mul! falls back to generic implementation

[rveltz]

Describe the bug
Calling mul! falls back to generic implementation unlike calling rmul!

To Reproduce
The Minimal Working Example (MWE) for this bug:

using CuArrays, LinearAlgebra
CuArrays.allowscalar(false)
rhs = rand(1000000)
rhs_d = cu(rhs)
#this works:
rmul!(rhs_d,1.0)

#this fails:
mul!(rhs_d,rhs_d,1.0)

Environment details (please complete this section)
Details on Julia:

# please post the output of:
versioninfo()

julia> versioninfo()
Julia Version 1.0.3
Commit 099e826241* (2018-12-18 01:34 UTC)
Platform Info:
OS: Linux (x86_64-redhat-linux)
CPU: Intel(R) Xeon(R) CPU E5-2637 v4 @ 3.50GHz
WORD_SIZE: 64
LIBM: libopenlibm

[MikeInnes]

I think mul! slightly predates CuArrays so there's probably just a dispatch case that's missed, should be easy to hook up.

[maleadt]

This works now.