#asahi-dev · 2022-09-06 13:37
jannauthere's already a PKGBUILD: https://archlinuxarm.org/packages/any/alsa-ucm-conf/files/PKGBUILD#asahi-dev · 2022-09-06 13:09
jannauI'd guess rc4, I haven't seen any conflicts since rc1 so keeping that as base for the bits/branches shouldn't be an issue#asahi-dev · 2022-09-05 22:59
jannaugood night#asahi-dev · 2022-09-05 22:59
jannautoo late, machine with the tree is already powered down#asahi-dev · 2022-09-05 22:56
jannauI've also pushed a refreshed bits/000-devicetree, the non sorted nodes avoided conflicts there so it's not all bad#asahi-dev · 2022-09-05 22:53
jannauiirc marcan was under the same impression#asahi-dev · 2022-09-05 22:42
jannauI plan to send the t600x devicetree out tomorrow based on this https://github.com/jannau/linux/tree/apple-t600x-dts-for-v6.1 (based on sound/for-next for the admac dt-bindings)#asahi-dev · 2022-09-05 20:26
jannaualso it looks lite the nodes are not ordered by address#asahi-dev · 2022-09-05 20:18
jannaupovik: did you forgot to run dtbs_check on "arm64: dts: apple: t8103: Add MCA and its support"? the nco_clkref can not be inside 'soc {}'#asahi-dev · 2022-09-05 17:51
jannauthe dock's display outputs are probably behind a DP MST hub which would explain why only one output works#asahi-dev · 2022-09-05 17:49
jannauI can redo a more rigorous test after dinner#asahi-dev · 2022-09-05 17:49
jannausven: with 4 external displays: https://gist.github.com/jannau/80a8330219920295d534b76c2b4cabec connected and disconnected displays so a little messy#asahi-dev · 2022-09-05 11:20
jannauI'll test tonight with additional displays connected through hdmi/2nd usb-c port#asahi-dev · 2022-09-05 11:04
jannausven: https://gist.github.com/jannau/b9d6826fa649b712463c10f7651dd01d#asahi-dev · 2022-09-05 11:02
jannauthunderbolt + DP works as dual display#asahi-dev · 2022-09-05 11:01
jannaunot as far as I can tell#asahi-dev · 2022-09-05 10:48
jannauHP claims "... a dock that supports up to four 4K displays" I guess that requires DP MST support#asahi-dev · 2022-09-05 10:46
jannautried a different monitor and the instability might be caused by mirroring the same mode to both display#asahi-dev · 2022-09-05 06:51
jannaubuilding the same software with -march=native will be a problem everywhere#asahi-dev · 2022-09-05 06:49
jannauthis is not a problem for other distros since they don't use -march=native to build packages#asahi-dev · 2022-09-05 06:48
jannauthis only becomes a problem when gcc codegen uses instructions from a extension #asahi-dev · 2022-09-05 06:47
jannauproblem is that gcc doesn't pass '-march=...' to the assembler for -march=native. probably because the assembler doesn't support native on aarch64#asahi-dev · 2022-09-04 15:28
jannauheadphone jack works on m1x and m2, mac studio pcie xhci works as well#asahi-dev · 2022-09-04 15:27
jannauapart from the dependencies on devicetree (no conflicts between povik's and my changes) all merges nicely on v6.0-rc3 and works as intended#asahi-dev · 2022-09-04 15:25
jannaupovik: https://github.com/AsahiLinux/linux/pull/50#asahi-dev · 2022-09-04 10:52
jannauis there anything audio related I should include patchset (based on 6.0-rc1)?#asahi-dev · 2022-09-04 10:51
jannaupovik: is your audio devicetree update stable? I'm currently preparing the t600x devicetrees for submission. I was planning to do a PR for bits/000-devicetree including you audio changes#asahi-dev · 2022-09-03 20:24
jannauit was worth a try to see if dcp does something different when spoken to over dcp ep vs. iboot protocol#asahi-dev · 2022-09-03 20:15
jannauI guess that means it can't communicate with the display#asahi-dev · 2022-09-03 20:12
jannauanything interesting in the dcpext dcp ep communication?#asahi-dev · 2022-09-03 20:11
jannauI just checked the regiser I was thinking of is pointing into pmp not pmgr and is not the one requested#asahi-dev · 2022-09-03 20:08
jannau:(#asahi-dev · 2022-09-03 17:50
jannausr_mapDeviceMemoryWithIndex() request the addr/length of the register#asahi-dev · 2022-09-03 17:48
jannaustill has rough edges and crashes dcp if it is already initilized#asahi-dev · 2022-09-03 17:48
jannauworth a shot if that works on dispext. dcp requests a MMIO maaping for pmgr so it could enable the dptx clocks#asahi-dev · 2022-09-03 17:46
jannausven: https://github.com/jannau/m1n1/tree/experiments_dcp_set_mode brings up the display on disp0 with m1n1's display_init() deactivated#asahi-dev · 2022-09-03 14:35
jannausigh, why does the same code (known differences accounted for) that works on the mac studio crashes dcp on the mac mini#asahi-dev · 2022-09-03 11:55
jannaulooks to me that way or is there valid MMIO at 0x493C00000?#asahi-dev · 2022-09-03 11:21
jannaudoes it work on all DARTs with bypass support?#asahi-dev · 2022-09-02 18:31
jannausven: dpxbar trace for a thunderbolt 3 display: https://gist.github.com/jannau/836902d87f115d03c635eeb128b834c9 as far as I can see nothing interesting#asahi-dev · 2022-09-01 22:00
jannaurobher: confirmed, thanks#asahi-dev · 2022-09-01 20:17
jannauI have a fix for that#asahi-dev · 2022-09-01 20:14
jannauthanks. note that t6000-j316s.dts is not in upstream. I'm preparing it for sending it out#asahi-dev · 2022-09-01 19:58
jannauand "arch/arm64/boot/dts/apple/t8103-j274.dtb: pcie@690000000: msi-ranges: [[1], [0]] is too long"#asahi-dev · 2022-09-01 19:57
jannaurobher: arch/arm64/boot/dts/apple/t6000-j316s.dtb: pcie@590000000: msi-ranges: [[1], [0], [0]] is too long#asahi-dev · 2022-09-01 17:56
jannaudoes someone understand how the dts validation for "msi-ranges" for PCIe is supposed to work? it fails for t8102 and t600x. error could be that it just expects a single value for "intspec" instead of 3 or 4 as the #interrupt-cells specify#asahi-dev · 2022-09-01 01:25
jannaudone#asahi-dev · 2022-09-01 01:09
jannauwill send the next interation in a moment#asahi-dev · 2022-09-01 01:09
jannaumarcan: the time since end of june for t6000-dart was my fault. I was very indecisive on which of Robin's review comments I wanted to address#asahi-dev · 2022-08-31 17:51
jannautoo late :(