sort returns nothing

[xaellison]

Describe the bug

sort returns nothing. Introduced by #1217

To reproduce

The Minimal Working Example (MWE) for this bug:

julia> begin
       a = rand(10000)
       c=CuArray(a)
       s = sort(c)
       typeof(s)
       end
Nothing

Manifest.toml

Paste your Manifest.toml here, or accurately describe which version of CUDA.jl and its dependencies (GPUArrays.jl, GPUCompiler.jl, LLVM.jl) you are using.

Expected behavior

Should return an array of same type as c the same way Base does.

julia> typeof(sort(a))
Vector{Float64} (alias for Array{Float64, 1})

Version info

Details on Julia:

Currently checked out head 9636915.

julia> versioninfo()
Julia Version 1.7.0
Commit 3bf9d17731 (2021-11-30 12:12 UTC)
Platform Info:
  OS: Windows (x86_64-w64-mingw32)
  CPU: Intel(R) Core(TM) i7-9700K CPU @ 3.60GHz
  WORD_SIZE: 64
  LIBM: libopenlibm
  LLVM: libLLVM-12.0.1 (ORCJIT, skylake)
Environment:
  JULIA_CUDA_NSYS = C:\Program Files\NVIDIA Corporation\Nsight Systems 2021.1.1\target-windows-x64\nsys.exe
  JULIA_CUDA_USE_BINARYBUILDER = false
  JULIA_CUDA_USE_COMPAT = false
  JULIA_NUM_THREADS = 3

Details on CUDA:

 CUDA.versioninfo()
CUDA toolkit 10.1, local installation
NVIDIA driver 496.49.0, for CUDA 11.5
CUDA driver 11.5

Libraries:
- CUBLAS: 10.2.1
- CURAND: 10.2.3
- CUFFT: 10.4.0
- CUSOLVER: 10.2.0
- CUSPARSE: 10.3.0
- CUPTI: 14.0.0
- NVML: 11.0.0+496.49
- CUDNN: missing
- CUTENSOR: missing

Toolchain:
- Julia: 1.7.0
- LLVM: 12.0.1
- PTX ISA support: 3.2, 4.0, 4.1, 4.2, 4.3, 5.0, 6.0, 6.1, 6.3, 6.4
- Device capability support: sm_35, sm_37, sm_50, sm_52, sm_53, sm_60, sm_61, sm_62, sm_70, sm_72, sm_75

Environment:
- JULIA_CUDA_NSYS: C:\Program Files\NVIDIA Corporation\Nsight Systems 2021.1.1\target-windows-x64\nsys.exe
- JULIA_CUDA_USE_BINARYBUILDER: false
- JULIA_CUDA_USE_COMPAT: false

1 device:
  0: NVIDIA GeForce RTX 2070 (sm_75, 5.424 GiB / 8.000 GiB available)

Additional context

Add any other context about the problem here.