<html>
<head>
<base href="http://llvm.org/bugs/" />
</head>
<body><table border="1" cellspacing="0" cellpadding="8">
<tr>
<th>Bug ID</th>
<td><a class="bz_bug_link
bz_status_NEW "
title="NEW --- - [loop vectorizer] failure to vectorize int to fp conversions for <256 bit vectors"
href="http://llvm.org/bugs/show_bug.cgi?id=21356">21356</a>
</td>
</tr>
<tr>
<th>Summary</th>
<td>[loop vectorizer] failure to vectorize int to fp conversions for <256 bit vectors
</td>
</tr>
<tr>
<th>Product</th>
<td>libraries
</td>
</tr>
<tr>
<th>Version</th>
<td>trunk
</td>
</tr>
<tr>
<th>Hardware</th>
<td>PC
</td>
</tr>
<tr>
<th>OS</th>
<td>All
</td>
</tr>
<tr>
<th>Status</th>
<td>NEW
</td>
</tr>
<tr>
<th>Severity</th>
<td>normal
</td>
</tr>
<tr>
<th>Priority</th>
<td>P
</td>
</tr>
<tr>
<th>Component</th>
<td>Loop Optimizer
</td>
</tr>
<tr>
<th>Assignee</th>
<td>unassignedbugs@nondot.org
</td>
</tr>
<tr>
<th>Reporter</th>
<td>spatel+llvm@rotateright.com
</td>
</tr>
<tr>
<th>CC</th>
<td>llvmbugs@cs.uiuc.edu
</td>
</tr>
<tr>
<th>Classification</th>
<td>Unclassified
</td>
</tr></table>
<p>
<div>
<pre>Using clang/tools built from r220352:
Original C source:
void foo(int *x, float *y) {
int i;
for (i=0; i<1024; i++) {
y[i] = x[i];
}
}
-----------------------------------------------------------------
Reduced IR test case:
$ cat lvbug.ll
target datalayout = "e-m:o-i64:64-f80:128-n8:16:32:64-S128"
target triple = "x86_64-apple-macosx10.9.0"
define void @foo(i32* nocapture readonly %x, float* nocapture %y) {
entry:
br label %for.body
for.body:
%indvars.iv = phi i64 [ 0, %entry ], [ %indvars.iv.next, %for.body ]
%arrayidx = getelementptr inbounds i32* %x, i64 %indvars.iv
%0 = load i32* %arrayidx, align 4
%conv = sitofp i32 %0 to float
%arrayidx2 = getelementptr inbounds float* %y, i64 %indvars.iv
store float %conv, float* %arrayidx2, align 4
%indvars.iv.next = add nuw nsw i64 %indvars.iv, 1
%exitcond = icmp eq i64 %indvars.iv, 1023
br i1 %exitcond, label %for.end, label %for.body
for.end:
ret void
}
----------------------------------------------------------------
$ ./opt -mcpu=core2 -loop-vectorize lvbug.ll -S | grep sitofp
%conv = sitofp i32 %0 to float
But with 256-bit vectors available:
$ ./opt -mcpu=core-avx2 -loop-vectorize lvbug.ll -S | grep sitofp
%104 = sitofp <8 x i32> %wide.load to <8 x float>
%105 = sitofp <8 x i32> %wide.load9 to <8 x float>
%106 = sitofp <8 x i32> %wide.load10 to <8 x float>
%107 = sitofp <8 x i32> %wide.load11 to <8 x float>
%conv = sitofp i32 %189 to float</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>