<div dir="ltr"><br><div class="gmail_extra"><br><div class="gmail_quote">On Tue, Mar 3, 2015 at 11:33 AM, Rafael Espíndola <span dir="ltr"><<a href="mailto:rafael.espindola@gmail.com" target="_blank">rafael.espindola@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><span class="">>> In this case the ComdatGV would be the function, which is not private, no?<br>
><br>
><br>
> I was referring to the case where the function has internal linkage (and we<br>
> have EmitUniquedSection enabled for data sections). The global would be<br>
> private and it wouldn't be possible to stick it in a COMDAT.<br>
<br>
</span>I don't follow, if the function is internal, ComdatGV will be the<br>
function and have internal linkage, no?<br></blockquote><div><br></div><div>The function will be internal but why would the ComdatGV be the function? I don't see how the function would be in a COMDAT in that case.</div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<br>
In any case, if there is a CodeGen limitation for COFF, using internal<br>
for COFF and private for MachO/ELF for now is probably fine.<br>
<br>
I will try to open a bug to track "private in comdat for COFF".<br></blockquote><div><br></div><div>Thanks!</div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<br>
Cheers,<br>
Rafael<br>
</blockquote></div><br></div></div>