[glibc/maskray/clang] x86: Define __HAVE_FLOAT128 for Clang and use __builtin_*f128 code path

[Fangrui Song]

https://sourceware.org/git/gitweb.cgi?p=glibc.git;h=31554aa3e35fc73d83998641f580c451bddd36b0

commit 31554aa3e35fc73d83998641f580c451bddd36b0
Author: Fangrui Song <maskray@google.com>
Date:   Sun Oct 10 14:38:00 2021 -0700

    x86: Define __HAVE_FLOAT128 for Clang and use __builtin_*f128 code path
    
    Clang supports __builtin_fabsf128 (despite not supporting _Float128) but
    _not _builtin_fabsq.
    By falling back to `typedef __float128 _Float128;`, the float128 code
    will be buildable with Clang.

Diff:
---
 sysdeps/x86/bits/floatn.h | 15 ++++++++-------
 1 file changed, 8 insertions(+), 7 deletions(-)

diff --git a/sysdeps/x86/bits/floatn.h b/sysdeps/x86/bits/floatn.h
index f0c51716a9..890f2fafc9 100644
--- a/sysdeps/x86/bits/floatn.h
+++ b/sysdeps/x86/bits/floatn.h
@@ -28,7 +28,8 @@
    support, for x86_64 and x86.  */
 #if (defined __x86_64__							\
      ? __GNUC_PREREQ (4, 3)						\
-     : (defined __GNU__ ? __GNUC_PREREQ (4, 5) : __GNUC_PREREQ (4, 4)))
+     : (defined __GNU__ ? __GNUC_PREREQ (4, 5) : __GNUC_PREREQ (4, 4))) \
+    || defined __clang__
 # define __HAVE_FLOAT128 1
 #else
 # define __HAVE_FLOAT128 0
@@ -58,7 +59,7 @@
 /* Defined to concatenate the literal suffix to be used with _Float128
    types, if __HAVE_FLOAT128 is 1. */
 # if __HAVE_FLOAT128
-#  if !__GNUC_PREREQ (7, 0) || defined __cplusplus
+#  if !__GNUC_PREREQ (7, 0) || defined __clang__ || defined __cplusplus
 /* The literal suffix f128 exists only since GCC 7.0.  */
 #   define __f128(x) x##q
 #  else
@@ -68,7 +69,7 @@
 
 /* Defined to a complex binary128 type if __HAVE_FLOAT128 is 1.  */
 # if __HAVE_FLOAT128
-#  if !__GNUC_PREREQ (7, 0) || defined __cplusplus
+#  if !__GNUC_PREREQ (7, 0) || defined __clang__ || defined __cplusplus
 /* Add a typedef for older GCC compilers which don't natively support
    _Complex _Float128.  */
 typedef _Complex float __cfloat128 __attribute__ ((__mode__ (__TC__)));
@@ -82,12 +83,12 @@ typedef _Complex float __cfloat128 __attribute__ ((__mode__ (__TC__)));
 # if __HAVE_FLOAT128
 
 /* The type _Float128 exists only since GCC 7.0.  */
-#  if !__GNUC_PREREQ (7, 0) || defined __cplusplus
+#  if !__GNUC_PREREQ (7, 0) || defined __clang__ || defined __cplusplus
 typedef __float128 _Float128;
 #  endif
 
 /* __builtin_huge_valf128 doesn't exist before GCC 7.0.  */
-#  if !__GNUC_PREREQ (7, 0)
+#  if !__GNUC_PREREQ (7, 0) && !defined __clang__
 #   define __builtin_huge_valf128() ((_Float128) __builtin_huge_val ())
 #  endif
 
@@ -96,7 +97,7 @@ typedef __float128 _Float128;
    Converting a narrower sNaN to _Float128 produces a quiet NaN, so
    attempts to use _Float128 sNaNs will not work properly with older
    compilers.  */
-#  if !__GNUC_PREREQ (7, 0)
+#  if !__GNUC_PREREQ (7, 0) && !defined __clang__
 #   define __builtin_copysignf128 __builtin_copysignq
 #   define __builtin_fabsf128 __builtin_fabsq
 #   define __builtin_inff128() ((_Float128) __builtin_inf ())
@@ -108,7 +109,7 @@ typedef __float128 _Float128;
    e.g.: __builtin_signbitf128, before GCC 6.  However, there has never
    been a __builtin_signbitf128 in GCC and the type-generic builtin is
    only available since GCC 6.  */
-#  if !__GNUC_PREREQ (6, 0)
+#  if !__GNUC_PREREQ (6, 0) && !defined __clang__
 #   define __builtin_signbitf128 __signbitf128
 #  endif