Community
Participate
Working Groups
Install latest El Capitan Beta 3 (10.11 - 15A216g) and run any Eclipse based environment such as Eclipse, MyEclipseIDE etc. The application will start slowly and then hang/freeze with the UI being unresponsive.
Can you run jstack on the hanging application and attach the dump here?
Created attachment 255084 [details] jstack dump of hanging eclipse I started Eclipse (eclipse-jee-mars-R-macosx-cocoa-x86_64) from the command line via ./eclipse. Eclipse will then start and show the workspace selection window with a spinning color-wheel. Eventually the colour wheel will disappear, but the workspace selection window will remain unresponsive and OS X shows Eclipse as "not responding".
Created attachment 255085 [details] jstack of MyEclipseIDE installer FWIW - I can also reproduce this on the MyEclipseIDE installer (http://www.genuitec.com/products/myeclipse/download/get/?118) I am not sure if holds any relevance, but if you compare both jstack files, both seem to hang in NSRunLoop: Eclipse: "main" #1 prio=6 os_prio=31 tid=0x000000010302b800 nid=0x50f runnable [0x00007fff5fbfd000] java.lang.Thread.State: RUNNABLE at org.eclipse.swt.internal.cocoa.OS.objc_msgSend_bool(Native Method) at org.eclipse.swt.internal.cocoa.NSRunLoop.runMode(NSRunLoop.java:42) at org.eclipse.swt.widgets.Display.sleep(Display.java:4740) at org.eclipse.jface.window.Window.runEventLoop(Window.java:828) at org.eclipse.jface.window.Window.open(Window.java:803) at org.eclipse.ui.internal.ide.ChooseWorkspaceDialog.prompt(ChooseWorkspaceDialog.java:93) at org.eclipse.ui.internal.ide.application.IDEApplication.promptForWorkspace(IDEApplication.java:343) at org.eclipse.ui.internal.ide.application.IDEApplication.checkInstanceLocation(IDEApplication.java:262) vs MyEclipseIDEInstaller: "main" #1 prio=5 os_prio=31 tid=0x0000000103041000 nid=0x50f runnable [0x00007fff5fbfd000] java.lang.Thread.State: RUNNABLE at org.eclipse.swt.internal.cocoa.OS.objc_msgSend_bool(Native Method) at org.eclipse.swt.internal.cocoa.NSRunLoop.runMode(NSRunLoop.java:42) at org.eclipse.swt.widgets.Display.sleep(Display.java:4610) at com.genuitec.pulse.installer.container.ui.base.BaseInstallerGUIContainer.waitForModule(Unknown Source) at com.genuitec.pulse.installer.module.ui.message.WelcomeModule.run(Unknown Source) at com.genuitec.pulse.installer.machine.internal.SequenceMachineImpl.runSequence(Unknown Source) at com.genuitec.pulse.installer.machine.internal.SequenceMachineImpl.run(Unknown Source) at com.genuitec.pulse.installer.application.CommonInstallerApplication.start(Unknown Source)
Thanks for trying out Eclipse on 10.11. Moving to SWT.
*** Bug 472300 has been marked as a duplicate of this bug. ***
Even the simplest shell snippet: public static void main(String[] args) { final Display display = new Display(); final Shell shell = new Shell(display); shell.setSize(300, 200); shell.open(); while (!shell.isDisposed()) { if (!display.readAndDispatch()) { display.sleep(); } } display.dispose(); } hangs at java.lang.Thread.State: RUNNABLE at org.eclipse.swt.internal.cocoa.OS.objc_msgSend_bool(Native Method) at org.eclipse.swt.internal.cocoa.NSRunLoop.runMode(NSRunLoop.java:42) at org.eclipse.swt.widgets.Display.sleep(Display.java:4740)
*** Bug 472432 has been marked as a duplicate of this bug. ***
(In reply to Thomas Singer from comment #6) > Even the simplest shell snippet: > > public static void main(String[] args) { > final Display display = new Display(); > > final Shell shell = new Shell(display); > > shell.setSize(300, 200); > shell.open(); > > while (!shell.isDisposed()) { > if (!display.readAndDispatch()) { > display.sleep(); > } > } > > display.dispose(); > } > > hangs at > > java.lang.Thread.State: RUNNABLE > at org.eclipse.swt.internal.cocoa.OS.objc_msgSend_bool(Native Method) > at org.eclipse.swt.internal.cocoa.NSRunLoop.runMode(NSRunLoop.java:42) > at org.eclipse.swt.widgets.Display.sleep(Display.java:4740) I currently don't have the 10.11 beta to try out. Can you pls check if there is any exception thrown when the snippet is launching?
10.11 is now in public beta. However, I have not confirmed the first public beta release matches beta 3. It may or may not manifest this issue.
(In reply to Lakshmi Shanmugam from comment #8) > Can you pls check if there is any exception thrown when the snippet is > launching? There is no exception thrown. According to the thread-dump it just hangs in OS.objc_msgSend_bool.
Same issue for me with the official beta
Executing from console I have this log: /System/Library/Frameworks/OpenCL.framework/Versions/A/lib/clang/3.2/include/cl_kernel.h:4842:21: note: candidate function not viable: requires 3 arguments, but 4 were provided float4 __OVERLOAD__ clamp(float4 x, float4 minval, float4 maxval); ^ /System/Library/Frameworks/OpenCL.framework/Versions/A/lib/clang/3.2/include/cl_kernel.h:4841:21: note: candidate function not viable: requires 3 arguments, but 4 were provided float3 __OVERLOAD__ clamp(float3 x, float3 minval, float3 maxval); ^ /System/Library/Frameworks/OpenCL.framework/Versions/A/lib/clang/3.2/include/cl_kernel.h:4840:21: note: candidate function not viable: requires 3 arguments, but 4 were provided float2 __OVERLOAD__ clamp(float2 x, float2 minval, float2 maxval); ^ /System/Library/Frameworks/OpenCL.framework/Versions/A/lib/clang/3.2/include/cl_kernel.h:4839:20: note: candidate function not viable: requires 3 arguments, but 4 were provided float __OVERLOAD__ clamp(float x, float minval, float maxval); ^ Unable to compile kernel. We should debug this. Here more information from other user that have the same problem. Someone says the problem is the new core UI access. (https://forums.developer.apple.com/thread/8776) Best Regards
I tried with the 10.11 official beta and it has the same problem. This seems to be a problem with the OS and is mentioned as a known issue in the release notes: "Some versions of Eclipse may hang during use". --> https://developer.apple.com/library/prerelease/mac/releasenotes/General/rn-osx-10.11/
Question is, if this is something which should/can be fixed within Eclipse or if Apple will actually acknowledge/fix. I posted the same bug to Apple with id=21741973 on 9th July and have yet to receive acknowledgement/comment from Apple.
And beta 4 fixes the issue
Fixed in beta 4.
*** Bug 473333 has been marked as a duplicate of this bug. ***
Confirmed that it is fixed in public beta2.
I just wanted to point out for the record that the resolve status of this bug is misleading as the cause of this is very likely to be an Eclipse bug, namely the same as bug 477979, where SWT relied on undocumented behavior. It is just that Apple worked around it by attempting to identify SWT applications and then revert to the old incorrect behavior. As noted in bug 477979, that identification attempt may fail.
