derek.holloway
7502bf64a4
CodeQL / Analyze (actions) (push) Canceled after 0s
CodeQL / Analyze (python) (push) Canceled after 0s
pre-commit / pre-commit (push) Canceled after 0s
Multi-Arch CI / setup (push) Canceled after 0s
Publish build_manylinux_x86_64 images / publish_build_manylinux_x86_64 (push) Canceled after 0s
Setup (Multi-Arch) / setup (push) Canceled after 0s
Multi-Arch CI / Linux::${{ fromJSON(needs.setup.outputs.linux_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubi10 image / publish_no_rocm_image_ubi10 (push) Canceled after 0s
Multi-Arch CI / Windows::${{ fromJSON(needs.setup.outputs.windows_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Multi-Arch CI / Manifest Diff (push) Canceled after 0s
TheRock Manifest Diff Report / Generate Manifest Diff Report (push) Canceled after 0s
Multi-Arch CI / CI Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_ocl_rt images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_rocgdb images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Unit Tests / Unit Tests :: ubuntu-24.04 (push) Canceled after 0s
Unit Tests / Unit Tests :: windows-2022 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_media (push) Canceled after 0s
Unit Tests / Unit Tests Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_openmpi (push) Canceled after 0s
925 lines
38 KiB
C++
925 lines
38 KiB
C++
/* **************************************************************************
|
|
* Copyright (C) 2020-2025 Advanced Micro Devices, Inc. All rights reserved.
|
|
*
|
|
* Redistribution and use in source and binary forms, with or without
|
|
* modification, are permitted provided that the following conditions
|
|
* are met:
|
|
*
|
|
* 1. Redistributions of source code must retain the above copyright
|
|
* notice, this list of conditions and the following disclaimer.
|
|
*
|
|
* 2. Redistributions in binary form must reproduce the above copyright
|
|
* notice, this list of conditions and the following disclaimer in the
|
|
* documentation and/or other materials provided with the distribution.
|
|
*
|
|
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
|
|
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
|
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
|
|
* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
|
|
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
|
|
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
|
|
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
|
|
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
|
|
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
|
|
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
|
|
* SUCH DAMAGE.
|
|
* *************************************************************************/
|
|
|
|
#pragma once
|
|
|
|
#include "common/misc/client_util.hpp"
|
|
#include "common/misc/clientcommon.hpp"
|
|
#include "common/misc/lapack_host_reference.hpp"
|
|
#include "common/misc/norm.hpp"
|
|
#include "common/misc/rocsolver.hpp"
|
|
#include "common/misc/rocsolver_arguments.hpp"
|
|
#include "common/misc/rocsolver_test.hpp"
|
|
#include "common/misc/rocsolver_timer.hpp"
|
|
|
|
template <bool STRIDED, typename T, typename TT, typename W, typename U>
|
|
void gesvd_checkBadArgs(const rocblas_handle handle,
|
|
const rocblas_svect left_svect,
|
|
const rocblas_svect right_svect,
|
|
const rocblas_int m,
|
|
const rocblas_int n,
|
|
W dA,
|
|
const rocblas_int lda,
|
|
const rocblas_stride stA,
|
|
TT dS,
|
|
const rocblas_stride stS,
|
|
T dU,
|
|
const rocblas_int ldu,
|
|
const rocblas_stride stU,
|
|
T dV,
|
|
const rocblas_int ldv,
|
|
const rocblas_stride stV,
|
|
TT dE,
|
|
const rocblas_stride stE,
|
|
const rocblas_workmode fa,
|
|
U dinfo,
|
|
const rocblas_int bc)
|
|
{
|
|
// handle
|
|
EXPECT_ROCBLAS_STATUS(rocsolver_gesvd(STRIDED, nullptr, left_svect, right_svect, m, n, dA, lda,
|
|
stA, dS, stS, dU, ldu, stU, dV, ldv, stV, dE, stE, fa,
|
|
dinfo, bc),
|
|
rocblas_status_invalid_handle);
|
|
|
|
// values
|
|
EXPECT_ROCBLAS_STATUS(rocsolver_gesvd(STRIDED, handle, rocblas_svect(0), right_svect, m, n, dA,
|
|
lda, stA, dS, stS, dU, ldu, stU, dV, ldv, stV, dE, stE,
|
|
fa, dinfo, bc),
|
|
rocblas_status_invalid_value);
|
|
EXPECT_ROCBLAS_STATUS(rocsolver_gesvd(STRIDED, handle, left_svect, rocblas_svect(0), m, n, dA,
|
|
lda, stA, dS, stS, dU, ldu, stU, dV, ldv, stV, dE, stE,
|
|
fa, dinfo, bc),
|
|
rocblas_status_invalid_value);
|
|
EXPECT_ROCBLAS_STATUS(rocsolver_gesvd(STRIDED, handle, rocblas_svect_overwrite,
|
|
rocblas_svect_overwrite, m, n, dA, lda, stA, dS, stS, dU,
|
|
ldu, stU, dV, ldv, stV, dE, stE, fa, dinfo, bc),
|
|
rocblas_status_invalid_value);
|
|
|
|
// sizes (only check batch_count if applicable)
|
|
if(STRIDED)
|
|
EXPECT_ROCBLAS_STATUS(rocsolver_gesvd(STRIDED, handle, left_svect, right_svect, m, n, dA,
|
|
lda, stA, dS, stS, dU, ldu, stU, dV, ldv, stV, dE,
|
|
stE, fa, dinfo, -1),
|
|
rocblas_status_invalid_size);
|
|
|
|
// pointers
|
|
EXPECT_ROCBLAS_STATUS(rocsolver_gesvd(STRIDED, handle, left_svect, right_svect, m, n,
|
|
(W) nullptr, lda, stA, dS, stS, dU, ldu, stU, dV, ldv,
|
|
stV, dE, stE, fa, dinfo, bc),
|
|
rocblas_status_invalid_pointer);
|
|
EXPECT_ROCBLAS_STATUS(rocsolver_gesvd(STRIDED, handle, left_svect, right_svect, m, n, dA, lda,
|
|
stA, (TT) nullptr, stS, dU, ldu, stU, dV, ldv, stV, dE,
|
|
stE, fa, dinfo, bc),
|
|
rocblas_status_invalid_pointer);
|
|
EXPECT_ROCBLAS_STATUS(rocsolver_gesvd(STRIDED, handle, left_svect, right_svect, m, n, dA, lda,
|
|
stA, dS, stS, (T) nullptr, ldu, stU, dV, ldv, stV, dE,
|
|
stE, fa, dinfo, bc),
|
|
rocblas_status_invalid_pointer);
|
|
EXPECT_ROCBLAS_STATUS(rocsolver_gesvd(STRIDED, handle, left_svect, right_svect, m, n, dA, lda,
|
|
stA, dS, stS, dU, ldu, stU, (T) nullptr, ldv, stV, dE,
|
|
stE, fa, dinfo, bc),
|
|
rocblas_status_invalid_pointer);
|
|
EXPECT_ROCBLAS_STATUS(rocsolver_gesvd(STRIDED, handle, left_svect, right_svect, m, n, dA, lda,
|
|
stA, dS, stS, dU, ldu, stU, dV, ldv, stV, (TT) nullptr,
|
|
stE, fa, dinfo, bc),
|
|
rocblas_status_invalid_pointer);
|
|
EXPECT_ROCBLAS_STATUS(rocsolver_gesvd(STRIDED, handle, left_svect, right_svect, m, n, dA, lda,
|
|
stA, dS, stS, dU, ldu, stU, dV, ldv, stV, dE, stE, fa,
|
|
(U) nullptr, bc),
|
|
rocblas_status_invalid_pointer);
|
|
|
|
// quick return with invalid pointers
|
|
EXPECT_ROCBLAS_STATUS(rocsolver_gesvd(STRIDED, handle, left_svect, right_svect, 0, n,
|
|
(W) nullptr, lda, stA, (TT) nullptr, stS, (T) nullptr,
|
|
ldu, stU, dV, ldv, stV, (TT) nullptr, stE, fa, dinfo, bc),
|
|
rocblas_status_success);
|
|
EXPECT_ROCBLAS_STATUS(rocsolver_gesvd(STRIDED, handle, left_svect, right_svect, m, 0,
|
|
(W) nullptr, lda, stA, (TT) nullptr, stS, dU, ldu, stU,
|
|
(T) nullptr, ldv, stV, (TT) nullptr, stE, fa, dinfo, bc),
|
|
rocblas_status_success);
|
|
|
|
// quick return with zero batch_count if applicable
|
|
if(STRIDED)
|
|
EXPECT_ROCBLAS_STATUS(rocsolver_gesvd(STRIDED, handle, left_svect, right_svect, m, n, dA,
|
|
lda, stA, dS, stS, dU, ldu, stU, dV, ldv, stV, dE,
|
|
stE, fa, (U) nullptr, 0),
|
|
rocblas_status_success);
|
|
}
|
|
|
|
template <bool BATCHED, bool STRIDED, typename T>
|
|
void testing_gesvd_bad_arg()
|
|
{
|
|
using S = decltype(std::real(T{}));
|
|
|
|
// safe arguments
|
|
rocblas_local_handle handle;
|
|
rocblas_svect left_svect = rocblas_svect_all;
|
|
rocblas_svect right_svect = rocblas_svect_all;
|
|
rocblas_int m = 2;
|
|
rocblas_int n = 2;
|
|
rocblas_int lda = 2;
|
|
rocblas_int ldu = 2;
|
|
rocblas_int ldv = 2;
|
|
rocblas_stride stA = 2;
|
|
rocblas_stride stS = 2;
|
|
rocblas_stride stU = 2;
|
|
rocblas_stride stV = 2;
|
|
rocblas_stride stE = 2;
|
|
rocblas_int bc = 1;
|
|
rocblas_workmode fa = rocblas_outofplace;
|
|
|
|
if(BATCHED)
|
|
{
|
|
// memory allocations
|
|
device_batch_vector<T> dA(1, 1, 1);
|
|
device_strided_batch_vector<S> dS(1, 1, 1, 1);
|
|
device_strided_batch_vector<T> dU(1, 1, 1, 1);
|
|
device_strided_batch_vector<T> dV(1, 1, 1, 1);
|
|
device_strided_batch_vector<S> dE(1, 1, 1, 1);
|
|
device_strided_batch_vector<rocblas_int> dinfo(1, 1, 1, 1);
|
|
CHECK_HIP_ERROR(dA.memcheck());
|
|
CHECK_HIP_ERROR(dS.memcheck());
|
|
CHECK_HIP_ERROR(dU.memcheck());
|
|
CHECK_HIP_ERROR(dV.memcheck());
|
|
CHECK_HIP_ERROR(dE.memcheck());
|
|
CHECK_HIP_ERROR(dinfo.memcheck());
|
|
|
|
// check bad arguments
|
|
gesvd_checkBadArgs<STRIDED>(handle, left_svect, right_svect, m, n, dA.data(), lda, stA,
|
|
dS.data(), stS, dU.data(), ldu, stU, dV.data(), ldv, stV,
|
|
dE.data(), stE, fa, dinfo.data(), bc);
|
|
}
|
|
else
|
|
{
|
|
// memory allocations
|
|
device_strided_batch_vector<T> dA(1, 1, 1, 1);
|
|
device_strided_batch_vector<S> dS(1, 1, 1, 1);
|
|
device_strided_batch_vector<T> dU(1, 1, 1, 1);
|
|
device_strided_batch_vector<T> dV(1, 1, 1, 1);
|
|
device_strided_batch_vector<S> dE(1, 1, 1, 1);
|
|
device_strided_batch_vector<rocblas_int> dinfo(1, 1, 1, 1);
|
|
CHECK_HIP_ERROR(dA.memcheck());
|
|
CHECK_HIP_ERROR(dS.memcheck());
|
|
CHECK_HIP_ERROR(dU.memcheck());
|
|
CHECK_HIP_ERROR(dV.memcheck());
|
|
CHECK_HIP_ERROR(dE.memcheck());
|
|
CHECK_HIP_ERROR(dinfo.memcheck());
|
|
|
|
// check bad arguments
|
|
gesvd_checkBadArgs<STRIDED>(handle, left_svect, right_svect, m, n, dA.data(), lda, stA,
|
|
dS.data(), stS, dU.data(), ldu, stU, dV.data(), ldv, stV,
|
|
dE.data(), stE, fa, dinfo.data(), bc);
|
|
}
|
|
}
|
|
|
|
template <bool CPU, bool GPU, typename T, typename Td, typename Th>
|
|
void gesvd_initData(const rocblas_handle handle,
|
|
const rocblas_svect left_svect,
|
|
const rocblas_svect right_svect,
|
|
const rocblas_int m,
|
|
const rocblas_int n,
|
|
Td& dA,
|
|
const rocblas_int lda,
|
|
const rocblas_int bc,
|
|
Th& hA,
|
|
std::vector<T>& A,
|
|
const bool test,
|
|
const bool singular)
|
|
{
|
|
if(CPU)
|
|
{
|
|
rocblas_init<T>(hA, true);
|
|
|
|
for(rocblas_int b = 0; b < bc; ++b)
|
|
{
|
|
if(!singular)
|
|
{
|
|
// scale A to avoid singularities
|
|
for(rocblas_int i = 0; i < m; i++)
|
|
{
|
|
for(rocblas_int j = 0; j < n; j++)
|
|
{
|
|
if(i == j)
|
|
hA[b][i + j * lda] += 400;
|
|
else
|
|
hA[b][i + j * lda] -= 4;
|
|
}
|
|
}
|
|
}
|
|
else
|
|
{
|
|
// form a singular matrix consisting of all ones
|
|
for(rocblas_int i = 0; i < m; i++)
|
|
{
|
|
for(rocblas_int j = 0; j < n; j++)
|
|
{
|
|
hA[b][i + j * lda] = 1;
|
|
}
|
|
}
|
|
}
|
|
|
|
// make copy of original data to test vectors if required
|
|
if(test && (left_svect != rocblas_svect_none || right_svect != rocblas_svect_none))
|
|
{
|
|
for(rocblas_int i = 0; i < m; i++)
|
|
{
|
|
for(rocblas_int j = 0; j < n; j++)
|
|
A[b * lda * n + i + j * lda] = hA[b][i + j * lda];
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
if(GPU)
|
|
{
|
|
// now copy to the GPU
|
|
CHECK_HIP_ERROR(dA.transfer_from(hA));
|
|
}
|
|
}
|
|
|
|
template <bool STRIDED, typename T, typename Wd, typename Td, typename Ud, typename Id, typename Wh, typename Th, typename Uh, typename Ih>
|
|
void gesvd_getError(const rocblas_handle handle,
|
|
const rocblas_svect left_svect,
|
|
const rocblas_svect right_svect,
|
|
const rocblas_int m,
|
|
const rocblas_int n,
|
|
Wd& dA,
|
|
const rocblas_int lda,
|
|
const rocblas_stride stA,
|
|
Td& dS,
|
|
const rocblas_stride stS,
|
|
Ud& dU,
|
|
const rocblas_int ldu,
|
|
const rocblas_stride stU,
|
|
Ud& dV,
|
|
const rocblas_int ldv,
|
|
const rocblas_stride stV,
|
|
Td& dE,
|
|
const rocblas_stride stE,
|
|
const rocblas_workmode fa,
|
|
Id& dinfo,
|
|
const rocblas_int bc,
|
|
const rocblas_svect left_svectT,
|
|
const rocblas_svect right_svectT,
|
|
const rocblas_int mT,
|
|
const rocblas_int nT,
|
|
Ud& dUT,
|
|
const rocblas_int lduT,
|
|
const rocblas_stride stUT,
|
|
Ud& dVT,
|
|
const rocblas_int ldvT,
|
|
const rocblas_stride stVT,
|
|
Wh& hA,
|
|
Th& hS,
|
|
Th& hSres,
|
|
Uh& hU,
|
|
Uh& Ures,
|
|
const rocblas_int ldures,
|
|
Uh& hV,
|
|
Uh& Vres,
|
|
const rocblas_int ldvres,
|
|
Ih& hinfo,
|
|
Ih& hinfoRes,
|
|
double* max_err,
|
|
double* max_errv,
|
|
const bool singular)
|
|
{
|
|
using W = decltype(std::real(T{}));
|
|
|
|
rocblas_int lwork = 5 * std::max(m, n);
|
|
rocblas_int lrwork = (rocblas_is_complex<T> ? 5 * std::min(m, n) : 0);
|
|
std::vector<T> work(lwork);
|
|
std::vector<W> rwork(lrwork);
|
|
std::vector<T> A(lda * n * bc);
|
|
|
|
// input data initialization
|
|
gesvd_initData<true, true, T>(handle, left_svect, right_svect, m, n, dA, lda, bc, hA, A, true,
|
|
singular);
|
|
|
|
// execute computations:
|
|
// complementary execution to compute all singular vectors if needed (always in-place to ensure
|
|
// we don't combine results computed by gemm_batched with results computed by gemm_strided_batched)
|
|
CHECK_ROCBLAS_ERROR(rocsolver_gesvd(STRIDED, handle, left_svectT, right_svectT, mT, nT,
|
|
dA.data(), lda, stA, dS.data(), stS, dUT.data(), lduT, stUT,
|
|
dVT.data(), ldvT, stVT, dE.data(), stE, rocblas_inplace,
|
|
dinfo.data(), bc));
|
|
|
|
if(left_svect == rocblas_svect_none && right_svect != rocblas_svect_none)
|
|
CHECK_HIP_ERROR(Ures.transfer_from(dUT));
|
|
if(right_svect == rocblas_svect_none && left_svect != rocblas_svect_none)
|
|
CHECK_HIP_ERROR(Vres.transfer_from(dVT));
|
|
|
|
gesvd_initData<false, true, T>(handle, left_svect, right_svect, m, n, dA, lda, bc, hA, A, true,
|
|
singular);
|
|
|
|
// CPU lapack
|
|
for(rocblas_int b = 0; b < bc; ++b)
|
|
cpu_gesvd(rocblas_svect_none, rocblas_svect_none, m, n, hA[b], lda, hS[b], hU[b], ldu,
|
|
hV[b], ldv, work.data(), lwork, rwork.data(), hinfo[b]);
|
|
|
|
// GPU lapack
|
|
CHECK_ROCBLAS_ERROR(rocsolver_gesvd(STRIDED, handle, left_svect, right_svect, m, n, dA.data(),
|
|
lda, stA, dS.data(), stS, dU.data(), ldu, stU, dV.data(),
|
|
ldv, stV, dE.data(), stE, fa, dinfo.data(), bc));
|
|
|
|
CHECK_HIP_ERROR(hSres.transfer_from(dS));
|
|
CHECK_HIP_ERROR(hinfoRes.transfer_from(dinfo));
|
|
|
|
if(left_svect == rocblas_svect_singular || left_svect == rocblas_svect_all)
|
|
CHECK_HIP_ERROR(Ures.transfer_from(dU));
|
|
if(right_svect == rocblas_svect_singular || right_svect == rocblas_svect_all)
|
|
CHECK_HIP_ERROR(Vres.transfer_from(dV));
|
|
|
|
if(left_svect == rocblas_svect_overwrite)
|
|
{
|
|
CHECK_HIP_ERROR(hA.transfer_from(dA));
|
|
for(rocblas_int b = 0; b < bc; ++b)
|
|
{
|
|
for(rocblas_int i = 0; i < m; i++)
|
|
{
|
|
for(rocblas_int j = 0; j < std::min(m, n); j++)
|
|
Ures[b][i + j * ldures] = hA[b][i + j * lda];
|
|
}
|
|
}
|
|
}
|
|
if(right_svect == rocblas_svect_overwrite)
|
|
{
|
|
CHECK_HIP_ERROR(hA.transfer_from(dA));
|
|
for(rocblas_int b = 0; b < bc; ++b)
|
|
{
|
|
for(rocblas_int i = 0; i < std::min(m, n); i++)
|
|
{
|
|
for(rocblas_int j = 0; j < n; j++)
|
|
Vres[b][i + j * ldvres] = hA[b][i + j * lda];
|
|
}
|
|
}
|
|
}
|
|
|
|
// Check info for non-convergence
|
|
*max_err = 0;
|
|
for(rocblas_int b = 0; b < bc; ++b)
|
|
{
|
|
EXPECT_EQ(hinfo[b][0], hinfoRes[b][0]) << "where b = " << b;
|
|
if(hinfo[b][0] != hinfoRes[b][0])
|
|
*max_err += 1;
|
|
}
|
|
|
|
// (We expect the used input matrices to always converge. Testing
|
|
// implicitly the equivalent non-converged matrix is very complicated and it boils
|
|
// down to essentially run the algorithm again and until convergence is achieved).
|
|
|
|
double err;
|
|
*max_errv = 0;
|
|
|
|
for(rocblas_int b = 0; b < bc; ++b)
|
|
{
|
|
// error is ||hS - hSres||
|
|
err = norm_error('F', 1, std::min(m, n), 1, hS[b], hSres[b]);
|
|
*max_err = err > *max_err ? err : *max_err;
|
|
|
|
// Check the singular vectors if required
|
|
if(hinfo[b][0] == 0 && (left_svect != rocblas_svect_none || right_svect != rocblas_svect_none))
|
|
{
|
|
err = 0;
|
|
// check singular vectors implicitly (A*v_k = s_k*u_k)
|
|
for(rocblas_int k = 0; k < std::min(m, n); ++k)
|
|
{
|
|
for(rocblas_int i = 0; i < m; ++i)
|
|
{
|
|
T tmp = 0;
|
|
for(rocblas_int j = 0; j < n; ++j)
|
|
tmp += A[b * lda * n + i + j * lda] * sconj(Vres[b][k + j * ldvres]);
|
|
tmp -= hSres[b][k] * Ures[b][i + k * ldures];
|
|
err += std::abs(tmp) * std::abs(tmp);
|
|
}
|
|
}
|
|
err = std::sqrt(err) / double(snorm('F', m, n, A.data() + b * lda * n, lda));
|
|
*max_errv = err > *max_errv ? err : *max_errv;
|
|
}
|
|
}
|
|
}
|
|
|
|
template <bool STRIDED, typename T, typename Wd, typename Td, typename Ud, typename Id, typename Wh, typename Th, typename Uh, typename Ih>
|
|
void gesvd_getPerfData(const rocblas_handle handle,
|
|
const rocblas_svect left_svect,
|
|
const rocblas_svect right_svect,
|
|
const rocblas_int m,
|
|
const rocblas_int n,
|
|
Wd& dA,
|
|
const rocblas_int lda,
|
|
const rocblas_stride stA,
|
|
Td& dS,
|
|
const rocblas_stride stS,
|
|
Ud& dU,
|
|
const rocblas_int ldu,
|
|
const rocblas_stride stU,
|
|
Ud& dV,
|
|
const rocblas_int ldv,
|
|
const rocblas_stride stV,
|
|
Td& dE,
|
|
const rocblas_stride stE,
|
|
const rocblas_workmode fa,
|
|
Id& dinfo,
|
|
const rocblas_int bc,
|
|
Wh& hA,
|
|
Th& hS,
|
|
Uh& hU,
|
|
Uh& hV,
|
|
Ih& hinfo,
|
|
double* gpu_time_used,
|
|
double* cpu_time_used,
|
|
const rocblas_int hot_calls,
|
|
const int profile,
|
|
const bool profile_kernels,
|
|
const bool perf,
|
|
const bool singular)
|
|
{
|
|
using W = decltype(std::real(T{}));
|
|
|
|
rocblas_int lwork = 5 * std::max(m, n);
|
|
rocblas_int lrwork = (rocblas_is_complex<T> ? 5 * std::min(m, n) : 0);
|
|
std::vector<T> work(lwork);
|
|
std::vector<W> rwork(lrwork);
|
|
std::vector<T> A;
|
|
|
|
if(!perf)
|
|
{
|
|
gesvd_initData<true, false, T>(handle, left_svect, right_svect, m, n, dA, lda, bc, hA, A,
|
|
false, singular);
|
|
|
|
// cpu-lapack performance (only if not in perf mode)
|
|
*cpu_time_used = get_time_us_no_sync();
|
|
for(rocblas_int b = 0; b < bc; ++b)
|
|
cpu_gesvd(left_svect, right_svect, m, n, hA[b], lda, hS[b], hU[b], ldu, hV[b], ldv,
|
|
work.data(), lwork, rwork.data(), hinfo[b]);
|
|
*cpu_time_used = get_time_us_no_sync() - *cpu_time_used;
|
|
}
|
|
|
|
gesvd_initData<true, false, T>(handle, left_svect, right_svect, m, n, dA, lda, bc, hA, A, false,
|
|
singular);
|
|
|
|
// cold calls
|
|
for(int iter = 0; iter < 2; iter++)
|
|
{
|
|
gesvd_initData<false, true, T>(handle, left_svect, right_svect, m, n, dA, lda, bc, hA, A,
|
|
false, singular);
|
|
|
|
CHECK_ROCBLAS_ERROR(rocsolver_gesvd(
|
|
STRIDED, handle, left_svect, right_svect, m, n, dA.data(), lda, stA, dS.data(), stS,
|
|
dU.data(), ldu, stU, dV.data(), ldv, stV, dE.data(), stE, fa, dinfo.data(), bc));
|
|
}
|
|
|
|
// gpu-lapack performance
|
|
hipStream_t stream;
|
|
CHECK_ROCBLAS_ERROR(rocblas_get_stream(handle, &stream));
|
|
rocsolver_timer timer;
|
|
|
|
if(profile > 0)
|
|
{
|
|
if(profile_kernels)
|
|
rocsolver_log_set_layer_mode(rocblas_layer_mode_log_profile
|
|
| rocblas_layer_mode_ex_log_kernel);
|
|
else
|
|
rocsolver_log_set_layer_mode(rocblas_layer_mode_log_profile);
|
|
rocsolver_log_set_max_levels(profile);
|
|
}
|
|
|
|
for(rocblas_int iter = 0; iter < hot_calls; iter++)
|
|
{
|
|
gesvd_initData<false, true, T>(handle, left_svect, right_svect, m, n, dA, lda, bc, hA, A,
|
|
false, singular);
|
|
|
|
timer.start(stream);
|
|
rocsolver_gesvd(STRIDED, handle, left_svect, right_svect, m, n, dA.data(), lda, stA,
|
|
dS.data(), stS, dU.data(), ldu, stU, dV.data(), ldv, stV, dE.data(), stE,
|
|
fa, dinfo.data(), bc);
|
|
timer.end(stream);
|
|
}
|
|
*gpu_time_used = timer.get_combined();
|
|
}
|
|
|
|
template <bool BATCHED, bool STRIDED, typename T>
|
|
void testing_gesvd(Arguments& argus)
|
|
{
|
|
using S = decltype(std::real(T{}));
|
|
|
|
// get arguments
|
|
rocblas_local_handle handle;
|
|
char leftvC = argus.get<char>("left_svect");
|
|
char rightvC = argus.get<char>("right_svect");
|
|
rocblas_int m = argus.get<rocblas_int>("m");
|
|
rocblas_int n = argus.get<rocblas_int>("n", m);
|
|
rocblas_int lda = argus.get<rocblas_int>("lda", m);
|
|
rocblas_int ldu = argus.get<rocblas_int>("ldu", m);
|
|
rocblas_int ldv = argus.get<rocblas_int>("ldv", (rightvC == 'A' ? n : std::min(m, n)));
|
|
rocblas_stride stA = argus.get<rocblas_stride>("strideA", lda * n);
|
|
rocblas_stride stS = argus.get<rocblas_stride>("strideS", std::min(m, n));
|
|
rocblas_stride stU = argus.get<rocblas_stride>("strideU", ldu * m);
|
|
rocblas_stride stV = argus.get<rocblas_stride>("strideV", ldv * n);
|
|
rocblas_stride stE = argus.get<rocblas_stride>("strideE", std::min(m, n) - 1);
|
|
char faC = argus.get<char>("fast_alg");
|
|
|
|
rocblas_svect leftv = char2rocblas_svect(leftvC);
|
|
rocblas_svect rightv = char2rocblas_svect(rightvC);
|
|
rocblas_workmode fa = char2rocblas_workmode(faC);
|
|
rocblas_int bc = argus.batch_count;
|
|
rocblas_int hot_calls = argus.iters;
|
|
|
|
if(argus.alg_mode == 1)
|
|
{
|
|
EXPECT_ROCBLAS_STATUS(
|
|
rocsolver_set_alg_mode(handle, rocsolver_function_gesvd, rocsolver_alg_mode_hybrid),
|
|
rocblas_status_success);
|
|
|
|
rocsolver_alg_mode alg_mode;
|
|
EXPECT_ROCBLAS_STATUS(rocsolver_get_alg_mode(handle, rocsolver_function_gesvd, &alg_mode),
|
|
rocblas_status_success);
|
|
|
|
EXPECT_EQ(alg_mode, rocsolver_alg_mode_hybrid);
|
|
}
|
|
|
|
// check non-supported values
|
|
if(rightv == rocblas_svect_overwrite && leftv == rocblas_svect_overwrite)
|
|
{
|
|
if(BATCHED)
|
|
EXPECT_ROCBLAS_STATUS(rocsolver_gesvd(STRIDED, handle, leftv, rightv, m, n,
|
|
(T* const*)nullptr, lda, stA, (S*)nullptr, stS,
|
|
(T*)nullptr, ldu, stU, (T*)nullptr, ldv, stV,
|
|
(S*)nullptr, stE, fa, (rocblas_int*)nullptr, bc),
|
|
rocblas_status_invalid_value);
|
|
else
|
|
EXPECT_ROCBLAS_STATUS(rocsolver_gesvd(STRIDED, handle, leftv, rightv, m, n, (T*)nullptr,
|
|
lda, stA, (S*)nullptr, stS, (T*)nullptr, ldu, stU,
|
|
(T*)nullptr, ldv, stV, (S*)nullptr, stE, fa,
|
|
(rocblas_int*)nullptr, bc),
|
|
rocblas_status_invalid_value);
|
|
|
|
if(argus.timing)
|
|
rocsolver_bench_inform(inform_invalid_args);
|
|
|
|
return;
|
|
}
|
|
|
|
/** TESTING OF SINGULAR VECTORS IS DONE IMPLICITLY, NOT EXPLICITLY COMPARING
|
|
WITH LAPACK. SO, WE ALWAYS NEED TO COMPUTE THE SAME NUMBER OF ELEMENTS OF
|
|
THE RIGHT AND LEFT VECTORS. WHILE DOING THIS, IF MORE VECTORS THAN THE
|
|
SPECIFIED IN THE MAIN CALL NEED TO BE COMPUTED, WE DO SO WITH AN EXTRA CALL **/
|
|
|
|
rocblas_svect leftvT = rocblas_svect_none;
|
|
rocblas_svect rightvT = rocblas_svect_none;
|
|
rocblas_int ldvT = 1;
|
|
rocblas_int lduT = 1;
|
|
rocblas_int mT = 0;
|
|
rocblas_int nT = 0;
|
|
bool svects = (leftv != rocblas_svect_none || rightv != rocblas_svect_none);
|
|
|
|
if(svects)
|
|
{
|
|
if(leftv == rocblas_svect_none)
|
|
{
|
|
leftvT = rocblas_svect_all;
|
|
lduT = m;
|
|
mT = m;
|
|
nT = n;
|
|
if((n > m && fa == rocblas_outofplace) || (n > m && rightv == rocblas_svect_overwrite))
|
|
rightvT = rocblas_svect_overwrite;
|
|
}
|
|
if(rightv == rocblas_svect_none)
|
|
{
|
|
rightvT = rocblas_svect_all;
|
|
ldvT = n;
|
|
mT = m;
|
|
nT = n;
|
|
if((m >= n && fa == rocblas_outofplace) || (m >= n && leftv == rocblas_svect_overwrite))
|
|
leftvT = rocblas_svect_overwrite;
|
|
}
|
|
}
|
|
|
|
// determine sizes
|
|
rocblas_int ldures = 1;
|
|
rocblas_int ldvres = 1;
|
|
size_t size_Sres = 0;
|
|
size_t size_Ures = 0;
|
|
size_t size_Vres = 0;
|
|
size_t size_UT = 0;
|
|
size_t size_VT = 0;
|
|
size_t size_A = size_t(lda) * n;
|
|
size_t size_S = size_t(std::min(m, n));
|
|
size_t size_E = size_t(std::min(m, n) - 1);
|
|
size_t size_V = size_t(ldv) * n;
|
|
size_t size_U = size_t(ldu) * m;
|
|
if(argus.unit_check || argus.norm_check)
|
|
{
|
|
size_VT = size_t(ldvT) * nT;
|
|
size_UT = size_t(lduT) * mT;
|
|
size_Sres = size_S;
|
|
if(svects)
|
|
{
|
|
if(leftv == rocblas_svect_none)
|
|
{
|
|
size_Ures = size_UT;
|
|
ldures = lduT;
|
|
}
|
|
else if(leftv == rocblas_svect_singular || leftv == rocblas_svect_all)
|
|
{
|
|
size_Ures = size_U;
|
|
ldures = ldu;
|
|
}
|
|
else
|
|
{
|
|
size_Ures = m * m;
|
|
ldures = m;
|
|
}
|
|
|
|
if(rightv == rocblas_svect_none)
|
|
{
|
|
size_Vres = size_VT;
|
|
ldvres = ldvT;
|
|
}
|
|
else if(rightv == rocblas_svect_singular || rightv == rocblas_svect_all)
|
|
{
|
|
size_Vres = size_V;
|
|
ldvres = ldv;
|
|
}
|
|
else
|
|
{
|
|
size_Vres = n * n;
|
|
ldvres = n;
|
|
}
|
|
}
|
|
}
|
|
rocblas_stride stUT = size_UT;
|
|
rocblas_stride stVT = size_VT;
|
|
rocblas_stride stUres = size_Ures;
|
|
rocblas_stride stVres = size_Vres;
|
|
|
|
double max_error = 0, gpu_time_used = 0, cpu_time_used = 0, max_errorv = 0;
|
|
|
|
// check invalid sizes
|
|
bool invalid_size = (n < 0 || m < 0 || lda < m || ldu < 1 || ldv < 1 || bc < 0)
|
|
|| ((leftv == rocblas_svect_all || leftv == rocblas_svect_singular) && ldu < m)
|
|
|| ((rightv == rocblas_svect_all && ldv < n)
|
|
|| (rightv == rocblas_svect_singular && ldv < std::min(m, n)));
|
|
|
|
if(invalid_size)
|
|
{
|
|
if(BATCHED)
|
|
EXPECT_ROCBLAS_STATUS(rocsolver_gesvd(STRIDED, handle, leftv, rightv, m, n,
|
|
(T* const*)nullptr, lda, stA, (S*)nullptr, stS,
|
|
(T*)nullptr, ldu, stU, (T*)nullptr, ldv, stV,
|
|
(S*)nullptr, stE, fa, (rocblas_int*)nullptr, bc),
|
|
rocblas_status_invalid_size);
|
|
else
|
|
EXPECT_ROCBLAS_STATUS(rocsolver_gesvd(STRIDED, handle, leftv, rightv, m, n, (T*)nullptr,
|
|
lda, stA, (S*)nullptr, stS, (T*)nullptr, ldu, stU,
|
|
(T*)nullptr, ldv, stV, (S*)nullptr, stE, fa,
|
|
(rocblas_int*)nullptr, bc),
|
|
rocblas_status_invalid_size);
|
|
|
|
if(argus.timing)
|
|
rocsolver_bench_inform(inform_invalid_size);
|
|
|
|
return;
|
|
}
|
|
|
|
// memory size query is necessary
|
|
if(argus.mem_query)
|
|
{
|
|
CHECK_ROCBLAS_ERROR(rocblas_start_device_memory_size_query(handle));
|
|
if(BATCHED)
|
|
{
|
|
CHECK_ALLOC_QUERY(rocsolver_gesvd(STRIDED, handle, leftv, rightv, m, n,
|
|
(T* const*)nullptr, lda, stA, (S*)nullptr, stS,
|
|
(T*)nullptr, ldu, stU, (T*)nullptr, ldv, stV,
|
|
(S*)nullptr, stE, fa, (rocblas_int*)nullptr, bc));
|
|
CHECK_ALLOC_QUERY(rocsolver_gesvd(STRIDED, handle, leftvT, rightvT, mT, nT,
|
|
(T* const*)nullptr, lda, stA, (S*)nullptr, stS,
|
|
(T*)nullptr, lduT, stUT, (T*)nullptr, ldvT, stVT,
|
|
(S*)nullptr, stE, fa, (rocblas_int*)nullptr, bc));
|
|
}
|
|
else
|
|
{
|
|
CHECK_ALLOC_QUERY(rocsolver_gesvd(STRIDED, handle, leftv, rightv, m, n, (T*)nullptr,
|
|
lda, stA, (S*)nullptr, stS, (T*)nullptr, ldu, stU,
|
|
(T*)nullptr, ldv, stV, (S*)nullptr, stE, fa,
|
|
(rocblas_int*)nullptr, bc));
|
|
CHECK_ALLOC_QUERY(rocsolver_gesvd(STRIDED, handle, leftvT, rightvT, mT, nT, (T*)nullptr,
|
|
lda, stA, (S*)nullptr, stS, (T*)nullptr, lduT, stUT,
|
|
(T*)nullptr, ldvT, stVT, (S*)nullptr, stE, fa,
|
|
(rocblas_int*)nullptr, bc));
|
|
}
|
|
|
|
size_t size;
|
|
CHECK_ROCBLAS_ERROR(rocblas_stop_device_memory_size_query(handle, &size));
|
|
|
|
rocsolver_bench_inform(inform_mem_query, size);
|
|
return;
|
|
}
|
|
|
|
// memory allocations (all cases)
|
|
// host
|
|
host_strided_batch_vector<S> hS(size_S, 1, stS, bc);
|
|
host_strided_batch_vector<T> hV(size_V, 1, stV, bc);
|
|
host_strided_batch_vector<T> hU(size_U, 1, stU, bc);
|
|
host_strided_batch_vector<rocblas_int> hinfo(1, 1, 1, bc);
|
|
host_strided_batch_vector<rocblas_int> hinfoRes(1, 1, 1, bc);
|
|
host_strided_batch_vector<S> hSres(size_Sres, 1, stS, bc);
|
|
host_strided_batch_vector<T> Vres(size_Vres, 1, stVres, bc);
|
|
host_strided_batch_vector<T> Ures(size_Ures, 1, stUres, bc);
|
|
// device
|
|
device_strided_batch_vector<S> dE(size_E, 1, stE, bc);
|
|
device_strided_batch_vector<S> dS(size_S, 1, stS, bc);
|
|
device_strided_batch_vector<T> dV(size_V, 1, stV, bc);
|
|
device_strided_batch_vector<T> dU(size_U, 1, stU, bc);
|
|
device_strided_batch_vector<rocblas_int> dinfo(1, 1, 1, bc);
|
|
device_strided_batch_vector<T> dVT(size_VT, 1, stVT, bc);
|
|
device_strided_batch_vector<T> dUT(size_UT, 1, stUT, bc);
|
|
if(size_VT)
|
|
CHECK_HIP_ERROR(dVT.memcheck());
|
|
if(size_UT)
|
|
CHECK_HIP_ERROR(dUT.memcheck());
|
|
if(size_E)
|
|
CHECK_HIP_ERROR(dE.memcheck());
|
|
if(size_S)
|
|
CHECK_HIP_ERROR(dS.memcheck());
|
|
if(size_V)
|
|
CHECK_HIP_ERROR(dV.memcheck());
|
|
if(size_U)
|
|
CHECK_HIP_ERROR(dU.memcheck());
|
|
CHECK_HIP_ERROR(dinfo.memcheck());
|
|
|
|
if(BATCHED)
|
|
{
|
|
// memory allocations
|
|
host_batch_vector<T> hA(size_A, 1, bc);
|
|
device_batch_vector<T> dA(size_A, 1, bc);
|
|
if(size_A)
|
|
CHECK_HIP_ERROR(dA.memcheck());
|
|
|
|
// check quick return
|
|
if(n == 0 || m == 0 || bc == 0)
|
|
{
|
|
EXPECT_ROCBLAS_STATUS(rocsolver_gesvd(STRIDED, handle, leftv, rightv, m, n, dA.data(),
|
|
lda, stA, dS.data(), stS, dU.data(), ldu, stU,
|
|
dV.data(), ldv, stV, dE.data(), stE, fa,
|
|
dinfo.data(), bc),
|
|
rocblas_status_success);
|
|
if(argus.timing)
|
|
rocsolver_bench_inform(inform_quick_return);
|
|
|
|
return;
|
|
}
|
|
|
|
// check computations
|
|
if(argus.unit_check || argus.norm_check)
|
|
{
|
|
gesvd_getError<STRIDED, T>(handle, leftv, rightv, m, n, dA, lda, stA, dS, stS, dU, ldu,
|
|
stU, dV, ldv, stV, dE, stE, fa, dinfo, bc, leftvT, rightvT,
|
|
mT, nT, dUT, lduT, stUT, dVT, ldvT, stVT, hA, hS, hSres, hU,
|
|
Ures, ldures, hV, Vres, ldvres, hinfo, hinfoRes, &max_error,
|
|
&max_errorv, argus.singular);
|
|
}
|
|
|
|
// collect performance data
|
|
if(argus.timing && hot_calls > 0)
|
|
{
|
|
gesvd_getPerfData<STRIDED, T>(
|
|
handle, leftv, rightv, m, n, dA, lda, stA, dS, stS, dU, ldu, stU, dV, ldv, stV, dE,
|
|
stE, fa, dinfo, bc, hA, hS, hU, hV, hinfo, &gpu_time_used, &cpu_time_used,
|
|
hot_calls, argus.profile, argus.profile_kernels, argus.perf, argus.singular);
|
|
}
|
|
}
|
|
|
|
else
|
|
{
|
|
// memory allocations
|
|
host_strided_batch_vector<T> hA(size_A, 1, stA, bc);
|
|
device_strided_batch_vector<T> dA(size_A, 1, stA, bc);
|
|
if(size_A)
|
|
CHECK_HIP_ERROR(dA.memcheck());
|
|
|
|
// check quick return
|
|
if(n == 0 || m == 0 || bc == 0)
|
|
{
|
|
EXPECT_ROCBLAS_STATUS(rocsolver_gesvd(STRIDED, handle, leftv, rightv, m, n, dA.data(),
|
|
lda, stA, dS.data(), stS, dU.data(), ldu, stU,
|
|
dV.data(), ldv, stV, dE.data(), stE, fa,
|
|
dinfo.data(), bc),
|
|
rocblas_status_success);
|
|
if(argus.timing)
|
|
rocsolver_bench_inform(inform_quick_return);
|
|
|
|
return;
|
|
}
|
|
|
|
// check computations
|
|
if(argus.unit_check || argus.norm_check)
|
|
{
|
|
gesvd_getError<STRIDED, T>(handle, leftv, rightv, m, n, dA, lda, stA, dS, stS, dU, ldu,
|
|
stU, dV, ldv, stV, dE, stE, fa, dinfo, bc, leftvT, rightvT,
|
|
mT, nT, dUT, lduT, stUT, dVT, ldvT, stVT, hA, hS, hSres, hU,
|
|
Ures, ldures, hV, Vres, ldvres, hinfo, hinfoRes, &max_error,
|
|
&max_errorv, argus.singular);
|
|
}
|
|
|
|
// collect performance data
|
|
if(argus.timing && hot_calls > 0)
|
|
{
|
|
gesvd_getPerfData<STRIDED, T>(
|
|
handle, leftv, rightv, m, n, dA, lda, stA, dS, stS, dU, ldu, stU, dV, ldv, stV, dE,
|
|
stE, fa, dinfo, bc, hA, hS, hU, hV, hinfo, &gpu_time_used, &cpu_time_used,
|
|
hot_calls, argus.profile, argus.profile_kernels, argus.perf, argus.singular);
|
|
}
|
|
}
|
|
|
|
// validate results for rocsolver-test
|
|
// using 2 * min(m, n) * machine_precision as tolerance
|
|
if(argus.unit_check)
|
|
{
|
|
ROCSOLVER_TEST_CHECK(T, max_error, 2 * std::min(m, n));
|
|
if(svects)
|
|
ROCSOLVER_TEST_CHECK(T, max_errorv, 2 * std::min(m, n));
|
|
}
|
|
|
|
// output results for rocsolver-bench
|
|
if(argus.timing)
|
|
{
|
|
if(svects)
|
|
max_error = (max_error >= max_errorv) ? max_error : max_errorv;
|
|
|
|
if(!argus.perf)
|
|
{
|
|
rocsolver_bench_header("Arguments:");
|
|
if(BATCHED)
|
|
{
|
|
rocsolver_bench_output("left_svect", "right_svect", "m", "n", "lda", "strideS",
|
|
"ldu", "strideU", "ldv", "strideV", "strideE", "batch_c");
|
|
rocsolver_bench_output(leftvC, rightvC, m, n, lda, stS, ldu, stU, ldv, stV, stE, bc);
|
|
}
|
|
else if(STRIDED)
|
|
{
|
|
rocsolver_bench_output("left_svect", "right_svect", "m", "n", "lda", "strideA",
|
|
"strideS", "ldu", "strideU", "ldv", "strideV", "strideE",
|
|
"batch_c");
|
|
rocsolver_bench_output(leftvC, rightvC, m, n, lda, stA, stS, ldu, stU, ldv, stV,
|
|
stE, bc);
|
|
}
|
|
else
|
|
{
|
|
rocsolver_bench_output("left_svect", "right_svect", "m", "n", "lda", "ldu", "ldv");
|
|
rocsolver_bench_output(leftvC, rightvC, m, n, lda, ldu, ldv);
|
|
}
|
|
rocsolver_bench_header("Results:");
|
|
if(argus.norm_check)
|
|
{
|
|
rocsolver_bench_output("cpu_time_us", "gpu_time_us", "error");
|
|
rocsolver_bench_output(cpu_time_used, gpu_time_used, max_error);
|
|
}
|
|
else
|
|
{
|
|
rocsolver_bench_output("cpu_time_us", "gpu_time_us");
|
|
rocsolver_bench_output(cpu_time_used, gpu_time_used);
|
|
}
|
|
rocsolver_bench_endl();
|
|
}
|
|
else
|
|
{
|
|
if(argus.norm_check)
|
|
rocsolver_bench_output(gpu_time_used, max_error);
|
|
else
|
|
rocsolver_bench_output(gpu_time_used);
|
|
}
|
|
}
|
|
|
|
// ensure all arguments were consumed
|
|
argus.validate_consumed();
|
|
}
|
|
|
|
#define EXTERN_TESTING_GESVD(...) extern template void testing_gesvd<__VA_ARGS__>(Arguments&);
|
|
|
|
INSTANTIATE(EXTERN_TESTING_GESVD, FOREACH_MATRIX_DATA_LAYOUT, FOREACH_SCALAR_TYPE, APPLY_STAMP)
|