#
✕ nick:jannau
#asahi-dev · 2022-10-08 12:08
jannaustill worth to looking closer at the swap_surface struct
#asahi-dev · 2022-10-08 12:07
jannaunothing obvious in swap_surface struct which could be the display brightness
#asahi-dev · 2022-10-08 12:05
jannauas I wrote above. I currently have no idea how macos controls the brightness. All I see currently in tracing is a surface swap and then callbacks from dcp which look like results of the brightness change
#asahi-dev · 2022-10-08 08:25
jannauI don't see how they would persist after chainloading. I have no problems, only time I see UartTimeouts is when something bad has happened on the system, i.e. after a macos panic for example
#asahi-dev · 2022-10-08 07:06
jannauAny idea what "function-pcc_update = 362:PCCU()" in the disp0 node could be? 362 is the phandle of the backlight node.
#asahi-dev · 2022-10-08 07:00
jannauit doesn't appear to be handled on dcpep/iomfb (0x37). All I see there is surface swap immediately followed what appears to be feedback from a brighness change
#asahi-dev · 2022-10-08 06:58
jannauI'm looking into display brightness control but I have no idea how that's realized
#asahi-dev · 2022-10-08 06:57
jannauotherwise I see a single report/notify with "backlight-calibration-paramters" in it on that endpoint
#asahi-dev · 2022-10-08 06:54
jannaurxbuf/txbuf don't like a dva and rxlen/txlen are random looking large values
#asahi-dev · 2022-10-08 06:53
jannausven: have you used trace_dcp.py on a laptop with your epic changes? endpoint 0x2c blows up in handle_reply with an EPICCmd which doesn't look like one
#asahi-dev · 2022-10-07 19:02
jannaupb17[m]: no, I didn't
#asahi-dev · 2022-10-07 19:02
jannaupovik: it fails since it's missing in the bindings
#asahi-dev · 2022-10-07 19:00
jannaupb17[m]: there is a typo 'reg = <0x3 0xb9030000 ...' should be 0x3 0x9b030000 it's correct in the unit address: fpwm@39b030000
#asahi-dev · 2022-10-07 17:31
jannauChaosPrincess: yes, it makes binds the pins from the pin controller for spi use
#asahi-dev · 2022-10-07 06:28
jannauto be verified on M2 with 12.4
#asahi-dev · 2022-10-07 06:28
jannauI ended up installing 12.4 on t8103 and based on tracing it looks like there are no obvious changes which would prevent using the same driver for 12.3 and 12.4
#asahi-dev · 2022-10-07 06:23
jannauah, I probably have '-v' since I trying to test the usb proxy with m1n1 from the installer
#asahi-dev · 2022-10-06 18:20
jannauthe latest ventura dev beta prints syslog to the display during boot
#asahi-dev · 2022-10-03 20:24
jannauwithout my usb fixes but at least there's some progress
#asahi-dev · 2022-10-03 14:01
jannauoh, I forgot about that
#asahi-dev · 2022-10-02 20:32
jannauJanKral[m]: please do
#asahi-dev · 2022-10-02 14:01
jannaunot sure how to proceed but I git merge should be able to handle commits with the same content gracefully
#asahi-dev · 2022-10-02 13:59
jannaumarcan: the admac dt-bindings iommus/power-domain patch is in dmaengine next: https://git.kernel.org/pub/scm/linux/kernel/git/vkoul/dmaengine.git/log/?h=next
#asahi-dev · 2022-10-02 12:08
jannautheir (v6.0-rcX) kernel config misses cs42l83
#asahi-dev · 2022-10-02 12:08
jannaubut this seems now already deleted. their testing on the m1 macbook air couldn't have that extensive given that they didn't notice that the headphone jack wasn't working
#asahi-dev · 2022-10-02 12:04
jannauchadmed[m]: lol
#asahi-dev · 2022-10-02 11:48
jannauwhat are stolen PKGBUILDs? copied but removed attibution?
#asahi-dev · 2022-10-01 16:21
jannauonly mildly annoying part is writing binary numbers from shell scripts like run_guest_kernel.sh
#asahi-dev · 2022-10-01 16:20
jannauthat makes it simple to support concatenated cpio archives and we can rely on the kernel for decompression (for zstd support)
#asahi-dev · 2022-10-01 16:19
jannaumarcan: any objections over using a custom header with the payload size for initramfs payloads?
#asahi-dev · 2022-10-01 15:23
jannaut8103 supports pro display xdr, so it should support the the studio display
#asahi-dev · 2022-10-01 15:14
jannauat least 2
#asahi-dev · 2022-10-01 15:10
jannaubut shouldn't there then be a second crossbar on the second die and it should be possible to drive 6 6k displays
#asahi-dev · 2022-10-01 14:52
jannauI guess we could try if the front thunderbolt ports support display out
#asahi-dev · 2022-10-01 14:51
jannauno, I think that should be possible with dcpextX,0:dcpextX,1 as dottedmag sees with the tiled thunderbolt display
#asahi-dev · 2022-10-01 14:50
jannauyes, using 2 stream. there are 6 thunderbolt 4 ports so it seems apple might have been lazy wrt to the number of supported displays
#asahi-dev · 2022-10-01 14:47
jannauapple claims for the studio without distinction between max and ultra to support four 6k displays
#asahi-dev · 2022-10-01 14:41
jannau['0x1ff', '0x1ff', '0x1ff', '0x1ff', '0x1ff', '0x1ff']
#asahi-dev · 2022-10-01 14:40
jannausven: all zero: ['0x0', '0x0', '0x0', '0x0', '0x0', '0x0']
#asahi-dev · 2022-10-01 14:33
jannauon both dies
#asahi-dev · 2022-10-01 14:29
jannauultra has dcpext7
#asahi-dev · 2022-10-01 14:27
jannauI think 4 per die on the m1 max
#asahi-dev · 2022-10-01 11:17
jannausven: yes, already fixed. I thought it's marked at some point as not connected
#asahi-dev · 2022-10-01 09:18
jannaua little concering is that the identify boxes are displayed on the hdmi connected display
#asahi-dev · 2022-10-01 09:17
jannaudcpext somehow confuses Xorg or kde. kde's display settings shows a second monitor with an made up 1024x768 mode
#asahi-dev · 2022-10-01 09:16
jannauconfirmed, doesn't happen with macos
#asahi-dev · 2022-10-01 09:12
jannaum1n1
#asahi-dev · 2022-10-01 08:57
jannauworked around that for now by ignoring it
#asahi-dev · 2022-10-01 08:57
jannauthe invalid magic is not the problem. I get tx.cmd's shorter than 64 bytes and empty replies instead of at least 16 bytes which breaks the parsing in handle_reply/handle_cmd
#asahi-dev · 2022-10-01 08:49
jannaunot sure, getting too short replies as well: https://gist.github.com/jannau/bcad09c0b103b14e9f92b94c5055b1a5

plus ›