**** BEGIN LOGGING AT Sun Mar 29 02:59:57 2009 Mar 29 09:11:07 * kgilmer needs a sanity check Mar 29 09:11:23 i run devshell on an arbitrary package that builds fine Mar 29 09:11:28 i echo $CC Mar 29 09:11:47 I try to run the result and get arm-angstrom-linux-gnueabi-gcc: No such file or directory Mar 29 09:11:59 shouldn't that work? is something in my configuration hosed? Mar 29 09:12:08 I see the cross tools in the TMPDIR Mar 29 09:14:23 hi everybody Mar 29 09:18:48 I'm back to libwebkit-1.0-1 - 0.1+svnr40814-r8.1 that works with epiphany (2.24.2) and midori (midori-0.1.2+r0+gitrf30dff433056e239ca775d16f343bdb249f1a338-r0), but I have to use ajax,youtube,pdf and doc plugin, all on beagleboard. anydoby has used gnash or swfdec on arm platform ? Mar 29 10:19:59 i get ERROR: QA Issue: gdk.pc failed sanity test, how do i rebuilt gdk Mar 29 10:24:10 hi everyone Mar 29 10:24:49 hi Mar 29 10:25:02 I'm trying to bitbake qt4-embedded-4.5.0, which appeared in the org.openembedded.dev repo about 2 days ago Mar 29 10:25:47 but I think there's something wrong: Multiple .bb files are due to be built which each provide qt4-embedded Mar 29 10:26:41 I can't seem to find the problem Mar 29 10:30:04 the only difference from the 4.5 bitbake file compared to the 4.4.3 file is (except the different patches): the line: PR = "r10" in 4.4.3 was changed to DEFAULT_PREFERENCE = "-1" Mar 29 10:30:10 in 4.5 Mar 29 10:36:41 does anyone know how to solve this? Mar 29 10:45:57 !log Mar 29 10:48:12 which package produces gdk.pc? Mar 29 11:16:40 03Koen Kooi  07org.openembedded.dev * r1d5d7f5fe5 10openembedded.git/ (15 files in 3 dirs): fluxbox: add 1.1.1, remove old versions Mar 29 11:17:02 03Koen Kooi  07org.openembedded.dev * r12ae476961 10openembedded.git/contrib/angstrom/build-feeds.sh: angstrom feed builder: add fluxbox Mar 29 11:42:20 03Leon Woestenberg  07org.openembedded.dev * r69d1bc9c60 10openembedded.git/contrib/dependsgraph/dependsgraph.sh: contrib/dependsgraph: Shell script showing simple depends graph. Mar 29 11:42:25 03Leon Woestenberg  07org.openembedded.dev * r1bdc66e108 10openembedded.git/recipes/ltp/ltp_20090131.bb: Mar 29 11:42:25 ltp: must use configure now; see INSTALL. Mar 29 11:42:25 This fixes a build issue. Mar 29 11:42:25 However, QA issues remain regarding GNU hash. Mar 29 12:39:16 Is there any 'mipsel' maintainer? Mar 29 12:45:38 ant__, don't know but there is a MAINTAINER file in oe Mar 29 12:46:59 empty..wrt wl500g machine Mar 29 12:47:06 I'll fix a typo Mar 29 12:47:38 has anyone tried to bitbake qt4-openembedded 4.5? here it says: Multiple .bb files are due to be built which each provide qt4-embedded Mar 29 12:47:43 03Stefan Schmidt  07fso/milestone5.5 * rf1c2d1af3e 10openembedded.git/conf/checksums.ini: checksums.ini: New sums for connman, wpa-supplicant and openbmap Mar 29 12:47:43 03Stefan Schmidt  07fso/milestone5.5 * r01ff4952c1 10openembedded.git/packages/connman/ (connman_0.14.bb connman_0.15.bb): connman: Add version 0.14 and 0.15 Mar 29 12:47:44 03Stefan Schmidt  07fso/milestone5.5 * rb096e29d45 10openembedded.git/packages/wpa-supplicant/wpa-supplicant_0.6.9.bb: wpa-supplicant: Add version 0.6.9 Mar 29 12:47:45 03Stefan Schmidt  07fso/milestone5.5 * r3c14dc1414 10openembedded.git/packages/openbmap/openbmap-logger_0.2.0.bb: openbmap: Add version 0.2.0 Mar 29 12:47:47 03Stefan Schmidt  07fso/milestone5.5 * r375d73f07a 10openembedded.git/conf/distro/include/preferred-om-2008-versions.inc: preferred-om-2008-versions: Bump wpa-supplicant and connman Mar 29 12:47:50 03Stefan Schmidt  07fso/milestone5.5 * r0b61b8085f 10openembedded.git/packages/connman/ (connman.inc connman_0.15.bb): connman: Package python test scripts into connman-test-utils Mar 29 12:47:53 03Stefan Schmidt  07fso/milestone5.5 * r7303b5d7cf 10openembedded.git/packages/tasks/task-fso-compliance.bb: task-fso-compliance: Install connman-test-utils in fso images Mar 29 12:48:21 hehe.. Mar 29 12:48:42 * Jay7 thinking about buying some wifi router/bridge Mar 29 12:49:01 may be get one for testing OE too? :) Mar 29 12:49:36 Broadcom 63xx is almost done (OpenWrt) Mar 29 12:49:43 would be nice Mar 29 12:50:05 very chap and popular routers Mar 29 12:50:13 well, cheap Mar 29 12:50:28 what models are based on this? Mar 29 12:51:25 http://wiki.openwrt.org/CategoryBCM63xx Mar 29 12:51:29 :-] Mar 29 13:01:56 03Stefan Schmidt  07org.openembedded.dev * r7777d1dda6 10openembedded.git/conf/checksums.ini: checksums.ini: New sums for connman, wpa-supplicant and openbmap Mar 29 13:01:56 03Stefan Schmidt  07org.openembedded.dev * r43a6942035 10openembedded.git/recipes/connman/ (connman_0.14.bb connman_0.15.bb): connman: Add version 0.14 and 0.15 Mar 29 13:01:57 03Stefan Schmidt  07org.openembedded.dev * r2c722ec1d6 10openembedded.git/recipes/wpa-supplicant/wpa-supplicant_0.6.9.bb: wpa-supplicant: Add version 0.6.9 Mar 29 13:01:58 03Stefan Schmidt  07org.openembedded.dev * r6374461dcf 10openembedded.git/recipes/openbmap/openbmap-logger_0.2.0.bb: openbmap: Add version 0.2.0 Mar 29 13:11:03 hello, is it normal that I have so much undefined here: http://rafb.net/p/DaUQp432.html Mar 29 13:12:06 with things like islower or isupper (ctypes.h in C but here it's c++) Mar 29 13:37:22 i always get this error: "ERROR: QA Issue: gdk.pc failed sanity test" this error is also listed a bug in oe. is there a workaround Mar 29 13:37:28 see: http://projects.linuxtogo.org/pipermail/openembedded-issues/2008-June/009916.html Mar 29 14:01:09 03Andrea Adami  07org.openembedded.dev * r8d94d2ea3b 10openembedded.git/conf/machine/wl500g.conf: wl500g.conf: fix include Mar 29 14:02:25 rb500.conf seems broken too Mar 29 14:02:40 which kernel for these mipsel (mips2)? Mar 29 14:46:11 is there anyone there? Mar 29 14:46:30 I think I've a big problem with boost_1.36.0: http://rafb.net/p/DaUQp432.html Mar 29 14:47:08 I've not one but 220 undefined...with standard things such as tolower, and toupper Mar 29 14:48:59 Why not just use the version of boost that actually builds? Mar 29 14:49:11 Don't you first need to the set LD_LIBRARY path the the root of your staging Mar 29 14:50:05 keesj, ah yes maybe...lol Mar 29 14:50:42 mwester, what version do you mean...I use the lastest one avaliable because rschuster told me to use it Mar 29 14:50:53 and because I prefer using the lastest one Mar 29 14:51:12 thanks a lot Mar 29 14:51:58 * mwester continues to wonder at the un-natural obsession people seem to have for "the latest release/patch/update" when they have no grounds to prefer that over any other. Mar 29 14:53:42 Gnutoo: grep for "boost" in openembedded/conf and openembedded/conf/include --- that'll give you a good idea of what others have found to work when the latest does not. Mar 29 14:53:59 mwester, gnuradio needs a recent boost :( Mar 29 14:54:00 ok thanks Mar 29 14:54:06 so I ahve to sort of care Mar 29 14:54:06 mwester doesn't understand "becasue it's there" Mar 29 14:54:32 mwester: have you an idea about mipsel state? Mar 29 14:54:32 http://www.askoxford.com/worldofwords/quotations/quotefrom/mallory/ Mar 29 14:54:43 * mwester DOES understand "because it's there", and /me works in field tech support, so he also knows all about stability and what it takes to make something actually work. Mar 29 14:54:50 :P Mar 29 14:54:54 ant__: No idea, I'm afraid. Mar 29 14:55:17 Crofton|work: gnuradio also requires an obsolete libusb, IIRC -- strange combination! Mar 29 14:55:23 well Mar 29 14:55:38 let's see we are in the transition from one libusb to another Mar 29 14:55:46 But clearly you *know* why you need the most recent boost Mar 29 14:55:49 and they peeked insde private structs Mar 29 14:56:01 eb likes some fancy boost class .... Mar 29 14:56:31 my comments are directed to those who blindly and glibly "upgrade" and then wonder why something broke, when they really should wonder why they upgraded if they don't know what the upgrade will get them... Mar 29 14:56:54 (Which pretty much encompasses every Windows system in home use) Mar 29 14:56:58 I will admit to blindly upgrading iperf Mar 29 14:57:20 but, it is not a core dependency Mar 29 14:57:34 you mean... like busybox? :-D Mar 29 14:57:40 :) Mar 29 14:57:44 (which breaks with pretty much every upgrade!) Mar 29 14:58:06 busybox changes should need 10 acks and 2 weeks of real testing :) Mar 29 14:58:13 at least! Mar 29 14:58:52 my attempt at OSK u-boot testing really broke things :) Mar 29 14:58:53 I feel we've just now gotten to the point where we know what works and doesn't with the current (outdated) busybox in OE... I guess that means its time for a change? Mar 29 14:59:29 But on a serious note re: busybox -- the removal of tinylogin is going to force some disruptive changes, I fear. :( Mar 29 15:00:07 IIRC, a non-tinylogin system will require busybox to run SETUID root Mar 29 15:00:41 I think we need 10 acks on a code-review of the busybox startup code before we can do that. Mar 29 15:00:50 yeah Mar 29 15:00:57 there has been a lot of talk Mar 29 15:01:10 need some action Mar 29 15:01:30 maybe we can switch minimal distro to bb login for testing Mar 29 15:01:48 * mwester adds "replace tinylogin" to his list of SlugOS changes for his experimental branch Mar 29 15:02:19 I actually had it working around November, but with an older busybox, and there were some serious problems with init. Mar 29 15:02:46 I'll have to dig up my notes and see what that was, and how we avoid those, and if they are specific to SlugOS Mar 29 15:03:37 With 2.6.29 kernel being much larger than the 2.6.27 kernel, I could use the space in flash anyway, so maybe this isn't a bad thing, assuming we feel ok about the security aspects. Mar 29 15:04:24 mwester: there are some lzma compressed kernel stuff patches around, (probably too experiment tho) Mar 29 15:07:43 http://www.embeddedlinuxconference.com/elc_2009/sessions.html#Gend Mar 29 15:10:30 top 3 *pains*? I wonder what the pain and delay would have been if I didn't have bitbake/oe :-) Mar 29 15:10:40 hi timtimred Mar 29 15:10:52 timtimred: bother bother ssd sheeva :-) Mar 29 15:13:12 oh right right Mar 29 15:16:53 03Leon Woestenberg  07org.openembedded.dev * rc9960f76ce 10openembedded.git/recipes/ltp/ltp_20090131.bb: Mar 29 15:16:53 ltp: Fix two QA issues; .debug leak and GNU_HASH. Mar 29 15:16:53 Thanks to Koen for the hint on GNU_HASH. Mar 29 15:17:04 03Leon Woestenberg  07org.openembedded.dev * r6980c14769 10openembedded.git/contrib/dependsgraph/dependsgraph.sh: contrib/dependsgraph: Use dot instead of neato. Mar 29 15:17:06 03Leon Woestenberg  07org.openembedded.dev * r75b5212bea 10openembedded.git/recipes/images/minimal-image.bb: Mar 29 15:17:06 minimal-image: respect ANGSTROM_EXTRA_DEPENDS. Mar 29 15:17:06 Although the minimal(ist) image is meant to be, well, Mar 29 15:17:06 minimal, we should not restrict users from adding a Mar 29 15:17:08 package using ANGSTROM_EXTRA_DEPENDS. Mar 29 15:17:10 Rationale: All images are meant as a senseful base Mar 29 15:17:12 for customization, and expansion by the user. Mar 29 15:42:18 mwester, anyway boost_1.33.1 fails with the same error Mar 29 15:43:02 s/boost_1.33.1 fails/wesnoth fails to link to boost/ Mar 29 15:43:55 :( Mar 29 15:44:22 is this the intrinsics issue? Mar 29 15:44:52 Crofton|work, no it's __sync_add_and_fetch_4 there is no such function for arm so I must provide one Mar 29 15:45:11 yes, the intrinsics issue Mar 29 15:45:24 ah Mar 29 15:45:41 still locking/atomic things Mar 29 15:46:38 Crofton|work, I've armv4...what should I do Mar 29 15:47:10 https://svn.boost.org/trac/boost/ticket/2255 Mar 29 15:47:28 Crofton|work, I was told by rschuster to steal the function from GCC but this could create licences problems Mar 29 15:47:35 ok i'll look thanks a lot Mar 29 15:48:21 http://gcc.gnu.org/onlinedocs/gcc-4.1.2/gcc/Atomic-Builtins.html Mar 29 15:48:53 the problem being, gcc defines them for intel only. Mar 29 15:49:02 I had to replace one for gnuradio Mar 29 15:49:11 the patch is in the meta-data Mar 29 15:49:19 Crofton|work, yes I read that Mar 29 15:49:36 Crofton|work, ok thanks a lot I'll look at gnuradio patches Mar 29 15:49:42 what arch are you buiulding for? Mar 29 15:49:48 this is for boost Mar 29 15:49:57 armv4 Mar 29 15:50:01 armv6 and up will use ldrex/strex Mar 29 15:50:02 armv4t Mar 29 15:50:10 earlier need swp I think Mar 29 15:50:15 yes Mar 29 15:51:21 there are swp,swpb,etc... Mar 29 15:54:01 Crofton|work, I only know spim(a mips emulator) assembly...if I replace ldrex and strex by swp will it work? Mar 29 15:54:38 * Gnutoo will look at the arm manual Mar 29 15:54:46 no :) Mar 29 15:55:48 ah ldrex and ltrex are load and store instructions...and swp are swap instrucions (both atomic) Mar 29 15:55:52 ouch... Mar 29 15:56:09 ldrex/strex do not lock as much stuff Mar 29 16:00:09 is there a quick tutorial for arm assembly for someone that knows spim? Mar 29 16:00:44 (I heard that arm is more complicated than mips) Mar 29 16:04:27 03Frans Meulenbroeks  07org.openembedded.dev * r02f5b72c6a 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Mar 29 16:04:38 03Frans Meulenbroeks  07org.openembedded.dev * rb573614a7e 10openembedded.git/recipes/linux/ (linux-omap-2.6.28/mmctiming.patch linux-omap_2.6.28.bb): Mar 29 16:04:38 linux-omap-2.6.28: relaxed timeout values for SD cards Mar 29 16:04:38 this makes it possible to use cards which are around the border of the spec Mar 29 16:15:59 hi everybody: compiling now gnome-keyring I have: "configure: error: asn1Parser tool is not installed" ,then apt-get install libtasn1-3-bin , and now works Mar 29 16:46:09 03Leon Woestenberg  07org.openembedded.dev * r4079b96a95 10openembedded.git/recipes/vlc/x264_r2245.bb: x264: Remove do_configure override. Mar 29 16:50:51 Hi, I am still learning how to use oe on my embedded arm920/at91 board. I can't find any way to remove opkg from image creation. Can you point me to the right direction? Mar 29 17:03:08 Flyser_: opkg remove nameofpackage Mar 29 17:04:36 if you want not to compile search inside your recipe: which recipe is? Mar 29 17:08:41 no, I don't want opkg in the generated image Mar 29 17:08:57 (I don't want udev either, but thats another story) Mar 29 17:09:57 and by opkg I mean the application called opkg not .ipk files or something else Mar 29 17:10:03 what distro? Mar 29 17:10:27 my own (very similar to the "minimal" distro) Mar 29 17:10:31 ok Mar 29 17:10:39 not sure then Mar 29 17:10:46 I was going to suggest minimal distro Mar 29 17:12:47 what would happen when I set PREFERRED_PROVIDER_opkg = "" Mar 29 17:17:08 okay .. PREFERRED_PROVIDER_opkg = "" does not seem to work Mar 29 17:30:12 hi XorA Mar 29 17:33:58 XorA: :) Mar 29 17:43:19 Flyser_: you don't want the opkg mechanism Mar 29 17:43:22 ok. Mar 29 17:43:54 recalcati_oe: right Mar 29 17:44:17 somwhere in your recipe there will be a reference to openembedded/recipes/opkg . Mar 29 17:44:24 maybe tak.base or similar Mar 29 17:46:50 grep opkg ../../../../../openembedded/recipes -r | grep DEPEN Mar 29 17:46:58 grep opkg ../../../../../openembedded/recipes -r | grep require Mar 29 17:47:05 or I don't know Mar 29 17:47:21 grep opkg openembedded/recipes -r | grep DEPEN Mar 29 17:47:29 the path is better like this Mar 29 17:47:43 morning Mar 29 17:48:18 anybody knows how to compile qt4 embedded demos and exmaples ? I have qt4 embedded working but without demos Mar 29 17:48:24 hrw: morning Mar 29 17:48:34 or better good evening Mar 29 17:48:42 ~ugt Mar 29 17:48:43 from memory, ugt is Universal Greeting Time. Created in #mipslinux, it is a rule that states that whenever somebody enters an IRC channel it is always morning, and it is always late when the person leaves. The local time of any other people in the channel, including the greeter, is irrelevant. http://www.total-knowledge.com/~ilya/mips/ugt.html Mar 29 17:48:51 it does not list packages that are installed in my image Mar 29 17:49:27 ibot: interesting .. I didn't know Mar 29 17:51:14 hi Mar 29 17:52:02 no Flyser_ you have to look where you compile Mar 29 17:52:04 I'm messing with boost recipes... I see python __anonymous () { ... there, is that just a block of python that gets automatically executed? Mar 29 17:52:11 upon parting the recipe? Mar 29 17:52:14 parSing Mar 29 17:53:17 recalcati_oe: what do you mean? its not like grep didn't show anything at all, but nothing that seems to be related to my distro/machine/image Mar 29 17:58:59 cd openembedded Mar 29 17:59:10 i know Mar 29 17:59:15 grep opkg . -r | grep DEPEN Mar 29 17:59:23 and try not to compile it Mar 29 18:00:04 dent: boost... scary thing Mar 29 18:00:10 03Koen Kooi  07org.openembedded.dev * rf3cc636c6d 10openembedded.git/recipes/gnome/gnome-keyring_2.26.0.bb: gnome-keyring: depend in libtasn1-native to get access to needed binaries Mar 29 18:00:21 03Koen Kooi  07org.openembedded.dev * rdfd310fbb6 10openembedded.git/recipes/gnutls/libtasn1-native_1.7.bb: libtasn1: add native version Mar 29 18:02:28 recalcati_oe: oh right ONLINE_PACKAGE_MANAGEMENT = "none" did the trick. thanks very much :-) Mar 29 18:03:49 donation time in angstrom-devel :) Mar 29 18:04:08 hrw: yeah Mar 29 18:04:12 Jay7: tosa one you mean? Mar 29 18:04:13 Bah, stupid ipk trick :( Mar 29 18:04:14 bluelighting donates his h**** Mar 29 18:04:22 ah.. and h36xx/38xx Mar 29 18:04:25 I wonder if ipk packages up a full ltp package now Mar 29 18:04:30 and... I'm a noob as for OE, but it seems to me there's quite a mess in the boost recipes :-/ Mar 29 18:04:33 deb/rpm are annoying me with other issues Mar 29 18:04:34 Jay7: he offered them long time ago Mar 29 18:04:51 dent: boost is nightmare to get it buil Mar 29 18:04:52 t Mar 29 18:04:54 hrw, hey, was it python-sqlite-native you added a recipe for, to fix my problem before? Mar 29 18:05:03 hrw: seems we need some database of donated devices/accessories on the web somewhere Mar 29 18:05:11 hrw: maybe... but the current recipe doesn't seem to me to make it any easier :-/ Mar 29 18:05:13 may be in wiki for first time Mar 29 18:05:18 Tartarus: yes, but it failed for you so I did not pushed it and dropped somewhere Mar 29 18:05:24 Jay7: wiki->ProjectDevices? Mar 29 18:05:26 hrw, still have it around? Mar 29 18:05:30 I wanna try it again, heh Mar 29 18:05:31 Tartarus: have to check Mar 29 18:06:31 Tartarus: http://cgit.openembedded.net/cgit.cgi?url=openembedded/commit/&h=hrw/poky-merge/add-rpm-packages-support&id=1ed68c0f560fa017e15e1d08b3a827c0cd76b90e Mar 29 18:06:53 thanks Mar 29 18:07:11 np Mar 29 18:07:31 Tartarus: I hope that you will get it working. I prefer to not spend time on rpm Mar 29 18:08:10 hm... beginner's question: I'm fiddling with a recipe. Is it ok if I adapt it, then run bitbake package -f -c compile ? Mar 29 18:08:42 in log it seems like I see previous values for variables... (?) Mar 29 18:08:52 as if the change wasn't taken into account Mar 29 18:12:08 dent: yes, it is good way Mar 29 18:12:58 oki, I see the problem... I'm editing another copy of the file, in a different tree :)) sigh! Mar 29 18:13:25 hrw: and as for python __anonymous () {} - those get executed upon parsing? Mar 29 18:14:28 yes Mar 29 18:14:50 thx Mar 29 18:15:10 so, if I do a print "YEAH" in such a block... do I see it anywhere? Mar 29 18:15:25 it seems to me the block for some reason is not executed; but probably just some local problem Mar 29 18:18:59 bye guys Mar 29 18:19:20 bye Mar 29 18:20:49 if all the depndencies are present, is there a difference if I do bitbake -b recipe.bb -f -c compile vs. bitbake package -f -c compile ? Mar 29 18:21:56 hi Mar 29 18:22:01 i'm newbie Mar 29 18:22:20 looking for started intro materials Mar 29 18:22:22 help :( Mar 29 18:23:19 I suggest poky docs: http://pokylinux.org/releases/pinky-3.1/doc/poky-handbook.html Mar 29 18:23:44 it's poky-focused, but as that is just OE with some extra "tuning", I found it ok Mar 29 18:25:49 03Koen Kooi  07org.openembedded.dev * r777c140275 10openembedded.git/recipes/dsplink/ (ti-dmai_svn.bb ti-paths.inc): Mar 29 18:25:49 dsplink: add support for all omap3 boards using the 'armv7a' override Mar 29 18:25:49 * needs revisiting when non-omap3 armv7a TIi SoCs become available Mar 29 18:26:07 hi dent Mar 29 18:26:16 do i need device to test out the examples there Mar 29 18:26:19 (I'm a newbie too, btw...) Mar 29 18:26:28 i'm newbie but very enthusiatic to learn embedded programming Mar 29 18:26:43 nope... + if you grab the real poky, it's easy to set up quemu simulation Mar 29 18:27:24 quemu? Mar 29 18:27:28 more details please Mar 29 18:27:34 :) Mar 29 18:27:42 what is quemu? Mar 29 18:27:55 http://en.wikipedia.org/wiki/QEMU Mar 29 18:29:02 i see Mar 29 18:29:03 it can emulate various archs (inc. x86 and arm) - you can build an image in OE and have quemu boot into that Mar 29 18:29:14 build an image? Mar 29 18:29:16 sorry? Mar 29 18:29:19 i.e. you can experiemtn with OE even without a real device Mar 29 18:29:31 hm, you're more newbie than I am :) Mar 29 18:29:33 fricking cool Mar 29 18:29:45 does it include parallel port and so on? Mar 29 18:30:27 don't know what all hw is emulated... Mar 29 18:30:37 it certainly has network :) Mar 29 18:31:31 what os do i need to experiment with that thingy? Mar 29 18:32:43 man it's all on the web, just read :-/ Mar 29 18:33:15 but basically linux would do fine for all of that ^, maybe other unix-like systems would do too Mar 29 18:35:35 i read thru it Mar 29 18:35:41 it works with ubuntu Mar 29 18:35:43 but it's not C Mar 29 18:35:51 linux device driver should all be in c? Mar 29 18:36:19 I don't think I follow... Mar 29 18:36:58 device driver is basically calling functions from kernel header file Mar 29 18:37:22 OE is not about kernel drivers... Mar 29 18:39:12 then how does it talk with device? Mar 29 18:40:04 http://en.wikipedia.org/wiki/Openembedded Mar 29 18:40:11 first read what OE actually is... Mar 29 18:40:55 i see. Mar 29 18:41:01 any linux device driver channel? Mar 29 19:59:58 Hello! bitbake minimal-image Mar 29 19:59:58 NOTE: Out of date cache found, rebuilding... Mar 29 19:59:58 ERROR: opening /home/s/my/oe/openembedded/packages/*/*.bb: [Errno 2] No such file or directory: '/home/s/my/oe/openembedded/packages/*/*.bb' Mar 29 19:59:58 ERROR: Nothing PROVIDES 'minimal-image' Mar 29 20:03:42 fpga: ls -l /home/s/my/oe Mar 29 20:04:52 fpga: either you forgot to git clone the openembedded repository or its in the wrong place Mar 29 20:05:22 drwxr-xr-x 9 s s 4096 2009-01-31 22:29 bitbake Mar 29 20:05:22 drwxr-xr-x 11 s s 4096 2009-03-07 03:52 build Mar 29 20:05:22 -rwxr--r-- 1 s s 97 2009-01-31 22:34 oe Mar 29 20:05:22 drwxr-xr-x 11 s s 4096 2009-03-29 23:52 openembedded Mar 29 20:05:22 drwxr-xr-x 4 s s 40960 2009-03-08 02:51 sources Mar 29 20:05:23 drwxr-xr-x 3 s s 4096 2009-03-01 23:04 tmp Mar 29 20:05:47 s@sh:~/my/oe/openembedded$ git pull Mar 29 20:05:47 Already up-to-date. Mar 29 20:06:23 and whats in there (pls don't paste more than ~3 lines) Mar 29 20:06:31 sorry Mar 29 20:08:44 there are really none packages dir!.. Mar 29 20:08:57 git clone solves my problem? Mar 29 20:11:17 uhm I don't understand your last sentence. sry^^ Mar 29 20:11:34 that is "there are really none packages dir" Mar 29 20:12:32 packages dir does not exists Mar 29 20:13:09 there is no packages dir Mar 29 20:13:16 yes Mar 29 20:13:16 "lasses conf contrib COPYING.MIT docs files MAINTAINERS README recipes removal.txt site" Mar 29 20:13:24 classes* Mar 29 20:13:34 classes - exists Mar 29 20:13:48 but why it want packages dir?.. Mar 29 20:14:06 oh thats right^^ Mar 29 20:14:25 your BBPATH env var is wrong I guess? Mar 29 20:16:01 I did not updated my openmbedded for 15 days and now I just git pull and bitbake minimal-image - strange... Mar 29 20:16:13 before this day my BBPATH was OK.. Mar 29 20:16:26 oh ... no idea then Mar 29 20:16:40 export BBPATH=/home/s/my/oe/openembedded Mar 29 20:16:51 looks okay to me Mar 29 20:16:55 thank for try to help... Mar 29 20:16:56 :) Mar 29 20:17:08 ^^ Mar 29 20:17:13 I deleted openembedded and performing clean pull Mar 29 20:17:28 *clone... Mar 29 20:17:57 good luck :) Mar 29 20:18:22 thanks :) Mar 29 21:01:14 Crofton|work, hi again...I've undersood nearly all the bits of the patch( http://rafb.net/p/F0LS0m34.html )...but I don't understand asm volatile ("1: \n\t" and specially the 1: ...is it specific to the inline assembly...because usually you can't start a label with a number....I also don't understand the bne(does it mean branch not equal) 1b...does it jump to 1 if the condition(is it the result of the prev Mar 29 21:01:15 ious cmp...it is said that it produce a condition)it's not equal Mar 29 21:02:01 (It didn't took so much but I ate and played tremulous) Mar 29 21:02:49 woglinde, hi Mar 29 21:03:18 hi gnutoo Mar 29 21:03:34 woglinde, do you know inline arm assembly? Mar 29 21:03:42 ah little bit Mar 29 21:04:05 woglinde, I've not understood the 1: and the bne 1b in http://rafb.net/p/F0LS0m34.html Mar 29 21:04:28 is 1: a label? if so why does it start by a number? Mar 29 21:04:31 ahah Mar 29 21:04:33 atomics again Mar 29 21:04:37 yes lol Mar 29 21:05:14 and does bne 1b jump to 1: if the result of the cmp is not 0? Mar 29 21:05:21 I think so Mar 29 21:05:26 ok thanks Mar 29 21:05:32 Crofton|work, for both? Mar 29 21:05:38 we have to see the whole file Mar 29 21:05:42 s/both/both questions/ Mar 29 21:05:47 but 1b seems a lable somwehre Mar 29 21:05:51 ok i'll pastebin it Mar 29 21:05:58 ok Mar 29 21:06:01 hm the code is stolen from libatomics anyway Mar 29 21:06:12 ok Mar 29 21:06:27 and 1_36 is old Mar 29 21:06:30 of boost Mar 29 21:06:49 woglinde, what is this libatomics? and can I steal code that makes me the same thing but with swp? Mar 29 21:07:19 libatomics is standalone or part of boehm garbage collector Mar 29 21:07:28 woglinde, should I update boost to 1.38? Mar 29 21:07:38 but newer version wit armfixes is in boehm-gc 7.1 Mar 29 21:07:50 ok thanks Mar 29 21:08:06 the cacao has a version in there mecurial Mar 29 21:08:31 ok thanks a lot Mar 29 21:08:50 http://mips.complang.tuwien.ac.at/hg/cacao/file/e64f89ac43c8/src/mm/boehm-gc/libatomic_ops-1.2/src/atomic_ops/sysdeps/gcc/arm.h Mar 29 21:09:02 thanks Mar 29 21:10:02 thanks a lot... Mar 29 21:10:29 you might want to read this article too http://lwn.net/Articles/314144/ Mar 29 21:12:21 ok I'll look at the article Mar 29 21:14:30 hm the comments Mar 29 21:24:58 yes...I'm reading http://0pointer.de/blog/projects/atomic-rt.html Mar 29 21:25:36 dont forget the comments from npitre Mar 29 21:25:42 about this Mar 29 21:27:08 ok Mar 29 21:33:56 03Andrea Adami  07org.openembedded.dev * racc50bd8c1 10openembedded.git/conf/machine/include/zaurus-2.6.inc: zaurus-2.6.inc: fix for KERNEL_IMAGETYPE (and allow uImage) Mar 29 21:34:07 03Andrea Adami  07org.openembedded.dev * ra1d8d7d470 10openembedded.git/conf/machine/spitz.conf: spitz.conf: set UBOOT_MACHINE_c7x0 = "akita_config" Mar 29 21:34:07 03Andrea Adami  07org.openembedded.dev * r7d75581c67 10openembedded.git/conf/machine/c7x0.conf: c7x0.conf: set UBOOT_MACHINE_c7x0 = "corgi_config" Mar 29 21:36:12 03Andrea Adami  07org.openembedded.dev * r526669b8a3 10openembedded.git/conf/machine/spitz.conf: spitz.conf: fix typo Mar 29 21:49:51 thanks a lot and bye...i'll try to do a patch tomorrow Mar 29 21:50:06 03Andrea Adami  07org.openembedded.dev * r809a403054 10openembedded.git/conf/machine/ (akita.conf c7x0.conf spitz.conf): zaurus-clamshells: add mtdparts info in $machine.conf Mar 29 21:57:28 03Andrea Adami  07org.openembedded.dev * r2af4308023 10openembedded.git/conf/machine/ (akita.conf c7x0.conf spitz.conf): zaurus-clamshells: minor cleanings (machine-overrides unneeded) Mar 29 21:57:39 gi tlog Mar 29 21:57:54 hmpf.. Mar 29 22:04:25 03Andrea Adami  07org.openembedded.dev * r7aa40e2922 10openembedded.git/conf/machine/spitz.conf: spitz.conf: add note abour max rootfs size Mar 29 22:21:14 03Andrea Adami  07org.openembedded.dev * r1b900d4896 10openembedded.git/conf/machine/include/zaurus-clamshell.inc: zaurus-clamshell.inc: add UBOOT_ENTRYPOINT and UBOOT_ADDRESS Mar 29 22:26:36 http://www.angstrom-distribution.org/unstable/sources/embryo_svn.enlightenment.org_.svn.e.trunk_39786_.tar.gz is not present Mar 29 22:51:49 * Jay7 -> sleep Mar 29 22:55:55 gn Mar 30 01:39:12 Hi ... I'm trying to define a new machine type and bitbake doesn't seem to be picking up PREFERRED_PROVIDER_virtual/kernel for some reason Mar 30 01:39:23 when I run bitbake -i, I see that it is defined Mar 30 01:39:38 but when I do 'which virtual/kernel' it doesn't seem to be using the one I set Mar 30 01:39:58 and it complains when I go to build that the preferred provider isn't set Mar 30 01:43:26 Did you use linux_2.6.SOMETHING.bb ? Mar 30 01:43:36 and if so, did you add DEFAULT_PREFERENCE_yournewmachine = "1" to it? Mar 30 01:43:50 actually I have a separate kernel recipe Mar 30 01:44:29 I'm trying to modify a distribution/machine combo that was given to me Mar 30 01:45:37 the kernel recipe is linux-ipac9302-2.6.22.bb Mar 30 01:46:09 the strange thing is that if I set MACHINE="ipac9302", it picks up the right kernel recipe Mar 30 01:46:27 but if I rename conf/machine/ipac9302.conf to conf/machine/somethingelse.conf and set MACHINE="somethingelse" Mar 30 01:47:01 it no longer gets the right provider, even though PREFERRED_PROVIDER_virtual/kernel is still set to "linux-ipac9302" in that config file Mar 30 01:58:43 oh. I see, it's because of the COMPATIBLE_MACHINE variable in linux-ipac9302-2.6.22.bb Mar 30 01:58:46 ok Mar 30 01:59:52 removing that variable makes it work again ... great **** ENDING LOGGING AT Mon Mar 30 02:59:57 2009