🚀 go-pugleaf

Jay Plett <jay@princeton.edu> writes the following to the sun-managers
mailing list:

> One of my users found a bug in the Sun optimizer.  The bug can cause
> programs to perform incorrectly with no indication of failure.

[...]

> Incidentally, the bug is also present in older versions of gcc.


I have verified that gcc 2.0 with all optimizations turned on (-O2
-funroll-all-loops) does indeed handle your sample code correctly, at
least on a sparc machine running SunOS 4.1.1:

     zerkalo% cat testme.c
     #include <stdio.h>

     main()
     {
	     int     m, n, p;
	     int     mhold;
	     int     clunk;

     #ifdef NOBUG
	     clunk = &mhold - &m;
     #endif
	     p = 1;
	     for (m = 1; m <= 2; m++)
		     for (n = 1; n <= 2; n++, p++)
		     {       if (p == 1)
			     {       mhold = m;
				     printf("(p==1)");
			     }
			     printf("\tp=%d, mhold=%d, m=%d\n", p, mhold, m);
		     }
     }


     zerkalo% gcc -O2 -funroll-all-loops testme.c
     zerkalo% ./a.out
     (p==1)  p=1, mhold=1, m=1
	     p=2, mhold=1, m=1
	     p=3, mhold=1, m=2
	     p=4, mhold=1, m=2
     zerkalo%


According to your mail message, this is indeed the correct output.

So gcc 2.0 users, relax!  Your optimizer is doing the right thing!

I also checked gcc 1.40 on sparc and SunOS 4.1.1.  This too produces
the correct output, with -O specified as the optimize switch.

Manavendra K. Thakur			 Internet: thakur@zerkalo.harvard.edu
Systems Programmer, High Energy Division BITNET:   thakur@cfa.BITNET
Harvard-Smithsonian Center for		 DECNET:   CFA::thakur
Astrophysics				 UUCP:	   ...!uunet!mit-eddie!thakur