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
772 lines
27 KiB
C++
772 lines
27 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 "testing_common.hpp"
|
|
|
|
/* ============================================================================================ */
|
|
|
|
template <typename T>
|
|
void testing_geam_batched_bad_arg(const Arguments& arg)
|
|
{
|
|
auto rocblas_geam_batched_fn
|
|
= arg.api & c_API_FORTRAN ? rocblas_geam_batched<T, true> : rocblas_geam_batched<T, false>;
|
|
|
|
auto rocblas_geam_batched_fn_64 = arg.api & c_API_FORTRAN ? rocblas_geam_batched_64<T, true>
|
|
: rocblas_geam_batched_64<T, false>;
|
|
|
|
for(auto pointer_mode : {rocblas_pointer_mode_host, rocblas_pointer_mode_device})
|
|
{
|
|
rocblas_local_handle handle{arg};
|
|
CHECK_ROCBLAS_ERROR(rocblas_set_pointer_mode(handle, pointer_mode));
|
|
|
|
const int64_t M = 100;
|
|
const int64_t N = 99;
|
|
|
|
const int64_t lda = 100;
|
|
const int64_t ldb = 100;
|
|
const int64_t ldc = 100;
|
|
|
|
const int64_t batch_count = 2;
|
|
|
|
DEVICE_MEMCHECK(device_vector<T>, alpha_d, (1));
|
|
DEVICE_MEMCHECK(device_vector<T>, beta_d, (1));
|
|
DEVICE_MEMCHECK(device_vector<T>, one_d, (1));
|
|
DEVICE_MEMCHECK(device_vector<T>, zero_d, (1));
|
|
|
|
const T alpha_h(1), beta_h(2), one_h(1), zero_h(0);
|
|
|
|
const T* alpha = &alpha_h;
|
|
const T* beta = &beta_h;
|
|
const T* one = &one_h;
|
|
const T* zero = &zero_h;
|
|
|
|
if(pointer_mode == rocblas_pointer_mode_device)
|
|
{
|
|
CHECK_HIP_ERROR(hipMemcpy(alpha_d, alpha, sizeof(*alpha), hipMemcpyHostToDevice));
|
|
alpha = alpha_d;
|
|
CHECK_HIP_ERROR(hipMemcpy(beta_d, beta, sizeof(*beta), hipMemcpyHostToDevice));
|
|
beta = beta_d;
|
|
CHECK_HIP_ERROR(hipMemcpy(one_d, one, sizeof(*one), hipMemcpyHostToDevice));
|
|
one = one_d;
|
|
CHECK_HIP_ERROR(hipMemcpy(zero_d, zero, sizeof(*zero), hipMemcpyHostToDevice));
|
|
zero = zero_d;
|
|
}
|
|
|
|
const rocblas_operation transA = rocblas_operation_none;
|
|
const rocblas_operation transB = rocblas_operation_none;
|
|
|
|
int64_t A_row = transA == rocblas_operation_none ? M : N;
|
|
int64_t A_col = transA == rocblas_operation_none ? N : M;
|
|
int64_t B_row = transB == rocblas_operation_none ? M : N;
|
|
int64_t B_col = transB == rocblas_operation_none ? N : M;
|
|
|
|
// Allocate device memory
|
|
DEVICE_MEMCHECK(device_batch_matrix<T>, dA, (A_row, A_col, lda, batch_count));
|
|
DEVICE_MEMCHECK(device_batch_matrix<T>, dB, (B_row, B_col, ldb, batch_count));
|
|
DEVICE_MEMCHECK(device_batch_matrix<T>, dC, (M, N, ldc, batch_count));
|
|
|
|
DAPI_EXPECT(rocblas_status_invalid_handle,
|
|
rocblas_geam_batched_fn,
|
|
(nullptr,
|
|
transA,
|
|
transB,
|
|
M,
|
|
N,
|
|
alpha,
|
|
dA.ptr_on_device(),
|
|
lda,
|
|
beta,
|
|
dB.ptr_on_device(),
|
|
ldb,
|
|
dC.ptr_on_device(),
|
|
ldc,
|
|
batch_count));
|
|
|
|
// invalid values
|
|
DAPI_EXPECT(rocblas_status_invalid_value,
|
|
rocblas_geam_batched_fn,
|
|
(handle,
|
|
(rocblas_operation)rocblas_fill_full,
|
|
transB,
|
|
M,
|
|
N,
|
|
alpha,
|
|
dA.ptr_on_device(),
|
|
lda,
|
|
beta,
|
|
dB.ptr_on_device(),
|
|
ldb,
|
|
dC.ptr_on_device(),
|
|
ldc,
|
|
batch_count));
|
|
|
|
DAPI_EXPECT(rocblas_status_invalid_value,
|
|
rocblas_geam_batched_fn,
|
|
(handle,
|
|
transA,
|
|
(rocblas_operation)rocblas_fill_full,
|
|
M,
|
|
N,
|
|
alpha,
|
|
dA.ptr_on_device(),
|
|
lda,
|
|
beta,
|
|
dB.ptr_on_device(),
|
|
ldb,
|
|
dC.ptr_on_device(),
|
|
ldc,
|
|
batch_count));
|
|
|
|
// invalid size in regular tests
|
|
|
|
// alpha/beta
|
|
DAPI_EXPECT(rocblas_status_invalid_pointer,
|
|
rocblas_geam_batched_fn,
|
|
(handle,
|
|
transA,
|
|
transB,
|
|
M,
|
|
N,
|
|
nullptr,
|
|
dA.ptr_on_device(),
|
|
lda,
|
|
beta,
|
|
dB.ptr_on_device(),
|
|
ldb,
|
|
dC.ptr_on_device(),
|
|
ldc,
|
|
batch_count));
|
|
|
|
DAPI_EXPECT(rocblas_status_invalid_pointer,
|
|
rocblas_geam_batched_fn,
|
|
(handle,
|
|
transA,
|
|
transB,
|
|
M,
|
|
N,
|
|
alpha,
|
|
dA.ptr_on_device(),
|
|
lda,
|
|
nullptr,
|
|
dB.ptr_on_device(),
|
|
ldb,
|
|
dC.ptr_on_device(),
|
|
ldc,
|
|
batch_count));
|
|
|
|
// invalid pointers
|
|
DAPI_EXPECT(rocblas_status_invalid_pointer,
|
|
rocblas_geam_batched_fn,
|
|
(handle,
|
|
transA,
|
|
transB,
|
|
M,
|
|
N,
|
|
alpha,
|
|
dA.ptr_on_device(),
|
|
lda,
|
|
beta,
|
|
dB.ptr_on_device(),
|
|
ldb,
|
|
nullptr,
|
|
ldc,
|
|
batch_count));
|
|
|
|
if(pointer_mode == rocblas_pointer_mode_host)
|
|
{
|
|
DAPI_EXPECT(rocblas_status_invalid_pointer,
|
|
rocblas_geam_batched_fn,
|
|
(handle,
|
|
transA,
|
|
transB,
|
|
M,
|
|
N,
|
|
alpha,
|
|
nullptr,
|
|
lda,
|
|
beta,
|
|
dB.ptr_on_device(),
|
|
ldb,
|
|
dC.ptr_on_device(),
|
|
ldc,
|
|
batch_count));
|
|
|
|
DAPI_EXPECT(rocblas_status_invalid_pointer,
|
|
rocblas_geam_batched_fn,
|
|
(handle,
|
|
transA,
|
|
transB,
|
|
M,
|
|
N,
|
|
alpha,
|
|
dA.ptr_on_device(),
|
|
lda,
|
|
beta,
|
|
nullptr,
|
|
ldb,
|
|
dC.ptr_on_device(),
|
|
ldc,
|
|
batch_count));
|
|
}
|
|
|
|
// batch_count==0 then all may be nullptr
|
|
DAPI_CHECK(rocblas_geam_batched_fn,
|
|
(handle,
|
|
transA,
|
|
transB,
|
|
M,
|
|
N,
|
|
nullptr,
|
|
nullptr,
|
|
lda,
|
|
nullptr,
|
|
nullptr,
|
|
ldb,
|
|
nullptr,
|
|
ldc,
|
|
0));
|
|
|
|
// M==0 then all may be nullptr
|
|
DAPI_CHECK(rocblas_geam_batched_fn,
|
|
(handle,
|
|
transA,
|
|
transB,
|
|
0,
|
|
N,
|
|
nullptr,
|
|
nullptr,
|
|
lda,
|
|
nullptr,
|
|
nullptr,
|
|
ldb,
|
|
nullptr,
|
|
ldc,
|
|
0));
|
|
|
|
// N==0 then all may be nullptr
|
|
DAPI_CHECK(rocblas_geam_batched_fn,
|
|
(handle,
|
|
transA,
|
|
transB,
|
|
M,
|
|
0,
|
|
nullptr,
|
|
nullptr,
|
|
lda,
|
|
nullptr,
|
|
nullptr,
|
|
ldb,
|
|
nullptr,
|
|
ldc,
|
|
0));
|
|
|
|
// alpha==0 then A may be nullptr
|
|
DAPI_CHECK(rocblas_geam_batched_fn,
|
|
(handle,
|
|
transA,
|
|
transB,
|
|
M,
|
|
N,
|
|
zero,
|
|
nullptr,
|
|
lda,
|
|
beta,
|
|
dB.ptr_on_device(),
|
|
ldb,
|
|
dC.ptr_on_device(),
|
|
ldc,
|
|
batch_count));
|
|
|
|
// beta==0 then B may be nullptr
|
|
DAPI_CHECK(rocblas_geam_batched_fn,
|
|
(handle,
|
|
transA,
|
|
transB,
|
|
M,
|
|
N,
|
|
alpha,
|
|
dA.ptr_on_device(),
|
|
lda,
|
|
zero,
|
|
nullptr,
|
|
ldb,
|
|
dC.ptr_on_device(),
|
|
ldc,
|
|
batch_count));
|
|
}
|
|
}
|
|
|
|
template <typename T>
|
|
void testing_geam_batched(const Arguments& arg)
|
|
{
|
|
auto rocblas_geam_batched_fn
|
|
= arg.api & c_API_FORTRAN ? rocblas_geam_batched<T, true> : rocblas_geam_batched<T, false>;
|
|
|
|
auto rocblas_geam_batched_fn_64 = arg.api & c_API_FORTRAN ? rocblas_geam_batched_64<T, true>
|
|
: rocblas_geam_batched_64<T, false>;
|
|
|
|
rocblas_operation transA = char2rocblas_operation(arg.transA);
|
|
rocblas_operation transB = char2rocblas_operation(arg.transB);
|
|
|
|
int64_t M = arg.M;
|
|
int64_t N = arg.N;
|
|
|
|
int64_t lda = arg.lda;
|
|
int64_t ldb = arg.ldb;
|
|
int64_t ldc = arg.ldc;
|
|
int64_t batch_count = arg.batch_count;
|
|
|
|
T alpha = arg.get_alpha<T>();
|
|
T beta = arg.get_beta<T>();
|
|
|
|
double gpu_time_used, cpu_time_used;
|
|
gpu_time_used = cpu_time_used = 0.0;
|
|
|
|
double rocblas_error_1 = std::numeric_limits<double>::max();
|
|
double rocblas_error_2 = std::numeric_limits<double>::max();
|
|
double rocblas_error = std::numeric_limits<double>::max();
|
|
|
|
rocblas_local_handle handle{arg};
|
|
|
|
int64_t A_row = transA == rocblas_operation_none ? M : N;
|
|
int64_t A_col = transA == rocblas_operation_none ? N : M;
|
|
int64_t B_row = transB == rocblas_operation_none ? M : N;
|
|
int64_t B_col = transB == rocblas_operation_none ? N : M;
|
|
|
|
size_t size_C = size_t(ldc) * size_t(N);
|
|
|
|
// argument sanity check before allocating invalid memory
|
|
bool invalid_size = M < 0 || N < 0 || lda < A_row || ldb < B_row || ldc < M || batch_count < 0;
|
|
if(invalid_size || !M || !N || !batch_count)
|
|
{
|
|
DAPI_EXPECT(invalid_size ? rocblas_status_invalid_size : rocblas_status_success,
|
|
rocblas_geam_batched_fn,
|
|
(handle,
|
|
transA,
|
|
transB,
|
|
M,
|
|
N,
|
|
nullptr,
|
|
nullptr,
|
|
lda,
|
|
nullptr,
|
|
nullptr,
|
|
ldb,
|
|
(T**)0x1, // defeat C==A or B leading dim invalid checks, C nullptr in bad_arg
|
|
ldc,
|
|
batch_count));
|
|
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, (A_row, A_col, lda, batch_count));
|
|
HOST_MEMCHECK(host_batch_matrix<T>, hA_copy, (A_row, A_col, lda, batch_count));
|
|
HOST_MEMCHECK(host_batch_matrix<T>, hB, (B_row, B_col, ldb, batch_count));
|
|
HOST_MEMCHECK(host_batch_matrix<T>, hB_copy, (B_row, B_col, ldb, batch_count));
|
|
HOST_MEMCHECK(host_batch_matrix<T>, hC, (M, N, ldc, batch_count));
|
|
HOST_MEMCHECK(host_batch_matrix<T>, hC_gold, (M, N, ldc, batch_count));
|
|
HOST_MEMCHECK(host_vector<T>, h_alpha, (1));
|
|
HOST_MEMCHECK(host_vector<T>, h_beta, (1));
|
|
|
|
// Allocate device memory
|
|
DEVICE_MEMCHECK(device_batch_matrix<T>, dA, (A_row, A_col, lda, batch_count));
|
|
DEVICE_MEMCHECK(device_batch_matrix<T>, dB, (B_row, B_col, ldb, batch_count));
|
|
DEVICE_MEMCHECK(device_batch_matrix<T>, dC, (M, N, ldc, batch_count));
|
|
DEVICE_MEMCHECK(device_batch_matrix<T>, dC_in_place, (M, N, ldc, batch_count));
|
|
DEVICE_MEMCHECK(device_vector<T>, d_alpha, (1));
|
|
DEVICE_MEMCHECK(device_vector<T>, d_beta, (1));
|
|
|
|
// Initial Data on CPU
|
|
h_alpha[0] = alpha;
|
|
h_beta[0] = beta;
|
|
|
|
// Initialize data on host memory
|
|
rocblas_init_matrix(
|
|
hA, arg, rocblas_client_alpha_sets_nan, rocblas_client_general_matrix, true);
|
|
rocblas_init_matrix(hB, arg, rocblas_client_beta_sets_nan, rocblas_client_general_matrix);
|
|
rocblas_init_matrix(hC, arg, rocblas_client_never_set_nan, rocblas_client_general_matrix);
|
|
|
|
hA_copy.copy_from(hA);
|
|
hB_copy.copy_from(hB);
|
|
hC_gold.copy_from(hC);
|
|
|
|
// copy data from CPU to device
|
|
|
|
CHECK_HIP_ERROR(dA.transfer_from(hA));
|
|
CHECK_HIP_ERROR(dB.transfer_from(hB));
|
|
|
|
if(arg.unit_check || arg.norm_check)
|
|
{
|
|
if(arg.pointer_mode_host)
|
|
{
|
|
CHECK_ROCBLAS_ERROR(rocblas_set_pointer_mode(handle, rocblas_pointer_mode_host));
|
|
handle.pre_test(arg);
|
|
DAPI_CHECK(rocblas_geam_batched_fn,
|
|
(handle,
|
|
transA,
|
|
transB,
|
|
M,
|
|
N,
|
|
&alpha,
|
|
dA.ptr_on_device(),
|
|
lda,
|
|
&beta,
|
|
dB.ptr_on_device(),
|
|
ldb,
|
|
dC.ptr_on_device(),
|
|
ldc,
|
|
batch_count));
|
|
handle.post_test(arg);
|
|
|
|
CHECK_HIP_ERROR(hC.transfer_from(dC));
|
|
}
|
|
|
|
if(arg.pointer_mode_device)
|
|
{
|
|
CHECK_HIP_ERROR(dC.transfer_from(hC_gold));
|
|
CHECK_HIP_ERROR(d_alpha.transfer_from(h_alpha));
|
|
CHECK_HIP_ERROR(d_beta.transfer_from(h_beta));
|
|
|
|
CHECK_ROCBLAS_ERROR(rocblas_set_pointer_mode(handle, rocblas_pointer_mode_device));
|
|
|
|
DAPI_CHECK(rocblas_geam_batched_fn,
|
|
(handle,
|
|
transA,
|
|
transB,
|
|
M,
|
|
N,
|
|
d_alpha,
|
|
dA.ptr_on_device(),
|
|
lda,
|
|
d_beta,
|
|
dB.ptr_on_device(),
|
|
ldb,
|
|
dC.ptr_on_device(),
|
|
ldc,
|
|
batch_count));
|
|
|
|
if(arg.repeatability_check)
|
|
{
|
|
HOST_MEMCHECK(host_batch_matrix<T>, hC_copy, (M, N, ldc, batch_count));
|
|
CHECK_HIP_ERROR(hC.transfer_from(dC));
|
|
// 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, (A_row, A_col, lda, batch_count));
|
|
DEVICE_MEMCHECK(
|
|
device_batch_matrix<T>, dB_copy, (B_row, B_col, ldb, batch_count));
|
|
DEVICE_MEMCHECK(device_batch_matrix<T>, dC_copy, (M, N, ldc, batch_count));
|
|
DEVICE_MEMCHECK(device_vector<T>, d_alpha_copy, (1));
|
|
DEVICE_MEMCHECK(device_vector<T>, d_beta_copy, (1));
|
|
|
|
// copy data from CPU to device
|
|
CHECK_HIP_ERROR(dA_copy.transfer_from(hA));
|
|
CHECK_HIP_ERROR(dB_copy.transfer_from(hB));
|
|
CHECK_HIP_ERROR(dC_copy.transfer_from(hC_gold));
|
|
CHECK_HIP_ERROR(d_alpha_copy.transfer_from(h_alpha));
|
|
CHECK_HIP_ERROR(d_beta_copy.transfer_from(h_beta));
|
|
|
|
CHECK_ROCBLAS_ERROR(
|
|
rocblas_set_pointer_mode(handle_copy, rocblas_pointer_mode_device));
|
|
|
|
for(int runs = 0; runs < arg.iters; runs++)
|
|
{
|
|
DAPI_CHECK(rocblas_geam_batched_fn,
|
|
(handle_copy,
|
|
transA,
|
|
transB,
|
|
M,
|
|
N,
|
|
d_alpha_copy,
|
|
dA_copy.ptr_on_device(),
|
|
lda,
|
|
d_beta_copy,
|
|
dB_copy.ptr_on_device(),
|
|
ldb,
|
|
dC_copy.ptr_on_device(),
|
|
ldc,
|
|
batch_count));
|
|
CHECK_HIP_ERROR(hC_copy.transfer_from(dC_copy));
|
|
unit_check_general<T>(M, N, ldc, hC, hC_copy, batch_count);
|
|
}
|
|
}
|
|
return;
|
|
}
|
|
}
|
|
|
|
// reference calculation for golden result
|
|
cpu_time_used = get_time_us_no_sync();
|
|
|
|
for(int b = 0; b < batch_count; b++)
|
|
{
|
|
auto hA_copy_p = hA_copy[b];
|
|
auto hB_copy_p = hB_copy[b];
|
|
auto hC_gold_p = hC_gold[b];
|
|
|
|
ref_geam(transA,
|
|
transB,
|
|
M,
|
|
N,
|
|
(T*)h_alpha,
|
|
(T*)hA_copy_p,
|
|
lda,
|
|
(T*)h_beta,
|
|
(T*)hB_copy_p,
|
|
ldb,
|
|
(T*)hC_gold_p,
|
|
ldc);
|
|
}
|
|
|
|
cpu_time_used = get_time_us_no_sync() - cpu_time_used;
|
|
|
|
if(arg.pointer_mode_host)
|
|
{
|
|
if(arg.unit_check)
|
|
{
|
|
unit_check_general<T>(M, N, ldc, hC_gold, hC, batch_count);
|
|
}
|
|
|
|
if(arg.norm_check)
|
|
{
|
|
rocblas_error_1 = norm_check_general<T>('F', M, N, ldc, hC_gold, hC, batch_count);
|
|
}
|
|
}
|
|
|
|
if(arg.pointer_mode_device)
|
|
{
|
|
// fetch GPU
|
|
CHECK_HIP_ERROR(hC.transfer_from(dC));
|
|
|
|
if(arg.unit_check)
|
|
{
|
|
unit_check_general<T>(M, N, ldc, hC_gold, hC, batch_count);
|
|
}
|
|
|
|
if(arg.norm_check)
|
|
{
|
|
rocblas_error_2 = norm_check_general<T>('F', M, N, ldc, hC_gold, hC, batch_count);
|
|
}
|
|
}
|
|
|
|
// inplace check for dC == dA
|
|
if(arg.pointer_mode_host)
|
|
{
|
|
bool invalid_size_in_place = lda != ldc || transA != rocblas_operation_none;
|
|
|
|
if((lda == ldc) && (transA == rocblas_operation_none))
|
|
CHECK_HIP_ERROR(dC_in_place.transfer_from(hA));
|
|
//
|
|
CHECK_ROCBLAS_ERROR(rocblas_set_pointer_mode(handle, rocblas_pointer_mode_host));
|
|
|
|
DAPI_EXPECT(invalid_size_in_place ? rocblas_status_invalid_size
|
|
: rocblas_status_success,
|
|
rocblas_geam_batched_fn,
|
|
(handle,
|
|
transA,
|
|
transB,
|
|
M,
|
|
N,
|
|
&alpha,
|
|
dC_in_place.ptr_on_device(),
|
|
lda,
|
|
&beta,
|
|
dB.ptr_on_device(),
|
|
ldb,
|
|
dC_in_place.ptr_on_device(),
|
|
ldc,
|
|
batch_count));
|
|
|
|
if(!invalid_size_in_place)
|
|
{
|
|
CHECK_HIP_ERROR(hC.transfer_from(dC_in_place));
|
|
// dA was clobbered by dC_in_place, so copy hA back to dA
|
|
CHECK_HIP_ERROR(dA.transfer_from(hA));
|
|
|
|
// reference calculation
|
|
for(size_t b = 0; b < batch_count; b++)
|
|
{
|
|
auto hA_copy_p = hA_copy[b];
|
|
auto hB_copy_p = hB_copy[b];
|
|
auto hC_gold_p = hC_gold[b];
|
|
|
|
ref_geam(transA,
|
|
transB,
|
|
M,
|
|
N,
|
|
(T*)h_alpha,
|
|
(T*)hA_copy_p,
|
|
lda,
|
|
(T*)h_beta,
|
|
(T*)hB_copy_p,
|
|
ldb,
|
|
(T*)hC_gold_p,
|
|
ldc);
|
|
}
|
|
|
|
if(arg.unit_check)
|
|
{
|
|
unit_check_general<T>(M, N, ldc, hC_gold, hC, batch_count);
|
|
}
|
|
|
|
if(arg.norm_check)
|
|
{
|
|
rocblas_error = norm_check_general<T>('F', M, N, ldc, hC_gold, hC, batch_count);
|
|
}
|
|
}
|
|
}
|
|
|
|
// inplace check for dC == dB
|
|
if(arg.pointer_mode_host)
|
|
{
|
|
bool invalid_size_in_place = ldb != ldc || transB != rocblas_operation_none;
|
|
|
|
if((ldb == ldc) && (transB == rocblas_operation_none))
|
|
CHECK_HIP_ERROR(dC_in_place.transfer_from(hB));
|
|
|
|
CHECK_ROCBLAS_ERROR(rocblas_set_pointer_mode(handle, rocblas_pointer_mode_host));
|
|
|
|
DAPI_EXPECT(invalid_size_in_place ? rocblas_status_invalid_size
|
|
: rocblas_status_success,
|
|
rocblas_geam_batched_fn,
|
|
(handle,
|
|
transA,
|
|
transB,
|
|
M,
|
|
N,
|
|
&alpha,
|
|
dA.ptr_on_device(),
|
|
lda,
|
|
&beta,
|
|
dC_in_place.ptr_on_device(),
|
|
ldb,
|
|
dC_in_place.ptr_on_device(),
|
|
ldc,
|
|
batch_count));
|
|
|
|
if(!invalid_size_in_place)
|
|
{
|
|
CHECK_HIP_ERROR(hC.transfer_from(dC_in_place));
|
|
|
|
// reference calculation
|
|
for(size_t b = 0; b < batch_count; b++)
|
|
{
|
|
auto hA_copy_p = hA_copy[b];
|
|
auto hB_copy_p = hB_copy[b];
|
|
auto hC_gold_p = hC_gold[b];
|
|
|
|
ref_geam(transA,
|
|
transB,
|
|
M,
|
|
N,
|
|
(T*)h_alpha,
|
|
(T*)hA_copy_p,
|
|
lda,
|
|
(T*)h_beta,
|
|
(T*)hB_copy_p,
|
|
ldb,
|
|
(T*)hC_gold_p,
|
|
ldc);
|
|
}
|
|
|
|
if(arg.unit_check)
|
|
{
|
|
unit_check_general<T>(M, N, ldc, hC_gold, hC, batch_count);
|
|
}
|
|
|
|
if(arg.norm_check)
|
|
{
|
|
rocblas_error = norm_check_general<T>('F', M, N, ldc, hC_gold, hC, batch_count);
|
|
}
|
|
}
|
|
}
|
|
|
|
} // end of if unit/norm check
|
|
|
|
if(arg.timing)
|
|
{
|
|
double gpu_time_used;
|
|
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 iter = 0; iter < total_calls; iter++)
|
|
{
|
|
if(iter == number_cold_calls)
|
|
gpu_time_used = get_time_us_sync(stream);
|
|
|
|
DAPI_DISPATCH(rocblas_geam_batched_fn,
|
|
(handle,
|
|
transA,
|
|
transB,
|
|
M,
|
|
N,
|
|
&alpha,
|
|
dA.ptr_on_device(),
|
|
lda,
|
|
&beta,
|
|
dB.ptr_on_device(),
|
|
ldb,
|
|
dC.ptr_on_device(),
|
|
ldc,
|
|
batch_count));
|
|
}
|
|
|
|
gpu_time_used = get_time_us_sync(stream) - gpu_time_used; // in microseconds
|
|
|
|
ArgumentModel<e_transA,
|
|
e_transB,
|
|
e_M,
|
|
e_N,
|
|
e_alpha,
|
|
e_lda,
|
|
e_beta,
|
|
e_ldb,
|
|
e_ldc,
|
|
e_batch_count>{}
|
|
.log_args<T>(rocblas_cout,
|
|
arg,
|
|
gpu_time_used,
|
|
geam_gflop_count<T>(M, N),
|
|
ArgumentLogging::NA_value,
|
|
cpu_time_used,
|
|
rocblas_error_1,
|
|
rocblas_error_2);
|
|
}
|
|
}
|