-
Enhancement
-
Resolution: Fixed
-
P4
-
25
-
b05
-
generic
SRC = Float16.valueOf(src[i]);
RES = Float16.max(SRC, Float16.NaN)
The reason is that the current scheme to detect Float16 operations for some of the operations is based on generic pattern matching. In the above case, the second argument gets constant folded to a Float constant, thus bypassing the existing pattern, which expects the inputs to be ConvHF2F.
- causes
-
JDK-8361582 AArch64: Some ConH values cannot be replicated with SVE
-
- Open
-
-
JDK-8361037 [ubsan] compiler/c2/irTests/TestFloat16ScalarOperations division by 0
-
- Resolved
-
-
JDK-8362594 Aarch64: Fix JTREG test - hotspot/jtreg/compiler/vectorization/TestFloat16VectorOperations.java failure on 256-bit SVE machines
-
- Closed
-
- duplicates
-
JDK-8352154 Folding optimizations for Float16
-
- Closed
-
- relates to
-
JDK-8342103 C2 compiler support for Float16 type and associated scalar operations
-
- Resolved
-
- links to
-
Commit(master) openjdk/jdk/a49ecb26
-
Review(master) openjdk/jdk/24179