Re: My first compiler bug: work around or redesign?

[onox <denkpadje@gmail.com>]

On Mar 29, 10:25 am, Ludovic Brenta <ludo...@ludovic-brenta.org>
wrote:
> Natasha Kerensikova wrote on comp.lang.ada:
>
> > Ludovic Brenta wrote:
> >> Report it to the FSF athttp://gcc.gnu.org/bugzilla. IMHO, a public bug
> >> database is worth a thousand private ones.
>
> >http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52752
>
> > Below is the minimal testcase I have worked out.
>
> > Is there anything I should do at this point? Or will everything leading
> > to a fix be handled from there?
>
> Just wait for someone to fix the bug and then consider backporting
> the fix into your distribution's version of GNAT.
>
> But Per Sandberg wrote the day before on comp.lang.ada:
>
> > Any way it is reported.
>
> which pointed me tohttp://gcc.gnu.org/PR52735
>
> So you should indicate that your bug is a duplicate with a more
> minimal test case.
>
> --
> Ludovic Brenta.

It seems the error is triggered in a different .c file with GPL 2010:

gnatmake -v -gnat05 -gnatd -O0 testcase.adb:

+===========================GNAT BUG
DETECTED==============================+
| GPL 2010 (20100603) (i686-pc-linux-gnu) GCC
error:                       |
| in simplify_subreg, at simplify-rtx.c:
4924                               |
| Error detected around test.adb:16
...
compilation abandoned

gnatmake -v -gnat05 -gnatd -O1 testcase.adb:

+===========================GNAT BUG
DETECTED==============================+
| GPL 2010 (20100603) (i686-pc-linux-gnu) GCC
error:                       |
| in simplify_subreg, at simplify-rtx.c:
4924                               |
| Error detected around test.adb:21
...
raised TYPES.UNRECOVERABLE_ERROR : comperr.adb:423

gnatmake -v -gnat05 -gnatd -O2 testcase.adb:

gcc -c -gnat05 -gnatd -O2 test.adb

raised STORAGE_ERROR : stack overflow (or erroneous memory access)

No gnat bug box with -O2 :/