Re: [8] Review request for 8015730: PIT: On Linux, OGL=true and fbobject=false leads to deadlock or infinite loop

2013-07-03 Thread Anthony Petrov
On 07/03/2013 08:35 PM, Artem Ananiev wrote: src/solaris/native/sun/java2d/opengl/GLXSurfaceData.c 398 XSync(awt_display, False); OK, got it. All other usages of the EXEC_WITH_XERROR_HANDLER() pass TRUE to the RESTORE_XERROR_HANDLER(), and hence call XSync(), too (in XErrorHandlerUtil.j

Re: [8] Review request for 8015730: PIT: On Linux, OGL=true and fbobject=false leads to deadlock or infinite loop

2013-07-03 Thread Artem Ananiev
On 7/3/2013 7:46 PM, Anthony Petrov wrote: On 07/03/2013 06:47 PM, Artem Ananiev wrote: On 7/1/2013 8:50 PM, Anthony Petrov wrote: Thanks for the additional information, Anton. Since this fix simply reverts the behavior in GLXSurfaceData.c back to the pre-8005607 era, it could probably be cons

Re: [8] Review request for 8015730: PIT: On Linux, OGL=true and fbobject=false leads to deadlock or infinite loop

2013-07-03 Thread Anthony Petrov
On 07/03/2013 06:47 PM, Artem Ananiev wrote: On 7/1/2013 8:50 PM, Anthony Petrov wrote: Thanks for the additional information, Anton. Since this fix simply reverts the behavior in GLXSurfaceData.c back to the pre-8005607 era, it could probably be considered a good interim solution for the proble

Re: [8] Review request for 8015730: PIT: On Linux, OGL=true and fbobject=false leads to deadlock or infinite loop

2013-07-03 Thread Anton Litvinov
Hello Artem, Thank you for the review of this fix. No, I am still convinced that the fix reverts XError handling part of the function "Java_sun_java2d_opengl_GLXSurfaceData_initPbuffer" in the file "jdk/src/solaris/native/sun/java2d/opengl/GLXSurfaceData.c" to not taking AWT lock for a call t

apple.laf.useScreenMenuBar does not work with shift modifier and functional keys

2013-07-03 Thread Denis Fokin
Hi AWT team, I recently field a bug http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=8019265 I am not sure whether you had a chance to look into the issue, so I want to share some additional details here. It seems the problem is caused by [NSEvent charactersIgnoringModifiers] method. You c

Re: [8] Review request for 8015730: PIT: On Linux, OGL=true and fbobject=false leads to deadlock or infinite loop

2013-07-03 Thread Artem Ananiev
On 7/1/2013 8:50 PM, Anthony Petrov wrote: Thanks for the additional information, Anton. Since this fix simply reverts the behavior in GLXSurfaceData.c back to the pre-8005607 era, it could probably be considered a good interim solution for the problem. I'd like to hear Artem's opinion on this,

[8] Request for review: JDK-8013611 Modal dialog fails to obtain keyboard focus

2013-07-03 Thread Anton V. Tarasov
Please, review a fix. jira: https://jbs.oracle.com/bugs/browse/JDK-8013611 webrev: http://cr.openjdk.java.net/~ant/JDK-8013611/webrev.0 The fix refines a condition by which a focus request is directed through the branches. Please refer to jira for more details. I've run reg tests on Windows/M