**** BEGIN LOGGING AT Thu Sep 22 02:59:56 2005 Sep 22 03:01:29 erkki: strange.. Sep 22 03:02:04 I have tried building, cleaning, etc the packages, still the same Sep 22 03:02:31 but as I'm trying to add a new machine to oe, and it's my first contact with oe, then possibly some enviroment or config issue Sep 22 03:13:22 morning Sep 22 03:13:52 morning XorA Sep 22 03:13:52 Hi XorA Sep 22 03:16:04 hi XorA Sep 22 03:18:19 while checking in the linux kernel patches for pxa27x-2.4.20 .. i notice that the nwfpe patch by tetsu.. is not there.. Sep 22 03:19:53 RP: I 'charged' my Zaurus last night, but it only got to 45% (and the battery is not *that* bad - it tends to start at 90%), I don't suppose you've seen any interesting charging bugs...? Sep 22 03:20:42 zuery: noone maintain those zaurus kernels probably - if you think it is worth adding and can provide patch which apply then go ahead Sep 22 03:22:03 if anyone could take a look at the following oe build failure, i'd appreciate it: http://erkki.itech.ee/loox/bootstrap_log Sep 22 03:22:56 erkki: DISTRO = familiar probably should be DISTRO = familiar-0.8.3 Sep 22 03:26:47 XorA: that's wierd, I have DISTRO = "familiar-0.8.3" Sep 22 03:26:51 .. in local.conf Sep 22 03:26:55 maybe it's overriden somewhere Sep 22 03:27:27 erkki: ah its ok, the numbers get chopped off somewhere before its printed Sep 22 03:27:35 erkki: just checked mine Sep 22 03:27:41 XorA: k Sep 22 03:28:02 erkki: no idea then, as long as you OE checkup is upto date it should work Sep 22 03:28:31 XorA, yeah, the checkout say's its up to date Sep 22 03:33:51 i think i got that problem before.. Sep 22 03:34:24 do all over again... Sep 22 03:35:00 i rm all the folder and do again .. with proper configuration Sep 22 03:35:38 zuery: I have built it already 2 times:p Sep 22 03:35:58 .. but as I tweaked some settings in-the-middle, maybe I should try once more Sep 22 03:36:21 erkki: ya.. don't change the settings... half way Sep 22 03:37:13 zuery: there's a change, though, that there are some problems in my config, as I'm trying to add a new machine to oe Sep 22 03:37:44 hello Sep 22 03:37:47 i got some problem... with the oe sometimes.. Sep 22 03:39:49 erkki: good luck... in bitbake.... Sep 22 03:39:56 gtg first... Sep 22 04:26:47 hey Sep 22 04:28:06 hey zecke Sep 22 04:39:12 hey zecke Sep 22 04:40:09 koen: is ewi fine again= Sep 22 04:40:38 zecke: I've informed my boss Sep 22 04:40:50 and he should send a reboot monkey into the server room Sep 22 05:01:32 03tmbinc 07org.oe.dreambox * r56c60ad9... 10/packages/dreambox/ (2 files in 2 dirs): dreambox-buildimage: new flash partition layout for dm702x Sep 22 05:01:36 03tmbinc 07org.oe.dreambox * rd0959e17... 10/packages/dreambox/dreambox-secondstage.bb: dreambox-secondstage: use version 35 for dm7020 in order to account for new flash partition layout Sep 22 05:01:40 03tmbinc 07org.oe.dreambox * r04aa0270... 10/packages/linux/linux-dm7020.bb: linux-dm7020: add patch to change partition layout Sep 22 05:19:58 RP: thanks for that CF mounting fix, udev hotplugging is now a dream on my machine Sep 22 05:20:10 RP: and works so much faster than what was there before Sep 22 05:24:59 RP: I get the idea that Alan Cox is misunderstanding the issue by confusing card and adapter Sep 22 05:26:36 bloody welsh :/ Sep 22 05:27:15 RP: is the CD line not sufficient for hotplug? Sep 22 05:55:03 koen: did you try aircrack on OS X? Sep 22 05:57:20 zecke_: nope Sep 22 05:57:34 I only use kismac Sep 22 05:57:38 Al deserves broadband internet ;) Sep 22 06:13:59 morning Sep 22 06:26:14 hi chouimat Sep 22 06:38:43 Hi. Is there a way to regenerate the ipk/Packages file after building a new ipk? Sep 22 06:39:50 ipkg-make-index Sep 22 06:39:57 rebuild bootstrap-image Sep 22 06:40:01 select any of them Sep 22 06:40:47 Thanks. Sep 22 06:43:34 np Sep 22 06:52:02 woglinde: hey Sep 22 07:21:00 03ph5 07org.oe.dev * r456405c1... 10/packages/bluez/ (4 files): bluez-libs,bluez-utils: new version 2.21 Sep 22 07:21:04 03ph5 07org.oe.dev * rab4108f7... 10/packages/pmount/ (files/no-hal.patch pmount_0.9.4.bb): pmount: add a recipe for pmount 0.9.4 Sep 22 07:21:15 RP: I just added pmount to oe. But it doesn't work because /dev/mmcblk0p1 is not marked as removable. Sep 22 07:39:06 morning Sep 22 07:39:54 damn storms. trees down all over the northern suburbs here, lost power till 3am last night Sep 22 07:39:57 heh Sep 22 07:40:02 glad we dont get those often Sep 22 07:41:38 hey mickeyl & kergoth Sep 22 07:43:17 where i can download oe full distribution in one archive? Sep 22 07:43:35 'oe full distribution'? Sep 22 07:43:45 hi mickeyl Sep 22 07:43:55 =/ Sep 22 07:44:11 * koen_ gets out his soldering iron to fix the serial console on his slug Sep 22 07:44:11 mickeyl: http://www.oesf.org/forums/index.php?showtopic=14897 Sep 22 07:46:12 hrw|work: http://bugs.openembedded.org/show_bug.cgi?id=339 Sep 22 07:46:53 lovely R DJ Sep 22 07:46:56 hrw|work: from ewi..nl/OE (once it is up) Sep 22 07:48:37 hrhr: it was for you Sep 22 07:49:53 hrw|work: yeah i know. I'll inspect the state of OZ 3.5.4 @ weekend and - if they're somewhat usable - release some testing images Sep 22 07:50:54 hm.. I have to check do we fit opie in 16M flash Sep 22 07:52:24 hrw|work: surely opie is just pim , do you need to fill 16M flash for pim ? *snigger snigger* Sep 22 07:52:52 Open Personal Information E-something Sep 22 07:53:48 ah open personal information, that's like my bank account number is .. my PIN is ... etc Sep 22 07:54:54 its amazing what a bad back can do for your humour, for that i am sorry Sep 22 07:55:27 two days staring at the ceiling drove me mad Sep 22 07:56:40 ade|desk: because they hit the framebuffer direct they use 16M to get the extreme speed Sep 22 07:58:26 lol Sep 22 07:59:54 question about bitbake: are errors like this normal Sep 22 08:00:10 NOTE: Handling BitBake files: - (0863/2754) [31 %]ERROR: kaffe.inc:76: unparsed line: 'pkg_postinst_${PN} () {' while parsing /opt/oe/org.openembedded.dev/packages/kaffe/kaffeh-native_cvs.bb Sep 22 08:00:30 I googled some earlier error messages and turned up some logs from this IRC chatroom Sep 22 08:00:52 however based on those it wasn't clear how to resolve unparsed line errors. I got the impression I might be able to just ignore them Sep 22 08:01:06 sounds like you need to upgrade bitbake or oe or both Sep 22 08:01:51 ah... I grabbed the latest packaged bitbake. I would download the latest in svn however I'm behind a firewall. Is the 1.2.1 package too old? Sep 22 08:02:04 MP0: 1.3.2.1 is current Sep 22 08:02:08 iirc Sep 22 08:03:03 MP0: yes, you need a more recent bitbake Sep 22 08:03:46 since http://developer.berlios.de/project/showfiles.php?group_id=2701 shows only 1.2.1, I guess I'll have to figure out how to get it out of svn. Thanks guys. Sep 22 08:05:12 zecke_: someone other then kergoth has rights to release bitbake? Sep 22 08:06:18 hrw|work: maybe phil Sep 22 08:06:32 http://handhelds.org/~zecke/downloads/bitbake/ Sep 22 08:06:52 I've just managed to build gpe-mini-browser and it's working great. Are there any config options? eg: loading images Sep 22 08:09:02 johbrahms: better ask in #gpe Sep 22 08:09:22 shall do Sep 22 08:09:33 re Sep 22 08:10:03 zecke_, much obliged! Sep 22 08:25:28 OE wiki (moinmoin) does not work.. Sep 22 08:26:43 yep Sep 22 08:26:48 /dev/ida/disc0/part8 89G 89G 0 100% /home Sep 22 08:26:56 thats never good Sep 22 08:29:46 * koen_ removes some familiar snapshots Sep 22 08:29:49 /dev/ida/disc0/part8 89G 81G 7.6G 92% /home Sep 22 08:29:53 much better :) Sep 22 08:31:31 heh.. that was you nasty boy... Sep 22 08:34:37 maybe I shouldn't have include a world feed for each snapshot :) Sep 22 08:36:39 how much snapshots you keep? Sep 22 08:36:53 s/much/many Sep 22 08:37:15 I removed the 0.8.1 and 0.8.2 snapshots + snapshot feeds Sep 22 08:38:18 so only http://familiar.handhelds.org/snapshots/v0.8.3/ remains Sep 22 08:53:58 RP: http://lkml.org/lkml/2005/9/22/176 <- success! Sep 22 08:54:08 (if I understand it correctly) Sep 22 09:11:42 koen_: r DJ bug closed. Sep 22 09:11:47 user error Sep 22 09:11:48 cu Sep 22 09:13:21 as expected Sep 22 09:49:37 hi Sep 22 09:49:44 hey hrw Sep 22 09:51:26 koen: Yes, I seem to have managed to convince Alan Cox. Just need to get a patch accepted now :) Sep 22 09:51:38 lardman|work: There are no known charging bugs Sep 22 09:52:08 pH5|away: I have a version of pmount outside OE but I'm not happy with it for the reasons you describe Sep 22 09:52:22 I will point out this bug to the developer... Sep 22 09:52:35 let see does opie-image for c7x0 will fill in 16M without extra stuff (I lack clean 'armv4' build) Sep 22 09:52:44 All: For reference, I'm away tomorrow and most of the weekend Sep 22 09:53:42 RP: mickeyl and I will covertly branch ;) Sep 22 09:54:03 actually I'll be away too :) Sep 22 09:54:09 pH5|away: I think we will end up using pmount. For now we can rely on the mounts being in fstab until pmount is useable. globbing in the pmount.allow file would solve some of our problems Sep 22 09:54:43 RP: thanks for your persistence on lkml! Sep 22 09:54:53 heh.. looks like its too big ;( Sep 22 09:55:36 mreimer: I do get a bit annoyed when people tell me things they "know" without reading my emails. The "GNOME and KDE work" comments were getting to me a bit ;-) Sep 22 09:56:24 If anyone ones to submit patches for any of the 4 bugs we now have listed, feel free - I probably can't work on it until the middle of next week now Sep 22 09:56:46 (that's listed on LKML related to my original one) Sep 22 09:56:57 RP: linuxexpo? Sep 22 09:57:06 koen: Not yet Sep 22 09:57:13 That's next month Sep 22 09:57:19 5th 6th IIRC Sep 22 09:57:40 I'm camping in Yorkshire :-/ Sep 22 09:58:11 sounds like rain Sep 22 09:59:06 Last time I did this, they lost all power to the village for several hours one night. Internet access is way beyond them ;-) Sep 22 10:00:49 camping... going to Peak? Sep 22 10:01:59 NAiL: No, its a social thing Sep 22 10:02:32 oh, ok Sep 22 10:03:22 just curious.. Been in the area a couple of times ;) Sep 22 10:05:06 I'll be in the Yorkshire Dales. Going by motorcycle - there's some lovely roads around there... Sep 22 10:26:58 cu Sep 22 10:28:37 RP: Okay, hopefully i just did something stupid like unplug the charger Sep 22 10:28:48 RP: Have a nice time camping Sep 22 10:28:50 night all Sep 22 10:28:54 hey zecke Sep 22 10:56:24 hmm Sep 22 10:56:35 koen, i love you Sep 22 11:04:17 dont worry zecke i love you too Sep 22 11:04:23 I asked about a wierd oe build problem here before, maybe there's someone here now who could help me... http://erkki.itech.ee/loox/bootstrap_log Sep 22 11:06:20 looks liek your missing your dependencies Sep 22 11:06:36 erkki_: is that a build from scratch? Sep 22 11:06:48 if not, rm the TMPDIR and build again Sep 22 11:07:16 koen: yes Sep 22 11:07:37 was the non-version familiar bb destrioyed with the other non-versions? Sep 22 11:08:08 emte: sorry, could you explain what you mean Sep 22 11:08:42 there is/was a push to make all non-version distros fail to build Sep 22 11:08:48 erkki_: can't the loox reuse the h3900 machine type? Sep 22 11:08:49 .. at first I tried to build with a custom kernel & machine conf, twice with the same error, then got the same error with ipaq3900 Sep 22 11:09:00 ask koen i am pretty sure he still has me on ignore Sep 22 11:09:09 apparently a tornado touched down like 2 miles from my house last night. Sep 22 11:09:12 koen: perhaps, I'm just getting used to oe Sep 22 11:09:28 re Sep 22 11:09:37 kergoth`food, you do live near tornado alley now dont you>? Sep 22 11:10:01 minnesota is pretty far north.. dont get them where i'm at all that often Sep 22 11:10:04 but it does happen Sep 22 11:10:10 http://www.startribune.com/stories/1608/5629018.html .. i'm right in the center of that diagram Sep 22 11:10:48 the packages, that seem to be missing, actually are all built Sep 22 11:11:18 erkki_: are they present in the deploy/ipk directory? Sep 22 11:11:21 then there's a wierd line: | ipkg: Cannot create directory `/usr/lib/ipkg': Permission denied Sep 22 11:11:28 koen: yes Sep 22 11:11:30 erkki_: that line is expected Sep 22 11:11:37 k Sep 22 11:13:43 would someone mind committing the patches in bugs 287 and 337? Sep 22 11:17:13 mreimer: package/pcmcia-cs/pcmcia-cs-3.2.8/etc/rc.pcmcia doesn't exist in my repo Sep 22 11:17:53 koen: oops, I think it's actually files/arm/pcmcia; checking... Sep 22 11:18:59 koen: the second patch in that bug should apply to packages/pcmcia-cs/pcmcia-cs-3.2.8/pcic-extra.patch Sep 22 11:19:14 koen: the first should apply to packages/pcmcia-cs/files/arm/pcmcia Sep 22 11:19:36 I applied the first Sep 22 11:19:40 the third fails Sep 22 11:19:52 there should be only two patches Sep 22 11:19:52 the second patch has been disabled in bugzilla Sep 22 11:20:06 yes, the second was obsoleted by the third Sep 22 11:20:31 so apply just the first and third Sep 22 11:21:21 http://bugs.openembedded.org/attachment.cgi?id=269 doesn't apply Sep 22 11:27:03 koen: ok, try http://bugs.openembedded.org/attachment.cgi?id=270 Sep 22 11:27:11 I obsoleted 269 Sep 22 11:27:54 patches to patches twist my brain Sep 22 11:29:03 like a gold brick wrapped in lemon :) Sep 22 11:35:02 koen: also, would you mind adding a blueprobe_0.15.bb? Sep 22 11:43:29 lets see if my new image will behave Sep 22 11:44:09 time to take drugs first Sep 22 11:44:27 koen: thanks for closing 287 Sep 22 11:52:03 wow Sep 22 11:52:22 i was expecting 2.6 to fail ... Sep 22 11:52:32 koen, thanks for closing 337 Sep 22 11:52:37 mreimer: np Sep 22 11:53:54 koen: would you mind committing a couple more things? Sep 22 11:54:14 fire away Sep 22 11:54:29 koen: would you mind adding a blueprobe_0.15.bb, same as the others? Sep 22 11:54:47 mreimer: monotone pull :) Sep 22 11:54:53 :-) ok Sep 22 11:55:02 http://monotone.vanille.de/viewmtn/revision.psp?id=f91c10780dc9c984133a4113600bb7070e4a46ad Sep 22 11:55:26 koen: also, would you add this line to xserver-kdrive_20050207.bb: Sep 22 11:55:33 koen: SRC_URI_append_h3900 = " file://disable-apm.patch;patch=1" Sep 22 11:55:35 03koen 07org.oe.dev * r7d3d2109... 10/packages/pcmcia-cs/ (files/arm/pcmcia pcmcia-cs-3.2.8/pcic-extra.patch): packages/pcmcia-cs/: apply patches from #287 Sep 22 11:55:39 03koen 07org.oe.dev * r2e0e7cea... 10/packages/pcmcia-cs/pcmcia-cs_3.2.8.bb: packages/pcmcia-cs/pcmcia-cs_3.2.8.bb: bump PR after last change Sep 22 11:55:43 03koen 07org.oe.dev * rf91c1078... 10/packages/blueprobe/blueprobe_0.15.bb: packages/blueprobe/: add 0.15 Sep 22 11:59:01 mreimer: anything else? Sep 22 12:00:23 koen: one more: change usb0 in packages/netbase/interfaces to a static config like usbf Sep 22 12:00:23 hmm Sep 22 12:00:34 03koen 07org.oe.dev * r840187fe... 10/packages/keylaunch/files/keylaunchrc: packages/keylaunch/files/keylaunchrc: apply patches from #337 Sep 22 12:00:53 2.6 seems to be working better than 2.4 ... Sep 22 12:00:55 koen: basically just copy the five usbf lines and s/usbf/usb0/ Sep 22 12:02:07 03koen 07org.oe.dev * r12c78fd2... 10/packages/xserver/xserver-kdrive_20050207.bb: packages/xserver/xserver-kdrive_20050207.bb: enable apm hack for h3900 too Sep 22 12:02:26 thanks koen! I think that'll do it for now. Sep 22 12:02:33 is udev working now? Sep 22 12:02:53 the h3900 has which chip? Sep 22 12:03:00 sa or pxa? Sep 22 12:04:16 found it pxa Sep 22 12:07:58 03koen 07org.oe.dev * r73e27d19... 10/packages/netbase/ (netbase/interfaces netbase_4.21.bb): packages/netbase/netbase/interfaces: make usb0 behave like it's 2.4 counterpart, bump PR in .bb Sep 22 12:09:16 bleh the horrible apm hack spreads Sep 22 12:09:33 it does indeed Sep 22 12:09:35 XorA|gone: yes :-( Sep 22 12:09:48 someone needs to fix 2.6 kernel :-) Sep 22 12:10:00 yes please! Sep 22 12:10:17 XorA|gone: is it a problem with the APM implementation? Sep 22 12:10:31 pH5: http://monotone.vanille.de/viewmtn/revision.psp?id=0baef97de8a3421a655e3081bf611dd1aa5c9717 Sep 22 12:10:56 koen: that was fast :) Sep 22 12:11:10 mreimer: two problems, 1) kernel puts any apm listener after apmd into D state and that stops suspend, 2) that old version of kdrive has a mutiple close bug on TS device Sep 22 12:11:33 03koen 07org.oe.dev * r0baef97d... 10/packages/blueprobe/ (blueprobe-0.15/hx4700.patch blueprobe_0.15.bb): packages/blueprobe/blueprobe_0.15.bb: add patch from http://handhelds.org/~bugzilla/show_bug.cgi?id=1398 Sep 22 12:12:03 XorA|gone: who's responsible for the kernel apm code? has anyone asked about this upstream? Sep 22 12:12:33 mreimer: not a clue Sep 22 12:12:56 mreimer: I beleive no 2 is fixed in later kdrives Sep 22 12:13:31 XorA|gone: yes, I think in 20050624 and later, but versions after 20050207 have a nasty bug where the X server probes ttyS0 looking for a mouse, messing up the serial console Sep 22 12:13:44 mreimer: ooops Sep 22 12:16:04 grr stupid usb cradle .. Sep 22 12:16:30 anyone else have the h36/37 usb cradle lock up thier device? Sep 22 12:21:45 mreimer: perhaps we should add a patch to oe and turn this off. i assume there is not one single device that has a mouse on ttyS0? Sep 22 12:22:07 pH5: I agree; let's do it Sep 22 12:22:17 pH5: I asked a couple of times on #xorg, but no answer Sep 22 12:22:54 pH5: I wouldnt even know where to buy serial mice anymore Sep 22 12:23:44 some ts screens (the one on CRTs) interface via serial Sep 22 12:23:58 but lets disable it till someone starts to scream :) Sep 22 12:24:26 disable it only in openzaurus and familiar distribs? Sep 22 12:24:37 disable it entirely Sep 22 12:24:47 since we have a getty on ttyS0 by default Sep 22 12:25:10 koen: should the getty not open the /dev/ttyS0 exclusively anyway? Sep 22 12:25:21 it should, but it doesn't Sep 22 12:25:29 sucks Sep 22 12:25:33 yeah Sep 22 12:26:17 I think the kernel debugging probably mess any serial mouse up anyway :-) Sep 22 12:27:58 Ok, how about this: http://bugs.openembedded.org/show_bug.cgi?id=340 Sep 22 12:28:43 mreimer: heh. looks good to me. Sep 22 12:30:45 03jbowler 07org.oe.dev * rf6740acf... 10/packages/upslug/ (6 files): upslug2: move to upslug2_4, move to file (not cvs) release Sep 22 12:32:30 mreimer: I think we can put it in under _familiar and _openzaurus overrides Sep 22 12:33:00 koen: so we wouldn't have to put it in all the .bb's? Sep 22 12:34:00 we put a SRC_URI_append_familiar += "serial.patch" in the kdrive.bb Sep 22 12:35:22 so it's in monotone now? I'll pull and build Sep 22 12:35:43 not yet Sep 22 12:39:35 ugh Sep 22 12:40:01 i can really notice the 2 orders of magnitudes smaller link Sep 22 12:40:16 100Mbit vs 2-ish mbit Sep 22 12:42:40 03jbowler 07org.oe.dev * r26162f44... 10/packages/devio/ (11 files): devio: move to devio_1.0, move to a file (not cvs) release Sep 22 12:42:44 03jbowler 07org.oe.dev * rca72f385... 10/packages/binutils/ (3 files in 2 dirs): (log message trimmed) Sep 22 12:42:45 binutils_2.16: patch thumb trampoline and glue code generation Sep 22 12:42:47 This patch fixes problems with the ARM thumb specific code generation. It Sep 22 12:42:49 also adds debugging to bfd/elf32-arm.c to detect ref counting problems - Sep 22 12:42:51 one error message can fire in theory in the absence of thumb support if Sep 22 12:42:53 a union refcount field is used to refcount when it is set to something else. Sep 22 12:42:55 The error should be ignorable, in that the code should not have changed in Sep 22 12:46:00 03jbowler 07org.oe.dev * r969c909f... 10/packages/gcc/ (7 files in 2 dirs): Sep 22 12:46:00 gcc_3.4.4: updated thumb patches Sep 22 12:46:00 The four previous thumb related patch files have been combined into one Sep 22 12:46:00 and simplified. Additional problems with call_via_rX functions being Sep 22 12:46:00 called via the PLT have been fixed (the functions can no longer be called Sep 22 12:46:01 this way - they are hidden). Unnecessary adds of exports of these Sep 22 12:46:03 functions from libgcc have been removed. Sep 22 12:50:40 03koen 07org.oe.dev * rdecc2032... 10/packages/xserver/ (2 files in 2 dirs): packages/xserver/xserver-kdrive_20050624.bb: add serial patch from #340 Sep 22 12:52:44 03jbowler 07org.oe.dev * rd8d3e2d9... 10/ (8 files in 3 dirs): Sep 22 12:52:44 uclibc_0.9.28: fix thumb support to allow thumb uclibc Sep 22 12:52:44 This, together with the fixes in gcc and binutils, allows a system to Sep 22 12:52:44 be build with thumb libgcc and libuClibc (etc). ucslugc is changed to Sep 22 12:52:45 release 2 and to use thumb compilation of these modules. Sep 22 12:54:41 jbowler: how much space do you win with thumb/ucslugc? Sep 22 12:54:45 600k? Sep 22 12:56:34 Probably not that much on disk, it was working out at about 80% of the text segment size, but on a raw system there's still quite a bit of data - maybe only 10%. Sep 22 12:56:53 still quite a lot Sep 22 12:57:16 Well, it counts more in the RAM - because there the instructions really are half the size (although there are more of them). Sep 22 12:57:42 On disk (at least in jffs2) there's compression, and 32 bit probably compresses better than 16 bit instructions... Sep 22 12:58:15 Now it's working I think I'll try a build without thumb and without interwork support, then I'll have a firm number. Sep 22 13:00:40 I get this crazy feeling it is faster too... but I think that might be because of what I just did to /lib/ld-uClibc_0.9.28.so in the thumb case (it now skips undefined symbols in places where it didn't before). Sep 22 13:01:03 nice Sep 22 13:01:22 I'm still thinking about reflashing to uclugc Sep 22 13:01:40 koen: how about using the serial patch for xserver-kdrive_cvs.bb too? Sep 22 13:02:15 jbowler: it can fetch two instructions at once ;) Sep 22 13:02:17 koen: if you rebuild now (after those changes I just committed) it is necessary to completely rebuild everything. Sep 22 13:02:24 so it is twice as fast (at least) Sep 22 13:02:25 I'm not terribly fond of adding patches to floating cvs stuff Sep 22 13:02:44 zecke: no, the core just fetches one at a time, but the I-cache line has twice as many Sep 22 13:02:49 jbowler: the only armeb build is on my slug :) Sep 22 13:03:23 Oh, gdb does not work - not at all. Sep 22 13:06:48 join #postfix Sep 22 13:07:16 koen: is the usb0 config going to work when the address/netmask/network/gateway settings are not indented? Sep 22 13:07:35 hmmm Sep 22 13:07:41 it isn't python Sep 22 13:07:48 mut I'll indent them to be sure Sep 22 13:08:06 not sure either Sep 22 13:08:50 I don't see any mention of whitespace in interfaces(5), so maybe it doesn't matter Sep 22 13:09:27 03jbowler 07org.oe.dev * rdcfd3843... 10/packages/linux/ (nslu2-kernel/2.6/thumb-swi.patch nslu2-kernel_2.6.12.2.bb): nslu2-kernel_2.6.12.2: add a patch for thumb breakpoints (via a thumb SWI) Sep 22 13:09:31 03jbowler 07org.oe.dev * rba30bfad... 10/packages/gdb/ (gdb-6.3/thumb-breakpoint.patch gdb_6.3.bb): Sep 22 13:09:31 gdb_6.3: patch the thumb handling to use a thumb swi (not an undefined instr) Sep 22 13:09:31 This doesn't fix thumb debugging (a thumb program run under gdb will crash Sep 22 13:09:31 right at the start) but it does avoid the illegal instruction crash by Sep 22 13:09:31 inserting a swi instead. Sep 22 13:12:52 03koen 07org.oe.dev * r3918e04d... 10/packages/netbase/netbase/interfaces: Sep 22 13:12:52 packages/netbase/netbase/interfaces: indent usb0 entry to make Sep 22 13:12:52 sure broken apps can still read it Sep 22 13:16:43 monotone: multiple update candidates: Sep 22 13:16:45 what's that mean? Sep 22 13:16:59 mreimer: the tree is not merged Sep 22 13:17:03 some schmuck forgot to merge Sep 22 13:17:06 so just wait a bit? Sep 22 13:17:07 (that would be me) Sep 22 13:17:11 :) Sep 22 13:17:43 koen and jbowler have been playing pass the head - I dropped it earlier too. Sep 22 13:19:33 you can see it in the pretty picture at http://monotone.vanille.de/viewmtn/revision.psp?id=a89d5956fffe61c303ff6a056247b5569871e5c5 Sep 22 13:27:13 any other changes? Sep 22 13:27:19 going once... Sep 22 13:27:26 not from me Sep 22 13:27:27 going twice...t Sep 22 13:27:33 * koen close ssh Sep 22 13:27:42 no more monotone for me tonight :) Sep 22 13:28:32 thanks koen! Sep 22 13:28:48 gets monotonous, doesn't it? Sep 22 13:30:02 heh Sep 22 13:31:46 monotone is the only thing which is too big for thumb - the address range of a thumb bl is +/- 22 bits and a PLT function call from the end of the .exe is just too far. Sep 22 13:54:10 03jbowler 07org.oe.dev * r8982ebf1... 10/packages/meta/ (ucslugc-native.bb ucslugc-packages.bb): Sep 22 13:54:10 ucslugc-packages: temporarily remove perl Sep 22 13:54:10 The .so link of libperl doesn't link with anything which provides the Sep 22 13:54:10 _call_via_rX functions, so it breaks on thumb (the previous version didn't Sep 22 13:54:10 work because it pulled _call_via_rX from libgcc_s.so.1 - and that doesn't Sep 22 13:54:11 always work.) Sep 22 14:02:00 Is there an OE package that provides startx? Sep 22 14:02:46 xserver-xorg would be my bet Sep 22 14:03:06 kdrive is configless, so it doesn't require startx-fu Sep 22 14:06:11 I'll put this another way, do we have an alternative for starting kdrive that's not gpe-dm? Sep 22 14:06:24 Xfbdev & Sep 22 14:07:10 cant you just call XFree86 or Xorg directly? Sep 22 14:08:27 RP: what's wrong with gpe-dm? Sep 22 14:09:01 * CosmicPenguin goes back to half heartedly fixing kexec-tools Sep 22 14:09:14 koen: We're not going there ;-) Sep 22 14:09:45 afaik gpe-image is the only image in OE that creates a working X environment for PDAs Sep 22 14:09:54 RP: I installed pmount, but it doesn't want to mount /dev/mmcblk0p1 because mmcblk0 is not marked as removable :-/ Sep 22 14:10:09 pH5: Yes, read my comments earlier Sep 22 14:10:14 pmount is broken Sep 22 14:10:29 that's what I wanted to hear :) Sep 22 14:10:57 koen: The problem with Xfbdev directly is it doesn't result in a useful system. No touchscreen calibration, doesn't start a display manager etc. Sep 22 14:11:27 like I said Sep 22 14:11:30 afaik gpe-image is the only image in OE that creates a working X environment for PDAs Sep 22 14:12:34 gpe-dm does a pretty nice you as display manager Sep 22 14:12:54 koen: We don't have anything for desktops? Sep 22 14:13:19 that uses Xorg Sep 22 14:13:29 bitbake mythfront-image iirc Sep 22 14:14:02 koen: ok, thanks. This needs some thought... Sep 22 14:14:06 RP: looks like fixing pmount is as easy as adding "mmc" to hotplug_buses[] in device_removable() in policy.c. Sep 22 14:14:36 RP: you could try to use pdm Sep 22 14:14:42 RP: I promise nothing though Sep 22 14:14:50 pH5: I'd be interested to know if that works... Sep 22 14:15:28 RP: I agree the situation isn't optimal Sep 22 14:15:47 but the situation right now is you need a bit of GPE to get a working kdrive image Sep 22 14:22:22 RP: it does. it fails on the next step though, because busybox mount doesn't understand all the options Sep 22 14:23:00 RP: also, it tries to mount at /media/mmcblk0p1, which doesn't sound as nice as /media/card Sep 22 14:26:10 is monotone.vanille.de alive? Sep 22 14:27:53 it is Sep 22 14:28:39 of course not - its an inanimate object Sep 22 14:29:09 monotone: read from fd 4 (peer monotone.vanille.de) failed, disconnecting Sep 22 14:35:26 netnix: you have monotone 0.22? Sep 22 14:38:30 pH5: I'm fairly sure it will fail later as the sysfs attribute is set as none removable Sep 22 14:38:41 I have patches for the mount options locally Sep 22 14:39:06 pH5: Let me talk to the pmount maintainer and see what can be negotiated Sep 22 14:39:07 RP: why should it fail later? it just tries to call mount. Sep 22 14:39:13 koen, now, all is ok, but connection is very slow. Sep 22 14:39:28 pH5: busybox mount, not utils-linux mount Sep 22 14:41:19 RP: yes, it would be good if the offending options could be turned off or made optional. Sep 22 14:41:37 or it makes a syscall instead of using mount Sep 22 14:47:46 RP: that's a nice idea. that code might even be smaller than constructing the options for /bin/mount ;) Sep 22 14:49:17 pH5: No, because you also have to write out /etc/mtab :-( Sep 22 14:49:38 don't we symlink that to /proc/mounts? Sep 22 14:50:00 koen: We do, most systems dont Sep 22 15:00:16 'night all Sep 22 15:00:23 'night koen Sep 22 15:01:44 03ph5 07org.oe.dev * ra60d3fd5... 10/packages/pmount/ (files/mmc-fix.patch pmount_0.9.4.bb): pmount: fix pmount to work with mmc block devices Sep 22 15:04:31 hmm zip231 took a walk off oesources? Sep 22 15:11:10 ah info-zip.org is down Sep 22 15:25:42