**** BEGIN LOGGING AT Thu Dec 27 02:59:57 2007 Dec 27 03:26:19 03freyther 07org.oe.dev * r1729c6fe... 10/ (1 packages/libqanava/libqanava_0.0.7.bb): packages/libqanava/libqanava_0.0.7.bb: Fix spelling of the description Dec 27 03:26:28 03freyther 07org.oe.dev * rd88d10fd... 10/ (7 files in 7 dirs): Dec 27 03:26:28 Qt4 based applications: Do not inherit qmake when building for Qt4/X11 Dec 27 03:26:28 inherit qmake will download Qt3 and build parts of it, we don't want that. Dec 27 03:26:34 03freyther 07org.oe.dev * r2039d1e2... 10/ (1 packages/qmake/qmake2-native/0001-fix-mkspecs.patch): Dec 27 03:26:34 packages/qmake/qmake2-native: Change the specs to find our resource compiler (rcc) Dec 27 03:26:34 The build of qsvn failed due invoking the wrong rcc. Change the qmake specs Dec 27 03:26:34 to set QMAKE_RCC and set it to OE_QMAKE_RCC. Dec 27 03:26:43 03freyther 07org.oe.dev * rfbd86972... 10/ (1 classes/qtopia4core.bbclass): Dec 27 03:26:43 classes/qtopia4core.bbclass: The current name of the include and libdir is qtopia Dec 27 03:26:43 Change the include and library path to end with qtopia as this is where Dec 27 03:26:43 qtopia-core is going to stage the headers in. The full blown Qtopia platform Dec 27 03:26:45 will use the same directory. Dec 27 03:26:48 03freyther 07org.oe.dev * re9f582e5... 10/ (4 files in 2 dirs): Dec 27 03:26:49 packages/qsvn: Experiment with building the same source in different variants Dec 27 03:26:51 Experiment with having the same source build with different versions of Dec 27 03:26:53 the underlying toolkit. In the best case we want to be installable in parallel. Dec 27 03:26:57 03freyther 07org.oe.dev * r3fc8eea0... 10/ (1 packages/qsvn/qsvn.inc): Dec 27 03:26:59 packages/qsvn/qsvn.inc: Propose a function to be able to set the FN Dec 27 03:27:01 With diverting packages ${PN} can sometimes not be used. In such cases Dec 27 03:27:03 we don't to use the filename of the currently included file. This can Dec 27 03:27:05 work with immediate assignments and a helper function like this one. Dec 27 03:27:07 03freyther 07org.oe.dev * r495a38aa... 10/ (1 classes/insane.bbclass): classes/insane.bbclass: Removal of whitespace and tabs Dec 27 03:27:12 03freyther 07org.oe.dev * r38f5d2ab... 10/ (1 MAINTAINERS): MAINTAINERS: Add qt and qmake to the recipes list. Dec 27 03:27:15 03freyther 07org.oe.dev * rdcd07d25... 10/ (1 classes/insane.bbclass): classes/insane.bbclass: Try to increase readability, remove deadcode, ask to make desktop errors fatal Dec 27 03:27:20 03freyther 07org.oe.dev * r162ce9b1... 10/ (1 classes/insane.bbclass): classes/insane.bbclass: Try to increase the readability Dec 27 03:27:26 03freyther 07org.oe.dev * rea1be019... 10/ (1 classes/insane.bbclass): Dec 27 03:27:26 classes/insane.bbclass: Make .la failures non-fatal the right way, split up the .la test in two parts Dec 27 03:27:27 -We have a function that can decide if a failure is fatal, use that instead of Dec 27 03:27:29 setting sane = True. Dec 27 03:27:33 -Test for installed and workdir separately. This allows us to ignore a certain error Dec 27 03:27:34 class and complain on others. Dec 27 03:27:37 03freyther 07org.oe.dev * rcf7ee5be... 10/ (1 classes/insane.bbclass): classes/insane.bbclass: Separate the workdir test and make it fatal Dec 27 03:27:44 03freyther 07org.oe.dev * r50619005... 10/ (1 conf/checksums.ini): conf/checksums.ini: Correct the URL of the Qt/X11 4.3.3 sourcecode Dec 27 03:27:51 03freyther 07org.oe.dev * re3d8733c... 10/ (1 classes/insane.bbclass): classes/insane.bbclass: Handle non-elf files gracefully (again) Dec 27 03:28:02 03freyther 07org.oe.dev * r39f105b1... 10/ (1 classes/insane.bbclass): classes/insane.bbclass: Decided to not make the workdir issues fatal Dec 27 03:28:12 03freyther 07org.oe.dev * r9374f526... 10/ (11 files in 6 dirs): Dec 27 03:28:12 Qt4/QtopiaCore4: Backport QT_LIBINFIX to add a suffix to libraries. Adapt qmake Dec 27 03:28:12 Add the 'E' suffix to QtopiaCore libraries. This allows to install both Dec 27 03:28:12 qtopiacore, qt/X11 at the same time. Dec 27 06:55:02 http://digg.com/2008_us_elections/Mike_Huckabee_If_you_vote_for_me_you_live_if_you_don_t Dec 27 06:55:02 Check out Mike Huckabee's latest psychotic utterance! Dec 27 07:54:48 when i build gtk with uclibc "libintl.h: No such file or directory" is reported. Dec 27 07:54:57 does it broken? Dec 27 09:53:54 readon: You probably need to build gettext Dec 27 09:55:59 morning all Dec 27 09:57:29 is there a reason for 'bitbake -c clean one_package' not to clean what one_package may have add to the staging directory ? Dec 27 09:58:52 * cyrilRomain think for example to .m4 files Dec 27 10:11:40 cyrilRomain: that's the whole idea of packaged staging I believe (i.e. it's a future feature) Dec 27 10:12:06 rwhitby: great Dec 27 10:12:20 morning Dec 27 10:12:31 cyrilRomain: tmp/ subdirs are not tracked yet Dec 27 10:12:45 hrw: morning Dec 27 10:13:05 ok, do you have an idea for when it is planned ? Dec 27 10:13:27 its work-in-progress Dec 27 10:13:45 hrw: it is 'safe' to switch ? Dec 27 10:13:50 add INHERIT += "packaged-staging2" into local.conf and do full-from-scratch build Dec 27 10:14:10 hrw: thanks Dec 27 10:14:41 * cyrilRomain read again RP mail Dec 27 10:16:47 03koen 07org.oe.dev * re3c90b11... 10/ (1 contrib/angstrom/build-feeds.sh): angstrom build-feed: add more packages and ppc405 machine Dec 27 10:22:27 re Dec 27 10:44:21 moin Dec 27 10:46:50 mornink Dec 27 10:47:28 * zecke kicks apr, apr-util and subversion's butts Dec 27 10:48:00 hi zecke Dec 27 10:48:11 moin :) Dec 27 10:52:54 morning Dec 27 10:54:24 hi CoreDump Dec 27 10:55:47 ~seen koen Dec 27 10:55:55 koen was last seen on IRC in channel #oe, 14d 49m 32s ago, saying: 'I forgot to commit that portion yesterday'. Dec 27 11:03:14 03koen 07org.oe.dev * rdf475b84... 10/ (1 packages/tasks/task-base.bb): task-base: bump PR Dec 27 11:03:22 03koen 07org.oe.angstrom-2007.12-stable * r58275f9d... 10/ (1 conf/distro/angstrom-2007.1.conf): angstrom: release 2007 Dec 27 11:11:05 zecke pasted "binconfig change for hrw" at http://paste.lisp.org/display/53101 Dec 27 11:11:20 hrw: is the above change okay for you? Dec 27 11:12:07 zecke: what for it is? Dec 27 11:12:42 zecke annotated #53101 with "user of it" at http://paste.lisp.org/display/53101#1 Dec 27 11:13:08 hrw: a user, I have a binconfig file that needs more treatment Dec 27 11:14:29 looks good Dec 27 11:14:38 hehe, it is labs.o-hand.com now? hehe Dec 27 11:21:44 koen went undercover? Dec 27 11:21:49 ~seen koen Dec 27 11:21:51 koen was last seen on IRC in channel #oe, 14d 1h 15m 28s ago, saying: 'I forgot to commit that portion yesterday'. Dec 27 11:22:15 hey likewise Dec 27 11:22:25 hi zecke Dec 27 11:22:51 hi Jin^eLD Dec 27 11:23:01 enjoying holidays? :> Dec 27 11:26:34 no, code sprinting at work :-) Dec 27 11:26:46 heh Dec 27 11:27:06 ~lart 2.6 kernel for being so big Dec 27 11:27:06 * ibot calls 2.6 kernel on the phone ... the lights are on but nobody's home for being so big Dec 27 11:28:45 =) Dec 27 11:45:45 Well, I had kexec commandlines working for the zaurus, then I wiped out the patch :( Dec 27 12:03:29 -rwxr-xr-x 1 hrw hrw 694484 gru 27 12:31 arch/arm/boot/zImage Dec 27 12:03:34 its hard to get lower Dec 27 12:05:32 next will be building 500K initramfs Dec 27 12:06:28 s/get lower/keep it that low in the future/ ;) Dec 27 12:06:57 CoreDump: its very stripped kernel Dec 27 12:28:12 Shoragan: http://wiki.openmoko.org/wiki/How_to_run_OpenMoko_Apps_on_PC Dec 27 12:50:38 hi Dec 27 12:50:51 tmbinc, hey Dec 27 12:51:03 zecke, are you at CCC already? Dec 27 12:51:17 gilligan_: he is Dec 27 12:51:28 ups, should not discover his privacy. ;) Dec 27 12:51:36 hehehe Dec 27 12:51:53 well I don't know what he looks like, so you didn't uncover much yet :) Dec 27 12:52:55 gilligan_: yes Dec 27 12:53:02 ;] Dec 27 12:53:05 gilligan_: we are in the hackcenter Dec 27 12:53:46 zecke, to be honest i have no idea where i am.. remember this is the first time for me at CCC .. i am in that "cellar room" where i think the lightning talks take place Dec 27 12:53:54 would that be the hackcenter? :) Dec 27 12:54:17 gilligan_: yes, cellar or suterain (whatever you want to call it) ;) Dec 27 12:55:16 zecke, :) where are you? I am sitting right at the entrance on the green matt which has "icmp 4" written on it Dec 27 12:56:14 gilligan_: hehe, do you see the OpenBSD table? Dec 27 12:56:54 uhm.. no? I suppose it has a big OpenBSD sign? :) Dec 27 12:58:00 gilligan_: is simon with you? Dec 27 12:58:21 03freyther 07org.oe.dev * r58fc861e... 10/ (4 files in 3 dirs): Dec 27 12:58:21 classes/binconfig.bbclass: Allow packages to add sed expressions Dec 27 12:58:21 packages/apr/apr-util,apr: patch the config files to be 'installed' and not point to the build and source-directory. Dec 27 12:58:27 03freyther 07org.oe.dev * rdcf75f3b... 10/ (1 packages/apr/apr-util_1.2.7.bb packages/apr/apr_1.2.7.bb): packages/apr/apr,apr-util: Stage the apr_rules.mk and install it as rules.mk in apr-util Dec 27 12:58:28 zecke: not yet.. he is still on his way Dec 27 12:58:32 03freyther 07org.oe.dev * r1b7a0ce6... 10/ (1 packages/apr/apr-util_1.2.7.bb): Dec 27 12:58:32 packages/apr/apr-util: Manually patch the .la file to not reference /usr/lib/FOO.la but use -lFOO instead Dec 27 12:58:32 subversion is still compiling and we have no more references to the workdir! Dec 27 12:58:37 zecke, as he chose to travel by train.. yuck.. Dec 27 12:58:50 gilligan_: go to the openbsd table and go left then Dec 27 12:58:56 gilligan_: I thought he arrived yesterday night? Dec 27 13:00:02 at my place.. but i live in the south of Germany Dec 27 13:00:12 zecke, and while I took the plane he took a train :) Dec 27 13:00:13 hehe Dec 27 13:01:01 well i'll go to the OpenBSD table now then.. i'm the one with glasses and a black/grey pullover and a "Schaal" looking puzzled and searching Dec 27 13:05:01 gilligan_: the word you were looking for was "scarf" ;) Dec 27 13:06:42 thanks ;]] Dec 27 13:07:00 and zecke i had no luck finding you.. left of OpenBSD was a bit of a lose description :) Dec 27 13:07:44 well i guess we'll run into each other later on Dec 27 13:19:16 hi all ... Dec 27 13:20:02 happy to see info about FOSDEM 2008 ... i'll be at there and happy to help ! Dec 27 13:26:06 zecke: In poky I changed qmake2-native into qmake2-cross since it installs things into the target staging area. Is that the right thing to do? Dec 27 13:30:07 RP: autobuilder is green Dec 27 13:31:10 we need to add FOSDEM 2008 to the wiki Dec 27 13:31:18 I can't edit the events page though Dec 27 13:58:10 03pH5 07org.oe.dev * re2bf9fb2... 10/ (1 conf/distro/angstrom-2008.1.conf): bump linux-handhelds-2.6 to 2.6.21-hh20 in angstrom 2008.1, too Dec 27 14:02:39 cbrake: hi Dec 27 14:02:55 hrw: hello Dec 27 14:03:14 cbrake: do you use mkfs.jffs(2) on device when you install mtd-utils? Dec 27 14:04:03 hrw: no, I generally make the jffs2 image on the workstation, and just use the flash/eraseall tools in mtd-utils Dec 27 14:04:12 cbrake: would be nice to move it to separate package as it depend on liblzo which is quite big (~170K) Dec 27 14:04:32 ops - 116K Dec 27 14:04:37 hrw: good point, that sounds worthwhile Dec 27 14:05:40 cbrake: how big is your rootfs for doing updates? I got 2.7M with uclibc and it is quite big Dec 27 14:05:59 4.5M uncompressed Dec 27 14:06:54 hrw: I have been using cpio.gz format (initramfs). Let me check ... Dec 27 14:07:00 3.4M without ipkg informations Dec 27 14:07:41 1.29M gzipped Dec 27 14:09:11 hrw: the last one I built a month ago was 1176159 bytes (cpio.gz format) Dec 27 14:10:06 hrw: I tried last week, and the build failed -- trying again now. I'm building Angstrom-minimalist-image-mtdutils-uclibc Dec 27 14:10:07 I want to go to ~500-600K Dec 27 14:10:43 hrw: I have 2.3MB of flash on the cm-x270, so I've not been motivated to squeeze it any more :-) Dec 27 14:10:53 hrw: starting my build again, and see what happens Dec 27 14:11:22 breakfast -- be back in a bit ... Dec 27 14:11:46 bon appetite Dec 27 14:36:26 cbrake_away: http://pastebin.ca/833672 Dec 27 15:08:38 hrw: looks good +1 for me Dec 27 15:10:18 ok Dec 27 15:10:22 pushing then Dec 27 15:19:02 any ideas why base image is not booring on the OSK, but console does? Dec 27 15:20:51 those remounting messages looks weird Dec 27 15:21:33 yeah Dec 27 15:21:43 I do not see them for console image Dec 27 15:22:13 you boot from flash... tried to boot base via nfs maybe? Dec 27 15:22:26 try to add debugs into boot process Dec 27 15:22:32 both were booted from flash Dec 27 15:22:40 no changes to kernel or u-boot env Dec 27 15:22:49 I am pretty sure it booted a few weeks ago Dec 27 15:22:52 er Dec 27 15:22:56 minimal booted Dec 27 15:23:43 would fstab be different for those images? Dec 27 15:25:28 three heads :) Dec 27 15:31:17 maybe I should reflash and see if the problem goes away :) Dec 27 15:56:07 ugh! the libtool binary name generated by the freetype do_configure differs from the current and the next generation packaged staging Dec 27 15:56:37 hrw, the log messages I posted where from a second boot Dec 27 15:56:53 looks like the first boot hangs when REmounting root filesystem Dec 27 15:57:03 remounting?? Dec 27 15:57:06 after doind teh /dev thing Dec 27 15:57:15 ah Dec 27 15:57:24 then the proc failure Dec 27 15:57:35 there is a delay in between those meesages Dec 27 15:58:06 I'm wondering what the differences are between base image and console? Dec 27 16:04:54 bye Dec 27 16:18:32 ~seen RP Dec 27 16:18:35 rp is currently on #maemo #handhelds #oe #openmoko #openzaurus. Has said a total of 16 messages. Is idling for 2h 52m 29s, last said: 'zecke: In poky I changed qmake2-native into qmake2-cross since it installs things into the target staging area. Is that the right thing to do?'. Dec 27 16:19:33 RP: could you send out your kexec-tools patch as well? I'll probably try out your patch later today. Dec 27 16:23:34 hvontres|work: hi Dec 27 16:25:40 hvontres|work: I've just pushed the kexec-tools patch into OE Dec 27 16:25:44 RP: morning :) Dec 27 16:26:15 RP: cool. i was hapy to see your patch in my inbox this morning :) Dec 27 16:26:40 hvontres|work: I said I'd get there ;-) Dec 27 16:26:43 RP: Is your patch intended to apply on top of mwester's original atag patch? Dec 27 16:27:02 hvontres|work: The patch I mailed applies against latest mainline git Dec 27 16:27:30 hvontres|work: It reverts most of the mwester patch from mainline (but I suspect he'll agree with it) Dec 27 16:27:59 RP: hehe.. Dec 27 16:28:36 * mwester has a number of concerns, but will wait until he has some time to look more closely. Dec 27 16:28:44 RP: Thanks a lot for looking into this :) Dec 27 16:28:53 mwester: What kind of concerns? Dec 27 16:29:45 mwester: I have tried to make sure it can do everything it did before but also solve several other scenarios... Dec 27 16:30:09 RP: heh... bugzilla just told me about the kexec-tools patch Dec 27 16:31:11 Ok, well then I best look more closely; there seemed to be a few cases that were no longer supported, and I don't see how ATAGs get to the new kernel at all. Dec 27 16:31:46 mwester: It needs a kexec-tools that is atags aware Dec 27 16:32:07 mwester: See the patch that just went into OE Dec 27 16:32:25 Yes, but user-space cannot put the atags into place as user-space AFAIK cannot deal with physical addresses Dec 27 16:32:49 mwester: How does the kernel image appear then? ;-) Dec 27 16:32:57 Kernel code Dec 27 16:33:09 mwester: The same relocation code adds the atags too now Dec 27 16:34:46 Ok, so the kexec userspace patch then also has a kernel component to add back in the code for the relocation? Dec 27 16:35:34 Also, what about devices that do not pass in ATAGs via R2, or devices that use the old-style bootparams? Dec 27 16:35:41 mwester: No, kexec has a powerful relocation mechanism built into it. We use that mechanism to make a custom set of atags appear in the right place Dec 27 16:36:43 And what happens if kexec = Y in defconfig, but the ATAGS = N? Unbootable kexec kernel? Dec 27 16:36:55 mwester: Whatever tags are used to boot the device are saved and are accessible from userspace by kexec. Any subsequently launched kernel uses the tags from userspace which are pointed at by R2. Dec 27 16:37:17 I guessed that. Dec 27 16:37:40 I'm wondering about the older bootloaders that don't set R2 - are they S.O.L. with this, then? Dec 27 16:37:59 mwester: kexec sets r2 and is the bootloader Dec 27 16:38:04 mwester: kexec = Y, ATAGS = N could lead to an unbootable kernel if the kernel doesn't have sane defaults and crashes without atags Dec 27 16:38:49 If you want to force atags on with kexec I have no strong preference, just mention it on the ARM list as I think some people won't like it :/ Dec 27 16:39:12 Ok, I must have missed the code that saves atags if they aren't pointed to by R2. Dec 27 16:39:18 mwester: The zaurus is an old device which doesn't use r2 in the bootloader. kexec works perfectly on the device Dec 27 16:40:02 Also, what insurance exists that would prevent a user from installing the new kernel but attempting to use it with the old kexec userspace binary (which wouldn't know to do the new atag thing)? Dec 27 16:40:05 about fosdem : i was just looking to book ... Astrid Hotel is already full ! :( Dec 27 16:40:55 mwester: It will just see invalid atags and hence skip them. If the kernel has sane defaults it will work, if it doesn't, that could cause problems Dec 27 16:41:37 Well, it would be a change in behavior, as the existing code preserves the boot atags, and doesn't assume the kernel is sane. Dec 27 16:41:39 mwester: The solution is to make sure the newer kexec-tools get into feeds ASAP I guess since its safe the other way around, userspace will just add some tags the kernel will ignore Dec 27 16:42:11 mwester: I've done my best to ensure everyone's interests are covered as best I can :/ Dec 27 16:43:31 03freyther 07org.oe.dev * rc6fac7ea... 10/ (1 packages/qt4/qt_staging.inc): packages/qt4/qt_staging.inc: Manually fix the .la files to say installed=no Dec 27 16:43:36 03freyther 07org.oe.dev * rfc6211a2... 10/ (1 classes/autotools.bbclass): Dec 27 16:43:36 classes/autotools.bbclass: Use bruteforce and set installed=yes to =no Dec 27 16:43:36 This is silenting issues with various libtool files that have installed=yes. Dec 27 16:43:42 03hrw 07org.oe.dev * ra4569078... 10/ (1 packages/mtd/mtd-utils_1.0.0+git.bb): mtd-utils: move mkfs.jffs(2) to separate packages - mtd-utils do not need LZO Dec 27 16:43:47 03rpurdie 07org.oe.dev * r28e2872e... 10/ (3 files in 3 dirs): kexec-tools: Add support for ARM atags allowing the commandline to be set (based on #3430 with changes) Dec 27 16:54:01 jsheldon: hi Dec 27 16:54:20 jsheldon: you mean you could not compile qemu 0.90 ? Dec 27 16:54:23 ant|work: hello Dec 27 16:54:44 ant|work: correct, same as in this bug http://bugs.openembedded.org/show_bug.cgi?id=2157 Dec 27 16:55:02 jsheldon: I moved on Gentoo and cannot compile binary locales anymore... Dec 27 16:55:18 jsheldon: binutils 2.18 and gcc 3.4.6 Dec 27 16:55:36 jskheldon: 3.3.6 won't work too Dec 27 16:56:05 jsheldon: so finally I put ASSUME_PROVIDED += "qemu-native" Dec 27 16:56:09 ant|work, jsheldon: What is your gcc 3.x called? Dec 27 16:56:22 jsheldon : but the build wasn't successful Dec 27 16:56:41 jsheldon: 3.3.6 ( I installed a binary) Dec 27 16:57:23 jsheldon: so now I'm skipping the binary locale generation... Dec 27 16:57:23 ant|work: yeah that bug says ASSUME_PROVIDED isn't a good solution Dec 27 16:57:44 jsheldon: I got it working on ArkLinux Dec 27 16:57:57 jsheldon: quemu was 0.82 at the time Dec 27 16:58:16 ant|work: ah, well there is a patch in that bug, I might work on adapting it and posting it into the bug Dec 27 16:58:40 ant|work: I'm just surprised that more people don't have the same problem Dec 27 16:58:43 jsheldon: what is your distro Dec 27 16:58:44 ? Dec 27 16:58:54 ant|work: debian etch Dec 27 16:59:34 The proper solution is going to be a gcc-native_3.4.4 :/ Dec 27 17:00:36 ant|work: I had that problem on Gentoo, but with gcc-config I switch back and forth to build qemu with gcc3.x, and anything else with gcc4.x Dec 27 17:00:59 hi mickeyl Dec 27 17:01:09 cyrilRomain: eek :/ Dec 27 17:01:12 hi guys Dec 27 17:01:16 RP:cyrilRomain: I installed Gentoo last week - FAILS! Dec 27 17:01:36 gcc 3.3.x for Gentoo are MASKED Dec 27 17:01:49 3.4.7 fails during compilation Dec 27 17:01:55 sorry 3.4.6 Dec 27 17:02:04 * * OE Bug 3430 has been RESOLVED (FIXED) by rpurdie(AT)rpsys.net Dec 27 17:02:07 * *  Patch for adding Cmdline support in kexec-tools Dec 27 17:02:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3430 Dec 27 17:03:11 cirylRomain: I mean you can install gcc 3.4 but then fails during qemu-compilation Dec 27 17:03:59 cirylRomain: and the same gcc fails generating binary locales Dec 27 17:04:09 ant|work: :( Dec 27 17:04:14 ant|work: but wait.. I don't have to switch anymore since few months ... Dec 27 17:04:30 cirylRomain: don't do! Dec 27 17:04:58 cirylRomain: tonite I'll test binutils-config Dec 27 17:05:28 cirylRomain: what's your binutils version? Dec 27 17:05:43 ant|work: 2.18-r1 Dec 27 17:06:29 cirylRomain: would you patebin emerge --info please? Dec 27 17:06:34 *pastebin Dec 27 17:07:31 ant|work: http://rafb.net/p/HL4f9v17.html Dec 27 17:07:46 cyrilRomain: thanks a lot Dec 27 17:08:44 cyrilRomain: i see ... I'm on "-march=prescott" Dec 27 17:09:06 cyrilRomain: not safe enough? Dec 27 17:09:20 ant|work: dunno Dec 27 17:09:36 cyrilRomain: Core2 Dual E6600 Dec 27 17:10:35 * ant is tempted to test on 64 bit but is already afraid... Dec 27 17:17:07 ant|work: qemu<0.9.0 needs gcc<4, but qemu-0.9.0 build fine with gcc4.x on Gentoo, and also in oe (at least for me) Dec 27 17:17:28 ant|work: but not for everyone :-/ Dec 27 17:17:51 cyrilRomain: now I see it clear: qemu was bumped to 0.90 Dec 27 17:18:20 cyilRomain: but now, with 0.90 bitbake qemu fails Dec 27 17:18:39 cyrilRomain: using the OE toolchain Dec 27 17:18:48 cyrilRomain: and gcc3 on host Dec 27 17:19:07 cyrilRomain: probably it is a gentoo problem... Dec 27 17:21:48 03freyther 07org.oe.dev * r334c3fe8... 10/ (3 files in 3 dirs): packages/qt4/qtopia-core: Force building of tools, demos and examples Dec 27 17:21:55 cyrilRomain: http://packages.gentoo.org/package/sys-devel/gcc?arches=all Dec 27 17:21:55 03freyther 07org.oe.dev * r16d9e54a... 10/ (3 files in 3 dirs): packages/qt4/qt_packaging.inc: Fix packaging of the QtUiTools${QT_LIBINFIX} Dec 27 17:22:03 03freyther 07org.oe.dev * r2c1b6fa0... 10/ (1 packages/qt4/qt_staging.inc): packages/qt4/qt_staging.inc: Stage the .prl and semi private Qt libraries. Dec 27 17:22:29 heh Dec 27 17:22:43 gentoo guys: how EXACTLY is your gcc3 BINARY named? Dec 27 17:22:52 gcc346? gcc-3.4.6? gcc-3.4? Dec 27 17:22:59 ant|work: not sure ... Most of issue I founs specific with Gentoo is wrong 'ar' and 'ranlib' of some packages (such errors do not occurs with other distro) Dec 27 17:23:02 hrw: I tried ALL the version I found on the net! Dec 27 17:23:07 gcc 3* Dec 27 17:23:13 ant|work: BINARY! Dec 27 17:23:26 /usr/bin/gcc-3.4 in Debian it is Dec 27 17:23:43 hrw : my pc is 10 miles away...later Dec 27 17:23:56 hrw: but read here about 3.3 Dec 27 17:23:57 jsheldon: same question Dec 27 17:24:00 http://packages.gentoo.org/package/sys-devel/gcc?arches=all Dec 27 17:24:17 hrw: i686-pc-linux-gnu-3.4.6 Dec 27 17:24:44 Drop keywords as this version doesnt built on modern systems anymore Dec 27 17:24:48 cyrilRomain: gracias Dec 27 17:25:06 so the solution is easy Dec 27 17:25:29 edit packages/qemu/qemu-gcc3-check.inc and add i686-pc-linux-gnu-3.4.6 into list Dec 27 17:26:19 hrw: I tricked this by renaming and symlinking..was not just name Dec 27 17:27:08 hrw: I read the whitelist Dec 27 17:27:38 hrw: and finished instaling a precompiled binary! on Gentoo ! Shame! Dec 27 17:28:22 hrw: but even the 3.3.6 doesn't install glibc (localedef error) Dec 27 17:28:36 hrw: I'm going for a build of org.openembedded.angstrom-2007.12-stable for the cm-x270, angstrom-2007.1.conf uses gcc 4.1.2 Dec 27 17:29:27 hrw: I was attempting as simple a build as possible, essentially all I changed in local.conf was the machine, then bitbaked minimal-image Dec 27 17:29:27 jsheldon: I even compiled all but qemu with gcc 4.3.0-ark Dec 27 17:30:28 bb in 30 mins Dec 27 17:30:49 I have gcc 4.2.x and 3.4.x in system and qemu just works Dec 27 17:31:06 /usr/bin/gcc is 4.2.x /usr/bin/gcc-3.4 is 3.4.x Dec 27 17:31:20 hrw, so qemu uses the host's gcc? Dec 27 17:32:08 in which case, my host's gcc is the same, 4.1.1 Dec 27 17:32:20 scratch that, 4.1.2 Dec 27 17:53:50 qemu-native as all -native and -cross recipes use host gcc Dec 27 17:59:03 !logs Dec 27 17:59:04 Channel logs for #oe are archived at: Dec 27 17:59:04 http://hentges.net/tmp/logs/irc/%23oe Dec 27 17:59:05 Live-logs are available at Dec 27 17:59:08 http://hentges.net/tmp/logs/irc/livelogs/%23oe.livelog Dec 27 17:59:10 See ?? help-logs for usage instructions Dec 27 18:24:13 bye all Dec 27 18:51:44 Just a heads up, the bugtracker might go down a few times over the next week or two as I do some hardware upgrades Dec 27 19:22:17 ggilbert: excelent timing :) Dec 27 19:22:26 Yo Dec 27 19:24:05 It's been running a bit low on RAM and I've continually put off driving up to install some more :) Dec 27 19:25:19 * hvontres|work thinks all server farms should have an exelent pub nearby :) Dec 27 20:56:38 hello all Dec 27 20:59:35 gm Dec 27 21:05:35 survived xmas? :-) Dec 27 21:30:35 * cyrilRomain survived xmas, but start to wonder if he will survives OE ... Dec 27 21:31:39 cyrilRomain: :-) tell me Dec 27 21:33:34 likewise: OE is so great that it shows the limit of the imperative model of software package management Dec 27 21:34:58 likewise: OE requires so much time and effort to use and to play with ... Dec 27 21:35:11 cyrilRomain: I know. Dec 27 21:35:52 cyrilRomain: other build systems target only a small set of packages, with limited tunables (C library, anyone) and a few machines. Dec 27 21:36:08 OE is like a mozilla monster. Dec 27 21:36:13 likewise: I know, and that's why OE is so great Dec 27 21:36:28 :-) Dec 27 21:36:38 but at the same time it is such a pain .. to only pick and assemble software Dec 27 21:36:43 it shouldn't be so Dec 27 21:36:52 OE rocks...after you get that bloody learning curve behind you ;) Dec 27 21:37:59 There is no such thing as having 'learnt' OE... Dec 27 21:38:17 CoreDump: Was the bug a missing DEPENDS btw? Dec 27 21:38:23 ^^^^ What he said!! :D Dec 27 21:39:03 Everytime I think I'm past the steep part of the curve, I realize that I just fell off the curve. Dec 27 21:39:08 RP: yes it was Dec 27 21:40:31 CoreDump: Was a fix for the DEPENDS pushed? Dec 27 21:40:36 hi Dodji Dec 27 21:40:40 yes, by hrw Dec 27 21:40:47 hey RP Dec 27 21:41:13 I will add them as I find them. There really isn't any other way Dec 27 21:41:33 CoreDump: No, sadly not :( Dec 27 21:41:51 CoreDump: As I said, I hope packaged staging will at least allow an exhaustive test... Dec 27 21:42:00 ;) Dec 27 21:42:25 Crofton|home: Hello. What was that patch/diff split util again? Dec 27 21:44:55 RP: looks like I'll need to massage your patch to work on 2.6.17 on the way home tonight :) Dec 27 21:45:04 * hvontres|work likes the train :) Dec 27 21:45:35 hvontres|work: I've just pushed a patch for 2.6.23 which will probably work better for 2.6.17 Dec 27 21:45:52 RP: in OE? Dec 27 21:46:15 hvontres|work: yes, but its also http://www.rpsys.net/openzaurus/patches/export_atags-r0.patch Dec 27 21:46:33 RP: muchas gracias :) Dec 27 21:47:57 RP: do you know of any other devices / dev boards that use pxafb on a pxa25x? I would like to get a better idea of waht is hapening in http://bugs.openembedded.org/show_bug.cgi?id=2120 Dec 27 21:55:24 hvontres|work: Offhand I can't think of anything 25x based Dec 27 21:55:36 hvontres|work: Does it happen with 2.6.19 or .18? Dec 27 21:56:31 RP: IIRC, I first saw it in .19. I haven't had a chance to try .18 yet. What is the easiest way to resurect a deleted recipie? Dec 27 21:58:33 RP: sorry, I am still a mtn newb.... Dec 27 21:58:33 hvontres|work: switch to a revision from back then, I guess you can find when it was deleted with mtn log Dec 27 21:58:45 RP: thanks Dec 27 21:59:24 03rpurdie 07org.oe.dev * rfbb56d5a... 10/ (14 files in 3 dirs): linux-rp-2.6.23: Add atags export patch for kexec use. Fix spitz suspend key issue. Sync kernel mirror uri tweak from poky. Dec 27 21:59:29 LOL Dec 27 21:59:32 load average: 22.37, 21.21, 13.14 Dec 27 21:59:44 BB_NUMBER_THREADS ftw Dec 27 22:00:20 And I am wondering why I can't do anything else on this box while BB is running ;) Dec 27 22:00:55 CoreDump: bitbake threading is one of the nicest stress tests I know :) Dec 27 22:01:29 a load of 22 is scary =\ Dec 27 22:01:33 CoreDump: ftw? Dec 27 22:01:44 for the win Dec 27 22:02:08 CoreDump: That is nothing ;-) Dec 27 22:02:23 hehe Dec 27 22:05:45 hvontres|work: 2.6.17 was the last kernel version I booted on poodle it seems :/ Dec 27 22:06:23 So, dumb question, but who would be the best person to talk to about oe bugzilla permissions? Dec 27 22:06:29 RP: I've noticed :) Dec 27 22:06:43 I've been looking for Koen for a while, but he doesn't seem to be about much Dec 27 22:06:50 ggilbert: OE core team I guess? Dec 27 22:06:59 Who'd that be these days? Dec 27 22:07:10 * hvontres|work thinks RP just volunteered Dec 27 22:07:16 haha Dec 27 22:07:28 ggilbert: openembedded-private@openembedded.org Dec 27 22:07:34 kk Dec 27 22:07:42 Thanks Dec 27 22:19:39 hvontres|work: h3900, h5000, h2200 are pxafb based with pxa25x iirc Dec 27 22:20:04 hvontres|work: gumstix-connex too but they do not have screen too often Dec 27 22:27:15 eww, I hope he doesn't try pdax Dec 27 22:27:51 he'll go balistic when he screws up the u-boot install =) (which he probably will) Dec 27 22:28:03 bgold? Dec 27 22:28:12 yeah Dec 27 22:35:33 CoreDump: do you think bgold is cyorxamp reincarneted? Dec 27 22:36:02 hrw: thanks . Dec 27 22:36:05 gods no! Dec 27 22:36:08 ugh I hope not Dec 27 22:36:37 CoreDump: wait, couldn't be...he dosn't swear enough :0 Dec 27 22:36:57 read the backlog ;) Dec 27 22:37:10 * hrw -> bed Dec 27 22:37:15 see you tomorrow guys Dec 27 22:37:20 n8 hrw Dec 27 22:37:37 my neo will run busybox 1.8.2 (from Poky) since today Dec 27 22:44:09 CoreDump: well, he swears a bit, but nothing compared to goold ol' cyorxamp :) Dec 27 22:46:06 =) /mode +b cyorxamp!* Dec 27 22:46:24 hi Dec 27 22:46:38 ah, he is still banned in #angstrom Dec 27 22:51:03 Crofton|home: ping Dec 27 22:57:27 hvontres|work: Does charging work for you on poodle? Dec 27 22:58:28 03rwhitby 07org.oe.dev * rbed1f45d... 10/ (1 packages/meta/slugos-packages.bb): slugos-packages: Added task-nas-server-* and task-proper-tools Dec 27 23:12:59 RP: as far as I can tell, yes. Dec 27 23:13:29 RP: but I have been running mostly 2.6.17 kernels due to the display bug Dec 27 23:14:06 hvontres|work: 2.6.23 is in a mess. There is an LED driver bug and resume doesn't happen at all :( Dec 27 23:14:16 2.6.17 doesn't appear to charge for me either... Dec 27 23:15:43 RP: hmmm, I'll have to check that when I get home tonight. No AC adapter here at work, just my usb-powered cradle :) Dec 27 23:16:28 RP: 2.6.17 on OZ seems to charge ok. (LED is off right now) Dec 27 23:18:08 RP: on 2.6.23 I found that the LED driver oops went away after I did a menuconfig to get the sound driver enabled again.. Dec 27 23:18:47 has anyone gotten the linksys WCF12 to work on any rom? If so, what did you do to get it to work? Dec 27 23:22:35 hvontres|work: That worries me slightly, it shouldn't matter about sound :/ Dec 27 23:23:45 RP: I think make oldconfig does something strange to the 2.6.17-based defconfig Dec 27 23:27:55 RP: btw, 2.6.21 also has no sound drivers using the standard defconfig. Dec 27 23:31:00 hvontres|work: I'm planning to concentrate on .23 ;-) Dec 27 23:32:38 likewise, pong Dec 27 23:32:47 I wasn't sure what uyou meant about teh patch thing Dec 27 23:33:16 and I am off to a birthday party Dec 27 23:33:18 l8r all Dec 27 23:33:20 Crofton||home: ah, never mind then Dec 27 23:33:22 have fun! Dec 27 23:33:27 c7x0 doesn't come out suspend with 2.6.23, neither does poodle :/ Dec 27 23:33:33 Crofton|work: enjoy! :) Dec 27 23:33:47 RP: c7x0? Dec 27 23:34:02 ah, zaurus Dec 27 23:49:34 RP: Btw, I noticed that the poodle section in zaurus-reboot-r3 is missing the code to toggle the LED to switch between halt and restart. I tried a quick hack, but the LED driver still baffles me :( Dec 27 23:50:49 03likewise 07org.oe.dev * rf3e47f94... 10/ (1 packages/linux/linux_2.6.23.bb): linux-2.6.23: Added patches for .12, CFS-v24.1 backport, and squashfs-lzma. Dec 27 23:51:40 anyone know which option on the maitenence screen is backup restore? Dec 27 23:55:15 time to go home. Dec 28 01:31:44 Yay, poodle suspend/resume working under 2.6.23 :) Dec 28 01:31:50 * RP -> Zzzz Dec 28 01:38:46 03rpurdie 07org.oe.dev * r3806d889... 10/ (1 packages/linux/linux-rp_2.6.23.bb): linux-rp-2.6.23: Fix suspend/resume problems for pxa25x devices. Fix poodle LCD resume issue with a temp hack until a proper patch can be created. Make 2.6.23 the default for poodle **** ENDING LOGGING AT Fri Dec 28 02:59:56 2007