bug in g77 -O2 optimization found -- resending with attachment

[Jonathan DeSena]

Bernd Schmidt wrote:

> On Mon, 8 Jan 2001, Jonathan DeSena wrote:
> 
> 
>> Synopsis
>>      g77 generates incorrect code for a particular looping construct
>> when optimization of -O2 or higher is used.
>> 
>> Release     2.95.3 19991030 (prerelease)
> 
> 
> Can you reproduce this problem with 2.95.2 and the more recent 2.95.3 test
> release as well?
> 
> 
> Bernd
I just tested this on a Sun Ultra5 (Solaris 2.7 SPARC IIi) with g77 
version 2.95.2 19991024.  The result was the same: the code produced the 
exact same numbers, which of course is the wrong result.  (This 
surprised me a little.)  I do not have the output of g77 -v now as I am 
writing this on my Linux PC, which is separately located from the Sun. 
If you need it, I could get this output and send it too.

Also I could try downloading a more recent version of 2.95.3 if necessary.

Thanks,
JTD