1
0
mirror of https://github.com/TomHarte/CLK.git synced 2024-11-03 08:05:40 +00:00
Commit Graph

8510 Commits

Author SHA1 Message Date
Thomas Harte
c5e8b547af Captures the attach flag and observes activation rule. 2021-10-22 11:21:58 -07:00
Thomas Harte
e67de90ad0 Starts to bring sprites inside DMADevice orthodoxy. 2021-10-21 21:57:46 -07:00
Thomas Harte
c3c84c88a1 Switch to ahead-of-time planar to chunky conversion. 2021-10-21 20:48:57 -07:00
Thomas Harte
0dc9c4cee1 Undo hard-coding of fetch window. 2021-10-19 15:18:39 -07:00
Thomas Harte
544c137cb0 Add updated intel. 2021-10-16 13:30:56 -07:00
Thomas Harte
b312a61a81 Add two dummy reads. 2021-10-16 13:30:45 -07:00
Thomas Harte
4917556a99 The shift goes the other way in descending mode. 2021-10-16 11:09:40 -07:00
Thomas Harte
15ed4a0d09 Introduce failing test case for sector decoding. 2021-10-16 10:48:32 -07:00
Thomas Harte
aa6b0f07b7 Correct filename. 2021-10-16 05:37:46 -07:00
Thomas Harte
d9d20d9d30 Walk back slightly. 2021-10-14 18:02:58 -07:00
Thomas Harte
689bfbbdb3 Be overt in initialiser list. 2021-10-14 16:57:26 -07:00
Thomas Harte
e27a10bde4 Simplify control flow. 2021-10-14 16:47:18 -07:00
Thomas Harte
253a199f27 Fire sync-match interrupt upon any match. 2021-10-14 16:36:17 -07:00
Thomas Harte
61e5702520 Remove dead TODO. 2021-10-14 16:09:11 -07:00
Thomas Harte
b12c640807 Makes drives non-copyable.
To avoid error in the future.
2021-10-14 12:37:55 -07:00
Thomas Harte
9be23ecc34 Add end-of-Blit interrupt.
Along with a slightly easier path for posting interrupts, in C++ compilation unit terms.
2021-10-13 15:09:19 -07:00
Thomas Harte
8960f471a0 Use unspread_bits for FM and MFM decoding. 2021-10-12 15:18:50 -07:00
Thomas Harte
955cb6411c Factor out bit spreading.
(And do a better job of it)
2021-10-12 14:49:01 -07:00
Thomas Harte
fc4ca4f8e3 I don't think there are sync words at the start of the track. 2021-10-12 10:38:15 -07:00
Thomas Harte
eec068914e Slightly improve logging. 2021-10-11 18:05:57 -07:00
Thomas Harte
a1f02d0cd8 Add track padding. 2021-10-11 18:05:37 -07:00
Thomas Harte
39b8285ba5 Trust the HRM on step bit, but catch rising edge. 2021-10-11 07:42:42 -07:00
Thomas Harte
7733fef3bd DSKLEN has to be written twice. 2021-10-11 06:16:01 -07:00
Thomas Harte
6acddfdb98 Add the sync match interrupt.
Albeit that it doesn't yet unblock disk DMA.
2021-10-11 03:37:56 -07:00
Thomas Harte
ec3d5c0b32 Increase maximum number of activity LEDs to eight. 2021-10-10 18:37:33 -07:00
Thomas Harte
99492c2ec2 Further tweak logging. 2021-10-10 18:19:50 -07:00
Thomas Harte
addf9f9af4 Moves block byte writes into Storage::Encodings::MFM::Encoder. 2021-10-10 16:06:51 -07:00
Thomas Harte
846b505d27 Reduce logging; disk data probably isn't the immediate obstacle. 2021-10-10 13:04:10 -07:00
Thomas Harte
c4cfcfab8e Checksums appear to be calculated as 32-bit quantities. 2021-10-10 12:58:10 -07:00
Thomas Harte
5e083426c5 Takes another run at checksums.
It turns out I'd read entirely the wrong section of the ADF FAQ. Am now trying to piece things together from various EAB threads.
2021-10-10 11:47:48 -07:00
Thomas Harte
8d43b4a98d Expands Disk DMA access window. 2021-10-10 11:47:02 -07:00
Thomas Harte
aeaea073c6 Switch both: (i) which bits are odd/even; and (ii) nibble ordering. 2021-10-09 13:45:19 -07:00
Thomas Harte
6b0dd19442 Name file appropriately: the logo comes from Kickstart. 2021-10-09 08:02:15 -07:00
Thomas Harte
9336ffe216 Take a stab at index-hole sync. 2021-10-09 08:01:02 -07:00
Thomas Harte
eb157f15f3 Adds index hole interrupt. 2021-10-09 04:08:59 -07:00
Thomas Harte
d6e2a3f425 Make a first attempt to spool into RAM. 2021-10-08 18:11:47 -07:00
Thomas Harte
b47ca13ed3 Push disk data onwards. 2021-10-08 17:18:11 -07:00
Thomas Harte
67546c4d6e Per the HRM, the index hole is connected to CIA B, potentially to raise an interrupt. 2021-10-08 17:12:37 -07:00
Thomas Harte
f72deb0a5c Correct RDY position. 2021-10-08 04:32:13 -07:00
Thomas Harte
616ccbb878 Correct ID bit placement, multiplex with motor state.
The latter per my reading of http://www.primrosebank.net/computers/amiga/upgrades/amiga_upgrades_storage_fdis.htm
2021-10-08 04:05:57 -07:00
Thomas Harte
5899af0038 Starts accumulating disk data. 2021-10-07 05:11:32 -07:00
Thomas Harte
ed303310bb Spell out slightly more; this makes debugging a touch easier. 2021-10-06 13:40:48 -07:00
Thomas Harte
33ff4f3b5c Eliminate drive copies. 2021-10-06 13:40:28 -07:00
Thomas Harte
20bad38d42 Add drive activity lights. 2021-10-06 04:54:40 -07:00
Thomas Harte
92a07398cd I think CHNG works the other way around. 2021-10-06 04:47:52 -07:00
Thomas Harte
ce8f782577 Corrects meaning of IBM-style RDY. 2021-10-06 04:42:44 -07:00
Thomas Harte
e961d0b4a3 Switch RDY type. 2021-10-06 04:41:09 -07:00
Thomas Harte
2253ff656a Adds route for inserting disks. 2021-10-05 16:12:30 -07:00
Thomas Harte
18631399ad Attempts to clock the disk controller. 2021-10-05 15:38:56 -07:00
Thomas Harte
ad4afcdcd5 Switch stepping direction.
Empirically, based on the actions of Kickstart, and assuming my confusion is because the relevant signal is active low.
2021-10-05 15:23:48 -07:00