On Wed, 4 Feb 2026 09:05:14 GMT, Sergey Bylokhov <[email protected]> wrote:
> ReleasePrimitiveArrayCritical is called with JNI_ABORT, which discards pixel > data when the JVM copies the array. Changed to 0 to always copy back FWIW I see the aarch64 version of the test is still pending. I thought it had passed. So if it fails there too it is probably a config issue and since these headless tests are run without desktop access, I wonder if this test in fact really needs it for Aqua. I'm running on my desktop so it has access so that isn't the headless environment. Maybe you need to make this test headful - at least for Mac. ------------- PR Comment: https://git.openjdk.org/jdk/pull/29563#issuecomment-3894039518
