Fix multiply at the input limits

Author: omus

src/FixedPointDecimals.jl

@@ -128,8 +128,8 @@ end
 # TODO: can we use floating point to speed this up? after we build a
 # correctness test suite.
 function *{T, f}(x::FD{T, f}, y::FD{T, f})
-    powt = T(10)^f
-    quotient, remainder = fldmod(Base.widemul(x.i, y.i), powt)
+    powt = coefficient(FD{T,f})
+    quotient, remainder = fldmod(widemul(x.i, y.i), powt)
     reinterpret(FD{T, f}, _round_to_even(quotient, remainder, powt))
 end
 

test/runtests.jl

@@ -229,6 +229,27 @@ end
         @test 20 * FD{Int8,1}(0.1) == FD{Int8,1}(2.0)
         @test FD{Int8,1}(0.1) * 20 == FD{Int8,1}(2.0)
     end
+
+    @testset "limits" begin
+        for T in CONTAINER_TYPES
+            x = FixedPointDecimals.max_exp10(T)
+            f = x + 1
+
+            @eval begin
+                scalar = reinterpret(FD{$T,$f}, $T(10)^$x)  # 0.1
+
+                # Since multiply will round the result we'll make sure our value does not
+                # always rounds down.
+                max_int = typemax($T) - (typemax($T) % 10)
+                min_int = typemin($T) - (typemin($T) % 10)
+
+                @test reinterpret(FD{$T,$f}, max_int) * scalar ==
+                    reinterpret(FD{$T,$f}, div(max_int, 10))
+                @test reinterpret(FD{$T,$f}, min_int) * scalar ==
+                    reinterpret(FD{$T,$f}, div(min_int, 10))
+            end
+        end
+    end
 end
 
 @testset "division" begin