Macintosh-HW
/
macsehw
mirror of https://github.com/quorten/macsehw.git
1
0
Fork 0
Code Issues Projects Releases Wiki Activity

Act on falling edge rather than rising edge. 

Programming defensively with knowledge of the ROM quirks, this seems a
safer bet.
This commit is contained in:
Andrew Makousky 2020-09-11 15:50:15 -05:00
parent 3079bd621f
commit fe1bb541c1
1 changed files with 23 additions and 14 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

37
firmware/rtc/MacRTC.c
View File

@ -416,16 +416,19 @@ void loop(void)
set_sleep_mode(0); // Sleep mode 0 == default, timers still running.
sleep_mode();
} else {
/* Normally we only perform an action on the rising edge of the
serial clock. The main exception is cleanup at the last cycle
of serial output, there we wait one full cycle and then until
the falling edge before switching the direction of the data pin
/* Normally we only perform an action on the falling edge of the
serial clock.
If we instead program to use the rising edge for almost all
actions, we have an exception: Cleanup at the last cycle of
serial output, there we wait one full cycle and then until the
falling edge before switching the direction of the data pin
back to an input. */
if (serClockFalling &&
/* if (serClockFalling &&
serialState == SENDING_DATA &&
serialBitNum >= 9) {
clearState();
} else if (serClockRising) {
} else */ if (serClockFalling) {
boolean writeRequest;
switch(serialState) {
case RECEIVING_COMMAND:
@ -526,15 +529,21 @@ void loop(void)
digitalWriteOD(SERIAL_DATA_PIN,
bitRead(serialData, 7 - serialBitNum));
serialBitNum++;
/* if (serialBitNum <= 7)
break; */
if (serialBitNum >= 9)
clearState();
/* NOTE: The last output cycle is treated specially, hold the
data line as an output for at least one full next cycle,
then until the falling edge of the serial clock, then
switch back to an input and reset the serial communication
state. It's for bug compatibility with the ROM, but with a
little bit of sanity too. */
/* NOTE: The last output cycle is treated specially if we act
on the rising edge of the clock, hold the data line as an
output for at least one full next cycle, then until the
falling edge of the serial clock, then switch back to an
input and reset the serial communication state. It's for
bug compatibility with the ROM, but with a little bit of
sanity too.
However, for the time being, I've changed the code so all
actuions are preformed on the falling edge of the clock.
This seems to make things more consistent/robust given the
documented errors in the Macintosh ROM. */
break;
#if !defined(NoXPRAM) || !NoXPRAM