Hi Elena,

Thanks for taking notes!

On Wed, 2007-02-14 at 11:26 -0500, Elena Zannoni wrote:
> Tim: stop the world or stop individual threads? Right now it's stop the 
> world.
> But underlying mechanism to do stop each thread is there.

What you can do at all times is add an Instruction TaskObserver.

     * Interface used to notify that a Task has executed a single
     * instruction. <code>updateExecuted</code> is called as soon as
     * the Instruction observer is added to the Task. And whenever the
     * Task starts running again (isn't blocked or suspended) it will
     * be called on each instruction being executed.
     * <p>
     * This TaskObserver can also be used for executing code that
     * needs the Task to be (temporarily) blocked or suspended as soon
     * as possible. <code>updateExecuted()</code> will be called as
     * soon as this observer has been properly added, and at that time
     * the Task is suspended to make it possible to inspect the Task
     * state. If no other action is request, the method can then just
     * delete the observer from the Task again.

> volatile attribute can make a variable stick around even with optimization.

And I only added volatile to make sure they wouldn't get optimized away.
Filed bug report #4048 for this. (Does the fhpd have a tracker bug?)

> Help <command> is totally broken

New bug #4050 This would be really useful!

> compatibility with gdb command set. it's needed, maybe a set of aliases.

New bug #4049

> 2. Request to open up the RH only conf call to all. For general status
>    of the project, who works on what, progress reports, and
>    dependencies/issue with other components (kernel, gcc, libunwind,
>    etc) RH has things that cannot be discussed openly, will need to
>    figure out how to avoid having 3 meetings.

Note that "cannot be discussed openly" is more "not very frysk related".
I would vote for less phone talk and more discussion on list by email.
Unless there are very good minutes made and posted (you make great
minutes btw!) there is big chance of things falling through the cracks.

If we discuss some things more on the list (just simple things like
'heay, has anybody recently tried a 2.6.20 kernel? Does it pass our
frysk-import tests?') that would be really good to have an automatic
record for everybody, even if you cannot make a call.

Cheers,

Mark