| T42 | <edp_17> Almost all of them. Except if I run them from fingerterm. In that case they do not crash. | 00:04 |
|---|---|---|
| r0kk3rz | ah so its an invoker issue | 00:04 |
| T42 | <edp_17> Yeah, this is what Elros also said. | 00:05 |
| r0kk3rz | its an old exynos device right? | 00:05 |
| T42 | <edp_17> yes, yes. still. 😄 | 00:08 |
| r0kk3rz | yeah ok, probably something libc or libhybris then | 00:10 |
| T42 | <edp_17> Do you mean glibc? | 00:11 |
| r0kk3rz | you could strace it to get more info | 00:13 |
| r0kk3rz | just make sure you run it through invoker | 00:13 |
| T42 | <edp_17> How can I run strace if I am not calling from command line? | 00:14 |
| r0kk3rz | well you are | 00:14 |
| T42 | <edp_17> So I need to call it from command line with strace but have to make sure it is using invoker. | 00:16 |
| r0kk3rz | yeah | 00:16 |
| r0kk3rz | so, strace invoker harbour-blah | 00:17 |
| T42 | <edp_17> It requires to define the type of app. Is it qt5 or qtquick2 or silica-qt5 or generic? | 00:20 |
| r0kk3rz | depends on the app | 00:21 |
| r0kk3rz | most are silica-qt5 | 00:21 |
| T42 | <edp_17> Thanks. | 00:21 |
| T42 | <edp_17> I have ran but I don't know how to transfer the log from the console to somewhere you can see. | 00:25 |
| r0kk3rz | lol | 00:25 |
| r0kk3rz | most people just copy paste, or you can dump it to a file | 00:26 |
| T42 | <edp_17> Yep. Lol. I used the pipe but it didn't generate log file. | 00:26 |
| T42 | <edp_17> It ran on the device from fingerterm, not on pc, therefore I cannot copy/paste. | 00:26 |
| r0kk3rz | use ssh | 00:27 |
| r0kk3rz | poking around with fingerterm is way too cumbersome | 00:27 |
| T42 | <edp_17> Are you saying if I login with ssh and run that command that works? I did that but it outputted a different message. (I guess because this app cannot be launched from pc via ssh.) | 00:29 |
| T42 | <edp_17> That's why I ran it on the device from fingerterm. | 00:29 |
| r0kk3rz | you can | 00:29 |
| r0kk3rz | so long as you are user nemo | 00:29 |
| r0kk3rz | i dont mean telnet, use ssh | 00:29 |
| T42 | <edp_17> Yes, I am logged in via ssh as nemo. | 00:30 |
| T42 | <edp_17> Sorry, I just realized that I have done the same things a few times. | 00:32 |
| T42 | <edp_17> Here is the log: https://pastebin.com/vbPX7EZc | 00:32 |
| T42 | <edp_17> Ops, I was playing with the camera and the device suddenly restarted. 😄 | 00:48 |
| T42 | <edp_17> Question. If I rebuild the kernel, do I need to copy over the wifi module crated for with the new kernel, or I can use the prior one? | 00:58 |
| r0kk3rz | try it | 01:01 |
| T42 | <edp_17> Just asking because I changed the kernel for the bluetooth and after I flashed this new kernel there is no wifi. The dhd.ko module is not loaded. | 01:02 |
| r0kk3rz | so, theres your answer | 01:03 |
| T42 | <edp_17> In the kernel out/.../wireless there is no wifi module generated. 😳 | 01:06 |
| T42 | <edp_17> I have found it. It was somewhere else. 😜 | 01:08 |
| T42 | <edp_17> Yep, that was the case. I transferred the new module, loaded with insmod and the wifi started. | 01:12 |
| T42 | <edp_17> The hadk-faq at 10.5 says I need to identify the bluetooth device name with stracing hciattach. | 01:21 |
| T42 | <edp_17> The strace hciattach gives me: "strace: Can't stat 'hciattach': No such file or directory" | 01:22 |
| T42 | <edp_17> What is missing? | 01:22 |
| r0kk3rz | bluez tools | 01:24 |
| r0kk3rz | you wont need to strace it | 01:24 |
| r0kk3rz | because you'll be using bluez4 not 5 | 01:24 |
| r0kk3rz | also, make sure you have bluez4 installed and not 5 | 01:25 |
| T42 | <edp_17> I don't know whether it is using bluez4 or 5 but in /etc I have a bluez5 folder | 01:28 |
| r0kk3rz | check your patterns | 01:29 |
| r0kk3rz | i think we default to bluez5 these days | 01:29 |
| T42 | <edp_17> It is bluez5 there in pattern indeed. | 01:30 |
| T42 | <edp_17> Is it good? I had problem with bluez5 on the hammerhead and had to use bluez4 instead. | 01:30 |
| T42 | <edp_17> The hadk-faq says: "Symlink your firmware file to /etc/firmware." | 01:31 |
| T42 | <edp_17> I don't have firmware in /etc | 01:31 |
| r0kk3rz | you need to use bluez4 | 01:32 |
| r0kk3rz | your kernel is waaaaay too old for bluez5 | 01:32 |
| T42 | <edp_17> Okay, so now I need to rebuild packages and install the libhybris*.rpm? Or something else? | 01:33 |
| r0kk3rz | its easiest to just fix the patterns, build configs, and reflash the new image | 01:36 |
| T42 | <edp_17> If I do a 'rpm/dhd/helpers/build_packages.sh' that will rebuild configs,packages and create the image too. Right? | 01:38 |
| T42 | <edp_17> Just asking because the hadk-faq 10.5 says something more about Build rfkill middleware and add to patterns | 01:39 |
| T42 | <edp_17> rpm/dhd/helpers/build_packages.sh --mw=https://github.com/mer-hybris/bluetooth-rfkill-event --spec=rpm/bluetooth-rfkill-event-hciattach.spec | 01:39 |
| T42 | <edp_17> and add configs: https://github.com/mer-hybris/droid-config-f5121/commit/afa01bdf4bdb8a0d16bbd34996ec7cac34bbbc55 | 01:39 |
| r0kk3rz | yeah | 01:40 |
| T42 | <edp_17> and if I rebuild a new image anyway, I may should include these too. What do you think? | 01:40 |
| T42 | <edp_17> Cool, I am at the 'Symlink your firmware file to /etc/firmware.' step and there is an example from your repo. 😄 | 01:41 |
| T42 | <edp_17> eg. https://github.com/r0kk3rz/droid-config-scorpion_windy/blob/master/sparse/etc/firmware/BCM4350C0.hcd | 01:41 |
| T42 | <edp_17> So, I have found the .hcd file in mer/android/droid/vendor/samsung/superior/proprietary/system/vendor/firmware/BCM4334.hcd | 01:42 |
| r0kk3rz | yes i wrote that faq section | 01:43 |
| T42 | <edp_17> What should I add into /etc/firmware/BCM4334.hcd ? | 01:43 |
| r0kk3rz | that is likely not quite the right file name | 01:44 |
| r0kk3rz | but like it says, you symlink it | 01:44 |
| T42 | <edp_17> I am not convinced the file name is BCM4334.hcd because there are some others too: BCM4334_murata.hcd and BCM4334_semco.hcd | 01:45 |
| T42 | <edp_17> Okay, I see that I need to create a symlink in /etc/firmware/xxx.hcd that points to the real xxx.hcd file. | 01:46 |
| T42 | <edp_17> But in your example the real .hcd file in the /system/etc/ while I don't have that in my tree. | 01:47 |
| r0kk3rz | you will on your device though | 01:47 |
| T42 | <edp_17> I only have found those three BCMxx.hcd files in 'mer/android/droid/vendor/samsung/superior/proprietary/system/vendor/firmware/' but I am not sure which one to use. | 01:48 |
| r0kk3rz | look on your device | 01:48 |
| T42 | <edp_17> On the device there is no firmware in /etc and /system/etc | 01:49 |
| r0kk3rz | it'll be under system somewhere | 01:50 |
| T42 | <edp_17> I have found a firmware in /system/vendor | 01:51 |
| T42 | <edp_17> But there is no BCMxxx.hcd file. | 01:51 |
| T42 | <edp_17> Actually there is no BCMxxx.hcd file on the device anywhere. | 01:51 |
| r0kk3rz | what bt device do you have? | 01:53 |
| r0kk3rz | you probably should've figured that out first | 01:53 |
| T42 | <edp_17> I think broadcomm but not sure. | 01:54 |
| r0kk3rz | then be sure | 01:54 |
| T42 | <edp_17> I think somebody form this forum helped me a while ago and said, this device uses broadcomm | 01:54 |
| r0kk3rz | seems to say firmware is under system/bin | 02:00 |
| r0kk3rz | https://github.com/CyanogenMod/android_device_samsung_galaxys2-common/blob/cm-14.1/bluetooth/vnd_smdk4210.txt | 02:01 |
| T42 | <edp_17> Very handy if you know how to search online. 😁 | 02:02 |
| T42 | <edp_17> Thank you. | 02:02 |
| r0kk3rz | well, if you think about it, nothing magically just happens | 02:02 |
| r0kk3rz | so its all in the android repos somewhere | 02:03 |
| T42 | <edp_17> I have found the three .hcd files in /system/bin: bcm4330B1.hcd, bcm4330B1_murata.hcd and bcm4330B1_semcosh.hcd | 02:03 |
| r0kk3rz | so much for no file anywhere on the device :P | 02:03 |
| r0kk3rz | link the first one | 02:03 |
| T42 | <edp_17> In my defense, I was looking with capital BCM as that was the name on PC. | 02:04 |
| T42 | <edp_17> But I know this is just an excuse. | 02:04 |
| T42 | <edp_17> I linked that into /etc/firmware on both the device and in sparse. In sparse that is a broken link. I guess this is not a problem? | 02:09 |
| r0kk3rz | so long as the path is right, the wont be broken on the device | 02:15 |
| r0kk3rz | now will it? | 02:15 |
| T42 | <edp_17> Yeah, I agree. | 02:16 |
| T42 | <edp_17> I ran the 'rpm/dhd/helpers/build_packages.sh --mw=https://github.com/mer-hybris/bluetooth-rfkill-event --spec=rpm/bluetooth-rfkill-event-hciattach.spec' but is is moaning about bluez5. | 02:17 |
| r0kk3rz | ignore that for now | 02:17 |
| T42 | <edp_17> I already changed the bluez5 to bluez4 in pattern but didn't rebuild packages. I think that's why it is complaining about it. | 02:18 |
| T42 | <edp_17> Okay, I ignore it, then what's next? | 02:18 |
| r0kk3rz | continue | 02:18 |
| T42 | <edp_17> To the last step? add configs: https://github.com/mer-hybris/droid-config-f5121/commit/afa01bdf4bdb8a0d16bbd34996ec7cac34bbbc55 | 02:19 |
| r0kk3rz | no, build the image | 02:19 |
| T42 | <edp_17> Okay. | 02:21 |
| T42 | <edp_17> While it is building the image I have a few questions you may know the answers or can advice where start looking. | 02:26 |
| T42 | <edp_17> 1. What can be the reason why cannot send text while it can receive? | 02:27 |
| T42 | <edp_17> 2. Mobile data doesn't work. It is switching on then immediately off, then on again, and so on. | 02:28 |
| T42 | <edp_17> 3. mtp doesnt work (I think this is kernel related but not sure what to add/set/change) | 02:28 |
| T42 | <birdzhang> mobile data does not work on 16.0 base vince and davinci too | 02:30 |
| T42 | <edp_17> 4. sound. It only has two stages: on or off. The volume rocker adjusts the volume bar but not the volume itself. Once I reach 0 the sound is gone, then when I pres the volume rocker up, I have sound but that doesn't increase when I am keep pressing the rocker. | 02:30 |
| T42 | <edp_17> Hi @birdzhang : This is cm12.1 base. | 02:32 |
| T42 | <edp_17> 5. it seems there is still problem with rild or ofono because if I keep the onono service on, the battery draing is huge! Even when the device is on charger, it is rapidly losing battery. | 02:33 |
| r0kk3rz | if its still spamming then yeah | 02:33 |
| T42 | <edp_17> How can I find that out? htop doesn't show high cpu usage. | 02:34 |
| T42 | <edp_17> 6. I mentioned the app crash issue which is liked to the invoker. | 02:35 |
| T42 | <edp_17> 7. This is the last I promise. 😄 There is a strange screen refresh "delay". | 02:36 |
| r0kk3rz | 8. one thing at a time please | 02:36 |
| T42 | <edp_17> I didn't listed these to solve them now. | 02:36 |
| T42 | <edp_17> I just mentioned IF you have a quick idea about any of them. | 02:37 |
| r0kk3rz | sure | 02:37 |
| T42 | <edp_17> I would like to go through them one by one. | 02:37 |
| r0kk3rz | its an old exynos device | 02:37 |
| r0kk3rz | all this is quite expected | 02:37 |
| T42 | <edp_17> I see. Well, how far this already gone is a success I think. | 02:38 |
| T42 | <edp_17> Things do not work and buggy but the system is booting up, there is wifi, sound and many other things. So, I think it can be considered as a successful port. | 02:39 |
| r0kk3rz | yeah if you set your bar that low | 02:40 |
| T42 | <edp_17> 😄 | 02:40 |
| T42 | <edp_17> Nope. That's why I would like to continue until all the missing parts work ok-ish. | 02:41 |
| r0kk3rz | it doesnt sound that usable for much if all the apps crash from the launcher | 02:41 |
| T42 | <edp_17> Yeah, that is a real pain as I need to start everything from command line. | 02:42 |
| T42 | <edp_17> And for a strange reason in this way there is no ambiences in the settings. That page is blank. | 02:42 |
| T42 | <edp_17> Oh, and something else. Because the hybris-boot didn't work, I had to flash the kernel image directly on the kernel partition. This means there is no recovery on the device and if I need to flash a new image, I need to flash the recovery first. 😔 | 02:46 |
| T42 | <edp_17> Do you have idea how can I combine a kernel with a recovery to workaround this issue? | 02:47 |
| r0kk3rz | the boot.img should work | 03:20 |
| r0kk3rz | if it doesnt, you should figure out why | 03:20 |
| T42 | <elros34> @edp_17 7: do you have high cpu usage/spam in udevadm monitor when you touch screen? | 09:02 |
| T42 | <edp_17> @r0kk3rz: The boot.img didn't work because the device needs a different format (Linux kernel ARM boot executable zImage (little-endian)). That's why I flashed the kernel directly to the kernel partition. But this way I cannot boot into recovery. | 11:11 |
| T42 | <edp_17> @elros34 : I will try this once got time and report. Thank you. | 11:12 |
| r0kk3rz | so format it properly | 11:13 |
| T42 | <edp_17> I wasn't aware that it possible. How? | 11:38 |
| T42 | <elros34> image format you flash to KERNEL partition doesn't really matter. Both hybris-boot.img and your zImage use kernel + hybris-boot's initramfs | 11:42 |
| T42 | <edp_17> @elros34 : I just said what I was told. I needed to change the initramfs path (I don't remember where) then I was advised to flash the file called kernel from the same folder where the hybris-boot.img created an then the device started booting up. | 11:56 |
| T42 | <elros34> Check how it's done in android for your device: https://github.com/edp17/android_kernel_samsung_smdk4412/blob/cm-12.1/usr/galaxys2_initramfs_files/init It loads recovery.cpio or boot.cpio based on bootmode. You replaced whole original initramfs with hybris-boot but you could try to replace only boot.cpio archive with hybris-boot inst | 12:07 |
| T42 | ead | 12:07 |
| T42 | <edp_17> @elros34 : I will try, thank you. | 12:38 |
| T42 | <alexisrev> I started earning the day I took the courage to invest in Bitcoin with Amos Bauer , now I'm glad I took that courage. I just got $5500 profits sent directly to my PayPal account. | 13:59 |
| T42 | <alexisrev> https://t.me/joinchat/AAAAAEeKU6RLm8meqJJqtw | 13:59 |
| T42 | <polanobi> Please Add Bot Rose | 15:37 |
| T42 | <edp_17> Hi guys, | 20:39 |
| T42 | <edp_17> I am trying make the bluetooth working on my S2 port. | 20:40 |
| T42 | <edp_17> I have followed the hadk-faq 10.5, and enabled the bt driver in kernel, then changed to bluez4 in the pattern. | 20:41 |
| T42 | <edp_17> I created a new image and flashed on the device, but it stuck on the samsung logo. | 20:41 |
| T42 | <edp_17> I checked /system is mounted. | 20:41 |
| T42 | <edp_17> the droid-hal-device.conf is ok, it uses EGL_PLATFORM=fbdev. | 20:43 |
| T42 | <edp_17> I created a logcat: https://pastebin.com/AnV2t8eQ | 20:44 |
| T42 | <edp_17> When I try to start lipstick: '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.' | 20:44 |
| T42 | <edp_17> systemctl status lipstick.service: Loaded: not-found (Reason: No such file or directory) , Active: inactive (dead) | 20:46 |
| T42 | <edp_17> journalctl -xe: https://pastebin.com/r3bRJ8u4 | 20:47 |
| T42 | <edp_17> I have found it! It's my bad. I entered fdbev instead of fbdev. Lol. | 20:53 |
| T42 | <edp_17> Now I have a new problem. I cannot go through the initial process because the app crashes. (The same issue as the earlier app crash issue but this time it is worse as cannot setup the device.) | 21:18 |
| T42 | <edp_17> I have tried to restart with 'jolla-startupwizard' but got 'Segmentation fault' | 21:21 |
| T42 | <edp_17> Okay, I have managed to go through the initial setup. I needed to launch it from PC (ssh terminal) with command: invoker --type=qt5 jolla-startupwizard | 22:22 |
| T42 | <edp_17> If I start an app with invoker --type=qt5, it doesn't crash! | 22:22 |
| T42 | <edp_17> Not elegantly but I have solved the app crash issue. I have removed the word 'silica' from '--type=silica-qt5' in all /usr/share/applications/*.desktop file. (Probably later I need a better solution but now this is sufficient.) | 23:21 |
| T42 | <edp_17> @elros34 : You asked about high cpu usage in udevadm monitor when I touch the screen. I don't have that udevadm monitor, but in htop I see high cpu usage of lipstick when I touch the screen. | 23:36 |
| T42 | <elros34> install it. I guess as a workaround you can remove mapplauncherd-booster-silica-qt5 instead changing every desktop file manually | 23:37 |
| T42 | <edp_17> The "/usr/bin/lipstick -plugin evdevtouch:/dev/input/event2 ...." usually uses 1%-6% but when I touch the screen (and keep touching) the cpu usage goes up to 50%-70%. Is it a problem? | 23:38 |
| T42 | <edp_17> @elros34 : I tried to install but zypper se didn't find it. | 23:38 |
| T42 | <elros34> right it's provided by systemd so its already installed | 23:39 |
| T42 | <edp_17> It might already be installed but there is no udevadm command. | 23:41 |
| T42 | <edp_17> About the other one (mapplauncherd-booster-silica-qt5), if I want to remove it, it gives me a huge list of packages which will be removed with this mapplauncherd-booster-silica-qt5. | 23:41 |
| T42 | <edp_17> Plus the mapplauncherd-booster-silica-qt5 is commented out in pattern. | 23:42 |
| T42 | <elros34> rpm -ql systemd | grep udevadm | 23:43 |
| T42 | <edp_17> Thanks, it is there indeed. | 23:45 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!