Re: [rfc, rfa/doc] Multi-threaded watchpoint improvements

[Eli Zaretskii <eliz@gnu.org>]

> From: Andreas Schwab <schwab@suse.de>
> Date: Wed, 23 Apr 2008 13:15:47 +0200
> 
> Here's a patch.  OK?
> 
> Andreas.
> 
> 2008-04-23  Andreas Schwab  <schwab@suse.de>
> 
> 	* target.h (struct target_ops): Add
> 	to_watchpoint_addr_within_range.
> 	(target_watchpoint_addr_within_range): New function.
> 	* target.c (update_current_target): Inherit
> 	to_watchpoint_addr_within_range, defaulting to
> 	default_watchpoint_addr_within_range.
> 	(default_watchpoint_addr_within_range): New function.
> 	(debug_to_watchpoint_addr_within_range): New function.
> 	(setup_target_debug): Set to_watchpoint_addr_within_range.
> 	* ppc-linux-nat.c (ppc_linux_watchpoint_addr_within_range):
> 	New function.
> 	(_initialize_ppc_linux_nat): Set to_watchpoint_addr_within_range.
> 	* breakpoint.c (watchpoints_triggered): Use
> 	target_watchpoint_addr_within_range.
> 
> doc/:
> 	* gdbint.texinfo (Algorithms): Describe
> 	target_watchpoint_addr_within_range.

The patch to gdbint.texinfo is approved, with one comment:

> +Check whether @var{addr} (as returned by target_stopped_data_address)

target_stopped_data_address should have the @code markup, since it's a
C symbol.

> +                              This only needs to be provided if the
> +granularity of a watchpoint is greater than one byte

Strictly speaking, I think it's needed only if watchpoint's
granularity is greater than the granularity of data addresses, but
that's nitpicking.