Trying to build qt-X11 for iMX28 EVK board we're evaluating.
using Ubuntu 9.04 per the ltib docs for this board.
TRY 1: Terminates at
Processing: qt-x11
====================
Build path taken because: no prebuilt rpm,
Testing network connectivity
OK GPP:
Try qt-everywhere-opensource-src-4.6.0.tar.gz.md5 from the GPP
http://bitshrine.org/gpp/qt-everywhere-opensource-src-4.6.0.tar.gz.md5:
20:33:23 ERROR 404: Not Found.
Try qt-everywhere-opensource-src-4.6.0.tar.gz from the GPP
http://bitshrine.org/gpp/qt-everywhere-opensource-src-4.6.0.tar.gz:
20:33:23 ERROR 404: Not Found.
Can't get: qt-everywhere-opensource-src-4.6.0.tar.gz at ./ltib line 802.
Okay, so another one where the specified file is not on bitshrine, sigh
download qt-everywhere-opensource-src-4.6.0.tar.gz from fedora
and install into /opt/freescale/pkgs
TRY2:
Okay, try a prep first
./ltib -m prep -p qt-everywhere-opensource-src
hmm, that succeeded. Let's try a ./ltib
No luck. Grepping for errors give me:
cc1plus: error: unrecognized command line option '-mmmx'verywhere-opensource-src-4.6.0/src/xmlpatterns/functions/qerrorfn_p.h
cc1plus: error: unrecognized command line option '-m3dnow'rywhere-opensource-src-4.6.0/src/xmlpatterns/data/qvalidationerror.ccc1plus: error: unrecognized command line option '-msse'
cc1plus: error: unrecognized command line option '-msse2'erywhere-opensource-src-4.6.0/src/xmlpatterns/data/qvalidationerror_plibmng.cpp:42:20: fatal error: libmng.h: No such file or directory
db2.cpp:42:20: fatal error: sqlcli.h: No such file or directoryre-opensource-src-4.6.0/src/scripttools/debugging/qscripterrorlibase.cpp:42:19: fatal error: ibase.h: No such file or directory
../mysql/mysql.cpp:42:19: fatal error: mysql.h: No such file or directoryrce-src-4.6.0/src/scripttools/debugging/qscripterrorlmysql.cpp:42:19: fatal error: mysql.h: No such file or directory
oci.cpp:42:17: fatal error: oci.h: No such file or directorywhere-opensource-src-4.6.0/src/scripttools/debugging/qscripterrorlodbc.cpp:42:17: fatal error: sql.h: No such file or directory
iodbc.cpp:42:17: fatal error: sql.h: No such file or directoryere-opensource-src-4.6.0/src/scripttools/debugging/qscripterrorlpsql.cpp:42:22: fatal error: libpq-fe.h: No such file or directory
sqlite2.cpp:42:20: fatal error: sqlite.h: No such file or directorypensource-src-4.6.0/src/scripttools/debugging/qscripterrorltds.cpp:42:22: fatal error: sybfront.h: No such file or directory
cups.cpp:42:23: fatal error: cups/cups.h: No such file or directory
gnu-libiconv.cpp:54:46: error: invalid conversion from 'const char**' to 'char**' [-fpermissive]
/home/jim/ltib/rootfs/usr/include/iconv.h:83:15: error: initializing argument 2 of 'size_t libiconv(libiconv_t, char**, size_t*, char**, size_t*)' [-fpermissive]
dbus.cpp:43:23: fatal error: dbus/dbus.h: No such file or directory
egl.cpp:42:21: fatal error: EGL/egl.h: No such file or directory
egl4gles1.cpp:42:22: fatal error: GLES/egl.h: No such file or directory
Has anyone gotten qt-x11 to build ok for MX28 ?
or any hints where I should look next ?
Or am I being stupid, and the MX28 can't support Qt on X11 ?
Thanks
Solved! Go to Solution.
The LTIB is a little hard to getting working sometimes depending in several issues. You may want to try Yocto.
Yocto has a good community around it and currently most commonly used SoCs from i.MX platform is already supported by the meta-fsl-arm layer.
You can find more information how to make a build at https://community.freescale.com/docs/DOC-94023
The LTIB is a little hard to getting working sometimes depending in several issues. You may want to try Yocto.
Yocto has a good community around it and currently most commonly used SoCs from i.MX platform is already supported by the meta-fsl-arm layer.
You can find more information how to make a build at https://community.freescale.com/docs/DOC-94023
It took a very long time for Yocto to get set up and compile, even on a quad-core. But it just WORKED!
Using ltib, I was about to give up on the imx28evk board as unsuitable. I am not stupid, but if the vendor-supplied tools and instructions weren't working, how would I ever determine if the fault was in my actions or my code, or if the process itself was too buggy to operate correctly.
Now that I have the build functioning, and QT demo programs working I can start coding and testing.
I think Freescale's docs should just point the user to yocto, and say "this will take some time to set up the first time, but it works smoothly after that.
Thank you for renewing my enthusiasm in this project.
I am glad it worked well; we've been working hard to make Yocto to work as smooth as possible. We still have some known issues but we're working on those. :-)
Good luck with your project.
Thank you very much. I think that it is, indeed, time to abandon ltib as a failed experiment. I'll give yocto a try.