Ticket #837 (closed bug: fixed)

Opened 2 years ago

Last modified 1 month ago

[app_server] crash in Desktop::Cursor()

Reported by: diver Assigned to: stippi
Priority: blocker Milestone: R1/alpha1
Component: Servers/app_server Version: R1 development
Cc: Platform: All

Description

On real hw, while plaing around with SoundPlay, dunno how to reproduce yet.

Segment violation

Desktop::Cursor()
Desktop::SetCursor()
ServerApp::SetCurrentCursor()
WindowsLayer::MouseMoved()
MouseFilter::Filter()
EventDispatcher::_EventLoop()
EventDispatcher::_eventLooper()

Change History

09/25/06 07:08:26 changed by wkornewald

  • version set to R1 development.
  • milestone set to R1.

10/06/06 08:55:39 changed by axeld

  • priority changed from normal to blocker.

10/02/07 05:51:11 changed by axeld

  • milestone changed from R1 to R1/alpha.

03/29/08 03:14:19 changed by nielx

This ticket is a little bit vague for a nomination as blocker of the alpha 1 release. Diver, can you try to find a consistent method for reproduction?

03/29/08 04:43:49 changed by diver

I didnt' sow this crash for year now. I think its fixed.

(follow-up: ↓ 8 ) 03/29/08 05:37:36 changed by axeld

It hasn't been fixed yet, the code is still problematic. It only shows rarely, though (I remember running into it about two month ago the last time).
Maybe it would be okay to move it to R1 instead, I dunno.

03/29/08 12:55:19 changed by stippi

I didn't see it either for a _long_ time, but yesterday I got a crash in HWInterface::_DrawCursor(), which I believe is originating from the same problem. I should write a test app which constantly changes the cursor shape, I think this is what triggers it, together with something else in the cursor code.

(in reply to: ↑ 6 ) 03/30/08 04:02:35 changed by nielx

  • owner changed from axeld to stippi.

Replying to axeld:

It hasn't been fixed yet, the code is still problematic. It only shows rarely, though (I remember running into it about two month ago the last time).
Maybe it would be okay to move it to R1 instead, I dunno.

Let's keep it here until we are at a later stage in the alpha 1 process, and then reevaluate. Anyway, assigning to stippi because he should write a test app :-)

04/02/08 06:14:31 changed by stippi

  • status changed from new to closed.
  • resolution set to fixed.

Hopefully fixed in r24741. The bug was/is very hard to reproduce, so I cannot verify that the problem I saw was the indeed cause. Please reopen if you encounter it again.