Skip to content

[Core] Add lowest() to platform numeric_limits<float> (issue #3509) #3621

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

Open
XFDG wants to merge 1 commit into
base: main
Choose a base branch
from
+28 −0
Open

[Core] Add lowest() to platform numeric_limits<float> (issue #3509) #3621

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
2 changes: 2 additions & 0 deletions include/cutlass/platform/platform.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -911,6 +911,8 @@ struct numeric_limits<float> {
static constexpr float infinity() noexcept { return bit_cast<float, int32_t>(0x7f800000);}
CUTLASS_HOST_DEVICE
static constexpr float max() noexcept { return bit_cast<float, int32_t>(0x7f7fffff);}
CUTLASS_HOST_DEVICE
static constexpr float lowest() noexcept { return -max();}
static constexpr bool is_integer = false;
static constexpr bool has_infinity = true;
};
Expand Down
26 changes: 26 additions & 0 deletions test/unit/core/functional.cu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -36,6 +36,7 @@

#include "cutlass/functional.h"
#include "cutlass/core_io.h"
#include "cutlass/epilogue/thread/linear_combination_clamp.h"

#include "cutlass/layout/matrix.h"
#include "cutlass/util/host_tensor.h"
Expand Down Expand Up @@ -428,6 +429,31 @@ TEST(Functional, multiply_add_bf16x17) {

/////////////////////////////////////////////////////////////////////////////////////////////////

TEST(Functional, linear_combination_clamp_float) {
using Fragment = cutlass::Array<float, 1>;
using Operator =
cutlass::epilogue::thread::LinearCombinationClamp<float, 1, float, float>;
using Tensor = cutlass::HostTensor<float, cutlass::layout::RowMajor>;

Fragment accumulator;
accumulator[0] = 2.0f;
EXPECT_EQ(Operator()(accumulator)[0], 2.0f);

Tensor output({1, 1});
Tensor input({1, 1});
input.host_data()[0] = 2.0f;
input.sync_device();

test::core::kernel::unary_operator<Fragment, Operator><<<1, 1>>>(
reinterpret_cast<Fragment *>(output.device_data()),
reinterpret_cast<Fragment const *>(input.device_data()));
output.sync_host();

EXPECT_EQ(output.host_data()[0], 2.0f);
}

/////////////////////////////////////////////////////////////////////////////////////////////////

template <typename T>
cutlass::Quaternion<T> random_quaternion(int range) {
return cutlass::Quaternion<T>{
Expand Down