This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
[patch 3.1/3] bpstat_what removal - addon gdb_assert
Hi Pedro,
On Thu, 24 Jun 2010 16:44:23 +0200, Jan Kratochvil wrote:
> attaching follow-up on your patch still using BPSTAT_WHAT_* symbols.
this is a gdb_assert followup with the goal of making the code both more clear
and more safe against future changes. I do not mind much about this patch.
No regressions on {x86_64,x86_64-m32,i686}-fedora13-linux-gnu.
Thanks,
Jan
2010-06-24 Jan Kratochvil <jan.kratochvil@redhat.com>
* breakpoint.c (bpstat_what): Pre-set this_action to -1. Set
this_action to BPSTAT_WHAT_KEEP_CHECKING in cases it has been left
unchanged. New gdb_assert.
--- a/gdb/breakpoint.c
+++ b/gdb/breakpoint.c
@@ -4234,7 +4234,7 @@ bpstat_what (bpstat bs)
{
/* Extract this BS's action. After processing each BS, we check
if its action overrides all we've seem so far. */
- enum bpstat_what_main_action this_action = BPSTAT_WHAT_KEEP_CHECKING;
+ enum bpstat_what_main_action this_action = -1;
enum bptype bptype;
if (bs->breakpoint_at == NULL)
@@ -4251,6 +4251,7 @@ bpstat_what (bpstat bs)
switch (bptype)
{
case bp_none:
+ this_action = BPSTAT_WHAT_KEEP_CHECKING;
break;
case bp_breakpoint:
case bp_hardware_breakpoint:
@@ -4281,6 +4282,7 @@ bpstat_what (bpstat bs)
{
/* There was a watchpoint, but we're not stopping.
This requires no further action. */
+ this_action = BPSTAT_WHAT_KEEP_CHECKING;
}
break;
case bp_longjmp:
@@ -4317,6 +4319,7 @@ bpstat_what (bpstat bs)
{
/* There was a catchpoint, but we're not stopping.
This requires no further action. */
+ this_action = BPSTAT_WHAT_KEEP_CHECKING;
}
break;
case bp_shlib_event:
@@ -4359,6 +4362,7 @@ bpstat_what (bpstat bs)
_("bpstat_what: unhandled bptype %d"), (int) bptype);
}
+ gdb_assert (this_action != -1);
retval.main_action = max (retval.main_action, this_action);
}