How const_double expression should be used?

Peryt, Sebastian
Thu Jun 1 06:28:00 GMT 2017


I would like to ask how properly const_double expression should be used inside const_vector? Currently I want to write RTL with 
vector consisting of 2 DF elements that are zeros, but documentation regarding const_double is not so clear for me.

>From const_vector description I know something like this can be done with const_int:

(const_vector:V2DF [(const_int 0) (const_int 0)])

but, const_double has different syntax, than const_int: (const_double:m i0 i1 ...). I don't understand what i0, i1 are and there are only two sentences 
mentioning them:
If m is VOIDmode, the bits of the value are stored in i0 and i1. i0 is customarily
accessed with the macro CONST_DOUBLE_LOW and i1 with CONST_DOUBLE_HIGH.

I tried to find any example in current GCC sources, but apparently there isn't one. I would greatly appreciate explanation on how to use const_double in such case.


More information about the Gcc-help mailing list