[PATCH] Improve DAG combine pass on certain IR vector patterns

[Eric Christopher]

On Fri Jan 16 2015 at 12:08:21 PM Quentin Colombet <qcolombet at apple.com>
wrote:

> Hi Fiona,
>
> The approach looks overall good to me, but I have a few remarks and
> questions.
> 1. Run your patch through clang-format please. The patch does not follow
> the LLVM formatting guidelines.
> 2. What is the impact of this on arm64 and armv7s generated code? Although
> the approach makes sense to me, I want to be sure we do not degrade other
> targets. Note that I do not expect you to run tests if you cannot :).
> 3. What are the runtime performance impact on x86_64, with and without
> -mavx2?
> 4. Add a run line for avx  too:
> +; RUN: llc < %s -march=x86-64 -mattr=+avx2 | FileCheck %s
> 5. Should we use a triple instead of march?
>
>
I'll reply to this part: Probably not. The code generation likely shouldn't
be to be target specific past x86_64. If it is then we've probably got
something else to look at ;)

-eric


> Thanks,
> -Quentin
>
>
> On Jan 16, 2015, at 11:24 AM, Fiona Glaser <fglaser at apple.com> wrote:
>
> > Talked about this one offline and decided to implement Mehdi’s strategy.
> >
> > Fiona
> >
> > <patch.diff>_______________________________________________
> > 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
> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20150116/b46ace09/attachment.html>