maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity

[OpenMP] Change AMDGCN to AMDGPU in the Cmake Module 

Summary:
Change the name for targeting AMD offloading.
This commit is contained in:
Joseph Huber 2021-07-20 12:52:00 -04:00 committed by Huber, Joseph
parent 0cc2346cbf
commit b917a1d713
3 changed files with 10 additions and 10 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
openmp/docs/SupportAndFAQ.rst
View File

@ -240,7 +240,7 @@ offloading provided by LLVM. It will attempt to find OpenMP target offloading
support for your compiler. The flags necessary for OpenMP target offloading will
be loaded into the ``OpenMPTarget::OpenMPTarget_<device>`` target or the
``OpenMPTarget_<device>_FLAGS`` variable if successful. Currently supported
devices are ``AMDGCN`` and ``NVPTX``.
devices are ``AMDGPU`` and ``NVPTX``.
To use this module, simply add the path to CMake's current module path and call
``find_package``. The module will be installed with your OpenMP installation by

16
openmp/tools/Modules/FindOpenMPTarget.cmake
View File

@ -27,7 +27,7 @@ already. It does not need to be included separately to get full OpenMP support.
Variables
^^^^^^^^^
The module exposes the components ``NVPTX`` and ``AMDGCN``. Each of these
The module exposes the components ``NVPTX`` and ``AMDGPU``. Each of these
controls the various offloading targets to search OpenMP target offloasing
support for.
@ -38,7 +38,7 @@ Depending on the enabled components the following variables will be set:
targets have been found.
This module will set the following variables per language in your
project, where ``<device>`` is one of NVPTX or AMDGCN
project, where ``<device>`` is one of NVPTX or AMDGPU
``OpenMPTarget_<device>_FOUND``
Variable indicating if OpenMP support for the ``<device>`` was detected.
@ -62,7 +62,7 @@ be used to override the standard flag searching for a given compiler.
``OpenMPTarget_<device>_ARCH``
Sets the architecture of ``<device>`` to compile for. Such as `sm_70` for NVPTX
or `gfx908` for AMDGCN.
or `gfx908` for AMDGPU.
``OpenMPTarget_<device>_DEVICE``
Sets the name of the device to offload to.
@ -124,7 +124,7 @@ function(_OPENMP_TARGET_DEVICE_CANDIDATES LANG DEVICE)
if(DEFINED OMPTarget_DEVICE_${CMAKE_${LANG}_COMPILER_ID})
set(OpenMPTarget_DEVICE_CANDIDATES "${OMPTarget_DEVICE_${CMAKE_${LANG}_COMPILER_ID}}")
endif()
elseif("${DEVICE}" STREQUAL "AMDGCN")
elseif("${DEVICE}" STREQUAL "AMDGPU")
set(OMPTarget_DEVICE_Clang "amdgcn-amd-amdhsa")
set(OMPTarget_DEVICE_GNU "hsa")
@ -227,7 +227,7 @@ foreach(LANG IN ITEMS C CXX)
set(OpenMP_${LANG}_VERSION ${OpenMP_${LANG}_VERSION}
CACHE STRING "OpenMP Version" FORCE)
mark_as_advanced(OpenMP_${LANG}_VERSION)
foreach(DEVICE IN ITEMS NVPTX AMDGCN)
foreach(DEVICE IN ITEMS NVPTX AMDGPU)
if(CMAKE_${LANG}_COMPILER_LOADED)
if(NOT DEFINED OpenMPTarget_${LANG}_FLAGS OR NOT DEFINED OpenMPTarget_${LANG}_DEVICE)
_OPENMP_TARGET_DEVICE_GET_FLAGS(${LANG} ${DEVICE}
@ -242,9 +242,9 @@ foreach(LANG IN ITEMS C CXX)
CACHE STRING "${DEVICE} target compile flags for OpenMP target offloading" FORCE)
set(OpenMPTarget_${DEVICE}_ARCH ${_OpenMPTarget_${DEVICE}_ARCHS}
CACHE STRING "${DEVICE} target architecture flags for OpenMP target offloading" FORCE)
set(OpenMPTarget_${DEVICE}_LIBS ${OpenMPTarget_${DEVICE}_LIBS_WORK}
set(OpenMPTarget_${DEVICE}_LIBRARIES ${OpenMPTarget_${DEVICE}_LIBS_WORK}
CACHE STRING "${DEVICE} target libraries for OpenMP target offloading" FORCE)
mark_as_advanced(OpenMPTarget_${DEVICE}_FLAGS OpenMPTarget_${DEVICE}_ARCH OpenMPTarget_${DEVICE}_LIBS)
mark_as_advanced(OpenMPTarget_${DEVICE}_FLAGS OpenMPTarget_${DEVICE}_ARCH OpenMPTarget_${DEVICE}_LIBRARIES)
endif()
endif()
endforeach()
@ -301,7 +301,7 @@ foreach(LANG IN ITEMS C CXX)
INTERFACE_INCLUDE_DIRECTORIES "$<BUILD_INTERFACE:${OpenMP_${LANG}_INCLUDE_DIRS}>")
set_property(TARGET OpenMPTarget::OpenMPTarget_${DEVICE} PROPERTY
INTERFACE_LINK_LIBRARIES
"${OpenMPTarget_${DEVICE}_LIBS}"
"${OpenMPTarget_${DEVICE}_LIBRARIES}"
"${OpenMP_${LANG}_LIBRARIES}")
# The offloading flags must also be passed during the linking phase so
# the compiler can pass the binary to the correct toolchain.

2
openmp/tools/Modules/README.rst
View File

@ -16,7 +16,7 @@ device. The module will attempt to compile a test program using known compiler
flags for each requested architecture. If successful, the flags required for
offloading will be loaded into the ``OpenMPTarget::OpenMPTarget_<device>``
target or the ``OpenMPTarget_<device>_FLAGS`` variable. Currently supported target
devices are ``NVPTX`` and ``AMDGCN``. This module is still under development so
devices are ``NVPTX`` and ``AMDGPU``. This module is still under development so
some features may be missing.
To use this module, simply add the path to CMake's current module path and call