From: Peter A. <pet...@ya...> - 2008-02-13 20:07:26
|
Hi, I'm trying to get java running on a gumstix. The instructions I have found online look like they are a little bit out of date since OE is being used now, but they are a good starting point. I was wondering if anyone has succeeded in getting java on a gumstix recently. Right now I'm trying to compile GNU classpath (version 0.96) , but I'm getting a "CROSS COMPILE Badness" error. The compiler I am using is "gumstix-oe/tmp/cross/bin/arm-angstrom-linux-gnueabi-gcc" and judging by the results of my google search this is an OE issue. However, I can't find any solutions to it. Any suggestions? Thanks, - Peter --------------------------------- Never miss a thing. Make Yahoo your homepage. |
From: Steve S. <sa...@gm...> - 2008-02-13 20:32:45
|
Peter, Looks like you've already set up gumstix-oe, so just type: bitbake classpath I believe it will build 0.95, but it is a good place to get started. I haven't had a chance to try any java stuff yet, so you will be a pioneer :-) Steve On Feb 13, 2008 12:07 PM, Peter Abeles <pet...@ya...> wrote: > Hi, > > I'm trying to get java running on a gumstix. The instructions I have found > online look like they are a little bit out of date since OE is being used > now, but they are a good starting point. I was wondering if anyone has > succeeded in getting java on a gumstix recently. > > Right now I'm trying to compile GNU classpath (version 0.96) , but I'm > getting a "CROSS COMPILE Badness" error. The compiler I am using is > "gumstix-oe/tmp/cross/bin/arm-angstrom-linux-gnueabi-gcc" and judging by the > results of my google search this is an OE issue. However, I can't find any > solutions to it. > > Any suggestions? > > Thanks, > - Peter > > > ________________________________ > Never miss a thing. Make Yahoo your homepage. > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2008. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > _______________________________________________ > gumstix-users mailing list > gum...@li... > https://lists.sourceforge.net/lists/listinfo/gumstix-users > > |
From: Peter A <pet...@ya...> - 2008-02-14 15:09:44
|
Steve, Thanks for the reply. I did as you said and entered "bitbake classpath" but the build failed. There was a bunch of errors, I have pasted some of them below: - Peter --------- ERROR MESSAGE START -------------------- Patch gconf_version.patch does not apply (enforce with -f) ERROR: Printing the environment of the function ERROR: Error in executing: ERROR: Exception:<class 'CmdError'> Message:Command Error: exit status: 1 Output: Applying patch gconf_version.patch can't find file to patch at input line 3 Perhaps you used the wrong -p or --strip option? The text leading up to this was: -------------------------- |--- classpath-0.93/configure.ac~ 2007-03-29 15:52:30.000000000 -0500 |+++ classpath-0.93/configure.ac 2007-03-29 15:52:30.000000000 -0500 -------------------------- No file to patch. Skipping patch. 1 out of 1 hunk ignored Patch gconf_version.patch does not apply (enforce with -f) ERROR: Printing the environment of the function ERROR: Build of /home/pabeles/gumstix/gumstix-oe/org.openembedded.snapshot/packages/classpath/classpath-minimal_0.90.bb do_patch failed Traceback (most recent call last): File "/home/pabeles/gumstix/gumstix-oe/bitbake/bin/bitbake", line 134, in <module> main() File "/home/pabeles/gumstix/gumstix-oe/bitbake/bin/bitbake", line 131, in main cooker.cook() File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/cooker.py", line 552, in cook return self.buildTargets(pkgs_to_build) File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/cooker.py", line 466, in buildTargets failures = rq.execute_runqueue() File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/runqueue.py", line 403, in execute_runqueue self.execute_runqueue_internal() File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/runqueue.py", line 531, in execute_runqueue_internal self.cooker.tryBuild(fn, False) File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/cooker.py", line 125, in tryBuild return self.tryBuildPackage(fn, item, self.configuration.cmd, the_data, build_depends) File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/cooker.py", line 98, in tryBuildPackage bb.build.exec_task('do_%s' % task, the_data) File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/build.py", line 292, in exec_task execute(None, task) File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/build.py", line 281, in execute exec_func(item, localdata) File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/build.py", line 98, in exec_func exec_func_python(func, d) File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/build.py", line 117, in exec_func_python utils.better_exec(comp,g,tmp, bb.data.getVar('FILE',d,1)) File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/utils.py", line 144, in better_exec exec code in context File "do_patch", line 4, in <module> File "do_patch", line 2, in do_patch File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/build.py", line 98, in exec_func exec_func_python(func, d) File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/build.py", line 117, in exec_func_python utils.better_exec(comp,g,tmp, bb.data.getVar('FILE',d,1)) File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/utils.py", line 144, in better_exec exec code in context File "base_do_patch", line 4, in <module> File "base_do_patch", line 2, in base_do_patch File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/build.py", line 98, in exec_func exec_func_python(func, d) File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/build.py", line 117, in exec_func_python utils.better_exec(comp,g,tmp, bb.data.getVar('FILE',d,1)) File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/utils.py", line 144, in better_exec exec code in context File "patch_do_patch", line 100, in <module> File "patch_do_patch", line 98, in patch_do_patch File "<bb>", line 339, in Resolve __builtin__.CmdError: Command Error: exit status: 1 Output: Applying patch gconf_version.patch can't find file to patch at input line 3 Perhaps you used the wrong -p or --strip option? The text leading up to this was: -------------------------- |--- classpath-0.93/configure.ac~ 2007-03-29 15:52:30.000000000 -0500 |+++ classpath-0.93/configure.ac 2007-03-29 15:52:30.000000000 -0500 -------------------------- No file to patch. Skipping patch. 1 out of 1 hunk ignored Patch gconf_version.patch does not apply (enforce with -f) ERROR: Task 202 (/home/pabeles/gumstix/gumstix-oe/org.openembedded.snapshot/packages/classpath/classpath-minimal_0.90.bb, do_patch) failed NOTE: Tasks Summary: Attempted 640 tasks of which 288 didn't need to be rerun and 1 failed. ERROR: '/home/pabeles/gumstix/gumstix-oe/org.openembedded.snapshot/packages/classpath/classpath-minimal_0.90.bb' failed -- View this message in context: http://www.nabble.com/gumstix-%2B-OE-%2B-java-tp15466829p15481201.html Sent from the Gumstix mailing list archive at Nabble.com. |
From: Steve S. <sa...@gm...> - 2008-02-14 15:46:19
|
Peter, Well, I said you were going to be a pioneer :-) I'll look at updating the classpath package. Stay tuned . . . Steve On Thu, Feb 14, 2008 at 7:09 AM, Peter A <pet...@ya...> wrote: > > Steve, > > Thanks for the reply. I did as you said and entered "bitbake classpath" but > the build failed. There was a bunch of errors, I have pasted some of them > below: > > - Peter > > --------- ERROR MESSAGE START -------------------- > > Patch gconf_version.patch does not apply (enforce with -f) > ERROR: Printing the environment of the function > ERROR: Error in executing: > ERROR: Exception:<class 'CmdError'> Message:Command Error: exit status: 1 > Output: > Applying patch gconf_version.patch > can't find file to patch at input line 3 > Perhaps you used the wrong -p or --strip option? > The text leading up to this was: > -------------------------- > |--- classpath-0.93/configure.ac~ 2007-03-29 15:52:30.000000000 -0500 > |+++ classpath-0.93/configure.ac 2007-03-29 15:52:30.000000000 -0500 > -------------------------- > No file to patch. Skipping patch. > 1 out of 1 hunk ignored > Patch gconf_version.patch does not apply (enforce with -f) > ERROR: Printing the environment of the function > ERROR: Build of > /home/pabeles/gumstix/gumstix-oe/org.openembedded.snapshot/packages/classpath/classpath-minimal_0.90.bb > do_patch failed > Traceback (most recent call last): > File "/home/pabeles/gumstix/gumstix-oe/bitbake/bin/bitbake", line 134, in > <module> > main() > File "/home/pabeles/gumstix/gumstix-oe/bitbake/bin/bitbake", line 131, in > main > cooker.cook() > File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/cooker.py", line > 552, in cook > return self.buildTargets(pkgs_to_build) > File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/cooker.py", line > 466, in buildTargets > failures = rq.execute_runqueue() > File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/runqueue.py", line > 403, in execute_runqueue > self.execute_runqueue_internal() > File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/runqueue.py", line > 531, in execute_runqueue_internal > self.cooker.tryBuild(fn, False) > File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/cooker.py", line > 125, in tryBuild > return self.tryBuildPackage(fn, item, self.configuration.cmd, the_data, > build_depends) > File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/cooker.py", line 98, > in tryBuildPackage > bb.build.exec_task('do_%s' % task, the_data) > File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/build.py", line 292, > in exec_task > execute(None, task) > File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/build.py", line 281, > in execute > exec_func(item, localdata) > File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/build.py", line 98, > in exec_func > exec_func_python(func, d) > File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/build.py", line 117, > in exec_func_python > utils.better_exec(comp,g,tmp, bb.data.getVar('FILE',d,1)) > File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/utils.py", line 144, > in better_exec > exec code in context > File "do_patch", line 4, in <module> > File "do_patch", line 2, in do_patch > File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/build.py", line 98, > in exec_func > exec_func_python(func, d) > File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/build.py", line 117, > in exec_func_python > utils.better_exec(comp,g,tmp, bb.data.getVar('FILE',d,1)) > File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/utils.py", line 144, > in better_exec > exec code in context > File "base_do_patch", line 4, in <module> > File "base_do_patch", line 2, in base_do_patch > File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/build.py", line 98, > in exec_func > exec_func_python(func, d) > File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/build.py", line 117, > in exec_func_python > utils.better_exec(comp,g,tmp, bb.data.getVar('FILE',d,1)) > File "/home/pabeles/gumstix/gumstix-oe/bitbake/lib/bb/utils.py", line 144, > in better_exec > exec code in context > File "patch_do_patch", line 100, in <module> > File "patch_do_patch", line 98, in patch_do_patch > File "<bb>", line 339, in Resolve > __builtin__.CmdError: Command Error: exit status: 1 Output: > Applying patch gconf_version.patch > can't find file to patch at input line 3 > Perhaps you used the wrong -p or --strip option? > The text leading up to this was: > -------------------------- > |--- classpath-0.93/configure.ac~ 2007-03-29 15:52:30.000000000 -0500 > |+++ classpath-0.93/configure.ac 2007-03-29 15:52:30.000000000 -0500 > -------------------------- > No file to patch. Skipping patch. > 1 out of 1 hunk ignored > Patch gconf_version.patch does not apply (enforce with -f) > ERROR: Task 202 > (/home/pabeles/gumstix/gumstix-oe/org.openembedded.snapshot/packages/classpath/classpath-minimal_0.90.bb, > do_patch) failed > NOTE: Tasks Summary: Attempted 640 tasks of which 288 didn't need to be > rerun and 1 failed. > ERROR: > '/home/pabeles/gumstix/gumstix-oe/org.openembedded.snapshot/packages/classpath/classpath-minimal_0.90.bb' > failed > > -- > View this message in context: http://www.nabble.com/gumstix-%2B-OE-%2B-java-tp15466829p15481201.html > Sent from the Gumstix mailing list archive at Nabble.com. > > > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2008. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > _______________________________________________ > gumstix-users mailing list > gum...@li... > https://lists.sourceforge.net/lists/listinfo/gumstix-users > |
From: Peter A <pet...@ya...> - 2008-02-14 22:17:21
|
Steve, FYI I tried to build classpath 0.95 from source code I downloaded from their website and got the same error that I did with 0.96. So unless I am doing something wrong, I suspect you will get the same error also once the build gets fixed. Just in case here is what I am doing and the error message. I was also going to post a bug report on OE's website, but they have yet to verify the e-mail address... Thanks, - Peter -------- COMMANDS --------------- $ cd classpath-0.95/ $ export PATH=$PATH:~/gumstix/gumstix-oe/tmp/cross/bin $ ./configure --prefix=/home/pabeles/gumstix/tmp CC=arm-angstrom-linux-gnueabi-gcc CPP=arm-angstrom-linux-gnueabi-cpp CXX=arm-angstrom-linux-gnueabi-g++ RANLIB=arm-angstrom-linux-gnueabi-ranlib --host=arm-linux --disable-gtk-peer --disable-plugin --with-ecj $ make ------- ERROR MESSAGE ------------- mkdir .libs arm-angstrom-linux-gnueabi-gcc -DHAVE_CONFIG_H -I. -I../../../include -I../../../include -I../../../native/jni/classpath -I../../../native/jni/native-lib -W -Wall -Wmissing-declarations -Wwrite-strings -Wmissing-prototypes -Wno-long-long -Werror -DORBIT2=1 -pthread -I/usr/include/gconf/2 -I/usr/include/orbit-2.0 -I/usr/include/glib-2.0 -I/usr/lib/glib-2.0/include -I/usr/include/gtk-2.0 -I/usr/lib/gtk-2.0/include -I/usr/include/pango-1.0 -I/usr/include/glib-2.0 -I/usr/lib/glib-2.0/include -I/usr/include/cairo -I/usr/include/freetype2 -I/usr/include/libpng12 -g -O2 -MT GConfNativePeer.lo -MD -MP -MF .deps/GConfNativePeer.Tpo -c GConfNativePeer.c -fPIC -DPIC -o .libs/GConfNativePeer.o CROSS COMPILE Badness: /usr/include in INCLUDEPATH: /usr/include/gconf/2 cc1: internal compiler error: in add_path, at c-incpath.c:362 Please submit a full bug report, with preprocessed source if appropriate. See <URL:http://gcc.gnu.org/bugs.html> for instructions. make[3]: *** [GConfNativePeer.lo] Error 1 make[3]: Leaving directory `/home/pabeles/gumstix/java/classpath-0.95/native/jni/gconf-peer' make[2]: *** [all-recursive] Error 1 make[2]: Leaving directory `/home/pabeles/gumstix/java/classpath-0.95/native/jni' make[1]: *** [all-recursive] Error 1 make[1]: Leaving directory `/home/pabeles/gumstix/java/classpath-0.95/native' -- View this message in context: http://www.nabble.com/gumstix-%2B-OE-%2B-java-tp15466829p15490517.html Sent from the Gumstix mailing list archive at Nabble.com. |
From: Steve S. <sa...@gm...> - 2008-02-14 22:55:09
|
Peter, Have you tried running with -O1 or -O0? That might get past the compiler error. Steve On Thu, Feb 14, 2008 at 2:17 PM, Peter A <pet...@ya...> wrote: > > Steve, > > FYI I tried to build classpath 0.95 from source code I downloaded from their > website and got the same error that I did with 0.96. So unless I am doing > something wrong, I suspect you will get the same error also once the build > gets fixed. Just in case here is what I am doing and the error message. I > was also going to post a bug report on OE's website, but they have yet to > verify the e-mail address... > > Thanks, > - Peter > > -------- COMMANDS --------------- > > $ cd classpath-0.95/ > $ export PATH=$PATH:~/gumstix/gumstix-oe/tmp/cross/bin > $ ./configure --prefix=/home/pabeles/gumstix/tmp > CC=arm-angstrom-linux-gnueabi-gcc CPP=arm-angstrom-linux-gnueabi-cpp > CXX=arm-angstrom-linux-gnueabi-g++ RANLIB=arm-angstrom-linux-gnueabi-ranlib > --host=arm-linux --disable-gtk-peer --disable-plugin --with-ecj > $ make > > ------- ERROR MESSAGE ------------- > > mkdir .libs > arm-angstrom-linux-gnueabi-gcc -DHAVE_CONFIG_H -I. -I../../../include > -I../../../include -I../../../native/jni/classpath > -I../../../native/jni/native-lib -W -Wall -Wmissing-declarations > -Wwrite-strings -Wmissing-prototypes -Wno-long-long -Werror -DORBIT2=1 > -pthread -I/usr/include/gconf/2 -I/usr/include/orbit-2.0 > -I/usr/include/glib-2.0 -I/usr/lib/glib-2.0/include -I/usr/include/gtk-2.0 > -I/usr/lib/gtk-2.0/include -I/usr/include/pango-1.0 -I/usr/include/glib-2.0 > -I/usr/lib/glib-2.0/include -I/usr/include/cairo -I/usr/include/freetype2 > -I/usr/include/libpng12 -g -O2 -MT GConfNativePeer.lo -MD -MP -MF > .deps/GConfNativePeer.Tpo -c GConfNativePeer.c -fPIC -DPIC -o > .libs/GConfNativePeer.o > CROSS COMPILE Badness: /usr/include in INCLUDEPATH: /usr/include/gconf/2 > cc1: internal compiler error: in add_path, at c-incpath.c:362 > Please submit a full bug report, > with preprocessed source if appropriate. > See <URL:http://gcc.gnu.org/bugs.html> for instructions. > make[3]: *** [GConfNativePeer.lo] Error 1 > make[3]: Leaving directory > `/home/pabeles/gumstix/java/classpath-0.95/native/jni/gconf-peer' > make[2]: *** [all-recursive] Error 1 > make[2]: Leaving directory > `/home/pabeles/gumstix/java/classpath-0.95/native/jni' > make[1]: *** [all-recursive] Error 1 > make[1]: Leaving directory > `/home/pabeles/gumstix/java/classpath-0.95/native' > > -- > View this message in context: http://www.nabble.com/gumstix-%2B-OE-%2B-java-tp15466829p15490517.html > > > Sent from the Gumstix mailing list archive at Nabble.com. > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2008. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > _______________________________________________ > gumstix-users mailing list > gum...@li... > https://lists.sourceforge.net/lists/listinfo/gumstix-users > |
From: Steve S. <sa...@gm...> - 2008-02-15 07:01:50
|
Peter, I updated the classpath recipe in gumstix-oe. The build for 0.95 succeeds for me now -- I'm not a Java guy, so I'll leave the testing to you :-) Just do an svn update to get it. Steve On Thu, Feb 14, 2008 at 2:17 PM, Peter A <pet...@ya...> wrote: > > Steve, > > FYI I tried to build classpath 0.95 from source code I downloaded from their > website and got the same error that I did with 0.96. So unless I am doing > something wrong, I suspect you will get the same error also once the build > gets fixed. Just in case here is what I am doing and the error message. I > was also going to post a bug report on OE's website, but they have yet to > verify the e-mail address... > > Thanks, > - Peter > > -------- COMMANDS --------------- > > $ cd classpath-0.95/ > $ export PATH=$PATH:~/gumstix/gumstix-oe/tmp/cross/bin > $ ./configure --prefix=/home/pabeles/gumstix/tmp > CC=arm-angstrom-linux-gnueabi-gcc CPP=arm-angstrom-linux-gnueabi-cpp > CXX=arm-angstrom-linux-gnueabi-g++ RANLIB=arm-angstrom-linux-gnueabi-ranlib > --host=arm-linux --disable-gtk-peer --disable-plugin --with-ecj > $ make > > ------- ERROR MESSAGE ------------- > > mkdir .libs > arm-angstrom-linux-gnueabi-gcc -DHAVE_CONFIG_H -I. -I../../../include > -I../../../include -I../../../native/jni/classpath > -I../../../native/jni/native-lib -W -Wall -Wmissing-declarations > -Wwrite-strings -Wmissing-prototypes -Wno-long-long -Werror -DORBIT2=1 > -pthread -I/usr/include/gconf/2 -I/usr/include/orbit-2.0 > -I/usr/include/glib-2.0 -I/usr/lib/glib-2.0/include -I/usr/include/gtk-2.0 > -I/usr/lib/gtk-2.0/include -I/usr/include/pango-1.0 -I/usr/include/glib-2.0 > -I/usr/lib/glib-2.0/include -I/usr/include/cairo -I/usr/include/freetype2 > -I/usr/include/libpng12 -g -O2 -MT GConfNativePeer.lo -MD -MP -MF > .deps/GConfNativePeer.Tpo -c GConfNativePeer.c -fPIC -DPIC -o > .libs/GConfNativePeer.o > CROSS COMPILE Badness: /usr/include in INCLUDEPATH: /usr/include/gconf/2 > cc1: internal compiler error: in add_path, at c-incpath.c:362 > Please submit a full bug report, > with preprocessed source if appropriate. > See <URL:http://gcc.gnu.org/bugs.html> for instructions. > make[3]: *** [GConfNativePeer.lo] Error 1 > make[3]: Leaving directory > `/home/pabeles/gumstix/java/classpath-0.95/native/jni/gconf-peer' > make[2]: *** [all-recursive] Error 1 > make[2]: Leaving directory > `/home/pabeles/gumstix/java/classpath-0.95/native/jni' > make[1]: *** [all-recursive] Error 1 > make[1]: Leaving directory > `/home/pabeles/gumstix/java/classpath-0.95/native' > > -- > View this message in context: http://www.nabble.com/gumstix-%2B-OE-%2B-java-tp15466829p15490517.html > > > Sent from the Gumstix mailing list archive at Nabble.com. > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2008. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > _______________________________________________ > gumstix-users mailing list > gum...@li... > https://lists.sourceforge.net/lists/listinfo/gumstix-users > |
From: Peter A <pet...@ya...> - 2008-02-15 20:45:26
|
Steve, I was able to build classpath, but it was too big. The .ipk files were about 9 megs total and I have 5 megs available. So I tried to build classpath-minimal. That appears to build, but it didn't generate an .ipk file. I did figure out why classpath compiles in OE but not when I build it. If you include the parameter "--disable-gconf-peer" when configuring the package it will compile. - Peter Peter, I updated the classpath recipe in gumstix-oe. The build for 0.95 succeeds for me now -- I'm not a Java guy, so I'll leave the testing to you :-) Just do an svn update to get it. Steve -- View this message in context: http://www.nabble.com/gumstix-%2B-OE-%2B-java-tp15466829p15509742.html Sent from the Gumstix mailing list archive at Nabble.com. |
From: /<evin <kev...@ya...> - 2009-02-20 15:48:31
|
Has anyone figured this issue out? I too am looking for classpath minimal. I updated my preferred provider in gumstix.inc and it looks like it is building classpath-minimal 0.95....but no IPKG to be found? I see a gumstix recipe for classpath-minimal 0.96.1 but I get the following error when doing bitbake classpath-minimal_0.96.1.bb: ERROR: Nothing PROVIDES 'classpath-minimal_0.96.1.bb' (but '[]' DEPENDS on or otherwise requires it) Has anyone else had success bitbaking classpath-minimal? (oh Verdex Pro XL6P and I would like to avoid mmc or manual classpath building/pruning if possible) Thanks, Kevin Steve, I was able to build classpath, but it was too big. The .ipk files were about 9 megs total and I have 5 megs available. So I tried to build classpath-minimal. That appears to build, but it didn't generate an .ipk file. I did figure out why classpath compiles in OE but not when I build it. If you include the parameter "--disable-gconf-peer" when configuring the package it will compile. - Peter Steve Sakoman wrote: > > Peter, > > I updated the classpath recipe in gumstix-oe. The build for 0.95 > succeeds for me now -- I'm not a Java guy, so I'll leave the testing > to you :-) > > Just do an svn update to get it. > > Steve > > -- View this message in context: http://www.nabble.com/gumstix-%2B-OE-%2B-java-tp15466829p22120454.html Sent from the Gumstix mailing list archive at Nabble.com. |
From: Steve S. <sa...@gm...> - 2008-02-15 20:49:52
|
Peter, Do you have an sd, mmc, or cf card on your gumstix? If so, you could install the package on the flash card with ipkg. I'll try to put some instructions together. Steve On Fri, Feb 15, 2008 at 12:45 PM, Peter A <pet...@ya...> wrote: > > Steve, > > I was able to build classpath, but it was too big. The .ipk files were > about 9 megs total and I have 5 megs available. So I tried to build > classpath-minimal. That appears to build, but it didn't generate an .ipk > file. > > I did figure out why classpath compiles in OE but not when I build it. If > you include the parameter "--disable-gconf-peer" when configuring the > package it will compile. > > - Peter > > > > Peter, > > I updated the classpath recipe in gumstix-oe. The build for 0.95 > succeeds for me now -- I'm not a Java guy, so I'll leave the testing > to you :-) > > Just do an svn update to get it. > > Steve > > -- > View this message in context: http://www.nabble.com/gumstix-%2B-OE-%2B-java-tp15466829p15509742.html > > > Sent from the Gumstix mailing list archive at Nabble.com. > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2008. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > _______________________________________________ > gumstix-users mailing list > gum...@li... > https://lists.sourceforge.net/lists/listinfo/gumstix-users > |
From: Peter A <pet...@ya...> - 2008-02-19 20:54:00
|
Steve, I got a java "hello world" running on my gumstix with OE now. Since I now have a microsd card the limited flash space is no longer a problem. 1) I set up my microsd using your instructions at: http://www.gumstix.net/User/view/User-how-to-s/Installing-packages-to-MMC/SD/microSD-memory-cards/110.html 2) I then installed classpath and jamvm by typing (on the gumstix): ipkg -d mmc install classpath ipkg -d mmc install jamvm 3) jamvm did not correctly set up the symbolic links. I'm not sure if that's my fault or not. I accidentally killed the install with control-c while it was installing, then tried to remove it to be safe, then told it to reinstall... 4) type "rm /usr/bin/linux" then type "ln -s /media/card/usr/bin/jamvm /usr/bin/java" 5) Other links are also screwed up, so you need to type the following to run the java program: java -Xbootclasspath:/media/card/usr/share/jamvm/classes.zip:/media/card/usr/share/classpath/glibj.zip -Dgnu.classpath.boot.library.path=/media/card/usr/lib/classpath HelloWorld Which is fairly painless. Manually building and editing classpath and jamvm will result in much smaller installations, but with 1gig on my microsd that's not needed. - Peter Steve Sakoman wrote: > > Peter, > > Do you have an sd, mmc, or cf card on your gumstix? > > If so, you could install the package on the flash card with ipkg. > > I'll try to put some instructions together. > > Steve > > On Fri, Feb 15, 2008 at 12:45 PM, Peter A <pet...@ya...> wrote: >> >> Steve, >> >> I was able to build classpath, but it was too big. The .ipk files were >> about 9 megs total and I have 5 megs available. So I tried to build >> classpath-minimal. That appears to build, but it didn't generate an >> .ipk >> file. >> >> I did figure out why classpath compiles in OE but not when I build it. >> If >> you include the parameter "--disable-gconf-peer" when configuring the >> package it will compile. >> >> - Peter >> >> >> >> Peter, >> >> I updated the classpath recipe in gumstix-oe. The build for 0.95 >> succeeds for me now -- I'm not a Java guy, so I'll leave the testing >> to you :-) >> >> Just do an svn update to get it. >> >> Steve >> >> -- >> View this message in context: >> http://www.nabble.com/gumstix-%2B-OE-%2B-java-tp15466829p15509742.html >> >> >> Sent from the Gumstix mailing list archive at Nabble.com. >> >> >> >> ------------------------------------------------------------------------- >> This SF.net email is sponsored by: Microsoft >> Defy all challenges. Microsoft(R) Visual Studio 2008. >> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ >> _______________________________________________ >> gumstix-users mailing list >> gum...@li... >> https://lists.sourceforge.net/lists/listinfo/gumstix-users >> > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2008. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > _______________________________________________ > gumstix-users mailing list > gum...@li... > https://lists.sourceforge.net/lists/listinfo/gumstix-users > > -- View this message in context: http://www.nabble.com/gumstix-%2B-OE-%2B-java-tp15466829p15566134.html Sent from the Gumstix mailing list archive at Nabble.com. |
From: Steve S. <sa...@gm...> - 2008-02-19 21:02:57
|
> I got a java "hello world" running on my gumstix with OE now. Since I now > have a microsd card the limited flash space is no longer a problem. Cool! > 2) I then installed classpath and jamvm by typing (on the gumstix): > > ipkg -d mmc install classpath > ipkg -d mmc install jamvm > > 3) jamvm did not correctly set up the symbolic links. I'm not sure if > that's my fault or not. It's your fault if you didn't do the final step in the instructions: ipkg-link mount /media/card That's what sets up the links :-) Steve |
From: Peter A <pet...@ya...> - 2008-02-20 14:35:26
|
Yep, your right. I forgot to do "ipkg-link mount /media/card". All the links are now set up correctly. Steve Sakoman wrote: > >> >> 3) jamvm did not correctly set up the symbolic links. I'm not sure if >> that's my fault or not. > > It's your fault if you didn't do the final step in the instructions: > ipkg-link mount /media/card > > That's what sets up the links :-) > > -- View this message in context: http://www.nabble.com/gumstix-%2B-OE-%2B-java-tp15466829p15590291.html Sent from the Gumstix mailing list archive at Nabble.com. |
From: Steve S. <sa...@gm...> - 2008-02-20 15:30:41
|
Excellent! Are you compiling on your desktop and then copying the output over? Guess the next step is to get a gumstix native java compiler in the repository :-) We need to fill up that gigabyte of flash! Steve On Feb 20, 2008 6:35 AM, Peter A <pet...@ya...> wrote: > > Yep, your right. I forgot to do "ipkg-link mount /media/card". All the links > are now set up correctly. > > > > Steve Sakoman wrote: > > > >> > >> 3) jamvm did not correctly set up the symbolic links. I'm not sure if > >> that's my fault or not. > > > > It's your fault if you didn't do the final step in the instructions: > > ipkg-link mount /media/card > > > > That's what sets up the links :-) > > > > > > > -- > View this message in context: http://www.nabble.com/gumstix-%2B-OE-%2B-java-tp15466829p15590291.html > > Sent from the Gumstix mailing list archive at Nabble.com. > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2008. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > _______________________________________________ > gumstix-users mailing list > gum...@li... > https://lists.sourceforge.net/lists/listinfo/gumstix-users > |
From: Peter A <pet...@ya...> - 2008-02-21 13:56:54
|
I'm compiling it on my desktop and copying it over. Now that I'm trying to run a real java program, I'm actually having a bunch of problems with classpath and jamvm being non-compliant. Most of the problems are related to where it searches for files. Not being able to print a float is also annoying... - Peter Steve Sakoman wrote: > > Excellent! > > Are you compiling on your desktop and then copying the output over? > > Guess the next step is to get a gumstix native java compiler in the > repository :-) > > We need to fill up that gigabyte of flash! > > Steve > > On Feb 20, 2008 6:35 AM, Peter A <pet...@ya...> wrote: >> >> Yep, your right. I forgot to do "ipkg-link mount /media/card". All the >> links >> are now set up correctly. >> >> >> >> Steve Sakoman wrote: >> > >> >> >> >> 3) jamvm did not correctly set up the symbolic links. I'm not sure if >> >> that's my fault or not. >> > >> > It's your fault if you didn't do the final step in the instructions: >> > ipkg-link mount /media/card >> > >> > That's what sets up the links :-) >> > >> > >> >> >> -- >> View this message in context: >> http://www.nabble.com/gumstix-%2B-OE-%2B-java-tp15466829p15590291.html >> >> Sent from the Gumstix mailing list archive at Nabble.com. >> >> >> ------------------------------------------------------------------------- >> This SF.net email is sponsored by: Microsoft >> Defy all challenges. Microsoft(R) Visual Studio 2008. >> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ >> _______________________________________________ >> gumstix-users mailing list >> gum...@li... >> https://lists.sourceforge.net/lists/listinfo/gumstix-users >> > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2008. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > _______________________________________________ > gumstix-users mailing list > gum...@li... > https://lists.sourceforge.net/lists/listinfo/gumstix-users > > -- View this message in context: http://www.nabble.com/gumstix-%2B-OE-%2B-java-tp15466829p15611837.html Sent from the Gumstix mailing list archive at Nabble.com. |
From: Steve S. <sa...@gm...> - 2008-02-21 15:47:31
|
Peter, I think there are a couple of other jvm packages available. Not sure if that would help though, since your issues seem to be more classpath related. Steve On Thu, Feb 21, 2008 at 5:56 AM, Peter A <pet...@ya...> wrote: > > I'm compiling it on my desktop and copying it over. > > Now that I'm trying to run a real java program, I'm actually having a bunch > of problems with classpath and jamvm being non-compliant. Most of the > problems are related to where it searches for files. Not being able to > print a float is also annoying... > > - Peter > > > > > Steve Sakoman wrote: > > > > Excellent! > > > > Are you compiling on your desktop and then copying the output over? > > > > Guess the next step is to get a gumstix native java compiler in the > > repository :-) > > > > We need to fill up that gigabyte of flash! > > > > Steve > > > > On Feb 20, 2008 6:35 AM, Peter A <pet...@ya...> wrote: > >> > >> Yep, your right. I forgot to do "ipkg-link mount /media/card". All the > >> links > >> are now set up correctly. > >> > >> > >> > >> Steve Sakoman wrote: > >> > > >> >> > >> >> 3) jamvm did not correctly set up the symbolic links. I'm not sure if > >> >> that's my fault or not. > >> > > >> > It's your fault if you didn't do the final step in the instructions: > >> > ipkg-link mount /media/card > >> > > >> > That's what sets up the links :-) > >> > > >> > > >> > >> > >> -- > >> View this message in context: > >> http://www.nabble.com/gumstix-%2B-OE-%2B-java-tp15466829p15590291.html > >> > >> Sent from the Gumstix mailing list archive at Nabble.com. > >> > >> > >> ------------------------------------------------------------------------- > >> This SF.net email is sponsored by: Microsoft > >> Defy all challenges. Microsoft(R) Visual Studio 2008. > >> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > >> _______________________________________________ > >> gumstix-users mailing list > >> gum...@li... > >> https://lists.sourceforge.net/lists/listinfo/gumstix-users > >> > > > > ------------------------------------------------------------------------- > > This SF.net email is sponsored by: Microsoft > > Defy all challenges. Microsoft(R) Visual Studio 2008. > > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > > _______________________________________________ > > gumstix-users mailing list > > gum...@li... > > https://lists.sourceforge.net/lists/listinfo/gumstix-users > > > > > > -- > View this message in context: http://www.nabble.com/gumstix-%2B-OE-%2B-java-tp15466829p15611837.html > > > Sent from the Gumstix mailing list archive at Nabble.com. > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2008. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > _______________________________________________ > gumstix-users mailing list > gum...@li... > https://lists.sourceforge.net/lists/listinfo/gumstix-users > |
From: Peter A <pet...@ya...> - 2008-02-21 21:15:02
|
Steve, The biggest problem I'm now having is that I can't bind a TCP socket to listen for incoming connections since it crashes when I do. When I compile GNU classpath for my desktop and run the code with jamvm it works. When compiled for the arm and run on the gumstix it crashes. After going through classpath's code I can see that it is crashing on a call to the java native interface (JNI). So I'm looking at the alternatives again and posting a bug report... - Peter Steve Sakoman wrote: > > Peter, > > I think there are a couple of other jvm packages available. Not sure > if that would help though, since your issues seem to be more classpath > related. > > Steve > > -- View this message in context: http://www.nabble.com/gumstix-%2B-OE-%2B-java-tp15466829p15618963.html Sent from the Gumstix mailing list archive at Nabble.com. |
From: Stephen B. <wid...@st...> - 2008-02-21 22:06:55
|
On Thursday 21 February 2008 21:15:07 Peter A wrote: > Steve, > > The biggest problem I'm now having is that I can't bind a TCP socket to > listen for incoming connections since it crashes when I do. When I compile > GNU classpath for my desktop and run the code with jamvm it works. When > compiled for the arm and run on the gumstix it crashes. After going > through classpath's code I can see that it is crashing on a call to the > java native interface (JNI). > > So I'm looking at the alternatives again and posting a bug report... I'm also trying to get some Java code working on an emulated connex (via qemu, I can't buy the hardware until I know it's going to work). I'm looking at two different problems. One is getting the code working with JamVM and gnu classpath, and the other is getting classpath on the gumstix. Now C/C++ is usually my poison of choice, and I wouldn't bother getting this Java on the system but I don't want to rewrite it myself :) On my Gentoo x86 host, I have got the code to almost-entirely work using JamVM 1.5.0. 1.4.2 didn't support java.net properly from what I can tell, and 1.4.5 got a bit further but still stopped. Interestingly 1.5.0 promises order-of-magnitude speedups as well. That's with classpath 0.93 as there's nothing higher on Gentoo in portage at the moment. I've tried to get a higher classpath in Gentoo, but it fails to compile with ecj and runs out of memory. The GNU website says ecj should be run with an argument to allow more memory to be used in its virtual machine when compiling but I haven't worked out how to do that yet. I can't compile classpath 0.95 in OE either, as I get an error from javac saying that "-1.5" isn't a valid flag. This seems to be due to the fact that the bitbake script uses "--with-ecj=javac" to tell classpath it can use my javac like ecj. I've made my own recipe to change this but I hit the OutOfMemory exception problem again. Using --with-javac seems to cause a different error that I don't understand, which I don't have to hand here. So just wondering what VM people are using to build classpath at the moment. Perhaps it's a Gentoo thing as Gentoo doesn't wrap ecj. Stephen. |
From: Steve S. <sa...@gm...> - 2008-02-21 22:21:51
|
Hmm . . . the gumstix autobuilder is building classpath 0.95 and jamvm 1.45 twice a day. The binary packages for both are available in the repository, too. Are you trying to build with the most recent gumstix-oe? Steve On Thu, Feb 21, 2008 at 2:06 PM, Stephen Bridges <wid...@st...> wrote: > On Thursday 21 February 2008 21:15:07 Peter A wrote: > > Steve, > > > > The biggest problem I'm now having is that I can't bind a TCP socket to > > listen for incoming connections since it crashes when I do. When I compile > > GNU classpath for my desktop and run the code with jamvm it works. When > > compiled for the arm and run on the gumstix it crashes. After going > > through classpath's code I can see that it is crashing on a call to the > > java native interface (JNI). > > > > So I'm looking at the alternatives again and posting a bug report... > > I'm also trying to get some Java code working on an emulated connex (via qemu, > I can't buy the hardware until I know it's going to work). I'm looking at > two different problems. One is getting the code working with JamVM and gnu > classpath, and the other is getting classpath on the gumstix. > > Now C/C++ is usually my poison of choice, and I wouldn't bother getting this > Java on the system but I don't want to rewrite it myself :) On my Gentoo x86 > host, I have got the code to almost-entirely work using JamVM 1.5.0. 1.4.2 > didn't support java.net properly from what I can tell, and 1.4.5 got a bit > further but still stopped. Interestingly 1.5.0 promises order-of-magnitude > speedups as well. That's with classpath 0.93 as there's nothing higher on > Gentoo in portage at the moment. > > I've tried to get a higher classpath in Gentoo, but it fails to compile with > ecj and runs out of memory. The GNU website says ecj should be run with an > argument to allow more memory to be used in its virtual machine when > compiling but I haven't worked out how to do that yet. > > I can't compile classpath 0.95 in OE either, as I get an error from javac > saying that "-1.5" isn't a valid flag. This seems to be due to the fact that > the bitbake script uses "--with-ecj=javac" to tell classpath it can use my > javac like ecj. I've made my own recipe to change this but I hit the > OutOfMemory exception problem again. Using --with-javac seems to cause a > different error that I don't understand, which I don't have to hand here. > > So just wondering what VM people are using to build classpath at the moment. > Perhaps it's a Gentoo thing as Gentoo doesn't wrap ecj. > > Stephen. > > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2008. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > _______________________________________________ > gumstix-users mailing list > gum...@li... > https://lists.sourceforge.net/lists/listinfo/gumstix-users > |
From: Stephen B. <wid...@st...> - 2008-02-21 23:26:19
|
On Thursday 21 February 2008 22:21:56 Steve Sakoman wrote: > Hmm . . . the gumstix autobuilder is building classpath 0.95 and jamvm > 1.45 twice a day. The binary packages for both are available in the > repository, too. > > Are you trying to build with the most recent gumstix-oe? Yeah, I checked out yesterday. I've had to modify classpath to build on my desktop natively anyway, so that's why it doesn't compile. But I assume that your build machine must have javac symlinked to ecj, and mine doesn't. Javac here is linked to the blackdown JDK which doesn't like the "-1.5" flag passed to it by classpath's make. Changing the bitbake recipe from --with-ecj=javac to --with-javac fixed that. Now I'm working on patch to lib/Makefile.in to add -J-Xmx512M to JCOMPILER for javac, but it kind of vanishes into bitbake, which informs me it has applied, but it's not in the temp directory. Stephen. |
From: Steve S. <sa...@gm...> - 2008-02-21 23:41:23
|
> Yeah, I checked out yesterday. I've had to modify classpath to build on my > desktop natively anyway, so that's why it doesn't compile. But I assume that > your build machine must have javac symlinked to ecj, and mine doesn't. Ah, that is correct. I do have ecj installed on my build machine. > Javac > here is linked to the blackdown JDK which doesn't like the "-1.5" flag passed > to it by classpath's make. Changing the bitbake recipe from --with-ecj=javac > to --with-javac fixed that. Makes sense. > Now I'm working on patch to lib/Makefile.in to add -J-Xmx512M to JCOMPILER for > javac, but it kind of vanishes into bitbake, which informs me it has applied, > but it's not in the temp directory. I assume you are doing a "bitbake -c rebuild" in order to force a rebuild? Must be if you are getting a message that the patch is being applied. So the source file that's supposed to be patched remains unchanged? Strange! Steve |
From: Stephen B. <wid...@st...> - 2008-02-21 23:55:14
|
On Thursday 21 February 2008 23:41:08 Steve Sakoman wrote: > Ah, that is correct. I do have ecj installed on my build machine. Gentoo doesn't symlink it for some reason - it must have a plan, but I couldn't get hold of a dev on IRC. > I assume you are doing a "bitbake -c rebuild" in order to force a > rebuild? Must be if you are getting a message that the patch is being > applied. No, I didn't find that command! I've been doing -f -c unpack; patch; configure; install; instead! That'll save some time. > So the source file that's supposed to be patched remains unchanged? > Strange! I was patching Makefile.in, just before it gets deleted. Patching Makefile.am is doing a much better job, and now the package has built. Now I can go to bed, and test this at work tomorrow! Stephen. --- lib/Makefile.am 2008-02-21 23:48:19.000000000 +0000 +++ lib/Makefile.am 2008-02-21 23:48:36.000000000 +0000 @@ -17,7 +17,7 @@ JCOMPILER = $(ECJ) -1.5 -warn:-deprecation,serial,typeHiding,unchecked,unused,varargsCast -proceedOnError -bootclasspath '' -classpath $(compile_classpath) -d . @classes else if FOUND_JAVAC -JCOMPILER = $(JAVAC) -bootclasspath '' -classpath $(compile_classpath) -d . @classes +JCOMPILER = $(JAVAC) -J-Xmx512M -bootclasspath '' -classpath $(compile_classpath) -d . @classes #else #if FOUND_GCJ ## This should never be used when gcj is the compiler. |
From: Peter A <pet...@ya...> - 2008-02-22 01:14:38
|
At first I tried compiling classpath with javac, but gave up since no matter how much memory I gave it never compiled. Now I just compile classpath with ecj. I'm surprised you got it to work with those flags. I tried to get javac to compile it with those flags earlier with no success. I have also been using classpath 0.96 and jamvm 1.5 on my gumstix, but I built them outside of OE. It also sounds like classpath 0.97 will be coming out in a week or two. Stephen Bridges-3 wrote: > > On Thursday 21 February 2008 23:41:08 Steve Sakoman wrote: > >> Ah, that is correct. I do have ecj installed on my build machine. > > Gentoo doesn't symlink it for some reason - it must have a plan, but I > couldn't get hold of a dev on IRC. > > >> I assume you are doing a "bitbake -c rebuild" in order to force a >> rebuild? Must be if you are getting a message that the patch is being >> applied. > > No, I didn't find that command! I've been doing -f -c unpack; patch; > configure; install; instead! That'll save some time. > >> So the source file that's supposed to be patched remains unchanged? >> Strange! > > I was patching Makefile.in, just before it gets deleted. Patching > Makefile.am > is doing a much better job, and now the package has built. Now I can go > to > bed, and test this at work tomorrow! > > Stephen. > > --- lib/Makefile.am 2008-02-21 23:48:19.000000000 +0000 > +++ lib/Makefile.am 2008-02-21 23:48:36.000000000 +0000 > @@ -17,7 +17,7 @@ > JCOMPILER = > $(ECJ) -1.5 > -warn:-deprecation,serial,typeHiding,unchecked,unused,varargsCast > -proceedOnError -bootclasspath '' -classpath > $(compile_classpath) -d . @classes > else > if FOUND_JAVAC > -JCOMPILER = $(JAVAC) -bootclasspath '' -classpath $(compile_classpath) -d > . > @classes > +JCOMPILER = $(JAVAC) -J-Xmx512M -bootclasspath '' -classpath > $(compile_classpath) -d . @classes > #else > #if FOUND_GCJ > ## This should never be used when gcj is the compiler. > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2008. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > _______________________________________________ > gumstix-users mailing list > gum...@li... > https://lists.sourceforge.net/lists/listinfo/gumstix-users > > -- View this message in context: http://www.nabble.com/gumstix-%2B-OE-%2B-java-tp15466829p15625415.html Sent from the Gumstix mailing list archive at Nabble.com. |
From: Steve S. <sa...@gm...> - 2008-02-22 01:32:08
|
Peter, > I have also been using classpath 0.96 and jamvm 1.5 on my gumstix, but I > built them outside of OE Do those work better? Should I bump the versions in gumstix-oe? Steve On Thu, Feb 21, 2008 at 5:14 PM, Peter A <pet...@ya...> wrote: > > At first I tried compiling classpath with javac, but gave up since no matter > how much memory I gave it never compiled. Now I just compile classpath with > ecj. I'm surprised you got it to work with those flags. I tried to get > javac to compile it with those flags earlier with no success. > > I have also been using classpath 0.96 and jamvm 1.5 on my gumstix, but I > built them outside of OE. It also sounds like classpath 0.97 will be coming > out in a week or two. > > > > > Stephen Bridges-3 wrote: > > > > On Thursday 21 February 2008 23:41:08 Steve Sakoman wrote: > > > >> Ah, that is correct. I do have ecj installed on my build machine. > > > > Gentoo doesn't symlink it for some reason - it must have a plan, but I > > couldn't get hold of a dev on IRC. > > > > > >> I assume you are doing a "bitbake -c rebuild" in order to force a > >> rebuild? Must be if you are getting a message that the patch is being > >> applied. > > > > No, I didn't find that command! I've been doing -f -c unpack; patch; > > configure; install; instead! That'll save some time. > > > >> So the source file that's supposed to be patched remains unchanged? > >> Strange! > > > > I was patching Makefile.in, just before it gets deleted. Patching > > Makefile.am > > is doing a much better job, and now the package has built. Now I can go > > to > > bed, and test this at work tomorrow! > > > > Stephen. > > > > --- lib/Makefile.am 2008-02-21 23:48:19.000000000 +0000 > > +++ lib/Makefile.am 2008-02-21 23:48:36.000000000 +0000 > > @@ -17,7 +17,7 @@ > > JCOMPILER = > > $(ECJ) -1.5 > > -warn:-deprecation,serial,typeHiding,unchecked,unused,varargsCast > > -proceedOnError -bootclasspath '' -classpath > > $(compile_classpath) -d . @classes > > else > > if FOUND_JAVAC > > -JCOMPILER = $(JAVAC) -bootclasspath '' -classpath $(compile_classpath) -d > > . > > @classes > > +JCOMPILER = $(JAVAC) -J-Xmx512M -bootclasspath '' -classpath > > $(compile_classpath) -d . @classes > > #else > > #if FOUND_GCJ > > ## This should never be used when gcj is the compiler. > > > > ------------------------------------------------------------------------- > > This SF.net email is sponsored by: Microsoft > > Defy all challenges. Microsoft(R) Visual Studio 2008. > > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > > _______________________________________________ > > gumstix-users mailing list > > gum...@li... > > https://lists.sourceforge.net/lists/listinfo/gumstix-users > > > > > > -- > View this message in context: http://www.nabble.com/gumstix-%2B-OE-%2B-java-tp15466829p15625415.html > > Sent from the Gumstix mailing list archive at Nabble.com. > > > ------------------------------------------------------------------------- > > > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2008. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > _______________________________________________ > gumstix-users mailing list > gum...@li... > https://lists.sourceforge.net/lists/listinfo/gumstix-users > |
From: Peter A <pet...@ya...> - 2008-02-22 15:39:47
|
I think that jamvm should be bumped to 1.5. This morning I realized I was not running 1.5 on the gumstix, but the version installed by ipkg, which is 1.45. When I started to use version 1.5 ServerSocket.bind() started to work. I have pasted the ./configure flags that I am using below: ./configure --prefix=<YOUR OPTIONAL INSTALL DIR> --host=arm-angstrom-linux-gnueabi --build=i686-linux --disable-gtk-peer --disable-gconf-peer --disable-plug I haven't noticed a big difference between classpath 0.95 and 0.96. They also just released 0.97 today. I'll give that a try and let you know if it's worth it. - Peter Steve Sakoman wrote: > > Peter, > >> I have also been using classpath 0.96 and jamvm 1.5 on my gumstix, but I >> built them outside of OE > > Do those work better? Should I bump the versions in gumstix-oe? > > Steve > > -- View this message in context: http://www.nabble.com/gumstix-%2B-OE-%2B-java-tp15466829p15634439.html Sent from the Gumstix mailing list archive at Nabble.com. |