#
✕ nick:_jannau_
#asahi-dev · 2022-10-05 11:26
_jannau_when starting from athe asahi distro config linked above
#asahi-dev · 2022-10-05 11:25
_jannau_you probably want to change a few things to built-in and disable a lot of modules if you want to use it for developing/usb-booting
#asahi-dev · 2022-10-05 11:24
_jannau_no, that is the arm64 defconfig and produces a 4k kernel. We're also not updating it in asahi (mostly don't care but also to avoid conflicts)
#asahi-dev · 2022-09-30 16:18
_jannau_I can check myself, not really a question which urgently needs an answer. thanks
#asahi-dev · 2022-09-30 16:16
_jannau_does macos offers a display selection if you have more screens connected than dcpexts?
#asahi-dev · 2022-09-30 16:06
_jannau_hardcode it in the beginning. I think it's reasonable to say during development external displays work only on port b
#asahi-dev · 2022-09-30 16:04
_jannau_i.e. it would only somewhat work on m1 max and ultra
#asahi-dev · 2022-09-30 16:04
_jannau_hardcoding doesn't work at all when we have fewer dcpext than ports
#asahi-dev · 2022-09-30 16:01
_jannau_yes, I was confused because I thought you were speaking about the exlectric hotplug
#asahi-dev · 2022-09-30 15:59
_jannau_agreed. did't come the hotplug event before the link training? i.e. you saw it even before the link training was working
#asahi-dev · 2022-09-30 15:57
_jannau_yes
#asahi-dev · 2022-09-30 15:57
_jannau_me too. starting from "I guess we actually only fire that hotplug event after we've already configured the crossbar"
#asahi-dev · 2022-09-30 15:55
_jannau_with hotplug you meant the actual display hotplug. I assumed it meant the electril port hotplug as hint that a dcpext needs to be connected to that port
#asahi-dev · 2022-09-30 15:50
_jannau_but how do we then know when/how to configure the crossbar?
#asahi-dev · 2022-09-30 15:42
_jannau_that looks right
#asahi-dev · 2022-09-30 15:32
_jannau_drm documentaion is not bad, only problem is that all the legacy stuff is documented as well. It would be nice if there was a stripped down version
#asahi-dev · 2022-09-30 15:30
_jannau_the internal display pipeline in drm: see https://www.kernel.org/doc/html/v4.15/gpu/drm-kms.html for some pictures
#asahi-dev · 2022-09-30 15:26
_jannau_mostly so that we do not call the internal display HDMI-A
#asahi-dev · 2022-09-30 15:26
_jannau_but dcp will have drm level connector which might make more sense to bind to dcp
#asahi-dev · 2022-09-30 15:24
_jannau_I guess for dispext it makes more sense if there are connected to the top level node
#asahi-dev · 2022-09-30 15:23
_jannau_I assume the "typec1: connector" nodes from tps6598x should somehow link to the display
#asahi-dev · 2022-09-30 15:21
_jannau_yes. I mean that too. not sure how he connectors are supposed to work
#asahi-dev · 2022-09-30 15:19
_jannau_the disp0 node
#asahi-dev · 2022-09-30 15:18
_jannau_I haven't looked how it is supposed to look in the dt but crtc should map to dcp/dcpext* so not the top level display system node
#asahi-dev · 2022-09-30 15:16
_jannau_I think the connectors should link against the dcps/crtcs in the dt
#asahi-dev · 2022-09-30 15:12
_jannau_I think for dcpext/dispext only dcp text and heap are pre-mapped, so that should be easy to add in m1n1
#asahi-dev · 2022-09-30 15:09
_jannau_sven: ^^^
#asahi-dev · 2022-09-30 15:09
_jannau_yes, asahi-dcp-september. m1n1 changes are in https://github.com/AsahiLinux/m1n1/pull/256
#asahi-dev · 2022-09-30 11:06
_jannau_I'd assume that is used to push ausio data
#asahi-dev · 2022-09-30 11:05
_jannau_dp-audio* has dma-parent/dma-channels
#asahi-dev · 2022-09-29 15:18
_jannau_yes, the drm driver will decide which dcp goes to which port. each dcp instance will be a crtc which can be connected to different outputs
#asahi-dev · 2022-09-29 15:16
_jannau_I guess tps6598x will inform us that there is an dp-altmode capable device
#asahi-dev · 2022-09-29 15:09
_jannau_do you already have an idea how to structure this. DP-xbar and everything dcp related belongs in the display driver
#asahi-dev · 2022-09-29 15:03
_jannau_sven: \o/!!! Did you missing something in some epic endpoint?
#asahi-dev · 2022-09-29 07:16
_jannau_nevermind, you're in #armlinux as well
#asahi-dev · 2022-09-29 07:14
_jannau_marcan: arnd | pulled everything I got for v6.1 now, the only thing that's missing is marcan's DT update
#asahi-dev · 2022-09-27 15:46
_jannau_I assumed it ment copying the firmware into a tmpfs instead of the root fs
#asahi-dev · 2022-09-27 15:37
_jannau_I use it as well to inject matching kernel modules
#asahi-dev · 2022-09-27 15:35
_jannau_I'm using that since a few days for development and it works in practice. that is of course independent of what we do with normal installed systems. I can always repackage the firmware
#asahi-dev · 2022-09-27 15:32
_jannau_and that makes it easy to update the firmware in the rotfs. the initramfs can update it before it switches to os root
#asahi-dev · 2022-09-27 15:30
_jannau_it makes it very easy to provide firmware in the initramfs since linux supports concatenated cpio archives
#asahi-dev · 2022-09-26 15:06
_jannau_otoh dcp would probably crash if an expected mapping doesn't exists
#asahi-dev · 2022-09-26 15:06
_jannau_have you checked the darts? does macos replicate the shared mem mappings iboot does for dcp,disp0 and dcp,piodma? iirc for dispext/dcpext only dcp text and heap are premapped. I could image thsoe mappings are done by driverc and not via dcp fw commands
#asahi-dev · 2022-09-26 14:56
_jannau_mcdp29xx doesn't make any sense except on the macbook pro hdmi connector. hdcp seems unlikely
#asahi-dev · 2022-09-26 12:34
_jannau_does windows for arm supports 16k pages? if not hanling the iommu will get more fun
#asahi-dev · 2022-09-26 12:23
_jannau_dcp on m2 will need ias == 36
#asahi-dev · 2022-09-26 12:22
_jannau_I kept it only for 4k support
#asahi-dev · 2022-09-26 12:17
_jannau_no functional changes but a little bit of sanity checking which should never been hit
#asahi-dev · 2022-09-26 12:16
_jannau_marcan: I would expect a change which prevents using more than 1 ttbr
#asahi-dev · 2022-09-26 12:14
_jannau_there should be minimal io-pgtable-dart changes too but I suppose those will be less problematic as well as we are more ore less solely responsible for it

plus ›