337899 – [debug view][non-stop] Process label is not updated properly

Some Eclipse Foundation services are deprecated, or will be soon. Please ensure you've read this important communication.

Bug 337899 - [debug view][non-stop] Process label is not updated properly

Summary: [debug view][non-stop] Process label is not updated properly

Status:	RESOLVED FIXED

Alias:	None

Product:	CDT
Classification:	Tools
Component:	cdt-debug-dsf-gdb (show other bugs)
Version:	8.0
Hardware:	PC Linux

Importance:	P3 normal (vote)
Target Milestone:	9.1.0
Assignee:	Marc Dumais
QA Contact:	Marc Khouzam

URL:
Whiteboard:
Keywords:

Duplicates (1):	410108 (view as bug list)
Depends on:
Blocks:

Reported:	2011-02-22 15:54 EST by Marc Khouzam
Modified:	2016-08-29 11:20 EDT (History)
CC List:	6 users (show)

See Also:	Gerrit Change Git Commit Gerrit Change Git Commit

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 Marc Khouzam

2011-02-22 15:54:31 EST

When in non-stop mode, if I resume the program without making it stop right away, the thread is shown as running but the process still shows as stopped in the debug view.

If I refresh the DV, the process icon is properly updated.

Sounds similar to Bug 274870

Comment 1 Marc Khouzam

2011-02-22 16:09:24 EST

I think this is because in non-stop, we send IResumed event for the different threads and not the container.  This means that we don't refresh the container in the debug view.

If all threads are running, we should show the container as running.  If one thread or more is suspended we should show the container as suspended.  This is what we see when I refresh the DV but not automatically.

I see two options:
1- when we get an IResumed event in ContainerVMNode, we specify that the container could have changed and must be refreshed
2- when we issue an IResumed event for a thread, we could also issue a IContainerResumed if all threads are now running.

#1 is simpler and safer
#2 is more efficient but potentially more risky as I think we made some assumptions about not having Container events in non-stop

Comment 2 Mario Prieto

2013-08-28 06:42:48 EDT

I think this entry is related to 
https://bugs.eclipse.org/bugs/show_bug.cgi?id=410108

Comment 3 Marc Dumais

2016-08-26 12:59:22 EDT

*** Bug 410108 has been marked as a duplicate of this bug. ***

Comment 4 Eclipse Genie

2016-08-26 13:04:12 EDT

New Gerrit change created: https://git.eclipse.org/r/79876

Comment 5 Marc Dumais

2016-08-26 13:08:52 EDT

(In reply to Marc Khouzam from comment #1)
> I see two options:
> 1- when we get an IResumed event in ContainerVMNode, we specify that the
> container could have changed and must be refreshed
> 2- when we issue an IResumed event for a thread, we could also issue a
> IContainerResumed if all threads are now running.

I went with solution #1 - a node label refresh is relatively light, compared to everything that goes on in CDT when a thread resumes.

Comment 6 Eclipse Genie

2016-08-29 07:23:32 EDT

Gerrit change https://git.eclipse.org/r/79876 was merged to [master].
Commit: http://git.eclipse.org/c/cdt/org.eclipse.cdt.git/commit/?id=a90d9dfbf8cef53d276484e8c53fe971559767c8

Comment 7 Eclipse Genie

2016-08-29 10:02:20 EDT

New Gerrit change created: https://git.eclipse.org/r/79942

Comment 8 Eclipse Genie

2016-08-29 11:02:56 EDT

Gerrit change https://git.eclipse.org/r/79942 was merged to [cdt_9_1].
Commit: http://git.eclipse.org/c/cdt/org.eclipse.cdt.git/commit/?id=0b6ca32767127249f7ce558fcf8db3525a214a01

Comment 9 Marc Dumais

2016-08-29 11:04:22 EDT

committed to master and 9_1 branches