**** BEGIN LOGGING AT Tue Jan 18 02:59:57 2011 Jan 18 03:13:53 What exactly are you looking for? Pre-build distro? Toolchain? Or do you already have those, and desire a means to use them to create your own distro for the board? Jan 18 03:18:53 03Khem Raj  07master * r260f78b3e5 10openembedded.git/recipes/tgt/tgt-1.0.12/stub-out-posix-fallocate-uclibc.patch: Jan 18 03:18:53 tgt-1.0.12: Fix the stub-out-posix-fallocate-uclibc.patch Jan 18 03:18:53 Signed-off-by: Khem Raj Jan 18 07:07:42 cn anyone help me how to compile multiple c files in openembedded Jan 18 07:30:36 03Martin Jansa  07master * r506ab59480 10openembedded.git/recipes/freesmartphone/frameworkd_git.bb: Jan 18 07:30:37 frameworkd: bump SRCREV to fix KeyError: 'message-volume' issue Jan 18 07:30:37 Signed-off-by: Martin Jansa Jan 18 07:30:48 03Martin Jansa  07master * r0823619e7d 10openembedded.git/recipes/openmoko-3rdparty/ (4 files in 2 dirs): (log message trimmed) Jan 18 07:30:48 iliwi: add patch for vala-0.11.4 and patch from Ben Thompson Jan 18 07:30:48 * fix wrong null position in my Jan 18 07:30:48 0001-adapt-to-elementary-genlist_item_insert_before-API-c.patch Jan 18 07:30:48 * Ben Thompson fixed WPA issue Jan 18 07:30:48 * Pau Espin Pedrol added another patch for newer vala-0.11.4 Jan 18 07:30:49 * all patches were sent upstream, no reply yet Jan 18 07:59:00 gm Jan 18 08:11:23 03Martin Jansa  07master * re595c05a11 10openembedded.git/recipes/tslib/tslib_git.bb: Jan 18 08:11:24 tslib: force old SRCREV for palmpre Jan 18 08:11:24 * playya fixed it in his tslib repo, waiting for pull to kergoth's branch Jan 18 08:11:24 Signed-off-by: Martin Jansa Jan 18 08:12:42 good morning Jan 18 08:22:01 OQ Jan 18 08:22:03 morning Jan 18 08:52:22 hey bluelightning Jan 18 08:57:32 good morning Jan 18 09:03:11 03Eric BENARD  07org.openembedded.dev * rc99881cc22 10openembedded.git/recipes/busybox/busybox_1.18.2.bb: (log message trimmed) Jan 18 09:03:11 busybox-1.18.2: add fix Jan 18 09:03:11 - insmod: check for module read errors Jan 18 09:03:11 bb's commit id : 094cc51e50bdb877fa4c245dbde47e4dfbf94387 Jan 18 09:03:11 - following Bernhard Reutner-Fisher's request, busybox patches Jan 18 09:03:11 don't have anymore timestamp inside which should make them fetchable Jan 18 09:03:12 instead of integrating them in OE so I try this way. Jan 18 09:09:40 03Martin Jansa  07master * rb28783434f 10openembedded.git/recipes/coreutils/ (11 files in 3 dirs): Jan 18 09:09:40 coreutils: remove old broken versions Jan 18 09:09:40 * for details read http://www.mail-archive.com/openembedded-devel@lists.openembedded.org/msg14003.html Jan 18 09:09:40 Acked-by: Frans Meulenbroeks Jan 18 09:09:40 Signed-off-by: Martin Jansa Jan 18 09:09:52 03Martin Jansa  07master * r02782d09e2 10openembedded.git/recipes/procps/procps.inc: Jan 18 09:09:52 procps: use u-a for pmap, otherwise conflicts with pmap from busybox Jan 18 09:09:52 Signed-off-by: Martin Jansa Jan 18 09:18:26 03Stefan Schmidt  07master * r34f1803ff6 10openembedded.git/classes/bug-java-library.bbclass: Jan 18 09:18:26 bug-java-library.bbclass: Add a class to install jar files without symlinks Jan 18 09:18:26 Signed-off-by: Stefan Schmidt Jan 18 09:18:40 03Stefan Schmidt  07master * rd2b084e284 10openembedded.git/classes/jni-library.bbclass: Jan 18 09:18:40 jni-library.bbclass: Add helper class for jni libraries Jan 18 09:18:40 Signed-off-by: Stefan Schmidt Jan 18 09:18:43 03Stefan Schmidt  07master * r4be33a03f2 10openembedded.git/classes/bug-image.bbclass: Jan 18 09:18:44 bug-image.bbclass: Add class for bug images Jan 18 09:18:44 Signed-off-by: Stefan Schmidt Jan 18 09:21:19 hi ant_work, all Jan 18 10:54:16 03Martin Jansa  07master * r62eebc4473 10openembedded.git/recipes/freesmartphone/fso-sounds.bb: Jan 18 10:54:16 fso-sounds: add fso-sounds-extras package with all remaining sounds and RSUGGEST it Jan 18 10:54:16 Signed-off-by: Martin Jansa Jan 18 10:54:19 03Martin Jansa  07master * r96f00ef431 10openembedded.git/recipes/freesmartphone/frameworkd_git.bb: Jan 18 10:54:19 frameworkd: frameworkd-config add RDEPEND on fso-sounds Jan 18 10:54:19 Signed-off-by: Martin Jansa Jan 18 11:00:06 03Koen Kooi  07org.openembedded.dev * rc8195fa9cf 10openembedded.git/ (124 files in 6 dirs): Jan 18 11:00:06 linux-omap 2.6.37: replace various patch with upstream versions and rediff Jan 18 11:00:06 * The beagle poweron patch was split into 3 and accepted upstream thanks to Nishant Jan 18 11:00:06 * the 2 720MHz patches were combined and pending acceptance thanks to Sanjeev Jan 18 11:00:06 * the boardfile 720MHz patches aren't needed anymore Jan 18 11:00:06 Signed-off-by: Koen Kooi Jan 18 12:45:30 RP: ping Jan 18 13:05:44 any TSC member online? Jan 18 13:06:42 spellcasting T S C Jan 18 13:11:18 * mwester lays out candles and chalk lines for the Rite of AshkEnte Jan 18 13:14:32 03Koen Kooi  07org.openembedded.dev * r23e2a9d9a2 10openembedded.git/recipes/linux/linux-omap-2.6.37/beagleboard/defconfig: Jan 18 13:14:32 linux-omap 2.6.37: switch default governor to userspace Jan 18 13:14:32 Signed-off-by: Koen Kooi Jan 18 13:16:00 hehe Jan 18 13:22:47 /joint #yocto Jan 18 13:22:51 oops Jan 18 13:29:29 JaMa|Wrk: I think it's now the moment to RIP for kdrive on zaurus Jan 18 13:29:51 nobody has time to adapt the old stuff against th daily changes Jan 18 13:31:30 we know xorg-xserver it's still a bit heavier and memory hungry but I really don't see alternatves Jan 18 13:32:08 worst will happen to c7x0 which will loose the hw "acceleration" of imageon / W100 Jan 18 13:33:16 ant_work, thanks for looking after the older stuff Jan 18 13:33:20 ant_work: I'm using it on spitz since xserver-1.5.. but lately spitz is at home getting dust Jan 18 13:33:20 ant_work: we have xorg driver for imageon Jan 18 13:33:46 really? like kdrive-imageon? Jan 18 13:33:51 ant_work: yes Jan 18 13:33:58 ant_work: I have to be carefull with dust while rebooting it while testing new kernels on it :) Jan 18 13:34:04 I used it when I wrote the xorg support for GTA02 Jan 18 13:34:05 I must have been drunk last time I checked Jan 18 13:34:29 JaMa|Wrk: the real issue is power management... Jan 18 13:34:50 ant_work: yes, mine is running from adapter Jan 18 13:35:09 he he..2.6.27 won't boot if AC is connected Jan 18 13:35:15 2.6.37 ! Jan 18 13:35:24 imagine... Jan 18 13:36:08 here it boots ok with 2.6.37 Jan 18 13:36:48 on my C3200 there are issues (linux-kexecboot) Jan 18 13:36:48 but already twice I lost wlan connection to it :/ Jan 18 13:37:17 after flashing I have to unbranch the AC, otherways it freezes Jan 18 13:37:26 and because I was at home to reboot it only twice this year I didn't debug it much.. Jan 18 13:37:36 * XorA puts hands over ears and goes lalalalala Jan 18 13:38:06 I spend most of my free-time filterng youtube videos for my kid's iphone :/ Jan 18 13:38:20 * JaMa|Wrk throwing couple of zauruses on XorA singing lalalalala Jan 18 13:38:35 * XorA is finally zaurus free Jan 18 13:38:41 iphone and tablet are evil with children around.... Jan 18 13:38:46 too easy to tap :) Jan 18 13:39:10 My godson is most confused by the fact that cameras don't have touchscreens. Jan 18 13:39:22 * JaMa|Wrk lost use-case for C3000 when received n900 Jan 18 13:41:02 is still a cheap embedded-arm testbed, with keyboard and serial Jan 18 13:41:12 it's sad... we're always just a few steps away from complete support for these devices :/ Jan 18 13:41:48 it will be the same for SOC, see the patch flood on lkml Jan 18 13:45:45 * ant_work is getting insane reading too much together from http://www.flameeyes.eu/autotools-mythbuster/index.html Jan 18 13:47:31 ant_work: you just need to get hand dirty and start kernel hacking **** BEGIN LOGGING AT Tue Jan 18 14:13:58 2011 Jan 18 14:48:30 03Chris Larson  07master * rdc0190ae96 10bitbake.git/lib/bb/cooker.py: (log message trimmed) Jan 18 14:48:30 cooker: also respond to SystemExit for execute_runqueue Jan 18 14:48:30 Without this, bitbake appears to hang when encountering an invalid task. In Jan 18 14:48:30 reality, the SystemExit flows up uncaught, the server may go away, and the UI Jan 18 14:48:30 doesn't know about it, so requires us to ^C to exit. Jan 18 14:48:31 This is a temporary measure. Poky's bitbake handles invalid tasks Jan 18 14:48:32 differently, and avoids this *particular* issue, but we do need to ensure that Jan 18 14:51:19 kergoth, heh, thanks for fixing it :) Jan 18 14:51:55 np Jan 18 14:52:20 uncovered quite the oversight in the server's exception handling Jan 18 14:53:25 hrm Jan 18 15:21:41 bbl Jan 18 15:29:26 03Chris Larson  07master * rad5ce1595e 10openembedded.git/recipes/devmem2/devmem2.bb: Jan 18 15:29:26 devmem2: enable strict alignment Jan 18 15:29:26 Signed-off-by: Chris Larson Jan 18 15:37:48 kergoth: ping Jan 18 15:38:28 morphis: pong Jan 18 15:39:14 kergoth: can you apply too patches for tslib, which fixes cy8mrln-palmpre plugin? Jan 18 15:39:26 kergoth: they are here: http://amethyst.openembedded.net/~morphis/tslib/ Jan 18 15:40:12 merging from you guys is proving to be a royal pain due to the fact that i had to rebuild/rebase the history to fix that odd git repository problem. no longer as simple as just merging from for-upstream, too many conflicts Jan 18 15:40:16 * kergoth looks Jan 18 15:40:44 kergoth: you right, the playya/tslib repository is not the best git repository ... Jan 18 15:41:55 but the two patches are quite simple Jan 18 15:42:34 one fixes compilation issues someone added without testing for compilation before ... and the second one is even something somebody added without thinking about it before Jan 18 15:43:37 * kergoth nods Jan 18 15:43:40 morphis: hheheh Jan 18 15:43:40 willd o Jan 18 15:44:04 kergoth: but I hope things will be better in the future, sorry for the much work ... Jan 18 15:45:10 it should just be a matter of 1) generating a merge commit where we're synced up, or 2) rebasing the playya branches on master Jan 18 15:45:37 jepp Jan 18 15:48:00 ugh,i just tried to sudo apt-get install autoconf-native Jan 18 15:48:03 hehe Jan 18 15:48:35 haha Jan 18 15:48:46 kergoth: too used to bb names Jan 18 15:48:53 apparently :) Jan 18 15:49:29 morphis: applied Jan 18 15:49:41 kergoth: thanks very much Jan 18 15:49:55 kergoth: will bump tslib version in OE too, ok? Jan 18 15:50:39 k Jan 18 16:04:04 Did someone ever ported SDKMACHINE support from Poky to OE? Jan 18 16:35:09 03Simon Busch  07org.openembedded.dev * r60625aa689 10openembedded.git/recipes/freesmartphone/libmsmcomm_git.bb: Jan 18 16:35:09 libmsmcomm: adjust for new version and bump PR Jan 18 16:35:09 Signed-off-by: Simon Busch Jan 18 16:35:17 03Simon Busch  07org.openembedded.dev * re7fc2cf3be 10openembedded.git/recipes/freesmartphone/ (msmcomm-specs_git.bb msmcommd-specs_git.bb): Jan 18 16:35:17 msmcommd-specs: rename to msmcomm-specs and adjust for new upstream version Jan 18 16:35:17 Signed-off-by: Simon Busch Jan 18 16:35:19 03Simon Busch  07org.openembedded.dev * r432345c0b7 10openembedded.git/recipes/freesmartphone/msmcomm.inc: Jan 18 16:35:19 msmcomm: bump SRCREV Jan 18 16:35:19 Signed-off-by: Simon Busch Jan 18 16:35:24 03Simon Busch  07org.openembedded.dev * r7629fb2c1a 10openembedded.git/recipes/tslib/tslib_git.bb: Jan 18 16:35:25 tslib: bump SRCREV and remove extra SRCREV for palmpre Jan 18 16:35:25 Signed-off-by: Simon Busch Jan 18 16:35:27 03Simon Busch  07org.openembedded.dev * r907565a5aa 10openembedded.git/recipes/freesmartphone/msmcommd_git.bb: Jan 18 16:35:28 msmcommd: adjust for new version Jan 18 16:35:28 Signed-off-by: Simon Busch Jan 18 16:43:32 kergoth, i think there's a hole in dependees logging: something pulls in udev/acl udev/attr udev (although i am using mdev) but it's not diagnosed *who* pulls it in (AFAICS) See what i mean? Jan 18 16:45:32 blindvt`: yep, that should be changed. of course, you can use -g, but it'd be nice to not have to. *why* things are being built is a pretty important and common thing to want to investigate Jan 18 16:45:51 kergoth, my local micro-image.bb uses IMAGE_DEV_MANAGER = "busybox-mdev" and i have set BBMASK = "(udev|acl|attr)" Jan 18 16:46:11 -g is your friend :) Jan 18 16:46:24 and NoProvider doesn't have ._dependees (or however it's called) Jan 18 16:47:21 03Simon Busch  07org.openembedded.dev * rb07183dc5f 10openembedded.git/recipes/freesmartphone/fsogsmd_git.bb: Jan 18 16:47:21 fsogsmd: should now depend on msmcomm-specs rather than msmcommd-specs Jan 18 16:47:21 Signed-off-by: Simon Busch Jan 18 16:48:19 kergoth, another thing (that i asked yesterday or so, perhaps we can do that today, didn't look yet): Re +FULL_OPTIMIZATION_powerpc = "-O2": How can we say that we want that for gcc pv <= 4.5.x ? Jan 18 16:48:35 s/today/already/ Jan 18 16:49:35 there's no way to do that Jan 18 16:50:33 FOO_powerpc_pn_gcc_pv_lt_4.6.0 =+ "something" # this would be ugly beyond words and wouldn't work out for let's say 3 such version conditions Jan 18 16:50:52 kergoth, ideas on how we could tackle this? Jan 18 16:52:40 there's oe.version_less_or_equal that could be used in the value, perhaps. I'll try to remember to play with it Jan 18 16:53:43 there's no way for one recipe to inspect anything about another recipe without emitting files to disk Jan 18 16:54:08 and the config metadata doesn't know about recipes at all, that isn't available until the parsing completes and the runqueue gets prepared Jan 18 16:55:08 what i was talking to khem about yesterday was adding an explicit hook after the runqueue is prepared, where an event handler could inspect what is going to be built Jan 18 16:55:17 so you can see after the preferences have been processed, and respond to htat Jan 18 17:46:44 kergoth: Also bothering me at the moment is the contexts a lot of the anonymous python run in. I wonder if they'd be better styled similar to event handlers hooked at particular points but ones which can change the datastore Jan 18 17:52:17 RP: I think we should change anonymous python functions internally into RecipeParsed event handlers, consolidating the code Jan 18 17:53:01 03Simon Busch  07org.openembedded.dev * rfcd2c3184e 10openembedded.git/conf/distro/include/fso-autorev.inc: Jan 18 17:53:01 fso-autorev: msmcommd-specs -> msmcomm-specs Jan 18 17:53:01 Signed-off-by: Simon Busch Jan 18 17:58:13 kergoth: that could work... Jan 18 17:58:31 the only extra bit would be passing e.data as d, but that should be pretty straightforward Jan 18 17:59:18 something to mull over / create a prototype of, anyway Jan 18 17:59:20 see how it looks after Jan 18 18:07:35 does angstrom provide pre-built 'images'? I see their packages but can't find various machine images - can't seem to get their online builder narcissus to work - trying to investigate a package they have built that won't currently build with OE Jan 18 19:23:06 kergoth, is it OK to add a vim modeline? Jan 18 19:26:00 playya: fine with me Jan 18 19:26:17 ok Jan 18 19:46:42 i'm confused about oe, it says it's for cross compiling, but it started building everything for i686, where do you tell it that you have a cross compiler? Jan 18 19:47:09 you don't need to Jan 18 19:47:15 it's building the tools it needs to run while building everything else Jan 18 19:47:27 hence the -native postfix on those recipes Jan 18 19:47:36 the i686 things you see being built are native tools needed for the cross-compilation etc Jan 18 19:47:52 git is needed for cross compilation? Jan 18 19:48:04 needed to fetch the sources.... Jan 18 19:48:05 it's needed to fetch sources Jan 18 19:48:09 and not all build machines have it Jan 18 19:48:19 if you do, and don't want it to build it, add ASSUME_PROVIDED += "git-native" Jan 18 19:49:12 (this is *not* safe for everything. there are 3 different reasons for natives to be built: 1) we can't rely on it existing, but any will do, 2) we patch it, so we *must* have our version, or 3) we need a specific version as a part of the crosscompile process, e.g. for perl, python Jan 18 19:49:16 ) Jan 18 19:49:27 only 1) can be safely added to ASSUME_PROVIDED Jan 18 19:49:57 erm, 4) we require a newer version, e.g. we need m4, but it has to be >= 1.4.6(iirc) Jan 18 19:52:25 let's assume that i built a rootfs with OE, and a kernel and busybox with code sourcery toolchain. I use qemu to boot into busybox, mount my rootfs image and chroot into the OE rootfs, but I get illegal instruction Jan 18 19:52:42 that's because of the gnu-gcc vs code sourcery deal? or something else? Jan 18 19:53:21 you can use external toolchain with OE if you wish... Jan 18 19:53:55 you can't really mix your externally built busybox with libs built by OE Jan 18 19:53:56 i'm trying to understand the nature of this error before I spend many more hours going down a path that might not be right Jan 18 19:54:17 is there a reason your trying to replace busybox built by OE with one you built using an external toolchain? Jan 18 19:54:19 i'm not married to code sourcery, it was just easiet to get and try out Jan 18 19:54:32 yeah, i don't have a busybox built by OE Jan 18 19:54:41 try just building a full filesystem/kernel/bootloaders using OE Jan 18 19:55:03 or if you just want to play around with stuff, download and install using pre-built images Jan 18 19:56:09 I don't see any reason why you couldn't build your kernel externally with other tools (as long as your not trying to use out-of-tree modules built within OE or other OE packages that depend on kernel) - but you said above you were using busybox built with CSL and that would have library conflicts Jan 18 19:56:11 the prebuilt images is what gives me the illegal instruction when i try to chroot to it Jan 18 19:56:41 but, even when i don't load busybox and just try to boot the kernel the /sbin/init won't run Jan 18 19:57:06 not sure if i'm doing somethign wrong, or a kernel built with one tool won't run a fs built with another Jan 18 19:57:20 your building OE image for x86 target and chrooting to it from an x86 host just to play with the filesystem? Jan 18 19:57:59 i haven't done anything with OE yet, just downloaded a built image Jan 18 19:58:09 trying to use oe, but it keeps breaking Jan 18 19:58:10 for what architecture/machine? Jan 18 19:58:17 arm smartq5 Jan 18 19:58:28 but, not really, arm qemu Jan 18 19:58:30 so your using qemu to emulate that arch? Jan 18 19:58:39 yeah Jan 18 19:58:50 ya... not something I've done myself, but others here have Jan 18 19:59:35 really i'm just looking for a way to build python and directfb for arm Jan 18 20:00:48 there's nothing wrong with the pre-built images likely - its probably something to do with the qemu config or support I would think Jan 18 20:00:55 gnu-config-native-git-r0 is failing checksum Jan 18 20:00:59 you don't have target hardware to work with? Jan 18 20:01:07 i do Jan 18 20:01:12 if the download failed that will fail checksum Jan 18 20:01:41 rm gnu-config-native-git-* and build again... watch the msgs carefully for the do_fetch process of that pkg if it fails again Jan 18 20:02:12 so have you tried pre-built images on your target (not qemu)? Jan 18 20:02:40 hcl2: oe defaults to building its own toolchain. csl is nowhere near easiest :P Jan 18 20:03:02 Function base_do_fetch failed Jan 18 20:03:16 but i can get csl, untar it, set some env variables, and build a kernel Jan 18 20:03:29 pastebin the output of the do_fetch failure Jan 18 20:03:30 rnel. Jan 18 20:03:33 er, OE isn't a kernel. Jan 18 20:03:39 you're comparing apples and oranges here Jan 18 20:03:49 yes... sounds like your not understanding what OE is Jan 18 20:03:54 wait, i didn't compare them Jan 18 20:04:04 its a linux build system - builds bootloaders, packages, filesystem images and the toolchains needed to build those Jan 18 20:04:05 you said "csl is nowhere near easiest" Jan 18 20:04:42 to use CSL with OE as external toolchain its not easiest - requires some extra config and I don't know that its clearly documented (lots of questions about it) Jan 18 20:05:16 your doing something even more complex - your building your kernel externally with CSL, and coupling it to a rootfs built with OE, then trying to run those together under qemu (if I follow you correctly) Jan 18 20:06:18 i'm just tryign to couple it with any rootfs that will work Jan 18 20:06:35 but all of them seem to throw illegal instruction or the /sbin/init doesn't run Jan 18 20:07:10 i just want to add pygame to this device and it's like this never ending spiral of going back one step until it turns out that you have to compile your own distro Jan 18 20:07:45 sounds like your perhaps mixing architectures - cflag diffs or something with your qemu machine. Not clear why your using qemu if you have the actual target hardware Jan 18 20:08:07 as kergoth said your not doing things the easy way.... Jan 18 20:17:41 anyone know who qt4_embedded is at 4.6.3 instead of 4.7.1? any problems with 4.7.1? Jan 18 20:20:54 03Michael 'Mickey' Lauer  07org.openembedded.dev * r7818ffe4e7 10openembedded.git/recipes/vala/ (vala_0.11.4.1.bb vala_0.11.4.2.bb): vala: 0.11.4.1 -> 0.11.4.2 Jan 18 20:28:29 anr78, likely your distro conf Jan 18 20:28:44 distro/minimal.conf has PREFERRED_QT_VERSION ?= "4.7.0" Jan 18 20:29:05 (I'm using release-2012-12 so I may be dated) Jan 18 20:29:25 tharvey: yeah. I just noticed the DEFAULT_PREFERENCE of -1 in the 4.7.1 recipe Jan 18 20:29:32 the 4.7.0 and 4.7.1 recipes have DEFAULT_PREFERENCE = "-1" so you have to use PREFERRED_VERSION_qt4-embedded to select Jan 18 20:29:34 yup Jan 18 20:31:30 tharvey: I'm running Angstrom @ Beagleboard, so I guess the angstrom conf is the place to look Jan 18 20:32:59 hmmm... I'm using angstrom as well Jan 18 20:34:06 tharvey: it says 4.6.3 in conf/distro/include/angstrom-2008-preferred-versions.inc Jan 18 20:34:12 distro/include/angstrom-2010-preferred-versions.inc Jan 18 20:34:25 ya, just saw that as well Jan 18 20:34:26 ah, yeah. Jan 18 20:34:30 use 2010? Jan 18 20:34:40 wonder if I have 2008 or 2010.... Jan 18 20:34:40 or set in a local.conf Jan 18 20:34:42 :) Jan 18 20:35:43 at home now, so I'll have to check at work tomorrow. Jan 18 20:37:17 anr78, ah... I'm using 2008 as well, but I do have a PREFERRED_VERSION_qt4-embedded in one of my local conf's Jan 18 20:38:27 tharvey: and it's working well? Jan 18 20:41:10 anr78, 4.7.0 is what I defined - haven't tried 4.7.1 yet for no real reason - working fine for me Jan 18 20:42:15 tharvey: ok Jan 18 20:48:11 03Michael Smith  07master * r9d92b4fd30 10openembedded.git/recipes/openjdk/openjdk-6-common.inc: Jan 18 20:48:11 openjdk-6-common.inc: remove reference to do_qa_configure Jan 18 20:48:11 do_qa_configure is only defined for distros that inherit insane. Jan 18 20:48:11 Signed-off-by: Michael Smith Jan 18 20:48:11 Cc: Henning Heinold Jan 18 20:48:25 03Michael Smith  07master * rffaf836d25 10openembedded.git/recipes/icedtea/ (3 files in 2 dirs): (log message trimmed) Jan 18 20:48:25 icedtea6-native 1.7.5: fix build on old build hosts Jan 18 20:48:25 Set ALT_FREETYPE_HEADERS_PATH so openjdk-ecj doesn't look in Jan 18 20:48:25 /usr/include for freetype2 headers. Jan 18 20:48:25 Build xawt sizer.32 on x86_64. A precompiled sizer.32 is included in Jan 18 20:48:25 the icedtea tarball, but this could be an accident, and it doesn't Jan 18 20:48:26 run on older build systems that can't handle DT_GNU_HASH executables; Jan 18 21:44:27 hello all Jan 18 22:11:45 03Khem Raj  07master * re6c26bacb7 10openembedded.git/recipes/uclibc/uclibc-git/nslu2be/uClibc.machine: Jan 18 22:11:45 uclibc-git: Add defconfig for nslu2be Jan 18 22:11:45 Signed-off-by: Khem Raj Jan 18 22:11:47 03Khem Raj  07master * raec2d12c9f 10openembedded.git/conf/distro/ (include/slugos.inc ucslugc.conf): Jan 18 22:11:47 ucslugc.conf: Revitalize, can not build slugos-image with uclibc Jan 18 22:11:47 Signed-off-by: Khem Raj Jan 18 22:11:48 03Khem Raj  07master * r2a31f77a17 10openembedded.git/recipes/uclibc/ (uclibc-git/mips64-no-waitpid.patch uclibc_git.bb): Jan 18 22:11:48 uclibc_git.bb: Fix the build on mips64/nptl Jan 18 22:11:48 * It does not yet boot though but one step at a time Jan 18 22:11:48 Signed-off-by: Khem Raj Jan 18 22:11:49 03Khem Raj  07master * r4526008b14 10openembedded.git/recipes/uclibc/uclibc-git/ixp4xx/ (uClibc.distro uClibc.machine): Jan 18 22:12:09 uclibc-git: delete ixp4xx configs Jan 18 22:12:09 Signed-off-by: Khem Raj Jan 18 22:29:18 khem, fun, that should fix almost all of my not buildings Jan 18 22:29:27 except p2020ds+minimal-uclibc :( Jan 18 22:55:15 Tartarus: p2020ds+minimal-uclibc whats issue there Jan 18 22:57:07 perl Jan 18 22:57:10 hmm its a e500 bigendial Jan 18 22:57:14 ah perl Jan 18 22:57:19 does 5.10.1 help ? Jan 18 22:57:22 nope Jan 18 22:57:30 then try 5.12 Jan 18 22:57:34 5.12.1 Jan 18 22:57:40 heh Jan 18 22:57:51 we dont have recipes for 5.12.x though Jan 18 22:58:19 is it a general issue with uclibc/ppc/be Jan 18 23:06:09 ok so slug can boot using uCslugC which is slugos image build with uclibc Jan 18 23:06:13 bingo Jan 18 23:09:39 ah, sorry, sec Jan 18 23:10:06 http://pastebin.com/7pngqXS4 Jan 18 23:10:09 5.10 Jan 18 23:15:10 hmmm uclibc does not have crypt_data actually Jan 18 23:15:14 yes Jan 18 23:15:18 And there's some mangle in there already Jan 18 23:15:19 so error is right Jan 18 23:15:20 That's not working Jan 18 23:15:26 ok Jan 18 23:15:28 I saw, got surprised and got distracted Jan 18 23:15:32 lemme look into perl Jan 18 23:15:35 but it's very reproducible here, yeah Jan 18 23:17:16 ifdef HAS_CRYPT Jan 18 23:17:26 how does it define that Jan 18 23:21:16 Tartarus: Does it happen only on this machine ? Jan 18 23:21:21 how about qemuppc ? Jan 18 23:24:43 only p2020ds i think Jan 18 23:25:12 aside from the whole gcc-cross-sdk + uclibc thing Jan 18 23:25:32 my only failures are p2020ds+minimal-uclibc (everything) and qemumimps64+minimal-uclibc (everything) Jan 18 23:25:42 and I see you pushed a build fix for qemumips64 Jan 18 23:26:02 mpc8315e-rdb and qemuppc work for minimal-uclibc+perl Jan 18 23:30:19 if it works for qemuppc it should have worked for p2020ds Jan 18 23:30:26 uclibc configs are same Jan 18 23:30:28 hmmm Jan 18 23:30:52 could be SPE has something to do here ? Jan 18 23:31:27 I do not know why the mangle isn't working when it really ought to have Jan 18 23:31:38 and keep getting distracted like right now when i need to go run an errand Jan 18 23:31:48 but it's a constantly showing up problem that 5.8.8 has too Jan 18 23:32:00 ok is the toolchain endinfg with -uclibcspe ? Jan 18 23:32:02 So if you have time it shouldn't be hard to reproduce Jan 18 23:32:08 yes Jan 18 23:32:13 ok Jan 18 23:32:17 i bet thats the reason Jan 18 23:32:21 something that should be using the libc-uclibc override? Jan 18 23:32:37 that override is still valid Jan 18 23:32:38 going, but i assume, yay, good quick debugging khem :) Jan 18 23:33:02 ok ttyl Jan 18 23:39:08 Tartarus: I am pushing a fix untested please try Jan 18 23:39:13 and report back Jan 18 23:39:34 03Khem Raj  07master * rccf1f88328 10openembedded.git/recipes/perl/ (perl_5.10.1.bb perl_5.8.8.bb): Jan 18 23:39:34 perl: Fix build for linux-uclibcspe targets Jan 18 23:39:34 Signed-off-by: Khem Raj Jan 18 23:45:52 03Khem Raj  07master * rf532fbe7d9 10openembedded.git/recipes/avahi/ (12 files): Jan 18 23:45:52 avahi: Fix build for uclibcspe targets and use INC_PR Jan 18 23:45:52 Signed-off-by: Khem Raj Jan 19 00:00:50 03Khem Raj  07master * r0781dd1886 10openembedded.git/recipes/appweb/appweb.inc: Jan 19 00:00:50 appweb.inc: Account for all different uclibc triplets Jan 19 00:00:50 Signed-off-by: Khem Raj Jan 19 00:10:41 khem, updating locally Jan 19 00:11:15 khem, fyi, the avahi.inc fix should probably be LDFLAGS_append_libc-uclibc Jan 19 00:12:21 and maybe we should change he if [ target_os = linux-uclibc ... to be a echo ${TARGET_OS} | grep -q linux-uclibc ] or so Jan 19 00:16:18 03Khem Raj  07master * rdb6d8bf17a 10openembedded.git/recipes/ (images/neuros-osd-base-image.bb tasks/task-openprotium.bb): Jan 19 00:16:18 neuros-osd-base-image.bb, task-openprotium.bb: Replace DISTRO_DEV_MANAGER with IMAGE_DEV_MANAGER Jan 19 00:16:18 Signed-off-by: Khem Raj Jan 19 00:16:25 03Khem Raj  07master * rc1de4ac61d 10openembedded.git/conf/distro/ucslugc.conf: Jan 19 00:16:25 ucslugc.conf: Define DISTRO_TYPE and IMAGE_DEV_MANAGER to be busybox-mdev Jan 19 00:16:25 Signed-off-by: Khem Raj Jan 19 00:18:20 khem, seems to have fixed perl tho, yay :) Jan 19 00:31:02 ok, doing a fixup for a buncha places that should use libc-uclibc Jan 19 00:31:07 since that'll fix 2 more build failures Jan 19 00:34:23 03Chris Larson  07master * re4193a20d7 10openembedded.git/recipes/netperf/ (files/vfork.patch netperf_2.4.4.bb): Jan 19 00:34:23 netperf: apply patch to fall back from fork to vfork Jan 19 00:34:23 Patch courtesy Freescale. Jan 19 00:34:23 Signed-off-by: Chris Larson Jan 19 00:34:30 03Chris Larson  07master * ra2539e0e25 10openembedded.git/recipes/ethtool/ (ethtool.inc ethtool/fifo.patch ethtool_2.6.35.bb): Jan 19 00:34:31 ethtool: apply patch to add 3 fifo TSEC params Jan 19 00:34:31 - fifo-tx-thr Jan 19 00:34:31 - fifo-tx-starve Jan 19 00:34:31 - fifo-tx-shutoff Jan 19 00:34:31 Patch courtesy Freescale. Jan 19 00:34:31 Signed-off-by: Chris Larson Jan 19 00:34:50 hrm Jan 19 00:43:01 So.. Jan 19 00:43:08 There's a few things that are actually harmfully bad, heh Jan 19 00:43:17 EXTRA_OECONF_append_linux-uclibc = " ${@base_contains('TARGET_ARCH', 'powerpc'... Jan 19 00:43:20 in gcc in a few places Jan 19 00:43:28 That'll not get evaluated when needed Jan 19 00:43:39 (for spe targets) Jan 19 00:43:47 fixing up to libc-uclibc will be a good thing there, heh Jan 19 00:50:59 heh Jan 19 01:07:17 And yeah, OK, CIA is gonna spam in a minute... Jan 19 01:07:21 34 commits Jan 19 01:08:15 sheesh Jan 19 01:09:35 I think 30 or so were broken on linux-uclibcspe in either trivial or non-trivial ways Jan 19 01:09:53 (which is to say some of these compiled but might have been otherwise broken and others were other build failures I saw) Jan 19 01:10:10 and 4 had the spe case already Jan 19 01:10:16 03Tom Rini  07org.openembedded.dev * r128143a5f8 10openembedded.git/recipes/ckermit/ckermit_211.bb: Jan 19 01:10:16 ckermit: Switch to using libc-uclibc Jan 19 01:10:16 Signed-off-by: Tom Rini Jan 19 01:10:23 03Tom Rini  07org.openembedded.dev * rcde17a27e2 10openembedded.git/recipes/diffutils/diffutils_2.8.1.bb: Jan 19 01:10:23 diffutils: Switch to libc-uclibc Jan 19 01:10:23 Signed-off-by: Tom Rini Jan 19 01:10:26 03Tom Rini  07org.openembedded.dev * rad838ec12c 10openembedded.git/recipes/gcc/gcc-4.4.2.inc: Jan 19 01:10:26 gcc-4.4.2: Switch to libc-uclibc Jan 19 01:10:26 Signed-off-by: Tom Rini Jan 19 01:10:27 03Tom Rini  07org.openembedded.dev * r400b890efc 10openembedded.git/recipes/eibd/eibd.inc: Jan 19 01:10:28 eibd: Switch to libc-uclibc Jan 19 01:10:28 Signed-off-by: Tom Rini Jan 19 01:10:30 03Tom Rini  07org.openembedded.dev * r32c61e9077 10openembedded.git/recipes/dt/dt_17.07.bb: Jan 19 01:10:30 dt: Switch to libc-uclibc Jan 19 01:10:30 Signed-off-by: Tom Rini Jan 19 01:10:31 03Tom Rini  07org.openembedded.dev * r22d7149bac 10openembedded.git/recipes/gcc/gcc-4.4.1.inc: Jan 19 01:10:32 gcc-4.4.1: Switch to libc-uclibc Jan 19 01:10:32 Signed-off-by: Tom Rini Jan 19 01:10:34 03Tom Rini  07org.openembedded.dev * r2ea67f98cc 10openembedded.git/recipes/gcc/gcc-4.4.4.inc: Jan 19 01:10:34 gcc-4.4.4: Switch to libc-uclibc Jan 19 01:10:34 Signed-off-by: Tom Rini Jan 19 01:10:37 03Tom Rini  07org.openembedded.dev * rca44b873f5 10openembedded.git/recipes/gcc/gcc-4.5.inc: Jan 19 01:10:37 gcc-svn: Switch to libc-uclibc Jan 19 01:10:37 Signed-off-by: Tom Rini Jan 19 01:10:39 03Tom Rini  07org.openembedded.dev * rbb4c670dfd 10openembedded.git/recipes/findutils/findutils.inc: Jan 19 01:10:39 findutils: Switch to libc-uclibc Jan 19 01:10:39 Signed-off-by: Tom Rini Jan 19 01:10:42 03Tom Rini  07org.openembedded.dev * r2a0753cc68 10openembedded.git/recipes/avahi/avahi.inc: Jan 19 01:10:43 avahi.inc: Switch to libc-uclibc Jan 19 01:10:43 As all current cases were caught, no PR change. Jan 19 01:10:43 Signed-off-by: Tom Rini Jan 19 01:10:44 03Tom Rini  07org.openembedded.dev * ra4a44637f9 10openembedded.git/recipes/gcc/gcc-svn.inc: Jan 19 01:10:45 gcc-svn.inc: Switch to libc-uclibc Jan 19 01:10:45 Signed-off-by: Tom Rini Jan 19 01:10:48 03Tom Rini  07org.openembedded.dev * r9f1729edd0 10openembedded.git/recipes/gdb/ (gdb-common.inc gdb.inc gdbserver.inc): Jan 19 01:10:49 gdb/gdbserver: Change RRECOMMENDS to use libc-{uclibc,glibc} Jan 19 01:10:49 This fixes certain cases like linux-uclibcspe Jan 19 01:10:49 Signed-off-by: Tom Rini Jan 19 01:10:50 03Tom Rini  07org.openembedded.dev * r49d88fab53 10openembedded.git/recipes/libaio/libaio-oracle_0.3.0.bb: Jan 19 01:10:50 libaio-oracle: Switch to libc-uclibc Jan 19 01:11:02 03Tom Rini  07org.openembedded.dev * r064a3d7464 10openembedded.git/recipes/pam/ (libpam_1.0.2.bb libpam_1.1.0.bb libpam_1.1.1.bb): Jan 19 01:11:02 pam: Switch to libc-uclibc Jan 19 01:11:02 Signed-off-by: Tom Rini Jan 19 01:11:06 03Tom Rini  07org.openembedded.dev * rc0e46e3f79 10openembedded.git/recipes/meta/slugos-native.bb: Jan 19 01:11:06 slugos-native: Switch to libc-uclibc Jan 19 01:11:06 Signed-off-by: Tom Rini Jan 19 01:11:08 03Tom Rini  07org.openembedded.dev * r22ca95a9cb 10openembedded.git/recipes/openjdk/ (openjdk-6-release-6b18.inc openjdk-6_6b18-1.8.bb): Jan 19 01:11:08 openjdk-6: Switch to libc-uclibc Jan 19 01:11:08 Signed-off-by: Tom Rini Jan 19 01:11:10 03Tom Rini  07org.openembedded.dev * ref963a0f16 10openembedded.git/recipes/meta/openprotium-packages.bb: Jan 19 01:11:10 openprotium-packages: Switch to libc-uclibc Jan 19 01:11:10 Signed-off-by: Tom Rini Jan 19 01:11:11 03Tom Rini  07org.openembedded.dev * r1bcffc772f 10openembedded.git/recipes/rhythmbox/rhythmbox_0.12.1.bb: Jan 19 01:11:12 rhythmbox: Switch to libc-uclibc Jan 19 01:11:12 Signed-off-by: Tom Rini Jan 19 01:11:13 03Tom Rini  07org.openembedded.dev * rb94cfdb521 10openembedded.git/recipes/ntpclient/ntpclient_2007_365.bb: Jan 19 01:11:13 ntpclient: Switch to libc-uclibc Jan 19 01:11:13 Signed-off-by: Tom Rini Jan 19 01:11:14 03Tom Rini  07org.openembedded.dev * rfeecc16175 10openembedded.git/recipes/samba/samba_3.0.34.bb: Jan 19 01:11:14 samba 3.0.34: Switch to libc-uclibc Jan 19 01:11:14 Signed-off-by: Tom Rini Jan 19 01:11:20 03Tom Rini  07org.openembedded.dev * redd3a1debd 10openembedded.git/recipes/tasks/task-sdk-bare.bb: Jan 19 01:11:20 task-sdk-bare: Switch to libc-uclibc / libc-glibc Jan 19 01:11:21 Signed-off-by: Tom Rini Jan 19 01:11:23 03Tom Rini  07org.openembedded.dev * rb12d6432a1 10openembedded.git/recipes/gcc/ (18 files): Jan 19 01:11:23 gcc-common.inc: Switch to libc-uclibc Jan 19 01:11:23 Bump INC_PR for everyone that I didn't just change. Jan 19 01:11:23 Signed-off-by: Tom Rini Jan 19 01:11:35 03Tom Rini  07org.openembedded.dev * r5e4a76196b 10openembedded.git/recipes/tasks/task-slugos-extd.bb: Jan 19 01:11:35 task-slugos-extd: Switch to libc-uclibc Jan 19 01:11:35 Signed-off-by: Tom Rini Jan 19 01:11:37 03Tom Rini  07org.openembedded.dev * ra7107a0201 10openembedded.git/recipes/sysstat/sysstat.inc: Jan 19 01:11:38 sysstat: Switch to libc-uclibc Jan 19 01:11:38 Don't change PR since it was already catching all cases. Jan 19 01:11:38 Signed-off-by: Tom Rini Jan 19 01:11:40 03Tom Rini  07org.openembedded.dev * r2a7ddea6f7 10openembedded.git/recipes/upslug/upslug2.inc: Jan 19 01:11:41 upslug2: Switch to libc-uclibc Jan 19 01:11:41 Signed-off-by: Tom Rini Jan 19 01:11:42 03Tom Rini  07org.openembedded.dev * r7190532c5a 10openembedded.git/recipes/vdr/vdr_1.7.16.bb: Jan 19 01:11:42 vdr: Switch to libc-uclibc Jan 19 01:11:42 Signed-off-by: Tom Rini Jan 19 01:11:43 03Tom Rini  07org.openembedded.dev * r1f64015313 10openembedded.git/recipes/gettext/ (gettext_0.14.1.bb gettext_0.17.bb): Jan 19 01:11:43 gettext: Switch to libc-uclibc Jan 19 01:11:43 Signed-off-by: Tom Rini Jan 19 01:11:45 03Tom Rini  07org.openembedded.dev * r192cb4fe24 10openembedded.git/recipes/powertop/powertop.inc: Jan 19 01:11:46 powertop: Switch to libc-uclibc Jan 19 01:11:46 Signed-off-by: Tom Rini Jan 19 01:11:47 03Tom Rini  07org.openembedded.dev * r0393bec46c 10openembedded.git/recipes/tcp-wrappers/tcp-wrappers_7.6.bb: Jan 19 01:11:47 tcp-wrappers: Switch to libc-uclibc Jan 19 01:11:47 Signed-off-by: Tom Rini Jan 19 01:59:08 can anyone tell me if ipkg/opkg will resume a download if it's interrupted? Jan 19 02:38:41 kgilmer, it wont Jan 19 02:44:01 ok, thanks grg :) Jan 19 02:47:41 that could probably be added though... Both libcurl and wget support such things. **** ENDING LOGGING AT Wed Jan 19 02:59:57 2011