**** BEGIN LOGGING AT Sat Aug 04 02:59:56 2007 Aug 04 04:33:29 Has the Collie SD patch been rolled into the kernel for Collie builds? Aug 04 05:43:47 drmikecrowe: please don't use 'PACKAGES = ', only 'PACKAGES =+ ' or 'PACKAGES =+ ' Aug 04 05:44:31 koen: hmm, I missed that. I'll update Aug 04 05:46:56 http://rafb.net/p/PoEQKe71.html # I just think openssl needs this on uclibc-gnueabi-arm. Aug 04 05:47:17 otherwise, it could not pass oe_conf. Aug 04 06:02:24 koen: As I build my system, I'm putting install stuff in my bb file, and compile stuff in my makefile. Not that it's a big deal, but any issues with setting it up that way? Aug 04 06:02:50 not that I know off Aug 04 06:04:03 koen: morning Aug 04 06:04:37 koen: I am able to build and boot OE angstrom-console-image finally for first time :) Aug 04 06:05:00 I did nfs booted it Aug 04 06:09:32 nice Aug 04 06:09:56 hrw|gone: 'inherit lib_package' in your poky speex recipe will give the same packaging, but less work :) Aug 04 06:12:34 koen: Do u know if I can force dropbear to accept empty password and log me in without asking for confirmation Aug 04 06:13:08 koen: I want to do some automated testing which needs it. It accepts empty passwords I saw that patch too Aug 04 06:13:42 but it asks to enter password so something like if password is empty then I do not have to confirm it Aug 04 06:13:44 the easiest way is to add you pubkey to .ssh/authorized_keys Aug 04 06:14:14 (a pubkey without a password ofcourse) Aug 04 06:15:13 koen: ok Aug 04 06:15:38 koen: I was thinking if there was something PermitEmptyPasswords like thing Aug 04 06:15:43 as it is in openssh Aug 04 06:16:11 koen: Did u see the patch I posted for omap1 kernel build ? Aug 04 06:20:09 I saw the patch and I replied Aug 04 06:20:44 (like 10 seconds ago :) ) Aug 04 06:21:01 * koen loathes uImages Aug 04 06:23:07 ok Aug 04 06:24:01 koen: unfortunately on omaps its the most prevalent image type Aug 04 06:25:00 koen: The patch looked ok to me and the resulting image booted fine for me on osk5912 the earlier build I was having problem with uncompressing Aug 04 06:25:23 The make uImage target is the right think to do in my opinion for omaps Aug 04 06:26:12 unfortunately 'make uImage' doesn't work on all kernels :( Aug 04 06:26:29 that's why I copied the code from the makefile into linux.inc Aug 04 06:27:03 but if 'make uImage' does work, we should use it Aug 04 06:27:20 * koen waits for Crofton|home to wake up Aug 04 06:27:30 I see ok however for omaps we should use it Aug 04 06:27:44 koen: he said he might not work over the weekend Aug 04 06:35:35 morning all Aug 04 06:36:09 uimage is an issue on ppc targets also Aug 04 06:36:26 not all kernels support it Aug 04 06:45:33 steliosk: ok this is only for arm/omap though Aug 04 06:46:29 03koen 07org.oe.dev * rdf1e6ea0... 10/ (9 files in 3 dirs): Hal: merge improvements from poky, remove stale versions Aug 04 06:47:55 khem : I am handling it inside the kernel.bb for ppc Aug 04 06:48:30 steliosk: for arm/omap I added it to linux-omap1.inc Aug 04 08:26:46 03koen 07org.oe.dev * rea8323f4... 10/ (3 files in 3 dirs): zaurusd: add patch for finding tslib 1.0 pkgconfig file Aug 04 10:27:18 koen, ping Aug 04 10:30:21 Crofton|home: pong Aug 04 10:30:32 I am "awake" Aug 04 10:30:45 I use the term losely though, no coffee yet Aug 04 10:43:31 Efika build failed at libusb again Aug 04 10:44:05 hi all Aug 04 10:44:52 hmmm Aug 04 10:45:03 I made a thinko with ANGSTROM_MODE=eglibc Aug 04 10:46:19 cool Aug 04 10:46:46 I suspect we should just do make Uimage Aug 04 10:47:08 I am curious why the manual opeations wouild not make a kernel that khem couldl boot Aug 04 10:48:18 I suspect a different entry point Aug 04 10:48:24 (the -e and -a switches) Aug 04 10:48:39 yeah Aug 04 10:48:49 (I don't know what an entrypoint is, but the kernel makefile specifies it) Aug 04 10:49:08 I had a problem with entry point at one stage Aug 04 10:50:23 my entry is ox8000 from bottom of ram Aug 04 10:56:02 typcial nerds, drive the chick out of defcon Aug 04 10:56:14 http://blog.wired.com/27bstroke6/2007/08/media-mole-at-d.html Aug 04 10:59:10 RP: what do you think of http://bugs.openembedded.org/attachment.cgi?id=1784&action=edit ? Aug 04 10:59:26 ok, trying to build bootstrap-image for osk Aug 04 11:01:51 Crofton|home: have a look at how linux.inc handles uImage Aug 04 11:03:48 Crofton|home: it should allow tweaking of option by adding a few if MACHINE= statements Aug 04 11:04:11 linux.inc converts vmlinux to uImage? Aug 04 11:04:23 looks like it always uses the same entry points Aug 04 11:07:26 We badly need to shake down TI for some dev boards to pass around Aug 04 11:07:40 amen Aug 04 11:07:50 * koen heads out to buy lunch ingredients Aug 04 11:07:55 just having some some more people can test builds Aug 04 11:07:57 heh Aug 04 11:07:59 hav efun Aug 04 11:08:12 I'll be gone in an hour or so Aug 04 11:08:15 must dig a hole Aug 04 11:37:29 hello all Aug 04 11:38:04 gm Aug 04 11:38:40 Crofton|home: Got enough inspiration to work on your thesis Aug 04 11:38:53 heh Aug 04 11:38:57 today I am not working Aug 04 11:39:01 tomorrow I work Aug 04 11:39:51 got around to bring the efika home? Aug 04 11:39:52 well, I am trying an osk bild atm Aug 04 11:39:56 yeah Aug 04 11:40:05 but now I can't build libusb for Efika Aug 04 11:40:10 C++ linake age issues Aug 04 11:40:31 possibly realted to me dorking with libtool version, but I do not think so Aug 04 11:40:47 so trying an OSK build to see if newer libtool works there Aug 04 11:41:00 and to look at kernel issue raised by khem Aug 04 11:42:46 Crofton|home: OK, my working build was with an older rev. (see that wiki) Aug 04 11:42:55 yeah Aug 04 11:42:58 I'll try to build it against current Aug 04 11:43:15 we need an autobuild system going Aug 04 11:43:40 that rebuilds lots of targets and lets us know when things break Aug 04 11:44:58 Crofton|home: yes, I'm not sure what is being run at OpenEmbedded.org for that Aug 04 11:45:24 At work, I now have a build server, from which the target(s) boot over the latest kernel/rootfs using TFTP/NFS. Aug 04 11:45:41 my build server at work died :) Aug 04 11:45:52 I need to fix something up Aug 04 11:45:57 To add, is a verification stage, that tests the availability of the rootfs application Aug 04 11:47:07 just pinning down when we introduce build bustage would be helpful :) Aug 04 11:47:16 verification just awesome Aug 04 11:47:58 I wonder how you would script the image flashing process and boot verification Aug 04 12:02:43 I just found an AVR group on facebookl Aug 04 12:02:54 and an embedded developers group Aug 04 12:03:42 autobuilds autobuilds autobuilds Aug 04 12:04:44 facebookl? Aug 04 12:06:03 Crofton|home: I can flash by scripts (using netcat to a BDI2000 flasher), I can flash by bootloader scripts (RedBoot), but I do not Flash. Most targets load kernel into RAM over TFTP and mount root over NFS. Aug 04 12:07:21 * koen discovers that http://www.j5live.com/?p=383 is quite yummie Aug 04 12:07:31 hi koen Aug 04 12:07:35 hey likewise Aug 04 12:08:07 Linksys NAS200 for anyone who is interested: http://www.nslu2-linux.org/wiki/NAS200/HomePage, pic at http://derekyoung.net/IMG_1720.JPG (ignore the purple, dyoung's digital camera is playing up) Aug 04 12:08:08 koen: you are vegie now, or nynke is? Aug 04 12:08:21 no vegie, I just like mushrooms a lot Aug 04 12:09:07 rwhitby: hi, RDC, is that the cripple x486? Aug 04 12:10:32 rwhitby: http://www.rdc.com.tw/Uploads/datasheet/R3210_Mbrief_20061121.pdf Aug 04 12:10:37 some 486 microcode emulated thing I think Aug 04 12:10:44 likewise: yep, seen that. Aug 04 12:11:00 it's basically an x86 slug with SATA Aug 04 12:11:30 rwhitby: nice casing, sadly only 100Mbit :( Aug 04 12:11:32 rwhitby: hmmm. I really dislike the x86 architecture coming in cheaper than embedded archs here... Aug 04 12:12:09 * rwhitby flashes two neo's using his router as the host. Aug 04 12:12:31 likewise: it's cheap because they can hire incompetent embedded 'engineers' ;) Aug 04 12:12:54 * rwhitby wonders if the RDC R3210 has the A20 gate thingo Aug 04 12:18:30 likewise, www.facebook.com Aug 04 12:18:53 * Crofton showers Aug 04 12:28:31 Crofton : I make uimage "manually" on the magicbox kernels Aug 04 12:29:07 Crofton : what's the issue with libusb/efika ? Aug 04 12:33:03 steliosk: you visiting Freescale Austin? Aug 04 12:33:40 likewise : Hope to. Haven't made any contacts yet Aug 04 12:34:22 steliosk: I saw some nice parts coming out soon Aug 04 12:34:25 steliosk, c++ linkage issues on libusb Aug 04 12:34:33 trying an OSK build atm Aug 04 12:34:47 Crofton: that was with current EFIkA, angstrom glibc? Aug 04 12:35:15 yes Aug 04 12:35:42 MACHINE=efika Aug 04 12:35:50 likewise : yeah. i am interested forthe powerquick + coprocessor Aug 04 12:35:52 DISTRO="angstrom-2007.1" Aug 04 12:36:22 Crofton|home : my last efika build was a couple of weeks ago and was fine Aug 04 12:36:30 steliosk: QUICC engine co-procesor? Aug 04 12:36:32 +s Aug 04 12:36:48 koen, mentions he has experienced random toochain breakage Aug 04 12:36:59 steliosk: 8314, 8315, 8372, 8377, 8321? Aug 04 12:37:13 ouch Aug 04 12:37:18 http://www.freescale.com/webapp/sps/site/overview.jsp?nodeId=02VS0l383359858308 Aug 04 12:37:38 ok, I am off Aug 04 12:37:53 maybe I'll take some pictures of crazy stuff we do for fun :) Aug 04 12:38:06 Crofton|home : strange Aug 04 12:38:13 yeah Aug 04 12:38:22 Crofton|home : will look at it Aug 04 12:38:28 poor handling of the lines for -lstdc++ Aug 04 12:38:31 thanks Aug 04 12:38:34 more eyes the better Aug 04 12:39:45 heh Aug 04 12:40:18 i am "pitching" angstrom/efika on the power.org presentation... Aug 04 12:40:28 that should be awesome Aug 04 12:40:28 * steliosk hopes it does not break by them Aug 04 12:40:34 heh Aug 04 12:40:49 we need support for stable dev env ... Aug 04 12:40:53 ok Aug 04 12:41:00 I have to go, friends actually leaving on time Aug 04 12:41:07 poof Aug 04 12:41:22 bye Aug 04 12:41:38 stable dev environment sounds like military intelligence :) Aug 04 12:42:08 contradiction in terms Aug 04 12:44:26 steliosk: we are planning to use the 83xx series, xx13 and xx14, xx72 maybe. Aug 04 12:44:44 steliosk: you know their ltib/bitshrine project? Aug 04 12:44:58 nop Aug 04 12:45:13 try to convince them to add Freescale support to OE, instead of LTIB (see bitshrine.org) Aug 04 12:45:34 Linux Target Image Builder is Freescale's "buildroot" Aug 04 12:46:59 bbiab Aug 04 12:49:42 likewise : interesting. will have a look. Have you used it ? Aug 04 12:50:16 by reading the docs it looks alot like openwrt stuff Aug 04 12:50:19 * rwhitby used to work for Freescale, and architected a PowerQUICC 3 device while there. Aug 04 12:51:37 rwhitby : heh Aug 04 12:51:51 so you will be the one we will torture for details :) Aug 04 12:51:59 only if it's an 8568 Aug 04 12:52:37 I know nothing about the 83xx series which isn't in the public databooks, since they were done elsewhere. Aug 04 13:11:01 rwhitby: you ever programmed the QUICC engines? Aug 04 13:11:11 likewise: nope Aug 04 13:11:11 * likewise is unpacking his dreambox 600 pvr Aug 04 13:11:36 me neither, just wondering if it is as "open" as is the NPE co-procs on the IXP Aug 04 13:16:40 likewise: I believe in both cases it's all just a matter of how many silicon unit's you're going to buy ;-) Aug 04 13:18:57 likewise : probably Aug 04 13:19:37 likewise : currently they list 3 companies with knowledge on them Aug 04 13:24:30 likewise : it looks like bitshrine is using pre-build toolchains, which means that they can be integrated with OE Aug 04 13:26:22 steliosk: yes, I would rather build those exact toolchains from source though. Aug 04 13:26:29 source RPM's are there too Aug 04 13:26:41 or leave both options Aug 04 13:40:55 does source rpm builds work ? Aug 04 13:41:16 i think there is/was a problem with rps handling in general Aug 04 13:41:23 s/rps/rmp Aug 04 13:41:33 s/rmp/rpm Aug 04 13:44:30 morning Aug 04 14:13:31 bbiab Aug 04 14:15:00 mickeyl: good morning! Aug 04 14:15:08 morning Aug 04 14:23:10 koen: hmm, adding this openmoko DISTRO override is work Aug 04 14:23:26 i would have preferred to go with the angstrom override for some time Aug 04 14:24:01 i wonder how much of the angstrom include files i can use Aug 04 14:24:17 i don't fancy copying too much and then trying to keep in sync Aug 04 14:36:15 03mickeyl 07org.oe.dev * r1e3fbad3... 10/ (5 files in 4 dirs): Aug 04 14:36:15 base-files: add openmoko specifics Aug 04 14:36:15 repair formatting, guys would you PLEASE STOP MIXING TABS AND SPACES!!!!!!!!!! Aug 04 14:36:25 03mickeyl 07org.oe.dev * rafab7b05... 10/ (7 files in 4 dirs): Aug 04 14:36:25 sysvinit: add openmoko specifics Aug 04 14:36:25 repair formatting. DON'T MIX TABS AND SPACES!!!!!! Aug 04 14:36:37 03mickeyl 07org.oe.dev * r6b47044e... 10/ (14 files in 13 dirs): busybox: add openmoko specifics Aug 04 14:36:48 03mickeyl 07org.oe.dev * r65905a94... 10/ (1 conf/distro/openmoko.conf): conf/openmoko.conf: change DISTRO name to make openmoko override active Aug 04 14:37:05 03mickeyl 07org.oe.dev * r75cc5b6d... 10/ (1 packages/tslib/tslib_1.0.bb): Aug 04 14:37:05 tslib: recommend pointercal also for openmoko Aug 04 14:37:05 fix formatting *sigh* Aug 04 15:47:10 mickeyl: symlinks Aug 04 15:47:30 hmm Aug 04 15:47:34 monotone supports symlinks ? Aug 04 15:47:37 that would be really cool Aug 04 15:49:10 * koen tries Aug 04 15:50:11 koen : is there a problem with OE site ? Aug 04 15:50:30 mickeyl: http://lists.linuxtogo.org/pipermail/openembedded-devel/2007-July/002623.html Aug 04 15:51:02 steliosk: apart from it being slow? Aug 04 15:51:24 94.4% wa Aug 04 15:52:15 unable to connect to database ? Aug 04 15:53:28 mickeyl: mtn dereferences symlinks :( Aug 04 15:54:21 "The attribute mechanism was originally motivated by the fact that some people like to store executable programs in version control systems, and would like the programs to remain executable when they check out a workspace. For example, the configure shell script commonly shipped with many programs should be executable. Similarly, some people would like to store devices, symbolic links, read-only files, and all manner of extra attributes of a file, not di Aug 04 15:54:54 bummer Aug 04 15:55:19 i really want angstrom to be a fallbakc Aug 04 15:55:42 so i guess the only thing for now is adding openmoko as override in front of angstrom Aug 04 15:55:54 only in openmoko.conf of course though Aug 04 15:55:59 * mickeyl tries tomorrow Aug 04 15:56:13 mickeyl: or add angstrom as an override in openmoko.conf :) Aug 04 15:57:24 mickeyl: btw, openmoko will suffer from http://bugs.openembedded.org/show_bug.cgi?id=2601 as well Aug 04 15:58:37 eeks, nasty Aug 04 15:59:20 one more reason to only allow utf8 ;) Aug 04 16:10:20 03koen 07org.oe.dev * r97361f57... 10/ (1 delete.txt): mtn: symlink test Aug 04 16:12:07 03koen 07org.oe.dev * r97361f57... 10/ (1 delete.txt): mtn: symlink test Aug 04 16:12:14 03koen 07org.oe.dev * rac74c69b... 10/ (1 packages/angstrom/angstrom-x11-base-depends.bb): angstrom-x11-base-depends: mark as machine specific Aug 04 16:12:21 03koen 07org.oe.dev * r373f8ffa... 10/ (3 files in 2 dirs): abiword 2.5: fix packaging Aug 04 16:20:28 hi all Aug 04 16:28:17 Hi folks, what would cause the error: arm-angstrom-linux-gnueabi-ld: unrecognized option '-Wl,-rpath-link,/opt/stuff/tmp/staging/arm-angstrom-linux-gnueabi/lib' Aug 04 16:28:22 ? Aug 04 17:20:12 Any help? http://pastebin.com/d1412baf8 Aug 04 17:57:45 * cbrake_away boots up a shiny new new1973 :-) Aug 04 18:06:28 koen: Let me have a look at that later on, there should be a neater way to fix that Aug 04 18:15:36 RP: thanks Aug 04 18:16:20 RP: http://www.angstrom-distribution.org/unstable/images/c7x0/20070803/ Aug 04 18:16:37 I'm happy, since I now have a working solution as local diff :) Aug 04 18:22:02 koen: Looks good :) Aug 04 18:23:05 according to murphies law the images itself will be broken now they are easy to install Aug 04 18:23:58 koen: Naturally ;-) Aug 04 18:24:15 * RP thinks he knows the cause of the bitbake slowdown and the sqlite locking issue Aug 04 18:24:30 Has anyone ever used the ${A} variable? Aug 04 18:24:43 I don;t even know what it means Aug 04 18:24:52 I do know where B is used, though Aug 04 18:25:06 koen: glibc? Aug 04 18:25:17 out-of-sourcetree builds Aug 04 18:25:23 glibc, binutils, etc Aug 04 18:25:53 A is a list of the local paths from SRC_URI Aug 04 18:26:16 ah Aug 04 18:26:26 now I know it, it's obvious Aug 04 18:26:28 * koen coughs Aug 04 18:27:03 * RP only knows of it since its setup in an obsure bit of bitbake Aug 04 18:28:05 I'm sure there was one user but I can't remember where :/ Aug 04 18:30:52 * RP might resort to the disable it and see what breaks approach Aug 04 18:32:18 [20:22] jcole verlaat de chatruimte. ("Powered by OE: www.openembedded.org") Aug 04 18:32:29 * koen forward ports that branding patch to pidgin Aug 04 18:33:25 * RP -> back shortly Aug 04 18:33:54 03koen 07org.oe.dev * r15632785... 10/ (3 files in 3 dirs): pidgin: port OE branding patch to libpurple Aug 04 19:30:20 koen, webket in angst feed ? Aug 04 19:30:28 bother Aug 04 19:31:08 libusb builds for OSK Aug 04 19:31:55 now switching machine to efika and trying bootstrap-image Aug 04 19:34:24 hi Aug 04 20:05:48 how can i install gpe and all kernel modules on my hx4700 Aug 04 20:05:52 koen, looks like the Efika toolchain is hosed Aug 04 20:06:01 my OSK build succeeded Aug 04 20:06:10 * Crofton|home leaves again Aug 04 21:28:56 03mickeyl 07org.oe.dev * r4d4a16da... 10/ (1 conf/machine/fic-gta01.conf conf/machine/fic-gta02.conf): fic-gta0[12].conf: always generate little endian jffs2 images Aug 04 21:31:22 NOTE: Handling BitBake files: | (3366/5000) [67 %] Aug 04 21:31:30 exactly 5k bb files in my local tree :) Aug 04 21:39:18 koen: :) Aug 04 21:45:27 Um, "branch org.openembedded.dev has multiple heads", should I be worried? Aug 04 21:52:56 naw, don't be overly worried. Aug 04 21:53:05 It'll resolve itself shortly, or someone will fix it. Aug 04 21:56:43 It looks like it has been that way for about 24 hours Aug 04 22:14:13 03mickeyl 07org.oe.dev * r4d4a16da... 10/ (1 conf/machine/fic-gta01.conf conf/machine/fic-gta02.conf): fic-gta0[12].conf: always generate little endian jffs2 images Aug 04 22:16:59 * * OE Bug 2384 has been RESOLVED (FIXED) by heinold(AT)inf.fu-berlin.de Aug 04 22:17:01 * *  wrong uClibc.machine for simpad Aug 04 22:17:03 * * http://bugs.openembedded.org/show_bug.cgi?id=2384 Aug 04 22:17:59 * * OE Bug 1955 has been RESOLVED (FIXED) by heinold(AT)inf.fu-berlin.de Aug 04 22:18:01 * *  quilt does not get built properly with bb1.8 multithreading Aug 04 22:18:03 * * http://bugs.openembedded.org/show_bug.cgi?id=1955 Aug 04 22:19:58 * * OE Bug 2512 has been RESOLVED (FIXED) by heinold(AT)inf.fu-berlin.de Aug 04 22:20:00 * *  base-files has simpad-specific files Aug 04 22:20:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2512 Aug 04 22:43:54 03mickeyl 07org.oe.dev * r4d4a16da... 10/ (1 conf/machine/fic-gta01.conf conf/machine/fic-gta02.conf): fic-gta0[12].conf: always generate little endian jffs2 images Aug 04 22:50:54 03mickeyl 07org.oe.dev * r4d4a16da... 10/ (1 conf/machine/fic-gta01.conf conf/machine/fic-gta02.conf): fic-gta0[12].conf: always generate little endian jffs2 images Aug 04 22:51:19 aeh Aug 04 22:51:27 cia ware you borken up? Aug 04 23:15:48 koen: Still around? Aug 04 23:16:07 koen: If so, can you confirm your python verison Aug 04 23:55:14 nas200 pics: http://dyoung.org/nas200/ Aug 04 23:59:45 03mickeyl 07org.oe.dev * r4d4a16da... 10/ (1 conf/machine/fic-gta01.conf conf/machine/fic-gta02.conf): fic-gta0[12].conf: always generate little endian jffs2 images Aug 05 00:04:05 03mickeyl 07org.oe.dev * r4d4a16da... 10/ (1 conf/machine/fic-gta01.conf conf/machine/fic-gta02.conf): fic-gta0[12].conf: always generate little endian jffs2 images Aug 05 00:26:45 03mickeyl 07org.oe.dev * r4d4a16da... 10/ (1 conf/machine/fic-gta01.conf conf/machine/fic-gta02.conf): fic-gta0[12].conf: always generate little endian jffs2 images Aug 05 00:54:58 Hi folks, anybody online that can help with this? http://pastebin.com/d1412baf8 Aug 05 01:46:52 03mickeyl 07org.oe.dev * r4d4a16da... 10/ (1 conf/machine/fic-gta01.conf conf/machine/fic-gta02.conf): fic-gta0[12].conf: always generate little endian jffs2 images Aug 05 02:29:30 03mickeyl 07org.oe.dev * r4d4a16da... 10/ (1 conf/machine/fic-gta01.conf conf/machine/fic-gta02.conf): fic-gta0[12].conf: always generate little endian jffs2 images **** ENDING LOGGING AT Sun Aug 05 02:59:56 2007