RFR: 8331560: Refactor Hotspot container detection code so that subsystem delegates to controllers [v4]

Severin Gehwolf sgehwolf at openjdk.org
Wed Jun 19 14:43:29 UTC 2024


On Mon, 17 Jun 2024 10:32:10 GMT, Johan Sjölen <jsjolen at openjdk.org> wrote:

>> Severin Gehwolf has updated the pull request with a new target base due to a merge or a rebase. The pull request now contains 36 commits:
>> 
>>  - Tabs
>>  - Templated class use references
>>  - initializer lists
>>  - Style nit
>>  - Get rid of multiple inheritance
>>  - Merge branch 'master' into jdk-8331560-cgroup-controller-delegation
>>  - Review feedback
>>  - Clean up
>>  - Appropriately handle version specific printing
>>  - Fix inheritance hierarchy
>>  - ... and 26 more: https://git.openjdk.org/jdk/compare/cc64aeac...6aff2076
>
> src/hotspot/os/linux/cgroupV1Subsystem_linux.cpp line 93:
> 
>> 91: 
>> 92: static inline
>> 93: void do_trace_log(julong read_mem_limit, julong host_mem) {
> 
> Should it be called trace when we log on a debug level?

renamed to `verbose_log`

> src/hotspot/os/linux/cgroupV1Subsystem_linux.hpp line 56:
> 
>> 54: 
>> 55:   private:
>> 56:     CgroupV1Controller* _reader;
> 
> It seems simpler to me to have this as a `CgroupV1Controller` and perhaps have a `reader()` accessor that returns a pointer. This class currently has a memory leak as the malloc-allocated `_reader` is never freed in the destructor. Up to you what you want to do.

Done that now.

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

PR Review Comment: https://git.openjdk.org/jdk/pull/19085#discussion_r1646324403
PR Review Comment: https://git.openjdk.org/jdk/pull/19085#discussion_r1646325964


More information about the hotspot-runtime-dev mailing list