Re: [RFC] GC: external thread suspension

[Bryce McKinlay <mckinlay at redhat dot com>]

Keith Seitz wrote:
> So now there are two routes. One (complex) patch which allows garbage 
> collection to run on suspended threads (or at least I think it does!) 
> and one (simple) patch which suspends garbage collection while the 
> thread is externally suspended.

If the JDWP back-end is partially written in Java, then it is important 
that GC can run when threads are suspended. If not, the back-end will be 
prone to deadlock during memory allocation. Eventually, I think JDWP 
implementation will need to be free of any Java code/locks in order to 
be 100% robust in the face of thread suspension, but currently that's 
not the case. Also, even for a JDWP implementation in pure C/C++ it may 
be desirable to make use of the GC.

I like how GC_suspend_self and GC_suspend_thread are handled in the same 
way in this patch, however. Perhaps the two approaches can be combined?

Bryce