Use rmul! instead of broadcast for kscal!

[amontoison]

rmul! will dispatch to BLAS routines on GPUs:

• https://github.com/JuliaGPU/CUDA.jl/blob/master/lib/cublas/linalg.jl#L9
• https://github.com/JuliaGPU/AMDGPU.jl/blob/master/src/blas/highlevel.jl#L7
• https://github.com/JuliaGPU/oneAPI.jl/blob/master/lib/mkl/linalg.jl#L14

It will be also faster for CPU types, the code is

function rmul!(X::AbstractArray, s::Number)
    @simd for I in eachindex(X)
        @inbounds X[I] *= s
    end
    X
end

in the source code of Julia (julia/stdlib/v1.11/LinearAlgebra/src/generic.jl).

[github-actions]

Package name	latest	stable
CaNNOLeS.jl
DCISolver.jl
FletcherPenaltySolver.jl
JSOSolvers.jl
LLSModels.jl
LinearSolve.jl
Percival.jl
RipQP.jl