**** BEGIN LOGGING AT Fri Sep 03 02:59:57 2010 Sep 03 03:00:00 for SH but just for some SoC Sep 03 03:00:23 can I create some kind of param? Sep 03 05:54:19 grg: usually the one with binutils should be used because thats the one that gets installed first Sep 03 05:54:38 development wise gcc one is developed and others sync with it Sep 03 05:55:08 khem, ok. I suppose i'll tell gdb not to install libiberty.a Sep 03 06:08:38 grg: thats right Sep 03 06:16:24 gm Sep 03 06:25:29 eFfeM_work, hello Sep 03 06:36:16 good morning Sep 03 07:19:47 morning Sep 03 07:20:31 gm all Sep 03 07:40:25 mornin Sep 03 07:41:56 zecke: did you send mail / get response on the sdk issue we discussed the other day? Sep 03 07:42:58 tasslehoff: I talked with RP yesterday... his shorter answer was he created nativesdk for these kind of issues and this is where my knowledge ends. :( Sep 03 07:43:13 tasslehoff: so I will have to figure out how to convert gcc-cross-sdk to nativesdk Sep 03 07:44:08 good morning Sep 03 07:44:38 zecke: My knowledge ended even earlier, cause I barely understood that :-) Sep 03 07:44:50 but a bug/problem in gcc-cross-sdk Sep 03 07:45:32 hej , anyone here knows how bitbake return values should be understood ? Sep 03 07:45:58 mlip: of the bin/bitbake binary? probably not at all Sep 03 07:46:00 thought that 0 would be success and 1 failure Sep 03 07:46:40 so there is no intentional meaning ? Sep 03 08:25:55 morning all Sep 03 08:27:11 good morning Sep 03 08:32:30 hi all, anyone cares to look at the soc-family patch I posted yesterday (and ack it if ok) Sep 03 08:36:57 eFfeM_work: please do not rush Sep 03 08:38:25 zecke: i won't but I feel class patches are often skipped/ignored/delayed and this fixes a problem for most machines/distro's Sep 03 08:39:47 eFfeM_work: currently testing your patch, build should finish in ~1h Sep 03 08:39:57 cool Sep 03 08:41:07 ericben: just to inform you about yesterdays buildbot experiment: it basically works; the build will mark as failure though, since bitbake returns 1 on success (but don't know whether that is guaranteed) Sep 03 08:41:12 new machine just arrived, mounting in progress and then this will be a buildbot machine so scheduling tests will be easier here ;-) Sep 03 08:41:33 actually without the patch if you have a machine that does not have SOC_FAMILY (e.g. sheevaplug) then you'll immediately see the problem when starting bitbake; it'll complain about not being able to resolve a dep to gcc2.95 because you are dragging in all those linux2.4 recipes Sep 03 08:41:59 angstrom solved it by defining SOC_FAMILY in the distro.conf file (to "Unknown" Sep 03 08:42:01 mlip: interesting, maybee we can call bitbake through a shell script to return the right value to make buildbot happy Sep 03 08:42:17 and JaMa worked around it by blacklisting the linux2.4 recipes Sep 03 08:42:42 ericben: i do currently call a shell wrapper that returns the bitbake value; making buildbot happy works, but won't let us get an "actual" result here Sep 03 08:42:51 ericben: if you use oestats that is no problem ofc Sep 03 08:43:20 eFfeM_work: ok that strange because I don't get this gcc2.95 thing here Sep 03 08:43:35 ericben: what machine/distro ? Sep 03 08:43:38 mlip: yes in that case we just don't use the buildbot results table Sep 03 08:43:49 eFfeM_work: angstrom / custom armv5 machine Sep 03 08:44:14 ericben angstrom does not have the issue, the angstrom conf file sets SOC_FAMILY Sep 03 08:44:43 eFfeM_work: OK so testing with minimal on an armv4 machine Sep 03 08:44:48 conf/distro/include/angstrom.inc:SOC_FAMILY ?= "Unknown" Sep 03 08:45:09 it showed for me doing minimal with sheevaplug (armv5t3) Sep 03 08:45:10 how can I integrate a new toolchain Sep 03 08:45:13 can I create some kind of param? Sep 03 08:45:17 for SH but just for some SoC Sep 03 08:45:47 mlip: did you have to do any magic to get buildbot running ? or are the files in contrib still up to date ? Sep 03 08:46:57 ericben: they are basically ok; I added an entry for bitbake, since it is not contained in oe anymore (i think it was in the past) Sep 03 08:47:22 ericben: currently struggling with git hang ups; the retry value of buildbot doesnt seem to work Sep 03 08:49:34 mlip: I've setup a mirror on a server which has a good connexion, do you want to try it ? Sep 03 08:49:41 Hi everybody, i would like juste do: task do_compile in some package without task do unpack because i modify manually a code so i would like just compile Sep 03 08:50:25 any idea please Sep 03 08:50:25 kamel, first do an unpack then modify the code then rerun the recipe Sep 03 08:50:31 ericben: If you mean as a slave; I am sorry, since I can't route the master server publicly (atm) Sep 03 08:50:45 Where can I find the do_compile instructions for a given .bb package if they are neither in the .bb file directly nor in an included file of the .bb ? Sep 03 08:51:11 anyone experience with this android tablet: http://www.linuxfordevices.com/c/a/News/LightInTheBoxcom-aPad-Android-Tablet/?kc=rss ? Sep 03 08:51:18 or suggestions for a better/other one? Sep 03 08:51:54 Radiao: in the class files Sep 03 08:51:58 thanks Sep 03 08:52:02 Radiao: look at the inherits statement and search the corresponding class file Sep 03 08:53:07 mlip: I was talking of the git server Sep 03 08:53:47 ericben: ah, if you have a mirror (also of testing-next) I can try that later yeah Sep 03 08:53:52 so if it inherits cmake, then it is classes/cmake.bbclass ? Sep 03 08:54:07 ericben: currently I need a few tries to get positive updates on testing-next AND bitbake in the same run Sep 03 08:56:33 eFfeM_work: ok using minimal you patch removes the error you mentionned Sep 03 08:56:44 I need the examples built of the "wt3" package, but per default only the wt3-libs are built -- any idea how I could tell bitbake to build these for the target environment? It uses CMake. Sep 03 08:57:40 anybody knows if eglibc is incompatible with PXA270 ? Sep 03 08:58:16 every filesystem I build don't work with a PXA270 board Sep 03 08:58:21 with AT91 works Sep 03 08:59:00 I tried minima-limage built for: mainstone, akita, c7x0. no way Sep 03 08:59:04 I could just go to tmp/work/mipsel-oe-linux/wt3-3.0.0-r0.1/wt-3.0.0/examples and type make there -- but it uses my i386 environment and fails with incompatible libs -- I need the correct environment there that bitbake sets up. Sep 03 08:59:22 VFS: Mounted root (jffs2 filesystem). Sep 03 08:59:22 Freeing init memory: 76K Sep 03 08:59:26 and freeze Sep 03 08:59:52 mckoan: is same build with glibc working ok? Sep 03 09:00:37 Jay7: I should try Sep 03 09:01:14 Jay7: same problem with Angstrom which uses glibc IIRC Sep 03 09:01:39 angstrom x11-image run ok on akita Sep 03 09:01:48 with glibc Sep 03 09:02:21 but it was built about 1-2 weeks ago Sep 03 09:02:25 mckoan: are youy using tune-xscale or tune-iwmmxt ? Sep 03 09:02:35 maybe you problem comes from iwmmxt Sep 03 09:03:00 ericben: I inherit it from the machine Sep 03 09:03:51 ericben: akita has MACHINE_FEATURES += "iwmmxt" Sep 03 09:05:21 mckoan: isn't MACHINE_FEATURES += iwmmxt is different than require include/tune-iwmmxt.inc ? Sep 03 09:06:33 ericben: grep tune-iwmmxt.inc conf/machine/* shows that I am not using it with mainstone nor akita Sep 03 09:07:00 ok this was just an idea :-) Sep 03 09:07:32 ericben: thx Sep 03 09:08:03 anyone knows if x-load that is generated by the recipe is used for flash and/or sdcard ? Sep 03 09:18:50 In command bitbake how just ia do task do _unpack Sep 03 09:18:59 or task do compile Sep 03 09:20:03 kamel I suggest start reading the doc. you might want to check out the -c opetion Sep 03 09:20:21 ahh ok thanks Sep 03 09:20:54 kamel: here's a link for you: http://bitbake.berlios.de/manual/ Sep 03 09:20:57 hint: ch 4 Sep 03 09:21:18 give a man a fish and he can eat for one day, learn a man to fish and he can eat all his lives Sep 03 09:21:43 so that's why i am pointing you to the place to get the answer instead of giving it directly Sep 03 09:26:20 eFfeM_work: thanks so mutch Sep 03 09:26:41 yw, hope you found the info you needed, and that it will help you in the future Sep 03 09:26:47 but if you get stuck don't be afraid to ask Sep 03 09:27:29 Can some one point me to git repos of other projects that use recipes. Other than git://git.openembedded.org/openembedded Sep 03 09:27:51 google for poky Sep 03 09:28:00 roger Sep 03 10:24:14 can I one way or another avoid that /lib/modules is created when I build an image; I don't have modules, my image file is stripped down to IMAGE_INSTALL = "kernel" and still I get this dreaded dir Sep 03 10:39:31 how can I introdue a new expand? Sep 03 10:39:48 basicaly I want to introduce new sub arch of SH Sep 03 10:40:08 and apply some patch to gcc, binutils & co only when use it Sep 03 10:40:53 I means a new append_ Sep 03 10:41:02 I want to do this Sep 03 10:41:10 SRC_URI_append_st40 Sep 03 10:41:17 or SRC_URI_append_sh_st40 Sep 03 10:42:24 pb_: hey, could you help me? Sep 03 10:42:43 zecke: I can try. What help do you require? Sep 03 10:45:27 pb_: it is about net-tools. I want to answer a question like: is eth* up, right now I look into /proc/net/route to get my answer. Is there any better way? I looked at net-tools and it is using the procfs file as well. Sep 03 10:46:43 I define SOC_FAMILY="st40" Sep 03 10:47:20 zecke: if what you want to know is "do I have an internet connection" then yes, inspecting /proc/net/route for a default gateway is as good a way as any. I think you can also ask netlink if you don't want to rely on /proc. Sep 03 10:47:48 zecke: if you actually want to know whether the physical eth* interfaces are up then you can use ioctl(SIOCGIFFLAGS) to find out. if they are up then they will have IFF_UP | IFF_RUNNING set. Sep 03 10:48:32 pb_: do you think the ioctl is any faster than open/read on /proc/net/route? Sep 03 10:48:52 yes, but I doubt you would notice the difference in most cases. Sep 03 10:51:50 zecke, has anyone tested tinderbox running on a 64bit instance? Sep 03 10:54:17 ka6sox4: i don't think so. Do you run into issues? I would assume python runs the same on 32 and 64 bit machines? Sep 03 10:54:46 it does Sep 03 10:55:37 okay I didn't see anything jump out at me with Ubuntu 10.04 LTS with regard to Tinderbox Sep 03 10:56:13 I'll clone melo and try an upgrade on the clone Sep 03 10:57:59 I think that garnet made the transition okay..I didn't hear that khem was having trouble with builds. Sep 03 10:58:32 ka6sox4: thanks! Sep 03 11:05:21 ka6sox4 are you also involved in our git server? Sep 03 11:08:07 * ka6sox4 doesn't know how to answer that....if its down or needs upgrading I do that...content is somebody elses baliwick. Sep 03 11:09:31 ka6sox4 well it seems to reject connection upon starting quite regularly and the speed reminds me of my modem days (I get around 100K/s so it *is* better than modem :-) ) Sep 03 11:11:33 eFfeM_work, we need to evaluate the link between the US and .eu Sep 03 11:11:44 here I see pretty high rates. Sep 03 11:11:54 hm strange Sep 03 11:12:13 i don't see this with other servers Sep 03 11:12:25 when you see others from .eu on ask them about the rates. Sep 03 11:12:42 likewise also mentioned getting poor performance Sep 03 11:13:30 okay I"ll take a look @ the monitoring graphs and see what is happening locally @ the server. Sep 03 11:13:42 great, thanks Sep 03 11:14:13 and if it helps if you want me to initiate a transfer or start a git clone or a wget or so let me know Sep 03 11:17:02 ka6sox4 did you see this thread on the ML? http://thread.gmane.org/gmane.comp.handhelds.openembedded/36455 Sep 03 11:23:12 okay...didn't see that one. I can't see any reason for this unless its a problem with Git itself...when khem and I were testing with garnet to melo (a Gb wire) we saw the same issue. Sep 03 11:23:51 there is only about 100Mb aggregate on our subnet that I can see. Sep 03 11:24:34 not in speed but in reliability of connections Sep 03 11:24:51 I'll ask rwhitby how the webos-internals git is acting. Sep 03 11:27:09 ok, thanks for giving this attention Sep 03 11:28:41 let me add another CPU to Melo. Sep 03 11:28:47 quick down and up Sep 03 11:31:48 of course its been 180days since its last fsck....figures Sep 03 11:32:59 ka6sox4, hm, guess this is a cluster, is it possible that one of the systems has a configuration issue causing it to hangup ? Sep 03 11:33:14 * eFfeM_work does not know too much about clusters, so apologies if that is a dumb question Sep 03 11:33:44 no its not a dumb question..I did look @ all the systems on the cluster and nothing appears to be out of whack Sep 03 11:35:00 okay you now have 2 CPUs and enough ram.... Sep 03 11:35:07 although its not swapping Sep 03 11:36:52 ok, let me kick off a clone Sep 03 11:37:11 no hangup Sep 03 11:37:52 still pulling at around 100 KiB/s, now dropping to 80-90 Sep 03 11:39:07 short test to local server gives around 1 MB/s up and down Sep 03 11:42:32 pb_: okay, using getifaddrs is slower than g_file_get_contents on /proc/net/route Sep 03 11:44:34 pb_: is there an easy ioctl to check if there is any default gateway set? Sep 03 11:44:36 hm, following instructions is extremely difficult, especially if you are the one who wrote the instruction in the first place Sep 03 12:05:25 eFfeM_work, are you stil doing that clone? Sep 03 12:05:49 yesm it is now recieving deltas Sep 03 12:06:08 it is finised now Sep 03 12:06:16 do you want me to redo it ? Sep 03 12:07:14 yes, please I want to set a trap on your connection to diagnosis it Sep 03 12:07:25 diagnose Sep 03 12:08:01 started Sep 03 12:08:26 no hangup Sep 03 12:08:43 speed now 90-100 kib-s Sep 03 12:08:46 KiB/s Sep 03 12:09:05 that jives with what I am seeing too. Sep 03 12:09:24 jives? is that agrees ? Sep 03 12:09:45 actually it is peddling between 80 and 120 or so, seems like a windowing thingie Sep 03 12:09:47 oh sorry yes Sep 03 12:09:52 np Sep 03 12:10:06 i've learned some words over the years Sep 03 12:10:45 I need someone in the US to do a clone Sep 03 12:11:21 most of the US hasn't even had their first cup of coffee Sep 03 12:13:46 I'd do a clone, except I'm not home and apparently my build machine has suffered a fan failure. Sep 03 12:14:45 hiya mwester...I'm doing one now Sep 03 12:14:58 Oh good. Sep 03 12:15:25 It's awful early out on the west coast for you Sep 03 12:15:43 yes, had an earache that has kept me up all night Sep 03 12:16:26 eFfeM_work, I'm only getting about 174kB/sec Sep 03 12:16:43 so not sure...I'll have to talk to the colo in a couple of hours. Sep 03 12:17:15 hm, that does not sound too good, but seems it is close to you then Sep 03 12:17:34 closer...dunno where it is yet. Sep 03 12:19:00 looks like 4 clones going on for only 400kB/sec of download. Sep 03 12:19:48 russia just woke up...but they are only getting about 25kB/sec Sep 03 12:20:08 and what kind of bandwidth does the system have ? Sep 03 12:20:20 supposedly Gb Sep 03 12:20:30 but in reality probably only 400Mb Sep 03 12:20:32 i'm still around 100 KiB Sep 03 12:20:42 ya, I see that Sep 03 12:21:02 400 is still nice, we lost a factor 1000 Sep 03 12:21:43 getting the error below when building the latest angstrom base-image with DISTRO=angstrom-2010-x Sep 03 12:21:44 ERROR: '/beagleboard/angstrom/source/sources/openembedded/recipes/linux/linux-omap-psp_2.6.32.bb' failed Sep 03 12:21:54 anyone faced this issue? Sep 03 12:22:35 somebody tripped over the fiber and as a fallback we have a 10Mbit network connection ;-) Sep 03 12:23:23 backhoe Sep 03 12:23:39 could be avian carrier :-) Sep 03 12:24:13 during Pigeon season the packet loss is murder Sep 03 12:26:40 especially in the hunting season ;-) Sep 03 12:28:18 hi Sep 03 12:29:20 i am merging a native recipe into non-native, i need to have different versions of do_compile_append for native and target Sep 03 12:30:23 but do_compile_append_virtclass-native does not get called, and the regular do_compile_append is called Sep 03 12:32:07 hi eFfeM_work Sep 03 12:34:14 hi fahad, Sep 03 12:35:07 no idea on your problem maybe it should be do_compile_virtclass-native_append instead, otherwise no idea and no time to look at it now Sep 03 12:35:08 sry Sep 03 12:35:09 is it not possible to over-ride the do_compile_append function using the virtclass-native? Sep 03 12:35:44 i also tried it this way, but still the regular do_install_append is called Sep 03 12:35:51 okay, thanks anyway Sep 03 12:37:15 eFfeM_work, I just cloned to a machine across the Gb wire between physical machines @ 15.5Mb/sec Sep 03 12:37:22 so its NOT a server issue Sep 03 12:37:50 good, Sep 03 12:37:57 bad Sep 03 12:38:01 fahad maybe someone else knows Sep 03 12:38:07 I have control over the server...but not the net Sep 03 12:38:35 ka6sox4 ah ok, but at least we have reduced a source of failure and limited it to the distance bewteen you and the server Sep 03 12:38:45 actually the part between server and where it splits to you and me Sep 03 12:38:51 guess it is close to the server Sep 03 12:39:21 I have 2 physical machines and went from 1 to the other with a clone... Sep 03 12:39:26 we have historically had slowness problems between git server and EU, and its always been fast in the US Sep 03 12:39:49 only 80-130Kb/sec now Sep 03 12:39:51 cbrake: apparently not any more Sep 03 12:39:54 pretty poor Sep 03 12:41:57 I'm getting between 200-500 KiB/sec on clone Sep 03 12:42:21 something just changed. Sep 03 12:42:36 rates are 5X what they were 30 seconds ago Sep 03 12:43:11 eFfeM_work: what about in EU? (I assume that is where you are?) Sep 03 12:43:39 i'm getting 80-120 KiB/sec Sep 03 12:45:05 well, I guess we get what we pay for -- if we want better performance, we probably need to be in a data center Sep 03 12:45:25 it *is* in a data center, though Sep 03 13:26:59 ka6sox4 now cloning from git://repo.or.cz/openembedded.git, thisone seems to give around 1 MiB Sep 03 13:27:15 actually there is some variance, saw it go to 2 at peaks Sep 03 13:27:45 might also be due to our network load, might try tonight when everyone is home Sep 03 13:34:11 OE GIT gives me 10 kB/sec on a clone Sep 03 13:47:39 re Sep 03 13:49:14 likewise: khem i also have issues with gcc4.5 busybox and my jtag program segfault, some other small progs do work Sep 03 13:52:58 no idea what is wrong, the jtag prog for 4.5 is 10k stripped, with 4.3.3 it is 15k stripped Sep 03 13:59:07 does exis any setting to build busybox with static libraries? Sep 03 14:00:46 i think there is a recipe for busybox-static Sep 03 14:07:30 eFfeM_work: thx indeed, and anly a few of images use it, trying micro-image now Sep 03 14:08:00 you can always roll your own image, that is not difficult Sep 03 14:08:02 * mckoan starts hating Marvell stuff Sep 03 14:12:50 they have some nice stuff, but docs often suck or are avaible only under nda Sep 03 14:14:32 morning Sep 03 14:57:20 03Steffen Sledz  07org.openembedded.dev * r8f341294a7 10openembedded.git/recipes/rt2x00/ (3 files in 2 dirs): Sep 03 14:57:20 rt2x00: add linux kernel driver for Ralink RT3572 Sep 03 14:57:20 Signed-off-by: Steffen Sledz Sep 03 15:21:23 yikes, i knew it was a while ago, but holy cow Sep 03 15:21:36 1.10/master branched off of 1.8 3.5 years ago Sep 03 15:28:50 heh Sep 03 15:34:52 omg Sep 03 15:35:42 * soltys yawns Sep 03 15:38:16 soltys: do not yawn, go to atm ;D Sep 03 15:39:16 wtf ?:) Sep 03 15:41:44 ;d Sep 03 15:48:43 i wish cgit had a "compare branch" function, the way github does Sep 03 15:57:59 re Sep 03 16:44:50 mckoan|away: wonder what's up with the marvell stuff? Sep 03 16:47:15 likewise: i feel the hw is often ok, but it is often also very hard to get info Sep 03 16:47:42 and support for devices like openrd somehwat sucks Sep 03 16:47:52 like getting a good u-boot Sep 03 17:00:49 agreed. would like to learn mckoan's experience on this./ Sep 03 17:28:13 I have a bitbake question; I'm trying to build the qt4 library for avahi; I've made a .bb for it, however my libavahi-qt ipk depends on the metapacakge "avahi-qt" .. when I want it to only depend on avahi. Sep 03 17:37:38 its being build using a library from avahi-qt Sep 03 17:38:00 now make sure that this library that it needs at buildtime/runtime Sep 03 17:38:24 is provided by avahi too Sep 03 17:41:32 so the dpkg-deb -I shows it depending on libavahi-common3, which is good, and what it should depend on. Sep 03 17:44:21 it's just that 'avahi-qt' is shown as well, which shouldn't be a package. Sep 03 17:46:47 is there a way to tell bitbake that isn't a package to be made Sep 03 17:47:50 it isn't actually making an ipk for it, since there are no files for it to package.. Sep 03 17:49:48 03Khem Raj  07org.openembedded.dev * r65c8f0a20e 10openembedded.git/recipes/classpathx/gnumail_1.1.2.bb: Sep 03 17:49:48 gnumail_1.1.2.bb: Point inetlib to java data dir in target staging area. Sep 03 17:49:48 Signed-off-by: Khem Raj Sep 03 17:49:49 03Khem Raj  07org.openembedded.dev * r499f4aacba 10openembedded.git/conf/distro/minimal.conf: Sep 03 17:49:49 minimal.conf: Select bluez4 for bluez-utils-dbus and bluez-utils Sep 03 17:49:49 * Set bluez-libs for PREFERRED_PROVIDER for bluez-libs Sep 03 17:49:49 Signed-off-by: Khem Raj Sep 03 17:49:59 03Khem Raj  07org.openembedded.dev * rf941f5c0aa 10openembedded.git/conf/distro/include/ (2 files): Sep 03 17:49:59 preferred-xorg-versions-X11R7.5-latest.inc, preferred-xorg-versions-X11R7.5.inc: Bump pinned versions of lndir and libxvmc Sep 03 17:49:59 * The already pinned versions of these packages are in obsolete Sep 03 17:49:59 therefore pin something thats available and supported. Sep 03 17:50:00 Signed-off-by: Khem Raj Sep 03 17:52:15 03Khem Raj  07org.openembedded.dev * r6f5892efa3 10openembedded.git/recipes/autoconf/autoconf213_2.13.bb: Sep 03 17:52:15 autoconf213_2.13.bb: Re-add. Sep 03 17:52:15 * Autoconf 2.13 is special and many current packages oe/non-oe uses it Sep 03 17:52:15 we need to keep it. Sep 03 17:52:15 Signed-off-by: Khem Raj Sep 03 18:24:19 hello Sep 03 18:24:22 me again Sep 03 18:24:57 any chance I can get pointed in the direction of the default bitbake config file Sep 03 18:30:44 conf/local.conf Sep 03 18:31:01 suggest to read some pages of the wiki (e.g. getting started) Sep 03 18:42:11 nm, I found it Sep 03 18:42:18 thanks Sep 03 18:52:55 this could take awhile, fo sho Sep 03 18:57:18 package browser on angstrom-distribution is broken, seems web server is full Sep 03 20:11:23 I build usual angstrom2008 with matchbox and GPE. GPE is by default. I run this image in my target, shuted down the gpe window manager and trying to start matchbox. .. anybody know how to do it? Sep 03 20:22:46 hi all - how do you define libraries to link against in openembedded? (trying to get fdisk to link against libtinfo) Sep 03 20:23:08 well, in bitbake, I guess is more correct to ask Sep 03 20:28:53 no, you were correct the first time Sep 03 20:30:42 dv__: matchbox-desktop iirc Sep 03 20:35:42 ahhh, good to know, thanks kergoth Sep 03 20:38:38 modified the Makefile to get Angstrom to keep compiling (blurgk), but at least it'll get me through until I know "the right way™" Sep 03 21:11:55 anyone know if there is there someway to extract that postinst scripts from an ipk Sep 03 21:20:36 an ipk is just an 'ar' archive with two tarballs inside Sep 03 21:20:39 you can extract it easily Sep 03 21:34:27 kergoth, are you experiencing any slowness with Git? Sep 03 21:38:34 is there a recipe already setup for XDirectFB? Sep 03 22:11:32 hello Sep 03 22:11:32 is there a recipe already setup for XDirectFB? Sep 03 22:11:48 Ther Sep 03 22:11:51 ops..sorry Sep 03 22:12:30 Where are found the directory of sources os PSPLASH to change Splash Image ? Sep 03 22:28:32 angelox_123: recipes/psplash read the recipes and you get the details to change the image Sep 03 22:55:08 have anyhow to emulate a image compiled into qemu ? Sep 03 22:56:21 depends on the target Sep 03 22:56:29 armv5te Sep 03 22:57:02 or you say of DISTRO ? Sep 03 22:59:37 no Sep 03 22:59:41 I meant the target Sep 03 22:59:54 qemu-system-arm --help will help alot Sep 03 23:00:06 thanks! Sep 03 23:00:29 but if i run: qemu-system-arm ZIMAGE ROOTFS it will work yes? Sep 03 23:00:51 no Sep 03 23:00:59 =( Sep 03 23:01:37 qemu-system-arm -kernel ZIMAGE, and depending the filetype of ROOTFS, must mount rootfs the way intended Sep 03 23:04:38 Image-a1200.bin: No such file or directory Sep 03 23:04:39 qemu: could not load kernel 'zImage-a1200.bin' Sep 03 23:04:45 but the file is on directory Sep 03 23:09:41 angelox_123: see contrib/qemu/run-qemu.sh Sep 03 23:09:52 and adapt it to your needs Sep 03 23:10:14 qemu boots OE very well on various arches Sep 03 23:11:23 thanks i will try Sep 03 23:16:00 here don't have kernel compiled to qemu..how to compile? (bitbake *** ) Sep 03 23:21:37 angelox_123: MACHINE=qemuarm DISTRO=minimal bitbake minimal-image Sep 03 23:21:53 thanks Sep 03 23:21:54 will generate a kernel and root file system which is bootable in qemu Sep 03 23:22:23 but i can't use another rootfs outside this ? Sep 03 23:24:31 you can Sep 03 23:26:03 cheers Sep 04 00:07:17 is there a XDirectFB recipe? Sep 04 00:07:34 NOTE: Couldn't find shared library provider for libgcc_s.so.1 Sep 04 00:07:34 ERROR: '/home/dev/oe/angstrom-setup-scripts/sources/openembedded/recipes/linux/linux-omap-psp_2.6.32.bb' failed Sep 04 00:07:41 and is there any help for that error? Sep 04 00:09:27 Module tun not found. Any help to run-qemu ? Sep 04 00:13:46 yes Sep 04 00:13:54 google tun/tap Sep 04 00:24:56 angelox_123: if you read comments in run-qemu.sh it should help you set it up too Sep 04 00:25:07 atleast on debian/ubuntu systems Sep 04 00:34:58 03Florian Boor  07org.openembedded.dev * r3c80d12dff 10openembedded.git/recipes/linux/ (3 files in 2 dirs): linux: Add dockstar support to 2.6.35 Sep 04 00:35:03 03Florian Boor  07org.openembedded.dev * r9569e129ba 10openembedded.git/conf/machine/dockstar.conf: dockstar.conf: Machine configuration for Seagate Dockstar Sep 04 01:02:00 why Openmoko isn't supported on OE ? Sep 04 01:07:06 angelox_123: what do you mean? Sep 04 01:07:26 angelox_123: the Openmoko distribution is gone, but the gta0x is well supported, e.g. by SHR Sep 04 01:07:52 on openembedded says isn't using Sep 04 01:10:45 angelox_123: what do you mean? Sep 04 01:36:32 hello all Sep 04 01:36:39 Seems I suffer from this error... Sep 04 01:36:41 http://permalink.gmane.org/gmane.comp.handhelds.openembedded/36494 Sep 04 01:42:06 hmmm Sep 04 01:44:44 any idea kergoth? **** ENDING LOGGING AT Sat Sep 04 02:59:57 2010