diff --git a/clang/include/clang/Basic/BuiltinsX86.def b/clang/include/clang/Basic/BuiltinsX86.def
--- a/clang/include/clang/Basic/BuiltinsX86.def
+++ b/clang/include/clang/Basic/BuiltinsX86.def
@@ -365,7 +365,7 @@
 TARGET_BUILTIN(__builtin_ia32_pslldqi128_byteshift, "V2OiV2OiIi", "ncV:128:", "sse2")
 TARGET_BUILTIN(__builtin_ia32_psrldqi128_byteshift, "V2OiV2OiIi", "ncV:128:", "sse2")
 
-TARGET_BUILTIN(__builtin_ia32_monitor, "vv*UiUi", "n", "sse3")
+TARGET_BUILTIN(__builtin_ia32_monitor, "vvC*UiUi", "n", "sse3")
 TARGET_BUILTIN(__builtin_ia32_mwait, "vUiUi", "n", "sse3")
 TARGET_BUILTIN(__builtin_ia32_lddqu, "V16ccC*", "nV:128:", "sse3")
 
@@ -1396,8 +1396,8 @@
 TARGET_BUILTIN(__builtin_ia32_psrlwi512, "V32sV32si", "ncV:512:", "avx512bw")
 TARGET_BUILTIN(__builtin_ia32_pslldqi512_byteshift, "V8OiV8OiIi", "ncV:512:", "avx512bw")
 TARGET_BUILTIN(__builtin_ia32_psrldqi512_byteshift, "V8OiV8OiIi", "ncV:512:", "avx512bw")
-TARGET_BUILTIN(__builtin_ia32_movdqa32load128_mask, "V4iV4i*V4iUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_movdqa32load256_mask, "V8iV8i*V8iUc", "nV:256:", "avx512vl")
+TARGET_BUILTIN(__builtin_ia32_movdqa32load128_mask, "V4iV4iC*V4iUc", "nV:128:", "avx512vl")
+TARGET_BUILTIN(__builtin_ia32_movdqa32load256_mask, "V8iV8iC*V8iUc", "nV:256:", "avx512vl")
 TARGET_BUILTIN(__builtin_ia32_movdqa32load512_mask, "V16iV16iC*V16iUs", "nV:512:", "avx512f")
 TARGET_BUILTIN(__builtin_ia32_movdqa32store512_mask, "vV16i*V16iUs", "nV:512:", "avx512f")
 TARGET_BUILTIN(__builtin_ia32_movdqa64load512_mask, "V8OiV8OiC*V8OiUc", "nV:512:", "avx512f")
@@ -1418,8 +1418,8 @@
 TARGET_BUILTIN(__builtin_ia32_vcomiss, "iV4fV4fIiIi", "ncV:128:", "avx512f")
 TARGET_BUILTIN(__builtin_ia32_kunpckdi, "UOiUOiUOi", "nc", "avx512bw")
 TARGET_BUILTIN(__builtin_ia32_kunpcksi, "UiUiUi", "nc", "avx512bw")
-TARGET_BUILTIN(__builtin_ia32_loaddquhi512_mask, "V32sV32s*V32sUi", "nV:512:", "avx512bw")
-TARGET_BUILTIN(__builtin_ia32_loaddquqi512_mask, "V64cV64c*V64cUOi", "nV:512:", "avx512bw")
+TARGET_BUILTIN(__builtin_ia32_loaddquhi512_mask, "V32sV32sC*V32sUi", "nV:512:", "avx512bw")
+TARGET_BUILTIN(__builtin_ia32_loaddquqi512_mask, "V64cV64cC*V64cUOi", "nV:512:", "avx512bw")
 TARGET_BUILTIN(__builtin_ia32_fixupimmpd512_mask, "V8dV8dV8dV8OiIiUcIi", "ncV:512:", "avx512f")
 TARGET_BUILTIN(__builtin_ia32_fixupimmpd512_maskz, "V8dV8dV8dV8OiIiUcIi", "ncV:512:", "avx512f")
 TARGET_BUILTIN(__builtin_ia32_fixupimmps512_mask, "V16fV16fV16fV16iIiUsIi", "ncV:512:", "avx512f")
@@ -1432,10 +1432,10 @@
 TARGET_BUILTIN(__builtin_ia32_getexpss128_round_mask, "V4fV4fV4fV4fUcIi", "ncV:128:", "avx512f")
 TARGET_BUILTIN(__builtin_ia32_getmantsd_round_mask, "V2dV2dV2dIiV2dUcIi", "ncV:128:", "avx512f")
 TARGET_BUILTIN(__builtin_ia32_getmantss_round_mask, "V4fV4fV4fIiV4fUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_loaddquhi128_mask, "V8sV8s*V8sUc", "nV:128:", "avx512bw,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_loaddquhi256_mask, "V16sV16s*V16sUs", "nV:256:", "avx512bw,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_loaddquqi128_mask, "V16cV16c*V16cUs", "nV:128:", "avx512bw,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_loaddquqi256_mask, "V32cV32c*V32cUi", "nV:256:", "avx512bw,avx512vl")
+TARGET_BUILTIN(__builtin_ia32_loaddquhi128_mask, "V8sV8sC*V8sUc", "nV:128:", "avx512bw,avx512vl")
+TARGET_BUILTIN(__builtin_ia32_loaddquhi256_mask, "V16sV16sC*V16sUs", "nV:256:", "avx512bw,avx512vl")
+TARGET_BUILTIN(__builtin_ia32_loaddquqi128_mask, "V16cV16cC*V16cUs", "nV:128:", "avx512bw,avx512vl")
+TARGET_BUILTIN(__builtin_ia32_loaddquqi256_mask, "V32cV32cC*V32cUi", "nV:256:", "avx512bw,avx512vl")
 TARGET_BUILTIN(__builtin_ia32_fixupimmpd128_mask, "V2dV2dV2dV2OiIiUc", "ncV:128:", "avx512vl")
 TARGET_BUILTIN(__builtin_ia32_fixupimmpd128_maskz, "V2dV2dV2dV2OiIiUc", "ncV:128:", "avx512vl")
 TARGET_BUILTIN(__builtin_ia32_fixupimmpd256_mask, "V4dV4dV4dV4OiIiUc", "ncV:256:", "avx512vl")
@@ -1444,20 +1444,20 @@
 TARGET_BUILTIN(__builtin_ia32_fixupimmps128_maskz, "V4fV4fV4fV4iIiUc", "ncV:128:", "avx512vl")
 TARGET_BUILTIN(__builtin_ia32_fixupimmps256_mask, "V8fV8fV8fV8iIiUc", "ncV:256:", "avx512vl")
 TARGET_BUILTIN(__builtin_ia32_fixupimmps256_maskz, "V8fV8fV8fV8iIiUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_loadapd128_mask, "V2dV2d*V2dUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_loadsd128_mask, "V2dV2d*V2dUc", "nV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_loadapd256_mask, "V4dV4d*V4dUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_loadaps128_mask, "V4fV4f*V4fUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_loadss128_mask, "V4fV4f*V4fUc", "nV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_loadaps256_mask, "V8fV8f*V8fUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_loaddqudi128_mask, "V2OiV2Oi*V2OiUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_loaddqudi256_mask, "V4OiV4Oi*V4OiUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_loaddqusi128_mask, "V4iV4i*V4iUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_loaddqusi256_mask, "V8iV8i*V8iUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_loadupd128_mask, "V2dV2d*V2dUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_loadupd256_mask, "V4dV4d*V4dUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_loadups128_mask, "V4fV4f*V4fUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_loadups256_mask, "V8fV8f*V8fUc", "nV:256:", "avx512vl")
+TARGET_BUILTIN(__builtin_ia32_loadapd128_mask, "V2dV2dC*V2dUc", "nV:128:", "avx512vl")
+TARGET_BUILTIN(__builtin_ia32_loadsd128_mask, "V2dV2dC*V2dUc", "nV:128:", "avx512f")
+TARGET_BUILTIN(__builtin_ia32_loadapd256_mask, "V4dV4dC*V4dUc", "nV:256:", "avx512vl")
+TARGET_BUILTIN(__builtin_ia32_loadaps128_mask, "V4fV4fC*V4fUc", "nV:128:", "avx512vl")
+TARGET_BUILTIN(__builtin_ia32_loadss128_mask, "V4fV4fC*V4fUc", "nV:128:", "avx512f")
+TARGET_BUILTIN(__builtin_ia32_loadaps256_mask, "V8fV8fC*V8fUc", "nV:256:", "avx512vl")
+TARGET_BUILTIN(__builtin_ia32_loaddqudi128_mask, "V2OiV2OiC*V2OiUc", "nV:128:", "avx512vl")
+TARGET_BUILTIN(__builtin_ia32_loaddqudi256_mask, "V4OiV4OiC*V4OiUc", "nV:256:", "avx512vl")
+TARGET_BUILTIN(__builtin_ia32_loaddqusi128_mask, "V4iV4iC*V4iUc", "nV:128:", "avx512vl")
+TARGET_BUILTIN(__builtin_ia32_loaddqusi256_mask, "V8iV8iC*V8iUc", "nV:256:", "avx512vl")
+TARGET_BUILTIN(__builtin_ia32_loadupd128_mask, "V2dV2dC*V2dUc", "nV:128:", "avx512vl")
+TARGET_BUILTIN(__builtin_ia32_loadupd256_mask, "V4dV4dC*V4dUc", "nV:256:", "avx512vl")
+TARGET_BUILTIN(__builtin_ia32_loadups128_mask, "V4fV4fC*V4fUc", "nV:128:", "avx512vl")
+TARGET_BUILTIN(__builtin_ia32_loadups256_mask, "V8fV8fC*V8fUc", "nV:256:", "avx512vl")
 TARGET_BUILTIN(__builtin_ia32_storedquhi512_mask, "vV32s*V32sUi", "nV:512:", "avx512bw")
 TARGET_BUILTIN(__builtin_ia32_storedquqi512_mask, "vV64c*V64cUOi", "nV:512:", "avx512bw")
 TARGET_BUILTIN(__builtin_ia32_storedquhi128_mask, "vV8s*V8sUc", "nV:128:", "avx512vl,avx512bw")
@@ -1872,11 +1872,11 @@
 TARGET_BUILTIN(__builtin_ia32_selectsd_128, "V2dUcV2dV2d", "ncV:128:", "avx512f")
 
 // MONITORX/MWAITX
-TARGET_BUILTIN(__builtin_ia32_monitorx, "vv*UiUi", "n", "mwaitx")
+TARGET_BUILTIN(__builtin_ia32_monitorx, "vvC*UiUi", "n", "mwaitx")
 TARGET_BUILTIN(__builtin_ia32_mwaitx, "vUiUiUi", "n", "mwaitx")
 
 // WAITPKG
-TARGET_BUILTIN(__builtin_ia32_umonitor, "vv*", "n", "waitpkg")
+TARGET_BUILTIN(__builtin_ia32_umonitor, "vvC*", "n", "waitpkg")
 TARGET_BUILTIN(__builtin_ia32_umwait, "UcUiUiUi", "n", "waitpkg")
 TARGET_BUILTIN(__builtin_ia32_tpause, "UcUiUiUi", "n", "waitpkg")
 
diff --git a/clang/lib/Headers/avx512bwintrin.h b/clang/lib/Headers/avx512bwintrin.h
--- a/clang/lib/Headers/avx512bwintrin.h
+++ b/clang/lib/Headers/avx512bwintrin.h
@@ -1731,13 +1731,13 @@
   struct __loadu_epi16 {
     __m512i_u __v;
   } __attribute__((__packed__, __may_alias__));
-  return ((struct __loadu_epi16*)__P)->__v;
+  return ((const struct __loadu_epi16*)__P)->__v;
 }
 
 static __inline__ __m512i __DEFAULT_FN_ATTRS512
 _mm512_mask_loadu_epi16 (__m512i __W, __mmask32 __U, void const *__P)
 {
-  return (__m512i) __builtin_ia32_loaddquhi512_mask ((__v32hi *) __P,
+  return (__m512i) __builtin_ia32_loaddquhi512_mask ((const __v32hi *) __P,
                  (__v32hi) __W,
                  (__mmask32) __U);
 }
@@ -1745,7 +1745,7 @@
 static __inline__ __m512i __DEFAULT_FN_ATTRS512
 _mm512_maskz_loadu_epi16 (__mmask32 __U, void const *__P)
 {
-  return (__m512i) __builtin_ia32_loaddquhi512_mask ((__v32hi *) __P,
+  return (__m512i) __builtin_ia32_loaddquhi512_mask ((const __v32hi *) __P,
                  (__v32hi)
                  _mm512_setzero_si512 (),
                  (__mmask32) __U);
@@ -1757,13 +1757,13 @@
   struct __loadu_epi8 {
     __m512i_u __v;
   } __attribute__((__packed__, __may_alias__));
-  return ((struct __loadu_epi8*)__P)->__v;
+  return ((const struct __loadu_epi8*)__P)->__v;
 }
 
 static __inline__ __m512i __DEFAULT_FN_ATTRS512
 _mm512_mask_loadu_epi8 (__m512i __W, __mmask64 __U, void const *__P)
 {
-  return (__m512i) __builtin_ia32_loaddquqi512_mask ((__v64qi *) __P,
+  return (__m512i) __builtin_ia32_loaddquqi512_mask ((const __v64qi *) __P,
                  (__v64qi) __W,
                  (__mmask64) __U);
 }
@@ -1771,7 +1771,7 @@
 static __inline__ __m512i __DEFAULT_FN_ATTRS512
 _mm512_maskz_loadu_epi8 (__mmask64 __U, void const *__P)
 {
-  return (__m512i) __builtin_ia32_loaddquqi512_mask ((__v64qi *) __P,
+  return (__m512i) __builtin_ia32_loaddquqi512_mask ((const __v64qi *) __P,
                  (__v64qi)
                  _mm512_setzero_si512 (),
                  (__mmask64) __U);
diff --git a/clang/lib/Headers/avx512fintrin.h b/clang/lib/Headers/avx512fintrin.h
--- a/clang/lib/Headers/avx512fintrin.h
+++ b/clang/lib/Headers/avx512fintrin.h
@@ -4305,7 +4305,7 @@
   struct __loadu_si512 {
     __m512i_u __v;
   } __attribute__((__packed__, __may_alias__));
-  return ((struct __loadu_si512*)__P)->__v;
+  return ((const struct __loadu_si512*)__P)->__v;
 }
 
 static __inline __m512i __DEFAULT_FN_ATTRS512
@@ -4314,7 +4314,7 @@
   struct __loadu_epi32 {
     __m512i_u __v;
   } __attribute__((__packed__, __may_alias__));
-  return ((struct __loadu_epi32*)__P)->__v;
+  return ((const struct __loadu_epi32*)__P)->__v;
 }
 
 static __inline __m512i __DEFAULT_FN_ATTRS512
@@ -4341,7 +4341,7 @@
   struct __loadu_epi64 {
     __m512i_u __v;
   } __attribute__((__packed__, __may_alias__));
-  return ((struct __loadu_epi64*)__P)->__v;
+  return ((const struct __loadu_epi64*)__P)->__v;
 }
 
 static __inline __m512i __DEFAULT_FN_ATTRS512
@@ -4401,7 +4401,7 @@
   struct __loadu_pd {
     __m512d_u __v;
   } __attribute__((__packed__, __may_alias__));
-  return ((struct __loadu_pd*)__p)->__v;
+  return ((const struct __loadu_pd*)__p)->__v;
 }
 
 static __inline __m512 __DEFAULT_FN_ATTRS512
@@ -4410,13 +4410,13 @@
   struct __loadu_ps {
     __m512_u __v;
   } __attribute__((__packed__, __may_alias__));
-  return ((struct __loadu_ps*)__p)->__v;
+  return ((const struct __loadu_ps*)__p)->__v;
 }
 
 static __inline __m512 __DEFAULT_FN_ATTRS512
 _mm512_load_ps(void const *__p)
 {
-  return *(__m512*)__p;
+  return *(const __m512*)__p;
 }
 
 static __inline __m512 __DEFAULT_FN_ATTRS512
@@ -4439,7 +4439,7 @@
 static __inline __m512d __DEFAULT_FN_ATTRS512
 _mm512_load_pd(void const *__p)
 {
-  return *(__m512d*)__p;
+  return *(const __m512d*)__p;
 }
 
 static __inline __m512d __DEFAULT_FN_ATTRS512
@@ -4462,19 +4462,19 @@
 static __inline __m512i __DEFAULT_FN_ATTRS512
 _mm512_load_si512 (void const *__P)
 {
-  return *(__m512i *) __P;
+  return *(const __m512i *) __P;
 }
 
 static __inline __m512i __DEFAULT_FN_ATTRS512
 _mm512_load_epi32 (void const *__P)
 {
-  return *(__m512i *) __P;
+  return *(const __m512i *) __P;
 }
 
 static __inline __m512i __DEFAULT_FN_ATTRS512
 _mm512_load_epi64 (void const *__P)
 {
-  return *(__m512i *) __P;
+  return *(const __m512i *) __P;
 }
 
 /* SIMD store ops */
@@ -8724,13 +8724,13 @@
                                                 (__v4sf)_mm_setzero_ps(),
                                                 0, 4, 4, 4);
 
-  return (__m128) __builtin_ia32_loadss128_mask ((__v4sf *) __A, src, __U & 1);
+  return (__m128) __builtin_ia32_loadss128_mask ((const __v4sf *) __A, src, __U & 1);
 }
 
 static __inline__ __m128 __DEFAULT_FN_ATTRS128
 _mm_maskz_load_ss (__mmask8 __U, const float* __A)
 {
-  return (__m128)__builtin_ia32_loadss128_mask ((__v4sf *) __A,
+  return (__m128)__builtin_ia32_loadss128_mask ((const __v4sf *) __A,
                                                 (__v4sf) _mm_setzero_ps(),
                                                 __U & 1);
 }
@@ -8742,13 +8742,13 @@
                                                  (__v2df)_mm_setzero_pd(),
                                                  0, 2);
 
-  return (__m128d) __builtin_ia32_loadsd128_mask ((__v2df *) __A, src, __U & 1);
+  return (__m128d) __builtin_ia32_loadsd128_mask ((const __v2df *) __A, src, __U & 1);
 }
 
 static __inline__ __m128d __DEFAULT_FN_ATTRS128
 _mm_maskz_load_sd (__mmask8 __U, const double* __A)
 {
-  return (__m128d) __builtin_ia32_loadsd128_mask ((__v2df *) __A,
+  return (__m128d) __builtin_ia32_loadsd128_mask ((const __v2df *) __A,
                                                   (__v2df) _mm_setzero_pd(),
                                                   __U & 1);
 }
diff --git a/clang/lib/Headers/avx512vlbwintrin.h b/clang/lib/Headers/avx512vlbwintrin.h
--- a/clang/lib/Headers/avx512vlbwintrin.h
+++ b/clang/lib/Headers/avx512vlbwintrin.h
@@ -2289,13 +2289,13 @@
   struct __loadu_epi16 {
     __m128i_u __v;
   } __attribute__((__packed__, __may_alias__));
-  return ((struct __loadu_epi16*)__P)->__v;
+  return ((const struct __loadu_epi16*)__P)->__v;
 }
 
 static __inline__ __m128i __DEFAULT_FN_ATTRS128
 _mm_mask_loadu_epi16 (__m128i __W, __mmask8 __U, void const *__P)
 {
-  return (__m128i) __builtin_ia32_loaddquhi128_mask ((__v8hi *) __P,
+  return (__m128i) __builtin_ia32_loaddquhi128_mask ((const __v8hi *) __P,
                  (__v8hi) __W,
                  (__mmask8) __U);
 }
@@ -2303,7 +2303,7 @@
 static __inline__ __m128i __DEFAULT_FN_ATTRS128
 _mm_maskz_loadu_epi16 (__mmask8 __U, void const *__P)
 {
-  return (__m128i) __builtin_ia32_loaddquhi128_mask ((__v8hi *) __P,
+  return (__m128i) __builtin_ia32_loaddquhi128_mask ((const __v8hi *) __P,
                  (__v8hi)
                  _mm_setzero_si128 (),
                  (__mmask8) __U);
@@ -2315,13 +2315,13 @@
   struct __loadu_epi16 {
     __m256i_u __v;
   } __attribute__((__packed__, __may_alias__));
-  return ((struct __loadu_epi16*)__P)->__v;
+  return ((const struct __loadu_epi16*)__P)->__v;
 }
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_mask_loadu_epi16 (__m256i __W, __mmask16 __U, void const *__P)
 {
-  return (__m256i) __builtin_ia32_loaddquhi256_mask ((__v16hi *) __P,
+  return (__m256i) __builtin_ia32_loaddquhi256_mask ((const __v16hi *) __P,
                  (__v16hi) __W,
                  (__mmask16) __U);
 }
@@ -2329,7 +2329,7 @@
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_maskz_loadu_epi16 (__mmask16 __U, void const *__P)
 {
-  return (__m256i) __builtin_ia32_loaddquhi256_mask ((__v16hi *) __P,
+  return (__m256i) __builtin_ia32_loaddquhi256_mask ((const __v16hi *) __P,
                  (__v16hi)
                  _mm256_setzero_si256 (),
                  (__mmask16) __U);
@@ -2341,13 +2341,13 @@
   struct __loadu_epi8 {
     __m128i_u __v;
   } __attribute__((__packed__, __may_alias__));
-  return ((struct __loadu_epi8*)__P)->__v;
+  return ((const struct __loadu_epi8*)__P)->__v;
 }
 
 static __inline__ __m128i __DEFAULT_FN_ATTRS128
 _mm_mask_loadu_epi8 (__m128i __W, __mmask16 __U, void const *__P)
 {
-  return (__m128i) __builtin_ia32_loaddquqi128_mask ((__v16qi *) __P,
+  return (__m128i) __builtin_ia32_loaddquqi128_mask ((const __v16qi *) __P,
                  (__v16qi) __W,
                  (__mmask16) __U);
 }
@@ -2355,7 +2355,7 @@
 static __inline__ __m128i __DEFAULT_FN_ATTRS128
 _mm_maskz_loadu_epi8 (__mmask16 __U, void const *__P)
 {
-  return (__m128i) __builtin_ia32_loaddquqi128_mask ((__v16qi *) __P,
+  return (__m128i) __builtin_ia32_loaddquqi128_mask ((const __v16qi *) __P,
                  (__v16qi)
                  _mm_setzero_si128 (),
                  (__mmask16) __U);
@@ -2367,13 +2367,13 @@
   struct __loadu_epi8 {
     __m256i_u __v;
   } __attribute__((__packed__, __may_alias__));
-  return ((struct __loadu_epi8*)__P)->__v;
+  return ((const struct __loadu_epi8*)__P)->__v;
 }
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_mask_loadu_epi8 (__m256i __W, __mmask32 __U, void const *__P)
 {
-  return (__m256i) __builtin_ia32_loaddquqi256_mask ((__v32qi *) __P,
+  return (__m256i) __builtin_ia32_loaddquqi256_mask ((const __v32qi *) __P,
                  (__v32qi) __W,
                  (__mmask32) __U);
 }
@@ -2381,7 +2381,7 @@
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_maskz_loadu_epi8 (__mmask32 __U, void const *__P)
 {
-  return (__m256i) __builtin_ia32_loaddquqi256_mask ((__v32qi *) __P,
+  return (__m256i) __builtin_ia32_loaddquqi256_mask ((const __v32qi *) __P,
                  (__v32qi)
                  _mm256_setzero_si256 (),
                  (__mmask32) __U);
diff --git a/clang/lib/Headers/avx512vlintrin.h b/clang/lib/Headers/avx512vlintrin.h
--- a/clang/lib/Headers/avx512vlintrin.h
+++ b/clang/lib/Headers/avx512vlintrin.h
@@ -2505,7 +2505,7 @@
 
 static __inline__ __m128d __DEFAULT_FN_ATTRS128
 _mm_mask_expandloadu_pd (__m128d __W, __mmask8 __U, void const *__P) {
-  return (__m128d) __builtin_ia32_expandloaddf128_mask ((__v2df *) __P,
+  return (__m128d) __builtin_ia32_expandloaddf128_mask ((const __v2df *) __P,
               (__v2df) __W,
               (__mmask8)
               __U);
@@ -2513,7 +2513,7 @@
 
 static __inline__ __m128d __DEFAULT_FN_ATTRS128
 _mm_maskz_expandloadu_pd (__mmask8 __U, void const *__P) {
-  return (__m128d) __builtin_ia32_expandloaddf128_mask ((__v2df *) __P,
+  return (__m128d) __builtin_ia32_expandloaddf128_mask ((const __v2df *) __P,
                (__v2df)
                _mm_setzero_pd (),
                (__mmask8)
@@ -2522,7 +2522,7 @@
 
 static __inline__ __m256d __DEFAULT_FN_ATTRS256
 _mm256_mask_expandloadu_pd (__m256d __W, __mmask8 __U, void const *__P) {
-  return (__m256d) __builtin_ia32_expandloaddf256_mask ((__v4df *) __P,
+  return (__m256d) __builtin_ia32_expandloaddf256_mask ((const __v4df *) __P,
               (__v4df) __W,
               (__mmask8)
               __U);
@@ -2530,7 +2530,7 @@
 
 static __inline__ __m256d __DEFAULT_FN_ATTRS256
 _mm256_maskz_expandloadu_pd (__mmask8 __U, void const *__P) {
-  return (__m256d) __builtin_ia32_expandloaddf256_mask ((__v4df *) __P,
+  return (__m256d) __builtin_ia32_expandloaddf256_mask ((const __v4df *) __P,
                (__v4df)
                _mm256_setzero_pd (),
                (__mmask8)
@@ -2539,7 +2539,7 @@
 
 static __inline__ __m128i __DEFAULT_FN_ATTRS128
 _mm_mask_expandloadu_epi64 (__m128i __W, __mmask8 __U, void const *__P) {
-  return (__m128i) __builtin_ia32_expandloaddi128_mask ((__v2di *) __P,
+  return (__m128i) __builtin_ia32_expandloaddi128_mask ((const __v2di *) __P,
               (__v2di) __W,
               (__mmask8)
               __U);
@@ -2547,7 +2547,7 @@
 
 static __inline__ __m128i __DEFAULT_FN_ATTRS128
 _mm_maskz_expandloadu_epi64 (__mmask8 __U, void const *__P) {
-  return (__m128i) __builtin_ia32_expandloaddi128_mask ((__v2di *) __P,
+  return (__m128i) __builtin_ia32_expandloaddi128_mask ((const __v2di *) __P,
                (__v2di)
                _mm_setzero_si128 (),
                (__mmask8)
@@ -2557,7 +2557,7 @@
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_mask_expandloadu_epi64 (__m256i __W, __mmask8 __U,
              void const *__P) {
-  return (__m256i) __builtin_ia32_expandloaddi256_mask ((__v4di *) __P,
+  return (__m256i) __builtin_ia32_expandloaddi256_mask ((const __v4di *) __P,
               (__v4di) __W,
               (__mmask8)
               __U);
@@ -2565,7 +2565,7 @@
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_maskz_expandloadu_epi64 (__mmask8 __U, void const *__P) {
-  return (__m256i) __builtin_ia32_expandloaddi256_mask ((__v4di *) __P,
+  return (__m256i) __builtin_ia32_expandloaddi256_mask ((const __v4di *) __P,
                (__v4di)
                _mm256_setzero_si256 (),
                (__mmask8)
@@ -2574,14 +2574,14 @@
 
 static __inline__ __m128 __DEFAULT_FN_ATTRS128
 _mm_mask_expandloadu_ps (__m128 __W, __mmask8 __U, void const *__P) {
-  return (__m128) __builtin_ia32_expandloadsf128_mask ((__v4sf *) __P,
+  return (__m128) __builtin_ia32_expandloadsf128_mask ((const __v4sf *) __P,
                    (__v4sf) __W,
                    (__mmask8) __U);
 }
 
 static __inline__ __m128 __DEFAULT_FN_ATTRS128
 _mm_maskz_expandloadu_ps (__mmask8 __U, void const *__P) {
-  return (__m128) __builtin_ia32_expandloadsf128_mask ((__v4sf *) __P,
+  return (__m128) __builtin_ia32_expandloadsf128_mask ((const __v4sf *) __P,
               (__v4sf)
               _mm_setzero_ps (),
               (__mmask8)
@@ -2590,14 +2590,14 @@
 
 static __inline__ __m256 __DEFAULT_FN_ATTRS256
 _mm256_mask_expandloadu_ps (__m256 __W, __mmask8 __U, void const *__P) {
-  return (__m256) __builtin_ia32_expandloadsf256_mask ((__v8sf *) __P,
+  return (__m256) __builtin_ia32_expandloadsf256_mask ((const __v8sf *) __P,
                    (__v8sf) __W,
                    (__mmask8) __U);
 }
 
 static __inline__ __m256 __DEFAULT_FN_ATTRS256
 _mm256_maskz_expandloadu_ps (__mmask8 __U, void const *__P) {
-  return (__m256) __builtin_ia32_expandloadsf256_mask ((__v8sf *) __P,
+  return (__m256) __builtin_ia32_expandloadsf256_mask ((const __v8sf *) __P,
               (__v8sf)
               _mm256_setzero_ps (),
               (__mmask8)
@@ -2606,7 +2606,7 @@
 
 static __inline__ __m128i __DEFAULT_FN_ATTRS128
 _mm_mask_expandloadu_epi32 (__m128i __W, __mmask8 __U, void const *__P) {
-  return (__m128i) __builtin_ia32_expandloadsi128_mask ((__v4si *) __P,
+  return (__m128i) __builtin_ia32_expandloadsi128_mask ((const __v4si *) __P,
               (__v4si) __W,
               (__mmask8)
               __U);
@@ -2614,7 +2614,7 @@
 
 static __inline__ __m128i __DEFAULT_FN_ATTRS128
 _mm_maskz_expandloadu_epi32 (__mmask8 __U, void const *__P) {
-  return (__m128i) __builtin_ia32_expandloadsi128_mask ((__v4si *) __P,
+  return (__m128i) __builtin_ia32_expandloadsi128_mask ((const __v4si *) __P,
                (__v4si)
                _mm_setzero_si128 (),
                (__mmask8)     __U);
@@ -2623,7 +2623,7 @@
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_mask_expandloadu_epi32 (__m256i __W, __mmask8 __U,
              void const *__P) {
-  return (__m256i) __builtin_ia32_expandloadsi256_mask ((__v8si *) __P,
+  return (__m256i) __builtin_ia32_expandloadsi256_mask ((const __v8si *) __P,
               (__v8si) __W,
               (__mmask8)
               __U);
@@ -2631,7 +2631,7 @@
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_maskz_expandloadu_epi32 (__mmask8 __U, void const *__P) {
-  return (__m256i) __builtin_ia32_expandloadsi256_mask ((__v8si *) __P,
+  return (__m256i) __builtin_ia32_expandloadsi256_mask ((const __v8si *) __P,
                (__v8si)
                _mm256_setzero_si256 (),
                (__mmask8)
@@ -5073,13 +5073,13 @@
 static __inline __m128i __DEFAULT_FN_ATTRS128
 _mm_load_epi32 (void const *__P)
 {
-  return *(__m128i *) __P;
+  return *(const __m128i *) __P;
 }
 
 static __inline__ __m128i __DEFAULT_FN_ATTRS128
 _mm_mask_load_epi32 (__m128i __W, __mmask8 __U, void const *__P)
 {
-  return (__m128i) __builtin_ia32_movdqa32load128_mask ((__v4si *) __P,
+  return (__m128i) __builtin_ia32_movdqa32load128_mask ((const __v4si *) __P,
               (__v4si) __W,
               (__mmask8)
               __U);
@@ -5088,7 +5088,7 @@
 static __inline__ __m128i __DEFAULT_FN_ATTRS128
 _mm_maskz_load_epi32 (__mmask8 __U, void const *__P)
 {
-  return (__m128i) __builtin_ia32_movdqa32load128_mask ((__v4si *) __P,
+  return (__m128i) __builtin_ia32_movdqa32load128_mask ((const __v4si *) __P,
               (__v4si)
               _mm_setzero_si128 (),
               (__mmask8)
@@ -5098,13 +5098,13 @@
 static __inline __m256i __DEFAULT_FN_ATTRS256
 _mm256_load_epi32 (void const *__P)
 {
-  return *(__m256i *) __P;
+  return *(const __m256i *) __P;
 }
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_mask_load_epi32 (__m256i __W, __mmask8 __U, void const *__P)
 {
-  return (__m256i) __builtin_ia32_movdqa32load256_mask ((__v8si *) __P,
+  return (__m256i) __builtin_ia32_movdqa32load256_mask ((const __v8si *) __P,
               (__v8si) __W,
               (__mmask8)
               __U);
@@ -5113,7 +5113,7 @@
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_maskz_load_epi32 (__mmask8 __U, void const *__P)
 {
-  return (__m256i) __builtin_ia32_movdqa32load256_mask ((__v8si *) __P,
+  return (__m256i) __builtin_ia32_movdqa32load256_mask ((const __v8si *) __P,
               (__v8si)
               _mm256_setzero_si256 (),
               (__mmask8)
@@ -5183,13 +5183,13 @@
 static __inline __m128i __DEFAULT_FN_ATTRS128
 _mm_load_epi64 (void const *__P)
 {
-  return *(__m128i *) __P;
+  return *(const __m128i *) __P;
 }
 
 static __inline__ __m128i __DEFAULT_FN_ATTRS128
 _mm_mask_load_epi64 (__m128i __W, __mmask8 __U, void const *__P)
 {
-  return (__m128i) __builtin_ia32_movdqa64load128_mask ((__v2di *) __P,
+  return (__m128i) __builtin_ia32_movdqa64load128_mask ((const __v2di *) __P,
               (__v2di) __W,
               (__mmask8)
               __U);
@@ -5198,7 +5198,7 @@
 static __inline__ __m128i __DEFAULT_FN_ATTRS128
 _mm_maskz_load_epi64 (__mmask8 __U, void const *__P)
 {
-  return (__m128i) __builtin_ia32_movdqa64load128_mask ((__v2di *) __P,
+  return (__m128i) __builtin_ia32_movdqa64load128_mask ((const __v2di *) __P,
               (__v2di)
               _mm_setzero_si128 (),
               (__mmask8)
@@ -5208,13 +5208,13 @@
 static __inline __m256i __DEFAULT_FN_ATTRS256
 _mm256_load_epi64 (void const *__P)
 {
-  return *(__m256i *) __P;
+  return *(const __m256i *) __P;
 }
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_mask_load_epi64 (__m256i __W, __mmask8 __U, void const *__P)
 {
-  return (__m256i) __builtin_ia32_movdqa64load256_mask ((__v4di *) __P,
+  return (__m256i) __builtin_ia32_movdqa64load256_mask ((const __v4di *) __P,
               (__v4di) __W,
               (__mmask8)
               __U);
@@ -5223,7 +5223,7 @@
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_maskz_load_epi64 (__mmask8 __U, void const *__P)
 {
-  return (__m256i) __builtin_ia32_movdqa64load256_mask ((__v4di *) __P,
+  return (__m256i) __builtin_ia32_movdqa64load256_mask ((const __v4di *) __P,
               (__v4di)
               _mm256_setzero_si256 (),
               (__mmask8)
@@ -5430,7 +5430,7 @@
 static __inline__ __m128d __DEFAULT_FN_ATTRS128
 _mm_mask_load_pd (__m128d __W, __mmask8 __U, void const *__P)
 {
-  return (__m128d) __builtin_ia32_loadapd128_mask ((__v2df *) __P,
+  return (__m128d) __builtin_ia32_loadapd128_mask ((const __v2df *) __P,
                (__v2df) __W,
                (__mmask8) __U);
 }
@@ -5438,7 +5438,7 @@
 static __inline__ __m128d __DEFAULT_FN_ATTRS128
 _mm_maskz_load_pd (__mmask8 __U, void const *__P)
 {
-  return (__m128d) __builtin_ia32_loadapd128_mask ((__v2df *) __P,
+  return (__m128d) __builtin_ia32_loadapd128_mask ((const __v2df *) __P,
                (__v2df)
                _mm_setzero_pd (),
                (__mmask8) __U);
@@ -5447,7 +5447,7 @@
 static __inline__ __m256d __DEFAULT_FN_ATTRS256
 _mm256_mask_load_pd (__m256d __W, __mmask8 __U, void const *__P)
 {
-  return (__m256d) __builtin_ia32_loadapd256_mask ((__v4df *) __P,
+  return (__m256d) __builtin_ia32_loadapd256_mask ((const __v4df *) __P,
                (__v4df) __W,
                (__mmask8) __U);
 }
@@ -5455,7 +5455,7 @@
 static __inline__ __m256d __DEFAULT_FN_ATTRS256
 _mm256_maskz_load_pd (__mmask8 __U, void const *__P)
 {
-  return (__m256d) __builtin_ia32_loadapd256_mask ((__v4df *) __P,
+  return (__m256d) __builtin_ia32_loadapd256_mask ((const __v4df *) __P,
                (__v4df)
                _mm256_setzero_pd (),
                (__mmask8) __U);
@@ -5464,7 +5464,7 @@
 static __inline__ __m128 __DEFAULT_FN_ATTRS128
 _mm_mask_load_ps (__m128 __W, __mmask8 __U, void const *__P)
 {
-  return (__m128) __builtin_ia32_loadaps128_mask ((__v4sf *) __P,
+  return (__m128) __builtin_ia32_loadaps128_mask ((const __v4sf *) __P,
               (__v4sf) __W,
               (__mmask8) __U);
 }
@@ -5472,7 +5472,7 @@
 static __inline__ __m128 __DEFAULT_FN_ATTRS128
 _mm_maskz_load_ps (__mmask8 __U, void const *__P)
 {
-  return (__m128) __builtin_ia32_loadaps128_mask ((__v4sf *) __P,
+  return (__m128) __builtin_ia32_loadaps128_mask ((const __v4sf *) __P,
               (__v4sf)
               _mm_setzero_ps (),
               (__mmask8) __U);
@@ -5481,7 +5481,7 @@
 static __inline__ __m256 __DEFAULT_FN_ATTRS256
 _mm256_mask_load_ps (__m256 __W, __mmask8 __U, void const *__P)
 {
-  return (__m256) __builtin_ia32_loadaps256_mask ((__v8sf *) __P,
+  return (__m256) __builtin_ia32_loadaps256_mask ((const __v8sf *) __P,
               (__v8sf) __W,
               (__mmask8) __U);
 }
@@ -5489,7 +5489,7 @@
 static __inline__ __m256 __DEFAULT_FN_ATTRS256
 _mm256_maskz_load_ps (__mmask8 __U, void const *__P)
 {
-  return (__m256) __builtin_ia32_loadaps256_mask ((__v8sf *) __P,
+  return (__m256) __builtin_ia32_loadaps256_mask ((const __v8sf *) __P,
               (__v8sf)
               _mm256_setzero_ps (),
               (__mmask8) __U);
@@ -5501,13 +5501,13 @@
   struct __loadu_epi64 {
     __m128i_u __v;
   } __attribute__((__packed__, __may_alias__));
-  return ((struct __loadu_epi64*)__P)->__v;
+  return ((const struct __loadu_epi64*)__P)->__v;
 }
 
 static __inline__ __m128i __DEFAULT_FN_ATTRS128
 _mm_mask_loadu_epi64 (__m128i __W, __mmask8 __U, void const *__P)
 {
-  return (__m128i) __builtin_ia32_loaddqudi128_mask ((__v2di *) __P,
+  return (__m128i) __builtin_ia32_loaddqudi128_mask ((const __v2di *) __P,
                  (__v2di) __W,
                  (__mmask8) __U);
 }
@@ -5515,7 +5515,7 @@
 static __inline__ __m128i __DEFAULT_FN_ATTRS128
 _mm_maskz_loadu_epi64 (__mmask8 __U, void const *__P)
 {
-  return (__m128i) __builtin_ia32_loaddqudi128_mask ((__v2di *) __P,
+  return (__m128i) __builtin_ia32_loaddqudi128_mask ((const __v2di *) __P,
                  (__v2di)
                  _mm_setzero_si128 (),
                  (__mmask8) __U);
@@ -5527,13 +5527,13 @@
   struct __loadu_epi64 {
     __m256i_u __v;
   } __attribute__((__packed__, __may_alias__));
-  return ((struct __loadu_epi64*)__P)->__v;
+  return ((const struct __loadu_epi64*)__P)->__v;
 }
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_mask_loadu_epi64 (__m256i __W, __mmask8 __U, void const *__P)
 {
-  return (__m256i) __builtin_ia32_loaddqudi256_mask ((__v4di *) __P,
+  return (__m256i) __builtin_ia32_loaddqudi256_mask ((const __v4di *) __P,
                  (__v4di) __W,
                  (__mmask8) __U);
 }
@@ -5541,7 +5541,7 @@
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_maskz_loadu_epi64 (__mmask8 __U, void const *__P)
 {
-  return (__m256i) __builtin_ia32_loaddqudi256_mask ((__v4di *) __P,
+  return (__m256i) __builtin_ia32_loaddqudi256_mask ((const __v4di *) __P,
                  (__v4di)
                  _mm256_setzero_si256 (),
                  (__mmask8) __U);
@@ -5553,13 +5553,13 @@
   struct __loadu_epi32 {
     __m128i_u __v;
   } __attribute__((__packed__, __may_alias__));
-  return ((struct __loadu_epi32*)__P)->__v;
+  return ((const struct __loadu_epi32*)__P)->__v;
 }
 
 static __inline__ __m128i __DEFAULT_FN_ATTRS128
 _mm_mask_loadu_epi32 (__m128i __W, __mmask8 __U, void const *__P)
 {
-  return (__m128i) __builtin_ia32_loaddqusi128_mask ((__v4si *) __P,
+  return (__m128i) __builtin_ia32_loaddqusi128_mask ((const __v4si *) __P,
                  (__v4si) __W,
                  (__mmask8) __U);
 }
@@ -5567,7 +5567,7 @@
 static __inline__ __m128i __DEFAULT_FN_ATTRS128
 _mm_maskz_loadu_epi32 (__mmask8 __U, void const *__P)
 {
-  return (__m128i) __builtin_ia32_loaddqusi128_mask ((__v4si *) __P,
+  return (__m128i) __builtin_ia32_loaddqusi128_mask ((const __v4si *) __P,
                  (__v4si)
                  _mm_setzero_si128 (),
                  (__mmask8) __U);
@@ -5579,13 +5579,13 @@
   struct __loadu_epi32 {
     __m256i_u __v;
   } __attribute__((__packed__, __may_alias__));
-  return ((struct __loadu_epi32*)__P)->__v;
+  return ((const struct __loadu_epi32*)__P)->__v;
 }
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_mask_loadu_epi32 (__m256i __W, __mmask8 __U, void const *__P)
 {
-  return (__m256i) __builtin_ia32_loaddqusi256_mask ((__v8si *) __P,
+  return (__m256i) __builtin_ia32_loaddqusi256_mask ((const __v8si *) __P,
                  (__v8si) __W,
                  (__mmask8) __U);
 }
@@ -5593,7 +5593,7 @@
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_maskz_loadu_epi32 (__mmask8 __U, void const *__P)
 {
-  return (__m256i) __builtin_ia32_loaddqusi256_mask ((__v8si *) __P,
+  return (__m256i) __builtin_ia32_loaddqusi256_mask ((const __v8si *) __P,
                  (__v8si)
                  _mm256_setzero_si256 (),
                  (__mmask8) __U);
@@ -5602,7 +5602,7 @@
 static __inline__ __m128d __DEFAULT_FN_ATTRS128
 _mm_mask_loadu_pd (__m128d __W, __mmask8 __U, void const *__P)
 {
-  return (__m128d) __builtin_ia32_loadupd128_mask ((__v2df *) __P,
+  return (__m128d) __builtin_ia32_loadupd128_mask ((const __v2df *) __P,
                (__v2df) __W,
                (__mmask8) __U);
 }
@@ -5610,7 +5610,7 @@
 static __inline__ __m128d __DEFAULT_FN_ATTRS128
 _mm_maskz_loadu_pd (__mmask8 __U, void const *__P)
 {
-  return (__m128d) __builtin_ia32_loadupd128_mask ((__v2df *) __P,
+  return (__m128d) __builtin_ia32_loadupd128_mask ((const __v2df *) __P,
                (__v2df)
                _mm_setzero_pd (),
                (__mmask8) __U);
@@ -5619,7 +5619,7 @@
 static __inline__ __m256d __DEFAULT_FN_ATTRS256
 _mm256_mask_loadu_pd (__m256d __W, __mmask8 __U, void const *__P)
 {
-  return (__m256d) __builtin_ia32_loadupd256_mask ((__v4df *) __P,
+  return (__m256d) __builtin_ia32_loadupd256_mask ((const __v4df *) __P,
                (__v4df) __W,
                (__mmask8) __U);
 }
@@ -5627,7 +5627,7 @@
 static __inline__ __m256d __DEFAULT_FN_ATTRS256
 _mm256_maskz_loadu_pd (__mmask8 __U, void const *__P)
 {
-  return (__m256d) __builtin_ia32_loadupd256_mask ((__v4df *) __P,
+  return (__m256d) __builtin_ia32_loadupd256_mask ((const __v4df *) __P,
                (__v4df)
                _mm256_setzero_pd (),
                (__mmask8) __U);
@@ -5636,7 +5636,7 @@
 static __inline__ __m128 __DEFAULT_FN_ATTRS128
 _mm_mask_loadu_ps (__m128 __W, __mmask8 __U, void const *__P)
 {
-  return (__m128) __builtin_ia32_loadups128_mask ((__v4sf *) __P,
+  return (__m128) __builtin_ia32_loadups128_mask ((const __v4sf *) __P,
               (__v4sf) __W,
               (__mmask8) __U);
 }
@@ -5644,7 +5644,7 @@
 static __inline__ __m128 __DEFAULT_FN_ATTRS128
 _mm_maskz_loadu_ps (__mmask8 __U, void const *__P)
 {
-  return (__m128) __builtin_ia32_loadups128_mask ((__v4sf *) __P,
+  return (__m128) __builtin_ia32_loadups128_mask ((const __v4sf *) __P,
               (__v4sf)
               _mm_setzero_ps (),
               (__mmask8) __U);
@@ -5653,7 +5653,7 @@
 static __inline__ __m256 __DEFAULT_FN_ATTRS256
 _mm256_mask_loadu_ps (__m256 __W, __mmask8 __U, void const *__P)
 {
-  return (__m256) __builtin_ia32_loadups256_mask ((__v8sf *) __P,
+  return (__m256) __builtin_ia32_loadups256_mask ((const __v8sf *) __P,
               (__v8sf) __W,
               (__mmask8) __U);
 }
@@ -5661,7 +5661,7 @@
 static __inline__ __m256 __DEFAULT_FN_ATTRS256
 _mm256_maskz_loadu_ps (__mmask8 __U, void const *__P)
 {
-  return (__m256) __builtin_ia32_loadups256_mask ((__v8sf *) __P,
+  return (__m256) __builtin_ia32_loadups256_mask ((const __v8sf *) __P,
               (__v8sf)
               _mm256_setzero_ps (),
               (__mmask8) __U);
diff --git a/clang/lib/Headers/avxintrin.h b/clang/lib/Headers/avxintrin.h
--- a/clang/lib/Headers/avxintrin.h
+++ b/clang/lib/Headers/avxintrin.h
@@ -3069,7 +3069,7 @@
 static __inline __m256d __DEFAULT_FN_ATTRS
 _mm256_load_pd(double const *__p)
 {
-  return *(__m256d *)__p;
+  return *(const __m256d *)__p;
 }
 
 /// Loads 8 single-precision floating point values from a 32-byte aligned
@@ -3085,7 +3085,7 @@
 static __inline __m256 __DEFAULT_FN_ATTRS
 _mm256_load_ps(float const *__p)
 {
-  return *(__m256 *)__p;
+  return *(const __m256 *)__p;
 }
 
 /// Loads 4 double-precision floating point values from an unaligned
@@ -3105,7 +3105,7 @@
   struct __loadu_pd {
     __m256d_u __v;
   } __attribute__((__packed__, __may_alias__));
-  return ((struct __loadu_pd*)__p)->__v;
+  return ((const struct __loadu_pd*)__p)->__v;
 }
 
 /// Loads 8 single-precision floating point values from an unaligned
@@ -3125,7 +3125,7 @@
   struct __loadu_ps {
     __m256_u __v;
   } __attribute__((__packed__, __may_alias__));
-  return ((struct __loadu_ps*)__p)->__v;
+  return ((const struct __loadu_ps*)__p)->__v;
 }
 
 /// Loads 256 bits of integer data from a 32-byte aligned memory
@@ -3161,7 +3161,7 @@
   struct __loadu_si256 {
     __m256i_u __v;
   } __attribute__((__packed__, __may_alias__));
-  return ((struct __loadu_si256*)__p)->__v;
+  return ((const struct __loadu_si256*)__p)->__v;
 }
 
 /// Loads 256 bits of integer data from an unaligned memory location
diff --git a/clang/lib/Headers/emmintrin.h b/clang/lib/Headers/emmintrin.h
--- a/clang/lib/Headers/emmintrin.h
+++ b/clang/lib/Headers/emmintrin.h
@@ -1578,7 +1578,7 @@
 static __inline__ __m128d __DEFAULT_FN_ATTRS
 _mm_load_pd(double const *__dp)
 {
-  return *(__m128d*)__dp;
+  return *(const __m128d*)__dp;
 }
 
 /// Loads a double-precision floating-point value from a specified memory
@@ -1599,7 +1599,7 @@
   struct __mm_load1_pd_struct {
     double __u;
   } __attribute__((__packed__, __may_alias__));
-  double __u = ((struct __mm_load1_pd_struct*)__dp)->__u;
+  double __u = ((const struct __mm_load1_pd_struct*)__dp)->__u;
   return __extension__ (__m128d){ __u, __u };
 }
 
@@ -1622,7 +1622,7 @@
 static __inline__ __m128d __DEFAULT_FN_ATTRS
 _mm_loadr_pd(double const *__dp)
 {
-  __m128d __u = *(__m128d*)__dp;
+  __m128d __u = *(const __m128d*)__dp;
   return __builtin_shufflevector((__v2df)__u, (__v2df)__u, 1, 0);
 }
 
@@ -1643,7 +1643,7 @@
   struct __loadu_pd {
     __m128d_u __v;
   } __attribute__((__packed__, __may_alias__));
-  return ((struct __loadu_pd*)__dp)->__v;
+  return ((const struct __loadu_pd*)__dp)->__v;
 }
 
 /// Loads a 64-bit integer value to the low element of a 128-bit integer
@@ -1663,7 +1663,7 @@
   struct __loadu_si64 {
     long long __v;
   } __attribute__((__packed__, __may_alias__));
-  long long __u = ((struct __loadu_si64*)__a)->__v;
+  long long __u = ((const struct __loadu_si64*)__a)->__v;
   return __extension__ (__m128i)(__v2di){__u, 0LL};
 }
 
@@ -1684,7 +1684,7 @@
   struct __loadu_si32 {
     int __v;
   } __attribute__((__packed__, __may_alias__));
-  int __u = ((struct __loadu_si32*)__a)->__v;
+  int __u = ((const struct __loadu_si32*)__a)->__v;
   return __extension__ (__m128i)(__v4si){__u, 0, 0, 0};
 }
 
@@ -1705,7 +1705,7 @@
   struct __loadu_si16 {
     short __v;
   } __attribute__((__packed__, __may_alias__));
-  short __u = ((struct __loadu_si16*)__a)->__v;
+  short __u = ((const struct __loadu_si16*)__a)->__v;
   return __extension__ (__m128i)(__v8hi){__u, 0, 0, 0, 0, 0, 0, 0};
 }
 
@@ -1726,7 +1726,7 @@
   struct __mm_load_sd_struct {
     double __u;
   } __attribute__((__packed__, __may_alias__));
-  double __u = ((struct __mm_load_sd_struct*)__dp)->__u;
+  double __u = ((const struct __mm_load_sd_struct*)__dp)->__u;
   return __extension__ (__m128d){ __u, 0 };
 }
 
@@ -1753,7 +1753,7 @@
   struct __mm_loadh_pd_struct {
     double __u;
   } __attribute__((__packed__, __may_alias__));
-  double __u = ((struct __mm_loadh_pd_struct*)__dp)->__u;
+  double __u = ((const struct __mm_loadh_pd_struct*)__dp)->__u;
   return __extension__ (__m128d){ __a[0], __u };
 }
 
@@ -1780,7 +1780,7 @@
   struct __mm_loadl_pd_struct {
     double __u;
   } __attribute__((__packed__, __may_alias__));
-  double __u = ((struct __mm_loadl_pd_struct*)__dp)->__u;
+  double __u = ((const struct __mm_loadl_pd_struct*)__dp)->__u;
   return __extension__ (__m128d){ __u, __a[1] };
 }
 
@@ -3550,7 +3550,7 @@
   struct __loadu_si128 {
     __m128i_u __v;
   } __attribute__((__packed__, __may_alias__));
-  return ((struct __loadu_si128*)__p)->__v;
+  return ((const struct __loadu_si128*)__p)->__v;
 }
 
 /// Returns a vector of [2 x i64] where the lower element is taken from
@@ -3571,7 +3571,7 @@
   struct __mm_loadl_epi64_struct {
     long long __u;
   } __attribute__((__packed__, __may_alias__));
-  return __extension__ (__m128i) { ((struct __mm_loadl_epi64_struct*)__p)->__u, 0};
+  return __extension__ (__m128i) { ((const struct __mm_loadl_epi64_struct*)__p)->__u, 0};
 }
 
 /// Generates a 128-bit vector of [4 x i32] with unspecified content.
diff --git a/clang/lib/Headers/immintrin.h b/clang/lib/Headers/immintrin.h
--- a/clang/lib/Headers/immintrin.h
+++ b/clang/lib/Headers/immintrin.h
@@ -301,7 +301,7 @@
   struct __loadu_i16 {
     short __v;
   } __attribute__((__packed__, __may_alias__));
-  return __builtin_bswap16(((struct __loadu_i16*)__P)->__v);
+  return __builtin_bswap16(((const struct __loadu_i16*)__P)->__v);
 }
 
 static __inline__ void __attribute__((__always_inline__, __nodebug__, __target__("movbe")))
@@ -317,7 +317,7 @@
   struct __loadu_i32 {
     int __v;
   } __attribute__((__packed__, __may_alias__));
-  return __builtin_bswap32(((struct __loadu_i32*)__P)->__v);
+  return __builtin_bswap32(((const struct __loadu_i32*)__P)->__v);
 }
 
 static __inline__ void __attribute__((__always_inline__, __nodebug__, __target__("movbe")))
@@ -334,7 +334,7 @@
   struct __loadu_i64 {
     long long __v;
   } __attribute__((__packed__, __may_alias__));
-  return __builtin_bswap64(((struct __loadu_i64*)__P)->__v);
+  return __builtin_bswap64(((const struct __loadu_i64*)__P)->__v);
 }
 
 static __inline__ void __attribute__((__always_inline__, __nodebug__, __target__("movbe")))
diff --git a/clang/lib/Headers/mwaitxintrin.h b/clang/lib/Headers/mwaitxintrin.h
--- a/clang/lib/Headers/mwaitxintrin.h
+++ b/clang/lib/Headers/mwaitxintrin.h
@@ -17,9 +17,9 @@
 /* Define the default attributes for the functions in this file. */
 #define __DEFAULT_FN_ATTRS __attribute__((__always_inline__, __nodebug__,  __target__("mwaitx")))
 static __inline__ void __DEFAULT_FN_ATTRS
-_mm_monitorx(void const * __p, unsigned __extensions, unsigned __hints)
+_mm_monitorx(void * __p, unsigned __extensions, unsigned __hints)
 {
-  __builtin_ia32_monitorx((void *)__p, __extensions, __hints);
+  __builtin_ia32_monitorx(__p, __extensions, __hints);
 }
 
 static __inline__ void __DEFAULT_FN_ATTRS
diff --git a/clang/lib/Headers/pmmintrin.h b/clang/lib/Headers/pmmintrin.h
--- a/clang/lib/Headers/pmmintrin.h
+++ b/clang/lib/Headers/pmmintrin.h
@@ -263,7 +263,7 @@
 static __inline__ void __DEFAULT_FN_ATTRS
 _mm_monitor(void const *__p, unsigned __extensions, unsigned __hints)
 {
-  __builtin_ia32_monitor((void *)__p, __extensions, __hints);
+  __builtin_ia32_monitor(__p, __extensions, __hints);
 }
 
 /// Used with the MONITOR instruction to wait while the processor is in
diff --git a/clang/lib/Headers/xmmintrin.h b/clang/lib/Headers/xmmintrin.h
--- a/clang/lib/Headers/xmmintrin.h
+++ b/clang/lib/Headers/xmmintrin.h
@@ -1627,7 +1627,7 @@
   struct __mm_loadh_pi_struct {
     __mm_loadh_pi_v2f32 __u;
   } __attribute__((__packed__, __may_alias__));
-  __mm_loadh_pi_v2f32 __b = ((struct __mm_loadh_pi_struct*)__p)->__u;
+  __mm_loadh_pi_v2f32 __b = ((const struct __mm_loadh_pi_struct*)__p)->__u;
   __m128 __bb = __builtin_shufflevector(__b, __b, 0, 1, 0, 1);
   return __builtin_shufflevector(__a, __bb, 0, 1, 4, 5);
 }
@@ -1654,7 +1654,7 @@
   struct __mm_loadl_pi_struct {
     __mm_loadl_pi_v2f32 __u;
   } __attribute__((__packed__, __may_alias__));
-  __mm_loadl_pi_v2f32 __b = ((struct __mm_loadl_pi_struct*)__p)->__u;
+  __mm_loadl_pi_v2f32 __b = ((const struct __mm_loadl_pi_struct*)__p)->__u;
   __m128 __bb = __builtin_shufflevector(__b, __b, 0, 1, 0, 1);
   return __builtin_shufflevector(__a, __bb, 4, 5, 2, 3);
 }
@@ -1680,7 +1680,7 @@
   struct __mm_load_ss_struct {
     float __u;
   } __attribute__((__packed__, __may_alias__));
-  float __u = ((struct __mm_load_ss_struct*)__p)->__u;
+  float __u = ((const struct __mm_load_ss_struct*)__p)->__u;
   return __extension__ (__m128){ __u, 0, 0, 0 };
 }
 
@@ -1702,7 +1702,7 @@
   struct __mm_load1_ps_struct {
     float __u;
   } __attribute__((__packed__, __may_alias__));
-  float __u = ((struct __mm_load1_ps_struct*)__p)->__u;
+  float __u = ((const struct __mm_load1_ps_struct*)__p)->__u;
   return __extension__ (__m128){ __u, __u, __u, __u };
 }
 
@@ -1722,7 +1722,7 @@
 static __inline__ __m128 __DEFAULT_FN_ATTRS
 _mm_load_ps(const float *__p)
 {
-  return *(__m128*)__p;
+  return *(const __m128*)__p;
 }
 
 /// Loads a 128-bit floating-point vector of [4 x float] from an
@@ -1742,7 +1742,7 @@
   struct __loadu_ps {
     __m128_u __v;
   } __attribute__((__packed__, __may_alias__));
-  return ((struct __loadu_ps*)__p)->__v;
+  return ((const struct __loadu_ps*)__p)->__v;
 }
 
 /// Loads four packed float values, in reverse order, from an aligned
@@ -2100,7 +2100,7 @@
 ///    be generated. \n
 ///    _MM_HINT_T2: Move data using the T2 hint. The PREFETCHT2 instruction will
 ///    be generated.
-#define _mm_prefetch(a, sel) (__builtin_prefetch((void *)(a), \
+#define _mm_prefetch(a, sel) (__builtin_prefetch((const void *)(a), \
                                                  ((sel) >> 2) & 1, (sel) & 0x3))
 #endif
 
diff --git a/clang/test/Headers/x86-intrinsics-headers-clean.cpp b/clang/test/Headers/x86-intrinsics-headers-clean.cpp
--- a/clang/test/Headers/x86-intrinsics-headers-clean.cpp
+++ b/clang/test/Headers/x86-intrinsics-headers-clean.cpp
@@ -1,7 +1,7 @@
 // Make sure the intrinsic headers compile cleanly with no warnings or errors.
 
 // RUN: %clang_cc1 -ffreestanding -triple x86_64-unknown-unknown -Wsystem-headers \
-// RUN:   -fsyntax-only -flax-vector-conversions=none -x c++ -verify %s
+// RUN:   -Wcast-qual -fsyntax-only -flax-vector-conversions=none -x c++ -verify %s
 
 // expected-no-diagnostics
 
diff --git a/clang/test/Headers/x86intrin-2.c b/clang/test/Headers/x86intrin-2.c
--- a/clang/test/Headers/x86intrin-2.c
+++ b/clang/test/Headers/x86intrin-2.c
@@ -1,6 +1,6 @@
-// RUN: %clang_cc1 -fsyntax-only -ffreestanding %s -verify
-// RUN: %clang_cc1 -fsyntax-only -ffreestanding -flax-vector-conversions=none %s -verify
-// RUN: %clang_cc1 -fsyntax-only -ffreestanding -x c++ %s -verify
+// RUN: %clang_cc1 -fsyntax-only -ffreestanding -Wcast-qual %s -verify
+// RUN: %clang_cc1 -fsyntax-only -ffreestanding -flax-vector-conversions=none -Wcast-qual %s -verify
+// RUN: %clang_cc1 -fsyntax-only -ffreestanding -Wcast-qual -x c++ %s -verify
 // expected-no-diagnostics
 
 #if defined(i386) || defined(__x86_64__)
@@ -16,6 +16,10 @@
   return _mm_add_ss(a, b);
 }
 
+void __attribute__((__target__("sse"))) mm_prefetch_wrap(const void *p) {
+  _mm_prefetch(p, 0x3);
+}
+
 __m128d __attribute__((__target__("sse2"))) mm_sqrt_sd_wrap(__m128d a, __m128d b) {
   return _mm_sqrt_sd(a, b);
 }