| *** ChanServ sets mode: +v T4 | 02:39 | |
| T4 | xreactx was added by: xreactx | 03:36 |
|---|---|---|
| Umeaboy | After the Hello World script turned out OK I noticed how extremely hard it is to push from /rpm to my git repo droid-hal-maple even if I force push. | 04:11 |
| Umeaboy | Can't I just push the normal way from outside of /rpm? | 04:12 |
| Umeaboy | Had to do git push --force Umeaboy HEAD:master in order to make it work. | 04:30 |
| T4 | <BusterBg_18> So I already builded a 3.0.1 zip with the devel repo and worked really nice, I repeat the steps for “testing:” right? | 06:01 |
| *** ChanServ sets mode: +v T4 | 08:53 | |
| T4 | <BusterBg_18> lbt: could you please patternise nemo:testing:hw:xiaomi:kenzo? | 09:01 |
| T4 | <MarcoDS_bit> mal: I commented out any usb related stuff in init.rc but still crashesh, should I try reverting to 3.0.0.8 like others suggested? | 09:38 |
| mal | did you get the logs I asked | 09:48 |
| T4 | <MarcoDS_bit> dmesg: https://paste.ubuntu.com/p/W6BYVtFbJc/ | 09:52 |
| T4 | <MarcoDS_bit> journalctl: https://paste.ubuntu.com/p/RbV3F6gWvx/ | 09:52 |
| T4 | <MarcoDS_bit> mal: the other one gave me error (?) | 09:52 |
| T4 | srmane4444 was added by: srmane4444 | 09:57 |
| mal | @MarcoDS_bit it seems to fail quite soon, are you sure selinux is properly disabled? | 10:31 |
| T4 | <MarcoDS_bit> mal: I left AUDIT=y and in kernelcmd set selinux=0 and audit=0 so yeah it should be disabled | 10:33 |
| mal | and enabled seliux bootparam option in kernel? | 10:34 |
| T4 | <MarcoDS_bit> yes | 10:34 |
| mal | run systemctl mask droid-hal-init on device (or ln -s /dev/null /etc/systemd/system/droid-hal-init.service) | 10:35 |
| T4 | <It_sMike> I'm back :) | 10:42 |
| T4 | <It_sMike> mal : not just the browser... have similar issue while working with other parts ( even with turned off wlan ) | 10:44 |
| T4 | <It_sMike> about the droid-version building, I already have that line in my droid-config*.spec, let me send you the logs | 10:46 |
| T4 | <MarcoDS_bit> mal: what should we see with that command? | 10:58 |
| T4 | <It_sMike> mal seems the problem is gone now ( conflict between two packages error while building droid-version ) | 11:00 |
| mal | @MarcoDS_bit it might prevent some reboot issues | 11:02 |
| mal | depending on what causes those | 11:02 |
| T4 | <MarcoDS_bit> mal: done but still freezes, should it be because of the way we installed it? So: Untarring on PC and then adb pushing all the files in the directory, because it failed to extract the tar on the device (gave tar: short read) | 11:07 |
| mal | shouldn't matter, assuming the rootfs is really ok | 11:08 |
| T4 | <MarcoDS_bit> It should, we can chroot sailfish os from recovery too | 11:12 |
| T4 | <MarcoDS_bit> https://paste.ubuntu.com/p/3cprWtKh5s/ similiar to this | 11:13 |
| mal | @MarcoDS_bit what exact command did you use for extracting the rootfs? | 11:14 |
| T4 | <MarcoDS_bit> sudo tar -xvjf sfchroot.tar.bz2 | 11:17 |
| T4 | <MarcoDS_bit> Then I did sudo chmod -R a=rwx | 11:17 |
| mal | the script uses "tar --numeric-owner -xvjf " | 11:18 |
| T4 | <MarcoDS_bit> Without the first sudo it gives me error | 11:18 |
| mal | you should use that | 11:18 |
| mal | you should never run chmod -R to rootfs | 11:18 |
| T4 | <MarcoDS_bit> Oh ok, retrying with that | 11:19 |
| mal | https://github.com/mer-hybris/hybris-boot/blob/master/updater-unpack.sh | 11:19 |
| mal | that is the script used | 11:20 |
| T4 | <MarcoDS_bit> Actually the script gives us error in recovery, as on my old phone where I had to install an old recovery or I had the exact same error | 11:22 |
| T4 | <It_sMike> Hmmm find out something new... | 11:26 |
| T4 | <It_sMike> Think I've finally find a working solution | 11:37 |
| T4 | <It_sMike> Made some new changes to kernel, will report back if it's worked | 11:37 |
| T4 | <It_sMike> Building right now... | 12:03 |
| piggz | jusa: afternoon .... i have BT audio \o/ | 12:47 |
| piggz | however, it is very bad.... | 12:48 |
| piggz | Jan 27 12:47:09 Sailfish pulseaudio[3781]: Skipping 91380 us (= 16116 bytes) in audio stream | 12:48 |
| piggz | Jan 27 12:47:09 Sailfish pulseaudio[3781]: Implicit underrun of ''Need Some1' by 'The Prodigy'' | 12:48 |
| piggz | Jan 27 12:47:09 Sailfish pulseaudio[3781]: Requesting rewind due to rewrite. | 12:48 |
| piggz | Jan 27 12:47:09 Sailfish pulseaudio[3781]: Requesting rewind due to rewrite. | 12:48 |
| r0kk3rz | bad how? | 12:49 |
| piggz | r0kk3rz: choppy, very | 12:52 |
| piggz | like it cant send enough data | 12:52 |
| piggz | im wondering about my BT baud rate? | 12:52 |
| piggz | its 115200 atm | 12:52 |
| piggz | however, other sppeds dont appear to work | 12:56 |
| r0kk3rz | hmm thats a bit slow i guess | 13:10 |
| r0kk3rz | mines like 3000000 | 13:11 |
| r0kk3rz | which is like 30x faster | 13:12 |
| piggz | r0kk3rz: i get | 13:16 |
| piggz | E3:B2 failed | 13:16 |
| piggz | Jan 27 13:16:03 Sailfish bluetooth_rfkill_event[3690]: attach_hci(bluetooth_rfkill_event.c:926): Failed to execute hciattach -d -m /dev/ttyHSU0 bcm43xx 3000000 noflow sleep 00:39:25:A4:E3:B2, exiting | 13:16 |
| r0kk3rz | got h4 enabled in kernel? | 13:24 |
| r0kk3rz | CONFIG_BT_HCIUART_H4=y | 13:25 |
| piggz | r0kk3rz: yup | 13:28 |
| r0kk3rz | hmmm | 13:28 |
| r0kk3rz | what bt dev was it again? | 13:28 |
| piggz | r0kk3rz: BCM4354 | 13:32 |
| piggz | (A2) | 13:32 |
| r0kk3rz | you did confirm it was loading firmware right? | 13:40 |
| piggz | r0kk3rz: seems some use brcm_patchrom_plus to init this card to 3000000 | 13:40 |
| r0kk3rz | thats an alternative firmware patcher yeah | 13:41 |
| r0kk3rz | i dont think you should need that, at least i dont | 13:41 |
| piggz | r0kk3rz: what is your chip? | 13:44 |
| r0kk3rz | bcm4350c0 | 13:46 |
| r0kk3rz | so, not massively different to yours i shouldnt think | 13:46 |
| mal | using hybris-15.1 would probably make bluetooth easier on some of the problematic chips | 13:53 |
| r0kk3rz | using bluebinder? | 13:53 |
| r0kk3rz | piggz: it could also be a bluesleep issue, theres some patches in sony kernel related to that | 13:55 |
| piggz | r0kk3rz: ok, lemme see | 13:55 |
| mal | r0kk3rz: yes, bluebinder makes things easier | 14:13 |
| piggz | r0kk3rz: well, i build brcm_patchram_plus and it works charm | 14:13 |
| T4 | <It_sMike> I've done everything that I could... still no luck | 14:13 |
| piggz | audio working great, speed is 3000000 | 14:13 |
| piggz | ./brcm_patchram_plus --patchram /etc/firmware/bt/BCM4354A2.hcd --no2bytes --baudrate 3000000 --use_baudrate_for_download /dev/ttyHSU0 --bd_addr 00:39:25:A4:E3:B2 --enable_hci | 14:13 |
| piggz | @eugenio_g7 ^^ | 14:14 |
| T4 | <It_sMike> just gonna give up for now :) | 14:14 |
| piggz | mal: it wont make it easier on a 'pure' port ;) | 14:15 |
| mal | piggz: true | 14:25 |
| mal | @It_sMike there must be a way to fix that, it might take some time to figure it out | 14:32 |
| T4 | <It_sMike> Sure :) | 14:33 |
| piggz | r0kk3rz: mal: so, would shipping brcm_patchrom_plus be an acceptible solution? | 14:35 |
| T4 | <It_sMike> Please let me in mal if you or other porters found a solution | 14:35 |
| mal | piggz: should be fine, you probably can build it on OBS in adaptation project | 14:36 |
| mal | @It_sMike it's possible to remove the problematic configs from the device the proper way as a workaround | 14:37 |
| mal | I mean that the configs are removed already in the build stage | 14:38 |
| T4 | <It_sMike> You mean disabling that netfilter thing ? from the kernel? | 14:40 |
| T4 | <It_sMike> The package that causing this problem is named : " connman-configs-sailfish* " | 14:41 |
| mal | not sure if disabling the kernel flag will cause problems | 14:42 |
| mal | but it's possible to override that package in config repo | 14:42 |
| T4 | <It_sMike> Yep... gonna do this, till I ( or others ) found a better solution to this... | 14:43 |
| T4 | <It_sMike> Will this action break something else? just asking to be sure :) | 14:44 |
| *** ChanServ sets mode: +v T4 | 14:46 | |
| mal | not sure | 14:46 |
| mal | you probably need to copy at least one file to sparse to prevent issues, that package provides also main.conf which is probably needed | 14:47 |
| T4 | <It_sMike> Yeah I see... | 14:48 |
| r0kk3rz | piggz: did it fix yo audio? | 14:49 |
| piggz | r0kk3rz: it did | 15:10 |
| piggz | mal: r0kk3rz: currently, i build using android.mk, so will go with droid-hal-latte | 15:11 |
| piggz | not sure if a linux version exists | 15:12 |
| piggz | suppose i could make one, its only one source file | 15:12 |
| r0kk3rz | good, that baud rate was quite slow, im surprised it worked at all | 15:13 |
| r0kk3rz | i wonder whats wrong with the bluez5 patcher | 15:14 |
| piggz | r0kk3rz: i could ask in #bluez, but this will do for now... | 15:15 |
| T4 | <BusterBg_18> mal is my testing repo all right? (besides missing patterns) | 16:16 |
| mal | seems reasonable | 16:29 |
| T4 | <MarcoDS_bit> so, we had a droid-hal-startup error that we solved by searching on the irc channel, but now it still freezes and we can't see any critical error :/ | 16:59 |
| T4 | <MarcoDS_bit> dmesg https://paste.ubuntu.com/p/T4gYdPc823/ | 16:59 |
| T4 | <MarcoDS_bit> journalctl -b https://paste.ubuntu.com/p/r3GYWWByz3/ | 16:59 |
| mal | what error? | 16:59 |
| T4 | <MarcoDS_bit> Sailfish droid-hal-init: Failed to initialize property area | 17:00 |
| mal | oh, were you missing detritus package or what? | 17:00 |
| T4 | <MarcoDS_bit> we changed last line of /usr/bin/droid/droid-hal-startup.sh to /sbin/droid-hal-init as we found here http://www.merproject.org/logs/%23sailfishos-porters/%23sailfishos-porters.2017-09-29.log.html | 17:01 |
| mal | that change is obsolete | 17:02 |
| T4 | <MarcoDS_bit> oh sad | 17:02 |
| T4 | <MarcoDS_bit> should I try to build detritus? | 17:02 |
| mal | pastebin your droid-hal spec | 17:02 |
| T4 | <MarcoDS_bit> https://paste.ubuntu.com/p/YX5JDvhs5h/ | 17:04 |
| mal | detritus is automatically built if you have some things in droid-hal spec | 17:04 |
| mal | hmm, you really didn't need to add anything to that file? like straggler files definition? it's not always needed but if build fails then usually | 17:05 |
| T4 | <MarcoDS_bit> so that's not the problem?.. | 17:05 |
| mal | do you have the .rc files in config repo as told in faq? | 17:06 |
| mal | was that 15.1 port? | 17:06 |
| T4 | <MarcoDS_bit> yes 15.1 | 17:06 |
| T4 | <MarcoDS_bit> wait I'm checking now for the .rc files | 17:06 |
| mal | @MarcoDS_bit where did you see the droid-hal-init error, I have never seen anything related to that in your logs | 17:09 |
| mal | also journalctl -f might be useful to follow the log until the end | 17:09 |
| T4 | <MarcoDS_bit> https://paste.ubuntu.com/p/Y373C4KQzz/ | 17:10 |
| T4 | <MarcoDS_bit> I don't know for the .rc files, can't find the right section in the faq sorry | 17:11 |
| T4 | <MarcoDS_bit> I'll try journalctl -f asap too | 17:11 |
| mal | search 15.1 in faq | 17:22 |
| mal | the one linked in channel topic https://public.etherpad-mozilla.org/p/faq-hadk | 17:23 |
| T4 | <MarcoDS_bit> Config repo is.. what precisely? | 17:25 |
| mal | @MarcoDS_bit why did you say you added audit=0 to kernel commandline, it should't be added | 17:25 |
| mal | hybris/droid-configs | 17:26 |
| T4 | <MarcoDS_bit> Ok | 17:26 |
| T4 | <MarcoDS_bit> Mer-check says that it has to be set to 0 if is set as "y" in the defconfig if I'm not wrong, or am I? | 17:27 |
| mal | you should ignore the whole audit part of mer check | 17:32 |
| T4 | <MarcoDS_bit> Ahn ok, so I should delete audit=0 and what about the defconfig? | 17:32 |
| T4 | <MarcoDS_bit> Btw rc files added and build_packages.sh -c done | 17:32 |
| mal | you need to have selinux bootparam enabled in defconfig and selinux=0 in kernel commandline, do not make any audit related changes to either, so do not use AUDIT=n or audit=0 | 17:38 |
| T4 | <MarcoDS_bit> Ok, so I'm rebuilding hybris-boot, than zip and then another test | 17:41 |
| mal | no need to build whole image if you just change kernel, you can flash the kernel with fastboot | 17:45 |
| mal | hybris-boot.img is in out/ | 17:45 |
| T4 | <MarcoDS_bit> Oh ok, btw done, only have to test it | 17:56 |
| T4 | <MarcoDS_bit> same crash as before | 19:36 |
| T4 | <MarcoDS_bit> https://paste.ubuntu.com/p/sS56Vqk9Jq/ | 19:37 |
| T4 | <MarcoDS_bit> journalctl -f https://paste.ubuntu.com/p/njxmVc4Kws/ | 19:48 |
| T4 | <MarcoDS_bit> mal: we got it to work with "systemctl mask usb-moded.service" (thanks TheKit) 🙂 no gui btw | 20:06 |
| T4 | <It_sMike> If anyone facing reboots while working with 3.0.1.x , read here : https://github.com/CancroSailfishOS/droid-config-cancro/commit/d8a6d9cfc2225177a142fb83326f6c83c6fc7be2 , Everything's working fine for me now... no problem so far. | 20:10 |
| mal | @It_sMike didn't you have to add any "Provides: connman-configs-sailfish" to droid-config spec? | 20:11 |
| T4 | <It_sMike> Just one thing, don't forget to avoid the installtion of connman-configs-sailfish* rpm package before building system image. | 20:11 |
| T4 | <It_sMike> mal : Uh... forget to commit it on github | 20:12 |
| *** ChanServ sets mode: +v T4 | 20:14 | |
| T4 | <It_sMike> *installation :D ( sorry for mistake ) | 20:14 |
| T4 | <It_sMike> And also this one after adding those changes : https://github.com/CancroSailfishOS/droid-config-cancro/commit/cbc10c7f2d9b3adffd5c3328de1c0e4e7daa950e | 20:16 |
| T4 | <MarcoDS_bit> is hwc2 still not supported? | 20:30 |
| mal | it is | 20:32 |
| T4 | <MarcoDS_bit> okk, so my problem is probably with surfaceflinger, I'll try to fix it tomorrow | 20:33 |
| piggz | mal: latest == 3.0.1.11? | 20:33 |
| mal | piggz: afaik not yet | 20:34 |
| piggz | mal: ok ... am i ok to add a 3.0.1 to me devel? | 20:35 |
| mal | if you want, I have it for fp2 | 20:35 |
| r0kk3rz | mal: lbt was waiting on you | 20:38 |
| mal | r0kk3rz: I already said to him to go ahead and change latest | 20:39 |
| T4 | <BusterBg_18> Question, Is lbt the only one that can patternise a repo? | 20:41 |
| r0kk3rz | mal: btw is there anything you check with the new targets? common? | 20:42 |
| mal | r0kk3rz: what do you mean? I always adjust common so there are no old packages built for each target | 20:45 |
| piggz | mal: does 3.0.1 need sensor fixes? | 21:08 |
| piggz | none of my sensors work | 21:08 |
| steeeve | What has changed on sailfish since i stopped dealing with it? | 21:10 |
| mal | piggz: you forgot to update sensorfw in you OBS | 21:12 |
| piggz | mal: ok | 21:12 |
| mal | piggz: I suggest you disable latest in devel for now to not break it | 21:12 |
| mal | just in case you want to use 3.0.0.8 again | 21:13 |
| mal | steeeve: when did you stop? | 21:37 |
| steeeve | October or november i think | 21:37 |
| mal | steeeve: was 3.0.0 out already then? | 21:38 |
| steeeve | Yes | 21:38 |
| mal | ok, nothing big has happened 3.0.1 was just released a while ago | 21:39 |
| steeeve | Oh, so, i dont have to update anything? | 21:40 |
| mal | sensorfw needs updating, otherwise no new versions, except some devices need a new kernel config flag in defconfig | 21:42 |
| steeeve | Ok, i will do it now | 21:44 |
| mal | steeeve: you should check that you have CONFIG_NETFILTER_XT_MATCH_CONNTRACK=y in kernel | 21:45 |
| mal | steeeve: there has been 3 devices with issues using 3.0.1 | 21:45 |
| mal | network related kernel panics | 21:45 |
| mal | because of firewall enabling | 21:45 |
| steeeve | Luckily i have it! | 21:47 |
| steeeve | Already set | 21:47 |
| mal | ok, which kernel version does your device have? | 21:48 |
| steeeve | 3.10.106 | 21:49 |
| steeeve | Someone built things on obs for me? | 21:53 |
| steeeve | Because last build was today | 21:54 |
| mal | steeeve: which project? | 21:54 |
| steeeve | A5ultexx stuff | 21:55 |
| T4 | <BusterBg_18> mal can you help me patternise testing please?:p or can I do it myself? | 21:55 |
| steeeve | Those things | 21:56 |
| mal | @BusterBg_18 lbt does patternizing | 21:56 |
| mal | steeeve: it doesn't have any recent builds | 21:57 |
| mal | steeeve: if you mean https://build.merproject.org/project/show/nemo:devel:hw:samsung:a5ultexx | 21:57 |
| T4 | <BusterBg_18> Oh ok, is he available? | 21:58 |
| steeeve | Then what "Updated at: 2019-01-27 21:58:15+00:00" is for? | 21:58 |
| mal | steeeve: where ? | 21:58 |
| steeeve | On the failed built package | 21:59 |
| steeeve | Gst | 21:59 |
| steeeve | Ah, maybe it's when i refreshed the page since that was 1 minute ago lol | 21:59 |
| mal | yes | 22:00 |
| steeeve | •_• | 22:00 |
| mal | steeeve: I fixed your gst-droid, it just needed to be updated | 22:02 |
| steeeve | Didnt that get fixed with https://github.com/sailfishos/gst-droid/commit/4bb59235ccf226353d4ae736c16397874a1377f3 ? | 22:02 |
| steeeve | At this point i trigger a rebuild of all packages | 22:03 |
| mal | steeeve: but you probably want to wait until tomorrow, hopefully then latest target will update to 3.0.1.11 and don't have to do hacky things | 22:03 |
| steeeve | Ah ok, if i remember i will hehe | 22:04 |
| mal | steeeve: don't do that, geoclue and sensorfw will break | 22:04 |
| steeeve | Kk, will wait | 22:04 |
| mal | steeeve: sensorfw will need to be updated after 3.0.1 target | 22:04 |
| steeeve | Tomorrow if it target gets updated i will do a full rebuild, since last one was few months ago | 22:06 |
| mal | ok | 22:09 |
| mal | steeeve: yes, you should also update droidmedia, it is too old for new gst-droid | 22:21 |
| steeeve | Sure, i will do a mass update :P | 22:23 |
| piggz | mal: sensors fixed ... anything else worth updating for 3.0.1? | 22:27 |
| mal | piggz: you seem to use your own version of gst-droid instead of upstream, pulseaudio droid is slightly old, so is qt5-qpa-hwcomposer-plugin, is mce-plugin-libybris intentionally old? | 22:31 |
| piggz | mal: probbaly none intentionally....my gst-droid was from when i was devleloping the camera changes... | 22:34 |
| mal | piggz: some devices had led issues (like fp2) with new mce-plugin-libhybris but with correct configuration it worked | 22:35 |
| piggz | mal: r0kk3rz: this is what i use for BT https://github.com/piggz/brcm_patchram_plus | 22:57 |
| mal | piggz: https://github.com/mer-hybris/broadcom-bluetooth ? | 23:05 |
| piggz | mal: ah | 23:05 |
| mal | piggz: assuming you used the same version of that program, worth a try at least | 23:06 |
| piggz | mal: well, i used a random version in the android source ... clearly worth a go though! | 23:06 |
| piggz | gnight! | 23:07 |
| mal | piggz: yes, no need to reinvent the wheel | 23:07 |
| mal | piggz: did the new version work from OBS build? | 23:07 |
| piggz | mal: to not invent the wheel, you need to know the wheel has been invented! | 23:08 |
| piggz | mal: i built in android sdk cos im lazy ;) | 23:08 |
| piggz | and its currently installed in libexec/droid-hybris.... | 23:09 |
| piggz | but, i'll change that | 23:09 |
| mal | piggz: I'm adding qtmultimedia to common for new camera features | 23:10 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!