**** BEGIN LOGGING AT Sun Jan 07 02:59:58 2007 Jan 07 04:29:25 03justinp 07org.oe.oz354x * r83e1ef9b... 10/ (10 files in 8 dirs): Jan 07 04:29:25 revert previous accidental commit Jan 07 04:29:25 applied changes from 87c2d837ea5586793e98a4604630ceb3fa70dc7c Jan 07 04:29:25 through d51c990067f86309a3aab1e6b76eaaa5c65f550a Jan 07 04:29:28 03justinp 07org.oe.oz354x * re2968ef1... 10/ (1 packages/glibc/glibc-package.bbclass): glibc-package.bbclass: fix qemu -L parameter (bug #1756) Jan 07 06:20:33 03rwhitby 07org.oe.dev * ra903c08f... 10/ (1 packages/meta/slugos-packages.bb): slugos-packages: changed bluez-utils-nodbus to just bluez-utils Jan 07 06:27:30 03pfalcon 07org.oe.dev * r1f5c493b... 10/ (1 packages/fbreader/fbreader_0.7.4q.bb): fbreader 0.7.4q: Latest version of FBReader. Jan 07 09:25:57 03koen 07org.oe.dev * rd919d63d... 10/ (13 files in 2 dirs): lots of machines: remove bogus architecture Jan 07 09:56:21 koen: I'm putting the nslu2-linux autobuilder machine back into auto-merge mode (instead of just choosing first head). I'm looking for reasons why I shouldn't make it push those auto-merges upstream as well. Note that if it's not an automatic merge, it doesn't touch it. Jan 07 09:58:05 rwhitby: why are you making it merge again? Jan 07 09:58:40 cause choosing the first head is not auto-building stuff that people are checking in. Jan 07 09:58:57 so the auto-builder has to merge to build the latest stuff (which is what people expect of the autobuilder) Jan 07 09:58:59 so make those people merge Jan 07 09:59:22 it's not them who are causing the multiple heads - it's the nature of distributed SCM Jan 07 09:59:44 right Jan 07 09:59:59 but people who can complain about the autobuilder can also merge themselves Jan 07 10:00:01 they merge, and commit, and then there are still multiple heads cause someone else committed on a different server at the same time. they expect the autobuilder to build the commit they just made, and it's not. Jan 07 10:00:25 so tell those people to commit to the same server Jan 07 10:00:42 ok, I'm telling all OE people to commit to monotone.nslu2-linux.org then :-) Jan 07 10:01:10 (just kidding) Jan 07 10:01:32 seriously, there is no way to stop multiple heads, and the people who are committing changes and expecting them to be built are not the problem. Jan 07 10:01:39 I still don't see the problem Jan 07 10:01:57 ok. (1) I commit a change to monotone.nslu2-linux.org Jan 07 10:02:01 if stuff doesn't get built, people should check for multiple heads instead of complaining Jan 07 10:02:11 it's not rocket science Jan 07 10:02:16 sorry, that's not an appropriate response. Jan 07 10:02:28 people should not need to work around the SCM Jan 07 10:02:54 if I commit something, and there is one single head after my commit, then I should not have to check again in 15 minutes when the autobuilder starts building. Jan 07 10:02:58 nslu2 should not push 170 bogus merges because people 'should not need to work around the SCM' Jan 07 10:03:11 koen: stick to the problem at hand please. Jan 07 10:03:15 nslu2 people should not push broken stuff Jan 07 10:03:25 nslu2 people should fix the bugs they introduce Jan 07 10:03:25 koen: stick to the problem at hand please. Jan 07 10:03:32 koen: stick to the problem at hand please. Jan 07 10:03:41 !oebug 1762 Jan 07 10:03:43 * * Bug 1762, Status: NEW, Created: 2007-01-06 19:35 Jan 07 10:03:43 * * papercrane(AT)reversefold.com: slugos-packages unparsed line Jan 07 10:03:45 * * http://bugs.openembedded.org/show_bug.cgi?id=1762 Jan 07 10:04:31 morning Jan 07 10:04:44 I think that automerging is correct way Jan 07 10:04:50 anyway, getting back to the discussion at hand. my contention stands that autobuilders should auto-merge. Jan 07 10:04:54 if it is auto merge+push Jan 07 10:05:12 thats how dscm works Jan 07 10:05:12 hrw: and what happens if it requires a manual merge? Jan 07 10:05:28 hrw: right - that's the discussion I'm having at the moment, and asking for feedback on that (not other unrelated stuff) Jan 07 10:05:29 koen: Note that if it's not an automatic merge, it doesn't touch it. Jan 07 10:05:43 its quote of rwhitby from some lines above Jan 07 10:05:57 hrw: so it doesn't merge, which is the same as picking the first head Jan 07 10:06:19 correct - and a non-automatic merge hasn't affect the autobuilder in at least 6 months. Jan 07 10:06:28 hrw: I've had my commit reverted by a nslu2 merge more that once, so sorry if I don't trust their automerger Jan 07 10:06:36 whereas it successfully did 170 automatic merges it seems Jan 07 10:07:02 koen: ops Jan 07 10:07:16 koen: characterising "nslu2 merge" vs "OE merge" is not a helpful way to discuss problems. Jan 07 10:08:35 Joe Dev can also do wrong merge Jan 07 10:08:35 sure Jan 07 10:08:41 the Master Makefile merge logic is open for your review. Please point out it's flaws. I believe you're remembering mistakes made by people trying to do manual merges back when monotone presented manual merges every day Jan 07 10:08:50 Processor Some Random V6 Processor rev 2 (v6l) Jan 07 10:08:59 CPU architecture 6TEJ - omap? Jan 07 10:09:19 univac: or i.mx, or monahan Jan 07 10:09:28 it's from new nokia n800 Jan 07 10:09:45 univac: omap 2420 afaik Jan 07 10:10:03 anyway, I'm still asking for feedback on automerge+push as the preferred mechanism. I haven't seen an on-topic argument against it yet, but I won't turn it on without agreement from the other server maintainers. Jan 07 10:10:11 ok, thx Jan 07 10:10:51 rwhitby: don't push automerges Jan 07 10:11:00 rwhitby: and since you don't push, don't merge Jan 07 10:11:06 koen: please explain why Jan 07 10:11:26 (without referring to unrelated topics in the past) Jan 07 10:12:02 so I can't refer to the 170 bogus merges nslu2 pushed? Jan 07 10:12:49 actually, that's a good argument for it. it would have been 170 helpful automerges (removing 170 instances of multiple heads for the users who inevitably ask why there are multiple heads) instead of 170 all at once. Jan 07 10:13:37 (i.e. 170 isolated merges spread over six months, instead of 170 all at once) Jan 07 10:13:44 since you can't distinguish an automerge from a manual merge, all merges should be done by developers Jan 07 10:13:53 just like the merge that hrw did earlier today. Jan 07 10:14:26 rwhitby: which one? I did lot of merges in last 36h Jan 07 10:14:47 hrw: any one of your helpful auto-merges of multiple heads Jan 07 10:15:24 ok, thanks for your input - I need to drive my mother in law home now, but will be back in about an hour. I'll check the logs to see if there is any further discussion on the topic. Jan 07 10:16:12 hrw: I was referring to your merge before koen's statement - his statement came in just as I sent my statment, and it made my statement look like it was calling your merge bad - I was actually calling your merge good. Jan 07 10:17:18 I'll be interested in seeing the argument against a machine running "monotone merge", accepting it only if there is no error (i.e. it was an automatic merge) and then pushing. Jan 07 10:18:04 I know that I don't each case of auto-merging that I do, and I doubt that anyone else does either. Jan 07 10:18:14 s/don't each/don't check each/ Jan 07 10:18:26 my version is: DSCM works in a way which require merging. monotone automerging is something which I trust (I never look how it merge if it does not call merge helper util). so if autobuilder need to merge and will NOT merge when it need human decision 'how to merge' then I'm ok with it but let it push merge after merging to not get lot of merges one day Jan 07 10:19:10 hrw: thank you for your concise description of my proposition. Jan 07 10:19:44 we cant get dscm without lot of merging Jan 07 10:19:53 * rwhitby will be back in 1 hour to see whether consensus is reached. Jan 07 10:20:09 hrw: right, and I say that looking at the track record of nslu2, I don't trust them with doing merge+push Jan 07 10:21:15 thats remind me thread about 'oe workflow' when people suggest 'create branch' when bbcollections or lot of merging can be used instead Jan 07 10:21:34 koen: then check their automerge scripts and help to make them better Jan 07 10:24:27 hrw: last time they asked about autobulder merges (months ago) I said merge+push Jan 07 10:25:39 ~change 400 usd to eur Jan 07 10:25:45 400.00 United States Dollar (USD) makes 307.338 Euro (EUR) (from http://www.xe.com/) Jan 07 10:26:09 hrw: anyway, monotone.openembedded.org now has an automerger Jan 07 10:26:30 and you trust it more then nslu2 one? Jan 07 10:26:37 yes Jan 07 10:26:43 why? Jan 07 10:27:09 because it's the same automerger I used last year on ewi Jan 07 10:27:30 then give it to nslu2 people to compare with their Jan 07 10:27:49 why? Jan 07 10:28:04 more than one automerger just begs for merge wars Jan 07 10:28:27 ok Jan 07 10:28:30 I see a point Jan 07 10:29:32 it still won't help against whiny people not being about to want >15m for the autobuilder to pick up that merge, but I can live with that Jan 07 10:29:34 so on hydras their autobuilder should do another pull to check does it got fixed Jan 07 10:30:01 come to think of it Jan 07 10:30:12 the solution to rwhitby problem is even easier Jan 07 10:30:39 their autobuilder just needs to pick the head created by @nslu2-linux Jan 07 10:31:35 gumstix config is broken Jan 07 10:31:51 PACKAGE_ARCH = "arm" Jan 07 10:32:22 a1200 lack kernel config Jan 07 10:33:24 right, that should come with the ezx8 patchset Jan 07 10:33:37 anyway I report bug Jan 07 10:34:00 the best way to make sure I don't forget :) Jan 07 10:34:11 !oebug 1766 Jan 07 10:34:12 * * Bug 1766, Status: NEW, Created: 2007-01-07 03:05 Jan 07 10:34:13 * * openembedded(AT)hrw.one.pl: a1200 lack kernel defconfig Jan 07 10:34:14 * * http://bugs.openembedded.org/show_bug.cgi?id=1766 Jan 07 10:36:09 | arm-angstrom-linux-gnueabi-ld: cannot represent machine `arm:mx31ads' Jan 07 10:36:14 have ever got it? Jan 07 10:36:22 its for amsdelta/angstrom Jan 07 10:36:30 h1910/angstrom Jan 07 10:36:48 hrw: yes you do MACHINE=foo bitbake bar Jan 07 10:36:59 we are missing some csets from poky to fix that Jan 07 10:37:07 in your case binutils messes up Jan 07 10:37:22 ok. so I will revert to conf/auto.conf way Jan 07 10:39:08 a bit longer due to parsing but working Jan 07 10:56:20 cu Jan 07 11:08:04 03koen 07org.oe.dev * r24c0c5c2... 10/ (3 files in 2 dirs): gcc 4.1.1, gcc3-build-cross.inc: fix libsupc++ and libstdc++ la files with some XorA sed magic Jan 07 11:10:17 ~hail XorA|gon1 Jan 07 11:10:21 * ibot bows down to XorA|gon1 and chants, "I'M NOT WORTHY!!" Jan 07 11:12:31 Hi Jan 07 11:12:51 hey noha Jan 07 11:15:08 Is there a way bitbake can show packages which depend on a package I give a a command line argument? Jan 07 11:15:29 I got the -g yesterday but I mean the other way round Jan 07 11:16:04 hot that I know off Jan 07 11:16:45 grep -rn . | grep DEPENDS Jan 07 11:16:46 How do you check dependencies you suspect to be mistaken? Jan 07 11:17:10 . that is Jan 07 11:23:24 In bitbake.conf there is MACHINE_TASK_PROVIDER ?= "task-bootstrap" but task-bootstrap is not existing anymore. What would be a good exchange value? Jan 07 11:29:54 norbert@neoflux:/opt/oe/build$ grep -rn task-bootstrap ../openembedded/* | grep RDEPENDS Jan 07 11:30:02 noha_: task-base Jan 07 11:30:03 norbert@neoflux:/opt/oe/build$ bitbake gpe-image Jan 07 11:30:03 N Jan 07 11:30:12 ERROR: Nothing provides runtime dependency task-bootstrap Jan 07 11:30:12 NOTE: no buildable providers for gpe-image Jan 07 11:30:12 n Jan 07 11:30:23 I don't understand this Jan 07 11:31:06 noha_: task-bootstrap was removed. You probably want task-base but the machine you're using might not have been setup for task-base. Which machine is this? Jan 07 11:31:16 simpad Jan 07 11:31:50 03rpurdie 07org.oe.dev * rba407b0d... 10/ (1 classes/kernel.bbclass): kernel.bbclass: Remove the KERNEL_RELEASE variable and set it to KERNEL_VERSION. This should removes a large source of kernel module problems as the variables then 'just work' Jan 07 11:31:55 03rpurdie 07org.oe.dev * rd9e32177... 10/ (23 files in 2 dirs): packages/linux: Remove uneeded KERNEL_RELEASE assignments and replace references with KERNEL_VERSION Jan 07 11:31:59 * RP frowns at koen Jan 07 11:32:10 * rwhitby returns, reads the backlog, and chuckles at how koen argues against automerge+push by bringing up issues unrelated to the technical item at hand, and then implements the proposal on his server. Jan 07 11:32:52 rwhitby: As I see it, the main potential problem is an automeger war. Why do we need two automergers? Jan 07 11:33:24 RP: I agree about merge wars (I brought up that issue when this was discussed last week) Jan 07 11:33:37 noha_: simpad was never converted to use task-base. Someone needs to convert it. Jan 07 11:33:42 so now that there *is* an automerger in place, there is no reason to have two. Jan 07 11:34:27 and koen's idea about choosing the "local" head in a hydra situation is a good one. Jan 07 11:34:52 RP: Ok, thanks. That means there it doesn't make much sense for me to move on Jan 07 11:36:05 koen: I'm sensing a lot of angst from you directed at nslu2-linux. is it simply the 170 changesets and the syntax error that I introduced yesterday, or is it something deeper than that which we need to discuss in depth ? Jan 07 11:36:41 (or is it just a bad morning in your timezone?) Jan 07 11:37:36 noha_: Have a look at the other machine.conf files for things like MACHINE_FEATURES. simpad needs converting and the conversion means adding entries like these Jan 07 11:37:44 What is a good machine config where I can have a look for changes to apply? Jan 07 11:38:29 noha_: The Zaurus machines are probably similar (say c7x0). In that case the changes are in a common include file but its the same idea Jan 07 11:38:47 Ok, I'll give it a try Jan 07 11:40:23 noha_: there is no reference to task-booststrap in bitbake.conf Jan 07 11:40:51 ok, sorry, I did a pull just a few minutes ago. My version was little bit older Jan 07 11:42:58 at least two weeks older :) Jan 07 11:47:37 oh wow, ich collect _ Jan 07 11:57:52 <[cc]smart> qemu doesn't fetch from http://sources.nslu2-linux.org/sources/qemu_cvs.savannah.nongnu.org__20070107.tar.gz , after redirect it 404s Jan 07 12:03:40 [cc]smart: that would be because cvs tarballs only appear on the source mirror after the autobuilder has built that package on that date and pushed the source. Jan 07 12:04:14 the problem really is why didn't your local bitbake session check out the CVS correctly after that. Jan 07 12:11:13 After I did an update from monotone bitbake is doing real slow again. Is the cache not being updated or not being used? Any things to do after doing an update? Jan 07 12:14:39 koen: cool, we shall all have buildable c++ now Jan 07 12:16:44 koen: anyway, if there is a deeper issue with nslu2-linux collaboration with OE, I'm happy to discuss it at any time to make sure there is no bad will. Jan 07 12:16:46 night all Jan 07 12:17:06 rwhitby: no bad will Jan 07 12:17:26 koen: that's good to hear. thanks. Jan 07 12:21:28 | /a/home/hrw/devel/build/angstrom/tmp/cross/bin/arm-angstrom-linux-gnueabi-ld: cannot represent machine `arm:mx31ads' Jan 07 12:21:44 it strikes me even when MACHINE is in auto.conf ;( Jan 07 12:22:09 hrw: unset MACHINE, rebuild binutils Jan 07 12:22:38 koen: Did you find out if your scripts were unsetting MACHINE? Jan 07 12:22:44 koen: ok Jan 07 12:23:04 RP: it seems the tweaks to kernel.bbclass and base.bbclass are in OE Jan 07 12:23:22 koen: The one in kernel.bbclass shouldn't even be needed anymore Jan 07 12:23:36 (but is harmless to leave in) Jan 07 12:24:06 RP: you want http://monotone.openembedded.org/revision.psp?id=24c0c5c26357241fc537a9d6e7deca64d83f1f50 in poky Jan 07 12:24:20 RP: tf will have fewer problems with abiword :) Jan 07 12:25:37 * RP is surprised tf hasn't requested abiword in poky yet :) Jan 07 12:27:11 RP: KERNEL_RELEASE is gone now? Jan 07 12:27:42 hrw: yes Jan 07 12:27:53 ok Jan 07 12:28:07 I just had to merge packages/linux_2.6.19.bb Jan 07 12:28:22 hrw: It caused too many build failures for no good reason Jan 07 12:30:10 Where can I get a list of MACHINE_FEATURES? Jan 07 12:30:34 noha_: Read task-base.bb? Jan 07 12:31:24 thanks Jan 07 12:34:12 03koen 07org.oe.dev * r6b3a4b78... 10/ (4 files in 2 dirs): h3600, h5000, hx4700: fix quoting of EXTRA_IMAGECMD Jan 07 12:38:09 03hrw 07org.oe.dev * rf66c5a13... 10/ (4 files in 3 dirs): linux: updated to 2.6.19, progear config updated, added progear backlight support Jan 07 12:41:05 RP: does collie use the upstream ucb1400 in 2.6.19? Jan 07 13:11:53 * koen looks at Angstrom-bootstrap-image-test-20070107-collie.rootfs.jffs2 Jan 07 13:13:09 <[cc]smart> rwhitby: it says cvs [checkout aborted]: cannot get working directory: No such file or directory and there's no qemu directory in tmp/work Jan 07 13:15:52 koen: does it really use 2.6? Jan 07 13:16:14 zImage-2.6.18-collie-20070107130906.bin Jan 07 13:16:30 ok Jan 07 13:17:02 I need new hdd.. Jan 07 13:18:51 koen: probably not yet Jan 07 13:36:26 Hi! Jan 07 13:36:42 hi posokolvsky Jan 07 13:37:43 koen: for what machine matchbox-panel bekage happens? I tested on both screen variants, wonder if that machine doesn't have it GUI_MACHINE_CLASS defined Jan 07 13:38:48 it's all in the snippet pasted Jan 07 13:39:03 /opt/oe/build/tmp/work/simpad-linux/bootstrap-image-1.0-r0/temp/run.do_rootfs.18341: 1: rootfs_tar_do_rootfs: not found Jan 07 13:39:22 Does anyone has a glue why this function is missing? Jan 07 13:40:03 noha_: We don't support making root filesystems from tarballs? Jan 07 13:40:25 noha_: Which package classes did you inherit in local.conf? Jan 07 13:40:31 Ah, ok, the function is just a construct. And the real error is a misconfiguration? Jan 07 13:41:11 noha_: correct Jan 07 13:41:35 So, maybe it is the IMAGE_FSTYPES = "jffs2 tar" ?? Jan 07 13:42:01 INHERIT is package_ipk and package_tar Jan 07 13:42:34 noha_: In which order. Please paste the INHERIT line exactly... Jan 07 13:43:08 INHERIT = "package_ipk" Jan 07 13:43:08 INHERIT = "package_tar" Jan 07 13:43:16 Ok, it should be only one Jan 07 13:43:37 noha_: INHERIT = "package_ipk package_tar" should work Jan 07 13:44:21 Ok, it's runnning :) Jan 07 13:44:31 IMAGE_FSTYPES = "jffs2 tar.gz" is ok? Jan 07 13:44:43 posokolvsky: http://ewi546.ewi.utwente.nl/tinderbox/showlog.pl?machine_id=177&logfile=20070107150654.log Jan 07 13:44:51 posokolvsky: h1910 fails to build Jan 07 13:45:08 | arch/arm/mach-pxa/built-in.o: In function `h4000_set_led': Jan 07 13:45:08 | dma_needs_bounce.c:(.text+0x1ca0): undefined reference to `ipaq_asic3_set_led' Jan 07 13:48:12 sweet! That does it! Jan 07 13:48:23 And bootstrap-image is supposed to be bootable? Jan 07 13:48:49 noha_: yes Jan 07 13:49:20 wow, that is cute. Jan 07 13:50:46 Oh, is it true, that bitbake reuses the cache only if one run succeeds? Jan 07 13:51:20 that's not too clever ;) Jan 07 13:51:36 .oO(I should tell zecke about this!) Jan 07 13:52:54 The xterminal-image is an orphan target? Jan 07 13:53:40 hi all Jan 07 13:54:05 hi mice Jan 07 13:55:20 niv one 1hree: hi Jan 07 13:57:24 Hmmm, why does xterminal-image depend on qte? and libsdl-qpe? Jan 07 14:01:21 noha_: The cache is always used Jan 07 14:02:06 What slows down the bitbake run? reparsing? Jan 07 14:04:24 noha_: what DISTRO are you using? Jan 07 14:04:51 openzaurus-unstable Jan 07 14:04:56 right Jan 07 14:05:02 that would mess up sdl Jan 07 14:05:38 ok, does this mean, it isn't a good idea to openzaurus-unstable? Jan 07 14:05:54 add PREFERRED_PROVIDER_virtual/libsdl = "libsdl-x11" to your local.conf Jan 07 14:11:13 koen: bluez-utils/libs 3.8 released.. Jan 07 14:16:13 "Release of bluez-libs-3.8 and bluez-utils-3.8 Jan 07 14:16:14 These releases add some minor enhancements to the stable API and fixes some bugs." Jan 07 14:18:10 bluez ? how said bluez? Jan 07 14:18:49 :-) Jan 07 14:35:41 03pfalcon 07org.oe.dev * r18d4581d... 10/ (3 files in 2 dirs): Jan 07 14:35:41 matchbox-panel: Test GUI_MACHINE_CLASS using base_conditional(), instead of Jan 07 14:35:41 double indirection trick, should be more robust and fix issues seen. Jan 07 14:39:24 03pfalcon 07org.oe.dev * re29766ec... 10/ (1 conf/bitbake.conf): bitbake.conf: Provide default value for GUI_MACHINE_CLASS. Jan 07 14:41:30 koen: from hh.org powered devices only h2200 have bootloader built during 'bitbake task-base'? Jan 07 14:41:52 hrw: and h4000 iirc Jan 07 14:42:06 those are the only devices using LAB I know of Jan 07 14:42:52 I'd like to check in the change to make do_deploy tasks run when making images. It will break users of bitbake 1.7 that isn't bitbake head. Does anyone object to that? Jan 07 14:43:15 RP: what about bitbake 1.6 users? Jan 07 14:43:26 hrw: unaffected Jan 07 14:43:44 RP: then ok from me - note 'trunk update needed' in commit anyway Jan 07 14:44:12 hrw: yes. 1.7.x users know its marked as unstable afterall :) Jan 07 14:44:24 yep Jan 07 14:45:24 btw - how big ram/storage space you guys have in build/desktop machines? I'm thinking of small upgrade in ~March/April here Jan 07 14:45:30 hrw, thanks for note, will check. Jan 07 14:46:00 koen, nope, h4000 doesn't have any bootloader support for now, just starting to ponder about what it would be ;-) Jan 07 14:46:10 posokolvsky: http://ewi546.ewi.utwente.nl/tinderbox/showbuilds.pl?tree=Angstrom - look at 'hrw-test-build-arm-sh3 Linux 2.6.9-34.0.1.EL' - it is 'all ARM/angstrom' build Jan 07 14:46:48 hrw My desktop has 0.75G memory and ~40GB of disk. I do most OE work over nfs onto a fileserver although I had a new harddisk on order to rectify this ;-) Jan 07 14:47:24 hrw-test-build-arm-sh3 :) Jan 07 14:47:51 koen: named after binconfig change test which was arm->sh3->arm build Jan 07 14:47:58 RP: Just FYI, I killed only staging/ after your refactor, and re-run do_stage tasks. it mostley works, but have lotsa: Jan 07 14:48:00 NOTE: Couldn't find shared library provider for libc.so.6 Jan 07 14:48:00 NOTE: Couldn't find shared library provider for libmb.so.1 Jan 07 14:48:00 NOTE: Couldn't find shared library provider for libjpeg.so.62 Jan 07 14:48:15 posokolvsky: objdump bug Jan 07 14:48:15 rebuild from scratch Jan 07 14:48:22 RP: so, unless, you have hint for me, whole tmp/ kill is required, after all Jan 07 14:48:30 ok Jan 07 14:48:44 posokolvsky: easier to rebuild from scratch then rebuilding all needed libs Jan 07 14:48:55 koen: any plans to improve p-s? Jan 07 14:49:11 posokolvsky: That suggests you wiped the package data structures which means you need need to repackage everything at the very least Jan 07 14:49:12 hrw: plans, yes, time, no Jan 07 14:49:25 hrw: see my mail about code sprints to the ml Jan 07 14:49:38 I saw that Jan 07 14:49:50 RP: sure, I did all that. so, just FYI, that you should tell people to rebuild from scratch, not hack ;-) Jan 07 14:50:34 posokolvsky: ok, thanks. I did say a partial wipe might not work. I didn't know for sure, I do now :) Jan 07 14:53:19 03rpurdie 07org.oe.dev * r47274cc0... 10/ (1 classes/image.bbclass): Jan 07 14:53:19 image.bbclass: Make images trigger all do_deploy tasks to run so kernels and Jan 07 14:53:19 things like updater.sh for the zaurus get deployed with the image. This change Jan 07 14:53:19 on affects bitbake 1.7.x users who *will* need to use the latest 1.7.x version Jan 07 14:53:19 from svn. Jan 07 14:54:41 mornnig Jan 07 14:57:22 posokolvsky: I will build matchbox-panel for my progear when it will finish current build Jan 07 14:57:38 meh Jan 07 14:57:39 17M Angstrom-gpe-image-test-20070107-collie.rootfs.jffs2 Jan 07 14:57:49 so much for gdb, strace and friends in the rootfs Jan 07 14:57:52 koen: ONLY 17m Jan 07 14:58:28 hrw, thanks. I guess issue was due to GUI_MACHINE_CLASS not being defined for collie... Jan 07 14:58:45 posokolvsky: my progear does not have it defined at all Jan 07 14:59:12 hrw, I made more robust check as well as set default. but you should define it ;-) Jan 07 14:59:33 we need someone who will move all those vars to MACHINE_GUI_* namespace with screensize, dpi etc Jan 07 14:59:58 or is that your router? ;-) Jan 07 15:00:33 I do not define progear as bigscreen as in OE bigscreen is vga/svga... progear has xga so it would be big or bigger? Jan 07 15:00:37 hrw, that may be too much for near time. so far, most apps offer just small and big icon sets, and we're covered ;-) Jan 07 15:00:53 hrw, bigscreen, that's all ;-) Jan 07 15:00:57 hi people Jan 07 15:01:00 posokolvsky: progear has xga (1024x768) 10.4" touchscreen Jan 07 15:01:02 hi philippe Jan 07 15:01:10 does OE have some tool to track files in staging? Jan 07 15:01:14 h1940, h2200 built Jan 07 15:01:25 I have Xlib.h disappearing at a given moment. Jan 07 15:02:15 and have no idea how to find out when this happens Jan 07 15:03:12 philippe: in theory packaged-staging had to help with it. In practice it is not finished Jan 07 15:03:36 hrw: ok.. so no real stuff to debug it then... Jan 07 15:03:54 philippe: chattr +i Xlib.h and rebui;d Jan 07 15:03:57 Oh well I'll keep using the libx11 rebuild trick Jan 07 15:04:24 NOTE: Removing failed build target virtual/arm-angstrom-linux-gnueabi-gcc-2.95 Jan 07 15:04:27 ;D Jan 07 15:05:57 koen: did not think about that. Let's hope OE barfs on that so I can identify the culprit Jan 07 15:07:03 koen: btw have you tried the gtk-webcore svn builds? Jan 07 15:07:05 posokolvsky: h3900 FAILS - lack of defconfig Jan 07 15:07:51 koen: I build an image but did not get it to boot decently on my Ipaq. The h3800 2.6 kernel is pretty broken. Jan 07 15:07:56 NOTE: Unpacking /a/home/hrw/devel/oe/org.openembedded.dev/packages/linux/linux-handhelds-2.6/defconfig to /a/home/hrw/devel/build/angstrom/tmp/work/h3900-angstrom-linux-gnueabi/linux-handhelds-2.6-2.6.19-hh7-r0/ Jan 07 15:08:03 cp: cannot stat `/a/home/hrw/devel/oe/org.openembedded.dev/packages/linux/linux-handhelds-2.6/defconfig': No such file or directory Jan 07 15:08:28 philippe: nope, too busy polishing angstrom-oabu Jan 07 15:08:36 oabi Jan 07 15:09:07 damn, and I do not have any other "supported" device... Jan 07 15:09:49 netbook does not really work because of missing kernel tools (and I do not have source for those, only florian has AFAIK), yopy has no decent kernel yet... Jan 07 15:11:03 philippe: have florian send you the tools Jan 07 15:11:34 I'm waiting for open tools to send off the netbook to a kernel hacker to get it ported to 2.6.recent Jan 07 15:12:40 * philippe knows a hacker with a netbook who might be willing to help with that too. Jan 07 15:14:39 implementing http://linuxtogo.org/~anonymous/netbook-booting.txt would be a start Jan 07 15:14:58 or convincing florian to build a 2.6.9 kernel with kexec support Jan 07 15:18:29 koen: for so complicated boot system kexec looks best Jan 07 15:18:46 btw - someone tried to build RPMs with OE? Jan 07 15:23:28 h4000 built Jan 07 15:23:41 hmm. I need ccache there Jan 07 15:25:27 03pfalcon 07org.oe.dev * rdea29766... 10/ (3 files in 3 dirs): linux-handhelds-2.6: Add nice clean defconfigman-generated defconfig for h3900. Jan 07 15:26:11 posokolvsky: can you add defconfigman recipe into OE? Jan 07 15:27:02 hrw: Good morning;) Have you had a chance to look at bug 1755? Jan 07 15:27:14 !oebug 1755 Jan 07 15:27:15 * * Bug 1755, Status: NEW, Created: 2007-01-06 15:04 Jan 07 15:27:16 * * dothedog(AT)comcast.net: mysql-4.1.18 build fails on gen_lex_hash Jan 07 15:27:17 * * http://bugs.openembedded.org/show_bug.cgi?id=1755 Jan 07 15:27:19 yes Jan 07 15:27:29 hrw, well, it's just a python script not even requiring installation ;-). And it's not clear how it would fit so far. I'd like to get some general feedback first ;-) Jan 07 15:28:23 posokolvsky: I like to have all OE related stuff available after some bitbake work ;D Jan 07 15:28:57 hrw, ok, then I'll add it ;-) Jan 07 15:29:20 hrw: what do I need to do? Jan 07 15:29:32 dothedog: w8 a moment - I will push some stuf Jan 07 15:29:42 hrw: excellent! Jan 07 15:30:09 posokolvsky: I build all needed stuff for host using OE Jan 07 15:30:18 posokolvsky: MACHINE=native bitbake ccache now Jan 07 15:30:25 nice ;-) Jan 07 15:31:13 CIA-4: BEWARE Jan 07 15:31:21 ~20 changest Jan 07 15:32:06 pushed Jan 07 15:32:18 hrw: I'll give it a shot. Jan 07 15:32:19 dothedog: pull/update and rebuild Jan 07 15:32:50 03hrw 07org.oe.oz354x * rf1f40c30... 10/ (1 classes/base.bbclass): Jan 07 15:32:50 base.bbclass: handle tar.Z archives - close #1433 (from .dev) Jan 07 15:32:50 applied changes from de17215ca0ec2c4f638399b95baf36a6e60b7bd4 Jan 07 15:32:50 through af3514ac399dfad05cfb417dadad1316f57a1f3c Jan 07 15:32:53 03hrw 07org.oe.oz354x * rfbced6b1... 10/ (1 classes/base.bbclass): Jan 07 15:32:53 base.bbclass: make showing display revision work with the mtn 0.30 workspace format (from .dev) Jan 07 15:32:54 applied changes from a0629352152e47a18449beec57ef52289d116470 Jan 07 15:32:56 through 98967f30d170009882b65d09d686fc199579b15f Jan 07 15:32:58 03hrw 07org.oe.oz354x * r4cbb1494... 10/ (1 packages/curl/curl_7.16.0.bb): Jan 07 15:33:00 curl: add 7.16.0, make linking against gnutls work (from .dev) Jan 07 15:33:03 applied changes from 74b62afa9d9b5f127553b347e236dccae5e5f546 Jan 07 15:33:04 through 6afc7d70704deaac82cc70ddb91183805e3f8802 Jan 07 15:33:06 03hrw 07org.oe.oz354x * rab154145... 10/ (7 files in 2 dirs): gpe-mini-browser: dropped not used versions Jan 07 15:33:11 03hrw 07org.oe.oz354x * r43a39660... 10/ (1 packages/gpe-mini-browser/gpe-mini-browser_0.19.bb): gpe-mini-browser: fixed SRC_URI to point to LTG, bumper PR due to curl update Jan 07 15:33:17 03hrw 07org.oe.oz354x * r52f9e3ab... 10/ (1 packages/gtk-webcore/osb-nrcit_20050430.bb): osb-nrcit: bump due to curl update Jan 07 15:33:19 03hrw 07org.oe.oz354x * r6a33f5f9... 10/ (4 files in 3 dirs): Jan 07 15:33:21 libgpg-error: added 1.3 (from .dev) Jan 07 15:33:25 applied changes from f980f926d4ed3ba38a98024e535fd402f176cd7b Jan 07 15:33:25 through 9cab53e99b0127c5e70d7658ff04bb3209a7e611 Jan 07 15:33:27 03hrw 07org.oe.oz354x * rece43327... 10/ (8 files in 4 dirs): libgpg-error: dropped old versions Jan 07 15:33:29 03hrw 07org.oe.oz354x * r2b607ac4... 10/ (19 files in 5 dirs): coreutils: added 5.3.0 from .dev, dropped older Jan 07 15:33:34 03hrw 07org.oe.oz354x * rcd9e3a05... 10/ (11 files in 4 dirs): libusb: update to 0.1.12 Jan 07 15:33:38 03hrw 07org.oe.oz354x * r96e195d5... 10/ (1 packages/libusb/libusb_0.1.12.bb): libusb: fix do_stage() Jan 07 15:33:40 03hrw 07org.oe.oz354x * rb36de34f... 10/ (53 files in 10 dirs): dbus/dbus-glib: updated to 1.0.2/0.72 and dropped older Jan 07 15:33:43 03hrw 07org.oe.oz354x * r48b7e771... 10/ (1 packages/curl/curl_7.16.0.bb): curl 7.16: package ca-certs (88kB), needed for ssl (e.g https) and RRECOMMEND it for libcurl (from .dev) Jan 07 15:33:46 03hrw 07org.oe.oz354x * r0a22f8dd... 10/ (4 files in 4 dirs): (log message trimmed) Jan 07 15:33:50 uicmoc 2.3.10, 3.3.5: add patch to make it build on 64bit platforms. Jan 07 15:33:52 Thanks Darren Edmundson for writing the patch and Eric Shattow for testing. Jan 07 15:33:55 uicmoc3: fix a buglet in the 64bit-cleanup patch Jan 07 15:33:56 applied changes from b10f1ed461d73562b22f666803d0299ed187ca8a Jan 07 15:33:58 through 0d418c5552d13b58959ab3334bcae1de8f796828 Jan 07 15:34:00 applied changes from bdfe3d888ff9f0ac9b669e49dc70d55553650a79 Jan 07 15:34:02 03hrw 07org.oe.oz354x * rb8c02e6b... 10/ (1 site/i686-linux): site/i686-linux: add ac_cv_sizeof_off_t to get mysql-native build - part of #1743 Jan 07 15:34:07 03hrw 07org.oe.oz354x * reeae3645... 10/ (1 packages/mysql/mysql-native_4.1.18.bb): mysql-native: disable innoDB - close #1743 Jan 07 15:34:10 03hrw 07org.oe.oz354x * r9552ef0d... 10/ (1 site/x86_64-linux): site/x86-64-linux: add ac_cv_sizeof_off_t to get mysql-native build - part of #1743 Jan 07 15:34:15 03hrw 07org.oe.oz354x * rd1ee7443... 10/ (1 packages/ipkg/ipkg-collateral.bb): ipkg-collateral: mark /etc/ipkg.conf as CONFFILE - close #1713 Jan 07 15:34:20 03hrw 07org.oe.oz354x * rd9960852... 10/ (10 files in 5 dirs): apt: dropped as OZ does not use DPKG/APT Jan 07 15:36:01 CIA-4: you are flooding Jan 07 15:36:27 03hrw 07org.oe.oz354x * r7a3ad4e2... 10/ (1 packages/ipkg/ipkg-collateral.bb): ipkg-collateral: its /etc/ not /etc/etc Jan 07 15:36:30 03hrw 07org.oe.oz354x * r5c0e6ace... 10/ (1 packages/curl/curl_7.16.0.bb): curl: fix EXTRA_OECONF to point to proper dir in staging Jan 07 15:37:04 now time to close some bugs Jan 07 15:37:57 hrw: impressive list of commits Jan 07 15:38:03 philippe: one day of work Jan 07 15:39:39 dothedog: remember to mark bugs as verified when you willg et mysql(native) built Jan 07 15:41:02 hrw: Stupid question. I get "/tmp/mtn.right.CHILDE" 18L, 286C Jan 07 15:41:02 Press ENTER or type command to continue Jan 07 15:41:14 how do I merge? Jan 07 15:42:01 dothedog: best way: revert your changes and do update again Jan 07 15:42:41 hrw: Never done that before, what is the command? Jan 07 15:43:34 mtn revert FILENAME Jan 07 15:44:31 hrw: got it. Jan 07 15:44:38 hrw: is rd1ee7443. in dev as well? Jan 07 15:44:52 h5000 built Jan 07 15:45:13 koen: which one it is? Jan 07 15:45:31 [17:03] CIA-4: hrw org.oe.oz354x * rd1ee7443... / (1 packages/ipkg/ipkg-collateral.bb): ipkg-collateral: mark /etc/ipkg.conf as CONFFILE - close #1713 Jan 07 15:46:04 koen: not yet Jan 07 15:46:31 hrw: I get this: mtn revert ../org.openembedded.oz354x/packages/mysql/mysql-native_4.1.18.bb Jan 07 15:46:31 mtn: misuse: workspace required but not found Jan 07 15:47:08 dothedog: cd ../org.openembedded.oz354x/packages/mysql/;mtn revert Jan 07 15:47:49 hrw: Ahhh, that worked. Jan 07 15:54:10 hrw: Got something else now: mtn: misuse: path _MTN/tmp/1 already exists Jan 07 15:54:30 dothedog: rm -rf _MTN/tmp;mtn update Jan 07 15:55:36 hrw: mtn: misuse: path 'packages/coreutils/coreutils-5.3.0' already exists, cannot create Jan 07 15:55:46 03hrw 07org.oe.dev * raf9734d1... 10/ (1 packages/ipkg/ipkg-collateral.bb): Jan 07 15:55:46 ipkg-collateral: added some tweaks from OpenZaurus Jan 07 15:55:46 - packaging as 'all' as this is not arch specific Jan 07 15:55:46 - marked ipkg.conf as CONFFILE to not get it removed on upgrade - close #1713 Jan 07 15:56:34 dothedog: remove Jan 07 15:58:11 hrw: now I get: mtn: misuse: 66 missing files Jan 07 15:58:20 Am I doing something stupid? Jan 07 15:58:35 dothedog: no. mtn does own stupid tricks on you Jan 07 15:58:58 dothedog: you have some files changed and monotone has problem on updating. Jan 07 15:59:45 dothedog: you can fight with it or checkout branch again and then look for your changes Jan 07 15:59:45 hrw: Should I do this? mtn: misuse: 'mtn revert --missing' Jan 07 16:00:08 dothedog: if you didnot changed things in repo then yes Jan 07 16:02:14 hrw: do you mean move the org.openembedded.oz354x to something else and checkout again? Jan 07 16:02:53 yes Jan 07 16:03:06 h6300 built Jan 07 16:06:02 posokolvsky: ERROR: Unable to parse conf/bitbake.conf (conf/machine/htcblueangel.conf:32: unparsed line: ' blueangel_battery blueangel_leds "') Jan 07 16:11:41 03hrw 07org.oe.oz354x * rf813ebcf... 10/ (10 files in 4 dirs): tin: updated to 1.9.1 Jan 07 16:11:45 03hrw 07org.oe.dev * r77b6bc0a... 10/ (1 conf/machine/htcblueangel.conf): htcblueangel: fix parse error Jan 07 16:15:04 posokolvsky: http://ewi546.ewi.utwente.nl/tinderbox/showlog.pl?machine_id=177&logfile=20070107173621.log - htcuniversal FAIL Jan 07 16:16:10 hrw: looks like wrong kernel version/wrong defconfig Jan 07 16:16:30 __NR_openat was only implemented for arm in 2.6.19 iirc Jan 07 16:16:48 koen: thats something for machine maintainers Jan 07 16:16:56 hrw: yep Jan 07 16:17:05 when this will end I will post to OE Jan 07 16:17:16 hrw: cool, thanks for testing Jan 07 16:17:27 'this week ARM broken machines are:' Jan 07 16:17:41 koen: thanks goes to CELF for machine Jan 07 16:22:05 hi! can someone please help me - i'm trying to compile opie-image (angstrom build) but it tells me that qemu needs gcc 3.x while i provide it with gcc 4.x. how do i tell it to compile it using gcc 3.x? Jan 07 16:22:50 shirour: install gcc-3.4 or gcc-3.3 in system Jan 07 16:23:17 what is in the nc file koen pushed yesterday? Jan 07 16:23:23 the .inc file? Jan 07 16:24:48 nope Jan 07 16:28:32 hrw: just built mysql-native and it works now. I put a comment in bug 1743. Question, who actually closes them? Jan 07 16:29:28 dothedog: you can as reporter Jan 07 16:29:42 hrw: will do Jan 07 16:29:47 thx Jan 07 16:30:44 hrw: marked as closed Jan 07 16:30:49 thx Jan 07 16:30:53 hrw: but how do i tell bitbake to use it for qemu's compilation process? Jan 07 16:31:37 shirour: qemu-native recipe has check for gcc-3.3, gcc-3.4, gcc32 inside - you just have to have them in system Jan 07 16:31:48 hrw: even better - can i just use premade package and turn its flag off in the conf file? Jan 07 16:32:09 koen: hx2000 is not supported in Angstrom? Jan 07 16:32:25 shirour: can. ASSUME_PROVIDED += "qemu-native" Jan 07 16:32:41 hrw: great..! and do you recommend of doing this? Jan 07 16:32:50 shirour: I recommend using OE version Jan 07 16:33:05 hrw: ok. so installing gcc3.4 as well is the best solution..? Jan 07 16:33:13 koen: supported as having kernel version set atleast Jan 07 16:33:18 shirour: yes Jan 07 16:33:53 hrw: I think the plan is to try and get some images going with it Jan 07 16:34:06 RP: git kernel got built Jan 07 16:34:43 hrw: as RP said :) Jan 07 16:35:02 hrw: RP and I agreed to sort the kernel version via default_preference Jan 07 16:35:23 ok Jan 07 16:35:35 NOTE: package task-base-1.0: completed Jan 07 16:35:36 The git kernel will become 2.6.20 at which point we'll probably make it the default Jan 07 16:36:17 hrw: ok, i've installed gcc3.4. in order to continue the compilation process from the same place i just need to run "bitbake opie-base" again? Jan 07 16:37:51 shirour: task-base will be enough for you now ;D Jan 07 16:38:46 hrw: you mean enough to get started? Jan 07 16:39:50 yep Jan 07 16:40:06 hrw: what does it contain..? Jan 07 16:41:02 <[cc]smart> where are fetches of sourcecode actually performed in OE ? Jan 07 16:41:08 hrw: btw, what dist. do you recommend? i have kernel 2.6.15 running Jan 07 16:41:22 hrw: so i'm afraid angstrom may not work Jan 07 16:41:27 shirour: I would update kernel to 2.6.16/18 and use angstrom Jan 07 16:41:45 shirour: or you can use generic or openzaurus-unstable Jan 07 16:41:53 hrw: what about familiar? Jan 07 16:42:08 shirour: submit machine config and kernel to OE so it will be included in my builds Jan 07 16:42:15 familiar is not supported here Jan 07 16:43:05 hrwl: currently i just use the omap one with a very few changes Jan 07 16:43:40 shirour: but own kernel Jan 07 16:44:06 hrw: yes. oh, you mean the kernel patch Jan 07 16:44:28 hrw: yeah, i'll do that soon. currently we're working on it a bit... Jan 07 16:44:37 hrw: mysql works now too. I marked 1767 as fixed. Jan 07 16:44:48 dothedog: thx for reporting and testing Jan 07 16:45:05 hrw: no problem. Jan 07 16:45:20 shirour: what do you feel about OE now after started to use it? was it worth? Jan 07 16:45:21 hrw: thanks for fixing it so fast! Jan 07 16:46:20 hrw: i really hope so, i hardly compiled anything yet Jan 07 16:46:55 hrw: the concept looks great, i wish it would work Jan 07 16:47:57 !oebug 1713 Jan 07 16:47:58 * * Bug 1713, Status: RESOLVED, Created: 2006-12-23 11:08 Jan 07 16:47:59 * * bugs.openembedded.org(AT)rolf.leggewie.biz: ipkg-collateral overwrites config file /etc/ipkg.conf Jan 07 16:48:00 * * http://bugs.openembedded.org/show_bug.cgi?id=1713 Jan 07 16:48:08 Laibsc1: fixed Jan 07 16:48:14 hrw: I know Jan 07 16:48:14 hrw: which dist. do you think i should choose- generic or openzaurus..? Jan 07 16:48:24 hrw: Did you push the fix into dev as well? Jan 07 16:48:33 hrw: Or is that not necessary? Jan 07 16:48:39 hrw: is there anyway by which i can build angstrom for older kernels? Jan 07 16:48:39 shirour, which one are you using now? Jan 07 16:48:39 shirour: for now I suggest generic and TARGET_FPU = "soft" Jan 07 16:48:51 shirour: ah.. yes.. Jan 07 16:49:08 shirour: DISTRO="angstrom-2007.1-oabi" is too new for me ;) Jan 07 16:49:23 Laibsc1: pushed into .dev too - forgot to comment Jan 07 16:49:33 OK Jan 07 16:49:37 nice, thanks. Jan 07 16:50:47 03rpurdie * r737 10bitbake/lib/bb/ (build.py runqueue.py): build.py: Merge stamp_is_current and stamp_is_current_cache into one function. This needs rethinking eventually. Jan 07 16:52:04 RP: When are you going to push trunk to stable? I start to loathe not being able to "configure somepackage" in bitbake -I Jan 07 16:52:13 Would be very nice. Jan 07 16:52:24 hrw: great.. thanks..! but the one that is marked by "oabi" is the one without it? Jan 07 16:53:36 03rpurdie * r738 10bitbake/lib/bb/build.py: build.py: Add del_stamp function to remove existing stamps (factoring common functionality into a shared function). Jan 07 16:55:27 shirour: eabi works in 2.6.16+ kernels. you have older so use -oabi version of angstrom Jan 07 16:55:33 koen probably can tell more Jan 07 16:55:44 03koen 07org.oe.dev * rbe981942... 10/ (1 packages/gcc/gcc3-build-cross.inc): gcc3-build-cross.inc: add safeguard for gcc-cross-initial Jan 07 16:55:50 you can patch 2.6.15 easily as well Jan 07 16:56:33 Laibsch: What do you mean? You mean a bug in bitbake 1.6? Jan 07 16:56:44 koen: so that it will support eabi? how can i do that? Jan 07 16:57:34 03rpurdie * r739 10bitbake/lib/bb/ (cooker.py runqueue.py shell.py): runqueue.py: Pass cooker to prepare_runqueue (for now) Jan 07 16:57:46 shirour: with the patches from http://www.arm.linux.org.uk/developer/patches/search.php?summary=eabi Jan 07 16:57:53 !oebug 1656 Jan 07 16:57:54 * * Bug 1656, Status: RESOLVED, Created: 2006-12-05 17:51 Jan 07 16:57:55 * * bugs.openembedded.org(AT)rolf.leggewie.biz: bitbake shell chokes on configure at times Jan 07 16:57:56 * * http://bugs.openembedded.org/show_bug.cgi?id=1656 Jan 07 16:58:07 RP: that one and bug 1654 Jan 07 16:58:25 They are not in trunk but in stable bitbake. Jan 07 16:59:48 koen: great, thanks..! so all i need to do is apply the patch for 2.6.15 and i'll have eabi?? Jan 07 17:00:10 4 or 5 patches, but yes Jan 07 17:00:25 3101 thru 3112 iirc Jan 07 17:01:41 koen: but isn't that for the 2.6.14? Jan 07 17:01:44 Laibsch: The problem is trunk has other issues :-( Jan 07 17:02:08 shirour: it will apply to 2.6.15 as well Jan 07 17:02:23 koen: great, thanks..! so 3101-3112 it is Jan 07 17:06:03 shirour: 3101/1, 31021, 3103/1, 3104/1, 3105/4, 3106/2, 3107/3 are the most important ones Jan 07 17:06:42 ok. i hope i wouldn't have problems... what's the best way to know if it works? Jan 07 17:06:53 boot it :) Jan 07 17:07:14 oh there's a kernel boot message? Jan 07 17:07:36 do i need to turn it on in menuconfig? Jan 07 17:08:35 hx4700 built Jan 07 17:09:33 shirour: CONFIG_AEABI=y Jan 07 17:09:40 CONFIG_OABI_COMPAT=y Jan 07 17:10:22 koen: ok, thanks..! Jan 07 17:10:34 this way you willhave kernel which will work with OABI and EABI rootfs Jan 07 17:10:52 which one is better? Jan 07 17:11:01 EABI is future for arm Jan 07 17:14:50 ks8695 built Jan 07 17:16:09 RP: bitbake trunk has var which holds count of tasks in progress? Jan 07 17:16:53 re Jan 07 17:16:56 RP: I'm thinking about adding small (local) patch which would give information how many tasks are in progress Jan 07 17:17:13 logicpd-pxa270 started Jan 07 17:20:12 03rpurdie * r740 10bitbake/lib/bb/ (build.py cooker.py runqueue.py): Make -f (force) flag handling work properly Jan 07 17:20:18 hrw: No, it doesn't keep count. I should change that as it would be simple to do... Jan 07 17:20:41 what is alchemy? Jan 07 17:20:54 Crofton: mips family of cpus Jan 07 17:21:02 previously AMD Jan 07 17:21:12 doh Jan 07 17:21:14 http://openalchemy.org/index.php/Main_Page Jan 07 17:21:46 Crofton: or UI for Amida Jan 07 17:21:49 Crofton: nice UI btw Jan 07 17:22:29 Crofton: I saw that on one event and UI was nice (amida itself was not so nice) Jan 07 17:23:17 * RP is happier with that set of changes out the way to bitbake Jan 07 17:24:18 RP: cleaning poky changesets? Jan 07 17:24:18 morning :-) Jan 07 17:24:21 hi likewise Jan 07 17:24:42 hrw: No, just plain fixing bugs. That one is also present in poky... Jan 07 17:28:26 logicpd-pxa270 built Jan 07 17:29:31 magician is HTC magician? Jan 07 17:30:43 03rpurdie * r741 10bitbake/lib/bb/shell.py: shell.py: Fix up force flag handling for directly named files Jan 07 17:30:47 http://www.linux-watch.com/news/NS7056717365.html Jan 07 17:32:01 Crofton: yes, we know about n800 ;D Jan 07 17:32:28 posokolvsky: can we rename magician to htcmagician? Jan 07 17:33:37 I am slow Jan 07 17:35:28 Crofton: well good excuse, you're lagging some timezones. Jan 07 17:36:04 (but we always work late :-) Jan 07 17:37:54 anyone know for sure what processor is in N800? Jan 07 17:38:18 rumours say omap2420 Jan 07 17:48:09 03rpurdie * r742 10bitbake/lib/bb/runqueue.py: runqueue.py: Give the user some idea of the number of tasks involved Jan 07 17:49:18 hrw: Just for you ;-) Jan 07 17:52:36 thx Jan 07 17:53:16 xf86-video-vesa-X11R7.1-1.2.0 is broken Jan 07 17:53:32 NOTE: Running task 811 of 968 (ID: 43, /home/hrw/devel/oe/org.openembedded/packages/xorg-driver/xf86-video-vesa_X11R7.1-1.2.0.bb, do_fetch) Jan 07 17:53:36 ;) Jan 07 17:53:52 mainstone built Jan 07 17:54:25 hrw: You'll understand the number when it finishes building... Jan 07 17:55:01 k Jan 07 17:55:44 hrw: Basically we can't know in advance how many stamps are going to be valid. It skipped the first 811 tasks in this case... Jan 07 17:56:03 understand Jan 07 17:56:22 mtx-3 config is broken - use 'arm' packagearch Jan 07 18:04:53 hmm, I'm running into an issue with base.bbclass, but I do not want to touch that Jan 07 18:07:54 koen: ping Jan 07 18:08:02 likewise: pong Jan 07 18:08:16 koen: http://www.openembedded.org/viewmtn/diff.psp?id1=77b6bc0a2e2c6465ab8132383765456617d1247f&id2=be981942fc1ce0d6a3e06da184fd2bdc2a662194 Jan 07 18:08:55 could you explain why this diff is there? I am running into a bug I am just debugging and it seems involved. Jan 07 18:09:18 !oebug #1769 Jan 07 18:09:19 * * Bug 1769, Status: NEW, Created: 2007-01-07 10:25 Jan 07 18:09:20 * * likewise(AT)gmx.net: libstdc++ .la files might be broken wrt dependency_libs line Jan 07 18:09:21 * * http://bugs.openembedded.org/show_bug.cgi?id=1769 Jan 07 18:10:39 koen: my current hypothesis is that there should be no surrounding spaces around the = sign. Jan 07 18:11:02 likewise: mtn pull :) Jan 07 18:11:47 03koen 07org.oe.dev * r5156b0a1... 10/ (1 packages/gcc/gcc3-build-cross.inc): gcc3-cross-build.inc: remove spurious spaces Jan 07 18:12:13 koen: hmm, so you think I am right? Jan 07 18:12:17 yes Jan 07 18:12:31 koen: but the spaces where already there before the diff I mentioned. Jan 07 18:12:50 koen: are was the file editing shortly before? Jan 07 18:12:55 s/are/or/ Jan 07 18:13:11 koen: starting a clean build for efika, just to be sure Jan 07 18:14:03 koen: that was a quick commit, btw Jan 07 18:19:18 NOTE: Tasks Summary: Attempted 968 tasks of which 920 didn't need to be rerun and 0 failed. Jan 07 18:19:23 nice summary Jan 07 18:19:50 omniorb does not build for efika Jan 07 18:20:17 Crofton: what error? Jan 07 18:20:30 koen: what do we do with armv4+ targets which has wrong package_arch? fix or report? Jan 07 18:20:34 I thin kI thin kto check some type sizes Jan 07 18:20:49 I am working on getting a running efika, then I will fix it Jan 07 18:21:08 looks very similar to a problem I had on arm, having to do with long double Jan 07 18:21:23 hrw: fix Jan 07 18:21:24 Crofton__: there are more packages that won't build for efika, musicpd/mdp is one I hit yesterday Jan 07 18:21:43 hrw: reporting an waiting for the fix just takes way too long Jan 07 18:21:48 * RP -> foog Jan 07 18:21:49 I am going to try and get mine running, then get my sdr stuff going on it Jan 07 18:21:50 food Jan 07 18:22:04 Crofton__: sdr? soft-radio? Jan 07 18:22:04 koen: gumstix and mtx-3 are broken then Jan 07 18:22:12 ~sdr Jan 07 18:22:15 Multicast Session Directory Tool. URL: http://www-mice.cs.ucl.ac.uk/multimedia/software/sdr/. Software Defined Radio Jan 07 18:22:31 software defined radio in my case :) Jan 07 18:22:39 Crofton__: My EFIKA runs fine, so if you get stuck, drop a line. Although I am not running from HDD or USB. Jan 07 18:22:58 I am going to try net boot this afternoon Jan 07 18:23:17 although I have come into work and looked for some sapcers so I can mount it in the carboard box it came in Jan 07 18:23:24 no luck though :( Jan 07 18:23:50 ~sapcers Jan 07 18:23:55 anyway, time to walk back home Jan 07 18:23:55 noha__: you are a towel! Jan 07 18:23:58 spacers Jan 07 18:24:02 dug Jan 07 18:24:04 duh Jan 07 18:24:07 :) Jan 07 18:24:08 mickey|squash: http://emea.kontron.com/index.php?id=53&detail_id=679 search for Lutz Jan 07 18:24:20 you're towel! Jan 07 18:25:07 morning Jan 07 18:25:18 morning XorA Jan 07 18:25:46 hey XorA Jan 07 18:26:16 03hrw 07org.oe.dev * r7679548c... 10/ (4 files in 3 dirs): xf86-video-vesa: fix CROSS COMPILE BADNESS Jan 07 18:26:27 ouch, that hurt my ears Jan 07 18:26:36 XorA: hi Jan 07 18:29:36 NOTE: Running task 2691 of 3084 (ID: 2053, /home/hrw/devel/oe/org.openembedded/packages/gnome/libsoup_2.2.98.bb, do_populate_staging) Jan 07 18:31:20 gqview is broken Jan 07 18:31:30 | gdk-pixbuf-csource --raw --extern --build-list folder_closed /home/hrw/devel/build/progear/tmp/work/i686-angstrom-linux/gqview-2.1.1-r1/gqview-2.1.1/src/icons/folder_closed.png folder_locked /home/hrw/devel/build/progear/tmp/work/i686-angstrom-linux/gqview-2.1.1-r1/gqview-2.1.1/src/icons/folder_locked.png folder_open /home/hrw/devel/build/progear/tmp/work/i686-angstrom-linux/gqview-2.1.1-r1/gqview-2.1.1/src/icons/folder_open.pn Jan 07 18:31:38 | /bin/sh: gdk-pixbuf-csource: command not found Jan 07 18:31:55 hrw: there are a lot more OE apps requiring that Jan 07 18:32:16 koen: so a lot more OE apps are broken? Jan 07 18:32:31 ooooh Nokia N800 Jan 07 18:32:37 libgtk-2.0-native time? Jan 07 18:32:49 XorA: you are another lagged user ;) Jan 07 18:33:09 hrw: not all of use are hardwired to internet yet :-) Jan 07 18:35:34 !oebug 1770 Jan 07 18:35:35 * * Bug 1770, Status: NEW, Created: 2007-01-07 11:07 Jan 07 18:35:36 * * openembedded(AT)hrw.one.pl: gqview need Native gdk-pixbuf-csource Jan 07 18:35:37 * * http://bugs.openembedded.org/show_bug.cgi?id=1770 Jan 07 18:35:48 koen: can I mark it as RC for angstrom? Jan 07 18:36:28 btw - mainstone built Jan 07 18:36:54 hrw: add gnumeric to 1770 Jan 07 18:37:21 XorA: you have www browser too ;) Jan 07 18:37:27 XorA: I'm not yet hit by gnumeric Jan 07 18:38:00 hrw: Ill forget by the time Ive fixed mplayer Jan 07 18:39:01 gerwinin: hi. sorry I did not read a msg from you. Jan 07 18:39:55 Likewise : ik gebruikte het volgende bordje : J7F4  Jan 07 18:40:15 Likewise : It seems I tried to send you the messages when I was offline Jan 07 18:40:53 Likewise: I received some mail from a company who is offering small embedded systems for 99 dollars Jan 07 18:40:57 Per 1 Jan 07 18:41:09 gerwinin: ok, nice board indeed. My Epia is suffering from a (well known) bad capacitors problem. Jan 07 18:41:14 hello Jan 07 18:41:18 gerwinin: nice boards? Jan 07 18:41:49 gerwinin: nice price.. Jan 07 18:42:26 I need some help and tweaks about eo Jan 07 18:42:54 i WANT TO CROSS COMPIL A GLIBC APPLICATION WITH THIS TOOL BUT I HAVE LOT OF PROBLEMS Jan 07 18:42:59 caps lock oupsi Jan 07 18:43:13 I use this "new" card http://www.linux-mips.org/wiki/Marvell_Firehawk Jan 07 18:43:19 Likewise , I had a similar problem Jan 07 18:43:41 Likewise : well the boards are okay, they are also offering webpads and stuff Jan 07 18:43:45 OE don't provide a 2.4 kernel Jan 07 18:43:50 Likewise but their webpads are a bit expensive Jan 07 18:43:53 Any hints which kernel is best to use for a simpad build? Jan 07 18:44:05 Likewise : 600 dollars Jan 07 18:44:11 gerwinin: Jan 07 18:44:22 gerwinin: Nokia N800 comes in at USD400 Jan 07 18:44:35 but is probably a tad smaller Jan 07 18:45:04 Likewise I can send you their folder Jan 07 18:45:25 gerwinin: I don't have money to spend on yet another embedded board right now, enough on my desk to play with Jan 07 18:45:52 gerwinin: can you send me copy? Jan 07 18:46:43 gerwinin: I am interested in what the $99 boards are though. Jan 07 18:46:52 hrw: yes of course Jan 07 18:47:29 Likewise : a bit of the same here Jan 07 18:47:37 hrw: I will send you their folder Jan 07 18:47:47 thx Jan 07 18:51:48 * XorA tries to expand days to 48 hours Jan 07 18:53:30 23 Jan 07 18:55:19 goxboxlive: htcuniversal is broken Jan 07 18:55:46 * philippe also wonders what you get for $99 Jan 07 18:56:13 100mhz cpu, 2m flash, 8m ram, serial + few gpio lines Jan 07 18:56:46 hrw: I know it is. My local is working, but what is the failure message? Jan 07 18:56:50 but linksys mipsel boards <100USD are mipsel, wifi, serial, ethernet etc.. Jan 07 18:56:52 goxboxlive: http://ewi546.ewi.utwente.nl/tinderbox/showlog.pl?machine_id=177&logfile=20070107173621.log Jan 07 18:59:40 XorA: abiword atleast is not affected Jan 07 18:59:57 h3900 built Jan 07 19:00:09 not time for htcblueangel Jan 07 19:00:13 s/not/now Jan 07 19:01:45 I speak in outer space.. no?? Jan 07 19:01:49 XorA: gnumeric failed. Jan 07 19:01:51 /bin/sh: raw: command not found Jan 07 19:01:51 make[4]: [../../src/pixmaps/gnumeric-stock-pixbufs.h] Error 127 (ignored) Jan 07 19:01:51 make[4]: Leaving directory `/home/hrw/devel/build/progear/tmp/work/i686-angstrom-linux/gnumeric-1.6.3-r0/gnumeric-1.6.3/src/pixmaps' Jan 07 19:02:02 XorA: http://ewi546.ewi.utwente.nl/tinderbox/showlog.pl?machine_id=181&logfile=20070107200731.log Jan 07 19:02:25 anyone setup dnsmasq to boot an efika? Jan 07 19:02:38 hrw: Yes, that is a bug in the .18-hh1 tag. It is fixed in the hh2 tag and above. But as i have said alot of times now, HTC Universal is not finished yet, it should only build the kernel from cvs and not any other tags. I will let you know when we have a fully working kernel, and then we can use tags. Jan 07 19:04:08 hrw: yeah thats bug 1770 Jan 07 19:04:23 But actually, that is not a universal bug, it was a bug in the c code of audit in the .18 kernel Jan 07 19:04:49 goxboxlive: then provide us with info which hh.org kernel release builds for universal Jan 07 19:05:25 goxboxlive: without kernel this machine will not get angstrom test images Jan 07 19:07:27 cu guys Jan 07 19:07:55 hrw: That will not be a problem, i diont think it is ready to have a release, we have to have fully working. But, as long I can have it in OE, and you commit my patches i and the rtest of us are happy with that. Jan 07 19:13:02 gerwinin: cris boards? Jan 07 19:13:32 cris^H^H^H^Haxis Jan 07 19:14:13 Cris ? Jan 07 19:14:21 Koen: cris boards ? Jan 07 19:14:41 your $99 embedded boards Jan 07 19:15:06 gerwinin: http://www.antratek.nl/Acme-Fox-board.html Jan 07 19:25:07 Koen: no this one is much more expensive Jan 07 19:25:20 Koen: I bought one and it costed me about 140 euro Jan 07 19:33:42 gerwinin: this board does not really compete with the EP93xx boards, I think Jan 07 19:35:38 (and we have no cris support I guess) Jan 07 19:37:51 what's this bitbake error (RunQueue, Index out of Range): http://pastebin.ca/309551 Jan 07 20:02:08 re Jan 07 20:02:12 likewise: update bitbake Jan 07 20:02:23 gerwinin: did you sent me infos? Jan 07 20:02:37 hrw: 1.7.2, rev 742 Jan 07 20:02:50 hrw: 1.6 seems to work ok. Jan 07 20:03:04 likewise: hm.. I just did some builds with 742 Jan 07 20:03:16 one of them is still going Jan 07 20:04:04 hrw: I was at 1.7.2, rev 735, which gave the error, then cleared the cache, went with 1.7.2 rev 742, which also gave the same error. Now 1.6 works. Jan 07 20:04:55 hrw: but there is something strange with the error I pasted: it has references to both the efika and ixp4xx build, so there might be something dangling here. Jan 07 20:05:09 likewise: BBPATH bad? Jan 07 20:05:51 hrw: no it is ok. I grepped for ixp4xx in the env output, and the config files. zero occurences. Jan 07 20:06:18 strange.. Jan 07 20:06:30 who maintain htcblueangel?? Jan 07 20:06:30 hrw: but I suspect the problem may be local, I will debug later Jan 07 20:09:14 hrw: rob_w is the blueangel dude Jan 07 20:09:43 I have one complaint to make Jan 07 20:10:05 B_Lizzard: we suck? Jan 07 20:10:10 we already know that :) Jan 07 20:10:14 :) Jan 07 20:10:45 ~botmail to rob_w HTC BlueAngel machine config in OE is broken - kernel modules need renaming. Jan 07 20:11:16 When installing multiple versions of GCC (like 4.1.1, and 3.3.4), the "gcc" and "g++" executables are not links to their respective binaries, but instaid, they're normal binaries Jan 07 20:11:26 Makes it harder to replace them and whatnot Jan 07 20:11:38 Switch between GCC versions Jan 07 20:12:26 koen: You drooling over the N800? Jan 07 20:12:31 :) Jan 07 20:13:38 B_Lizzard: yep Jan 07 20:13:52 I'm loving it Jan 07 20:14:21 Anyways, stupid complaint, but I guess it might be a valid thought or something like that Jan 07 20:14:42 mtx-3 is problematic machine Jan 07 20:14:46 bbiab Jan 07 20:17:28 03koen 07org.oe.dev * r1dec36a8... 10/ (1 conf/machine/htcblueangel.conf): htcblueangel: fix kernel-module naming Jan 07 20:25:47 Hello! Jan 07 20:31:52 hrw, those "safeguards" for gcc3, would they be needed for gcc4? anyone built gcc4 after today's koen patches? I got libsupc++ not found in gcc-cross-initial so far. Jan 07 20:33:55 posokolvsky: that's what the safeguards are for Jan 07 20:42:22 bbl Jan 07 20:49:29 What python package is imap.py in Jan 07 20:58:19 ok, after latest pull gcc4-initial built ok Jan 07 21:01:29 Hi Jan 07 21:02:02 sirfred|c760: hey Jan 07 21:02:07 sirfred|c760: nice xserver Jan 07 21:02:38 XorA: Thnx. Did you tested it? Jan 07 21:03:30 sirfred|c760: I am right now Jan 07 21:03:47 sirfred|c760: only thing I find so far, is it would be nice if xrandr supported 320x240 Jan 07 21:04:38 XorA: Yes. I have plans to do so. I think it's supported at framebuffer level. Jan 07 21:05:15 XorA: I just need to register that mode in rr config and test, I suppose Jan 07 21:05:37 XorA: Did you tested xv? Jan 07 21:06:11 * sirfred|c760 likes gaim on husky Jan 07 21:06:48 03xora 07org.oe.dev * r87665cf1... 10/ (1 packages/mplayer/mplayer_0.0+1.0rc1.bb): Jan 07 21:06:48 mplayer_0.0+1.0rc1.bb : enable x11 and xv output now we have a kdrive Jan 07 21:06:48 that does them Jan 07 21:07:38 XorA: I'm also thinking about implementing some of the XRender accel. Perhaps some blending ops can be done on hw Jan 07 21:07:49 sirfred|c760: yes, I just pushed the mplayer as you can see :-) Jan 07 21:08:56 anyone know a good reference to get efika to net boot? Jan 07 21:09:00 Great. Anyway there is an strange offset. The overlay surface is a little shifted from the colorkey rectangle (a small yellow zone in the upper side) Jan 07 21:09:00 good morning Jan 07 21:09:11 morning florian Jan 07 21:10:37 XorA: I've already made some changes, to detect the screen orientation instead of assuming it's 90 degrees rotated on startup Jan 07 21:11:55 XorA: A weird thing about mplayer is that it always start maximized, under matchbox wm Jan 07 21:12:59 XorA: The only way I was able to have it render the video at its original size was running without matchbox wm Jan 07 21:14:50 sirfred|c760: that is not weird. If mplayer opens an X11 window, matchbox will fullscreen that window (minus taskbar if there is one). It is default matchbox behaviour Jan 07 21:15:44 sirfred|c760: yes, that is a fault of mb-wm Jan 07 21:15:58 philippe: So, its a feature of matchbox Jan 07 21:16:04 sirfred|c760: I get a yellow border top and bottom of my test video Jan 07 21:16:27 sirfred|c760: my video is 320x176 Jan 07 21:16:34 doh Jan 07 21:16:34 XorA: Yes, some misscalculated offset I was not able to detect Jan 07 21:16:46 must actually connect ethernet to efika ..... Jan 07 21:17:33 XorA: I tested with 388x210 or so and only got the upper one. Jan 07 21:17:47 sirfred|c760: not to do with having to round to 4/8 pixels? Jan 07 21:19:39 XorA: The overlay surface is rounded to 16, but its painted black. So, it should cover the colorkey rectangle (yellow). I thought I was just placing the overlay badly, but if you get upper and down bars... Jan 07 21:20:30 XorA: What you are seeing is the colorkey rectangle that should be completely shadowed by the overlay Jan 07 21:21:04 And as the overlay surface is upper rounded... it's strange Jan 07 21:21:16 sirfred|c760: yeah, I am familiar with that effect Jan 07 21:21:53 XorA: You can always set the overlay colour to black. ;-) Jan 07 21:22:00 sirfred|c760: :-D Jan 07 21:23:23 | checking for .preinit_array/.init_array/.fini_array support... no Jan 07 21:23:23 | configure: error: Need linker with .init_array/.fini_array support. Jan 07 21:23:25 fsck Jan 07 21:23:32 XorA: I've also exposed an xv attr to limit the max overlay surface size that the driver should use. It tries to downscale the surface to make it fit into that size. Jan 07 21:23:56 XV_MAXOVERLAYSIZE IIRC Jan 07 21:24:16 XorA: You can see it using xvinfo Jan 07 21:27:50 sirfred|c760: Im just crunching a 320x240 to see how that looks Jan 07 21:28:30 sirfred|c760: I dont suppose there is any chance of vsync support? Jan 07 21:29:39 XorA: We can wait vsync, don't know how could it affect performance Jan 07 21:30:15 XorA: Anyway I didn't notice too much flickering in my tests Jan 07 21:30:36 sirfred|c760: I am seeing quite bad tearing Jan 07 21:30:46 XorA: Did you set fastsysclk to 1? Jan 07 21:32:56 sirfred|c760: ahah, I forgot all about that, yes thats much better Jan 07 21:33:22 XorA: I think it should be the default. Jan 07 21:35:25 sirfred|c760: yes, it looks like it Jan 07 21:35:43 re again Jan 07 21:35:54 goxboxlive: I will look at your mail tomorrow Jan 07 21:36:31 XorA: I will experiment with vsync, anyway Jan 07 21:36:49 hrw: Thats great, koen has allready commented it. Jan 07 21:37:03 goxboxlive: I know - I looked at those mails Jan 07 21:37:29 goxboxlive: some parts of patch probably can be cleaned and accepted Jan 07 21:38:11 Do you know of any place where I could buy a battery for my husky? Jan 07 21:38:28 It looks it's starting to drain faster Jan 07 21:38:43 hrw: thx Jan 07 21:38:47 sirfred|c760: laptopforless.com, brando shop Jan 07 21:39:14 hrw: thx. I will take a look. Jan 07 21:39:30 * sirfred|c760 tries with gpe-mini-browser Jan 07 21:42:05 goxboxlive: btw - I do not understand people which do not use name surname in mails Jan 07 21:43:11 hrw: well, i am afraid of koen, he might send a missile after me one day :-) Jan 07 21:43:51 goxboxlive: this is childish - you are under 20? Jan 07 21:44:18 :-) Jan 07 21:44:49 No way. Only "Error loading page" Jan 07 21:46:30 hrw: name/nick gives no more info in an email than each other if the person only knows you from IRC Jan 07 21:46:45 hrw: there are more people know me as XorA than my real name Jan 07 21:49:42 XorA: "goxboxlive " is.... Jan 07 21:50:06 :-) Jan 07 21:50:12 XorA: there is something like signature where I used to put infos like 'IRC: Hrw" Jan 07 21:50:34 XorA is what I send out a lot Jan 07 21:51:03 I'm seeing that when I close the lid of my husky, the backlight keeps on. Is there kernel support for the switch that detects the lid closing? Jan 07 21:51:34 sirfred|c760: there is Jan 07 21:51:38 sirfred|c760: yes, it emits an input subsystem event Jan 07 21:53:26 So, we should be able to switch off the light, shouldn't we? What level should that be implemented, zaurusd? Jan 07 21:53:47 sirfred|c760: zaurusd policy would be the natural choice Jan 07 21:54:21 sirfred|c760: controlled by a config file, which chooses suspend/lcdoff/nothing Jan 07 21:55:17 Perhaps w100fb could expose a file to do that, if it's not already doing Jan 07 21:58:20 Good night Jan 07 21:58:40 ok. diet-x11_X11R7.1-1.0.1.bb makes include/X11/Xlib.h disappear.... Jan 07 21:58:47 which is not good.... Jan 07 22:05:45 goxboxlive: sent answer Jan 07 22:05:48 hi Rod Jan 07 22:05:58 hey hrw Jan 07 22:07:00 mx21ads built Jan 07 22:11:46 hrw: Bug 1732 Jan 07 22:13:40 goxboxlive: pester us about bugs from time to time Jan 07 22:14:32 hrw: I forget it, since it works here localy. Jan 07 22:17:04 goxboxlive gamer tag ? :) Jan 07 22:20:01 ? Jan 07 22:22:20 nevermind. Jan 07 22:26:41 i have own feed, how can i regenerate a Package file? Jan 07 22:27:37 ipkg-make-index Jan 07 22:39:57 ~lart monotone for its SLOWAGE Jan 07 22:39:57 * ibot turns monotone into a lifesized tux doll for its SLOWAGE Jan 07 22:43:44 time to sleep Jan 07 22:43:45 cu Jan 07 23:25:09 likewise, ? Jan 07 23:25:16 Crofton|home: yes Jan 07 23:25:28 I can load my efika kernel vis tftp Jan 07 23:25:36 having trouble mounting nfs root Jan 07 23:26:01 seems like I am having trouble getting the path in Jan 07 23:26:09 bootserver=192.168.1.16, rootserver=192.168.1.106, rootpath= Jan 07 23:26:51 you use ip=dhcp on the kernel boot line? Jan 07 23:27:00 yeah Jan 07 23:27:12 um, I may have nfs server issues ... Jan 07 23:27:18 can'tshowmount from another machine Jan 07 23:29:55 This is my NFS root mount: http://pastebin.ca/309786 Jan 07 23:30:38 I suspect I buggered the nfs setup Jan 07 23:30:49 what does exportfs -rafv show? Jan 07 23:31:37 my /etc/exports: /nfsroot 192.168.*.*(sync,rw,no_root_squash,no_all_squash,subtree_check) Jan 07 23:31:40 nothing .... Jan 07 23:31:47 .oO(hilight) Jan 07 23:32:41 Crofton: then re-edit /etc/exports and run sudo exportfs -rafv to re-export. nfsd4 is running? Jan 07 23:32:42 give me a few minutes Jan 07 23:33:00 Crofton: it's bed time for me, but alas :-) Jan 07 23:33:00 yeah Jan 07 23:33:17 ok Jan 07 23:33:25 I think it is a nfs server sissue Jan 07 23:33:32 I should be able to figure it out Jan 07 23:37:48 if in doubt, disable the firewall :) Jan 07 23:38:06 Crofton: ok, let us know how it worked out. Here's my EFIKA local.conf (added some mp3 players to test sound): http://pastebin.ca/309791 Jan 07 23:38:12 :-) Jan 07 23:38:43 I am off to bed. good nite for later. Jan 07 23:39:03 gn Jan 07 23:39:05 ok Jan 07 23:39:09 looks like I found root Jan 07 23:40:05 no serial port login Jan 07 23:40:12 I can get in via ssh though Jan 08 00:00:08 efika has booted from teh hard drive Jan 08 00:03:14 good night **** ENDING LOGGING AT Mon Jan 08 02:59:57 2007