Friday, 2020-05-08

T42冯鑫淘 %lastname% was added by: 冯鑫淘 %lastname%01:42
UmeaboyHi! I'm now following the HADK pdf and so far so good, but I was wondering..... if I use the hybris branch for aosp 8.1, would I still need to do the kernel check when hybris-hal is built?02:40
r0kk3rzwell yeah02:51
UmeaboyOK. Good to know.03:05
Umeaboyr0kk3rz: Here are all the things that needs to be fixed. Some I know what value to set, but some I don't. Can you help me fix these? https://pastebin.com/qF8hTWrj03:44
UmeaboyWhat's good about enabling kernel namespaces for systemd-nspawn containers ?03:45
UmeaboyI know it says optional, but the HADK pdf tells you to rinse lather and repeat until there are no errors or warnings.03:45
UmeaboyAnd.........If I open the defconfig for maple and search for CONFIG_IPC_NS I get no hit.03:47
UmeaboyShould it give a warning for something that isn't even there?03:48
UmeaboyIs it safe to add it to the defconfig?03:48
UmeaboyHow can I check if bluez is used as a bluetooth stack?04:17
UmeaboyI need to set CONFIG_BT_RFCOMM to either y or !04:17
UmeaboyI guess y is prefered.04:17
r0kk3rzignore bt for now04:17
r0kk3rzandroid defconfigs dont have everything usually, so just add a line with the required config code04:18
UmeaboyOK.04:19
UmeaboyI did # CONFIG_BT_RFCOMM is not set04:19
UmeaboyIs that alright to do?04:19
r0kk3rzit wont do anything04:19
UmeaboyOK.04:19
r0kk3rzall the sonys have broadcomm bt if i remember04:20
r0kk3rzbut with newer ones you'll be using bluebinder anyway04:20
r0kk3rzso it should have all the config flags it needs by default04:20
UmeaboyOK.04:35
UmeaboyHow do I know if the device uses watchdogs with dsme ?04:35
UmeaboyI see that it's good to have it, but I don't know if this is valid for maple.04:36
r0kk3rzturn it on04:41
UmeaboyDo I need to do make clean or make clobber each time before I do make hybris-hal?11:29
UmeaboyAfter fixing the warnings/errors.11:30
r0kk3rzno11:31
UmeaboyIt seems like I got the same result after I rebuilt with the changes.11:33
r0kk3rzmaybe you did it wrong11:36
UmeaboyHere are the things I changed in the defconfig: https://pastebin.com/zie6b1z111:42
UmeaboyAnd the first issue still says WARNING: CONFIG_NETPRIO_CGROUP is invalid11:43
UmeaboyIt is unset11:43
UmeaboyIt is NOT unset.11:43
UmeaboyLine 10011:43
UmeaboyIt clearly says y.11:43
kernlingUmeaboy, what kernel version?11:46
Umeaboy4.411:46
kernlingUmeaboy, that config was renamed: https://github.com/torvalds/linux/commit/86f8515f9721fa171483f0fe0391968fbb949cc911:48
kernlingwould be good to patch mer-kernel-check i guess11:48
UmeaboySo I just replace the current line with CONFIG_CGROUP_NET_{PRIO,CLASSID}=y ?11:50
T42<edp_17> @Umeaboy: Just an idea. Have you checked the depencies of that CONFIG_NETPRIO_CGROUP? There might some other flags need to be set/unset and that's why the compiler unsets it.12:35
UmeaboyHow can I check dependencies?12:41
deathmistUmeaboy: search on https://cateee.net/lkddb/web-lkddb/ or run `menuconfig` / whatever you prefer inside the kernel directory with a proper cross-compiler setup (e.g. in habuild env)13:10
Umeaboydeathmist: It says that it's found in 3.3–3.13 and I'm building 4.413:16
UmeaboySo it doesn't make sense.13:16
deathmistwhat is?13:16
UmeaboyCONFIG_NETPRIO_CGROUP13:16
Umeaboyhttps://cateee.net/lkddb/web-lkddb/NETPRIO_CGROUP.html13:16
deathmistas kernling pointed out before already, that config option got renamed to `CGROUP_NET_PRIO` starting with kernel release 3.1413:17
UmeaboyYeah, but that still doesn't explain why Sony didn't change that in this kernel though.13:27
deathmistI don't understand what the issue is, if you set `NETPRIO_CGROUP=y` and it dissapears from final generated .config -> it's working as intended because the option is deprecated in favor of `CGROUP_NET_PRIO`13:32
Umeaboydeathmist: So, I should delete the line with CGROUP_NET_PRIO and set NETPRIO_CGROUP=y and then rebuild?13:56
T42<ankaos> CONFIG_SECURITY_SELINUX_BOOTPARAM_VALUE=0 maybe fix my problem14:01
UmeaboyI had that set to y.14:18
UmeaboyI mean for CONFIG_SECURITY_SELINUX_BOOTPARAM14:19
UmeaboyGotta go for now.14:28
deathmistmal: any idea why following https://jolla.zendesk.com/hc/en-us/articles/203548266#usb-inst on 3.3 doesn't do anything and gives me the following journal? https://pastebin.com/raw/5bSt2bGH14:34
maldeathmist: does your device have configfs usb?14:36
deathmistyep14:36
malnot sure if usb tethering works with that14:37
deathmistit looks to be broken on LOS too, explains why it wouldn't work on SFOS either way lol. checked all the functions in Android and only USB tethering mode reported empty on `readlink /config/usb_gadget/g1/configs/b.1/f1`15:04
deathmistnvm it works fine, a `vendor.usb.rndis.func.name` property was empty on Android because the WLAN driver didn't load, found out for some reason setting up some of the USB controller configuration is done `on property:wlan.driver.status=ok`..... what kind of logic is that?16:19
deathmisteither way that should help on SFOS as well because `getprop wlan.driver.status` currently returns nothing for me16:20
deathmistnvm again, remembered all of the USB stuff is completely different in SFOS16:25
deathmistunder `/config/usb_gadget/g1/functions/` I see `ffs.mtp`, `mass_storage.usb0` & `rndis.usb0`. it uses `mass_storage.usb0` when in "Charging only" mode for whatever reason, shouldn't the `rndis.usb0` function also work the other way around for internet sharing?16:28
*** redj_ is now known as redj16:38
piggzjusa: hi, could you PR your PP P changes so we can merge them .... they work reasonably well, i was able to make a call17:44
T42<edp_17> @lbt: Hi, can you create an OBS account for me, please? (User: edp17, email: edp17@pm.me) Thanks.18:48
T42Paul El %lastname% was added by: Paul El %lastname%20:30

Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!