diff options
author | 2023-10-13 10:28:36 +0000 | |
---|---|---|
committer | 2023-10-17 11:56:00 +0000 | |
commit | d5c097bcda44e237ecabcdba9b3dca2348289138 (patch) | |
tree | 46be9f64c07e806c8c1f68e19ac09b0d41f8a635 /compiler/optimizing/intrinsics.h | |
parent | f7bd87edf3b80ce3bbd6e571fd119c878cb79992 (diff) |
riscv64: Implement `CriticalNativeAbiFixupRiscv64`.
And pass integral stack args sign-extended to 64 bits for
direct @CriticalNative calls. Enable direct @CriticalNative
call codegen unconditionally and also enable `HClinitCheck`
codegen and extend the 178-app-image-native-method run-test
to properly test these use cases.
Test: # Edit `run-test` to disable checker, then
testrunner.py --target --64 --ndebug --optimizing
# Ignore 6 pre-existing failures (down from 7).
Bug: 283082089
Change-Id: Ia514c62006c7079b04182cc39e413eb2deb089c1
Diffstat (limited to 'compiler/optimizing/intrinsics.h')
-rw-r--r-- | compiler/optimizing/intrinsics.h | 5 |
1 files changed, 5 insertions, 0 deletions
diff --git a/compiler/optimizing/intrinsics.h b/compiler/optimizing/intrinsics.h index b6c7e1b997..a16b93dfc6 100644 --- a/compiler/optimizing/intrinsics.h +++ b/compiler/optimizing/intrinsics.h @@ -328,6 +328,11 @@ bool IsCallFreeIntrinsic(HInvoke* invoke, Codegenerator* codegen) { return false; } +// Insert a `Float.floatToRawIntBits()` or `Double.doubleToRawLongBits()` intrinsic for a +// given input. These fake calls are needed on arm and riscv64 to satisfy type consistency +// checks while passing certain FP args in core registers for direct @CriticalNative calls. +void InsertFpToIntegralIntrinsic(HInvokeStaticOrDirect* invoke, size_t input_index); + } // namespace art #endif // ART_COMPILER_OPTIMIZING_INTRINSICS_H_ |