RFR: Cleaning up human-readable output
Staffan Friberg
staffan.friberg at oracle.com
Fri Aug 30 07:40:53 PDT 2013
I like it. Much cleaner and readable.
What does the Fork line refer to? Number of executed benchmarks in
forked mode?
If that is the case perhaps change to something like.
Run Progress: 9 of 10, [Forked | In Process]
//Staffan
On 2013-08-30 02:15, Aleksey Shipilev wrote:
> Hi,
>
> The human-readable output seems to be overloaded with the info. Let's
> prune it! I have the change which transforms the output like this:
>
> ------------------ 8< ------------------------------------------------
>
> # Runtime (per iteration): 100ms
> # Iterations: 5
> # Thread count: 1
> # Threads will synchronize iterations
> # Benchmark mode: Throughput, ops/time
> # Running: org.openjdk.jmh.samples.JMHSample_01_HelloWorld.hello
> # Warmup Iteration 1 (100ms in 1 thread): 0.634 ops/nsec
> # Warmup Iteration 2 (100ms in 1 thread): 0.743 ops/nsec
> # Warmup Iteration 3 (100ms in 1 thread): 0.844 ops/nsec
> # Warmup Iteration 4 (100ms in 1 thread): 0.841 ops/nsec
> # Warmup Iteration 5 (100ms in 1 thread): 0.843 ops/nsec
> Iteration 1 (100ms in 1 thread): 0.831 ops/nsec
> Iteration 2 (100ms in 1 thread): 0.810 ops/nsec
> Iteration 3 (100ms in 1 thread): 0.923 ops/nsec
> Iteration 4 (100ms in 1 thread): 0.830 ops/nsec
> Iteration 5 (100ms in 1 thread): 0.830 ops/nsec
>
> Run result "***": 0.845 ±(95%) 0.056 ±(99%) 0.093 ops/nsec
> Run statistics "***": min = 0.810, avg = 0.845, max = 0.923, stdev = 0.045
> Run confidence intervals "***": 95% [0.789, 0.900], 99% [0.752, 0.937]
>
> ------------------ 8< ------------------------------------------------
>
> ...into this...
>
> ------------------ 8< ------------------------------------------------
>
> # Fork: 9 of 10
> # Running: org.openjdk.jmh.samples.JMHSample_01_HelloWorld.hello
> # Warmup: 5 iterations, 100 ms each
> # Measurement: 5 iterations, 100 ms each
> # Threads: 1 thread, will synchronize iterations
> # Benchmark mode: Throughput, ops/time
> # Warmup Iteration 1: 0.089 ops/nsec
> # Warmup Iteration 2: 1.857 ops/nsec
> # Warmup Iteration 3: 1.959 ops/nsec
> # Warmup Iteration 4: 1.964 ops/nsec
> # Warmup Iteration 5: 1.953 ops/nsec
> Iteration 1: 1.964 ops/nsec
> Iteration 2: 1.964 ops/nsec
> Iteration 3: 1.964 ops/nsec
> Iteration 4: 1.968 ops/nsec
> Iteration 5: 1.971 ops/nsec
>
> Result: 1.966 ±(95%) 0.004 ±(99%) 0.006 ops/nsec
> Statistics: (min, avg, max) = (1.964, 1.966, 1.971), stdev = 0.003
> Confidence intervals: 95% [1.962, 1.970], 99% [1.960, 1.972]
>
> ------------------ 8< ------------------------------------------------
>
> Any objections for the change?
>
> -Aleksey.
More information about the jmh-dev
mailing list