diff --git a/renderscript/clang-include/Intrin.h b/renderscript/clang-include/Intrin.h
index e365abe..ff6d278 100644
--- a/renderscript/clang-include/Intrin.h
+++ b/renderscript/clang-include/Intrin.h
@@ -100,11 +100,9 @@
 unsigned int __popcnt(unsigned int);
 static __inline__
 unsigned short __popcnt16(unsigned short);
-static __inline__
-unsigned __int64 __rdtsc(void);
-unsigned __int64 __rdtscp(unsigned int *);
 unsigned long __readcr0(void);
 unsigned long __readcr2(void);
+static __inline__
 unsigned long __readcr3(void);
 unsigned long __readcr4(void);
 unsigned long __readcr8(void);
@@ -119,6 +117,7 @@
 static __inline__
 unsigned short __readfsword(unsigned long);
 #endif
+static __inline__
 unsigned __int64 __readmsr(unsigned long);
 unsigned __int64 __readpmc(unsigned long);
 unsigned long __segmentlimit(unsigned long);
@@ -143,6 +142,7 @@
 void __vmx_vmptrst(unsigned __int64 *);
 void __wbinvd(void);
 void __writecr0(unsigned int);
+static __inline__
 void __writecr3(unsigned int);
 void __writecr4(unsigned int);
 void __writecr8(unsigned int);
@@ -825,14 +825,12 @@
   return _Comparand;
 }
 #endif
-#ifdef __x86_64__
 static __inline__ __int64 __attribute__((__always_inline__, __nodebug__))
 _InterlockedCompareExchange64(__int64 volatile *_Destination,
                               __int64 _Exchange, __int64 _Comparand) {
   __atomic_compare_exchange(_Destination, &_Comparand, &_Exchange, 0, 0, 0);
   return _Comparand;
 }
-#endif
 /*----------------------------------------------------------------------------*\
 |* Barriers
 \*----------------------------------------------------------------------------*/
@@ -981,6 +979,35 @@
   __asm__ volatile ("hlt");
 }
 
+/*----------------------------------------------------------------------------*\
+|* Privileged intrinsics
+\*----------------------------------------------------------------------------*/
+static __inline__ unsigned __int64 __attribute__((__always_inline__, __nodebug__))
+__readmsr(unsigned long __register) {
+  // Loads the contents of a 64-bit model specific register (MSR) specified in
+  // the ECX register into registers EDX:EAX. The EDX register is loaded with
+  // the high-order 32 bits of the MSR and the EAX register is loaded with the
+  // low-order 32 bits. If less than 64 bits are implemented in the MSR being
+  // read, the values returned to EDX:EAX in unimplemented bit locations are
+  // undefined.
+  unsigned long __edx;
+  unsigned long __eax;
+  __asm__ ("rdmsr" : "=d"(__edx), "=a"(__eax) : "c"(__register));
+  return (((unsigned __int64)__edx) << 32) | (unsigned __int64)__eax;
+}
+
+static __inline__ unsigned long __attribute__((always_inline, __nodebug__))
+__readcr3(void) {
+  unsigned long __cr3_val;
+  __asm__ __volatile__ ("mov %%cr3, %0" : "=q"(__cr3_val) : : "memory");
+  return __cr3_val;
+}
+
+static __inline__ void __attribute__((always_inline, __nodebug__))
+__writecr3(unsigned int __cr3_val) {
+  __asm__ ("mov %0, %%cr3" : : "q"(__cr3_val) : "memory");
+}
+
 #ifdef __cplusplus
 }
 #endif
diff --git a/renderscript/clang-include/avx2intrin.h b/renderscript/clang-include/avx2intrin.h
index 9574469..394fdfe 100644
--- a/renderscript/clang-include/avx2intrin.h
+++ b/renderscript/clang-include/avx2intrin.h
@@ -160,7 +160,23 @@
 #define _mm256_blend_epi16(V1, V2, M) __extension__ ({ \
   __m256i __V1 = (V1); \
   __m256i __V2 = (V2); \
-  (__m256i)__builtin_ia32_pblendw256((__v16hi)__V1, (__v16hi)__V2, (M)); })
+  (__m256d)__builtin_shufflevector((__v16hi)__V1, (__v16hi)__V2, \
+                                   (((M) & 0x01) ? 16 : 0), \
+                                   (((M) & 0x02) ? 17 : 1), \
+                                   (((M) & 0x04) ? 18 : 2), \
+                                   (((M) & 0x08) ? 19 : 3), \
+                                   (((M) & 0x10) ? 20 : 4), \
+                                   (((M) & 0x20) ? 21 : 5), \
+                                   (((M) & 0x40) ? 22 : 6), \
+                                   (((M) & 0x80) ? 23 : 7), \
+                                   (((M) & 0x01) ? 24 : 8), \
+                                   (((M) & 0x02) ? 25 : 9), \
+                                   (((M) & 0x04) ? 26 : 10), \
+                                   (((M) & 0x08) ? 27 : 11), \
+                                   (((M) & 0x10) ? 28 : 12), \
+                                   (((M) & 0x20) ? 29 : 13), \
+                                   (((M) & 0x40) ? 30 : 14), \
+                                   (((M) & 0x80) ? 31 : 15)); })
 
 static __inline__ __m256i __attribute__((__always_inline__, __nodebug__))
 _mm256_cmpeq_epi8(__m256i __a, __m256i __b)
@@ -761,12 +777,24 @@
 #define _mm_blend_epi32(V1, V2, M) __extension__ ({ \
   __m128i __V1 = (V1); \
   __m128i __V2 = (V2); \
-  (__m128i)__builtin_ia32_pblendd128((__v4si)__V1, (__v4si)__V2, (M)); })
+  (__m128i)__builtin_shufflevector((__v4si)__V1, (__v4si)__V2, \
+                                   (((M) & 0x01) ? 4 : 0), \
+                                   (((M) & 0x02) ? 5 : 1), \
+                                   (((M) & 0x04) ? 6 : 2), \
+                                   (((M) & 0x08) ? 7 : 3)); })
 
 #define _mm256_blend_epi32(V1, V2, M) __extension__ ({ \
   __m256i __V1 = (V1); \
   __m256i __V2 = (V2); \
-  (__m256i)__builtin_ia32_pblendd256((__v8si)__V1, (__v8si)__V2, (M)); })
+  (__m256i)__builtin_shufflevector((__v8si)__V1, (__v8si)__V2, \
+                                   (((M) & 0x01) ?  8 : 0), \
+                                   (((M) & 0x02) ?  9 : 1), \
+                                   (((M) & 0x04) ? 10 : 2), \
+                                   (((M) & 0x08) ? 11 : 3), \
+                                   (((M) & 0x10) ? 12 : 4), \
+                                   (((M) & 0x20) ? 13 : 5), \
+                                   (((M) & 0x40) ? 14 : 6), \
+                                   (((M) & 0x80) ? 15 : 7)); })
 
 static __inline__ __m256i __attribute__((__always_inline__, __nodebug__))
 _mm256_broadcastb_epi8(__m128i __X)
diff --git a/renderscript/clang-include/avxintrin.h b/renderscript/clang-include/avxintrin.h
index 141c4d9..3d50439 100644
--- a/renderscript/clang-include/avxintrin.h
+++ b/renderscript/clang-include/avxintrin.h
@@ -308,12 +308,24 @@
 #define _mm256_blend_pd(V1, V2, M) __extension__ ({ \
   __m256d __V1 = (V1); \
   __m256d __V2 = (V2); \
-  (__m256d)__builtin_ia32_blendpd256((__v4df)__V1, (__v4df)__V2, (M)); })
+  (__m256d)__builtin_shufflevector((__v4df)__V1, (__v4df)__V2, \
+                                   (((M) & 0x01) ? 4 : 0), \
+                                   (((M) & 0x02) ? 5 : 1), \
+                                   (((M) & 0x04) ? 6 : 2), \
+                                   (((M) & 0x08) ? 7 : 3)); })
 
 #define _mm256_blend_ps(V1, V2, M) __extension__ ({ \
   __m256 __V1 = (V1); \
   __m256 __V2 = (V2); \
-  (__m256)__builtin_ia32_blendps256((__v8sf)__V1, (__v8sf)__V2, (M)); })
+  (__m256)__builtin_shufflevector((__v8sf)__V1, (__v8sf)__V2, \
+                                  (((M) & 0x01) ?  8 : 0), \
+                                  (((M) & 0x02) ?  9 : 1), \
+                                  (((M) & 0x04) ? 10 : 2), \
+                                  (((M) & 0x08) ? 11 : 3), \
+                                  (((M) & 0x10) ? 12 : 4), \
+                                  (((M) & 0x20) ? 13 : 5), \
+                                  (((M) & 0x40) ? 14 : 6), \
+                                  (((M) & 0x80) ? 15 : 7)); })
 
 static __inline __m256d __attribute__((__always_inline__, __nodebug__))
 _mm256_blendv_pd(__m256d __a, __m256d __b, __m256d __c)
diff --git a/renderscript/clang-include/ia32intrin.h b/renderscript/clang-include/ia32intrin.h
index a5985f6..55c2247 100644
--- a/renderscript/clang-include/ia32intrin.h
+++ b/renderscript/clang-include/ia32intrin.h
@@ -82,9 +82,13 @@
 /* __rdtsc */
 static __inline__ unsigned long long __attribute__((__always_inline__, __nodebug__))
 __rdtsc(void) {
-  unsigned int __eax, __edx;
-  __asm__ ("rdtsc" : "=a" (__eax), "=d" (__edx));
-  return ((unsigned long long)__edx << 32) | __eax;
+  return __builtin_ia32_rdtsc();
+}
+
+/* __rdtscp */
+static __inline__ unsigned long long __attribute__((__always_inline__, __nodebug__))
+__rdtscp(unsigned int *__A) {
+  return __builtin_ia32_rdtscp(__A);
 }
 
 #define _rdtsc() __rdtsc()
diff --git a/renderscript/clang-include/smmintrin.h b/renderscript/clang-include/smmintrin.h
index 53b3ccb..6e35734 100644
--- a/renderscript/clang-include/smmintrin.h
+++ b/renderscript/clang-include/smmintrin.h
@@ -79,12 +79,18 @@
 #define _mm_blend_pd(V1, V2, M) __extension__ ({ \
   __m128d __V1 = (V1); \
   __m128d __V2 = (V2); \
-  (__m128d) __builtin_ia32_blendpd ((__v2df)__V1, (__v2df)__V2, (M)); })
+  (__m128d)__builtin_shufflevector((__v2df)__V1, (__v2df)__V2, \
+                                   (((M) & 0x01) ? 2 : 0), \
+                                   (((M) & 0x02) ? 3 : 1)); })
 
 #define _mm_blend_ps(V1, V2, M) __extension__ ({ \
   __m128 __V1 = (V1); \
   __m128 __V2 = (V2); \
-  (__m128) __builtin_ia32_blendps ((__v4sf)__V1, (__v4sf)__V2, (M)); })
+  (__m128)__builtin_shufflevector((__v4sf)__V1, (__v4sf)__V2, \
+                                  (((M) & 0x01) ? 4 : 0), \
+                                  (((M) & 0x02) ? 5 : 1), \
+                                  (((M) & 0x04) ? 6 : 2), \
+                                  (((M) & 0x08) ? 7 : 3)); })
 
 static __inline__ __m128d __attribute__((__always_inline__, __nodebug__))
 _mm_blendv_pd (__m128d __V1, __m128d __V2, __m128d __M)
@@ -110,7 +116,15 @@
 #define _mm_blend_epi16(V1, V2, M) __extension__ ({ \
   __m128i __V1 = (V1); \
   __m128i __V2 = (V2); \
-  (__m128i) __builtin_ia32_pblendw128 ((__v8hi)__V1, (__v8hi)__V2, (M)); })
+  (__m128i)__builtin_shufflevector((__v8hi)__V1, (__v8hi)__V2, \
+                                   (((M) & 0x01) ?  8 : 0), \
+                                   (((M) & 0x02) ?  9 : 1), \
+                                   (((M) & 0x04) ? 10 : 2), \
+                                   (((M) & 0x08) ? 11 : 3), \
+                                   (((M) & 0x10) ? 12 : 4), \
+                                   (((M) & 0x20) ? 13 : 5), \
+                                   (((M) & 0x40) ? 14 : 6), \
+                                   (((M) & 0x80) ? 15 : 7)); })
 
 /* SSE4 Dword Multiply Instructions.  */
 static __inline__  __m128i __attribute__((__always_inline__, __nodebug__))
diff --git a/renderscript/clang-include/stddef.h b/renderscript/clang-include/stddef.h
index 97126ed..2dfe0a2 100644
--- a/renderscript/clang-include/stddef.h
+++ b/renderscript/clang-include/stddef.h
@@ -23,9 +23,22 @@
  *===-----------------------------------------------------------------------===
  */
 
-#ifndef __STDDEF_H
-#define __STDDEF_H
+#if !defined(__STDDEF_H) || defined(__need_ptrdiff_t) ||                       \
+    defined(__need_size_t) || defined(__need_wchar_t) ||                       \
+    defined(__need_NULL) || defined(__need_wint_t)
 
+#if !defined(__need_ptrdiff_t) && !defined(__need_size_t) &&                   \
+    !defined(__need_wchar_t) && !defined(__need_NULL) &&                       \
+    !defined(__need_wint_t)
+#define __STDDEF_H
+#define __need_ptrdiff_t
+#define __need_size_t
+#define __need_wchar_t
+#define __need_NULL
+/* __need_wint_t is intentionally not defined here. */
+#endif
+
+#if defined(__need_ptrdiff_t)
 #if !defined(_PTRDIFF_T) || __has_feature(modules)
 /* Always define ptrdiff_t when modules are available. */
 #if !__has_feature(modules)
@@ -33,7 +46,10 @@
 #endif
 typedef __PTRDIFF_TYPE__ ptrdiff_t;
 #endif
+#undef __need_ptrdiff_t
+#endif /* defined(__need_ptrdiff_t) */
 
+#if defined(__need_size_t)
 #if !defined(_SIZE_T) || __has_feature(modules)
 /* Always define size_t when modules are available. */
 #if !__has_feature(modules)
@@ -41,7 +57,10 @@
 #endif
 typedef __SIZE_TYPE__ size_t;
 #endif
+#undef __need_size_t
+#endif /*defined(__need_size_t) */
 
+#if defined(__STDDEF_H)
 /* ISO9899:2011 7.20 (C11 Annex K): Define rsize_t if __STDC_WANT_LIB_EXT1__ is
  * enabled. */
 #if (defined(__STDC_WANT_LIB_EXT1__) && __STDC_WANT_LIB_EXT1__ >= 1 && \
@@ -52,7 +71,9 @@
 #endif
 typedef __SIZE_TYPE__ rsize_t;
 #endif
+#endif /* defined(__STDDEF_H) */
 
+#if defined(__need_wchar_t)
 #ifndef __cplusplus
 /* Always define wchar_t when modules are available. */
 #if !defined(_WCHAR_T) || __has_feature(modules)
@@ -65,7 +86,10 @@
 typedef __WCHAR_TYPE__ wchar_t;
 #endif
 #endif
+#undef __need_wchar_t
+#endif /* defined(__need_wchar_t) */
 
+#if defined(__need_NULL)
 #undef NULL
 #ifdef __cplusplus
 #  if !defined(__MINGW32__) && !defined(_MSC_VER)
@@ -76,15 +100,19 @@
 #else
 #  define NULL ((void*)0)
 #endif
-
 #ifdef __cplusplus
 #if defined(_MSC_EXTENSIONS) && defined(_NATIVE_NULLPTR_SUPPORTED)
 namespace std { typedef decltype(nullptr) nullptr_t; }
 using ::std::nullptr_t;
 #endif
 #endif
+#undef __need_NULL
+#endif /* defined(__need_NULL) */
+
+#if defined(__STDDEF_H)
 
 #if __STDC_VERSION__ >= 201112L || __cplusplus >= 201103L
+#if !defined(__CLANG_MAX_ALIGN_T_DEFINED) || __has_feature(modules)
 #ifndef _MSC_VER
 typedef struct {
   long long __clang_max_align_nonce1
@@ -97,10 +125,10 @@
 #endif
 #define __CLANG_MAX_ALIGN_T_DEFINED
 #endif
+#endif
 
 #define offsetof(t, d) __builtin_offsetof(t, d)
-
-#endif /* __STDDEF_H */
+#endif  /* __STDDEF_H */
 
 /* Some C libraries expect to see a wint_t here. Others (notably MinGW) will use
 __WINT_TYPE__ directly; accommodate both by requiring __need_wint_t */
@@ -114,3 +142,5 @@
 #endif
 #undef __need_wint_t
 #endif /* __need_wint_t */
+
+#endif
diff --git a/renderscript/clang-include/xmmintrin.h b/renderscript/clang-include/xmmintrin.h
index c8765a7..e777ec0 100644
--- a/renderscript/clang-include/xmmintrin.h
+++ b/renderscript/clang-include/xmmintrin.h
@@ -905,7 +905,7 @@
   __a = _mm_movehl_ps(__a, __a);
   __c = _mm_cvtps_pi32(__a);
   
-  return _mm_packs_pi16(__b, __c);
+  return _mm_packs_pi32(__b, __c);
 }
 
 static __inline__ __m64 __attribute__((__always_inline__, __nodebug__))