**** BEGIN LOGGING AT Mon Mar 10 02:59:58 2014
Mar 10 06:09:48 <sanny> hai all,  am trying to build cross compiler for at91sam9g35. Am getting the following issue http://pastebin.com/9BkhUwkE
Mar 10 08:40:54 <mckoan> good morning
Mar 10 08:41:37 <lpapp> good morning
Mar 10 09:09:30 <Jin^eLD> mornin
Mar 10 09:39:24 <bluelightning> morning all
Mar 10 09:41:51 <Jin^eLD> morning bluelightning
Mar 10 09:43:12 <SorenHolm> Hi
Mar 10 09:44:20 <SorenHolm> Have anyone here tried to build Qt-applications using cmake. Apparently it ends up linking with the native version...
Mar 10 09:52:57 <bluelightning> SorenHolm: we have example recipes that do that, e.g. see the recipe for qmmp
Mar 10 09:53:20 <SorenHolm> bluelightning: I'll have a look thanks..
Mar 10 09:55:53 <lpapp> SorenHolm: I built them, yeah.
Mar 10 09:56:07 <lpapp> in fact, the whole KDE stack is like that.
Mar 10 10:04:03 <SorenHolm> bluelightning: hmmm it looks pretting much what I'm doing except that I have qt4-x11-free in DEPENDS. I'll try to inherit qt4x11 and see if it helps.
Mar 10 10:05:19 <lpapp> SorenHolm: why are you still using qt4 ? :O
Mar 10 10:06:20 <SorenHolm> lpapp: "If it works don't fix it". But we aare actually moving foward within the next weeks.
Mar 10 10:06:27 <bluelightning> SorenHolm: inheriting qt4x11 will already put qt4-x11-free in DEPENDS
Mar 10 10:07:10 <SorenHolm> bluelightning: thought so.
Mar 10 10:07:16 <lpapp> SorenHolm: I do not see how it works... but ok.
Mar 10 10:07:36 <lpapp> it is slower, more unflexible, and unreliable.
Mar 10 10:07:45 <bluelightning> lpapp: please don't start this again
Mar 10 10:07:45 <lpapp> and it is also unmaintained, pretty much.
Mar 10 10:08:04 <SorenHolm> lpapp: We have an application that uses some on the thing that are removed from qt4.
Mar 10 10:08:23 <lpapp> like?
Mar 10 10:08:24 <SorenHolm> lpapp: But within the next weeks we are moving it.
Mar 10 10:08:40 <SorenHolm> lpapp: qfilesystemmodel afair.
Mar 10 10:08:56 <lpapp> http://qt-project.org/doc/qt-5/QFileSystemModel.html
Mar 10 10:10:55 <SorenHolm> lpapp: oh well - my memory apparently  is not helping me here :D ... but there where somehing.
Mar 10 10:12:21 <lpapp> SorenHolm: if you do not need to release the software within the next few weeks, I would not personally bother putting more work in to qt 4, but this is just my opinion.
Mar 10 10:15:09 <SorenHolm> lpapp: ok, thanks. I won't.
Mar 10 10:17:13 <lpapp> SorenHolm: http://layers.openembedded.org/layerindex/branch/master/layer/meta-qt5/
Mar 10 10:17:34 <lpapp> (and http://qt-project.org/doc/qt-5/portingguide.html)
Mar 10 10:17:46 <SorenHolm> lpapp: yup - I already have it pulled in.
Mar 10 10:18:11 <lpapp> ok. :)
Mar 10 10:19:42 <lpapp> hmm, I am surprised it depends on meta-ruby ...
Mar 10 10:19:53 <lpapp> qt itself does not have any ruby dependency for sure.
Mar 10 10:20:27 <lpapp> eh, they introduced it in webkit? :O
Mar 10 10:21:02 <lpapp> (ok, it is optional... not sure how optional layer dependencies are handled)
Mar 10 10:21:17 <sanny> hai all,  am trying to build cross compiler for at91sam9g35. Am getting the following issue http://pastebin.com/9BkhUwkE
Mar 10 10:21:54 <lpapp> This paste has been removed!
Mar 10 10:41:47 <Net147> I am getting some hangs on shutdown with systemd 208. is it planned to upgrade to systemd 210 for Yocto 1.6?
Mar 10 10:43:31 <bluelightning> Net147: I'm not directly involved in systemd work but I had thought so, yes
Mar 10 10:44:47 <bluelightning> Net147: btw, re bug 5918 - is that something you'd be able to look into?
Mar 10 10:44:48 <yocti> Bug https://bugzilla.yoctoproject.org/show_bug.cgi?id=5918 normal, Medium, 1.6, richard.purdie, NEW , git submodule init doesn't work on target
Mar 10 10:50:33 <Net147> yocti: busy with other things at the moment, but a simple sed would probably fix it...
Mar 10 10:50:35 <yocti> Net147: Error: "busy" is not a valid command.
Mar 10 10:50:43 <Net147> bluelightning: ^
Mar 10 10:50:43 <lpapp> bluelightning: ^
Mar 10 10:52:52 <bluelightning> Net147: ok... we'll get to it when we get to it, but I figured if it was something you needed a bit more urgently... ;)
Mar 10 11:06:32 <Net147> bluelightning: cross development using Qt 5.2.1 + Qt Creator to Yocto-based target device seems to work well enough for me
Mar 10 11:13:26 <Net147> khem: ping
Mar 10 11:21:56 <Jin^eLD> for multimachine configurations... is PERSISTENT_DIR supposed to be shared or should it be set per machine?
Mar 10 11:39:36 <bluelightning> Jin^eLD: shared
Mar 10 11:39:55 <Jin^eLD> thx
Mar 10 11:40:02 <bluelightning> in most cases you shouldn't need to set it at all
Mar 10 11:40:23 <Jin^eLD> well it somehow creates it the top dir of my build
Mar 10 11:40:33 <Jin^eLD> *in
Mar 10 11:40:45 <bluelightning> the top dir meaning which directory exactly? your build directory?
Mar 10 11:41:04 <Jin^eLD> no, outside the actual build directory
Mar 10 11:41:27 <Jin^eLD> I have an $OETREE set and the cache is somehow created under $OETREE/cache
Mar 10 11:41:50 <Jin^eLD> $OETREE being my top directory below which I have all checkouts etc
Mar 10 11:41:55 <bluelightning> er, OETREE?
Mar 10 11:42:03 <bluelightning> I don't think that's a standard variable
Mar 10 11:42:14 <Jin^eLD> me neither :)
Mar 10 11:42:33 <bluelightning> well, I'd need to know what standard variable that does correspond to...
Mar 10 11:43:06 <Jin^eLD> I use it to set various standard variables relative to that
Mar 10 11:43:20 <bluelightning> FWIW, PERSISTENT_DIR is intended to be saved in your build directory, so that e.g. the persistent version sqlite database is preserved even if you throw out TMPDIR
Mar 10 11:43:20 <Jin^eLD> i.e. export DL_DIR=${OETREE}/downloads
Mar 10 11:43:22 <Jin^eLD> and similar
Mar 10 11:43:41 <Jin^eLD> I did not do any settings of PERSISTENT_DIR myself so I wonder how it landed outside
Mar 10 11:43:58 <Jin^eLD> but I'll just set PERSISTENT_DIR accordingly and it'll be fine I guess
Mar 10 11:44:38 <bluelightning> the default is ${TOPDIR}/cache
Mar 10 11:45:04 <bluelightning> normally TOPDIR points to the build directory
Mar 10 11:45:11 <Jin^eLD> aha, there it is...
Mar 10 11:45:20 <Jin^eLD> my TOPDIR points to the top directory of the checkout
Mar 10 11:45:27 <bluelightning> right, if you've set things up in an alternative layout you'll find things look different ;)
Mar 10 11:46:13 <bluelightning> I suspect we assume that TOPDIR is appropriate to write things into in other places, so I would caution against setting it to somewhere where that's not appropriate
Mar 10 11:46:24 <Jin^eLD> well I was assuming that TOPDIR is the shared top directory, i.e. judging from the example #DL_DIR ?= "${TOPDIR}/downloads"
Mar 10 11:46:51 <Jin^eLD> idea being that downloads directory is shared between all possible builds and configurations
Mar 10 11:47:10 <Jin^eLD>      This variable points to the Build Directory. BitBake automatically sets this variable.
Mar 10 11:47:13 <bluelightning> that isn't the default though
Mar 10 11:47:13 <Jin^eLD> ok, so I should fix that
Mar 10 11:47:16 <Jin^eLD> and override downloads
Mar 10 11:47:20 <Jin^eLD> thanks
Mar 10 15:22:13 <rigaku_> I am using denzil and I use the psplash utiliy at boot up and it seems to work but I still have a break between the splash screen and Xorg starting that displays text output from the SysVinit. Does anyone know how to solve this?
Mar 10 16:00:44 <rigaku_> anyone out there?
Mar 10 16:02:12 <kergoth> no
Mar 10 16:02:45 <rigaku_> thats what I thought
Mar 10 16:03:16 * BCMM lurks
Mar 10 16:05:21 <kergoth> IRC is asynchronous. all you can do is speak and await a response. not everyone sits staring at a quiet irc channel when they have day jobs and the like
Mar 10 16:06:56 <LetoThe2nd> damn day jobs! :/
Mar 10 16:08:08 <BCMM> depends how big the channel is really
Mar 10 16:08:26 <BCMM> ask a question in #ubuntu and something has gone wrong if you don't have five wrong answers in the first minute
Mar 10 16:10:13 <rburton> haha
Mar 10 16:10:30 <rburton> is one of them 'install mint"
Mar 10 16:11:02 <kergoth> hah
Mar 10 16:11:14 <LetoThe2nd> usually its more like "did your mint install break this, or are you actually using an ubuntu"
Mar 10 16:29:55 <SnookEE> Hello everyone, I'm trying to setup a rootfs with php support. I've built the php package successfully, but my rootfs does not have the php-cgi and php populated in the bin directory.
Mar 10 16:31:05 <rburton> you'll need to add the packages containing them to the image
Mar 10 16:31:23 <rburton> just building them doesn't put anything into an image
Mar 10 16:32:14 <SnookEE> is this why is creates the rpm packages?
Mar 10 16:32:33 <SnookEE> when i look inside of the rpm package that is also missing the binary
Mar 10 16:33:07 <rburton> it possibly put the binaries in a package you're not looking at.  i've never used php though, so i can't really help.
Mar 10 16:33:13 <SnookEE> oh i see what you are saying, yes I did add them to the image
Mar 10 16:33:30 <SnookEE> IMAGE_INSTALL += "php"
Mar 10 16:33:42 <SnookEE> something along those lines is what you are saying needed to be done?
Mar 10 16:33:47 <rburton> yes
Mar 10 16:34:06 <SnookEE> yes, I believe I've done that correctly so and when i bitbake that it appears to build it
Mar 10 16:34:30 <SnookEE> also the php.inc in the recipe seems to specify that the php and php-cgi should be added to the bindir
Mar 10 16:34:35 <SnookEE> but somehow it fails to do that
Mar 10 16:34:36 <mckoan> SnookEE: add php-cgi to IMAGE_INSTALL
Mar 10 16:34:54 <mckoan> SnookEE: IMAGE_INSTALL += "php-cgi"
Mar 10 16:34:58 <rburton> yeah, looking at the recipe there's a load of packages you probably care about
Mar 10 16:35:13 <SnookEE> IMAGE_INSTALL = "udev-cache update-rc.d util-linux-libblkid packagegroup-core-boot sysvinit-inittab base-passwd modutils-initscripts update-alternatives-cworth initscripts libkmod netbase busybox-udhcpc eglibc udev sysvinit base-files udev-utils util-linux-libuuid busybox busybox-syslog busybox-hwclock init-ifupdown sysvinit-pidof run-postinsts "
Mar 10 16:35:13 <SnookEE> IMAGE_INSTALL += " bash libxml2 openssh openssl sqlite3 lighttpd php fuse grep gzip perl vim gdb file cronie sshfs-fuse util-linux sed time "
Mar 10 16:35:13 <SnookEE> IMAGE_INSTALL += " zlib lemon libc-client php-cgi"
Mar 10 16:35:13 <SnookEE> DESCRIPTION = "Image recipe for FFM main CPU"
Mar 10 16:35:16 <SnookEE> #/home/msnook/freescale/mx6_stuff/yocto_tryout/fsl-community-bsp/sources/poky/meta/
Mar 10 16:35:18 <SnookEE> sorry about the big paste
Mar 10 16:35:26 <mckoan> SnookEE: use pastebin
Mar 10 16:36:34 <mckoan> SnookEE: try pastebin the recipe
Mar 10 16:36:56 <SnookEE> sorry one sec
Mar 10 16:37:19 <rburton> SnookEE: you just need to add php-cgi, as mckoan said
Mar 10 16:37:30 <SnookEE> ok doing that now
Mar 10 16:38:22 <SnookEE> isn't the php binary also something that should have been added? or is it just php-cgi?
Mar 10 16:38:47 <rburton> have a look at the recipe
Mar 10 16:38:57 <rburton> you'll see that php-cgi is the package that contains the php binary
Mar 10 16:40:15 <SnookEE> oh i think i see that
Mar 10 16:40:16 <SnookEE> FILES_${PN}-cli = "${bindir}/php"
Mar 10 16:40:16 <SnookEE> FILES_${PN}-cgi = "${bindir}/php-cgi"
Mar 10 16:40:30 <SnookEE> is that specifying a sub package of some sort?
Mar 10 16:40:52 <SnookEE> ahhh and this part
Mar 10 16:40:53 <SnookEE> FILES_${PN} = "${libdir}/php"
Mar 10 16:41:09 <SnookEE> I said just "php" so all it did was give me the library stuff
Mar 10 16:42:45 <SnookEE> you guys are the best! thanks
Mar 10 19:22:39 <MatthieuG> Hi !
Mar 10 19:22:41 <MatthieuG> There is way to determine with bitbake if a recipe include another ?
Mar 10 19:23:18 <MatthieuG> I want to know if  angstrom console-image use boost recipe
Mar 10 19:25:10 <Crofton> probably not
Mar 10 19:28:45 <MatthieuG> so if I include in my recipe boost and console-image there may be an duplicated entry
Mar 10 19:28:56 <Crofton> dupes are not an issue
Mar 10 19:30:36 <MatthieuG> ok
Mar 10 19:31:41 <MatthieuG> But, If I add boost in my recipe, bitbake will detect a change in my recipe, so does it rebuild some dependencies ?
Mar 10 19:32:00 <MatthieuG> and so build boost twice
Mar 10 19:45:43 <khem> MatthieuG: well, a dependency can come from many packages
Mar 10 19:46:06 <khem> but bitbake's dependency resolver takes care of where to converge
Mar 10 19:46:38 <khem> in this case both deps will terminate at 1 package i.e. boost
Mar 10 19:46:55 <kergoth> gah, gettext version mismatch. anyone else doing builds without gplv3 recently?
Mar 10 19:47:17 <khem> so bitbake will make sure that boost is built and staged before all recipes depending on it are built
Mar 10 19:47:27 <khem> kergoth: w/o gplv3 ?
Mar 10 19:47:36 <khem> so you use external tc
Mar 10 19:47:43 <khem> like gcc 4.2 or somesuch
Mar 10 19:48:13 <kergoth> it doesn't exclude it from native/cross, since those don't end up on the target
Mar 10 19:48:37 <khem> how about gcc-runtime ?
Mar 10 19:49:35 <kergoth> libgcc is gplv3? that seems unlikely, since itd infect everything you build, no?
Mar 10 19:50:40 <khem> kergoth: its gplv3+exceptions :)
Mar 10 19:51:00 <khem> but the fact is its license was changed when switch was made from v2 to v3
Mar 10 19:51:36 <khem> although the exception clause remained same until gcc 4.4
Mar 10 19:53:32 <khem> my hunch is FSF was expecting apple to return to gcc development/use with the exception rework
Mar 10 19:55:38 <MatthieuG> khem: thank you for explain
Mar 10 23:30:02 <khem> bitbake has gripes fetching some gitouious repos
Mar 10 23:30:07 <khem> e.g. git://gitorious.org/qt/qtlocation.git
Mar 10 23:30:22 <khem> I can clone it fine on same build system outside bitbake
Mar 10 23:30:58 <khem> but bitbake
Mar 10 23:31:00 <khem> ERROR: Fetcher failure: Unable to find revision f28408346243cf090326f4738fd838219c21e00f in branch master even from upstream                                    ERROR: Function failed: Fetcher failure for URL: 'git://qt.gitorious.org/qt/qtlocation.git;branch=master'. Unable to fetch URL from any source.
Mar 10 23:31:37 <khem> I have had issues with gitorious + bitbake in past
Mar 11 00:24:42 <mr_science> khem: how is the recipe git stuff specified?
Mar 11 00:25:09 <mr_science> did you try branch=master and src_rev=commit_hash ?
Mar 11 00:25:43 <mr_science> git fetch stuff changed recently and doesn't seem to support src_rev=tag_name any more
**** ENDING LOGGING AT Tue Mar 11 02:59:59 2014