youmuliu/stgsharp
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Daily update on 20250921-173333 by Nitload Youmu. Additional Message : New design of matrix enumeration, capable for parallel task publication.

Browse Source 
Signed-off-by: Nitload <YoumuLiu_NSI@whv0t.onmicrosoft.com>
This commit is contained in:
Nitload
2025-09-21 17:33:33 +08:00
parent 0404ecf044
commit b8a51c13dd
15 changed files with 463 additions and 304 deletions
Download Patch File Download Diff File Expand all files Collapse all files
StgSharp.Common/Mathematics/Linear/MatrixParallel.Execute.cs
+43 -43
View File
@@ -33,42 +33,43 @@ using System;
using System.Collections.Generic;
using System.Formats.Asn1;
using System.Linq;
using System.Numerics;
using System.Runtime.CompilerServices;
using System.Text;
using System.Threading.Tasks;
namespace StgSharp.Mathematics
{
internal static partial class MatrixParallel
public static partial class MatrixParallel
{
[MethodImpl(MethodImplOptions.AggressiveInlining)]
internal static unsafe void Compute_Binary(MatrixParallelTaskPackage* p)
internal static unsafe void Compute_Binary(MatrixParallelTaskPackageNonGeneric* p)
{
MatrixKernel* l = p->Left, r = p->Right, a = p->Result;
long
lCount = p->LeftPrimOffset * p->LeftPrimStride,
rCount = p->RightPrimOffset * p->RightPrimStride,
aCount = p->ResultPrimOffset * p->ResultPrimStride,
pCount = p->PrimCount;
int eSize = p->ElementSize;
nint l = p->Left, r = p->Right, a = p->Result;
long
lCount = (long)p->LeftPrimOffset * p->LeftPrimStride,
rCount = (long)p->RightPrimOffset * p->RightPrimStride,
aCount = (long)p->ResultPrimOffset * p->ResultPrimStride,
pCount = p->PrimCount,
lLimit = pCount - p->LeftPrimOffset,
rLimit = pCount - p->RightPrimOffset,
aLimit = pCount - p->ResultPrimOffset;
delegate* unmanaged[Cdecl]<MatrixKernel*, MatrixKernel*, MatrixKernel*, void> op =
(delegate* unmanaged[Cdecl]<MatrixKernel*, MatrixKernel*, MatrixKernel*, void>)p->ComputeHandle;
delegate* unmanaged[Cdecl]<IntPtr, IntPtr, IntPtr, void> op =
(delegate* unmanaged[Cdecl]<IntPtr, IntPtr, IntPtr, void>)p->ComputeHandle;
for (long i = 0; i < pCount; i++)
{
lCount = i < lLimit ? lCount + p->LeftPrimOffset : 0;
rCount = i < rLimit ? rCount + p->RightPrimOffset : 0;
aCount = i < rLimit ? aCount + p->ResultPrimOffset : 0;
MatrixKernel* left = l + lCount;
MatrixKernel* right = r + rCount;
MatrixKernel* ans = a + aCount;
nint left = l + (nint)(lCount * eSize);
nint right = r + (nint)(rCount * eSize);
nint ans = a + (nint)(aCount * eSize);
long
lc = p->LeftSecOffset * p->LeftSecStride,
rc = p->RightSecOffset * p->RightSecStride,
ac = p->ResultSecOffset * p->ResultSecStride,
lc = (long)p->LeftSecOffset * p->LeftSecStride,
rc = (long)p->RightSecOffset * p->RightSecStride,
ac = (long)p->ResultSecOffset * p->ResultSecStride,
sCount = p->SecCount;
long
ll = sCount - p->LeftSecOffset,
@@ -79,15 +80,15 @@ namespace StgSharp.Mathematics
lc = j < ll ? lc + p->LeftSecOffset : 0;
rc = j < rl ? rc + p->RightSecOffset : 0;
ac = j < al ? ac + p->ResultSecOffset : 0;
op(left + lc, right + rc, ans + ac);
op(left + (nint)lc, right + (nint)rc, ans + (nint)ac);
}
}
}
[MethodImpl(MethodImplOptions.AggressiveInlining)]
internal static unsafe void Compute_BinaryScalar(MatrixParallelTaskPackage* p)
internal static unsafe void Compute_BinaryScalar(MatrixParallelTaskPackageNonGeneric* p)
{
MatrixKernel* l = p->Left, r = p->Right, a = p->Result;
nint l = p->Left, r = p->Right, a = p->Result;
long
lCount = p->LeftPrimOffset * p->LeftPrimStride,
rCount = p->RightPrimOffset * p->RightPrimStride,
@@ -97,23 +98,23 @@ namespace StgSharp.Mathematics
lLimit = pCount - p->LeftPrimOffset,
rLimit = pCount - p->RightPrimOffset,
aLimit = pCount - p->ResultPrimOffset;
delegate* unmanaged[Cdecl]<MatrixKernel*, MatrixKernel*, MatrixKernel*, ScalarPacket*, void> op =
(delegate* unmanaged[Cdecl]<MatrixKernel*, MatrixKernel*, MatrixKernel*, ScalarPacket*, void>)p->ComputeHandle;
delegate* unmanaged[Cdecl]<nint,nint,nint, ScalarPacket*, void> op =
(delegate* unmanaged[Cdecl]<nint, nint, nint, ScalarPacket*, void>)p->ComputeHandle;
ScalarPacket* scalar = p->Scalar;
for (long i = 0; i < pCount; i++)
{
lCount = i < lLimit ? lCount + p->LeftPrimOffset : 0;
rCount = i < rLimit ? rCount + p->RightPrimOffset : 0;
aCount = i < rLimit ? aCount + p->ResultPrimOffset : 0;
MatrixKernel* left = l + lCount;
MatrixKernel* right = r + rCount;
MatrixKernel* ans = a + aCount;
nint
left = l + (nint)lCount,
right = r + (nint)rCount,
ans = a + (nint)aCount;
long
lc = p->LeftSecOffset * p->LeftSecStride,
rc = p->RightSecOffset * p->RightSecStride,
ac = p->ResultSecOffset * p->ResultSecStride,
sCount = p->SecCount;
long
sCount = p->SecCount,
ll = sCount - p->LeftSecOffset,
rl = sCount - p->RightSecOffset,
al = sCount - p->ResultSecOffset;
@@ -122,28 +123,27 @@ namespace StgSharp.Mathematics
lc = j < ll ? lc + p->LeftSecOffset : 0;
rc = j < rl ? rc + p->RightSecOffset : 0;
ac = j < al ? ac + p->ResultSecOffset : 0;
op(left + lc, right + rc, ans + ac, scalar);
op(left + (nint)lc, right + (nint)rc, ans + (nint)ac, scalar);
}
}
}
[MethodImpl(MethodImplOptions.AggressiveInlining)]
internal static unsafe void Compute_Unary(MatrixParallelTaskPackage* p)
internal static unsafe void Compute_Unary(MatrixParallelTaskPackageNonGeneric* p)
{
MatrixKernel* s = p->Right, a = p->Result;
nint s = p->Right, a = p->Result;
long rCount = p->RightPrimOffset * p->RightPrimStride,
aCount = p->ResultPrimOffset * p->ResultPrimStride,
pCount = p->PrimCount;
aCount = p->ResultPrimOffset * p->ResultPrimStride,
pCount = p->PrimCount;
long rLimit = pCount - p->RightPrimOffset,
aLimit = pCount - p->ResultPrimOffset;
delegate* unmanaged[Cdecl]<MatrixKernel*, MatrixKernel*, void> op =
(delegate* unmanaged[Cdecl]<MatrixKernel*, MatrixKernel*, void>)p->ComputeHandle;
delegate* unmanaged[Cdecl]<nint, nint, void> op =
(delegate* unmanaged[Cdecl]<nint, nint, void>)p->ComputeHandle;
for (long i = 0; i < pCount; i++)
{
rCount = i < rLimit ? rCount + p->RightPrimOffset : 0;
aCount = i < rLimit ? aCount + p->ResultPrimOffset : 0;
MatrixKernel* source = s + rCount;
MatrixKernel* ans = a + aCount;
nint source = s + (nint)rCount, ans = a + (nint)aCount;
long rc = p->RightSecOffset * p->RightSecStride,
ac = p->ResultSecOffset * p->ResultSecStride,
sCount = p->SecCount;
@@ -153,29 +153,29 @@ namespace StgSharp.Mathematics
{
rc = j < rl ? rc + p->RightSecOffset : 0;
ac = j < al ? ac + p->ResultSecOffset : 0;
op(source + rc, ans + ac);
op(source + (nint)rc, ans + (nint)ac);
}
}
}
[MethodImpl(MethodImplOptions.AggressiveInlining)]
internal static unsafe void Compute_UnaryScalar(MatrixParallelTaskPackage* p)
internal static unsafe void Compute_UnaryScalar(MatrixParallelTaskPackageNonGeneric* p)
{
MatrixKernel* s = p->Right, a = p->Result;
nint s = p->Right, a = p->Result;
long rCount = p->RightPrimOffset * p->RightPrimStride,
aCount = p->ResultPrimOffset * p->ResultPrimStride,
pCount = p->PrimCount;
long rLimit = pCount - p->RightPrimOffset,
aLimit = pCount - p->ResultPrimOffset;
delegate* unmanaged[Cdecl]<MatrixKernel*, MatrixKernel*, ScalarPacket*, void> op =
(delegate* unmanaged[Cdecl]<MatrixKernel*, MatrixKernel*, ScalarPacket*, void>)p->ComputeHandle;
delegate* unmanaged[Cdecl]<nint, nint, ScalarPacket*, void> op =
(delegate* unmanaged[Cdecl]<nint, nint, ScalarPacket*, void>)p->ComputeHandle;
ScalarPacket* scalar = p->Scalar;
for (long i = 0; i < pCount; i++)
{
rCount = i < rLimit ? rCount + p->RightPrimOffset : 0;
aCount = i < rLimit ? aCount + p->ResultPrimOffset : 0;
MatrixKernel* source = s + rCount;
MatrixKernel* ans = a + aCount;
nint source = s + (nint)rCount;
nint ans = a + (nint)aCount;
long rc = p->RightSecOffset * p->RightSecStride,
ac = p->ResultSecOffset * p->ResultSecStride,
sCount = p->SecCount;
@@ -185,7 +185,7 @@ namespace StgSharp.Mathematics
{
rc = j < rl ? rc + p->RightSecOffset : 0;
ac = j < al ? ac + p->ResultSecOffset : 0;
op(source + rc, ans + ac, scalar);
op(source + (nint)rc, ans + (nint)ac, scalar);
}
}
}