Add proper CMake build for extension_parallel

Previously it was copied in several places per executorch_srcs.cmake.

Needed for #8932

Test Plan: Compare cmake-out/executorch_srcs.cmake before/after for my usual testing cmake config with "all the CPU stuff" on; found that thread_parallel.cpp is now duplicated only in one place instead of multiple (it's in llama_runner, which needs a general fixup because it's duplicating several extensions).

ghstack-source-id: 333d1c9
ghstack-comment-id: 2698715235
Pull Request resolved: #8938

Author: swolchok

CMakeLists.txt

@@ -749,9 +749,9 @@ endif()
 
 if(EXECUTORCH_BUILD_PTHREADPOOL
    AND EXECUTORCH_BUILD_CPUINFO
-   AND CMAKE_CXX_STANDARD GREATER_EQUAL 14
 )
   add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/extension/threadpool)
+  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/extension/parallel)
 endif()
 
 if(EXECUTORCH_BUILD_PYBIND)

build/cmake_deps.toml

@@ -73,6 +73,7 @@ excludes = [
 deps = [
   "executorch",
   "executorch_core",
+  "extension_parallel",
   "extension_threadpool",
   "portable_kernels",
 ]
@@ -115,6 +116,7 @@ excludes = [
 deps = [
   "executorch_core",
   "executorch",
+  "extension_parallel",
 ]
 
 [targets.optimized_native_cpu_ops]
@@ -129,6 +131,8 @@ excludes = [
 deps = [
   "executorch_core",
   "executorch",
+  "extension_parallel",
+  "extension_threadpool",
   "portable_kernels",
 ]
 # ---------------------------------- core end ----------------------------------
@@ -208,6 +212,19 @@ deps = [
   "extension_runner_util",
 ]
 
+[targets.extension_parallel]
+buck_targets = [
+  "//extension/parallel:thread_parallel",
+]
+filters = [
+  ".cpp$",
+]
+deps = [
+  "executorch",
+  "executorch_core",
+  "extension_threadpool",
+]
+
 [targets.extension_tensor]
 buck_targets = [
   "//extension/tensor:tensor",
@@ -395,6 +412,7 @@ deps = [
   "executorch",
   "executorch_core",
   "optimized_kernels",
+  "extension_parallel",
   "extension_threadpool",
   "xnnpack_backend",
 ]

kernels/optimized/CMakeLists.txt

@@ -43,7 +43,7 @@ endif()
 list(TRANSFORM _optimized_cpublas__srcs PREPEND "${EXECUTORCH_ROOT}/")
 add_library(cpublas STATIC ${_optimized_cpublas__srcs})
 target_link_libraries(
-  cpublas PRIVATE executorch_core eigen_blas extension_threadpool
+  cpublas PRIVATE executorch_core eigen_blas extension_parallel extension_threadpool
 )
 target_compile_options(cpublas PUBLIC ${_common_compile_options})