Re: Watchpoint on an unloaded shared library(1)

[Emi SUZUKI <emi-suzuki@tjsys.co.jp>]

Hello Joel, 

From: Joel Brobecker <brobecker at adacore.com>
Subject: Re: Watchpoint on an unloaded shared library(1)
Date: Mon, 22 Dec 2008 07:27:58 +0400

>> > > How about replace all the code that's inside
>> > > 
>> > >   if (bpt->type == bp_watchpoint || 
>> > >       bpt->type == bp_hardware_watchpoint ||
>> > >       bpt->type == bp_read_watchpoint ||
>> > >       bpt->type == bp_access_watchpoint)
>> > >     {
>> > >       [re-create a lot of stuff for our watchpoint...]
>> > > 
>> > > by a call to update_watchpoint (bpt, 1)?  Would that work in your case? 
>> > 
>> > It works without merging the missing code each other, as long as we
>> > don't have to care the hardware watchpoint resource count here: if the
>> > user sets other watchpoints while the disabled hardware watchpoints
>> > exist, re-enabling the disabled ones might fail out of the shortage of
>> > resources.  
>> 
>> That's a good point.
> 
> As it turns out, I think we just got lucky thanks to Jan:
> http://www.sourceware.org/ml/gdb-patches/2008-12/msg00363.html.
> 
> update_watchpoint now automagically downgrades hardware watchpoints
> into a software watchpoints if needed. So we can get rid of that part.
> So all we have to do, now is:
> 
>>    if (bpt->type == bp_watchpoint || 
>>        bpt->type == bp_hardware_watchpoint ||
>>        bpt->type == bp_read_watchpoint ||
>>        bpt->type == bp_access_watchpoint)
>>      {
>>        struct gdb_exception e;
>> 
>>        TRY_CATCH (e, RETURN_MASK_ALL)
>>          {
>>            update_watchpoint (...);
>>          }
>>        if (e.reason < 0)
>>          {
>>            exception_fprintf (gdb_strerr, e, "Cannot enable watchpoint %d:",
>>                               b->number);
>>            return;
>>          }

This does work for me, except one duplication of the codes (that you
may have noticed already):

  static void
  do_enable_breakpoint (struct breakpoint *bpt, enum bpdisp disposition)
  {
    int target_resources_ok, other_type_used;
    struct value *mark;
  
    if (bpt->type == bp_hardware_breakpoint)
      {
        int i;
        i = hw_breakpoint_used_count ();
        target_resources_ok = 
          TARGET_CAN_USE_HARDWARE_WATCHPOINT (bp_hardware_breakpoint, 
                                              i + 1, 0);
        if (target_resources_ok == 0)
          error (_("No hardware breakpoint support in the target."));
        else if (target_resources_ok < 0)
          error (_("Hardware breakpoints used exceeds limit."));
      }

That should be done in update_watchpoint.  

I would try to address incapabilities on PPC after those codes are
committed.  Thank you for taking care of it!


I wonder if I should greet for Christmas here...
-- 
Emi SUZUKI / emi-suzuki at tjsys.co.jp