**** BEGIN LOGGING AT Tue Aug 07 02:59:57 2007 Aug 07 06:47:42 hi I am munging some packaging and running into dependency problems Aug 07 06:47:59 it says This is probably a circular dependency (the chain might be printed above). Aug 07 06:48:30 now the message above says only package numbers how can I associate that to a package name Aug 07 06:49:15 ERROR: Task 14 (/local/sandbox/openembedded/org.openembedded.dev/packages/shasum/shasum-native.bb, do_install) (depends: Set([13])) Aug 07 06:49:30 now how do I figure which is task 13 and which is task 14 Aug 07 06:52:32 morning Aug 07 06:52:44 hrw: mornig Aug 07 06:52:59 hrw: can you help me out in the problem I just posted above Aug 07 06:53:07 khem: did you use DEPENDS/RDEPENDS/DISTRO_EXTRA_* vars in local.conf? Aug 07 06:53:46 hrw: no Aug 07 06:53:57 hrw: but I am adding eglibc package Aug 07 06:54:10 so I added eglibc-initial_svn.bb Aug 07 06:54:37 pastebin it then Aug 07 06:54:45 ok Aug 07 06:55:04 khem: koen already added eglibc-initial (untested) - that one you use? Aug 07 06:55:05 bb is not giving me which packages are having circular dep Aug 07 06:55:12 yes thats one I am using Aug 07 06:55:43 let me update Aug 07 06:55:52 but it needed some fixes and I figured that the toolchain for NPTL can be built is little differently Aug 07 06:55:56 which I know for sure Aug 07 06:56:07 so I wanted to give my sequence a shot Aug 07 06:56:51 you can always use "-b" switch to bitbake but this ignore deps and rquire knowledge Aug 07 06:57:40 hah no I want to solve the dep issue Aug 07 06:59:00 let me check Aug 07 06:59:42 * khem tries to upload on paste.debian.net Aug 07 07:03:44 hmmm something wrong with my uplink Aug 07 07:14:05 hrw: http://rafb.net/p/8ykQYI24.html Aug 07 07:20:40 khem: add -DDD to bitbake and look at output - you will get info which providers are chosen Aug 07 07:26:59 hrw: How can I associate task numbers Aug 07 07:27:14 hrw: it does not have an association table or something Aug 07 07:27:27 ERROR: Task 14 (/local/sandbox/openembedded/org.openembedded.dev/packages/shasum/shasum-native.bb, do_install) (depends: Set([13])) Aug 07 07:27:40 now how do I find out what task 14 is Aug 07 07:27:53 khem: if shasum even has problem then you have totoally wrong deps Aug 07 07:27:53 rather task 13 Aug 07 07:28:05 hmmm Aug 07 07:28:09 khem: here I do not have that problem - ANGSTROM_MODE = "eglibc" Aug 07 07:28:21 hrw: OK Aug 07 07:28:27 but I modified it a bit Aug 07 07:28:39 eglibc-intermediate needs adding Aug 07 07:28:41 if not modified than it does not Aug 07 07:28:49 then how you modded? Aug 07 07:29:04 hrw: actually I added a step in eglibc-initial Aug 07 07:29:17 show me diff Aug 07 07:29:54 do_stage () { Aug 07 07:29:54 - oe_runmake cross-compiling=yes install_root=${CROSS_DIR}/${TARGET_SYS} prefix="" install-headers Aug 07 07:29:54 - Aug 07 07:29:54 + oe_runmake cross-compiling=yes install_root=${CROSS_DIR}/${TARGET_SYS} prefix="" csu/subdir_lib install-headers Aug 07 07:29:54 + cp ${S}/csu/crt1.o ${S}/csu/crti.o ${S}/csu/crtn.o ${CROSS_DIR}/${TARGET_SYS}/usr/lib Aug 07 07:29:56 + ${CC} -nostdlib -nostartfiles -shared -x c /dev/null -o ${CROSS_DIR}/${TARGET_SYS}/usr/lib/libc.so Aug 07 07:30:53 hrw: what do glibc-initial and glibc-intermediate do Aug 07 07:31:05 that should not be problem.. Aug 07 07:31:05 I think I know glibc-initial is prolly just glibc-headers Aug 07 07:31:19 khem: to bootstrap toolchain you need them Aug 07 07:31:21 but I dont know about glibc-intermediate Aug 07 07:31:35 itnermediate is needed for ntpl Aug 07 07:32:36 hrw: so its not final glibc but some sort of intermediate stuff Aug 07 07:32:48 do you know why it was needed Aug 07 07:33:13 I build this way Aug 07 07:33:35 http://source.mvista.com/~kraj/toolchain-eglibc.sh Aug 07 07:34:19 for NPTL toolchains same way I did in crosstools too Aug 07 07:34:55 khem: you also work for montavista? (like Chris Larson) Aug 07 07:34:57 hrw: what does this mean Aug 07 07:34:59 -PREFERRED_PROVIDER_virtual/arm-angstrom-linux-gnueabi-libc-for-gcc Aug 07 07:35:19 hrw: Yes I do someone need to pay my bills Aug 07 07:35:49 virtual/arm-angstrom-linux-gnueabi-libc-for-gcc is glibc-intermediate Aug 07 07:36:25 khem: I am not so famiiar with building toolchains - always used OE for it Aug 07 07:36:45 hrw ok :) I have done is quite a bit Aug 07 07:37:01 and thats what I did initially for montavista too Aug 07 07:37:16 hrw: how do I discard my changeset in mtn Aug 07 07:37:20 khem: we do: glibc-initial -> gcc-initial -> glibc-intermediate -> gcc-cross -> glibc Aug 07 07:37:24 mtn disapprove Aug 07 07:38:15 03koen 07org.oe.dev * r2029e85b... 10/ (1 packages/glibc/eglibc-intermediate_svn.bb): eglibc-intermediate: add completely untested eglibc-intermediate Aug 07 07:38:49 binutils-cross->cross-gcc-initial->glibc-initial->cross-gcc-intermediate->glibc->cross-gcc Aug 07 07:38:53 this is what I do Aug 07 07:39:00 actually I would add another step Aug 07 07:39:37 binutils-cross->cross-gcc-initial->glibc-initial->cross-gcc-intermediate->glibc-intermediate->cross-gcc->glibc Aug 07 07:40:36 that way we get the unwind info straightened out Aug 07 07:40:48 currently I think OE must be faking libgcc Aug 07 07:42:07 as I told I am not familiar with toolchain build sequence Aug 07 07:43:08 moin Aug 07 07:45:07 hrw: but you helped me a lot :) Aug 07 07:48:07 ;) Aug 07 07:49:11 khem: you are from SF area? Aug 07 07:50:35 good morning all Aug 07 07:51:21 hrw: yes Aug 07 07:51:31 koen: moin Aug 07 07:51:44 koen: I was trying eglibc Aug 07 07:51:57 figured out that eglibc-intermediate is missing Aug 07 07:52:37 koen: then I went down the rat hole while trying ot teach OE my way of building toolchain Aug 07 07:53:14 good morning all Aug 07 07:54:09 * hrw -> post office Aug 07 07:54:18 hi Stelios Aug 07 07:56:09 hrw|gone : hi ! Going to pickup the package ? Aug 07 07:56:18 koen: what does glibc-intermediate step do exactly Aug 07 08:06:14 afaik build a libc so gcc doesn't barf on missing thread support Aug 07 08:06:26 ah i see Aug 07 08:06:58 I made local changes which I want to discard mtn disapprove does not do it Aug 07 08:07:03 is there any other way Aug 07 08:07:40 good morning Aug 07 08:07:46 something like mtn discard Aug 07 08:08:35 how can I supress the merge of a task if I have more then one PATH for BBFILES? Aug 07 08:08:38 mtn revert Aug 07 08:11:28 morning all Aug 07 08:12:10 or is this not intended? Aug 07 08:15:25 zarath: It shouldn't merge them (and doesn't here for me) Aug 07 08:15:58 it always merges them here Aug 07 08:17:00 if I look in the dependecies (bitbake -g) I see references to both bb files Aug 07 08:22:18 zarath: Building both is not merging. If they provide different things its perfectly possible for that to happen Aug 07 08:24:20 ok, but what can i do if i don't want all the things that the other task provides? Aug 07 08:29:59 steliosk: alix arrived Aug 07 08:30:50 steliosk: its TINY Aug 07 08:35:50 morning Aug 07 08:36:06 hey XorA Aug 07 08:36:35 zarath: Then don't build things that depend on them Aug 07 08:36:53 zarath: Move the other file out the way and you'll find out why it wants to build it Aug 07 08:37:44 zarath: It is not a bitbake problem though, its doing exactly what its supposed to by the sounds of it Aug 07 08:38:43 so there is comming up a dependency back to the task from an underlying package... Aug 07 08:38:53 zarath: right Aug 07 08:39:18 hrw : hehe. all new babies are :) Aug 07 08:40:52 steliosk: one thing PC Engines should consider for future - rear ATX panel for it so it will be easier to use in miniitx cases Aug 07 08:45:16 RP: what is the priority if I have 2 BBFILE pathes Aug 07 08:46:30 03koen 07org.oe.dev * r7f1ed839... 10/ (1 conf/distro/angstrom-2007.1.conf): angstrom: move include so it can override more parts Aug 07 08:53:16 koen: ping Aug 07 08:53:37 koen: I think eglibc-initial should be built before eglibc-intermediate right ? Aug 07 08:55:04 zarath: Whatever priority you define to those paths Aug 07 08:57:11 is it like in shell evaluation, so that the later one overides the prior? Aug 07 08:58:24 zarath: No, you use regexs to assign priorities to them Aug 07 08:59:12 zarath: Read the top of http://svn.o-hand.com/view/poky/trunk/build/conf/local.conf.sample?rev=2285&view=markup Aug 07 08:59:24 khem: yes Aug 07 08:59:45 it seems not to be called then Aug 07 09:00:10 and /gcc-cross dies Aug 07 09:00:35 local/sandbox/openembedded/build/tmp/cross/bin/arm-angstrom-linux-gnueabi-ld: crti.o: No such file: No such file or directory Aug 07 09:01:10 RP higher value -> higher priority? Aug 07 09:02:15 zarath: I think so, I can't remember for certain Aug 07 09:02:36 koen: someone should depend on gcc-cross-initial Aug 07 09:02:46 # BBFILE_PATTERN_extras = "^${OEROOT}/meta/" <- shouldn't this be # BBFILE_PATTERN_extras = "^${OEROOT}/meta-extras/" in the example? Aug 07 09:09:28 zarath: yes Aug 07 09:10:21 zarath: I've fixed that ;-) Aug 07 09:11:57 ok, one thing less to ponder on Aug 07 09:14:22 hi what does this translate to DEPENDS += "${@['virtual/${TARGET_PREFIX}libc-initial',''][bb.data.getVar('TARGET_ARCH', d, 1) in ['arm', 'armeb', 'mips', 'mipsel']]}" Aug 07 09:14:38 is it a conditional depend ? Aug 07 09:15:06 which only will be added if arch is not one of the listed Aug 07 09:20:09 ~fish XorA Aug 07 09:20:10 * ibot slaps XorA around with a large trout Aug 07 09:21:21 btw as assumed, higher priority value results in higher priority Aug 07 09:31:22 hey lrg Aug 07 09:31:39 hey koen. Aug 07 09:32:12 koen: can you email me your postal address. I will have something to send you next week. Aug 07 09:36:25 khem: yes Aug 07 09:36:40 lrg: sent Aug 07 09:37:05 RP: so for arm no glibc-initial is performed? Aug 07 09:37:52 khem: Sorry, it will be added for the arches listed I thnk Aug 07 09:38:17 ok but it doesnt seems so Aug 07 09:38:27 and that doesn't translate to glibc-initial, it translates to something like arm-linux-libc-initial Aug 07 09:38:47 * khem just figured out that do_stage is missing for eglibc Aug 07 09:43:34 koen, RP : angstrom-x11-image builds now after the latest commits Aug 07 09:44:04 koen, RP : i'll install the image to see if it works also Aug 07 09:44:38 btw eglibc works or is "under development" ? Aug 07 09:47:33 steliosk: "completely untested" Aug 07 09:49:12 RP, mickey|zzZZzz, florian: Packaged contents of hildon-1 into /data/build/koen/OE/build/tmp/angstrom/deploy/glibc/ipk/armv4t/libhildon-1-0_1.0.12+svnr13015-r0_armv4t.ipk Aug 07 09:50:12 it only needed this patch: Aug 07 09:50:14 - GTK_BUTTONBOX_CENTER); Aug 07 09:50:15 + GTK_BUTTONBOX_DEFAULT); Aug 07 09:50:49 steliosk: did you managed to boot alix over pxe? Aug 07 09:50:50 koen: nice Aug 07 09:53:40 ~lart the person that broke the htcuniversal defconfig Aug 07 09:53:41 * ibot changes the person that broke the htcuniversal defconfig's permissions to 0777 and tells the world Aug 07 09:56:58 good morning btw :-) Aug 07 10:01:25 steliosk: did you managed to boot alix over pxe? here is looks like never even trying Aug 07 10:07:35 hrw : did not try pxe. just used a CF with Oplinux for x86 i had Aug 07 10:08:12 hrw : there was a bios update, that Pascal has send me a message Aug 07 10:08:35 hrw : have i send you the link ? Aug 07 10:09:05 export LDFLAGS += "-lstdc++" Aug 07 10:09:13 versus CXXFLAGS Aug 07 10:09:26 cxxflags gets the option passed Aug 07 10:09:29 ldflags does Aug 07 10:09:30 not Aug 07 10:10:15 steliosk: you sent me link - I will update when get board booted from cf Aug 07 10:16:02 koen: how big / partition is on c700? Aug 07 10:16:30 20 megs or so Aug 07 10:21:19 Here is the problem statement: Aug 07 10:21:36 Something is fishy with c++ linking with -lstdc++ Aug 07 10:21:55 and it is different on the OSK versus the Efika .. Aug 07 10:23:52 who provides -lstdc++? Aug 07 10:24:05 gcc iirc Aug 07 10:24:45 03koen 07org.oe.dev * r64cb3a9b... 10/ (4 files in 3 dirs): osso-ic: add internet connection lib for maemo Aug 07 10:25:24 good call! Aug 07 10:25:32 http://gcc.gnu.org/libstdc++/ Aug 07 10:28:11 ~lart fedora Aug 07 10:28:11 * ibot moos at fedora Aug 07 10:30:52 ~lart defconfigs Aug 07 10:30:52 * ibot hurls dozens of incontinent, insomniac, hungry kittens with tiny little razor-sharp claws and a wide variety of contagious intestinal parasites at defconfigs Aug 07 10:37:06 anyone tried mtn 0.36 Aug 07 10:45:44 ohhhh, that's gross Aug 07 10:50:15 how to filter some packages out from DISTRO_EXTRA_RDEPENDS Aug 07 10:55:40 what is the difference between libtool-native and libtool-cross? Aug 07 10:56:09 libtool-native is for building -native packages AIUI Aug 07 10:56:28 libtool-native runs on host and produces host libraries Aug 07 10:56:33 what is strange is that I think libtool-native is used for libusb Aug 07 10:56:36 libtool-cross runs on host and produces cross libraries Aug 07 10:56:55 becasue iof the rpath-control patch Aug 07 10:57:30 god working with fedora core is difficult Aug 07 10:57:51 * Crofton|home uses Fedora Aug 07 10:58:37 * eekee stays away from all the major distros :] Aug 07 10:58:49 OE: Not hardcoding '/home/balister/oe/tmp/cross/lib/gcc/powerpc-angstrom-linux/4 Aug 07 10:58:49 .1.2/../../../../powerpc-angstrom-linux/lib/' into rpath Aug 07 10:59:01 this line is produced by rpath-contrl patch Aug 07 10:59:09 hmm, does cross include native Aug 07 10:59:37 03mickeyl 07org.oe.dev * r37353422... 10/ (3 files in 3 dirs): psplash: smooth bar corners in OpenMoko version Aug 07 10:59:42 03mickeyl 07org.oe.dev * r82da4ff9... 10/ (6 files in 2 dirs): Aug 07 10:59:42 gstreamer update: Aug 07 10:59:42 * gstreamer 0.10.14 Aug 07 10:59:42 * gst-plugins-base 0.10.14 Aug 07 10:59:42 * gst-plugins-good 0.10.6 Aug 07 10:59:43 * gst-plugins-bad 0.10.5 (doesn't compile yet) Aug 07 10:59:45 * gst-plugins-ugly 0.10.6 Aug 07 10:59:49 03mickeyl 07org.oe.dev * r00c60fb6... 10/ (5 files in 3 dirs): add dbus-c++[-native], a native C++ binding to the DBUS API Aug 07 10:59:54 03mickeyl 07org.oe.dev * r3bc048bb... 10/ (4 files in 3 dirs): gst-plugins-bad 0.10.5 fix compiling (see upstream bug #454078) Aug 07 10:59:59 03florian 07org.oe.dev * rdb0c4ff9... 10/ (1 packages/tasks/task-gpe-base.bb): task-gpe-base: fix RDEPENDS Aug 07 11:00:04 03florian 07org.oe.dev * r614f3cd9... 10/ (1 packages/gpesyncd/gpesyncd_2.0.bb): gpesyncd: Add 2.0 Aug 07 11:01:13 something is fishy in libtool land Aug 07 11:01:20 wouldn't be the first time Aug 07 11:01:25 * Crofton|home grins Aug 07 11:01:37 I swear -native is running when building libusb Aug 07 11:01:42 not -cross Aug 07 11:02:01 unless somethig happens I do not understand Aug 07 11:02:49 Crofotn|home : That's really strange. I just build it for Efika Aug 07 11:02:57 ~hail qemu Aug 07 11:02:58 * ibot bows down to qemu and chants, "I'M NOT WORTHY!!" Aug 07 11:03:09 * hrw booted visage-linux in qemu to prepare cf card for alix Aug 07 11:03:15 steliosk, with libtool 1.5.10? Aug 07 11:03:35 Crofton|home : let me check. Aug 07 11:03:37 that also works for me, I am testing 1.5.24 Aug 07 11:03:51 and I found a libtool-native patch not in 24 Aug 07 11:04:10 but the patch is for native, but the message appears in the libusb build log Aug 07 11:04:25 I am going to update the 1.5.24 stuff with the patch Aug 07 11:04:51 03koen 07org.oe.dev * ra9c56872... 10/ (1 packages/gtkhtml/gtkhtml.inc): gtkhtml: fix QA issues Aug 07 11:04:56 yeah i am using 1.5.10 Aug 07 11:06:30 I need 1.5.24 for another package Aug 07 11:06:44 and, I really think we should move forward Aug 07 11:06:55 ala pkgconfig and autoconf Aug 07 11:07:52 * Crofton|home tests mtn 0.36 Aug 07 11:11:08 how to filter some packages out from DISTRO_EXTRA_RDEPENDS? i don't want avahi be compiled in my images Aug 07 11:14:31 steliosk: booted Aug 07 11:18:46 * hrw needs to connect few leds to board but not today Aug 07 11:19:02 readon: just remove it from DISTRO_EXTRA_RDEPENDS locally. Aug 07 11:19:09 I think avahi should go into task-something, eventually. task-base-extended, perhaps? Aug 07 11:19:41 its in task-distro-base now Aug 07 11:20:52 hrw: yes, through ${DISTRO_EXTRA_RDEPENDS} Aug 07 11:21:59 I'd prefer to move it directly into task-base.bb somewhere. Aug 07 11:22:44 pH5: no Aug 07 11:23:15 pH5: if I want to build for my own distro-which-do-not-like-avahi I want to not have to remove it from something Aug 07 11:23:44 hrw: then let's have a distro feature "dns-sd" or "zeroconf" or something like that Aug 07 11:28:55 im tring compile e-wm but i got this error Aug 07 11:28:59 in makedepend-native Aug 07 11:29:00 make[1]: *** No rule to make target `makedepend.@APP_MAN_SUFFIX@', needed by `all-am'. Stop. Aug 07 11:29:45 psycoman: can you build xorg-macros-native and then rebuild makedepend-native? Aug 07 11:29:54 if it helps then give sign Aug 07 11:30:30 hmm could it be that the bitbake 1.6.8 git fetcher doesn't know about 'tag=' yet ? Aug 07 11:34:00 03koen 07org.oe.dev * r072d3a52... 10/ (1 packages/maemo3/libhildonhelp_1.9.1.bb): libhildonhelp: add 1.9.1 Aug 07 11:34:05 03koen 07org.oe.dev * r4731d848... 10/ (1 packages/maemo3/libosso-help_2.1.2.bb): libosso-help: add 2.1.2 Aug 07 11:34:11 03koen 07org.oe.dev * r4852f621... 10/ (1 packages/maemo3/hildon-libs_0.15.1.bb): hildon-libs: add 0.15.1 (from launchpad, the nokia site is too hard to navigate) Aug 07 11:34:17 03koen 07org.oe.dev * r8fef55ee... 10/ (1 packages/maemo3/osso-thumbnail_0.7.bb): osso-thumbnail: add 0.7 Aug 07 11:35:17 hrw: ERROR: No providers of build target xorg-macros-native (for []) Aug 07 11:38:06 * Crofton|home rebuilds libtool-native and hopes his problems magically disappear Aug 07 11:41:34 ops Aug 07 11:45:46 mickeyl: The git fetcher knew about tag from the beginning Aug 07 11:46:50 Crofton|home: still have problem im makedepend Aug 07 11:47:02 s/im/in Aug 07 11:47:09 heh Aug 07 11:47:17 that was a while back, having other problems atm Aug 07 11:47:32 heh :) Aug 07 11:47:35 libusb on Efika with new libtool version Aug 07 11:47:50 i see now :(, then... Aug 07 11:49:33 now to rebuild everything again to see if this helps Aug 07 11:51:49 RP: interesting. then it's something different that breaks. probably our svnnow patch Aug 07 11:54:33 koen: Did the changes in bitbake trunk resolve those sqlite locking issues btw? Aug 07 12:04:06 i have to open a bug fo makedepends problem or clean-up my satagin ? Aug 07 12:04:28 s/stagin/stagin Aug 07 12:04:58 RP: yes! Aug 07 12:05:14 koen: Great :) Aug 07 12:05:42 RP: http://rafb.net/p/WTGoH137.html <- scary Aug 07 12:06:50 koen: Scary indeed. Its building though? Aug 07 12:06:57 yes Aug 07 12:08:05 03koen 07org.oe.dev * r55e79a8c... 10/ (1 packages/maemo3/osso-gwconnect_1.0.8.bb): osso-gwconnect: add 1.0.8 Aug 07 12:08:11 03koen 07org.oe.dev * r3b28085c... 10/ (1 packages/maemo3/libhildonmime_1.9.5.bb): libhildonmime: add 1.9.5 Aug 07 12:08:16 another etch->sid upgrade.. Aug 07 12:08:16 03koen 07org.oe.dev * ra32f961c... 10/ (1 packages/maemo3/mce-dev_1.5.6.bb): mce-dev: add 1.5.6 Aug 07 12:10:45 RP: everything upto hildon-fm builds against a stock gtk Aug 07 12:11:04 hildon-fm needs the tap-n-hold + filechooser patches Aug 07 12:13:21 koen: so when angstrom-hildon-image? Aug 07 12:13:36 hrw: when they fix that $&*@)$*@) hildon-fm Aug 07 12:20:57 koen: hmm. Its better than it was I guess Aug 07 12:38:29 koen: I will add linux_2.6.22.bb (def_pref=-1 for all !alix) Aug 07 12:38:44 NOTE: multiple providers are available for virtual/i686-angstrom-linux-libc-initial (glibc-initial, eglibc-initial); Aug 07 12:38:47 NOTE: consider defining PREFERRED_PROVIDER_virtual/i686-angstrom-linux-libc-initial Aug 07 12:38:59 hrw: hold on, let me commit mine first :) Aug 07 12:39:41 koen: linux_2.6.22? ok Aug 07 12:46:34 how do I switch VTs on a spitz? Aug 07 12:48:44 eekee: chvt 1/2/3/4/5/6 etc. Aug 07 12:48:52 or fn-left/right, iirc Aug 07 12:49:04 ah great ty Aug 07 12:49:04 hrw: could you merge in http://rafb.net/p/K5BFpp42.html into your .bb? Aug 07 12:49:25 my 2.6.22 recipe is in a tree that I can't push to mtn.oe.org :( Aug 07 12:52:41 koen: ok Aug 07 12:53:13 koen: Hi, I have a little patch for at91 Xserver here : http://rafb.net/p/YYgi8a34.html Aug 07 12:54:15 koen: Hi, I have a tiny patch for Xserver on AT91 here : http://rafb.net/p/YYgi8a34.html Aug 07 12:54:15 do you whant me to send an email on the mailing list ? Aug 07 12:54:28 please attach it to bugzilla Aug 07 12:54:42 oops sory for double post :-( Aug 07 12:54:50 noglitch: and xserver-common needs to get patched as well Aug 07 12:59:21 koen: bugzilla ok, new bug ? Aug 07 12:59:43 yes Aug 07 13:02:13 koen: xserver-common ? where ? Aug 07 13:04:58 noglitch: gpe... i'll apply it Aug 07 13:12:52 koen: pushed Aug 07 13:15:55 steliosk: http://www.openembedded.org/projectdevices Aug 07 13:16:08 03hrw 07org.oe.dev * r0eb49adf... 10/ (6 files in 5 dirs): linux: added 2.6.22 with config for PC Engines Alix1c board Aug 07 13:16:14 03koen 07org.oe.dev * rd8e888f7... 10/ (1 packages/linux/linux_2.6.22.bb): linux 2.6.22: add AVR32 support Aug 07 13:16:20 03hrw 07org.oe.dev * ra8f105d4... 10/ (1 conf/machine/alix.conf): alix: added machine config for PC Engines Alix1c board Aug 07 13:17:43 noglitch: oh, that's for xserver-kdrive-common... well, i'll add these lines anyway :-) Aug 07 13:27:49 florian: cool thanks ! Aug 07 13:47:24 hello, my client was crashed, so i missed a notice. would you resend it? Aug 07 13:49:27 i am readon Aug 07 13:50:42 ~lart web developers Aug 07 13:50:43 * ibot brandishes Excalibur! "With this sword, I vanquish thee, web developers!" and lops off web developers's head Aug 07 13:51:38 03florian 07org.oe.dev * r240100bf... 10/ (1 packages/tasks/task-gpephone.bb): task-gpephone: Get rid of obsolete dependency task. Aug 07 13:51:45 03florian 07org.oe.dev * rb8fa0cf2... 10/ (1 packages/meta/meta-gpephone.bb): meta-gpephone: Remove obsolete dependency. Aug 07 13:51:52 03hrw 07org.oe.dev * r03be9e5a... 10/ (1 packages/linux/linux.inc packages/linux/linux_2.6.22.bb): linux 2.6.22: Alix boards use bzImage kernels Aug 07 13:55:34 i still have error in make depends Aug 07 13:55:36 http://pastebin.com/d636a1474 Aug 07 13:55:51 maybe a celan up work dir can work ? Aug 07 14:03:30 another strange datapoint in my libtool saga, Why does libtool-cross get built in ppc603e-angstrom-linux ? Aug 07 14:04:32 koen: In machine config file, there is this statement : PACKAGE_EXTRA_ARCHS = "armv4t armv5te" Aug 07 14:04:32 I thought it will be much accurat with simply : PACKAGE_EXTRA_ARCHS = "armv5te" Aug 07 14:09:31 noglitch: no Aug 07 14:09:41 noglitch: armv5 machine can use armv4 packages Aug 07 14:10:09 morning Aug 07 14:12:08 hrw: ok, I understand. and the require conf/machine/include/tune-arm926ejs.conf line tells which compilation config to use. Aug 07 14:12:23 yes Aug 07 14:19:20 another super tiny patch for at91sam9263 machine conf : http://rafb.net/p/SrbOmu24.html Aug 07 14:25:26 noglitch: is there an at91 patch for 2.6.22 yet? Aug 07 14:26:44 koen: no - 2.6.22-rc1 is latest Aug 07 14:30:48 koen, hrw: true, not yet... Aug 07 14:33:14 I think that it can be used with 2.6.22 - maybe some parts will not apply Aug 07 14:34:09 8 rejects Aug 07 14:34:12 hrw, koen: if you whant, I have a .22-r5 patchset which may apply better Aug 07 14:34:14 arrrg Aug 07 14:34:46 sorry, waiting on libusn to compile, finished configuring and moved on to something else ... Aug 07 14:34:48 noglitch: I do not have at91 boards Aug 07 14:37:33 * Crofton|home is convinced the task ordering is maximizes his annoyance ... Aug 07 14:39:16 bye Aug 07 14:57:18 03nicolas.ferre 07org.oe.dev * r639e1e03... 10/ (1 conf/machine/at91sam9263ek.conf): at91sam9263ek.conf: tweak MACHINE_FEATURES and EXTRA_IMAGECMD Aug 07 15:17:59 hi all Aug 07 15:21:56 * koen is annoyed by being unable to reach parts of the internet Aug 07 15:22:01 ~lart routers Aug 07 15:22:01 * ibot gives routers a "free" copy of Windows and then charges double for "Upgrades" Aug 07 15:26:35 heh heh Aug 07 15:31:26 03rpurdie 07org.oe.dev * re609b6a9... 10/ (1 classes/package_deb.bbclass classes/rootfs_deb.bbclass): package_deb/rootfs_deb.bbclass: Sync fixes from Poky Aug 07 15:32:41 RP: are the do_strip changes ready for OE? Aug 07 15:35:29 I give up on libtool/ppc603e C++ linking again Aug 07 15:35:39 koen: Yes, they seem to work nicely apart from making it spit a warning out into the build that used to hide in logfiles Aug 07 15:35:43 need to figure out why it works for OSK Aug 07 15:35:52 koen: Not much of a space gain but there is a tiny one Aug 07 15:36:07 there is already something foul with C++ linkage in PPC603 tool chain Aug 07 15:36:11 koen: I've asked on the binutils list about the warning Aug 07 15:36:43 is that why you ugraded to binutils .12? Aug 07 15:36:56 koen: I wondered if it would help, it doesn't Aug 07 15:37:23 koen: On the other hand it doesn't appear to break anything so I committed it... Aug 07 15:38:02 .12 works quite well, but hrw|gone will kill me if I bump PREF_VER for binutils again :) Aug 07 15:38:47 koen: --author rp@oe.org :-) Aug 07 15:40:55 RP: hey, as long as you are here and talking Aug 07 15:41:20 RP: dwmw2 was looking for you in #olpc - he wants to talk about swapfiles on jffs2 or something like that Aug 07 15:45:04 CosmicPenguin: I found him, thanks Aug 07 15:45:24 excellent Aug 07 15:45:32 * CosmicPenguin melts back into the background Aug 07 15:45:32 CosmicPenguin: I think it was jg who wanted me and I never did get to see the original conversation though... Aug 07 15:46:18 RP: yeah - its all based on trying to set up a temporary swap file solution to relieve the memory pressure on the old boards until they can be replaced Aug 07 15:48:27 CosmicPenguin: There are patches on LKML that will let them do just that. I just wish mainline would care about the bugs they exposed in the swap layer :/ Aug 07 15:48:45 * RP proposed fixes but nobody dares to apply them... Aug 07 15:49:36 CosmicPenguin: how about replacing all python code by C? Aug 07 15:49:38 * koen runs Aug 07 15:49:43 I'm with you man Aug 07 15:50:06 rwerite the kernel in python Aug 07 15:51:40 But I do admit, that pygtk and pycairo has been really useful a few times when I've been debugging rendering problems in X Aug 07 15:51:56 Its way easier to hack a python script on target Aug 07 15:53:46 hi CosmicPenguin Crofton|home koen Aug 07 15:54:16 hey chouimat Aug 07 15:54:22 gm Aug 07 15:54:43 * chouimat tries to decide on which OS to base his new project Aug 07 15:54:51 wince Aug 07 15:54:56 heh Aug 07 15:55:06 Crofton|home: it's not embedded ... Aug 07 15:55:16 vista! Aug 07 15:55:36 hehe ... maybe one of the BSDs Aug 07 15:56:34 Crofton|home. You now the word "svista" in Greek means, "turn it off, erase it", so if someone asks about windows we say "svista" Aug 07 15:57:00 steliosk: hehe Aug 07 15:57:27 * steliosk wonders how many MB embedded vista will be Aug 07 15:59:59 koen: binutils are fast to rebuild ;) Aug 07 16:00:01 bye Aug 07 16:01:30 steliosk: way too much Aug 07 16:25:03 i have a proposal regarding the list of events in the openembedded.org page. They should be reversed i.e future/current events at the top, oldest at the bottom Aug 07 16:25:16 any objections ? Aug 07 16:40:13 steliosk: go for it Aug 07 16:41:16 wow http://os2linux.sourceforge.net/ Aug 07 16:46:23 ok, done. added OEDEM dates also Aug 07 17:34:56 Fic 1973 in brazil: http://forumpcs.com.br/coluna.php?b=216541 Aug 07 17:40:15 Hello Aug 07 17:40:24 In a bb file, what does PE mean? Aug 07 17:40:28 precedence? Aug 07 17:40:29 :/ Aug 07 17:51:27 Got a proposal to do a presentation on OpenMoko/FIC for a local linux club. Is anyone wishing to provide one of the dev phones for a couple of weeks ? Aug 07 17:53:46 steliosk: you want a bv4 or a bv3? Aug 07 17:54:16 koen : anything. Just to get familiar with it and have something to saw them Aug 07 17:54:37 Hmm, is there no longer any X11R7.1-whatever files? Aug 07 17:54:55 Is X11R7.2 the only choice now? Aug 07 17:55:03 I'd better change the distro file Aug 07 17:55:06 steliosk: email your adress and I'll send you a bv3 + battery Aug 07 17:55:35 koen : Great thanks :) Aug 07 17:55:39 I see that a lot of X11R7.1-something files have been dropped... Aug 07 17:58:11 steliosk: you need a mini-usb cable to charge it, do you have one? Aug 07 17:58:41 koen : yes, my qtek pda/phone uses the same Aug 07 17:59:52 steliosk: I'd advice you to flash a build from .dev when you get it Aug 07 18:00:38 Um, maybe I'm just being stupid or something, but where did all the packages with -X11R7.1- in their PV go? Aug 07 18:00:47 I see no xedit, no xlogo Aug 07 18:00:54 Why were those dropped? Aug 07 18:01:20 Um, actually, most xorg utilities are gone Aug 07 18:01:34 (or seem to have gone) Aug 07 18:01:49 B_Lizzard: that's the usual for PDA distros, as far as I know Aug 07 18:02:08 I mean the bb files Aug 07 18:02:22 ohh, that I wouldn't know Aug 07 18:03:52 I'm worried things will be broken Aug 07 18:03:54 :/ Aug 07 18:05:28 maybe there's an xorg-utils package, but I'm pretty sure nothing depends on xedit, xlogo, or xeyes ;) Aug 07 18:06:16 Well, whereas before my recent update, xorg-util was full, now it's almost completely empty Aug 07 18:06:26 Lemme check upstream Aug 07 18:06:35 Maybe my toolchain is borked Aug 07 18:07:25 oh, we dont' do window maker Aug 07 18:07:47 or do we... *digs out the update instructions Aug 07 18:09:12 Same in upstream Aug 07 18:09:37 Did someone go on a binge? Aug 07 18:09:49 I mean, that's a lot of things to remove Aug 07 18:09:57 There must be a good reason :/ Aug 07 18:11:21 Well it seems to be traditional to pare out everything that isn't obviously and absolutely completely essential to the focussed task of making the machine a true Personal Digital Assistant, LOL Aug 07 18:11:47 Kinda annoys me actually, especially when it's on a machine that can handle more than the minimum Aug 07 18:12:18 still, there's usually options to add stuff back, so yeah, I'm wondering... Aug 07 18:12:34 Um, I'm only acting like this because some of the X11R7.1 files worked better than the 1.0.x files for me Aug 07 18:13:11 For instance, in xserver-kdrive-1.0.2, I couldn't change the mouse button Aug 07 18:13:18 ahh *nod* Aug 07 18:13:23 Neither with xmodmap nor with xmonobut Aug 07 18:13:55 And encodings-1.0.2 is missing some essential encodings compared to encodings X11R7.1-1.0.1 Aug 07 18:14:09 And a lot of crap like that Aug 07 18:14:43 ah, that's nasty Aug 07 18:15:05 Yeap Aug 07 18:15:22 Took me years to find the source of those bugs :/ Aug 07 18:15:46 damn :/ Aug 07 18:54:49 ibot: test Aug 07 18:54:50 Test Passed! Aug 07 18:57:43 ibot: fail Aug 07 18:57:44 * ibot gives eekee the worst grade imaginable, an A minus... minus! Aug 07 18:57:47 ;.; Aug 07 18:57:56 hehe Aug 07 19:20:23 Hi all, I'm trying to generate a i386 version of mamona and I'm getting the following error on glibc-intermidiate: Aug 07 19:20:26 checking installed Linux kernel header files... 2.0.10 or later Aug 07 19:20:26 | checking for kernel header at least 2.6.17... too old! Aug 07 19:20:26 | configure: error: *** The available kernel headers are older than the requested Aug 07 19:20:26 | *** compatible kernel version Aug 07 19:20:26 | FATAL: oe_runconf failed Aug 07 19:21:05 any idea what can be wrong? Aug 07 19:21:38 2.6.17 << 2.0.10 ???????? is it crazy? Aug 07 19:21:51 checking for kernel header at least 2.6.17... too old! Aug 07 19:22:02 you don't have kernel headers of 2.6.17 or higher Aug 07 19:22:05 what is the question? Aug 07 19:23:49 uhm I missunderstood... but I have it Aug 07 19:24:01 i have the 2.6.17 and the 2.6.20 Aug 07 19:24:09 where it is looking for? Aug 07 19:29:14 what kernel headers do you have installed? Aug 07 19:36:57 I'm coming back to OE after an enforced pause, and hitting the fact that libcap and tcpdump source sites were trojaned a while back. Can someone give me a brief pointer on how to work around downed source repositories? I'm building task-base for an ARM dev board. Aug 07 19:37:50 russf: DISTRO=angstrom-2007.1 Aug 07 19:38:04 vivijim: try i586 instead of i386 Aug 07 19:38:13 OK. thanks - I'll give it a try Aug 07 19:39:44 koen I'm using the x86 machine that uses i486... should I change it in x86 machine or use another machine description? Aug 07 19:40:16 koen: i586-generic one? is it better? Aug 07 19:41:16 I remember something about i486 and nptl, but I'm fuzzy on the details Aug 07 19:41:27 I know i586 works for me Aug 07 19:41:34 never tried to build i486 Aug 07 19:42:29 2.6.22 is booting on the cm-x270, now to finish the integration ... Aug 07 19:43:23 Um, I was asking before about the reasons behind the removal of all X11R7.1-x.x.x bb files from the repo Aug 07 19:43:36 It seems a bit...um, drastic Aug 07 19:44:10 koenok, I'll try it Aug 07 19:44:13 Are they going to be brought back? Aug 07 19:53:11 I'll take that as a no Aug 07 20:21:34 03mickeyl 07org.oe.dev * r2f3182e2... 10/ (3 files in 3 dirs): uboot-openmoko|linux-gta01: softlink to the latest binary in the deploy directory Aug 07 20:23:45 B_Lizzard: They're not useful any more - all the X packages are now modular and have their own version number - trying to stick with the monolithic versioning is problematic Aug 07 20:29:03 Eh, I've had some problems with the "modular" versions before Aug 07 20:29:33 As I said before, "encodings-X11R7.1-1.0.1" has a few more encodings than 1.0.2 Aug 07 20:29:54 And I've had my share of problems with xserver-kdrive-1.0.2 Aug 07 20:30:34 Plus, I can understand removing the X11R7.1 packages that had monolithic alternatives available, but as I can see, a lot of packages were simply removed Aug 07 20:30:40 Not a good practice at all Aug 07 20:31:16 koen|away: I had exactly the same error using the i586-generic :( Aug 07 20:32:01 7.1 was nearly two years ago - if the core X code doesn't have what you need today, then your beef isn't here Aug 07 20:32:38 re Aug 07 20:35:21 hi zecke Aug 07 20:35:47 hey Aug 07 20:35:50 I was thinking more of "bug", not "beef" Aug 07 20:36:37 zecke: I forgot to tell you that it is confirmed, I'll attend the oedem... Aug 07 20:36:52 awesome Aug 07 20:38:39 Who is the tool chain guru? Aug 07 20:38:50 er sucker? Aug 07 20:40:05 Eh? Aug 07 20:40:10 :) Aug 07 20:41:14 There is something fishy in ppc-603 tool chain wrt to C++ Aug 07 20:41:39 I think libtool breaks whatever hacks we used to work around the base problem Aug 08 00:02:02 http://www.ohloh.net/ Aug 08 00:02:08 no monotone support Aug 08 00:08:49 03jluebbe 07org.oe.dev * ra37614a8... 10/ (1 packages/openmoko2/openmoko-today2_svn.bb): packages/openmoko2/openmoko-today2_svn.bb: RDEPEND on eds ecal server to fix crashes Aug 08 00:12:26 ibot: test Aug 08 00:12:27 Oh, no! There's a test and I haven't studied! Aug 08 00:15:17 heh Aug 08 00:15:34 you need a cot in your lab Aug 08 00:38:15 koen|away, ping **** ENDING LOGGING AT Wed Aug 08 02:59:57 2007