Add exhaustive tests for Float16 unary functions

This tests every `Float16` value for the following functions. `sin,cos,tan,asin,acos,atan,sinh,cosh,tanh,asinh,acosh,atanh,exp,exp2,exp10,expm1,log,log2,log10,log1p,sqrt,cbrt`
Specifically, I think it is reasonable that any Float16 implementation in Base be accurate to .5 ULP since Float16 already is a very unstable type. This test takes about 5 seconds on my computer, so before merging, it would be a good idea to make sure this is acceptable for CI.

test/float16.jl

@@ -4,6 +4,13 @@ using Test
 
 f = Float16(2.)
 g = Float16(1.)
+
+struct FloatIterator{T} end
+function Base.iterate(it::FloatIterator{T}, el=T(-Inf)) where T
+    return el == T(Inf) ? nothing : (el, nextfloat(el))
+end
+Base.length(::FloatIterator{Float16}) = 2^16 - 2^11
+
 @testset "comparisons" begin
     @test f >= g
     @test f > g
@@ -74,12 +81,12 @@ end
 @testset "unary ops" begin
     @test -f === Float16(-2.)
     @test Float16(0.5f0)^2 ≈ Float16(0.5f0^2)
-    @test sin(f) ≈ sin(2f0)
-    @test log10(Float16(100)) == Float16(2.0)
-    @test sin(ComplexF16(f)) ≈ sin(complex(2f0))
-
-    # no domain error is thrown for negative values
-    @test cbrt(Float16(-1.0)) == -1.0
+    for func in (sin,cos,tan,asin,acos,atan,sinh,cosh,tanh,asinh,acosh,atanh,
+                 exp,exp2,exp10,expm1,log,log2,log10,log1p,sqrt,cbrt)
+        for x in FloatIterator{Float16}()
+            @test try func(x) === Float16(func(widen(x))) catch; true end
+        end
+    end
 end
 @testset "binary ops" begin
     @test f+g === Float16(3f0)