#asahi-dev · 2021-01-30 15:07
marcanI'll leave this pulling crash/nocrash bits overnight#asahi-dev · 2021-01-30 15:07
marcananyway, going to call it a day#asahi-dev · 2021-01-30 15:06
marcankettenis: yeah, figured that would work (or I can just make m1n1 lie about it)#asahi-dev · 2021-01-30 15:05
marcanbut they suck in many ways#asahi-dev · 2021-01-30 15:05
marcanthough they will *probably* work for this#asahi-dev · 2021-01-30 15:05
marcanalso, I would avoid "bidirectional" level translators#asahi-dev · 2021-01-30 15:05
marcanyes, you can't use resistor dividers for power supplies unless you use really low value resistors :p#asahi-dev · 2021-01-30 14:56
marcanBluerise: that looks like it should work#asahi-dev · 2021-01-30 14:54
marcanis it hard to just make it support x2r10g10b10?#asahi-dev · 2021-01-30 14:54
marcanI mean, we could do that if we needed to#asahi-dev · 2021-01-30 14:54
marcankettenis: I saw corellium were patching that in the display controller#asahi-dev · 2021-01-30 14:49
marcanwell, you need a 1.8V supply for the muart, right? :p#asahi-dev · 2021-01-30 14:46
marcanBluerise: a 220 ohm + 470 ohm divider will give you ~1.2V out of a 1.8V signal#asahi-dev · 2021-01-30 14:46
marcanthen I guess all you need is some 1.8V reference voltage from somewhere#asahi-dev · 2021-01-30 14:45
marcanBluerise: cool :)#asahi-dev · 2021-01-30 14:36
marcanI have it running in a loop incrementing base addresses by 64K, I'll leave it overnight and see if I see patterns tomorrow :)#asahi-dev · 2021-01-30 14:34
marcanarnd: 16K right now (in m1n1)#asahi-dev · 2021-01-30 14:30
marcanthat part is critical#asahi-dev · 2021-01-30 14:30
marcanwith SOP'DEBUG#asahi-dev · 2021-01-30 14:29
marcan1.8V converters are fine (3.3V won't work, threshold is too high) if you do that#asahi-dev · 2021-01-30 14:29
marcanso you would need to run it at 1.8V and use a resistor divider on the TX direction#asahi-dev · 2021-01-30 14:29
marcanthat UART you linked does down to 1.8V#asahi-dev · 2021-01-30 14:29
marcanbut yeah, if you can do compliant DFP negotiation and send SOP'DEBUG/SOP''DEBUG VDMs it should work#asahi-dev · 2021-01-30 14:28
marcanpretty sure it should be possible with a standard type C cable#asahi-dev · 2021-01-30 14:28
marcanactually one thing I need to try out, if I find some time tomorrow, is to make this work with a simple M1 - M1#asahi-dev · 2021-01-30 14:28
marcanalso you need to be a DFP#asahi-dev · 2021-01-30 14:27
marcanincluding sending SOP'DEBUG VDMs?#asahi-dev · 2021-01-30 14:27
marcanBluerise: do you have the USB-PD stuff?#asahi-dev · 2021-01-30 14:26
marcanglasgow in 1.65V mode and a resistor divider for the TX side#asahi-dev · 2021-01-30 14:25
marcanthat screams like a cache/TLB problem to me#asahi-dev · 2021-01-30 14:25
marcanand whether it happens depends on the load address of m1n1#asahi-dev · 2021-01-30 14:25
marcanit's a fun one: leaving that register enabled, I get SError faults (after a few times, an endless loop) after normal page faults - but not always#asahi-dev · 2021-01-30 14:24
marcanI'll test it as soon as I'm done with the current oddysey#asahi-dev · 2021-01-30 07:33
marcancorellium sets that one to zero, disabling all faults, thus shoving all address issues under the carpet :)#asahi-dev · 2021-01-30 07:33
marcandocumenting another mystery register: https://github.com/AsahiLinux/docs/wiki/HW:ARM-System-Registers#sys_l2c_err_sts#asahi-dev · 2021-01-29 18:18
marcangood night ;)#asahi-dev · 2021-01-29 18:18
marcantomorrow smp bringup#asahi-dev · 2021-01-29 18:17
marcanpassing the basic memory/fb info from adt to fdt#asahi-dev · 2021-01-29 18:17
marcanit can boot linux properly now#asahi-dev · 2021-01-29 18:16
marcan(see the commit)#asahi-dev · 2021-01-29 18:16
marcanmight as well, you know, use them#asahi-dev · 2021-01-29 18:15
marcanand it's not about checking, we have the lengths, and I had just added some fancy foreach macros that *do* in facto do it all properly#asahi-dev · 2021-01-29 18:15
marcanmodwizcode: fixed#asahi-dev · 2021-01-29 18:11
marcanoh yeah that's buggy heh#asahi-dev · 2021-01-29 18:09
marcanthe thing is passing around the size to everything is a PITA#asahi-dev · 2021-01-29 18:09
marcanwonder if there's a terminator guaranteed#asahi-dev · 2021-01-29 18:09
marcanyeah it doesn't use the size right now#asahi-dev · 2021-01-29 18:08
marcanI don't hardcode addresses any more except for UART and AIC right now#asahi-dev · 2021-01-29 18:08
marcanas of a few commits back it's required#asahi-dev · 2021-01-29 18:08
marcanoh, well, um, yes, it does kind of expect to have a real devicetree