[Webkit-unassigned] [Bug 155371] build fails with memory model cannot be stronger than success memory model for ‘__atomic_compare_exchange’

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=155371

--- Comment #2 from Guillaume Emont <guijemont at igalia.com> ---
Created attachment 282039
  --> https://bugs.webkit.org/attachment.cgi?id=282039&action=review
Safe workaround, but probably suboptimal.

After a quick inquiry, it seems to me that there is an issue with gcc and/or libstdc++. The relevan code in libstdc++ doesn't look suspicious though, so I'd lean towards an issue with gcc's -Os under some specific conditions that we meet, though I could be wrong. I have tried to make a simple test case that reproduces that to understand better what's up but have not succeeded so far. In the meantime, this seems to work as a workaround (it compiles and looks safe, though locks might get slower than they could because of this).

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.webkit.org/pipermail/webkit-unassigned/attachments/20160625/0cad9fd2/attachment-0001.html>