[PATCH] Reassociate: optimize I1 + x - I2 into x when I1 and I2 are identical
chandlerc at google.com
Sun Oct 5 16:13:37 PDT 2014
LGTM. I feel like we might want to be doing something more systematic to
ensure that identical sub-expressions are correctly modeled throughout the
reassociate pass, but this seems like an obvious strict improvement. =]
On Thu, Oct 2, 2014 at 9:15 AM, Dmitri Shtilman <dshtilman at icloud.com>
> Please review.
> On Sep 24, 2014, at 3:49 PM, Dmitri Shtilman <dshtilman at icloud.com> wrote:
> > Hi all,
> > The following small patch gives the Reassociate pass a bit more
> flexibility and autonomy when optimizing expressions.
> > Particularly, it addresses cases where Reassociate breaks Subtracts but
> then fails to optimize combinations like I1 + -I2 where I1 and I2 have the
> same rank and are identical.
> > Please review.
> > Dmitri
> > <Reassociate.patch>_______________________________________________
> > llvm-commits mailing list
> > llvm-commits at cs.uiuc.edu
> > http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
> llvm-commits mailing list
> llvm-commits at cs.uiuc.edu
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the llvm-commits