Bug ID: JDK-8015270 @Contended: fix multiple issues in the layout code

JDK-8015270 : @Contended: fix multiple issues in the layout code

Type: Bug
Component: hotspot
Sub-Component: runtime
Affected Version: hs25

Priority: P3
Status: Closed
Resolution: Fixed

Submitted: 2013-05-23
Updated: 2014-06-26
Resolved: 2013-05-27

Versions (Unresolved/Resolved/Fixed)

The Version table provides details related to the release that this issue/RFE will be addressed.

Unresolved : Release in which this issue/RFE will be addressed.
Resolved: Release in which this issue/RFE has been resolved.
Fixed : Release in which this issue/RFE has been fixed. The release containing this fix may be available for download as an Early Access Release or a General Availability Release.

To download the current JDK release, click here.

JDK 8	Other
8Fixed	hs25Fixed

Related Reports

Duplicate :	JDK-8014488 - @Contended classes may trigger 'guarantee(false) failed: boom' during verify
Duplicate :	JDK-8014498 - @Contended classes triggers 'guarantee(PageArmed == 0) failed: invariant' on linux-amd64
Duplicate :	JDK-8014517 - @Contended: * glibc detected * java: double free or corruption (!prev) ***
Duplicate :	JDK-8014964 - @Contended breaks has_nonstatic_fields invariant
Duplicate :	JDK-8014529 - @Contended: SIGBUS during class parsing
Duplicate :	JDK-8014886 - @Contended fields can overrun oop maps
Duplicate :	JDK-8014526 - @Contended: ShouldNotReachHere()
Relates :	JDK-8014529 - @Contended: SIGBUS during class parsing
Relates :	JDK-8014517 - @Contended: * glibc detected * java: double free or corruption (!prev) ***
Relates :	JDK-8014498 - @Contended classes triggers 'guarantee(PageArmed == 0) failed: invariant' on linux-amd64
Relates :	JDK-8014488 - @Contended classes may trigger 'guarantee(false) failed: boom' during verify

Description

This is the umbrella issue for various @Contended touchups.

The issues arguably manifest in the VM crashes/corruptions when specially-crafted @Contended classes are available. See linked issues for the observed failure modes. 

The more concrete failures are only happening on new hotspot/test/runtime/contended/{HasNonStatic,OopMaps} regression tests *with* the new asserts. This opens up the way to test the fix: revert the change, put the asserts back in, and watch the ClassFileParser::layout_fields asserts fire on fastdebug build.

Comments

Added more thorough comment. Update: thanks!
27-05-2013
Just to be clear: How do the new tests behave if run with a VM/JDK without the entire fix proposed for this bug? For example, if these tests are run against HSX-25-B33/JDK8-B91, then what happens? Do they crash in terrible searing pain? Cause a solar eclipse? If they only fail with the new asserts in place, then the VM/SQE engineer is going to need to know that.
27-05-2013
Related failures.
27-05-2013
The priority for one of the duplicates was wrong, restoring back to P3.
27-05-2013
The following bug under the umbrella is a P2: JDK-8014886 @Contended fields can overrun oop maps so that makes this bug a P2.
27-05-2013