search

tensorflow / mlir

"Multi-Level Intermediate Representation" Compiler Infrastructure
1.74k stars 259 forks source link

Moving the GPUIndexIntrinsicOpLowering template to a common location #163

Closed deven-amd closed 5 years ago

deven-amd commented 5 years ago

The GPUIndexIntrinsicOpLowering template is currently is used by the code in both the GPUToNVVM and GPUToROCDL dirs. Moving it to a common location to remove code duplication.

This is a followup to the request made in PR #154

@joker-eph @sherhut . Not quite sure whether I got this right...please review.

ftynse commented 5 years ago

~@deven-amd could you please rebase?~

done by @sherhut locally, no further action needed, thanks!