Sun 09 May 2010 01:47:32 PM UTC, original submission:
sometimes "something" causes the parallel port to be left in a state where the reset pin or data pins are driven high by the
STK200 programmer
In this state there is no way to reset the port using avrdude
as the -E parameter is being ignored.
This seems like an oversight, for although the STK200 by design cannot power the target AVR it can interfere with its operation.
it seems there is a "setpin" where there should be "setmany"
here's a patch :)
==== BEGIN
diff --git a/par.c b/par.c
index 269c6c8..6bb064e 100644
--- a/par.c
+++ b/par.c
@@ -285,7 +285,7 @@ static void par_close(PROGRAMMER * pgm)
ppi_setall(&pgm->fd, PPIDATA, pgm->ppidata);
ppi_setall(&pgm->fd, PPICTRL, pgm->ppictrl);
- par_setpin(pgm, pgm->pinno[PPI_AVR_BUFF], 1);
+ par_setmany(pgm, pgm->pinno[PPI_AVR_BUFF], 1);
/*
* Handle exit specs.
==== END
IMO havinf setpin and setmany able to me interchanged
without generating a compiler diagnostic is a mistake.
|