Some Eclipse Foundation services are deprecated, or will be soon. Please ensure you've read this important communication.
Bug 15212 - Object shows up when I launch
Summary: Object shows up when I launch
Status: VERIFIED FIXED
Alias: None
Product: JDT
Classification: Eclipse Project
Component: Debug (show other bugs)
Version: 2.0   Edit
Hardware: PC Windows 2000
: P2 normal (vote)
Target Milestone: 2.0 F3   Edit
Assignee: Darin Wright CLA
QA Contact:
URL:
Whiteboard:
Keywords:
: 16292 17866 18113 (view as bug list)
Depends on:
Blocks:
 
Reported: 2002-05-03 12:52 EDT by Darin Wright CLA
Modified: 2002-06-07 10:09 EDT (History)
3 users (show)

See Also:

Attachments
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 Darin Wright CLA 2002-05-03 12:52:14 EDT 
Often when I start a debug session, the source for java.lang.Object shows up in 
the editor pane. For some reason, the debugger is openning this editor.
Comment 1 Darin Wright CLA 2002-05-14 22:06:09 EDT 
Cannot make this occurr anymore with fix for 12740. Closing.
Comment 2 Jared Burns CLA 2002-05-17 14:33:56 EDT 
*** Bug 16292 has been marked as a duplicate of this bug. ***
Comment 3 Darin Wright CLA 2002-05-18 10:31:18 EDT 
This is happenning with our latest plug-ins
Comment 4 Darin Wright CLA 2002-05-18 11:09:56 EDT 
This appears to be a "luck of the draw" scenario. When a VM starts, its threads 
can be suspended. The launch view attempts to expand and display stack frames 
for suspended threads. In this case a thread is suspended in Object.wait 
(transiently). The thread then resumes, and all is well. 

Not sure if there is a fix for this.
Comment 5 Jared Burns CLA 2002-05-22 09:46:20 EDT 
I've started seeing this in build 20020521 on Linux-GTK. I never saw it before.
Comment 6 Jared Burns CLA 2002-05-22 10:12:10 EDT 
The problem is that we're asking the VM for threads and adding them to the 
view before we receive ThreadStart notifications.

We could probably fix this problem by only adding threads to the view for 
which we've received ThreadStart notifications.
Comment 7 Darin Wright CLA 2002-05-22 10:21:05 EDT 
I don't believe that works in a remote attach - there is no guarentee that you 
will receive a thread start for already "started" threads.
Comment 8 Darin Wright CLA 2002-05-27 09:51:55 EDT 
*** Bug 17866 has been marked as a duplicate of this bug. ***
Comment 9 Jared Burns CLA 2002-05-30 15:59:26 EDT 
The JDWP spec actually says that the threads returned by the VM shouldn't 
include threads which haven't been started yet:
"Threads that have not yet been started and threads that have completed their 
execution are not included in the returned list."

This bug could be caused by the VM(s?) not honoring this contract.
Comment 10 Jared Burns CLA 2002-06-03 10:20:59 EDT 
According to specification, all threads that the VM returns at startup are 
running.

Updated JDIDebugTarget#initializeState() to initialize our JDIThreads in the 
running state. This is safe because of the facts that the threads must be 
running when we connect to the VM and we ask for the threads before we install 
any breakpoints.

This change does not affect threads which are created later.
Comment 11 Darin Wright CLA 2002-06-03 16:33:46 EDT 
*** Bug 18113 has been marked as a duplicate of this bug. ***
Comment 12 Darin Wright CLA 2002-06-05 14:28:16 EDT 
Fixed. Needs F3 verification.
Comment 13 Darin Wright CLA 2002-06-07 10:09:00 EDT 
Verified.