[jdk8u-dev] RFR: 8280963: Incorrect PrintFlags formatting on Windows

[Alex Kasko]

On Wed, 20 Apr 2022 21:01:34 GMT, Alex Kasko <akasko at openjdk.org> wrote:

> A fix to 8u-only Windows-specific problem with -XX:+PrintFlags* output.
> 
> Problem description on Stack Overflow by Andrei Pangin: https://stackoverflow.com/a/63309395
> 
> The problem was fixed in jdk9 as part of [JDK-8042893](https://bugs.openjdk.java.net/browse/JDK-8042893), but this [jdk9 change](https://hg.openjdk.java.net/jdk9/jdk9/hotspot/rev/115188e14c15) doesn't look like a good candidate for the 8u backport according to [8u Best Practices](https://mail.openjdk.java.net/pipermail/jdk8u-dev/2020-June/012002.html). Still the problem manifests itself as a customer-visible bug on Windows because people do use -XX:+PrintFlagsFinal to find out the actual flags picked up by JVM and can be confused by unexpected zeros in the output. Proposed patch cherry-picks the minimal change from [JDK-8042893 commit](https://hg.openjdk.java.net/jdk9/jdk9/hotspot/rev/115188e14c15#l53.1).
> 
> Mailing list review: https://mail.openjdk.java.net/pipermail/jdk8u-dev/2022-February/014532.html
> 
> Testing:
> 
>  - [x] regression test is included with the proposed patch
>  - [x] checked that it builds on Windows 64-bit and 32-bit on VS2017 and VS2010
>  - [x] ran jtreg:hotspot/test/runtime on Windows and Linux

Would it be better if I add a `PRAGMA_FORMAT_MUTE_WARNINGS_FOR_GCC` change as a separate backport?

I understand that it is a part of the same original change, just the motivation behind the change is completely different from the format strings change that are included now. I am fine with doing it either way, just wanted to point out that the inclusion of pragma here may be confusing.

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

PR: https://git.openjdk.java.net/jdk8u-dev/pull/45