[llvm] [tsan] Don't use `enum __tsan_memory_order` in tsan interface (PR #114724)

[Vitaly Buka via llvm-commits]

vitalybuka wrote:

This is breaking the greendragon bots, could you revert or submit a fix if it's a quick one?

> /Users/ec2-user/jenkins/workspace/[llvm.org/as-lldb-cmake/llvm-project/compiler-rt/lib/tsan/rtl/tsan_interceptors_mac.cpp:103:1](http://llvm.org/as-lldb-cmake/llvm-project/compiler-rt/lib/tsan/rtl/tsan_interceptors_mac.cpp:103:1): error: no matching function for call to '__tsan_atomic32_fetch_add' 103 | OSATOMIC_INTERCEPTORS_ARITHMETIC(OSAtomicAdd, fetch_add, | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 104 | OSATOMIC_INTERCEPTOR_PLUS_X) | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~ /Users/ec2-user/jenkins/workspace/[llvm.org/as-lldb-cmake/llvm-project/compiler-rt/lib/tsan/rtl/tsan_interceptors_mac.cpp:81:37](http://llvm.org/as-lldb-cmake/llvm-project/compiler-rt/lib/tsan/rtl/tsan_interceptors_mac.cpp:81:37): note: expanded from macro 'OSATOMIC_INTERCEPTORS_ARITHMETIC' 81 | m(int32_t, int32_t, a32, f##32, __tsan_atomic32_##tsan_atomic_f, | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ :6:1: note: expanded from here 6 | __tsan_atomic32_fetch_add | ^~~~~~~~~~~~~~~~~~~~~~~~~ /Users/ec2-user/jenkins/workspace/[llvm.org/as-lldb-cmake/llvm-project/compiler-rt/lib/tsan/rtl/tsan_interceptors_mac.cpp:63:14](http://llvm.org/as-lldb-cmake/llvm-project/compiler-rt/lib/tsan/rtl/tsan_interceptors_mac.cpp:63:14): note: expanded from macro 'OSATOMIC_INTERCEPTOR_PLUS_X' 63 | return tsan_atomic_f((volatile tsan_t *)ptr, x, mo) + x; | ^~~~~~~~~~~~~ /Users/ec2-user/jenkins/workspace/[llvm.org/as-lldb-cmake/llvm-project/compiler-rt/lib/tsan/rtl/tsan_interface.h:278:5](http://llvm.org/as-lldb-cmake/llvm-project/compiler-rt/lib/tsan/rtl/tsan_interface.h:278:5): note: candidate function not viable: no known conversion from 'const morder' to 'int' for 3rd argument 278 | a32 __tsan_atomic32_fetch_add(volatile a32 *a, a32 v, int mo); | ^ ~~~~~~ https://green.lab.llvm.org/job/llvm.org/view/LLDB/job/as-lldb-cmake/14686/console



https://github.com/llvm/llvm-project/pull/114724