**** BEGIN LOGGING AT Sat May 27 02:59:56 2006 May 27 05:04:44 sup people? May 27 05:08:03 everything's cool :-) May 27 05:17:07 any ipaq that has fpu ? May 27 06:10:51 CSMan, they all have soft May 27 06:11:04 the pxa ones i belive use hardware tho May 27 06:21:11 XScale-PXA255 rev 6 (v5l) May 27 06:21:19 emte: it'd be nice if they had fpu May 27 06:21:35 it'd drain the battery thou May 27 06:51:44 CSMan, i am not sure that logic makes sense May 27 06:52:30 fpu in hardware would be more efficient than in software battery-wise May 27 06:52:37 i'd expect May 27 06:52:52 but it would cost more to make the chips with builtin fpu handling May 27 06:53:09 emte: absolutely, i meant if that feature is added in hardware it'd increase the ipaq's current power consumption May 27 06:53:32 which is around 1mA May 27 06:53:32 i doubt that myself May 27 06:53:56 that hardware wouldnt get used unless called May 27 06:54:18 simularily the soft code doesnt get called unless used either May 27 06:54:32 which increases cpu load and power draw May 27 06:55:35 if you factor in interrupts and software timing i would think that soft fpu would use more power May 27 06:56:25 since hardware is quick and atonomous for the most part. dump it in, puck it up when done May 27 06:56:31 pick* May 27 06:56:50 no, start - stop - resume stuff May 27 06:56:54 right May 27 06:57:43 so overall, I think it is better to have fpus on handheld devices May 27 06:57:57 that PSP has a pretty fancy config May 27 06:58:24 yes, but if your a company and it costs $5 extra for fpu and your shipping 10M units what would you do? May 27 06:58:51 $50M extra is hard to justify to your shareholders May 27 07:12:30 Morning all May 27 07:19:09 Hi guys. anyone working on putting iPaq rx3715 stuff int OE? May 27 07:20:37 I got some stuff of my own that I could contribute if nobody else is. May 27 07:22:26 myopiate, have you looked at the wiki to see where they are at? May 27 07:22:46 yes, I'm a regular looker at that May 27 07:23:35 we have two guys who are hacking the kernel. but they seem to be working on vanilla kernels rather than hh kernels May 27 07:24:25 we've also been twiddling with the h2200 images, but we should be really building our own. May 27 07:24:31 i think you missunderstand how kernel dev works May 27 07:25:04 i'm here to be enlightened ;-) May 27 07:25:17 general kernel dev is against a vanilla kernel May 27 07:25:44 there is no such thing as a hh kernel until after those patches are applied to a vanilla kernel May 27 07:26:57 as well there is a lot of device crossover in many areas May 27 07:27:31 so you make one proper set instaed of one kernel per device May 27 07:27:51 and call the proper modules and configurations May 27 07:28:20 it is no different from a PC kernel in this respect May 27 07:29:03 embedded hardware is just specilized hardware May 27 07:30:03 the closest PC compairison i can think of are Laptops, which you see special drivers/modules for different manuf May 27 07:34:48 handhelds.org has a CVS repository with kernel source in it. I've been digging around in there and i have been assumimg that familiar release kernels are built from that. May 27 07:35:36 Hang on a sec... let me look at oe and see where it actually get's it's kernel source from. May 27 07:36:47 My issue at the moment is that we have two people working (quite well together) on their own individual kernels and just trickle patches to people interested May 27 07:36:57 myopiate, we get it from kernel.org May 27 07:37:05 and apply patches May 27 07:37:29 altho oesources caches as well May 27 07:37:36 etc May 27 07:40:30 myopiate: familiar does build the kernels for most devices from the handhelds.org cvs tree, but it doesn't necessarily have to be that way for rx3715 May 27 07:44:01 i am guessing that might change with thier adoption of git ... May 27 07:44:17 hard to say tho May 27 07:45:09 in your opinions, would it be worthwhile incorporating some of the existing rx3715 patches into the hh.org cvs tree? May 27 07:45:25 speaking of hh kernel devs, we could have asked gremlin[it] May 27 07:45:41 A lot of the stuff for the rx3715 is the same as the the h1940 May 27 07:45:49 being of similar SoC heritage May 27 07:49:56 morning all May 27 07:50:31 What I need say probably is. We are getting some work done on the rx3715 port of familar. Having our work join in with the greater familiar project maybe beneficial to everybody. May 27 07:50:41 How do we do it? May 27 07:51:05 you need to try and get a hold of the familiar crew May 27 07:51:24 which might be hard to do before 0.8.4 is released May 27 07:52:50 but you can always submit your stuff to .dev in the meantime May 27 07:54:52 Greetings ! May 27 07:55:26 hi Ifaistos May 27 07:58:56 rp: Hi !. May 27 08:00:04 RP: missed OE deveping these past few days.... it's much easier to find/solve bugs than to handle customers :/ May 27 08:00:32 thanks emte, I'll lurk in the familiar channel for a while. main rx3715 kernel hacker wants to wait a little longer. May 27 08:00:40 Ifaistos: I've also found myself busy with other things :-/ May 27 08:01:09 myopiate, that means your working towards familiar 0.9 i guess May 27 08:01:41 RP: Got to make a living also :) May 27 08:01:54 btw has anyone worked with Mono ? May 27 08:02:07 emte, I'll have a look at the release schedule for familiar, and see what everyone else says. May 27 08:04:09 03rpurdie 07org.oe.dev * r39086741... 10/packages/qemu/ (4 files in 2 dirs): qemu cvs: Remove now unneeded docs patch and add a patch to fix an SDL mouse bug and enable ARM PL110 RGB colour handling May 27 08:04:37 myopiate: yes, merging your code into handhelds.org cvs would probably be ideal. As well as the h1940, you can probably share at least some code with the h2200 and h5x00. May 27 08:04:44 morning May 27 08:05:23 pb_ it has just come to my attention that the h1940 guy wants to merg into vanilla kernel May 27 08:06:04 Anyone here familiar with qemu-native build ? "strip : file not found" message : cannot get rid of that. May 27 08:08:11 uv1: Which version of qemu? May 27 08:08:21 uv1: The cvs version has a patch for that May 27 08:09:06 RP: 0.7.0 May 27 08:09:38 uv1: You could try the cvs configure patch May 27 08:09:54 RP: Thanks, will try ... May 27 08:30:45 Sometimes when i am building images, the process keeps hanging on 'do_rootfs' . I have runned it in debug mode, and it seems like it hanging when it's doing the Packages files. How can i fix this.? May 27 08:31:16 packages.gz, Packages.filesystems and Packages May 27 08:32:02 What do you mean by "doing" the files? May 27 08:32:32 I know it's creating the files, because i moved the old ones, and it created the new ones. But it refuse to go further. ( making the packages files) May 27 08:32:40 If ipkg-make-index is hanging, I guess you should use strace to find out what's up. May 27 08:33:05 * pb_ heading off to the frozen north now May 27 08:33:06 later all May 27 08:33:42 how do i run do a 'strace'? May 27 08:35:09 * Philippe is back (gone 36:03:52) May 27 08:40:11 03koen 07org.oe.dev * rcc68e6a6... 10/packages/monotone/files/configure.ac-no-sync-with-stdio-0.26.patch: monotone: add missing patch May 27 08:49:55 Does anyone get this error message when parsing bb files ? http://oe.pastebin.com/740831 May 27 08:53:34 morning May 27 08:54:46 I have similar error with mesa, klibc, gnome-themes, ublibc-initial, uclibc_svn, uclibc and a DEPEND error on lsof :-/ May 27 08:56:41 03rpurdie 07org.oe.dev * rcf975cbb... 10/classes/ (manifest.bbclass base.bbclass): base.bbclass: Move manifest code into manifest.bbclass and fix a bug in it (import bb.manifest) May 27 08:57:06 cyrilRomain: That message probably means that TARGET_CPPFLAGS isn't set and your setup isn't correctly configured somehow May 27 08:57:47 RP: ah... I changed linux to linux-gnueabi in TARGET_OS ... May 27 08:59:05 RP: hmm... its something else (in my conf i guess) May 27 09:02:23 RP: I got it (it was my '-isystem' line comments in bitbake.conf), thx ! May 27 09:44:35 03rpurdie 07org.oe.dev * r5f5194d5... 10/ (53 files in 6 dirs): Clean up IPKG_ARCHS as it was being used inconsistently, often with duplication. Create IPKG_EXTRA_ARCHS to add to the standard variable. Update documentation.conf with the details. May 27 10:10:55 hm is mini_fo an option for some devices? May 27 10:10:56 http://www.denx.de/wiki/Know/MiniFOFAQ May 27 10:24:16 03lenehan 07org.oe.dev * rd200e691... 10/packages/pmacct/pmacct_0.7.9.bb: May 27 10:24:16 pmacct 0.7.9: Don't look for pcap headers in /usr/include, look for them in May 27 10:24:16 the staging area. Closes bug #712. May 27 10:52:33 +IPKG_EXTRA_ARCHS = "armv5te openmn" May 27 10:52:38 morning May 27 10:52:52 hi hrw May 27 10:53:01 ops.. pasted wrongly May 27 10:53:48 conf/machine/geodegx.conf: +IPKG_EXTRA_ARCHS = "geodegx" is useless May 27 10:53:55 or rather not needed May 27 10:54:22 hrw: right. I probably missed that one :) May 27 10:55:11 1910 has armv5te but 1930/1940 has also armv4/v4t/5e - did not they have same cpu? May 27 10:56:23 Z clamshells does not support armv4t? May 27 10:57:26 RP: I like that changeset May 27 10:59:03 hrw: clamshells can support armv4t - there are probably a few archs we can add to the Zaurus May 27 10:59:23 hrw: I've been meaning to clean it up for a while. Feel free to improve on it :) May 27 11:00:11 RP: sure - when find time May 27 11:00:41 now time to boot 2.6 on collie ;) May 27 11:00:51 hrw: ah, time. I know this problem... May 27 11:01:17 collie/2.6 or beer with friend - what will be faster... May 27 11:02:11 and need time to create better routing on my wrt54 May 27 11:02:31 some priorities add etc.. May 27 11:05:14 hu hrw May 27 11:05:17 hu rp May 27 11:05:27 hu woglinde May 27 11:08:32 unlimited gprs access + ipaq + gprs pcmcia card == raports from road.. May 27 11:09:07 hi woglinde May 27 11:12:09 how is sysrq pn collie 2.6? May 27 11:12:10 hey people May 27 11:12:24 hi mithro May 27 11:12:32 hrw: I'm not sure which key it maps to... May 27 11:12:38 anyone know something like http://discomatic.sourceforge.net/ which is more up to date? May 27 11:13:40 mithro: cd audio ripper? May 27 11:15:10 no May 27 11:15:23 program which allows me to queue up discs to be burnt May 27 11:15:39 i've been looking everywhere and can't find anything May 27 11:17:45 hmm, why am i registerd twice ? May 27 11:30:16 hi zecke May 27 11:30:31 hey zecke May 27 11:30:41 hey May 27 11:30:46 hi zecke May 27 11:31:13 running emerge --regen is no fun on the Shark :} May 27 11:31:16 hi zecke May 27 11:31:36 zecke: your shark does not run OE distro yet? May 27 11:31:44 hrw: yet May 27 11:33:07 2.5G in DL_DIR... May 27 11:34:38 moin zecke May 27 11:35:31 hrw: and i wanted to play with emerge... May 27 11:38:05 koen : I now have Office for OSX as well :( May 27 11:40:59 03rpurdie 07org.oe.dev * r66dc5a29... 10/packages/libtool/ (3 files in 2 dirs): libtool-native/cross: Extend the tag patch to libttdl fixing build failures in jpeg when the compiler tuning options change (from poky) May 27 11:41:03 03rpurdie 07org.oe.dev * r6af43359... 10/packages/xournal/ (files/no-printing.diff xournal_0.2.bb): xournal 0.2: Add patch to remove libgnomeprintui dependency (from poky) May 27 11:41:07 03rpurdie 07org.oe.dev * r9d5b4ae2... 10/packages/leafpad/ (leafpad_0.8.4.bb leafpad_0.8.9.bb): leafpad 0.8.4 -> 0.8.9 and fix DEPENDS (from poky) May 27 11:41:11 03rpurdie 07org.oe.dev * r15f6c085... 10/packages/kf/kf_0.5.4.1.bb: kf: Fix pkgconfig related build issue (from poky) May 27 11:41:15 03rpurdie 07org.oe.dev * r8cc2a41c... 10/packages/js/js_1.5.bb: js: Set soname correctly so OE's shlibs functionality can correctly detect the library. In turn this means js is correctly added to the RDEPENDS of web (from poky) May 27 11:41:20 03rpurdie 07org.oe.dev * rb4a401df... 10/packages/db/ (db-native_4.2.52.bb db_4.2.52.bb): db: Add 4.2.52 (and a native version). This is known to work better than 4.3 with eds (from poky) May 27 11:41:24 03rpurdie 07org.oe.dev * r63da6f40... 10/packages/libglade/ (libglade-2.5.1/no-deprecation.patch libglade_2.5.1.bb): libglade 2.5.1: Enable depreciated functions (from poky) May 27 11:41:28 03rpurdie 07org.oe.dev * rcc79274e... 10/packages/udev/ (12 files in 6 dirs): udev: Add /dev/ram to the automount blacklists for all devices. Enable 089 as the default version. May 27 11:41:33 03rpurdie 07org.oe.dev * r54778fb3... 10/packages/linux/ (linux-openzaurus_2.6.16+git.bb linux-openzaurus_2.6.16.bb): linux-openzaurus: Add a hack to the integrator to enable RGB so qemuarm runs in RGB instead of BGR by default (from poky) May 27 11:41:37 03rpurdie 07org.oe.dev * re6c055d4... 10/packages/gcc/gcc_4.0.2.bb: gcc 4.0.2: Fix DEPENDS (from poky) May 27 11:41:41 03rpurdie 07org.oe.dev * re978f145... 10/packages/avahi/avahi_0.6.10.bb: avahi 0.6.10: Only run the postinst on the target device (from poky) May 27 11:41:45 03rpurdie 07org.oe.dev * r707934bd... 10/conf/machine/nokia770.conf: nokia770.conf: Set tslib PREFERRED_PROVIDER and enable machine specific tuning flags May 27 11:41:49 03rpurdie 07org.oe.dev * r6e7c8ef1... 10/conf/machine/qemuarm.conf: qemuarm.conf: Default to tar.bz2 images May 27 11:41:54 03rpurdie 07org.oe.dev * r1b1f8f5f... 10/conf/machine/ (geodegx.conf h1910.conf include/zaurus-clamshell.conf): Further IPKG_EXTRA_ARCHS cleanup May 27 11:42:07 RP is on fire ! May 27 11:43:07 RP: RGB? triggered too many matchbox errors? May 27 11:43:55 zecke: poky has matchbox patches that fix those BGR errors but for various reasons we'd prefer an RGB emulation environment... May 27 11:44:12 zecke: It also means opie might manage... May 27 11:44:33 RP: that is not challenging anymore :} May 27 11:45:00 zecke: Feel free not to apply that patch ;-) (its only a one liner) May 27 11:45:27 zecke: I've also added a patch to qemu to enable it to emulate RGB correctly as the code wasn't present to emulate that part of the hardware May 27 11:49:28 when building gpe and detect stylus. Should i use libx11 or diet-x11? May 27 11:49:52 goxboxlive_: for GPE both will work May 27 11:50:08 goxboxlive_: if you want E17 you need the full blown X11 May 27 11:50:20 What is E17? May 27 11:50:34 goxboxlive_: Enligthenment (once again spelled wrong) May 27 11:50:53 zecke: Ok thx May 27 11:51:06 goxboxlive_: http://www.enlightenment.org/ May 27 11:51:57 goxboxlive_: http://www.enlightenment.org/Enlightenment/DR17/ May 27 11:53:41 Is that an bb file in the OE? So i can make it for my universal? May 27 11:54:07 ah LinuxDevices survey is completed May 27 11:54:30 OT: what do you think about 'current release' infos on right side of openzaurus.org? May 27 11:56:13 hrw: errier? May 27 11:56:15 ~errier May 27 11:58:02 03rpurdie 07org.oe.dev * r0b9e4521... 10/packages/libmatchbox/libmatchbox_svn.bb: libmatchbox svn: Add 16bpp fixes patch May 27 11:58:06 03rpurdie 07org.oe.dev * rdfc91bed... 10/packages/libmatchbox/libmatchbox.inc: libmatchbox: Disable pango as it doesn't bring much benefit May 27 11:58:07 hrw: Does the release read like this: 3.5.4.1-rc4 is not available for collie and tosa? May 27 11:58:49 it seems like autofoo.patch isent available in the diet-x11 May 27 11:59:05 goxboxlive_: that would be new :( May 27 12:00:16 can i just remove the line in the diet-x11 bitbake file? Will it be working after building? May 27 12:01:01 goxboxlive_: lol, you love randomly deleting stuff, don't you? May 27 12:01:08 zecke: yes May 27 12:01:45 zecke: switched to terrier. ans 3541 is for clamshells May 27 12:01:55 zecke: :-) Well that was the problem with the obex-ftp file. May 27 12:01:57 cu May 27 12:02:19 goxboxlive_: you can use BBMASK = " *diet-x11*" or such ;) May 27 12:02:40 goxboxlive_: monotone ls missing | grep autofoo.patch May 27 12:03:05 ok May 27 12:31:57 morning May 27 12:33:03 CoreDump|home: morning May 27 12:53:32 * chouimat is back. May 27 12:53:42 morning May 27 12:53:53 morning chouimat May 27 12:54:28 chouimat: morning May 27 12:57:50 hi chouimat May 27 12:58:12 ok this mplayer-compiled-in-opie-buildroot bug is positively annyoing May 27 12:58:34 Opie buildroot? May 27 12:58:50 coredump what happend? May 27 12:59:37 if you build an opie-image after after that build mplayer, mplayer starts demanding opie libraries. Which sort of stinks if your are now building a GPE image... May 27 12:59:49 s/after after/and after May 27 13:00:21 03coredump 07org.oe.oz354x * rbd011b54... 10/packages/samba/ (5 files in 2 dirs): May 27 13:00:21 samba: samba-sessential: The next step on my way to a fool-proof samba server. May 27 13:00:21 This one starts a wide open p/w-less server with a single share containing a May 27 13:00:21 helpfile explaining how to configure a samba user+password. Once that's done the May 27 13:00:21 _real_ secure samba server is activated grating access to the targets filesystem May 27 13:00:22 via encrypted passwords May 27 13:00:26 03coredump 07org.oe.dev * r2d741a20... 10/packages/samba/ (5 files in 2 dirs): May 27 13:00:28 samba: samba-sessential: The next step on my way to a fool-proof samba server. May 27 13:00:30 This one starts a wide open p/w-less server with a single share containing a May 27 13:00:32 helpfile explaining how to configure a samba user+password. Once that's done the May 27 13:00:34 _real_ secure samba server is activated grating access to the targets filesystem May 27 13:00:36 via encrypted passwords May 27 13:01:09 coredump whats now the trick with windows-xp and samba and no password? May 27 13:01:27 woglinde: the trick is an ugly work-around May 27 13:01:40 urgs May 27 13:01:57 in security = share windows doesn't always send a password =) May 27 13:02:15 but you have to type someone May 27 13:02:18 right? May 27 13:02:19 so I start a server which is totally unprotected with a single read-only share May 27 13:02:37 woglinde: no, windows does not ask for user/pw May 27 13:02:42 hm May 27 13:03:21 and once the user created a user+password for samba, I start samba with another "secure" configuration that requires a password May 27 13:04:08 this is as fool-proof and user-friendly as it gets... May 27 13:05:38 hm May 27 13:06:05 can't expect from a user to know how to configure a samba password, really. May 27 13:06:18 Even _I_ had to look it up ;) May 27 13:06:22 yes May 27 13:07:03 Now they get a nice Help.txt on how to do that without having to enter any passwords at first May 27 13:08:08 And I dare say, a working NFS/Samba server is a bloody nice thing to have on a PDA, FTP / scp suck =) May 27 13:08:45 CoreDump|home: next step. Plan9 Network filesystem? May 27 13:08:57 zecke: heh May 27 13:09:02 coredump I like scp May 27 13:09:15 but I know what I am doing most of the time May 27 13:09:28 woglinde: scp is nice for copying files, but I really like to "browse" my Z May 27 13:09:49 CoreDump|home: fish://? May 27 13:09:53 with the crappy windows exlporer? May 27 13:10:00 zecke: doesn't work for some reason... May 27 13:10:15 woglinde: I don't run windows. With Nautilus =) May 27 13:10:38 uh nautilus the big fat ugly think from gnome? *g* May 27 13:14:10 hi all May 27 13:14:14 I like it =) It is reasonably fast and feature-rich May 27 13:14:27 morning koen May 27 13:14:52 CoreDump|home: nautilus OOMed on thumbnailing the world-depend.svg May 27 13:14:54 hey woglinde May 27 13:15:04 koen *g* May 27 13:15:22 koen: =D May 27 13:15:29 I haven't tried ksvg yet May 27 13:15:45 lol May 27 13:15:52 today next try to get a pushchair May 27 13:16:13 RP: nice work! May 27 13:16:37 http://hentges.net/tmp/screenshots/Zaurus/Akita/Misc/samba-essential/ May 27 13:16:59 That's what users expect to be able to on a modern PDA I guess May 27 13:17:26 woglinde: pushchair? is that the name? May 27 13:18:11 coredump did you mention about workgroup May 27 13:18:30 that can be a problem May 27 13:18:50 because silly window can only be part of one May 27 13:18:50 woglinde: even windows doesn't give a damn about the workgroup. The Zaurus is going to show up in the workgroup "OPENZAURUS" May 27 13:19:12 but you can browse them all ;) And use shares from any May 27 13:19:19 yes May 27 13:19:32 but that most of the users dont know May 27 13:19:39 hm May 27 13:19:44 or do May 27 13:19:47 right May 27 13:19:50 they see it when they browse the "entire network" May 27 13:19:53 yep May 27 13:19:54 sorry May 27 13:19:57 ;) May 27 13:20:15 thought about the ip way May 27 13:20:17 I had a good deep look at windows smb uglyness yesterday.... May 27 13:20:53 Could anyone try and build parted and tell me if it worked May 27 13:21:07 okay May 27 13:21:12 bye for now May 27 13:21:44 Ken|JLime: will try May 27 13:21:50 bye woglinde May 27 13:22:10 CoreDump|home: thx May 27 13:23:22 CoreDump|home: could you put packages/samba/files/smb-essential-inactive.conf in files/openzaurus? May 27 13:23:42 CoreDump|home: I don't think !OZ distros will like 'workgroup = OPENZAURUS' in their config files May 27 13:24:03 koen: it could print "ownede by OpenZaurus" May 27 13:24:05 owned May 27 13:24:09 'some people' could start reverting.... May 27 13:24:13 * koen makes no further comments May 27 13:24:15 koen: psst May 27 13:24:30 koen sure May 27 13:24:52 well, we could just say workgroup = OpenEmbeddeded =D May 27 13:25:25 yeah, or echo $DISTRO_NAME into it May 27 13:25:30 indeed May 27 13:25:35 I'll do that May 27 13:26:40 Ken|JLime: NOTE: package parted-1.6.23: completed <- that was on .oz354x May 27 13:26:52 hmmm May 27 13:27:07 Yet another weird bug then May 27 13:27:38 maybe a missing dependency? My /tmp is rather large May 27 13:28:05 CoreDump|home: will create a bugreport about it, getting some undeclared stuff. May 27 13:29:48 for the records, Enlightenment doesn't need full-x11. May 27 13:29:51 * mickey|zzZZzz goes shopping May 27 13:29:52 later May 27 13:29:58 mickey|zzZZzz: since when? May 27 13:30:03 since ever May 27 13:30:19 i had build failures due diet-x11 May 27 13:30:35 ya. see bugtracker entry. the people who upgraded x11 removed my patch that stops diet-x11 from lying about its features, that's all May 27 13:30:44 i added patch again May 27 13:30:46 ah okay May 27 13:30:47 volia, it builds May 27 13:30:49 l8er May 27 13:30:50 :) May 27 13:35:01 http://www.yankodesign.com/images/image.php?img=Design_News/2006/May_2/transparent_toaster.jpg&osCsid=4a63d79ec5c878c08947377c853f0cb3 May 27 13:35:03 cool May 27 13:36:18 hehe May 27 13:37:08 Ken|JLime: Remove ac_cv_header_getopt_h from the site file May 27 13:37:56 now you will always remove the toast in the right time ;) May 27 13:40:55 v8jlene: fixes parted? May 27 13:41:07 Ken|JLime: Yes. May 27 13:41:14 v8jlene: :D goodie thx May 27 13:41:44 Ken|JLime: Looks like parted bug, only includes that struct def if it's using getopt.... May 27 13:42:14 Ken|JLime: They are listed in the site file as being for minicom, but I've tested minicom as well and it doesn't need them. Only sh3 and sh4 site files have them... May 27 13:45:55 Ken|JLime: The three entries for minicom can be removed from the site file, the current version of minicom doesn't use any of them. Let me know if it works for you and I'll push the change. May 27 13:47:31 03coredump 07org.oe.oz354x * r0d94da18... 10/packages/samba/ (3 files in 2 dirs): samba: samba-essential: Set workgroup to DISTRO May 27 13:47:35 03coredump 07org.oe.dev * raedc3275... 10/packages/samba/ (3 files in 2 dirs): samba: samba-essential: Set workgroup to DISTRO May 27 13:52:31 03philippe 07org.oe.dev * r126ec629... 10/packages/gtk-webcore/ (files/fix-loading-error-report.patch osb-nrcit_20050430.bb): May 27 13:52:31 osb-nrcit : adding patch that fixes error reporting May 27 13:52:31 * packages/gtk-webcore/files/fix-loading-error-report.patch : patch that fixes the error reporting May 27 13:52:31 * packages/gtk-webcore/osb-nrcit_20050430.bb : added patch usage + bumped up PR May 27 13:57:20 v8jlene: worked, thx May 27 13:58:34 * koen merges May 27 14:03:27 03lenehan 07org.oe.dev * r75c68da9... 10/site/ (sh3-linux sh4-linux): May 27 14:03:27 site/sh: Fix parted for sh3/sh4 by removing the minicom entries. The minicom May 27 14:03:27 entries do not exist in other site files and are not used by the current May 27 14:03:27 version of minicom. May 27 14:10:17 I tried it in #openzaurus but there was no answer yet. Maybe its more an #oe problem? ince wednesday I try to repair my openzaurus installation on the C1000. First I tried it with OZ 3.5.3 (hentges). This doesn't work so I switched to 3.5.4.1 from openzaurus yesterday evening - it doesn't work either. I flash with a CF-card, "--------Zaurus CF-Updater --- " shows up, then "MODEL:SL-C1000" then "kernel", "0% 100%" and "zImage.bin:1156968bytes". The May 27 14:11:05 odo: hmm May 27 14:11:29 zecke: so you say ... :-) May 27 14:16:27 odo: make sure the FS on that CF card isn't damaged and verify the md5 sums of zImage and initrd.bin May 27 14:30:08 koen: thanks. Hopefuly a few useful fixes in there and nothing too controvershal :) May 27 14:31:04 let's hope so :) May 27 14:31:31 I'll have to submit the result of my SoC for merging into .dev after the summer May 27 14:31:45 that's going to be a lot of 'fun' May 27 14:32:28 Lets just hope you get it right :) May 27 14:33:06 I know I'm not an official mentor but if you want to discuss the implementation, I'm happy to assist if I can May 27 14:33:37 Its probably better to communicate what you're doing as people may have ideas which can help May 27 14:34:21 yeah May 27 14:34:42 that's why I have it in a seperate branch which I sync regularly May 27 14:35:00 to avoid a novell like code drop May 27 14:35:15 masterhacker zecke is watching me like a hawk :) May 27 14:35:22 I've noticed it - I'll have to start pulling that :) May 27 14:41:44 koen: lol May 27 14:41:53 koen: we need to continue to talk :) May 27 14:42:12 koen: the most easy way: Each package emits exactly one pstage package May 27 14:42:20 koen: plan b, get the data from BitBake May 27 14:42:30 ~lart thunderbird May 27 14:42:30 * ibot chops thunderbird in half with a free AOL CD May 27 14:42:52 The plan used to be to install base package and the -dev package into staging? May 27 14:43:05 that's the ultimate goal May 27 14:43:31 right, good :) May 27 14:44:32 I've been playing with the on device development environment and its working quite well. I've not stressed it to much yet but the basics work May 27 14:44:51 I should release a qemu arm image... May 27 14:45:57 ah cool May 27 14:46:09 you eliminated 3 points from my todo list May 27 14:46:11 RP: libtool and pkg-config need special handling May 27 14:46:11 RP: do you solved the persistence problem (initrd) May 27 14:46:11 did you solve... May 27 14:46:11 man what is wrong with me May 27 14:46:55 zecke: pkgconfig just worked in my test... May 27 14:47:24 RP: where did you install the packages to? /? May 27 14:47:36 zecke: Yes May 27 14:47:56 zecke: I built them into the image... May 27 14:47:56 we should get a few sponsored SBCs into a colo-site for OE May 27 14:48:07 SBCs? May 27 14:48:12 colo? May 27 14:48:19 RP: ah well that is working ;) May 27 14:48:22 single board computer and colocation May 27 14:48:34 zecke: dev-cluster with fat pipe May 27 14:48:45 zecke: It was nice to have it in poky though ;-) May 27 14:48:47 RP: libtool (.la) has installed=yes (which is true) and pkg-config :) May 27 14:48:56 koen: what do you put in the pipe? May 27 14:49:01 zecke: bits May 27 14:49:09 bits of what? May 27 14:49:33 RP: have you seen the Opie and GPE build dependency graph from yesterday? May 27 14:49:40 zecke: no May 27 14:50:00 RP: http://ewi546.ewi.utwente.nl/OE/gpe-build.png May 27 14:50:00 RP: http://ewi546.ewi.utwente.nl/OE/opie-build.png May 27 14:50:09 RP: http://ewi546.ewi.utwente.nl/OE/matchbox-build.png May 27 14:50:18 jesus May 27 14:50:30 zecke: how do you generate these? May 27 14:50:44 RP: and I wonder why matchbox-panel needs virtual/kernel May 27 14:51:06 zecke: I'll have to make the parallel build mode for bitbake :} May 27 14:51:09 cbrake: hehe, bitbake -g matchbox-panel; dot -Tpng -o matchbox.png depends.dot May 27 14:51:15 probably some applet that depends on a module... May 27 14:51:29 battery applet probably May 27 14:51:41 cbrake: and add -I virtual/libc -I quilt-native to omit deps May 27 14:52:25 zecke: nice. May 27 14:52:38 zecke: -g must be a new bitbake option? May 27 14:53:00 yes May 27 14:53:11 a bitbake 1.5 feature May 27 14:53:46 cbrake: it will generate depends.dot, rdepends.dot and alldepends.dot May 27 14:53:46 darn, my initrd.bin for Akita is getting tight May 27 14:53:54 cbrake: http://svn.berlios.de/viewcvs/bitbake?rev=517&view=rev May 27 14:53:56 zecke: excellent -- I've often thought a dependency output would be a nice feature. May 27 14:54:20 koen: :-) May 27 14:54:35 zecke: I like it :) May 27 14:55:29 lol May 27 14:56:01 "You are running 2.6.16-3.5.4.1-rc4" I expect 2.4.18-crappix May 27 14:59:23 okay enough haskell for today (more tomorrow) back to hacking May 27 14:59:40 wonderful my NSLu2 now has hugs installed May 27 14:59:55 03cbrake 07org.oe.dev * r7ec369fd... 10/packages/cups/ (cups-1.1.23/strftime_fix.patch cups_1.1.23.bb): May 27 14:59:55 cups 1.1.23: fix strftime seg fault issue May 27 14:59:55 - when strftime is called w/ a NULL format string, it is segfaulting May 27 14:59:55 in glibc May 27 14:59:55 - integrate code from latest cups svn to fix May 27 15:00:57 * cbrake is trying to get cups working on his NSLU2 ... May 27 15:01:09 re May 27 15:01:17 wb hrw May 27 15:01:21 hrw: hello May 27 15:01:22 I have to buy some dvdr - out of space on hdd May 27 15:01:57 120G hdd is so small.. ;D May 27 15:01:58 grrrrr May 27 15:02:19 sound stopped working on SL-Cxx00, probably a missing module... May 27 15:03:46 ah, just the usual "no auto-depmod" bug... May 27 15:04:15 that ipkg bug? May 27 15:04:31 dunno who is to blame but it gets annoying ;) May 27 15:05:33 yeah, the intercept code is a malfunctioning a bit May 27 15:05:47 I do not like ipkg at all.. May 27 15:06:11 hehe May 27 15:06:27 hmm the volume is way too low now May 27 15:06:36 its slow even on my desktop May 27 15:06:43 its slow even on ewi May 27 15:06:55 even the tho the PCM level is maxed May 27 15:07:28 hrw: yes May 27 15:09:31 ~curse cardctl May 27 15:09:37 May the fleas of a thousand camels infest your most sensitive regions, cardctl ! May 27 15:12:48 hrw: nice, now udev and cardct fight over wlan heh. sometimes it's eth0, sometimes it's wlan0 May 27 15:13:32 mickey|shopping: ping, do we have dante mirrors? May 27 15:13:54 CoreDump|home: cool.................... May 27 15:14:25 let someone destroy 2.6 pcmcia subsystem... May 27 15:16:02 RP: could you take a look at the BitBake backport bug? May 27 15:16:21 RP: I have a solution for the INHERIT with event lister bug May 27 15:17:47 zecke: I had a look at that - it looks reasonable to me although I don't claim to understand that area of code 100% :) May 27 15:20:15 no one does :} May 27 15:20:30 RP: it will fix the tinder reports of the NSLU2 people May 27 15:21:17 Presuming you've tested it and it doesn't cause bitbake to explode, I'd say apply it :) May 27 15:21:59 okay covered my ass ;) May 27 15:22:13 zecke: ;-) May 27 15:22:51 hrw: Have you any objection to changing the the logos patch to add all the logos in one patch and select the right one in the defconfigs? May 27 15:23:20 RP: no obkections May 27 15:23:26 RP: no objections May 27 15:23:55 I have to build kernel for collie which will boot and then add logo into it.. May 27 15:23:59 hrw: The OZ logo is going to cause me problems in poky and has just prevented me syncing things. I'll rework it to be a distro based setting... May 27 15:24:08 hrw: collie doesn't boot atm? May 27 15:24:26 * RP wonders what magic pavel has in his tree... May 27 15:24:57 RP: does not May 27 15:25:30 hrw: No idea why? May 27 15:25:44 RP: doesn't pavel have some git thingie? May 27 15:25:56 koen: correct, he does May 27 15:26:04 koen: Anything useful in it should be in mainline though May 27 15:26:08 RP: is poky still (ab)using OZ feeds? May 27 15:26:23 lol May 27 15:26:45 koen: feeds can not be abused May 27 15:26:47 koen: It might reference them although that was never intentional and as far as I know, nobody has ever used that May 27 15:26:48 RP: nothing appear in serial and I never tried 2.6 on collie before May 27 15:27:16 hrw: Neither have I... May 27 15:27:27 RP: just checking who uses OZ feeds in case I restructure ewi's disk layout May 27 15:28:23 koen: Ah, ok. I'd not worry about poky in that case :) May 27 15:28:36 man, the uni mailserver is breaking again May 27 15:29:27 Ok I checked ist twice, the md5s are ok. The JFSS2 error remains, the modprobe error stays and hter are some "rm: unable to stat `/tmp/*ยด: Input/output error" messages. The teh booting stops. May 27 15:30:02 odo: In that case I'd recommend a NAND restore May 27 15:30:13 OT: LOL at people who resized their mtd partitions but don't know how to change it back May 27 15:30:36 thank god OZ doesn't recommend such stuff May 27 15:30:56 koen: ot doesn't but it should be more forgiving =) May 27 15:31:13 I thought about that myself ... so in the case nobody has another good idea ...? May 27 15:31:43 odo: just do it. It is pretty safe on SL-Cxx00 May 27 15:32:06 hmmm May 27 15:32:13 I should really teach OE to generate .debs May 27 15:32:32 and hand in my bachelor thesis next week May 27 15:32:37 I only have the CACKO-ROM for NAND-restore yet ... have to do the whole way back to OPIE then ... May 27 15:32:49 .debs are on my todo list as well. Fairly well down it :-/ May 27 15:33:04 RP: hehe, we should publish these lists :} May 27 15:33:42 RP: initial support is fairly easy, just absorb ipkg-build.sh into OE and make it output .debs instead of .ipkgs May 27 15:33:54 ~20 lines of shell May 27 15:34:04 koen: I suspected as much May 27 15:34:14 integrating apt + dpkg is a lot more work May 27 15:34:29 indeed :) May 27 15:34:42 debootstrap? May 27 15:34:47 adding offline mode to dpkg? May 27 15:37:12 Is bitbake 1.3.3 still the minimum version for OE these days? May 27 15:37:26 no May 27 15:37:29 1.4.2 May 27 15:37:48 That's not what sanity.conf says? May 27 15:38:04 :( May 27 15:38:16 I'll fix... May 27 15:38:17 sanity is insane in that case May 27 15:38:34 Just to make shure: http://ewi546.ewi.utwente.nl/tmp/hrw/3.5.4.1-test/rc4/images/akita/opie-image-3.5.4.1-rc4-akita.rootfs.img has to be renamed to initrd.bin? May 27 15:39:08 odo: yes May 27 15:39:59 zecke: I'm putting Detlef's docu into monotone now May 27 15:40:09 (after it finishes pulling) May 27 15:40:10 koen: good :) May 27 15:40:26 koen: do you mind if I put your GSoC project in the wiki? May 27 15:40:35 zecke: not at all May 27 15:41:07 cu May 27 15:41:17 * hrw goes shpping.. some dvdr media May 27 15:41:23 hrw|gone: enjoy May 27 15:41:32 /dev/hda6 88G 88G 217M 100% /home May 27 15:42:08 =) May 27 15:42:49 koen: I think I do not like the keyboard of a macbook May 27 15:43:34 zecke: I hate the glary screen May 27 15:43:57 03rpurdie 07org.oe.dev * rc74480da... 10/conf/sanity.conf: sanity.conf: Update bitbake minimum version requirement to 1.4.2 May 27 15:46:49 03koen 07org.oe.documentation * r816188d4... 10/usermanual/embworld-oe.dbk: usermanual/embworld-oe.dbk: Add Detlef Vollmann's docu May 27 15:48:10 zecke: I cc:ed you on the mail to detlef about the copyrights May 27 15:50:47 koen: did he answer? May 27 15:51:17 not yet May 27 15:53:55 monotone: 1.1 M | 165.8 k | 759 | 184 | 0 May 27 15:54:17 * koen starts making dinner while monotone is cruching May 27 15:54:22 crunching* May 27 15:54:39 still at mtn 0.25? May 27 16:02:06 i'm being kind of lazy, i could install mtn 0.25 i guess May 27 16:02:13 oh great, /opt/Qtpalmtop in my GPE image. I hate this idiotic libSDL behaviour May 27 16:02:19 crouching montone hidden bitbake May 27 16:04:02 03freyther 07org.oe.dev * r683e48c1... 10/classes/base.bbclass: May 27 16:04:02 openembedded/classes/base.bbclass: Add CTAN mirror May 27 16:04:02 Add CTAN mirror. A German/Berlin, Japan and an May 27 16:04:02 Australian one. May 27 16:06:07 wasn't libSDL supposed to be binary compatible bewteen opie and X? May 27 16:06:26 yes, as long as you don't link SDLmain May 27 16:06:49 well, why would it depend on opiecore2 then grr May 27 16:08:07 CoreDump|home: hehe May 27 16:08:19 CoreDump|home: from the outside both libraries should be the same May 27 16:08:27 internally they link to... May 27 16:08:53 not funny, the SDL and maplayer dependency issues makes it impossible to compile GPE and opie in the same tmp May 27 16:08:55 CoreDump|home: hmm we should care more for SDL and look at the shlibs deps May 27 16:09:21 CoreDump|home: due the shlib deps code... May 27 16:09:44 well mplayer is actually needing opie libs... May 27 16:09:50 not just via DEPENDS May 27 16:10:01 it won't run in GPE ;) May 27 16:10:58 CoreDump|home: objdump and show me the headers of mplayer (SO_NEEDED) May 27 16:14:49 zecke: http://openembedded.pastebin.com/741333 May 27 16:15:00 i hope "-x" was right... May 27 16:15:24 oh that RPATH sucks May 27 16:15:40 has mplayer been fixed when libSDL became binary compatible? May 27 16:15:55 CoreDump|home: good mplayer does not link to Qt or libopie May 27 16:16:02 CoreDump|home: you simply have the wrong SDL installed :} May 27 16:16:27 how can there be a wrong SDL when they are compatible *sigh* May 27 16:16:39 CoreDump|home: hehe May 27 16:16:50 I'll rebuild the SDL stuff and try again May 27 16:16:55 CoreDump|home: they expose the same ABI/API to clients (applications using them) May 27 16:17:09 CoreDump|home: they are different internal, but in a way not noticed by the outside May 27 16:17:20 CoreDump|home: you could exchange libsdl opie with libsdl gpe May 27 16:19:22 IMO there should be a libsdl-x11 and a libsdl-qpe so the _image_ gets the correct one installed May 27 16:21:09 CoreDump|home: this is the possible shlib deps bug May 27 16:21:21 CoreDump|home: when a package links to libSDL it should depen on either of them May 27 16:21:30 CoreDump|home: currently it depends on one (the last built one) May 27 16:22:01 zecke: but libsdl-x11 and libsdl-qpe.bb always produce libsdl.ipk. There is no way to tell which one is in deploy May 27 16:22:16 CoreDump|home: that is the bug ;) May 27 16:23:09 zecke: thanks, installing the correct libSDL unb0rks mplayer May 27 16:59:34 zecke: detlef ok'd to include his stuff into usermanual.xml May 27 17:00:22 okay May 27 17:02:17 zecke: could you respond to his email about the planning? May 27 17:02:22 you being or doc-meister and all May 27 17:03:13 koen: hehe, what makes 2.5 better? May 27 17:03:20 * koen has no idea May 27 17:03:53 wow Perl Jam used CC for their video May 27 17:04:02 yeah May 27 17:04:14 but flash crashes firefox on my ubuntu May 27 17:05:02 koen: I would have assumed the vote Republicans by now May 27 17:08:25 zecke: You appear to have broken the whitespace in tinderbox.bbclass May 27 17:09:16 03rwhitby 07org.oe.dev * rabf00cce... 10/packages/meta/slugos-packages.bb: slugos-packages: Added asterisk (compiles, is untested) May 27 17:09:22 03nail 07org.oe.dev * r56ee470f... 10/packages/linux/ (ixp4xx-kernel/2.6.16/defconfig ixp4xx-kernel_2.6.16.bb): ixp4xx-kernel: Set CONFIG_EVDEV=y. Buzzer won't work when it's a module May 27 17:09:55 damn May 27 17:10:34 RP: will adjust... I have copy and pasted the patch from a machine behind a fascist firewall May 27 17:12:36 zecke: I think there's just one line broken but I can't easily push a correction atm :-/ May 27 17:16:23 koen: updating to version 2.5 makes sense right? May 27 17:17:16 RP: do you know the line? May 27 17:29:30 http://docbook.sourceforge.net/release/xsl/current/html/refentry.xsl:1: parser error : Document is empty May 27 17:29:33 hmm May 27 17:29:43 zecke: When the do_rootfs hangs ( it's hanging when doing the three Packages.gz, Packages.filesystem and packages.) when i am building a opie-image. How can i fix it? I May 27 17:30:01 goxboxlive_: hangs? dunno, fix ipkg? May 27 17:30:11 yes May 27 17:31:40 goxboxlive_: find out where ipkg hangs May 27 17:31:59 Ok, i'll try to build with -DDDD May 27 17:33:53 goxboxlive_: hmm May 27 17:33:56 ps xua | grep ipk May 27 17:34:05 gdb ipkg-cl `pidof ipkg-cl` May 27 17:34:06 bt May 27 17:34:12 otherwise it can't hang... May 27 17:34:15 oh May 27 17:34:26 zecke: 296 May 27 17:34:31 wait you mean it hangs on Packages.gz? and what is packages.filesystem? May 27 17:35:03 goxboxlive_: BTW: it is getting harder and harder to help you. And this is all due you deleting files at random May 27 17:35:20 goxboxlive_: really stop that! otherwise I will stop listening May 27 17:36:03 zecke: I am not deleting files in the fam083 branch. It was just in the dev branch. But i can se your point. May 27 17:38:44 goxboxlive_: if it hangs, do ps xua and see if ipkg is running May 27 17:39:15 zecke: And actually, i have never ever deleted files, just been in the *bb files and deleted som lines. I know that's not the way to go, but what a heck, it might work. And the only thing i have added is support for the HTC Universal since it not support yet. May 27 17:40:42 I havent commedthat far yet. It's building with -DDDD option. It takes some time. May 27 17:43:34 zecke: I want to stay away from licensing issues, but moving to 2.5 is fine by me May 27 17:44:18 koen: hehe May 27 17:49:42 hm multiple heads again May 27 17:53:33 http://docbook.sourceforge.net/release/xsl/current/common/pi.xsl:1: parser error : Start tag expected, '<' not found May 27 17:53:33 HTTP/1.1 200 OK May 27 17:53:34 wtf is wrong here :} May 27 17:54:10 zecke: check the base href tag in detlef's docu May 27 17:54:25 that is with the bitbake manual... May 27 17:54:29 hmmm May 27 17:55:05 manually loading this file is working... May 27 17:55:39 zecke: 98 entries in the survey, time to close it? May 27 17:56:15 koen: so how goes the staging :) May 27 17:56:39 mithro: I got a small set of metadata to play with, but haven't done a lot this weekend May 27 17:56:52 koen: send a mail we will close it end of the month May 27 17:57:00 mithro: http://ewi546.ewi.utwente.nl/tmp/viewmtn/branch.psp?branch=org.openembedded.packaged-staging May 27 18:00:06 mail sent May 27 18:00:50 I wonder why fbset doesn't change a thing on Akita May 27 18:01:21 CoreDump|home: you are root? May 27 18:01:25 yeah May 27 18:01:45 I've messed up a VT and can't set the correct mode :\ May 27 18:04:23 welp, I should be off to bed May 27 18:04:24 gnight May 27 18:04:42 n8 mithro May 27 18:04:48 'night mithro May 27 18:05:25 CoreDump|home: how big is samba-essential for arm? May 27 18:06:00 -rw-r--r-- 1 mhentges mhentges 2.3M May 27 15:54 samba-essential_3.0.20-r3_armv5te.ipk May 27 18:06:31 that could probably be lowered by hacking smbd and removing all the uneeded printer and domain stuff May 27 18:07:15 hey Bernardo May 27 18:07:32 hi koen, hi guys May 27 18:07:34 zecke: i saw undercover germans today May 27 18:07:52 zecke: wearing dutch WM soccer hats May 27 18:08:06 undercover germans? May 27 18:08:06 =) May 27 18:08:18 anyio May 27 18:08:20 aargh May 27 18:08:35 koen hehe May 27 18:08:49 anyone wearing bright orange "hup, holland, hup" gear is either dutch or undercover ;) May 27 18:09:28 hm I think I will be for holland again May 27 18:09:45 who is hentges? May 27 18:09:50 mithro: CoreDump|home May 27 18:10:04 ahh May 27 18:10:36 CoreDump|home: why do you maintain a seperate rom? May 27 18:11:03 because I'm not sattisfied with the default ROM in certain points May 27 18:11:33 many of these points were worked out in 3.5.4+, not all of them =) May 27 18:11:48 rom? May 27 18:11:56 koen *g* May 27 18:11:57 that is so 1990's May 27 18:11:57 * CoreDump|home chuckles May 27 18:12:00 oesf talk May 27 18:12:04 ~users May 27 18:12:06 so this is what we get for sharing our unpaid volunteer work with you? complaints and accuses? well done, this clearly supports our motivation to continue working on open source projects. May 27 18:12:14 ~rom May 27 18:12:16 well, rom is Read Only Memory May 27 18:12:24 :D May 27 18:12:31 ok, looks like busybox's fbset can't change anything May 27 18:12:42 now in which packages hides the real thing hmm May 27 18:13:06 coredump are ypu sure that the kernel really allows fbset's? May 27 18:13:19 i tried to think up something cool another solution for ROM but didn't think of any May 27 18:13:20 woglinde: uhh, no clue really May 27 18:13:26 busy box fbset doesn't work May 27 18:13:38 i know this from experiance May 27 18:13:43 mithro: :\ May 27 18:13:55 accoring to goole that is indeed the case May 27 18:13:59 google May 27 18:14:21 i did a patch for it a while back - must have gotten lost somewhere May 27 18:14:32 CoreDump|home: http://ftp.debian.org/debian/pool/main/f/fbset/fbset_2.1.orig.tar.gz May 27 18:14:52 hey EvilDevil May 27 18:14:55 ehm May 27 18:14:58 hey eFfeM May 27 18:15:03 hi koen May 27 18:22:11 I'm pushing, stay away please ;) May 27 18:23:45 okay May 27 18:24:10 zecke bee aware multiple heads May 27 18:24:25 morning May 27 18:24:29 hi xora May 27 18:24:30 multiple heads, schmultiple heads May 27 18:24:33 03freyther 07org.oe.dev * reab2d205... 10/classes/tinderclient.bbclass: May 27 18:24:33 class/tinderclient.bbclass: Change the tabs again May 27 18:24:33 Reported by Richard, fixing the indention level May 27 18:24:33 of this class. May 27 18:24:35 hey mister benidorm May 27 18:25:49 hey koen May 27 18:26:20 I snuck in a new epdfview and gimp while you weren't looking ;) May 27 18:26:37 koen: cool May 27 18:26:47 koen: just syncing now May 27 18:28:54 only 338 revs, koen youve been slacking May 27 18:29:41 yeah, sorry May 27 18:29:58 we added some new devs to cover up our slacking May 27 18:34:45 koen: cool, Im sure when Im back at work on monday I can catch up on all the cool shit May 27 18:36:32 just to check (or even to better understand OE): is it normal that gcc-4.1.x/arm*.patch contains 'gcc-4.0.2' token instead of 'gcc-4.1.x' ? May 27 18:39:25 cyril yes sometime May 27 18:39:54 if the patch changed nothing what has changed in the 4.1 release May 27 18:39:58 woglinde: when building arm gcc-4.1.x with gcc-4.0.2, right ? May 27 18:41:20 s/gcc-4.0.2/cross gcc-4.0.2/ May 27 18:41:52 oh what token you mean? May 27 18:42:02 I thought you meant the name of a patchfile May 27 18:42:15 no, the internal content May 27 18:42:36 could you paste it? May 27 18:43:14 hmm I wonder if remove dropbear is smart May 27 18:43:48 no it wasn't :( May 27 18:43:58 lol May 27 18:45:35 woglinde: I thnk my NSlu2 is now secure! May 27 18:45:40 hehe May 27 18:45:45 netconsole May 27 18:47:50 woglinde: well, for example the content of the gcc-4.1.1/arm-softfloat.patch (http://oe.pastebin.com/741578) contains a patch for gcc-4.0.2 May 27 18:48:31 yo no prob eith that May 27 18:48:45 than this is the same issue in gcc-4.0.2 and gcc-4.1 May 27 18:48:58 and nothing in that area changed the abi May 27 18:49:52 those are just pathnames, patch -p1 will ignore those May 27 18:50:06 koen yepp that too May 27 18:50:22 well thanks for the clarification May 27 18:55:31 hmm, does anyone actively do OE builds for PowerPC? May 27 18:55:42 joshua_: not yet May 27 18:55:55 joshua_: I have a PowerPC board next to me May 27 18:56:07 have three mpc823 next too me May 27 18:56:13 but no elan May 27 18:56:18 woglinde: elan? May 27 18:56:25 E Learning Framework? May 27 18:57:03 woglinde, RPX? May 27 18:57:26 joshua no famous german settopbox dbox2 May 27 18:57:44 I have a pair of RPX Lites here that are bricks until I can either hack libmpc8xx to do the right thing or write my own code... May 27 18:58:29 how much flash does the dbox2 have? May 27 18:58:47 let me think May 27 18:58:51 16 or 32 May 27 18:58:55 mom May 27 19:00:15 oh only 8 mb May 27 19:00:17 32 mb ram May 27 19:00:41 ah May 27 19:00:43 never flashed only worked with nfsroot May 27 19:01:15 this rpx is 16mb/64mb. I had it working with nfsroot until I tried hacking MTD to use jffs2 on it... looks like I got it wrong, since it now has no bootloader... May 27 19:01:49 joshua jtag May 27 19:01:51 that map should've been right, too. May 27 19:01:58 yeah, bdm is supposed to be the right way to do it on these things May 27 19:02:00 hmm May 27 19:03:32 maybe I will build a bdmbox later today May 27 19:04:10 damn, i'm cursed ;-( be it the dev or ozXfamX branch... same problems with running init - kernel can't find it May 27 19:04:42 frg hm huh maybee the same problem as zecke wrong floatingpoint support May 27 19:05:05 how would that be connected? May 27 19:08:10 i run a diff booting_stock_familiar082 booting_my_083_build .... the only difference except different dir entries and frag entries is this: May 27 19:08:10 < +file_offset increments = 0000146B May 27 19:08:10 < loaded file of size = 0x0009C1B0 at location 0xC0008000 May 27 19:08:10 > +file_offset increments = 00002082 May 27 19:08:10 > loaded file of size = 0x0009C1D8 at location 0xC0008000 May 27 19:08:45 other kernel messages identical May 27 19:09:35 and i have no clue where to look at now :/ May 27 19:13:56 Frg: try with the familiar guys May 27 19:14:06 Frg: and preferable try one of the prebuilt familiar images May 27 19:14:15 Frg: they are close to a release May 27 19:15:26 if all fails... :) May 27 19:22:13 muhaha, the real fbset works =) May 27 19:22:40 coredump nice May 27 19:25:57 hmm tricky... May 27 19:26:07 should I add the depends of the RDEPENDS :} May 27 19:27:02 zecke: madness ;) May 27 19:27:19 CoreDump|home: oh in the graph generation May 27 19:28:20 hmm how do I avoid code duplication... May 27 19:28:32 woglinde: BTW: I love infinite lists and partial evaluation May 27 19:28:37 woglinde: partial function binding May 27 19:32:05 morning May 27 19:32:16 I start to like usb 2.0 bandwidth May 27 19:32:28 dvd+r x8 over usb 2.0 May 27 19:32:33 nice May 27 19:32:40 did you get the drive today? May 27 19:33:01 zecke: friend lent me his usb2.0 dvdrw drive May 27 19:33:25 or borrow - I never followed difference of that words May 27 19:33:38 ve to run dict May 27 19:34:25 checked May 27 19:37:44 I forgot it :( May 27 19:37:52 hrw: I prefer firewire though May 27 19:40:02 zecke: I lack fireware cables (has controler onboard) May 27 19:50:39 xora: how was vacations? May 27 19:50:41 dvd 1x = cd 8x, right? May 27 19:50:55 koen: is not x10? May 27 19:50:55 and 1x is 175kB/s iirc May 27 19:51:09 hrw: could be May 27 19:51:22 1x is 150 May 27 19:51:33 so dvd 8x = 64-80 x 150kB/s May 27 19:51:53 that is well below the 480 Mbit/s of usb2 May 27 19:52:11 but above 12Mbit/s of usb1.1 May 27 19:53:19 ofcourse - but those dvds are x8 only May 27 19:53:38 hrw: hot, but fun, Im glad to be in rainy old scotland again May 27 19:54:24 XorA_: thx again for battery cover - c760 is so small with 950mAh battery May 27 19:54:35 hrw: cool it arrived safe then May 27 19:56:24 XorA|gpo: btw, mardy has w100 accel'ed kdrive working(ish) May 27 19:56:35 time to get dinner May 27 20:04:55 koen: I saw the email, that is very cool May 27 20:06:16 XorA|gone: smacznego May 27 20:07:38 hrw|tv: that wierd speccy music guy? May 27 20:08:01 ~bon appetit May 27 20:08:07 bon appetit is probably smacznego. Guten Appetit. Eet Smakelijk. God Appetitt. Buon Appetito. Buen apetito Bom Apetite. buen apetito May 27 20:08:49 'smac' = dutch word for spam May 27 20:09:17 http://www.cynthiavandepeppel.nl/smac.jpg May 27 20:09:32 spam as mielonka? (meat spam) May 27 20:09:43 yes May 27 20:12:03 CoreDump, zecke: After NAND-restore, flashing of CACKO 1.23 (for partitioning), reflashing of OPIE an dOZ it looks like it's working again. Let's see how it looks after removing the opie-pim and installing KO/Pi and KA/Pi, zbedic and qpegps (if that has a chance to run on opie ...). That's the bread and butter. Thnanks for your help! May 27 20:16:08 Have to restore a network config with bootp and tftpbook I last needed one and a half year ago :-( That will force me offline some time. Bye. May 27 20:16:10 oops had some unpushed stuff in my svk tree :} May 27 20:19:24 is it possible to run tasks within tasks? May 27 20:19:52 like do_foo() { blah; do_bar}; May 27 20:23:06 yes, but do_bar wont get its stamps checked or set, as that occurs outside the shell May 27 20:27:24 heh.. my 8years old toshiba dvd drive has problems with just burnt dvd+r... May 27 20:27:39 koen: the graphviz OSX is nice May 27 20:27:50 I suppose dvd drives can only read dvd-r May 27 20:27:54 not dvd+r May 27 20:27:57 koen: bb.build.exec_task May 27 20:28:14 can anybody see a good reason why EINVAL should be undeclared as long as errno.h is included (on compiling the apmd package)? May 27 20:30:38 i am thinking about repartitioning my poodle to increase /dev/mtdblock2... am i stupid? May 27 20:34:40 * cyrilRomain dropbear finally compiled, thanks to the zecke's star i guess :) May 27 20:41:42 I begin wondering if the LCD on Akita is even capable of running in 320x240... May 27 20:43:46 CoreDump|home: sash made a patch for it... May 27 20:43:57 heh May 27 20:44:03 figures May 27 20:48:21 hello May 27 20:48:30 CoreDump|home: under 2.6? May 27 20:48:42 hello hhhcz May 27 20:48:42 i'm trying to build openzaurus 3.5.4.1 May 27 20:48:49 RP: yep May 27 20:48:55 and the build fails on qemu-native May 27 20:48:56 03coredump 07org.oe.oz354x * r8ee07b62... 10/packages/fbset/ (fbset_2.1.bb files/makefile.patch): fbset: Add fbset May 27 20:48:58 CoreDump|home: Is is capable, the kernel doesn't support it atm May 27 20:49:01 03coredump 07org.oe.dev * r56a0604f... 10/packages/fbset/ (fbset_2.1.bb files/makefile.patch): fbset: Add fbset May 27 20:49:14 i wonder why qemu is required May 27 20:49:18 RP: ahh, that explains it, thanks! May 27 20:49:23 hhhcz: you lack gcc 3.x on host? May 27 20:49:33 hhhcz: pregenerated locales req qemu May 27 20:49:38 my host is amd64 May 27 20:49:52 it seems to be why it fails May 27 20:50:01 hhhcz: you have gcc3 on host? May 27 20:50:10 yes May 27 20:50:47 http://pastebin.com/741785 May 27 20:52:04 i don't know aht is that gcc32 thing... May 27 20:52:11 what May 27 20:55:01 * hrw|tv cannot help - x86 only May 27 20:57:40 * chouimat is away: Bedtime is early today May 27 21:02:19 is there a central place where sash hides his patches? May 27 21:02:35 pdaxrom-builder archive May 27 21:03:05 i think i see the problem May 27 21:03:24 in qemu-native_0.7.0.bb May 27 21:03:36 hrw|tv: the sdk thingy? May 27 21:03:40 i thires to find a specific version of gcc May 27 21:03:47 yes May 27 21:03:54 hmm May 27 21:03:58 but in gentto 64, gcc32 is the 32 bit version of gcc May 27 21:04:03 and not gcc 3.2 May 27 21:04:12 what is the dir above ${S} called? May 27 21:04:12 nice, 404 on all downloads May 27 21:04:31 hhhcz: gcc-3.4 or gcc-3.3 is checked first May 27 21:04:37 CoreDump|home: moment... May 27 21:04:38 ie, ${S}/../ ;) May 27 21:05:11 hehe they have a "donato-meter" =D May 27 21:05:23 hrw|tv: i only have gcc and gcc-3.4.6 in my //usr/bin ... May 27 21:05:50 and also gcc32 and gcc64 which are hardlinks to gcc May 27 21:06:00 hhhcz: then do ln -sf gcc-3.4.6 gcc-3.4 May 27 21:06:31 yes, that's what i was going to do :) May 27 21:06:32 hhhcz: oz 3.5.4.1 use pregenerated locales and it need qemu native working which need gcc 3.3 or 3.4 or 3.2 May 27 21:07:26 ok, thanks :) May 27 21:10:55 NAiL: WORKDIR? May 27 21:11:01 hrw|tv: I've found the builder, looks like they are reworking their site... May 27 21:11:12 CoreDump|home: they do that May 27 21:11:25 CoreDump|home: on ewi in tmp/hrw/kernel-atches/pdax/ May 27 21:11:59 'http://ewi546.ewi.utwente.nl/tmp/hrw/kernel-patches/pdax/' May 27 21:12:32 hrw|tv: great, thanks! May 27 21:12:53 0221_linux-2.6.16-spitz-usbcable-events.patch <- mehh they already have that May 27 21:13:34 anyone willing to get a couple of AMD X2's for the OE crew :} May 27 21:16:43 someone has broken collie? working mainboard is needed May 27 21:25:08 zecke: thanks, figured it out after a while May 27 21:25:19 now I'm having problems with the linker :-P May 27 21:28:35 hmm emerge --sync runs for the last couple of hours (12h) May 27 21:29:12 zecke: something is going terribly wrong, i suppose May 27 21:29:29 or the link is damn slow... May 27 21:29:37 zecke: hanging in cache rebuild ? but 12h ... 8-/ May 27 21:30:16 ixoth: rsync is done ;) May 27 21:30:32 cyrilRomain: it spent 8 hours getting to 50% May 27 21:30:36 updating portage cache? May 27 21:30:44 zecke: the new portage version correct that May 27 21:30:50 cyrilRomain: the last 4 hours were spent to to get to 52% May 27 21:31:03 Can anyone help me figure out how to fix this? May 27 21:31:03 /tmp/ccF33a3n.o: In function `main': May 27 21:31:04 serialfilter.c:(.text+0x128): undefined reference to `MAX' May 27 21:31:15 cyrilRomain: it looks like emerge scales worse than bitbake May 27 21:31:35 NAiL: check the build log May 27 21:31:42 NAiL: like implicit declaration of... May 27 21:31:45 zecke: well try portage 2.1, this problem was corrected May 27 21:32:05 cyrilRomain: how to get portage 2.1? this is the latest ARM stage3 May 27 21:32:43 oh my... zecke, are you installing gentoo on collie?? May 27 21:32:48 zecke: no such thing... May 27 21:32:59 there's only one more line in the log, and that's "collect2: ld returned 1 exit status May 27 21:33:19 zecke: on ARM ? ah dunno if 2.1_rc1 is available even masked .... May 27 21:33:23 ixoth: on Shark using NFSroot and swap on network block device May 27 21:34:07 lets see if the kernel still compiles =) May 27 21:34:12 zecke: portage-2.1_rc3 is availablefor ~arm May 27 21:34:36 cyrilRomain: is that a c implementation? May 27 21:34:48 zecke: no May 27 21:35:11 zecke: at least a think May 27 21:35:18 s/a/I/ May 27 21:35:55 hmmm... updating portage cache can take a LOT of time even on fast machines, so I suppose 20 hours is OK May 27 21:36:24 ....on such a device May 27 21:36:44 zecke: mad ;-) May 27 21:36:49 cyrilRomain: how can I update to portage 2.1? May 27 21:37:04 ixoth: I only have 400mb swap May 27 21:37:07 zecke: no, depends on python (I looked into the .ebuild). But where are the portage C implementation ? Is there a special package name ? a special USE flag ? May 27 21:37:12 RP: well the NFS and swap is served by the NSLu2 May 27 21:37:37 cyrilRomain: that was a question? I wondered how they fixed the 'performance' issue May 27 21:37:52 ixoth: I thought I would rsync the dep graph as well May 27 21:38:10 zecke: I think in algorithm, like in bitbake 1.4 ;) May 27 21:38:30 cyrilRomain: all due inline assembly ;) May 27 21:38:34 zecke: How do mutltimachine and tinderbox get on? If any given build channel supposed to represent one target or not? May 27 21:38:47 s/If/Is/ May 27 21:39:27 RP: ATM nothing is there to split them up May 27 21:39:46 zecke: Right. You have plans to change this? May 27 21:39:54 zecke: to update to portage 2.1, 'echo sys-apps/portage ~arm >> /etc/portage/package.keywords' and 'emerge portage' should do the trick May 27 21:40:03 zecke: ouch May 27 21:40:21 RP: sure May 27 21:40:21 zecke: to update to portage 2.1, "echo 'sys-apps/portage ~arm' >> /etc/portage/package.keywords" May 27 21:40:22 so, to use portage 2.1 you need to add a line like "sys-apps/portage ~arm" to /etc/portage/package.keywords May 27 21:40:33 oh damn.... my typing is sooooo slow May 27 21:40:56 zecke: I'm feeding some data into tinderbox. I'm not 100% sure I'm doing the right thing though :) May 27 21:40:57 can't I do ebuild to get portage 2.1 first May 27 21:41:23 RP: http://ewi546.ewi.utwente.nl/tinderbox/showlog.pl?machine_id=42&logfile=20060527220214.log ? May 27 21:41:59 zecke: Still running... May 27 21:42:32 that is a fast machine? May 27 21:42:46 zecke: Yes, but its doing the binary locale generation May 27 21:43:26 people should just learn one sane language, german, and the locale mess is avoided... May 27 21:43:33 zecke: If you consider a dual 2.8Ghz Xeon fast anyway... May 27 21:43:47 certainly not ;) May 27 21:43:48 zecke: you don't need to create an ebuild: since you emerge --sync you should already have one for portage 2.1 May 27 21:43:57 zecke: I should have applied my locale limitation patch May 27 21:44:19 zecke: Its just finishing as we speak :) May 27 21:44:29 cyrilRomain: sure, but I would like to kill that sync May 27 21:44:29 cyrilRomain: get the new portage May 27 21:44:29 cyrilRomain: sync again May 27 21:44:56 I need X2, my precious May 27 21:45:16 pb__: give me precious :} May 27 21:45:24 zecke: ok, then just download the ebuild here: http://gentoo-portage.com/sys-apps/portage May 27 21:45:38 cyrilRomain: can I just do ebuild portage.ebuild May 27 21:46:03 what do you mean by 'do' ? creating the ebuild ? May 27 21:46:22 cyrilRomain: no, doing the build May 27 21:46:41 zecke: 'emerge portage' then May 27 21:47:21 cyrilRomain: he probably wants to skip all dependency calculations etc May 27 21:47:25 cyrilRomain: but I have to wait for the generation of the cache to complete? May 27 21:47:47 zecke: I never Ctrl+C on that :-/ May 27 21:48:14 zecke: dunno if it can break something .... May 27 21:48:17 cyrilRomain: well within another hour it goes OOM anyway May 27 21:48:22 that would be something like: $ ebuild /path/to/file/portage.ebuild merge May 27 21:48:58 in fact, i Ctrl-C-ed on it a couple of times May 27 21:49:15 everything was OK, as far as I remember May 27 21:49:27 muaha May 27 21:49:52 320x240 works with the patch, but I'm getting nasty screen flickering on X start May 27 21:51:25 okay let us see how far i get May 27 21:55:09 Arrgh, ipkg segfault http://ewi546.ewi.utwente.nl/tinderbox/showlog.pl?machine_id=41&logfile=20060527230630.log May 27 21:55:53 cyrilRomain: space in the build dir? May 27 21:56:14 cyrilRomain: tosa-linux-gnueabi no '-' allowed May 27 21:56:34 zecke: no space May 27 21:56:50 I meant dash, it is getting late May 27 21:57:01 zecke: no dot either May 27 21:57:08 DASH May 27 21:57:09 DASH May 27 21:57:10 DASH May 27 21:57:13 '-' May 27 21:57:16 '-' May 27 21:57:16 '-' May 27 21:57:31 tosaDASHlinuxDASHgnueabi May 27 21:57:31 * cyrilRomain put his glasses May 27 21:57:32 DASH May 27 21:57:39 zecke: ok May 27 21:57:44 good :) May 27 21:57:56 another ipkg bug, but I didn't know it until this week May 27 21:59:30 :) May 27 21:59:39 zecke: well, it will be difficult if dash are not allowed anymore :-/ May 27 22:02:21 kergoth: the stamps aren't the problem May 27 22:02:39 cyrilRomain: it is allowed, ipkg just dies May 27 22:02:44 kergoth: I want to run do_stage twice on different ${STAGING} May 27 22:03:01 koen: bb.build.exec_task May 27 22:03:08 zecke: thanks May 27 22:03:16 koen: but see above ;) May 27 22:04:17 CoreDump|home: why did you add the GPL header? ;) May 27 22:04:32 hmm? May 27 22:05:18 zecke: I looked for ipkg modification since it worked 2 weeks ago. ipkg.bb was modified but to add a slugo thing. I have no idea why it started to segfault May 27 22:05:44 CoreDump|home: to fbset, I would prefer to have one license for all files May 27 22:06:09 zecke: like for the glibc loop, I can look in the past, but it take (too) many time. May 27 22:06:10 The .bb file is GPL, and so is fbset May 27 22:06:25 CoreDump|home: OE metadata is MIT licensed May 27 22:06:34 hrw|tv: all but fbset.bb is MIT May 27 22:06:41 ah heh May 27 22:06:48 zecke: iirc some AMD recipes was gpled... May 27 22:07:30 OE has mit as a blanket license, but everyone is free to choose a license for their stuff May 27 22:07:46 but OE strongly encourages GPL May 27 22:07:54 ehm May 27 22:07:55 koen *g* May 27 22:07:55 MIT May 27 22:07:58 koen: hmm May 27 22:08:20 koen: we should keep MIT otherwise we add uncertainity to ISVs :} May 27 22:08:37 koen: well we use any license, so check each bbfile before using OE :} May 27 22:08:37 <- wrong message May 27 22:08:37 zecke: I'm so not going to say what I want to say May 27 22:08:51 koen: msg me :) May 27 22:09:31 give it a rest guys, it's a bloody header template of mine May 27 22:09:46 CoreDump|home: hehe May 27 22:12:19 zecke: boostrap-image built correctly on 05/13/06 and started to segfault on 05/15/06. May 27 22:13:17 I prefer postcard-ware as a 'license' :) May 27 22:13:46 koen: like bluefish? when I started with Free Software I wanted to get postcards May 27 22:13:50 koen: until now i have none May 27 22:13:58 yeah, like bluefish May 27 22:17:57 cu May 27 22:18:14 ciao hrw May 27 22:20:41 shame that "OEREV = 'unknown'" in tinderbox build report. Having the revision number would help to track down bug spawning. May 27 22:21:12 cyrilRomain: hehe May 27 22:21:32 cyrilRomain: my first build bm-zecke segfaulted as well May 27 22:21:40 first time I have used '-' in my build May 27 22:23:14 but in /home/cyril/OpenEmbedded/build/tmp/work/tosa-linux-gnueabi, tosa-linux-gnueabi is created for me. May 27 22:23:32 zecke: can i safely rename tosa-linux-gnueabi to tosalinuxgnueabi ? May 27 22:23:38 zecke: a quick question (learning proper commit stuff), If I want to replace a file with a new one, instead of doing a diff (that will be quite large). Should I remove it and add the new one, or is there a simpler "replace" command? May 27 22:23:55 yes, you could fix ipkg as well May 27 22:24:04 Ken|JLime: rename? May 27 22:24:19 Ken|JLime: rename --execute and then adding changes on top of that? May 27 22:24:25 like a version bump? May 27 22:24:53 zecke: Well, I want to replace a kernel config file, I got the one that works here and want to replace the old one. May 27 22:30:06 'night mickey|zzZZzz :) May 27 22:32:46 * v8jlene is back (gone 08:20:33) May 27 22:34:45 darn kdrive doesn't have support for more than one screen res May 27 22:35:05 * cyrilRomain don't think segfault occurs because of '-' characters (the 05/13/06 build confirm that since I use '-' and '.' for build dir). May 27 22:35:33 cyrilRomain: use gdb and valgrind on the ipkg run May 27 22:35:36 03coredump 07org.oe.oz354x * r4fd31070... 10/packages/fbset/fbset_2.1.bb: fbset: Change meta-data to MIT license May 27 22:35:40 03coredump 07org.oe.dev * rfaf532c4... 10/packages/fbset/fbset_2.1.bb: fbset: Change meta-data to MIT license May 27 22:35:40 :} May 27 22:39:51 zecke: how can I use them within the bitbake run ? otherwise how can I get the context (i.e. with environment variables) from which ipkg is called ? May 27 22:40:34 cyrilRomain: ulimit -c May 27 22:40:43 cyrilRomain: should make ipkg CoreDump|home May 27 22:41:00 ? May 27 22:41:14 CoreDump|home: tab completion was shorter than writing coredump May 27 22:41:21 lol May 27 22:41:25 :) May 27 22:41:29 well it is now defeated May 27 22:42:40 * cyrilRomain does not know ulimit :-[ May 27 22:42:52 * cyrilRomain man ulimit May 27 22:42:59 or info ulimit May 27 22:44:49 good night everyone May 27 22:53:53 03lenehan 07org.oe.dev * r3f8a8e38... 10/packages/rsync/ (rsync_2.6.8.bb rsync_2.6.7.bb): rsync: Update SRC_URI for 2.6.7 and add 2.6.8. Closes bug# 947. May 27 23:08:16 zecke: how is going your portage 2.1 ? May 27 23:15:23 cyrilRomain: it built May 27 23:17:58 cyrilRomain: it is updating the cache now May 27 23:18:35 zecke: you should see the speed difference May 27 23:19:10 not yet May 27 23:28:10 O.o hum... May 27 23:28:35 I looked on the Gentoo forum, someone suggest to run 'emerge --metadata' after emerging portage May 27 23:29:23 cyrilRomain: yeah, I'm doing that atm May 27 23:30:35 I remember I did not run it and got some dump message in console, but only at the first emerge run May 27 23:31:01 32% May 27 23:33:22 'night all May 27 23:33:30 night koen May 27 23:36:54 48% May 27 23:36:59 a lot better May 27 23:37:04 zecke: suspence now May 27 23:37:25 zecke: the 50% wall should not be a wall anymore May 27 23:37:53 looks good, the screen is still swimming May 27 23:38:01 I need to patch this shark... May 27 23:41:02 cyrilRomain: 55% May 27 23:41:53 'night all May 27 23:41:54 zecke: great, should accelerate a bit now until 100% May 27 23:42:19 'night RP May 27 23:42:21 nite May 27 23:42:45 -g is now complete May 27 23:42:47 zecke: on which machine do you use this embedded Gentoo ? May 27 23:42:53 Shark May 27 23:42:58 'embedded' is relative May 27 23:43:21 cyrilRomain: it is called DNARD or Shark May 27 23:43:30 cyrilRomain: it was the NetworkComputer by DEC/DIGITAL May 27 23:44:36 zecke: how does a Shark look like, any photos ? May 27 23:45:36 cyrilRomain: http://www.cs.utah.edu/~danderse/dnard/ May 27 23:46:38 http://www.cs.utah.edu/~danderse/dnard/dnard-2.jpg May 27 23:47:24 oh god, it looks like a 2001 space odysee machine ! May 27 23:48:21 ;) May 27 23:48:33 75% May 27 23:52:42 * cyrilRomain should learn how to use gdb now. Just know how 'run' and 'backtrace' commands. May 27 23:53:00 got a coredump? May 27 23:53:01 laugh not accepted ;) May 27 23:53:08 gdb ipkg coredump May 27 23:53:11 bt May 27 23:53:23 well I can p,b,c,r ;) May 27 23:53:28 zecke: i tried 'gdb ipkg-cl' May 27 23:53:37 zecke: then run -f blabla May 27 23:53:39 bt May 27 23:53:57 ouch #0 0xb7e3fa5a in memcpy () from /lib/libc.so.6 May 27 23:54:24 hmm then use valgrind May 27 23:54:24 much better in this case May 27 23:54:27 http://oe.pastebin.com/742062 May 27 23:54:49 I try 'gdb ipkg-cl coredump' right now May 27 23:55:53 same May 27 23:56:13 * cyrilRomain should now learn how to use valgrind, never used. May 27 23:57:30 valgrind --tool=memcheck --num-callers=20 ipkg ... May 27 23:57:48 wow 100% May 27 23:57:58 cool May 27 23:58:01 thanks May 28 00:02:34 hum ... I'll re-run it with '> dump', so many LEAK SUMMARY and ERROR SUMMARY ! May 28 00:03:06 it ends with http://oe.pastebin.com/742076 May 28 00:04:02 cyrilRomain: the ending is not interesting the errors are May 28 00:04:12 yes, -v ! May 28 00:04:15 cyrilRomain: it is unlikely to segfault using valgrind May 28 00:04:34 the ERROR invalid read of size 4 May 28 00:04:37 is more interesting May 28 00:05:14 it has segfaulted though May 28 00:06:29 then it is really bad :) May 28 00:11:10 with -v: http://oe.pastebin.com/742089 May 28 00:11:49 not more informative for my eyes :-/ May 28 00:12:30 well just copy everything... May 28 00:12:51 the end is not interesting... May 28 00:16:26 http://oe.pastebin.com/742098 May 28 00:16:35 well I think the beginning lack May 28 00:16:49 s/lack/is missing May 28 00:17:29 (top of console buffer), redirection fail... I'll try with `` May 28 00:18:28 is it too hard to run valgrind May 28 00:18:29 hum ... does not work with valgrind `ipkg...` > dump May 28 00:18:33 and paste the complete log? May 28 00:19:08 ah, --log= :) May 28 00:25:02 uploading 800k dump May 28 00:25:31 AAhhhhaaAAHaAhhh : Fatal error: Allowed memory size of 8388608 bytes exhausted (tried to allocate 1863613 bytes) May 28 00:29:20 so the start: http://oe.pastebin.com/742105 and the end: http://oe.pastebin.com/742104 May 28 00:30:37 Ken|JLime: I put that commit example on my wiki - just for future reference and while I had the example handy: http://oe.twibble.org/mediawiki/index.php/Working_with_Monotone May 28 00:32:11 cyrilRomain: this lock looks torally fucked up... May 28 00:32:16 how many runs did you mix? May 28 00:32:19 cyrilRomain: one run May 28 00:32:20 no -v May 28 00:32:23 attach the log May 28 00:32:28 or file a bug at hh.org on ipkg May 28 00:34:14 zecke: well, one run but true with -v May 28 00:34:16 anyway cya May 28 00:34:25 yes, good night May 28 00:49:43 cyrilRomain: found something else? :-) May 28 01:04:46 03kristoffer 07org.oe.dev * rf4b1b8ad... 10/ (2 files in 2 dirs): May 28 01:04:46 jlime-donkey.conf: Updates distro file May 28 01:04:46 * gcc 4.1.1 May 28 01:04:46 * opie version 1.2.2 + include file May 28 01:04:46 linux-jlime-sh3-2.6.11/defconfig_jlime: fixes build issues May 28 01:04:46 * Removed i2c support and one wifi card. May 28 01:10:00 interesting, jlime must not have a decent lithion monitor May 28 01:10:16 emte: ? May 28 01:10:54 most common lithion monitoring chips are i2c May 28 01:11:27 few are onewire, but in most cases i2c is moreefficent for multiple devices May 28 01:12:11 aha, this build error is probobly more due to us using gcc 4.1.1, or somewhat atleast May 28 01:12:13 probably a few rare spi ones around still as well May 28 01:13:17 Ken|JLime: Have you tested a kernel built with gcc 4.1.1 yet? I know that sh kernels don't work with 4.0... don't think I've seen anyone try 4.1 yet. May 28 01:13:40 4.1.1 works perfect May 28 01:13:41 jnc was messing with it a while ago to make it work May 28 01:13:52 4.0.x though is shitty May 28 01:14:13 odd, i thought it was the reverse May 28 01:14:27 used 3.4.4 before, and had a couple of segfault issues with it May 28 01:14:31 got fixed with 4.1.1 May 28 01:14:43 Ken|JLime: I'll have to give it a go on my sh4 board then. May 28 01:15:04 v8jlene: do that, builds for sh3 (and works good) so shouldnt be a problem for sh4 May 28 01:20:55 03kristoffer 07org.oe.dev * r2ed8309a... 10/packages/pcmcia-cs/ (pcmcia-cs-3.2.8/superh-noO3.patch pcmcia-cs_3.2.8.bb): May 28 01:20:55 packages/pcmcia-cs/pcmcia-cs-3.2.8/superh-noO3.patch: Lowers Optimization May 28 01:20:55 * Sets -O2 instead of -O3 (otherwise 'pcrel too far' error) May 28 01:20:55 packages/pcmcia-cs/pcmcia-cs_3.2.8.bb: Including patch for sh3 May 28 01:20:55 * Sets usage of patch specific to sh3 May 28 01:31:59 koen: you awake man? May 28 01:32:14 <[g2]> Ken|JLime I think he's sleeping May 28 01:32:23 Just read the oe mailinglist May 28 01:32:31 whats up with that handhelds.org thing? May 28 01:32:56 <[g2]> I dunno, I haven't seen the ml May 28 01:33:43 Enough is enough -- you are no May 28 01:33:43 longer welcome here. You may no longer utilize handhelds.org or the services May 28 01:33:43 that it provides in any manner. May 28 01:33:56 pretty harsh words May 28 01:34:12 umm who stated that? May 28 01:34:23 George France May 28 01:34:23 Director May 28 01:34:23 Handhelds.org, Inc. May 28 01:34:36 then the lawyers made thier decision May 28 01:35:07 this "conflict" been in the making awhile then? May 28 01:35:24 says something about giving koen "several warnings" May 28 01:35:45 yes, a few months now May 28 01:36:01 odd, i dont have taht messsage from the oe list May 28 01:36:13 what is the title? May 28 01:36:19 <[g2]> what's the date and time ? May 28 01:36:20 Sent : Sunday, May 28, 2006 1:32 AM May 28 01:36:25 Re: newly built h3900 image fails booting... May 28 01:36:25 Oe Digest, Vol 18, Issue 68 May 28 01:37:36 hmm May 28 01:37:39 i dont have it ... May 28 01:38:19 emte: http://openembedded.pastebin.com/742181 May 28 01:38:33 thnks May 28 01:38:47 i wonder if my mailserver is holding it for soem reason May 28 01:39:17 whats the main issue of this conflict? May 28 01:39:36 personal attacks May 28 01:40:10 no virus, no spamassassin tags May 28 01:40:57 might just be load qued on the mail server May 28 01:41:46 or a problematic postfix filter ? ;) Man I sure hate those May 28 01:42:00 s/postfix/procmail May 28 01:42:39 if it were my actuall server i'd check May 28 01:42:55 its a hosted service tho May 28 01:43:02 i c May 28 01:43:30 well, maybe, just maybe oe should look for another webserver... May 28 01:43:33 too many problems with telus and running certain servers on res accounts May 28 01:44:00 that message looked like it regarded only one person May 28 01:44:10 correct May 28 01:44:11 it wasnt addressed to the projects May 28 01:45:14 meh qemu-nate in .oz354x if b0rked May 28 01:45:33 * emte wonders if it's time to turn the ribs May 28 01:46:41 few more min May 28 01:47:50 03kristoffer 07org.oe.dev * rb4098f1c... 10/site/sh3-linux: May 28 01:47:50 site/sh3-linux: Makes lzo build nicely May 28 01:47:50 * addition of ac_cv_sizeof_char_p=${ac_cv_sizeof_char_p=4} May 28 02:02:16 ~botmail for zecke: finally got wiki.openembedded.org pointed at the berlios site. let me know if you need anything else done, dns wise May 28 02:05:16 * [g2] hugs katossi__ May 28 02:05:19 <[g2]> doh! May 28 02:05:27 <[g2]> fscking tab complete May 28 02:05:30 hehe. hey [g2]. how ya been? May 28 02:05:42 <[g2]> that's who I was aiming for :) May 28 02:05:51 <[g2]> kergoth I'm quite well thx and you ? May 28 02:07:22 <[g2]> kergoth I've just build a Dapper Live CD that runs OE out of the box May 28 02:07:45 <[g2]> just wget the nslu2-linux Makefile, have some storage and build May 28 02:08:02 very nice May 28 02:08:16 doing pretty well here, though somewhat annoyed by the politics involved in working on a government project May 28 02:08:42 <[g2]> heh... You know there are only two phases of a govt project :) May 28 02:08:52 <[g2]> a) too early to tell May 28 02:08:58 <[g2]> b) too late to stop May 28 02:08:59 <[g2]> :0 May 28 02:09:02 hehe May 28 02:09:25 * [g2] was an Ada bigot in a former lifetime May 28 02:09:44 ugh, we've got so much ada in the tree, it scares me May 28 02:09:46 :) May 28 02:09:58 <[g2]> it'd scare me too :) May 28 02:11:42 atleast being gov you dont have to worry about money or deadlines :P May 28 02:12:11 <[g2]> emte no, just contract renewals May 28 02:12:41 <[g2]> and _THE MAN_ :) May 28 02:12:47 lol May 28 02:12:54 * [g2] too May 28 02:15:44 n8 all May 28 02:15:54 <[g2]> nite May 28 02:16:27 03coredump 07org.oe.oz354x * r39fec794... 10/packages/fbset/ (fbset_2.1.bb files/makefile.patch): fbset: Do not hard-code arm-gcc May 28 02:16:31 03coredump 07org.oe.dev * r9163a4b2... 10/packages/fbset/ (fbset_2.1.bb files/makefile.patch): fbset: Do not hard-code arm-gcc May 28 02:24:37 Coredump creates multiple heads and then runs away... Nice one ;) I'll merge. May 28 02:33:16 haha May 28 02:37:06 emte, heh, the deadlines are very tight, highly schedule driven May 28 02:43:03 calling it a night, cya tommorow May 28 02:55:28 hey peeps **** ENDING LOGGING AT Sun May 28 02:59:56 2006