<html>
<head>
<meta http-equiv="Content-Type" content="text/html;
charset=windows-1252">
</head>
<body text="#000000" bgcolor="#FFFFFF">
<div class="moz-cite-prefix">On 1/12/2018 1:02 AM, Sam Parker via
llvm-dev wrote:<br>
</div>
<blockquote type="cite"
cite="mid:AM4PR0801MB147643A16AA9DA8F2FA0DE9885170@AM4PR0801MB1476.eurprd08.prod.outlook.com">
<meta http-equiv="Content-Type" content="text/html;
charset=windows-1252">
<style type="text/css" style="display:none;"><!-- P {margin-top:0;margin-bottom:0;} --></style>
<div id="divtagdefaultwrapper" dir="ltr" style="font-size: 12pt;
color: rgb(0, 0, 0); font-family: Calibri, Helvetica,
sans-serif, EmojiFont, "Apple Color Emoji",
"Segoe UI Emoji", NotoColorEmoji, "Segoe UI
Symbol", "Android Emoji", EmojiSymbols;">
<p style="margin-top:0; margin-bottom:0">Hi Tim and all,</p>
<p style="margin-top:0; margin-bottom:0"><br>
</p>
<p style="margin-top:0; margin-bottom:0">A recent commit to the
type legalizer has highlighted AArch64's use of v1 data types,
could you shine some light on why they've been declared as
legal? And any particularly awkward barriers that prevent
them from being removed as legal?</p>
</div>
</blockquote>
<br>
v1i64 specifically is used to indicate that a value should be stored
in a SIMD register (as opposed to an integer register, like we use
for i64). This allows instruction selection to select instructions
which produce and use i64 values in SIMD registers, like loads,
stores, and i64<->f64 conversions.<br>
<br>
This isn't the only possible approach, but it seems to work
reasonably well as-is.<br>
<br>
-Eli<br>
<p><br>
</p>
<pre class="moz-signature" cols="72">--
Employee of Qualcomm Innovation Center, Inc.
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, a Linux Foundation Collaborative Project</pre>
</body>
</html>