URGENT RFR(XS): 8213487 [BACKOUT] 8213414 Fix incorrect copy constructors in hotspot

Daniel D. Daugherty daniel.daugherty at oracle.com
Wed Nov 7 16:42:34 UTC 2018


Greetings,

This is an urgent fix for solving the jdk/jdk CI failures in tier[1-3] 
testing.

     JDK-8213487 [BACKOUT] 8213414 Fix incorrect copy constructors in 
hotspot
     https://bugs.openjdk.java.net/browse/JDK-8213487

Here's the backout webrev:

http://cr.openjdk.java.net/~dcubed/8213487-webrev/0_for_jdk_jdk/


Here's the original bug fix being backed out:

     JDK-8213414 Fix incorrect copy constructors in hotspot
     https://bugs.openjdk.java.net/browse/JDK-8213414

Here's the original webrev:

http://cr.openjdk.java.net/~ihse/JDK-8213414-fix-hotspot-copy-constructors/webrev.01


Here's the failure caused by JDK-8213414:

     JDK-8213481 tests failing "assert(~(_allocation_t[0] | 
allocation_mask) != (uintptr_t)this
                 || !is_type_set()) failed: embedded or stack only"
     https://bugs.openjdk.java.net/browse/JDK-8213481


The backout was generated using:

     $ hg backout 44f34d2c3243
     3 files updated, 0 files merged, 0 files removed, 0 files unresolved

The resulting repo was tested with the following test:

     runtime/logging/ClassInitializationTest.java

which was used to triage JDK-8213481. With the backout in place,
runtime/logging/ClassInitializationTest.java no longer fails.

The backout repo is also being tested via a 
"builds-tier1,hs-tier1,jdk-tier1"
Mach5 job.

Thanks, in advance, for any comments, suggestions or questions.

Dan



More information about the hotspot-runtime-dev mailing list