Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!utgpu!water!watmath!clyde!rutgers!ames!sdcsvax!ucbvax!wnre.aecl.CDN!delaney
From: delaney@wnre.aecl.CDN.UUCP
Newsgroups: comp.sys.apple
Subject: (none)
Message-ID: <775*delaney@wnre.aecl.cdn>
Date: Thu, 11-Jun-87 03:09:33 EDT
Article-I.D.: wnre.775*delaney
Posted: Thu Jun 11 03:09:33 1987
Date-Received: Sat, 13-Jun-87 08:19:03 EDT
Sender: daemon@ucbvax.BERKELEY.EDU
Distribution: world
Organization: The ARPA Internet
Lines: 41

Subject: apple tech note on Super Serial bug

 Miscellaneous Technical Note #3
     Super Serial Card Firmware Bug
 Written by:      Cameron Birse          11/85
 ________________________________________________________________________
 There is a bug (or omission) in the Super Serial Card firmware that has 
 become evident recently with the combination of Access //, UniDisk 3.5, 
 and a Super Serial Card used as a printer interface. The problem occurs 
 when you try to use the Access // utilities to print a file. When 
 Access retrieves data from the UniDisk and then tries to print it to 
 the Super Serial Card, a slot data bus conflict occurs. The reason for 
 this conflict is that when Access goes from the UniDisk operation to 
 the Super Serial Card operation, the UniDisk still 'owns' the $C800 ROM 
 space. The 'bug' in the serial firmware is that it does not make a 
 reference to $CFFF to clear the $C800 space before it jumps to it's 
 routines in that space. Therefore both the UniDisk 3.5 Interface card 
 ROM and the Super Serial Card ROM are enabled on the $C800 space at the 
 same time resulting in 'garbage' code where there should be Super 
 Serial Card firmware routines. An easy way to see this problem is to 
 look at the Super Serial Card firmware listing, specifically, the 
 PWRITE routine starting at $Cs97 (page 62 of the Super Serial Card 
 manual). At that address, the first instruction is a JMP $C9AA. This 
 jump is to the $C800 space without disabling any other slot ROMs. This 
 is the same for PREAD, PINIT, and PSTATUS. The BASIC entry point does 
 not have this problem because there is a STA $CFFF instruction at 
 location $Cs1B. An easy workaround to the problem would be to reference 
 $CFFF before you call any of the Pascal entry points on the Super 
 Serial Card, for example:
 Newslot:   STA   $CFFF      ;reset slot ROM space
            LDA   Char       ;Char = character to output
            LDX   #$Cn       ;n = slot #           LDY   #$n0
            STX   Mslot      ;Mslot = $7F8, in case of interupt
            JSR   Pwrite     ;now call the routine of your choice
 This bug did not happen with the Disk // or the Profile because the 
 Disk // does not have $C800 ROM and the Profile card data bus driver 
 'lost' the battle with the Super Serial Card. The Profile relies only 
 on the drive capability of the 2716 ROM to drive the bus whereas the 
 UniDisk uses a 74LS245 octal bus driver to interface to the data bus. 
 The Super Serial Card also uses a 74LS245 to interface to the data bus, 
 so the UniDisk and SSC have equal drive capability, and neither wins.