NERSC · jrmadsen · Jul 11, 2021 · Jun 30, 2021 · Jul 6, 2021 · Jul 7, 2021
diff --git a/.circleci/config.yml b/.circleci/config.yml
@@ -241,7 +241,7 @@ jobs:
             python ./pyctest-runner.py -SF --pyctest-site=CircleCI --pyctest-model=Continuous --pyctest-build-type=Release -j 1
             --build-libs shared --python --tools avail timem --cxx-standard=17 --compile-time-perf ${HOME}/ctp
             -- -V --output-on-failure
-            -- -DTIMEMORY_BUILD_{CALIPER,COMPILER_INSTRUMENTATION}=OFF -DPYTHON_EXECUTABLE=$(which python) -DTIMEMORY_BUILD_PYTHON_{HATCHET,TIMEMORY_BUILD_PYTHON_LINE_PROFILER}=OFF -DCMAKE_INSTALL_PREFIX=${HOME}/timemory-install
+            -- -DTIMEMORY_BUILD_{CALIPER,COMPILER_INSTRUMENTATION}=OFF -DPYTHON_EXECUTABLE=$(which python) -DTIMEMORY_BUILD_PYTHON_{HATCHET,LINE_PROFILER}=OFF -DCMAKE_INSTALL_PREFIX=${HOME}/timemory-install
       - run:
           name: install
           command: >

diff --git a/.requirements/build.txt b/.requirements/build.txt
@@ -1,4 +1,3 @@
 Cython
 scikit-build
 cmake
-ninja
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -1,5 +1,30 @@
 # timemory
 
+## Version 3.2.1
+
+> Date: Fri Jul 9 16:55:33 2021 -0500
+
+- pytimem fix
+  - fix missing import of component_bundle and component_tuple
+- added additional python tests
+- Ability to build with static libraries: python bindings, mpip library, mallocp library, ompt library, ncclp library, KokkosP libraries
+- Setting TIMEMORY_BUILD_PYTHON to OFF now results in searching for external pybind11 install
+- Renamed some CMake files in cmake/Modules
+- Updated caliper and gotcha submodules to support {CALIPER,GOTCHA}_INSTALL_{CONFIG,HEADER} options
+- Added TIMEMORY_INSTALL_PYTHON option
+- Fixed BUILD_STATIC_LIBS=ON + CMAKE_POSITION_INDEPENDENT_CODE=ON
+- Fixed TIMEMORY_USE_CUDA=ON + TIMEMORY_REQUIRE_PACKAGES=ON to fail
+- If TIMEMORY_REQUIRED_PACKAGES=OFF, search for packages first before adding submodule
+- Extended setup.py to support more options and support non-development install (no headers or cmake config)
+- Removed TIMEMORY_EMBED_PYTHON option
+- Disable timemory-jump when no shared libraries are built since dlopen isn't possible
+- Replaced allocator member functions construct, destroy, allocate, deallocate with calls to static functions of allocator traits
+- added support for CMAKE_ARGS env variable in setup.py
+- remove absolute rpath when SKBUILD/SPACK_BUILD (since these have staging directories)
+- timemory-{c,cxx,fortran} alias libraries in build tree
+- toggled python function profiler to not include line number by default
+  - This can cause strange results when generators are used
+
 ## Version 3.2.0
 
 > Date: Sun Jun 27 21:10:57 2021 -0500

diff --git a/CMakeLists.txt b/CMakeLists.txt
@@ -14,11 +14,11 @@ if(CMAKE_SOURCE_DIR STREQUAL CMAKE_BINARY_DIR AND
     message(AUTHOR_WARNING "In-source build")
 endif()
 
-if(APPLE AND NOT "$ENV{CONDA_PYTHON_EXE}" STREQUAL "")
-    # disable by default bc conda will set these and cause problem with python bindings
-    set(CMAKE_C_FLAGS "" CACHE STRING "")
-    set(CMAKE_CXX_FLAGS "" CACHE STRING "")
-endif()
+# if(APPLE AND NOT "$ENV{CONDA_PYTHON_EXE}" STREQUAL "")
+#    # disable by default bc conda will set these and cause problem with python bindings
+#    set(CMAKE_C_FLAGS "" CACHE STRING "")
+#    set(CMAKE_CXX_FLAGS "" CACHE STRING "")
+# endif()
 
 #----------------------------------------------------------------------------------------#
 #   project
@@ -70,33 +70,17 @@ else()
     # set(CMAKE_POSITION_INDEPENDENT_CODE ON)
 endif()
 
-set(TIMEMORY_GIT_DESCRIBE "unknown")
-set(TIMEMORY_GIT_REVISION "unknown")
-
-find_package(Git QUIET)
-if(Git_FOUND)
-    execute_process(
-        COMMAND             ${GIT_EXECUTABLE} describe --tags
-        WORKING_DIRECTORY   ${PROJECT_SOURCE_DIR}
-        OUTPUT_VARIABLE     TIMEMORY_GIT_DESCRIBE
-        ERROR_QUIET
-        OUTPUT_STRIP_TRAILING_WHITESPACE)
-    execute_process(
-        COMMAND             ${GIT_EXECUTABLE} rev-parse HEAD
-        WORKING_DIRECTORY   ${PROJECT_SOURCE_DIR}
-        OUTPUT_VARIABLE     TIMEMORY_GIT_REVISION
-        ERROR_QUIET
-        OUTPUT_STRIP_TRAILING_WHITESPACE)
-endif()
-
 # install directories
 include(GNUInstallDirs)
 # cmake installation folder -- change CMAKE_INSTALL_DATAROOTDIR to tweak this
 set(CMAKE_INSTALL_CONFIGDIR  ${CMAKE_INSTALL_DATAROOTDIR}/cmake/${PROJECT_NAME})
 
 if(UNIX AND NOT APPLE)
     set(CMAKE_INSTALL_RPATH
-        "\$ORIGIN/../${CMAKE_INSTALL_LIBDIR}:\$ORIGIN/timemory:\$ORIGIN:${CMAKE_INSTALL_PREFIX}/${CMAKE_INSTALL_LIBDIR}:${CMAKE_INSTALL_PREFIX}/${CMAKE_INSTALL_LIBDIR}/timemory")
+        "\$ORIGIN/../${CMAKE_INSTALL_LIBDIR}:\$ORIGIN/timemory:\$ORIGIN")
+    if(NOT SKBUILD AND NOT SPACK_BUILD)
+        set(CMAKE_INSTALL_RPATH "${CMAKE_INSTALL_RPATH}:${CMAKE_INSTALL_PREFIX}/${CMAKE_INSTALL_LIBDIR}:${CMAKE_INSTALL_PREFIX}/${CMAKE_INSTALL_LIBDIR}/timemory")
+    endif()
 endif()
 
 # create the full path version and generic path versions
@@ -183,14 +167,12 @@ option(TIMEMORY_QUIET_CONFIG "Make timemory configuration quieter" OFF)
 mark_as_advanced(TIMEMORY_QUIET_CONFIG)
 
 if(SKBUILD)
-    set(_BUILD_SHARED_CXX ON)
+    if(BUILD_STATIC_LIBS AND TIMEMORY_USE_PYTHON)
+        set(CMAKE_POSITION_INDEPENDENT_CODE ON)
+    endif()
 else()
-    if((TIMEMORY_BUILD_PYTHON OR TIMEMORY_USE_PYTHON) AND NOT BUILD_SHARED_LIBS)
-        if(NOT TIMEMORY_BUILD_QUIET)
-            message(AUTHOR_WARNING "BUILD_SHARED_LIBS=OFF --> disabling TIMEMORY_BUILD_PYTHON...")
-        endif()
-        set(TIMEMORY_BUILD_PYTHON OFF)
-        set(TIMEMORY_USE_PYTHON OFF)
+    if(BUILD_STATIC_LIBS AND NOT BUILD_SHARED_LIBS AND TIMEMORY_USE_PYTHON AND NOT CMAKE_POSITION_INDEPENDENT_CODE)
+        message(FATAL_ERROR "Error! Python compilation with static libraries requires CMAKE_POSITION_INDEPENDENT_CODE to be ON")
     endif()
 endif()
 
@@ -202,7 +184,7 @@ endif()
 set(_TLS_DESCRIPT "Thread-local static model: 'global-dynamic', 'local-dynamic', 'initial-exec', 'local-exec'")
 set(_TLS_OPTIONS "global-dynamic" "local-dynamic" "initial-exec" "local-exec")
 
-if(SKBUILD OR TIMEMORY_BUILD_PYTHON OR TIMEMORY_USE_PYTHON OR TIMEMORY_USE_DYNINST)
+if(SKBUILD OR TIMEMORY_USE_PYTHON OR TIMEMORY_USE_DYNINST)
     set(TIMEMORY_TLS_MODEL "global-dynamic" CACHE STRING "${_TLS_DESCRIPT}")
     # ensure local override
     set(TIMEMORY_TLS_MODEL "global-dynamic")

diff --git a/MANIFEST.in b/MANIFEST.in
@@ -46,3 +46,7 @@ recursive-exclude _skbuild *
 recursive-exclude build *
 recursive-exclude dist *
 recursive-exclude external/hatchet/build *
+
+# bypass global exclude of git files
+
+include docs/.gitinfo
diff --git a/VERSION b/VERSION
@@ -1 +1 @@
-3.2.0
+3.2.1
diff --git a/cmake/Modules/BuildSettings.cmake b/cmake/Modules/BuildSettings.cmake
@@ -10,19 +10,54 @@ include_guard(DIRECTORY)
 
 include(GNUInstallDirs)
 include(Compilers)
+include(FindPackageHandleStandardArgs)
 
 target_compile_definitions(timemory-compile-options INTERFACE $<$<CONFIG:DEBUG>:DEBUG>)
 
-if(CMAKE_DL_LIBS)
-    set(dl_LIBRARY ${CMAKE_DL_LIBS})
-    target_link_libraries(timemory-compile-options INTERFACE ${CMAKE_DL_LIBS})
-else()
-    find_library(dl_LIBRARY NAMES dl)
+#----------------------------------------------------------------------------------------#
+#   dynamic linking and runtime libraries
+#
+if(CMAKE_DL_LIBS AND NOT "${CMAKE_DL_LIBS}" STREQUAL "dl")
+    # if cmake provides dl library, use that
+    set(dl_LIBRARY ${CMAKE_DL_LIBS} CACHE FILEPATH "dynamic linking system library")
+endif()
+
+foreach(_TYPE dl rt dw)
+    if(NOT ${_TYPE}_LIBRARY)
+        find_library(${_TYPE}_LIBRARY NAMES ${_TYPE})
+    endif()
+endforeach()
+
+find_package_handle_standard_args(dl-library REQUIRED_VARS dl_LIBRARY)
+find_package_handle_standard_args(rt-library REQUIRED_VARS rt_LIBRARY)
+# find_package_handle_standard_args(dw-library REQUIRED_VARS dw_LIBRARY)
+
+if(TIMEMORY_BUILD_PORTABLE)
     if(dl_LIBRARY)
-        target_link_libraries(timemory-compile-options INTERFACE ${dl_LIBRARY})
+        set(dl_LIBRARY dl)
+    else()
+        set(dl_LIBRARY)
+    endif()
+elseif(dl_LIBRARY)
+    add_rpath(${dl_LIBRARY})
+endif()
+
+if(dl_LIBRARY)
+    target_link_libraries(timemory-compile-options INTERFACE ${dl_LIBRARY})
+endif()
+
+if(TIMEMORY_BUILD_PORTABLE)
+    if(rt_LIBRARY)
+        set(rt_LIBRARY rt)
+    else()
+        set(rt_LIBRARY)
     endif()
+elseif(rt_LIBRARY)
+    add_rpath(${rt_LIBRARY})
 endif()
 
+#----------------------------------------------------------------------------------------#
+
 if(WIN32)
     set(OS_FLAG "/bigobj")
 else()
@@ -72,15 +107,6 @@ endif()
 add_interface_library(timemory-compile-debuginfo
     "Attempts to set best flags for more expressive profiling information in debug or optimized binaries")
 
-# if cmake provides dl library, use that
-if(CMAKE_DL_LIBS)
-    set(dl_LIBRARY "${CMAKE_DL_LIBS}" CACHE STRING "dynamic linking libraries")
-endif()
-
-find_library(rt_LIBRARY NAMES rt)
-find_library(dl_LIBRARY NAMES dl)
-find_library(dw_LIBRARY NAMES dw)
-
 add_target_flag_if_avail(timemory-compile-debuginfo
     "-g"
     "-fno-omit-frame-pointer"
@@ -333,7 +359,7 @@ set(VECTOR_DEFINITION               TIMEMORY_VEC)
 set(VECTOR_INTERFACE_TARGET         timemory-vector)
 set(ARCH_INTERFACE_TARGET           timemory-arch)
 
-include(ArchConfig)
+include(ConfigCpuArch)
 
 add_cmake_defines(TIMEMORY_VEC VALUE)
 

diff --git a/cmake/Modules/ClangFormat.cmake b/cmake/Modules/ClangFormat.cmake
@@ -113,14 +113,18 @@ if(CLANG_FORMATTER)
             COMMAND ${CLANG_FORMATTER} -i ${examples})
     endif()
 
+    set(_MSG "'${CLANG_FORMATTER}'")
     if(BLACK_FORMATTER)
         set(_COMMAND ${_COMMAND}
             COMMAND ${BLACK_FORMATTER} -q ${PROJECT_SOURCE_DIR})
+        set(_MSG "${_MSG} and '${BLACK_FORMATTER}'")
     endif()
 
     add_custom_target(${FORMAT_NAME}
         ${_COMMAND}
         WORKING_DIRECTORY ${PROJECT_SOURCE_DIR}
-        COMMENT "[${PROJECT_NAME}] Running '${CLANG_FORMATTER}'..."
+        COMMENT "[${PROJECT_NAME}] Running ${_MSG}..."
         SOURCES ${headers} ${sources} ${examples})
+
+    unset(_MSG)
 endif()
diff --git a/cmake/Modules/Compilers.cmake b/cmake/Modules/Compilers.cmake
@@ -174,21 +174,25 @@ macro(ADD_C_FLAG_IF_AVAIL FLAG)
         string(REPLACE "-" "_" FLAG_NAME "${FLAG_NAME}")
         string(REPLACE " " "_" FLAG_NAME "${FLAG_NAME}")
         string(REPLACE "=" "_" FLAG_NAME "${FLAG_NAME}")
-        timemory_begin_flag_check()
-        check_c_compiler_flag("-Werror" c_werror)
-        if(c_werror)
-            check_c_compiler_flag("${FLAG} -Werror" ${FLAG_NAME})
+        if(NOT TIMEMORY_BUILD_C)
+            set(${FLAG_NAME} ON)
         else()
-            check_c_compiler_flag("${FLAG}" ${FLAG_NAME})
-        endif()
-        timemory_end_flag_check()
-        if(${FLAG_NAME})
-            if("${_LTARG}" STREQUAL "")
-                list(APPEND ${PROJECT_NAME}_C_FLAGS "${FLAG}")
-                list(APPEND ${PROJECT_NAME}_C_COMPILE_OPTIONS "${FLAG}")
-                add_target_c_flag(${LIBNAME}-compile-options ${FLAG})
+            timemory_begin_flag_check()
+            check_c_compiler_flag("-Werror" c_werror)
+            if(c_werror)
+                check_c_compiler_flag("${FLAG} -Werror" ${FLAG_NAME})
             else()
-                add_target_c_flag(${_TARG} ${FLAG})
+                check_c_compiler_flag("${FLAG}" ${FLAG_NAME})
+            endif()
+            timemory_end_flag_check()
+            if(${FLAG_NAME})
+                if("${_LTARG}" STREQUAL "")
+                    list(APPEND ${PROJECT_NAME}_C_FLAGS "${FLAG}")
+                    list(APPEND ${PROJECT_NAME}_C_COMPILE_OPTIONS "${FLAG}")
+                    add_target_c_flag(${LIBNAME}-compile-options ${FLAG})
+                else()
+                    add_target_c_flag(${_TARG} ${FLAG})
+                endif()
             endif()
         endif()
     endif()

diff --git a/cmake/Modules/CUDAConfig.cmake → cmake/Modules/ConfigCUDA.cmake b/cmake/Modules/CUDAConfig.cmake → cmake/Modules/ConfigCUDA.cmake
@@ -157,28 +157,38 @@ if("CUDA" IN_LIST LANGUAGES)
         ${CUDA_INCLUDE_DIRS}
         ${CMAKE_CUDA_TOOLKIT_INCLUDE_DIRECTORIES})
 
-    find_library(CUDA_dl_LIBRARY
-        NAMES dl)
+    if(NOT CUDA_dl_LIBRARY)
+        if(dl_LIBRARY)
+            set(CUDA_dl_LIBRARY ${dl_LIBRARY})
+        else()
+            find_library(CUDA_dl_LIBRARY NAMES dl)
+        endif()
+    endif()
 
-    target_link_libraries(${PROJECT_CUDA_INTERFACE_PREFIX}-cudart INTERFACE
-        ${CUDA_CUDART_LIBRARY} ${CUDA_rt_LIBRARY})
+    if(NOT CUDA_rt_LIBRARY)
+        if(rt_LIBRARY)
+            set(CUDA_rt_LIBRARY ${rt_LIBRARY})
+        else()
+            find_library(CUDA_rt_LIBRARY NAMES rt)
+        endif()
+    endif()
 
-    target_link_libraries(${PROJECT_CUDA_INTERFACE_PREFIX}-cudart-device INTERFACE
-        ${CUDA_cudadevrt_LIBRARY} ${CUDA_rt_LIBRARY})
+    if(NOT CUDA_dl_LIBRARY)
+        set(CUDA_dl_LIBRARY)
+    endif()
 
-    target_link_libraries(${PROJECT_CUDA_INTERFACE_PREFIX}-cudart-static INTERFACE
-        ${CUDA_cudart_static_LIBRARY} ${CUDA_rt_LIBRARY})
+    if(NOT CUDA_rt_LIBRARY)
+        set(CUDA_rt_LIBRARY)
+    endif()
 
-    if(CUDA_dl_LIBRARY)
-        target_link_libraries(${PROJECT_CUDA_INTERFACE_PREFIX}-cudart INTERFACE
-            ${CUDA_dl_LIBRARY})
+    target_link_libraries(${PROJECT_CUDA_INTERFACE_PREFIX}-cudart INTERFACE
+        ${CUDA_CUDART_LIBRARY} ${CUDA_rt_LIBRARY} ${CUDA_dl_LIBRARY})
 
-        target_link_libraries(${PROJECT_CUDA_INTERFACE_PREFIX}-cudart-device INTERFACE
-            ${CUDA_dl_LIBRARY})
+    target_link_libraries(${PROJECT_CUDA_INTERFACE_PREFIX}-cudart-device INTERFACE
+        ${CUDA_cudadevrt_LIBRARY} ${CUDA_rt_LIBRARY} ${CUDA_dl_LIBRARY})
 
-        target_link_libraries(${PROJECT_CUDA_INTERFACE_PREFIX}-cudart-static INTERFACE
-            ${CUDA_dl_LIBRARY})
-    endif()
+    target_link_libraries(${PROJECT_CUDA_INTERFACE_PREFIX}-cudart-static INTERFACE
+        ${CUDA_cudart_static_LIBRARY} ${CUDA_rt_LIBRARY} ${CUDA_dl_LIBRARY})
 
 else()
     message(FATAL_ERROR

diff --git a/cmake/Modules/CaliperDepends.cmake → cmake/Modules/ConfigCaliper.cmake b/cmake/Modules/CaliperDepends.cmake → cmake/Modules/ConfigCaliper.cmake
@@ -1,7 +1,7 @@
 # Try to find the libraries and headers for Caliper optional dependencies
 # Usage of this module is as follows
 #
-#     include(CaliperDepends)
+#     include(ConfigCaliper)
 #
 
 set(CALIPER_OPTION_PREFIX ON CACHE INTERNAL "Prefix caliper options with CALIPER_")
@@ -11,9 +11,9 @@ set(CALIPER_WITH_PAPI       OFF     CACHE BOOL "Enable PAPI in Caliper")
 set(CALIPER_WITH_MPI        OFF     CACHE BOOL "Enable MPI in Caliper")
 set(CALIPER_WITH_CUPTI      OFF     CACHE BOOL "Enable CUPTI in Caliper")
 
-if(TIMEMORY_USE_CUPTI)
-    set(CALIPER_WITH_CUPTI OFF CACHE BOOL "Enable cupti in Caliper")
-endif()
+# always sync with timemory settings
+set(CALIPER_INSTALL_CONFIG  ${TIMEMORY_INSTALL_CONFIG} CACHE BOOL "Install cmake and pkg-config files" FORCE)
+set(CALIPER_INSTALL_HEADERS ${TIMEMORY_INSTALL_HEADERS} CACHE BOOL "Install caliper headers" FORCE)
 
 find_path(LIBUNWIND_INCLUDE_DIR
     NAMES           unwind.h libunwind.h

diff --git a/cmake/Modules/ArchConfig.cmake → cmake/Modules/ConfigCpuArch.cmake b/cmake/Modules/ArchConfig.cmake → cmake/Modules/ConfigCpuArch.cmake
@@ -2,7 +2,17 @@
 # Configures architecture options
 #
 
-find_package(CpuArch)
+set(_CpuArch_COMPONENTS)
+if(TIMEMORY_BUILD_PORTABLE)
+    set(CpuArch_FIND_DEFAULT ON)
+    if(MSVC)
+        set(_CpuArch_COMPONENTS OPTIONAL_COMPONENTS sse sse2 avx avx2)
+    else()
+        set(_CpuArch_COMPONENTS OPTIONAL_COMPONENTS sse sse2 sse3 ssse3 sse4 sse4_1 sse4_2 fma avx avx2 altivec)
+    endif()
+endif()
+
+find_package(CpuArch ${_CpuArch_COMPONENTS})
 
 if(CpuArch_FOUND)