-
Enhancement
-
Resolution: Duplicate
-
P3
-
12
In the Graal version the hottest part is the temporary buffer clearing related to JDK-8207775 but that part is not as hot in the profile on C2.
Run just this single micro with:
java -jar jmh-jdk-microbenchmarks-1.0-SNAPSHOT.jar crypto.full.AESBench.decrypt -f 1 -wi 8 -p algorithm=AES/CBC/PKCS5Padding
Run just this single micro with:
java -jar jmh-jdk-microbenchmarks-1.0-SNAPSHOT.jar crypto.full.AESBench.decrypt -f 1 -wi 8 -p algorithm=AES/CBC/PKCS5Padding
- duplicates
-
JDK-8238565 [Graal] No vectorized Arrays.fill causes several regressions to C2
- Open
- relates to
-
JDK-8237401 [Graal] 50% regression to C2 on crypto.small.CipherBench.GCM.decrypt
- Closed
-
JDK-8213319 [Graal] 30% regression to C2 on AESBench.encrypt AES/ECB/PKCS5Padding
- Closed