[Bug tree-optimization/22041] Reverse loop order for increased efficiency
rakdver at atrey dot karlin dot mff dot cuni dot cz
gcc-bugzilla@gcc.gnu.org
Sun May 28 14:25:00 GMT 2006
------- Comment #5 from rakdver at atrey dot karlin dot mff dot cuni dot cz 2006-05-28 14:25 -------
Subject: Re: Reverse loop order for increased efficiency
> ------- Comment #4 from tkoenig at gcc dot gnu dot org 2006-05-28 14:10 -------
> (In reply to comment #3)
> > Dunno if tree loop reversal is already there, but Zdenek probably likes to know
> > what bugs he is fixing...
>
> Loop reversal doesn't occur yet.
>
> gcc version 4.2.0 20060521 (experimental) gives us:
>
> L4:
> movl (%ecx), %eax
> addl $1, %ebx
> addl %edi, %ecx
> movl %eax, (%edx)
> addl %esi, %edx
> cmpl 16(%ebp), %ebx
> jne L4
loop reversal was removed with the old loop optimizer; the tree level
loop reversal patch
(http://gcc.gnu.org/ml/gcc-patches/2006-01/msg01851.html) was not
reviewed.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22041
More information about the Gcc-bugs
mailing list