| Breakpoint on AssertionError exception [message #276389] |
Mon, 22 November 2004 20:24  |
Eclipse User |
|
|
|
Originally posted by: cortell.cox.net
This is a bit strange. I'm using assert() in the plugin I'm developing. I
pass -ea to the runtime-workbench VM and I have a breakpoint set on
AssertionError. When an assertion fails in my plugin code, the regular
Eclipse session apparently reacts--but not as expected. The expected
behavior is that the debugger shows the thread suspended at the assert()
statement. Eclipse does no such thing. The thread where the assert occurs
continues to be reported in the running state. If I explicitly suspenpend
the thread at that point, then the debugger shows that the execution is
suspended at the assert() statement. I said that the debugger non-the-less
reacts, and by that I mean that the debugger seems to want "attention"
when the assertion fails. I'm doing all this on Windows and what I mean is
that the debugger's entry in the Windows task bar starts blinking--just as
it does when a breakpoint is hit. So, clearly, it's reacting to the assert
failure, but it apparently fails to report that the thread is suspended.
Now here's where things get really strange. If I force a rebuild of a
basic "Hello World" project in my workspace then retry the
workspace-runtime debug session, the debugger begins reacting as
expected--showing the suspended thread when the assert is hit. Stanger
yet: if I instead (first) rebuild the plugin itself, it does not correct
the behavior. Rebuilding the non-plugin Java project is the only thing
that seems to make the debugger work right.
Like I said, it's bizarre. Not sure anyone will have an explanation or
solution. I've relaunched Eclipse many times and tried tons of variations
to nail down the behavior. There's nothing intermittent about this. The
behavior is as described 100% of the time. Just figured I'd post this and
see. Maybe this is a manifestation of a known issue. I'm using Eclipse 3.0
with J2SDK 1.4.1_01 on Windows 2000.
|
|
|
| Re: Breakpoint on AssertionError exception [message #276440 is a reply to message #276389] |
Tue, 23 November 2004 17:47  |
 Darin Wright
Messages: 454
Registered: July 2009 |
Senior Member |
|
|
Please file a bug with an example test case.
Thanks,
Darin
"John Cortell" <cortell@cox.net> wrote in message
news:cnthun$sj6$1@www.eclipse.org...
> This is a bit strange. I'm using assert() in the plugin I'm developing. I
> pass -ea to the runtime-workbench VM and I have a breakpoint set on
> AssertionError. When an assertion fails in my plugin code, the regular
> Eclipse session apparently reacts--but not as expected. The expected
> behavior is that the debugger shows the thread suspended at the assert()
> statement. Eclipse does no such thing. The thread where the assert occurs
> continues to be reported in the running state. If I explicitly suspenpend
> the thread at that point, then the debugger shows that the execution is
> suspended at the assert() statement. I said that the debugger non-the-less
> reacts, and by that I mean that the debugger seems to want "attention"
> when the assertion fails. I'm doing all this on Windows and what I mean is
> that the debugger's entry in the Windows task bar starts blinking--just as
> it does when a breakpoint is hit. So, clearly, it's reacting to the assert
> failure, but it apparently fails to report that the thread is suspended.
>
> Now here's where things get really strange. If I force a rebuild of a
> basic "Hello World" project in my workspace then retry the
> workspace-runtime debug session, the debugger begins reacting as
> expected--showing the suspended thread when the assert is hit. Stanger
> yet: if I instead (first) rebuild the plugin itself, it does not correct
> the behavior. Rebuilding the non-plugin Java project is the only thing
> that seems to make the debugger work right.
>
> Like I said, it's bizarre. Not sure anyone will have an explanation or
> solution. I've relaunched Eclipse many times and tried tons of variations
> to nail down the behavior. There's nothing intermittent about this. The
> behavior is as described 100% of the time. Just figured I'd post this and
> see. Maybe this is a manifestation of a known issue. I'm using Eclipse 3.0
> with J2SDK 1.4.1_01 on Windows 2000.
>
|
|
|
Powered by
FUDForum. Page generated in 0.03324 seconds
] 
Search
Help
Register
Login
Home
