**** BEGIN LOGGING AT Fri Nov 26 02:59:58 2010 Nov 26 03:27:27 i have a working libgcc_s.so.1 and a nonworking. Taking an objdump -S and looking at the diff... http://pastebin.com/W9uaJxAH Nov 26 03:27:41 there is a singular change. Puzzling Nov 26 03:30:32 It looks like my bad libgcc_s.so was built with binutils-cross-2.19.51 and the good with binutils-cross-2.20.1 Nov 26 04:36:46 khem: hi, any chance, that you've fix for that in your pocket? it's on release branch http://tinderbox.openembedded.net/public/logs/task/11263961.txt for http://tinderbox.openembedded.net/builds/106442/ Nov 26 04:38:35 khem: I've fix for that failing x264 on uclibc in angstrom-2008.1 and I'll send it to the mailing list after the confirmed rebuild Nov 26 04:39:05 khem: so you can cherry-pick it then to the release branch Nov 26 04:40:35 khem: this is that x264 failure http://tinderbox.openembedded.net/public/logs/task/11290581.txt Nov 26 05:12:25 Hi, if i successfully bitbake console-image for my platform, where is the cross compiler (toolchain) located once compile is complete? Nov 26 05:45:49 guyvdb_, I'm building on x86_64 and my target arch is mipsel. My compiler lives at: tmp/sysroots/x86_64-linux/mipsel/bin/mipsel-oe-linux-uclibc-gcc Nov 26 05:47:02 grg thx Nov 26 05:47:18 guyvdb_, your's might be in sysroots/$HOST/usr/$TARGET/bin instead Nov 26 05:47:25 s/'// Nov 26 05:47:59 ok found it x86_64-linux/usr/arch/bin Nov 26 05:48:21 thx Nov 26 05:51:57 I want to build a base system (console) + qt-embedded. What image should I inherit from? console-image ? Nov 26 05:53:14 i haven't used qt with oe before, but i see a qt4e-base-image.bb which might be good to look at Nov 26 05:54:54 k Nov 26 05:57:07 also i am using DISTRO = "angstrom-2008.1" is there a newer angstrom that i should be using? Nov 26 06:07:08 guyvdb_, there is angstrom-2010.x Nov 26 06:07:56 that will give you a more recent toolchain, i don't use angstrom these days so i can't tell you if its worth using one over the other for any particular scenario Nov 26 06:08:44 what is a best minimal distro? Nov 26 06:09:17 best for what? Nov 26 06:09:28 See http://wiki.openembedded.net/index.php/Testing to see what people are testing Nov 26 06:09:47 minimal gets lots of testing so most things should work well Nov 26 06:10:08 I am looking to build a single application (Qt4) embedded device Nov 26 06:10:12 ok Nov 26 06:10:40 angstrom gets lots of testing from texas instruments developers, so if you have, e.g. an omap device, you'll probably find angstrom to be a good fit Nov 26 06:10:56 but others will work fine too Nov 26 06:11:01 no i am on a mini2440 clone Nov 26 06:11:15 Em2440 Nov 26 06:12:41 that's a samsung arm, right? Nov 26 06:13:00 yup Nov 26 06:13:29 S3C2440 Nov 26 06:13:46 what are you building on? Nov 26 06:14:13 jz4730 and jz4740. mipsel boards. I just use the qemumipsel target Nov 26 06:15:17 from the testing wiki page, i see that qt4e-demo-image has been tested for angstrom-2010.x and minimal. So either of those should be usable for you Nov 26 06:16:07 I will build both to check out... takes forever to build :) Nov 26 06:16:32 you probably won't get many differences between them Nov 26 06:17:00 but I am happy i have been able to build angstrom 2008 and get it running on target. I have only had board for a day. But did a lot of reading before board arrived Nov 26 06:17:49 I have tried both OE and poky... eventually coming back to OE Nov 26 06:22:07 what is the size of -minimal for armel (armv7)? Nov 26 06:26:22 got to go... thx for help grg Nov 26 06:26:31 guyvdb_, no problem Nov 26 06:26:54 ka6sox, sorry i can't help on that... i haven't done an arm build this year Nov 26 06:27:17 seems quiet today... does .eu have a holiday too? Nov 26 06:27:32 thanks grg, np...I need to make a recovery rootfs for the pre. Nov 26 06:27:38 (ramfs that is) Nov 26 06:27:52 i suspect sizes are not dissimilar across platforms Nov 26 06:28:20 arm has various thumb bits, so its size will be smaller than, e.g. mips Nov 26 06:29:18 mipsel/ minimal-image .ext2 is ~7mb. Nov 26 06:29:59 no, wiat. that's crap.... Nov 26 06:30:07 :) Nov 26 06:30:49 i'll extract a few tarballs... Nov 26 06:31:00 okay certainly small enough to fit easily in a ramfs. Nov 26 06:31:23 I'll need things like lvm2 and dd to make it work. Nov 26 06:35:09 ka6sox, Extracted rootfs sizes (du -hs), minimal/minimal-image is 8.5M, minimal-uclibc/minimal-image is 5.8M. micro/micro-image is 4.7M, micro-uclibc/micro-image is 2.4M Nov 26 06:35:29 so arm and x86 will be smaller Nov 26 06:35:43 other RISC, probably similar to mips Nov 26 06:36:14 grg, thanks. Nov 26 06:36:18 thats a good start Nov 26 06:36:44 if you use squashfs, you could easily get micro-image under 1m Nov 26 06:49:56 gm Nov 26 06:53:18 khem, yesterday when you were not able to get into the git server...it was because there were 15 people already @ the trough. Nov 26 06:53:38 sucking 40Mb of traffic Nov 26 07:13:47 ka6sox: hi, are you the person who can check why I can't push to git server? :) Nov 26 07:14:04 ka6sox: tried contact mickey and cbrake, but no luck... Nov 26 07:20:08 ynezz, who did you send your key too? Nov 26 07:20:24 cbrake, mickey Nov 26 07:20:51 okay what is the sig on the key? Nov 26 07:21:38 khem: this is fix for that mesa failure, is it ok? (I'm not sure about that __UCLIBC define) http://ynezz.ibawizard.net/private/0001-mesa-fix-build-error-on-uclibc.patch Nov 26 07:22:39 khem: and this is the fix for the x264 failure http://ynezz.ibawizard.net/private/0004-angstrom-2008.1-bump-uclibc-to-0.9.30.3-which-unbrea.patch Nov 26 07:23:06 khem: if it's ok with you, I'll send them for the review to the mailing list Nov 26 07:23:16 ka6sox: checking, mmnt Nov 26 07:23:20 kk Nov 26 07:24:42 ka6sox: ssh-rsa AAAAB3NzaC1yc2EAAAABIwAAAQEAqyG9sQi8z2eZdGppqcGsm3TZzpixz58gozCAS8wjiYfsB0SOUV0chLSuG9ZgJSxzW594kv3iJafWmU17ylmWO7adm (truncated) Nov 26 07:25:20 I just need the sig ... ie ynezz@ or just ynez? (at the end) Nov 26 07:25:30 ynezz@ntbk Nov 26 07:25:38 kk..let me look Nov 26 07:25:59 I thought, that the signature is the whole thing :) Nov 26 07:26:21 to find it I just need the last little bit first Nov 26 07:26:28 then I'll ask you for the entire key Nov 26 07:29:42 Hi All, To running of opkg command on Readonly file system, getting following error: Nov 26 07:29:42 opkg-cl list-installed Nov 26 07:29:42 Collected errors: Nov 26 07:29:42 * Could not lock /usr/lib/opkg/lock: Read-only file system Nov 26 07:29:42 Please let me know your suggestion Nov 26 07:30:56 remount as rw :) Nov 26 07:32:12 mount / -o remount,rw or so Nov 26 07:32:23 unless ofc the rfs is squashfs or so Nov 26 07:38:50 eFfeM_work: rfs is nfs mount point using as mount -o remount,ro rootfs/ Nov 26 07:39:19 ro means readonly :-) Nov 26 07:39:25 change that to rw Nov 26 07:39:37 ok Nov 26 07:39:43 and make sure your nfs server allows write (/etc/exports or so) Nov 26 07:43:11 03Martin Jansa  07master * rc83376a666 10openembedded.git/recipes/e17/eve_svn.bb: Nov 26 07:43:11 eve: bump PR to rebuild after last webkit-efl upgrade Nov 26 07:43:11 * otherwise old build ends with Illegal instruction on armv4t Nov 26 08:19:22 ynezz, ping? Nov 26 08:28:21 good morning Nov 26 08:28:46 mood groaning Nov 26 08:32:35 ynezz, try it..no guarantees...it was pretty messed up. Nov 26 08:33:07 morning Nov 26 08:33:23 good morning and good night Nov 26 08:41:10 snowing Nov 26 08:48:15 I've started an initramfs fs in which i've included the nand write executable. But i can't use it, it tells me not found. http://pastebin.com/PV4gLL6w Any idea what could be the problem ? Nov 26 08:50:00 melchior: check permissions (ls -al) Nov 26 08:51:18 -rwxr-xr-x 1 root root 81745 Jan 1 1970 nandwrite Nov 26 08:51:57 seems good to me Nov 26 08:54:41 hmmm Nov 26 08:55:00 you could try adding strace to your image and then call strace nandwrite Nov 26 08:56:08 i'll try, is it included in busybox ? Nov 26 08:56:44 no is a separate package Nov 26 09:02:07 i've included strace in my initramfs fs and get /bin/sh: strace: not found Nov 26 09:02:54 I've compiled busybox statically ofr my initramfs fs, could it be a missing lib for nandwrite and strace ? Nov 26 09:07:35 melchior: IMHO it could be a missing lib Nov 26 09:42:37 hello all iam tyring to build openjdk, in the process of building it downloads so many packages , i have some tar balls in my box how can i give it oe so that it doesnot download it again ? Nov 26 09:42:47 any ideas..? Nov 26 09:43:09 vlrk: use a common download directory for your builds Nov 26 09:44:59 you mean to say that the sources directory should be same for my tar balls and the oe ! am i right? Nov 26 09:45:47 i mean DL_DIR Nov 26 09:45:51 vlrk: I think so Nov 26 09:45:54 parameter in local.conf Nov 26 09:46:21 so if i move / copy or give soft link to this directory it will not download again ? Nov 26 09:46:41 vlrk: move or copy, I don't know if a link works Nov 26 09:46:50 ok .. Nov 26 09:46:58 in the mean time .. Nov 26 09:47:14 can we get info related to how oe can be worked with prebuilt chain Nov 26 09:47:43 vlrk: check the wiki and google Nov 26 09:49:53 ynezz: or at least 0.9.30.3 (but my build broke on e2fsprogs) Nov 26 09:54:51 hey anyone on... anyone that is "in charge" Nov 26 09:55:22 * XorA is IN CHARGE Nov 26 09:56:15 I mean as in an OE core member, or someone on the board Nov 26 09:56:24 that would be me as well Nov 26 09:56:41 really, awesome Nov 26 09:56:48 I will privmsg Nov 26 09:56:54 there are no core members anymore Nov 26 09:56:56 here I'm actually in discharge after a short night of charge :-D Nov 26 09:57:04 eV replaced the concept of core Nov 26 09:58:12 ynezz: btw I had too old binutils, probably Nov 26 09:58:32 we have to ask Angstrom guys to bump both uclibc and binutils Nov 26 10:10:39 done Nov 26 10:23:09 ant_work: angstrom-next is using much newer uclibc and binutils I think Nov 26 10:23:18 ant_work: will be a while before angstrom does Nov 26 10:23:58 angstrom-2008 fails to build uclibc as-it-is Nov 26 10:24:14 it used to work, done it many times for avr32 Nov 26 10:24:22 ant_work: I've patches for that, will send them for RFC/ACK Nov 26 10:24:30 now x264 and ffmpeg are pulled in and are failing Nov 26 10:24:44 XorA: console-image fails on ts72xx angstrom 2008/2010 Nov 26 10:24:56 mesa and x264 Nov 26 10:25:37 http://ynezz.ibawizard.net/private/0001-mesa-fix-build-error-on-uclibc.patch Nov 26 10:25:42 XorA: actually only x264 is pulled in my zaurus images (console, x11, x11-gpe) but I know freempg will fail too Nov 26 10:25:43 http://ynezz.ibawizard.net/private/0004-angstrom-2008.1-bump-uclibc-to-0.9.30.3-which-unbrea.patch Nov 26 10:26:13 ynezz: 0.9.30.3 will then fail on e2fsprogs Nov 26 10:26:17 :/ Nov 26 10:26:20 ah, cool :D Nov 26 10:26:26 bump bnutils Nov 26 10:26:46 03Eric BĂ©nard  07master * r26002b31b7 10openembedded.git/classes/qt4x11.bbclass: (log message trimmed) Nov 26 10:26:46 qt4x11: set QT_LIBINFIX and QT_DIR_NAME Nov 26 10:26:46 this will fix the following issue with qwt 5.2.1 where obj dir is: Nov 26 10:26:46 work/armv4t-oe-linux-gnueabi/qwt-5.2.1-r0/qwt-5.2.1/src/obj${QT_LIBINFIX}/ Nov 26 10:26:46 instead of : Nov 26 10:26:47 work/armv4t-oe-linux-gnueabi/qwt-5.2.1-r0/qwt-5.2.1/src/obj/ Nov 26 10:26:47 From: Martin Jansa Nov 26 10:27:03 ant_work: you know about those issues and you don't fix it? :) Nov 26 10:27:09 and use uclibc_git Nov 26 10:27:19 it's release-2010.12 Nov 26 10:27:27 so it depends on khem, not me Nov 26 10:27:48 see http://lists.linuxtogo.org/pipermail/angstrom-distro-devel/2010-November/004255.html thread Nov 26 10:27:56 and since 2 combinations fails for me I would like to fix them Nov 26 10:28:00 we are waiting for 0.9.32 Nov 26 10:28:21 until thenm use uclibc_git (0.9.31) Nov 26 10:29:07 what? Nov 26 10:29:15 and binutils 2.20.1 Nov 26 10:29:21 do you know about the release-2010.12? Nov 26 10:29:36 I simply don't get what are you talking about Nov 26 10:29:37 :) Nov 26 10:29:38 I tested master but surely release is broken too in this regard Nov 26 10:29:51 ANGSTROM_BINUTILS_VERSION ?= "2.20.1" Nov 26 10:29:57 angstrom distro cannot build uclibc images today Nov 26 10:30:01 ok, isn't it easier to fix e2fsprog? Nov 26 10:30:06 in release I mean Nov 26 10:30:31 ynezz: I asked khem and he's not inclined to fix older uclibc Nov 26 10:30:40 ok, nice release then Nov 26 10:30:41 :) Nov 26 10:30:52 perhaps I misunderstood ;) Nov 26 10:31:00 good to know, I can spent my time better Nov 26 10:31:07 ant_work: please submit patches to bump uclibc to angstrom2010 Nov 26 10:31:24 is not big matter, just use the defaults of minmal / angstrom next Nov 26 10:31:39 XorA: wil test another build and do Nov 26 10:32:14 ant_work: angstrom2010 is playground for new toolchains Nov 26 10:32:24 ant_work: 2008 is meant to be pretty stable Nov 26 10:32:39 stable but it doesn't even build here :) Nov 26 10:32:52 at least uclibc, probably it isn't tested that much Nov 26 10:36:25 so khem is going to make a release with 0.9.32, right? Nov 26 10:39:09 it looks like Nov 26 10:39:33 you have to ask blindvt and khem Nov 26 10:39:42 but yes, I understood Nov 26 10:42:28 uclibc breakage is new as I was using it for avr32 development Nov 26 10:46:45 first failure is x264, which I think is relatively new in my Zaurus rootfs's Nov 26 10:48:49 ant_work: there are still zaurus users? Nov 26 10:49:03 not sure Nov 26 10:49:47 is still good to test arm images Nov 26 10:50:14 the poor Z is exposed to any sort of breakage: Xserver, bluez, keymaps,... :P Nov 26 10:50:22 good testbed Nov 26 10:50:29 ;D Nov 26 10:52:29 there is even an h2200 user :-) Nov 26 10:52:54 I think Zaurus has been the first machine in OE booting lzma-compressed kernels Nov 26 10:53:10 ant_work: you are size restricted - thats why Nov 26 10:53:41 ant_work: kexecboot kernels use ext4_as_ext[23] thing probably too Nov 26 10:54:44 hrw: at least lzo should be used for all kernels instead of .gz Nov 26 11:18:00 hrw: u-boot is reaching our size and will surpass us soon ;) Nov 26 11:24:12 ;) Nov 26 11:40:35 03Koen Kooi  07org.openembedded.dev * r8c95a9e182 10openembedded.git/recipes/linux/linux-omap_2.6.37rc.bb: Nov 26 11:40:35 linux-omap 2.6.37rc: move to rc3 Nov 26 11:40:35 Signed-off-by: Koen Kooi Nov 26 12:24:05 hi all Nov 26 12:25:28 Can anyone say, how to build a customized kernel for BeagboardxM using Openembedded .. or can it be done using openembedded ? Nov 26 13:16:42 hi there Nov 26 13:17:01 i git the following message after trying to bitbake nano/vim Nov 26 13:17:07 http://pastebin.com/XUJAMqWU Nov 26 13:23:32 jannau: not sure what happened with mythtv, yesterday evening the problem was reproducable at home, i've kicked off a build on the autobuilder (also svn 27202) and that one builds, guess I'll try a clean build at home overnight Nov 26 13:25:43 any ideas about this: http://pastebin.com/XUJAMqWU ? Nov 26 13:28:54 ant_work: did you ever used elementary_test on you akita? Nov 26 13:31:04 cazze: what do you mean? Basic runtime tests? Nov 26 13:32:22 no, elementary from enlightenment Nov 26 13:32:40 no then Nov 26 13:33:12 ant_work: ok thx Nov 26 13:41:29 hi, how to build a customized kernel for BeagboardxM using Openembedded .. or can it be done using openembedded ? Nov 26 13:49:56 eFfeM_work: maybe a old set of mythtv libraries in the sysroot? Nov 26 14:15:27 03Koen Kooi  07org.openembedded.dev * r41a2fe452d 10openembedded.git/recipes/dropbear/ (dropbear.inc dropbear/init): Nov 26 14:15:27 dropbear: remove 'set -e' from initscript Nov 26 14:15:27 This fixes dropbear startup issues reported in recent builds Nov 26 14:15:27 Signed-off-by: Koen Kooi Nov 26 15:31:03 hello all , if i want to exclude any package from build , where can i do that ..? Nov 26 15:38:52 whi is that in oesf.org forum I don't have permissions to do anything? Nov 26 15:39:17 I registered and can login but cannot for instance post a reply or even contact the admins... Nov 26 15:41:29 alkopop79: the correct website is http://wiki.openembedded.org Nov 26 15:41:43 thank you Nov 26 15:41:56 but still cannot do anything in the forum Nov 26 15:42:01 this is silly Nov 26 15:42:46 can't find any forums in the wiki page Nov 26 15:46:25 there is no forum in the wiki Nov 26 15:46:34 what are you looking for ? Nov 26 15:49:44 I think he wants to use the oesf forums. Nov 26 15:50:23 ah what is oesf vs the ev ? Nov 26 15:50:31 unrelated, afaik Nov 26 15:50:44 I think oesf.org is just some bunch of guys. Nov 26 15:53:28 the problem is Nov 26 15:53:42 i don't get any help with my bitbake trouble Nov 26 15:53:45 so Nov 26 15:53:49 registered to the forum Nov 26 15:53:52 can login Nov 26 15:54:00 but cannot post or start a topic Nov 26 15:54:04 or contact the admins Nov 26 15:54:22 it says i don't have permission Nov 26 15:54:45 anyway Nov 26 15:54:47 alkopop79: oesf forums are (largely) unrelated to openembedded Nov 26 15:54:55 ok Nov 26 15:54:57 yea oesf has nothing to do with OE Nov 26 15:54:57 cool Nov 26 15:54:57 ] Nov 26 15:55:06 great Nov 26 15:55:07 thx Nov 26 15:55:19 alkopop1979: your best bet is to try the oe mailing list Nov 26 15:55:21 alkopop79: if you want to discuss something, either here or the mailing list are the ideal places Nov 26 15:55:38 that's what i do but don't get any answers Nov 26 15:55:56 which reminds me Nov 26 15:56:16 can u guys please have a look at this: http://pastebin.com/Kg6Ma35M ? Nov 26 15:56:24 i wonder what's this about Nov 26 15:56:29 it drives me insane Nov 26 15:56:36 I have everything installed Nov 26 15:56:49 yet bitbake moans about the dependencies Nov 26 15:57:10 do i need to change something in the config file? Nov 26 15:59:25 alkopop79: tell us more about your try, which machine, distro, image ? Nov 26 15:59:34 sorry Nov 26 15:59:40 alkopop79: host machine distro, bitbake version? Nov 26 15:59:46 ubuntu to angstrom Nov 26 16:00:14 alkopop79: which angstrom version? Nov 26 16:00:23 alkopop79: which OE branch? Nov 26 16:00:31 just a sec... Nov 26 16:00:31 alkopop79: there are endless possibilities Nov 26 16:00:57 and only A FEW are tested Nov 26 16:00:59 just figuring out how to check the software version Nov 26 16:01:23 angstrom tends to have better test coverage Nov 26 16:01:37 alkopop79: collect your details and come back once you have them, I suggest to post to ML too Nov 26 16:02:04 cool Nov 26 16:02:05 thx Nov 26 16:03:22 that sounds like BBCLASSEXTEND lossage. Nov 26 16:03:33 "perl-native-runtime-native", probably not a good thing Nov 26 16:04:11 I suspect that appending perl-native-runtime-native to ASSUME_PROVIDED would probably be enough to work around the issue, but clearly that isn't a fix. Nov 26 16:06:05 alternatively, patching native.bbclass to recognise that perl-native-runtime is already native would probably solve it too, but that also doesn't sound very nice. Nov 26 16:06:28 probably the "right" answer is to fix the naming of the package in question, make it perl-runtime-native and then I think everything would work out right. Nov 26 16:06:34 i changed recipe , how can i make oe to take the new recipe while building? Nov 26 16:06:47 kergoth: morning Nov 26 16:19:29 guys, apologies Nov 26 16:20:30 but cannot find a way to check the angstrom distro's version Nov 26 16:20:38 i tried terminal commands Nov 26 16:20:42 but nothing Nov 26 16:20:53 is there a way to check angstrom distro version? Nov 26 16:20:58 look in /etc Nov 26 16:21:10 cat /etc/issue Nov 26 16:21:44 thx! Nov 26 16:37:01 alkopop79: in your local.conf DISTRO= Nov 26 16:37:21 ok Nov 26 16:37:23 check it Nov 26 16:40:39 according to the local.conf the DISTRO is angstrom 2008.1 Nov 26 16:41:13 on the other hand the actual version on beagleboard is 2009.X-test-20100104 Nov 26 16:41:27 bitbake is 1.8.18 Nov 26 16:41:43 on ubuntu 10.04 Nov 26 16:42:23 did you try what I suggested earlier? Nov 26 16:43:04 this one: ix the naming of the package in question, make it perl-runtime-native and then I think everything would work out right ? Nov 26 16:43:19 well, try the ASSUME_PROVIDED one first Nov 26 16:43:24 ok Nov 26 16:44:14 this is the actual assume provided: qemu-native Nov 26 16:44:25 should i change it to perl-native...? Nov 26 16:44:39 can any body help me what i need to do if i want to exclude one task during building in oe? Nov 26 16:44:48 alkopop1979: just append to it Nov 26 16:44:56 ASSUME_PROVIDED += "perl-native-runtime-native" Nov 26 16:50:49 thx Nov 26 17:10:16 pb_, just trying what u recommended Nov 26 17:10:38 anyone know if there's a way I can build OE so that it's not dependent on host utilities? I frequently find my builds broken after I do a host upgrade Nov 26 17:12:41 use a chroot or something. Nov 26 17:12:59 we build things, to build things you have to run things like make. you're not going to magically make it be able to build without those Nov 26 17:13:37 hi, is SITE_ENDIANESS available in all recipes? or should I inherit something for getting it? Nov 26 17:13:44 siteinfo Nov 26 17:14:11 ah ok Nov 26 17:14:14 thanks a lot Nov 26 17:15:16 np Nov 26 17:16:26 03Chris Larson  07master * r97e92abe49 10bitbake.git/bin/ (bitbake bitdoc): Nov 26 17:16:26 Use __file__, not sys.argv[0] Nov 26 17:16:26 Signed-off-by: Chris Larson Nov 26 17:24:09 pb_ bitbake did not quit building in 2 mins, no error messages and still running Nov 26 17:24:11 thx! Nov 26 17:35:32 ouch xmame dates from monotone era Nov 26 17:35:39 I wonder why it uses sdl and not X? Nov 26 17:35:45 s/?// Nov 26 17:36:33 there's an sdl backend for X Nov 26 17:36:51 presumably the mame people felt that sdl was more portable, and that's not an entirely unreasonable point of view. Nov 26 17:37:11 yes but X seem to have xv Nov 26 17:37:22 maybe X is faster on device with omapfb which have xv Nov 26 17:37:25 I should try Nov 26 17:39:26 doesn't sdl do xv as well? Nov 26 17:39:36 that said, I don't think the majority of mame targets would be able to make a great deal of use of xv. Nov 26 17:39:59 was there a particular thing you were hoping it would help with? Nov 26 17:41:35 speed Nov 26 17:41:48 basically I tried a game it worked but it was not fullscreen Nov 26 17:41:55 then I put it fullscreen with scalling Nov 26 17:42:08 and then it uses 100% CPU and sound is problematic Nov 26 17:44:23 oh right, scaling. yeah, I expect xv would help with that. Nov 26 17:44:37 but, as I say, I would have thought that it would work fine already with the sdl backend. Nov 26 18:16:59 what is the best efl engine to use on kdrive? Nov 26 18:17:11 x11? x11-16? Nov 26 18:17:17 xrender? Nov 26 18:17:35 i'm working on a very old ipaq h2200 Nov 26 18:18:04 and elementary test is always hitting 87% Nov 26 18:23:41 khem put quiite some jobs on our autobuilder, will try to update the wiki asap (i've been quite busy last week or so). need tor run now, later Nov 26 19:14:37 x11 backend is a lot faster Nov 26 19:24:47 Crofton|road, Crofton|work ping? Nov 26 19:25:24 pong Nov 26 19:26:36 hiya...I think we now have the mystery of the git not working right understood... Nov 26 19:26:45 cool Nov 26 19:27:11 any patchwork info? Nov 26 19:27:26 Maybe we should have a wiki page on these sorts of things Nov 26 19:28:15 patchwork also needs help... Nov 26 19:55:30 Crofton|road, you mean a status page? Nov 26 19:56:07 more a history page Nov 26 19:56:24 a log of problems and solutions Nov 26 19:57:42 hi valhalla Nov 26 19:58:04 did you get your openpandora? Nov 26 19:59:16 pb_, seem a lot faster with x11 backend Nov 26 20:14:24 GNUtoo|laptop: is your sdl backend xv-enabled? Nov 26 20:14:33 there's no particular reason that x11 should be inherently faster Nov 26 20:14:54 I'm not sure SDL supports xv and MIT-SHM Nov 26 20:15:33 but indeed XV+SHM were enabled with X11 Nov 26 20:16:23 btw who knows what games are present in oe? Nov 26 20:16:41 maybe I should grep for the section Nov 26 20:16:55 s/for/for games in/ Nov 26 20:35:58 GNUtoo|laptop: #jlime people have some ports Nov 26 20:37:48 ok thanks Nov 26 20:39:32 GNUtoo|laptop: imho, that would be great to have special category Nov 26 20:39:48 games are fun :) Nov 26 20:40:06 Jay7, maybe it'll come with the new cathegories thing Nov 26 20:40:12 I don't remmber how it's called Nov 26 20:41:59 hm.. Nov 26 20:42:13 I've thought we have CATEGORIES field in recipes.. Nov 26 20:52:42 Jay7, I don't remember how it's called Nov 26 20:52:45 maybe bblayers Nov 26 20:52:48 or something like that Nov 26 20:52:53 the idea is to have layers Nov 26 20:53:10 organized by categories Nov 26 20:53:13 like the main layer Nov 26 20:53:16 the game layer Nov 26 20:53:25 the omap3 layer Nov 26 20:53:27 etc... Nov 26 20:57:28 btw, now meta-toolchains are built ok Nov 26 20:57:34 only -arago one is failing Nov 26 20:58:00 so it would be great to have patches on meta-toolchains in release Nov 26 21:12:34 https://gist.github.com/717211 - no idea how the performance is, but was fun to play with Nov 26 21:19:24 03Chris Larson  07master * r5a56b4120b 10bitbake.git/lib/bb/utils.py: Nov 26 21:19:24 vercmp: don't choke on empty first version components Nov 26 21:19:24 Signed-off-by: Chris Larson Nov 26 21:19:45 03Chris Larson  07master * re75aa94e94 10bitbake.git/lib/bb/utils.py: Nov 26 21:19:45 vercmp: don't choke on empty first version components Nov 26 21:19:45 Signed-off-by: Chris Larson Nov 26 21:25:21 kergoth: what is right way to run multiple processes from python with waiting for completion? Nov 26 21:25:38 i.e. os.system()-like but with fork before Nov 26 21:26:49 subprocess or multiprocessing, depending on what you need to do. subprocess if you need to fork+exec, multiprocessing if you're running python code Nov 26 21:27:02 I need exec() Nov 26 21:27:15 ok, I'll read about subprocess Nov 26 21:28:14 if you're operating within the metadata, see also lib/oe/process.py for convenience wrappers around subprocess Nov 26 21:28:37 I'm still about locale generation :) Nov 26 21:28:47 I wish to make this parallel Nov 26 21:41:14 hmmm, how to benchmark this.. think i need to alter lib/bb/parse/parse_py to be able to *only* parse, and return the ast nodes, not doing any finalizing operations Nov 26 21:41:22 then compare that Nov 26 21:45:14 kergoth, what is source/m4 repo? Nov 26 21:52:16 Hi all. Shouldn't I be able to disable udev by adding ANGSTROM_BLACKLIST_pn_udev = "something" to my image recipe? Nov 26 21:54:19 ka6sox libsamplerate on sources.openembedded.org seems to have an md5sum that does not match the recipe Nov 26 21:55:05 eFfeM got a source with the right md5sum? Nov 26 21:55:22 which version of libsamplerate? Nov 26 21:56:18 oops. maybe ANGSTROM_BLACKLIST_pn-udev would be better. Nov 26 21:59:41 ka6sox '/home/hudson/jobs/FM_TEST/workspace/downloads/libsamplerate-0.1.7.tar.gz' Nov 26 21:59:48 ka6sox: I'd like to use it to experiment with a different way of handling our sources for recipes, if the repo is functional that is, i forgot about it Nov 26 21:59:50 heh Nov 26 22:00:29 ka6sox the tarball changed upstream and the checksums in the recipe have been modified Nov 26 22:01:22 ka6sox http://cgit.openembedded.org/cgit.cgi/openembedded/commit/?id=40e5976b29d5ae95a5c483e87c53ad23e518c7c5 Nov 26 22:01:36 that is the patch that changed the checksums Nov 26 22:05:51 kergoth the repo isn't present but the config says its there...let me create the repo to match Nov 26 22:05:58 that way we can fix gitosis. Nov 26 22:06:13 ah, right, i forgot that gitosis can't handle creations :) Nov 26 22:06:21 yep yep Nov 26 22:07:09 eFfeM no difference from 0.1.7 on that lib with the "fixed" one Nov 26 22:07:38 other than the mismatched md5sum Nov 26 22:12:27 eFfeM I could use a little help solving this mystery... Nov 26 22:17:15 ka6sox was afk the patck seems to indicate there is a diff Nov 26 22:17:49 probably only in the changelog Nov 26 22:18:18 okay. so do we just change the bb to reflect the correct md5sum or ? Nov 26 22:18:44 probably should test it first. Nov 26 22:19:17 ka6sox my understanding is that the recipe is changed for the upstream tarball Nov 26 22:19:38 i got a checksum mismatch because I added sources.openembedded.org as a premirror Nov 26 22:19:54 wanted to do some release testing and figured I needed it for the hh.org files Nov 26 22:20:24 okay where is the correct file and what is the checksum you are expecting? Nov 26 22:22:00 ka6sox source is at http://www.mega-nerd.com/SRC/ sums are: Nov 26 22:22:09 SRC_URI[md5sum] = "6731a81cb0c622c483b28c0d7f90867d" Nov 26 22:22:09 SRC_URI[sha256sum] = "78ed5d9ff1bf162c4a078f6a3e7432a537dd2f22dc58872b081fb01156027fcc" Nov 26 22:22:13 that is what the recipe says Nov 26 22:22:24 0.1.7 version Nov 26 22:24:00 okay so http://www.mega-nerd.com/SRC/libsamplerate-0.1.7.tar.gz Nov 26 22:24:26 yes Nov 26 22:24:52 okay let me get that in the sourcemirror and verify md5sums Nov 26 22:25:09 great Nov 26 22:25:12 is there any other bb that uses the OLD one? Nov 26 22:25:18 or just that one? Nov 26 22:25:23 just that one Nov 26 22:25:47 03Frans Meulenbroeks  07org.openembedded.dev * rf7403d0b48 10openembedded.git/recipes/mythtv/ (3 files): Nov 26 22:25:47 mythtv: moved to rev 27202 Nov 26 22:25:47 this fixes the compilation issue with qt 4.7 Nov 26 22:25:47 Signed-off-by: Frans Meulenbroeks Nov 26 22:25:48 jannau: rebuild mythtv with 27202 from scratch, now it builds fine Nov 26 22:25:50 there it is Nov 26 22:26:21 khem, can you get the above patch for mythtv in release, the previous one did not build after moving to qt 4.6 Nov 26 22:26:46 eFfeM: great, I still can't build angstrom Nov 26 22:28:25 eFfeM try now to build it Nov 26 22:29:05 6731a81cb0c622c483b28c0d7f90867d Nov 26 22:29:17 matches Nov 26 22:30:13 jannau, I'm trying to get james-image for minimal to build with release Nov 26 22:30:21 that mostly works now Nov 26 22:42:16 eFfeM: ok Nov 26 22:52:19 too much food mmm Nov 26 22:57:03 khem: you have been assimilated Nov 26 22:57:13 heh yes Nov 26 22:57:37 Crofton|road: have dinner at a friends so more is coming Nov 26 22:57:49 cool Nov 26 22:57:53 I am flying home Nov 26 22:57:59 msut mkae a presentatoin for next week Nov 26 22:58:10 spend Thanksgiving with my wife's family Nov 26 22:59:00 hmm Nov 26 22:59:03 need to find earphones Nov 26 22:59:05 bye all Nov 26 22:59:23 happy journey Nov 26 23:13:51 khem: bon appetit :-) Nov 26 23:37:55 calling it a day, cya tomorrow ! Nov 26 23:38:00 khem: http://tinderbox.openembedded.net/packages/1096135/ Nov 26 23:38:02 eFfeM: nite Nov 26 23:38:19 nite eFfeM did that work? Nov 26 23:38:20 khem: efikamx_kernel_do_unpack failed again Nov 26 23:39:38 khem: but only for meta-toolchain-gpe Nov 26 23:40:01 ka6sox haven't tried actually restarted the autobuilder without sources.openembedded.org as premirror Nov 26 23:40:15 butif the md5sum is ok i guess it is fine Nov 26 23:40:23 cya.... Nov 27 00:24:26 I've started meta-toolchains build against release branch Nov 27 00:24:31 * Jay7 -> sleep Nov 27 01:41:35 03Khem Raj  07master * ra00e4c89ba 10openembedded.git/recipes/gcc/gcc-configure-sdk.inc: Nov 27 01:41:35 gcc-configure-sdk.inc: Remove special do_configure Nov 27 01:41:35 Signed-off-by: Khem Raj Nov 27 01:41:36 03Khem Raj  07master * rf592eeb55d 10openembedded.git/recipes/uclibc/ (uclibc-git/mips64/uClibc.machine uclibc.inc): Nov 27 01:41:36 uclibc.inc: Dont use gnuhash on mips64 Nov 27 01:41:36 * Add a machine config for mips64 based machines Nov 27 01:41:36 Signed-off-by: Khem Raj Nov 27 01:41:38 03Khem Raj  07master * ref5ba269a5 10openembedded.git/classes/sdk.bbclass: Nov 27 01:41:38 sdk.bbclass: Define target_includedir Nov 27 01:41:38 This var is used in configuring gcc recipes so we define it here for sdk Nov 27 01:41:38 as well best is to have this defined globally so we can configure target Nov 27 01:41:39 cross and sdk recipes properly Nov 27 01:41:39 Signed-off-by: Khem Raj **** ENDING LOGGING AT Sat Nov 27 02:59:58 2010