Re: Switch statement code generation

Joshua Cranmer <Pidgeot18@verizon.invalid> writes:
>I'm trying to put together a list of various methods to do code generate
>for switch statements.
...
>Are there any other methods worth mentioning?


Hash-table-based jump tables.


IIRC the following paper discusses these methods:


@Article{journals/spe/KannanP94,
    title = "Short Communication: Correction to 'Producing Good
Code for the case Statement'",
    author = "Sampath Kannan and Todd A. Proebsting",
    journal = "Softw, Pract. Exper",
    year = "1994",
    number = "2",
    volume = "24",
    bibdate = "2003-11-25",
    bibsource = "DBLP,
http://dblp.uni-trier.de/db/journals/spe/spe24.html#KannanP94",
    pages = "233",
}


and this refers to:


@Article{bernstein:85a,
    author = "R. L. Bernstein",
    title = "Producing good code for the case statement",
    journal = "Software -- Practice \& Experience",
    volume = "15",
    number = "10",
    pages = "1021--1024",
    month = oct,
    year = "1985",
}


These days, with the varying amounts of branch prediction hardware,
the best way probably depends on the hardware (which branch predictors
are implemented in the hardware?) and on the run-time characteristics
of the program (how effective are the branch predictors on this
particular usage pattern).


- anton
--
M. Anton Ertl
anton@mips.complang.tuwien.ac.at
http://www.complang.tuwien.ac.at/anton/

Return to the comp.compilers page.
Search the comp.compilers archives again.