby <guerby@acm.org> writes:
> 
> > I got the same error when bootstraping with a 3.13 GNAT, but got
> > past it in another bootstrap by using of my previous GNAT build from
> > CVS GCC as build compiler instead (my build is currently in stage
> > 3). I'll try to work out the details of what works.
> 
> Some additional data points:
> 
>   - Bootstrapping with ACT GNAT 3.14p binaries works on Debian/x86
>     unstable and Red Hat 7.1 (using the 2.8.1-based C compiler
>     supplied by ACT).
> 
>   - Bootstrapping with CC=gnatgcc works on Debian/x86 unstable.
> 
>   - Bootstrapping with CC=gcc-3.0 does NOT work on Debian/x86 unstable.
> 
>   - Bootstrapping with CC=gcc (2.95.4 20011002 (Debian prerelease))
>     does NOT work either.
> 
> Since people can use CC=gnatgcc (thus using the 2.8.1-based C
> compiler), or, if it is not available, GCC 3.1, I don't think this is
> a real problem.  We only have to document this somewhere.
>
Is this the problem you are seeing?

../stage1/xgcc -B../stage1/ -B/usr/local/i686-pc-linux-gnu/bin/ -c -g -O2 -W -Wall -Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes -Wtraditional -pedantic -Wno-long-long  -gnatpg -gnata -I- -I. -I/usr/src/cvs/gcc/gcc/ada /usr/src/cvs/gcc/gcc/ada/namet.adb
+===========================GNAT BUG DETECTED==============================+
| 3.1 (20020212) (i686-pc-linux-gnu) internal error: Illegal instruction   |
| No source file position information available                            |
| Please submit bug report by email to gcc-bugs@gcc.gnu.org.               |
| Use a subject line meaningful to you and us to track the bug.            |
| Include the entire contents of this bug box in the report.               |
| Include the exact gcc or gnatmake command that you entered.              |
| Also include sources listed below in gnatchop format                     |
| (concatenated together with no headers between files).                   |
| (use plain ASCII or MIME attachment).                                    |
| See gnatinfo.txt for full info on procedure for submitting bugs.         |
+==========================================================================+

Please include these source files with error report

/usr/src/cvs/gcc/gcc/ada/namet.adb
/usr/src/cvs/gcc/gcc/ada/namet.ads
/usr/src/cvs/gcc/gcc/ada/alloc.ads
/usr/src/cvs/gcc/gcc/ada/table.ads
/usr/src/cvs/gcc/gcc/ada/types.ads
/usr/src/cvs/gcc/gcc/ada/debug.ads
/usr/src/cvs/gcc/gcc/ada/output.ads
/usr/src/cvs/gcc/gcc/ada/tree_io.ads
/usr/src/cvs/gcc/gcc/ada/widechar.ads
xgcc: Internal error: Segmentation fault (program gnat1)
Please submit a full bug report.
See <URL:http://www.gnu.org/software/gcc/bugs.html> for instructions.
make[3]: *** [namet.o] Error 1

I've been getting this myself since around 9 Mar. I test 3.0 and 3.1
branches daily but have not completed a bootstrap on 3.1 since 8 Mar.
Other possible ... Ada problems:
grep 'Error 1' build.log
make[2]: [stage1-start] Error 1 (ignored)
make[2]: [stage1-start] Error 1 (ignored)
make[2]: [c++.stage1] Error 1 (ignored)
make[2]: [f77.stage1] Error 1 (ignored)
make[2]: [java.stage1] Error 1 (ignored)
make[2]: [objc.stage1] Error 1 (ignored)
make[2]: [objc.stage1] Error 1 (ignored)
make[3]: *** [namet.o] Error 1
make[3]: *** [snames.o] Error 1
make[3]: *** [treepr.o] Error 1
make[3]: *** [namet.o] Error 1
make[3]: *** [snames.o] Error 1

All of this is Debian unstable updated daily, which includes gnat-3.14p
and hasn't changed since mid February.

-- 
Gordon Sadler