**** BEGIN LOGGING AT Fri Jul 22 02:59:57 2011 Jul 22 06:16:38 moin Jul 22 07:35:31 SHR: 03Martin.Jansa 07shr-chroot * r15ce1666f805 10/ (93 files in 14 dirs): system upgrade Jul 22 09:39:39 hi Jul 22 09:52:58 is there a clean way to remove udev? Jul 22 09:53:21 it seems to be a dependency of xserver-xorg, however sshd misbehaves if i have udev installed Jul 22 09:54:10 marcellus1: only libudev is pulled by xserver-xorg to image Jul 22 09:54:18 not udevd Jul 22 09:54:23 hmm Jul 22 09:54:35 I did just erase everything with *udev* off my phone Jul 22 09:54:42 and now its kind of broken Jul 22 09:54:48 of course Jul 22 09:54:52 libudev is needed Jul 22 09:56:23 I want to prevent udev from being built in the first place Jul 22 09:56:56 I mean udevd Jul 22 09:57:22 it got somehow pulled in Jul 22 10:02:39 I had this error http://permalink.gmane.org/gmane.comp.handhelds.openembedded.scm/46401 and also added the line to shr-image.inc however udevd still gets pulled in Jul 22 10:20:29 GNUtoo|laptop: good morning Jul 22 10:21:58 hi Jul 22 10:23:09 I was just playing around with udevd/libudev Jul 22 10:23:16 and apperantly did something wrong Jul 22 10:23:27 we use devtmpfs Jul 22 10:23:31 when we can Jul 22 10:24:03 else we use udev-compat for older devices Jul 22 10:24:08 like palm-pre Jul 22 10:24:29 I needed to remove udevd because of http://permalink.gmane.org/gmane.comp.handhelds.openembedded.scm/46401 Jul 22 10:25:01 I did add IMAGE_DEV_MANAGER_gp-zero ="" to shr-image.inc however udev was still added Jul 22 10:25:13 try that: Jul 22 10:25:16 = "" Jul 22 10:25:20 instead of ="" Jul 22 10:26:12 and besides breaking my current running system I have apperantly removed too much of the tmp/deploy directory and now get http://paste.pocoo.org/show/444178/ Jul 22 10:26:39 I don't know for the libudev0 Jul 22 10:26:43 maybe ask JaMa Jul 22 10:28:16 JaMa: ? Jul 22 10:29:55 GNUtoo|laptop: but thats not that tragic Jul 22 10:30:06 marcellus1: hmm build "udev" Jul 22 10:30:08 ok Jul 22 10:30:10 GNUtoo|laptop: I can just change the xorg recipe to compile it without libudev support Jul 22 10:30:19 JaMa: I want to get rid of udev Jul 22 10:30:20 udevd is one of produced packages, libudev0 is another Jul 22 10:30:54 there is an inc for xorg which decide which one to build between: Jul 22 10:30:56 *none Jul 22 10:30:57 *udev Jul 22 10:30:58 when the system uses devtmpfs libudev supposely only confuses xorg Jul 22 10:30:58 I know but libudev0 is fine and is not causing any troubles.. you need to get rid of udevd on _target_ only Jul 22 10:31:03 *hal Jul 22 10:31:33 DISTRO_XORG_CONFIG_MANAGER Jul 22 10:31:39 yes that one Jul 22 10:31:51 but again if you have udev in DISTRO_XORG_CONFIG_MANAGER then it pulls only libudev and that's fine Jul 22 10:32:16 because it allows autoadd deviced based on names provided by kernel Jul 22 10:32:20 marcellus1, how can you be sure your problem comes from udev? Jul 22 10:32:49 oops Jul 22 10:32:53 I meant: Jul 22 10:32:59 http://permalink.gmane.org/gmane.comp.handhelds.openembedded.scm/46401 Jul 22 10:33:02 it fixed it Jul 22 10:33:04 are you really sure that your problem comes from udev Jul 22 10:33:08 only I did remove too much Jul 22 10:33:14 marcellus1: and there is more packages depending on libudev.. so don't expect to get rid of it :) Jul 22 10:33:22 I mean the Xorg problem Jul 22 10:33:22 :( Jul 22 10:33:25 with white screen Jul 22 10:33:32 the screen is still blank Jul 22 10:33:37 marcellus1: and not even building it is a lot more difficult Jul 22 10:33:52 however after removing udev from the runlevel i can get bash with jobs control Jul 22 10:33:55 yes, however you have the log of xorg Jul 22 10:34:12 thats still ongoing Jul 22 10:34:36 ok Jul 22 10:34:47 I think I messed something up with the kernel config since ootb x was coming up Jul 22 10:34:54 and now I just have a blank screen Jul 22 10:35:07 ok Jul 22 10:35:10 quite possible Jul 22 10:35:30 but rebuilding takes a lot of time :( Jul 22 10:35:35 ah ok Jul 22 10:35:47 then I've a trick for you Jul 22 10:35:55 bitbake devshell Jul 22 10:36:01 source it like that: Jul 22 10:36:35 source ~/embedded/oe/oetmps/shr/deploy/addons/arm-oe-linux-gnueabi-shr-om-gta02-devshell Jul 22 10:36:40 marcellus1: opkg remove udevd Jul 22 10:36:42 replace it with the path you use Jul 22 10:36:50 and go in the kernel dir Jul 22 10:36:52 marcellus1: then it won't be in runlevel at all :) Jul 22 10:36:52 and do: Jul 22 10:37:22 I could have just removed /etc/rc?.d/*udevd Jul 22 10:37:39 make ARCH=arm CROSS_COMPILE="arm-oe-linux-gnueabi-" LDFLAGS="" Jul 22 10:39:23 that sounds interesting, however I still need to wait a moment Jul 22 10:40:06 yes but each rebuild won't be that long Jul 22 10:40:23 then you can even do a script to automatize that Jul 22 10:40:28 like make ... Jul 22 10:40:33 :) Jul 22 10:40:43 fastboot /path/to/kernel Jul 22 10:40:49 etc... Jul 22 10:40:56 *fastboot boot Jul 22 10:42:45 my ram is full before parsing recipes is done, I think it will still take a while Jul 22 10:43:47 however devshell is compiling, time for a smoke, brb Jul 22 10:50:15 GNUtoo|laptop: am I supposed to run the devshell as root? Jul 22 10:50:26 I get Jul 22 10:50:26 $ source tmp/deploy/addons/arm-oe-linux-gnueabi-shr-gp-zero-devshell Jul 22 10:50:26 -su: alias: `./configure': invalid alias name Jul 22 10:51:35 it does however compile :) Jul 22 10:52:52 no Jul 22 10:52:56 that's normal Jul 22 10:53:02 ignore the warninig Jul 22 10:53:32 hell, I forgot to copy .config to defconfig Jul 22 10:55:01 ok Jul 22 10:55:11 but here since you're doing it manually Jul 22 10:55:15 you don't need to do so Jul 22 10:55:21 you just fastboot the zImage Jul 22 10:55:43 i need the shr image as well Jul 22 10:55:58 since I deleted quite everything with *udev* Jul 22 10:56:07 and opkg doesn't seem to like that a lot Jul 22 10:56:38 do what JaMa said then Jul 22 10:57:21 I can't remove, since there is a lot of dependencies on libudev0 Jul 22 10:57:32 and when i try to force opkg segfaults Jul 22 10:57:44 but thats a little bit out of the use case I guess Jul 22 10:58:44 don't remove Jul 22 10:58:47 as said JaMa Jul 22 10:58:56 just remove udevd Jul 22 10:58:59 not libudev0 Jul 22 10:59:04 but I can't remove udev Jul 22 10:59:13 why? Jul 22 10:59:18 what does it tell? Jul 22 11:00:02 I can't tell right now, since the phone is down right now, but don't worry I 'll just rebuild the image and reflash Jul 22 11:00:09 ok Jul 22 11:04:40 if you've set IMAGE_DEV_MANAGER_gp-zero to empty after task-base was built then task-base is probably depending on udevd, but that's fine to -force-depends Jul 22 11:06:52 how -force-depends? Jul 22 11:07:03 do i need to rebuild task-base to get rid of udev? Jul 22 11:25:16 yes if you didn't set IMAGE_DEV_MANAGER_gp-zero soon enough Jul 22 11:26:36 hmm maybe not Jul 22 11:26:56 it should be fine without rebuild Jul 22 12:03:10 hi everyone Jul 22 12:40:23 compilation finished, udev was not included Jul 22 12:40:53 however I removed X from all runlevels and the screen blanking doesn't seem to be connected to X at all Jul 22 12:40:54 udev*d* right? Jul 22 12:40:59 yeah Jul 22 12:41:07 sry Jul 22 12:42:25 I can't find the systemlog Jul 22 12:42:58 /var/volatile/log only has lastlog which is 1 line long Jul 22 12:46:18 dmesg shows Jul 22 12:46:18 [ 602.255921] lcdc_ili9325sim_panel_off *** Jul 22 12:46:18 [ 602.256069] lcdc un-config Jul 22 12:46:18 [ 602.277923] bl_lv 0 bl_on 0 panel_on 0 Jul 22 12:46:25 no luck with google by now Jul 22 12:48:20 any idea? Jul 22 13:08:43 thanks to recent updates infobot doesnt reackt anymore to "thanks .* apt !" posts? Jul 22 13:08:55 react Jul 22 13:09:21 thank you apt Jul 22 13:09:33 thanks apt Jul 22 13:09:47 better that, it was a bit weird Jul 22 13:13:24 thanks to recent kde updates my kmail scrambles non-ascii chars when replying to encrypted mails... not good :-( Jul 22 13:24:00 mrmoku, pfft, I told you already. Jul 22 13:24:03 mrmoku, kde sucks :P Jul 22 13:24:17 I think I've been telling you that for years now. Jul 22 13:24:49 fluxbox is better but doesn't include mail application :) Jul 22 13:25:05 afaik (never looking for one as mutt is fine) Jul 22 13:26:55 JaMa: I'm using mutt as secondary mailer Jul 22 13:27:16 JaMa: though it has two big downsides Jul 22 13:27:40 JaMa: a) loading big folders is quite slow (though that maybe does not hold anymore with current mutt and header caching) Jul 22 13:27:48 b) you can only see one mail Jul 22 13:31:29 mrmoku: I was happy with mutt speed as long as I was using synchronized maildir stored locally Jul 22 13:32:03 mrmoku: laterly I've changed config to read gmail account directly over imap and sometimes it takes quite long :/ Jul 22 13:32:33 mrmoku: but kmail/thunderbird is not running very well in screen, so not an option for me :) Jul 22 13:33:20 and b) yeah sometimes it's pain and I start 2nd mutt in next screen window.. Jul 22 13:34:54 yeah, which is said to be not a good thing though Jul 22 13:35:30 I'm ussing offlineimap to sync my IMAP boxen and use the header caching thing mutt has integrated these days Jul 22 14:12:54 library for msm modems in samsung phones: https://github.com/ius/samsung_h1_libmsm Jul 22 14:12:59 hi btw Jul 22 14:14:12 hi Jul 22 14:18:19 is there a command line tool to interact with lcdc? Jul 22 14:33:06 lcdc? Jul 22 14:35:58 thats appearently somehow connected to my panel Jul 22 14:37:08 when I boot up the panel lights up until it blanks, at that time I get Jul 22 14:37:08 [ 602.255921] lcdc_ili9325sim_panel_off *** Jul 22 14:37:08 [ 602.256069] lcdc un-config Jul 22 14:37:08 [ 602.277923] bl_lv 0 bl_on 0 panel_on 0 Jul 22 15:58:29 marcellus1, diff your config with cyanogen's Jul 22 15:58:34 with diff -u Jul 22 16:04:34 I started with the cyanogen's config Jul 22 16:05:26 I only changed paranoid network and added usb_net support Jul 22 16:09:58 GNUtoo|laptop: also the default config is not easily accessible, I'd need to reboot into android and copy it from there Jul 22 16:10:34 ok Jul 22 16:10:55 you said xorg worked before, right? Jul 22 16:11:10 yeah Jul 22 16:11:23 however only for 2seconds until the screen went blank Jul 22 16:11:30 ah ok Jul 22 16:11:35 hmmm Jul 22 16:11:44 dmesg? Jul 22 16:12:10 [ 602.255921] lcdc_ili9325sim_panel_off *** Jul 22 16:12:11 [ 602.256069] lcdc un-config Jul 22 16:12:11 [ 602.277923] bl_lv 0 bl_on 0 panel_on 0 Jul 22 16:12:12 maybe you need a refresh thread Jul 22 16:12:19 ? Jul 22 16:12:20 ok Jul 22 16:12:32 try to port the refresh thread patch to your kernel Jul 22 16:14:28 where do I find the patch? Jul 22 16:14:29 http://gitorious.org/htc-msm-2-6-32/leviathan-incoming/commit/accc8020388d71d0df6189eb1227ffe25a022323 Jul 22 16:52:41 GNUtoo|laptop: is linux-leviathan an android fork? Jul 22 16:53:04 it's a kernel only Jul 22 16:53:06 the patch doesn't apply, there is no msmfb_update function Jul 22 16:53:15 ah? Jul 22 16:53:16 however I found Jul 22 16:53:16 /* ToDo: possible conflict with android which doesn't expect sw refresher */ Jul 22 16:53:16 /* Jul 22 16:53:16 if (!mfd->hw_refresh) Jul 22 16:53:16 { Jul 22 16:53:17 if ((ret = msm_fb_resume_sw_refresher(mfd)) != 0) Jul 22 16:53:17 { Jul 22 16:53:18 MSM_FB_INFO("msm_fb_blank_sub: msm_fb_resume_sw_refresher failed = %d!\n",ret); Jul 22 16:53:18 } Jul 22 16:53:19 } Jul 22 16:53:19 */ Jul 22 16:53:47 try to activate the software refresher Jul 22 16:57:41 and remove the traces of the patch you tried to apply if there are some Jul 22 16:58:00 that could break compilation if you have some stuff remaining from the patch I pointed you too Jul 22 17:04:21 what was the git command to squash a commit again? Jul 22 17:05:03 git rebase -i HEAD~N Jul 22 17:05:07 with N beeing a number Jul 22 17:08:56 thx Jul 22 17:33:01 GNUtoo|laptop: :) Jul 22 17:33:03 works Jul 22 17:33:10 what works? Jul 22 17:33:12 the rebase? Jul 22 17:33:17 Xorg? Jul 22 17:33:26 the screen now turns dim, not off Jul 22 17:33:34 ok nice Jul 22 17:33:36 and it turns back on, when I touch the screen :) Jul 22 17:33:40 nice Jul 22 17:33:51 you have an image? Jul 22 17:33:59 on the screen Jul 22 17:34:10 I disabled x before Jul 22 17:34:14 wait, I'll try Jul 22 17:34:15 ok Jul 22 17:34:42 btw since you use fsodeviced and that android uses the LED API the screen could turn black Jul 22 17:34:54 because it would consider the screen as a led Jul 22 17:35:11 crashes nearly instantly Jul 22 17:35:13 you have to tell it not to ignore the screen Jul 22 17:35:14 ok Jul 22 17:35:15 how? Jul 22 17:35:21 log Jul 22 17:35:55 must be the session, the terminal works fine Jul 22 17:36:56 http://paste.pocoo.org/show/444389/ Jul 22 17:37:09 it switches back to the terminal when dimming Jul 22 17:38:38 what is Touchsceen: Close supposed to mean? Jul 22 17:39:20 that is better http://paste.pocoo.org/show/444391/ Jul 22 17:40:51 hmmm Jul 22 17:40:58 maybe it's xorg + xsession Jul 22 17:41:04 I suspect the input Jul 22 17:41:13 try to disable the touchscreen input Jul 22 17:41:17 and look if it crashes Jul 22 17:42:10 [1390865.181] (II) XINPUT: Adding extended input device "Touchscreen" (type: KEYBOARD) Jul 22 17:42:15 fix your xorg.conf Jul 22 17:42:16 and retry Jul 22 17:42:51 seems steady Jul 22 17:43:05 hint: Jul 22 17:43:07 try evtest Jul 22 17:43:12 evtest /dev/input/foo Jul 22 17:43:13 the screen turns light, when i touch it, but the taps don't reach to X Jul 22 17:43:15 I did Jul 22 17:43:19 it'll print its name Jul 22 17:43:21 it should be the right device Jul 22 17:43:25 ah Jul 22 17:43:26 ? Jul 22 17:44:33 event3 Jul 22 17:44:47 the other devices don't output anything while touching Jul 22 17:44:58 and event1 seems to be the volume buttons Jul 22 17:45:29 however I was wondering how to calibrate Jul 22 17:45:54 there is this script, but the home/back/search/menu buttons are also capacitive Jul 22 17:46:15 is the max_y supposed to be the edge of the screen, or the edge of the field Jul 22 17:46:34 or wich way can I use the buttons afterwards? Jul 22 17:52:59 xinput_calibrator Jul 22 17:53:33 usually android devices are already calibrated Jul 22 17:54:18 it looks pretty good Jul 22 17:54:32 its nearly 1:1 exact as the pixels lay Jul 22 18:03:22 is there a simple touchscreen testing program Jul 22 18:03:35 like where i can paint or something like that Jul 22 18:04:15 Xorg -retro Jul 22 18:04:37 that does not show anything, only complains about the touchscreen Jul 22 18:04:48 however the terminal does work Jul 22 18:05:00 well as far a terminal on a device with a keyboard can work Jul 22 18:05:20 i mean vala-terminal Jul 22 18:05:23 I tought the geeksphone zero has no keyboard Jul 22 18:05:30 it has no keyboard Jul 22 18:05:33 ok Jul 22 18:05:47 but I sticked vala-terminal into ~/.xinitrc to see if its the x server or something else Jul 22 18:05:56 ok Jul 22 18:41:36 is there any input driver for touchchscreens besides evdev? Jul 22 18:46:28 oktober? ffs Jul 22 18:46:37 wrong tab :( Jul 22 19:12:18 tslib Jul 22 19:12:23 i guess Jul 22 19:40:27 I'll go bye Jul 22 19:40:53 there is xf86-input-tslib there is also evtouch etc...but evdev is the preferred one Jul 22 19:40:59 if it doubble touch Jul 22 19:41:07 modify the kernel not to do so Jul 22 21:20:07 ohh i hate this Jul 22 21:21:03 i have installed archlinux from scratch in another partition and cornucopia stills dont compiling Jul 22 21:28:50 and if i conpile libnl from fso i get unrecognized option -nl-genl Jul 22 21:28:55 compile* Jul 22 21:31:42 wow Jul 22 21:31:47 fixed :) Jul 22 21:38:08 there's some variable to export that makes vala applications searchs for vapi files in folder exported? Jul 22 21:38:16 i don't think if that makes senses Jul 22 21:38:58 :) Jul 22 21:41:17 something like "./configure --with-valavapidir=$SOMEWHERE" Jul 22 22:03:10 maybe if i add them in .pc files... Jul 22 22:05:42 no... Jul 22 23:16:02 XDG_DATA_DIRS helped ;) Jul 23 02:27:03 freesmartphone.org: 03angelo 07angelox/testing * r7e667b46a902 10aurora/aurora-applications/app-settings/Makefile.am: aurora-applications: big update in Makefile to add recent settings app changes Jul 23 02:37:38 good night everyone **** ENDING LOGGING AT Sat Jul 23 02:59:57 2011