| T42 | <k1gen> @elros34 no, I didn't. how do I do what you wrote? | 05:52 |
|---|---|---|
| T42 | <adampigg> PSA connman in 4.5 prioritises ethernet devices, so, if your port exposes a virtual ethernet device and wont autoconnect to wifi or cellular, this is why | 06:31 |
| T42 | <k1gen> @elros34 I ran `cat /proc/<gdb pa pid>/maps` and there's nothing at 7fb5469140: https://vomitb.in/raw/6sAy8oFmGB | 06:45 |
| Crabster | morning | 06:56 |
| T42 | <k1gen> o/ | 07:05 |
| piggz_ | mal: adding "ifb" to the end of the blacklisted devices helps | 08:09 |
| piggz_ | mal: main.conf is in project jolla-common-configuration, which i guess isnt on GH, so I cant PR it | 08:13 |
| Mister_Magister | piggz_: whatsup, how's portin going | 08:18 |
| piggz_ | Mister_Magister: so ... on 4.5, connman prioritises ethernet devices over wlan and cellular. if your port has an ethernet device (volla) then it wont auto-connect to any other network | 08:19 |
| piggz_ | much debugging was needed | 08:19 |
| Mister_Magister | lol | 08:19 |
| Mister_Magister | why would phone have ethernet | 08:19 |
| piggz_ | volla is built with CONFIG_IFB, which is a virtual ethernet device | 08:20 |
| Mister_Magister | but at the same time with same logic, why would phone not have mmcblk0 :) | 08:20 |
| piggz_ | made me wonder if the connman chance was to support some ethernet on a jolla in-vehicle type SoC | 08:20 |
| Mister_Magister | can't u just yeet IFB | 08:22 |
| piggz_ | on the plus side, i did learn a bit more about connman and connmanctl :) | 08:22 |
| Mister_Magister | piggz_: i mean technially you can just connect ethernet on usb | 08:22 |
| Mister_Magister | I would prioritize that over wlan | 08:23 |
| Mister_Magister | piggz_: knowledge gained is always a win | 08:23 |
| piggz_ | Mister_Magister: yeah, in that case it is correct to do that, but probably not ifb devices | 08:23 |
| Mister_Magister | yee | 08:23 |
| piggz_ | other issues is its not easy for porters to override connman conf | 08:23 |
| piggz_ | i do it in pinephone, but its messy | 08:23 |
| T42 | <abranson> IFB looks to be some sort of traffic aggregating QoS device? Is it really needed? | 08:24 |
| Mister_Magister | ye why do you need it in volla in the first place | 08:25 |
| piggz_ | abr: probably not, likely only used on android, but never been an issue for 2 years :) | 08:25 |
| T42 | <abranson> If disabling it in the kernel config works, then it might be better to add that to the kernel-check rather than bother connman with it. no need to have random unused networks interfaces littering up the place. | 08:28 |
| piggz_ | abr: there are 32 interfaces exposed by "ip addr" :) | 08:29 |
| piggz_ | phones are weir dlike that :D | 08:30 |
| T42 | <abranson> 32! i'm surprised you only get trouble from that one then. my x10iii only has 16, and that's with wlan and tethering on | 08:31 |
| piggz_ | https://paste.mozilla.org/SQv1sLrq | 08:31 |
| piggz_ | there are 2 that cause the issue, ifb0 and ifb1 | 08:32 |
| T42 | <abranson> i guess they're called mediatek not networktek | 08:33 |
| piggz_ | main thing was that at least it was figured out | 08:34 |
| T42 | <abranson> yeah really cool. there were a lot of changes between 4.4 and 4.5 so that could have been a real needle in a haystack | 08:35 |
| T42 | <elros34> @k1gen this address will be different for every crash so I hope you have check maps with correct one | 08:49 |
| T42 | <k1gen> @elros34 nothing with new thread address either: | 09:02 |
| T42 | <k1gen> [root@flame defaultuser]# awk -F'[ -]' -v s=7fb5417140 '$1<=s&&s<$2' /proc/8012/maps | 09:02 |
| T42 | <k1gen> 7fb5218000-7fb5418000 rw-p 00000000 00:00 0 | 09:02 |
| T42 | <elros34> ok so that will not help so try hidl_compat, there is small chance that it will change something | 09:04 |
| T42 | <k1gen> where do I start with hidl_compat? | 09:05 |
| T42 | <elros34> IIRC you do not start it (it's library) but bind mount, there should be some pointers in docs. Did you build it yet | 09:07 |
| T42 | <k1gen> no, not yet, searching for it in tg. https://github.com/Halium/android_vendor_halium_hardware, right? | 09:10 |
| T42 | <k1gen> clone to vendor/halium/hardware and make hidl_compat? | 09:11 |
| T42 | <elros34> something like that, check branches so it builds for your android version | 09:12 |
| T42 | <k1gen> well, there's only master, halium-10.0 and halium-12.0. I'm on hybris-18.1 | 09:13 |
| T42 | <elros34> 10 sounds good but you will see | 09:14 |
| T42 | <k1gen> it's building. what should I do with audio.hidl_compat.default.so then? | 09:30 |
| T42 | <elros34> wasn't there some systemd mount unit in fp4 repo with read to use solution? | 09:32 |
| T42 | <k1gen> https://github.com/mlehtima/droid-config-fp4/blob/devel/sparse/usr/lib/systemd/system/vendor-lib64-hw-audio.primary.default.so.mount this one? | 09:35 |
| T42 | <k1gen> hidl_compat build, now I have out/target/product/flame/system/lib/hw/audio.hidl_compat.default.so | 09:38 |
| T42 | <k1gen> after starting the service successfully and running pulseaudio, another segfault. wait for stdout, gdb and logcat | 09:43 |
| T42 | <elros34> I assume you follow instruction and audioserver is up otherwise do not paste anything? | 09:44 |
| T42 | <elros34> aduio service* | 09:45 |
| T42 | <k1gen> pulseaudio is masked if that's what you are asking, I'm starting it manually | 09:48 |
| T42 | <k1gen> I also can't find 'hidl' in hadk, hadk-hot and hadk-faq, so I don't quite understand what instructions you are talking about either | 09:49 |
| T42 | <elros34> basic instruction is in hadk-hot, including service name to reenable in disabled_services.rc | 09:52 |
| T42 | <k1gen> so delete the service audioserver audioserver_HYBRIS_DISABLED line and reboot? | 09:54 |
| T42 | <elros34> obviously | 09:55 |
| T42 | <k1gen> should I also delete the service vendor.audio-hal /vendor/bin/hw/android.hardware.audio.service_HYBRIS_DISABLED line? | 09:56 |
| T42 | <elros34> ah so this one instead (audio.service) | 09:57 |
| T42 | <k1gen> deleted service vendor.audio-hal /vendor/bin/hw/android.hardware.audio.service_HYBRIS_DISABLED and recovered service audioserver audioserver_HYBRIS_DISABLED, rebooting | 09:58 |
| T42 | <k1gen> 2659 audioser {audio.service} /vendor/bin/hw/android.hardware.audio.service in ps aux, pulse still segfaults | 10:01 |
| T42 | <k1gen> @elros34 stdout: https://vomitb.in/raw/8JZbwuU43M; gdb: https://vomitb.in/raw/djnWZ6vZ1t | 10:06 |
| Mister_Magister | It's nice that there's qt 6, i can just add 5. in the url to get docs for sfos | 10:06 |
| T42 | <k1gen> still nothing in maps: | 10:08 |
| T42 | <k1gen> Thread 5 "droid-sink-prim" received signal SIGSEGV, Segmentation fault. | 10:08 |
| T42 | <k1gen> [Switching to Thread 0x7fb53a2140 (LWP 5580)] | 10:08 |
| T42 | <k1gen> | 10:08 |
| T42 | <k1gen> # pgrep 'pulseaudio' | 10:08 |
| T42 | <k1gen> 5564 | 10:08 |
| T42 | <k1gen> | 10:08 |
| T42 | <k1gen> # awk -F'[ -]' -v s=7fb53a2140 '$1<=s&&s<$2' /proc/5564/maps | 10:08 |
| T42 | <k1gen> 7fb51a3000-7fb53a3000 rw-p 00000000 00:00 0 | 10:08 |
| T42 | <k1gen> @elros34 sorry, forgot to enable vendor-lib64-hw-audio.primary.default.so.mount before rebooting. pa stdout withaudio.hidl_compat.default.so bind mounted to audio.primary.default.so: https://vomitb.in/raw/gjv07Opdgg; gdb: https://vomitb.in/raw/mcn7SnQpEV | 10:15 |
| mal | that still loaded the original module, quite sure default module is only used as fallback when nothing else is found | 10:19 |
| T42 | <k1gen> how do I load hidl_compat.so then? | 10:21 |
| T42 | <elros34> then you need to change bind mount unit. If you run 'strace -e file' then you should see what is really opened, next time you do such a test confirm with starce | 10:21 |
| T42 | <k1gen> strace: https://0x0.st/H8rf.txt | 10:24 |
| mal | it loads /vendor/lib64/hw/audio.primary.msmnile.so | 10:25 |
| T42 | <k1gen> I see, because the platform is msmnile. I saw this name playing with caf long ago | 10:26 |
| T42 | <k1gen> so I should rename mount unit to vendor-lib64-hw-audio.primary.msmnile.so.mount and change Where= value? | 10:28 |
| T42 | <k1gen> What=/usr/libexec/droid-hybris/system/lib64/hw/audio.hidl_compat.default.so | 10:30 |
| T42 | <k1gen> Where=/vendor/lib64/hw/audio.primary.msmnile.so | 10:30 |
| T42 | <k1gen> just to be sure | 10:30 |
| T42 | <k1gen> mal: https://vomitb.in/raw/U8BoG6TGEI | 10:31 |
| mal | are you sure you copied the correct library ""/vendor/lib64/hw/audio.primary.msmnile.so" is 32-bit instead of 64-bit" | 10:34 |
| T42 | <k1gen> mal: HABUILD_SDK [flame] [k1gen@think hadk]$ file out/target/product/flame/system/lib/hw/audio.hidl_compat.default.so | 10:35 |
| T42 | <k1gen> out/target/product/flame/system/lib/hw/audio.hidl_compat.default.so: ELF 32-bit LSB shared object, ARM, EABI5 version 1 (SYSV), dynamically linked, BuildID[md5/uuid]=ab267c7c58c2662e2a10798e949f29d6, stripped | 10:35 |
| T42 | <k1gen> HABUILD_SDK [flame] [k1gen@think hadk]$ file out/target/product/flame/system/lib/hw64/audio.hidl_compat.default.so | 10:35 |
| T42 | <k1gen> out/target/product/flame/system/lib/hw64/audio.hidl_compat.default.so: cannot open `out/target/product/flame/system/lib/hw64/audio.hidl_compat.default.so' (No such file or directory) | 10:35 |
| mal | out/target/product/flame/system/lib64/hw/audio.hidl_compat.default.so | 10:36 |
| mal | not lib/hw64 | 10:36 |
| T42 | <k1gen> yeah, sorry. out/target/product/flame/system/lib64/hw/audio.hidl_compat.default.so: ELF 64-bit LSB shared object, ARM aarch64, version 1 (SYSV), dynamically linked, BuildID[md5/uuid]=98ddbc173b8b7aaf6bb95b16cdf54a33, stripped | 10:36 |
| T42 | <k1gen> different error after copying 64-bit lib | 10:37 |
| T42 | <k1gen> mal: https://vomitb.in/raw/88Xr0VY5Pe | 10:38 |
| T42 | <Verevka86> What is your device? don't you have sound? | 10:40 |
| mal | which pulse droid module are you using? | 10:40 |
| mal | @Verevka86 pulseaudio is crashing on that device | 10:40 |
| T42 | <k1gen> mal: how do I check? | 10:40 |
| T42 | <k1gen> @Verevka86 google pixel 4 (flame), don't have sound | 10:41 |
| T42 | <Verevka86> Base los? (re @k1gen: @Verevka86 google pi...) | 10:41 |
| T42 | <k1gen> don't reply. 18.1 | 10:42 |
| T42 | <Verevka86> https://github.com/SailfishOS-vayu/android_vendor_xiaomi/commit/e8271e5137c5e89ec17120e160a7b20937f5d7c5 (re @k1gen: don't reply. 18.1) | 10:42 |
| T42 | <k1gen> don't reply please, it doesn't look good in irc. what do I do with what you sent? | 10:43 |
| T42 | <Verevka86> Sorry 😇 | 10:43 |
| T42 | <Verevka86> find these libraries in the stock firmware and add when building los 18.1 | 10:44 |
| T42 | <k1gen> I'm sorry, "building lineageos"? I deleted lineage repo, because I didn't have enough space. can I replace them in sailfish directly? | 10:45 |
| T42 | <TheVancedGamer> you can try, mer-hybris repo has at least half of AOSP/lineage repos | 10:47 |
| T42 | <k1gen> so I can't just replace .so libs in those paths on my device running sailfish with git ones @Verevka86 sent? | 10:49 |
| T42 | <Verevka86> You need your libraries, you can add them with -bind directly to sailfish | 10:53 |
| T42 | <Verevka86> https://github.com/craigacgomez/android_vendor_google/blob/r11x/flame/proprietary/vendor/lib/hw/audio.primary.msmnile.so | 10:57 |
| T42 | <k1gen> sorry, I'm lost, I don't understand what should I do. "find these libraries in the stock firmware and add", "You need your libraries, you can add them with -bind directly to sailfish". what should I do exactly? | 10:57 |
| T42 | <Verevka86> https://github.com/craigacgomez/android_vendor_google/blob/r11x/flame/proprietary/vendor/lib64/hw/audio.primary.msmnile.so | 10:57 |
| T42 | <Verevka86> Find these libraries in the official firmware | 10:58 |
| mal | @Verevka86 it crashes even with audio.primary.msmnile.so loaded, we were attempting to use hidl compat to see if there is some issue in 64-bit lib | 10:59 |
| T42 | <k1gen> mal: thank god you're here, I understand you explanations better. did you see new pulseaudio crash stdout, any ideas? any use for these links Verevka86 sent? | 11:01 |
| mal | @k1gen I mean do you have pulseaudio-modules-droid or pulseaudio-modules-droid-jb2q? | 11:01 |
| T42 | <k1gen> mal: pulseaudio-modules-droid | 11:02 |
| T42 | <k1gen> pulseaudio-modules-droid-hidl is also installed | 11:02 |
| T42 | <Verevka86> audio_policy_configuration.xml fixed? | 11:03 |
| mal | I think pulseaudio-modules-droid does not work with hidl compat | 11:03 |
| mal | unless you have the patches I have locally :) | 11:03 |
| T42 | <k1gen> I don't think I have them | 11:03 |
| mal | try installing jb2q | 11:04 |
| T42 | <k1gen> again? | 11:04 |
| T42 | <Verevka86> NullPointerException: show your audio_policy_configuration.xml | 11:05 |
| T42 | <k1gen> stupid question, sorry. give me a minute | 11:05 |
| T42 | <k1gen> @Verevka86: https://vomitb.in/raw/Oi70H8NJcv | 11:07 |
| T42 | <k1gen> mal: built it, is the command to install them right: devel-su zypper in pulseaudio-modules-droid-jb2q-14.2.101-1.aarch64.rpm pulseaudio-modules-droid-jb2q-common-14.2.101-1.aarch64.rpm -pulseaudio-modules-droid -pulseaudio-modules-droid-common ? | 11:10 |
| T42 | <k1gen> I mean, is it everything that's needed | 11:12 |
| T42 | <Verevka86> Try it- https://github.com/SailfishOS-vayu/android_device_xiaomi_sm8150-common/commit/080423907c28d5460af0389c908b0f276f12f30f | 11:13 |
| T42 | <Verevka86> though I'm not sure that this has not been fixed in pulseaudio | 11:13 |
| T42 | <k1gen> will this work with jb2q? | 11:13 |
| T42 | <k1gen> nvm, /vendor/etc/audio_policy_configuration.xml is unwritable | 11:14 |
| T42 | <k1gen> fixed it with bind mount | 11:16 |
| T42 | <Verevka86> Yes | 11:16 |
| mal | @Verevka86 that was fixed already in pulseaudio https://github.com/mer-hybris/pulseaudio-modules-droid/commit/f0ac6091a3596447753087764b595ef71eebf1bd | 11:19 |
| T42 | <k1gen> @Verevka86: https://vomitb.in/raw/cQml3ISO4I | 11:20 |
| T42 | <k1gen> mal: I will revert this now, also is the command to install -jb2q right? | 11:21 |
| mal | I don't really know what you have now in there | 11:21 |
| mal | was that last one with or without hidl compat | 11:21 |
| mal | that command for installing jb2q seemed ok | 11:22 |
| T42 | <k1gen> mal: https://vomitb.in/raw/o5yt0JKfOh | 11:23 |
| mal | @Verevka86 I use 24-bit audio on fp4 https://github.com/mlehtima/droid-config-fp4/blob/devel/sparse/etc/pulse/arm_droid_card_custom.pa | 11:23 |
| mal | @k1gen did you install jb2q or not? | 11:24 |
| mal | you might need to use --force if it won't otherwise | 11:24 |
| T42 | <k1gen> not yet, reverting these changes Verevka suggested | 11:24 |
| T42 | <k1gen> The following 2 NEW packages are going to be installed: | 11:25 |
| T42 | <k1gen> pulseaudio-modules-droid-jb2q pulseaudio-modules-droid-jb2q-common | 11:25 |
| T42 | <k1gen> | 11:25 |
| mal | @Verevka86 both pulseaudio-modules-droid and pulseaudio-modules-droid-jb2q have the 24-bit audio fix | 11:26 |
| T42 | <k1gen> The following 2 packages are going to be REMOVED: | 11:26 |
| T42 | <k1gen> pulseaudio-modules-droid pulseaudio-modules-droid-common | 11:26 |
| mal | you really don't need to paste such trivial things | 11:26 |
| T42 | <k1gen> I see, sorry | 11:26 |
| T42 | <k1gen> a lot of output from pulse, still running | 11:27 |
| T42 | <k1gen> let me kill it and show you stdout | 11:27 |
| mal | did audio work? | 11:27 |
| T42 | <k1gen> https://vomitb.in/raw/EITSu7CS4M | 11:28 |
| T42 | <k1gen> let me play something | 11:28 |
| T42 | <k1gen> SOUND WORKS | 11:29 |
| T42 | <k1gen> thank you mal and elros! | 11:30 |
| mal | still would be nice to know why the 64-bit hal fails | 11:30 |
| T42 | <k1gen> yeah, would be nice. any way to know that? if even the 8xxMb log didn't help | 11:34 |
| T42 | <k1gen> I noticed a bug: if you press play/pause a couple of times, volume resets and device starts playing on max volume. can I do anything about it? | 11:47 |
| mal | have you tried disabling hw volume control using the quirk | 11:48 |
| T42 | <k1gen> usually this happens on lockscreen, it's enough to press play/pause button one-four times to trigger this, if device is unlocked - it's much more rare | 11:49 |
| mal | hmm | 11:49 |
| T42 | <k1gen> mal: no, not yet. I remember you sent a quirk for pulseaudio-modules-droid, but not for pulseaudio-modules-droid-jb2q | 11:50 |
| T42 | <k1gen> yeah, it happens pretty much every time on lockscreen if music wasn't playing for some time | 11:51 |
| T42 | <k1gen> it can be fixed with volume change using side buttons and touching play/pause button two times | 11:51 |
| T42 | <k1gen> if device is unlocked, issue is almost the same, with one difference - volume is maxed out for a fraction of a second, then it becomes ok again. kind of audio-jumpscare | 11:53 |
| mal | how about reading the readme in jb2q repo | 11:55 |
| T42 | <k1gen> no_hw_volume? | 11:57 |
| T42 | <k1gen> mal: load-module module-droid-card rate=48000 quirks=+no_hw_volume in /etc/pulse/arm_droid_card_custom.pa helped me, thanks | 12:06 |
| T42 | <k1gen> only camera left, but I think most difficult stuff is behind | 12:08 |
| T42 | <edp_17> mal : After this success, do you have idea why audio doesn't work on my motog7? :) | 13:16 |
| T42 | <k1gen> sound in calls doesn't work. I don't hear them and they don't hear me. calls work fine though, as does cellular and hotspot | 13:35 |
| T42 | <edp_17> Does the device ring for incoming calls? | 13:39 |
| T42 | <k1gen> let me ask someone to call me | 13:40 |
| T42 | <k1gen> @edp_17 it does | 13:43 |
| T42 | <edp_17> Good for you. Mine doesn't :( | 13:44 |
| T42 | <k1gen> well, I don't have audio in calls, so it doesn't really matter that much if it rings or not | 13:44 |
| mal | @k1gen so similar issue that I have on fp4 | 13:45 |
| T42 | <k1gen> huh | 13:45 |
| T42 | <edp_17> @k1gen : It seems now we have similar problem. | 13:45 |
| T42 | <k1gen> nice, three guys stuck on same issue | 13:46 |
| mal | @edp_17 does audio work at all for you? | 13:46 |
| T42 | <edp_17> I also use hidl_compat and bind mount, have sound when playing music via audio jack or BT, but no sound in calls. | 13:46 |
| T42 | <k1gen> exactly same situation | 13:46 |
| T42 | <k1gen> don't have jack though | 13:47 |
| T42 | <edp_17> That's why I asked whether you have audio when receiving a call. | 13:47 |
| T42 | <k1gen> I have ringtone sound on incoming calls, but no audio in calls | 13:48 |
| T42 | <edp_17> Oh, my device doesn't ring when receiving a call. | 13:48 |
| mal | @edp_17 somehow feels like this issue is related to hidl_compat | 13:48 |
| T42 | <edp_17> Feels like, yes. | 13:49 |
| T42 | <edp_17> We might need to ask NotKit whether he has idea. | 13:50 |
| T42 | <k1gen> who's NotKit? | 13:50 |
| T42 | <edp_17> I use halium-10.0 branch. Which one are you using? | 13:50 |
| T42 | <k1gen> same | 13:52 |
| T42 | <k1gen> you suggest to try 12.0? | 13:52 |
| T42 | <edp_17> I was thinking about. | 13:53 |
| T42 | <k1gen> mal: what do you think? | 13:53 |
| mal | @edp_17 I have a custom hidl_compat | 13:54 |
| mal | mainly 12 but with some additional things so I can use droid module instead of jb2q | 13:55 |
| T42 | <k1gen> can you share your repo? | 13:55 |
| mal | I haven't pushed it anywhere | 13:55 |
| mal | yet | 13:55 |
| mal | but it should not matter right now | 13:56 |
| T42 | <k1gen> hmm, so no matter hidl_compat branch and changes - we're still in the same situation | 13:59 |
| mal | yes | 13:59 |
| T42 | <k1gen> then maybe it's not hidl_compat to blame? just thinking | 14:00 |
| mal | well since that is the only thing common in the devices and I haven't heard of others having similar issue hidl_compat seems like strongest candidate for the reason | 14:01 |
| T42 | <edp_17> Agree. Sound worked out of the box on all my other devices that were not using hidl_compat. | 14:06 |
| T42 | <k1gen> @NotKit hey, sorry for ping. do you have any ideas on this hidl_compat issue me, mal and edp_17 are having where there's sound in system but not in calls? | 14:09 |
| Mister_Magister | ehh building zips is still broken? | 18:19 |
| T42 | <k1gen> what should I check first if sleep is broken? | 18:21 |
| Mister_Magister | wifi | 18:24 |
| T42 | <Mister_Magister> https://github.com/mer-hybris/hadk-faq/blob/master/README.rst#bcmdhd-wlan-driver-sleep-suspend-issues (re @k1gen: what should I check ...) | 18:24 |
| mal | also some kernels have auto sleep disabled and need this https://github.com/mer-hybris/android_kernel_sony_msm/blob/hybris-sony-aosp/LA.UM.9.12.r1/arch/arm64/configs/aosp_lena_pdx213_defconfig#L936 | 18:26 |
| mal | also other possible reasons | 18:28 |
| mal | logs should tell if it even tries to suspend and if something is preventing it | 18:28 |
| T42 | <k1gen> mal: yeah: | 18:30 |
| T42 | <k1gen> # zcat /proc/config.gz | grep 'PM_AUTOSLEEP' | 18:30 |
| T42 | <k1gen> # CONFIG_PM_AUTOSLEEP is not set | 18:30 |
| T42 | <k1gen> also, nothing suspicious in journalctl on sleep/suspend | 18:31 |
| mal | then try that config option and see what happens | 18:31 |
| T42 | <k1gen> any mtp-related defconfig keys I should add while I'm here? | 18:32 |
| mal | depends on what usb driver your device uses | 18:32 |
| T42 | <k1gen> pixel mtp mounting fails on my laptop | 18:32 |
| Mister_Magister | mal: any fix for "line 9: $IMG_OUT_DIR/os-release: ambiguous redirect" | 18:33 |
| T42 | <k1gen> lsmod doesn't work :| | 18:33 |
| mal | Mister_Magister: odd, when does that happen? | 18:34 |
| Mister_Magister | mal: i'm trying to build zip | 18:34 |
| Mister_Magister | i can give you entire log | 18:34 |
| Mister_Magister | https://paste.opensuse.org/pastes/c9799f64331e | 18:34 |
| T42 | <k1gen> how can I check which usb driver my device uses? | 18:38 |
| mal | check what usb options are enabled in kernel | 18:39 |
| T42 | <k1gen> mal: https://vomitb.in/raw/JGXMoJIpCN | 18:44 |
| mal | Mister_Magister: just to be sure, do you have correct sdk and tooling and target? also latest version is 4.5.0.19 | 18:50 |
| Mister_Magister | i don't have tooling nor target as they're not needed for mic. sdk i just installed | 18:50 |
| mal | ok, which version of sdk | 18:52 |
| Mister_Magister | uhhh | 18:52 |
| Mister_Magister | latest? | 18:52 |
| Mister_Magister | waait | 18:52 |
| Mister_Magister | ye 4.5 | 18:53 |
| mal | do you build using build_packages.sh -i or using some custom .ks? | 18:53 |
| Mister_Magister | i don't use build_packages, i launch mic by hand, and i snagged .ks from rpm file | 18:54 |
| T42 | <k1gen> mal: I guess CONFIG_USB_CONFIGFS_F_FS and CONFIG_USB_CONFIGFS_F_MTP will be enough, and if yes - they're already enabled | 19:03 |
| mal | well you need to check which functions are available in /config | 19:03 |
| mal | on device | 19:03 |
| mal | somewhere under that | 19:04 |
| T42 | <k1gen> # ls -l /config/usb_gadget/g1/functions/ | 19:04 |
| T42 | <k1gen> drwxr-xr-x 2 root root 0 Apr 13 15:03 accessory.gs2 | 19:04 |
| T42 | <k1gen> drwxr-xr-x 2 root root 0 Apr 13 15:03 audio_source.gs3 | 19:04 |
| T42 | <k1gen> drwxr-xr-x 2 root root 0 Apr 13 15:03 cser.dun.0 | 19:04 |
| T42 | <k1gen> drwxr-xr-x 2 root root 0 Apr 13 15:03 cser.nmea.1 | 19:04 |
| T42 | <k1gen> drwxr-xr-x 2 root root 0 Apr 13 15:03 diag.diag | 19:04 |
| T42 | <k1gen> drwxr-xr-x 2 root root 0 Apr 13 15:03 diag.diag_mdm | 19:04 |
| Mister_Magister | NO | 19:04 |
| T42 | <k1gen> drwxr-xr-x 2 root root 0 Apr 13 15:03 ffs.adb | 19:04 |
| T42 | <k1gen> drwxr-xr-x 2 root root 0 Apr 13 15:03 ffs.mtp | 19:04 |
| Mister_Magister | JESUS | 19:04 |
| T42 | <k1gen> drwxr-xr-x 2 root root 0 Apr 13 15:03 ffs.ptp | 19:04 |
| T42 | <k1gen> drwxr-xr-x 2 root root 0 Apr 13 15:03 gsi.dpl | 19:04 |
| T42 | <k1gen> drwxr-xr-x 2 root root 0 Apr 13 15:03 gsi.rmnet | 19:04 |
| T42 | <k1gen> drwxr-xr-x 2 root root 0 Apr 13 15:03 gsi.rndis | 19:04 |
| T42 | <k1gen> drwxr-xr-x 3 root root 0 Apr 13 15:03 mass_storage.usb0 | 19:04 |
| T42 | <k1gen> drwxr-xr-x 2 root root 0 Apr 13 15:03 midi.gs5 | 19:04 |
| T42 | <k1gen> drwxr-xr-x 2 root root 0 Apr 13 15:03 qdss.qdss | 19:04 |
| T42 | <k1gen> drwxr-xr-x 2 root root 0 Apr 13 15:03 qdss.qdss_mdm | 19:04 |
| T42 | <k1gen> drwxr-xr-x 3 root root 0 Jan 29 1970 rndis.usb0 | 19:04 |
| T42 | <Mister_Magister> @k1gen great spam you did there | 19:05 |
| T42 | <XAP2P> lmao | 19:05 |
| T42 | <k1gen> sorry, if I delete it here, will it be deleted on irc side? | 19:05 |
| T42 | <Mister_Magister> paste.opensuse.org <- off there you go | 19:05 |
| T42 | <Mister_Magister> irc has no delete function | 19:05 |
| T42 | <k1gen> I see. sorry again | 19:05 |
| Mister_Magister | mal: i haven't been able to build image for couple years now | 19:06 |
| Mister_Magister | and i jsut wanna build image from repos | 19:06 |
| mal | and you are sure the .ks is ok? | 19:06 |
| Mister_Magister | i copied it 1:1 from rpm file so sure? | 19:06 |
| T42 | <k1gen> mal: ls -l of /config/usb_gadget/g1/functions/: https://vomitb.in/raw/JhylUtMlfP | 19:06 |
| Mister_Magister | https://paste.opensuse.org/pastes/ee6c5579ad3e here's ks if you wanna take a look | 19:06 |
| mal | @k1gen try adding file /etc/usb-moded/usb-moded-configfs.ini with this content https://github.com/mer-hybris/droid-config-sony-seine/blob/master/sparse/etc/usb-moded/usb-moded-configfs-seine.ini | 19:10 |
| T42 | <k1gen> thank you. I saw that file searching for mtp but wasn't sure I needed it | 19:11 |
| T42 | <k1gen> LIBMTP PANIC: Unable to find interface & endpoints of device with this file. also, device is recognized as Google Inc (for Fairphone): Fairphone 2 (18d1:0a07) :D | 19:21 |
| mal | Mister_Magister: show the mic command you use | 19:24 |
| Mister_Magister | sudo mic create fs --arch=$PORT_ARCH --tokenmap=ARCH:$PORT_ARCH,RELEASE:$RELEASE,EXTRA_NAME:$EXTRA_NAME --record-pkgs=name,url --outdir=sfe-$DEVICE-$RELEASE$EXTRA_NAME --pack-to=sfe-$DEVICE-$RELEASE$EXTRA_NAME.tar.bz2 Jolla-\@RELEASE\@-$DEVICE-\@ARCH\@.ks --verbose --debug | 19:24 |
| T42 | <K31j0> mal: How do I build 18.1 hybris-hal? | 19:25 |
| T42 | <edp_17> @Mister_Magister : I've changed bzip2 to gzip in .ks file AND applied this: https://github.com/sailfishos-on-tucana/hybris-boot/commit/ab283965e5adcc9566c4e6e4450eb681b5d92752 | 19:25 |
| mal | @K31j0 check faq https://github.com/mer-hybris/hadk-faq#hybris-17-1 | 19:26 |
| T42 | <edp_17> And I used this to build image: | 19:28 |
| T42 | <edp_17> sudo mic create fs --arch=$PORT_ARCH --tokenmap=ARCH:$PORT_ARCH,RELEASE:$RELEASE,EXTRA_NAME:$EXTRA_NAME --record-pkgs=name,url --outdir=sfe-$DEVICE-$RELEASE$EXTRA_NAME --pack-to=sfe-$DEVICE-$RELEASE$EXTRA_NAME.tar.gz $ANDROID_ROOT/Jolla-@RELEASE@-$DEVICE-@ARCH@.ks #--verbose #--debug | 19:28 |
| T42 | <edp_17> You need to replace tar.bz2 with tar.gz in the mic command too. | 19:29 |
| Mister_Magister | @edp_17 will changing bzip2 to gzip really affect anything? | 19:30 |
| Mister_Magister | it looks like something wrong with ks file | 19:30 |
| mal | Mister_Magister: I see some diff compared to my .ks | 19:30 |
| Mister_Magister | mal: huh | 19:30 |
| T42 | <k1gen> supposedly fixing twrp installation | 19:30 |
| Mister_Magister | i don't have issue with instalation, i can't even build bzip2 | 19:31 |
| T42 | <edp_17> @Mister_Magister : that's how I moved to gzip. | 19:32 |
| T42 | <edp_17> Oh, I see. Then you should check your .ks file. | 19:32 |
| Mister_Magister | ye i didn't modify it | 19:32 |
| T42 | <edp_17> Save the current .ks, and let it regenerate. Them compare the two. | 19:33 |
| mal | Mister_Magister: I have this repo in list repo --name=customer-jolla-@RELEASE@ --baseurl=https://releases.jolla.com/features/@RELEASE@/customers/jolla/@ARCH@/ and this diff https://pastebin.com/UDfqCxCX | 19:33 |
| Mister_Magister | i'm using correct droid-configs for 4.5 tho | 19:33 |
| mal | can't remember when I regenerated my .ks | 19:34 |
| Mister_Magister | and your diff is not affecting the area where error exists | 19:34 |
| mal | yeah, still you should try | 19:34 |
| mal | add that repo and diff and see what happens | 19:34 |
| Mister_Magister | especially doubt repo will change anything as on device i don't have customer-jolla repo | 19:35 |
| Mister_Magister | mal: can u give me your entire ks file and device/vendor/os version? | 19:35 |
| Mister_Magister | i'll try building it | 19:35 |
| Mister_Magister | and see if it works | 19:35 |
| mal | can't you just do the changes, I only removed the device specific parts | 19:38 |
| Mister_Magister | i could, but that way we would eliminate any variables | 19:38 |
| T42 | <K31j0> Oh by the way, where’s my manners. Hello after a long break, mal! I was porting sirius (Xperia Z2) and lux (Moto X Play) with Mister_Magister | 19:43 |
| T42 | <K31j0> A looong while back :D | 19:43 |
| Mister_Magister | and g3 | 19:43 |
| T42 | <K31j0> Last commit on sirius and lux is from 2019, took a while | 19:44 |
| Mister_Magister | meanwhile mal be like "i don't even know who you are" | 19:47 |
| T42 | <k1gen> mal: sleep works with CONFIG_PM_AUTOSLEEP=y, thanks for suggestion. mtp still doesn't though, even with that file you suggested | 20:03 |
| T42 | <elros34> @Mister_Magister ks depends on sdk target so make you use correct onee | 20:15 |
| Mister_Magister | @elros34 no it doesn't | 20:16 |
| Mister_Magister | mic doesn't use target | 20:16 |
| T42 | <elros34> I am not talking about mic but generated ks file | 20:17 |
| Mister_Magister | that i took from obs | 20:17 |
| mal | Mister_Magister: did the modified .ks help? | 20:51 |
| Mister_Magister | didn't try yet | 20:51 |
| T42 | <K31j0> mal: any idea what to do if the branch is in the git repo, but the repo tool says that the revision isn’t found? | 21:17 |
| mal | which repo? | 21:26 |
| mal | if you have some issue like that when doing repo sync then don't use --fetch-submodules | 21:27 |
| T42 | <K31j0> I’m using repo sync -c -j16 | 21:27 |
| T42 | <K31j0> I’m trying to sync TheMuppets/proprietary_vendor_asus into vendor/asus from revision lineage-18.1 | 21:27 |
| mal | make sure you don't have some invisible character in the line somewhere | 21:29 |
| T42 | <K31j0> It started working out of the blue | 21:29 |
| T42 | <K31j0> Gotta love repo | 21:29 |
| Mister_Magister | just keep sycncing, just keep syncing… | 21:29 |
| Mister_Magister | mal: 18.1 is usable right? | 21:51 |
| Mister_Magister | i 'member about 19 being experimental iirc | 21:52 |
| T42 | <k1gen> ofc it's usable | 21:53 |
| Mister_Magister | nice | 21:56 |
| Mister_Magister | its just lacks instructions | 21:56 |
| Mister_Magister | but imma guide @K31j0 instead xd | 21:57 |
| mal | Mister_Magister: there is no 19 at the moment, I have locally hybris-20 though | 22:04 |
| Mister_Magister | xd | 22:04 |
| mal | and device booting fine with hybris-20 | 22:04 |
| mal | everything except call audio, mobile data and video recording works | 22:05 |
| Mister_Magister | nice | 22:05 |
| Mister_Magister | mal: so basically situation looks like this: @K31j0 be porting zenfone 8 on 18.1, there are no instructions but imma guide, and if anything we'll harass you, okiday? xdddd | 22:06 |
| Mister_Magister | i might get same phone later | 22:06 |
| * Mister_Magister verdandiTEAM back in TEAM | 22:06 | |
| mal | good starting point following faq and checking hadk-hot for extra hints like submodule revisions etc | 22:06 |
| T42 | <K31j0> How hard is porting hybris to newer bases? Google has been changing so much stuff lately. | 22:07 |
| Mister_Magister | better and worse | 22:07 |
| Mister_Magister | we got binder which makes things easier | 22:07 |
| Mister_Magister | but bootloader is brainfuck | 22:07 |
| mal | hybris-20 was surprisingly easy, trying to remember how long basic stuff took, maybe like couple of days at most | 22:07 |
| mal | could have been even less | 22:08 |
| mal | droidmedia took a bit time because android changed things again | 22:09 |
| mal | that reminds me, I probably should make some PRs | 22:09 |
| Mister_Magister | mal: that reminds me | 22:09 |
| Mister_Magister | hows h264 seeking going xd | 22:09 |
| mal | I didn't yet have time to look into that | 22:10 |
| Mister_Magister | ahh | 22:10 |
| Mister_Magister | mal: when you're bored plz create nemo:deve:hw:asus:sake | 22:37 |
| * Mister_Magister tfw many of my repos were still created by sledges ;-; miss him | 22:38 | |
| T42 | <K31j0> What happened to him? :< | 22:40 |
| Mister_Magister | nobody knows | 22:40 |
| Mister_Magister | back in my day… | 22:41 |
| mal | https://build.sailfishos.org/project/show/nemo:devel:hw:asus:sake and https://build.merproject.org/project/show/nemo:testing:hw:asus:sake | 22:48 |
| Mister_Magister | thank! | 22:48 |
| Mister_Magister | lol sledges doesn't work for jolla since 3 years already | 23:03 |
| Mister_Magister | daaang i'm old | 23:03 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!