#
✕ nick:sven
#asahi-dev · 2022-11-21 14:51
svenit's completely harmless
#asahi-dev · 2022-11-21 14:50
svenyou can drop 73d651c07433712e97cb0c7121e8d50492397639 since 6d8fc203b28ff8f6115fbe5eaf584de8b824f4fa is upstream now fwiw
#asahi-dev · 2022-11-21 10:58
svenmarcan: btw. can you review and send https://lore.kernel.org/asahi/20221104154153.79373-1-sven@svenpeter.dev/T/#u and https://lore.kernel.org/asahi/20221104153902.78999-1-sven@svenpeter.dev/T/#u to arnd?
#asahi-dev · 2022-11-20 21:49
svenplease move this to #asahi
#asahi-dev · 2022-11-20 20:45
svenbut I don’t remember what it was
#asahi-dev · 2022-11-20 20:45
svenI remember seeing that somewhere else as well
#asahi-dev · 2022-11-20 20:38
svenlooks like I can close it :D
#asahi-dev · 2022-11-20 19:59
svennice!
#asahi-dev · 2022-11-20 18:23
svenbut essentially only devices that support >32 bit dma should do that and those should all work fine if the iommu also supports that I think
#asahi-dev · 2022-11-20 18:21
svenbut dwc3 will do that
#asahi-dev · 2022-11-20 18:21
svenyes
#asahi-dev · 2022-11-20 18:21
svennice!
#asahi-dev · 2022-11-20 18:12
svenluckily the iova allocator starts handing out allocations at the very end of its aperture so it'll be very obvious if this breaks anything else :D
#asahi-dev · 2022-11-20 18:10
svenif so i'd just set the domain aperture to that for the M2 darts
#asahi-dev · 2022-11-20 18:08
svendo all DARTs on the M2 support 36 bit? or is there a register where you can get the aperture from?
#asahi-dev · 2022-11-20 18:07
svendidn't the pcie ones have that as well?
#asahi-dev · 2022-11-20 18:07
svencan't you mark that region as reserved if it actually causes problems?
#asahi-dev · 2022-11-20 18:06
svenquite a few darts have those vm-start/vm-size properties and so far we just ignore all of them
#asahi-dev · 2022-11-20 18:05
svenbut if it's not required i'd just skip it
#asahi-dev · 2022-11-20 18:05
svenI don't think there's any documented property for that. iirc there is one iommu that has a custom property ("xxxx,dma-window" maybe)
#asahi-dev · 2022-11-20 18:03
svenhrm, so you actually need to enforce vm-start and vm-size?
#asahi-dev · 2022-11-20 17:55
svenah, i remember "complaining" about that ;)
#asahi-dev · 2022-11-20 17:54
svenwhy will they make me happy? :D
#asahi-dev · 2022-11-20 17:54
sven\o/
#asahi-dev · 2022-11-19 12:56
svenI bet a role switch event gets lost
#asahi-dev · 2022-11-19 12:56
svenyeah, looks like a very quick unplug/plug/unplug cycle that the mini somehow does triggers it
#asahi-dev · 2022-11-19 12:50
sven404
#asahi-dev · 2022-11-19 12:49
svenlooks it sees unplug/plug/unplug very quickly there
#asahi-dev · 2022-11-19 12:35
svenmaybe it was just the usb curse then
#asahi-dev · 2022-11-19 12:35
svenhuh
#asahi-dev · 2022-11-19 12:23
svenI suspect that somehow m1n1 in device mode triggers an additional phy_power_on that’s not matched with a phy_power_off fwiw
#asahi-dev · 2022-11-19 12:16
svencan you paste the entire dmesg since boot with those prints?
#asahi-dev · 2022-11-19 12:14
svenand xhci did shut down just before, didn’t it?
#asahi-dev · 2022-11-19 12:13
svenit looks like there’s an imbalanced call to phy_power_on somewhere
#asahi-dev · 2022-11-19 12:12
svenjust to confirm, this is the latest asahi-wip?
#asahi-dev · 2022-11-19 12:11
svenjannau: okay, so it looks like for some reason the power off callback is never called from the phy core
#asahi-dev · 2022-11-19 12:05
svenjannau: can you paste the exact location you put the prints for the atcphy issue?
#asahi-dev · 2022-11-19 12:00
svendidn’t one coprocessor need a special message that would break the others? or was that for shutdown?
#asahi-dev · 2022-11-19 11:17
svenso I guess we still do something slightly different when talking to it :/
#asahi-dev · 2022-11-19 11:16
svenhuh. that sounds weird. I’d expect smc to be fully backwards compatible
#asahi-dev · 2022-11-19 11:04
svenI “just” want thunderbolt so that I can use my dock :<
#asahi-dev · 2022-11-18 23:45
svenwasn’t your official stance that DCP is all jannau? :p
#asahi-dev · 2022-11-18 22:39
svenyeah, let’s please blame it all on dcpext and not atcphy :p
#asahi-dev · 2022-11-18 22:05
sven\o/
#asahi-dev · 2022-11-18 22:00
svenugh. that looks like it doesn’t get the atcphy mutex ij power off for some reason :/
#asahi-dev · 2022-11-18 15:27
svenyeah, I don’t think we’ve ever needed an update to stage 1 so far. fingers crossed it stays that way ;)
#asahi-dev · 2022-11-18 09:06
svenhrm, i assume no one worked on getting these VDM commands to reboot another mac supported in linux yet?
#asahi-dev · 2022-11-18 08:26
svenok. Would be interesting to see when the phy callbacks actually run because I have no idea why mux_set doesn’t see that dwc3 is already down
#asahi-dev · 2022-11-18 08:09
svenjannau: if you can reproduce this reliably, can you add tracepoints or prints to phy_power_on/off in the core phy code and inside atcphy?
#asahi-dev · 2022-11-17 22:09
svenso for some reason phy_power_off isn’t called

plus ›