Mister_Magister | mal: https://github.com/mlehtima/backports-bluetooth can you remember what's this all about? | 03:48 |
---|---|---|
Mister_Magister | i have it in my moto x 2014/victara port and i don't remember why | 03:48 |
T42 | <rodrisola> Mal: lineage os for my device not support ext4... And think about of bootloop | 04:01 |
Mister_Magister | you can add ext4 support in kernel | 04:02 |
Mister_Magister | https://wiki.merproject.org/wiki/Adaptations/faq-hadk#Using_backported_Bluetooth_drivers_in_3.4_kernel_for_devices_with_Qualcomm_bluetooth_chip_using_hci_smd_driver ahh | 04:31 |
Mister_Magister | tbh this faq should be linked in topid | 04:34 |
T42 | <rodrisola> In defconfig? (re @SailfishFreenodeIRCBridgeBot: <Mister_Magister>you...) | 04:37 |
Mister_Magister | yes | 04:38 |
T42 | <foxclore> mal: @elros34 regarding yesterday's issue: I did "make clean" then rebuilt droidmedia, but issue still remains. There is no libdroidmedia in out ( | 07:25 |
Mister_Magister | yeah xperias had some fix for the wlan because wlan was blocking the sleep and i had it applied in my 5z previously but i no longer have and my phone never sleeps | 08:45 |
Mister_Magister | mal: do you remember it maybe? | 08:45 |
mal | Mister_Magister: well there have been some suspend fixed, one if the connman suspend plugin, then in some new kernels enabling auto suspend in kernel was needed etc | 08:46 |
Mister_Magister | mal: i think its the first one | 08:47 |
Mister_Magister | my kernel is 4.9 and only the connman plugin rings some bell | 08:47 |
mal | Mister_Magister: the bluetooth backports is to use bluez5 on old kernel which doesn't have support for it | 08:47 |
Mister_Magister | mhm found it | 08:47 |
mal | Mister_Magister: connman plugin is probably in adaptation-common | 08:48 |
Mister_Magister | sailfish-connman-plugin-suspend? | 08:48 |
mal | yes | 08:49 |
Mister_Magister | thanks | 08:49 |
Mister_Magister | will add it to patterns | 08:49 |
Mister_Magister | lets see if it suspends now | 08:49 |
Mister_Magister | hmm its second device i'm OTA'ing to 4.2 and 4.3 and kernel doesn't get flashed | 09:04 |
Mister_Magister | for whatever reason | 09:04 |
Mister_Magister | seems like its sleeping, thanks mal <3 | 09:06 |
Mister_Magister | mal: trying to boot 4.3 on 3.4 kernel and i only see this in logcat https://paste.opensuse.org/00e23933 i'm not sure if i didn't miss some patches in kernel or something | 09:09 |
Mister_Magister | yeah dmesg confirms that its just that | 09:09 |
Mister_Magister | idk what i broke :/ | 09:12 |
Mister_Magister | looking into kernel config changes | 09:16 |
Mister_Magister | can't find anything obvious | 09:18 |
Mister_Magister | what the hell did you do magister | 09:23 |
Mister_Magister | mal: do you remember some patch that was changing fget to fget_raw | 09:25 |
mal | yes, there was some fix for that needed in some kernel versions | 09:26 |
Mister_Magister | mal: yeah i do know that but i don't know the patch, my g2 had it only changed in one place | 09:27 |
Mister_Magister | they're both 3.4 kernel | 09:28 |
Mister_Magister | i suspect i need to change it in some more places but the patch would be useful | 09:28 |
Mister_Magister | as a reference | 09:28 |
Mister_Magister | can't find anything in logs about it tho | 09:28 |
mal | that depends a bit on the exact kernel version afaik | 09:29 |
Mister_Magister | do i just need to replace all fget with fget_raw? | 09:30 |
Mister_Magister | nah doesn't seem so | 09:30 |
mal | fp2 kernel has something related to that, the hybris-21.05.0-rel branch | 09:30 |
Mister_Magister | but maybe i'm searching in wrong place, do you have any issue about adsprpc issue? | 09:31 |
Mister_Magister | any clue* | 09:31 |
Mister_Magister | its getting permission issue :/ | 09:32 |
Mister_Magister | i don't think that fget is issue for the adspprpc its sailjail fix | 09:34 |
Mister_Magister | so i'm looking in wrong place | 09:35 |
Mister_Magister | idk what the hell changed | 09:36 |
Mister_Magister | i'm so confused | 09:38 |
Mister_Magister | mal: here's strace, it cannot open the device for some reason https://paste.opensuse.org/6278dbc3 | 09:44 |
Mister_Magister | lets do sanity check and boot older kernel even tho it doesn't have fixes for sailjail | 09:46 |
mal | are the permissions of that device node correct? | 09:48 |
Mister_Magister | i thought they're weird but when i changed them, nothing changed | 09:50 |
Mister_Magister | on old kernel same issue | 09:51 |
Mister_Magister | so it isn't kernel then | 09:52 |
Mister_Magister | even when, as root, i do cat /dev/adsprpc-smd i get permission denied | 09:54 |
Mister_Magister | something's not right | 09:54 |
Mister_Magister | its not selinux as selinux is disabled | 09:55 |
Mister_Magister | while doing cat /dev/adsprpc-smd on other devices it returns invalid argument which would be correct | 09:59 |
Mister_Magister | even on 4.3 | 09:59 |
T42 | <elros34> @foxclore but what was the output? It was al about output, after make droidmedia, you should have printed path to built library like I said | 10:02 |
Mister_Magister | mal: could it be sailjail causing issue? | 10:04 |
mal | Mister_Magister: sailjail is only relevant for UI apps | 10:05 |
Mister_Magister | mal: oh :< any idea then? You're my only hope :< | 10:05 |
T42 | <elros34> have you tired most obvious chmod a+rw /dev/adsprpc-smd? what is the owner of it? | 10:06 |
Mister_Magister | @elros34 i tried chmod 777 and changing owner to system:system, for some reason the owner is media camera while my other device has system:system | 10:07 |
Mister_Magister | and since i'm root permissions shouldn't matter | 10:08 |
Mister_Magister | so there is something else blocking access i thimk | 10:08 |
Mister_Magister | i removed the line that chages it to media:camera | 10:12 |
Mister_Magister | and no change | 10:13 |
T42 | <elros34> I would search in init*rc files for adsp and in kernel. Could be that it needs some initialization or special group when started | 10:16 |
T42 | <foxclore> But there was nothing in output regarding droidmedia, that the thing (re @elros34: @foxclore but what w...) | 10:18 |
T42 | <elros34> its hard to believe that make droidmedia finished without any related output. Btw did you apply hybris patches? | 10:19 |
mal | Mister_Magister: sorry, maybe later today, I need to work | 10:20 |
T42 | <foxclore> Yes (re @elros34: its hard to believe ...) | 10:21 |
Mister_Magister | mal: ahh its okay no worries :) | 10:21 |
T42 | <foxclore> Well I'll paste all logs once I'll get home (re @elros34: its hard to believe ...) | 10:21 |
Mister_Magister | @elros34 at that point i might as well turn the kernel upside down, worst part is that on previous kernel same issue happens so it's related to the fs and not kernel | 10:22 |
Mister_Magister | hmm might as well build zip and diff them | 10:36 |
T42 | <foxclore> @elros34 https://pastebin.com/eG8AUkWG | 13:52 |
T42 | <foxclore> Logs | 13:52 |
T42 | <rodrisola> A question, when creating the zip, of course there is no error, I install it on my phone, when I finish installing of course it does not give me an error, start the phone and seconds it enters fast boot, it does not start. Why do you think? | 14:46 |
mal | @rodrisola which android base? are you sure kernel contains valid init | 15:22 |
T42 | <rodrisola> Mmm 18.1 (re @SailfishFreenodeIRCBridgeBot: <mal>@rodrisola whic...) | 15:39 |
T42 | <rodrisola> Mal: android base 18.1 | 15:40 |
T42 | <rodrisola> Mal: How do I check if you have a valid init? | 15:40 |
mal | mostly two ways, trying to see from build log if it complains about cpio or other way is to break apart hybris-boot.img using bootimgtools or something to see the size of the init | 15:42 |
poetaster | mal: it looks like FF port, like the volla port, have issues with Sailjail and GPS? | 15:44 |
mal | poetaster: ? has anyone checked strace for what it tries to open? | 15:52 |
poetaster | mal: was actually two issues; Starting it from terminal i got an error about the ‘ApplicationName’. | 15:55 |
mal | hmm | 15:55 |
poetaster | mai: I changed it at /usr/share/applications/harbour-dwd.desktop from harbour-dwd to HarbourDWD and now it works as expexted. I | 15:55 |
mal | I'm confused | 15:55 |
poetaster | two issues: one, on FF, nameing seems to be an issue (works elsewhere), two GPS doesn't work. but it also doesn't work on volla. | 15:56 |
poetaster | mal: https://forum.sailfishos.org/t/gps-location-data-not-available-in-sailjailed-app | 15:56 |
poetaster | mal: the FF info I got via private chat. | 15:57 |
Thaodan | poetaster: If there's a path needs to be whitelist that could be an issue | 16:01 |
Thaodan | That's why I said run the app with sailjail --trace | 16:01 |
poetaster | Thaodan: --trace didn't provide any further data. Network was available. | 16:04 |
Thaodan | Ah meh. I thought there will be access denied on some of those vendor libs or paths that's that are only used on volla | 16:05 |
poetaster | Thaodan: I've pinged piggz, but it seems to be the same issue on FF2 (but that's 4.2!) | 16:05 |
Thaodan | We did something similar here: some | 16:06 |
Thaodan | https://github.com/sailfishos/sailjail-permissions/pull/114 | 16:06 |
poetaster | mal: I'll try strace on Jailconfiged app later. | 16:08 |
poetaster | mal: problem is my FF2 is still 3.4 and I'm snowed under. | 16:08 |
T42 | <elros34> @foxclore why do refuse to include command you use/ command prompt? How do I know what you have done: make droimedia or hybris-hal, did you export required env variable and so on | 16:13 |
T42 | <foxclore> make droidmedia (re @elros34: @foxclore why do ref...) | 16:15 |
T42 | <foxclore> The only thing exported was that disable path thingy in hadk-faq | 16:16 |
T42 | <elros34> and that was right after make clean? | 16:16 |
T42 | <foxclore> Yes (re @elros34: and that was right a...) | 16:16 |
T42 | <elros34> do you have PORT_ARCH set? echo prints something? | 16:45 |
T42 | <rodrisola> mal: i installed bootimgtool.. but what extract… kernel, ramdisk? (re @SailfishFreenodeIRCBridgeBot: <mal>mostly two ways...) | 17:29 |
T42 | <elros34> use abootimg -i hybris-boot.img | 17:30 |
piggz | poetaster: so, whats not workin, puremaps? | 17:31 |
poetaster | piggz: no it's just Location in Sailjail context. puremaps is fine. I had just released harbour-dwd with sailjail the other day on chum. | 17:36 |
Mister_Magister | poetaster: ping, you got time now? | 17:39 |
Mister_Magister | sorry i mean mal | 17:39 |
piggz | poetaster: best idea is strace the process while running in jail to see whats going wrong | 17:40 |
mal | Mister_Magister: so what exactly are you trying to fix? | 17:41 |
T42 | <rodrisola> mal: this is my init.rc https://www.toptal.com/developers/hastebin/tedoropaji.bash | 17:41 |
mal | I just meant that if the init is something else than empty then it's fine | 17:42 |
mal | maybe you need the usual revert in kernel for skip init thingy | 17:42 |
mal | something like this https://github.com/mer-hybris/android_kernel_sony_msm/commit/d73196c847c6e25734be44cdcdadb56da5c5d4f1 | 17:42 |
mal | so you need to find the commit in your kernel which adds skip_initramfs option and then revert it | 17:43 |
Mister_Magister | mal: so, i have my moto x 2014, it was on 3.4 so i wanted to update it to 4.3. I've updated packages on obs, updated the kernel for 4.3, updated droidmedia, updated audioflingerglue, the base is 13.0, when i was still on 3.4 i booted the new kernel to confirm fdinfo, everything worked fine, so i updated to 4.3 and now adsprpc is not starting blocking entire droid-hal-init and i can't open /dev/adsprpc-smd which is created by | 17:43 |
Mister_Magister | kernel, but even with the 3.4 kernel i can't open it so something else than kernel is blocking it i think | 17:43 |
mal | what problems does that cause? | 17:44 |
Mister_Magister | em, droid-hal-init is not working so nothing is working | 17:44 |
Mister_Magister | its like if you tried to boot with masked dhi | 17:45 |
Mister_Magister | not going to happen | 17:45 |
Mister_Magister | like, adsprpc is needed for dhi to start | 17:47 |
Mister_Magister | question is why can't i open/access it | 17:47 |
Mister_Magister | like the only thing that changed in this whole equation is dhd package and sfos version | 17:50 |
* Mister_Magister that's quite a lot | 17:50 | |
mal | how did you update to 4.3? directly from old version or did you reflash? | 17:53 |
Mister_Magister | directly | 17:53 |
mal | anything in logs about firmware loading? | 17:53 |
Mister_Magister | ill get you full dmesg asap | 17:54 |
mal | skipping stop releases is not a good idea | 17:54 |
Mister_Magister | mal: oh but isn't like every version after 4.0 a stop release | 17:54 |
mal | yeah, it's been a bit more common recently, there is always some reason for it | 17:55 |
Mister_Magister | i've updated g2 from 3.4 to 4.2 without issues | 17:55 |
Mister_Magister | tho doubt that would cause the problem really | 17:55 |
T42 | <rodrisola> mal: I just noticed an error in my init.rc, My partition data Sda34, in the init is as /dev/sda34, but when making ls -l in /dev/ No partition exists Sd*, if it is not existing in /dev/block | 18:00 |
T42 | <rodrisola> I have to change... Fixmountpoint | 18:00 |
mal | where did you check that? | 18:01 |
mal | I'm a bit confused what is working and what is not on your device | 18:01 |
T42 | <elros34> init.rc is not same as init-scripts you are probably talking. Even with wrong partition you would not boot directly to fastboot mode. Can you show abootimg output? | 18:02 |
mal | @elros34 direct boot to fastboot mode sounds like it needs that skip_initramfs revert so it can boot to sfos kernel | 18:05 |
T42 | <elros34> could be or more likely missing initramfs, that is why I asked about: abootimg -i hybris-boot.img but in vain | 18:06 |
mal | @elros34 I assumed the init shown earlier was from the initramfs | 18:07 |
T42 | <elros34> it could be but it can be also from out/ so we can guess until @rodrisola clarify it | 18:08 |
T42 | <rodrisola> mal: in the init.rc log, it indicates that the data partition is / dev / SDA34, in my fixmountpoin, it is SDA34, but when in TWRP, I execute the command "ls -l / dev / *", there is no SD *, but I run "ls -l / dev / block / *" there if the SD* partitions exist | 18:10 |
T42 | <elros34> again this is not your issue and there is no such a thing as init.rc log | 18:12 |
mal | @rodrisola also comparing what you see in recovery is not always accurate when you compare to other environments | 18:22 |
mal | @rodrisola did you check the initramfs you extracted from kernel, assuming you did that | 18:22 |
T42 | <rodrisola> abootimg -i hybris-boot.img https://www.toptal.com/developers/hastebin/ugocesukok.yaml | 18:27 |
T42 | <elros34> looks good enough to boot, try dmesg -w on host, then flash hybris-boot.img and check whether you get any log at first boot | 18:30 |
T42 | <rodrisola> mal: yes i extract | 18:39 |
T42 | <rodrisola> @elros34 when you mean host, do you mean my ubuntu, or device? | 18:40 |
T42 | <elros34> your pc: ubuntu or whatever you use, it shhould show some usb related messages which can give some hints whether device try to boot android init, hybris init or just directly boot to fastboot | 18:44 |
T42 | <rodrisola> @elros34 https://www.toptal.com/developers/hastebin/asavafasom.yaml | 18:52 |
T42 | <rodrisola> just reboot in fastboot | 18:52 |
T42 | <elros34> which line is 1th boot? | 18:58 |
T42 | <elros34> not sure but: 'New USB device found, idVendor=18d1, idProduct=d00d and Product: Android' looks more like android init than fastboot mode. Have you tired advice from mal about skip_initramfs? | 19:03 |
T42 | <rodrisola> first in twrp mode for flash adb sideload, and reboot | 19:04 |
T42 | <rodrisola> https://www.toptal.com/developers/hastebin/utowaluzok.yaml | 19:04 |
Mister_Magister | mal: https://paste.opensuse.org/8630865 | 19:05 |
T42 | <elros34> @rodrisola why don't you use fastboot flash boot hybris-boot.img? | 19:05 |
Mister_Magister | error while handling import on line '11' in '/init.rc' | 19:07 |
Mister_Magister | HMMM | 19:07 |
Mister_Magister | or more like "/init.rc: 30: invalid command 'setcon'" | 19:17 |
Mister_Magister | mal: ^ i think that breaks init.rc | 19:17 |
Mister_Magister | and it would make sense as its the userspace change | 19:17 |
T42 | <rodrisola> @elros34 Because I had deleted all the system and data, now fastboot flash boot hybris-boot.img, how can I disable the skip_initramfs. I do not find information on the internet, excuse the ignorance | 19:18 |
Mister_Magister | hmmm maybe too new libhybris? | 19:19 |
Mister_Magister | or something | 19:19 |
T42 | <elros34> @rodrisola: mal explained it to you less than 1h ago, you can also find instruction in hadk-hot. I do not understand why you deleted system? You need it to boot lineage and to boot sfos also. | 19:24 |
Mister_Magister | mal: why would setcon not be available? | 19:27 |
Mister_Magister | or is it just warning? | 19:27 |
Mister_Magister | https://paste.opensuse.org/66059670 heres longer dmesg | 19:31 |
T42 | <elros34> maybe this one: "adsp: Failed to locate adsp.mdt" | 19:39 |
Mister_Magister | @elros34 how did i miss it | 19:40 |
Mister_Magister | @elros34 ill investigate it and i'll get back to you on that | 19:40 |
Mister_Magister | i think its valid clue | 19:40 |
Mister_Magister | it's in /firmware/image/adsp.mdt | 19:41 |
T42 | <elros34> maybe it would be easier to notice error if you would disable audit | 19:44 |
Mister_Magister | it is disabled twice | 19:45 |
Mister_Magister | wait | 19:45 |
Mister_Magister | itsnot | 19:45 |
Mister_Magister | but anyway, @elros34 do you maybe know where can i check what locations its checking? | 19:45 |
T42 | <elros34> no but this source match other adsp errors: https://android.googlesource.com/kernel/msm/+/android-5.1.0_r0.6/arch/arm/mach-msm/qdsp6v2/adsp-loader.c | 19:48 |
Mister_Magister | hmmmm | 19:52 |
Mister_Magister | i found the function where it returns failed to locate | 19:54 |
korppi | Mister_Magister: cat /lib/udev/rules.d/* |grep firmware | 20:01 |
Mister_Magister | korppi: nothing | 20:02 |
Mister_Magister | hmmmm | 20:02 |
korppi | so you are missing udev rule | 20:02 |
Mister_Magister | would seem so, hold i got an idea | 20:02 |
korppi | Mister_Magister: SUBSYSTEM=="firmware", ACTION=="add", RUN="/usr/bin/droid/droid-load-firmware.sh" | 20:03 |
Mister_Magister | thanks | 20:03 |
T42 | <elros34> droid-load-firmware always prints firmware loading attempts, so if it fails you should have it in journal | 20:05 |
Mister_Magister | @elros34 the problem is that its not being run | 20:06 |
Mister_Magister | is what korppi is saing | 20:06 |
Mister_Magister | question is why is that udev rule missing | 20:06 |
Mister_Magister | who adds it | 20:06 |
Mister_Magister | droid-configs | 20:07 |
Mister_Magister | yeah last line is missing for whateve reason | 20:07 |
Mister_Magister | its present in the package so why… | 20:09 |
korppi | I had that too... | 20:10 |
Mister_Magister | i know why | 20:10 |
Mister_Magister | I'm such a fucking idiot | 20:10 |
Mister_Magister | i had obs repo set to 3.4 | 20:11 |
Mister_Magister | so it didn't update any device-specific package | 20:11 |
Mister_Magister | how do i recover from this hmmm | 20:12 |
korppi | Ok. Different reason for my device then. | 20:12 |
Mister_Magister | i need internet but the wlan module won't load xd | 20:13 |
Mister_Magister | also jsut adding that line didn't fix it xd | 20:13 |
korppi | Nice. Usb with some firewall magic then... | 20:14 |
Mister_Magister | yeah the firewall magic is the real struggle | 20:15 |
Mister_Magister | back in the day hadk had tips on the usb routing | 20:38 |
Mister_Magister | but it doesn't work anymore | 20:39 |
Mister_Magister | oh it works | 20:40 |
Mister_Magister | we are going places | 20:41 |
Mister_Magister | if it didn't reboot while im refreshing repos that would be ultra helpful | 20:43 |
Mister_Magister | hmm i think i can mask shi | 20:43 |
Mister_Magister | dhi | 20:43 |
Mister_Magister | nothing provides 'gmp-droid' needed by the to be installed patterns-sailfish-device-adaptation-victara-1.1+master.20211213072731.19.ga8d660a-1.24.2.jolla.armv7hl | 20:45 |
Mister_Magister | im missing a repo am i not | 20:45 |
Mister_Magister | yep, adaptation-common is missing | 20:47 |
Mister_Magister | but why, it is present in /usr/share/ssu/repos.d/10-jolla.ini | 20:49 |
Mister_Magister | why nothing provides gmp-droid even with adaptation-common | 20:52 |
Mister_Magister | wasn't gmp-droid supposed to be provided by adaptation-common since 4.3? | 20:55 |
Mister_Magister | on my other 4.3 device it is in adaptation-common | 20:55 |
Mister_Magister | what the heck is going on | 20:55 |
Mister_Magister | maybe just adding repo url wasn't the solution then | 20:56 |
T42 | <elros34> why did you add repo? Didn't we figure out why repo is missing several days ago? | 20:57 |
T42 | <elros34> tip: ssu cache | 20:57 |
Mister_Magister | no i don't think we did, maybe you got wrong person | 20:57 |
Mister_Magister | but yes rm -rf /var/cache/ssu/ ssu ur is the solutin | 20:58 |
T42 | <elros34> :) I remeber it good because I was supprised that bug was ignore even with your report and release was out then | 20:58 |
Mister_Magister | i haven't talked about it in "several days ago" timeline, more like "several months ago" | 20:59 |
Mister_Magister | but ye, it got me thinking about that issue so i went and read logs and fixed it properly | 20:59 |
T42 | <elros34> right it was more than several days | 21:00 |
Mister_Magister | i'm now worried about how users are supposed to upgrade to 4.3 without destroying everything | 21:00 |
Mister_Magister | like me | 21:00 |
Mister_Magister | seems like its not the end of my issues tho | 21:03 |
Mister_Magister | oh we got ui, thats something | 21:03 |
Mister_Magister | nice anything about media just crashes apps | 21:14 |
Mister_Magister | ah i had old gstdroid | 21:15 |
Mister_Magister | hmm vibrations don't work | 21:18 |
T42 | <adampigg> perms (re @SailfishFreenodeIRCBridgeBot: <Mister_Magister>hmm...) | 21:18 |
Mister_Magister | @adampigg elaborate plz | 21:19 |
Mister_Magister | i seen it somewhere but couldn't find it | 21:19 |
Mister_Magister | like i remember doing something about it but i don't remember where | 21:20 |
Mister_Magister | @adampigg ? | 21:23 |
Mister_Magister | https://github.com/VerdandiTeam/android_device_motorola_msm8226-common/commit/a98a2933e60387ab4dc2861004522ba839b47b13 found it | 21:23 |
Mister_Magister | ugh i can't build any zip now | 22:22 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!