| T42 | <nc1x72> Hello all, | 11:17 |
|---|---|---|
| T42 | <nc1x72> So I compiled sfos for Nothing Phone 1. But it reboots after OEM splash screen (doesn’t show SailfishOS logo before rebooting). | 11:17 |
| T42 | <nc1x72> Anyone got an idea on what could possibly be wrong? | 11:17 |
| T42 | <elros34> a lot of, start with https://hadk.sailfishos.org/getting-in/ and figure out whether device shows up in host dmesg | 11:34 |
| mal | @nc1x72 how soon does it reboot? | 12:38 |
| mal | which android base? | 12:38 |
| T42 | <nc1x72> In less than a minute. LineageOS 21 base (which is what your fp5 build is based on I believe) | 12:40 |
| mal | if it's about a minute then it suggest usb is not being setup correctly | 12:47 |
| mal | does your device have which boot partitions? boot, vendor_boot? does it have init_boot also? | 12:48 |
| mal | which partitions did you flash? | 12:48 |
| T42 | <nc1x72> Same config and same platform as fp5 (qcom lahaina). boot and vendor_boot. Rootfs flashed to userdata | 12:54 |
| mal | and you flashed also vendor_boot now? | 13:21 |
| T42 | <nc1x72> Yes | 14:11 |
| T42 | <nc1x72> Ok I set selinux=0 in CMDLINE and it doesn’t reboot anymore. But it shows just a black screen about 20seconds after SailfishOS logo. Still not detected when plugged into the computer. I built lvm image that I flashed to userdata partition | 16:46 |
| T42 | <Umeaman> Kernel issue? | 16:55 |
| mal | I think selinux should not be disabled completely | 17:04 |
| mal | @nc1x72 are you sure you have proper selinux configs? | 17:04 |
| mal | how did you handle system/vendor partitions things? | 17:05 |
| T42 | <nc1x72> Hete is kernel defconfig https://github.com/SailfishOSSpacewar/android_kernel_nothing_sm7325/blob/hybris-23.2/arch/arm64/configs/vendor/lahaina-qgki_defconfig | 17:27 |
| T42 | <nc1x72> system, system_ext, odm and vendor contents are inside the lvm rootfs. | 17:29 |
| mal | have you checked that the symlinks under here https://github.com/SailfishOSSpacewar/droid-config-Spacewar/tree/master/sparse/etc/selinux are correct? | 17:59 |
| mal | there are 2 of those in subfolders | 17:59 |
| mal | not sure if it's still an issue but at least before it caused some issues to have device code containing capital letters | 17:59 |
| mal | *device codename | 18:00 |
| mal | you can see in fp5 repo that I handled that like this https://github.com/mlehtima/droid-hal-fp5/blob/master/droid-hal-fp5.spec | 18:01 |
| mal | device vs rpm_device | 18:01 |
| mal | can't remember at which point it might have been issue, probably not at early boot | 18:02 |
| T42 | <nc1x72> The symlinks targets exist. I have checked. | 18:19 |
| T42 | <nc1x72> | 18:19 |
| T42 | <nc1x72> Regarding device codename, the build scripts didn’t complain about it. So, I’m guessing it’s not a blocking point. | 18:19 |
| mal | did you do lvm based build or community style build? | 18:33 |
| mal | lvm based in a bit more difficult to debug | 18:33 |
| mal | and you checked that the generated mount services look correct? | 18:34 |
| mal | i.e. fixup-mountpoints worked correctly | 18:34 |
| T42 | <nc1x72> I went with lvm build indeed. Not sure about the other mount services but I fixed fixup-mountpoints locally (forgot to clone that repo to make changes) | 18:53 |
| mal | you can check those mount services in droid-config rpms | 19:38 |
| mal | wait, droid-hal rpms | 19:38 |
| mal | sorry | 19:38 |
| T42 | <nc1x72> In /usr/lib/systemd/system, I have these (relating to mounts: | 23:49 |
| T42 | <nc1x72> - metadata.mount | 23:49 |
| T42 | <nc1x72> - mnt-vendor-persist.moint | 23:49 |
| T42 | <nc1x72> - mnt-vendor-qcms.mount | 23:49 |
| T42 | <nc1x72> - vendor-bt_firmware.mount | 23:49 |
| T42 | <nc1x72> - vendor-dsp.mount | 23:50 |
| T42 | <nc1x72> - vendor-firmware_mnt.mount | 23:50 |
| T42 | <nc1x72> | 23:50 |
| T42 | <nc1x72> They all point to the correct partitions. Made the changes locally before building, just like with fixup-mountpoins | 23:50 |
| T42 | <nc1x72> | 23:50 |
| T42 | <nc1x72> I did discover that usb-moded might be broken, that’s why I don’t get a detection. Will try to feed it the IDs from stock usb rc files and see, because phone does charge when plugged into an outlet. | 23:50 |
| mal | you mean that gadget_udc_device ? | 23:50 |
| mal | in usb-moded config | 23:52 |
| T42 | <nc1x72> I’m talking about the IDs defined in /etc/usb-moded | 23:54 |
| T42 | <nc1x72> That’s where I fixed usb-moded in UT | 23:54 |
| T42 | <nc1x72> Is it different on SFOS? | 23:54 |
| mal | not sure what ids you mean | 23:54 |
| mal | the config is quite simple | 23:55 |
| mal | you mean the function names? | 23:55 |
| mal | does your device use usb0 or rndis0 for the interface? | 23:56 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!