and many more 6.9.* reports from http://statistics.netbeans.org/exceptions/detail.do?id=153648
shows an OOM from the method above.
Petr Pisl has already addressed the issue by throwing AssertionError when a cycle in the method happens but the OOM may still happen in the production build with assertions disabled
Fixed on my disk, when the repository will be open, I will commit to the web-main.
I don't know how the QA could verify the fix, because the cases where it can happen should be already fixed. This is a prevention before OOM if there will be a case (in the code) that is not covered yet.
(In reply to comment #3)
> I don't know how the QA could verify the fix, because the cases where it can
> happen should be already fixed. This is a prevention before OOM if there will
> be a case (in the code) that is not covered yet.
Ok integrate this into 7.0 and we'll do sanity test to try to catch regressions.
(Lada, please do sanity test once this fix is integrated into 7.0 RC1, thanks in advance)
Sure, I'll test it once it will be in 7.0 RC1
Integrated into 'main-golden', will be available in build *201103110400* on http://bits.netbeans.org/dev/nightly/ (upload may still be in progress)
User: Petr Pisl <email@example.com>
Log: #196493 - OOM in org.netbeans.modules.php.editor.indent.FormatVisitor.addFormatToken
Problem didn't appear in sanity test, verified
Product Version: NetBeans IDE Dev (Build 201103130400)
Java: 1.6.0_24; Java HotSpot(TM) Client VM 19.1-b02
System: Windows XP version 5.1 running on x86; Cp1250; cs_CZ (nb)
The patch transplanted into releases repository, release70 branch