Merge pull request #4 from PolyCam/jgibson/neon-custom-kernels

optimized: add NEON grid_sampler_2d.out and Vectorized<float> sum.IntList_out

Author: jgibson2

kernels/optimized/CMakeLists.txt

@@ -75,6 +75,21 @@ target_link_libraries(
                            kernels_util_all_deps
 )
 target_compile_options(optimized_kernels PUBLIC ${_common_compile_options})
+
+# op_grid_sampler_2d.cpp uses ARMv8.2-a+fp16 NEON intrinsics
+# (vcvt_f32_f16 / vld1_f16) when compiled for aarch64. Scope the extra
+# `-march` flag to just that source so non-arm64 targets (e.g. x86_64 on
+# Android) are unaffected — the kernel itself has `#ifdef __aarch64__`
+# guards and falls through to the portable kernel otherwise.
+if(CMAKE_SYSTEM_PROCESSOR MATCHES "aarch64|arm64"
+   OR ANDROID_ABI STREQUAL "arm64-v8a"
+)
+  set_source_files_properties(
+    ${EXECUTORCH_ROOT}/kernels/optimized/cpu/op_grid_sampler_2d.cpp
+    PROPERTIES COMPILE_OPTIONS "-march=armv8.2-a+fp16"
+  )
+endif()
+
 # Build a library for _optimized_kernels_srcs
 #
 # optimized_ops_lib: Register optimized ops kernels into Executorch runtime
@@ -83,6 +98,30 @@ gen_operators_lib(
   executorch_core
 )
 
+# On-device verifier for optimized grid_sampler_2d / sum.IntList_out.
+# Opt-in via -DEXECUTORCH_BUILD_OPTIMIZED_VERIFY=ON so it doesn't affect
+# default AAR / library builds. Cross-checks both ops against an fp32
+# reference derived from the portable kernel; non-zero exit on divergence.
+if(EXECUTORCH_BUILD_OPTIMIZED_VERIFY)
+  add_executable(
+    verify_optimized_kernels ${EXECUTORCH_ROOT}/kernels/optimized/verify.cpp
+  )
+  target_link_libraries(
+    verify_optimized_kernels
+    PRIVATE optimized_kernels portable_kernels executorch_core
+  )
+  target_compile_options(
+    verify_optimized_kernels PRIVATE ${_common_compile_options}
+  )
+  if(CMAKE_SYSTEM_PROCESSOR MATCHES "aarch64|arm64"
+     OR ANDROID_ABI STREQUAL "arm64-v8a"
+  )
+    target_compile_options(
+      verify_optimized_kernels PRIVATE -march=armv8.2-a+fp16
+    )
+  endif()
+endif()
+
 install(
   # eigen_blas doesn't export itself, so we have to do our own install to export
   # it.