Revert "restore performance of integer shift operators."

This reverts commit da0dc58.

This was causing this failure, but I'm not sure why:

JULIA test/numbers
 * numbers             SplitVectorResult #0: 0x7faf84a4df10: v2i128 = insert_subvector 0x7faf8438d610, 0x7faf84338610, 0x7faf843a6110 [ID=0]

LLVM ERROR: Do not know how to split the result of this operator!

make[1]: *** [numbers] Error 1
make: *** [test-numbers] Error 2

base/operators.jl

@@ -111,9 +111,9 @@ const .≠ = .!=
 <<(x,y::Int32) = no_op_err("<<", typeof(x))
 >>(x,y::Int32) = no_op_err(">>", typeof(x))
 >>>(x,y::Int32) = no_op_err(">>>", typeof(x))
-<<(x,y::Integer) = x << (y % Int32)
->>(x,y::Integer) = x >> (y % Int32)
->>>(x,y::Integer) = x >>> (y % Int32)
+<<(x,y::Integer) = x << convert(Int32,y)
+>>(x,y::Integer) = x >> convert(Int32,y)
+>>>(x,y::Integer) = x >>> convert(Int32,y)
 
 # fallback div, fld, and cld implementations
 # NOTE: C89 fmod() and x87 FPREM implicitly provide truncating float division,