**** BEGIN LOGGING AT Fri Jan 30 02:59:57 2009 Jan 30 08:22:11 I'm just debugging what is memleaking on my system, so I wonder if is it enough to monitor RSS(Resident Memory Size) in /proc/*/stat ? Jan 30 09:00:06 03Sergey Lapin  07ieee80215 * r147f5f8925 10openembedded.git/packages/libnl/ (files/fix-includes.patch libnl_1.1.bb): Jan 30 09:00:06 libnl build fix Jan 30 09:00:06 Signed-off-by: Sergey Lapin Jan 30 09:00:18 03Sergey Lapin  07ieee80215 * r68e81d1817 10openembedded.git/packages/ (3 files in 2 dirs): Jan 30 09:00:18 Revert "Merge branch 'ieee80215' of git@git.openembedded.net:openembedded into ieee80215" Jan 30 09:00:18 This reverts commit 7eee8a7d69244d34dfa8a7ba709a8e9faeb6c5b2. Jan 30 09:00:18 03Sergey Lapin  07ieee80215 * r7880fad12f 10openembedded.git/ (conf/checksums.ini packages/libsmi/libsmi_0.4.8.bb): checksums.ini: added checksum for libsmi Jan 30 09:00:21 03Sergey Lapin  07ieee80215 * r95899c3c46 10openembedded.git/packages/lowpan-utils/lowpan-utils_git.bb: Jan 30 09:00:24 lowpan-utils: added (IEEE802.15 userspace) Jan 30 09:00:26 Signed-off-by: Sergey Lapin Jan 30 09:00:28 03Sergey Lapin  07ieee80215 * r87497add6b 10openembedded.git/packages/wireshark/wireshark_1.0.5.bb: Jan 30 09:00:31 wireshark: added 1.0.5 Jan 30 09:00:33 Now both tshark and wireshark frontends are build in the Jan 30 09:00:35 same run. Packaging is more fine-grained. Jan 30 09:11:06 morning folks Jan 30 09:11:42 is there relevant performance data of embedded processors which have "java capabilites" in how far they benefit the application? Jan 30 09:11:53 or like some pointers where to look for such data? Jan 30 09:24:12 morning all Jan 30 09:27:47 <_boto> good morning Jan 30 09:28:57 morning Jan 30 09:36:04 hi, in the past i defined DISTRO_FEED_PREFIX/URI to generate customized feeds conf. this does no longer work now, does someone know how to do that now? Jan 30 09:38:36 morning pplz :) Jan 30 09:38:37 hi Jay7 Jan 30 09:38:43 thesing: hey :) Jan 30 09:39:06 thesing: I have no pull access to your git repo :) Jan 30 09:39:26 whats the error message? Jan 30 09:41:11 moirning Jan 30 09:41:32 grr.. can't switch to origin branch right now to check.. I have some uncommited changes Jan 30 09:42:01 thesing: what proto should I use? git://? Jan 30 09:42:17 thesing: is there any special authentication? Jan 30 09:42:35 hrw: morning :) Jan 30 09:43:50 Jay7: I have thesing@collie.linuxtogo.org:kexecboot.git in config as origin Jan 30 09:43:59 btw, good url about embedded linux developement http://free-electrons.com/docs/ Jan 30 09:44:19 hm.. I should check this Jan 30 09:44:25 you may need to point it to my home dir. Jan 30 09:44:44 whoa its a thesing :-) Jan 30 09:44:50 thesing: I've tried with ssh:// scheme but then I have no filesystem permissions Jan 30 09:46:08 thesing: long time no see Jan 30 09:47:52 XorA: I have to finish my thesis till monday, so I didn't (and don't) have much time. Jan 30 09:48:03 I forbid myself irc ;) Jan 30 09:48:16 thesing: ahah that explains all, I shall bug you next week then Jan 30 09:48:23 thesing: at FOSDEM? Jan 30 09:49:15 no. I have to defend the thesis near that time. Jan 30 09:49:30 thesing: fatal: The remote end hung up unexpectedly Jan 30 09:49:56 seems that I should write something about me from @git.linuxtogo.org or @collie.linuxtogo.org Jan 30 09:50:19 thesing: do you using git:// scheme in url? Jan 30 09:51:58 Jay7: git clone ssh://thesing@git.linuxtogo.org/home/thesing/kexecboot.git worked for me Jan 30 09:52:05 hehe.. Jan 30 09:52:29 thesing: with ssh:// I have no filesystem permissions :) Jan 30 09:52:35 you are owner Jan 30 09:53:09 yes but you have an account there. and the permissions should be right. Jan 30 09:53:19 I can fetch but can't push Jan 30 09:53:54 ah.. in beginning I've written 'pull'.. Jan 30 09:53:56 I mean push Jan 30 09:54:12 ok. the permissions are wrong. Jan 30 09:55:29 Jay7: try again Jan 30 09:57:57 seems working Jan 30 09:57:59 thanks Jan 30 09:58:30 I'll push my changes towards evening Jan 30 10:00:00 Jay7: there are some patches to psplash in oe that add support for 1 bit framebuffers. You could add these to kexecboot (since I used the psplash framebuffer code) Jan 30 10:00:41 thesing: hm.. yesterday we have moved fb_render to 16bit writes.. Jan 30 10:00:56 to fix tosa Jan 30 10:01:04 but I'll check that Jan 30 10:01:30 fb_render wasn't in pspash. they draw directly to the screen. Jan 30 10:01:37 Jay7: even 1 bit fb should divide by sixteen evenly Jan 30 10:01:40 ah, ok Jan 30 10:02:44 XorA: I have some time now because I do some benchmarks atm (have to type some numbers in a table every half minute ) Jan 30 10:02:48 is it psplash_grayscale.patch? Jan 30 10:03:17 thesing: just wondering what was the state of the collie SD driver, it seems non functional on any of my SD cards :-( Jan 30 10:05:11 Jay7: psplash_1bpp.patch I think. But I don't know if the greyscale patch is in. Jan 30 10:06:01 I have not seen that file in packages/psplash/files/* Jan 30 10:07:08 XorA: it works for all my cards. Aren't they recognized or do you only get fserrors sometimes? Jan 30 10:07:53 mmcblk0 error -22 read/write (or something similar to that) Jan 30 10:08:11 they start booting fine, and kexecboot works, but after init starts they fail Jan 30 10:10:38 maybe it helps if we limit the speed. Jan 30 10:12:33 My cards also give this error but the mmc subsystem retries and it succeeds later. Jan 30 10:14:28 thesing: I see it retry, never succeeds Jan 30 10:14:36 thesing: I shall have to poke the kernel then Jan 30 10:14:49 thesing: are you using high speed cards? Jan 30 10:17:02 no. And it shouldn't matter. The cards are used in spi-mode and thats slow for all cards. It can only got to 25MHz spi clock. Jan 30 10:17:15 hmm, wonder what the issue is Jan 30 10:19:00 thesing: I have plenty of the failing cards, remind me to post you one when your thesis is submitted Jan 30 10:19:09 I think its an issue of the mmc-spi driver and not the collie-spi driver. Jan 30 10:20:04 XorA: let's see how much time I have after that. My happy student life ends and I have to get a real job ... Jan 30 10:20:19 thesing: hehe, yeah, I need one of those too Jan 30 10:24:49 hmm, I wonder if OM kernel has any mmc-spi fixes Jan 30 10:25:36 XorA: btw. there is a driver for the buzzer now. Some guy wrote it for simpad. But I didn't test it yes. Jan 30 10:25:51 and it's an oss driver I think. Jan 30 10:26:19 thesing: should be easy enough to knock it about into alsa Jan 30 10:26:46 XorA: do you have enough free time and asoc experience? Jan 30 10:27:37 thesing: I have a little asoc experience :-) and possibly too much free time Jan 30 10:31:25 XorA: http://www.linuxtogo.org/~thesing/linux-2.6.24-SIMpad-ucb1x00-audio.patch the old ucb-audio driver is in the audio branch of my git repository at linuxtogo Jan 30 10:32:10 XorA: I think once the dma engine of SA1100 is supported by asoc the headphones should just work. Jan 30 10:33:22 thesing: cool Jan 30 10:34:48 thesing: ucb1x00 driver? Jan 30 10:35:48 hrw: yes. Jan 30 10:37:41 re Jan 30 10:40:59 hi florian Jan 30 10:41:38 hey pb_ Jan 30 11:15:45 03Sergey Lapin  07org.openembedded.dev * r2a50d92523 10openembedded.git/packages/libnl/ (files/fix-includes.patch libnl_1.1.bb): Jan 30 11:15:45 libnl build fix Jan 30 11:15:45 Signed-off-by: Sergey Lapin Jan 30 11:19:46 03Koen Kooi  07org.openembedded.dev * refb46ab0f4 10openembedded.git/packages/kexecboot/linux-kexecboot_2.6.21-hh20.bb: linux-kexecboot 2.6.21: do postprocessing on initramfs in do_configure_append Jan 30 11:19:57 03Koen Kooi  07org.openembedded.dev * rfc03e2be0b 10openembedded.git/classes/kernel.bbclass: kernel.bbclass: move initramfs stuff to configure so we can do postprocessing on it with do_configure_append Jan 30 11:31:17 what are uImage and xuImage kernel image types Jan 30 11:31:18 ? Jan 30 11:31:47 03Koen Kooi  07org.openembedded.dev * rc49f410e88 10openembedded.git/packages/kexecboot/ (3 files in 2 dirs): linux-kexecboot: make hx4700 work Jan 30 11:32:23 uImage is for u-boot, I think you are misreading the script for xuImage :-) Jan 30 11:34:07 xuImage is the right image type I think (it occures in many oe files) Jan 30 11:36:57 booxter: xuImage seems uImage + vmlinux (xreference) although I've never used it Jan 30 11:40:01 I try to build my uImage kernel. For this I applied patch with uImage make target that seems to be executed after built-in kernel parts are compiled (I see it in bitbake logs), then kernel modules are built, and then - wtf? - uImage is rebuilt again with some "default" (?) entry and load addresses - 0x20008000. WTF? Jan 30 11:42:50 it seems to use UBOOT_ENTRYPOINT and UBOOT_LOADADDRESS defined in classes/kernel.bbclass Jan 30 11:43:39 but I don't need these values. I want them to be generated automatically for every new kernel version. Jan 30 11:44:53 booxter: but those values are not dynamic Jan 30 11:45:10 XorA: I can generate them in my uImage makefile target Jan 30 11:46:29 LA=`$(OBJDUMP) -p vmlinux | sed -n 's/LOAD.*vaddr \([^ ]*\).*/\1/p'`; EN=`$(OBJDUMP) -f vmlinux | sed -n 's/.*start address \([^ ]*\).*/\1/p'`; mkimage -A mips -O linux -T kernel -C gzip -a $$LA -e $$EN -n "Linux Kernel Image" -d vmlinux.bin.gz uImage Jan 30 11:46:49 I parse vmlinux to find these addresses - successfully :) Jan 30 11:49:50 well submit a patch then :-) Jan 30 11:50:22 to completely remove this UBOOT_* vars?:) Jan 30 11:50:45 I think I should speak with some devs at first :) Jan 30 11:50:51 if you think you got a better way, then always supply a patch Jan 30 11:50:54 hrw: what'd you think about the idea? Jan 30 11:50:59 you are talking to a dev :-D Jan 30 11:51:16 ow man, ok, I'll look into it :) Jan 30 11:51:29 never assume we know everything :-D Jan 30 12:05:41 * Jay7 is away for some hours.. Jan 30 12:09:38 03Felix Domke  07org.openembedded.dreambox * re842caf60a 10openembedded.git/packages/tuxbox/tuxbox-common.bb: tuxbox-common: update to r11 of common data Jan 30 12:16:22 booxter: XorA is core dev ;) Jan 30 12:16:58 hrw: ok, i'm in an hour from sending the patch :) Jan 30 12:17:44 hi fgomez Jan 30 12:17:57 * XorA has just been AWOL a bit, but now Im back, shudder in pheer :-) Jan 30 12:18:03 it's a madness that for every kernel and platform you need to specify these addresses :) Jan 30 12:19:01 XorA: which pacakge in OE contains alsa state file? Jan 30 12:19:39 alsa-states or openmoko-states Jan 30 12:19:55 ok Jan 30 12:21:41 I need to buy small speakers Jan 30 12:21:47 booxter: yeah, agreed, it is silly to have to specify the same addresses in two places Jan 30 12:21:50 2x30mm ones Jan 30 12:22:45 booxter: but, even better than your plan would be to teach u-boot to load elf images directly rather than requiring this crazy uimage format, then it could just read the addresses from the elf header directly. Jan 30 12:23:57 pb_: that's not in my todo for this weekend :-D Jan 30 12:24:02 heh Jan 30 12:24:09 oh well :-} Jan 30 12:24:45 quick pb_ hack his todo list Jan 30 12:25:54 any thoughts on this: http://rafb.net/p/wq48NU15.html Jan 30 12:26:08 03Koen Kooi  07org.openembedded.dev * rc74e1d46b6 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Jan 30 12:26:18 03Koen Kooi  07org.openembedded.dev * r867cb32a4c 10openembedded.git/packages/pointercal/ (files/hx4700/pointercal pointercal_0.0.bb): pointercal: reinstate hx4700 pointercal Jan 30 12:26:19 03Henry von Tresckow  07org.openembedded.dev * r2466a29b60 10openembedded.git/ (3 files in 3 dirs): sg3-utils: add 1.24 Jan 30 12:27:34 Crofton: looks like a threading issue Jan 30 12:27:57 yeah Jan 30 12:28:06 I am suspecting a rebuild will work Jan 30 12:28:10 suggestions? Jan 30 12:28:39 no idea, I very very occasionally see craziness like that, but when you go back to look it just works Jan 30 12:29:34 yeah Jan 30 12:29:40 it has that sense to it Jan 30 12:29:52 this is my multi thread machone Jan 30 12:33:06 * hrw found new use for devboard Jan 30 12:33:16 playing music for children Jan 30 12:33:43 hehe Jan 30 12:34:12 hi hrw Jan 30 12:34:19 btw - someone know how to teach busybox ash (and/or minicom) to provide LINES/COLUMNS variables? Jan 30 12:34:46 I am connecting to devboard with minicom and would prefer screen fullsize instead of 80x25 Jan 30 12:36:01 hrw, my god-daughters like http://www.snap-circuits.com/phpstore/index.php Jan 30 12:37:13 Crofton|work: nice stuff Jan 30 12:37:39 but children music was on my pendrive which I plugged to board so thats why it was played Jan 30 12:40:46 root@at91sam9263ek:~# opkg install umbaumba Jan 30 12:40:46 An error ocurred, return value: -1092642420. Jan 30 12:40:47 wtf? Jan 30 12:40:50 opkg r197 Jan 30 12:40:57 hrw, it was a riot watching a 5 and 7 year old assemble a RC car and drive it around Jan 30 12:41:08 upkg update Jan 30 12:41:10 ? Jan 30 12:42:15 Crofton|work: I get this each time when ask for not existing package Jan 30 12:42:22 ah Jan 30 12:42:33 I think opkg is not so good at reporting errors Jan 30 12:43:07 03Marcin Juszkiewicz  07org.openembedded.dev * rebb79c1fe7 10openembedded.git/packages/alsa/ (alsa-state.bb alsa-state/at91sam9263ek/asound.state): alsa-state: added state for at91sam9263ek board Jan 30 12:49:18 03Koen Kooi  07org.openembedded.dev * r35f919236a 10openembedded.git/files/device_table-minimal.txt: device-table minimal: use increment feature instead of listing each entry Jan 30 12:51:24 hi Cliff Jan 30 12:52:48 hrw: Hello Marcin Jan 30 12:52:59 hrw: you make any progress with mono? Jan 30 12:57:27 cbrake: not yet Jan 30 13:18:19 03Ihar Hrachyshka  07org.openembedded.dev * reff32d62b9 10openembedded.git/ (4 files in 4 dirs): Jan 30 13:18:19 Reworked stb225 machine configuration file; build uImage for stb225. Jan 30 13:18:19 Signed-off-by: Ihar Hrachyshka Jan 30 13:27:03 03Graeme Gregory  07org.openembedded.dev * r9fba646c97 10openembedded.git/classes/ (package_deb.bbclass rootfs_deb.bbclass): package_deb.bbclass, rootfs_deb.bbclass : merge in poky changes to these classes that make sure correct files end up on rootfs Jan 30 13:44:24 03Koen Kooi  07org.openembedded.dev * r267aa2df85 10openembedded.git/packages/kexecboot/ (device_table-oldmmc.txt initramfs-kexecboot-image.bb): initramfs-kexecboot-image: specify custom device tables from hx4700 and h2200 since those are from the pre dynamic mmc dark ages Jan 30 13:48:51 03Graeme Gregory  07org.openembedded.dev * r32e4a25abe 10openembedded.git/classes/rootfs_deb.bbclass: rootfs_deb.bbclass : fix merge problems where I removed ${libdir} Jan 30 13:52:06 03Koen Kooi  07org.openembedded.dev * r86bb1d7cd7 10openembedded.git/ (2 files in 2 dirs): devicetables: when using autoincrement don't specify device number Jan 30 13:52:06 03Koen Kooi  07org.openembedded.dev * rbbb4f7c6e5 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Jan 30 13:53:26 should minimal-image start a shell on successfull bootup? Jan 30 14:06:33 hi zecke Jan 30 14:09:06 ho ho ho Jan 30 14:31:06 hi all Jan 30 14:31:25 booxter: should Jan 30 14:31:28 iirc Jan 30 14:31:52 I still can't get shell on my stb225 board. Stucked. Jan 30 14:33:33 dmesg shows that it initialised ttyS0 and ttyS1 for my serial UART device. I manually (!) created ttyS0 for my major:minors Jan 30 14:34:07 I specified SERIAL_CONSOLE ?= "115200 ttyS0" and got the following string in my inittab: Jan 30 14:34:27 S:2345:respawn:/sbin/getty 115200 ttyS0 Jan 30 14:35:32 but anyway I don't get any shell on my serial port though it works ok (I see kernel, userspace init messages, minicom reacts on my input - Enter, other symbols..) Jan 30 14:36:03 hrw: any suggestions? Jan 30 14:37:33 Ow... After some time (I went away from my dev machine for a while) it successfully started getty! What's the matter with this delay? Jan 30 14:38:04 firstboot setup? Jan 30 14:38:28 maybe, but there is no ssh setup step in it. Jan 30 14:39:21 i think i got it... i boot from my nfs, and maybe after dhcp lease my board changes its ip address, losts nfs mount - and bingo! - timeout until remount... Jan 30 14:40:39 booxter: use static ip then? Jan 30 14:41:38 hrw: ow, i got it. Or maybe I can start with ip=dhcp, not a static setup (i used it until now) Jan 30 15:00:53 03Felix Domke  07org.openembedded.dreambox * r62fc2b0eaf 10openembedded.git/packages/base-files/ (base-files/dm7025/fstab base-files_3.0.14.bb): base-files: fix fstab for dm7025 Jan 30 15:04:14 booxter: are you sure that your serial UART is named ttyS0? PLease show major:minors you created Jan 30 15:09:26 mckoan: I reapplied my tty UART patch for fixing the mess in generic stb225 linux kernel driver. That's what I have now: http://pastebin.com/m7ad271c0 Jan 30 15:09:28 and why you had to create them... Jan 30 15:17:28 booxter: IMHO you problem is not ttyS related Jan 30 15:20:55 anyone here with the openmoko devr crew? Jan 30 15:23:01 prpplague, they are probably just shy, I also think #openmoko exists Jan 30 15:24:28 hey all Jan 30 15:24:40 OE e.V. is now official Jan 30 15:24:51 Crofton|work: yea well, i've found that #openmoko doesn't seem populated with the most friendly or qualified devrs Jan 30 15:25:02 rschuster: e.V. ?? Jan 30 15:25:02 prpplague: you probably want #openmoko-cdevel Jan 30 15:25:03 papers from the court arrived me yesterday Jan 30 15:25:24 pb_: ahh dandy Jan 30 15:25:28 #openmoko is the user channel, generally only good for asking whether anybody has heard yet that android was released.# Jan 30 15:25:32 rschuster, are we official? Jan 30 15:25:43 pb_, rofl :) Jan 30 15:25:48 Crofton|work: we are :) Jan 30 15:26:16 right in time for next FOSDEM ;) Jan 30 15:26:20 I suggest a celebration is in order Jan 30 15:26:23 rschuster: what is e.V. ? Jan 30 15:26:44 very roughly, german version of a non-profit Jan 30 15:27:15 Didn't quite take a year :) Jan 30 15:27:22 Crofton|work: ahh Jan 30 15:27:24 rschuster, what next? Jan 30 15:27:49 What is the official name? Jan 30 15:28:07 hi prpplague Jan 30 15:28:41 OpenEmbedded e.V. Jan 30 15:28:49 and OE e.V. in short Jan 30 15:28:49 prpplague: there was also #openmoko-devel but with key so access was limited to openmoko coreteam + cooperating people Jan 30 15:28:52 both are valid Jan 30 15:29:20 mckoan: hrw: I managed to get shell and sent the patch to ML. Tnx for help. Jan 30 15:29:58 booxter: what was the problem? Jan 30 15:30:05 Crofton|work: next stop is getting the tax exemption privilege Jan 30 15:30:12 rschuster: übercool! Jan 30 15:30:13 ah Jan 30 15:30:32 so we can start shaking down companies benfiting from our work :) Jan 30 15:31:13 hehe! Jan 30 15:31:14 mckoan: I don't know exactly. I think in my board network driver which can drop some packets while on NFS. Jan 30 15:32:21 rschuster, how long do you think this will take? Jan 30 15:33:40 now.. who will write about e.V. to ML and on website? Jan 30 15:38:35 mwester: will you kindly find time to help us with the arm-bootloader thing and old-parameters? Was posted in oe-dev ML a couple of weeks ago... Jan 30 15:40:31 Crofton|work: No idea. I also don't know what needs to be done for this actually. Need to get discussed with florian, mickey and holger Jan 30 15:41:44 rschuster: hey cool - that's good news! Jan 30 15:48:39 rschuster: what is the benefit of having OE OE e.V.? Jan 30 15:51:57 mckoan: http://www.fsfeurope.org/projects/ftf/building-legal-infrastructure.en.html - the stuff below "Project Governance" Jan 30 15:53:24 think of administering stuff like marketing and/or training material Jan 30 15:53:56 or companies could now donate hw to OE e.V. instead of having to hand it out to individuals Jan 30 15:54:37 with tax exemption in place things get even better Jan 30 15:55:27 rschuster: I just found instructions about tax exemption... looks pretty straight forward. Jan 30 15:55:50 florian: would you mind preparing a letter? Jan 30 15:56:41 rschuster: Just answering your email :-) Jan 30 16:00:48 rschuster: and once companies donated hw or money to OE e.V., OE share among members? Or better can finance OE internal projects? Jan 30 16:01:56 mckoan: if money is donated it could be used to pay someones travel expenses for example Jan 30 16:03:19 mckoan: hw is already in project's hands: http://wiki.openembedded.net/index.php/ProjectDevices Jan 30 16:05:19 rschuster: so money (or other stuff) donated in case can't be used for OE development, is it correct? Jan 30 16:07:33 mckoan: not sure. IMO it would be better if any of the OE developers is hired by such a company for a specific job Jan 30 16:12:47 rschuster: I understand, so OE e.V. purpose is 'only' Project governance and Copyright consolidation Jan 30 16:14:31 rschuster: but this do have a cost, how OE plan to uphold it? Jan 30 16:15:03 mckoan: donations? :) Jan 30 16:15:55 bbl Jan 30 16:16:19 florian: of course, but donations are untrustworthy Jan 30 16:17:10 mckoan: we'll see if we can do it in this way. if not we have the option to decide to introduce membership fees. Jan 30 16:20:01 florian: you are right. I don't want to be pessimistic, I simply want to consider various aspects of this matter. Jan 30 16:21:04 mckoan: Right now it is quite easy since the only thing we need to have is a little bit of time to organise things. Jan 30 16:22:44 florian: time? it's rarer than gold! ;-) Jan 30 16:28:21 mckoan: indeed... that's one onf the disadvantages right now. Jan 30 16:28:59 mckoan: but if we find some people who are not developers but are interested in helping oe in this way things could become much better Jan 30 16:32:58 bye Jan 30 16:36:06 hrw|gone: doh, sorry was in a meeting Jan 30 16:36:10 hrw|gone: later bud Jan 30 16:37:18 pb_: yea there is a minor bug in the s3cmci.c i wanted to let them know about, but i can't seem to get in contact with anyone there at openmoko Jan 30 16:41:45 prpplague: doh Jan 30 16:43:05 mickey|dinner ? Jan 30 16:59:57 hi Jan 30 17:06:40 thorsten1: hi Jan 30 17:12:56 how do you force rebuild of a single package? I need to rebuild kernel? Jan 30 17:13:03 how do you force rebuild of a single package? I need to rebuild kernel Jan 30 17:13:48 bitbake -c rebuild linux Jan 30 17:14:07 bitbake virtual/kernel -c rebuild Jan 30 17:14:10 bitbake -cclean linux is the same? Jan 30 17:14:30 rebuild does clean and build Jan 30 17:14:31 :P Jan 30 17:14:50 thx Jan 30 17:16:01 Jay7: kexecboot is taking over the world :-) Jan 30 17:17:01 what's diff between bitbake -crebuild linux or virtual/kernel ? Jan 30 17:17:20 personally, I do bitbake linux-omap Jan 30 17:17:36 depends if your virtual/kernel is linux or not Jan 30 17:17:46 I think virtual/kernel resolves to the proper bb file, which I do "manually" Jan 30 17:18:31 can anyone point me to an example recipe that uses mercurial SCM? if I understand the bitbake/lib/bb/fetch/hg.py SRC_URI would have a protocol=hg but I don't see that anywhere (unless I'm blind) Jan 30 17:18:31 XorA: what other kernel could be? Jan 30 17:18:56 mckoan: there are hundreds in package/linux Jan 30 17:20:13 XorA: but in package/linux there are only linux kernels Jan 30 17:20:55 hi, all! Jan 30 17:21:07 mckoan: but they are not all called linux Jan 30 17:22:11 I try to run qemux86 and have mouse problems with kdrive, lik it is being at various offset from host mouse and it does not allow access to all screen controls Jan 30 17:22:30 is there some solution out there to fix this problem? Jan 30 17:23:27 who broke e-image :-) Jan 30 17:25:28 XorA: I see, thx Jan 30 17:37:18 anyone every try to build wxpython with OE? Jan 30 17:38:31 Crofton|work: never tied, but did you try wxWidgets only with OE? Jan 30 17:38:41 s/tied/tried Jan 30 17:38:49 I have some code that wants wxpython Jan 30 17:39:23 Crofton|work: I don't know, sorry Jan 30 17:39:42 Crofton|work: I asked because I need to test wxWidgets soon Jan 30 17:40:05 I suppose I may look at it, but it will be after fosdem :) Jan 30 17:40:20 Crofton|work: BTW wxPython needs wxWidgets Jan 30 17:40:34 right :) Jan 30 17:41:06 re Jan 30 17:44:41 wxwidgets is a PITA in OE Jan 30 17:44:53 how bad? Jan 30 17:44:58 what's pita? Jan 30 17:44:58 its got one crazy ass build syste, Jan 30 17:45:10 but I think the recipe in OE built last I tried it Jan 30 17:45:48 yeah Jan 30 17:46:07 I try wxpython once, and had no luck, but that might have been my fault Jan 30 17:46:29 heh, there is only one person to blame for everything with python in the name :-) Jan 30 17:50:18 XorA: I call it 'empire of kexecboot' :)) Jan 30 17:50:47 now to get it running on a pandora :-) Jan 30 17:51:32 and could be even cooler to build in ability to flash UBI/JFFS2/YAFS partitions from kexecboot :- Jan 30 17:51:35 :-) Jan 30 17:53:12 but thats feature creep :-) Jan 30 17:54:24 I've thinking about installer/flasher ;) Jan 30 17:57:31 Jay7: I think this should be a different thing. you could boot a kernel with an installer-initramfs with kexecboot Jan 30 17:58:14 thesing: yes, me and Andrea decide the same Jan 30 17:59:45 thesing: well done for shooting down the feature creep :-) Jan 30 18:00:12 I don't even know what the current new features are. Jan 30 18:00:25 Jay7: what did you change when I was away ? Jan 30 18:01:02 http://linuxtogo.org/gowiki/AngstromKexecboot Jan 30 18:01:17 there is some todo and wishlist Jan 30 18:01:42 seems that Andrea already do some updates :) Jan 30 18:01:54 XorA: do you have an openmoko phone? if kexecboot would run on it, some guys from om community might be interested. Jan 30 18:02:17 thesing: I have 4 of them, Ill see if I can work up motivation Jan 30 18:02:28 thesing: good thing is the FreeRunner has an 8M kernel partition Jan 30 18:03:01 hehe.. Jan 30 18:03:14 XorA: send me /proc/cpuinfo from FreeRunner too then :) Jan 30 18:03:30 Jay7: have to boot it sometime Jan 30 18:05:56 I'm not sure if the configfile in /boot is a good idea. with it you have to worry about packaging it. Jan 30 18:06:23 It should at least be optional so we can boot distros with kexecboot that don't know about it. Jan 30 18:06:35 thesing: Omegamoon needed it in zubuntu Jan 30 18:07:15 forget what I said. The kernel needs to be placed in /boot and a cmdline anyway. Jan 30 18:07:35 So the distro has to know about kexecboot. Doesn't hurt to have a configfile. Jan 30 18:08:06 distro can have no config file Jan 30 18:08:13 then it will work as before Jan 30 18:08:17 the question is from which device do you use the configfile if more then one are present? Jan 30 18:09:05 thesing: from last atm Jan 30 18:09:17 timeout should go into NAND Jan 30 18:09:27 may be in 0.6 Jan 30 18:10:21 in pure nand or a file in a partition? Jan 30 18:10:37 pure nand imho Jan 30 18:10:52 but there can be some problems with other devices.. Jan 30 18:11:15 we will read/write mtdparts info into nand as sharp/cacko installer does Jan 30 18:14:12 thesing: I see that timeout/default/orded seems 'broken by design' Jan 30 18:14:31 but I see no better solution now Jan 30 18:15:33 s/orded/order/ Jan 30 18:16:38 we could use some part of flash from the kernel partition that we don't need. Jan 30 18:17:21 thesing: we can use 'parameters area' too Jan 30 18:17:34 partitions info is stored there Jan 30 18:17:46 there is some reserved areas Jan 30 18:18:02 we can reuse 'em Jan 30 18:18:03 But the configfiles seem to be easiest way. They could have a priority value inside. Jan 30 18:18:05 should just well document a searched for set of paths for config with priorities Jan 30 18:18:21 thesing: yes, ORDER is the one Jan 30 18:18:30 so jffs2 then SD then CF then USB then NFS Jan 30 18:18:48 XorA: in order from /proc/partitions atm Jan 30 18:19:04 Jay7: which is effectively random Jan 30 18:19:10 yeah.. Jan 30 18:21:25 Koen reports about mmc detections problems on his ipaq too Jan 30 18:21:50 so we can move to event-driven device scanning/adding to list.. Jan 30 18:22:06 scary move :) Jan 30 18:22:26 can probably steal code for that from mdev Jan 30 18:22:31 I don't think this is necessary. Jan 30 18:22:46 We should add code to debug this problem. Jan 30 18:23:09 thesing: I hope too.. but sleep does not help Jan 30 18:23:38 thesing: device just not in /proc/partitions at moment of reading Jan 30 18:24:00 do we have a reread option? Jan 30 18:24:18 we have very fast reboot option instead :) Jan 30 18:25:07 btw we can try to use inotify on /proc/partitions Jan 30 18:25:16 I think user triggered rescan could solve the problem. Jan 30 18:25:37 and make using usb sticks more comfortable. Jan 30 18:25:58 hm.. let's try :) Jan 30 18:26:05 do we create device node dynamically? Jan 30 18:26:33 atm all devices is precreated with device-table-* Jan 30 18:26:49 yes I remember ;) Jan 30 18:26:57 i can do mknod() according to /proc/partitions Jan 30 18:27:34 that's what I thought. Jan 30 18:27:46 I was lazy back then ;) Jan 30 18:28:09 ok, we thinking similar :) Jan 30 18:28:40 will mark in TODO Jan 30 18:29:22 btw Jan 30 18:29:26 * XorA leaves for dinner Jan 30 18:30:04 thesing: about psplash_1bpp.patch.. where I can find it? Jan 30 18:30:31 isn't it in oe/packages/psplash/files ? Jan 30 18:31:24 I should update my oe-tree first. Jan 30 18:31:37 maybe the greyscale patch is the successor. Jan 30 18:32:51 thesing: can you check this? Jan 30 18:33:56 It is. Jan 30 18:34:17 ok, I'll apply this to our fb.c Jan 30 18:35:05 don't forget to rename the new psplash_fb_foo function. Jan 30 18:35:28 03Koen Kooi  07org.openembedded.dev * r2a288bbc36 10openembedded.git/conf/machine/ (h2200.conf hx4700.conf): hx4700, h2200: band-aid device table to account for the prehistoric .21 kernel Jan 30 18:35:29 we don't want to make it too obvious where we stole the code from ;) Jan 30 18:35:36 :) Jan 30 18:38:48 I'm doing some refactoring now Jan 30 18:38:56 util.c/h added Jan 30 18:39:50 ant__: wb :) Jan 30 18:39:57 hey Jan 30 18:41:47 I have a question.. in init-mode on other hardware can eventif be not "/dev/event0"? Jan 30 18:42:15 should this parameter be hardware dependend? Jan 30 18:42:37 Jay7: Zubuntu lacks /dev/event0, I think is oe/kexecboot specific Jan 30 18:43:04 ant__: which device used in zubuntu? Jan 30 18:43:19 iirc /dev/input/event0 Jan 30 18:43:29 hm.. Jan 30 18:43:33 I have tarball here..1mom Jan 30 18:44:00 Jay7: I created /dev/event0 because I couldn't get the device-tables.txt create /dev/input/event0 for me ... Jan 30 18:44:21 hi thesing Jan 30 18:44:26 ant__: hi. Jan 30 18:44:36 someone told us udev creates /dev/init/eventX Jan 30 18:45:17 Jay7: nothing in /dev (Zubuntu) Jan 30 18:45:25 is a dynamic thing then Jan 30 18:45:40 strange.. Jan 30 18:45:51 anyway why use kexecboot as stand-alone? Jan 30 18:45:53 yes as everything in dev is in these days. Jan 30 18:46:18 seems that zubuntu run udev/mdev before kexecboot Jan 30 18:46:35 maybe as a rebootmanager? Jan 30 18:46:48 don't be confused..I'm talking about the rootfs Jan 30 18:47:03 Omegamoon used poky's kernel Jan 30 18:47:07 as base Jan 30 18:47:10 btw, I can create /dev/event0 by mknod() too :) Jan 30 18:47:14 nothing more Jan 30 18:47:27 Jay7: I did :-) Jan 30 18:47:39 then it will be the same on all platforms Jan 30 18:47:40 but the point is..why kexecboot and not reboot? Jan 30 18:48:06 I'm unsure the devices are unmounted cleanly... Jan 30 18:48:16 ant__: because you wan't a graphical rebootmanager. Jan 30 18:48:27 ant__: no kexec unmounts them. Jan 30 18:48:40 ok.. Jan 30 18:49:06 well, one could just 'test' kexecboot whithout having it as bootloader Jan 30 18:49:09 some of next stages is to create libkexec and link to it :) Jan 30 18:49:38 *g* Jan 30 18:49:38 ala crunchgen or busybox :) Jan 30 18:50:12 ah, what about the mini-logo? Jan 30 18:50:23 as patch in OE or in ltg? Jan 30 18:50:35 btw, about logo's in general.. Jan 30 18:50:48 what graphical format is easier to parse now? Jan 30 18:51:13 I can remeber how to parse pcx but it is rare now :) Jan 30 18:51:39 no idea. Jan 30 18:52:18 what format uses kernel for splash? Jan 30 18:52:23 well, the idea is making it using distro-overrides like for kernel bootlogo Jan 30 18:52:31 Jay7: clut224.ppm Jan 30 18:53:05 well, I'll check for ppm parser then :) Jan 30 18:53:16 look in kernel, scripts Jan 30 18:53:20 bbl Jan 30 18:54:36 If we keep getting new features at this rate we need to make them compile-time-configurable. Jan 30 18:57:13 that is good idea but I'm not autotools guru :) Jan 30 18:57:37 I can try do add #ifdef's around code :) Jan 30 19:07:33 03Koen Kooi  07org.openembedded.dev * raf7f93c01d 10openembedded.git/packages/bonjour/ (bonjour_107.6.bb files/bonjour.conf files/bonjour.init): bonjour: add 107.6 (from gumstix) Jan 30 19:17:20 03Koen Kooi  07org.openembedded.dev * rf874f88548 10openembedded.git/conf/checksums.ini: checksums: add bonjour Jan 30 19:21:15 join #beagle Jan 30 19:40:56 hi woglinde Jan 30 19:41:16 hi mrdata Jan 30 19:41:57 for now, i play with skeypad sl and kernel2624 Jan 30 19:45:33 mrdata hehe Jan 30 19:45:39 I got 2.6.29 booting Jan 30 19:45:54 but didnt test suspend/resume yet Jan 30 19:48:01 woglinde: bäh Jan 30 19:48:01 re florian Jan 30 19:48:58 woglinde 2.6.29-rc3 ? Jan 30 19:50:52 03Koen Kooi  07org.openembedded.dev * r0246f0a44c 10openembedded.git/packages/bonjour/ (bonjour_107.6.bb files/bonjour.conf): bonjour: make the config file more generic, fix GNU_HASH QA problems Jan 30 19:54:18 mrdata ups sorry 2.6.28 Jan 30 19:54:41 hi gnutoo Jan 30 19:54:50 woglinde, hi Jan 30 19:55:04 woglinde, did you see the big buck bunny recipees...lol Jan 30 19:55:05 woglinde okay, thats fine Jan 30 19:56:06 woglinde, i was away for a while because of exams and data-loss(i retstore my >=6 month old backup tomorrow) Jan 30 20:03:28 woglinde, I hope I still have my ssh key inside my old backup... Jan 30 20:03:58 gnutoo hm Jan 30 20:04:05 otherwise you can send mickeyl a new one Jan 30 20:04:16 woglinde, ok Jan 30 20:47:01 03Koen Kooi  07org.openembedded.dev * r82bc1c85b8 10openembedded.git/packages/gnome/gnome-icon-theme_2.24.0.bb: gnome-icon-theme: update icon cache in postinst Jan 30 21:17:25 Jay7: wb Jan 30 21:18:42 ant__: you too :) Jan 30 21:18:55 refactoring is close to end Jan 30 21:19:07 ok, I'd look at the mini-logo Jan 30 21:19:15 just resizingthe big ones Jan 30 21:19:41 hi ant Jan 30 21:19:45 ant__: btw, customizing that logo according to distro is good idea.. Jan 30 21:19:49 hi woglinde Jan 30 21:20:38 we can replace logo-img.h with one from SRC_URI Jan 30 21:20:41 woglinde: we would need to know FB_ANGLE of simpad :-) Jan 30 21:20:49 hm Jan 30 21:20:57 where do I have to look up? Jan 30 21:21:00 hehe... I have #define KXB_FBANGLE Jan 30 21:21:39 woglinde: do you add some rotation at all? Jan 30 21:21:50 ant__: you can add EXTRA_OEMAKE += "KXB_FBANGLE=" Jan 30 21:21:56 opie can rotate Jan 30 21:22:19 ant__: KXB_EVENTIF too :) Jan 30 21:22:29 woglinde: ok, let say, is 0,0 on top-left corner? Jan 30 21:22:43 hm Jan 30 21:22:44 can be Jan 30 21:23:19 here is wrong :-) http://www.flickr.com/photos/koenkooi/3238253191/ Jan 30 21:23:42 you see what I mean? Jan 30 21:25:37 hm this would be right on the simpad Jan 30 21:25:46 Jay7: btw look at the (steady) curor in the image there Jan 30 21:25:53 *cursor Jan 30 21:26:06 wow Jan 30 21:26:50 how is on akita? cursor blinking on 0,0 only during bootlogo? Jan 30 21:27:05 in this screen I don't have any cursor on c7x0 Jan 30 21:27:07 hello, where can i find information about supporting a lcd? Jan 30 21:27:50 it is run with windowsce, but i want to run the system with linux Jan 30 21:28:05 ant__: depends on console.. Jan 30 21:28:10 woglinde: I suspect the simpad has 0 rotation (like c7x0) Jan 30 21:28:31 ant__: let's decide about consoles.. what should be master console? Jan 30 21:28:34 it has a PXA270 intel processor and the lcd is an optrex Jan 30 21:28:38 Jay7: he Jan 30 21:28:53 I'm for stderr output on ttyS0 Jan 30 21:29:10 and not on tty1 Jan 30 21:29:39 but swapping console brings the cursor back on framebuffer... Jan 30 21:29:47 at least on c7x0 Jan 30 21:29:50 brolin: we don't know :) try ask later/in workdays Jan 30 21:30:40 Jay7: ok, i will come back to ask Jan 30 21:31:18 brolin I suggest first you try the tool haret(its a windowstool) to gether some more infos Jan 30 21:31:47 and If I am not complety wrong you can boot a linuxkernel with haret too Jan 30 21:32:56 eh, woglinde thanks for the clue Jan 30 21:33:03 brolin: you need to know a number of settings about the lcd to configure the pxa framebuffer driver correctly Jan 30 21:33:13 i'm gonna check Jan 30 21:33:24 with some kernels you can pass the settings as command line args at boot time Jan 30 21:34:47 i can read these from the datasheet scruggs, do you remember which data struct to check Jan 30 21:34:53 ? Jan 30 21:35:26 I think I have a note handy... will check Jan 30 21:37:25 haret looks good Jan 30 21:37:54 ant__: I'll digg into console tools.. Jan 30 21:38:18 noblink? Jan 30 21:38:46 i'm a newbie, but you will see me here more frecuently Jan 30 21:39:02 Jay7: uh... noblink is 1995 ! Jan 30 21:39:16 brolin: this is the patch that gumstix uses for the verdex (pxa270 based) Jan 30 21:39:24 ant__: yep.. can it be applied to current kernels? Jan 30 21:39:33 most of it is setup in the machine file, but you'll get the idea Jan 30 21:40:00 Jay7: didn't try yet...unsure about how useful Jan 30 21:40:35 yes, all info is util Jan 30 21:40:57 useful Jan 30 21:41:30 english is not my native language, I think tha you note this ;D Jan 30 21:46:26 scruggs: where is the patch? or i don't understand Jan 30 21:47:02 brolin: sorry, looks like I didn't paste the link! here it it http://pastebin.com/m32fd0833 Jan 30 21:48:00 ok, thanks Jan 30 21:48:01 starting with the 2.6.19 kernel you can't specify the framebuffer settings on the kernel command line, so you'll need to do something similar to what the patch does Jan 30 21:48:03 gdk-pixbuf-csource --macros logo.png Jan 30 21:48:07 oops sorry Jan 30 21:48:31 you really help me, good path to start Jan 30 21:57:19 ant__: is tosa fb angle 0? Jan 30 21:57:24 yes Jan 30 21:57:37 I'm almost finished with mini-logos :-) Jan 30 21:58:10 just dubious about Gimp interpolation ;-) Jan 30 21:58:22 I'm unsure about Poodle, Corgi and Collie fb angle Jan 30 21:58:29 now it is 270 Jan 30 21:58:34 collie should be 270.... Jan 30 21:58:45 thesing thought c7x0 was the exception Jan 30 21:59:04 please set 0 as default :-) Jan 30 21:59:07 shepherd, husky and tosa is 0 now Jan 30 21:59:15 yes I've do already Jan 30 21:59:54 what about poodle and corgi? :) Jan 30 22:00:08 well, corgi is like c7x0 Jan 30 22:00:13 so 0 Jan 30 22:00:49 poodle..dunno...probably 270 Jan 30 22:01:45 hm.. I have forgot Terrier :)