[PATCH] D35788: [DAGCombiner] Extending pattern detection for vector shuffle.

[Jatin Bhateja via llvm-commits]

Hi Zvi,

Review patch was generated against my sandbox which was last updated on
17/07/2017. Will rebase with trunk top.

This diff makes vector argument splitting efficient as it uses only minimal
required subvector needed to create a vector_shuffle.

Regards,
Jatin


On Tue, Jul 25, 2017 at 10:27 AM, Soft Dev <soft.devl81 at gmail.com> wrote:

> Hi Zvi,
>
> Review patch was generated against my sandbox which was last updated on
> 17/07/2017. Will rebase with trunk top.
>
> This diff makes vector argument splitting efficient as it uses only
> minimal required subvector needed to create a vector_shuffle.
>
> Regards.
>
> On Tue, Jul 25, 2017 at 1:41 AM, Zvi Rackover via Phabricator <
> reviews at reviews.llvm.org> wrote:
>
>> zvi added a comment.
>>
>> See https://reviews.llvm.org/D35700 where pr33784 is handled a bit
>> differently.
>>
>>
>>
>> ================
>> Comment at: test/CodeGen/X86/pr33784-vector-shuffle.ll:1
>> +; NOTE: Assertions have been autogenerated by
>> utils/update_llc_test_checks.py
>> +
>> ----------------
>> RKSimon wrote:
>> > zvi wrote:
>> > > Can you please commit the test and re-apply the patch so that we can
>> see the change in the generated code?
>> > Please can you add this test to shuffle-vs-trunc-512.ll instead of
>> creating a new test file?
>> This case already exists in test/CodeGen/X86/shuffle-strid
>> ed-with-offset-512.ll?
>> Can you please verify you are rebased on top-of-trunk?
>>
>>
>> https://reviews.llvm.org/D35788
>>
>>
>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20170725/e75a4681/attachment.html>