Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

CI test #99871

Closed
wants to merge 3 commits into from
Closed

CI test #99871

wants to merge 3 commits into from

Conversation

EgorBo
Copy link
Member

@EgorBo EgorBo commented Mar 16, 2024

No description provided.

@dotnet-issue-labeler dotnet-issue-labeler bot added the area-CodeGen-coreclr CLR JIT compiler in src/coreclr/src/jit and related components such as SuperPMI label Mar 16, 2024
@dotnet dotnet deleted a comment from EgorBot Mar 16, 2024
@dotnet dotnet deleted a comment from EgorBot Mar 17, 2024
@EgorBo EgorBo closed this Mar 17, 2024
@EgorBo EgorBo reopened this Mar 17, 2024
@EgorBo EgorBo closed this Mar 17, 2024
@github-actions github-actions bot locked and limited conversation to collaborators Apr 17, 2024
@EgorBo EgorBo reopened this May 3, 2024
@dotnet dotnet unlocked this conversation May 3, 2024
@EgorBo EgorBo closed this May 11, 2024
@EgorBo EgorBo reopened this May 18, 2024
@EgorBo EgorBo closed this May 19, 2024
@dotnet dotnet deleted a comment from EgorBot May 19, 2024
@dotnet dotnet deleted a comment from EgorBot May 19, 2024
@dotnet dotnet deleted a comment from EgorBot May 19, 2024
@dotnet dotnet deleted a comment from EgorBot May 19, 2024
@dotnet dotnet deleted a comment from EgorBot May 19, 2024
@dotnet dotnet deleted a comment from EgorBot May 19, 2024
@dotnet dotnet deleted a comment from EgorBot May 19, 2024
@dotnet dotnet deleted a comment from EgorBot May 19, 2024
@EgorBo EgorBo reopened this May 20, 2024
@dotnet dotnet deleted a comment from EgorBot May 20, 2024
@dotnet dotnet deleted a comment from EgorBot May 20, 2024
@dotnet dotnet deleted a comment from EgorBot May 20, 2024
@dotnet dotnet deleted a comment from EgorBot May 20, 2024
@dotnet dotnet deleted a comment from EgorBot May 20, 2024
@dotnet dotnet deleted a comment from EgorBot May 20, 2024
@dotnet dotnet deleted a comment from EgorBot May 20, 2024
@dotnet dotnet deleted a comment from EgorBot Jun 28, 2024
@dotnet dotnet deleted a comment from EgorBot Jun 28, 2024
@dotnet dotnet deleted a comment from EgorBot Jun 28, 2024
@EgorBo
Copy link
Member Author

EgorBo commented Jun 28, 2024

@EgorBot -intel

using BenchmarkDotNet.Attributes;
using BenchmarkDotNet.Running;
using System.Runtime.InteropServices;

BenchmarkRunner.Run<Utf8Bench>(args: args);

public unsafe class Utf8Bench
{
    byte* _ptr1 = null;
    byte* _ptr2 = null;

    [GlobalSetup]
    public void GlobalSetup()
    {
        _ptr1 = (byte*)NativeMemory.AlignedAlloc(300 * 1024 * 1024, 64);
        _ptr2 = (byte*)NativeMemory.AlignedAlloc(300 * 1024 * 1024, 64);
    }

    [GlobalCleanup]
    public void GlobalCleanup()
    {
        NativeMemory.Free(_ptr1);
        NativeMemory.Free(_ptr2);
    }

    public static IEnumerable<long> Sizes()
    {
        yield return     100_000;
    }

    [Benchmark]
    [ArgumentsSource(nameof(Sizes))]
    public void Managed(long size) => Buffer.MemoryCopy(_ptr1, _ptr2, size, size);
}

4 similar comments
@EgorBo
Copy link
Member Author

EgorBo commented Jun 28, 2024

@EgorBot -intel

using BenchmarkDotNet.Attributes;
using BenchmarkDotNet.Running;
using System.Runtime.InteropServices;

BenchmarkRunner.Run<Utf8Bench>(args: args);

public unsafe class Utf8Bench
{
    byte* _ptr1 = null;
    byte* _ptr2 = null;

    [GlobalSetup]
    public void GlobalSetup()
    {
        _ptr1 = (byte*)NativeMemory.AlignedAlloc(300 * 1024 * 1024, 64);
        _ptr2 = (byte*)NativeMemory.AlignedAlloc(300 * 1024 * 1024, 64);
    }

    [GlobalCleanup]
    public void GlobalCleanup()
    {
        NativeMemory.Free(_ptr1);
        NativeMemory.Free(_ptr2);
    }

    public static IEnumerable<long> Sizes()
    {
        yield return     100_000;
    }

    [Benchmark]
    [ArgumentsSource(nameof(Sizes))]
    public void Managed(long size) => Buffer.MemoryCopy(_ptr1, _ptr2, size, size);
}

@EgorBo
Copy link
Member Author

EgorBo commented Jun 28, 2024

@EgorBot -intel

using BenchmarkDotNet.Attributes;
using BenchmarkDotNet.Running;
using System.Runtime.InteropServices;

BenchmarkRunner.Run<Utf8Bench>(args: args);

public unsafe class Utf8Bench
{
    byte* _ptr1 = null;
    byte* _ptr2 = null;

    [GlobalSetup]
    public void GlobalSetup()
    {
        _ptr1 = (byte*)NativeMemory.AlignedAlloc(300 * 1024 * 1024, 64);
        _ptr2 = (byte*)NativeMemory.AlignedAlloc(300 * 1024 * 1024, 64);
    }

    [GlobalCleanup]
    public void GlobalCleanup()
    {
        NativeMemory.Free(_ptr1);
        NativeMemory.Free(_ptr2);
    }

    public static IEnumerable<long> Sizes()
    {
        yield return     100_000;
    }

    [Benchmark]
    [ArgumentsSource(nameof(Sizes))]
    public void Managed(long size) => Buffer.MemoryCopy(_ptr1, _ptr2, size, size);
}

@EgorBo
Copy link
Member Author

EgorBo commented Jun 28, 2024

@EgorBot -intel

using BenchmarkDotNet.Attributes;
using BenchmarkDotNet.Running;
using System.Runtime.InteropServices;

BenchmarkRunner.Run<Utf8Bench>(args: args);

public unsafe class Utf8Bench
{
    byte* _ptr1 = null;
    byte* _ptr2 = null;

    [GlobalSetup]
    public void GlobalSetup()
    {
        _ptr1 = (byte*)NativeMemory.AlignedAlloc(300 * 1024 * 1024, 64);
        _ptr2 = (byte*)NativeMemory.AlignedAlloc(300 * 1024 * 1024, 64);
    }

    [GlobalCleanup]
    public void GlobalCleanup()
    {
        NativeMemory.Free(_ptr1);
        NativeMemory.Free(_ptr2);
    }

    public static IEnumerable<long> Sizes()
    {
        yield return     100_000;
    }

    [Benchmark]
    [ArgumentsSource(nameof(Sizes))]
    public void Managed(long size) => Buffer.MemoryCopy(_ptr1, _ptr2, size, size);
}

@EgorBo
Copy link
Member Author

EgorBo commented Jun 28, 2024

@EgorBot -intel

using BenchmarkDotNet.Attributes;
using BenchmarkDotNet.Running;
using System.Runtime.InteropServices;

BenchmarkRunner.Run<Utf8Bench>(args: args);

public unsafe class Utf8Bench
{
    byte* _ptr1 = null;
    byte* _ptr2 = null;

    [GlobalSetup]
    public void GlobalSetup()
    {
        _ptr1 = (byte*)NativeMemory.AlignedAlloc(300 * 1024 * 1024, 64);
        _ptr2 = (byte*)NativeMemory.AlignedAlloc(300 * 1024 * 1024, 64);
    }

    [GlobalCleanup]
    public void GlobalCleanup()
    {
        NativeMemory.Free(_ptr1);
        NativeMemory.Free(_ptr2);
    }

    public static IEnumerable<long> Sizes()
    {
        yield return     100_000;
    }

    [Benchmark]
    [ArgumentsSource(nameof(Sizes))]
    public void Managed(long size) => Buffer.MemoryCopy(_ptr1, _ptr2, size, size);
}

@EgorBot
Copy link

EgorBot commented Jun 28, 2024

❌ Benchmark failed on Intel
publishing results failed

@EgorBo
Copy link
Member Author

EgorBo commented Jun 28, 2024

@EgorBot -intel

using BenchmarkDotNet.Attributes;
using BenchmarkDotNet.Running;
using System.Runtime.InteropServices;

BenchmarkRunner.Run<Utf8Bench>(args: args);

public unsafe class Utf8Bench
{
    byte* _ptr1 = null;
    byte* _ptr2 = null;

    [GlobalSetup]
    public void GlobalSetup()
    {
        _ptr1 = (byte*)NativeMemory.AlignedAlloc(300 * 1024 * 1024, 64);
        _ptr2 = (byte*)NativeMemory.AlignedAlloc(300 * 1024 * 1024, 64);
    }

    [GlobalCleanup]
    public void GlobalCleanup()
    {
        NativeMemory.Free(_ptr1);
        NativeMemory.Free(_ptr2);
    }

    public static IEnumerable<long> Sizes()
    {
        yield return     100_000;
    }

    [Benchmark]
    [ArgumentsSource(nameof(Sizes))]
    public void Managed(long size) => Buffer.MemoryCopy(_ptr1, _ptr2, size, size);
}

@EgorBot
Copy link

EgorBot commented Jun 28, 2024

❌ Benchmark failed on Intel
publishing results failed

@EgorBo
Copy link
Member Author

EgorBo commented Jun 28, 2024

@EgorBot -intel

using BenchmarkDotNet.Attributes;
using BenchmarkDotNet.Running;
using System.Runtime.InteropServices;

BenchmarkRunner.Run<Utf8Bench>(args: args);

public unsafe class Utf8Bench
{
    byte* _ptr1 = null;
    byte* _ptr2 = null;

    [GlobalSetup]
    public void GlobalSetup()
    {
        _ptr1 = (byte*)NativeMemory.AlignedAlloc(300 * 1024 * 1024, 64);
        _ptr2 = (byte*)NativeMemory.AlignedAlloc(300 * 1024 * 1024, 64);
    }

    [GlobalCleanup]
    public void GlobalCleanup()
    {
        NativeMemory.Free(_ptr1);
        NativeMemory.Free(_ptr2);
    }

    public static IEnumerable<long> Sizes()
    {
        yield return     100_000;
    }

    [Benchmark]
    [ArgumentsSource(nameof(Sizes))]
    public void Managed(long size) => Buffer.MemoryCopy(_ptr1, _ptr2, size, size);
}

@EgorBot
Copy link

EgorBot commented Jun 28, 2024

Benchmark results on Intel
BenchmarkDotNet v0.13.12, Ubuntu 22.04.4 LTS (Jammy Jellyfish)
Intel Xeon Platinum 8370C CPU 2.80GHz, 1 CPU, 8 logical and 4 physical cores
  Job-FRYCZO : .NET 9.0.0 (42.42.42.42424), X64 RyuJIT AVX-512F+CD+BW+DQ+VL+VBMI
  Job-VHWOZG : .NET 9.0.0 (42.42.42.42424), X64 RyuJIT AVX-512F+CD+BW+DQ+VL+VBMI
Method Toolchain size Mean Error Ratio
Managed Main 100000 2.167 μs 0.0004 μs 1.00
Managed PR 100000 2.265 μs 0.0002 μs 1.05

BDN_Artifacts.zip

@EgorBo EgorBo closed this Jun 29, 2024
@github-actions github-actions bot locked and limited conversation to collaborators Jul 31, 2024
@EgorBo
Copy link
Member Author

EgorBo commented Aug 13, 2024

@EgorBot -intel

using BenchmarkDotNet.Attributes;
using BenchmarkDotNet.Running;

BenchmarkRunner.Run<MyBench>(args: args); // args must be forwarded to BDN!

public class MyBench
{
    string _str = "dotnet/RUNTIME";

    [Benchmark]
    public bool Test() =>
        // JIT is expected to unroll this with SIMD:
        _str.Equals("dotnet/runtime", StringComparison.OrdinalIgnoreCase);
}

@dotnet dotnet unlocked this conversation Aug 13, 2024
@EgorBo
Copy link
Member Author

EgorBo commented Aug 13, 2024

@EgorBot -intel

using BenchmarkDotNet.Attributes;
using BenchmarkDotNet.Running;

BenchmarkRunner.Run<MyBench>(args: args); // args must be forwarded to BDN!

public class MyBench
{
    string _str = "dotnet/RUNTIME";

    [Benchmark]
    public bool Test() =>
        // JIT is expected to unroll this with SIMD:
        _str.Equals("dotnet/runtime", StringComparison.OrdinalIgnoreCase);
}

@EgorBot
Copy link

EgorBot commented Aug 13, 2024

Benchmark results on Intel
BenchmarkDotNet v0.14.0, Ubuntu 22.04.4 LTS (Jammy Jellyfish)
Intel Xeon Platinum 8370C CPU 2.80GHz, 1 CPU, 16 logical and 8 physical cores
  Job-PYUHGT : .NET 9.0.0 (42.42.42.42424), X64 RyuJIT AVX-512F+CD+BW+DQ+VL+VBMI
  Job-TVEMWZ : .NET 9.0.0 (42.42.42.42424), X64 RyuJIT AVX-512F+CD+BW+DQ+VL+VBMI
StdDev=0.0002 ns
Method Toolchain Mean Error Ratio
Test Main 0.5734 ns 0.0002 ns 1.00
Test PR 0.5738 ns 0.0002 ns 1.00

BDN_Artifacts.zip

@EgorBo
Copy link
Member Author

EgorBo commented Aug 13, 2024

@EgorBot -arm64 -profile

using System.Runtime.CompilerServices;
using BenchmarkDotNet.Attributes;
using BenchmarkDotNet.Running;

BenchmarkRunner.Run<Bench>(args: args);

public class Bench
{
    // struct in heap
    MyStruct MS = new();

    [Benchmark]
    public void CheckedWB_InHeap()
    {
        string str = new string('x', 1);
        for (long i = 0; i < 200000000; i++)
            CheckedWriteBarrier(ref MS, str);
    }

    [MethodImpl(MethodImplOptions.NoInlining)]
    void CheckedWriteBarrier(ref MyStruct ms, string str) => ms.Str = str;
}

public record struct MyStruct(string Str);

internal class Foo
{
    public volatile Foo x;
}

@EgorBot
Copy link

EgorBot commented Aug 13, 2024

Benchmark results on Arm64
BenchmarkDotNet v0.14.0, Ubuntu 22.04.4 LTS (Jammy Jellyfish)
Unknown processor
  Job-ISALJM : .NET 9.0.0 (42.42.42.42424), Arm64 RyuJIT AdvSIMD
  Job-BBFFUH : .NET 9.0.0 (42.42.42.42424), Arm64 RyuJIT AdvSIMD
Method Toolchain Mean Error Ratio
CheckedWB_InHeap Main 801.7 ms 0.42 ms 1.00
CheckedWB_InHeap PR 801.6 ms 0.48 ms 1.00

BDN_Artifacts.zip

Flame graphs: Main vs PR 🔥
Hot asm: Main vs PR
Hot functions: Main vs PR

For clean perf results, make sure you have just one [Benchmark] in your app.

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
area-CodeGen-coreclr CLR JIT compiler in src/coreclr/src/jit and related components such as SuperPMI
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants