Loop unroll fixes

Mark Mitchell mark@codesourcery.com
Thu Oct 25 10:36:00 GMT 2001


--On Thursday, October 25, 2001 09:28:48 AM -0500 Zoltan Hidvegi 
<hzoli@hzoli.2y.net> wrote:

>> You're saying that GCC has a long-standing bug whereby some infinite
>
> Earlier yoday, I showed a testcase for that infinite loop, but that
> was using ints.  The same example can be converted to unsigned ints,
> when gcc-2.95.x no longer fails, but gcc-3.0 and later still does.
> gcc-2.95.x only used the ctr if it was an increment by 1 loop with
> signed loop variable.  The unsigned example is shown below.

I'm sorry to be dense; your mind goes quicker than mine.

If I understand correctly, this reflects a regression in 3.0 relative
to 2.95.x.  Correct?

That's bad; we should try to fix the bug.

Independently of that, does your change make the problem worse, or
just leave it the same?  If the latter, the patch is OK -- but you
should change the comments to indicate that there is this bug, and
we know about it.

Thanks,

--
Mark Mitchell                   mark@codesourcery.com
CodeSourcery, LLC               http://www.codesourcery.com



More information about the Gcc-patches mailing list