**** BEGIN LOGGING AT Mon Feb 26 03:00:02 2018 Feb 26 08:32:45 Morning! Feb 26 08:47:06 saidinesh5: You around? Feb 26 10:41:08 bshah: You around? Feb 26 10:49:30 barely Feb 26 10:52:40 bshah: Still struggling getting something useful on my mido's display..... test_hwcomposer works, but my luna-next fails still... logcat gave me: libEGL : validate_display:99 error 3008 (EGL_BAD_DISPLAY) Feb 26 10:52:55 libEGL : call to OpenGL ES API with no current context (logged once per thread) and ServiceManager: service 'display.qservice' died Feb 26 10:54:05 what is the environment variable for luna-next? Feb 26 10:54:26 i.e how do you run luna-next? Feb 26 10:55:59 QT_QPA_PLATFORM=hwcomposer Feb 26 10:56:29 Some searching pointed me to https://github.com/mer-hybris/qt5-qpa-hwcomposer-plugin/blob/294e3267a1c34545d78e85b1da45b8f26b515c33/hwcomposer/hwcomposer.pro#L44 Feb 26 10:57:46 And another one that pointed me here: https://github.com/AsteroidOS/meta-asteroid/blob/0ffbbed54ff1b369dd274f923ecc5b0e9b627ace/recipes-qt/qt5/qt5-qpa-hwcomposer-plugin/0003-Fix-build-with-Qt-5.9.patch Feb 26 10:59:53 In my dmesg I see "Feb 12 08:54:01 mido kernel: QSGRenderThread[3422]: unhandled input address range fault (11) at 0x10000000018, esr 0x92000004" Feb 26 11:00:02 So I'm tempted to try the 2nd link from asteroid guys Feb 26 11:00:09 Since we're on QT 5.9.x too Feb 26 11:00:22 I don't really know much about qt qpa hwcomposer sadly Feb 26 11:00:34 (we (KDE) don't use it) Feb 26 11:01:21 bshah: OK Feb 26 11:01:30 sorry :) Feb 26 11:01:33 I guess I'll trial and error with the AsteroidOS patch then Feb 26 11:15:11 Ah seems I had a bit outdated qt5-qpa-hwcomposer-plugin in my build Feb 26 11:15:15 Trying a new one now Feb 26 11:25:06 ooh got UI! Feb 26 11:26:03 \o/ Feb 26 11:26:11 Buggy as hell but well Feb 26 11:26:22 I.e. missing letters and elements Feb 26 11:26:23 But well Feb 26 11:26:29 missing letters! Feb 26 11:26:32 I know what to dod Feb 26 11:26:34 :P Feb 26 11:27:02 But that might be my environment.conf Feb 26 11:27:08 I added some stuff there Feb 26 11:27:12 Let me remove that first Feb 26 11:27:16 no no Feb 26 11:28:59 Well I added stuff like QT_QPA_EGLFS_PHYSICAL_WIDTH, QT_QPA_EGLFS_PHYSICAL_HEIGHT, QPA_HWC_BUFFER_COUNT, QT_QPA_EGLFS_WIDTH, QT_QPA_EGLFS_HEIGHT Feb 26 11:29:04 Which we never needed before Feb 26 11:29:22 Without these it's the same Feb 26 11:29:28 well.. you're hitting specific bug Feb 26 11:29:28 Seems I'm missing text randomly Feb 26 11:29:42 moment Feb 26 11:29:43 And also background image but that could be my config Feb 26 11:29:46 :P Feb 26 11:31:01 Herrie|Laptop: which qt version? Feb 26 11:31:46 Herrie|Laptop: try: QT_ENABLE_GLYPH_CACHE_WORKAROUND=1 Feb 26 11:32:40 if it doesn't help maybe you need to backport this I think: https://codereview.qt-project.org/#/c/185259/ I am not sure Feb 26 11:33:03 * bshah goes off now for while Feb 26 11:35:12 bshah: OK where I should put this? Feb 26 11:36:39 OK seems environment.conf Feb 26 11:39:26 bshah: That worked! Text is back! Feb 26 11:39:49 picture or didn't happen Feb 26 11:39:55 :P Feb 26 11:43:30 bshah: Before: http://tinypic.com/r/vwppag/9 and after: http://tinypic.com/r/27ycy9e/9 Feb 26 11:43:53 Still have other issues to sort.... I.e. screen turns off, maliit doesn't behave but it's at least progress Feb 26 11:44:01 yay.. (I was more interested in after pic) :P Feb 26 11:44:38 Need to get wifi & ril to behave but that shouldn't be rocket science Feb 26 11:56:17 Herrie|Laptop: \o/ Feb 26 11:58:50 Tofe: Yeah it's progress Feb 26 11:58:54 Screen turns off though Feb 26 11:59:14 I guess I want to kill qseeproxydaemon since it spams logs etc Feb 26 12:46:41 bshah: Do we care about qseeproxydaemon at all or can just disable it in init? Feb 26 12:46:58 just disable it Feb 26 13:01:07 bshah: OK Feb 26 13:31:19 JaMa: Hi Feb 26 13:31:25 We got some work to merge for you ;) Feb 26 13:31:30 https://github.com/shr-distribution/meta-smartphone/pulls Feb 26 14:02:42 bshah: Hmmz this QT_ENABLE_GLYPH_CACHE_WORKAROUND=1 works for my card shell, but not for apps I open inside this cardshell Feb 26 14:02:50 I.e. if I open QML settings app it's there Feb 26 14:02:53 Tofe: Any clues? Feb 26 14:04:08 What "is there" ? Feb 26 14:04:29 Tofe: Well garbled text as per http://tinypic.com/view.php?pic=vwppag&s=9 Feb 26 14:04:50 Herrie|Laptop: that means this variable should be set for all Qt apps, i.e. all apps Feb 26 14:04:59 Tofe: Seems so Feb 26 14:05:05 Is there a way to do this systemwide? Feb 26 14:05:29 I.e. it should be in luna-qml-launcher? Feb 26 14:05:50 I think that would be the way yes; and also in luna-webappmanager Feb 26 14:06:03 but that's quite invasive Feb 26 14:08:53 Tofe: We could add it here conditionally based on a flag we could pass per target in the bitbake recipe? Feb 26 14:08:53 https://github.com/webOS-ports/luna-qml-launcher/blob/46646bc05256835b15c50c3f14e1d803e1e2103e/src/main.cpp Feb 26 14:09:39 Maybe, but I'm looking at https://git.qt.io/consulting-usa/qtbase-xcb-rendering/commit/9ae028f507a22bd03c861e9d14c2efc4aa2efeda and I'm wondering what GPU you have Feb 26 14:09:55 Because it seems this workaround is pretty common actually in Qt Feb 26 14:10:35 Snapdragon 625 = Adreno 506 Feb 26 14:11:26 So probably a rendering string of "Adreno 4xx" Feb 26 14:11:35 let me see if they included that in later Qt versions Feb 26 14:12:32 https://github.com/qt/qtbase/blob/5.11/src/gui/kernel/qopenglcontext.cpp#L1001 not yet at least Feb 26 14:13:27 I'm tempted to suggest patching Qt Feb 26 14:14:06 You mean to add 5xx ? Feb 26 14:14:18 yup Feb 26 14:15:02 Seems there's also 6xx already Feb 26 14:15:07 So yeah patching QT makes sense Feb 26 14:15:09 We are probably amongst the first people to experience it Feb 26 14:15:14 LOL Feb 26 14:15:24 So simply adding the 5xx and 6xx should solve it in QT Feb 26 14:15:31 yes Feb 26 14:16:54 I can propose you a patch tonight, or validate one of your if feel like doing one Feb 26 14:16:59 yours* Feb 26 14:17:20 Tofe: Let me kick something off Feb 26 14:17:36 It's pretty straightforward: duplicate the lines with 4xx, and replace 4 with 5 :) Feb 26 14:18:08 Yeah will add 6xx while I'm at it Feb 26 14:18:11 Just in case Feb 26 14:18:16 ok :) Feb 26 14:23:28 Herrie|Laptop: on it Feb 26 14:44:36 Tofe: https://github.com/Herrie82/qtbase/commit/f66be04cd0bb38266dadaa190fef6aaa1193ed4e Feb 26 14:44:38 That should do it Feb 26 14:49:24 Having a build running locally now with the patch Feb 26 15:14:00 JaMa: Thnx for merge, 1 more for now: https://github.com/shr-distribution/meta-smartphone/pull/60 Feb 26 15:31:24 Herrie|Laptop: done as well Feb 26 15:41:40 JaMa: Thnx Feb 26 15:41:51 There will probably be some more over the next week, we'll see :) Feb 26 15:42:00 Tying up some loose ends for various targets Feb 26 16:02:05 JaMa: And another one for Huawei Angler (Google Nexus 6P): https://github.com/shr-distribution/meta-smartphone/pull/61 Feb 26 16:08:25 heh, ok then :) Feb 26 16:09:23 Herrie|Laptop: who would be the tester for this one? I don't have any Nexus 6P :p Feb 26 16:09:40 Tofe: The linux-unix_10 guy Feb 26 16:09:49 oh ok, good Feb 26 16:10:00 Got an image ready for him Feb 26 16:10:27 I guess we'll see him around here soon :) Feb 26 16:10:58 Pinged him on the PR Feb 26 16:11:05 JaMa: Sorry for spam :P Feb 26 16:11:19 Nexus should be straight forward Feb 26 16:11:22 He had Halium running already Feb 26 16:11:33 So all defconfig was OK already Feb 26 16:11:40 Herrie|Laptop: np, thanks for all the work Feb 26 16:11:42 Just needed to fix our gcc patches mainly Feb 26 16:12:18 I.e. last 5 commits: https://github.com/Herrie82/android_kernel_huawei_angler/commits/cm-14.1 Feb 26 18:16:26 thejsa: I have a dt.img from the Halium build, you could try that together with the other file. Not sure it'll work though since dt.img is build with gcc 4.x toolchain and kernel with gcc 6.3 Feb 26 18:17:01 But worth a try. If it doesn't work I could possibly get dt.img to build with GCC 6.3 too Feb 26 18:25:19 Tofe: new build almost ready locally. If successful will upstream patch to QT Feb 26 18:27:57 ok good **** ENDING LOGGING AT Tue Feb 27 03:00:01 2018