[Bug target/87555] There is no need for UNSPEC_FMADDSUB

crazylht at gmail dot com gcc-bugzilla@gcc.gnu.org
Fri Jun 18 06:34:08 GMT 2021


https://gcc.gnu.org/bugzilla/show_bug.cgi?id=87555

--- Comment #8 from Hongtao.liu <crazylht at gmail dot com> ---
(In reply to Hongtao.liu from comment #7)
> (In reply to Hongtao.liu from comment #6)
> > > So it there any canonical rtx for vec_merge? 
> > > (vec_merge (A B const_int 10) should abviously equal to (vec_merge B A
> > > const_int 5)
> > 
> > A and B here are 4-element vector.
> 
> Similar for 8-element vector C, D.
> (vec_merge (C D const_int 170 [0xaa])) equal to (vec_merge (D C const_int 85
> [0x55]))
> 
> 16-element vector [0xaaaa] and [0x5555]

And also 

(vec_merge (C D const_int 170 [0xaa])) equal to (vec_merge (D C const_int 85
[0x55]))
is equal to
 (vec_select (vec_concat D C) (0,9,2,11,4,13,6,15))


More information about the Gcc-bugs mailing list