add chainrules for r2r, dct

[vpuri3]

No description provided.

[vpuri3]

how is the gradient computed for plan_dct, if there's not rrule dor dct??

using LinearAlgebra, FFTW, Zygote

x = rand(4)
C = plan_dct(x)

f(x) = C \ (C * x) |> norm
g(x) = x |> dct |> idct |> norm
h(x) = plan_dct(x) \ (plan_dct(x) * x) |> norm

@show Zygote.gradient(f, x) # ([0.7499995699183157, 0.5170775887690442, 0.3522881598130941, 0.2145331321046639],)
@show Zygote.gradient(g, x) # errors
@show Zygote.gradient(h, x) # errors

error message:

julia> Zygote.gradient(f, x)                                                          
ERROR: Compiling Tuple{Type{FFTW.r2rFFTWPlan{Float64, Any, false, 1}}, Vector{Float64}, FFTW.FakeArray{Float64, 1}, UnitRange{Int64}, Int64, UInt32, Float64}: try/catch is n
ot supported.                                                                         
Refer to the Zygote documentation for fixes.                                          
https://fluxml.ai/Zygote.jl/latest/limitations