**** BEGIN LOGGING AT Tue Dec 25 02:59:59 2018 Dec 25 11:36:57 Hi everybody! … I'm getting this error when running 'make mkbootimg': … https://pastebin.com/yXVRF8w8 … Device: Pixel XL (marlin) … halium-7.1 Dec 25 11:37:49 https://pastebin.com/raw/yXVRF8w8 Dec 25 11:38:00 This might be better (cleaner layout) Dec 25 11:38:35 I tried Ubuntu's pastebin service but it seems to have a bug with larger pastes that hides the 'Paste!' button Dec 25 11:40:23 The Halium wiki says that one should re-run 'breakfast'. … But the output in their error sample is different (their sample can't find fstab vs. mine which does find it) … And when I do, this strange phenomenon happens: … https://pastebin.com/EC6ig9aH Dec 25 11:40:57 @Waseem Alkurdi, Check fixup-mountpoints file Dec 25 11:41:40 @Waseem Alkurdi, With halium 7.1 you need to use breakfast command Dec 25 11:41:56 @Mattia990, Did this Dec 25 11:42:08 The problem is that /boot isn't mounted in fstab Dec 25 11:42:19 Wait a sec ... I'll paste it Dec 25 11:45:55 Ah damn ... Forgot to say that there are a couple fstab files here ... … One is recovery.fstab … And you have fstab.aosp_common (which is what I used, no /boot there) … And you also have vold.fstab and fstab.common Dec 25 11:47:13 There *is* a /boot entry under recovery.fstab, actually Dec 25 11:47:57 Should I use it even if (I presume) that isn't the file used by Android itself? Dec 25 11:48:35 It is of type "emmc" and the Wiki says we should skip tgese Dec 25 11:48:38 These* Dec 25 11:54:03 Ah DAMN!😓 Dec 25 11:54:34 Both 'userdata' and 'boot' are on /dev/block/sda35 ! Dec 25 11:55:26 Oh ... my fault😅 Dec 25 11:55:30 No they aren't Dec 25 11:56:03 Still, it's an 'emmc' partition ... it's supposed to be skipped if we're to follow the wiki Dec 25 11:58:05 Same error Dec 25 11:58:40 I've included the path I found for /boot in fixup-mountpoints and still throws the same error Dec 25 11:59:18 There's already port for marlin no? Dec 25 12:00:02 @bhushanshah, I want to build from totally scratch as to learn how to do things myself😁 Dec 25 12:00:08 @Waseem Alkurdi, Send it pls Dec 25 12:00:21 @Waseem Alkurdi, Sure. But you don't have to redo most of the things Dec 25 12:00:43 Point of halium is that work is shared :) Dec 25 12:01:30 @Mattia990 … pastebin.com/raw/sb3mNyWT … They turned out not to be on the same partition … But still, even if I added /boot, same error Dec 25 12:01:52 @bhushanshah, If I just clone what was done, his am I going to learn how things tick?😅 Dec 25 12:02:19 @Waseem Alkurdi, I think userdata is wrong Dec 25 12:02:30 The sda35? Dec 25 12:02:34 I'm not sure completely redoing things is a way to learn things Dec 25 12:02:46 If anything it's NIH syndrome Dec 25 12:02:52 😛 Dec 25 12:03:06 @Waseem Alkurdi, Yep Dec 25 12:03:42 Also fwiw since marlin is device with a/b partition, hybris-boot won't work with it Dec 25 12:03:55 @bhushanshah, 😂😂😂😂😂😂😂 … I understand your point, but no, I really want to reinvent the wheel as to learn how it was invented Dec 25 12:04:06 @bhushanshah, But @Sergobot did it Dec 25 12:04:28 @Mattia990 Double-checked it Dec 25 12:04:40 It is correct (the sda35) Dec 25 12:05:13 @Waseem Alkurdi, It is not using hybris-boot Dec 25 12:05:22 It is using halium-boot Dec 25 12:05:32 https://gist.github.com/Sergobot/e201cfcbc13eae3c3c9fef6ade9ad185 Dec 25 12:06:13 @bhushanshah, Ouch😓 Dec 25 12:06:33 Is it documented in the official wiki? Dec 25 12:06:43 Build instructions, etc? Dec 25 12:07:46 Oh, found it Dec 25 12:07:55 Not yet, as the port isn't finished yet Dec 25 12:08:29 @Sergobot, Marlin or halium-boot? Dec 25 12:08:36 I was planning to make a device page from it in the docs when the graphics starts working but never got to that point >< Dec 25 12:08:42 @Waseem Alkurdi, The port Dec 25 12:08:52 @Sergobot how similar is marlin and crosshatch? Dec 25 12:09:03 Crosshatch 🤔 Dec 25 12:09:26 Pixel something something Dec 25 12:09:34 @Sergobot, Exactly Dec 25 12:09:45 @Sergobot, Cross-something + something-gatch Dec 25 12:09:47 Hatch* Dec 25 12:09:52 Pixel 3 and 3 XL Dec 25 12:10:05 Combination of the two codenames Dec 25 12:10:10 Yeah, that seems to be XL 3 Dec 25 12:10:40 @Waseem Alkurdi, Oh never mind that Dec 25 12:10:47 @bhushanshah, Never reasearched the latter one Dec 25 12:11:01 It has Pie only, no? Dec 25 12:11:12 @Sergobot, Yes … Aren't they based on a common kernel? Dec 25 12:11:20 Like marlin and sailfish are? Dec 25 12:11:55 The term 'bluecross' refers to thqt Dec 25 12:11:57 That* Dec 25 12:12:03 Blue line+crosshatch Dec 25 12:12:21 Ugh Dec 25 12:12:35 (Photo, 720x1280) https://irc.ubports.com/q3apiFK7.png Dec 25 12:13:16 @Sergobot reason I'm asking is because https://wiki.postmarketos.org/wiki/Google_Pixel_3_XL_(google-crosshatch) Dec 25 12:14:03 @bhushanshah, Interesting Dec 25 12:14:45 Not sure if it will help much with marlin port Dec 25 12:14:50 There's also https://gitlab.com/postmarketOS/pmaports/issues/153 Dec 25 12:21:24 It's wonderful pmOS has this much progress on a very recent device. Though, malrin is 2 years older and I'm sure things are kinda different there :( Dec 25 12:21:48 What's SoC of marlin? Dec 25 12:21:52 @Sergobot, But we have some support in mainline, right? Dec 25 12:21:58 @bhushanshah, SDM821 Dec 25 12:22:04 @bhushanshah, Snapdragon 821 Dec 25 12:22:14 I've seen that it has mainline support Dec 25 12:22:23 But there was a catch somewhere Dec 25 12:24:05 @Waseem Alkurdi, Never heard of that Dec 25 12:24:06 Hmm no sdm821 dts in mainline trre Dec 25 12:24:22 @Sergobot, Wait … Let me find the link where I've seen this Dec 25 12:25:29 Wait … Not the 821, but the 820, and particularly on the DragonBoard 820c Dec 25 12:26:03 https://www.reddit.com/r/postmarketOS/comments/74l7fs/would_snapdragon_820821_devices_1st_gen_pixel_lg/ Dec 25 12:26:54 ["Currently they have basic ](https://git.linaro.org/landing-teams/working/qualcomm/kernel.git/tree/arch/arm64/boot/dts/qcom/apq8096-db820c.dtsi?h=release/db820c/qcomlt-4.11#n316)"Currently they have basic [support in mainline](https://github.com/torvalds/linux/blob/master/arch/arm64/boot/dts/qcom/apq8096-db820c.dtsi#L125) fo Dec 25 12:26:54 r boot and USB, and Linaro's landing repository has additional support for the screen on a 4.11 kernel." Dec 25 12:30:31 @Daniel, Okay, good to know. A patch is already available, I just need to apply it when I arrive back home in a few days. Dec 25 12:31:01 @JBBgameich, I'll do it later today or tomorrow :) Dec 25 12:40:08 What's wrong now? … Tried to build, now with hybris-boot disabled with a # in build/core/main.mk … mkbootimg builds fine, but halium-boot fails with an error: … ninja: error: unknown target 'halium-boot' Dec 25 12:40:17 The whole build then fails Dec 25 12:41:33 Is there a dependency on hybris-boot? How could there be one when hybris-boot doesn't even build? Dec 25 12:44:10 You don't need to comment it out Dec 25 12:44:25 What's the issue with hybris-boot? Dec 25 12:45:13 @Waseem Alkurdi, This Dec 25 12:45:17 This happens when I build it Dec 25 12:45:49 Has anyone stared tinkering with halium-8 maybe? Dec 25 12:48:49 @Waseem Alkurdi, This should help … https://github.com/Sergobot/android_device_google_marlin/commit/741b8b341525f919ca52a14ac587cad6acf429ca Dec 25 12:51:13 @Sergobot, Simply rename recovery.fstab to another.fstab? Dec 25 12:51:23 Or do I need to apply it properly? Dec 25 12:51:31 Well, I copied Dec 25 12:52:07 Because some script of hybris-boot ignores fstabs with recovery in their names Dec 25 12:52:36 But it's the same file inside, right? Dec 25 12:53:00 Yep Dec 25 12:53:12 So should I take out the entry for /boot from fixup-mountpoints then? Dec 25 12:53:32 I added that as per suggestion and I knew it should not be in there Dec 25 12:53:53 I didn't use fixup-mountpoints, as halium-boot doesn't rely on them Dec 25 12:54:46 Great! I'm trying to build now ... Dec 25 12:57:17 Ah, it worked!😍😍😍 Dec 25 12:57:39 (Photo, 720x1280) https://irc.ubports.com/0ngftrF3.png Dec 25 12:57:42 Building!😁 Dec 25 12:58:00 Wait! What's it downloading now? Dec 25 12:58:15 An initrd.img-touch-arm64 Dec 25 12:58:51 Is it a heavy file? Dec 25 13:00:02 no Dec 25 13:00:34 Like ~4 megabytes Dec 25 13:00:56 But it's stuck on it for some reason Dec 25 13:01:06 Did a du -h and it is exactly 4.0 megs Dec 25 13:01:17 But it's just stuck there Dec 25 13:01:24 Hmm Dec 25 13:01:35 Anything else happening? Dec 25 13:01:54 @Sergobot, Temperature is high, some fan noise Dec 25 13:02:03 But the terminal is stuck at that point Dec 25 13:02:14 Ctrl-C and try again? Dec 25 13:02:25 Just wait Dec 25 13:03:04 @Sergobot, Not a particularly patient type😭 Dec 25 13:03:44 It's probably compiling something Dec 25 13:03:54 A ccl process is in 'top' Dec 25 13:04:37 Ah, it's aborted Dec 25 13:05:02 All the output just came through at once Dec 25 13:05:52 @bhushanshah, Thanks. Just create a new changelog entry and the fixed version will be directly pushed to the repository :) Dec 25 13:06:11 @Waseem Alkurdi, No space left on the damn SSD😓 Dec 25 13:06:29 (Sticker, 512x512) https://irc.ubports.com/QRLMs0YG.webp Dec 25 13:07:49 @Waseem Alkurdi, Wait, it is building the boot image Dec 25 13:08:55 @Mattia990, But it was silent … Until it failed, when all the output came through at once … A quirk of 'mka' instead of 'make'? Dec 25 13:49:39 @Waseem Alkurdi, It is right, check the error and resolve it Dec 25 14:22:18 And finally, successfully built halium-boot ... only to be faced with an error building system image😓 … deciduous trees  Dec 25 14:22:23 Wait what Dec 25 14:22:52 I meant to paste this: … http://pastebin.com/raw/j9HgHiuQ Dec 25 14:24:38 Do I just 'touch' the missing file? Dec 25 14:29:14 Not one file ... the whole SHARED_LIBRARIES directory is missing Dec 25 14:30:19 @Waseem Alkurdi, No, something is missing and you need to add to your manifest and download it Dec 25 14:30:53 @Mattia990, And how do I know what is missing in this case? Dec 25 15:04:30 @Mattia990 @Sergobot … Anybody?😭 Dec 25 15:05:45 libcurl Dec 25 15:06:03 You showed us the instructions and did not read them? Dec 25 15:06:35 @minlexx, Of course I did ... Spent a whole day doing just that!😥 Dec 25 15:06:42 I even have them offline Dec 25 15:06:59 @aleasto, As a dependency? Dec 25 15:07:04 Wait Dec 25 15:07:15 I'll double-check them Dec 25 15:07:24 Yeah, you need to download it too Dec 25 15:07:36 At the moment libcurl isn't included by default Dec 25 15:07:54 @WaseemAlkurdi, `As of 08/01/2018 libcurl isn't downloaded by default. So, a couple more patches: repo_manifest.patch and halium_build_libcurl.patch Apply them (you already know how to) in .repo/manifest and build directories respectively. Finally, run repo sync -c again for these two last patches to take effect.` just from th Dec 25 15:07:55 e instruction Dec 25 15:08:28 Thanks, Alexey :) Dec 25 15:08:57 @minlexx, What section is this note from? @alexeysergey Dec 25 15:10:49 @WaseemAlkurdi, AH SHIT!😪 Dec 25 15:11:06 That's from @Sergobot 's instructions😭 Dec 25 15:11:35 Okay ... No more not reading the instructions from now on! I promise😅 Dec 25 15:12:06 Hey, I have just got a stupid question iny head … Can Ubuntu Touch support phones with a notch ? Dec 25 15:12:43 @aurnytoraink, The part behind the notch won't be visible Dec 25 15:13:02 But just get Halium running before thinking of that😂 Dec 25 15:13:24 @WaseemAlkurdi, 😢 Dec 25 15:13:57 @WaseemAlkurdi, Okay, so they need to do an alternative Dec 25 15:14:51 @aurnytoraink, When there are a whole lot of notched phones out there *running* *the* *OS* successfully, it would be trivial to fix Dec 25 15:15:27 Just get it running ... and these small details would be taken care of in time Dec 25 15:16:40 Okay Dec 25 16:32:55 Another failure ... … https://pastebin.com/raw/xnwwjLsN Dec 25 16:33:43 flex-2.5.39: loadlocale.c:130: _nl_intern_locale_data: Assertion `cnt < (sizeof (_nl_value_type_LC_TIME) / sizeof (_nl_value_type_LC_TIME[0]))' failed. Dec 25 16:33:52 This in particular is the error Dec 25 16:34:28 LC_TIME however is declared properly as end_US.UTF-8 Dec 25 16:51:04 Flex error is a documented error Dec 25 16:52:07 what is documented solution though? Dec 25 16:52:16 easiest is LANG=C make -jwhateverf Dec 25 16:52:19 [Edit] easiest is LANG=C make -jwhatever Dec 25 16:52:53 @UniversalSuperBox, Ouch ... Dec 25 16:53:11 Broke my promise once more😪 Dec 25 16:53:16 No big deal. :) Dec 25 16:53:42 People do miss it often though. I wonder if there's a better way to show it off. Dec 25 17:09:31 @UniversalSuperBox, 😂😂😂😂😂😂😂 … Writing it in BIG BOLD RED letters possibly? Dec 25 17:25:56 Didn't work ... Dec 25 17:25:59 Same error Dec 25 17:26:35 But checking using 'locale' shows that everything is still set to 'en_US.utf8' except for LANG Dec 25 17:27:42 Wait … It's 'en_US.utf8' vs. 'en_US.UTF-8' when not setting LANG=C Dec 25 17:29:22 And on another tangent, in the Wiki, this error also results in an 'Aborted (core dumped)' message after the main error. This didn't happen here. Dec 25 17:30:16 Setting USE_HOST_FLEX=true doesn't work either Dec 25 17:31:46 Nor it's actually using the host flex Dec 25 17:32:58 The error shows flex version 2.5.39 … But the host flex is version 2.6.4 Dec 25 17:36:35 How about hardcoding the value of the assertion(Or hardcoding a path to the host flex)? Dec 25 17:42:22 https://pastebin.com/raw/VEYfzVjK Dec 25 17:42:44 Passed the flex assertion error, but there, another error Dec 25 17:47:26 @WaseemAlkurdi, system/sepolicy/untrusted_app.te:210:ERROR 'unknown type exfat' at token ';' on line 19585: … }:{ dir { { chr_file blk_file } { file lnk_file sock_file fifo_file } } } { create unlink }; … #line 210 … checkpolicy: error(s) encountered while parsing configuration Dec 25 17:47:40 It clearly says that it's an sepolicy error Dec 25 17:48:07 @theimpulson, Don't tell me it's documented or I'll shoot myself Dec 25 17:48:17 Moreover your defconfig has a lot of duplicate flags. Fix them as well. Dec 25 17:48:48 @theimpulson, That's probably because I've taken it from LineageOS Dec 25 17:48:58 Leftovers from their work Dec 25 17:49:00 @WaseemAlkurdi, Did you removed sepolicy inclusion in board? Dec 25 17:49:22 @theimpulson, Nope ... Dec 25 17:49:33 BoardConfig.mk? Dec 25 17:49:37 Yes Dec 25 17:49:55 No one told me to 😥 Dec 25 17:50:07 Will give it the once-over now Dec 25 17:50:34 You are supposed to remove sepolicy completely. Halium doesn't supports it and it's based OS uses AppArmor. Dec 25 17:51:02 @theimpulson, But if I remember closely, the device requires SELinux to be enabled Dec 25 17:51:05 Permissive won't boot Dec 25 17:51:29 (But again, is that why we have to disable verity?) Dec 25 17:52:00 Ah screw Verizon Dec 25 17:52:11 It's their damned name in here Dec 25 17:52:25 (Photo, 720x1280) https://irc.ubports.com/ZU492XKU.png Dec 25 17:52:49 Should I also go and remove everything related to that from the device tree? Dec 25 17:53:03 Commenting out should do Dec 25 17:53:14 Try building now Dec 25 17:53:25 But this seems to be a path in there Dec 25 17:53:46 Pointing at device/google/marlin/sepolicy/verizon Dec 25 17:53:59 Or is it unimportant? Dec 25 17:54:38 Commenting out = Doesn't matter what it is, it won't get included during building Dec 25 17:55:00 Built again, same error Dec 25 17:55:09 Let me check BoardConfigVendor.mk Dec 25 17:55:45 Empty except for comments Dec 25 17:56:12 Same sepolicy error Dec 25 17:56:17 Again Dec 25 17:56:35 Should I make clean? Dec 25 18:05:39 Not sure but Dec 25 18:05:49 Try reverting this commit Dec 25 18:06:09 https://github.com/LineageOS/android_system_sepolicy/commit/83840c7f863cd30bd0df3c6db3701e39eee2a097 Dec 25 18:06:48 Before we do though … Could it be related to the fact that I'm using the host's flex instead of the prebuilt? Dec 25 18:07:09 Nope Dec 25 18:09:57 Do I do a got revert or a got reset though? Dec 25 18:10:08 Probably just a revert, right Dec 25 18:10:24 Not reverting Dec 25 18:17:07 Ah what's wrong?! Dec 25 18:17:14 The thing isn't reverting Dec 25 18:18:06 It's complaining because the .got directory is gone Dec 25 18:19:52 .git * Dec 25 19:32:43 Merry Christmas Dec 25 21:53:40 Monty was added by: Monty Dec 25 23:34:24 Merry Christmas 🎄🎁 to everybody ! Dec 25 23:58:54 Congratulations for everyone’s efforts ! You are building the future of open source for mobile devices ! Long life to Halium ! 🥳 Dec 26 02:06:49 Actura was added by: Actura **** ENDING LOGGING AT Wed Dec 26 03:00:01 2018