Микроконтроллеры

Открываем FM A consequence of implicitly GO’ing to a breakpoint on main() as described in 14) above is that two GO’es are required to cause the RELEASE JTAG ON GO to work when working with assembler. RELEASE JTAG ON GO requests C-SPY to effectively detach itself from the device, thus freeing the JTAG pins to their port functions (if applicable). However, once C-SPY releases its JTAG interface to the device, C-SPY looses the status of the device. If C-SPY was “blind” following a GO, it could not determine when the breakpoint at main() was reached. Therefore, it was decided that the JTAG pins would not be released until after the second GO was executed (the first GO was required to cause the CPU to execute from the first line to main().). Once the first (implicit) breakpoint is reached at main(), a second GO releases the CPU and disables the JTAG control. In assembler, however, there is no implicit initial breakpoint set. Thus, following a RESET, even if RELEASE JTAG ON GO is selected and GO is asserted, the JTAG controls will not be released. To cause the JTAG controls to be released when using assembler, the device must be STOP’ped (from a running state), and then restarted using GO. Do NOT reset the device before restarting. Thus, the sequence RESET-GO-STOP-GO is required to start the CPU with JTAG released (assuming that RELEASE JTAG ON GO is selected) when working in assembler.