Re: [PATCH]: Make Linux use the new unified x86 watchpoint support

[Michael Snyder <msnyder@cygnus.com>]

Eli Zaretskii wrote:
> 
> > Date: Tue, 27 Mar 2001 08:54:14 -0800
> > From: Michael Snyder <msnyder@redhat.com>
> >
> > > >    This possibility has been discussed back in November, but the
> > > >    conclusion was that it's not a good idea.  I don't remember the
> > > >    details, but the reasons had something to do with threads and how
> > > >    the register cache is used in conjunction with threads.  (I can dig
> > > >    out the URLs of the relevant messages, if you want to read them.)
> > > >
> > > > I suggested doing this, but several people objected to exposing the
> > > > debug registers in this way.  Threads have nothing to do with it
> > >
> > > I'm quite sure they did.  IIRC, the issue was that debug registers are
> > > global, whereas normal registers are per thread.
> >
> > Mmmmm... that's probably true.  However the effect would just be
> > that all threads appear to have the same values for debug regs.
> > If you change them in one, they'll change in all.
> 
> As far as I understood, GDB's regcache cannot do such magic.  It
> assumes that all the registers are per thread, so if you change them
> in one thread, the other threads' registers remain unchanged in the
> cache.

Currently there is only one cache (not one per thread).
When you change threads, the cache is flushed.  However, 
there are plans for a per-thread cache, so your concern
is not ill-placed.


> In other words, this kind of global registers is (or was when we
> discussed it) incompatible with how regcache works.  That is why I was
> advised not to introduce the debug registers into the regcache.

They could be pseudo-registers, which means the debugger would
have the opportunity to handle them specially.