[jdk21u-dev] RFR: 8348625: [21u, 17u, 11u] Revert JDK-8185862 to restore old java.awt.headless behavior on Windows

[Severin Gehwolf]

On Sun, 26 Jan 2025 12:35:01 GMT, Christoph Langer <clanger at openjdk.org> wrote:

> This basically reverts 78e35ea113bed4a5f66225836c887a45696bbf44 to restore old behavior with regards to headless handling on Windows.

Differences to a revert of 78e35ea113bed4a5f66225836c887a45696bbf44 are:

- `src/java.desktop/share/classes/java/awt/GraphicsEnvironment.java` unused import isn't back re-introduced. OK
- `src/java.desktop/windows/classes/sun/awt/Win32GraphicsEnvironment.java` extra space in comment not back re-introduced. OK.
- `src/java.desktop/windows/native/libawt/windows/awt_Win32GraphicsDevice.cpp`: `VERIFY` statements kept. Basically an assert, which should be fine.
- `src/java.desktop/windows/native/libawt/windows/awt_Win32GraphicsEnv.cpp` the `return` in the failure case is not re-introduced. We are done with the method anyway. OK.

Seems fine to me. But it would be good if a windows person could take a look and test it as well. @fthevenet perhaps?

@RealCLanger Please fix the title mismatch. Thanks!

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

Marked as reviewed by sgehwolf (Reviewer).

PR Review: https://git.openjdk.org/jdk21u-dev/pull/1359#pullrequestreview-2575903899
PR Comment: https://git.openjdk.org/jdk21u-dev/pull/1359#issuecomment-2616235036