On Wed, 27 Oct 2021 12:42:50 GMT, Prasanta Sadhukhan <psadhuk...@openjdk.org> wrote:
> As per JDK-8252813, some tests fails recurringly in CI macos system. This is > an attempt to fix the swing tests. > It was seen from the logs that we have color mismatch in these tests. > > For example, in PressedIcon test, we had following log > > ----------System.out:(1/75)---------- > color java.awt.Color[r=55,g=55,b=55] COLOR_1Xjava.awt.Color[r=255,g=0,b=0] > ----------System.err:(13/842)---------- > java.lang.RuntimeException: Colors is different for scale=1! > at PressedIconTest.main(PressedIconTest.java:97) > at > java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native Method) > at > java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:76) > at > java.base/jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:51) > at java.base/java.lang.reflect.Method.invoke(Method.java:569) > at > com.sun.javatest.regtest.agent.MainWrapper$MainThread.run(MainWrapper.java:127) > at java.base/java.lang.Thread.run(Thread.java:833) > > and JInternalFrame test, we had > ----------System.err:(15/939)---------- > FRAME_COLOR Red: 255; Green: 200; Blue: 0 > Pixel color Red: 55; Green: 55; Blue: 55 > java.lang.RuntimeException: Internal frame is not correctly dragged! > at bug8069348.main(bug8069348.java:96) > at > java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native Method) > at > java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:76) > at > java.base/jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:51) > at java.base/java.lang.reflect.Method.invoke(Method.java:569) > at > com.sun.javatest.regtest.agent.MainWrapper$MainThread.run(MainWrapper.java:127) > at java.base/java.lang.Thread.run(Thread.java:833) > > where color is obtained as 55,55,55 instead of required color. The png image > generated at failure also did not reveal anything abnormal apart from > presence of mouse cursor around the same area where the getPixelColor > location is pointing to. And since mouse cursor is also grayish black, it > seems robot was wrongly picking up cursor pixels instead of background color. > Modified tests to use location.x-10, location.y-10 to obtain the pixel color. > It does not hamper the test as the whole area around the location is coloured > with same color in the test so getting pixel color from a bit wide of the > centre will not affect the test. > Modified swing tests are passing in the affected system and also in other > macos systems and platforms. Link in JBS. > > The awt tests that are failing seems to have different root cause and will be > handled separately. This pull request has now been integrated. Changeset: 91607436 Author: Prasanta Sadhukhan <psadhuk...@openjdk.org> URL: https://git.openjdk.java.net/jdk/commit/91607436b79126ccb999deaf38a98209dbfe6ec1 Stats: 126 lines in 5 files changed: 79 ins; 7 del; 40 mod 8276058: Some swing test fails on specific CI macos system Reviewed-by: prr, kizune ------------- PR: https://git.openjdk.java.net/jdk/pull/6140
