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
151 lines
4.2 KiB
C++
151 lines
4.2 KiB
C++
// Copyright Advanced Micro Devices, Inc., or its affiliates.
|
|
// SPDX-License-Identifier: MIT
|
|
|
|
#include <rocRoller/Operations/BlockScale.hpp>
|
|
#include <rocRoller/Operations/Command.hpp>
|
|
#include <rocRoller/Operations/Operation.hpp>
|
|
|
|
namespace rocRoller
|
|
{
|
|
namespace Operations
|
|
{
|
|
BlockScale::BlockScale(OperationTag data,
|
|
int dimensions,
|
|
std::optional<OperationTag> scale,
|
|
std::vector<size_t> strides)
|
|
: BaseOperation()
|
|
, m_data(data)
|
|
, m_scale(scale)
|
|
{
|
|
if(!strides.empty())
|
|
{
|
|
m_strides = std::move(strides);
|
|
}
|
|
|
|
if(m_strides.empty())
|
|
{
|
|
// Default value for first stride based on hardware arch
|
|
m_strides.push_back(32);
|
|
}
|
|
|
|
if(m_strides.size() != dimensions)
|
|
{
|
|
m_strides.resize(dimensions, 1);
|
|
}
|
|
}
|
|
|
|
std::unordered_set<OperationTag> BlockScale::getInputs() const
|
|
{
|
|
if(scaleMode() == ScaleMode::Inline)
|
|
return {m_data};
|
|
return {m_data, m_scale.value()};
|
|
}
|
|
|
|
OperationTag BlockScale::data() const
|
|
{
|
|
return m_data;
|
|
}
|
|
|
|
std::optional<OperationTag> BlockScale::scale() const
|
|
{
|
|
return m_scale;
|
|
}
|
|
|
|
std::string BlockScale::toString() const
|
|
{
|
|
std::ostringstream rv;
|
|
|
|
rv << "BlockScale(" << scaleMode() << ", {";
|
|
streamJoin(rv, m_strides, ", ");
|
|
rv << "}): Data: " << m_data;
|
|
|
|
if(m_scale)
|
|
{
|
|
rv << ", Scale: " << *m_scale;
|
|
}
|
|
|
|
return rv.str();
|
|
}
|
|
|
|
ScaleMode BlockScale::scaleMode() const
|
|
{
|
|
if(m_strides.empty())
|
|
return ScaleMode::SingleScale;
|
|
return m_scale.has_value() ? ScaleMode::Separate : ScaleMode::Inline;
|
|
}
|
|
|
|
const std::vector<size_t>& BlockScale::strides() const
|
|
{
|
|
return m_strides;
|
|
}
|
|
|
|
bool BlockScale::operator==(BlockScale const& rhs) const
|
|
{
|
|
return m_tag == rhs.m_tag && m_data == rhs.m_data && m_scale == rhs.m_scale
|
|
&& m_strides == rhs.m_strides;
|
|
}
|
|
|
|
std::string toString(ScaleMode const& mode)
|
|
{
|
|
switch(mode)
|
|
{
|
|
case ScaleMode::None:
|
|
return "None";
|
|
case ScaleMode::SingleScale:
|
|
return "SingleScale";
|
|
case ScaleMode::Separate:
|
|
return "Separate";
|
|
case ScaleMode::Inline:
|
|
return "Inline";
|
|
case ScaleMode::Count:;
|
|
}
|
|
|
|
return "Invalid";
|
|
}
|
|
|
|
std::ostream& operator<<(std::ostream& stream, ScaleMode const& mode)
|
|
{
|
|
return stream << toString(mode);
|
|
}
|
|
|
|
SubTileTranspose::SubTileTranspose(OperationTag input,
|
|
std::vector<size_t> tileDimensions,
|
|
bool transpose)
|
|
: m_input(input)
|
|
, m_tileDimensions(std::move(tileDimensions))
|
|
, m_transpose(transpose)
|
|
{
|
|
}
|
|
|
|
std::unordered_set<OperationTag> SubTileTranspose::getInputs() const
|
|
{
|
|
return {m_input};
|
|
}
|
|
std::string SubTileTranspose::toString() const
|
|
{
|
|
return fmt::format(
|
|
"SubTileTranspose({}: input {})", concatenate(m_tileDimensions), m_input.value);
|
|
}
|
|
|
|
const std::vector<size_t>& SubTileTranspose::tileDimensions() const
|
|
{
|
|
return m_tileDimensions;
|
|
}
|
|
|
|
bool SubTileTranspose::isTranspose() const
|
|
{
|
|
return m_transpose;
|
|
}
|
|
|
|
bool SubTileTranspose::operator==(SubTileTranspose const& other) const
|
|
{
|
|
return (*this <=> other) == std::strong_ordering::equal;
|
|
}
|
|
|
|
OperationTag SubTileTranspose::input() const
|
|
{
|
|
return m_input;
|
|
}
|
|
}
|
|
}
|