RFR: 8256265: G1: Improve parallelism in regions that failed evacuation [v2]

[Thomas Schatzl]

On Tue, 13 Sep 2022 09:44:35 GMT, Thomas Schatzl <tschatzl at openjdk.org> wrote:

>> src/hotspot/share/gc/g1/g1YoungGCPostEvacuateTasks.cpp line 373:
>> 
>>> 371:     bool do_heap_region(HeapRegion* r) override {
>>> 372:       G1CollectedHeap::heap()->clear_bitmap_for_region(r, false /* update_tams */);
>>> 373:       return false;
>> 
>> I wonder if it's possible to assert `tams == bottom` here.
>
> No, after some thinking. G1 clears TAMSes during concurrent clearing of the bitmap. This marking phase is already concurrent to the mixed phase. We could reset TAMS anyway though.

On second thought, yes bottom should be tams for regions in the collection set (of which retained regions are a subset).

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

PR: https://git.openjdk.org/jdk/pull/9980