Update on "Use c10 version of half/bfloat16 in executorch"

swolchok · swolchok · commit af153384e454 · 2024-12-18T10:44:35.000-08:00
Accomplished by importing relevant files from c10 into executorch/runtime/core/portable_type/c10, and then using `using` in the top-level ExecuTorch headers. This approach should keep the ExecuTorch build hermetic for embedded use cases. In the future, we should add a CI job to ensure the c10 files stay identical to the PyTorch ones. Differential Revision: [D66106969](https://our.internmc.facebook.com/intern/diff/D66106969/) [ghstack-poisoned]
diff --git a/backends/qualcomm/CMakeLists.txt b/backends/qualcomm/CMakeLists.txt
@@ -53,6 +53,7 @@ add_custom_command(
 )
 
 add_compile_options("-Wall" "-Werror" "-Wno-sign-compare")
+add_compile_definitions(C10_USING_CUSTOM_GENERATED_MACROS)
 
 # GNU emit wanring for ignored attributes Unfortunately, we use [[maybe_unused]]
 # which can be ignored by GNU. So we make it a warning, not an error in GNU.

Original file line number	Diff line number	Diff line change
`@@ -53,6 +53,7 @@ add_custom_command(`
`53`	`53`	`)`
`54`	`54`
`55`	`55`	`add_compile_options("-Wall" "-Werror" "-Wno-sign-compare")`
	`56`	`+add_compile_definitions(C10_USING_CUSTOM_GENERATED_MACROS)`
`56`	`57`
`57`	`58`	`# GNU emit wanring for ignored attributes Unfortunately, we use [[maybe_unused]]`
`58`	`59`	`# which can be ignored by GNU. So we make it a warning, not an error in GNU.`