RFR: 8229147: Linux os::create_thread() overcounts guardpage size with newer glibc (>=2.27)

[David Holmes]

On Fri, 21 Apr 2023 13:51:46 GMT, Aleksey Shipilev <shade at openjdk.org> wrote:

>> src/hotspot/os/linux/os_linux.hpp line 152:
>> 
>>> 150:   static size_t default_guard_size(os::ThreadType thr_type);
>>> 151: 
>>> 152:   static bool AdjustStackSizeForGuardPages; // See comments in os_linux.cpp
>> 
>> With the capitalization like that, this looks like a global flag. Other statics in this file have different style. Suggestion: `_adjust_stacksize_for_guard_pages`.
>
> Can we also assert the lifecycle for this flag is correct? E.g.:
> 
> 
> private:
>  static bool _init_adjust_...; // Maybe simpler name?
>  static bool _adjust_...;
> 
>  static void init_...() {
>    ...
>    _init_adjust_... = true;
>  }
> 
> public:
>  static bool adjust_stacksize_for_guard_pages() {
>    assert(_init_adjust..., "Should be initialized");
>    return _adjust_...;
>  }

> With the capitalization like that, this looks like a global flag. Other statics in this file have different style. Suggestion: `_adjust_stacksize_for_guard_pages`.

The use of initial underscore is for "local" variables, this one has to be exported via `os::Linux:AdjustStackSizeForGuardPages`. I also thought it matched nicely with `AdjustStackSizeForTLS` even though not actually a global.

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

PR Review Comment: https://git.openjdk.org/jdk/pull/13571#discussion_r1174241769