Kunshan Wang
Kunshan Wang
Interesting. This workaround takes advantage of the fact that the underlying languages, JS and Python, are dynamically typed. I think similar things already happened for function parameters. We use `欲行是術。必先得三數。曰「甲」。曰「乙」。曰「丙」。`...
> 今天突然又想到一个词，“志曰”，这在史书中常见，用于表示打印日志也正合宜。 > 不过在缺少操作数时，“志之”听起来不良，或可用“志焉”。 “焉”其实就是“于之”的意思，就像“诸”是“之于”的意思。这里我觉得就是“志之”而不是“志于之”，因为“之”这里指代那句话。 其实“志之”并没有什么。论语里也有“默而识之”，“识”通“志”。
Yes. It needs to be merged/rebased with the `master` version because the architecture and the API have changed a lot. I also see `#[cfg(feature="malloc")]` macros everywhere. The code is significantly...
There is one more problem. If multiple inter-connected objects die at the same time, they will all be enqueued for finalization. However, before the next GC, the mutator may only...
@qinsoon (1) is not necessary. Vanilla Ruby does only (2), and will refuse to report an object as root if the header is cleared. But having (1) allows a stack...
> The issue with (2) is that it is not sound w.r.t. the "valid object" semantics. Some random memory may return a false positive. If the "valid object" semantics means...
@qinsoon You are right. The existing 'alloc bit' API (`is_mmtk_object`) is **sufficient** to support Ruby. This proposed API change is an *optimization* so that we can rule out "invalid" objects...
From our discussion in the evening, the "cyclic dead objects" problem can be solved by requiring the GC to inspect the VO bit before attempting to scan an object. Alternatively,...
@steveblackburn What name should we give to the thing returned by `alloc`? It is "a region of memory where the VM is allowed to write in, and is managed by...
In https://github.com/mmtk/mmtk-core/pull/606#discussion_r955626986, Wenyu suggested moving `VMEdge` to a different trait. But doing so will make it hard to access for the same reason mentioned here. With this change, it is...