Bugzilla – Full Text Bug Listing
| Summary: | [CommonNavigator] Test failure (disposed font) in org.eclipse.ui.tests.navigator_win32.win32.x86 | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Product: | [Eclipse Project] Platform | Reporter: | Boris Bokowski <bokowski> | ||||||
| Component: | UI | Assignee: | Francis Upton IV <francisu> | ||||||
| Status: | VERIFIED FIXED | QA Contact: | |||||||
| Severity: | normal | ||||||||
| Priority: | P3 | CC: | eclipse, francisu, mdelder, pwebster | ||||||
| Version: | 3.4 | ||||||||
| Target Milestone: | 3.5 M6 | ||||||||
| Hardware: | PC | ||||||||
| OS: | Windows XP | ||||||||
| Whiteboard: | |||||||||
| Bug Depends on: | 226191 | ||||||||
| Bug Blocks: | |||||||||
| Attachments: |
|
||||||||
|
Description
Boris Bokowski
Created attachment 96965 [details]
ugly patch
This should fix the problem for the next build, but it's a hack. The problem is that the common navigator calls labelProvider.dispose() before doing a refresh. The label provider's font is still being used by one of the tree items (from the paint callback), but at that time, it is already disposed. A better fix would be to dispose of the label provider after performing the refresh, but I will have to see if this is something we can do without API changes.
Created attachment 96966 [details]
mylyn/context/zip
Ugh. I cannot commit this change because my CVS permissions are screwed up. See also bug 226191 comment 4. Paul or Kim, if one of you sees this in time for the integration build, could you please try to release the patch and either move the tag for this one file or do a new build submission? Thanks! (In reply to comment #3) > Ugh. I cannot commit this change because my CVS permissions are screwed up. See > also bug 226191 comment 4. > > Paul or Kim, if one of you sees this in time for the integration build, could > you please try to release the patch and either move the tag for this one file > or do a new build submission? Thanks! > I will version before I come in to work. Done. This hack can probably be backed out with the fix to bug 265317, but this needs to be tested on Windows to make sure it will work. (In reply to comment #6) > This hack can probably be backed out with the fix to bug 265317, but this needs > to be tested on Windows to make sure it will work. > Removed the hack and verified passes on Windows. Verified all CNF tests passed in I20090427-1800 |