**** BEGIN LOGGING AT Sat May 30 02:59:59 2020 May 30 04:35:10 Morning May 30 04:36:52 JaMa: I've forced push some minor fixes, nothing shocking though. Forgot some inherit webos_app for testr and added a few more missing permissions in core-apps and app-services that I saw during runtime testing. Nothing major. Doing another round of testing shortly to see if there's other things that need fixing, but should be in pretty good shape already. May 30 05:47:10 Tofe: Seems I might have found the issue with C+DAV connector. It needed to have access to https://github.com/webOS-ports/app-services/blob/herrie/acg/com.palm.service.accounts/files/sysbus/com.palm.service.accounts.api.json#L5 however in perms I only specified "applications", so now added "applications.internal" which might help to solve things May 30 06:08:47 No that's not it, but it will help at least May 30 08:09:09 Morning! May 30 08:16:52 Tofe: for the ls-monitor flooding we could simply limit the number of calls it does here: https://github.com/webOS-ports/luna-next-cardshell/blob/master/qml/Connectors/TelephonyService.qml#L59 May 30 08:18:26 Or query the device name and if it contains qemux86 simply not start the timer May 30 08:20:03 Herrie: but here, the backend simply isn't there, isn't it ? there must be a way to isolate this use-case May 30 08:20:55 I'll have a coffee first May 30 08:25:00 Tofe: The timer works fine on phones when there might be some delay in getting telephony service May 30 08:25:55 But for qemux86(-64) that gives some issues. Not having oFono and telephonyd gave other issues though I don't recall what exactly May 30 08:27:21 And issue tracker is still down so cannot find it easily May 30 08:29:10 getting the device name only for qemu (i.e. in 95% of use-case it's not useful) seems like a lot of "test"-code inside normal code May 30 08:29:39 I'd prefer a buildtime thing, but I'm not fully awake yet May 30 08:30:28 Tofe: well we had this previously and I reworked that. Commit message contains part of the reasoning but I remember there was more: https://github.com/webOS-ports/meta-webos-ports/commit/5b7f52824d68db48d54977bece3e5ed6d97a67a8#diff-c97edbed1c6225a88e460a687143aa3c May 30 08:35:01 I found this: https://github.com/webOS-ports/meta-webos-ports/pull/197 May 30 08:35:42 and the commit can be seen here in the history of commits: https://github.com/webOS-ports/meta-webos-ports/commits/zeus?after=a12df5bfc5a3b11bfbdcbd0ed0c4a44ea52c9138+1159 May 30 08:56:07 Tofe: Yes, but there was more details in the bug tracker I remember even May 30 08:56:17 Herrie: we should actually subscribe for any change in the backend, instead of probing ourself May 30 08:57:15 Tofe: If it allows subscribing that would be nicer v.s. checking every 0,5s ;0 May 30 08:57:17 ;) May 30 08:57:50 what implements it ? webos-telephonyd ? May 30 08:59:36 Hi. all the sstate synced downloas are failing in between, generally after downloading 20-25MB. Can i do something to tackle this? May 30 08:59:55 Suddenly connection is closed from server side as per logs May 30 09:00:40 Seems so: https://github.com/webOS-ports/webos-telephonyd/blob/herrie/acg/files/sysbus/webos-telephonyd.api.json.in#L6 May 30 09:00:52 I have downloaded those manually via firefox, by resuming again and again. Are those files usable I mean, can i copypaste those files in sstate folder and will bb recognise them as okay? May 30 09:01:46 or ay other solution by confguring some files.... it will save my day May 30 09:04:04 BT40: Weird issues you have there.... May 30 09:04:12 I'm happy to build something for you if needed ;) May 30 09:04:46 okay sure.... even this seems logical to me no. struggling so much..... ) May 30 09:04:52 n=no May 30 09:04:58 no=now May 30 09:05:14 i will provide all those config files May 30 09:06:29 BT40: Infra guys are working hard to keep things running & working, but recently they had quite some DDOS from China May 30 09:06:31 shall i upload those files needed on dropbox and share? Or shall i push somewhere like github..... which method is conenient for you Herrie May 30 09:06:40 Github is better May 30 09:06:47 okay May 30 09:07:16 I am not familiar with github. Will go through it and share you the folder. May 30 09:07:23 sorry, share you the details May 30 09:07:24 Also they are moving quite some things around & improving infrastructue... All is done for free and in spare time, so there's no SLA's etc as you can understand May 30 09:07:32 We're alreayd happy to have infra available for free :0 May 30 09:07:33 yeah May 30 09:07:45 BT40: Well dropbox can also work May 30 09:09:10 Herrie, Since my compilation is 99% complete, 6450 out of 6507 are completed, already a lot work is done. one last try i want to give. May 30 09:09:27 So i wanna know, will manually downloading work? May 30 09:09:40 i have alreadydownloaded sstate files via firefox May 30 09:09:53 so if i can use them somehow, work will be completed May 30 09:10:18 BT40: there must be som md5sum or something that comes along May 30 09:10:38 If not, the saviour you guys are here. I will share with you and you can compile it for me May 30 09:10:45 okay May 30 09:10:58 so i will push changes to github May 30 09:11:32 will take a couple of hours to learn github, i will come back aand discuss in detail how to push efficiently May 30 09:12:00 BT40: pushing your current work is a good idea, we never know what can happen :) May 30 09:12:17 ok done . Will push today May 30 09:12:47 Herrie: looking at the source of webos-telephonyd, as soon as com.palm.telephony exists we should just be able to start subscribing and waiting, no need to probe May 30 09:12:58 Though I guess the probing is there for a reason... May 30 09:14:08 https://github.com/webOS-ports/luna-next-cardshell/commit/953bfb08c377027d4a910731e07cff209c5222d0 it all came in one go May 30 09:15:03 BT40: you were interested in the compile log for a complete qtwebengine build? May 30 09:15:24 yeah May 30 09:15:42 Tofe, yeah, that will be good to see May 30 09:15:53 one moment then, I rebuilt qtwebengine 5.15 for Xiaomi Tissot yesterday evening May 30 09:15:57 (twice...) May 30 09:18:18 BT40: https://www.dropbox.com/s/5yfuuq5a583i3v2/log.do_compile_qtwebengine_5_15.txt?dl=0 May 30 09:18:25 I thought there would be timestamps :/ May 30 09:19:51 but configure task ended at 18:34, and compile task at 21:03, so just do the math :) May 30 09:20:06 Tofe: maybe the systemd order was wrong and that's why morphis decided to do the probing May 30 09:20:19 3 hrs compiletime....wao May 30 09:20:29 We fixed quite some of the systemd startup order things so probing might not be needed May 30 09:20:33 Herrie: but the probing only begins when the servicename is available on the ls2 bus May 30 09:21:05 I think he wasn't sure if "subscribe" would also work when the service isn't available yet May 30 09:21:14 let's try ! May 30 09:26:02 Herrie: he might have been fooled by https://github.com/webOS-ports/webos-telephonyd/blob/webOS-ports/webOS-OSE/src/telephonyservice_misc.c#L65 May 30 09:26:31 but it's actually fine, we don't want to be called unless the service is initialized May 30 09:29:55 Tofe: yes might be May 30 09:32:37 ah, no, he was right (as often), the subscription fails is the service isn't initialized May 30 09:32:47 but I think it shouldn't fail May 30 09:40:46 Herrie: I'll patch webos-telephonyd a little May 30 10:00:29 Tofe: Ok May 30 10:04:07 yup, that works May 30 10:08:07 here we go. May 30 10:11:34 Lucky enough...QtWeb Engine completed via sstate :) May 30 10:11:51 6497/6507 completed May 30 10:12:12 ERROR: luneos-dev-image-1.0-r0 do_rootfs: Unable to install packages. May 30 10:12:21 Collected errors: May 30 10:12:21 * opkg_prepare_url_for_install: Couldn't find anything to satisfy 'kernel-module-wlan'. May 30 10:13:33 Details: https://pastebin.com/zsXY6yjx May 30 10:13:56 BT40: ah, that's easier May 30 10:14:51 in your land.conf file, there is a MACHINE_EXTRA_RDEPENDS May 30 10:15:21 here, you can specify which kernel modules to add on the final image May 30 10:15:36 but these modules, of course, must match the defconfig chosen for your kernel May 30 10:16:11 in this case, I think you included the wlan feature directly in the kernel, not as a module May 30 10:16:11 although it is giving error, strange thing is it find sstate. How is this possible? May 30 10:16:44 it's an error for luneos-dev-image, qtwebengine is fine now May 30 10:17:18 Tofe: nice workaround like this. This is more proper vs previous solution May 30 10:18:19 land.conf...in meta-smartphone folder? May 30 10:18:45 BT40: yes; I suggest you don't touch your kernel recipe, and remove kernel-module-wlan from MACHINE_EXTRA_RDEPENDS May 30 10:19:11 maybe wifi won't work, but that's a minor issue if it boots at all :) May 30 10:19:24 ok May 30 10:19:34 can we attempt wifi later on? May 30 10:20:12 sure we will May 30 10:20:21 ok May 30 10:20:58 yes; I suggest you don't touch your kernel recipe..... Shall i remove kernel-module-wlan line from machine.conf? May 30 10:22:08 * BT40 is confused at prior sentence having words machine extra repends May 30 10:22:16 from land.conf yes; I just meant that if you modify linux-xiaomi-land.bb it'll trigger a whole lot of rebuilds May 30 10:22:27 okay May 30 10:22:32 will remove this line May 30 10:23:13 do i need to setup environment again or any other sub command to implement changes? May 30 10:23:20 or bb will automatically do? May 30 10:23:42 no, remove the line, save, and start bb luneos-dev-package once more May 30 10:23:55 started May 30 10:24:15 bb will automatically detect what should be rebuilt; in this case, only luneos-dev-image May 30 10:25:37 luneos-dev-image:do_rootfs is one of the last tasks; you'll soon have a .zip to sideload on the device May 30 10:26:13 (but it manipulates big 700MB files, so it might take some minutes) May 30 10:29:57 i do have a concern. During my current first build, there has been 4 times power cuts and sudden computer off: (I mean i didn't press ctrl c to close, neither system shut down. Direct power resulted in sudden PC out of power ). I fear this might have corrupted some downloads or compile files. May 30 10:30:05 Will bb take care of this? May 30 10:30:37 i hope final image will not be corrupted May 30 10:31:16 Because although I can build twice, but pushing corrupted things on Github will be a foolishness by me. So thats why i am asking May 30 10:31:18 There are several md5sums along the way, for ssttate, for packages, etc May 30 10:31:38 for GitHub you'll just push text files May 30 10:32:08 and for the Halium image you built, there is both md5 ans sha256 checks May 30 10:32:47 30 warnings generated like this: WARNING: luneos-dev-image-1.0-r0 do_rootfs: QA Issue: Unable to check the binary /usr/palm/applications/org.webosports.app.settings-common/org.webosports.app.settings.wifi mentioned in LS2 role files May 30 10:33:29 WARNING: luneos-dev-image-1.0-r0 do_rootfs: QA Issue: Unable to check the binary /usr/palm/applications/org.webosports.app.settings-common/org.webosports.app.settings.backup mentioned in LS2 role files May 30 10:33:29 So yes, there might be concerns about the integrity of your image, but for the source recipes files I don't think that'll be a problem. Though it's one more reason to push this to a safe place :p May 30 10:33:57 Herrie: jenkins is finishing the build, we can merge the PR now and restart them May 30 10:34:11 * JaMa reads backlog May 30 10:34:12 BT40: We'll handle these QA warning later on, they're not critical May 30 10:34:26 okay May 30 10:34:49 and they're the consequence of our recent work with Herrie on the ACG migration (security model used by LG's webOS) May 30 10:36:09 JaMa, Herrie: I'll push the big green button May 30 10:36:47 completed May 30 10:36:49 Tasks Summary: Attempted 6507 tasks of which 6496 didn't need to be rerun and all succeeded. May 30 10:37:15 JaMa, Herrie: ACG migration merged ! May 30 10:37:24 No error, only warnings May 30 10:37:31 Hope image is built May 30 10:37:35 BT40: congrats, you have now an image May 30 10:37:47 Tofe: BT40: kernel modules should always go to MACHINE_EXTRA_RRECOMMENDS not RDEPENDS May 30 10:38:26 JaMa: but will it include them anyway ? it looks like RRECOMMENDS isn't a strong word May 30 10:38:35 Congrats....This is completely because of you 4 guys, all credit to you. THANK YOU FROM MY HEART May 30 10:39:04 Well thanks but this isn't the end of the road yet :) now let's see if that boots May 30 10:39:28 Tofe: it's strong enough May 30 10:39:28 BT40: look into tmp-glibc/deploy/images/land/ May 30 10:39:42 JaMa: ok, we made the mistake for several machines I think May 30 10:39:44 * BT40 wishes even if his image is corrupted due to some issue, phone doesn't get hard bricked May 30 10:40:03 Tofe: RSUGGESTS is weak, just to recommend something when you install something by hand with opkg May 30 10:40:23 but RRECOMMENDS are installed by default and to remove some you need to add them to BAD_RECOMMENDATION May 30 10:40:58 JaMa: suggest recommends, recommends installs... Yocto could have made it a bit better wording :p May 30 10:41:02 BT40: please be aware that phone can get hard bricked even if the image is otherwise completely fine May 30 10:41:33 JaMa, yupp. Hope everything goes fine.... May 30 10:41:47 Tofe: this goes down to opkg, deb format using Suggests/Depends/Recommends... OE just added R :) May 30 10:41:57 BT40: so. you have a working recovery already, right ? May 30 10:42:14 TWRP yes May 30 10:42:33 ok, great; boot the phone into that May 30 10:42:45 should be power + VolumeUp keys May 30 10:43:06 will need some time. Need to backup data. This is my primary phone.... May 30 10:43:28 Oh, it's your primary phone... yes, be careful with your data May 30 10:45:16 It's not ideal to do that kind of hacking on a primary phone, the whole thing is a bit risky, whatever precaution we take May 30 10:45:38 But I guess you alraedy know that May 30 10:46:11 Tofe: Yahoo :) May 30 10:46:20 That's a big one out of the way May 30 10:46:49 If you tested the webos-telephonyd and lnc I'm OK to merge them May 30 10:46:52 BT40: I suggest you first download a factory xiaomi-land rom image, and flash it, to be sure you have a way back first May 30 10:47:10 okay May 30 10:47:13 Herrie: yes, it's tested on tissot May 30 10:47:15 good idea May 30 10:47:20 but not on qemu May 30 10:48:10 should I wait for some changes or trigger jenkins right now? May 30 10:48:27 BT40: but from a "partition" point-of-view, the only partition we rewrite is the boot partition. We don't erase userdata, or system, or any other. May 30 10:48:47 JaMa: maybe we can bump webos-telephony and luna-next-cardshell just before May 30 10:48:53 ok May 30 10:49:09 it's a minor fix, but the logs should be less spammed May 30 10:50:29 Herrie: JaMa: one remark though: I did my webos-telephonyd PR towards the main branch, not the acg one... May 30 10:51:00 (though I did test with acg modifications enabled) May 30 10:51:35 Herrie: I suggest we merge both my PR and herrie/acg into the main branch May 30 10:51:51 and bump SRCREV and switch back to the usual branch in the recipe May 30 10:53:34 Tofe: yes May 30 10:53:59 Does device charging work while flashing?? I mean does phone charges side by side during pricess? May 30 10:54:16 although i would charge 100%, still knowing is helpful May 30 10:55:14 BT40: that's a good question, actually I don't know. But I think so. May 30 10:55:38 , the only partition we rewrite is the boot ......... Doesn't webOS has its native apps? I mean isn't webos independent of android? May 30 10:55:52 i understand halium is needed for compatibility May 30 10:55:59 but as an Os May 30 10:56:10 i thought it is completely independent May 30 10:56:30 native apps=native tools and files i meen May 30 10:57:37 * BT40 's understanding is half baked :) May 30 10:58:06 It is independant, yes; we don't reuse anything of Android binaries, except what we build in Halium May 30 10:58:30 usr, system, etc folders then should be different May 30 10:58:47 isn't boot directory used only for booting? May 30 10:59:11 in desktop distros, various other folders play huge role May 30 10:59:38 . / moont i wanna say May 30 10:59:45 mount* May 30 10:59:52 what we do today is that usr/ etc/ var/ and so on go in to the userdata partition (which is already ext4) under the luneos/ folder. And during boot we do a chroot into that directory, so that it's used as the root for the rest of the system May 30 11:00:23 ok May 30 11:00:55 We would be able to just reformat userdata brutally and put ourselves in the root, but it's more convenient for debug to still have access to an android system if all goes wrong May 30 11:01:20 On the pinephone, we don't have any other system, so we just create our own partition table, for instance May 30 11:02:14 Actually if we wanted to install luneos definitely on an android phone, we would choose the system partition, of course :) May 30 11:03:03 Also the partitionning is not up to us, because some partitions are directly used by the hardware, like the GSM modem, GPS, Wifi, etc May 30 11:15:54 shall i flash lineage 14 instead of stock one? It would give some idea which sensors are supported. May 30 11:16:19 Also i'd be able to properly recognise as it will be different from stock one May 30 11:17:48 BT40: I recommended stock one because it's possible to flash it even without a working recovery May 30 11:18:02 okay May 30 11:18:32 but that's just a failsafe, you can download both :) May 30 11:32:04 https://github.com/webOS-ports/meta-webos-ports/pull/435 ready here May 30 11:33:56 ^Merged. JaMa: if you want to launch new builds, go ahead May 30 11:37:26 how much time does flashing rom takes ... rough idea only. 15-20 mins, 30-45 minutes.... or hour? May 30 11:37:47 i need to charge battery accordingly :) May 30 11:38:10 if moe than hour, will connect charger side by side May 30 11:38:47 device: snapdragon 430, 2Gb ram, 16gb rom May 30 11:39:38 Tofe: thanks, build in queue May 30 11:42:05 Tofe: is webOS-ports/webOS-OSE branch completely different history in webos-telephonyd or can we merge that as well to webOS-ports/master? May 30 11:42:45 JaMa: That one should be mergable May 30 11:43:17 The ones that are different are the forks from OWO that also exist in OSE mainly I would say May 30 11:45:53 BT40: 2-3mins May 30 11:47:10 Anything that we developed ourselves should have a mergeable history I'd say May 30 12:17:13 JaMa: So eventually we'll need to keep quite some of the webOS-ports/webOS-OSE branches. Might be good to add something for that in a bbclass, so we don't specify the branch all the time? May 30 12:19:33 Herrie: if the history is completely different (because based on different repo) then it should be in different repository as well May 30 12:20:14 then we can have bbclass for webOS-ports forks from OWO and different one from OSE while both can use the same webOS-ports/master branches by default May 30 12:29:26 JaMa: Problem is that many repos have exactly the same name in OWO and OSE May 30 12:29:55 I'm open for suggestions on how to deal with this in a more proper way May 30 12:32:24 JaMa: They share the same history, just the history in OSE is missing ;) May 30 12:32:47 So if all the history would be there they would be mergable in most cases :P May 30 12:32:55 But LG decided to do a full dump without history May 30 12:37:52 from git POV it's different history, because there are no common object, so the repo just gets significantly bigger May 30 12:39:33 we can append something to BPN e.g. -OSE in the bbclass to make it clear that the git history is based on OSE fork not OWO as it is in many other cases May 30 12:40:19 we could also rename the old branches into different repo, but that would break reproducibility of all old builds (I don't care much about that, but still something to consider) May 30 12:41:46 or we can continue as it's now :) many of these repos probably aren't big enough to matter much when they were doubled May 30 12:46:13 JaMa: Nothing matters for those May 30 12:47:03 They should all be fairly small May 30 12:47:10 It are not kernel repos or something May 30 12:48:13 I tend to cleanup and delete branches that are merged and not longer needed May 30 12:48:37 just tried to reflash original rom using twrp, suddenly fone becomes power off without any warning. Suspicious. May 30 12:48:50 now again i have entered recovery May 30 12:51:57 second attempt: mount o0f /firmware failed May 30 12:52:37 failed to mount /dev/block/bootdevice/byname/splash May 30 12:54:40 second attempt flash complete, but when i press repoot, it says no OS installed. strange May 30 12:55:24 Third attempt zip signature verification failed, error installing rom May 30 13:22:34 so scary. I thought i've bricked phone.....somehow manage to start. Initially it didn't get normal start from power button. 10 second press made it to start May 30 13:23:10 now everything seems normal May 30 13:23:49 good thing was, i downloaded stock rom. Lineage os didn't get installed, gave errors. May 30 13:24:16 Now ready to flash WebOs/LuneOS May 30 13:24:51 is the procedure same? Just copying the file and selecting install from TWRP?? May 30 13:25:10 (after factory data reset wipe) May 30 14:19:02 Bt40 yes May 30 14:19:10 Sorry out and about a bit May 30 14:38:35 shall i flash luneos-dev-package-land-0-0.zip file with twrp install option? May 30 14:38:49 eager for first webos install May 30 15:04:18 * BT40 wonders is installing via adb sideload necessary.... wouldn't twrp be enough? May 30 15:16:35 BT40: o/ May 30 15:21:02 BT40: don't bother with reset swipe; however, if there is encryption of userdata, you'll have to reformat userdata May 30 15:21:32 BT40: looking at your various failed flash attempts, is your usb cable ok ? May 30 15:21:47 yes May 30 15:21:59 i was flashing from internal memory May 30 15:22:04 direct via twrp May 30 15:22:09 not via adb May 30 15:22:35 i first copied image to internal storage, then installed via twrp install button May 30 15:23:27 Miui stock rom causes bootlop problems, found internet archieve. so switched to lineage os some other image. May 30 15:23:42 right now ready to flash May 30 15:23:58 all data backed up. just awaiting for final steps May 30 15:25:00 generally i wiped to factory data reset in twrp to be on safe side. after that, i flashed new rom May 30 15:25:38 also wiped /system, /cache, /dalvik and /data (maybe, nots sure of fourth) May 30 15:38:54 Tofe: For merging all the ACG changes into the "normal" branches for each repo. Should I create a new branch, let's say herrie/acg-clean where I take herrie/acg and squash stuff and then PR from there? May 30 15:39:09 This way I won't break current builds with force pushes while squashing in herrie/acg May 30 15:40:38 Since some repos have a ton of ACG commits and fixups, makes sense to me to squash them? May 30 15:42:05 BT40: adb sideload or TWRP doesn't matter. I tried both, both work in general May 30 15:42:20 ok May 30 15:42:26 going to flash right now May 30 15:42:51 will first implement factory data reset in twrp, then opt for install button. May 30 15:43:09 Hope the file is correct one: May 30 15:43:24 luneos-dev-package-land-0-0.zip fil May 30 15:44:10 Yes that should be fine May 30 15:44:31 Herrie: herrie/acg-clean sounds good for preparing the PR for webOS-ports/master branch May 30 15:45:20 JaMa: OK that was my idea as well, this way can keep current stuff building and clean up in the meanwhile too, so we have hopefully single commits per repo for ACG instead of 10+ in some cases :P May 30 15:46:06 Herrie: well I'm not against multiple commits, as long as their topic is clearly stated in the title; but multiple "fixup" commit should be squashed yes May 30 15:46:32 flashing kernel May 30 15:46:59 complete....rebooting May 30 15:47:37 oh no. stuck to boot loop. 3 times booted logo May 30 15:48:57 back to recovery... Didn't boot May 30 15:49:16 was stuck at mi splash screen and looped 8 times May 30 15:59:01 network issue...irc got disconnected. Hope i didn't missed some confrontation May 30 15:59:41 BT40: In that case you might want to check the following file: May 30 16:00:10 Check: /sys/fs/pstore/console-ramoops May 30 16:00:22 That should have the log from previous boot and hopefully some clues May 30 16:00:47 how to access that file? via adb? May 30 16:01:47 via adb, when you are in twrp May 30 16:07:01 adb shell May 30 16:07:02 land:/ # ls /sys/fs/pstore/console-ramoops May 30 16:07:02 ls: /sys/fs/pstore/console-ramoops: No such file or directory May 30 16:07:21 is ls command appropriate? May 30 16:07:26 or adb pull? May 30 16:11:22 ls is fine, you just don't have any log available May 30 16:11:37 that's going to be a bit tricky I guess May 30 16:11:46 adb shell ls /sys/fs/pstore May 30 16:11:46 Returns nothing. May 30 16:12:25 ah, twrp doesn't even provide pstore feature... May 30 16:12:49 can you adb pull /proc/config.gz ? May 30 16:13:13 mmh no anyway it won't help much May 30 16:13:16 pulled May 30 16:13:49 well that's twrp's defconfig, I just wanted to make sure they don't provide pstore May 30 16:14:27 what's the output of "zgrep PSTORE config.gz" ? May 30 16:15:32 https://www.dropbox.com/s/3a6hqehlojqrri0/config?dl=0 May 30 16:15:35 ok May 30 16:15:40 will grep and tell May 30 16:16:22 CONFIG_PSTORE=y May 30 16:16:22 CONFIG_PSTORE_CONSOLE=y May 30 16:16:22 CONFIG_PSTORE_PMSG=y May 30 16:16:22 CONFIG_PSTORE_RAM=y May 30 16:16:37 ok, so it has it... May 30 16:17:16 in adb, can you do "mount -t pstore pstore /sys/fs/pstore" May 30 16:18:04 adb shell mount -t pstore pstore /sys/fs/pstore May 30 16:18:04 mount: 'pstore'->'/sys/fs/pstore': Device or resource busy May 30 16:19:43 and adb shell mount | grep pstore ? May 30 16:21:43 pstore on /sys/fs/pstore type pstore (rw,seclabel,relatime) May 30 16:22:34 ok so no pstore from the luneos kernel May 30 16:22:59 where is your kernel, on github ? May 30 16:23:18 i didn't pushed changes yet May 30 16:23:39 well the recipe does take the kernel from somewhere May 30 16:23:45 okay May 30 16:23:55 will give you the details May 30 16:24:59 can this folder name help? May 30 16:25:01 github.com.kskarthik.android_kernel_xiaomi_msm8937.git May 30 16:26:48 this kernel was used in halium May 30 16:27:22 yes, that's fine; it's also what you used in linux-xiaomi-land.bb I guess ? May 30 16:29:43 git://github.com/kskarthik/android_kernel_xiaomi_msm8937.git;branch=halium-7.1 \ May 30 16:30:00 ok, fine May 30 16:30:05 this is written in file linux_xiaomi_land_git.bb May 30 16:30:45 I think a good next step would be for you to push your modifications of meta-webos-ports and meta-smartphone on github May 30 16:30:54 it'll be easier to analyze and help May 30 16:31:41 ok. will do this tomorrow. Its night time here. 10pm May 30 16:32:07 if you don't yet have a github account you'll need one; then you'll need to fork meta-webos-ports and meta-smartphone from our repository, and last you'll need to push the changes. But yes, tomorrow then :) May 30 16:32:14 is there some way to turn off my phone? May 30 16:32:24 otherwise battery will drain in bootloop May 30 16:32:35 from twrp in the "reboot" submenu you should have "poweroff" May 30 16:32:47 or i have to flash android right now which was working earlier May 30 16:33:21 power off again started boot loop May 30 16:33:50 ah, it's because you have the usb cable still connected May 30 16:33:58 go to twrp, unplug the cable, poweroff May 30 16:34:13 :) May 30 16:34:15 ok May 30 16:35:12 Tofe, did your rosy device boot up in first attempt? i mean during your first build? May 30 16:35:28 or you also did some changes? May 30 16:35:36 asking as both are redmi May 30 16:35:47 nope, I got a bootloop too May 30 16:36:17 and what was causing that? just a rough idea... May 30 16:36:23 Same for my mido May 30 16:36:29 otherwise i'd find hard to sleep tonight May 30 16:36:30 I forgot... but it's in irc logs for sure :) May 30 16:36:33 It's usually kernel May 30 16:36:48 Hope its fixeable May 30 16:36:55 Be sure to check the defconfig with the mer verify kernel script May 30 16:37:09 BT40: there isn't yet a device we couldn't boot on May 30 16:37:22 Should've been done as part of halium but could be they forgot things May 30 16:37:36 thats the spirit .....Thats the confidence. May 30 16:37:41 We need same settings as Mer/SFOS in general May 30 16:37:48 Yes, the kernel defconfig is a good start May 30 16:38:31 * BT40 thinks you guys already knew that i'd also face bootloop......hahaha May 30 16:38:55 * BT40 even before i flashed...... hahaha. May 30 16:39:15 :) May 30 16:39:21 It's a big classic May 30 16:40:09 * BT40 scrolls and finds Tofe> Well thanks but this isn't the end of the road yet :) now let's see if that boots May 30 16:42:47 Nice to be here, Thank you, will join tomorrow and work on bootloop. May 30 16:43:04 * BT40 is preparing to shut down /sleep May 30 20:56:20 Tofe: As for emails not showing up right away I saw this in logs: May 30 21:42:34 qemux86-64 ls-hubd[274]: [2020-05-30 21:42:34:374] [0x7fe9c4d2e740] [error] [com.palm.imap.client]: Exception in command DeleteActivitiesCommand: 'details flag' must be specified (MojErr 22) -- /usr/src/debug/mojomail-imap/2.0.0-99+gitAUTOINC+446c53ce81-r0/git/imap/src/commands/DeleteActivitiesCommand.cpp:54 May 30 20:56:20 May 30 21:42:34 qemux86-64 ls-hubd[274]: [2020-05-30 21:42:34:374] [0x7fe9c4d2e740] [error] [com.palm.imap.client]: Exception in command ScheduleRetryCommand: 'details flag' must be specified (MojErr 22) -- /usr/src/debug/mojomail-imap/2.0.0-99+gitAUTOINC+446c53ce81-r0/git/imap/src/commands/DeleteActivitiesCommand.cpp:54 May 30 20:57:19 Seems this payload here is missing some info: https://github.com/webOS-ports/mojomail/blob/herrie/acg/imap/src/commands/DeleteActivitiesCommand.cpp#L44 May 30 20:57:26 Any idea how to debug/add logging there? May 31 02:37:22 do we need those do_convert_webos_preferred_gfx_image_format tasks? it's breaking the build, because sstate filenames are too long :/ May 31 02:39:27 I think we should drop them again, it was added in https://github.com/webOS-ports/meta-webos-ports/commit/2d43c815e60126f1360692c596071281b43d73e7#diff-4ae4558ddb800a4b278322396a40e9a2R7 May 31 02:40:58 webos_qmake5.bbclass webos_prerelease_dep.bbclass webos_preferred_gfx_image_format.bbclass probably aren't needed as well **** ENDING LOGGING AT Sun May 31 02:59:57 2020