**** BEGIN LOGGING AT Thu Jun 29 02:59:57 2006 Jun 29 06:15:45 anyone awake? Jun 29 06:18:41 goxboxlive: Yeah... Jun 29 06:27:29 v8jlene: do you know how i can patch qte so my keyboard on the htc universal will work in Opie? Jun 29 06:28:57 goxboxlive: Heh, no idea... my work is all for a router distro - so no UI stuff at all. So no idea about qte or opie or keyboards. Jun 29 06:28:58 Someone mension kernel-keyboardmap.patch for me, but when i try to aplly it in qte-common-2.3.10.inc the compilation failes about missing kernelkeyboard.cpp Jun 29 06:29:07 ok Jun 29 06:41:06 03lenehan 07org.oe.dev * rea9b2aa1... 10/packages/linux/ (3 files in 2 dirs): (log message trimmed) Jun 29 06:41:06 linux-titan-sh4 2.6.16: Updates: Jun 29 06:41:06 - Add support for getgeo to the ssfdc driver so lilo will work with onboard Jun 29 06:41:06 flash with recent kernels. Jun 29 06:41:06 - Override the KERNEL_RELEASE to be KERNEL_VERSION - since sh kernels add Jun 29 06:41:06 -sh to the end and this isn't being picked up. Jun 29 06:41:08 - Change the kernel imagetype to zImage since lilo boots that fine and the Jun 29 07:13:34 celf@thinkpanda ~/newbase/org.openembedded.dev $ monotone update Jun 29 07:13:35 monotone: already up to date at c559a14f1e4e3c5114c6501d8124accafc74740f Jun 29 07:13:51 hi, OE ppls, is that the latest OE.db i can get? Jun 29 07:14:06 seems more than 5 days didn't updated? Jun 29 07:16:51 minipanda: did you do a monotone -pull ? Jun 29 07:18:33 hvontres: sorry, i've done monotone -pull, but seems enter a wrong org.openembedded.dev/ directory :-( Jun 29 07:18:55 hvontres: ok, i will try later and redo a checkout Jun 29 07:19:27 BTW: where shall i report a bug for Bitbake? Jun 29 07:19:35 good morning all Jun 29 07:19:37 bugs.openembedded.org Jun 29 07:19:45 minipanda: make sure it is a bitbake bug though Jun 29 07:19:45 morning Jun 29 07:20:03 koen: 29 00:01 < rwhitby> I see koen mailed me asking to remove the tinderbox.nslu2-linux.org alias. Where has tinderbox moved to, or has the functionality been withdrawn? Jun 29 07:20:20 i just tried to use 'pastelog' but then bb forced to quit Jun 29 07:20:34 rwhitby: ewi546 musn't have any alias!!! :) Jun 29 07:20:48 rwhitby: if we do, ewi546 will be terminated Jun 29 07:20:55 whatever :} Jun 29 07:21:41 minipanda: try ~/newbase $ monotone --db=/stuff/OE.db pull monotone.openembedded.org org.openembedded.dev Jun 29 07:22:05 asuming OE.db is your db file.. Jun 29 07:22:14 hvontres: thanks, 'rev written' now ;-) Jun 29 07:22:21 np Jun 29 07:24:18 ooops :Query failure: Table 'pastebin' was not locked with LOCK TABLES Jun 29 07:24:18 insert into pastebin (poster, domain, posted, format, code, parent_pid, expires,expiry_flag) values ('minipanda', 'oe', now(), 'text', '\[celf@xsara build\]$ bitbake -i\r\nBitBake Shell Version 0\.5\.3\.1 \(C\) 2005 Michael \'Mickey\' Lauer \\r\nType \'help\' for more information, press CTRL-D to exit\.\r\nBB\>\> build bootstrap-image\r\nSHELL\: This command needs to parse bbfiles\.\.\.\r\nNOTE\: Using cache in Jun 29 07:24:33 seems something running on oe.pastebin.com was crashed Jun 29 07:25:07 so when i tried 'pastelog' in my 'bitbake -i' shell, will crashed too Jun 29 07:29:49 night all Jun 29 07:29:58 bye... Jun 29 07:30:20 o.. pavelm was Jun 29 07:30:31 good luck keeping ewi546 up...after all the trouble we had last week ;) Jun 29 07:31:42 morning Jun 29 07:32:45 hi XorA Jun 29 07:36:24 morning Jun 29 07:37:08 hi, can anybody help me take a look at http://pastebot.nd.edu/320 Jun 29 07:37:20 i really stucked with this issue for two days and really have no clue :-( Jun 29 07:38:09 seems the linux-omap1-2.6.12-rc2 pacakge build alright, Jun 29 07:38:35 but failed at do_deploy Jun 29 07:41:23 minipanda: you need to look into the log file Jun 29 07:42:55 zecke: well, the log file only contain one line "/home/celf/newbase/individual/celf/101/OE-20060622_20060627/omap5912osk_familiar-unstable_bootstrap-image/build/tmp/work/omap5912osk-linux/linux-omap1-2.6.12-rc2-r3/temp/run.do_deploy.15919: line 685: do_deploy: command not found Jun 29 07:42:56 " Jun 29 07:43:12 koen: ping Jun 29 07:43:14 that is bad Jun 29 07:43:27 zecke: and the line 685 of run.do_deploy is just "do_deploy" Jun 29 07:44:16 hrw|work: pong Jun 29 07:44:44 minipanda: that is bad :) Jun 29 07:44:49 hey koen Jun 29 07:44:51 zecke: hi, i've read some other .bb file for kernel in /pacakges/linux/* Jun 29 07:45:12 koen: http://test.hentges.net/ is mod_rewrite round-robin test for resolving feeds.oz.org - can you look? Jun 29 07:45:13 zecke: and seems they also call the same sub-function do_deploy as well Jun 29 07:45:45 minipanda: e.g. the linux-openzaurus.inc has a custom do_deploy Jun 29 07:45:52 minipanda: do you have one for your OMAP kernel? Jun 29 07:45:54 koen: I want to know does ewi can or cannot be used in such way Jun 29 07:45:54 hrw|work: seems to work Jun 29 07:46:00 hrw|work: sure Jun 29 07:46:07 koen: thx Jun 29 07:46:09 hrw|work: anything that doesn't involve dns is fine Jun 29 07:46:17 hrw|work: thats pretty sweet, is the code available for interest? Jun 29 07:46:17 ok Jun 29 07:46:24 XorA: ask CoreDump Jun 29 07:46:33 RP: ping, I got a poky and OE 'fix' Jun 29 07:46:45 CoreDump|home: is the code available for your sweet mod rewrite haxoring? Jun 29 07:46:48 RP: RRECOMMENDS_${PN} = "" in autoconf-native.bb Jun 29 07:46:58 XorA: sure thing, hold on Jun 29 07:47:06 zecke: sorry, i don't know exactly your meaning? Jun 29 07:47:23 minipanda: well, I have no clue about this linux-omap file Jun 29 07:47:42 minipanda: all I know is do_deploy is shell, an empty shell method should at least contain a ':' Jun 29 07:47:56 minipanda: I don't know why you would want to leave do_deploy blank Jun 29 07:48:11 zecke: well, not me leave it blank Jun 29 07:48:32 zecke: this same .bb file can be build in OLD OE.db Jun 29 07:48:45 minipanda: well, do you use a omap5912osk? Jun 29 07:48:53 zecke: yes, i am Jun 29 07:49:03 minipanda: if it is the same bb file, where it comes from does not matter Jun 29 07:49:21 minipanda: what about looking into this bb file once? Jun 29 07:49:29 zecke: well, i just curious, why the same .bb file can be build few months ago, but failed now Jun 29 07:49:38 minipanda: In my outdated copy of OE I have do_deploy_omap5912osk( Jun 29 07:49:48 XorA: http://rafb.net/paste/results/f9nP1f83.html Jun 29 07:50:01 minipanda: this means if omap5912osk is in your OVERRIDES (e.g. from MACHINES/DISTRO) it will use this method Jun 29 07:50:02 zecke: ah, could you post that .bb file somewhere on pastebot.org? Jun 29 07:50:07 minipanda: otherwise it will crash Jun 29 07:50:22 minipanda: use your OE.db to extract that version ;) Jun 29 07:50:29 minipanda: we have SCM for a reason Jun 29 07:50:43 CoreDump|home: cool, that rewrite engine is a lot smarter than I ever thought Jun 29 07:51:23 zecke: ok, i just use -checkout each time, and i don't know how to get the exactly verison of that .bb file from OE.db Jun 29 07:51:30 zecke: and what's the 'SCM' meaning? Jun 29 07:51:40 XorA: yeah, the module is pretty nifty and the documentation is very good Jun 29 07:51:43 minipanda: dc2bd7bf38445d4f691591f85e37ab1524251d28 this is my manifest Jun 29 07:51:54 minipanda: 12b4da6c57d22b540df4a906937d3c7d4fa403fc this is the old revision Jun 29 07:52:11 * XorA keeps that snipper for reference Jun 29 07:52:15 minipanda: ah 12b4da6c57d22b540df4a906937d3c7d4fa403fc <- the current revision Jun 29 07:54:14 zecke: so could you let me know how can i -checkout "dc2bd7bf38445d4f691591f85e37ab1524251d28" from my local OE.db? Jun 29 07:54:28 minipanda: e.g. you could do monotone cat -r 12b4da6c57d22b540df4a906937d3c7d4fa403fc packages/linux/linux-omap1_2.6.12-rc2.bb Jun 29 07:54:40 minipanda: sorry, my time is up, need to start working now Jun 29 07:54:42 later guys Jun 29 07:58:23 XorA: apache.org has very good docs with examples for mod_rewrite Jun 29 08:44:37 hey pH5 Jun 29 08:44:54 hey koen Jun 29 09:07:36 hi polynoumous Jun 29 09:11:48 ARGH!! Jun 29 09:11:55 KILLL!! KILL!! ipkg Jun 29 09:12:05 hrw|work: oops Jun 29 09:12:18 koen: got hit by efl++ probably Jun 29 09:12:35 hrw what happens? Jun 29 09:12:45 mv *.ipk morgue/ Jun 29 09:13:08 good that it was not 'rm *.ipk' Jun 29 09:22:03 hrw|work: it looks like it's the new tar Jun 29 09:22:17 hrw|work: apt-get dist-upgrade seems to have broken a batch of OE assumptions Jun 29 09:25:29 koen? Jun 29 09:25:51 woglinde: new version tools, new version of OE breakages Jun 29 09:26:02 s/tools/of tools/ Jun 29 09:26:36 hrw|work: + tends to break ipkg-make-index in many cases, +cvsXXXXX always ranks lower than any version number in make-index, but works as expected in ipkg Jun 29 09:30:19 maybe Jun 29 09:31:09 ipkg(-utils) looks like something which was written for much smaller amount of data Jun 29 09:39:28 XorA: that's probably just a bug in the python version compare routine. it should be easy enough to fix. Jun 29 10:26:02 tar: *control: Not found in archivetar: Error exit delayed from previous errorstar: Pattern matching characters used in file names. Please,tar: use --wildcards to enable pattern matching, or --no-wildcards totar: suppress this warning. Jun 29 10:26:09 hrw|work: you get the same error? Jun 29 10:27:08 koen: I did not built images yet Jun 29 10:27:29 running ipkg-make-index is the problem Jun 29 10:37:04 hrw|work: try http://ewi546.ewi.utwente.nl/tmp/ipkg-utils-fix.patch Jun 29 10:41:05 will do but not now Jun 29 11:00:35 03hrw 07org.oe.oz354x * r08e61c47... 10/packages/initscripts/initscripts-openslug_1.0.bb: dropped initscripts for openslug Jun 29 11:00:39 03hrw 07org.oe.oz354x * r4cd47ee4... 10/packages/debianutils/ (3 files): debianutils: added 2.16.1, dropped older not fetchable ones Jun 29 11:00:43 03hrw 07org.oe.dev * r6d7a9d2f... 10/packages/debianutils/debianutils_2.16.1.bb: debianutils: added 2.16.1 Jun 29 11:29:20 what is the difference between do_install and do_stage? Jun 29 11:30:18 do_install is for making packages and do_stage just copies stuff into staging/ Jun 29 11:30:41 speaking about staging... Jun 29 11:30:47 koen: how goes your SoC? Jun 29 11:31:08 glibc suxx0rs Jun 29 11:31:33 I see Jun 29 11:31:43 I've been trying to get glibc to stage properly without needing 'make install' a zillion times Jun 29 11:35:59 so if I want to manually compile a program I should go to its work directory, run do_compile followed by do_stage, then go to the work directory of my main meta file and run do_rootfs and I should have a rootfs that has the newly compiled program installed, correct? Jun 29 11:36:25 when I refer to do_ I mean the scripts in the temp directories :-) Jun 29 11:37:22 bitbake -b file.bb -c compile -f ? Jun 29 11:38:03 koen, will that put the program into the the tmp/rootfs area? Jun 29 11:38:16 no, that will compile it Jun 29 11:38:21 that is what I thought. Jun 29 11:41:13 I want to get the new compiled program into the rootfs dir so I can manually run mkfs.jffs2. I think the only way to do this is run the scripts in the temp directories manually. Jun 29 11:42:56 it is a real nice feature of OE having the scripts in the temp directories :-) Jun 29 11:47:59 Gerrath: add package to IPKG_INSTALL of image Jun 29 11:48:08 Build statistics: Attempted builds: 157 Failed builds: 152 Dependencies not satisfied: 73 Jun 29 11:48:13 world ended Jun 29 11:49:37 hrw|work, that will just generate an ipkg to download to the target correct? Jun 29 11:49:44 oh noooes the worl ended Jun 29 11:50:04 Gerrath: you use bitbake + OE to build packages and then use packages to generate images Jun 29 11:52:40 I thought the only use for ipkg was to provide a (.deb like) file that you would then download to the target and use ipkg to install it. Jun 29 11:54:13 Gerrath: that too Jun 29 11:55:59 hrw|work, I guess I will have to read up on ipkg as I have no idea how to build an image with it :-) Jun 29 11:56:20 Gerrath: packages/meta/*-image.bb Jun 29 11:58:36 hrw|work, ok I think I understand what you are saying. I already wrote a meta file for my project so your saying to compile the packaged I modifed using bitbake then rebuild my project using my meta file (through bitbake). Jun 29 11:59:03 sort of that Jun 29 12:11:37 03koen 07org.oe.dev * r744da369... 10/packages/suspend-desktop/suspend-desktop_1.0.bb: Jun 29 12:11:37 suspend-desktop: fixups Jun 29 12:11:37 * fix do_install to not break (see oe bug #1023, #1024 and #1025) Jun 29 12:11:37 * set package_arch to all Jun 29 12:16:20 hi Cliff Jun 29 12:17:08 hello Marcin Jun 29 12:19:08 03hrw 07org.oe.oz354x * r22638cb9... 10/packages/fbreader/ (3 files in 2 dirs): fbreader: added 0.7.3d Jun 29 12:19:12 03hrw 07org.oe.oz354x * rce7d852f... 10/conf/licenses.conf: licenses.conf: added another variants of GPL v2 Jun 29 12:19:17 03hrw 07org.oe.dev * r44c112a1... 10/packages/fbreader/fbreader_0.7.3d.bb: fbreader 0.7.3d: fixed SRC_URI Jun 29 12:22:30 good morning cbrake. Jun 29 12:23:03 150 failed packages Jun 29 12:23:17 Gerrath: morning. How are things? Jun 29 12:23:28 Could anyone help me how i patch the qte-mt or qte tp qork with a usb keyboard? ( On HTC Universal it's a USB keyboard) . Jun 29 12:24:39 I have heard from lpoetter that QTE can be patched with kernel-kaymap.patch, but i have tried in the wte-common.inc file, but it failes when building Jun 29 12:25:00 morning all Jun 29 12:25:01 hi Dirk Jun 29 12:25:07 hey Marcin Jun 29 12:25:10 goxboxlive: I've got my zaurus working with a usb keyboard.. loadkeys w/ a standard qwerty keymap did the job.. Jun 29 12:25:12 goxboxlive: qte in OE contain kernel keymap patch Jun 29 12:26:01 hrw: I know that, i have tried to define the keymap patch, but it failes buidliding. It s probaly because i am doing some thing wrong Jun 29 12:29:56 how to define alias with params in bash? Jun 29 12:30:11 alias bbnew='bitbake -cclean $1;bitbake $1 Jun 29 12:30:11 I am adding kernel-keymap.patch in to the list in qte-common.inc file, but it complaines about a missing file. I'll buidl again and i let you know what it says. Jun 29 12:30:26 alias bbnew='bitbake -cclean $1;bitbake $1' got autoreplaced $1 ;( Jun 29 12:33:43 I have to use this in the /etc/init.d/opie files to get it to work ( export QWS_KEYBOARD=USB:/dev//input/event1) . But that is with a precomiled qte-mt lib. I just want it to work right out of the OE. Jun 29 12:34:47 goxboxlive: if you build qte with OE and set QWS_KEYBOARD=USB:/dev//input/event1 then it works? Jun 29 12:35:26 Ok here is the failings message i get when i am adding the patch in the qte-common file: http://pastebin.ca/74777 Jun 29 12:36:23 hrw|work: Only when i am aplying the prebuild qte lib. There was a guy who made it, but he has left so i cant get in touch with him. Jun 29 12:36:31 ;( Jun 29 12:37:28 But lpotter said to me that Opie could get the keyboard for the kernel. It was just to aplply the kernell-keymap.patch. But i ahve problem to build it as you can see. Jun 29 12:37:43 hi all Jun 29 12:37:51 also to write correct english :-) Jun 29 12:37:51 hey mr_nice_slacker Jun 29 12:38:17 goxboxlive: I'm using .oz354x mostly so do not follow .dev changes.. Jun 29 12:39:10 What is the flile kernelkeyboard.cpp? Where can i get that file? It seems like it's missing. Is it something from my cvs kernel? Jun 29 12:39:52 no - its from qte patches Jun 29 12:40:11 hrw|work: use a function?: function bbnew { bitbake -cclean $1; bitbake $1; } Jun 29 12:40:14 This is so strangem because with GPE the keyboard work right out of OE. Jun 29 12:40:47 goxboxlive: its not strange Jun 29 12:40:56 ok Jun 29 12:41:01 qte is very limited when it comes to device support out-of-box Jun 29 12:41:12 v8jlene: will check Jun 29 12:42:12 i c, but is HTC universal the only device that have a integrated USB keyboard. If i just knew another device with the same setup i could just tried to do the same. Jun 29 12:47:24 2.3.10 or 2.3.12? Jun 29 12:47:37 2.3.10 Jun 29 12:48:02 universal do not apply kernelkeymap currently.. Jun 29 12:48:05 hi Skyhusker Jun 29 12:48:48 hey hrw|work Jun 29 12:49:53 hrw|work: what do you think about using swap on microdrive models? Jun 29 12:49:56 hey Skyhusker Jun 29 12:50:05 ok, so i am using a wrong patch. Which patch should i then use? There have to be a way to solve this. We now have allmost fully support for the HTC Universal in the dev branch, but we should also had a working keyboard :-) . Jun 29 12:50:21 koen: many people do it already Jun 29 12:50:35 hrw|work: by default I mean Jun 29 12:50:41 zecke my hero, how is the mood today? Jun 29 12:50:54 hey koen Jun 29 12:50:56 hrw|work: iirc ubuntu has a udev rule that detects and uses swap partitions automagically Jun 29 12:51:02 hey zecke Jun 29 12:51:10 koen: o.. thats interesting Jun 29 12:51:52 I say iirc, since I hosed my ubuntu install today Jun 29 12:51:58 goxboxlive: busy :( Jun 29 12:52:05 note: do not use a 20MB /boot partition on ubuntu Jun 29 12:52:11 goxboxlive: and the sister of my grandmother is about to pass away Jun 29 12:52:14 goxboxlive: so bad mood Jun 29 12:52:25 lpotter_: okay. While I have asked to GPL Qtopia Jun 29 12:52:36 lpotter_: you have made three Qtopia releases to your customers Jun 29 12:52:36 zecke: ok, sorry to hear that. Jun 29 12:52:53 lpotter_: so, I think it is time to push Gtk+ more harder Jun 29 12:59:33 zecke: who are you, and where is the real zecke? Jun 29 13:00:30 btw, are there any plans to move to monotone >=0.26? 0.25 doesn't work for me since i moved to gcc 4.1 Jun 29 13:00:49 Skyhusker: see openembedded.org Jun 29 13:00:52 pb_: hehe, well. If TT says Community piss of we don't want you Jun 29 13:00:57 koen: thanks :) Jun 29 13:01:03 pb_: I will try to get less addicted to Qt Jun 29 13:01:26 zecke: well come to the light side Jun 29 13:01:28 zecke: try efl++ ;) Jun 29 13:03:45 koen: I want something cross platform :) Jun 29 13:04:01 koen: and I want it before HURD is finished Jun 29 13:04:14 :) Jun 29 13:04:50 koen: I didn't like the GSoC interview Jun 29 13:05:11 zecke: you have plenty of time, so ;) Jun 29 13:05:15 there's a stream of complaints about that Jun 29 13:06:27 zecke: which interview was that? Jun 29 13:08:29 pb_: sorry, survey Jun 29 13:08:55 pb_: it has that fishy job interview smell Jun 29 13:09:09 pb_: could he or his student be right for google... Jun 29 13:10:04 zecke: ah, I see Jun 29 13:10:07 that isn't too good Jun 29 13:11:19 "would you like to work in a small cubicle at google?" Jun 29 13:11:27 You didn't answer the required question... Jun 29 13:11:37 comes up if you don't answer... Jun 29 13:12:13 heh Jun 29 13:12:36 ah and it only has the "Yes" button ;) Jun 29 13:12:58 :-} Jun 29 13:14:18 sounds kinda like microsoft recruiting tactics Jun 29 13:23:33 aargh Jun 29 13:23:36 now i'm pissed off Jun 29 13:23:59 someone broke my code, fucked my identation, didn't keep my coding style and committed Jun 29 13:24:21 giel: details? Jun 29 13:24:27 no Jun 29 13:24:30 not interesting Jun 29 13:25:54 ~lart enca for -I/usr/include Jun 29 13:25:54 * ibot slaps a compatible dib on enca's head for -I/usr/include Jun 29 13:34:06 I noticed that all the files in my dev/ directory under rootfs are just files, shouldn't they actually be devices? Jun 29 13:38:14 03pH5 07org.oe.dev * rc7ec779d... 10/packages/misc-binary-only/acx-firmware_1.0.bb: acx-firmware: recommend kernel modules, don't depend on them Jun 29 13:39:00 Gerrath: which image? Jun 29 13:40:10 who is john bloom? Jun 29 13:41:02 XorA: johnX iirc Jun 29 13:41:03 * chouimat is back. Jun 29 13:41:13 koen: cheers Jun 29 13:41:15 ~seen johnX Jun 29 13:41:26 johnx was last seen on IRC in channel #openzaurus, 10d 4h 34m 46s ago, saying: 'I guess copy what you can, but it looks like the mmc/sd driver itself has run into trouble, and that's what caused this in the first place'. Jun 29 13:41:36 morning Jun 29 13:41:41 horrible hack in enca.bb mixes my arm headers with x86_64 with bad effect Jun 29 13:43:58 do13_, generic. Jun 29 13:53:47 Gerrath: It looks like generic doesn't generate the devices. Have a look at image_ipkg.bbclass and tro to inherit this Jun 29 13:53:51 hmmm Jun 29 13:54:04 what's the magic incantation to put w100 fb in 320x240? Jun 29 13:54:18 mornin Jun 29 13:54:27 koen: fbset? Jun 29 13:54:36 that's what I thought Jun 29 13:54:44 but busybox fbset is nonworking Jun 29 13:56:22 koen: it works, you need to remember to pass the arg to say actually change, not just give me timins Jun 29 13:56:31 koen: I cant rember which arg that is offhand Jun 29 13:57:04 Xfbdev -screen 320x240@90 :) Jun 29 13:57:07 fbset -n -xres 240 -yres 320 Jun 29 13:58:09 do13, Thanks. Jun 29 13:59:05 Gerrath: does it works now? Jun 29 14:04:09 03hrw 07org.oe.oz354x * radcc491a... 10/packages/supertux/supertux_0.1.2.bb: supertux: updated to version from .dev Jun 29 14:04:14 03hrw 07org.oe.oz354x * r8b23d3f9... 10/packages/silo/ (cross.patch silo_1.4.8.bb): dropped silo Jun 29 14:04:18 03hrw 07org.oe.oz354x * rba8df928... 10/packages/freeze/ (files/freeze files/unfreeze freeze.bb unfreeze.bb): dropped freeze/unfreeze Jun 29 14:04:22 03hrw 07org.oe.oz354x * r48bc4c5f... 10/conf/distro/openzaurus-3.5.4.1.conf: openzaurus 3.5.4.1: add oz-bootsplash into images Jun 29 14:08:03 zecke:hi, if you have some time could you tell me where to download your 2.6.0 and 2.6.10 kernels? if not just write me a botmail if you have some time for it. Jun 29 14:08:26 after building the bootstrap image, the root fs does not have an HDA device in /dev Jun 29 14:08:48 I added support for IDE in the defconfig file.... does anything else need to be done? Jun 29 14:10:17 do13_, well not really but that is because I have a slew of other issues to iron out first :-) Jun 29 14:13:42 zecke, any update on the bitbake issue? Jun 29 14:19:52 have a bitbake problem, any guru around. Just got some logical questions Jun 29 14:21:04 don't ask to ask, ask Jun 29 14:21:11 ah, roger Jun 29 14:21:29 Trying to get the kernel to be built by gcc 3.4.3 Jun 29 14:22:00 Ive set a preferred providers virtual/{TARGET_PREFIX}gcc-3.4.3:the-gcc Jun 29 14:22:16 But, I must also put a provides the-gcc, in the gcc.bb file right? Jun 29 14:22:30 Good evening Jun 29 14:22:58 I use CCsuffix="-3.4.3" in kernel.bb to get it to use that gcc Jun 29 14:29:58 nice device: http://www.titanomni.com/player.html Jun 29 14:30:18 koen: I had to the ep93xx-kernel_2.6.16.bb file so that it would run depmod... can that be added to the base bitbake file? also although it build the modules and the dependency list its not included in the rootfs crteated by bootstrap-image Jun 29 14:31:05 jasuris: that should be handled by kernel.bbclass Jun 29 14:35:18 mr_nice_slacker: wow ... coool device Jun 29 14:38:23 for some reason the created function pkg_postinst_modules is never called in do_install, so i had to modify the bb to include a call to it... Jun 29 14:39:31 zecke think I got it now, was several mistakes on my part Jun 29 14:40:36 Gerrath: That's actual normal - as a normal use you can't create device nodes, but fakeroot keeps track of them. So any images created from it such as a .tar.gz or the like will have the proper device nodes in it. Jun 29 14:41:13 Gerrath: user not use... Jun 29 14:56:04 Getting lots of nbd related booting errors, trying to mount stuff. How can I make that go away? Jun 29 15:10:54 03eFfeM 07org.oe.dev * raa18ee8e... 10/ (3 files in 2 dirs): pvrusb2-mci: added missing hotplug files Jun 29 15:15:31 v8jlene, thanks. Jun 29 15:28:07 which package provides libsocket? libgcrypt fails to build because it is missing -lsocket Jun 29 15:37:42 gm Jun 29 15:38:04 Im making changes to my defconfig file, but they are not making it to the .config file Jun 29 15:38:09 anyone know why? Jun 29 15:38:16 hmm Jun 29 15:38:28 are you completely rebuliding the package every time? Jun 29 15:38:32 do you clean after changes Jun 29 15:39:15 i do a rebuild..... I think that cleans everything.... Jun 29 15:39:21 um Jun 29 15:39:24 define rebuild? Jun 29 15:39:43 i do bitbake -i Jun 29 15:39:50 rm -rf tmp;bitbake bootstrap image Jun 29 15:39:51 and then rebuild ep93xx-kernel Jun 29 15:40:06 trying clean see if that works Jun 29 15:40:38 I've never used interactive mode Jun 29 15:41:01 Look in the log, make sure its copying the defconfig from the right locadtion Jun 29 15:41:18 I suspect clean will fix problem Jun 29 15:41:35 I second that suspect Jun 29 15:41:45 now we shoudl start betting Jun 29 15:41:51 hehe Jun 29 15:41:58 ok thanks Jun 29 15:42:24 that fix it? Jun 29 15:42:45 i just started the build.... let you know in a sec Jun 29 15:42:53 Well, a rebuild is a clean, so that really doesn't help Jun 29 15:43:23 hmmm Jun 29 15:43:35 that did not fix the problem Jun 29 15:43:38 CosmicPenguin, have you used -i much? Jun 29 15:43:41 ok Jun 29 15:43:49 check the logs Jun 29 15:43:52 You have to look at the logs - figure out which defconfig is being copied Jun 29 15:43:55 I was unawre of -i Jun 29 15:44:09 Crofton: yeah - I use it pretty reguarly - rebuild is very useful, especially when I am in heavy debug Jun 29 15:44:20 yeah Jun 29 15:44:23 I can see that Jun 29 15:44:27 And since parse is so fast now, its painless Jun 29 16:24:33 I've been struggling with a problem during user space booting for about a week. Here is my boot output: http://www.pastebin.ca/74908 If I download the root file system from ftp://ftp.arm.linux.org.uk/pub/armlinux/people/xscale/mainstone/ for the most part the system boots and works, if I generate a rootfs using OE I get the errors listed in my pastebin post. Both rootfs's use glibc. Jun 29 16:25:35 I've double checked the mkfs.jffs2 command being run in my OE enviroment and it is identical to the one I use when generating the mainstone rootfs from the ftp site. Jun 29 16:26:16 one difference is the mainstone ftp rootfs is using busybox for init were the OE rootfs is using init.sysvinit Jun 29 16:27:54 hi andersee Jun 29 16:28:38 mr_nice_slacker: hehe Jun 29 16:28:44 he zecke Jun 29 16:29:05 mr_nice_slacker: well internet. I have asked people to test and finish the last bits at 2.6.10 and 2.6.0 Jun 29 16:29:22 mr_nice_slacker: instead of helping people preferred to stick with 2.4er kernel Jun 29 16:29:28 mr_nice_slacker: now my patches might be either mirrored somewhere, or they are not :) Jun 29 16:30:58 mr_nice_slacker: anyway 2.6.18 will work way better :) Jun 29 16:31:40 zecke, I'm going to repost a problem I posted a couple minutes ago (before you were logged on) since you have experiance with a PXA270 :-) Jun 29 16:31:56 Gerrath: well, I'm about to go offline again Jun 29 16:32:09 zecke, ok, I won't repost then :-) Jun 29 16:32:16 hehe :) Jun 29 16:32:27 my windoze is almost updated Jun 29 16:32:50 okay done :) Jun 29 16:32:59 zecke, the error thing I talked about last night when trying to get a different gcc-cross for the kernel, turned out to be my fault. Works fine now Jun 29 16:33:06 D O H Jun 29 16:33:10 woglinde: howdy Jun 29 16:34:07 andersee what are the future plans for uclibc? Jun 29 16:36:28 woglinde: I need to get a release out sometime soon... Jun 29 16:36:54 woglinde: I've been trying to track down a uCibc + mips + gcc 4.1.x problem for about a week now.... Jun 29 16:37:45 andersee oh :( Jun 29 16:38:14 woglinde: what I really could use is a spare $10 million or so, so I could retire and work on it full time. :-) Jun 29 16:39:06 woglinde: real work has been taking up nearly all of my time for months Jun 29 16:39:11 andersee yeah thats what everyone dreams of, but we are not marc shuttleworth Jun 29 16:39:24 sucks, doesnt it Jun 29 16:39:26 :-) Jun 29 16:39:34 andersee are you on hawaii again? Jun 29 16:39:44 heh, nope Jun 29 16:40:03 sitting at home in my basement office Jun 29 16:40:07 re thejapa Jun 29 16:40:27 woglinde: hi! Jun 29 16:46:23 RP: you tried to hide on Cwiis's pictures? Jun 29 16:50:47 zecke:hi, thx - so i will try the 2.6.17 - may i ask you about a .config file - i can't get a valid kernel allways this unrecognized kernel image. shuld i use another toolchain? actually i use the 3.4.1 toolchain from handhelds.org. btw booting from cf: boot vfat :) Jun 29 16:51:20 zecke, the issue we talked about last night, using different gcc for kernel. Turned out it was my fault, works fine now Jun 29 16:51:27 so no need to blame bitbake Jun 29 16:51:35 Kristoffer: what was the issue? Jun 29 16:51:49 using 3.4.3 gcc for compiling kernel, while using 4.1.1 for everything else Jun 29 16:52:01 03kristoffer 07org.oe.dev * rec186ae9... 10/ (2 files in 2 dirs): Jun 29 16:52:01 conf/distro/jlime-donkey.conf : +kernel-modules to image +fix kernel build Jun 29 16:52:01 * Adds Kernel-modules to all images Jun 29 16:52:01 * Fixes Kernel issue so that kernel is built using 3.4.4 Jun 29 16:52:01 packages/linux/linux-jlime-sh3_2.6.16.20.bb: 3.4.3 -> 3.4.4 Jun 29 16:52:03 * Fixes so it uses 3.4.4 instead of 3.4.3 due to 3.4.3 not being Jun 29 16:52:04 availabel. Jun 29 16:53:58 Kristoffer: file a bug report, say what you did run and ask for bitbake to detect it Jun 29 16:55:07 mr_nice use my config Jun 29 16:55:12 http://page.mi.fu-berlin.de/~heinold/.config Jun 29 16:55:46 woglinde: how is your wife doing? Jun 29 16:57:38 zecke fine Jun 29 17:00:15 woglinde: many thx! Jun 29 17:01:01 zecke, it worked :) so no bug Jun 29 17:01:09 was a fault on my part for not checking correctly Jun 29 17:01:23 Kristoffer: what did you not check correctly Jun 29 17:01:37 first off I used 3.4.3 which is not availabel as a gcc-cross-kernel package Jun 29 17:01:45 Kristoffer: My todo for next month look like: Hack SIMpad, Hack TinderBox, Hack BitBake Jun 29 17:01:58 Well, it worked fine so dont bother with this atleast Jun 29 17:02:02 zecke: ehm.. Jun 29 17:02:08 zecke: you forget "get filthy rich" Jun 29 17:03:11 zecke:the oder looks verry good *lol Jun 29 17:03:21 koen: didn't you know? I decided to get a flithy rich wife Jun 29 17:03:46 zecke I should tell al Jun 29 17:03:47 zecke: ah Jun 29 17:03:56 woglinde: to get flithy rich? Jun 29 17:04:10 koen: 0800-zecke-boy at your service Jun 29 17:04:16 zecke, one thing I thought of, why not create a cross-gcc-kernel-PV for every gcc availabel.. would take away any chance of people messing it up Jun 29 17:04:18 gosh VS .NET takes ages to install Jun 29 17:04:43 zecke: Half of the install time is finding the right CD to put in.. Jun 29 17:04:49 zecke 4 dvd's? Jun 29 17:05:14 NAbyss: worse :) Jun 29 17:05:31 zecke: I gave up with VS2k3.. haven't tried since then.. it's gone downhill? :) Jun 29 17:05:48 NAbyss: this is the VS C++ edition (VS2k3) Jun 29 17:06:09 Ahh.. Jun 29 17:06:12 NAbyss: It is so badly translated... your need to test the 2^3 options to get IIS installed Jun 29 17:06:18 NAbyss: and then you can continue Jun 29 17:06:36 NAbyss: insert 'Required Stuff CD' msg box comes up: Start with CD1 Jun 29 17:06:47 NAbyss: Insert CD1, Install required software Jun 29 17:06:54 NAbyss: hit 'skip' switch CDs... Jun 29 17:06:54 what is oz-devel? the mailing list... Jun 29 17:07:03 zecke: Don't remind me.. I put that nightmare behind me after many a visit to the psychologist ;) Jun 29 17:07:44 NAbyss: in Uni VS2k5 is dog sloow and is crashing a lot Jun 29 17:08:02 NAbyss: using a 'Makefile' project, compile, click on the error -> boom Jun 29 17:08:15 Yeah, I've heard as much.. I'm sticking with delphi 7.. I've got a copy of delphi 2005, and they've just copied all of vs2k3 and 2k5's mistakes.. Jun 29 17:08:21 Whoops. Jun 29 17:11:47 delphi? wasnt that a oracle? Jun 29 17:11:48 lol Jun 29 17:12:59 woglinde: think vc++, but pascal.. Jun 29 17:13:17 delphi is a lot like jbuilder Jun 29 17:13:23 or the other way around Jun 29 17:13:52 Yeah.. delphi was the original, IIRC, then they brought out c++builder, then jbuilder Jun 29 17:14:17 nabyss I know what delphi is, but I want make this joke Jun 29 17:14:35 woglinde: ah, okies =) Jun 29 17:17:17 would this by any means by a known issue of any kind? | /mnt/ext0/poli/4700/oe/build/tmp/angstrom/work/ipaq-pxa270-linux-gnueabi/bootstrap-image-1.0-r0/temp/run.do_rootfs.14147: line 660: 28076 Segmentation fault Jun 29 17:17:29 | ipkg: couldn't write Jun 29 17:20:09 poli: stop using the '-' in the path name Jun 29 17:20:15 poli: or upgrade ipkg Jun 29 17:20:33 ~mornings Jun 29 17:20:35 Mornings MUST be destroyed! (see also http://www.destroymornings.com/) Jun 29 17:21:12 zecke: ummm that was bitbake bootstrap-image :( Jun 29 17:22:11 poli: oh well, upgrade ipkg Jun 29 17:23:43 zecke: and how do I do that? (I never installed it in the first place, I guess OE did that for me, didnt it?) Jun 29 17:24:19 poli: a good chanche to learn and give back? Jun 29 17:24:57 zecke: possibly. :) Jun 29 17:36:52 zecke: qt sucks: http://rafb.net/paste/results/QXZtIq60.html Jun 29 17:38:25 koen: Quicktime? yes it sucks, it is proprietary Jun 29 17:38:36 zecke *g* Jun 29 17:39:23 koen: I would not say Qt sucks Jun 29 17:39:38 koen: it would suck, if it would compile and wouldn't run Jun 29 17:39:42 zecke: you would say that I need to pay TT loads of money to 'fix' it? Jun 29 17:39:56 TT would say that Jun 29 17:40:00 koen: no Jun 29 17:40:16 koen: Maybe their ARM Assembly does not work with EABI Jun 29 17:40:33 koen: so I understand that they refuse to cross compile to unknown platforms Jun 29 17:40:59 heh Jun 29 17:41:13 I doubt qt4/x11 has arm assembly Jun 29 17:41:27 koen: it has Jun 29 17:41:35 zecke oh Jun 29 17:41:41 koen: it has Assembly for x86, arm, sparc, mips for atomar test and set Jun 29 17:42:37 koen: the ARM assembly requires cores with at least 'swp' Jun 29 17:43:00 hm no asm for ppc Jun 29 17:43:08 and ppc Jun 29 17:43:24 last time I try to compile Jun 29 17:43:35 ppc was not supported platform Jun 29 17:43:41 zecke: well, it sucks that they don't have plain C fallbacks Jun 29 17:44:02 they have... Jun 29 17:44:18 but these use full blown locks, and I think it is not enabled by default Jun 29 18:04:35 zecke: it seems it doesn't like linuxgnueabi, since it expects 'linux' Jun 29 18:08:22 koen: right Jun 29 18:13:11 using the linux-g++ entries gives me: | {standard input}:243: Error: Rn must not overlap other operands -- `swpb r1,r0,[r1]' Jun 29 18:13:12 looks like you need to create a linux-gnueabi-g++ mkspec Jun 29 18:13:49 which is weird, since the mkspec doesn't seems arch dependant Jun 29 18:14:12 they can be Jun 29 18:15:08 03koen 07org.oe.dev * ra0bb6704... 10/packages/ipkg/ipkg-native_0.99.163.bb: ipkg-native: add 0.99.163, fetched from tarball instead of cvs Jun 29 18:19:14 * koen delegates the problem to mickeyl Jun 29 18:20:41 koen, wb Jun 29 18:20:51 hey joshua_ Jun 29 18:21:21 enjoy the trip? Jun 29 18:23:50 certainly Jun 29 18:23:59 excellent Jun 29 18:36:46 bbl... Jun 29 18:48:15 zecke, if you have a moment, look at my following posts and let me know if you have any ideas/suggestions: Jun 29 18:48:20 I've been struggling with a problem during user space booting for about a week. Here is my boot output: http://www.pastebin.ca/74908 If I download the root file system from ftp://ftp.arm.linux.org.uk/pub/armlinux/people/xscale/mainstone/ for the most part the system boots and works, if I generate a rootfs using OE I get the errors listed in my pastebin post. Both rootfs's use glibc. Jun 29 18:48:28 I've double checked the mkfs.jffs2 command being run in my OE enviroment and it is identical to the one I use when generating the mainstone rootfs from the ftp site. Jun 29 18:48:48 one difference is the mainstone ftp rootfs is using busybox for init were the OE rootfs is using init.sysvinit Jun 29 18:50:57 does anyone have a website for 02 wireless? Jun 29 18:50:59 Gerrath: well, I think we have proposed to try nfsroot two days ago Jun 29 18:51:14 Gerrath: and did you succeed in mounting this jffs2 file locally? Jun 29 18:51:20 Gerrath: how big are the libraries etc? Jun 29 18:51:57 Gerrath: I think these empty flash messages and the error relates :) Jun 29 18:52:11 zecke, I tried the nfs mount but had the same results, the files seem to be reasonable sizes. Jun 29 18:52:36 Gerrath: well, if it complains about JFFS2 you did not use nfsroot Jun 29 18:53:07 zecke, I better check that again.. I have run many tests over the past several days.. be back. Jun 29 18:58:03 zecke, ugh, I'm glad I checked again, I was using the rootfs in tmp/rootfs This rootfs is not correct since it has file names instead of devices under /dev I will point my nfs mount to my jffs2 RAM mount for the real jffs2 image :-) Be back.. Jun 29 18:58:41 Gerrath: the /dev stuff should not hurt due udev and such Jun 29 19:02:41 zecke, it does mess things up, I get a can't open console error if I use that rootfs, This was the same error I got when I made my own jffs2 image against the tmp/rootfs directory. uhm could udev be my issue? Jun 29 19:07:26 zecke, I do not have udev installed. Jun 29 19:07:35 minipanda: any luck getting your oe.db updated? Jun 29 19:09:17 ~seen RP Jun 29 19:09:20 rp was last seen on IRC in channel #oe, 1d 9h 43m 54s ago, saying: 'It can't yet but it shouldn't be diiffcult too d'. Jun 29 19:09:37 RP is on GUADEC. Jun 29 19:10:02 mickey|office: ??? Jun 29 19:10:14 i thought you were interested Jun 29 19:10:22 thanks Jun 29 19:10:24 np Jun 29 19:10:40 mickey|office:Maybe you know this: Jun 29 19:11:06 Where in the .oz454x branch is the kernel version set? Jun 29 19:11:33 if it's anyhow similar to .dev, then in conf/machine/.conf Jun 29 19:11:41 I was trying to build a 2.6.17 image for my poodle but it always comes out with a 2.6.16 kernel Jun 29 19:12:00 check whether 2.6.17 has a DEFAULT_PREFERENCE of -1 Jun 29 19:12:12 yes it does... Jun 29 19:12:19 then increase it Jun 29 19:12:37 e.g. Jun 29 19:12:38 PREFERRED_VERSION_linux-openzaurus = "2.6.17" Jun 29 19:12:40 in your local.conf Jun 29 19:12:50 Ahhh, thank you very much.. you are wise in the ways of OE :) Jun 29 19:12:50 or edit the file and remove it :D Jun 29 19:12:55 hehe Jun 29 19:13:25 I will give my Dual PII's another workout tonight then..:) Jun 29 19:13:50 excellent. let them work Jun 29 19:13:57 hvontres|work: well, even after updated with latest OE.db Jun 29 19:13:58 it's too hot here to power on my devel machine Jun 29 19:14:06 the dual xeons just emit too much heat :/ Jun 29 19:14:08 hvontres|work: i have the same issue like before Jun 29 19:14:32 How bad is it over there? Jun 29 19:14:48 34 C in my room Jun 29 19:14:53 24 out there Jun 29 19:15:07 OPEN a window..:) Jun 29 19:15:15 oh that's what they are for? Jun 29 19:15:29 i couldnt run my shuttle xpc until the hotel replaced our air conditioner. damn arizona weather :P Jun 29 19:15:35 heh Jun 29 19:16:32 how hot is it now kergoth ? Jun 29 19:16:43 hi Jun 29 19:17:08 hey hrw@tosa Jun 29 19:17:10 wif config on tosa is too complocated Jun 29 19:17:21 ya, that usbwlan sucks Jun 29 19:17:22 it can be made simpler Jun 29 19:17:40 i removed all pre/post stuff Jun 29 19:18:07 plan to move it to scripts in /etc/network/ Jun 29 19:18:19 Crofton: pretty cool this week, only 103F today.. course thats mostly due to the chance of thunderstorms lately. last week it was 108F+ all week Jun 29 19:18:51 heh Jun 29 19:19:06 mickeyl: i will get free hostcable :) Jun 29 19:19:21 I remember coming out of the plant around 5:30 and you could feel the radiant heat from eveything Jun 29 19:19:41 Juck, only about 90F (30C) Out in Oceanside..:) But kinda humid lately Jun 29 19:19:58 hrw-tosa: congrats :) Jun 29 19:20:16 mickeyl: second goes to dirk Jun 29 19:20:28 my paypal balance is too low. i need to work more on stuff that gets released to get donations ;) Jun 29 19:21:20 yep.. most of my donation account went to customs Jun 29 19:21:52 time to fix something on tosa Jun 29 19:22:42 for example for me wifi works after suspend Jun 29 19:23:49 but removal of if-pre/post removed disabling of wifi Jun 29 19:24:59 25% battery left Jun 29 19:25:21 tosa screen roxx Jun 29 19:25:25 no kidding Jun 29 19:25:40 lowest brightness and readable Jun 29 19:26:01 mickel:? Jun 29 19:26:11 ya? Jun 29 19:26:34 'no kidding'? Jun 29 19:26:38 ah Jun 29 19:26:38 no kidding = you're damn right Jun 29 19:26:45 (at least that's how i use it) Jun 29 19:26:54 ah Jun 29 19:27:08 trying to improve my english by using phrases :) Jun 29 19:27:09 for me its reverse Jun 29 19:27:47 that would be "no, kidding" Jun 29 19:27:53 (notice the ',') :) Jun 29 19:27:58 for me 'no kidding' = "dont fuck" Jun 29 19:28:01 heh Jun 29 19:28:01 oka Jun 29 19:28:37 cu. movie in tv Jun 29 19:28:42 one minute Jun 29 19:28:47 what would you prefer for our OE berlin meeting Jun 29 19:28:51 september or oktober? Jun 29 19:29:17 yes? Jun 29 19:29:22 which month? Jun 29 19:30:07 hard to tell now Jun 29 19:30:14 hrw-tosa: oz-devel means the developers mailing list? Jun 29 19:30:16 ok, np. if you know, just give me ring Jun 29 19:30:26 nev:yes Jun 29 19:30:28 l8er Jun 29 19:30:32 * mickey|office tv as well Jun 29 19:30:38 ok mickey Jun 29 19:30:51 thanks...I will send an email... Jun 29 19:30:52 cu - suspend Jun 29 19:31:38 zecke, ok, I still get the exact same errors with the nfs mount as I do with the flash file system, minus the jffs2 warnings. The errors are the lib files are too small: http://pastebin.ca/75805 Jun 29 19:32:14 zecke, so it is not related to flash or jffs2 at all. Jun 29 19:32:17 invalid id Jun 29 19:32:26 Gerrath: wait Jun 29 19:32:39 Gerrath: are these files too small on the host? Jun 29 19:32:43 Gerrath: what is file saying? Jun 29 19:34:14 zecke, Here is "ls -l" from the host: http://pastebin.ca/75087 File sizes look good to me. Jun 29 19:34:44 Gerrath: I refuse to look at it, Are these files fine, or are they not? Jun 29 19:35:12 Gerrath: e.g. is the libc for ARM, does it carry 'enough' symbols, does it have an ELF header? Jun 29 19:35:22 Anybody manage to kexec from 2.6 back to 2.4 and start up OZ 3.5.4 ? Jun 29 19:35:39 hvontres|work: I think CoreDump|afk did Jun 29 19:35:58 Gerrath: if all files in the nfsroot are fine Jun 29 19:36:09 Gerrath: and they are wrong on the kernel... then... Jun 29 19:37:09 Gerrath: and your 75805 is too incomplete Jun 29 19:38:59 zecke, I did a readelf -h and the header looks good. Jun 29 19:39:30 Gerrath: okay, could you consider posting a complete bootlog with nfsroot? Jun 29 19:39:41 zecke, please bare with me I'm very new to kernel dev work :-) I'm trying to get up to speed quick. Jun 29 19:40:05 Gerrath: also, what was proposed days ago: Using a known working kernel on nfs root. Did you try that? Jun 29 19:40:16 Gerrath: yeah, I just want you to think :) Jun 29 19:40:33 generally, I'm lazy most of the time as well, but thinking is fun Jun 29 19:42:04 Gerrath: Some general hints: Compile ELF into the Kernel, use a working Floating Point Emulation, add the stuff you need to the kernel :) Jun 29 19:44:05 zecke, I do not have a known working kernel, thanks for the advice and guidence. Here is the entire boot log along with an elf header output at the end: http://pastebin.ca/75100 Jun 29 19:45:49 mickeyl: many apps need hires fixes.. tosa show it from other side then c7x0 Jun 29 19:46:14 Gerrath: okay stupid question (I have no clue there as well) Jun 29 19:46:24 Gerrath: if fork fails with No such device... Jun 29 19:46:45 Gerrath: do you have an EABI rootfs, and a non EABI kernel? or the other way around? Jun 29 19:47:45 mickeyl: look at my scaps Jun 29 19:51:02 zecke, my kernel does not have AEABI enabled, I'm not quite sure how to check what the rootfs is.. I'll google some I guess :-) Jun 29 19:52:18 "Starting portmap daemon: portmapportmap: fork: No such device." Jun 29 19:52:24 that looks weird Jun 29 19:52:59 zecke, I just looked in the elf header again, it says right there that it is EABI! Jun 29 19:53:25 zecke, I'll recompile the kernel with AEABI enabled. Jun 29 19:53:43 Gerrath: where did you get that rootfs? Jun 29 19:53:51 EABI rootfs'es are kind of rare Jun 29 19:54:38 koen, I built it with OE, I'm using the generic distro. Jun 29 19:54:53 hmmm Jun 29 19:54:57 that shouldn't use EABI Jun 29 19:54:59 scaps from tosa suxx too Jun 29 19:55:24 afaik, you can't even boot an eabi rootfs with an OABI kernel Jun 29 19:55:30 koen, I'm not even sure how to disable it :-) This is a fresh OE dev invorment. Jun 29 19:56:02 do you have objdump on that rootfs? Jun 29 19:57:18 could you do an 'arm-linux-objdump -p ' ? Jun 29 19:57:30 replace arm-linux with the arch you're targeting Jun 29 19:57:46 an eabi file would say something like 'private flags = 4000002: [Version4 EABI] [has entry point]" Jun 29 20:00:27 koen, I used readelf on one of the rootfs lib files and here is the flags: 0x202, has entry point, GNU EABI, software FP Jun 29 20:00:43 I don't really trust readelf Jun 29 20:00:47 koen, I will have to read up more on objdump before I can give you something usefull form it. Jun 29 20:01:29 go to tmp/cross//bin Jun 29 20:02:24 do ./-objdump -p /path/to/rootfs/lib/lib.so Jun 29 20:04:31 like this: koen@bitbake:/data/build/koen/OE/build/tmp/angstrom/cross/arm-angstrom-linux-gnueabi/bin$ ./objdump -p ../../../staging/arm-angstrom-linux-gnueabi/lib/libc-2.4.so Jun 29 20:04:48 the last line lists the private flags Jun 29 20:05:26 koen, does not look like EABI. private flags = 202: [APCS-32] [FPA float format] [software FP] [has entry point] Jun 29 20:05:37 that's what I thought Jun 29 20:06:12 for reference, my eabi builds say: private flags = 4000002: [Version4 EABI] [has entry point] Jun 29 20:07:12 hmm busted :} Jun 29 20:07:27 that's why I don't trust readelf Jun 29 20:07:55 koen, ok so elfread is not very accurate :-) Jun 29 20:08:53 koen, so I guess I'm back to square 1 which is no where. Jun 29 20:09:08 you ruled out some more possibilities Jun 29 20:09:15 koen, up. Jun 29 20:09:22 koen, I mean yup. Jun 29 20:09:53 well I guess I need to start working on why portmap is not forking. Jun 29 20:10:12 or why it is saying no such device. Jun 29 20:10:53 Gerrath: init=/bin/sh and see where it is going Jun 29 20:11:13 zecke: that doesn't work with busybox Jun 29 20:11:24 'init=/bin/busybosh ash' Jun 29 20:11:28 ehm Jun 29 20:11:44 'init=/bin/busybox ash' Jun 29 20:11:44 busybosh? Jun 29 20:11:48 man - that chick has such a busybosh! Jun 29 20:11:50 zecke, yes I tried the init=/bin/sh before and got a kernel panic.. thanks koen I will try that.. Jun 29 20:13:39 koen, zecke thanks for the help, I think you gave me enough input to start a second round of troubleshooting. Talk to you later :-) time to debug. Jun 29 20:14:08 Gerrath: have fun. I'm sure you will find it Jun 29 20:17:37 lol Jun 29 20:17:43 NASA is awesome... Jun 29 20:17:59 laser equipment to scan for damage through trash in the space... Jun 29 20:18:24 NASA is great Jun 29 20:18:32 capturing birds before the launch Jun 29 20:18:44 koen: taking pictures and videos from the ISS Jun 29 20:18:52 they have cool tech(tm) Jun 29 20:20:10 anyone know where I could get a .deb for qt/embedded-3.3.4? Jun 29 20:20:22 bgat: qt/e 3 doesn't exist Jun 29 20:20:29 only 2 and 4 Jun 29 20:20:32 koen: it does exist Jun 29 20:20:37 koen: schurig is using it :) Jun 29 20:20:46 koen: better talk to Trolltech about their website, then. :) Jun 29 20:20:52 zecke: ah, but it was FUDded to death right? Jun 29 20:21:09 tho, 2.x is much more popular, I'm trying to switch a debian/stable machine to qte and that seems the most direct path... Jun 29 20:21:14 koen: declared too huge, and produced a business deadlock Jun 29 20:21:15 I remember hearing something about TT not wanting opie to use or something Jun 29 20:21:17 ... to use 3.3.4, I mean. Jun 29 20:21:25 koen: Sharp didn't want to use Qt3 when no other is using Qt3 Jun 29 20:21:37 koen: others didn't want to use Qt3 when Sharp didn't use it... Jun 29 20:21:52 the above is just my impression Jun 29 20:21:54 to clarify Jun 29 20:22:38 so, does anyone know of a package, before I start building it on my own? Jun 29 20:22:58 bgat: Debian? Jun 29 20:23:09 zecke: nasa also has cool animals: http://dominion.kabel.utwente.nl/koen/cms/index.php/v/Koen/buitenland/amerika2005/ksc/IMG_1390.jpg.html?q=gallery Jun 29 20:23:47 arg, sf is broken again :/ Jun 29 20:23:47 zecke: yes. debian-arm, if you must know. :) Jun 29 20:24:08 I know I'm not exactly in the right forum, but people here seem to understand embedded a little better than #debian :) Jun 29 20:24:15 bgat: not the right channel though, did you try packages.debian.org? Jun 29 20:24:51 koen: what a small alligator, don't think i've ever seen one that small Jun 29 20:24:58 zecke: I checked there first, either I don't have the right search or there isn't one... Jun 29 20:25:26 mithro: they come in all sizes :) Jun 29 20:25:41 bgat: yeah, didn't find one myself Jun 29 20:25:48 mithro: I hope to see some OZ salt-water crocodiles one day Jun 29 20:26:00 but that would probably involve going to oz Jun 29 20:26:06 can somebody check http://sourceforge.net/export/rss2_projsummary.php?group_id=132078 and see if it dead halfway through Jun 29 20:26:13 bgat: looks like you need to build one, or teach OE to create debs Jun 29 20:26:26 zecke: let's just say I'm not keen on building up my own if someone else has already... Jun 29 20:26:41 mithro: error Jun 29 20:26:54 mithro: I get an error, am I owned yet? Jun 29 20:27:03 okay Jun 29 20:27:11 zecke: owned? Jun 29 20:28:50 hey mithro Jun 29 20:28:57 hey lpotter_ Jun 29 20:29:02 why you up this hour? Jun 29 20:30:08 lpotter_: hey Jun 29 20:30:09 any able to get to http://sourceforge.net/project/stats/?group_id=132078&ugn=thousandparsec ? Jun 29 20:30:29 mithro: no, I just get a blank page Jun 29 20:30:35 lpotter_: I have seen you have relaesed Qtopia 4.1.2, 4.1.3 and just 4.1.4 Jun 29 20:31:03 lpotter_: any ETA when the dual licensed, award winning Qtopia, gets a third license (GPL) and is made publically available? Jun 29 20:31:27 bgat: thanks Jun 29 20:31:28 lpotter_: give us the source, help us to add it to OE, create a linuxdevices.com story, success :) Jun 29 20:31:38 lpotter_: any chance you will be coming to LCA 2007? Jun 29 20:38:15 are there any non-embedded OS using BitBake? Jun 29 20:38:28 not that I know of Jun 29 20:38:55 do you know of any reason why not? or it just hasn't happened? Jun 29 20:39:13 whats a non-embedded OS anyway/ Jun 29 20:40:32 CosmicPenguin: desktop Jun 29 20:42:03 Ohh s/OS/distribution/ Jun 29 20:42:18 And sure - people build desktop style images all the time Jun 29 20:42:27 Windows and FreeBSD are distributions? Jun 29 20:42:45 "distribution" is an incorrect label applied to Linux-based OS Jun 29 20:42:50 No Jun 29 20:43:01 OS is an operating system - distribution is a collection of packages Jun 29 20:43:20 What packages are collected make all the difference in the world Jun 29 20:43:29 ok, so you have a point that most Linux-based OS are also distributions Jun 29 20:43:36 but I was primarily asking about OS in general Jun 29 20:43:57 with or without the accompanying collection of pkgs Jun 29 20:44:14 Ok - so bascially, you're asking if anyone has ever used bitbake to compile a kernel other then Linux Jun 29 20:44:22 I guess I'm just wondering if anyone's used BitBake as one would use Portage, and if not why Jun 29 20:44:37 And the answer is, some people have considered it, but I don't know if they were successful Jun 29 20:44:59 having Linux as a kernel doesn't make an OS any less an OS Jun 29 20:54:34 mithro: I am sure I could come to LCA2007 Jun 29 20:55:14 i need speakers for the Embedded Miniconf there Jun 29 20:55:44 ok Jun 29 21:01:00 ljp: did you see my questions? :) Jun 29 21:01:12 ljp: will bitching at mr. ettrich help? Jun 29 21:05:10 nope. Jun 29 21:06:04 they need a commercial product behind any GPL release, and so they are refactoring qtopia a bit, so we can release the phone interface without telephony Jun 29 21:08:58 too bad that it's taking so long. meanwhile one or the other oss embracing ODM will consider alternatives Jun 29 21:09:10 i agree. Jun 29 21:17:32 all the options suck in a way. i will have to decide between recommending basing something own based on qtop^h^h/E or using gtk+ or efl then :/ Jun 29 21:17:59 so much for a common linux phone platform Jun 29 21:20:48 * mickey|tv signs http://osnews.com/story.php?news_id=15040 and sighs :/ Jun 29 21:23:28 ermmm.. each symbian version is incompatible with the next.. whats the diff? Jun 29 21:23:50 the difference is that i don't care about symbian. i care about linux and we _had_ a chance to do better Jun 29 21:24:00 but it looks like it's already too late Jun 29 21:24:01 mickey|tv, has efl actually become something usable? Jun 29 21:24:56 ggilbert: efl is great. the widget sets on top of evas, ecore, and edje are getting better every day. Although there a lot of quirks and still lots of things to do, they are an alternative for someone who wants to have a unique UI. Jun 29 21:25:01 actually, most manufactures hide the fact it runs linux Jun 29 21:26:45 eugenia like bashing linux Jun 29 21:27:08 yes, but she has a point Jun 29 21:27:20 eugenia? Jun 29 21:27:27 she doesnt understand that manus and phone operators do not want users to run native apps Jun 29 21:27:48 not yet. Jun 29 21:27:52 I believe this will change Jun 29 21:28:13 not all ODMs are loving j2me Jun 29 21:28:16 Doesn't anyone care about the customers? :p Jun 29 21:28:18 they are some with a clue Jun 29 21:28:21 thankfully Jun 29 21:28:37 ljp: well, they want to control the number of native apps Jun 29 21:28:40 and that it is actually illegal (in some places) to mess with phones Jun 29 21:28:54 ljp: and well, they use j2me anyway, so it works on windows mobile but they are not too locked in Jun 29 21:29:09 * ggilbert_ rather loves the fact that he doesn't need to go hunting for some phone specific version of an app when I'm looking for a bit of functionality Jun 29 21:29:14 ljp: I think one needs to differentiate Jun 29 21:29:37 ljp: In any country: If you have a wireless module (separate core), that is certified independandly Jun 29 21:29:57 ljp: you can mess with the 'phone' as much as you want Jun 29 21:29:59 *nod* GSM cert, modem interface - period. Jun 29 21:30:19 ljp: e.g. Siemens, or the stuff you use the most, Wavecom provide these modules :) Jun 29 21:30:23 but a lot of phones do not have a seperate baseband processor Jun 29 21:30:51 well they are more expensive and a lot of people want to do everything in one processor to decrease the BOM Jun 29 21:31:17 but these phones won't be picked up by the community anyway (too low end) Jun 29 21:31:33 the smartphone class you're pushing will most certainly have Jun 29 21:31:40 and as seen on linuxdevices.com, Qtopia isn't serving this low level market, which is getting bigger, well Jun 29 21:32:20 actually, no. as I said, not every phone has that feature Jun 29 21:33:14 ljp: if you release GPL source, and these vendors get scary Jun 29 21:33:21 you either have serve security issues within the code Jun 29 21:33:30 (which you have old rsync, zlib...) Jun 29 21:33:33 png Jun 29 21:33:51 but the solution is not not to GPL Jun 29 21:34:30 I hope I have enough time tomorrow to sketch my akademy paper Jun 29 21:34:44 "The lost momentum", I will demonstrate GPE and display migration Jun 29 21:34:56 pb_: what is the state of display migration nowadays? Jun 29 21:35:18 zecke: is it that teleport app? Jun 29 21:35:32 thejapa: it is just a GUI Jun 29 21:35:43 thejapa: now assume you can hook it up with your usb device Jun 29 21:35:44 oops :) Jun 29 21:36:02 with your bluetooth device... Jun 29 21:36:08 and you move around in the room :) Jun 29 21:36:20 mmm... bluetooth Jun 29 21:36:24 zigbee Jun 29 21:36:41 cool Jun 29 21:37:12 I find it ridicoulous that 'desktop' projects try to copy the copy of OSX :) Jun 29 21:37:27 i need a big screen for that. 240x320 isn't enough Jun 29 21:37:27 to get into that market, while another huge, even bigger market is there Jun 29 21:37:48 thejapa: well. In Opie we had 'BigGUI' classes Jun 29 21:38:00 thejapa: depending on the Desktop Widget Size, the GUI changed Jun 29 21:38:08 ahhh Jun 29 21:38:13 thejapa: you could teleport your app to a hughe display, and update the GUI Jun 29 21:38:17 (should I say) ohhh Jun 29 21:38:36 thejapa: this would mean you could carry your small bubble gum sized mobile device with you Jun 29 21:38:59 thejapa: which got GSM/UMTS/WLAN/BNEP to connect to your rack to store and backup your data Jun 29 21:39:14 while giving you the experience of using your normal office applications locally :) Jun 29 21:39:48 actually is just creating this platform, they only need a webbrowser for their platform to work Jun 29 21:40:09 ljp: when will this Qt Terminal Server system be published? Jun 29 21:40:17 zecke: "The lost momentum" -- that inspires me to blog about opie and qtopia asap. Jun 29 21:40:38 mickey|tv: hehe, I wanted to rant, but did not have enough time :) Jun 29 21:40:44 :) Jun 29 21:40:56 i will reserve a 1h time slot for that tomorrow Jun 29 21:41:08 mickey|tv, zecke: have either of you guys worked with nokia's stuff yet? Jun 29 21:41:20 the mystical "next man month" Jun 29 21:41:31 ggilbert_: not much, enough to get some feeling Jun 29 21:41:49 ggilbert: yes. it's cool. it's not perfect, but it's cool. I'm still waiting for a complete C++ binding though. I can't stand C. Jun 29 21:41:52 ggilbert_: it is promising, specially with the stuff o-hand is cooking Jun 29 21:42:01 mickey|tv: use the python bindings Jun 29 21:42:21 zecke: I'm afraid that'll be too slow. the N770 is not powerful enough. Jun 29 21:42:33 Promising is a good sign :p Jun 29 21:42:38 mickey|tv: use IronPython then? Jun 29 21:42:42 particularly if they stick with it Jun 29 21:42:55 zecke: *aufstampf* I WANT C++ Jun 29 21:43:13 mickey|tv: me too, I have heard one can mix c and c++ Jun 29 21:43:19 huh? Jun 29 21:43:20 that's just a rumour. Jun 29 21:43:26 i have yet to see a prove Jun 29 21:43:26 isn't C++ the same thing as D? Jun 29 21:43:42 okay back to work Jun 29 21:43:49 good idea. cu Jun 29 21:43:53 * mickey|tv returns to tv Jun 29 21:43:54 mickey|tv: have you ever seen displaymigration Jun 29 21:44:04 or does it work, the way it does with academic grades? Jun 29 21:44:07 zecke: not yet. you have to show me when we meet Jun 29 21:44:10 C++ == B Jun 29 21:50:24 hi Jun 29 22:00:02 g'night Jun 29 23:07:42 hmmm. my simpad no longer calibrates the screen Jun 29 23:38:53 cya later Jun 30 01:04:12 03lenehan 07org.oe.dev * r3ffe0b7f... 10/packages/linux/linux-titan-sh4-2.6.16/titan-config: Jun 30 01:04:12 linux-titan-sh4 2.6.16: Disable IP auto configuration for the kernel. This Jun 30 01:04:12 kernel goes in the flash (and not nfsroot) and so shouldn't be auto Jun 30 01:04:12 configuring itself. Jun 30 02:12:51 * chouimat is away: Zzzzzz **** ENDING LOGGING AT Fri Jun 30 02:59:57 2006