[hcoop/debian/mlton.git] / doc / guide / localhost / GoogleSummerOfCode2013

<!DOCTYPE html>\r
<html lang="en">\r
<head>\r
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">\r
<meta name="generator" content="AsciiDoc 8.6.9">\r
<title>Google Summer of Code (2013)</title>\r
<link rel="stylesheet" href="./asciidoc.css" type="text/css">\r
<link rel="stylesheet" href="./pygments.css" type="text/css">\r
\r
\r
<script type="text/javascript" src="./asciidoc.js"></script>\r
<script type="text/javascript">\r
/*<![CDATA[*/\r
asciidoc.install(2);\r
/*]]>*/\r
</script>\r
<link rel="stylesheet" href="./mlton.css" type="text/css">\r
</head>\r
<body class="article">\r
<div id="banner">\r
<div id="banner-home">\r
<a href="./Home">MLton 20180207</a>\r
</div>\r
</div>\r
<div id="header">\r
<h1>Google Summer of Code (2013)</h1>\r
<div id="toc">
  <div id="toctitle">Table of Contents</div>
  <noscript><p><b>JavaScript must be enabled in your browser to display the table of contents.</b></p></noscript>
</div>\r
</div>\r
<div id="content">\r
<div class="sect1">\r
<h2 id="_mentors">Mentors</h2>\r
<div class="sectionbody">\r
<div class="paragraph"><p>The following developers have agreed to serve as mentors for the 2013 Google Summer of Code:</p></div>\r
<div class="ulist"><ul>\r
<li>\r
<p>\r
<a href="http://www.cs.rit.edu/%7Emtf">Matthew Fluet</a>\r
</p>\r
</li>\r
<li>\r
<p>\r
<a href="http://www.cse.buffalo.edu/%7Elziarek/">Lukasz (Luke) Ziarek</a>\r
</p>\r
</li>\r
<li>\r
<p>\r
<a href="http://www.cs.purdue.edu/homes/suresh/">Suresh Jagannathan</a>\r
</p>\r
</li>\r
</ul></div>\r
</div>\r
</div>\r
<div class="sect1">\r
<h2 id="_ideas_list">Ideas List</h2>\r
<div class="sectionbody">\r
<div class="sect2">\r
<h3 id="_implement_a_partial_redundancy_elimination_pre_optimization">Implement a Partial Redundancy Elimination (PRE) Optimization</h3>\r
<div class="paragraph"><p>Partial redundancy elimination (PRE) is a program transformation that\r
removes operations that are redundant on some, but not necessarily all\r
paths, through the program.  PRE can subsume both common subexpression\r
elimination and loop-invariant code motion, and is therefore a\r
potentially powerful optimization.  However, a na&iuml;ve\r
implementation of PRE on a program in static single assignment (SSA)\r
form is unlikely to be effective.  This project aims to adapt and\r
implement the SSAPRE algorithm(s) of Thomas VanDrunen in MLton&#8217;s SSA\r
intermediate language.</p></div>\r
<div class="paragraph"><p>Background:</p></div>\r
<div class="openblock">\r
<div class="content">\r
<div class="ulist"><ul>\r
<li>\r
<p>\r
<a href="http://onlinelibrary.wiley.com/doi/10.1002/spe.618/abstract">Anticipation-based partial redundancy elimination for static single assignment form</a>; Thomas VanDrunen and Antony L. Hosking\r
</p>\r
</li>\r
<li>\r
<p>\r
<a href="http://cs.wheaton.edu/%7Etvandrun/writings/thesis.pdf">Partial Redundancy Elimination for Global Value Numbering</a>; Thomas VanDrunen\r
</p>\r
</li>\r
<li>\r
<p>\r
<a href="http://www.springerlink.com/content/w06m3cw453nphm1u/">Value-Based Partial Redundancy Elimination</a>; Thomas VanDrunen and Antony L. Hosking\r
</p>\r
</li>\r
<li>\r
<p>\r
<a href="http://portal.acm.org/citation.cfm?doid=319301.319348">Partial redundancy elimination in SSA form</a>; Robert Kennedy, Sun Chan, Shin-Ming Liu, Raymond Lo, Peng Tu, and Fred Chow\r
</p>\r
</li>\r
</ul></div>\r
</div></div>\r
<div class="paragraph"><p>Recommended Skills: SML programming experience; some middle-end compiler experience</p></div>\r
</div>\r
<div class="sect2">\r
<h3 id="_design_and_implement_a_heap_profiler">Design and Implement a Heap Profiler</h3>\r
<div class="paragraph"><p>A heap profile is a description of the space usage of a program.  A\r
heap profile is concerned with the allocation, retention, and\r
deallocation (via garbage collection) of heap data during the\r
execution of a program.  A heap profile can be used to diagnose\r
performance problems in a functional program that arise from space\r
leaks.  This project aims to design and implement a heap profiler for\r
MLton compiled programs.</p></div>\r
<div class="paragraph"><p>Background:</p></div>\r
<div class="openblock">\r
<div class="content">\r
<div class="ulist"><ul>\r
<li>\r
<p>\r
<a href="http://portal.acm.org/citation.cfm?doid=583854.582451">GCspy: an adaptable heap visualisation framework</a>; Tony Printezis and Richard Jones\r
</p>\r
</li>\r
<li>\r
<p>\r
<a href="http://journals.cambridge.org/action/displayAbstract?aid=1349892">New dimensions in heap profiling</a>; Colin Runciman and Niklas R&ouml;jemo\r
</p>\r
</li>\r
<li>\r
<p>\r
<a href="http://www.springerlink.com/content/710501660722gw37/">Heap profiling for space efficiency</a>; Colin Runciman and Niklas R&ouml;jemo\r
</p>\r
</li>\r
<li>\r
<p>\r
<a href="http://journals.cambridge.org/action/displayAbstract?aid=1323096">Heap profiling of lazy functional programs</a>; Colin Runciman and David Wakeling\r
</p>\r
</li>\r
</ul></div>\r
</div></div>\r
<div class="paragraph"><p>Recommended Skills: C and SML programming experience; some experience with UI and visualization</p></div>\r
</div>\r
<div class="sect2">\r
<h3 id="_garbage_collector_improvements">Garbage Collector Improvements</h3>\r
<div class="paragraph"><p>The garbage collector plays a significant role in the performance of\r
functional languages.  Garbage collect too often, and program\r
performance suffers due to the excessive time spent in the garbage\r
collector.  Garbage collect not often enough, and program performance\r
suffers due to the excessive space used by the uncollected garbage.\r
One particular issue is ensuring that a program utilizing a garbage\r
collector "plays nice" with other processes on the system, by not\r
using too much or too little physical memory.  While there are some\r
reasonable theoretical results about garbage collections with heaps of\r
fixed size, there seems to be insufficient work that really looks\r
carefully at the question of dynamically resizing the heap in response\r
to the live data demands of the application and, similarly, in\r
response to the behavior of the operating system and other processes.\r
This project aims to investigate improvements to the memory behavior of\r
MLton compiled programs through better tuning of the garbage\r
collector.</p></div>\r
<div class="paragraph"><p>Background:</p></div>\r
<div class="openblock">\r
<div class="content">\r
<div class="ulist"><ul>\r
<li>\r
<p>\r
<a href="http://www.dcs.gla.ac.uk/%7Ewhited/papers/automated_heap_sizing.pdf">Automated Heap Sizing in the Poly/ML Runtime (Position Paper)</a>; David White, Jeremy Singer, Jonathan Aitken, and David Matthews\r
</p>\r
</li>\r
<li>\r
<p>\r
<a href="http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=4145125">Isla Vista Heap Sizing: Using Feedback to Avoid Paging</a>; Chris Grzegorczyk, Sunil Soman, Chandra Krintz, and Rich Wolski\r
</p>\r
</li>\r
<li>\r
<p>\r
<a href="http://portal.acm.org/citation.cfm?doid=1152649.1152652">Controlling garbage collection and heap growth to reduce the execution time of Java applications</a>; Tim Brecht, Eshrat Arjomandi, Chang Li, and Hang Pham\r
</p>\r
</li>\r
<li>\r
<p>\r
<a href="http://portal.acm.org/citation.cfm?doid=1065010.1065028">Garbage collection without paging</a>; Matthew Hertz, Yi Feng, and Emery D. Berger\r
</p>\r
</li>\r
<li>\r
<p>\r
<a href="http://portal.acm.org/citation.cfm?doid=1029873.1029881">Automatic heap sizing: taking real memory into account</a>; Ting Yang, Matthew Hertz, Emery D. Berger, Scott F. Kaplan, and J. Eliot B. Moss\r
</p>\r
</li>\r
</ul></div>\r
</div></div>\r
<div class="paragraph"><p>Recommended Skills: C programming experience; some operating systems and/or systems programming experience; some compiler and garbage collector experience</p></div>\r
</div>\r
<div class="sect2">\r
<h3 id="_implement_successor_160_ml_language_features">Implement Successor&#160;ML Language Features</h3>\r
<div class="paragraph"><p>Any programming language, including Standard&#160;ML, can be improved.\r
The community has identified a number of modest extensions and\r
revisions to the Standard&#160;ML programming language that would\r
likely prove useful in practice.  This project aims to implement these\r
language features in the MLton compiler.</p></div>\r
<div class="paragraph"><p>Background:</p></div>\r
<div class="openblock">\r
<div class="content">\r
<div class="ulist"><ul>\r
<li>\r
<p>\r
<a href="http://successor-ml.org/index.php?title=Main_Page">Successor&#160;ML</a>\r
</p>\r
</li>\r
<li>\r
<p>\r
<a href="http://www.mpi-sws.org/%7Erossberg/hamlet/index.html#successor-ml">HaMLet (Successor&#160;ML)</a>\r
</p>\r
</li>\r
<li>\r
<p>\r
<a href="http://journals.cambridge.org/action/displayAbstract?aid=1322628">A critique of Standard&#160;ML</a>; Andrew W. Appel\r
</p>\r
</li>\r
</ul></div>\r
</div></div>\r
<div class="paragraph"><p>Recommended Skills: SML programming experience; some front-end compiler experience (i.e., scanners and parsers)</p></div>\r
</div>\r
<div class="sect2">\r
<h3 id="_implement_source_level_debugging">Implement Source-level Debugging</h3>\r
<div class="paragraph"><p>Debugging is a fact of programming life.  Unfortunately, most SML\r
implementations (including MLton) provide little to no source-level\r
debugging support.  This project aims to add basic to intermediate\r
source-level debugging support to the MLton compiler.  MLton already\r
supports source-level profiling, which can be used to attribute bytes\r
allocated or time spent in source functions.  It should be relatively\r
straightforward to leverage this source-level information into basic\r
source-level debugging support, with the ability to set/unset\r
breakpoints and step through declarations and functions.  It may be\r
possible to also provide intermediate source-level debugging support,\r
with the ability to inspect in-scope variables of basic types (e.g.,\r
types compatible with MLton&#8217;s foreign function interface).</p></div>\r
<div class="paragraph"><p>Background:</p></div>\r
<div class="openblock">\r
<div class="content">\r
<div class="ulist"><ul>\r
<li>\r
<p>\r
<a href="http://mlton.org/HowProfilingWorks">MLton&#8201;&#8212;&#8201;How Profiling Works</a>\r
</p>\r
</li>\r
<li>\r
<p>\r
<a href="http://mlton.org/ForeignFunctionInterfaceTypes">MLton&#8201;&#8212;&#8201;Foreign Function Interface Types</a>\r
</p>\r
</li>\r
<li>\r
<p>\r
<a href="http://dwarfstd.org/">DWARF Debugging Standard</a>\r
</p>\r
</li>\r
<li>\r
<p>\r
<a href="http://sourceware.org/gdb/current/onlinedocs/stabs/index.html">STABS Debugging Format</a>\r
</p>\r
</li>\r
</ul></div>\r
</div></div>\r
<div class="paragraph"><p>Recommended Skills: SML programming experience; some compiler experience</p></div>\r
</div>\r
<div class="sect2">\r
<h3 id="_simd_primitives">SIMD Primitives</h3>\r
<div class="paragraph"><p>Most modern processors offer some direct support for SIMD (Single\r
Instruction, Multiple Data) operations, such as Intel&#8217;s MMX/SSE\r
instructions, AMD&#8217;s 3DNow!  instructions, and IBM&#8217;s AltiVec.  Such\r
instructions are particularly useful for multimedia, scientific, and\r
cryptographic applications.  This project aims to add preliminary\r
support for vector data and vector operations to the MLton compiler.\r
Ideally, after surveying SIMD instruction sets and SIMD support in\r
other compilers, a core set of SIMD primitives with broad architecture\r
and compiler support can be identified.  After adding SIMD primitives\r
to the core compiler and carrying them through to the various\r
backends, there will be opportunities to design and implement an SML\r
library that exposes the primitives to the SML programmer as well as\r
opportunities to design and implement auto-vectorization\r
optimizations.</p></div>\r
<div class="paragraph"><p>Background:</p></div>\r
<div class="openblock">\r
<div class="content">\r
<div class="ulist"><ul>\r
<li>\r
<p>\r
<a href="http://en.wikipedia.org/wiki/SIMD">SIMD</a>\r
</p>\r
</li>\r
<li>\r
<p>\r
<a href="http://gcc.gnu.org/projects/tree-ssa/vectorization.html">Auto-vectorization in GCC</a>\r
</p>\r
</li>\r
<li>\r
<p>\r
<a href="http://llvm.org/docs/Vectorizers.html">Auto-vectorization in LLVM</a>\r
</p>\r
</li>\r
</ul></div>\r
</div></div>\r
<div class="paragraph"><p>Recommended Skills: SML programming experience; some compiler experience; some computer architecture experience</p></div>\r
</div>\r
<div class="sect2">\r
<h3 id="_rtos_support">RTOS Support</h3>\r
<div class="paragraph"><p>This project entails porting the MLton compiler to RTOSs such as:\r
RTEMS, RT Linux, and FreeRTOS.  The project will include modifications\r
to the MLton build and configuration process.  Students will need to\r
extend the MLton configuration process for each of the RTOSs.  The\r
MLton compilation process will need to be extended to invoke the C\r
cross compilers the RTOSs provide for embedded support.  Test scripts\r
for validation will be necessary and these will need to be run in\r
emulators for supported architectures.</p></div>\r
<div class="paragraph"><p>Recommended Skills: C programming experience; some scripting experience</p></div>\r
</div>\r
<div class="sect2">\r
<h3 id="_region_based_memory_management">Region Based Memory Management</h3>\r
<div class="paragraph"><p>Region based memory management is an alternative automatic memory\r
management scheme to garbage collection.  Regions can be inferred by\r
the compiler (e.g., Cyclone and MLKit) or provided to the programmer\r
through a library.  Since many students do not have extensive\r
experience with compilers we plan on adopting the later approach.\r
Creating a viable region based memory solution requires the removal of\r
the GC and changes to the allocator.  Additionally, write barriers\r
will be necessary to ensure references between two ML objects is never\r
established if the left hand side of the assignment has a longer\r
lifetime than the right hand side.  Students will need to come up with\r
an appropriate interface for creating, entering, and exiting regions\r
(examples include RTSJ scoped memory and SCJ scoped memory).</p></div>\r
<div class="paragraph"><p>Background:</p></div>\r
<div class="openblock">\r
<div class="content">\r
<div class="ulist"><ul>\r
<li>\r
<p>\r
Cyclone\r
</p>\r
</li>\r
<li>\r
<p>\r
MLKit\r
</p>\r
</li>\r
<li>\r
<p>\r
RTSJ + SCJ scopes\r
</p>\r
</li>\r
</ul></div>\r
</div></div>\r
<div class="paragraph"><p>Recommended Skills: SML programming experience; C programming experience; some compiler and garbage collector experience</p></div>\r
</div>\r
<div class="sect2">\r
<h3 id="_integration_of_multi_mlton">Integration of Multi-MLton</h3>\r
<div class="paragraph"><p><a href="http://multimlton.cs.purdue.edu">MultiMLton</a> is a compiler and runtime\r
environment that targets scalable multicore platforms.  It is an\r
extension of MLton.  It combines new language abstractions and\r
associated compiler analyses for expressing and implementing various\r
kinds of fine-grained parallelism (safe futures, speculation,\r
transactions, etc.), along with a sophisticated runtime system tuned\r
to efficiently handle large numbers of lightweight threads.  The core\r
stable features of MultiMLton will need to be integrated with the\r
latest MLton public release.  Certain experimental features, such as\r
support for the Intel SCC and distributed runtime will be omitted.\r
This project requires students to understand the delta between the\r
MultiMLton code base and the MLton code base.  Students will need to\r
create build and configuration scripts for MLton to enable MultiMLton\r
features.</p></div>\r
<div class="paragraph"><p>Background</p></div>\r
<div class="openblock">\r
<div class="content">\r
<div class="ulist"><ul>\r
<li>\r
<p>\r
<a href="http://multimlton.cs.purdue.edu/mML/Publications.html">MultiMLton&#8201;&#8212;&#8201;Publications</a>\r
</p>\r
</li>\r
</ul></div>\r
</div></div>\r
<div class="paragraph"><p>Recommended Skills: SML programming experience; C programming experience; some compiler experience</p></div>\r
</div>\r
</div>\r
</div>\r
</div>\r
<div id="footnotes"><hr></div>\r
<div id="footer">\r
<div id="footer-text">\r
</div>\r
<div id="footer-badges">\r
</div>\r
</div>\r
</body>\r
</html>\r
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>Google Summer of Code (2013)</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(2);\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>Google Summer of Code (2013)</h1>\r
	27	<div id="toc">
	28	<div id="toctitle">Table of Contents</div>
	29	<noscript><p><b>JavaScript must be enabled in your browser to display the table of contents.</b></p></noscript>
	30	</div>\r
	31	</div>\r
	32	<div id="content">\r
	33	<div class="sect1">\r
	34	<h2 id="_mentors">Mentors</h2>\r
	35	<div class="sectionbody">\r
	36	<div class="paragraph"><p>The following developers have agreed to serve as mentors for the 2013 Google Summer of Code:</p></div>\r
	37	<div class="ulist"><ul>\r
	38	<li>\r
	39	<p>\r
	40	<a href="http://www.cs.rit.edu/%7Emtf">Matthew Fluet</a>\r
	41	</p>\r
	42	</li>\r
	43	<li>\r
	44	<p>\r
	45	<a href="http://www.cse.buffalo.edu/%7Elziarek/">Lukasz (Luke) Ziarek</a>\r
	46	</p>\r
	47	</li>\r
	48	<li>\r
	49	<p>\r
	50	<a href="http://www.cs.purdue.edu/homes/suresh/">Suresh Jagannathan</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="_ideas_list">Ideas List</h2>\r
	58	<div class="sectionbody">\r
	59	<div class="sect2">\r
	60	<h3 id="_implement_a_partial_redundancy_elimination_pre_optimization">Implement a Partial Redundancy Elimination (PRE) Optimization</h3>\r
	61	<div class="paragraph"><p>Partial redundancy elimination (PRE) is a program transformation that\r
	62	removes operations that are redundant on some, but not necessarily all\r
	63	paths, through the program. PRE can subsume both common subexpression\r
	64	elimination and loop-invariant code motion, and is therefore a\r
65	potentially powerful optimization. However, a naïve\r
66	implementation of PRE on a program in static single assignment (SSA)\r
67	form is unlikely to be effective. This project aims to adapt and\r
68	implement the SSAPRE algorithm(s) of Thomas VanDrunen in MLton’s SSA\r
69	intermediate language.</p></div>\r
70	<div class="paragraph"><p>Background:</p></div>\r
71	<div class="openblock">\r
72	<div class="content">\r
73	<div class="ulist"><ul>\r
74	<li>\r
75	<p>\r
76	<a href="http://onlinelibrary.wiley.com/doi/10.1002/spe.618/abstract">Anticipation-based partial redundancy elimination for static single assignment form</a>; Thomas VanDrunen and Antony L. Hosking\r
77	</p>\r
78	</li>\r
79	<li>\r
80	<p>\r
81	<a href="http://cs.wheaton.edu/%7Etvandrun/writings/thesis.pdf">Partial Redundancy Elimination for Global Value Numbering</a>; Thomas VanDrunen\r
82	</p>\r
83	</li>\r
84	<li>\r
85	<p>\r
86	<a href="http://www.springerlink.com/content/w06m3cw453nphm1u/">Value-Based Partial Redundancy Elimination</a>; Thomas VanDrunen and Antony L. Hosking\r
87	</p>\r
88	</li>\r
89	<li>\r
90	<p>\r
91	<a href="http://portal.acm.org/citation.cfm?doid=319301.319348">Partial redundancy elimination in SSA form</a>; Robert Kennedy, Sun Chan, Shin-Ming Liu, Raymond Lo, Peng Tu, and Fred Chow\r
92	</p>\r
93	</li>\r
94	</ul></div>\r
95	</div></div>\r
96	<div class="paragraph"><p>Recommended Skills: SML programming experience; some middle-end compiler experience</p></div>\r
97	</div>\r
98	<div class="sect2">\r
99	<h3 id="_design_and_implement_a_heap_profiler">Design and Implement a Heap Profiler</h3>\r
100	<div class="paragraph"><p>A heap profile is a description of the space usage of a program. A\r
101	heap profile is concerned with the allocation, retention, and\r
102	deallocation (via garbage collection) of heap data during the\r
103	execution of a program. A heap profile can be used to diagnose\r
104	performance problems in a functional program that arise from space\r
105	leaks. This project aims to design and implement a heap profiler for\r
106	MLton compiled programs.</p></div>\r
107	<div class="paragraph"><p>Background:</p></div>\r
108	<div class="openblock">\r
109	<div class="content">\r
110	<div class="ulist"><ul>\r
111	<li>\r
112	<p>\r
113	<a href="http://portal.acm.org/citation.cfm?doid=583854.582451">GCspy: an adaptable heap visualisation framework</a>; Tony Printezis and Richard Jones\r
114	</p>\r
115	</li>\r
116	<li>\r
117	<p>\r
118	<a href="http://journals.cambridge.org/action/displayAbstract?aid=1349892">New dimensions in heap profiling</a>; Colin Runciman and Niklas Röjemo\r
119	</p>\r
120	</li>\r
121	<li>\r
122	<p>\r
123	<a href="http://www.springerlink.com/content/710501660722gw37/">Heap profiling for space efficiency</a>; Colin Runciman and Niklas Röjemo\r
124	</p>\r
125	</li>\r
126	<li>\r
127	<p>\r
128	<a href="http://journals.cambridge.org/action/displayAbstract?aid=1323096">Heap profiling of lazy functional programs</a>; Colin Runciman and David Wakeling\r
129	</p>\r
130	</li>\r
131	</ul></div>\r
132	</div></div>\r
133	<div class="paragraph"><p>Recommended Skills: C and SML programming experience; some experience with UI and visualization</p></div>\r
134	</div>\r
135	<div class="sect2">\r
136	<h3 id="_garbage_collector_improvements">Garbage Collector Improvements</h3>\r
137	<div class="paragraph"><p>The garbage collector plays a significant role in the performance of\r
138	functional languages. Garbage collect too often, and program\r
139	performance suffers due to the excessive time spent in the garbage\r
140	collector. Garbage collect not often enough, and program performance\r
141	suffers due to the excessive space used by the uncollected garbage.\r
142	One particular issue is ensuring that a program utilizing a garbage\r
143	collector "plays nice" with other processes on the system, by not\r
144	using too much or too little physical memory. While there are some\r
145	reasonable theoretical results about garbage collections with heaps of\r
146	fixed size, there seems to be insufficient work that really looks\r
147	carefully at the question of dynamically resizing the heap in response\r
148	to the live data demands of the application and, similarly, in\r
149	response to the behavior of the operating system and other processes.\r
150	This project aims to investigate improvements to the memory behavior of\r
151	MLton compiled programs through better tuning of the garbage\r
152	collector.</p></div>\r
153	<div class="paragraph"><p>Background:</p></div>\r
154	<div class="openblock">\r
155	<div class="content">\r
156	<div class="ulist"><ul>\r
157	<li>\r
158	<p>\r
159	<a href="http://www.dcs.gla.ac.uk/%7Ewhited/papers/automated_heap_sizing.pdf">Automated Heap Sizing in the Poly/ML Runtime (Position Paper)</a>; David White, Jeremy Singer, Jonathan Aitken, and David Matthews\r
160	</p>\r
161	</li>\r
162	<li>\r
163	<p>\r
164	<a href="http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=4145125">Isla Vista Heap Sizing: Using Feedback to Avoid Paging</a>; Chris Grzegorczyk, Sunil Soman, Chandra Krintz, and Rich Wolski\r
165	</p>\r
166	</li>\r
167	<li>\r
168	<p>\r
169	<a href="http://portal.acm.org/citation.cfm?doid=1152649.1152652">Controlling garbage collection and heap growth to reduce the execution time of Java applications</a>; Tim Brecht, Eshrat Arjomandi, Chang Li, and Hang Pham\r
170	</p>\r
171	</li>\r
172	<li>\r
173	<p>\r
174	<a href="http://portal.acm.org/citation.cfm?doid=1065010.1065028">Garbage collection without paging</a>; Matthew Hertz, Yi Feng, and Emery D. Berger\r
175	</p>\r
176	</li>\r
177	<li>\r
178	<p>\r
179	<a href="http://portal.acm.org/citation.cfm?doid=1029873.1029881">Automatic heap sizing: taking real memory into account</a>; Ting Yang, Matthew Hertz, Emery D. Berger, Scott F. Kaplan, and J. Eliot B. Moss\r
180	</p>\r
181	</li>\r
182	</ul></div>\r
183	</div></div>\r
184	<div class="paragraph"><p>Recommended Skills: C programming experience; some operating systems and/or systems programming experience; some compiler and garbage collector experience</p></div>\r
185	</div>\r
186	<div class="sect2">\r
187	<h3 id="_implement_successor_160_ml_language_features">Implement Successor ML Language Features</h3>\r
188	<div class="paragraph"><p>Any programming language, including Standard ML, can be improved.\r
189	The community has identified a number of modest extensions and\r
190	revisions to the Standard ML programming language that would\r
191	likely prove useful in practice. This project aims to implement these\r
192	language features in the MLton compiler.</p></div>\r
193	<div class="paragraph"><p>Background:</p></div>\r
194	<div class="openblock">\r
195	<div class="content">\r
196	<div class="ulist"><ul>\r
197	<li>\r
198	<p>\r
199	<a href="http://successor-ml.org/index.php?title=Main_Page">Successor ML</a>\r
200	</p>\r
201	</li>\r
202	<li>\r
203	<p>\r
204	<a href="http://www.mpi-sws.org/%7Erossberg/hamlet/index.html#successor-ml">HaMLet (Successor ML)</a>\r
205	</p>\r
206	</li>\r
207	<li>\r
208	<p>\r
209	<a href="http://journals.cambridge.org/action/displayAbstract?aid=1322628">A critique of Standard ML</a>; Andrew W. Appel\r
210	</p>\r
211	</li>\r
212	</ul></div>\r
213	</div></div>\r
214	<div class="paragraph"><p>Recommended Skills: SML programming experience; some front-end compiler experience (i.e., scanners and parsers)</p></div>\r
215	</div>\r
216	<div class="sect2">\r
217	<h3 id="_implement_source_level_debugging">Implement Source-level Debugging</h3>\r
218	<div class="paragraph"><p>Debugging is a fact of programming life. Unfortunately, most SML\r
219	implementations (including MLton) provide little to no source-level\r
220	debugging support. This project aims to add basic to intermediate\r
221	source-level debugging support to the MLton compiler. MLton already\r
222	supports source-level profiling, which can be used to attribute bytes\r
223	allocated or time spent in source functions. It should be relatively\r
224	straightforward to leverage this source-level information into basic\r
225	source-level debugging support, with the ability to set/unset\r
226	breakpoints and step through declarations and functions. It may be\r
227	possible to also provide intermediate source-level debugging support,\r
228	with the ability to inspect in-scope variables of basic types (e.g.,\r
229	types compatible with MLton’s foreign function interface).</p></div>\r
230	<div class="paragraph"><p>Background:</p></div>\r
231	<div class="openblock">\r
232	<div class="content">\r
233	<div class="ulist"><ul>\r
234	<li>\r
235	<p>\r
236	<a href="http://mlton.org/HowProfilingWorks">MLton — How Profiling Works</a>\r
237	</p>\r
238	</li>\r
239	<li>\r
240	<p>\r
241	<a href="http://mlton.org/ForeignFunctionInterfaceTypes">MLton — Foreign Function Interface Types</a>\r
242	</p>\r
243	</li>\r
244	<li>\r
245	<p>\r
246	<a href="http://dwarfstd.org/">DWARF Debugging Standard</a>\r
247	</p>\r
248	</li>\r
249	<li>\r
250	<p>\r
251	<a href="http://sourceware.org/gdb/current/onlinedocs/stabs/index.html">STABS Debugging Format</a>\r
252	</p>\r
253	</li>\r
254	</ul></div>\r
255	</div></div>\r
256	<div class="paragraph"><p>Recommended Skills: SML programming experience; some compiler experience</p></div>\r
257	</div>\r
258	<div class="sect2">\r
259	<h3 id="_simd_primitives">SIMD Primitives</h3>\r
260	<div class="paragraph"><p>Most modern processors offer some direct support for SIMD (Single\r
261	Instruction, Multiple Data) operations, such as Intel’s MMX/SSE\r
262	instructions, AMD’s 3DNow! instructions, and IBM’s AltiVec. Such\r
263	instructions are particularly useful for multimedia, scientific, and\r
264	cryptographic applications. This project aims to add preliminary\r
265	support for vector data and vector operations to the MLton compiler.\r
266	Ideally, after surveying SIMD instruction sets and SIMD support in\r
267	other compilers, a core set of SIMD primitives with broad architecture\r
268	and compiler support can be identified. After adding SIMD primitives\r
269	to the core compiler and carrying them through to the various\r
270	backends, there will be opportunities to design and implement an SML\r
271	library that exposes the primitives to the SML programmer as well as\r
272	opportunities to design and implement auto-vectorization\r
273	optimizations.</p></div>\r
274	<div class="paragraph"><p>Background:</p></div>\r
275	<div class="openblock">\r
276	<div class="content">\r
277	<div class="ulist"><ul>\r
278	<li>\r
279	<p>\r
280	<a href="http://en.wikipedia.org/wiki/SIMD">SIMD</a>\r
281	</p>\r
282	</li>\r
283	<li>\r
284	<p>\r
285	<a href="http://gcc.gnu.org/projects/tree-ssa/vectorization.html">Auto-vectorization in GCC</a>\r
286	</p>\r
287	</li>\r
288	<li>\r
289	<p>\r
290	<a href="http://llvm.org/docs/Vectorizers.html">Auto-vectorization in LLVM</a>\r
291	</p>\r
292	</li>\r
293	</ul></div>\r
294	</div></div>\r
295	<div class="paragraph"><p>Recommended Skills: SML programming experience; some compiler experience; some computer architecture experience</p></div>\r
296	</div>\r
297	<div class="sect2">\r
298	<h3 id="_rtos_support">RTOS Support</h3>\r
299	<div class="paragraph"><p>This project entails porting the MLton compiler to RTOSs such as:\r
300	RTEMS, RT Linux, and FreeRTOS. The project will include modifications\r
301	to the MLton build and configuration process. Students will need to\r
302	extend the MLton configuration process for each of the RTOSs. The\r
303	MLton compilation process will need to be extended to invoke the C\r
304	cross compilers the RTOSs provide for embedded support. Test scripts\r
305	for validation will be necessary and these will need to be run in\r
306	emulators for supported architectures.</p></div>\r
307	<div class="paragraph"><p>Recommended Skills: C programming experience; some scripting experience</p></div>\r
308	</div>\r
309	<div class="sect2">\r
310	<h3 id="_region_based_memory_management">Region Based Memory Management</h3>\r
311	<div class="paragraph"><p>Region based memory management is an alternative automatic memory\r
312	management scheme to garbage collection. Regions can be inferred by\r
313	the compiler (e.g., Cyclone and MLKit) or provided to the programmer\r
314	through a library. Since many students do not have extensive\r
315	experience with compilers we plan on adopting the later approach.\r
316	Creating a viable region based memory solution requires the removal of\r
317	the GC and changes to the allocator. Additionally, write barriers\r
318	will be necessary to ensure references between two ML objects is never\r
319	established if the left hand side of the assignment has a longer\r
320	lifetime than the right hand side. Students will need to come up with\r
321	an appropriate interface for creating, entering, and exiting regions\r
322	(examples include RTSJ scoped memory and SCJ scoped memory).</p></div>\r
323	<div class="paragraph"><p>Background:</p></div>\r
324	<div class="openblock">\r
325	<div class="content">\r
326	<div class="ulist"><ul>\r
327	<li>\r
328	<p>\r
329	Cyclone\r
330	</p>\r
331	</li>\r
332	<li>\r
333	<p>\r
334	MLKit\r
335	</p>\r
336	</li>\r
337	<li>\r
338	<p>\r
339	RTSJ + SCJ scopes\r
340	</p>\r
341	</li>\r
342	</ul></div>\r
343	</div></div>\r
344	<div class="paragraph"><p>Recommended Skills: SML programming experience; C programming experience; some compiler and garbage collector experience</p></div>\r
345	</div>\r
346	<div class="sect2">\r
347	<h3 id="_integration_of_multi_mlton">Integration of Multi-MLton</h3>\r
348	<div class="paragraph"><p><a href="http://multimlton.cs.purdue.edu">MultiMLton</a> is a compiler and runtime\r
349	environment that targets scalable multicore platforms. It is an\r
350	extension of MLton. It combines new language abstractions and\r
351	associated compiler analyses for expressing and implementing various\r
352	kinds of fine-grained parallelism (safe futures, speculation,\r
353	transactions, etc.), along with a sophisticated runtime system tuned\r
354	to efficiently handle large numbers of lightweight threads. The core\r
355	stable features of MultiMLton will need to be integrated with the\r
356	latest MLton public release. Certain experimental features, such as\r
357	support for the Intel SCC and distributed runtime will be omitted.\r
358	This project requires students to understand the delta between the\r
359	MultiMLton code base and the MLton code base. Students will need to\r
360	create build and configuration scripts for MLton to enable MultiMLton\r
361	features.</p></div>\r
362	<div class="paragraph"><p>Background</p></div>\r
363	<div class="openblock">\r
364	<div class="content">\r
365	<div class="ulist"><ul>\r
366	<li>\r
367	<p>\r
368	<a href="http://multimlton.cs.purdue.edu/mML/Publications.html">MultiMLton — Publications</a>\r
369	</p>\r
370	</li>\r
371	</ul></div>\r
372	</div></div>\r
373	<div class="paragraph"><p>Recommended Skills: SML programming experience; C programming experience; some compiler experience</p></div>\r
374	</div>\r
375	</div>\r
376	</div>\r
377	</div>\r
378	<div id="footnotes"><hr></div>\r
379	<div id="footer">\r
380	<div id="footer-text">\r
381	</div>\r
382	<div id="footer-badges">\r
383	</div>\r
384	</div>\r
385	</body>\r
386	</html>\r