Some Eclipse Foundation services are deprecated, or will be soon. Please ensure you've read this important communication.
Bug 327906 - Display#wake does not work reliably during lifecycle
Summary: Display#wake does not work reliably during lifecycle
Status: RESOLVED FIXED
Alias: None
Product: RAP
Classification: RT
Component: RWT (show other bugs)
Version: 1.3   Edit
Hardware: All All
: P3 normal (vote)
Target Milestone: 1.4 M3   Edit
Assignee: Project Inbox CLA
QA Contact:
URL:
Whiteboard: sr132
Keywords:
: 317908 (view as bug list)
Depends on:
Blocks: 315418
  Show dependency tree
 
Reported: 2010-10-15 11:24 EDT by Ralf Sternberg CLA
Modified: 2010-10-26 06:58 EDT (History)
1 user (show)

See Also:

Attachments
Test case (1.76 KB, patch)
2010-10-15 11:33 EDT, Ralf Sternberg CLA 		no flags Details | Diff
Patch for 1.3 and HEAD (854 bytes, patch)
2010-10-18 07:08 EDT, Ralf Sternberg CLA 		no flags Details | Diff
View All Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Ralf Sternberg CLA 2010-10-15 11:24:53 EDT 
When Display#wake is called from a background thread while the UIThread is currently running, it will have no effect, even though the UICallBack is active.
This is caused by the implementation of UICallBackManager#sendUICallBack, which does nothing when the UIThread is running. The idea behind this implementation is that when the UIThread finishes ( in notifyUIThreadEnd ), the UICallBack request is answered if the Synchronizer has asyncRunnables in the queue. But in case of wake, there are no runnables.

JFace uses Display#wake in ModalContext ( indirectly called through asyncExec( null ) ), where this bug prevents the ModalContext from finishing correctly (which in turn causes bug 315418).
Comment 1 Ralf Sternberg CLA 2010-10-15 11:33:08 EDT 
Created attachment 180973 [details]
Test case
Comment 2 Ralf Sternberg CLA 2010-10-18 07:08:33 EDT 
Created attachment 181081 [details]
Patch for 1.3 and HEAD

With this patch, Display#wake causes the UICallBack request to unblock immediately. This should solve the problem reliably, but could still be optimized for cases where Display#wake is called multiple times in one request.
Comment 3 Ralf Sternberg CLA 2010-10-18 07:15:59 EDT 
Applied patch to CVS HEAD. We should consider applying it to 1.3 maintenance branch as well.
Comment 4 Ralf Sternberg CLA 2010-10-18 10:38:43 EDT 
*** Bug 317908 has been marked as a duplicate of this bug. ***
Comment 5 Ralf Sternberg CLA 2010-10-26 06:58:12 EDT 
Applied the patch also to the 1.3 maintenance branch.