**** BEGIN LOGGING AT Wed Nov 27 02:59:57 2019 Nov 27 05:19:34 Morning Nov 27 05:19:40 Tofe: Some clues from log: 2019-11-26T21:24:44.339953Z [165.735256879] user.warning WebAppMgr [] WAM PALMSYSTEM {"APP_ID":"com.palm.launcher","PID":0} webOSSystem is not initialized. key:launchParams, value:{ "launchedAtBoot": true } Nov 27 05:19:41 2019-11-26T21:24:44.341886Z [165.737190771] user.warning WebAppMgr [] WAM QWARNING {} (null), QObject::connect: No such slot WebAppWayland::webPageTitleChangedSlot() Nov 27 12:37:13 JaMa: btw, what will be the next name for Yocto ?... Nov 27 12:37:49 it's alphabetical and we're at zeus :p Nov 27 12:38:42 oh, "Dunfell". Kind of disappointed :p Nov 27 13:25:23 :) Nov 27 16:24:59 Tofe: You seen my logs earlier today? Nov 27 16:25:19 They actually seem to be false alarms, since I ahve those on OSE as well ;) Nov 27 16:26:57 But this one looks more suspicious: Nov 27 12:45:11 qemux86 kernel: Chrome_InProcGp[836]: segfault at 0 ip aef4a985 sp a3ffab08 error 4 in libwayland-client.so.0.3.0[aef47000+b000] Nov 27 16:26:57 Nov 27 12:45:11 qemux86 kernel: Code: 8d 74 26 00 90 55 57 56 53 e8 d7 f2 ff ff 81 c3 8b 96 00 00 83 ec 5c 8b 7c 24 70 8d 44 24 7c 8b 74 24 74 50 6a 14 8d 6c 24 08 <8b> 17 55 8d 04 76 8b 4a 0c 8d 04 81 ff 70 04 e8 07 21 00 00 ff b4 Nov 27 16:27:07 And also Nov 27 12:45:34 qemux86 kernel: WebAppMgr[1050]: segfault at 7 ip b218837b sp bf9e3450 error 4 in libcbe.so[b030f000+781b000] Nov 27 16:27:07 Nov 27 12:45:34 qemux86 kernel: Code: 83 ec 3c 8b 75 0c 8b 7d 08 8d 46 ff 8d 93 80 5b 00 00 80 3a 00 0f 84 cc 00 00 00 80 bf 25 01 00 00 00 0f 84 bf 00 00 00 8b 00 <66> 83 78 07 7f 0f 87 1a 01 00 00 c7 46 03 03 00 00 00 8b 75 0c 80 Nov 27 16:27:19 I'll try to build for Hammerhead now and see what it does there Nov 27 16:27:41 Could be just some qemux86 specific quirk, since you said you would at least get a black window Nov 27 16:27:44 I get zero ;) Nov 27 16:30:40 ah yes, that one in libwayland seems much more likely Nov 27 16:31:16 it would be a wayland protocol version issue again Nov 27 17:11:59 Tofe: Any idea how/where to check? Nov 27 17:12:14 Tofe: Any idea how/where to check? Nov 27 17:50:13 Herrie: weeell, I'd say, "install webruntime-dbg and run wam with gdb" Nov 27 17:50:42 that way we'll have at least the stack Nov 27 17:51:34 Tofe: ooh that will be fun LOL Nov 27 17:55:03 I can guide you through that :) Nov 27 18:36:44 Tofe: If you can paste the steps I'll pick it up when I find time. Nov 27 18:36:56 Install the dbg ipk I guess Nov 27 18:47:24 yes, that's the first Nov 27 18:47:59 then, in the .sh for wam, in the last line where is starts wam, just add "gdb --args " at the beginning Nov 27 18:49:19 then start the .sh, and when gdb is started and you have input, enter "c" to continue Nov 27 18:49:25 that'll start wam itself Nov 27 18:49:30 which will crash, of course Nov 27 18:49:57 at the crash, you'll get input again in gdb, and enter "bt" for backtrace Nov 27 18:50:59 then just paste the output somewhere and we'll discuss it Nov 27 19:01:43 Tofe: OK, saves me digging through logs Nov 27 19:02:03 We'll see if I find time later tonight or early tomorrow morning Nov 27 20:40:38 BTW: doing this with remote gdb from debugfs saves the painful step of installing -dbg packages and more importantly on slower MACHINEs it prevent OOMK from killing the gdb on target before it loads e.g. chromium debug symbols Nov 27 21:27:15 JaMa: Any guidance on how to do that? Nov 27 21:40:30 Herrie: let me try the docs Nov 27 21:44:16 https://www.jann.cc/_downloads/ypdd-oct2014-advanced-class-iot.pdf around page 28 still makes sense, except for debugfs you can just enable debugfs generation from rootfs automatically Nov 27 21:44:53 #IMAGE_GEN_DEBUGFS = "1" Nov 27 21:44:53 #IMAGE_FSTYPES_DEBUGFS = "tar.gz" Nov 27 21:49:17 Tofe: https://paste.ubuntu.com/p/ZsSJ2tCnZ5/ Nov 27 21:50:53 Seems there was more: https://paste.ubuntu.com/p/34KQM6t6MB/ **** ENDING LOGGING AT Thu Nov 28 02:59:57 2019