<div dir="ltr">On Tue, Jul 16, 2013 at 1:10 PM, Greg Clayton <span dir="ltr"><<a href="mailto:gclayton@apple.com" target="_blank">gclayton@apple.com</a>></span> wrote:<br><div class="gmail_extra"><div class="gmail_quote">
<blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex"><div class="im"><br>
On Jul 15, 2013, at 5:24 PM, Michael Sartain <<a href="mailto:mikesart@gmail.com">mikesart@gmail.com</a>> wrote:<br>
<br>
</div><div class="im">> On Mon, Jul 15, 2013 at 4:15 PM, Greg Clayton <<a href="mailto:gclayton@apple.com">gclayton@apple.com</a>> wrote:<br>
> You might think about storing the breakpoint ID instead of a shared pointer to the breakpoint.<br>
><br>
><br>
> Can I ask why? There should only be one of these I believe.<br>
<br>
</div>It really shouldn't matter really. Just to make sure no strong reference to a breakpoint stays around longer than required.<br>
<div class="im"><br>
> I think ThreadPlanStepRange.cpp does something similar with m_next_branch_bp_sp. Does that make sense to switch to IDs as well?<br>
<br>
</div>I can see the thread plans doing this because they will be accessing the breakpoint as soon as you stop. For the dynamic loaders, we tend to create the breakpoint and then the callback will be called automatically, so we aren't digging up the breakpoint each time we hit it.<br>
<br>
This doesn't need to be done, this is just out the MacOSX dynamic loader does it.</blockquote><div><br></div><div>Got it - I just wanted to understand the reasoning. I've switched over, re-ran the tests, and posted the new patch here. Please let me know if this looks ok when you get a chance and I'll commit.</div>
<div><br></div><div> <a href="http://llvm-reviews.chandlerc.com/D1145">http://llvm-reviews.chandlerc.com/D1145</a></div><div><br></div><div>Thanks for info Greg.</div><div> -Mike</div></div></div></div>