**** BEGIN LOGGING AT Thu Feb 18 02:59:57 2010 Feb 18 03:27:15 JesusMcCloud: no me though. I'm a 24/7 insomniac :-P Feb 18 03:28:02 DocScrutinizer, i'm when playing around with edje... it works better at night Feb 18 04:48:25 ok, who do i throw rocks at for closing #318? Feb 18 06:00:45 moinmoin Feb 18 06:03:02 how to add new i2c device from user space in openmoko Feb 18 06:04:42 DocScrutinizer, moin, do you sleep... ever? Feb 18 06:05:01 nah, sleeping is overrated Feb 18 06:05:35 http://www.mjmwired.net/kernel/Documentation/i2c/instantiating-devices Feb 18 06:06:07 the doc explains abt adding new i2c device from the user space Feb 18 06:06:42 but i don't see the bus listed under /sys/bus/i2c/devices/ Feb 18 06:06:53 i only see the devices under it Feb 18 06:07:05 nasty Feb 18 06:07:11 echo eeprom 0x50 > /sys/bus/i2c/devices/i2c-3/new_device Feb 18 06:07:28 the above is the example given in the doc Feb 18 06:08:41 ask larsc Feb 18 06:09:04 or PaulFertser Feb 18 06:12:38 larsc: PaulFertser : can u plz respond to my question ? Feb 18 06:27:09 JesusMcCloud: hehe instead of slaping you should ask me to update them.. because this dependencis made sense in time I added them.. Feb 18 06:27:32 JaMa|GoNe, i have an offer for you: Feb 18 06:27:50 JesusMcCloud: that shr-theme-neo (meta package) pulls old libframeworkd-phonegui-efl-theme-neo and gpe-theme-neo which then pulls those gpe-scap and old libframeworkd stuff Feb 18 06:28:05 JaMa|GoNe, i never tried to install metapackage Feb 18 06:28:54 I have to go.. later Feb 18 06:28:55 JaMa|GoNe, the deal is: i have a new nEo theme ready. (just polishing some roug edges and testint mrmoku's themign mechanism)... if you fix the recipes, i will push it into git Feb 18 06:47:55 exit Feb 18 06:52:01 wolfyg: :-D Feb 18 06:59:37 DocScrutinizer: yeah, sorry ;-) Feb 18 06:59:53 wasn't even the channel I meant, just another zsh window... Feb 18 06:59:57 np Feb 18 07:18:19 DocScrutinizer: The debug board I2C signals are came out from S3C2442 arm processor, it that correct ? Feb 18 07:29:17 satish yes Feb 18 07:33:18 roh : Is it possible to have any gpio pins to get on Debug boards Feb 18 07:33:25 roh: I don't see on schematics Feb 18 07:33:46 roh: for my project i need 2 gpio pins .. Feb 18 07:34:01 there is only i2c and i think one irq capable gpio Feb 18 07:35:22 there is also an spi bus Feb 18 07:35:36 and a line calles SS0 Feb 18 07:36:39 http://downloads.openmoko.org/developer/schematics/GTA02/Schematics_Freerunner-GTA02_A5-A7cumulative_public_RC0.pdf Feb 18 07:36:44 page 7 Feb 18 07:36:54 lower right corner has the pinout of the debug conn. Feb 18 07:37:26 http://downloads.openmoko.org/developer/schematics/debugboards/OpenMoKo_Debug_Board_V3_MP.pdf shows the side of the debug board Feb 18 07:38:15 debugboard V3 has the stuff on a connector (usually not soldered in by default) labeled J11 Feb 18 07:39:12 but these (i2c and spi) are not 'free', they are shared with other devices inside the phone, like the audio codec (on i2c) and the accels (on spi) Feb 18 07:39:31 so addressing / chip select is important Feb 18 07:42:52 i'm not sure if the schem is correct, but it says spi0 is exclusively on the debug conn, and spi1 is shared between both accel. Feb 18 07:43:34 i2c is shared with the wolfson (audio codec, for config) and the pmu. as long as your stuff has different addresses it shouldnt be a problem Feb 18 07:54:22 mrmoku|away, either i screwed up, or theming phoneui does not work as it should Feb 18 07:58:28 satish: hi Feb 18 07:59:26 satish: i thought you wanted to write a userspace spi driver, is it a different question now? Feb 18 08:07:09 JaMa|GoNe: have you seen this when building xserver-xorg? configure.ac:42: error: must install fontutil 1.1 or later before running autoconf/autogen Feb 18 08:22:43 Weiss: yes it should be fixed in my repo.. font-util-native is new dependency for xserver-xorg Feb 18 08:23:04 Weiss: be warned, that I rebase shr-experimental repo *a log* Feb 18 08:23:06 Weiss: be warned, that I rebase shr-experimental repo *a lot* Feb 18 08:23:46 JesusMcCloud: great! can you please push it to repo first? So I can update recipes and SRCREVs used? Feb 18 08:24:15 JaMa|GoNe, as soon as i fixed my botched e17 config Feb 18 08:24:21 JesusMcCloud: if installing unnneeded stuff was the only problem you can install separate *neo* recipes.. Feb 18 08:24:38 JaMa|GoNe, i did *NOT* use the metapackage Feb 18 08:24:47 and I'll move those old neo recipes from RDEPENDS to RSUGGESTS Feb 18 08:24:58 JesusMcCloud: so which package pulled that old cruft? Feb 18 08:25:35 JaMa|GoNe, either elementary or illume theme Feb 18 08:28:12 JaMa|GoNe, there's only asmall problem atm: as of now, phoneui-theming doenst work (or i screwed up) overwriting default theme works... but i am sure mrmoku will fix this ;-) Feb 18 08:28:48 JesusMcCloud: yes.. because they also RRECOMMENDS other *neo* recipes.. already replaced that too with RSUGGEST (not installed by default with package) Feb 18 08:29:27 stupid bb commands... Feb 18 08:45:47 PaulFertser: Hi, yes it is different question...my co proc has I2C interface also...so i want to probe using I2C Feb 18 08:46:22 PaulFertser: as I2C has user space ioctls implemented as per kernel documentation Feb 18 08:46:36 satish: spi seemed to be an easier way since you do not have to share the bus with anybody Feb 18 08:47:26 PaulFertser: but i2c is easier to probe...i no need to implement anything extra....i can just asign device address and probe Feb 18 08:48:52 PaulFertser: if i'm able to register my co proc as new I2C device .....i feel it's easy to probe as i have i2c-0 device listed under /dev/ fs Feb 18 08:50:11 PaulFertser: in the fast i have some problem with I2C ....like while connecting the new device phone display is going off ...that's way i shifted to SPI Feb 18 08:50:21 PaulFertser: now the problem got solved... Feb 18 08:51:15 PaulFertser: Please let me know how to add new i2c device from user space.. Feb 18 08:51:39 JaMa|GoNe, do oyu have permissions to grant me git access ? (somehow i dont have access anymore) Feb 18 08:51:42 PaulFertser: i'm trying this http://www.mjmwired.net/kernel/Documentation/i2c/instantiating-devices Feb 18 08:51:55 satish: ok, let me see Feb 18 08:55:18 satish: which branch are you using? Feb 18 08:56:43 PaulFertser: andy tracking branch Feb 18 08:57:19 satish: why are you using a more modern documentation then, what makes you think it should work? Feb 18 08:58:42 JesusMcCloud: no and no, ask mrmoku|away (not sure if someone else has grant) Feb 18 08:58:48 PaulFertser: ok... Feb 18 08:59:07 PaulFertser: if i upgrade to the latest kernel 2.6.32 Feb 18 08:59:27 JaMa, i will email him my public key again Feb 18 08:59:30 PaulFertser: do openmoko has any known problems or bugs...? Feb 18 09:00:01 PaulFertser: if not, i will try to upgrade and see will it work ? Feb 18 09:00:46 satish: i guess there're known problems but i do not know exactly which. Feb 18 09:01:42 PaulFertser: ok....is it possible to get done in andy tracking branch...i mean adding new I2C device from user space Feb 18 09:01:56 JesusMcCloud: i see you there as bernd, is it you? Feb 18 09:01:59 satish: no idea yet Feb 18 09:02:37 JaMa, yup Feb 18 09:02:47 JaMa, waita sex Feb 18 09:03:03 s/sex/sec/ Feb 18 09:03:13 * JaMa awaiting sex Feb 18 09:03:17 nope Feb 18 09:03:21 not with Jesus.. but.. Feb 18 09:03:34 staying up all night has its side effects... Feb 18 09:03:45 JesusMcCloud: are you pushing over ssh://git? Feb 18 09:03:52 i am trying Feb 18 09:04:17 you can e-mail me the key and I can compare them.. Feb 18 09:04:27 but I guess you would remember changing key lately Feb 18 09:04:58 JaMa, will do (although i think i shoudl be listed as jesus, not bernd) Feb 18 09:05:57 no jesus in shr git config Feb 18 09:06:25 and bernd is for shr-themes and libphone-ui-shr repo Feb 18 09:06:52 JaMa, well then.. email is sent Feb 18 09:07:41 JesusMcCloud: yes same key.. so maybe you have to specify username for git+ssh :) Feb 18 09:08:10 JaMa, i never had to do this... hmm will try Feb 18 09:09:44 JaMa, permission denied Feb 18 09:11:16 JesusMcCloud: hmm I don't know where gitosis logs auth failures.. so I cannot help you.. mrmoku probably can Feb 18 09:12:16 i'll regenerate my public key (maybe i did screw up sth when creating a couple of keys for university) Feb 18 09:32:22 Weiss: only issue still remaining is while building xf86-video-glamo I had to remove -Werror, because in edid.h in new xorg are bit fields and warning is emitted that it's only gcc extension Feb 18 09:32:40 ok, cool Feb 18 09:32:54 thanks for taking care of all of this Feb 18 09:33:19 JaMa, http://scap.linuxtogo.org/files/224745ab1570e89094058702e05c1ed3.png http://scap.linuxtogo.org/files/9f1ccc1b1a12bc9b3b29def10b2a0555.png Feb 18 09:33:37 you're welcome.. if it saves you few mins for glamo developement, than it's win for all :) Feb 18 09:34:04 JesusMcCloud: great! looks nice Feb 18 09:35:05 neo theme is now even more complete than gry Feb 18 09:36:03 only thing missing is menu and map, everythign else is nice and green Feb 18 09:36:43 i need some sleep... gn8 Feb 18 09:36:51 does that theme partecipate in the contest? Feb 18 09:37:00 graziano, dunno Feb 18 09:37:38 it took almost half ayear until people began to realize how much performance gain is possible through smart themes Feb 18 09:37:53 the current default is participating Feb 18 09:38:08 bed time now... Feb 18 09:40:05 I like the blue one Feb 18 10:39:32 SHR: 03jesus 07shr-themes * r19f74990c4ab 10/phoneui-shr/phoneui-shr-theme-neo/ (20 files): neo theme for phoneui Feb 18 10:39:33 SHR: 03jesus 07shr-themes * r570c1cb6d9a3 10/elementary/elementary-theme-neo/ (43 files): complete rewrite Feb 18 10:39:35 SHR: 03jesus 07shr-themes * r66ebdbaed125 10/e-wm/e-wm-theme-illume-neo/ (8 files): fixed flaunch Feb 18 10:45:17 hello , there is problem with intone-video , and it does not run in gui ? Feb 18 10:48:26 leviathan: any prediction about SHR calling on htc dream? Feb 18 11:03:31 PaulFertser: Is it possible to upgrade andy tracking branch with git checkout to linux 2.6.32 kernel Feb 18 11:06:55 what formats does mplayer support on arm ? itone-video just plays flv format Feb 18 11:13:46 how to restart X in shr-t ? Feb 18 11:15:48 alabd: ssh in and run /etc/init.d/xserver-nodm stop and then start Feb 18 11:16:13 hanks Feb 18 11:16:16 thanks Feb 18 11:18:03 JaMa: what video formats does intone-video support on shr-t ? Feb 18 11:18:06 satish: you're supposed to checkout the branch you need Feb 18 11:18:39 satish: in fact to create a local branch based on origin/blabla (remote branch) Feb 18 11:20:13 alabd: almost all as mplayer on desktop Feb 18 11:20:22 alabd: except stuff like win32 etc ofc Feb 18 11:21:21 PaulFertser: for 2.6.32 did any one in openmoko list made a branch and customized to the phone ? Feb 18 11:21:48 PaulFertser: openmoko freerunner phone Feb 18 11:22:08 satish: there is 2.6.32 branch in git.openmoko.org Feb 18 11:22:25 JaMa: ok Feb 18 11:22:34 JaMa: intone-video does not play 3gp , mp4 Feb 18 11:24:50 any other complete video player Feb 18 11:27:30 JaMa: what is the branch name ...any idea ? Feb 18 11:28:10 om-gta02-2.6.32 looks quite self-expanatory to me Feb 18 11:30:15 JaMa: ok .. Feb 18 11:31:04 JaMa: actually there three like gta02-machine-2.6.32....so i confused ... :) Feb 18 11:32:50 <[Rui]> alabd: fix mplayer, not intone. Feb 18 11:33:18 satish: you can see what gets merged to om-gta02- and it will be clear Feb 18 11:33:35 satish: or there was en e-mail about new branches about a year ago in ML Feb 18 11:34:17 JaMa: ok.. Feb 18 11:37:42 [Rui]: thanks how to fix ? http://pastebin.com/m5d532d1 this is /usr/etc/mplayer/codecs.conf Feb 18 11:38:16 <[Rui]> alabd: maybe you need to fix the build so it builds 3gp support, I don't think that's a runtime thing :) Feb 18 11:41:29 JaMa: http://patchwork.dev.bearstech.com/patch/671/ but i forgot to add some credit to you, so please add it :) Feb 18 11:42:41 JaMa: should not those formats that are in this list be played ? http://pastebin.com/m5d532d1 this is /usr/etc/mplayer/codecs.conf Feb 18 11:42:56 alabd: 3gp was supported in amrwb/amrnb now on desktop there is usually opencore-amr used instead.. Feb 18 11:43:33 alabd: I didn't check opencore-amr availability for arm and didn't changed mplayer_git recipe for that.. Feb 18 11:44:36 Heinervdm: no credit needed, thanks for patch.. I'll try it Feb 18 11:45:04 JaMa: i removed libpng and zlib from my host to test it, it works :) Feb 18 11:45:52 good :) Feb 18 11:46:15 Heinervdm: I guess you're sure that png and zlib is not used later on target Feb 18 11:47:01 JaMa: i don't understand Feb 18 11:48:06 Heinervdm: I've seen you removed dependency on png/zlib and changed that to -native only Feb 18 11:48:41 JaMa: the resulting package has an automatically generated depencie on libpng and zlib Feb 18 11:48:41 Heinervdm: so on FR there won't be png/zlib installed in imaginary case when nothing pulled it in before aceofpenguins Feb 18 11:48:46 i checked taht Feb 18 11:48:55 ah ok Feb 18 11:49:14 so every -native dependency creates runtime dependency for non-native package? Feb 18 11:49:16 i think that's some bitbake magic Feb 18 11:49:22 I didn't know that.. Feb 18 11:49:47 and there is also few cases when it's not possible (non-native doesnt exist) so maybe those cases are ignored then.. Feb 18 11:50:52 not sure if it's _always right_ magic.. as ie librsvg-native is used for preparing navit-icons, but then is not needed on device where you get already only .png files Feb 18 11:51:49 JaMa: i think that this is only working if you don't set RDEPENDS, but i'm not sure Feb 18 11:53:45 I've seen that every DEPENDS is added to RDEPENDS if exists (in some bbclass), but didn't know that it also loose it's -native postfix if non-native package exists.. Feb 18 11:54:35 Heinervdm: btw did you read that eglibc: reply on oe-devel? Feb 18 11:55:18 JaMa: there were only 2 Feb 18 11:55:53 and someone suggested using #!/bin/sh as some older glibc recipes does Feb 18 11:56:11 ah, no haven't read that Feb 18 11:56:14 which I've seen confirmed to work in some SHR ML a wbefore.. Feb 18 11:56:56 so we should check that Feb 18 11:57:33 Heinervdm: yes.. imho better solution than bash dependency.. Feb 18 11:57:52 yes Feb 18 12:00:14 NOTE: Not creating empty archive for curl-certs-7.19.7-r3.4 Feb 18 12:02:22 Heinervdm: seen that before.. when checking why it's not in feed.. then I left that as curl-certs doesn't look like best solution for spreading all certs Feb 18 12:03:19 JaMa: it seems that they were removed form curl sources Feb 18 12:03:23 Heinervdm: imho it should be handled with some better policy what certs are thrusted by default etc.. so maybe following debian ca-certs package closely would be good start Feb 18 12:03:36 perhaps we can create a ca-certificates packages from that file: http://curl.haxx.se/ca/cacert.pem Feb 18 12:04:29 Heinervdm: I was thinking about this http://packages.debian.org/lenny/ca-certificates Feb 18 12:05:02 JaMa: i can't find a real source for that package Feb 18 12:05:11 use their tar.gz as SRC_URI and then just repack it to .ipk Feb 18 12:06:23 I would expect that there is no real upstream package for them.. just prepared tar.gz from certs they consider usefull and valid :) Feb 18 12:07:21 TAsn: thnx for the fix. Now contact-list is ok. Feb 18 12:07:53 mrmoku|away: yesterday i've got a black screen during the phone call. 8( Feb 18 12:09:56 Heinervdm: but in case some cert is in worse case compromised.. than we should be able to remove it quick.. so I guess that we should choose good upstream for security issues Feb 18 12:10:40 JaMa: now what should humble do to play mp4 or 3gp ? Feb 18 12:11:25 alabd: check opencore-amr availability for arm and change mplayer_git recipe for that.. if possible Feb 18 12:11:47 alabd: probably also new recipes for opencore-arm would be needed Feb 18 12:12:13 any other player with most format support on shr ? vlc ? Feb 18 12:17:11 he forgot: ask GOD for amr packaging Feb 18 12:18:10 ok i asked it just now Feb 18 12:19:34 ah good, then we can just wait for commit :) Feb 18 12:27:20 sure Feb 18 12:31:05 SHR: 03jesus 07shr-themes * rf847105c5ef5 10/icons/icon-theme-neo/86x86/apps/ (16 files): more icons for nEo theme Feb 18 12:59:03 leviathan, ping Feb 18 13:03:26 help Feb 18 13:04:06 is still valid that battery bug that fail to charge the freerunner if you discharge all the battery? Feb 18 13:05:02 graziano: please rephrase the question Feb 18 13:05:10 om-gta02-2.6.32 branch don't have build env like andy branch, Is there any doc for build guidlines ? Feb 18 13:07:05 PaulFertser: I remember that there was a bug. It was related to this: the user had not to let the battery discahrge completely because then it couldn't be recharged. Because to recharge it you had to switch it on first Feb 18 13:07:36 PaulFertser, he wonders if when your battery is totally or nearly totally discharged,you could still boot the freerunner in some way in order to be able to charge it again Feb 18 13:07:49 PaulFertser, btw I solved my wifi firmware problem... Feb 18 13:08:06 GNUtoo: thanks :) Feb 18 13:08:08 I was told to install the missing udev rules and scripts Feb 18 13:08:13 and it worked Feb 18 13:08:19 GNUtoo: yeah, saw that Feb 18 13:08:46 GNUtoo: though "symbol version mismatch" can't be exlained by that Feb 18 13:08:59 PaulFertser, I solved that too Feb 18 13:09:11 PaulFertser, I simply extracted *all* modules and it worked Feb 18 13:09:18 instead of just libertas Feb 18 13:09:40 btw,now libertas wifi fails a lot less often...I think I still will have to bugreport Feb 18 13:10:07 GNUtoo: i guess that confirms my suspicion. You had "regular" *80211 installed but tried to run the driver from compat-wireless or something like that. Feb 18 13:10:17 indeed Feb 18 13:10:37 opkg upgrade overwritten my compat-wireless things Feb 18 13:10:46 graziano: nothing's changed, depends on hardware revision mostly. A6 can be booted without a battery altogether but software support for that is suboptimal. Feb 18 13:10:48 even If I removed libertas in defconfig Feb 18 13:11:10 PaulFertser: I was thinking about getting an A7+ Feb 18 13:11:18 graziano: nothing's changed with it Feb 18 13:11:28 In this regard Feb 18 13:11:52 well but Feb 18 13:12:00 you said it can be booted without battery :) Feb 18 13:12:24 which bootloader implement that(booting without battery) ? Feb 18 13:13:54 GNUtoo: Qi doesn't care so one can say it does. Also nor u-boot sometimes manages to negotiate on 500mA before turning on the backlight so sometimes one can boot the system this way too. But in the Qi case the kernel turning on BL too early will ruin everything. I'm not sure i this's fixed in .32 but at least larsc reported success in his tests. Feb 18 13:14:22 ok thanks a lot Feb 18 13:17:33 PaulFertser, so if we boot with QI on micro-sd card a kernel without CONFIG_BACKLIGHT or something like that it could work? Feb 18 13:18:24 GNUtoo: i haven't tried it but it should work Feb 18 13:18:36 ok thanks Feb 18 13:31:51 GNUtoo: i guess there was some pmu BL register access hardcoded, so probably disabling BACKLIGHT is not enough, at least on andy-tracking. Feb 18 13:32:10 ok Feb 18 13:32:42 so what should be done? booting a dummy thing like grub-invaders? Feb 18 13:33:13 ah that would not negoziate 500mA Feb 18 13:34:16 s/should/could Feb 18 13:34:45 or maybe modify kenrel sources like the BSP files Feb 18 13:35:52 hmm? Feb 18 13:35:56 grub invaders? Feb 18 13:36:02 on htcdream?? Feb 18 13:36:06 leviathan, no Feb 18 13:36:12 leviathan, hi I've a fix for alsa Feb 18 13:36:18 leviathan, is mono ok? Feb 18 13:36:30 uhm Feb 18 13:36:36 why? Feb 18 13:36:48 because mono fixes buffer underruns Feb 18 13:36:54 s/fix/workarrounds Feb 18 13:37:03 yes Feb 18 13:37:07 its just a workaround Feb 18 13:37:09 sound is cristal clear Feb 18 13:37:13 with mono Feb 18 13:37:17 yes Feb 18 13:37:21 but mono is crap Feb 18 13:37:24 I've also a fix for removing your dirty hack Feb 18 13:37:29 ok Feb 18 13:37:33 thats interesting Feb 18 13:37:42 can you please push it? Feb 18 13:37:42 yes but it's the same in the driver Feb 18 13:37:44 ok Feb 18 13:37:56 I'll clean a bit and push Feb 18 13:38:13 mono is done with /etc/asoundrc Feb 18 13:38:54 leviathan, which branch should I push to? mine or yours Feb 18 13:39:20 uhm, mine Feb 18 13:42:14 ok Feb 18 13:43:19 GNUtoo: i think .32 already has that fixed, no need for custom bootloaders or anything. Feb 18 13:43:33 ok thanks Feb 18 13:46:36 GNUtoo: didnt you push yet? Feb 18 13:46:42 I didnt see anything... Feb 18 13:47:00 I'm working on it Feb 18 13:47:04 ok Feb 18 13:47:14 I have to import stuff I did else it will be very dirty Feb 18 13:47:19 lots of debug stuff etc... Feb 18 13:47:22 ok Feb 18 13:47:30 take your time Feb 18 14:25:52 leviathan, pushed Feb 18 14:26:07 Weiss: tested that segfault on drm kernel, with old xserver Feb 18 14:26:24 Weiss: with upgraded libdrm it still works Feb 18 14:26:39 Weiss: so problem is in either mesa-dri or xserver itself Feb 18 14:27:13 Weiss: but I cannot upgrade xserver without mesa (due to dependency) and if I upgrade only mesa-dri package on system with old xserver I get this error Feb 18 14:27:16 Weiss: (EE) AIGLX error: glamo exports no extensions (/usr/lib/dri/glamo_dri.so: undefined symbol: __driDriverExtensions) Feb 18 14:28:19 leviathan, also try that: http://pastebin.com/d5485541f Feb 18 14:38:21 doh, openembedded seems like political kindergarden Feb 18 14:39:39 spaetz: what did you expect to find? Feb 18 14:40:09 nature of OE almost guarantees that Feb 18 14:45:30 GNUtoo: ok Feb 18 15:22:47 GNUtoo|oeee: hi Feb 18 15:22:53 following thing: wifi Feb 18 15:23:08 seems that they are using wrong frequency Feb 18 15:23:17 so the wifi chip is unable to initialize Feb 18 15:23:21 and gives an error Feb 18 15:23:22 ... Feb 18 15:23:52 leviathan: who are using wrong frequency? Feb 18 15:24:18 uhm wl1251, GNU/linux on htcdream ;) Feb 18 15:24:55 ah, so openmoko people are safe :) Feb 18 15:25:42 http://pastebin.com/m77fd3a56 Feb 18 15:25:50 look thats it Feb 18 15:25:52 uhhm Feb 18 15:26:02 nope: its openmoko on the htcdream now ;D Feb 18 15:26:13 okeee, well SHR Feb 18 15:26:22 but I keep SHR calling OpenMoko too Feb 18 15:26:28 ^^ Feb 18 15:27:01 the important is the idea, not the name or the person itself Feb 18 15:27:32 that you have at the end of the day something done, which is worth the time of your life, you've invested Feb 18 15:32:14 GNUtoo: http://pastebin.com/m77fd3a56 Feb 18 15:32:46 line 189 Feb 18 15:32:52 theres the bug Feb 18 15:32:53 :) Feb 18 15:35:00 ok Feb 18 15:35:36 Weiss: that segfault is AFAIK from DRI2ScreenInit maybe introduced by this commit http://cgit.freedesktop.org/xorg/xserver/commit/?id=04a54f69a8085ab3fe11a8713bd8b6b16ed1db27 Feb 18 15:39:03 hm. fso-abyss is not responding Feb 18 15:39:49 Weiss: simple patch and I have Xorg again :) - dri2info.version = DRI2INFOREC_VERSION; Feb 18 15:39:52 + dri2info.version = 3; Feb 18 15:41:13 Weiss: IIRC we already talked, that xf86-video-glamo shouldn't follow upstream so closely (by setting his version with included constant) Feb 18 15:41:56 hmm? Feb 18 15:42:41 mickeyl: fso-abyss seems segfaulted during that: 2010.02.18 18:37:01.59 ogsmd INFO : responding OK Feb 18 15:42:42 2010.02.18 18:37:01.69 ogsmd.modems.ti_calypso INFO Requesting new channel from 'fso-abyss' Feb 18 15:43:01 Weiss: IIRC there was very similar error before Feb 18 15:43:26 Weiss: and I asked why is the version set from dri2proto header file through DRI2INFOREC_VERSION constant Feb 18 15:43:51 Weiss: as dri2proto can change without needed changes in ddx code Feb 18 15:44:32 mickey|office: ^^ Feb 18 15:45:16 ah, I see.. we should declare the version we use, and not pretend it's the latest Feb 18 15:45:25 I wonder where that dependency came from Feb 18 15:46:28 Weiss: yes, that's how I understand it.. as that DRI2ScreenInit function handle different versions differently (and doesn't segfault) Feb 18 15:47:36 Weiss: we probably talked about it before http://git.bitwiz.org.uk/?p=xf86-video-glamo.git;a=commit;h=7ba80e79f33439d5a97fd20a69c1ff159e7a6440 Feb 18 15:49:03 hmm yes.. I think I remember Feb 18 15:49:50 btw with this fixed I get the same error as with older xserver before.. (EE) AIGLX error: glamo exports no extensions (/usr/lib/dri/glamo_dri.so: undefined symbol: __driDriverExtensions) Feb 18 15:50:21 Something wrong with frameworkd and abyss. Feb 18 15:50:23 2010.02.18 18:46:16.821 ogsmd.modems.ti_calypso INFO Requesting new channel from 'fso-abyss' Feb 18 15:50:24 2010.02.18 18:46:26.711 ogsmd.channel INFO : initializing Feb 18 15:50:24 2010.02.18 18:46:26.731 ogsmd INFO : responding OK Feb 18 15:50:24 2010.02.18 18:46:26.741 ogsmd.modems.ti_calypso INFO Requesting new channel from 'fso-abyss' Feb 18 15:50:24 2010.02.18 18:46:27.901 ogsmd.channel INFO : initializing Feb 18 15:50:25 2010.02.18 18:46:27.919 ogsmd INFO : responding OK Feb 18 15:50:28 2010.02.18 18:46:27.930 ogsmd.modems.ti_calypso INFO Requesting new channel from 'fso-abyss' Feb 18 15:50:30 2010.02.18 18:46:33.6 ogsmd.modems.abstract.channel WARNING UNHANDLED INTERMEDIATE: OK Feb 18 15:50:32 2010.02.18 18:49:29.330 frameworkd.resource INFO setting resource status for GSM from enabling to disabling Feb 18 15:50:35 2010.02.18 18:49:30.301 frameworkd.resource INFO setting resource status for GSM from disabling to disabled Feb 18 15:50:38 sorry Feb 18 15:50:40 missed the win Feb 18 15:51:16 Weiss: every other mesa-dri driver exports that.. so I'll add it too Feb 18 15:53:29 sure.. need to check what the values mean though Feb 18 15:56:05 Weiss: http://gitorious.org/mesa/mesa/commit/1ac166895fef47806c9e9286d2a6356b4db8398d Feb 18 15:56:29 Weiss: in that upstream commit it's explained a bit Feb 18 15:57:44 ok, cool Feb 18 15:58:05 I wonder if the core does magic that means it currently supports DRI1 Feb 18 16:05:09 Weiss: after this change new and old xserver started ok (with glamo driver - that error only forced Xorg to fallback to swrast which is much better than segfault) Feb 18 16:05:40 Weiss: so if you can merge this last commit and publish new libdrm, I would push my patches to oe.dev Feb 18 16:06:56 Weiss: well I also should check that something is actually drawn on screen.. now just checking xorg log from work.. Feb 18 16:46:32 ffs power out Feb 18 16:46:33 :| Feb 18 16:47:05 q-master, what fix? I think you are talking about something mrmoku did :P Feb 18 16:47:14 mrmoku|away, did you merge it with master? Feb 18 17:00:26 hi Feb 18 17:01:06 I see there are ubifs images for SHR? What are the advantages/drawbacks compared to jffs2? Feb 18 17:01:48 drawback: much more experimental Feb 18 17:02:33 advantage: not sure, but I think it does better wear-leveling and mounting is faster Feb 18 17:03:43 hm ok Feb 18 17:29:02 Hey all... I'm trying to make my first bitbake recipe for SHR :) (just some silly small test app to get to know the toolchain). Now, I had an error in the install stage of my recipe, but now bitbake won't do the package from scratch again. How can I force it do redo all of my package tasks? Feb 18 17:30:06 beniwtv: add -c clean option Feb 18 17:32:03 beniwtv: bitbake -c clean and then again bitbake Feb 18 17:32:35 Heinervdm, ESI: Thanks Feb 18 17:47:14 heyho Feb 18 17:48:32 ah, hey Feb 18 17:55:30 :) Feb 18 18:01:31 Anyone noticed that 'file' in SHR reports it can't find any magic files? Feb 18 18:01:45 <[Rui]> hi, pin is not showing up again (I did no updates) Feb 18 18:04:17 <[Rui]> 2010.02.18 18:04:29.77 ogsmd.modems.ti_calypso CRITICAL Error alocating channel ogsmd.call : org.freesmartphone.GSM.MUX.NoChannel: Modem does not provide this channel. Feb 18 18:06:28 woot! My recipe builds and makes a nice .ipk :) Feb 18 18:11:08 ThibG: I find UBI, at least on 2.6.32, to have almost no resilience against unclean shutdowns Feb 18 18:12:43 Weiss, hum ok Feb 18 18:13:00 btw, is there some fsck for ffs2 or something alike? Feb 18 18:16:55 mickeyl, what's the easiest way to know the framework's "rootdir" path from an extrenal script? Feb 18 18:19:37 <[Rui]> well, at least after the upgrade it now asked me for the pin... Feb 18 18:43:44 JaMa, i am tesing my bb recipes for the themes atm. is it ok if i email them to you if they work? Feb 18 19:02:16 phew :P Feb 18 19:02:21 freesmartphone.org: 03tom 07opimd-redesign * r7acfaaf23b89 10framework/framework/subsystems/opimd/convert_db/convert_db.py: Changed the script to work with the new type enforcement - i.e make sure fields are of a certain type Feb 18 19:02:22 freesmartphone.org: 03tom 07opimd-redesign * ra060b69851c3 10framework/framework/subsystems/opimd/ (TODO db_handler.py pimd_generic.py type_manager.py): Feb 18 19:02:22 freesmartphone.org: Added real type support (not only strings). Feb 18 19:02:22 freesmartphone.org: Made database creation more generic Feb 18 19:02:22 freesmartphone.org: Restricted adding reserved fields completely Feb 18 19:02:22 freesmartphone.org: Added specific python types per opimd type Feb 18 19:02:24 freesmartphone.org: 03tom 07opimd-redesign * rd94ff84f2902 10framework/framework/subsystems/opimd/ (convert_db/convert_db.py pimd_generic.py): Fixed a typo that caused conversion/opimd not to work Feb 18 19:02:24 freesmartphone.org: 03tom 07opimd-redesign * rc241dea41f61 10framework/framework/subsystems/opimd/ (TODO pimd_calls.py pimd_messages.py): Feb 18 19:02:24 freesmartphone.org: Unread messages count works Feb 18 19:02:24 freesmartphone.org: New missed calls count works Feb 18 19:02:25 freesmartphone.org: Updated TODO Feb 18 19:07:02 TAsn, do oyu have a working bb enviroment? Feb 18 19:07:22 YES Feb 18 19:07:28 oops Feb 18 19:07:28 yes Feb 18 19:07:29 :P Feb 18 19:08:28 ok, well i want to try my bb recipe for phoneui-theme Feb 18 19:09:32 TAsn, i have working bb enviroment aswell, but "do-fetch" fails.. so i want to know if it is ma fault or not. woudl you mind trying http://pastebin.org/93724 Feb 18 19:11:09 sec Feb 18 19:11:12