Re: [cp-testresults] classpath daily snapshot 20061009 FAILED

[Michael Koch]

On Mon, Oct 09, 2006 at 08:16:38AM +0200, Mark Wielaard wrote:
> Hi Michael,
> 
> On Mon, 2006-10-09 at 05:53 +0200, Michael Koch wrote:
> >  gcc -DHAVE_CONFIG_H -I. -I../../../../native/jni/java-net 
> > -I../../../include -I../../../../include -I../../../../native/jni/classpath 
> > -I../../../../native/jni/native-lib -W -Wall -Wmissing-declarations 
> > -Wwrite-strings -Wmissing-prototypes -Wno-long-long -Wstrict-prototypes 
> > -pedantic -g -O2 -MT java_net_VMInetAddress.lo -MD -MP -MF 
> > .deps/java_net_VMInetAddress.Tpo -c 
> > ../../../../native/jni/java-net/java_net_VMInetAddress.c  -fPIC -DPIC -o 
> > .libs/java_net_VMInetAddress.o
> > if /bin/sh ../../../libtool --mode=compile --tag=CC gcc -DHAVE_CONFIG_H -I. 
> > -I../../../../native/jni/java-net -I../../../include  -I../../../../include 
> > -I../../../../native/jni/classpath -I../../../../native/jni/native-lib  -W 
> > -Wall -Wmissing-declarations -Wwrite-strings -Wmissing-prototypes 
> > -Wno-long-long -Wstrict-prototypes -pedantic  -g -O2 -MT 
> > java_net_VMNetworkInterface.lo -MD -MP -MF 
> > ".deps/java_net_VMNetworkInterface.Tpo" -c -o 
> > java_net_VMNetworkInterface.lo 
> > ../../../../native/jni/java-net/java_net_VMNetworkInterface.c; \
> > then mv -f ".deps/java_net_VMNetworkInterface.Tpo" 
> > ".deps/java_net_VMNetworkInterface.Plo"; else rm -f 
> > ".deps/java_net_VMNetworkInterface.Tpo"; exit 1; fi
> >  gcc -DHAVE_CONFIG_H -I. -I../../../../native/jni/java-net 
> > -I../../../include -I../../../../include -I../../../../native/jni/classpath 
> > -I../../../../native/jni/native-lib -W -Wall -Wmissing-declarations 
> > -Wwrite-strings -Wmissing-prototypes -Wno-long-long -Wstrict-prototypes 
> > -pedantic -g -O2 -MT java_net_VMNetworkInterface.lo -MD -MP -MF 
> > .deps/java_net_VMNetworkInterface.Tpo -c 
> > ../../../../native/jni/java-net/java_net_VMNetworkInterface.c  -fPIC -DPIC 
> > -o .libs/java_net_VMNetworkInterface.o
> > ../../../../native/jni/java-net/java_net_VMNetworkInterface.c:44:21: 
> > ifaddrs.h: No such file or directory
> > ../../../../native/jni/java-net/java_net_VMNetworkInterface.c: In function 
> > `Java_java_net_VMNetworkInterface_getVMInterfaces':
> > ../../../../native/jni/java-net/java_net_VMNetworkInterface.c:112: warning: 
> > unused parameter `clazz'
> > /usr/include/bits/string2.h: At top level:
> > ../../../../native/jni/java-net/java_net_VMNetworkInterface.c:99: warning: 
> > `free_netif_list' defined but not used
> > make[3]: *** [java_net_VMNetworkInterface.lo] Error 1
> > make[3]: Leaving directory 
> > `/home/mkoch/src/classpath/build/native/jni/java-net'
> > make[2]: *** [all-recursive] Error 1
> > make[2]: Leaving directory `/home/mkoch/src/classpath/build/native/jni'
> > make[1]: *** [all-recursive] Error 1
> > make[1]: Leaving directory `/home/mkoch/src/classpath/build/native'
> > make: *** [all-recursive] Error 1
> 
> According to this error message we do detect that your system doesn't
> provide getifaddrs, but don't condition the includes correctly. What
> kind of system is this? 

That is Patrik Reali's Machine.

address@hidden:~> cat /etc/SuSE-release
SuSE Linux 8.1 (i386)
VERSION = 8.1

It uses glibc 2.2.5

address@hidden:/usr/include> rpm -qf /lib/libc.so.6
glibc-2.2.5-163

The header ifaddrs.h doesn't exist in this libc version. And the method
getifaddrs is not defined in any other header. This method seems to be
glibc 2.3 or newer only.

The question is now: Do we want to support glibc (and probably other
libc's for embedded targets) that lack this method? Perhaps we can just
throw some exception on these systems.


Cheers,
Michael
-- 
http://www.worldforge.org/