[00:00] -bash: /dev/wcnss_wlan: Bad address [00:00] yeah, not surprised [00:02] cwayne: anything in dmesg after that echo? [00:03] rsalveti, nope, but it failed with -bash: /dev/wcnss_wlan: Bad address [00:06] seems you're indeed missing a few files in there [00:06] like it's missing the latest commit from the device repo [00:07] ah, i wonder how old this image is [00:07] i tried his nightly but it didn't boot [00:09] alright i'll go get dinner and come back to this after.. rsalveti if you have a fix in mind ill be happy to build/test later tonight [00:10] i'm wicked excited about this, would love to at least have wifi + data working [00:10] cwayne: oh, by old I mean just the android image [00:11] rsalveti, yeah [00:11] rootfs is like 2 days old [00:11] build failed [00:11] Target boot image: /home/cwayne/Projects/ports/oneplus/phablet/out/target/product/bacon/boot.img [00:11] usage: mkbootimg [00:11] --kernel [00:11] --ramdisk [00:11] [ --second <2ndbootloader-filename> ] [00:11] [ --cmdline ] [00:11] [ --board ] [00:11] [ --base

] [00:11] [ --pagesize ] [00:11] -o|--output [00:11] device/oneplus/bacon/mkbootimg.mk:47: recipe for target '/home/cwayne/Projects/ports/oneplus/phablet/out/target/product/bacon/boot.img' failed [00:14] ah ok there we go [00:14] build success [00:15] rsalveti, what output did you want? [00:15] hm, failed to generate the boot.img though [00:16] probably wrong argument or missing one [00:16] i rebuilt, there was a github issue with mkbootimg, i copied his over and it seemed to work this time [00:16] oh, cool [00:16] so there should be a dir called root, or rootfs under out/target/product/bacon [00:17] cwayne: just give me the list of files in there [00:17] cwayne@boomer:~/Projects/ports/oneplus/phablet/out/target/product/bacon/root$ lscharger init init.usb.rc sepolicy [00:17] data init.bacon.rc proc sys [00:17] default.prop init.environ.rc property_contexts system [00:17] dev init.qcom-common.rc res ueventd.qcom.rc [00:17] file_contexts init.rc sbin ueventd.rc [00:17] fstab.bacon init.trace.rc seapp_contexts [00:18] rsalveti, ^ [00:18] cwayne: cool, yeah, with fstab and stuff [00:18] cwayne: now just fastboot flash boot out/target/product/bacon/boot.img [00:19] might need to update the system image as well [00:19] if you got conn_init [00:19] check for that [00:19] under out/target/product/bacon/system, look for conn_init [00:20] yeap [00:20] https://forums.oneplus.net/threads/wip-ubuntu-touch-for-oneplus-one.266170/ [00:20] to flash everything [00:20] rsalveti, just flashed boot.img and its now booting [00:21] right, that will help the fstab issue, but might still be missing the conn_init one (as I think that is under system instead) [00:21] cwayne: give me your dmesg output after booting [00:21] yeah, good point [00:21] oops was just gonna go into recovery, let me boot back real quick [00:21] in case updating system.img breaks shit [00:21] right [00:21] just want to know if it was able to find the fstab [00:23] rsalveti, http://paste.ubuntu.com/10630968/ [00:24] now trying to update system.img [00:25] cwayne: oh, sorry, yeah, needs a new system.img [00:25] cwayne: the android initrd lives inside the system.img [00:25] so dmesg didn't change much actually [00:26] ah right [00:26] here goes nothin [00:28] howdy [00:28] i want to mount the internal storage of aquaris but libmtp doesnt seem to recognize the device. what am i missing? [00:30] rsalveti, damnit, no boot [00:30] cwayne: how did you update the system.img? [00:30] re-ran rootstock [00:31] maybe the system.img is sparse, and you need to convert that first [00:31] not sure if his script is doing that [00:32] let me check [00:34] it actually converts that [00:34] cwayne: do you have a working recovery which you can use adb shell? [00:34] rsalveti, yea [00:35] cwayne: run: ./build/tools/update-system-img.sh out/target/product/bacon/system.img [00:35] cwayne: from recovery [00:35] cwayne: under the android build system dir [00:38] can anyone help me out please? im running libmtp 1.1.8 and my aquaris doesnt get recognized [00:38] Danny_D: how are you trying to mount the mtp device? [00:38] using it with nautilus or trying to manually mount it? [00:39] also, output of tail -f /var/log/syslog when connecting the cable to your host machine should also help [00:39] ugh rsalveti now i cant get the phone to power on at all [00:39] Danny_D: and another question, what are you using as host? [00:39] ah there we go [00:39] cwayne: :D [00:40] cwayne: any better? [00:41] rsalveti: using the mtpfs command, libmtp reports device unknown, uses android flags o_0 [00:41] on gentoo [00:41] Device 0 (VID=2a47 and PID=2008) is UNKNOWN. [00:42] right [00:42] that is probably why [00:42] let me point you to one patch we used to get it recognized [00:43] which reminds me we can already push the patch upstream, as the device is now public [00:43] rsalveti, not yet, this phone suuuuuuuucks sometimes [00:43] cwayne: what is the issue? [00:44] sometimes it just straight up refuses to power on [00:44] cwayne: weird [00:44] so it wasnt pushed to libmtp yet? ok i already thought so even though i find it a little strange since people actually using it :) [00:45] rsalveti, yeah, and super obnoxious [00:45] Danny_D: http://paste.ubuntu.com/10631100/ [00:45] let me take this opportunity to send that upstream [00:48] anyone know what causes the power mgmt problems on hammerhead? [00:48] rsalveti, yeah i cant get this thing to boot, probably a sign I need a break :) [00:49] will try and run that script after dinner [00:49] cwayne: alright, the reason might be that your system.img is busted [00:49] rsalveti: is there a way to do mount it and tell libmtp about the device specific options without recompiling? [00:49] Danny_D: not that I know [00:49] rsalveti, i doubt it, I dont even get it to power on at all [00:49] weird, put it to charge for a bit, make it relax a bit more [00:50] yeah [00:50] its weird, its happened a few times now [00:50] and i just got it back from an RMA today :/ [00:52] oh well === cwayne is now known as cwayne-afk [00:52] will check back in later [00:55] rsalveti: is the cause known of the power mgmt problems on hammerhead? [00:55] cwayne-afk: oh, ok [00:55] DonkeyHotei: which power mgmt issues? [00:56] well, most reports are that battery life on ubuntu is much wor [00:56] well, most reports are that battery life on ubuntu is much wose than on android [00:56] for hammerhead [00:58] rsalveti: ^ [01:00] well, it depends on a lot of things, first need to check if the device is indeed deep suspending [01:00] then check if something is behaving badly on that hardware [01:00] as we improved/fixed most of the known issues we had [01:01] is there a definitive test for deep suspend? [01:02] you can check that usually with syslog, but cking has a bunch of tools to investigate that [01:02] https://wiki.ubuntu.com/Kernel/PowerManagement/IdentifyingIssues [01:08] Danny_D: ok, sent upstream [01:19] Suspends: [01:19] 3 suspends aborted (13.04%). [01:19] 20 suspends succeeded (86.96%). [01:19] rsalveti: ^ === randy is now known as Guest98999 [01:27] I seem to have lost the ability to su within adb shell. How is that obtained? I've been trying, unsuccessfully, to boot ubuntu-touch. [01:28] is that while booted into android, or ubuntu? [01:29] trying to boot into ubuntu on my device. the device partially boots, I can adb shell, but then can't get much done due to permission denied. [01:30] i actually was able at one point run dmesg, but now I can't [01:32] i wasn't sure what the output of my kernel should be, so i re-flashed a build of cyanogenmod to compare, then when reflashing ubuntu, i lost su. [01:37] which device? [01:37] galaxy nexus [01:38] i'm not sure that port is maintained [01:38] yeah, I'm trying to get it running as a little project of mine. [01:39] since the device is not supported anymore, i tried to run through the porting guide. [01:43] well, look at that. adb root works. :) [01:44] if you can get the touch ui to display, the screenlock passwd you set will be used for sudo [01:45] the boot process never mounts the ubuntu container, so it just sits here with a blank screen. [01:45] is there anyway to get more output from the kernel? I've tried adding debug to the kernel cmdline but that doesn't seem to do anything. [01:46] are you using an ubuntu kernel image? [01:49] i originally built the kernel using the cyanogenmod tuna kernel source. i figured that was the culprit, so i used the pre-built maguro kernel and it gave the exact same output. [01:50] i think i even tried building the ubuntu kernel but that also gave the same outcome. [01:51] certain options in kernel config must be enabled, per the porting guide [01:52] rsalveti, still wont boot [01:52] hooray [01:52] cwayne-afk: try reflashing the system image that was published by the port owner [01:52] just to see if you can get it to boot again [01:52] rsalveti, i cant get it to fastboot or recovery [01:52] it literally doesnt power on [01:52] ouch [01:53] maybe disconnecting battery? [01:53] that's where i thought things went wrong. i tried enabling the kernel options, but every time I ran make, the .config file gets re-written [01:53] we had that issue with mako before [01:53] rsalveti, tamper-proof screws to be able to remove battery [01:53] cwayne-afk: that sucks [01:53] i have a set of tamper-proof drivers [01:54] holy shit i got it [01:54] kinda === cwayne-afk is now known as cwayne [01:54] if you hold the power button while plugging and unplugging usb it turns out [01:54] that makes, literally zero sense. [01:55] rsalveti: i got it to work, using libmtp 1.1.8 & thunar-volman with ur supplied patch [01:55] Danny_D: great [01:56] cwayne: :-) [01:56] stable hardware [01:57] A++ [01:57] randy1: are you sure you're actually flashing a full system image? [01:57] holy shit it booted [01:59] rsalveti, dmesg http://paste.ubuntu.com/10631404/ [02:00] cwayne: nice [02:00] cwayne: [02:00] [ 15.517777] initrd: mounting /dev/disk/by-partlabel/modem as /root/android//firmware [02:00] rsalveti, yeah, I have a /firmware, but it appears to be empty [02:00] so now /firmware should have everything [02:00] hm [02:00] cwayne: check /android/firmware [02:00] hm ls: cannot open directory .: Permission denied [02:00] /firmware should just be a syslink to /android/firmware [02:00] yo [02:00] also empty [02:01] where do ringtones need to be placed and messaging tones? [02:01] cwayne: weird, paste me the output of mount [02:01] rsalveti, http://paste.ubuntu.com/10631407/ [02:02] cwayne: yeah, it's not actually mounting it [02:02] yea [02:03] but i have /system/bin/conn_init [02:03] cwayne: paste me the output of /var/lib/lxc/android/rootfs/fstab* [02:04] /var/lib/lxc/android/rootfs/fstab.bacon [02:04] cwayne: sorry, cat [02:06] rsalveti, http://paste.ubuntu.com/10631431/ [02:09] cwayne: run the following, as root: [02:09] mkdir -p /android/firmware [02:09] mount /dev/disk/by-partlabel/modem /android/firmware -t vfat -o ro,shortname=lower,uid=1000,gid=1026,dmask=227,fmask=337,context=u:object_r:firmware_file:s0 [02:09] I think the amount of options is probably what is causing the error [02:11] cwayne: http://paste.ubuntu.com/10631455/ [02:11] didnt seem to complain [02:11] what it tries to execute in the initrd [02:11] cwayne: did it mount it? [02:11] yep [02:12] hm, that's super weird then [02:13] damnit, super weird is never easy fixes :) [02:15] cwayne: maybe the link from udev might be causing this issue [02:15] saw that before [02:15] /dev/block/platform/msm_sdcc.1/by-name/modem should be pointing out to a device [02:16] cwayne: like /dev/mmcblk0p5 [02:16] in the end [02:16] i think its /dev/mmcblk0p1 [02:16] /dev/mmcblk0p1 on /android/firmware type vfat (ro,relatime,uid=1000,gid=1026,fmask=0337,dmask=0227,codepage=cp437,iocharset=iso8859-1,shortname=lower,errors=remount-ro) [02:16] cwayne: right, can you replace the fstab lines by giving the mmcblk0 instead? [02:17] like replacing /dev/block/platform/msm_sdcc.1/by-name/modem with /dev/mmcblk0p1 [02:17] and so on [02:17] then rebuild the image, flash and reboot [02:18] how do i find out all of them? [02:18] i got that one just by running mount [02:19] cwayne: ls -l /dev/disk/by-partlabel [02:19] they are all links [02:19] ah right [02:20] rsalveti, do i just edit it in out/target/product/bacon/system/blah or is there somewhere else [02:20] cwayne: no, need to edit the original file [02:20] cwayne: probably under device/lge/bacon [02:20] or similar [02:20] ah there we go [02:20] let me find the gitrepo [02:20] rsalveti: wakelocks of interest are wlan_rx_wake, wlan_wd_wake, wlan_wake, and qpnp-vadc-ee189e00, in that order [02:21] device/oneplus/bacon/ubuntu-configs [02:21] cwayne: yeah, guess that's the one [02:21] cwayne: just rebuild and check under out/target/product/bacon/root/fstab [02:21] if it got the new values, then you changed the right file [02:21] DonkeyHotei: yeah, these are all kernel related [02:22] usually we can't do much for that ones [02:32] wtf [02:32] http://paste.ubuntu.com/10631554/ [02:33] oh goddamnit did i forget to run lunch [02:34] yep [02:34] ok there we go [02:34] rsalveti, so it looks right, just update the system.img? [02:39] cwayne: yeah [02:39] rsalveti, no difference [02:42] cwayne: as usual, dmesg and cat /var/lib/lxc/android/rootfs/fstab* [02:42] cwayne: and output of mount [02:43] rsalveti, mount: http://paste.ubuntu.com/10631592/ [02:43] rsalveti, http://paste.ubuntu.com/10631596/ fstab [02:44] wtf [02:44] dmesg [02:44] http://paste.ubuntu.com/10631601/ [02:45] rsalveti, ? [02:48] cwayne: run the following script: http://paste.ubuntu.com/10631610/ [02:48] cwayne: as root, with bash -x [02:48] bash -x ./script.sh [02:48] Which module will call function "media_codec_create_by_codec_name" of hybris/media/media.c? [02:50] jivin: gst-plugins-bad1.0 [02:50] the hybris plugin of that package [02:50] jivin: https://github.com/jhodapp/gst-plugins-bad/blob/master/sys/androidmedia/gstamchybris.c#L235 [02:50] rsalveti, http://paste.ubuntu.com/10631639/ [02:51] jivin, why do you ask? [02:51] Thanks.However, when I start application "MediaPlayer-app" to play a video, function "media_codec_create_by_codec_name" isn't called [02:51] cwayne: crap, there is a bug in our code [02:52] in the initrd? [02:53] cwayne: yeah [02:53] just a sec [02:55] jhodapp, I just want to figure out the full call procedures [02:55] jivin, it has to be called [02:56] cwayne: run the following: http://paste.ubuntu.com/10631651/ [02:56] cwayne: same thing, bash -x as root [02:56] Whether some media player application doesn't use gst-hybris? [02:57] For example, built-in application "Gallery"? [02:57] jivin, it definitely does, the only thing that doesn't use the hardware decoder yet are videos in the web browser [02:57] rsalveti, http://paste.ubuntu.com/10631655/ [02:58] jivin, it's called via: https://github.com/jhodapp/gst-plugins-bad/blob/master/sys/androidmedia/gstamcvideodechybris.c#L510 [02:58] cwayne: can you paste me your mount output now? [02:58] jivin, which is called by gstreamer automatically for the decoder plugin interface: https://github.com/jhodapp/gst-plugins-bad/blob/master/sys/androidmedia/gstamcvideodechybris.c#L390 [02:58] rsalveti, http://paste.ubuntu.com/10631660/ [02:58] looks better [02:58] I hook function "media_codec_create_by_codec_name" and add debug codes. However, there is no expected log when i play a video via "gallery"? [02:59] jivin, you're probably not compiling it correctly...you have to make sure to enable the hybris plugins, not the default gstamc only ones [02:59] cwayne: right, will upload this small fix to the archive, if you try to build it tomorrow it should hopefully fix your issue [02:59] but we still need to understand why the label is a problem [02:59] might be the same issue we had with snappy [02:59] had to call wait-for-root on every label [02:59] rsalveti, ! thanks [03:00] seems kind of a race with udev [03:00] so just repo sync then rebuild once its in? [03:00] cwayne: yup [03:01] rsalveti, cool, will you shoot me a mail when it's in? [03:01] cwayne: sure [03:01] thanks [03:01] so i shouldn't expect it to work even though stuffs mounted now right? [03:03] cwayne: well, you can still try now running that other scripts [03:03] like conn_init and family [03:04] still get -bash: /dev/wcnss_wlan: Bad address when i try to echo 1 to it [03:04] right, call conn_init first, under the android container [03:05] no change [03:08] cwayne: anything new in dmesg? [03:09] not that i can see [03:10] hm, that original aosp based repo has nothing as well [03:10] let me check the CM one [03:10] * cwayne needs to run for tonight [03:10] thanks for the help rsalveti [03:11] will pick it up again tomrrow am [03:11] alright, later! [03:11] rsalveti: here is the wakelock list http://paste.ubuntu.com/10631710/ [03:13] another problem! gst-hybris plugin will call android::MediaCodecs indirectly, and then it will access OMX hardware, which means application will have access to hardware device. So the problem:Does application have permission to do this? [03:14] cwayne: the one used by CM: https://github.com/CyanogenMod/android_device_oppo_msm8974-common/blob/cm-12.0/msm8974.mk#L162 [03:14] so yeah, they don't use conn_init it seems [03:14] but wcnss_service instead [03:14] with the blobs in place [03:15] jivin: we have media-hub doing that hardware related work [03:15] the app will only talk to media-hub [03:16] is media-hub a service in android or ubuntu? [03:18] ubuntu === salem_ is now known as _salem [03:36] Is media-hub a global service in ubuntu?And it serve all media application for media acceleration? And the service has permission to access to hardware device via gst-hybris plugin? [03:37] or it is just a library? [03:40] phablet 2481 0.0 1.3 359812 25648 ? Ssl Mar19 0:40 media-hub-serve [03:42] "phablet"(media-hub-server user) have access to hardware device? [04:02] how to install skype/ viber on ubuntu touch? | http://askubuntu.com/q/598968 [05:10] Ubuntu Touch Wireless PEAP Workaround | http://askubuntu.com/q/598980 === ecloud_wfh is now known as ecloud [07:42] How to install google chrome for ubuntu- touch? | http://askubuntu.com/q/599021 [08:21] good morning [08:30] dpm: https://wiki.ubuntu.com/Touch/ReleaseNotes [08:30] this page is over a year out of date, we should either update it or delete it. [08:31] I'm poking you because you have more edits in the history than anyone else, so I think you own it :) [08:34] ha [08:34] I just happened to modify the SSH access bits [08:34] but should be fine to delete, I guess [08:35] especially as it has advice on it we probably don't want people doing [08:35] and contributes to people doing stuff like this http://www.reddit.com/r/Ubuntu/comments/2zljaw/ubuntu_phone_aquaris_e45_ubuntu_edition/ [08:38] How to use ubuntu-touch as a desktop? | http://askubuntu.com/q/599039 [08:41] popey, i improved http://askubuntu.com/questions/348714/ubuntu-touch-apt-get-install-openssh-server-error a little, that was also heavily out of date [08:42] ooh, nice update [08:42] everyone upvote http://askubuntu.com/a/599041/612 :) [08:44] * popey deletes the section about ssh from the ReleaseNotes page and links to that instead [08:44] might be worth doing that for other sections [08:47] might be worth doing that for all our documentation :P [08:47] most of it is out of date i think [08:49] it would also be good to have a "why is the image writeable, why don't we use dpkg" page ... [08:49] err [08:49] *not writeable [08:50] ogra_: is it because you hate users' freedom? [08:50] what's going on with my krillin? used 23% battery only over night [08:51] that's good, but weird because it's usually not that great ;-) [08:51] nhaines, yeah, there is a longstanding internal conspiracy in canonical to make the deb format die, in fact thats the actual reason Ubuntu was founded :P [08:51] I knew it! I'm telling OMG Ubuntu and also The Register. ;) [08:52] seb128, i have the feeling your device is always a week behind :) [08:52] * ogra_ has awesome battery life since last weekend [08:52] well, I've been travelling this week [08:53] so power usage is not the same when you are commuting [08:53] ah right ... your GSM is more busy and does a lot of sending too [08:53] right [08:53] It's because movement generates static electricity and that affects battery discharge rate. [08:53] lol [08:54] yay ... friday \o/ [08:54] it is totally foggy outside ... no eclipse for me :( [08:54] oh, right [08:54] same here [08:54] we had wonderful weather the whole week ... just not today ... [09:03] Hmm, for the past few days, the Weather Channel scope has been really spotty, only returning directed search results and not local search results. Is there somewhere I could look to see if it's just unhappy with the GPS or something? [09:07] ogra_: dpm ok, i have dramatically trimmed down https://wiki.ubuntu.com/Touch/ReleaseNotes [09:08] Are there release notes for 14.10? [09:08] Devices with 1080 screens [09:08] Set the GRID_UNIT_PX environment variable to an appropriate value for the device (i.e. 30 for a 5” handset screen) in the /usr/bin/ubuntu-session script and in the /home/phablet/.bashrc file. [09:09] rip that out [09:09] (bottom) [09:09] endusers shouldnt play with GRID_UNIT_PX [09:12] ok [09:12] nhaines: not that I am aware of [09:32] could someone help me out please? I was hit by yesterdays unbootable devel proposed image. But i cant seem to roll back to the previous image as suggested in the email exchange about that. [09:33] matv1, roll forward ... the broken image was fixed [09:33] ogra_ image 142 right? tried that as well just now. getting the same error [09:34] .cache/ubuntuimages/ubuntu-touch/devel-proposed/mako/version-142.tar.xz.asc to device: free space on /cache/recovery is 542M [09:34] ogra_ so i am running ubuntu-device-flash --clean-cache touch [09:34] which gives me a nice "Cleaning prevously downloaded content" [09:35] but i still get exactly the same error [09:36] do you know what thats about? [09:37] oh hang on I can still adb shell into the device [09:39] are you in recovery ? [09:40] ogra_ yes [09:40] and you are on which device ? N4 ? [09:40] yes Mako [09:41] are yu also in recovery when running ubuntu-device-flash ? [09:45] ogra_ yes [09:47] ogra_ as I am able to adb shell I should be able to apt-get right? could that sort this out? [09:47] you are on 141 ? i think Mirv made his install work by installing the missing package, so yeah, that might work [09:51] ogra_ I am going to try that. Thnks fr now! [09:54] matv1: yes, if you have writable image you can do apt update + apt upgrade [09:54] and reboot [09:57] Mirv oh I did not make my image writable. [09:57] can i still do that now? [09:58] matv1: I think it should still be possible phablet-config writable-image [09:58]