**** BEGIN LOGGING AT Tue Aug 17 02:59:57 2010 Aug 17 04:42:17 03Khem Raj 07org.openembedded.dev * re6584742e0 10openembedded.git/recipes/docbook-sgml-dtd/docbook-sgml-dtd-4.5-native.bb: Aug 17 04:42:17 docbook-sgml-dtd-4.5-native.bb: New recipe. Aug 17 04:42:17 Signed-off-by: Khem Raj Aug 17 04:42:20 03Khem Raj 07org.openembedded.dev * r353f19af2d 10openembedded.git/recipes/docbook-utils/docbook-utils-native_0.6.14.bb: Aug 17 04:42:20 docbook-utils-native_0.6.14.bb: Get rid of do_stage Aug 17 04:42:20 Signed-off-by: Khem Raj Aug 17 06:24:58 03Martin Jansa 07org.openembedded.dev * r9980198df3 10openembedded.git/conf/distro/include/sane-toolchain-eglibc.inc: Aug 17 06:24:59 eglibc: limit -ggdb3 work-around to arm (as suggested by khem) Aug 17 06:24:59 * -ggdb3 breaks build on mips (as reported by Graham) Aug 17 06:24:59 Signed-off-by: Martin Jansa Aug 17 06:50:12 gm Aug 17 06:50:30 gm Aug 17 06:51:07 JaMa: nice work on python Aug 17 06:51:15 not sure if I have time tonight to test it Aug 17 06:52:13 I'll ask mickey about -devel/-dev change and if he agree I'll push it Aug 17 06:58:19 good plan Aug 17 07:02:04 i'll also try to test the x patches. They look ok to me. also this will be wed or thu Aug 17 07:05:31 I am using this command: CROSS_COMPILE=arm-angstrom-linux-gnueabi- make Aug 17 07:05:32 And i got this error: /home/ndemir/angstrom/angstrom-setup-scripts/build/tmp-angstrom_2008_1/sysroots/i686-linux/usr/armv7a/lib/gcc/arm-angstrom-linux-gnueabi/4.3.3/../../../../arm-angstrom-linux-gnueabi/lib/libgcc_s.so, not found Aug 17 07:09:35 ndemir: so that file does not exist Aug 17 07:10:04 why? i just installed the toolchain Aug 17 07:10:12 ndemir: and that is not really an OE question, in OE you do not invoke things with CROSS_COMPILE and make Aug 17 07:10:28 if you want to build something with oe use bitbake and bake a recipe Aug 17 07:10:35 manually i compiled the package to understand the problem Aug 17 07:11:00 eFfeM_work: it says libc.so not found in another error. Aug 17 07:11:07 do you have and idea? Aug 17 07:11:32 ndemir: it says ligbcc_s.so not found Aug 17 07:11:45 warning: libc.so, needed by /home/ndemir/angstrom/angstrom-setup-scripts/build/tmp-angstrom_2008_1/sysroots/i686-linux/usr/armv7a/lib/gcc/arm-angstrom-linux-gnueabi/4.3.3/../../../../arm-angstrom-linux-gnueabi/lib/libgcc_s.so, not found (try using -rpath or -rpath-link) Aug 17 07:12:09 did you build glibc (or eglibc or uclibc) ? Aug 17 07:12:17 and installed it ? Aug 17 07:18:11 eFfeM_work: should i use the command: bitbake glibc? Aug 17 07:18:31 maybe, don't know exactly what you did before Aug 17 07:18:44 better do something like bitbake helloworld Aug 17 07:18:58 that will compile and install the complete toolchain Aug 17 07:19:25 and whatever you want to compile might require additonal libs Aug 17 07:19:43 * eFfeM_work strongly suggests to make a recipe for it, peek at other recipes to get started Aug 17 07:22:18 eFfeM_work: look at this -> http://pastebin.com/Upm2bu03 Aug 17 07:22:25 03Martin Jansa 07org.openembedded.dev * ra5052d3889 10openembedded.git/recipes/gnutls/ (4 files in 2 dirs): Aug 17 07:22:26 gnutls: add version 2.10.1 Aug 17 07:22:26 Signed-off-by: Martin Jansa Aug 17 07:24:29 ndemir it has some problems with the compiler, i've no idea what you did upfront but I would suggest rm -rf /home/ndemir/angstrom/angstrom-setup-scripts/build/tmp-angstrom_2008_1 Aug 17 07:24:30 then Aug 17 07:24:35 bitbake helloworld Aug 17 07:24:48 eFfeM_work: i will do that Aug 17 07:24:56 it will give you a clean start Aug 17 07:25:03 guess you are building angstrom for beagleboard Aug 17 07:25:38 taht is build quite often and should work without hassle Aug 17 07:25:45 gl Aug 17 07:37:18 eFfeM_work: i am trying to compile lshw, http://packages.ubuntu.com/karmic/lshw. you can see depends on that page. what should i write for DEPENDS in bb file? Aug 17 07:39:27 ndemir: DEPENDS = "pciutils usbutils" Aug 17 07:39:44 no need to add depends for libc or so, these can be assumed to be present Aug 17 07:40:03 hımm ok thanks eFfeM_work Aug 17 07:40:10 yw Aug 17 07:41:19 ndemir: btw the first package is the hardest one :-) Aug 17 07:41:52 but unlike school it is allowed to cheat and look at other recipes :-) Aug 17 07:44:36 eFfeM_work: :) Aug 17 07:52:40 Noor: saw your patches but most likely I won't have time to look at them today (and push them) Aug 17 08:00:39 OK Aug 17 08:00:42 no problem Aug 17 08:10:00 hi all Aug 17 08:14:41 Noor: but don't let that stop you from submitting new patches Aug 17 08:14:43 hi gremlin[it] Aug 17 08:22:22 eFfeM_work: no I am sending new ones Aug 17 08:22:34 ok Aug 17 08:32:51 good morning Aug 17 08:35:05 03Martin Jansa 07org.openembedded.dev * r9debd2fea1 10openembedded.git/ (31 files in 5 dirs): Aug 17 08:35:06 python: move 2.6.[124] to obsolete, move all distributions to 2.6.5 Aug 17 08:35:06 Signed-off-by: Martin Jansa Aug 17 08:35:06 Acked-by: Michael 'Mickey' Lauer Aug 17 08:35:07 florian: moin Aug 17 08:35:21 03Martin Jansa 07org.openembedded.dev * rb16899ec9e 10openembedded.git/recipes/python/ (18 files in 2 dirs): Aug 17 08:35:21 python: import 2.6.5 from poky Aug 17 08:35:21 * without legacy staging Aug 17 08:35:21 * fix QA issue libpython2: non -dev package contains symlink .so: libpython2 path Aug 17 08:35:22 * use FILESPATHPKG Aug 17 08:35:22 Signed-off-by: Martin Jansa Aug 17 08:35:23 Acked-by: Michael 'Mickey' Lauer Aug 17 08:35:37 florian: btw do you remember if there is a way to upload screenshot with gpe-scap in non-interactive way? Aug 17 08:36:41 JaMa: hum no I don't think so but this might be easy to add Aug 17 08:38:12 florian: quite often I'm testing something remotely without seeing what's shown on display, so it would be usefull Aug 17 08:39:32 JaMa: yes true, I missed such a function some time ago too. Aug 17 08:46:09 hi eFfeM_work Aug 17 08:46:15 hi florian ... Aug 17 08:46:32 hi gremlin[it] Aug 17 08:47:57 I have a problem, module-init-tools-cross refuse to compile cause a "/usr/bin/ld: cannot find -lc" ... target is atngw100 angstrom uclibc ... i notice some other had simila trouble but i don't find a solution ... Aug 17 08:48:01 hi florian ! Aug 17 08:50:05 eFfeM_work: how does the bitbake compile? does it use make CROSS_COMPILE=arm-angstrom-linux-gnueabi- command? Aug 17 08:50:05 Aug 17 08:50:43 it does lots of things internally but it make sure paths and prefixes are correct Aug 17 08:50:49 do you have a problem? Aug 17 08:52:24 eFfem_work: ERROR: runstrip: ''arm-angstrom-linux-gnueabi-strip' --remove-section=.comment --remove-section=.note '/home/ndemir/angstrom/angstrom-setup-scripts/build/tmp-angstrom_2008_1/work/armv7a-angstrom-linux-gnueabi/lshw-B.02.14-r0.0/package/usr/sbin/lshw'' strip command failed Aug 17 08:52:27 i have this error Aug 17 08:52:58 guess the compile then worked Aug 17 08:53:06 i checked the lshw binary. it is not a binary works in arm Aug 17 08:53:07 can you pastebin the whole log ? Aug 17 08:53:15 ah ok Aug 17 08:53:26 then thelog is probaby not that useful Aug 17 08:53:37 eFfeM_work: yes. you are right. Aug 17 08:53:48 eFfem_work: what should i do? Aug 17 08:53:54 hn Aug 17 08:53:56 hm Aug 17 08:54:20 does the package use automake Aug 17 08:54:26 and can you pastebin your recipe ? Aug 17 08:54:35 it has makefile Aug 17 08:55:19 it has only makefile not ./configure Aug 17 08:55:28 ok then show the recipe Aug 17 08:55:31 ok Aug 17 08:55:39 good morning Aug 17 08:56:48 hi ant_work Aug 17 08:56:58 03Dmitry Eremin-Solenikov 07org.openembedded.dev * r3dfdd75198 10openembedded.git/recipes/lowpan-tools/ (lowpan-tools_0.2.1.bb lowpan-tools_0.2.2.bb): Aug 17 08:56:58 lowpan-tools: update to 0.2.2 Aug 17 08:56:58 Signed-off-by: Dmitry Eremin-Solenikov Aug 17 08:57:32 eFfem_work: lshw.inc -> http://pastebin.com/3wgW2vXN Aug 17 08:57:53 eFfem_work: lshw_B.02.14.bb -> http://pastebin.com/ABVNFrkg Aug 17 08:58:24 ndemir: inherit autotools should not be there if there are no autotools files (no .ac or .m4 or .in files or so) Aug 17 08:58:46 ok. what should i write? automake? Aug 17 08:58:56 no, one sec Aug 17 09:00:10 eFfeM_work: ok. i am waiting. Aug 17 09:00:20 trying to find a good example Aug 17 09:01:09 ok Aug 17 09:01:43 can't find a good example but you need to add do_compile and do_install functions Aug 17 09:01:54 e.g Aug 17 09:01:55 do_compile() { Aug 17 09:01:55 oe_runmake 'CXX=${CXX}' Aug 17 09:01:55 } Aug 17 09:02:03 the CXX part is probably not needed Aug 17 09:02:20 i tried it before. but i will try it again. Aug 17 09:02:42 eFfeM_work: what if there is configure script and not makefile? Aug 17 09:02:46 and do_install should call oe_runmake install Aug 17 09:03:21 if there is a configure script and also things like configure.in you probably need to use autotools Aug 17 09:04:32 but you do not need it for thsi Aug 17 09:06:18 eFfeM_work: i added do_compile and do_install. here is the error -> http://pastebin.com/FmZbRk3n Aug 17 09:07:12 there should be more, it should mention a logfile Aug 17 09:08:50 eFfeM_work: should i use inherit for this? Aug 17 09:09:11 you are missing a lib path Aug 17 09:10:34 it might be that your recipe uses a hardcoded path, then you need a patch it might also be that you need to set a dir Aug 17 09:11:06 eFfeM_work: you saw my bb file. Aug 17 09:11:09 that you need to set an override to a dir that specifies the place where the linkfile is found Aug 17 09:11:18 yeah, but that is in the makefile, it depends on the package Aug 17 09:11:49 it finds but says "incompatible" Aug 17 09:11:59 eFfem_work: /home/ndemir/angstrom/angstrom-setup-scripts/build/tmp-angstrom_2008_1/sysroots/i686-linux/usr/armv7a/lib/gcc/arm-angstrom-linux-gnueabi/4.3.3/../../../../arm-angstrom-linux-gnueabi/bin/ld: skipping incompatible ./core//liblshw.a when searching for -llshw Aug 17 09:12:32 either you compiled with the wrong CC or you are using a wrong path Aug 17 09:13:00 in do_compile try: Aug 17 09:13:09 oe_runmake CC="${CC} ${LDFLAGS}" Aug 17 09:13:24 do a bitbake -c clean lshw Aug 17 09:13:27 then bitbake lshw Aug 17 09:14:31 ok now i am trying Aug 17 09:18:50 any clue form my trouble ? http://tinderbox.openembedded.net/packages/696809/ Aug 17 09:22:29 gremlin[it]: cannot find -lc, do you have glibc buidl and installed Aug 17 09:23:54 eFfeM_work, i'm using uclibc ... libc.a .so is not present into the dir where gcc look for it ... but uclibc and uclibc-initial are both build with success ... Aug 17 09:25:49 gremlin[it]: hm, strange, there is a libc.a somewhere else in sysroots (don't know too much abut uclibc Aug 17 09:26:37 sysroots/avr32-angstrom-linux-uclibc/usr/lib/libc.a Aug 17 09:27:03 and on same place also shared ones Aug 17 09:27:37 and where does gcc want it ? Aug 17 09:27:46 actually have a similar issue Aug 17 09:28:02 afk for a moment Aug 17 09:28:23 sysroots/x86_64-linux ... (my host is a x86_64 Aug 17 09:36:48 gremlin[it]: you can find out the paths used by gcc with Aug 17 09:36:50 echo "" | gcc -c -E -v - Aug 17 09:37:11 and gcc -print-search-dirs Aug 17 09:38:07 inside devshell ? Aug 17 09:41:28 Anyone an idea why xtscal doesn't seem to work any more? Aug 17 09:45:00 florian: what goes wrong with it? Aug 17 09:47:24 pb_: The cross wasn't displayed and I got an error message when I tapped on the screen last night. Aug 17 09:47:45 hrm... left the affected device at home Aug 17 09:49:55 It was somethign with XReply Assertion !dpy->xcb->reply_data Aug 17 09:50:06 hi lrg Aug 17 09:52:10 eFfeM_work, plain (native) gcc obviously search in my host paths ... if you see the log there is full specification of search paths on command line ... Aug 17 09:52:28 seem there isn't a "cross uclibc" library installed ... Aug 17 09:54:27 hm, not really an idea (not an uclibc/toolchain lib) Aug 17 09:55:01 for the cmds I gave, guess you can do it in devshell, I've also used them directly and replaced gcc by with the full path of the cross gcc Aug 17 09:58:00 wolla wolla wolla Aug 17 09:58:10 hi XorA ! Aug 17 09:58:13 wolla ? Aug 17 10:00:04 florian: about xtscal some days ago: http://apt.rikers.org/%23oe/20100701.html.gz Aug 17 10:00:12 on the top Aug 17 10:00:41 ant_work: yes this looks familiar Aug 17 10:01:12 some changes to libxcb? added to xorg? Aug 17 10:01:23 btw opie is working again Aug 17 10:01:34 just gpe was failing when I tried Aug 17 10:01:52 I tested opie just a few days ago Aug 17 10:02:50 I doubt Opie would use libxcb Aug 17 10:03:12 iirc was some udev side-issue Aug 17 10:03:16 or xtscal :-D Aug 17 10:03:29 opie had the crosshairs :p Aug 17 10:04:23 This reminds me to a couple of xtscal patches Aug 17 10:04:47 florian: hm, right. that sounds like an xcb bug I guess. Aug 17 10:05:01 possibly the XCALIBRATE extension is doing something that it doesn't like. Aug 17 10:10:43 Sounds like a lot of fun debugging... a job for later. Aug 17 10:11:22 I need to get Android running on a board first *sigh* Aug 17 10:15:06 mickey|office: good morning Aug 17 10:15:18 good morning pb_ Aug 17 10:15:32 pb_: did anything important happen on the last TSC meeting? Aug 17 10:15:52 eFfeM_work: which package should i bitbake to obtain libc? Aug 17 10:16:18 NB: took me 6 hours to get back from munich... weather was really lousy and there were dozens of construction sites on the Autobahn Aug 17 10:17:51 mickey|office: no, nothing very noteworthy. I don't think we had much business to discuss. Aug 17 10:18:01 mickey|office: oh dear. how long should it take normally? Aug 17 10:18:11 3.40 - 4.00 Aug 17 10:18:18 depending on traffic Aug 17 10:18:43 it's ~ 400 km Aug 17 10:18:59 ah right. Aug 17 10:19:16 only an average of 100km/h? that doesn't sound right for a bmw on the autobahn :-} Aug 17 10:19:41 heh Aug 17 10:19:48 surely with your new vehicle you should be able to travel that distance in about 1h! Aug 17 10:20:04 hehehe Aug 17 10:20:13 indeed, the new 330i is a monster Aug 17 10:20:24 pity i usually don't drive that much of Autobahn Aug 17 10:20:55 i might change that now :) Aug 17 10:20:58 heh Aug 17 10:21:55 :-) Aug 17 10:23:13 I read a review of the 330d some time ago... the 330i might be even more fun :) Aug 17 10:26:49 ya, it's quite amazing acceleration-wise. although the convertible with the steel roof is quite heavy Aug 17 10:29:17 eFfeM_work: I need some help regarding sgml-common-native recipe Aug 17 10:29:38 Noor: about to go for lunch will be back in 30 mins Aug 17 10:29:43 OK Aug 17 10:30:01 but feel free to describe your issue; maybe someone else can help and otherwise I'll read it when I am back Aug 17 10:30:10 BTW in which country you are living? Aug 17 10:30:40 * florian needs a new car too Aug 17 10:31:23 I am removing old style staging and replacing do_stage with do_install function in sgml-common-native recipe Aug 17 10:33:05 during compilation a script is created which is called install-catalog. In this script the path to tmp/sysroot.i686-linux/etc/sgml is hardcoded using during do_compile function Aug 17 10:33:51 so when this script runs its creats a file in tmp/sysroot/i686-linux/stc/sgml with name catalog Aug 17 10:34:36 in old scenario the updated files in sysroot goes in staging_pkg directory Aug 17 10:37:17 but in new style staging this does not happen. rather content from sysroot-destdir are copied to staging-pkg directory Aug 17 10:37:54 so I am not having catalog file in ipk file Aug 17 10:38:49 because thing are not coming to staging-pkg directory from tmp/sysroot but from sysroot-destdir Aug 17 10:40:22 sounds like you need to fix that script to install into the correct place. Aug 17 10:40:28 i.e. ${D} Aug 17 10:40:54 is it right to add ${D} in script Aug 17 10:42:07 i was confused because if this file will be used from anyother utility it will create catalog at the sgml-common-native image directory Aug 17 10:42:15 isn't it Aug 17 10:42:36 oh, the script itself is installed? Aug 17 10:42:40 that is a bit of a nuisance Aug 17 10:42:59 I guess you could have it read ${D} from the environment rather than expanding it inline. Aug 17 10:45:10 I think I got it Aug 17 10:45:30 i'll ping u again if i face any problem Aug 17 10:46:03 ok, cool Aug 17 11:50:58 pb_: I change the install-catalog script to echo "$SGML_POINTER $SGML_CENTRALIZED" >> ${D}/home/nahsan/SB/open_comm/tmp/sysroots/i686-linux/etc/sgml/catalog Aug 17 11:51:23 previosly it was echo "$SGML_POINTER $SGML_CENTRALIZED" >> /home/nahsan/SB/open_comm/tmp/sysroots/i686-linux/etc/sgml/catalog without ${D} Aug 17 11:52:00 but it is still creating catalog file at /home/nahsan/SB/open_comm/tmp/sysroots/i686-linux/etc/sgml instead of image Aug 17 11:55:38 when ipk file is being created the function parses and replaces all staging_dir paths with 'FIXMESTAGINGDIR' Aug 17 11:56:33 fahad usman here? pls see my reply on your cacaoh patch Aug 17 11:57:36 so when i changes install-catalog --add ${sysconfdir}/sgml/sgml-ent.cat ${STAGING_DATADIR}/sgml/sgml-iso-entities-8879.1986/catalog with install-catalog --add ${sysconfdir}/sgml/sgml-ent.cat ${D}${datadir}/sgml/sgml-iso-entities-8879.1986/catalog Aug 17 11:57:45 Noor, don't know anything about this recipe or catalog files, but guess you want to redefine FIXMESTAGINGDIR Aug 17 11:58:28 but don't think you should explicitly have dirs like i686-linux in a patch, this will fail on 64 bit Aug 17 11:58:35 03Simon Busch 07org.openembedded.dev * r3a3ffb7d7b 10openembedded.git/conf/machine/palmpre.conf: Aug 17 11:58:35 palmpre: use omap3530beagle config for x-load Aug 17 11:58:35 We currently have no palmpre support for x-load. As long as we miss it we use the config Aug 17 11:58:35 from the beagleboard machine. Aug 17 11:58:35 Signed-off-by: Simon Busch Aug 17 11:58:36 Acked-by: Stefan Schmidt Aug 17 11:58:44 03Simon Busch 07org.openembedded.dev * r7225589793 10openembedded.git/recipes/u-boot/u-boot_git.bb: Aug 17 11:58:45 u-boot: bump git revision for palmpre machine Aug 17 11:58:45 The current git revision for palmpre machine fails to compile. I updated it to the latest Aug 17 11:58:45 revision available as u-boot is currently not usable on the palmpre machine at all. Aug 17 11:58:45 Signed-off-by: Simon Busch Aug 17 11:58:45 Acked-by: Stefan Schmidt Aug 17 11:58:46 03Simon Busch 07org.openembedded.dev * ra6e9154445 10openembedded.git/recipes/linux/ (linux-palmpre/defconfig linux-palmpre_git.bb): Aug 17 11:58:46 linux-palmpre: add git version Aug 17 11:58:47 The sources from Palm for the Palm Pre has beed modified with little improvements and are Aug 17 11:58:47 rebased against the latest version from Palm found at http://opensource.palm.com from time Aug 17 11:58:48 to time. The modified kernel tree is maintained by the FSO team. Aug 17 11:58:48 Signed-off-by: Simon Busch Aug 17 11:58:49 Acked-by: Stefan Schmidt Aug 17 11:58:57 03Simon Busch 07org.openembedded.dev * r7f397fe647 10openembedded.git/recipes/tslib/ (tslib.inc tslib/palmpre/ts.conf tslib/palmpre/tslib.sh): Aug 17 11:58:57 tslib: add config files for palmpre machine Aug 17 11:58:57 Signed-off-by: Simon Busch Aug 17 11:58:57 Acked-by: Stefan Schmidt Aug 17 12:00:13 this is coming from variable Aug 17 12:00:41 sysconfdir Aug 17 12:01:01 hi all Aug 17 12:01:06 i have a quick question Aug 17 12:01:13 hi Fahad_ Aug 17 12:01:14 ask Aug 17 12:01:17 i am modifying some recipes Aug 17 12:01:23 btw send an email in reply to your latest patch Aug 17 12:01:30 i am removing the legacy staging Aug 17 12:01:37 from a bunch of native recipes Aug 17 12:01:52 and also passing them through oe-stylize Aug 17 12:01:56 add Aug 17 12:01:57 NATIVE_INSTALL_WORKS = "1" Aug 17 12:02:11 so, should i subit one patch for both the changes or two seperate patches Aug 17 12:02:20 where you removed legacy staging Aug 17 12:02:27 yes, i am doing that already Aug 17 12:02:27 better 2, as they are not related Aug 17 12:02:34 or only loosely Aug 17 12:02:51 didn't see the NATIVE_INSTALL_WORKS in your patch Aug 17 12:05:40 okay, should i send another patch for cacaoh-native? Aug 17 12:06:28 eFfeM_work: FIXMESTAGINGDIR is just a string Aug 17 12:06:39 ah ok Aug 17 12:07:29 I think it is replaced with STAGING_DIR Aug 17 12:07:31 in staging_packager all files that are to be packages are parsed and staging dir path is replaced with this string Aug 17 12:07:35 yeah Aug 17 12:07:51 found these two lines: Aug 17 12:07:53 classes/packaged-staging.bbclass: sed -i -e s:${STAGING_DIR}:FIXMESTAGINGDIR:g $i Aug 17 12:07:53 classes/packaged-staging.bbclass: os.system("sed -i -e s:FIXMESTAGINGDIR:%s:g %s" % (staging, tmpdir + '/' + file)) Aug 17 12:08:08 but don't have any experience in that area (and currently not too much time) Aug 17 12:08:48 OK Aug 17 12:11:03 Can you refer any other person Aug 17 12:12:08 well if there is someone here that can help, I think they will speak up Aug 17 12:12:37 Fahad_: if the line is missing send a new patch (after testing etc). please label it v2 Aug 17 12:56:48 eFfeM_work: should i still add NATIVE_INSTALL_WORKS even if it does not affect the resulting .ipk file and i don't see any message in the log indicating that legacy staging is being used Aug 17 13:09:13 Fahad_: are the files copied to the stagin dir if you have no NATIVE_INSTALL_WORKS ? Aug 17 13:09:42 let me verify, just a min Aug 17 13:11:46 Fahad, here is what kergoth said about it: http://lists.linuxtogo.org/pipermail/openembedded-devel/2010-August/022885.html Aug 17 13:12:39 btw kergoth is also with mentor and has far more knowledge on this than I have Aug 17 13:16:35 eFfeM_work: i get exactly the same contents in the staging directory with or without NATIVE_INSTALL_WORKS = "1" Aug 17 13:17:11 Ah ok, then afaik it is ok to leave it out (unless kergoth__ or someone else says otherwise) Aug 17 13:18:19 eFfeM_work: okay, then i think i should wait for their response, in the meanwhile i am replying to the mailing list Aug 17 13:18:56 ok, i probably won't have time to test and commit it today anyway Aug 17 13:19:16 but ofc if someone else says it is ok and pushes it that is equally fine with me Aug 17 13:20:57 eFfeM_work: okay, thanks Aug 17 13:39:23 03Koen Kooi 07org.openembedded.dev * r65cd015e62 10openembedded.git/recipes/vnc/x11vnc_0.9.11.bb: x11vnc: add 0.9.11 Aug 17 14:16:10 anybody know how to get OE to add an initramfs to the kernel? Aug 17 14:19:57 Not 100% but there's recipes for that dance, yes Aug 17 14:20:06 kexec related at least Aug 17 15:11:40 eFfeM_work: in uicmoc3-native recipe OE_QMAKE_MOC variable is used and this variable is define in qmake.bbclass to ${STAGING_BINDIR_NATIVE}/moc Aug 17 15:12:09 is it right to replace this variable with ${D}{bindir}/moc in recipe file? Aug 17 15:31:07 Noor: this variable assumes different dirs based on what it is compiling for Aug 17 15:31:34 Noor: in this case you could replace it with ${D}{bindir}/moc3 Aug 17 15:31:52 in the recipe file? Aug 17 15:32:19 yes Aug 17 15:32:45 Thanks. Thats what I did just want to confirm Aug 17 15:44:47 Noor, I'm gone again in a few minutes, but I saw khem gave you a better answer than I could Aug 17 15:45:05 not sure when I'm back so not sure whether I have time to peek at your patches Aug 17 15:45:17 No problem Aug 17 15:45:33 Noor, btw if you want to you could also try to convert some recipes to BBCLASSEXTEND = "native" Aug 17 15:45:35 afk Aug 17 16:18:33 hi all ! Aug 17 16:18:57 gremlin[it]: hello long time Aug 17 16:19:12 yes really long time ! Aug 17 16:19:20 what have you been upto Aug 17 16:20:13 mhhh many things, health trouble, lot of work, a broken relationship ... Aug 17 16:20:56 at now i have a little trouble with uclibc and NGW100 ... talking about OE :P Aug 17 16:23:50 gremlin[it]: I see. I hope all is well now Aug 17 16:24:04 what sort of trouble with uclibc Aug 17 16:24:05 khem, yes seem yes ... Aug 17 16:24:40 khem, http://tinderbox.openembedded.net/builds/83079/ Aug 17 16:25:55 gremlin[it]: thats your host issue Aug 17 16:27:14 khem, ... i need libc.a on my distro ? Aug 17 16:31:13 gremlin[it]: probably Aug 17 16:32:29 khem, let's try ... i was stucked looking who in OE could provide uclibc'd libc.a ... hahaha Aug 17 16:42:47 khem, you was right !!! Aug 17 16:43:50 heh Aug 17 16:43:56 I did that yesterday also Aug 17 16:44:03 we should add that check to sanity Aug 17 16:46:33 Crofton, yes was a subtle error ... Aug 17 16:47:06 Crofton, maybe i try to do something ... Aug 17 16:48:44 yeah Aug 17 16:48:56 I bites everytime you install on a new machine Aug 17 16:49:56 Crofton, what recipe i can check to learn how to do a sanity check ? Aug 17 16:51:32 I think those checks are done in the sanity class Aug 17 16:51:48 mhh ok ... i try to take a look Aug 17 16:52:56 hmmm Aug 17 16:52:58 | ./libpython2.6.so: undefined reference to `_PyParser_Grammar' Aug 17 16:53:01 that rings a bell Aug 17 16:53:31 indeed, i have seen this before Aug 17 16:53:57 maybe something like ld vs gcc used for linking? Aug 17 16:54:27 florian, no seem do with gcc --> http://tinderbox.openembedded.net/public/logs/task/7202085.txt Aug 17 16:54:41 at least with module-init-tools-cross Aug 17 16:55:36 gremlin[it]: this doesn't have to count for python Aug 17 16:56:20 ok Aug 17 16:57:05 have to go, bye Aug 17 16:57:57 GNUtoo|laptop: what version of python? Aug 17 16:58:28 I've seen that as possibly a build time race in ptyhon which i thought was fixed in more recent (ie not 2.6.[12]) versions Aug 17 16:58:39 ah parallel build issues it seem Aug 17 16:58:44 this time I bet it built Aug 17 16:58:54 I'll look the version Aug 17 17:01:07 work/x86_64-linux/python-native-2.6.5-ml11.0/temp$ tail log.do_compile.31953 =>./libpython2.6.so: undefined reference to `_PyParser_Grammar' Aug 17 17:01:44 the other do_compile has that: Aug 17 17:02:06 Yeah, damn Aug 17 17:02:10 Still there :( Aug 17 17:02:12 many warnings and some failed module Aug 17 17:02:20 Time to go debug the makefiles and fix it Aug 17 17:02:22