This patch adjusts the inf-ttrace target to always register the main
thread in GDB's thread table.  Again, I'm using the new thread_change_ptid
function to update inferior_ptid.

( I haven't touched the hpux-thread.c target, as it is still
waiting for someone to rewrite it as a thread_stratum target, and
remove its dependency on deprecated_child_ops.  That is, it's dead broken
code currently.  While we're on to it, does anyone still care for 
HP-UX / ptrace (not ttrace) ? )

I have no means to test this.  Can anybody with HP-UX access check
if I didn't break anything?  This patch only depends on patch 1 of
the series.

-- 
Pedro Alves