vec_permx
Purpose
Performs a partial permute of the first two arguments, which form an aligned 32-byte section of an emulated vector up to 256 bytes wide, using the partial permute control vector in the third argument. The fourth argument identifies which 32-byte section of the emulated vector is contained in the first two arguments.
Syntax
d=vec_permx(a,b,c,e)Result and argument types
The following table describes the types of the returned value and the function arguments.
| d | a | b | c | e |
|---|---|---|---|---|
| vector signed char | vector signed char | vector signed char | vector unsigned char | const int |
| vector unsigned char | vector unsigned char | vector unsigned char | vector unsigned char | const int |
| vector signed short | vector signed short | vector signed short | vector unsigned char | const int |
| vector unsigned short | vector unsigned short | vector unsigned short | vector unsigned char | const int |
| vector signed int | vector signed int | vector signed int | vector unsigned char | const int |
| vector unsigned int | vector unsigned int | vector unsigned int | vector unsigned char | const int |
| vector signed long long | vector signed long long | vector signed long long | vector unsigned char | const int |
| vector unsigned long long | vector unsigned long long | vector unsigned long long | vector unsigned char | const int |
| vector float | vector float | vector float | vector unsigned char | const int |
| vector double | vector double | vector double | vector unsigned char | const int |
e is constrained to
values of 0 to 7, inclusive.Result value
Let s be the concatenation of a and b. For each integer value i from 0 to 15, do the following: let
j be the contents of bits 3 through 7 of byte element i of
c. If e is equal to the contents of bits 0 through 2 of byte
element i of c, the contents of byte element j of s are placed
into byte element i of d. Otherwise, the contents of byte element i of
d are set to zero.
This built-in function can be used to emulate permutes on vectors up to 256 bytes in length, and can also be used to perform a parallel table lookup on up to 256-bytes tables.