Re: [PATCH] Stop considering hw and sw breakpoints duplicates (Re: [PATCH] breakpoint: Make sure location types match before swapping)

[Pedro Alves <palves@redhat.com>]

On 4/21/20 5:24 PM, Christian Biesinger via Gdb-patches wrote:
> On Sun, Apr 19, 2020 at 1:49 PM Pedro Alves via Gdb-patches
> <gdb-patches@sourceware.org> wrote:
>> From 10d0944768e6ce59861c1522ed48449422a76736 Mon Sep 17 00:00:00 2001
>> From: Pedro Alves <palves@redhat.com>
>> Date: Sun, 19 Apr 2020 18:25:55 +0100
>> Subject: [PATCH] Stop considering hw and sw breakpoints duplicates
>>
>> ---
>>  gdb/breakpoint.c                                   | 269 ++++++++++++++-------
>>  gdb/testsuite/gdb.base/breakpoint-in-ro-region.exp |   2 +-
>>  2 files changed, 183 insertions(+), 88 deletions(-)
>>
>> diff --git a/gdb/breakpoint.c b/gdb/breakpoint.c
>> index e49025461ba..27799e89807 100644
>> --- a/gdb/breakpoint.c
>> +++ b/gdb/breakpoint.c
>> @@ -144,6 +144,10 @@ static void describe_other_breakpoints (struct gdbarch *,
>>  static int watchpoint_locations_match (struct bp_location *loc1,
>>                                        struct bp_location *loc2);
>>
>> +static int breakpoint_locations_match (struct bp_location *loc1,
>> +                                      struct bp_location *loc2,
>> +                                      bool sw_hw_bps_match = false);
>> +
> 
> Should this return a bool?
> 
>>  static int breakpoint_location_address_match (struct bp_location *bl,
>>                                               const struct address_space *aspace,
>>                                               CORE_ADDR addr);
> 
> And this, I guess.

Yeah, they should.  Old, pre-C++ code.
It feels like an orthogonal change though, particularly since I think we
could change all of those two, plus watchpoint_locations_match at the
same time.

Thanks,
Pedro Alves