This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: [PATCH 3/4] Support ranged and masked watchpoints
- From: Luis Machado <luisgpm at linux dot vnet dot ibm dot com>
- To: Joel Brobecker <brobecker at adacore dot com>
- Cc: Thiago Jung Bauermann <bauerman at br dot ibm dot com>, gdb-patches at sourceware dot org, Matt Tyrlik <tyrlik at us dot ibm dot com>
- Date: Tue, 12 Jan 2010 10:01:41 -0200
- Subject: Re: [PATCH 3/4] Support ranged and masked watchpoints
- References: <200912232231.06331.bauerman@br.ibm.com> <200912311520.36690.bauerman@br.ibm.com> <20100112105942.GL2007@adacore.com>
- Reply-to: luisgpm at linux dot vnet dot ibm dot com
On Tue, 2010-01-12 at 14:59 +0400, Joel Brobecker wrote:
> > > Adds support for the following types of watchpoints:
> [range watchpoints]
>
> In this instance as well, I would like to see if we could push all
> the logic to the target, by providing all the information the target
> needs in order to make that decision.
>
> One nice upside is that this will allow the target to use both range
> watchpoint and condition hardware acceleration if the target allows it.
> Or if it doesn't, not separating the two types of features with
> separate target watchpoint_insert/remove routines allows the target
> to choose which one makes most sense if a choice needs to be made.
>
> What do you think?
>
The overall idea of pushing all the code that is closer to the target
somewhere more suitable sounds good to me. That way we can keep the
generic portion of GDB clean, which is less painful to maintain.
This would probably depend on the target vector facelift?
Regards,
Luis