Arm64/SVE: JitStress/JitStressRegs fixes #102543
Merged
Conversation
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
…s LOW_PREDICATE flag
…_LowMaskedOperation
ghost
added
the
area-CodeGen-coreclr
Contributor
Author
|
@dotnet/arm64-contrib |
Contributor
|
Tagging subscribers to this area: @JulieLeeMSFT, @jakobbotsch |
if falseReg == embMaskOp2Reg, we simply generate:
```
sel z16.s, p7, z9.s, z10.s
mla z16.s, p7/m, z10.s, z11.s
```
Here `z10` holds `falseReg` and `embMaskOp2Reg`.
a74nh
reviewed
May 22, 2024
TIHan
approved these changes
May 22, 2024
Contributor
TIHan
left a comment
TIHan
left a comment
There was a problem hiding this comment.
LGTM. Glad we are catching these.
steveharter
pushed a commit
to steveharter/runtime
that referenced
this pull request
May 28, 2024
* handle case for FMA where falseReg == embMaskOp1Reg
* workaround because predicateRegister/vectorRegister are same
* When intrinsic is wrapped in ConditionalSelect, make sure to check its LOW_PREDICATE flag
* Mark AddAcross with HW_Flag_LowMaskedOperation
* double check if ConditionalSelect's op2 is hwintrinsic
* Mark Max with HW_Flag_LowMaskedOperation
* Mark MaxAcross with HW_Flag_LowMaskedOperation
* Mark MinNumber/MaxNumber/MinNumberAcross/MaxNumberAcross with HW_Flag_LowMaskedOperation
* Mark Min/MinAcross with HW_Flag_LowMaskedOperation
* fix the workaround for predicate vs. vector register
* fix the predicate mask preference
* Introduce INS_SCALABLE_OPTS_PREDICATE_MERGE_MOV
* jit format
* revert change to csproj
* remove assert that can not happen for FMA
if falseReg == embMaskOp2Reg, we simply generate:
```
sel z16.s, p7, z9.s, z10.s
mla z16.s, p7/m, z10.s, z11.s
```
Here `z10` holds `falseReg` and `embMaskOp2Reg`.
* revert a condition added for workaround of predicate == vector register
* remove the extra comment
Ruihan-Yin
pushed a commit
to Ruihan-Yin/runtime
that referenced
this pull request
May 30, 2024
* handle case for FMA where falseReg == embMaskOp1Reg
* workaround because predicateRegister/vectorRegister are same
* When intrinsic is wrapped in ConditionalSelect, make sure to check its LOW_PREDICATE flag
* Mark AddAcross with HW_Flag_LowMaskedOperation
* double check if ConditionalSelect's op2 is hwintrinsic
* Mark Max with HW_Flag_LowMaskedOperation
* Mark MaxAcross with HW_Flag_LowMaskedOperation
* Mark MinNumber/MaxNumber/MinNumberAcross/MaxNumberAcross with HW_Flag_LowMaskedOperation
* Mark Min/MinAcross with HW_Flag_LowMaskedOperation
* fix the workaround for predicate vs. vector register
* fix the predicate mask preference
* Introduce INS_SCALABLE_OPTS_PREDICATE_MERGE_MOV
* jit format
* revert change to csproj
* remove assert that can not happen for FMA
if falseReg == embMaskOp2Reg, we simply generate:
```
sel z16.s, p7, z9.s, z10.s
mla z16.s, p7/m, z10.s, z11.s
```
Here `z10` holds `falseReg` and `embMaskOp2Reg`.
* revert a condition added for workaround of predicate == vector register
* remove the extra comment
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
3 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.
falseReg == embMaskOp1Regmovwhen used as alias forselbecause predicateRegister/vectorRegister are sameConditionalSelect, check the intrinsic flag if it needs low register maskHW_Flag_LowMaskedOperationThere are still some functional failures with JitStress/JitStressRegs, but wanted to send this out.