[jdk17u-dev] RFR: 8314225: SIGSEGV in JavaThread::is_lock_owned [v2]

[Richard Reingruber]

On Fri, 10 Jan 2025 11:09:25 GMT, Goetz Lindenmaier <goetz at openjdk.org> wrote:

>> I backport this for parity with 17.0.15-oracle.
>> 
>> I had to resolve the larger part of this change. 
>> None of my edits change the code essentially, though.
>> 
>> src/hotspot/share/jfr/leakprofiler/checkpoint/rootResolver.cpp
>> Resolved because of NULL/nullptr difference
>> 
>> src/hotspot/share/runtime/javaThread.cpp|hpp
>> These files don't exist in 17, they were split off from thread.cpp|hpp.
>> The modiefied code looks very similar in 17, so I could easily identify 
>> the changed code and apply the patches.  Usual minor nullptr diffs etc.
>> 
>> src/hotspot/share/runtime/monitorChunk.cpp
>> Resolved because of NULL/nullptr difference
>> 
>> src/hotspot/share/runtime/monitorChunk.hpp
>> Resolved because of context diff.
>> 
>> src/hotspot/share/runtime/synchronizer.cpp
>> Resolved. Different checks in if condition.
>> One is for the new locking mode in head, the other whether the mask has a locker.
>> ... check ...
>> 
>> src/hotspot/share/runtime/thread.cpp
>> Resolved because new locking modes added an assertion in removed function is_lock_owned.
>> 
>> src/hotspot/share/runtime/vframeArray.cpp
>> Resolved. Code differs because 21 has biased locking removed.
>> 
>> 
>> 
>> To make it build, I had to replace
>> Javathread::cast() by as_Java_thread() in synchronizer.cpp.
>> 
>> Also I removed is_linked() from monitorChunk.hpp.
>
> Goetz Lindenmaier has updated the pull request incrementally with one additional commit since the last revision:
> 
>   Fix assertion message.

Marked as reviewed by rrich (Reviewer).

-------------

PR Review: https://git.openjdk.org/jdk17u-dev/pull/3177#pullrequestreview-2542175283