Thomas Harte
b8c6d4b153
Rips out my high-level ADB microcontroller protocol implementation.
...
Adds just enough that the main computer validates the ADB controller as present and talking.
2021-01-30 17:53:27 -05:00
Thomas Harte
f50e8b5106
If I'm going to maintain the max_address approach, & is 'correct'.
...
% +1 would be 'more correct', but I think this approach is probably misguided.
2021-01-27 18:31:11 -05:00
Thomas Harte
dcc2fe0990
Improves M50470 entry-point detection, adds test output.
2021-01-26 21:29:17 -05:00
Thomas Harte
56111c75ae
Makes first efforts towards disassembly.
2021-01-26 19:52:30 -05:00
Thomas Harte
fc4bda0047
Experimentally flipping interpretation of the output bit gives something closer to coherent.
2021-01-25 22:02:39 -05:00
Thomas Harte
c8beb59172
Attempts properly to track ADB bus activity.
...
Output is not yet a valid ADB stream. Work to do.
2021-01-25 17:43:22 -05:00
Thomas Harte
8789ffda15
Corrects performer storage, RMW/W confusion, implicit casts, port readback.
2021-01-24 22:30:42 -05:00
Thomas Harte
e8e604dc3c
Attempts to wire up M50470 and GLU.
...
Resulting in an unexpected interest in R15. Bugs to find, I guess.
2021-01-24 18:07:05 -05:00
Thomas Harte
57e0fdfadc
Ensures ADB microcontroller is clocked.
...
And runs at the 'correct' speed (i.e. modulo my instruction-by-instruction implementation).
2021-01-23 22:55:12 -05:00
Thomas Harte
ec0018df79
Routes in the ADB keyboard ROM. This should get as far as parsing.
2021-01-18 16:59:49 -05:00
Thomas Harte
12784a71e2
A stab in the dark: does the IOLC inhibit also affect vector fetches?
2020-12-29 20:53:56 -05:00
Thomas Harte
114d48b076
This register appears to be read/write.
2020-12-11 21:43:34 -05:00
Thomas Harte
159924dcc0
More clarity tweaks.
2020-12-10 22:47:11 -05:00
Thomas Harte
5d8f284757
Makes minor style improvements.
2020-12-10 22:11:53 -05:00
Thomas Harte
c978a95463
Increases asserts and adds a test.
...
Thereby discovers and fixes a problem with set_main_paging().
2020-12-10 21:49:23 -05:00
Thomas Harte
1928c955d9
Ensures safe startup of the Ensoniq.
2020-12-09 19:46:32 -05:00
Thomas Harte
049a78c667
Slightly restricts video flushing test.
2020-12-08 18:47:15 -05:00
Thomas Harte
65ca931e83
Throws in a new assert, against the unimplemented bit 0 of new video.
2020-12-06 20:26:24 -05:00
Thomas Harte
6cb71eb11b
This needs explicitly to be a bool for the table lookups to work.
2020-12-06 16:43:07 -05:00
Thomas Harte
43251193ee
The actual maximum line length is now 656.
2020-12-06 16:42:43 -05:00
Thomas Harte
55de98fb46
Adds a new statement of intent.
...
Now I need to try to decide whether I like my current all-in-one mapping for shadowing + paging, or whether it's better to split the things. I'm tending towards the latter at least until the functionality works.
2020-12-05 19:09:21 -05:00
Thomas Harte
6273ef8ba2
Adds means to force specific ROM 03 self tests.
2020-12-02 20:48:19 -05:00
Thomas Harte
3c6f09a898
Corrects super high-res aspect ratio and placement.
2020-12-02 20:47:26 -05:00
Thomas Harte
24fcb0c24b
Corrects video counter values.
...
The built-in speed test now passes.
2020-12-01 18:35:55 -05:00
Thomas Harte
03e2b6a265
Makes a slightly more rigorous attempt at discerning 1Mhz and 2.8Mhz operation.
2020-12-01 17:46:30 -05:00
Thomas Harte
ee22cf7ca1
Ensures that PAGE2 propagates from the state register to video.
2020-11-30 22:56:19 -05:00
Thomas Harte
187f507532
The soft switch is LCBANK2, not LCBANK1.
...
[This also jimmys the IIgs into always entering its extended self test, for now]
2020-11-30 22:35:51 -05:00
Thomas Harte
6000bd3a5e
Adds a bonus debugging assert. Let's see.
2020-11-30 18:15:02 -05:00
Thomas Harte
87069da3dd
Improves exposition, eliminates a couple of redundant map adjustments.
2020-11-30 18:07:03 -05:00
Thomas Harte
5cb4077576
Switches from modulo to and.
2020-11-30 17:47:57 -05:00
Thomas Harte
e9c7e0b9dd
Provisionally reverses meaning of language card RAM bank select.
2020-11-29 21:57:17 -05:00
Thomas Harte
35aa7612bb
Ensures that auxiliary/language-card soft switches don't trigger my assert.
2020-11-29 21:32:24 -05:00
Thomas Harte
acaa841822
Adds guaranteed trip to ROM for vector pulls.
2020-11-29 21:29:15 -05:00
Thomas Harte
46c1c9b5ee
CLRVBLINT calls it 3.75Hz. Which makes the arithmetic nicer.
2020-11-29 21:25:06 -05:00
Thomas Harte
4bdbca64b2
Takes a shot at the Mega II-style video interrupts.
2020-11-29 21:21:46 -05:00
Thomas Harte
11fe8ab6db
Corrects counter scales, adds a read for $c032.
...
Albeit that I have no idea what that's supposed to read as.
2020-11-29 20:08:59 -05:00
Thomas Harte
a9ce43d244
Takes a shot at the two video counter registers.
2020-11-29 19:57:35 -05:00
Thomas Harte
310282b7c9
Ensures extra_border_length always has a defined value.
2020-11-27 10:31:04 -05:00
Thomas Harte
af667c718e
Gets a bit more rigorous in remaining missing parts.
2020-11-26 22:36:32 -05:00
Thomas Harte
950f5b1691
Closes the loop on interrupts.
2020-11-26 19:56:42 -05:00
Thomas Harte
cbc0d848ad
Implements most of get_data.
2020-11-26 17:25:27 -05:00
Thomas Harte
f4d13d1f6f
Takes a run at the bus side of honouring Ensoniq sequence points.
2020-11-26 17:14:46 -05:00
Thomas Harte
6808ad6f5d
Adds a getter for the interrupt line.
2020-11-26 16:44:35 -05:00
Thomas Harte
7a8920ee38
Takes a stab at next_sequence_point.
2020-11-26 16:41:11 -05:00
Thomas Harte
4870506f6e
Implements skip_audio.
2020-11-26 16:24:48 -05:00
Thomas Harte
6f47f9d67c
Corrects placement of address bits.
2020-11-26 16:15:40 -05:00
Thomas Harte
8093f67173
Ensures video interrupts can't be missed by a suitably-timed access.
2020-11-26 16:11:03 -05:00
Thomas Harte
72884f37c3
It's still interrupt-deficient, but fills in additional Ensoniq audio generation.
2020-11-26 16:03:28 -05:00
Thomas Harte
8edb3fcd5f
Attempts to obey accumulator size in determining sample end.
2020-11-26 15:07:29 -05:00
Thomas Harte
fdd102df52
Resolves border colour resets.
2020-11-26 13:13:48 -05:00
Thomas Harte
03a893dc74
Quick refactor: this clearly isn't a VideoBase, it's the full implementation.
2020-11-26 12:54:20 -05:00
Thomas Harte
56de2512ae
Adds a further safety assert.
2020-11-25 23:34:30 -05:00
Thomas Harte
cdc2311045
Enables fuzzing, adds a definite no-op write.
2020-11-25 23:33:55 -05:00
Thomas Harte
eec27c3406
Reaches for marginally more coherent ADB data.
2020-11-25 17:34:00 -05:00
Thomas Harte
098a22aa95
Avoid out-of-bounds access of double_bytes.
2020-11-24 09:38:07 -05:00
Thomas Harte
7ede3d2b9e
Corrects collection of palettes other than palette 0.
2020-11-23 21:00:26 -05:00
Thomas Harte
e7160fe3c3
Rounds out the IIgs video hardware, bugs aside.
2020-11-23 20:58:32 -05:00
Thomas Harte
9d61665014
Attempts to add colour double [low/high] resolution output.
2020-11-23 19:05:18 -05:00
Thomas Harte
d2938ad7c8
Eliminate magic constants.
2020-11-23 18:36:44 -05:00
Thomas Harte
46f7ff07f7
Adds support for fill mode.
2020-11-22 21:55:21 -05:00
Thomas Harte
a34f294ba8
Pulls out commonalities re: NTSC colour, ensures mixed modes on a line works.
2020-11-22 21:29:40 -05:00
Thomas Harte
cd7d080b7a
Corrects low-resolution mode.
2020-11-22 20:52:42 -05:00
Thomas Harte
b0936b6ef4
Resolves high-resolution output.
...
Yet to optimise, but working.
2020-11-22 19:10:05 -05:00
Thomas Harte
8fae74f93e
Reintroduces delay bit, reverses phase.
...
There are stray columns of errors, but otherwise this is almost correct.
2020-11-22 11:06:14 -05:00
Thomas Harte
fca48e4b66
Makes hasty attempt to shift 'NTSC' in the most natural direction.
2020-11-21 23:39:58 -05:00
Thomas Harte
3b2ea37428
Slightly cleans up.
2020-11-21 22:53:26 -05:00
Thomas Harte
3cba3a5ac0
Corrects card mask test outside of bank $00.
2020-11-21 22:22:27 -05:00
Thomas Harte
4b024c5787
Starts to make some attempt at classic II modes.
2020-11-21 18:07:51 -05:00
Thomas Harte
4a42de4f18
Attempts to add 5.25" drive support to the IIgs.
...
I want to try some classic software.
2020-11-20 21:37:17 -05:00
Thomas Harte
d00e5d23ef
Takes a second shot at the MemoryWrite constructor complaint.
2020-11-19 22:28:10 -05:00
Thomas Harte
2c9ce116a2
Resolves various GCC-reported issues.
2020-11-19 22:21:20 -05:00
Thomas Harte
3512352c32
Attempt to use the most-significant relevant bits for sample position.
2020-11-19 22:13:09 -05:00
Thomas Harte
4d9372c52f
Also takes a stab at swap mode.
2020-11-19 21:56:49 -05:00
Thomas Harte
1d288b08b6
Attempts the two most basic forms of DOC output.
...
Sans interrupts. Or register reads of any variety.
2020-11-19 21:19:27 -05:00
Thomas Harte
f3c7c11772
Register writes now reach the audio thread.
2020-11-18 21:52:03 -05:00
Thomas Harte
4b9fe805e9
Sets up a queue to push memory writes onto the audio thread.
2020-11-18 21:40:56 -05:00
Thomas Harte
a7051e4e42
Strip this forceinline until I've satisfied myself that it works in declarations.
2020-11-18 21:40:25 -05:00
Thomas Harte
34794223b4
For now, at least, c800–cfff is always built-in ROM.
...
Otherwise I probably need to extend my c3 logic to cover the other built-in cards (?)
2020-11-18 19:49:45 -05:00
Thomas Harte
96cf617ee6
Advances slightly. I think I need a custom queue for RAM writes.
2020-11-18 19:48:53 -05:00
Thomas Harte
69dddf34b9
Adds bonus sanity check.
2020-11-18 19:47:56 -05:00
Thomas Harte
8f4597f742
Hacks in double text.
...
Actually, only one error: it should start half a column earlier. All 'double' output should. TODO.
2020-11-18 19:47:22 -05:00
Thomas Harte
98347cb1c3
Starts in the direction of audio support.
2020-11-18 18:39:11 -05:00
Thomas Harte
bb80e53021
Reduces frequency of video flushes.
2020-11-16 21:55:41 -05:00
Thomas Harte
952891d1b6
Improves commentary.
2020-11-16 21:46:35 -05:00
Thomas Harte
6dfad6a44b
Slightly reduces logging.
...
Hopefully soon I can tear the whole lot out.
2020-11-16 21:46:19 -05:00
Thomas Harte
e4c5bfdd5c
Takes a repeat shot at proper shadowing.
...
I think the Apple IIgs Technical Reference explains how these bits interact, and I just had inhibit_all_pages off all on my own.
2020-11-16 19:54:12 -05:00
Thomas Harte
9a55eb56ea
Attempts to provide saner sequence point behaviour.
2020-11-16 19:00:11 -05:00
Thomas Harte
9206ab5dc3
Adds notes to self; implements get_next_sequence_point for video, allowing per-line interrupts.
2020-11-16 14:42:50 -05:00
Thomas Harte
7e39550fc0
Attempts to make JustInTimeActor sequence-point aware.
...
With the objective of chopping out a lot of future boilerplate.
2020-11-15 21:58:18 -05:00
Thomas Harte
c3f5fbd300
Picks a better framing compromise for classic and new video modes.
2020-11-15 19:14:43 -05:00
Thomas Harte
1db713fec1
Attempts more meaningful super high-res pixel output.
...
With a timing hack as noted.
2020-11-15 18:36:24 -05:00
Thomas Harte
68ba73bee0
Ensures I get some sort of feedback for non-text modes.
2020-11-15 17:16:52 -05:00
Thomas Harte
cdacf280e1
After much extra logging, corrects destination bank for MVN and MVP.
2020-11-15 16:08:29 -05:00
Thomas Harte
1538a02e18
Better explains concern.
2020-11-14 19:27:20 -05:00
Thomas Harte
f9cec9a102
Attempts also to implement 1Mhz access costs.
...
Subject to TODO, and same observation as before: this is as to my current understanding only.
2020-11-14 19:23:01 -05:00
Thomas Harte
adda3d8f42
Attempts a 'full' model of 2.8Mhz access timing.
...
i.e. full to my current understanding.
2020-11-14 19:10:41 -05:00
Thomas Harte
ec3ff0da12
Steps towards proper calculation of time.
2020-11-14 18:39:16 -05:00
Thomas Harte
73c38b3b0d
Collapses nested conditionals.
2020-11-14 18:23:31 -05:00
Thomas Harte
edc8050b36
Adds activity indicators.
2020-11-14 18:00:06 -05:00
Thomas Harte
715a1b9cd6
Ensures safe shutdown.
2020-11-12 21:44:51 -05:00
