devtools/eigen
1
0
Fork 0
mirror of https://gitlab.com/libeigen/eigen.git synced 2026-04-10 11:34:33 +08:00
Code Issues Packages Projects Releases Wiki Activity

Extend support for Packet16b:

Browse Source 
* Add ptranspose<*,4> to support matmul and add unit test for Matrix<bool> * Matrix<bool>
* work around a bug in slicing of Tensor<bool>.
* Add tensor tests

This speeds up matmul for boolean matrices by about 10x

name                            old time/op             new time/op             delta
BM_MatMul<bool>/8                267ns ± 0%              479ns ± 0%  +79.25%          (p=0.008 n=5+5)
BM_MatMul<bool>/32              6.42µs ± 0%             0.87µs ± 0%  -86.50%          (p=0.008 n=5+5)
BM_MatMul<bool>/64              43.3µs ± 0%              5.9µs ± 0%  -86.42%          (p=0.008 n=5+5)
BM_MatMul<bool>/128              315µs ± 0%               44µs ± 0%  -85.98%          (p=0.008 n=5+5)
BM_MatMul<bool>/256             2.41ms ± 0%             0.34ms ± 0%  -85.68%          (p=0.008 n=5+5)
BM_MatMul<bool>/512             18.8ms ± 0%              2.7ms ± 0%  -85.53%          (p=0.008 n=5+5)
BM_MatMul<bool>/1k               149ms ± 0%               22ms ± 0%  -85.40%          (p=0.008 n=5+5)
This commit is contained in:
Rasmus Munk Larsen
2020-04-24 17:29:25 -07:00
parent b47c777993
commit ab773c7e91
10 changed files with 267 additions and 162 deletions
Download Patch File Download Diff File Expand all files Collapse all files

29
test/product_small.cpp
View File

@@ -56,6 +56,31 @@ test_lazy_single(int rows, int cols, int depth)
VERIFY_IS_APPROX(C+=A.lazyProduct(B), ref_prod(D,A,B));
}
template<typename T>
void test_dynamic_exact()
{
int rows = internal::random<int>(1,64);
int cols = internal::random<int>(1,64);
int depth = internal::random<int>(1,65);
typedef Matrix<T,Dynamic,Dynamic> MatrixX;
MatrixX A(rows,depth); A.setRandom();
MatrixX B(depth,cols); B.setRandom();
MatrixX C(rows,cols); C.setRandom();
MatrixX D(C);
for(Index i=0;i<C.rows();++i)
for(Index j=0;j<C.cols();++j)
for(Index k=0;k<A.cols();++k)
D.coeffRef(i,j) |= A.coeff(i,k) & B.coeff(k,j);
C += A * B;
VERIFY_IS_EQUAL(C, D);
MatrixX E = B.transpose();
for(Index i=0;i<B.rows();++i)
for(Index j=0;j<B.cols();++j)
VERIFY_IS_EQUAL(B(i,j), E(j,i));
}
template<typename T, int Rows, int Cols, int Depth, int OC, int OA, int OB>
typename internal::enable_if< ( (Rows ==1&&Depth!=1&&OA==ColMajor)
|| (Depth==1&&Rows !=1&&OA==RowMajor)
@@ -78,7 +103,7 @@ void test_lazy_all_layout(int rows=Rows, int cols=Cols, int depth=Depth)
CALL_SUBTEST(( test_lazy_single<T,Rows,Cols,Depth,RowMajor,ColMajor,RowMajor>(rows,cols,depth) ));
CALL_SUBTEST(( test_lazy_single<T,Rows,Cols,Depth,ColMajor,RowMajor,RowMajor>(rows,cols,depth) ));
CALL_SUBTEST(( test_lazy_single<T,Rows,Cols,Depth,RowMajor,RowMajor,RowMajor>(rows,cols,depth) ));
}
}
template<typename T>
void test_lazy_l1()
@@ -291,6 +316,8 @@ EIGEN_DECLARE_TEST(product_small)
CALL_SUBTEST_6( bug_1311<3>() );
CALL_SUBTEST_6( bug_1311<5>() );
CALL_SUBTEST_9( test_dynamic_exact<bool>() );
}
CALL_SUBTEST_6( product_small_regressions<0>() );