LLVM.Extra.Extension.X86

Some special operations on X86 processors. If you want to use them in algorithms you will always have to prepare an alternative implementation in terms of plain LLVM instructions. You will then run them with run and this driver function then selects the most advanced of both implementations. Functions that are written this way can be found in LLVM.Extra.Vector. Availability of extensions is checked with the CPUID instruction. However this does only work if you compile code for the host machine, that is cross compilation will fail! For cross compilation we would need access to the SubTarget detection of LLVM that is only available in the C++ interface in version 2.6.

Documentation

maxss :: T (V4Float -> V4Float -> CodeGenFunction r V4Float)

minss :: T (V4Float -> V4Float -> CodeGenFunction r V4Float)

maxps :: T (V4Float -> V4Float -> CodeGenFunction r V4Float)

minps :: T (V4Float -> V4Float -> CodeGenFunction r V4Float)

maxsd :: T (V2Double -> V2Double -> CodeGenFunction r V2Double)

minsd :: T (V2Double -> V2Double -> CodeGenFunction r V2Double)

maxpd :: T (V2Double -> V2Double -> CodeGenFunction r V2Double)

minpd :: T (V2Double -> V2Double -> CodeGenFunction r V2Double)

cmpss :: T (FPPredicate -> V4Float -> V4Float -> CodeGenFunction r V4Int32)

cmpps :: T (FPPredicate -> V4Float -> V4Float -> CodeGenFunction r V4Int32)

cmpsd :: T (FPPredicate -> V2Double -> V2Double -> CodeGenFunction r V2Int64)

cmppd :: T (FPPredicate -> V2Double -> V2Double -> CodeGenFunction r V2Int64)

cmpps256 :: T (FPPredicate -> V8Float -> V8Float -> CodeGenFunction r V8Int32)

cmppd256 :: T (FPPredicate -> V4Double -> V4Double -> CodeGenFunction r V4Int64)

pcmpgtb :: T (V16Int8 -> V16Int8 -> CodeGenFunction r V16Int8)

pcmpgtw :: T (V8Int16 -> V8Int16 -> CodeGenFunction r V8Int16)

pcmpgtd :: T (V4Int32 -> V4Int32 -> CodeGenFunction r V4Int32)

pcmpgtq :: T (V2Int64 -> V2Int64 -> CodeGenFunction r V2Int64)

pcmpugtb :: T (V16Word8 -> V16Word8 -> CodeGenFunction r V16Word8)

pcmpugtw :: T (V8Word16 -> V8Word16 -> CodeGenFunction r V8Word16)

pcmpugtd :: T (V4Word32 -> V4Word32 -> CodeGenFunction r V4Word32)

pcmpugtq :: T (V2Word64 -> V2Word64 -> CodeGenFunction r V2Word64)

pminsb :: T (V16Int8 -> V16Int8 -> CodeGenFunction r V16Int8)

pminsw :: T (V8Int16 -> V8Int16 -> CodeGenFunction r V8Int16)

pminsd :: T (V4Int32 -> V4Int32 -> CodeGenFunction r V4Int32)

pmaxsb :: T (V16Int8 -> V16Int8 -> CodeGenFunction r V16Int8)

pmaxsw :: T (V8Int16 -> V8Int16 -> CodeGenFunction r V8Int16)

pmaxsd :: T (V4Int32 -> V4Int32 -> CodeGenFunction r V4Int32)

pminub :: T (V16Word8 -> V16Word8 -> CodeGenFunction r V16Word8)

pminuw :: T (V8Word16 -> V8Word16 -> CodeGenFunction r V8Word16)

pminud :: T (V4Word32 -> V4Word32 -> CodeGenFunction r V4Word32)

pmaxub :: T (V16Word8 -> V16Word8 -> CodeGenFunction r V16Word8)

pmaxuw :: T (V8Word16 -> V8Word16 -> CodeGenFunction r V8Word16)

pmaxud :: T (V4Word32 -> V4Word32 -> CodeGenFunction r V4Word32)

pabsb :: T (V16Int8 -> CodeGenFunction r V16Int8)

pabsw :: T (V8Int16 -> CodeGenFunction r V8Int16)

pabsd :: T (V4Int32 -> CodeGenFunction r V4Int32)

pmuludq :: T (V4Word32 -> V4Word32 -> CodeGenFunction r V2Word64)

pmuldq :: T (V4Int32 -> V4Int32 -> CodeGenFunction r V2Int64)

pmulld :: T (V4Word32 -> V4Word32 -> CodeGenFunction r V4Word32)

cvtps2dq :: T (V4Float -> CodeGenFunction r V4Int32)

cvtpd2dq :: T (V2Double -> CodeGenFunction r V4Int32)

the upper two integers are set to zero, there is no instruction that converts to Int64

cvtdq2ps :: T (V4Int32 -> CodeGenFunction r V4Float)

cvtdq2pd :: T (V4Int32 -> CodeGenFunction r V2Double)

the upper two integers are ignored, there is no instruction that converts from Int64

ldmxcsr :: T (Value (Ptr Word32) -> CodeGenFunction r ())

MXCSR is not really supported by LLVM-2.6. LLVM does not know about the dependency of all floating point operations on this status register.

stmxcsr :: T (Value (Ptr Word32) -> CodeGenFunction r ())

withMXCSR :: Word32 -> T (CodeGenFunction r a -> CodeGenFunction r a)

haddps :: T (V4Float -> V4Float -> CodeGenFunction r V4Float)

haddpd :: T (V2Double -> V2Double -> CodeGenFunction r V2Double)

dpps :: T (V4Float -> V4Float -> Value Int8 -> CodeGenFunction r V4Float)

dppd :: T (V2Double -> V2Double -> Value Int8 -> CodeGenFunction r V2Double)

roundss :: T (V4Float -> Value Word32 -> CodeGenFunction r V4Float)

roundps :: T (V4Float -> Value Word32 -> CodeGenFunction r V4Float)

roundsd :: T (V2Double -> Value Word32 -> CodeGenFunction r V2Double)

roundpd :: T (V2Double -> Value Word32 -> CodeGenFunction r V2Double)

absss :: T (V4Float -> CodeGenFunction r V4Float)

abssd :: T (V2Double -> CodeGenFunction r V2Double)

absps :: Positive n => T (Value (Vector n Float) -> CodeGenFunction r (Value (Vector n Float)))

abspd :: Positive n => T (Value (Vector n Double) -> CodeGenFunction r (Value (Vector n Double)))

psllw128 :: T (V8Int16 -> V8Int16 -> CodeGenFunction r V8Int16)

pslld128 :: T (V4Int32 -> V4Int32 -> CodeGenFunction r V4Int32)

psllq128 :: T (V2Int64 -> V2Int64 -> CodeGenFunction r V2Int64)

psrlw128 :: T (V8Int16 -> V8Int16 -> CodeGenFunction r V8Int16)

psrld128 :: T (V4Int32 -> V4Int32 -> CodeGenFunction r V4Int32)

psrlq128 :: T (V2Int64 -> V2Int64 -> CodeGenFunction r V2Int64)

psraw128 :: T (V8Int16 -> V8Int16 -> CodeGenFunction r V8Int16)

psrad128 :: T (V4Int32 -> V4Int32 -> CodeGenFunction r V4Int32)

paddsb128 :: T (V16Int8 -> V16Int8 -> CodeGenFunction r V16Int8)

paddsw128 :: T (V8Int16 -> V8Int16 -> CodeGenFunction r V8Int16)

paddusb128 :: T (V16Word8 -> V16Word8 -> CodeGenFunction r V16Word8)

paddusw128 :: T (V8Word16 -> V8Word16 -> CodeGenFunction r V8Word16)

psubsb128 :: T (V16Int8 -> V16Int8 -> CodeGenFunction r V16Int8)

psubsw128 :: T (V8Int16 -> V8Int16 -> CodeGenFunction r V8Int16)

psubusb128 :: T (V16Word8 -> V16Word8 -> CodeGenFunction r V16Word8)

psubusw128 :: T (V8Word16 -> V8Word16 -> CodeGenFunction r V8Word16)