Jimmy K

I have some more tests that look like they are failing now. So I'll need to figure out the problem before this can be merged.

I found the problem. https://github.com/eclipse/omr/blob/8c233cdf9151d5b097733a426537d4fd6c37051f/compiler/optimizer/VPHandlers.cpp#L5410 `isUnsafeCAS()` does not pass in a parameter. https://github.com/eclipse/omr/blob/8c233cdf9151d5b097733a426537d4fd6c37051f/compiler/compile/OMRMethod.hpp#L153 So NULL is passed in as the `TR::Compilation *`. https://github.com/IBMJimmyk/openj9/blob/8f7d8c2ce7914c5ae4a80413002fa8221077f7cb/runtime/compiler/env/j9method.cpp#L5545-L5548 This is the temporary platform check that...

I retested after I added the fix to `isUnsafeCAS` so it doesn't use a NULL comp object and the tests pass now. @hzongaro Can you look over the latest changes?...

I fixed the commit message.

The Windows failure looks to be this: ``` [WS-CLEANUP] Deleting project workspace... 16:39:01 [WS-CLEANUP] Deferred wipeout is disabled by the job configuration... 16:39:19 ERROR: Cannot delete workspace :Unable to delete...

I'll take a look at the sanity.functional failures for Windows.

All of the JDK8 Windows failures fall into the same pattern. The test workload itself is successful and prints the expected result for the test and shortly afterwards there is...

I was able to reproduce the JDK8 Windows failures on a build that excluded both my Unsafe CAS work and the `j9shr29.dll` crash fix. So I can say those failures...

I can try taking a look at this.

I tried running this workload locally on a ppcle Linux machine and observed hangs even under `-Xint`. Machine: `pjavaubun02.rtp.raleigh.ibm.com` Java build is a JDK24 with recent source code: ``` openjdk...