This is the mail archive of the gcc-bugs@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: ICE: Does anyone care at all?


Michael Meissner wrote:
> 
> On Mon, Jul 10, 2000 at 03:07:07PM -0700, Mike Stump wrote:
> > I don't like the patch.  If people want to fix the bug, or implement
> > nonfaulting functionality that is nicer than faulting, that would be
> > ok.  Explaining the fault, then flauting, is, icky.
> >
> > > Date: Mon, 10 Jul 2000 13:37:17 -0700
> > > From: Bruce Korb <bkorb@sco.COM>
> > > Organization: Santa Cruz Operations
> >
> > > !         fputs ("We know about this.  Someday, it may get fixed.\n", stderr );
> > > !         abort ();
> 
> At least abort() is now usually defined to call fancy_abort() with the __FILE__
> and __LINE__ information that says internal compiler error instead of just
> segfaulting.

Yes.  It is nice.  That is how I found the correct "abort ()" line.
However, it does not tell anyone that the problem has already been
found, reported and not yet fixed.  Maybe add another macro:

  #define known_abort() do{ error (z_known_ice); abort (); }while(0)

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]