**** BEGIN LOGGING AT Mon Jan 10 02:59:58 2011 Jan 10 05:08:57 hiii Jan 10 05:09:29 does any know how to convert .tar.gz to .jffs2 image Jan 10 05:11:17 sanket: with mkfs.jffs2 Jan 10 05:13:27 i had used that..... but the image is not working on my mini2440 board Jan 10 05:14:11 something is wrong with sector size and page size Jan 10 05:14:15 there's a bunch of options for it from memory, like for block size and things, maybe they need tweaking Jan 10 05:16:02 right now i m directly trying to port .tar.gz on mini2440... letc... if it works then it good enough Jan 10 05:16:03 I seem to remember that you have to erase the flash before you write the image as well if it wasn't already erased (but it's been a while and I'm not really an expert :P ) Jan 10 05:16:55 i m using online builder of http://www.angstrom-distribution.org/narcissus/ to build a sample image for mini2440 Jan 10 05:18:13 doesn't that give you a jffs2 image option? Jan 10 05:18:40 no... Jan 10 05:19:12 i had already build an .jffs2 image using oe.... Jan 10 05:19:41 but I m not getting X ... to run my application Jan 10 05:20:15 what about undewr "advanced" options, then "Select the type of image you want."? Jan 10 05:23:27 it provides an unstable image but then to i had tried that .. it take a lot of time. I had almost waited for 2 hrs... doesnt knowing that the process is going or not.... Jan 10 05:28:35 I am reading the cgit.oe Jan 10 05:28:45 What is the status of linux-omap3-2.6.37? Jan 10 06:16:21 03Khem Raj  07master * r5c6591a4e5 10openembedded.git/conf/distro/include/preferred-slugos-versions.inc: Jan 10 06:16:21 preferred-slugos-versions.inc: Lock newer versions of linux-libc-headers and dbus Jan 10 06:16:21 Signed-off-by: Khem Raj Jan 10 06:35:19 sanket: if you want to get jffs2 image add IMAGE_FSTYPES += "jffs2" to your local.conf Jan 10 06:35:33 and then rerun bitbake Jan 10 06:37:46 i had already build jffs2 image... but the problem is .... i m not able to run X application. I m only getting console.... Jan 10 06:39:00 when i run #X ... i m getting some driver are missing... Jan 10 08:11:01 03Klaus Kurzmann  07org.openembedded.dev * r2dc9b0098a 10openembedded.git/recipes/freesmartphone/cornucopia.inc: Jan 10 08:11:01 cornucopia.inc: bump SRCREV to get fixed fsodevice.conf for gta02 Jan 10 08:11:01 Signed-off-by: Klaus Kurzmann Jan 10 08:16:56 03Martin Jansa  07master * r9ef31e4aa6 10openembedded.git/recipes/efl1/azy_svn.bb: Jan 10 08:16:56 azy: add mysql5 to depends it's used in azy/src/tests/T_SQL.azy_server_stubs.h if mysql-config is found Jan 10 08:16:56 Acked-by: Khem Raj Jan 10 08:16:56 Signed-off-by: Martin Jansa Jan 10 08:20:46 03Martin Jansa  07master * r16a1f2e538 10openembedded.git/recipes/linux/ (5 files in 3 dirs): Jan 10 08:20:46 linux-2.6.37: add om-gta02 support Jan 10 08:20:46 Signed-off-by: Martin Jansa Jan 10 08:45:02 gm Jan 10 08:56:22 03Martin Jansa  07master * rdaaefcb00a 10openembedded.git/recipes/freesmartphone/ (fso-specs_git.bb libfso-glib_git.bb): Jan 10 08:56:23 fso-specs,lifso-glib: bump SRCREV Jan 10 08:56:23 Signed-off-by: Martin Jansa Jan 10 10:11:22 03Eric BENARD  07org.openembedded.dev * r24db17681a 10openembedded.git/recipes/busybox/ (3 files in 2 dirs): Jan 10 10:11:22 busybox-1.18.1: add latest fix Jan 10 10:11:22 modprobe-small: support compressed modules in insmodmaster Jan 10 10:11:22 http://git.busybox.net/busybox/commit/?id=e9d12b57bfb941e754d09a7d5f54c183ca7dbb25 Jan 10 10:11:22 Signed-off-by: Eric Bénard Jan 10 10:11:23 Acked-by: Khem Raj Jan 10 10:28:08 yo Jan 10 10:28:09 hello Jan 10 10:29:21 had anyone build a filesystem using angstrom online builder for mini2440 board Jan 10 10:59:48 sanket: some people have, I dont have access to that hardware though Jan 10 11:01:57 ya ok.... I got it.... I had successfully launched my X application.. Jan 10 11:08:23 sanket: cool Jan 10 11:09:31 but I m not getting touch screen access Jan 10 11:11:38 should be simple enough to debug, check device was probed, check /dev/input/touch* was created Jan 10 11:11:48 check X logs that it used the touchscreen driver Jan 10 13:05:04 good morning Jan 10 13:29:29 03Richard Purdie  07master * r644f130d7e 10bitbake.git/lib/bb/siggen.py: Jan 10 13:29:29 bitbake/siggen.py: Add a signature generator which adds the hash to the stamp files Jan 10 13:29:29 (From Poky rev: 97789a160ad21ccc614f976fa5e2b306d816aabf) Jan 10 13:29:29 Signed-off-by: Richard Purdie Jan 10 13:29:37 03Richard Purdie  07master * r61897c0f57 10bitbake.git/lib/bb/ui/knotty.py: Jan 10 13:29:37 bitbake/knotty.py: Don't show log messages of NOTE and below for task processes Jan 10 13:29:37 This cleans up the knotty console messages to be a lot quieter and cleaning, Jan 10 13:29:37 in keeping with the expectations of most users. Jan 10 13:29:37 (From Poky rev: b22e345e05efcc3f66278af8f09fb083afe32b68) Jan 10 13:29:37 Signed-off-by: Richard Purdie Jan 10 13:29:38 03Richard Purdie  07master * r8d39a9d204 10bitbake.git/lib/bb/fetch/__init__.py: Jan 10 13:29:38 fetch/__init__.py: Add missing debug level argument Jan 10 13:29:39 Signed-off-by: Richard Purdie Jan 10 13:29:40 03Richard Purdie  07master * r2d97dedfb9 10bitbake.git/lib/bb/msg.py: Jan 10 13:29:40 bitbake/msg.py: Add named definitions for all logging levels Jan 10 13:29:40 (From Poky rev: a13352cfaf4b5d19196151e362851b1516e6af15) Jan 10 13:29:41 Signed-off-by: Richard Purdie Jan 10 13:29:42 03Richard Purdie  07master * rcf4a309cce 10bitbake.git/lib/bb/__init__.py: Jan 10 13:29:43 bitbake/__init__.py: Add taskpid to all LogRecords (subclassed to be BBLogRecords) Jan 10 13:29:43 This allows us to identify which task messages are from. Jan 10 13:29:58 (From Poky rev: 66c6200ff34e8eeca5d1a689bbf9d6a83818248f) Jan 10 13:29:59 Signed-off-by: Richard Purdie Jan 10 13:29:59 03Richard Purdie  07master * r0b7bcf9bc5 10bitbake.git/lib/bb/runqueue.py: Jan 10 13:29:59 runqueue.py: Fix up some lost range->xrange conversions misplaced as part of the poky sync up Jan 10 13:29:59 Signed-off-by: Richard Purdie Jan 10 13:30:00 03Richard Purdie  07master * re2a39605e3 10bitbake.git/lib/bb/persist_data.py: Jan 10 13:30:00 persist_data: Don't loop forever waiting on database locks Jan 10 13:30:01 Signed-off-by: Richard Purdie Jan 10 13:30:01 03Richard Purdie  07master * ra1adbe7a3c 10bitbake.git/lib/bb/runqueue.py: Jan 10 13:30:02 bitbake/runqueue.py: Fix a bug where do_setscene dependencies would be ignored Jan 10 13:30:04 (From Poky rev: df8569b4d89ce83e3cafd87f2f37b795d1bfbd6d) Jan 10 13:30:04 Signed-off-by: Richard Purdie Jan 10 13:30:04 03Richard Purdie  07master * r43b1781b45 10bitbake.git/lib/bb/runqueue.py: Jan 10 13:30:04 runqueue.py: Add debug to help resolve partial log message failure Jan 10 13:30:05 Signed-off-by: Richard Purdie Jan 10 13:30:05 03Richard Purdie  07master * r12b3aeeafc 10bitbake.git/lib/bb/utils.py: Jan 10 13:30:06 utils.py: Update log message to use the logger calls Jan 10 13:30:06 Signed-off-by: Richard Purdie Jan 10 13:37:41 has anyone worked with friendlyarm mini6410 board?? Jan 10 14:25:49 03Koen Kooi  07org.openembedded.dev * r8d35525233 10openembedded.git/recipes/gstreamer/gst-tracelib_git.bb: Jan 10 14:25:49 gst-tracelib: add git version Jan 10 14:25:49 Signed-off-by: Koen Kooi Jan 10 14:46:48 looks like OE is still on hols :-D Jan 10 14:47:02 mostly Jan 10 14:47:13 except the wikispammers never take a holiday Jan 10 14:47:51 best cure for that would be send out a spam, then assasinate everyone that answers, then repeat Jan 10 14:47:58 eventually all stupid people would be dead Jan 10 14:48:42 after a night of reading BOFH archives...not such a bad idea. Jan 10 14:48:53 once stupid people are dead spam will not be so lucrative Jan 10 14:49:00 XorA: about 2011 plans...what is worst/better a) old kernel boots but udev issues arise or b) new kernel boots well but no suspend/resume/recharge ? Jan 10 14:49:17 ant_work: fix new kernel :-D Jan 10 14:50:12 you know what, I know who to bother (if you send him the cable) Jan 10 14:50:18 pavel Jan 10 14:50:28 * mwester thinks the idea could be applied to many other parts of life, starting with automobile drivers. Jan 10 14:51:26 the world gets to be a quiet place soon. Jan 10 14:53:46 anyway, some research showed the amount of spam decreased in 2010 Jan 10 14:54:05 almost unbelievable Jan 10 14:54:27 better isp filters? Jan 10 14:55:05 I can hardly imagine the number of spammers has decreased :/ Jan 10 14:55:15 I'd be happy to be contradicted Jan 10 14:57:59 ant_work, most SA's would be glad if we could go spammer plinking Jan 10 14:58:54 heh Jan 10 15:01:46 ant_work: if we consider spam as a business activity (for someone), crisis is affecting them too ;-) Jan 10 15:07:08 ant_work: if you get me an address to send it to I can Jan 10 15:12:09 ok, thx, mail sent Jan 10 15:13:12 * XorA is officially out of the zaurus game :-D Jan 10 15:14:08 is just an hobby now Jan 10 15:14:49 :-D Jan 10 15:15:08 ant_work: going to FOSDEM? Jan 10 15:15:20 don't know yet Jan 10 15:15:26 lot of things crossing :/ Jan 10 15:16:06 always the way, I hope to make it Jan 10 15:16:11 need to get organising Jan 10 15:16:16 * mwester wishes there were some embedded and/or linux events in the US midwest, from time-to-time... Jan 10 15:16:28 I'm in France the week before..let see Jan 10 15:16:30 mwester, organize one ... Jan 10 15:16:39 hehe! :) Jan 10 15:16:40 ant_work: short walk to brussels then Jan 10 15:16:54 of course I'm not sure how much embedded stuff is in the midwest Jan 10 15:17:17 RP, RP__, ping Jan 10 15:17:19 less and less anymore Jan 10 15:17:37 we're reverting to an agricultural economy, I think. Jan 10 15:17:57 So I shall be inventing 'OE for cows' Jan 10 15:17:59 :p Jan 10 15:18:48 blindvt`: pong Jan 10 15:19:28 RP__, lib/bb/build.py includes utils twice. Not sure which variant to delete? Jan 10 15:21:06 blindvt`: Either should work :) Jan 10 15:21:07 RP__, there was a PEP that (IIRC) said to use one import per line, and ISTR that one of 'import bb.utils' or 'from bb import utils' was preferred but i don't remember offhand which one Jan 10 15:21:46 blindvt`: I'm not sure which is my pythonic tbh Jan 10 15:22:12 mwester: bitbake cowdung :-D Jan 10 15:23:05 :D I prefer to bitbake steak, and then add a garbage collection phase to bitbake (to handle the dung part) Jan 10 15:23:44 RP__, something completely different. Sometime ago kergoth said that a kconfig interface to the data would be potential nice to have. I've wasted^Wthought about this during the weekend, FYA: uclibc.org/~aldot/bitbake/ the fwd-kconfig.00l2* stuff Jan 10 15:25:19 RP__, this uses pn (and not packages) ATM though and just emits depends as "select", which is not really nice with pristine kconfig at least Jan 10 15:26:40 blindvt`: Whats the purpose for doing this? Jan 10 15:26:59 while playing around, i stumbled across these two categories of bugs in OE, in case somebody wants to take care of them: Jan 10 15:27:01 git grep "[^$%]{[A-Z][A-Z]*}" | egrep -v "\.(diff|patch|rules|c):" Jan 10 15:27:01 grep ".#" recipes/libqglviewer/libqglviewer* Jan 10 15:28:08 RP__, not sure, TBH. Perhaps to post-process it for a list of packages, or to show the users which packages are available and what they do? Jan 10 15:29:17 blindvt`: Its certainly interesting code, I'm just struggling to work out how it would help people :) Jan 10 15:30:11 03Martin Jansa  07master * rf3b77424a8 10openembedded.git/ (3 files in 3 dirs): Jan 10 15:30:11 e-base: bump EFL_SRCREV Jan 10 15:30:11 Signed-off-by: Martin Jansa Jan 10 15:30:12 RP__, please tell me if you find out ;) But serious. Let me ask kergoth what he was aiming at Jan 10 15:31:15 03Martin Jansa  07master * r6a8e0ff179 10openembedded.git/recipes/linux/ (6 files in 4 dirs): Jan 10 15:31:15 linux-2.6.37: update openmoko and shr patch, add support for om-gta01 Jan 10 15:31:15 Signed-off-by: Martin Jansa Jan 10 15:31:23 03Martin Jansa  07master * r5057daf836 10openembedded.git/recipes/mesa/mesa-7.9.1.inc: Jan 10 15:31:23 mesa-7.9.1: change configure.ac to check for python not python2 as python-native does not stage python2 Jan 10 15:31:23 * should we fix it in python recipes instead? maybe more projects are looking for python2 and finds is on host not sysroot Jan 10 15:31:23 Signed-off-by: Martin Jansa Jan 10 15:31:24 03Martin Jansa  07master * r1143b63b95 10openembedded.git/recipes/mesa/ (6 files in 2 dirs): Jan 10 15:31:24 mesa: drop 7.8.2 version Jan 10 15:31:25 Signed-off-by: Martin Jansa Jan 10 15:31:26 03Martin Jansa  07master * r8beda76108 10openembedded.git/conf/distro/shr.conf: Jan 10 15:31:26 SHR: drop PREFERRED_PROVIDER_virtual/libusb0 and use default libusb-compat Jan 10 15:31:26 * libusb has legacy staging Jan 10 15:31:27 * libusb-compat is better maintained and newer Jan 10 15:31:27 Signed-off-by: Martin Jansa Jan 10 15:31:28 03Martin Jansa  07master * rdedd6ad322 10openembedded.git/conf/distro/include/angstrom-2010-preferred-versions.inc: Jan 10 15:31:29 angstrom-2010: prefer newer mesa 7.9.1 Jan 10 15:31:29 Signed-off-by: Martin Jansa Jan 10 15:31:30 Acked-by: Koen Kooi Jan 10 15:35:34 03Daniele Ricci  07master * re2517b6e6a 10openembedded.git/recipes/mokosuite2/mokophone_git.bb: Jan 10 15:35:35 mokophone: SRCREV bump Jan 10 15:35:35 Reject button now acts also as mute button if ringtone is playing Jan 10 15:35:35 Signed-off-by: Daniele Ricci Jan 10 15:35:35 Signed-off-by: Martin Jansa Jan 10 15:35:48 03Martin Jansa  07master * r3b2d2d4319 10openembedded.git/recipes/shr/libphone-ui-shr_git.bb: Jan 10 15:35:48 libphone-ui-shr: bump SRCREV Jan 10 15:35:48 Signed-off-by: Martin Jansa Jan 10 15:36:04 zecke: ping Jan 10 15:36:21 zecke: can you see my message in the private chat? Jan 10 15:38:44 rschus: just saw the hi. how comes you changed your nick? Jan 10 15:39:14 zecke: yeah, working from my laptop nowadays Jan 10 15:39:45 zecke: I dont have my office PC anymore since I am working from home :) Jan 10 15:48:45 Does x11-image fire up X on any of the qemus currently? Jan 10 16:00:03 It looks like only qemux86 atm Jan 10 16:11:37 03Martin Jansa  07master * r48d527cd24 10openembedded.git/recipes/tasks/task-shr-feed.bb: Jan 10 16:11:38 task-shr-feed: add emtooth2 Jan 10 16:11:38 Signed-off-by: Martin Jansa Jan 10 16:11:49 03Pau Espin Pedrol  07master * r72f22a0a9e 10openembedded.git/recipes/openmoko-3rdparty/emtooth2_svn.bb: Jan 10 16:11:49 emtooth2: add recipe for emtooth2 (new vala version of the app) Jan 10 16:11:49 Signed-off-by: Martin Jansa Jan 10 16:21:00 I haven't missed anyone fixing the libnl2 race have I? Jan 10 16:29:09 * Tartarus posts 2 patches Jan 10 16:30:32 And anyone else awake that might want to review Noor's patch for packaged-staging? Chris and I are happy with it, but want a non-mentor ack if possible :) Jan 10 16:51:10 concerning races, I met several problem (libxslt-native, libxml2) which are failing in do_configure and work once I relaunch bitbake (using bitbake master) Jan 10 16:51:27 That's odd, yeah Jan 10 16:52:11 Tartarus: your lastest 2 patches are fine please install them Jan 10 16:52:30 I have to try again with bitbake 1.10.x. Did anyone met something like this ? Jan 10 16:54:43 khem, I will sometime today, hoping to collect one or two more acks Jan 10 16:54:48 each time the error is the same : perl: error while loading shared libraries: libperl.so: cannot open shared object file: No such file or directory Jan 10 16:54:54 and : aclocal: /home/ebenard/OEUKREA/tmp_angstrom-2010.x_eukrea-cpuimx25/sysroots/i686-linux/usr/bin/autom4te failed with e Jan 10 16:54:58 xit status: 127 Jan 10 16:55:19 ericben: which distro? Jan 10 16:55:28 mckoan: angstrom 2010.x Jan 10 16:55:51 ericben: on host Jan 10 16:55:56 it seems it runs configure for libxml2-native or libxslt-native before perl-native was ready or something like this Jan 10 16:56:41 mckoan: ubuntu server 9.10 on x86 host Jan 10 16:57:07 ericben: and your problem is during bitbake? Jan 10 16:57:10 mckoan: it was working fine with older release of OE/bitbake on the same host Jan 10 16:57:41 mckoan: it's when building an image from scratch. With 4 BB threads. Jan 10 16:58:04 ericben: I had a similar problem some time ago and was related to the host distro Jan 10 16:58:11 http://www.kaeilos.com/debian-lenny-and-git-native Jan 10 16:58:22 I don't know if may help Jan 10 16:58:29 ;-) Jan 10 16:59:02 mckoan: thanks but I don't think this is the same problem. Here it seesm bitbake tries to configure these package to early, because relaunching bitbake finish the build. Jan 10 17:01:06 ericben: np Jan 10 17:01:52 ericben: actually I am using BB_NUMBER_THREADS = "2" for ages to be sure to get a bitbake successful end Jan 10 17:02:27 s/am/have been Jan 10 17:06:56 03Martin Jansa  07master * rb329bb2420 10openembedded.git/ (10 files in 3 dirs): (log message trimmed) Jan 10 17:06:57 xorg: upgrade to X11R7.6 and remove preferred-xorg-versions*.inc files Jan 10 17:06:57 * X11R7.6-RC1 was already released Jan 10 17:06:57 http://www.x.org/releases/X11R7.6-RC1/doc/xorg-docs/ReleaseNotes.html Jan 10 17:06:57 so it's IMHO time to start migration there Jan 10 17:06:57 * preferred-xorg-versions-live.inc was tracking latest stable released Jan 10 17:06:57 tarballs, but only reason I was maintaining it was to overwrite older Jan 10 17:09:27 http://pastebin.com/SzkJp1Li works but is that right? Jan 10 17:09:33 I think so from my quick googling Jan 10 17:10:32 (The problem is that route/pktloc.c which is not autogenerated needs route/pktloc_syntax.h) Jan 10 17:12:10 That looks about right to me. Jan 10 17:14:02 thanks Jan 10 17:15:10 03Tom Rini  07master * rd344300c57 10openembedded.git/conf/machine/include/qemu.inc: (log message trimmed) Jan 10 17:15:10 qemu.inc: Switch to xserver-xorg Jan 10 17:15:10 Keeping with xserver-kdrive means that we have a sometimes race Jan 10 17:15:10 in that xserver-kdrive is too old for the other xorg components that Jan 10 17:15:10 are built (with X11R7.5) to build. So we switch to xserver-xorg and Jan 10 17:15:10 appropriate modules. Tested on qemux86 and x11-image. Jan 10 17:15:11 Acked-by: Khem Raj Jan 10 17:15:14 03Tom Rini  07master * r9fd078b63b 10openembedded.git/conf/distro/minimal.conf: Jan 10 17:15:14 minimal.conf: Fully switch to bluez4 Jan 10 17:15:14 We still had bluez-libs being set for bluez-libs rather than bluez4 which Jan 10 17:15:14 meant that both were being built. Jan 10 17:15:14 Acked-by: Khem Raj Jan 10 17:15:14 Signed-off-by: Tom Rini Jan 10 17:15:16 03Tom Rini  07master * r8229cb1214 10openembedded.git/recipes/libnl/ (3 files in 2 dirs): Jan 10 17:15:16 libnl2: Fix a race on route/pktloc_syntax.h Jan 10 17:15:16 At issue is that route/pktloc.c (not generated) depends on Jan 10 18:02:14 03Richard Purdie  07master * r79f972e9af 10bitbake.git/lib/bb/fetch2/git.py: Jan 10 18:02:15 fetch2/git: Remove duplicated code only added for backwards compatibility Jan 10 18:02:15 Signed-off-by: Richard Purdie Jan 10 18:02:25 03Yu Ke  07master * r7c567d55f4 10bitbake.git/lib/bb/fetch2/ (13 files): Jan 10 18:02:26 bitbake: copy bb.fetch to bb.fetch2 as initial code base for fetcher overhaul Jan 10 18:02:26 Signed-off-by: Yu Ke Jan 10 18:02:26 Signed-off-by: Richard Purdie Jan 10 18:03:06 OK, http://wiki.openembedded.org/index.php/Testing#Test_combinations updated Jan 10 18:05:40 03Khem Raj  07master * r8308b77fe7 10openembedded.git/conf/distro/slugos.conf: Jan 10 18:05:40 slugos.conf: Use thumb mode to save some more space Jan 10 18:05:40 Signed-off-by: Khem Raj Jan 10 18:06:39 :) Progress for SlugOS... Jan 10 18:08:21 https://github.com/kergoth/openembedded/commit/d4b8f04 Jan 10 18:08:23 thoughts? Jan 10 18:14:09 oh.. seems zauruses should be switched to xorg too :\ Jan 10 18:15:32 they did not yet? Jan 10 18:16:06 Jay7, a lot of stuff is on kdrive still :( Jan 10 18:16:24 hrw: akita isn't iirc Jan 10 18:16:32 may be other clamshells as well Jan 10 18:17:00 khem, btw, lots of stuff isn't building for uclibc :( Jan 10 18:17:05 Or anyone else that's a uclibc person Jan 10 18:20:04 03Richard Purdie  07master * rde2518ab76 10bitbake.git/lib/bb/__init__.py: Jan 10 18:20:04 Switch to fetch2 fetcher code based on the environment variable BBFETCH2 Jan 10 18:20:04 (From Poky rev: af06e4a8efbcaff657c397c34f2cb893d2d7f7f7) Jan 10 18:20:04 Signed-off-by: Richard Purdie Jan 10 18:20:13 03Yu Ke  07master * r614d6fa073 10bitbake.git/lib/bb/fetch2/ (13 files): Jan 10 18:20:13 bb.fetch2: replace bb.fetch with bb.fetch2 in the bb.fetch Jan 10 18:20:13 bb.fetch2 is copied from bb.fetch, and has many bb.fetch referrence. Jan 10 18:20:13 Fix these referrence with bb.fetch2 referrence Jan 10 18:20:13 (From Poky rev: c64b445f38179a117d39319f762f9dc2435da0ea) Jan 10 18:20:14 Signed-off-by: Yu Ke Jan 10 18:20:14 Signed-off-by: Richard Purdie Jan 10 18:46:15 mwester: So now nslu is booted of nfs but nfsroot does not have turnup installed so how do I switch back to say usb again Jan 10 18:46:29 mwester: I see that usb is mounted in /mnt Jan 10 18:47:00 nfsroot should have /sbin/turnup -- IIRC "turnup flash" should do it Jan 10 18:48:01 mwester: I have booted a console-image Jan 10 18:48:06 Then reboot, and you can then issue the turnup command to turnup to a flash device: "turnup memstick -i /dev/sda1" Jan 10 18:48:11 mwester: so that probably does not have it Jan 10 18:48:23 Ah, yeah - console-image != slugos Jan 10 18:48:28 :) Jan 10 18:48:43 but I can build a slug image Jan 10 18:48:51 is it nslu2-base-image ? Jan 10 18:48:57 slugos-image Jan 10 18:49:03 hmm Jan 10 18:49:49 slugos-image.bb is still nslu2-specific, but I'm trying to abstract it so that I can build images for MACHINE=qemuarm and MACHINE=sheevaplug Jan 10 18:50:37 mwester: k Jan 10 18:50:42 It is very unfortunate that the versatilepb machine has no on-board flash. :( I'd much rather test with an emulated device that would let me test turnup and related tools, but that requres flash. Jan 10 18:50:47 yeah it would be nice to make it processor agnostic Jan 10 18:51:12 oh flash Jan 10 18:51:37 but you should be able to test memstick Jan 10 18:52:13 Yep, but parts of the initial boot depend on jffs2 filesystems, which we can't support on qemuarm. Jan 10 18:52:55 i.e. sysconf and related mechanism to remember config information across reflashing of the device. Jan 10 18:54:53 I can invest some effort to make turnup and sysconf more generic, perhaps based on a config file, so that I can simulate the various flash partitions and filesystems using (for example) a partitioned virtual disk on qemuarm. That might also help if I wanted to try something a bit more modern for the flash filesystem (jffs2 is a bit dated) Jan 10 18:55:05 I think qemu just deals with block devices prolly not with fs Jan 10 18:55:43 Yes, ordinary block devices. Jan 10 18:55:55 And block devices cannot support jffs2 Jan 10 18:56:20 mwester: hmm yes Jan 10 18:56:23 so there is a catch Jan 10 18:56:29 always. Jan 10 18:56:56 but qemu can do a lot of work otherwise Jan 10 18:57:32 so far I have been holding slugos by neck and trying to make it compile console-image and I succeeded too :) Jan 10 18:57:55 yes -- in particular, I think I can add automated testing of the image to the autobuilder, at least I expect that I can run through the standard "new user experience" per the nslu2 wiki in a completely automated fashion. Jan 10 18:58:34 hmm, i should push these bitbake vim syntax fixups to the bitbake repo Jan 10 18:58:35 :D SlugOS needs to be grasped by the scruff of its neck, and spoken to firmly! It's a bit dated... but coming along nicely now. :) Jan 10 18:58:55 must run off to a meeting now - back later. Jan 10 19:00:35 03Chris Larson  07master * r0153ace246 10bitbake.git/lib/bb/ (__init__.py event.py): Jan 10 19:00:35 Inject taskpid into log records via our log handler Jan 10 19:00:35 It turns out that while log filters added with addFilter are only associated Jan 10 19:00:35 with that logger, and not its children, handlers are inherited, and handlers Jan 10 19:00:35 can be filters. So, let's add filtering to our existing LogHandler class Jan 10 19:00:36 which dispatches our log records as bitbake events. Jan 10 19:00:37 Signed-off-by: Chris Larson Jan 10 19:00:46 03Chris Larson  07master * rf49da31206 10bitbake.git/bin/bitbake: Jan 10 19:00:46 bitbake: handle IOError from event_queue.get Jan 10 19:00:46 These can occur in a particular ^C shutdown, so just break out of the event Jan 10 19:00:46 queue empty loop if we encounter it. Jan 10 19:00:46 Signed-off-by: Chris Larson Jan 10 19:05:43 03Chris Larson  07master * r4210eb0b78 10bitbake.git/contrib/vim/syntax/bitbake.vim: Jan 10 19:05:43 vim: properly highlight python inside of ${@} Jan 10 19:05:43 Signed-off-by: Chris Larson Jan 10 19:05:44 03Chris Larson  07master * rf36354a1bf 10bitbake.git/contrib/vim/syntax/bitbake.vim: Jan 10 19:05:45 vim: handle highlighting exports without assignment Jan 10 19:05:45 Signed-off-by: Chris Larson Jan 10 19:05:52 03Chris Larson  07master * r29ce70ac85 10bitbake.git/contrib/vim/ftplugin/bitbake.vim: Jan 10 19:05:52 vim: add an ftplugin for indentation settings Jan 10 19:05:52 Signed-off-by: Chris Larson Jan 10 19:09:14 03Chris Larson  07master * r1676a2c880 10openembedded.git/conf/bitbake.conf: Jan 10 19:09:14 bitbake.conf: fix quote usage in COMBINED_FEATURES Jan 10 19:09:14 Signed-off-by: Chris Larson Jan 10 19:18:25 * kergoth` thinks itd be useful to add a warning from package_write/package_write_ipk when the only packages emitted were -dev/-dbg Jan 10 19:18:39 re Jan 10 19:18:47 kergoth`: indeed Jan 10 19:19:59 better yet, warn if the only packages emitted were empty, or packages with ALLOW_EMPTY=1, rather than an explicit list Jan 10 19:20:06 heh Jan 10 19:27:10 03Chris Larson  07master * r89fb37db1d 10openembedded.git/ (3 files in 3 dirs): Jan 10 19:27:10 procinfo-ng: new recipe, version 2.0.304 Jan 10 19:27:10 Signed-off-by: Chris Larson Jan 10 19:27:47 what was the -ng naming again for ? Jan 10 19:28:00 hi kergoth` I suspect bitbake master brings several problem which were not present around 1 or 2 weeks ago : from a clean tmp, once libxslt-native & once libxml2-native failed with libperl.so not found (and relaunching bitbake "solves" the problem so it seems bitbake tries to configure these libs before a depend is finished to being run) all this with 4 BB_THREADS. Then perl failed with ./miniperl command not found (this seems to be the sa Jan 10 19:29:32 if you have a race, changing shit is going to affect it. that doesn't imply a causal relationship Jan 10 19:29:52 try a bisect if you suspect a commit in bitbake Jan 10 19:30:17 rob_w: it's the one that's actually been touched more recently than 2001. Jan 10 19:30:20 :P Jan 10 19:30:25 and supports kernels newer than like 2.2 Jan 10 19:30:57 didnt the ng stand for some words ? Jan 10 19:31:14 oh, its commonly used for rewrites, refers to next generation as far as i know Jan 10 19:31:24 ah ok Jan 10 19:31:34 thats easy to remeber thx Jan 10 19:31:58 kergoth`: ok I didn't have these race before but I'll check if perl-native is missing from some dependencies. Jan 10 19:32:53 the fact that you never saw it before doesn't necessarily mean it wasn't a race. Jan 10 19:33:01 it j ust means it happened to work for you, that's all Jan 10 19:35:53 03Chris Larson  07master * r8385bfb7da 10bitbake.git/lib/bb/msg.py: Jan 10 19:35:53 msg: no need to hardcode the logging levels Jan 10 19:35:53 Signed-off-by: Chris Larson Jan 10 19:37:15 would it be possilbke and wished to create those cia message as a link ? Jan 10 19:37:49 cia has built in support for links to scm browser interfaces, it just needs to be passed / configured in the hook Jan 10 19:38:44 for a newbee like that would be cool to just follow the link if a checking sounds interesting Jan 10 19:38:52 like me ^ Jan 10 19:39:13 yeah, it's a good idea Jan 10 19:49:07 gm Jan 10 19:54:16 hi likewise Jan 10 20:14:03 03Chris Larson  07master * re0728e4283 10openembedded.git/recipes/cramfs/cramfs_cvs.bb: Jan 10 20:14:03 cramfs: add cvs version Jan 10 20:14:03 Signed-off-by: Chris Larson Jan 10 20:21:26 kergoth`: seems perl-native is missing from autoconf.inc's DEPENDS_virtclass-native & RDEPENDS_virtclass-native Jan 10 20:21:46 which could explain the race met when configuring Jan 10 20:22:07 perl-native is a nightmare. talk to Tartarus Jan 10 20:23:09 perl is listed as DEPENDS & RDEPENDS so perl-native seems necessary to virtclass-native Jan 10 20:23:51 Yeah, no Jan 10 20:23:57 We don't use perl-native for autoconf/make Jan 10 20:24:30 Tartarus: perl-native-runtime is listed as RDEPENDS_automake(native Jan 10 20:24:34 ericben, I would however encourage you to give moving OE towards using perl-native for everything a shot Jan 10 20:24:44 It was however when I dug into it about a year ago, a nightmare Jan 10 20:24:55 ericben, yes, perl-native-runtime is then in ASSUME_PROVIDED in bitbake.conf Jan 10 20:25:12 It's a "We don't want to open this can of worms so..." kind of thing Jan 10 20:25:32 And there's a fair number of problems that a build can hit once we do have perl-native built Jan 10 20:25:34 Tartarus: ok, maybe you have an idea then why I get "libperl.so not found" when bitbake tries to configure some libs (to early) Jan 10 20:25:38 So I'm not against fixing up the hell we have, really Jan 10 20:25:40 It's just a mess Jan 10 20:25:51 ericben, has perl-native been built? Jan 10 20:26:32 Tartarus: I think not yet, that's what I was suspecting a problem in bitbake but kergoth` is right this is more likely a race in oe which is exhibited by recent bitbake on my build system Jan 10 20:26:51 Tartarus: moreover launching bitbake a second time "fixes" the problem Jan 10 20:27:07 ericben, what distro again? Jan 10 20:27:24 Actually, hmm, wth Jan 10 20:27:40 Tartarus: on the host : ubuntu server karmic Jan 10 20:27:41 When it does fail can you pastebin the config.log about whats going on? Jan 10 20:29:02 Tartarus: http://pastebin.com/ky43sr77 Jan 10 20:30:47 And your machine is otherwise known sane? Jan 10 20:30:59 ie memtest is fine, fsck doesn't say "woah, your screwed.." Jan 10 20:31:00 etc Jan 10 20:31:30 (and if you run head on /home/ebenard/OEUKREA/tmp_angstrom-2010.x_eukrea-cpuimx25/sysroots/i686-linux/usr/bin/autom4te it's trying to use /usr/bin/perl right?) Jan 10 20:31:59 Tartarus: #! /usr/bin/env perl Jan 10 20:32:19 ok Jan 10 20:32:23 also should be fine, heh Jan 10 20:32:29 since there's no perl-native you said Jan 10 20:32:41 Tartarus: the machine was sane until this problem ;) Jan 10 20:32:59 Tartarus: perl-native's build was is process when that failed Jan 10 20:33:11 in progress sorry Jan 10 20:36:07 kergoth`: concerning my problem building perl on my machine which is setup in french. You fixed it here : http://cgit.openembedded.net/cgit.cgi/bitbake/commit/?id=bebdb554b5affc960b462bf2c59657fb78357f50 and I think it's broken again here : http://cgit.openembedded.net/cgit.cgi/bitbake/commit/lib/bb/build.py?id=315b83d1de954e51d0d077ff00140a7ec41ffbc7 Jan 10 20:36:56 perl has a hack in the makefile (if I remember well) which doesn't support french output of the tools Jan 10 20:38:42 ah, right, that. Jan 10 20:38:55 OE should be setting export LC_ALL = "C" in bitbake.conf or base.bbclass Jan 10 20:39:42 how far in progress? Jan 10 20:39:46 was perl-native Jan 10 20:41:51 Tartarus: once my actual build finish I'll launch it again and tell you. It seems I can't reproduce it on a second build host. Jan 10 20:42:28 kergoth`: ok thanks Jan 10 20:47:29 hello, help me pls, I am trying to build somthing in oe, but can't Jan 10 20:47:40 I have a lot of errors like this: ERROR: EOL while scanning string literal (, line 1) while parsing /stuff/openembedded/recipes/nxproxy/nxproxy_3.3.0-2.bb NOTE: :EOL while scanning string literal (, line 1) while evaluating: ${@oe.utils.ifelse(bool(d.getVar('MACHINE', True)), '${MACHINE}', 'BASE_PACKAGE_ARCH')} ERROR: Error in executing: /stuff/openembedded/recipes/live555/live555_20101015.bb Jan 10 20:47:55 wrong MACHINE? Jan 10 20:48:22 config: cat build/conf/local.conf DL_DIR = "${HOME}/sources" BBFILES = "/stuff/openembedded/recipes/*/*.bb" TARGET_ARCH = "mipsel" TARGET_OS = "linux-uclibc" INHERIT = "package_tar" #DISTRO = "minimal-uclibc" Jan 10 20:48:44 cat build/conf/local.conf | wgetpaste Your paste can be seen here: http://paste.pocoo.org/show/318527/ Jan 10 20:49:35 Sunsun: why you don't set MACHINE there? Jan 10 20:49:58 there is not config for my machine, so I choose architecture, os.... Jan 10 20:50:52 you can't do this way Jan 10 20:51:07 you should write machine config then Jan 10 20:51:15 look into conf/machine/* Jan 10 20:51:45 so, at first I should create config, and than use it? ok, I will try, thx Jan 10 20:52:11 yes, create machine config then add MACHINE= Jan 10 20:52:17 w/o .conf Jan 10 20:52:37 look into other machine files Jan 10 20:52:52 ok Jan 10 20:54:48 * Jay7 is thinking about freebsd OE port again.. Jan 10 21:07:05 Jay7 из корбины? :) спасибо, создал конфиг, сборка пошла Jan 10 21:07:29 Sunsun: this is english-speaking channel ;) Jan 10 21:07:40 ok) Jan 10 21:07:52 are you understanding me? Jan 10 21:07:54 Sunsun: anyway, what machine you have? Jan 10 21:07:59 yes, I'm russian :) Jan 10 21:08:05 iconbit hds41l Jan 10 21:08:49 can you send config to oe devel mailing list to share your work? :) Jan 10 21:09:18 maybe Jan 10 21:09:21 better as git-format-patch output or even with git-send-email Jan 10 21:10:19 Sunsun: btw, there is russian OE list.. almost dead but.. Jan 10 21:10:39 Sunsun: http://groups.google.com/group/openembedded-ru Jan 10 21:12:59 thx Jan 10 21:20:51 i changed DISTRO_FEED_URI and rebuild distro-feed-configs afterwards, but configs contain still the old URI, what else should i do to change the feed? Jan 10 22:12:03 03Richard Purdie  07master * r7c0f59cd41 10bitbake.git/lib/bb/runqueue.py: Jan 10 22:12:03 bitbake/runqueue.py: Fix del_stamp calls and -f option to bitbake with the BasicHash siggen code Jan 10 22:12:03 (From Poky rev: 48e567bb477ad468d) Jan 10 22:12:03 Signed-off-by: Richard Purdie Jan 10 22:38:43 03Richard Purdie  07master * r84ea542671 10bitbake.git/lib/bb/runqueue.py: Jan 10 22:38:44 runqueue.py: Call bb.cache.Cache.loadDataFull function directly when dumping signatures as per the rest of runqueue Jan 10 22:38:44 Signed-off-by: Richard Purdie **** ENDING LOGGING AT Tue Jan 11 02:59:58 2011