Porting request/PR out: G1: Improve Application Throughput with a More Efficient Write-Barrier [Was: Re: Porting request for JEP Reduce Latency of G1 Write Barriers]

[Thomas Schatzl]

Hi all,

 > On 22.10.24 10:26, Thomas Schatzl wrote:
 >> Hi all,
 >>
 >>    we in the GC team have been experimenting with new post-write
 >> barriers for G1 to reduce the throughput difference between G1 and
 >> Parallel GC.
 >>
 >> After quite a few attempts we think we found a balanced solution
 >> between overall complexity and impact on throughput and latency.
 >>
 >> [...]

On 08.11.24 16:39, Thomas Schatzl wrote:
> Hi all,
> 
 >[...]
> 
> The implementation CR also already contains some more information about 
> the implemetation (https://bugs.openjdk.org/browse/JDK-8342382), which 
> will also be updated a little before asking for reviews.

   we in the Oracle gc team decided that due to the size of the change 
(+3k, -4.4k LOC) are not waiting any longer for the JEP moving to 
candidate before asking for reviews.

There is a PR out for review for this change now [4]. I would like to 
ask you maintainers to have a look/test.

M. Doerr from SAP already confirmed that the change is still working 
:)(thanks!).

I am not expecting any or any serious issues for the other platforms 
since not much changed since asking for the ports. However in particular 
I hope for feedback from maintainers for the ARM (32 bit) and the RISCV 
port, as I have only been able to perform very limited testing using 
cross-compilers and emulators there.

Another confirmation for the S390 port that things still work would be 
appreciated too. :)

Thanks,
   Thomas

[4] https://github.com/openjdk/jdk/pull/23739