Date: Wed, 29 Jun 2005 14:11:53 -0400 From: Stephan Uphoff <ups@tree.com> To: Marcel Moolenaar <marcel@xcllnt.net> Cc: cvs-src@freebsd.org, src-committers@freebsd.org, Bruce Evans <bde@zeta.org.au>, cvs-all@freebsd.org Subject: Re: cvs commit: src/sys/i386/i386 trap.c src/sys/amd64/amd64 trap.c Message-ID: <1120068712.77984.18622.camel@palm> In-Reply-To: <b1b2f68bf241cca18df281a369574619@xcllnt.net> References: <200506252214.j5PMEgip000380@repoman.freebsd.org> <7c65ef6b909f86fa7f5a8aa041773a72@xcllnt.net> <20050627050618.W34733@delplex.bde.org> <1120052829.77984.17793.camel@palm> <b1b2f68bf241cca18df281a369574619@xcllnt.net>
next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, 2005-06-29 at 11:42, Marcel Moolenaar wrote:
> On Jun 29, 2005, at 6:47 AM, Stephan Uphoff wrote:
>
> > this is just a quick fix to get basic debugging capabilities back for
> > some common environments. I plan to migrate parts or all of kdb_trap
> > back to MD code to deal with SMP race conditions.
>
> That is not a good idea. The overall behaviour of entering the
> debugger in inherently MI. The MD specifics are in the details,
> which require nothing more than some MD callback functions to
> fill in the blanks or create the right abstraction.
> Race conditions are not MD phenomena either, but may require MD
> techniques to prevent them. Hence, to fix race conditions you
> don't have to degenerate MI code to MD code, provided you have
> proper MD callback functions to fill in the blanks or create
> the right abstractions.
Don't panic !
The last thing I want to do is to totally dismantle the current kdb_trap
or sprinkle MI code all over the different architecture directories.
The console stuff definitely belongs in the MI part.
However for readability I would rather have:
kdb_trap_md() {
md code;
....
if (really_enter_debugger)
kdb_trap_mi();
....
md code;
}
instead of:
kdb_trap_mi() {
really_enter_debugger = md_debugger_enter_code();
if (!really_enter_debugger)
return
.. stuff ..
md_debugger_exit_code();
}
I am all for MI code and callback functions.
However kdb_trap is only called from MD code.
Guaranteeing a certain execution environment for the MI code is all I
care about. There is no abstraction gained by using MI callbacks on
entering/exiting the MD kdb trap function.
This being said I am not religious about it and can use callbacks if you
feel strongly about this. I just think callbacks would be a bit silly in
this context.
Stephan
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?1120068712.77984.18622.camel>