| r0kk3rz | @edp_17: dont use that one, you should use the one from themuppets | 00:00 |
|---|---|---|
| T42 | <edp_17> @r0kk3rz: Is this the one: https://github.com/TheMuppets/proprietary_vendor_samsung | 00:01 |
| r0kk3rz | yeah | 00:01 |
| mautzone | r0kk3rz: What are the differnces from the two repos? | 00:02 |
| T42 | <edp_17> Ok, so I need to change this in local_manifest, then do a repo sync fetch submodules, then rebuild hybris-hal? | 00:02 |
| r0kk3rz | mautzone: hard to know exactly, because those repos are filled with blobs | 00:03 |
| mautzone | r0kk3rz: Ok. | 00:03 |
| r0kk3rz | but if you build things against the wrong blobs, you're going to get segfaults | 00:03 |
| mautzone | r0kk3rz: Seems to be worth a try. | 00:03 |
| r0kk3rz | so you really need to use the same source as the base image | 00:03 |
| T42 | <edp_17> Ok, I'll try then. | 00:04 |
| r0kk3rz | that repo is huge btw, so maybe just download the hero folder | 00:05 |
| mautzone | What are the advantages of Bluez5 instead of Bluez4 in SFOS? | 00:05 |
| T42 | <edp_17> @r0kk3rz: so I need to define 'TheMuppets/proprietary_vendor_samsung/hero-common' rather than 'TheMuppets/proprietary_vendor_samsung' ? | 00:06 |
| r0kk3rz | you cant do that | 00:07 |
| T42 | <edp_17> Yep, I just realized. | 00:07 |
| T42 | <edp_17> Then how can I do that? | 00:07 |
| r0kk3rz | mautzone: lots of things really, but if you've got an old kernel you need bluez4 | 00:07 |
| T42 | <edp_17> What does it mean old? All android phone use old kernel. :( | 00:08 |
| T42 | <edp_17> 3.4 is it old. I think so. | 00:08 |
| T42 | <edp_17> 3.18, this one too. | 00:08 |
| r0kk3rz | 3.4 is too old for bluez5 | 00:08 |
| r0kk3rz | 3.18 isnt | 00:09 |
| mautzone | r0kk3rz: I cherry picked the necessary changes to get bluez5 working on 3.4 kernel from ubports. | 00:09 |
| T42 | <edp_17> Ok, thanks. | 00:09 |
| r0kk3rz | yeah you can backport the whole bt stack | 00:09 |
| r0kk3rz | mal did that on fp2 | 00:09 |
| mautzone | r0kk3rz: But i think i went into another problem. I can set up the bt device with btattach, but it takes several minutes to find other bt devices and then the connection fails. | 00:10 |
| mautzone | r0kk3rz: Ialso tried it with mals backports as the hadk-faq suggests, but this didn't work. | 00:11 |
| T42 | <edp_17> @r0kk3rz: before I start the fetch submodules, is it not causing problem if there are stuff in there from the other repo? | 00:11 |
| T42 | <edp_17> I have just checked what's in my vendor/samsung/hero-common. I have only that stuff what's in here: https://github.com/TheMuppets/proprietary_vendor_samsung/tree/cm-14.1/hero-common | 00:14 |
| T42 | <edp_17> But how can I use only this bit from the huge TheMuppets/proprietary_vendor_samsung tree? | 00:14 |
| r0kk3rz | @edp_17 i suspect thats why driim made his own repo, and copied the blobs he needed into that | 00:15 |
| T42 | <edp_17> I see. Manually. | 00:16 |
| r0kk3rz | but for now, so long as the files are in the folder, you dont need to resync | 00:16 |
| T42 | <edp_17> Oh boy, lots of files especially in the proprietary sub folder. :( The lots of file means, lots of chance to make mistakes. | 00:17 |
| T42 | <edp_17> Is there a way to save a file from github with firefox? I don't want to copy/paste all files' content because I will surely make mistakes. | 00:21 |
| r0kk3rz | mautzone: what kind of bt you have? | 00:21 |
| T42 | <edp_17> Nevermind, I know how. | 00:22 |
| mautzone | r0kk3rz: What do you mean? Chipset? | 00:23 |
| r0kk3rz | mautzone: yeah | 00:23 |
| mautzone | bcm 4335 | 00:23 |
| r0kk3rz | ah broadcomm | 00:24 |
| r0kk3rz | was it working before the backport? | 00:24 |
| mautzone | r0kk3rz: Yes. | 00:26 |
| r0kk3rz | there might be some bluesleep fixes you need now | 00:27 |
| mautzone | r0kk3rz: Ok, but is there any kind of log for bt? | 00:28 |
| r0kk3rz | theres a bunch of stuff in bluez5-tools for probing | 00:28 |
| mautzone | r0kk3rz: You mean like bluetoothctl, hciconfig, btmon and so on? | 00:29 |
| r0kk3rz | yeah | 00:29 |
| T42 | <edp_17> @mautzone: If this helps. When I did the hammerhed build I needed to install the connman-tool because the bluetooth_sleep uses a function from there. The BT was broken until I installed the connman-tools, then it worked fine. | 00:30 |
| T42 | <edp_17> Bluez4 of course. | 00:30 |
| mautzone | r0kk3rz:Everything seems ok, and since it found my XA2 after several minutes of searching, i think that it is working...kind of. But the XA2 was not able to find the Nexus. | 00:32 |
| r0kk3rz | yeah sounds like the bluesleep issues we had | 00:33 |
| r0kk3rz | you installed the rfkill service? | 00:33 |
| mautzone | r0kk3rz: Yes, but i don't use it. I haven't figured out what it does actually. Does it only block or unblock or does it manage the conection via hciattach? | 00:35 |
| r0kk3rz | it is mostly used to load the firmware on startup | 00:35 |
| r0kk3rz | in a consistent way | 00:35 |
| mautzone | r0kk3rz: Id there some documentation? | 00:37 |
| r0kk3rz | not really, other than the stuff in the faq i wrote | 00:37 |
| r0kk3rz | the config file is fairly self explanatory | 00:37 |
| r0kk3rz | you would have the details already for however you're using hciattach | 00:38 |
| mautzone | r0kk3rz: Using btattach. | 00:38 |
| Simdol | How do you fix: Fatal: 'oneplus-enchilada-armv7hl' is not a known build target | 00:42 |
| Simdol | for build_packages.sh | 00:42 |
| r0kk3rz | Simdol: you made the target? did you give it the right name? | 00:43 |
| Simdol | r0kk3rz: I don't think so, how do you do that? | 00:43 |
| mautzone | Simdol: Wrong target name. | 00:43 |
| r0kk3rz | Simdol: it says in the hadk | 00:43 |
| Simdol | mautzone: is it oneplus-enchilada-arm64 | 00:44 |
| Simdol | instead of armv7hl? | 00:44 |
| r0kk3rz | chapter 6, setting up scratchbox2 target | 00:44 |
| mautzone | Output of sdk-assistant list? | 00:44 |
| Simdol | SailfishOS-latest | 00:44 |
| Simdol | ├─SailfishOS-latest-armv7hl | 00:44 |
| Simdol | └─SailfishOS-latest-i486 | 00:44 |
| r0kk3rz | so | 00:44 |
| r0kk3rz | you ignored the box with 'warning' written on it | 00:44 |
| Simdol | oh shit lol | 00:45 |
| Simdol | i just read it | 00:45 |
| r0kk3rz | there you go | 00:45 |
| Simdol | is that really reason why it doesn't build? | 00:45 |
| r0kk3rz | if your thing has the wrong name | 00:45 |
| r0kk3rz | yes | 00:45 |
| Simdol | r0kk3rz: so i assume the naming is done through file name? | 00:46 |
| Simdol | so line oneplus-enchilada-armv7hl.7z | 00:47 |
| Simdol | *like | 00:47 |
| T42 | <edp_17> @r0kk3rz: I have a concern. Some files has some hieroglyph like '# Low Power Range calibration 시에만 enable' but when I see as raw on github I see '# Low Power Range�� start value: 0dBm' | 00:47 |
| T42 | <edp_17> So those chars have converted to ? Will it cause any issues? | 00:47 |
| r0kk3rz | edp_17, what on earth are you doing? | 00:48 |
| T42 | <edp_17> From github TheMuppets/hero-common I am manually saving all files into my vendor/samsung/hero-common | 00:49 |
| r0kk3rz | yeah... dont do that | 00:49 |
| T42 | <edp_17> This is what you said that Driim did the same because there is no way to sync a part of a repo. | 00:49 |
| T42 | <edp_17> Ok then, what else I can do? | 00:50 |
| r0kk3rz | maybe you just need to download the whole thing, it was just a suggestion | 00:50 |
| T42 | <edp_17> In Driim's repo, that I forked, was only this hero-common. So I definetaly don't need all. | 00:51 |
| r0kk3rz | no you dont | 00:51 |
| r0kk3rz | its annoying they put it all in a mega repo, for other vendors they dont do that | 00:52 |
| T42 | <edp_17> Yeah, I agree but they did for this. :( | 00:52 |
| T42 | <edp_17> So, my options: 1. sync the whole, 2. manually download all from hero-common sub folder. 3. ?? | 00:53 |
| r0kk3rz | see what you can google up | 00:54 |
| T42 | <edp_17> Ok, I have updated my local manifest to point to TheMuppets and I do sync. I download them all. While I am googling it is downloading. 😁 | 00:55 |
| T42 | <edp_17> Ops, it seems I need to push some changes first. | 00:57 |
| mautzone | edp_17: You could just do a git clone and copy the files, no need to put them in your manifest if your not resyncing. | 01:01 |
| T42 | <edp_17> @mautzone: Good point, I do this now. Once I pushed my other changes to github. Thanks! | 01:04 |
| mautzone | edp_17: Or you could also downlaod the zip file. | 01:05 |
| T42 | <edp_17> I have started cloning. | 01:05 |
| T42 | <edp_17> Thanks. | 01:05 |
| mautzone | r0kk3rz: Does rf kill service also set up the bt device? Or do i need to di this manually? | 01:06 |
| r0kk3rz | mautzone: it sets it up, thats basically its only job | 01:07 |
| mautzone | r0kk3rz: Then i must have an error in the config file. | 01:12 |
| mautzone | Damn, so late again.... | 01:14 |
| mautzone | r0kk3rz: Thanks for your help! | 01:15 |
| T42 | <edp_17> @mautzone: Yep, this is my problem too. getting tired and so much things to do... | 01:15 |
| mautzone | edp_17: Just take a day off this porting stuff :-P | 01:15 |
| T42 | <edp_17> @mautzone: I already have taken 2. Now I feel I shouldn't but I had to. :) | 01:17 |
| T42 | <edp_17> @mautzone: Good night. | 01:17 |
| Simdol | Does anyone know how to solve this dependency issue? http://dpaste.com/1R0Z48R | 01:20 |
| r0kk3rz | Simdol: you need this https://github.com/mer-hybris/libncicore | 01:22 |
| Simdol | r0kk3rz: Okay, how do I install it? | 01:23 |
| Simdol | just make && make install? | 01:23 |
| r0kk3rz | theres a build_packages.sh command | 01:23 |
| r0kk3rz | https://github.com/mer-hybris/droid-hal-device/blob/master/helpers/build_packages.sh#L42 | 01:24 |
| T42 | <edp_17> @r0kk3rz: I have cloned the vendor/samsung from the TheMuppets, then deleted everything from my local vendor/samsung/hero-common, then copied everything here from themuppets/hero-common, then pushed everything to github. The only difference between TheMuppets and Driim's repo is this: https://github.com/edp17/proprietary_vendor_samsung | 01:26 |
| T42 | /commit/45a6cb1f0954f6f86d213fd6d8395d174d0466a6 | 01:26 |
| T42 | <edp_17> Is it possible that this is the reason why the gui isn't starting? | 01:27 |
| T42 | <edp_17> This is wifi related but my build is not really yet there. | 01:27 |
| r0kk3rz | doubtful | 01:28 |
| T42 | <edp_17> I though so. | 01:28 |
| T42 | <edp_17> Then I am stuck again. | 01:28 |
| mautzone | Good nght everyone. | 01:29 |
| T42 | <edp_17> @mautzone: Sleep well. | 01:30 |
| Simdol | r0kk3rz: sorry, still can't figure it out how to install | 01:35 |
| Simdol | build_packages.sh -m libncicore is not the way to do it, i assume? | 01:35 |
| r0kk3rz | no | 01:36 |
| r0kk3rz | --mw=[github repo url] | 01:36 |
| Simdol | oh you specify the url | 01:37 |
| r0kk3rz | yes | 01:37 |
| T42 | <edp_17> @r0kk3rz: What I spotted in Driim's repo and I do no have is a file: sparse/usr/bin/droid/droid-hal-startup.sh | 01:40 |
| T42 | <edp_17> If I do not have droid-hal-startup.sh in my sparse, is there anywhere else by default what the process put in when building the image? Or need to manually put this into sparse? | 01:43 |
| Simdol | Still have this error: | 01:47 |
| Simdol | error: Failed build dependencies: | 01:47 |
| Simdol | pkgconfig(libnciplugin) is needed by nfcd-binder-plugin-1.1.1-0.armv7hl | 01:47 |
| T42 | <edp_17> Well, comparing the droid-hal-startup.sh what I have on my device and what is in Driim's repo, the difference is that he added a condition for aarch64 or 32 | 01:47 |
| Simdol | pkgconfig(nfcd-plugin) >= 1.0.20 is needed by nfcd-binder-plugin-1.1.1-0.armv7hl | 01:47 |
| T42 | <edp_17> So, it is pointless to use his one. | 01:47 |
| mautzone | Simdol: I think you need to install the package. | 01:48 |
| r0kk3rz | mautzone: it should get pulled in by the requires | 01:50 |
| r0kk3rz | @edp_17 id ignore that one for now | 01:50 |
| T42 | <edp_17> @mautzone: As you are still here rather than counting the sheep on a nice green field 😀 Can you confirm whether my hammerhead build was good for OBS, please? | 01:50 |
| T42 | <edp_17> @r0kk3rz: Thanks, I have ignored. | 01:51 |
| r0kk3rz | https://github.com/mer-hybris/libnciplugin | 01:51 |
| mautzone | edp_17: Could you update without problems? | 01:51 |
| T42 | <edp_17> @mautzone: Yes I did. I thought mentioned this on maemo. | 01:52 |
| T42 | <edp_17> I have checked the files in your test (or dev I don't remember) repo and seen you were using those files I sent you. (based on the timestamps) | 01:53 |
| Simdol | r0kk3rz: for the life of me, I can't find nfcd-plugin. I can only find nfcd-binder-plugin, which won't install. | 01:54 |
| Simdol | Anyone know where to find nfcd-plugin? | 01:59 |
| r0kk3rz | maybe you need this as well https://git.sailfishos.org/mer-core/nfcd/tree/master | 01:59 |
| mautzone | edp_17: Can you build this? build_packages.sh --mw=https://github.com/mautz-kernel/qt5-qpa-hwcomposer-plugin | 02:02 |
| Simdol | r0kk3rz: http://dpaste.com/3HZGHF9 | 02:04 |
| T42 | <edp_17> @mautzone: Ok, I am trying this out now. | 02:04 |
| T42 | <edp_17> Wait a second! Should this qt5-qpa-hwcomposer-plugin be somewhere in droid-local-repo? | 02:07 |
| T42 | <edp_17> Newermind, I found it. | 02:08 |
| mautzone | edp_17:Yes, but this version is 29 commits behind, which shoukd be the version driim has used. | 02:11 |
| T42 | <edp_17> @mautzone: 👍 Thank you! One day please teach me how to build these old packages. 😀 | 02:12 |
| T42 | <edp_17> I am running the mic now. Fingers crossed! | 02:13 |
| mautzone | edp_17: Did the package built? | 02:14 |
| T42 | <edp_17> Yes, without any issues. | 02:14 |
| mautzone | Maybe it is enough to install this package on your device. | 02:15 |
| T42 | <edp_17> I have already wiped. :( | 02:15 |
| mautzone | :-D | 02:15 |
| T42 | <edp_17> I know, I know. :) | 02:16 |
| T42 | <edp_17> You should set your nick. It would be easier to respond. :P | 02:17 |
| mautzone | ? | 02:17 |
| T42 | <edp_17> When I enter @ and start typing maut... it doesn't bring up your nick. | 02:18 |
| T42 | <edp_17> If you set this as username, it will do that. edp17 was booked so I became edp_17. | 02:18 |
| mautzone | This is my username. | 02:19 |
| mautzone | Maybe it's due to this Telegram transport or whatever it is. | 02:19 |
| T42 | <edp_17> Oh, then I am sorry. Are you using a telegram client? | 02:19 |
| T42 | <edp_17> Yes, I just guessed. I am on the web version. | 02:20 |
| T42 | <edp_17> By the way, why are your trying to use bluez5 instead of bluez4? | 02:20 |
| mautzone | Nope, Xchat. | 02:20 |
| mautzone | Yes, i took the bluez5 route. | 02:21 |
| T42 | <edp_17> What's the benefit against bluez4? | 02:21 |
| mautzone | I hope it will give better quality during calls. | 02:22 |
| T42 | <edp_17> I see. | 02:23 |
| r0kk3rz | we're on irc, not telespam | 02:23 |
| r0kk3rz | so it wont autocomplete us | 02:24 |
| mautzone | r0kk3rz: I know, but isn't T42 a telegram transport? | 02:24 |
| r0kk3rz | it is | 02:24 |
| mautzone | r0kk3rz: After several tries i was able to pair the Nexus5 and my XA2 over bluetooth...so, i think the backported driver is basically working | 02:26 |
| T42 | <edp_17> I have tried to ask questions on irc then I gave up. I couldn't send a single message, then when I registered my user, no one really responded. I am using the telegram web client and it is much much better. for me at least. | 02:26 |
| T42 | <edp_17> @mautzone: Whoohoo. Well done! | 02:26 |
| mautzone | edp_17: Nah, this took about 5 minutes :-D | 02:27 |
| r0kk3rz | well its bluetooth, sorta working and actually working can be different | 02:27 |
| r0kk3rz | most people want bt headphones to work at least | 02:27 |
| mautzone | r0kk3rz: I know why i never use it... | 02:28 |
| mautzone | r0kk3rz: First i'll try to get the rfkill service manage the bluetooth device. | 02:29 |
| r0kk3rz | getting things to pair is usually ok, but then theres stuff about profile selection, various middleware integrations .etc | 02:29 |
| T42 | <edp_17> Speaking about that, on my hammerhead port the sound is good when I play music and connected to a bluetooth speaker or radio transmitter or even headset but for calls ... not so great. What can be the reason behind this? | 02:29 |
| r0kk3rz | headset profile is different from a2dp profile | 02:30 |
| T42 | <edp_17> I understand that but when I was connected to the radio transmitter playing the music was great, then I made a phone call and that sounded different. Did it use a different protocol or something for audio calls? | 02:32 |
| r0kk3rz | yes | 02:33 |
| T42 | <edp_17> @mautzone: Wish me luck I am installing the image with your qt5-qpa-hwcomposer-plugin. | 02:33 |
| r0kk3rz | depending on which profile it ends up selecting | 02:33 |
| T42 | <edp_17> @r0kk3rz: Thanks. That explains then. | 02:34 |
| r0kk3rz | i think theres a few generations for call hardware, so it might select an ancient one | 02:34 |
| r0kk3rz | with tight bandwidth restrictions | 02:34 |
| T42 | <edp_17> @0kk3rz: Can I somehow set that profile on the device? | 02:34 |
| T42 | <edp_17> Try which sounds better? | 02:35 |
| r0kk3rz | you should first probe it to see what profiles the device says it supports | 02:35 |
| r0kk3rz | and which one it ends up using | 02:35 |
| T42 | <edp_17> @r0kk3rz: Ok, thanks I may come back to this later if I have something is working. | 02:36 |
| T42 | <edp_17> I booted up and the led light up then stopped. Still no gui. | 02:37 |
| T42 | <edp_17> But after a time the samsung logo is dimmed and double tapping on the screen brighten it up again. | 02:38 |
| T42 | <edp_17> And the 'EGL_PLATFORM=hwcomposer /system/xbin/strace /usr/lib/qt5/bin/qmlscene -platform hwcomposer main.qml' gives me the same 'Segmentation fault' | 02:39 |
| T42 | <edp_17> What is gralloc? | 02:39 |
| T42 | <edp_17> 'SIGSEGV {si_signo=SIGSEGV, si_code=SEGV_MAPERR, si_addr=0}' | 02:40 |
| r0kk3rz | https://source.android.com/devices/graphics/arch-bq-gralloc | 02:40 |
| T42 | <edp_17> Thanks. | 02:42 |
| T42 | <edp_17> Wovwovwov, Now the 'systemctl --user restart lipstick' gives me 'Job for lipstick.service failed because a fatal signal was delivered to the control process. See "systemctl status lipstick.service" and "journalctl -xe" for details. | 02:43 |
| T42 | <edp_17> ' | 02:43 |
| T42 | <edp_17> And the journalctl -xe: https://pastebin.com/MqnfnLqb | 02:44 |
| T42 | <edp_17> I don't see any obvious wrong in there. | 02:45 |
| T42 | <edp_17> The 'systemctl status lipstick.service' gives me: 'lipstick.service | 02:46 |
| T42 | <edp_17> Loaded: not-found (Reason: No such file or directory) | 02:46 |
| T42 | <edp_17> Active: inactive (dead) | 02:46 |
| T42 | <edp_17> ' | 02:46 |
| r0kk3rz | huh? the issue hasnt changed | 02:48 |
| T42 | <edp_17> Now the error message is different when I try to start lipstick. | 02:49 |
| r0kk3rz | run test_hwcomposer | 02:50 |
| T42 | <edp_17> Ok, and here is the relevant (I think) part from journal: https://pastebin.com/qvuRNFe6 | 02:51 |
| T42 | <edp_17> 'EGL_PLATFORM=hwcomposer /system/xbin/strace /usr/lib/qt5/bin/qmlscene -platform hwcomposer main.qml': https://pastebin.com/10Cieay9 | 02:52 |
| r0kk3rz | is it still segfaulting? | 02:53 |
| T42 | <edp_17> 'EGL_PLATFORM=hwcomposer /system/xbin/strace test_hwcomposer': https://pastebin.com/Uv4EPBZ3 | 02:54 |
| T42 | <edp_17> Yes, it still segfaulting. :( | 02:54 |
| T42 | <edp_17> The same message: 'EGL_PLATFORM=hwcomposer /system/xbin/strace test_hwcomposer' | 02:55 |
| T42 | <edp_17> what the heck is this libgui.so? | 02:55 |
| T42 | <edp_17> and the RTLD_NOLOAD? | 02:55 |
| T42 | <edp_17> I found this: https://together.jolla.com/question/211500/how-to-resolve-an-overlay-error-in-release-31011/ | 02:56 |
| T42 | <edp_17> Maybe this is because I built and installed the droid-camres and gstreamer packages? | 03:02 |
| T42 | <edp_17> Okay, I am out. Good night. | 03:13 |
| T42 | <Evans %lastname%> (Photo, 890x976) https://irc.thaodan.de/.imgstore/aly6lrxn0G.png 联系方式 @fafalea | 07:55 |
| masya_ | I'm stucked at vendor logo after installing SFOS. Trying to get logs - https://pastebin.com/e7Whu7eC . LOS-16.0, Xiaomi Mi A2 (jasmine_sprout), A/B partition scheme. | 11:00 |
| masya_ | From LOS get mountpoints https://pastebin.com/6rukMejZ | 11:02 |
| masya_ | Fixup-mountpoints https://pastebin.com/z6hDn3ed | 11:03 |
| KALUBE | masya_: for treble devices you need to use the _a slot partition but remove the _a suffix like this: https://github.com/sailfish-oneplus6/hybris-boot/blob/fbd90f5fad090762adf140a58f68f9d4f25cc5fe/fixup-mountpoints#L812 | 11:48 |
| masya_ | Ok, will try this. Thanks | 11:49 |
| deathmist | masya_: could you link your kernel repo, you probably have a bad commit which will make it so your device never loads hybris-boot ramdisk thanks to a commit. you cannot get any useful logs from SFOS chroot in TWRP | 11:51 |
| deathmist | any reason why buteo-mtp.service would require dev-mtp.mount to be NOT masked to even launch? that mount unit just fails for me since I use configfs instead of functionfs & I just found this https://git.sailfishos.org/mer-core/buteo-mtp/blob/master/rpm/buteo-mtp-qt5.spec#L121 which doesn't explain much (to me anyway) | 13:24 |
| deathmist | meh masking the unit causes even more spam, probably better to implement better configfs support to buteo-mtp at some point so the unit is not obligatory | 13:34 |
| deathmist | spiiroin: you around? I've added sailfish_hw_key_enabled=true dconf thing & https://git.sailfishos.org/mer-core/mce/blob/master/inifiles/buttonbacklight.ini both in fresh images, yet I have to manually invoke "echo 1 > /sys/class/leds/button-backlight/brightness" after boots to enable backlight on physical keys (back/recents) | 14:58 |
| deathmist | output of "mcetool" https://pastebin.com/raw/5kLuEg90 "Button backlight" is reported as disabled | 14:59 |
| deathmist | I've also enabled ButtonBacklight in /usr/share/csd/settings.d/hw-settings.ini of course https://git.io/Jedml | 15:02 |
| spiiroin | deathmist: have to go now, but I'll check what you wrote later on ... | 15:04 |
| deathmist | alright :) | 15:04 |
| vknecht | how could I "emulate" on property:init.svc.surfaceflinger=running from an rc file, so as to start hwcomposer service only after minisf is running ? | 15:04 |
| deathmist | vknecht: I'd guess something like "setprop init.svc.surfaceflinger running" under an "on boot" section in some init.rc file | 15:06 |
| deathmist | oh after, ehm I'll try to look into it xD | 15:07 |
| deathmist | wonder if you can setprop something from a "service" definition in an android init rc file, because you could just add it under that setprop under minisf service in /usr/libexec/droid-hybris/system/etc/init/servicemanager.rc somehow | 15:15 |
| deathmist | s/under that setprop// | 15:18 |
| * vknecht tries a naive "on boot" "sleep 5" "start vendor.hwcomposer-2-1 | 15:33 | |
| vknecht | bingo :-) | 15:34 |
| vknecht | allows lang selection screen, but reboots after sfos logo screen and some delay | 15:36 |
| vknecht | guess have to add some explicit onrestart wrt. minisf, or something | 15:37 |
| vknecht | https://framapic.org/wUMyoCO4hV9Q/hEKVIssXjYWZ.png | 15:57 |
| deathmist | nice ^^ | 16:03 |
| deathmist | mal: I assume I can disable building on the 3.2.0 repo somehow so it doesn't start building stuff always for every single version I have added on testing? https://build.sailfishos.org/project/show/nemo:testing:hw:oneplus:cheeseburger | 16:45 |
| mal | deathmist: https://build.merproject.org/project/show/nemo:testing:hw:fairphone:fp2-sibon check that for example | 17:10 |
| mal | deathmist: the repositories subpage has toggle for each target | 17:11 |
| mal | deathmist: the build flag part in there | 17:11 |
| deathmist | ah very nice, got it :D | 17:13 |
| deathmist | ayyy OTA using sfos-upgrade went perfectly, now on 3.2.1 from 3.2.0 :) | 18:04 |
| mal | it seemed like a simple update | 18:09 |
| deathmist | yeah I read about it, mostly bug fixing but that's welcome too especially if they're annoying ones like notif/ticker bar breaking when trying to uninstall apps etc | 18:10 |
| spiiroin | deathmist: have you added button backlight to hw settings? i.e. you should have somewhere in /usr/share/csd/settings.d/*.ini -> [features] ButtonBacklight=1 | 18:22 |
| deathmist | yep, I linked it before | 18:22 |
| spiiroin | oh, mm, yes ;-) | 18:23 |
| spiiroin | that should enable csd test, does it do something sane? | 18:23 |
| spiiroin | .. pastebin links do not seem to work. server outage? | 18:24 |
| deathmist | spiiroin: damn seems like it, here's fresh mcetool output from 3.2.1 where behavior is still identical https://dpaste.org/Jd1h/raw | 18:28 |
| mal | deathmist: what problem are you trying to fix | 18:29 |
| deathmist | mal: backlight for physical buttons not working without echo 1 to sysfs manually http://www.merproject.org/logs/%23sailfishos-porters/%23sailfishos-porters.2019-12-20.log.html#t2019-12-20T14:58:03 | 18:30 |
| mal | hmm | 18:30 |
| mal | deathmist: can you give a link to your repo | 18:30 |
| spiiroin | deathmist: have you tried that csd test? it should bypass all the lipstick related things | 18:31 |
| deathmist | droid-config here, others under same "organization" https://github.com/sailfishos-oneplus5/droid-config-cheeseburger | 18:31 |
| deathmist | I've never ran a csd test before, how would I do that? | 18:31 |
| spiiroin | deathmist: settings -> about product -> tap on build number 5 times | 18:32 |
| mal | deathmist: I always check the hardware csd tests before a new release | 18:33 |
| spiiroin | deathmist: buteo mtp needs to have those device nodes in place (end point config = writing to control ep). that symlink just makes mtp device node mounting part of local-fs target | 18:38 |
| spiiroin | at least earlier one needed to add some functionfs patches to kernel. afaik configfs is just another way to take functionfs features in use | 18:39 |
| spiiroin | but mal / sledges might actually know more about how those kernel / configfs setup things go | 18:40 |
| deathmist | spiiroin: I meant that MTP works perfectly fine, that units just seemingly fails because I don't have functionfs stuff configured, but instead use configfs for it https://github.com/sailfishos-oneplus5/droid-config-cheeseburger/blob/master/sparse/etc/usb-moded/10-setup-configfs.ini | 18:41 |
| deathmist | either way csd test for button backlight passes, it flashes | 18:41 |
| mal | deathmist: strange | 18:41 |
| spiiroin | deathmist: you have added also the android home/menu/back keys to hwconfig? they should show up in keys csd test | 18:46 |
| spiiroin | ... do you get key events from them? with evdev_trace i.e. do they act as keys or just and aread of touchscreen? | 18:46 |
| masya_ | deathmist I'm using this kernel repo https://github.com/LineageOS/android_kernel_xiaomi_sdm660 | 18:47 |
| T42 | <elros34> masya: there is no 16.0 branch in this kernel | 18:49 |
| deathmist | spiiroin: probably not, they do not show up in key tests, only vol +/-. I never bothered to dig deeper as the back and recents (I remapped recents to back too in synaptics driver xD) and home key all functioned out of the box, here's evdev_trace tho https://dpaste.org/CCf3 | 18:49 |
| deathmist | here's how I did the remap on kernel https://git.io/JedZ8 very ugly yes I know, but it works exactly how I want it to work ^^ | 18:52 |
| spiiroin | deathmist: tail end of /usr/share/lipstick-jolla-home-qt5/compositor.qml -> you need to have Key_HomePage in hw settings for lipstick to handle those android keys | 19:08 |
| spiiroin | "ssu-sysinfo --has-hw-key=Key_HomePage && echo YES || echo NO" -> hw settings ok when you get YES from that | 19:09 |
| spiiroin | [Keys] Keys=0x01000072, 0x01000070, 0x010000dc, 0x01000090, 0x01000061 -> would be vol keys + menu/home/back | 19:10 |
| piggz | spiiroin: oh neat, thats something i can do on mido too then | 19:21 |
| piggz | spiiroin: oh, thats awesome, why did i not know that was possible | 19:27 |
| deathmist | spiiroin: thank you! that made it work as intended :D no need for this stupid garbo https://git.io/JedZN then that I've been doing for a very loooooong time | 19:33 |
| deathmist | it also behaved in a weird way where the LEDs started blinking occasionally on suspend | 19:35 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
