On Tuesday 18 March 2008 01:20:41 Daniel Jacobowitz wrote:
> On Sat, Mar 15, 2008 at 01:12:14PM +0300, Vladimir Prus wrote:
> > > This patch is OK, if you will add the missing pieces: a NEWS entry and
> > > documentation for the new command, and a testcase that enables
> > > always-inserted and verifies that it isn't a brick.
> > 
> > I've added NEW and documentation.
> > 
> > I'm still not sure how to test this adequately.
> 
> This version is OK when Eli is happy with the documentation.
> 
> For a testcase, I agree that we need to run the whole testsuite in
> this mode to get full coverage.  But it's useful to have a single
> test that runs in this mode even if the rest of the testsuite
> doesn't, in case someone completely breaks it.  That's why the
> gdb.server tests run for native targets that build gdbserver,
> even when the rest of the testsuite uses "run".
> 
> Could you write a test that uses "set break always-inserted" and does
> something simple like running to a function and then continuing past
> it to another breakpoint?

How about the attached? It depends on my previous patch to introduce
some testsuite helper functions.

> > +	  /* We're allowing a thread to run past a breakpoint it has
> > +             hit, by single-stepping the thread with the breakpoint
> > +             removed.  In which case, we need to single-step only this
> > +             thread, and keep others stopped, as they can miss this
> > +             breakpoint if allowed to run.
> > +
> > +	     The current code actually removes all breakpoints when
> > +	     doing this, not just the one being stepped over, so if we
> > +	     let other threads run, we can actually miss any
> > +	     breakpoint, not just the one at PC.  */
> 
> Extra spacing here?

Yes, I've fixed this.

- Volodya