mal | yay | 10:28 |
---|---|---|
kimmoli | !hello | 10:29 |
merbot | kimmoli: Error: "hello" is not a valid command. | 10:29 |
ghosalmartin | morning :) | 11:19 |
ghosalmartin | hmm if i build bootimage by it self it breaks, but it completes as part of hybris-boot...is there a chance hybris-boot is actually failing silently? | 11:59 |
ghosalmartin | mal: when you disabled selinux in system/core. did that stop the relience on libsepol? | 12:01 |
mal | ghosalmartin: are you sure you have the correct busybox, i.e. the prebuilt one used for android 8 | 12:05 |
ghosalmartin | mal...no | 12:05 |
ghosalmartin | is that now required :P | 12:05 |
mal | ghosalmartin: check external/busybox to see which one you have | 12:05 |
mal | manifest has it correctly but not sure if your repo has been synced | 12:06 |
ghosalmartin | mal: 32bit one. i synced using the hybris-15.1 manifest? | 12:06 |
mal | so that points to mer-hybris/android_external_busybox_prebuilt ? | 12:06 |
mal | I mean the local copy | 12:06 |
ghosalmartin | mal: yes | 12:06 |
mal | ok | 12:07 |
ghosalmartin | am just trying to build a boot android kernel first | 12:07 |
ghosalmartin | mal: also theres some weird thing with the pixel phones that the recovery is in the same partition as boot | 12:10 |
mal | ghosalmartin: also some older sonys had recovery built into the kernel | 12:11 |
mal | not sure about newer sonys | 12:11 |
ghosalmartin | mal: hmm wonder why their moving to this kind of arch. it seemed much cleaner just having a separate partition | 12:11 |
ghosalmartin | could be that they want 2 boot partitions and 2 systems partitions for completely seamless OTA | 12:12 |
mal | ghosalmartin: sorry that I haven't made the test build yet for your device, I have some issues with disk space, I can only have out folder of one build at the time | 12:15 |
ghosalmartin | mal: its alright, this is a new device :P | 12:15 |
mal | oh, not the same annoying one | 12:15 |
ghosalmartin | pretty much :P its the Pixel Xl this time | 12:16 |
ghosalmartin | marlin | 12:16 |
mal | ok | 12:16 |
mal | wondering if I should rebase hybris-15.1 | 12:16 |
ghosalmartin | think its outta date? | 12:20 |
mal | probably not too much but can't be sure, hybris-15.1 was based on r26 but lineage now uses r43 aosp as base | 12:21 |
ghosalmartin | mal: ahh i dont mind rebasing it, got some spare time today | 12:21 |
ghosalmartin | its complaining about some libsepol things when attempting a build, so could be relate | 12:22 |
mal | ghosalmartin: need to think how to handle that, a rebase or merge, rebase is cleaner probably | 12:22 |
ghosalmartin | rebase is always cleaner. i actively avoid merging nowadays unless others | 12:23 |
mal | ghosalmartin: for me the rebase is easier as I can just force push the branches :) | 12:23 |
ghosalmartin | there is that :P | 12:24 |
ghosalmartin | mal: what would be the easiest way to get the changes in? make them on a fork and get that taken in? | 12:31 |
mal | then it would be a merge | 12:32 |
ghosalmartin | like this https://github.com/mer-hybris/android_build/pull/10 | 12:33 |
ghosalmartin | but isnt there a rebase option when selecting the merge? | 12:33 |
ghosalmartin | always been curious if that button actually doesnt anything useful | 12:34 |
ghosalmartin | yeah seems to work okay according to this quick test https://github.com/ghosalmartin/rebase_test/commits/master | 12:40 |
r0kk3rz | yeah there is, i think i tried it once | 12:40 |
ghosalmartin | it seems to work well. but am curious how itd work in a large team 8-10 people | 12:43 |
ghosalmartin | would we start to get some crazy conflicts coming in from it. am trying convert the team to rebasing...still :P | 12:43 |
ghosalmartin | also anyone know what happens if /boot isnt declared inside the fstab? do I just add an entry? | 12:44 |
Mikaela | T4 is now storing Telegram attachments for 24 hours, but I cannot make any stability or uptime or anything promises, I am not paid for this or anything, and I may not be able to remove confidential accidental uploads immediately, so please refrain from making those. | 12:47 |
ghosalmartin | i have a nice rebasing script that pulls in all the appropriate remotes, and pushes to mine after its done | 12:56 |
ghosalmartin | the only limiting factor is now my internet speed :P | 12:56 |
echo_salik | hi mal | 12:57 |
echo_salik | hi r0kk3rz | 12:58 |
mal | ghosalmartin: nice | 12:58 |
echo_salik | how are you... | 12:58 |
ghosalmartin | https://pastebin.com/NuUPrwZW boom | 12:59 |
ghosalmartin | if you have no conflicts its a slam dunk | 12:59 |
mal | not very well, almost getting a headache today also and now I need to go and complain about a harddisk to a store, it broke down completely yesterday and it still has warranty | 13:00 |
ghosalmartin | hopefully no data loss | 13:01 |
echo_salik | aww... good luck... | 13:01 |
mal | ghosalmartin: that was part of a raid 5 setup so should be fine | 13:05 |
mal | ghosalmartin: if that would have been without raid then all data would have been lost, it doesn't even detect properly anymore | 13:06 |
ghosalmartin | mal: ahh good good, raid never ceases to amaze me | 13:07 |
ghosalmartin | mal: before i open the prs, would using githubs rebase be a good way to handle the rebase to latest lineageos? if so all rebased and ready to merge | 13:18 |
ghosalmartin | or wanna test em first :P | 13:18 |
mal | I'll have a look and discuss with others | 13:19 |
ghosalmartin | okays :) | 13:19 |
ghosalmartin | btw the hybris-boot-part. is that actually required to boot to stage 1? | 13:21 |
mal | which one? | 13:21 |
ghosalmartin | when building hybris boot its searching for /boot. is that required to get to telnet. or is that only for systemd? | 13:23 |
ghosalmartin | this error ./hybris/hybris-boot/Android.mk:71: error: There should be a one and only one device entry for HYBRIS_BOOT_PART and HYBRIS_DATA_PART. | 13:23 |
ghosalmartin | HYBRIS_BOOT_PART doesnt have a /boot in the fstab for whatever reason | 13:23 |
mal | on one device I added /dev/null as /boot in fstab https://github.com/mlehtima/android_device_semc_msm7x30-common/commit/e1a57373b370ae7c4a3bc4541c09e751c567f6da | 13:24 |
ghosalmartin | that looks perfect ta :) | 13:26 |
ghosalmartin | boom am into telnet by disabling dm-verity | 13:41 |
ghosalmartin | woop woop :D | 13:41 |
mal | nice | 13:42 |
ghosalmartin | shame hte same didnt work with the mi mix 2, but ah well | 13:42 |
ghosalmartin | although it might be worth a shot, they both use the same hybris-15.1 codebase | 13:43 |
ghosalmartin | CONFIG_VT strikes again breaking yet another kernel | 14:39 |
mal | ghosalmartin: easy warranty change, I got a new disk right away | 14:43 |
ghosalmartin | mal: thats good. its nice when stores just sort it | 14:43 |
r0kk3rz | ghosalmartin: you have all the luck | 15:15 |
ghosalmartin | r0kk3rz, who knows it may get better :P | 15:17 |
r0kk3rz | or worse :) | 15:19 |
ghosalmartin | well atleast this device theres something to work with | 15:22 |
ghosalmartin | mal, is the current method of reaching ui just using the qpa-surfaceflinger plugin? | 17:04 |
mal | ghosalmartin: just building the plugin and changing the conf | 17:05 |
mal | ghosalmartin: assuming everything else is ok | 17:05 |
Mister_Magister | whatcha doing guys? | 17:05 |
ghosalmartin | mal: awesome :) thanks. am currently hoping CONFIG_VT isn't all that important | 17:05 |
mal | ghosalmartin: for 15.1 you need some things in config repo | 17:05 |
ghosalmartin | i can get to first stage kernel fine now | 17:06 |
ghosalmartin | i guess next battle of rootfs | 17:06 |
ghosalmartin | Mister_Magister, that Pixel XL | 17:06 |
Mister_Magister | ghosalmartin: oooohh with android8 right? | 17:06 |
ghosalmartin | Mister_Magister, yahh, its to new for cm14 job | 17:07 |
Mister_Magister | ghosalmartin: you are using cm as a base? | 17:08 |
ghosalmartin | Mister_Magister, nah lineageos | 17:08 |
mal | ghosalmartin: to get everything working on 15.1 needs some magic things and unmerged stuff | 17:08 |
Mister_Magister | yeah i mean ln xD | 17:08 |
Mister_Magister | mal: likeā¦ everything? | 17:08 |
ghosalmartin | mal: for now think id settle for a pulse...and i dont mean pulseaudio :P | 17:09 |
Mister_Magister | 100% everything? | 17:09 |
mal | well most things can be done: audio, calls, wifi, gnss and camera are all possible | 17:09 |
Mister_Magister | niiice | 17:09 |
ghosalmartin | mal: oh yeah did you get a chance to try out the bt through bindeR? | 17:10 |
mal | sensors also, depending on which method it uses | 17:10 |
mal | ghosalmartin: not yet, I should try that, I did some testing to see I can get nfc working | 17:10 |
mal | no luck with nfc yet though | 17:10 |
ghosalmartin | will we need nfc in sfos in the future? :P | 17:10 |
mal | maybe | 17:11 |
mal | I think it should not be too difficult, just that in android kernel the drivers don't use the linux way for communicating with userspace so we might need some custom middleware/Qt plugin | 17:12 |
mal | other option would be to change the driver in linux | 17:12 |
ghosalmartin | dont we need that for everything :P | 17:13 |
mal | ghosalmartin: can you check if your device has old style android hal for sensors or does it use the new hidl interface already? | 17:14 |
ghosalmartin | mal: how would I do that? | 17:14 |
mal | check if it has some sensors*.so in some android lib folder | 17:14 |
ghosalmartin | hmm according to this no https://github.com/TheMuppets/proprietary_vendor_google/tree/lineage-15.1/marlin/proprietary/lib | 17:15 |
ghosalmartin | but i need to take a proper look at the device i htink | 17:15 |
ghosalmartin | ah mal is this the one libsensorservicehidl.so | 17:17 |
mal | ghosalmartin: you should see something like android.hardware.sensors@1.0-impl.so | 17:20 |
Mister_Magister | mal: btw how ya feelin today | 17:20 |
ghosalmartin | if this port works am defo buying a sailfish to put sailfish on it :P | 17:20 |
mal | Mister_Magister: not very well, had some small headache most of the day | 17:21 |
Mister_Magister | maybe not enough water? | 17:21 |
Mister_Magister | try drinking more :P that helps with headache | 17:24 |
mal | it's just bad sleeping | 17:24 |
mal | I had it since I woke up | 17:25 |
ghosalmartin | anyone figured out an easy way to clone projects on obs? | 17:27 |
ghosalmartin | or do I just checkout one with osc and then push to another? | 17:27 |
mal | ghosalmartin: osc copypac? | 17:29 |
mal | I just have bunch of those for syncing devices | 17:30 |
ghosalmartin | mal: yeah I remember trying a script with that and a for each but had some weird errors | 17:30 |
mal | ghosalmartin: in case your device only has the hidl interface for sensors then I can create a sensorfw version for you, I know now quite well how to use the libgbinder so I can probably add the support in a couple of hours | 17:49 |
ghosalmartin | mal: ahh I see, for now lemme get it up and running first :P i havent even finished building hybris-hal yet | 17:50 |
mal | sure | 17:50 |
guhl | hi, does somebody have experience with assemble_vintf failing when building hybris-hal on hybris-15 -> http://paste.opensuse.org/20003405 | 19:32 |
guhl | somehow it does not know where the kernel sources are | 19:33 |
ghosalmartin | guhl, yooo | 21:26 |
ghosalmartin | were you been :P | 21:26 |
ghosalmartin | how did you get mi mix 2 to boot | 21:26 |
ghosalmartin | ive been tryna get it working with the latest lineageos but CONFIG_VT seems to corrupt the kernel to bits | 21:27 |
ghosalmartin | mal: when building libhybris for 15.1 was anything special required? its complaining it cant find sync.h but its in the devel? | 21:30 |
mal | ghosalmartin: https://github.com/krnlyng/libhybris-1/tree/android8-initial | 21:33 |
ghosalmartin | mal: ta :) | 21:37 |
mal | ghosalmartin: you'll also need these https://github.com/mer-hybris/droid-config-sony-nile/tree/master/sparse/usr/libexec/droid-hybris/system/etc/init | 21:40 |
mal | not related to libhybris but you still need those at some point | 21:41 |
ghosalmartin | mal: well fingers crossed I make it far that tonight :P | 21:42 |
mal | better add those now before you build the first image | 21:42 |
mal | also don't forget detritus package from patterns | 21:42 |
guhl | ghosalmartin, hi | 21:49 |
mal | ghosalmartin: I probably should write the 15.1 porting instructions somewhere so I don't have to tell everyone separately | 21:49 |
ghosalmartin | guhl, been waiting a while for you :P how did you get he mi mix 2 to boot with a broken config_vt? :P | 21:50 |
ghosalmartin | mal: guess a piratepad never hurt anyone :P | 21:50 |
mal | ghosalmartin: I'm see if I manage to wrtie something this week | 21:50 |
guhl | have been totally covered with work, some times i have to abandon everything because of work | 21:50 |
ghosalmartin | guhl, ah fair enough :) | 21:51 |
guhl | you are working on the chiron? | 21:51 |
guhl | mal that would be very nice | 21:52 |
ghosalmartin | trying to. ive just managed to drain the battery and get it stuck in a boot loop on a dead kernel bleurgh | 21:52 |
ghosalmartin | i started some work but on the latest kernel CONFIG_VT seemed to just nuke the kernel to the point where I had to do a clean and rebuild | 21:53 |
ghosalmartin | but that could of been some issue with dm-verity. device won't boot to test :P | 21:53 |
guhl | i just restarted the work based on LOS 15.1 | 21:55 |
ghosalmartin | ah it just needed a decent charger | 21:55 |
ghosalmartin | guhl, can you get to telnet? | 21:55 |
guhl | currently working on the mer friendly kernel | 21:56 |
guhl | nothing to boot yet | 21:56 |
guhl | it took me ages to get my old LOS 14 based to build to boot so that i could telnet into it | 21:58 |
ghosalmartin | well if it reboots instantly disable CONFIG_VT, if it takes a bit longer...duno :P | 21:59 |
mal | ghosalmartin: it seems you need to fix libhybris a bit, the spec needs some small fixes for hybris-15.1 | 22:09 |
ghosalmartin | mal: yeah i noticed, shouldnt be too hard but am all excited by the mi mix 2 device now :P | 22:10 |
ghosalmartin | https://pastebin.com/H8PLCWah | 22:11 |
ghosalmartin | after I unplugged and replugged usb it rebooted oops | 22:11 |
guhl | do you have any function enabled of the config_fs | 22:13 |
ghosalmartin | guhl, yes, hybris-boot latest master has config_fs. am trying your specific configfs flags now | 22:15 |
guhl | i mean in the kernel config | 22:17 |
guhl | as it says: Config c/1 of g1 needs at least one function. | 22:17 |
guhl | configfs-gadget a800000.dwc3: failed to start g1: -22 | 22:17 |
ghosalmartin | like CONFIG_USB_CONFIGFS_UEVENT? | 22:18 |
guhl | like CONFIG_USB_CONFIGFS_F_FS=y | 22:21 |
guhl | i am currently struggling with the rndis function | 22:21 |
ghosalmartin | well all of them were enabled when I attempted that | 22:22 |
guhl | ok | 22:22 |
ghosalmartin | guhl, some success | 22:23 |
ghosalmartin | telnet is up | 22:23 |
ghosalmartin | https://pastebin.com/MiMwqCv2 | 22:24 |
ghosalmartin | today has been very productive it seems :P | 22:26 |
ghosalmartin | guhl, CONFIG_USB_CONFIGFS_F_GSI is the one breaking hte rest | 22:36 |
mal | hmm you need such strange usb changes, I only needed some simpler things | 22:40 |
ghosalmartin | mal: tbh i can disable most of those and it still works fine | 22:46 |
mal | ok | 22:46 |
ghosalmartin | am just tryna be like a bull in a china shop...but not knock too many things over. its been months with this device not booting :P | 22:47 |
guhl | thats the same i also had in the config for the LOS 14 build | 22:48 |
ghosalmartin | thanks to this device and the pixel xl both being hybris 15.1 i can try and do both at the same time | 22:49 |
guhl | i am looking forward to seeing the wifi and modem coming up :-) | 22:49 |
ghosalmartin | am determined now to make this device my daily driver :P | 22:50 |
guhl | i would be more than happy my jolla c is falling apart | 22:51 |
ghosalmartin | my X is actually doing pretty well...but i never bothered glueing the back on when I took it off and now it randomly falls off :P | 22:51 |
guhl | i just inherited a Z3C from my wife (where the back also falls of) might also be a nice device | 22:53 |
guhl | when i try to build with the config_usb flags like this the kernel build fails with | 22:55 |
ghosalmartin | oh another thing, the SELINUX_BOOTPARAM config has been breaking on my Pixel XL when I enable it | 22:56 |
guhl | https://pastebin.com/kr5x7qae | 22:56 |
ghosalmartin | guhl, https://github.com/ghosalmartin/android_kernel_xiaomi_msm8998/commits/hybris-15.1 | 22:56 |
ghosalmartin | thats fully building and conforms to friendly mer | 22:56 |
ghosalmartin | and runs | 22:56 |
guhl | i'll grab it | 22:58 |
guhl | should we somehow join the effort and not have two separate github repositories? | 23:01 |
ghosalmartin | guhl, possibly :P | 23:03 |
ghosalmartin | ive added you as collab | 23:03 |
ghosalmartin | guess ill just have to remember to branch off hybris-15.1 and stop force pushing :P | 23:04 |
ghosalmartin | hmm i updated my sdk and ubu-chroot has gone :/ everytime I get close to anything porting this happens :P | 23:06 |
guhl | k, we'll talk about that, today i have to hit the sack. | 23:07 |
ghosalmartin | night man | 23:07 |
guhl | night | 23:07 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!