On Tue, 18 Aug 2026 13:23:39 GMT, Casper Norrbin <[email protected]> wrote:

> I don't think there is any dependancy on the current (previous) format. The 
> only JDK reference is the test testing the command, and the command has only 
> been visible for a very short while. Keeping both formats supported was not 
> the intention, rather a coincidence since we need internal names when probing 
> the symbol table.

Oh I misread this a bit.  Yes here inside the probe method it's internal format.

We can get away with swapping / for . specifically in class names I suppose.  
This command ends up being a bit permissive in what it accepts, in a way that  
VM.class_hierarchy is not.  8-)

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

PR Comment: https://git.openjdk.org/jdk/pull/32396#issuecomment-5329097324

Reply via email to