Am 06/11/2011 01:32 AM, schrieb Peter Stuge: > Tomek CEDRO wrote: >>>> If you switch output into Hi-Z the pull-up or pull-down resistor will >>>> take care of the on the pin state. >>> I think Michael's point is that *which* state it is is not known (and >>> can not be known) by generic ft2232 code in OpenOCD. >>> >>> And that just because the code does not have information, it should >>> not make guesses. >> How it cannot have this information? Who does then? When? > Well, it depends on what is between the FT2232 and the CPU, so > information from both interface and target board is needed. It could > certainly be added to OpenOCD, but I'm not sure if it is actually > available yet? Probably not - and I don't think this would be useful. If we want known state on the pins, we can simply drive them to that state (or keep them driven and not turn them off).
>> The goal of OpenOCD is to interfere with target to obtain expected >> results, after the job is done no interaction is necessary, or we >> simply not quit the program to enforce some specific behavior... > Yes, a good point. Michael, is the point to leave a known debug state > driven, or just to leave with a known pin state? I would like to keep the target in the same state. No matter if it is running or safely halted, if I quit OpenOCD, that should not change. Now aside from that, there might be situations where it is useful to logically detach the JTAG signals from the target (making them all High-Z), for example to see how the target reacts to a manual reset without intervention by OpenOCD, but that should be done by issuing some command. cu Michael _______________________________________________ Openocd-development mailing list [email protected] https://lists.berlios.de/mailman/listinfo/openocd-development
