[Bug tree-optimization/123123] `a != 0 ? 1 : a + 1` is not always optimized to 1

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=123123

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |matz at gcc dot gnu.org,
                   |                            |rguenth at gcc dot gnu.org

--- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> ---
Possibly for all constants though, a != CST ? CST +- CST2 : a +- CST2.  I'll
note this is conditional constant propagation of 'a' on the a == CST path
even for larger expressions than a +- CST2.  IIRC there's that SSA-w/o-a-CFG
from Burg(?) where the two conditional forms ( :? vs. if()) can be handled
the same.

It might be interesting to have a constant/copy/folding (aka forwprop) pass
on such an intermediate representation.