**** BEGIN LOGGING AT Sun Jul 11 02:59:57 2010 Jul 11 03:17:56 03Khem Raj  07org.openembedded.dev * r1f373112f4 10openembedded.git/recipes/quilt/ (5 files in 2 dirs): (log message trimmed) Jul 11 03:17:56 quilt: Add option QUILT_NO_RM_SYMLINKS so follow symlinks with quilt refresh Jul 11 03:17:56 * OE now creates symlinks into metadata. This patch allows quilt to Jul 11 03:17:56 follow the symlinks so when a patch is refreshed then the result Jul 11 03:17:56 happens in the metadata location. Jul 11 06:53:33 gm Jul 11 07:04:37 pb_ ping Jul 11 08:06:50 JFYI: OE profile on ohloh https://www.ohloh.net/p/3928 Jul 11 08:30:13 03Koen Kooi  07org.openembedded.dev * ra9b120b534 10openembedded.git/recipes/gcc/ (2 files): gcc csl 2007q3: hack to fix build Jul 11 09:04:39 eFfeM: hello Jul 11 09:06:42 pb__ wnated to discuss with you if it should be base.bbclass or cross.bbclass Jul 11 09:07:14 inbetween i did send an email as I'm probably gone in a few minutes and only have little time later today Jul 11 09:09:48 and wheter we should also introduce COMPATIBLE_ARCH Jul 11 09:10:32 yeah, I just replied to your mail Jul 11 09:10:42 I don't understand why COMPATIBLE_ARCH is necessary. Jul 11 09:11:58 well e.g. most gcc versions do not support nios2 in any flavour that's why _ARCH could be handy Jul 11 09:12:34 need to go now, back in 30-60 mins Jul 11 09:13:42 right, but you can check that with TARGET_SYS anyway Jul 11 09:13:58 (since TARGET_SYS is, in general, defined in terms of TARGET_ARCH + TARGET_OS) Jul 11 09:14:12 there doesn't seem any need to add a separate check for that Jul 11 10:06:47 yeah, figured that out about half an hour ago Jul 11 10:08:58 pb__ just read your email, somehow I assumed that cross.bbclas would be for recipes running on the host doing things for the target Jul 11 10:09:10 is this a wrong assumption? Jul 11 10:16:19 COMPATIBLE_MACHINE is also there and if COMPATIBLE_TARGET_SYS is only used for cross, it is still possible that the system tries to build a non-supported recipe (e.g. for if COMPATIBLE_TARGET_SYS gcc 4.x says it is for sys S only then non S systems will not build gcc-cross 4.x but with a gcc-cross it is still possible to build the gcc 4.x for S Jul 11 10:16:30 at least that is my understanding but I can be wrong here Jul 11 11:09:32 eFfeM: you're correct that cross.bbclass is for recipes running on BUILD_SYS and generating code for TARGET_SYS. Jul 11 11:19:27 good day! Jul 11 11:19:27 I have made an 2.6.28 image... now I want a 2.6.33 image.. how can I change OE to change the default kernel version when doing images? Jul 11 11:20:25 aszpain, in local.conf put something like Jul 11 11:20:26 PREFERRED_VERSION_ "2.6.33" Jul 11 11:20:47 my kernel package? Jul 11 11:20:54 aszpain, what is your machine ? Jul 11 11:20:59 IGEPv2 Jul 11 11:21:04 igep0020 Jul 11 11:21:51 aszpain, all kernel recipes can be found in recipes/linux/ Jul 11 11:22:07 or set DEFAULT_PREFERENCE_="1" in appropriate kernel recipe Jul 11 11:22:19 but local.conf is better Jul 11 11:22:47 so u cant upgrade to ANY kernel u want just the ones in the recipes? Jul 11 11:23:19 aszpain, sure you can. but you need to find out what .bb is providing your kernel first Jul 11 11:24:00 aszpain, obviously PREFERRED_VERSION_ "2.6.33" assumes that you have 2.6.33 version of that recipe present in OE Jul 11 11:24:19 hmm let me chekc if it is here Jul 11 11:24:28 recipes/linux/ u said right ? ok Jul 11 11:26:52 where is not found Jul 11 11:27:05 it only has the recipe for 28 Jul 11 11:27:09 :( Jul 11 11:27:24 so how can I make / get that .BB file? Jul 11 11:31:12 dcordes.... cant you just copy the .bb file of the 28 as a base for the 2.6.33 ? Jul 11 11:31:45 meta/packages/linux/linux-igep2_2.6.28.10-0.bb Jul 11 11:31:48 aszpain, that's what I usually do Jul 11 11:31:58 and... does it works? Jul 11 11:32:28 aszpain, look at the .bb file content and just try Jul 11 11:32:59 aszpain, on a site note the fact that your recipes are in packages rather than recipes shows that you are not using org.openembdded.dev branch of the oe source tree Jul 11 11:33:24 aszpain, so you are probably about to do double work Jul 11 11:34:02 hmmmm Jul 11 11:34:35 aszpain, 'git branch' to check Jul 11 11:34:38 git clone http://repo.or.cz/r/openembedded.git Jul 11 11:34:47 omg thats what my wiki said Jul 11 11:35:12 why my board provider said me to dl oe from here? Jul 11 11:35:29 please use 'git branch' to check what branch you on. it's the one with the asterisk Jul 11 11:36:21 * org.openembedded.dev Jul 11 11:36:24 yes it is then Jul 11 11:36:46 its ok Jul 11 11:37:56 pb_ but shouldn't COMPATIBLE_TARGET_SYS also avoid building target recipes that make no sense (e.g. building a binutils or gcc for a target that does not support that target also breaks things Jul 11 11:39:46 yeah, I guess so Jul 11 11:40:00 in absence of a generic toolchain.bbclass, maybe base.bbclass really is the least bad place for that check. Jul 11 11:40:16 it is a bit of a shame to introduce the extra overhead for the 99% of packages that won't benefit from it, but still. Jul 11 11:40:58 pb_ i have no better idea Jul 11 11:40:58 03Paul Eggleton  07org.openembedded.dev * r078d6eadc4 10openembedded.git/conf/machine/ (h4000.conf hx4700.conf): conf/machine: add keymaps to recommended list for h4000/hx4700 to fix record button Jul 11 11:41:09 03Paul Eggleton  07org.openembedded.dev * r4857a88f7e 10openembedded.git/recipes/keymaps/ (3 files in 3 dirs): keymaps: add keymaps for h4000 & hx4700 to make record button usable Jul 11 11:41:19 though, that said, building gcc for an unsupported target would choke fairly quickly during configure, so I am not quite sure what the extra check really buys you. Jul 11 11:41:57 pb_ if you agree with the patch in base.bbclass can you ack the original patch ? Jul 11 11:42:23 pb_ I want to avoid that distro's pull in a recipe that is not supported for that target Jul 11 11:42:48 aszpain, are you uptodate ? 'git pull' Jul 11 11:43:23 ok, fair enough I guess Jul 11 11:43:34 aszpain, you really should not have a packages dir Jul 11 11:44:55 is it just gcc and binutils that are going to do this? if so, maybe their own .inc files would be a better place. Jul 11 11:48:05 03Paul Eggleton  07org.openembedded.dev * r2cc5b6a073 10openembedded.git/recipes/opie-init/opie-init.inc: opie-init: disable special init handling for c7x0 (for w100 which is currently disabled) Jul 11 11:48:33 mickey|bbl, I am subscribed to the devel ml. few hours ago I sent couple of mails which are not visible in the archive yet. can you check ? Jul 11 11:50:37 could be that your server is greylisted, it will do that if you haven't posted before. Jul 11 11:51:31 pb_, it were the first mails I sent Jul 11 11:52:32 that's probably it then Jul 11 11:55:10 pb_ it is as far as I see it now gcc, binutils and *libc; haven't considered doing it locally, maybe it is better to have a generic solution, I can imagine it is useful elsewehre too Jul 11 11:55:37 eFfeM: libc isn't a cross tool, so you can just check for COMPATIBLE_HOST there. Jul 11 11:56:14 it's only the cross ones that are slightly complex. if that's just gcc and binutils then I would suggest doing it locally in their own recipes Jul 11 11:56:59 or in a compatible-toolchain.bbclass that they all inherit, if you prefer Jul 11 11:57:21 pb_ this is not only about cross tools, libc has target specific code it it and if libc does not run on an arch it is pointless to select it as a package or build it Jul 11 11:57:31 eFfeM: right, but you can check that with COMPATIBLE_HOST Jul 11 11:57:46 i.e. there is already a mechanism to do what you need for libc, there is no need to add a new one Jul 11 11:58:19 pb_ I agree with the last sentence but thougth COMPATIBLE_HOST only said something about hosts Jul 11 11:58:59 e.g. what would I add to glibc 2.4 if I want to specify this is not for nios2 Jul 11 11:59:10 COMPATIBLE_HOST = "(?!nios2)" Jul 11 11:59:39 or whatever the notation was for negative regexps Jul 11 12:00:33 your negate is ok, just thought COMPATIBLE_HOST was for saying that things do not run on the build host Jul 11 12:00:50 i think COMPATIBLE_HOST is not documented Jul 11 12:00:58 that would be COMPATIBLE_BUILD_SYS or something, which doesn't exist. Jul 11 12:01:06 COMPATIBLE_HOST just tests against ${HOST_SYS} Jul 11 12:01:20 pb_ found that Jul 11 12:02:13 thought HOST_SYS was for host Jul 11 12:02:30 it is. I suspect you are confused about what "host" means in this context. Jul 11 12:02:34 yes Jul 11 12:02:44 will try this later, got visitors Jul 11 12:02:50 thanks alot for your info Jul 11 12:03:21 see the autoconf manual, it has some explanation of the "build"/"host"/"target" terminology Jul 11 12:03:29 pb_, can you whitelist my mailserver then ? Jul 11 12:03:38 dcordes: no, but the greylist will sort it out eventually Jul 11 12:03:44 ok ^^ Jul 11 12:04:05 pb_, how long does that usually take ? Jul 11 12:04:16 dcordes: a few hours, I forget exactly how it's configured on discovery Jul 11 12:04:34 I don't have my ssh keys with me now but I can look at it later if your mail still hasn't been delivered. Jul 11 12:04:56 pb_, I sent the first mail 03:21:44 CET Jul 11 12:05:12 ah, it shouldn't have taken that long Jul 11 12:05:17 you didn't get a bounce, presumably? Jul 11 12:05:21 pb_, nope Jul 11 12:05:43 pb_, I have a discovery unprivileged user Jul 11 12:06:24 okay. if you ping me again later when I am back at home, I will have a poke at the logs and see what has gone wrong. unless mickey|bbl returns and sorts you out before then. Jul 11 12:06:32 or flo_lap, he also has root on discovery Jul 11 12:09:11 pb_, ok thanks a lot Jul 11 12:10:03 pb_ will look at the manual, will get back to this later Jul 11 12:10:06 bye Jul 11 12:10:13 righto Jul 11 12:16:43 newb question Jul 11 12:17:13 I have built the angstrom image for x86 Jul 11 12:17:24 I have the image files with the root fs Jul 11 12:17:59 now I want to add a package so that it is included in the image Jul 11 12:18:44 I tried adding grub using local.conf Jul 11 12:19:09 but grub wasn't included on the image until I removed my tmp folder and built everything again Jul 11 12:19:16 which took 6 hours Jul 11 12:19:37 karmakop, you used ANGSTROM_EXTRA_INSTALL ? Jul 11 12:20:25 DISTRO_EXTRA_RDEPENDS += "grub" Jul 11 12:20:49 that's how I added grub Jul 11 12:21:10 hi zecke Jul 11 12:22:06 looking up ANGSTROM_EXTRA_INSTALL Jul 11 12:22:53 karmakop, that's what I use. usually when I rebuild the image then stuff will get built and included without any rm -rf tmp Jul 11 12:23:55 pb_: hi! Jul 11 12:23:55 pb_: what is the temperature like in Cambridge (or close to it)? Jul 11 12:24:35 zecke: mid-high 20s, I think. we had about 30C yesterday but I think today is a bit cooler. Jul 11 12:24:37 let me look Jul 11 12:25:12 http://www.wunderground.com/cgi-bin/findweather/getForecast?query=zmw:00000.1.03571 says 26.4C today Jul 11 12:25:57 yesterday's high was 31.8C apparently Jul 11 12:30:00 dcordes: Thanks, looking for the right .bb to add that line in Jul 11 12:30:53 karmakop, many image recipes already have a line that includes what you put into ANGSTROM_EXTRA_INSTALL so you can just put it in local.conf Jul 11 12:31:28 karmakop, the easiest way is prolly just adding grub and everything else you want directly in the image recipe into IMAGE_INSTALL Jul 11 12:40:56 dcordes: I'm floundering trying to find the right .bb Jul 11 12:41:12 karmakop, what's the bitbake commandline you use for the image ? Jul 11 12:44:33 bitbake x11-image Jul 11 12:45:04 dcordes: sorry, dealing with a 4 year old covered in syrup Jul 11 12:46:46 khem, glib-2.0 also failing with PREFERRED_VERSION_glib-2.0-native = "2.24.0" http://tinderbox.openembedded.net/public/logs/task/6482496.txt Jul 11 12:46:53 karmakop, :) Jul 11 12:47:30 karmakop, recipes/images/x11-image.bb Jul 11 12:48:45 dcordes: ahh, missing the images subfolder, just found it with find Jul 11 12:49:22 going to try lilo, grub crashed with a floating point exception Jul 11 12:51:17 oops Jul 11 12:58:24 dcordes imagine i have my .bb file 2.6.33... what do I have todo for getting bitbake x11-image working with that kernel version? Jul 11 13:05:58 dcordes: thanks, building now Jul 11 13:06:54 aszpain, aszpain, are you uptodate ? 'git pull' aszpain, you really should not have a packages dir Jul 11 13:19:03 mickey|sofa, hey there :) can you take a look at my devel ml problem ? Jul 11 13:20:25 ? Jul 11 13:34:00 mickey|bbl, mickey|bbl, I am subscribed to the devel ml. few hours ago I sent couple of mails which are not visible in the archive yet. can you check ? Jul 11 13:34:32 mickey|sofa, pb_, I sent the first mail 03:21:44 CET Jul 11 13:35:49 let me check logs, from which email address did you send?` Jul 11 13:38:00 lukas at htc-linux.org Jul 11 13:45:13 hmm, i don't see any entries Jul 11 13:46:01 do you have a transcript from your mail server that they have been successfully received by linuxtogo.org` Jul 11 13:46:02 ? Jul 11 13:49:23 mickey|sofa, hm they in sent box Jul 11 13:50:19 mickey|sofa, I found the problem Jul 11 13:51:00 mickey|sofa, my mailserver destroys my address. it appends my second name somehow, although I receive mails to address without it Jul 11 13:51:05 mickey|sofa, very strange Jul 11 13:51:30 that explains why nothing ever arrived on tslib and arm linux lists Jul 11 13:52:41 :) Jul 11 13:53:02 which mailserver are oyu using? Jul 11 13:54:10 mickey|sofa, horde ? Jul 11 13:54:56 horde have own mail server? :) Jul 11 13:55:16 03Michael 'Mickey' Lauer  07org.openembedded.dev * r58abfe62fe 10openembedded.git/ (4 files in 2 dirs): Jul 11 13:55:16 python 2.6: add python-numbers subpackage adding support for numerical APIs: Jul 11 13:55:16 Included for now are the modules numbers and decimal Jul 11 13:55:59 Jay7, I don't know that's just the interface. all I know is what it does is pretty braindead Jul 11 15:48:36 khem: Don't suppose you want to go edit the mipsel site files and see what's mipsel-linux, what's glibc and what's uclibc? :) Jul 11 15:55:39 does the command bitbake -c build recompiles the packets? or it just takes the binaries and pack them? Jul 11 17:34:41 in my x11-image.bb file I have modified the ANGSTROM_EXTRA_INSTALL ?= "" to be ANGSTROM_EXTRA_INSTALL ?= "lilo-sh" Jul 11 17:37:10 but it doesn't like lilo-sh Jul 11 17:37:42 so now I'm back to figuring out the grub problem Jul 11 17:38:14 can I paste like 5 lines in here or is that against the rules? Jul 11 17:40:19 when you rebuild a kernel changing the .config and you choose one driver to be loaded as module .ko for ex: CONFIG_X=m is thar .ko modules automatically be loaded or you have to manual insmod it? Jul 11 17:45:41 aszpain: lsmod after boot and see if it's there? Jul 11 17:47:20 ok Jul 11 17:47:46 I will warn you that I'm a total newb Jul 11 17:48:17 I have yet to get my image to boot :) Jul 11 17:48:19 one question: I did a bitbake x11-image... and I have all the images and so... but now I want to change some kernel property like CONFIG_HID=y how can I easily do it with oe? Jul 11 17:48:31 :) Jul 11 17:49:25 Here is something I found that is very similar to my grub issue if not exactly the same: Jul 11 17:49:30 Using the latest grub version (0.97-15) I'm unable to install grub on a ext2 formatted device using the following command: Jul 11 17:49:31 grub> setup (hd1) Jul 11 17:49:31 results in: Jul 11 17:49:31 grub> setup (hd1)Floating point exception Jul 11 17:49:31 Recompiling grub without the grub-inode-size.patch resolves this issue. Jul 11 17:50:04 so doing a find in the openembedded folder I get: Jul 11 17:51:13 ./recipes/grub Jul 11 17:51:13 ./recipes/grub/grub-0.97 Jul 11 17:51:13 ./recipes/grub/grub-0.97/grub-support-256byte-inode.diff Jul 11 17:51:13 ./recipes/grub/grub-0.97/grub-0.97-disk_geometry-1.patch Jul 11 17:51:13 ./recipes/grub/grub_0.97.bb Jul 11 17:52:03 since there is a patch there I am wondering if it has already been applied and if I can undo it and try without Jul 11 19:59:43 Tartarus: what do you mean Jul 11 19:59:59 * khem is gripped to final atm Jul 11 21:15:38 khem: Frequently, imho, sites/* has generic linux, or generic $arch stuff inside of the arch-linux-libc file Jul 11 21:58:07 Tartarus: ah true. However I did not want to clean the mess this time. Jul 11 21:58:33 Tartarus: the define I added is arch specific infact Jul 11 21:58:47 someday we should redo the sites Jul 11 22:34:43 khem, others hi Jul 11 22:35:19 khem: I started the build again with the proposed glib-2.0-cross downgrade and gcc-cross_4.5 Jul 11 22:35:29 khem: same libiconv error on glib Jul 11 22:36:28 khem: it would be good to finish the build so I can show koen not only 4.4.4 compiled angstrom works well but even 4.5 :) Jul 11 22:42:50 khem: I always urge folks when they touch one to clean 'em up Jul 11 22:42:54 at least the one they touched :) Jul 11 22:43:12 since I bet you'll hit a few more like that, needing to be copy/pasted from mips-linux, and os on Jul 11 22:47:21 Tartarus: its like a covered shit and I try to put mud over it to cover it Jul 11 22:47:50 Tartarus: the problem is if fix something I should morally build that many combo's to verify Jul 11 22:48:10 but yes I think it makes sense to clean them up Jul 11 22:48:23 dcordes: hmmm Jul 11 22:48:37 dcordes: so this is a new machine Jul 11 22:48:39 right Jul 11 22:49:07 dcordes: I would suggest that install libiconv on this box and then it will work Jul 11 22:49:16 thats easiest way around Jul 11 22:50:21 khem: ok but I bet others will run into same problem Jul 11 22:58:32 dcordes: I know someone should fix the recipe in some way Jul 11 22:58:48 dcordes: I dont know why I dont have this problem yet Jul 11 23:00:30 khem: if it is a host dependency and there is no libiconv-native, how would it be fixed in the recipe ? Jul 11 23:00:47 removing the dep Jul 11 23:00:53 in someway Jul 11 23:01:06 it should be doable if someone spends time looking into it Jul 11 23:01:52 khem: ok. it was seen as I put it on the ml.. so I will work around it Jul 11 23:02:23 khem: question is how. is it easy to create native recipe from the oe libiconv ? Jul 11 23:07:40 khem: it seems the host already has a libiconv*.so installed Jul 11 23:07:57 khem: do you know by any chance which package is needed in ubuntu ? Jul 11 23:13:44 khem: ok we installed all packages containing libiconv in name and still get this line during compile Jul 11 23:13:49 gconvert.c:55:2: error: #error GNU libiconv not in use but included iconv.h is from libiconv Jul 11 23:14:09 http://tinderbox.openembedded.net/public/logs/task/6487205.txt Jul 11 23:32:12 dcordes: hmm Jul 11 23:32:26 dcordes: you are using angstrom right Jul 11 23:37:21 03Khem Raj  07org.openembedded.dev * red52cfe058 10openembedded.git/contrib/qemu/run-qemu.sh: Jul 11 23:37:21 run-qemu.sh: Add qemumipsel target Jul 11 23:37:21 Signed-off-by: Khem Raj Jul 11 23:38:25 khem: yea Jul 11 23:39:05 dcordes: can you add OE --with-libiconv=gnu to glibc-2.0_2.24.1.bb Jul 11 23:40:00 khem: hm I just reverted to some commit from july 8th and rebuilding now .. Jul 11 23:40:44 EXTRA_OECONF += " --with-libiconv=gnu " Jul 11 23:40:50 try that too Jul 11 23:40:51 khem: what exactly does that line do ? and why changing in glibc not glib Jul 11 23:41:03 I meant glib Jul 11 23:41:17 glib-2.0_2.24.1.bb Jul 11 23:45:50 dcordes: btw do you have iconv.h on /usr/local/include Jul 11 23:46:38 dcordes: infact you should remove all the stuff you have installed in /usr/local Jul 11 23:53:40 khem: man... it exists. that stupid admin program messes up entire OE Jul 11 23:55:20 khem: I fear we can't just remove it Jul 11 23:55:37 khem: is there a way to tell the operating system to not use these libs or only use them fro the admin prgoram ? Jul 11 23:55:42 program Jul 11 23:59:58 dcordes: EXTRA_OECONF += " --with-libiconv=gnu " should help you Jul 12 00:00:06 dcordes: btw. if it solves the issue Jul 12 00:00:16 please respond to your thread about the real problem Jul 12 00:00:19 and the solution Jul 12 00:00:43 actually glib needs iconv but it should use glibc provided iconv Jul 12 00:00:52 or libiconv provided one Jul 12 00:00:59 ok so I wll add that line in local.conf ? Jul 12 00:01:06 no Jul 12 00:01:16 in glib-2.0_2.24.1.bb Jul 12 00:01:31 I best add it in that plus the native one, right ? Jul 12 00:01:42 no need on native one Jul 12 00:01:45 ok Jul 12 00:01:45 actually you know what Jul 12 00:01:50 you only need it for native Jul 12 00:01:55 just a moment Jul 12 00:02:58 EXTRA_OECONF_append_virtclass-native = " --with-libiconv=gnu " Jul 12 00:03:01 is what you need Jul 12 00:05:36 is there another boot loader besides grub I should be using? I can't get past a floating point error Jul 12 00:06:51 karmakop: explain the problem in more detail dont expect answers to vague questions Jul 12 00:07:25 I was pretty specific above but I can repost, or maybe a forum is a better option Jul 12 00:08:14 http://www.cs.nmsu.edu/~pfeiffer/grub.html Jul 12 00:08:22 I followed the instructions there Jul 12 00:08:46 grub> setup (hd0) Jul 12 00:09:00 on that step I get a Floating Point exception Jul 12 00:09:44 karmakop: yes mailing list is always a better place for longer discussions Jul 12 00:09:47 people come and go Jul 12 00:09:58 and not many know what the last episode was Jul 12 00:10:07 but some dont know Jul 12 00:10:13 like me in this case Jul 12 00:10:29 seems your grub is busted Jul 12 00:10:44 which arch Jul 12 00:11:01 x86 Jul 12 00:11:45 but it has done the same whether I am using my workstation with a chroot or on the actual hardware I plan on using with a livecd Jul 12 00:12:32 ok Did you try to look deeper into the grub execution Jul 12 00:13:24 no Jul 12 00:17:00 I find similar instances of people having the same problem but outside of oe Jul 12 00:17:46 karmakop: I think one solution is to upgrade grub to grub2 Jul 12 00:18:02 karmakop: write a recipe for grub2 and that should help you Jul 12 00:18:17 oh my :) Jul 12 00:18:33 or try to add this patch http://bugs.debian.org/cgi-bin/bugreport.cgi?msg=15;filename=grub-support-256byte-inode.diff;att=1;bug=463123 Jul 12 00:18:41 I dont know if it will work right away Jul 12 00:18:57 grub has to go hand in hand with e2fsprogs Jul 12 00:19:16 and it seems we have newer enough e2fsprogs but not grub Jul 12 00:19:23 there was a patch in grub already Jul 12 00:19:31 ok Jul 12 00:20:27 very similar to that name but I'm going by memory Jul 12 00:21:55 karmakop: are you using .dev ? Jul 12 00:22:02 I see that this patch is there Jul 12 00:22:16 I used the getting start instructions Jul 12 00:22:33 karmakop: which branch of OE are you using ? Jul 12 00:22:39 are you on stable Jul 12 00:23:05 or master aka org.openembedded.dev Jul 12 00:24:04 git clone git://git.openembedded.org/openembedded Jul 12 00:24:22 then did a 'git pull' a couple of days ago Jul 12 00:24:35 I don't know what that puts me at to be honest Jul 12 00:26:15 mrave@fresh:~/stuff/openembedded/recipes/grub/grub-0.97$ ls Jul 12 00:26:16 automake-1.10.patch grub-0.97-disk_geometry-1.patch menu.lst Jul 12 00:26:16 eee701 grub-support-256byte-inode.diff Jul 12 00:27:00 looks like whatever I'm using has the patch Jul 12 00:27:00 karmakop: git branch would tell you Jul 12 00:27:10 karmakop: ok so you seem to be on .dev Jul 12 00:27:36 I was actually wondering if I should try without the patch Jul 12 00:27:48 yeah you could try it Jul 12 00:27:52 and narrow down the issue Jul 12 00:28:06 I find myself having to rebuild everything though Jul 12 00:28:26 can I just delete that .diff and .patch? Jul 12 00:28:44 and then 'bitbake console-image' ? Jul 12 00:31:10 khem: http://privatepaste.com/58e6c199dc Jul 12 00:31:18 khem: can you check if I added the line correctly ? Jul 12 00:33:38 khem: seems I might want to try ext3 Jul 12 01:22:59 khem: I changed the image to ext3 and no Floating Point exception Jul 12 01:23:12 khem: Thanks for your help Jul 12 02:25:29 'bitbake gcc-cross' builds perl too? **** ENDING LOGGING AT Mon Jul 12 02:59:57 2010