Hello!
I extract JDK 9 code from String, StringLatin1, StringUTF16 and optimize some
functions https://github.com/volodin-aa/openjdk-benchmark
My optimization:
1. Create fullMatchesCI() methods with simple for (...) loop
https://github.com/volodin-aa/openjdk-benchmark/blob/master/src/su/tweak/
Hi Andrey,
On 04/27/2016 12:57 PM, Andrey wrote:
> I publish my JMH benchmark at github
> https://github.com/volodin-aa/openjdk-benchmark
Please note that you really should compete with JDK 9 String, not with
JDK 8. String.equalsIgnoreCase is different in JDK 9, and the obvious
improvement one c
Results for long strings
private static final String _STRING = "This test needs to be better. Always
comparing String and string doesn't tell us much. These strings are extremely
short, and a few more cases would be useful. Please change your units to
something more sensible. I suggest nano
On 27/04/16 10:57, Andrey wrote:
> I publish my JMH benchmark at github
> https://github.com/volodin-aa/openjdk-benchmark
This test needs to be better. Always comparing "String" and "string"
doesn't tell us much. These strings are extremely short, and a few
more cases would be useful.
Please c
I publish my JMH benchmark at github
https://github.com/volodin-aa/openjdk-benchmark
27.04.2016, 01:23, "Andrey" :
> I create simple benchmark (attached). Optimized version more faster:
>
> # JMH 1.12 (released 26 days ago)
> # VM version: JDK 1.8.0_91, VM 25.91-b14
> ..
> # Run complete. Total t
I create simple benchmark (attached). Optimized version more faster:
# JMH 1.12 (released 26 days ago)
# VM version: JDK 1.8.0_91, VM 25.91-b14
...
# Run complete. Total time: 00:04:02
BenchmarkMode Cnt ScoreError Units
StringBenchmark.constConst
On 04/26/2016 03:25 PM, Andrey wrote:
> May be can create optimized regionMatches(...) for use in
> equalsIgnoreCase(...)?
When the HotSpot JVM's just-in-time compiler optimizes this code it
inlines all of these tests, realizes that they are constant values,
and generates no code for them. All t
