**** BEGIN LOGGING AT Wed Jan 07 02:59:57 2009 Jan 07 05:36:24 archae0pteryxAFK: ping Jan 07 05:41:19 anyone used installed the jack package before? Jan 07 07:22:18 is there anyone online that has access to kill git branches ? Jan 07 07:22:30 good morning everybody Jan 07 07:25:51 hi Jan 07 07:26:28 I have download the m4 source code by using git. But I met error when I build the console-image. Jan 07 07:26:43 The error is FATAL: no configure script found Jan 07 07:27:00 could somebody help me? Jan 07 07:28:26 morning all Jan 07 07:34:40 could anybody help me? Jan 07 07:43:58 morning Jan 07 07:51:18 03Tick Chen  07org.openembedded.dev * rf68ec51f8f 10openembedded.git/ (2 files in 2 dirs): comic-reader: adding new bb file for comic-reader Jan 07 08:47:45 03Koen Kooi  07org.openembedded.dev * r37696e523b 10openembedded.git/conf/checksums.ini: checksums.ini: add sgx checksums Jan 07 09:32:56 03Koen Kooi  07org.openembedded.dev * rf197d4375f 10openembedded.git/packages/v4l2apps/libv4l_0.5.7.bb: Jan 07 09:32:56 libv4l: add 0.5.7. This is needed for e.g. gspca cams to work with newer kernels: Jan 07 09:32:56 * http://n2.nabble.com/zc3xx-webcam-%28041e%3A4034-Creative-Webcam-Instant%29-stopped-working-some-time-ago-%28since-gspca-kernel-integration-%29-td1667749.html#none Jan 07 10:07:01 hi florian Jan 07 10:07:08 lrg: ehlo Jan 07 10:07:11 morning Jan 07 10:19:15 morning Jan 07 10:29:58 hrw: moin & happy new year Jan 07 10:32:20 schurig: danke and hny too Jan 07 10:56:31 hi all, I met an error when building the gpephone-image Jan 07 10:56:44 sh: rm: command not found Jan 07 10:56:45 | + for i in locale-base-en-us Jan 07 10:56:45 | + opkg-cl -f /home/cch/embedded/stable/tmp/staging/x86_64-linux/etc/opkg.conf -o /home/cch/embedded/stable/tmp/rootfs install locale-base-en-us Jan 07 10:56:45 | Collected errors: Jan 07 10:56:45 | * Cannot find package locale-base-en-us Jan 07 11:14:54 03Koen Kooi  07org.openembedded.dev * rd0d237ed3d 10openembedded.git/packages/clutter/clutter.inc: clutter: add support for omap3evm Jan 07 11:21:06 * * OE Bug 4968 has been created by broonie(AT)sirena.org.uk Jan 07 11:21:08 * * Fix mainstone build Jan 07 11:21:10 * * http://bugs.openembedded.net/show_bug.cgi?id=4968 Jan 07 11:22:38 * broonie notes that it'd be rather helpful to stop requiring a description for attachments. Jan 07 11:32:10 broonie: did not you have r/w? Jan 07 11:32:17 hrw: No. Jan 07 11:32:31 I have r (well, everyone does) but not w. Jan 07 11:32:34 o.. I thought that you ahve Jan 07 11:34:23 71 machines set PACKAGE_EXTRA_ARCHS... Jan 07 11:34:45 Yes, I've already fixed at least one other. Jan 07 11:34:53 Looks like it's been cut'n'pasted all over. Jan 07 11:39:32 :) Jan 07 11:42:32 03Mark Brown  07org.openembedded.dev * r02ddc78cc8 10openembedded.git/conf/machine/mainstone.conf: mainstone: do not set PACKAGE_EXTRA_ARCHS - closes #4968 Jan 07 11:43:23 could you hrw fix that in the ts72xx.conf also, or should I file a bug? Jan 07 11:44:06 * * OE Bug 4968 has been RESOLVED (FIXED) by openembedded(AT)haerwu.biz Jan 07 11:44:08 * * Fix mainstone build Jan 07 11:44:10 * * http://bugs.openembedded.net/show_bug.cgi?id=4968 Jan 07 11:44:55 will do Jan 07 11:45:12 thanks Jan 07 11:45:55 03Marcin Juszkiewicz  07org.openembedded.dev * ra8a0ce7efd 10openembedded.git/conf/machine/ts72xx.conf: ts72xx: do not set PACKAGE_EXTRA_ARCHS Jan 07 12:16:06 03Florian Boor  07org.openembedded.dev * r7e891e410a 10openembedded.git/packages/maemo/ (xsp/xsp-fix-pc.patch xsp_1.0.0-8.bb): xsp: Fix pc file - passes sanity check now. Jan 07 12:16:17 03Florian Boor  07org.openembedded.dev * r5571d638e7 10openembedded.git/packages/maemo/xsp/xsp-fix-pc.patch: xsp: Remove whitespace from patch Jan 07 12:22:30 03Robert Schuster  07org.openembedded.dev * rf9754fa7c5 10openembedded.git/packages/llvm/ (llvm-2.4/fix-build.patch llvm_2.4.bb): llmvm 2.4: New recipe. Jan 07 12:22:42 03Robert Schuster  07org.openembedded.dev * r1864f2e0a1 10openembedded.git/conf/checksums.ini: checksums.ini: Added llvm 2.4 entry. Jan 07 12:36:42 64 files changed, 17 insertions(+), 58 deletions(-) Jan 07 12:36:47 conf/machine cleanup Jan 07 12:37:39 * broonie didn't want to do that without testing in case they did something useful Jan 07 12:44:33 hrw: good guy :) Jan 07 12:46:11 broonie: thats why I sent patch to oe ml instead of pushing Jan 07 13:02:06 * * OE Bug 4969 has been created by broonie(AT)sirena.org.uk Jan 07 13:02:08 * * Bump ALSA to 1.0.18 Jan 07 13:02:10 * * http://bugs.openembedded.net/show_bug.cgi?id=4969 Jan 07 13:03:14 * XorA suspects broonie should get a +w to his name :-D Jan 07 13:04:05 * * OE Bug 4970 has been created by sander(AT)3v8.net Jan 07 13:04:07 * * no dependencies generated on libpng3 package Jan 07 13:04:09 * * http://bugs.openembedded.net/show_bug.cgi?id=4970 Jan 07 13:15:45 03Koen Kooi  07org.openembedded.dev * r00b402ea10 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Jan 07 13:15:45 03Lukas Gorris  07org.openembedded.dev * r0062d5870c 10openembedded.git/packages/linux/ (4 files in 4 dirs): linux-msm7xxxx: update defconfigs and SRCREV Jan 07 13:15:45 03Koen Kooi  07org.openembedded.dev * rb5fe107851 10openembedded.git/conf/checksums.ini: checksums: add libv4l Jan 07 13:20:51 nice: http://www.qnap.com/pro_detail_feature.asp?p_id=108 Jan 07 13:21:27 03Koen Kooi  07org.openembedded.dev * r1289a8e309 10openembedded.git/packages/powervr-drivers/omap3-sgx-modules_1.3.13.1397.bb: omap3-sgx-modules: install into proper place Jan 07 13:57:19 03Koen Kooi  07org.openembedded.dev * r7c93a979f7 10openembedded.git/packages/powervr-drivers/omap3-sgx-modules_1.3.13.1397.bb: omap3-sgx-modules: really fix the install path Jan 07 14:07:04 hello Jan 07 14:12:37 03Koen Kooi  07org.openembedded.dev * rd45651ebb1 10openembedded.git/packages/linux/ (18 files in 4 dirs): linux-omap 2.6.28: update to newer DSS2 snapshot Jan 07 14:29:08 03Cliff Brake  07org.openembedded.dev * rf6876205bc 10openembedded.git/packages/mono/ (4 files): mono: disable thumb as incompatible inline ASM is used Jan 07 15:12:08 cbrake: is there a lot of asm in there? it would be a bit sad to disable thumb for the whole package if there are just a few files with assembler. Jan 07 15:12:28 pb_: I don't know Jan 07 15:12:32 plus, of course, half the time the eleet inline asm is just wrong and should be removed anyway :-} Jan 07 15:12:52 pb_: they generate ASM in the trampoline code as well, so I suspect there is other breakage Jan 07 15:13:04 ah, I see Jan 07 15:13:04 pb_: but I've not done an audit Jan 07 15:13:15 in that case, yah, running it as arm is probably a sensible thing to do for the moment. Jan 07 15:13:47 pb_: they are also still using the old syscal format in their trampoline code so I can't disable OABI in the kernel either :-( Jan 07 15:13:51 sigh Jan 07 15:14:04 doh Jan 07 15:14:49 that should be easy enough to fix, though, the two syscall ABIs aren't all that different. Jan 07 15:15:25 are they just using SYS_cacheflush or is there something more sophisticated going on? Jan 07 15:15:44 pb_: agreed -- that would probably be worth doing. Let me see if I can find the code ... Jan 07 15:17:20 g'day kergoth Jan 07 15:17:23 morning Jan 07 15:17:36 hey pb_ Jan 07 15:17:38 how's it going? Jan 07 15:18:16 not bad Jan 07 15:18:25 just back from holiday, still getting over that initial "back to work" shock Jan 07 15:19:02 ahh yeah.. the past few days havent quite got me back in the swing of things yet, myself Jan 07 15:19:06 morning Jan 07 15:19:10 hey chouimat|away Jan 07 15:19:11 what does PACKAGE =+ "jack-lib jack-server jack-examples" mean in a recipe? Jan 07 15:19:20 you mean PACKAGES? Jan 07 15:19:37 yes Jan 07 15:19:56 those are the binary packages that will be emitted by the packaging classes Jan 07 15:20:07 hey cosmicpenguin Jan 07 15:20:24 kergoth: have you used the jack recipe before> Jan 07 15:21:03 iirc =+ is a prepend or something... i'm a bit rusty :) Jan 07 15:26:44 ok I'm just having issues using the jack recipe as a dependancy for a recipe I'm building... it seems to build the jack recipe but those packages are not in the rootfs and they should be Jan 07 15:27:36 pb_: mono trampoline ARM ASM: http://anonsvn.mono-project.com/viewvc/trunk/mono/mono/arch/arm/arm-codegen.h?view=markup Jan 07 15:32:17 pb_: actually it was the mono_arch_flush_icache that was failing in this module: http://anonsvn.mono-project.com/viewvc/trunk/mono/mono/mini/mini-arm.c?revision=122416&view=markup Jan 07 15:32:26 pb_: and looks like it has already been fixed :-) Jan 07 15:32:31 heh :-) Jan 07 15:35:12 03Koen Kooi  07org.openembedded.dev * r952fbbaccb 10openembedded.git/packages/linux/ (6 files in 4 dirs): linux-omap 2.6.28: add patches to make XV work better with DSS2 Jan 07 15:49:01 what is the current status of ipkg/opkg ? opkg is used by default now (for Angstrom for instance) ? Jan 07 15:50:17 yes Jan 07 15:50:41 ok, but the packages still are .ipk files ? Jan 07 15:50:55 yes Jan 07 15:51:24 ok thank you Jan 07 16:06:15 .deb/.ipk/.opk all the same format, no sense inventing a new one for opkg Jan 07 16:11:26 do I then need to copy what I have mx31moboard in ipkg-collateral to opkg-collateral ? I guess yes Jan 07 16:36:43 hello, what's the difference between a funcion (def function_name) and something like "python package_do_split_locales() {" in the classes(for instance package.bbclass) Jan 07 16:39:27 one is a function, the other one is a task Jan 07 16:39:39 tasks can be shell or python Jan 07 16:39:49 ah ok thanks a lot Jan 07 16:41:15 03Koen Kooi  07org.openembedded.dev * rdc9c9617f7 10openembedded.git/packages/tasks/task-beagleboard-demo.bb: task beagleboard-demo: ship 'places' module and x11vnc-init Jan 07 16:46:44 03Koen Kooi  07org.openembedded.dev * rf202615bc4 10openembedded.git/packages/angstrom/e-wm-config-angstrom.bb: e-wm-config-angstrom: rrecommend places module Jan 07 17:14:07 * * OE Bug 4971 has been created by valentin.longchamp(AT)epfl.ch Jan 07 17:14:09 * * mobots image Jan 07 17:14:11 * * http://bugs.openembedded.net/show_bug.cgi?id=4971 Jan 07 17:14:11 mabe i understand...the package's name contains _...so mabe it just doesn't create the variables etc... Jan 07 17:17:44 mmm...it uses legitimize_package_name so...should not be that Jan 07 17:40:02 03Cliff Brake  07org.openembedded.dev * r8835b22aa1 10openembedded.git/ (10 files in 4 dirs): (log message trimmed) Jan 07 17:40:02 mono 2.2: update from preview 3 to RC1, add EABI patch Jan 07 17:40:02 This changeset renames recipes to better reflect mono versioning. Jan 07 17:40:02 Note, the previous mono 2.2 recipe was mis-named Jan 07 17:40:04 rc2 and was actually preview 2. If you built the Jan 07 17:40:06 misnamed 2.2.0-rc2 package, it will take precedence Jan 07 17:40:08 over the new recipe and you must manually remove Jan 07 17:50:47 cbrake: use PE then? Jan 07 17:51:58 anyone used the jack recipe? Jan 07 17:55:06 * * OE Bug 4972 has been created by valentin.longchamp(AT)epfl.ch Jan 07 17:55:08 * * the toolchain that would go with the mobots-image Jan 07 17:55:10 * * http://bugs.openembedded.net/show_bug.cgi?id=4972 Jan 07 17:58:45 mabe i know why it doesn't get exectued...there is no populate_packages....so populate_packages_prepend doesn't get exected...mabe that's because i do not inherit autotools Jan 07 18:23:50 03Koen Kooi  07org.openembedded.dev * rca084dba42 10openembedded.git/packages/powervr-drivers/libgles-omap3.inc: libgles-omap3: fix RRECOMMENDS and make binlocation more flexible Jan 07 18:25:33 \ Jan 07 18:25:39 ! Jan 07 18:26:30 HEH Jan 07 18:26:32 er Jan 07 18:40:43 hello, here's the recipe: http://rafb.net/p/0ZMHMc25.html and here's the include file: http://rafb.net/p/CevTdE32.html and here's the interesting debug part: http://rafb.net/p/Zf88VO36.html and here's what directories are created into the install directory: boost boost-dbg boost-dev boost-dev.shlibdeps boost-doc boost-locale Jan 07 18:41:26 and it doesn't create boost-$libname_sanitized Jan 07 18:41:53 i've been trying to debug this during days... Jan 07 18:43:01 is there anyone there? Jan 07 18:43:53 evry file is packaged into the boost-dbg... Jan 07 18:45:02 s/dbg/dev Jan 07 18:51:47 hrw: yes, I could use PE, but I don't think mono 2.2 has many users but me as it is not the default in Angstrom Jan 07 18:54:16 03Cliff Brake  07org.openembedded.dev * rc929ffc653 10openembedded.git/ (10 files in 4 dirs): (log message trimmed) Jan 07 18:54:16 mono 2.2: update from preview 3 to RC1, add EABI patch Jan 07 18:54:16 This changeset renames recipes to better reflect mono versioning. Jan 07 18:54:16 Note, the previous mono 2.2 recipe was mis-named Jan 07 18:54:16 rc2 and was actually preview 2. If you built the Jan 07 18:54:18 misnamed 2.2.0-rc2 package, it will take precedence Jan 07 18:54:22 over the new recipe and you must manually remove Jan 07 18:54:24 03Cliff Brake  07org.openembedded.dev * r6d843b2fba 10openembedded.git/conf/checksums.ini: checksums.ini: add checksums for devmem2 Jan 07 18:57:31 garren you are still having trouble? Jan 07 19:06:05 archae0pteryx: making progress wrote my own package and it seems to be working Jan 07 19:06:22 cool Jan 07 19:11:41 mickey, glad that youve all moved to git style repo. I assume the old mtn passes dont work anymore? Jan 07 19:13:47 I am trying to build a rootfs using angstrom and oabi the minimal-image works but "console-image" returns it can not find the recommended glibc package http://www.paste-it.net/private/o913bf5/ Jan 07 19:14:15 kristoffer: heh, you've been living under a rock? :) developers have been requested to send ssh keys for git multiple times via the mailing list Jan 07 19:14:43 I tried different things but a am now really lost Jan 07 19:17:28 mickey, Ive been under a rock of spam lately, so most stuff has been put in the trashcan :( linux-kernel mailinglist has been killing me. Jan 07 19:20:06 kristoffer: i see. send me an ssh key and you get access again Jan 07 19:20:33 mickey, thx, will get right to it. Jan 07 19:21:55 lkml kills Jan 07 19:22:04 perl discussion? Jan 07 19:38:05 * archae0pteryx maintains perl every few months Jan 07 19:44:57 mickey, Ive sent the email. Jan 07 19:45:29 archae0pteryx: so you can be good person for OE perl Jan 07 19:56:41 I want to use ruby actually Jan 07 20:04:27 by the way can we make dillo_0.8.6 depend on wget or should someone make a patch for busybox's wget because otherwise downloads doesn't work Jan 07 20:55:07 i'm having a similiar issue to this one http://projects.linuxtogo.org/pipermail/openembedded-devel/2008-October/006376.html can anyone help me resolve it? Jan 07 21:02:51 what package? Jan 07 21:04:28 lisppaste7: url Jan 07 21:04:28 To use the lisppaste bot, visit http://paste.lisp.org/new/oe and enter your paste. Jan 07 21:04:38 scant: you can use the above url to paste your log files... Jan 07 21:04:45 php-5.2.6 Jan 07 21:08:53 log.do_configure: http://pastebin.ca/1302985 run.do_configure: http://pastebin.ca/1302988 config.log: http://pastebin.ca/1302991 Jan 07 21:09:33 okay, but pastebin.ca doesn't work on the side of the worl I'm Jan 07 21:09:48 ? Jan 07 21:10:22 s/worl/world/.... opening pastebin.ca will simply timeout, so good luck with the n-1 other people in the channel Jan 07 21:10:27 scant pasted "log.do_configure" at http://paste.lisp.org/display/73251 Jan 07 21:11:51 scant annotated #73251 with "config.log" at http://paste.lisp.org/display/73251#1 Jan 07 21:12:29 scant annotated #73251 with "run.do_configure" at http://paste.lisp.org/display/73251#2 Jan 07 21:16:20 i have a new .bb file to build my task, it builds and installs the binaries to pkg/image/pkg/path but i haven't explicitly specd out how to make an .ipk so when baking my image i get an error about the my package missing, any ideas? or do i have to spec out how to make an ipk? Jan 07 21:19:02 hmm, actually i found a 'dev' ip for it in here deploy/glibc/ipk/i486/xxx.ipk, it shoudl find this right? Jan 07 21:20:12 zecke: r u still here? Jan 07 21:21:56 configure:3183: ccache armeb-linux-gnueabi-gcc -march=armv5te -mtune=xscale -mno-thumb-interwork -mno-thumb -mno-thumb-interwork -mno-thumb --version >&5 Jan 07 21:21:59 ccache: error while loading shared libraries: /slug/slugos/tmp/staging/armv5teb-linux-gnueabi/usr/lib/libz.so.1: ELF file data encoding not little-endian Jan 07 21:22:04 this comes from executing the compiler right? Jan 07 21:22:10 is LD_LIBRARY_PATH or such set? Jan 07 21:23:08 i'm using bitbake Jan 07 21:23:44 scant: oh well... that is a good start Jan 07 21:23:50 * cbrake takes another shot at a workaround for the broken PXA270 SD controller ... Jan 07 21:24:32 scant: I can not tell you what the problem is (otherwise it would be already fixed)... I can help you to find that out though... Jan 07 21:24:33 php-native_5.2.6.bb includes php-native.inc and php-native.inc contains: "export LD_LIBRARY_PATH = "${STAGING_LIBDIR_NATIVE}"" Jan 07 21:25:30 php_5.2.6.bb contains: "export LD_LIBRARY_PATH = "${STAGING_LIBDIR}"" and export PHP_NATIVE_DIR = "${STAGING_BINDIR_NATIVE}" Jan 07 21:26:18 scant: what happens is that you try to execute your compile (and it links to libz) and tries to resolve the library... Jan 07 21:26:26 scant: using LD_LIBRARY_PATH... Jan 07 21:29:00 zecke: i think since I'm cross compiling, the libraries are in the wrong format Jan 07 21:29:56 scant: well, what of the above is not understandable? Jan 07 21:32:03 is my LD_LIBRARY_PATH wrong? Jan 07 21:33:09 scant: if you type ldd /what/ever/your/path/is/armeb-linux-gnueabi-gcc Jan 07 21:33:25 zecke: does PHP need loaders Jan 07 21:33:41 scant: you will see it is linkging to libz... with LD_LIBRARY_PATH pointing to /slug/slugos/tmp/staging/armv5teb-linux-gnueabi/usr/lib/ Jan 07 21:33:42 then on cross compile it might be using the native loader Jan 07 21:33:59 i'm issue is very similiar to the web post I first mentioned Jan 07 21:34:02 scant: you instruct the dynmaic linker to search for libz in this directory... Jan 07 21:34:23 scant: zecke you can try the command from config.log manually Jan 07 21:34:36 i'm using bitbake trying to cross compile php-5.2.6 Jan 07 21:34:37 see if you get same error Jan 07 21:35:13 thats ok we need to figure out this particular configure test which is failing Jan 07 21:35:14 khem: php-5.2.6 seems to have both a native bb file Jan 07 21:35:35 and a regular php_5.2.6.bb file Jan 07 21:35:56 khem: ??? any invocation of gcc fails due a hostile LD_LIBRARY_PATH... anyway... I will leave you two to resolve it... Jan 07 21:37:33 khem: it's finding big-endian libraries, which as the web post suggests is correct, because these libraries are for the target, but i don't know how to change things so, *i think* the compile will use the native libaries, eventhough i'm cross-compiling, dunno why it needs native i386 libraries Jan 07 21:38:12 zecke: i don't know what to test/try, if you can help me resolve this, i would appreciate it Jan 07 21:38:14 scant: ldd /what/ever/your/path/is/armeb-linux-gnueabi-gcc Jan 07 21:39:44 scant annotated #73251 with "ldd /what/ever/your/path/is/armeb-linux-gnueabi-gcc" at http://paste.lisp.org/display/73251#3 Jan 07 21:40:00 zecke: is that what you're asking for? Jan 07 21:40:16 scant: no, I'm asking you to think and understand what is happening Jan 07 21:40:49 huh? Jan 07 21:40:53 scant: In your config.log (which _you_ should read), you see that any invocation of the -gcc is failing Jan 07 21:41:18 weird Jan 07 21:41:50 config.log is kind of really long, i'm not sure what i need to look for Jan 07 21:42:31 scant: configure fails? search for thet text string in config.log Jan 07 21:42:49 scant: execute /what/ever/.../armeb-linux-gnueabi-gcc -march=armv5te -mtune=xscale -mno-thumb-interwork -mno-thumb -mno-thumb-interwork -mno-thumb --version Jan 07 21:44:56 I'm having troubble fetching the mtd utils, the fetch is failing with a 404 Jan 07 21:45:11 zecke: i've located the errors in config.log, but i still don't know how to resolve the issue Jan 07 21:45:38 scant: you don't understand the issue... understand it first :) Jan 07 21:45:51 scant: I asked you to execute a simple command and three minutes passed... Jan 07 21:46:43 zecke: actually ccache needs libz.so Jan 07 21:46:59 and in scant's case Jan 07 21:47:01 khem: ah thanks, makes more sense Jan 07 21:47:14 because LD_LIBRARY_PATH is messed up Jan 07 21:47:22 no ccache is paralysed Jan 07 21:47:28 s/no/now Jan 07 21:47:37 zecke: armeb-linux-gnueabi-gcc (GCC) 4.2.4 is what that command returns Jan 07 21:48:00 scant: see the difference "ccache" and "LD_LIBRARY_PATH=/slug/slugos/tmp/staging/armv5teb-linux-gnueabi/usr/lib ccache" Jan 07 21:48:02 I wonder why we need export LD_LIBRARY_PATH = "${STAGING_LIBDIR}" in php_5.2.6.bb Jan 07 21:48:53 zecke: i don't understand what you mean Jan 07 21:49:19 try to execute above command Jan 07 21:49:52 which command? Jan 07 21:50:02 LD_LIBRARY_PATH=/slug/slugos/tmp/staging/armv5teb-linux-gnueabi/usr/lib ccache Jan 07 21:50:14 ... Jan 07 21:50:23 Anyone have a moment to help with a failure to fetch the mtd package? It *used* to work! Jan 07 21:50:37 it returns: ccache: error while loading shared libraries: /slug/slugos/tmp/staging/armv5teb-linux-gnueabi/usr/lib/libz.so.1: ELF file data encoding not little-endian Jan 07 21:50:49 bingo thats your problem Jan 07 21:51:12 should I clear my ccache? Jan 07 21:51:16 scant: no Jan 07 21:51:22 I just ran ccache -C Jan 07 21:51:36 maybe that's what caused this issue Jan 07 21:51:52 scant: You can comment out export LD_LIBRARY_PATH = "${STAGING_LIBDIR}" from php_5.2.6.bb Jan 07 21:52:01 and rebuild php Jan 07 21:52:09 khem: ok, let me try Jan 07 21:52:16 its a wild guess though Jan 07 21:52:25 NOTE: fetch http://www.angstrom-distribution.org/unstable/sources/git_git.infradead.org.mtd-utils.git_9845d92440bd87739c89edd000fd6e0c47fab185.tar.gz Jan 07 21:52:25 --13:49:08-- http://www.angstrom-distribution.org/unstable/sources/git_git.infradead.org.mtd-utils.git_9845d92440bd87739c89edd000fd6e0c47fab185.tar.gz Jan 07 21:52:25 Resolving www.angstrom-distribution.org... 85.214.40.226 Jan 07 21:52:25 Connecting to www.angstrom-distribution.org|85.214.40.226|:80... connected. Jan 07 21:52:26 HTTP request sent, awaiting response... 404 Not Found Jan 07 21:52:27 you might stumble elesewhere Jan 07 21:52:28 13:49:08 ERROR 404: Not Found. Jan 07 21:53:34 khem: I'm rebuilding now... Jan 07 21:54:21 Good luck scant. Jan 07 21:54:43 khem, any advice about my problem with mtd? Jan 07 21:54:51 thank you, i doubt it will work Jan 07 21:56:27 Patience is helpful in dealing with large distributed systems like OE :-) Jan 07 21:56:47 robtow: the file is not there Jan 07 21:57:38 Indeed. So should I perhaps look at the archive, and update the bb file - and submit the bb recipe back as a bugfix? I'm unsure about the protocol. Jan 07 21:57:44 khem: it sucessfully completed configuring and is now compiling, but i still doubt it will work Jan 07 21:58:02 It *did* fetch a couple of weeks ago. Jan 07 21:58:28 scant: many packages are not cross compilation friendly php could be one of them Jan 07 21:59:10 robtow: Are you building mtd-utils_1.0.0+git.bb Jan 07 22:00:12 Yes, khem; that appears to be what is invoked. I'm trying to do a complete build for beagleboard; something I've done in the past, all the way through loading and executing on the board. But now it's failing on this fetch. Jan 07 22:00:43 did you lose your local source cache? Jan 07 22:01:07 not that you should be having a problem, but it help mask problems like this :) Jan 07 22:01:27 may be delete your local source cache Jan 07 22:01:33 and rebuild it Jan 07 22:01:34 Crofton - I am deliberately doing a complete build NOT depending on a local cache. It's a requirement. Jan 07 22:01:36 maybe Jan 07 22:01:45 check if the mtd sources are there Jan 07 22:01:47 ah Jan 07 22:01:48 ok Jan 07 22:01:58 it should have checked out git tree Jan 07 22:02:07 so maybe the git tree is missing? Jan 07 22:02:09 I made a complete new directory tree and install, and this is the result. Jan 07 22:02:14 ok Jan 07 22:02:16 I understand Jan 07 22:02:27 possibly a transient failure on the git server Jan 07 22:02:38 the angstrom source mirror may not have a copy Jan 07 22:02:55 It appears to reference a repo on www.infradead.com, and that site is there... I have't looked deep into the tree there. Jan 07 22:03:24 khem: it's still compiling... Jan 07 22:03:29 I'm in the middle of something (as in this is procrastinatyion) Jan 07 22:03:43 scant: enjoy the anxious moments Jan 07 22:03:59 if I had time, I'd try rebuilding it locally and delete/mv the mtd-util source from my oe-sources .... Jan 07 22:04:15 I'll check later, after local success on my work task ... Jan 07 22:04:38 if you kept the sources, you could add it to the new source dir a s a test Jan 07 22:05:07 Pardon? Are you offering to build mtd and share it? I'd be happy to just have a pointer to something I could actually fetch.... like I said, I'm puzzled by the "protocol" in cases like this. Jan 07 22:05:33 Crofton - I have the sources from other successful builds, in other trees. Jan 07 22:05:45 khem: i've patched and built some bb files to cross-compile, but i don't understand the native and regular bb files, can you tell me what the difference/purpose is? Jan 07 22:09:50 no, just check the fetch at my end Jan 07 22:10:05 just making some suggestions you can use to narrow down the problem Jan 07 22:13:07 Crofton - a fetch from infradead.com works: http://git.infradead.org/mtd-utils.git?a=snapshot;h=06a91971e49dbcbd51854b2f9858a41e0837f993;sf=tgz Jan 07 22:14:25 So is there a problem with the angstrom source mirror? Jan 07 22:15:36 How can I best fix that? Jan 07 22:17:24 hmmm Jan 07 22:17:37 khem: the ipkg build, but when I try to install it: Collected errors: * Package (null) has no Architecture defined. Jan 07 22:17:39 if t fails, it should fail over ..... Jan 07 22:18:09 robtow, can you pastebin the entire log message ... Jan 07 22:18:16 ok Jan 07 22:18:30 I think the angstrom source mirror is a red herring Jan 07 22:20:08 It's a bit long, but here goes. I *do* appreciate the advice and help Jan 07 22:20:10 Here: Jan 07 22:20:17 bitbake console-image Jan 07 22:20:17 NOTE: Handling BitBake files: \ (1610/1610) [100 %] Jan 07 22:20:17 NOTE: Parsing finished. 1439 cached, 0 parsed, 171 skipped, 0 masked. Jan 07 22:20:17 NOTE: build 200901071418: started Jan 07 22:20:17 OE Build Configuration: Jan 07 22:20:38 methinks rob needs to learn about pastebins Jan 07 22:21:19 kergoth: and you can teach him :) Jan 07 22:21:56 robtow: don't paste more than 5 lines into an irc channel. google for a pastebin, there are a lot of sites specifically for pasting large quanitites of data Jan 07 22:22:25 Kergoth - thanks for the tip. Not familiar with "pastebin" Will do next time. Jan 07 22:23:05 its a site that takes a large paste and hands you a url to give someone Jan 07 22:23:11 often with syntax highlighting of the content Jan 07 22:23:21 NIce :-) I like it. Will do now. Jan 07 22:23:26 ~pastebin Jan 07 22:23:27 [~pastebin] A "pastebin" is a web-based service where you can paste anything over 3 lines without flooding the channel. Here are links to a few : http://www.pastebin.com , http://pastebin.ca , http://channels.debian.net/paste , http://paste.lisp.org , http://www.rafb.net/paste Jan 07 22:23:38 ah, right, ibot :) Jan 07 22:23:44 * kergoth hugs ibot Jan 07 22:24:48 here is the url to the pastebin: http://rafb.net/p/CoRxfx37.html Jan 07 22:26:03 the actual failure is that git fetch message, and the fetch command issued after it. the git fetcher isn'tw orking right with that new version of git, there Jan 07 22:26:24 it checked angstrom due to your PREMIRRORS probably? /me shrugs Jan 07 22:26:36 Hmmm. Jan 07 22:26:44 PREMIRRORS is defaulted. Jan 07 22:27:23 right Jan 07 22:27:43 it avoids beating some servers to death Jan 07 22:27:53 there are a couple of git repos on slow lines :) Jan 07 22:28:04 That sounds sensible. Jan 07 22:28:20 In *this* case, what would be an appropriate change to PREMIRRORS? Jan 07 22:28:32 the problem is line 134 Jan 07 22:28:40 nothing to do with the source mirror Jan 07 22:28:40 premirrors isnt the problem Jan 07 22:28:52 only that the source mirror does not have the source you need Jan 07 22:28:52 i just told you, its choking due to a git syntax change in the newer version of git Jan 07 22:28:52 oh? Jan 07 22:29:04 Arg. Jan 07 22:29:13 how do we fix this? Jan 07 22:29:16 robtow: mtd-utils build fine for me Jan 07 22:29:27 robtow: it pulled out git tree Jan 07 22:29:32 khem, you have local copy of source :) Jan 07 22:29:36 no Jan 07 22:29:38 Did it fetch ok as well as build, khem? Jan 07 22:29:39 hmmm Jan 07 22:29:44 Exactly. Jan 07 22:29:45 Crofton|work: it should probably do a git fetch --tags after the pull to get the tags, instead of using pull --tags, afaik Jan 07 22:29:48 * Crofton|work goes back to work :) Jan 07 22:29:54 * kergoth shrugs Jan 07 22:30:13 I am required to do the fetch (for external reasons I won't go into). Depending on a local cache isn't sufficient for my task. Jan 07 22:30:31 robtaylor, I understand Jan 07 22:30:38 this stuff should work for all cases Jan 07 22:30:42 er robtow Jan 07 22:30:48 "should" :-) Jan 07 22:30:52 i think fetch --tags always worked, o it shouldnt break compatibility Jan 07 22:30:57 s/o it/so it/ Jan 07 22:31:09 are we using git from git-native Jan 07 22:31:12 or hiost git? Jan 07 22:31:16 good question Jan 07 22:31:42 Well, I'm easy; just need to know how to fix the bb recipe, I guess, to use a "gbetter" syntax? Jan 07 22:31:49 robtow, what is your build machine running distro wise Jan 07 22:31:59 this is the git fetcher Jan 07 22:32:01 It is a Centos machine Jan 07 22:32:07 in bitbake I think Jan 07 22:32:07 maybe try a more recent bitbake, who knows, couldve been fixed already Jan 07 22:32:08 kraj pasted "mtd-utils git" at http://paste.lisp.org/display/73256 Jan 07 22:32:10 ah Jan 07 22:32:11 old git Jan 07 22:32:17 i think Jan 07 22:32:20 Centos 5.2, to be precise Jan 07 22:32:23 robtow git --version Jan 07 22:32:48 [rtow4@rtow4 build]$ git --version Jan 07 22:32:49 git version 1.5.2.1 Jan 07 22:33:10 the oldest I have is 1.5.3.3 Jan 07 22:33:12 hmmm Jan 07 22:33:23 Interesting. Jan 07 22:33:38 So perhaps I should update git, and try again, is that implied? Jan 07 22:33:46 worth a shot Jan 07 22:33:48 I am guessing ..... Jan 07 22:33:57 we can assume you have git on build machine Jan 07 22:34:06 since you need to use git to fetch oe metadata Jan 07 22:34:18 but, I suspect fetcher needs a minimum rev Jan 07 22:34:26 it would be good to ask on the ml Jan 07 22:34:44 and I really need to finish what i am supposed to be doing :) Jan 07 22:34:44 Well, I have used that version on a couple of machines, that DID do fetches and builds up through NOv/Dec. Jan 07 22:35:00 Did you update bitbake? Jan 07 22:35:27 I'm using bitbake-1.8.10 Jan 07 22:35:44 oe builds it own git for the host Jan 07 22:35:50 unless you ask it not to do so Jan 07 22:36:06 hmm. Jan 07 22:36:08 currently git-1.6.0.4 Jan 07 22:36:18 1.8.13 here for bitbake Jan 07 22:36:44 same here Jan 07 22:36:57 update bitbake first Jan 07 22:37:09 watch for BB_ENV_EXTRAWHITE issues Jan 07 22:37:13 Well, I could try the later version. And then tell my team mates to upgrade as well fi that works. Jan 07 22:37:20 * Crofton|work goes to the other machine Jan 07 22:37:22 right Jan 07 22:37:32 Was trying to keep things stable and shared.... but this problem cropped up, and it's stalled things. Jan 07 22:37:35 khem: I have a php ipkg now, but it doesn't install Jan 07 22:37:46 scant: why Jan 07 22:37:54 kergoth: at Openmoko we had the question why was tslib created and why not do the calibration, dejittering in kernel. do you know answers? Jan 07 22:39:01 khem: Collected errors: * Package (null) has no Architecture defined Jan 07 22:40:03 zecke: i expect its the usual kernel vs userland debate, but i couldn't say for sure, rmk would know, he started tslib Jan 07 22:40:14 i just took over after douglas lowder, who in turn had taken it over from rmk Jan 07 22:41:04 one could argue what filtering you do is a policy decision, and kernel is mechanism, not policy, i suppose Jan 07 22:41:04 * kergoth shrugs Jan 07 22:42:18 fair enough Jan 07 22:43:29 * Crofton|work does not like CORBA Jan 07 22:45:14 I see bitbake 1.8.12 as the latest on developer.berlios.de/projects/bitbake (14 Dec 2008) - not 1.8.13 Jan 07 22:45:34 I am pulling from svn I suspect Jan 07 22:45:44 robtow: odd numbers are development versions from svn Jan 07 22:45:45 try .12 :) Jan 07 22:45:53 ok. Jan 07 22:45:53 robtow: released will be even Jan 07 22:46:02 New day, new problem... how do I resolve this one when doing a git pull for openembedded ? Jan 07 22:46:05 ~/moko/openembedded$ git pull Jan 07 22:46:06 fatal: refs/remotes/origin/john_lee/openmoko-toolchain tracks both refs/heads/origin/john_lee/openmoko-toolchain and refs/heads/john_lee/openmoko-toolchain Jan 07 22:46:06 I'll make a new clean tree, and try it. Jan 07 22:46:44 robtow, it is always good to hear from people doing clean builds, they find these issues first Jan 07 22:47:26 Yes, Crofton, that's the nature of the beast :-) And it helps to have a relaxed attitude! Jan 07 22:47:44 At least I have a very fast machine for builds :-) Jan 07 22:47:46 :) Jan 07 22:47:51 I've searched the mailing list archive for January 2009 without seeing it being reported/resolved, any clues ? Jan 07 22:49:13 Can it be a configuration problem in .git/config i.e. with following lines: Jan 07 22:49:13 fetch = +refs/heads/*:refs/remotes/origin/* Jan 07 22:49:13 fetch = +refs/heads/*:refs/remotes/* Jan 07 22:49:13 ? Jan 07 22:49:59 (sorry about the flooding... :-/ ) Jan 07 22:51:13 ../common/.git/config only has one 'fetch = +refs/heads/*:refs/remotes/origin/*' line Jan 07 22:55:38 Hmm, I commented out the latter of 'fetch = +refs/heads/*:refs/remotes/*' in openembedded/.git/config and the 'git pull' succeeded without error, is it a 50% chance I got it right then or what do you think ? Jan 07 23:07:43 03Koen Kooi  07org.openembedded.dev * rbbbf56d7c1 10openembedded.git/packages/linux/ (linux-omap-2.6.27/nand.patch linux-omap_2.6.27.bb): linux-omap 2.6.27: fix NAND access **** ENDING LOGGING AT Thu Jan 08 02:59:57 2009