**** BEGIN LOGGING AT Wed Oct 07 14:49:23 2009 Oct 07 15:03:11 can i create a bb file which is only altering the rootfs a bit and coping files into it . So it also doesnt have SRC_URI but lives out of the mybbpkg/files directory ? Oct 07 15:07:41 rob_w: uh, files in the files directory are in SRC_URI. they'r ethe file:// urls Oct 07 15:08:16 right i saw that now .. so it should be any problem at all having ONLY file:// stuff going in Oct 07 15:08:27 right Oct 07 15:08:31 oe doesn't care where the stuff comes from Oct 07 15:08:43 everything in SRC_URI is fetched by the do_fetch task provided by base.bbclass Oct 07 15:08:59 you could, in fact, provide a custom do_fetch, however, which did nothing, or what you want directly, rather than relying on the default Oct 07 15:09:07 but that's really unnecessary, i expect Oct 07 15:30:55 hi Oct 07 15:31:21 I'm trying to build angstrom toolchain and I bumped into an error Oct 07 15:31:28 I get http://linux.fjfi.cvut.cz/~zub/binutils-cross-sdk.log Oct 07 15:32:16 anybody know why is that happening? Oct 07 15:39:05 (i.e. what is it that I ...ed :) Oct 07 15:50:33 I think probably the sdk recipes are just a bit broken. Oct 07 15:51:27 ok, I'll try to look into it a bit Oct 07 15:51:33 This worked for me a short time ago... Oct 07 15:51:49 But I currently get a similar error wth Python Oct 07 15:51:53 kergoth: heh, my beagleboard arrived. it is rather smaller than I was expecting, quite cute. Oct 07 15:52:03 sadly it seems to lack a video output port which is a bit of a show-stopper for my purposes, doh Oct 07 15:52:41 ah, no, I see, that little connector labelled dvi-d must be the one. Oct 07 15:52:45 florian: I also build that ok some ~ 2 weeks ago on different box (and arch) Oct 07 15:52:50 I wonder what sort of plug this is. Oct 07 15:52:50 I'll try on that other box Oct 07 15:53:56 pb__: HDMI it is Oct 07 15:54:19 pb__: yep... that's just DVI with the small connector used for hdmi Oct 07 15:54:38 right, thanks Oct 07 15:54:43 but it lacks some hdmi features... Oct 07 15:55:02 luckily you can get adapters to regular dvi very cheap Oct 07 15:55:55 ok, great Oct 07 15:56:02 DVI-D connector was too big so they used HDMI one Oct 07 15:56:13 right Oct 07 15:56:23 seems a bit weird to fit the huge s-video connector if they were worried about size, but still. Oct 07 15:56:25 pb__: beagleboard manual is handy btw Oct 07 15:57:22 pb__: heh.. true, I had to remove it to fit it in a case for some experiments :) Oct 07 15:58:54 right, I have ordered an hdmi-to-dvi cable. £2.20 from ebay, I guess that's not so bad. Oct 07 16:00:35 I use hdmi<>hdmi + hdmi-dvi adapter Oct 07 16:00:42 I didn't manage to wait for it and bought one nearby ;) Oct 07 16:00:56 heh Oct 07 16:01:22 yeah, I am too lazy to go to the computer store. Oct 07 16:02:07 :-) Oct 07 16:07:51 hm, stupid gthumb, why doesn't it cache my photos locally? Oct 07 16:08:01 very tiresome that it insists on re-indexing them every time I plug the camera in Oct 07 16:08:33 pb__: you should feel lucky it doesnt just crash after random numbers of photos :-) Oct 07 16:08:34 seems like an obvious thing for them to have missed Oct 07 16:08:39 comes from not talking to your userbase much, i expect Oct 07 16:08:41 heh Oct 07 16:08:56 heh, probably Oct 07 16:09:08 it is one of my least favourite pieces of software, but I don't know of anything better Oct 07 16:09:11 Possibly it's just difficult for them to detect if the photos have changed. Oct 07 16:09:44 broonie: you can make the assumption same name means no change then index in background Oct 07 16:09:52 mm, maybe. but, how often do you change the photo without changing its name? Oct 07 16:09:58 XorA: I wouldn't trust that for a camera. Oct 07 16:09:59 broonie: I havent met a camera yet that re-used numbers Oct 07 16:10:14 I would be happy with an outdated thumbnail if I had hacked the images on the card by hand. Oct 07 16:10:46 XorA: or you did not made enough photos Oct 07 16:10:48 XorA: I'd be entirely unsurprised if there were some, though. Oct 07 16:11:12 broonie: then they are not common, and the common case is fast, the uncommon case take a little longer to update thumbnail Oct 07 16:11:34 The trouble is detecting when the uncommon case happened, though. Oct 07 16:11:59 broonie: if you're going to reindex in the background anyway (per xora's suggestion) then you don't need to detect it. Oct 07 16:12:02 for .jpg file size is probably enough information. Oct 07 16:12:13 the only effect would be that it takes a moment for the thumbnail to become correct, which is not the end of the world Oct 07 16:13:11 anyway, while I was venting my spleen there it has finally finished indexing my photos so I can now get on with my intended task :-} Oct 07 16:13:20 hehe Oct 07 16:13:21 it'd be way more than a moment but yeah. Like I say, I think looking for filesize changes would do the trick anyway. Oct 07 16:14:48 yes, that would work Oct 07 16:16:06 maybe I should send patches and/or hatemail to the gthumb people Oct 07 16:18:16 not really familiar with gthumb, but i've heard good things about F-spot. maybe see how that is? Oct 07 16:18:32 ah, never heard of that Oct 07 16:18:36 I'll give it a try Oct 07 16:18:38 Yeah, I use f-spot for that sort of stuff - seems to work pretty well. Oct 07 16:19:26 generates pretty web galleries :-) Oct 07 16:20:42 * pb__ waits for slow apt-get Oct 07 16:20:58 * XorA yawns and decides to call it a day Oct 07 16:22:23 anyone played with the fish shell? Oct 07 16:22:25 seems kind of cute Oct 07 16:24:20 I am lazy and use bash Oct 07 16:25:07 it does sound cute Oct 07 16:25:34 judging from the wikipedia article, anyway :-} Oct 07 16:25:36 * eFfeM thinks the fish shell is somewhere in the C shell Oct 07 16:25:57 the "universal variables" thing sounds a bit freaky though. does it really allow you to assign arbitrary things in other people's shells? Oct 07 16:26:07 03Marcin Juszkiewicz  07stable/2009 * rf923e40b45 10openembedded.git/ (3 files in 3 dirs): Oct 07 16:26:07 linux-bug: updated to newer svn revision, tweaked kernel configuration Oct 07 16:26:07 Signed-off-by: Marcin Juszkiewicz Oct 07 16:26:07 Acked-by: Philip Balister Oct 07 16:26:24 oh, maybe this is just crummy wikipedia editing, I guess they are missing an apostrophe Oct 07 16:26:54 "... universal variables, which allow a user to permanently assign a value to a variable across all the users running fish shells" Oct 07 16:27:05 presumably that should be "... all the user's running fish shells" Oct 07 16:27:20 they're universal to you, it just persists them to a file in your xdg config dir for fish, ~/.config/fish or whtaever Oct 07 16:27:32 just means you dont have to screw with rc files if you just want to tweak something Oct 07 16:27:52 right, that does sound handy Oct 07 16:27:53 what's kind of weird about it is they truly emphasize simplicity, so they drop posix compliance, remove the many ways of doing something in favor of one Oct 07 16:28:14 so it's not if foo; then bar; fi, it's if foo; bar; end Oct 07 16:28:30 yeah, that doesn't sound too unreasonable. initially I was a bit startled by that, but thinking about it more I can't really see why I care about posix compliance for my interactive shell Oct 07 16:28:33 and no `` and $(), just () Oct 07 16:28:39 i think i like it for interactive, yeah Oct 07 16:28:44 can always make a .sh for the real stuff Oct 07 16:28:48 right Oct 07 16:29:16 the only thing I can think of that would be annoying is projects that require you to "source complicated_bash_setup_script.sh" before doing anything. Oct 07 16:29:32 hehe, yeah, i just had to create a setup.fish in my mvl6 project, actually.. Oct 07 16:29:47 you'd end up having to use bash for interactive use in those kinds of projects (or cloning the setup script) Oct 07 16:29:59 yep Oct 07 16:34:59 anyone here that uses vixie cron via openembedded? is appending a task to /var/cron/tabs/ the right approach to get something done by cron ? Oct 07 16:35:19 and is this done via pkg_postinst() routine ? Oct 07 16:36:05 I'm not quite sure what you mean by "via openembedded". are you talking about installing cron in an oe-build rootfs? Oct 07 16:36:24 hrw: if you are applying those FILESPATHPKG glibc changes to stable you may also want 3bfbef2d3ab4e6ca62be3c509850b1d6793c193b, without it glibc-2.3.2 breaks Oct 07 16:36:51 if you want to have a cronjob included in the shipped system, your best option is probably to put it either in /etc/crontab or /etc/cron.d or some such. /var/cron/.. would probably be ok too. Oct 07 16:37:16 pb__: yeah "via openembedded" => installed in the rootfs Oct 07 16:37:46 msmith_: fetch mail Oct 07 16:38:13 pb__: i tried installing in /etc/cron/crontabs/ as it is done the same way by ntp(date), but it didn't work; /var/... worked, AFTER i did a reboot on the device, although pkg_postinst() had /etc/init.d/cron restart at its end ... Oct 07 16:38:42 so i was wondering wether /var/... and pkg_postinst() was the right approach Oct 07 16:40:18 ~seen likewise Oct 07 16:40:26 likewise was last seen on IRC in channel #oe, 8d 17h 53m 28s ago, saying: 'khem: hack lightly, and enjoy your evening'. Oct 07 17:22:02 hi, anyone can help me adding support for openclient in oe? I've added openrd-clienet.conf set MACHINE, but if I bitbake virtual/kernel I get multiple providers are available. When I select linux I get the wrong bb file (there is a linux-kirkwood bb file that I need, but setting PREFERRED_PROVIDER_virtual/kernel to linux-kirkwood does not give results Oct 07 17:22:29 linux-kirkwood is also not listed in the suggestions for virtual/kernel Oct 07 17:22:34 anyone a suggestion? Oct 07 18:29:30 nevermind found it Oct 07 19:19:49 03Steve Sakoman  07org.openembedded.dev * r9db2af0e53 10openembedded.git/recipes/atk/ (atk-native_1.27.90.bb atk-native_1.28.0.bb): atk-native: add 1.28.0 Oct 07 19:19:53 03Steve Sakoman  07org.openembedded.dev * r1ca3681257 10openembedded.git/ (conf/checksums.ini recipes/gcalctool/gcalctool_5.27.5.bb): gcalctool: add 5.28.0 Oct 07 19:19:54 03Koen Kooi  07org.openembedded.dev * rf9044c8c64 10openembedded.git/recipes/ (4 files in 2 dirs): enlightenment: some modules moved in the svn tree, account for that Oct 07 19:19:56 03Steve Sakoman  07org.openembedded.dev * ra0d8f66f72 10openembedded.git/ (conf/checksums.ini recipes/gnome/gnome-icon-theme_2.28.0.bb): gnome-icon-theme: add 2.28.0 Oct 07 19:20:03 03Steve Sakoman  07org.openembedded.dev * r0fe4a6841b 10openembedded.git/conf/distro/include/angstrom-2008-preferred-versions.inc: angstrom-2008-preferred-versions.inc: use wpa-supplicant 0.6.9 Oct 07 19:20:04 03Steve Sakoman  07org.openembedded.dev * re8e386a5f8 10openembedded.git/ (conf/checksums.ini recipes/atk/atk_1.27.90.bb): atk: add 1.28.0 Oct 07 19:27:50 mickey|dinner: ping Oct 07 19:27:59 anyone tried building python for a 64 bit target with OE? Oct 07 19:28:08 it's exploding on me here, maybe a multilib issue? Oct 07 19:28:41 kergoth: long back I attempted ppc64 but I dont think I remember building python Oct 07 19:28:50 hmm, okay, thanks anyway Oct 07 19:29:01 python builds are greek to me.. :) Oct 07 19:29:14 what are you seeing ? Oct 07 19:29:19 error I mean Oct 07 19:30:08 distutils.errors.DistutilsPlatformError: invalid Python installation: Oct 07 19:30:09 unable to open ... staging/x86_64-mv-linux/usr/lib64/lib/python2.6/config/Makefile Oct 07 19:30:18 /lib64/lib/ looks wrong Oct 07 19:32:00 yeah I think its appending lib to lindir Oct 07 19:32:37 also ahve failures in strace, mesa, xf86-video-xgi, etc Oct 07 19:32:44 obviously 64 bit targets need love :) Oct 07 19:32:56 yeah Oct 07 19:33:33 I think amd64 would be right one to begin with and then move on to ppc64 and mips64 Oct 07 19:34:07 and also sort out multilib as they all will need it Oct 07 19:34:28 yeah Oct 07 19:34:33 building glibc thrice (mips64) wont be fun Oct 07 19:34:48 looks like roman's done work on the multilib bits already on a branch, wonder what the status of it is Oct 07 19:35:02 I havent followed it too much Oct 07 19:35:56 * kergoth either Oct 07 19:36:15 ah he has a branch hmm Oct 07 19:36:24 I can look at the patches in free time Oct 07 19:36:59 * kergoth fires off x86_64 build and waits Oct 07 19:39:38 anyone an idea how i properly get my config file in a git kernel build; I added file://defconfig in the SRC_URI but the file ends up in the top level dir while S = "${WORKDIR}/git" Oct 07 19:40:30 "top level dir"? Oct 07 19:40:43 all urls in SRC_URI are "unpacked" to WORKDIR Oct 07 19:40:49 so thats exactly where the defconfig should go Oct 07 19:41:14 then a preconfigure or configure task or similar should copy it to ${S}/.config prior to the build Oct 07 19:41:23 * kergoth isn't sure what kernel.bbclass is doing nowadays, though Oct 07 19:41:35 03Khem Raj  07org.openembedded.dev * rd685ab28a4 10openembedded.git/classes/insane.bbclass: Oct 07 19:41:35 insane.bbclass: Use linux-gnu for TARGET_OS Oct 07 19:41:35 * Add linux-uclibcspe Oct 07 19:41:35 Signed-off-by: Søren Holm Oct 07 19:41:35 Signed-off-by: Khem Raj Oct 07 19:42:18 kergoth: thanks, thought the copy would be done automatically Oct 07 19:42:33 maybe check kernel.bbclass, see what it's up to? Oct 07 19:42:36 thats what i'd do Oct 07 19:45:31 well kernel.bbclass does not copy so i will do so myself Oct 07 19:49:15 found some code from another bb file to copy :-) Oct 07 19:51:03 eFfeM: which machine/distro/ and OE branch are u using Oct 07 19:51:50 khem, i'm trying to add a machine (openrd-client) this is a kirkwood variant, I'm using angstrom on the main head Oct 07 19:54:43 hm, copying to .config did not help, do_configre starts with Oct 07 19:54:44 echo "" > /home/frans/oe/tmp_angstrom/work/openrd-client-angstrom-linux-gnueabi/linux-kirkwood-2.6.31-r10/git/.config Oct 07 19:54:51 then apparently also does things with defconfig Oct 07 19:55:13 with some sed script Oct 07 19:55:23 do_configure_prepend Oct 07 19:55:28 is where it all happens Oct 07 19:56:12 i'll try again with a defconfig, expected that to work, it didn't but after reading do_configure I would assume it should Oct 07 20:00:54 morning all Oct 07 20:01:01 hi Oct 07 20:03:13 khem kergoth think i found it, re retrying :-) Oct 07 20:10:21 ~seen timtimred Oct 07 20:10:23 timtimred is currently on #oe (1h 27m 33s), last said: 'i only have access to the database resources'. Oct 07 20:10:34 mnwg Oct 07 20:10:37 oops hi Oct 07 20:10:55 timtimred: hi, tried to wake you on #openslug before Oct 07 20:11:11 plug! Oct 07 20:11:12 :) Oct 07 20:11:21 do you feel involved in kirkwood support in oe Oct 07 20:11:36 i havnt touched it much recently Oct 07 20:11:39 yes plug, oops, typo, have been involved in slug related projects ttoo in the past Oct 07 20:12:06 saw you and likewise worked on it, added openrd client support Oct 07 20:12:14 cool Oct 07 20:12:17 but niticed the git srcrev was no good any more Oct 07 20:12:29 ahhh... Oct 07 20:12:31 only the client parts that have been posted so no vga or sound yet Oct 07 20:12:54 and noticed that the head of the marvell git did not contain the openrd_base code Oct 07 20:13:05 cool did you update it so it pulls? or do you want the git repo uploaded? Oct 07 20:13:16 so basically ended up using the last base related commit as starting point :-( Oct 07 20:13:24 i think most of it is in vanilla kernel now Oct 07 20:13:26 ? Oct 07 20:13:34 for plug anyway Oct 07 20:13:49 yeah, but then i would expect it also to be in the marvell git Oct 07 20:13:51 ive not been too active recently due to work :( Oct 07 20:14:59 ah, relooking at it, guess i took a side branch, went for .31.1 Oct 07 20:15:05 but the head indeed has the base support Oct 07 20:15:40 if it doesnt pull we should update the git one to one that does Oct 07 20:16:16 well the head one does have the support Oct 07 20:16:38 2.6.31.1 did not have openrd base Oct 07 20:17:29 ahhh Oct 07 20:20:20 will try to move to 2.6.32-rc1 Oct 07 20:22:39 sure. Oct 07 20:32:53 eFfeM: what was it Oct 07 20:33:19 eFfeM: probably defconfig is being used correctly to generate .config it must be something else Oct 07 20:34:39 khem defconfig was used properly but there was an error in my defconfigg :-( Oct 07 20:40:18 I see thats what I thought too :) Oct 07 20:40:30 so all hunky dory now ? Oct 07 20:42:17 hope so, it builds Oct 07 20:42:24 keeping fingers crossed :-) Oct 07 20:43:13 03Andrea Adami  07org.openembedded.dev * rf5d88d0f92 10openembedded.git/recipes/kexec/files/no-getline-no-fscanf.patch: kexec-tools_2.0.1: rework/fix no-getline-no-fscanf.patch. Credits to Yuri Bushmelev. Oct 07 20:48:25 yay Oct 07 20:48:26 ooting the kernel. Oct 07 20:48:26 Linux version 2.6.32-rc2 (frans@linux-suse) (gcc version 4.3.3 (GCC) ) #1 PREEMPT Wed Oct 7 22:28:08 CEST 2009 Oct 07 20:49:46 not fully there yet, but made a dent Oct 07 20:49:50 kernel oops after Loading the saved-state of the serial devices... Oct 07 20:50:07 guess that's for tomorrow though Oct 07 20:52:11 good job ! have fun Oct 07 20:54:17 ty, calling it a day for now, nite everyone Oct 07 20:58:06 hmm... Oct 07 20:58:18 wpa_supplicant fails Oct 07 20:58:44 action_wpa.sh: no such file or directory Oct 07 21:09:07 Jay7: you might try reverting 6aea479c251050e0d13ae01b98c7a35da393b41f Oct 07 21:09:18 sakoman: I've found bug Oct 07 21:09:33 this file is in wpa-supplicant_0.6.7/ Oct 07 21:09:45 can you move or copy it into files/ Oct 07 21:10:03 I think the best fix is to revert the patch mentioned above Oct 07 21:11:15 that patch seems to have issues, so it is best removed Oct 07 21:12:00 i.e. remove all that *.sh files? Oct 07 21:12:33 (revert will do that) Oct 07 21:13:14 Jay7: yes Oct 07 21:13:23 sakoman: ok, reverting then Oct 07 21:13:44 there os a more recent patch that moves to 0.6.9, which sees to work well Oct 07 21:14:25 jo Oct 07 21:14:30 ~seen khem Oct 07 21:14:32 khem is currently on #oe (1d 6h 4s) #uclibc (1d 6h 4s). Has said a total of 65 messages. Is idling for 22m 21s, last said: 'good job ! have fun'. Oct 07 21:14:43 woglinde: hey Oct 07 21:14:46 he ant Oct 07 21:14:53 saw your new getline patch Oct 07 21:14:54 hehe Oct 07 21:15:04 is it smaller now? Oct 07 21:15:26 couplke of bits..the patch is not enlarging it Oct 07 21:15:34 I'll rebuild dietlibc now Oct 07 21:15:38 sakoman: I'll test buildability now Oct 07 21:15:54 but will not check 'workability' :) Oct 07 21:16:23 * XorA|gone is driven mad by wrt54g Oct 07 21:16:44 ant__: do we have nandlogical now in console-image for all zauruses? Oct 07 21:17:25 hmm..yes /usr/bin/nandlogical Oct 07 21:17:27 xora *g* Oct 07 21:17:34 hms Oct 07 21:17:44 I have to patch gcc again for armv4 with uclibc Oct 07 21:17:53 **** ***** **** **** **** **** Oct 07 21:17:56 damn floatsymbols Oct 07 21:18:01 ~lart XorA Oct 07 21:18:01 * ibot squeezes XorA till XorA turns blue like papa smurf Oct 07 21:18:10 ~curse xora Oct 07 21:18:11 May you be reincarnated as a Windows XP administrator, xora ! Oct 07 21:19:22 * XorA|gone is a grade A idiot Oct 07 21:21:01 xora haha Oct 07 21:21:05 what happend? Oct 07 21:21:22 ping ping khem Oct 07 21:21:30 jo florian Oct 07 21:21:59 re Oct 07 21:22:08 hms when will hit 2.6.31 debian unstable Oct 07 21:23:47 03Koen Kooi  07org.openembedded.dev * rcc3a658201 10openembedded.git/ (5 files in 3 dirs): gstreamer: update some bits Oct 07 21:23:47 03Koen Kooi  07org.openembedded.dev * r312d954c46 10openembedded.git/recipes/wpa-supplicant/wpa-supplicant-0.6.inc: wpa-supplicant: fix merge error Oct 07 21:24:26 hm some one knows if qemu integrator is armv4? Oct 07 21:24:49 woglinde: forgot to fix the default route on one box Oct 07 21:25:01 woglinde: so forwarding ports meant it didnt know where to send the reply Oct 07 21:25:04 xora *g* Oct 07 21:25:51 woglinde: whats up Oct 07 21:26:17 khem we need to twek gcc-4.4.1 to get uclibc compiled for armv4 Oct 07 21:26:35 oh the default is armv5te Oct 07 21:26:40 ok no problems Oct 07 21:26:40 its the old gcc/config/arm/t-linux stuf Oct 07 21:26:50 hehe.. Oct 07 21:26:53 oh Oct 07 21:27:00 woglinde: how about moving to eabi Oct 07 21:27:07 sakoman: seems koen fix this too.. Oct 07 21:27:12 now gcc 4.4 supports eabi for armv4 Oct 07 21:27:15 khem is gcc and binutils ready now? Oct 07 21:27:19 ah Oct 07 21:27:20 yes Oct 07 21:27:26 give it a shot :) Oct 07 21:27:29 yes Oct 07 21:27:34 defenitly Oct 07 21:27:35 and I will try to be *helpful* Oct 07 21:27:45 hm I am using micro now Oct 07 21:27:54 but I have to tweak something Oct 07 21:27:58 so oe let it got Oct 07 21:28:03 I use minimal Oct 07 21:28:20 micro is too pita with changes in layout Oct 07 21:28:35 not all apps would compile Oct 07 21:28:39 sakoman: reverted version builds ok Oct 07 21:28:45 hm I think I only need to tweak machine conf Oct 07 21:28:47 but on positive we find some real bugs Oct 07 21:28:55 Jay7: koen pulls some of my changes into oe.dev -- I think this problem cropped up when he missed one of my commits Oct 07 21:29:28 Jay7: left things in a broken state Oct 07 21:29:41 khem hm whats the kernel status about armv4 eabi? Oct 07 21:29:55 do I need newer libc-header? Oct 07 21:29:58 XorA|gone: so when do we change angstrom to use gcc 4.4? Oct 07 21:30:11 sakoman: please fix it appropriately :) Oct 07 21:30:12 woglinde: I think existing headers would do Oct 07 21:30:19 thesing hehe you want testing for collie? Oct 07 21:30:23 khem 2.6.23? Oct 07 21:30:28 should be ok Oct 07 21:30:37 args Oct 07 21:30:37 I usually use closer to real kernel Oct 07 21:30:46 why I remove oetmp Oct 07 21:30:56 If somebody gives me a patch I can test. Oct 07 21:30:57 Jay7: things should be fine after koen's latest commit Oct 07 21:31:05 its good excuse to have a day break and heat the room up :) Oct 07 21:31:06 ok Oct 07 21:31:10 thesing will test here for micro and simpad Oct 07 21:31:27 aeh micro-uclibc Oct 07 21:31:28 woglinde: which armv4 machine should I use Oct 07 21:31:34 simpad Oct 07 21:31:41 hm or collie Oct 07 21:31:48 I will try compile only test Oct 07 21:31:56 oh collie ok Oct 07 21:32:08 kernel should be newer Oct 07 21:32:14 simpad is at 2.6.27 Oct 07 21:32:33 whatever it pulls in by default should be ok isnt it Oct 07 21:32:52 khem do you know somethin about the arm-machines qemu supports? Oct 07 21:33:11 it supports upto armv6 Oct 07 21:33:17 I am not sure about armv7 Oct 07 21:33:26 collie is currently as .26 Oct 07 21:33:33 thesing oh Oct 07 21:33:43 but I could push it to .32-rc3 tomorrow Oct 07 21:33:44 woglinde: one of platform from arm is emulated well Oct 07 21:34:00 khem what about integrator? Oct 07 21:34:00 it was integrator or somethin else dont remember Oct 07 21:34:03 isnt it armv4? Oct 07 21:34:34 oh integrator comes with many types of arms Oct 07 21:34:51 currently in oe qemuarm is v5te Oct 07 21:34:53 or so Oct 07 21:36:59 runs with -M versatilepb Oct 07 21:38:29 hm what was akita? Oct 07 21:39:13 pxa 270 Oct 07 21:39:15 okay Oct 07 21:55:19 hrms Oct 07 21:55:21 sane-toolchains Oct 07 21:59:16 lol Oct 07 21:59:23 I had an error in strongarm-tune Oct 07 21:59:27 now its showing Oct 07 22:21:46 Hi Oct 07 22:23:20 woglinde: I'm having funny results patching the Makefile Oct 07 22:23:25 of dietlibc Oct 07 22:23:54 oh..my bad... Oct 07 22:24:01 found Oct 07 22:24:43 khem still there? Oct 07 22:28:50 Just a git-question. How can I see where a specific local branch is branched out from ? Oct 07 22:32:24 man git-branch Oct 07 22:32:25 -v, --verbose Show sha1 and commit subject line for each head, along with relationship to upstream branch (if any). If given twice, print the name of the upstream branch, as well. Oct 07 22:36:12 woglinde: aaaah "-v -v" missed that one. Thanks. Oct 07 22:38:25 cool Oct 07 22:38:33 armv4-eabi toolchain compiled Oct 07 22:38:42 thats a major step Oct 07 22:48:42 woglinde: for dietlibc is not so trivial fix :/ Oct 07 23:00:05 ant hm? Oct 07 23:07:48 arm-angstrom-linux-gnueabi-strip -R .comment -R .note bin-i386/diet Oct 07 23:08:02 why does it want to strip an x86 binary :P Oct 07 23:08:35 that's my testing... Oct 07 23:09:36 woglinde: if I wreck CROSS insane things happen Oct 07 23:10:04 ...the kiss of the x86 Oct 07 23:10:11 lethal for arm :) Oct 07 23:16:38 woglinde: vanilla dietlibc_0.32 failing Oct 07 23:16:40 http://tinderbox.openembedded.org/builds/32833/ Oct 07 23:19:29 gn Oct 08 00:45:15 03Otavio Salvador  07org.openembedded.dev * r2b7891e0ff 10openembedded.git/recipes/squashfs-tools/ (5 files): Oct 08 00:45:15 squashfs-tools: refactor 3.3 and 4.0 recipes and use INC_PR Oct 08 00:45:15 Signed-off-by: Otavio Salvador Oct 08 00:45:16 03Otavio Salvador  07org.openembedded.dev * rdf9d3c175d 10openembedded.git/recipes/squashfs-tools/ (16 files in 2 dirs): Oct 08 00:45:19 squashfs-tools: drop 2.0, 2.2r2 and 3.1r2 versions Oct 08 00:45:21 Signed-off-by: Otavio Salvador Oct 08 00:45:23 03Otavio Salvador  07org.openembedded.dev * r2df87bacf1 10openembedded.git/conf/machine/qemux86.conf: Oct 08 00:45:26 machine/qemux86: x86 is MACHINE_FEATURES already set in i686-generic machine Oct 08 00:45:28 Signed-off-by: Otavio Salvador Oct 08 00:45:30 03Otavio Salvador  07org.openembedded.dev * r19bfd19e4c 10openembedded.git/recipes/wpa-supplicant/ (5 files): Oct 08 00:45:33 wpa-supplicant: use INC_PR Oct 08 00:45:35 Signed-off-by: Otavio Salvador Oct 08 00:50:57 khem just as info kernel runs Oct 08 01:04:16 jo kergoth Oct 08 01:04:22 hey Oct 08 01:04:27 armv4-eabi really works Oct 08 01:05:18 kernel at least Oct 08 01:05:26 image test will follow tomorrow Oct 08 01:05:49 good nite Oct 08 01:26:46 so i recently became aware of the 8 character password limit when i mistyped the last few of my pass and found myself with a '$' Oct 08 01:27:04 i checked tinylogin and wasn't able to find a limit there....is this uclibc? busybox? Oct 08 01:29:43 i see 'CONFIG_PASSWORD_MINLEN=6' in busybox defconfig Oct 08 01:32:42 m4t: it's most likely libc Oct 08 01:33:02 if the app uses getpassphrase() instead of providing its own function for the purpose Oct 08 01:34:39 k Oct 08 01:34:45 i'll keep looking Oct 08 01:41:32 tinylogin uses getpass() Oct 08 01:44:10 it gives char pass[200];, but specifies (through output) a max length of 8 **** ENDING LOGGING AT Thu Oct 08 02:59:58 2009