Optimize the product of a householder-sequence with the identity, and optimize the evaluation of a HouseholderSequence to a dense matrix using faster blocked product.
diff --git a/CTestConfig.cmake b/CTestConfig.cmake index 4c00278..8b4cd79 100644 --- a/CTestConfig.cmake +++ b/CTestConfig.cmake
@@ -11,7 +11,7 @@ set(CTEST_DROP_SITE "manao.inria.fr") set(CTEST_DROP_LOCATION "/CDash/submit.php?project=Eigen") set(CTEST_DROP_SITE_CDASH TRUE) -set(CTEST_PROJECT_SUBPROJECTS -Official -Unsupported -) +#set(CTEST_PROJECT_SUBPROJECTS +#Official +#Unsupported +#)
diff --git a/Eigen/src/Householder/HouseholderSequence.h b/Eigen/src/Householder/HouseholderSequence.h index a4f40b7..af19701 100644 --- a/Eigen/src/Householder/HouseholderSequence.h +++ b/Eigen/src/Householder/HouseholderSequence.h
@@ -295,6 +295,14 @@ for(Index k = 0; k<cols()-vecs ; ++k) dst.col(k).tail(rows()-k-1).setZero(); } + else if(m_length>BlockSize) + { + dst.setIdentity(rows(), rows()); + if(m_reverse) + applyThisOnTheLeft(dst,workspace,true); + else + applyThisOnTheLeft(dst,workspace,true); + } else { dst.setIdentity(rows(), rows()); @@ -332,24 +340,27 @@ } /** \internal */ - template<typename Dest> inline void applyThisOnTheLeft(Dest& dst) const + template<typename Dest> inline void applyThisOnTheLeft(Dest& dst, bool inputIsIdentity = false) const { Matrix<Scalar,1,Dest::ColsAtCompileTime,RowMajor,1,Dest::MaxColsAtCompileTime> workspace; - applyThisOnTheLeft(dst, workspace); + applyThisOnTheLeft(dst, workspace, inputIsIdentity); } /** \internal */ template<typename Dest, typename Workspace> - inline void applyThisOnTheLeft(Dest& dst, Workspace& workspace) const + inline void applyThisOnTheLeft(Dest& dst, Workspace& workspace, bool inputIsIdentity = false) const { - const Index BlockSize = 48; + if(inputIsIdentity && m_reverse) + inputIsIdentity = false; // if the entries are large enough, then apply the reflectors by block if(m_length>=BlockSize && dst.cols()>1) { - for(Index i = 0; i < m_length; i+=BlockSize) + // Make sure we have at least 2 useful blocks, otherwise it is point-less: + Index blockSize = m_length<2*BlockSize ? (m_length+1)/2 : BlockSize; + for(Index i = 0; i < m_length; i+=blockSize) { - Index end = m_reverse ? (std::min)(m_length,i+BlockSize) : m_length-i; - Index k = m_reverse ? i : (std::max)(Index(0),end-BlockSize); + Index end = m_reverse ? (std::min)(m_length,i+blockSize) : m_length-i; + Index k = m_reverse ? i : (std::max)(Index(0),end-blockSize); Index bs = end-k; Index start = k + m_shift; @@ -359,7 +370,14 @@ Side==OnTheRight ? bs : m_vectors.rows()-start, Side==OnTheRight ? m_vectors.cols()-start : bs); typename internal::conditional<Side==OnTheRight, Transpose<SubVectorsType>, SubVectorsType&>::type sub_vecs(sub_vecs1); - Block<Dest,Dynamic,Dynamic> sub_dst(dst,dst.rows()-rows()+m_shift+k,0, rows()-m_shift-k,dst.cols()); + + Index dstStart = dst.rows()-rows()+m_shift+k; + Index dstRows = rows()-m_shift-k; + Block<Dest,Dynamic,Dynamic> sub_dst(dst, + dstStart, + inputIsIdentity ? dstStart : 0, + dstRows, + inputIsIdentity ? dstRows : dst.cols()); apply_block_householder_on_the_left(sub_dst, sub_vecs, m_coeffs.segment(k, bs), !m_reverse); } } @@ -369,7 +387,8 @@ for(Index k = 0; k < m_length; ++k) { Index actual_k = m_reverse ? k : m_length-k-1; - dst.bottomRows(rows()-m_shift-actual_k) + Index dstStart = rows()-m_shift-actual_k; + dst.bottomRightCorner(dstStart, inputIsIdentity ? dstStart : dst.cols()) .applyHouseholderOnTheLeft(essentialVector(actual_k), m_coeffs.coeff(actual_k), workspace.data()); } } @@ -387,7 +406,7 @@ { typename internal::matrix_type_times_scalar_type<Scalar, OtherDerived>::Type res(other.template cast<typename internal::matrix_type_times_scalar_type<Scalar,OtherDerived>::ResultScalar>()); - applyThisOnTheLeft(res); + applyThisOnTheLeft(res, internal::is_identity<OtherDerived>::value && res.rows()==res.cols()); return res; } @@ -461,6 +480,7 @@ bool m_reverse; Index m_length; Index m_shift; + enum { BlockSize = 48 }; }; /** \brief Computes the product of a matrix with a Householder sequence.
diff --git a/bench/bench_gemm.cpp b/bench/bench_gemm.cpp index 8528c55..688d99c 100644 --- a/bench/bench_gemm.cpp +++ b/bench/bench_gemm.cpp
@@ -129,7 +129,7 @@ template<typename A, typename B, typename C> EIGEN_DONT_INLINE void gemm(const A& a, const B& b, C& c) { - c.noalias() += a * b; + c.noalias() += a * b; } int main(int argc, char ** argv)