Re: RFR: 8283996: Reduce cost of year and month calculations
On Wed, 30 Mar 2022 12:06:39 GMT, Claes Redestad wrote: > A few integer divisions and multiplications can be replaced with test + > addition, leading to a decent speed-up on java.time microbenchmarks such as > `GetYearBench`. Numbers from my local x86 workstation, seeing similar > speed-up on aarch64 and other x86 setups. > > Baseline: > > BenchmarkMode Cnt Score > Error Units > GetYearBench.getYearFromMillisZoneOffsetthrpt 15 18.492 ± > 0.017 ops/ms > GetYearBench.getYearFromMillisZoneRegionthrpt 15 6.121 ± > 0.135 ops/ms > GetYearBench.getYearFromMillisZoneRegionNormalized thrpt 15 18.936 ± > 0.012 ops/ms > GetYearBench.getYearFromMillisZoneRegionUTC thrpt 15 9.283 ± > 0.222 ops/ms > > Patched: > > BenchmarkMode Cnt Score > Error Units > GetYearBench.getYearFromMillisZoneOffsetthrpt 15 20.931 ± > 0.013 ops/ms > GetYearBench.getYearFromMillisZoneRegionthrpt 15 6.858 ± > 0.167 ops/ms > GetYearBench.getYearFromMillisZoneRegionNormalized thrpt 15 20.923 ± > 0.017 ops/ms > GetYearBench.getYearFromMillisZoneRegionUTC thrpt 15 10.028 ± > 0.182 ops/ms > > > Testing: java.time tests locally, CI tier1+2 ongoing. Marked as reviewed by rriggs (Reviewer). - PR: https://git.openjdk.java.net/jdk/pull/8039
Re: RFR: 8283996: Reduce cost of year and month calculations
On Wed, 30 Mar 2022 12:06:39 GMT, Claes Redestad wrote: > A few integer divisions and multiplications can be replaced with test + > addition, leading to a decent speed-up on java.time microbenchmarks such as > `GetYearBench`. Numbers from my local x86 workstation, seeing similar > speed-up on aarch64 and other x86 setups. > > Baseline: > > BenchmarkMode Cnt Score > Error Units > GetYearBench.getYearFromMillisZoneOffsetthrpt 15 18.492 ± > 0.017 ops/ms > GetYearBench.getYearFromMillisZoneRegionthrpt 15 6.121 ± > 0.135 ops/ms > GetYearBench.getYearFromMillisZoneRegionNormalized thrpt 15 18.936 ± > 0.012 ops/ms > GetYearBench.getYearFromMillisZoneRegionUTC thrpt 15 9.283 ± > 0.222 ops/ms > > Patched: > > BenchmarkMode Cnt Score > Error Units > GetYearBench.getYearFromMillisZoneOffsetthrpt 15 20.931 ± > 0.013 ops/ms > GetYearBench.getYearFromMillisZoneRegionthrpt 15 6.858 ± > 0.167 ops/ms > GetYearBench.getYearFromMillisZoneRegionNormalized thrpt 15 20.923 ± > 0.017 ops/ms > GetYearBench.getYearFromMillisZoneRegionUTC thrpt 15 10.028 ± > 0.182 ops/ms > > > Testing: java.time tests locally, CI tier1+2 ongoing. Marked as reviewed by naoto (Reviewer). - PR: https://git.openjdk.java.net/jdk/pull/8039
Re: RFR: 8283996: Reduce cost of year and month calculations
On Wed, 30 Mar 2022 12:06:39 GMT, Claes Redestad wrote: > A few integer divisions and multiplications can be replaced with test + > addition, leading to a decent speed-up on java.time microbenchmarks such as > `GetYearBench`. Numbers from my local x86 workstation, seeing similar > speed-up on aarch64 and other x86 setups. > > Baseline: > > BenchmarkMode Cnt Score > Error Units > GetYearBench.getYearFromMillisZoneOffsetthrpt 15 18.492 ± > 0.017 ops/ms > GetYearBench.getYearFromMillisZoneRegionthrpt 15 6.121 ± > 0.135 ops/ms > GetYearBench.getYearFromMillisZoneRegionNormalized thrpt 15 18.936 ± > 0.012 ops/ms > GetYearBench.getYearFromMillisZoneRegionUTC thrpt 15 9.283 ± > 0.222 ops/ms > > Patched: > > BenchmarkMode Cnt Score > Error Units > GetYearBench.getYearFromMillisZoneOffsetthrpt 15 20.931 ± > 0.013 ops/ms > GetYearBench.getYearFromMillisZoneRegionthrpt 15 6.858 ± > 0.167 ops/ms > GetYearBench.getYearFromMillisZoneRegionNormalized thrpt 15 20.923 ± > 0.017 ops/ms > GetYearBench.getYearFromMillisZoneRegionUTC thrpt 15 10.028 ± > 0.182 ops/ms > > > Testing: java.time tests locally, CI tier1+2 ongoing. I'm happy, based on the assertion that it produces the same result and is faster. - Marked as reviewed by scolebourne (Author). PR: https://git.openjdk.java.net/jdk/pull/8039
Re: RFR: 8283996: Reduce cost of year and month calculations
On Wed, 30 Mar 2022 16:08:15 GMT, Brian Burkhalter wrote: > Looks all right assuming tests pass. Thanks! Tier1+2 testing passed. - PR: https://git.openjdk.java.net/jdk/pull/8039
Re: RFR: 8283996: Reduce cost of year and month calculations
On Wed, 30 Mar 2022 12:06:39 GMT, Claes Redestad wrote: > A few integer divisions and multiplications can be replaced with test + > addition, leading to a decent speed-up on java.time microbenchmarks such as > `GetYearBench`. Numbers from my local x86 workstation, seeing similar > speed-up on aarch64 and other x86 setups. > > Baseline: > > BenchmarkMode Cnt Score > Error Units > GetYearBench.getYearFromMillisZoneOffsetthrpt 15 18.492 ± > 0.017 ops/ms > GetYearBench.getYearFromMillisZoneRegionthrpt 15 6.121 ± > 0.135 ops/ms > GetYearBench.getYearFromMillisZoneRegionNormalized thrpt 15 18.936 ± > 0.012 ops/ms > GetYearBench.getYearFromMillisZoneRegionUTC thrpt 15 9.283 ± > 0.222 ops/ms > > Patched: > > BenchmarkMode Cnt Score > Error Units > GetYearBench.getYearFromMillisZoneOffsetthrpt 15 20.931 ± > 0.013 ops/ms > GetYearBench.getYearFromMillisZoneRegionthrpt 15 6.858 ± > 0.167 ops/ms > GetYearBench.getYearFromMillisZoneRegionNormalized thrpt 15 20.923 ± > 0.017 ops/ms > GetYearBench.getYearFromMillisZoneRegionUTC thrpt 15 10.028 ± > 0.182 ops/ms > > > Testing: java.time tests locally, CI tier1+2 ongoing. Looks all right assuming tests pass. - Marked as reviewed by bpb (Reviewer). PR: https://git.openjdk.java.net/jdk/pull/8039
RFR: 8283996: Reduce cost of year and month calculations
A few integer divisions and multiplications can be replaced with test + addition, leading to a decent speed-up on java.time microbenchmarks such as `GetYearBench`. Numbers from my local x86 workstation, seeing similar speed-up on aarch64 and other x86 setups. Baseline: BenchmarkMode Cnt Score Error Units GetYearBench.getYearFromMillisZoneOffsetthrpt 15 18.492 ± 0.017 ops/ms GetYearBench.getYearFromMillisZoneRegionthrpt 15 6.121 ± 0.135 ops/ms GetYearBench.getYearFromMillisZoneRegionNormalized thrpt 15 18.936 ± 0.012 ops/ms GetYearBench.getYearFromMillisZoneRegionUTC thrpt 15 9.283 ± 0.222 ops/ms Patched: BenchmarkMode Cnt Score Error Units GetYearBench.getYearFromMillisZoneOffsetthrpt 15 20.931 ± 0.013 ops/ms GetYearBench.getYearFromMillisZoneRegionthrpt 15 6.858 ± 0.167 ops/ms GetYearBench.getYearFromMillisZoneRegionNormalized thrpt 15 20.923 ± 0.017 ops/ms GetYearBench.getYearFromMillisZoneRegionUTC thrpt 15 10.028 ± 0.182 ops/ms Testing: java.time tests locally, CI tier1+2 ongoing. - Commit messages: - Reduce cost of year and month calculations Changes: https://git.openjdk.java.net/jdk/pull/8039/files Webrev: https://webrevs.openjdk.java.net/?repo=jdk=8039=00 Issue: https://bugs.openjdk.java.net/browse/JDK-8283996 Stats: 12 lines in 2 files changed: 6 ins; 1 del; 5 mod Patch: https://git.openjdk.java.net/jdk/pull/8039.diff Fetch: git fetch https://git.openjdk.java.net/jdk pull/8039/head:pull/8039 PR: https://git.openjdk.java.net/jdk/pull/8039
