T42 | <DylanVanAssche> Okay :) also the pixel ratio can be set as well in the rpm for testing. The covers are 1/4 of the screen :p The keyboard is like 1/2 of the screen :p | 06:23 |
---|---|---|
T42 | <DylanVanAssche> @adampigg I was wrong, the pixel ratio is fine, the icons are too big :P | 06:57 |
T42 | <DylanVanAssche> However, for some reason, I can't force the orientation to landscape. I added the `default_allowed_orientations=15` though | 07:01 |
T42 | <DylanVanAssche> Works from Settings, but not from the quick actions thing | 07:01 |
T42 | <DylanVanAssche> * you can set it in Settings, but it's ignored | 07:02 |
T42 | <adampigg> Does the sensor work? | 07:11 |
T42 | <adampigg> Mine is booting but screen stays off | 07:12 |
T42 | <DylanVanAssche> No, there's something wrong with the sensor driver. It seems missing, even when compiled, needs further investigation. In the Qt5 Components app, orientation forced works. So it's probably the sensor that's doing difficult | 07:13 |
T42 | <DylanVanAssche> For the PineTab: pixel ratio = 1.0 and icons = z1.25 seems to be nice. z1.0 seems to be blurred and too small for the status bar. @adampigg can you verify this a bit? You ported more device than I :P | 07:15 |
T42 | <adampigg> Yes, sensor is needed iirc | 07:16 |
T42 | <adampigg> Do u know why my screen wont be turning on? | 07:16 |
T42 | <adampigg> Correct kernel boots i think | 07:17 |
T42 | <DylanVanAssche> Well did you compiled the right kernel? The pine64-kernel branch with pine64-defconfig. | 07:18 |
T42 | <DylanVanAssche> Also, make sure that you have the modules copied over. They are different than our previous kernel, especially display. | 07:18 |
T42 | <DylanVanAssche> If you done all that, a log would be nice | 07:18 |
T42 | <adampigg> Thats what i thought i did, i will denug more later, it was after midnight when i last tried :) | 07:22 |
T42 | <DylanVanAssche> Okay ;) After I did those things, it worked out of the box :) | 07:29 |
T42 | <adampigg> Typical :D | 07:31 |
T42 | <DylanVanAssche> Hahah :P | 07:32 |
T42 | <DylanVanAssche> Well, the PineTab has a battery drain issue due to the WiFi. Powersaving is broken atm in the driver :/ | 07:32 |
T42 | <DylanVanAssche> Also charing is broken, the charger keeps turning on/off in SFOS. Works fine in Android | 07:33 |
T42 | <DylanVanAssche> However, that's broken as well in the PinePhone | 07:34 |
T42 | <adampigg> @dylanvanassche when your tab arrived, what os was on it, and in what state? | 07:53 |
T42 | <adampigg> When mine booted on the serial console, it was a linux kernel, but you get a login prompt, but there was nothing on the screen | 07:54 |
T42 | <adampigg> @DylanVanAssche journalonly says this about lipstick | 08:27 |
T42 | <adampigg> Jul 26 02:43:23 Sailfish systemd[4272]: Starting The lipstick UI... | 08:27 |
T42 | <adampigg> Jul 26 02:43:23 Sailfish lipstick[4426]: [W] unknown:0 - drmModeGetResources failed | 08:27 |
T42 | <adampigg> Jul 26 02:43:23 Sailfish systemd[4272]: lipstick.service: Main process exited, code=killed, status=11/SEGV | 08:27 |
T42 | <adampigg> @DylanVanAssche this looks important https://bpaste.net/show/rMA4 | 08:30 |
vknecht | piggz, glad you like the share button :-) but why are you cutting your supercing ? | 08:37 |
vknecht | *supercinq | 08:37 |
piggz | vknecht: its a french car it has holes! | 08:38 |
vknecht | ah, fixing rust damages ? | 08:38 |
vknecht | how old is it ? 30, 35 years old ? | 08:40 |
piggz | vknecht: 30 | 08:41 |
vknecht | impressive, not sure when I last saw one here... but... does it run sfos ? :-) | 08:46 |
piggz | vknecht: well, maybe it could witha new head unit, but im torn between keeping it original, and upgrades! | 08:48 |
piggz | infact, it doesnt even have an ecu | 08:48 |
vknecht | go cyberpunk :-) | 08:48 |
piggz | i did read that some models did have injectors ... i think they musy have been european only, never heard of them over here | 08:49 |
vknecht | I hope finding parts is not too difficult | 08:56 |
piggz | vknecht: no, seems fine so far, its in good condition...i had to make a new radiator fan brancket and fan though | 09:01 |
piggz | waseasier than buying a real one ... motor technology had improved in 30 years,its much smaller | 09:01 |
piggz | mal: vknecht: does one on the X devices display portrait selfie pictures in the wrong orientation, in the gallery or viewer ? | 10:22 |
piggz | works fine on mido | 10:23 |
T42 | <DylanVanAssche> piggz: failed to bound message isn't happening on mine. Mine has Android on it and boots fine into that when no SDcard with SFOS is available | 10:26 |
T42 | <DylanVanAssche> I have also 2GB RAM :p | 10:26 |
piggz | hmmm | 10:26 |
T42 | <DylanVanAssche> The Android image is fine to run FB for example. Everything is working there | 10:27 |
T42 | <adampigg> i wonder if mine is faulty | 10:29 |
piggz | mal: what enables the tablet-ui ? | 11:32 |
rinigus | is there any order in filling up fixup-mountpoints? | 11:37 |
piggz | i dont think so | 11:39 |
rinigus | piggz: thanks! | 11:39 |
rinigus | one more question (still have to check on aosp): if `ls -l /dev/block/bootdevice/by-name/` returns system_b -> /dev/block/sda52, do I fill sda52 or block/sda52 in the fixup-mountpoints? | 11:41 |
piggz | rinigus: not sure, does this help https://github.com/sailfish-on-fxtecpro1/hybris-boot/blob/master/fixup-mountpoints#L78 | 11:42 |
piggz | that is an a/b device | 11:43 |
rinigus | piggz: I presume sony xz2 is also a/b device. do you have sdaXX in /dev or under /dev/block? | 11:49 |
vknecht | piggz, not tried on X yet, but on idol3 selfies seems flipped horizontally, both for portrait and landscape | 11:49 |
rinigus | ... when reading comment in the header of fixup-mountpoints, it sounds like map pointer should correspond to device under /dev | 11:52 |
vknecht | strange, bt worked on aosp8-based X after boot (tho file transfer failed), and failed again when stop-restarting it | 12:00 |
vknecht | on X it seems selfies in the viewfinder are flipped horizontally (unlike on idol3), and indeed in gallery there's a vertical flip | 12:08 |
vknecht | has to do with the way the sensor is physically mounted/rotated, I guess ? | 12:15 |
T42 | <adampigg> Damn different phones!!!! | 12:45 |
T42 | <adampigg> Maybe we should use the page rotation, and kill 2 birds with one stone by properly supporting portrait | 12:45 |
*** electro is now known as Guest7643 | 14:00 | |
T42 | <DylanVanAssche> If you have only an accelerometer, you have only to set `GSensor=1` in the `hw-settings.ini` file of the CSD tool? Or any other switches in their? | 14:24 |
vknecht | hmm, is it up to the cam app to fill metadata ? seems pics/vid taken with jolla-cam have a bit more than those taken with pov | 16:28 |
vknecht | at least for focuslength and flash | 16:31 |
mal | @adampigg usually tablet ui is enable based on display size, not completely sure | 16:32 |
jellypony | Hello everyone! Having problem then building android core. One of the problem is that "/boot (/data also) appears to live on...blablabla". I used fixup-mountpoints but it didn't help. What can cause it? | 17:54 |
vknecht | jellypony, to clarify you could tell which device, what and how you're building, and which mod you made to fixup-mountpoints | 18:11 |
jellypony | Xiaomi nitrogen (mi max 3). Following official hadk. Im using community's LineageOS 16. Device tree is here: github.com/kirillinda . Building of an android base. Log: pastebin.com/xLFZpiSG . fixup-mountpoints : pastebin.com/26URUT7w . There is a problem with missing dependencies too, but i want to fix mounting first. | 18:36 |
mal | jellypony: btw, hybris-16.0 requires one extra step to build, official hakd pdf doesn't have it | 18:46 |
mal | jellypony: you need to run "hybris-patches/apply-patches.sh --mb" before you run make hybris-hal | 18:46 |
mal | the fixup probably wasn't the problem but that missing command | 18:49 |
*** jellypony1 is now known as jellypony | 19:36 | |
T42 | <BusterBg_18> mal: I successfully built hybris16 for Xiaomi-Sirius and already fixed the config errors thrown by kernel checker | 19:37 |
T42 | <BusterBg_18> Is there another extra step before running build-packages? | 19:38 |
mal | you might want to add these https://github.com/mer-hybris/droid-config-sony-ganges-pie/tree/1e174da014db99f5d4253556ca38ab0cefc49dd4/sparse/usr/libexec/droid-hybris/system/etc/init to your config repo, make sure you use the ones from that commit, newer versions have some changes which probably might not work without some other changes | 19:41 |
mal | jellypony: did you see my message about your build issue? | 19:44 |
T42 | <BusterBg_18> mal: build-packages fails at Middleware https://pastebin.com/uFUDdDLw | 21:29 |
mal | cd hybris/mw/ofono-ril-binder-plugin/ and then git reset --hard 1.0.6 and then rpm/dhd/helpers/build_packages.sh -b hybris/mw/ofono-ril-binder-plugin | 21:30 |
deathmist | @BusterBg_18 cd hybris/mw/ofono-ril-binder-plugin/ && git reset --hard 1.0.7 && cd - | 21:30 |
deathmist | xD that was quick, 1.0.7 works for me as well | 21:31 |
mal | ok, probably works, 1.0.6 is in 3.1.0 release I think | 21:31 |
T42 | <BusterBg_18> Cool, that did it | 21:36 |
T42 | <BusterBg_18> Now nfcd-binder-plugin also failed | 21:36 |
deathmist | same thing, except reset to 1.0.4 | 21:37 |
T42 | <BusterBg_18> Another git reset I presume | 21:37 |
mal | in similar case a good way is to check the latest official changelog on tjc and check which version is listed there and do git reset using that | 21:38 |
mal | and if the package is not listed in the latest changelog the check earlier versions | 21:38 |
mal | of course channel logs often might be a quicker way to check those versions | 21:42 |
deathmist | @BusterBg_18 you may want to bookmark https://piggz.co.uk/sailfishos-porters-archive/ if you don't want to setup a full local copy of the log archive ;) | 21:44 |
T42 | <BusterBg_18> deathmist: cool, thanks for the link, I used to do the lazy way of just searching on Telegram | 21:47 |
T42 | <BusterBg_18> One more thing, according to logs I should skip qt5-qpa-surfaceflinger-plugin, right? | 21:48 |
deathmist | yep, most (if not all) new devices use hwcomposer | 21:48 |
T42 | <BusterBg_18> building packages was successful after reset --hard and skipping surfaceflinger-plugin now it's time for building the rootfs | 21:54 |
T42 | <BusterBg_18> But I have one question | 21:55 |
T42 | <BusterBg_18> Why didn't it asked me for straggler files? | 21:56 |
T42 | <BusterBg_18> I remember them being needed for xiaomi-kenzo/kate | 21:57 |
mal | @BusterBg_18 there is something wrong now in droid-hal build, it doesn't seem to complain about straggler files anymore, need to investigate why | 22:06 |
mal | you probably have to manually check the root folder somewhere under out/ for *_context files and add those to straggler files manually | 22:07 |
T42 | <BusterBg_18> Ok I'm going to try that, and after that I suppose I have to rebuild droid-hal | 22:09 |
jellypony | Oh, mal, thanks a lot. Now it compiles but i still got a warning about improper boot and data mounting. Also a bunch of new ones. Cant figure out whats wrong :C Log: https://pastebin.com/EzahdCuK | 22:10 |
jellypony | As I said, I already used fixup-mountpoints script | 22:10 |
jellypony | D: | 22:11 |
mal | hmm, wondering if the incorrect indentation in the line before "nitrogen" in fixup-mountpoint could cause issues | 22:24 |
*** juiceme is now known as Guest38757 | 22:43 | |
T42 | <BusterBg_18> mal: ok im at a loss, I cant find anything related to _context | 22:49 |
T42 | <BusterBg_18> everything thats under out/*/root https://pastebin.com/5R8RsK6K | 23:04 |
mal | @BusterBg_18 ok, then you don't need those, you will probably need these: bt_firmware, dsp, firmware, persist | 23:07 |
T42 | <BusterBg_18> What happens if I add more like vendor for example? | 23:13 |
mal | I assume you have /vendor as a mount point already | 23:13 |
mal | in which case it shouldn't be in straggler files | 23:14 |
mal | some of those are automatically ignored | 23:14 |
T42 | <BusterBg_18> Oh ok, it's just that on kenzo all of these were needed: | 23:16 |
T42 | <BusterBg_18> /bugreports\/d\/file_contexts.bin\/init.qcom.sh\/property_contexts\/sdcard\/selinux_version\/service_contexts\/vendor\ | 23:16 |
T42 | <BusterBg_18> So that's where my doubts come from | 23:16 |
T42 | <BusterBg_18> But I will give it a shot | 23:16 |
mal | well I think it might not be a problem if it's there, not completely sure | 23:17 |
mal | which device is that+ | 23:18 |
T42 | <BusterBg_18> Kenzo? Or the one I'm currently working on? | 23:21 |
T42 | <BusterBg_18> Kenzo=Xiaomi redmi note 3 | 23:21 |
T42 | <BusterBg_18> Sirius=Xiaomi mi 8 SE | 23:21 |
mal | just had a look at the fstab, /vendor is a mount point so probably you can skip it from that list for sirius but as the folder is empty it should be fine also the have it | 23:30 |
T42 | <BusterBg_18> Ok so I added the ones you said plus bugreports and d just to be safe | 23:40 |
T42 | <BusterBg_18> Currently redoing build_packages | 23:41 |
mal | you only need to run "build_packages.sh -d" if you do that kind of changes | 23:46 |
T42 | <BusterBg_18> Oh, I thought everything needed to be rebuilt | 23:46 |
T42 | <BusterBg_18> BTW what does packages.sh -v builds exactly? | 23:47 |
T42 | <BusterBg_18> I don't remember tbh | 23:47 |
mal | -v builds droid-hal-version rpms | 23:48 |
T42 | <BusterBg_18> I just know it's needed after -m | 23:48 |
mal | it depends on what you have changed to know what is needed to be built again | 23:48 |
mal | those straggler files changes are just packaging things and won't affect anything else, if you edit config repo then you need to do -c | 23:49 |
mal | if you rebuild hybris-hal or do any other bigger changes to android side then you should build also middleware | 23:50 |
mal | I did once make a flowchart for those which would give some idea when to use what but I never made that public | 23:51 |
T42 | <BusterBg_18> That would have been useful for when I was doing kenzo, I just rebuilded everything whenever I changed something | 23:52 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!