> > +@item state
> > +The current state of the task.
> 
> Would it be a good idea to enumerate the possible states here?  Or is
> that common knowledge for anyone who is an Ada programmer?

It's a bit of both. But the list of states is already provided
inside the documentation of the `info tasks' command, so I added
a couple of references to that section of the manual:
  - One next to the part where I mention the `info tasks' command;
  - One inside the description of the `state' field, mentioning
    explicitly that the list of possible states is described there.

> > +@smallexample
> > +-thread-info
> > +^done,tasks=@{nr_rows="3",nr_cols="8",
> 
> You are documenting -ada-task-info, so why the example is for a
> different command?

That's a good catch - copy/paste typo. The output is correct, though.

> > +@item ada-task-info
> > +Indicates the presence of the @code{-ada-task-info} command.
> 
> I would use "availability of" or "support for", rather than "presence
> of".  (Yes, I know the other items use the same language you did; I
> think they all should be changed.)

OK. Fixed as suggested.  I also went ahead and modified the others
(that's the second patch, which can be applied independently of
the first).

[Ada/doco] Document the new -ada-task-info GDB/MI command.
gdb/doc/ChangeLog:

        * gdb.texinfo (GDB/MI Ada Tasking Commands): New node.
        (GDB/MI Miscellaneous Commands): Add `ada-task-info'
        as possible feature returned by the `-list-features'
        command.

[doco] rephrase `presence of' with `support for'.
gdb/doc/ChangeLog:

        * gdb.texinfo (GDB/MI Miscellaneous Commands): Minor
        rephasing.

-- 
Joel