RFR: JDK-8319314: NMT detail report slow or hangs for large number of mappings

[Thomas Stuefe]

On Fri, 3 Nov 2023 09:25:27 GMT, Johan Sjölen <jsjolen at openjdk.org> wrote:

>> If a lot of mappings have been registered with NMT, NMT detail report may get so slow as to hang seemingly. 
>> 
>> 100 k mappings will take > 1 minute
>> 1 million mappings never returns (I canceled the test after 90 minutes)
>> 
>> We burn all the time doing a sorted insert into the target array of the MemBaseline object. That sorted insert is pointless since the source array is already sorted by region base address.
>> 
>> Omitting the pointless sort drops the runtime for 1 million mappings from > 90 minutes to 48 seconds.
>> 
>> Tests: Tested locally. GHAs (Windows error unrelated).
>
> src/hotspot/share/nmt/memBaseline.cpp line 130:
> 
>> 128: 
>> 129:   bool do_allocation_site(const ReservedMemoryRegion* rgn)  {
>> 130:     assert(rgn->base() >= _last_base, "region unordered?");
> 
> It seems like we don't assign to `_last_base`?

Ah, missed that. Thanks.

Wonder why no mails are sent to MLs though.

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

PR Review Comment: https://git.openjdk.org/jdk/pull/16480#discussion_r1381443268