**** BEGIN LOGGING AT Fri Jun 16 02:59:56 2006 Jun 16 06:41:12 bonjour Jun 16 06:56:10 morning Jun 16 06:56:24 http://www.joemonster.org/i/o/atlas.jpg - for our German devels Jun 16 07:19:36 :)) Jun 16 07:47:17 morning all Jun 16 07:48:08 hi RP Jun 16 07:48:29 RP: can you test 3541 images? (I built them already) Jun 16 07:49:00 hrw: Where are they? Jun 16 07:50:10 http://ewi546.ewi.utwente.nl/tmp/hrw/3.5.4.1-test/.final/ Jun 16 07:50:14 I found the problem with the git kernels - the mmc oops patch from rmk is causing all the problems with the CF card umounting... Jun 16 07:50:35 I have to build opie/e-core for c7x0/spitz now Jun 16 07:51:09 I have to check which one we use - there was few version of it Jun 16 07:53:41 RP: http://www.kernel.org/git/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=032ebf2620ef99a4fedaa0f77dc2272095ac5863 Jun 16 07:54:07 RP: this should be used instead Jun 16 07:54:20 hi XorA Jun 16 07:54:33 XorA: want to test 3541 final? Jun 16 07:54:43 03pH5 07org.oe.dev * ra56d7408... 10/packages/minicom/ (minicom-2.1/gcc4-scope.patch minicom_2.1.bb): minicom-2.1: build fix for gcc, closes #1087 Jun 16 07:54:54 hrw: Right. That isn't the patch we're using... Jun 16 07:55:10 hrw: I might be able to, its a bad day at work, half the company just closed down, I have to start planning to decommision two foriegn sites Jun 16 07:55:33 XorA: http://ewi546.ewi.utwente.nl/tmp/hrw/3.5.4.1-test/.final/ contain images Jun 16 07:57:31 RP: will you change it or want me to do it? Jun 16 07:58:10 hrw: I'm about to push Jun 16 07:58:13 ok Jun 16 07:58:39 hrw: did you rejig abiword? Jun 16 07:59:15 XorA: iirc not yet - can you remind me changeset? Jun 16 08:02:54 XorA: thats glib-2.0 change which force recompilation of everything? Jun 16 08:03:27 CIA: linux-oz-2.6: Swap brokenmmc oops patch for a corrected version and remove from the git kernel Jun 16 08:04:40 thx RP Jun 16 08:04:54 I forgot to bump PR :-/ Jun 16 08:05:04 np - I have one change to do ;) Jun 16 08:05:17 00-hostap is merged Jun 16 08:05:22 and add one more Jun 16 08:13:31 hrw: I didnt keep the emails, you need the latest rev of whichever glib-2.0 branch uses that dev currently has, then rebuild glib-2.0 and abiword Jun 16 08:13:56 hrw: or short version is LDFLAGS += "-Wl,--export-dynamic" in abiword.bb Jun 16 08:14:12 hrw: sorry got called away for meeting there Jun 16 08:17:45 ok Jun 16 08:21:35 RP: 10-pcnet got added (merged) Jun 16 08:22:33 XorA: this change in glib-2.0 does not force need to recompile every glib-2.0 using packages? Jun 16 08:24:33 hrw: only ones that use gmodule and go wrong, abiword is the only one we have spotted yet Jun 16 08:24:44 ok Jun 16 08:24:49 NOTE: package linux-openzaurus-2.6.16-r38: task do_patch: completed Jun 16 08:27:27 linux-openzaurus_2.6.16.bb | 21 ++++++++++++++------- Jun 16 08:27:27 1 file changed, 14 insertions(+), 7 deletions(-) Jun 16 08:27:53 RP: looks scary but it is not so scary inside Jun 16 08:31:08 RP: can I add myself to maintainer list of linux-oz? Jun 16 08:31:53 hrw: there are two kernsl for 3541?? Jun 16 08:32:03 XorA: why two? Jun 16 08:32:14 hrw: in your directory there are two Jun 16 08:32:21 ah.. I had r35 and r36... Jun 16 08:32:32 will rebuild r38 kernel and images Jun 16 08:34:29 hrw: Why not :) Jun 16 08:39:52 linux-openzaurus: added pcnet_cs patch, added me to maintainers Jun 16 08:39:52 - 10-pcnet.patch adds TrendNet Ethernet card into pcnet_cs driver (merged upstream) Jun 16 08:39:52 - added myself into list of Maintainers Jun 16 08:39:52 - changed SRC_URI a bit: - restructured order of few hack patches - added 'status' field for few patches Jun 16 08:40:10 RP: I hope you will like new SRC_URI look Jun 16 08:40:42 03hrw 07org.oe.oz354x * r26924f1f... 10/packages/linux/ (3 files in 2 dirs): Jun 16 08:40:42 linux-openzaurus: added pcnet_cs patch, added me to maintainers Jun 16 08:40:42 - 10-pcnet.patch adds TrendNet Ethernet card into pcnet_cs driver (merged upstream) Jun 16 08:40:42 - added myself into list of Maintainers Jun 16 08:40:42 - changed SRC_URI a bit: Jun 16 08:40:43 - restructured order of few hack patches Jun 16 08:40:45 - added 'status' field for few patches Jun 16 08:40:47 03rpurdie 07org.oe.dev * r975b9382... 10/packages/linux/ (linux-openzaurus_2.6.16+git.bb linux-openzaurus_2.6.16.bb): linux-oz-2.6: Swap broken mmc oops patch for a corrected version and remove from the git kernel Jun 16 08:40:52 03hrw 07org.oe.dev * rd7293f33... 10/packages/linux/ (3 files in 2 dirs): Jun 16 08:40:54 linux-openzaurus: added pcnet_cs patch, added me to maintainers Jun 16 08:40:56 - 10-pcnet.patch adds TrendNet Ethernet card into pcnet_cs driver (merged upstream) Jun 16 08:40:58 - added myself into list of Maintainers Jun 16 08:41:00 - changed SRC_URI a bit: Jun 16 08:41:02 - restructured order of few hack patches Jun 16 08:41:04 - added 'status' field for few patches Jun 16 08:41:25 o.. cia is back Jun 16 08:44:09 hi alan Jun 16 08:44:49 hey alan|home Jun 16 08:45:10 hrw: I'm not sure I like the blank lines but otherwise its fine. There are probably a few other patches that should be marked as hacks... Jun 16 08:45:30 hey XorA Jun 16 08:45:33 hi all ! Jun 16 08:46:30 hi alan|home Jun 16 08:47:49 RP: its easier to spot sections with blank lines Jun 16 08:48:12 hrw: I know why you did it, I just don't like it :) Jun 16 08:48:17 ;) Jun 16 08:52:34 hrw: As an offical maintainer, you need to start thinking about keeping changes in sync with the +git.bb btw ;-) Jun 16 08:53:05 RP: ;) Jun 16 08:53:10 RP: will remember Jun 16 09:15:53 hi zecke Jun 16 09:16:04 zecke: how was exam Jun 16 09:16:06 ? Jun 16 09:19:50 hmm Jun 16 09:19:52 2.0 Jun 16 09:20:00 should i be happy? :} Jun 16 09:20:42 zecke: I dont know which system is in Germany. Polish has 2.0 (worst, failed) - 5.0 (best) Jun 16 09:21:38 schools (not studies) also has 1.0 (totally bad) and 6.0 (really best) Jun 16 09:22:38 schools Jun 16 09:24:33 zecke: ;( Jun 16 09:24:59 oh wait Jun 16 09:25:08 1.0 really best Jun 16 09:25:13 6.0 totally bad Jun 16 09:26:00 ah... then gratulations zecke Jun 16 09:29:31 mickey|brkfast: smacznego Jun 16 09:34:21 http://piwo.informatyka.org/media/PIWO_06_foto/target38.html - /me Jun 16 10:02:16 zecke: I seems to me that I found what part of linphone needs -I/usr/include Jun 16 10:02:30 It seems Jun 16 10:10:01 03xora 07org.oe.dev * rd8cfd65d... 10/packages/epdfview/epdfview_0.1.5.bb: epdfview_0.1.5.bb : new version Jun 16 10:13:43 zecke: can you help me to fix this error? Jun 16 10:18:42 justalad: hi Jun 16 10:19:50 zecke: Good day :-) Jun 16 10:20:28 justalad: well, which part is failing? Jun 16 10:21:44 linphone-0.12.2/speex/src/speexenc.c Jun 16 10:22:50 justalad: lol Jun 16 10:22:56 in linphone-0.12.2/speex/src/Makefile I see the lines OGG_INCLUDES = -I/usr/include and OGG_LDFLAGS = -L/usr/lib Jun 16 10:23:10 justalad: okay, and is that Makefile generated? Jun 16 10:23:55 from Makefile.in isn't it? Jun 16 10:24:30 justalad: well, I don't have the source here :) Jun 16 10:24:41 justalad: so you have Makefile.am and Makefile.in? Jun 16 10:25:03 yes I have Jun 16 10:25:26 justalad: Makefile and makefile.in are generated Jun 16 10:25:34 justalad: look at Makefile.am Jun 16 10:25:58 INCLUDES = -I$(top_srcdir)/libspeex $(OGG_INCLUDES) Jun 16 10:27:22 justalad: well Jun 16 10:27:25 linphone ship own speex? Jun 16 10:27:32 justalad: then as I said yesterday Jun 16 10:27:35 yes Jun 16 10:27:37 justalad: take a look into config.log Jun 16 10:29:52 * XorA bets it needs a --with-ogg or similar argument Jun 16 10:30:55 ~curse zautrix Jun 16 10:30:59 May you be reincarnated as a Windows XP administrator, zautrix ! Jun 16 10:32:00 kdepimpi cannot be build for qt/e-mt Jun 16 10:34:22 hrw: how's it failing? Jun 16 10:34:49 hrw: where's that photo taken? Jun 16 10:35:06 mickeyl: first it need qmutex.h but its qt3 header. changed to qthread.h fails because qt2 lack QMutexLocker Jun 16 10:35:12 XorA linphone compiles with ogg by default Jun 16 10:35:32 mickeyl: Poznan, P.I.W.O. conference Jun 16 10:35:40 mickeyl: 2006.05.20 Jun 16 10:36:30 3541 final images recreated - who can test? Jun 16 10:36:40 http://ewi546.ewi.utwente.nl/tmp/hrw/3.5.4.1-test/.final/ Jun 16 10:37:23 justalad: yeah, but --with-ogg=${STAGING_DIR} or similar arguments are quite common to stop configure assumptions about /usr/include Jun 16 10:38:45 hrw: I can test a couple of c7x0 images Jun 16 10:38:54 or should I go for cxx00? Jun 16 10:38:58 mickeyl: I prefer cxx00 testing Jun 16 10:39:07 ok Jun 16 10:39:13 mickeyl: c7x0 is machine which I have so I can test Jun 16 10:40:23 and looks like spitz donor forgot about it Jun 16 10:40:30 oh Jun 16 10:40:51 when you don't have one until we meet in berlin you will get mine Jun 16 10:41:26 thx Jun 16 10:41:56 I do not like to remind in situation like this - 3 weeks after last email contact Jun 16 10:42:17 *nod* Jun 16 10:42:31 wait one week more and then a slight ping :) Jun 16 10:42:37 ;D Jun 16 10:43:09 mickeyl: after release with mail: 'how does 3.5.4.1 works for you?' Jun 16 10:43:14 ;P Jun 16 10:43:44 hehe, righto Jun 16 10:44:06 mickeyl: last time I reminded after rc4 Jun 16 10:44:06 zecke: what I must see in config.log? Jun 16 10:44:33 justalad: as I said yesterday. config.log will tell you which test 'dragged' /usr/include 'in' Jun 16 10:44:51 hrw: I can also test cxx00 Jun 16 10:45:08 great Jun 16 10:45:13 well, c3000 anyway Jun 16 10:45:44 RP: borzoi can be flashed with akita stuf - right? Jun 16 10:46:11 NOTE: package gconf-native-2.6.1-r0: task do_distribute_sources: started Jun 16 10:46:11 NOTE: Licenses (['unknown']) are not all listed in SRC_DISTRIBUTE_LICENSES, skipping source distribution Jun 16 10:46:15 not nice.. Jun 16 10:46:20 hrw: No, spitz Jun 16 10:46:49 RP: I mean - can borzoi be flashed with akita rootfs/kernel to boot from flash? Jun 16 10:47:20 hrw: No Jun 16 10:47:21 not with the akita kernel Jun 16 10:47:52 CoreDump|afk is the only one i know with an akita Jun 16 10:47:54 oh and bernado Jun 16 10:47:56 ~seen bernado Jun 16 10:48:02 ok Jun 16 10:48:10 mickeyl: i haven't seen 'bernado' Jun 16 10:48:11 wanted to be sure Jun 16 10:48:16 ~seen bernardo Jun 16 10:48:18 bernardo was last seen on IRC in channel #oe, 7d 4h 54m 49s ago, saying: 'good morning'. Jun 16 10:48:36 'Life's not fair. But the root password helps.' Jun 16 10:48:45 The akita kernel can be made compatible with borzoi but for size reasons I split akita from spitz/borzoi Jun 16 10:48:57 ko Jun 16 10:50:45 hi woglinde Jun 16 10:50:52 hi hrw Jun 16 10:56:09 linphone dependencies are scary... Jun 16 11:04:44 JustinP: e-image-core booting... Jun 16 11:07:01 justalad: bottom of mediastreamer/Makefile.am two direct includes of /usr/include and /usr/X11R6 Jun 16 11:07:11 justalad: those are totally wrong Jun 16 11:11:45 justalad: configure in speex directory must be passed --with-ogg=${STAGING_DIR} to find our version of ogg, otherwise it assumes /usr/include fix these two issues and linphone should compile Jun 16 11:12:12 hi mickeyl Jun 16 11:12:29 XorA: I would even try to build it with OE copy of speex Jun 16 11:12:46 hey pb_ Jun 16 11:12:50 hrw: that would be the best plan Jun 16 11:12:53 hi pb Jun 16 11:14:31 justalad: try adding speex to DEPENDS line and issuing --with-speex=${STAGING_DIR} Jun 16 11:16:40 hrw: The spitz gpe image seems to work. I've booted it, logo is fine, autorotate works, sound works Jun 16 11:16:55 RP: great Jun 16 11:17:08 e-image-core boot on c7x0 Jun 16 11:18:54 hrw: Is this a new version? Jun 16 11:19:02 lardman: 3541 final Jun 16 11:19:17 hello everybody. Jun 16 11:19:19 hrw: What happens if you reboot? Can you login? Jun 16 11:20:03 lardman: yes Jun 16 11:20:10 hrw: good good :) Jun 16 11:22:02 Ideally we should have a gap between the softkeys and the side of the screen I guess :) Jun 16 11:23:25 hrw: wifi (hostap+spectrum) work Jun 16 11:23:38 cool Jun 16 11:23:56 Although it could be using orinoco for my hostap card... Jun 16 11:24:25 RP: let me guess - card 0x0156, 0x0002 or d601, 0005? Jun 16 11:25:38 hrw: 026f,030b Jun 16 11:27:23 hrw: It depends on the order I load the modules. I'm sure hostap has a line for this specific card so I'd say the pcmcia subsystem is buggy somewhere :-/ Jun 16 11:27:41 RP: send me modalias of that card with its cardctl ident Jun 16 11:28:19 hrw: If orinoco is loaded and hostap itsn't should it use orinoco or wait for hostap to load? Jun 16 11:28:33 RP: depends Jun 16 11:28:59 RP: if pcmcia-cs is used then it should use module which does cardmgr tell to use Jun 16 11:29:19 oo 3451 final ? maybe I'll get dma back :) Jun 16 11:29:30 RP: but we can still have few cards wrong configured (wrong == orinoco for prism2) Jun 16 11:29:34 RP: you might even want to consider turning the soft keys off in landscape mode Jun 16 11:29:49 since they're really not that much of a use in that mode Jun 16 11:30:36 mickeyl: I think it will work much better if we alter the map to introduce a small gap. If they're working properly, they can probably stay emabled Jun 16 11:30:50 RP: k, worth a try Jun 16 11:31:12 hrw: I submitted the ID for this card long ago when pcmciautils first appeared so I know its in hostap Jun 16 11:32:04 * mickeyl flashing opie-3.5.4.1.final Jun 16 11:32:33 RP: do you know how i can calculate the proper pixclock value from a given lccr3 ? Jun 16 11:32:37 RP: ok Jun 16 11:32:52 RP: i have this braindead MOTO code that submits totally inconsistent values in the pxafb structure Jun 16 11:33:07 mickeyl: hehe Jun 16 11:33:11 mickeyl: You need to reverse the calculation performed by pxafb Jun 16 11:33:26 RP: i was afraid of you saying that... how :D Jun 16 11:33:59 mickeyl: reading the intel manual :) Jun 16 11:34:12 ok, i have that printed here Jun 16 11:34:31 heh, neat logo Jun 16 11:35:05 standard backlight still way too high IMO Jun 16 11:35:07 hrw: http://www.rpsys.net/openzaurus/temp/cardid.txt if you want to check up on it Jun 16 11:35:17 RP: thx Jun 16 11:35:34 XorA: --with-speex=${STAGING_DIR} not helps Jun 16 11:35:41 mickeyl: The values in mainline are from before we corrected the brightness curve :-/ Jun 16 11:35:54 RP: i see. no biggi Jun 16 11:36:22 mickeyl: We need to do one cleanup of things like that - I don't really want to do it piecemeal Jun 16 11:36:31 RP: you got orinoco with this card? Jun 16 11:36:53 hrw: orinoco was loaded, hostap wasn't, orinoco took the card Jun 16 11:37:12 hrw: If I unloaded orinoco, hostap took it Jun 16 11:37:23 RP: ok. Jun 16 11:37:24 mmh Jun 16 11:37:26 opie is nice Jun 16 11:37:32 * mickeyl didn't use it for long Jun 16 11:37:48 a pity we don't have just a couple of killer apps working fine Jun 16 11:37:52 *sigh* Jun 16 11:37:57 RP: add into /etc/modutils/hostap_cs.conf: "modalias hostap_cs" and update-modules + eject/insert of card Jun 16 11:38:15 mickeyl: or working package manager Jun 16 11:38:53 ya Jun 16 11:40:09 hrw: That fixed it Jun 16 11:40:56 what is LCLK ? Jun 16 11:41:08 mickeyl: LCD Clock? Jun 16 11:41:10 RP: thx Jun 16 11:41:11 or rather... how do i find the value for my board Jun 16 11:41:28 RP: so I have another set of module aliases to add.. Jun 16 11:41:37 LCLK is not the pixel clock or is it? Jun 16 11:42:11 mickeyl: Its the LCD core frequency and varies with cpu frequency (or can do) Jun 16 11:42:40 RP: i see. is there a standard on a pxa270? Jun 16 11:43:23 i think it runs @ 312Mhz Jun 16 11:43:33 mickeyl: Have a look at mach-pxa/pxa27x.c - get_lcdclk...() - Your device probably prints it as it boot Jun 16 11:44:44 Hey guys Jun 16 11:44:54 hmm. no Jun 16 11:44:59 guess i need to enable -DDEBUG Jun 16 11:45:21 Decided to try and get started with OE and running into a problem compiling binutils-cross Jun 16 11:45:46 Is this a known issue (build host is FC5) Jun 16 11:46:36 slaine_: which issue? Jun 16 11:47:00 mickeyl: I never run anything but debug kernels :) Jun 16 11:47:08 excuse the paste Jun 16 11:47:10 /home/glen/src/oe/build/tmp/work/i586-linux/binutils-cross-2.15.94.0.1-r2/binutils-2.15.94.0.1/gas/config/tc-i386.h:441: error: array type has incomplete element type Jun 16 11:47:18 slaine_: use pastebin Jun 16 11:47:21 ~pastebin Jun 16 11:47:24 * RP -> lunch Jun 16 11:47:28 pastebin is probably a place to paste your stuff without flooding the channel - try http://pastebin.com/, or http://pastebin.ca, or http://channels.debian.net/paste, or for #oe use http://oe.pastebin.com, or http://bzflag.pastebin.ca/ Jun 16 11:47:32 RP: :) how do i enable debug at compile time? Jun 16 11:47:41 ~bon appetit Jun 16 11:47:43 somebody said bon appetit was smacznego. Guten Appetit. Eet Smakelijk. God Appetitt. Buon Appetito. Buen apetito Bom Apetite. buen apetito Jun 16 11:47:47 hrw, if it was any longer I would have :) Jun 16 11:47:49 merci. Jun 16 11:50:04 I'm investigating if OpenEmbedded would make a good platform going forward for our product (http://www.lincor.com/marketssolutions1.php) Jun 16 11:51:11 slaine_: nice project Jun 16 11:51:19 thanks Jun 16 11:51:57 slaine_: which version of gcc do you have on your system? Jun 16 11:52:18 slaine_: and could it be that the selected version of binutils is too old for binutils? Jun 16 11:52:19 saline nice Jun 16 11:52:29 We're currently based off of FC2 binaries, but with a completely customizied set of initscripts and our own software stack (though UI is based on the mozilla-1.7.3 browser) Jun 16 11:52:31 slaine_: I have not seen this error yet, did you try to google? Jun 16 11:52:49 zecke, I did, but only post I found was in French. Jun 16 11:53:05 I've got the epia.conf for the machine and generic for the hardware Jun 16 11:53:06 I speak french, if I can help.. Jun 16 11:53:33 gcc -v gives "gcc version 4.1.1 20060525 (Red Hat 4.1.1-1)" Jun 16 11:53:34 magnet hm looks like the host utils are broken Jun 16 11:53:40 slaine_: If I'm not mistaken it is using 'gcc' (host gcc) to compile binutils) Jun 16 11:54:19 slaine_: do you have a gcc-4.0 or gcc-3.4 or even gcc-3.3 on your system? Jun 16 11:54:47 slaine the device is x86 based? Jun 16 11:55:18 woglinde, Yup, Via C3 1GHz, pasively cooled Jun 16 11:55:38 zecke, gcc 4.1.1 as per Fedora Core 5 Jun 16 11:55:48 woglinde: yep Jun 16 11:56:16 Guys, I've to pop out for a few mins, I'll bbs Jun 16 11:59:32 ~lart gcc on x86-64 running as x86 Jun 16 11:59:32 * ibot blasts gcc on x86-64 running as x86 with a huge firehose then strangles gcc on x86-64 running as x86 with it Jun 16 11:59:54 poor x86_64 Jun 16 12:00:50 mickeyl: I have an akita too Jun 16 12:00:52 in other words: fscking ewi which cannot build gcc 3.4.4 for arm Jun 16 12:06:11 umm I'm still missing dma on 3.5.4.1 final :( Jun 16 12:06:23 odo_: good. then you can test Jun 16 12:08:09 mickeyl: what do i need out of that directory? updater.sh, zImage, opie-image is clear, but what are the others for (gpe i know too). Jun 16 12:08:35 odo_: ignore other Jun 16 12:08:49 odo_: afaik, you just need a zImage, opie-image, updater.sh and gnu-tar Jun 16 12:09:35 NAbyss: for akita he does not need gnu-tar Jun 16 12:12:25 JustinP: e-image provided too Jun 16 12:13:36 hi Cliff Jun 16 12:13:47 hrw: morning Jun 16 12:17:01 mickeyl: CONFIG_DEBUG_KERNEL Jun 16 12:18:23 morning :) Jun 16 12:18:41 RP: oh, of course :/ thanks :)) Jun 16 12:18:42 hrw: ah, okies Jun 16 12:18:56 Gerrath: morning Jun 16 12:20:08 RP: isn't pr_debug and friends common now or why are there stil so many handcrafted macros doing the same ? Jun 16 12:20:23 Right, I'm back Jun 16 12:20:24 mickeyl: Its slowly taking over :) Jun 16 12:20:45 So, any suggestions on getting my binutils-cross compile error ? Jun 16 12:20:56 Is FC5's gcc too bleeding edge ? Jun 16 12:21:44 Or should I be creating some sort of bootstrap image first and then looking at doing the gpe-image Jun 16 12:22:23 mickeyl: Its being printed at KERN_INFO level. I'd have thought that would appear on the console by default. You could always hack it to KERN_ERR if you can't find it :) Jun 16 12:22:43 mickeyl: or use sysrq-8 Jun 16 12:23:03 I assume I should just be able to build gpe-image "off the bat" Jun 16 12:23:16 hrw: Its a boot message so he'd have to be fast :) Jun 16 12:23:22 ah. Jun 16 12:23:34 RP: dmesg buffer too small? Jun 16 12:23:35 slaine_: gpe-image should work. bootstrap image is simpler though so perhaps easier to start with Jun 16 12:23:36 Would it be caught my dmesg ? Jun 16 12:23:46 RP, I'll try that so, thanks Jun 16 12:24:00 hrw: Perhaps. I don't know what mickeyl has looked at :) Jun 16 12:24:22 hmm, got just as far and barfed Jun 16 12:24:46 How come it doesn't compile with it's own gcc ? Jun 16 12:25:16 could that be related to the machine/distro I've chosen in my config ? Jun 16 12:25:50 slaine_: You could set some ASSUME_PROVIDED lines to make it use your existing compiler Jun 16 12:26:16 slaine_: I suspect since its failing in binutils, the choice of gpe-image vs bootstrap-image won't help :-/ Jun 16 12:26:23 indeed Jun 16 12:27:13 RP, is it trying to build binutils-cross to create a bootstrapped compiler toolchain ? Jun 16 12:27:23 slaine_: For example of the ASSUME_PROVIDED lines you could use see conf/machine/native.conf Jun 16 12:27:37 slaine_: Yes, it will try and cross compile Jun 16 12:27:59 Ah right, so it is attempting to build it's own toolchain and is failing there Jun 16 12:28:01 I get it Jun 16 12:28:33 lame Jun 16 12:28:45 The failure, not the platform :) Jun 16 12:30:51 http://www.sunix.com.tw/it/en/Product_Detail.php?cate=2&class_a_id=34&sid=447 Jun 16 12:30:51 err Jun 16 12:30:51 sorry Jun 16 12:40:23 Hmm, binutils-2.15.94.0.1 is pretty old Jun 16 12:41:22 * chouimat is back. Jun 16 12:41:37 morning Jun 16 12:41:43 saline yes Jun 16 12:41:47 slaine even Jun 16 12:45:53 slugos and angstrom use 2.16 versions Jun 16 12:46:39 probaly for x86 the newer binutils are better Jun 16 12:47:16 try PREFERRED_VERSION_binutils ?= "2.16.1" in your localconf Jun 16 12:49:19 Yeah, just looking over the angstrom settings and they use pretty much upto date everything Jun 16 12:49:44 I can only assume that the generic one is bit-rotting ? Jun 16 12:50:53 no Jun 16 12:50:57 the generic one is a fail safe Jun 16 12:51:05 which isn't using bleeding edge Jun 16 12:51:25 jepp works excellent for arm devices Jun 16 12:51:40 if it doesn't work for other archs, lets patch/fix it Jun 16 12:52:50 well it ain't bootstrapping on FC5, I'm sure as a failsafe it just works once it's built :) Jun 16 12:55:17 woglinde, Do I need to take a copy of the distro conf file I want to tweak and put it in my build/conf/ path ? Jun 16 12:55:34 slaine yes Jun 16 12:55:38 Simply adding the above to my build/conf/local.conf made no diff Jun 16 12:56:06 slaine yes add it to the local.conf Jun 16 12:56:20 before or after the DISTRO= part ? Jun 16 12:56:31 doesnt matter Jun 16 12:56:55 maybee you should remove the ? Jun 16 12:57:08 * slaine_ hates being a n00b Jun 16 12:57:33 slaine hm we all were noob sometimes Jun 16 12:57:57 slaine_: try a newer binutils PREFFERED_VERSION... Jun 16 12:58:20 zecke I already posted Jun 16 12:58:21 slaine_: and building OE in more configurations is on our todolist as well (mine) Jun 16 12:58:23 ah okay Jun 16 12:59:04 zecke, working on it, just not getting it recognised, it's still picking up the old one and failing after I added PREFERRED_VERSION_binutils ?= "2.16.1" to the local.conf Jun 16 12:59:33 cu Jun 16 12:59:44 I may take a look at the angstrom-2006-9 distro config and base a new one on that Jun 16 13:00:02 basically, try and x86'ify it :) Jun 16 13:00:25 slaine_: cp oe/conf/distro/generic.conf build/conf/distro/generic.conf and edit it Jun 16 13:00:39 hrw|gone, cheers Jun 16 13:07:23 Right, I've made a custom distro.conf, wish me luck :) Jun 16 13:36:34 geebus, libc from CVS. That's a bit much, isn't there a tar.bz2 of it. Would be far quicker to download Jun 16 13:37:12 Of course, it could just be my conf file specifying the wrong things. Jun 16 13:40:16 slaine_: do you have a SRC_TARBALL_STASH set? Jun 16 13:40:35 this cvs co might be available as snapshot to download directly Jun 16 13:41:09 zecke, Nope, glad to see it's possible to specify it though Jun 16 13:41:29 I've only started looking into OE today, so I'm sure I'll be asking lots of stupid questions Jun 16 13:42:22 these are very welcome Jun 16 13:43:02 slaine_: and you happen to hit a less often used (and less often tested) configuration on host and target :} Jun 16 13:43:25 hehe Jun 16 13:43:37 don't tell anyone here or they'll rope me into testing Jun 16 13:45:07 slaine_: oh well, testing is one of my interests... I was just too busy with other stuff Jun 16 13:45:21 zecke, I guess as a matter of history, most people are focused on the Zarius platform ? Jun 16 13:45:23 I hope we can work out the rough edges you will notive Jun 16 13:45:37 I'm sure we can Jun 16 13:45:47 slaine_: not that narrow, more like ARM is far more tested Jun 16 13:46:35 My main goal is to be able to produce an x86/embedded distro with the latest Xorg and GTK+/Cairo stuff, that I can run our software stack on. And have it so that I can easily make a build host, as we've more than one dev team. Jun 16 13:47:04 From what I've seen, that should be achievable Jun 16 13:47:29 Our current UI is served from Apache/PHP to our own Gecko based browser Jun 16 13:47:55 The browser is written in GTK, as is out media app, which sits on top of VLC Jun 16 13:48:39 we do not have cross-compiled apache ;( Jun 16 13:48:42 re Jun 16 13:49:07 slaine_: the UI bits should not be a problem Jun 16 13:49:38 slaine_: the other stuff should be 'minor' as well, it just takes some time and asking questions (we slowly work on the documentation) Jun 16 13:53:00 hrw|gone, haven't you gone somewhere ;) Jun 16 13:53:24 Ah, just spotted the nick change, you must be back now :) Jun 16 13:53:39 If there's no Apache, I'm sure we can work on that with you. Jun 16 13:53:57 Or add it ourselves with our own software Jun 16 13:54:52 I'm not sure if we'll replace the current OS on the product I posted before. I'm investigating some scaled down versions that have no integrated LCD and are effectively Via based STB's. Jun 16 13:55:26 If we do an STB version, the HTML based UI won't be any good for a remote control Jun 16 13:55:27 slaine_: OE contains cherokee -- depending what you need this may work. Jun 16 13:56:08 so we'd probably just do a replacement app in SDL or something. Jun 16 13:56:53 cbrake, Does cherokee do PHP ? That's the only reason we have Apache. The UI needs to be somewhat dyamic and PHP is ideal for that Jun 16 13:57:09 no php is security hole Jun 16 13:58:19 slaine_: cherokee can do php via fastcgi Jun 16 14:00:02 Just flashed 3.5.4.1 final: 1. Settings:appearance (the icon with the star) brings a small window with an information-bubble and a small button, that's all. Jun 16 14:01:26 2. the icon right to the side of it shows some tabs - that's all you can't do anything with it Jun 16 14:02:00 why are ther no more descriptions (names) below the icons? Jun 16 14:02:30 works just fine here. you don't have an old Settings directory on your /home partition or do you? Jun 16 14:02:49 just look .... Jun 16 14:04:02 which icon is the console? They are so small now. Jun 16 14:05:55 ok, now i am sure you didn't follow the installation instructions and removed $HOME/Settings Jun 16 14:06:31 installation instructions? Jun 16 14:06:37 installation instructions. Jun 16 14:07:24 not in that directory: http://ewi546.ewi.utwente.nl/tmp/hrw/3.5.4.1-test/.final/akita/ Jun 16 14:09:11 http://wiki.openzaurus.org/Akita/Install Jun 16 14:09:18 linked from openzaurus.org Jun 16 14:09:30 they not very good though, there is just Jun 16 14:09:36 "Note that the flashing process does not erase your /home directory, therefore settings from other systems may remain there and possibly cause (sapace) problems Jun 16 14:09:37 " Jun 16 14:09:42 which is not clear Jun 16 14:09:52 when i wrote the instructions for earlier releases i said Jun 16 14:10:04 rm -rf $HOME/Settings to get the new qpe.conf installed on first boot Jun 16 14:10:13 (before flashing, of course) Jun 16 14:10:26 It can be altered of course Jun 16 14:10:34 can it? no, it's a wiki Jun 16 14:10:38 heh Jun 16 14:10:39 ;) Jun 16 14:11:13 you won't be able to even estimate how often old /home/Settings was a reason for dissing earlier releases Jun 16 14:11:29 s/dissing/moaning about/ Jun 16 14:11:29 Yes, I remember Jun 16 14:12:08 What are the things which need to be deleted? /root/Settings/ or is it /home/root/Settings/? Jun 16 14:13:27 That's not the only problem with the description - things like "Download files: zImage.bin, initrd.bin, updater.sh. If the filenames are different, change them to the above or else the zaurus will not be able to find them. The filenames are somewhat case sensitive. Either use the capital I in zImage.bin, or all lower, or all caps." Jun 16 14:13:37 confuse those who don't know what has to be renamed Jun 16 14:14:14 lardman: /root gets linked to /home/root since a couple of OZ releases. that's the root of the problems Jun 16 14:14:17 (sic!) Jun 16 14:15:16 We should write a better updater.sh... Jun 16 14:15:24 This is true Jun 16 14:15:34 "Your Settings directory is present. Was this intentional?" Jun 16 14:16:09 RP: See your fanclub on oesf thinking about paying you to implement overlays? ;) Jun 16 14:16:27 lardman: I haven't! Jun 16 14:16:42 RP: I'll find the url, hang on Jun 16 14:17:19 http://www.oesf.org/forums/index.php?showtopic=19976 and Jun 16 14:17:45 it came from here iirc: http://www.oesf.org/forums/index.php?showtopic=19923 (interesting reading this one) Jun 16 14:21:44 Any gpe people here? What needs to be deleted before a gpe install? ~/.gpe/ ? Jun 16 14:22:00 I've got a free Sunday, so I'll try to give the instructions some love Jun 16 14:22:43 lardman: You probably want to ask that in #gpe. Hopefully they managed not to break upgrades :) Jun 16 14:22:55 RP: true Jun 16 14:23:57 lardman: I ipkg upgrade a lot and its never broken for me Jun 16 14:24:07 lardman: I wasn't ware overlay demand is quite as strong as it is. Its time thats the issue, not money :-/ Jun 16 14:24:16 RP: I realise that Jun 16 14:24:27 RP: people want their pr0n divx Jun 16 14:24:38 heh Jun 16 14:24:52 RP: I'm sure it'd make people happy in all sorts of ways Jun 16 14:25:05 RP: but mainly the "I've got it too" way Jun 16 14:25:15 i'm glad people finally start to honor that RP's changes may take long but are done in the right way. Jun 16 14:25:23 What isn't realised is how much time even keeping the kernel stable takes. 6 hours yeserday to work out why the git kernel was suffering PCMCIA trouble :-/ Jun 16 14:25:36 rp :( Jun 16 14:25:49 rmk's testing seems to be virtually non-existant these days Jun 16 14:25:54 The payback is as soon as 2.6.17 comes out, we'll have a perfectly stable OZ kernel for it now Jun 16 14:25:56 RP: blog it, or post it somewhere and then we can point people at it Jun 16 14:25:59 that was different in earlier times Jun 16 14:26:05 RP: excellent. Jun 16 14:26:41 lardman: I keep getting told to blog more :) Jun 16 14:26:52 RP: yeah, that'd rock. Jun 16 14:27:01 * mickeyl just started a new blog as well. Jun 16 14:27:06 dumped my handcoded one Jun 16 14:27:14 to get something that people can drag onto their planet sites Jun 16 14:27:25 planet.oe? Jun 16 14:27:34 mickeyl: How's your thesis coming along? Jun 16 14:27:34 planet.linuxtogo.org Jun 16 14:27:50 mickeyl: Your comment about rmk is ironic as it was a patch from rmk that was the problem :) Jun 16 14:28:05 It was later reverted from mainline but I didn't notice :-/ Jun 16 14:28:05 ~lart MATLAB 7 Jun 16 14:28:11 lardman: a really strong lack of motivation paired with WM is making it very difficult. I still hope i can submit it in the next 10 weeks Jun 16 14:28:19 RP: *nod* Jun 16 14:28:43 mickeyl: Cool. I'm just trying to catch up with everything after doing my corrections. Feeling very tired Jun 16 14:28:59 i can understand that :D Jun 16 14:29:20 mickeyl: On the plus side we aren't seeing anywhere near as much breakage as we used to believe it or not. Jun 16 14:29:52 I've also done some things like testing and debugging the new irq changes so when its merged, we should be unaffected Jun 16 14:30:32 I guess users care more about features than the latest kernels though :) Jun 16 14:31:11 RP: that's right. (all three sentences) Jun 16 14:31:31 multimedia is one of the last things we really suck a bit Jun 16 14:31:39 compared to cacko and pdaX Jun 16 14:31:50 but imo it's low priority Jun 16 14:31:54 compare to a rock solid base system Jun 16 14:31:55 we just need a nice frontend to mplayer and we're sorted Jun 16 14:32:12 mickeyl: But from the users' point of view, multimedia is a big thing Jun 16 14:32:13 lardman: we need TCPMP Jun 16 14:32:32 * XorA waits patiently for 1.0 to be released Jun 16 14:32:53 XorA: It runs on Linux too? Jun 16 14:33:12 lardman: www.tcpmp.com lists GPE and QTe as targets Jun 16 14:33:18 I guess this is to be expected. We've worked on creating a really solid foundation and we have that. Now we can building some nice things on it :) Jun 16 14:33:42 XorA: coming in July... Jun 16 14:34:05 cool Jun 16 14:34:16 lardman: and on my 190Mhz omap PocketPC plays better than mplayer on 400Mhz Z Jun 16 14:34:40 XorA: They have a lot of hw accel stuff built into the player iirc Jun 16 14:35:10 omap... I wonder if it would work on the Nokia 770... Jun 16 14:35:29 lardman: I dont think my phone has much hardware accell Jun 16 14:38:24 morning all Jun 16 14:39:07 hey dirk Jun 16 14:39:14 how was your vacation? Jun 16 14:39:38 hey mickeyl. very good Jun 16 14:40:24 Good evening Jun 16 14:40:42 hey Dirk Jun 16 14:40:51 hey Liam Jun 16 14:40:54 hey sirfred Jun 16 14:41:24 lrg: hi Jun 16 14:41:31 hi sirfred. got your mail. will process on weekend :) Jun 16 14:41:39 mickeyl: great. Jun 16 14:41:47 mickeyl: Well, I would like to comment with you the problem. Jun 16 14:41:57 mickeyl: what's about your motorola phone? can you make phone calls with 2.6 kernel? Jun 16 14:42:06 mickeyl: It seems that sometimes the qcop message is got and sometimes isn't. Jun 16 14:42:11 mickeyl: Any idea about the reason? Jun 16 14:42:37 A little offtopic, what about a HP hx4700 at 315 euro ? Jun 16 14:42:38 do13: no no, we're not that far yet. AP/BP communication is not working at all yet. The main kernel guy is a bit busy with other things and I'm concentrating on the AP drivers atm. Jun 16 14:42:44 A good price? Jun 16 14:42:46 sirfred: no that's really surprising me Jun 16 14:43:01 sirfred: i think so. perhaps a bit steep Jun 16 14:43:05 for a used one at leat Jun 16 14:43:07 least Jun 16 14:43:15 It's a new one. Jun 16 14:43:27 in that case it sounds good to me Jun 16 14:43:31 but i'm not an ipaq expert Jun 16 14:43:37 try to ask on #familiar Jun 16 14:43:45 I should better keep with my zaurus. Jun 16 14:43:46 or #handhelds.org Jun 16 14:43:53 I have to avoid the temptation. Jun 16 14:43:59 sirfred: buy more zaurus models! :) Jun 16 14:44:09 we could use pxafb accelleration ;)) Jun 16 14:44:21 mickeyl: Hmm, what's that? Jun 16 14:44:32 I deleted /home/root/Settings and flashed again and the icons are bigger now ... but there are still no names and the labels within teh applications are without text Jun 16 14:44:41 sirfred: pxafb is the framebuffer interface used on most new zaurus models Jun 16 14:44:49 sirfred: no own graphics card, just a generic interface Jun 16 14:44:54 mickeyl: fast? Jun 16 14:45:39 sirfred: i think it's more or less on par with the w100, but it has a documented(!) overlay interface with hardware YUV and RGB conversion + alphablending layer Jun 16 14:45:49 documented.. oooh Jun 16 14:46:24 I have been too much time in the dark. Jun 16 14:46:40 Perhaps I couldn't resist the light of documentation. Jun 16 14:46:45 :D Jun 16 14:47:00 mickeyl: I've been looking for a way to improve the w100 driver. Jun 16 14:47:19 mickeyl: First thing I've tried is to disable acceleration when the number of cliprects is high. Jun 16 14:47:42 mickeyl: Unfortunately, switching from hardware to software rendering needs to wait for the graphic chips to be idle. Jun 16 14:47:49 So , it's radically slower. Jun 16 14:48:14 The other idea is trying to accellerate bitblitting. Jun 16 14:48:32 Actually, most of bitblitting is not accellerated because only 16 bpp bitmaps are supported. Jun 16 14:49:30 I think bitblt is the key to have a decent accelleration. Jun 16 14:50:39 * chouimat is away: boring stuff Jun 16 14:51:52 why isn't it possible to just blow up bitmaps with lower depths and then use btblt? Jun 16 14:53:24 mickeyl: I was thinking about that. Jun 16 14:53:41 mickeyl: Well, when the bitmap is created, qte asks for space for it. Jun 16 14:54:12 mickeyl: But AFAIR it's stored in its original depth. Jun 16 14:54:34 mickeyl: The idea would be that, just downgrade it to 16bpp and be able to hardware bitblt it. Jun 16 14:55:21 mickeyl: Just doing that transformation every time the bitmap is going to be drawn seems not a good path to go. Jun 16 14:55:29 Because we will need to transfer it to the graphics memory. Jun 16 14:55:37 Every time Jun 16 14:57:32 i see Jun 16 14:58:14 I get an "jff2 error: because ino #799 and #1752 appear to be hard links" but it still boots Jun 16 14:58:28 sirfred: I got my hx4700 for 350€ about a year ago. Jun 16 14:58:51 pH5: So, perhaps it's not a very good price. Jun 16 15:02:34 * chouimat is back. Jun 16 15:03:19 sirfred: It was a very good price at that time, and as there is no replacement model, the prices keep high. Jun 16 15:03:42 pH5: Thanks. Jun 16 15:04:06 sirfred: I see used ones on ebay for >300€, so maybe it's not so bad after all. Jun 16 15:04:38 pH5: What's your experience with the machine? Jun 16 15:04:43 pH5: Satisfactory? Jun 16 15:05:13 hi Jun 16 15:05:53 http://www.cbc.ca/story/canada/national/2006/06/15/street-racing.html Jun 16 15:06:28 sirfred: both frustrating and satisfactory :) Jun 16 15:08:50 sirfred: now that wlan and power management mostly work it's much better than before, but I'm still missing EABI and w3220 acceleration very much. Jun 16 15:09:26 pH5: I understand Jun 16 15:12:08 ~lart orinoco for forcing 0x0156, 0x0002 Jun 16 15:12:27 gah, build still failing Jun 16 15:12:31 This time on glibc Jun 16 15:12:53 It's trying to get a 2.3.5 CVS version even though I've specified 2.4 Jun 16 15:17:23 looks like 'ipkg remove -recursive *orinoco*' hint will stay in 3.5.4.1 releasenotes Jun 16 15:21:55 hi Marcos Jun 16 15:22:07 Hi Hrw! Jun 16 15:22:24 I've stopped using bitbake/oe, things are finally flowing Jun 16 15:22:48 I can build gpe apps to test on desktop now. Jun 16 15:23:17 Did you get the mail I forwarded? Jun 16 15:23:26 thejapa: google one? yes Jun 16 15:23:54 I'm not sure how they found me, the person didn't mention SoC. Complete luck, I guess. Jun 16 15:24:50 RP: nice post Jun 16 15:25:20 hrw: I've tested ubuntu 64 on my desktop. It rocks, really. Jun 16 15:26:03 thejapa: good to know Jun 16 15:26:09 * hrw is a bit busy now Jun 16 15:28:18 lardman: Hopefully it comes accross right... Jun 16 15:28:29 RP: hello Jun 16 15:28:35 hi sirfred Jun 16 15:29:02 RP: did you talked Mardy lately? Jun 16 15:29:20 sirfred: Depends how lately you mean? Jun 16 15:29:25 :) Jun 16 15:29:28 One week? Jun 16 15:29:41 sirfred: Not quite that recently. Why? Jun 16 15:29:52 I would like to comment with him about bitblitting. Jun 16 15:30:29 RP: I've seen his kdrive code and is not supporting 32bpp bitblitting. As expected from our knowledge from aticore library Jun 16 15:30:54 sirfred: Yes, w100 is 16 bit only Jun 16 15:31:25 RP: I find extrange anyway, that blitting from 32 bpp to 16bpp is not implemented. Jun 16 15:31:42 RP: In a chip that's suposedly able to make IDCT, for example. Jun 16 15:32:01 It's a very common operation. Jun 16 15:32:22 sirfred: If we could find a way to do it, it would be a real performance boost :) Jun 16 15:32:42 I'm thinking about playing a little with the gmc_src_datatype field. Jun 16 15:33:07 And wanted to know if Mardy was playing this game before. Jun 16 15:33:40 sirfred: I think both of use have at various points but we never had much luck. Now we know more about things, we might stand a batter chance though Jun 16 15:34:19 RP: I will give it a try. Jun 16 15:34:29 I'll keep you informed. Jun 16 15:34:31 sirfred: I hope you find it :) Jun 16 15:34:38 RP: thanks. Jun 16 15:37:25 rp: thanks Jun 16 15:45:51 Guys, where does glibc-initial come from ? Jun 16 15:46:16 By default it's trying to get some 2.3.5 +cvs2004.... version Jun 16 15:46:56 I've not specified anything for glibc-initial, but I've got 2.4 for glibc and glibc-intermediate Jun 16 15:47:03 any tips ? Jun 16 15:48:03 isn't bitbake saying something about multiple providers Jun 16 15:48:55 zecke, I'll check in a second Jun 16 15:48:57 slaine_ afaik glibc2.4's DEFAULT_PREFERENCE is still lower than 2.3.5's Jun 16 15:49:24 I've kicked it off again, but specidied glibc-initial as 2.4 Jun 16 15:50:09 is there actually a 2.4 version of glibc-inital in the packages directory? Jun 16 15:51:02 pb_: there is, but it's not set as PREFERRED_VERSION in angstrom-something.conf Jun 16 15:51:11 pb_, seems to be, it's downloading now Jun 16 15:51:24 pH5: ah, right Jun 16 15:51:53 probably no reason that 2.4 couldn't just be the global defauit for -initial, actually Jun 16 15:52:05 I wonder why. Forgetfulness? Jun 16 15:52:45 pb_: when is time to remove DEFAULT_PREFERENCE = "-1" from glib_2.4.bb? Jun 16 15:54:31 whoops, I think I hit a wrong button there Jun 16 15:54:38 affs Jun 16 15:55:00 glibc-initial-2.4 won't compile for x86 'cause it's trying to apple arm patchs Jun 16 15:55:19 it should be safe enough to apple those patches on any architecture Jun 16 15:55:25 apple, Freudian slip there :) Jun 16 15:56:25 ~pastebin Jun 16 15:56:41 ~ pastebin Jun 16 15:57:15 hmm, ibot gone ? Jun 16 15:57:44 somebody said pastebin was a place to paste your stuff without flooding the channel - try http://pastebin.com/, or http://pastebin.ca, or http://channels.debian.net/paste Jun 16 15:57:56 where is ibot when you need it Jun 16 15:58:28 http://oe.pastebin.com/713221 Jun 16 15:58:49 pb_, "should" was the key word in there sentence. Jun 16 16:00:19 do you have any evidence to suggest that it isn't? Jun 16 16:00:29 iirc, those arm patches don't touch any files outside of sysdeps/arm Jun 16 16:01:23 The pastebin shows that the arm_longlong.patch doesn't apply Jun 16 16:01:35 right, but it doesn't look like that's due to a wrong architecture Jun 16 16:01:41 judging from that output, it wouldn't apply on arm either Jun 16 16:01:50 it looks like the patch is just plain broken Jun 16 16:02:06 :( Jun 16 16:02:32 Assuming stdlib/longlong.h isn't arch specific, then you'd be right Jun 16 16:02:34 I didn't knew we have a glibc-initial yet Jun 16 16:03:42 oh yeah. obviously I was wrong about the patch touching files outside of sysdeps/arm, but the piece of longlong.h that gets patched is inside #ifdef __arm__. Jun 16 16:04:01 zecke: er? we've had glibc-initial since... forever. Jun 16 16:05:36 version 2.4. Jun 16 16:05:45 oh right Jun 16 16:05:50 yeah, I didn't know we had that either Jun 16 16:06:05 *busy* cya later Jun 16 16:13:18 looks like I'm getting WPA working again.. Jun 16 16:15:44 hrw hehe Jun 16 16:18:31 pb_, I think I see the problem Jun 16 16:19:01 the do_patchcmd script that got created is pointing to the patch file from glibc-cvs as opposed to glibc-2.4 Jun 16 16:19:12 ah, that's no good Jun 16 16:19:23 is there a FILESPATH in the .bb file? Jun 16 16:19:51 jez.... I do not like way in which gpe handle font sizes.. Jun 16 16:20:07 pb_, just reading over it today Jun 16 16:20:13 I mean now Jun 16 16:20:44 damn this stream of consciousness that keeps coming out via my fingers ;) Jun 16 16:21:01 heh Jun 16 16:21:20 FILESPATH = "${@base_set_filespath([ '${FILE_DIRNAME}/glibc-cvs', '${FILE_DIRNAME}/glibc', '${FILE_DIRNAME}/files', '${FILE_DIRNAME}' ], d)}" Jun 16 16:21:25 There we go Jun 16 16:21:32 right Jun 16 16:21:34 that looks bad Jun 16 16:21:36 I assume I should replace -cvs with -2.4 Jun 16 16:21:43 yeah, probably Jun 16 16:21:50 but how do I push something like that back up stream (assuming it works) Jun 16 16:22:06 file a patch in bugzilla, ideally Jun 16 16:22:21 use monotone log to find out who added the broken file, and assign the bug to them Jun 16 16:27:02 03hrw 07org.oe.oz354x * rf4886475... 10/packages/twisted/twisted_1.2.0.bb: twisted: fix SRC_URI - close #1077 Jun 16 16:27:07 03hrw 07org.oe.oz354x * rc293620b... 10/packages/glib-2.0/ (files/configure-libtool.patch glib-2.0_2.6.4.bb): Jun 16 16:27:07 glib-2.0: add a patch to change configure to call $host_alias-libtool as this is what is generated. This means that gmodule-2.0.pc gets populated with the correct flags. Jun 16 16:27:07 Found this while tracking down why abiword-plugins didnt work. So at least Jun 16 16:27:07 abiword needs to be rebuilt. Maybe other gmodule using apps that were Jun 16 16:27:09 malfunctioning as well. Jun 16 16:27:11 taken from .dev Jun 16 16:27:13 03hrw 07org.oe.oz354x * r87273600... 10/packages/glib-2.0/ (18 files in 3 dirs): glib-2.0: removed not used versions Jun 16 16:27:16 03hrw 07org.oe.oz354x * rd16386fb... 10/packages/abiword/ (abiword-plugins_2.4.4.bb abiword_2.4.4.bb): abiword: bumped PR due to glib-2.0 change Jun 16 16:27:19 03hrw 07org.oe.oz354x * r66ff139c... 10/packages/wpa-supplicant/wpa-supplicant_0.4.8.bb: wpa-supplicant: add wpa_supplicant.conf to CONFFILES, me as maintainer Jun 16 16:27:24 03hrw 07org.oe.dev * r68f7b61f... 10/packages/twisted/twisted_1.2.0.bb: twisted: fix SRC_URI - close #1077 Jun 16 16:36:51 pb_, Right, next question Jun 16 16:37:02 arm_longlong.patch is getting included from somewhere Jun 16 16:37:11 But I haven't figured out where yet Jun 16 16:37:17 somewhere in the SRC_URI, probably Jun 16 16:37:32 It's not mentioned in glbic-initial_2.4.bb file at all Jun 16 16:37:33 use 'bitbake -c showdata -b ...' to see the expanded variables, that sometimes helps Jun 16 16:37:48 yah, probably getting set in some other file that's included from this one Jun 16 16:37:54 okay Jun 16 16:39:07 looking at the .bb file, I'm guessing that include glibc_${PV}.bb will include the glibc_2.4.bb file ? Jun 16 16:39:20 If so, that's probably where its getting it from Jun 16 16:40:05 yeah Jun 16 16:41:39 bye chaps Jun 16 16:58:28 pb_, Getting closer now Jun 16 16:58:53 should .../work/i586-linux/glibc-initial-2.4-r5/ have an extracted glibc-2.4 archive in it (to be patched) Jun 16 16:59:32 I think that's the problem, it's trying to patch against an empty dir Jun 16 17:09:18 mickey|fifa2006: after match can you boot opie image on c7x0? Jun 16 17:09:30 hrw *g* Jun 16 17:20:51 * chouimat is away: lunchtime Jun 16 17:30:08 CoreDump|home: hi Jun 16 17:30:20 hey hrw Jun 16 17:30:35 CoreDump|home: can you test akita 3541 images? Jun 16 17:31:05 CoreDump|home: http://ewi546.ewi.utwente.nl/tmp/hrw/3.5.4.1-test/.final/ Jun 16 17:31:08 sure Jun 16 17:31:15 thx Jun 16 17:34:41 hrw: :-) Jun 16 17:34:50 ouch, ewi is kinda slow today Jun 16 17:36:30 JustinP: e-image/e-image-core provided so expect bugraports Jun 16 17:44:46 hrw: ok Jun 16 17:46:36 Has anyone noticed that qpegps (OE build) doesnt work with gpsd (OE build). I know gpsd work's, becuse i have ran it with my BT GPS, but qpegps refuse to find my GPS device. Jun 16 17:46:47 *dev branch* Jun 16 17:51:19 goxboxlive: qpegps team did not updated to new API of gpsd iirc Jun 16 17:52:05 hrw: I se, but what do i have to do to get it to work? Jun 16 17:53:32 goxboxlive: find working pair and add it into OE or pester qpegps guys to update or write that code Jun 16 17:54:55 ok, i can try to send them a mail. Jun 16 18:04:27 * chouimat is back. Jun 16 18:16:48 RP: ping Jun 16 18:39:29 jo tmbinc Jun 16 19:02:20 hrw|tv: unfortunately, since I can't even fix the other devs' problems..... Jun 16 19:02:31 hrw|tv: it just works for me...I don't know why it wouldn't for other people Jun 16 19:03:49 * lpotter beats his head against the wall Jun 16 19:04:33 anyone familar with the pm.c of the pxa in the kernel? Jun 16 19:07:26 Gerrath: well, I wrote it back in the SA1110 days :-) Jun 16 19:08:50 hi lpotter ... Jun 16 19:08:59 hey Jun 16 19:09:16 lpotter: qtopia4 gpl soon? Jun 16 19:09:27 * lpotter beats his head against the wall Jun 16 19:09:33 =D Jun 16 19:09:35 hope so Jun 16 19:10:12 lpotter: ok ... Jun 16 19:10:14 some people just dont get it, unfortunetly Jun 16 19:10:26 cbrake, oh, ok. Jun 16 19:11:06 lpotter: I just want to remove the qt2 ifdef ... Jun 16 19:12:07 Gerrath: have not looked at in awhile -- looks like currently only the traditional sleep/suspend mode is supported. Jun 16 19:13:01 cbrake, from what I read the pm functions are controlled by a demon (apmd or acpid) correct? Jun 16 19:13:23 daemon I mean Jun 16 19:13:27 lol Jun 16 19:13:41 cant really say much during our "quite period" Jun 16 19:14:10 lpotter: I will wait since I can't do anything else about it Jun 16 19:14:29 Gerrath: on arm only apmd - this platform lack acpi Jun 16 19:14:52 hrw|tv, ok, thanks. Jun 16 19:15:16 even the apm support is emulated. the daemon only exists to be able to call out userspace scripts when power managemetn events occur Jun 16 19:15:23 so i wouldnt say the functions are "controlled by" them Jun 16 19:15:26 more the other way around Jun 16 19:16:19 cbrake, I'm assuming it would be pretty easy to add the standby and and deep idle states, this code looks pretty straight forward. Jun 16 19:17:26 Gerrath: standby seems to fit into the suspend model pretty well. I'm not so sure about deep idle without doing a little research first. Jun 16 19:18:19 cbrake, deep idle interests me the most since there is sig. power savings and only 1us to wake up from it. Jun 16 19:18:33 i wouldnt think idling would be hugely beneficial given that linux is jiffies driven, continually ticking, unless there's enough time between ticks Jun 16 19:18:46 * kergoth is still interested in the idea of a jiffies-less linux kernel Jun 16 19:19:21 Gerrath: yeah 1us is quick. Resume from suspend usually takes couple 100 ms. Jun 16 19:19:57 * lpotter doesnt see what the kernel has to do with bad peanut butter Jun 16 19:20:51 * kergoth throws a blob of jiffy at lpotter Jun 16 19:21:37 on another note, http://david.navi.cx/blog/?p=94 is cool Jun 16 19:24:45 you know, we've never really figured out a good way to do initscripts Jun 16 19:27:05 initng looks promising Jun 16 19:27:34 kergoth concurent starting of services? Jun 16 19:28:07 yep. the idea has been around for some time. the old page about need(5) based inits, the gentoo hack on top of sysvinit, etc Jun 16 19:28:14 but initng looks like it might not be a horrible implementation Jun 16 19:28:18 need to mess with it more Jun 16 19:28:28 heh Jun 16 19:28:33 I agree, but I was talking about OE itself Jun 16 19:28:40 hm minit is already good Jun 16 19:28:46 having initscripts_1.0.bb full of if [ distro ] just isn't working Jun 16 19:29:19 hrm. historically, when/where did /etc/sysconfig originate? Jun 16 19:29:35 pretty sure it wasn't in BSD. I don't remember seeing it in old sysv Jun 16 19:29:54 redhat innovation? or something I'm not remembering right? Jun 16 19:30:08 raduga yes Jun 16 19:30:19 yeah, it was redhat Jun 16 19:30:52 i do remember being annoyed/surprised by it, the first time I saw it Jun 16 19:30:58 but it seems to work, moderately ok :/ Jun 16 19:31:01 CosmicPenguin: ah, right, yeah, thats a mess.. i'd like to have a generic format for init script metadata that oe then uses to emit the correct format for the init being used. Jun 16 19:31:07 cbrake, how does the kernel handle the ram in the different states? is there a state where it switches it to self-refresh mode? Jun 16 19:31:36 if it didnt switch to self refresh mode, suspend wouldnt exactly be very effective :P Jun 16 19:31:46 Gerrath: in suspend(sleep), the RAM switches to self refresh Jun 16 19:31:57 Gerrath: I'm not sure about standby for the PXA Jun 16 19:33:10 kergoth, good point :-) hard to recover when everything is gone, LOL. Jun 16 19:33:17 indeed ;) Jun 16 19:33:28 Gerrath: suspend on the PXA is a lot of fun, because during resume, the BL starts at the reset vector and you have to bring the SDRAM out of self refresh and then jump to where you left off -- with lots of other minor details. Jun 16 19:33:48 * cbrake looks at PXA standby a little more ... Jun 16 19:34:08 cbrake, but that is already handled by the current routine written in pxa/pm.c, correct? Jun 16 19:34:29 Gerrath: what is missing is the BL portion -- there is a lot of work that happens in the BL. Jun 16 19:34:59 Gerrath: and some of it is platform dependent. Jun 16 19:37:27 Gerrath: getting a new system to resume can lead to grey hair (or maybe it was the kids) :-) Jun 16 19:38:07 heh - we probably can't talk about it officially, but be it known here that the first root filesystem booted with LinuxBIOS on the OLPC boards was generated by OE.. :) Jun 16 19:39:29 cosmic hehe Jun 16 19:50:02 CosmicPenguin: interesting Jun 16 19:51:14 cbrake: heh, at least 2.6 has the driver model, sucks a lot less (getting resume to work) Jun 16 19:51:29 i'm still blown away by stock 2.4 running the resume hooks in the same order as teh suspend ones Jun 16 19:51:32 heh Jun 16 19:52:00 kergoth: yeah, the driver model is a big improvement Jun 16 19:52:22 i would like to see something more capable than cpufreq though Jun 16 19:52:37 a proper representation of the clock tree with individual nodes being scalable would be slick Jun 16 19:54:14 * CosmicPenguin hugs bluetooth Jun 16 19:54:58 sort of Jun 16 19:55:52 http://photos.jibble.org/GeekPr0n30 Jun 16 20:00:41 mickey|fifa2006: few things about your new site: 1. link logo to / of service. 2. provide some kind of blog / so it will be easier to read it Jun 16 20:01:28 <[g2]> anyone know the state of the iwmmx support for pxa27x in OE ? Jun 16 20:01:54 * [g2] wonders what state this is in for use in OpenEZX Jun 16 20:01:54 [g2]: include conf/machine/tune-iwmmxt.conf? Jun 16 20:02:52 <[g2]> hrw|tv anyone use that ? I've seen stuff in the repo for mainstone, but I don't know how much air time it's got Jun 16 20:03:47 [g2]: I suspect that angstrom will use it when it will finish starting Jun 16 20:13:41 * cbrake needs to dust off the mainstone build ... Jun 16 20:16:42 Gerrath: RAM used to be a big portion of the suspend current, but w/ 1.8V parts this is probably changed. Jun 16 20:38:35 * chouimat is away: pub!! Jun 16 20:45:32 cu Jun 16 23:57:26 Hi everyone Jun 17 00:03:43 jacques: hey Jun 17 00:03:57 howdy kergoth Jun 17 00:04:12 how ya been? Jun 17 00:06:01 Hi kergoth Jun 17 00:06:08 pretty good. have a job I really like. living near old friends. Jun 17 00:06:20 how about you? didn't you just move in march? Jun 17 00:06:33 What is your job, jacques? Jun 17 00:06:35 nice Jun 17 00:06:39 i move quite a bit Jun 17 00:06:46 went from texas back to mn, then from mn to phoenix Jun 17 00:06:48 tis hot Jun 17 00:07:19 heh, but at least it's a dry heat (unlike here) Jun 17 00:07:30 yea, just feel like you're slowing baking Jun 17 00:07:34 rather than having to swim to the car Jun 17 00:07:37 bioeng, I'm doing embedded linux now Jun 17 00:07:49 You're working for a company? Jun 17 00:08:05 Getting paid? Jun 17 00:08:12 Or is it volunteer? Jun 17 00:09:25 I'm actually working for a real company, getting paid :-) Jun 17 00:09:48 and you actually like it, hold on to that ;) Jun 17 00:11:03 yeah, I like what I'm doing, people I'm working with, etc. Jun 17 00:11:08 it's nice Jun 17 00:11:19 You're a software engineer? Jun 17 00:11:37 bioeng, yes, that's my position title :-) Jun 17 00:11:45 Wow! Jun 17 00:11:54 My friend is an embedded software engineer at Boeing Jun 17 00:11:58 I really look up to her Jun 17 00:12:08 because she is so successful Jun 17 00:12:18 i like my coworkers here, but i cant say i'm happy with the work environment Jun 17 00:12:36 jacques: i'm working with russ from #elinux nowadays Jun 17 00:13:01 kergoth, ah cool - I should visit #elinux more Jun 17 00:13:15 I'm working with a really good friend of 20 years Jun 17 00:14:59 Well, I'm working to graduate with my EE degree Jun 17 00:15:05 jacques: ah nice Jun 17 00:15:10 today has sucked Jun 17 00:15:15 yes Jun 17 00:15:15 everything just stopped working Jun 17 00:15:18 :( Jun 17 00:15:20 ouch Jun 17 00:15:30 it was so good this morning too ... :) Jun 17 00:15:33 I am familiar with that phenomenon. Jun 17 00:15:33 http://www.infradead.org/~jcrouse/olpc-screenshot.jpg Jun 17 00:15:35 Since I admire my friend so much because everyone has treated her so nicely in school (better than me), I'm working towards getting into embedded Jun 17 00:15:47 i should leave early, stuck here for another 1.5 hours Jun 17 00:15:58 CosmicPenguin: nice! Jun 17 00:16:07 kergoth: notice the banner? :) Jun 17 00:16:21 hehe, yep Jun 17 00:16:37 Is an EE degree good for embedded, guys? Jun 17 00:16:48 doesn't hurg Jun 17 00:16:51 s/hurg/hurt/ Jun 17 00:16:56 hehe, source: not found. typical shell script under busybox error Jun 17 00:17:08 yeah - thats my bad Jun 17 00:17:23 EE would help on the writing drivers side of things. i've often wished i knew hardware better Jun 17 00:17:25 at least I killed the script that was trying to rm -rf /tmp Jun 17 00:17:26 Is Boeing a good place to work? Jun 17 00:20:12 Would Boeing be a good place to work? Jun 17 00:20:28 bioeng, like all huge companies there are probably great divisions to work for and not so great Jun 17 00:20:29 * kergoth doesnt know Jun 17 00:21:07 other than that I have no first hand knowledge Jun 17 00:21:10 I need to do embedded so people will treat me as wonderful as they did my friend Jun 17 00:21:22 They treated her like royalty Jun 17 00:21:28 heh Jun 17 00:21:37 The administration made sure she got a scholarship and job from them Jun 17 00:21:46 The engineering college administration that is Jun 17 00:21:49 well at my company we get free donuts on fridays Jun 17 00:21:50 thats scary embedded Jun 17 00:21:55 embedded folks are in demand, i'll say that much. i've never had trouble finding a job, though it often involves moving Jun 17 00:21:58 like, you screw up, people die type of embedded Jun 17 00:22:03 heh, scary indeed Jun 17 00:22:19 I screw up, a MP3 skipps Jun 17 00:22:21 My friend majored in computer engineering in college Jun 17 00:22:25 if I screw up, oil might cost more Jun 17 00:23:08 I'm struggling because my dad does not approve of me being an EE Jun 17 00:23:20 He makes me work as a construction laborer all the time Jun 17 00:23:33 heh Jun 17 00:23:36 you gotta do what makes you happy Jun 17 00:23:44 my dad would be dancing for joy if i was an EE Jun 17 00:23:47 once you are an adult, you are responsible for yourself Jun 17 00:23:55 What makes me happy is to emulate my friend Jun 17 00:24:17 in 1992, my parents questioned my choice to be a CS major, in 1998, they took all the credit Jun 17 00:24:23 hahah Jun 17 00:24:24 I don't mean to sound weird or anything Jun 17 00:24:24 lol Jun 17 00:24:45 It's just that people at school did not treat me very well Jun 17 00:24:48 I started as EE but switched to CS after two years Jun 17 00:24:52 And this is a college Jun 17 00:24:59 They all acted like kids Jun 17 00:25:28 They started by telling the engineering administration untrue things about me Jun 17 00:25:44 That got them wanting to kick me out according to some Jun 17 00:25:59 Then I got invited to the engineering honor society Tau Beta Pi Jun 17 00:26:23 No one gave me a chance, they wouldn't meet me or talk to me, and they rejected me because "we don't know you" Jun 17 00:26:29 School has been hard Jun 17 00:27:23 But I'm working to put my life back together Jun 17 00:27:52 And I have to hand it to the IRC community: you guys have treated me much better than the people at school Jun 17 00:28:09 So what do you think, guys? Jun 17 00:28:14 I hope I didn't offend anyone Jun 17 00:29:24 I'm sorry Jun 17 00:29:25 personally, i think you should be more realistic. once you're out of college, what are you go into isn't going to suddenly get you treated better. embedded development isnt some magic bullet to a better life, its just a career. Jun 17 00:30:27 Well Jun 17 00:30:47 It's looking more like I'll be an entrepreneur after graduation Jun 17 00:30:57 with my civil and electrical engineering degrees in a year and a half Jun 17 00:32:30 The same guy who got my friend her job at Boeing told me I need to be an entrepreneur Jun 17 00:33:52 I'm looking into doing something with embedded in my entreprenurial career Jun 17 00:34:23 possibly in a business in structural system monitoring using electronics Jun 17 00:34:51 I'm also looking into going to UIUC for a master's Jun 17 00:36:06 Did I offend anyone? Jun 17 00:36:09 I'm sorry if I did Jun 17 00:37:27 Can someone respond please? Jun 17 00:37:30 I'm very sorry Jun 17 00:37:36 I didn't mean to offend anyone Jun 17 00:38:36 not offended, just dunno what to say Jun 17 00:38:45 OK Jun 17 00:38:51 Hmmmm Jun 17 00:39:15 What kinds of things is embedded systems used in? Jun 17 00:40:37 * CosmicPenguin goes home Jun 17 00:40:45 I can't cause any more damage around this popsicle stand Jun 17 00:41:32 jacques, what kinds of products are you working on? Jun 17 00:44:20 seismic acquisition systems Jun 17 00:45:07 That kind of relates to civil engineering Jun 17 00:45:18 I'm a civil engineering major as well Jun 17 00:45:38 Of course, I'd like to use embedded systems in civil engineering monitoring systems Jun 17 00:45:56 yes I can see the similarities Jun 17 00:46:00 But I do want to take advantage of every entrepreneurial opportunity I can with both engineering degrees Jun 17 00:46:20 Would you know anything about structural monitoring systems? Jun 17 00:48:00 no, sorry. I don't really know much abou seismic other than what I've picked up in the last few months Jun 17 00:53:22 You had to pick it up in order to program the algorithims that drive the equipment? Jun 17 00:54:24 well we have a whole dept of scientists and mathematicians to design the algos Jun 17 00:54:49 Oh Jun 17 00:55:17 They communicate the requirements to you and you program the algos? Jun 17 00:57:11 yep Jun 17 00:58:10 What programming language do you use? Jun 17 00:59:01 For programming that stuff Jun 17 00:59:36 C Jun 17 00:59:57 some assembler Jun 17 01:00:30 What architechture? Jun 17 01:00:49 And is the company you work for a large company? Jun 17 01:01:48 medium sized, powerpc (at the moment) Jun 17 01:34:41 ... Jun 17 01:35:46 bioeng is getting a degree in elec eng and is asking what sort of devices are embedded? Jun 17 01:36:04 must be a high quality degree **** ENDING LOGGING AT Sat Jun 17 02:59:57 2006