On Tue, Apr 5, 2022, 7:02 AM Undiscussed Horrific Abuse, One Victim of
   Many <[1]gmkarl@gmail.com> wrote:

   notes: in the flasher source, it's designed for more than one
   interface, and use of the lpc chip is guarded by this:
   uint8_t lpc_test(void) {
           nibble_hw_init();

   FRAME is driven high for 1us and then allowed to drop for 1us.

           lpc_init();

   calls nibble_init():
   INIT is briefly dropped and then raised.
   CLK is raised.
   FRAME is raised.
   calls nibble_write(0):
   DATA lines are set high.
   calls clock_cycle() 24 times.
   The code for clock_cycle drops CLK then immediately raises it.
   It goes on with more port writes. I may have made an error in
   transcribing the above.
   I'm guessing the clock is happening faster than I have my oscilloscope
   set. Maybe I can spend some time learning it better and see the 1us
   FRAME behavior.
           if (lpc_read_address(0xFFFFFFFF)==-1) return 0;

           return 1;
   }
   flashrom itself doesn't seem to care about that check (although I
   haven't looked in the source): it just probes for a bunch of chips
   anyway.
   so I guess the thing to do would be to look for the first lines that
   are engaged, and see where failure starts.

References

   1. mailto:gmkarl@gmail.com