Skip to content

Complete gfx1151 Tensile support #696

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

Merged
amd-mtrifuno merged 2 commits into from
Jul 16, 2025
Merged

Complete gfx1151 Tensile support #696

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
6811c90
Complete gfx1151 Tensile support
amd-mtrifuno Jul 16, 2025
50ded4a
Update file names for gfx1151
amd-mtrifuno Jul 16, 2025
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: 1 addition & 1 deletion shared/tensile/Tensile/Common.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -325,7 +325,7 @@ class DeveloperWarning(Warning):
'gfx1010':'navi10', 'gfx1011':'navi12', 'gfx1012':'navi14',
'gfx1030':'navi21', 'gfx1031':'navi22', 'gfx1032':'navi23', 'gfx1034':'navi24', 'gfx1035':'rembrandt',
'gfx1100':'navi31', 'gfx1101':'navi32', 'gfx1102':'navi33', 'gfx1103':'gfx1103',
'gfx1150':'strixpoint', 'gfx1151':'gfx1151',
'gfx1150':'strixpoint', 'gfx1151':'strixhalo',
'gfx1200':'gfx1200',
'gfx1201':'gfx1201'
}
Expand Down
4 changes: 2 additions & 2 deletions shared/tensile/Tensile/Source/CMakeLists.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -51,9 +51,9 @@ if(CMAKE_CXX_COMPILER MATCHES ".*/hipcc$" OR CMAKE_CXX_COMPILER MATCHES ".*clang
endif()

if(CMAKE_CXX_COMPILER STREQUAL "hipcc")
set(TENSILE_GPU_ARCHS gfx803 gfx900 gfx906:xnack- gfx908:xnack- gfx90a:xnack- gfx1010 gfx1011 gfx1012 gfx1030 gfx1031 gfx1032 gfx1034 gfx1035 gfx1100 gfx1101 gfx1102 gfx1103 gfx1150 CACHE STRING "GPU architectures")
set(TENSILE_GPU_ARCHS gfx803 gfx900 gfx906:xnack- gfx908:xnack- gfx90a:xnack- gfx1010 gfx1011 gfx1012 gfx1030 gfx1031 gfx1032 gfx1034 gfx1035 gfx1100 gfx1101 gfx1102 gfx1103 gfx1150 gfx1151 CACHE STRING "GPU architectures")
else()
set(TENSILE_GPU_ARCHS gfx803 gfx900 gfx906 gfx908 gfx90a gfx1010 gfx1011 gfx1012 gfx1030 gfx1031 gfx1032 gfx1034 gfx1035 gfx1100 gfx1101 gfx1102 gfx1103 gfx1150 CACHE STRING "GPU architectures")
set(TENSILE_GPU_ARCHS gfx803 gfx900 gfx906 gfx908 gfx90a gfx1010 gfx1011 gfx1012 gfx1030 gfx1031 gfx1032 gfx1034 gfx1035 gfx1100 gfx1101 gfx1102 gfx1103 gfx1150 gfx1151 CACHE STRING "GPU architectures")
endif()

include(CMakeDependentOption)
Expand Down
9 changes: 8 additions & 1 deletion shared/tensile/Tensile/Source/lib/include/Tensile/AMDGPU.hpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -74,7 +74,8 @@ namespace Tensile
gfx1101 = 1101,
gfx1102 = 1102,
gfx1103 = 1103,
gfx1150 = 1150
gfx1150 = 1150,
gfx1151 = 1151
};

static std::string toString(Processor p)
Expand Down Expand Up @@ -121,6 +122,8 @@ namespace Tensile
return "gfx1103";
case AMDGPU::Processor::gfx1150:
return "gfx1150";
case AMDGPU::Processor::gfx1151:
return "gfx1151";
}
return "";
}
Expand Down Expand Up @@ -191,6 +194,10 @@ namespace Tensile
{
return AMDGPU::Processor::gfx1150;
}
else if(deviceString.find("gfx1151") != std::string::npos)
{
return AMDGPU::Processor::gfx1151;
}
else
{
return static_cast<AMDGPU::Processor>(0);
Expand Down
5 changes: 5 additions & 0 deletions shared/tensile/Tensile/Source/lib/include/Tensile/PlaceholderLibrary.hpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -59,6 +59,7 @@ namespace Tensile
gfx1102,
gfx1103,
gfx1150,
gfx1151,
All
};

Expand Down Expand Up @@ -107,6 +108,10 @@ namespace Tensile
return "TensileLibrary_*_gfx1102";
case LazyLoadingInit::gfx1103:
return "TensileLibrary_*_gfx1103";
case LazyLoadingInit::gfx1150:
return "TensileLibrary_*_gfx1150";
case LazyLoadingInit::gfx1151:
return "TensileLibrary_*_gfx1151";
case LazyLoadingInit::None:
return "";
}
Expand Down
1 change: 1 addition & 0 deletions shared/tensile/Tensile/Source/lib/include/Tensile/Serialization/Predicates.hpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -235,6 +235,7 @@ namespace Tensile
iot::enumCase(io, value, "gfx1102", AMDGPU::Processor::gfx1102);
iot::enumCase(io, value, "gfx1103", AMDGPU::Processor::gfx1103);
iot::enumCase(io, value, "gfx1150", AMDGPU::Processor::gfx1150);
iot::enumCase(io, value, "gfx1151", AMDGPU::Processor::gfx1151);
}
};

Expand Down
1 change: 1 addition & 0 deletions shared/tensile/pytest.ini
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -109,6 +109,7 @@ markers =
xfail-gfx1102: architecture
xfail-gfx1103: architecture
xfail-gfx1150: architecture
xfail-gfx1151: architecture
skip-gfx000: architecture
skip-gfx900: architecture
skip-gfx906: architecture
Expand Down