<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
</head>
<body text="#000000" bgcolor="#FFFFFF">
<p><br>
</p>
<div class="moz-cite-prefix">On 1/15/19 6:07 AM, Zheng CZ Chen via llvm-dev wrote:<br>
</div>
<blockquote type="cite" cite="mid:OF0B3D1FF6.102A7389-ON48258383.003EA927-48258383.0042A0F2@notes.na.collabserv.com">
<p><font size="2">Hi,</font><br>
<br>
<font size="2">There are some compilers with a aggressive optimization which restricts function pointer parameters. Let's say opt restrict_args. When restrict_args is turned on, compiler will treat all function pointer parameters as restrict one.</font><br>
</p>
</blockquote>
<p><br>
</p>
<p>I certainly understand the use case, in a general sense. In my experience, these options are used with (fairly old) pre-C99 code bases (and specifically C, not C++), which follow something akin to a one-function-per-source-file model and which can't be modified
(e.g., for licensing reasons). Using these options are certainly considered bad practice, and they only apply to certain legacy code bases. Does this match your experience and expected usage?<br>
</p>
<p>In an engineering sense, this seems like a trivial feature to support. I don't object to supporting it, but if we do, we probably want to:</p>
<p> 1. Restrict it's application to C (e.g., it should be an error to use with C++, OpenCL, CUDA, and any other languages that Clang supports).</p>
<p> 2. When used with C99 or later language standards, the use of this flag generates a warning on each function definition with a fixit hint showing where the restrict keyword should be placed (we can then, optionally of course, use these fixits to automatically
upgrade code where possible using our corresponding infrastructure). This warning should have a separate flag, and is disabled by default for pre-C99 standard modes, and enabled by default otherwise, but can be toggled independently.<br>
</p>
<p> -Hal<br>
</p>
<blockquote type="cite" cite="mid:OF0B3D1FF6.102A7389-ON48258383.003EA927-48258383.0042A0F2@notes.na.collabserv.com">
<p><br>
<i><font size="2">int foo(int * a) </font></i><font size="2">+ restrict_args opt </font>
<br>
<br>
<font size="2">equals to:</font><br>
<br>
<i><font size="2">int foo(int * restrict a)</font></i><br>
<br>
<br>
<font size="2">Here is a complete example:</font><br>
<font size="2">source code:</font><br>
<font size="2">extern int num;</font><br>
<i><font size="2">int foo(int * a)</font></i><br>
<i><font size="2">{</font></i><br>
<i><font size="2">(*a) = 10;</font></i><br>
<i><font size="2">num++;</font></i><br>
<i><font size="2">(*a)++;</font></i><br>
<i><font size="2"></font></i><br>
<i><font size="2">return *a; </font></i><br>
<i><font size="2">}</font></i><br>
<br>
<font size="2">Using IBM xlc compiler with option -qrestrict at -O2, we get result:</font><br>
<br>
<font size="2">0000000000000000 <foo>:</font><br>
<font size="2">0: 00 00 4c 3c addis r2,r12,0</font><br>
<font size="2">4: 00 00 42 38 addi r2,r2,0</font><br>
<font size="2">8: 00 00 a2 3c addis r5,r2,0</font><br>
<font size="2">c: 00 00 a5 e8 ld r5,0(r5)</font><br>
<font size="2">10: 0b 00 00 38 li r0,11</font><br>
<font size="2">14: 00 00 03 90 stw r0,0(r3)</font><br>
<font size="2">18: 00 00 85 80 lwz r4,0(r5)</font><br>
<font size="2"></font><b><font size="2">1c: 0b 00 60 38 li r3,11 ------>since we confirm num will not change the content where pointer to, compiler can directly return 11.</font></b><br>
<font size="2">20: 01 00 04 38 addi r0,r4,1</font><br>
<font size="2">24: 00 00 05 90 stw r0,0(r5)</font><br>
<font size="2">28: 20 00 80 4e blr</font><br>
<br>
<font size="2">Seems clang does not have such optimization. And I don't find similar option in gcc either.</font><br>
<br>
<font size="2">Is it possible to add this optimization into clang?</font><br>
<br>
<font size="2">Thanks.</font><br>
<br>
<font size="2">BRS//</font><br>
<font size="2">Chen Zheng</font><br>
<font size="2">Power Compiler Backend Developer</font><br>
<br>
</p>
<br>
<fieldset class="mimeAttachmentHeader"></fieldset>
<pre class="moz-quote-pre" wrap="">_______________________________________________
LLVM Developers mailing list
<a class="moz-txt-link-abbreviated" href="mailto:llvm-dev@lists.llvm.org">llvm-dev@lists.llvm.org</a>
<a class="moz-txt-link-freetext" href="http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev">http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev</a>
</pre>
</blockquote>
<pre class="moz-signature" cols="72">--
Hal Finkel
Lead, Compiler Technology and Programming Languages
Leadership Computing Facility
Argonne National Laboratory</pre>
</body>
</html>