RFR: 8287024: G1: Improve the API boundary between HeapRegionRemSet and G1CardSet

[Albert Mingkun Yang]

On Thu, 19 May 2022 14:53:59 GMT, Thomas Schatzl <tschatzl at openjdk.org> wrote:

> Hi all,
> 
>   can I get reviews for this small refactoring that clarifies the boundaries of what kind of data `HeapRegionRemSet` and `G1CardSet` should operate on?
> 
> * `HeapRegionRemSet` deals with addresses, heap regions
> * `G1CardSet` deals with card indexes in the heap, card regions and cards within card regions
> 
> The `G1CardSet::add_card()` and `G1CardSet::contains_card()` methods are fuzzy about that, they require the caller (`HeapRegionRemSet`) to know about card regions and cards within regions. 
> 
> Testing: gha
> 
> Thanks,
>   Thomas

A very minor comment.

src/hotspot/share/gc/g1/g1CardSet.cpp line 703:

> 701: }
> 702: 
> 703: void G1CardSet::split_card(uintptr_t offset, uint& card_region, uint& card_within_region) const {

I think `card` is more specific than `offset`, IMO.

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

Marked as reviewed by ayang (Reviewer).

PR: https://git.openjdk.java.net/jdk/pull/8790