**** BEGIN LOGGING AT Sun Aug 12 02:59:57 2007 Aug 12 03:19:12 Hi all. i am trying to build the latest version of org.openembedded.dev on an x86-64 machine with slamd64 linux. last week I had built opie-image with not much problems. but today I get this error when i updated and did a rebuild. http://cmh.hackndev.com/pastebin/29/ Aug 12 03:21:13 what should $OPIE_CVS_PV be ? Aug 12 03:21:25 Any suggestions ? Aug 12 03:22:17 BobOfDoom, any idea ? who is the right person to ping for this ? Aug 12 03:40:41 sakiv: have you got "require conf/distro/include/preferred-opie-versions-1.2.4-pre.inc" in local.conf? Aug 12 03:41:02 BobOfDoom, no i removed the last line Aug 12 03:41:26 preferred-opie-versions-1.2.4-pre.inc sets OPIE_CVS_PV Aug 12 03:42:05 BobOfDoom, in fact the local.conf says # EDIT THIS FILE and then remove the line below before using! Aug 12 03:42:06 #require conf/distro/include/preferred-opie-versions-1.2.3-pre.inc Aug 12 03:42:28 so I commented the line out. when i update OE, do I have to update the conf file as well ? Aug 12 03:43:04 ah.. well OPIE 1.2.3 was just released recently, so that means the version numbers have been bumped in a lot of places Aug 12 03:43:21 to build opie you will need some preferred-opie-versions file Aug 12 03:43:43 see http://www.linuxtogo.org/gowiki/OpieWithAngstrom Aug 12 03:43:55 I've been using 1.2.4-pre without problems Aug 12 03:43:58 ok Aug 12 03:44:36 ok i will try it out thanks a lot. i was just getting frustrated all day with trying to fix various errors in building opie Aug 12 04:44:29 i modified the preferred_provider of a packages in .bb recipe, but it has no effect. even after i update a new PR. Aug 12 06:44:22 03mwester 07org.oe.dev * r5ed171e3... 10/ (1 packages/openssh/openssh_4.6p1.bb): Aug 12 06:44:22 openssh 4.6p1: /var/run is volatile; removed this from the packages to avoid Aug 12 06:44:22 conflicts with the volatiles mechanism. Aug 12 06:55:27 good morning Aug 12 08:03:03 good morning all Aug 12 08:03:22 moorning koen Aug 12 08:22:00 good morning Aug 12 08:22:08 03koen 07org.oe.dev * rfeb4e917... 10/ (5 files in 2 dirs): webkit: rename to webkit-gtk Aug 12 08:22:12 03koen 07org.oe.dev * r910ef122... 10/ (3 files in 3 dirs): midori, openmoko-feedreader2: catch up with webkit rename Aug 12 08:22:24 hey pH5 Aug 12 08:23:33 hej koen, do you know why freetype-2.3.4 is DEFAULT_PREFERENCE="-1"? Aug 12 08:29:15 iirc small api change Aug 12 08:31:40 hi koen, pH5 Aug 12 08:31:47 hey Bernardo Aug 12 08:32:21 koen: ah, ok. just wondering because it's default in angstrom for so long. Aug 12 08:32:24 hi Bernardo Aug 12 08:32:37 pH5: because it had a major speedup Aug 12 08:33:01 but iirc some older versions of apps didn't compile against it Aug 12 08:37:17 morning all Aug 12 08:37:27 hey RP Aug 12 08:38:11 RP: http://rafb.net/p/bv6YZU80.html :) Aug 12 08:38:38 hi RP Aug 12 08:39:53 hi RP Aug 12 08:42:35 pH5: When you tested the depchains was that before or after my commit last night? Aug 12 08:42:53 pH5: I synced the depchains function with poky last night so some of the issues may or may not be fixed Aug 12 08:43:23 RP: that was before. I just got your changes this morning. Aug 12 08:43:37 koen: and did it work? :) Aug 12 08:44:17 pH5: ok, I'm not 100% sure if the issues you've mentioned are fixed or not. I suspect at least some are... Aug 12 08:45:57 RP: works perfectly Aug 12 08:48:28 RP: ok, will check. Aug 12 08:48:34 You didn't touch the ALLOW_EMPTY part, right? Aug 12 08:48:41 http://en.pastebin.ca/654302 <-- makes sense? Aug 12 08:48:58 I notice _tar and _rpm don't support ALLOW_EMPTY at all. Aug 12 08:49:36 pH5: I didn't touch ALLOW_EMPTY Aug 12 08:49:41 * Bernardo kicks makedepend-native Aug 12 08:50:04 pH5: I'm just replying to your mail, it seems we do add all the -dev and -dbg varients by hand in poky to PACKAGES Aug 12 08:50:08 anyone seen this: *** No rule to make target `makedepend.@APP_MAN_SUFFIX@', needed by `all-am'. Stop. Aug 12 08:54:52 03koen 07org.oe.dev * r2b515ec2... 10/ (1 packages/uclibc/uclibc_svn.bb): uclibc svn: fixes Aug 12 08:56:44 pH5: Yes, I think the ALLOW_EMPTY change makes sense Aug 12 09:06:46 hey mickeyl Aug 12 09:16:00 good morning! Aug 12 09:16:10 bbiab, breakfast Aug 12 09:21:30 * koen seperates his local diff into org.oe.dev.bfin, org.oe.dev.avr32 and org.oe.dev.commit Aug 12 09:21:53 that just leaves a load of SRCDATE -> rSRCREV changes Aug 12 09:23:41 morning mickeyl Aug 12 09:25:07 We only have 97 immediate expansions in OE which is quite nice :) Aug 12 09:29:07 03koen 07org.oe.dev * r0f0f2234... 10/ (4 files in 3 dirs): linux-linux-headers: add 2.6.22 Aug 12 09:29:11 03koen 07org.oe.dev * r6deaf509... 10/ (1 conf/distro/angstrom-2007.1.conf): angstrom: add FEED_ARCH for blackfin Aug 12 09:29:17 03koen 07org.oe.dev * r9ccdfedb... 10/ (1 site/arm-common): site/arm-common: add unaligned access for gstreamer Aug 12 09:37:09 hi mickeyl Aug 12 09:38:26 RP: the ALLOW_EMPTY change is committed. I'll try to cook up some -dbg/-dev/-doc package autogeneration based on PACKAGES. Would this be useful for more than task packages? Some distro could decide to split up -dbg packages, for example. Aug 12 09:40:37 RP: I'm not sure about the RDEPENDS_${PN} vs RDEPENDS issue. It seems to work as you said, so maybe I just looked at a -doc package instead of a -dev package accidentally. Aug 12 09:46:50 pH5: I think we should confine it to task.bbclass for now. If we want other uses we can consider that Aug 12 09:47:17 pH5: I've yet to be convinced split up -dbg packages are useful Aug 12 09:48:22 RP: should I replace PACKAGEFUNCS with something like "task_autopackages package_depchains emit_pkgdata" in task.bbclass? Aug 12 09:49:03 RP: say we split out a gui and a library out of a recipe Aug 12 09:49:17 RP: the libs depends on libc, the gui on bonobo Aug 12 09:49:45 I don't want to drag in bonobo + dbg just to debug the lib on e.g. a router Aug 12 09:51:00 pH5: No, I'd just add an anonymouse function that extends PACKAGES Aug 12 09:52:36 koen: We have the option of splitting dbg packages in cases like that. Those cases are not the common case... Aug 12 09:52:38 what are the things that currently hold up a bitbake 1.8.8 release? Aug 12 09:52:49 koen: Testing Aug 12 09:53:00 RP: do I have to make sure that the anonymous function is called before package_depchains (and if so, how)? Aug 12 09:53:27 pH5: anonymous functions execute immediately after the data is parsed Aug 12 09:53:32 RP: ok Aug 12 10:04:55 03pH5 07org.oe.dev * rfd98012f... 10/ (1 classes/package_deb.bbclass classes/package_ipk.bbclass): Aug 12 10:04:55 package_{deb,ipk}.bbclass: change ALLOW_EMPTY check Aug 12 10:04:55 - before an unset ALLOW_EMPTY meant False, while any string value meant True, Aug 12 10:04:55 now "1" means True and any other value or unset means False Aug 12 10:04:55 - this allows to set ALLOW_EMPTY = "1" and explicitly unset it for a Aug 12 10:04:55 subpackage with ALLOW_EMPTY_somepackage = "0" Aug 12 10:05:01 03koen 07org.oe.dev * r4a4c4562... 10/ (4 files in 2 dirs): uclibc: add blackfin configs Aug 12 10:06:26 03koen 07org.oe.dev * r1d466223... 10/ (1 packages/gcc/gcc-cross-initial_4.2.1.bb): gcc: add initial cross 4.2.1 Aug 12 10:24:02 huh, anonymous python functions cause parse errors if they are not indented with four spaces? Aug 12 10:24:35 pH5: yes Aug 12 10:52:17 RP: prototype: http://en.pastebin.ca/654379 Aug 12 10:53:06 we might need to add a mechanism to not create empty packages with empty RDEPENDS/RRECOMMENDS after package_depchains though. Aug 12 10:53:12 bbl Aug 12 10:53:32 pH5: ALLOW_EMPTY_${PN}-doc = "0" is missing Aug 12 10:53:56 koen: I built -doc task packages just for fun :) Aug 12 10:54:06 pH5: line 17 Aug 12 10:58:06 hi all Aug 12 11:04:28 pH5: Due to bitbake.conf, every -dev/-dbg will have some RDEPENDS/RECOMMENDS and must be present to fill in gaps in the chains Aug 12 11:07:00 pH5: The final version shouldn't have -doc in ;-) Aug 12 11:07:48 pH5: I suspect we should just simplify this and overwrite PACKAGES in task files... Aug 12 11:08:01 that reminds me Aug 12 11:08:10 pH5: The function then just needs to add -dev/-dbg versions to everything in PACKAGES Aug 12 11:08:13 one of these days, I'm writing a proper 'collect-.debug-stuff' task Aug 12 11:08:45 mickeyl: To workaround the deficiencies of the FILES_ variable? Aug 12 11:08:46 mickeyl: ubuntu has it in /usr/lib/debug//foo Aug 12 11:09:51 mickeyl: We could declare a FILES2_ which used proper regexps :} Aug 12 11:12:05 RP: that sounds very appealing Aug 12 11:12:15 [FILES2] Aug 12 11:12:22 not necessarily with this name though Aug 12 11:12:27 mickeyl: It would solve the problem neatly and give us a powerful new tool Aug 12 11:12:31 agreed Aug 12 11:12:43 mickeyl: right, a better name would be good, it was more the idea I wanted to convey :) Aug 12 11:12:43 there's more than just the .debug files to it Aug 12 11:12:57 almost everything turning into pluggable these days Aug 12 11:13:07 makes FILES unnecessarily complicated Aug 12 11:13:09 so I'm all for it Aug 12 11:13:13 Bernardo, ping Aug 12 11:13:39 mickeyl: The SRC_URI_OVERRIDES_PACKAGE_FUNC is still bothering me. I added a profile option (-P) to bitbake and its showed I do 67,000 file stat calls just parsing poky :/ Aug 12 11:13:55 mickeyl: about 60,000 of them are due to SRC_URI_OVERRIDES_PACKAGE_FUNC Aug 12 11:15:51 eeks Aug 12 11:15:55 that's _too_ much Aug 12 11:16:42 . o O (OE metadata ramdisk) Aug 12 11:16:53 mickeyl: yes. When you look at the code its understandable though. Its converting file:// uris into paths. To do that it goes through FILESPATH with a ton of os.path.exists() calls... Aug 12 11:18:14 (python's exists uses stat behind the scenes) Aug 12 11:18:15 RP: yeah. i doubt that we can simplify this a lot. Aug 12 11:18:50 mickeyl: 60,000 stats for 700 .bb files is too high though :/ Aug 12 11:18:58 oops, right Aug 12 11:19:00 poky is just 700 Aug 12 11:19:10 well, even for 4500 it's too much Aug 12 11:19:21 03rpurdie 07org.oe.dev * r5fa855df... 10/ (1 packages/tasks/task-maemo.bb packages/meta/meta-maemo.bb): meta-maemo: Split into a tasks package and remove uneeded immediate expansions Aug 12 11:19:39 .dev will be much worse, I just benchmarked poky as its quicker and will reveal the same issues Aug 12 11:19:42 but it's inherently O(bbfiles*#overrides), isn't it? Aug 12 11:20:15 mickeyl: basically, yes Aug 12 11:20:44 RP: are we losing much by going from python to C in that case? We might think about adding a C function that does all in one python call Aug 12 11:21:18 mickeyl: Its the stat calls themselves that take the time as its a syscall which might hit the disk Aug 12 11:21:59 At the very least you're throwing in 60,000 context switches :/ Aug 12 11:22:34 mickeyl: Out of interest we make about 5.6 million function calls in total parsing poky :) Aug 12 11:23:59 mickeyl: I tried to add a cache around the bb.which function (which is used to evaluate PATH variables) but only 12 calls are the same :/ Aug 12 11:26:25 mmh Aug 12 11:26:58 Another little problem is the "local" fetcher isn't splitting the ;patch=1 off the end of the filenames :/ Aug 12 11:27:09 * RP has no idea how it works at all Aug 12 11:30:41 03raj.khem 07org.oe.dev * rc9f63e9e... 10/ (1 packages/binutils/binutils-cross.inc): binutils-cross.inc: enable installation of libbfd Aug 12 11:30:45 03koen 07org.oe.dev * r2aa5f6c6... 10/ (6 files in 2 dirs): binutils cross 2.17.x: make use of binutils-cross.inc Aug 12 11:30:54 If we remove that, it can actually find the file and we're down to 47,000 stats... Aug 12 11:31:05 OE's unpack function hides this little bug... Aug 12 11:31:18 ~lart hiding bugs Aug 12 11:31:18 * ibot pulls out his louisville slugger and uses hiding bugs's head to break the homerun record Aug 12 11:34:40 "don't need any parameters for extraction, strip them off". Why would localpath give something with parameters? :/ Aug 12 11:35:18 who did this masterpiece? :D Aug 12 11:36:11 Should we consult bitkeeper? :) Aug 12 11:36:38 ya, i'm almost sure it wasn't me ;) Aug 12 11:38:04 mickeyl: kergoth in version 1.1 of base.bbclass :) Aug 12 11:38:16 hehehe Aug 12 11:38:21 wow, that's quite old Aug 12 11:38:37 The function hasn't changed much :) Aug 12 11:40:09 and I think I've just thought of a way we can avoid some of these stat calls without breaking anything Aug 12 11:41:00 mickeyl: When we switch to 1.8.8 I'll nuke that bit of base.bbclass and stop it masking problems... Aug 12 11:41:22 RP: ok, good plan Aug 12 11:45:59 and I have this down to about 7,000 stat calls :) Aug 12 11:47:39 14% speedup for poky :) Aug 12 11:48:07 It also took the number of path.join calls down from 60,000 to more like 7,000 Aug 12 11:51:35 maybe we need less meta-data :) Aug 12 11:52:05 Crofton|home: That was one of Poky's objectives hence its mere 700 files ;-) Aug 12 11:53:11 03rpurdie 07org.oe.dev * re8982cdc... 10/ (1 classes/base.bbclass): base.bbclass: Optimise the SRC_URI_OVERRIDES_PACKAGE_ARCH code so localpath is only called if its possible there are machine override matches. Also document some code hiding bugs for post 1.8.8 removal Aug 12 11:53:12 collections helps reduce meta-data proliferation Aug 12 11:53:54 I wonder if this some group of meta-adta we could split into a collection Aug 12 11:54:43 "code hiding bugs"? I think I'd better do something else for a while :) Aug 12 11:55:52 Crofton|home: I have wondered whether we should have a meta-opie, meta-gpe and meta-omoko since those data sets are fairly independent Aug 12 11:56:29 and maybe a meta-core... Aug 12 11:56:51 I was planning to put that on the agenda for OEDEM Aug 12 12:00:43 RP: at least a meta-toolchain having gcc, autotools and friends Aug 12 12:00:48 agreed Aug 12 12:01:51 that would help prevent mass bustage from a bad commit Aug 12 12:01:56 which is good and bad Aug 12 12:02:04 good in that reduces bustage Aug 12 12:02:17 bad in that people do not notice and fix as quickly Aug 12 12:02:25 * Crofton|home chants autobuilds Aug 12 12:02:45 autobuilds are wonderful Aug 12 12:02:52 * RP hugs the poky autobuilder Aug 12 12:02:59 we need to reduce dependence on people to find stupid problems Aug 12 12:03:25 so, who volunteers to fix seppuku.bbclass then? Aug 12 12:03:37 !oebug 2722 Aug 12 12:03:39 * * Bug 2722, Status: NEW, Created: 2007-08-08 04:08 Aug 12 12:03:40 * * : Seppuku.bbclass triggers recursion error Aug 12 12:03:41 * * http://bugs.openembedded.org/show_bug.cgi?id=2722 Aug 12 12:04:07 what is that anyways? Aug 12 12:04:23 koen: When did that start happening? Aug 12 12:04:23 the class that files bugs for autobuilds Aug 12 12:05:00 RP: 5 minutes before I filed the bug and shortly after I fixed the python error Aug 12 12:05:27 koen: bitbake related or just some python problem? Aug 12 12:05:41 wrong indentation of an if statement Aug 12 12:09:18 03rpurdie * r877 10/ (4 files in 4 dirs): fetch/local.py: Fix local fetcher's localpath return values (remove parameters) Aug 12 12:09:37 mickeyl: all the openmoko autobuild buggreports are generated by seppuku,bbclass Aug 12 12:09:54 03rpurdie * r878 10/ (2 files in 2 dirs): bb/__init__.py: Improve which function Aug 12 12:10:00 koen: good. then i like it :) Aug 12 12:13:39 mickeyl: it's broken now, as #2722 details Aug 12 12:16:30 koen: Its the bb.note lines Aug 12 12:16:59 koen: Its incompatible with the changes in 1.8 where notes trigger the event handlers, hence the recursion Aug 12 12:17:48 what should I change it too? Aug 12 12:17:52 s/too/to/ Aug 12 12:19:44 koen: I'll commit a fix Aug 12 12:20:29 It still won't work but you'll get a message saying why ;-) Aug 12 12:23:57 koen: pushed Aug 12 12:24:05 thanks Aug 12 12:27:46 03koen 07org.oe.dev * rb3274bff... 10/ (1 packages/binutils/binutils-cross.inc): binutils-cross: fix typo Aug 12 12:27:51 03rpurdie 07org.oe.dev * r68653f0e... 10/ (1 classes/seppuku.bbclass): seppuku.bbclass: Avoid recursion problem (#2722) Aug 12 12:37:15 hi all Aug 12 12:37:18 koen: The only remaning package.bbclass patch from poky is the runstrip one. I'll leave that to you Aug 12 12:37:21 hi steliosk Aug 12 12:38:59 RP : hi! Aug 12 12:43:38 * steliosk wonders why people want mug-shots with presenter short bios... Aug 12 12:48:57 * * OE Bug 2722 has been RESOLVED (FIXED) by rpurdie(AT)rpsys.net Aug 12 12:48:59 * *  Seppuku.bbclass triggers recursion error Aug 12 12:49:01 * * http://bugs.openembedded.org/show_bug.cgi?id=2722 Aug 12 12:57:12 steliosk, how is the presentation? Aug 12 13:00:19 steliosk, we shoudl try and remove the C++ linking hack b4 your talk :) Aug 12 13:06:55 Crofton|home : i have send the draft, final should be mail by the end of the month Aug 12 13:07:42 great Aug 12 13:07:49 Crofton|home : what's the "c++ linking hack "? Aug 12 13:07:54 looks like a good conference program Aug 12 13:08:10 well, for c++ apps we need to add -lstdc++ to bb file Aug 12 13:08:15 for ppc Aug 12 13:08:18 but not OSK Aug 12 13:08:27 with libtool-1.5.10 Aug 12 13:08:36 of course, 1.5.24 breaks this hack Aug 12 13:08:45 which is my hidden motive Aug 12 13:09:07 but my argument is hat we should not require custom settings based on target Aug 12 13:09:10 does the efika/angstrom build from .dev works or not ? Aug 12 13:09:17 well, it works Aug 12 13:09:35 but every c++ package needs the -lstdc++ line Aug 12 13:09:39 only for efika Aug 12 13:09:54 i have seen a strange issue with asterisk 1.4 also Aug 12 13:10:00 c++ related? Aug 12 13:10:05 nop Aug 12 13:10:12 what is the issue? Aug 12 13:10:44 does asterisk use libtool? Aug 12 13:10:47 If you try to build asterisk there is an app that needs to run on the host (if ever) Aug 12 13:11:00 it builds on arm,xscale etc etc Aug 12 13:11:02 libtool links with -nostdlib Aug 12 13:11:08 fails on ppc Aug 12 13:11:22 because its try to link the host app with target libs Aug 12 13:11:34 same bb, no changes etc Aug 12 13:12:09 which should not be happening at the first place Aug 12 13:12:20 hmm Aug 12 13:12:40 I think we need to review the tool chain construction carefully Aug 12 13:12:49 but given your c++/linking problem makes me think that there is a problem and it manifests it self Aug 12 13:12:55 yeah Aug 12 13:13:18 I do not understand why the line libtool feeds to gcc fails, since it is very close to teh OSK line that succeeds Aug 12 13:14:33 maybe gcc was not build with the exact same switches (not very possible) or the ppc expects something else Aug 12 13:14:38 I need to get breakfast and shoower Aug 12 13:15:06 yeah, I guess we should start reviewing gcc-cross carefully Aug 12 13:15:15 Hopefully that is the correct package Aug 12 13:16:04 Can you try to use another binary toolchain with OE and see if the problem shows again ? Aug 12 13:16:24 Crofton|home : i.e denx Aug 12 13:16:32 I'll try and look into this Aug 12 13:16:43 do we have instructions on how to do this? Aug 12 13:17:06 cbrake has posted a blog/wiki about it Aug 12 13:17:09 ok Aug 12 13:17:12 mickeyl, RP: http://bugs.openembedded.org/show_bug.cgi?id=2733 Aug 12 13:17:17 Crofton|home: sharprom is one example, there is also support for it in poky Aug 12 13:17:28 I'll add to my procrastination tasks Aug 12 13:17:42 thesis is job #1 Aug 12 13:17:57 I say that in case my wife ever figures out irc logs are online :) Aug 12 13:18:08 koen: The logs are missing? Aug 12 13:18:09 and I hope the leading space supresses logging Aug 12 13:18:10 heh Aug 12 13:18:19 Crofton|home: ;-) Aug 12 13:18:20 RP: yes, but it created a bug! Aug 12 13:18:21 bbiab Aug 12 13:18:34 koen: :) Aug 12 13:25:57 koen : Did you try to rebuild asterisk 1.4 after the last changes ? Aug 12 13:26:04 yes Aug 12 13:30:24 koen: What was the status of the irex bugs? Aug 12 13:31:03 RP: I only merged the machine.conf Aug 12 13:31:15 the kernel and distro still need to get cleaned up and added Aug 12 13:32:36 koen: 2663 is therefore fixed? Aug 12 13:33:16 RP: yes, I don't care much about the renaming Aug 12 13:35:35 where is cbrake's blog? Aug 12 13:36:20 Crofton|home: http://bec-systems.com/web Aug 12 13:37:14 RP: seppuku.bbclass:330: bugname = "noprovider for %s runtime: %s" % (event.getItem, event.getisRuntime) Aug 12 13:37:53 koen: ? Aug 12 13:37:55 RP: that gives 'AttributeError: NoProvider instance has no attribute 'getisRuntime' Aug 12 13:37:56 ' Aug 12 13:38:09 is that a bitbake bug or a supuku bug? Aug 12 13:38:16 (it works when used with -b) Aug 12 13:40:20 koen: Its a seppuku bug, there are separate events for noprovider and norprovider now Aug 12 13:49:35 I see Aug 12 13:49:40 * koen tries a fix Aug 12 13:52:59 What can i try doing if bitbake doesn't want to bake me any images? Aug 12 13:53:19 I've met errors while bitbaking gpe, opie images; other like helloworld seems to not have versions for my model. Aug 12 13:53:36 bla: what DISTRO and MACHINE? Aug 12 13:53:39 Get a new oven, or run it in -D mode to work out why it's not running. Have you set the environment variables appropriately? Aug 12 13:53:58 * * OE Bug 2733 has been created by autobuild(AT)openembedded.org Aug 12 13:54:00 * * libpcap-0.9.6-r1-do_configure Aug 12 13:54:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2733 Aug 12 13:54:08 koen, angstrom-2007.1 / hx2000 Aug 12 13:54:14 koen, host machine is amd64 Gentoo Aug 12 13:54:23 re Aug 12 13:54:35 bla: don't build 'gpe-image' or 'opie-image', those are broken Aug 12 13:54:39 It finishes work after 2xxx tasks done. Aug 12 13:54:44 koen, heh, nice to know. Which should I try? Aug 12 13:54:50 bla: try 'angstrom-x11-image' or 'angstrom-console-image' Aug 12 13:54:58 * * OE Bug 2734 has been RESOLVED (INVALID) by Aug 12 13:55:00 * *  libnet-1.0.2a-r1-do_configure Aug 12 13:55:02 koen, Ok, I'll try in a minute. Aug 12 13:55:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2734 Aug 12 13:55:10 * * OE Bug 2712 has been RESOLVED (INVALID) by rpurdie(AT)rpsys.net Aug 12 13:55:12 * *  does fakeroot work? Aug 12 13:55:14 * * http://bugs.openembedded.org/show_bug.cgi?id=2712 Aug 12 13:55:21 bla: angstrom-XYZ-image Aug 12 13:55:22 * * OE Bug 2734 has been created by autobuild(AT)openembedded.org Aug 12 13:55:24 * * libnet-1.0.2a-r1-do_configure Aug 12 13:55:26 * * http://bugs.openembedded.org/show_bug.cgi?id=2734 Aug 12 13:55:31 Just after it finishes caching new update. (There was a syntax error in in some .bb file) Aug 12 13:56:04 It might be fixed by now. Aug 12 13:56:22 03rpurdie 07org.oe.dev * r2e20c0e4... 10/ (1 classes/sanity.bbclass): sanity.bbclass: Remove git and bison, we build them if needed Aug 12 13:56:30 RP: seppuku works now, I'm starting a build for it Aug 12 13:56:32 03koen 07org.oe.dev * r3f2a2726... 10/ (1 classes/seppuku.bbclass): seppuku: quick fix for Noprovider not having isRuntime anymore (use Norprovider for that) Aug 12 13:56:56 Are there other distros for this machine I can try? Aug 12 13:57:14 no Aug 12 14:02:46 http://rafb.net/p/US4ob340.html Aug 12 14:02:56 Error while caching. (Freshly updated mtn repos) Aug 12 14:03:36 try using the bitbake 1.8 branch from svn Aug 12 14:03:58 * * OE Bug 2735 has been created by autobuild(AT)openembedded.org Aug 12 14:04:00 * * noprovider for > Aug 12 14:04:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2735 Aug 12 14:07:35 is anyone going to get upset if I start a linux-omap.inc file to support omap1,2,3 and davinci boards? Aug 12 14:07:49 Crofton|home: we'll cheer you on Aug 12 14:07:49 ass opposed to extending linux.inc? Aug 12 14:07:55 er as :) Aug 12 14:07:56 ok Aug 12 14:08:14 omap uses a different kernel tree instead of a patch, right? Aug 12 14:08:24 both Aug 12 14:08:44 there are git trees, and patches against releases Aug 12 14:09:12 basically, I want to keep the file small and simple enough we can undertand it :) Aug 12 14:09:45 and we can make changes without impacting too many machines :) Aug 12 14:10:02 koen, mkay. Aug 12 14:10:13 However it worked till I synced today. Aug 12 14:12:38 hi Aug 12 14:12:57 is there are a program for wake up on lan in OE? Aug 12 14:15:10 koen, BitBake Build Tool Core version 1.8.7, bitbake version 1.8.7 Aug 12 14:15:15 koen, is it ok version? Aug 12 14:15:51 revision 878? Aug 12 14:16:45 Hm, it must've been older. Aug 12 14:17:14 Trying with 878. Aug 12 14:17:25 Ok, passed. Aug 12 14:17:30 ;) Aug 12 14:18:40 RP, mickeyl: http://dominion.kabel.utwente.nl/koen/cms/automated-build-and-bugreports Aug 12 14:19:59 * * OE Bug 2733 has been RESOLVED (INVALID) by Aug 12 14:20:00 * *  libpcap-0.9.6-r1-do_configure Aug 12 14:20:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2733 Aug 12 14:20:17 why not adapt buildbot? Aug 12 14:20:41 because it is unsuited for the task? Aug 12 14:20:47 ok Aug 12 14:21:51 Crofton|home: I started an osk build with seppuku now Aug 12 14:22:00 If I want to rerun some task should I just remove stamps files? Aug 12 14:22:14 no, you use -c rebuild Aug 12 14:22:23 I still get error while patching. But I find current bitbake can handle alread done tasks better... Aug 12 14:22:27 Ok. Aug 12 14:26:01 Like this: http://rafb.net/p/q5MEif27.html Aug 12 14:27:10 koen: looks good :) Aug 12 14:28:12 what's the minimum version of monotone to use at the moment? Aug 12 14:30:12 hmm, I have been slacking, there are omap patches for 21 and 22 Aug 12 14:33:58 * * OE Bug 2735 has been RESOLVED (FIXED) by Aug 12 14:34:00 * *  noprovider for > Aug 12 14:34:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2735 Aug 12 14:35:15 morning Aug 12 14:42:42 1,1 M | 940,3 k | 28/2.904 | 6/718 looks like I didn't update in a long time Aug 12 14:43:39 Crofton|home: pong Aug 12 14:43:58 you where having a problem with a missing macro? Aug 12 14:44:14 I ahd that a while ago, but have not stumbled across it again Aug 12 14:44:29 hello, I know I'm lazy and not using bugzilla, but please someone fix gpe-fsi, so it would depend on fsi-client:) Aug 12 14:44:37 I have a sticky note with "xorg-macros" on it Aug 12 14:45:01 I think the package that is not building for you needs to depend on this package ... Aug 12 14:45:53 I've checked bug 2723 per Laibsch suggestion Aug 12 14:46:08 ~oebug 2723 Aug 12 14:46:33 it seems to need a dependency to util-macros-native - I'm trying that now, building that one before makedepend-native Aug 12 14:46:42 ok Aug 12 14:47:02 I have had other issues I have been working on, so it did not have the problem again Aug 12 14:48:11 if I get the problem again, I will check this bug ans try the fix Aug 12 14:48:56 is there a way to mark bugs so that you can build a list of bugs you care about? Aug 12 14:50:59 Crofton|home : can you pull the from the efika git/kernel rep ? Aug 12 14:53:28 I haben't tried, is there any reason why I couldn't? Aug 12 14:53:49 i am getting errors while trying to pull Aug 12 14:54:08 ah Aug 12 14:54:11 fatal: unexpected EOF Aug 12 14:54:14 what is the command? Aug 12 14:54:40 git fetch git://git.secretlab.ca/git/linux-2.6.git lite5200:sl-lite5200 Aug 12 14:54:43 I usually add myself to the cc list Aug 12 14:54:57 I'll need to clone first right? Aug 12 14:55:04 but there might be a better/easier way to track bugs you're interested in Aug 12 14:55:14 git clone git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git Aug 12 14:55:18 Yeah, I want to avoid teh emaisl :) Aug 12 14:55:32 ~logs Aug 12 14:55:35 hmm... logs is apt/ibot/infobot/jbot/purl all log daily to http://ibot.rikers.org// where channelname is html encoded ie: %23debian | lines that start with a space are not shown | some channels have stats at http://ibot.rikers.org/stats/.html.gz Aug 12 14:55:43 bbiab Aug 12 14:55:50 Crofton|home : clone the main 2.6 Aug 12 14:56:04 ok Aug 12 14:56:08 then what? Aug 12 14:56:34 the other command will get a branch? Aug 12 14:56:48 * Crofton|home has a 2.6.22 kernel for the OSK now Aug 12 14:57:54 Can I omit package during bitbaking somehow? Aug 12 14:58:02 I can't get pass this patch error. Aug 12 15:06:33 "Bill Nye Booed in Waco for pointing out Moon reflects the Sun" Aug 12 15:11:19 * Crofton|home is very happy the bible does not discuss electricity or computers Aug 12 15:14:06 * Bernardo is away: Ausente por agora. Aug 12 15:32:47 Crofton|home: hmmm? Aug 12 15:33:14 koen mentioend Bill Nye getting booed Aug 12 15:34:04 Crofton|home: ouch Aug 12 15:37:46 steliosk, cd into linux-2.6.git and run the fetch leads to failure Aug 12 15:43:51 03crofton 07org.oe.dev * re6c2f3b6... 10/ (8 files in 2 dirs): Aug 12 15:43:51 linux-omap : Move linux-omap1 to linux-omap. linux-omap should work for Aug 12 15:43:51 OMAP{1,2,3} and davinci processors. Aug 12 15:43:57 03crofton 07org.oe.dev * r31df792e... 10/ (1 packages/linux/linux-omap1_2.6.x+git.bb): linux-omap : Convert bb file to use linux-omap from linux-omap1. Aug 12 15:44:05 03crofton 07org.oe.dev * r6b3dffb4... 10/ (5 files in 3 dirs): linux-omap1_2.6.22 : A support for linux-omap1 version 2.6.22 Aug 12 15:48:23 Bill Nye, "The Science Guy" got boo'd ? Aug 12 15:48:33 google Aug 12 15:49:41 Sigh. Never attempt to discuss *anything* with a religous zealot. Aug 12 15:49:46 heh Aug 12 15:49:47 Crofton|home : it seems there is a problem with the server :( Aug 12 15:50:23 steliosk, so the fetch would integrate a new branch? Aug 12 15:50:52 * RP just managed to overwrite every .bb file in his local repo with binary data :} Aug 12 15:51:12 oops Aug 12 15:51:23 Ouch! Aug 12 15:51:39 stupid python path join... Aug 12 15:51:39 Every one? Now that takes a special effort! :-) Aug 12 15:51:51 or a program run amok Aug 12 15:51:59 os.path.join("something", fn) = fn if fn starts with a / Aug 12 15:52:37 Its a special version of bitbake I made to do some tests, I just forgot the above "feature of join" :/ Aug 12 15:52:58 * RP wonders how much data is lost :/ Aug 12 15:53:03 any sufficently advanced feature is indistinguishable from magic Aug 12 15:53:22 the "black" variety, in this case. Aug 12 15:57:54 aahrg. Its got everything including my local metadata and unchecked in files :-( Aug 12 16:10:26 Crofton|home : yes Aug 12 16:12:16 should I call the davinci evn board machine file "davinci_dvevm.conf" Aug 12 16:35:54 RP: OE has base_path_join iirc Aug 12 16:35:57 u-boot_git is failing in fetch Aug 12 16:36:13 koen: Yes, bitbake doesn't though :/ Aug 12 16:36:29 ah, right Aug 12 16:36:58 * * OE Bug 2554 has been RESOLVED (FIXED) by mwester(AT)dls.net Aug 12 16:37:00 * *  broken cross build environment / fix ncurses build Aug 12 16:37:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2554 Aug 12 16:40:46 is there a change needed by git recipes? Aug 12 16:42:31 Crofton|home: In what way? Aug 12 16:42:46 not sure, I thought I some some discussion about git Aug 12 16:43:43 Crofton|home: Two things. We now support SRCREV with git so the version can autoupdate when the repo changes. Also, git-native is auto added as a dependency for reciepes using git Aug 12 16:43:58 ok Aug 12 16:44:21 do old recipes need any changes? Aug 12 16:44:45 no, nothing changes Aug 12 16:44:48 I trying to build u-boot from git and it hangs at remote: 100% (41565/41565) done Aug 12 16:46:02 I doubt its hanging, its probably creating tarballs Aug 12 16:46:22 no tar runnign Aug 12 16:46:37 :/ Aug 12 16:47:15 I need to start writing soon, I lleave it for a bit Aug 12 16:52:47 hmmm Aug 12 16:52:56 I have a u-boot-davinci_dvevm-git-r1.bin Aug 12 16:53:07 I wonder if it works .... Aug 12 16:53:25 to bad those TI guys haven't sent me an evm .... Aug 12 16:53:31 * Crofton|home leave to work on thesis Aug 12 17:05:56 03rpurdie * r879 10/ (4 files in 4 dirs): ConfHandler.py: Apply OVERRIDES before performing immediate expansions Aug 12 17:07:32 RP: does that fix http://bugs.openembedded.org/show_bug.cgi?id=2601 ? Aug 12 17:22:58 * * OE Bug 2736 has been created by autobuild(AT)openembedded.org Aug 12 17:23:00 * * noprovider for > Aug 12 17:23:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2736 Aug 12 17:27:43 hmm, that bug is not helpful Aug 12 18:00:45 good morning Aug 12 18:06:32 <_law_> hi flo_lap Aug 12 18:14:18 * koen opens a bottle of leffe radieusse Aug 12 18:14:38 koen, I have made a uboot for the dv evm Aug 12 18:14:48 Crofton: cool Aug 12 18:14:54 I am trying to get a kernel done Aug 12 18:14:58 let me drink a veer on that Aug 12 18:15:02 beer* Aug 12 18:15:04 heh Aug 12 18:15:41 I'm hoping to get some form of images together that I can talk someone into flashing Aug 12 18:19:44 the nice thing is incremental work to build stuff for DV is very low, since it is smae ARM as OSK Aug 12 18:30:46 RP: ok, simple. if we make overwriting PACKAGES mandatory, we can as well put PACKAGES="${PN}" back into the bbclass for single-package recipes and remove the ALLOW_EMPTY_... = "0" overrides. Aug 12 18:30:48 http://en.pastebin.ca/654716 Aug 12 18:39:44 But for task-base that creates empty task-base-touchscreen-dbg and task-base-touchscreen-dbg packages, for example. They don't even seem to Recommend task base-touchscreen itself, maybe because task-base-touchscreen Depends on tslib-tests and tslib-calibrate, both of which don't have corresponding -dev/-dbg packages. Aug 12 18:42:55 Crofton: the autobuilder has a few osk bugs for you Aug 12 18:50:13 argh Aug 12 18:50:33 hmm Aug 12 18:51:45 not much info .... Aug 12 18:52:22 the log attacher seems to be a bit buggy Aug 12 18:52:35 :) Aug 12 18:52:54 I only find 2739 Aug 12 18:52:57 are there more? Aug 12 18:53:03 I am searhcing for omap Aug 12 18:54:31 Crofton: http://tinyurl.com/yur8me Aug 12 18:54:31 http://pastebin.ca/654734 Aug 12 18:55:54 those tasks complete now Aug 12 18:57:22 um, I'm going to need a crystal ball to figure out what those mean .... Aug 12 19:02:09 can you add something so searching by machine finds bugs from the autobuilder? Aug 12 19:02:30 ie search for omap5912osk Aug 12 19:06:58 * * OE Bug 2737 has been created by autobuild(AT)openembedded.org Aug 12 19:07:00 * * noprovider for > Aug 12 19:07:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2737 Aug 12 19:07:10 * * OE Bug 2739 has been created by autobuild(AT)openembedded.org Aug 12 19:07:12 * * linux-omap1-2.6.20-omap1-r0-do_deploy Aug 12 19:07:14 * * http://bugs.openembedded.org/show_bug.cgi?id=2739 Aug 12 19:07:21 Crofton: like this: http://rafb.net/p/NifQFg92.html ? Aug 12 19:07:22 * * OE Bug 2740 has been created by autobuild(AT)openembedded.org Aug 12 19:07:24 * * noprovider for > Aug 12 19:07:26 * * http://bugs.openembedded.org/show_bug.cgi?id=2740 Aug 12 19:07:34 * * OE Bug 2738 has been created by autobuild(AT)openembedded.org Aug 12 19:07:36 * * u-boot-1.2.0-r2-do_deploy Aug 12 19:07:38 * * http://bugs.openembedded.org/show_bug.cgi?id=2738 Aug 12 19:08:04 that looks better Aug 12 19:09:11 and we need the logs so we can see what is bad Aug 12 19:09:21 trying to fix that Aug 12 19:09:30 :) Aug 12 19:09:40 right now you are spamming BZ Aug 12 19:17:43 aha Aug 12 19:17:51 bugzilla changed the wording Aug 12 19:18:55 http://rafb.net/p/bNqdLR72.html Aug 12 19:22:58 * * OE Bug 2736 has been marked as DUPLICATE of bug 2737 by Aug 12 19:23:01 * *  noprovider for > Aug 12 19:23:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2736 Aug 12 19:36:58 * * OE Bug 2738 has been RESOLVED (FIXED) by Aug 12 19:37:00 * *  u-boot-1.2.0-r2-do_deploy Aug 12 19:37:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2738 Aug 12 19:37:10 * * OE Bug 2742 has been created by  Aug 12 19:37:12 * * test bug, please ignore Aug 12 19:37:14 * * http://bugs.openembedded.org/show_bug.cgi?id=2742 Aug 12 19:37:22 * * OE Bug 2741 has been created by autobuild(AT)openembedded.org Aug 12 19:37:23 heh Aug 12 19:37:24 * * noprovider for > Aug 12 19:37:26 * * http://bugs.openembedded.org/show_bug.cgi?id=2741 Aug 12 19:37:45 koen resolving non-existent bugs should not increase your karma :) Aug 12 19:38:06 :) Aug 12 19:41:39 trying to fetch the davinci linux kernel is taking forever Aug 12 19:52:58 * * OE Bug 2743 has been created by autobuild(AT)openembedded.org Aug 12 19:53:00 * * noprovider for > Aug 12 19:53:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2743 Aug 12 19:53:10 * * OE Bug 2743 has been RESOLVED (INVALID) by Aug 12 19:53:12 bugmail is lagging Aug 12 19:53:12 * *  noprovider for > Aug 12 19:53:15 * * http://bugs.openembedded.org/show_bug.cgi?id=2743 Aug 12 19:53:22 * * OE Bug 2737 has been RESOLVED (INVALID) by Aug 12 19:53:24 * *  noprovider for > Aug 12 19:53:26 * * http://bugs.openembedded.org/show_bug.cgi?id=2737 Aug 12 19:53:43 03koen 07org.oe.dev * r9a4862c9... 10/ (1 classes/seppuku.bbclass): Aug 12 19:53:43 seppuku: misc fixes Aug 12 19:53:43 * better error messages Aug 12 19:53:43 * fixed bug creation and attachment handling Aug 12 19:53:44 * now reports machine in bug content (till we have a machine section in bugzilla) Aug 12 19:53:58 * * OE Bug 2744 has been RESOLVED (FIXED) by Aug 12 19:54:00 * *  synergy-1.3.1-r0-do_configure Aug 12 19:54:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2744 Aug 12 19:54:10 * * OE Bug 2744 has been created by autobuild(AT)openembedded.org Aug 12 19:54:12 * * synergy-1.3.1-r0-do_configure Aug 12 19:54:14 * * http://bugs.openembedded.org/show_bug.cgi?id=2744 Aug 12 19:54:21 any git fetcher gurus about? Aug 12 19:54:22 * * OE Bug 2740 has been RESOLVED (INVALID) by Aug 12 19:54:24 * *  noprovider for > Aug 12 19:54:26 * * http://bugs.openembedded.org/show_bug.cgi?id=2740 Aug 12 19:55:54 Crofton|work: RP would be your best bet Aug 12 19:56:50 I had protocol=http Aug 12 19:57:01 maybe need protocol=git .... Aug 12 19:57:24 I should be writing anyway Aug 12 19:58:00 http://dealextreme.com/details.dx/sku.4355~r.96035782 Aug 12 19:58:08 I need one in the movie theatre ... Aug 12 19:59:38 those things are neat, till you need to call 911 Aug 12 20:01:12 * Crofton grew up the the pre-911 world Aug 12 20:01:28 and he means that in the telephone way Aug 12 20:03:41 I wonder how unhealthy that thing is, compared to the phone signals :> Aug 12 20:04:26 what I'd like to see is an RFID chip destroyer :P Aug 12 20:07:58 * * OE Bug 2745 has been created by autobuild(AT)openembedded.org Aug 12 20:08:00 * * gtk+-2.10.14-r3-do_install Aug 12 20:08:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2745 Aug 12 20:08:58 * * OE Bug 2741 has been RESOLVED (INVALID) by Aug 12 20:09:00 * *  noprovider for > Aug 12 20:09:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2741 Aug 12 20:09:10 * * OE Bug 2745 has been RESOLVED (INVALID) by Aug 12 20:09:12 * *  gtk+-2.10.14-r3-do_install Aug 12 20:09:14 * * http://bugs.openembedded.org/show_bug.cgi?id=2745 Aug 12 20:11:35 I think I smoothed most of the rough edges of seppuku.bbclass Aug 12 20:14:13 any bugs I need to look at? Aug 12 20:14:39 is 2739 valid? Aug 12 20:15:05 it did fail the first time Aug 12 20:18:37 do you know Dirk Behme? He is active on the omap and davinci lists Aug 12 20:19:10 I only heard his name a few times Aug 12 20:19:22 it is all over mailing lists Aug 12 20:35:58 03koen 07org.oe.dev * r7097bc1a... 10/ (1 packages/gtk+/gtk+_2.10.14.bb packages/gtk+/gtk-2.10.inc): gtk+: work around rrecommends bug Aug 12 20:49:25 looks like I fixed my git fetching problem Aug 12 20:49:53 Hopefully, the git kernel builds for the evm :), that will be later Aug 12 20:56:15 Is today a bug-hunting day, then? Aug 12 21:01:32 bug creation day more like it, I think koen is running some autobuilder/autobugreporter tests Aug 12 21:06:10 * RP returns Aug 12 21:06:23 koen: I'm not sure about 2601, I've not looked at that one Aug 12 21:06:30 just like General McArthur Aug 12 21:06:46 Crofton: What was the problem? :) Aug 12 21:06:53 mine Aug 12 21:07:04 I had proto-http for somethign using git Aug 12 21:07:27 couldn't find HEAD Aug 12 21:08:00 Crofton: ah, that would explain it Aug 12 21:09:44 is there a make config that gets you a curses kernel configering Aug 12 21:10:20 Crofton: No, but we probably should add one... Aug 12 21:11:08 RP, I need to edit a kernel config, usually I do make xconfig Aug 12 21:11:40 Crofton: Ah, make menuconfig Aug 12 21:11:52 thanks Aug 12 21:11:56 Crofton: I thought you meant a task that ran that... Aug 12 21:12:17 heh, no a simple case Aug 12 21:13:01 I am working on gettting davinci support into OE, even though I don't have an evm Aug 12 21:14:10 Use the ARM EABI to compiel the kernel turns on EABI, right? Aug 12 21:14:30 Crofton: No, you turn it on in the defconfig Aug 12 21:15:15 that is in the config Aug 12 21:15:32 the question is bad :) Aug 12 21:15:54 I've seen the BB files that add it, they make me nervous Aug 12 21:16:03 I should try again though Aug 12 21:16:33 I keep meaning to add that to kernel.bbclass itself Aug 12 21:17:57 03rpurdie * r880 10/ (4 files in 4 dirs): cooker.py: Allow the -b -e option combination to take regular expressions Aug 12 21:18:17 * RP finally got irritated with the -b -e option combination to fix it... Aug 12 21:18:53 anyway, lets see if I can build a kernel for a davinci EVM Aug 12 21:20:14 The sad thing is a can't remember what I wanted to test with the -b -e options now :} Aug 12 21:20:31 RP: I worked around the append problem Aug 12 21:22:24 koen: I know, it needs fixing properly though Aug 12 21:23:23 Crofton: something is wonkey with the osk kernel Aug 12 21:23:32 it just tried to build the jlime kernel Aug 12 21:23:35 define "wonkey" Aug 12 21:23:46 ok, that is wonkey Aug 12 21:24:00 bitbake linux-omap1 Aug 12 21:24:15 bitbake angstrom-x11-image failed with a weird error that was caused by the jlime kernel bb Aug 12 21:24:36 * koen sets a compatible machine on that Aug 12 21:24:56 I renamed linux-omap1 to linux-omap is all Aug 12 21:25:29 did you adjust hte machine.conf? Aug 12 21:25:48 I do not thinks so Aug 12 21:25:54 it still uses linux-omap1 Aug 12 21:26:19 I don't normally do X builds Aug 12 21:26:35 although I should as I have an OSK with a screen Aug 12 21:26:58 ok davinvi git kernel building Aug 12 21:32:13 koen: I'm having to work a bit outside the bug bit I think bitbake is working. The thing to understand is that FOO_angstrom overrides FOO_append_linux Aug 12 21:33:02 neither ends up in the package Aug 12 21:33:11 What you'd need if FOO_angstrom_append_linux :/ Aug 12 21:33:16 s/if/is/ Aug 12 21:33:27 not FOO_append_angstrom or FOO_append_linux Aug 12 21:33:36 uImage-2.6.x+git-davinci-davinci_dvevm-20070812212531 Aug 12 21:33:36 koen: Can you show a full example? Aug 12 21:34:18 koen: Florian's bug gives one override but its meanlingless without knowing what else is set Aug 12 21:35:20 RP: http://rafb.net/p/8rEO3v53.html Aug 12 21:35:56 marcin and I used the distro override because someone objected to making it global Aug 12 21:36:15 (I just made it global, bring on the hatemail) Aug 12 21:36:30 ok, angstrom-minimal-image created for the evm Aug 12 21:36:41 now, I need an evm Aug 12 21:36:45 or Aug 12 21:36:47 a sucker Aug 12 21:36:53 no 'or' Aug 12 21:36:57 you need an evm Aug 12 21:37:02 :) Aug 12 21:37:10 well, I think I know a sucker Aug 12 21:37:30 and rishi is not replying to emails Aug 12 21:38:50 koen: It doesn't give nothing, it gives " ttf-dejavu-sans" here Aug 12 21:39:03 koen: If you remove the ${PN} it works more as expected Aug 12 21:39:22 Which hints at an expandKeys problem :/ Aug 12 21:39:30 ~summon zecke Aug 12 21:39:31 apt takes out 20 clean, identical-looking phones, some extra hands, and pretends to be a telemarketer for a large corporation, so he gets delivered a phonelist containing zecke's coordinates Aug 12 21:40:05 bleh Aug 12 21:40:14 html parsing with regexps in python Aug 12 21:40:37 my eyes just imploded Aug 12 21:40:55 time for bed Aug 12 21:40:57 'night all Aug 12 21:41:06 I'm gn Aug 12 21:41:08 'night koen Aug 12 21:41:13 and Crofton Aug 12 21:43:02 03koen 07org.oe.dev * rfc146673... 10/ (1 conf/distro/include/preferred-xorg-versions.inc): preferred xorg versions: fix xserver version Aug 12 21:43:08 03koen 07org.oe.dev * r65fb29ae... 10/ (1 packages/linux/linux-jlime-current.bb): linux-jlime-current: set bogus COMPATIBLE machine so the maintainer adds a proper one Aug 12 23:19:52 03crofton 07org.oe.dev * radcdf1b1... 10/ (9 files in 2 dirs): Aug 12 23:19:52 linux-omap1 : Move compatible machine from generic include file to specific Aug 12 23:19:52 version files. Aug 12 23:19:57 03crofton 07org.oe.dev * r75e743b3... 10/ (5 files in 3 dirs): Aug 12 23:19:57 linux-davinci : Add support for build TI Davinci processor kernels. Currently Aug 12 23:19:57 can build a kernel for the Davinci EVM board. Aug 12 23:20:01 03crofton 07org.oe.dev * ra3c9cfee... 10/ (1 conf/machine/davinci_dvevm.conf): davinci-dvevm.conf : Add machine file for the TI Davinci EVM board. Aug 12 23:27:58 * * OE Bug 2746 has been created by marcel(AT)ziswiler.com Aug 12 23:28:00 * * libgsmd_svn.bb uses SRCREV instead of SRCDATE Aug 12 23:28:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2746 Aug 12 23:54:04 03rpurdie 07org.oe.dev * r5fd8123e... 10/ (1 classes/sanity.bbclass): sanity.bbclass: Sync with Poky for umask check Aug 13 01:43:32 I am getting this error error: pick-rref: HEAD not found Aug 13 01:43:54 when I it is trying to get kernel from git.mvista.com Aug 13 01:44:03 has someone else seen such an issue Aug 13 01:47:35 khem, what bb file? Aug 13 01:48:02 I had an issue where there was proto=http, not proto=git Aug 13 01:48:16 khem: Looks to me like the upstream git (version control) doesn't have a head.. might have an unmerged head.. haven't had that error before Aug 13 01:48:42 Crofton|home: I am adding new one which is based on linux-omap1_2.6.x+git.bb Aug 13 01:48:51 SRC_URI = "git://source.mvista.com/git/linux-davinci-2.6.git;protocol=git \ Aug 13 01:48:56 heh Aug 13 01:49:06 mine is Aug 13 01:49:35 SRC_URI = "git://source.mvista.com/git/linux-omap-2.6.git;protocol=http \ Aug 13 01:49:41 try changing to protocol=git Aug 13 01:49:49 ah there you go Aug 13 01:49:59 I had teh exact issue earlier today with fetching the davinci kernel Aug 13 01:50:32 that fixes it Aug 13 01:50:46 ok, I'll push fix Aug 13 01:50:49 but that also means there is a bug in above said bb file :) Aug 13 01:51:17 cool thanks Aug 13 01:51:41 Crofton|home: I am almost done with omap2430sdp Aug 13 01:51:54 Crofton|home: I will post patch to ml plz Aug 13 01:51:57 have a look Aug 13 01:52:53 -SRC_URI = "git://source.mvista.com/git/linux-omap-2.6.git;protocol=http \ Aug 13 01:52:54 +SRC_URI = "git://source.mvista.com/git/linux-omap-2.6.git;protocol=git \ Aug 13 01:52:54 file://defconfig" Aug 13 01:53:19 I need to sort through the linux-omap1 kernels Aug 13 01:53:39 you saw I renamed linux-omap1.inc to linux-omap.inc? Aug 13 01:54:16 Crofton|home: I feel like the structure could be better .... Aug 13 01:54:45 yes saw that :) Aug 13 01:55:09 Crofton|home: omap1 omap2 omap3 etc. are families Aug 13 01:55:28 so as far as kernel is concerned we can not have a common omap1 kernel e.g. Aug 13 01:55:48 so I think we do need to remove the family like bb files Aug 13 01:56:00 and add board specific bb files Aug 13 01:56:07 I left the root is linux-omap.inc Aug 13 01:56:21 yeah thats true Aug 13 01:56:34 so we can work on it in a small group Aug 13 01:56:44 there is linux.inc Aug 13 01:56:46 however I was talking about e.g. linux-omap1_2.6.20-omap1.bb Aug 13 01:56:54 and that is corect long term Aug 13 01:57:01 this config wont work on all omap1 boards Aug 13 01:57:02 yeah Aug 13 01:57:12 so I would be specific Aug 13 01:57:35 and would rename it to something like linux-omap5912osk_2.6.20-omap1.bb Aug 13 01:57:47 we can have "general" stuff in linux-omap.inc and board spcific in specialiaztion files Aug 13 01:58:02 thats ok Aug 13 01:58:27 well, to some extent we can cover multiple machines in the same bb file Aug 13 01:58:37 look at linux.inc Aug 13 01:58:41 hmmm maybe Aug 13 01:58:54 but I would still like to add one for each board :) Aug 13 01:59:06 but I want to keep it simpler at thr moment Aug 13 01:59:19 thats fine Aug 13 01:59:20 I understand Aug 13 01:59:44 but if we can support several boards from one bb file, that is OK also Aug 13 02:00:16 thats true but it may end up with multiple defconfig files Aug 13 02:00:18 bt the important thing is we start providing excellent support for xisting boards Aug 13 02:00:24 yes Aug 13 02:00:27 so its same thing in some different structure Aug 13 02:00:33 managing defconfig is tricky Aug 13 02:00:43 mine approach I feel is a bit readable Aug 13 02:01:03 this is why I want to work from linux-omap.inc Aug 13 02:01:08 but I think you are right we have to contain the variety Aug 13 02:01:16 so we can figure out the balance we need Aug 13 02:01:29 without trippig up other builds :) Aug 13 02:01:41 I think have omap1 omap2 and omap3 Aug 13 02:01:46 yeah Aug 13 02:01:49 and then have all defconfigs inside Aug 13 02:02:06 omap5912 omap1510 omap1610 omap1710 are all omap1 Aug 13 02:02:24 do you have any need for old ABI builds? Aug 13 02:04:56 the hard thing for us is lack of boards to test with Aug 13 02:05:06 I am shooting blind on the EVM Aug 13 02:05:58 we might need on omap1 Aug 13 02:06:07 if someone is really upto old stuff Aug 13 02:06:12 ok Aug 13 02:06:21 but I would prefer that we drop it from dev Aug 13 02:06:22 we have selectors in Aug 13 02:06:27 as its not future :) Aug 13 02:06:31 :) Aug 13 02:06:32 good Aug 13 02:06:43 I need to go now Aug 13 02:06:48 talk to you tomorrow Aug 13 02:06:53 me too Aug 13 02:06:55 cu Aug 13 02:06:57 l8r Aug 13 02:07:09 I will post my patch if I am ready Aug 13 02:11:17 hello all! Aug 13 02:11:35 im getting trouble with a bluetooth on zire72 Aug 13 02:12:10 im using this files: http://www.palmlinux.cz/modules.php?name=Downloads&d_op=viewdownload&cid=7 Aug 13 02:12:59 ive tried to build from mtn of OE, but i cant understand the process and the image generated by compiling only linux4palm dont book (initrd cannot mount the partition) Aug 13 02:13:03 so, someone can guide me? Aug 13 02:34:59 03crofton 07org.oe.dev * rc4c89e77... 10/ (1 packages/linux/linux-omap1_2.6.x+git.bb): linux-omap1 : Fix git source line. **** ENDING LOGGING AT Mon Aug 13 02:59:56 2007