**** BEGIN LOGGING AT Wed Feb 15 10:59:56 2006 Feb 15 11:03:50 03mreimer 07org.oe.oz354fam083 * rc1116e6e... 10/packages/gpe-bootsplash/gpe-bootsplash_1.15.bb: Feb 15 11:03:50 gpe-bootsplash: add version 1.15, and start a bit later in the boot process Feb 15 11:03:50 to get rid of a couple of error messages. Feb 15 11:03:55 03mreimer 07org.oe.oz354fam083 * r4d9375e9... 10/packages/gpe-bootsplash/gpe-bootsplash_1.15.bb: disapproval of revision 'c1116e6ea9e731cefda34ac72410890c992618e3' Feb 15 11:03:59 03mreimer 07org.oe.oz354fam083 * r8f37e157... 10/packages/irda-utils/irda-utils_0.9.16.bb: irda-utils 0.9.16: add support for hx4700. Feb 15 11:04:04 03mreimer 07org.oe.oz354fam083 * r46ba424d... 10/packages/linux/ (handhelds-pxa-2.6/h2200_pm.patch handhelds-pxa-2.6_cvs.bb): Feb 15 11:04:04 handhelds-pxa-2.6_cvs.bb: Add a patch for h2200 to retry loading the Feb 15 11:04:04 bootloader firmware every 30 seconds until successful. Also, I changed the Feb 15 11:04:04 bootloader firmware filename to the more appropriate h2200_bootloader.bin. Feb 15 11:04:08 03mreimer 07org.oe.oz354fam083 * r6a200652... 10/packages/h2200-bootloader/h2200-bootloader.bb: Feb 15 11:04:08 h2200-bootloader.bb: Make it a bit quieter, and use the new firmware filename Feb 15 11:04:09 h2200_bootloader.bin. Feb 15 11:04:13 03mreimer 07org.oe.oz354fam083 * r8fd5d232... 10/packages/linux/handhelds-pxa-2.6/h2200/defconfig: linux/handhelds-pxa-2.6/h2200/defconfig: Build in the h2200-specific drivers. Feb 15 11:04:16 03mreimer 07org.oe.oz354fam083 * rc5b92303... 10/conf/machine/h2200.conf: Feb 15 11:04:18 h2200.conf: Fix EXTRA_IMAGECMD_jffs2 to generate an image that works on Feb 15 11:04:20 h2200. Also remove unnecessary module_autoload directives now that we're Feb 15 11:04:22 building in the h2200-specific modules. Feb 15 11:04:26 03mreimer 07org.oe.oz354fam083 * raaf10993... 10/conf/machine/handheld-common.conf: Feb 15 11:04:28 handheld-common.conf: Add hidp and irtty-sir to HANDHELD_MODULES so these Feb 15 11:04:30 modules get included in the rootfs. Feb 15 11:04:32 03mreimer 07org.oe.oz354fam083 * r20b05a0d... 10/conf/machine/handheld-common.conf: handheld-common.conf: Remove irtty-sir. Feb 15 11:04:35 03hrw 07org.oe.oz354fam083 * r1f6027f0... 10/conf/machine/poodle-2.4.conf: poodle-2.4.conf: fix virtual/kernel name Feb 15 11:04:40 03justinp 07org.oe.oz354fam083 * r31445fb0... 10/packages/meta/task-e-x11.bb: task-e-x11: add DEPENDS Feb 15 11:04:44 03justinp 07org.oe.oz354fam083 * r77a8dee5... 10/packages/meta/e-image.bb: e-image: Add gdk-pixmap-loader packages (how do these get pulled in in a GPE image?) Feb 15 11:04:47 03justinp 07org.oe.dev * rc70cc766... 10/packages/meta/e-image.bb: e-image: Add gdk-pixmap-loader packages (how do these get pulled in in a GPE image?) Feb 15 11:04:50 03philippe 07org.oe.dev * r53cc53e9... 10/packages/gpe-plucker/ (gpe-plucker/gpe-plucker-desktop.patch gpe-plucker_0.2.bb): Feb 15 11:04:55 gpe-plucker: packages/gpe-plucker/gpe-plucker/gpe-plucker-desktop.patch Feb 15 11:04:57 * patch to fix bug #1531 (handhelds.org) Feb 15 11:04:59 gpe-plucker: packages/gpe-plucker/gpe-plucker/gpe-plucker-desktop.patch Feb 15 11:05:01 * add usage of patch Feb 15 11:05:03 03justinp 07org.oe.dev * r37212ffa... 10/packages/gpe-confd/gpe-confd_0.16.bb: gpe-confd: add missing DEPENDS on sqlite Feb 15 11:05:06 03mreimer 07org.oe.dev * r34851fe4... 10/packages/linux/handhelds-pxa-2.6/h2200/defconfig: Feb 15 11:05:10 packages/linux/handhelds-pxa-2.6/h2200/defconfig: Turn off DEBUG_INFO, which Feb 15 11:05:12 makes the modules enormous. Feb 15 11:05:18 03pH5 07org.oe.dev * rb85360b6... 10/packages/xmonobut/ (xmonobut_0.4.1.bb xmonobut_0.4.bb): xmonobut: depend on x11 instead of diet-x11 Feb 15 11:36:02 anyone who has a clue on libtool got an opinion on http://pastebin.com/555761 ? Feb 15 11:39:49 <_law_> hi Feb 15 11:40:34 <_law_> is there a kernelfunction to check if clamshell is open/close ? Feb 15 11:44:41 _law_: One of the input system ioctls should help - look for the status of the SW bits Feb 15 11:45:08 <_law_> RP, where can i find this Feb 15 11:47:04 <_law_> "/sys/class/input/input0/capabilities" Feb 15 11:47:15 _law_: http://svn.o-hand.com/view/misc/trunk/chkhinge26/chkhinge26.c?rev=56&view=markup might give you hints Feb 15 11:47:34 I'm in the process of rewriting it as it happens Feb 15 12:15:02 03rwhitby 07org.oe.dev * ra7dedfaf... 10/packages/linux/ixp4xx-kernel/2.6.16/94-nas100d-setup.patch: ixp4xx-kernel: Changed nas100d rtc from probe to force. Feb 15 12:15:06 03rwhitby 07org.oe.dev * r8c50e11a... 10/packages/linux/ixp4xx-kernel/2.6.16/96-nas100d-leds.patch: ixp4xx-kernel: Changed nas100d led name from disk-1 to disk. Feb 15 12:17:59 good aternoon ... Feb 15 12:18:57 hello rp.... Feb 15 12:19:19 RP: you know if the recent .dev gpe image building? Feb 15 12:24:32 greentux: I haven't tried one for a while. Are you seeing a problem? Feb 15 12:26:41 RP: lala has end december version and has some problems... he said its not so easy to upgrade to recent versions (some major changes were done in th elast weeks...) Feb 15 12:27:17 gpe builds on .dev for me until yesterday, a change in xext cvs busted me today Feb 15 12:27:18 greentux: I tested the build after those changes and it worked Feb 15 12:27:34 RP: good to know, tnx Feb 15 12:28:30 greentux: I'm aware of one small issue with a current build from scratch but I can tell you how to fix it - I need to check a fix into bitbake for that which will happen as soon as I get a free moment Feb 15 12:37:25 03rwhitby 07org.oe.dev * r6dfbf762... 10/packages/linux/ (2 files in 2 dirs): ixp4xx-kernel: added the remove-extraversion patch, to do what the bb file used to do, but for some reason no longer does. Feb 15 12:44:41 RP: what issue? Feb 15 12:45:01 XorA: bitbake isn't building all RDEPENDS at the moment Feb 15 12:46:15 RP: ok, nothing thats hit me then Feb 15 12:47:38 XorA: No, it was highlighted in some external files I was testing Feb 15 12:48:16 only issue I see is gdbm refusing to stage, but I cant figure out WTF is wrong with it Feb 15 12:51:14 does already packages for localisation of the keyboard input exist? Feb 15 12:54:44 <_law_> hrw|work, does opie turn off backlight if i close the clamshell (with kernel 2.6) ? Feb 15 13:01:08 I checked that fix into bitbake Feb 15 13:16:51 lardman: iirc it can do it Feb 15 13:17:27 _law_: I would have to check it - no opie now on Z Feb 15 13:23:23 hey Feb 15 13:31:19 zecke: good afternoon Feb 15 13:32:00 hey Feb 15 13:37:20 i want to build a 2.6 kernel for tosa Feb 15 13:37:53 but when i start bitbake with "bitbake virtual/kernel" i got many warnings and errors Feb 15 13:38:14 what is the best way to test the 2.6 Kernel? Feb 15 13:40:50 Bedman: The 2.6 kernel in oe isn't up to date yet. Feb 15 13:41:03 ok Feb 15 13:41:16 i habe to patch it by hand? Feb 15 13:41:20 have Feb 15 13:42:01 I suggest wait a few days. I have to sort out some problems Feb 15 13:42:08 ok ;-) Feb 15 13:43:51 After activating the battery measurement the touchscreen detects random pen down events. Also ohci needs pm Feb 15 14:03:46 afternoon Feb 15 14:11:11 I'm trying to find the oe source, and can't seem to find it. Feb 15 14:19:16 ~google openembedded repository Feb 15 14:19:26 never mind. I found a snapshot Feb 15 14:19:43 reenoo: I can't get to the repo. Is it down? Feb 15 14:21:41 I get 404 error from http://monotone.vanille.de/viewmtn Feb 15 14:22:09 well, that's just viewmtn Feb 15 14:24:06 or vanille is down Feb 15 14:24:20 no, I'm pulling from there as we speak Feb 15 14:30:41 ~ewi Feb 15 14:30:46 rumour has it, ewi is ewi546.ewi.utwente.nl - main backup server for OE monotone. Hosts database snapshot: http://ewi546.ewi.utwente.nl/OE/OE.db.bz2 Feb 15 14:31:17 reenoo: http://ewi546.ewi.utwente.nl/tmp/viewmtn/ Feb 15 14:33:08 hrw|work: the point being? I'm not looking for a viewmtn Feb 15 14:41:13 morning Feb 15 14:43:48 ok Feb 15 14:43:55 reenoo: my mistake Feb 15 14:44:00 kaenat: http://ewi546.ewi.utwente.nl/tmp/viewmtn/ Feb 15 14:44:34 hrw|work: ok :) Feb 15 15:41:07 hi Feb 15 15:48:56 hmm I'm having a problem with kernel module dependencies. task-bootstrap depends on $module_name, but in deploy/ipk the modules have a "-2.6" suffix (ie: $module_name-2.6) Feb 15 15:49:21 CoreDump|home: the modules ought to Provide: the non-suffixed name Feb 15 15:49:32 hmmmm Feb 15 15:49:38 kernel-module-wacom-2.6_2.6.15-r6_akita.ipk Feb 15 15:49:38 kernel-module-wp512-2.6_2.6.15-r6_akita.ipk Feb 15 15:49:41 just an example Feb 15 15:50:06 indeed Feb 15 15:50:07 so, how did I break that again Feb 15 15:50:13 ahhhh Feb 15 15:50:20 Provide: blahh Feb 15 15:50:27 ;) Feb 15 15:50:27 * CoreDump|home checks Feb 15 15:51:33 Provides: kernel-module-xpad-2.6 Feb 15 15:51:35 mehh Feb 15 15:52:11 oh, that's no good Feb 15 15:52:35 * CoreDump|home checks the kernel bbclass Feb 15 15:53:25 http://pastebin.com/556078 is from akita 2.4 kernel module Feb 15 15:53:44 Package: kernel-module-sco-2.4 + Provides: kernel-module-sco Feb 15 15:53:59 hmm Feb 15 15:54:39 well, I was compiling kernels by running "tmp/run.do_compile", maybe that b0rked something Feb 15 15:54:53 no, that was another tree Feb 15 15:56:22 meh, the interesting parts are python in kernel.bbclass Feb 15 15:57:13 speaking about modules... could someone look at #524 in bugtracker? Feb 15 15:58:27 ouch Feb 15 15:58:44 one more reason I use different trees for different builds =) Feb 15 16:00:35 * CoreDump|home looks at python populate_packages_prepend () and sighs Feb 15 16:04:55 <[lala]> hi Feb 15 16:08:53 * CoreDump|home sets PARALLEL_INSTALL_MODULES = 1 and rebuilds the kernel Feb 15 16:14:54 bye Feb 15 16:31:00 <_law_> RP, chkhinge26 works fine now thanks Feb 15 16:43:55 <[lala]> hmm, is there any working monotone server i can sync from? tried ewi and vanille - both non working atm Feb 15 16:45:49 ewi was working a minite ago Feb 15 16:45:57 and it's pull, not sync Feb 15 16:46:24 * france is away: Away Feb 15 16:46:25 check the website for the other servers Feb 15 16:47:58 <[lala]> last time i checked there were no others Feb 15 16:49:16 there are 3 or 4 Feb 15 16:49:22 have been since th ebeginning Feb 15 16:51:48 cu Feb 15 16:56:04 <[lala]> well, monotone.nslu2-linux.org is marked as not-up-to-date on the webpage Feb 15 16:56:12 <[lala]> is that still the case? Feb 15 16:56:15 * Philippe is back (gone 19:55:19) Feb 15 16:57:04 <[lala]> dominion seems to be working :-) Feb 15 17:09:03 RP: ping Feb 15 17:09:06 CoreDump|afk: ping Feb 15 17:10:26 JustinP: pong Feb 15 17:11:38 RP: to turn on kexec do I just CONFIG_KEXEC=y? Feb 15 17:11:52 JustinP: I'd guess so. I've not tried it yet Feb 15 17:11:55 RP: when I did that my screen started in portrait mode (which I found strange) and the root was NULL Feb 15 17:12:05 RP: so it didn't boot... Feb 15 17:12:25 doesn't sound good :-/ Feb 15 17:12:36 hmmm... Feb 15 17:12:45 I ask because CoreDump|afk said that he was able to kexec.... Feb 15 17:13:15 I'd suspect something else broke Feb 15 17:14:14 hmmm... Feb 15 17:14:29 perhaps I should try it without kexec and make sure the current version is working normally Feb 15 17:35:21 re Feb 15 17:35:52 hi hrw Feb 15 18:06:42 RP: a "normal" 2.6 kernel build is still in portrait and has no root= for me.... Feb 15 18:06:50 RP: perhaps it's because I'm using bitbake -b? Feb 15 18:06:57 * JustinP tries in bitbake -i Feb 15 18:08:38 I know it *shouldn't* make a difference.... Feb 15 18:09:55 RP: was the kernel perhaps changed lately to be in portrait rotation for spitz? Feb 15 18:10:03 don't suppose it would be... Feb 15 18:10:19 * JustinP wishes commit e-mails were working Feb 15 18:12:44 does anybody know if the kexec patch is going to be committed to mainline? Feb 15 18:14:28 * JustinP has no idea Feb 15 18:14:37 RP probably knows Feb 15 18:23:20 * CosmicPenguin wonders if the hwsensors folks have an irc channel Feb 15 18:30:21 is ewi down? Feb 15 18:34:00 morning kergoth Feb 15 18:35:56 * JustinP grumbles Feb 15 18:36:09 kergoth: monotone.vanille.de? Feb 15 18:36:58 hi mreimer Feb 15 18:37:04 hi pb__ Feb 15 18:37:08 * JustinP would very much like commit e-mails... Feb 15 18:39:55 cu Feb 15 18:47:15 peek linux-openzaurus CMDLINE Feb 15 18:47:15 console=ttyS0,115200n8 console=tty1 noinitrd root=/dev/hda1 rootfstype=ext3 rw ${CMDLINE_MEM} fbcon=rotate:1 dyntick=enable debug Feb 15 18:47:21 WTF is going on.... Feb 15 18:49:33 oh wait...I had the wrong one...hmm....why is the shell building on an old zImage date.... Feb 15 18:49:36 :-| Feb 15 18:51:22 hey Feb 15 18:52:31 flocking sheep Feb 15 18:53:00 why isn't my kernel getting a CMDLINE? Feb 15 18:53:11 apparently there's something wrong with my freebsd, i did the same GettingStarted procedure on a mdk box and...it's doing the job Feb 15 18:54:01 i wonder if it has something to do with python Feb 15 18:55:25 it could have something to do with the version.... Feb 15 19:29:01 I am getting REAL tired of trying to make abiword work! Feb 15 19:29:17 poli: really? it built for me the other day Feb 15 19:29:45 mreimer: I just build abiword and gtk+ with symbols and non-stripped... Bug #1520 Feb 15 19:29:46 poli: http://sermons.desiringGod.org/oe/ipk/abiword_2.3.99-r5_armv5te.ipk Feb 15 19:30:04 ah Feb 15 19:30:31 mreimer: Now I can't get gdb to debug it!!! gdb dies just before the error happens. Unbelievable. Feb 15 19:30:39 grr Feb 15 19:31:47 poli: does it work at all, or does it crash on startup? it's working on my hx4700 Feb 15 19:32:36 mreimer: starts, splash screen, opens the main editor screen and crashes. Feb 15 19:32:54 poli: you could try my ipk Feb 15 19:33:16 poli: might help determine if it's a toolchain/dependency issue Feb 15 19:33:33 mreimer: I will... I just reliazed that I don't know how the armv5te is different from the plain arm ipk. Feb 15 19:34:10 poli: pxa255 and pxa27x are armv5te, so it'll work on your hx4700 Feb 15 19:34:34 mreimer: I installed and got OE to work because of that bug... I can't get even the familiar-shipping one to work. Feb 15 19:36:57 mreimer: is there a way to force ipkg to install that package when it thinks it is already up-to-date? Feb 15 19:36:58 ~lart CMDLINE Feb 15 19:36:58 * ibot teaches CMDLINE that M$ Access is a database. No, really, a database. A real live multi-user... well, ok, not multi-user, but a database. Yeah, that sounds right. Feb 15 19:37:05 poli: -f Feb 15 19:37:29 JustinP: wan't -f the conf file option? Feb 15 19:38:08 poli: ipkg help install perhaps? Feb 15 19:38:33 poli: -force-reinstall Feb 15 19:38:54 ~lart bad Indian coders Feb 15 19:38:54 * ibot whips out a shotgun, trudges over to bad Indian coders, and goes postal Feb 15 19:39:02 oh... it was right under my nose. Thanks, mreimer. Feb 15 19:41:10 mreimer: that is weird... I am getting a dependency problem... apperently ipkg doesn't consider my installed libfontconfig1 to be a valid dependency :( Feb 15 19:41:32 -force-depends? Feb 15 19:41:45 poli: yeah, I've had troubles with libfontconfig1 too. Try what JustinP said Feb 15 19:42:49 That is soooo weird... fontconfig wasn't installed at all. Feb 15 19:43:00 Was it a dependency from the start?! Feb 15 19:44:26 Well, at least we know it isn't a dependency problem... Feb 15 19:44:36 Same behaviour as bug #1520 Feb 15 19:44:49 crash on start up? Feb 15 19:45:25 poli: are you running from .dev or .oz354fam083? Feb 15 19:45:28 splash-screen, main editor window and immediate crash Feb 15 19:46:00 mreimer: I built mine from oz354fam083, but it has the same behaviour since I flashed the image... Feb 15 19:46:14 mreimer: I actually got to compile it myself because it didn't work at first... Feb 15 19:46:31 poli: http://handhelds.org/scap/port.11737.png Feb 15 19:47:27 poli: try 'ipkg update; ipkg -test upgrade' to see what might get updated. you might also try adding my feed Feb 15 19:47:39 mreimer: I have everything up-tp-date.. Feb 15 19:47:54 mreimer: I do get an error, take a look at #1520 Feb 15 19:54:56 cya later Feb 15 20:07:32 i have an error that says "Coulnd't find magic file" Feb 15 20:07:41 what is that related too ? Feb 15 20:07:43 ohhh magic! Feb 15 20:07:45 I love magic Feb 15 20:07:45 am I missing a tool ? Feb 15 20:09:43 lol Feb 15 20:11:27 file, probably Feb 15 20:30:43 RP: it turns out that my problem was that I had no newline at the end of my defconfig Feb 15 20:31:04 RP: could I suggest adding a newline before the CMDLINE added in linux-openzaurus.inc? Feb 15 21:16:34 re Feb 15 21:29:02 03justinp 07org.oe.dev * r868a692e... 10/packages/e17/ (2 files in 2 dirs): e17-gpe-menu-convert: don't create eaps which already exist, don't add items to menus when they are already in there, re-added postinst, fixed DEPENDS Feb 15 21:31:56 JustinP: great - can you push it into .ox354fam083? Feb 15 21:31:59 cu Feb 15 21:37:44 hrw|gone: yeah, going to do that right now Feb 15 21:37:51 hrw|gone: went to lunch ;-) Feb 15 21:41:38 hrw|gone: done Feb 15 22:01:11 03justinp 07org.oe.oz354fam083 * r0756fb4a... 10/packages/e17/ (2 files in 2 dirs): e17-gpe-menu-convert: don't create eaps which already exist, don't add items to menus when they are already in there, re-added postinst, fixed DEPENDS Feb 15 22:02:10 * Philippe is away: visual contact - melancholic dreams Feb 15 22:28:18 pwd Feb 15 22:29:51 #oe Feb 15 22:30:38 RP: why does my spitz power off when I "init 6"? Feb 15 22:36:15 does anyone have an example of a package in OE that needs to build a binary which it runs later in the build? Feb 15 22:36:19 if RP responded I didn't see it....darn screen.... Feb 15 22:36:34 JustinP: he didn't Feb 15 22:36:42 johnX: I have an example of a package which needs that... Feb 15 22:36:43 johnX: thanks Feb 15 22:37:34 great...I'm trying to get enca to build without resorting to qemu-arm/binfmt_misc like I do now Feb 15 22:38:02 * JustinP has no idea what that is... Feb 15 22:38:07 I'd like emacs to be in OE Feb 15 22:38:16 enca is a dependency of fbreader Feb 15 22:38:18 but it needs to compile itself with....itself.... Feb 15 22:38:30 I actually compiled it on my zaurus Feb 15 22:38:41 took 12 hours or so Feb 15 22:38:45 maybe more Feb 15 22:38:53 ah Feb 15 22:39:02 and I have ipks which I created, but I don't really like that solution... Feb 15 22:39:31 well I use qemu-arm, which allows ARM binaries to run on x86 transparently Feb 15 22:39:31 hey Feb 15 22:39:37 hi Feb 15 22:40:01 (as long as binfmt_misc is setup right, and you have the libs they need someplace where they can find them) Feb 15 22:41:52 I see... Feb 15 22:42:05 in terms of an example, I was thinking more along the lines of a package that already works, so that I can see how to do this with enca Feb 15 22:44:29 I know Feb 15 22:44:33 I was just mentioning it... Feb 15 22:44:35 ah Feb 15 22:45:06 you might be able to make qemu do what you need... Feb 15 22:45:42 is enca in OE? Feb 15 22:45:56 nope Feb 15 22:45:59 ah Feb 15 22:46:09 well, I'd appreciate the example Feb 15 22:46:11 i made a .bb for it so I could build fbreader Feb 15 22:46:18 of qemu in action? Feb 15 22:46:19 suppose I'd need to install qemnu-arm as well? Feb 15 22:46:25 brb Feb 15 22:46:27 k Feb 15 22:52:15 I have to get to class now, but I'll be back on later...as for qemu, in the way I'm using it, it's entirely transparent Feb 15 22:52:42 ie just execute an arm binary and it runs like it's native Feb 15 22:52:47 later Feb 15 23:00:02 RP: kexec works for me botting from 2.6 into 2.4. However, I have to -f in order for it to work as the 2.6 kernel can't successfully reboot on my spitz... Feb 15 23:08:45 tr Feb 15 23:08:48 re even Feb 15 23:35:12 * france is back (gone 06:48:50) Feb 16 00:00:54 well i hose familiar today on my ipaq Feb 16 00:00:57 hosed* Feb 16 00:01:32 seems suspending with a CF card in the sleeve does some bad juju Feb 16 00:02:28 any ideas on how to recover short of a reinstall? Feb 16 00:03:01 as much fun as it is staring at the "X" cursor is Feb 16 00:07:24 emte: Ctr-Alt-Backspace :-/ Feb 16 00:07:30 emte: Fn-Left? Feb 16 00:07:36 * JustinP has no ipaq Feb 16 00:07:45 the h36 has rhose keys? Feb 16 00:07:50 those* Feb 16 00:07:55 and of course the lack of keyboard kind of....sucks.... Feb 16 00:08:00 heh, I have no idea Feb 16 00:08:02 I assume not Feb 16 00:08:21 without a keyboard I think you're SOL Feb 16 00:08:36 there is a hard reset but it doent help Feb 16 00:08:58 you need to make it go to shell....is there a key you can press to stop X? Feb 16 00:09:05 of course that wouldn't help Feb 16 00:09:13 without a keyboard you can't do anything outside of X Feb 16 00:09:24 you'll have to reinstall if you don't have some kind of keyboard Feb 16 00:09:31 (AFAIK) Feb 16 00:09:53 lol Feb 16 00:10:17 if i thought the shell would help ... Feb 16 00:11:30 :-( Feb 16 00:12:43 using serial cable now to find out if it even gets that far Feb 16 00:18:56 what was the command to re-enable console messages? Feb 16 00:20:04 dunno Feb 16 00:20:22 well i've a log file full of: Feb 16 00:20:24 Feb 13 13:51:31 h3600 daemon.info gpe-dm: gpe-dm: session pid 1670 exited with status 1 Feb 16 00:20:24 Feb 13 13:51:32 h3600 user.debug kernel: dma period = 1398056 ps, clock = 206400 kHz Feb 16 00:20:24 Feb 13 13:51:32 h3600 user.debug kernel: dma period = 1398056 ps, clock = 206400 kHz Feb 16 00:20:24 Feb 13 13:51:33 h3600 user.debug kernel: dma period = 1398056 ps, clock = 206400 kHz Feb 16 00:20:31 thoughts? Feb 16 00:22:02 why is it killing gpe-dm ... Feb 16 00:22:21 is your / mounting ro perhaps? Feb 16 00:22:46 rootfs / rootfs rw 0 0 Feb 16 00:23:04 and Feb 16 00:23:06 /dev/root / jffs2 rw 0 0 Feb 16 00:23:07 no, I mean is the system mounting it ro for some reason Feb 16 00:23:19 thats from mtab Feb 16 00:23:21 I've had problems where the dm won't load if / is ro Feb 16 00:23:24 ah Feb 16 00:23:29 how are you accessing this? Feb 16 00:23:34 serial Feb 16 00:24:07 ah Feb 16 00:30:15 hrm ... Feb 16 00:30:19 what is happening Feb 16 00:34:35 i wonder ... Feb 16 00:36:46 The good news is that trying to get something working in Gentoo is just as annoying as getting it to work in OE Feb 16 00:37:12 sounds about right Feb 16 00:37:30 So if thats the level of sophistication we're going for, mission accomplished Feb 16 00:38:48 hmm Feb 16 00:39:00 well i got X to come up ... but nothing works Feb 16 00:40:36 well .. i just got a kernel fualt and backtrace :( Feb 16 00:44:30 http://oe.pastebin.com/556933 Feb 16 00:44:36 stack overflow? Feb 16 00:47:39 oe.pastebin.com - how did we get that? Feb 16 00:48:15 ahh - cute Feb 16 00:50:26 by using freenode Feb 16 00:50:36 its automatically allocated Feb 16 00:51:16 any thoughts yet how to fix this? Feb 16 00:57:07 emte: don't know what to tell you...I suggest a reflash.... Feb 16 00:57:45 yeah, but a reflash doesnt help fix the problem Feb 16 00:57:57 theis the familiar 0.8.3 RC2 Feb 16 00:58:00 this* Feb 16 01:02:11 this is first boot? Feb 16 01:02:20 no Feb 16 01:02:28 well...you can try to fix it...I have no idea how, though.... Feb 16 01:02:44 all i did was suspend with the CF card in the sleeve Feb 16 01:02:47 good nite Feb 16 01:02:53 ok Feb 16 01:02:55 well.... Feb 16 01:03:04 can you see if any of your data is corrupt? Feb 16 01:03:25 not found any so far Feb 16 01:06:25 can you get a terminal somehow? Feb 16 01:07:35 using the terminal via serial Feb 16 01:07:51 then try starting gpe-dm and see what happens Feb 16 01:07:55 what gets output Feb 16 01:08:45 Feb 13 14:11:59 h3600 daemon.info gpe-dm: gpe-dm: session pid 6590 exited with status 1 Feb 16 01:08:45 Feb 13 14:11:59 h3600 user.debug kernel: dma period = 1398056 ps, clock = 206400 kHz Feb 16 01:08:45 Feb 13 14:11:59 h3600 user.debug kernel: dma period = 1398056 ps, clock = 206400 kHz Feb 16 01:08:45 Feb 13 14:12:00 h3600 user.debug kernel: dma period = 1398056 ps, clock = 206400 kHz Feb 16 01:08:50 loops forever Feb 16 01:09:00 not overly helpfull Feb 16 01:09:12 I mean what does it output in the *shell* Feb 16 01:09:24 I mean a shell, not kernel output Feb 16 01:10:14 nothing Feb 16 01:10:33 i am still trying to figure out how to enable shell messages again Feb 16 01:14:32 root@h3600:~# sh /etc/rc5.d/S99gpe-dm stop Feb 16 01:14:32 Stopping GPE display manager: gpe-dm Feb 16 01:14:32 Segmentation fault Feb 16 01:14:32 Segmentation fault Feb 16 01:14:35 nice Feb 16 01:15:07 do an strace Feb 16 01:15:27 wtf ... Feb 16 01:15:39 3621 root 572 S /bin/sh /etc/X11/Xinit.d/11zaurus Feb 16 01:15:39 3623 root 572 R /bin/sh /etc/X11/Xinit.d/11zaurus Feb 16 01:15:43 bbiab Feb 16 01:15:53 hmmm... Feb 16 01:16:00 now its trying to load zaurus hing stuff ... Feb 16 01:16:04 hinge* Feb 16 01:40:20 emte: you sure nothing got truncated or anything? Feb 16 02:08:25 ok what am i doing wrong? pth is trying to be built and it says this | pth_mctx.c:476:2: #error "Unsupported Linux (g)libc version and/or platform" Feb 16 02:16:20 simcop2387: look at the error...is your libc or platform unsupported? Feb 16 02:16:28 simcop2387: perhaps you're trying uclibc and it doesn't like it? Feb 16 02:17:00 i'm using glibc last i checked Feb 16 02:17:13 i'm going to check what triggered that exactly in a moment Feb 16 02:18:48 JustinP, i've no clue what is happening now .. nor what that zaurus file would be on an ipaq in the first place Feb 16 02:19:00 s/what /why Feb 16 02:19:57 emte: ah, that didn't stand out to me before....good point Feb 16 02:20:50 emte: 11zaurus is a shell script (as are all Xinit.d files) which sets the initial X direction on hinged zaurii Feb 16 02:20:56 yup Feb 16 02:21:00 emte: it shouldn't do anything on your ipaq Feb 16 02:21:10 it shouldnt be running either Feb 16 02:21:16 emte: why not? Feb 16 02:21:27 somehow it sponataneously started Feb 16 02:21:27 emte: oh, is it currently running? Feb 16 02:21:33 yes, twice Feb 16 02:21:34 it's not spontaneous Feb 16 02:21:38 it's part of Xinit Feb 16 02:21:47 likely your Xinit is getting re-run.... Feb 16 02:21:50 it hasnt ran in the past that i know of Feb 16 02:21:56 of course it does Feb 16 02:22:03 Xinit always runs when X is siarted Feb 16 02:22:13 no, i mean the zaurus script Feb 16 02:22:15 it's what starts gpe-dm Feb 16 02:22:31 the zaurus script is part of Xinit, it would also be run normally Feb 16 02:22:38 but it should just exit right out on non-zaurii Feb 16 02:22:47 emte: try which chkinge Feb 16 02:23:19 no reply Feb 16 02:23:37 then 11zaurus shouldn't do anything, it should just exit Feb 16 02:23:45 you're sure that it's the same instance still running? Feb 16 02:23:47 interesting ... Feb 16 02:23:58 it should *run* but not do anything Feb 16 02:24:01 now my ipaq is asleep while i am using it Feb 16 02:24:16 12318 root 752 S /usr/bin/ipaq-sleep Feb 16 02:24:16 12342 root 728 R ps -ef Feb 16 02:24:32 it's not asleep or you wouldn't be able to do that Feb 16 02:24:42 lol Feb 16 02:24:48 are your /dev entries all ok? Feb 16 02:24:53 is /proc mounted? Feb 16 02:24:57 /sys? Feb 16 02:25:02 * emte has no clue Feb 16 02:25:10 the thing is hosed Feb 16 02:25:12 type "mount" Feb 16 02:25:40 /sys doesnt exist Feb 16 02:25:49 but yeah proc and dev are fine Feb 16 02:28:33 hmm where is my CF Feb 16 02:29:28 emte: is your CF card still plugged in? Feb 16 02:29:36 and your sleeve? Feb 16 02:29:49 they were not Feb 16 02:29:52 are you on a 2.4 or 2.6 kernel? Feb 16 02:29:59 what is the output of "mount"? Feb 16 02:30:06 Linux h3600 2.4.19-rmk6-pxa1-hh37 #1 Sat Jan 28 02:47:27 CET 2006 armv4l unknown Feb 16 02:30:22 looks liek 0.8.3 is 2.4.19 for h36 Feb 16 02:30:38 /dev/root on / type jffs2 (rw) Feb 16 02:30:38 none on /dev type devfs (rw) Feb 16 02:30:38 proc on /proc type proc (rw) Feb 16 02:30:38 tmpfs on /var type tmpfs (rw) Feb 16 02:30:38 tmpfs on /media/ram type tmpfs (rw) Feb 16 02:32:10 looks ok...althought I don't know why it would be /dev/root.... Feb 16 02:33:28 would have to ask koen i guess Feb 16 02:34:03 root Feb 16 02:37:58 ls -al /dev/root Feb 16 02:38:52 sec i'll have to reboot into it Feb 16 02:40:42 k Feb 16 02:41:35 lr-xr-xr-x 1 root root 10 Feb 13 15:28 /dev/root -> mtdblock/1 Feb 16 02:42:14 ah, ok Feb 16 02:42:19 that makes sense then Feb 16 02:42:28 I'd say do an fsck but that's kind of impossible Feb 16 02:42:56 lol Feb 16 02:43:07 i am just reflashing now Feb 16 02:43:47 too many odd things happening to pinpoint anything Feb 16 02:43:58 load root Feb 16 02:44:08 ok Feb 16 02:45:03 hmm Feb 16 02:45:15 wonder if the CF reflash works ... Feb 16 03:53:48 ahh what happened to http://www.openembedded.org/ btw appears down for at least a week now =0 Feb 16 03:57:59 longer than that Feb 16 04:00:52 http://www.youtube.com/watch?v=zp-y3ZNaCqs Feb 16 04:00:59 this is crazy ... Feb 16 04:01:56 yep, saw that Feb 16 04:17:27 RP: ping Feb 16 05:38:30 my patches are still in bug tracker :/ Feb 16 05:40:58 rwhitby-laid-off: what happened to ya mate ? Feb 16 05:59:01 shadows: Freescale has shut down the Adelaide System-On-Chip Design Centre. I was chief architect there, architecting the next generation PowerQuicc III (MPC85xx) design. Feb 16 05:59:08 shadows: eek...you're patching the bb files? Feb 16 05:59:24 Got a good redundancy package (over a year's salary), so it's good news :-) Feb 16 06:02:36 shadows: so do you want both of those sqlite things in? Feb 16 06:08:43 yes Feb 16 06:09:03 from my previous experience as a Gentoo dev, i have been following with OE also Feb 16 06:09:10 that is, to provide only the changes Feb 16 06:09:50 rwhitby-laid-off: :/ Feb 16 06:09:55 both methods are used Feb 16 06:10:00 motolayofforola Feb 16 06:10:01 depends on who it is Feb 16 06:10:46 rwhitby-laid-off: more time to spend on girlfriend/wife/openembedded right? Feb 16 06:16:23 shadows: ok, I'm checking in your stuff... Feb 16 06:16:54 JustinP: who's the best? oh yeah, that's you. Feb 16 06:17:32 and please, the gtk-webcore stuff too. Feb 16 06:18:33 shadows: yep. Got an nslu2-linux presentation at LinuxSA to work on for Tuesday Feb 16 06:18:56 shadows: http://www.linuxsa.org.au/meetings/ Feb 16 06:20:03 super sweet =) Feb 16 06:21:53 hmm Feb 16 06:22:18 how do you figure out the dbus server names ? Feb 16 06:26:04 shadows: please check in real files next time...your patches aren't right.... Feb 16 06:26:30 shadows: when you're altering a file in MT patches are good/fine, but when it's new files, just attach those Feb 16 06:26:54 hmm Feb 16 06:27:03 silly thing Feb 16 06:27:16 no hay servidor -> there is no servant Feb 16 06:27:30 but why is there no server ... Feb 16 06:28:37 shadows: for example, neither of the main.mk patches is needed if I copy the main.mk from 2.8.16, which I figured is what you did.... Feb 16 06:29:23 shadows: wait...or... Feb 16 06:29:23 grr Feb 16 06:29:50 shadows: argh, nm.... Feb 16 06:39:34 geez, that compiled quick Feb 16 06:40:43 I asked earlier today, but I might as well try again: Does anyone know how to package something that builds a binary and then tries to use that binary for a later part of the build? Feb 16 06:41:14 isnt that how most libraries are used? Feb 16 06:42:08 well, in this case it builds a little c program, then tries to run it to create a file that's used later in the build Feb 16 06:42:27 taht should be handled by it's build system Feb 16 06:42:37 ie, make, smake, etc Feb 16 06:43:57 shadows: please have RP look at 676, I'd rather not check in kernel patches.... Feb 16 06:44:30 emte: well, it's build system compiles it into an ARM binary, then tries to run it in my OE tree on an x86 Feb 16 06:45:53 hmm Feb 16 06:46:24 I think there must be some other package like that...I just need some kind of an example to go buy Feb 16 06:46:36 what are you building? Feb 16 06:46:57 i can only think of packages taht have tests in them offhand Feb 16 06:47:18 enca (the extremely naive charset analyzer) which is a dependency of fbreader Feb 16 06:47:27 johnX: oe sets BUILD_CC to the build machine's compiler. generally you'd modify its buildsystem to seperate build from target, compiler wise, so you can pass that in to build that tool thats ued in the build Feb 16 06:47:40 s/ued/used/ Feb 16 06:47:40 kergoth meant: johnX: oe sets BUILD_CC to the build machine's compiler. generally you'd modify its buildsystem to seperate build from target, compiler wise, so you can pass that in to build that tool thats used in the build Feb 16 06:47:49 hey kergoth Feb 16 06:48:01 hey Feb 16 06:48:33 i dont suppose you'd have any idea where i'd find eth dbus name and port numbers do ya? Feb 16 06:48:47 ok, that gets me halfway there...I just have no idea where to start in modifying the buildsystem Feb 16 06:49:10 I'm a little out of my depth in hacking up autoconf based stuff Feb 16 07:08:33 all checked in but not yet pushed Feb 16 07:08:43 shadows: going to test building the osb stuff Feb 16 07:08:59 JustinP: okay, it should be alright. good that you are testing it Feb 16 07:09:30 will take a while to compile them, though.... Feb 16 07:09:49 JustinP: that kernel patch has net effect zero Feb 16 07:10:04 JustinP: i respect that you don't want to touch it, but it's something RP already knows about Feb 16 07:10:11 he's not willing to spend time to check in a fix Feb 16 07:11:02 so, that's your call if you want to commit a fix, it will just be reversed later when there is a new kernel pushed into the tree Feb 16 07:11:57 the patch moves code around in the same file scope, so really the only effect it has is to make that file compile without errors, and the errors were only showing up on gcc4 Feb 16 07:12:49 i'd make the argument that gcc 4.0.2 should be what everyone is targeting if not gcc 2.95, though this may not be realistic as a goal for the next month, perhaps in two months it would be achievable Feb 16 07:14:05 shadows: what do you mean about 2.95? Feb 16 07:14:10 shadows: we are NOT NOT NOT using 2.95 Feb 16 07:14:18 yes, we are for the embeddix kernel Feb 16 07:14:26 well yes, for that POS Feb 16 07:14:33 but that's not what we use for the rest of the programs! Feb 16 07:14:45 ;) Feb 16 07:15:03 that is what i mean, if not for that exception, it should be the rule to use gcc4 Feb 16 07:15:04 they use 3.4.4 I think (the -cross package...) Feb 16 07:15:11 oh, I see Feb 16 07:16:18 my experience with making gpe-image build on a gcc4 cross initial, lends me to believing that gcc4 is where we should be doing development with as a cross initial and cross Feb 16 07:16:54 the differences are minor, and where there were differences it was a matter of the programmer who wrote nonsense code Feb 16 07:17:40 * emte still maintains that 4.0.x is too new Feb 16 07:17:44 fix the nonsense code, or enable the gcc option which provides backward compatibility, and you've just fixed or documented the buggy code Feb 16 07:17:56 when you look at all the initail problems there were with 3.0.x Feb 16 07:18:01 initial* Feb 16 07:18:13 no, gcc3 was a completely new system from gcc2 Feb 16 07:18:30 that is unfair to compare gcc2->gcc3 against gcc3->gcc4 Feb 16 07:18:34 as is gcc4 from what i understand Feb 16 07:18:41 not so much Feb 16 07:19:07 with gcc2->gcc3 you had the integration of egcs and mainline Gnu C Compiler Feb 16 07:19:33 the focus shift away from the C compiler and into a generic framework for language compilers Feb 16 07:19:59 brand new support structures from funded developers (redhat et. al) Feb 16 07:20:12 it was a lot of change in not just the version numbers Feb 16 07:20:52 now the changes are more in the platform support, not so much in the addition or removal of missing-and-needed features Feb 16 07:21:25 shadows: I'll check in the kerne update Feb 16 07:22:17 my key argument that gcc4 should be primary, is the use of amd64 cpu computers; amd64 platform is the most cost effective i've seen to date and hilights code which makes stupid assumptions Feb 16 07:22:34 hi Feb 16 07:22:41 hey Feb 16 07:22:44 JustinP: okay Feb 16 07:22:53 shadows: gcc4 means also patching stuff to get it built Feb 16 07:24:00 i don't see the problem with this Feb 16 07:24:29 I see - lack of devels to patch everything Feb 16 07:24:34 many of us simply don't know what the differences are between gcc3 and 4 Feb 16 07:24:52 as a meta-distro tool, OE devs and project devs should have tight communication with upstream devs Feb 16 07:24:58 for example we cannot ship konqueror-embedded because it fails under gcc4 (already patched to gcc34) Feb 16 07:25:13 hrw: want me to look at it? Feb 16 07:25:43 right now these things are not looked at by anyone i know of, they are silently ignored and treated as impossible Feb 16 07:25:54 this silent treatment is a bad thing Feb 16 07:26:01 shadows: no need - we plan to build OZ/Fam with gcc 3.4 and nothing else use it Feb 16 07:26:04 it works on what we have Feb 16 07:26:11 and it works on hybrid systems Feb 16 07:26:23 it's not the kind of ignoring you're talking about... Feb 16 07:26:25 shadows, i am curious how you classify amd64 as cost effective ? Feb 16 07:26:37 what it takes is someone with an amd64-only machine to fix it...that's you right now, shadows Feb 16 07:26:42 emte: per performance, it is the fastest Feb 16 07:26:50 heh Feb 16 07:26:57 i need to hear about these issues though Feb 16 07:27:09 if a developer says "no, we are not going to ever use gcc4" Feb 16 07:27:14 then i am not going to spend time to do it Feb 16 07:27:16 shadows: per performance we should use opterons 875 in 8-cpu configs Feb 16 07:27:21 who ever said that? Feb 16 07:27:33 hrw: i meant to say cost per performance Feb 16 07:28:00 * emte sees no benifits of an amd64 platform unless i am runnig a large db Feb 16 07:28:32 the athlon64 3000 xp+ chip i'm running, as an example, runs 32-bit code faster than a comperable 32-bit processor Feb 16 07:28:38 the memory controller is on die Feb 16 07:29:14 shadows: I know a64 speed - use ewi for builds (2xOpt) Feb 16 07:29:29 * france is away: Away Feb 16 07:29:35 but at home I have 1.8GHz athlonxp, at work 1.4G axp Feb 16 07:29:46 and this will not change in next few years Feb 16 07:29:52 hrw: if you're using ewi for builds, then you may be aware that gcc3 outputs incorrect code when the host cross initial is gcc3? Feb 16 07:30:25 most of the time it is a larger number of instructions that do absolutely nothing useful Feb 16 07:30:30 shadows: it's probably a hybrid system Feb 16 07:30:32 but sometimes it is wrong Feb 16 07:30:51 shadows: I had no problems on a Gentoo amd64 hybrid system before.... Feb 16 07:30:57 mmhm Feb 16 07:31:06 shadows: currenly I used opie and gpe from this machine and no problems. same with over 100 testers Feb 16 07:31:12 shadows: (well, I had some problems with system includes, but that was pretty simple) Feb 16 07:31:29 32/64 may not give the same results as a true 64 system Feb 16 07:31:40 exactly Feb 16 07:31:42 oh hmm Feb 16 07:31:57 shadows: grrrr Feb 16 07:32:05 JustinP: trouble? Feb 16 07:32:09 shadows: cc1plus: error: unrecognized command line option "-fno-threadsafe-statics" Feb 16 07:32:17 must be a gcc4-only option Feb 16 07:32:18 guh Feb 16 07:32:24 that's my fault Feb 16 07:32:35 how do i insert logic to only do that if it's GCC4 ? Feb 16 07:32:44 you test that your using gcc4 Feb 16 07:32:56 and limit only to gcc4 Feb 16 07:32:56 emte: that's not helping... Feb 16 07:33:01 by the way, that option is 100% equivilent to the options the actual Apple developers use Feb 16 07:33:06 they are using gcc4 i think Feb 16 07:33:27 JustinP, its an option to keep it away from the working 3.4.4 stuff Feb 16 07:33:35 emte: what option Feb 16 07:33:47 to test and override if passed Feb 16 07:33:47 emte: he was asking *how* to test Feb 16 07:34:06 ah Feb 16 07:34:18 python magic in the bb file Feb 16 07:34:26 emte: we need a patch but only for a specific gcc version... Feb 16 07:34:32 right Feb 16 07:34:46 so you test for it Feb 16 07:34:53 or you can add a flag and use that Feb 16 07:34:55 JustinP: can we make that version of the bb dependent on gcc4 being used as the cross? Feb 16 07:35:00 i would like to do that Feb 16 07:35:05 shadows: don't know.... Feb 16 07:35:16 shadows: python magic again.... Feb 16 07:35:47 why make that version GCC4-only? seems we just need to remove that patch for non-GCC4.....makes sense to include the newer version for other people too Feb 16 07:36:03 no it doesn't Feb 16 07:36:10 why not? Feb 16 07:36:12 the newer version depends on that fix Feb 16 07:36:25 JustinP, just test PREFERRED_VERSION_gcc or PREFERRED_VERSION_gcc-cross Feb 16 07:36:25 ....for gcc4.... Feb 16 07:36:41 yeah. the old version is fine for gcc3 Feb 16 07:36:51 the only reason you'd want the newer version is for gcc4 Feb 16 07:36:52 are you sure the new version won't work with gcc3? Feb 16 07:37:23 i wouldn't know, gcc3 is generally broken on my setup Feb 16 07:37:33 exactly Feb 16 07:37:39 so what was wrong without that compile option? Feb 16 07:37:52 it seems to me that that option would only be required to fix the compile in gcc4.... Feb 16 07:38:10 was it a compile-time problem or runtime? Feb 16 07:38:15 either way I can test it Feb 16 07:38:17 without the compile option, you are compiling code which uses statics in threads Feb 16 07:38:21 which is very bad Feb 16 07:38:38 and if you don't document that, it will cause problems you will not be able to find a reason for having Feb 16 07:38:56 it was compile time and link time Feb 16 07:39:08 and you're sure that gcc3 will compile it with statics in threads? Feb 16 07:39:09 the code would compile, it would even link sometimes, but if you ran that, it failed Feb 16 07:39:14 sometimes it would not link Feb 16 07:39:26 gcc3 default behavior for that case is undefined Feb 16 07:39:31 so have fun with it :( Feb 16 07:39:35 hmmmm Feb 16 07:40:00 i asked the gcc devs, and a few told me that they weren't willing to answer questions about gcc3 Feb 16 07:41:05 grr Feb 16 07:41:14 * emte curses at the code Feb 16 07:41:36 if it gives me "no hay servidor" one more time .... Feb 16 07:41:48 * emte grumbles Feb 16 07:44:28 * JustinP tries some magic... Feb 16 07:45:11 you must roll a 3 or higher for the magic to succeed Feb 16 07:45:17 * hrw updates 3.5.4-test5 feeds Feb 16 07:45:21 * JustinP rolls a D20 Feb 16 07:46:03 * JustinP gets a 6 Feb 16 07:46:09 lol Feb 16 07:46:30 i think i am rolling a d101 and getting 1 Feb 16 07:46:40 hehe Feb 16 07:47:03 emte: I feel your pain Feb 16 07:47:08 thing is its not my code ... Feb 16 07:47:17 its turran's test code Feb 16 07:47:40 and i "think" my proble is an incorrect name Feb 16 07:47:44 problem Feb 16 07:47:59 but its hard to tell cuz everything is in spanish Feb 16 07:49:21 hmm, that could be a problem Feb 16 07:49:49 meanwhile, all I want to do is have everything in a certain subdirectory be compiled with BUILD_CC, but I can't figure out where to set CC = $(BUILD_CC) so that it sticks Feb 16 07:49:58 the code could help JustinP with some of eth e17 stuff when i get it figured out actually Feb 16 07:50:27 help reduce some overhead Feb 16 07:51:30 johnX, not sure what you mean ... Feb 16 07:51:59 wouldnt CC = $(BUILD_CC) just set CC to the build compiler you want to use? Feb 16 07:52:11 johnX: you can always patch the Makefile (or sed -i it) in do_compile_prepend Feb 16 07:52:33 JustinP: That might be just what I'm looking for Feb 16 07:55:49 johnX: hehe. I used sed fixes to fix most of the problems I found.... Feb 16 07:56:12 johnX: you might be able to patch Makefile.in in that dir as well... Feb 16 07:56:15 ir Makefile.am Feb 16 07:56:41 I thought about that...and tried it Feb 16 07:57:27 I set CC = $(BUILD_CC) right near the top of Makefile.in and it gets wiped out during configure I think Feb 16 07:58:07 * emte gives up and goes to bed Feb 16 07:58:08 well a do_compile_prepend hack is fine by me.... Feb 16 07:58:44 a real fix would be better of course, but for us mortals who don't understand the inner working of autofoo.... Feb 16 07:59:11 * JustinP tries more magic Feb 16 07:59:19 great...cause otherwise I'll stick a little note in the .bb that says "to compile this go get qemu and an ARM chroot" Feb 16 07:59:42 which is significantly more of a hack than patching one line Feb 16 08:01:48 cu Feb 16 08:02:52 wow, there's a great app roadnav for gps Feb 16 08:02:57 never heard of it before Feb 16 08:03:07 i'm trying it out on my desktop computer, it is amazingly effective Feb 16 08:03:29 are there any python/bb file gurus around? Feb 16 08:05:37 JustinP: i would leave it in as-is with a comment that says "# FIXME: bug number (bug no) " Feb 16 08:05:50 and have the bug number in there, reopened Feb 16 08:06:09 if someone is bothered enough by it, they can write the python to fix it Feb 16 08:06:28 or change their code to depend on older version of the bb Feb 16 08:06:35 i think that is acceptable Feb 16 08:06:46 I suppose I could just make it DEFAULT_PREFERENCE="-5" or so.... Feb 16 08:07:02 why? Feb 16 08:07:18 because according to you it won't work with gcc3 Feb 16 08:07:35 so we can't just let it go building with gcc3, can we? Feb 16 08:08:20 no according to me the developers of the code that gtk-webcore is a port of develop with gcc4 Feb 16 08:08:23 I have something that *should* work.... Feb 16 08:08:35 and?> Feb 16 08:09:02 what's your point? most of us are still using gcc3.... Feb 16 08:09:14 yes, so the point is that it must be fixed to work with gcc3 Feb 16 08:09:26 not DEFAULT_PREFERENCE'd into obscurity Feb 16 08:09:43 well something has to be done... Feb 16 08:10:26 I think I could make the patch conditional on the gcc version....but again you said that it won't work with gcc3... Feb 16 08:10:27 we can specify versions of bb packages to use in the meta builds yes? Feb 16 08:10:46 it's not going to break a build of some task if the task defines the earlier version Feb 16 08:11:04 so the question is more one of, do we want the work-around to be in the tasks which need to be built Feb 16 08:11:07 or in the package files Feb 16 08:11:17 i don't think this point is properly addressed yet Feb 16 08:11:36 the build itself should be conditional on the gcc cross being used Feb 16 08:12:16 applying patches conditionally is unacceptable Feb 16 08:12:37 the source code must not change in a way that is not accounted for by the SCM Feb 16 08:21:25 Anyone have any ideas on why libsvg-cairo might fail to configure, It tells me 'No package 'libsvg' found', so I guess it's more of a problem with libsvg missing, so the problem is why is libsvg missing? :) Feb 16 08:23:23 morning all Feb 16 08:23:44 ashak: I suggest pulling and updating, then rebuilding libsvg Feb 16 08:23:58 ashak: libsvg wasn't staging due to some missing () Feb 16 08:25:11 JustinP: I pulled and updated about 30 mins ago, although I have a feeling that I didn't clean libsvg... which could explain it. Feb 16 08:26:16 ashak: yeah, clean and rebuild libsvg Feb 16 08:26:42 JustinP: thanks Feb 16 08:28:35 shadows: DEFAULT_PREFERENCE = "${@['-1', '1'][bb.data.getVar('PREFERRED_VERSION_gcc-cross', d, 1).split('.')[0] == '4']}" Feb 16 08:28:55 morning Dirk Feb 16 08:29:02 JustinP: Cool, that built fine. Feb 16 08:30:25 shadows: will push in a minute Feb 16 08:32:29 morning Feb 16 08:32:31 hi dirk Feb 16 08:32:37 shadows: your patches are pushed Feb 16 08:32:47 (well, they're being pushed) Feb 16 08:35:43 03justinp 07org.oe.dev * r13e08f59... 10/packages/sqlite/sqlite_2.8.16.bb: sqlite: bug #687 fix INTPTR_TYPE on 64-bit hosts Feb 16 08:35:47 03justinp 07org.oe.dev * rccca9b7b... 10/packages/sqlite/ (4 files in 2 dirs): sqlite: Add 2.8.17 from bug #688, thanks to Eric Shattow Feb 16 08:35:51 03justinp 07org.oe.dev * r19b5a455... 10/packages/gtk-webcore/osb-jscore_20060212.bb: osb-jscore: add new version to fix bug 681, thanks to Eric Shattow Feb 16 08:35:55 03justinp 07org.oe.dev * r6279f0e9... 10/packages/gtk-webcore/osb-nrcore_20060212.bb: osb-nrcore: add 20060212 to fix bug # 682, thanks to Eric Shattow Feb 16 08:35:59 03justinp 07org.oe.dev * raf5b6e64... 10/packages/gtk-webcore/osb-nrcit_20060212.bb: osb-nrcit: Add 20060212, fixes bug #683, thanks to Eric Shattow Feb 16 08:36:04 03justinp 07org.oe.dev * rabf3b27d... 10/packages/gtk-webcore/ (6 files in 2 dirs): gtk-webcore: add gcc4 patches for jscore, nrcit, and nrcore, fixes bug #684, thanks to Eric Shattow Feb 16 08:36:08 03justinp 07org.oe.dev * r75fa541b... 10/packages/gtk-webcore/osb-browser_20060212.bb: osb-browser: add version 20060212, fixes bug #685, thanks to Eric Shattow Feb 16 08:36:12 03justinp 07org.oe.dev * rb537e561... 10/packages/linux/ (2 files in 2 dirs): linux-openzaurus-2.6.15: fix gcc4 compilation, fixes bug #676, thanks to Eric Shattow Feb 16 08:36:16 03justinp 07org.oe.dev * r1b1428d3... 10/packages/gtk-webcore/ (4 files): gtk-webcore: make newer versions only compile with GCC4 by default Feb 16 08:38:28 shadows: and there you go Feb 16 08:39:10 hi Marcin, Liam Feb 16 08:39:38 hrw|work: Hey Marcin Feb 16 08:39:45 hi Liam Feb 16 08:56:53 shadows: ok, the new versions aren't build for me, so it's working right Feb 16 08:58:36 <_law_> i think www.hwaci.com is down :-( cant download sqlite-2.8.17.tar.gz Feb 16 08:59:01 pfff Feb 16 08:59:08 it worked for me 1 hour ago Feb 16 09:04:43 Damn, it appears taht something that's built within gpe-image requires that my kernel be built. Feb 16 09:15:23 morning all Feb 16 09:16:27 RP: morning Richard Feb 16 09:17:18 morning Richard Feb 16 09:22:57 hi Richard Feb 16 09:24:07 <_law_> www.hwaci.com works again Feb 16 09:40:01 hi all Feb 16 09:41:03 morning Feb 16 09:41:56 hi koen, XorA Feb 16 09:42:50 hey RP Feb 16 09:43:59 <_law_> any ideas why gpe-image doesnt build? http://pastebin.com/557367 Feb 16 09:44:50 looks like someone RDEPENDS on kernel-modules and OE can't find them Feb 16 09:45:47 <_law_> koen, i get this problem since linux-openzaurus-2.6.15-r6 Feb 16 09:47:20 hi koen, XorA Feb 16 09:47:31 hey do13_ Feb 16 09:47:52 hi XorA koen Feb 16 09:48:02 hey hrw|work Feb 16 09:49:23 _law_: It looks like your kernel didn't get built for some reason... Feb 16 09:49:45 <_law_> RP, should i try to rebuild the kernel? Feb 16 09:52:02 _law_: Probably - you could have a look into tmp/deploy/ipk and see if the kernel modules are there (and check tmp/work to see if a kernel built at all) Feb 16 10:01:52 <_law_> RP, NOTE: package linux-openzaurus-2.6.15: completed Feb 16 10:01:59 <_law_> try to build gpe-image Feb 16 10:12:19 <_law_> RP, ipk also in /deploy but gpe-image doesnt work Feb 16 10:13:37 morning Feb 16 10:15:12 hi reenoo Feb 16 10:15:59 hey pb__ Feb 16 10:23:13 pb__: I have a patch for #1462 (Media category) btw. will check it in tonight. Feb 16 10:24:29 reenoo: awesome, thanks Feb 16 10:25:24 <_law_> here is the new gpe-image problem log :-) : http://pastebin.com/557390 Feb 16 10:39:32 morning all Feb 16 10:40:21 reenoo: hi Feb 16 10:45:37 hi katossi_uni Feb 16 10:46:04 reenoo: anything new in the java/gcj realms? or are you done with that? Feb 16 10:46:59 I'm not exactly done with it, but I don't have the time to work on it either ATM Feb 16 10:51:32 RP: kernel module for kexec like switch? http://www.piro.hopto.org/~piro/pukiwiki/pukiwiki.php?%5B%5BKernelSwicher%A4%F2%BB%C8%A4%A6%5D%5D Feb 16 10:51:37 japanese **** ENDING LOGGING AT Thu Feb 16 10:59:56 2006