Re: eclipse crashes on startup
Thomas J. Zeeman <tjzeeman@cs.vu.nl> writes:
> On Thu, 15 Jan 2004, Eckhart [iso-8859-1] Guthöhrlein wrote:
>
>> Current Java thread:
>> at java.util.zip.Inflater.end(Native Method)
> Same here but on a selfcompiled version of eclipse (using debian
> package and apt-src) on testing.
Can you try an upstream build?
I just tried eclipse-SDK-2.1.2-linux-gtk.zip with 1.4.1-01 and
1.4.2-rc1 (both gcc-2.95 and gcc-3.2 variants) and everything worked
fine.
> Some bugs in the BTS suggest that the Blackdown JVM is causing this
> and that either the newer BD version (no debian package available
> from BD themselves; check the archives for some links to some
> created by others) or a version from Sun remedies this.
> Unfortunately that didn't help for me. I have used the latest BD
> version (turned into a .deb with mpkg-j2se) and a recent Sun version
> (also turned into a .deb). Both still crash eclipse quite fast and
> somewhat unpredictably. The latest BD version even crashes when I
> try to recompile the eclipse-packages (it crashes with a similar
> error message as you posted) I'll try to redo that with a completely
> up-to-date Sun version this weekend.
>
> As the error suggests
You mean this?
,----
| NOTE: We are unable to locate the function name symbol for the error
| just occurred. Please refer to release documentation for possible
| reason and solutions.
`----
That just means java couldn't find out the *name* of the function where
the SIGSEGV happened. That's why it says 'Function=(null)+0x400EF0C4'
instead of something like 'Function=foobar+<offset>'.
> that some symbols are no longer exported from glibc
You would get some ld error in that case.
> you could, in the mean time, try to go back to an older version of
> glibc that still does that. Then again, I do not remember seeing an
> update for glibc since November or so...
Juergen
--
Juergen Kreileder, Blackdown Java-Linux Team
http://www.blackdown.org/java-linux/java2-status/
Reply to: