RFR: 8328278: Do not print the tenuring threshold in AgeTable::print_on [v3]

[Guoxiong Li]

On Tue, 2 Apr 2024 10:15:21 GMT, Albert Mingkun Yang <ayang at openjdk.org> wrote:

>>> It looks as if you are now suppressing the tenuring threshold printing entirely? In a sense that is a crucial part of what one looks at in GC logs. May be I am missing something.
>> 
>> When computing the threshold, the threshold will be printed. Please read the method `AgeTable::compute_tenuring_threshold`.
>> 
>>> Could you share for both G1 and for Serial, the before and after logs with `+PrintTenuringDistribution` ?
>> 
>> Currently, the option `-XX:+PrintTenuringDistribution` is not available in mainline, we can use `-Xlog:gc+age*=trace` instead.
>
>> Could you share for both G1 and for Serial, the before and after logs with +PrintTenuringDistribution ?
> 
> Using `-Xlog:gc+age=trace`:
> 
> 
> ## baseline
> [5.527s][debug][gc,age] GC(0) Desired survivor size 6979584 bytes, new threshold 1 (max threshold 15)
> [5.527s][trace][gc,age] GC(0) Age table with threshold 1 (max threshold 15)
> [5.527s][trace][gc,age] GC(0) - age   1:   13959168 bytes,   13959168 total
> 
> ## new
> [4.196s][debug][gc,age] GC(0) Desired survivor size 6979584 bytes, new threshold 1 (max threshold 15)
> [4.196s][trace][gc,age] GC(0) Age table:
> [4.196s][trace][gc,age] GC(0) - age   1:   13959160 bytes,   13959160 total
> 
> 
> @lgxbslgx Could you merge master?

@albertnetymk @ysramakrishna Thanks for the review.

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

PR Comment: https://git.openjdk.org/jdk/pull/18329#issuecomment-2033798820