**** BEGIN LOGGING AT Sat Aug 23 02:59:57 2008 Aug 23 07:46:05 * * OE Bug 4509 has been RESOLVED (FIXED) by raj.khem(AT)gmail.com Aug 23 07:46:07 * * task-native-sdk is missing cc1 Aug 23 07:46:09 * * http://bugs.openembedded.net/show_bug.cgi?id=4509 Aug 23 10:05:01 Hi there! Aug 23 10:05:51 I started bitbake qt4-x11-free and after a day of compiling I got this http://pastebin.com/d67120b46 :( Aug 23 10:16:27 ~seen khem Aug 23 10:16:28 khem was last seen on IRC in channel #oe, 1d 8h 49m ago, saying: 'Crofton: what does spartan have'. Aug 23 10:42:08 woglinde: hi Aug 23 10:43:59 hi mr_nice Aug 23 10:44:19 woglinde: do you have some time for me? Aug 23 10:44:52 woglinde: it's about staging dirs Aug 23 10:47:35 mr_nice yes I have a little bit time Aug 23 10:50:09 woglinde: I want to build aufs (it is a kernel module) for that I think I need the linux kernel staging dir which is located at bla/staging/simpad-angstrom-linux/kernel. If I use STAGING_KERNEL_DIR it points to arm-oabi-angstrom-linux/kernel which is wron Aug 23 10:50:10 g Aug 23 10:51:15 woglinde: is STAGING_KERNEL_DIR the wrong choice? Aug 23 11:01:13 mr_nice hm you could look at the madwifi driver Aug 23 11:01:38 woglinde: ok I will do so Aug 23 11:02:15 hm there is STAGING_KERNEL_DIR Aug 23 11:02:26 woglinde: yes Aug 23 11:02:38 woglinde: but it points to the wrong direction Aug 23 11:02:46 woglinde: on my machine Aug 23 11:03:52 hm Aug 23 11:04:11 mom Aug 23 11:04:20 woglinde: with hardcoding the direction I am able to compile the module :( Aug 23 11:04:41 hm Aug 23 11:04:44 her it says Aug 23 11:04:46 /devel/arm/oetmp2/staging/arm-oabi-angstrom-linux-uclibc/kernel Aug 23 11:04:53 which should be okay Aug 23 11:05:11 maybee you put something wrong in your local.conf? Aug 23 11:05:16 on my machine it says arm-oabi-angstrom-linux/kernel Aug 23 11:05:42 but the kernel staging dir is located at staging/simpad-angstrom-linux/kernel Aug 23 11:06:05 hm I took the local.conf from angstrom-distribution Aug 23 11:08:31 this is my local.conf http://pastebin.ca/1182345 Aug 23 11:09:04 mr_nice are you using stable or dev? Aug 23 11:09:10 woglinde: stable Aug 23 11:09:31 ah Aug 23 11:09:32 hm Aug 23 11:09:38 woglinde: I don't like the dev tree because I am mostly not able to build anything there Aug 23 11:09:39 sorry Aug 23 11:09:43 dont know about stable Aug 23 11:09:50 ???? Aug 23 11:10:02 I can build a lot off stuff Aug 23 11:10:06 the stuff I can build Aug 23 11:10:09 I try to fix Aug 23 11:11:06 hi laibsch Aug 23 11:14:45 woglinde: I prefer to use the older stuff :) it is less pain for me. Aug 23 11:16:48 stable has many bugs Aug 23 11:16:56 which no one will be fixing Aug 23 11:17:23 woglinde: well thats because of the too strict policy to get something into stable Aug 23 11:17:56 wogline: I have tried it many times and in most cases I only got one or no response Aug 23 11:18:15 woglinde: but you need two core devs to ak Aug 23 11:18:39 mr_nice hm maybee I am on the wrong list Aug 23 11:18:46 I never saw something of you Aug 23 11:19:34 woglinde: maybee I am using the wrong list? hm I use openembedded-stablebranch@openembedded.org Aug 23 11:20:11 hm yes Aug 23 11:20:17 I am not subscribed to this Aug 23 11:20:24 I think many others too Aug 23 11:20:32 so its clear you didnt get acks Aug 23 11:21:42 woglinde: hm, better to cc that stuff to oe-devel ? Aug 23 11:22:05 mr_nice could be a quick solution Aug 23 11:25:52 woglinde: I will do that in the future. maybe it improves the famosity of the stable mailing list, too Aug 23 12:07:20 mr_nice: If you sent stuff to oe-stable and no response, that is an issue in itself that you should raise Aug 23 12:08:42 I've repeatedly said that it is a problem if OE does not live up to the standards it seemingly offers Aug 23 12:09:18 That being said, my seldom observations seem to indicate that mails to oe-stable are indeed answered in a timley fashion Aug 23 12:09:33 If that was not the case for you, that would be a new situation. Aug 23 12:15:36 * RP wishes python supported cross compiling better :/ Aug 23 12:16:47 RP, gm Aug 23 12:17:12 Who is the meta-toolchain guru? Aug 23 12:17:55 hi Crofton Aug 23 12:17:59 probably me Aug 23 12:20:25 Laibsch: hm, I consider about writing to oe-devel about it but I think it is like woglinde sayd that most devels don't subscribe the stable mailing list Aug 23 12:21:08 well, in that case and if presents problems, we would need to revise policy Aug 23 12:21:41 Even though I don't use stable, I would help out with it if the bugtracker was somehow integrated into the workflow Aug 23 12:21:48 I deal with the bugtracker anyhow Aug 23 12:22:01 I certainly don't want to deal with another mailling list Aug 23 12:22:20 Show me the mails you sent that were left unanswered Aug 23 12:23:47 RP, can you look into something for me? Aug 23 12:24:12 Crofton: If you mention the problem I may be able to help but no promises ;-) Aug 23 12:24:24 I'm hunting for the email :) Aug 23 12:24:39 I hope it is easy for some how knows their way arounf the bb file Aug 23 12:26:05 Laibsch: it was only one unanswered mail until now http://lists.linuxtogo.org/pipermail/openembedded-stablebranch/2008-August/000150.html koen normaly cares a lot about the stable list Aug 23 12:26:46 Laibsch: here is one which needs another dev to ack http://lists.linuxtogo.org/pipermail/openembedded-stablebranch/2008-August/000148.html Aug 23 12:27:06 send a ping to the list? Aug 23 12:27:26 hm Aug 23 12:27:28 Laibsch: ? Aug 23 12:27:28 I'm reluctant to ack things I do not follow closely Aug 23 12:27:35 * woglinde wonders how to build uclibc_nptl Aug 23 12:27:40 so I am part of the problem Aug 23 12:28:15 mr_nice: a "bump" in forum speak, a reminder Aug 23 12:29:40 rp, http://groups.google.com/group/beagleboard/browse_thread/thread/95085a9d0e8f0a68/f5764bc60cfdb028?lnk=gst&q=sdk#f5764bc60cfdb028 Aug 23 12:29:56 basically running the SDK on another machine does not work Aug 23 12:30:07 mr_nice: I don't see "systematic neglect", so I would say you need to take your chances with the current process and be persistent until you either loose interest or succeed Aug 23 12:32:27 Laibsch: hm, I will try it Aug 23 12:33:09 mr_nice, ping the list for a second ack Aug 23 12:33:27 if no one else does, I will likely do it since I trust koen Aug 23 12:33:50 but I do not want to get in the habit of using that as the basis for ack'ing things :) Aug 23 12:33:52 Crofton: ok, I will do so Aug 23 12:39:21 he florian Aug 23 12:53:56 ~seen thebohemian Aug 23 12:53:57 thebohemian was last seen on IRC in channel #oe, 1d 23h 53m 25s ago, saying: 'hi henning'. Aug 23 12:54:08 ~seen rschuster Aug 23 12:54:09 rschuster was last seen on IRC in channel #oe, 13d 18h 54m 58s ago, saying: 'woglinde: this is in openssl.inc for example. can't believe that no one else has seen this'. Aug 23 12:54:29 ~seen zecke Aug 23 12:54:29 zecke was last seen on IRC in channel #oe, 2d 22h 49m 20s ago, saying: 'cool'. Aug 23 12:55:54 rp, nag :) did you see the link to the SDK problem Aug 23 12:58:08 Crofton: I did, then I got distracted about 5 times since then :( Aug 23 12:58:13 RP, mickey|zzZZzz: Can you take a look at the cia bot commit hook? Aug 23 12:58:26 Laibsch: monotone or git? Aug 23 12:58:36 ok, just wanted to make sure you didn't ask for it again in five minutes Aug 23 12:58:36 not sure how it works now Aug 23 12:58:46 although I can find it easily now :) Aug 23 12:58:48 RP: there are no cia messages here Aug 23 12:59:20 Laibsch: Thats usually a CIA problem then and we probably need to check the status of the CIA servers Aug 23 13:00:10 * Crofton wishes it was not called CIA Aug 23 13:00:22 Crofton: Basically, yes the toolchain shouldn't have those paths in it. It sounds like we're missing some sdk packages somewhere Aug 23 13:00:48 Crofton: offhand I have no idea why, it needs a debugging session Aug 23 13:01:27 We don't have an mfpr-sdk package that I remember which would be the start of a pointer to the problem though? Aug 23 13:01:35 at least I can tel lthe guy someone who understands acknowledges there is a problem Aug 23 13:01:36 RP: http://cia.vc/stats/total/commits has commits from other projects from a couple of seconds ago Aug 23 13:01:54 Laibsch: but none from ours? Aug 23 13:02:01 no Aug 23 13:02:06 since wednesday, it seems Aug 23 13:02:18 then it sounds like a monotone server issue Aug 23 13:02:41 http://cia.vc/stats/project/openembedded Aug 23 13:02:51 OK, that means koen or mickey? Aug 23 13:03:21 what update the thing that updates the git mirror? Aug 23 13:03:26 sakoman, was asking about that Aug 23 13:03:59 Laibsch, there should be a wiki page documenting these services, and who to ping when they fail Aug 23 13:04:53 Crofton: You mean from the cia services? Aug 23 13:05:06 I guess we should establish first we are indeed sending out those mails Aug 23 13:05:18 and git mirror Aug 23 13:05:20 It looks more like a problem on our end to me atm Aug 23 13:05:30 just a general OE it infrastructure page Aug 23 13:06:05 Unfortunately, the whole web site has been taken down as a response to the new wiki being brought live Aug 23 13:06:18 if it is not in the new wiki, I don't think it can be accessed Aug 23 13:06:27 I don't remember such a page in the wiki Aug 23 13:06:50 I'm suggesting we create one :) Aug 23 13:08:12 yes Aug 23 13:08:21 I suppport the suggestion ;-) Aug 23 13:10:43 then we can figure out how to mirror it of the website for viewing when the website is down :) Aug 23 13:11:42 RP: BTW, oe-commits is down, too Aug 23 13:12:02 so indeed it looks like an mtn problem Aug 23 13:12:33 mickey|zzZZzz: Can you take a look when you wake up? Aug 23 13:49:44 Well, I guess mickey is awake, but at FROSCON Aug 23 14:09:26 oe-commits is down? :D That would explain the lack of comments on some of my recent commits, then! Aug 23 14:58:25 I'm trying to build the latest stuff for collie on ubuntu, so I followed the OE wiki and grabbed org.openembedded.dev... When I run bitbake it fails with "cp: cannot stat `Angstrom-console-image-glibc-ipk-2008.1-test-20080823-collie.rootfs.tar.gz': No such file or directory" Aug 23 14:58:36 but there is a Angstrom-console-image-glibc-ipk-2008.1-test-20080823-collie.rootfs.tar Aug 23 14:58:47 any clue as to what I'm doing wrong? Aug 23 15:03:50 it runs for a long time b4 failing Aug 23 15:04:04 can you pastebin the test text leading up to the error? Aug 23 15:04:47 Crofton: I know the reason for the error Aug 23 15:04:59 ok, tell us :) Aug 23 15:05:00 probably give me a clue on pastebin? is that an IRC command? Aug 23 15:05:08 ~pastebin Aug 23 15:05:09 ~pastebin Aug 23 15:05:09 [~pastebin] A "pastebin" is a web-based service where you can paste anything over 3 lines without flooding the channel. Here are links to a few : http://www.pastebin.com , http://pastebin.ca , http://channels.debian.net/paste , http://paste.lisp.org , http://www.rafb.net/paste Aug 23 15:05:15 thesing's recent work for booting collie Aug 23 15:05:42 beattie_97211: You need something like "bitbake kexec-boot" or something like that Aug 23 15:05:49 one clue, if I gzip the file, and run bitbake again it completes Aug 23 15:05:54 I don't remember 100% what it was Aug 23 15:06:11 beattie_97211: You certainly don't want to try and pull tricks Aug 23 15:06:37 I know about kexecboot Aug 23 15:09:23 Well, there is something you need to bitbake and it will produce the file you need up above Aug 23 15:09:59 thesing has been gone from irc for a few days, I'd like to talk to him about smoothing over that glitch Aug 23 15:11:07 I hope I have this pasetbin thing right http://pastebin.com/d4cf3b679 Aug 23 15:12:52 I can probably limp along by running the gzip by hand Aug 23 15:16:02 Try "bitbake initramfs-kexecboot" Aug 23 15:16:14 I again advise you not to fidlle Aug 23 15:16:34 I did yesterday and it had a similar problem, I'll try again now Aug 23 15:17:55 actually yesterday I did "bitbake linux-kexecboot" I'll try yours just for grins Aug 23 15:18:40 I understadn the state of the kernel code for collie Aug 23 15:19:44 I honestly don't remember which target it was you needed to bake Aug 23 15:20:14 I understand no harm in trying I may juts have to wait for thesing to show up Aug 23 15:20:14 Laibsch, Crofton hello :) Aug 23 15:20:24 hi Aug 23 15:20:31 is it possible to create a blank wiki page without creating a link from an existing page? Aug 23 15:20:57 i've always just created the link first and then the wiki software will generate the new page for me.. Aug 23 15:22:08 gm Aug 23 15:22:13 kgilmer: sure Aug 23 15:22:21 thousands of ways Aug 23 15:22:23 here is one Aug 23 15:23:02 create the link on another page, click preview, click the red link for you new page, then just don't save your changes from the first step Aug 23 15:23:11 ah, nice Aug 23 15:23:15 or manipulate the URL bar directly if you know what you are doing Aug 23 15:23:21 that is also possible Aug 23 15:23:29 no that first suggestion is good. Aug 23 15:23:29 thx Aug 23 15:28:10 Is there a policie or something for the name scheme of external modules? I just created a aufs_cvs .bb file and wonder if I need to name it aufs_cvs_2.6.24.bb becuase it is for 2.6.24 only Aug 23 15:59:07 when you say "2.6.24 only", do you mean that this specific source code will only work with 2.6.24, not any older or newer version? if so, you should call it "aufs-2.6.24_cvs.bb" Aug 23 16:06:46 pb__: you need to modifiy the makefile for different kernelversions there is ${KERNEL_VERSION} and patching the file with bitbake code would be nice but I am not experienced enought to do that. so I decided to patch with a simple diff. Aug 23 16:08:05 ah, I see. well, I would really encourage you to figure out how to do the patching "properly" based on ${KERNEL_VERSION}. It isn't all that hard, there should be plenty of examples in the existing packages. Aug 23 16:11:52 pb__: hm, ok do you remember one? else I will serarch the packages for ${KERNEL_VERSION} Aug 23 16:13:21 not offhand, though I know that u-boot_git.bb does something similar with flash sizes. Aug 23 16:16:01 pb__ : ok thanks Aug 23 22:51:18 Somebody there who can help me? http://pastebin.com/d67120b46 Aug 23 23:41:05 Come on Guys! Somebody step up, kick my ass and tell me how to do it right. :) Aug 23 23:42:11 I don't know where I have to look for the mistake I've made Aug 23 23:42:17 I took a quick look,, Aug 23 23:42:55 I'm not real familiar with building qt Aug 23 23:43:35 what is the rcc4 program? Aug 23 23:43:56 I don't know :( Aug 23 23:44:11 I just installed bitbake and openembedded Aug 23 23:44:23 made the local.conf file Aug 23 23:44:41 and started bitbake qt4-x11-free Aug 23 23:44:58 can you try something simple like console-image forst Aug 23 23:45:07 the qt stuff can be painful Aug 23 23:45:16 at least this is my uinderstanding Aug 23 23:45:23 yeah I see :( Aug 23 23:45:42 but what means console-image forst? Aug 23 23:45:58 aahh Aug 23 23:46:02 first Aug 23 23:46:09 and how to do that? Aug 23 23:48:57 actually, did I bitbake a hole image? Aug 23 23:49:21 I only want to cross-compile the qt libraries Aug 23 23:54:22 I don't think Aug 23 23:54:26 try asking on the list Aug 23 23:56:14 what list? Aug 23 23:59:31 oe dev list Aug 24 02:31:06 man its late/early **** ENDING LOGGING AT Sun Aug 24 02:59:56 2008