[Octave-bug-tracker] [bug #53733] Panic/segfault when calling Java method after forge:im[read|write]

[Mike Miller]

Follow-up Comment #62, bug #53733 (project octave):

Yes it does. I get the following.


(gdb) r
Starting program: /opt/gnu/octave/bin/octave-cli-4.2.2 --norc --silent --eval
testscr
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".

Program received signal SIGSEGV, Segmentation fault.
[New Thread 0x7fffd0e80700 (LWP 20364)]
[New Thread 0x7fffd0d7f700 (LWP 20365)]
[New Thread 0x7fffd0c7e700 (LWP 20366)]
[New Thread 0x7fffd0b7d700 (LWP 20367)]
[New Thread 0x7fffd0a7c700 (LWP 20368)]
[New Thread 0x7fffd097b700 (LWP 20369)]
[New Thread 0x7fffd087a700 (LWP 20370)]
[New Thread 0x7fffd0779700 (LWP 20371)]
[New Thread 0x7fffc9fe7700 (LWP 20372)]
[New Thread 0x7fffc9ee6700 (LWP 20373)]
[New Thread 0x7fffc9de5700 (LWP 20374)]
[New Thread 0x7fffc9a65700 (LWP 20375)]
[New Thread 0x7fffc9964700 (LWP 20376)]
[New Thread 0x7fffc9863700 (LWP 20377)]
[New Thread 0x7fffc9762700 (LWP 20378)]
[New Thread 0x7fffc9661700 (LWP 20379)]
[New Thread 0x7fffc9560700 (LWP 20380)]
[New Thread 0x7fffc945f700 (LWP 20381)]
N =  10
N =  100
N =  1000
N =  10000
N =  100000
N =  1000000
N =  10000000
N =  100000000
N =    1.0000e+09

Thread 1 "octave-cli-4.2." received signal SIGSEGV, Segmentation fault.


followed by an endless supply of SIGSEGV lines.

When N gets to be 10^8, it takes maybe 10 seconds of 100% busy before it
prints 1e+09. At 10^9 it is still 100% busy, and after a few seconds starts
printing endless screenfuls of SIGSEGV.

>From what I have seen, the main difference between Java 8 and Java 9 is that
Java 8 always produces a segfault when the argument is of length 10^9, whether
a GraphicsMagick function has been called yet or not. That's the same behavior
I see with Java 9 when a GraphicsMagick function has *not* been called. With
Java 9, the size to produce a segfault is reduced to 10^5 after a
GraphicsMagick function has been called. That's basically what I described in
comment #60.

The "endless supply of SIGSEGV lines" symptom only occurs when running with
gdb when imformats or some other GraphicsMagick function has been called. If
no image functions are called, the segmentation fault always occurs at N=10^9,
and it is only a single SIGSEGV message along with Octave's "stopping myself"
diagnostic message.

    _______________________________________________________

Reply to this item at:

  <http://savannah.gnu.org/bugs/?53733>

_______________________________________________
  Message sent via Savannah
  https://savannah.gnu.org/