<html xmlns:v="urn:schemas-microsoft-com:vml" xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:w="urn:schemas-microsoft-com:office:word" xmlns:m="http://schemas.microsoft.com/office/2004/12/omml" xmlns="http://www.w3.org/TR/REC-html40">
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
<meta name="Generator" content="Microsoft Word 15 (filtered medium)">
<style><!--
/* Font Definitions */
@font-face
{font-family:"Cambria Math";
panose-1:2 4 5 3 5 4 6 3 2 4;}
@font-face
{font-family:Calibri;
panose-1:2 15 5 2 2 2 4 3 2 4;}
/* Style Definitions */
p.MsoNormal, li.MsoNormal, div.MsoNormal
{margin:0cm;
margin-bottom:.0001pt;
font-size:11.0pt;
font-family:"Calibri",sans-serif;}
a:link, span.MsoHyperlink
{mso-style-priority:99;
color:blue;
text-decoration:underline;}
a:visited, span.MsoHyperlinkFollowed
{mso-style-priority:99;
color:purple;
text-decoration:underline;}
p.msonormal0, li.msonormal0, div.msonormal0
{mso-style-name:msonormal;
mso-margin-top-alt:auto;
margin-right:0cm;
mso-margin-bottom-alt:auto;
margin-left:0cm;
font-size:11.0pt;
font-family:"Calibri",sans-serif;}
span.EmailStyle18
{mso-style-type:personal-reply;
font-family:"Calibri",sans-serif;
color:windowtext;}
.MsoChpDefault
{mso-style-type:export-only;}
@page WordSection1
{size:612.0pt 792.0pt;
margin:72.0pt 72.0pt 72.0pt 72.0pt;}
div.WordSection1
{page:WordSection1;}
--></style><!--[if gte mso 9]><xml>
<o:shapedefaults v:ext="edit" spidmax="1026" />
</xml><![endif]--><!--[if gte mso 9]><xml>
<o:shapelayout v:ext="edit">
<o:idmap v:ext="edit" data="1" />
</o:shapelayout></xml><![endif]-->
</head>
<body lang="EN-US" link="blue" vlink="purple">
<div class="WordSection1">
<p class="MsoNormal">Thanks Sanjay for confirming that this seem to be broken.<o:p></o:p></p>
<p class="MsoNormal">I’ll check with Mikael to make sure we write a new PR (probably have to wait until tomorrow).<o:p></o:p></p>
<p class="MsoNormal"><o:p> </o:p></p>
<p class="MsoNormal">PS. This kind of confirms that DAGTypeLegalizer::PromoteIntRes_BITCAST is doing the wrong thing for big-endian as well, see <a href="https://bugs.llvm.org/show_bug.cgi?id=44135">https://bugs.llvm.org/show_bug.cgi?id=44135</a> , so maybe
we can set that PR to “confirmed” (and then move forward with making a proper patch based on the workaround presented in that PR).<o:p></o:p></p>
<p class="MsoNormal"><o:p> </o:p></p>
<p class="MsoNormal">/Björn<o:p></o:p></p>
<p class="MsoNormal"><o:p> </o:p></p>
<div style="border:none;border-left:solid blue 1.5pt;padding:0cm 0cm 0cm 4.0pt">
<div>
<div style="border:none;border-top:solid #E1E1E1 1.0pt;padding:3.0pt 0cm 0cm 0cm">
<p class="MsoNormal"><b>From:</b> llvm-dev <llvm-dev-bounces@lists.llvm.org> <b>On Behalf Of
</b>Sanjay Patel via llvm-dev<br>
<b>Sent:</b> den 28 november 2019 16:55<br>
<b>To:</b> Mikael Holmén <mikael.holmen@ericsson.com>; Chris Lattner <clattner@nondot.org><br>
<b>Cc:</b> llvm-dev@lists.llvm.org<br>
<b>Subject:</b> Re: [llvm-dev] Instcombine and bitcast of vector. Wrong CHECKs in cast.ll, miscompile in instcombine?<o:p></o:p></p>
</div>
</div>
<p class="MsoNormal"><o:p> </o:p></p>
<div>
<div>
<p class="MsoNormal">Looks broken to me - we need to consider big/little-endian datalayout when bitcasting to/from vectors.<o:p></o:p></p>
</div>
<div>
<p class="MsoNormal">We should have some documentation for this in the LangRef, but I don't see anything currently.<o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal">The transform in question was added here:<o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><a href="https://protect2.fireeye.com/v1/url?k=e6ff41bf-ba756d27-e6ff0124-0cc47ad93eae-0e34bc4531ded90b&q=1&e=7537b523-36d8-43aa-9f2e-412e89ef8c5f&u=https%3A%2F%2Freviews.llvm.org%2FrL103354">https://reviews.llvm.org/rL103354</a><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal">You can find other vector bitcast transforms that (hopefully correctly...) account for the datalayout difference for vector elements.<o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal">Example:<o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><a href="https://protect2.fireeye.com/v1/url?k=4a660b74-16ec27ec-4a664bef-0cc47ad93eae-98051a8097706eba&q=1&e=7537b523-36d8-43aa-9f2e-412e89ef8c5f&u=https%3A%2F%2Fgithub.com%2Fllvm%2Fllvm-project%2Fblob%2Fmaster%2Fllvm%2Ftest%2FTransforms%2FInstCombine%2Fbitcast-bigendian.ll%23L10">https://github.com/llvm/llvm-project/blob/master/llvm/test/Transforms/InstCombine/bitcast-bigendian.ll#L10</a><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><a href="https://protect2.fireeye.com/v1/url?k=7e024c7d-228860e5-7e020ce6-0cc47ad93eae-c3ed7bf3f1423e80&q=1&e=7537b523-36d8-43aa-9f2e-412e89ef8c5f&u=https%3A%2F%2Fgithub.com%2Fllvm%2Fllvm-project%2Fblob%2Fmaster%2Fllvm%2Ftest%2FTransforms%2FInstCombine%2Fbitcast.ll%23L290">https://github.com/llvm/llvm-project/blob/master/llvm/test/Transforms/InstCombine/bitcast.ll#L290</a><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal">So we need something like this:<o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><a href="https://protect2.fireeye.com/v1/url?k=eb400e4d-b7ca22d5-eb404ed6-0cc47ad93eae-33441214035703e7&q=1&e=7537b523-36d8-43aa-9f2e-412e89ef8c5f&u=https%3A%2F%2Fgithub.com%2Fllvm%2Fllvm-project%2Fblob%2Fmaster%2Fllvm%2Flib%2FTransforms%2FInstCombine%2FInstCombineCasts.cpp%23L487">https://github.com/llvm/llvm-project/blob/master/llvm/lib/Transforms/InstCombine/InstCombineCasts.cpp#L487</a><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
</div>
<p class="MsoNormal"><o:p> </o:p></p>
<div>
<div>
<p class="MsoNormal">On Thu, Nov 28, 2019 at 9:12 AM Mikael Holmén via llvm-dev <<a href="mailto:llvm-dev@lists.llvm.org" target="_blank">llvm-dev@lists.llvm.org</a>> wrote:<o:p></o:p></p>
</div>
<blockquote style="border:none;border-left:solid #CCCCCC 1.0pt;padding:0cm 0cm 0cm 6.0pt;margin-left:4.8pt;margin-right:0cm">
<p class="MsoNormal">Hi,<br>
<br>
In<br>
llvm/test/Transforms/InstCombine/cast.ll<br>
there is a test like this:<br>
<br>
target datalayout = "E-p:64:64:64-p1:32:32:32-p2:64:64:64-p3:64:64:64-<br>
a0:0:8-f32:32:32-f64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-<br>
v64:64:64-v128:128:128-n8:16:32:64"<br>
<br>
[...]<br>
<br>
define <3 x i32> @test60(<4 x i32> %call4) {<br>
; CHECK-LABEL: @test60(<br>
; CHECK-NEXT: [[P10:%.*]] = shufflevector <4 x i32> [[CALL4:%.*]],<br>
<4 x i32> undef, <3 x i32> <i32 0, i32 1, i32 2><br>
; CHECK-NEXT: ret <3 x i32> [[P10]]<br>
;<br>
%p11 = bitcast <4 x i32> %call4 to i128<br>
%p9 = trunc i128 %p11 to i96<br>
%p10 = bitcast i96 %p9 to <3 x i32><br>
ret <3 x i32> %p10<br>
<br>
}<br>
<br>
If we assume the input vector is e.g. <1, 2, 3, 4> then I assume %p11<br>
would be the (hex) value 1234, %p9 would be the 234 and %p10 would then<br>
be the vector <2, 3, 4>.<br>
<br>
Am I right, or am I missing something here? Note that the datalayout<br>
says we're using big endian.<br>
<br>
But the CHECK-NEXT checks that the result is made up of the elements at<br>
index 0, 1 and 2 from the input vector, which would be <1, 2, 3>.<br>
<br>
So, broken testcase or am I missing something?<br>
<br>
Thanks,<br>
Mikael<br>
<br>
_______________________________________________<br>
LLVM Developers mailing list<br>
<a href="mailto:llvm-dev@lists.llvm.org" target="_blank">llvm-dev@lists.llvm.org</a><br>
<a href="https://protect2.fireeye.com/v1/url?k=616c8bcf-3de6a757-616ccb54-0cc47ad93eae-ec99014218a008a2&q=1&e=7537b523-36d8-43aa-9f2e-412e89ef8c5f&u=https%3A%2F%2Flists.llvm.org%2Fcgi-bin%2Fmailman%2Flistinfo%2Fllvm-dev" target="_blank">https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev</a><o:p></o:p></p>
</blockquote>
</div>
</div>
</div>
</body>
</html>