Region-Based Memory Management
DIKU International Summer School
Monday 18 August - Friday 22 August, 1997
Copenhagen
From: http://www.diku.dk/research-groups/topps/activities/kit2/summerschool.html
Abstract
The ML Kit with Regions was released
in April 1997. The release followed a five-year
effort to develop region-based implementation
technology. The summer school is a
chance to learn about region-based memory management and its
realisation in the ML Kit from people who have worked on developing it.
(Added 21 August 1997: slides and exercises used during the course.)
The summer school will be concerned with the theory and implementation
of region-based memory management. In addition, there will be programming
experiments, in which participants can experience programming with regions.
Intended audience
The summer school is primarily targeted at Ph.D. students with an interest in
programming languages.
Participants are assumed to know Standard ML (or a similar programming
language). The theoretical parts assume knowledge of operational semantics.
No prior knowledge of regions is assumed.
More senior researchers are of course welcome too!
Speakers:
- Mads Tofte
- Lars Birkedal
- Martin Elsman
- Tommy Højfeld Olesen
- Peter Sestoft
Course Components
- Theory: Region Inference rules (soundness and algorithms),
Multiplicity Inference, Storage Mode Analysis, Physical Size Inference, Modules and
Separate Compilation.
- Implementation: The Design of the ML Kit, the Kit Abstract
Machine, Region Profiling, Garbage Collecting Regions.
- Programming: Basic Values, Lists, First-Order Functions,
Value Declarations, Static Detection of Space Leaks, References, Recursive
Data Types, Exceptions, Resetting Regions, Higher-Order Functions, Efficient
Function Calls.
Timetable (Postscript)
Course Materials
Venue
Department of Computer Science (DIKU),
University of Copenhagen,
Universitetsparken 1,
DK-2100 Copenhagen
Travel information
Maintained by tofte@itu.dk.
Secretary for the Summer School: Karin Outzen karino@diku.dk