# Detecting actual CPU count: 32 detected
# JMH 1.6.3 (released 5 days ago)
# VM invoker: /aurora/clredest/contended/jdk-cache64/jre/bin/java
# VM options: -XX:-UseBiasedLocking -XX:+UseHeavyMonitors
# Warmup: <none>
# Measurement: 1 iterations, 500 ms each, 4000000 calls per op
# Timeout: 10 min per iteration
# Threads: 32 threads
# Benchmark mode: Single shot invocation time
# Benchmark: org.openjdk.micros.contended.ContendedMonitorBench.synchronizeAndAccessContendedHeader
# Parameters: (inflate = false, monitors = 8, step = 8)

# Run progress: 0.00% complete, ETA 00:00:00
# Fork: 1 of 1
# Preparing profilers: perf 
# Profilers consume stderr from target VM, use -v EXTRA to copy to console
Iteration   1: 2387167.241 ±(99.9%) 153517.324 us/op
# Processing profiler results: perf 


Run result: 2387167.24 us/op (<= 2 samples)

Perf stats:
--------------------------------------------------

      35843.198075 task-clock                #   11.386 CPUs utilized          
         1,310,023 context-switches          #    0.037 M/sec                  
             2,767 CPU-migrations            #    0.000 M/sec                  
            14,126 page-faults               #    0.000 M/sec                  
    96,761,112,269 cycles                    #    2.700 GHz                     [25.66%]
    71,391,196,948 stalled-cycles-frontend   #   73.78% frontend cycles idle    [25.85%]
    56,586,750,391 stalled-cycles-backend    #   58.48% backend  cycles idle    [25.77%]
    58,668,717,159 instructions              #    0.61  insns per cycle        
                                             #    1.22  stalled cycles per insn [32.28%]
    11,568,806,175 branches                  #  322.762 M/sec                   [32.12%]
        78,827,679 branch-misses             #    0.68% of all branches         [32.18%]
    18,912,682,144 L1-dcache-loads           #  527.651 M/sec                   [32.21%]
       525,297,700 L1-dcache-load-misses     #    2.78% of all L1-dcache hits   [32.15%]
         5,370,104 LLC-loads                 #    0.150 M/sec                   [25.78%]
         4,861,731 LLC-load-misses           #   90.53% of all LL-cache hits    [25.63%]
   <not supported> L1-icache-loads         
       407,589,489 L1-icache-load-misses     #    0.00% of all L1-icache hits   [25.20%]
    19,511,345,970 dTLB-loads                #  544.353 M/sec                   [25.03%]
        30,539,251 dTLB-load-misses          #    0.16% of all dTLB cache hits  [24.47%]
        42,440,332 iTLB-loads                #    1.184 M/sec                   [24.60%]
         8,746,537 iTLB-load-misses          #   20.61% of all iTLB cache hits  [25.10%]
   <not supported> L1-dcache-prefetches    
        62,138,434 L1-dcache-prefetch-misses #    1.734 M/sec                   [25.26%]

       3.147968501 seconds time elapsed


WARNING: Your system uses old "perf", which can not delay data collection.
Therefore, perf performance data includes benchmark warmup.


# JMH 1.6.3 (released 5 days ago)
# VM invoker: /aurora/clredest/contended/jdk-cache64/jre/bin/java
# VM options: -XX:-UseBiasedLocking -XX:+UseHeavyMonitors
# Warmup: <none>
# Measurement: 1 iterations, 500 ms each, 4000000 calls per op
# Timeout: 10 min per iteration
# Threads: 32 threads
# Benchmark mode: Single shot invocation time
# Benchmark: org.openjdk.micros.contended.ContendedMonitorBench.synchronizeAndAccessOwnHeader
# Parameters: (inflate = false, monitors = 8, step = 8)

# Run progress: 33.33% complete, ETA 00:00:06
# Fork: 1 of 1
# Preparing profilers: perf 
# Profilers consume stderr from target VM, use -v EXTRA to copy to console
Iteration   1: 2361877.927 ±(99.9%) 162831.352 us/op
# Processing profiler results: perf 


Run result: 2361877.93 us/op (<= 2 samples)

Perf stats:
--------------------------------------------------

      37135.144090 task-clock                #   12.441 CPUs utilized          
         1,186,197 context-switches          #    0.032 M/sec                  
             5,229 CPU-migrations            #    0.000 M/sec                  
            14,013 page-faults               #    0.000 M/sec                  
   100,624,391,684 cycles                    #    2.710 GHz                     [25.68%]
    75,391,018,854 stalled-cycles-frontend   #   74.92% frontend cycles idle    [25.74%]
    59,451,562,882 stalled-cycles-backend    #   59.08% backend  cycles idle    [25.46%]
    59,412,217,830 instructions              #    0.59  insns per cycle        
                                             #    1.27  stalled cycles per insn [31.64%]
    11,929,948,669 branches                  #  321.258 M/sec                   [31.50%]
        79,294,224 branch-misses             #    0.66% of all branches         [31.51%]
    19,540,476,252 L1-dcache-loads           #  526.199 M/sec                   [31.60%]
       525,727,005 L1-dcache-load-misses     #    2.69% of all L1-dcache hits   [31.84%]
         6,789,167 LLC-loads                 #    0.183 M/sec                   [25.39%]
         6,187,867 LLC-load-misses           #   91.14% of all LL-cache hits    [25.64%]
   <not supported> L1-icache-loads         
       378,453,307 L1-icache-load-misses     #    0.00% of all L1-icache hits   [25.49%]
    19,415,651,490 dTLB-loads                #  522.838 M/sec                   [25.42%]
        24,459,129 dTLB-load-misses          #    0.13% of all dTLB cache hits  [25.31%]
        43,343,167 iTLB-loads                #    1.167 M/sec                   [24.99%]
        10,300,115 iTLB-load-misses          #   23.76% of all iTLB cache hits  [25.24%]
   <not supported> L1-dcache-prefetches    
        55,573,689 L1-dcache-prefetch-misses #    1.497 M/sec                   [25.38%]

       2.984814188 seconds time elapsed


WARNING: Your system uses old "perf", which can not delay data collection.
Therefore, perf performance data includes benchmark warmup.


# JMH 1.6.3 (released 5 days ago)
# VM invoker: /aurora/clredest/contended/jdk-cache64/jre/bin/java
# VM options: -XX:-UseBiasedLocking -XX:+UseHeavyMonitors
# Warmup: <none>
# Measurement: 1 iterations, 500 ms each, 4000000 calls per op
# Timeout: 10 min per iteration
# Threads: 32 threads
# Benchmark mode: Single shot invocation time
# Benchmark: org.openjdk.micros.contended.ContendedMonitorBench.synchronizeAndAccessUncontendedHeader
# Parameters: (inflate = false, monitors = 8, step = 8)

# Run progress: 66.67% complete, ETA 00:00:03
# Fork: 1 of 1
# Preparing profilers: perf 
# Profilers consume stderr from target VM, use -v EXTRA to copy to console
Iteration   1: 1206646.505 ±(99.9%) 137258.445 us/op
# Processing profiler results: perf 


Run result: 1206646.51 us/op (<= 2 samples)

Perf stats:
--------------------------------------------------

      25260.103374 task-clock                #   14.370 CPUs utilized          
           623,661 context-switches          #    0.025 M/sec                  
             3,031 CPU-migrations            #    0.000 M/sec                  
            14,091 page-faults               #    0.001 M/sec                  
    68,404,639,175 cycles                    #    2.708 GHz                     [25.66%]
    56,194,455,899 stalled-cycles-frontend   #   82.15% frontend cycles idle    [25.43%]
    47,056,256,110 stalled-cycles-backend    #   68.79% backend  cycles idle    [25.58%]
    26,211,665,002 instructions              #    0.38  insns per cycle        
                                             #    2.14  stalled cycles per insn [32.20%]
     6,054,661,669 branches                  #  239.693 M/sec                   [32.11%]
        70,870,988 branch-misses             #    1.17% of all branches         [32.25%]
     8,151,194,805 L1-dcache-loads           #  322.690 M/sec                   [32.20%]
       326,879,711 L1-dcache-load-misses     #    4.01% of all L1-dcache hits   [31.87%]
         4,400,494 LLC-loads                 #    0.174 M/sec                   [25.47%]
         3,236,307 LLC-load-misses           #   73.54% of all LL-cache hits    [25.91%]
   <not supported> L1-icache-loads         
       231,585,868 L1-icache-load-misses     #    0.00% of all L1-icache hits   [25.97%]
     8,165,470,910 dTLB-loads                #  323.256 M/sec                   [25.87%]
        16,716,439 dTLB-load-misses          #    0.20% of all dTLB cache hits  [25.43%]
        36,582,319 iTLB-loads                #    1.448 M/sec                   [25.28%]
         4,170,048 iTLB-load-misses          #   11.40% of all iTLB cache hits  [25.34%]
   <not supported> L1-dcache-prefetches    
        45,212,751 L1-dcache-prefetch-misses #    1.790 M/sec                   [25.29%]

       1.757890482 seconds time elapsed


WARNING: Your system uses old "perf", which can not delay data collection.
Therefore, perf performance data includes benchmark warmup.


# Run complete. Total time: 00:00:07

Benchmark                                                         (inflate)  (monitors)  (step)  Mode  Cnt        Score   Error  Units
ContendedMonitorBench.synchronizeAndAccessContendedHeader             false           8       8    ss    1  2387167.241 ±   NaN  us/op
ContendedMonitorBench.synchronizeAndAccessContendedHeader:@cpi        false           8       8    ss    1        1.649 ±   NaN    CPI
ContendedMonitorBench.synchronizeAndAccessOwnHeader                   false           8       8    ss    1  2361877.927 ±   NaN  us/op
ContendedMonitorBench.synchronizeAndAccessOwnHeader:@cpi              false           8       8    ss    1        1.694 ±   NaN    CPI
ContendedMonitorBench.synchronizeAndAccessUncontendedHeader           false           8       8    ss    1  1206646.505 ±   NaN  us/op
ContendedMonitorBench.synchronizeAndAccessUncontendedHeader:@cpi      false           8       8    ss    1        2.610 ±   NaN    CPI