#
✕ nick:sven
#asahi · 2021-06-06 10:49
svendamnit. i hoped we could ignore most of the usb phy for now inside the kernel
#asahi · 2021-06-06 10:48
svenok. so then we'll probably have to do the phy reset sequence whenever we heard from that tps chip as well
#asahi · 2021-06-06 09:41
sventhat should be in the adt then
#asahi · 2021-06-06 09:41
sven*has
#asahi · 2021-06-06 09:41
svenso it's none of the drd irqs, it's not the i2c irq and the phy itself as not interrupts
#asahi · 2021-06-06 09:40
sveni'm out of ideas now :D
#asahi · 2021-06-06 09:40
svenhrm
#asahi · 2021-06-06 09:38
svenguess it was spamming the i2c bus the whole time ;)
#asahi · 2021-06-06 09:38
svenheh
#asahi · 2021-06-06 09:20
svenstrange
#asahi · 2021-06-06 09:18
sveni think the 16k iommu issue is worse ;)
#asahi · 2021-06-06 09:15
svenso maybe also write 0xff 0xff 0xff... to 0x18 and 0x19? :D
#asahi · 2021-06-06 09:15
svenhm... and then we might also need to ack the irqs once
#asahi · 2021-06-06 09:12
svenoh.. you mean in mac os and see if replugging breaks then as well?
#asahi · 2021-06-06 09:12
svenhow would masking irqs help?
#asahi · 2021-06-06 09:12
svencould be the usb pd i2c interrupt
#asahi · 2021-06-06 09:12
svenyeah, i'm pretty sure about that
#asahi · 2021-06-06 09:09
svenbut i don't know enough yet to figure out a sane way to handle this in linux
#asahi · 2021-06-06 09:09
sveni can't detect unplugging in device mode but the linux xhci driver seems to get disconnect events
#asahi · 2021-06-06 09:09
svenno idea
#asahi · 2021-06-06 09:08
svenbut that reset sequence above seems to be enough at least for usb2
#asahi · 2021-06-06 09:08
svenwell, or if it's replugged. i think mac os actually goes through the whole phy init sequence
#asahi · 2021-06-06 09:08
sven"once in a while" = when the device is unplugged
#asahi · 2021-06-06 09:04
svenfound that yesterday thanks to jannau's mmiotrace :)
#asahi · 2021-06-06 09:03
sveni just don't have any irq to trigger that so far
#asahi · 2021-06-06 09:03
svennot sure if you saw, but set32(0x502a90004, 3); clear32(0x502a90004, 3) fixes it as well
#asahi · 2021-06-06 09:01
svenyeah, but that's also true for the data connection
#asahi · 2021-06-06 09:00
sven(well, except for the usual don't replug the port or everything breaks issue)
#asahi · 2021-06-06 09:00
svenif you boot a linux kernel with just a dwc3 node charging on both ports still works
#asahi · 2021-06-06 08:59
sventhat's... weird
#asahi · 2021-06-06 08:58
sventhat should disable interrupts for that chip. i think.
#asahi · 2021-06-06 08:58
svenjannau: you can try to take my PR and write 8 times 0x00 to address 0x16
#asahi · 2021-06-06 08:57
svenso no
#asahi · 2021-06-06 08:57
svenand the PR i have open just sends the SPSS command
#asahi · 2021-06-06 08:57
sventhe currentl m1n1 doesn't touch that chip at all
#asahi · 2021-06-06 08:57
svenso it tries to handle it over and over again
#asahi · 2021-06-06 08:57
svenit constantly thinks it sees an irq but there isn't one on that chip
#asahi · 2021-06-06 08:56
svenand what pipcet[m] said would explain all of this
#asahi · 2021-06-06 08:56
sventhe IRQ comes from the gpio controller
#asahi · 2021-06-06 08:55
sven(https://www.ti.com/lit/ug/slvuan1a/slvuan1a.pdf?ts=1622969685240&ref_url=https%253A%252F%252Fwww.ti.com%252Fproduct%252FTPS65982)
#asahi · 2021-06-06 08:55
svenlooks like it's reading register 0x14 which is IntEvent1
#asahi · 2021-06-05 23:30
svenit's probably talking to some co-processor
#asahi · 2021-06-05 23:30
svenwhy do you reuse the iop,ascwrap-v4 compatibility for so many things
#asahi · 2021-06-05 23:29
svenokay, wtf apple
#asahi · 2021-06-05 23:29
svenhuh. wait.
#asahi · 2021-06-05 23:29
svendon't think that's nvme
#asahi · 2021-06-05 23:28
sventhat's pmp apparently
#asahi · 2021-06-05 23:28
sveniop,ascwrap-v4
#asahi · 2021-06-05 23:26
svenit's also just running on one efficiency core afaik
#asahi · 2021-06-05 23:24
svenand no idea what pmp[17] is

plus ›