-
Bug
-
Resolution: Fixed
-
P4
-
25
-
b18
-
riscv
On Riscv, it does generate 2 SubHF which means it's not folded, but there is an extra SubF generated in slow path when TLAB runs out. Although it does not impact correctness.
- caused by
-
JDK-8351515 C2 incorrectly removes double negation for double and float
-
- Resolved
-
- duplicates
-
JDK-8353693 RISC-V: compiler/floatingpoint/TestSubNodeFloatDoubleNegation.java is failing with Zfh
-
- Closed
-
-
JDK-8353693 RISC-V: compiler/floatingpoint/TestSubNodeFloatDoubleNegation.java is failing with Zfh
-
- Closed
-
- relates to
-
JDK-8353732 TestSubNodeFloatDoubleNegation.java generates an extra useless SubF
-
- Open
-
-
JDK-8353730 TestSubNodeFloatDoubleNegation.java fails with native Float16 support
-
- Resolved
-
- links to
-
Commit(master) openjdk/jdk/21db0fdb
-
Review(master) openjdk/jdk/24421