Lack-of-inlining bug with type that has extra parameter

[timholy]

This is on 0.5.0-rc3:

julia> versioninfo()
Julia Version 0.5.0-rc3+0
Commit e6f843b* (2016-08-22 23:43 UTC)
Platform Info:
  System: Linux (x86_64-linux-gnu)
  CPU: Intel(R) Core(TM) i7-5500U CPU @ 2.40GHz
  WORD_SIZE: 64
  BLAS: libopenblas (USE64BITINT DYNAMIC_ARCH NO_AFFINITY Haswell)
  LAPACK: libopenblas64_
  LIBM: libopenlibm
  LLVM: libLLVM-3.7.1 (ORCJIT, broadwell)

Suppose we have some types that we want to scan for in a tuple:

immutable Triggs end
immutable TriggsContainer{T,V}
    obj::V
end

hastriggs(kernel::Tuple) = _hastriggs(false, kernel...)
_hastriggs(ret) = ret
@inline _hastriggs(ret, kern, kernel...) = _hastriggs(ret, kernel...)
@inline _hastriggs{T,V<:Triggs}(ret, kern::Union{V,TriggsContainer{T,V}}, kernel...) = true

If you run julia normally, this works fine:

tim@diva:/tmp$ julia -q
julia> include("inlining.jl")
_hastriggs (generic function with 3 methods)

julia> hastriggs((rand(2), rand(2)))
false

julia> hastriggs((rand(2), Triggs()))
true

julia> hastriggs((Triggs(), rand(2)))
true

But woe be to the person who dares turn off inlining:

tim@diva:/tmp$ julia -q --inline=no
julia> include("inlining.jl")
_hastriggs (generic function with 3 methods)

julia> hastriggs((rand(2), rand(2)))
false

julia> hastriggs((rand(2), Triggs()))
false

julia> hastriggs((Triggs(), rand(2)))
false

If you get rid of the (here useless, but not in my real application)T parameter, then it works correctly with or without inlining.

Interestingly, with inlining off we have:

julia> @which _hastriggs(false, Triggs())
ERROR: no unique matching method for the specified argument types
 in error(::String) at ./error.jl:21
 in which(::Any, ::Any) at ./reflection.jl:442
 in eval(::Module, ::Any) at ./boot.jl:234

and it seems that it likes either one:

julia> methods(_hastriggs, Tuple{Bool, Triggs})
#2 methods for generic function "_hastriggs":
_hastriggs{T,V<:Triggs}(ret, kern::Union{TriggsContainer{T,V},V}, kernel...) at /tmp/inlining.jl:9
_hastriggs(ret, kern, kernel...) at /tmp/inlining.jl:8

[timholy]

Interestingly, the @which component of this persists even when inlining is on.

[vtjnash]

Looks like an inliner bug, since the correct answer is false in all of the examples given.

[timholy]

Am I blind? (Probably, yes.) Note that Union includes V<:Triggs as one of the options.

[vtjnash]

yes, but it also has T as one of the requirements, and that can't be satisfied by Triggs

[timholy]

So this is really the same as #18110.

[vtjnash]

similar, yes. I have a patch for this bug and am just writing the test.