[Bug middle-end/82905] vector shift forced to 32 bits

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82905

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
          Component|tree-optimization           |middle-end

--- Comment #2 from Richard Biener <rguenth at gcc dot gnu.org> ---
We lack the narrowing pattern, only some FEs do that themselves.  There's a dup
for avr somewhere.