Date: Fri, 7 May 1993 13:17 EST
From: "Wonkoo Kim, EE, U. of Pittsburgh" <WKIM AT vms DOT cis DOT pitt DOT edu>
Subject: Re: gcc hangs: again
To: miller AT hac2arpa DOT hac DOT COM, djgpp AT sun DOT soe DOT clarkson DOT EDU

Peter E. Miller <PMILLER AT dpc2 DOT hdos DOT hac DOT COM> writes:

>> I posted a problem that gcc is stucked when compiling a certain program.
>> In verifying the program, I found a new fact.  My file contains several
>> functions and a function has several return() statements at the end of the
>> function, ...
>>   ...
>> When I changed this conditional branches into if/else if/else way, then gcc
>> did hang.  I.e.
>>   ...
>> didn't hang the gcc.
>
>I took both pieces of code and made the two source files at the end
>of this note. I used
>  gcc -Wall -S test#.c
>on both. During compilation of test1.c a spurious character
>appeared at the top of my screen. Compilation of test2.c doesn't do
>this.
>
>When I used
>  gcc -Wall -O -S test#.c
>a spurious character appeared on my screen during compilation of both,
>at the same location. These tests are repeatable for me.
>
>I have been using DJGPP 1.09 (2.2.2) ever since it was released and
>never noticed a problem like this. I've used it on a 386-16 ISA and a
>486-50 EISA without problems.
>
>---- test1.c --------------------------------------------------------
>#include <float.h>
>#include <math.h>
>
>double
>test (double x, double y)
>{
>  if (y == 0.) return (23.6);
>  if (x == 0.) return (DBL_MAX);
>  return (10.*log10(y / x));
>}
>---- test2.c --------------------------------------------------------
>#include <float.h>
>#include <math.h>
>
>double
>test (double x, double y)
>{
>  double p;
>
>  if (y == 0.)
>    p = 23.6;
>  else if (x == 0.)
>    p = DBL_MAX;
>  else
>    p = 10.*log10(y/x);
>  return(p);
>}
>---------------------------------------------------------------------

These two test programs are short and good enough to produce the problem
occurred on my system as I posted before.  On your machine, gcc behaves
differently, but on my machine, test1.c always makes the gcc hang, but test2.c
is okay without optimazation but gcc hangs with -O option. The results ware the
same with or without -S option. 

I didn't expect this short code could reproduce the same problem!
(So far, only 2 systems(you and I) reported the problem.)

Thanks.

Wonkoo.