Skip to content
Snippets Groups Projects
New look: Off

linux-postmarketos-allwinner: external HDMI is broken

    • Closed Imported Issue created by Administrator @root

    @MartijnBraam and I are looking into it right now, trying to resolve it ASAP.

    To upload designs, you'll need to enable LFS and have an admin enable hashed storage. More information

    Child items 0

    • No child items are currently assigned. Use child items to break down this issue into smaller parts.

    Activity

    • Administrator changed milestone to %[device] PinePhone support · Imported

      changed milestone to %[device] PinePhone support

    • Administrator added type::bug label · Imported

      added type::bug label

    • Administrator mentioned in merge request !1523 (closed) · Imported

      mentioned in merge request !1523 (closed)

      By Oliver Smith on 2020-09-01T14:46:09

    • Administrator mentioned in merge request !1524 (closed) · Imported

      mentioned in merge request !1524 (closed)

      By Oliver Smith on 2020-09-01T14:59:33

      • Administrator
        Administrator @root · Imported
        Author Owner

        Fix is merged. Binary packages for v20.05 branch and master are building now. I'll close this issue as soon as people can just hit "update" on their phones and install the fix.

        By Oliver Smith on 2020-09-02T09:03:30

        Edited by Ghost User
      • Administrator
        Administrator @root · Imported
        Author Owner

        Nice. Was wondering what was wrong with it!

        By nolim1t - f6287b82CC84bcbd on 2020-09-02T09:03:30

      • Administrator
        Administrator @root · Imported
        Author Owner

        Not working unfortunately, but is true that with megi’s kernel works at least for Gnome and Phosh. Even with such monster like in attach ;)EE907486-EEEF-4652-8887-7937404AC6A5

        By Jerzy Skafander on 2020-09-23T18:15:39

      • Please register or sign in to reply
      • Administrator
        Administrator @root · Imported
        Author Owner

        The binary packages are published for the stable channel. Unfortunately, it appears to be only working on some monitors and not all of them.

        • It works for @z3ntu on an "AOC Q3279VWFD8" monitor, see his photo: https://fosstodon.org/@z3ntu/104790471532689018
        • @MartijnBraam reported that it's working on the HDMI port of his DELL monitor, but not on another 10" monitor.
        • It does not work on my Philips monitor.

        More feedback whether it works for you or not is appreciated (just reply to this specific message).

        Newer versions of the kernel with megi's patches should have better HDMI support, but we will need to adjust some things to keep the modem working. This is the path forward, but it will take some time to be done properly, and to have everything tested before we ship it as update.

        (Before anybody complains... we wrote "for Linux enthusiasts" all over, for a reason. This is a community project, we are doing our best with the few resources we have and help is appreciated.)

        By Oliver Smith on 2020-09-01T20:02:43

        Edited by Administrator
      • Administrator
        Administrator @root · Imported
        Author Owner

        Tested on:

        • Vizio TV via USB-C dock that shipped with the convergence edition PinePhone
        • NexDock 2 via USB-C cable (tried two different cables to be sure)

        ...after an apk update && apk upgrade on my PinePhone. Neither seemed to work; the phone display appeared to switch to a console after plugging into the NexDock, but I didn't get output on the NexDock's display.

        By Jonathan Frederickson on 2020-09-01T20:02:43

      • Administrator
        Administrator @root · Imported
        Author Owner

        Tried on my Coby 19" with HDMI and it did not work.

        By ChrisM3568 on 2020-09-02T13:52:31

      • Administrator
        Administrator @root · Imported
        Author Owner

        Tested on my 4K Philips 43PUS6162 TV, it does not seem to be recognized at all.

        By Bart Ribbers on 2020-09-02T17:31:19

      • Administrator
        Administrator @root · Imported
        Author Owner

        I have tested it on a BenQ GW2480, it showed the wallpaper for 3 seconds then stopped working. I also tested it on ASUS and Dell monitors with non of them working at all. It worked perfectly on an ACER SB220Q with no issues.

        By mhovet on 2020-09-02T20:40:13

      • Administrator
        Administrator @root · Imported
        Author Owner

        After updating to the latest software as of 2020-09-05 in the Software GUI, I get zero output on a Samsung LN46D630. It also causes a total and complete display crash on the main touchscreen (down to '^@' and high-pitch mono speaker scream -- oof)

        Not much log output on the dmesg before that. Something gets set to 0x00.

        By bujiraso on 2020-09-05T15:51:48

        Edited by Administrator
      • Administrator
        Administrator @root · Imported
        Author Owner

        Also works on a Philips Brilliance 258B (1440p) and doesn't work on a 1080p AOC

        By Luca Weiss on 2020-09-08T19:41:29

        Edited by Administrator
      • Administrator
        Administrator @root · Imported
        Author Owner

        I updated my recently purchased convergence Pinephone and tried a number of monitors with no success: Dell U2212 xwide and Compaq 5017m.

        When plugging in the usb-c with the hdmi monitor attached, the monitor will flicker and the phone's screen goes blank, looking hopeful for a few seconds, sometimes there's two or three odd at-sigh/@ characters in the upper left of the phone's screen. Then the monitor goes blank and the phone's screen returns with the Postmarketos x/gui interface. So tantalizingly close!!

        ALSO: When re-starting the phone with the usb-c-via-hdmi monitor attached the external monitor actual works in the initial text mode! See attached screen shot from the Compaq 5017m monitor, with the Pinephones' encrypted pw prompt. But as soon as the Postmarketos x/gui comes up the monitor goes blank and only the phone's screen displays.

        Hope that helps, best of luck in finding a solution!!

        For what it's worth below is the dmesg output that's logged when I plug the usb-c/hdmi monitor in:

        pp_hdmi_startup_text_screen_works_v3

        Dmesg output that's logged when I plug the usb-c/hdmi monitor in:

        `

        [ 169.594859] anx7688 1-0028: plug irq (cd=1)

        [ 169.607491] anx7688 1-0028: cable inserted

        [ 169.647505] anx7688 1-0028: power enabled

        [ 169.711915] anx7688 1-0028: eeprom0 = 0x03

        [ 169.711934] anx7688 1-0028: fw loaded after 40 ms

        [ 169.712478] anx7688 1-0028: OCM firmware loaded (version 0x2312)

        [ 169.715690] anx7688 1-0028: send pd packet cmd=0 05 00 32 90 01 26 12

        [ 169.762140] anx7688 1-0028: send pd packet cmd=1 05 01 2c 91 01 26 16

        [ 169.764012] anx7688 1-0028: send pd packet cmd=2 11 02 00 00 00 ec 00 00 00 00 00 00 00 00 39 00 00 51 77

        [ 169.766872] anx7688 1-0028: send pd packet cmd=3 05 03 00 00 01 ff f8

        [ 169.768612] anx7688 1-0028: OCM configuration completed

        [ 169.771956] anx7688 1-0028: status changed to 0x00

        [ 169.771973] anx7688 1-0028: cc_status changed to CC1 = SRC.Open CC2 = SRC.Open

        [ 169.771983] anx7688 1-0028: dp state changed to 0x00

        [ 169.772008] anx7688 1-0028: data role change requested to ufp

        [ 170.203236] anx7688 1-0028: status irq

        [ 170.204832] anx7688 1-0028: soft = 0x38

        [ 170.206089] anx7688 1-0028: cc_status changed to CC1 = SNK.Power3.0 CC2 = SRC.Open

        [ 170.278375] anx7688 1-0028: status irq

        [ 170.279989] anx7688 1-0028: soft = 0x01

        [ 170.283594] anx7688 1-0028: recv ocm message cmd=0 05 00 c8 90 01 2e 74

        [ 170.284073] anx7688 1-0028: status irq

        [ 170.285644] anx7688 1-0028: soft = 0x01

        [ 170.289238] anx7688 1-0028: recv ocm message cmd=240 03 f0 16 00 f7

        [ 170.313228] anx7688 1-0028: status irq

        [ 170.314845] anx7688 1-0028: soft = 0x40

        [ 170.891495] anx7688 1-0028: status irq

        [ 170.893176] anx7688 1-0028: soft = 0x21

        [ 170.896858] anx7688 1-0028: recv ocm message cmd=17 01 11 ee

        [ 170.898160] anx7688 1-0028: status changed to 0x20

        [ 170.898297] anx7688 1-0028: data role change requested to dfp

        [ 170.898337] phy id_det change to host

        [ 170.945317] anx7688 1-0028: status irq

        [ 170.946991] anx7688 1-0028: soft = 0x05

        [ 170.950647] anx7688 1-0028: recv ocm message cmd=240 03 f0 13 00 fa

        [ 170.951946] anx7688 1-0028: status changed to 0x24

        [ 170.969747] anx7688 1-0028: status irq

        [ 170.971370] anx7688 1-0028: soft = 0x01

        [ 170.975027] anx7688 1-0028: recv ocm message cmd=2 11 02 39 06 00 6c 00 00 00 00 12 02 20 a0 db 00 00 00 93

        [ 170.975560] anx7688 1-0028: status irq

        [ 170.977175] anx7688 1-0028: soft = 0x01

        [ 170.980825] anx7688 1-0028: recv ocm message cmd=3 09 03 39 06 01 ff 00 00 00 00 b5

        [ 170.982024] anx7688 1-0028: status irq

        [ 170.983640] anx7688 1-0028: soft = 0x01

        [ 170.987307] anx7688 1-0028: recv ocm message cmd=39 09 27 03 80 01 ff 05 0c 00 00 3c

        [ 170.988189] anx7688 1-0028: status irq

        [ 170.989816] anx7688 1-0028: soft = 0x01

        [ 170.993469] anx7688 1-0028: recv ocm message cmd=25 01 19 e6

        [ 171.131526] usb 1-1: new high-speed USB device number 8 using ehci-platform

        [ 171.289149] hub 1-1:1.0: USB hub found

        [ 171.289278] hub 1-1:1.0: 4 ports detected

        [ 171.675488] usb 1-1.1: new full-speed USB device number 9 using ehci-platform

        [ 172.067511] usb 1-1.2: new high-speed USB device number 10 using ehci-platform

        [ 172.467566] usb 1-1.2: reset high-speed USB device number 10 using ehci-platform

        [ 172.706213] r8152 1-1.2:1.0: skip request firmware

        [ 172.737402] r8152 1-1.2:1.0 eth0: v1.11.11

        `

        By Tom Ja on 2020-09-15T14:06:42

        Edited by Administrator
      • Administrator
        Administrator @root · Imported
        Author Owner

        Hello, thanks for your work on this! Another report:

        Recently upgraded to Linux pinephone 5.8.0 #4 (closed)-postmarketos-allwinner via the apk upgrade.

        Still no luck with the external monitor, when plugged in the phone goes blank for a few seconds, but then returns, no display on the Dell U2212 (native: 1920x1080) monitor.

        On Discord there's mention of 'megi's kernel having fixed this for PMos: ( Install megi's kernel... or use pboot to test out postmarketOS with a 5.9 kernel https://megous.com/git/p-boot/tree/README )

        Will these 5.9 changes be coming to the official PM apk repo?

        Thanks again! FWIW below is the dmesg output when plugging in the external monitor under 5.8:

        [126635.233364] anx7688 1-0028: plug irq (cd=1)

        [126635.248345] anx7688 1-0028: cable inserted

        [126635.288352] anx7688 1-0028: power enabled

        [126635.352798] anx7688 1-0028: eeprom0 = 0x03

        [126635.352810] anx7688 1-0028: fw loaded after 40 ms

        [126635.353345] anx7688 1-0028: OCM firmware loaded (version 0x2312)

        [126635.356476] anx7688 1-0028: send pd packet cmd=0 05 00 32 90 01 26 12

        [126635.401580] anx7688 1-0028: send pd packet cmd=1 05 01 2c 91 01 26 16

        [126635.403310] anx7688 1-0028: send pd packet cmd=2 11 02 00 00 00 ec 00 00 00 00 00 00 00 00 39 00 00 51 77

        [126635.406163] anx7688 1-0028: send pd packet cmd=3 05 03 00 00 01 ff f8

        [126635.407901] anx7688 1-0028: OCM configuration completed

        [126635.410494] anx7688 1-0028: status changed to 0x00

        [126635.410512] anx7688 1-0028: cc_status changed to CC1 = SRC.Open CC2 = SRC.Open

        [126635.410522] anx7688 1-0028: dp state changed to 0x00

        [126635.410548] anx7688 1-0028: data role change requested to ufp

        [126635.844040] anx7688 1-0028: status irq

        [126635.845615] anx7688 1-0028: soft = 0x38

        [126635.846868] anx7688 1-0028: cc_status changed to CC1 = SRC.Open CC2 = SNK.Power3.0

        [126635.919985] anx7688 1-0028: status irq

        [126635.926027] anx7688 1-0028: soft = 0x01

        [126635.930463] anx7688 1-0028: recv ocm message cmd=0 05 00 c8 90 01 2e 74

        [126636.829817] anx7688 1-0028: status changed to 0x24

        [126636.830085] anx7688 1-0028: data role change requested to dfp

        [126636.830183] phy id_det change to host

        [126637.068389] usb 1-1: new high-speed USB device number 33 using ehci-platform

        [126637.226675] hub 1-1:1.0: USB hub found

        [126637.227024] hub 1-1:1.0: 4 ports detected

        [126637.612341] usb 1-1.1: new full-speed USB device number 34 using ehci-platform

        [126638.012383] usb 1-1.2: new high-speed USB device number 35 using ehci-platform

        [126638.412450] usb 1-1.2: reset high-speed USB device number 35 using ehci-platform

        [126638.596518] rfkill: input handler enabled

        [126638.650428] r8152 1-1.2:1.0: skip request firmware

        [126638.682482] r8152 1-1.2:1.0 eth0: v1.11.11

        [126638.877704] anx7688 1-0028: dp state changed to 0x04

        [126640.001534] elogind-daemon[2641]: Removed session c10.

        [126640.243533] elogind-daemon[2641]: Failed to create inotify watch on /dev/null/utmp, ignoring: Not a directory [126640.244911] elogind-daemon[2641]: New session c11 of user user.

        [126640.925640] anx7688 1-0028: dp state changed to 0x05

        [126642.973579] anx7688 1-0028: dp state changed to 0x04

        [126643.997586] anx7688 1-0028: dp state changed to 0x05

        [126644.803658] rfkill: input handler disabled

        By Tom Ja on 2020-09-18T19:42:32

        Edited by Administrator
      • Please register or sign in to reply
    • Administrator added category::device label · Imported

      added category::device label

    • Administrator
      Administrator @root · Imported
      Author Owner

      Ive tested on my Philips 4K TV and also on a Fujitsu monitor which sway lists as "Fujitsu Siemens Computers GmbH B22W-6 LED pG YV4G149090".

      None of them manage to get output working on my PinePhone pmOS CE 3GB edition and supplied USB-C dock.

      Before anybody complains... we wrote "for Linux enthusiasts" all over, for a reason

      Just to be clear: Not complaining. This is clearly an enthusiast project, and I'm just trying to help as I can, by reporting in what I can reproduce with the equipment I have :smile:

      By Jostein Kjønigsen on 2020-09-02T09:06:41

      Edited by Administrator
    • Administrator
      Administrator @root · Imported
      Author Owner

      Tried on my dell D2719HGF with no luck. Also a smaller 24in dell monitor didn't work either. Nor did my Samsung S23C570.

      By Izaic on 2020-09-02T04:06:19

      Edited by Administrator
    • Administrator
      Administrator @root · Imported
      Author Owner

      I now tested on another monitor (Samsung S24C450) and I'm still getting the same behaviour: no output.

      This is what I'm seeing in dmesg when connecting and disconnecting the hub with the HDMI-cable already connected:

      On connect:

      [ 1705.104806] anx7688 1-0028: plug irq (cd=1)
[ 1705.120387] anx7688 1-0028: cable inserted
[ 1705.160413] anx7688 1-0028: power enabled
[ 1705.224882] anx7688 1-0028: eeprom0 = 0x03
[ 1705.224902] anx7688 1-0028: fw loaded after 40 ms
[ 1705.225444] anx7688 1-0028: OCM firmware loaded (version 0x2312)
[ 1705.228645] anx7688 1-0028: send pd packet cmd=0 05 00 32 90 01 26 12
[ 1705.273716] anx7688 1-0028: send pd packet cmd=1 05 01 2c 91 01 26 16
[ 1705.275473] anx7688 1-0028: send pd packet cmd=2 11 02 00 00 00 ec 00 00 00 00 00 00 00 00 39 00 00 51 77
[ 1705.278370] anx7688 1-0028: send pd packet cmd=3 05 03 00 00 01 ff f8
[ 1705.280117] anx7688 1-0028: OCM configuration completed
[ 1705.281715] anx7688 1-0028: status changed to 0x00
[ 1705.281732] anx7688 1-0028: cc_status changed to CC1 = SRC.Open CC2 = SRC.Open
[ 1705.281741] anx7688 1-0028: dp state changed to 0x00
[ 1705.376672] anx7688 1-0028: status irq
[ 1705.378320] anx7688 1-0028:   soft = 0x0c
[ 1705.379622] anx7688 1-0028: status changed to 0x0c
[ 1705.716269] anx7688 1-0028: status irq
[ 1705.717964] anx7688 1-0028:   soft = 0x30
[ 1705.719228] anx7688 1-0028: status changed to 0x2c
[ 1705.719234] anx7688 1-0028: cc_status changed to CC1 = SRC.Ra CC2 = SRC.Rd
[ 1705.719347] anx7688 1-0028: data role change requested to dfp
[ 1705.719378] phy id_det change to host
[ 1705.952390] usb 1-1: new high-speed USB device number 2 using ehci-platform
[ 1706.110982] hub 1-1:1.0: USB hub found
[ 1706.111173] hub 1-1:1.0: 4 ports detected
[ 1706.229040] anx7688 1-0028: status irq
[ 1706.230766] anx7688 1-0028:   soft = 0x01
[ 1706.234443] anx7688 1-0028: recv ocm message cmd=22 05 16 32 c8 00 12 d9
[ 1706.359877] anx7688 1-0028: status irq
[ 1706.361566] anx7688 1-0028:   soft = 0x01
[ 1706.365225] anx7688 1-0028: recv ocm message cmd=1 05 01 2c 91 01 05 37
[ 1706.382478] anx7688 1-0028: status irq
[ 1706.384081] anx7688 1-0028:   soft = 0x01
[ 1706.387803] anx7688 1-0028: recv ocm message cmd=2 11 02 39 06 00 6c 00 00 00 00 12 02 20 a0 db 00 00 00 93
[ 1706.388382] anx7688 1-0028: status irq
[ 1706.390012] anx7688 1-0028:   soft = 0x01
[ 1706.393671] anx7688 1-0028: recv ocm message cmd=3 09 03 39 06 01 ff 00 00 00 00 b5
[ 1706.395111] anx7688 1-0028: status irq
[ 1706.396744] anx7688 1-0028:   soft = 0x01
[ 1706.400402] anx7688 1-0028: recv ocm message cmd=39 09 27 03 80 01 ff 05 0c 00 00 3c
[ 1706.401425] anx7688 1-0028: status irq
[ 1706.403042] anx7688 1-0028:   soft = 0x01
[ 1706.406697] anx7688 1-0028: recv ocm message cmd=25 01 19 e6
[ 1706.500415] usb 1-1.1: new full-speed USB device number 3 using ehci-platform
[ 1707.040436] usb 1-1.2: new high-speed USB device number 4 using ehci-platform
[ 1707.312554] usbcore: registered new interface driver r8152
[ 1707.322729] usbcore: registered new interface driver cdc_ether
[ 1707.504488] usb 1-1.2: reset high-speed USB device number 4 using ehci-platform
[ 1707.747668] r8152 1-1.2:1.0: skip request firmware
[ 1707.778474] r8152 1-1.2:1.0 eth0: v1.11.11
[ 1708.725781] anx7688 1-0028: dp state changed to 0x04
[ 1709.749768] anx7688 1-0028: dp state changed to 0x05

      On disconnect:

      [ 1754.411401] usb 1-1: USB disconnect, device number 2
[ 1754.411420] usb 1-1.1: USB disconnect, device number 3
[ 1754.412289] usb 1-1.2: USB disconnect, device number 4
[ 1754.414842] r8152 1-1.2:1.0 eth0: Stop submitting intr, status -108
[ 1754.425217] anx7688 1-0028: status irq
[ 1754.431055] anx7688 1-0028:   soft = 0x0c
[ 1754.432505] anx7688 1-0028: status changed to 0x20
[ 1755.094527] anx7688 1-0028: plug irq (cd=0)
[ 1755.095160] anx7688 1-0028: status irq
[ 1755.096867] anx7688 1-0028:   soft = 0x20
[ 1755.098164] anx7688 1-0028: status changed to 0x00
[ 1755.098320] anx7688 1-0028: data role change requested to ufp
[ 1755.098365] phy id_det change to device
[ 1755.108006] anx7688 1-0028: cable removed
[ 1755.124078] anx7688 1-0028: power disabled

      By Jostein Kjønigsen on 2020-09-02T16:17:42

    • Administrator
      Administrator @root · Imported
      Author Owner

      When plugging in HDMI, I get the following from dmesg

      [  206.207721] anx7688 1-0028: dp state changed to 0x00
[  211.327774] anx7688 1-0028: dp state changed to 0x03
[  211.796689] rfkill: input handler enabled
[  213.150112] elogind-daemon[2754]: Removed session c4.
[  213.363936] elogind-daemon[2754]: Failed to create inotify watch on /dev/null/utmp, ignoring: Not a directory
[  213.365214] elogind-daemon[2754]: New session c5 of user user.
[  213.375797] anx7688 1-0028: dp state changed to 0x04
[  214.399760] anx7688 1-0028: dp state changed to 0x05
[  217.893116] rfkill: input handler disabled

      By Noah Zalev on 2020-09-02T18:25:41

    • Administrator
      Administrator @root · Imported
      Author Owner

      I have the PostmarketOS CE Pinephone 3gb. I have tried 4 different HDMI cables(Amazon Basic, Monster, RocketFish, Generic) with the provided adapter(powered) and also a 3rd party adapter(Anker). It was tried on 2 tvs(Samsung, Toshiba), 2 moniters(Asus, Assure) and 1 projector(Samsung). No display out though HDMI still. Thank you for the effort you have been putting into this project.

      By Zachary Johnson on 2020-09-05T12:59:53

      Edited by Administrator
    • Administrator
      Administrator @root · Imported
      Author Owner

      Although it's also broken for me, i'm experimenting with xrandr, and the hdmi monitor is listed as XWAYLAND1 and seems correct. Any tips on if this is a fruitful path? If you can fill us in on the best way to work with the hdmi out in future that would be helpful. Thanks.

      By Argon Wight on 2020-09-07T07:21:40

    • Administrator
      Administrator @root · Imported
      Author Owner

      @MakeSaltLake Along those same lines, I installed wlr-randr as the Wayland equivalent and it seems to detect that the HDMI output is plugged in and even thinks it's enabled (in my case, it's a Vizio TV). But there's no actual output.

      By ಚಿರಾಗ್ ನಟರಾಜ್ on 2020-09-07T22:42:24

      • Administrator
        Administrator @root · Imported
        Author Owner

        Reading Megi's blog, he experimented with some HW-accelerated video on the PinePhone and video-out, and I noticed he had the same convergence kit we all have.

        At the end of his post he has the following observation:

        Another issue is a heavy power consumption (4W) from the dock when HDMI output is enabled. Also dock's display output doesn't seem to work when powering the dock from the USB-C power supply.

        I must admit I've always powered the dock when doing these tests (and couldn't get anything to work). Can anyone who has made this work (or nor work) verify their findings against if they've powered the HUB or not?

        Maybe that's why we're all seeing different, seemingly random observations here?

        By Jostein Kjønigsen on 2020-09-18T19:41:18

        Edited by Administrator
      • Administrator
        Administrator @root · Imported
        Author Owner

        FWIW: I've tried to get the external hdmi monitor working with the dock powered and unpowered, unsuccessful either way thus far.

        When the dock is powered I'm using a 5v 2.1 amp brick to the usb-c port on the dock.

        Thanks!

        By Tom Ja on 2020-09-18T19:41:18

      • Administrator
        Administrator @root · Imported
        Author Owner

        With power, I see

        [  738.843461] anx7688 1-0028: dp state changed to 0x00
[  754.203239] anx7688 1-0028: dp state changed to 0x03
[  754.517595] rfkill: input handler enabled
[  755.227285] anx7688 1-0028: dp state changed to 0x04
[  756.030390] elogind-daemon[2774]: Removed session c4.
[  756.252978] elogind-daemon[2774]: Failed to create inotify watch on /dev/null/utmp, ignoring: Not a directory
[  756.254322] elogind-daemon[2774]: New session c5 of user user.
[  757.275294] anx7688 1-0028: dp state changed to 0x05
[  760.347132] anx7688 1-0028: dp state changed to 0x04
[  761.371347] anx7688 1-0028: dp state changed to 0x05
[  762.159746] rfkill: input handler disabled

        Without power is looks almost the same. Two thing I eye spy -- why is /dev/null being treated like a directory here? Only seems to happen on HDMI plug-in. Maybe something wrong there. I'm guessing that "null" there is from a variable and that nobody hand-typed /dev/null/something Next thing I notice is the dp state changes to 3, 4, or 5 on plug in (kind just rolls across them) and then back to 0x00 when disconnecting. So... the hardware is working, it looks like.

        Absolutely nothing on the display btw no matter what I tried. Most strangely is that when I do plug in the display I can leave out the right side of the pine display and if I come back quickly enough I can even spot the small size mouse. I think the display is connected in software somehow, despite nothing being sent to it.

        Hope these thoughts help!

        By bujiraso on 2020-09-18T23:29:41

      • Please register or sign in to reply
    • Administrator mentioned in issue #794 (closed) · Imported

      mentioned in issue #794 (closed)

      By Emi Vasilek on 2020-09-18T18:06:34

      • Administrator
        Administrator @root · Imported
        Author Owner

        HDMI output works on pmOS Gnome on megi's 13 distribution image [0] which is using megi's kernel. So, maybe packaging it alongside the normal kernel and letting users optionally install it would be ideal? I can see that it will not be the default !1533 (closed)

        [0] https://xnux.eu/p-boot-demo/

        By Emi Vasilek on 2020-09-19T20:46:56

        Edited by Administrator
      • Administrator
        Administrator @root · Imported
        Author Owner

        Confirming megi's kernel works for me on the pBoot 13 distro image from my SD card. PMOS Phosh and Gnome both confirmed to show output -- whoo hoo!

        I'm with whatever plan integrates that patch or kernel ASAP :)

        By bujiraso on 2020-09-19T20:46:56

      • Administrator
        Administrator @root · Imported
        Author Owner

        Yes, megi's 13-pBoot works for me too with the dock and external monitor! (Just have to figure out how to expand the friggin BTRFS partition, getting unsupported optional features (10), but that's another story).

        It's great to see the display working, but any idea how to get apps to display on it? Title-bar dragging doesn't work, or is not available.

        On standard Debian I can disable my laptop's built-in display when working with a monitor via setup. I did not see that functionality with megi's 13 on either Mobian or PMos.

        Chess and the flashlight app work on Mobian, how cool is that?

        Kind of amazed I got this far with PinePhone and the convergence package.

        Not complaining! Thanks again for any help!

        By Tom Ja on 2020-09-20T15:40:17

      • Administrator
        Administrator @root · Imported
        Author Owner

        I'm happy to read that it works with megi's current branch!

        There are still different opinions on what's the right kernel to ship in the core team (who are the people who will essentially maintain this kernel in postmarketOS), so I'll bring this up in the next team meeting and try to make it so that we can ship an updated kernel next week where this works by default. And also have a clear path forward to which kernel versions to update to.

        Thank you all in this thread very much for all the invaluable feedback! :tada: :tada: :tada:

        By Oliver Smith on 2020-09-21T15:16:37

      • Administrator
        Administrator @root · Imported
        Author Owner

        @ollieparanoid About the choice between megi's kernel and the one found on pine64's gitlab,
        I think megi's kernel may also fix manual suspend on 1.0 and Braveheart (#774 (closed)) because it's working on megi's p-boot (13 distro) demo image (pmos+phosh).

        By Federico Amedeo Izzo on 2020-09-22T10:54:12

      • Administrator
        Administrator @root · Imported
        Author Owner

        I've reproduced that HDMI works with megi's multiboot image on three monitors.

        So after lots of discussion, we will likely just ship megi's kernel tree. As you all know, he drives PinePhone kernel related development forward like no other, his tree and changes are well documented on his homepage and he frequently publishes tested "build this one" commits with the latest improvements. So it should make a faster and less time consuming workflow to simply package these versions (after a smoke test) into pmOS edge and after some time (to catch regressions) ship the same version in pmOS stable.

        With that being said, I'm working on an update since yesterday and unfortunately HDMI does not simply work when building the kernel from source as our linux-postmarketos-allwinner package. It seems to be related to our kernel config, so I've tried to build with pinephone_defconfig from megi's branch (+pmOS changes to make pmbootstrap kconfig check happy), but same result. It might get fixed when changing a bunch of kernel modules from =m (build as module) to =y (build into the kernel), as it was done in pinephone_multidist_defconfig (the one used in the multiboot image). Other things are also broken, we'll need to adjust some alsa related things etc. So it will take some more time until it's actually shipped, but I'm on it.

        WIP MR: !1614 (closed)

        By Oliver Smith on 2020-10-01T15:55:07

        Edited by Administrator
      • Administrator
        Administrator @root · Imported
        Author Owner

        This approach sounds good to me!

        Megi's image does have a lot of key functionality. Kudos for looking at merging those changes, perhaps switch back when appropriate and all is working in your branch? Until then Megi's image looks good to go.

        Thanks again to you and your team (and Megi!!) for your hard work and keeping PMos w/PPhone on track.

        By Tom Ja on 2020-10-01T18:14:46

      • Please register or sign in to reply
    • Administrator mentioned in merge request !1614 (closed) · Imported

      mentioned in merge request !1614 (closed)

      By Oliver Smith on 2020-10-01T15:54:16

      • Administrator
        Administrator @root · Imported
        Author Owner

        Update from https://xnux.eu/log/#021:

        [...]

        External display fixes

        With help from ollieparanoid, I noticed that whether external display works in my kernel or not depends on whether the user uses p-boot display mode or not (p-boot without display support, or without splash screen, or u-boot).

        I fixed this issue. It related to what parent the MIPI-DSI clock had on boot.

        There's still one issue where external display only works if the phone is booted with dock connected but not when the dock is first connected after boot. That one still remains to be investigated.

        The symptoms are that TCON1 clock is not enabled when the dock is connected only after boot, for some reason.

        After this is fixed, the use of external display should become much more reliable.

        Hopefully, I'll manage to figure out the fix prior to 5.9 release next week.

        So once the complete fix is out, we should be able to ship it in postmarketOS :)

        By Oliver Smith on 2020-10-05T13:55:47

        Edited by Ghost User
      • Administrator
        Administrator @root · Imported
        Author Owner

        And I suspected that bootloader was the reason..

        By Alexey Min on 2020-10-05T13:55:47

      • Administrator
        Administrator @root · Imported
        Author Owner

        Interesting, not sure why it just works for me then with u-boot. Hopefully that makes it work on the rest of the monitors.

        By Martijn Braam on 2020-10-05T15:29:01

      • Please register or sign in to reply
    • Administrator closed · Imported

      closed

    • Administrator
      Administrator @root · Imported
      Author Owner

      FYI: I tested it out with the 5.9.1 kernel on stable and my displays (1920x1080) are now working fine using the USB dock! :tada:

      By Dylan Van Assche on 2020-10-25T06:20:05

      • Administrator
        Administrator @root · Imported
        Author Owner

        I connected the dock to the running phone and then a BENQ GL 2450. Previously this didn't work at all, now it shows a portion of the background image. I have yet to see whether it's working correctly when I have some more time but it's definitely progress.

        By murks on 2020-10-25T07:39:40

        Edited by Ghost User
      • Administrator
        Administrator @root · Imported
        Author Owner

        Try connecting a keyboard, open an application, and then press Super ("Windows key") + Shift + Right arrow key. If this doesn't work, try a different arrow key.

        By Newbyte on 2020-10-25T07:39:40

      • Please register or sign in to reply
    Please register or sign in to reply