#
#asahi-dev · 2024-09-24 15:38
nickchanencoded as a short
#asahi-dev · 2024-09-24 15:37
nickchanbrightness 0 - 2047
#asahi-dev · 2024-09-24 15:37
nickchan2 byte write
#asahi-dev · 2024-09-24 15:37
nickchanchaos_princess: figured it out
#asahi-dev · 2024-09-24 15:36
chaos_princessalso, remind me later today, i might have some of my research notes saved
#asahi-dev · 2024-09-24 15:34
chaos_princesshttps://github.com/AsahiLinux/linux/blob/asahi/drivers/gpu/drm/adp/adp_drv.c#L72 - here are the relevant offsets
#asahi-dev · 2024-09-24 15:33
chaos_princessonly one mmio address there
#asahi-dev · 2024-09-24 15:33
chaos_princessin m1 device trees the relevant one is described as mipi-dsim
#asahi-dev · 2024-09-24 15:32
nickchanchaos_princess: yes
#asahi-dev · 2024-09-24 15:31
chaos_princessdo you have a device tree?
#asahi-dev · 2024-09-24 15:30
nickchannavigate layers and layers of helper functions
#asahi-dev · 2024-09-24 15:29
nickchan... which I have no idea how to
#asahi-dev · 2024-09-24 15:19
nickchanso I can do random things in m1n1
#asahi-dev · 2024-09-24 15:19
nickchanokay now i just need to find the mmio that's being poked
#asahi-dev · 2024-09-24 14:55
chaos_princessm1 display should accept a 2 byte command too in theory
#asahi-dev · 2024-09-24 14:55
chaos_princessand fill it from there
#asahi-dev · 2024-09-24 14:55
chaos_princesshttps://github.com/AsahiLinux/linux/blob/bd0a1a7d465fcb60685a2360565ed424bafff354/drivers/gpu/drm/adp/panel-summit.c#L49
#asahi-dev · 2024-09-24 14:55
chaos_princessNo need, just add a of property
#asahi-dev · 2024-09-24 14:55
nickchan"apple,t8015-summit"
#asahi-dev · 2024-09-24 14:54
nickchanI guess I need another compatible then
#asahi-dev · 2024-09-24 14:53
chaos_princessunfortunately, i do not have the pdf of the version of mipi dcs spec that has that command
#asahi-dev · 2024-09-24 14:53
chaos_princessyea, i guess it has extra bits on primary display, and other devices send 2 bytes there https://elixir.bootlin.com/linux/v6.11/source/drivers/gpu/drm/drm_mipi_dsi.c#L1342
#asahi-dev · 2024-09-24 14:51
nickchana11 /arm-io/mipi-dsim/lcd backlight-level-encoding = 11
#asahi-dev · 2024-09-24 14:51
nickchana11 /arm-io/mipi-dsim/lcd = 11
#asahi-dev · 2024-09-24 14:50
nickchanm1 /arm-io/mipi-dsim/lcddfr backlight-level-encoding = 8
#asahi-dev · 2024-09-24 14:50
nickchanchaos_princess: i think the brightness encoding is a bit wrong on a11
#asahi-dev · 2024-09-23 18:44
fl0_idnickchan thx for the explainer
#asahi-dev · 2024-09-23 17:49
nickchans8000 samsung is not vulnerable
#asahi-dev · 2024-09-23 17:48
nickchans8001 / s8003 is first to have it
#asahi-dev · 2024-09-23 17:48
nickchanabout the bug that marcan found where s3_5_c15_c10_1 is accessible from EL0
#asahi-dev · 2024-09-23 17:45
nickchaneven though the individual hardware blocks are still A10 blocks
#asahi-dev · 2024-09-23 17:44
nickchanhow the interrupts and mmios locations are wired up on T2 is also very different from the rest
#asahi-dev · 2024-09-23 17:44
nickchanT2 is very weird it is basically a messed up A10, but due to some issues it doesn't run m1n1 properly yet
#asahi-dev · 2024-09-23 17:43
nickchanA8X, A9X, A10X is M1-class
#asahi-dev · 2024-09-23 17:43
nickchanA7, A8, A9, A10, A11 is the lowest powered chip
#asahi-dev · 2024-09-23 17:43
nickchanfl0_id: right
#asahi-dev · 2024-09-23 17:29
fl0_idnickchan these are mobile chips, right?
#asahi-dev · 2024-09-23 16:27
nickchanokay, sent a PR to m1n1 for a7-a11 support, hopefully the code isn't too garbage
#asahi-dev · 2024-09-23 14:25
tobheguess that means I updated m1n1 as recommended for the last dcp bug
#asahi-dev · 2024-09-23 14:24
tobhethat would be iBoot-11881.1.1 and m1n1 v1.4.14
#asahi-dev · 2024-09-23 14:19
tobheright, sequoia is what I meant. to many names starting with s
#asahi-dev · 2024-09-23 14:16
jannautobhe: what do you mean by sonoma? The changes in m1n1 1.4.15 should only apply to sequoia. please report system fw version ( /proc/device-tree/chosen/asahi,iboot1-version ) amd 1st stage m1n1 version (probably requires running the installer)
#asahi-dev · 2024-09-23 12:05
tobhe_jannau: it looks like I can also trigger the sonoma black screen issue on my m2 air with m1n1 1.4.15. happy to test diffs or provide debug output.
#asahi-dev · 2024-09-23 10:36
nickchanmay send the new version a bit sooner to update some trivial like tags as well
#asahi-dev · 2024-09-23 07:25
nickchanit should also be possible to stick the bootrom in the devicetree as a MTD device on A7-A11 but that's for later
#asahi-dev · 2024-09-23 05:18
LarstiQwhereby 'point this out' doesn't refer to it being weekend ;)
#asahi-dev · 2024-09-23 03:53
nickchanin order to not generate too many emails
#asahi-dev · 2024-09-23 03:51
nickchanI will send a v5 whenever one of these happened: 1. the commits adding the actual dts files get reviewed 2. it is weekend 3. someone points this out on the ML
#asahi-dev · 2024-09-23 03:50
nickchanthe comment about the order is wrong
#asahi-dev · 2024-09-23 03:49
nickchanfwiw in v4 of the device tree series I sent there's a slight error in the commit message on the commit that added a7-a11 cpu cores

plus ›