On Wed, 18 Jun 2025 16:20:14 GMT, Manukumar V S <m...@openjdk.org> wrote:
> Issue: > MyanmarTextTest.java produces a false positive result when some of the test > preconditions are not met. It checks whether certain fonts are present in the > system, for example, 'Padauk' on Linux. If the required font is missing, the > test simply returns early, and the test ends up passing, which is incorrect. > Ideally, it should throw a jtreg.SkippedException when the necessary > preconditions are not satisfied. > > Another scenario is that the test passes on headless machines even though it > creates GUI components. Ideally, when GUI components are created in code > running on a headless machine, a HeadlessException should be thrown. However, > since MyanmarTextTest.java exits before reaching the point where the GUI is > created (due to unmet preconditions), it incorrectly reports a pass. This > behavior may lead to a misinterpretation of the test as being headless, which > it is not. > > Fix: > Need to throw jtreg.SkippedException in cases where some pre-conditions for > running the test are not met. > > Testing: > Tested using mach5 in all available platforms and got full PASS. test/jdk/java/awt/font/TextLayout/MyanmarTextTest.java line 56: > 54: > 55: private static final String FONT_WINDOWS = "Myanmar Text"; > 56: private static final String FONT_LINUX = "Padauk"; A more useful update might be to update the test to look for more fonts on Linux. I don't see Padauk on my Ubuntu 24.04 system. I do have various Noto ones - eg "Noto Sans Myanmar" So on Linux I would look for BOTH fonts before giving up. ------------- PR Review Comment: https://git.openjdk.org/jdk/pull/25879#discussion_r2155245925
