<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
</head>
<body text="#000000" bgcolor="#FFFFFF">
<p>Oh, yes, it is a typo, thanks for catching it!<br>
</p>
<pre class="moz-signature" cols="72">-------------
Best regards,
Alexey Bataev</pre>
<div class="moz-cite-prefix">26.07.2018 13:59, Jonas Hahnfeld via
cfe-commits пишет:<br>
</div>
<blockquote type="cite"
cite="mid:%3C42cceb247ebd6935380c852ac30a7878@hahnjo.de%3E">Hi
Alexey,
<br>
<br>
On 2018-07-26 19:53, Alexey Bataev via cfe-commits wrote:
<br>
<blockquote type="cite">Author: abataev
<br>
Date: Thu Jul 26 10:53:45 2018
<br>
New Revision: 338049
<br>
<br>
URL: <a class="moz-txt-link-freetext" href="http://llvm.org/viewvc/llvm-project?rev=338049&view=rev">http://llvm.org/viewvc/llvm-project?rev=338049&view=rev</a>
<br>
Log:
<br>
[OPENMP] What's new for OpenMP in clang.
<br>
<br>
Updated ReleaseNotes + Status of the OpenMP support in clang.
<br>
<br>
Modified:
<br>
cfe/trunk/docs/OpenMPSupport.rst
<br>
cfe/trunk/docs/ReleaseNotes.rst
<br>
<br>
Modified: cfe/trunk/docs/OpenMPSupport.rst
<br>
[...]
<br>
<br>
Modified: cfe/trunk/docs/ReleaseNotes.rst
<br>
URL:
<br>
<a class="moz-txt-link-freetext" href="http://llvm.org/viewvc/llvm-project/cfe/trunk/docs/ReleaseNotes.rst?rev=338049&r1=338048&r2=338049&view=diff">http://llvm.org/viewvc/llvm-project/cfe/trunk/docs/ReleaseNotes.rst?rev=338049&r1=338048&r2=338049&view=diff</a>
<br>
==============================================================================
<br>
--- cfe/trunk/docs/ReleaseNotes.rst (original)
<br>
+++ cfe/trunk/docs/ReleaseNotes.rst Thu Jul 26 10:53:45 2018
<br>
@@ -216,7 +216,21 @@ OpenCL C Language Changes in Clang
<br>
OpenMP Support in Clang
<br>
----------------------------------
<br>
<br>
-- ...
<br>
+- Clang gained basic support for OpenMP 4.5 offloading for
NVPTX target.
<br>
+ To compile your program for NVPTX target use the following
options:
<br>
+ `-fopenmp -fopenmp-targets=nvptx64-nvidia-cuda` for 64 bit
platforms or
<br>
+ `-fopenmp -fopenmp-targets=nvptx-nvidia-cuda` for 32 bit
platform.
<br>
+
<br>
+- Passing options to the OpenMP device offloading toolchain can
be done using
<br>
+ the `-Xopenmp-target=<triple> -opt=val` flag. In this
way the `-opt=val`
<br>
+ option will be forwarded to the respective OpenMP device
offloading toolchain
<br>
+ described by the triple. For example passing the compute
capability to
<br>
+ the OpenMP NVPTX offloading toolchain can be done as follows:
<br>
+ `-Xopenmp-target=nvptx62-nvidia-cuda -march=sm_60`.
<br>
</blockquote>
<br>
Is that a typo and should say "nvptx64"?
<br>
<br>
</blockquote>
<br>
</body>
</html>