This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: [patch 2/3] Implement support for PowerPC BookE ranged watchpoints
- From: Thiago Jung Bauermann <bauerman at br dot ibm dot com>
- To: Pedro Alves <pedro at codesourcery dot com>
- Cc: gdb-patches at sourceware dot org, Jan Kratochvil <jan dot kratochvil at redhat dot com>, Joel Brobecker <brobecker at adacore dot com>, Eli Zaretskii <eliz at gnu dot org>
- Date: Wed, 08 Dec 2010 23:43:59 -0200
- Subject: Re: [patch 2/3] Implement support for PowerPC BookE ranged watchpoints
- References: <1290549100.3164.47.camel@hactar> <201011251731.58135.pedro@codesourcery.com> <1290806122.3009.37.camel@hactar> <201011271747.39053.pedro@codesourcery.com>
On Sat, 2010-11-27 at 17:47 +0000, Pedro Alves wrote:
> On Friday 26 November 2010 21:15:22, Thiago Jung Bauermann wrote:
> > So in your opinion, the watch command should always use two watchpoint
> > registers to set up a ranged watchpoint in BookE ppc? I'm a bit
> > reluctant to use all the watchpoint registers to set up one
> > watchpoint...
>
> Blame the hardware designers, not me. :-)
Well, you know hardware designers. They're transistor counters. :-)
> WDYT? Sounds reasonable? I'd prefer doing these changes
> as first step.
Thank you very much for the detailed plans! I'm very sorry about my
delay answering this. I'm waiting for internal feedback about this
option.
> I haven't looked yet at what kind of masks
> masked watchpoints support, but if they only support masks in the form
> of (binary) 11110000, 1111111100, etc. (no alternating 1s and 0s),
> then no target_insert_watchpoint or remote protocol change is
> required either for those.
Masked watchpoints support arbitrary masks, so they warrant their own
place under the sun. :-) They be considered independently of ranged
watchpoints then.
--
[]'s
Thiago Jung Bauermann
IBM Linux Technology Center