R63 EDITION 13 Page 1 ************* RMA additions *************** Release 2 (Edition 13): This is release 2 for the patch. It corrects errors in codes: 1. Fixed bug which did not allow tfr/exg with "W" 2. Disallows illegal codes original did not barf on: codes such as asr, and, bit, eor, adc, or, neg, asr, lsl, asl, rol with W,E,F 3. "lsl" was omitted from the list of commands in original Release 4. opt w?? now operational. See below 5. It was intended that opts be included on the same line, ala fcb,fdb's. The original worked as long as single-character opts were used. However, as soon as a multi-char opt (such as "w" was used, anything after was treated as comment. 6. A cosmetic thing, for listings, opt d, w, (or v for r63) would include part of the last line that was longer than the opt line, (try opt d?? just after the psect line). 7. r63 now will delete the ROF if an error has occurred unless the "k" option (either on command line or in an opt line) has been specified. In this archive is the .ipc file to modify the "rma" from the Level 2 Development Pak. This archive consists of two files -- this "readme" and rma.ipc. These modifications will add the commands for the new - well, new to us - Hitachi 6309 chip. The utility "ipatch" will be required for this modification, along with the factory-stock version of "rma" CRC = $F83DD9 (Edition 11). An ident of the two files should produce: Original New file ----------- ----------- Header for: rma Header for: r63 Module size: $4EAF #20143 Module size: $5C27 #23591 Module CRC: $F83DD9 (Good) Module CRC: $11BE26 (Good) Hdr parity: $C9 Hdr parity: $53 Exec. off: $001A #26 Exec. off: $001A #26 Data Size: $190D #6413 Data Size: $1A98 #6808 Edition: $0B #11 Edition: $0D #13 Ty/La At/Rv: $11 $81 Ty/La At/Rv: $11 $81 Prog mod, 6809 obj, re-en, R/O Prog mod, 6809 obj, re-en, R/O The new version works identically to the original, with the exception of the added mnemonics and one other addition. A new option has been added - "vx", where valid values can be either 0 or 1. This sets the RMA version number, so now, R63 EDITION 13 Page 2 r.link - compatible code can be generated ( v0 ) for public library functions. The default is version 1, version 1 can be specified, but is not necessary. Of course this can also be specified in the source " opt vx". Any value other than 0 or 1 will be ignored if it is on the command line, but will generate an error if in the source code. SYNTAX: same as 'd'epth or 'w'idth. "opt k" 'K'eep ROF. If this opt is not specified, the ROF will be deleted in case of error. SYNTAX: same as other 1-char opts. Opt "W" now works. A line will print up to one less character than the width specified. This will permit printing the carriage return without skipping a line if the length is >= the max width for the device. The default is 80. However, regardless of what width is specified, everything through the operand field will be printed. One side-effect of this is that if a width of less than about 50 is specified, then no comments will be displayed at all. The header line is not affected by the width spec. It will default to a minimum of 80, expanded further for a long pathlist to the filename. In this event, a page-depth of one less than max (or opt f) may need to be specified. The opts can be specified with more than one per line. As with fcb/fdb, several can be listed on a single line, separated by a comma, and no spaces. This feature worked on the original till w or d was encountered (and v in r63). Then anything after was treated as comment. As a little side note, this fix also fixed the printing of portions of the previous line as comment. One other change, the program name as supplied is "r63". This is in line with the later OSK rma's ("r68"). However, if you wish, the name can be changed to any other 3-character-or-less name. The name is located at $0D, $0E, and $0F. Remember that the last character must have the MS- bit set. The name also appears at the end, at $5C20-$5C22. The last char here does NOT have the MSB set, but the next char must be a NULL. To use "cc1", it will also be necessary to patch "cc1" for the new name. I will not attemt to describe the added features of the 6309. The best references currently available is "The 6309 Book", from Burke & Burke. I would heartily recommend it. A note of caution should be interjected here. I have not really and thoroughly tested the new program. To the best of my knowledge, the new modes operate correctly, but I may have missed some points. Also it could be possible that some of the old modes may have some bugs in them, as this was a project of reverse-engineering, so to speak, and some alterations had to be made in a few of the data structures. R63 EDITION 13 Page 3 This program should be used with some degree of caution, being considered to be still in the testing stages, for a little while. Some test programs should be compiled and compared with "rma" output for the old modes, and if you have it, compare the new modes with those from the "xsm" code. If any bugs are found, please get in touch with me. Also if you have any comments or suggestions for improving the general performance of the program, I would welcome them, although I would prefer not to dwell too much on extensively altering the main program. To contact me, I may be contacted on Compuserve (PIN 72330,2051) either through e-mail or in the OS9 Forum, or via internet:72330.2051@compuserve.com. I may also be contacted by U.S. Mail at the following address: David Breeding 336 Breeding Rd. Russell Springs, KY 42642-9330 This patch may be freely distributed as long as this disclaimer remains with it. ALSO: If you have passed previous versions of this patch along to any other service or BBS, would you please follow up with this patch.