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
936 lines
34 KiB
C++
936 lines
34 KiB
C++
/* ************************************************************************
|
|
* Copyright (C) 2026 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
|
|
* copies 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
|
|
* IMPLIED, 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 CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
|
|
* THE SOFTWARE.
|
|
*
|
|
* ************************************************************************ */
|
|
#include "rocsparse_clients_dnvec_descr.hpp"
|
|
#include "rocsparse_clients_objects.hpp"
|
|
#include "rocsparse_clients_spmat_descr.hpp"
|
|
#include "rocsparse_clients_sptrsv.hpp"
|
|
#include "testing.hpp"
|
|
|
|
template <typename I, typename J, typename T>
|
|
static void host_csr_lsolve(J M,
|
|
T alpha,
|
|
const I* csr_row_ptr,
|
|
const J* csr_col_ind,
|
|
const T* csr_val,
|
|
const T* x,
|
|
int64_t x_inc,
|
|
T* y,
|
|
rocsparse_diag_type diag_type,
|
|
rocsparse_index_base base,
|
|
int64_t* struct_pivot,
|
|
int64_t* numeric_pivot)
|
|
{
|
|
ROCSPARSE_CLIENTS_ROUTINE_TRACE;
|
|
|
|
// Get device properties
|
|
int dev;
|
|
hipDeviceProp_t prop;
|
|
|
|
std::ignore = hipGetDevice(&dev);
|
|
std::ignore = hipGetDeviceProperties(&prop, dev);
|
|
|
|
std::vector<T> temp(prop.warpSize);
|
|
|
|
// Process lower triangular part
|
|
for(J row = 0; row < M; ++row)
|
|
{
|
|
temp.assign(prop.warpSize, static_cast<T>(0));
|
|
temp[0] = alpha * x[x_inc * row];
|
|
|
|
I diag = -1;
|
|
I row_begin = csr_row_ptr[row] - base;
|
|
I row_end = csr_row_ptr[row + 1] - base;
|
|
|
|
T diag_val = static_cast<T>(0);
|
|
|
|
for(I l = row_begin; l < row_end; l += prop.warpSize)
|
|
{
|
|
for(uint32_t k = 0; k < prop.warpSize; ++k)
|
|
{
|
|
I j = l + k;
|
|
|
|
// Do not run out of bounds
|
|
if(j >= row_end)
|
|
{
|
|
break;
|
|
}
|
|
|
|
J local_col = csr_col_ind[j] - base;
|
|
T local_val = csr_val[j];
|
|
|
|
if(local_val == static_cast<T>(0) && local_col == row
|
|
&& diag_type == rocsparse_diag_type_non_unit)
|
|
{
|
|
// Numerical zero pivot found, avoid division by 0
|
|
// and store index for later use.
|
|
*numeric_pivot = std::min(*numeric_pivot, int64_t(row + base));
|
|
local_val = static_cast<T>(1);
|
|
}
|
|
|
|
// Ignore all entries that are above the diagonal
|
|
if(local_col > row)
|
|
{
|
|
break;
|
|
}
|
|
|
|
// Diagonal entry
|
|
if(local_col == row)
|
|
{
|
|
// If diagonal type is non unit, do division by diagonal entry
|
|
// This is not required for unit diagonal for obvious reasons
|
|
if(diag_type == rocsparse_diag_type_non_unit)
|
|
{
|
|
diag = j;
|
|
diag_val = static_cast<T>(1) / local_val;
|
|
}
|
|
|
|
break;
|
|
}
|
|
|
|
// Lower triangular part
|
|
temp[k] = std::fma(-local_val, y[local_col], temp[k]);
|
|
}
|
|
}
|
|
|
|
for(uint32_t j = 1; j < prop.warpSize; j <<= 1)
|
|
{
|
|
for(uint32_t k = 0; k < prop.warpSize - j; ++k)
|
|
{
|
|
temp[k] += temp[k + j];
|
|
}
|
|
}
|
|
|
|
if(diag_type == rocsparse_diag_type_non_unit)
|
|
{
|
|
if(diag == -1)
|
|
{
|
|
*struct_pivot = std::min(*struct_pivot, int64_t(row + base));
|
|
}
|
|
|
|
y[row] = temp[0] * diag_val;
|
|
}
|
|
else
|
|
{
|
|
y[row] = temp[0];
|
|
}
|
|
}
|
|
}
|
|
|
|
template <typename I, typename J, typename T>
|
|
static void host_csr_usolve(J M,
|
|
T alpha,
|
|
const I* csr_row_ptr,
|
|
const J* csr_col_ind,
|
|
const T* csr_val,
|
|
const T* x,
|
|
int64_t x_inc,
|
|
T* y,
|
|
rocsparse_diag_type diag_type,
|
|
rocsparse_index_base base,
|
|
int64_t* struct_pivot,
|
|
int64_t* numeric_pivot)
|
|
{
|
|
ROCSPARSE_CLIENTS_ROUTINE_TRACE;
|
|
|
|
// Get device properties
|
|
int dev;
|
|
hipDeviceProp_t prop;
|
|
|
|
std::ignore = hipGetDevice(&dev);
|
|
std::ignore = hipGetDeviceProperties(&prop, dev);
|
|
|
|
std::vector<T> temp(prop.warpSize);
|
|
|
|
// Process upper triangular part
|
|
for(J row = M - 1; row >= 0; --row)
|
|
{
|
|
|
|
temp.assign(prop.warpSize, static_cast<T>(0));
|
|
temp[0] = alpha * x[x_inc * row];
|
|
|
|
I diag = -1;
|
|
I row_begin = csr_row_ptr[row] - base;
|
|
I row_end = csr_row_ptr[row + 1] - base;
|
|
|
|
T diag_val = static_cast<T>(0);
|
|
|
|
for(I l = row_end - 1; l >= row_begin; l -= prop.warpSize)
|
|
{
|
|
for(uint32_t k = 0; k < prop.warpSize; ++k)
|
|
{
|
|
I j = l - k;
|
|
|
|
// Do not run out of bounds
|
|
if(j < row_begin)
|
|
{
|
|
break;
|
|
}
|
|
|
|
J local_col = csr_col_ind[j] - base;
|
|
T local_val = csr_val[j];
|
|
|
|
// Ignore all entries that are below the diagonal
|
|
if(local_col < row)
|
|
{
|
|
continue;
|
|
}
|
|
|
|
// Diagonal entry
|
|
if(local_col == row)
|
|
{
|
|
if(diag_type == rocsparse_diag_type_non_unit)
|
|
{
|
|
// Check for numerical zero
|
|
if(local_val == static_cast<T>(0))
|
|
{
|
|
*numeric_pivot = std::min(*numeric_pivot, int64_t(row + base));
|
|
local_val = static_cast<T>(1);
|
|
}
|
|
|
|
diag = j;
|
|
diag_val = static_cast<T>(1) / local_val;
|
|
}
|
|
|
|
continue;
|
|
}
|
|
|
|
// Upper triangular part
|
|
temp[k] = std::fma(-local_val, y[local_col], temp[k]);
|
|
}
|
|
}
|
|
|
|
for(uint32_t j = 1; j < prop.warpSize; j <<= 1)
|
|
{
|
|
for(uint32_t k = 0; k < prop.warpSize - j; ++k)
|
|
{
|
|
temp[k] += temp[k + j];
|
|
}
|
|
}
|
|
|
|
if(diag_type == rocsparse_diag_type_non_unit)
|
|
{
|
|
if(diag == -1)
|
|
{
|
|
*struct_pivot = std::min(*struct_pivot, int64_t(row + base));
|
|
}
|
|
|
|
y[row] = temp[0] * diag_val;
|
|
}
|
|
else
|
|
{
|
|
y[row] = temp[0];
|
|
}
|
|
}
|
|
}
|
|
|
|
template <typename I, typename J, typename T>
|
|
void cpu_csrsv(rocsparse_operation trans,
|
|
J M,
|
|
I nnz,
|
|
T alpha,
|
|
const I* csr_row_ptr,
|
|
const J* csr_col_ind,
|
|
const T* csr_val,
|
|
const T* x,
|
|
int64_t x_inc,
|
|
T* y,
|
|
rocsparse_diag_type diag_type,
|
|
rocsparse_fill_mode fill_mode,
|
|
rocsparse_index_base base,
|
|
int64_t* struct_pivot,
|
|
int64_t* numeric_pivot)
|
|
{
|
|
ROCSPARSE_CLIENTS_ROUTINE_TRACE;
|
|
|
|
// Initialize pivot
|
|
*struct_pivot = M + 1;
|
|
*numeric_pivot = M + 1;
|
|
|
|
if(trans == rocsparse_operation_none)
|
|
{
|
|
if(fill_mode == rocsparse_fill_mode_lower)
|
|
{
|
|
host_csr_lsolve(M,
|
|
alpha,
|
|
csr_row_ptr,
|
|
csr_col_ind,
|
|
csr_val,
|
|
x,
|
|
x_inc,
|
|
y,
|
|
diag_type,
|
|
base,
|
|
struct_pivot,
|
|
numeric_pivot);
|
|
}
|
|
else
|
|
{
|
|
host_csr_usolve(M,
|
|
alpha,
|
|
csr_row_ptr,
|
|
csr_col_ind,
|
|
csr_val,
|
|
x,
|
|
x_inc,
|
|
y,
|
|
diag_type,
|
|
base,
|
|
struct_pivot,
|
|
numeric_pivot);
|
|
}
|
|
}
|
|
else if(trans == rocsparse_operation_transpose
|
|
|| trans == rocsparse_operation_conjugate_transpose)
|
|
{
|
|
// Transpose matrix
|
|
std::vector<I> csrt_row_ptr(M + 1);
|
|
std::vector<J> csrt_col_ind(nnz);
|
|
std::vector<T> csrt_val(nnz);
|
|
|
|
host_csr_to_csc(M,
|
|
M,
|
|
nnz,
|
|
csr_row_ptr,
|
|
csr_col_ind,
|
|
csr_val,
|
|
csrt_col_ind,
|
|
csrt_row_ptr,
|
|
csrt_val,
|
|
rocsparse_action_numeric,
|
|
base);
|
|
|
|
if(trans == rocsparse_operation_conjugate_transpose)
|
|
{
|
|
for(size_t i = 0; i < csrt_val.size(); i++)
|
|
{
|
|
csrt_val[i] = rocsparse_conj(csrt_val[i]);
|
|
}
|
|
}
|
|
|
|
if(fill_mode == rocsparse_fill_mode_lower)
|
|
{
|
|
host_csr_usolve(M,
|
|
alpha,
|
|
csrt_row_ptr.data(),
|
|
csrt_col_ind.data(),
|
|
csrt_val.data(),
|
|
x,
|
|
x_inc,
|
|
y,
|
|
diag_type,
|
|
base,
|
|
struct_pivot,
|
|
numeric_pivot);
|
|
}
|
|
else
|
|
{
|
|
host_csr_lsolve(M,
|
|
alpha,
|
|
csrt_row_ptr.data(),
|
|
csrt_col_ind.data(),
|
|
csrt_val.data(),
|
|
x,
|
|
x_inc,
|
|
y,
|
|
diag_type,
|
|
base,
|
|
struct_pivot,
|
|
numeric_pivot);
|
|
}
|
|
}
|
|
|
|
*numeric_pivot = std::min(*numeric_pivot, *struct_pivot);
|
|
|
|
*struct_pivot = (*struct_pivot == M + 1) ? -1 : *struct_pivot;
|
|
*numeric_pivot = (*numeric_pivot == M + 1) ? -1 : *numeric_pivot;
|
|
}
|
|
|
|
template <typename I, typename T>
|
|
void cpu_coosv(rocsparse_operation trans,
|
|
I M,
|
|
int64_t nnz,
|
|
T alpha,
|
|
const I* coo_row_ind,
|
|
const I* coo_col_ind,
|
|
const T* coo_val,
|
|
const T* x,
|
|
T* y,
|
|
rocsparse_diag_type diag_type,
|
|
rocsparse_fill_mode fill_mode,
|
|
rocsparse_index_base base,
|
|
int64_t* struct_pivot,
|
|
int64_t* numeric_pivot)
|
|
{
|
|
ROCSPARSE_CLIENTS_ROUTINE_TRACE;
|
|
|
|
if(std::is_same<I, int32_t>() && nnz < std::numeric_limits<int32_t>::max())
|
|
{
|
|
std::vector<int32_t> csr_row_ptr(M + 1);
|
|
|
|
host_coo_to_csr<int32_t, I>(M, nnz, coo_row_ind, csr_row_ptr.data(), base);
|
|
|
|
cpu_csrsv<int32_t, I>(trans,
|
|
M,
|
|
nnz,
|
|
alpha,
|
|
csr_row_ptr.data(),
|
|
coo_col_ind,
|
|
coo_val,
|
|
x,
|
|
(int64_t)1,
|
|
y,
|
|
diag_type,
|
|
fill_mode,
|
|
base,
|
|
struct_pivot,
|
|
numeric_pivot);
|
|
}
|
|
else
|
|
{
|
|
std::vector<int64_t> csr_row_ptr(M + 1);
|
|
|
|
host_coo_to_csr(M, nnz, coo_row_ind, csr_row_ptr.data(), base);
|
|
|
|
cpu_csrsv(trans,
|
|
M,
|
|
nnz,
|
|
alpha,
|
|
csr_row_ptr.data(),
|
|
coo_col_ind,
|
|
coo_val,
|
|
x,
|
|
(int64_t)1,
|
|
y,
|
|
diag_type,
|
|
fill_mode,
|
|
base,
|
|
struct_pivot,
|
|
numeric_pivot);
|
|
}
|
|
}
|
|
|
|
namespace rocsparse_clients
|
|
{
|
|
|
|
class sptrsv_descr
|
|
{
|
|
private:
|
|
rocsparse_handle m_handle{};
|
|
rocsparse_sptrsv_descr m_descr{};
|
|
|
|
public:
|
|
struct config_t
|
|
{
|
|
rocsparse_operation operation;
|
|
rocsparse_sptrsv_alg alg;
|
|
rocsparse_datatype scalar_datatype;
|
|
rocsparse_datatype compute_datatype;
|
|
rocsparse_analysis_policy apol;
|
|
};
|
|
|
|
config_t config;
|
|
|
|
void set(rocsparse_handle handle)
|
|
{
|
|
rocsparse_create_sptrsv_descr(&this->m_descr);
|
|
rocsparse_error p_error[1] = {nullptr};
|
|
rocsparse_sptrsv_set_input(handle,
|
|
this->m_descr,
|
|
rocsparse_sptrsv_input_operation,
|
|
&config.operation,
|
|
sizeof(config.operation),
|
|
p_error);
|
|
|
|
rocsparse_sptrsv_set_input(handle,
|
|
this->m_descr,
|
|
rocsparse_sptrsv_input_alg,
|
|
&config.alg,
|
|
sizeof(config.alg),
|
|
p_error);
|
|
|
|
rocsparse_sptrsv_set_input(handle,
|
|
this->m_descr,
|
|
rocsparse_sptrsv_input_scalar_datatype,
|
|
&config.scalar_datatype,
|
|
sizeof(config.scalar_datatype),
|
|
p_error);
|
|
|
|
rocsparse_sptrsv_set_input(handle,
|
|
this->m_descr,
|
|
rocsparse_sptrsv_input_compute_datatype,
|
|
&config.compute_datatype,
|
|
sizeof(config.compute_datatype),
|
|
p_error);
|
|
|
|
rocsparse_sptrsv_set_input(handle,
|
|
this->m_descr,
|
|
rocsparse_sptrsv_input_analysis_policy,
|
|
&config.apol,
|
|
sizeof(config.apol),
|
|
p_error);
|
|
}
|
|
|
|
sptrsv_descr(rocsparse_handle handle,
|
|
int64_t batch_count,
|
|
const rocsparse_operation operation,
|
|
const rocsparse_sptrsv_alg alg,
|
|
const rocsparse_datatype scalar_datatype,
|
|
const rocsparse_datatype compute_datatype,
|
|
const rocsparse_analysis_policy apol)
|
|
: m_handle(handle)
|
|
, config({operation, alg, scalar_datatype, compute_datatype, apol})
|
|
{
|
|
ROCSPARSE_CLIENTS_ROUTINE_TRACE;
|
|
rocsparse_error* p_error = nullptr;
|
|
const rocsparse_status status
|
|
= rocsparse_sptrsv_descr_create(handle, &this->m_descr, p_error);
|
|
if(status != rocsparse_status_success)
|
|
{
|
|
throw(status);
|
|
}
|
|
this->set(handle);
|
|
}
|
|
|
|
~sptrsv_descr()
|
|
{
|
|
ROCSPARSE_CLIENTS_ROUTINE_TRACE;
|
|
rocsparse_error* p_error = nullptr;
|
|
std::ignore = rocsparse_sptrsv_descr_destroy(this->m_handle, this->m_descr, p_error);
|
|
}
|
|
|
|
inline operator rocsparse_sptrsv_descr&()
|
|
{
|
|
return this->m_descr;
|
|
}
|
|
|
|
inline operator const rocsparse_sptrsv_descr&() const
|
|
{
|
|
return this->m_descr;
|
|
}
|
|
};
|
|
|
|
template <typename T, typename I, typename J = I>
|
|
void sptrsv_host(int64_t batch_count,
|
|
rocsparse_clients::sptrsv_descr& sptrsv_descr,
|
|
const T* halpha,
|
|
rocsparse_operation operation,
|
|
rocsparse_clients::spmat_descr<T, I, J>& A,
|
|
const rocsparse_clients::dnvec_descr<T>& x,
|
|
rocsparse_clients::dnvec_descr<T>& y,
|
|
const rocsparse_diag_type diag,
|
|
const rocsparse_fill_mode uplo,
|
|
int64_t* symbolic,
|
|
int64_t* exact)
|
|
{
|
|
const rocsparse_format format = A.get_format();
|
|
switch(format)
|
|
{
|
|
case rocsparse_format_coo:
|
|
{
|
|
auto& host = A.template as<rocsparse_format_coo>().host();
|
|
|
|
for(int64_t i = 0; i < batch_count; ++i)
|
|
{
|
|
const T* p = host.val.data() + i * A.get_stride();
|
|
const T* p_hx = x.host().data() + i * x.get_stride();
|
|
T* p_hy = y.host().data() + i * y.get_stride();
|
|
|
|
cpu_coosv<I, T>(operation,
|
|
host.m,
|
|
host.nnz,
|
|
*halpha,
|
|
host.row_ind,
|
|
host.col_ind,
|
|
p,
|
|
p_hx,
|
|
p_hy,
|
|
diag,
|
|
uplo,
|
|
host.base,
|
|
symbolic + i,
|
|
exact + i);
|
|
}
|
|
break;
|
|
}
|
|
case rocsparse_format_csr:
|
|
{
|
|
auto& host = A.template as<rocsparse_format_csr>().host();
|
|
for(int64_t i = 0; i < batch_count; ++i)
|
|
{
|
|
const T* p = host.val.data() + i * A.get_stride();
|
|
const T* p_hx = x.host().data() + i * x.get_stride();
|
|
T* p_hy = y.host().data() + i * y.get_stride();
|
|
cpu_csrsv<I, J, T>(operation,
|
|
host.m,
|
|
host.nnz,
|
|
*halpha,
|
|
host.ptr,
|
|
host.ind,
|
|
p,
|
|
p_hx,
|
|
(int64_t)1,
|
|
p_hy,
|
|
diag,
|
|
uplo,
|
|
host.base,
|
|
symbolic + i,
|
|
exact + i);
|
|
}
|
|
|
|
break;
|
|
}
|
|
|
|
case rocsparse_format_csc:
|
|
{
|
|
#ifdef ROCSPARSE_WITH_CSC_TRSV
|
|
// A CSC matrix is the transpose of the CSR matrix sharing the same
|
|
// arrays, so host_cscsv forwards to host_csrsv with the transpose
|
|
// operation and the fill mode flipped (and, for the conjugate
|
|
// transpose, the values conjugated). Each batch entry is solved
|
|
// independently using the per-batch strides of the matrix values and
|
|
// of the x / y dense vectors.
|
|
auto& host = A.template as<rocsparse_format_csc>().host();
|
|
for(int64_t i = 0; i < batch_count; ++i)
|
|
{
|
|
const T* p = host.val.data() + i * A.get_stride();
|
|
const T* p_hx = x.host().data() + i * x.get_stride();
|
|
T* p_hy = y.host().data() + i * y.get_stride();
|
|
|
|
J analysis_pivot = -1;
|
|
J solve_pivot = -1;
|
|
host_cscsv<I, J, T>(operation,
|
|
host.m,
|
|
host.nnz,
|
|
*halpha,
|
|
host.ptr,
|
|
host.ind,
|
|
p,
|
|
p_hx,
|
|
(int64_t)1,
|
|
p_hy,
|
|
diag,
|
|
uplo,
|
|
host.base,
|
|
&analysis_pivot,
|
|
&solve_pivot);
|
|
|
|
symbolic[i] = analysis_pivot;
|
|
exact[i] = solve_pivot;
|
|
}
|
|
#endif
|
|
break;
|
|
}
|
|
|
|
case rocsparse_format_bsr:
|
|
case rocsparse_format_ell:
|
|
case rocsparse_format_sell:
|
|
case rocsparse_format_bell:
|
|
case rocsparse_format_coo_aos:
|
|
{
|
|
break;
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
template <typename I, typename J, typename T>
|
|
void testing_sptrsv_bad_arg(const Arguments& arg)
|
|
{
|
|
}
|
|
|
|
template <typename I, typename J, typename T>
|
|
void testing_sptrsv(const Arguments& arg)
|
|
{
|
|
|
|
rocsparse_error* p_error = nullptr;
|
|
|
|
if(arg.M != arg.N)
|
|
{
|
|
return;
|
|
}
|
|
|
|
#ifndef ROCSPARSE_WITH_CSC_TRSV
|
|
// CSC triangular solve support can be disabled at build time
|
|
// (BUILD_WITH_CSC_TRSV=OFF); skip the CSC cases in that configuration.
|
|
if(arg.formatA == rocsparse_format_csc)
|
|
{
|
|
return;
|
|
}
|
|
#endif
|
|
|
|
const int64_t batch_count = (arg.batch_count > 1) ? arg.batch_count : 1;
|
|
const int64_t batch_count_A = (arg.batch_count_A > 0) ? arg.batch_count_A : batch_count;
|
|
const int64_t batch_count_x = (arg.batch_count_B > 0) ? arg.batch_count_B : batch_count;
|
|
|
|
bool full_rank = true;
|
|
|
|
rocsparse_clients::spmat_descr<T, I, J> A(arg, batch_count_A, full_rank);
|
|
if(false == A.is_square())
|
|
{
|
|
return;
|
|
}
|
|
const int64_t M = A.get_nrows();
|
|
rocsparse_clients::dnvec_descr<T> x(M, batch_count_x, M);
|
|
rocsparse_clients::dnvec_descr<T> y(M, batch_count, M);
|
|
|
|
host_scalar<T> halpha(arg.get_alpha<T>());
|
|
device_scalar<T> dalpha(halpha);
|
|
|
|
const rocsparse_analysis_policy apol = arg.apol;
|
|
const rocsparse_datatype ttype = get_datatype<T>();
|
|
const rocsparse_operation operation = arg.transA;
|
|
const rocsparse_sptrsv_alg alg = arg.sptrsv_alg;
|
|
const rocsparse_diag_type diag = arg.diag;
|
|
const rocsparse_fill_mode uplo = arg.uplo;
|
|
const rocsparse_matrix_type matrix_type = arg.matrix_type;
|
|
|
|
CHECK_ROCSPARSE_ERROR(
|
|
rocsparse_spmat_set_attribute(A, rocsparse_spmat_fill_mode, &uplo, sizeof(uplo)));
|
|
CHECK_ROCSPARSE_ERROR(
|
|
rocsparse_spmat_set_attribute(A, rocsparse_spmat_diag_type, &diag, sizeof(diag)));
|
|
CHECK_ROCSPARSE_ERROR(rocsparse_spmat_set_attribute(
|
|
A, rocsparse_spmat_matrix_type, &matrix_type, sizeof(matrix_type)));
|
|
|
|
//
|
|
// Create handle.
|
|
//
|
|
rocsparse_local_handle handle(arg);
|
|
hipStream_t stream{};
|
|
CHECK_ROCSPARSE_ERROR(rocsparse_get_stream(handle, &stream));
|
|
rocsparse_clients::sptrsv_descr sptrsv_descr(
|
|
handle, batch_count, operation, alg, ttype, ttype, apol);
|
|
|
|
rocsparse_clients::sptrsv_analysis(handle, sptrsv_descr, A, x, y, p_error);
|
|
|
|
host_dense_vector<int64_t> host_symbolic_position(batch_count);
|
|
CHECK_ROCSPARSE_ERROR(rocsparse_set_pointer_mode(handle, rocsparse_pointer_mode_host));
|
|
CHECK_ROCSPARSE_ERROR(rocsparse_sptrsv_get_output(handle,
|
|
sptrsv_descr,
|
|
rocsparse_sptrsv_output_singularity_position,
|
|
host_symbolic_position,
|
|
sizeof(int64_t),
|
|
p_error));
|
|
|
|
{
|
|
device_dense_vector<int64_t> device_symbolic_position(batch_count);
|
|
CHECK_ROCSPARSE_ERROR(rocsparse_set_pointer_mode(handle, rocsparse_pointer_mode_device));
|
|
CHECK_ROCSPARSE_ERROR(
|
|
rocsparse_sptrsv_get_output(handle,
|
|
sptrsv_descr,
|
|
rocsparse_sptrsv_output_singularity_position,
|
|
device_symbolic_position,
|
|
sizeof(int64_t),
|
|
p_error));
|
|
CHECK_HIP_ERROR(hipStreamSynchronize(stream));
|
|
host_symbolic_position.unit_check(device_symbolic_position);
|
|
}
|
|
|
|
if(arg.unit_check)
|
|
{
|
|
host_dense_vector<int64_t> cpu_symbolic_position(batch_count);
|
|
host_dense_vector<int64_t> cpu_numeric_position(batch_count);
|
|
|
|
rocsparse_clients::sptrsv_host<T, I, J>(batch_count,
|
|
sptrsv_descr,
|
|
halpha,
|
|
operation,
|
|
A,
|
|
x,
|
|
y,
|
|
diag,
|
|
uplo,
|
|
cpu_symbolic_position,
|
|
cpu_numeric_position);
|
|
|
|
for(auto mode : {rocsparse_pointer_mode_host, rocsparse_pointer_mode_device})
|
|
{
|
|
void* alpha = (mode == rocsparse_pointer_mode_host) ? halpha : dalpha;
|
|
|
|
rocsparse_clients::sptrsv_compute(handle, sptrsv_descr, A, x, y, mode, alpha, p_error);
|
|
|
|
host_dense_vector<int64_t> host_numeric_position(batch_count);
|
|
CHECK_ROCSPARSE_ERROR(rocsparse_set_pointer_mode(handle, rocsparse_pointer_mode_host));
|
|
CHECK_ROCSPARSE_ERROR(
|
|
rocsparse_sptrsv_get_output(handle,
|
|
sptrsv_descr,
|
|
rocsparse_sptrsv_output_singularity_position,
|
|
host_numeric_position,
|
|
sizeof(int64_t),
|
|
p_error));
|
|
{
|
|
device_dense_vector<int64_t> device_numeric_position(batch_count);
|
|
|
|
CHECK_ROCSPARSE_ERROR(
|
|
rocsparse_set_pointer_mode(handle, rocsparse_pointer_mode_device));
|
|
CHECK_ROCSPARSE_ERROR(
|
|
rocsparse_sptrsv_get_output(handle,
|
|
sptrsv_descr,
|
|
rocsparse_sptrsv_output_singularity_position,
|
|
device_numeric_position,
|
|
sizeof(int64_t),
|
|
p_error));
|
|
|
|
CHECK_HIP_ERROR(hipStreamSynchronize(stream));
|
|
host_numeric_position.unit_check(device_numeric_position);
|
|
}
|
|
|
|
cpu_symbolic_position.unit_check(host_symbolic_position);
|
|
cpu_numeric_position.unit_check(host_numeric_position);
|
|
|
|
if(ROCSPARSE_REPRODUCIBILITY)
|
|
{
|
|
if(rocsparse_pointer_mode_host == mode)
|
|
{
|
|
rocsparse_reproducibility::save("Y pointer mode host", y.device());
|
|
}
|
|
else
|
|
{
|
|
rocsparse_reproducibility::save("Y pointer mode device", y.device());
|
|
}
|
|
}
|
|
|
|
y.near_check_values(host_symbolic_position, host_numeric_position);
|
|
}
|
|
}
|
|
|
|
if(arg.timing)
|
|
{
|
|
size_t buffer_size;
|
|
CHECK_ROCSPARSE_ERROR(rocsparse_sptrsv_buffer_size(
|
|
handle, sptrsv_descr, A, x, y, rocsparse_sptrsv_stage_compute, &buffer_size, p_error));
|
|
device_dense_vector<char> buffer(buffer_size);
|
|
|
|
CHECK_ROCSPARSE_ERROR(rocsparse_set_pointer_mode(handle, rocsparse_pointer_mode_host));
|
|
CHECK_ROCSPARSE_ERROR(rocsparse_sptrsv_set_input(handle,
|
|
sptrsv_descr,
|
|
rocsparse_sptrsv_input_scalar_alpha,
|
|
halpha,
|
|
sizeof(halpha.data()),
|
|
p_error));
|
|
|
|
const double gpu_time_used
|
|
= rocsparse_clients::run_benchmark(arg,
|
|
rocsparse_sptrsv,
|
|
handle,
|
|
sptrsv_descr,
|
|
A,
|
|
x,
|
|
y,
|
|
rocsparse_sptrsv_stage_compute,
|
|
buffer_size,
|
|
buffer,
|
|
p_error);
|
|
|
|
int64_t A_m = A.get_nrows();
|
|
int64_t A_nnz = 0;
|
|
int64_t A_bdim = 1;
|
|
double gbyte_count = 0;
|
|
const rocsparse_format format = A.get_format();
|
|
switch(format)
|
|
{
|
|
|
|
case rocsparse_format_ell:
|
|
case rocsparse_format_bell:
|
|
case rocsparse_format_sell:
|
|
case rocsparse_format_coo_aos:
|
|
{
|
|
break;
|
|
}
|
|
|
|
case rocsparse_format_csc:
|
|
{
|
|
auto& device = A.template as<rocsparse_format_csc>().device();
|
|
gbyte_count = csrsv_gbyte_count<T>(device.m, device.nnz);
|
|
A_nnz = device.nnz;
|
|
break;
|
|
}
|
|
|
|
case rocsparse_format_coo:
|
|
{
|
|
auto& device = A.template as<rocsparse_format_coo>().device();
|
|
gbyte_count = coosv_gbyte_count<T>(device.m, device.nnz);
|
|
A_nnz = device.nnz;
|
|
break;
|
|
}
|
|
|
|
case rocsparse_format_csr:
|
|
{
|
|
auto& device = A.template as<rocsparse_format_csr>().device();
|
|
gbyte_count = csrsv_gbyte_count<T>(device.m, device.nnz);
|
|
A_nnz = device.nnz;
|
|
break;
|
|
}
|
|
|
|
case rocsparse_format_bsr:
|
|
{
|
|
auto& device = A.template as<rocsparse_format_bsr>().device();
|
|
gbyte_count
|
|
= csrsv_gbyte_count<T>(device.mb * device.row_block_dim,
|
|
device.nnzb * device.row_block_dim * device.col_block_dim);
|
|
A_nnz = device.nnzb * device.row_block_dim * device.col_block_dim;
|
|
A_bdim = device.row_block_dim;
|
|
break;
|
|
}
|
|
}
|
|
|
|
const double gflop_count = spsv_gflop_count(A_m, A_nnz, diag);
|
|
const double gpu_gflops = get_gpu_gflops(gpu_time_used, gflop_count);
|
|
const double gpu_gbyte = get_gpu_gbyte(gpu_time_used, gbyte_count);
|
|
display_timing_info(display_key_t::M,
|
|
A_m,
|
|
display_key_t::nnz_A,
|
|
A_nnz,
|
|
display_key_t::bdim_A,
|
|
A_bdim,
|
|
display_key_t::alpha,
|
|
halpha,
|
|
display_key_t::algorithm,
|
|
rocsparse_sptrsvalg2string(alg),
|
|
display_key_t::gflops,
|
|
gpu_gflops,
|
|
display_key_t::bandwidth,
|
|
gpu_gbyte,
|
|
display_key_t::time_ms,
|
|
get_gpu_time_msec(gpu_time_used));
|
|
}
|
|
}
|
|
|
|
#define INSTANTIATE(ITYPE, JTYPE, TTYPE) \
|
|
template void testing_sptrsv_bad_arg<ITYPE, JTYPE, TTYPE>(const Arguments& arg); \
|
|
template void testing_sptrsv<ITYPE, JTYPE, TTYPE>(const Arguments& arg)
|
|
|
|
INSTANTIATE(int32_t, int32_t, float);
|
|
INSTANTIATE(int32_t, int32_t, double);
|
|
INSTANTIATE(int32_t, int32_t, rocsparse_float_complex);
|
|
INSTANTIATE(int32_t, int32_t, rocsparse_double_complex);
|
|
INSTANTIATE(int64_t, int32_t, float);
|
|
INSTANTIATE(int64_t, int32_t, double);
|
|
INSTANTIATE(int64_t, int32_t, rocsparse_float_complex);
|
|
INSTANTIATE(int64_t, int32_t, rocsparse_double_complex);
|
|
INSTANTIATE(int64_t, int64_t, float);
|
|
INSTANTIATE(int64_t, int64_t, double);
|
|
INSTANTIATE(int64_t, int64_t, rocsparse_float_complex);
|
|
INSTANTIATE(int64_t, int64_t, rocsparse_double_complex);
|
|
|
|
void testing_sptrsv_extra(const Arguments& arg) {}
|