<html>
<head>
<base href="https://bugs.webkit.org/" />
</head>
<body>
<p>
<div>
<b><a class="bz_bug_link
bz_status_NEW "
title="NEW - Air::IRC doesn't need to have a special-case for uncolored Tmps since they all get colored"
href="https://bugs.webkit.org/show_bug.cgi?id=163548#c1">Comment # 1</a>
on <a class="bz_bug_link
bz_status_NEW "
title="NEW - Air::IRC doesn't need to have a special-case for uncolored Tmps since they all get colored"
href="https://bugs.webkit.org/show_bug.cgi?id=163548">bug 163548</a>
from <span class="vcard"><a class="email" href="mailto:fpizlo@apple.com" title="Filip Pizlo <fpizlo@apple.com>"> <span class="fn">Filip Pizlo</span></a>
</span></b>
<pre>Here's a horrible bug. Say you have a Patchpoint that late-clobbers all regs followed by a Patchpoint that wants a scratch. If you do this, the scratch is uncolorable.
We have this bug right now because we don't have callee-save FPRs on x86. So if you have a CCall or JS call Patchpoint followed by a Patchpoint that wants FP scratches, then the FP scratch will not get colored and it will be left as %xmm0 at the end of IRC. That may be wrong, if the return value of the CCall or Patchpoint just before was %xmm0, for example.</pre>
</div>
</p>
<hr>
<span>You are receiving this mail because:</span>
<ul>
<li>You are the assignee for the bug.</li>
</ul>
</body>
</html>