**** BEGIN LOGGING AT Wed Feb 03 02:59:57 2021 Feb 03 08:20:48 Morning! Feb 03 08:24:39 Morning! Feb 03 08:49:10 morning Feb 03 10:39:35 Tofe: Am I right in assuming we're not using initramfs-scripts-android anymore and just initramfs-scripts-halium, or am I wrong there? Feb 03 10:40:14 Herrie: the usecase for -android is a mainlined android device Feb 03 10:40:32 hammerhead could become one of these Feb 03 10:41:15 but today, yes, only -halium is used Feb 03 10:41:45 Tofe: OK Feb 03 10:41:51 Clear Feb 03 10:42:21 There's also a -simple one for pinephone, maybe it could be merged with -android, not sure yet Feb 03 10:43:00 Seems that we would need to add something similar to https://github.com/Halium/initramfs-tools-halium/commit/2e95f3a55d99c3126fa21731d28acc59f44ffc83 into https://github.com/Tofee/meta-smartphone-1/blob/testing/halium-9.0/meta-android/recipes-core/initrdscripts/initramfs-scripts-halium/init.sh#L25 Feb 03 10:43:05 So we can debug with newer kernels Feb 03 10:43:21 Seems 4.4+ kernels use configfs and no longer functionfs Feb 03 10:43:29 So ADB fails on those on panic Feb 03 10:44:02 This was the log on Sagit: https://paste.ubuntu.com/p/zBRFN7r59C/ Feb 03 10:44:13 [ 4.37778U ili4pd: PANIC for reason: Could.'d dald data partition. Feb 03 10:44:28 Which is weird, because in TWRP there's a /data/luneos/ etc Feb 03 10:44:45 But the partitions for this device are a bit weird (sda/sde) Feb 03 11:00:42 I thought we had something like that in the -android version Feb 03 11:01:12 Tofe: In Android version yes, but not in the Halium one Feb 03 11:01:56 ok; then yes we'll need to copy it there, and find a way to use one or another Feb 03 11:02:16 Well seems it takes care of that too Feb 03 11:02:36 if [ -d $ANDROID_USB ]; then etc Feb 03 11:07:02 true Feb 03 11:07:55 for sagit, maybe there's no partition name ? or not the usual names ? that can be seen in twrp, if needed Feb 03 11:33:30 Tofe: They have these: https://github.com/SailfishOS-sagit/hybris-boot/commit/4ddd60e3f0bbe3990f12f60fdd6e4dc50d02ce4a Feb 03 12:06:59 I should get a full list of partitions later today Feb 03 12:07:03 From twrp Feb 03 12:07:11 I don't see where we prefix anything Feb 03 12:07:45 But it could be our ancient update-binary (https://github.com/webOS-ports/android-update-package/tree/master/META-INF/com/google/android) is the problem somehow ;) Feb 03 12:39:27 Herrie: so there is a "userdata" partition... it should be detected by halium's initrd, normally Feb 03 12:39:56 Tofe: Yeah.... Feb 03 12:40:04 It seems that in TWRP he has /data/luneos etc Feb 03 12:40:09 And it has data Feb 03 12:40:13 So that doesn't seem to be the issue Feb 03 20:32:44 Tofe: You have any clue why /dev might not be mounted? Feb 03 20:32:54 Or at least why the halium initscript cannot find userdata? Feb 03 20:37:00 I.e. somehow this part doesn't return anything: https://github.com/Tofee/initramfs-tools-halium/blob/tofe/halium-9.0/scripts/halium#L417 Feb 03 21:17:32 Herrie: not really, unfortunately Feb 03 21:21:01 Tofe; OK, will try to get some more details Feb 03 21:21:07 Regarding parition layout Feb 03 21:21:44 Herrie: maybe we could try overriding data partition with cmdline ? Feb 03 21:22:29 like datapart=/dev/sda17 Feb 03 21:24:09 when in initrd, we are running busybox's mdev to populate /dev Feb 03 21:24:29 so maybe mdev gets confused - but no idea why Feb 03 21:29:26 Tofe: Yeah I saw the datapart cmdline will try that Feb 03 21:31:22 Herrie: what is that "sagit" device from, by the way ? another person interested in a port ? or a new old device you're trying ? **** ENDING LOGGING AT Thu Feb 04 03:02:57 2021