Commit | Line | Data |
---|---|---|
7f918cf1 CE |
1 | <!DOCTYPE html>\r |
2 | <html lang="en">\r | |
3 | <head>\r | |
4 | <meta http-equiv="Content-Type" content="text/html; charset=UTF-8">\r | |
5 | <meta name="generator" content="AsciiDoc 8.6.9">\r | |
6 | <title>DeepFlatten</title>\r | |
7 | <link rel="stylesheet" href="./asciidoc.css" type="text/css">\r | |
8 | <link rel="stylesheet" href="./pygments.css" type="text/css">\r | |
9 | \r | |
10 | \r | |
11 | <script type="text/javascript" src="./asciidoc.js"></script>\r | |
12 | <script type="text/javascript">\r | |
13 | /*<![CDATA[*/\r | |
14 | asciidoc.install();\r | |
15 | /*]]>*/\r | |
16 | </script>\r | |
17 | <link rel="stylesheet" href="./mlton.css" type="text/css">\r | |
18 | </head>\r | |
19 | <body class="article">\r | |
20 | <div id="banner">\r | |
21 | <div id="banner-home">\r | |
22 | <a href="./Home">MLton 20180207</a>\r | |
23 | </div>\r | |
24 | </div>\r | |
25 | <div id="header">\r | |
26 | <h1>DeepFlatten</h1>\r | |
27 | </div>\r | |
28 | <div id="content">\r | |
29 | <div id="preamble">\r | |
30 | <div class="sectionbody">\r | |
31 | <div class="paragraph"><p><a href="DeepFlatten">DeepFlatten</a> is an optimization pass for the <a href="SSA2">SSA2</a>\r | |
32 | <a href="IntermediateLanguage">IntermediateLanguage</a>, invoked from <a href="SSA2Simplify">SSA2Simplify</a>.</p></div>\r | |
33 | </div>\r | |
34 | </div>\r | |
35 | <div class="sect1">\r | |
36 | <h2 id="_description">Description</h2>\r | |
37 | <div class="sectionbody">\r | |
38 | <div class="paragraph"><p>This pass flattens into mutable fields of objects and into vectors.</p></div>\r | |
39 | <div class="paragraph"><p>For example, an <span class="monospaced">(int * int) ref</span> is represented by a 2 word\r | |
40 | object, and an <span class="monospaced">(int * int) array</span> contains pairs of <span class="monospaced">int</span>-s,\r | |
41 | rather than pointers to pairs of <span class="monospaced">int</span>-s.</p></div>\r | |
42 | </div>\r | |
43 | </div>\r | |
44 | <div class="sect1">\r | |
45 | <h2 id="_implementation">Implementation</h2>\r | |
46 | <div class="sectionbody">\r | |
47 | <div class="ulist"><ul>\r | |
48 | <li>\r | |
49 | <p>\r | |
50 | <a href="https://github.com/MLton/mlton/blob/master/mlton/ssa/deep-flatten.fun"><span class="monospaced">deep-flatten.fun</span></a>\r | |
51 | </p>\r | |
52 | </li>\r | |
53 | </ul></div>\r | |
54 | </div>\r | |
55 | </div>\r | |
56 | <div class="sect1">\r | |
57 | <h2 id="_details_and_notes">Details and Notes</h2>\r | |
58 | <div class="sectionbody">\r | |
59 | <div class="paragraph"><p>There are some performance issues with the deep flatten pass, where it\r | |
60 | consumes an excessive amount of memory.</p></div>\r | |
61 | <div class="ulist"><ul>\r | |
62 | <li>\r | |
63 | <p>\r | |
64 | <a href="http://www.mlton.org/pipermail/mlton/2005-April/026990.html">http://www.mlton.org/pipermail/mlton/2005-April/026990.html</a>\r | |
65 | </p>\r | |
66 | </li>\r | |
67 | <li>\r | |
68 | <p>\r | |
69 | <a href="http://www.mlton.org/pipermail/mlton-user/2010-June/001626.html">http://www.mlton.org/pipermail/mlton-user/2010-June/001626.html</a>\r | |
70 | </p>\r | |
71 | </li>\r | |
72 | <li>\r | |
73 | <p>\r | |
74 | <a href="http://www.mlton.org/pipermail/mlton/2010-December/030876.html">http://www.mlton.org/pipermail/mlton/2010-December/030876.html</a>\r | |
75 | </p>\r | |
76 | </li>\r | |
77 | </ul></div>\r | |
78 | <div class="paragraph"><p>A number of applications require compilation with\r | |
79 | <span class="monospaced">-disable-pass deepFlatten</span> to avoid exceeding available memory. It is\r | |
80 | often asked whether the deep flatten pass usually has a significant\r | |
81 | impact on performance. The standard benchmark suite was run with and\r | |
82 | without the deep flatten pass enabled when the pass was first\r | |
83 | introduced:</p></div>\r | |
84 | <div class="ulist"><ul>\r | |
85 | <li>\r | |
86 | <p>\r | |
87 | <a href="http://www.mlton.org/pipermail/mlton/2004-August/025760.html">http://www.mlton.org/pipermail/mlton/2004-August/025760.html</a>\r | |
88 | </p>\r | |
89 | </li>\r | |
90 | </ul></div>\r | |
91 | <div class="paragraph"><p>The conclusion is that it does not have a significant impact.\r | |
92 | However, these are micro benchmarks; other applications may derive\r | |
93 | greater benefit from the pass.</p></div>\r | |
94 | </div>\r | |
95 | </div>\r | |
96 | </div>\r | |
97 | <div id="footnotes"><hr></div>\r | |
98 | <div id="footer">\r | |
99 | <div id="footer-text">\r | |
100 | </div>\r | |
101 | <div id="footer-badges">\r | |
102 | </div>\r | |
103 | </div>\r | |
104 | </body>\r | |
105 | </html>\r |