PATCH: [3.3 branch] IA64 bootstrap failure

[H. J. Lu]

On Tue, Jul 08, 2003 at 03:00:01PM -0700, Zack Weinberg wrote:
> "H. J. Lu" <hjl@lucon.org> writes:
> 
> >> in the generated insn-conditions.c. It means the stage1 compiler may
> >> be different from the stage2 compiler, depending on what compiler is
> >> used to bootstrap gcc. I think insn-conditions.c should be the same
> >> regardless the compiler used to boostrap.
> >> 
> >> 
> >
> > I am testing this patch now.
> >
> >
> > H.J.
> >
> > 2003-07-08  H.J. Lu <hongjiu.lu@intel.com>
> >
> > 	* aclocal.m4 (gcc_AC_BUILTIN_CONSTANT_P): New.
> > 	* configure.in: Use it.
> > 	* configure: Regenerated.
> > 	* config.in: Likewise.
> >
> > 	* genconditions.c (write_header): Use __builtin_constant_p if
> > 	HAVE_BUILTIN_CONSTANT_P is defined.
> 
> Rejected, even if it fixes the bug.  Your autoconf test is wrong -
> __builtin_constant_p existed all the way back to 2.7 or earlier, but
> did not do what insn-conditions.c needs until 3.0.1 - and furthermore

Look closely, there is nothing wrong with my autoconf test, which
tests static array initializers.

> this merely papers over the real bug, which is somewhere else.

Without my patch, I can't even compare the dumps between stage1 and
stage2 compilers since they come from different source files.


H.J.