[Bug middle-end/85237] missed optimisation opportunity for large/negative shifts

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=85237

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2018-04-06
          Component|rtl-optimization            |middle-end
     Ever confirmed|0                           |1

--- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> ---
Yeah, we don't really take advantage of UB in shifts.  Instead we really treat
them implementation-defined aka 100 >> 1000 is zero, likewise we do (did?)
treat 100 >> -2 as 100 << 2.

Note historically RTL doesn't have any UB, it matches target implementation
behavior for shifts (see that stupid SHIFT_COUNT_TRUNCATED thing...).