**** BEGIN LOGGING AT Mon Jun 05 02:59:57 2006 Jun 05 03:16:02 jnc, was more a comment in respect to the fact that apperently the oe bugtracker is no longer at hh.o Jun 05 03:16:53 hopefully its still linked to it, but i havent tried to verify that Jun 05 03:17:43 but france saw it so it's all good Jun 05 03:17:46 :) Jun 05 03:19:08 i don't know what's going on Jun 05 03:20:08 i'm trying to use the beta on my zaurus, you know, to replace the SharpROM. the install process isn't exactly working, and also the install instructions listed require that i give internet access to my zaurus Jun 05 03:20:39 i was trying to do all this with a wifi card and my WPA2 protected network, which sort of looks like it could work, if only wpa_supplicant were functional Jun 05 03:22:44 so following the advices, i hit the link to file an GPE category bug report, and it throws me into the hh.org bugtracker which says hey search for a bug before you file it, i search for the bug and it says that search is broken in that version of the bugtracker Jun 05 03:23:00 i _can't_ be the only person running into this particular problem, can i? Jun 05 03:23:42 not sure Jun 05 03:23:57 i dont think i've ever had a problem searching it that i can remember Jun 05 03:24:38 i am curious tho, how does wpa_supplicant become a gpe bug? Jun 05 03:24:58 i'm following the given directions, is all what i'm doing Jun 05 03:25:24 the wpa_supplicant thing really becomes an OE bug Jun 05 03:25:45 i know about it, i've fixed it, and the fixes were reverted because someone didn't like it Jun 05 03:30:12 so it would be really nice if i could at least follow the given directions. Jun 05 03:38:40 Hmm - ipkg upgrade of netbase is not very friendly. Jun 05 03:39:04 (it kills all your network interfaces, *before* downloading the newer version of the netbase package for installation). Jun 05 03:39:18 It also seems to delete /etc/network/interfaces just for fun too ... Jun 05 03:40:36 which is strange, cause it is listed in CONFFILES. Jun 05 03:51:51 03rwhitby 07org.oe.dev * rec42c808... 10/packages/meta/slugos-packages.bb: slugos-packages: Demoted asterisk - fails to build on uclibc. Jun 05 04:28:56 03justinp 07org.oe.dev * r407ec36a... 10/ (5 files in 4 dirs): mtnpatch: replace split with regex Jun 05 05:14:19 * emte wonders what effect a buffer underflow will have when encoding video files ... Jun 05 05:25:06 ah .. possible frame drop on physical device players Jun 05 06:17:49 Is the monotone server on ewi down? Jun 05 06:40:55 DataBeaver: it should be up Jun 05 06:42:20 Anyone know a way to probe the results of a MM operation after the MACHINE_START MAPIO does it's iotable_init stuff? And I mean from userland. Jun 05 06:42:35 I'd like to see what physical addresses were maped to which virtual addresses Jun 05 06:43:17 koen: Then why do I get this: "mtn: error: I/O failure while talking to peer ewi546.ewi.utwente.nl, disconnecting" Jun 05 06:48:24 You guys know any good channels to ask questions around linux kernel porting to ARM's? Jun 05 06:49:50 Malakai, this is probably as close as you get Jun 05 06:50:18 the core kernel crew dont do arm work for some time now Jun 05 06:50:49 hrmmm Jun 05 06:51:04 and what your asking for you can lookup in the headers Jun 05 06:51:13 assume i don't have them :( Jun 05 06:51:24 i can dissable a dump of the ROM images, and find it Jun 05 06:51:32 but i was hoping I could find it interactively Jun 05 06:51:44 just look in the kernel or in gcc Jun 05 06:52:04 the chip headers with register references are there somewhere Jun 05 06:52:24 Toshiba compiled the kernel. It's running on their MP3 player. The source they gave out as per GPL is missing the values for all the addresses they map to iotable_init Jun 05 06:52:35 most likly in gcc's headers actually Jun 05 06:52:47 I figured most of them out based on the chip's reference manual. but it has 5 AUX addresess i have no clue what they are Jun 05 06:53:20 Malakai: Under the terms of the GPL, doesn't the source they provide have to be complete? Jun 05 06:53:28 probably chip specific features or remapping of device area Jun 05 06:53:39 they don't feel configure options to be 'source' Jun 05 06:53:52 and the AUX1 -> AUX5 addys are kernel configure variables Jun 05 06:54:08 Ah Jun 05 06:54:15 as are their IOCS1, IOCS2, FLASH_START and others, but we can figure thouse out. Jun 05 06:54:30 i'll take another looks the the from image Jun 05 06:54:44 maybe it's obvious now that i know the addys of the other locations Jun 05 06:54:59 there is a chance they could be bus addresses as well Jun 05 06:55:41 hard to say without knowing the schem/devices onboard Jun 05 06:56:14 even knowing its still hard. heh. we have it documented pretty well: http://www.rockbox.org/twiki/bin/view/Main/GigabeatInfo Jun 05 06:56:52 Someone got the i2c to the LED backlight figured out. The inputs for all the buttons/touchpads are mapped. Jun 05 06:57:05 We can also force it into USB ATA mode on demand. we're gettting close. Jun 05 06:57:45 I'd like to put something other than 2.4.18-rmk6 on it. So I don't have to be dependent on the Toshiba OHCI driver they wrote as 'proprietary'. No source. Jun 05 06:58:08 pretty sure near full support for the S3C2440 is in a 2.6 kernel Jun 05 06:59:39 i kind of doubt that Jun 05 06:59:51 but thats just me Jun 05 07:00:18 2.6 arm kernels are pretty rare Jun 05 07:00:23 the s3c24xx is pretty well supported Jun 05 07:00:34 even the familiar ones for ipaqs arnt all there Jun 05 07:00:42 emte: yes, since all the changes are in mainline now Jun 05 07:00:44 i'm not sure on the kernel number Jun 05 07:00:48 but i saw a lot of patches on S3C2440 Jun 05 07:01:23 assuming I can build and run 2.4.18-rmk6 with the Toshiba patches, then i'll look at puting on a higher version kernel with cleaner S3C2440 support Jun 05 07:01:38 the Toshiba support for it looks nothing like the later implementations. very hackish Jun 05 07:02:00 Malakai: I know the wolfson sound chip is well supported :) Jun 05 07:02:03 ~seen lrg Jun 05 07:02:19 lrg was last seen on IRC in channel #oe, 2d 16h 3m 48s ago, saying: 'hrw|work, mickeyl: I'll be getting my server in a month or two. I'm fine for disk space, but what's the bandwith like per month ?'. Jun 05 07:02:23 We haven't played with the DAC yet. Jun 05 07:03:01 I know the RockBox people will have to implement their own support for everything. They aren't Linux based. I'd like to just use the thing as my own embedded linux play-thing. Jun 05 07:03:28 Malakai: lrg works at wolfson and has written alsa drivers for their chips Jun 05 07:03:39 and the 8571 is pretty common Jun 05 07:04:00 cool. i'll make sure the team knows Jun 05 07:04:09 so we don't spend too much time trying hack at that Jun 05 07:05:19 yawn ... nuff bitbanging protocols tonight Jun 05 07:05:28 time for sleep Jun 05 07:09:24 morning Jun 05 07:11:44 hey XorA Jun 05 07:13:12 koen: what yah broken? Jun 05 07:14:33 lots o'stuff :) Jun 05 07:15:34 koen: noticed you summoned the Liam :-) Jun 05 07:15:39 bitbake is based on gentoo portage? Jun 05 07:15:58 univac: it *was* based on portage long ago Jun 05 07:16:06 note the past tense Jun 05 07:16:21 mhm Jun 05 07:16:23 now it is far better :-D Jun 05 07:19:24 yeah, bitbake -funrollloops -fOMG1337 didn't work to well Jun 05 07:19:40 :)) Jun 05 07:19:51 < gentoo user ;p Jun 05 07:20:04 instead we CFLAGS="-Osecretarmoptims" bitbake :-) Jun 05 07:20:42 ok back to making abiword work Jun 05 07:21:38 XorA: did you add your insights to the abiword bug you opened? Jun 05 07:22:16 koen: not yet, I know exactly why it fails to work, but Im trying to work out why the configure script is creating that situation on arm and not x86/amd64 Jun 05 07:27:02 morning Jun 05 07:27:10 hey eFfeM Jun 05 07:27:24 ~lart redboot Jun 05 07:27:24 * ibot installs PocketPC on redboot's PDA Jun 05 07:28:03 koen, you're early (or I just plain late) Jun 05 07:28:21 eFfeM: yeah, I needed to call expansys at 8 am Jun 05 07:28:48 * eFfeM is restoring yesterdays mess, moved the downloads dir for the gcc 4.1.1. test then nuked it) Jun 05 07:29:07 pff, i try to avoid real work on a free day. Jun 05 07:29:24 8am? where are they? bad time to call to the US Jun 05 07:29:33 I'm trying to get me a780 delivered before I leave :) Jun 05 07:29:46 holiday? Jun 05 07:29:57 yes Jun 05 07:30:40 expansys.nl is in haarlem (023 area code) Jun 05 07:30:49 aren't they closed today ? Jun 05 07:31:03 going to some exotic location? Jun 05 07:31:09 I was affraid so, but I just got of the phone with them :) Jun 05 07:31:09 * eFfeM still has 5 weeks to go Jun 05 07:31:50 eFfeM: barbados Jun 05 07:32:30 2.5 weeks without internet and gadgets Jun 05 07:32:47 hmm. not bad at all Jun 05 07:32:47 (except when I empty my cam to my pwerbook) Jun 05 07:32:50 koen: heh heh Jun 05 07:32:58 koen: take GPRS :-D Jun 05 07:33:02 :) Jun 05 07:33:04 EDGE! Jun 05 07:33:24 i'll must buy phone with EDGE Jun 05 07:33:29 been to canada last year, I actually maintained a small weblog for family and friends whenever I encountered a hotel with internet Jun 05 07:33:38 UTMS! Jun 05 07:34:25 maybe I'll say hi here at #oe when I'm at prpplague's place Jun 05 07:35:07 koen, we'll miss you here Jun 05 07:36:50 too bad I can't escape the soccer stuff Jun 05 07:37:16 he he Jun 05 07:37:54 * eFfeM is going to India june 20. no soccer there (india is cricket country) Jun 05 07:38:09 hehe :) Jun 05 07:38:30 barbados will host the world championship of cricket next year :) Jun 05 07:40:39 morning Jun 05 07:40:43 koen: well in 2.5 weeks I should be able to fix abiword :-) Jun 05 07:40:56 hey hrw|work Jun 05 07:41:33 heh.. two days ago machine died Jun 05 07:42:16 should a package depend on install-native or is that so generic it should really be broght in by base.bbclass or suchlike? Jun 05 07:42:56 I suspect base.bbclass Jun 05 07:43:26 koen: the interesting fails you get when you build apps without doing bitbake gpe-image first Jun 05 07:44:31 hrw|work: which? Jun 05 07:44:57 univac: my at work - I had to reboot it Jun 05 07:45:07 x86? Jun 05 07:46:44 yep Jun 05 08:14:01 ~lart out of tree modules Jun 05 08:14:01 * ibot slaps out of tree modules upside and over the head with one freakishly huge killer whale named hugh Jun 05 08:14:32 well thats the end of OE then Jun 05 08:18:44 XorA: what's the end? Jun 05 08:28:32 koen: you squashed it with a whale Jun 05 08:28:43 ah Jun 05 08:28:55 I wanted to squash the sn9c102 modules Jun 05 08:30:06 koen: just try again, but please save the whales Jun 05 08:30:18 ~lart sn9c102 Jun 05 08:30:19 * ibot chops sn9c102 in half with a free AOL CD Jun 05 08:32:02 koen: there is no way you can do fine work with a whale as your weapon of choice :-) Jun 05 08:33:25 XorA: When all you've got is a whale... every problem looks like fish.... Jun 05 08:36:35 ... or a Captain Ahab ... Jun 05 08:39:42 03koen 07org.oe.dev * r64a93d31... 10/classes/kernel.bbclass: classes/kernel.bbclass: also stage include/media Jun 05 08:41:42 hrm Jun 05 08:41:43 | arm-linux-ld: unrecognized option '-Wl,-rpath-link,/data/build/koen/OE/build/tmp/openzaurus/staging/arm-linux/lib' Jun 05 08:42:36 eFfeM: I vaguely remember someone saying he had a new .bb for asterisk Jun 05 08:42:44 eFfeM: but it could be my imagination Jun 05 08:44:59 koen, he said in a followup that mt 0.25.2 works. I'm not sure if I've seen a msg about a bb file. Given the fact that he was just struggling with the right mt version I would be surprised if he came up with a bb file Jun 05 08:45:37 anyway, if there is a newer version, I'll happily add it for him or if he has done a good job you can of course give him access Jun 05 08:47:02 koen: wrt your ld message: the , after rpath-link looks suspect, I would have expected an = there Jun 05 08:47:13 but I don't know the ld syntax by heart Jun 05 08:47:31 LDFLAGS="" solved it Jun 05 08:52:16 ~seen CoreDump|home Jun 05 08:52:29 coredump|home is currently on #gpe #oe #openzaurus #handhelds.org. Has said a total of 322 messages. Is idling for 2d 15h 12m 1s, last said: 'killing ipaq-sleep should help'. Jun 05 08:53:36 * koen is victorious Jun 05 08:54:47 03koen 07org.oe.dev * r84d44ee2... 10/packages/sn9c102/sn9c102_1.32.bb: sn9c102: add 1.32, needs >= 2.6.16 to build Jun 05 08:55:43 03koen 07org.oe.dev * r84d44ee2... 10/packages/sn9c102/sn9c102_1.32.bb: sn9c102: add 1.32, needs >= 2.6.16 to build Jun 05 09:03:44 /msg nickserv set unfiltered on Jun 05 09:05:21 hi Liam Jun 05 09:05:32 hey Marcin Jun 05 09:07:11 hey liam Jun 05 09:07:26 hey Koen Jun 05 09:08:38 oh dear Jun 05 09:08:43 tomorrow is 6-6-6 Jun 05 09:10:02 eFfeM: no, that comma after rpath-link is correct Jun 05 09:10:19 ok Jun 05 09:10:31 as I said I am not an expert, it just looked suspect Jun 05 09:10:42 ibot, botmail for zecke: hail zecke. no, never tried the vm_swappiness thing since most gpe platforms don't have any swap. I'd be interested in your experiences if you decide to try it out though. Jun 05 09:12:25 XorA: a package should depend on install-native if it particularly needs that (patched) version of "install" rather than the normal one from the host. generally, that's only true for old style gpe programs. Jun 05 09:13:22 pb__: ok, cheers for that info Jun 05 09:13:46 pb__: this is cups I fell over on Jun 05 09:14:04 As long as you have a USB 1.1 host controller on a SoC, you can use it for keyboard support, correct? Host Controller for keyboard, device controller side for being plugged into anther host controller. right? Jun 05 09:14:39 "The groups for sale include: [..] Intel's communications processor business, including the IXP network processor [..] Intel's Xscale business Jun 05 09:14:47 " Jun 05 09:14:58 koen: got a few million spare? Jun 05 09:15:05 Malakai: yes Jun 05 09:15:51 XorA: let's persuade cosmicpenguin to let amd buy it :) Jun 05 09:15:54 interesting. that means the gigabeat mp3 player could have a keyboard or mous attached to it Jun 05 09:17:19 koen: where did you see this ? Jun 05 09:17:24 Malakai: if it provide power on usb host connector Jun 05 09:17:40 lrg: http://www.mercurynews.com/mld/mercurynews/business/technology/14733041.htm Jun 05 09:18:59 hrmm Jun 05 09:19:27 lrg: but no pressrelease from intel Jun 05 09:20:26 koen: thanks. Jun 05 09:21:02 hew|work: do some USB type A connectors not provide power? Jun 05 09:22:49 Malakai: some devices don't provide power to their bus Jun 05 09:23:23 you can hookup and external +5V to the vcc line if you want Jun 05 09:24:10 well, next time i take apart the cradle, i'll measure the connector for power Jun 05 09:24:18 the cradle itself is powered Jun 05 09:25:49 wow, my cc toolchain finished building. that only took 12 hours Jun 05 09:26:17 and 4,200 patches. Jun 05 09:27:35 Malakai: does this mp3 player really have a host controller? Seems pretty odd Jun 05 09:28:03 hi rwhitby Jun 05 09:28:12 this mp3 player is so not a mp3 player. they only cloak it as one Jun 05 09:28:24 http://www.rockbox.org/twiki/bin/view/Main/GigabeatInfo Jun 05 09:28:28 that's the details we got on it. Jun 05 09:28:50 the S3C2440 SoC gives us the OHCI Jun 05 09:29:13 EZ-USB there's a AT2LP 2.0 ATA bridge on it too Jun 05 09:29:23 for 2.0 speed straight to hard drive Jun 05 09:33:33 Malakai: looks good, but i can't find any availability in NL Jun 05 09:34:18 they are phasing them out for the WinCE models Jun 05 09:36:21 bad. the f40s is $ 199 on amazon.com Jun 05 09:36:23 :-| Jun 05 09:37:45 you'll see cheap MEG-Fxx and MEG-Xxx come on the market, as fanboys rush to the new S series Jun 05 09:39:41 I guess they switched to wince because it's far superior to linux ? Jun 05 09:39:47 yeah but probably not in NL, might be worthwhile to look at them on ebay though Jun 05 09:40:12 jacques: probably they move to wince because of better support for windows drm Jun 05 09:40:32 I think they switched to wince because we kept nagging them for the GPL code they owe us Jun 05 09:42:19 eFfeM, yes, that makes sense. sigh. Jun 05 09:42:33 They had their own DRM on it before. Jun 05 09:42:43 everything gets put onto it in SAT encryption Jun 05 09:43:13 you can't just hook it up to your machine as a USB device and pull off the files. Jun 05 09:43:31 Malakai, I love how companies *cough* iomega nas100d *cough* think that killing a product or changing its OS can excuse them from complying with the GPL. Jun 05 09:43:37 you have to go through some special usb driver/filter. which is not supported on linux. Jun 05 09:43:44 so the thing runs linux, but can't be access from it. Jun 05 09:43:47 hmm too bad Jun 05 09:43:51 Thus, the need for gigahacking. Jun 05 09:43:53 hi zecke Jun 05 09:44:27 so it is not hacked yet Jun 05 09:44:28 Someone found out that if you plug in a USB NIC with the pegasus chipset, it'll DHCP and open samba shares to it's drives. Jun 05 09:44:31 that was it's downfall. Jun 05 09:44:42 oh it's hacked, just not completely Jun 05 09:44:44 ah, ok Jun 05 09:44:46 pb__: I agree that most platforms lack swap, but the kernel got swap compiled in Jun 05 09:44:56 the from was encrypted, that got cracked about 2 weeks ago Jun 05 09:45:09 so now new images can be placed on the HD, and will be loaded in during boot Jun 05 09:45:09 zecke: on our 2.4 kernels you can't easily avoid that Jun 05 09:45:09 pb__: and from a quick glance vm_swappiness not only controls swap but also trhowing away unused pages Jun 05 09:45:19 well my usb pegasus cable had its id's just added in 2.6.16 Jun 05 09:45:27 zecke: ah, right Jun 05 09:45:39 if you make a mistake though, you gotta open it up, unplug the HD, turn it on, then plug in the hd, it'll be in ata mode Jun 05 09:45:46 and you can fix your mistake. Jun 05 09:45:52 pb__: I will ask CoreDump|home to experiment :) Jun 05 09:45:58 need to get u-boot or something on it. Jun 05 09:46:04 zecke: very good Jun 05 09:46:10 it did not stop my Shark to OOM :} Jun 05 09:46:38 Malakai: this asks for a boot script (like the flag that used to be there on the nslu) and/or kexec Jun 05 09:47:10 eFfeM: well, this is shipped with 2.4.18 Jun 05 09:47:44 does it ask for a bootscript? i dont know Jun 05 09:47:59 oh well, might dig into this later Jun 05 09:48:03 it has it's own boot loader written by Toshiba. we've disassbled enough to know it has xmode support and serial stuff Jun 05 09:48:08 but the serial ports aren't exposed Jun 05 09:48:18 * eFfeM heads back to reading this boring thesis Jun 05 09:48:32 Malakai: can't you compile your own modules and load them ? Jun 05 09:48:42 if you make a dir on it's harddrive called "update" and put a special named .sh file in it, then you can have it load stuf fon boot Jun 05 09:49:03 eFfeM: they didn't give us there kernel config options as part of the GPL release Jun 05 09:49:13 and it reference 5 AUX addresses that we have no idea what they are Jun 05 09:49:28 we're experimenting and trying to rebuild the kernel it shipped with. but it's slow going Jun 05 09:49:53 eventually i'd like to put 2.6 on it which has native S3C2440 support Jun 05 09:49:58 ah yes, that are slow processes with a lot of trial and error Jun 05 09:50:31 i need a way to examine the iotable from useland via a .sh script Jun 05 09:50:38 Malakai: /proc/config.gz? Jun 05 09:50:54 if you can get kexec to run on it testing it becomes a lot simpler, but for now I did not manage to get kexec going on my nslu2 with arm be Jun 05 09:50:56 will that show me the physical/virtual maping given during MAPIO in MACHINE_START? Jun 05 09:50:58 NAbyss_: that's a 2.6 feature Jun 05 09:51:05 koen: Ah, smeg. Jun 05 09:51:12 Malakai: devmem2? Jun 05 09:51:56 looking it up now. never heard of it. Jun 05 09:52:20 was used on ipaqs Jun 05 09:52:26 Malakai: http://lists.arm.linux.org.uk/pipermail/linux-arm-kernel/2000-September/000018.html and http://www.lart.tudelft.nl/lartware/port/devmem2.c Jun 05 09:52:29 quite useful Jun 05 09:52:35 See, i don't need to read/write, i need to know what they passed to iotable_init Jun 05 09:52:57 What's the physical offsets passed in for AUX1, AUX2, AUX3...to 5 Jun 05 09:54:14 if you look here: http://rafb.net/paste/results/JbXzZu72.html Jun 05 09:54:17 line 67 Jun 05 09:54:35 that structure, most of it comes from options put into the kernel configure Jun 05 09:55:49 the user manual for the S3C2440 gives us most the obvious onces (IO_START, FLASH_START, SRAM_START, IOCS_START, IOCS3_START, IOCS4_START) Jun 05 09:55:57 but the AUX stuff is mystifing Jun 05 09:56:33 can't you just look at map.h? Jun 05 09:56:44 roaming around in /proc (and if available /sys) can sometimes be *very* informative Jun 05 09:57:13 I don't think those mappings are exposed anywhere convenient in user space. You'd need to poke around the swapper pagetable by hand, which would be a bit tedious. Jun 05 09:58:40 zecke: check our pagerank Jun 05 09:59:23 for now i just put those values in la-la land. I'm hoping something secondary like the DAC uses them. Jun 05 09:59:52 if i can get a self compiled kernel on, then i can set those addr's to a known place and put break points on them Jun 05 10:00:12 well, assuming i can get usbdnet to work in 2.4.18 Jun 05 10:00:27 otherwise i'll be writing a log file to the hd, and bricking it each time to reset it Jun 05 10:00:30 koen: how do I do that again :} Jun 05 10:00:45 zecke: google.com -> "openembedded" Jun 05 10:01:48 koen: need to improve cross compile Jun 05 10:01:51 koen: ARM Linux :} Jun 05 10:02:08 i think my crosscompile toolchain failed: Jun 05 10:02:18 hello: ELF 32-bit LSB executable, Intel 80386, version 1 (SYSV), for GNU/Linux 2.2.0, dynamically linked (uses shared libs), not stripped Jun 05 10:02:45 my hello.c with CC=arm-linux-gcc, ended with 80386 instead of ARM Jun 05 10:02:46 Malakai: you wanted to build for i386? Jun 05 10:02:49 no Jun 05 10:02:56 i thought the CC export would handle that Jun 05 10:03:03 do i have to use gcc options? Jun 05 10:03:05 how did you compile this hello.c? Jun 05 10:03:25 export CC=arm-linux-gcc Jun 05 10:03:25 gcc -o hello hello.c Jun 05 10:03:48 Malakai: and you think those things relate? Jun 05 10:04:01 $CC -o hello hello.c would do it Jun 05 10:04:22 i haven't looked what CC does Jun 05 10:04:23 Malakai: gcc is a C Compiler, why should it consider looking at CC? :) Jun 05 10:04:26 i was basing this off of: http://72.14.209.104/search?q=cache:YcThliqx09UJ:www.ailis.de/~k/docs/crosscompiling/toolchain.php Jun 05 10:04:34 never done anything like this before. Jun 05 10:05:29 "That's it. You should now have a working cross-compile toolchain for the ARM platform. If you want to cross-compile a program just set the CC environment-variable to arm-linux-gcc and compile the program as usual. " Jun 05 10:05:31 Malakai: use arm-linux-gcc, it is the compiler Jun 05 10:05:45 Malakai: but I fear this is totally Off Topic here Jun 05 10:06:09 Malakai: this site assumes you have Makefiles where $(CC) is used in a rule... Jun 05 10:06:38 yeah, i was just looking for a channel that has some people knowledgable in CC'ing Jun 05 10:07:49 Malakai: and now you are fine? Jun 05 10:08:22 yup. not sure what the CC variable is for though, if you have to explictly call /usr/local/bin/arm-linux-gcc directly Jun 05 10:09:08 Malakai: CC is shell.... Jun 05 10:09:14 Malakai: used in Makefiles... Jun 05 10:09:22 Malakai: shell is that thing that Unix has Jun 05 10:09:47 zecke: and shellfish Jun 05 10:09:58 right Jun 05 10:10:06 i see. i'm really tired for building and patching all this. Jun 05 10:10:18 Malakai: this is why we created OpenEmbedded Jun 05 10:10:31 Malakai: patch stuff exactly one, by one or two persons Jun 05 10:10:42 Malakai: always be able to build it, everywhere, anytime Jun 05 10:11:19 well good. when we finish opening up the Gigabeat, it can benefit from that. Jun 05 10:11:41 until then, i'm trying to build the kernel like their developers did to eliminate variables. Jun 05 10:14:55 cya later Jun 05 10:33:07 good morning Jun 05 10:33:38 florian: good morning! Jun 05 10:42:35 03xora 07org.oe.dev * r815a6b1e... 10/packages/abiword/abiword_2.4.4.bb: Jun 05 10:42:35 abiword_2.4.4.bb : add -Wl,--export-dynamic to the LDFLAGS in the .bb Jun 05 10:42:35 file. On x86 builds it seems to get this argument by luck from gmodule Jun 05 10:42:35 build arguments. This is needed to get abiword plugins to see symbols Jun 05 10:42:35 defined within abiword, and thus abiword-plugins now work. Jun 05 10:43:45 hrw|work: did oz 3.5.4.1 release yet? Jun 05 10:44:09 XorA: have to release it Jun 05 10:44:39 hrw|work: that change might be a nice one to include then, opendocument plugin should hopefully work now, I lack any opendocument docs :-) Jun 05 10:45:20 morning Jun 05 10:45:39 XorA: I don't think it's luck: the gmodule pkg-config data sounds like the right place for that option. If it isn't showing up there on arm then anything that uses gmodule is likely to break. Jun 05 10:48:34 pb__: ok, so its a bug in our gmodule, I shall look into that then, that option is certainly in my amd64 version but not my arm version Jun 05 10:48:36 hi lardman Jun 05 10:49:48 XorA: yeah, seems that way Jun 05 10:51:40 nick cbrake Jun 05 10:51:42 * XorA dreams of a configure debugger Jun 05 11:10:18 * koen reads http://www.openembedded.org/viewmtn/revision.psp?id=815a6b1e0db931b81db4fe7f65116e911d84d8a4 Jun 05 11:12:10 XorA: yay! Jun 05 11:14:41 ~change 12 gbp to pln Jun 05 11:14:47 12.00 Pound (GBP) makes 69.0815 Zloty (PLN) (from http://www.xe.com/) Jun 05 11:16:29 :) Jun 05 11:21:16 pb__: I think Ive found the problem, glib-2.0 called ./libtool in configure (comes from configure.in) but the glib build directory doesnt have a ./libtool Jun 05 11:21:43 pb__: as far as I can gather thats how it adds the -Wl,--export-dynamic to the LDFLAGS Jun 05 11:23:22 ah Jun 05 11:23:33 yeah, that could be a problem Jun 05 11:23:46 can you teach it to call ${HOST_SYS}-libtool instead? Jun 05 11:24:01 pb__: yes, that would be a simple patch Jun 05 11:24:42 actually, that might fail in a native build. maybe it needs to only use the HOST_SYS version when configured cross (i.e. HOST_SYS != BUILD_SYS) Jun 05 11:24:58 hail mickeyl Jun 05 11:25:03 hey pb_ Jun 05 11:25:57 you are trained reading schematics? Jun 05 11:26:12 pb__: is there a simple way to do that logic in .bb language Jun 05 11:26:49 mickeyl: linux-ezx if now fixed up properly Jun 05 11:26:50 could you have a quick look at http://vanille.de/temp/keypad.jpeg and tell me what the symbol next to INDICATOR_B means? Jun 05 11:26:53 koen: thanks! Jun 05 11:27:55 mickeyl: it seems OE unpacks the patches in do_patch, so it's legal to rm ${S}/patches before that Jun 05 11:28:21 koen: that's pretty cool. it eases my maintenance since I'm going to have more and more patches on top of ezx6 Jun 05 11:28:38 hi mickeyl Jun 05 11:28:40 mickeyl: earth Jun 05 11:28:45 mickeyl: the right-pointing arrow-tag thing? it means that a signal is coming into this schematic from somewhere else. Jun 05 11:29:03 mickeyl: below is a ground symbol (0 volts). Jun 05 11:29:52 hmm.... B+ comes from somewhere else with +4.42v. So if I want to turn the INDICATOR_B on, then I just need to enable B+ ? Jun 05 11:30:09 OMG someone ported MegaZeux to Linux Jun 05 11:30:09 XorA: probably best to put the logic in the configure script itself Jun 05 11:30:18 finally a new use for my Z now that I'm not in college! Jun 05 11:30:44 pb__: ok, I shall attempt some shell scripting Jun 05 11:30:44 bbiab, gotta shut down my pc for electrical testing Jun 05 11:30:48 XorA: good luck Jun 05 11:31:33 mickeyl: J606 is probably a connector? Jun 05 11:32:28 cbrake: yes, i think j606-3 and j606-2 are connected to a LED. I just can't find how to enable this programmatically Jun 05 11:35:17 mickeyl: hard to tell. So can you measure INDICATOR_B and verify you can set it high and low? Jun 05 11:35:43 cbrake: unfortunately not even that. INDICATOR_B and B+ are connected to an ASIC where I just have fragmentary docs Jun 05 11:36:13 mickeyl: e680 or a780? Jun 05 11:36:18 koen: e680 Jun 05 11:36:34 i can't find where they are setting the blue LED Jun 05 11:36:38 it's not in the kernel sources Jun 05 11:36:41 or I am blind Jun 05 11:36:46 (the latter being more likely) Jun 05 11:37:16 * NAiL gives mickeyl his glasses Jun 05 11:37:18 makes me crazy seeing that LED being set but I can't do it Jun 05 11:39:09 mickeyl: VAB_IO may also be involved. Jun 05 11:43:21 *sigh* Jun 05 11:43:29 ~lart ASICs without specs Jun 05 11:43:29 * ibot takes ASICs without specs to the vet for a "special" visit Jun 05 11:43:43 the schematics don't buy me too much Jun 05 11:45:16 do you think it would be worth the effort asking MOTO for specs? now that they are seemingly so open source friendly? Jun 05 11:45:32 mickeyl: Can't hurt.. Jun 05 12:15:48 mickeyl: It could be the led is hardwired to some functionality and can't be controlled independently? Jun 05 12:20:46 RP: on the e680 i have a "power save" feature in the UI. When I turn this off, the blue LED goes on. Jun 05 12:21:11 of course, i have no idea what the UI does by changing this value Jun 05 12:21:44 mickeyl: Do you suspect its a GPIO or connected to something else? Jun 05 12:22:15 You have source code for controlling the ASIC under 2.4? Jun 05 12:22:17 RP: I'm sure it's not connected to a pxa-gpio. I'm afraid it's connected to the PCAP ASIC Jun 05 12:22:32 RP: Yes, I have the 2.4 source code, but I can't find any reference to this LED :/ Jun 05 12:22:53 mickeyl: What about any kind of power saving mode? Jun 05 12:23:02 03xora 07org.oe.dev * r274c7a01... 10/packages/abiword/abiword_2.4.4.bb: disapproval of revision '815a6b1e0db931b81db4fe7f65116e911d84d8a4' Jun 05 12:23:12 mickeyl: Can you query the status of the ASIC? Jun 05 12:23:46 mickeyl: I'm wondering if you can write a 2.4 module that would read back the status so you could see what changed when you enable the power saving mode Jun 05 12:24:02 RP: right. that would work. i'm afraid that's my only option Jun 05 12:24:21 this kind of reverse engineering is very tedious :/ Jun 05 12:24:31 hats off to all people who don't have source! Jun 05 12:24:42 You can see why kernel development takes a while :) Jun 05 12:24:50 yeah :) Jun 05 12:25:18 This is why the hx2750 is a pain. We know nothing apart from the few facts I could get out of Jamey :-/ Jun 05 12:26:01 RP: hx2750 boot only from haret? Jun 05 12:27:08 hrw|work: yes Jun 05 12:27:57 hrw|work: Its a hard device to get at the JTAG pins on and the standard bootloaders have no recovery option Jun 05 12:30:22 Sounds like the era of linux ipaqs is over :-( Jun 05 12:31:49 florian: we need to sentence all PHB's to death Jun 05 12:46:45 RP: just checking me mail system is working, did you get the mail I sent about Liam's latest patch rc9 Jun 05 12:47:56 XorA: did it work ? Jun 05 12:48:12 lrg: I assume you didnt get email Jun 05 12:48:31 lrg: it allows playback, but is noisy as hell when you close the audio device Jun 05 12:48:39 lrg: havent had time to check record yet Jun 05 12:48:54 lrg: and speaker output doesnt work (silent no matter the mixer settings) Jun 05 12:50:08 XorA: thanks, although I'm not sure why spk is broken..... don't remember changing it :-\ Jun 05 12:50:59 XorA: what sort of noise do you hear at close ? Jun 05 12:51:09 is it a pop ? Jun 05 12:51:28 lrg: a burst of pops very close together, which is really quite loud static sounding, then 4/5 pops about 1s apart Jun 05 12:52:43 XorA: can you come to westfield one night ? Jun 05 12:53:14 lrg: probably, I think you said that was out murryfield way? Jun 05 12:53:49 XorA: yep, it's right next to murryfield. Jun 05 12:53:58 lrg: have bicycle :-) Jun 05 12:54:14 XorA: let me set up some corgi builds first then I'll get back to you... Jun 05 12:54:43 XorA: sorry about this, I have no way to test this..... Jun 05 12:54:55 lrg: I know, but at least I am in the local :-) Jun 05 12:55:10 XorA: indeed :) Jun 05 12:55:26 mickeyl: did someone not mention a spare c7x0 device available for developing at moment? Jun 05 12:57:01 XorA: http://www.openembedded.org/wiki/ProjectDevices Jun 05 12:58:04 koen: I think I have the correct fix for abiword now, just waiting for pb to rejoin Jun 05 12:58:15 even better Jun 05 12:58:24 koen: bug in glib build Jun 05 13:01:13 koen: edited it a bit Jun 05 13:01:42 lrg, XorA: I'd offer to help debug but I have no time atm :-( Jun 05 13:01:55 hrw|work: cool Jun 05 13:02:27 koen: links should not be shown as urls Jun 05 13:02:51 yeah, leftovers from the old wiki Jun 05 13:03:08 I did the same for http://www.openembedded.org/wiki/PapersAndPresentations Jun 05 13:03:08 RP: np, I think its my fault. Jun 05 13:05:15 XorA: yes, I think hrw has one available atm. I have as well, but mine is scheduled to go to zecke soon Jun 05 13:33:17 mickeyl: do you know how to rescue bricked collie? user reported that even diagmenus does not work Jun 05 13:34:38 jtag Jun 05 13:34:52 that's the only thing that can help bricked devices Jun 05 13:35:01 (otherwise it isn't bricked) Jun 05 13:35:17 koen: yep Jun 05 14:02:05 03xora 07org.oe.dev * r520e9fe2... 10/packages/cups/cups_1.1.23.bb: cups_1.1.23.bb : needs install-native to build Jun 05 14:02:10 03xora 07org.oe.dev * r37934d96... 10/packages/libidl/libidl_0.8.6.bb: libidl_0.8.6.bb : needs libidl-native to build Jun 05 14:02:14 03xora 07org.oe.dev * r063a83fd... 10/packages/fbpanel/fbpanel_4.3.bb: fbpanel_4.3.bb : add missing xlibs dependancies Jun 05 14:19:33 I'm using the PXA270 and I would like to add support for miniSD, is it already supported in the Linux kernel for the PXA270 or will I need to add it? Jun 05 14:21:11 IIRC the generic code is already in place, you will just need to add your board glue code. Jun 05 14:22:19 mickeyl, ok and I just grep'ed the arm/pxa source for the kernel and it looks like at least some of the glue code is there for the PXA MMC (SD) controller so I guess it should be pretty easy. Jun 05 14:22:45 where is the feed for the 3.5.4.1 openzaurus image in testing? Jun 05 14:23:01 i'm looking for a newer version of wpa_supplicant that will actually work Jun 05 14:23:32 (i.e. the wpa_supplicant that gets installed with the GPE image is incompatible to the kernel interfaces, I think) Jun 05 14:30:08 jnc: see #openzaurus or www.openzaurus.org Jun 05 14:30:21 mickeyl, well I guess I was wrong, the glue logic was for the mainstone specific board so I will have to do it for my board, which should be still easy since I can base it off of the mainstone. Thanks for your help. Jun 05 15:08:59 Gerrath: minisd == sd == mmc - only size differ Jun 05 15:09:40 Gerrath: from kernel point of view there is mmc or sd card - does not care about size or extra features Jun 05 15:16:11 CosmicPenguin: FYI, rmk is saying nobody has reported block size issues with mmc/sd to him btw Jun 05 15:17:26 RP: Well, you would need a large card (2G or 4G) and you would need to try to use them on several different machines Jun 05 15:17:54 RP: its really easy to reproduce - take a large card, format it on a Windows box or USB reader, and then try to read it on a native SD controller Jun 05 15:18:11 CosmicPenguin: I know. I guess we just need to convince rmk of this :-/ Jun 05 15:19:02 RP: When I first sent out the patch, his main gripe was that he didn't think cards with a larger block size would be able to write 512 byte packets if the card didn't support partial block write Jun 05 15:19:30 But subsequently, the new SD 1.1 spec that was released showed that all cards must support 512 byte transactions regardless of their native block size Jun 05 15:20:09 CosmicPenguin: Right, and Pierre wants to implement that but rmk seems to think we need some more confirmation of this... Jun 05 15:21:32 I'll try to find it in the PDF Jun 05 15:22:28 CosmicPenguin: http://lkml.org/lkml/2006/5/29/88 Jun 05 15:23:02 I'll add a reply adding my voice to it but the more the merrier :) Jun 05 15:24:09 Does anyone have a 2GB they can run past his patch? Jun 05 15:24:49 I do, but I don't have a setup sitting around at the moment Jun 05 15:24:59 all the alchemy boards I have are busy elsewhere Jun 05 15:25:37 No 2GB cards here :-/ Jun 05 15:26:36 ok time for initrd number 35 :( Jun 05 15:27:53 RP: for further proof - section 3.4.3 (page 18) of the simplified spec states: Jun 05 15:28:08 "rmk is wrong" Jun 05 15:28:09 A card supporting block Jun 05 15:28:09 write shall be required that Block Length set by CMD16 shall be 512 bytes regardless of Jun 05 15:28:09 WRITE_BL_LEN is set to 1k or 2k bytes. Jun 05 15:28:38 ok. I have to go - cu all Jun 05 15:30:52 CosmicPenguin: That looks quite specific :) Jun 05 16:01:32 cbrake, how did my LogicPD PXA720 board support files look? Jun 05 16:18:40 Hey all Jun 05 16:19:31 hey ken|Jlime Jun 05 16:19:57 hey man :D Jun 05 16:20:55 bye chaps Jun 05 16:22:27 Anyone built a successful gpe-image lately? Jun 05 16:27:27 03koen 07org.oe.dev * re754f038... 10/packages/libeventdb/libeventdb_0.30.bb: libeventdb: lower default pref Jun 05 16:30:57 Im seriosly thinking of changing distro, currently using gentoo and switching between 'emerge sync' to update package list and also doing monotone pull... takes alot of time :) Jun 05 16:33:17 that'd be wise. I never could understand the hang of gentoo. strikes me as backwards to wait hours to get seconds of speed improvement. *shrug* Jun 05 16:33:33 heh, right Jun 05 16:35:27 Currently Im loosing all possible speed gains by going through this endless update frenzy Jun 05 16:36:25 I mostly agree though having a few machines running distcc (even windows with cygwin) and not having to deal with incompatible library issues does make up for some of the headaches. Jun 05 16:39:03 Gerrath: looked good. I'll send you something to review this afternoon. Jun 05 16:40:36 03mickeyl 07org.oe.dev * rbd32d9c5... 10/packages/linux/ (6 files in 2 dirs): linux-ezx: add LED drivers for A780 and E680 Jun 05 16:42:41 mickeyl: I use gentoo because it is fairly current, has lots of packages, and is a continous upgrade path. Jun 05 16:43:01 mickeyl: but, I'm coming from fedora -- have not seriously used debian a lot yet. Jun 05 16:43:04 mickeyl: what do you use? Jun 05 16:43:46 mandriva here, especially because mandriva cooker is updated regularly as well Jun 05 16:53:49 pb__: I think the correct fix for that glib problem is $host_alias-libtool as that seems to be used to reference libtool throughout the script, and for native compiles we do pass a --host arg Jun 05 16:56:01 way too nice out today :-) Jun 05 16:57:32 Pffft, going to be 112F/44.4C today Jun 05 16:58:20 eek, where do you live? Sahara? Jun 05 16:58:51 Near Phoneix, Arizona, USA Jun 05 16:59:09 Ah, didnt realize it was so hot there Jun 05 16:59:16 joshin: 112 is nice, 121 is hot. ;-) Jun 05 16:59:32 :) Jun 05 16:59:35 france_, It is still just early June. :) Jun 05 17:00:09 joshin: but not unusual for phoenix in june Jun 05 17:01:33 france_, It is a little hotter than normal but within the standard deviation. Last year was 119 on field day (amateur radio event at the end of June) Jun 05 17:02:29 that sounds more like the phoenix I remember..and 180+ in the car Jun 05 17:02:58 Yup. Have oven mitts in the car to hold the steering wheel. :) Jun 05 17:03:15 Jun 05 17:09:42 ~weather eheh Jun 05 17:09:45 Eindhoven, Netherlands; (EHEH) 51-27N 005-25E 28M; last updated: 2006.06.05 1625 UTC; Dew Point: 42 F (6 C); Pressure (altimeter): 30.24 in. Hg (1024 hPa); Relative Humidity: 54%; Sky conditions: mostly cloudy; Temperature: 59 F (15 C); Visibility: greater than 7 mile(s); Wind: from the NNW (340 degrees) at 15 MPH (13 KT) Jun 05 17:10:13 15c. this is june ... Jun 05 17:10:45 * joshin plans a summer trip to Eindhoven. Jun 05 17:16:51 joshin, you'll bore to death in Eindhoven ... Jun 05 17:17:13 my boss used to work in Phoenix Jun 05 17:19:10 eFfeM, boring is underrated. Jun 05 17:21:15 yeah, actually if you live around phoenix you might like it around here, it is quite crowded (hey it is in the netherlands) but there is a fair amount of forests Jun 05 17:21:47 i wouldn't mind getting an additional 10C from where you live.... Jun 05 17:23:59 I'll happily trade you if you throw in 6-12cm of rain over the course of a year Jun 05 17:25:40 I would not mind the forests Jun 05 17:25:56 or a few fewer degrees C, or a few more cm of rain Jun 05 17:26:09 * XorA has rain to spare Jun 05 17:28:23 ah together we can make a better world ... Jun 05 17:28:34 we have enough rain here Jun 05 17:29:10 raduga: where do you live? Jun 05 17:29:49 southern california Jun 05 17:30:03 it's cool enough most of the time, but gets a little too warm in the summer Jun 05 17:30:07 but never, ever enough rain Jun 05 17:30:47 I worked for years in a supermarket that had a song about S. California and rain as muzak Jun 05 17:31:51 koen, it can always be worse, I know someone who worked in summer in the Efteling, I think he still gets nightmares from Messieur Carnivale Jun 05 17:31:59 actually, on average, we get some (maybe 10-15 days) of rain a year, Jun 05 17:32:08 but in the last decade, its been more like 3-4 days Jun 05 17:32:12 actually had to wait 20mins and already was bored Jun 05 17:32:53 s. cal is indeed pretty dry especially if you do not live near the coast. Jun 05 17:33:08 i live 2km from the coast Jun 05 17:33:14 its still dry :( Jun 05 17:33:18 It's been a while since I've been there Jun 05 17:33:29 * raduga does a rain dance Jun 05 17:33:47 * eFfeM was thinking about palm springs :-) Jun 05 17:33:55 palm springs is desert Jun 05 17:33:59 i know Jun 05 17:34:02 only old people and fools live there Jun 05 17:34:08 lol yes Jun 05 17:34:20 sounds like florida Jun 05 17:34:24 actually that was the reason i didn't go there Jun 05 17:34:35 koen, believe me palm springs is much worse Jun 05 17:34:36 effem: the old people, the fools, or the desert? Jun 05 17:34:44 all of them ... Jun 05 17:34:58 eFfeM: drove thru palm springs ~10 years ago :) Jun 05 17:34:59 well, florida weather might be worse, if you don't like heat + humidity. Palm Springs is dry Jun 05 17:35:15 but palm springs is a barren wasteland, desolate and lifeless, in every regard Jun 05 17:35:33 florida has gators, and some wildlife Jun 05 17:35:43 neither one is a center of civilization Jun 05 17:35:55 did a tour la -> sfo > yosemite -> death valley -> vegas -> zion -> bryce -> page -> grand canyon -> phoenix a few years back. Decided that the drive from phx to lax was not worth my time Jun 05 17:36:22 although i would have loved to see london bridge ... Jun 05 17:36:45 * eFfeM thinks koen is rushing to buy tickets .... Jun 05 17:37:31 effem: i like SF, yosemite, zion and bryce canyons Jun 05 17:37:47 I can't imagine seeing them all in one trip though Jun 05 17:38:15 * raduga was in San Francisco, last week Jun 05 17:38:33 hey, i'm from europe, had to stuff them into a 4 week holiday, it is like americans doing europe in 14 days Jun 05 17:38:48 bryce has become my favourite Jun 05 17:38:53 yes :) Jun 05 17:39:14 did you see any mormons? Jun 05 17:39:53 well, not really in full glory, but loved the names of the waitresses in s. utah. dignity etc Jun 05 17:40:10 i've seen the pennsylvania dutch last year though Jun 05 17:40:26 btw my west coast trip is at http://www.dse.nl/~meulenbr/holiday2000/ Jun 05 17:44:10 back later Jun 05 17:45:22 MACHINE = "collie" | for an ipaq running familiar / gpe what should I use? "familiar" Jun 05 17:46:37 Servo888: look in org.openembedded.dev/conf/machine/ for nearest ft Jun 05 17:46:39 fit Jun 05 17:47:33 There we go h3600.conf. Though would be nice to have an h3800 but they're the same machine anyways. Jun 05 17:49:29 ls Jun 05 17:56:02 RP: ping? Jun 05 18:12:47 I just dont get i t. Why wouldnt gpe-calender build. It allways failing when im atrying to build a gpe-image. It complains about event-db. Has anayone manage to build it the couple of last weeks? Jun 05 18:22:15 mickeyl, got a quick question about libart-lgpl Jun 05 18:22:30 goxboxlive: hrm... which version is this and what's the error? Jun 05 18:22:37 ken: yeah? Jun 05 18:23:13 it grabs the art_config.h from the {HOST_ARCH} var, but im considering whats the best approach for sh3 Jun 05 18:23:26 either setting up sh4,sh3 folders Jun 05 18:23:38 or set all shx to grab from sh Jun 05 18:24:30 since we split sh3 and sh4 in all other areas i vote for the first approach Jun 05 18:24:43 oki, easiest to fix also Jun 05 18:24:48 yeah Jun 05 18:24:49 spectrum-fw is a bit broken Jun 05 18:24:52 it doesn't fetch in do_fetch Jun 05 18:24:56 it fetches in do_compile Jun 05 18:24:57 hey kergie Jun 05 18:25:01 hey Jun 05 18:25:49 heh, right Jun 05 18:25:53 that should be fixed Jun 05 18:26:11 * mickeyl does Jun 05 18:26:43 * kergoth was trying to use -c fetch to populate his sources dir for a bootstrap image, and didnt even get that far :) Jun 05 18:26:48 heh Jun 05 18:33:15 florian: I'll post at pastebin. It is the dev branch, bitbake version 1.4.3. Hold on i'll post the link when i am ready. Jun 05 18:33:33 goxboxlive: ok Jun 05 18:35:39 florian: Here it is http://pastebin.com/760387 Jun 05 18:36:23 03mickeyl 07org.oe.dev * r4214be31... 10/packages/orinoco/spectrum-fw.bb: spectrum-fw: fetch in do_fetch (sic!) Jun 05 18:38:11 florian: Koen said something to me that i had to either use a newer version of gpe-calendar or i had to use a older version of libeventdb. But i have no clue how to use a older version. I have tried to build earlier versions of libeventdb with bibake -b but it has not helped. Jun 05 18:38:34 goxboxlive: use an older libeventdb Jun 05 18:39:01 florian: How do i use a older libevendb? Jun 05 18:42:02 florian:, goxboxlive : http://www.openembedded.org/viewmtn/revision.psp?id=e754f038eb03281981bba6e48f154314cd9f8a79 Jun 05 18:42:03 goxboxlive: hmm yes, that will get you as soon as you don't use -b anymore. you need to make sure oe doesn't try to use the latest release. you can just set DEFAULT_PREFERENCE=-1 in its bb. all older releases are expected to work. Jun 05 18:43:05 koen: perfect :) Jun 05 18:48:57 where can i find osb-jscore? Jun 05 18:51:03 gtk-webcore.sf.net Jun 05 18:59:23 mickeyl, I managed to create my first multiple head Jun 05 18:59:25 03kristoffer 07org.oe.dev * rf0fb7993... 10/packages/gnome/libart-lgpl/ (sh3/art_config.h sh4/art_config.h): Jun 05 18:59:25 gnome/libart-lgpl/sh3/art_config.h: Adding support for sh3 Jun 05 18:59:25 Setting up new sh3 folder with including art_config.h Jun 05 18:59:25 gnome/libart-lgpl/sh4/art_config.h: Adding support for sh4 Jun 05 18:59:25 Setting up new sh4 folder with including art_config. Jun 05 19:00:20 mickeyl, your spectrum-fw.bb update came between my pull and push Jun 05 19:01:41 'monotone merge ; monotone sync' Jun 05 19:02:35 oki Jun 05 19:04:18 florian: I still have the same problem. I have cleaned the libeventdb, and when i am building it it builds version 0.21 ( not 0.30) so i guess that's right: Here is my pastebin look at top of it: http://pastebin.com/760460 Jun 05 19:04:21 koen, btw any good way to locate where a package is located? Looking for osb-jscore but having abit of a trouble finding it Jun 05 19:04:37 03cbrake 07org.oe.dev * r6eb59522... 10/ (3 files in 3 dirs): logicpd-pxa270: initial support for Logicpd PXA270, provided by Shane Volpe Jun 05 19:06:39 ken: find | xargs grep jscore Jun 05 19:07:06 cbrake, thx man Jun 05 19:07:08 ken: its nice the OE meta data is separate from the build -- text searches are very quick :-) Jun 05 19:07:19 yeah noticed :) Jun 05 19:11:08 I've built my gcc without gcj support, and im now experiencing problems with building osb-jscore Jun 05 19:11:51 getting "final link failed: Nonrepresentable section on output" Jun 05 19:15:21 * Malakai is back... generic back msg **** BEGIN LOGGING AT Mon Jun 05 19:47:55 2006 Jun 05 19:53:17 hi Jun 05 19:53:25 hey Jun 05 19:57:06 hey hrw Jun 05 20:08:09 looks like I will work on another kernel addon soon.. Jun 05 20:08:24 module_conf_hostap_cs = "lot of aliases" Jun 05 20:08:41 die-orinoco-die.patch? Jun 05 20:09:01 but done in proper way Jun 05 20:12:24 so you're going to whitelist cards for hostap? Jun 05 20:13:35 yes Jun 05 20:13:39 root@c7x0:~# cat `find /sys/devices/platform/sharp-scoop/pxa2xx-pcmcia/ -name modalias` Jun 05 20:15:54 and run that on each card? Jun 05 20:16:08 yep Jun 05 20:18:30 * Harvy is back Jun 05 20:18:58 * Harvy is back Jun 05 20:44:29 Hola CP Jun 05 20:47:14 hey Jun 05 20:54:12 cu Jun 05 21:14:43 RP: did you respond to RMK? Jun 05 21:56:56 * Philippe is away: visual contact - melancholic dreams Jun 05 23:05:20 Does anyone know Ben Dooks IRC nick? Jun 06 00:58:48 I'm having problems compiling wget-1.9.1. I get this error: Jun 06 00:58:49 CROSS COMPILE Badness: /opt/include in INCLUDEPATH: /opt/include Jun 06 00:58:49 cc1: internal compiler error: Aborted Jun 06 00:58:49 Please submit a full bug report, Jun 06 01:02:46 NAiL: In .dev? When did you last pull from monotone? I fixed that a few days ago... Jun 06 01:06:17 hmm Jun 06 01:06:34 might be me out of sync then ;-) Jun 06 01:06:46 NAiL: The change was just adding this to the wget .bb: EXTRA_OECONF += "--without-ssl" Jun 06 01:08:04 I've got that line Jun 06 01:08:15 wget_1.9.1.bb:EXTRA_OECONF += "--without-ssl" Jun 06 01:08:53 NAiL: Hmm, might be searching for something else. Stick a copy of your do_configure log into pastebin. Jun 06 01:10:30 http://pastebin.com/761250 Jun 06 01:13:07 NAiL: It didn't pass the "--without-ssl" to configure... wtf... Jun 06 01:14:08 no mention of ssl in the run.do_configure file Jun 06 01:14:27 oh, sorry Jun 06 01:14:34 my bad Jun 06 01:14:51 (I think) Jun 06 01:15:22 there's a EXTRA_OECONF_slugos in there. Guess that overrides the following += Jun 06 01:16:24 NAiL: I would not have thought so... but I'll test... Jun 06 01:17:25 NAiL: Your right... that's the problem... now what's the best way to deal with that? Jun 06 01:17:37 good q ;-) Jun 06 01:18:17 I guess the *best* way would be to actually fix wget. But that's kinda beyond me... Jun 06 01:18:44 just setting the _slugos to "--without-ssl" would probably work though Jun 06 01:19:02 NAiL: Yeah, I'm thinking this: http://oe.pastebin.com/761261 Jun 06 01:19:18 NAiL: Disbaling ssl is the correct fix for wget ;) Jun 06 01:23:12 NAiL: I'll push that fix now. Jun 06 01:24:08 v8jlene: thanks :) Jun 06 01:27:19 03lenehan 07org.oe.dev * rd85adbaa... 10/packages/wget/wget_1.9.1.bb: Jun 06 01:27:19 wget 1.9.1: The fix to stop wget searching the system paths didn't work for Jun 06 01:27:19 unslung/slugsos since an append to EXTRA_OECONF appears to happen before a Jun 06 01:27:19 condidition setting even though the append happens later. We need to pass Jun 06 01:27:20 the same option to disable ssl for unslung/slugsos directly instead of Jun 06 01:27:21 relying on the append. Jun 06 02:07:43 "It is recommended to run bitbake without installing it, as a sibling directory of org.openembedded.dev/ and build/ directories. Indeed, as bitbake is written in python it does not need compilation for being used. You'll just have to set the PATH variable so that the BitBake tools are accessible (see #envsetup Setup the Environement section). " Jun 06 02:07:50 can someone translate that? Jun 06 02:08:32 theres a link on your wiki GettingStarted pointing to http://www.openembedded.org/wiki/%23envsetup_Setup_the_Environement which doesn't exists. So I'm not sure what path variables I'm supposed to us Jun 06 02:09:08 i checked out bitbake from svn though Jun 06 02:15:24 Malakai: I presume it means add the bin directory from bitbake to your path, ie for me it'd be /data/oe/src/bitbake/bitbake-1.4.2/bin Jun 06 02:15:35 thanks Jun 06 02:15:45 Malakai: I used the 1.4.2 release (as subject of this channel says to use) and built an rpm from it. Jun 06 02:16:17 i got the same one. Jun 06 02:16:35 i'm getting monotone now. i used apt-get but got a .18 version. doing a manual get for .25 Jun 06 02:20:51 Yep, fedora core and debian testing have .26 which is also no good at the moment... needs to be .25. Jun 06 02:28:56 speaking of .26 any idea of when that will be used Jun 06 02:31:43 mamay: Not really. There is some incompatiblity with .25 so everyone has to upgrade at the same time... and if you have local changes in monotone you need to extract those and then re-apply aftre upgrading. All a bit of mess really. Jun 06 02:31:59 * v8jlene goes to get lunch **** ENDING LOGGING AT Tue Jun 06 02:59:57 2006