| * Danct12 sent a long message: < https://matrix.org/_matrix/media/v1/download/matrix.org/VQzbBSnYXdowBZyBbSTTwzVD > | 01:51 | |
| Danct12 | I got a error while building the device package. | 01:51 |
|---|---|---|
| Danct12 | * Building rpm/droid-config-santoni.spec | 01:51 |
| Danct12 | the log is above, can anyone help me? | 01:51 |
| asriel[m] | faq-hadk has a section for that | 02:00 |
| mccreary | piggz: It seems nothing has changed in the Qualcomm boot_control HAL since nougat, so I pulled the pre-HIDL source. It seems to work on my Pro1, can you give it a try? https://drive.google.com/open?id=1VqLgQvutOxX3SpevA_P16ps9FpoX8QOG | 02:27 |
| Danct12 | got that fixed and now I got a new error when running mic | 02:27 |
| Danct12 | ``` | 02:27 |
| Danct12 | Warning[07/21 02:24:10] : repo problem: nothing provides pulseaudio-modules-droid needed by pattern:jolla-hw-adaptation-santoni-1-1.noarch, | 02:27 |
| Danct12 | ``` | 02:27 |
| mccreary | With luck that will bypass the 14-character truncation bug | 02:28 |
| Danct12 | pulseaudio-modules-droid needed by pattern:jolla-hw-adaptation-santoni-1-1.noarch | 02:30 |
| asriel[m] | <Danct12 "pulseaudio-modules-droid needed "> Have you completely build_packages.sh? | 03:17 |
| Danct12 | asriel: that did it | 03:51 |
| * Danct12 sent a long message: < https://matrix.org/_matrix/media/v1/download/matrix.org/hkAqKGVsQMVpRryIXgrXvryw > | 03:51 | |
| Danct12 | thanks | 03:51 |
| asriel[m] | Alright, so I just flashed SFOS and now I got this telnet shell | 04:50 |
| * asriel[m] sent a long message: < https://matrix.org/_matrix/media/v1/download/matrix.org/BwnkEUVZTVGoNIahrRhAudqL > | 04:50 | |
| asriel[m] | What can I do next? | 04:53 |
| liux | asriel[m], UI already shows on device? | 04:55 |
| asriel[m] | Only stuck at the bootloader logo, and only the white led shows up at the bottom of the device | 05:03 |
| asriel[m] | And only telnet | 05:04 |
| liux | asriel[m], how about https://wiki.merproject.org/wiki/Adaptations/faq-hadk#lipstick_segfaults.2Fno_display | 05:06 |
| liux | asriel[m], and try to print log for journal, dmesg and logcat | 05:06 |
| asriel[m] | I don't have wifi so I cannot do network based stuffs | 05:10 |
| asriel[m] | but will send logs, hold on | 05:10 |
| asriel[m] | and now for some reason it just rebooted to fastboot | 05:10 |
| liux | asriel[m], can't you get logs with telnet? | 05:11 |
| asriel[m] | I could, using netcat of course. | 05:14 |
| asriel[m] | dmesg: https://paste.ubuntu.com/p/k7HDX4Rnb4/ | 05:14 |
| liux | asriel[m], no certain reason for reboot. for instance, my device reboot because lack of patch of cgroup in kernel, then lack of defination in liblog. most of them can be found in hadk document or your logs | 05:14 |
| asriel[m] | logcat: https://paste.ubuntu.com/p/Y9txdxf2tF/ | 05:15 |
| asriel[m] | and journalctl: https://paste.ubuntu.com/p/2FnTJ5KJ3k/ | 05:16 |
| liux | asriel[m], seems hwservicemanager are killed many times | 05:21 |
| asriel[m] | <liux "asriel, seems hwservicemanager a"> yep seems like everything is depending on hwservicemanager | 05:26 |
| liux | asriel[m], yes so it caused pausing or reboot. i've never met hwservicemanager issue so can't give any advise now. | 05:28 |
| liux | asriel[m], you may dig it yourself and ask others if experienced. or search logs of porters channel | 05:29 |
| electro | elros34, hi an idea with log ? | 08:36 |
| T42 | <elros34> still the same: while true; do dmesg -c;done and continue to boot in another window | 08:40 |
| electro | thk | 08:42 |
| electro | i don't understand all | 08:42 |
| electro | dmesg -c just have output | 08:44 |
| electro | but if i launch cmd "continue" | 08:44 |
| electro | i lost telnet connection | 08:44 |
| electro | and no possible to relaunch dmesg | 08:44 |
| electro | Can you explain to me more presicly elros34 ? | 08:46 |
| electro | thanks a lot | 08:46 |
| T42 | <elros34> it's NOT dmesg -c, copy WHOLE command I gave you and execute in telnet then open new terminal window, telnet to device and continue booting then pasebin whole output of dmesg | 08:48 |
| electro | this cmd ? | 08:50 |
| electro | while true; do dmesg -c;done | 08:50 |
| electro | ok, i will try | 08:51 |
| T42 | <elros34> yes, normally you would use dmesg -w (follow) but for some reason it doesn't work in sfos at least for me | 08:51 |
| liux | mal, I found the video playback is not really "frozen". After pressing play button more than 30 seconds (sometimes up to 2 minitus), it plays video as normal | 08:55 |
| electro5751 | so elros34, i launch this cmd : while true; do dmesg -c; done | 09:06 |
| electro5751 | and this on another terminal : echo "continue" > /init-ctl/stdin | 09:07 |
| electro5751 | or the full cmd : echo "continue" > /init-ctl/stdin; sleep 5; cat /init.log; journalctl —no-page | 09:08 |
| electro5751 | when i launch the cmd continue elros34, i have this output dmesg : [ 268.556409] c3 EXT4-fs (mmcblk0p16): re-mounted. Opts: (null) | 09:10 |
| electro5751 | and nothing more -> battery picture of the device | 09:10 |
| T42 | <elros34> try to disable WATCHDOG_NOWAYOUT in defconfig | 09:15 |
| electro5751 | ok | 09:32 |
| electro5751 | i must go out, i will try this later, thanks elros34 => I must recompile all package with this changes ? | 09:33 |
| electro5751 | WATCHDOG_NOWAYOUT | 09:33 |
| T42 | <elros34> you should already learn something, when you change kernel defconfig then make hybris-boot is enough | 09:36 |
| liux | I can dial but the micphone does not work..... | 09:59 |
| T42 | <elros34> do you have audioflingerglue installed? | 10:06 |
| liux | elros34, yes, and i can hear from phone or video play | 10:07 |
| liux | elros34, just not mic input | 10:08 |
| asriel[m] | Oh, so a further investigation, I found that hwservicemanager was segfault when LD_PRELOAD as in faq-hadk | 10:21 |
| asriel[m] | https://github.com/mer-hybris/droid-config-sony-nile/blob/master/sparse/usr/libexec/droid-hybris/system/etc/init/hwservicemanager.rc | 10:21 |
| mal | asriel[m]: I assume hybris-15.1 base? do you have the selinux_stubs patch? | 10:22 |
| asriel[m] | Which patch? I never saw any patches in the faq so not sure | 10:22 |
| asriel[m] | or maybe I didn't read carefully enough | 10:23 |
| asriel[m] | and yes, hybris-15.1 | 10:23 |
| mal | I think nobody added it to faq | 10:24 |
| mal | asriel[m]: https://piggz.co.uk/sailfishos-porters-archive/index.php?log=2019-05-27.txt#line24 | 10:26 |
| asriel[m] | Probably, I tried to perform a search on the faq-hadk just to find Bluetooth patches which isn't related to this hwservicemanager error | 10:26 |
| mal | I probably should add that directly to selinux_stubs repo anyway | 10:26 |
| piggz | \o/ my webservice! | 10:27 |
| asriel[m] | Riggght, so apply this patch and re-run hybris-hal and recreate a new rootfs then I should.. get something new? Let's see. :P | 10:28 |
| mal | I added that to 15.1 porting notes in faq until I push the patch to selinux_stubs repo | 10:33 |
| liux | does anyone have met no sound input issue? micphone seems not work | 10:35 |
| mal | logs | 10:37 |
| vknecht | csd :-) | 10:39 |
| vknecht | confirmed a broken mike with it once :-/ | 10:40 |
| liux | Unable to open log device 'main'. after rm /dev/log and relink to /dev/alog, the logcat can only print logs for new started process | 10:46 |
| mal | that should not be needed anymore | 10:48 |
| asriel[m] | mal: Oh, and by rebuilding hybris-hal.. you mean rm -rf the output folder and run it again? | 11:03 |
| asriel[m] | aka clean build | 11:04 |
| mal | no need, just run make hybris-hal again, it should rebuild needed things, hopefully | 11:10 |
| asriel[m] | Can I just flash a new sailfishos rom on top of the old one? | 11:38 |
| mal | just install the zip | 11:40 |
| asriel[m] | mal: heeeeey, that worked! | 12:06 |
| * asriel[m] sent a long message: < https://matrix.org/_matrix/media/v1/download/matrix.org/MpHcGGdcnzxZWfZkfhfCtJdv > | 12:06 | |
| asriel[m] | now i'll deal with another problem that the firmware failed to load | 12:06 |
| asriel[m] | oh right, as a temporary workaround i symlink'd | 12:15 |
| asriel[m] | anyway, here comes another problem | 12:16 |
| asriel[m] | The UI seems to crash and it just took away the network interface with it | 12:16 |
| asriel[m] | now all i got was a charging led (it seems) and black screen | 12:17 |
| asriel[m] | now it no longer come back :( | 12:21 |
| asriel[m] | even though i can still turn on the screen by pressing the power button, it's just black | 12:21 |
| mal | asriel[m]: can you telnet to the device before that happens and check output of "mount" and "systemctl status" also journalctl -a would be useful | 12:23 |
| mal | asriel[m]: did you see UI already or did it fail before that? | 12:24 |
| asriel[m] | <mal "asriel: did you see UI already o"> On the first boot, yes, asking me to do normal setup stuffs, then during the timezone setup it mysteriously vanishes to nothingness | 12:25 |
| asriel[m] | All it left was the loading icon | 12:25 |
| asriel[m] | Now it no longer disables the interface, but the screen is still black | 12:26 |
| mal | check logs, dmesg, journalctl and /usr/libexec/droid-hybris/system/bin/logcat | 12:26 |
| asriel[m] | btw here is "mount": https://paste.ubuntu.com/p/w82rVg7YZ5/ | 12:30 |
| mal | ok, that seems reasonable | 12:30 |
| asriel[m] | dmesg: https://paste.ubuntu.com/p/NVPhnBdtFR/ | 12:34 |
| asriel[m] | journalctl -a: https://paste.ubuntu.com/p/7Cqsgbz3YG/ | 12:34 |
| asriel[m] | and the list of services https://paste.ubuntu.com/p/DfjH9XBkMX/ | 12:35 |
| asriel[m] | the logcat is spammed with adsprpcd errors | 12:36 |
| * asriel[m] sent a long message: < https://matrix.org/_matrix/media/v1/download/matrix.org/EEIwjFGnmCSBHKpCBlLWnBya > | 12:36 | |
| mal | asriel[m]: hmm, could you check if you can see any fstab and uevent*.rc files anywhere under $ANDROID_ROOT/out in ouyr build env | 12:38 |
| asriel[m] | yes, fstab is in out/target/product/santoni/vendor/etc/fstab.qcom | 12:41 |
| asriel[m] | and both ueventd.rc is in out/target/product/santoni/root/ueventd.rc and out/target/product/santoni/vendor/ueventd.rc | 12:42 |
| mal | ok, then it's probably ok | 12:42 |
| mal | some devices have had issues that those files are not copied to out/ which then leads to all kinds of issues | 12:43 |
| mal | asriel[m]: is that some normal qcom device? | 12:46 |
| asriel[m] | yes, it's a qualcomm device | 12:46 |
| mal | asriel[m]: can you check android-config.h file in droid-hal-$DEVICE-devel rpm (n in droid-hal-repo) for anything likeQ QCOM_BSP | 12:47 |
| mal | sorry for the amount of typos, the keyboard on my laptop is breaking apart and causing a lot of errors | 12:49 |
| asriel[m] | hadk/rpm/droid-hal-santoni.spec ? | 12:55 |
| mal | asriel[m]: I mean the actual rpm file generated by build | 12:55 |
| mal | asriel[m]: somewhere under $ANDROID_ROOT/droid-local-repo/$DEVICE | 12:56 |
| mal | asriel[m]: I think I might know the reason for the problem, that's why I ask for you to check the header | 12:57 |
| asriel[m] | /* Added by automatic QCOM_BSP detection */ | 12:58 |
| asriel[m] | #define QCOM_BSP 1 | 12:58 |
| asriel[m] | #define QTI_BSP 1 | 12:58 |
| mal | hmm, so you have those there, then my guess was wrong | 12:59 |
| mal | asriel[m]: can you try to see if logcat shows any other messages if you try to restart lipstick.service in another terminal | 13:03 |
| asriel[m] | can't because of Failed to get D-Bus connection: No such file or directory | 13:06 |
| mal | asriel[m]: those adsp errors in logcat look bad anyway, probably some missing or not found firmware, can you check where you have some firmare files, probably /vendor/firmware and/or /system/etc/firmware, which those | 13:06 |
| mal | asriel[m]: systemctl-user restart lipstick.service (or was it systemctl --user) | 13:07 |
| * asriel[m] sent a long message: < https://matrix.org/_matrix/media/v1/download/matrix.org/BaZwxRTjVvEIUzBUpMWTSaKa > | 13:09 | |
| asriel[m] | Oh that's weird, I see the UI now | 13:09 |
| mal | does it seem stable? | 13:10 |
| asriel[m] | I don't have the touch screen yet sadly | 13:10 |
| asriel[m] | and now no more network interface but the UI is still waiting for me to unlock it through the non-working touch screen | 13:11 |
| asriel[m] | Actually I can swipe the top bar down | 13:12 |
| mal | asriel[m]: fix the input device in https://github.com/mlehtima/droid-config-fp2-sibon/blob/master/sparse/var/lib/environment/compositor/droid-hal-device.conf#L9 | 13:12 |
| asriel[m] | Regarding the ADSP error, it tries to look adsp.mdt which is in /firmware | 13:13 |
| mal | asriel[m]: run "ln -s /firmware /lib/firmware" on device | 13:15 |
| mal | then reboot | 13:16 |
| mal | asriel[m]: oh, if you can swipe from top that probably mean the startup wizards have not been finished properly, or something else | 13:17 |
| asriel[m] | Anyway, I catched the lipstick error | 13:19 |
| * asriel[m] sent a long message: < https://matrix.org/_matrix/media/v1/download/matrix.org/VWszgKtiqxJmghHzwGsSmFxX > | 13:19 | |
| asriel[m] | and yes, it does show the setup wizard for good few seconds then just crash | 13:20 |
| mal | hmm, wondering what is causing the crash | 13:28 |
| mal | asriel[m]: just to be sure, which branch of libhybris do you use, I assume the one mentioned in faq, i.e. master branch | 13:29 |
| asriel[m] | It's master | 13:29 |
| asriel[m] | Also for the firmware error, would it be better to just add the paths to the firmware path like this? https://github.com/ubports-santoni/android_kernel_xiaomi_santoni/commit/72084768fe9a169b8505db7781d544cfcb828087 | 13:31 |
| mal | asriel[m]: that is also possible if needed | 13:43 |
| mal | asriel[m]: usually i's enough to add some symlink | 13:43 |
| mal | but if the firmware files are in many places the patching kernel might be the easier way | 13:44 |
| *** asriel[m] is now known as asriel | 13:47 | |
| asriel | Yep, they're in /vendor/firmware and /firmware/image | 13:48 |
| mal | asriel: oh, then you should try to replace the symlink I mentioned earlier with "ln -s /firmware/image /lib/firmware" | 13:50 |
| mal | could you give list of files in those folders?= | 13:50 |
| asriel | But then that'll also took away 3D because the Adreno firmwares are in /vendor/firmware | 13:51 |
| mal | ah | 13:51 |
| asriel | The paths I could found contain very important firmware files for modem, fingerprint and some other stuffs are in /firmware/image | 13:51 |
| mal | give ls -lR output of both /vendor/firmware and /firmware/ | 13:51 |
| mal | yes, might be easier to patch kernel then | 13:52 |
| mal | try that | 13:52 |
| asriel | There is also this /system/etc/firmware/wlan/prima/WCNSS_cfg.dat too which probably is the configuration | 13:53 |
| asriel | <mal "give ls -lR output of both /vend"> https://paste.ubuntu.com/p/svKFGt7YmJ/ | 13:55 |
| mal | you can try to add symlink "ln -s /system/etc/firmware /etc/firmware" some devices need that | 13:55 |
| mal | what do you have in /system/etc/firmware? some config stuff only or wlan firmware? | 13:55 |
| asriel | <mal "what do you have in /system/etc/"> only that file I mentioned above: /system/etc/firmware/wlan/prima/WCNSS_cfg.dat | 13:56 |
| mal | ok | 13:56 |
| mal | maybe try patching kernel with teh firmware loading path change and see how it works after that | 13:57 |
| mal | I hate it when android now has firmware files in so many places | 13:59 |
| mal | on some devices there are even more than two places | 13:59 |
| asriel | Heey, the patch got rid of the firmware errors | 14:33 |
| asriel | And the UI seems to be stable now | 14:33 |
| asriel | Sounds also works too! \o/ | 14:34 |
| deathmist | hey, any known reason I can't overwrite (disable) vendor time_daemon and qti services from an rc file in /usr/libexec/droid-hybris/system/etc/init/ ? they start regardless unless I edit the source /vendor/etc/init/hw/init.qcom.rc file https://pastebin.com/hDvnbpp0 | 15:32 |
| deathmist | here's the "disabled_services.rc" file I've made https://pastebin.com/KSTnDEzm | 15:39 |
| electro5751 | elros34, CONFIG_WATCHDOG_NOWAYOUT is not set | 15:52 |
| electro5751 | i try to disable this config but i'm note sure that do anything | 15:53 |
| electro5751 | i think that don't change my problem, i try | 15:53 |
| mal | asriel: very good | 15:54 |
| mal | deathmist: time_daemon is probably in /init.rc which is loaded first | 15:55 |
| mal | deathmist: there was something about some things being loaded in wrong order | 15:56 |
| deathmist | mal: just checked, neither of them are in /init.rc | 15:56 |
| mal | deathmist: so it's in /vendor then? | 15:59 |
| deathmist | yeah, /vendor/etc/init/hw/init.qcom.rc | 15:59 |
| mal | deathmist: probably the issue is that the file is imported via /init.rc and that is loaded first, not sure though | 15:59 |
| T42 | <elros34> check init.target.rc or all /init*rc | 16:00 |
| deathmist | it totally is added by /init.rc, in logs it's the second thing it adds to import list, so I should import it at the end of my disabled_services.rc file or something similar? | 16:02 |
| T42 | <elros34> elector5751: I saw it in yor earlier defconfig changes so how it looks like now? paste: "zcat /proc/config.gz". Also check what happen when you type: echo "V" > /dev/watchdog | 16:03 |
| deathmist | what I just said earlier seems to have worked perfectly, thanks for the tips :) | 16:15 |
| asriel | alright so without internet this is getting bored | 16:21 |
| asriel | does anyone knows how to get wifi? | 16:22 |
| asriel | This device uses WCNSS for wifi, and I already found a way to do it, however just asking if anyone here have better ways to do that | 16:25 |
| mal | asriel: what kernel driver does it use for wifi? | 16:26 |
| mal | the config option | 16:26 |
| asriel | CONFIG_WCNSS_CORE | 16:27 |
| asriel | or CONFIG_WCNSS_* | 16:28 |
| mal | asriel: can you give a link to the kernel source repo, I don't remember what device that | 16:29 |
| mal | *that is | 16:29 |
| asriel | <mal "asriel: can you give a link to t"> https://github.com/sailfish-santoni/android_kernel_xiaomi_msm8937/blob/hybris-15.1 | 16:30 |
| electro5751 | elros34 : how have full output with this cmd -> zcat /proc/config.gz | 16:30 |
| electro5751 | https://dpaste.de/QHDm | 16:31 |
| electro5751 | elros34 : my defconfig have already CONFIG_WATCHDOG | 16:32 |
| asriel | <electro5751 "elros34 : how have full output w"> pipe the command | 16:32 |
| asriel | zcat /proc/config.gz > kernel_config | 16:32 |
| asriel | and you'll have the full kernel config output in the new file | 16:33 |
| electro5751 | thanks | 16:33 |
| electro5751 | how to download file with telnet | 16:33 |
| asriel | i would say netcat is an option but sfos does not come with it | 16:33 |
| electro5751 | ok | 16:34 |
| asriel | However you have SSH if I'm sure so just scp it | 16:34 |
| electro5751 | ok, ssh port 22 ? | 16:34 |
| asriel | yeah the default port, and the same ip used for telnet | 16:34 |
| asriel | Unless you're booting to a recovery environment then this of course doesn't work I'm sure | 16:35 |
| electro5751 | avec user ? | 16:35 |
| electro5751 | yes, adb with recovery TWRP | 16:35 |
| asriel | user is nemo | 16:35 |
| electro5751 | ok thk | 16:35 |
| electro5751 | ssh: connect to host 192.168.2.15 port 22: Connection refused | 16:35 |
| asriel | Oh | 16:36 |
| asriel | I thought you're using hybris-boot | 16:36 |
| electro5751 | hum, yes | 16:36 |
| Danct12 | asriel: what about echo sta to the /sys/module/wlan/parameters/fwpath? | 16:39 |
| asriel | <Danct12 "asriel: what about echo sta to t"> I already knew that, just it may not be the best way? | 16:39 |
| electro5751 | elros34 : echo "V" > /dev/watchdog -> device display battery picture | 16:40 |
| Danct12 | bitrvmpd used it in his sailfishos port and personally it's the best way | 16:40 |
| electro5751 | mal, CONFIG_WATCHDOG must be enable or disable ? | 16:47 |
| mal | depends on device | 16:50 |
| electro5751 | i have again access to telnet | 16:50 |
| electro5751 | with echo "V" > /dev/watchdog | 16:50 |
| electro5751 | but what can i do again | 16:50 |
| electro5751 | i don't know how debug step by step | 16:51 |
| mal | asriel: no need for hci smd in hybris-15.1 ports | 16:51 |
| mal | asriel: instead you should add CONFIG_BT_HCIVHCI=y | 16:52 |
| mal | asriel: you seem to have pronto_wlan in defconfig | 16:53 |
| mal | asriel: check faq related to that | 16:53 |
| electro5751 | anyone can help me to debug the startup of a device ? i have a telnet access | 16:55 |
| Danct12 | got sfos booted | 16:56 |
| Danct12 | anyway, i dont have a modem | 16:56 |
| Danct12 | here is the modem log | 17:03 |
| Danct12 | https://paste.ubuntu.com/p/z4qKT7w8vV/ | 17:03 |
| Danct12 | logcat* | 17:03 |
| mal | Danct12: what device? | 17:04 |
| Danct12 | the same device as asriel since i work on the port too | 17:04 |
| mal | ok | 17:04 |
| mal | Danct12: in case you haven't done already I suggest you add ofono-ril-binder-plugin and bluebinder to patterns like here https://github.com/mer-hybris/droid-config-sony-nile/blob/master/patterns/jolla-hw-adaptation-pioneer.yaml#L14 | 17:06 |
| mal | those are needed for modem and bluetooth on hybris-15.1 based ports | 17:07 |
| Danct12 | oh so the hci smd patch isn't needed | 17:07 |
| Danct12 | just only CONFIG_BT_HCIVHCI as you mentioned above? | 17:07 |
| mal | Danct12: yes, in hybris-15.1 we do bluetooth via android hal | 17:08 |
| mal | of course hci smd is possible if the chip works with that, but that depends completely which bt chip the device has | 17:08 |
| mal | bluebinder should usually work on all (although some devices might some rfkill related patching of kernel but well see about that later) | 17:09 |
| Danct12 | as for the wifi, that can also working on hybris-15.1 by compiling them as a module? | 17:09 |
| Danct12 | wcnss* | 17:10 |
| mal | Danct12: faq mentions what to build as module, in case on pronto_wlan on that driver needs to be a module | 17:11 |
| deathmist | Danct12: for BT I also needed BT_BNEP enabled on top of BT_HCIVHCI in case that alone doesn't work :) | 17:11 |
| mal | yes, good point | 17:11 |
| mal | check that also | 17:11 |
| mal | deathmist: thata device has BNEP already enabled | 17:12 |
| mal | Danct12: in your case you should make CONFIG_PRONTO_WLAN=m for wlan and then try the wlan systemd service for modprobing it | 17:13 |
| mal | it usually should be enough | 17:13 |
| deathmist | mal alright, nice :p I finally also fixed cpusets by modifying /init.rc and some vendor init.target.performance.rc init files, they were both using "cpus" & "mems" instead of the "cpuset.cpus" and "cpuset.mems" paths on all the files for me on SFOS | 17:15 |
| Danct12 | ubports seems to still use "cpus" instead of "cpuset.cpus" | 17:20 |
| electro | repo error : fatal: unable to auto-detect email address | 17:29 |
| electro | with this command : repo init -u git://github.com/mer-hybris/android.git -b hybris-14.1 | 17:29 |
| Danct12 | you need to set user.name and user.email in git | 17:30 |
| electro | i have already done this | 17:30 |
| electro | !! | 17:31 |
| merbot | electro: Error: "!" is not a valid command. | 17:31 |
| mal | electro: maybe you need set those like here https://git-scm.com/book/en/v2/Getting-Started-First-Time-Git-Setup the "Your identity part" | 17:31 |
| electro | i have already done this | 17:31 |
| electro | $ git config --global user.name "John Doe" | 17:32 |
| electro | $ git config --global user.email johndoe@example.com | 17:32 |
| Danct12 | maybe you need to quote the email address | 17:32 |
| electro | o yes maybe | 17:33 |
| mal | does the $HOME/.gitconfig look ok | 17:33 |
| electro | humm, no ! what is quote ? | 17:33 |
| Danct12 | "this is quote" | 17:33 |
| electro | look like good for file | 17:34 |
| Danct12 | also please consider using an actual user.email and user.name for uploading your work to git | 17:34 |
| deathmist | how important are the /init.zygote*.rc files and /sbin/charger symlink on /charger for Sailfish? just asking whether I should include them in dcd straggler files | 17:35 |
| mal | deathmist: any file that is not packaged (i.e. build complains about those) should be in straggler files | 17:36 |
| mal | some are not used at all though | 17:36 |
| deathmist | build doesn't complain, so I assume unneeded | 17:36 |
| mal | deathmist: then where do you see those? | 17:37 |
| mal | deathmist: you only care about that the droid-hal build complains about, straggler files are for the files that are no packaged by default and it will complain about those in build log | 17:37 |
| deathmist | just wandering around the device root dir in out/ and wanted to be sure | 17:38 |
| mal | well droid-hal build ignores some things by default | 17:39 |
| mal | so you only check the build log for unpackaged files | 17:39 |
| electro | sorry but why i have this reponse | 17:44 |
| electro | fatal: unable to auto-detect email address (got 'sailfish-dev@jopcbureau.(none)') | 17:44 |
| mal | maybe exit the build env and go back | 17:45 |
| T42 | <elros34> electro: about whole log: increase buffer in your terminal to infinity | 17:45 |
| T42 | <elros34> how fast does it reboot after writing "magic close" V to watchdog | 17:46 |
| electro5751 | elros34 : ok thk | 17:46 |
| electro5751 | humm, don't understand elros34 the last sentence | 17:47 |
| electro5751 | when i writed V to watchdof -> my device reboot ? | 17:47 |
| electro5751 | i have don't see this but yes telnet was again available | 17:47 |
| T42 | <elros34> how many seconds it takes to lose telnet connection after you write "V"? 1s, 5s, 15s? | 17:50 |
| electro | elros34 -> i don't loose telnet connection after write "V" ! | 17:51 |
| electro | i loose telnet with the cmd "echo continue, ..." | 17:51 |
| electro | but the phone displayed battery picture | 17:51 |
| electro | no message on dmesg | 17:51 |
| electro | no more message on dmesg | 17:51 |
| Danct12 | sounds like it drops it to charging mode | 17:52 |
| Danct12 | anyway to make build_packages.sh runs on multiple threads? | 17:53 |
| electro5751 | this cmd : echo "continue" > /init-ctl/stdin | 17:53 |
| electro5751 | close the telnet connection between my computer and the device | 17:53 |
| Danct12 | everything is built using one thread :( | 17:53 |
| deathmist | Danct12: I was wondering why it's so slow, guess a R7 3700X won't help much then :/ thankfully you almost always only have to rebuild with -d and -c | 17:55 |
| Danct12 | yeah because everything is done on a single thread :\ | 17:56 |
| electro | what is this elros34 ? "magic close" V to watchdog | 17:58 |
| electro5751 | this cmd : echo "V" > /dev/watchdog | 17:59 |
| electro | i don't have ~/.config/ in my new user sailfish-dev | 18:06 |
| electro | how can i configure this ? | 18:06 |
| Danct12 | mal: right, so the pronto wlan module thing works | 18:39 |
| Danct12 | i modprobe it by hand | 18:39 |
| Danct12 | bluetooth still doesn't work | 18:39 |
| Danct12 | I have HCIVHCI and BNEP enabled | 18:39 |
| Danct12 | systemctl status of bluebinder shows this May 22 12:50:39 Sailfish systemd[1]: Failed to start Simple proxy for using android binder based bluetooth through vhci.. | 18:42 |
| electro | elros34, i'm trying to compile all things on another computer, i will see the result | 18:44 |
| deathmist | Danct12 : check full journalctl, maybe grep for "blue" to see what exactly is failing | 18:47 |
| * Danct12 sent a long message: < https://matrix.org/_matrix/media/v1/download/matrix.org/iaYEkoqyxyuqJzwddxbPGWXS > | 18:48 | |
| Danct12 | I think this is the cause | 18:49 |
| mal | Danct12: did you verify that vhci is enabled now in kernel | 18:49 |
| Danct12 | sh-3.2# zcat /proc/config.gz | grep VHCI | 18:51 |
| Danct12 | CONFIG_BT_HCIVHCI=y | 18:51 |
| Danct12 | BNEP is also enabled too | 18:52 |
| Danct12 | CONFIG_BT_BNEP=y | 18:52 |
| Danct12 | CONFIG_BT_BNEP_MC_FILTER=y | 18:52 |
| Danct12 | CONFIG_BT_BNEP_PROTO_FILTER=y | 18:52 |
| mal | ok, do you see anything interesting in logcat? | 18:52 |
| * Danct12 sent a long message: < https://matrix.org/_matrix/media/v1/download/matrix.org/LPeZxAessCiIjkQShbGuTBwa > | 18:53 | |
| mal | Danct12: can you pastebin whole logcat | 18:54 |
| Danct12 | sure | 18:58 |
| Danct12 | https://paste.ubuntu.com/p/bWXVYNTkCC/ | 19:00 |
| Danct12 | there are a bunch of time services errors | 19:01 |
| mal | Danct12: any luck with modem? | 19:01 |
| Danct12 | <mal "Danct12: any luck with modem?"> nope as it seems, there is no /dev/socket/rild too | 19:02 |
| Danct12 | but rild is running with `-c 2` | 19:02 |
| mal | Danct12: only rild with -c 2? no normal one? | 19:05 |
| Danct12 | radio 2714 0.0 0.5 81340 14788 ? Sl 12:39 0:00 /vendor/bin/hw/rild -c 2 | 19:06 |
| Danct12 | radio 2751 0.0 0.5 84496 14920 ? Sl 12:39 0:00 /vendor/bin/hw/rild | 19:06 |
| mal | ok, btw, hybris-15.1 uses modem differently from others | 19:06 |
| electro | why this : error: src refspec hybris-14.1 does not match any. | 19:07 |
| mal | which repo? | 19:08 |
| mal | show your local manifest | 19:08 |
| electro | it's a git remote | 19:08 |
| electro | git remote add electro575 https://github.com/user/android_kernel_samsung_t0lte_bureau | 19:09 |
| mal | so that command gives the error or some other? | 19:09 |
| electro | it's during this cmd : git push electro575 hybris-14.1 | 19:10 |
| electro | forget git checkout | 19:11 |
| mal | Danct12: I suggest you do chapter 13.2 and 13.3 from hadk pdf to get camera and stuff working and it will also remove the sensorservice messages from logcat | 19:12 |
| electro | it's okey mal, sorry | 19:12 |
| electro | qemu: Unsupported syscall: 384 | 19:19 |
| Danct12 | that one is fine | 19:19 |
| electro | after launch this cmd : | 19:19 |
| Danct12 | i got it too | 19:19 |
| electro | sdk-assistant create $VENDOR-$DEVICE-$PORT_ARCH http://releases.sailfishos.org/sdk/targets/Sailfish_OSarget-armv7hl.tar.7z_T | 19:19 |
| mal | those qemu messages are normal | 19:21 |
| mal | Danct12: about modem, do you see anything from ofono in journalctl? | 19:23 |
| Danct12 | i'll obtain logs tomorrow though | 19:24 |
| Danct12 | going to get some sleep anyway | 19:24 |
| mal | ok | 19:25 |
| electro | mal, Package 'ssu-kickstart-configuration' not found. | 19:41 |
| electro | when i launch : rpm/dhd/helpers/build_packages.sh | 19:42 |
| mal | faq, didn't you fix that already once? | 19:42 |
| electro | yes but i restart all config in another computer | 19:42 |
| mal | why? | 19:42 |
| electro | maybe i try the solution | 19:42 |
| electro | i don't have another solution in this moment | 19:42 |
| electro | nobody can help me to debug my device | 19:43 |
| mal | you cannot expect everything to be solved in one day | 19:43 |
| mal | there is no reason to start from beginning | 19:44 |
| electro | i don't know | 19:44 |
| electro | but have you an idea for this : https://dpaste.de/9pcP | 19:44 |
| mal | I told you already, read faq | 19:45 |
| electro | yes | 19:45 |
| electro | in the pdf ? | 19:45 |
| electro | or the website | 19:45 |
| mal | the link in channel topic | 19:45 |
| mal | hadk pdf is not faq | 19:45 |
| electro | ok | 19:46 |
| mal | but like I said, there is no reason to start from beginning, just copy the things you did already and continue | 19:46 |
| electro | continue to do what ? i don't know where i must start to debug | 19:47 |
| mal | first by giving all logs from device and telling what you have done so far | 19:48 |
| mal | so how does the device behave? | 19:48 |
| electro | humm, ok, bootloop without a ligne debug2 | 19:51 |
| electro | with this line, i can use telnet | 19:51 |
| electro | try to continue but when i launch the command to continue, the telnet session is broken | 19:51 |
| mal | electro: so did you disable droid-hal-init as I asked? | 19:53 |
| mal | electro: do you lose telnet right after that or after a moment? | 19:53 |
| electro | right after | 19:54 |
| electro | disable droid-hal-init, no | 19:54 |
| electro | how disable this ? | 19:54 |
| electro | how is created this file "Jolla-@RELEASE@-t0lte-@ARCH@.ks" ? | 19:55 |
| mal | to disable droid-hal-init run this on device: ln -s /dev/null /etc/systemd/system/droid-hal-init.service | 19:55 |
| electro | hum, i have tried this one moment | 19:57 |
| mal | also show output of "cat /proc/cmdline" from device | 19:57 |
| electro | ok | 19:57 |
| mal | and output of "zcat /proc/config.gz | grep SELINUX" | 19:57 |
| electro5751 | ln: creating symbolic link `/etc/systemd/system/droid-hal-init.service': File exists | 19:57 |
| mal | then check "ls -l /etc/systemd/system/droid-hal-init.service" to see if it's correct | 19:58 |
| electro5751 | /etc/systemd/system/droid-hal-init.service -> /dev/null | 19:59 |
| mal | ok | 19:59 |
| mal | is the also user@100000.service in the same folder | 19:59 |
| electro5751 | https://dpaste.de/4R0E | 19:59 |
| mal | pointing to /dev/null | 19:59 |
| electro5751 | ok | 20:00 |
| mal | so based on that /proc/cmdline you once again didn't do what I asked earlier | 20:00 |
| electro5751 | what ? i have paste the /proc/cmdline for you | 20:00 |
| electro5751 | https://dpaste.de/nCS9 | 20:01 |
| mal | electro5751: if you did what I asked day or two ago you should have selinux=0 in that output | 20:01 |
| electro5751 | yes | 20:02 |
| electro5751 | so disable all ? | 20:02 |
| mal | NOOOOOOOOO | 20:02 |
| electro5751 | but i have write selinux=0 in the cmd | 20:02 |
| mal | electro5751: show where did you add it | 20:02 |
| electro5751 | i don't remember the file | 20:03 |
| mal | probably not the correct place then | 20:03 |
| mal | also what did you mean by "disable all", disable what? | 20:03 |
| electro5751 | CONFIG_SELINUX=n | 20:04 |
| electro5751 | ... | 20:04 |
| electro5751 | all CONFIG_SELINUX=n | 20:04 |
| mal | of course not, where the hell did you get that idea? | 20:04 |
| mal | I told you nothing about defconfig | 20:04 |
| electro5751 | yes i know | 20:04 |
| electro5751 | i show you in 10 s | 20:04 |
| electro5751 | i need this too or not ? | 20:05 |
| electro5751 | http://repo.merproject.org/obs/home:/elros34:/glibc/sailfishos_3.0.3.9/ | 20:05 |
| mal | I really cannot trust anything you have done, you might have disabled many things that might be needed or done some random changes | 20:05 |
| mal | well yes if elros34 told you to use it | 20:06 |
| electro5751 | chapter 9.3.1 about bootloops | 20:07 |
| electro5751 | selinux=0 | 20:08 |
| electro5751 | to | 20:08 |
| electro5751 | your | 20:08 |
| electro5751 | kernel | 20:08 |
| electro5751 | command | 20:08 |
| electro5751 | line | 20:08 |
| electro5751 | (usually | 20:08 |
| electro5751 | in | 20:08 |
| electro5751 | BOARD_KERNEL_CMDLINE | 20:08 |
| electro5751 | under $ANDROID_ROOT/device/$VENDOR/*/BoardConfig*.mk) | 20:08 |
| electro5751 | O sorry | 20:08 |
| electro5751 | mal, BOARD_KERNEL_CMDLINE := console=ttySAC2,115200 selinux=0 | 20:11 |
| electro5751 | in the file : ~/hadk/device/samsung/smdk4412-common/BoardCommonConfig.mk | 20:11 |
| mal | so did you rebuild kernel after adding that? | 20:16 |
| electro | yes mal | 20:19 |
| electro | a website said too audit=0 | 20:21 |
| electro | it's necessary ? | 20:21 |
| electro | i will write a text file with changes | 20:23 |
| mal | no need to add audit=0 | 20:33 |
| mal | if you say you added selinux=0 and rebuilt kernel then why is it not in /proc/cmdline, either you did something wrong or didn't flash new kernel | 20:34 |
| electro | i have add selinux=0 maybe in the wrong file, i will see that | 20:45 |
| electro | must have this dependencies ? | 20:45 |
| electro | sudo apt-get install bsdmainutils rsync vim unzip imagemagick software-properties-common python-software-properties | 20:46 |
| electro | and this : sudo apt-get install openjdk-8-jdk | 20:46 |
| electro | mal | 20:46 |
| mal | I can't remember, do what hadk pdf says | 20:46 |
| electro | ok | 20:51 |
| electro | why i have this | 20:54 |
| electro | with mic | 20:54 |
| electro | URLGrabber error: http://repo.merproject.org/obs/nemo:/devel:/hw:/common/sailfish_latest_armv7l/repodata/repomd.xml - e.errno: 14 | 20:54 |
| mal | electro: you did something wrong, arch is armv7hl not armv7l | 20:56 |
| electro | arf | 20:57 |
| mal | no idea where you did the mistake | 20:57 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!