[PATCH] Improve DAG combine pass on certain IR vector patterns
chandlerc at google.com
Fri Jan 16 15:13:06 PST 2015
Cool, this looks good to me provided your data indicates this pattern works
well on other targets as well. =]
Thanks for working on it! Any chance you can also look at the fact that we
use vmovq here rather than vmovlpd?
We also at some point need to do a post-processing of the shuffles and
replace ones that use packed double type when there is an equivalent for
packed single type and it removes a bitcast.... It would be really awesome
to get the "obvious" code of vmovlps + vmovhps here (or some variant of
vmovlps that still targeted the floating point vector unit and didn't have
an input dependency... mayb vxorps + vmovlps + vmovhps would be best)
On Fri, Jan 16, 2015 at 3:06 PM, Fiona Glaser <fglaser at apple.com> wrote:
> > You probably want vector-shuffle-256-v8.ll, as your tests are producing
> v8f32 vectors.
> Oops, my mistake, I was thinking 32 was the data type size.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the llvm-commits