https://bugs.kde.org/show_bug.cgi?id=496552

            Bug ID: 496552
           Summary: intermittent false positives with JNI code?
    Classification: Developer tools
           Product: valgrind
           Version: 3.23.0
          Platform: Compiled Sources
                OS: Linux
            Status: REPORTED
          Severity: normal
          Priority: NOR
         Component: memcheck
          Assignee: jsew...@acm.org
          Reporter: wallstp...@gmail.com
  Target Milestone: ---

We regularly regression-test our code with valgrind, including some JNI code,
and those tests always pass at this point.  Except yesterday, when we got a
bunch of "Illegal read/write" errors.

I've satisfied myself as best I can that these are false positives by examining
the source, but at the same time I don't recall ever seeing a true false
positive from valgrind before, so I'm suspicious of that explanation.

I know this is a long shot, but I've attached the report file and I'd be very
grateful if someone has the time to take a look and see if anything odd jumps
out.  As you can see, we're running memcheck on the JVM, which generates a LOT
of output (we use the scripts from https://github.com/nyfix/memleaktest to pare
that down to something manageable).  

The bits that are relevant to this discussion are at lines #9655-10497.  

Any suggestions much appreciated -- thanks in advance!

-- 
You are receiving this mail because:
You are watching all bug changes.

Reply via email to