Yi Lin
Yi Lin
Currently we added a plan-specific flag `may_trace_duplicate_edges`. For marksweep, it is set to true, and for other plans, it is false. When a plan allows tracing duplicate edges, we simply...
> This could just be implemented by object pinning, right? If you are using conservative roots, then we can just pin the roots. Right. But pinning requires using object metadata...
I think `Mutator` is `Send` because we move it to a new thread when we create a new mutator.
That is interesting and makes sense. Probably we should use `Mutex` for `GCWork`. `GCWork` should be `Send` and not `Sync`.
With a discussion with Steve, this does not seem to be an issue. We could just use the min of size limits of generational and immix and use it as...
Wenyu mentioned that the object size may change during copying. For example, JikesRVM may add an extra word to objects during copying. This means an object whose size is fine...
Another overlooked point in the current implementation is the allocation semantic during copying. Basically we are ignoring that, as we only copy from nursery to mature. If we are going...
We have APIs to allow VMs to do malloc with MMTk (and count that into MMTk's heap). This PR allows VM to manage some allocation by themselves, but also count...
Will https://github.com/mmtk/mmtk-core/issues/573 solve this issue as well?
binding-refs OPENJDK_BINDING_REPO=qinsoon/mmtk-openjdk OPENJDK_BINDING_REF=update-pr-632 JIKESRVM_BINDING_REPO=qinsoon/mmtk-jikesrvm JIKESRVM_BINDING_REF=update-pr-632