The following patch gets threaded watchpoint support working for the x86.  On 
x86 linux, the dr_status register is thread-specific.  This means that the 
current method which uses the PID to call PTRACE is wrong.  I have changed this 
to use the current lwp for the inferior_ptid.  Corresponding to this, the 
i386_stopped_data_address function switches the inferior_ptid to the trap_ptid. 
  Thus, we can see if we really stopped for a watchpoint or hardware breakpoint.

Because the thread-db.c code changes the trap_ptid into a high-level ptid (pid + 
tid), I had to add a new target vector interface which gives back the lwp for a 
given ptid.  This is used by the low level dr get routine.

With this change plus a change to breakpoint.c, threaded watchpoints work on the 
x86.  Part 2 will be the breakpoint.c change which is not x86-specific.  Part 3 
will be a test case when I figure out how to code it properly.  If you want to 
test it with the two patches applied, just debug gdb.threads/schedlock and issue:

b main
run
watch args[0]
watch args[1]
continue
continue
continue
etc...

Ok?

-- Jeff J.

2004-06-11  Jeff Johnston  <jjohnstn@redhat.com>

         * i386-linux-nat.c (i386_linux_dr_get): Use target_get_lwp
         function to get lwp of current ptid to use in the PTRACE
         call.
         * i386-nat.c: Add comments regarding the DR status register.
         (i386_stopped_data_address): Switch to the trap_ptid before
         getting the dr_status register.
         (i386_stopped_by_hwbp): Ditto.
         * lin-lwp.c (child_wait): Set the trap_ptid when appropriate.
         * target.h (struct target ops): Add new function to_get_lwp.
         (target_get_lwp): New macro.
         * target.c (INHERIT): Inherit to_get_lwp.
         (update_current_target): Default to_get_lwp to ptid_get_lwp.
         (init_dummy_target): Ditto.
         * thread-db.c (thread_db_get_lwp): New function.
         (thread_db_init): Expose thread_db_get_lwp in target vector.