Move where intrinsic function attributes are set#5795
Merged
dsharletg merged 8 commits intohalide:masterfrom Mar 10, 2021
Merged
Move where intrinsic function attributes are set#5795dsharletg merged 8 commits intohalide:masterfrom
dsharletg merged 8 commits intohalide:masterfrom
Conversation
dsharletg
reviewed
Mar 9, 2021
src/CodeGen_ARM.cpp
Outdated
|
|
||
| // ABDL - Widening absolute difference | ||
| // Need to be able to handle both signed and unsigned outputs for signed inputs. | ||
| {"vabdl_i8x8", "vabdl_i8x8", Int(16, 8), "vabdl", {Int(8, 8), Int(8, 8)}, ArmIntrinsic::NoMangle | ArmIntrinsic::NoPrefix}, |
Contributor
There was a problem hiding this comment.
I think these should be named "widening_absd"
Contributor
|
Going to start the build bots because the only change I suggested is a rename. |
pranavb-ca
approved these changes
Mar 9, 2021
Contributor
|
Ready to land? |
dsharletg
approved these changes
Mar 10, 2021
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
5 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Currently Halide always sets the
ReadNoneandNoUnwindattributes to all intrinsic calls insideCodeGen_LLVM::call_intrin. The AMX intrinsics introduced in #5780 do access memory, so by setting theReadNoneattribute LLVM can do some invalid optimizations.Instead of setting the attributes on the call to the intrinsic, we can set the attributes on the intrinsic function itself when registering it.
There were some raw calls to
call_intrinin the ARM and Hexagon CodeGen backends. The ARM calls have been converted to usecall_intrin_overload, while the Hexagon calls already use a wrapper function so this has been modified to set the attributes.