**** BEGIN LOGGING AT Tue Apr 14 02:59:57 2009 Apr 14 03:04:17 what's if programs that compiled with sysv? Apr 14 04:40:23 03Denys Dmytriyenko  07org.openembedded.dev * rb37579152e 10openembedded.git/packages/asterisk/ (6 files in 2 dirs): asterisk 1.4.23.1: move to recipes/ from accidentally commited packages/ Apr 14 04:54:52 !seen XorA Apr 14 04:54:54 XorA (n=XorA@94-193-93-235.zone7.bethere.co.uk) was last seen quitting from #oe 10 hours, 45 minutes ago stating ({"Leaving"}). Apr 14 04:55:16 anybody alive? Apr 14 04:55:51 anybody else seeing e2fsprogs breaking after http://cgit.openembedded.net/cgit.cgi?url=openembedded/commit/&id=d187eec071bb0afe8bda13cbf7b89170fe99e384 Apr 14 06:24:23 ok so what do i do when a bibake pkg fails? Apr 14 06:40:44 good morning Apr 14 07:13:24 oi Apr 14 07:24:48 morning Apr 14 07:31:41 03Koen Kooi  07org.openembedded.dev * r111c0c70c1 10openembedded.git/recipes/gnome/gnome-bluetooth_2.27.1.bb: gnome-bluetooth: add obex-data-server to rrecommends Apr 14 07:31:42 03Koen Kooi  07org.openembedded.dev * r7ac45fd7ee 10openembedded.git/ (conf/checksums.ini recipes/gnome/gvfs_1.2.1.bb): gvfs: update to 1.2.1 Apr 14 07:31:43 03Koen Kooi  07org.openembedded.dev * rd2282642ae 10openembedded.git/ (2 files in 2 dirs): obex-data-server: add 0.4.4 Apr 14 07:31:43 03Koen Kooi  07org.openembedded.dev * r1a9bfdf30e 10openembedded.git/ (3 files in 3 dirs): nautilus: add 2.26.1 Apr 14 07:31:46 03Koen Kooi  07org.openembedded.dev * rca723f8911 10openembedded.git/ (conf/checksums.ini recipes/obexftp/obexftp_0.23.bb): obexftp: update to 0.23 Apr 14 07:31:49 03Koen Kooi  07org.openembedded.dev * rac6ff675b7 10openembedded.git/recipes/gdbm/gdbm_1.8.3.bb: gdbm: use LDFLAGS (fixes QA error) Apr 14 07:31:52 03Koen Kooi  07org.openembedded.dev * r37b2a0e9e2 10openembedded.git/ (2 files in 2 dirs): gnome-packagekit: update to 2.27.1 Apr 14 07:31:57 03Koen Kooi  07org.openembedded.dev * r645e6805fc 10openembedded.git/recipes/gnome/gnome-bluetooth_2.27.1.bb: gnome-bluetooth: update to 2.27.3 Apr 14 07:32:00 03Koen Kooi  07org.openembedded.dev * r968b109556 10openembedded.git/ (conf/checksums.ini recipes/gnome/gnome-menus_2.26.0.bb): gnome-menus: update to 2.26.0 Apr 14 07:32:05 03Koen Kooi  07org.openembedded.dev * r18ef4a1270 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Apr 14 07:32:08 03Koen Kooi  07org.openembedded.dev * r7c9bef982a 10openembedded.git/ (conf/checksums.ini recipes/packagekit/packagekit_0.4.5.bb): packagekit: update to 0.4.6 Apr 14 08:11:31 03Tom Rini  07org.openembedded.dev * ra8ba77b61d 10openembedded.git/recipes/busybox/ (8 files in 2 dirs): Apr 14 08:11:31 busybox 1.13.2: Add more upstream patches, bump PR Apr 14 08:11:31 This adds the awk, killall, printf, syslogd, top24, unzip and wget patches Apr 14 08:11:31 from http://busybox.net/downloads/fixes-1.13.2/ Apr 14 08:11:31 03Marcin Juszkiewicz  07org.openembedded.dev * rf5fa15933e 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Apr 14 08:11:35 03Marcin Juszkiewicz  07org.openembedded.dev * r6382819648 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Apr 14 08:11:38 03Marcin Juszkiewicz  07org.openembedded.dev * r4e3ef28c36 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Apr 14 08:11:41 03Marcin Juszkiewicz  07org.openembedded.dev * r3b82e8b272 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Apr 14 08:11:44 03Marcin Juszkiewicz  07org.openembedded.dev * r0bc69c83a4 10openembedded.git/recipes/linux/ (4 files in 4 dirs): linux: fixed defconfigs for at91sam* devboards Apr 14 08:13:07 morning Apr 14 08:13:31 hi :-) Apr 14 08:14:12 * hrw needs uclibc mastah Apr 14 08:29:40 03Koen Kooi  07org.openembedded.dev * r01be9642c5 10openembedded.git/ (conf/checksums.ini recipes/gnome/gnome-icon-theme_2.26.0.bb): gnome-icon-theme: update to 2.26 Apr 14 08:29:50 03Koen Kooi  07org.openembedded.dev * rb579bdce5d 10openembedded.git/classes/gtk-icon-cache.bbclass: gtk-icon-cache bbclass: build caches for all icondirs in ${datadir}/icons Apr 14 08:29:51 03Koen Kooi  07org.openembedded.dev * r56fe494547 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Apr 14 08:31:39 khem: can you share which toolchain and which machine gave you uclibc/thumb working? it fails for me in stable/2009 for at91sam9263ek/angstrom-2008.1 Apr 14 08:48:10 florian: good morning Apr 14 08:48:15 good morning Apr 14 09:03:16 hello all, I'm trying to compile kernel for mini2440 (from wiki.linuxmce.com/index.php/Mini2440), I got problem when booting using the kernel. the board did not continue after displaying: Uncompressing Linux............................................................................................................................ done, booting the kernel. Apr 14 09:04:23 I'm using filesystem downloaded form andahammer.com, only the kernel that I re-compile. Apr 14 09:06:47 it is the file /kernel-bin/arch/arm/boot/zImage that I write to the board. is it correct? Apr 14 09:07:42 chen101: did you recompile the kernel using OE? Apr 14 09:20:35 hi, I recompile ke kernel directly. Apr 14 09:21:11 it is from git clone git://repo.or.cz/linux-2.6/mini2440.git Apr 14 09:22:54 chen101: strange Apr 14 09:23:01 chen101: you use u-boot? Apr 14 09:23:13 hrw: I'm interested to know which toolchain and which machine give you uclibc working Apr 14 09:23:59 mckoan: angstrom-2008.1 gives uclibc for avr32, armv7a, i586. for arm4/5 we need to set it to not use thumb Apr 14 09:24:00 I'm testing beagleboard with commit 31c5753ea1a29750060a002b867362354ae42669, and Angstrom distro but I have some bugs. Apr 14 09:24:31 recalcati_irc: stable/2009? Apr 14 09:24:33 I'm in stable/2009 branch Apr 14 09:25:03 hrw: I think no, how to make sure? Apr 14 09:25:38 before, I download the binary from andahammer.com (both kernel and qtopia). it works. then, I recompile ke kernel, but it can't boot. Apr 14 09:25:46 epiphany doens't start: Failed to create directory "/home/recalcati/.gnome2/epiphany" Apr 14 09:25:53 jkridner: hello Apr 14 09:26:00 if I write the kernel from andahammer.com again, it works. Apr 14 09:26:22 chen101: which bootloader it use? Apr 14 09:27:36 sorry, how to find out? Apr 14 09:27:40 midori starts, but has a lot of problem with gmail in full version, not html base Apr 14 09:28:32 midori in gmail can't write e-mail Apr 14 09:29:54 chen101: connect serial cable and reset device? Apr 14 09:30:07 chen101: then pastebin output from device Apr 14 09:30:27 chen101: ok, it use u-boot Apr 14 09:30:33 chen101: you need uImage not zImage Apr 14 09:30:34 I tried to debug it, but I have still problems in mapping shared libraries in gdb. I can't point to webkit files. I use"directory" command and "solib-absolute-prefix" Apr 14 09:30:57 chen101: or you have vivi bootloader which you can use to flash uboot Apr 14 09:31:06 chen101: all is described on page which you gave Apr 14 09:31:27 oh yes, I'm using vivi when transfer the kernel. it's just I don't modify the u-boot. Apr 14 09:32:02 firefox can send e-mail with gmail full version, but I need more cpu, so for example how to go 600Mhz in stable/2009 ? How can I do it ? Apr 14 09:32:10 I'll try to transfer the uImage then, and try reboot the device. Apr 14 09:32:22 recalcati_irc: ask on #beagle? Apr 14 09:32:39 * hrw -> coffee break Apr 14 09:33:28 hrw: you told me to ask here Apr 14 09:33:51 if you prefer I write a complete report in order not to disturb all you Apr 14 09:34:33 hrw: as you're the poky expert, could you explain a little bit the "overlay" mechanism? I'd like to be able to do it with OE too Apr 14 09:41:19 03Tick Chen  07org.openembedded.dev * ra2151a9e98 10openembedded.git/ (3 files in 3 dirs): madbutterfly: adding madbutterfly bb file Apr 14 09:43:40 hrw: hi, here is the pastebin( http://pastebin.com/m17be679 ) , if I transfer the uImage, it is said that it is not a compressed linux kernel image Apr 14 09:43:56 btw, I'm using dwn (windows based) for transferring the binary. Apr 14 09:44:06 dnw* Apr 14 09:45:38 anybody built x86 "bitbake x11-image" since now? Apr 14 09:46:07 if I transfer the zImage, here is the pastebin: http://pastebin.com/m40ed27a0 Apr 14 09:46:19 stop after displaying booting the kernel. Apr 14 09:50:21 chen101: if you use uImage you have to call it from u-boot bootloader Apr 14 09:50:41 chen101: if you use zImage you have to configure command line parameters properly Apr 14 09:51:12 oh.. Apr 14 09:51:17 chen101: the problem you are facing to is mostly caused by your bootloader Apr 14 09:51:55 chen101: moreover, id your S3C2440A based board a custom one or a Freescale one? Apr 14 09:52:25 mckoan: how to find out? Apr 14 09:52:53 chen101: which board is? where does it come from? Apr 14 09:53:38 the board is bought from here: http://www.csun.co.jp/SHOP/200812021.html Apr 14 09:54:00 it comes with old kernel and qtopia 1.7.0 Apr 14 09:54:21 then I download the binary kernel+qtopia from andahammer.com Apr 14 09:57:29 and since I want to customize the kernel (add iptables functionality and ppp), I recompile the kernel. Apr 14 09:58:26 mckoan: Freescale s3c24xx?????? Apr 14 09:58:49 looks more like a Samsung :) Apr 14 10:09:06 XorA: sorry my mistake :-D Apr 14 10:09:39 mckoan: careful there, you'll get the stock market excited :-D Apr 14 10:10:29 chen101: MINI2440 is a japanese custom product. where your bootloader and kernel come from? Apr 14 10:12:13 sorry, I don't know where it came from. but I have just downloaded the new binary from andahammer.com and put it inside the board. Apr 14 10:12:35 but I already backup the original, just in case if it is needed. Apr 14 10:13:17 is it different board? Apr 14 10:14:21 hello :) Apr 14 10:14:24 ReaperOfSouls: 11:22 < hrw> recalcati: oe ml Apr 14 10:14:26 ups Apr 14 10:14:28 recalcati_irc: 11:22 < hrw> recalcati: oe ml Apr 14 10:14:35 recalcati_irc: oe ml != #oe Apr 14 10:15:15 hi flameman Apr 14 10:15:20 recalcati_irc: scrivilo in mailing list ;-) Apr 14 10:15:36 hrw: oe ml http://wiki.openembedded.net/index.php/Mailing_Lists which one? Apr 14 10:15:41 methril|work: overlay is just another dir with classes/recipes/confs Apr 14 10:15:45 recalcati_irc: -devel Apr 14 10:15:56 mckoan, hrw : ok, thx Apr 14 10:16:03 methril|work: you add it into BBPATH and BBFILES Apr 14 10:17:09 hrw: yes, i know, i was guessing what happens if you write some recipes in that directory to make some changes in the original ones (that could not be in OE mainline) Apr 14 10:18:03 i am trying to use the irda of zaurus c1000 akita, not working at the moment, anyone experienced ? Apr 14 10:18:05 methril|work: read about bitbake collections. you can give priorities for them Apr 14 10:18:29 hrw: thanks Apr 14 10:49:03 hrw: I just send an email to openembedded-devel@lists.openembedded.org ? Once upon a time I tried to subscribe, but I didn't receive answer Apr 14 10:55:04 flameman: what you trying to do? Apr 14 11:00:25 Hello everybody! Apr 14 11:00:57 Hi Apu Apr 14 11:01:11 I sent now the subscription to openembedded-devel@lists.openembedded.org . Is there any chace to be accepted ? Apr 14 11:01:29 I'd like to contribute to stable/2009 Apr 14 11:01:54 first of all doing tests on beagleboard with angstrom Apr 14 11:13:39 03Koen Kooi  07org.openembedded.dev * rdd3d60afc9 10openembedded.git/recipes/gnome/gnome-bluetooth_2.27.3.bb: gnome-bluetooth: add fix for properties crash Apr 14 11:13:39 03Koen Kooi  07org.openembedded.dev * r2419aee437 10openembedded.git/recipes/angstrom/angstrom-gpe-task-base.bb: angstrom gpe task-base: gnome-bluetooth now does that gpe-bluetooth and bluez-gnome tried to do, so replace them Apr 14 11:13:40 03Koen Kooi  07org.openembedded.dev * r1360f47d9e 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Apr 14 11:15:16 03Koen Kooi  07org.openembedded.dev * r54aad17440 10openembedded.git/recipes/gnome/gnome-bluetooth_2.27.3.bb: gnome-bluetooth: remove debugging leftovers Apr 14 11:23:56 bye all. thanks.. Apr 14 11:38:45 jkridner: ping Apr 14 12:28:01 * Gnutoo think he will try uclibc Apr 14 12:28:06 hehe Apr 14 12:28:24 woglinde, hi Apr 14 12:28:41 woglinde, it's because of nptl...I've a 2.4 kernel Apr 14 12:29:03 oh Apr 14 12:29:04 hehe Apr 14 12:29:11 why this olde kernel? Apr 14 12:29:17 and arm? Apr 14 12:29:27 then you can test version 0.9.30.1 Apr 14 12:29:30 woglinde, yes arm+2.4.19 kernel Apr 14 12:29:48 with linuxthread instead of default linuxthread.old Apr 14 12:29:49 because it's a pma430 and has proprietary drivers... Apr 14 12:29:54 ok Apr 14 12:30:39 is TARGET_OS = "linux-uclibc" sufficent? Apr 14 12:30:48 nope Apr 14 12:30:50 + ANGSTROMLIBC = "uclibc" Apr 14 12:30:55 use angstroem Apr 14 12:31:03 and ANGSTROMLIBC = "uclibc" Apr 14 12:31:13 ok Apr 14 12:31:22 what is angstroem? Apr 14 12:31:30 hoe? Apr 14 12:31:40 ah ok typo Apr 14 12:31:55 DISTRO = "angstrom-2008.1" Apr 14 12:31:56 ANGSTROMLIBC = "uclibc" Apr 14 12:32:10 sorry that I can't type the "sonderzeichen" Apr 14 12:32:13 *g* Apr 14 12:33:03 ok Apr 14 12:33:03 hallo Apr 14 12:33:23 is there something I have to do for oe to take in account some changes on base.bbclass ? Apr 14 12:33:54 ptitjes send your patches to the oe-devel mailinglist Apr 14 12:34:12 if you send enough and good patches Apr 14 12:34:14 woglinde: sorry I was not clear :) Apr 14 12:34:22 someday you will get write account Apr 14 12:34:27 I meant so that bitbake takes my changes in account Apr 14 12:34:27 :D Apr 14 12:34:41 ??? Apr 14 12:34:54 they are used immedately Apr 14 12:35:00 I applied http://cgit.openembedded.net/cgit.cgi?url=openembedded/commit/&h=koen/darwin-host-fixes&id=28fcf1e8988f5e6ba8f05f39c94086f824c3d14f Apr 14 12:35:15 and it seems to not be taken in account as it still make the same error Apr 14 12:35:19 ptitjes, where are located your modified bbclass? Apr 14 12:35:34 openembedded/classes Apr 14 12:35:43 ok Apr 14 12:35:57 isn't that right ? Apr 14 12:36:27 normaly yes... Apr 14 12:36:31 woglinde: is there a way to flush the cache ? Apr 14 12:36:54 touch local.conf ? Apr 14 12:40:03 Gnutoo: so at last how looks like your local.conf? Apr 14 12:40:32 mckoan, i'v put things in a freepma distro Apr 14 12:40:44 Gnutoo: thanks Apr 14 12:40:50 but I'll pastebin it if you want Apr 14 12:40:52 Gnutoo: is it in OE tree ? Apr 14 12:40:57 no Apr 14 12:41:26 Gnutoo: any link please? Apr 14 12:41:27 I did it because some local.conf things weren't taken into account Apr 14 12:41:36 5s I finish my local.conf Apr 14 12:44:51 doesn't work yet: http://rafb.net/p/tfBmTJ60.html Apr 14 12:45:06 cause I forgetten TARGET_OS in my local.conf Apr 14 12:47:47 ah no no need it worked without Apr 14 12:48:19 Gnutoo: thx, I needed just to see ho to manage a uclibc build ;-) Apr 14 12:48:31 :) Apr 14 12:48:51 I've a trick if you're interrested: Apr 14 12:48:57 BBFILES = "" Apr 14 12:49:10 and so you don't have to wait the parse to see if it's ok Apr 14 12:49:57 so it print the oe build configuration faster: http://rafb.net/p/17u6vg98.html Apr 14 12:50:11 and after that restore the BBFILES Apr 14 12:50:16 and bitbake what you want Apr 14 12:55:05 Gnutoo: thank you, cool hint Apr 14 13:02:04 /nick cbrake Apr 14 13:02:32 hi Cliff Apr 14 13:03:34 hi hrw Apr 14 13:12:28 Hey, guys, could anyone tell me what was this fakeroot "no space left on device" issue, and how to solve it, again? Apr 14 13:15:03 Ainulindale: its not fakeroot got out of space Apr 14 13:15:27 Well there is plenty of space here :-) Apr 14 13:15:48 so give more info Apr 14 13:17:20 hrw: http://tinderbox.openembedded.net/public/logs/4354823.txt Apr 14 13:17:26 This isn't very informative, is it? :-) Apr 14 13:18:35 We had this issue once, although I can't remember how we solved it Apr 14 13:20:28 Ainulindale: no idea how to help ;( Apr 14 13:20:39 Well, thanks anyway :-) Apr 14 13:25:46 03Carsten Haitzler  07org.openembedded.dev * r2859cd17d7 10openembedded.git/conf/machine/palmt650.conf: palmt650: use palmt650-keyboard-modmap on t650 to make kbd work right Apr 14 13:25:47 03Carsten Haitzler  07org.openembedded.dev * rb7f7293617 10openembedded.git/recipes/e2fsprogs/ (3 files in 2 dirs): e2fsprogs: fix compile to use 3 args so it works. Apr 14 13:25:49 03Carsten Haitzler  07org.openembedded.dev * r660bc2c08f 10openembedded.git/recipes/e17/elementary-sms_svn.bb: elementary-sms: add recipe Apr 14 13:25:49 03Carsten Haitzler  07org.openembedded.dev * rb4ea283587 10openembedded.git/recipes/pointercal/files/palmt650/pointercal: pointercal: add pre-calibration for palmt650 Apr 14 13:25:51 03Carsten Haitzler  07org.openembedded.dev * re71ef2ab3c 10openembedded.git/recipes/palmt650/ (3 files in 2 dirs): palmt650-keyboard-modmap: keyboard map for palmt650 so it works right Apr 14 13:25:54 03Carsten Haitzler  07org.openembedded.dev * r6348c34c39 10openembedded.git/conf/distro/include/preferred-om-2008-versions.inc: preferred-om-2008-versions: use nfs-utils version that works Apr 14 13:25:57 03Carsten Haitzler  07org.openembedded.dev * rf5a114a9ee 10openembedded.git/recipes/xserver-kdrive-common/xserver-kdrive-common/30xTs_Calibrate: xserver-kdrive-common: fix tscalibrate script to not run if ts dev not exist Apr 14 13:25:58 Armageddon! Apr 14 13:26:03 03Carsten Haitzler  07org.openembedded.dev * rbde1b8002c 10openembedded.git/recipes/images/illume-image.bb: illume-image: update packages to be cleaner Apr 14 13:26:06 03Carsten Haitzler  07org.openembedded.dev * rb7555d7d3a 10openembedded.git/recipes/xserver-kdrive-common/ (xserver-kdrive-common_0.1.bb xserver-nodm-init/xserver-nodm): xserver-kdrive-common: fix start script to suport background opt and up pkgrev Apr 14 13:26:12 03Carsten Haitzler  07org.openembedded.dev * re57036c6b4 10openembedded.git/recipes/glibc/glibc_2.6.1.bb: Apr 14 13:26:15 glibc: write ld.so.cache to usual /etc location so it can be cached and speed Apr 14 13:26:17 up boot (so it doesnt need to re-generate every boot!) Apr 14 13:26:19 03Carsten Haitzler  07org.openembedded.dev * r1c1e92afe0 10openembedded.git/conf/distro/include/sane-srcrevs.inc: efl: update svnrev Apr 14 13:26:23 03Carsten Haitzler  07org.openembedded.dev * rd845d444ca 10openembedded.git/recipes/ (7 files in 2 dirs): efl: updates to have improved packaging and fewer... bugs! Apr 14 13:26:27 03Carsten Haitzler  07org.openembedded.dev * r631decbb48 10openembedded.git/recipes/xserver-kdrive-common/xserver-kdrive-common/ (Xserver openmoko/Xserver): xserver-kdrive-common: fix Xserver script to properly choose background PPM Apr 14 13:36:08 ouch... Apr 14 13:36:33 woglinde,hi again I've ld: ERROR: /home/embedded/oetmp_pma/cross/arm-oabi/lib/gcc/arm-angstrom-linux-uclibc/4.2.4/libgcc.a(_clz.o) uses hardware FP, whereas lib/libuClibc-0.9.30.1.so uses software FP Apr 14 13:36:56 Ainulindale: die oe.dev! die! Apr 14 13:36:57 :) Apr 14 13:37:20 raster nope Apr 14 13:37:23 raster: :-) Apr 14 13:37:40 gnutoo arm oabi with hardware fp? Apr 14 13:37:53 woglinde, yes worst case scenario...lol Apr 14 13:37:58 never saw such a proc Apr 14 13:38:01 woglinde, I want to be able to chroot Apr 14 13:38:04 hehehe Apr 14 13:38:31 gnutoo seems you have to modifiy your machine.conf Apr 14 13:38:35 or uclibc.conf Apr 14 13:38:56 woglinde, if I cross-compile a hello world with hardware fp and make an addition between 2 floats....with hard float it works on the machine...so... Apr 14 13:39:08 woglinde, thanks I'll change uclibc.conf Apr 14 13:42:06 just to be shure hard float means that if the device doesn't have hardware support the kernel catches the exception,compute the operation and return the result(wich is very slow) Apr 14 13:45:15 (I've TARGET_FPU = "hard") Apr 14 13:45:24 so it must be an uclibc config file... Apr 14 13:46:22 mmm...minimal-uclibc.conf:TARGET_FPU_arm Apr 14 13:46:33 oops Apr 14 13:46:35 minimal-uclibc.conf:TARGET_FPU_arm = "soft" Apr 14 13:46:40 I'll look into the file Apr 14 13:56:47 03Marcin Juszkiewicz  07org.openembedded.dev * r86b32dfe8e 10openembedded.git/conf/machine/atngw100.conf: atngw100: do not set kernel version in machine config Apr 14 13:56:57 03Marcin Juszkiewicz  07org.openembedded.dev * r1f05937e03 10openembedded.git/recipes/linux/ (linux/atngw100/defconfig linux_2.6.29.bb): linux: atngw100 uses 2.6.29 now (config updated with lot of changes) Apr 14 13:56:58 03Marcin Juszkiewicz  07org.openembedded.dev * rff97dad855 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Apr 14 13:58:55 03Koen Kooi  07org.openembedded.dev * r1549a85764 10openembedded.git/ (10 files in 3 dirs): linux-omap 2.6.29: add more DSS2 patches Apr 14 13:58:55 03Koen Kooi  07org.openembedded.dev * r9f42f94918 10openembedded.git/conf/machine/ (omap3-pandora.conf overo.conf): overo, pandora: remove traces of u-boot-omap3 Apr 14 13:58:56 03Koen Kooi  07org.openembedded.dev * r52b2c9f92f 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Apr 14 14:07:15 ~seen koen Apr 14 14:07:18 koen was last seen on IRC in channel #oe, 488d 4h 54s ago, saying: 'I forgot to commit that portion yesterday'. Apr 14 14:07:25 lrg: go to #beagle Apr 14 14:07:31 hrw: thanks Apr 14 14:11:40 hi lrg Apr 14 14:11:50 hi woglinde Apr 14 14:11:53 hi pb Apr 14 14:11:57 have to go Apr 14 14:11:59 till later Apr 14 14:12:39 hey pb_ Apr 14 14:22:28 ~seen rschuster Apr 14 14:22:30 rschuster was last seen on IRC in channel #oe, 16d 22h 18m 12s ago, saying: 'you need to apply a trick with EXTRA_OEMAKE (the value for normal make) to get it work'. Apr 14 14:22:58 hmmm Apr 14 14:23:03 what is this qa err... Apr 14 14:23:10 ERROR: QA Issue: No GNU_HASH in the elf binary: '/home/raster/oe/tmp.angstrom.ov Apr 14 14:23:10 ero/work/armv7a-angstrom-linux-gnueabi/ltrace-0.4-r2/install/ltrace/usr/bin/ltra Apr 14 14:23:10 ce' Apr 14 14:23:14 no GNU_HASH ? Apr 14 14:23:39 raster, don't know..I've often that...it may be related to ldfags Apr 14 14:23:59 hrrrmmmm Apr 14 14:24:09 I'll look into the mailing list Apr 14 14:25:14 as related i mean that you don't have ${LDFLAGS} in oe_runmake Apr 14 14:26:46 raster, I looked at the mailing list and both commit that silence the GNU hash had +${LDFLAGS} Apr 14 14:29:23 how do I silence that: ld: ERROR: /home/embedded/oetmp_pma/cross/arm-oabi/lib/gcc/arm-angstrom-linux-uclibc/4.2.4/libgcc.a(_clz.o) uses hardware FP, whereas lib/libuClibc-0.9.30.1.so uses software FP Apr 14 14:29:24 ? Apr 14 14:29:29 I've already tried that: Apr 14 14:30:12 TARGET_FPU_arm ?= "soft" instead of TARGET_FPU_arm = "soft" in uclibc-minimal.conf Apr 14 14:30:31 s/uclibc-minimal.conf/minimal-uclibc.conf// Apr 14 14:33:51 anyone? I rebuilt my tmpdir but I had this error Apr 14 14:34:02 I want to be shure before rebuilding it again Apr 14 14:34:22 or if I bitbkae -c clean all uclibc* Apr 14 14:35:16 Gnutoo: doing it :) Apr 14 14:35:40 lol Apr 14 14:36:54 03Carsten Haitzler (Rasterman)  07org.openembedded.dev * r2d4157ddaa 10openembedded.git/recipes/ltrace/ltrace_0.4.bb: ltrace: fix build (at least for omap3) Apr 14 14:36:54 03Carsten Haitzler (Rasterman  07org.openembedded.dev * ra40f32d467 10openembedded.git/conf/distro/include/sane-srcrevs.inc: sane-sercrevs: fix duplicate of essential-dailler and libefso Apr 14 14:37:10 but I don't know what to change to make it hard fpu Apr 14 14:37:18 it prints: Apr 14 14:37:50 have a ncie day Apr 14 14:38:12 TARGET_FPU = "hard" Apr 14 14:38:29 s/hard fpu/full hard fpu Apr 14 14:38:49 here it's a mix of both Apr 14 14:38:55 and I'm oabi... Apr 14 14:39:00 => that doesn't work Apr 14 14:39:07 s/doesn't/won't Apr 14 14:39:43 ah libgcc Apr 14 14:39:54 ah no Apr 14 14:43:02 morning Apr 14 14:51:10 mmm...UCLIBC_HAS_FLOATS=y UCLIBC_HAS_SOFT_FLOAT=y Apr 14 14:51:26 hello kergoth Apr 14 14:56:23 Hmmm weird issue Apr 14 14:56:34 I set HOST_NONSYSV = 1 for fakeroot Apr 14 14:56:36 rebuilt fakeroot Apr 14 14:56:40 retried to build my image Apr 14 14:56:45 it still complains about IPC Apr 14 14:57:29 mmm...it copies uclibc.machine Apr 14 15:00:55 how do fix it in a commitable way(review needed first I know) Apr 14 15:00:58 ? Apr 14 15:01:06 maybe I should talk to woglinde Apr 14 15:04:44 03Angus Ainslie  07fso/milestone5.5 * rf94b5a7e04 10openembedded.git/conf/distro/include/sane-srcrevs.inc: sane-srcrevs : bump efl version to get system startup directories processed Apr 14 15:04:45 03Angus Ainslie  07fso/milestone5.5 * r9cc4bb04be 10openembedded.git/packages/openmoko-projects/paroli_git.bb: paroli : add autostart, sounds and theme Apr 14 15:04:45 03Angus Ainslie  07fso/milestone5.5 * rbf73b89343 10openembedded.git/recipes/freesmartphone/libeflvala_git.bb: libeflvala : put recipe in the correct location Apr 14 15:04:46 03Angus Ainslie  07fso/milestone5.5 * r86176a44de 10openembedded.git/conf/distro/include/sane-srcrevs.inc: sane-srcrevs : bump frameworkd to get unknown operator bug fixed Apr 14 15:04:49 03Michael 'Mickey' Lauer  07fso/milestone5.5 * r9063911027 10openembedded.git/packages/images/ (7 files): fso-images: add fso-paroli-image; refactor fso-image into .inc; add fso-zhone-image Apr 14 15:04:51 03Angus Ainslie  07fso/milestone5.5 * r679a78f3b4 10openembedded.git/packages/openmoko-projects/paroli_git.bb: Apr 14 15:04:55 paroli: overwrite some config files with our versions Apr 14 15:04:57 some configs belong to paroli-theme and paroli-sounds Apr 14 15:04:58 03Angus Ainslie  07fso/milestone5.5 * r48f282afcb 10openembedded.git/ (3 files in 3 dirs): linphone : add a new version 3.1 recipe Apr 14 15:05:04 03Angus Ainslie  07fso/milestone5.5 * r672084e8e9 10openembedded.git/packages/python/ (files/ecore_window_free.patch python-ecore_svn.bb): Apr 14 15:05:07 python-ecore: fixing ecore bindings Apr 14 15:05:09 ecore_x_window_del was renamed to ecore_x_window_free Apr 14 15:05:11 03Angus Ainslie  07fso/milestone5.5 * r18842046e0 10openembedded.git/packages/tasks/task-openmoko-feed.bb: task-openmoko-feed : add dbus-x11 for dbus-launch so there is a session dbus Apr 14 15:11:43 I don't have a tmp/work///temp/log.do_package_stage.xxxxx file Apr 14 15:11:52 only do_package_stage_all Apr 14 15:12:13 how can I know what happened wrong during the do_package_stage phase ? Apr 14 15:23:46 grr...I don't know what happend with BBFILES... Apr 14 15:24:09 when I put only my overlay I've: 0073/6664 Apr 14 15:24:18 and I've not 6000 files Apr 14 15:24:47 did something change in the BBFILES handling? Apr 14 15:24:53 maybe the collections? Apr 14 15:25:15 I didn't understood or have forgetten what collections was Apr 14 15:27:07 Gnutoo: that shouldn't be happening. what's your BBFILES set to? Apr 14 15:27:31 BBFILES = "${HOME}/oe/org.openembedded.dev/recipes/*/*.bb" Apr 14 15:27:41 ah maybe it lacks the : Apr 14 15:27:52 ah no sorry Apr 14 15:28:23 I'm not sure what you're expecting there, if your BBFILES points at upstream ,thats what its going to parse Apr 14 15:28:40 I'm stupid and not attentive sorry Apr 14 15:28:46 hehe, no worries Apr 14 15:29:00 but anyway it didn't took my uclibc changes... Apr 14 15:29:04 I'll look again Apr 14 15:29:57 it won't prefer your overlay to upstream unless you're setting BBFILE_* or COLLECTIONS, and i think the versions need to be the same too, if you have an old version in the overlay and a new version in upstream and don't have that version locked down with a pref, itll use upstream Apr 14 15:37:13 khem: ping Apr 14 15:38:32 guys I have ar constantly saying: Apr 14 15:38:43 ar: illegal option -- f Apr 14 15:38:56 anyone has an idea of what I should do ? Apr 14 15:39:22 also I applied http://cgit.openembedded.net/cgit.cgi?url=openembedded/commit/&h=koen/darwin-host-fixes&id=28fcf1e8988f5e6ba8f05f39c94086f824c3d14f Apr 14 15:39:54 but still I have "cp: illegal option -- d" during do_package_stage phases Apr 14 15:41:26 thanks a lot it's parsing...I hope collections will work Apr 14 15:41:34 the Mac OS X bugs have to be fixed by someone with Mac OS X actively working on it. don't expect someone in here to be fixing them all for you, you should look into fixing them yourself, or wait until someone with it goes through and fixes more of them. Apr 14 15:44:19 kergoth: I fixed a lot of them actually these two days Apr 14 15:44:49 kergoth: I'm asking for advice from some people that do know the oe|bitbake code base better than I do Apr 14 15:45:01 ar and cp commandline options have nothing to do with oe. Apr 14 15:45:12 well, very little Apr 14 15:45:19 kergoth: they are called somewhere in oe Apr 14 15:45:36 yes, they are. grep is your friend Apr 14 15:45:56 kergoth: in fact, the cp problem, is already corected (I think) in the commit above Apr 14 15:46:23 kergoth: however for some reason I don't understand, changes to my base.bbclass seems to not be taken in account Apr 14 15:46:36 thanks a lot the collection tip worked Apr 14 15:47:12 kergoth: I hope I don't mind you asking question on this channel Apr 14 15:47:13 ptitjes: once again,g rep is your friend Apr 14 15:47:24 ptitjes: i just found *3* places that call cp with -d in classes. Apr 14 15:47:26 not difficult Apr 14 15:47:32 packaged-staging.bbclass calls it in two places Apr 14 15:47:49 ho Apr 14 15:47:49 lines 199 and 418 Apr 14 15:47:53 that might explain Apr 14 15:48:03 it calls it as 'cp -dpR' Apr 14 15:48:11 probably want to fix that the same way the above commit does Apr 14 15:48:57 thanks Apr 14 15:49:06 np Apr 14 15:49:18 I thought (maybe stupidely) this was handled in the base.bbclass Apr 14 15:49:36 and not even noticed there was a package-staging.bbclass Apr 14 15:50:00 packaged staging use is a distro policy decision, its up to the distro whether or not to INHERIT it, and whether or not to enable it Apr 14 15:50:01 kergoth: please forgive me I'm a newbie Apr 14 15:50:05 not a problem Apr 14 15:50:19 sorry if i seemed harsh Apr 14 15:50:22 clearly i haven't had my caffeine yet today Apr 14 15:55:14 * l403 gives kergoth a coffee bean Apr 14 15:55:39 hurray Apr 14 16:12:46 kergoth: I might myself also be a little harsh Apr 14 16:12:58 sorry for that Apr 14 16:13:15 it seems you gave me the right advice Apr 14 16:13:23 glad to hear it Apr 14 16:14:04 if the cp works I think it will correct my missing termio.h in /usr/lib Apr 14 16:14:06 :D Apr 14 16:35:41 hi mickeyl Apr 14 16:36:04 hey pb_ Apr 14 16:37:33 glad to hear you have some sun over there ;-} Apr 14 16:39:30 hehe, quite a bit. nice 24C here atm. Apr 14 16:39:40 after this long winter, it's very appreciated Apr 14 16:39:58 why do we get errors such as "ERROR: QA run found fatal errors. Please consider fixing them." Apr 14 16:41:54 I'm betting its because the QA run found fatal errors Apr 14 16:42:33 re Apr 14 16:42:45 * mwester bets kergoth is right, but has a feeling that's not the answer vsr was looking for Apr 14 16:43:23 yes, so whats the answer that will fix this problem Apr 14 16:43:23 mwester: depmod patch! Apr 14 16:43:59 * mwester has changed "git pull" in his update-openembedded script to "git pull || git pull || git pull || git pull" ----- but wonders if the correct solution isn't to find out what is choking the git server. Apr 14 16:44:12 Ainulindale: what depmod patch, and what would you like to do about it? Apr 14 16:44:13 :) Apr 14 16:44:19 Grmmmmbl. Apr 14 16:44:29 Well some way to depmod -a at image generation Apr 14 16:44:33 Then at kernel upgrade Apr 14 16:44:38 It is done already. Apr 14 16:44:45 Commited? Apr 14 16:44:50 I didn't see a thing about that in your commit Apr 14 16:44:51 vsr: that completely depends on what errors it found. sadly, we aren't as good as reading your mind as we'd like Apr 14 16:44:55 It is a patch in the SHR overlay. Apr 14 16:45:16 Yeah I saw that Apr 14 16:45:39 But I had to do depmod -a manually here Apr 14 16:45:39 kergoth: http://www.pastebin.ca/1391850 Apr 14 16:45:40 With latest Apr 14 16:45:52 Ainulindale: basically, the postinst script for each kernel-module-*.ipk does a depmod; the problem was that the kernel version was 2.6.29-rc3 but the recipe said 2.6.28 -- so the depmod tried to use the wrong /lib/modules/ path. Apr 14 16:45:57 And looking at your patch I can't see how it'd produce depmod -a Apr 14 16:46:06 hmmm Apr 14 16:46:30 Well then I can't explain why I had to run it manually (first boot) Apr 14 16:46:42 Ainulindale: the long-term fix cannot be committed to the FSO branch without assistance from mickeyl, since it is a tricky problem to get right (a different SRCREV will result in the kernel changing version again) Apr 14 16:46:45 Maybe I didn't have to, but something else was p reventing usb from being detected Apr 14 16:46:56 Ainulindale: THe USB is a different problem. Apr 14 16:47:23 That is? Apr 14 16:47:25 That has to do with something I have fought against the whole time; you cannot make the USB Netrwork driver a module and expect to have it work reliably. Apr 14 16:47:57 I do not have a debug board on a gta02 so I cannot figure out what is going wrong, but it seems quite clear that there will have to be special handling for that kernel module. Apr 14 16:48:06 Well, I'm quite open to have it built-in Apr 14 16:49:09 As am I. My opiniion on that is the same as the 100,000 kernel modules being installed in the image by default --- the base image should provide a BASIC set of modules, and it should work 100% reliably, and those who wish to change that should opkg install what they want. Apr 14 16:49:40 Well, agreed Apr 14 16:49:47 Fancy doing that? :-> Apr 14 16:49:54 As a distro matures and stabilizes, that can be re-evaluated, but it is silly to have so many kernel modules in the base image, and we simply create reliability issues by having everything in the kernel be a module anyway. Apr 14 16:50:01 I'll be happy to do that. Apr 14 16:50:16 03Michael 'Mickey' Lauer  07org.openembedded.dev * ra0239e4e33 10openembedded.git/recipes/linux/linux-msm7xxxx_git.bb: linux-msm7xxx: bump to HEAD which seems to be working again Apr 14 16:50:20 Ok, good :-) Apr 14 16:50:53 kergoth, exactly where do we identify where is the problem in .bb file?: http://www.pastebin.ca/1391856 Apr 14 16:50:56 If we create an SHR kernel image, that is a copy (or even includes) the same bb file used in FSO, we will be free to repackage as needed, and will be able to do many hacks that would otherwise mess up mickeyl's life ;) Apr 14 16:51:19 mwester: well again, I trust you on this matter, I don't have enough technical details to be sure about what I'll do Apr 14 16:51:30 So if you're willing to handle that, I'll be more than happy Apr 14 16:51:37 (of course, just switching to the dev kernel would solve a number of the issues!) Apr 14 16:51:40 vsr: the buildsystem in question isn't obeying our LDFLAGS. how you make it do so depends on the buildsystem Apr 14 16:52:11 Ainulindale: It shall be done. Also, we need to take a look at that overlay and especially at those OE patches, and see what we can get into OE. Apr 14 16:52:24 (easing my life) sounds good! Apr 14 16:52:38 mwester: yep I agree Apr 14 16:52:39 vsr: there are lots of recent commits to OpenEmbedded to fix that problem in a variety of recipes, perhaps one of those will give you ideas on how to fix it for that package Apr 14 16:52:43 But I'd like to keep this overlay that way Apr 14 16:52:50 you mean my environmental variables are causing the problem Apr 14 16:52:54 okay Apr 14 16:52:59 Removing stuff is a good thing indeed, but this build architecture suits us well Apr 14 16:52:59 mickeyl: no, not really; we're just transferring pain around. A trick most politicians are quite expert at... :-D Apr 14 16:53:02 vsr: no Apr 14 16:53:09 mwester: heheh Apr 14 16:53:26 vsr: i mean exactly what i say. the buildsystem of the project being built by that recipe isn't obeying all of openembedded's variables, from its metadata. Apr 14 16:53:27 But I'd like to keep the overlay the way it is (minus packages commitable) as it'll be a sandbox, and I like that Apr 14 16:53:47 hrw|gone: you should not use binutils 2.19x Apr 14 16:53:54 This way we'll be able to "open" it up to people, in order for us to ease the package transition into OE (less access rights needed, etc...) Apr 14 16:54:11 hrw|gone: gcc 4.3.3 binitils 2.18 should work Apr 14 16:54:16 kergoth: maybe you know this... Apr 14 16:54:54 kergoth: how can I set the PREFERRED_PROVIDER for a _runtime_ package? Apr 14 16:55:00 Ainulindale: Then I shall like to convert some of those patches into SHR-specific recipes. Patching the openembedded tree is just too ugly for words... Apr 14 16:55:49 denix: it depends on the context Apr 14 16:55:57 mwester: well yes, but how would you do that then? Apr 14 16:56:18 kergoth: NOTE: multiple providers are available for runtime update-alternatives (update-alternatives-cworth, update-alternatives-dpkg); Apr 14 16:56:55 kergoth: nothing seems to work to fix that Apr 14 16:56:56 if you're talking about bitbake's code that figures out which to build based on RDEPENDS< then you set it the same way you normally do, its smart enough to handle it Apr 14 16:56:58 Ainulindale: overrides Apr 14 16:57:02 according to RP anyway Apr 14 16:57:05 i asked him about that last week Apr 14 16:57:23 mwester: well I'd like to avoid duplication as much as I can Apr 14 16:57:24 Ainulindale: I'll take a look more closely, and offer a suggestion for you and mickeyl to review. Apr 14 16:57:26 That's the only "condition" I have :-) Apr 14 16:57:30 mwester: yep :-) Apr 14 16:57:44 mwester: great Apr 14 16:57:46 thanks Apr 14 16:58:59 kergoth: Angstrom sets this PREFERRED_PROVIDER_virtual/update-alternatives = "update-alternatives-cworth" but it doesn't seem to eliminate the above message from bitbake Apr 14 16:59:30 try PREFERED_PROVIDER_update-alternatives = "update-alternatives-cworth", perhaps? Apr 14 16:59:42 kergoth: tried that - still no go Apr 14 16:59:49 I'd say ask RP Apr 14 16:59:56 I'm just going by what he said about that code Apr 14 17:00:20 kergoth: I noticed that only _runtime_ packages have this issue Apr 14 17:00:29 that's not surprising. Apr 14 17:00:36 it's handled differently Apr 14 17:00:43 ok, I'll try to ask RP :) Apr 14 17:00:52 and thats code I've never touched, and was written after i left Apr 14 17:00:54 heh Apr 14 17:01:21 thanks anyway :) Apr 14 17:02:29 btw, how's adoption of bitbake going on? :) Apr 14 17:03:50 ? Apr 14 17:07:04 kergoth: getting bb into oe git? Apr 14 17:07:27 ah, that, haven't had the time. should be able to this week unless someone else gets to it Apr 14 17:08:03 ah, no problem. just checking :) Apr 14 17:08:15 * kergoth nods Apr 14 17:10:12 mmm...GNU_HASH is not anymore optional? QA fails because of it...I should fix uclibc's ldflags Apr 14 17:10:24 (before it only warned) Apr 14 17:10:37 pretty sure there was a mailing list thread about that Apr 14 17:12:07 I saw so much fixing in the commits... Apr 14 17:12:19 but I should have forgetten about the mailing list post Apr 14 17:12:24 yes, there was - I had to fix my QA issues :) Apr 14 17:12:42 lol ok Apr 14 17:13:37 hum... I need your advices again men Apr 14 17:14:17 it seems openssl-native request me a /usr/include/termio.h Apr 14 17:14:25 this files seems to be from glibc Apr 14 17:14:35 but there is no glibc-native Apr 14 17:15:30 do you think I should create a glibc-native or do you think there might something simpler to do ? Apr 14 17:15:49 ptitjes: no Apr 14 17:16:16 ok so I'll go for the glibc-native then Apr 14 17:16:19 Gnutoo: http://thread.gmane.org/gmane.comp.handhelds.openembedded/22491 Apr 14 17:16:24 thanks Apr 14 17:17:11 ptitjes: you need to install -dev package for glibc on your host Apr 14 17:17:16 would be easier Apr 14 17:17:52 khem but then it won't be on the include search path of OE Apr 14 17:17:54 ? Apr 14 17:18:30 03Koen Kooi  07org.openembedded.dev * r13e3e087e4 10openembedded.git/recipes/dsplink/ (4 files): dsplink: add INSANE_SKIP till the codec build gets fixed upstream Apr 14 17:19:48 khem: am I wrong ? Apr 14 17:20:06 ptitjes: it will look for native packages Apr 14 17:20:31 ha ok Apr 14 17:20:34 thanks Apr 14 17:24:19 ptitjes: look at http://wiki.openembedded.net/index.php/OEandYourDistro Apr 14 17:24:41 and try to have those relevant packages installed on your host Apr 14 17:25:38 khem: arf I'm half on the Gentoo chair and half on the MacOS chair :) Apr 14 17:27:01 That must be uncomfortable. Apr 14 17:27:09 héhéhé :) Apr 14 17:27:18 lol Apr 14 17:28:16 ptitjes: heh well stach em up and sit on one Apr 14 17:29:05 ptitjes, by the way I'm curious...why are you on MacOSX? Apr 14 17:29:20 thats like saying why are you on linux? Apr 14 17:29:44 or windows - personal preference! Apr 14 17:29:47 ok Apr 14 17:29:53 (i prefer os x too!) Apr 14 17:30:26 however, its not the best choice for OE, unless you are working towards making it work on OE... :D Apr 14 17:30:36 khem: in fact, the Gentoo instructions does not state glibc Apr 14 17:30:51 I asked because some people don't want to be on windows or OSX but are like hostages of the platform(for instance they must use it because they must use a windows program or something like this) Apr 14 17:30:53 or making OE work on it even Apr 14 17:30:55 khem: maybe one of the listed packages are dependeing on glibc Apr 14 17:30:57 ptitjes: just saw that Apr 14 17:31:09 ptitjes: it must be pulled by others Apr 14 17:31:15 yeah Apr 14 17:31:23 and on gentoo you cant build much with out libc-dev Apr 14 17:31:27 unfortunately there is no glibc for prefix Apr 14 17:31:48 so I assume its meant to be there if you have a properly emerged gentoo box beforehand Apr 14 17:32:23 ptitjes: I am not a gentoo user let alone this prefix Apr 14 17:32:33 I don't know prefix proceed but I don't have any glibc for sure Apr 14 17:32:46 +how Apr 14 17:32:51 look for libc Apr 14 17:33:25 I see dietlibc Apr 14 17:33:42 but I'm not sure that it would containt my needed termio.h Apr 14 17:34:32 ugh Apr 14 17:35:00 khem: asking this on prefix channel, it seems it uses the host's libc Apr 14 17:35:08 so the one of MacOS Apr 14 17:35:11 for me Apr 14 17:35:21 well thats runtime Apr 14 17:35:41 here you need something for compile time. So yeah glibc-native would be your solution Apr 14 17:35:51 but why are you using such crippled system Apr 14 17:35:52 and this one provides a termios.h and not termio.h :D Apr 14 17:36:17 cant you use some other distro on top of parallels or vmplayer Apr 14 17:36:50 even normal gentoo will be ok Apr 14 17:36:52 khem: I don't know well how parallels work Apr 14 17:37:23 khem: but it seems to me it is a virtual machine but does not uses proc virtualization Apr 14 17:37:27 am I wrong ? Apr 14 17:37:28 then get a vmplayer and get a ubuntu image for it Apr 14 17:37:37 its all there at no charge Apr 14 17:37:54 I use Parallels on a Mac; run both Xp and Ubuntu using it. Works fine. Apr 14 17:38:36 robtow: so you can tell me... it emulates a single processor, does it ? Apr 14 17:38:52 It doesn't emulate a processor. Apr 14 17:39:41 It allows guest OS to run, under a hypervisor. launched from within Mac OS. Apr 14 17:39:45 robtow: so why did my x86_64 distro told me that it can't install on a x86 proc ? Apr 14 17:40:12 when tried to run under parallels Apr 14 17:40:44 I have no idea. Apr 14 17:40:57 Is your underlying processor a 64 bit or 32 bit processor? Apr 14 17:41:02 hi khem Apr 14 17:41:04 64bit Apr 14 17:41:23 wait I check Apr 14 17:41:37 well, perhaps it's because your version of Mac OS is only 32 bit. Apr 14 17:41:38 yes EM64T Apr 14 17:42:10 And Parallels is running under that host OS, and is constrained by its addressing limits. Apr 14 17:42:17 I don't think so but is there a way I can check that ? Apr 14 17:42:30 some headers ? Apr 14 17:42:41 You don't really need more than 4GB of RAM for OE anyway, so you can just install a 32-bit guest OS. Apr 14 17:43:30 I agree. Just install Ubuntu, and you're good to go. Apr 14 17:44:29 robtow: another question is parallels bound to install an OS in a file ? Apr 14 17:44:32 :) Or Fedora. Apr 14 17:44:41 yeah or gentoo Apr 14 17:44:49 I'm used to that so Apr 14 17:44:57 According to the parallels website, they do not support 64 bit guest operating systems even when running under 64 bit Mac OS (Lepeord). So there you are. Apr 14 17:45:27 there it is :) Apr 14 17:45:34 ptitjes - you should go over to the parallels site, and browse the fora there. Apr 14 17:45:42 yep Apr 14 17:45:47 thank robtow Apr 14 17:45:53 you're welcome. Apr 14 17:46:33 but I think I won't go this way Apr 14 17:46:42 almost until they support 64bit OS Apr 14 17:47:56 because I would have to maintain three OS (the mac, the real gentoo, and the gentoo for //s) Apr 14 17:49:00 I do most of my OE work under 64 bit Centos 5.2 quad core machine; also have a 32 bit Centos 5.2 sandbox, and have previously also done some OE under Ubuntu on a old 32 bit x86 laptop. It all works well enough. Apr 14 17:49:54 hey woglinde wie gehts Apr 14 17:50:22 ptitjes - not a big deal once you set it up. Apr 14 17:50:43 my mac's battery died after 1.5 years and old viao from ~2003 still runs for 30 mins Apr 14 17:50:59 and new one is 130 Apr 14 17:51:01 crap Apr 14 17:52:02 khem - maybe you've been cycling the Mac's battery up and down through more cycles Apr 14 17:53:03 men I just do know how I'll do it Apr 14 17:53:16 I'll ASSUME_PROVIDED openssl Apr 14 17:53:19 :D Apr 14 17:53:25 that is much more simpler Apr 14 18:02:06 but doesn't work Apr 14 18:02:08 :( Apr 14 18:02:13 arf Apr 14 18:02:37 * ptitjes goes to have apetizers Apr 14 18:02:45 robtow: I aint gonna buy a new mac anymore problem solved Apr 14 18:05:04 khem - as long as you ahve tools that solve your problems it's all cool. Apr 14 18:05:22 <--is sooo totaly NOt into the Mac vs PC vs Linux religions Apr 14 18:05:41 robtow: I am not too Apr 14 18:05:54 robtow: but I like Macs Apr 14 18:06:12 robtow: I didn't have before I can't afford one Apr 14 18:06:25 <---got started on PDP-7 using punched tape and teletype; then IBM-360 with punched cards. Apr 14 18:07:43 I simply had so many problem with mine may be I am unlucky in dec I had to replace the adapter (another 90 bucks) Apr 14 18:07:50 <---has had Macs since the original 128K Mac in 1984. Apr 14 18:08:19 03Denys Dmytriyenko  07fso/milestone5.5 * r79af66272f 10openembedded.git/packages/ (16801 files in 4640 dirs): (log message trimmed) Apr 14 18:08:19 rename packages/ to recipes/ per earlier agreement Apr 14 18:08:19 See links below for more details: Apr 14 18:08:19 http://thread.gmane.org/gmane.comp.handhelds.openembedded/21326 Apr 14 18:08:19 http://thread.gmane.org/gmane.comp.handhelds.openembedded/21816 Apr 14 18:08:21 Signed-off-by: Denys Dmytriyenko Apr 14 18:08:25 Acked-by: Mike Westerhof Apr 14 18:08:30 03Angus Ainslie  07fso/milestone5.5 * r7adc430194 10openembedded.git/ (2 files in 2 dirs): Apr 14 18:08:30 task-openmoko-feed : add linphone to the feed Apr 14 18:08:32 sane-srcrevs : build the latest linphone Apr 14 18:08:34 03Denys Dmytriyenko  07fso/milestone5.5 * r044fdb4816 10openembedded.git/recipes/ (5 files in 3 dirs): (log message trimmed) Apr 14 18:08:37 recipes: fix hardcoded instances of packages/ Apr 14 18:08:41 Signed-off-by: Denys Dmytriyenko Apr 14 18:08:43 Acked-by: Mike Westerhof Apr 14 18:08:45 Acked-by: Philip Balister Apr 14 18:08:47 Acked-by: Khem Raj Apr 14 18:08:49 Acked-by: Marcin Juszkiewicz Apr 14 18:08:53 03Theodore A. Roth  07fso/milestone5.5 * rcd133984ff 10openembedded.git/conf/local.conf.sample: local.conf.sample: Replace packages/ with recipes/. Apr 14 18:08:58 03Denys Dmytriyenko  07fso/milestone5.5 * r9ca67723c3 10openembedded.git/classes/base.bbclass: (log message trimmed) Apr 14 18:09:01 base.bbclass: fix base_get_scmbasepath() to look for recipes/ and fall back to packages/ while parsing BBFILES Apr 14 18:09:03 Signed-off-by: Denys Dmytriyenko Apr 14 18:09:05 Acked-by: Mike Westerhof Apr 14 18:09:07 Acked-by: Philip Balister Apr 14 18:09:11 Acked-by: Khem Raj Apr 14 18:09:13 Acked-by: Marcin Juszkiewicz Apr 14 18:12:10 <---fondly remembers the Xerox Dorado personal computer Apr 14 18:13:09 khem - stuff wears out. I've worn out HP, Toshiba, Macs, etc. It's like cars; it just wears out, and you ahve to replace parts or whole systems. Apr 14 18:14:00 yeah may be I was expecting more given my older systems Apr 14 18:14:37 hello, I've EXTRA_OEMAKE = "${OEMAKE_NO_CC} 'CC=${CC}' 'HOSTCFLAGS=-I${STAGING_INCDIR_NATIVE} ${LDFLAGS}'" but the GNU HASH QA problem doesn't want to go away Apr 14 18:15:53 Gnutoo: which package Apr 14 18:16:17 uclibc_0.9.30.1.bb Apr 14 18:16:44 hmmm what do you want to pass specifically Apr 14 18:16:54 ah the gnu_hash stuff Apr 14 18:17:07 yes,ldd and ldconfig don't have gnu_hash Apr 14 18:17:13 but I'm unshure if they need it Apr 14 18:17:35 if they need to run on target they need it Apr 14 18:17:44 ok Apr 14 18:18:36 ah they get compiled in do_install(){ Apr 14 18:19:38 here the command that was executed: http://rafb.net/p/x7Iyaa91.html Apr 14 18:21:04 khem: thx Apr 14 18:32:45 03Koen Kooi  07stable/2009 * r87fa0d07ab 10openembedded.git/recipes/gstreamer/gst-plugins-base_0.10.22.bb: Apr 14 18:32:45 gst-plugins-base: depend on libtheora so gst-plugin theora gets built Apr 14 18:32:45 Signed-off-by: Koen Kooi Apr 14 18:32:45 Acked-by: Marcin Juszkiewicz Apr 14 18:32:55 03Chris Larson  07stable/2009 * r598231b1e3 10openembedded.git/recipes/sato-icon-theme/sato-icon-theme.inc: Apr 14 18:32:55 sato-icon-theme: I doubt sato-icon-theme.inc including sato-icon-theme.inc is helpful. Apr 14 18:32:55 Signed-off-by: Chris Larson Apr 14 18:32:55 Signed-off-by: Koen Kooi Apr 14 18:32:57 Acked-by: Marcin Juszkiewicz Apr 14 18:54:06 what should I do then? Apr 14 18:55:34 woglinde, gu are you still there? Apr 14 18:56:06 s/gu/hi/ Apr 14 18:58:15 I'll try the verbose swithc Apr 14 18:58:20 s/hc/ch Apr 14 19:00:35 ah ok Apr 14 19:01:41 Under Ubuntu 9.04 bitbake gives me: bitbake/lib/bb/COW.py:29: DeprecationWarning: the sets module is deprecated Apr 14 19:02:05 python2.6 stuff Apr 14 19:02:30 hrw: will we fix bitbake or have bitbake depend on other python? Apr 14 19:02:38 hi, I am having this problem with stable/2009: http://pastebin.com/m36c05ba Apr 14 19:02:52 likewise: they are warnings not errors Apr 14 19:02:53 has it been fixed in dev? this is actually someone else's pastebin from two weeks ago Apr 14 19:03:02 (curl-sdk does not build) Apr 14 19:03:23 dhd: search ml Apr 14 19:03:28 dhd: iirc something was done Apr 14 19:03:34 this has been discussed and potential solution is there Apr 14 19:04:36 ah I see Apr 14 19:04:38 thanks Apr 14 19:06:03 hrw: armv5te/omap5912 was the target arch that I made working on uclibc/thumb Apr 14 19:06:08 mmm....http://rafb.net/p/3kfVDz14.html Apr 14 19:07:13 khem: GNU ar (Linux/GNU Binutils) 2.18.50.0.7.20080502 Apr 14 19:07:30 hrw: I used 2.18 release Apr 14 19:07:39 Angstrom uses 2.18.50.0.7 Apr 14 19:08:13 Did not try that usually thse HJ Lu releases are good for x86 lnx Apr 14 19:08:37 I have had issues with them on non x86 machines so I prefer to use releases Apr 14 19:08:42 oh by the way, there is a kernel patch that's needed for that binutils on ARM Apr 14 19:08:47 otherwise modules will not load Apr 14 19:08:47 when 2.19.1 did not work I chose 2.18 Apr 14 19:09:21 khem: can you look at 2.18.50.0.7 then? Apr 14 19:09:23 hrw: I will try 2.18.50.0.7 Apr 14 19:09:26 thx Apr 14 19:09:26 (at least on armv4t) Apr 14 19:10:33 Gnutoo: good. Seems to work Apr 14 19:11:25 ??? Apr 14 19:11:34 12:06 < Gnutoo> mmm....http://rafb.net/p/3kfVDz14.html Apr 14 19:11:39 yes I know Apr 14 19:11:43 but it still doesn't work Apr 14 19:11:54 so it should be the build system Apr 14 19:12:03 I'll check the hash manually Apr 14 19:12:45 Gnutoo: build with V=1 Apr 14 19:12:49 and post the build log Apr 14 19:12:53 ok thanks a lot... Apr 14 19:13:21 thanks I didn't know what to put after the V=(true,1) or -DV Apr 14 19:15:21 denix: perhaps the config qa task should check for AM_GNU_GETTEXT in configure.{in,ac} and abort/warn if the recipe isn't inheriting gettext Apr 14 19:16:08 03Stanislav Brabec  07org.openembedded.dev * rf4a7e18834 10openembedded.git/ (conf/checksums.ini recipes/mtkbabel/mtkbabel_0.7.bb): mtkbabel: Update to version 0.8. Apr 14 19:16:18 ~curse me Apr 14 19:16:19 03Stanislav Brabec  07org.openembedded.dev * rd9cb855047 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Apr 14 19:16:19 May the fleas of a thousand camels infest your most sensitive regions, hrw ! Apr 14 19:20:18 03Denys Dmytriyenko  07org.openembedded.dev * ra2fa7e7164 10openembedded.git/recipes/e2fsprogs/e2fsprogs_1.41.4.bb: e2fsprogs 1.41.4: depends on gettext-native, due to the removal of do_configure Apr 14 19:21:13 03Denys Dmytriyenko  07org.openembedded.dev * rf6e91a7cca 10openembedded.git/recipes/udev/ (udev-124/libvolume-id-soname.patch udev_124.bb): udev 124: fix the absolute soname link in libvolume for SDK Apr 14 19:22:02 kergoth: that would be a nice precaution measure Apr 14 19:23:28 easy to implement, too. just need to call autoconf -t AM_GNU_GETTEXT. if the output is non-empty, and not bb.data.inherits_class("gettext"), warn Apr 14 19:25:31 http://rafb.net/p/tgNOr297.html Apr 14 19:25:56 ah, there is gettext.bbclass... I was just adding DEPENDS for that :) Apr 14 19:26:49 ~hail tftp Apr 14 19:26:50 * ibot bows down to tftp and chants, "I'M NOT WORTHY!!" Apr 14 19:28:12 khem, http://rafb.net/p/tgNOr297.html <---here's the part with ldd on the V=1 build Apr 14 19:28:52 I'm having troulbe building meta-toolchain, because task-sdk-bare depends on uclibc-thread-db which isn't being built for some reason... Apr 14 19:29:26 denix: aye, gettext.bbclass = appropriate deps + enable/disable nls in EXTRA_AUTOCONF, iirc Apr 14 19:30:55 shit.. still no dataflash ;( Apr 14 19:32:01 oh... I see, thread debugging is not enabled in uclibc config Apr 14 19:32:11 is there a good reason for this? Apr 14 19:32:23 kergoth: right. will change it to inherit Apr 14 19:32:43 Gnutoo: make CROSS="arm-angstrom-linux-uclibc-" CC="ccache arm-angstrom-linux-uclibc-gcc -march=armv4" -C utils Apr 14 19:32:45 i think there are a number of packages missing that inherit Apr 14 19:32:51 so the flags are not passed Apr 14 19:32:52 i spotted a few the other day Apr 14 19:33:18 Gnutoo: enclose whole CC in quotes CC=".." Apr 14 19:33:25 ok Apr 14 19:33:59 dhd: the thread debugging you are seeing in there is not what you are looking for Apr 14 19:34:14 those options print messages on console like kernel kprintf Apr 14 19:34:31 03Koen Kooi  07fso/milestone5.5 * r92cc465f7d 10openembedded.git/recipes/mozilla/ (fennec/jsautocfg-dontoverwrite.patch fennec_hg.bb): fennec: update from alpha to beta release, work around latest mozilla bugs (filed upstream) Apr 14 19:34:33 khem: PTHREADS_DEBUG_SUPPORT in uclibc you mean? Apr 14 19:34:38 yes Apr 14 19:34:49 does a working libthread_db in uclibc require nptl then? Apr 14 19:35:04 if so I guess I can just remove uclibc-thread-db from the SDK task Apr 14 19:35:30 dhd: I was talking about DODEBUG_PT Apr 14 19:35:48 kergoth: huh, grep 'gettext-native' recipes/* -r|grep DEPENDS Apr 14 19:36:01 oh... in any case, task-sdk-base expects the uclibc-thread-db package to exist, but it does not Apr 14 19:36:08 (in stable/2009) Apr 14 19:36:36 dhd: infact we should enable PTHREADS_DEBUG_SUPPORT Apr 14 19:37:03 hmm weird, the uclibc package is definitely building libthread_db-0.9.30.so, but no ipk is being built Apr 14 19:37:52 khem: yeah, that's what I thought... Apr 14 19:44:56 uf.. Apr 14 19:45:07 ngw100 finally has all entries in /proc/mtd Apr 14 19:45:25 oh. because of debian-versions the thread-db package ends up being called libthread-db1 Apr 14 19:45:30 how can this be fixed? Apr 14 19:45:47 (at least I assume this is the debian versions thing) Apr 14 19:48:31 are the SDK tasks incompatible with INHERIT += "debian"? Apr 14 19:53:45 oh, okay, one must simply add DEBIANNAME_uclibc-thread-db = "uclibc-thread-db" to uclibc.inc Apr 14 19:55:23 * dhd wonders if glibc need a similar fix? Apr 14 19:57:16 03Marcin Juszkiewicz  07org.openembedded.dev * r5f97fcad85 10openembedded.git/recipes/linux/ (linux/atngw100/defconfig linux_2.6.29.bb): (log message trimmed) Apr 14 19:57:16 linux: enabled lot of options for Atmel NGW100 device Apr 14 19:57:16 - netfilter support (everything) Apr 14 19:57:16 - 8MB Dataflash chip is now recognized Apr 14 19:57:16 - power management Apr 14 19:57:18 - watchdog Apr 14 19:57:20 - dma engine Apr 14 19:57:41 hi gremlin[it] :-D Apr 14 19:58:01 hi mckoan|away Apr 14 19:58:29 nice features from hrw ... so i can test it ! Apr 14 19:58:51 gremlin[it]: which one? Apr 14 19:59:13 mckoan|away, i'm looking about kaleios features and or personality (NAS / router / PDA) Apr 14 19:59:29 last one about NGW100 ... i received a NGW100 ... Apr 14 19:59:49 great! what about the damn sheevaplug ? Apr 14 20:00:00 mckoan|away, no news at today Apr 14 20:01:09 mckoan|away, a couple of questiona bout kaleios, differences from angstrom ? is suitable as blink (no display) base distro ? Apr 14 20:01:42 can i consider kaleios more like angstrom or 'general' OE distribution ? Apr 14 20:02:07 gremlin[it]: at the moment is still likely angstrom Apr 14 20:02:25 ATNGW100 Network Gateway Kit looks cool device Apr 14 20:03:26 yes is quite cool and i found some hardware hacking to play with, such as display and wifi Apr 14 20:03:41 kaleios target features will be faster boot and less unrequired services for embedded systems Apr 14 20:04:11 mhh good ... I'm trying it for qemu target at now Apr 14 20:05:09 on stable/2009 Apr 14 20:06:05 thanks a lot it worked!!! Apr 14 20:16:36 mmm...| gettextP.h:29:21: error: iconv.h: No such file or directory (uclibc hasn't iconv and I must bitbake it) Apr 14 20:16:52 uclibc uses libiconv Apr 14 20:17:35 yes I'm bitbaking it right now Apr 14 20:18:04 gnutoo hm didnt merge my gettext-branch yet Apr 14 20:18:13 then the support is better with libiconv Apr 14 20:18:31 but shouldn't it compile it automagically before compiling libgbgerror Apr 14 20:19:07 theoretical yes Apr 14 20:19:18 if libgpgerror is inheriting gettext, it should be, yes Apr 14 20:19:53 so where should I put de DEPEND+= Apr 14 20:20:04 nope Apr 14 20:20:08 @{} Apr 14 20:20:10 inherit gettext Apr 14 20:20:11 of course Apr 14 20:20:13 ah ok Apr 14 20:20:24 ouch in the class? Apr 14 20:20:30 * kergoth notes that lots of recipes that use gettext don't seem to be inheriting it at the moment Apr 14 20:20:44 kergoth right Apr 14 20:21:02 i'll check if it is inheriting gettext Apr 14 20:21:43 woglinde: i just suggested to denix earlier that we make configure_qa check for it and warn/abort. if its autotools and calls AM_GNU_GETTEXT but isnt inheriting gettext, freak out. easily done, and may save headaches Apr 14 20:22:32 have a nice rest of day Apr 14 20:22:46 good nite hrw Apr 14 20:24:27 night hrw|gone Apr 14 20:25:45 is there an easy way to get OE to build native binaries as 32-bit on an amd64 system? Apr 14 20:26:03 (by "easy", meaning "easier than setting up a chroot and doing it there") Apr 14 20:26:12 * dhd needs to build a 32-bit SDK Apr 14 20:26:49 BUILD_CFLAGS += "-m32" or something? Apr 14 20:26:56 * kergoth shrugs Apr 14 20:27:04 yeah something like that... Apr 14 20:27:09 I suppose that I could just try it... Apr 14 20:27:17 * mwester isn't sure he'd trust a simple flag over a chroot Apr 14 20:27:21 thanks Apr 14 20:27:28 * kergoth either, but what the hell Apr 14 20:27:47 would BUILD_CFLAGS be the appropriate variable for host binaries, then? Apr 14 20:27:55 "host" is ambiguous Apr 14 20:27:59 BUILD_ is the machine you're building on Apr 14 20:28:03 oh, yes, sorry :) Apr 14 20:28:09 I mean for the build machine Apr 14 20:28:26 ok will try it... Apr 14 20:38:12 woglinde: since you currently work on gettext, would you fix all the occurrences of: grep 'gettext-native' recipes/* -r|grep DEPENDS Apr 14 20:40:30 denix hm sometimes you need the direct dep of gettext-native Apr 14 20:41:42 re pb Apr 14 20:42:51 woglinde, can I commit the inherit gettext or does it need review? Apr 14 20:43:08 gnutoo which package? Apr 14 20:43:18 did you really test from empty tmp? Apr 14 20:44:09 libgpg-error Apr 14 20:44:34 woglinde: those are just the ones that get built before glibc, so they need the actual libintl, right? Apr 14 20:44:47 no but I had made a clean on libiconv and it pulled it Apr 14 20:45:01 gnutoo okay Apr 14 20:45:06 kergoth hm? Apr 14 20:45:13 explicit gettext-native dep Apr 14 20:45:22 what packages would need that/ Apr 14 20:45:34 kergoth hm if I remeber correctly the need some binary or so Apr 14 20:45:39 hmm, k Apr 14 20:45:40 ( http://rafb.net/p/k64f2n61.html ) <- oops there is my mail in it Apr 14 20:45:54 but dont know it exactly anymore Apr 14 20:46:11 maybee I mix it with libintl usage Apr 14 20:46:23 which wouldnt be gettext-mative Apr 14 20:47:30 * kergoth nods Apr 14 20:47:52 we're using gettext-native for its utils and m4 macros right Apr 14 20:48:04 gettext confuses me, probably cause I've never actually written an app myself that used it :) Apr 14 20:48:11 *g* Apr 14 20:48:33 kergoth and the host-binaries to compile the .po files Apr 14 20:48:37 * kergoth nods Apr 14 20:48:40 hi ant Apr 14 20:48:41 woglinde, was the okay of the commit or just for saying that you understood what I told(I always say ok to acknowledge that I read what the others say) Apr 14 20:48:52 gnutoo commit it Apr 14 20:48:55 ok thanks Apr 14 20:48:57 if it is bad Apr 14 20:49:07 show with the finger at me Apr 14 20:49:11 ok Apr 14 20:53:02 hello woglinde Apr 14 20:54:50 ptitjes: just for the records...every Gentoo has a glibc installed...and gcc too :) Apr 14 21:00:01 ant__: unless you emerge -C glibc gcc :) Apr 14 21:00:11 :) Apr 14 21:01:02 has anybody found the time to test the new bluez4.x stack? Apr 14 21:01:34 03Leon Woestenberg  07org.openembedded.dev * r2c241e8606 10openembedded.git/recipes/util-linux-ng/ (3 files): util-linux-ng: Fixed dangling symlinks. Apr 14 21:02:01 nope. I don't speak "dbus", so it's tough -- I've been using the older bluez stuff instead. :( Apr 14 21:02:13 ;-) Apr 14 21:08:02 http://www.rafb.net/cohen/blog/2009/02/13/interviewing-part-2-architects-and-chefs/ (because there are many cooks here creating recipes) Apr 14 21:12:59 gn Apr 14 21:17:19 keesj: nice! Apr 14 21:17:44 good post Apr 14 21:36:04 anyone working on the stable/2009 branch here ? I got directfb crashing, including the sample apps and tools.. I'd like to cross check with someone else using stable... Apr 14 21:37:48 03Denis 'Gnutoo' Carikli  07org.openembedded.dev * r06b6c395da 10openembedded.git/recipes/libgpg-error/libgpg-error.inc: Apr 14 21:37:48 libgpg-error: add inherit gettext in order to have libiconv as dependency with uclibc Apr 14 21:37:48 thanks woglinde for the info Apr 14 21:38:33 buserror hm which machineß Apr 14 21:39:50 woglinde, arm920t board. angstrom armv4t glibc Apr 14 21:41:38 by the way what's the difference between arm920t and Arm925Tsid? I saw no tune file for 925tsid Apr 14 21:42:37 gnutoo sid? Apr 14 21:43:02 I knew what it was...but I don't remember Apr 14 21:43:06 maybe d is for debug Apr 14 21:43:36 anyway I meant 925t Apr 14 21:44:07 hm shouldnt be no difference Apr 14 21:44:14 args Apr 14 21:44:19 double no Apr 14 21:44:42 ok Apr 14 21:47:22 anyway, directfb crashes in libdirectfb_joystick.so -- console claims an alignment problem on address 0xffffffff (which explains :>) Apr 14 21:47:33 hm Apr 14 21:47:41 do you need joystick? Apr 14 21:47:56 otherwise disable it in .directfbrc Apr 14 21:49:01 mmm....| sys_nim.c:48: error: expected declaration specifiers or '...' before 'sys_init_module' Apr 14 21:49:08 I'll search for what it means Apr 14 21:49:22 missing include Apr 14 21:50:28 ok Apr 14 21:55:11 woglinde, I disabled a bunch of modules, all the the ones gdb was crashing into. problem persists, but I hsve no backtrace left to know what is happening :/ Apr 14 21:55:27 feels like bad code generation Apr 14 21:55:30 buserror uh hm Apr 14 21:59:38 I'm going to try the same package in qemu. qemu is more lax than the real 920 Apr 14 22:00:55 buserror I only tested directfb on x86 Apr 14 22:03:15 I could try on a real one this week... Apr 14 22:04:03 hi florian :-) Apr 14 22:04:11 * Jay7 -> sleep Apr 14 22:05:02 at the end of the month I will get the beagleboard from robert Apr 14 22:05:08 hi BusError Apr 14 22:05:42 BusError: did you ever got my mail? :) Apr 14 22:06:25 woglinde: that's not an arm920t ;) Apr 14 22:06:34 florian right Apr 14 22:06:38 but its an arm Apr 14 22:07:15 florian, oops. tought I had replied to you sorry ! I closed #mini2440 because of the harrasment. it was really getting silly. Apr 14 22:07:57 BusError: heh... you need a good drink ;) Apr 14 22:08:12 woglinde, OK I can confirm the same exact -image works in qemu. do it must be the tuning flags not passed to the directfb modules Apr 14 22:08:42 florian, sorry about that. for 5% of good people. the rest was just too irritating :> feel free to ping me anytime tho :-) Apr 14 22:08:49 buserror hm you could check the logs Apr 14 22:11:17 woglinde, going to try to dig that out. wish me luck in there :> Apr 14 22:11:24 * BusError takes a deep breath Apr 14 22:11:55 buserror hm If I remeber correctly you have to pass a verbose flag to cmake Apr 14 22:12:47 woglinde, hmm I must say I have a vague idea where the logs live, but thats the extent of my knowledge :/ Apr 14 22:12:57 *g* Apr 14 22:13:19 oetmp/work/$arch/$recipename/tmp/log_do Apr 14 22:13:51 I'm of the "bitbake and pray hard" openembeded church :> Apr 14 22:18:54 woglinde, are the "-march" and other flags used and rolled into the "CC" or are they part of a CFLAGS ? (I assume the CC) Apr 14 22:19:20 they are part of CC Apr 14 22:20:14 OK. I suspected so thanks Apr 14 22:23:20 mmm...I don't find which include I should include Apr 14 22:23:36 by the way I don't need modutils at first Apr 14 22:23:48 (cause I will need different modutils) Apr 14 22:23:59 (cause I've a 2.4 kernel) Apr 14 22:28:02 woglinde, the compile log looks correct. altho it uses a few extra optimisation flags that were not passed in the configure Apr 14 22:28:12 (/oetmp_pma/staging/arm-oabi-angstrom-linux-uclibc/usr/include $ grep sys_init_module -r * gives nothing) Apr 14 22:29:15 ah module.h Apr 14 22:29:18 mabe it's there Apr 14 22:30:37 buserror hm yes directfb is very agressive at this point Apr 14 22:30:58 maybe you could try do remove the configure flags from the Makefiles Apr 14 22:31:01 and try again Apr 14 22:31:57 it looks like gcc is generating bad 920 code for one of these flags Apr 14 22:33:54 mmm....module.h is included Apr 14 22:34:10 and if it didn't found it would have told it Apr 14 22:34:10 woglinde, http://pastebin.com/m6283dbfd <- the "expensive-optimizations" and crew Apr 14 22:35:02 buserror ah Apr 14 22:35:08 you are compiling thumb Apr 14 22:35:26 maybee thats the root of this evil Apr 14 22:36:11 well. the tune file forces thumb somehow Apr 14 22:36:13 could you try with ARM_INSTRUCTION_SET = "arm" Apr 14 22:36:28 in the directfb recipes Apr 14 22:36:40 and I have to go to sleep now Apr 14 22:36:49 thanks anyway Apr 14 22:39:00 i'll go bye Apr 14 22:59:26 hmm "arm" crashes with an illegal instruction. seems one of the optimisation disables -marvh restrictions in both thumb and arm Apr 14 23:00:08 How do I control which kernel modules get included in my rootfs? Apr 14 23:02:09 BusError okay good to know Apr 14 23:03:20 woglinde, I'll try to see how I can disable it somehow. how can I bitbake equivalent of "make clean, make all " without wiping the work repipe dir ? Apr 14 23:03:33 hm Apr 14 23:03:44 buserror you can only delete the stampfile Apr 14 23:05:03 which is under oetmp/stamps/$arch/$recipename.do_compile and others Apr 14 23:08:35 okay good nite Apr 14 23:08:38 now really Apr 14 23:28:03 Okay. I know I ask stupid questions, but I'm stumped again. Apr 14 23:28:30 I build several dozen modules for my kernel, but almost none of them are actually getting included in my rootfs. Apr 14 23:28:49 Adding "kernel-modules" to my EXTRA_RRECOMMENDS in my machine's conf does nothing. Apr 14 23:28:54 Any suggestions? Apr 14 23:30:20 * khem should also look for a beagleboard Apr 14 23:34:21 MostAwesomeDude: you could specify them one by one Apr 14 23:34:37 like MACHINE_EXTRA_RRECOMMENDS += "kernel-module-ide-cs ..." Apr 14 23:34:44 khem: I'd rather not, considering that there's several dozen of them... Apr 14 23:35:07 Also it's rather braindead that omapfb isn't magically included. But whatever. I'd just like for it to work. Apr 14 23:36:06 MACHINE_EXTRA_RRECOMMENDS = " kernel-modules" Apr 14 23:36:12 doesnt work ? Apr 14 23:38:33 khem: Nope. Apr 14 23:48:28 mmm....how can I get the query_module function...I should have it in uclibc Apr 14 23:49:56 s/get/use Apr 14 23:50:52 maybe it's not exported? Apr 14 23:59:44 Gnutoo: isnt it a lnx kernel function Apr 15 00:00:10 it's for loading 2.4 kernel modules Apr 15 00:00:17 I think so Apr 15 00:01:10 yes it is Apr 15 00:01:48 its source is in /uClibc-0.9.30.1/libc/sysdeps/linux/common/query_module.c Apr 15 00:02:45 http://rafb.net/p/xHYCee65.html Apr 15 00:03:44 ah ok I understand...it should be because I've built with 2.6 headers but I was told that there were no problem to do so Apr 15 00:03:46 but: Apr 15 00:03:52 *I had nptl problems Apr 15 00:03:57 *and now I've this problem Apr 15 00:04:06 I would say if you have kernel-header problems Apr 15 00:04:20 hmmm Apr 15 00:04:28 2.6 and 2.4 is wider gap Apr 15 00:04:31 what should I do? recompile uclibc? change the kernel headers? Apr 15 00:04:37 both Apr 15 00:04:54 I'm stuck with 2.4 on my device...I can't change the kernel... Apr 15 00:04:55 use kernel headers which are more closer to your real kernel Apr 15 00:05:05 ok thanks Apr 15 00:05:05 and then recompile uclibc using those headers Apr 15 00:05:12 ok Apr 15 00:05:22 thanks a lot Apr 15 00:05:25 I am sure there are santized headers for 2.4 Apr 15 00:05:42 look around. If you dont find we can cobble something together Apr 15 00:05:44 what do you mean by sanitized? Apr 15 00:05:50 * khem leaves Apr 15 00:05:51 ok thanks a lot Apr 15 00:06:41 fixed the directfb problem, by upgrading to directfb 1.3 Apr 15 00:47:45 dangit how do i fix a bitbake md5sum oopz **** ENDING LOGGING AT Wed Apr 15 02:59:57 2009