RFR: 8311993: Test serviceability/sa/UniqueVtableTest.java failed: duplicate vtables detected [v2]

[Alex Menkov]

On Wed, 28 Aug 2024 03:42:27 GMT, Chris Plummer <cjplummer at openjdk.org> wrote:

> ```
>   public boolean isSharingEnabled() {
>     if (sharingEnabled == null) {
>         Address address = VM.getVM().getDebugger().lookup(null, "UseSharedSpaces");
>         if (address == null && getOS().equals("win32")) {
>             // On Win32 symbols are prefixed with the dll name. So look for
>             // UseSharedSpaces as a symbol in jvm.dll.
>             address = VM.getVM().getDebugger().lookup(null, "jvm!UseSharedSpaces");
>         }
>         sharingEnabled = address.getJBooleanAt(0);
>     }
>     return sharingEnabled.booleanValue();
>   }
> ```
> 
> I'm not sure why this is failing. Based on the existing code and comments, it seems at some point SA worked without relying on the windbg native symbol support. Code like isSharingEnabled() probably got added afterwards and was never tested with it disabled.

"UseSharedSpaces" is exported from jvm.dll, but the issue here on Windows SymbolLookup searches for decorated symbols (i.e. "??_7UseSharedSpaces@@6B@") and the symbol is exported undecorated.
>From globalDefinition.hpp (note extern "C"):


extern "C" {
// Make sure UseSharedSpaces is accessible to the serviceability agent.
extern JNIEXPORT jboolean UseSharedSpaces;
}

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

PR Comment: https://git.openjdk.org/jdk/pull/20684#issuecomment-2316420789