**** BEGIN LOGGING AT Fri Sep 07 02:59:59 2012
Sep 07 05:51:24 <vitus> allo ka6sox
Sep 07 05:51:42 <vitus> sorry, wrong window
Sep 07 05:52:20 <ka6sox> okay
Sep 07 07:05:38 <bluelightning> morning all
Sep 07 07:05:47 <woglinde> hi  bl
Sep 07 07:26:01 <mckoan|away> good morning
Sep 07 08:13:14 <silvio> morning all
Sep 07 08:26:28 <mickey_office> morning
Sep 07 08:26:54 <woglinde> hw mickeyl
Sep 07 08:30:54 <silvio> I have the qt 4.8.1, but in openembedded-core I have this line;
Sep 07 08:30:58 <silvio> openembedded-core/meta/conf/distro/include/distro_tracking_fields.inc:RECIPE_LATEST_VERSION_pn-qt4-native = "4.8.0"
Sep 07 08:31:15 <silvio> so I obtain these type of warnings:
Sep 07 08:31:16 <woglinde> and?
Sep 07 08:31:20 <woglinde> chnage it
Sep 07 08:31:26 <silvio> NOTE: preferred version 4.8.0 of qt4-x11-free not available (for item libqtpvrqwswsegl4-dbg)
Sep 07 08:31:27 <silvio> NOTE: versions of qt4-x11-free available: 4.7.4 4.8.1
Sep 07 08:31:28 <woglinde> anf off you go
Sep 07 08:31:39 <silvio> can I canghe it there or in other layers?
Sep 07 08:31:51 <silvio> I mean it secure canghe it there?
Sep 07 08:32:29 <woglinde> o.O
Sep 07 08:32:32 <silvio> the right way to do is change it in openembedded-core layer?
Sep 07 08:32:48 <woglinde> yes
Sep 07 08:32:56 <silvio> ok thanks woglinde 
Sep 07 08:33:15 <woglinde> send patch request do oe-core
Sep 07 08:40:49 <woglinde> hurray khem send patches for qemu 1.2
Sep 07 08:48:11 <silvio> I changed these linees, but not solve my warnings 
Sep 07 08:51:10 <silvio> I hve to rebuild somethings? I still have same errors
Sep 07 08:58:11 <woglinde> do git grep for 4.8.0
Sep 07 08:58:25 <woglinde> in meta-oe and oe-core
Sep 07 08:58:34 <woglinde> and maybee other layers you are using
Sep 07 08:58:56 <silvio> i did grep 4.8.0 in stuff, now i try git grep
Sep 07 09:03:12 <woglinde> hi zecke
Sep 07 09:03:39 <silvio> woglinde, I found something in file patches
Sep 07 09:05:33 <silvio> could be this?:
Sep 07 09:05:36 <silvio> meta-angstrom/conf/distro/include/angstrom-2010-preferred-versions.inc:ANGSTROM_QT_VERSION ?= "4.8.0"
Sep 07 09:05:43 <stefan_schmidt_w> woglinde, zecke: hey guys! How are you?
Sep 07 09:07:42 <woglinde> silvio yes but as you see its a week reference you can set it in your local.conf to 4.8.1
Sep 07 09:07:51 <woglinde> he stefan_schmidt fine so far
Sep 07 09:11:00 <woglinde> hm I should update icedtead to 1.11.4
Sep 07 09:11:13 <silvio> woglinde, which is the right line to set QT at 4.8.1 that I have to add?
Sep 07 09:12:23 <woglinde> ANGSTROM_QT_VERSION = "4.8.1" maybee?
Sep 07 09:17:58 <silvio>  ANGSTROM_QT_VERSION ?= "4.8.1  I put also the question mark, seams work, thanks
Sep 07 09:18:30 <woglinde> there is nothing which can override local.conf
Sep 07 09:22:35 <silvio> right, so the messages that I had are only warnings that are come from the  set of ANGSTROM_QT_VERSION ?= "4.8.0" in the file meta-angstrom/conf/distro/include/angstrom-2010-preferred-versions.inc, i m right?
Sep 07 09:30:40 <woglinde> yes
Sep 07 09:30:55 <pb_> morning all
Sep 07 09:34:06 <silvio> hi
Sep 07 10:26:45 <RP> morning all
Sep 07 10:27:49 <DarkPlayer> hi
Sep 07 10:31:20 <DarkPlayer> i have a problem compiling a 3.5/3.6 kernel for the hx4700 ipaq. I am using a the arm-angstrom-linux-gnueabi- cross compiling toolchain and can't get a kernel to boot. My config is based on the 3.0.1 config from angstrom-linux, which works.
Sep 07 10:32:03 <bluelightning> DarkPlayer: hi... I'm the maintainer of the meta-handheld layer for OE which supports hx4700
Sep 07 10:32:21 <bluelightning> DarkPlayer: I have been trying for a while now to get a booting system on hx4700 with not much success :(
Sep 07 10:32:52 <bluelightning> last I spoke to the author of the recent hx4700 kernel work he suggested my problems were not kernel-related however
Sep 07 10:33:07 <DarkPlayer> bluelightning: my kernel seems to crash after loading
Sep 07 10:33:35 <DarkPlayer> bluelightning: the screen does not init and no LED turns on
Sep 07 10:33:41 <bluelightning> 3.0.1 is quite old, most of the hx4700 support has been merged into mainline now
Sep 07 10:33:56 <bluelightning> so what we really ought to do is update to a newer kernel...
Sep 07 10:37:50 <DarkPlayer> bluelightning: there is no kernel > 3.0.1 known to work ?
Sep 07 10:38:04 <bluelightning> DarkPlayer: the author says it works fine for him
Sep 07 10:38:29 <bluelightning> I haven't been able to verify that, but I 
Sep 07 10:38:44 <bluelightning> .. I probably haven't spent enough time on it
Sep 07 11:20:29 <mertsas> I'm looking at where the devshell is chosen. We try to set OE_TERMINAL="xterm", but we get the UnsupportedTerminal exception
Sep 07 11:28:09 <bluelightning> mertsas: the actual terminal options are implemented in meta/lib/oe/terminal.py FYI
Sep 07 11:43:52 <DarkPlayer> bluelightning: you said there is some version newer than 3.0.1, but i cant find any build scripts for hx4700 and a newer version, which source did you mean ?
Sep 07 12:03:18 <mickey_office> yo z.!
Sep 07 12:24:06 <pb_> hi mickey_office
Sep 07 12:25:14 <mickey_office> hey pb_ ! how are things? didn't you plan to stop by in Frankfurt some weeks ago?
Sep 07 12:25:57 <eFfeM> khem, saw your eglibc reply, I still suspect upstream, note that the URI also has an additional /svn in it; also the uri as I made it is the same as the one from oe core
Sep 07 12:28:35 <pb_> mickey_office: yeah, that plan didn't quite work out.  the other friends that we were going to visit in Frankfurt had to take their daughter to hospital for a skull operation (!!) and then Ruth reached the point where it was impossible to travel due to her huge size.
Sep 07 12:29:03 <mickey_office> when is the new baby due?
Sep 07 12:29:13 <mickey_office> and did the skull operation work out ok?
Sep 07 12:29:17 <mickey_office> that sounds scary
Sep 07 12:29:19 <pb_> well, her due date was actually today
Sep 07 12:29:30 <pb_> but, as it happens, she was born on the 18th, about 3 weeks early :-}
Sep 07 12:29:36 <mickey_office> oooh
Sep 07 12:29:39 <mickey_office> congratulations!!!
Sep 07 12:29:43 <mickey_office> what's her name?
Sep 07 12:29:46 <pb_> Sophie
Sep 07 12:29:49 <pb_> thanks :-)
Sep 07 12:29:56 <mickey_office> awesome, welcome on board.
Sep 07 12:30:13 <pb_> see, um, http://pbcl.net/gallery/index.php/Sophie
Sep 07 12:30:33 <pb_> yeah, skull op apparently went ok
Sep 07 12:31:55 <mickey_office> very nice. how's Emma treating her new sister?
Sep 07 12:32:38 <pb_> she's very excited to have a new sister.  not always 100% reliable at remembering that babies are more fragile than dolls and can't be dropped, squeezed etc.
Sep 07 12:33:28 <mickey_office> right, i think that's ok given her age ;)
Sep 07 12:33:36 <mickey_office> how's the house coming up?
Sep 07 12:33:36 <bluelightning> DarkPlayer: we don't have recipes for the newer kernel yet
Sep 07 12:33:46 <mickey_office> are you done yet?
Sep 07 12:35:42 <pb_> not finished, though we have some more rooms done.
Sep 07 12:35:55 <pb_> we're now sleeping in our actual bedroom, which means I have my office back.
Sep 07 12:37:54 <mickey_office> heh, very good
Sep 07 12:38:12 <eFfeM> pb_:  gratz with the baby & my sympathy for the housing
Sep 07 12:38:26 * eFfeM is in the middle of a renovation project too
Sep 07 12:40:49 <pb_> eFfeM: thanks
Sep 07 12:41:27 <eFfeM> pb_ enjoy them while you can, time goes fast (mine are already 23)
Sep 07 12:41:54 <eFfeM> btw due to the renovation I probably won't go to ELCE
Sep 07 12:42:14 <mickey_office> hrw: back home safely?
Sep 07 12:42:18 <pb_> yeah, I won't go to ELC-E either, due to the baby in my case
Sep 07 12:43:07 <hrw> mickey_office: on a train to Poznań now. just left Wrocław
Sep 07 12:43:19 <mickey_office> hrw: ah, so still underway
Sep 07 12:43:36 <hrw> mickey_office: friend's wedding tomorrow. home on monday
Sep 07 12:43:55 <mickey_office> ah, right, the wedding, i remember
Sep 07 12:44:35 <hrw> pb_: I will be in Cambridge 1-6 Oct. ale?
Sep 07 12:48:30 <pb_> sounds like a good plan.  ping me closer to the time and we can sort something out.
Sep 07 12:49:26 <hrw> pb_: ok, will do
Sep 07 12:54:14 <silvio> I put in my local.conf option to use make with -j8 option, but seams that some recipes doesn't use all cores, someone know why? (eg. QWT)
Sep 07 12:58:04 <bluelightning> silvio: does the recipe set PARALLEL_MAKE = "" ?
Sep 07 13:00:34 <silvio> In the recipe isn't set, but in build/local.conf is: PARALLEL_MAKE = "-j 8"
Sep 07 13:00:52 <silvio> bluelightning, In the recipe isn't set, but in build/local.conf is: PARALLEL_MAKE = "-j 8"
Sep 07 13:01:14 <hrw> silvio: some recipes do not use make but other build systems
Sep 07 13:01:16 <bluelightning> silvio: so it could be that the way the build is invoked, it does not get that option passed in
Sep 07 13:03:10 <silvio> at me seams qwt use make, so is in devshell, but pass option in other way, there is a reason for that choice? Because compiling time increase a lot
Sep 07 13:03:38 <bluelightning> silvio: I doubt it's deliberate
Sep 07 13:03:49 <bluelightning> that it doesn't use PARALLEL_MAKE I mean
Sep 07 13:04:36 <silvio> ok, so is better don't try to force it?
Sep 07 13:06:34 <bluelightning> silvio: hmm, looking at the recipe it doesn't seem to me that it's overriding the default make commands...
Sep 07 13:08:51 <silvio> bluelightning, when compiling I do this  "ps axww|grep make" , and seam that doesn't consider -j option
Sep 07 13:11:21 <silvio> bluelightning, is not so important, is just curiosity to understand something more 
Sep 07 13:15:54 <bluelightning> silvio: PARALLEL_MAKE gets passed to make in oe_runmake in base.bbclass, via EXTRA_OEMAKE which has PARALLEL_MAKE added to it in conf/bitbake.conf for the compile task
Sep 07 13:16:41 <bluelightning> so it would be worth checking the value of both PARALLEL_MAKE and EXTRA_OEMAKE for qwt (using bitbake -e qwt ...)
Sep 07 13:28:30 <silvio> # PARALLEL_MAKE=-j 8
Sep 07 13:28:30 <silvio> PARALLEL_MAKE="-j 8"
Sep 07 13:29:22 <silvio> but it should be:  export PARALLEL_MAKE="-j 8"
Sep 07 13:32:20 <bluelightning> silvio: no, that's not needed
Sep 07 13:32:48 <bluelightning> silvio: what is EXTRA_OEMAKE for qwt?
Sep 07 13:32:55 <bluelightning> the value of it I mean
Sep 07 13:33:46 <silvio> bluelightning, EXTRA_OEMAKE="MAKEFLAGS="
Sep 07 13:34:37 <silvio> bluelightning, make  MAKEFLAGS=  "$@" || die "oe_runmake failed"
Sep 07 13:35:51 <bluelightning> silvio: what about: bitbake -e -c compile qwt | grep EXTRA_OEMAKE
Sep 07 13:37:03 <silvio> bluelightning, # EXTRA_OEMAKE= MAKEFLAGS= 
Sep 07 13:37:04 <silvio> EXTRA_OEMAKE="MAKEFLAGS="
Sep 07 13:37:04 <silvio> # OE_TERMINAL_EXPORTS=XAUTHORITY SHELL DBUS_SESSION_BUS_ADDRESS DISPLAY EXTRA_OEMAKE SCREENDIR
Sep 07 13:37:04 <silvio> OE_TERMINAL_EXPORTS="XAUTHORITY SHELL DBUS_SESSION_BUS_ADDRESS DISPLAY EXTRA_OEMAKE SCREENDIR"
Sep 07 13:38:40 <silvio> bluelightning, to say where are my header to a recipe the right variable to use is filepath or filedir?
Sep 07 13:38:56 <bluelightning> silvio: sorry I don't understand the question...
Sep 07 13:39:44 <silvio> bluelightning, when i compile my recipe I obtain this error:
Sep 07 13:39:46 <silvio> ./ui/tsampler.h:19:21: fatal error: qscpecg.h: No such file or directory
Sep 07 13:40:21 <silvio> so I need to specify where can be found the file *.h, who is the right way in a recipe?
Sep 07 13:40:34 <silvio> so I need to specify where can be found the file *.h, which is the right way in a recipe?
Sep 07 13:40:51 <silvio> so I need to specify where can be found the file *.h, which is the right way do specify it in a recipe?
Sep 07 13:41:12 <bluelightning> depends on the build system being used by the source you are building...
Sep 07 13:42:02 <bluelightning> plain makefile? qmake? cmake? autotools?
Sep 07 13:42:23 <silvio> bluelightning, qmake
Sep 07 13:42:51 <silvio> from a qtxxxx.pro  project
Sep 07 13:43:01 <bluelightning> silvio: are you doing inherit qmake in your recipe?
Sep 07 13:43:12 <bluelightning> er, inherit qmake2 I mean
Sep 07 13:44:31 <silvio> bluelightning, I inherit only qt4x11 in my recipe, then in do_configure I use qmke2 to obtain make file from file.pro
Sep 07 13:45:09 <bluelightning> right, qt4x11 inherits qmake2 already, ok
Sep 07 13:45:41 <silvio> bluelightning, right :)
Sep 07 13:46:09 <bluelightning> silvio: er running qmake is already handled, you don't need to do that yourself
Sep 07 13:47:56 <silvio> so i s not necessaary : qmake2 -makefile project.pro ?
Sep 07 13:50:03 <woglinde> re
Sep 07 13:53:03 <bluelightning> silvio: no, if you look in qmake_base.bbclass which is inherited by qmake2.bbclass, it already does this
Sep 07 13:53:09 <bluelightning> silvio: this was true in OE-Classic as well
Sep 07 13:53:59 <silvio> bluelightning, ok, thanks, and for headers I need use staging?
Sep 07 13:54:37 <bluelightning> silvio: well, headers from recipe A that are needed by recipe B need to be staged by recipe A yes
Sep 07 13:55:22 <bluelightning> but staging is accomplished just by making sure the headers are installed during do_install, there's no extra steps beyond that
Sep 07 13:57:16 <silvio> bluelightning, but in do_install  I have only this:    oe_runmake install INSTALL_ROOT=${D}, but then headers are not found, something wrong i think
Sep 07 13:58:29 <bluelightning> silvio: so there are two possibilities - (a) headers were not installed at all or (b) headers were installed but the path is not the expected one
Sep 07 13:58:51 <silvio> find is running ....
Sep 07 14:39:13 <silvio> bluelightning, sorry i can't find but headers go there: install -d ${D}/${exec_prefix}/include/ ?
Sep 07 14:45:55 <bluelightning> silvio: ${includedir}
Sep 07 18:42:36 <kergoth> i kind of hate the new packagegroup- stuff. i named the image features mechanism that largely because i couldn't think of anything else to call it, but i never did like the name :\
Sep 07 20:33:16 <woglinde_> hm the new bitbake task output is nice
Sep 07 20:44:59 <khem> woglinde_: yes 
Sep 07 20:45:06 <khem> woglinde_: hard to catch you these days 
Sep 07 20:45:18 <khem> seems busy for you
Sep 07 20:49:29 <woglinde_> hm I am here
Sep 07 20:49:34 <woglinde_> everyday
Sep 07 20:49:36 <woglinde_> and I have an email
Sep 07 20:53:05 <khem> cool. 
Sep 07 20:53:12 <khem> I posted some openjdk-7 patches
Sep 07 20:53:17 <khem> some time ago
Sep 07 20:53:24 <khem> did you have change to look at them
Sep 07 21:00:20 <khem> woglinde: ?
Sep 07 21:11:03 <prahal2> I get a gcc segfault while I compile : "#include <float.h>\n#include <math.h>\n int main () {double tmp7 =  log10(31.0); return 0;}"  gcc -S input_file.c -lm ; on armv4t gnueabi native ( shr freerunner)
Sep 07 21:46:47 <woglinde> khem yes and I already told you that I will merge the branch into master
Sep 07 21:46:52 <woglinde> and good nite
Sep 07 22:07:26 <khem> prahal2: which gcc version
Sep 07 22:08:04 <prahal2> gcc version 4.7.2 20120706 (prerelease) (GCC)
Sep 07 22:09:58 <khem> prahal2: ok and this is when compiling on freerunner not cross-compile
Sep 07 22:10:25 <prahal2> mind it compiles fine on amrv7a-vfp
Sep 07 22:10:31 <prahal2> yes
Sep 07 22:10:38 <prahal2> cross compile seems fine
Sep 07 22:10:47 <prahal2> it is
Sep 07 22:11:46 <prahal2> yes and same release ... should not cross suffer the same bug ?
Sep 07 22:12:50 <prahal2> on phoenux gta04 (same gcc release)  it is fine too
Sep 07 22:13:31 <prahal2> my guess was that the 31.0 was not seen as a double by the compiler due to a broken patch
Sep 07 22:14:46 <prahal2> because if I do "double a = 31.0; double tmp7 = log10 (a);" all is fine
Sep 07 22:26:57 <prahal2> might also be a local issue (I had an opkg issue which I resolved by reinstall most packages
Sep 07 22:28:16 <prahal2> ie on a full upgrade the states are in /tmp and as tmp is 62M only tmpfs it can get full pretty easily . Then one loose all install state for the packages which were upgrade (which on a full upgrade means pretty much everything
Sep 07 22:43:56 <khem> prahal2: I do not have any armv4t device to test it but if you can reproduce it consistently then file a bug in yocto bugzilla 
Sep 07 22:44:11 <khem> thats the best way to track it
Sep 07 22:51:41 <prahal2> thanks
Sep 07 23:02:04 <prahal2> it turns out the segfault happens in gmp 5.0.5  looking for  a git repo I stumbled upon a github binary repo where one tell "rebuild armeabi binary with -mthumb per NDK specs" itis Rupan/gmp 
**** ENDING LOGGING AT Sat Sep 08 02:59:58 2012