**** BEGIN LOGGING AT Wed Feb 03 02:59:57 2010 Feb 03 05:50:10 . Feb 03 06:28:31 Laibsch: Hi, pls can you take a look at http://patchwork.openembedded.org/patch/1525/ mainly the part about mirroring this tarball somewhere on OE, thanks Feb 03 06:29:32 Once it is in OE the mirroring will happen automatically (eventually) Feb 03 06:30:00 But please don't push URI that you know will be gone soon Feb 03 06:35:29 Laibsch: is it possible to put it somewhere manualy, I can host it for a while, but I don't know how much traffic I can handle Feb 03 06:36:04 there's a ton of options Feb 03 06:45:56 Laibsch: what do you mean? put it on github for example? Feb 03 06:46:32 github is obviously for git repos Feb 03 06:46:48 if you want to host tarballs, sf.net is just one option among thousands Feb 03 06:47:10 ok Feb 03 06:47:35 if you have just a small project I doubt many people will load your hoster anyway Feb 03 06:48:13 load = saturate Feb 03 06:48:29 I expect bandwidth demand to be reasonable Feb 03 06:48:31 in most cases Feb 03 06:48:49 it's a wifi driver for new USB wifi sticks Feb 03 06:48:54 ralink 3070 Feb 03 06:49:02 for example touchbook has this one also Feb 03 06:51:10 are you writing the driver? Feb 03 06:51:21 do you plan to include it in the kernel eventually? Feb 03 06:51:25 no, it's just recipe Feb 03 06:51:33 did you looked at that URL? :) Feb 03 06:51:39 http://patchwork.openembedded.org/patch/1525/ Feb 03 06:51:58 > +SRC_URI = "http://ynezz.ibawizard.net/dump/${TMP}.tar.bz2;name=tarball \ Feb 03 06:52:00 sure Feb 03 06:52:14 what's that tarball, then? Feb 03 06:52:34 source code of that driver Feb 03 06:53:06 http://bugs.openembedded.org/show_bug.cgi?id=5322 Feb 03 06:53:13 maybe you can look at that, too Feb 03 06:53:20 and fix it while you're at it Feb 03 06:53:41 http://www.ralinktech.com.tw/data/drivers/2009_0525_RT3070_Linux_STA_v${PV}.tar.bz2 Feb 03 06:53:46 something like that? Feb 03 06:54:05 yep Feb 03 06:54:12 where did you originally get the source code? Feb 03 06:54:18 that's what should be in SRC_URI Feb 03 06:54:56 from ralink Feb 03 06:55:07 but it's unstable URL Feb 03 06:55:40 does it works for you? Feb 03 06:58:08 I've had this recipe for about a two months in my branch Feb 03 06:58:35 I tried to build it before posting the patch to OE Feb 03 06:59:05 and it wasn't possible to download it from that ralink URL anymore Feb 03 06:59:50 does it work for me? I don't care about that particular source. I don't have the hardware, so why should I download the source? Feb 03 07:00:15 ok, nevermind then Feb 03 07:00:23 Have you checked for newer drivers? Feb 03 07:00:31 maybe they removed the old one Feb 03 07:00:39 yes, it's the latest Feb 03 07:01:15 you are telling me that the manufacturer pulled the driver source code for a particular, current hardware? Feb 03 07:01:20 * Laibsch finds that hard to believe Feb 03 07:02:04 well it's possible, but without direct URL anymore Feb 03 07:02:33 youhttp://www.ralinktech.com/license_us.php?n=2&p=0&t=U0wyRnpjMlYwY3k4eU1EQTVMekV4THpJd0wyUnZkMjVzYjJGa01EVTRNalU1TlRjek1DNWllakk5UFQweU1EQTVYekV4TVRCZlVsUXpNRGN3WDB4cGJuVjRYMU5VUVY5Mk1pNHhMakl1TUM1MFlYST1D Feb 03 07:02:42 you mean they mask the real URI with some Javascript stuff? Feb 03 07:02:43 you need to agree with license first Feb 03 07:02:50 some PHP stuff Feb 03 07:02:51 I see Feb 03 07:03:08 I guess they may want you to agree with the license for a reason? Feb 03 07:03:23 We don't mirror the TA stuff, for example Feb 03 07:03:24 GNU GPL license? Feb 03 07:03:26 TI Feb 03 07:03:37 so, they are just stupid ;-) ? Feb 03 07:03:51 as GPL would be freely redistributable Feb 03 07:03:57 great, now you get it :) Feb 03 07:04:20 well, you never mentioned obfuscation or something like that Feb 03 07:04:46 yep, I started from the wrong end, sorry :) Feb 03 07:06:26 how tight is your bandwidth? Feb 03 07:06:43 it's only 0.5M and people won't be pulling this by the minute Feb 03 07:06:52 I expect very low BW requirements Feb 03 07:08:06 ok, so no problem then Feb 03 07:08:17 it's 100mbit Feb 03 07:08:50 But let's stop wasting time ;-) Feb 03 07:09:05 but I don't know what would they say for 200GB/month traffic :) Feb 03 07:09:08 I would have had no issue hosting this on my 20K upstream Feb 03 07:09:20 and you were worried when you have 100Mbit? Feb 03 07:09:22 ok, zecke told me different Feb 03 07:09:23 Buhahaha Feb 03 07:09:34 and said to ask you Feb 03 07:09:46 so I don't know, never hosted anything for OE Feb 03 07:09:52 200GB/month = 400 million downloads Feb 03 07:09:57 you gotta be kidding! Feb 03 07:10:12 40 million Feb 03 07:10:13 but still Feb 03 07:10:14 well, I don't know how many users use OE Feb 03 07:10:25 you believe OE has 40 million users? Feb 03 07:10:33 you believe they all have this HW? Feb 03 07:10:43 no Feb 03 07:10:44 you believe they kill their source dir every mont? Feb 03 07:10:46 month Feb 03 07:10:49 NO! Feb 03 07:10:56 ok, you've point Feb 03 07:11:02 no problem then Feb 03 07:11:04 thanks :) Feb 03 07:11:15 Anyway Feb 03 07:11:19 Since you were scared Feb 03 07:11:21 http://paste.debian.net/58528/ Feb 03 07:11:23 ;-) Feb 03 07:11:39 http://sources.openembedded.org/2009_1110_RT3070_Linux_STA_v2.1.2.0.tar.bz2 should now work Feb 03 07:12:15 d'oh Feb 03 07:12:22 thanks! Feb 03 07:12:31 please, be reasonable next time ;-) Feb 03 07:12:35 thank you for your contributions Feb 03 07:13:21 I wonder how to reply to this http://patchwork.openembedded.org/patch/1089/ Feb 03 07:13:58 Message-Id: <1254668193.18167.4.camel@leeloo.somedomain> Feb 03 07:14:11 would it be enough to forge that in Reply-To ? Feb 03 07:14:40 In-Reply-To even Feb 03 07:17:56 RP: wrt yesterdays problem: could use advice on whether also to patch the rpm and deb part, ipk part seems to work (actually the eating of the pudding will be done in a few minutes; rm TMPDIR and rebuild console-image) Feb 03 07:19:48 ynezz: that's the same as the bug ticket I gave you earlier Feb 03 07:19:52 you can comment there Feb 03 07:21:15 ok Feb 03 07:53:48 good morning Feb 03 08:08:22 should openembedded be useable by calling bitbake with "setarch i686"? Feb 03 08:25:20 Weaselweb: if you mean MACHINE=i686-generic, it should Feb 03 08:26:26 mckoan: so "setarch i686 bitbake console-image" should work? Feb 03 08:31:18 Weaselweb: never used setarch Feb 03 08:32:04 well, i need to create a i686 toolchain on a x86_64 machine, but a virtual machine is unuseable for that Feb 03 08:35:02 Weaselweb: MACHINE=i686-generic bitbake minimal-image Feb 03 08:39:53 mckoan: it is still building x86-64-linux Feb 03 08:40:07 should I remove some already build directories? stamps, etc? Feb 03 08:51:24 Weaselweb: not sure I understand what you want to get Feb 03 08:52:26 I'm currently build an angstrom image, but on my x86_64 it (obviously) creates 64bit binaries for toolchain. but I need a 32bit toolchain, so it can be used on other computers Feb 03 08:54:58 Weaselweb: aha, now it's clear. You have to use Canadian cross compilation, and I never knew anybody who used it successfully Feb 03 08:55:09 :( Feb 03 08:56:54 Weaselweb: just download a 32 bit linux vm somewhere and build in there Feb 03 08:57:41 eFfeM_work: this is not useable it took about 12 hours only to build until mpfr, not even gcc... Feb 03 08:58:18 mh, I created a new build dir and run "linux32 bitbake qt-image" and this at least creates a i686-linux stamp directory. it seems I cannot not mix architectures Feb 03 09:00:40 Weaselweb: you are on 64 bit or on 64 Mhz ? I have a 3 Ghz dual core from 4 years ago (D920) started clean build on it for two architectures (armv5te and armv7) yesterday at around 10 pm, was finished this morning around 4 am, bott times bitbake console-image Feb 03 09:01:37 I have a crappy mainboard which slows the harddisk access massivly Feb 03 09:15:42 Weaselweb: Poky can do that;-) Feb 03 09:16:19 RP: fyi, just finished rebuilding console image from packaged staging without any build problem (haven't actually loaded it into hw) Feb 03 09:17:23 eFfeM_work: great :) Feb 03 09:23:07 RP: should I also adapt the rpm and deb parts for using DISTRO_PR ? Never used those parts so no idea if they use DISTRO_PR Feb 03 09:33:24 morning Feb 03 09:35:38 hi hrw Feb 03 10:00:42 eFfeM_work: See what the package_deb and package_rpm classes use when creating them? Feb 03 10:08:01 RP will check that, good plan. Guess eFfeM_home will work on it tonight :-) Feb 03 10:21:31 * mckoan wonders if freenode.net spam issues are still necessary, mickey|zzZzzz could we remove +R ? Feb 03 10:23:03 mckoan: you need to bugger kergoth about it Feb 03 10:23:12 but he seems to be deaf on that ear Feb 03 10:32:22 yeah, stupid freenode Feb 03 10:34:50 good morning Feb 03 10:38:56 stupid glibc Feb 03 11:16:36 bye Feb 03 12:38:25 errr hi... need some help with managing packages... Feb 03 12:38:35 bitbake specifically... Feb 03 12:42:24 Managing packages in which way? Feb 03 12:42:52 making use of the 'BBEXTENDCLASS' functionality to create native packages Feb 03 12:56:01 i am trying to create native packages using the BBCLASSEXTEND... it works okif i change the main recipe of the package... can it be done using the amend.inc ?? Feb 03 14:31:33 re Feb 03 14:32:22 ;) Feb 03 14:33:45 guys... need some help with BBCLASSEXTEND and amend.inc ! Feb 03 14:34:29 * kergoth 's never tested those two together, but don't see any reason it wouldn't work. run bitbake -e foo | grep \^BBCLASSEXTEND= and see if its set the way you think it is Feb 03 14:37:30 kergoth: no environment output... just the error of "Nothing Provides" Feb 03 14:37:37 no.. Feb 03 14:37:43 bitbake -e on the original recipe, not the -native Feb 03 14:37:51 okk Feb 03 14:37:53 you can't get -e output on a provider that doesn't exist. Feb 03 14:42:02 kergoth: i get the following errors Feb 03 14:42:03 NOTE: Runtime target 'eglibc' is unbuildable, removing... Missing or unbuildable dependency chain was: ['eglibc', 'subversion-native', 'neon-native', 'gnutls-native', 'libgcrypt-native'] Feb 03 14:43:34 though BBCLASSEXTEND="native" does show up in the env output Feb 03 14:46:36 kergoth: if i do a dry run, i get the following error Feb 03 14:47:10 Parsing .bb files, please wait...ERROR: 'virtual:native:/home/al/release/openembedded/recipes/libgcrypt/libgcrypt_1.1.91.bb' while parsing /home/al/release-/openembedded/recipes/libgcrypt/libgcrypt_1.1.91.bb Feb 03 15:56:38 Hi, all. Within a recipe, I need to find out the version of another package, which was built before. Is this possible? How? Feb 03 15:57:51 nope Feb 03 16:10:14 sure? reason for asking is: I have initramfs image recipe, which depends on the kernel (modules get installed into initramfs), afterwards I create a uImage of kernel+initramfs and there I would like to have the version of the previously built kernel package Feb 03 16:14:17 bye Feb 03 17:19:22 hmm, would be nice if a missing RRECOMMEND didn't make a target unbuildable Feb 03 17:34:26 kergoth: I'd call that a bug in the metadata though ;-) Feb 03 17:35:07 It says recommend instead of depend? :P Feb 03 17:35:15 well, think about it from the perspective of someone starting something like poky. if they don't want the rrecommended bits, they'd have to patch up the recipe and deviate further from upstream Feb 03 17:39:53 kergoth: I understand why it'd cause problems in that scenario Feb 03 17:40:24 RP__: I've always seen Recommends as a packaging thing though and expected the higher level .bb files to be consistent Feb 03 18:30:00 GNUtoo: eee701: does that boot from SD and/or USB or only from internal Feb 03 18:30:10 * eFfeM1 is building it now :-) Feb 03 18:48:18 XorA, I'm going to try and reproduce the problem they are blaming on you, but providing zero data :) Feb 03 18:58:20 Crofton: I can reproduce "it doesnt build" simply by rm -rf org.openembedded.dev Feb 03 18:59:29 :) Feb 03 18:59:34 I pulled tmp Feb 03 19:00:03 we'll see what happens Feb 03 19:00:34 I did a build from scratch after glib changes Feb 03 19:01:03 I know a little about the guy having issues, so I feel a little guilty Feb 03 19:01:30 I wouldnt mind looking at it, but the email doesnt even tell me the recipe or error seen Feb 03 19:03:34 anyway time to do the socialising thing Feb 03 19:03:53 What sort of geek are you? Feb 03 19:04:00 Hand in your badge this instant! Feb 03 19:04:02 failed and glad Feb 03 19:04:24 * XorA is going to where the chicks is at Feb 03 19:04:35 :) Feb 03 19:11:17 where? Feb 03 19:19:36 Crofton|work: XorA good luck :) Feb 03 19:23:34 re Feb 03 19:36:02 XorA: just did a build of glibc for eeepc from the head w/o problem; no problems for beagle or sheeva either yesterday (not sure if that was exactly dev hea though) Feb 03 19:53:13 eFfeM1: you changed accidentally also recipes/u-boot/u-boot_1.1.4.bb in your last commit.. Feb 03 19:59:40 JaMa|Gone: ouch, thought i checked it so carefully Feb 03 19:59:50 JaMa|Gone will fix Feb 03 20:00:18 JaMa|Gone: tought you were gone but i have an issue with libx11, for 1.3.3. you commented out a patch but I think it is needed Feb 03 20:00:43 eFfeM1: which one? Feb 03 20:00:44 libx11 does not build for me for x686 machines (eeepc) Feb 03 20:00:50 the one for libtool Feb 03 20:01:10 eFfeM1, sd usb and internal are displayed in the bios boot menu Feb 03 20:01:32 this one file://dolt-fix.patch;patch=1 \ Feb 03 20:01:43 eFfeM1: if it helps for you.. feel free to uncomment it.. Feb 03 20:01:55 JaMa|Gone: I'm testing it Feb 03 20:03:40 JaMa|Gone: git revert HEAD undoes my commit right ? Feb 03 20:04:46 eFfeM1, git revert hash Feb 03 20:04:52 with hash beeing a git hash Feb 03 20:05:17 eFfeM1, ahh this one...I didn't commit the fix yet Feb 03 20:05:37 eFfeM, you need another better dolt patch Feb 03 20:05:53 simply remove DOLT from configure.ac or configure.in Feb 03 20:07:11 eFfeM1: if you can partially revert it I think you have to commit it as new one removing that u-boot part Feb 03 20:07:25 s/can /want / Feb 03 20:07:34 JaMa|Gone: i will fully revert it and recommit Feb 03 20:07:51 then as GNUtoo said Feb 03 20:07:54 GNUtoo: there is a patch in the recipe but that was commented out Feb 03 20:08:02 eFfeM1, for x11 should I commit the fix? (will take some time to check) Feb 03 20:08:28 no I didn't mean that...I mean remove the commented patch and make your own with the instructions above Feb 03 20:08:40 beacuse the commented one may be not good Feb 03 20:09:46 I didn't test the commented patch...but it didn't apply if I remember well Feb 03 20:10:09 but I know for shure that my way works well.... Feb 03 20:10:22 with all arches etc... Feb 03 20:10:50 JaMa|Gone: change has been reverted and recommitted w/o the u-boot change Feb 03 20:11:09 eFfeM1, btw for reverting a commit: git log gives you the commit list...pick the hash...then git revert picked_hash Feb 03 20:11:16 ok Feb 03 20:11:41 eFfeM1: thanks Feb 03 20:11:41 I take time to commit and check extensively commits but sometime screw things because I do errors with git Feb 03 20:14:02 GNUtoo: the patch that was commented out works fine and patches acinclude.m4, just tested it for eee701, works fine for me, now testing if it also still builds for beagle, if so, I see no reason to make a new patch, this one looks quite fine to me Feb 03 20:14:18 ok Feb 03 20:14:22 JaMa|Gone: thanks for pointing me to the problem Feb 03 20:14:25 if it works Feb 03 20:14:39 you're talking about libx11 right? Feb 03 20:14:45 GNUtoo: what did you build for eeepc? x11-image ? Feb 03 20:14:49 GNUtoo: yes, libx11 Feb 03 20:14:57 ok so we're talking about the same thing Feb 03 20:15:01 no xfce46-image Feb 03 20:15:05 I wrote a mail about that Feb 03 20:15:22 actually i was somehow surprised that libx11 got build as I was building console-image Feb 03 20:15:29 someone asked about the inclusion of eee701 into angstrom Feb 03 20:15:32 yeah saw your email Feb 03 20:15:35 ok Feb 03 20:16:08 GNUtoo: saw that one too, actually I have all the stuff needed to make the autobuilder changes Feb 03 20:16:17 ok Feb 03 20:16:31 decided I wanted to try to build first (which is now with libx11 proven to be a good plan :-) ) Feb 03 20:16:40 GNUtoo: btw you were coming to fosdem ? Feb 03 20:16:43 yes Feb 03 20:16:50 if things go right Feb 03 20:17:02 if I get things running for eeepc i might bring mine as well Feb 03 20:17:16 * eFfeM1 as a 1st gen one with chinese kbd :-) Feb 03 20:17:42 I will bring the one I have,because my regular laptop screen backlight is broken Feb 03 20:17:55 I use it as a build system now... Feb 03 20:18:03 (the regular laptop) Feb 03 20:18:17 pentium M 2GHz@less(heat problems) Feb 03 20:18:43 so I heard that someone will bring a regular laptop to show oe Feb 03 20:18:59 s/so/and Feb 03 20:19:45 mine had 2*500GB hdd(raid0 => 1TB) Feb 03 20:20:14 (usefull if you have different machine with different distros) Feb 03 20:20:54 i don't have a laptop, so can't bring it, had to return it when i lost my job and my new employer only gave me a desktop (but then again it is a fast one) Feb 03 20:21:13 pacakged staging also greatly helps to reduce disk usage Feb 03 20:21:47 ok :( Feb 03 20:22:51 what's that? Feb 03 20:23:34 GNUtoo: read this: http://marcin.juszkiewicz.com.pl/2008/07/01/packaged-staging-and-what-it-gives/ Feb 03 20:24:48 thanks a lt Feb 03 20:24:52 *lot Feb 03 20:25:00 rm_work is not an option for me Feb 03 20:25:05 so that may be a lot better Feb 03 20:44:57 eFfeM, thanks a lot...do you have an idea about where the staging tarballs end? Feb 03 20:45:20 in the DEPLOY_DIR_PSTAGE dir Feb 03 20:45:24 for me pstage Feb 03 20:45:32 ah ok I should set it I think Feb 03 20:46:09 note that you'll need a full rebuild so better set up something for overnight Feb 03 20:46:25 and make sure you're at dev head, you need the patch I pushed earlier this evening Feb 03 20:46:36 have been fighting for several days to get it running Feb 03 20:47:03 ok so I'll enable it later Feb 03 20:47:07 thanks a lot Feb 03 20:47:09 my pstage with 3 architectures (armv5, armv7, i686) in it is about 2.5 GB Feb 03 20:47:11 np Feb 03 20:52:47 re Feb 03 20:53:18 jo hrw Feb 03 20:53:54 mmm e-image-core.bb depends on kdrive... Feb 03 20:54:46 IMAGE_INSTALL add that: xserver-kdrive-fbdev Feb 03 20:55:21 and of course there is also that: xserver-kdrive in DEPENDS Feb 03 20:55:39 should it be changed for XSERVER or something like that? Feb 03 20:56:55 what is faster? beagle board or sheevaplug? Feb 03 20:57:09 good question Feb 03 20:57:12 eFfeM1, ^^^ (for the speed question) Feb 03 20:58:21 because one is armv7 but has 600 or 720(igepv2 ) Mhz Feb 03 20:58:34 GNUtoo depends on what you want to do I guess, didn't really bench them Feb 03 20:58:45 and the other is armv5 and bug has 1.0 or 1.2 GHz (I don't remember) Feb 03 20:58:51 ok Feb 03 20:59:12 actually not even sure if my beagle is running at 500 or 600 seem to recall that in the early days it was 500, not sure if I moved to 600 Feb 03 20:59:19 ok Feb 03 20:59:19 sheeva is 1.2 ghz Feb 03 20:59:23 maybe bootloader settings Feb 03 20:59:24 ok Feb 03 20:59:40 saw a notice of someone who overclocked his beagle to 900 mhz but that was not too stable Feb 03 20:59:46 lol Feb 03 20:59:46 haven't tried it mysefl Feb 03 20:59:50 better not Feb 03 21:00:10 i've clocked my 701 to 900 iso the default 630 Feb 03 21:00:15 but that is not really overclocking Feb 03 21:00:20 me too p4-module or something Feb 03 21:00:25 yup Feb 03 21:01:17 btw what should I do for e-image-core> Feb 03 21:01:28 s/>/?/ Feb 03 21:14:32 hi all Feb 03 21:15:26 i would like to know what of the hardware i have will be interesting for fosdem booth ... Feb 03 21:16:46 gremlin[it]: take EVBeagle Feb 03 21:16:54 for sure ... Feb 03 21:17:09 most people knew normal BB so blue one will be something different Feb 03 21:17:35 I broke my expansion board for BB so will not take any BB Feb 03 21:17:42 hrw ... mainly i'm looking if i carry my laptop with me or not ... if yes ... i have no space for the gift in the return fly :D Feb 03 21:17:45 can evbeagle is a beagle board without 3d? Feb 03 21:17:49 s/can/an Feb 03 21:18:03 GNUtoo: no, it is german clone Feb 03 21:18:10 GNUtoo, no should be full Feb 03 21:18:14 ok Feb 03 21:18:53 If I buy something like that I would prefer igepv2 with 512M of ram and 720Mhz cpu Feb 03 21:19:29 GNUtoo: sure, but what if you will get BB for free? Feb 03 21:19:35 GNUtoo, i agree ... i had it for a cheap price Feb 03 21:19:40 ok Feb 03 21:19:45 there are free BB? Feb 03 21:19:57 GNUtoo: not quite Feb 03 21:19:59 I missed the one in the electronic shop that gave free ones Feb 03 21:20:26 GNUtoo: I got B7 and C3+BeagleBug from Bug Labs and then C3 from TI Germany Feb 03 21:20:29 https://www.sparkfun.com/commerce/news.php?id=305 Feb 03 21:20:34 ok Feb 03 21:21:08 ... more ... the NGW100 could be interesting ... it run micro-image Feb 03 21:21:26 too bad they also sold breakout htc cables....that would have given me serial on htcdream which would have given me kgdb which would have saved me much time Feb 03 21:21:42 maybe time to order one when I comes back from fosdem Feb 03 21:22:09 also I was thinking the igepv2 is spanish...so it's in the european union...so no custom taxes Feb 03 21:22:57 I don't know if my wrt54gsv4 can be usefull and I don't know if I can bring it but it has serial cable Feb 03 21:24:02 http://gnutoo.homelinux.org/blog/images/open_router.jpg Feb 03 21:24:39 but it seem a dangerous device while open Feb 03 21:25:03 it gives me some little headaches when the wifi comes on Feb 03 21:25:10 not present when it's closed Feb 03 21:26:39 hi. how can i get a list of recipe tasks which are actually executed for a specific recipe and in what order? Feb 03 21:28:19 matgnt: you cant Feb 03 21:28:48 matgnt: you can find them in tmp/stamps/ anyway but OE can run some of tasks in any order Feb 03 21:30:35 btw what should I do for e-image-core? it hardcodes kdrie Feb 03 21:30:39 *kdrive Feb 03 21:30:49 GNUtoo: wrt54gsv4 - do we have any useful image for it? Feb 03 21:31:01 GNUtoo: image hardcodes it? Feb 03 21:31:31 GNUtoo: remove it and send patch Feb 03 21:31:42 hrw: thanks for the hint with stamps. that is enough for my needs. Feb 03 21:32:07 hrw, ok thanks Feb 03 21:32:15 for wrt I used micro Feb 03 21:32:33 micro-base-image from NFS or something like that Feb 03 21:32:35 so not very usefull Feb 03 21:32:52 ok Feb 03 21:33:06 I have wrt54gs v1.1 so 8MB flash ;) Feb 03 21:33:15 and time to go for me Feb 03 21:34:17 ok Feb 03 21:58:30 i have a bb without the beagleboard lettering does that also count as intresting? Feb 03 21:59:14 eFfeM: hey..seems you did a great job wrt packaged-staging Feb 03 21:59:36 have you solved the arch mismatch too? Feb 03 22:03:04 ant__: not sure, haven't seen it since Feb 03 22:03:20 might be fixed as side effect Feb 03 22:03:33 i did look at it and the Architecture line was there Feb 03 22:03:39 will see if I get the problem again Feb 03 22:04:17 ..because staging-sysvinit-beagleboard-angstrom-linux-gnueabi_2.86-r57_i686-linux.ipk Feb 03 22:04:23 is still wrong Feb 03 22:04:56 ok, I'll RFS Feb 03 22:05:43 define wrong and expand RFS Feb 03 22:06:17 beagleboard is not i686 Feb 03 22:06:26 Rebuild From Scratch Feb 03 22:06:34 (a.k.a. the hard way) Feb 03 22:07:07 you'll indeed need a rebuild from scratch Feb 03 22:07:14 :) Feb 03 22:08:00 The i686 is still there in the name, thougth it was needed because it is a staging package to be expanded on i686 Feb 03 22:08:17 RP can you say something about that ? Feb 03 22:08:31 the old problem: Feb 03 22:08:32 * Package staging-glibc-armv5te-angstrom-linux-gnueabi version 2.9-r35.2 has no architecture specified, defaulting to i686-linux. Feb 03 22:08:51 * Package staging-linux-rp-c7x0-angstrom-linux-gnueabi version 2.6.26-r11 has no architecture specified, defaulting to i686-linux. Feb 03 22:09:02 multiarch and machine specific Feb 03 22:09:28 were still defaulting to buildhodt-arch (i686 here) Feb 03 22:09:52 the odd thing is that the control file does specify an architecture Feb 03 22:09:53 that was last week..crossing fingers.... Feb 03 22:10:09 Anyone having success with .32 kernel and dss2 on beagleboard? Feb 03 22:14:52 eFfeM1: for the moment pstage resides inside tmpdir... Feb 03 22:21:31 reboot Feb 03 22:23:17 khem, can you recommend an arm (preferably softfp, mmu, eabi) conf/machine/*.conf that i could use to see if nptl-ppc broke something outside of ppc? Feb 03 22:23:23 ant__: I moved pstage out of tmpdir as per hrw's instructions on http://marcin.juszkiewicz.com.pl/2008/07/01/packaged-staging-and-what-it-gives/ Feb 03 22:24:01 eFfeM: my builds are hit by package_Stagefile_shell Feb 03 22:26:17 ant, had no problem there, don't really understand it, but denix has a patch Feb 03 22:27:13 see this: http://blog.denix.org/2008/09/getting-even-more-dangerous-with-pstage.html Feb 03 22:27:19 but didn't run into problems there Feb 03 22:27:45 anyway, calling it a day for now. Feb 03 22:27:48 nite all Feb 03 22:31:09 'nite Feb 03 22:31:38 btw it was me replying in the ML :) Feb 03 22:40:02 I wonder what's the best illume/enlightenment image...I'll try to bitbake e-image with kdrive->${XSERVER} Feb 03 22:46:27 define "best" Feb 03 22:48:48 mmm something usable on all devices Feb 03 22:49:04 something that let you choose the profile Feb 03 22:49:10 (maybe they all do that) Feb 03 22:49:33 and that has only illume/enlightenment **** ENDING LOGGING AT Thu Feb 04 02:59:57 2010