/* ************************************************************************ * Copyright 2020-2021 Advanced Micro Devices, Inc. * ************************************************************************ */ #pragma once #include "clientcommon.hpp" template void ormqr_unmqr_checkBadArgs(const hipsolverHandle_t handle, const hipsolverSideMode_t side, const hipsolverOperation_t trans, const int m, const int n, const int k, T dA, const int lda, T dIpiv, T dC, const int ldc, T dWork, const int lwork, U dInfo) { // handle EXPECT_ROCBLAS_STATUS( hipsolver_ormqr_unmqr( FORTRAN, nullptr, side, trans, m, n, k, dA, lda, dIpiv, dC, ldc, dWork, lwork, dInfo), HIPSOLVER_STATUS_NOT_INITIALIZED); // values EXPECT_ROCBLAS_STATUS(hipsolver_ormqr_unmqr(FORTRAN, handle, hipsolverSideMode_t(-1), trans, m, n, k, dA, lda, dIpiv, dC, ldc, dWork, lwork, dInfo), HIPSOLVER_STATUS_INVALID_ENUM); EXPECT_ROCBLAS_STATUS(hipsolver_ormqr_unmqr(FORTRAN, handle, side, hipsolverOperation_t(-1), m, n, k, dA, lda, dIpiv, dC, ldc, dWork, lwork, dInfo), HIPSOLVER_STATUS_INVALID_ENUM); if(COMPLEX) EXPECT_ROCBLAS_STATUS(hipsolver_ormqr_unmqr(FORTRAN, handle, side, HIPSOLVER_OP_T, m, n, k, dA, lda, dIpiv, dC, ldc, dWork, lwork, dInfo), HIPSOLVER_STATUS_INVALID_VALUE); else EXPECT_ROCBLAS_STATUS(hipsolver_ormqr_unmqr(FORTRAN, handle, side, HIPSOLVER_OP_C, m, n, k, dA, lda, dIpiv, dC, ldc, dWork, lwork, dInfo), HIPSOLVER_STATUS_INVALID_VALUE); #if defined(__HIP_PLATFORM_HCC__) || defined(__HIP_PLATFORM_AMD__) // pointers EXPECT_ROCBLAS_STATUS(hipsolver_ormqr_unmqr(FORTRAN, handle, side, trans, m, n, k, (T) nullptr, lda, dIpiv, dC, ldc, dWork, lwork, dInfo), HIPSOLVER_STATUS_INVALID_VALUE); EXPECT_ROCBLAS_STATUS(hipsolver_ormqr_unmqr(FORTRAN, handle, side, trans, m, n, k, dA, lda, (T) nullptr, dC, ldc, dWork, lwork, dInfo), HIPSOLVER_STATUS_INVALID_VALUE); EXPECT_ROCBLAS_STATUS(hipsolver_ormqr_unmqr(FORTRAN, handle, side, trans, m, n, k, dA, lda, dIpiv, (T) nullptr, ldc, dWork, lwork, dInfo), HIPSOLVER_STATUS_INVALID_VALUE); #endif } template > void testing_ormqr_unmqr_bad_arg() { // safe arguments hipsolver_local_handle handle; hipsolverSideMode_t side = HIPSOLVER_SIDE_LEFT; hipsolverOperation_t trans = HIPSOLVER_OP_N; int k = 1; int m = 1; int n = 1; int lda = 1; int ldc = 1; // memory allocation device_strided_batch_vector dA(1, 1, 1, 1); device_strided_batch_vector dIpiv(1, 1, 1, 1); device_strided_batch_vector dC(1, 1, 1, 1); device_strided_batch_vector dInfo(1, 1, 1, 1); CHECK_HIP_ERROR(dA.memcheck()); CHECK_HIP_ERROR(dIpiv.memcheck()); CHECK_HIP_ERROR(dC.memcheck()); CHECK_HIP_ERROR(dInfo.memcheck()); int size_W; hipsolver_ormqr_unmqr_bufferSize(FORTRAN, handle, side, trans, m, n, k, dA.data(), lda, dIpiv.data(), dC.data(), ldc, &size_W); device_strided_batch_vector dWork(size_W, 1, size_W, 1); if(size_W) CHECK_HIP_ERROR(dWork.memcheck()); // check bad arguments ormqr_unmqr_checkBadArgs(handle, side, trans, m, n, k, dA.data(), lda, dIpiv.data(), dC.data(), ldc, dWork.data(), size_W, dInfo.data()); } template void ormqr_unmqr_initData(const hipsolverHandle_t handle, const hipsolverSideMode_t side, const hipsolverOperation_t trans, const int m, const int n, const int k, Td& dA, const int lda, Td& dIpiv, Td& dC, const int ldc, Th& hA, Th& hIpiv, Th& hC, std::vector& hW, size_t size_W) { if(CPU) { int nq = (side == HIPSOLVER_SIDE_LEFT) ? m : n; rocblas_init(hA, true); rocblas_init(hIpiv, true); rocblas_init(hC, true); // scale to avoid singularities for(int i = 0; i < nq; ++i) { for(int j = 0; j < k; ++j) { if(i == j) hA[0][i + j * lda] += 400; else hA[0][i + j * lda] -= 4; } } // compute QR factorization cblas_geqrf(nq, k, hA[0], lda, hIpiv[0], hW.data(), size_W); } if(GPU) { // copy data from CPU to device CHECK_HIP_ERROR(dA.transfer_from(hA)); CHECK_HIP_ERROR(dIpiv.transfer_from(hIpiv)); CHECK_HIP_ERROR(dC.transfer_from(hC)); } } template void ormqr_unmqr_getError(const hipsolverHandle_t handle, const hipsolverSideMode_t side, const hipsolverOperation_t trans, const int m, const int n, const int k, Td& dA, const int lda, Td& dIpiv, Td& dC, const int ldc, Td& dWork, const int lwork, Ud& dInfo, Th& hA, Th& hIpiv, Th& hC, Th& hCr, Uh& hInfo, double* max_err) { size_t size_W = max(max(m, n), k); std::vector hW(size_W); // initialize data ormqr_unmqr_initData( handle, side, trans, m, n, k, dA, lda, dIpiv, dC, ldc, hA, hIpiv, hC, hW, size_W); // execute computations // GPU lapack CHECK_ROCBLAS_ERROR(hipsolver_ormqr_unmqr(FORTRAN, handle, side, trans, m, n, k, dA.data(), lda, dIpiv.data(), dC.data(), ldc, dWork.data(), lwork, dInfo.data())); CHECK_HIP_ERROR(hCr.transfer_from(dC)); // CPU lapack cblas_ormqr_unmqr(side, trans, m, n, k, hA[0], lda, hIpiv[0], hC[0], ldc, hW.data(), size_W); // error is ||hC - hCr|| / ||hC|| // (THIS DOES NOT ACCOUNT FOR NUMERICAL REPRODUCIBILITY ISSUES. // IT MIGHT BE REVISITED IN THE FUTURE) // using frobenius norm *max_err = norm_error('F', m, n, ldc, hC[0], hCr[0]); } template void ormqr_unmqr_getPerfData(const hipsolverHandle_t handle, const hipsolverSideMode_t side, const hipsolverOperation_t trans, const int m, const int n, const int k, Td& dA, const int lda, Td& dIpiv, Td& dC, const int ldc, Td& dWork, const int lwork, Ud& dInfo, Th& hA, Th& hIpiv, Th& hC, Uh& hInfo, double* gpu_time_used, double* cpu_time_used, const int hot_calls, const bool perf) { size_t size_W = max(max(m, n), k); std::vector hW(size_W); if(!perf) { ormqr_unmqr_initData( handle, side, trans, m, n, k, dA, lda, dIpiv, dC, ldc, hA, hIpiv, hC, hW, size_W); // cpu-lapack performance (only if not in perf mode) *cpu_time_used = get_time_us_no_sync(); cblas_ormqr_unmqr( side, trans, m, n, k, hA[0], lda, hIpiv[0], hC[0], ldc, hW.data(), size_W); *cpu_time_used = get_time_us_no_sync() - *cpu_time_used; } ormqr_unmqr_initData( handle, side, trans, m, n, k, dA, lda, dIpiv, dC, ldc, hA, hIpiv, hC, hW, size_W); // cold calls for(int iter = 0; iter < 2; iter++) { ormqr_unmqr_initData( handle, side, trans, m, n, k, dA, lda, dIpiv, dC, ldc, hA, hIpiv, hC, hW, size_W); CHECK_ROCBLAS_ERROR(hipsolver_ormqr_unmqr(FORTRAN, handle, side, trans, m, n, k, dA.data(), lda, dIpiv.data(), dC.data(), ldc, dWork.data(), lwork, dInfo.data())); } // gpu-lapack performance hipStream_t stream; CHECK_ROCBLAS_ERROR(hipsolverGetStream(handle, &stream)); double start; for(int iter = 0; iter < hot_calls; iter++) { ormqr_unmqr_initData( handle, side, trans, m, n, k, dA, lda, dIpiv, dC, ldc, hA, hIpiv, hC, hW, size_W); start = get_time_us_sync(stream); hipsolver_ormqr_unmqr(FORTRAN, handle, side, trans, m, n, k, dA.data(), lda, dIpiv.data(), dC.data(), ldc, dWork.data(), lwork, dInfo.data()); *gpu_time_used += get_time_us_sync(stream) - start; } *gpu_time_used /= hot_calls; } template > void testing_ormqr_unmqr(Arguments& argus) { // get arguments hipsolver_local_handle handle; char sideC = argus.get("side"); char transC = argus.get("trans"); int m, n, k; if(sideC == 'L') { m = argus.get("m"); n = argus.get("n", m); k = argus.get("k", m); } else { n = argus.get("n"); m = argus.get("m", n); k = argus.get("k", n); } int lda = argus.get("lda", sideC == 'L' ? m : n); int ldc = argus.get("ldc", m); hipsolverSideMode_t side = char2hipsolver_side(sideC); hipsolverOperation_t trans = char2hipsolver_operation(transC); int hot_calls = argus.iters; // check non-supported values bool invalid_value = ((COMPLEX && trans == HIPSOLVER_OP_T) || (!COMPLEX && trans == HIPSOLVER_OP_C)); if(invalid_value) { EXPECT_ROCBLAS_STATUS(hipsolver_ormqr_unmqr(FORTRAN, handle, side, trans, m, n, k, (T*)nullptr, lda, (T*)nullptr, (T*)nullptr, ldc, (T*)nullptr, 0, (int*)nullptr), HIPSOLVER_STATUS_INVALID_VALUE); if(argus.timing) ROCSOLVER_BENCH_INFORM(2); return; } // determine sizes bool left = (side == HIPSOLVER_SIDE_LEFT); size_t size_A = size_t(lda) * k; size_t size_P = size_t(k); size_t size_C = size_t(ldc) * n; double max_error = 0, gpu_time_used = 0, cpu_time_used = 0; size_t size_Cr = (argus.unit_check || argus.norm_check) ? size_C : 0; // check invalid sizes bool invalid_size = ((m < 0 || n < 0 || k < 0 || ldc < m) || (left && (lda < m || k > m)) || (!left && (lda < n || k > n))); if(invalid_size) { EXPECT_ROCBLAS_STATUS(hipsolver_ormqr_unmqr(FORTRAN, handle, side, trans, m, n, k, (T*)nullptr, lda, (T*)nullptr, (T*)nullptr, ldc, (T*)nullptr, 0, (int*)nullptr), HIPSOLVER_STATUS_INVALID_VALUE); if(argus.timing) ROCSOLVER_BENCH_INFORM(1); return; } // memory allocations host_strided_batch_vector hC(size_C, 1, size_C, 1); host_strided_batch_vector hCr(size_Cr, 1, size_Cr, 1); host_strided_batch_vector hIpiv(size_P, 1, size_P, 1); host_strided_batch_vector hA(size_A, 1, size_A, 1); host_strided_batch_vector hInfo(1, 1, 1, 1); device_strided_batch_vector dC(size_C, 1, size_C, 1); device_strided_batch_vector dIpiv(size_P, 1, size_P, 1); device_strided_batch_vector dA(size_A, 1, size_A, 1); device_strided_batch_vector dInfo(1, 1, 1, 1); if(size_A) CHECK_HIP_ERROR(dA.memcheck()); if(size_P) CHECK_HIP_ERROR(dIpiv.memcheck()); if(size_C) CHECK_HIP_ERROR(dC.memcheck()); CHECK_HIP_ERROR(dInfo.memcheck()); int size_W; hipsolver_ormqr_unmqr_bufferSize(FORTRAN, handle, side, trans, m, n, k, dA.data(), lda, dIpiv.data(), dC.data(), ldc, &size_W); device_strided_batch_vector dWork(size_W, 1, size_W, 1); if(size_W) CHECK_HIP_ERROR(dWork.memcheck()); // check computations if(argus.unit_check || argus.norm_check) ormqr_unmqr_getError(handle, side, trans, m, n, k, dA, lda, dIpiv, dC, ldc, dWork, size_W, dInfo, hA, hIpiv, hC, hCr, hInfo, &max_error); // collect performance data if(argus.timing) ormqr_unmqr_getPerfData(handle, side, trans, m, n, k, dA, lda, dIpiv, dC, ldc, dWork, size_W, dInfo, hA, hIpiv, hC, hInfo, &gpu_time_used, &cpu_time_used, hot_calls, argus.perf); // validate results for rocsolver-test // using s * machine_precision as tolerance int s = left ? m : n; if(argus.unit_check) ROCSOLVER_TEST_CHECK(T, max_error, s); // output results for rocsolver-bench if(argus.timing) { if(!argus.perf) { std::cerr << "\n============================================\n"; std::cerr << "Arguments:\n"; std::cerr << "============================================\n"; rocsolver_bench_output("side", "trans", "m", "n", "k", "lda", "ldc"); rocsolver_bench_output(sideC, transC, m, n, k, lda, ldc); std::cerr << "\n============================================\n"; std::cerr << "Results:\n"; std::cerr << "============================================\n"; if(argus.norm_check) { rocsolver_bench_output("cpu_time", "gpu_time", "error"); rocsolver_bench_output(cpu_time_used, gpu_time_used, max_error); } else { rocsolver_bench_output("cpu_time", "gpu_time"); rocsolver_bench_output(cpu_time_used, gpu_time_used); } std::cerr << std::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(); }