Hi Rick, You don't have to rely on TRST, you should be able to reset the TAP by sending five clocks with TMS high.
If you meant to say that you can't enter debug mode without TRST, that's indeed the case for XScale chips (and maybe some others). However, you should still be able to use boundary scan and dump the flash via it (albeit slowly). OpenOCD does not have good support for boundary scan so you might have better luck with UrJTAG. If you have an MCU board with some GPIOs or can control them from PC, you can try my JTAG search program to confirm the pin layout. http://mbed.org/users/igorsk/programs/JTAG_Search/5z843/docs/main_8cpp_source.html It's written for mbed but should be easy to port to anything with some GPIOs. That said, a pull-down resistor should not prevent you from using the pin. You just need to provide some voltage to override it. On Thu, Nov 4, 2010 at 13:40, Rick van Rein <[email protected]> wrote: > Hi, > > I'm reverse engineering a SIP phone so I can reposess it and develop open > source > firmware for it. I think I found why JTAG was failing. > > It appears that my TRST is wired to GND through 12 Ohm resistor. As a result, > I cannot reset the chip into JTAG mode -- it needs Vtref on TRST to function, > AFAIK. > > Is this a well-known pattern? Any more hints from experience? > > > Thanks, > -Rick > _______________________________________________ > Openocd-development mailing list > [email protected] > https://lists.berlios.de/mailman/listinfo/openocd-development > -- WBR, Igor _______________________________________________ Openocd-development mailing list [email protected] https://lists.berlios.de/mailman/listinfo/openocd-development
