Changes all the kernel files so that they are placed into const char*'s within each kernel file. Doxygen works but the object holding the kernel code is undocumented. All of the kernel structs are moved into the respective kernel file

functions through kernel_cl.func(inputs).

Adds kernel_cl across gpu functions. Getting seg_fault on tranpose test. Pretty sure it's due to how I'm making the compiler options for the kernel

split gpu checks in to separate files, fixed the header guards, added tests for check diagonal zeros