|Stack based-Register Based email@example.com (Vinoth Kumar) (2001-01-19)|
|Re: Stack based-Register Based firstname.lastname@example.org (2001-01-20)|
|Re: Stack based-Register Based email@example.com (2001-01-20)|
|Re: Stack based-Register Based firstname.lastname@example.org (Roedy Green) (2001-01-26)|
|Re: Stack based-Register Based email@example.com (2001-01-26)|
|Re: Stack based-Register Based firstname.lastname@example.org (Andrew Berg) (2001-01-28)|
|Re: Stack based-Register Based email@example.com (2001-02-01)|
|Re: Stack based-Register Based firstname.lastname@example.org (Jonathan Barker) (2001-02-01)|
|Re: Stack based-Register Based email@example.com (2001-02-04)|
|Re: Stack based-Register Based Martin.Ward@durham.ac.uk (2001-02-04)|
|[7 later articles]|
|From:||Roedy Green <firstname.lastname@example.org>|
|Date:||26 Jan 2001 16:45:54 -0500|
|Organization:||Canadian Mind Products|
|Posted-Date:||26 Jan 2001 16:45:54 EST|
On 20 Jan 2001 16:13:22 -0500, email@example.com (Anton
Ertl) wrote or quoted :
>for interpreters stack machines are a good choice
These are easy to generate code for and easy to implement. Lots of
hobbyists have implemented a Forth interpreter for example, which is
very similar to a JVM. It ensures rapid proliferation to many
platforms, at least of a simple interpreter.
You don't need to deal with the problem of a limited supply of
registers. Since the number of registers and the way they work can
vary so much between platforms, it is best to keep any assumptions
about how they work out of the JVM design. That is really more a
matter of optimising and fine tuning.
A stack machine that almost fully implements the JVM stack machine
directly is quite possible. You don't need the overhead of a JIT.
This is important is something as tiny as a cellphone.
For the JAVA GLOSSARY see http://www.mindprod.com/jgloss.html
Roedy Green, Canadian Mind Products
Custom computer programming since 1963
Return to the
Search the comp.compilers archives again.