**** BEGIN LOGGING AT Sun May 20 02:59:57 2007 May 20 09:10:40 I have another PPC issue. May 20 09:10:48 | ccache gcc -isystem/data/oetmp/staging/ppc-linux/include -O2 -DLINUX_PPDEV -o ppt.o -c ppt.c May 20 09:10:49 | ppt.c:8:20: error: sys/io.h: No such file or directory May 20 09:11:18 This file doesn't exists on FC6 on PPC. May 20 09:12:52 hi all May 20 09:15:59 holtmann: that is a -native packages that fails? May 20 09:16:14 sjf2410-linux-native-20060807-r1 May 20 09:17:29 http://packages.debian.org/cgi-bin/search_contents.pl?word=sys%2Fio.h&searchmode=searchfilesanddirs&case=insensitive&version=stable&arch=powerpc debian lacks, sys/io.h as well May 20 09:17:53 I have it in Ubuntu Feisty, but not on Fedora Core 6. May 20 09:18:11 And glibc-headers and kernel-headers plus -devel is installed. May 20 09:18:14 holtmann: can I talk you into using quilt? e.g. changedir into the source directory and call ../../temp/run.do_compile May 20 09:18:50 Tell me the commands. May 20 09:19:53 holtmann: remove the sys/io.h from ppt.c and see if it compiles May 20 09:20:02 holtmann: quilt new make-itcompile.patch May 20 09:20:05 holtmann: quilt edit ppt.c May 20 09:20:19 From which directory? May 20 09:20:26 holtmann: compile and two outcomes are possible (but you know that) May 20 09:21:07 holtmann: let me guess /data/oetmp/work/ppc-linux/sjf2410-linux-native-20060807-r1/sjf2410-linux May 20 09:21:57 holtmann: I assume it will compile as ioperm is #ifdefed May 20 09:24:20 It compiles. So this packages needs to check for sys/io.h and then have an ifdef. May 20 09:24:58 What next with quilt. I never used it. May 20 09:25:23 holtmann: well, it doesn't need sys/io.h at all May 20 09:25:28 holtmann: now you can do quilt refresh May 20 09:25:39 holtmann: and you will find a patch in patches/ send it to harald May 20 09:26:37 Okay. Can you put it into OE anyway. May 20 09:26:45 sure May 20 09:30:53 feel free to pastebin the patch May 20 10:22:51 psokolovsky, hello May 20 10:23:02 Hi! May 20 10:24:29 , Look at this http://cmh.hackndev.com/pastebin/41/ , please. Is it bug? May 20 10:25:48 "rebuild and clean " was not help resolving problem May 20 10:26:00 step2back: iirc, there was some talk about lirc problems. anyway, you shouldn't need lirc to build a normal image. May 20 10:27:44 hmmm... "bitbake opie-image" stop building at this. Hot to leave out this package? May 20 10:29:15 step2back: search what pulls it in May 20 10:29:58 step2back: I haev to admit I didn't build opie-image for a week, it indeed could break. will start builing soon. May 20 10:30:02 and? May 20 10:31:55 Should I wait new "revision" by you? How can I understand ready for building opie-image? May 20 10:32:44 step2back: no, you should not wait, should try to solve problem yourself, and report via bugtracker. May 20 10:33:24 you said that exist some package manager. Where is it? May 20 10:34:06 I mean "how to leave out it"? May 20 10:36:20 step2back: leave package manager out? it's completely senseless May 20 10:40:10 You said that not need to me (opie-image)! May 20 10:44:07 step2back: package manager is important, basic part of any linux-based distro May 20 10:46:27 Oh... I think (in past) that you said about manager of "org.openembedded.dev/packages". May 20 10:50:32 step2back: no, manager of binary installable packages. May 20 10:51:47 Так простой выход для меня сделать обрезание lirc из репозитория? May 20 10:57:42 step2back: here, don't use non-english for sure ;-) May 20 10:58:04 step2back: yes, just try to find out what recipe pulls it in, and comment out for now. May 20 10:58:10 I can't say it in eng May 20 10:58:45 step2back: use pm please May 20 10:58:59 bb in 15min May 20 10:59:14 What is a reason for save defconfig in oe.monotone? May 20 11:10:51 step2back: so OE-based distros have well-known, deterministic config, under tehir control. May 20 11:14:15 there are wrong defconfig for palmld. I cuted from *.bb using default defconfig. I think it's better to use defconfig from svn fro H&D. May 20 11:19:45 step2back: no. please submit correct defconfig for OE. yes better, use defconfigman. svn/cvs defconfigs can't be used by exactly that reason - they can be wrong at any time. May 20 11:22:37 how? May 20 11:26:47 Connecting to ftp.rpm.org|209.132.176.171|:21... failed: Connection refused. May 20 11:26:47 NOTE: Task failed: Fetch failed: ftp://ftp.rpm.org/pub/rpm/dist/rpm-4.1.x/popt-1.7.tar.gz May 20 11:26:49 :( May 20 11:28:10 holtmann: you can d/l it manually from anywhere, and put in you d/l directory. May 20 11:29:08 What happened to the idea of a big download server that mirrors all used files. May 20 11:29:55 for familiar it's exist. May 20 11:31:06 Do I have to create a md5 or something after downloading the file and put it into the sources directory. May 20 11:32:29 holtmann: no, jsut d/l it, that's all May 20 11:32:53 holtmann: as for mirror, there're mirrors, of various up-to-date status. May 20 11:33:21 I have all the source in /data/sources, but putting it there iseems not enough. Creating the *.md5 is doing the trick. May 20 12:12:12 03rpurdie * r822 10/ (4 files in 4 dirs): runqueue.py: Fix bug when target was in ASSUME_PROVIDED (#2236) May 20 12:16:03 RP: got a minute ? want to ask something about some patches in your kernels May 20 12:16:30 ifaistos: feel free to ask, if I haven't time I will answer later May 20 12:16:48 RP: OK May 20 12:18:57 RP: basically i was looking at the lzo_ series of patches, and especially the lzo_jffs2 and lzo_kernel May 20 12:19:17 ifaistos: ok May 20 12:20:11 RP: trying to see if i need to configure lzo compression of jffs2 i could not find any option ? does it work by default or am i blind ? May 20 12:20:36 ifaistos: There should be an option for it May 20 12:21:10 RP: though so... probably its me been blind (again) May 20 12:21:47 ifaistos: CONFIG_JFFS2_LZO May 20 12:21:58 RP: also for the lzo kernel image does it build by "default" in OE or do i need to inherit something ? (haven't look at it with detail to tell you the truth) May 20 12:22:43 ifaistos: You mean does it build lzo jffs2 images? May 20 12:22:54 yes May 20 12:23:11 ifaistos: No, it doesn't. The trick is to remove the "-x lzo" on the jffs2 commandline May 20 12:23:24 (it means exclude lzo) May 20 12:23:46 how is it in terms of compression with jffs2 ? May 20 12:24:10 ifaistos: I found it to give ~5% less compression than zlib May 20 12:25:23 RP: so does it make sense to use it ? I have a machine with limited flash (2MB) and i am trying to squize things as much as possible May 20 12:25:39 ifaistos: Maybe not in your case. The point of it was speed May 20 12:25:48 Its much faster than zlib May 20 12:26:52 RP: speed is an issue also but size is a bigger one... at least in my case May 20 12:27:07 ifaistos: Like everything, its a compromise May 20 12:28:19 RP: Richard thanks ! May 20 12:30:12 03rpurdie 07org.oe.dev * r8f7fdc42... 10/ (1 packages/clucene packages/nonworking/clucene): clucene: Move to nonworking. its cvs doesn't exist any more May 20 12:57:04 :)) mtn: renaming packages/clucene to packages/nonworking/clucene May 20 13:30:40 03rwhitby 07org.oe.dev * r69c4f0ca... 10/ (5 files in 4 dirs): ixp4xx-kernel: Updated to 2.6.21.1 May 20 13:31:18 hi May 20 13:37:20 koen: can we remove gqview from x11-image? It's not compiling May 20 13:37:28 !oebug 2336 May 20 13:37:30 * * Bug 2336, Status: NEW, Created: 2007-05-18 11:50 May 20 13:37:31 * * pinguozz(AT)netsons.org: fail compile gqview May 20 13:37:32 * * http://bugs.openembedded.org/show_bug.cgi?id=2336 May 20 13:39:43 . do you leave out lirc from packages? May 20 13:40:11 step2back: I'm compiling the default a-x11-image May 20 13:41:01 lirc has not been built for my machine (yet?) May 20 13:42:08 hmmm.... I feel my self like a one of universe who want to build opie-image May 20 13:42:36 Servers that are down or not responding are a pain in the ass if you are building from scratch. May 20 13:42:38 My is Palm LifeDrive. not build too. May 20 13:47:11 holtmann: that's drawback of opensource. and not only that. for example, one sweet day, you may find that your OS' hardware support stack suddenly started to depend on some GUI-spawned userspace bloat, without offering backwards compatibility with what it did previously ;-E May 20 13:48:43 I think that a really huge farm of mirrors dedicated for OE would solve this problem. May 20 13:49:37 holtmann : Thinking about donating one to OE ? May 20 13:50:02 If I had the money, I actually would. May 20 13:50:38 :) May 20 13:51:56 ~seen RP May 20 13:51:58 rp is currently on #handhelds #oe #openmoko #openzaurus. Has said a total of 11 messages. Is idling for 1h 24m 51s, last said: 'ifaistos: Like everything, its a compromise'. May 20 13:54:34 ~ping RP May 20 13:54:36 pong RP May 20 13:58:08 My main issue currently is that downloads.sourceforge.net points to kent.dl.sourceforge.net for me right now. And that server is terrible slow. May 20 13:59:27 holtmann, You should be able to specify SF dl server. May 20 13:59:43 holtmann: grep -Hrn for SOURCEFORGE_MIRROR and replace it May 20 14:01:22 * Bernardo is back. May 20 14:03:15 hi cedricb May 20 14:04:56 RP: hi May 20 14:06:17 i'm just trying to test sound (alsa) with mplayer (fbdev) and I'm getting nothing on the jack... May 20 14:06:59 cedricb: Try changing the controls in alsamixer May 20 14:09:01 RP: ok I've got master to green, jack function to headphone May 20 14:11:07 Hi All! Just downloaded bitbake from SVN and not able to run it due to the following error: May 20 14:11:10 ERROR: Unable to open conf/bitbake.conf May 20 14:11:19 RP: and no sound May 20 14:11:38 RP: any config file I can send u? May 20 14:11:53 strace bitbake shows, stat bitbake searches for bitbake.conf in /usr... May 20 14:13:31 RP: speaker is off May 20 14:14:01 cedricb: I haven't got a poodle handy atm :/ May 20 14:14:39 cedricb: There is some control that you need to make sure is activated, I can't remember the name off the top of my head May 20 14:15:05 RP: I'm using the mad library with alsa May 20 14:15:22 Ah, I don't think that works May 20 14:15:31 ah ok May 20 14:15:48 so which one should I use then? May 20 14:16:26 cedricb: oss works iirc May 20 14:18:03 nope, oss+mad -> nothing May 20 14:19:04 Have you set "Output Mixer HiFi Payback Switch" ? May 20 14:21:05 I can't set it on... May 20 14:21:38 cedricb: You should be able to :/ May 20 14:22:15 with which key? up/dow nothing May 20 14:22:23 cedricb: space? May 20 14:22:27 s/dow/down May 20 14:22:33 cedricb: maybe 'm' May 20 14:22:43 yes May 20 14:25:39 RP: thx work -> mad+alsa May 20 14:25:47 with that switch May 20 14:26:15 cedricb: great :) May 20 14:26:16 did u have any chance to read my email? :-) May 20 14:26:31 Which email was that? May 20 14:27:17 email about SDL May 20 14:28:11 cedricb: Ah, its ended up in my to deal with queue :/ May 20 14:28:41 :-) May 20 14:31:47 morning May 20 14:35:54 cedricb: I've replied May 20 14:38:37 RP: interested in kernel bugs on akita wake from suspend? May 20 14:38:59 Bernardo: Yes May 20 14:39:24 I got a interesting "unbalanced IRQ 31 wake disable"at kernel/irq/manage.c: 167 May 20 14:39:41 I'll try to output dmesg to a file and copy it here to the pc May 20 14:39:51 Bernardo: Is it reproducable? May 20 14:40:43 so far it happened at least twice on a angstrom opie-kdepim-image I generated today May 20 14:40:53 I think it also happened before May 20 14:41:48 I'm not sure if it happens every time or just the first suspend since bootup May 20 14:42:52 http://pastebin.ca/498547 May 20 14:44:04 Bernardo: Ah, its something using /dev/rtc May 20 14:44:28 I think I only suspended this akita twice since boot, so it happens every suspend/resume May 20 14:44:29 Bernardo: That explains why I haven't seen it since I don't have much using that device... May 20 14:44:58 yes, it looks like a rtc problem May 20 14:45:01 Bernardo: It looks like the zaurus suspend/resume code is conflicting with the kernel rtc driver... May 20 14:45:50 that might explain some of the suspend/resume problems people have been complaining with kernel 2.6.x on the akita May 20 14:46:27 Bernardo: I doubt it would stop the suspend/resume, its just a BUG_ON May 20 14:47:40 so it wouldn't be responsible for the crashes during resume some people complained of? May 20 14:48:01 Bernardo: I doubt it May 20 15:01:58 * * OE Bug 2184 has been RESOLVED (DUPLICATE) by oe@hentges.net May 20 15:01:59 * * [Bug 2184] gqview-2.1.1 fails to build error in May 20 15:02:00 * * http://bugs.openembedded.org/show_bug.cgi?id=2184 May 20 15:10:09 good day all May 20 15:18:21 RP: thx for the reply May 20 15:18:52 gm May 20 15:19:08 RP, are you the buildbot guru? May 20 15:19:31 Crofton: I use buildbot with poky May 20 15:19:40 cedricb: Hopefully it makes sense? May 20 15:19:59 can you have buildbot install stuff on the system, such as libraries May 20 15:20:09 so other branches on svn can link against them? May 20 15:21:00 Crofton: I don't understand? May 20 15:21:17 I have pieces of a larger project in svn May 20 15:21:18 03ifaistos 07org.oe.dev * r13698b01... 10/ (1 packages/aiostress/aiostress_0.0.bb): packages/aiostress/aiostress_0.0.bb: Change URI as the file has disappeared from the Suse ftp May 20 15:21:31 I start with some libraries that I want to build and install May 20 15:21:47 then rebuild programs that depend on these libraries May 20 15:22:06 Crofton: You should be able to make it do that, yes May 20 15:22:41 Crofton: buildbot is just a task exectutor, a bit like bitbake in that sense. It depends if you can write a script to do what needs to be done May 20 15:22:47 ok May 20 15:24:28 zecke: ping May 20 15:26:25 How to make bitbake OOM == touch test_test_test_0.1.bb :/ May 20 15:27:48 ~lart automatic emails from cycling teams May 20 15:27:49 * ibot chops automatic emails from cycling teams in half with a free AOL CD May 20 15:45:02 hmmm Using Angstrom distro (althow I have tryed also with generic) I am not able to compile anything... The output is just folllowing: May 20 15:45:04 NOTE: Tasks Summary: Attempted 0 tasks of which 0 didn't need to be rerun and 0 failed. May 20 15:45:43 Can someone help me in this problem? May 20 15:46:28 DeeKey: well, paste the maningul bits, or better read them! May 20 15:46:38 DeeKey: You have either a wrong BBPATH or a wrong BBFILES... May 20 15:48:45 Suppose we want to support building kernels with builtin initramfs: initramfs depends on kernel-modules depends on kernel depends on initramfs. May 20 15:49:05 zecke: "only" 30 bitbake bugs :) May 20 15:49:14 RP: I see you are cleaning up May 20 15:49:16 sorry, bugs + feature requests May 20 15:49:27 !oebug 1148 May 20 15:49:28 * * Bug 1148, Status: RESOLVED (WONTFIX), Created: 2006-07-08 00:42 May 20 15:49:29 * * papercrane(AT)reversefold.com: Segfault when not using the shell May 20 15:49:30 * * http://bugs.openembedded.org/show_bug.cgi?id=1148 May 20 15:49:43 well, we can't fix Gentoo May 20 15:49:55 It was bitbake 1.4 May 20 15:50:19 RP: still, bitbake1.4 doesn't segfault. Unless the python interpreter segfaults due miscompilation May 20 15:50:47 zecke: ok, the resoltuion is still valid as we don't have a CANTFIX :) May 20 15:51:02 :) May 20 15:51:13 !oebug 1488 May 20 15:51:15 * * Bug 1488, Status: NEW, Created: 2006-10-14 16:54 May 20 15:51:16 * * likewise(AT)gmx.net: Treat --ignore-deps argument as regexp May 20 15:51:16 * * http://bugs.openembedded.org/show_bug.cgi?id=1488 May 20 15:51:47 Got a patch to go with that, but it probably does not apply to any recent bitbake's... Care if I recreate the patch? May 20 15:52:11 likewise: Half that bug was fixed - bitbake 1.8 doesn't show ERRORs when generating dep graphs now May 20 15:52:31 likewise: If you have a recent patch, I might consider applying it :) May 20 15:52:37 RP: ok, only the feature remains then (regex support for ignores)? May 20 15:53:08 likewise: yes May 20 15:53:31 RP: ok, I'll consider refreshing it :-) May 20 15:53:57 * zecke|study continious filling his brain with math... May 20 15:55:28 zecke|study: Suppose there are five pirates who have to divide a loot of 100 gold pieces... May 20 15:55:58 * * OE Bug 2345 has been created by bero@arklinux.org May 20 15:55:59 * * cups 1.2.10 fails to build May 20 15:56:00 * * http://bugs.openembedded.org/show_bug.cgi?id=2345 May 20 15:56:18 I love bero May 20 15:56:55 zecke|study: Did you see my comment on the bitbake OOM bug btw? May 20 15:57:12 RP: which one that test_0.1.bb doesn't trigger it? May 20 15:57:30 zecke|study: Yes, do we support multiple underscores in PN? May 20 15:57:39 zecke|study: I know we don't in the package name output May 20 15:57:43 RP: well, then it is time inspect the PN in conf/bitbake.conf May 20 15:57:59 sorry, we don't in PV, not sure about PN :/ May 20 15:58:05 RP: well, I think we should support them in filenames. May 20 15:58:36 hi, May 20 15:58:36 I guess I'm switching from the old ARM ABI to EABI when I change my oe DISTRO variable from "generic" to "angstrom-2007.1", will it be necessary to do some cleanup before running "bitbake" to avoid mixing with already build ipks or will it rebuild everything automatically? May 20 15:58:38 PN = "${@bb.parse.BBHandler.vars_from_file(bb.data.getVar('FILE',d),d)[0] or 'defaultpkgname'}" May 20 15:59:16 AlGe: I'd start a new tmp dir to be honest, everything will need to rebuilt May 20 15:59:28 ok, thanks May 20 16:00:19 zecke|study: That function looks scary :) May 20 16:00:47 zecke|study: exp = 3 - len(parts) May 20 16:00:58 RP: scary == infinite loop? May 20 16:01:24 RP: you might want to improve your python foo by writing a unit test for it? May 20 16:01:38 RP: take a look at data.py on how to do it (doctest) May 20 16:01:46 zecke|study: don't think the problem is there, there is just a problem I noticed May 20 16:02:12 (axon_fpga_bitstream_syhq2 becomes axon_fpg) May 20 16:02:14 RP: well, either the parser chokes on empty (0 byte files), or on parsing the filename May 20 16:02:39 and pV gets? May 20 16:04:24 hmmm the existing apt in OE seems to have been replaced in Debian mirrors by a newer versions May 20 16:04:40 should we move it to unsuported ? May 20 16:05:47 * Bernardo is away: Ausente por agora. May 20 16:05:57 zecke|study: It is that function, s/3/10/ and it at least returns... May 20 16:07:15 bbl :} May 20 16:07:55 stelios_: That happens a lot, we should probably just upgrade it May 20 16:08:35 RP: De default patches fail if i upgrade but haven't looked why May 20 16:24:25 03rpurdie * r823 10/ (6 files in 6 dirs): BBHandler.py: Raise ParseError for filenames with multiple underscores instead of infinitely looping (1.6, 1.8, trunk) May 20 16:30:52 beecrypt.bb has a copyright notice and no licence is that correct ? May 20 16:35:29 03rpurdie * r824 10/ (6 files in 6 dirs): cooker.py: Fix invalid regexp in BBMASK error handling (missing import) (#1124) May 20 17:16:14 gm m May 20 17:16:16 gm mickeyl May 20 17:17:16 hey May 20 17:20:44 * Crofton curses koen May 20 17:21:15 I should be working out register values for a DDR2 controller, but instead I am looking at project Euler May 20 17:21:31 which he pointed out yesterady :) May 20 17:34:50 Crofton what is it? May 20 17:37:57 www.project-euler.net May 20 17:38:33 www.projecteuler.net May 20 17:38:51 hmm website is sick now May 20 17:39:15 http://projecteuler.net/index.php?section=about May 20 17:40:20 Crofton thanks ... May 20 18:09:33 hi mickeyl May 20 18:18:08 03likewise 07org.oe.dev * rf929edf1... 10/ (1 packages/hdparm/hdparm-6.9/uclibc.patch): hdparm: Fix building against uclibc. May 20 18:18:36 likewise: is libiconv failing for you as well? May 20 18:18:54 zecke|study: insane-wise, I think yes. May 20 18:19:23 likewise: good, I'm working on a solution but m4/lib-link.m4 is giving me a hardtime May 20 18:19:46 let me look along by building it from scratch. May 20 18:19:49 likewise: neither --enable-relocatable nor --disable--rpath really work... May 20 18:20:10 likewise: --disable-rpath=yes might be try but libtool is totally stupid May 20 18:20:35 the tools get built on do_install and it doesn't honor prefix but hardcodes DESTDIR+prefix in the rpath May 20 18:20:43 and it is doing it on purpose... May 20 18:21:29 zecke|study: on purpose? why? May 20 18:27:57 I don't know May 20 18:37:23 | + ipkg-cl -f /data/oetmp/work/fic-gta01-linux/openmoko-devel-image-1.0-r0/temp/ipkg.conf -o /data/oetmp/rootfs update May 20 18:37:23 | /data/oetmp/work/fic-gta01-linux/openmoko-devel-image-1.0-r0/temp/run.do_rootfs.4970: line 299: ipkg-cl: command not found May 20 18:37:23 NOTE: Task failed: /data/oetmp/work/fic-gta01-linux/openmoko-devel-image-1.0-r0/temp/log.do_rootfs.4970 May 20 18:37:36 Any idea why this fails? May 20 18:37:51 holtmann: It didn't build ipkg-native for some reason May 20 18:41:53 holtmann: org.openembedded.dev/classes/rootfs_ipkg.bbclass should have PACKAGE_INSTALL += "ipkg ipkg-collateral" May 20 18:43:58 likewise: I'm wondering if this is that bitbake 1.6 inherits_cache bug? May 20 18:45:54 Building ipkg-native and then building the image seems to work. However I have no idea how to fix it. May 20 18:46:23 hmm, are you building multithreaded bitbake? May 20 18:46:48 building *with a* multi... May 20 18:46:57 holtmann: I think its a bitbake 1.6 bug (you're using openmoko?) May 20 18:47:59 No. I am using the OE, but I am building the openmoko image. May 20 18:48:14 holtmann: ok, which bitbake version are you using? May 20 18:48:32 However itBitBake Build Tool Core version 1.6.6, bitbake version 1.6.6 May 20 18:48:37 The one that comes with Fedora Core 6. May 20 18:49:39 That version of bitbake is broken and your either need to patch it (patches have been on the mailing list), update to the latest bitbake svn on the 1.6 branch or switch to the 1.8 series May 20 18:49:39 However finally it seems I got an image. May 20 18:50:04 We haven't made a 1.6 release with the fix in yet May 20 18:50:23 This bug is not that a big problem, but might be worth to fix it and do a new release. May 20 18:50:54 Mickeyl are you there ? May 20 18:50:56 I applied the fix to svn head yesterday, a release might follow once I finish working through a backlog of bug reports ;-) May 20 18:52:10 * RP -> afk May 20 18:55:38 03likewise 07org.oe.dev * r4f6fcc60... 10/ (1 packages/uclibc/uclibc-0.9.29/angstrom/uClibc.distro): uclibc: 0.9.29/angstrom/uClibc.distro needs UCLIBC_SUSV3_LEGACY=y for perl May 20 18:58:27 So since I have my first home build openmoko image, how to I tell it to use bluez-utils-3.10.1 instead of bluez-utils-3.9. May 20 19:10:45 holtmann: remove the DEFAULT_PREFERENCE = "-1" line from bluez-utils_3.10.1.bb May 20 19:14:16 holtmann: fresh packages get a lower (negative preference) so that only knowledgable users get to discover new bugs May 20 19:14:50 holtmann: so you have now officially applied for beta testing ;-) May 20 19:20:46 I know. Looks good so far. Had to hack around the dependencies a little, but it seems the new image is building now. May 20 19:21:01 monotone.openembedded.org is very slow today... May 20 19:28:32 03ifaistos 07org.oe.dev * r443963b3... 10/ (1 conf/checksums.ini): conf/checksums.ini : Fix boost-asio checksums May 20 20:05:51 stelios_: ping May 20 20:06:07 likewise : hey May 20 20:06:29 stelios_: hey, I'm tracing a set of bugs similar to yours: #2001 May 20 20:06:48 stelios_: do you happen to know which kernel header version that was? May 20 20:08:02 http://bugs.openembedded.org/show_bug.cgi?id=2001 May 20 20:08:35 likewise : hmmm let me check May 20 20:11:21 likewise : it was either 2.6.17.9 or 2.6.20 (but i think it was 2.6.17.9) May 20 20:11:52 stelios_: these errors are related to 2.6.18+ May 20 20:12:17 stelios_: and have to do with the removal of syscall macro's. May 20 20:12:50 likewise : there a re a few more apps that failed with similar errors May 20 20:13:00 stelios_: I have some patches in the queue, although I'm not 100% sure they will work for everyone (old headers, new headers, glibc/uclibc) May 20 20:13:15 stelios_: so far, I found vsftpd May 20 20:14:11 likewise : have to look at the bugs i have filed. can't remember which other(s) May 20 20:21:15 likewise : if i can remember the other apps were also network related and had pcap (if iremember correctly) failing butgoogling around i found it was also a syscall issue May 20 20:23:42 hhmm... Does anyone know how to fix this error? ERROR: No buildable providers available for required build target meta-maemo May 20 20:25:09 stelios_: I will post patches to the bugs, as I'm not fully confident yet. May 20 20:27:11 likewise : nice. i'll try them out and let you know. i am creating a oplinux-packages.bb file and has a long list of apps that don't compile... vsftpd is one of them... May 20 20:27:34 stelios_: are you using uclibc or glibc? May 20 20:27:58 * * OE Bug 2001 has been created by likewise@gmx.net May 20 20:27:59 * * libcap fails to build May 20 20:27:59 * * http://bugs.openembedded.org/show_bug.cgi?id=2001 May 20 21:03:53 zecke: Have you a moment to comment on #1110? May 20 21:04:09 !oebug 1110 May 20 21:04:10 * * Bug 1110, Status: NEW, Created: 2006-06-20 06:49 May 20 21:04:11 * * yamaken(AT)bp.iij4u.or.jp: env variable MAKEFLAGS should not be propagated into bitbake subprocesses May 20 21:04:12 * * http://bugs.openembedded.org/show_bug.cgi?id=1110 May 20 21:04:46 zecke: Would you say its the user's responsibility to provide a sane environment? May 20 21:04:58 I like comment #7 May 20 21:05:16 :) May 20 21:05:36 RP: well, it doesn't matter what I believe (I still think some people would have a better life using a typewriter) May 20 21:05:51 we have seen this with MACHINE breaking binutils, MAKEFLAGS breaking stuff May 20 21:06:10 zecke: The difference is that MACHINE is a variable we use, MAKEFLAGS is not May 20 21:06:33 we do use MAKEFLAGS (take a look at the scripts we generate) May 20 21:07:01 RP: I think we should allow to filter variables from our initial config May 20 21:07:33 zecke: We do allow it through the unexport option May 20 21:08:15 zecke: There is a difference between allowing the user to set something like MACHINE (and hence needing an unexport) and having users random variables breaking bitbake May 20 21:08:20 we should add an unexport for MAKEFLAGS then May 20 21:08:42 zecke: Can you see the difference between MACHINE and MAKEFLAGS? May 20 21:09:01 RP: yes, both start with M,A but then there is a difference May 20 21:09:23 RP: but I tend to agree (now), the user can unexport them in site.conf May 20 21:09:23 * RP sighs May 20 21:09:51 I don't think having a policy of randomly unexporting a set of variables is good May 20 21:10:04 Who knows what we'll be asked to unexport next :/ May 20 21:10:16 Perhaps we should unexport everything by default? May 20 21:10:25 only allow things marked as export May 20 21:10:57 this will give us the same issue May 20 21:11:09 should we export LD_LIBRARY_PATH (influecing gcc), or SVN_SSH May 20 21:11:31 let us ask the user to a) have a clean environment or b) clean it in site/auto.conf May 20 21:11:42 ok, agreed May 20 21:11:58 So I can close that suggesting the user unexports in site.conf May 20 21:16:09 zecke: Thanks (it does matter what you believe) May 20 21:30:57 * * OE Bug 1110 has been RESOLVED (FIXED) by rpurdie@rpsys.net May 20 21:30:58 * * [Bug 1110] env variable MAKEFLAGS should not be May 20 21:30:59 * * http://bugs.openembedded.org/show_bug.cgi?id=1110 May 20 21:33:18 likewise: I suck, sorry May 20 21:33:35 koen: ?? May 20 21:33:45 koen: good evening May 20 21:34:01 I was in Reusel today, but we left the restaurant to late to pick up the server at gerwins place May 20 21:34:08 good evening all May 20 21:34:34 ah, well, I haven't heard from gerwinin yet. I'm not sure if he's at home often. May 20 21:34:49 and I zoab May 20 21:34:50 koen: any good restaurant? May 20 21:35:07 "de wekker" in reusel is a pretty good restaurant May 20 21:35:21 4 course menu for 28 euros May 20 21:36:16 koen: my brother has worked on zowab (zeer open waterafvoerende asfalt beton), which does *not* ice during winters. There still running tests on small high ways strips with that. May 20 21:37:19 koen: it is deployed with a natural heath capacitor (in underground layers), so that the highways heatens the water during summers, then de-ices the road during winters (from the underground water heat capacitor) May 20 21:37:29 cool May 20 21:37:36 or rather, "hot" May 20 21:37:56 koen: that about *all* I know about it. Now he works at Nefit/Deventer designing heating boilers :-) May 20 21:38:19 I think both markets will plummet with global warming :-) May 20 21:42:09 * RP looks at #2101 and cries May 20 21:42:16 2108 May 20 21:42:26 !oebug 2108 May 20 21:42:27 * * Bug 2108, Status: NEW, Created: 2007-04-13 07:15 May 20 21:42:28 * * florian.boor(AT)kernelconcepts.de: Dependency error lacks information May 20 21:42:29 * * http://bugs.openembedded.org/show_bug.cgi?id=2108 May 20 21:43:22 http://www.pastebin.ca/499114 is why May 20 21:43:47 This is why it doesn't just print a note all the time, some are "known" to fail and thats acceptable May 20 21:44:20 hehe May 20 21:44:24 arm-angstrom-linux-gnueabi-gcc-2.95 May 20 21:44:32 I feel dirty just by looking at that May 20 21:49:51 hhe May 20 21:50:05 I managed to speak to the DDR2 Ram on the sffsdr today May 20 21:50:10 * koen looks at http://www.jeroenhoek.nl/Files/Angstrom/passer.png May 20 21:50:58 * * OE Bug 2347 has been created by likewise@gmx.net May 20 21:50:59 * * beware of upstream bug in uclibc 0.9.29 May 20 21:51:00 * * http://bugs.openembedded.org/show_bug.cgi?id=2347 May 20 21:51:16 holtmann: I'll look at the bluez(-gnome) stuff tomorrow May 20 21:51:20 * RP notes that 1.8 svn is broken as is trunk :-( May 20 21:54:16 * koen go ZZZZZ now May 20 21:55:35 sleep tight May 20 21:56:58 * * OE Bug 2348 has been created by likewise@gmx.net May 20 21:56:59 * * alsa-utils build fails against uclibc May 20 21:57:00 * * http://bugs.openembedded.org/show_bug.cgi?id=2348 May 20 21:57:39 Crofton: congrats. got your head out of the hole that is called libtool? ;-) May 20 21:58:02 Anyway, enough fixing and submitting new bugs, zzz now May 20 21:58:08 cu! May 20 21:59:31 * Crofton has been avoiding libtool May 20 22:49:40 03rpurdie * r825 10/ (2 files in 2 dirs): runqueue.py: Use correct variable for task length May 20 23:04:40 kergoth: why was france kick/banned? May 20 23:07:32 ~slap kergoth May 20 23:07:34 * ibot slaps kergoth, keep your grubby fingers to yourself! May 20 23:13:52 kergoth: Is something going on? May 20 23:16:58 * * OE Bug 2349 has been created by pmiscml@gmail.com May 20 23:16:59 * * Speed up ipkg-make-index few times, try 2. May 20 23:17:00 * * http://bugs.openembedded.org/show_bug.cgi?id=2349 May 20 23:19:39 03rpurdie 07bitbake-1.8 * r826 10/lib/bb/runqueue.py: runqueue.py: Don't run sanity checks on incomplete builds May 20 23:20:45 RP: ^^^ May 20 23:21:00 psokolovsky: I saw that, thanks :) May 20 23:21:16 RP: I get the feeling that one of the other children sat on kergoth's sand castle. May 20 23:21:56 mzb: I'm making no comments until I hear whats happened... May 20 23:22:11 fair enough May 20 23:50:33 03rpurdie * r827 10/ (4 files in 4 dirs): taskdata.py: Promote certain warnings from debug to note 2 level May 21 01:25:55 !oebug 2350 May 21 01:25:57 * * Bug 2350, Status: NEW, Created: 2007-05-20 18:25 May 21 01:25:58 * * r.farina(AT)adelphia.net: Aircrack-ng May 21 01:25:59 * * http://bugs.openembedded.org/show_bug.cgi?id=2350 May 21 01:26:11 version bump if someone would be so kind as to commit this May 21 01:26:58 * * OE Bug 2350 has been created by r.farina@adelphia.net May 21 01:26:59 * * Aircrack-ng May 21 01:27:00 * * http://bugs.openembedded.org/show_bug.cgi?id=2350 May 21 02:56:02 Hi anybody ... I am a OpenEmbedded newbie and would like to build a custom distro booting from a CompactFlash image for an carPC that features a VIA Epia MII-12000 board ... I am running into several problems .... I need some hand-holding! I am on the right channel? May 21 02:58:16 for example: the gcc 4.1.1 does not fetch because checksum does not match ... uclic-0.9.28 does not install (seems to be a makefile problem) ... etc etc **** ENDING LOGGING AT Mon May 21 02:59:57 2007