RFR: 8308094: Add a compilation timeout flag to catch long running compilations [v5]

[Manuel Hässig]

On Wed, 13 Aug 2025 08:31:11 GMT, Christian Hagedorn <chagedorn at openjdk.org> wrote:

>> Manuel Hässig has updated the pull request with a new target base due to a merge or a rebase. The incremental webrev excludes the unrelated changes brought in by the merge/rebase. The pull request contains 11 additional commits since the last revision:
>> 
>>  - Merge branch 'master' into JDK-8308094-timeout
>>  - Rename _timer
>>  - remove _timeout_armed
>>  - ASSERT
>>  - Merge branch 'master' into JDK-8308094-timeout
>>  - No acquire release semantics
>>  - Factor Linux specific timeout functionality out of share/
>>  - Move timeout disarm above if
>>  - Merge branch 'master' into JDK-8308094-timeout
>>  - Fix SIGALRM test
>>  - ... and 1 more: https://git.openjdk.org/jdk/compare/098f25d4...8bb5eb7a
>
> src/hotspot/os/linux/compilerThreadTimeout_linux.hpp line 46:
> 
>> 44: #endif // !PRODUCT
>> 45:  public:
>> 46:   CompilerThreadTimeoutLinux() NOT_PRODUCT(DEBUG_ONLY(: _timer(nullptr))) {};
> 
> Why do you need the `NOT_PRODUCT`? It only wraps `DEBUG_ONLY`. If that's not set, the `NOT_PRODUCT` wraps nothing.

The initialization list should only be generated if `!PRODUCT && ASSERT`, so it does not appear in `optimized`. This is one way of expressing this conjunction in macros.

> src/hotspot/share/compiler/compilerThread.hpp line 54:
> 
>> 52:   void disarm() { return; };
>> 53:   bool init_timeout() { return true; };
>> 54: };
> 
> Should we also guard this with `ifndef LINUX` since it's only used for non-Linux?

Sounds reasonable.

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

PR Review Comment: https://git.openjdk.org/jdk/pull/26023#discussion_r2276419674
PR Review Comment: https://git.openjdk.org/jdk/pull/26023#discussion_r2276420718