[compiler-rt] [compiler-rt] Fix incorrect usages of check_cxx_compiler_flag (PR #83779)

[Alexander Richardson via llvm-commits]

================
@@ -150,21 +150,21 @@ check_cxx_compiler_flag(/wd4391 COMPILER_RT_HAS_WD4391_FLAG)
 check_cxx_compiler_flag(/wd4722 COMPILER_RT_HAS_WD4722_FLAG)
 check_cxx_compiler_flag(/wd4800 COMPILER_RT_HAS_WD4800_FLAG)
 
-check_cxx_compiler_flag(-Werror -Warray-bounds COMPILER_RT_HAS_ARRAY_BOUNDS_FLAG)
-check_cxx_compiler_flag(-Werror -Wuninitialized COMPILER_RT_HAS_UNINITIALIZED_FLAG)
-check_cxx_compiler_flag(-Werror -Wshadow COMPILER_RT_HAS_SHADOW_FLAG)
-check_cxx_compiler_flag(-Werror -Wempty-body COMPILER_RT_HAS_EMPTY_BODY_FLAG)
-check_cxx_compiler_flag(-Werror -Wsizeof-pointer-memaccess COMPILER_RT_HAS_SIZEOF_POINTER_MEMACCESS_FLAG)
-check_cxx_compiler_flag(-Werror -Wsizeof-array-argument COMPILER_RT_HAS_SIZEOF_ARRAY_ARGUMENT_FLAG)
-check_cxx_compiler_flag(-Werror -Wsuspicious-memaccess COMPILER_RT_HAS_SUSPICIOUS_MEMACCESS_FLAG)
-check_cxx_compiler_flag(-Werror -Wbuiltin-memcpy-chk-size COMPILER_RT_HAS_BUILTIN_MEMCPY_CHK_SIZE_FLAG)
-check_cxx_compiler_flag(-Werror -Warray-bounds-pointer-arithmetic COMPILER_RT_HAS_ARRAY_BOUNDS_POINTER_ARITHMETIC_FLAG)
-check_cxx_compiler_flag(-Werror -Wreturn-stack-address COMPILER_RT_HAS_RETURN_STACK_ADDRESS_FLAG)
-check_cxx_compiler_flag(-Werror -Wsizeof-array-decay COMPILER_RT_HAS_SIZEOF_ARRAY_DECAY_FLAG)
-check_cxx_compiler_flag(-Werror -Wformat-insufficient-args COMPILER_RT_HAS_FORMAT_INSUFFICIENT_ARGS_FLAG)
-check_cxx_compiler_flag(-Werror -Wformat-security COMPILER_RT_HAS_BUILTIN_FORMAL_SECURITY_FLAG)
-check_cxx_compiler_flag(-Werror -Wsizeof-array-div COMPILER_RT_HAS_SIZEOF_ARRAY_DIV_FLAG)
-check_cxx_compiler_flag(-Werror -Wsizeof-pointer-div COMPILER_RT_HAS_SIZEOF_POINTER_DIV_FLAG)
+check_cxx_compiler_flag("-Werror -Warray-bounds" COMPILER_RT_HAS_ARRAY_BOUNDS_FLAG)
----------------
arichardson wrote:

Thanks for looking into this. I tried x86 and aarch64 and didn't see any problems so assumed this was safe. Maybe related to big endian?

https://github.com/llvm/llvm-project/pull/83779