GC.AllocateUninitializedArray has regressed on macOS #65198
Description
GC.AllocateUninitializedArray benchmarks have regressed by 15% for "smaller size" (1000 elements) only on macOS (x64, I don't have arm64 data). Other Unixes are not affected and larger sizes (10000 elements) are not affected..
Details
.NET 6:
| Type | Method | length | pinned | Mean | Error | StdDev | Median | Min | Max | Gen 0 | Gen 1 | Gen 2 | Allocated |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Perf_GC<Byte> | AllocateUninitializedArray | 1000 | False | 99.14 ns | 2.948 ns | 3.276 ns | 97.28 ns | 96.46 ns | 108.0 ns | 0.4891 | - | - | 1 KB |
| Perf_GC<Char> | AllocateUninitializedArray | 1000 | False | 192.16 ns | 0.779 ns | 0.729 ns | 191.89 ns | 191.14 ns | 193.4 ns | 0.9669 | - | - | 1.98 KB |
| Perf_GC<Byte> | AllocateUninitializedArray | 1000 | True | 378.25 ns | 4.336 ns | 3.843 ns | 376.98 ns | 373.63 ns | 385.4 ns | 1.9396 | 1.9396 | 0.3246 | 1 KB |
| Perf_GC<Char> | AllocateUninitializedArray | 1000 | True | 527.51 ns | 35.474 ns | 34.840 ns | 537.93 ns | 440.44 ns | 558.6 ns | 3.8341 | 3.8341 | 0.6409 | 1.98 KB |
| Perf_GC<Byte> | AllocateUninitializedArray | 10000 | False | 460.30 ns | 13.229 ns | 12.993 ns | 465.62 ns | 433.22 ns | 472.1 ns | 4.7617 | - | - | 9.79 KB |
| Perf_GC<Char> | AllocateUninitializedArray | 10000 | False | 863.26 ns | 28.192 ns | 31.336 ns | 868.93 ns | 799.42 ns | 905.1 ns | 9.5207 | - | - | 19.55 KB |
| Perf_GC<Byte> | AllocateUninitializedArray | 10000 | True | 2,079.43 ns | 193.755 ns | 215.358 ns | 2,184.01 ns | 1,616.18 ns | 2,218.1 ns | 18.7916 | 18.7916 | 3.1319 | 9.79 KB |
| Perf_GC<Char> | AllocateUninitializedArray | 10000 | True | 4,294.48 ns | 339.558 ns | 391.035 ns | 4,411.58 ns | 3,194.01 ns | 4,566.0 ns | 36.7647 | 36.7647 | 6.1275 | 19.56 KB |
.NET 7 Preview 1:
| Type | Method | length | pinned | Mean | Error | StdDev | Median | Min | Max | Gen 0 | Gen 1 | Gen 2 | Allocated |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Perf_GC<Byte> | AllocateUninitializedArray | 1000 | False | 115.0 ns | 3.62 ns | 4.02 ns | 115.7 ns | 107.7 ns | 120.7 ns | 0.4894 | - | - | 1 KB |
| Perf_GC<Char> | AllocateUninitializedArray | 1000 | False | 218.5 ns | 10.22 ns | 11.76 ns | 221.6 ns | 197.3 ns | 238.9 ns | 0.9663 | - | - | 1.98 KB |
| Perf_GC<Byte> | AllocateUninitializedArray | 1000 | True | 408.4 ns | 44.27 ns | 49.21 ns | 403.0 ns | 314.0 ns | 498.6 ns | 1.9366 | 1.9366 | 0.3239 | 1 KB |
| Perf_GC<Char> | AllocateUninitializedArray | 1000 | True | 622.0 ns | 7.05 ns | 6.25 ns | 621.5 ns | 607.3 ns | 630.9 ns | 3.8293 | 3.8293 | 0.6402 | 1.98 KB |
| Perf_GC<Byte> | AllocateUninitializedArray | 10000 | False | 502.1 ns | 36.32 ns | 40.37 ns | 500.5 ns | 441.9 ns | 573.4 ns | 4.7603 | - | - | 9.79 KB |
| Perf_GC<Char> | AllocateUninitializedArray | 10000 | False | 834.6 ns | 24.70 ns | 25.36 ns | 844.3 ns | 785.4 ns | 860.9 ns | 9.5210 | - | - | 19.55 KB |
| Perf_GC<Byte> | AllocateUninitializedArray | 10000 | True | 2,328.8 ns | 272.49 ns | 313.79 ns | 2,413.5 ns | 1,747.3 ns | 2,679.2 ns | 18.7466 | 18.7466 | 3.1301 | 9.79 KB |
| Perf_GC<Char> | AllocateUninitializedArray | 10000 | True | 4,128.7 ns | 395.66 ns | 439.78 ns | 4,314.0 ns | 3,130.7 ns | 4,385.4 ns | 36.7561 | 36.7561 | 6.1260 | 19.56 KB |
Repro:
git clone https://github.com/dotnet/performance.git
python3 ./performance/scripts/benchmarks_ci.py -f net6.0 net7.0 --filter '*AllocateUninitializedArray' --bdn-arguments "--join true"Details
System.Tests.Perf_GC.AllocateArray(length: 10000, pinned: False)
| Result | Base | Diff | Ratio | Alloc Delta | Modality | Operating System | Bit | Processor Name | Base V | Diff V |
|---|---|---|---|---|---|---|---|---|---|---|
| Same | 248.92 | 243.83 | 1.02 | +0 | Windows 11 | X64 | AMD Ryzen Threadripper PRO 3945WX 12-Cores | 6.0.121.56705 | 7.0.22.7608 | |
| Same | 179.83 | 184.15 | 0.98 | +0 | Windows 11 | X64 | AMD Ryzen 9 5900X | 6.0.121.56705 | 7.0.22.7608 | |
| Same | 459.82 | 461.10 | 1.00 | +0 | Windows 10 | X64 | Intel Xeon CPU E5-1650 v4 3.60GHz | 6.0.121.56705 | 7.0.22.7608 | |
| Faster | 914.81 | 709.96 | 1.29 | +0 | Windows 11 | X64 | Intel Core i5-4300U CPU 1.90GHz (Haswell) | 6.0.121.56705 | 7.0.22.7608 | |
| Same | 312.96 | 304.02 | 1.03 | +0 | bimodal | Windows 10 | X64 | Intel Core i7-7700 CPU 3.60GHz (Kaby Lake) | 6.0.121.56705 | 7.0.22.7608 |
| Same | 298.12 | 327.61 | 0.91 | +0 | Windows 11 | X64 | Intel Core i7-8700 CPU 3.20GHz (Coffee Lake) | 6.0.121.56705 | 7.0.22.7608 | |
| Same | 307.02 | 307.68 | 1.00 | +0 | several? | Windows 11 | X64 | Intel Core i9-9900T CPU 2.10GHz | 6.0.121.56705 | 7.0.22.7608 |
| Same | 584.23 | 542.31 | 1.08 | +0 | several? | Windows 11 | X64 | Unknown processor | 6.0.121.56705 | 7.0.22.7608 |
| Same | 420.18 | 431.82 | 0.97 | +0 | Windows 11 | X64 | Unknown processor | 6.0.121.56705 | 7.0.22.7608 | |
| Same | 301.75 | 307.80 | 0.98 | +0 | ubuntu 20.04 | X64 | AMD Ryzen 9 5900X | 6.0.121.56705 | 7.0.22.7608 | |
| Same | 483.25 | 487.37 | 0.99 | +0 | ubuntu 18.04 | X64 | Intel Xeon CPU E5-1650 v4 3.60GHz | 6.0.121.56705 | 7.0.22.7608 | |
| Same | 1248.19 | 1157.88 | 1.08 | +0 | centos 7 | X64 | Intel Xeon CPU E5530 2.40GHz | 6.0.121.56705 | 7.0.22.7608 | |
| Same | 807.84 | 747.88 | 1.08 | +0 | ubuntu 18.04 | X64 | Intel Core i7-2720QM CPU 2.20GHz (Sandy Bridge) | 6.0.121.56705 | 7.0.22.7608 | |
| Same | 372.58 | 389.82 | 0.96 | +0 | alpine 3.13 | X64 | Intel Core i7-7700 CPU 3.60GHz (Kaby Lake) | 6.0.121.56705 | 7.0.22.7608 | |
| Same | 391.98 | 388.72 | 1.01 | +0 | ubuntu 18.04 | X64 | Intel Core i7-7700 CPU 3.60GHz (Kaby Lake) | 6.0.121.56705 | 7.0.22.7608 | |
| Same | 365.02 | 365.98 | 1.00 | +0 | ubuntu 20.04 | X64 | Intel Core i7-7700 CPU 3.60GHz (Kaby Lake) | 6.0.121.56705 | 7.0.22.7608 | |
| Same | 610.43 | 598.45 | 1.02 | +0 | Windows 10 | Arm64 | Microsoft SQ1 3.0 GHz | 6.0.121.56705 | 7.0.22.7608 | |
| Same | 269.72 | 275.66 | 0.98 | +0 | Windows 11 | X86 | AMD Ryzen Threadripper PRO 3945WX 12-Cores | 6.0.121.56705 | 7.0.22.7608 | |
| Same | 489.28 | 492.36 | 0.99 | +0 | Windows 10 | X86 | Intel Xeon CPU E5-1650 v4 3.60GHz | 6.0.121.56705 | 7.0.22.7608 | |
| Same | 827.06 | 821.72 | 1.01 | +0 | Windows 10 | Arm | Microsoft SQ1 3.0 GHz | 6.0.121.56705 | 7.0.22.7608 | |
| Slower | 853.39 | 1028.78 | 0.83 | +0 | macOS Big Sur 11.6.3 | X64 | Intel Core i5-4278U CPU 2.60GHz (Haswell) | 6.0.121.56705 | 7.0.22.7608 | |
| Slower | 734.86 | 859.05 | 0.86 | +0 | macOS Big Sur 11.4 | X64 | Intel Core i7-5557U CPU 3.10GHz (Broadwell) | 6.0.121.56705 | 7.0.22.7608 |
System.Tests.Perf_GC.AllocateUninitializedArray(length: 1000, pinned: False)
| Result | Base | Diff | Ratio | Alloc Delta | Modality | Operating System | Bit | Processor Name | Base V | Diff V |
|---|---|---|---|---|---|---|---|---|---|---|
| Same | 28.63 | 28.43 | 1.01 | +0 | Windows 11 | X64 | AMD Ryzen Threadripper PRO 3945WX 12-Cores | 6.0.121.56705 | 7.0.22.7608 | |
| Same | 19.49 | 20.59 | 0.95 | +0 | Windows 11 | X64 | AMD Ryzen 9 5900X | 6.0.121.56705 | 7.0.22.7608 | |
| Same | 47.97 | 48.42 | 0.99 | +0 | Windows 10 | X64 | Intel Xeon CPU E5-1650 v4 3.60GHz | 6.0.121.56705 | 7.0.22.7608 | |
| Faster | 96.60 | 73.23 | 1.32 | +0 | Windows 11 | X64 | Intel Core i5-4300U CPU 1.90GHz (Haswell) | 6.0.121.56705 | 7.0.22.7608 | |
| Same | 32.77 | 32.21 | 1.02 | +0 | several? | Windows 10 | X64 | Intel Core i7-7700 CPU 3.60GHz (Kaby Lake) | 6.0.121.56705 | 7.0.22.7608 |
| Same | 30.69 | 32.90 | 0.93 | +0 | Windows 11 | X64 | Intel Core i7-8700 CPU 3.20GHz (Coffee Lake) | 6.0.121.56705 | 7.0.22.7608 | |
| Same | 31.55 | 32.59 | 0.97 | +0 | several? | Windows 11 | X64 | Intel Core i9-9900T CPU 2.10GHz | 6.0.121.56705 | 7.0.22.7608 |
| Same | 59.60 | 62.74 | 0.95 | +0 | Windows 11 | X64 | Unknown processor | 6.0.121.56705 | 7.0.22.7608 | |
| Same | 42.11 | 46.20 | 0.91 | +0 | several? | Windows 11 | X64 | Unknown processor | 6.0.121.56705 | 7.0.22.7608 |
| Same | 37.46 | 38.63 | 0.97 | +0 | ubuntu 20.04 | X64 | AMD Ryzen 9 5900X | 6.0.121.56705 | 7.0.22.7608 | |
| Same | 53.49 | 54.05 | 0.99 | +0 | ubuntu 18.04 | X64 | Intel Xeon CPU E5-1650 v4 3.60GHz | 6.0.121.56705 | 7.0.22.7608 | |
| Same | 148.24 | 138.95 | 1.07 | +0 | centos 7 | X64 | Intel Xeon CPU E5530 2.40GHz | 6.0.121.56705 | 7.0.22.7608 | |
| Same | 87.90 | 85.15 | 1.03 | +0 | ubuntu 18.04 | X64 | Intel Core i7-2720QM CPU 2.20GHz (Sandy Bridge) | 6.0.121.56705 | 7.0.22.7608 | |
| Same | 43.12 | 43.76 | 0.99 | +0 | alpine 3.13 | X64 | Intel Core i7-7700 CPU 3.60GHz (Kaby Lake) | 6.0.121.56705 | 7.0.22.7608 | |
| Same | 50.21 | 48.88 | 1.03 | +0 | ubuntu 18.04 | X64 | Intel Core i7-7700 CPU 3.60GHz (Kaby Lake) | 6.0.121.56705 | 7.0.22.7608 | |
| Same | 46.33 | 46.50 | 1.00 | +0 | multimodal | ubuntu 20.04 | X64 | Intel Core i7-7700 CPU 3.60GHz (Kaby Lake) | 6.0.121.56705 | 7.0.22.7608 |
| Same | 64.90 | 65.37 | 0.99 | +0 | Windows 10 | Arm64 | Microsoft SQ1 3.0 GHz | 6.0.121.56705 | 7.0.22.7608 | |
| Same | 29.11 | 30.75 | 0.95 | +0 | Windows 11 | X86 | AMD Ryzen Threadripper PRO 3945WX 12-Cores | 6.0.121.56705 | 7.0.22.7608 | |
| Same | 50.59 | 52.71 | 0.96 | +0 | Windows 10 | X86 | Intel Xeon CPU E5-1650 v4 3.60GHz | 6.0.121.56705 | 7.0.22.7608 | |
| Same | 94.16 | 96.78 | 0.97 | +0 | Windows 10 | Arm | Microsoft SQ1 3.0 GHz | 6.0.121.56705 | 7.0.22.7608 | |
| Slower | 92.65 | 109.98 | 0.84 | +0 | macOS Big Sur 11.6.3 | X64 | Intel Core i5-4278U CPU 2.60GHz (Haswell) | 6.0.121.56705 | 7.0.22.7608 | |
| Slower | 77.08 | 91.51 | 0.84 | +0 | macOS Big Sur 11.4 | X64 | Intel Core i7-5557U CPU 3.10GHz (Broadwell) | 6.0.121.56705 | 7.0.22.7608 |
cc @VSadov
Metadata
Metadata
Type
Projects
Status