RFR: 8348890: Fix docs for -XX:AOT* options in java man page [v2]
Ioi Lam
iklam at openjdk.org
Tue Jan 28 22:12:47 UTC 2025
On Tue, 28 Jan 2025 20:57:56 GMT, John R Rose <jrose at openjdk.org> wrote:
>> Ioi Lam has updated the pull request incrementally with one additional commit since the last revision:
>>
>> @rose00 review comment
>
> src/java.base/share/man/java.md line 4014:
>
>> 4012: warm-up performance of Java applications.
>> 4013:
>> 4014: The Ahead-of-Time Cache (AOT Cache) is a container introduced in JDK 24 for
>
> s/Cache/cache/ here also
Fixed.
> src/java.base/share/man/java.md line 4019:
>
>> 4017: artifacts, such as execution profiles and compiled methods.
>> 4018:
>> 4019: An AOT cache is specific to a combination of the following:
>
> of the following
> +application configuration options/data/states/conditions:
>
> (Or something specific like that.)
Actually the AOT cache is more permissive with changes in command-line arguments. However, the three items below are absolute requirements that must be met. That's why I discuss the command-line option compatibility in the "loading of an AOT cache can fail for a number of reasons ..." section below. See the rendered text in https://github.com/iklam/jdk/blob/8348890-fix-xx-aot-docs-in-man-page/src/java.base/share/man/java.md#ahead-of-time-cache
> src/java.base/share/man/java.md line 4025:
>
>> 4023: - A particular OS and CPU architecture.
>> 4024:
>> 4025: If any of the above changes, you must recreate the AOT cache.
>
> Maybe add "The main class may be changed."
You can change the main class or parameters to the application and will still be able to use the cache, although if you execute new code that was cover during the training run then you will have less benefit from the cache.
-------------
PR Review Comment: https://git.openjdk.org/jdk/pull/23343#discussion_r1932930334
PR Review Comment: https://git.openjdk.org/jdk/pull/23343#discussion_r1932930227
PR Review Comment: https://git.openjdk.org/jdk/pull/23343#discussion_r1932930183
More information about the hotspot-runtime-dev
mailing list