<html><head><meta http-equiv="Content-Type" content="text/html charset=us-ascii"><meta http-equiv="Content-Type" content="text/html charset=us-ascii"></head><body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space;"><br><div><div>On Jul 28, 2014, at 9:36 , Manuel Klimek <<a href="mailto:klimek@google.com">klimek@google.com</a>> wrote:</div><br class="Apple-interchange-newline"><blockquote type="cite"><div dir="ltr"><div class="gmail_extra"><div class="gmail_quote">On Mon, Jul 28, 2014 at 6:27 PM, Jordan Rose <span dir="ltr"><<a href="mailto:jordan_rose@apple.com" target="_blank">jordan_rose@apple.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div style="word-wrap:break-word"><div class=""><br><div><div>On Jul 28, 2014, at 7:56 , Manuel Klimek <<a href="mailto:klimek@google.com" target="_blank">klimek@google.com</a>> wrote:</div>
<br><blockquote type="cite"><div dir="ltr"><div class="gmail_extra"><div class="gmail_quote">On Mon, Jul 28, 2014 at 4:01 PM, Manuel Klimek <span dir="ltr"><<a href="mailto:klimek@google.com" target="_blank">klimek@google.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr"><div class="gmail_extra"><div class="gmail_quote"><div>On Mon, Jul 28, 2014 at 3:43 PM, Manuel Klimek <span dir="ltr"><<a href="mailto:klimek@google.com" target="_blank">klimek@google.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr">Some more context:<div>It seems like we already need to do that when building the CFG: for the lifetime extended object we must not emit the destructor at the end of the full expression, which we currently do.</div>
</div></blockquote><div><br></div></div><div>And it looks like we're already trying to do that by handing BindToTemporary down. I'm investigating some more...</div></div></div></div></blockquote><div><br></div><div>
Found bug, separate patch sent (although right now I think it's also still wrong).</div><div>I think we need to fix how we handle lifetime extended temporaries before we can really do anything here - the main bug in lifetime extended temporary handling is in connecting the MaterializeTemporaryExpr to the right CXXBindTemporaryExpr (if it exists). Once that problem is solved, we can simple mark the CXXBindTemporary that is lifetime extended once we hit the MaterializeTempraryExpr in the static analyzer.</div>
</div></div></div></blockquote><br></div></div><div>I'd really like the CFG to be correct, not something that we have to reconstruct correctness for in the analyzer. </div></div></blockquote><div><br></div><div>Yes, but it's currently not correct.</div>
<div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div style="word-wrap:break-word"><div>We should be able to statically tell which temporary a MaterializeTemporaryExpr is going to bind to, right? Or rather, we should be able to tell whether or not a given CXXBindTemporaryExpr will be destroyed or bound if it is executed.</div>
</div></blockquote><div><br></div><div>Yes, fully agreed, but first we have to fix how to handle lifetime extended temporaries - the fix *is* getting that information correctly out of the AST.</div><div>I'd rather not mix this patch with a fix for lifetime extended temporaries, and I'm not sure continuing to maintain this patch until after I fixed lifetime extended temps is a good idea.</div>
<div> </div></div></div></div></blockquote><br></div><div>Agreed. I was just objecting to "mark the CXXBindTemporary...once we hit the MaterializeTemporaryExpr in the static analyzer". It seems like we would <i>never</i> add a destructor for that temporary in the first place, so we wouldn't have to mark it as skipped. Or did you mean "when building the CFG"?</div><div><br></div><div>Jordan</div><br></body></html>