16:03 <gamiee> hello. I am implementing proper reset, and for this to work, I need also reset GPIO controller, which thus resets pinmux, so OpenOCD losts JTAG connection. After the GPIO controller reset, I need to trigger system reset (through registers), but since JTAG is already cut off, I can't do that. One idea is to construct two jump instructions, jump to them and then just wait until OpenOCD will be able to connect again. But I can't do this through TCL, since only

16:03 <gamiee> from C code I can run algorithm. Any ideas?

16:35 <zapb_> gamiee, can you provide more context? Not sure what are you trying to do / implement

16:35 <zapb_> What is your target?

16:39 <gamiee> RISC-V, BL702.

16:40 <gamiee> zapb_ : In short, I need to set register which resets GPIO pinmux conf (so JTAG connection is lost) and then perform CPU reset through register.

16:41 <gamiee> But I can't reset CPU, because I will lose JTAG. So my idea is to put some trampoline code into RAM, which will reset GPIO and then perform CPU.

17:43 <zapb_> gamiee, and what's the problem? You can write this code via Tcl into your device, execute it, wait some time and try to examine the target again

17:43 <zapb_> Probably you need to call "poll off", otherwise OpenOCD tries to check the target status periodically

18:15 <gamiee> zapb_ : so I will just write it by writing bytearray? Like, I can do that, for sure... Just I thought there is some better solution, or I can call flash bank driver functions, that maybe would work too.

18:16 <gamiee> But I have bigger problem right now, I have another method, but my target doesn't support abstract commands... :(

18:16 <gamiee> so if I use sysbus for writing register, OpenOCD will be not able to read out result of that.

18:28 <zapb_> gamiee, it should work like that but is not the "best" solution probably. Are you trying to write a driver / support for this device? I'm not very familiar with the risc-v code in OpenOCD unfortunately

19:22 <gamiee> zapb_ : I already upstreamed basic support for the chip. now I am working on Flash driver (almost done), and now I am testing new, better reset, which seems to reset whole chip, as my previous method didn't.

19:22 <gamiee> I solved it by writing instructions into RAM and jumping to that.

19:23 <gamiee> it works so far, so I am happy.

19:43 <zapb_> gamiee, why do you need to reset the device for flash operations?