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
471 lines
21 KiB
C++
471 lines
21 KiB
C++
/* ************************************************************************
|
|
* Copyright (C) 2018-2025 Advanced Micro Devices, Inc. All rights reserved.
|
|
*
|
|
* Permission is hereby granted, free of charge, to any person obtaining a copy
|
|
* of this software and associated documentation files (the "Software"), to deal
|
|
* in the Software without restriction, including without limitation the rights
|
|
* to use, copy, modify, merge, publish, distribute, sublicense, and/or sell cop-
|
|
* ies of the Software, and to permit persons to whom the Software is furnished
|
|
* to do so, subject to the following conditions:
|
|
*
|
|
* The above copyright notice and this permission notice shall be included in all
|
|
* copies or substantial portions of the Software.
|
|
*
|
|
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IM-
|
|
* PLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS
|
|
* FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR
|
|
* COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER
|
|
* IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNE-
|
|
* CTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
|
|
*
|
|
* ************************************************************************ */
|
|
|
|
#pragma once
|
|
|
|
#include "cblas_interface.hpp"
|
|
#include "client_utility.hpp"
|
|
#include "flops.hpp"
|
|
#include "near.hpp"
|
|
#include "norm.hpp"
|
|
#include "rocblas.hpp"
|
|
#include "rocblas_init.hpp"
|
|
#include "rocblas_math.hpp"
|
|
#include "rocblas_matrix.hpp"
|
|
#include "rocblas_random.hpp"
|
|
#include "rocblas_test.hpp"
|
|
#include "rocblas_vector.hpp"
|
|
#include "unit.hpp"
|
|
|
|
#include "blas3/rocblas_trtri.hpp"
|
|
|
|
template <typename T>
|
|
void testing_trtri_batched_bad_arg(const Arguments& arg)
|
|
{
|
|
auto rocblas_trtri_batched_fn = arg.api & c_API_FORTRAN ? rocblas_trtri_batched<T, true>
|
|
: rocblas_trtri_batched<T, false>;
|
|
|
|
rocblas_local_handle handle{arg};
|
|
|
|
const rocblas_int N = 100;
|
|
const rocblas_int lda = 100;
|
|
const rocblas_int batch_count = 2;
|
|
|
|
const rocblas_fill uplo = rocblas_fill_upper;
|
|
const rocblas_diagonal diag = rocblas_diagonal_non_unit;
|
|
|
|
// Allocate device memory
|
|
DEVICE_MEMCHECK(device_batch_matrix<T>, dA, (N, N, lda, batch_count));
|
|
DEVICE_MEMCHECK(device_batch_matrix<T>, dinvA, (N, N, lda, batch_count));
|
|
|
|
EXPECT_ROCBLAS_STATUS(rocblas_trtri_batched_fn(handle,
|
|
rocblas_fill_full,
|
|
diag,
|
|
N,
|
|
dA.ptr_on_device(),
|
|
lda,
|
|
dinvA.ptr_on_device(),
|
|
lda,
|
|
batch_count),
|
|
rocblas_status_invalid_value);
|
|
|
|
EXPECT_ROCBLAS_STATUS(rocblas_trtri_batched_fn(handle,
|
|
uplo,
|
|
(rocblas_diagonal)rocblas_side_both,
|
|
N,
|
|
dA.ptr_on_device(),
|
|
lda,
|
|
dinvA.ptr_on_device(),
|
|
lda,
|
|
batch_count),
|
|
rocblas_status_invalid_value);
|
|
|
|
// check for invalid sizes
|
|
EXPECT_ROCBLAS_STATUS(rocblas_trtri_batched_fn(handle,
|
|
uplo,
|
|
diag,
|
|
-1,
|
|
dA.ptr_on_device(),
|
|
lda,
|
|
dinvA.ptr_on_device(),
|
|
lda,
|
|
batch_count),
|
|
rocblas_status_invalid_size);
|
|
|
|
EXPECT_ROCBLAS_STATUS(
|
|
rocblas_trtri_batched_fn(
|
|
handle, uplo, diag, N, dA.ptr_on_device(), lda, dinvA.ptr_on_device(), lda, -1),
|
|
rocblas_status_invalid_size);
|
|
|
|
EXPECT_ROCBLAS_STATUS(rocblas_trtri_batched_fn(handle,
|
|
uplo,
|
|
diag,
|
|
N,
|
|
dA.ptr_on_device(),
|
|
lda - 1,
|
|
dinvA.ptr_on_device(),
|
|
lda,
|
|
batch_count),
|
|
rocblas_status_invalid_size);
|
|
|
|
EXPECT_ROCBLAS_STATUS(rocblas_trtri_batched_fn(handle,
|
|
uplo,
|
|
diag,
|
|
N,
|
|
dA.ptr_on_device(),
|
|
lda,
|
|
dinvA.ptr_on_device(),
|
|
lda - 1,
|
|
batch_count),
|
|
rocblas_status_invalid_size);
|
|
|
|
// nullptr tests
|
|
EXPECT_ROCBLAS_STATUS(rocblas_trtri_batched_fn(nullptr,
|
|
uplo,
|
|
diag,
|
|
N,
|
|
dA.ptr_on_device(),
|
|
lda,
|
|
dinvA.ptr_on_device(),
|
|
lda,
|
|
batch_count),
|
|
rocblas_status_invalid_handle);
|
|
|
|
EXPECT_ROCBLAS_STATUS(
|
|
rocblas_trtri_batched_fn(
|
|
handle, uplo, diag, N, nullptr, lda, dinvA.ptr_on_device(), lda, batch_count),
|
|
rocblas_status_invalid_pointer);
|
|
|
|
EXPECT_ROCBLAS_STATUS(
|
|
rocblas_trtri_batched_fn(
|
|
handle, uplo, diag, N, dA.ptr_on_device(), lda, nullptr, lda, batch_count),
|
|
rocblas_status_invalid_pointer);
|
|
|
|
// quick return: If N==0, then all pointers can be nullptr without error
|
|
EXPECT_ROCBLAS_STATUS(
|
|
rocblas_trtri_batched_fn(handle, uplo, diag, 0, nullptr, lda, nullptr, lda, batch_count),
|
|
rocblas_status_success);
|
|
|
|
// quick return: If batch_count==0, then all pointers can be nullptr without error
|
|
EXPECT_ROCBLAS_STATUS(
|
|
rocblas_trtri_batched_fn(handle, uplo, diag, N, nullptr, lda, nullptr, lda, 0),
|
|
rocblas_status_success);
|
|
}
|
|
|
|
template <typename T>
|
|
void testing_trtri_batched(const Arguments& arg)
|
|
{
|
|
auto rocblas_trtri_batched_fn = arg.api & c_API_FORTRAN ? rocblas_trtri_batched<T, true>
|
|
: rocblas_trtri_batched<T, false>;
|
|
|
|
rocblas_int N = arg.N;
|
|
rocblas_int lda = arg.lda;
|
|
rocblas_int batch_count = arg.batch_count;
|
|
|
|
char char_uplo = arg.uplo;
|
|
char char_diag = arg.diag;
|
|
|
|
// char_uplo = 'U';
|
|
rocblas_fill uplo = char2rocblas_fill(char_uplo);
|
|
rocblas_diagonal diag = char2rocblas_diagonal(char_diag);
|
|
|
|
// for internal interface testing: using ldc/ldd as offsets
|
|
const bool internal_api = arg.api == INTERNAL;
|
|
rocblas_stride offsetA = internal_api ? arg.ldc : 0;
|
|
rocblas_stride offsetinvA = internal_api ? arg.ldd : 0;
|
|
|
|
rocblas_local_handle handle{arg};
|
|
|
|
// argument sanity check, quick return if input parameters are invalid before allocating invalid
|
|
bool invalid_size = N < 0 || lda < N || batch_count < 0;
|
|
if(invalid_size || batch_count == 0)
|
|
{
|
|
EXPECT_ROCBLAS_STATUS(rocblas_trtri_batched_fn(
|
|
handle, uplo, diag, N, nullptr, lda, nullptr, lda, batch_count),
|
|
invalid_size ? rocblas_status_invalid_size : rocblas_status_success);
|
|
return;
|
|
}
|
|
|
|
// Naming: `h` is in CPU (host) memory(eg hA), `d` is in GPU (device) memory (eg dA).
|
|
// Allocate host memory
|
|
HOST_MEMCHECK(host_batch_matrix<T>, hA, (N, N, lda, batch_count));
|
|
HOST_MEMCHECK(host_batch_matrix<T>, hB, (N, N, lda, batch_count));
|
|
HOST_MEMCHECK(host_batch_matrix<T>, hA_2, (N, N, lda, batch_count));
|
|
|
|
// Allocate device memory
|
|
DEVICE_MEMCHECK(device_batch_matrix<T>, dA, (N, N, lda, batch_count, false, offsetA));
|
|
DEVICE_MEMCHECK(device_batch_matrix<T>, dinvA, (N, N, lda, batch_count, false, offsetinvA));
|
|
|
|
// Initial Data on CPU
|
|
//Explicitly set the unused side of matrix `hA` to 0 when using it for temp storage.
|
|
//Used rocblas_client_triangular_matrix type initialization, which will ensure the unused side is set to 0 or could be done manually
|
|
rocblas_init_matrix(
|
|
hA, arg, rocblas_client_never_set_nan, rocblas_client_triangular_matrix, true, true);
|
|
|
|
for(size_t b = 0; b < batch_count; b++)
|
|
{
|
|
for(size_t i = 0; i < N; i++)
|
|
{
|
|
for(size_t j = 0; j < N; j++)
|
|
{
|
|
T* A = hA[b];
|
|
A[i + j * lda] *= 0.01;
|
|
|
|
if(j % 2)
|
|
A[i + j * lda] *= -1;
|
|
|
|
if(i == j)
|
|
{
|
|
if(diag == rocblas_diagonal_unit)
|
|
A[i + j * lda] = 1.0; // need to preprocess matrix for clbas_trtri
|
|
else
|
|
A[i + j * lda] *= 100.0;
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
hB.copy_from(hA);
|
|
|
|
// copy data from CPU to device
|
|
CHECK_HIP_ERROR(dA.transfer_from(hA));
|
|
CHECK_HIP_ERROR(dinvA.transfer_from(hA));
|
|
|
|
double gpu_time_used, cpu_time_used;
|
|
gpu_time_used = cpu_time_used = 0.0;
|
|
double rocblas_error_out, rocblas_error_in;
|
|
|
|
if(!ROCBLAS_REALLOC_ON_DEMAND)
|
|
{
|
|
// Compute size
|
|
CHECK_ROCBLAS_ERROR(rocblas_start_device_memory_size_query(handle));
|
|
|
|
CHECK_ALLOC_QUERY(rocblas_trtri_batched_fn(handle,
|
|
uplo,
|
|
diag,
|
|
N,
|
|
dA.ptr_on_device(),
|
|
lda,
|
|
dinvA.ptr_on_device(),
|
|
lda,
|
|
batch_count));
|
|
|
|
// Test in place
|
|
CHECK_ALLOC_QUERY(rocblas_trtri_batched_fn(
|
|
handle, uplo, diag, N, dA.ptr_on_device(), lda, dA.ptr_on_device(), lda, batch_count));
|
|
|
|
size_t size;
|
|
CHECK_ROCBLAS_ERROR(rocblas_stop_device_memory_size_query(handle, &size));
|
|
|
|
// Allocate memory
|
|
CHECK_ROCBLAS_ERROR(rocblas_set_device_memory_size(handle, size));
|
|
}
|
|
|
|
/* =====================================================================
|
|
ROCBLAS
|
|
=================================================================== */
|
|
if(arg.unit_check || arg.norm_check)
|
|
{
|
|
// Test out of place
|
|
handle.pre_test(arg);
|
|
if(arg.api != INTERNAL)
|
|
{
|
|
CHECK_ROCBLAS_ERROR(rocblas_trtri_batched_fn(handle,
|
|
uplo,
|
|
diag,
|
|
N,
|
|
dA.ptr_on_device(),
|
|
lda,
|
|
dinvA.ptr_on_device(),
|
|
lda,
|
|
batch_count));
|
|
if(arg.repeatability_check)
|
|
{
|
|
HOST_MEMCHECK(host_batch_matrix<T>, hA_copy, (N, N, lda, batch_count));
|
|
CHECK_HIP_ERROR(hA.transfer_from(dinvA));
|
|
|
|
// multi-GPU support
|
|
int device_id, device_count;
|
|
CHECK_HIP_ERROR(limit_device_count(device_count, (int)arg.devices));
|
|
|
|
for(int dev_id = 0; dev_id < device_count; dev_id++)
|
|
{
|
|
CHECK_HIP_ERROR(hipGetDevice(&device_id));
|
|
if(device_id != dev_id)
|
|
CHECK_HIP_ERROR(hipSetDevice(dev_id));
|
|
|
|
//New rocblas handle for new device
|
|
rocblas_local_handle handle_copy{arg};
|
|
|
|
//Allocate device memory in new device
|
|
DEVICE_MEMCHECK(
|
|
device_batch_matrix<T>, dA_copy, (N, N, lda, batch_count, false, offsetA));
|
|
DEVICE_MEMCHECK(device_batch_matrix<T>,
|
|
dinvA_copy,
|
|
(N, N, lda, batch_count, false, offsetinvA));
|
|
|
|
for(int runs = 0; runs < arg.iters; runs++)
|
|
{
|
|
CHECK_HIP_ERROR(dA_copy.transfer_from(hB));
|
|
CHECK_HIP_ERROR(dinvA_copy.transfer_from(hB));
|
|
CHECK_ROCBLAS_ERROR(rocblas_trtri_batched_fn(handle_copy,
|
|
uplo,
|
|
diag,
|
|
N,
|
|
dA_copy.ptr_on_device(),
|
|
lda,
|
|
dinvA_copy.ptr_on_device(),
|
|
lda,
|
|
batch_count));
|
|
CHECK_HIP_ERROR(hA_copy.transfer_from(dinvA_copy));
|
|
unit_check_general<T>(N, N, lda, hA, hA_copy, batch_count);
|
|
}
|
|
}
|
|
return;
|
|
}
|
|
}
|
|
else
|
|
{
|
|
rocblas_stride strideA = arg.stride_a;
|
|
rocblas_stride strideinvA = arg.stride_b;
|
|
rocblas_stride subStride = 0;
|
|
rocblas_int sub_batch_count = 1;
|
|
|
|
size_t work_el = rocblas_internal_trtri_temp_elements(N, batch_count);
|
|
DEVICE_MEMCHECK(device_batch_vector<T>, workspace_arr, (work_el, 1, batch_count));
|
|
|
|
CHECK_ROCBLAS_ERROR(
|
|
rocblas_internal_trtri_batched_template(handle,
|
|
uplo,
|
|
diag,
|
|
N,
|
|
(const T* const*)dA.ptr_on_device(),
|
|
-offsetA,
|
|
lda,
|
|
strideA,
|
|
subStride,
|
|
(T* const*)dinvA.ptr_on_device(),
|
|
-offsetinvA,
|
|
lda,
|
|
strideinvA,
|
|
subStride,
|
|
batch_count,
|
|
sub_batch_count,
|
|
(T* const*)workspace_arr.ptr_on_device()));
|
|
}
|
|
handle.post_test(arg);
|
|
|
|
// Test in place
|
|
handle.pre_test(arg);
|
|
if(arg.api != INTERNAL)
|
|
{
|
|
CHECK_ROCBLAS_ERROR(rocblas_trtri_batched_fn(handle,
|
|
uplo,
|
|
diag,
|
|
N,
|
|
dA.ptr_on_device(),
|
|
lda,
|
|
dA.ptr_on_device(),
|
|
lda,
|
|
batch_count));
|
|
}
|
|
else
|
|
{
|
|
rocblas_stride strideA = arg.stride_a;
|
|
rocblas_stride subStride = 0;
|
|
rocblas_int sub_batch_count = 1;
|
|
|
|
size_t work_el = rocblas_internal_trtri_temp_elements(N, batch_count);
|
|
DEVICE_MEMCHECK(device_batch_vector<T>, workspace_arr, (work_el, 1, batch_count));
|
|
|
|
CHECK_ROCBLAS_ERROR(
|
|
rocblas_internal_trtri_batched_template(handle,
|
|
uplo,
|
|
diag,
|
|
N,
|
|
(const T* const*)dA.ptr_on_device(),
|
|
-offsetA,
|
|
lda,
|
|
strideA,
|
|
subStride,
|
|
(T* const*)dA.ptr_on_device(),
|
|
-offsetA,
|
|
lda,
|
|
strideA,
|
|
subStride,
|
|
batch_count,
|
|
sub_batch_count,
|
|
(T* const*)workspace_arr.ptr_on_device()));
|
|
}
|
|
handle.post_test(arg);
|
|
|
|
// copy output from device to CPU
|
|
CHECK_HIP_ERROR(hA.transfer_from(dinvA));
|
|
|
|
/* =====================================================================
|
|
CPU BLAS
|
|
=================================================================== */
|
|
if(arg.timing)
|
|
cpu_time_used = get_time_us_no_sync();
|
|
|
|
// CBLAS doesn't have trtri implementation so using the LAPACK trtri
|
|
for(size_t b = 0; b < batch_count; b++)
|
|
ref_lapack_xtrtri<T>(char_uplo, char_diag, N, hB[b], lda);
|
|
|
|
if(arg.timing)
|
|
cpu_time_used = get_time_us_no_sync() - cpu_time_used;
|
|
|
|
// test out-of-place
|
|
const double rel_error = trtri_tolerance<T>(N);
|
|
if(arg.unit_check)
|
|
near_check_general<T>(N, N, lda, hB, hA, batch_count, rel_error);
|
|
|
|
if(arg.norm_check)
|
|
rocblas_error_out
|
|
= norm_check_symmetric<T>('F', char_uplo, N, lda, hB, hA, batch_count);
|
|
|
|
// test in-place
|
|
CHECK_HIP_ERROR(hA.transfer_from(dA));
|
|
if(arg.unit_check)
|
|
near_check_general<T>(N, N, lda, hB, hA, batch_count, rel_error);
|
|
|
|
if(arg.norm_check)
|
|
rocblas_error_in = norm_check_symmetric<T>('F', char_uplo, N, lda, hB, hA, batch_count);
|
|
}
|
|
|
|
if(arg.timing)
|
|
{
|
|
int number_cold_calls = arg.cold_iters;
|
|
int total_calls = number_cold_calls + arg.iters;
|
|
|
|
hipStream_t stream;
|
|
CHECK_ROCBLAS_ERROR(rocblas_get_stream(handle, &stream));
|
|
for(int i = 0; i < total_calls; i++)
|
|
{
|
|
if(i == number_cold_calls)
|
|
gpu_time_used = get_time_us_sync(stream);
|
|
|
|
rocblas_trtri_batched_fn(handle,
|
|
uplo,
|
|
diag,
|
|
N,
|
|
dA.ptr_on_device(),
|
|
lda,
|
|
dinvA.ptr_on_device(),
|
|
lda,
|
|
batch_count);
|
|
}
|
|
|
|
gpu_time_used = get_time_us_sync(stream) - gpu_time_used;
|
|
|
|
ArgumentModel<e_uplo, e_diag, e_N, e_lda, e_batch_count>{}.log_args<T>(
|
|
rocblas_cout,
|
|
arg,
|
|
gpu_time_used,
|
|
trtri_gflop_count<T>(N),
|
|
ArgumentLogging::NA_value,
|
|
cpu_time_used,
|
|
rocblas_error_out,
|
|
rocblas_error_in);
|
|
}
|
|
}
|