Sprinter/mame
2
0
Fork 0
mirror of https://github.com/holub/mame synced 2025-04-23 00:39:36 +03:00
Code Issues Actions 1 Packages Projects Releases Wiki Activity

(MESS) cat.c(Canon Cat): added another v2.40 romset dumped from two original developer's machines and made it the parent set. Lots and lots of documentation updates. [Lord Nightmare, Dwight Elvey, Sandy Bumgarner]

Browse Source
This commit is contained in:
Jonathan Gevaryahu 2013-02-04 10:37:32 +00:00
parent db6b7972f7
commit 9d25be3252
1 changed files with 182 additions and 37 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

219
src/mess/drivers/cat.c
View File

@ -1,12 +1,51 @@
/***************************************************************************
Canon Cat driver by Miodrag Milanovic and Lord Nightmare
Canon Cat, Model V777
IAI Swyft Model P0001
driver by Miodrag Milanovic and Lord Nightmare
With information and help from John "Sandy" Bumgarner, Dwight Elvey,
Charles Springer, Terry Holmes, Jonathan Sand, Aza Raskin and others.
This driver is dedicated in memory of Jef Raskin and Dave Boulton
12/06/2009 Skeleton driver.
15/06/2009 Working driver
Pictures: http://www.regnirps.com/Apple6502stuff/apple_iie_cat.htm
Canon cat blue "Use Front Labels":
` (really degree, +-, paragraph-mark and section-mark): LEFT MARGIN
-_: INDENT
+=: RIGHT MARGIN
UNDO: SPELL CHECK LEAP (this is actually a red LEAP label)
PERM SPACE/TAB: SET/CLEAR TAB
Q: UNDER LINE
W: BOLD
E: CAPS
T: Paragraph STYLE
U: LINE SPACE
O: ADD SPELLING
[ (really 1/4, 1/2, double-underline and |): SETUP
Backspace (really ERASE): ANSWER
LOCK: DOCUMENT LOCK
A: COPY
D: SEND CONTROL
G: CALC
J: PRINT
L: DISK
Colon: FORMAT/WIPE DISK (only when wheel! mode is enabled, see below)
"': PHONE
Enter (really RETURN): SEND
X: LOCAL LEAP
V: LEARN
N: EXPLAIN
,: SORT
?/: KBI/II
Pgup (really DOCUMENT/PAGE): TITLES
Leap Left: LEAP AGAIN (left)
Leap Right: LEAP AGAIN (right)
How to enable the FORTH interpreter (http://canoncat.org/canoncat/enableforth.html)
The definitive instructions of going Forth on a Cat, tested on a living Cat.
@ -25,9 +64,62 @@ In MESS, to activate it as above:
* when the Cat boots, type (without quotes) "Enable Forth Language"
* hold left-alt(leap left) and type E n a, release left-alt (the left cursor is now at the first character)
* simultaneously press both alt keys for a moment and release both (the whole "Enable Forth Language" line will be selected)
* press control(use front) and press backspace(Erase) (If beeping actually worked the cat would beep here)
* press control(use front) and press backspace(ERASE) (If beeping actually worked the cat would beep here)
* press control(use front), shift, and space (the cursor should stop blinking)
* press enter and the forth "ok" prompt should appear. you can type 'page' and enter to clear the screen
Optional further steps:
* type without quotes "-1 wheel! savesetup re" at the forth prompt to permanently
enable shift + use front + space to dump to forth mode easily
* change the keyboard setting in the setup menu (use front + [ ) to ASCII so you can type < and >
* after doing the -1 wheel! thing, you can compile a selected forth program in the editor
by selecting it and hitting ANSWER (use front + ERASE)
If ever in forth mode you can return to the editor with the forth word (without quotes) "re"
Canon cat credits easter egg:
* hold either leap key, then simultaneously hold shift, then type Q W E R A S D F Z X C V and release all the keys
* hit EXPLAIN (use front + N) and the credits screen will be displayed
Canon Cat credits details: (WIP)
Scott Kim - responsible for fonts on swyft and cat
Swyft versions:
There are at least 4 variants of machines called 'swyft':
* The earliest desktop units which use plexi or rubber-tooled case and an
angled monitor; about a dozen were made and at least two of clear plexi.
These are sometimes called "wrinkled" swyfts. 5.25" drive, may be able to
read Apple2 Swyftware and Swyftcard-created disks.
It is possible no prototypes of this type got beyond the 'runs forth console only' stage.
http://archive.computerhistory.org/resources/access/physical-object/2011/09/102746929.01.01.lg.JPG
http://www.digibarn.com/collections/systems/swyft/Swyft-No2-05-1271.jpg
http://www.digibarn.com/friends/jef-raskin/slides/iai/A%20-687%20SWYFTPRO.JPG
* The early "flat cat" or "roadkill" portable LCD screen units with a white
case and just a keyboard. Model SP0001
http://www.digibarn.com/collections/systems/swyft/Image82.jpg
* The later "ur-cat" desktop units which use a machine tooled case and look
more or less like the canon cat. about 100-200 were made. 3.5" drive.
These have a fully functional EDDE editor as the cat does, and can even compile forth programs.
(the 'swyft' driver is based on one of these)
* The very late portable LCD units with a dark grey case and a row of hotkey
buttons below the screen.
http://www.digibarn.com/collections/systems/swyft/swyft.jpg
Canon Cat versions:
There is really only one version of the cat which saw wide release, the US version.
* It is possible a very small number of UK/European units were released as a test.
If so, these will have slightly different keyboard key caps and different
system and spellcheck roms.
As for prototypes/dev cat machines, a few minor variants exist:
* Prototype cat motherboards used 16k*4bit drams instead of 64k*4bit as the
final system did and hence max out at 128k of dram instead of 512k.
The final system included 256k of dram and can be upgraded to 512k.
* At least some developer units were modified to have an external BNC
connector, ostensibly to display the internal screen's video externally.
http://www.digibarn.com/collections/systems/canon-cat/Image55.jpg
Canon Cat:
<insert guru-diagram here once drawn>
@ -36,18 +128,23 @@ X1: 19.968Mhz, used by GA2 (plus a PLL to multiply by 2?), and divide by 4 for c
X2: 3.579545Mhz, used by the DTMF generator chip AMI S2579 at IC40
X3: 2.4576Mhz, used by the modem chip AMI S35213 at IC37
IAI Swyft:
<insert guru-diagram here once drawn>
ToDo:
* Canon Cat
- Find the mirrors for the write-only video control register and figure out
what the writes actually do; hook these up properly to screen timing etc
- The 2.40 (bios 0) firmware gets annoyed and thinks it has a phone call at
- The 2.40 firmwares both get annoyed and think they have a phone call at
random.
(hit ctrl a few times to make it shut up for a bit or go into forth mode);
The 1.74 (bios 1) firmware doesn't have this issue.
(hit shift or usefront a few times to make it shut up for a bit or go into
forth mode); The 1.74 firmware doesn't have this issue.
Figure out what causes it and make it stop. May be OFFHOOK or more likely
the DUART thinks the phone is ringing constantly.
the DUART thinks the phone is ringing constantly due to incomplete duart or
an ip bit hooked up wrong.
You can alsp make it stop by going to SETUP and change the serial port from
'alternate printer' to SEND.
- Floppy drive (3.5", Single Sided Double Density MFM, ~400kb)
* Cat has very low level control of data being read or written, much like
the Amiga does
@ -71,18 +168,32 @@ ToDo:
- Hook duart IP2 up to the 6ms timer
- Correctly hook the duart interrupt to the 68k, including autovector using the vector register on the duart
- Watchdog timer/powerfail at 0x85xxxx
- Canon Cat released versions known: 1.74 (dumped), 2.40 (dumped), 2.42 (NEED DUMP)
- Canon Cat released versions known: 1.74 US (dumped), 2.40 US (dumped; is this actually original, or compiled from the released source code?), 2.42 (NEED DUMP)
It is possible a few prototype UK 1.74 or 2.40 units were produced; the code roms of these will differ (they contain different spellcheck "core" code) as well as the spellcheck roms, the keyboard id and the keycaps.
- Known Spellcheck roms: NH7-0684 (US, dumped); NH7-0724 (UK, NEED DUMP);
NH7-0813/0814 (Quebec/France, NEED DUMP); NH7-1019/1020/1021 (Germany, NEED DUMP)
It is possible the non-US roms were never officially released.
Wordlist sources: American Heritage (US and UK), Librarie Larousse (FR), Langenscheidt (DE)
- (would-be-really-nice-but-totally-unnecessary feature): due to open bus, the
svrom1 and svrom2 checksums in diagnostics read as 01A80000 and 01020000
respectively on a real machine (and hence appear inverted/'fail'-state).
This requires sub-cycle accurate 68k open bus emulation to pull off, as well
as emulating the fact that UDS/LDS are ?not connected? (unclear because this
happens inside an asic) for the SVROMS (or the svram or the code roms, for
that matter!)
- Hook Battery Low input to a dipswitch.
- Document what every IPx and OPx bit on the DUART connects to.
* Swyft
- Figure out the keyboard (interrupts are involved? or maybe an NMI on a
timer/vblank?)
- Communications ports (Duart? or some other plain UART?)
- Beeper
- Communications port (Duart? or some other plain UART?)
- Floppy (probably similar to the Cat)
- Centronics port (probably similar to the Cat)
- Joystick port
- Forth button (keep in mind shift-usefront-space ALWAYS enables forth on a swyft)
- Multple undumped firmware revisions exist
****************************************************************************/
@ -428,7 +539,7 @@ READ16_MEMBER( cat_state::cat_6ms_counter_r )
}
/* 0x840001: 'opr' Output Port Register
* writing 0x1c (or possibly anything) here resets the watchdog
* writing 0x1c (or probably anything with bit 3 set) here resets the watchdog
* if the watchdog expires an NMI is sent to the cpu
*/
WRITE16_MEMBER( cat_state::cat_opr_w )
@ -445,17 +556,16 @@ WRITE16_MEMBER( cat_state::cat_opr_w )
* \--------- ?
*/
#ifdef DEBUG_VIDEO_ENABLE_W
//if ((data&0xef)!= 0x0c)
fprintf(stderr, "Video enable reg write: offset %06X, data %04X\n", 0x840000+(offset<<1), data);
#endif
m_video_enable = BIT( data, 2 );
m_video_invert = 1-BIT( data, 4 );
}
// 0x850000: watchdog timer/video status
// 0x850000: 'wdt' watchdog timer/video status
// implement me!
// 0x860000: test regitser
// 0x860000: 'tcb' test regitser
// the power fail status reset bit also lives here somewhere?
WRITE16_MEMBER( cat_state::cat_tcb_w )
{
@ -495,7 +605,7 @@ a23 a22 a21 a20 a19 a18 a17 a16 a15 a14 a13 a12 a11 a10 a9 a8 a7 a6 a5 a4
0 0 1 x x 0 * * * * * * * * * * * * * * * * * 1 R SVROM 0 ic6 (MASK ROM tc531000) [controlled via GA2 /SVCS0]
0 0 1 x x 1 * * * * * * * * * * * * * * * * * 0 O OPEN BUS (reads as 0x2e) [controlled via GA2 /SVCS1] *SEE BELOW*
0 0 1 x x 1 * * * * * * * * * * * * * * * * * 1 R SVROM 1 ic8 (not present on cat as sold, open bus reads as 0x80) [controlled via GA2 /SVCS1] *SEE BELOW*
*NOTE: on developer units, two 128K SRAMS are mapped in place of the two entries immediately above!* (this involves some creative wiring+sockets or an official IAI 'shadow ram board')
*NOTE: on Dwight E's user-made developer unit, two 128K SRAMS are mapped in place of the two entries immediately above!* (this involves some creative wiring+sockets); the official IAI 'shadow ram board' maps the ram to the A00000-A3FFFF area instead)
0 1 * * * * * * * * * * * * * * * * * * * * * * *BOTH GATE ARRAYS 1 and 2 DECODE THIS AREA; 2 DEALS WITH ADDR AND 1 WITH DATA/CAS/RAS*
0 1 0 x x a b * * * * * * * * * * * * * * * * * RW VIDEO/SYSTEM DRAM (ab: 00=row 0, ic26-29; 01=row 1, ic22-25; 10=row 2; ic18-21; 11=row 3; ic14-17)
*NOTE: DRAM rows 2 and 3 above are only usually populated in cat developer units!*
@ -520,8 +630,8 @@ a23 a22 a21 a20 a19 a18 a17 a16 a15 a14 a13 a12 a11 a10 a9 a8 a7 a6 a5 a4
1 0 0 x x 1 1 0 x x x x x x x x x x x x x x x * R?W {'tcb'} test control bits: powerfail status in bit <?> (reads as 0x0000)
1 0 0 x x 1 1 1 x x x x x x x x x x x x x x x * ? Unknown (reads as 0x2e80)
1 0 1 x x x x x x x x x x x x x x x x x x x x x O OPEN BUS (reads as 0x2e80)
1 1 x x x x x x x x x x x x x x x x x x x x x x O OPEN BUS (reads as 0x2e80)
1 0 1 x x x x x x x x x x x x x x x x x x x x x O OPEN BUS (reads as 0x2e80) [68k DTACK is asserted by gate array 1 when accessing this area, for testing?] On real IAI shadow rom board, at least 0x40000 of ram lives here.
1 1 x x x x x x x x x x x x x x x x x x x x x x O OPEN BUS (reads as 0x2e80) [68k VPA is asserted by gate array 1 when accessing this area, for testing?]
*/
@ -548,8 +658,8 @@ static ADDRESS_MAP_START(cat_mem, AS_PROGRAM, 16, cat_state)
//AM_RANGE(0x850000, 0x850001) AM_READ(cat_video_status) AM_MIRROR(0x18FFFE) // video status and watchdog read: hblank, vblank or draw?
AM_RANGE(0x860000, 0x860001) AM_READWRITE(cat_0000_r, cat_tcb_w) AM_MIRROR(0x18FFFE) // Test mode
AM_RANGE(0x870000, 0x870001) AM_READ(cat_2e80_r) AM_MIRROR(0x18FFFE) // Open bus?
AM_RANGE(0xA00000, 0xA00001) AM_READ(cat_2e80_r) AM_MIRROR(0x1FFFFE) // Open bus?
AM_RANGE(0xC00000, 0xC00001) AM_READ(cat_2e80_r) AM_MIRROR(0x3FFFFE) // Open bus?
AM_RANGE(0xA00000, 0xA00001) AM_READ(cat_2e80_r) AM_MIRROR(0x1FFFFE) // Open bus/dtack? The 0xA00000-0xA3ffff area is ram used for shadow rom storage on cat developer machines, which is either banked over top of, or jumped to instead of the normal rom
AM_RANGE(0xC00000, 0xC00001) AM_READ(cat_2e80_r) AM_MIRROR(0x3FFFFE) // Open bus/vme?
ADDRESS_MAP_END
static ADDRESS_MAP_START(swyft_mem, AS_PROGRAM, 16, cat_state)
@ -891,29 +1001,64 @@ ROM_END
ROM_START( cat )
ROM_REGION( 0x40000, "maincpu", ROMREGION_ERASEFF )
// SYS ROM
ROM_SYSTEM_BIOS( 0, "r240", "Canon Cat V2.40 Firmware")
ROMX_LOAD( "r240l0.ic2", 0x00001, 0x10000, CRC(1b89bdc4) SHA1(39c639587dc30f9d6636b46d0465f06272838432), ROM_SKIP(1) | ROM_BIOS(1))
ROMX_LOAD( "r240h0.ic4", 0x00000, 0x10000, CRC(94f89b8c) SHA1(6c336bc30636a02c625d31f3057ec86bf4d155fc), ROM_SKIP(1) | ROM_BIOS(1))
ROMX_LOAD( "r240l1.ic3", 0x20001, 0x10000, CRC(1a73be4f) SHA1(e2de2cb485f78963368fb8ceba8fb66ca56dba34), ROM_SKIP(1) | ROM_BIOS(1))
ROMX_LOAD( "r240h1.ic5", 0x20000, 0x10000, CRC(898dd9f6) SHA1(93e791dd4ed7e4afa47a04df6fdde359e41c2075), ROM_SKIP(1) | ROM_BIOS(1))
ROM_SYSTEM_BIOS( 1, "r174", "Canon Cat V1.74 Firmware")
// Canon cat v1.74 roms are labeled as r74; they only added the major number to the rom label after 2.0
ROMX_LOAD( "r74__0l__c18c.blue.ic2", 0x00001, 0x10000, CRC(b19aa0c8) SHA1(85b3e549cfb91bd3dd32335e02eaaf9350e80900), ROM_SKIP(1) | ROM_BIOS(2))
ROMX_LOAD( "r74__0h__75a6.yellow.ic4", 0x00000, 0x10000, CRC(75281f77) SHA1(ed8b5e37713892ee83413d23c839d09e2fd2c1a9), ROM_SKIP(1) | ROM_BIOS(2))
ROMX_LOAD( "r74__1l__c8a3.green.ic3", 0x20001, 0x10000, CRC(93275558) SHA1(f690077a87076fd51ae385ac5a455804cbc43c8f), ROM_SKIP(1) | ROM_BIOS(2))
ROMX_LOAD( "r74__1h__3c37.white.ic5", 0x20000, 0x10000, CRC(5d7c3962) SHA1(8335993583fdd30b894c01c1a7a6aca61cd81bb4), ROM_SKIP(1) | ROM_BIOS(2))
/* This 2.40 code came from two development cat machines owned or formerly
* owned by former IAI employees Sandy Bumgarner and Dave Boulton.
* Dave Boulton's machine is interesting in that it has a prototype cat
* motherboard in it, which it has less space for dram than a 'released'
* cat does: it uses 16k*4 dram chips instead of 64k*4 as in the final
* cat, and hence can only support 128k of ram with all 4 rows of drams
* populated, as opposed to 256k-standard (2 rows) and 512k-max with all
* 4 rows populated on a "released" cat.
*/
ROM_SYSTEM_BIOS( 0, "r240", "Canon Cat V2.40 US Firmware")
ROMX_LOAD( "boultl0.ic2", 0x00001, 0x10000, CRC(77b66208) SHA1(9D718C0A521FEFE4F86EF328805B7921BADE9D89), ROM_SKIP(1) | ROM_BIOS(1))
ROMX_LOAD( "boulth0.ic4", 0x00000, 0x10000, CRC(f1e1361a) SHA1(0A85385527E2CC55790DE9F9919EB44AC32D7F62), ROM_SKIP(1) | ROM_BIOS(1))
ROMX_LOAD( "boultl1.ic3", 0x20001, 0x10000, CRC(c61dafb0) SHA1(93216c26c2d5fc71412acc548c96046a996ea668), ROM_SKIP(1) | ROM_BIOS(1))
ROMX_LOAD( "boulth1.ic5", 0x20000, 0x10000, CRC(bed1f761) SHA1(D177E1D3A39B005DD94A6BDA186221D597129AF4), ROM_SKIP(1) | ROM_BIOS(1))
/* This 2.40 code was compiled by Dwight Elvey based on the v2.40 source
* code disks recovered around 2004. It does NOT exactly match the above
* set exactly but has a few small differences.
* It is as of yet unknown whether it is earlier or later code than the
* set above.
*/
ROM_SYSTEM_BIOS( 1, "r240r", "Canon Cat V2.40 US Firmware compiled from recovered source code")
ROMX_LOAD( "r240l0.ic2", 0x00001, 0x10000, CRC(1b89bdc4) SHA1(39c639587dc30f9d6636b46d0465f06272838432), ROM_SKIP(1) | ROM_BIOS(2))
ROMX_LOAD( "r240h0.ic4", 0x00000, 0x10000, CRC(94f89b8c) SHA1(6c336bc30636a02c625d31f3057ec86bf4d155fc), ROM_SKIP(1) | ROM_BIOS(2))
ROMX_LOAD( "r240l1.ic3", 0x20001, 0x10000, CRC(1a73be4f) SHA1(e2de2cb485f78963368fb8ceba8fb66ca56dba34), ROM_SKIP(1) | ROM_BIOS(2))
ROMX_LOAD( "r240h1.ic5", 0x20000, 0x10000, CRC(898dd9f6) SHA1(93e791dd4ed7e4afa47a04df6fdde359e41c2075), ROM_SKIP(1) | ROM_BIOS(2))
/* This v1.74 code comes from (probably) the 'main us release' of first-run
* Canon cats, and was dumped from machine serial number R12014979
* Canon cat v1.74 roms are labeled as r74; they only added the major number
* to the rom label after v2.0?
*/
ROM_SYSTEM_BIOS( 2, "r174", "Canon Cat V1.74 US Firmware")
ROMX_LOAD( "r74__0l__c18c.blue.ic2", 0x00001, 0x10000, CRC(b19aa0c8) SHA1(85b3e549cfb91bd3dd32335e02eaaf9350e80900), ROM_SKIP(1) | ROM_BIOS(3))
ROMX_LOAD( "r74__0h__75a6.yellow.ic4", 0x00000, 0x10000, CRC(75281f77) SHA1(ed8b5e37713892ee83413d23c839d09e2fd2c1a9), ROM_SKIP(1) | ROM_BIOS(3))
ROMX_LOAD( "r74__1l__c8a3.green.ic3", 0x20001, 0x10000, CRC(93275558) SHA1(f690077a87076fd51ae385ac5a455804cbc43c8f), ROM_SKIP(1) | ROM_BIOS(3))
ROMX_LOAD( "r74__1h__3c37.white.ic5", 0x20000, 0x10000, CRC(5d7c3962) SHA1(8335993583fdd30b894c01c1a7a6aca61cd81bb4), ROM_SKIP(1) | ROM_BIOS(3))
ROM_REGION( 0x80000, "svrom", ROMREGION_ERASE00 )
// SPELLING VERIFICATION ROM (SVROM)
// since ROM_FILL16BE(0x0, 0x80000, 0x2e80) doesn't exist, the even bytes and latter chunk of the svrom space are filled in in DRIVER_INIT
// Romspace here is a little strange: there are 3 rom sockets on the board:
// svrom-0 maps to 200000-21ffff every ODD byte (d8-d0)
ROMX_LOAD( "uv1__nh7-0684__hn62301apc11__7h1.ic6", 0x00000, 0x20000, CRC(229ca210) SHA1(564b57647a34acdd82159993a3990a412233da14), ROM_SKIP(1)) // this is a 28pin tc531000 mask rom, 128KB long
// svrom-1 maps to 200000-21ffff every EVEN byte (d15-d7)
// no rom is in the socket; it reads as open bus 0x2E
// svrom-2 maps to 240000-25ffff every ODD byte (d8-d0)
// no rom is in the socket; it reads as open bus 0x80
// there is no svrom-3; 240000-25ffff EVEN always reads as 0x2E
/* Romspace here is a little strange: there are 3 rom sockets on the board:
* svrom-0 maps to 200000-21ffff every ODD byte (d8-d0)
* svrom-1 maps to 200000-21ffff every EVEN byte (d15-d7)
* (since no rom is in the socket; it reads as open bus 0x2E)
* svrom-2 maps to 240000-25ffff every ODD byte (d8-d0)
* (since no rom is in the socket; it reads as open bus 0x80)
* there is no svrom-3; 240000-25ffff EVEN always reads as 0x2E
* since ROM_FILL16BE(0x0, 0x80000, 0x2e80) doesn't exist, the
* even bytes and latter chunk of the svrom space need to be filled in
* DRIVER_INIT or some other means needs to be found to declare them as
* 'open bus' once the mame/mess core supports that.
* NOTE: there are at least 6 more SVROMS which existed (possibly in
* limited form), and are not dumped:
* UK (1 rom)
* French (2 roms)
* German (3 roms)
* Each of these will also have its own code romset as well.
*/
ROMX_LOAD( "uv1__nh7-0684__hn62301apc11__7h1.ic6", 0x00000, 0x20000, CRC(229ca210) SHA1(564b57647a34acdd82159993a3990a412233da14), ROM_SKIP(1)) // this is a 28pin tc531000 mask rom, 128KB long; "US" SVROM
ROM_END
/* Driver */