[Perf] Linux/x64: 71 Regressions on 7/2/2024 7:47:18 PM

[performanceautofiler]

Run Information

Name	Value
Architecture	x64
OS	ubuntu 22.04
Queue	TigerUbuntu
Baseline	b59816c21a6138174bf84fcc5b88a6ece8e1e41c
Compare	1d16fd14fa772621c7b09ebd00f83610207b1abf
Diff	Diff
Configs	CompilationMode:wasm, RunKind:micro

Regressions in System.Buffers.Tests.SearchValuesCharTests

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio
LastIndexOfAny - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	725.42 ns	807.08 ns	1.11	0.08	False
ContainsAny - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	138.16 ns	268.08 ns	1.94	0.14	True
IndexOfAnyExcept - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	156.34 ns	319.61 ns	2.04	0.17	True
LastIndexOfAnyExcept - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	166.35 ns	263.59 ns	1.58	0.20	True
LastIndexOfAnyExcept - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	138.02 ns	252.80 ns	1.83	0.21	True
IndexOfAny - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	131.17 ns	275.82 ns	2.10	0.20	True
ContainsAny - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	837.22 ns	925.71 ns	1.11	0.11	False
IndexOfAnyExcept - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	147.03 ns	277.80 ns	1.89	0.15	True

Test Report

Repro

General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.Buffers.Tests.SearchValuesCharTests*'

System.Buffers.Tests.SearchValuesCharTests.LastIndexOfAny(Values: "abcdefABCDEF0123456789Ü")

ETL Files

Histogram

JIT Disasms

System.Buffers.Tests.SearchValuesCharTests.ContainsAny(Values: "abcdefABCDEF0123456789")

ETL Files

Histogram

JIT Disasms

System.Buffers.Tests.SearchValuesCharTests.IndexOfAnyExcept(Values: "abcdefABCDEF0123456789Ü")

ETL Files

Histogram

JIT Disasms

System.Buffers.Tests.SearchValuesCharTests.LastIndexOfAnyExcept(Values: "abcdefABCDEF0123456789Ü")

ETL Files

Histogram

JIT Disasms

System.Buffers.Tests.SearchValuesCharTests.LastIndexOfAnyExcept(Values: "abcdefABCDEF0123456789")

ETL Files

Histogram

JIT Disasms

System.Buffers.Tests.SearchValuesCharTests.IndexOfAny(Values: "abcdefABCDEF0123456789")

ETL Files

Histogram

JIT Disasms

System.Buffers.Tests.SearchValuesCharTests.ContainsAny(Values: "abcdefABCDEF0123456789Ü")

ETL Files

Histogram

JIT Disasms

System.Buffers.Tests.SearchValuesCharTests.IndexOfAnyExcept(Values: "abcdefABCDEF0123456789")

ETL Files

Histogram

JIT Disasms

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository

---

Run Information

Name	Value
Architecture	x64
OS	ubuntu 22.04
Queue	TigerUbuntu
Baseline	b59816c21a6138174bf84fcc5b88a6ece8e1e41c
Compare	1fe7d189db4a49bc676ddb206456709e089c2293
Diff	Diff
Configs	CompilationMode:wasm, RunKind:micro

Regressions in System.Text.Json.Serialization.Tests.ReadJson<BinaryData>

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio
DeserializeFromString - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	7.82 μs	8.64 μs	1.10	0.06	False

Test Report

Repro

General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.Text.Json.Serialization.Tests.ReadJson<BinaryData>*'

System.Text.Json.Serialization.Tests.ReadJson<BinaryData>.DeserializeFromString(Mode: Reflection)

ETL Files

Histogram

JIT Disasms

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository

---

Run Information

Name	Value
Architecture	x64
OS	ubuntu 22.04
Queue	TigerUbuntu
Baseline	c07a81f90938a8c3ec5b056b78ddd1718dc9931c
Compare	2ea6ae57874c452923af059cbcb57d109564353c
Diff	Diff
Configs	CompilationMode:wasm, RunKind:micro

Regressions in System.Text.Perf_Ascii

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio
ToLower_Chars - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	219.56 ns	456.69 ns	2.08	0.17	True
EqualsIgnoreCase_ExactlyTheSame_Bytes_Chars - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	684.59 ns	723.76 ns	1.06	0.03	False
ToUpper_Chars - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	217.83 ns	458.69 ns	2.11	0.21	True
EqualsIgnoreCase_ExactlyTheSame_Chars - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	152.54 ns	219.22 ns	1.44	0.12	True
ToLowerInPlace_Bytes - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	112.60 ns	224.91 ns	2.00	0.14	True
ToLowerInPlace_Chars - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	178.62 ns	393.88 ns	2.21	0.10	True
IsValid_Bytes - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	46.45 ns	52.29 ns	1.13	0.21	False
ToLower_Bytes_Chars - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	180.32 ns	337.35 ns	1.87	0.19	True
Equals_Bytes - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	88.60 ns	109.05 ns	1.23	0.15	False
FromUtf16 - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	118.22 ns	179.64 ns	1.52	0.19	False
ToUpperInPlace_Chars - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	173.87 ns	403.13 ns	2.32	0.13	True
EqualsIgnoreCase_ExactlyTheSame_Bytes - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	91.84 ns	117.59 ns	1.28	0.22	False
EqualsIgnoreCase_DifferentCase_Bytes - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	350.86 ns	456.87 ns	1.30	0.07	False
Equals_Chars - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	145.98 ns	185.94 ns	1.27	0.15	False
ToLower_Bytes - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	150.55 ns	274.61 ns	1.82	0.17	False
EqualsIgnoreCase_DifferentCase_Chars - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	451.97 ns	515.99 ns	1.14	0.06	False

Test Report

Repro

General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.Text.Perf_Ascii*'

System.Text.Perf_Ascii.ToLower_Chars(Size: 128)

ETL Files

Histogram

JIT Disasms

System.Text.Perf_Ascii.EqualsIgnoreCase_ExactlyTheSame_Bytes_Chars(Size: 128)

ETL Files

Histogram

JIT Disasms

System.Text.Perf_Ascii.ToUpper_Chars(Size: 128)

ETL Files

Histogram

JIT Disasms

System.Text.Perf_Ascii.EqualsIgnoreCase_ExactlyTheSame_Chars(Size: 128)

ETL Files

Histogram

JIT Disasms

System.Text.Perf_Ascii.ToLowerInPlace_Bytes(Size: 128)

ETL Files

Histogram

JIT Disasms

System.Text.Perf_Ascii.ToLowerInPlace_Chars(Size: 128)

ETL Files

Histogram

JIT Disasms

System.Text.Perf_Ascii.IsValid_Bytes(Size: 128)

ETL Files

Histogram

JIT Disasms

System.Text.Perf_Ascii.ToLower_Bytes_Chars(Size: 128)

ETL Files

Histogram

JIT Disasms

System.Text.Perf_Ascii.Equals_Bytes(Size: 128)

ETL Files

Histogram

JIT Disasms

System.Text.Perf_Ascii.FromUtf16(Size: 128)

ETL Files

Histogram

JIT Disasms

System.Text.Perf_Ascii.ToUpperInPlace_Chars(Size: 128)

ETL Files

Histogram

JIT Disasms

System.Text.Perf_Ascii.EqualsIgnoreCase_ExactlyTheSame_Bytes(Size: 128)

ETL Files

Histogram

JIT Disasms

System.Text.Perf_Ascii.EqualsIgnoreCase_DifferentCase_Bytes(Size: 128)

ETL Files

Histogram

JIT Disasms

System.Text.Perf_Ascii.Equals_Chars(Size: 128)

ETL Files

Histogram

JIT Disasms

System.Text.Perf_Ascii.ToLower_Bytes(Size: 128)

ETL Files

Histogram

JIT Disasms

System.Text.Perf_Ascii.EqualsIgnoreCase_DifferentCase_Chars(Size: 128)

ETL Files

Histogram

JIT Disasms

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository

---

Run Information

Name	Value
Architecture	x64
OS	ubuntu 22.04
Queue	TigerUbuntu
Baseline	c07a81f90938a8c3ec5b056b78ddd1718dc9931c
Compare	2ea6ae57874c452923af059cbcb57d109564353c
Diff	Diff
Configs	CompilationMode:wasm, RunKind:micro

Regressions in System.Numerics.Tensors.Tests.Perf_BinaryIntegerTensorPrimitives<Byte>

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio
TrailingZeroCount - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	798.08 ns	1.85 μs	2.32	0.09	True
PopCount - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	136.70 ns	189.95 ns	1.39	0.24	False
PopCount - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	511.20 ns	1.55 μs	3.03	0.14	True
TrailingZeroCount - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	156.13 ns	218.94 ns	1.40	0.25	False

Test Report

Repro

General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.Numerics.Tensors.Tests.Perf_BinaryIntegerTensorPrimitives<Byte>*'

System.Numerics.Tensors.Tests.Perf_BinaryIntegerTensorPrimitives<Byte>.TrailingZeroCount(BufferLength: 3079)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_BinaryIntegerTensorPrimitives<Byte>.PopCount(BufferLength: 128)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_BinaryIntegerTensorPrimitives<Byte>.PopCount(BufferLength: 3079)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_BinaryIntegerTensorPrimitives<Byte>.TrailingZeroCount(BufferLength: 128)

ETL Files

Histogram

JIT Disasms

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository

---

Run Information

Name	Value
Architecture	x64
OS	ubuntu 22.04
Queue	TigerUbuntu
Baseline	b59816c21a6138174bf84fcc5b88a6ece8e1e41c
Compare	2ea6ae57874c452923af059cbcb57d109564353c
Diff	Diff
Configs	CompilationMode:wasm, RunKind:micro

Regressions in System.Text.Tests.Perf_Encoding

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio
GetBytes - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	487.12 ns	718.18 ns	1.47	0.07	True
GetBytes - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	896.02 ns	1.15 μs	1.28	0.06	False

Test Report

Repro

General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.Text.Tests.Perf_Encoding*'

System.Text.Tests.Perf_Encoding.GetBytes(size: 512, encName: "ascii")

ETL Files

Histogram

JIT Disasms

System.Text.Tests.Perf_Encoding.GetBytes(size: 512, encName: "utf-8")

ETL Files

Histogram

JIT Disasms

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository

[performanceautofiler]

Run Information

Name	Value
Architecture	x64
OS	ubuntu 22.04
Queue	TigerUbuntu
Baseline	c07a81f90938a8c3ec5b056b78ddd1718dc9931c
Compare	1d16fd14fa772621c7b09ebd00f83610207b1abf
Diff	Diff
Configs	CompilationMode:wasm, RunKind:micro

Regressions in System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Single>

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio
Sin - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	13.36 μs	20.03 μs	1.50	0.10	False
FusedMultiplyAdd_ScalarAddend - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	3.89 μs	4.28 μs	1.10	0.16	False
Truncate - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	4.57 μs	4.87 μs	1.07	0.18	False
Sin - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	304.49 μs	478.38 μs	1.57	0.09	False
FusedMultiplyAdd_Vectors - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	3.96 μs	4.38 μs	1.10	0.15	False
FusedMultiplyAdd_ScalarAddend - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	86.76 μs	97.45 μs	1.12	0.13	False
FusedMultiplyAdd_ScalarMultiplier - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	3.92 μs	4.35 μs	1.11	0.21	False
FusedMultiplyAdd_Vectors - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	90.85 μs	98.39 μs	1.08	0.15	False

Test Report

Repro

General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Single>*'

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Single>.Sin(BufferLength: 128)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Single>.FusedMultiplyAdd_ScalarAddend(BufferLength: 128)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Single>.Truncate(BufferLength: 128)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Single>.Sin(BufferLength: 3079)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Single>.FusedMultiplyAdd_Vectors(BufferLength: 128)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Single>.FusedMultiplyAdd_ScalarAddend(BufferLength: 3079)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Single>.FusedMultiplyAdd_ScalarMultiplier(BufferLength: 128)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Single>.FusedMultiplyAdd_Vectors(BufferLength: 3079)

ETL Files

Histogram

JIT Disasms

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository

---

Run Information

Name	Value
Architecture	x64
OS	ubuntu 22.04
Queue	TigerUbuntu
Baseline	c07a81f90938a8c3ec5b056b78ddd1718dc9931c
Compare	1fe7d189db4a49bc676ddb206456709e089c2293
Diff	Diff
Configs	CompilationMode:wasm, RunKind:micro

Regressions in System.Numerics.Tensors.Tests.Perf_NumberTensorPrimitives<Single>

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio
Max_Vector - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	90.64 μs	105.35 μs	1.16	0.14	False
Max_Vector - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	3.59 μs	4.16 μs	1.16	0.13	False
Max_Scalar - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	3.86 μs	4.47 μs	1.16	0.13	False

Test Report

Repro

General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.Numerics.Tensors.Tests.Perf_NumberTensorPrimitives<Single>*'

System.Numerics.Tensors.Tests.Perf_NumberTensorPrimitives<Single>.Max_Vector(BufferLength: 3079)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_NumberTensorPrimitives<Single>.Max_Vector(BufferLength: 128)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_NumberTensorPrimitives<Single>.Max_Scalar(BufferLength: 128)

ETL Files

Histogram

JIT Disasms

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository

---

Run Information

Name	Value
Architecture	x64
OS	ubuntu 22.04
Queue	TigerUbuntu
Baseline	c07a81f90938a8c3ec5b056b78ddd1718dc9931c
Compare	1fe7d189db4a49bc676ddb206456709e089c2293
Diff	Diff
Configs	CompilationMode:wasm, RunKind:micro

Regressions in System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Double>

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio
CosineSimilarity - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	3.58 μs	6.77 μs	1.89	0.11	False
FusedMultiplyAdd_Vectors - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	206.87 μs	230.29 μs	1.11	0.21	False
Truncate - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	143.77 μs	176.83 μs	1.23	0.17	False
FusedMultiplyAdd_ScalarMultiplier - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	211.13 μs	226.94 μs	1.07	0.17	False
Sin - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	1.14 ms	1.32 ms	1.16	0.20	False
FusedMultiplyAdd_ScalarMultiplier - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	8.77 μs	9.84 μs	1.12	0.24	False
Truncate - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	6.21 μs	7.17 μs	1.16	0.17	False
Sin - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	49.16 μs	56.53 μs	1.15	0.17	False

Test Report

Repro

General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Double>*'

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Double>.CosineSimilarity(BufferLength: 128)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Double>.FusedMultiplyAdd_Vectors(BufferLength: 3079)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Double>.Truncate(BufferLength: 3079)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Double>.FusedMultiplyAdd_ScalarMultiplier(BufferLength: 3079)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Double>.Sin(BufferLength: 3079)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Double>.FusedMultiplyAdd_ScalarMultiplier(BufferLength: 128)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Double>.Truncate(BufferLength: 128)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Double>.Sin(BufferLength: 128)

ETL Files

Histogram

JIT Disasms

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository

---

Run Information

Name	Value
Architecture	x64
OS	ubuntu 22.04
Queue	TigerUbuntu
Baseline	c07a81f90938a8c3ec5b056b78ddd1718dc9931c
Compare	1fe7d189db4a49bc676ddb206456709e089c2293
Diff	Diff
Configs	CompilationMode:wasm, RunKind:micro

Regressions in System.Text.Perf_Utf8Encoding

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio
GetString - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	645.81 μs	683.89 μs	1.06	0.00	False
GetBytes - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	182.48 μs	246.25 μs	1.35	0.01	True
GetString - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	504.39 μs	733.11 μs	1.45	0.10	True

Test Report

Repro

General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.Text.Perf_Utf8Encoding*'

System.Text.Perf_Utf8Encoding.GetString(Input: Greek)

ETL Files

Histogram

JIT Disasms

System.Text.Perf_Utf8Encoding.GetBytes(Input: EnglishAllAscii)

ETL Files

Histogram

JIT Disasms

System.Text.Perf_Utf8Encoding.GetString(Input: EnglishMostlyAscii)

ETL Files

Histogram

JIT Disasms

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository

---

Run Information

Name	Value
Architecture	x64
OS	ubuntu 22.04
Queue	TigerUbuntu
Baseline	c07a81f90938a8c3ec5b056b78ddd1718dc9931c
Compare	1d16fd14fa772621c7b09ebd00f83610207b1abf
Diff	Diff
Configs	CompilationMode:wasm, RunKind:micro

Regressions in System.Numerics.Tensors.Tests.Perf_NumberTensorPrimitives<Int32>

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio
IndexOfMax - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	893.42 ns	967.69 ns	1.08	0.02	False

Test Report

Repro

General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.Numerics.Tensors.Tests.Perf_NumberTensorPrimitives<Int32>*'

System.Numerics.Tensors.Tests.Perf_NumberTensorPrimitives<Int32>.IndexOfMax(BufferLength: 128)

ETL Files

Histogram

JIT Disasms

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository

---

Run Information

Name	Value
Architecture	x64
OS	ubuntu 22.04
Queue	TigerUbuntu
Baseline	c07a81f90938a8c3ec5b056b78ddd1718dc9931c
Compare	1fe7d189db4a49bc676ddb206456709e089c2293
Diff	Diff
Configs	CompilationMode:wasm, RunKind:micro

Regressions in System.Buffers.Tests.SearchValuesByteTests

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio
IndexOfAnyExcept - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	128.91 ns	178.23 ns	1.38	0.16	False
IndexOfAny - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	121.22 ns	143.62 ns	1.18	0.16	False
IndexOfAnyExcept - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	631.78 ns	761.59 ns	1.21	0.08	False
ContainsAny - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	118.70 ns	140.84 ns	1.19	0.14	False
ContainsAny - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	615.06 ns	785.21 ns	1.28	0.07	False

Test Report

Repro

General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.Buffers.Tests.SearchValuesByteTests*'

System.Buffers.Tests.SearchValuesByteTests.IndexOfAnyExcept(Values: "abcdefABCDEF0123456789")

ETL Files

Histogram

JIT Disasms

System.Buffers.Tests.SearchValuesByteTests.IndexOfAny(Values: "abcdefABCDEF0123456789")

ETL Files

Histogram

JIT Disasms

System.Buffers.Tests.SearchValuesByteTests.IndexOfAnyExcept(Values: "abcdefABCDEF0123456789Ü")

ETL Files

Histogram

JIT Disasms

System.Buffers.Tests.SearchValuesByteTests.ContainsAny(Values: "abcdefABCDEF0123456789")

ETL Files

Histogram

JIT Disasms

System.Buffers.Tests.SearchValuesByteTests.ContainsAny(Values: "abcdefABCDEF0123456789Ü")

ETL Files

Histogram

JIT Disasms

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository

[performanceautofiler]

Run Information

Name	Value
Architecture	x64
OS	ubuntu 22.04
Queue	TigerUbuntu
Baseline	b59816c21a6138174bf84fcc5b88a6ece8e1e41c
Compare	1d16fd14fa772621c7b09ebd00f83610207b1abf
Diff	Diff
Configs	CompilationMode:wasm, RunKind:micro

Regressions in System.Globalization.Tests.StringSearch

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio
LastIndexOf_Word_NotFound - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	275.95 ns	543.38 ns	1.97	0.10	True
IndexOf_Word_NotFound - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	269.45 ns	371.70 ns	1.38	0.11	True

Test Report

Repro

General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.Globalization.Tests.StringSearch*'

System.Globalization.Tests.StringSearch.LastIndexOf_Word_NotFound(Options: (en-US, Ordinal, False))

ETL Files

Histogram

JIT Disasms

System.Globalization.Tests.StringSearch.IndexOf_Word_NotFound(Options: (en-US, Ordinal, False))

ETL Files

Histogram

JIT Disasms

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository

---

Run Information

Name	Value
Architecture	x64
OS	ubuntu 22.04
Queue	TigerUbuntu
Baseline	c07a81f90938a8c3ec5b056b78ddd1718dc9931c
Compare	1fe7d189db4a49bc676ddb206456709e089c2293
Diff	Diff
Configs	CompilationMode:wasm, RunKind:micro

Regressions in System.Buffers.Text.Tests.Base64Tests

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio
Base64IsValid - Duration of single invocation ADX - Test Multi Config Graph	535.87 ns	799.17 ns	1.49	0.05	True

Test Report

Repro

General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.Buffers.Text.Tests.Base64Tests*'

System.Buffers.Text.Tests.Base64Tests.Base64IsValid(NumberOfBytes: 1000)

ETL Files

Histogram

JIT Disasms

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository

---

Run Information

Name	Value
Architecture	x64
OS	ubuntu 22.04
Queue	TigerUbuntu
Baseline	c07a81f90938a8c3ec5b056b78ddd1718dc9931c
Compare	1fe7d189db4a49bc676ddb206456709e089c2293
Diff	Diff
Configs	CompilationMode:wasm, RunKind:micro

Regressions in System.Tests.Perf_Uri

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio
EscapeDataString - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	675.88 ns	1.47 μs	2.18	0.11	True

Test Report

Repro

General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.Tests.Perf_Uri*'

System.Tests.Perf_Uri.EscapeDataString(input: "aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa")

ETL Files

Histogram

JIT Disasms

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository

---

Run Information

Name	Value
Architecture	x64
OS	ubuntu 22.04
Queue	TigerUbuntu
Baseline	c07a81f90938a8c3ec5b056b78ddd1718dc9931c
Compare	1fe7d189db4a49bc676ddb206456709e089c2293
Diff	Diff
Configs	CompilationMode:wasm, RunKind:micro

Regressions in System.IO.Tests.Perf_File

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio
WriteAllText - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	41.72 μs	47.66 μs	1.14	0.27	False
AppendAllText - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	63.70 μs	69.16 μs	1.09	0.03	False

Test Report

Repro

General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.IO.Tests.Perf_File*'

System.IO.Tests.Perf_File.WriteAllText(size: 10000)

ETL Files

Histogram

JIT Disasms

System.IO.Tests.Perf_File.AppendAllText(size: 10000)

ETL Files

Histogram

JIT Disasms

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository

---

Run Information

Name	Value
Architecture	x64
OS	ubuntu 22.04
Queue	TigerUbuntu
Baseline	b59816c21a6138174bf84fcc5b88a6ece8e1e41c
Compare	1fe7d189db4a49bc676ddb206456709e089c2293
Diff	Diff
Configs	CompilationMode:wasm, RunKind:micro

Regressions in System.IO.Tests.BinaryWriterExtendedTests

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio
WriteAsciiString - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	3.73 μs	6.91 μs	1.85	0.08	True
WriteAsciiString - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	1.67 ms	2.53 ms	1.52	0.02	False
WriteAsciiCharArray - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	818.77 μs	1.64 ms	2.00	0.03	True

Test Report

Repro

General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.IO.Tests.BinaryWriterExtendedTests*'

System.IO.Tests.BinaryWriterExtendedTests.WriteAsciiString(StringLengthInChars: 8000)

ETL Files

Histogram

JIT Disasms

System.IO.Tests.BinaryWriterExtendedTests.WriteAsciiString(StringLengthInChars: 2000000)

ETL Files

Histogram

JIT Disasms

System.IO.Tests.BinaryWriterExtendedTests.WriteAsciiCharArray(StringLengthInChars: 2000000)

ETL Files

Histogram

JIT Disasms

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository

---

Run Information

Name	Value
Architecture	x64
OS	ubuntu 22.04
Queue	TigerUbuntu
Baseline	b59816c21a6138174bf84fcc5b88a6ece8e1e41c
Compare	1d16fd14fa772621c7b09ebd00f83610207b1abf
Diff	Diff
Configs	CompilationMode:wasm, RunKind:micro

Regressions in System.Numerics.Tensors.Tests.Perf_BinaryIntegerTensorPrimitives<Int32>

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio
TrailingZeroCount - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	34.74 μs	42.96 μs	1.24	0.37	False
PopCount - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	859.48 ns	1.51 μs	1.76	0.17	True
PopCount - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	18.08 μs	30.05 μs	1.66	0.18	False

Test Report

Repro

General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.Numerics.Tensors.Tests.Perf_BinaryIntegerTensorPrimitives<Int32>*'

System.Numerics.Tensors.Tests.Perf_BinaryIntegerTensorPrimitives<Int32>.TrailingZeroCount(BufferLength: 3079)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_BinaryIntegerTensorPrimitives<Int32>.PopCount(BufferLength: 128)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_BinaryIntegerTensorPrimitives<Int32>.PopCount(BufferLength: 3079)

ETL Files

Histogram

JIT Disasms

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository

[lewing]

@tannergooding is this from dotnet/runtime@c65d921 ? range is dotnet/runtime@745b776...c8e4f2c

[tannergooding]

It could be, but its a bit surprising if that's the case?

Maybe this is a case where the jiterpreter needs to be updated to handle it as well, CC. @kg ?

[kg]

Adding interp intrinsics without updating the jiterpreter will usually regress performance, yes.

[tannergooding]

Adding interp intrinsics without updating the jiterpreter will usually regress performance, yes.

I see, that's a bit unfortunate. I'll keep it in mind for future PRs, however.

Is there something we can do in the meantime to improve the handling? I don't see anything in https://github.com/dotnet/runtime/blob/main/src/mono/browser/runtime/jiterpreter-trace-generator.ts or related files that looks to handle the concept of "bitcasting". However, I don't believe WASM needs to track the actual type and the PackedSimd support instead interprets the 128-bit value based on the opcode its passed into so perhaps we just need to handle it as a "no-op" and simply return the input value untouched?

Based on other code, this looks like it'd be roughly:

append_ldloc(builder, getArgU16(ip, 2), WasmOpcode.PREFIX_simd, WasmSimdOpcode.v128_load);
append_simd_store(builder, ip);

handled as part of emit_simd_2 (which it looks like means it has two operands on the stack where arg 1 is the destination and arg 2 is the source)?

[kg]

Adding interp intrinsics without updating the jiterpreter will usually regress performance, yes.

I see, that's a bit unfortunate. I'll keep it in mind for future PRs, however.

Is there something we can do in the meantime to improve the handling? I don't see anything in https://github.com/dotnet/runtime/blob/main/src/mono/browser/runtime/jiterpreter-trace-generator.ts or related files that looks to handle the concept of "bitcasting". However, I don't believe WASM needs to track the actual type and the PackedSimd support instead interprets the 128-bit value based on the opcode its passed into so perhaps we just need to handle it as a "no-op" and simply return the input value untouched?

Based on other code, this looks like it'd be roughly:

append_ldloc(builder, getArgU16(ip, 2), WasmOpcode.PREFIX_simd, WasmSimdOpcode.v128_load);
append_simd_store(builder, ip);

handled as part of emit_simd_2 (which it looks like means it has two operands on the stack where arg 1 is the destination and arg 2 is the source)?

That all looks right to me. A bitcast would indeed be a no-op in the WASM SIMD128 model since it's all just 'v128'. If the src and dest addresses are the same you can skip the operation entirely, but I don't know how often that would happen.

[tannergooding]

Are you aware of any other locations that might need to be updated, or is it mostly relegated to jiterpreter-trace-generator.ts?

[kg]

Are you aware of any other locations that might need to be updated, or is it mostly relegated to jiterpreter-trace-generator.ts?

The trace generator is the only part of the jiterpreter you should need to modify for most SIMD stuff. The table of intrinsics is automatically generated during the build process.

The jiterpreter's statistics system will automatically track which SIMD intrinsics were unhandled during the JIT process (if it's turned on) and you can use that to try and identify where there are coverage gaps.