/*
 * [/nfs/mediocris/home/mid/dspc/sim/TODO.txt     Fri Jul  9 12:08:15 UTC 1999]
 *
 * Your average TODO list...  in the form of a big C comment...
 *
 * (Copied from a random OfficeMax-brand Canary notepad I found laying around)
 *
 *
 *  
 *  * Paged DM/PM accesses -- DM done, PM not started
 *
 *  * Stacks...
 *      * PC Stack -- Should be working fine although support for the 
 *                    TOPPCSTACK attrocity has yet to commence (pg 3-26/27)
 *      * Stat{e,us} Stack -- Should be good (sorry about the naming confusion)
 *      * Loop/Counter Stacks -- Not yet.  
 *
 *  * Zero-overhead looping (ie, DO UNTIL) -- not started (I need to play
 *                                            with this a bit before I 
 *                                            implement...looks to be 
 *                                            a rather confusing feature) 
 *
 *  * Indirect Jumps/Calls -- this is a one-liner.  It just needs to be done
 * 
 *  * ALU conditions (ASTAT) should be done, although since few of the
 *                           ALU operations really work, many of them
 *                           are hardwired as false right now, some work,
 *                           and the rest work but maybe not correctly.
 *
 *  * The ALU operations.... also need to start storing things in
 *                           Real 2181 formats (twos complement, etc)
 *
 *  * Idle(n) instruction isn't implemented the Right Way.  
 * 
 *  * A whole load of other instructions (I can't think of ANY multifunctions
 *                                        that are even started, also many
 *                                        registers dont have pointers, as
 *                                        well as the MR and other oddball
 *                                        spaces, along with autoincrement
 *                                        writes and related ops)
 *  
 *  * BDMA -- transfers with BTYPE of 0x02 or 0x03 are not supported
 *
 *  * Soft Reboot may not be perfect -- I'm not real clear on what should
 *                                      get reset and what shouldn't, even
 *                                      though theres that table on pg 9-10.
 *
 *  * SPORTx's -- We should have some sort of external stream source here.
 *                Or at least make it look like we do.  As Dan has pointed
 *                out, Real Time connection to external devices is impossible
 *                (not just barely impossible, but impossible to the point
 *                of unimaginable and heathenist).  We need to be able to at
 *                least emulate the tick coming from the codec.
 *
 *  * IO() space -- Not real sure what we need here.
 *
 *  * Timer -- We probably don't need this at all.  But it might be amusing
 *             to implement anyway.
 *
 *  
 *
 * So it basically comes down to this...(ADI's ADSP-2100 Family User's Manual)
 * 
 *       Chapter  2 -- Computational Units -- Not even close.
 *       Chapter  3 -- Program Control -- Everything done but DO UNTIL
 *       Chapter  4 -- Data Transfer -- Basics done, need the details
 *       Chapter  5 -- Serial Ports -- Havent even looked at it
 *       Chapter  6 -- Timer -- Same
 *       Chapter  7 -- Host Interface Port -- Not applicable (eg, no way)
 *       Chapter  8 -- Analog Interface -- Don't have a 21msp58/59
 *       Chapter  9 -- System Interface -- Mostly non-applicable (what is 
 *                                         applicable should be done)
 *       Chapter 10 -- Memory Interface -- Need PM Paging, etc 
 *       Chapter 11 -- DMA -- BDMA should be done (except above limitations)
 *                            IDMA is not applicable, thank gode.
 *      
 * And the rest of them are useless chapters.  Except the half-complete,
 * partially incorrect Appendix A (Instruction Coding).  App B (Division
 * Exceptions) will be important for when I finally do the ALU ops correctly.
 * App C (Numeric Formats) is stuff I should've been doing all along, but 
 * forgot about.  Appendix D is just an interrupt vector table.  App E
 * is yet another half-assed attempt at a half-complete register table.
 * (Probably the most technically correct thing in this manual is what
 * comes after App E...the Index. Though I've never used it to find out.)
 *
 */