[X86] Add gcc rotate intrinsics to ia32intrin.h

[craig.topper]

Summary

This is another attempt at what Erich Keane tried to do in r355322.a

This adds rolb, rolw, rold, rolq and their ror equivalent as always_inline wrappers around __builtin_rotate* which will lower to funnel shift intrinsics in IR.

Additionally, when _MSC_VER is not defined we will define _rotl, _lrotl, _rotr, _lrotr as macros to one of the always_inline intrinsics mentioned above. Making sure that _lrotl/_lrotr use either 32 or 64 bit based on the size of long. These need to be macros because we have builtins with the same name for MS compatibility, but _MSC_VER isn't always defined when those builtins are enabled.

We also define _rotwl and _rotwr as macros aliasing to rolw/rorw just like gcc to complete the set. These don't need to be gated with _MSC_VER because these aren't MS builtins.

I've added tests both for non-MS and -ms-extensions with and without _MSC_VER being defined.

[craig.topper]

Add the test file

[craig.topper]

Fix bad comment copy/paste

[rnk]

Looks good to me. Let's wait for @jyknight though.

lib/Headers/ia32intrin.h

120

Can you add comments here indicating that these are already provided as builtins for MSVC, so these macros are unnecessary?

133

And, add a comment here indicating that these are not builtins, so they are provided in all modes.

[craig.topper]

Add comments based on rnk's review

[jyknight]

Looks good.

lib/Headers/ia32intrin.h

 
 #define _rdpmc(A) __rdpmc(A)
 
 static __inline__ void __attribute__((__always_inline__, __nodebug__))
 _wbinvd(void) {
   __builtin_ia32_wbinvd();
 }
 
+static __inline__ unsigned char __attribute__((__always_inline__, __nodebug__))
+__rolb(unsigned char __X, int __C) {
+  return __builtin_rotateleft8(__X, __C);
+}
+
+static __inline__ unsigned char __attribute__((__always_inline__, __nodebug__))
+__rorb(unsigned char __X, int __C) {
+  return __builtin_rotateright8(__X, __C);
+}
+
+static __inline__ unsigned short __attribute__((__always_inline__, __nodebug__))
+__rolw(unsigned short __X, int __C) {
+  return __builtin_rotateleft16(__X, __C);
+}
+
+static __inline__ unsigned short __attribute__((__always_inline__, __nodebug__))
+__rorw(unsigned short __X, int __C) {
+  return __builtin_rotateright16(__X, __C);
+}
+
+static __inline__ unsigned int __attribute__((__always_inline__, __nodebug__))
+__rold(unsigned int __X, int __C) {
+  return __builtin_rotateleft32(__X, __C);
+}
+
+static __inline__ unsigned int __attribute__((__always_inline__, __nodebug__))
+__rord(unsigned int __X, int __C) {
+  return __builtin_rotateright32(__X, __C);
+}
+
+#ifdef __x86_64__
+static __inline__ unsigned long long __attribute__((__always_inline__, __nodebug__))
+__rolq(unsigned long long __X, int __C) {
+  return __builtin_rotateleft64(__X, __C);
+}
+
+static __inline__ unsigned long long __attribute__((__always_inline__, __nodebug__))
+__rorq(unsigned long long __X, int __C) {
+  return __builtin_rotateright64(__X, __C);
+}
+#endif /* !__x86_64__ */
+
+#ifndef _MSC_VER
+/* Select the correct function based on the size of long. */
+#ifdef __LP64__
+#define _lrotl(a,b) __rolq((a), (b))
+#define _lrotr(a,b) __rorq((a), (b))
+#else
+#define _lrotl(a,b) __rold((a), (b))
+#define _lrotr(a,b) __rord((a), (b))
+#endif
+#define _rotl(a,b) __rold((a), (b))
+#define _rotr(a,b) __rord((a), (b))
+#endif // _MSC_VER
+
+#define _rotwl(a,b) __rolw((a), (b))
+#define _rotwr(a,b) __rorw((a), (b))
+
 #endif /* __IA32INTRIN_H */