Jimmy K

I am currently in the middle of making a build that includes https://github.com/eclipse-openj9/openj9/pull/22370 to see if the problem still occurs.

Unfortunately, it did not help. I ran an internal grinder to perform 1000 runs and got 1 failure which can be seen here: https://hyc-runtimes-jenkins.swg-devops.com/job/Grinder_iteration_1/6098/ The error message is different from...

That looks to be the case: ``` (kca) j9c 0x6C9200 Class Path/Name: {jdk/internal/foreign/BufferStack$PerThread$Frame} J9Class 0x00000000006c9200 ClassObject: 0x00000000e03f7578 Access: Final Synchronized (30) ClassLoader: 0x00007898a4140768 Object: 0x00000000e0040330 {jdk/internal/loader/ClassLoaders$BootClassLoader}T SubClassLink: 0x0000000000634e00 {java/lang/invoke/LambdaForm$VH/0x00000000880049e0} Instance...

From the javacore file: ``` 2XHREGISTER R2: 0000000000000000 2XHREGISTER R3: 00000000FA2F9360 ... 2XHREGISTER R30: 00000000E047EA60 ``` From the corefile: ``` 0x7898877c93b0 {jdk/.../SlicingAllocator.canAllocate} +15 10007e80 lwz r3, 0x10(r30) //

Looking at the same corefile, r3 is supposed to be an object of type `jdk/internal/foreign/NativeMemorySegmentImpl`. But it actually points in the middle of the object: ``` > whatis 0xfa2f9360 heap...

Also, I'm looking at the result of `!stackslots 0x68F900` when looking at a corefile and I'm trying to understand what I'm seeing: ``` JIT frame: bp = 0x0000000000775858, pc =...

Getting a `tracefull` log sounds like a good next step. I will try that and see what it says.

`0x00007898877C93C0` is where the segmentation fault occurred. r30 is changed a little bit before that. ``` 0x7898877c939c {jdk/.../SlicingAllocator.canAllocate} +10 2800cefb std r30, 0x28(r14) //

`r3` holds a bad value. It holds the value of `0xfa2f9360` and that address is in the middle of a `NativeMemorySegmentImpl` object. ``` > whatis 0xfa2f9360 heap #1 - name:...