BUG Community

Welcome! Log In

Forums BUG SDK File `jni' does not exist

Subscribe to File `jni' does not exist  22 posts, 6 voices

Log in to reply to this topic
 
May 27, 2009 12:35pm
User_img_not_found_03_med minusw 13 posts

First build, day 3.  I was able to get the kernel to build after resolving my symlink issue.

Now I am trying to build the fiile system:  bitbake bug-image-production

I get the same error on both r1.4 and r1.5.  Here are a few of the messages:

NOTE: package com.buglabs.bug.jni.common-1.0+svnr8657-r3: task do_jni_compile: started

ERROR: function do_jni_compile failed

|  File `jni’ does not exist.

| javadoc: error – fatal error

any thoughts, bk…

May 27, 2009 2:07pm
Medium Bug Labs team jconnolly 285 posts

I’d like to see the content of your logs.  Mine live here:

/home/jconnolly/buglabs/dev/bugbuild/com.buglabs.build.oe/build/tmp/work/bug-poky-linux-gnueabi/com.buglabs.bug.jni.common-1.0+svnr8657-r3/temp 

Example:

 

jconnolly@jconnolly-linux:~/buglabs/dev/bugbuild/com.buglabs.build.oe/build/tmp/work/bug-poky-linux-gnueabi/com.buglabs.bug.jni.common-1.0+svnr8657-r3/temp$ ls log.do_compile.15487 log.do_jni_compile.15500 log.do_removebinaries.20115 run.do_configure.15484 run.do_jni_install.13748 run.do_stage.15610 log.do_configure.15484 log.do_jni_install.13748 log.do_stage.15610 run.do_install.15589 run.do_jni_staging.20039 log.do_install.15589 log.do_jni_staging.20039 run.do_compile.15487 run.do_jni_compile.15500 run.do_removebinaries.20115 jconnolly@jconnolly-linux:~/buglabs/dev/bugbuild/com.buglabs.build.oe/build/tmp/work/bug-poky-linux-gnueabi/com.buglabs.bug.jni.common-1.0+svnr8657-r3/temp$ cat log.do_jni_compile.15500 NOTE: make -e MAKEFLAGS= PHONEME_HOME=/home/jconnolly/buglabs/dev/bugbuild/com.buglabs.build.oe/build/tmp/staging/arm-poky-linux-gnueabi/phoneme-advanced-personal/cdc/build/linux-arm-generic JAVAH_OUT=/home/jconnolly/buglabs/dev/bugbuild/com.buglabs.build.oe/build/tmp/work/bug-poky-linux-gnueabi/com.buglabs.bug.jni.common-1.0+svnr8657-r3/image/usr/lib/jni LIB_DIR=/home/jconnolly/buglabs/dev/bugbuild/com.buglabs.build.oe/build/tmp/work/bug-poky-linux-gnueabi/com.buglabs.bug.jni.common-1.0+svnr8657-r3/image/usr/lib/jni JAVAH_CLASSPATH=/home/jconnolly/buglabs/dev/bugbuild/com.buglabs.build.oe/build/tmp/work/bug-poky-linux-gnueabi/com.buglabs.bug.jni.common-1.0+svnr8657-r3/com.buglabs.bug.jni.common/build --debug COMMON_CLASSPATH=/home/jconnolly/buglabs/dev/bugbuild/com.buglabs.build.oe/build/tmp/staging/arm-poky-linux-gnueabi/usr/lib/com.buglabs.bug.jni.common.jar STAGING_BUG_INCDIR=/home/jconnolly/buglabs/dev/bugbuild/com.buglabs.build.oe/build/tmp/staging/arm-poky-linux-gnueabi/usr/include/bug BUG_LINUX_SRC=/home/jconnolly/buglabs/dev/bugbuild/com.buglabs.build.oe/build/tmp/staging/bug-poky-linux-gnueabi/kernel/ STAGING_INCDIR=/home/jconnolly/buglabs/dev/bugbuild/com.buglabs.build.oe/build/tmp/staging/arm-poky-linux-gnueabi/usr/include jni GNU Make 3.81 Copyright (C) 2006 Free Software Foundation, Inc. This is free software; see the source for copying conditions. There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. This program built for i486-pc-linux-gnu Reading makefiles... Updating goal targets.... File `jni' does not exist. Must remake target `jni'. javah -d /home/jconnolly/buglabs/dev/bugbuild/com.buglabs.build.oe/build/tmp/work/bug-poky-linux-gnueabi/com.buglabs.bug.jni.common-1.0+svnr8657-r3/image/usr/lib/jni -classpath /home/jconnolly/buglabs/dev/bugbuild/com.buglabs.build.oe/build/tmp/work/bug-poky-linux-gnueabi/com.buglabs.bug.jni.common-1.0+svnr8657-r3/com.buglabs.bug.jni.common/build:. com.buglabs.bug.jni.common.CharDevice Successfully remade target file `jni'.

 


May 27, 2009 4:59pm
User_img_not_found_03_med minusw 13 posts

This is becoming very fustrating.  Three days in and nothing to show for it.  Big demo coming up and no progress.

I worked on the first port of unix onto a micro back in the late 70’s and have been writing drivers for unix/linux since.

I can not change my development system at this time.  I am pretty much stuck on this fc9 box for now.

What do you suggest for my best odds of getting a build to succeed.

Should I pull 1.4 or 1.5.

What order should I build:

1. linux-bug

2. phoneme-advanced-foundation

3. phoneme-advanced-personal

4. bitbake bug-image-production

Are there steps I need to take between these builds?

Do I need all the above builds.  I was guessing the phoneme would help resolve a few issues.

bk…

May 27, 2009 5:24pm
Medium Bug Labs team jconnolly 285 posts

This is becoming very fustrating.  Three days in and nothing to show for it.  Big demo coming up and no progress.

I can not change my development system at this time.  I am pretty much stuck on this fc9 box for now.

From:  http://bugcommunity.com/wiki/index.php/Kernel_and_Rootfs_Build_System

System Requirements


Currently we test on Gentoo and Ubuntu GNU/Linux. Support is claimed for other distros.

Sorry you’re experiencing troubles, but just like any other open source project, we need information from you to get through troubleshooting.  Lamentations don’t help me to help you.  Give me logs.

bitbake bug-image-production.

bug-image-testing is not maintained and is a carry-over from early Poky migration.  I would also see poky: http://www.pokylinux.org/doc/poky-handbook.html   (not also that ubuntu/debian based systems are supported, but successful builds are also claimed on Fedora).

bug-image-production includes dependencies for linux-bug, so building linux-bug is extraneous.  If you have problems, myself, kgilmer, finsprings, and especially hrw are knowledgable about stepping through build problems.  Again, any of us are going to ask for your logs.  If you want realtime help, IRC is the best bet.  hrw is in poland and is most often available in the early hours of the morning.  I would also drop into #poky and #oe about build problems on other platforms.

I have also had success building from within virtualbox on ubuntu 8.04 (the most highly supported build OS, what we use on our automated build system).  The build itself, even from within the bloat of a virtual machine still only takes around 6 hours.

As for what you should build, again, like most open source projects, trunk is unstable and bleeding edge.  We are currently most actively working on svn://svn.buglabs.net/bug/branches/R1.4/qa/com.buglabs.build.oe.  So, following the instructions from the wiki:

svn co svn://svn.buglabs.net/bug/branches/R1.4/qa/com.buglabs.build.oe

cd com.buglabs.build.oe

source reinstate-build-env

bitbake bug-image-production

 

Besides the fun of building your own rootfs and kernel, what demo is this for?  Perhaps you’d be better suited with a prebuilt binary.

May 27, 2009 5:53pm
User_img_not_found_03_med minusw 13 posts

Great, I’ll give the 1.4 qa and just the bug-image-production a try.

Now it should matter that I use a different base name when I check the code out.

svn co svn://svn.buglabs.net/bug/branches/R1.4/qa/com.buglabs.build.oe r1.4-qa

So my base dir starts with r1.4-qa and not com.buglabs.build.oe.

 

I need to add a device driver to the kernel for the project I am working on.

That is why I am trying to build from source.

 

The past few times I tried to build the kernel, the build seem to have problems finding fakeroot.

It seem to be searching all over the world and would slowley keep retrying somewhere in japan (I think).

I would have to kill the build and then it would usually succeed the next time.

 

Off for my next try.  I’ll let you know how it goes.

 

thanks for you support, bk…

May 27, 2009 6:34pm
Medium Bug Labs team jconnolly 285 posts

No, it should not matter what your base-dir is.  And now I understand what you mean about fakeroot.  The recipe for fakeroot is using a SRCURI that is either slow or down altogether.  Typically what I do is find the recipe and edit the SRC_URI to be a location I know is quick and reliable.  for example:

 vi meta/packages/fakeroot/fakeroot_1.7.1.bb

changing

${DEBIAN_MIRROR}/main/f/fakeroot/fakeroot${PV}.tar.gz \

to

ftp://64.50.236.52/.1/nslu2/sources/fakeroot_1.7.1.tar.gz \

(don’t forget the trailing slash)

Then bitbake will fetch fakeroot’s source from your provided SRC_URI.

Again, any problems, drop into IRC and one of us will help.

May 27, 2009 9:19pm
User_img_not_found_03_med minusw 13 posts

This time I just tried the bug-image-production.  It did run a lot longer.

I also had another ftp issue and was able to change the url in the meta file and then it worked.

Same issue that I had before.  Here is the last part of my console output:

—-

NOTE: Running task 1023 of 3819 (ID: 371, /work/bamlabs/buglabs/svn/r1.4-qa/meta-bug/packages/bug-osgi/com.buglabs.bug.jni.common.bb, do_jni_compile)
NOTE: package com.buglabs.bug.jni.common-1.0+svnr9157: started
NOTE: package com.buglabs.bug.jni.common-1.0+svnr9157-r5: task do_jni_compile: started
ERROR: function do_jni_compile failed
ERROR: log data follows (/work/bamlabs/buglabs/svn/r1.4-qa/build/tmp/work/bug-poky-linux-gnueabi/com.buglabs.bug.jni.common-1.0+svnr9157-r5/temp/log.do_jni_compile.23624)
| NOTE: make -e MAKEFLAGS= PHONEME_HOME=/work/bamlabs/buglabs/svn/r1.4-qa/build/tmp/staging/arm-poky-linux-gnueabi/phoneme-advanced-personal/cdc/build/linux-arm-generic JAVAH_OUT=/work/bamlabs/buglabs/svn/r1.4-qa/build/tmp/work/bug-poky-linux-gnueabi/com.buglabs.bug.jni.common-1.0+svnr9157-r5/image/usr/lib/jni LIB_DIR=/work/bamlabs/buglabs/svn/r1.4-qa/build/tmp/work/bug-poky-linux-gnueabi/com.buglabs.bug.jni.common-1.0+svnr9157-r5/image/usr/lib/jni JAVAH_CLASSPATH=/work/bamlabs/buglabs/svn/r1.4-qa/build/tmp/work/bug-poky-linux-gnueabi/com.buglabs.bug.jni.common-1.0+svnr9157-r5/com.buglabs.bug.jni.common/build—debug COMMON_CLASSPATH=/work/bamlabs/buglabs/svn/r1.4-qa/build/tmp/staging/arm-poky-linux-gnueabi/usr/lib/com.buglabs.bug.jni.common.jar STAGING_BUG_INCDIR=/work/bamlabs/buglabs/svn/r1.4-qa/build/tmp/staging/arm-poky-linux-gnueabi/usr/include/bug BUG_LINUX_SRC=/work/bamlabs/buglabs/svn/r1.4-qa/build/tmp/staging/bug-poky-linux-gnueabi/kernel/ STAGING_INCDIR=/work/bamlabs/buglabs/svn/r1.4-qa/build/tmp/staging/arm-poky-linux-gnueabi/usr/include jni
| GNU Make 3.81
| Copyright© 2006  Free Software Foundation, Inc.
| This is free software; see the source for copying conditions.
| There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A
| PARTICULAR PURPOSE.
|
| This program built for i386-redhat-linux-gnu
| Reading makefiles…
| Updating goal targets….
|  File `jni’ does not exist.
| Must remake target `jni’.
| javah  -d /work/bamlabs/buglabs/svn/r1.4-qa/build/tmp/work/bug-poky-linux-gnueabi/com.buglabs.bug.jni.common-1.0+svnr9157-r5/image/usr/lib/jni -classpath /work/bamlabs/buglabs/svn/r1.4-qa/build/tmp/work/bug-poky-linux-gnueabi/com.buglabs.bug.jni.common-1.0+svnr9157-r5/com.buglabs.bug.jni.common/build:. com.buglabs.bug.jni.common.CharDevice
| java.lang.NoClassDefFoundError: com.sun.source.tree.Tree not found in java.lang.ClassLoader$1{urls=[file:/work/bamlabs/buglabs/svn/r1.4-qa/build/tmp/staging/i686-linux/usr/lib/tools.jar], parent=null}
|    at java.lang.VMClassLoader.defineClass(Native Method)
|    at java.lang.VMClassLoader.defineClassWithTransformers(VMClassLoader.java:454)
|    at java.lang.ClassLoader.defineClass(ClassLoader.java:471)
|    at java.security.SecureClassLoader.defineClass(SecureClassLoader.java:83)
|    at java.net.URLClassLoader.findClass(URLClassLoader.java:615)
|    at java.lang.ClassLoader.loadClass(ClassLoader.java:341)
|    at java.lang.ClassLoader$1.loadClass(ClassLoader.java:1112)
|    at java.lang.ClassLoader.loadClass(ClassLoader.java:293)
|    at com.sun.tools.javac.tree.JCTree$Visitor.<clinit>(JCTree.java:2129)
|    at com.sun.tools.javadoc.JavadocTool.make0(JavadocTool.java:89)
|    at com.sun.tools.javadoc.Start.parseAndExecute(Start.java:340)
|    at com.sun.tools.javadoc.Start.begin(Start.java:146)
|    at com.sun.tools.javadoc.Main.execute(Main.java:84)
|    at com.sun.tools.javah.Main.main(Main.java:153)
| Caused by: java.lang.ClassNotFoundException: com.sun.source.tree.Tree not found in java.lang.ClassLoader$1{urls=[file:/work/bamlabs/buglabs/svn/r1.4-qa/build/tmp/staging/i686-linux/usr/lib/tools.jar], parent=null}
|    at java.net.URLClassLoader.findClass(URLClassLoader.java:529)
|    at java.lang.ClassLoader.loadClass(ClassLoader.java:341)
|    at java.lang.ClassLoader$1.loadClass(ClassLoader.java:1112)
|    at java.lang.ClassLoader.loadClass(ClassLoader.java:293)
|    at java.lang.VMClassLoader.defineClass(Native Method)
|    ...13 more
| javadoc: error – fatal error
| 1 error
| make: [jni] Error 1
| FATAL: oe_runmake failed
NOTE: Task failed: /work/bamlabs/buglabs/svn/r1.4-qa/build/tmp/work/bug-poky-linux-gnueabi/com.buglabs.bug.jni.common-1.0+svnr9157-r5/temp/log.do_jni_compile.23624
NOTE: package com.buglabs.bug.jni.common-1.0+svnr9157-r5: task do_jni_compile: failed
ERROR: TaskFailed event exception, aborting
NOTE: package com.buglabs.bug.jni.common-1.0+svnr9157: failed
ERROR: Build of /work/bamlabs/buglabs/svn/r1.4-qa/meta-bug/packages/bug-osgi/com.buglabs.bug.jni.common.bb do_jni_compile failed
ERROR: Task 371 (/work/bamlabs/buglabs/svn/r1.4-qa/meta-bug/packages/bug-osgi/com.buglabs.bug.jni.common.bb, do_jni_compile) failed
NOTE: Tasks Summary: Attempted 1022 tasks of which 0 didn’t need to be rerun and 1 failed.
ERROR: ’/work/bamlabs/buglabs/svn/r1.4-qa/meta-bug/packages/bug-osgi/com.buglabs.bug.jni.common.bb’ failed

——

and the log file is:

 

NOTE: make -e MAKEFLAGS= PHONEME_HOME=/work/bamlabs/buglabs/svn/r1.4-qa/build/tmp/staging/arm-poky-linux-gnueabi/phoneme-advanced-personal/cdc/build/linux-arm-generic JAVAH_OUT=/work/bamlabs/buglabs/svn/r1.4-qa/build/tmp/work/bug-poky-linux-gnueabi/com.buglabs.bug.jni.common-1.0+svnr9157-r5/image/usr/lib/jni LIB_DIR=/work/bamlabs/buglabs/svn/r1.4-qa/build/tmp/work/bug-poky-linux-gnueabi/com.buglabs.bug.jni.common-1.0+svnr9157-r5/image/usr/lib/jni JAVAH_CLASSPATH=/work/bamlabs/buglabs/svn/r1.4-qa/build/tmp/work/bug-poky-linux-gnueabi/com.buglabs.bug.jni.common-1.0+svnr9157-r5/com.buglabs.bug.jni.common/build—debug COMMON_CLASSPATH=/work/bamlabs/buglabs/svn/r1.4-qa/build/tmp/staging/arm-poky-linux-gnueabi/usr/lib/com.buglabs.bug.jni.common.jar STAGING_BUG_INCDIR=/work/bamlabs/buglabs/svn/r1.4-qa/build/tmp/staging/arm-poky-linux-gnueabi/usr/include/bug BUG_LINUX_SRC=/work/bamlabs/buglabs/svn/r1.4-qa/build/tmp/staging/bug-poky-linux-gnueabi/kernel/ STAGING_INCDIR=/work/bamlabs/buglabs/svn/r1.4-qa/build/tmp/staging/arm-poky-linux-gnueabi/usr/include jni
GNU Make 3.81
Copyright© 2006  Free Software Foundation, Inc.
This is free software; see the source for copying conditions.
There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A
PARTICULAR PURPOSE.

This program built for i386-redhat-linux-gnu
Reading makefiles…
Updating goal targets….
 File `jni’ does not exist.
Must remake target `jni’.
javah  -d /work/bamlabs/buglabs/svn/r1.4-qa/build/tmp/work/bug-poky-linux-gnueabi/com.buglabs.bug.jni.common-1.0+svnr9157-r5/image/usr/lib/jni -classpath /work/bamlabs/buglabs/svn/r1.4-qa/build/tmp/work/bug-poky-linux-gnueabi/com.buglabs.bug.jni.common-1.0+svnr9157-r5/com.buglabs.bug.jni.common/build:. com.buglabs.bug.jni.common.CharDevice
java.lang.NoClassDefFoundError: com.sun.source.tree.Tree not found in java.lang.ClassLoader$1{urls=[file:/work/bamlabs/buglabs/svn/r1.4-qa/build/tmp/staging/i686-linux/usr/lib/tools.jar], parent=null}
   at java.lang.VMClassLoader.defineClass(Native Method)
   at java.lang.VMClassLoader.defineClassWithTransformers(VMClassLoader.java:454)
   at java.lang.ClassLoader.defineClass(ClassLoader.java:471)
   at java.security.SecureClassLoader.defineClass(SecureClassLoader.java:83)
   at java.net.URLClassLoader.findClass(URLClassLoader.java:615)
   at java.lang.ClassLoader.loadClass(ClassLoader.java:341)
   at java.lang.ClassLoader$1.loadClass(ClassLoader.java:1112)
   at java.lang.ClassLoader.loadClass(ClassLoader.java:293)
   at com.sun.tools.javac.tree.JCTree$Visitor.<clinit>(JCTree.java:2129)
   at com.sun.tools.javadoc.JavadocTool.make0(JavadocTool.java:89)
   at com.sun.tools.javadoc.Start.parseAndExecute(Start.java:340)
   at com.sun.tools.javadoc.Start.begin(Start.java:146)
   at com.sun.tools.javadoc.Main.execute(Main.java:84)
   at com.sun.tools.javah.Main.main(Main.java:153)
Caused by: java.lang.ClassNotFoundException: com.sun.source.tree.Tree not found in java.lang.ClassLoader$1{urls=[file:/work/bamlabs/buglabs/svn/r1.4-qa/build/tmp/staging/i686-linux/usr/lib/tools.jar], parent=null}
   at java.net.URLClassLoader.findClass(URLClassLoader.java:529)
   at java.lang.ClassLoader.loadClass(ClassLoader.java:341)
   at java.lang.ClassLoader$1.loadClass(ClassLoader.java:1112)
   at java.lang.ClassLoader.loadClass(ClassLoader.java:293)
   at java.lang.VMClassLoader.defineClass(Native Method)
   ...13 more
javadoc: error – fatal error
1 error
make:
[jni] Error 1
FATAL: oe_runmake failed


May 27, 2009 10:14pm
Medium Bug Labs team jconnolly 285 posts

yeah, hrw has dealt with this.  it’s the result of another java recipe staging its tools.jar over the appropriate one.  it may help to bitbake phoneme-advanced-personal -c rebuild, but I’ll get back to you on this tomorrow AM.

May 28, 2009 11:27am
Medium Bug Labs team jconnolly 285 posts

yeah, hrw has dealt with this.  it’s the result of another java recipe staging its tools.jar over the appropriate one.  it may help to bitbake phoneme-advanced-personal -c rebuild, but I’ll get back to you on this tomorrow AM.

Ok I have a few simple steps to try:

bitbake openjdk-langtools-native -c rebuild

then

bitbake com.buglabs.bug.jni.common -c rebuild

if it completes,

bitbake bug-image-production

 

If not, from within your r1.4-qa dir:

rm build/tmp/staging/i686-linux/usr/lib/tools.jar

cp build/tmp/work/i686-linux/openjdk-langtools-native-0.0+jdk7-b31-r3/icepick-0.0+hg20080118/tools.jar build/tmp/staging/i686-linux/usr/lib/

then

bitbake com.buglabs.bug.jni.common -c rebuild

if it completes,

bitbake bug-image-production

 

If this doesn’t work, let me know and  I’ll put my tools.jar up for you to put into build/tmp/staging/i686-linux/usr/lib/

May 28, 2009 11:49am
User_img_not_found_03_med minusw 13 posts

Both com.buglabs.bug.jni.common rebuilds failed:

Let me know if you want or need logs.

and way big thanks for looking into this.  bk…

May 28, 2009 11:54am
Medium Bug Labs team jconnolly 285 posts

ok I put my tools.jar up on bugcommunity.com for you.  http://bugcommunity.com/downloads/files/tools.jar .  and it should go into build/tmp/staging/i686-linux/usr/lib/

your openjdk-langtools-native is b0rked but this should fix it.

May 28, 2009 12:12pm
User_img_not_found_03_med minusw 13 posts

Yeah, back up and running again at 1050 of 3819 tasks.

Was this an issue with a tool on my box outside of your tree?

Is this something I should fix on my box?

way way big thanks again and I will let you know if the build ever completes.

btw: if you want to contact me and not clutter up this forum,

my email: bk@minusw.com

 

May 28, 2009 12:30pm
Medium Bug Labs team jconnolly 285 posts

No, this is a problem related to the ordering of tasks by bitbake in the build.  Something is subtly broken within the build whereby if one recipe executes before the other, the result is the problem you’re seeing.  It’s not a function of your system.  Also I’m happy to help.  I’m going to put up a blogpost about how to compile the kernel as part of a one-off (not part of the ginormous build system) if you’ve got our arm toolchain.  This should help you more easily prototype your driver.

The .bug.jni. are the easiest to break in my experience, so fingers crossed you should be home free after this hiccup.

 

 

May 28, 2009 1:13pm
User_img_not_found_03_med minusw 13 posts

Getting a little closer. curl-native is the next thing failing:

NOTE: Running task 1185 of 3819 (ID: 3774, /work/bamlabs/buglabs/svn/r1.4-qa/meta/packages/curl/curl-native_7.16.2.bb, do_compile)
NOTE: package curl-native-7.16.2: started
NOTE: package curl-native-7.16.2-r2: task do_compile: started
ERROR: function do_compile failed

—-

| ccache gcc -isystem/work/bamlabs/buglabs/svn/r1.4-qa/build/tmp/staging/i686-linux/usr/include -O2 -Wl,-rpath-link -Wl,/work/bamlabs/buglabs/svn/r1.4-qa/build/tmp/staging/i686-linux/usr/lib -Wl,-rpath -Wl,/work/bamlabs/buglabs/svn/r1.4-qa/build/tmp/staging/i686-linux/usr/lib -Wl,-O1 -o .libs/curl main.o hugehelp.o urlglob.o writeout.o writeenv.o getpass.o homedir.o curlutil.o strtoofft.o strdup.o  -L/work/bamlabs/buglabs/svn/r1.4-qa/build/tmp/staging/i686-linux/usr/lib -L/work/bamlabs/buglabs/svn/r1.4-qa/build/tmp/staging/i686-linux/usr//lib ../lib/.libs/libcurl.so -lssh2 -ldl /work/bamlabs/buglabs/svn/r1.4-qa/build/tmp/staging/i686-linux/usr/lib/libz.so -Wl,—rpath -Wl,/work/bamlabs/buglabs/svn/r1.4-qa/build/tmp/staging/i686-linux/usr/lib
| ../lib/.libs/libcurl.so: undefined reference to `libssh2_channel_writenb’
| ../lib/.libs/libcurl.so: undefined reference to `libssh2_sftp_readnb’
| ../lib/.libs/libcurl.so: undefined reference to `libssh2_channel_readnb’

May 28, 2009 1:24pm
Medium Bug Labs team jconnolly 285 posts

looks like it’s an issue with curl and libssh2 versions… http://bugs.sourcemage.org/show_bug.cgi?id=13963

let me see if I can reproduce your problem locally.

May 28, 2009 1:45pm
Medium Bug Labs team jconnolly 285 posts

for some reason it’s trying to link libssh2 when it shouldn’t be.  the recipe specifies: ”—without-ssl”.   I think your do_configure stage didn’t go right.  You can try editing the curl makefile yourself.  grep through build/tmp/work/i686-linux/curl-native-7.16.2-r2/curl-7.16.2/Makefile for ssh2, then remove -lssh2 from the makefile.  Try

bitbake curl-native (NOT -c rebuild) and try again.  That should help.  I’m also on AIM as john {AT} buglabs.net if you want.

 

May 28, 2009 5:28pm
Medium Bug Labs team jconnolly 285 posts

for the record, the curl recipe needed to be modified:

vi meta/packages/curl/curl_7.16.2.bb
after—without-ssl \
add—without-libssh2 \
bitbake curl-native -c rebuild

Aug 14, 2012 2:10am
User_img_not_found_06_med rubing 103 posts

Isn’t it funny how we www.coachfactorystoreec.com/ Coach Factory Store always want what we can’t have? The girl with the round and pert backside wishes that her rear were 5 times smaller. Meanwhile, the girl with the non existent posterior wishes to achieve a larger butt. Well, if the grass is greener on the other side, just water yours! If you’ve got a small behind, you can create the illusion of a larger one with butt enhancing jeans. Here’s how to do it. Look for light weight denim – This material is very stiff and heavy. If you try to stuff a flat butt into denim jeans, it will only create a flatter looking behind. A little stretch goes a long way – Your butt enhancing jeans should additionally contain some stretch. A mixture of 98 percent cotton and 2 percent lycra is best. This combination of lycra and cotton, especially when it’s in a nice tight-weave fabric that’s fitted through the hip and thigh area will help lift your buttocks. Too much lycra and your jeans may start to ripple like a pair of leggings. If there’s not enough cotton content, there won’t be enough structure to hold you up. It’s all in the pockets – The most effective trick for creating the illusion of a bigger butt is to choose jeans with flap pockets. The added bulk of the flap will enhance and add dimension to your tiny tush. True Religion is notoriously famous for their flap pocket designs. You can also try Hudson, Rock & Republic, Frankie B and Paige for addition flap pocket styles. Pocket placement – If you’re not a fan of flap pockets, or if you’ve already bought too many of them, you can also try denim with regular pockets. Be sure to aim for a pair with smaller pockets that sit higher and this will help make your rear look lifted. Cut is also important – Look for jeans with a slight tapering at www.coachfactoryonliness.com/ Coach Factory Online the knee, such as a straight leg or skinny style. This cut will make your bum look a bit larger in comparison to the rest of your body. Great styles to try are True Religion Billy, Julie or Stella, Siwy Harmah or Rose, and Hudson Emily or Paige. How low can you go – Actually, you don’t want to go too low when it comes to the jean rise. The goal is to project more, not less. A pair of low rise denim will only reduce the size of your rear. Instead, opt for a pair with a medium rise. Color me right – Since dark colors minimize, dark denim and indigo will create a slimming effect. You do not want this kind of effect on your barely there behind! Try to find light colored or medium blue denim. Grey is very flattering and will also work. Like a great bra, butt enhancing jeans can also lift and www.coachfactoryoutletusa.com/ Coach Factory Outlet separate what little you have. So it wasn’t in your genes to have a nice, round bottom. Big deal. Be happy with what you’ve got and just improvise!

Oct 3, 2012 2:40am
User_img_not_found_02_med wunai 103 posts

Thanks to the answers was possible to Tods shoes evaluate the intensity of use and the question of the distinction in purchasing. Women need a bag to carry their belongings, that’s for sure. Leather bags, women nowTods UK  have the flexibility and class beyond what I ever thought. The serious fashion lovers will love the leather bags thanks to its versatility, high quality and elegant style. With brands like Tod tods women shoes, Cole Haan,Tods Outlet   Rebecca Minkoff, among many others, is nothing but leather bags and upscale beyond amazing. Hogan shoes outlet The design is based on a conceptLouis Vuitton UK of minimal interference between your feet and the floor.

Oct 24, 2012 1:46am
User_img_not_found_01_med qazqazqaz 469 posts

As a perfect combination of classic and modern fashion,Coach Outlet Online Julia Bags can show the customers'unique personality.Tiffany UK grabbed the globe, stands out as the celeb on their favorite stars luxury type, granted an innovative simple move the pavement luxury, complete with luxury.In terms of the quality and superior design that make more and more customers are satisfied to Coach Factory Online.Welcome!hat experts claim Coach Outlet Online shopping is in the changes they are available in.As long as you open our Coach Factory Outlet webpage, you will view various Coach New Arrivals which are the most popular also the most fashionable in this year.On the date when the Louis Vuitton UK on sale, all people will pour into the stores. At this time, it on the booths will be wiped out, and only one or two days the seller will declare that there is no stock anymore.

Oct 24, 2012 4:16am
User_img_not_found_01_med qazqazqaz 469 posts

Remember that the right accessories can enhance your alluring look with your Canada Goose Parkas Outlet.Baggy, ovoid shaped Canada Goose Jackets Outlet that are rounded around the shoulders look fabulous and can be a perfect match for you.This excellent conviction so that you class gets to amateur Canada Goose Outlet In a number of Parka, for example Young Canada Goose affordable Cardigan.Luxury has always been synonymous with fur and leather and this year a faux fur Canada Goose Jackets Outlet will surely be the key piece of your wardrobe.

Nov 1, 2012 3:52am
User_img_not_found_01_med yiyi 74 posts

seven bad years (I don’t normally believe in hocus-pocus horoscopes), so far it
has been right,” Rausing wrote. “I’m “I’m cheap jordan shoes “I’m
hoping for seven good years starting 2007.”The couple have children
together.Nazi War Criminal Laszlo Csatary Taken Into Custody in in cheap jordans in
BudapestLaszlo Csatary Lived Openly Under His Own Name in DRAGANA JOVANOVICJuly
18, 2012— The past may have finally caught caught

cheap replica designer
handbags

caught up with a 97-year-old Nazi war criminal when Laszlo was
taken into custody today in Budapest.Csatary, who has been been moncler outlet been
convicted in absentia and sentenced to death, was picked early today, Bettina
Bagoly, spokeswoman for the Budapest prosecutor, told told

knockoff handbags

told ABC
News.The elderly former Nazi “was accused with war crimes,” according to a
statement released by the prosecutor’s prosecutor’s

cheap jordan
shoes

prosecutor’s office.Csatary will be taken before an investigative
judge later and the “prosecution office will initiate the house arrest,” the the

cheap moncler
jackets

the statement said.Bagoly said the prosecutor’s office is still
investigating against Csatary, and the office statement indicated that the probe
probe moncler down
jackets
probe began in September 2011 at the instigation of the
Wiesenthal Center in Israel which specializes in hunting Nazi war war

cheap moncler
jackets

war criminals.The prosecutor’s statement said they were warned
by the Center that Csatary “will leave the country as soon as as cheap real air jordans as he
realizes that there is ongoing investigation in the prosecutors said the
investigation is complicated by the age of of

knockoff handbags

of the crime
and that it was committed in Slovakia, are in Israel

Log in to reply to this topic
Forums BUG SDK File `jni' does not exist

Powered by Community Engine

Top
Login
Close
Bottom