<div dir="ltr"><div class="gmail_extra"><div class="gmail_quote">On Tue, Jul 15, 2014 at 4:10 AM, Janusz Sobczak <span dir="ltr"><<a href="mailto:janusz.sobczak@mobica.com" target="_blank">janusz.sobczak@mobica.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="">>>! In D4492#5, @rsmith wrote:<br>
> This makes Clang's constant expression evaluator and its IR generation disagree about the value of `__builtin_constant_p`.<br>
</div>How would you address this issue? My concern is that there are cases, where we need to fold to a constant early (i.e. global variable initialisation) and other cases where we can emit IR code and make the decision about 'constantness' at a later stage.<br>
<div class=""><br>
><br>
> What's the motivation behind this change?<br>
</div>Matching GCC behaviour as closely as we can.</blockquote><div><br></div><div>That seems like a means to an end, rather than a goal in itself. Also, we currently match GCC's -O0 behavior pretty closely. </div></div>
</div></div>