#asahi-dev · 2021-12-05 08:15
marcansven: heh, I've been testing stuff and sometimes it gets into a state where after a successful boot, it will send the first message with the versions, then get stuck#asahi-dev · 2021-12-05 07:45
marcanthe other bits in the cpu control reg do nothing?#asahi-dev · 2021-12-05 07:41
marcanso surely there is *some* way to reset these things?#asahi-dev · 2021-12-05 07:41
marcansven: the interesting thing is that since iBoot shuts down ANS2 (entirely it seems, it powers it down), the .data thing can't be a problem#asahi-dev · 2021-12-03 17:23
marcan(unfortunately NVMe is...)#asahi-dev · 2021-12-03 17:23
marcananyway, 'night :)#asahi-dev · 2021-12-03 17:23
marcanI tried PMP earlier and it didn't work; might try manually resetting .data since that one is not locked later#asahi-dev · 2021-12-03 17:22
marcanthat would mean we can't actually do clean resets#asahi-dev · 2021-12-03 17:22
marcanmaybe iBoot loads it and that's all you get#asahi-dev · 2021-12-03 17:22
marcanmaybe it isn't#asahi-dev · 2021-12-03 17:22
marcanI've always wondered how the .data section is reset when you reboot an ASC#asahi-dev · 2021-12-03 17:22
marcanalso, this could be something dumb#asahi-dev · 2021-12-03 17:21
marcanwonder if we can get any debug as to what these ASCs are doing#asahi-dev · 2021-12-03 17:21
marcanI'll have to look into this#asahi-dev · 2021-12-03 17:21
marcanhmm, yeah...#asahi-dev · 2021-12-03 17:16
marcanapcie_st specifically#asahi-dev · 2021-12-03 17:16
marcanapcie?#asahi-dev · 2021-12-03 17:16
marcanmaybe something else needs resetting#asahi-dev · 2021-12-03 17:16
marcanha#asahi-dev · 2021-12-03 17:12
marcanI gotta sleep, have fun :)#asahi-dev · 2021-12-03 17:10
marcanm1n1 polls of course#asahi-dev · 2021-12-03 17:10
marcanmaybe check the mailbox status manually#asahi-dev · 2021-12-03 17:10
marcanalso, if the mailbox driver relies on IRQs and it's already initialized by now, maybe that config is lost#asahi-dev · 2021-12-03 17:07
marcan(well, the whole asc)#asahi-dev · 2021-12-03 17:07
marcanmaybe trace the mailbox too#asahi-dev · 2021-12-03 17:07
marcanand that way you know when it's booting#asahi-dev · 2021-12-03 17:07
marcanso, make sure that bit is 0, *then* reset, *then* set it to 1#asahi-dev · 2021-12-03 17:06
marcanmaybe you're dropping the mailbox message in that case?#asahi-dev · 2021-12-03 17:06
marcanif it's left at 1, the thing will boot on its own#asahi-dev · 2021-12-03 17:06
marcansven: I have an idea#asahi-dev · 2021-12-03 17:06
marcanso it looks like it's a cpu_run flag or whatever#asahi-dev · 2021-12-03 17:05
marcanif 1 it does#asahi-dev · 2021-12-03 17:05
marcanif it's 0 it doesn't boot after a pmgr reset#asahi-dev · 2021-12-03 17:05
marcanactually, as long as the bit is 1 and left at 1 it works :D#asahi-dev · 2021-12-03 17:05
marcanwell, to be specific, just this is enough: self.asc.CPU_CONTROL.set(RUN=1)#asahi-dev · 2021-12-03 17:03
marcaneven if it's already 1 (as long as you reset)#asahi-dev · 2021-12-03 17:03
marcanbut even if you just make it a set32 to only write 1, it still works#asahi-dev · 2021-12-03 17:02
marcannote that the m1n1 asc code does 0->1->0 on that cpu run bit#asahi-dev · 2021-12-03 17:02
marcanbut it still works#asahi-dev · 2021-12-03 17:02
marcansven: oh, the reset doesn't clear the bit#asahi-dev · 2021-12-03 17:01
marcan:D#asahi-dev · 2021-12-03 17:00
marcanI need to get some sleep, but I hope that leads you somewhere :)#asahi-dev · 2021-12-03 17:00
marcanbut the reset works#asahi-dev · 2021-12-03 17:00
marcanyeah, it's not doing SART#asahi-dev · 2021-12-03 16:56
marcanhttps://hub.marcan.st/t/ans_test.py#asahi-dev · 2021-12-03 16:56
marcansven: it does work for me, found the old code#asahi-dev · 2021-12-03 16:44
marcanit did work with PMP IIRC#asahi-dev · 2021-12-03 16:44
marcansven: hm, I'll have to look into that#asahi-dev · 2021-12-03 16:29
marcansven: can you hvtrace it to make sure I don't fail it with the regmap stuff?#asahi-dev · 2021-12-03 14:44
marcanyup