**** BEGIN LOGGING AT Fri Oct 19 03:00:01 2012 Oct 19 03:01:01 Can I ask here about Open webOS for GNex building problems? Oct 19 03:06:51 isn't that a question itself? Oct 19 03:07:52 ~asktoask Oct 19 03:07:52 hmm... asktoask is Just ask your question and if someone's around, they'll be glad to help, if they can. Oct 19 03:08:06 =P this is the right place. i know nothing about it though, if you ask i'm sure others here will help eventually Oct 19 03:08:17 +1 ^ Oct 19 03:10:57 Thanks. I have fresh Ubuntu 12.04 on VM and trying to build webos-image Oct 19 03:11:37 Yesterday there was not enough space in hard drive, so I added a new hard drive and continued Oct 19 03:12:33 Today I got another error: "open CFI at the end of file; missing .cfi_endproc directive" Oct 19 03:12:57 And GCC stopped. I think that's memory problem..but not sure Oct 19 03:13:23 hmm, I haven't seen that error before Oct 19 03:13:55 google has some results though Oct 19 03:15:06 Premist, how far did the build go before you got that error? Oct 19 03:15:44 does it say anything about an internal compiler error or ICE ? Oct 19 03:16:46 The bulid process stuck on webkit-webos something something on 1817/4559 Oct 19 03:16:55 No error outputs Oct 19 03:17:04 oh that' really far Oct 19 03:17:12 This is my full output when I got an CFI something error http://pastebin.com/6n1eUGRP Oct 19 03:17:32 webkit-webos is probably the most resource-intensive build Oct 19 03:17:45 it takes about 3.8GB DRAm to link Oct 19 03:18:14 Premist, how much memory did you assign to the VM? Oct 19 03:18:28 First time, 1GB. I expanded to 2GB Oct 19 03:18:37 Maybe I should get physical Ubuntu machine... Oct 19 03:18:44 "arm-webos-linux-gnueabi-g++: internal compiler error: Killed" Oct 19 03:19:09 that looks very much like the VM ran out of memory and the OOM killer reaped g++ Oct 19 03:19:29 Premist, is it possible to assign 4GB (or more) to the VM? Oct 19 03:19:39 you also need a couple of GB of swap Oct 19 03:19:51 Because my host machine has 4gig memory, not possible Oct 19 03:20:03 otherwise, a physical machine would be better (if it has at least 4GB DRAM) Oct 19 03:20:27 My MacBook Air has 8GB RAM, so I can move VM to my Air and continue jobs Oct 19 03:21:00 that might work - I would give the VM 4GB RAM + 2GB swap or 6GB RAM Oct 19 03:21:04 :-\ Oct 19 03:22:10 Thanks for your help :) Oct 19 03:22:27 that might work - I would give the VM 4GB RAM + 2GB swap or 6GB RAM Oct 19 03:22:36 Premist, let us know how it goes Oct 19 03:22:59 Got it Oct 19 03:39:39 Premist: Physical machine helps a lot. VM is tough. Oct 19 03:40:04 penk: Very nice. Oct 19 03:40:34 Roy__: thanks :D Oct 19 03:42:56 * penk planned to get a Nexus 7 Oct 19 03:55:45 hi I am getting this error while building openwebos http://pastebin.com/D8jyZPQ1 Oct 19 03:58:30 harsha, how much memory does your build environment have? Oct 19 04:00:43 Can I build open webos at OS X? My host machine has 4GB of RAM so I think it would be better.. Oct 19 04:03:04 Premist, not natively in OSX, you would need to run a VM on the OSX box, parallels or something Oct 19 04:03:29 Hmm OK. Oct 19 04:04:24 @jacques: 1 GB in virtualbox Oct 19 04:05:18 harsha, that's not nearly enough to link webkit-webos - it requires 4GB RAM to link, plus a couple GB of swap (or 6GB RAM) Oct 19 04:06:31 @jacques: Oh is that why I am getting error for webkit !! Oct 19 04:06:53 harsha, yes Oct 19 04:32:37 I just tried to join http://webos-ports.org but I got this error http://i.imgur.com/B3eAf.png Oct 19 04:32:41 Maybe database problems? Oct 19 04:59:56 Premist, talk to scoutcamper|away when he's awake. Oct 19 05:06:39 Premist, taking a look now Oct 19 05:06:57 Premist, you were trying to register an account when you got that? Oct 19 05:20:24 Premist, it's fixed Oct 19 06:08:54 cryptk, Thanks. Oct 19 06:11:37 jacues, trying to build with 5GB RAM + 2GB swap. http://i.imgur.com/kuHi2.png Oct 19 06:24:42 morning Oct 19 06:25:11 m0rn Oct 19 06:42:32 Premist, if you get past webkit-webos do_compile you are probably good. Oct 19 07:02:44 looks like there are other little port suprises hidden in LunaSysMgr :) Oct 19 07:03:44 I think I'll just fork webos-ports for the moment, we'll see later how to merge my little modifs Oct 19 07:09:04 jacques, task 1817 completed, now installing (do_install) Oct 19 07:09:26 Premist, good Oct 19 07:10:54 Thanks for your help. I'll notify when I finished building. Oct 19 07:11:41 Premist, you're welcome. I'm going to sleep now so I probably won't reply, but I'll see it in the morning. Oct 19 07:11:51 Yep Oct 19 09:55:21 oh wow isis browser is slow Oct 19 11:45:47 why do meta-webos-ports have a higher priority than the meta-smartphones layers ? The latter do have more specialized versions, don't they ? Oct 19 12:04:50 Tofe: bbappends are applied in order of layers in bblayers.conf Oct 19 12:05:01 Tofe: priority are for overlayed recipes Oct 19 12:05:07 (same recipe in multiple layers) Oct 19 12:05:15 which is not the case for meta-webos-ports Oct 19 12:11:40 Tofe: read http://patches.openembedded.org/patch/21841/ for detail Oct 19 12:11:41 s Oct 19 12:14:56 JaMa: well... I fear this is not the case: I introduced a luna-sysmgr.bbappend in my acer layer, which is after meta-webos-ports and meta-webos in the bblayers.conf file. My bbappend was parsed before the one in meta-webos-ports... and I solved the issue by changing the priority of my acer layer. Oct 19 12:15:22 Anyway, even for overlayed recipes, my question would apply, I think Oct 19 12:15:50 meta-smartphone was used mostly for SHR distribution Oct 19 12:16:06 meta-webos-ports is layer to customize that for webos-ports needs (and webos distro) Oct 19 12:16:18 so it makes sense to have higher priority then meta-smartphone Oct 19 12:16:58 Tofe: check bitbake -e luna-sysmgr to see how luna-sysmgr gets parsed with meta-acer before and after meta-webos-ports Oct 19 12:17:51 I'll try. When I introduced my bbappend, bitbake told me this: Oct 19 12:17:51 NOTE: ['/media/Volatile/OpenWebOS/webos-ports-env/webos-ports/meta-smartphone/meta-acer/recipes-extended/luna-sysmgr/luna-sysmgr.bbappend', '/media/Volatile/OpenWebOS/webos-ports-env/webos-ports/meta-webos-ports/recipes-webos/luna-sysmgr/luna-sysmgr.bbappend'] to ['/media/Volatile/OpenWebOS/webos-ports-env/webos-ports/meta-webos-ports/recipes-webos/luna-sysmgr/luna-sysmgr.bbappend'] Oct 19 12:18:14 now I think it means my bbappend was parsed before the other one Oct 19 12:18:34 (sorry for the looong paths) Oct 19 12:18:35 and check order of BBPATH variable Oct 19 12:18:41 from -e output too Oct 19 12:19:31 ah you're right with order of bbappends Oct 19 12:19:38 I always forgot that.. http://lists.linuxtogo.org/pipermail/openembedded-core/2012-July/026401.html Oct 19 12:19:52 Yeppee ! Tofe 1 - JaMa 548 Oct 19 12:21:08 Ah, and I now have a full (maybe working?) build with my a500 machine Oct 19 12:21:14 0.5 because they are parsed in order of BBPATH in case priority is the same, so I wasn't completely wrong :) Oct 19 12:21:26 :) Oct 19 12:21:50 but use machine override to "force" your .bbappend to win Oct 19 12:22:18 it's safer way then priority/BBPATH order Oct 19 12:22:41 here it was not possible: I wanted to use my fork of the luna-sysmgr, so I had to override the SRCURI variable itself Oct 19 12:22:53 or can I use SRCURI_a500 ? Oct 19 12:23:33 (I had to fork it to add my a500 machine in the source code, unfortunately) Oct 19 12:23:55 in general case you should never use MACHINE variable or MACHINE as override in recipe which is not MACHINE_ARCH Oct 19 12:24:24 but because luna-sysmgr is already using MACHINE variable (making it MACHINE_ARCH from sstate checksum POV) Oct 19 12:24:36 then it's "ok" to use SRC_URI_a500 Oct 19 12:24:51 and PACKAGE_ARCH = MACHINE_ARCH should be added to upstream luna-sysmgr recipe Oct 19 12:25:02 anyway the future real solution will be to use the same source code for this package :) Oct 19 12:25:17 right now everything with qt4-webos in dependency tree should be PACKAGE_ARCH = MACHINE_ARCH Oct 19 12:25:39 ok Oct 19 12:25:48 and move all MACHINE specific settings to separate recipe luna-sysmgr-config Oct 19 12:26:03 to keep luna-sysmgr TUNE_PKGARCH Oct 19 12:26:10 and only small luna-sysmgr-config MACHINE_ARCH Oct 19 12:26:34 but for that to work we have to fix qt4-webos, webkit-webos and all other broken recipes in meta-webos :/ Oct 19 12:26:49 quite some work, eh Oct 19 12:27:18 true.. Oct 19 12:27:33 it's better to force opengl for whole armv7a PKGARCH Oct 19 12:27:49 then use MACHINE variable to enable/disable opengl Oct 19 12:28:13 downside is that every armv7a device we would like to support must come with opengl support Oct 19 12:29:29 well, all current devices support opengl, it's just that the drivers are not open Oct 19 12:30:31 yes, that's what I meant.. in other words every BSP we would like to support must provide opengl support in some way Oct 19 12:31:06 yep Oct 19 12:31:36 but because MACHINE is used also inside qt source repo it's not only about fixing metadata : Oct 19 12:48:02 JaMa: I could build right target boost binaries for tuna, interesting :) Oct 19 12:52:29 panda-z: just rebuilding that fixed it? Oct 19 12:54:32 JaMa: no, i'm not rebuilding it. i'm using the webos-ports repo and building for tuna. Oct 19 12:55:45 panda-z: so did you have to change something to fix it or not? Oct 19 12:56:15 ah it was for qemuarm before Oct 19 12:56:23 so it works for tuna but not for qemuarm? Oct 19 12:57:10 JaMa: yes, when using build-webos, it generate native binaries for qemuarm Oct 19 12:57:46 and by "webos-ports repo" you mean our webos-ports-setup scripts right? Oct 19 12:57:58 JaMa: yup Oct 19 12:58:16 ah ok, understood now Oct 19 12:58:31 well build-webos is weird in many ways Oct 19 13:00:33 now i think it(openembedded) works well with zsh, though i don't have enough space to link webkit :( Oct 19 13:00:45 "could not bind GL_ES API" ? Oct 19 13:01:31 panda-z: not enough space for swap? Oct 19 13:02:27 ah, eglBindAPI failed. How sad. Oct 19 13:02:42 JaMa: my 120GB partition is full filled with building stuffs for qemuarm and tuna Oct 19 13:03:21 needs to extend more space :P Oct 19 13:03:42 panda-z: you can enable rm_work to save some space in WORKDIR Oct 19 13:05:13 JaMa: you mean add INHERIT += "rm_work" to local.conf? Oct 19 13:06:06 panda-z: yes, it should be there already, just uncomment it Oct 19 13:07:42 JaMa: that make sense, ty Oct 19 13:32:15 morning all Oct 19 13:32:42 did someone ping me in here…? colloquy says they did, but the scrollback disagrees Oct 19 13:33:18 ShiftyAxel: 20:45:46 < PatrickC_TP> rwhitby, ShiftyAxel|Away, ka6sox-away: updated Phoenix cut-copy-paste repo. https://github.com/phxdevices/LunaSysMgr Oct 19 13:33:41 ty JaMa Oct 19 13:34:37 ShiftyAxel: btw there were some changes in upstream LunaSysMgr repo, maybe you want to update webos-ports repo with that Oct 19 13:34:50 so that our packaged version 0.930 matches to what upstream has Oct 19 13:35:11 ya, probably a good idea, but I think I've been breaking the build accidentally with my current setup, need to have a chat with morphis about it Oct 19 13:35:36 breaking the build as OE issues? Oct 19 13:36:24 git issues I believe, I'm not 100% sure what the problem is Oct 19 13:36:51 but ka6sox mentioned it to me a couple of days ago, and morphis pinged me yesterday saying he needed a word Oct 19 13:37:10 so I'm kind of worried that I'll make everything explode again if I start committing without getting my setup sorted out :P Oct 19 13:39:48 oh no, do I have to recompile webkit-webos each time I modify a bit the libgles-a500 package !? that will be a very long convergence... Oct 19 14:01:07 Tofe: yes with OEBasicHash enabled.. Oct 19 14:01:27 Tofe: but you can force OEBasic for development Oct 19 14:01:38 how can I do that ? Oct 19 14:01:41 ShiftyAxel: I can help you with it if you want Oct 19 14:02:16 Tofe: add BB_SIGNATURE_HANDLER_forcevariable = "OEBasic" to your local.conf Oct 19 14:02:33 JaMa: that would be good, though knowing what the problem is requires morphis or ka6sox I believe Oct 19 14:03:21 wasn't it about rebasing that repo? Oct 19 14:03:45 possibly, but I've got the hang of rebasing now Oct 19 14:03:59 I think it has something to do with luna-sysmgr/webOS-ports/master Oct 19 14:04:04 it would cause http://issues.webos-ports.org/issues/259 if you do rebase it Oct 19 14:04:16 without changing SRCREV in meta-webos-ports Oct 19 14:04:39 this sounds complex Oct 19 14:05:08 not really, rebase will change git hash of that commit you had in SRCREV before Oct 19 14:05:15 so if you rebase then you have to update SRCREV Oct 19 14:05:37 if you merge upstream branch, then you still have to update SRCREV to get new commits Oct 19 14:05:47 but old SRCREV will stay valid Oct 19 14:09:39 right, so where is SRCREV located? Oct 19 14:09:53 currently I'm only committing to webOS-ports/luna-sysmgr Oct 19 14:11:42 ShiftyAxel: bbappend in meta-webos-ports Oct 19 14:12:17 ./recipes-webos/luna-sysmgr/luna-sysmgr.bbappend Oct 19 14:12:26 now there is Oct 19 14:12:27 SRCREV = "0671dd7e62b5104c4e427c47be68b5a8af0fd377" Oct 19 14:12:57 and it's using webOS-ports/master branch in webOS-ports/luna-sysmgr repo Oct 19 14:13:42 anyone knows how to get wifi ether address on webOS before that sign up wizard? Oct 19 14:14:06 ty, i'll take a look at it Oct 19 14:14:20 so whenever I commit a change, update SRCREV and all will be fine? Oct 19 14:16:01 you do just commit or rebase? Oct 19 14:16:42 looks like not possible to get MAC before signing in :/ http://forums.webosnation.com/hp-touchpad-tips-information-resources/292661-how-get-touchpad-mac-address.html Oct 19 14:19:43 I tend to create my commits and then rebase against upstream before pushing Oct 19 14:21:55 then yes you have to change SRCREV to whatever commit you want after every rebase Oct 19 14:22:04 usually the latest Oct 19 14:34:01 I have many "hooked symbols" when I try test_egl: is that normal ? Oct 19 14:34:10 doesn't look very reliable Oct 19 14:36:40 yes Oct 19 14:37:25 ok. But I guess the "Illegal instruction" at the end isn't wanted :p Oct 19 14:38:14 I'm quite unsure what libc.so should be used in there Oct 19 14:39:07 I have the one from my android, but it looks like tuna takes the one of binaries-android, which is built againt I-don't-know-what Oct 19 14:41:58 its waaaaaay too early Oct 19 14:44:01 morning ka6sox Oct 19 14:44:11 morning JaMa Oct 19 14:50:42 ~coffee ka6sox Oct 19 14:50:45 morning Oct 19 14:50:52 morning Oct 19 14:51:14 morning all... Oct 19 14:52:59 oh no, I have to rebuild libc.so Oct 19 14:57:49 good luck with that. Oct 19 14:58:02 i had to build libc on a gentoo box once…that was running on a pentium 2. Oct 19 14:58:41 well, here it is the android "bionic" libc, to be built against the android SDK Oct 19 15:04:29 I don't understand why this bionic libc couldn't as well be a simple .bb Oct 19 15:05:20 there is a RFP for dealing with external SDKs in OE that hasn't been accepted yet Oct 19 15:05:47 but here it is the libc, which depends on nothing: it could be made a simple package Oct 19 15:06:09 its still using an external SDK and toolchain. Oct 19 15:06:45 ah, it's not standard gcc arm-linux-eabi ... Oct 19 15:15:10 bb30m Oct 19 15:35:29 getting the whole ICS source code just for building my little patched libc... oh yeah. Oct 19 16:16:37 is there a simple reason that the nexus was chosen for porting? Oct 19 16:16:53 open source drivers avaliable. Oct 19 16:17:49 and sufficiently modern hardware Oct 19 16:18:09 thanks. Oct 19 16:18:22 and its carrier free Oct 19 16:23:14 open source drivers are like the holy grail Oct 19 16:23:22 hey Loudergood! Oct 19 16:23:24 long time no see Oct 19 16:23:50 Hello Oct 19 16:26:51 for some things we have FOSS drivers..others we need to wrapper Oct 19 16:40:13 is there a nice way to shut down a gnex running webos? reboot and halt don't work; had to pull battery Oct 19 16:40:33 welcome to pre-alpha... Oct 19 16:40:33 push the power button for 5sec ? Oct 19 16:40:58 nope Oct 19 16:41:13 tried that... pulling battery was the only thing that I found that helped. Oct 19 16:41:25 same here Oct 19 16:41:32 reboot sometimes works, but it's very rare Oct 19 16:41:34 "shutdown -h now" works like a charm for my tablet :) Oct 19 16:41:43 glad it's not just me :-D Oct 19 16:42:04 I'll try shutdown next time Oct 19 16:42:08 never had a reboot work.. Oct 19 16:42:49 in F17 I have to do "shutdown -P" if I want the machine to poweroff Oct 19 16:42:54 well, let's just humbly say that my port is more advanced than the nexus on that point ;) Oct 19 16:43:08 why they changed that, I do not know. Oct 19 16:44:34 jacques, webos-dev-image build completed without issue(s), and successfully flashed to my GNex. Oct 19 16:45:40 Premist, congratulations :-) Oct 19 16:49:35 Premist, congrats. Oct 19 16:55:44 * Tofe is building a toolchain for building basic tools for building libc with a little patch. Oct 19 16:58:05 wow awesome Premist! Oct 19 17:06:26 Hi All Oct 19 17:06:40 morning Herrie Oct 19 17:06:53 It's evening here LOL ;) But morning to you ;) Oct 19 17:07:01 ~ugt Oct 19 17:07:02 ugt is, like, Universal Greeting Time. Created in #mipslinux, it is a rule that states that whenever somebody enters an IRC channel it is always morning, and it is always late when the person leaves. The local time of any other people in the channel, including the greeter, is irrelevant. http://www.total-knowledge.com/~ilya/mips/ugt.html Oct 19 17:07:03 ~ugt Oct 19 17:07:08 lol Oct 19 17:07:20 oh snap, UGT rush! Oct 19 17:07:22 LOL Oct 19 17:07:43 seems its smart enough to deal with that. Oct 19 17:07:49 yeah Oct 19 17:07:57 I'm wondering where to start when I would want to port OWO to a device. Is there anywhere a 101 on building custom kernels, including drivers, how to use qemu etc? Oct 19 17:42:14 mornin Oct 19 17:42:34 hiya Oct 19 17:52:46 JaMa: ping? Oct 19 17:55:28 I have new stuff locally to be pushed, but whenever I push it also wants to update webOS-ports master, despite the fact that i've not done any work in there Oct 19 17:55:40 not really sure what's going on with that... Oct 19 18:02:15 ShiftyAxel: you need to specify local branch and remote branch Oct 19 18:22:01 panda-z: and then pull down the latest webOS-ports/master to get my local in line with the remote? Oct 19 18:23:40 ShiftyAxel: you can just push your working branch to remote Oct 19 18:25:45 git remote -v will show your remote repository Oct 19 18:27:26 ShiftyAxel: you can push your working branch like this: git push : Oct 19 18:30:22 cool, ty panda-z Oct 19 18:30:35 yw Oct 19 18:42:12 goonite Oct 19 20:11:03 ShiftyAxel: did you push your changes? I don't see them Oct 19 20:11:30 he was waiting Oct 19 20:12:08 for someone to make sure he didn't cause an issue Oct 19 20:12:35 ShiftyAxel: go for it, I'll check as soon as you push it Oct 19 22:09:02 111MB/s scp over gigabit is beautiful *sniff* Oct 20 02:17:54 folks Oct 20 02:18:56 bhuey Oct 20 02:20:03 * bhuey is catch up on sleep **** ENDING LOGGING AT Sat Oct 20 03:00:02 2012