haraldwolff/CLBlast
1
0
Fork 0
mirror of https://github.com/CNugteren/CLBlast.git synced 2024-07-15 19:05:44 +02:00
Code Issues Packages Projects Releases Wiki Activity

Made it possible to build the OMATCOPY test and client in case only clBLAS is present

Browse source
This commit is contained in:
CNugteren 2016-06-28 16:36:01 +02:00
parent 9171f1c160
commit 2c031f3e1d
1 changed files with 35 additions and 40 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

75
test/routines/levelx/xomatcopy.hpp
View file

@ -83,52 +83,47 @@ class TestXomatcopy {
// Describes how to run a naive version of the routine (for correctness/performance comparison). // Describes how to run a naive version of the routine (for correctness/performance comparison).
// Note that a proper clBLAS or CPU BLAS comparison is not available for non-BLAS routines. // Note that a proper clBLAS or CPU BLAS comparison is not available for non-BLAS routines.
static StatusCode RunReference1(const Arguments<T> &args, Buffers<T> &buffers, Queue &queue) {
return RunReference2(args, buffers, queue);
}
#ifdef CLBLAST_REF_CLBLAS static StatusCode RunReference2(const Arguments<T> &args, Buffers<T> &buffers, Queue &queue) {
static StatusCode RunReference1(const Arguments<T> &args, Buffers<T> &buffers, Queue &queue) {
return RunReference2(args, buffers, queue);
}
#endif
#ifdef CLBLAST_REF_CBLAS // Data transfer from OpenCL to std::vector
static StatusCode RunReference2(const Arguments<T> &args, Buffers<T> &buffers, Queue &queue) { std::vector<T> a_mat_cpu(args.a_size, static_cast<T>(0));
std::vector<T> b_mat_cpu(args.b_size, static_cast<T>(0));
buffers.a_mat.Read(queue, args.a_size, a_mat_cpu);
buffers.b_mat.Read(queue, args.b_size, b_mat_cpu);
// Data transfer from OpenCL to std::vector // Checking for invalid arguments
std::vector<T> a_mat_cpu(args.a_size, static_cast<T>(0)); const auto a_rotated = (args.layout == Layout::kRowMajor);
std::vector<T> b_mat_cpu(args.b_size, static_cast<T>(0)); const auto b_rotated = (args.layout == Layout::kColMajor && args.a_transpose != Transpose::kNo) ||
buffers.a_mat.Read(queue, args.a_size, a_mat_cpu); (args.layout == Layout::kRowMajor && args.a_transpose == Transpose::kNo);
buffers.b_mat.Read(queue, args.b_size, b_mat_cpu); const auto a_base = (a_rotated) ? args.a_ld*(args.m-1) + args.n : args.a_ld*(args.n-1) + args.m;
const auto b_base = (b_rotated) ? args.b_ld*(args.m-1) + args.n : args.b_ld*(args.n-1) + args.m;
if ((args.m == 0) || (args.n == 0)) { return StatusCode::kInvalidDimension; }
if ((args.a_ld < args.m && !a_rotated) || (args.a_ld < args.n && a_rotated)) { return StatusCode::kInvalidLeadDimA; }
if ((args.b_ld < args.m && !b_rotated) || (args.b_ld < args.n && b_rotated)) { return StatusCode::kInvalidLeadDimB; }
if (buffers.a_mat.GetSize() < (a_base + args.a_offset) * sizeof(T)) { return StatusCode::kInsufficientMemoryA; }
if (buffers.b_mat.GetSize() < (b_base + args.b_offset) * sizeof(T)) { return StatusCode::kInsufficientMemoryB; }
// Checking for invalid arguments // Matrix copy, scaling, and/or transpose
const auto a_rotated = (args.layout == Layout::kRowMajor); for (auto id1 = size_t{0}; id1 < args.m; ++id1) {
const auto b_rotated = (args.layout == Layout::kColMajor && args.a_transpose != Transpose::kNo) || for (auto id2 = size_t{0}; id2 < args.n; ++id2) {
(args.layout == Layout::kRowMajor && args.a_transpose == Transpose::kNo); const auto a_one = (a_rotated) ? id2 : id1;
const auto a_base = (a_rotated) ? args.a_ld*(args.m-1) + args.n : args.a_ld*(args.n-1) + args.m; const auto a_two = (a_rotated) ? id1 : id2;
const auto b_base = (b_rotated) ? args.b_ld*(args.m-1) + args.n : args.b_ld*(args.n-1) + args.m; const auto b_one = (b_rotated) ? id2 : id1;
if ((args.m == 0) || (args.n == 0)) { return StatusCode::kInvalidDimension; } const auto b_two = (b_rotated) ? id1 : id2;
if ((args.a_ld < args.m && !a_rotated) || (args.a_ld < args.n && a_rotated)) { return StatusCode::kInvalidLeadDimA; } const auto a_index = a_two * args.a_ld + a_one + args.a_offset;
if ((args.b_ld < args.m && !b_rotated) || (args.b_ld < args.n && b_rotated)) { return StatusCode::kInvalidLeadDimB; } const auto b_index = b_two * args.b_ld + b_one + args.b_offset;
if (buffers.a_mat.GetSize() < (a_base + args.a_offset) * sizeof(T)) { return StatusCode::kInsufficientMemoryA; } b_mat_cpu[b_index] = args.alpha * a_mat_cpu[a_index];
if (buffers.b_mat.GetSize() < (b_base + args.b_offset) * sizeof(T)) { return StatusCode::kInsufficientMemoryB; }
// Matrix copy, scaling, and/or transpose
for (auto id1 = size_t{0}; id1 < args.m; ++id1) {
for (auto id2 = size_t{0}; id2 < args.n; ++id2) {
const auto a_one = (a_rotated) ? id2 : id1;
const auto a_two = (a_rotated) ? id1 : id2;
const auto b_one = (b_rotated) ? id2 : id1;
const auto b_two = (b_rotated) ? id1 : id2;
const auto a_index = a_two * args.a_ld + a_one + args.a_offset;
const auto b_index = b_two * args.b_ld + b_one + args.b_offset;
b_mat_cpu[b_index] = args.alpha * a_mat_cpu[a_index];
}
} }
// Data transfer back to OpenCL
buffers.b_mat.Write(queue, args.b_size, b_mat_cpu);
return StatusCode::kSuccess;
} }
#endif
// Data transfer back to OpenCL
buffers.b_mat.Write(queue, args.b_size, b_mat_cpu);
return StatusCode::kSuccess;
}
// Describes how to download the results of the computation (more importantly: which buffer) // Describes how to download the results of the computation (more importantly: which buffer)
static std::vector<T> DownloadResult(const Arguments<T> &args, Buffers<T> &buffers, Queue &queue) { static std::vector<T> DownloadResult(const Arguments<T> &args, Buffers<T> &buffers, Queue &queue) {