gcc: x86 Built-in Functions

1 
1 6.59.33 x86 Built-in Functions
1 ------------------------------
1 
1 These built-in functions are available for the x86-32 and x86-64 family
1 of computers, depending on the command-line switches used.
1 
1  If you specify command-line switches such as '-msse', the compiler
1 could use the extended instruction sets even if the built-ins are not
1 used explicitly in the program.  For this reason, applications that
1 perform run-time CPU detection must compile separate files for each
1 supported architecture, using the appropriate flags.  In particular, the
1 file containing the CPU detection code should be compiled without these
1 options.
1 
1  The following machine modes are available for use with MMX built-in
1 functions (⇒Vector Extensions): 'V2SI' for a vector of two 32-bit
1 integers, 'V4HI' for a vector of four 16-bit integers, and 'V8QI' for a
1 vector of eight 8-bit integers.  Some of the built-in functions operate
1 on MMX registers as a whole 64-bit entity, these use 'V1DI' as their
1 mode.
1 
1  If 3DNow! extensions are enabled, 'V2SF' is used as a mode for a vector
1 of two 32-bit floating-point values.
1 
1  If SSE extensions are enabled, 'V4SF' is used for a vector of four
1 32-bit floating-point values.  Some instructions use a vector of four
1 32-bit integers, these use 'V4SI'.  Finally, some instructions operate
1 on an entire vector register, interpreting it as a 128-bit integer,
1 these use mode 'TI'.
1 
1  The x86-32 and x86-64 family of processors use additional built-in
1 functions for efficient use of 'TF' ('__float128') 128-bit floating
1 point and 'TC' 128-bit complex floating-point values.
1 
1  The following floating-point built-in functions are always available.
1 All of them implement the function that is part of the name.
1 
1      __float128 __builtin_fabsq (__float128)
1      __float128 __builtin_copysignq (__float128, __float128)
1 
1  The following built-in functions are always available.
1 
1 '__float128 __builtin_infq (void)'
1      Similar to '__builtin_inf', except the return type is '__float128'.
1 
1 '__float128 __builtin_huge_valq (void)'
1      Similar to '__builtin_huge_val', except the return type is
1      '__float128'.
1 
1 '__float128 __builtin_nanq (void)'
1      Similar to '__builtin_nan', except the return type is '__float128'.
1 
1 '__float128 __builtin_nansq (void)'
1      Similar to '__builtin_nans', except the return type is
1      '__float128'.
1 
1  The following built-in function is always available.
1 
1 'void __builtin_ia32_pause (void)'
1      Generates the 'pause' machine instruction with a compiler memory
1      barrier.
1 
1  The following built-in functions are always available and can be used
1 to check the target platform type.
1 
1  -- Built-in Function: void __builtin_cpu_init (void)
1      This function runs the CPU detection code to check the type of CPU
1      and the features supported.  This built-in function needs to be
1      invoked along with the built-in functions to check CPU type and
1      features, '__builtin_cpu_is' and '__builtin_cpu_supports', only
1      when used in a function that is executed before any constructors
1      are called.  The CPU detection code is automatically executed in a
1      very high priority constructor.
1 
1      For example, this function has to be used in 'ifunc' resolvers that
1      check for CPU type using the built-in functions '__builtin_cpu_is'
1      and '__builtin_cpu_supports', or in constructors on targets that
1      don't support constructor priority.
1 
1           static void (*resolve_memcpy (void)) (void)
1           {
1             // ifunc resolvers fire before constructors, explicitly call the init
1             // function.
1             __builtin_cpu_init ();
1             if (__builtin_cpu_supports ("ssse3"))
1               return ssse3_memcpy; // super fast memcpy with ssse3 instructions.
1             else
1               return default_memcpy;
1           }
1 
1           void *memcpy (void *, const void *, size_t)
1                __attribute__ ((ifunc ("resolve_memcpy")));
1 
1  -- Built-in Function: int __builtin_cpu_is (const char *CPUNAME)
1      This function returns a positive integer if the run-time CPU is of
1      type CPUNAME and returns '0' otherwise.  The following CPU names
1      can be detected:
1 
1      'intel'
1           Intel CPU.
1 
1      'atom'
1           Intel Atom CPU.
1 
1      'core2'
1           Intel Core 2 CPU.
1 
1      'corei7'
1           Intel Core i7 CPU.
1 
1      'nehalem'
1           Intel Core i7 Nehalem CPU.
1 
1      'westmere'
1           Intel Core i7 Westmere CPU.
1 
1      'sandybridge'
1           Intel Core i7 Sandy Bridge CPU.
1 
1      'amd'
1           AMD CPU.
1 
1      'amdfam10h'
1           AMD Family 10h CPU.
1 
1      'barcelona'
1           AMD Family 10h Barcelona CPU.
1 
1      'shanghai'
1           AMD Family 10h Shanghai CPU.
1 
1      'istanbul'
1           AMD Family 10h Istanbul CPU.
1 
1      'btver1'
1           AMD Family 14h CPU.
1 
1      'amdfam15h'
1           AMD Family 15h CPU.
1 
1      'bdver1'
1           AMD Family 15h Bulldozer version 1.
1 
1      'bdver2'
1           AMD Family 15h Bulldozer version 2.
1 
1      'bdver3'
1           AMD Family 15h Bulldozer version 3.
1 
1      'bdver4'
1           AMD Family 15h Bulldozer version 4.
1 
1      'btver2'
1           AMD Family 16h CPU.
1 
1      'amdfam17h'
1           AMD Family 17h CPU.
1 
1      'znver1'
1           AMD Family 17h Zen version 1.
1 
1      Here is an example:
1           if (__builtin_cpu_is ("corei7"))
1             {
1                do_corei7 (); // Core i7 specific implementation.
1             }
1           else
1             {
1                do_generic (); // Generic implementation.
1             }
1 
1  -- Built-in Function: int __builtin_cpu_supports (const char *FEATURE)
1      This function returns a positive integer if the run-time CPU
1      supports FEATURE and returns '0' otherwise.  The following features
1      can be detected:
1 
1      'cmov'
1           CMOV instruction.
1      'mmx'
1           MMX instructions.
1      'popcnt'
1           POPCNT instruction.
1      'sse'
1           SSE instructions.
1      'sse2'
1           SSE2 instructions.
1      'sse3'
1           SSE3 instructions.
1      'ssse3'
1           SSSE3 instructions.
1      'sse4.1'
1           SSE4.1 instructions.
1      'sse4.2'
1           SSE4.2 instructions.
1      'avx'
1           AVX instructions.
1      'avx2'
1           AVX2 instructions.
1      'avx512f'
1           AVX512F instructions.
1 
1      Here is an example:
1           if (__builtin_cpu_supports ("popcnt"))
1             {
1                asm("popcnt %1,%0" : "=r"(count) : "rm"(n) : "cc");
1             }
1           else
1             {
1                count = generic_countbits (n); //generic implementation.
1             }
1 
1  The following built-in functions are made available by '-mmmx'.  All of
1 them generate the machine instruction that is part of the name.
1 
1      v8qi __builtin_ia32_paddb (v8qi, v8qi)
1      v4hi __builtin_ia32_paddw (v4hi, v4hi)
1      v2si __builtin_ia32_paddd (v2si, v2si)
1      v8qi __builtin_ia32_psubb (v8qi, v8qi)
1      v4hi __builtin_ia32_psubw (v4hi, v4hi)
1      v2si __builtin_ia32_psubd (v2si, v2si)
1      v8qi __builtin_ia32_paddsb (v8qi, v8qi)
1      v4hi __builtin_ia32_paddsw (v4hi, v4hi)
1      v8qi __builtin_ia32_psubsb (v8qi, v8qi)
1      v4hi __builtin_ia32_psubsw (v4hi, v4hi)
1      v8qi __builtin_ia32_paddusb (v8qi, v8qi)
1      v4hi __builtin_ia32_paddusw (v4hi, v4hi)
1      v8qi __builtin_ia32_psubusb (v8qi, v8qi)
1      v4hi __builtin_ia32_psubusw (v4hi, v4hi)
1      v4hi __builtin_ia32_pmullw (v4hi, v4hi)
1      v4hi __builtin_ia32_pmulhw (v4hi, v4hi)
1      di __builtin_ia32_pand (di, di)
1      di __builtin_ia32_pandn (di,di)
1      di __builtin_ia32_por (di, di)
1      di __builtin_ia32_pxor (di, di)
1      v8qi __builtin_ia32_pcmpeqb (v8qi, v8qi)
1      v4hi __builtin_ia32_pcmpeqw (v4hi, v4hi)
1      v2si __builtin_ia32_pcmpeqd (v2si, v2si)
1      v8qi __builtin_ia32_pcmpgtb (v8qi, v8qi)
1      v4hi __builtin_ia32_pcmpgtw (v4hi, v4hi)
1      v2si __builtin_ia32_pcmpgtd (v2si, v2si)
1      v8qi __builtin_ia32_punpckhbw (v8qi, v8qi)
1      v4hi __builtin_ia32_punpckhwd (v4hi, v4hi)
1      v2si __builtin_ia32_punpckhdq (v2si, v2si)
1      v8qi __builtin_ia32_punpcklbw (v8qi, v8qi)
1      v4hi __builtin_ia32_punpcklwd (v4hi, v4hi)
1      v2si __builtin_ia32_punpckldq (v2si, v2si)
1      v8qi __builtin_ia32_packsswb (v4hi, v4hi)
1      v4hi __builtin_ia32_packssdw (v2si, v2si)
1      v8qi __builtin_ia32_packuswb (v4hi, v4hi)
1 
1      v4hi __builtin_ia32_psllw (v4hi, v4hi)
1      v2si __builtin_ia32_pslld (v2si, v2si)
1      v1di __builtin_ia32_psllq (v1di, v1di)
1      v4hi __builtin_ia32_psrlw (v4hi, v4hi)
1      v2si __builtin_ia32_psrld (v2si, v2si)
1      v1di __builtin_ia32_psrlq (v1di, v1di)
1      v4hi __builtin_ia32_psraw (v4hi, v4hi)
1      v2si __builtin_ia32_psrad (v2si, v2si)
1      v4hi __builtin_ia32_psllwi (v4hi, int)
1      v2si __builtin_ia32_pslldi (v2si, int)
1      v1di __builtin_ia32_psllqi (v1di, int)
1      v4hi __builtin_ia32_psrlwi (v4hi, int)
1      v2si __builtin_ia32_psrldi (v2si, int)
1      v1di __builtin_ia32_psrlqi (v1di, int)
1      v4hi __builtin_ia32_psrawi (v4hi, int)
1      v2si __builtin_ia32_psradi (v2si, int)
1 
1 
1  The following built-in functions are made available either with
1 '-msse', or with '-m3dnowa'.  All of them generate the machine
1 instruction that is part of the name.
1 
1      v4hi __builtin_ia32_pmulhuw (v4hi, v4hi)
1      v8qi __builtin_ia32_pavgb (v8qi, v8qi)
1      v4hi __builtin_ia32_pavgw (v4hi, v4hi)
1      v1di __builtin_ia32_psadbw (v8qi, v8qi)
1      v8qi __builtin_ia32_pmaxub (v8qi, v8qi)
1      v4hi __builtin_ia32_pmaxsw (v4hi, v4hi)
1      v8qi __builtin_ia32_pminub (v8qi, v8qi)
1      v4hi __builtin_ia32_pminsw (v4hi, v4hi)
1      int __builtin_ia32_pmovmskb (v8qi)
1      void __builtin_ia32_maskmovq (v8qi, v8qi, char *)
1      void __builtin_ia32_movntq (di *, di)
1      void __builtin_ia32_sfence (void)
1 
1  The following built-in functions are available when '-msse' is used.
1 All of them generate the machine instruction that is part of the name.
1 
1      int __builtin_ia32_comieq (v4sf, v4sf)
1      int __builtin_ia32_comineq (v4sf, v4sf)
1      int __builtin_ia32_comilt (v4sf, v4sf)
1      int __builtin_ia32_comile (v4sf, v4sf)
1      int __builtin_ia32_comigt (v4sf, v4sf)
1      int __builtin_ia32_comige (v4sf, v4sf)
1      int __builtin_ia32_ucomieq (v4sf, v4sf)
1      int __builtin_ia32_ucomineq (v4sf, v4sf)
1      int __builtin_ia32_ucomilt (v4sf, v4sf)
1      int __builtin_ia32_ucomile (v4sf, v4sf)
1      int __builtin_ia32_ucomigt (v4sf, v4sf)
1      int __builtin_ia32_ucomige (v4sf, v4sf)
1      v4sf __builtin_ia32_addps (v4sf, v4sf)
1      v4sf __builtin_ia32_subps (v4sf, v4sf)
1      v4sf __builtin_ia32_mulps (v4sf, v4sf)
1      v4sf __builtin_ia32_divps (v4sf, v4sf)
1      v4sf __builtin_ia32_addss (v4sf, v4sf)
1      v4sf __builtin_ia32_subss (v4sf, v4sf)
1      v4sf __builtin_ia32_mulss (v4sf, v4sf)
1      v4sf __builtin_ia32_divss (v4sf, v4sf)
1      v4sf __builtin_ia32_cmpeqps (v4sf, v4sf)
1      v4sf __builtin_ia32_cmpltps (v4sf, v4sf)
1      v4sf __builtin_ia32_cmpleps (v4sf, v4sf)
1      v4sf __builtin_ia32_cmpgtps (v4sf, v4sf)
1      v4sf __builtin_ia32_cmpgeps (v4sf, v4sf)
1      v4sf __builtin_ia32_cmpunordps (v4sf, v4sf)
1      v4sf __builtin_ia32_cmpneqps (v4sf, v4sf)
1      v4sf __builtin_ia32_cmpnltps (v4sf, v4sf)
1      v4sf __builtin_ia32_cmpnleps (v4sf, v4sf)
1      v4sf __builtin_ia32_cmpngtps (v4sf, v4sf)
1      v4sf __builtin_ia32_cmpngeps (v4sf, v4sf)
1      v4sf __builtin_ia32_cmpordps (v4sf, v4sf)
1      v4sf __builtin_ia32_cmpeqss (v4sf, v4sf)
1      v4sf __builtin_ia32_cmpltss (v4sf, v4sf)
1      v4sf __builtin_ia32_cmpless (v4sf, v4sf)
1      v4sf __builtin_ia32_cmpunordss (v4sf, v4sf)
1      v4sf __builtin_ia32_cmpneqss (v4sf, v4sf)
1      v4sf __builtin_ia32_cmpnltss (v4sf, v4sf)
1      v4sf __builtin_ia32_cmpnless (v4sf, v4sf)
1      v4sf __builtin_ia32_cmpordss (v4sf, v4sf)
1      v4sf __builtin_ia32_maxps (v4sf, v4sf)
1      v4sf __builtin_ia32_maxss (v4sf, v4sf)
1      v4sf __builtin_ia32_minps (v4sf, v4sf)
1      v4sf __builtin_ia32_minss (v4sf, v4sf)
1      v4sf __builtin_ia32_andps (v4sf, v4sf)
1      v4sf __builtin_ia32_andnps (v4sf, v4sf)
1      v4sf __builtin_ia32_orps (v4sf, v4sf)
1      v4sf __builtin_ia32_xorps (v4sf, v4sf)
1      v4sf __builtin_ia32_movss (v4sf, v4sf)
1      v4sf __builtin_ia32_movhlps (v4sf, v4sf)
1      v4sf __builtin_ia32_movlhps (v4sf, v4sf)
1      v4sf __builtin_ia32_unpckhps (v4sf, v4sf)
1      v4sf __builtin_ia32_unpcklps (v4sf, v4sf)
1      v4sf __builtin_ia32_cvtpi2ps (v4sf, v2si)
1      v4sf __builtin_ia32_cvtsi2ss (v4sf, int)
1      v2si __builtin_ia32_cvtps2pi (v4sf)
1      int __builtin_ia32_cvtss2si (v4sf)
1      v2si __builtin_ia32_cvttps2pi (v4sf)
1      int __builtin_ia32_cvttss2si (v4sf)
1      v4sf __builtin_ia32_rcpps (v4sf)
1      v4sf __builtin_ia32_rsqrtps (v4sf)
1      v4sf __builtin_ia32_sqrtps (v4sf)
1      v4sf __builtin_ia32_rcpss (v4sf)
1      v4sf __builtin_ia32_rsqrtss (v4sf)
1      v4sf __builtin_ia32_sqrtss (v4sf)
1      v4sf __builtin_ia32_shufps (v4sf, v4sf, int)
1      void __builtin_ia32_movntps (float *, v4sf)
1      int __builtin_ia32_movmskps (v4sf)
1 
1  The following built-in functions are available when '-msse' is used.
1 
1 'v4sf __builtin_ia32_loadups (float *)'
1      Generates the 'movups' machine instruction as a load from memory.
1 'void __builtin_ia32_storeups (float *, v4sf)'
1      Generates the 'movups' machine instruction as a store to memory.
1 'v4sf __builtin_ia32_loadss (float *)'
1      Generates the 'movss' machine instruction as a load from memory.
1 'v4sf __builtin_ia32_loadhps (v4sf, const v2sf *)'
1      Generates the 'movhps' machine instruction as a load from memory.
1 'v4sf __builtin_ia32_loadlps (v4sf, const v2sf *)'
1      Generates the 'movlps' machine instruction as a load from memory
1 'void __builtin_ia32_storehps (v2sf *, v4sf)'
1      Generates the 'movhps' machine instruction as a store to memory.
1 'void __builtin_ia32_storelps (v2sf *, v4sf)'
1      Generates the 'movlps' machine instruction as a store to memory.
1 
1  The following built-in functions are available when '-msse2' is used.
1 All of them generate the machine instruction that is part of the name.
1 
1      int __builtin_ia32_comisdeq (v2df, v2df)
1      int __builtin_ia32_comisdlt (v2df, v2df)
1      int __builtin_ia32_comisdle (v2df, v2df)
1      int __builtin_ia32_comisdgt (v2df, v2df)
1      int __builtin_ia32_comisdge (v2df, v2df)
1      int __builtin_ia32_comisdneq (v2df, v2df)
1      int __builtin_ia32_ucomisdeq (v2df, v2df)
1      int __builtin_ia32_ucomisdlt (v2df, v2df)
1      int __builtin_ia32_ucomisdle (v2df, v2df)
1      int __builtin_ia32_ucomisdgt (v2df, v2df)
1      int __builtin_ia32_ucomisdge (v2df, v2df)
1      int __builtin_ia32_ucomisdneq (v2df, v2df)
1      v2df __builtin_ia32_cmpeqpd (v2df, v2df)
1      v2df __builtin_ia32_cmpltpd (v2df, v2df)
1      v2df __builtin_ia32_cmplepd (v2df, v2df)
1      v2df __builtin_ia32_cmpgtpd (v2df, v2df)
1      v2df __builtin_ia32_cmpgepd (v2df, v2df)
1      v2df __builtin_ia32_cmpunordpd (v2df, v2df)
1      v2df __builtin_ia32_cmpneqpd (v2df, v2df)
1      v2df __builtin_ia32_cmpnltpd (v2df, v2df)
1      v2df __builtin_ia32_cmpnlepd (v2df, v2df)
1      v2df __builtin_ia32_cmpngtpd (v2df, v2df)
1      v2df __builtin_ia32_cmpngepd (v2df, v2df)
1      v2df __builtin_ia32_cmpordpd (v2df, v2df)
1      v2df __builtin_ia32_cmpeqsd (v2df, v2df)
1      v2df __builtin_ia32_cmpltsd (v2df, v2df)
1      v2df __builtin_ia32_cmplesd (v2df, v2df)
1      v2df __builtin_ia32_cmpunordsd (v2df, v2df)
1      v2df __builtin_ia32_cmpneqsd (v2df, v2df)
1      v2df __builtin_ia32_cmpnltsd (v2df, v2df)
1      v2df __builtin_ia32_cmpnlesd (v2df, v2df)
1      v2df __builtin_ia32_cmpordsd (v2df, v2df)
1      v2di __builtin_ia32_paddq (v2di, v2di)
1      v2di __builtin_ia32_psubq (v2di, v2di)
1      v2df __builtin_ia32_addpd (v2df, v2df)
1      v2df __builtin_ia32_subpd (v2df, v2df)
1      v2df __builtin_ia32_mulpd (v2df, v2df)
1      v2df __builtin_ia32_divpd (v2df, v2df)
1      v2df __builtin_ia32_addsd (v2df, v2df)
1      v2df __builtin_ia32_subsd (v2df, v2df)
1      v2df __builtin_ia32_mulsd (v2df, v2df)
1      v2df __builtin_ia32_divsd (v2df, v2df)
1      v2df __builtin_ia32_minpd (v2df, v2df)
1      v2df __builtin_ia32_maxpd (v2df, v2df)
1      v2df __builtin_ia32_minsd (v2df, v2df)
1      v2df __builtin_ia32_maxsd (v2df, v2df)
1      v2df __builtin_ia32_andpd (v2df, v2df)
1      v2df __builtin_ia32_andnpd (v2df, v2df)
1      v2df __builtin_ia32_orpd (v2df, v2df)
1      v2df __builtin_ia32_xorpd (v2df, v2df)
1      v2df __builtin_ia32_movsd (v2df, v2df)
1      v2df __builtin_ia32_unpckhpd (v2df, v2df)
1      v2df __builtin_ia32_unpcklpd (v2df, v2df)
1      v16qi __builtin_ia32_paddb128 (v16qi, v16qi)
1      v8hi __builtin_ia32_paddw128 (v8hi, v8hi)
1      v4si __builtin_ia32_paddd128 (v4si, v4si)
1      v2di __builtin_ia32_paddq128 (v2di, v2di)
1      v16qi __builtin_ia32_psubb128 (v16qi, v16qi)
1      v8hi __builtin_ia32_psubw128 (v8hi, v8hi)
1      v4si __builtin_ia32_psubd128 (v4si, v4si)
1      v2di __builtin_ia32_psubq128 (v2di, v2di)
1      v8hi __builtin_ia32_pmullw128 (v8hi, v8hi)
1      v8hi __builtin_ia32_pmulhw128 (v8hi, v8hi)
1      v2di __builtin_ia32_pand128 (v2di, v2di)
1      v2di __builtin_ia32_pandn128 (v2di, v2di)
1      v2di __builtin_ia32_por128 (v2di, v2di)
1      v2di __builtin_ia32_pxor128 (v2di, v2di)
1      v16qi __builtin_ia32_pavgb128 (v16qi, v16qi)
1      v8hi __builtin_ia32_pavgw128 (v8hi, v8hi)
1      v16qi __builtin_ia32_pcmpeqb128 (v16qi, v16qi)
1      v8hi __builtin_ia32_pcmpeqw128 (v8hi, v8hi)
1      v4si __builtin_ia32_pcmpeqd128 (v4si, v4si)
1      v16qi __builtin_ia32_pcmpgtb128 (v16qi, v16qi)
1      v8hi __builtin_ia32_pcmpgtw128 (v8hi, v8hi)
1      v4si __builtin_ia32_pcmpgtd128 (v4si, v4si)
1      v16qi __builtin_ia32_pmaxub128 (v16qi, v16qi)
1      v8hi __builtin_ia32_pmaxsw128 (v8hi, v8hi)
1      v16qi __builtin_ia32_pminub128 (v16qi, v16qi)
1      v8hi __builtin_ia32_pminsw128 (v8hi, v8hi)
1      v16qi __builtin_ia32_punpckhbw128 (v16qi, v16qi)
1      v8hi __builtin_ia32_punpckhwd128 (v8hi, v8hi)
1      v4si __builtin_ia32_punpckhdq128 (v4si, v4si)
1      v2di __builtin_ia32_punpckhqdq128 (v2di, v2di)
1      v16qi __builtin_ia32_punpcklbw128 (v16qi, v16qi)
1      v8hi __builtin_ia32_punpcklwd128 (v8hi, v8hi)
1      v4si __builtin_ia32_punpckldq128 (v4si, v4si)
1      v2di __builtin_ia32_punpcklqdq128 (v2di, v2di)
1      v16qi __builtin_ia32_packsswb128 (v8hi, v8hi)
1      v8hi __builtin_ia32_packssdw128 (v4si, v4si)
1      v16qi __builtin_ia32_packuswb128 (v8hi, v8hi)
1      v8hi __builtin_ia32_pmulhuw128 (v8hi, v8hi)
1      void __builtin_ia32_maskmovdqu (v16qi, v16qi)
1      v2df __builtin_ia32_loadupd (double *)
1      void __builtin_ia32_storeupd (double *, v2df)
1      v2df __builtin_ia32_loadhpd (v2df, double const *)
1      v2df __builtin_ia32_loadlpd (v2df, double const *)
1      int __builtin_ia32_movmskpd (v2df)
1      int __builtin_ia32_pmovmskb128 (v16qi)
1      void __builtin_ia32_movnti (int *, int)
1      void __builtin_ia32_movnti64 (long long int *, long long int)
1      void __builtin_ia32_movntpd (double *, v2df)
1      void __builtin_ia32_movntdq (v2df *, v2df)
1      v4si __builtin_ia32_pshufd (v4si, int)
1      v8hi __builtin_ia32_pshuflw (v8hi, int)
1      v8hi __builtin_ia32_pshufhw (v8hi, int)
1      v2di __builtin_ia32_psadbw128 (v16qi, v16qi)
1      v2df __builtin_ia32_sqrtpd (v2df)
1      v2df __builtin_ia32_sqrtsd (v2df)
1      v2df __builtin_ia32_shufpd (v2df, v2df, int)
1      v2df __builtin_ia32_cvtdq2pd (v4si)
1      v4sf __builtin_ia32_cvtdq2ps (v4si)
1      v4si __builtin_ia32_cvtpd2dq (v2df)
1      v2si __builtin_ia32_cvtpd2pi (v2df)
1      v4sf __builtin_ia32_cvtpd2ps (v2df)
1      v4si __builtin_ia32_cvttpd2dq (v2df)
1      v2si __builtin_ia32_cvttpd2pi (v2df)
1      v2df __builtin_ia32_cvtpi2pd (v2si)
1      int __builtin_ia32_cvtsd2si (v2df)
1      int __builtin_ia32_cvttsd2si (v2df)
1      long long __builtin_ia32_cvtsd2si64 (v2df)
1      long long __builtin_ia32_cvttsd2si64 (v2df)
1      v4si __builtin_ia32_cvtps2dq (v4sf)
1      v2df __builtin_ia32_cvtps2pd (v4sf)
1      v4si __builtin_ia32_cvttps2dq (v4sf)
1      v2df __builtin_ia32_cvtsi2sd (v2df, int)
1      v2df __builtin_ia32_cvtsi642sd (v2df, long long)
1      v4sf __builtin_ia32_cvtsd2ss (v4sf, v2df)
1      v2df __builtin_ia32_cvtss2sd (v2df, v4sf)
1      void __builtin_ia32_clflush (const void *)
1      void __builtin_ia32_lfence (void)
1      void __builtin_ia32_mfence (void)
1      v16qi __builtin_ia32_loaddqu (const char *)
1      void __builtin_ia32_storedqu (char *, v16qi)
1      v1di __builtin_ia32_pmuludq (v2si, v2si)
1      v2di __builtin_ia32_pmuludq128 (v4si, v4si)
1      v8hi __builtin_ia32_psllw128 (v8hi, v8hi)
1      v4si __builtin_ia32_pslld128 (v4si, v4si)
1      v2di __builtin_ia32_psllq128 (v2di, v2di)
1      v8hi __builtin_ia32_psrlw128 (v8hi, v8hi)
1      v4si __builtin_ia32_psrld128 (v4si, v4si)
1      v2di __builtin_ia32_psrlq128 (v2di, v2di)
1      v8hi __builtin_ia32_psraw128 (v8hi, v8hi)
1      v4si __builtin_ia32_psrad128 (v4si, v4si)
1      v2di __builtin_ia32_pslldqi128 (v2di, int)
1      v8hi __builtin_ia32_psllwi128 (v8hi, int)
1      v4si __builtin_ia32_pslldi128 (v4si, int)
1      v2di __builtin_ia32_psllqi128 (v2di, int)
1      v2di __builtin_ia32_psrldqi128 (v2di, int)
1      v8hi __builtin_ia32_psrlwi128 (v8hi, int)
1      v4si __builtin_ia32_psrldi128 (v4si, int)
1      v2di __builtin_ia32_psrlqi128 (v2di, int)
1      v8hi __builtin_ia32_psrawi128 (v8hi, int)
1      v4si __builtin_ia32_psradi128 (v4si, int)
1      v4si __builtin_ia32_pmaddwd128 (v8hi, v8hi)
1      v2di __builtin_ia32_movq128 (v2di)
1 
1  The following built-in functions are available when '-msse3' is used.
1 All of them generate the machine instruction that is part of the name.
1 
1      v2df __builtin_ia32_addsubpd (v2df, v2df)
1      v4sf __builtin_ia32_addsubps (v4sf, v4sf)
1      v2df __builtin_ia32_haddpd (v2df, v2df)
1      v4sf __builtin_ia32_haddps (v4sf, v4sf)
1      v2df __builtin_ia32_hsubpd (v2df, v2df)
1      v4sf __builtin_ia32_hsubps (v4sf, v4sf)
1      v16qi __builtin_ia32_lddqu (char const *)
1      void __builtin_ia32_monitor (void *, unsigned int, unsigned int)
1      v4sf __builtin_ia32_movshdup (v4sf)
1      v4sf __builtin_ia32_movsldup (v4sf)
1      void __builtin_ia32_mwait (unsigned int, unsigned int)
1 
1  The following built-in functions are available when '-mssse3' is used.
1 All of them generate the machine instruction that is part of the name.
1 
1      v2si __builtin_ia32_phaddd (v2si, v2si)
1      v4hi __builtin_ia32_phaddw (v4hi, v4hi)
1      v4hi __builtin_ia32_phaddsw (v4hi, v4hi)
1      v2si __builtin_ia32_phsubd (v2si, v2si)
1      v4hi __builtin_ia32_phsubw (v4hi, v4hi)
1      v4hi __builtin_ia32_phsubsw (v4hi, v4hi)
1      v4hi __builtin_ia32_pmaddubsw (v8qi, v8qi)
1      v4hi __builtin_ia32_pmulhrsw (v4hi, v4hi)
1      v8qi __builtin_ia32_pshufb (v8qi, v8qi)
1      v8qi __builtin_ia32_psignb (v8qi, v8qi)
1      v2si __builtin_ia32_psignd (v2si, v2si)
1      v4hi __builtin_ia32_psignw (v4hi, v4hi)
1      v1di __builtin_ia32_palignr (v1di, v1di, int)
1      v8qi __builtin_ia32_pabsb (v8qi)
1      v2si __builtin_ia32_pabsd (v2si)
1      v4hi __builtin_ia32_pabsw (v4hi)
1 
1  The following built-in functions are available when '-mssse3' is used.
1 All of them generate the machine instruction that is part of the name.
1 
1      v4si __builtin_ia32_phaddd128 (v4si, v4si)
1      v8hi __builtin_ia32_phaddw128 (v8hi, v8hi)
1      v8hi __builtin_ia32_phaddsw128 (v8hi, v8hi)
1      v4si __builtin_ia32_phsubd128 (v4si, v4si)
1      v8hi __builtin_ia32_phsubw128 (v8hi, v8hi)
1      v8hi __builtin_ia32_phsubsw128 (v8hi, v8hi)
1      v8hi __builtin_ia32_pmaddubsw128 (v16qi, v16qi)
1      v8hi __builtin_ia32_pmulhrsw128 (v8hi, v8hi)
1      v16qi __builtin_ia32_pshufb128 (v16qi, v16qi)
1      v16qi __builtin_ia32_psignb128 (v16qi, v16qi)
1      v4si __builtin_ia32_psignd128 (v4si, v4si)
1      v8hi __builtin_ia32_psignw128 (v8hi, v8hi)
1      v2di __builtin_ia32_palignr128 (v2di, v2di, int)
1      v16qi __builtin_ia32_pabsb128 (v16qi)
1      v4si __builtin_ia32_pabsd128 (v4si)
1      v8hi __builtin_ia32_pabsw128 (v8hi)
1 
1  The following built-in functions are available when '-msse4.1' is used.
1 All of them generate the machine instruction that is part of the name.
1 
1      v2df __builtin_ia32_blendpd (v2df, v2df, const int)
1      v4sf __builtin_ia32_blendps (v4sf, v4sf, const int)
1      v2df __builtin_ia32_blendvpd (v2df, v2df, v2df)
1      v4sf __builtin_ia32_blendvps (v4sf, v4sf, v4sf)
1      v2df __builtin_ia32_dppd (v2df, v2df, const int)
1      v4sf __builtin_ia32_dpps (v4sf, v4sf, const int)
1      v4sf __builtin_ia32_insertps128 (v4sf, v4sf, const int)
1      v2di __builtin_ia32_movntdqa (v2di *);
1      v16qi __builtin_ia32_mpsadbw128 (v16qi, v16qi, const int)
1      v8hi __builtin_ia32_packusdw128 (v4si, v4si)
1      v16qi __builtin_ia32_pblendvb128 (v16qi, v16qi, v16qi)
1      v8hi __builtin_ia32_pblendw128 (v8hi, v8hi, const int)
1      v2di __builtin_ia32_pcmpeqq (v2di, v2di)
1      v8hi __builtin_ia32_phminposuw128 (v8hi)
1      v16qi __builtin_ia32_pmaxsb128 (v16qi, v16qi)
1      v4si __builtin_ia32_pmaxsd128 (v4si, v4si)
1      v4si __builtin_ia32_pmaxud128 (v4si, v4si)
1      v8hi __builtin_ia32_pmaxuw128 (v8hi, v8hi)
1      v16qi __builtin_ia32_pminsb128 (v16qi, v16qi)
1      v4si __builtin_ia32_pminsd128 (v4si, v4si)
1      v4si __builtin_ia32_pminud128 (v4si, v4si)
1      v8hi __builtin_ia32_pminuw128 (v8hi, v8hi)
1      v4si __builtin_ia32_pmovsxbd128 (v16qi)
1      v2di __builtin_ia32_pmovsxbq128 (v16qi)
1      v8hi __builtin_ia32_pmovsxbw128 (v16qi)
1      v2di __builtin_ia32_pmovsxdq128 (v4si)
1      v4si __builtin_ia32_pmovsxwd128 (v8hi)
1      v2di __builtin_ia32_pmovsxwq128 (v8hi)
1      v4si __builtin_ia32_pmovzxbd128 (v16qi)
1      v2di __builtin_ia32_pmovzxbq128 (v16qi)
1      v8hi __builtin_ia32_pmovzxbw128 (v16qi)
1      v2di __builtin_ia32_pmovzxdq128 (v4si)
1      v4si __builtin_ia32_pmovzxwd128 (v8hi)
1      v2di __builtin_ia32_pmovzxwq128 (v8hi)
1      v2di __builtin_ia32_pmuldq128 (v4si, v4si)
1      v4si __builtin_ia32_pmulld128 (v4si, v4si)
1      int __builtin_ia32_ptestc128 (v2di, v2di)
1      int __builtin_ia32_ptestnzc128 (v2di, v2di)
1      int __builtin_ia32_ptestz128 (v2di, v2di)
1      v2df __builtin_ia32_roundpd (v2df, const int)
1      v4sf __builtin_ia32_roundps (v4sf, const int)
1      v2df __builtin_ia32_roundsd (v2df, v2df, const int)
1      v4sf __builtin_ia32_roundss (v4sf, v4sf, const int)
1 
1  The following built-in functions are available when '-msse4.1' is used.
1 
1 'v4sf __builtin_ia32_vec_set_v4sf (v4sf, float, const int)'
1      Generates the 'insertps' machine instruction.
1 'int __builtin_ia32_vec_ext_v16qi (v16qi, const int)'
1      Generates the 'pextrb' machine instruction.
1 'v16qi __builtin_ia32_vec_set_v16qi (v16qi, int, const int)'
1      Generates the 'pinsrb' machine instruction.
1 'v4si __builtin_ia32_vec_set_v4si (v4si, int, const int)'
1      Generates the 'pinsrd' machine instruction.
1 'v2di __builtin_ia32_vec_set_v2di (v2di, long long, const int)'
1      Generates the 'pinsrq' machine instruction in 64bit mode.
1 
1  The following built-in functions are changed to generate new SSE4.1
1 instructions when '-msse4.1' is used.
1 
1 'float __builtin_ia32_vec_ext_v4sf (v4sf, const int)'
1      Generates the 'extractps' machine instruction.
1 'int __builtin_ia32_vec_ext_v4si (v4si, const int)'
1      Generates the 'pextrd' machine instruction.
1 'long long __builtin_ia32_vec_ext_v2di (v2di, const int)'
1      Generates the 'pextrq' machine instruction in 64bit mode.
1 
1  The following built-in functions are available when '-msse4.2' is used.
1 All of them generate the machine instruction that is part of the name.
1 
1      v16qi __builtin_ia32_pcmpestrm128 (v16qi, int, v16qi, int, const int)
1      int __builtin_ia32_pcmpestri128 (v16qi, int, v16qi, int, const int)
1      int __builtin_ia32_pcmpestria128 (v16qi, int, v16qi, int, const int)
1      int __builtin_ia32_pcmpestric128 (v16qi, int, v16qi, int, const int)
1      int __builtin_ia32_pcmpestrio128 (v16qi, int, v16qi, int, const int)
1      int __builtin_ia32_pcmpestris128 (v16qi, int, v16qi, int, const int)
1      int __builtin_ia32_pcmpestriz128 (v16qi, int, v16qi, int, const int)
1      v16qi __builtin_ia32_pcmpistrm128 (v16qi, v16qi, const int)
1      int __builtin_ia32_pcmpistri128 (v16qi, v16qi, const int)
1      int __builtin_ia32_pcmpistria128 (v16qi, v16qi, const int)
1      int __builtin_ia32_pcmpistric128 (v16qi, v16qi, const int)
1      int __builtin_ia32_pcmpistrio128 (v16qi, v16qi, const int)
1      int __builtin_ia32_pcmpistris128 (v16qi, v16qi, const int)
1      int __builtin_ia32_pcmpistriz128 (v16qi, v16qi, const int)
1      v2di __builtin_ia32_pcmpgtq (v2di, v2di)
1 
1  The following built-in functions are available when '-msse4.2' is used.
1 
1 'unsigned int __builtin_ia32_crc32qi (unsigned int, unsigned char)'
1      Generates the 'crc32b' machine instruction.
1 'unsigned int __builtin_ia32_crc32hi (unsigned int, unsigned short)'
1      Generates the 'crc32w' machine instruction.
1 'unsigned int __builtin_ia32_crc32si (unsigned int, unsigned int)'
1      Generates the 'crc32l' machine instruction.
1 'unsigned long long __builtin_ia32_crc32di (unsigned long long, unsigned long long)'
1      Generates the 'crc32q' machine instruction.
1 
1  The following built-in functions are changed to generate new SSE4.2
1 instructions when '-msse4.2' is used.
1 
1 'int __builtin_popcount (unsigned int)'
1      Generates the 'popcntl' machine instruction.
1 'int __builtin_popcountl (unsigned long)'
1      Generates the 'popcntl' or 'popcntq' machine instruction, depending
1      on the size of 'unsigned long'.
1 'int __builtin_popcountll (unsigned long long)'
1      Generates the 'popcntq' machine instruction.
1 
1  The following built-in functions are available when '-mavx' is used.
1 All of them generate the machine instruction that is part of the name.
1 
1      v4df __builtin_ia32_addpd256 (v4df,v4df)
1      v8sf __builtin_ia32_addps256 (v8sf,v8sf)
1      v4df __builtin_ia32_addsubpd256 (v4df,v4df)
1      v8sf __builtin_ia32_addsubps256 (v8sf,v8sf)
1      v4df __builtin_ia32_andnpd256 (v4df,v4df)
1      v8sf __builtin_ia32_andnps256 (v8sf,v8sf)
1      v4df __builtin_ia32_andpd256 (v4df,v4df)
1      v8sf __builtin_ia32_andps256 (v8sf,v8sf)
1      v4df __builtin_ia32_blendpd256 (v4df,v4df,int)
1      v8sf __builtin_ia32_blendps256 (v8sf,v8sf,int)
1      v4df __builtin_ia32_blendvpd256 (v4df,v4df,v4df)
1      v8sf __builtin_ia32_blendvps256 (v8sf,v8sf,v8sf)
1      v2df __builtin_ia32_cmppd (v2df,v2df,int)
1      v4df __builtin_ia32_cmppd256 (v4df,v4df,int)
1      v4sf __builtin_ia32_cmpps (v4sf,v4sf,int)
1      v8sf __builtin_ia32_cmpps256 (v8sf,v8sf,int)
1      v2df __builtin_ia32_cmpsd (v2df,v2df,int)
1      v4sf __builtin_ia32_cmpss (v4sf,v4sf,int)
1      v4df __builtin_ia32_cvtdq2pd256 (v4si)
1      v8sf __builtin_ia32_cvtdq2ps256 (v8si)
1      v4si __builtin_ia32_cvtpd2dq256 (v4df)
1      v4sf __builtin_ia32_cvtpd2ps256 (v4df)
1      v8si __builtin_ia32_cvtps2dq256 (v8sf)
1      v4df __builtin_ia32_cvtps2pd256 (v4sf)
1      v4si __builtin_ia32_cvttpd2dq256 (v4df)
1      v8si __builtin_ia32_cvttps2dq256 (v8sf)
1      v4df __builtin_ia32_divpd256 (v4df,v4df)
1      v8sf __builtin_ia32_divps256 (v8sf,v8sf)
1      v8sf __builtin_ia32_dpps256 (v8sf,v8sf,int)
1      v4df __builtin_ia32_haddpd256 (v4df,v4df)
1      v8sf __builtin_ia32_haddps256 (v8sf,v8sf)
1      v4df __builtin_ia32_hsubpd256 (v4df,v4df)
1      v8sf __builtin_ia32_hsubps256 (v8sf,v8sf)
1      v32qi __builtin_ia32_lddqu256 (pcchar)
1      v32qi __builtin_ia32_loaddqu256 (pcchar)
1      v4df __builtin_ia32_loadupd256 (pcdouble)
1      v8sf __builtin_ia32_loadups256 (pcfloat)
1      v2df __builtin_ia32_maskloadpd (pcv2df,v2df)
1      v4df __builtin_ia32_maskloadpd256 (pcv4df,v4df)
1      v4sf __builtin_ia32_maskloadps (pcv4sf,v4sf)
1      v8sf __builtin_ia32_maskloadps256 (pcv8sf,v8sf)
1      void __builtin_ia32_maskstorepd (pv2df,v2df,v2df)
1      void __builtin_ia32_maskstorepd256 (pv4df,v4df,v4df)
1      void __builtin_ia32_maskstoreps (pv4sf,v4sf,v4sf)
1      void __builtin_ia32_maskstoreps256 (pv8sf,v8sf,v8sf)
1      v4df __builtin_ia32_maxpd256 (v4df,v4df)
1      v8sf __builtin_ia32_maxps256 (v8sf,v8sf)
1      v4df __builtin_ia32_minpd256 (v4df,v4df)
1      v8sf __builtin_ia32_minps256 (v8sf,v8sf)
1      v4df __builtin_ia32_movddup256 (v4df)
1      int __builtin_ia32_movmskpd256 (v4df)
1      int __builtin_ia32_movmskps256 (v8sf)
1      v8sf __builtin_ia32_movshdup256 (v8sf)
1      v8sf __builtin_ia32_movsldup256 (v8sf)
1      v4df __builtin_ia32_mulpd256 (v4df,v4df)
1      v8sf __builtin_ia32_mulps256 (v8sf,v8sf)
1      v4df __builtin_ia32_orpd256 (v4df,v4df)
1      v8sf __builtin_ia32_orps256 (v8sf,v8sf)
1      v2df __builtin_ia32_pd_pd256 (v4df)
1      v4df __builtin_ia32_pd256_pd (v2df)
1      v4sf __builtin_ia32_ps_ps256 (v8sf)
1      v8sf __builtin_ia32_ps256_ps (v4sf)
1      int __builtin_ia32_ptestc256 (v4di,v4di,ptest)
1      int __builtin_ia32_ptestnzc256 (v4di,v4di,ptest)
1      int __builtin_ia32_ptestz256 (v4di,v4di,ptest)
1      v8sf __builtin_ia32_rcpps256 (v8sf)
1      v4df __builtin_ia32_roundpd256 (v4df,int)
1      v8sf __builtin_ia32_roundps256 (v8sf,int)
1      v8sf __builtin_ia32_rsqrtps_nr256 (v8sf)
1      v8sf __builtin_ia32_rsqrtps256 (v8sf)
1      v4df __builtin_ia32_shufpd256 (v4df,v4df,int)
1      v8sf __builtin_ia32_shufps256 (v8sf,v8sf,int)
1      v4si __builtin_ia32_si_si256 (v8si)
1      v8si __builtin_ia32_si256_si (v4si)
1      v4df __builtin_ia32_sqrtpd256 (v4df)
1      v8sf __builtin_ia32_sqrtps_nr256 (v8sf)
1      v8sf __builtin_ia32_sqrtps256 (v8sf)
1      void __builtin_ia32_storedqu256 (pchar,v32qi)
1      void __builtin_ia32_storeupd256 (pdouble,v4df)
1      void __builtin_ia32_storeups256 (pfloat,v8sf)
1      v4df __builtin_ia32_subpd256 (v4df,v4df)
1      v8sf __builtin_ia32_subps256 (v8sf,v8sf)
1      v4df __builtin_ia32_unpckhpd256 (v4df,v4df)
1      v8sf __builtin_ia32_unpckhps256 (v8sf,v8sf)
1      v4df __builtin_ia32_unpcklpd256 (v4df,v4df)
1      v8sf __builtin_ia32_unpcklps256 (v8sf,v8sf)
1      v4df __builtin_ia32_vbroadcastf128_pd256 (pcv2df)
1      v8sf __builtin_ia32_vbroadcastf128_ps256 (pcv4sf)
1      v4df __builtin_ia32_vbroadcastsd256 (pcdouble)
1      v4sf __builtin_ia32_vbroadcastss (pcfloat)
1      v8sf __builtin_ia32_vbroadcastss256 (pcfloat)
1      v2df __builtin_ia32_vextractf128_pd256 (v4df,int)
1      v4sf __builtin_ia32_vextractf128_ps256 (v8sf,int)
1      v4si __builtin_ia32_vextractf128_si256 (v8si,int)
1      v4df __builtin_ia32_vinsertf128_pd256 (v4df,v2df,int)
1      v8sf __builtin_ia32_vinsertf128_ps256 (v8sf,v4sf,int)
1      v8si __builtin_ia32_vinsertf128_si256 (v8si,v4si,int)
1      v4df __builtin_ia32_vperm2f128_pd256 (v4df,v4df,int)
1      v8sf __builtin_ia32_vperm2f128_ps256 (v8sf,v8sf,int)
1      v8si __builtin_ia32_vperm2f128_si256 (v8si,v8si,int)
1      v2df __builtin_ia32_vpermil2pd (v2df,v2df,v2di,int)
1      v4df __builtin_ia32_vpermil2pd256 (v4df,v4df,v4di,int)
1      v4sf __builtin_ia32_vpermil2ps (v4sf,v4sf,v4si,int)
1      v8sf __builtin_ia32_vpermil2ps256 (v8sf,v8sf,v8si,int)
1      v2df __builtin_ia32_vpermilpd (v2df,int)
1      v4df __builtin_ia32_vpermilpd256 (v4df,int)
1      v4sf __builtin_ia32_vpermilps (v4sf,int)
1      v8sf __builtin_ia32_vpermilps256 (v8sf,int)
1      v2df __builtin_ia32_vpermilvarpd (v2df,v2di)
1      v4df __builtin_ia32_vpermilvarpd256 (v4df,v4di)
1      v4sf __builtin_ia32_vpermilvarps (v4sf,v4si)
1      v8sf __builtin_ia32_vpermilvarps256 (v8sf,v8si)
1      int __builtin_ia32_vtestcpd (v2df,v2df,ptest)
1      int __builtin_ia32_vtestcpd256 (v4df,v4df,ptest)
1      int __builtin_ia32_vtestcps (v4sf,v4sf,ptest)
1      int __builtin_ia32_vtestcps256 (v8sf,v8sf,ptest)
1      int __builtin_ia32_vtestnzcpd (v2df,v2df,ptest)
1      int __builtin_ia32_vtestnzcpd256 (v4df,v4df,ptest)
1      int __builtin_ia32_vtestnzcps (v4sf,v4sf,ptest)
1      int __builtin_ia32_vtestnzcps256 (v8sf,v8sf,ptest)
1      int __builtin_ia32_vtestzpd (v2df,v2df,ptest)
1      int __builtin_ia32_vtestzpd256 (v4df,v4df,ptest)
1      int __builtin_ia32_vtestzps (v4sf,v4sf,ptest)
1      int __builtin_ia32_vtestzps256 (v8sf,v8sf,ptest)
1      void __builtin_ia32_vzeroall (void)
1      void __builtin_ia32_vzeroupper (void)
1      v4df __builtin_ia32_xorpd256 (v4df,v4df)
1      v8sf __builtin_ia32_xorps256 (v8sf,v8sf)
1 
1  The following built-in functions are available when '-mavx2' is used.
1 All of them generate the machine instruction that is part of the name.
1 
1      v32qi __builtin_ia32_mpsadbw256 (v32qi,v32qi,int)
1      v32qi __builtin_ia32_pabsb256 (v32qi)
1      v16hi __builtin_ia32_pabsw256 (v16hi)
1      v8si __builtin_ia32_pabsd256 (v8si)
1      v16hi __builtin_ia32_packssdw256 (v8si,v8si)
1      v32qi __builtin_ia32_packsswb256 (v16hi,v16hi)
1      v16hi __builtin_ia32_packusdw256 (v8si,v8si)
1      v32qi __builtin_ia32_packuswb256 (v16hi,v16hi)
1      v32qi __builtin_ia32_paddb256 (v32qi,v32qi)
1      v16hi __builtin_ia32_paddw256 (v16hi,v16hi)
1      v8si __builtin_ia32_paddd256 (v8si,v8si)
1      v4di __builtin_ia32_paddq256 (v4di,v4di)
1      v32qi __builtin_ia32_paddsb256 (v32qi,v32qi)
1      v16hi __builtin_ia32_paddsw256 (v16hi,v16hi)
1      v32qi __builtin_ia32_paddusb256 (v32qi,v32qi)
1      v16hi __builtin_ia32_paddusw256 (v16hi,v16hi)
1      v4di __builtin_ia32_palignr256 (v4di,v4di,int)
1      v4di __builtin_ia32_andsi256 (v4di,v4di)
1      v4di __builtin_ia32_andnotsi256 (v4di,v4di)
1      v32qi __builtin_ia32_pavgb256 (v32qi,v32qi)
1      v16hi __builtin_ia32_pavgw256 (v16hi,v16hi)
1      v32qi __builtin_ia32_pblendvb256 (v32qi,v32qi,v32qi)
1      v16hi __builtin_ia32_pblendw256 (v16hi,v16hi,int)
1      v32qi __builtin_ia32_pcmpeqb256 (v32qi,v32qi)
1      v16hi __builtin_ia32_pcmpeqw256 (v16hi,v16hi)
1      v8si __builtin_ia32_pcmpeqd256 (c8si,v8si)
1      v4di __builtin_ia32_pcmpeqq256 (v4di,v4di)
1      v32qi __builtin_ia32_pcmpgtb256 (v32qi,v32qi)
1      v16hi __builtin_ia32_pcmpgtw256 (16hi,v16hi)
1      v8si __builtin_ia32_pcmpgtd256 (v8si,v8si)
1      v4di __builtin_ia32_pcmpgtq256 (v4di,v4di)
1      v16hi __builtin_ia32_phaddw256 (v16hi,v16hi)
1      v8si __builtin_ia32_phaddd256 (v8si,v8si)
1      v16hi __builtin_ia32_phaddsw256 (v16hi,v16hi)
1      v16hi __builtin_ia32_phsubw256 (v16hi,v16hi)
1      v8si __builtin_ia32_phsubd256 (v8si,v8si)
1      v16hi __builtin_ia32_phsubsw256 (v16hi,v16hi)
1      v32qi __builtin_ia32_pmaddubsw256 (v32qi,v32qi)
1      v16hi __builtin_ia32_pmaddwd256 (v16hi,v16hi)
1      v32qi __builtin_ia32_pmaxsb256 (v32qi,v32qi)
1      v16hi __builtin_ia32_pmaxsw256 (v16hi,v16hi)
1      v8si __builtin_ia32_pmaxsd256 (v8si,v8si)
1      v32qi __builtin_ia32_pmaxub256 (v32qi,v32qi)
1      v16hi __builtin_ia32_pmaxuw256 (v16hi,v16hi)
1      v8si __builtin_ia32_pmaxud256 (v8si,v8si)
1      v32qi __builtin_ia32_pminsb256 (v32qi,v32qi)
1      v16hi __builtin_ia32_pminsw256 (v16hi,v16hi)
1      v8si __builtin_ia32_pminsd256 (v8si,v8si)
1      v32qi __builtin_ia32_pminub256 (v32qi,v32qi)
1      v16hi __builtin_ia32_pminuw256 (v16hi,v16hi)
1      v8si __builtin_ia32_pminud256 (v8si,v8si)
1      int __builtin_ia32_pmovmskb256 (v32qi)
1      v16hi __builtin_ia32_pmovsxbw256 (v16qi)
1      v8si __builtin_ia32_pmovsxbd256 (v16qi)
1      v4di __builtin_ia32_pmovsxbq256 (v16qi)
1      v8si __builtin_ia32_pmovsxwd256 (v8hi)
1      v4di __builtin_ia32_pmovsxwq256 (v8hi)
1      v4di __builtin_ia32_pmovsxdq256 (v4si)
1      v16hi __builtin_ia32_pmovzxbw256 (v16qi)
1      v8si __builtin_ia32_pmovzxbd256 (v16qi)
1      v4di __builtin_ia32_pmovzxbq256 (v16qi)
1      v8si __builtin_ia32_pmovzxwd256 (v8hi)
1      v4di __builtin_ia32_pmovzxwq256 (v8hi)
1      v4di __builtin_ia32_pmovzxdq256 (v4si)
1      v4di __builtin_ia32_pmuldq256 (v8si,v8si)
1      v16hi __builtin_ia32_pmulhrsw256 (v16hi, v16hi)
1      v16hi __builtin_ia32_pmulhuw256 (v16hi,v16hi)
1      v16hi __builtin_ia32_pmulhw256 (v16hi,v16hi)
1      v16hi __builtin_ia32_pmullw256 (v16hi,v16hi)
1      v8si __builtin_ia32_pmulld256 (v8si,v8si)
1      v4di __builtin_ia32_pmuludq256 (v8si,v8si)
1      v4di __builtin_ia32_por256 (v4di,v4di)
1      v16hi __builtin_ia32_psadbw256 (v32qi,v32qi)
1      v32qi __builtin_ia32_pshufb256 (v32qi,v32qi)
1      v8si __builtin_ia32_pshufd256 (v8si,int)
1      v16hi __builtin_ia32_pshufhw256 (v16hi,int)
1      v16hi __builtin_ia32_pshuflw256 (v16hi,int)
1      v32qi __builtin_ia32_psignb256 (v32qi,v32qi)
1      v16hi __builtin_ia32_psignw256 (v16hi,v16hi)
1      v8si __builtin_ia32_psignd256 (v8si,v8si)
1      v4di __builtin_ia32_pslldqi256 (v4di,int)
1      v16hi __builtin_ia32_psllwi256 (16hi,int)
1      v16hi __builtin_ia32_psllw256(v16hi,v8hi)
1      v8si __builtin_ia32_pslldi256 (v8si,int)
1      v8si __builtin_ia32_pslld256(v8si,v4si)
1      v4di __builtin_ia32_psllqi256 (v4di,int)
1      v4di __builtin_ia32_psllq256(v4di,v2di)
1      v16hi __builtin_ia32_psrawi256 (v16hi,int)
1      v16hi __builtin_ia32_psraw256 (v16hi,v8hi)
1      v8si __builtin_ia32_psradi256 (v8si,int)
1      v8si __builtin_ia32_psrad256 (v8si,v4si)
1      v4di __builtin_ia32_psrldqi256 (v4di, int)
1      v16hi __builtin_ia32_psrlwi256 (v16hi,int)
1      v16hi __builtin_ia32_psrlw256 (v16hi,v8hi)
1      v8si __builtin_ia32_psrldi256 (v8si,int)
1      v8si __builtin_ia32_psrld256 (v8si,v4si)
1      v4di __builtin_ia32_psrlqi256 (v4di,int)
1      v4di __builtin_ia32_psrlq256(v4di,v2di)
1      v32qi __builtin_ia32_psubb256 (v32qi,v32qi)
1      v32hi __builtin_ia32_psubw256 (v16hi,v16hi)
1      v8si __builtin_ia32_psubd256 (v8si,v8si)
1      v4di __builtin_ia32_psubq256 (v4di,v4di)
1      v32qi __builtin_ia32_psubsb256 (v32qi,v32qi)
1      v16hi __builtin_ia32_psubsw256 (v16hi,v16hi)
1      v32qi __builtin_ia32_psubusb256 (v32qi,v32qi)
1      v16hi __builtin_ia32_psubusw256 (v16hi,v16hi)
1      v32qi __builtin_ia32_punpckhbw256 (v32qi,v32qi)
1      v16hi __builtin_ia32_punpckhwd256 (v16hi,v16hi)
1      v8si __builtin_ia32_punpckhdq256 (v8si,v8si)
1      v4di __builtin_ia32_punpckhqdq256 (v4di,v4di)
1      v32qi __builtin_ia32_punpcklbw256 (v32qi,v32qi)
1      v16hi __builtin_ia32_punpcklwd256 (v16hi,v16hi)
1      v8si __builtin_ia32_punpckldq256 (v8si,v8si)
1      v4di __builtin_ia32_punpcklqdq256 (v4di,v4di)
1      v4di __builtin_ia32_pxor256 (v4di,v4di)
1      v4di __builtin_ia32_movntdqa256 (pv4di)
1      v4sf __builtin_ia32_vbroadcastss_ps (v4sf)
1      v8sf __builtin_ia32_vbroadcastss_ps256 (v4sf)
1      v4df __builtin_ia32_vbroadcastsd_pd256 (v2df)
1      v4di __builtin_ia32_vbroadcastsi256 (v2di)
1      v4si __builtin_ia32_pblendd128 (v4si,v4si)
1      v8si __builtin_ia32_pblendd256 (v8si,v8si)
1      v32qi __builtin_ia32_pbroadcastb256 (v16qi)
1      v16hi __builtin_ia32_pbroadcastw256 (v8hi)
1      v8si __builtin_ia32_pbroadcastd256 (v4si)
1      v4di __builtin_ia32_pbroadcastq256 (v2di)
1      v16qi __builtin_ia32_pbroadcastb128 (v16qi)
1      v8hi __builtin_ia32_pbroadcastw128 (v8hi)
1      v4si __builtin_ia32_pbroadcastd128 (v4si)
1      v2di __builtin_ia32_pbroadcastq128 (v2di)
1      v8si __builtin_ia32_permvarsi256 (v8si,v8si)
1      v4df __builtin_ia32_permdf256 (v4df,int)
1      v8sf __builtin_ia32_permvarsf256 (v8sf,v8sf)
1      v4di __builtin_ia32_permdi256 (v4di,int)
1      v4di __builtin_ia32_permti256 (v4di,v4di,int)
1      v4di __builtin_ia32_extract128i256 (v4di,int)
1      v4di __builtin_ia32_insert128i256 (v4di,v2di,int)
1      v8si __builtin_ia32_maskloadd256 (pcv8si,v8si)
1      v4di __builtin_ia32_maskloadq256 (pcv4di,v4di)
1      v4si __builtin_ia32_maskloadd (pcv4si,v4si)
1      v2di __builtin_ia32_maskloadq (pcv2di,v2di)
1      void __builtin_ia32_maskstored256 (pv8si,v8si,v8si)
1      void __builtin_ia32_maskstoreq256 (pv4di,v4di,v4di)
1      void __builtin_ia32_maskstored (pv4si,v4si,v4si)
1      void __builtin_ia32_maskstoreq (pv2di,v2di,v2di)
1      v8si __builtin_ia32_psllv8si (v8si,v8si)
1      v4si __builtin_ia32_psllv4si (v4si,v4si)
1      v4di __builtin_ia32_psllv4di (v4di,v4di)
1      v2di __builtin_ia32_psllv2di (v2di,v2di)
1      v8si __builtin_ia32_psrav8si (v8si,v8si)
1      v4si __builtin_ia32_psrav4si (v4si,v4si)
1      v8si __builtin_ia32_psrlv8si (v8si,v8si)
1      v4si __builtin_ia32_psrlv4si (v4si,v4si)
1      v4di __builtin_ia32_psrlv4di (v4di,v4di)
1      v2di __builtin_ia32_psrlv2di (v2di,v2di)
1      v2df __builtin_ia32_gathersiv2df (v2df, pcdouble,v4si,v2df,int)
1      v4df __builtin_ia32_gathersiv4df (v4df, pcdouble,v4si,v4df,int)
1      v2df __builtin_ia32_gatherdiv2df (v2df, pcdouble,v2di,v2df,int)
1      v4df __builtin_ia32_gatherdiv4df (v4df, pcdouble,v4di,v4df,int)
1      v4sf __builtin_ia32_gathersiv4sf (v4sf, pcfloat,v4si,v4sf,int)
1      v8sf __builtin_ia32_gathersiv8sf (v8sf, pcfloat,v8si,v8sf,int)
1      v4sf __builtin_ia32_gatherdiv4sf (v4sf, pcfloat,v2di,v4sf,int)
1      v4sf __builtin_ia32_gatherdiv4sf256 (v4sf, pcfloat,v4di,v4sf,int)
1      v2di __builtin_ia32_gathersiv2di (v2di, pcint64,v4si,v2di,int)
1      v4di __builtin_ia32_gathersiv4di (v4di, pcint64,v4si,v4di,int)
1      v2di __builtin_ia32_gatherdiv2di (v2di, pcint64,v2di,v2di,int)
1      v4di __builtin_ia32_gatherdiv4di (v4di, pcint64,v4di,v4di,int)
1      v4si __builtin_ia32_gathersiv4si (v4si, pcint,v4si,v4si,int)
1      v8si __builtin_ia32_gathersiv8si (v8si, pcint,v8si,v8si,int)
1      v4si __builtin_ia32_gatherdiv4si (v4si, pcint,v2di,v4si,int)
1      v4si __builtin_ia32_gatherdiv4si256 (v4si, pcint,v4di,v4si,int)
1 
1  The following built-in functions are available when '-maes' is used.
1 All of them generate the machine instruction that is part of the name.
1 
1      v2di __builtin_ia32_aesenc128 (v2di, v2di)
1      v2di __builtin_ia32_aesenclast128 (v2di, v2di)
1      v2di __builtin_ia32_aesdec128 (v2di, v2di)
1      v2di __builtin_ia32_aesdeclast128 (v2di, v2di)
1      v2di __builtin_ia32_aeskeygenassist128 (v2di, const int)
1      v2di __builtin_ia32_aesimc128 (v2di)
1 
1  The following built-in function is available when '-mpclmul' is used.
1 
1 'v2di __builtin_ia32_pclmulqdq128 (v2di, v2di, const int)'
1      Generates the 'pclmulqdq' machine instruction.
1 
1  The following built-in function is available when '-mfsgsbase' is used.
1 All of them generate the machine instruction that is part of the name.
1 
1      unsigned int __builtin_ia32_rdfsbase32 (void)
1      unsigned long long __builtin_ia32_rdfsbase64 (void)
1      unsigned int __builtin_ia32_rdgsbase32 (void)
1      unsigned long long __builtin_ia32_rdgsbase64 (void)
1      void _writefsbase_u32 (unsigned int)
1      void _writefsbase_u64 (unsigned long long)
1      void _writegsbase_u32 (unsigned int)
1      void _writegsbase_u64 (unsigned long long)
1 
1  The following built-in function is available when '-mrdrnd' is used.
1 All of them generate the machine instruction that is part of the name.
1 
1      unsigned int __builtin_ia32_rdrand16_step (unsigned short *)
1      unsigned int __builtin_ia32_rdrand32_step (unsigned int *)
1      unsigned int __builtin_ia32_rdrand64_step (unsigned long long *)
1 
1  The following built-in functions are available when '-msse4a' is used.
1 All of them generate the machine instruction that is part of the name.
1 
1      void __builtin_ia32_movntsd (double *, v2df)
1      void __builtin_ia32_movntss (float *, v4sf)
1      v2di __builtin_ia32_extrq  (v2di, v16qi)
1      v2di __builtin_ia32_extrqi (v2di, const unsigned int, const unsigned int)
1      v2di __builtin_ia32_insertq (v2di, v2di)
1      v2di __builtin_ia32_insertqi (v2di, v2di, const unsigned int, const unsigned int)
1 
1  The following built-in functions are available when '-mxop' is used.
1      v2df __builtin_ia32_vfrczpd (v2df)
1      v4sf __builtin_ia32_vfrczps (v4sf)
1      v2df __builtin_ia32_vfrczsd (v2df)
1      v4sf __builtin_ia32_vfrczss (v4sf)
1      v4df __builtin_ia32_vfrczpd256 (v4df)
1      v8sf __builtin_ia32_vfrczps256 (v8sf)
1      v2di __builtin_ia32_vpcmov (v2di, v2di, v2di)
1      v2di __builtin_ia32_vpcmov_v2di (v2di, v2di, v2di)
1      v4si __builtin_ia32_vpcmov_v4si (v4si, v4si, v4si)
1      v8hi __builtin_ia32_vpcmov_v8hi (v8hi, v8hi, v8hi)
1      v16qi __builtin_ia32_vpcmov_v16qi (v16qi, v16qi, v16qi)
1      v2df __builtin_ia32_vpcmov_v2df (v2df, v2df, v2df)
1      v4sf __builtin_ia32_vpcmov_v4sf (v4sf, v4sf, v4sf)
1      v4di __builtin_ia32_vpcmov_v4di256 (v4di, v4di, v4di)
1      v8si __builtin_ia32_vpcmov_v8si256 (v8si, v8si, v8si)
1      v16hi __builtin_ia32_vpcmov_v16hi256 (v16hi, v16hi, v16hi)
1      v32qi __builtin_ia32_vpcmov_v32qi256 (v32qi, v32qi, v32qi)
1      v4df __builtin_ia32_vpcmov_v4df256 (v4df, v4df, v4df)
1      v8sf __builtin_ia32_vpcmov_v8sf256 (v8sf, v8sf, v8sf)
1      v16qi __builtin_ia32_vpcomeqb (v16qi, v16qi)
1      v8hi __builtin_ia32_vpcomeqw (v8hi, v8hi)
1      v4si __builtin_ia32_vpcomeqd (v4si, v4si)
1      v2di __builtin_ia32_vpcomeqq (v2di, v2di)
1      v16qi __builtin_ia32_vpcomequb (v16qi, v16qi)
1      v4si __builtin_ia32_vpcomequd (v4si, v4si)
1      v2di __builtin_ia32_vpcomequq (v2di, v2di)
1      v8hi __builtin_ia32_vpcomequw (v8hi, v8hi)
1      v8hi __builtin_ia32_vpcomeqw (v8hi, v8hi)
1      v16qi __builtin_ia32_vpcomfalseb (v16qi, v16qi)
1      v4si __builtin_ia32_vpcomfalsed (v4si, v4si)
1      v2di __builtin_ia32_vpcomfalseq (v2di, v2di)
1      v16qi __builtin_ia32_vpcomfalseub (v16qi, v16qi)
1      v4si __builtin_ia32_vpcomfalseud (v4si, v4si)
1      v2di __builtin_ia32_vpcomfalseuq (v2di, v2di)
1      v8hi __builtin_ia32_vpcomfalseuw (v8hi, v8hi)
1      v8hi __builtin_ia32_vpcomfalsew (v8hi, v8hi)
1      v16qi __builtin_ia32_vpcomgeb (v16qi, v16qi)
1      v4si __builtin_ia32_vpcomged (v4si, v4si)
1      v2di __builtin_ia32_vpcomgeq (v2di, v2di)
1      v16qi __builtin_ia32_vpcomgeub (v16qi, v16qi)
1      v4si __builtin_ia32_vpcomgeud (v4si, v4si)
1      v2di __builtin_ia32_vpcomgeuq (v2di, v2di)
1      v8hi __builtin_ia32_vpcomgeuw (v8hi, v8hi)
1      v8hi __builtin_ia32_vpcomgew (v8hi, v8hi)
1      v16qi __builtin_ia32_vpcomgtb (v16qi, v16qi)
1      v4si __builtin_ia32_vpcomgtd (v4si, v4si)
1      v2di __builtin_ia32_vpcomgtq (v2di, v2di)
1      v16qi __builtin_ia32_vpcomgtub (v16qi, v16qi)
1      v4si __builtin_ia32_vpcomgtud (v4si, v4si)
1      v2di __builtin_ia32_vpcomgtuq (v2di, v2di)
1      v8hi __builtin_ia32_vpcomgtuw (v8hi, v8hi)
1      v8hi __builtin_ia32_vpcomgtw (v8hi, v8hi)
1      v16qi __builtin_ia32_vpcomleb (v16qi, v16qi)
1      v4si __builtin_ia32_vpcomled (v4si, v4si)
1      v2di __builtin_ia32_vpcomleq (v2di, v2di)
1      v16qi __builtin_ia32_vpcomleub (v16qi, v16qi)
1      v4si __builtin_ia32_vpcomleud (v4si, v4si)
1      v2di __builtin_ia32_vpcomleuq (v2di, v2di)
1      v8hi __builtin_ia32_vpcomleuw (v8hi, v8hi)
1      v8hi __builtin_ia32_vpcomlew (v8hi, v8hi)
1      v16qi __builtin_ia32_vpcomltb (v16qi, v16qi)
1      v4si __builtin_ia32_vpcomltd (v4si, v4si)
1      v2di __builtin_ia32_vpcomltq (v2di, v2di)
1      v16qi __builtin_ia32_vpcomltub (v16qi, v16qi)
1      v4si __builtin_ia32_vpcomltud (v4si, v4si)
1      v2di __builtin_ia32_vpcomltuq (v2di, v2di)
1      v8hi __builtin_ia32_vpcomltuw (v8hi, v8hi)
1      v8hi __builtin_ia32_vpcomltw (v8hi, v8hi)
1      v16qi __builtin_ia32_vpcomneb (v16qi, v16qi)
1      v4si __builtin_ia32_vpcomned (v4si, v4si)
1      v2di __builtin_ia32_vpcomneq (v2di, v2di)
1      v16qi __builtin_ia32_vpcomneub (v16qi, v16qi)
1      v4si __builtin_ia32_vpcomneud (v4si, v4si)
1      v2di __builtin_ia32_vpcomneuq (v2di, v2di)
1      v8hi __builtin_ia32_vpcomneuw (v8hi, v8hi)
1      v8hi __builtin_ia32_vpcomnew (v8hi, v8hi)
1      v16qi __builtin_ia32_vpcomtrueb (v16qi, v16qi)
1      v4si __builtin_ia32_vpcomtrued (v4si, v4si)
1      v2di __builtin_ia32_vpcomtrueq (v2di, v2di)
1      v16qi __builtin_ia32_vpcomtrueub (v16qi, v16qi)
1      v4si __builtin_ia32_vpcomtrueud (v4si, v4si)
1      v2di __builtin_ia32_vpcomtrueuq (v2di, v2di)
1      v8hi __builtin_ia32_vpcomtrueuw (v8hi, v8hi)
1      v8hi __builtin_ia32_vpcomtruew (v8hi, v8hi)
1      v4si __builtin_ia32_vphaddbd (v16qi)
1      v2di __builtin_ia32_vphaddbq (v16qi)
1      v8hi __builtin_ia32_vphaddbw (v16qi)
1      v2di __builtin_ia32_vphadddq (v4si)
1      v4si __builtin_ia32_vphaddubd (v16qi)
1      v2di __builtin_ia32_vphaddubq (v16qi)
1      v8hi __builtin_ia32_vphaddubw (v16qi)
1      v2di __builtin_ia32_vphaddudq (v4si)
1      v4si __builtin_ia32_vphadduwd (v8hi)
1      v2di __builtin_ia32_vphadduwq (v8hi)
1      v4si __builtin_ia32_vphaddwd (v8hi)
1      v2di __builtin_ia32_vphaddwq (v8hi)
1      v8hi __builtin_ia32_vphsubbw (v16qi)
1      v2di __builtin_ia32_vphsubdq (v4si)
1      v4si __builtin_ia32_vphsubwd (v8hi)
1      v4si __builtin_ia32_vpmacsdd (v4si, v4si, v4si)
1      v2di __builtin_ia32_vpmacsdqh (v4si, v4si, v2di)
1      v2di __builtin_ia32_vpmacsdql (v4si, v4si, v2di)
1      v4si __builtin_ia32_vpmacssdd (v4si, v4si, v4si)
1      v2di __builtin_ia32_vpmacssdqh (v4si, v4si, v2di)
1      v2di __builtin_ia32_vpmacssdql (v4si, v4si, v2di)
1      v4si __builtin_ia32_vpmacsswd (v8hi, v8hi, v4si)
1      v8hi __builtin_ia32_vpmacssww (v8hi, v8hi, v8hi)
1      v4si __builtin_ia32_vpmacswd (v8hi, v8hi, v4si)
1      v8hi __builtin_ia32_vpmacsww (v8hi, v8hi, v8hi)
1      v4si __builtin_ia32_vpmadcsswd (v8hi, v8hi, v4si)
1      v4si __builtin_ia32_vpmadcswd (v8hi, v8hi, v4si)
1      v16qi __builtin_ia32_vpperm (v16qi, v16qi, v16qi)
1      v16qi __builtin_ia32_vprotb (v16qi, v16qi)
1      v4si __builtin_ia32_vprotd (v4si, v4si)
1      v2di __builtin_ia32_vprotq (v2di, v2di)
1      v8hi __builtin_ia32_vprotw (v8hi, v8hi)
1      v16qi __builtin_ia32_vpshab (v16qi, v16qi)
1      v4si __builtin_ia32_vpshad (v4si, v4si)
1      v2di __builtin_ia32_vpshaq (v2di, v2di)
1      v8hi __builtin_ia32_vpshaw (v8hi, v8hi)
1      v16qi __builtin_ia32_vpshlb (v16qi, v16qi)
1      v4si __builtin_ia32_vpshld (v4si, v4si)
1      v2di __builtin_ia32_vpshlq (v2di, v2di)
1      v8hi __builtin_ia32_vpshlw (v8hi, v8hi)
1 
1  The following built-in functions are available when '-mfma4' is used.
1 All of them generate the machine instruction that is part of the name.
1 
1      v2df __builtin_ia32_vfmaddpd (v2df, v2df, v2df)
1      v4sf __builtin_ia32_vfmaddps (v4sf, v4sf, v4sf)
1      v2df __builtin_ia32_vfmaddsd (v2df, v2df, v2df)
1      v4sf __builtin_ia32_vfmaddss (v4sf, v4sf, v4sf)
1      v2df __builtin_ia32_vfmsubpd (v2df, v2df, v2df)
1      v4sf __builtin_ia32_vfmsubps (v4sf, v4sf, v4sf)
1      v2df __builtin_ia32_vfmsubsd (v2df, v2df, v2df)
1      v4sf __builtin_ia32_vfmsubss (v4sf, v4sf, v4sf)
1      v2df __builtin_ia32_vfnmaddpd (v2df, v2df, v2df)
1      v4sf __builtin_ia32_vfnmaddps (v4sf, v4sf, v4sf)
1      v2df __builtin_ia32_vfnmaddsd (v2df, v2df, v2df)
1      v4sf __builtin_ia32_vfnmaddss (v4sf, v4sf, v4sf)
1      v2df __builtin_ia32_vfnmsubpd (v2df, v2df, v2df)
1      v4sf __builtin_ia32_vfnmsubps (v4sf, v4sf, v4sf)
1      v2df __builtin_ia32_vfnmsubsd (v2df, v2df, v2df)
1      v4sf __builtin_ia32_vfnmsubss (v4sf, v4sf, v4sf)
1      v2df __builtin_ia32_vfmaddsubpd  (v2df, v2df, v2df)
1      v4sf __builtin_ia32_vfmaddsubps  (v4sf, v4sf, v4sf)
1      v2df __builtin_ia32_vfmsubaddpd  (v2df, v2df, v2df)
1      v4sf __builtin_ia32_vfmsubaddps  (v4sf, v4sf, v4sf)
1      v4df __builtin_ia32_vfmaddpd256 (v4df, v4df, v4df)
1      v8sf __builtin_ia32_vfmaddps256 (v8sf, v8sf, v8sf)
1      v4df __builtin_ia32_vfmsubpd256 (v4df, v4df, v4df)
1      v8sf __builtin_ia32_vfmsubps256 (v8sf, v8sf, v8sf)
1      v4df __builtin_ia32_vfnmaddpd256 (v4df, v4df, v4df)
1      v8sf __builtin_ia32_vfnmaddps256 (v8sf, v8sf, v8sf)
1      v4df __builtin_ia32_vfnmsubpd256 (v4df, v4df, v4df)
1      v8sf __builtin_ia32_vfnmsubps256 (v8sf, v8sf, v8sf)
1      v4df __builtin_ia32_vfmaddsubpd256 (v4df, v4df, v4df)
1      v8sf __builtin_ia32_vfmaddsubps256 (v8sf, v8sf, v8sf)
1      v4df __builtin_ia32_vfmsubaddpd256 (v4df, v4df, v4df)
1      v8sf __builtin_ia32_vfmsubaddps256 (v8sf, v8sf, v8sf)
1 
1 
1  The following built-in functions are available when '-mlwp' is used.
1 
1      void __builtin_ia32_llwpcb16 (void *);
1      void __builtin_ia32_llwpcb32 (void *);
1      void __builtin_ia32_llwpcb64 (void *);
1      void * __builtin_ia32_llwpcb16 (void);
1      void * __builtin_ia32_llwpcb32 (void);
1      void * __builtin_ia32_llwpcb64 (void);
1      void __builtin_ia32_lwpval16 (unsigned short, unsigned int, unsigned short)
1      void __builtin_ia32_lwpval32 (unsigned int, unsigned int, unsigned int)
1      void __builtin_ia32_lwpval64 (unsigned __int64, unsigned int, unsigned int)
1      unsigned char __builtin_ia32_lwpins16 (unsigned short, unsigned int, unsigned short)
1      unsigned char __builtin_ia32_lwpins32 (unsigned int, unsigned int, unsigned int)
1      unsigned char __builtin_ia32_lwpins64 (unsigned __int64, unsigned int, unsigned int)
1 
1  The following built-in functions are available when '-mbmi' is used.
1 All of them generate the machine instruction that is part of the name.
1      unsigned int __builtin_ia32_bextr_u32(unsigned int, unsigned int);
1      unsigned long long __builtin_ia32_bextr_u64 (unsigned long long, unsigned long long);
1 
1  The following built-in functions are available when '-mbmi2' is used.
1 All of them generate the machine instruction that is part of the name.
1      unsigned int _bzhi_u32 (unsigned int, unsigned int)
1      unsigned int _pdep_u32 (unsigned int, unsigned int)
1      unsigned int _pext_u32 (unsigned int, unsigned int)
1      unsigned long long _bzhi_u64 (unsigned long long, unsigned long long)
1      unsigned long long _pdep_u64 (unsigned long long, unsigned long long)
1      unsigned long long _pext_u64 (unsigned long long, unsigned long long)
1 
1  The following built-in functions are available when '-mlzcnt' is used.
1 All of them generate the machine instruction that is part of the name.
1      unsigned short __builtin_ia32_lzcnt_u16(unsigned short);
1      unsigned int __builtin_ia32_lzcnt_u32(unsigned int);
1      unsigned long long __builtin_ia32_lzcnt_u64 (unsigned long long);
1 
1  The following built-in functions are available when '-mfxsr' is used.
1 All of them generate the machine instruction that is part of the name.
1      void __builtin_ia32_fxsave (void *)
1      void __builtin_ia32_fxrstor (void *)
1      void __builtin_ia32_fxsave64 (void *)
1      void __builtin_ia32_fxrstor64 (void *)
1 
1  The following built-in functions are available when '-mxsave' is used.
1 All of them generate the machine instruction that is part of the name.
1      void __builtin_ia32_xsave (void *, long long)
1      void __builtin_ia32_xrstor (void *, long long)
1      void __builtin_ia32_xsave64 (void *, long long)
1      void __builtin_ia32_xrstor64 (void *, long long)
1 
1  The following built-in functions are available when '-mxsaveopt' is
1 used.  All of them generate the machine instruction that is part of the
1 name.
1      void __builtin_ia32_xsaveopt (void *, long long)
1      void __builtin_ia32_xsaveopt64 (void *, long long)
1 
1  The following built-in functions are available when '-mtbm' is used.
1 Both of them generate the immediate form of the bextr machine
1 instruction.
1      unsigned int __builtin_ia32_bextri_u32 (unsigned int,
1                                              const unsigned int);
1      unsigned long long __builtin_ia32_bextri_u64 (unsigned long long,
1                                                    const unsigned long long);
1 
1  The following built-in functions are available when '-m3dnow' is used.
1 All of them generate the machine instruction that is part of the name.
1 
1      void __builtin_ia32_femms (void)
1      v8qi __builtin_ia32_pavgusb (v8qi, v8qi)
1      v2si __builtin_ia32_pf2id (v2sf)
1      v2sf __builtin_ia32_pfacc (v2sf, v2sf)
1      v2sf __builtin_ia32_pfadd (v2sf, v2sf)
1      v2si __builtin_ia32_pfcmpeq (v2sf, v2sf)
1      v2si __builtin_ia32_pfcmpge (v2sf, v2sf)
1      v2si __builtin_ia32_pfcmpgt (v2sf, v2sf)
1      v2sf __builtin_ia32_pfmax (v2sf, v2sf)
1      v2sf __builtin_ia32_pfmin (v2sf, v2sf)
1      v2sf __builtin_ia32_pfmul (v2sf, v2sf)
1      v2sf __builtin_ia32_pfrcp (v2sf)
1      v2sf __builtin_ia32_pfrcpit1 (v2sf, v2sf)
1      v2sf __builtin_ia32_pfrcpit2 (v2sf, v2sf)
1      v2sf __builtin_ia32_pfrsqrt (v2sf)
1      v2sf __builtin_ia32_pfsub (v2sf, v2sf)
1      v2sf __builtin_ia32_pfsubr (v2sf, v2sf)
1      v2sf __builtin_ia32_pi2fd (v2si)
1      v4hi __builtin_ia32_pmulhrw (v4hi, v4hi)
1 
1  The following built-in functions are available when '-m3dnowa' is used.
1 All of them generate the machine instruction that is part of the name.
1 
1      v2si __builtin_ia32_pf2iw (v2sf)
1      v2sf __builtin_ia32_pfnacc (v2sf, v2sf)
1      v2sf __builtin_ia32_pfpnacc (v2sf, v2sf)
1      v2sf __builtin_ia32_pi2fw (v2si)
1      v2sf __builtin_ia32_pswapdsf (v2sf)
1      v2si __builtin_ia32_pswapdsi (v2si)
1 
1  The following built-in functions are available when '-mrtm' is used
1 They are used for restricted transactional memory.  These are the
11 internal low level functions.  Normally the functions in ⇒x86
 transactional memory intrinsics should be used instead.
1 
1      int __builtin_ia32_xbegin ()
1      void __builtin_ia32_xend ()
1      void __builtin_ia32_xabort (status)
1      int __builtin_ia32_xtest ()
1 
1  The following built-in functions are available when '-mmwaitx' is used.
1 All of them generate the machine instruction that is part of the name.
1      void __builtin_ia32_monitorx (void *, unsigned int, unsigned int)
1      void __builtin_ia32_mwaitx (unsigned int, unsigned int, unsigned int)
1 
1  The following built-in functions are available when '-mclzero' is used.
1 All of them generate the machine instruction that is part of the name.
1      void __builtin_i32_clzero (void *)
1 
1  The following built-in functions are available when '-mpku' is used.
1 They generate reads and writes to PKRU.
1      void __builtin_ia32_wrpkru (unsigned int)
1      unsigned int __builtin_ia32_rdpkru ()
1 
1  The following built-in functions are available when '-mcet' or
1 '-mshstk' option is used.  They support shadow stack machine
1 instructions from Intel Control-flow Enforcement Technology (CET). Each
1 built-in function generates the machine instruction that is part of the
1 function's name.  These are the internal low-level functions.  Normally
1 the functions in ⇒x86 control-flow protection intrinsics should
1 be used instead.
1 
1      unsigned int __builtin_ia32_rdsspd (void)
1      unsigned long long __builtin_ia32_rdsspq (void)
1      void __builtin_ia32_incsspd (unsigned int)
1      void __builtin_ia32_incsspq (unsigned long long)
1      void __builtin_ia32_saveprevssp(void);
1      void __builtin_ia32_rstorssp(void *);
1      void __builtin_ia32_wrssd(unsigned int, void *);
1      void __builtin_ia32_wrssq(unsigned long long, void *);
1      void __builtin_ia32_wrussd(unsigned int, void *);
1      void __builtin_ia32_wrussq(unsigned long long, void *);
1      void __builtin_ia32_setssbsy(void);
1      void __builtin_ia32_clrssbsy(void *);
1