**** BEGIN LOGGING AT Mon Sep 19 02:59:57 2011 Sep 19 04:57:21 03Richard Purdie  07master * r38a598731b 10bitbake.git/lib/bb/fetch2/git.py: Sep 19 04:57:21 git.py: Fix logging vs logger typo Sep 19 04:57:21 Signed-off-by: Richard Purdie Sep 19 04:57:22 03Richard Purdie  07master * r6ee69ab425 10bitbake.git/lib/bb/cooker.py: (log message trimmed) Sep 19 04:57:22 cooker.py: Fix key expansion issues in showVersions Sep 19 04:57:22 bitbake -s was not displaying correct version information when the Sep 19 04:57:23 PREFERRED_VERSION string contains other variables. The actual built Sep 19 04:57:23 versions would differ since the providers.py functions were called Sep 19 04:57:24 with expanded keys at this point. Sep 19 04:57:24 This patch expands keys for showVersions bringing everything into Sep 19 06:47:39 good morning Sep 19 06:48:34 Is there a "quick" fix to avoid having python-pycairo_1.8.0 link against the host libraries? - according to the log it uses -L/usr/lib and get "/usr/lib/libgcc_s.so.1: file not recognized: File format not recognized" Sep 19 07:34:41 03Steffen Sledz  07org.openembedded.dev * r508b49eedc 10openembedded.git/recipes/linux/ (linux-2.6.28/make-3.82.patch linux_2.6.28.bb): Sep 19 07:34:41 linux-2.6.28: allow building with GNU make >= 3.82 Sep 19 07:34:41 GNU make >= 3.82 does not allow mixed implicit and normal rules Sep 19 07:34:41 Signed-off-by: Steffen Sledz Sep 19 08:04:15 hi all Sep 19 08:05:43 i can only see cache files generated for populate_sysroot, package and populate_lic Sep 19 08:06:11 does bitbake also see checksum of compiler and configure tasks? Sep 19 08:06:23 if yes then how does it compare the checksum Sep 19 08:09:26 morning bluelightning Sep 19 08:09:34 hi Noor, all Sep 19 08:11:22 I cant see the cache file for configure/compile ans install task .... so I am assuming that the data of these tasks is not cached .... is it correct? Sep 19 08:24:16 How do I fix python-pycairo_1.8.0 from linking against the host libraries? - according to the log it uses -L/usr/lib and get "/usr/lib/libgcc_s.so.1: file not recognized: File format not recognized" ? Sep 19 08:29:00 Noor: if you mean sstate cache, no, it only needs to cache the result of populating the sysroot, deploydir, licensedir & packaging Sep 19 08:32:08 bluelightning: yeah talking about sstate cache ------ can you point me the logic where it handles that instead of calling do_configure and related task it will pick stuff from cache Sep 19 08:32:48 bluelightning: have looked runqueu.py in bitbake and sstate.bbclass .... if you can point be on exact code that will be a great help Sep 19 08:33:41 I am really curious that how bitbake decides that he has got the cache now he dont need to run configure/compile/install tasks Sep 19 08:34:30 Noor: it begins with the "setscene" tasks for each task that is sstate-cached Sep 19 08:34:45 there it determines whether the sstate cache file exists I think Sep 19 08:35:01 if it does, it extracts it and creates a stamp file for the real task Sep 19 10:26:06 someone can help me pls? i get similar errors while bitbaking smt. as can be seen here http://pastebin.com/ugF4JXTY Sep 19 10:27:50 or like this: http://pastebin.com/XgYH73Ty while running " $ bitbake console-image" Sep 19 10:38:28 juhsis: http://www.openembedded.org/index.php/OEandYourDistro Sep 19 10:41:21 this two lines of code worked: #sudo rm /bin/sh #sudo ln -s /bin/bash /bin/sh Sep 19 10:43:20 juhsis: please try this : env | grep SHELL Sep 19 10:44:49 for what? Sep 19 10:44:58 juhsis: check Sep 19 10:45:03 ok Sep 19 10:45:24 SHELL=/bin/bash Sep 19 10:45:41 i think it was /bin/dash before Sep 19 10:46:05 juhsis: ok, I suggest you to delete your OE tmp and retry a bitbake Sep 19 10:46:44 i m retryng now Sep 19 10:46:55 juhsis: in case post more details: distro, machine, ... Sep 19 10:47:15 http://code.google.com/p/opencv-dsp-acceleration/wiki/Instruction_For_Building_Examples Sep 19 10:47:15 * mckoan -> bbl Sep 19 10:47:25 i did this commands line by line Sep 19 10:48:00 i have completed 1st dtep Sep 19 10:48:04 step** Sep 19 10:48:13 tnks Sep 19 10:48:45 ups i have no space in machine :) Sep 19 10:48:58 how to delete tmps as you said? Sep 19 10:49:08 juhsis: rm tmp Sep 19 10:49:20 rm -fR tmp (be careful!!!) Sep 19 10:49:28 * mckoan -> lunch Sep 19 10:51:35 ok i found tmp files under build Sep 19 10:51:38 and erase them Sep 19 11:02:06 hi all Sep 19 11:25:55 03Martin Jansa  07master * r9de432fe23 10bitbake.git/lib/bb/fetch2/git.py: Sep 19 11:25:55 fetch2/git: fix logger.debug Sep 19 11:25:55 * logging/logger typo was fixed in 38a598731b49c8a0ba0ede570adc33eb1e848235 Sep 19 11:25:55 but debug level is still missing Sep 19 11:25:55 Signed-off-by: Martin Jansa Sep 19 11:25:56 Signed-off-by: Richard Purdie Sep 19 11:26:00 03Joshua Lock  07master * r5896d8ff7b 10bitbake.git/lib/bb/fetch2/wget.py: Sep 19 11:26:00 fetch2/wget: make checkstatus() quieter Sep 19 11:26:00 Change the wget fetcher to not emit the commands output when calling checkstatus, this Sep 19 11:26:00 matches the behaviour of the git fetchers checkstatus() method. Sep 19 11:26:01 Signed-off-by: Joshua Lock Sep 19 11:26:01 Signed-off-by: Richard Purdie Sep 19 11:32:03 03Martin Jansa  07master * r3a54dcc09a 10bitbake.git/lib/bb/fetch2/git.py: Sep 19 11:32:03 fetch2/git: be more carefull in _contains_ref when checking git log output Sep 19 11:32:03 * in some cases there could be output like this Sep 19 11:32:03 ERROR: ld.so: object 'libpseudo.so' from LD_PRELOAD cannot be preloaded: ignored. Sep 19 11:32:04 before wc -l output and returned 'output.split()[0] != 0' is always True Sep 19 11:32:04 Signed-off-by: Martin Jansa Sep 19 11:32:05 Signed-off-by: Richard Purdie Sep 19 11:34:38 03Paul Eggleton  07master * r3e7f8afeac 10bitbake.git/lib/bb/fetch2/git.py: (log message trimmed) Sep 19 11:34:38 fetch2/git: fix subpath destination directory Sep 19 11:34:38 Make the git fetcher's subpath (path within the git repo to fetch) Sep 19 11:34:38 option set the destsuffix (destination directory) option by default. Sep 19 11:34:38 This reverts the behaviour of subpath to the same as when it was Sep 19 11:34:39 introduced. Sep 19 11:34:39 Based on a patch by Dmitry Eremin-Solenikov Sep 19 11:36:42 Stumbled upon a problem with eglibc-nativesdk 2.12. do_install fails with _linking_ errors on nscd Sep 19 11:47:44 hi if I was to invalidate sstate-cache when my own global varaible's value is change how can I dot that Sep 19 11:48:25 I dont see a variable like whitelist which invalidates the sstate-cache if global varaible's value gets changed Sep 19 11:48:52 like if want to invalidate cache when compiler version is changed ... how it is possible Sep 19 12:35:33 Noor: Just do something like export GLOBAL_VERSION = "X" and then change X to invalidate all sstate files Sep 19 12:36:15 Noor: For the compiler version, just remove cross packages from the default dependency whitelist Sep 19 12:37:24 RP__: that mean when values are changes in exported global variables caches are automatiocally invalidated? Sep 19 12:37:42 Noor: yes Sep 19 12:37:52 OK ..... thanks Sep 19 12:38:01 thats what I was looking for Sep 19 12:39:20 hello Sep 19 12:40:03 i have an error which i cant get rid of Sep 19 12:40:21 NOTE: package ksymoops-2.4.11-r0: task Fetch failed: Unable to fetch URL http://kernel.org/pub/linux/utils/kernel/ksymoops/v2.4/ksymoops-2.4.11.tar.bz2 from any source. Sep 19 12:43:13 masterdenz91: I think kernel.org is down these days that is why u r hitting this issue Sep 19 12:46:48 ohh yes i think youre right im now reading that it has been hacked... :S Sep 19 12:52:44 bluelightning ping Sep 19 12:53:22 masterdenz91: if you are using oe-core you can add INHERIT += "own-mirrors" and SOURCE_MIRROR_URL = "http://autobuilder.yoctoproject.org/sources/" to your local.conf Sep 19 12:54:04 hello woglinde Sep 19 12:54:49 hi ant Sep 19 12:56:42 @ mr_nice i dont really understand what your saying since im not that experienced in all of this but when i click the link it says: object not found... Sep 19 12:59:08 masterdenz91: if you use the new oe-core you have a configuration file called local.conf Sep 19 12:59:26 masterdenz91: if you add the lines it will try to fetch the sources from the yocto mirror first Sep 19 13:08:56 ericben: hi; did you prepare the pull request with the qt4 changes? Sep 19 13:19:01 hi all Sep 19 13:22:37 hi Sep 19 13:23:09 hi pb_ Sep 19 13:23:30 I want to cross compile gdbserver to use on my target device Sep 19 13:24:01 which version of gdbserver should I use? any idea? Sep 19 13:24:26 I don't think gdbserver has changed for ages. Probably doesn't matter. Sep 19 13:26:02 mr_nice: apperently there are more then 1 local.conf files in my overo-eo subdirectories, I edited the /overo-oe/org.openembedded.dev/contrib/angstrom/local.conf file like you said but when i do bitbake omap3-palmtop-image it still tries to fetch from kernel.org Sep 19 13:26:53 should I call gdb-cross for cross compiling? Sep 19 13:28:06 for cross debugging, you mean? Sep 19 13:28:53 yes Sep 19 13:29:40 hi pb Sep 19 13:29:42 I need to debug app on target with gdbserver Sep 19 13:29:50 ? Sep 19 13:29:52 need Sep 19 13:30:49 yes Sep 19 13:31:05 so yes, gdb-cross Sep 19 13:31:15 mgundes: seems to b e the easiest way since you can use your rootfs with debug information from your NFS root Sep 19 13:32:02 thanks pb_ I will try it Sep 19 13:33:11 yes otavio, once I success to get cross gdbserver I will use NFS to use rootfs Sep 19 13:40:25 bluelightning: seems I got another failure at combo-layer Sep 19 13:40:32 [13:39:07] Applying: meta-openembedded: rfkill 0.4: import from OE .dev Sep 19 13:40:33 fatal: corrupt patch at line 21 Sep 19 13:42:26 otavio: hmm... what does line 21 contain? Sep 19 13:43:07 bluelightning: +Date: Fri, 11 Feb 2011 10:35:40 +0100 Sep 19 13:44:34 bluelightning: running it byhand also fails Sep 19 13:44:36 look: Sep 19 13:44:37 (devel)~/hacking/ossystems/embedded-linux% git am --dir meta-openembedded /home/otavio/hacking/ossystems/embedded-linux/patch-00bf84ab-5457-4ee1-9819-5dcfd7a0ee6a/meta-openembedded/0011-rfkill-0.4-import-from-OE-.dev.patch Sep 19 13:44:55 Applying: meta-openembedded: rfkill 0.4: import from OE .dev Sep 19 13:44:56 fatal: corrupt patch at line 21 Sep 19 13:44:56 Patch failed at 0001 meta-openembedded: rfkill 0.4: import from OE .dev Sep 19 13:45:55 otavio: hmm I suspect I know what could be the problem... could you pastebin that file? Sep 19 13:46:27 bluelightning: http://paste.debian.net/131003/ Sep 19 13:49:06 bluelightning: I'll be out for next 2hs ... please leave a message if you want me to try something Sep 19 13:49:38 otavio: I thought so... it's the hook script mangling the patch again Sep 19 13:49:51 otavio: this time triggered by Koen forgetting to add signed-off-by Sep 19 13:50:34 bluelightning: OK; if you want me to test a fix, ping me :) Sep 19 14:21:11 otavio: fixed, have just sent a patch to the oe-core list Sep 19 14:35:57 * ant_work is further discovering how meta-oe and oe-core are forming an ill-matched couple :/ Sep 19 14:57:09 ant hm? Sep 19 14:57:14 ant its in the flow Sep 19 14:59:03 ant_work: what's going wrong? Sep 19 14:59:37 oh, just too many duplicates and the delta is growing Sep 19 15:00:10 ah, and little or no-review in general Sep 19 15:00:34 so we have 2 tzdata/tzcode, 2 udev, ... Sep 19 15:01:12 now, it looks to me we should merge all the possible recipes in oe-core asap Sep 19 15:01:33 ant yes only one should be there Sep 19 15:06:54 one problem is almost nobody in the oe-core/yocto crowd remeber to look at the "newer" recipes in meta-oe or even in oe-classic before updating the corresponding recipe in oe-core Sep 19 15:07:00 a few do ;) Sep 19 15:07:26 s/remeber/remember/ Sep 19 15:07:33 s/remember/bother/ Sep 19 15:08:13 s/a few/one or two/ Sep 19 15:20:29 hmmm Sep 19 15:20:48 not to count duplications in other layers Sep 19 15:25:56 GNUtoo|work: ant_work if the recipe has reached good quality and does not regress oe-core it could be merged Sep 19 15:26:30 generally layers host different versions since they need it for one reason or other it may not be super necessary for oe-core Sep 19 15:27:36 generally meta-oe gets newer versions and they get hardened and eventually moved into oe-core Sep 19 15:27:55 example: tslib_git Sep 19 15:27:59 but that has been merged Sep 19 15:28:20 (in meta-oe) Sep 19 15:30:18 ok Sep 19 15:33:46 khem: I'm not convinced that's the best approach... if there's a valuable fix for a recipe in oe-core it ought to go straight to OE-core Sep 19 15:34:07 tslib_git fixes an issue Sep 19 15:34:21 else you have some error messages and it fails to work Sep 19 15:34:40 I don't remember well but maybe it's a version mismatch Sep 19 15:49:08 bluelightning: Its not a process but an observation Sep 19 15:49:26 khem: the original sin is that oe-core started with obsolete versions from poky Sep 19 15:49:32 since it gets tested in some distros which use meta-oe and can generate better confidence Sep 19 15:49:37 as I said hardening Sep 19 15:50:29 khem: hardened = last year's versions Sep 19 15:50:41 ant_work: obsolete versions are better than non working ones Sep 19 15:51:29 khem: the hardened branch should be kept separated, I understand Sep 19 15:51:42 and needs maintainers Sep 19 15:52:34 but we still need an official dev branch Sep 19 15:53:26 ant_work: dev branch for what ? Sep 19 15:53:37 isnt oe-core/master enough for dev Sep 19 15:53:44 khem: oe-core-next Sep 19 15:54:00 ant_work: why do u feel need for that Sep 19 15:54:10 are there lot of pending patches Sep 19 15:54:11 because the lack of review outside oe-core Sep 19 15:54:34 and the sparse development splited onmany layers Sep 19 15:54:44 hmm Sep 19 15:55:01 let the distro have theirs but keep common stuff together Sep 19 15:56:00 isnt it that way now Sep 19 15:58:00 khem: I know because today distro=buildsystem Sep 19 15:58:26 add some glue to that Sep 19 16:04:36 ant_work: in general I though oe-core updated recipes quite frquently Sep 19 16:04:50 so if you have any in mind that need update post patches Sep 19 16:04:58 yes, butthey overlook oe-classic and meta-oe Sep 19 16:05:03 see mtdutils case Sep 19 16:05:31 maybe we should pester more on the ML Sep 19 16:08:13 when a patch is posted you could review it and ask them to look at oe.dev for existing recipes Sep 19 16:08:29 but I will make a note of it and probably educate devs to peek into oe.dev Sep 19 16:08:35 before they commence an update Sep 19 16:08:37 khem: today I'm so angry because I've noticed how udev is still broken when using oe-core + meta-oe, *unless* (maybe) you use Angstrom Sep 19 16:08:53 now, oe-classic has the solutions (Tartarus) Sep 19 16:09:14 sometimes no one knows better so post patch as I said Sep 19 16:09:17 What'd I do? :) Sep 19 16:09:23 *g* Sep 19 16:09:28 the /tmp/shm/ patch Sep 19 16:09:53 ant_work: Oh, and on an unrelated note, I flew over you a few weeks ago, took Alitalia to fly to Greece Sep 19 16:10:18 tartarus hehe Sep 19 16:10:19 ant_work: it is a little bit sad but I am not sure what one can realistically do about it. I'm not sure it is reasonable to ask folks working on oe-core, who might not even use meta-oe, to check there every time they want to update a recipe. Sep 19 16:10:20 saw a blue ghost flying over my head... Sep 19 16:10:32 alitalia, I flew once Sep 19 16:10:37 greece is bankrupt Sep 19 16:10:40 more or less Sep 19 16:10:50 buy an island ;) Sep 19 16:11:01 I will bid for crete Sep 19 16:11:43 pb_: yes, it looks to me meta-oe is out of sight for many Sep 19 16:12:18 Well, the trip made me wish I had bought Flight Tracker Pro sooner and looked at these flights ontime percentage Sep 19 16:12:33 Oh well, was fun anyhow Sep 19 16:12:41 but, for meta-oe at least, I guess the other question is why these recipes were added to meta-oe in the first place rather than just updating the versions in oe-core. Sep 19 16:12:46 * Tartarus got to recall he almost speaks a tiny amount of italian :) Sep 19 16:12:56 heh Sep 19 16:12:57 pb_: we cannot change the world..I'd merge meta-oe and RIP Sep 19 16:13:08 the good bits of it Sep 19 16:13:32 let it be an abstruse dependency for some 433t distros Sep 19 16:13:38 Tartarus: you need gestures more in Italy :) Sep 19 16:16:31 pb_: eventually duplication should get smaller. I think it was a migration thing Sep 19 16:17:12 Hi, all! What is oe-classic everyone speaks about? Sep 19 16:17:29 slapin: the old monolithic oe tree Sep 19 16:17:48 (where recipes were listed alphabetically) Sep 19 16:18:02 ant_work: I guess that was more a feature of "ls" :-) Sep 19 16:18:07 * ant_work tears Sep 19 16:18:08 does it still maintained and updated? Sep 19 16:18:09 *g* Sep 19 16:18:19 slapin yes Sep 19 16:18:22 just as long as our layers don't end up leaning like the tower in Pisa Sep 19 16:18:30 pb_: there are undocumented tools to inspect the layers, though Sep 19 16:18:33 :) Sep 19 16:18:45 ant convice koen to test something else than angstroem Sep 19 16:18:53 but that is impossible Sep 19 16:18:53 ^_^ Sep 19 16:19:21 hi ka6sox Sep 19 16:19:29 morning woglinde Sep 19 16:19:36 tartarus holiday or buisness trip? Sep 19 16:19:37 that probably is impossible, but in koen's defence (!!) there is no reason that somebody else couldn't do that testing Sep 19 16:19:53 it does seem a bit much to expect him to test every distro single-handedly Sep 19 16:19:57 pb_, we have testing hardware doming... Sep 19 16:19:58 pb_: it's always bad if you don't build Angstrom for beagle Sep 19 16:20:01 pb yes Sep 19 16:20:04 s/doming/coming/ Sep 19 16:20:09 so ant should stop whining Sep 19 16:20:11 * woglinde runs Sep 19 16:20:11 ka6sox-away: cool Sep 19 16:20:12 heh Sep 19 16:20:22 ant_work: you say undocumented, but they do have built-in help ;) Sep 19 16:20:39 or rather, it does (bitbake-layers) Sep 19 16:20:49 woglinde: holiday Sep 19 16:20:52 bluelightning: sorry :) let say we (ol OE crowd) did not know them Sep 19 16:21:00 bluelightning: I think the word for that is "self-documenting" :-} Sep 19 16:21:01 ant_work: right ok Sep 19 16:21:06 pb_: :) Sep 19 16:21:08 Tartarus only italian? Sep 19 16:21:22 bluelightning: good kept secret :p Sep 19 16:21:46 I came in late...are we talking about testing meta-oe? Sep 19 16:21:55 meta-what ? Sep 19 16:22:00 ka6sox yes Sep 19 16:22:27 for the 5 qemu-* arches? Sep 19 16:22:32 ka6sox and the oe-core/meta-oe recipes overhang Sep 19 16:22:37 ups overlap Sep 19 16:22:43 right...we need to do that. Sep 19 16:23:05 ka6sox and more testing distros besides angstroem Sep 19 16:23:07 like micro Sep 19 16:23:25 woglinde: holiday in Greece Sep 19 16:23:35 woglinde: fwiw bluelightning did an awesome job putting meta-handheld together. We need simpad Sep 19 16:23:38 but we had stop-overs (longer than intended) in italy Sep 19 16:23:50 ant *g* Sep 19 16:24:17 ant_work: woglinde: simpad is in there, but probably needs work/testing Sep 19 16:24:33 right Sep 19 16:28:29 once we get the hardware checked out and setup automated testing of oe-core we can see what it will take to create a testing programme for meta-oe Sep 19 16:30:18 ka6sox-away is intel/linuxfoundation sponsoring the hw? Sep 19 16:30:36 LF is to my knowledge Sep 19 16:32:12 ah okay Sep 19 16:32:31 hm only nearly one month left to elce Sep 19 16:34:28 woglinde: I will be flying over EU Sep 19 16:34:32 during that time Sep 19 16:34:41 we'll look out for you Sep 19 16:34:45 I will try to look though binocularss :) Sep 19 16:35:47 if I use mmc2 on an omap3, does it appear as /dev/mmcblk1 or does it depend on detection order like USB flash drives? Sep 19 16:40:32 toofar what? Sep 19 16:40:46 what you mean wit mmc2? Sep 19 16:42:55 well, I have a craneboard Sep 19 16:43:03 it uses mmc1, and /dev/mmcblk1 shows up Sep 19 16:43:11 I have a custom board that uses mmc2 Sep 19 16:43:25 ? Sep 19 16:43:30 craneboard mmc1 -> /dev/mmcblk0 * sorry Sep 19 16:43:40 now you see Sep 19 16:43:44 custom board uses mmc2, so will it be be /dev/mmcblk1 Sep 19 16:44:03 you mean mmcX in u-boot values? Sep 19 16:45:25 nope, kernel values Sep 19 16:45:34 u-boot is difference and can see both mmc cards fine Sep 19 16:45:38 different* Sep 19 16:46:01 *sigh* Sep 19 16:46:14 I meant where did you get the mmc1 or mmc2 from? Sep 19 16:46:18 hw labeled? Sep 19 16:46:46 yeah, those are what the periphs are called in the datasheet Sep 19 16:47:38 okay Sep 19 16:47:51 depends on the soc kernel writer Sep 19 16:48:01 how they end up in the kernel enumeration Sep 19 16:48:29 ups sorry Sep 19 16:48:59 I meant depends on hacker who writes the boardfile Sep 19 16:49:30 so look always at the kernel source code if you are unsure Sep 19 16:49:54 for newer kernel it might be the new dts file Sep 19 16:49:58 so till later Sep 19 16:50:03 heading home now Sep 19 16:52:04 I wrote the board file Sep 19 16:55:43 bluelightning: worked fine Sep 19 16:55:47 bluelightning: thx by fixing it Sep 19 16:55:53 otavio: no problem :) Sep 19 16:59:02 toofar: the "first" card should always be mmcblk0 no matter what slot it's in Sep 19 16:59:15 but, well, can't you just try it and see? Sep 19 16:59:32 pb_: it's the rootfs and I can't get it to boot correctly Sep 19 17:00:22 pb_: on an evm I observe the behavior you describe Sep 19 17:01:46 ah. boot it from an alternate rootfs? initramfs? nfs? Sep 19 17:02:16 pb_: haven't thought about initramfs, nfs is hosed because I haven't configured the ethernet stuff yet on the custom board I'm trying to get up Sep 19 17:03:17 I think my real problem is that it just isn't detecting it even though I hve gpio_cd set correct in the board file Sep 19 17:03:49 and I know the board works since x-loader, u-boot, and kernel are loaded off the mmc card Sep 19 17:22:58 pb_: I got it working, I had a SN74AVCA406 transceiver configured wrong between the uP and the SD card Sep 19 17:22:59 thanks Sep 19 17:27:13 Strange. I need -lssp in eglibc-nativesdk and I don't need it in plain eglibc to build nscd correctly Sep 19 17:29:26 Moreover if I ask eglibc build to use -lssp, building fails for plain eglibc. Sep 19 17:54:58 lumag: eglibc needs to poke for ssp support and it does it with a runtime test IIRC Sep 19 17:55:13 lumag: that may not work with cross builds we do Sep 19 17:55:17 In both cases tests succeeds Sep 19 17:55:29 that may be the problem Sep 19 17:55:41 test is as simple as -c -x c -o /dev/null /dev/null Sep 19 17:55:57 yes it should do a link time test Sep 19 17:59:10 Hmmm. For target eglibc I really have __stack_chk_guard in ld-linux.so.3 Sep 19 17:59:23 for eglibc-nativesdk no that symbol in ld-linux Sep 19 18:01:01 Does anyone know of a way, in an oe(-core) recipe, to set dynamic runtime provides? The problem is that busybox, through it's potentially changing configuration file, produces a file "/etc/busybox.links" as an artifact of do_install Sep 19 18:01:27 Other packages may end up with dependencies on /usr/bin/env, which is generated by the pkg_postinst based on the contents of that file.. Sep 19 18:01:37 so we need a manual RPROVIDE of /usr/bin/env Sep 19 18:01:51 hi, what kind of environment are usable on zaurus sl-c3000? I wonder abuot the ram, there is only 64M Sep 19 18:02:03 I looked at kenrel and it seem very mainline Sep 19 18:02:10 now the problem is, how do I add the RPROVIDES based on the contents of /etc/busybox.links, and do_install runs, but before packaging? Sep 19 18:02:14 any suggestions? Sep 19 18:02:25 I looked at machine config and it use xf86-video-fbdev for xorg.... Sep 19 18:02:33 gnutoo, I use opie on my sl-6000. X11 + gpe are also useable Sep 19 18:02:43 ok nice Sep 19 18:03:07 opie is now up to date thanks to bluelightning Sep 19 18:03:36 I tried it on htcdream, and it seem that even sdl games can be played, if you use a special sdl version Sep 19 18:04:32 I may get one from mickeyl who doesn't want his old machines anymore(I think they takes space...) Sep 19 18:05:01 so before accepting I look if it can be usefull for me Sep 19 18:05:04 khem, it might be related to my icecc. I have to check Sep 19 18:05:24 is it still possible to view movies fullscreen? Sep 19 18:05:35 zbook says it can but it's kind of old Sep 19 18:05:46 and I guess there was a kdrive->xorg switch Sep 19 18:05:50 gnutoo, not now, we don't (yet) have libxine in oe-core infrastructure Sep 19 18:05:51 that lost the acceleration Sep 19 18:06:00 ok Sep 19 18:06:07 maybe I could import mplayer Sep 19 18:06:13 I need it for intone anyway Sep 19 18:06:34 and mplayer is immx If I remember well Sep 19 18:06:55 angelox had an issue with mplayer and immx in oe-dev some time ago.... Sep 19 18:07:25 khem, no, icecc is disabled for eglibc :( Sep 19 18:08:01 I guess the onmly remaining acceleration is immx and that bvdd Sep 19 18:08:04 is gone Sep 19 18:09:21 gnutoo, mplayer might be able to use pxa27x overlay on it's own Sep 19 18:09:41 ok nice Sep 19 18:10:00 like -vo omapfb for instance Sep 19 18:10:32 gnutoo, I don't know if it is/was supported anytime. Sep 19 18:10:55 ok Sep 19 18:11:06 zbook say there is a -vo bvdd Sep 19 18:13:08 so there was something Sep 19 18:13:49 Hi people Sep 19 18:13:57 hi Sep 19 18:14:17 Where can i get tzdata2011i.tar.gz? sources.openembedded.org doesn't seems to have it Sep 19 18:15:38 angelox|laptop: I've got it: c7a86ec34f30f8d6aa77ef94902a3047 tzdata2011i.tar.gz Sep 19 18:18:24 gnutoo: the thing is,the website from where the recipe is cloning doesn't have it no more (ftp://elsie.nci.nih.gov/pub/) Sep 19 18:19:06 i did download one from a debian mirror,let's check if md5sum matches Sep 19 18:19:51 ok Sep 19 18:21:33 it matches,thank you Sep 19 18:22:38 np, I did nothing... Sep 19 18:23:13 you did confirm me that the file exists and i'm not crazy by reading that name :) Sep 19 18:24:57 ok Sep 19 18:29:59 btw angelox|laptop do you know any details about pxa acceleration for graphics/video? Sep 19 18:30:14 since you have an ezx you might know.... Sep 19 18:33:10 gnutoo: no,sorry..that wasn't my work field... Sep 19 18:36:32 ok Sep 19 18:37:08 I was rather looking for generic infos like in what program it is supported, if it's mainline etc... Sep 19 18:37:36 not on the internals of how it works... Sep 19 18:42:45 what class of program? opengl/sdl stuff? **** ENDING LOGGING AT Tue Sep 20 02:59:56 2011