**** BEGIN LOGGING AT Fri Mar 29 02:59:57 2019 Mar 29 07:23:50 Hi ! Mar 29 07:24:31 * Tofe has started a warrior tissot build before leaving for work Mar 29 07:24:49 so, let me read the backlog... Mar 29 07:27:30 JaMa: it's find to have another name, really. I chose by-partname because it matched the uevent property, that's all Mar 29 07:27:40 s/find/fine/ Mar 29 09:24:28 Morning! Mar 29 09:47:56 morning Mar 29 09:48:09 Tofe: ok, I'm merging it now Mar 29 09:48:36 or rather after coffee to not break everything Mar 29 09:49:24 any objection to https://github.com/webOS-ports/android-update-package/pull/5 ? Mar 29 09:49:35 it doesn't make it much uglier :) Mar 29 09:59:49 go ahead :) Mar 29 10:39:45 heh I've pruned some unused old branches locally and now git gc made meta-webos-ports bigger (28MB -> 38MB) and meta-smartphone smaller (33MB -> 21MB) :) Mar 29 10:44:10 Tofe: when you get to rebasing your hammerhead-mainline changes, please compare with jansa/warrior branch and we can discuss them Mar 29 10:44:30 I mean rebasing on top warrior Mar 29 12:01:46 JaMa: sure Mar 29 12:02:01 my tissot warrior build just ended Mar 29 12:02:05 (successfully) Mar 29 12:06:54 cool Mar 29 12:07:12 btw: I've removed those CMDLINEs from kernel recipes you mentioned yesterday Mar 29 12:07:36 ah great, it'll make tinkering with the cmdline much faster Mar 29 12:10:45 heh, didn't have enough coffee in the morning Mar 29 12:11:03 fixing android-update-package SRCREV now.. Mar 29 13:45:26 now when we're on newer Yocto I'm updating local.conf a bit in webos-ports-setup Mar 29 17:59:12 JaMa: we definitively need the by-partlabel change, even on warrior Mar 29 18:05:21 Tofe: it's everywhere now (sumo, thud, warrior, master) Mar 29 18:05:40 new images with it are on jenkins already Mar 29 18:08:35 good Mar 29 18:08:55 strange: upon reboot, it fails to boot, black screen, no adb... Mar 29 18:10:04 reboot triggered from running system Mar 29 18:10:06 ? Mar 29 18:10:36 yes, to fastboot, then boot via cmdline Mar 29 18:11:01 but I have a pstore, I'm looking through the logs Mar 29 18:11:09 and this is after including the by-partlabel changes? Mar 29 18:11:28 "ipa ipa2_uc_state_check:301 uC is not loaded" / "Rebooting in 5 seconds.." Mar 29 18:11:50 the rebuild should be quick, just busybox, kernel and image + db8 repackaged Mar 29 18:12:47 is there a way to only reinstall busybox (withtout rebuild) when modifying mdev-partname.sh ?... Mar 29 18:13:30 I was rebuilding the kernel (with initramfs) Mar 29 18:14:13 takes about the same time as manually unbootimg, modify cpio, abootimg Mar 29 18:15:02 mmh "Unable to handle kernel NULL pointer dereference at virtual address 00000040" Mar 29 18:15:04 and building whole image with changes from today shouldn't take much longer (nothing should invalidate the big stuff) Mar 29 18:20:00 kernel crash seems to be triggered by ueventd (from android lxc container, if I understand correctly) Mar 29 18:20:22 in "firmware_direct_write+0x50/0xc0" Mar 29 18:38:47 I've added 'echo "" > /sys/kernel/uevent_helper' in stop_mdev(), and it seems to work for the immediate next boot -- I'll reboot to see if it's stable Mar 29 18:38:59 anyway it's a good idea to have it Mar 29 18:41:35 3 consecutive successful boots... let me PR that one Mar 29 18:49:17 good, makes sense Mar 29 18:53:12 btw the issue in qt 5.12.2 got fixed today, I'm retesting it on top of warrior Mar 29 18:54:15 https://github.com/webOS-ports/meta-webos-ports/pull/322 Mar 29 19:34:27 Tofe: any objection merging those anbox WIP changes (just the recipes without including it in the image)? Mar 29 19:35:11 Tofe: gtest upgrade seems to be useful anyway and I have some pending fixes which were needed with newer Yocto (like boost compatibility for dbus-cpp) Mar 29 19:35:34 it might be useful to merge it to warrior once we switch testing to it Mar 29 19:36:36 will send PR for Qt and Anbox changes tomorrow if everything builds ok over night Mar 29 19:43:07 JaMa: no, we can as well include these changes, that way the maintainance will be easier in the long run Mar 29 19:43:22 but let's keep it outside of images for now yes Mar 29 19:43:45 maybe we could introduce a new target image with anbox inside, I don't know Mar 29 20:20:15 with RDEPENDS/RRECOMMENDS defined inside the anbox recipe, it's really just anbox in IMAGE_INSTALL, so anybody can do it locally in local.conf Mar 29 20:20:44 creating new image would require new -package and -emulator-appliance recipes as well and that's more work for just 1 package Mar 29 20:49:49 right **** ENDING LOGGING AT Sat Mar 30 02:59:57 2019