GCC provides intrinsics to access the SIMD instructions provided by the
MSA MIPS SIMD Architecture. The interface is made available by including
<msa.h> and using -mmsa -mhard-float -mfp64 -mnan=2008.
For each __builtin_msa_*, there is a shortened name of the intrinsic,
__msa_*.
MSA implements 128-bit wide vector registers, operating on 8-, 16-, 32- and
64-bit integer, 16- and 32-bit fixed-point, or 32- and 64-bit floating point
data elements. The following vectors typedefs are included in msa.h:
v16i8, a vector of sixteen signed 8-bit integers;
v16u8, a vector of sixteen unsigned 8-bit integers;
v8i16, a vector of eight signed 16-bit integers;
v8u16, a vector of eight unsigned 16-bit integers;
v4i32, a vector of four signed 32-bit integers;
v4u32, a vector of four unsigned 32-bit integers;
v2i64, a vector of two signed 64-bit integers;
v2u64, a vector of two unsigned 64-bit integers;
v4f32, a vector of four 32-bit floats;
v2f64, a vector of two 64-bit doubles.
Instructions and corresponding built-ins may have additional restrictions and/or input/output values manipulated:
imm0_1, an integer literal in range 0 to 1;
imm0_3, an integer literal in range 0 to 3;
imm0_7, an integer literal in range 0 to 7;
imm0_15, an integer literal in range 0 to 15;
imm0_31, an integer literal in range 0 to 31;
imm0_63, an integer literal in range 0 to 63;
imm0_255, an integer literal in range 0 to 255;
imm_n16_15, an integer literal in range -16 to 15;
imm_n512_511, an integer literal in range -512 to 511;
imm_n1024_1022, an integer literal in range -512 to 511 left
shifted by 1 bit, i.e., -1024, -1022, …, 1020, 1022;
imm_n2048_2044, an integer literal in range -512 to 511 left
shifted by 2 bits, i.e., -2048, -2044, …, 2040, 2044;
imm_n4096_4088, an integer literal in range -512 to 511 left
shifted by 3 bits, i.e., -4096, -4088, …, 4080, 4088;
imm1_4, an integer literal in range 1 to 4;
i32, i64, u32, u64, f32, f64, defined as follows:
{
typedef int i32;
#if __LONG_MAX__ == __LONG_LONG_MAX__
typedef long i64;
#else
typedef long long i64;
#endif
typedef unsigned int u32;
#if __LONG_MAX__ == __LONG_LONG_MAX__
typedef unsigned long u64;
#else
typedef unsigned long long u64;
#endif
typedef double f64;
typedef float f32;
}