**** BEGIN LOGGING AT Tue Apr 25 09:59:57 2006 Apr 25 10:00:12 RP: Someone told me that some months ago, and claimed they used a builtin one Apr 25 10:00:30 that's the extent I know about a builtin mic Apr 25 10:00:41 luke-jr_: As I told you last time, it does not have one. Some devices like poodle do. The c760 does not. Apr 25 10:00:58 ok, I didn't expect it to Apr 25 10:01:33 luke-jr_: You can connect a mic or headset (mic + speaker) to the headphone jack Apr 25 10:01:41 RP: that's what I'd like to work, yes Apr 25 10:01:46 <_law_> does the current .dev gpe-image build? Apr 25 10:02:54 luke-jr_: The device can cope with a mic or a mono line in source - the difference is in software Apr 25 10:03:14 _law_: I built a gpe image for poodle recently Apr 25 10:03:32 <_law_> RP, with gcc 4.0.2 ? Apr 25 10:03:48 <_law_> i currently build one for my 760 Apr 25 10:03:51 _law_: No :) Apr 25 10:03:57 RP: I know; that's why I was checking on the kernel software status for it ;) Apr 25 10:04:01 hey zecke Apr 25 10:06:40 hi zecke Apr 25 10:07:10 RP: do you plan to fix the bug spotted by lrg? Apr 25 10:07:14 <_law_> whats the recommend gcc version ? Apr 25 10:07:26 RP: or is your limited time spend on something else? Apr 25 10:07:51 zecke: I'll fix it, I'm just trying to doo too many things at once atm :-/ Apr 25 10:08:02 wow Apr 25 10:08:08 my monotone diff is huge Apr 25 10:08:08 RP: I can look into it (now) as well Apr 25 10:08:23 zecke: Feel free to fix it :) Apr 25 10:11:06 koen: where did you get the OE, TI mexico news from? Apr 25 10:11:57 meh, i can't get irattach to work Apr 25 10:12:41 zecke: Sandino Flores Moreno aka tigrux Apr 25 10:12:44 ~seen tigrux Apr 25 10:12:53 tigrux was last seen on IRC in channel #handhelds.org, 6h 53m 3s ago, saying: 'Leeds: Because I should have thought that.'. Apr 25 10:13:23 mreimer: I fixed the h2200 jffs2 problem Apr 25 10:14:50 hey obergix[home] Apr 25 10:15:45 03koen 07org.oe.dev * rcffd8c3e... 10/packages/udev/udev_089.bb: udev 089: add links.conf Apr 25 10:15:50 zecke: good morning Apr 25 10:15:54 koen: what monotone diff? Apr 25 10:15:56 pb__: hey? Apr 25 10:16:01 03koen 07org.oe.dev * rcffd8c3e... 10/packages/udev/udev_089.bb: udev 089: add links.conf Apr 25 10:16:10 03koen 07org.oe.dev * r93e8c9aa... 10/packages/gtkmm/gtkmm_2.4.8.bb: gtkmm: ship headers in 2.4.8 dev package Apr 25 10:16:13 hey pb__ Apr 25 10:16:13 03koen 07org.oe.dev * r6bf3edc1... 10/packages/gnome/orbit2_2.13.2.bb: orbit2: disable parallel make Apr 25 10:16:15 zecke: working copy <> OE.db Apr 25 10:16:16 hi lrg Apr 25 10:16:17 03koen 07org.oe.dev * r7bea1481... 10/conf/distro/preferred-gpe-versions-2.8.inc: gpe 2.8: prefer gaim 2.00-mco1 Apr 25 10:16:20 hi Apr 25 10:16:21 03koen 07org.oe.dev * rbf6be079... 10/conf/machine/h2200.conf: Apr 25 10:16:21 h2200.conf: fix to support newer mtdutils: Apr 25 10:16:21 mkfs.jffs2: unrecognized option `--no-cleanmarkers' Apr 25 10:16:36 zecke: but it's a lot smaller now :) Apr 25 10:17:28 koen: I just compared poky to OE - 250k+ of diffs :-/ Apr 25 10:17:28 lrg: do you know if anybody is already working on an ASoC data pump driver for the s3c2410? One of our guys has just started updating our kernel bits to 2.6 and he's in the process of trying to get ALSA sound working. Apr 25 10:18:13 RP: that was to be expected Apr 25 10:18:51 RP: 4 files changed, 170 insertions(+), 1 deletion(-) Apr 25 10:18:56 RP: packages/linux/ Apr 25 10:19:18 koen: I know. Its mostly changes in OE as poky hasn't changed much. Its just depressing :-/ Apr 25 10:19:41 hrw|work: What's that with? Apr 25 10:20:17 RP: and mallum kept saying "it's just a temporary hack" Apr 25 10:20:37 RP: my hostap changes Apr 25 10:20:48 RP: you need svk to do the merges? Apr 25 10:20:52 hi, what the meaning of those info after build ?? Apr 25 10:20:53 Packaged contents of pthreads into /home/celf/netspectrum/celf/OE-20060305_omap5912osk_familiar-unstable_11_20060424/build/tmp/deploy/ipk/pthreads_0.1-r0_armv5te.ipk Apr 25 10:20:53 NOTE: Not creating empty archive for pthreads-doc-0.1-r0 Apr 25 10:20:53 NOTE: Not creating empty archive for pthreads-dev-0.1-r0 Apr 25 10:20:53 NOTE: Not creating empty archive for pthreads-locale-0.1-r0 Apr 25 10:21:06 celf? Apr 25 10:21:14 zekoZeko: the toaster guys Apr 25 10:21:16 ehm Apr 25 10:21:22 s/zekozeko/zecke/ Apr 25 10:21:27 minipanda: during package install Apr 25 10:21:31 why create something addtional like -doc? -dev -locale? Apr 25 10:21:54 why not? Apr 25 10:21:56 minipanda: e.g. -dev if you want to do development on your host or create an sdk Apr 25 10:22:16 minipanda: -locale if your application should have another locale/translation than bangladesh Apr 25 10:22:25 minipanda: -doc if you want to have man pages installed Apr 25 10:22:30 pb_, I think fluff is working on this. He asked for the latest code about a month ago I don't know of the progress though Apr 25 10:23:04 zecke: is that inherited from gentoo? never heard of that before Apr 25 10:23:10 minipanda: most of the stuff won't be needed for lets say a GUI less device without much storage Apr 25 10:23:14 lrg: okay, thanks Apr 25 10:23:20 minipanda: not that I know Apr 25 10:23:30 pb_: what codec to you have btw ? Apr 25 10:23:37 minipanda: -dev packages are pretty common on debian as well Apr 25 10:23:45 zecke: koen why call celf as toaster? Apr 25 10:24:03 minipanda debian does this for a long time Apr 25 10:24:04 minipanda: a toaster is consumer electronics, right? Apr 25 10:24:12 hehe Apr 25 10:24:27 lrg: mostly wm8721 right now, though we have a couple of boards with Philips and MOS Analog parts. Apr 25 10:24:28 woglinde: debian doesn't go as far as -locale and -doc alot of the time Apr 25 10:24:39 and gentoo has no packages ;) Apr 25 10:24:44 koen: well, sorry for my lack of american-english vocaulury :( Apr 25 10:24:50 mithro sure? Apr 25 10:25:02 koen: gentoo can generate binary packages too Apr 25 10:25:03 mithro look at the xorg transition Apr 25 10:25:21 pb__: btw, i think it will be fairly trivial to port the wm8731 driver to wm8721 Apr 25 10:25:22 minipanda: the point is to give a Distribution Maker a granular choiche on the to be installed stuff Apr 25 10:25:33 koen: Did mallum say what the long term aim was? :) Apr 25 10:25:35 minipanda: maybe every bit counts ;) Apr 25 10:25:39 lrg: yeah, I hope so. the 8731 is basically a superset of the 8721, right? Apr 25 10:25:50 hm suse goes the other way Apr 25 10:25:59 lrg: we also have the 8711 on one board, which I'm guessing is a similar deal Apr 25 10:26:01 yast-core depends on wireless tools Apr 25 10:26:02 woglinde: are you done with working yet? Apr 25 10:26:04 pb__: i think so Apr 25 10:26:10 woglinde: dev documentation is generally seperate, but most things ship with there readme, manual, help-text etc in the core package i've found Apr 25 10:26:11 woglinde: I'm getting hungry, and I need to go the copy shop Apr 25 10:26:12 * lrg checks Apr 25 10:26:18 zecke around 12:50 Apr 25 10:26:34 ETA 12:50, *prepares self-shooting traps* Apr 25 10:26:38 * minipanda spoke :" i use gentoo, pilgrim ... " Apr 25 10:26:46 zecke lets meet at 13:05 at the "china mann" Apr 25 10:27:16 citywok (shitty wok) Apr 25 10:27:22 oh, China mann? #!$!#!!@#$&^%(*^% ... where is it? Apr 25 10:27:39 minipanda its dahlem dorf Apr 25 10:27:47 minipanda: a food court next to our university Apr 25 10:28:04 minipanda: cheap and plenty of food, tasting quite good Apr 25 10:28:44 ahha, it's 3 AM in East Coast? Apr 25 10:28:45 pb__: your right, the codecs are all from the same family and should be straight forward to port using the 31 driver. Apr 25 10:29:07 minipanda: We are near Elb-Coast (Berlin/Germany/Europe) Apr 25 10:29:31 minipanda: was your question answered, or do you have further questions on how the magic works? Apr 25 10:30:05 minipanda: US east coast? no, it's about 0640 there. Apr 25 10:30:12 zecke: oh! great, you speak english quite good ... really Apr 25 10:30:17 zecke: what magic you mean? Apr 25 10:32:33 minipanda: how it creates packages, and decides which files to put into the packages Apr 25 10:32:48 zecke: no, i still confused ... Apr 25 10:32:54 zecke: it's toooooo tricky ... Apr 25 10:33:59 pb__: where are you? Apr 25 10:34:04 zecke: you eat paper? Apr 25 10:34:30 zecke: at my office Apr 25 10:34:52 mithro: did I claim that by accident? Apr 25 10:35:13 some could interperate that to mean so :P [20:06] woglinde: I'm getting hungry, and I need to go the copy shop Apr 25 10:35:30 zecke: for some reason, i want my test package run automatically or run it one by one Apr 25 10:35:30 mithro: I used to eat paper ;) Apr 25 10:35:34 mithro: it is a habbit Apr 25 10:35:41 zecke: so is that mean i need to patch /etc/rc.S/ or something Apr 25 10:36:03 minipanda: inherit update-rc.d (or similiar) Apr 25 10:36:20 minipanda: you can say which start script to run at which levels (and when to stop them) Apr 25 10:37:18 zecke: well, the question is when i installl all the test package to same '/root', they will not replace each others modificaiton again update-rc.d? Apr 25 10:38:12 minipanda: oh well, they can#t have the same file names :} Apr 25 10:38:15 minipanda: ipkg won# Apr 25 10:38:24 won't allow to install the packages then Apr 25 10:38:26 so bye alle Apr 25 10:38:51 RP: is pxaficp-ir the IR chipset driver? Apr 25 10:38:56 zecke: ok, when two ipkg want to modify a same configuration file, what will happened then? Apr 25 10:39:04 CoreDump|home: yes Apr 25 10:39:08 hmm Apr 25 10:39:14 minipanda: modify a file is a different story ;) Apr 25 10:39:15 e.g. auto-test.sh is unique in whole system Apr 25 10:39:28 I can't get irattach to work, always complaining "no such device" Apr 25 10:39:36 CoreDump|home: which device? Apr 25 10:39:42 anyway time for food Apr 25 10:39:47 that it doesn't say =) Apr 25 10:40:25 http://openembedded.pastebin.com/680714 Apr 25 10:41:07 do you need irattach at all Apr 25 10:41:10 anyway cya later Apr 25 10:41:40 hmm Apr 25 10:41:49 i think so yeah Apr 25 10:42:19 cya Apr 25 10:43:36 morning Apr 25 10:44:46 hi chouimat Apr 25 10:45:31 hmm.. xserver-kdrive/tosa is armv5te (other Z has it machine depend). matchbox-wm does not have poodle mappings.. Apr 25 10:49:18 I don't think xserver-kdrive should be machine dependent at all. Apr 25 10:49:35 What are the other zauruses doing that requires it to be specific? Apr 25 10:50:10 check the .bb and it will be become crystal clear Apr 25 10:51:09 pb__: patches: disable-apm.patch for most + faster-rotated.patch for some Apr 25 10:51:53 hrw|work: ah, right, I see Apr 25 10:53:08 It'd be nice to figure out a better way to deal with the faster-rotated thing at some point. Apr 25 10:53:32 as for apm, I suppose the right way to deal with that is to have an --no-apm switch on the command line, rather than having to patch it out at build time. Apr 25 10:53:47 the xserver args are already highly machine specific, so that wouldn't be much of an encumbrance. Apr 25 10:54:20 eh, no irdadump in oe? Apr 25 10:55:17 CoreDump|home: fix it and I will bless you Apr 25 10:56:30 03hrw 07org.oe.oz354x * r93a2c527... 10/packages/linux/ (4 files in 2 dirs): linux-openzaurus: added some hostap patches (experimental) - all was sent to proper maintainer in mainline (still no answer anyway) Apr 25 10:56:31 hrw|work: faster-rotated is a win on machines with UMA framebuffer, right? I guess the fb driver could export some flag to advertise that fact. Apr 25 10:56:34 03hrw 07org.oe.oz354x * rdb0bc18c... 10/packages/xserver/xserver-kdrive_20050207.bb: xserver-kdrive: bumped PR due to feed error Apr 25 10:56:44 pb_: Or fix the xserver so a patch wasn't needed ;-) Apr 25 10:57:02 RP: fix what, exactly? iirc, the apm problem is due to a bug in the kernel; the xserver is behaving correctly. Apr 25 10:57:25 pb__: when I wrote the patch, there was a bug both in kernel and X Apr 25 10:57:26 pb_: Well, fix whatever the problem is that requires the patch Apr 25 10:57:35 pb__: never got round to retesting the newer release Apr 25 10:57:40 pb__: dont know - I was opie user when it was added Apr 25 10:57:54 hmm its Makefile isn't generated Apr 25 10:57:57 XorA: yeah, I think the bug in X was fixed ages ago. that was the segfault on resume due to double ts_close(), right? Apr 25 10:58:08 pb__: yes, and I beleive that is fixed Apr 25 10:58:30 yeah. there's definitely a patch in there that's meant to fix it, and I'm pretty sure it works. Apr 25 10:58:48 pb__: if I get time tomorrow, (playing with postfix today) then Ill try it again Apr 25 10:58:55 cool, thanks Apr 25 10:59:07 I think the other apm problem needs some attention from the kernel h4x0rs. Apr 25 10:59:14 pb__: try and knock that irritating on eon the head Apr 25 10:59:14 XorA: .dev builds 20060312 by default, which shouldn't need the fix Apr 25 10:59:23 pb__: Why does the xserver need to know about suspend/resume anyway? Apr 25 10:59:30 koen: I know, I am currently busy atm Apr 25 10:59:36 RP: to repaint the framebuffer on machines that require it. Apr 25 10:59:38 XorA: same here :) Apr 25 10:59:50 RP: i.e. on hardware where the contents of vram isn't preserved during sleep Apr 25 11:00:05 pb__: Shouldn't the kernel be handling that? Apr 25 11:00:20 RP: no, the xserver can do it better Apr 25 11:00:45 03hrw 07org.oe.oz354x * r059a97dc... 10/packages/meta/task-bootstrap.bb: task-bootstrap: bumped due to config changes Apr 25 11:00:49 pb__: Better maybe, but I'd say the kernel not doing that would be a bug... Apr 25 11:01:02 RP: no, the kernel shouldn't be messing with that kind of thing. Apr 25 11:01:15 RP: the kernel bug was 2 listeners on /dev/apm_bios, second one enters D state when a suspend event happens Apr 25 11:01:35 pb__: So the kernel isn't responsible for returning the system in the state it was when it suspended? Apr 25 11:01:41 RP: although it would be useful if the kernel did export a flag to tell the xserver when a repaint is required. Apr 25 11:02:51 RP: no, not in this case. The kernel is responsible for restoring the state of the bits of the system that are under its control, but the framebuffer contents don't fall into that category. Apr 25 11:04:10 having the kernel restore the framebuffer from a backing copy would be a waste of time and memory. Apr 25 11:06:01 There's two different ways to look at that... Apr 25 11:07:51 ~seen pigi Apr 25 11:07:57 pigi was last seen on IRC in channel #oe, 4d 13h 22m 8s ago, saying: 'later all'. Apr 25 11:15:54 pb__: Thinking futher about this, I think Linux works around this whole issue now by switching vc before and after a suspend/resume. If thats the case, X shouldn't need to care about apm events anymore? Apr 25 11:17:08 RP: do you know if that objdump binutils patch you added is already in binutils 2.16+? Apr 25 11:17:40 koen: Offhand, I don't. If its not, we should apply it though. Apr 25 11:17:43 RP: what about non framebuffer backed devices, or cards not on a VT? Apr 25 11:18:19 RP: it failed to apply to 2.16.9 in OE Apr 25 11:20:00 `seen zecke Apr 25 11:20:04 I'm beginning to think the irda driver doesn't work on Akita. There's not even a dmesg message when the module loads Apr 25 11:20:10 !seen zecke Apr 25 11:20:17 ~seen zecke Apr 25 11:20:20 zecke was last seen on IRC in channel #oe, 38m 1s ago, saying: 'cya'. Apr 25 11:21:09 Thanks, CoreDump|home Apr 25 11:21:14 ;-) Apr 25 11:21:17 np ;) Apr 25 11:21:27 XorA: Do those devices use videomemory? Apr 25 11:21:37 CoreDump|home: /sys/devices/platform/pxa2xx-ir/ exist? Apr 25 11:21:49 koen: That suggests it was applied for 2.16. I might have a copy of that I can test for the bug Apr 25 11:22:12 hrw|work: yes Apr 25 11:22:17 RP: matrox second heads, XGA CF cards, who's to know what info they lose on suspend Apr 25 11:23:07 yay! Apr 25 11:23:11 draft mailed! Apr 25 11:23:11 CoreDump|home: /sys/devices/platform/pxa2xx-ir/driver/ exist too? Apr 25 11:23:17 uh oh Apr 25 11:23:28 * XorA goes out for a bit in the rare scottish sun Apr 25 11:23:41 # ls /sys/devices/platform/pxa2xx-ir/driver/ Apr 25 11:23:41 bind pxa2xx-ir unbind Apr 25 11:23:56 I think hentges.net just rebooted or something Apr 25 11:24:15 koen: 2.16 doesn't have the bug Apr 25 11:24:20 hmm no Apr 25 11:25:49 hrw|work: my problem is that discovery-log doesn't show anything, I can't use irdadump as it doesn't build and nothing what so ever indicates that irda is working Apr 25 11:26:39 XorA: Looking at the xserver, the apm code seems to assume attachment to a VT and I suspect it will see a VT change on suspend/resume. I therefore don't see APM bringing it any benefit... Apr 25 11:27:42 RP: cool, thanks for testing Apr 25 11:28:09 CoreDump|home: do you have a digicam nearby? Apr 25 11:28:37 CoreDump|home: gpe-beam just received vcard from my phone via irda (c760) Apr 25 11:28:39 koen: hmm yes, but I'm not sure the batteries will do another photo Apr 25 11:28:43 CoreDump|home: most digicam are sensitive to IR and show purple blips on the lcd Apr 25 11:29:00 * RP has never actually tried irda on the Z, I just made the patches to enable it :) Apr 25 11:29:29 hrw|work: is irattach running on the c7x0? Apr 25 11:29:46 koen: nice idea,thank! Apr 25 11:30:07 CoreDump|home: no Apr 25 11:30:07 CoreDump|home: that's how we debugged IR on the h2200 eons ago :) Apr 25 11:30:14 heh Apr 25 11:30:18 hrw|work: hmmmmm Apr 25 11:30:46 koen: I used webcam to test does TV remote works after repair Apr 25 11:30:46 CoreDump|home: we had serial console in IR before we found out to enable the serial connector Apr 25 11:30:56 If someone figures out the commands to make two Z's talk to each other via IR, let me know... :) Apr 25 11:31:09 koen: someone killed ewi connection? Apr 25 11:31:37 koen: ouch, sounds painful see how IR usually sucks heh Apr 25 11:31:41 hrw|work: which connection? Apr 25 11:31:50 debug1: Connecting to ewi546.ewi.utwente.nl [130.89.145.130] port 22. Apr 25 11:31:54 koen: stalled Apr 25 11:32:52 hrw|work: it should be working Apr 25 11:33:18 bad luck then I have Apr 25 11:34:34 * koen moves 500m north Apr 25 11:34:43 later all Apr 25 11:34:56 koen|uni: yep. local problem - connected via hh.org Apr 25 11:41:16 RP: few things which we can remove from kernel: cpufreq governors -> modules, disable 'verbose BUG() (70K)' Apr 25 11:41:57 hrw|work: I'd like verbose bug left in - helps if users report oopes Apr 25 11:42:10 ok - just noticed Apr 25 11:42:29 I'm rebuilding kernel for testing now anyway.. 64M sd card.. Apr 25 11:43:25 hrw|work: We can make the cpufreq governors modular though - that's fine with me Apr 25 11:43:38 nope irda just doesn't work Apr 25 11:43:54 socket(PF_IRDA, SOCK_STREAM, 0) = 3 Apr 25 11:43:54 getsockopt(3, SOL_IRDA, 1, 0xbea8ba88, 0xbea8b534) = -1 EAGAIN (Resource temporarily unavailable) Apr 25 11:46:51 CoreDump|home: Looking at the kernel, akita irda handling is a tiny bit different from anything else as it uses the IO expander to turn it on/off. I hope I haven't reversed the logic in there. You could also check the defconfig has all the bits you need enabled - it should have though... Apr 25 11:48:49 CoreDump|home: Nothing jumps out at me from a diff of the defconfig with c7x0 Apr 25 11:48:52 no SIR device drivers configured Apr 25 11:48:56 hmm Apr 25 11:49:22 Is some module not getting loaded? Apr 25 11:49:32 no clue, quite possible Apr 25 11:49:41 hrw|work: could you pastebin a lsmod for me? Apr 25 11:49:52 03mickeyl 07org.oe.dev * r4f1dfff4... 10/packages/openttd/openttd_0.4.0.1.bb: openttd: add opie support via sdl.bbclass Apr 25 11:49:57 03mickeyl 07org.oe.dev * rb483f521... 10/packages/bochs/bochs_2.1.bb: bochs: flag as broken Apr 25 11:50:01 03mickeyl 07org.oe.dev * r61aa408f... 10/packages/ (sliderulez/sliderulez_0.92.bb meta/task-qpe.bb): add sliderulez, an advanced RPN pocket calculator for Opie Apr 25 11:50:06 If it was the power enable, I'd not exect the errors you' Apr 25 11:50:09 re seeing Apr 25 11:50:17 CoreDump|home: moment Apr 25 11:50:56 http://openembedded.pastebin.com/680777 <- my lsmod Apr 25 11:51:02 http://oe.pastebin.com/680778 Apr 25 11:51:32 irda :) Apr 25 11:51:56 hang on, that is in both :-/ Apr 25 11:52:07 yeah Apr 25 11:52:26 he got rfcomm loaded Apr 25 11:52:48 and you have serial_cs Apr 25 11:53:03 yeah for testings sake Apr 25 11:53:13 though maybe the serialports need to be setup Apr 25 11:53:17 thought Apr 25 11:54:11 nah rfcomm is bluetooth Apr 25 11:56:01 hrw|work: could you rmmod + modprobe pxaficp_ir and check for dmesg messages? Apr 25 11:56:36 CoreDump|home: no messages iirc Apr 25 11:56:42 hmpf Apr 25 11:57:27 CoreDump|home: No messages on spitz. No idea if IR works on it either though Apr 25 11:57:32 =) Apr 25 11:57:57 FWIW the Akita is not picked up from irdadump on a known-to-be-working machine Apr 25 11:58:50 irxfer just errors out when trying to send something Apr 25 11:58:58 and hängs when trying to recieve something Apr 25 11:59:03 errr hangs Apr 25 11:59:41 anyways, gotta go to work, cya! Apr 25 12:08:29 Hi all Apr 25 12:08:37 hi dirk Apr 25 12:08:57 hey dirk Apr 25 12:09:23 hey Marcin, Liam Apr 25 12:14:44 hi, hrw|work koen|uni Apr 25 12:15:52 what's the meaning of INITSCRIPT_NAME="netperf" in netperf.bb? Apr 25 12:16:16 minipanda: look into OE/documentation.conf Apr 25 12:27:41 hrw|work: em, seems do_install not really install binaries correctly :( Apr 25 12:28:00 [hzhang@labserver1 rootfs]$ find | grep pth Apr 25 12:28:00 ./usr/lib/ipkg/info/pthreads.control Apr 25 12:28:00 ./usr/lib/ipkg/info/pthreads.list Apr 25 12:28:22 hrw|work: but the actually biniaries should be 'pth_XXX' or something Apr 25 12:29:08 minipanda: maybe its time to look inside package? Apr 25 12:29:34 but seems BB_DEBUG only have one level? ... Apr 25 12:29:41 RP: http://oe.pastebin.com/680819 - what you think? Apr 25 12:29:44 * minipanda expecting something like printk ... Apr 25 12:31:38 minipanda: I can say that I'm expecting minimal knowledge about packages managment tools.. Apr 25 12:31:48 koen|uni: could you install pyrex on ewi? Apr 25 12:31:49 re zecke Apr 25 12:35:01 hrw|work: sorry ... :( Apr 25 12:35:52 *bored* Apr 25 12:36:28 minipanda: dpkg-deb -c package.ipk Apr 25 12:37:08 hrw|work: ? dpke-deb -c equals ar x ? Apr 25 12:37:34 minipanda: no - it show content of package not extract it Apr 25 12:37:44 zecke: i'm totally insane about howto add a simple custom pacakge for almost whole night ... :( Apr 25 12:43:16 ~botmail for zecke: pyrex-2.4 installed Apr 25 12:45:22 minipanda: you just choose a custom package with a sucky makefile Apr 25 12:46:06 koen|uni: yes, maybe LTP is not design for a newbie Apr 25 12:48:20 whats LTP? Apr 25 12:49:02 mithro: Linux Test Project ... consist of a lot of testcase Apr 25 12:49:16 An NSLU2 appears to be sitting on my desk :) Apr 25 12:49:29 hrw|work: I'm not sure I like it :-/ Apr 25 12:49:31 RP: it does, yes, but only since 2.6. X still needs to support APM for compatibility with older kernels. Apr 25 12:49:44 RP: that was quick Apr 25 12:50:07 RP: nice Apr 25 12:50:22 RP: also, that VC switch is kind of undesirable because it forces a repaint even on UMA machines where it isn't required. Apr 25 12:50:39 rwhitby: I'm quite surprised as I wasn't expecting it until the end of the week :) Apr 25 12:50:40 RP: I do not push it Apr 25 12:51:38 hrw|work: It needs thought. The alternative might be to drop pcmcia-cs support for the 2.6 kernels. We'll be forced into that soon anyway Apr 25 12:52:24 pb__: I suspect the kernel did this to totally avoid the issue of restoring the framebuffer while removing the need for userspace suspend/resume notification Apr 25 12:53:00 RP: I recommend building DebianSlug (or OpenSlug, depending on whether you want LE or BE) and replacing the stock firmware as quickly as possible. Apr 25 12:53:37 RP: I think the kernel does that to make suspend-to-disk work properly on i386. Which is kind of a worthy goal but it would be nice if it could be done without penalising other people. Apr 25 12:53:42 rwhitby: You sound like me when someone gets a Zaurus :) Apr 25 12:53:46 i should pester lennert for armeb eabi support Apr 25 12:54:17 ~change 65 eur to pln Apr 25 12:54:22 65.00 Euro (EUR) makes 251.140 Zloty (PLN) (from http://www.xe.com/) Apr 25 12:54:47 hrw|work: hi, here is the default install: section in my Makefile Apr 25 12:54:55 hrw|work: install: Apr 25 12:54:56 @set -e; for i in $(TARGETS) ; do ln -f $$i ../../../bin/$$i ; done ; Apr 25 12:55:40 rwhitby: I'd better not play with it at the moment as I've work I'm supposed to be doing. I might set a build away though ;-) Apr 25 12:55:45 hrw|work: so ... use the directory reference directly ../../../bin/blah will broken the installation? Apr 25 12:56:10 * minipanda still can't fully understood magic of OE Apr 25 12:56:37 RP: do you want the MasterMakefile instructions, or are you just going to set up your own build area ? Apr 25 12:56:42 pb__: It wouldn't seem that is penalising that many people if most framebuffers don't restore state Apr 25 12:57:14 rwhitby: Can I build it "normally" from OE or is there external magic involved? Apr 25 12:57:46 RP: no external magic. the MasterMakefile just makes it easy for people not familiar with OE. Apr 25 12:57:53 (which of course is not you) Apr 25 12:58:03 RP: well, most of the ipaqs are UMA, for example. Apr 25 12:58:21 DISTRO=debianslug|openslug, MACHINE=nslu2 Apr 25 12:59:15 debianslug and openslug are identical modulo the endianness Apr 25 12:59:39 pb__: The apm code in the xserver already penalises them though? Apr 25 13:00:02 rwhitby: I'll go with little endian given thats what the rest of my devices are... Apr 25 13:00:30 RP: I run all my slugos devices little endian too. Apr 25 13:00:45 (just so I can run debian arm packages on them) Apr 25 13:01:09 rwhitby: How much will DISTRO="openzaurus-unstable" break? :) Apr 25 13:02:05 RP: my theory is - under Linux, X can really be considered a userspace driver - hence it should handle suspend/resume for the devices it drives Apr 25 13:02:10 RP: it won't boot for a start. Have a look at linux/ixp4xx-kernel.inc for the devio prepends needed to fool the kernel that it is actually on an nslu2 and not an ixdp425. Apr 25 13:03:08 RP: apart from that non-obvious thing, I expect you'd be able to sort out any other differences. Apr 25 13:04:40 RP: yeah, it does; that's why I was saying earlier it would be nice to have the kernel export a flag so that this could be avoided. Apr 25 13:04:56 RP: having the kernel do this VC switch makes the behaviour more entrenched, which seems like a retrograde step. Apr 25 13:05:34 rwhitby: I'd be using MACHINE="nslu2" and your kernel, just with a different distro (which would save me an extra build directory). I'll play it safe for now and experiment later :) Apr 25 13:06:29 RP: any news on that .bb for the nokia770 kernel? Apr 25 13:06:30 pb__: I don't like the idea of framebuffer devices losing their contents unless they make the VC switch in which case its acceptable. Apr 25 13:06:58 koen: My laptop is back in action and those files are intact. I just need to clean one up Apr 25 13:07:20 RP: I guess that if the kernel could refrain from doing the VC switch on UMA, that would be an acceptable solution. Apr 25 13:07:29 RP: I'd be happy to help you get a different distro to work with nslu2 and ixp4xx-kernel. It would be a good way to make sure they are generic and not bound to the slugos-derived distros. Apr 25 13:08:23 pb, RP: what is UMA? Apr 25 13:08:43 03koen 07org.oe.dev * rd847f936... 10/conf/machine/nokia770.conf: nokia770.conf: make it work with a newer mtd utils Apr 25 13:08:53 Unified memory arch Apr 25 13:08:59 ah thx Apr 25 13:09:12 rwhitby: It would probably make a good test for the other distros too. I'll get it working to get more of a feel for the device, then think about breaking things :) Apr 25 13:10:09 pb__: It sounds like a good idea in principle, I suspect convincing mainline of its value would be hard though... Apr 25 13:10:33 RP: we'd welcome any comments you have on ixp4xx-kernel too. Just remember there are a lot of things we were forced to do by the broken bootloader on the slug. We chose to fix things in the kernel, rather than ask people to risk bricking their device by replacing the bootloader. Apr 25 13:11:23 rwhitby: I fully appreciate that - The Zaurus kernel has its hacks to deal with the bootloader too (some in mainline :-/). Apr 25 13:11:39 RP: yeah, regrettably I long since gave up trying to convince those guys of anything. I guess we can patch our kernels locally though. Apr 25 13:12:05 hrw|work: unified memory architecture, ie frame buffer in main memory. Like on collie, I think. Apr 25 13:12:14 pb__: I'm slowly winning a few battles against them ;-) Apr 25 13:12:21 hrw|work: and as opposed to an offchip framebuffer like the c7x0 has Apr 25 13:12:26 RP: heh Apr 25 13:12:48 The c7x0 is the only zaurus that isn't UMA (not sure about tosa) Apr 25 13:12:55 RP: since you want LE, DISTRO=debianslug is the one you want. Apr 25 13:13:37 RP: but first you'll want to put a serial console header on it, and de-underclock it. Apr 25 13:14:29 (you won't need JTAG - I bought a cable and soldered a slug ready for it, but have never needed to use it) Apr 25 13:15:37 rwhitby: Which target do I want sluos-image? Apr 25 13:16:14 RP: yep, slugos-image will do it. That will set it up to download packages from the slugos-lag feed ("lag" = little-endian, arm, glibc) Apr 25 13:19:25 RP: and the best way to flash it is by holding in the reset button at power-on until the ready led goes red, then release and it will flash red/green which means your in RedBoot upgrade mode ready to send an image over from linux with upslug2 or from windows with the SerComm upgrade utility. Apr 25 13:20:02 ~lart redboot Apr 25 13:20:02 * ibot chops redboot in half with a free AOL CD Apr 25 13:20:31 RP: or if you have the serial console, just ^C and type "upgrade", or you can load from a tftp server and write the flash directly. Apr 25 13:21:18 I personally just type upgrade and then use the SerComm tool from windows. I do all my development in a VMWare session, and upslug2 can't get past the NAT setup I have to VMWare. Apr 25 13:21:33 XorA: http://emma-soft.com/projects/epdfview/changeset/109 Apr 25 13:22:17 RP: The serial consoles of my slugs and ludes are connected to an Asus wl500gx router's usb ports, so I can flash wirelessly. Apr 25 13:22:41 rwhitby: Thanks for the info, I was going to need to look all that up :). I haven't booted windows for a few months so I suspect I'll be using the linux util... Apr 25 13:23:07 * RP ran out of serial ports a while ago. My desk it a mess of cables... Apr 25 13:23:53 RP: A usb hub, and lots of old usb mobile phone data cables does the trick. Apr 25 13:24:32 rwhitby: I think I need some old USB mobile phone cables :) Apr 25 13:25:52 RP: http://nslu2-linux.thestuffguy.com/gallery/Serial_adapters Apr 25 13:26:04 03rwhitby 07org.oe.dev * r15a80a37... 10/packages/slugos-init/ (files/reflash files/sysconf files/turnup slugos-init_0.10.bb): slugos-init: Updated reflash, sysconf and turnup to work with the sysconfig partition on an NAS 100d. Apr 25 13:27:32 RP: have you o-hand guys looked at adding a new fontconfig to OE? Apr 25 13:27:55 RP: and http://www.nslu2-linux.org/wiki/HowTo/AddASerialPort Apr 25 13:28:00 RP: OE's copy is about 2 years old.... Apr 25 13:29:23 Do bitbake make a the modules? I cant se that it makes all the modules i have defined in the defconfig file. I have also made a copy of the defconfig to .config but it just dont builds all the modules. Do i have to run make modules my self before i make a bootstrap-image? Apr 25 13:29:38 wb Apr 25 13:30:06 goxboxlive: OE build everything mentioned in the .config Apr 25 13:30:12 builds* Apr 25 13:30:50 RP: dyoung may have a serial cable he could send you Apr 25 13:31:08 Hi Apr 25 13:31:22 koen: We haven't Apr 25 13:31:41 RP: do you like USB/Serial cables over normal serial cable ? Apr 25 13:32:26 dyoung: A USB cable might be the most useful Apr 25 13:33:06 dyoung, rwhitby: I might have an old mobile phone serial data lead I can use. I'll just have to try and find it :) Apr 25 13:33:15 Okay. I have "a few" extra ones. :-) But I'll need to put the proper connector on, and of course make sure it works. Apr 25 13:34:08 (a classic case of, "why buy one when you can buy 10 for 10 times the price") Apr 25 13:34:20 koen: Hmm that is strange, because when i manually build my kernel with the modules i have a lot more files in lib/modules/kerne-**** then when i build with bitbake. Apr 25 13:34:55 goxboxlive: I didn't say OE install all the modules.... Apr 25 13:35:40 RP: if you cannot find your cable, let me know dyoung@nslu2-linux.org and I'll get one to you. Apr 25 13:37:18 dyoung: ok, thanks. I've just looked at a map and I suspect postage costs would be more than a lead would cost to buy here - I'm sure I'll work something out :) Apr 25 13:37:21 koen: How do i make bitbake install all the modules ? I also hget this message when i am making a bootstrap-image: http://pastebin.com/680910 (look at line 23) Apr 25 13:37:40 RP: what is your geographic location? Apr 25 13:37:50 dyoung: UK Apr 25 13:38:22 $5 or so. Apr 25 13:38:42 No problem, I'll get the nslu2-linux project to reimburse me. Apr 25 13:39:11 RP: what kind of cables do you need, I can check me server room Apr 25 13:39:20 RP: the new fontconfig should be a lot faster due to the mmappable cache Apr 25 13:40:07 koen: woo woo, I love that developer Apr 25 13:40:08 XorA: He'll be needing a USB->LVTTL serial adaptor. Apr 25 13:40:29 dyoung: ah like the nokia/motorola ones Apr 25 13:41:44 ~botmail for zecke: http://guitone.berlios.de/wiki/index.php/Main_Page Apr 25 13:42:03 XorA: yes Apr 25 13:42:28 RP: I think I might have an old motorola one at home, Ill see and post it to you, it'll be like 40p max Apr 25 13:42:32 pl2303? Apr 25 13:42:46 hrw|work: pl2303 with 3.3V output will do fine. Apr 25 13:44:14 XorA: If its unwanted, I have a use for it :). I'm sure I can find something if not though. Apr 25 13:44:28 RP: well I smashed the phone for being shit Apr 25 13:45:01 RP: cant remember if I threw the lead away when I moved, I will look though Apr 25 13:45:36 night all. RP: feel free to ping me tomorrow - I usually idle here while I'm at work. Apr 25 13:45:38 RP: I'm gonna go back to sleep now; please dont hesitate to speak up if you n eed/want one. Apr 25 13:48:16 hmm, my wife might have an old samsung one as well, I rember that cable being odd Apr 25 13:48:30 dyoung: ok, thanks Apr 25 13:48:43 rwhitby: will do, thanks Apr 25 13:48:54 * RP wishes he was idle at work :) Apr 25 13:49:39 RP: beleive me, I'm not idle there either. but if you ping I will eventually see it, Apr 25 13:49:41 XorA: I'll also have a look tonight so hopefully one of us will find something, thanks :) Apr 25 13:52:45 rwhitby: I misread that, sorry :) (missed the word here out which changes the meaning somewhat!) Apr 25 13:54:35 XorA: NOTE: package epdfview-0.1.3-r1: task do_configure: started Apr 25 13:54:50 XorA: NOTE: package epdfview-0.1.3: completed Apr 25 13:55:04 koen: already installing it :-) Apr 25 13:58:37 03koen 07org.oe.dev * r92f18635... 10/packages/fontconfig/ (4 files in 2 dirs): fontconfig: update to 2.3.91 for some much needed speed improvements Apr 25 13:58:42 03koen 07org.oe.dev * rf1bf6327... 10/packages/epdfview/epdfview_0.1.3.bb: epdfview: update to 0.1.3 Apr 25 13:59:17 koen: god kdrive needs bitblt :-) Apr 25 13:59:34 koen: but epdfview_1.3 works well Apr 25 14:00:13 with stylus scrolling :) Apr 25 14:00:22 * koen while true ; do bitbake -b gaim/gaim_cvs.bb -c fetch ; sleep 30 ; done Apr 25 14:03:11 koen: cvs.sf.net? Apr 25 14:03:32 Hmm. We should probably have PKGDIR set in bitbake.conf give its use by the slug builds... Apr 25 14:05:14 RP: I use PKGDIR and HRW_PKGDIR Apr 25 14:06:39 hrw|work: I've been using collections... Apr 25 14:09:01 koen: s/r1/r0/ Apr 25 14:29:12 jo tmbinc Apr 25 14:36:26 hrw|work: yes, sf.net Apr 25 14:41:36 I've just added a madwifi-ng_svn.bb to OE, but it uses a snapshot - if someone would be kind enough to convert it to use SVN proper (I don't know the syntax off-hand) I'd appreciate it. Apr 25 14:42:45 koen: are any rules defined for fri/sat BSP? Apr 25 14:43:01 rwhitby: http://bitbake.berlios.de/manual/ch03s06.html Apr 25 14:43:10 rwhitby: what is lacking (what kind of info) Apr 25 14:44:14 hrw|work: not that I know off Apr 25 14:44:56 zecke: probably just the fact that it's after midnight and I couldn't be bothered looking for the info :-) Apr 25 14:45:01 welp I'm down to needing about 200 more words Apr 25 14:45:21 mithro: come on you are almost done Apr 25 14:45:59 zecke: do you want to discuss the opie exclusion in angstrom further? Apr 25 14:46:27 koen: not now Apr 25 14:46:30 200 words and a few little fixes Apr 25 14:46:48 koen: need to bring my car to the garage soon Apr 25 14:46:55 zecke: join us in #angstrom when you're ready Apr 25 14:47:00 zecke: you have a car?!? Apr 25 14:48:13 koen well sponsored by grandparents (now grandma) Apr 25 14:48:18 rwhitby: http://viewmtn.angrygoats.net/branch.psp?branch=net.venge.monotone.trac-plugin Apr 25 14:49:03 need to stop for a bit Apr 25 14:49:46 if the essay wasn't due yesterday I would stop and continue again tommorrow Apr 25 14:50:02 koen: now we really do need to get trac running on monotone.nslu2-linux.org :-) Apr 25 14:51:41 03koen 07org.oe.dev * rcaff9d77... 10/packages/epdfview/epdfview_0.1.3.bb: epdfview: fix PR Apr 25 14:54:31 * mithro notices his nslu2 is sitting unused Apr 25 14:55:04 mine is still in the box :p Apr 25 14:55:52 03rwhitby 07org.oe.dev * r7480e1b4... 10/ (5 files in 4 dirs): madwifi: Added madwifi-ng-modules, and included it in slugos-packages. Apr 25 14:57:02 bugs,oe,org is likely broken atm :p Apr 25 14:59:38 cya later Apr 25 15:03:58 ~seen kergoth Apr 25 15:04:02 kergoth is currently on #oe #wowace #elinux. Has said a total of 49 messages. Is idling for 11h 25m 49s, last said: '"You're DEAD! You're all dead!" "Whew. Good. He thinks we're zombies. He'll leave us alone."'. Apr 25 15:04:33 the upside is, it booted Apr 25 15:04:38 So uh I'll sort bugzilla shortly Apr 25 15:09:39 * koen starts experimenting the monotone + trac Apr 25 15:19:54 treke|home: no . key? Apr 25 15:23:34 How can i stop bitbake from generating all locales, can just set it up to generate english? Apr 25 15:24:28 edit the patchfile Apr 25 15:24:42 and we have trac integration: http://ewi546.ewi.utwente.nl:3000/trac/browser Apr 25 15:24:52 (running an old rosterified database) Apr 25 15:25:46 why are we using trac now instead of mntview? Apr 25 15:25:54 Rosterified? LIke Bob Marley? Apr 25 15:26:01 cu Apr 25 15:26:19 emte: 17:19 * koen starts experimenting the monotone + trac Apr 25 15:26:51 ah, for fun Apr 25 15:27:06 koen: http://ewi546.ewi.utwente.nl:3000/trac/browser trac detected an internal error Apr 25 15:27:23 XorA: the db locking still sucks a bit Apr 25 15:27:33 and it errors out on most operations Apr 25 15:28:10 * emte is curious why all the directories say "Add new machine qemuarm (builds images and a kernel compatible with ?" Apr 25 15:29:05 emte: because the plugin is higly alpha Apr 25 15:33:43 well atleast it only always complains about the same key Apr 25 15:34:36 cu Apr 25 15:38:24 hi hi Apr 25 15:38:41 * raduga just recently (yesterday) updated a several-month-old OE chain Apr 25 15:38:57 it won't build now, so apparently I'm doing something off; or missed upgrading something Apr 25 15:39:22 raduga: There have been a lot of changes in the past few months Apr 25 15:40:31 RP: yep. I'm using mt 0.25, bitbake 1.4.1, snap from ewi546 (+ mt update) Apr 25 15:41:14 dying during build of ipkg-native Apr 25 15:41:38 during link of libbb.a Apr 25 15:41:59 In function `__i686.get_pc_thunk.bx': Apr 25 15:41:59 | : multiple definition of `__i686.get_pc_thunk.bx' Apr 25 15:43:01 should be a clean tmp space, clean src and org.openembedded.dev strees Apr 25 15:43:32 RP: anything obvious you can think of? Apr 25 15:45:45 I've just installed OpenEmbedded by following the instruction in the wiki... Everything seems to be working except when I gets to the glibc-2.3.5 package which is building correctly up to the point when it is generating the locale. Apr 25 15:46:28 it returns me "NOTE: task failed: localedef returned an error." Apr 25 15:46:58 is there anybody who have a fix for this problem? Apr 25 15:47:35 what is the locale set to, on your build box? Apr 25 15:49:23 RP: using gcc 3.3.5 natively. are there any known conflicts with recent gcc? Apr 25 15:53:49 ehe. Apr 25 15:54:29 koen: hey, I thoungt it would look more fancy Apr 25 15:55:35 marcos: we've already established that that is a bug in gcc or glibc. Apr 25 15:55:46 enough already Apr 25 15:56:22 * raduga whines at RMS, to fix his buggy compiler Apr 25 15:57:09 Hey guys, is it not possible to have OE files reside on an NFS server? http://rafb.net/paste/results/ljgtHU13.html shows the log to my failing OE build. Apr 25 15:57:57 Laibsch: autom4te: cannot lock autom4te.cache/requests with mode 2 Apr 25 15:58:07 Laibsch: not a OE bug at all Apr 25 15:58:35 Laibsch: which nfs server, which client, which NFS version Apr 25 15:59:47 I never said it was a bug. ;-) Apr 25 16:00:31 Laibsch: well, locking on NFS can work on some systems Apr 25 16:00:53 Working on getting a newer release of bugzilla up Apr 25 16:01:05 the current release wont use the version of mysql I have Apr 25 16:01:13 well the currently installed release Apr 25 16:01:16 Anyone having any issues with hermes based pcmcia cards, with new 2.6.16 kernel? Apr 25 16:01:33 koen: lol NeoOffice wants to charge us with 100$ :} Apr 25 16:03:02 iWork is only 79$ Apr 25 16:03:26 * chouimat is away: lunch Apr 25 16:03:48 can anyone else not see oesources.org ? Apr 25 16:04:02 it seems to be down atm Apr 25 16:04:06 zecke: server side config is at http://rafb.net/paste/results/zAKQ9v63.html. I am not sure what you mean by client. Client is ubuntu, but dpkg -l nfs* does not show anything about nfs-client :-\ Apr 25 16:04:11 lrg: treke is your man :) Apr 25 16:04:27 lrg: alias ggilbert (or the other way around) Apr 25 16:04:37 zecke: thanks Apr 25 16:05:10 ggilbert, I'm having no luck downloading from oesources, is it down atm ? Apr 25 16:05:16 Laibsch: line 7? anyway, need to go now Apr 25 16:05:22 zecke: I was thinking about that newsletter thing. I'd certainly want to help out. But I cannot promise too much as far as regular input. So maybe working in a team would be better. How would I get the information about what to write anyway? Apr 25 16:05:39 OK, let's talk about that newsletter thing later. Apr 25 16:05:53 Laibsch: awesome, I will in the gym for the next couple of hours Apr 25 16:06:11 zecke: line 7 shows no client because it is the server! Apr 25 16:06:14 re Apr 25 16:06:38 YES Apr 25 16:06:45 oops Apr 25 16:06:58 Entire webserver is turned off while I upgrade bugzilla :p Apr 25 16:08:41 ggilbert, np Apr 25 16:13:19 yay 4200 words! Apr 25 16:14:17 raduga: I'm afraid I've not seen that problem before and don't know what would cause it :-( Apr 25 16:15:06 RP: want uboot on Z? Apr 25 16:15:08 mithro: yay :) Apr 25 16:15:16 RP: ancient IRC logs suggest its an obscure gcc bug Apr 25 16:15:31 raduga: Can you try a different gcc? Apr 25 16:15:36 RP, maybe a different cpu architecture (i386?) or gcc would help Apr 25 16:15:39 rp: i'll try Apr 25 16:16:29 using stock debian etch at the moment Apr 25 16:17:01 gcc version 3.3.5 (Debian 1:3.3.5-13) Apr 25 16:17:27 ouch, it's 2am Apr 25 16:17:48 what is etch? Apr 25 16:18:09 etch == unstable, stable, testing? Apr 25 16:19:58 testing, I think Apr 25 16:21:30 The code name for the next major Debian release after sarge is “etch”. Apr 25 16:21:42 so that would make it testing, since unstable is perpetually sid Apr 25 16:22:39 ricje: I have exactly the same problem you mentioned a while ago Apr 25 16:23:50 This is the output (adding some additional -D didn't help to get more): Apr 25 16:24:11 NOTE: generating locale es_NI (UTF-8) Apr 25 16:24:11 qemu: uncaught target signal 11 (Segmentation fault) - exiting Apr 25 16:24:11 N Apr 25 16:25:36 heh. here we go with gcc 4.0.3 Apr 25 16:25:45 see what that breaks :) Apr 25 16:27:31 kokos: do you have enough memory? Apr 25 16:27:35 are you heavy into swap? Apr 25 16:28:53 Comment on slashdot: thank you AT&T! WOW server downtime is saving my marriage. Apr 25 16:28:55 :P Apr 25 16:29:21 kokos: yeah, one of my machines does that too. I think something about the build environment breaks qemu. Apr 25 16:29:26 presumably it hates either my gcc or glibc versions. Apr 25 16:29:41 iirc, the machine that breaks is running some version of Fedora Core; the ones that work are running Debian Apr 25 16:29:43 mithro: hear, hear! Apr 25 16:30:54 ggilbert: i seem to remeber being told that sid was not going to be perpetually unstable, guess they where lieing Apr 25 16:31:13 mithro: who said that? Apr 25 16:31:15 sid has always been unstable up until now, and I doubt that would change. Apr 25 16:31:41 ggilbert: my memory is fuzzy - so I guess i'll say noone for now :P Apr 25 16:32:13 hehe :p. vaguelly remember something about sid being chosen since he was the one who kept breaking the toys in toy story :p Apr 25 16:32:25 also, bugzilla is nearly up now Apr 25 16:32:30 rock Apr 25 16:32:32 ggilbert: hah :P Apr 25 16:32:45 Just need to sort out a problem that breaks it horribly Apr 25 16:33:35 ah, it's a template problem Apr 25 16:34:01 I've dumped my custom template for the moment and reverted to stock. I'll get it ported over, but you can access bugzilla again Apr 25 16:35:56 strange: NOTE: package gpe-image-1.0-r22: task do_fetch: completed Apr 25 16:35:56 NOTE: package gpe-image-1.0-r22: task do_rootfs: started Apr 25 16:35:56 then, ipkg-cl just sits there chewing up cpu cycles. (Sorry if I missed this in any e-mail exchange.) Apr 25 16:38:05 anyone seen that on bitbake gpe-image? Apr 25 16:38:10 sorry, but its one of the reasons Debian and pdaXrom are cool, and OE and its related distributions are not cool Apr 25 16:38:25 Debian releases have inscrutable names, from Toy Story Apr 25 16:38:43 pdaXrom releases have inscrutable names, from Sash's ex-girlfriends Apr 25 16:38:53 OZ/Familiar/etc releases are just plain numbers :( Apr 25 16:39:23 i should report that to the bug tracker Apr 25 16:40:31 poky releases are named after pacman characters :) Apr 25 16:41:51 why don't we use kergoth's enemies list to name OE releases? Apr 25 16:42:07 3.5.4, the "Choking glibc" release Apr 25 16:42:38 3.6, "Stabbing libtool" Apr 25 16:43:51 raduga: (sorry for the delay) I think I should have enough memory 1G RAM 2G Swap, and at least 50G on the Disk free Apr 25 16:44:42 3.5.4.1 'die pdaX, die' release Apr 25 16:44:52 3.5.4.2 'collies really suxx' release Apr 25 16:44:59 hrw: ;-) Apr 25 16:45:20 pb__: (sorry for the delay) Thank you for that info. I have Gentoo here and have installed the OE script through python. But could that affect it? Apr 25 16:45:47 eh... through portage, sorry :) Apr 25 16:46:16 RP: or 3.5.4.1 'welcome in a real world' release even ;) Apr 25 16:50:56 i think 'welcome in a real world' should be 1.0 Apr 25 16:51:04 hello all. Apr 25 16:51:15 tnb: did you figure out the ipkg-cl problem? I don't see that. Apr 25 16:51:24 3.5.4.1 could be 'welcome to a world better than the real one' Apr 25 16:51:27 hey mreimer Apr 25 16:51:45 btw, Apr 25 16:51:45 hey koen. did RP's EABI patch help make your angstrom builds more stable? Apr 25 16:51:57 gcc 4.0.3 fixed the "problem" Apr 25 16:51:59 for now ;) Apr 25 16:52:05 mreimer: haven't tried that patch yet Apr 25 16:52:28 mreimer: no, just hangs at that point Apr 25 16:52:53 hi koen, pb__ Apr 25 16:53:11 tnb: what's the exact ipkg-cl command? build/tmp/work/arm-linux/gpe-image-1.0-r22/temp/log.do_rootfs* Apr 25 16:53:19 hey tnb Apr 25 16:53:38 mreimer: well, it's killed now. I'll let you know... Apr 25 16:54:27 would it be ok to edit the RequiredSoftware wiki, to suggest "gcc 4.0+ might help, if you have obscure build failures"? Apr 25 16:54:33 s/ok/reasonable/ Apr 25 16:55:06 raduga: gcc4 breaks qemu which need to be resolved by user manually by setting one var in local.conf Apr 25 16:55:51 hrw: thanks (sigh!) Apr 25 16:56:33 hrw: i won't bother the wiki then Apr 25 16:59:57 Anyone know how to reset (null) the linux passwd file ..? Apr 25 17:00:16 I got access to the system, through init=/bin/sh but cannot login (doesnt accept any passwd) Apr 25 17:00:20 'x'? Apr 25 17:00:35 how you mean? Apr 25 17:01:25 if ken had asked that same question on EFnet #linux, he'd be permbanned. Apr 25 17:01:28 probably for life ;) Apr 25 17:01:54 ken: zaurus or desktop box? Apr 25 17:02:10 lol, Its a serious question. I've tried to port the slackware to my jornada 6xx arch, and its giving me some headaches trying to login Apr 25 17:02:46 vi /etc/passwd and empty the password-field of root Apr 25 17:02:57 Oki, will give that a try Apr 25 17:03:10 or, well, running "passwd" in init=-/bin/sh might do the trock, too Apr 25 17:03:17 trick Apr 25 17:03:37 "Passwd: Cannot determine your user name." Apr 25 17:03:48 vi it is then Apr 25 17:03:48 Ken|JLime: passwd root Apr 25 17:04:01 Getting same answer Apr 25 17:04:48 ken: set the password in /etc/passwd to null with: Apr 25 17:05:04 root::0:0:root:/root:/bin/bash Apr 25 17:05:22 the first field *after* the username is the password file Apr 25 17:05:24 field Apr 25 17:05:31 Already tried that once, didnt do any difference Apr 25 17:05:36 ok Apr 25 17:05:45 what does the shadow entry for root look like? Apr 25 17:05:57 looks nice and encrypted Apr 25 17:06:00 what is your UID when logged in right now? Apr 25 17:06:13 uid=0, gid=0 Apr 25 17:06:17 so should be root, Apr 25 17:06:27 change the shadow password value to null, too Apr 25 17:07:06 then you *should* be able to log on, from a local console, as root, with no password Apr 25 17:07:27 I believe i've tried this before, if it doesnt work what other issues could there be? Apr 25 17:07:31 but if your uid is confused right now, you might not be able to run '/bin/passwd' until you log in properly as root Apr 25 17:07:55 when trying to su to root, it says "sorry" Apr 25 17:09:53 Could the login program in some way be malfunctioning? Apr 25 17:10:34 Or do the kernel need to support the md5sum encryption? (not sure if the current kernel is) Apr 25 17:15:39 hi tnb Apr 25 17:16:10 sorry, this is a repost... Apr 25 17:16:11 I've just installed OpenEmbedded by following the instruction in the wiki... Everything seems to be working except when I gets to the glibc-2.3.5 package which is building correctly up to the point when it is generating the locale. Apr 25 17:16:15 it returns me "NOTE: task failed: localedef returned an error." Apr 25 17:16:20 is there anybody who have a fix for this problem? Apr 25 17:16:48 you'll need to debug that a bit to find out why localedef is failing. the usual cause seems to be that qemu is crashing, generally with a segfault. Apr 25 17:17:00 what version of gcc is in your host environment? Apr 25 17:17:17 4.0.2 Apr 25 17:17:36 but I compiled qemu with gcc 3.2 Apr 25 17:19:04 why would qemu be needed while compiling glibc?? Apr 25 17:19:14 it's needed to generate the locale packages Apr 25 17:19:21 you can turn that off if you want Apr 25 17:19:27 * mickeyl yawns and waves Apr 25 17:19:32 mickeyl: good morning Apr 25 17:19:36 how do I turn that off? Apr 25 17:19:41 hehe, "morning" Apr 25 17:19:49 lazy germans Apr 25 17:20:23 *cough* hey we own the tradition of being a pretty busy species don't we? :) Apr 25 17:20:25 ricje: set ENABLE_BINARY_LOCALE_GENERATION = "0" in local.conf Apr 25 17:20:34 s/busy/disciplined/ or so Apr 25 17:20:38 mickeyl: hehe, right Apr 25 17:20:48 I guess we in the UK are famed for our laziness Apr 25 17:21:00 thanks pb_, I'm tryin that Apr 25 17:22:09 * mickeyl waits patiently for a new incarnition of task-python-everything to compile Apr 25 17:22:09 reenoo|a1k: unfortunately suppressing those rpaths wasn't as easy as I hoped: numerous packages are infected. I'll look at it some more tomorrow. Apr 25 17:23:53 pb_: thanks for tackling that. I hope you don't contract anything from libtool Apr 25 17:24:53 yeah, me too Apr 25 17:25:11 what rpaths? Apr 25 17:25:14 I think I have a patch for libtool now which eliminates the problem at its (most common) source, but there's something funny going on with freetype Apr 25 17:25:29 kergoth: loads of binaries are getting linked with -Wl,-rpath,${STAGING_LIBDIR}, which... sucks. Apr 25 17:25:38 Ken|JLime: hi Apr 25 17:25:46 ahh Apr 25 17:25:52 argh Apr 25 17:25:55 kergoth: hi Apr 25 17:25:59 hey Apr 25 17:26:08 kergoth: for example, if you strace gpe-calendar, you see it trying to open every library in a variety of stupid paths which don't exist on the target system Apr 25 17:26:11 kergoth: can you add CNAME wiki.openzaurus.org -> openzaurus.berlios.de? Apr 25 17:27:32 mreimer: if you feel intrepid, handhelds.org/~pb/hardcode.patch is the patch for libtool. it'd be interesting to know how many binaries still have problems if you do a clean build with that applied. Apr 25 17:27:46 pb_: yeah, i remember seeing that. i presume its taking the libdir in the .la and setting that as the rpath? Apr 25 17:27:52 hrw: k Apr 25 17:28:06 pb_: I dunno, I just took a shower... Apr 25 17:28:13 kergoth: that seems to be the main problem, but freetype's pkg-config data is somehow also ending up with -rpath in the Libs section Apr 25 17:28:15 kergoth: thx Apr 25 17:28:32 ah Apr 25 17:28:40 kergoth: so, anything which links with freetype automatically loses, and anything (like cairo) which inherits freetype's pkg-config stuff is also in trouble Apr 25 17:28:48 * kergoth nods Apr 25 17:28:57 I'm not sure where that's coming from, yet. didn't have time to track it down yesterday. Apr 25 17:29:00 "someone" needs to replace libtool with something maintainable Apr 25 17:29:04 someday. Apr 25 17:29:05 yah Apr 25 17:29:13 not me, though, and not today Apr 25 17:29:18 pb_: so apply this for libtool, or libtool-native and libtool-cross also? Apr 25 17:29:22 pb_: does the freetype I added today to .dev help with the pkgconfig trouble? Apr 25 17:29:27 mreimer: just libtool-native, I think Apr 25 17:29:53 I'm not exactly sure what libtool-cross is for Apr 25 17:29:57 koen: what kind of performance increases does that fontconfig give? Is that the scanning-lots-of-fonts mmap speedup? Apr 25 17:30:07 oh, I guess that provides arm-linux-libtool in staging Apr 25 17:30:10 probably want to patch that one too Apr 25 17:30:12 mreimer: yes, it's the mmap speedup Apr 25 17:30:32 koen: no idea, I haven't done any testing today. maybe mreimer can try it out. Apr 25 17:30:37 * pb_ going out to dinner now Apr 25 17:30:40 later all Apr 25 17:30:44 ~bon appetit Apr 25 17:30:47 methinks bon appetit is smacznego. Guten Appetit. Eet Smakelijk. God Appetitt. Buon Appetito. Buen apetito Bom Apetite. buen apetito Apr 25 17:30:53 heh Apr 25 17:30:57 mickeyl: thanks :-) Apr 25 17:31:27 I think it should be kergoth and thursday Apr 25 17:31:39 pb_: afaik only libjpeg uses the actual generated libtool in staging (libtool-cross), because we dont have its configure.{in,ac} to do an autotools regenerate with our libtool bits Apr 25 17:36:23 * koen mtn --db=OE-26.db db rosterify Apr 25 17:36:29 let's see how long that takes Apr 25 17:47:48 In a nutshell, what is the relatioin between maemo.org and OE? I guess OE will start produce software and maybe even releases for maemo, but will maemo ever work on non-Nokia HW? Why did Nokia not just contribut to OE, what are the (supposed) advantages of creating their own platform? Apr 25 17:49:05 when nokia started the work on maemo, OE was not mature enough. iirc that's whey they took scratchbox, which was the most promising at that time Apr 25 17:49:05 arm-linux/bin/ld: WARNING: /local/pkg/oe/c7x0/tmp/cross/lib/gcc/arm-linux/4.0.2/libgcc.a(_udivsi3.o): 816b71c(__udivsi3): PLT refcount was -1 (set to 1) Apr 25 17:49:05 wtf is that? Apr 25 17:49:26 nokia is still interested in OE Apr 25 17:49:32 nowadays, we have good relations to the maemo.org team. some are improving maemo in or Apr 25 17:49:35 s/or/oe/ Apr 25 17:50:13 mickeyl, i would think the fact that scratchbox has a development enviroment/ide might have been a factor as well ... Apr 25 17:50:39 evan as much as most people hate eclipse it is now used a lot Apr 25 17:50:43 even* Apr 25 17:50:52 * JustinP doesn't hate eclipse Apr 25 17:50:55 I talked to Sampsa about OE, and he seemed pretty happy with Scratchbox Apr 25 17:51:12 Laibsch: It looks like they are working from a debian base. If I had to guess, I'd say they went with what the original development team was familiar with Apr 25 17:51:40 it would be nice if we could get an eclipse plugin for OE that would setup and handle a project for atleast one target Apr 25 17:51:57 http://ewi546.ewi.utwente.nl:3000/trac/changeset/030f55becd1b1ba32e39732bceed7c82c79f6411 Apr 25 17:52:00 yay! Apr 25 17:52:01 it works better now Apr 25 17:52:23 it would help new people using OE in the very least Apr 25 17:53:00 OK, thanks for the replies. Apr 25 17:53:15 I have not heard about scratchbox. I'll have alook. Apr 25 18:08:14 heh Apr 25 18:08:22 mtn used 103% cpu just now Apr 25 18:08:35 must be a bug in top Apr 25 18:09:49 mreimer: http://ewi546.ewi.utwente.nl/tmp/viewmtn/revision.psp?id=bf6be0792ee7f592df6a7940fb7b0bf86f833157 Apr 25 18:11:17 | gtktreeview.override:1911: error: invalid storage class for function 'pygtk_tree_foreach_marshal' Apr 25 18:11:19 koen: does mkfs.jffs2 no longer have the --no-cleanmarkers option? Apr 25 18:11:20 hmm ? Apr 25 18:11:31 mreimer: it seems so Apr 25 18:11:38 koen: ok Apr 25 18:11:50 mreimer: I couldn't find anything resembling that in the new help output Apr 25 18:14:29 RP: did ASoC make mainline? Apr 25 18:14:58 cu Apr 25 18:16:12 Scratchbox sounds a lot more complicated than OE. All those specifically set-up compiler-chains don't really scale, do they? It is nice that OE takes the hassle out of cross-compilation for all kinds of HW (if I wanted to do that). Thanks guys. Apr 25 18:16:51 :) Apr 25 18:17:04 we play better once we improve our docs and once we have better qemu integration Apr 25 18:17:04 scratchbox's method of handling autoconf tests is slick. we should really have a .bbclass to let us choose to use it as an alternativ e to the presupplied test results Apr 25 18:17:05 OE and scratchbox have different goals Apr 25 18:17:12 mickeyl: :) Apr 25 18:17:12 * mickeyl nods to kergoth Apr 25 18:17:35 kergoth: we have qemu-native in OE now Apr 25 18:17:49 kergoth: so technically the hard part has been done :) Apr 25 18:18:15 just need the binfmt_misc bits Apr 25 18:19:24 when I chatted with the nokia people they wanted to use a hybrid approach: Apr 25 18:19:39 use OE for configuration management and building and scratchbox as an sdk fro users Apr 25 18:20:41 thatd make sense, given we pretty much suck as a development environment Apr 25 18:24:16 if we could fold the 'run qemu on the image' bits onto OE/bitbake we would have a nicer techdemo Apr 25 18:27:31 koen: what's the goal of scratchbox? Apr 25 18:28:06 also, btw, what needs to be done to build qemu, with gcc4? Apr 25 18:28:16 aside from "downgrade to gcc3" Apr 25 18:28:36 sorry if its becoming a FAQ :( Apr 25 18:29:10 raduga: build qemu-qop- Apr 25 18:29:36 qemu-qop-nogfx-native_svn.bb Apr 25 18:29:40 that should work with gcc4 Apr 25 18:29:47 thanks Apr 25 18:38:05 hold your horses.... massive python upgrades coming up Apr 25 18:43:38 * koen holds horses Apr 25 18:44:33 hmm three heads Apr 25 18:44:38 * mickeyl waits for automerge Apr 25 18:44:49 automerge? Apr 25 18:45:00 well... non manual merge :) Apr 25 18:45:04 ah Apr 25 18:45:55 03mickeyl 07org.oe.dev * r840ee9d2... 10/conf/distro/sharprom-compatible.conf: sharprom-compatible.conf: upgrade and add the new meta-fields @MENTOR and @REQUIREMENTS Apr 25 18:46:00 03mickeyl 07org.oe.dev * r7066fb78... 10/packages/python/python-pyqt_3.13.bb: python-pyqt: fix virtual/libqpe1 dependency Apr 25 18:46:04 03mickeyl 07org.oe.dev * rdf18bf8f... 10/classes/palmtop.bbclass: palmtop.bbclass: add a special case for sharprom compatible distros -- gcc 2.95 has no idea about libsupc++ Apr 25 18:46:09 03mickeyl 07org.oe.dev * rf010b7ae... 10/packages/meta/ (2 files): task-python-sharprom: update and switch to date based versioning Apr 25 18:46:13 03mickeyl 07org.oe.dev * r7c67e406... 10/packages/python/python-sip_4.1.1.bb: python-sip: inherit palmtop to make it qte/qte-mt logic working Apr 25 18:46:18 03mickeyl 07org.oe.dev * rb195eef6... 10/packages/python/python-pyqt_3.13.bb: python-pyqt: inherit palmtop to use the qte/qte-mt logic and to remove duplicated stuff Apr 25 18:46:22 03mickeyl 07org.oe.dev * r8e73f90a... 10/packages/smpeg/smpeg_0.4.4+0.4.5cvs20040311.bb: smpeg: fix staging, improve packaging Apr 25 18:46:26 03mickeyl 07org.oe.dev * r8f8e76a5... 10/packages/python/python-pyqwt_3.10.bb: python-pyqwt: make it build again, make it work w/ qte/qte-mt Apr 25 18:46:30 03mickeyl 07org.oe.dev * r168adcbd... 10/packages/qwt/qwt_4.2.0rc1.bb: qwt: make it work w/ qte/qte-mt Apr 25 18:46:34 03mickeyl 07org.oe.dev * r4b3a04ee... 10/packages/python/python_2.4.3.bb: python 2.4.3: remove tcl and tk for the sharprom Apr 25 18:46:38 03mickeyl 07org.oe.dev * r01d13ac2... 10/packages/python/ (5 files): python-pygtk2: upgrade to 2.8.6 and remove obsolete versions Apr 25 18:48:05 Spam Spam Spam and Beans Apr 25 18:48:20 can I have that without the spam? Apr 25 18:49:12 hehe Apr 25 18:49:31 This login thing is giving massive headache Apr 25 18:49:49 If its not due to passwd,group or anything like that. Could the login program itself be faulty? Apr 25 18:54:36 evening Apr 25 18:55:14 hey Apr 25 18:55:48 pb_: ah, thanks for checking. Apr 25 18:56:11 yo zecke Apr 25 18:57:35 koen: The problem is you can only mount qemu images read write via an nfsroot :-/ Apr 25 18:58:00 problem, schmoblem Apr 25 18:58:10 document it and call it a feature Apr 25 18:58:29 koen: http://www.rpsys.net/openzaurus/qemu/ Apr 25 18:58:34 *depressed* Apr 25 18:59:37 zecke: :-( Apr 25 19:03:53 * CosmicPenguin starts building an x86_64 image in OE Apr 25 19:04:00 on a x86_64 box - everybody duck! Apr 25 19:04:14 quack Apr 25 19:05:36 Hmm while doing "chown root:root user/" i get "invalid user" Apr 25 19:05:41 thats really weird Apr 25 19:07:43 CosmicPenguin: I'm going to get an AM2 once released Apr 25 19:10:37 RP: I have the cache issue fixed, do we want to consider backports to 1.4.1? Apr 25 19:10:46 rp: is this intended for running an arm-linux machine on a non-arm host? Apr 25 19:10:59 or running non-arm code, under arm? Apr 25 19:11:15 zecke: That fix is one that should be applied to the 1.4.x branch Apr 25 19:11:27 raduga: Running an arm image on your desktop system Apr 25 19:11:29 koen: that should be fun.. :) Apr 25 19:11:34 err tather, zecke Apr 25 19:13:35 zecke: is cia for bitbake working already? Apr 25 19:14:57 koen: well, I'm sending mails ;) Apr 25 19:15:05 koen: but they are likely to be wrong Apr 25 19:18:30 zecke: I resolved my earlier problems by disabling parallel make (-j) Apr 25 19:20:12 baking is slower than I remember it, Apr 25 19:20:19 but its definitely cooking ;) Apr 25 19:20:44 * raduga got through gcc, glib, kernel, all without bitbake barfing Apr 25 19:20:47 * raduga is happy Apr 25 19:22:16 to build an image using a different kernel than default (say, 2.4 instead of 2.6) Apr 25 19:22:20 is there a bitbake invocation? Apr 25 19:22:30 or do you have to manually edit the conf file? Apr 25 19:23:11 raduga: conf files Apr 25 19:24:57 Can somebody please verify that the bb file for anthy in bug 623 does not compile on a clean install (no previous crosscompiling-gcc, etc.)? It looks to me some build-dependency is missing. I had a completely new install but "bitbake anthy" went straight to compiling anthy when of course it should have compiled a lot of stuff first. Apr 25 19:26:53 Laibsch, did you or do you need to inherit one of the task bb bases? Apr 25 19:28:03 base.bbclass is automatically inherites Apr 25 19:28:54 the bb file inherits autotools and pkgconfig Apr 25 19:33:43 I left a note in bug 623. But maybe something else is wrong? I think this is strange that OE wanted to go straight to compiling that package. I am doing a different package with an almost empty bb file now and that seems at least to start compiling some other packages first. Apr 25 19:38:49 Laibsch, i would suspect anthy has more depends than only inheriting types of building tools Apr 25 19:39:19 Laibsch: how did you 'start over' Apr 25 19:39:30 Laibsch: and what was missing? Apr 25 19:39:44 Laibsch: it is almost certain you automatically depend on virtual/libc Apr 25 19:42:07 anthy by the looks minimally depends on libc6 > 2.3.2 Apr 25 19:50:35 stupid question: can stock C do 'references' as well? Apr 25 19:51:00 I don't know what you mean by "references" Apr 25 19:51:10 int foo(int &moo) {} Apr 25 19:51:25 Oh - no, I don't think so Apr 25 19:51:49 * zecke couldn't remember that C sucked so much ;) Apr 25 19:53:10 I hate that usage though - I far prefer your usual foo(int *blah) - why do I need to hide what blah really is? Apr 25 19:53:47 zecke, you can sometimes get away with it ... Apr 25 19:53:58 zecke, depending on the compiler Apr 25 19:54:31 * chouimat is back. Apr 25 19:54:53 CosmicPenguin: http://oe.pastebin.com/681637 Apr 25 19:55:04 CosmicPenguin: I hate to take a pointer from a temporary Apr 25 19:55:15 CosmicPenguin: You can have evil effects ;) Apr 25 19:55:34 * CosmicPenguin shrugs Apr 25 19:55:58 evil effects are indeed possible, but no eviler then char buffer[256] on the stack Apr 25 19:56:01 using pointers with structures within structures gets convoluted as well Apr 25 19:56:12 CosmicPenguin: well, Send is async Apr 25 19:56:24 CosmicPenguin: luckily the network stack is not zero copy, so the packet gets copied :} Apr 25 19:56:59 But on the other hand, there are those would would be annoyed to have an extra copy "just happen" Apr 25 19:57:17 CosmicPenguin: right, it is a loss loss situation Apr 25 19:57:30 CosmicPenguin: the signature - from a software technique point of view - sucks Apr 25 19:57:43 CosmicPenguin: and the implementations consumes more memory than necessary Apr 25 19:57:50 heh Apr 25 19:58:04 well that is a msp with 2k RAM Apr 25 19:58:16 luckily you have not too many packages in the queue Apr 25 20:07:09 zecke: I went almost from scratch. DL OE.db.bz2, removed all dirs from tmp and org.embedded.dev, checked out oe.db after doing a pull, then added the patch from 623 and compiled away. Apr 25 20:07:26 Laibsch: what about the tmp directory? Apr 25 20:07:32 Laibsch: ah right Apr 25 20:07:34 hmm Apr 25 20:08:49 Just asking this in case some guru have a nifty answer. The problem is this, Cannot su, login, passwd..or anything. The user root doesnt seem to exists (hence my problems i guess). I have exact same errors as Apr 25 20:08:54 http://linuxfromscratch.org/pipermail/lfs-dev/2000-November/007514.html Apr 25 20:09:14 Which leads me to believe that libnss is involved somehow, but the thing is that I got those files Apr 25 20:10:41 Aside from the su,login,passwd programs everything else works fine Apr 25 20:11:04 sorry no Apr 25 20:11:09 * CosmicPenguin hugs softlockup detection Apr 25 20:22:29 RP: hi, did you see with Dirk about the conflicting kernel 2.6 patches for tosa (http://oe.pastebin.com/677483) ? Apr 25 20:35:24 cyrilRomain: He's not been around :-/ Apr 25 20:35:34 cyrilRomain: I'll look into it and try and resolve it Apr 25 20:35:42 RP: :-/ I'm so impatient to dive into angstrom ;) Apr 25 20:36:04 RP: thanks, I'll stay tuned :) Apr 25 20:36:16 cyrilRomain: You have a tosa? Apr 25 20:36:22 RP: yes Apr 25 20:37:21 cyrilRomain: ok, just wondered :) Apr 25 20:38:45 RP: how knows, Zaurus are so great that anyone here may own more than one model :) Apr 25 20:41:45 cyrilRomain: At the moment I'm having a big clear out of my system to free some space. Once that's complete, I might be able to get some builds running again :) Apr 25 20:43:47 i have no zaurus Apr 25 20:44:10 cyrilRomain: I have only a SL5000d atm Apr 25 20:47:56 * RP counts 4 :-/. At least I put them to good use I guess... Apr 25 20:52:15 koen: rmk applied that patch Apr 25 20:54:06 RP: has ASoC made it into mainline? Apr 25 20:54:57 koen: should we add ideas to for SoC to 'old' ideas? Apr 25 20:56:22 RP: flashed that slug yet? Apr 25 21:06:04 zecke: yes Apr 25 21:06:22 koen: added two BitBake ideas Apr 25 21:06:23 ... Apr 25 21:08:46 zecke: did you see the monotone trac plugin? Apr 25 21:08:56 koen: I have not looked at it Apr 25 21:09:07 zecke: http://ewi546.ewi.utwente.nl:3000/trac/browser Apr 25 21:09:09 koen: From my experience trac doesn't scale too well Apr 25 21:09:17 I'm rosterifying a db now to test with Apr 25 21:09:34 trac integration with svn is nice Apr 25 21:09:40 s/svn/scm/ Apr 25 21:10:08 mreimer: yes but only if you use wiki, bug tracker and the rest of the gang Apr 25 21:10:22 koen: do you know if GNOME is using a patched viewcvs? Apr 25 21:10:33 zecke: I think so Apr 25 21:10:34 zecke: yep, that's the integration I'm talking about Apr 25 21:10:35 koen: their cvs and bugzilla is integrated (at least one direction) Apr 25 21:10:41 zecke: all the gnome tools are pretty custom Apr 25 21:10:57 koen: where are the patches ;) Apr 25 21:11:11 I think you'd have to ask Olav for the bugzilla stuff Apr 25 21:11:34 koen: KDE is much better, you can just svn co the bugzilla code Apr 25 21:11:49 could be Apr 25 21:11:53 haven't looked at that Apr 25 21:12:30 zecke: http://maemo.org/pipermail/maemo-developers/2006-April/003795.html Apr 25 21:13:38 "In principle even Apr 25 21:13:38 compiling and using them on your own machine would be Apr 25 21:13:39 illegal." Apr 25 21:13:43 hmm Apr 25 21:13:51 not being a lawyer I doubt that Apr 25 21:15:34 zecke: the "mere aggregation" clause is the one which defeats the argument of that post. Apr 25 21:16:00 (as long as you don't link against GPL) Apr 25 21:16:31 rwhitby: in private I can link anything to GPL Apr 25 21:16:38 rwhitby: or am I distributing it to myself? Apr 25 21:16:43 zecke: that's true too. Apr 25 21:17:21 good nite guys Apr 25 21:17:26 (that and mere aggregation is why I agree with your interpretation) Apr 25 21:17:36 mickeyl: are you there? Apr 25 21:18:43 nite zecke Apr 25 21:19:14 woglinde: and please ask ritter if he gives you credit points for putting a msp430 toolchain into OE Apr 25 21:21:04 ibot: botmail for mickeyl I would like to give an OE tutorial at the Linux Congress, or any other occassion Apr 25 21:21:17 zecke: you coming to OLS? Apr 25 21:22:53 zecke: I'm here Apr 25 21:23:07 "n a message to its shareholders, the current iPod nano system-on-a-chip (SoC) supplier announced that it had not been selected to supply its chips for new flash-based iPods due in the second half of the year." Apr 25 21:23:14 looks like lrg was right Apr 25 21:23:17 zecke: that sounds like a good idea Apr 25 21:24:04 CosmicPenguin: do they have some kind of 'sponsor' program? Apr 25 21:24:08 koen: what did lrg say? Apr 25 21:24:38 mreimer: we wanted to have a bet how soon apple would ditch portalplayer to use a pxa Apr 25 21:24:50 zecke: heh, I doubt it Apr 25 21:24:52 aha! Apr 25 21:24:59 zecke: whats most expensive - the registration, or the plane ticket? Apr 25 21:24:59 koen: that is old news Apr 25 21:25:14 CosmicPenguin: I had assume the plane Apr 25 21:25:17 koen: wasn't joshua thinking of doing an Apple internship with the iPod division? Apr 25 21:25:20 CosmicPenguin: now looking at the reg costs Apr 25 21:25:47 mreimer: no idea, haven't heard from joshua in a while Apr 25 21:25:55 oh not affordable atm Apr 25 21:26:11 zecke: doh Apr 25 21:26:41 zecke: well, if any of the core OE folks are going to be there, I'm sure that if you posted an inpromptu OE BOF, that you would get lots of folks to attend Apr 25 21:27:10 zecke: what are your plans for robocup btw? Apr 25 21:27:58 RP, do13: http://oe.pastebin.com/681848 (probably the same stuff cromain mentioned) Apr 25 21:28:00 reenoo: I think woglinde and me are waiting to be invented to play technical support Apr 25 21:28:17 hm Apr 25 21:28:25 dont know If zecke has the time Apr 25 21:28:41 woglinde: time, I never have enough of that Apr 25 21:28:43 and I will not have the time to travel to bremen Apr 25 21:29:02 because my son will be born in first july week Apr 25 21:29:06 reenoo: Fu-Fighters are a bit chaotic, I will not know until some days in advance Apr 25 21:29:13 hopefully Apr 25 21:29:14 woglinde: son? Apr 25 21:29:23 zecke son Apr 25 21:29:39 I thought I told you Apr 25 21:29:49 nope Apr 25 21:30:02 woglinde: oh! Apr 25 21:30:19 reenoo: sadly I don't know yet, I hope they have a place for me Apr 25 21:30:49 okay cya Apr 25 21:30:53 zecke I think so, but you have to do soemthing Apr 25 21:31:00 nite zecke Apr 25 21:31:14 zecke: I'll be there. drop me a note if you make it to bremen. Apr 25 21:31:28 woglinde: hehe, now we have automatic color calibration Apr 25 21:31:39 woglinde: I think they don't need one man per robot Apr 25 21:32:06 zecke: the registration deadline was sometime last week. maybe check with the person doing registrations for your team Apr 25 21:32:34 hm I dont think they forgot the registration Apr 25 21:32:51 reenoo the goal is to beat tribots Apr 25 21:32:58 no, but you have to register individual participants Apr 25 21:33:13 since when? Apr 25 21:33:13 haha, haven't seen that on the mailing list Apr 25 21:33:23 hm Apr 25 21:33:39 maybee the world cups rules where diffrent Apr 25 21:33:54 but for the german open there was no personal registration Apr 25 21:34:13 woglinde: dunno. we had to register team members so they can access the "backstage" area. Apr 25 21:34:27 for robocup 2006 that is Apr 25 21:34:54 reenoo hm then we will change the badges so waht Apr 25 21:35:05 well FU Fighters are chaotic, I bet raul forgot to mention it to simon... Apr 25 21:35:58 03rwhitby 07org.oe.dev * r7e6d5167... 10/packages/madwifi/madwifi-ng-modules_svn.bb: madwifi-ng-modules: Updated to latest upstream snapshot. Apr 25 21:41:02 mreimer: ASoC isn't in mainline as pushing it hasn't been attempted yet. Its on Liam's todo list Apr 25 21:41:34 mreimer: It will happen though and is agreed with the ALSA people in principle Apr 25 21:42:28 03mickeyl 07org.oe.dev * r1c63aa4a... 10/packages/visiscript/ (visiscript_0.3.1.bb visiscript_0.3.2.bb): visiscript: catch up with qte/qte-mt changes Apr 25 21:42:32 03mickeyl 07org.oe.dev * r8ece4b6d... 10/packages/anthy/ (6 files in 2 dirs): Apr 25 21:42:32 add anthy, a japanese input method backend library Apr 25 21:42:32 patches courtesy ph5 and Rolf Leggewie Apr 25 21:43:03 RP: thanks for the info Apr 25 21:46:56 rwhitby: My system ran out of space building the image so I don't have anything to flash yet :-/ Apr 25 21:47:07 RP: bummer Apr 25 21:49:19 re Apr 25 21:53:09 koen: For reference, we have patches/hacks in poky to use glib 2.10 with gtk 2.6.8 Apr 25 21:54:37 urghs Apr 25 21:54:41 someone destroy lots of my changes Apr 25 21:54:47 in opie.bbclass and sharprom distro Apr 25 21:54:53 ~lart merges Apr 25 21:54:53 * ibot gets a hotmal account and SPAMs merges Apr 25 21:55:09 mickeyl: :-( Apr 25 21:56:42 mickeyl uh Apr 25 21:57:46 ah here we are :/ Apr 25 21:57:47 http://ewi546.ewi.utwente.nl/tmp/viewmtn/revision.psp?id=f230ea166fd00577cf73bcb2cb81ab9767e16d4c Apr 25 21:57:53 ~lart florian Apr 25 21:57:53 * ibot executes killall -KILL florian Apr 25 21:58:00 * mickeyl tries to disapprove the merge Apr 25 21:58:23 darn Apr 25 21:58:29 monotone: misuse: revision 'f230ea166fd00577cf73bcb2cb81ab9767e16d4c' has 2 changesets, cannot invert Apr 25 21:59:11 any recommendations? Apr 25 22:00:09 hmm, ok i can extract the diffs and apply them again Apr 25 22:00:27 manual merges - just say no. Apr 25 22:00:43 especially when the person is not on irc Apr 25 22:01:32 * CoreDump|home gives up on task-essential-irda for now Apr 25 22:02:02 This might be a good reason for 0.26 as I think the merge algorithms are a lot better? Apr 25 22:02:21 If I can do an automatic merge, or a merge -lca, then I blow away my database and start again. Then at least I can blame monotone if anything goes wrong :-) Apr 25 22:02:36 s/can/can't/ Apr 25 22:02:43 (first one only) Apr 25 22:03:17 good night all ! Apr 25 22:04:42 hmm how can i reverse a diff? Apr 25 22:04:52 ah, just apply it again Apr 25 22:05:39 arghs Apr 25 22:05:42 hunk Failed Apr 25 22:05:42 or use -R Apr 25 22:05:47 mmmh Apr 25 22:05:58 *sigh* Apr 25 22:10:10 hmm out of ideas Apr 25 22:10:19 how can we completely remove the latest changeset? Apr 25 22:14:12 please hold off further commits as I'm trying to revert to r58d737a9422a93415e99e0ad16e64549264b4bb2 Apr 25 22:21:51 njs: It might be useful Apr 25 22:22:03 ite Apr 25 22:22:05 nite Apr 25 22:22:16 njs: It might be useful if you could somehow check the result of a merge before it was committed :-/ Apr 25 22:22:21 'night woglinde Apr 25 22:24:24 RP: do you remember by chance which package contains /etc/modutils/$somefile to configure USB Host or client modules in /etc/modules? Apr 25 22:25:00 CoreDump|home: I'm afaid I don't Apr 25 22:25:13 np Apr 25 22:25:23 CoreDump|home: There is an outstanding bug about some modutils directory (modutils.d) being missing that 2.6 looks for Apr 25 22:28:01 03mickeyl 07org.oe.dev * r6d0cb1a1... 10/ (3 files in 2 dirs): Apr 25 22:28:01 repair fatal damage introduced by florian's changeset Apr 25 22:28:01 http://ewi546.ewi.utwente.nl/tmp/viewmtn/revision.psp?id=f230ea166fd00577cf73bcb2cb81ab9767e16d4c Apr 25 22:28:01 how could that happen??? Apr 25 22:33:02 * raduga added his first bug to the bugtracker Apr 25 22:33:04 yay for me :/ Apr 25 22:33:39 a bug, and a suggested fix for the bug, too. #906 Apr 25 22:33:59 at least, I assume this is the appropriate place to file that. Apr 25 22:36:14 what severity/priority should be given to a bug that breaks building, but is trivial to fix if you know what you're doing? Apr 25 22:36:32 * CoreDump|home dunnos Apr 25 22:36:36 i.e. fetch fails because provider moved location of the source tarball Apr 25 22:38:08 lowest possible imo Apr 25 22:38:10 as long as oesources was up, its unlikely anyone would have noticed Apr 25 22:38:13 mickeyl: thanks Apr 25 22:38:19 np, thanks for reporting Apr 25 22:41:48 g'night Apr 25 22:42:08 'night mickey|zzZZzz Apr 25 22:42:08 'night mickey|zzZZzz Apr 25 23:00:09 'night reenoo|zZzZ Apr 25 23:40:51 stupid uicmoc-native Apr 26 02:46:23 received my RH-2 remote today Apr 26 02:48:25 got it from streamlinecups, and they are awesome (in case of event that someone is interested) Apr 26 04:45:34 RP: yeah. "workspace merge" will do this. It's been "todo" for years (literally), but I think worst case, I think I can get a SoC person to do it now. Apr 26 05:29:44 ping Apr 26 06:32:44 moin Apr 26 06:38:01 http://openzaurus.org/wordpress/2006/04/26/first-release-candidate-of-openzaurus-3541/ Apr 26 06:38:20 zecke: I used quilt on org.openembedded.dev. mickey added anthy to OE yesterday and my daily pull and update via cron incorporated those changes. Now I wonder how I can tell quilt to forget about those patches without doing "quilt push" which removes them? Apr 26 06:39:49 Laibsch: there is a way to skip a patch, but I don#t know how ;) Apr 26 06:42:35 zecke: After I remove all anthy patches, the directory fo course becomes emtpy again. Can't I just do a monotone update and restore things to normal? Apr 26 06:44:21 Laibsch: which patches, what has it do with quilt? Apr 26 06:44:32 Laibsch: tmp/work is not under monotone control Apr 26 06:46:30 I used quilt to import the initial patch from pH5 to "create" the anthy bb files etc in OE. Then I created my own patch so I could have sent that to the bugtracker later on and made my changes to the files via quilt edit. I was working in org.openembedded.dev not tmp/work. Should I have worked in tmp/work? I have no clue what kinds of things are below tmp Apr 26 06:48:13 monotone of course issues warning about missing files. I will use the revert command and I guess then things should be peachy more or less. Apr 26 06:52:07 Laibsch: well, use monotone for that... not quilt Apr 26 06:52:14 the right tool, at the right location Apr 26 06:57:16 Hehe, it is hard for a blind man to judge the flowers ;-) if you know what I am talking about Apr 26 06:57:42 Where is quilt's place? Apr 26 07:01:13 google? ;) Apr 26 07:01:26 somewhere on savannah, or what do you mean by place? Apr 26 07:03:28 hey zecke & Laibsch Apr 26 07:03:41 zecke: I had no problems rosterifying yesterdays .db :) Apr 26 07:03:54 koen: that fast on ewi? Apr 26 07:04:31 zecke: it to a few hours Apr 26 07:04:42 took* Apr 26 07:05:02 morning Apr 26 07:05:40 zecke: http://ewi546.ewi.utwente.nl:3000/trac/timeline Apr 26 07:06:18 that's with yesterday's .db Apr 26 07:06:40 port 3000 is blocked Apr 26 07:06:52 zecke: You said right tool for the right location. I was wondering where that location is with regards to quilt and OE. Apr 26 07:06:59 moin koen. How are you? Apr 26 07:07:09 Laibsch: tmp/work is quilts domain Apr 26 07:07:30 Laibsch: monotone co's should be used with monotone Apr 26 07:07:54 koen: do note that that plugin is all of 48 hours old, of course :-) Apr 26 07:08:54 njs: I know :) Apr 26 07:08:57 what kind of work would one do in tmp/work? Looks like the original sources are there? Apr 26 07:09:10 Laibsch: we do compiling there Apr 26 07:09:10 njs: I'm running the latest and greatest head :) Apr 26 07:09:14 Laibsch: including applying patches Apr 26 07:11:40 OK. But I guess for working on bb files and other OE stuff, I should use monotone. Apr 26 07:12:08 Can I do that without write access to one of the master servers? Apr 26 07:12:17 Is there a real dumb guys intro to using monotone with OE? I know enough to use it in read-only mode. But using it for tracking my changes I do not know about that yet. Apr 26 07:12:34 http://oe.handhelds.org/cgi-bin/moin.cgi/MonotonePhraseBook#preview does not seeem to apply for my situation (no write acces) Apr 26 07:13:02 Laibsch: yes, you can commit locally, create branches Apr 26 07:13:20 Laibsch: e.g I would create a laibsch branch, test my bb files there Apr 26 07:15:05 And I can then apply and unapply patches and create a diff to the current official sources? There are lots of commands in the man page. What are the relevant ones for my case? Apr 26 07:15:35 Laibsch: ah right, this might be more tricky Apr 26 07:15:51 Laibsch: you can do monotone diff and status if you have not comitted it Apr 26 07:16:10 Laibsch: and you can diff between two revisions Apr 26 07:17:35 Hm, but then I should not update in between working on OE, right? I do that daily via cron. How do I create my own branch? Apr 26 07:18:29 morning Apr 26 07:19:06 morning hrw|work Apr 26 07:19:23 Laibsch: http://venge.net/monotone/docs/Branching-and-Merging.html#Branching-and-Merging Apr 26 07:20:56 hey hrw|work Apr 26 07:21:26 zecke: ewi now has a monotone 0.26 server using yesterdays .db Apr 26 07:21:54 koen: http://ewi546.ewi.utwente.nl/tinderbox/showbuilds.pl?tree=Angstrom Apr 26 07:22:49 cute Apr 26 07:22:58 now we need to do regular builds Apr 26 07:23:02 quilt sure was nice, being simple and all. Apr 26 07:23:11 Laibsch: you do regular builds as well? Apr 26 07:23:22 * luke-jr_ used to do regular builds Apr 26 07:23:23 Laibsch: well, then continue to use it (in contrast to what I said) Apr 26 07:25:38 morning Apr 26 07:29:06 bleh Apr 26 07:29:07 | include/asm/arch/tosa.h:190: error: previous declaration of 'tosabl_device' was here Apr 26 07:29:10 | make[1]: *** [arch/arm/mach-pxa/tosa_lcd.o] Error 1 Apr 26 07:31:48 koen: iirc it was discussed during night Apr 26 07:33:11 I can only find reference to patches not applying Apr 26 07:34:26 25 22:32 < cyrilRomain> RP: hi, did you see with Dirk about the conflicting kernel 2.6 patches for tosa (http://oe.pastebin.com/677483) ? Apr 26 07:34:29 25 22:45 < RP> cyrilRomain: He's not been around :-/ Apr 26 07:34:31 25 22:45 < RP> cyrilRomain: I'll look into it and try and resolve it Apr 26 07:34:43 'try and resolve' Apr 26 07:34:52 yeah, the patches apply now :) Apr 26 07:46:05 bonjour Apr 26 07:46:47 zecke: I do not build that much. I run OE on my DSL router. This is not a powerhorse. Apr 26 07:47:10 ah okay Apr 26 07:47:35 I do it to get Japanese support in OZ. Apr 26 07:47:52 But I can of course run an odd build or two. Apr 26 07:50:13 RP, do you think kernel 2.4 could works on poky ? Apr 26 07:51:06 Genesis: Not out the box Apr 26 07:51:45 what could the pb be ? Apr 26 07:51:56 i fear about kdrive for example Apr 26 07:52:10 but my hardware only support 2.4 Apr 26 07:52:56 Your main issues will be the fstab, the lack of 2.4 module tools etc. kdrive should be ok Apr 26 07:54:50 hum i don't know what you mean but i'll try Apr 26 07:55:25 Genesis: everything that smelt like 2.4 was removed from poky Apr 26 07:55:34 Genesis: Which device are you talking about? One already in OE or something new? Apr 26 07:55:36 :// Apr 26 07:55:46 something new Apr 26 07:55:53 the i.mx21 ads Apr 26 07:55:59 something new uses a Linux2.4 kernel? Apr 26 07:56:15 zecke: "new" to OE Apr 26 07:56:19 the board is not really new Apr 26 07:56:35 I did remove everything 2.4 from poky as poky wasn't intended to support 2.4 Apr 26 07:56:43 Genesis: mreimer has worked on 2.6 support for i.mx Apr 26 07:56:59 he doesnt't answer to my queston Apr 26 07:57:13 i think he doest'n work on the mx21 Apr 26 07:57:18 but on the mx1 Apr 26 07:58:27 sure i'd prefer that mx21 was supported by the 2.6 Apr 26 07:58:44 but i work since more than a month , now i hate freescale Apr 26 07:59:14 treke|home: ping Apr 26 07:59:33 oe sources looks rather broken as its serving a php script for everything :-/ Apr 26 07:59:34 RP: Am I right to assume that linux-hotplug should be completly remove from kernel 2.6 images? Apr 26 07:59:42 all seems be done to make people buy a Linux from theirs partners like montavista Apr 26 07:59:55 CoreDump|home: Yes, as its all disabled. I'm not sure if it breaks anything though Apr 26 08:00:27 RP: treke.net got rebooted last night Apr 26 08:00:31 RP: it does break auto-loading of USB Host modules when removed. ohci-hcd needs a /etc/modutils entry to fix that Apr 26 08:00:48 what could i do with my 2.4 ? openembedded ? Apr 26 08:02:47 Genesis: You either add support for it to poky, importing any bits you need from OE, or use OE directly Apr 26 08:03:49 go :) Apr 26 08:04:16 Genesis: What makes poky more attractive to you (just asking)? Apr 26 08:04:19 CoreDump|home: It would be nice if ohci-hcd only got laoded when it detected a connection Apr 26 08:04:46 RP: the problem is that udev ignored USB unless ohci is loaded Apr 26 08:05:39 hey do13 Apr 26 08:06:25 hi dirk Apr 26 08:07:02 CoreDump|home: Tricky. I guess we could add it to the kernel's autoload list Apr 26 08:07:36 or just use /etc/modutils ;) Apr 26 08:07:37 CoreDump|home: but I think I would prefer adding something from the kernel to indicate whether host or client is present, then we can load the appropriate driver Apr 26 08:07:51 RP: that would be neat yeah Apr 26 08:07:53 hi Dirk Apr 26 08:08:13 but AFAICT the usb-host module doesn't mess w/ usb-client stuff Apr 26 08:08:36 CoreDump|home: They use the same output pins... Apr 26 08:09:00 a modprobe g_ether is enough to make it work. no need to rmmod ohci Apr 26 08:09:08 * CoreDump|home dunnos Apr 26 08:09:18 * RP doesn't trust the kernel driver that much Apr 26 08:09:24 =) Apr 26 08:09:50 I guess I know too much about how that code could go wrong :) Apr 26 08:09:59 heh yeah Apr 26 08:10:13 I used g_ether for about two days. Never had any problems Apr 26 08:11:19 I guess the ultimate goal would be to have g_ether and ohci-hcd loaded at the same time :) Apr 26 08:11:29 switching depending on the cable... Apr 26 08:11:32 that's what I was doing, really Apr 26 08:11:55 It can run as an hcd with g_ether loaded? Apr 26 08:12:16 IIRC yes, I'll try in a few minuted Apr 26 08:12:18 minutes Apr 26 08:13:31 yay, Angstrom machine config for SL-Cxx00 is done Apr 26 08:16:20 rebuilding "Packages" of a large feed takes too darn long :\ Apr 26 08:20:01 zecke, i began with OE and i find a lot of pb with it , with poky i succeed to make many different image Apr 26 08:20:25 that seems more usable Apr 26 08:20:34 really. Apr 26 08:21:14 hmm Apr 26 08:21:19 with OE you used .dev? Apr 26 08:24:55 don't know Apr 26 08:25:02 just try to follow the doc Apr 26 08:25:29 just try to find a good doc too Apr 26 08:35:21 RP: Ok, i checked, USB host does not work w/ g_ether loaded... Apr 26 08:37:43 CoreDump|home: That doesn't entirely surprise me Apr 26 08:38:56 zecke, monotone update -rjan2006prebreakage ( i used it ) , when i was trying OE i should install git monotone and a lot of stuff , not easy , not efficient i find Apr 26 08:39:16 i look after gentoo-embedded also Apr 26 08:39:26 git? Apr 26 08:39:56 perharps for debug another packet , don't remember sorry Apr 26 08:41:12 Genesis: okay, to summarise: primary: docs second: getting the source? Apr 26 08:41:29 Genesis: any other topics? Apr 26 08:43:41 perharps some scripts as setdevenv script Apr 26 08:43:56 documentation is the most pb i think Apr 26 08:44:07 perharps i'm a bit silly to find good information Apr 26 08:44:23 Genesis: it is two sided, we need to improve docs for sure Apr 26 08:44:34 Genesis: on the other hand 'automatic' stuff would be too much Apr 26 08:45:05 but as a gentoo user , i'm accustomed with clear doc . Apr 26 08:45:31 Genesis: clear doc yes, it is needed. Cookbooks... Apr 26 08:46:09 zecke: setdevenv is a script in poky that mostly configures bitbake/OE for you Apr 26 08:46:35 RP: ah well :) Apr 26 08:46:55 svn update is very simple Apr 26 08:47:46 i think i see in scratchbox is that we could build in the same time for many different targets Apr 26 08:47:46 Genesis: monotone pull, up was too difficult? Apr 26 08:48:18 RP: I would get sharp serial cable probably Apr 26 08:48:29 i've a lot of work on the board , and i spend my time to read documentation about monotone etc Apr 26 08:48:39 nothing is difficult Apr 26 08:48:43 hrw|work: is the one from SL-5500 working? Apr 26 08:48:58 but documentation is the master word. Apr 26 08:49:10 Genesis: more doc, more time spend reading Apr 26 08:49:21 Genesis: did you know svn from other projects? Apr 26 08:49:26 no Apr 26 08:49:29 zecke: yep - all Z has same I/O port - cxx00 lack usb there and have one-way serial only iirc Apr 26 08:49:31 hrw|work: Sounds good. I wish I had a proper Sharp cable :) Apr 26 08:49:51 hrw|work: I can send you my bulky serial cable Apr 26 08:49:54 RP: my handmade one does not work Apr 26 08:50:18 hrw|work: :-( Apr 26 08:50:33 RP: so when you said mic doesn't work, did you mean using the audio jack as a mic, or the supposed internal mic? Apr 26 08:50:58 hrw|work: http://www.expansys.com/product.asp?code=CE-170TS <- this is the one I have Apr 26 08:51:05 hrw|work: when it might work, I can ship it to you Apr 26 08:51:15 zecke: and I will have this one Apr 26 08:51:15 luke-jr_: Given I believe there is no internal mic, its not going to work, is it? :). I meant the jack Apr 26 08:51:40 RP: so how would I test "line in"? Apr 26 08:51:44 zecke: one user gave it with collie for his 62y father - I would get it from him Apr 26 08:51:46 hrw|work: ah you will? Apr 26 08:51:58 RP: do you want my serial cable? Apr 26 08:52:31 luke-jr_: Connect a line source to it or a mic and try recording? Apr 26 08:52:43 luke-jr_: It doesn't work though, I know that much Apr 26 08:53:07 RP: hrm, and no idea why? Apr 26 08:53:13 zecke: Might you not end up with a Zaurus at some point? Apr 26 08:53:14 RP: reboot seems to work on c700 Apr 26 08:53:21 zecke, : what can i say more , i try first OE , i spend a lot of time to make it works , i failed, i try poky , it works each time , i don't remember exactly what was my pbs with OE but that give me a wrong impression Apr 26 08:53:41 koen: and if I'm right, halt doesn't? Apr 26 08:53:42 RP: I have a SL-5500, no immediate plans to require a serial cable Apr 26 08:53:48 * zecke had it for hacking opie-console Apr 26 08:54:06 koen: Did the RTC survive? Apr 26 08:54:18 Genesis: well that is fine Apr 26 08:54:49 RP: on c760 looks like halt work. Apr 26 08:54:54 treke's mirror broken? Apr 26 08:54:57 zecke: I have a serialio lead which doesn't work with Spitz. Whilst I would like to aquire a proper sharp lead for spitz at some point, I also have no immediate need for it :) Apr 26 08:55:24 hrw|work: I don't think c700 supports halt (750 and 760 bootloaders are different) Apr 26 08:55:29 luke-jr_: yes Apr 26 08:55:31 RP: ok Apr 26 08:55:44 what's the best way to disable using it? Apr 26 08:56:01 RP: haven't looked at the clock Apr 26 08:56:13 RP: I want my ratoc card to work first :) Apr 26 08:56:45 RP: if you need one, I would be glad to ship it to you Apr 26 08:59:02 morning all Apr 26 08:59:56 NOTE: Update cvs://anonymous@cvs.sv.gnu.org/cvsroot/config;module=config;method=pserver;date=20050701 Apr 26 08:59:56 CVS moved to cvs.savannah.[non]gnu.org Apr 26 08:59:57 -.- Apr 26 09:00:02 Genesis: e.g. your build conf should just work Apr 26 09:00:12 luke-jr_: rm your cvs directory... Apr 26 09:00:55 zecke: thanks... fun, now I have more free space Apr 26 09:05:55 hi Dirk Apr 26 09:06:08 zecke: How do i make bitbake install all the modules defined in the defconfig file in the kernel. It only installs some, not all. Apr 26 09:06:20 zecke: I'll keep it in mind if I do need one, thanks. Apr 26 09:06:25 'ipkg install kernel-modules' Apr 26 09:07:22 Hey Richard Apr 26 09:07:39 goxboxlive: I doubt the defconfig file is even used Apr 26 09:08:37 luke-jr_: really? Apr 26 09:08:54 zecke: But i have made a line that it shall do it in the xanadux.bb file. It makes some of the modules, but not all. Apr 26 09:10:16 hrw|work: I'm assuming we have configs for the various devices and don't use defaults... Apr 26 09:10:35 Probably it makes all the modules, but it isent installing all the modules. Apr 26 09:12:31 goxboxlive: you define in machine config which should go into images Apr 26 09:14:10 luke-jr_: would you prefer that bitbake automatically wipes the offending directory? Apr 26 09:16:47 ok, i c. thx. Apr 26 09:26:27 RP: how'd the build go? Apr 26 09:29:01 rwhitby: It completed about 5 minutes ago after a few bumps :) Apr 26 09:29:21 rwhitby: I "overclocked" the device last night :) Apr 26 09:30:22 RP: anything I need to fix in the build? Apr 26 09:30:34 RP: "de-underclocked" is the correct term :-) Apr 26 09:31:13 rwhitby: No, it failed due to the sources mirror becomming broken, then the missing intel zip files Apr 26 09:31:34 yeah, nothing we can do about the intel files - it has to be a manual download Apr 26 09:31:59 there is also a sources.nslu2-linux.org which has everything we've ever needed for all previous builds. Apr 26 09:32:02 rwhitby: I realise that. I've done it now so shouldn't have to do it again for a while :) Apr 26 09:39:09 03coredump 07org.oe.oz354x * rbecabfdc... 10/conf/machine/ (akita.conf spitz.conf): machine: Add Angstrom machine definition for SL-Cxx00 Apr 26 09:39:13 03coredump 07org.oe.oz354x * r5c9d6b6a... 10/packages/meta/ (3 files): meta: Angstrom: Add task-essential-to-boot, task-essential-parts and a sample essential-machine-image.bb, the latter could use some pyhton magic to auto-select the correct parts Apr 26 09:39:20 03coredump 07org.oe.oz354x * r7c3283ec... 10/packages/irda-utils/ (files/init irda-utils_0.9.16.bb): irda-utils: Create /etc/sysconfig if it doesn't exist Apr 26 09:39:21 03coredump 07org.oe.oz354x * rcad73bbd... 10/packages/udev/ (6 files in 6 dirs): udev: Add /dev/loop* to mount blacklist Apr 26 09:39:25 03coredump 07org.oe.oz354x * r6a045307... 10/packages/xserver-common/ (files/softkeys-slcxxxx-xmodmap.patch xserver-common_1.8.bb): xserver-common: Remove comment from xmodmap Apr 26 09:39:29 03rpurdie 07org.oe.dev * r29a0eca7... 10/packages/linux/linux-openzaurus_2.6.16.bb: linux-oz-2.6: Fix tosa build failure Apr 26 09:39:33 03coredump 07org.oe.dev * r94c998d7... 10/packages/irda-utils/ (files/init irda-utils_0.9.16.bb): irda-utils: Create /etc/sysconfig if it doesn't exist Apr 26 09:39:37 03coredump 07org.oe.dev * r4851cae9... 10/packages/udev/ (6 files in 6 dirs): udev: Add /dev/loop* to mount blacklist Apr 26 09:41:13 03mickeyl 07org.oe.dev * r24f21a5b... 10/packages/ (4 files in 4 dirs): Apr 26 09:41:13 add qpobox and poboxserver, japanese input method stuff. poboxserver is not quite ready yet and needs someone familiar with Apr 26 09:41:13 that stuff to look into Apr 26 09:42:06 hrw|work: could you put xserver-common into upgrades? I err...forgot to push a changeset Apr 26 09:45:57 mickey was nice enough to push anthy into OE yesterday. It fails to build with an error that nothing provides anthy-native. Where do I need to look? Apr 26 09:47:04 03coredump 07org.oe.dev * r8d5c66be... 10/packages/meta/ (4 files): meta: Add Angstrom files from oz354x Apr 26 09:48:25 Laibsch: I assume anthy depends on some native stuff but there isnt an anthy-native .bb, if so and you have neccessary stuff on your host you can add ASSUME_PROVIDED += "anthy-native" in conf/local.conf Apr 26 09:49:25 No, I do not. I will need to look into this further. pH5 said his version compiled. Maybe mickey forgot to include something. He reworked a few things. Apr 26 09:52:51 yes, it looks like the patch from pH5 had an anthy-native_6801.bb whereas the version from mickey does not. Apr 26 09:53:06 Or maybe my checked out version if borked. Apr 26 09:54:45 Laibsch: I have no anthy-native either, try adding ph5's one Apr 26 09:55:05 Laibsch: until mickey|zzZZzz awakes Apr 26 09:55:10 ;-) Apr 26 09:55:16 Yes. Apr 26 09:57:49 RP: 'optimise for size' gave 70K on kernel for c7x0 **** ENDING LOGGING AT Wed Apr 26 09:59:56 2006