Batch and SVM (#2)

* toggle interleaving using cmd line args, svm init

* modified tests based on interleaving args

* enabled fast emulation

* Working svm host code

* Buffer locations attribute in fft3d kernel

* working batch fft3d svm

* working 3D FFT with bitrev diagonal transpose

* loop coalesced fft kernels

* fft3d: single loop dbl buffered fetch

* fft3d: fetch_transpose working

* Buffer location in stores

- buffer location to enable DDR stores in SVM,

- print iteration wise performance measurments

- modify kernel names in host to match the kernel

* FFT2d: bram dbl buf and single loop

* correct svm buffer flags, init output buffer, fft1d iter

* clFinish stms in DDR buffer reads

* FFT3D: separated svm and ddr

* FFT3D: Working opt bram transpose, bool instead of int

* FFT3D: renamed bram transpose files, dependencies

* CI: fixed host executable names

* FFT1d: working svm

* CI: corrected build, artifact paths

* Test: fix path to 2d bram

* CI: Adding release debug to restrict prints

* Test: fixed path for 3d FFT DDR

* modified verification function

* Working DDR batch of 2

* working DDR batch for all

* Split to individual files

* Hw Counters and const inp

* Reordered host kernel calls, removed channel depth

* batched fft3d with safelen

* fixed host code for svm, svm batch

* working svm batch

* no verify param

* free forgotten

* porting example to cpp, using cxxopts, cmake files to folder

* removed argparse, refactored timing

* verify bitrev fft1d, using opt bram fft3d

* working ddr batch

* streamlined synth path

* identical fft3d ddr and svm kernels

* updated tests

* removed extern modules

* removed common cl files to common dir

* gitlab ci update

* updated readme

* updated changelog

* fixed args in fft1d, fft2d

* fixed ddr batch and renamed queues

* added additional variants to application selection

* expressive emulation and report generation

* cleaned up application files

* fixed emulation dependency in test

* draft userguide

* fixed gitlab ci

* added power scripts

* minor change

* updated changelog

* ci: test fft1d

* updated ci with 2d, 3d variants

* removed fft3d bram due to lack of cpu resources

* cmake build type selection

* removed caching cmake variables for sdk and bsp

* modified readme and updated userguide

Authored-by: Arjun Ramaswami <arjun.ramaswami@uni-paderborn.de>

v0.1

Time the entire API call