| T42 | <BNONA_20> Fwd from Hfx: ๐๐๐ฌ๐ข๐๐๐ฅ๐ฅ๐ฒ ๐ญ๐ก๐จ๐ฎ๐ ๐ก๐ญ ๐ข๐ญ ๐ฐ๐๐ฌ ๐ ๐ฌ๐๐๐ฆ ๐ฎ๐ง๐ญ๐ข๐ฅ ๐ ๐๐๐๐ข๐๐๐ ๐ญ๐จ ๐ ๐ข๐ฏ๐ ๐ข๐ญ ๐ ๐ญ๐ซ๐ฒ ๐ฐ๐ข๐ญ๐ก ๐ ๐ฅ๐ข๐ญ๐ญ๐ฅ๐ ๐ฌ๐ญ๐๐ซ๐ญ๐ฎ๐ฉ ๐จ๐ ๐๐๐200 ๐๐ง๐ ๐ ๐ ๐จ๐ญ ๐๐๐๐ค ๐๐๐2250 24 ๐ก๐จ๐ฎ๐ซ๐ฌ ๐จ๐ ๐ญ๐ซ๐๐๏ฟฝ | 05:53 |
|---|---|---|
| T42 | <BNONA_20> ๐๐๐๐๐ก ๐ก๐ข๐ฆ ๐ฏ๐ข๐ ๐ก๐ข๐ฌ ๐ญ๐๐ฅ๐๐ ๐ซ๐๐ฆ ๐๐ก๐๐ง๐ง๐๐ฅ ๐๐๐๐๐๐ | 05:53 |
| T42 | <BNONA_20> https://t.me/joinchat/kitZi9rmmFwyOWY0 | 05:53 |
| T42 | <edp_17> mal: yes, I am still doing the hammerhead. | 07:19 |
| T42 | <edp_17> @Chrisme69 : I'll find you with direct message to avoid spamming this channel with that boot issue. | 07:20 |
| T42 | <Chrisme69> Thanks | 07:20 |
| T42 | <Chrisme69> A lot | 07:21 |
| srxl | Hello all! Looking to start work on a port for a device I have. I have a working (close to) mainline kernel for my device, although I can't seem to find any information on getting an SFOS build going without using Android sources. Is it possible to just build on top of a mainline kernel? | 08:11 |
| mal | srxl: there are at least 3 devices that boot to UI without android using a linux kernel | 09:14 |
| T42 | <edp_17> Hi all, I have an OBS related question. How can I add a repository with a custome name? (Currently I only have latest_arm7hl but I would like to create one for sfos3.4.0.24 and another one for 4.0.1.48 and some other for older versions.) | 10:36 |
| techgeekster | So I erased everything and started from scratch. It went smoothly this time until the --mic step. Error <creator>[03/12 10:32:56] : Unable to find package: patterns-sailfish-device-configuration-raphael | 10:37 |
| Thaodan | edp_17: You should do that per project since the repository is defined by its packages e.g. when you need to stop updating a package for version x but need to for version y. | 10:38 |
| techgeekster | Note: In my .ks file it does not have the patterns-sailfish-device-configuration-raphael packege in the %packages section | 10:41 |
| mal | @edp_17 so you really want a testing repo? | 10:45 |
| mal | @edp_17 similar to this https://build.sailfishos.org/project/show/nemo:testing:hw:fairphone:fp2-sibon just let me know for which device you need it and I can add such a repo | 10:46 |
| T42 | <edp_17> mal: idk I just wanted to add a separate repo for every sfos version I built. | 10:46 |
| T42 | <edp_17> mal: yes, I'd like to something like that. | 10:47 |
| mal | @edp_17 that is what testing is meant for, releases from public should be made from testing repos not devel | 10:47 |
| mal | *for public | 10:47 |
| T42 | <edp_17> mal: Oh, I see. | 10:48 |
| mal | because that way the repos can remain intact for older releases and won't get broken if you change something | 10:48 |
| mal | especially when you disable build for the version you have already released | 10:48 |
| T42 | <edp_17> mal: yes, this is what I need. | 10:49 |
| mal | so tell me which device you such a repo and I will add it, I will be out for a moment but will add after that | 10:49 |
| T42 | <edp_17> mal: I have project for i9100 with one repo in there latest_armv7hl. There I'd like to create a repo for 3.2.1.20, 3.3.0.16, 3.4.0.24 and 4.0.1.48 | 10:50 |
| mal | @edp_17 I don't see such device on OBS yet at all, which vendor is that | 10:51 |
| mal | ah, looked at wrong place, will add the repo once I'm back | 10:52 |
| T42 | <edp_17> mal: I also have a treltexx project where currently there is no repo at all. Here I'd like to add 3.3.0.16, 3.4.0.24 and 4.0.1.48 | 10:52 |
| T42 | <edp_17> mal: okay, no rush with that! Thanks. | 10:52 |
| T42 | <edp_17> mal: plus I'd like to have two new projects (hammerhead and n7000) Can I create those somehow? | 10:53 |
| mal | creating new projects on OBS is limited to few people like me, I can handle those also, what is vendor of n7000? | 10:54 |
| T42 | <edp_17> mal: Thanks! The vendor od n7000 is also Samsung | 10:55 |
| T42 | <edp_17> mal: for the n7000 Id like to have one repo so far: 4.0.1.48 | 10:58 |
| *** Mister_Magister_ is now known as Mister_Magister | 11:01 | |
| T42 | <edp_17> mal: For just in case I let you know the repos I'd like to have for the hammerhead: 3.2.0.12, 3.2.1.20, 3.3.0.16, 3.4.0.24 and 4.0.1.48 | 11:02 |
| T42 | <edp_17> Thanks for your help in advance! | 11:03 |
| techgeekster | In patterns I added sailfish-connman-plugin-suspend and now in mic it says that it cannot find it (Warning[03/12 11:27:41] : repo problem: nothing provides sailfish-connman-plugin-suspend needed by patterns-sailfish-device-adaptation-raphael-1-202103120934.aarch64, ) | 11:28 |
| techgeekster | Does anyone know how can I add it? | 11:28 |
| T42 | <zinstack625> techgeekster: you have to manually clone and compile it | 11:46 |
| T42 | <edp_17> Another question regarding OBS. Should I add all packages that I have in local repo $ANDROID_ROOT/hybris/mw ? | 11:47 |
| techgeekster | Thank you. For the logs, I did it with: rpm/dhd/helpers/build_packages.sh --mw=https://github.com/mer-hybris/sailfish-connman-plugin-suspend.git | 11:48 |
| T42 | <edp_17> @edp_17 [Another question regarding OBS. Should I add a โฆ], Or should I list all of them what I listed in _pattern/jolla-configuartion-$DEVICE and _pattern/jolla-hw-adaptation-$DEVICE and _pattern/sailfish-porter-tools? | 11:53 |
| srxl | mal: awesome, so it's doable! are there any resources on how to do it, or anything from those devices I can take a look at to get an idea as do how it goes? | 11:56 |
| T42 | <zinstack625> How's the state of hybris 17? | 11:57 |
| deathmist_ | mal: could you take a look at and fix compilation of https://github.com/piggz/sailfish-fpd-community/pull/35? I encountered the same ".../file.list: No such file or directory" under aarch64 | 13:11 |
| swanux | I'm rebuilding now, and I faced that gst-photography error that was mentioned above. I tried the recommended revert (to 20201104) but common/ is empty (so the build fails). Where should I look for common/gst-autogen.sh? | 13:12 |
| mal | swanux: how did downgrade gst-droid? | 13:17 |
| deathmist_ | swanux: it seems to be a submodule of gst-droid and I actually had to install git on the target tooling using "sb2 -t $VENDOR-$DEVICE-$PORT_ARCH -m sdk-install -R zypper in -y git" before I could run build_packages.sh --gg successfully | 13:17 |
| mal | deathmist_: you can't really use --gg now because latest gst-droid won't work with 4.0.1 | 13:18 |
| mal | or anything older either | 13:18 |
| swanux | mal: I followed your description (later, as common was empty I tried from release tags also) | 13:19 |
| swanux | deathmist_: I'll try it then, thanks | 13:19 |
| deathmist_ | right, well with my patch (https://github.com/mer-hybris/droid-hal-device/pull/288) and a modified build_packages.sh I can :) | 13:19 |
| mal | swanux: did you follow the instructions I posted couple of days ago: go to hybris/mw/gst-droid and run "git clean -x -f -d" and then "git reset --hard 0.20201104.0" and then check that the common/ folder in there has content, then back to $ANDROID_ROOT and run rpm/dhd/helpers/build_packages.sh -b hybris/mw/gst-droid | 13:19 |
| swanux | mal Exactly | 13:19 |
| mal | if common folder is empty try git submodule update | 13:20 |
| swanux | Ok | 13:20 |
| mal | if that doesn't help then I have other ideas | 13:20 |
| swanux | Common still empty | 13:23 |
| mal | git submodule update --init --recursive --remote | 13:25 |
| mal | then check git diff | 13:25 |
| piggz | mal: i notived ofono plugin had a mjor version jump too, for a new api level 1.2? | 13:28 |
| swanux | The first command solved it. I successfully built it. Thank you! How should I proceed? (as -gg isn't usable) | 13:29 |
| mal | piggz: yes | 13:30 |
| mal | piggz: it got support for the 1.2 version of the android radio hidl api | 13:31 |
| piggz | mal: could do with better support for the pinephone modem! | 13:32 |
| mal | piggz: what? how is that in any way related to pinephone? | 13:35 |
| piggz | mal: not the ril driver .. but the pinephone ofono plugin in general could do with love | 13:35 |
| mal | yeah, probably, where is the latest version of that? | 13:36 |
| piggz | mal: the version we are shipping is here https://git.sailfishos.org/piggz/ofono/tree/pinephone_plugin mostly from @neochapay | 13:37 |
| piggz | mal: other distros are either using modem manger, or the newer upstream ofono | 13:52 |
| piggz | iirc, sailfish ofono is uniquely old :D | 13:52 |
| T42 | <calebccff> Does hybris-17.1 make it to UI for anyone? | 14:09 |
| mal | yes | 14:13 |
| T42 | <calebccff> o.o right | 14:18 |
| mal | my device has only issues with fingerprint sensor and camera | 14:22 |
| mal | camera I know how to make work in a hacky way | 14:22 |
| calebccff | mal: is it aarch64? | 14:25 |
| mal | yes | 14:25 |
| deathmist_ | speaking of fingerprint and aarch64, does someone have ideas about https://pastebin.com/raw/hfyKs2wF? I'm using the community fpd sources from https://github.com/piggz/sailfish-fpd-community/pull/35 with @rinigus' patch applied so the RPM builds pass, but now the issue appears to be with sailfish-devicelock-fpd? | 14:27 |
| piggz | solution 1 seems safe | 14:28 |
| mal | yes solution 1 looks correct | 14:33 |
| T42 | <elros34> c option seems more correct. patterns-sailfish-device-configuration-$DEVICE will basically pull whole rootfs to your target . Why do you want to install everything in target? | 14:38 |
| rinigus | deathmist_: nut why do you install like that? | 14:38 |
| rinigus | (but) | 14:38 |
| deathmist_ | ok that doesn't seem to be what I should be doing anyway; I've got "Unable to find package: patterns-sailfish-device-configuration-$DEVICE" during "build_packages.sh --mic" on aarch64, droid-config repo at https://github.com/sailfishos-oneplus5/droid-config-cheeseburger, what could I be missing? this didn't seem to happen on my previous arm32 builds | 14:39 |
| rinigus | deathmist_: you have to compile it, but not install | 14:41 |
| mal | deathmist_: you do know about the bug with aarch64 and --mic | 14:41 |
| mal | ? | 14:41 |
| deathmist_ | mal: there was an older issue which I believe was fixed (the adaptation repo being in KS, which doesn't exist for me) | 14:42 |
| deathmist_ | that was weeks ago at this point though | 14:43 |
| rinigus | deathmist_: maybe useful. this is a hadk that I am using for tama aarch64 (based on seine): https://github.com/sailfishos-sony-tama/main/blob/hybris-10/hadk-sony-xz2.md | 14:43 |
| mal | deathmist_: yes, my fix however broke some sed magic in build_packages.sh and now there are no adaptation repos in .ks | 14:43 |
| deathmist_ | mal: so what should I do, revert the patch and remove the extra repo that doesn't exist again manually? | 14:46 |
| piggz | rinigus: hadk as md is actually a great idea | 14:46 |
| deathmist_ | ^ +1 | 14:46 |
| rinigus | piggz: that's the one based on seine wiki; earlier one that I used for aosp9 is based on ganges. so, idea was from jolla originally | 14:47 |
| rinigus | and does help to keep all small details for later. already helped me few times :) | 14:47 |
| mal | deathmist_: hack build_packages.sh so it works again | 14:49 |
| swanux | mal: I finished repackaging with your patch. Unfortunately /system/etc/ld.config.28.txt mountpoint still doesn't exist, and lipstick still dies. Any ideas? | 14:55 |
| *** techgeekster_ is now known as techgeekster | 14:55 | |
| mal | swanux: your device doesn't need that | 14:56 |
| mal | swanux: does strace look the same? | 14:58 |
| mal | swanux: with or without the libGLESv3.so change? | 14:58 |
| swanux | with libGLESv3.so it was just worse (even test_hwcomposer died early). Without it, it's as usual | 14:59 |
| swanux | mal: This was the problem with it : https://pastebin.ubuntu.com/p/9rtghQYr47/ | 15:00 |
| swanux | (related strace - https://pastebin.ubuntu.com/p/vXS6pqDD7f/) | 15:00 |
| mal | swanux: ok, I have an idea, copy the ld.config.txt from /system/etc to /usr/libexec/droid-hybris/system/etc/ then edit the file and add some changes there, under [system] section you see list of lines namespace.default.search.paths add a new line with "namespace.default.permitted.paths += /usr/libexec/droid-hybris/system/${LIB}" after those lines, then same for the [vendor] section | 15:10 |
| mal | swanux: oops, sorry the file is /system/etc/ld.config.vndk_lite.txt | 15:11 |
| mal | then create a similar systemd mount service you did earlier but that vndk_lite in every filename and file content instead of 28 | 15:12 |
| mal | hopefully you understand what I tried to say | 15:13 |
| swanux | mal: Yes, I got it. Just one question: in [system] and [vendor] sections is it okay if I put it at the end? | 15:14 |
| deathmist_ | mal: what exactly broke the sed? I could attempt fixing that and make a PR | 15:15 |
| mal | swanux: you have to put it after the other namespace.default.search.paths line in the section | 15:16 |
| swanux | Ok | 15:17 |
| mal | swanux: like this https://github.com/mer-hybris/droid-config-sony-ganges-pie/blob/master/sparse/usr/libexec/droid-hybris/system/etc/ld.config.28.txt#L269 | 15:17 |
| mal | deathmist_: the sed assumes there is a common adaptation repo | 15:17 |
| mal | deathmist_: https://github.com/mer-hybris/droid-hal-device/blob/master/helpers/build_packages.sh#L421 | 15:18 |
| mal | deathmist_: I'm considering removing the common repo from local builds completely | 15:18 |
| swanux | mal: Done. Now reboot? | 15:20 |
| mal | swanux: can you show the mount service you made | 15:20 |
| swanux | mal: https://pastebin.ubuntu.com/p/D6QcgfmY4S/ | 15:21 |
| mal | swanux: ok, and the symlink enabling that is also correct? | 15:21 |
| mal | is so then reboot | 15:21 |
| swanux | Yes. ln -s to ../$FILE. Now rebooting then | 15:22 |
| swanux | Reboot done | 15:22 |
| techgeekster | How can I get the journal from the boot. Device stays on oem boot logo | 15:23 |
| swanux | mal: That's my new mount https://pastebin.ubuntu.com/p/cQ2n9VHRyZ/ | 15:23 |
| mal | techgeekster: is telnet working? | 15:23 |
| mal | swanux: any help with graphics side? | 15:24 |
| techgeekster | A new USB Ethernet device apears in my system but I dow not get an ip | 15:24 |
| techgeekster | And telnet does not work | 15:24 |
| mal | techgeekster: so what all things have you done for the device? could you push the config repo to github | 15:25 |
| techgeekster | Wait a sec. I'll push the Android Source and the fixup_mountpoints | 15:26 |
| swanux | mal: No improvement so far. Only difference I noticed is when running /system/bin/surfaceflinger (I do it so my oem logo wont burn in) it now outputs this: WARNING: linker: /system/etc/ld.config.vndk_lite.txt:45: warning: appending to undefined property "namespace.default.permitted.paths" (treating as assignment) | 15:26 |
| mal | swanux: show the file | 15:26 |
| mal | swanux: ah, I think I made a mistake | 15:27 |
| T42 | <ankaos> hฤฑ. how fix this problem? https://paste.ubuntu.com/p/ygP5MZ5gfy/ | 15:27 |
| mal | @ankaos which android base? | 15:28 |
| T42 | <ankaos> 16.0 android version | 15:28 |
| mal | did you apply the patches correctly? | 15:28 |
| T42 | <ankaos> @mal [did you apply the patches correctly?], yes | 15:28 |
| calebccff | techgeekster: some linux distros don't play nice with host config. try manually configuring your IP to be `192.168.2.20` on the interface that appears, then try telnet | 15:29 |
| techgeekster | Ok. Will do | 15:30 |
| swanux | mal: https://pastebin.ubuntu.com/p/Gngrp4sZDd/ | 15:30 |
| mal | swanux: edit the new lines you added by replacing namespace.default.permitted.paths with namespace.default.search.paths | 15:31 |
| swanux | Ok | 15:32 |
| techgeekster | calebccff: I got telnet now. Thank you | 15:33 |
| techgeekster | mal : This is the logcat, if you can find something. https://paste.gnome.org/pzgcxj0ve. I am git pushing as we speak | 15:35 |
| swanux | mal: Ok, warning disappeared. Nothing on GUI front though | 15:36 |
| mal | swanux: strace | 15:39 |
| mal | techgeekster: get also output of journalctl -b --no-pager | 15:40 |
| techgeekster | mal: This is init.log https://paste.gnome.org/ptzxwrjfk. JournalCtl says No journal files were found. | 15:41 |
| swanux | mal: This is for lipstick - https://pastebin.ubuntu.com/p/nQBjTsGy52/ | 15:41 |
| mal | techgeekster: which kernel version does your device have? | 15:42 |
| mal | swanux: did you remove the /usr/libexec/droid-hybris/system/lib64/libGLESv3.so ? | 15:43 |
| techgeekster | mal: 4.14.83 | 15:43 |
| swanux | mal: Yes | 15:43 |
| mal | swanux: add it back | 15:44 |
| T42 | <ankaos> mal: how fix error? | 15:44 |
| swanux | mal: Copying v2 to v3? | 15:45 |
| mal | swanux: yes | 15:46 |
| mal | @ankaos not sure yet | 15:46 |
| swanux | mal: This outputs the same - https://pastebin.ubuntu.com/p/fbDHGjxvwn/ | 15:46 |
| T42 | <elros34> @techgeekster if you created init_enter_debug2 then to continue booting you need to follow instruction you can see when you login | 15:52 |
| techgeekster | T42: I removed the init_enter_debug2, still on oem logo. New init log https://paste.gnome.org/p6ejlwah8 | 15:59 |
| T42 | <calebccff> weird that /etc/ doesn't exist? | 16:00 |
| T42 | <elros34> init.log is usually useless, get journal | 16:00 |
| mal | which telnet port? | 16:01 |
| T42 | <calebccff> techgeekster: fyi, t42 is just the telegram bridge bot, the actual person is at the start of the message. For people in TG do "@USERNAME" | 16:01 |
| deathmist_ | mal: good news is that my aarch64 rootfs.tar.bz2 at least has /var/run as a symlink. if the filesystem package install issue didn't happen to anyone else here on arm32 base either I'd say it's safe to say it was because there were too many layers of containers/chroots going on when I built the last one. btw here's the PR to fix local aarch64 community builds: | 16:01 |
| deathmist_ | https://github.com/mer-hybris/droid-hal-device/pull/299 | 16:01 |
| techgeekster | @Calebccff Oh ok. Did not knew that. | 16:02 |
| techgeekster | @elros34 due to not understanding irc, I am sending it again. I removed the init_enter_debug2, still on oem logo. New init log https://paste.gnome.org/p6ejlwah8 | 16:06 |
| mal | I'll ask again, which telnet port? | 16:07 |
| T42 | <elros34> @techgeekster and I will repeat again get journalctl. init.log is not very useful | 16:07 |
| mal | ah, that seems 2323 probably | 16:08 |
| techgeekster | @elros34 Ok, what could I give that would be more useful as a log? I am now finishing the github repo https://github.com/sailfish-raphael | 16:08 |
| T42 | <elros34> type: journalctl --no-tail -f | 16:09 |
| mal | does "systemctl" show anything | 16:11 |
| techgeekster | @elros34 This is the result that repeats itself Feb 09 21:27:23 Mi9TPro kernel: pm8150b_charger: smblib_get_prop_typec_power_role: TYPE_C_MODE_CFG_REG = 0x10 | 16:19 |
| techgeekster | Feb 09 21:27:23 Mi9TPro kernel: pm8150b_charger: smblib_get_apsd_result: read APSD_RESULT_STATUS stat=0x1 | 16:19 |
| techgeekster | and Feb 09 21:28:38 Mi9TPro kernel: binder: 3452:3452 transaction failed 29189/-22, size 32-0 line 3020 | 16:20 |
| techgeekster | This is the whole log https://paste.gnome.org/p8e7d7x4a | 16:22 |
| T42 | <elros34> this log lacks early boot messages. You need to reboot and then try again so journal contains all relevant information since beginning of boot | 16:25 |
| mal | swanux: try to reorder the search paths in ld.config file, move the libexec on first, note the need to change the assignment += and = when needed | 16:25 |
| mal | @elros34 persistent journal would help | 16:26 |
| T42 | <elros34> @techgeekster or ^ but then also use -b0 argument | 16:26 |
| techgeekster | Ok i'll try in a bit and i'll give you the results | 16:27 |
| swanux | mal: Changed the order, now I got this with surfaceflinger - https://pastebin.ubuntu.com/p/y2mC6zTjnZ/ | 16:35 |
| mal | strace | 16:36 |
| swanux | (and the device got into bootloop after that) | 16:37 |
| swanux | surfaceflinger or the others? | 16:37 |
| mal | anything | 16:37 |
| swanux | mal: From surfaceflinger - https://pastebin.ubuntu.com/p/4TZx8ykchq/ | 16:40 |
| swanux | And it seems the bootloop isn't because of surfaceflinger / test_hwcomposer - It occurs after ~40s from boot (probably lipstick?) | 16:42 |
| T42 | <elros34> swanux did you mask user@100000 before testing test_hwcomposer? | 16:43 |
| T42 | <elros34> or surfaceflinger | 16:43 |
| swanux | @elros34 No. Should I? (as surfaceflinger worked before, even without masking) | 16:44 |
| T42 | <elros34> I think so otherwise you will have restarting/hanging lipstick which might interfere with your tests | 16:46 |
| swanux | And it worked before because lipstick died at early state anyway (?) | 16:47 |
| T42 | <elros34> didn't some of patches you use break surfaceflinger? | 16:49 |
| swanux | I don't think so. I mean, at the end of the day, yeah, they did, but I think it's because what you said, lipstick is still in the game (which is good I guess). I just applied those ld.config ... things following advice from mal | 16:51 |
| mal | @elros34 yes, a hacky attempt at ld.config.vndk_lite.txt | 16:51 |
| mal | swanux: wait, are you saing lipstick runs now? | 16:51 |
| swanux | I don't know, but otherwise what would cause the bootloop / crash in that state? | 16:52 |
| deathmist_ | mal: https://github.com/mer-hybris/hybris-patches/pull/16 implements the cg2_bpf removal patch for hybris-16.0 | 17:20 |
| poetaster | damn. | 17:39 |
| mal | swanux: what do you see in logs | 17:39 |
| poetaster | Just helped a dev. debug for the Volla port and I'm confused. | 17:42 |
| poetaster | should a community port provide a writable: /system/etc/? Is that not an android thing? | 17:43 |
| poetaster | context is this thread: https://forum.sailfishos.org/t/defender-app-update-harbour-defender-my-very-own-sailfish-devember/3670/66 | 17:43 |
| swanux | mal: Here is the journalctl until the first bootloop / crash - https://pastebin.ubuntu.com/p/NHZPr8FYGV/ | 17:44 |
| poetaster | sorry, noise. context is Vollaphone. piggz ting. I'll waite. | 17:45 |
| mal | poetaster: I don't understand why you would need to do anything in /system/etc | 17:53 |
| poetaster | I don't. app called defender (hosts and cookie jar) ad blocking. | 17:55 |
| mal | poetaster: so what is the problem with it, too much text in that thread | 17:56 |
| poetaster | on the vollaphone, it's a read only file system. his app installs but fails to write (lame). | 17:57 |
| poetaster | I'd say, he's doing 'too much'. | 17:57 |
| poetaster | userland guy says, sigh, it's not me, it's the community port. | 17:58 |
| poetaster | I say, you make assumptions that are not warranted. | 17:58 |
| T42 | <elros34> depends on port but most use /system from lineage and that is mounted as read-only. | 17:59 |
| mal | yes, very few community ports have /system as rw | 18:00 |
| poetaster | hmm. so he's shooting self in foot. his app will fail on many ports. | 18:00 |
| poetaster | I'll tell him. | 18:00 |
| poetaster | should I be nice about it :) | 18:00 |
| poetaster | mal: any luck on fp2 3.4 ? I need a second device for dev work. | 18:09 |
| poetaster | sigh: [ -f %{_a1configdir}/hosts.editable ] && echo โ%{_a1configdir}/hosts.editable existsโ || cp %{_a1configdir}/hosts %{_a1configdir}/hosts.editable 2>/dev/null || : | 18:11 |
| swanux | mal: Do the logs say something useful? | 18:14 |
| mal | poetaster: well the only issue is updating to 3.4.0 | 18:24 |
| mal | swanux: that tell android side process are failing | 18:25 |
| poetaster | mail: thanks. let me know if I can held. | 18:25 |
| mal | poetaster: there is way around it but the real reason is a kernel bug | 18:25 |
| poetaster | s/held/help/ | 18:28 |
| poetaster | mal: 'way around?' | 18:29 |
| mal | poetaster: I mean disable a certain feature for a certain systemd service so it won't crash the kernel | 18:29 |
| poetaster | mal: dsme still. | 18:36 |
| mal | yes | 18:36 |
| poetaster | which feature? | 18:36 |
| mal | PrivateNetwork sandboxing | 18:37 |
| rinigus | just released 4.0.1.48 for sony tama aosp9/arm32, thank you all for help! now just need to fix few issues with my apps and will be back with aosp10/aarch64 porting. assuming that nothing major will happen with the current release. | 18:37 |
| poetaster | ah. remember. turn it off :) | 18:38 |
| mal | poetaster: what only helps for update to 3.4.0, for 4.0.1 I need patched kernel anyway so it would better to have in 3.4.0 a kernel which would then work with 4.0.1 also | 18:40 |
| mal | *that | 18:40 |
| poetaster | mal: got it, got it. I'm interested in a 3.4.0 | 18:41 |
| mal | poetaster: well it's possible to update to it if you really want, you just might need to do some extra things when updating to 4.0.1 later | 18:41 |
| poetaster | I don't even have SFOS on the FP2 yet. I'm running SFOS on a volla. | 18:42 |
| poetaster | But i'd like to install 3.4 on an FP2. | 18:43 |
| mal | poetaster: I can give you a 3.4.0 image for fp2 if you really want | 18:43 |
| poetaster | Why not. On top of 18, not 19, or? | 18:44 |
| mal | it's the same base as in the instructions | 18:44 |
| poetaster | you mean merproj wiki? | 18:45 |
| mal | yes | 18:47 |
| mal | I need to copy the image to I made some time ago to some place | 18:48 |
| poetaster | mal: wait. maybe it's a time waster. | 18:48 |
| mal | poetaster: ? | 18:52 |
| poetaster | mal: just realized 19 is on it. I can do a 18.04.1 install on top? | 18:52 |
| poetaster | It's an old phone (new battery :) | 18:53 |
| mal | yes, you can install the old version on top of it | 18:54 |
| poetaster | ok. will do. feed me an image. | 19:21 |
| mal | poetaster: did you get the private message with the link? | 19:38 |
| poetaster | mal: yes I did. | 19:48 |
| poetaster | biab | 19:49 |
| mal | rinigus: I pushed the fix to community fpd aarch64 PR | 21:03 |
| rinigus | mal: thank you very much! it should have fixed it, will be able to check tomorrow. | 21:29 |
| mal | rinigus: I can test again also | 21:30 |
| techgeekster | @elros34 I did everything from scratch and even implemented some fixes from the channel logs. Still the same problem. Here is the full journal log https://paste.gnome.org/pami2gl83 | 22:08 |
| mal | I can't seem to access that paste | 22:11 |
| techgeekster | Let me reupload it to my gdrive | 22:11 |
| techgeekster | @mal https://drive.google.com/file/d/16YJeB8F6Mw6VZlS51czOUS1tPUnTcEUg/view?usp=sharing | 22:12 |
| *** techgeekster_ is now known as techgeekster | 22:21 | |
| mal | techgeekster: was that hybris-17.1? | 22:22 |
| mal | techgeekster: you seem to have some issue with selinux | 22:22 |
| techgeekster | @mal no hybris16 | 22:22 |
| techgeekster | I have deactivate it in the boot args | 22:23 |
| techgeekster | Should I deactivate it in the kernel too? | 22:23 |
| mal | did you see any unpackaged files in droid-hal rpm build? | 22:24 |
| techgeekster | Let me check | 22:24 |
| mal | Feb 09 20:05:47 Mi9TPro droid-hal-init: Could not read properties from '/plat_property_contexts': No such file or directory | 22:24 |
| mal | Feb 09 20:05:47 Mi9TPro droid-hal-init: Failed to initialize property area | 22:24 |
| mal | do you see that file anywhere in out/ | 22:25 |
| mal | in your build environment | 22:25 |
| techgeekster | this is the droid-hal build https://paste.gnome.org/pxaybfugo | 22:26 |
| mal | techgeekster_: can you check if you have plat_property_contexts anywhere in $ANDROID_ROOT/out if it's there it's in some subfolder deep down there | 22:28 |
| techgeekster_ | As far as I can see it is not found. Checking more as we speak | 22:29 |
| mal | this should tell if it's there: find out/ -name plat_property_contexts | 22:29 |
| techgeekster_ | @mal returned empty | 22:30 |
| mal | hmm | 22:30 |
| *** techgeekster_ is now known as techgeekster | 22:31 | |
| techgeekster | @mal In the whole $ANDROID_ROOT folder no results are found for plat_property_contexts | 22:34 |
| mal | techgeekster: so you didn't need straggler files in droid-hal spec? | 22:34 |
| techgeekster | @mal nope | 22:35 |
| techgeekster | You can see the files I used for the build here (https://github.com/sailfish-raphael). The repos are one commit behind, but mostly the same | 22:36 |
| mal | do you see that file on device in /system or /vendor? | 22:41 |
| mal | techgeekster: ^ | 22:43 |
| techgeekster | @mal No | 22:43 |
| techgeekster | neither in /system nor in /vendor | 22:43 |
| techgeekster | Should I search the phone for this folder? I mean the LineageOs + SFOS combo that is installed right now | 22:46 |
| mal | maybe | 22:46 |
| techgeekster | Ok wait a sec | 22:46 |
| mal | not sure if that is the real issue or not | 22:46 |
| mal | but the property area failure is a problem | 22:46 |
| T42 | <elros34> @techgeekster check your $ANDROID_ROOT/droid-hal-$DEVICE.log. Looks like your partitioons are not mounted. | 22:47 |
| mal | hmm, indeed no sign of system mount | 22:48 |
| techgeekster | I cant see any errors of fault. Just some warning. Let me upload the file | 22:49 |
| T42 | <elros34> also missing all hybris-16 *.rc files mentioned in hadk-faq | 22:49 |
| techgeekster | Oh ok. I did all the steps of the FAQ. Maybe they were not implemented? Will do those again | 22:51 |
| mal | the first problem to solve is the missing mounts | 22:51 |
| techgeekster | @elros34 This is the droid-hal-raphael-log https://drive.google.com/file/d/1WXPgk5lPT7z2vtu1Q6tARB99i77vEwSL/view?usp=sharing | 22:51 |
| techgeekster | @mal no such thing on the device either | 22:52 |
| mal | techgeekster: the problem is those other things we just noticed | 22:52 |
| mal | techgeekster: can you show your fixup-mountpoints changes | 22:53 |
| T42 | <elros34> unrelated but does sailfish-browser (sfos 4.0.1) also crash for you on that google drive site or it's only my device? | 22:53 |
| mal | I haven't tried to use that on sailfish-browser | 22:54 |
| mal | maybe I can test tomorrow | 22:54 |
| techgeekster | @mal these are the mountpoints https://drive.google.com/file/d/1aGYTrhEdJo0s8Be-kAluatgqi-A-sZTr/view?usp=sharing | 22:55 |
| mal | techgeekster: I think I know the problem | 22:55 |
| mal | maybe | 22:55 |
| techgeekster | @mal Nothing to lose! Tell me! | 22:56 |
| mal | can you run "make modules" in the HABUILD_SDK where you built make hybris-hal | 22:56 |
| techgeekster | @mal Yes, wait a second | 22:56 |
| mal | and show the list of modules it prints | 22:56 |
| mal | quite sure you bumped into the issue that random devices don't copy everything automatically to out/ like .rc files and fstab file | 22:57 |
| techgeekster | @mal should I share the result via gdrive or pastebin? | 23:00 |
| mal | either is fine | 23:00 |
| techgeekster | @mal https://paste.gnome.org/pyg8kfsx5 | 23:00 |
| mal | techgeekster: do you see any fstab files in out, use this: find out/ -name '*fstab*' | 23:05 |
| techgeekster | Yes, 5 files, 2 in hybris https://paste.gnome.org/puhkiscgu | 23:05 |
| mal | how about init.qcom.rc | 23:06 |
| techgeekster | @mal No results for that | 23:07 |
| mal | techgeekster: ok, copy file device/xiaomi/raphael/rootdir/etc/fstab.qcom to out/target/product/raphael/root/fstab.qcom | 23:09 |
| techgeekster | @mal. Scratch that. One result but not in /out, rather in device/xiaomi/raphael/rootdir/etc | 23:10 |
| mal | techgeekster: then also run make init.qcom.rc in the HABUILD_SDK (make sure you run the needed initial stuff before that i.e. breakfast) | 23:10 |
| mal | after doing that go to platform sdk and use "rpm/dhd/helpers/build_packages.sh -d" to repackage the droid-hal rpms | 23:11 |
| mal | then show again the droid-hal rpm build log | 23:12 |
| techgeekster | Ok. Should I copy and the init.qcom.rc file? | 23:12 |
| mal | that make init.qcom.rc should copy it | 23:13 |
| techgeekster | @mal Oh ok! Will show you the droid-hal build log when it's finished | 23:14 |
| mal | techgeekster: you can check again if the .rc file appeared in out/ | 23:14 |
| techgeekster | The script put it in hadk/out/target/product/raphael/system/etc/init/init.qcom.rc | 23:16 |
| mal | that is good | 23:17 |
| techgeekster | @mal New droid-hal build log https://drive.google.com/file/d/1oT_7VjueA0mZ57ZY7QDbBZKfldakJ9V1/view?usp=sharing | 23:20 |
| mal | techgeekster: ok, you need to do some changes still but looking better, in the droid-hal spec you have the line with makefstab_skip_entries, in that line replace the /vendor with / | 23:26 |
| techgeekster | @mal I have bunch of entries, including / (%define makefstab_skip_entries / /vendor /dev/stune /dev/cpuset /sys/fs/pstore /dev/cpuctl) | 23:28 |
| mal | remove the /vendor | 23:28 |
| techgeekster | Oh ok | 23:28 |
| mal | then rebuild droid-hal rpms again | 23:29 |
| techgeekster | Okk, doing in now | 23:29 |
| mal | techgeekster: after you have done that also make sure you copy the .rc files mentioned here https://github.com/mer-hybris/hadk-faq#hybris-16-0 to correct place in your config repo in case you haven't done it already | 23:30 |
| mal | then it should be time to build a new image and see what happens | 23:31 |
| mal | (of course if you change config repo then first run build_packages.sh -c) | 23:31 |
| techgeekster | @mal Okk. I'll do all that and inform you if I get a boot this time. | 23:31 |
| techgeekster | @mal Thank you very much for your time! | 23:32 |
| deathmist_ | mal: could you review https://github.com/mer-hybris/droid-hal-device/pull/299 & https://github.com/mer-hybris/hybris-patches/pull/16? | 23:33 |
| mal | deathmist_: yeah, tomorrow | 23:37 |
| deathmist_ | cool, those are kinda needed fixes :p | 23:49 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!