[PATCH] D15857: [InstCombine] Defend against worst-case exponential execution time

[Joerg Sonnenberger via llvm-commits]

> I don't think we even need any other limit - this would limit the number of
> function invocations to O(bitwidth) as after that many ORs we're guaranteed
> to hit a bailout.

I thought it is O(bitwidth * funsize)? Where funsize is the number of
expressions in the functions? That can be pretty large, but it isn't
needed, the limit would certainly be optional.

Joerg