[Coco] Question on 6309 TFM command

[L. Curtis Boyle]

Has anyone tested the TFM command when going from a single memory location, writing to a range (TFM r0,r1+), to see if it takes the same # of cycles as it would with both source and destination registers updating (TFM r0+,r1+). That’s what some of the 6309 docs say it does, but wondering if it is more optimized than that if it doesn’t have to update the source ptr (or possibly even the data byte itself, if it knows it is outside of the I/O range, perhaps).

If it is in fact the same speed either way, then filling a contiguous, large chunk of memory with a single byte would be faster using a stack blast, 6809 style (although it would take more RAM).

L. Curtis Boyle
curtisboyle at sasktel.net

TRS-80 Color Computer Games website
http://www.lcurtisboyle.com/nitros9/coco_game_list.html