<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
A key point here is that not all targets will have a "generic"
address space. This is a new concept which is bring introduced and
needs to be defined carefully.<br>
<br>
Philip<br>
<br>
<div class="moz-cite-prefix">On 03/31/2014 02:08 PM, Jingyue Wu
wrote:<br>
</div>
<blockquote
cite="mid:CAMROOrHfoRnvVXn20CRck5SVkJMGZ1h6n_hokMg3GfgdqCadiw@mail.gmail.com"
type="cite">
<div dir="ltr">Hi Matt,
<div><br>
</div>
<div>I actually think it should go to lib/Transforms/InstCombine
when it gets generalized. Also, <span
style="font-family:arial,sans-serif;font-size:13px">TTI</span> would
need to expose which address space is generic and whether such
optimization would benefit at all.
<div>
<br>
</div>
<div>Jingyue</div>
</div>
</div>
<div class="gmail_extra"><br>
<br>
<div class="gmail_quote">On Mon, Mar 31, 2014 at 2:00 PM, Matt
Arsenault <span dir="ltr"><<a moz-do-not-send="true"
href="mailto:Matthew.Arsenault@amd.com" target="_blank">Matthew.Arsenault@amd.com</a>></span>
wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0
.8ex;border-left:1px #ccc solid;padding-left:1ex"><br>
This looks like it doesn't handle vectors of pointers,
although that's probably not very important.<br>
<br>
I don't see anything tying this to NVPTX except the use of
0 for generic. Where should this go when generalized? Put it
in lib/CodeGen and add getGenericAddressSpace() to TTI?<br>
<br>
<br>
================<br>
Comment at:
lib/Target/NVPTX/NVPTXFavorNonGenericAddrSpaces.cpp:67-72<br>
@@ +66,8 @@<br>
+<br>
+ Value *Src = Cast->getOperand(0);<br>
+ PointerType *SrcTy =
cast<PointerType>(Src->getType());<br>
+ PointerType *DestTy =
cast<PointerType>(Cast->getType());<br>
+ // We cannot remove the addrspacecast if the element type
changes.<br>
+ if (SrcTy->getElementType() !=
DestTy->getElementType())<br>
+ return false;<br>
+<br>
----------------<br>
If the element type changes, you can still do it, but you'll
need to do the bitcast separately. I had a patch a while ago
(<a moz-do-not-send="true"
href="http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20131118/195642.html"
target="_blank">http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20131118/195642.html</a>)
to canonicalize to splitting the bitcast and addrspacecast
into separate operations. It might be better to handle it
that way than trying to figure it out here.<br>
<br>
<br>
<a moz-do-not-send="true"
href="http://llvm-reviews.chandlerc.com/D3235"
target="_blank">http://llvm-reviews.chandlerc.com/D3235</a><br>
</blockquote>
</div>
<br>
</div>
<br>
<fieldset class="mimeAttachmentHeader"></fieldset>
<br>
<pre wrap="">_______________________________________________
llvm-commits mailing list
<a class="moz-txt-link-abbreviated" href="mailto:llvm-commits@cs.uiuc.edu">llvm-commits@cs.uiuc.edu</a>
<a class="moz-txt-link-freetext" href="http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits">http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits</a>
</pre>
</blockquote>
<br>
</body>
</html>