danalbert added a comment.
In https://reviews.llvm.org/D24690#545523, @compnerd wrote:
> So, the only thing that Im confused about is where does `__BIONIC__` get
It's in Bionic's `<sys/cdefs.h>`, which gets pulled in via `<features.h>`.
Comment at: include/__config:340
@@ -339,3 +344,1 @@
-# include <features.h>
-#if __GLIBC_PREREQ(2, 15)
> What happened to this include? I believe it's needed to get `__GLIBC_PREREQ`.
Included much earlier now (L90). We always needed this for Linux, and having it
at the top of the file means we won't accidentally forget to include it for an
Comment at: include/__config:766
@@ -761,3 +765,3 @@
// Most unix variants have catopen. These are the specific ones that don't.
-#if !defined(_WIN32) && !defined(__ANDROID__) && !defined(_NEWLIB_VERSION)
+#if !defined(_WIN32) && !defined(__BIONIC__) && !defined(_NEWLIB_VERSION)
#define _LIBCPP_HAS_CATOPEN 1
> Not your fault, but `_WIN32` and `__unix__`? Im not sure if MinGW or cygwin
> define both.
That is odd. I would assume that `_WIN32` implies `!__unix__`.
cfe-commits mailing list