[Bug 520] Program compiled by ecj causes icedtea to segfault

[bugzilla-daemon at icedtea.classpath.org]

http://icedtea.classpath.org/bugzilla/show_bug.cgi?id=520





------- Comment #10 from jon.vanalten at redhat.com  2010-07-15 19:52 -------
(In reply to comment #8)
> I referred to running the same class file crashing with a segfault on some runs
> and working perfectly on others. I haven't reproduced this lately either but it
> happened.

Odd. I have now reproduced the behavior where it sometimes faults and sometimes
not.  But, when running in Eclipse, it runs fine (ie does not seg fault) when
in debug mode *if* there is at least one breakpoint set.  See also below.

I added a couple of (ugh) debug println statements and found the spot where,
consistently (for me) it faults.  So contrary to my prior thought in comment #3
it does manage to run part of the program.  It is at the beginning of the while
loop for backtracking.  The first checkpoint always prints, the line in the
loop never prints.  Strangely, a breakpoint set even after this point will
allow the program to run fine.  Next will be to look for oddities in the
bytecode, or at least differences from that produced by javac.


-- 
Configure bugmail: http://icedtea.classpath.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.