RFR: 8294847: Fix calculation of G1 effective scanned cards prediction [v2]

Thomas Schatzl tschatzl at openjdk.org
Wed Oct 12 07:33:24 UTC 2022


On Tue, 11 Oct 2022 18:10:34 GMT, Kim Barrett <kbarrett at openjdk.org> wrote:

>> Thomas Schatzl has updated the pull request incrementally with one additional commit since the last revision:
>> 
>>   kbarrett review
>
> src/hotspot/share/gc/g1/g1Analytics.hpp line 144:
> 
>> 142: 
>> 143:   // Predict how many of the given remembered set of length rs_length will add to
>> 144:   // the number of total cards scanned.
> 
> I'm having a very hard time parsing that comment.  I _think_ it is equivalent to "Predict how many cards in a remembered set of length rs_length will need to be scanned."

This predictor is a bit complicated: it predicts for that given remembered set length, what part of it adds to the number of scanned cards - assuming that the pending log buffers are always scanned, we need to find out how many additional cards the remembered set will dirty (and need to be scanned). I.e. the remembered set cards that do not duplicate the log buffer cards.
I tried to reformulate that, please let me know if it is more understandable now.

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

PR: https://git.openjdk.org/jdk/pull/10647



More information about the hotspot-gc-dev mailing list