<html>
<head>
<base href="https://bugs.llvm.org/">
</head>
<body><span class="vcard"><a class="email" href="mailto:spatel+llvm@rotateright.com" title="Sanjay Patel <spatel+llvm@rotateright.com>"> <span class="fn">Sanjay Patel</span></a>
</span> changed
<a class="bz_bug_link
bz_status_RESOLVED bz_closed"
title="RESOLVED FIXED - [x86] large constant offsets used for memory accesses in loop"
href="https://bugs.llvm.org/show_bug.cgi?id=35681">bug 35681</a>
<br>
<table border="1" cellspacing="0" cellpadding="8">
<tr>
<th>What</th>
<th>Removed</th>
<th>Added</th>
</tr>
<tr>
<td style="text-align:right;">Fixed By Commit(s)</td>
<td>
</td>
<td>324289
</td>
</tr>
<tr>
<td style="text-align:right;">Resolution</td>
<td>---
</td>
<td>FIXED
</td>
</tr>
<tr>
<td style="text-align:right;">Status</td>
<td>NEW
</td>
<td>RESOLVED
</td>
</tr>
<tr>
<td style="text-align:right;">See Also</td>
<td>
</td>
<td>https://bugs.llvm.org/show_bug.cgi?id=36180
</td>
</tr></table>
<p>
<div>
<b><a class="bz_bug_link
bz_status_RESOLVED bz_closed"
title="RESOLVED FIXED - [x86] large constant offsets used for memory accesses in loop"
href="https://bugs.llvm.org/show_bug.cgi?id=35681#c15">Comment # 15</a>
on <a class="bz_bug_link
bz_status_RESOLVED bz_closed"
title="RESOLVED FIXED - [x86] large constant offsets used for memory accesses in loop"
href="https://bugs.llvm.org/show_bug.cgi?id=35681">bug 35681</a>
from <span class="vcard"><a class="email" href="mailto:spatel+llvm@rotateright.com" title="Sanjay Patel <spatel+llvm@rotateright.com>"> <span class="fn">Sanjay Patel</span></a>
</span></b>
<pre>Should be fixed with:
<a href="https://reviews.llvm.org/rL324289">https://reviews.llvm.org/rL324289</a>
xorl %eax, %eax
.p2align 4, 0x90
LBB0_1:
movupd (%rsi,%rax,8), %xmm0
movupd (%rdi,%rax,8), %xmm1
maxpd %xmm1, %xmm0
movupd %xmm0, (%rdi,%rax,8)
addq $2, %rax
cmpq $65536, %rax ## imm = 0x10000
jne LBB0_1
retq
-----------------------------------------------------------------------------
Note that there is a potentially much bigger problem for some Intel uarch with
AVX code and indexed addressing tracked by <a class="bz_bug_link
bz_status_NEW "
title="NEW - [x86] scalar FP code runs ~15% slower on Haswell when compiled with -mavx"
href="show_bug.cgi?id=36180">bug 36180</a>.</pre>
</div>
</p>
<hr>
<span>You are receiving this mail because:</span>
<ul>
<li>You are on the CC list for the bug.</li>
</ul>
</body>
</html>