#asahi-dev · 2023-03-02 12:09
jannauprobably ok as workaround as long as we don't support it. A little annoying when support is added as the gpio is on at boot#asahi-dev · 2023-03-02 12:07
jannauskipping the port setup for ports with "pwren-gpios" in u-boot fixes the usb hci#asahi-dev · 2023-03-02 10:52
jannaucycling pwren is not enough I guess the pcie port needs to be reset as well to recover#asahi-dev · 2023-03-02 10:47
jannauBus xhci_pci: Register 400087f NbrPorts 4#asahi-dev · 2023-03-02 10:47
jannauyes#asahi-dev · 2023-03-02 10:42
jannaunot missing on j375#asahi-dev · 2023-03-02 10:39
jannauthe asmedia hci still doesn't work. the link is not detected as up. I'll try to cycle its missing pwren-gpio#asahi-dev · 2023-03-02 10:37
jannaushort circuiting #PERST if the link is already up works for ethernet#asahi-dev · 2023-03-02 09:41
jannauit does rmw_set(PORT_PERST_OFF, port->base + PORT_PERST); but never clears the bit#asahi-dev · 2023-03-02 09:35
jannauwhat do you mean by perst cycle? the port setup in apple_pcie_setup_port() or something else?#asahi-dev · 2023-03-02 09:25
jannaufor pci@3,0 as well#asahi-dev · 2023-03-02 09:24
jannaujust "pcie-apple 590000000.pcie: /soc/die0/pcie@590000000/pci@2,0 link didn't come up"#asahi-dev · 2023-03-02 09:14
jannauusb hci and ethernet work with PCI_APPLE disabled#asahi-dev · 2023-03-02 08:47
jannaukettenis: u-boot with pcie support doesn't fail over the firmware less asmedia usb hci but possibly breaks the pcie links for the hci and ethernet adapter on the studio#asahi-dev · 2023-03-01 21:18
jannaulooking for examples how to model display ports for dcp in the devicetree#asahi-dev · 2023-03-01 21:17
jannaukettenis: do you know if qualcomm's 8cx gen3 has usb4 ports? I see in bindings/clock/qcom,gcc-sc8280xp.yaml USB4 DP/PCIe phy clocks but thinkpad x13s, ms window dev kit and surface pro 9 list only usb 3.2 with dp-altmode#asahi-dev · 2023-03-01 10:27
jannauprobably both. it's not common so no out of the box fix but it's easy to find the relevant kernel parameters offered as fix/workaround #asahi-dev · 2023-03-01 10:23
jannaudecember 2020, good luck#asahi-dev · 2023-02-28 22:17
jannauwould anyone object making external display's and possible dcp-audio 13.2 only features?#asahi-dev · 2023-02-28 22:15
jannauannoying that the dptx service names in 13.2 are now "dispext%u:dcpdptx-port-epic:%u"#asahi-dev · 2023-02-28 22:14
jannauis still amused by 'dispext4294967295:dcpdptx-port-' for disp0 compared to 'dispext0:dcpdptx-port-epic:0' for dispext0#asahi-dev · 2023-02-26 15:20
jannaumarcan: we also have https://github.com/AsahiLinux/linux/pull/106 still open. shall I recreate that in gitlab?#asahi-dev · 2023-02-26 15:18
jannaumarcan: yes, it appears so. 200-dcp should be good up to 125aeecbc011d but it saw little testing#asahi-dev · 2023-02-26 15:16
jannausven: order of operations in atcphy? macos and linux traces looked different#asahi-dev · 2023-02-24 15:05
jannaukettenis: it's transmitted during video blanking and max channels, samplerate and resolution is specified. I missed that it was increased to 32 channels in dp 1.4#asahi-dev · 2023-02-24 14:51
jannaubandwidth (you can have more than 7.1 channels) and you need missing knowledge of the speaker config (and room) to mix it correctly#asahi-dev · 2023-02-24 14:28
jannaumsm, mediatek or rockchip might be better eaxample with displayport/usb-altmode support#asahi-dev · 2023-02-24 14:28
jannauvc4 is probably bad example since it's display out is afaik hdmi only and has no dynamic switching#asahi-dev · 2023-02-24 14:07
jannauah, that makes sense. naming the cards after their dcp(ext) would work then#asahi-dev · 2023-02-24 14:05
jannaunot even sure if the distinction between usb-c and thunderbolt makes sense#asahi-dev · 2023-02-24 14:04
jannauso probably only three distinct options: hdmi, usb-c, thunderbolt#asahi-dev · 2023-02-24 14:03
jannauI think it would make more sense to have dynamic names based on which port dcp is muxed to#asahi-dev · 2023-02-24 14:00
jannauso derived from EDID#asahi-dev · 2023-02-24 13:58
jannau?#asahi-dev · 2023-02-24 13:58
jannauthat's the display's name#asahi-dev · 2023-02-24 13:54
jannauthat sounds liek we should try to keep paerser with drm/apple#asahi-dev · 2023-02-24 13:51
jannaunaming is a little weird since the mapping between dcp and output port will become deynamic#asahi-dev · 2023-02-24 13:50
jannaumaybe afk/epic/parser should be moved to drivers/soc/apple if the audio driver depends on parser#asahi-dev · 2023-02-24 13:42
jannauI guess we need to start thinking about how we model the output ports for the display subsystem in the device tree#asahi-dev · 2023-02-24 13:40
jannauhow does that work from macaudio PoV. do we a single display audio output device or one for each dcp?#asahi-dev · 2023-02-24 13:39
jannaupovik: \o/#asahi-dev · 2023-02-23 21:47
jannaudcp and gpu use more than that but they might coalesce into 8 regions or less#asahi-dev · 2023-02-23 21:46
jannaumaybe we're just running out of regions? CONFIG_LMB_MAX_REGIONS seems to default to 8#asahi-dev · 2023-02-23 21:40
jannauerr, I can't read#asahi-dev · 2023-02-23 21:26
jannausee boot_fdt_add_mem_rsv_regions() in boot/image-fdt.c#asahi-dev · 2023-02-23 21:26
jannaupovik: there has to be something wrong with "no-map". the function producing the error message should not be called#asahi-dev · 2023-02-23 21:18
jannauno, we don't do anything for the carveouts there. see fdt_add_mem_rsv()#asahi-dev · 2023-02-23 21:14
jannauthat would exclude reserved memory from the carveouts too but there's the reserved memory map (not visible)#asahi-dev · 2023-02-23 19:39
jannauI'm not sure I looked if u-boot complains about the reallocated framebuffer. not you need to manually configure a hgiher display resolution for m1n1 to trigger the reallocation#asahi-dev · 2023-02-23 19:36
jannaudoes it boot to linux? I find inspecting the dtb with `dtc -I fs -O dts -o - /proc/device-tree | less -I` helpful