**** BEGIN LOGGING AT Thu Apr 23 02:59:57 2009 Apr 23 03:02:01 khem: ok, it seems your build is not pulling libc-headers Apr 23 03:03:20 well they are not being put into the correct directory. They must have been pulled. There has to be a staging directory populated. It seems like this target is wanting a different staging directory. Apr 23 03:04:42 A couple of hours ago I looked at a gcc command line from a log, and it was referencing an empty directory. Apr 23 03:05:13 ending in "include" Apr 23 03:08:08 dickelbeck: /diskb/oestuff/build/tmp/staging/armv4t-oe-linux-gnueabi/usr/include does this dir exist Apr 23 03:15:19 yes, and it is empty Apr 23 03:16:16 kergoth: I am getting this error lately http://pastey.net/112898 Apr 23 03:16:40 from staging on down, if I: find . -name include, I get: Apr 23 03:16:43 ./arm-linux/usr/include Apr 23 03:16:43 ./ks8695-linux-gnueabi/kernel/include Apr 23 03:16:43 ./i686-linux-sdk/usr/include Apr 23 03:16:43 ./armv4t-linux-gnueabi/usr/include Apr 23 03:16:43 ./arm-oe-linux-gnueabi/usr/include Apr 23 03:16:44 ./arm-linux-gnueabi/usr/lib/glib-2.0/include Apr 23 03:16:46 ./arm-linux-gnueabi/usr/include Apr 23 03:16:48 ./i686-linux/usr/include Apr 23 03:17:02 khem: no idea, I'm afraid. never messed with ipkg.py Apr 23 03:17:17 dickelbeck: hmm as long as the dir is there it should be ok Apr 23 03:17:47 kergoth: ok it seems I have to reluctantly look into it :) Apr 23 03:17:52 hehe, have fun Apr 23 03:18:35 khem: arm-linux-gnueabi/usr/include is not empty Apr 23 03:19:56 and this is suspiciously close to the directory that is empty Apr 23 03:20:52 arm-linux-gnueabi vs armv4t-linux-gnueabi Apr 23 03:21:34 dickelbeck: I think this is cross canadian problem Apr 23 03:22:51 What happened to Tartarus? Is he responsible for this? Anything I am doing wrong? Apr 23 03:24:47 I have this line in my local.conf: Apr 23 03:24:50 PACKAGE_EXTRA_ARCHS_ks8695 = "armv4t" Apr 23 03:26:24 dickelbeck: thats ok. Can you paste your configure command somewhere for gcc gcc-cross-sdk Apr 23 03:26:38 gettext fail http://tinderbox.openembedded.net/public/logs/4609653.txt "no rule to make target" 8| Apr 23 03:28:15 dickelbeck: should be something run.do_configure.xxxxxx Apr 23 03:31:00 maybe i should clean my build dir and start over Apr 23 03:34:15 khem: http://pastebin.com/m79f55833 Apr 23 03:56:29 dickelbeck: The sysroot points to armv4t-oe-linux-gnueabi Apr 23 03:56:41 but linux-headers are installed elsewhere Apr 23 04:00:37 glibc headers also right? Apr 23 04:02:17 yeah Apr 23 04:02:50 dickelbeck: I will try to reproduce it here and if I do get it. I will let you know Apr 23 04:08:52 thank you. I will be stepping away for awhile then. Apr 23 06:38:16 03Koen Kooi  07org.openembedded.dev * re06cd7c84b 10openembedded.git/recipes/iputils/ (10 files in 3 dirs): iputils: add 20071127 Apr 23 06:38:16 03Koen Kooi  07org.openembedded.dev * r35d0a5ed20 10openembedded.git/conf/checksums.ini: checksums: add more checksums Apr 23 06:38:17 03Khem Raj  07org.openembedded.dev * r971539b2bb 10openembedded.git/ (13 files in 3 dirs): Apr 23 06:38:17 binutils-2.19.51.0.3: Add the latest HJ Lu's binutils release 2.19.51.0.3 recipes. Apr 23 06:38:19 * With these binutils we can build uclibc in thumb mode correctly. Apr 23 06:38:21 Signed-off-by: Khem Raj Apr 23 06:38:23 Signed-off-by: Koen Kooi Apr 23 06:38:28 03Koen Kooi  07org.openembedded.dev * r8f6996cef5 10openembedded.git/contrib/ (angstrom/upload-packages.sh patchwork/pw-am.sh): patchword apply helper: use git-am -s and remove file when done Apr 23 06:43:04 good morning Apr 23 06:50:32 good morning Apr 23 06:53:48 03Dmitry Baryshkov  07org.openembedded.dev * r6685e5012a 10openembedded.git/conf/machine/include/tune-ppce500.inc: Apr 23 06:53:48 ppce500: add ppce500 to PACKAGE_EXTRA_ARCHS Apr 23 06:53:48 ppce500 was broken because generated opkg.conf didn't contain a reference Apr 23 06:53:48 to the core feeds. Fix it by providing PACKAGE_EXTRA_ARCHS as all other Apr 23 06:53:48 PPC tune files do. Apr 23 06:53:49 Signed-off-by: Dmitry Baryshkov Apr 23 06:53:58 03Dmitry Baryshkov  07org.openembedded.dev * r18e058e835 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Apr 23 09:13:13 03Steffen Sledz  07org.openembedded.dev * r4a9cec9c1a 10openembedded.git/recipes/linux/linux-2.6.24/hipox/defconfig: linux-2.6.24: enable kexec in defconfig for hipox machine Apr 23 09:56:58 morning Apr 23 10:05:35 Morning, thanks for help with my gumstix building.. Apr 23 10:16:30 03Koen Kooi  07org.openembedded.dev * r520b3470cb 10openembedded.git/ (conf/checksums.ini recipes/bluez/bluez4_4.37.bb): bluez4: add 4.37 Apr 23 10:16:33 03Koen Kooi  07org.openembedded.dev * r2747c78c7c 10openembedded.git/ (15 files in 3 dirs): linux-omap 2.6.29: add more dss2 patches, should solve problems when using u-boot splash Apr 23 10:16:43 03Koen Kooi  07org.openembedded.dev * r4c312174e1 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Apr 23 10:59:12 03Michael 'Mickey' Lauer  07org.openembedded.dev * rb72810f7fc 10openembedded.git/ (2 files in 2 dirs): openezx-kernel: bump to 2.6.30-rc2; this fixes the touchscreen issues we had for a while Apr 23 11:52:13 I want to scream and shoot in direction of few debian develoeprs Apr 23 12:00:05 03Koen Kooi  07org.openembedded.dev * r6d446348ff 10openembedded.git/recipes/powertop/powertop_1.11.bb: powertop: fix ldflags problem with omap3 machines Apr 23 12:00:06 03Koen Kooi  07org.openembedded.dev * r3f7b4e51b0 10openembedded.git/recipes/linux/ (100 files in 4 dirs): linux-omap-pm git: update to latest git, refresh DSS2 patches Apr 23 12:00:16 03Koen Kooi  07org.openembedded.dev * r3ec8b29b0f 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Apr 23 12:37:28 where can i get the latest u-boot sources? Apr 23 12:37:33 for beagle Apr 23 12:39:06 MACHINE=beagleboard bitbake u-boot? Apr 23 12:49:29 03Koen Kooi  07org.openembedded.dev * r1496aea759 10openembedded.git/recipes/efl1/ (ecore.inc ecore_svn.bb): Apr 23 12:49:29 ecore: fix packaging after e17 soname change upstream Apr 23 12:49:29 * this makes e work again after 1e71714dd3e693c64ac0d8c1f863e2efc326f30c broke it Apr 23 12:49:56 03Koen Kooi  07org.openembedded.dev * r8df21d4d8c 10openembedded.git/recipes/ffmpeg/omapfbplay_git.bb: omapfbplay: update for 2.6.29 header location change Apr 23 12:52:32 03Koen Kooi  07org.openembedded.dev * ra292887f0a 10openembedded.git/conf/distro/angstrom-2008.1.conf: Apr 23 12:52:32 angstrom 2009.X: move armv7a and vortex to gcc 4.3.3 Apr 23 12:52:32 * runtime tested on beagleboard and overo Apr 23 13:34:30 hello! Apr 23 13:34:58 http://git.openembedded.org/openembedded that page looks very dead to me Apr 23 13:35:39 cgit.openembedded.net Apr 23 13:35:58 nice the wiki is up-to-date ... *not* Apr 23 13:37:53 i'm sorry to say it, but oe looks very messy Apr 23 13:41:29 were are just humans ;( Apr 23 13:42:15 i know, but there are better organized project :P Apr 23 13:42:22 hey hrw Apr 23 13:42:30 or project that look better organized :P Apr 23 13:42:35 *projects Apr 23 13:42:36 can I please have your buildbot cfg and such? :) Apr 23 13:42:48 Tartarus: oe.dev/contrib/buildbot/ Apr 23 13:42:56 bah, i thought i looked :( Apr 23 13:43:16 recent push? Apr 23 13:44:06 yesterday Apr 23 13:44:13 that'd do it :) Apr 23 13:44:28 box i'm doing stuff on is on last friday still atm Apr 23 13:44:29 thanks Apr 23 13:45:14 Steffanx: http://git.openembedded.org/openembedded - where you got that link? Apr 23 13:45:38 http://wiki.openembedded.net/index.php/Getting_Started#Obtaining_OpenEmbedded_using_GIT Apr 23 13:46:25 Steffanx: there is no such link Apr 23 13:46:43 there is git://git.openembedded.net/openembedded Apr 23 13:46:49 but thats other protocol Apr 23 13:47:07 other protocol? Apr 23 13:47:17 ah ok Apr 23 13:47:42 nevermind, but it failed to 'fetch' that repro .. Apr 23 13:47:49 hrw, ok, reading that now, thanks alot :) Apr 23 13:47:57 Gonna get some frequent builds of meta-toolchain stuff going Apr 23 13:48:20 Steffanx: added link to web interface there just in case Apr 23 13:48:21 or it is my mistake hrw Apr 23 13:48:23 Initialized empty Git repository in /home/steffan/openembedded/.git/ Apr 23 13:48:23 fatal: The remote end hung up unexpectedly Apr 23 13:48:33 Steffanx: our machine is overloaded again Apr 23 13:48:49 the "firewall-version" works Apr 23 13:48:51 better Apr 23 13:49:10 anyone working on zoneminder recipe for OE? Apr 23 13:49:15 * hrw -> food Apr 23 13:49:37 i've seen http://tinderbox.openembedded.net/packages/zoneminder-image/ but nothing on git Apr 23 13:52:44 ~seen rschuster Apr 23 13:52:48 rschuster was last seen on IRC in channel #oe, 25d 21h 48m 30s ago, saying: 'you need to apply a trick with EXTRA_OEMAKE (the value for normal make) to get it work'. Apr 23 14:01:10 03Koen Kooi  07org.openembedded.dev * r7209ba7dd7 10openembedded.git/ (5 files in 2 dirs): Apr 23 14:01:10 ekiga: update to 3.2.0 Apr 23 14:01:10 opal: update to 3.6.1, fix QA Apr 23 14:01:10 ptlib: update to 2.6.1 Apr 23 14:02:06 03Roman I Khimov  07org.openembedded.dev * r2343808f73 10openembedded.git/ (classes/image.bbclass conf/bitbake.conf): Apr 23 14:02:06 bitbake.conf: use rootfs/${PN} for IMAGE_ROOTFS Apr 23 14:02:06 Allows parallel build for different images. Apr 23 14:02:06 Acked-by: Otavio Salvador Apr 23 14:02:07 03Roman I Khimov  07org.openembedded.dev * r1f1566b8c3 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Apr 23 14:28:34 Bah Apr 23 14:28:42 Trying to use OE with a real minimal dev table sucks :) Apr 23 14:29:20 psplash doesn't check for / make fb0 Apr 23 14:29:34 banner doesn't check for $vtmaster Apr 23 15:26:22 re Apr 23 15:26:39 wb# Apr 23 15:27:42 Tartarus: psplash needs to learn that some devices have more then one fb ;) Apr 23 15:27:51 Yeah, that too Apr 23 15:27:58 BUG has 2 framebuffers and psplash should use fb1 AND fb2 Apr 23 15:28:05 or none of them Apr 23 15:28:11 But OE just isnt' happy with a minimal static /dev Apr 23 15:28:19 Just gave up and put fb0 and tty[0-9] in Apr 23 15:28:24 maybe its too minimal? Apr 23 15:28:31 Well Apr 23 15:28:39 Last I knew, null and console were all you must have Apr 23 15:28:46 Tartarus: busybox-mdev is too big for that rootfs? Apr 23 15:29:04 hrw, the problem is stuff in S[0123] Apr 23 15:29:08 since udev is S04 Apr 23 15:29:14 mdev too I would assume Apr 23 15:29:26 right Apr 23 15:29:46 So psplash was giving me fits on needing fb0 and tty0 Apr 23 15:29:54 banner was trying to use tty0 as well Apr 23 15:30:05 and not checking if it existed and trying to make it Apr 23 15:30:07 banner can be fixed to work without tty Apr 23 15:30:12 Just gaveup and went the SlugOS way for banner Apr 23 15:31:23 which is? Apr 23 15:32:16 Not try and be smart Apr 23 15:32:21 echo "Booting" works just fine Apr 23 15:32:25 and ends up on console Apr 23 15:32:39 maybe we should do that in normal Apr 23 15:33:04 I'd +1 that :) Apr 23 15:33:11 rfc that Apr 23 15:33:35 morning Apr 23 15:47:47 morning Apr 23 16:09:41 Crofton, hi I hace oe commit access but I'm not sure how to access patchwork Apr 23 16:11:00 nytowl: register there first Apr 23 16:11:12 ok Apr 23 16:12:33 hmm Apr 23 16:12:36 452, '4.3.1 Insufficient system storage', 'Patchwork ' Apr 23 16:15:16 can't build meta-toolchain on oe-git: ERROR: QA Issue: non -dev package contains symlink .so: zlib-sdk path '/work/i686-i686-sdk-angstrom-linux/zlib-sdk-1.2.3-r5/install/zlib-sdk/usr/local/angstrom/i686/lib/libz.so' - what changed? Apr 23 16:15:47 Also gcc won't package due to GNU_HASH, I believe Apr 23 16:17:18 But, more QA issues have been made fatal w/o trying meta-toolchain :) Apr 23 16:23:36 Tartarus, so what are you saying... does the meta-toolchain recipe need to be fixed, or should I disable QA or what? Apr 23 16:25:38 03Marcin Juszkiewicz  07org.openembedded.dev * rfcffc3712b 10openembedded.git/recipes/tasks/task-base.bb: task-base: added btusb driver to list of recommended BT drivers Apr 23 16:26:04 hi. I have a compilation problem found in OpenMoko which is related to your glibc package. Apr 23 16:26:17 http://rafb.net/p/XX51kp93.html Apr 23 16:26:47 here is the output. I sometimes can compile a language file, sometimes the same file abort. Debian Unstable :) Apr 23 16:29:02 tharvey, zlib-sdk needs fixing, most likely Apr 23 16:29:21 It's on my TODO list after this big loop finishes so I can get some stuff into stable Apr 23 16:29:24 then repeat :( Apr 23 16:31:14 Crofton, hrw the registration crashed with an SMTP error , my account seems to have been created but when I try to login it tells me the account is inactive Apr 23 16:33:00 03Jeremy Lainé  07org.openembedded.dev * r58dec3db21 10openembedded.git/conf/checksums.ini: checksums.ini: add entries for Qt 4.5.1 Apr 23 16:33:10 03Jeremy Lainé  07org.openembedded.dev * rd9558b5034 10openembedded.git/: Merge branch 'org.openembedded.dev' of ssh://git@git.openembedded.net/openembedded into org.openembedded.dev Apr 23 16:35:37 should I better report this on the bugtracker? Apr 23 16:38:09 Tartarus, that was just the first error... there are many packages that report that same QA error about a symlink .so - I'm not sure I understand it... why can't a regular package have a symlink for a .so? I would think thats standard Apr 23 16:39:21 tharvey, not quite, no, iirc Apr 23 16:39:28 libfoo.so goes into -dev Apr 23 16:39:37 Tartarus, looking back to before this issue the same QA Issues existed when building meta-toolchain and like you said now I guess those have been changed to fatal Apr 23 16:39:50 morning all Apr 23 16:39:51 yeah Apr 23 16:39:55 hey RP! Apr 23 16:40:03 bombardbombardbombard :) Apr 23 16:41:18 Tartarus, still dont' follow on the so's - package libz for example contains /usr/lib/libz.so.1.2.3 and /usr/lib/libz.so.1 which is a symlink Apr 23 16:41:27 Yeah Apr 23 16:41:31 But not libz.so Apr 23 16:41:35 which is I think the error one Apr 23 16:42:53 yes, thats true... didn't realize '.so' should be a -dev vs '.so.*' - not clear why but then I'm not a distro packager either :P Apr 23 16:45:00 Tartarus, looking at zlib-sdk recipe it just includes zlib recipe - since there seem to be many of these 'non -dev package containing .so' errors, perhaps its something that should be handled by the sdk class? Apr 23 16:45:13 tharvey, hmm Apr 23 16:45:23 Lets go looking :) Apr 23 16:45:45 Yeah, I think so Apr 23 16:45:51 Lets see, I think I see... Apr 23 16:46:32 actually there is only a few of those errors for packages included in meta-toolchain.... zlib-sdk ncurses-sdk and gcc-cross-sdk Apr 23 16:46:44 I'll have a patch in a sec Apr 23 16:48:48 and the GNU_HASH err exists in libgcc, ncurses, gcc-cross-sdk, gcc, g++, cpp, gconf, bzip2, wireless-tools, etc - way more of those Apr 23 16:49:02 Some of those should be fixed by now Apr 23 16:49:05 Some aren't Apr 23 16:49:12 gcc ones aren't Apr 23 16:49:21 others are since fs images use those Apr 23 16:49:32 ah your right... on oe-git its only a few of them left Apr 23 16:49:59 Actually, try this.. Apr 23 16:50:03 open up classes/sdk.bbclass Apr 23 16:50:11 remove the FILES_${PN} stuff Apr 23 16:50:28 * Tartarus wonders if RP will perk up now :) Apr 23 16:51:00 I think the QA error being fatal means we can't get away with the -sdk packages being lazy and not splitting things up Apr 23 16:51:11 Or we have to make the QA check not check on sdk Apr 23 16:51:24 Removing those lines will confirm we need to do one or the other I think Apr 23 16:51:41 (So then -c clean zlib-sdk and build zlib-sdk again) Apr 23 16:52:08 Tartarus: Which bit of this should be perking me up? :) Apr 23 16:52:29 where can I change the QA Errors back to non fatal for the time being? Apr 23 16:52:38 trying your patch now Apr 23 16:52:39 * RP doesn't remember the sdk packages needing the split things up? Apr 23 16:53:06 03Angus Ainslie  07fso/milestone5.5 * rd47c00157e 10openembedded.git/conf/distro/include/sane-srcrevs.inc: sanae-srcrevs : bump efl to latest recommended version Apr 23 16:53:09 The correct answer is probably to teach insane.bbclass that -sdk packages are special... Apr 23 16:53:14 03Angus Ainslie  07fso/milestone5.5 * r973480fa99 10openembedded.git/: Merge branch 'fso/milestone5.5' of git@git.openembedded.net:openembedded into fso/milestone5.5 Apr 23 16:53:15 03Angus Ainslie  07fso/milestone5.5 * r20c9abbf71 10openembedded.git/ (4 files in 2 dirs): Apr 23 16:53:15 e.bbclass : catch up to enlightenments interest new .so naming convention Apr 23 16:53:15 also affects ecore.inc edje_svn.bb and evas.inc Apr 23 16:53:39 Tartarus, yes, that patch removes the 'non -dev package contains symlink .so' errors Apr 23 16:57:50 RP, so, the QA check for libfoo.so in the main package thing is fatal now, thus breaking all lib-sdk recipes Apr 23 16:58:16 RP, split the -sdk packages up like "normal" or skip that QA check on inherting sdk||canadian-sdk class? Apr 23 16:58:32 Ah, I see yo answered :) Apr 23 16:58:41 OK, I'll make a patch and post Apr 23 17:01:25 RP, I think we also need, to be safe, to change sdk.bbclass to glob %{prefix}/*/.debug Apr 23 17:01:29 does that sound right? Apr 23 17:01:41 Or wait until something breaks? Apr 23 17:06:25 tharvey, I'll have a patch for you in a min Apr 23 17:07:04 Tartarus, thanks for the help! Apr 23 17:10:25 tharvey, RP, http://pastebin.com/d11bd3738 Apr 23 17:13:22 re Apr 23 17:14:18 stupid git question: patch doesn't like that format, is there a 'git-patch' that will apply it directly without me patching by hand? Apr 23 17:14:44 tharvey: git am Apr 23 17:16:08 03Angus Ainslie  07fso/milestone5.5 * r19344e91d9 10openembedded.git/recipes/linux/linux-openmoko-stable_git.bb: linux-openmoko-stable : correct kernel version for modprobe Apr 23 17:16:13 03Angus Ainslie  07fso/milestone5.5 * r0ee83d98af 10openembedded.git/recipes/xorg-lib/libxext-native_1.0.4.bb: libext-native : fix provides Apr 23 17:16:46 git-am doesn't like it either but I'm really not that up on git - probably doing something wrong. git-am wants a mail-msg format with a signoff Apr 23 17:17:03 signoff is not required Apr 23 17:17:35 hmm Apr 23 17:17:37 paste error? Apr 23 17:17:41 what's your email Apr 23 17:20:01 I simply pasted it from the pastebin into a file and did a 'git-am < ../patches/insane.patch' which gives 'Patch does not have a valid e-mail address' Apr 23 17:20:22 sorry... just seems silly to have to manually apply the patch when I know there has to be an easy way to do it in git Apr 23 17:21:43 same error on the e-mailed version - I'll just do it manually - I can't seem to get the hang of git Apr 23 17:21:46 git am patch - no need to git am ya... 'Patch does not have a valid e-mail address' Apr 23 17:22:29 then add one by hand? Apr 23 17:24:21 heh, too much work for a quick test of a patch Apr 23 17:24:28 no kidding... Apr 23 17:25:06 doesn't like me adding e-mail by hand - patch is 'not' in the format of the patches posted to maillist - guess thats it Apr 23 17:30:54 So, does the change itself work? That's the important thing to find out atm :) Apr 23 17:43:23 Tartarus, checking it now - got sidetracked on something Apr 23 17:50:56 Tartarus, that didn't fix 'non -dev package contains symlink .so: gcc' but fixed the other symlinks Apr 23 17:53:28 Tartarus, ah... I see the issue I had with trying to apply the patch - your pastbin doesn't have the '@@' before the line numbers for some reason Apr 23 17:54:26 gcc or gcc-cross-sdk ? Apr 23 17:54:44 gcc Apr 23 17:54:53 Yeah, gcc will need some love :) Apr 23 17:54:56 RP, still around? Apr 23 17:55:47 and of course the GNU_HASH still exists with gcc/g++/cpp/gcov Apr 23 18:01:18 03Marcin Juszkiewicz  07stable/2009 * r64f22bc77c 10openembedded.git/recipes/tasks/task-base.bb: Apr 23 18:01:18 task-base: added btusb driver to list of recommended BT drivers Apr 23 18:01:18 'btusb' driver replaced 'hci-usb' one few kernel releases ago. We recommend Apr 23 18:01:18 both as some devices use older kernels. Apr 23 18:01:18 Signed-off-by: Marcin Juszkiewicz Apr 23 18:01:19 Acked-by: Koen Kooi Apr 23 18:01:23 Acked-by: Tom Rini Apr 23 18:12:55 03Marcin Juszkiewicz  07org.openembedded.dev * r92e83440eb 10openembedded.git/conf/machine/bug.conf: Apr 23 18:12:55 bug: added BUG Labs device Apr 23 18:12:55 BUG is i.mx31 based device which is often called 'LEGO for IT'. Apr 23 18:12:55 More info: http://buglabs.net/ Apr 23 18:12:58 03Marcin Juszkiewicz  07org.openembedded.dev * rb5b0c24801 10openembedded.git/ (3 files in 3 dirs): Apr 23 18:12:58 linux-bug: added 2.6.27.2 from BUG Labs svn Apr 23 18:13:00 By default r8918 is used - can be overriden in local.conf by: Apr 23 18:13:02 SRCREV_pn-linux-bug = "${AUTOREV}" Apr 23 18:13:04 In BUG Linux we used AUTOREV by default. Apr 23 18:13:08 03Marcin Juszkiewicz  07org.openembedded.dev * r7ea3026aa3 10openembedded.git/recipes/firmwares/ (3 files in 2 dirs): Apr 23 18:13:11 marvel-gspi-fw: added firmware for WiFi bugmodule Apr 23 18:13:13 This is firmware for not-yet-released-into-sales wifi/bt bugmodule. Apr 23 18:13:15 03Marcin Juszkiewicz  07org.openembedded.dev * rd9b64f479b 10openembedded.git/recipes/netbase/ (netbase/bug/hosts netbase/bug/interfaces netbase_4.21.bb): Apr 23 18:13:20 netbase: added BUG support Apr 23 18:13:22 We use other network range for usb0 then OE default and this will not Apr 23 18:13:24 change soon. Also support for onboard Ethernet, Wifi module and Apr 23 18:13:26 USB-Ethernet adapters is present. Apr 23 18:38:46 03Koen Kooi  07org.openembedded.dev * rb0775dcad5 10openembedded.git/contrib/angstrom/sort.sh: angstrom feed sorter: add support for a few more machines Apr 23 18:43:44 re Apr 23 18:44:25 hi florian khem Apr 23 18:44:49 hey khem Apr 23 18:44:58 Can you look into GNU_HASH problems with 'gcc' please? Apr 23 18:45:44 I can't seem to figure out where the change was to make QA errors fatal Apr 23 18:46:49 tharvey: git log classes/insane.bbclass Apr 23 18:47:54 ah... now I see it thx Apr 23 18:52:07 Tartarus: where does it complain ? Apr 23 18:52:10 libgcc ? Apr 23 18:54:10 do we have some picture viewer for raw framebuffer? Apr 23 18:54:36 khem, gcov & others Apr 23 18:54:42 10:55 and of course the GNU_HASH still exists with gcc/g++/cpp/gcov Apr 23 18:55:02 hrw: frame buffer capture utility? Apr 23 18:55:08 hrw, sec... Apr 23 18:55:23 cbrake: no, displaying pictures on framebuffer Apr 23 18:55:26 hrw, fbida, iirc Apr 23 18:55:29 but that's big Apr 23 18:55:34 * Tartarus checks something else Apr 23 18:56:47 2 pastebins in a sec Apr 23 18:57:54 http://pastebin.com/m31bf89e http://pastebin.com/m3da99bab Apr 23 18:58:15 didn't push way back when since I didn't want to bother figuring out what would be the right set of graphic libs to enable for public Apr 23 18:58:32 but for our needs I just went with png so we could display an image without the convert and cat fun Apr 23 18:58:46 Tartarus: I suggested to change defaults to use GNU_HASH in toolchain Apr 23 18:58:57 no one seemed to be interested in idea Apr 23 18:59:00 Tartarus: push it ;) Apr 23 18:59:36 khem, it's needed now Apr 23 18:59:39 please push it or repost :) Apr 23 18:59:43 hrw, heh Apr 23 18:59:56 Tartarus: I need to cook it well. I just had done few tests Apr 23 19:00:16 khem, k Apr 23 19:00:22 khem, keep in mind 'gcc' doesn't build at all now Apr 23 19:00:26 or rather, fails to package :) Apr 23 19:00:46 hrw, I lack checksums for that around :) Apr 23 19:00:53 Grabbed that from an old snapshot Apr 23 19:01:03 thats not problem Apr 23 19:01:27 heh Apr 23 19:01:36 cat tmp/checksums.ini >>../oe/conf/checksums.ini Apr 23 19:02:07 Bah, fine :) Apr 23 19:06:03 Tartarus: thx! Apr 23 19:08:02 03Tom Rini  07org.openembedded.dev * r0bdb7654c8 10openembedded.git/ (3 files in 3 dirs): Apr 23 19:08:02 fbv: Add a framebuffer-based png viewer. Apr 23 19:08:02 bmp/gif is supported but disabled. Apr 23 19:08:47 also jpeg is supported Apr 23 19:08:52 heh Apr 23 19:08:59 See, you should have pushed :) Apr 23 19:09:04 ;D Apr 23 19:09:58 Also, heh, I see I wrote that before I learned that you're really supposed to start at r0 Apr 23 19:16:02 03Fraxinas  07org.openembedded.dreambox * r3c319a5a7d 10openembedded.git/packages/enigma2/enigma2.bb: enigma2: add gst-plugins-decodebin2 as RDEPENDS Apr 23 19:16:06 03Fraxinas  07org.openembedded.dreambox * r573c129a4b 10openembedded.git/packages/enigma2/enigma2.bb: enigma2: add packages for mediaplayer update using gstreamer decodebin2 Apr 23 19:16:08 03Fraxinas  07org.openembedded.dreambox * r7c29f48774 10openembedded.git/packages/gstreamer/gst-plugin-dvbmediasink.bb: gst-plugin-dvbmediasink: increment srcdate Apr 23 19:30:59 sed, mx31 or general? Apr 23 19:31:00 ga Apr 23 19:36:56 * kergoth wonders when bitbake/oe will switch to python 2.5 Apr 23 19:41:55 kergoth: as minimum? Apr 23 19:42:01 * kergoth nods Apr 23 19:42:12 what would it gives? Apr 23 19:42:55 i like certain things from 2.5, like the 'with' statement, personally Apr 23 19:53:20 hallo Apr 23 19:54:54 i am currently setting up openembedded on my new phat workstation, for my beagleboard, gta01/gta02, and hopefully soon enough, pandora.. Apr 23 20:08:22 bye Apr 23 20:17:38 hello all Apr 23 20:18:27 mickeyl, florian, anyone: I could use a bit of advice from you guys; is there simple command line tooling to upload a picture or WML text to a phone using obex over bluetooth? Apr 23 20:19:33 hi likewise Apr 23 20:19:40 hi florian Apr 23 20:20:33 hrm yes, there was something liike this... iirc part of the openobex stuff Apr 23 20:20:44 likewise: obexftp? Apr 23 20:21:23 or something like obexpush Apr 23 20:21:34 Jay7: did you actually use these tools? Apr 23 20:21:52 likewise: I've used some time ago (~1-2 year) :) Apr 23 20:22:24 ok, well I didn't get them to work (my fault). I also tried "obexd" with "send-files". Apr 23 20:22:48 hm Apr 23 20:23:11 likewise: I can check it with my Nokia E50 if you wish :) Apr 23 20:23:51 * keesj used gnokii for such things Apr 23 20:24:38 a WHILE ago Apr 23 20:25:08 Use case: An embedded device (terminal) outputs a blob of info, then the user want a summary of that on his phone. Apr 23 20:25:40 User holds phone near device, enters the pin code given by the embedded device and voila. Apr 23 20:25:47 obexapp is second Apr 23 20:26:32 Jay7: I tried *obex* in Ubuntu 9.04, even compiled obexd, no luck so far with my 6300 Apr 23 20:27:12 likewise: thats strange Apr 23 21:03:57 Tartarus: back now for a bit Apr 23 21:04:11 Tartarus: Pulling in the .debug packages seems unneccessary? Apr 23 21:04:57 Into SDK? Yeah, probably most people wouldn't want that, but I could see needing it from time to time Apr 23 21:05:11 What I was saying toh Apr 23 21:05:24 Was change the glob from %{prefix}/bin/.debug to %{prefix}/*/.debug Apr 23 21:05:36 But, lets just wait and see if something fails first Apr 23 21:06:06 Tartarus: ah, right. Yes, lets see... Apr 23 21:06:30 I'd just note that full .debug is huge Apr 23 21:06:40 * RP enabled that by default in Poky recently Apr 23 21:07:33 03Stanislav Brabec  07org.openembedded.dev * r9ce2add5af 10openembedded.git/recipes/man/man_1.5p.bb: man: GNU_HASH QA fix. Apr 23 21:18:35 hi woglinde Apr 23 21:21:52 hi mrdata Apr 23 21:22:46 woglinde send mail to you *done* ;-) Apr 23 21:26:58 mrdata yes thanks Apr 23 21:27:14 i hope i can test it on sunday Apr 23 21:28:13 is okay for me, also if it take longer time to test Apr 23 21:29:13 night all Apr 23 21:29:29 nite Apr 23 21:30:14 Is zecke around much these days? Apr 23 21:30:26 rp he is back in berlin Apr 23 21:30:34 I meet him today Apr 23 21:30:41 ah rp Apr 23 21:30:53 woglinde: I should catch him online sometime... Apr 23 21:31:01 how is the poulsbo work going? Apr 23 21:31:20 can we asume something in the near future Apr 23 21:31:31 * RP looks innocent and asks "what Poulsbo work?" ;-) Apr 23 21:31:45 hm Apr 23 21:31:48 are tw richards Apr 23 21:31:56 there two Apr 23 21:32:03 I guess from LKML its obvious I've at least been looking at that code... Apr 23 21:32:44 hm Apr 23 21:32:59 okay Apr 23 21:33:20 I ported the other intel driver to > 2.6.27 Apr 23 21:33:45 rp, try to keep the bike a bit more upright .... Apr 23 21:33:49 * Jay7 -> sleep Apr 23 21:34:15 Crofton|work: yes, the weekend hurt :( Apr 23 21:49:49 03Andrea Adami  07org.openembedded.dev * rf822ec3f5a 10openembedded.git/recipes/images/ (console-base-image.bb console-image.bb gnuradio-image.bb): Apr 23 21:49:49 console-base-image,console-image,gnuradio-image: readd psplash-zap Apr 23 21:49:49 - for distros without psplash keep the old behaviour and check 'screen' Apr 23 21:49:57 Hmm Apr 23 21:50:20 If I RFC'd that every distro sets SOMETHINGORANOTHERLIBC, the same, would anyone object? Apr 23 21:50:23 angstrom has one Apr 23 21:50:26 minimal another Apr 23 21:50:50 and if we also did OVERRIDES .= "${LIBC}" we could make uclibc stuff easier Apr 23 21:51:13 no more linux-uclibc || linux-uclibcgnueabi || linux-bfin (once that works again) Apr 23 21:52:39 03Frederik Sdun  07fso/milestone5.5 * rcfd273c117 10openembedded.git/recipes/ (ewww/ewww_git.bb webkit/webkit-efl_git.bb): Apr 23 21:52:39 ewww : add untested ewww recipe Apr 23 21:52:39 Signed-off-by: Sebastian Spaeth Apr 23 21:52:39 Signed-off-by: Angus Ainslie Apr 23 21:52:43 03Angus Ainslie  07fso/milestone5.5 * r3dcf455a79 10openembedded.git/classes/efl.bbclass: efl.bbclass : catch up to enlightenments interesting new .so naming convention Apr 23 21:58:19 03Andrea Adami  07org.openembedded.dev * r3c33e2e729 10openembedded.git/recipes/kexecboot/initramfs-kexecboot-image.bb: initramfs-kexecboot-image: minor text cleanings Apr 23 22:00:23 Tartarus: sounds good, i'm all for cleaning up distros and moving angstromisms commonisms Apr 23 22:00:28 +to Apr 23 22:00:52 OK Apr 23 22:03:15 woglinde: I'm still trying to build libgles-omap3-3.00.00.06. Updating to 2.6.29 kernel gave me other problems, this time, missing trace/kmemtrace.h in include/linux/slab_def.h Apr 23 22:04:38 mickeyl: I think there are race-issues with the initscripts :( Apr 23 22:04:50 not unlikely Apr 23 22:04:57 hoj did you update? Apr 23 22:04:59 I patched the kernel header to config-out the kmemtrace trouble in slab_def.h, but got undeclared symbol errors like for omap_dispc_register_isr, DISPC_IRQ_VSYNC etc Apr 23 22:05:09 koen check something more on today Apr 23 22:05:36 more in Apr 23 22:05:40 mickeyl: the strange messagebus thing on first boot..disappeared after some other fixes to psplash... Apr 23 22:06:03 (S99zzapsplash: not found) Apr 23 22:06:14 woglinde you mean do a git pull? Not yet today.. Apr 23 22:06:39 jepp Apr 23 22:06:43 try git pull Apr 23 22:06:50 ok Apr 23 22:09:55 hoglinde: after pull, clean and rebuild kernel then libgles? Apr 23 22:11:21 oops, woglinde: after pull, clean and rebuild kernel then libgles? Apr 23 22:11:24 sorry Apr 23 22:12:01 My latest musings about OpenEmbedded stuff: all a recipe really needs to know how to do, action wise, against its sources are: clean, produce binary artifacts in a structure such that they can be used (basically, compile + install into a usable structure). everything else isn't something specific to that particular recipe. packaging, etc, is just taking those deployed artifacts and doing things with them, and fetching is just obtaining the recipe and source Apr 23 22:12:11 * kergoth ponders Apr 23 22:12:53 hoj hm the kernel pr changed Apr 23 22:13:03 so the newer kernel should be build automagically Apr 23 22:13:49 still need 'PREFFERED_VERSION_linux-omap = "2.6.29"' in local.conf? Apr 23 22:13:59 hoj yes Apr 23 22:14:15 k Apr 23 22:21:35 * Crofton|work shudders at kergoth using the word "artifacts" Apr 23 22:21:42 * kergoth chuckles Apr 23 22:21:48 well, thats what CM folk like to call em Apr 23 22:21:52 yeah Apr 23 22:22:06 and thats a good enough term. i think the concept is a useful one, tracing what artifacts go into and come out of each step in the process Apr 23 22:22:10 some guy called some code I worked on an artifact and I felt like punching him Apr 23 22:22:13 better tracability, and a more functional aspect Apr 23 22:22:14 haha Apr 23 22:22:59 yeah Apr 23 22:23:01 well, I'm trying to think of various possible ways of improving the system, and it seems like we shove a lot into classes that really isn't in any way controlled by the recipe's contents Apr 23 22:23:02 03Stanislav Brabec  07org.openembedded.dev * ree95838567 10openembedded.git/recipes/fsi-client/fsi-client_20040803.bb: fsi-client: GNU_HASH QA fix. Apr 23 22:23:02 I suppose Apr 23 22:23:46 woglinde: omap3-sgx-modules_1.3.13.1397.bb failed, error in omaplfb_linux.c is "mach/display.h: No such file or directory" Apr 23 22:23:57 and go enjoy Phoenix before it turns into a furnace ... Apr 23 22:24:02 all a recipe really needs to contain is descriptive metadata as received by/for upstream, and an exertion of control over the 2-3 actions, cleaning the tree, producing binaries, and deploying those binaries into a structure where their assumptions will be satisfied, and it'll function Apr 23 22:24:04 * Crofton|work goes to mow Apr 23 22:24:23 basically, I think you are correct Apr 23 22:24:39 hoj *sigh* Apr 23 22:24:50 but since there are no consistent rules for building systems, the crud is there to handle the special cass Apr 23 22:24:51 * kergoth 's not really implying anything by this, persay, but thinks its good to define such things so we know what options we have for improving things without breaking things Apr 23 22:24:57 yeah Apr 23 22:25:28 the problem is the people that do not play nicely, such as people who create their own build systems Apr 23 22:25:45 woglinde I think I've been here before :) maybe bad build or oe source state? Apr 23 22:26:15 hoj hm try a clean Apr 23 22:40:28 Hmm, this is gonna be a bit big :) Apr 23 22:46:25 woglinde: did clean of kernel and libgles, rebuilt kernel (ok), rebuilt libgles (nok): "trace/kmemtrace.h: No such file or directory" Apr 23 22:48:36 * Tartarus kicks off a loop to see if all distros parse still Apr 23 22:50:54 hoj try do change that line into linux/kmemtrace.h Apr 23 22:51:00 http://patchwork.kernel.org/patch/17965/ Apr 23 23:02:23 * kergoth ponders Apr 23 23:02:37 woglinde: I patched kernel source that I found in tmp/work/omap3evm-angstrom-linux-gnueabi/linux-2.6.29+2.6.30-rc1-r1/linux-2.6.29 That must not be right, since I get the same compile errors. How does libgles get its kernel headers? I need to add the patch to a recipe maybe? Apr 23 23:16:56 anyone know of a package (with an oe recipe) that reads/writes to a 24cxx i2c eeprom? I know of eeprog and can create a package but if something else is more standard I would rather use that Apr 23 23:35:50 03Stanislav Brabec  07org.openembedded.dev * rb3c9508c51 10openembedded.git/recipes/unzip/unzip_552.bb: unzip: GNU_HASH QA fix. Apr 23 23:42:46 03Stanislav Brabec  07org.openembedded.dev * rf8f96581f7 10openembedded.git/recipes/pxaregs/pxaregs_1.14.bb: pxaregs: GNU_HASH QA fix. Apr 24 00:04:26 hmmmm Apr 24 01:40:37 03Angus Ainslie  07fso/milestone5.5 * r71ca6b08d6 10openembedded.git/ (6 files in 2 dirs): (log message trimmed) Apr 24 01:40:37 e.bbclass and efl.bbclass : more changes to work with the Apr 24 01:40:37 new .so naming scheme for enlightenment Apr 24 01:40:37 These files are also affect Apr 24 01:40:37 recipes/efl1/ecore.inc Apr 24 01:40:37 recipes/efl1/edje_svn.bb Apr 24 01:40:39 recipes/efl1/eina_svn.bb Apr 24 01:59:28 thought it was something like that Apr 24 01:59:29 ga Apr 24 02:42:02 03Denys Dmytriyenko  07org.openembedded.dev * r5677ff24df 10openembedded.git/recipes/udev/ (udev-124/default udev-124/init udev_124.bb): Apr 24 02:42:02 udev 124: add cache invalidation logic on kernel change or its bootargs/cmdline/atags Apr 24 02:42:02 also add /etc/default/udev with the option to disable device caching Apr 24 02:42:02 Signed-off-by: Denys Dmytriyenko Apr 24 02:42:02 Acked-by: Tom Rini Apr 24 02:42:06 Acked-by: Koen Kooi **** ENDING LOGGING AT Fri Apr 24 02:59:57 2009