Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Rotg fixes for issue #1577 #1578

Merged
merged 2 commits into from
Oct 30, 2022
Merged

Rotg fixes for issue #1577 #1578

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
a80d58c
ROTG: fixing TPL layer and its ETI, see issue #1577
lucbv Oct 29, 2022
a798539
ROTG: applying clang-format
lucbv Oct 29, 2022
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 2 additions & 2 deletions blas/tpls/KokkosBlas1_rotg_tpl_spec_avail.hpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -139,7 +139,7 @@ KOKKOSBLAS1_ROTG_TPL_SPEC_AVAIL_CUBLAS(double, Kokkos::LayoutRight,
Kokkos::Cuda, Kokkos::CudaSpace)
KOKKOSBLAS1_ROTG_TPL_SPEC_AVAIL_CUBLAS(float, Kokkos::LayoutLeft, Kokkos::Cuda,
Kokkos::CudaSpace)
KOKKOSBLAS1_ROTG_TPL_SPEC_AVAIL_CUBLAS(float, Kokkos::LayoutLeft, Kokkos::Cuda,
KOKKOSBLAS1_ROTG_TPL_SPEC_AVAIL_CUBLAS(float, Kokkos::LayoutRight, Kokkos::Cuda,
Kokkos::CudaSpace)
KOKKOSBLAS1_ROTG_TPL_SPEC_AVAIL_CUBLAS(Kokkos::complex<double>,
Kokkos::LayoutLeft, Kokkos::Cuda,
Expand All @@ -160,7 +160,7 @@ KOKKOSBLAS1_ROTG_TPL_SPEC_AVAIL_CUBLAS(double, Kokkos::LayoutRight,
Kokkos::Cuda, Kokkos::CudaUVMSpace)
KOKKOSBLAS1_ROTG_TPL_SPEC_AVAIL_CUBLAS(float, Kokkos::LayoutLeft, Kokkos::Cuda,
Kokkos::CudaUVMSpace)
KOKKOSBLAS1_ROTG_TPL_SPEC_AVAIL_CUBLAS(float, Kokkos::LayoutLeft, Kokkos::Cuda,
KOKKOSBLAS1_ROTG_TPL_SPEC_AVAIL_CUBLAS(float, Kokkos::LayoutRight, Kokkos::Cuda,
Kokkos::CudaUVMSpace)
KOKKOSBLAS1_ROTG_TPL_SPEC_AVAIL_CUBLAS(Kokkos::complex<double>,
Kokkos::LayoutLeft, Kokkos::Cuda,
Expand Down
72 changes: 36 additions & 36 deletions blas/tpls/KokkosBlas1_rotg_tpl_spec_decl.hpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -318,44 +318,44 @@ namespace Impl {
} \
};

#define KOKKOSBLAS1_ZROTG_TPL_SPEC_DECL_CUBLAS(LAYOUT, EXECSPACE, MEMSPACE, \
ETI_SPEC_AVAIL) \
template <class MEMSPACE> \
struct Rotg< \
EXECSPACE, \
Kokkos::View<Kokkos::complex<double>, LAYOUT, \
Kokkos::Device<EXECSPACE, MEMSPACE>, \
Kokkos::MemoryTraits<Kokkos::Unmanaged>>, \
Kokkos::View<double, LAYOUT, Kokkos::Device<EXECSPACE, MEMSPACE>, \
Kokkos::MemoryTraits<Kokkos::Unmanaged>>, \
true, ETI_SPEC_AVAIL> { \
using SViewTypeKokkos::View<Kokkos::complex<double>, LAYOUT, \
Kokkos::Device<EXECSPACE, MEMSPACE>, \
Kokkos::MemoryTraits<Kokkos::Unmanaged>>; \
using MViewType = \
Kokkos::View<double, LAYOUT, Kokkos::Device<EXECSPACE, MEMSPACE>, \
Kokkos::MemoryTraits<Kokkos::Unmanaged>>; \
static void rotg(EXECSPACE const& space, SViewType const& a, \
SViewType const& b, MViewType const& c, \
SViewType const& s) { \
Kokkos::Profiling::pushRegion( \
"KokkosBlas::nrm1[TPL_CUBLAS,complex<double>]"); \
rotg_print_specialization<Kokkos::complex<double>, EXECSPACE>(); \
KokkosBlas::Impl::CudaBlasSingleton& singleton = \
KokkosBlas::Impl::CudaBlasSingleton::singleton(); \
KOKKOS_CUBLAS_SAFE_CALL_IMPL( \
cublasSetStream(singleton.handle, space.cuda_stream())); \
cublasZrotg(singleton.handle, \
reinterpret_cast<cuDoubleComplex*>(a.data()), \
reinterpret_cast<cuDoubleComplex*>(b.data()), c.data(), \
reinterpret_cast<cuDoubleComplex*>(s.data())); \
Kokkos::Profiling::popRegion(); \
} \
#define KOKKOSBLAS1_ZROTG_TPL_SPEC_DECL_CUBLAS(LAYOUT, EXECSPACE, MEMSPACE, \
ETI_SPEC_AVAIL) \
template <> \
struct Rotg< \
EXECSPACE, \
Kokkos::View<Kokkos::complex<double>, LAYOUT, \
Kokkos::Device<EXECSPACE, MEMSPACE>, \
Kokkos::MemoryTraits<Kokkos::Unmanaged>>, \
Kokkos::View<double, LAYOUT, Kokkos::Device<EXECSPACE, MEMSPACE>, \
Kokkos::MemoryTraits<Kokkos::Unmanaged>>, \
true, ETI_SPEC_AVAIL> { \
using SViewType = Kokkos::View<Kokkos::complex<double>, LAYOUT, \
Kokkos::Device<EXECSPACE, MEMSPACE>, \
Kokkos::MemoryTraits<Kokkos::Unmanaged>>; \
using MViewType = \
Kokkos::View<double, LAYOUT, Kokkos::Device<EXECSPACE, MEMSPACE>, \
Kokkos::MemoryTraits<Kokkos::Unmanaged>>; \
static void rotg(EXECSPACE const& space, SViewType const& a, \
SViewType const& b, MViewType const& c, \
SViewType const& s) { \
Kokkos::Profiling::pushRegion( \
"KokkosBlas::nrm1[TPL_CUBLAS,complex<double>]"); \
rotg_print_specialization<Kokkos::complex<double>, EXECSPACE>(); \
KokkosBlas::Impl::CudaBlasSingleton& singleton = \
KokkosBlas::Impl::CudaBlasSingleton::singleton(); \
KOKKOS_CUBLAS_SAFE_CALL_IMPL( \
cublasSetStream(singleton.handle, space.cuda_stream())); \
cublasZrotg(singleton.handle, \
reinterpret_cast<cuDoubleComplex*>(a.data()), \
reinterpret_cast<cuDoubleComplex*>(b.data()), c.data(), \
reinterpret_cast<cuDoubleComplex*>(s.data())); \
Kokkos::Profiling::popRegion(); \
} \
};

#define KOKKOSBLAS1_CROTG_TPL_SPEC_DECL_CUBLAS(LAYOUT, EXECSPACE, MEMSPACE, \
ETI_SPEC_AVAIL) \
template <class MEMSPACE> \
template <> \
struct Rotg<EXECSPACE, \
Kokkos::View<Kokkos::complex<float>, LAYOUT, \
Kokkos::Device<EXECSPACE, MEMSPACE>, \
Expand Down Expand Up @@ -529,7 +529,7 @@ namespace Impl {

#define KOKKOSBLAS1_ZROTG_TPL_SPEC_DECL_ROCBLAS(LAYOUT, EXECSPACE, MEMSPACE, \
ETI_SPEC_AVAIL) \
template <class MEMSPACE> \
template <> \
struct Rotg< \
EXECSPACE, \
Kokkos::View<Kokkos::complex<double>, LAYOUT, \
Expand Down Expand Up @@ -565,7 +565,7 @@ namespace Impl {

#define KOKKOSBLAS1_CROTG_TPL_SPEC_DECL_ROCBLAS(LAYOUT, EXECSPACE, MEMSPACE, \
ETI_SPEC_AVAIL) \
template <class MEMSPACE> \
template <> \
struct Rotg<EXECSPACE, \
Kokkos::View<Kokkos::complex<float>, LAYOUT, \
Kokkos::Device<EXECSPACE, MEMSPACE>, \
Expand Down