RFR: Some more descriptive names, removed redundant asserts. Request for comments. [v2]

[Ron Pressler]

On Tue, 5 Apr 2022 23:23:59 GMT, Coleen Phillimore <coleenp at openjdk.org> wrote:

>> src/hotspot/share/runtime/continuation.cpp line 2222:
>> 
>>> 2220:   }
>>> 2221: 
>>> 2222:   // Comment needed: why is 500? Related to 300 in can_thaw_fast?
>> 
>> It's a heuristic. Below this number, we thaw the whole chunk. Above it, we thaw one frame. 500 seemed to work ok, but might need to be adjusted after deeper profiling.
>
> Unrelated to 300 above then?  Seems like we've stack overflow checked for 300 words so that would be a good slop number for both?

It's unrelated. The 300 is just an estimate for the stack size taken by the native code. 500 is just a heuristic of when to thaw one frame or the entire chunk (so refers to Java frames). The function always makes room for the entire chunk.

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

PR: https://git.openjdk.java.net/loom/pull/139