On Sunday 28 February 2010, Arne Pagel wrote: > > Hmm, I didn't observe that at all. I wonder if that's specific > > to the "recover" procedure, or whether other power cycle sequences > > would do the same thing? > This has nothing to do with the recover procedure, > this always happens if I power cycle the target hardware. > > I compiled with --enable-ft2232_libftdi, did you use ftd2xx?
Just libftdi ... none of that proprietary stuff for me, when I have the choice! ;) > > Regardless, note that the action you're prompted to do is not just a > > power cycle ... it's "power cycle Stellaris chip, then restart OpenOCD." > > > > So one *could* argue the segfault is just fortune's way to make sure > > you really follow the second half of those instructions... ;) > > You are right, I didn't follow the second part of the instructions exactly, > but then I was forced to to so :) I recall in a few tests that I saw strange messages about not being able to issue MPSSE commands (to the ft2232) ... strange. They weren't repeatable. > It could be more fortune to stop openocd automatically at the end of the > procedure. > After the procedure the user-interface reacts very slow, to enter a shutdown > command needs some > patience. Very true. Maybe you could see how the "shutdown" command does its magic, and then make that command do the same thing? I think it wasn't much more than a magic return value... (That should be a much gentler introduction to modifying the innards of OpenOCD than trying to get this command to work!) - Dave _______________________________________________ Openocd-development mailing list Openocd-development@lists.berlios.de https://lists.berlios.de/mailman/listinfo/openocd-development
