[InstCombine] Missed optimization for trunc (Pow2 >> X) to i1

[nikic]

https://alive2.llvm.org/ce/z/jf8swi

define i1 @src(i8 %x) {
  %lshr = lshr i8 4, %x
  %trunc = trunc i8 %lshr to i1
  ret i1 %trunc
}

define i1 @tgt(i8 %x) {
  %cmp = icmp eq i8 %x, 2
  ret i1 %cmp
}

[nikic]

cc @andjo403 as you're working on trunc i1 opts.

[nikic]

There are also generalizations beyond the pow2 pattern, for example this: https://alive2.llvm.org/ce/z/Zr98JP

[kper]

I would like to work on that :)