6681 failures on Windows since I20260114-0210

[iloveeclipse]

See

• https://download.eclipse.org/eclipse/downloads/drops4/I20260115-1800/tests.html
• https://download.eclipse.org/eclipse/downloads/drops4/I20260115-1800/testresults/html/org.eclipse.swt.tests_ep439I-unit-win32-x86_64-java21_win32.win32.x86_64_21.html

Most of problems due "no more handles"

[HeikoKlare]

I am quite sure the failures are caused by:

• Execute tests with non-default autoscaling settings in I-Build #2967

The execution of the additional test suite with a different autoscaling configuration does not succeed (runs into a timeout) and probably leaves behind a bad state:

  [java] DumpStackTracesTimer almost reached timeout '7200000'.
     [java] totalMemory:            268435456
     [java] freeMemory (before GC): 111295632
     [java] freeMemory (after GC):  237283400
     [java] Thread dump 0 at 2026-01-16 01:47:13 +0000:
 
...

 [java] Java Result: -1
     [java] Timeout: killed the sub-process

collect-results:
[junitreport] the file C:\Users\genie.releng\workspace\AutomatedTests\ep439I-unit-win32-x86_64-java21\workarea\I20260115-1800\eclipse-testing\test-eclipse\eclipse\org.eclipse.swt.tests.junit.AllNonBrowserTests_AutoscaleOsNonDefaults.xml is empty.
[junitreport] This can be caused by the test JVM exiting unexpectedly

I cannot reproduce any issues with the configuration of that test suite locally. Might be specific to how the Ant-based test execution or the Jenkins environment. I will try to figure out.
As last resort I would revert the PR with the additional test suite. We can then see if changing the SWT defaults (as proposed by the follow PR) lead to the same issue or if this is just caused by basically every SWT test being executed twice:

• [Win32] Enable monitor-specific scaling by default #2955

[HeikoKlare]

Taking a deeper look, I see that the the first test bundle with UI (the Ant UI one) already fails with "no more handles":

org.eclipse.swt.SWTError: No more handles
	at org.eclipse.swt.SWT.error(SWT.java:4980)
	at org.eclipse.swt.SWT.error(SWT.java:4865)
	at org.eclipse.swt.SWT.error(SWT.java:4836)
	at org.eclipse.swt.widgets.Widget.error(Widget.java:510)
	at org.eclipse.swt.widgets.Control.createHandle(Control.java:690)

This is nothing that can be caused by:

• Execute tests with non-default autoscaling settings in I-Build #2967

I have double checked the other PR and don't see how this can change any behavior if the system property, for which handling was added, is not set:

• [Win32] Allow to customize DPI awareness #2970

To me, this more and more looks like another occurrence of:

• https://gitlab.eclipse.org/eclipsefdn/helpdesk/-/issues/6527
• https://gitlab.eclipse.org/eclipsefdn/helpdesk/-/issues/5214

I will create a helpdesk issue.
I have created a helpdesk issue: https://gitlab.eclipse.org/eclipsefdn/helpdesk/-/issues/7047

[HeikoKlare]

Just saw that this issue is not even correct. The failures already started to happen with: I20260114-0210

[iloveeclipse]

Just saw that this issue is not even correct. The failures already started to happen with: I20260114-0210

My bad, Windows tests were late yesterday morning so I didn't noticed failures

[HeikoKlare]

My bad, Windows tests were late yesterday morning so I didn't noticed failures

Same for me, so I feel just as "guilty" 🙂

[HeikoKlare]

Restart resolved the issue; reexecution worked fine again: https://ci.eclipse.org/releng/job/AutomatedTests/job/ep439I-unit-win32-x86_64-java21/56/

[HannesWell]

Thanks for taking care of it.