`lmul!` fails with Component Array on GPU

julia> lmul!(false, ComponentArray((; x=CUDA.ones(2))))
ERROR: Scalar indexing is disallowed.                                                                                                                    
Invocation of getindex resulted in scalar indexing of a GPU array.                                                                                       
This is typically caused by calling an iterating implementation of a method.          
Such implementations *do not* execute on the GPU, but very slowly on the CPU,
and therefore are only permitted from the REPL for prototyping purposes.                                                                                 
If you did intend to index this array, annotate the caller with @allowscalar.
Stacktrace:                                                                                                                                              
 [1] error(s::String)  
   @ Base ./error.jl:35                                                     
 [2] assertscalar(op::String)                                                                                                                            
   @ GPUArraysCore ~/.julia/packages/GPUArraysCore/uOYfN/src/GPUArraysCore.jl:103                                                                        
 [3] getindex                                                                                                                                            
   @ LinearAlgebra ~/.julia/packages/GPUArrays/dAUOE/src/host/indexing.jl:48 [inlined]
 [4] getindex                                                                                                                                            
   @ LinearAlgebra ~/.julia/dev/ComponentArrays/src/array_interface.jl:94 [inlined]
 [5] macro expansion                                                                                                                                     
   @ LinearAlgebra ~/.julia/juliaup/julia-1.10.0-beta3+0.x64.linux.gnu/share/julia/stdlib/v1.10/LinearAlgebra/src/generic.jl:221 [inlined]
 [6] macro expansion
   @ LinearAlgebra ./simdloop.jl:77 [inlined]
 [7] lmul!(s::Bool, X::ComponentVector{Float32, CuArray{Float32, 1, CUDA.Mem.DeviceBuffer}, Tuple{Axis{(x = 1:2,)}}})
   @ LinearAlgebra ~/.julia/juliaup/julia-1.10.0-beta3+0.x64.linux.gnu/share/julia/stdlib/v1.10/LinearAlgebra/src/generic.jl:220
 [8] top-level scope
   @ REPL[16]:1
 [9] top-level scope
   @ ~/.julia/packages/CUDA/nbRJk/src/initialization.jl:205

Reproduce the failing lmul!(false, ComponentArray((; x=CUDA.ones(2)))) call, then inspect src/array_interface.jl at the getindex entry shown in the stack trace and LinearAlgebra's generic lmul! implementation. Done means this GPU-backed ComponentArray operation completes without triggering scalar indexing.