tag:blogger.com,1999:blog-1215830887795601932.post3914040849103473456..comments2024-03-24T03:16:50.870-04:00Comments on Jeff Tranter's Blog: A 6809 Single Board Computer: Instruction Trace/Step FunctionJeff Tranterhttp://www.blogger.com/profile/12906110561479326057noreply@blogger.comBlogger2125tag:blogger.com,1999:blog-1215830887795601932.post-35205416665725080332020-09-28T03:58:35.805-04:002020-09-28T03:58:35.805-04:00I’m confused by this blog post, as the original AS...I’m confused by this blog post, as the original ASSIST09 monitor (published in The 1981 Motorola MC6809 Programming Manual), already has Trace command and functionality. This is already implemented as ‘T’ for Trace number of instructions. Also ‘.’ for Trace one instruction. There is also ‘S’ for altering the stack trace level value. ASSIST09 implements this Trace functionality via timer interrupt, by default using timer 1 interrupt of a standard Motorola PTM (ie. the MC6840). Note the original published ASSIST09 listing uses the 6850 ACIA at address $E008, with the 6840 PTM at $E000. With no PTM present the built-in Trace functionality is of course non-functional. But adding a PTM and using the existing ASSIST09 Trace implementation would seem the way to go?GregChttps://www.blogger.com/profile/09962161141715257407noreply@blogger.comtag:blogger.com,1999:blog-1215830887795601932.post-54337334269155211022019-10-06T22:38:22.476-04:002019-10-06T22:38:22.476-04:00Agree with what you say here but suggest one small...Agree with what you say here but suggest one small optimisation: instead of knowing the length of the instruction and placing the return to control sequence immediately after the instruction, just place the instruction in a 5-byte buffer preloaded with NOPs. That's what we used to do at Acorn with the single step monitor for the 6502. (well, that one was 3 bytes but same principle). May simplify and shorten the code a tiny bit.Ghttps://www.blogger.com/profile/04079389970653405030noreply@blogger.com