On Fri, Jan 21, 2011 at 5:57 AM, Fernando Boaglio <boag...@gmail.com> wrote: > Hi, > > There is a recent update which broke my Java IDE (Eclipse - not from > portage) . > > I've seen this strange behavior in 2 different machines, both ADM64, but I'm > not sure which package should I inform to this error: > > > *** glibc detected *** /opt/sun-jdk-1.6.0.23/bin/java: free(): invalid > pointer: 0x00000000456f30d0 *** > ======= Backtrace: ========= > /lib/libc.so.6(+0x783c6)[0x7f7c039593c6] > /opt/sun-jdk-1.6.0.23/jre/lib/amd64/server/libjvm.so(+0x61b879)[0x7f7c0341a879] > /opt/sun-jdk-1.6.0.23/jre/lib/amd64/server/libjvm.so(+0x43d44f)[0x7f7c0323c44f] > /home/fb/eclipseWTP3.2.3/configuration/org.eclipse.osgi/bundles/151/1/.cp/libswt-pi-gtk-3657.so(Java_org_eclipse_swt_internal_gtk_OS__1g_1data_1input_1stream_1read_1line+0xe7)[0x7f7bf5267d04] > [0x7f7bfebeaca8]
> Is it glibc ? glibc is reporting to you about the error, FYI you can control its behavior (ignore, warn, abort) with the MALLOC_CHECK_ variable. See "man malloc" and scroll to the last paragraph. Maybe running as root operated under a different malloc mode? Looks like maybe your file /home/fb/eclipseWTP3.2.3/configuration/org.eclipse.osgi/bundles/151/1/.cp/libswt-pi-gtk-3657.so is the common thread... I think the only way to really know exactly where it's crashing is to build everything involved with debug symbols and debug the coredump or use valgrind or something like that. If Eclipse is a binary install then I guess report it to whoever makes it (I don't know anything about Java or Eclipse).
