Comment # 5
on bug 7028
from Evan Huus
(In reply to comment #4)
> (In reply to comment #3)
> > The capture Jakub attached works for me on latest trunk. Scott, is your case
> > fixed as well?
>
> You can always make this XML file XXXX times bigger....
> I think we should have limit for reasembly and/or tvb length.
I am unclear if the crashes reported in this bug (and bug #7499) are simple
out-of-memory crashes or if there was actually an error in the code?
The sample captures on both bugs reassemble correctly for me, and while they
take a large amount of memory neither is particularly slow. Specifically, I am
*not* seeing the "more than 1000000 items in the tree" issue which would
definitely be a bug.
If there was an actual bug it's fixed now; if this is simply Wireshark trying
to allocate more memory than is available then it's basically the same issue
that's been around forever, just a new way of triggering high memory usage.
Aborting reassembly for extra large packets may be a good work-around, though I
suspect this issue will disappear entirely if we end up doing virtual
reassembly (as discussed on -dev in thread starting
https://www.wireshark.org/lists/wireshark-dev/201304/msg00159.html).
Thoughts?
You are receiving this mail because:
- You are watching all bug changes.