**** BEGIN LOGGING AT Tue Jun 07 02:59:57 2011 Jun 07 04:51:13 trying to build splashutils Jun 07 04:51:22 getting some error Jun 07 07:29:58 good morning Jun 07 08:24:04 pb_: my bad there was a mismatch wrt symbols. Now, looking at the code, all makes sense. Jun 07 08:29:33 if you mind to give me a final confirmation I was doing it sanely, logged efforts around 22pm are at http://logs.nslu2-linux.org/livelogs/kexecboot-prev.txt Jun 07 08:30:17 thanks in advance Jun 07 08:32:01 * ant_work moves back to EANCOM fun Jun 07 09:06:34 ant_work: right, that looks much more sensible. so, menu->current->current is wrong. Jun 07 09:07:08 in fact program freezes when entering first submenu... Jun 07 09:07:40 now Jay7x knows where to debug Jun 07 09:07:50 okay, very good Jun 07 09:08:13 still strange this crash happens *only* on poodle (3mMb) Jun 07 09:08:18 *32 Jun 07 09:09:40 probably the other devices with 64Mb have a different memorylayout sothat the misalignment is not fatal Jun 07 09:12:32 (same pxa255 I think) Jun 07 10:37:42 03Koen Kooi  07org.openembedded.dev * r46399f12a0 10openembedded.git/recipes/x-load/ (x-load-git/0001-Beagle-Rev-C5-support.patch x-load_git.bb): Jun 07 10:37:42 x-load: sync with meta-texasinstruments for beagle C5 support Jun 07 10:37:42 Signed-off-by: Koen Kooi Jun 07 10:39:43 c5, hm Jun 07 10:46:23 yes Jun 07 10:46:38 was discussed yesterday Jun 07 11:18:38 hi mickey|office Jun 07 11:19:00 surprised to find you in the office today Jun 07 11:20:01 how do I ask the folks in europe to pull from one of the mirrors and let the folks pushing in the early am use git.oe.org? Jun 07 11:20:20 ip mask? Jun 07 11:20:52 I could turn off git protocol and make people pushing use ssh. Jun 07 11:21:08 and let the mirrors handle pulls that way Jun 07 11:21:31 do you still see high overhead? Jun 07 11:21:46 what I see is pushes being blocked or slowed down considerably waiting for slots to work. Jun 07 11:22:03 thath's even worse Jun 07 11:22:18 while 10-30 people are being piggies at the trough trying to pull. Jun 07 11:22:47 yeah, I think blocking git protocol on the master server isprobably a good idea Jun 07 11:24:38 we put in these mirrors so .eu wouldn't have to wait... Jun 07 11:24:47 but everyone still uses git.oe.org Jun 07 11:26:51 have you had any luck getting new hardware out of the board? sounds like that is maybe the real solution. Jun 07 11:27:08 so I guess I'll try again on the ML and see if we can get folks using the mirrors for pulling and leave the master for pushing. Jun 07 11:27:37 hardware is "coming" I think... Jun 07 11:28:15 however git daemon isn't good for dealing with this kind of load...which is why everyone uses mirrors for pulling and only pushes to master. Jun 07 11:28:32 and the master has hooks to keep the mirrors in sync. Jun 07 11:29:10 kernel.org doesn't let people pull EVER from master...only a mirror. Jun 07 11:29:31 otherwise changes would never get propigated. Jun 07 11:32:44 hello--------????????? Jun 07 11:32:47 hello--------????????? Jun 07 11:32:59 ++++++++++!!!!!!!!!!!!!!++++++++ Jun 07 11:33:41 ;) Jun 07 11:34:22 i am compiling a kernel for friendly arm mini6410 ,in which there i think it is probaly a bug Jun 07 11:34:35 the details Jun 07 11:34:58 WARNING: modpost: Found 1 section mismatch(es). Jun 07 11:34:58 To see full details build your kernel with: Jun 07 11:34:58 'make CONFIG_DEBUG_SECTION_MISMATCH=y' Jun 07 11:34:58 GEN .version Jun 07 11:34:58 CHK include/linux/compile.h Jun 07 11:34:58 UPD include/linux/compile.h Jun 07 11:35:00 CC init/version.o Jun 07 11:35:02 LD init/built-in.o Jun 07 11:35:04 LD .tmp_vmlinux1 Jun 07 11:35:06 init/built-in.o: In function `do_one_initcall': Jun 07 11:35:08 /home/rahul/linux-2.6.28.6/init/main.c:704: undefined reference to `__gnu_mcount_nc' Jun 07 11:35:10 init/built-in.o: In function `run_init_process': Jun 07 11:35:12 /home/rahul/linux-2.6.28.6/init/main.c:787: undefined reference to `__gnu_mcount_nc' Jun 07 11:35:14 init/built-in.o: In function `init_post': Jun 07 11:35:16 /home/rahul/linux-2.6.28.6/init/main.c:796: undefined reference to `__gnu_mcount_nc' Jun 07 11:35:18 init/built-in.o: In function `name_to_dev_t': Jun 07 11:35:20 /home/rahul/linux-2.6.28.6/init/do_mounts.c:76: undefined reference to `__gnu_mcount_nc' Jun 07 11:35:22 init/built-in.o: In function `rest_init': Jun 07 11:35:26 /home/rahul/linux-2.6.28.6/init/main.c:465: undefined reference to `__gnu_mcount_nc' Jun 07 11:35:28 init/built-in.o:/home/rahul/linux-2.6.28.6/init/calibrate.c:123: more undefined references to `__gnu_mcount_nc' follow Jun 07 11:35:31 make: *** [.tmp_vmlinux1] Error 1 Jun 07 11:35:33 root@rahul-desktop:/home/rahul/linux-2.6.28.6# echo $? Jun 07 11:35:35 2 Jun 07 11:35:37 i searched the net and found out that it is a bug Jun 07 11:35:46 very good Jun 07 11:35:58 plz suggest a suitable patch no for it Jun 07 11:36:27 sounds like what you really want is a machine support channel Jun 07 11:36:58 i am using kernel linux-2.6.28.6 ,suggest patch for this version Jun 07 11:39:13 ultimate_warrior, 1 minute of using google tells me that your toolchain is too new for a kernel that old. You need a kernel Jun 07 11:39:13 having 181f817eaaca4c1f8a9c265d339d2b96de8b245d. Jun 07 11:41:13 http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=181f817eaaca4c1f8a9c265d339d2b96de8b245d Jun 07 11:41:44 i am using tool chain version no 4.4.3 Jun 07 11:42:31 hello Jun 07 11:43:05 can anyone help me.... how splash image work on angstrom oe base image for mini2440... Jun 07 11:43:30 ultimate_warrior, SINCE you have 4.4.3 this thing applies to you and you have to switch to kernel containing that commit, OR switch to a 4.3 based toolchain Jun 07 11:46:31 sanket: #angstrom? since thats angstrom specific? Jun 07 12:04:31 ka6sox-away, I builded image using oe with angstrom distribution for mini2440... Jun 07 12:05:52 right, and thats a distribution question, not one for development of OE. Jun 07 12:07:50 yeah, exactly. check the topic for details of what this channel is not. Jun 07 12:08:08 ka6sox-away, actually I want to know how splash loading bar works.... Jun 07 12:17:34 sanket: check /etc/init.d/rcS script Jun 07 12:17:48 sanket: or even better check files which are inside of psplash package Jun 07 12:18:57 hrw, when run psplash I m getting the image... I want to know how loading bar work..... Jun 07 12:19:18 mlip: I am still facing the problem the patch is not matching succesfully Jun 07 12:22:36 sanket: bar is part of psplash Jun 07 12:23:03 sanket: sysvinit script runs script after script and sends % via pipe to psplash process Jun 07 12:23:32 ok, lunch time Jun 07 12:23:46 MY tool chain version no 4.4.3,i am using kernel 2.6.28 ,cant use higher version than that,plz suggest a patch for this version Jun 07 12:28:23 ka6sox-away: you may use shaping or throttling Jun 07 12:28:41 but disable git access is easier, sure Jun 07 12:28:51 pb_: hi pb_, yah, Sabine literally threw me out of the hospital. "you can't help me anyways as long as it doesn't start... so get to work and come back when something happens" ;) Jun 07 12:30:36 mickey|office, my wife did same thing. Jun 07 12:41:58 ultimate_warrior, once more because I had a laugh when you entered the chat ;) Jun 07 12:42:00 http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=log;h=181f817eaaca4c1f8a9c265d339d2b96de8b245d Jun 07 12:42:14 this thing shows that you need probably at least 2.6.31 to apply this patch Jun 07 12:42:53 so either downgrade you toolchain , OR switch to atleast 2.6.31 and apply the patch; you most certainly wont be able to patch the 4.4.3 toolchain until you find a real wiz. about this Jun 07 12:50:07 mlip : actually i am compliing the kernel for mini2440 ,which does not support any higher linux version than that Jun 07 12:51:05 and tool chain version of 4.3 is not currently available Jun 07 12:51:21 hm.. Jun 07 12:51:57 pb_: is it safe to cast int to unsigned long long from aligment point of view? Jun 07 12:52:15 then you somehow have to try to hack it, or try to get in touch with the ppl responsible for that => in a machine support channel/ml, like already suggested Jun 07 12:54:31 hmm, oe website just died for me.. :( Jun 07 12:56:13 ok, it's back... just a blip Jun 07 13:29:56 03sledz  07org.openembedded.dev * rbd6fa71311 10openembedded.git/recipes/linux/ (linux-2.6.24/ftdi-4n-galaxy.patch linux_2.6.24.bb): Jun 07 13:29:56 linux-2.6.24: add another 4N-GALAXY.DE PID to ftdi_sio driver Jun 07 13:29:56 Signed-off-by: Steffen Sledz Jun 07 13:59:24 just created this: http://www.openembedded.org/index.php/LayerIndex Jun 07 13:59:29 comments? Jun 07 14:00:44 nice. Jun 07 14:01:01 might want a link to gitweb or whatever too, browsable Jun 07 14:01:35 kergoth: there's a link in the first column, perhaps it's not obvious enough Jun 07 14:01:45 I'll split it out actually Jun 07 14:01:53 ah, right. i'm blind :) Jun 07 14:02:07 maybe i'm just used to other sites having separated links for repo / browse Jun 07 14:02:10 * kergoth scratches head Jun 07 14:02:35 looks great though, we definitely needed something like that Jun 07 14:03:11 bluelightning: I wonder if you should mention / cover poky/meta-yocto somewhere Jun 07 14:03:24 even though its not its own repo at this point Jun 07 14:03:44 hmm yes I wondered about that Jun 07 14:03:50 * kergoth thinks Jun 07 14:06:49 huh, i always thought namespace packages was a setuptools thing, didn't realize it was its own PEP and can be used without setuptools Jun 07 14:06:50 nifty Jun 07 15:05:39 bluelightning: now that you see it in shr, pls put meta-zaurus in meta-handelds Jun 07 15:10:09 bluelightning: can I add meta-shr do Distribution layers with not that it's actually included in meta-smartphone repo as one of layers? Jun 07 15:11:10 JaMa|Wrk: yes I think meta-shr should be added... do you have a wiki account? or should I add it? Jun 07 15:13:20 bluelightning: nice, looking very good. i'm sure a lot of people will find this helpful Jun 07 15:14:30 Now the question is, if someone looks there and says okay, i want this distro with that bsp layer, how do we best help them.. either we should really well document how to do it with the default setup, or perhaps we should link them to the various setup scripts (e.g. angstrom's, or shr's makefiles, or..) Jun 07 15:14:33 * kergoth ponders Jun 07 15:17:25 kergoth: that's also a good question Jun 07 15:18:16 I kind of hope eventually we can move away from people's custom scripts (i.e. find a way to provide the same functionality in a standard way) Jun 07 15:18:33 until then though we should point to them somehow from there Jun 07 15:20:40 or rather, where applicable point to some other page that describes the recommended procedure Jun 07 15:23:39 JaMa|Wrk: meta-shr added Jun 07 15:25:04 bluelightning: hopefully this evening I'll commit some minor defconfig changes for Z Jun 07 15:25:28 ant_work: cool Jun 07 15:25:59 what do you think if we move all to xorg-xserver? kdrive is rally stale :/ Jun 07 15:26:42 ant_work: xorg-xserver+xf86-video-fbdev does not support xrandr Jun 07 15:26:52 wasn't readded? Jun 07 15:27:32 JaMa|Wrk: ? Jun 07 15:27:47 ant_work: I don't know if it was removed at some point, but for now it does not work, Xfbdev does rotation fine Jun 07 15:28:10 ok, kdrive-fbdev then Jun 07 15:28:55 stanislav tested memory use and found big improvements in xorg Jun 07 15:29:06 so I'm hoping we can move soon Jun 07 15:30:30 bluelightning: i dunno, anytime you start to dabble in scm specifics there are problems, not everyone uses the same tools. even if upstreams use git, it doesn't mean a company won't throw it all in a clearcase view :) Jun 07 15:32:39 kergoth: well, we should use our fetcher mechanisms such that we support any SCM that that supports; IIRC someone posted a patch for clearcase support ;) Jun 07 15:33:08 point taken, we can't necessarily fit into everyone's workflow Jun 07 15:33:20 that's not going to happen. companies aren't going to be fetching from the network, a lot of them want everything local Jun 07 15:33:35 which means storing upstream layers in local scms, not necessarily local git clones either Jun 07 15:33:41 yeah, there are a lot of use cases to consider Jun 07 15:33:46 its tough to meet them all with a single implementaiton Jun 07 15:34:28 oh, i see what you meant now, bitbake is installed, fetches all the layers from their local clearcase view Jun 07 15:34:34 bluelightning: thanks, but it's meta-smartphone/meta-shr Jun 07 15:34:39 the thing is, clearcase uses especially are used to *living* in their clearcase views Jun 07 15:34:46 its likely they'll expect their whole workspace to reside there Jun 07 15:35:13 and clearcase provides built in mechanisms to pull this piece from this branch and that from there and bring it together into a local setup Jun 07 15:35:15 JaMa|Wrk: ah, ok... so the one I have added is not used anymore? Jun 07 15:35:18 * kergoth ponders Jun 07 15:35:42 bluelightning: nope, it was replaced by that shared meta-smartphone (see last commit there) Jun 07 15:37:04 JaMa|Wrk: ok, so would it be appropriate to list meta-shr and then say "see meta-smartphone above" ? Jun 07 15:37:19 or should we explode out all of these repos that have multiple layers? Jun 07 15:37:26 I'm starting to think the latter Jun 07 15:37:27 bluelightning: yes something like that Jun 07 15:37:36 later seems ok too Jun 07 15:38:10 bluelightning: btw if you decide to link "setup scripts" for distributions then there is for SHR: http://wiki.shr-project.org/trac/wiki/Building%20SHR Jun 07 15:38:19 JaMa|Wrk: yep found that :) Jun 07 15:38:54 I was also trying to find the angstrom one but there doesn't seem to be one specifically for the new oe-core-based version Jun 07 15:39:02 guess I will need to check with koen Jun 07 15:39:06 iirc there is Jun 07 15:39:36 kergoth: we are doing a combo layer tool to support the poky repo style system but we've limited it to git for now Jun 07 15:40:33 bluelightning: btw similar page for machine support would be great Jun 07 15:40:38 there's also Jeremy's tools that he posted yesterday, have been looking over those today Jun 07 15:41:08 bluelightning: list machine "PR" name, then name of config in OE and then list bsp layers supporting it Jun 07 15:41:13 JaMa|Wrk: list of machines with pointers to which layer, maintainers etc. you mean? Jun 07 15:41:21 right ok Jun 07 15:41:37 maybe also some sort of status field Jun 07 15:41:40 agreed that would be useful Jun 07 15:41:41 i really need to check my patch labels in gmail more often (I auto-archive them to keep the inbox trim) Jun 07 15:42:06 like I know about some people working on htcbravo and htcleo support for meta-smartphone/meta-htc Jun 07 15:42:30 but it's not pushed yet, so it could be listed there with layer, but as WIP Jun 07 16:48:32 Hey all. Jun 07 16:52:44 gm Jun 07 16:52:58 anyone know how to install final configuration files, e.g. /etc/ssh/sshd_config, that conflict with base packages? Jun 07 16:57:55 ka6sox-away: btw, change http://www.openembedded.org/index.php/Getting_started page to reflect mirrors Jun 07 16:59:14 btw, where to find oe git mirrors? Jun 07 17:00:04 Jay7: There's one on repo.or.cz and on github. Jun 07 17:00:58 I see only reference about http mirror on repo.or.cz Jun 07 17:03:45 Jay7, github has the 2 mirrors currently...I am working to setup 4 new ones that will become the primaries... Jun 07 17:04:08 ka6sox-away: reflect this on wiki please Jun 07 17:04:10 then git.oe.org can become the "master" that people push to... Jun 07 17:04:23 Jay7, on that Getting started page? Jun 07 17:04:47 imho, good idea to create mirrors page and reference to mirrors from getting started Jun 07 17:05:03 Jay7, agreed. Jun 07 17:05:06 and use one as default for getting started page Jun 07 17:07:12 we should tell them to clone the github. Jun 07 17:11:12 Jay7, could you please put in the github info? I"m fixing an infra issue now... Jun 07 17:25:14 Jay7, so you want to continue to use the git:// method? Jun 07 17:27:29 Jay7, okay updated. Jun 07 17:27:40 is berlios still the place to get bitbake? Jun 07 17:36:32 ka6sox-away: OE use bb from oe's git Jun 07 17:36:54 ka6sox-away: say me git clone urls available Jun 07 17:37:10 I'll fix wiki info then Jun 07 17:38:08 the url I gave you for openembedded is the right one... Jun 07 17:38:42 I think we might have a mirror of bb out on github...but maybe its a development branch... I don't know. Jun 07 17:40:17 bb isn't frequently pulled by users Jun 07 17:40:33 ya, here is the git URL for bb: git@github.com:openembedded/bitbake.git Jun 07 17:40:34 and it's smaller than full OE repo Jun 07 17:40:44 343MB for full OE. Jun 07 17:40:54 compressed. Jun 07 17:42:24 what .eu mirrors exists? Jun 07 17:42:29 or.cz? Jun 07 17:42:42 the cz + whatever github has... Jun 07 17:42:58 hm.. is github load-balanced? Jun 07 17:42:59 I am working on getting mirroring with 2 US + 2 EU. Jun 07 17:43:10 Jay7, pretty sure it is... Jun 07 17:43:10 or should I change url somehow? Jun 07 17:43:20 not seen any URL change for .eu Jun 07 17:43:33 the new ones are geoIP based. Jun 07 17:43:48 well, then I'll change getting started to use github mirror Jun 07 17:44:04 thanks...that will help a bit... Jun 07 17:44:07 (not sure it's ok from marketing POV).. Jun 07 17:44:25 angstrom-distribution already does that. Jun 07 17:44:47 I would love to get the final configuration soon...but its not here yet. Jun 07 17:45:00 that should be transparent for users. Jun 07 17:45:36 is it possible to change the rootfs before creating its image ? Jun 07 17:46:48 yes Jun 07 17:46:51 there is a hook Jun 07 17:47:04 grep the recipes/images dir Jun 07 17:47:19 Jay7, I understood the github was so people could clone and play with it. Jun 07 17:47:38 woglinde: thanks Jun 07 17:47:44 ka6sox and mirror Jun 07 17:48:16 btw, github have https access as well Jun 07 17:48:26 may be no need to specify repo.or.cz in getting started then Jun 07 17:48:41 just use https://github.com/openembedded/openembedded.git Jun 07 17:48:52 true. Jun 07 17:48:58 thats a RO checkout right? Jun 07 17:49:01 yes Jun 07 17:49:11 then for the FW challenged that should work. Jun 07 17:49:16 anyway to get write access they should ask :) Jun 07 17:49:32 right, pull from mirrors, push to master Jun 07 17:49:44 well, http proto is working ok as well Jun 07 17:49:53 kk Jun 07 17:50:01 so that can be method #2 Jun 07 17:50:04 why http? Jun 07 17:50:08 git is too Jun 07 17:50:13 because FW's can block git Jun 07 17:50:33 but most don't block 80 or 443 Jun 07 17:50:39 woglinde: github is working with git, http and https Jun 07 17:50:48 excatcly Jun 07 17:50:59 http should work everywhere Jun 07 17:51:09 even via proxy (I hope) Jun 07 17:51:34 * Jay7 remembers oops-cache proxy which fails to pass svn.. Jun 07 17:51:55 well.. Jun 07 17:52:13 I'll just add all 3 methods Jun 07 17:52:15 Jay7, which is why I have a ssl-svn kludge still for nslu2-linux. Jun 07 17:54:08 sometime today I'm converting git from sv (which is semi-borken) to inetd Jun 07 17:54:17 probably after .eu goes to bed. Jun 07 17:54:36 sv... is it runit/daemon-tools? Jun 07 17:54:48 git-daemon-run yes Jun 07 17:54:53 hm.. Jun 07 17:55:00 borken for 2yrs still. Jun 07 17:55:40 I know what to do but I want cbrake to get a fresh pull before I change it. Jun 07 17:56:13 cbrake, can you do that late late today? Jun 07 18:06:19 hm.. repo.or.cz seems outdated a bit Jun 07 18:06:29 seems it updated by cron, not by hooks Jun 07 18:11:21 http://wiki.openembedded.org/index.php/Mirrors Jun 07 18:11:28 check it and improve it ;) Jun 07 18:35:14 Jay7, that looks good... Jun 07 18:35:34 let me bold a few things...to make them more visible. Jun 07 18:37:31 Jay7, done. Jun 07 18:38:50 * woglinde is going to install win server 2008 Jun 07 18:38:51 *g* Jun 07 18:39:01 woglinde: oh Jun 07 18:39:12 [sh]it happens Jun 07 18:39:41 woglinde, my condolences. Jun 07 18:40:13 ? Jun 07 18:40:23 I need to test some SSO stuff with ad Jun 07 18:41:44 ka6sox-away: yes, I'll be here till 9PM EDT today, so just let me know when you're ready Jun 07 18:42:22 perfect...so about 8pm would be perfect...gives me an hour to work in it. Jun 07 18:43:09 it should just "happen" Jun 07 18:50:59 gogo Jun 07 19:26:36 woglinde: you referred to a hook to change the rootfs before creating the image... Jun 07 19:27:06 are you referring to the ROOTFS_POSTPROCESS_COMMAND? Jun 07 19:27:51 yes Jun 07 19:28:03 hi stefan Jun 07 19:28:11 woglinde: just checking. I was asking a similar question earlier. Jun 07 19:28:38 woglinde: I'm trying to add config files in as the last step. Is that the best way to do so? Jun 07 19:38:36 hi woglinde Jun 07 19:47:41 hi stefan_schmidt Jun 07 19:47:44 hi woglinde Jun 07 19:48:30 hi GNUtoo Jun 07 19:49:04 hi obi Jun 07 20:02:40 03Martin Jansa  07org.openembedded.dev * r357d78b99b 10openembedded.git/recipes/webm/libvpx.inc: Jun 07 20:02:40 libvpx: force arm mode to be able to link with ffmpeg where arm mode already is forced Jun 07 20:02:40 * v2: updated INC_PR after 0478950960e5c600efc77c25559fc7eee311f686 Jun 07 20:02:40 Signed-off-by: Denis 'GNUtoo' Carikli Jun 07 20:02:50 03Denis 'GNUtoo' Carikli  07org.openembedded.dev * r08fdc869f0 10openembedded.git/: Merge branch 'master' of git.openembedded.net:openembedded into org.openembedded.dev Jun 07 20:02:51 03Denis 'GNUtoo' Carikli  07org.openembedded.dev * rc33f30c9df 10openembedded.git/recipes/linux/ (linux-nexusone/defconfig linux-nexusone_git.bb): Jun 07 20:02:51 linux-nexusone: new recipe Jun 07 20:02:51 This kernel is based on cyanogenmod kernel and will have fixes for Jun 07 20:02:51 having more standard userland<->kernel interfaces. Jun 07 20:02:52 The defconfig was minimized with savedefconfig Jun 07 20:02:52 Signed-off-by: Denis 'GNUtoo' Carikli Jun 07 20:02:53 03Denis 'GNUtoo' Carikli  07org.openembedded.dev * r910b2a21a4 10openembedded.git/: Merge branch 'master' of git.openembedded.net:openembedded into org.openembedded.dev Jun 07 20:02:53 03Denis 'GNUtoo' Carikli  07org.openembedded.dev * r70c1923b61 10openembedded.git/conf/machine/nexusone.conf: Jun 07 20:02:54 Add new nexusone machine configuration Jun 07 20:02:54 Signed-off-by: Denis 'GNUtoo' Carikli Jun 07 20:04:23 ouch ouch Jun 07 20:04:26 what's that: Jun 07 20:04:30 http://cgit.openembedded.org/cgit.cgi/openembedded/commit/?h=org.openembedded.dev&id=910b2a21a44e9d6d17a51e24a1cf9352a64f3d7d Jun 07 20:04:53 03Jeff Lance  07master * r8e81951f20 10openembedded.git/recipes/ti/ (3 files): (log message trimmed) Jun 07 20:04:53 matrix_gui: Update SRCREV to revision 257 Jun 07 20:04:53 * Adding arm only multimedia apps to am37x Jun 07 20:04:53 * Wifi updates to am37x, am180x Jun 07 20:04:53 * Adding to am180x, am37x, am35x, beagle: Jun 07 20:04:53 * Oprofile example Jun 07 20:04:53 Signed-off-by: Jeff Lance Jun 07 20:06:13 910b2a21a44e9d6d17a51e24a1cf9352a64f3d7d is empty on my laptop Jun 07 20:06:28 I always tried to avoid such commits Jun 07 20:06:33 cherry-picking Jun 07 20:06:53 but here I forgett to cherry-pick Jun 07 20:08:06 stefan_schmidt, do you know git well? Jun 07 20:08:13 did I do a mistake? Jun 07 20:08:16 how are people building i686 sdks vs x86_64 sdks? Are they doing it with two unique hosts, one 32bit and other 64bit linux? Jun 07 20:08:38 would be nice if I could build a 32bit SDK on a 64bit system, but not clear how to do that Jun 07 20:09:22 tharvey: have you considered a 32-bit chroot? Jun 07 20:09:45 tharvey, I think there are ways Jun 07 20:09:54 darknighte, thats what I think people are using, I looked into it once but couldn't find good clean documentation on setting one up Jun 07 20:09:59 I didn't look into nativesdk yet Jun 07 20:10:32 woglinde, hi Jun 07 20:11:25 you could always checkout out http://www.debian-administration.org/articles/566 Jun 07 20:11:29 tharvey, btw do you know git well? Jun 07 20:11:41 It's Debian specific and uses the schroot tool, but it might work for your needs. Jun 07 20:11:56 GNUtoo: I'm no expert, but what's the issue? Jun 07 20:12:13 that: Jun 07 20:12:25 http://cgit.openembedded.org/cgit.cgi/openembedded/commit/?id=910b2a21a44e9d6d17a51e24a1cf9352a64f3d7d Jun 07 20:12:33 is it a problem? Jun 07 20:13:59 GNUtoo, I'm not that good with git, but what is your specific question? Jun 07 20:14:32 darknighte, thanks I'll look into chroot some more - I think thats the popular choice Jun 07 20:14:53 tharvey: once you get it working, it's pretty slick. Jun 07 20:14:54 tharvey, is that commit ok/normal? Jun 07 20:15:03 or is it an issue I created? Jun 07 20:15:14 hi gnutoo Jun 07 20:15:16 hi Jun 07 20:15:28 GNUtoo: not sure what's the concern. Jun 07 20:15:36 maybe woglinde would know Jun 07 20:15:54 woglinde, I forgott to cherry-pick on top of master and pushed that : Jun 07 20:16:01 http://cgit.openembedded.org/cgit.cgi/openembedded/commit/?id=08fdc869f0fad6aebde1eb317fb0e15e596e5d4f Jun 07 20:16:03 and that: Jun 07 20:16:11 http://cgit.openembedded.org/cgit.cgi/openembedded/commit/?id=910b2a21a44e9d6d17a51e24a1cf9352a64f3d7d Jun 07 20:16:51 is it a big issue? Jun 07 20:16:59 or is it normal in the git operations Jun 07 20:17:19 GNUtoo, commits look fine to me - can't speak to the content of the patch files Jun 07 20:17:27 I always tried to avoid that thing Jun 07 20:17:58 GNUtoo: did you mean to commit a subset of the files, or what? The first commit is pretty sizeable patch set, but otherwise, I don't know what the issue is. Jun 07 20:18:16 .dev and master seems to be out of sync Jun 07 20:18:16 basically I didn't meant to commit that Jun 07 20:18:20 ah ok Jun 07 20:18:27 ouch Jun 07 20:18:32 or what you mean? Jun 07 20:18:50 03Eyal Reizer  07master * r00ae0e2a48 10openembedded.git/recipes/ti/ti-wifi-utils_git.bb: (log message trimmed) Jun 07 20:18:50 ti-wifi-utils: install calibrator ini files to rootfs Jun 07 20:18:50 In order to perform wifi calibration the calibrator needs to be Jun 07 20:18:50 provided with a reference NVS file that matches the wifi chip and the Jun 07 20:18:50 radio module that is being used. Jun 07 20:18:50 This patch install these ini files to the root file system along with Jun 07 20:18:51 the calibrator itself. Jun 07 20:18:53 you can even cherry-pick from one of your branches Jun 07 20:18:55 basically I mean that commit with the following message are created: Jun 07 20:19:11 Merge branch 'master' of git.openembedded.net:openembedded into org.openembedded.dev Jun 07 20:19:16 and I wonder why Jun 07 20:19:23 I didn't create that commit Jun 07 20:19:32 it was created automatically Jun 07 20:19:46 I always cherry-picked to avoid that kind of commits Jun 07 20:19:51 but this time I forgott Jun 07 20:20:45 I think it means you pushed all the changes that you had pulled into your branch. Jun 07 20:20:45 I think git does that for you to bring the branches in sync - afaik its frowned upon as you loose the individual msgs with each commit, but people still do it anyway Jun 07 20:20:59 ok Jun 07 20:21:07 Essentially, it look slike you pushed the changes from git.openembedded.net:openembedded to .dev. Jun 07 20:21:07 I saw other people did that too Jun 07 20:21:19 .net??? Jun 07 20:21:37 * darknighte shrugs Jun 07 20:21:42 there is no such branch Jun 07 20:21:56 that's what the commit message says... Jun 07 20:22:32 I guess theses things are created with git pull Jun 07 20:22:44 since git pull applied fine Jun 07 20:22:47 and needed no rebase Jun 07 20:22:54 I git add+git commited Jun 07 20:22:56 and pushed Jun 07 20:22:56 sorry, I cut and paste to little, [the branch 'master' of git.openembedded.net:openembedded into org.openembedded.dev] Jun 07 20:24:20 Maybe you pulled from one branch (with an implicit merge) and pushed to another? e.g. master vs. .dev? Jun 07 20:24:24 anyway since no one knows and that a lot of people does it Jun 07 20:24:29 I guess it's not problematic Jun 07 20:24:35 no Jun 07 20:24:38 unlikely Jun 07 20:24:49 How do I tell my linux kernel recipe not to try and strip native executables? Jun 07 20:27:04 GNUtoo: where should i send the patch of framworkd Jun 07 20:27:46 lxsameer, to freesmartphone Jun 07 20:27:58 ask mrmoku Jun 07 20:28:50 GNUtoo: I wouldn't worry about it either way, as long as the "extra" files would eventually have been in .dev anyway. seems fine to me. Jun 07 20:29:11 ok Jun 07 20:30:49 lxsameer: patches to frameworkd go to the smartphones userland list... see http://wiki.freesmartphone.org/index.php/Community_infrastructure Jun 07 21:01:00 anyone recall what's the bitbake command to remove all remnants of a recipe? Jun 07 21:01:16 bitbake clean doesn't do it and I've forgotten the other. Jun 07 21:02:05 sorry, I meant to say the downloaded files, as well. Jun 07 21:03:17 darknighte, perhaps you need the -f to force it? Jun 07 21:03:47 clean will indeed blow away WORKDIR Jun 07 21:04:34 yes, but it leaves the tarball. in the sources dir. Jun 07 21:04:51 oh, well. I'll just do another fetch. Jun 07 21:05:08 hmmm... don't know of any way to clean that, just rm it, the md5 and the .lock Jun 07 21:05:23 distclean Jun 07 21:05:30 the distclean task is clean + downloaded sources removal Jun 07 21:05:39 good to know! Jun 07 21:05:50 kergoth: thx. I knew there was one out there. just had a mental block. Jun 07 21:06:47 np Jun 07 21:07:06 i always mixed it and mrproper up, mrproper was a terrible task that wiped all of DL_DIR even though the task was per recipe Jun 07 21:07:11 useless and confusing Jun 07 21:07:20 apparently distclean has existed since 2008 :) Jun 07 21:07:53 kergoth: good to know. has mrproper been removed? Jun 07 21:12:19 darknighte: yep Jun 07 21:13:08 kergoth: cool Jun 07 21:17:57 03Andrea Adami  07org.openembedded.dev * r6edca740aa 10openembedded.git/recipes/linux/linux-2.6.39/poodle/defconfig: Jun 07 21:17:57 linux: update 2.6.39 defconfig for poodle Jun 07 21:17:57 * remove CONFIG_PXA_SHARPSL_DETECT_MACH_ID=y Jun 07 21:17:57 * machine-id is passed by kexec Jun 07 21:17:57 * follow other Zaurus devices Jun 07 21:17:57 Signed-off-by: Andrea Adami Jun 07 21:25:54 hmm. any idea what might be causing bitbake to choke here: Jun 07 21:25:55 jenkins@dauber:/opt/jenkins/workspace/oe-buglabs-sw2.1-scratch$ bitbake bug-image-production Jun 07 21:25:58 Parsing recipes: 0% | | ETA: --:--:-- Jun 07 21:26:56 jconnolly, does it hang indefinately? that was a bitbake bug some time back - what ver of bitbake? Jun 07 21:27:05 sec tharvey Jun 07 21:27:20 bitbake-1.12.0.tar.gz Jun 07 21:27:40 ya, I think I had to move up to 1.12 maintenance release to fix that Jun 07 21:27:44 er branch Jun 07 21:28:19 interesting. this is on a build server with python 2.6.2 Jun 07 21:28:40 my host machine has 2.7.1+ and I don't have the issue. anyway, yet another reason to wipe and restore the build machine Jun 07 21:28:47 and stick with ubuntu LTR's Jun 07 21:29:04 03Joshua Lock  07master * r15f051220a 10bitbake.git/lib/bb/cooker.py: Jun 07 21:29:04 bitbake/cooker: don't error in prepareTreeData for unbuildable targets Jun 07 21:29:04 Set abort to False in prepareTreeData so that unbuildable targets do not Jun 07 21:29:04 raise an exception. Jun 07 21:29:04 (From Poky rev: 3c113f355bce012d55dac873189a410657016a62) Jun 07 21:29:04 Signed-off-by: Joshua Lock Jun 07 21:29:04 Signed-off-by: Richard Purdie Jun 07 21:29:13 looking for the reference.. I can't recall what hte issue was exactly - kergoth do you recall the bitbake hang while caching issue? Jun 07 21:29:13 03Saul Wold  07master * r3afe85a6cf 10bitbake.git/lib/bb/ (cache.py cooker.py): (log message trimmed) Jun 07 21:29:13 universe target: add a new target to collect all recipe targets Jun 07 21:29:13 This new universe target is not intended to be used for Jun 07 21:29:13 compiling or building everything, it use is for sanity checking Jun 07 21:29:13 and other tasks that need to find all targets. This does not Jun 07 21:29:13 exclude any broken or virtual targets. Jun 07 21:29:14 (From Poky rev: 28e7041a9c110be2ac5dea1eb1f55ca8f056111e) Jun 07 21:30:04 tharvey: the only issue i know of wasn't a hang, it was a db locking error. Jun 07 21:30:08 03Denis 'GNUtoo' Carikli  07org.openembedded.dev * ra83c1cf658 10openembedded.git/recipes/netbase/ (netbase/nexusone/interfaces netbase_4.21.bb): Jun 07 21:30:08 netbase: add support for nexusone Jun 07 21:30:08 Without that patch usb0 doesn't come up at first boot, Jun 07 21:30:08 and so we can't ssh into the phone Jun 07 21:30:08 Signed-off-by: Denis 'GNUtoo' Carikli Jun 07 21:30:13 jconnolly, I originally ran into it when moving to a new build system as well Jun 07 21:30:18 03Denis 'GNUtoo' Carikli  07org.openembedded.dev * rb86cc02703 10openembedded.git/recipes/images/shr-image.inc: Jun 07 21:30:18 shr-image: select devtmpfs for the nexusone Jun 07 21:30:18 Devtmpfs allows us to boot faster, and udev is replaced by the Jun 07 21:30:18 freesmarphone middleware. Jun 07 21:30:18 Signed-off-by: Denis 'GNUtoo' Carikli Jun 07 21:30:19 03Denis 'GNUtoo' Carikli  07org.openembedded.dev * rcb6c40ce36 10openembedded.git/recipes/freesmartphone/frameworkd_git.bb: Jun 07 21:30:20 frameworkd: bump revision to get nexusone configurations. Jun 07 21:30:20 Signed-off-by: Denis 'GNUtoo' Carikli Jun 07 21:30:20 03Denis 'GNUtoo' Carikli  07org.openembedded.dev * ra215eb3933 10openembedded.git/recipes/xserver-common/ (files/Xserver.nexusone.patch xserver-common_1.34.bb): Jun 07 21:30:20 xserver-common: add nexusone DPI Jun 07 21:30:21 Without that patch the incorrect default DPI makes applications Jun 07 21:30:21 look too small on the phone's screen. Jun 07 21:30:22 Signed-off-by: Denis 'GNUtoo' Carikli Jun 07 21:30:25 jconnolly, hi Jun 07 21:30:52 hi GNUtoo Jun 07 21:31:06 03Denis 'GNUtoo' Carikli  07org.openembedded.dev * r7cadae3dfb 10openembedded.git/recipes/xorg-xserver/ (2 files in 2 dirs): Jun 07 21:31:06 xserver-xorg-conf: add nexusone xorg.conf and bump PR Jun 07 21:31:06 Signed-off-by: Denis 'GNUtoo' Carikli Jun 07 21:31:59 jconnolly, kergoth here it is http://cgit.openembedded.net/cgit.cgi/bitbake/commit/?h=1.12&id=8885b2597e76daf13bdb10cb4351283161c8aff3 Jun 07 21:32:47 wow great tharvey thanks Jun 07 21:33:13 oh, that, that was worked around a long time ago, didn't realize you were talking about that one. that didn't have to do with the caching, but the parsing. Jun 07 21:33:38 if htat's not in a 1.12 release, we need to throw one out Jun 07 21:34:07 right... my memory was off but I recalled the symptom. may have been a long time ago but the issue exists with latest 'released' 1.12 bitbake Jun 07 21:35:12 btw for bug1 I've to backport a kernel commit, that should give us udev171 Jun 07 21:44:01 thanks tharvey that worked for me Jun 07 21:46:50 03Richard Purdie  07master * r314759a047 10bitbake.git/lib/bb/cache.py: Jun 07 21:46:50 cache.py: Sync what amounts to whitespace with bitbake in poky, pending merge of cache patches Jun 07 21:46:50 Signed-off-by: Richard Purdie Jun 07 21:47:01 03Liping Ke  07master * rf0f5350692 10bitbake.git/lib/bb/cache.py: (log message trimmed) Jun 07 21:47:01 cache.py: Refactory Current Cache implementation Jun 07 21:47:01 This patch is for refactorying current cache implementation, the Jun 07 21:47:01 main reason is for introducing extra cache fields requests for Jun 07 21:47:02 image creator as well as other users. The refactory parts include: Jun 07 21:47:02 Move cache data retrieve methods out of Cache Data Fields Jun 07 21:47:02 Definition. Since this retrieve methods will be shared for Jun 07 21:47:03 03Liping Ke  07master * r75d9add923 10bitbake.git/lib/bb/cache_extra.py: (log message trimmed) Jun 07 21:47:04 cache: Introduce extra cache class for image creator Jun 07 21:47:04 Extra RecipeInfo will be all defined in this file. Currently, Jun 07 21:47:05 Only Hob (Image Creator) Requests some extra fields. So Jun 07 21:47:05 HobRecipeInfo is defined. It's named HobRecipeInfo because it Jun 07 21:47:06 is introduced by 'hob'. Users could also introduce other Jun 07 21:47:06 RecipeInfo or simply use those already defined RecipeInfo. Jun 07 21:47:07 03Liping Ke  07master * rd50389ae69 10bitbake.git/lib/bb/ (cache.py cooker.py ui/hob.py): (log message trimmed) Jun 07 21:47:08 cache: Introduce new param caches_array into Cache impl. Jun 07 21:55:10 Is there a mercurial mirror of the oe git tree? Jun 07 22:00:05 pb_, ping Jun 08 01:02:30 03Eyal Reizer  07master * r7af43d821a 10openembedded.git/recipes/crda/crda_1.1.1.bb: Jun 08 01:02:30 crda: fix regulatory domain agent for building with libnl2 Jun 08 01:02:30 This patch allows building the wireless regulatory domain agent with Jun 08 01:02:30 libnl version 2. Jun 08 01:02:30 Signed-off-by: Eyal Reizer Jun 08 01:02:31 Acked-by: Koen Kooi Jun 08 01:02:31 Signed-off-by: Denys Dmytriyenko Jun 08 01:03:21 rats...now I am behind.... Jun 08 01:03:38 I should have turned it off.... Jun 08 01:04:28 cbrake, have you left yet? Jun 08 01:06:25 ka6sox-home: nope still here Jun 08 01:06:37 ka6sox-home: I did not run a full rsync backup, but I did pull oe Jun 08 01:06:42 ka6sox-home: which repos you changing? Jun 08 01:07:15 all of them...the server is going to be running inetd...but data shouldn't change at all..or access as far as users are concerned. Jun 08 01:07:34 this is to try to get more performance out of it. Jun 08 01:07:40 ka6sox-home: ok, go for it then Jun 08 01:07:56 okay, did you pull all of them? Jun 08 01:08:01 or just OE? Jun 08 01:08:04 ka6sox-home: no, I did not Jun 08 01:08:15 pull them all please....then I will shut it down. Jun 08 01:08:16 why don't we just copy the /home/git dir on the server Jun 08 01:08:37 okay let me shut it down...then you copy what we need please. Jun 08 01:09:22 ok Jun 08 01:10:11 okay git daemon is down. Jun 08 01:10:46 ka6sox-home: ok, copy is done. There is now a /home/git_2011-06-07 Jun 08 01:10:52 that should cover us Jun 08 01:11:19 ty...okay let me get rid of git-daemon-run and make it run from inetd. Jun 08 01:11:26 excellent Jun 08 01:11:39 * cbrake goes to tuck in kids Jun 08 01:15:52 mickey|office everything good? Jun 08 01:25:37 can someone please try to pull from git.openembedded.org please? Jun 08 01:25:47 I hope this works right. Jun 08 01:28:21 worked for me. Jun 08 01:28:31 * mwester dashes off for dinner Jun 08 01:28:32 mwester, thanks! Jun 08 01:28:35 enjoy. Jun 08 01:29:07 now I just need someone to push something and all is "good" Jun 08 01:35:56 has anyone got anything to push? Jun 08 01:47:26 ka6sox-home: I'll try a test commit Jun 08 01:47:39 thanks Jun 08 01:48:21 ka6sox-home: the git protocol is not working Jun 08 01:48:33 hmmm... Jun 08 01:48:35 ka6sox-home: git clone git://git.openembedded.org/bitbake Jun 08 01:48:39 okay let me check something... Jun 08 01:48:55 let me check inetd services. Jun 08 01:49:01 and see if the port is bound. Jun 08 01:49:26 ka6sox-home: ssh access is still working, that probably never went down Jun 08 01:49:50 ya, just the daemon...its being somewhat of a pill...jas. Jun 08 01:59:44 try now please? Jun 08 02:03:02 03Anand Balagopalakrishnan  07master * r7f68af9308 10openembedded.git/recipes/powervr-drivers/libgles-omap3.inc: (log message trimmed) Jun 08 02:03:02 libgles-omap3: Specify the Window System to be updated in powervr.ini Jun 08 02:03:02 * SGX driver looks at /etc/powervr.ini for the Window System to be used. Jun 08 02:03:02 This patch introduces a variable to specify which Window System should Jun 08 02:03:03 be specified in powervr.ini file. Jun 08 02:03:03 * By having this variable, it's easy to override the default Window System Jun 08 02:03:04 through a config file or amendment Jun 08 02:03:13 03Denys Dmytriyenko  07master * re992a948b8 10openembedded.git/recipes/qt4/ (qt-4.7.3/0001-wsegl2-support.patch qt-4.6.3.inc): (log message trimmed) Jun 08 02:03:13 qt-4.6.3: update qt's powervr graphics driver plugin to work with SGX DDK v1.6 Jun 08 02:03:13 * The wsegl2 patch is required to build Qt with the latest SGX PowerVR drivers Jun 08 02:03:13 * Moved the wsegl2 patch from qt4/qt-4.7.3 to qt4/files Jun 08 02:03:13 * Added the wsegl2 patch to qt-4.6.3 Jun 08 02:03:13 Note: This is the "massaged in" version of the original patch from Anand Jun 08 02:03:14 Balagopalakrishnan as it was mangled and couldn't be applied. Jun 08 02:08:04 so are they using ssh or git? Jun 08 02:24:15 okay fixed and working... Jun 08 02:39:54 03Eyal Reizer  07master * r0720d95fcb 10openembedded.git/recipes/wpa-supplicant/wpa-gui-e_git.bb: Jun 08 02:39:54 wpa-gui-e: build the latest version from git sources Jun 08 02:39:54 Builds wpa-gui for qt4-embedded from the latest wpa-supplicant (0.8x) Jun 08 02:39:54 git sources Jun 08 02:39:54 Signed-off-by: Eyal Reizer Jun 08 02:39:54 Signed-off-by: Denys Dmytriyenko Jun 08 02:50:23 03Siddharth Heroor  07master * r4c32184588 10openembedded.git/recipes/sysvinit/sysvinit_2.86.bb: Jun 08 02:50:23 sysvinit: bump PR for ti816x serial console change Jun 08 02:50:23 * A new inittab is needed, so bump PR to regenerate it. Jun 08 02:50:23 Signed-off-by: Siddharth Heroor Jun 08 02:50:23 Signed-off-by: Denys Dmytriyenko Jun 08 02:50:32 03Siddharth Heroor  07master * r22d9cb5c2c 10openembedded.git/conf/machine/include/ti816x.inc: Jun 08 02:50:32 ti816x machine: Fix JFFS Image command Jun 08 02:50:32 * Fixes a bug in the command passed for JFFS2 Image creation. Jun 08 02:50:32 Signed-off-by: Siddharth Heroor Jun 08 02:50:33 Signed-off-by: Denys Dmytriyenko Jun 08 02:50:34 03Siddharth Heroor  07master * r0985607b85 10openembedded.git/conf/machine/include/ti816x.inc: Jun 08 02:50:34 ti816x machine: Update for latest kernel version Jun 08 02:50:34 * In 2.6.37 the console is tied to ttyO2. Jun 08 02:50:35 Signed-off-by: Siddharth Heroor Jun 08 02:50:35 Signed-off-by: Denys Dmytriyenko Jun 08 02:50:57 * ka6sox-home pulls so he can test that... **** ENDING LOGGING AT Wed Jun 08 02:59:56 2011