Why does Eclipse not stop at breakpoint in class with debug? [message #509531] |
Fri, 22 January 2010 12:25  |
Eclipse User |
|
|
|
Using Eclipse 3.5.1 and JDK 1.6.
I use the debugger a lot. I step through my application code, I step through code in open-source frameworks that I use. I usually connect to a remote JVM on the same box, running WebLogic. This works perfectly fine, almost all of the time.
Today, for some reason, I found a scenario where it's refusing to stop at certain breakpoints.
My application uses OpenJPA 1.2.2. At application startup it does some metadata analysis and generation. Because of some problem that we don't understand yet, it throws an exception at one point. It shows the line number, and I can view that line in the OpenJPA source and it appears to correspond to the source. I set a breakpoint on that line. I set a breakpoint on the method containing the line. Neither of those breakpoints are hit when I run my test case, even though I'm certain it's getting there (the exception stack trace confirms that). I verified through multiple avenues that the loaded classes have "debug" on.
This code is executed at startup of the server and application, but I know I've stepped through breakpoints in framework code in this situation before.
It seems like there's some threads issue here, but I'm not sure what it might be.
|
|
|
Re: Why does Eclipse not stop at breakpoint in class with debug? [message #509541 is a reply to message #509531] |
Fri, 22 January 2010 13:20   |
Eclipse User |
|
|
|
David M. Karr wrote:
> Using Eclipse 3.5.1 and JDK 1.6.
>
> I use the debugger a lot. I step through my application code, I step
> through code in open-source frameworks that I use. I usually connect to
> a remote JVM on the same box, running WebLogic. This works perfectly
> fine, almost all of the time.
>
> Today, for some reason, I found a scenario where it's refusing to stop
> at certain breakpoints.
>
> My application uses OpenJPA 1.2.2. At application startup it does some
> metadata analysis and generation. Because of some problem that we don't
> understand yet, it throws an exception at one point. It shows the line
> number, and I can view that line in the OpenJPA source and it appears to
> correspond to the source. I set a breakpoint on that line. I set a
> breakpoint on the method containing the line. Neither of those
> breakpoints are hit when I run my test case, even though I'm certain
> it's getting there (the exception stack trace confirms that). I verified
> through multiple avenues that the loaded classes have "debug" on.
>
> This code is executed at startup of the server and application, but I
> know I've stepped through breakpoints in framework code in this
> situation before.
>
> It seems like there's some threads issue here, but I'm not sure what it
> might be.
I doubt it's to do with threads; more likely it's a classpath issue, i.e.
Eclipse is associating a different source file with that code than the one you
put the breakpoint in. If you open an editor on the code containing the
breakpoint, and then you step in the debugger until you get to the same code,
does it open another (different) copy of the file?
|
|
|
|
|
|
|
Powered by
FUDForum. Page generated in 0.03858 seconds