Re: Caller/Callee saved Registers

hbaker@netcom.com (Henry G. Baker) writes:
>[When I first saw 'execute' instructions I thought "Why?"]


I've long been philosophizing about the `meaning' of dynamic code and I
like to view the `execute' instruction as an early attempt at structured
self-modifying code. I guess the form of the `execute' instruction, as
compared to some more general support for dynamic compilation, was
motivated by hardware designs of the day (easy to move a register in to
the internal instruction register), and its fall from favor was driven in
part by the trend towards more regular instruction sets that do less work
on each instruction (e.g., no implied looping); these changes make it hard
to pack big commands like "walk the following memory and apply this op. to
each nonzero element" in to a single register.


>[I've rarely seen it used.]


Mimic S/370 simulator lacks general support for dynamic compilation and
SMC (I-cache flushing and moving bits from D-space to I-space) but
`execute' is widely enough that Mimic supports it.


%A Cathy May
%T Mimic: A Fast S/370 Simulator
%J Proceedings of the ACM SIGPLAN 1987 Symposium on Interpreters and
Interpretive Techniques; SIGPLAN Notices
%V 22
%N 6
%C St. Paul, MN
%D June 1987
%P 1-13


;-D on ( Coming soon: self-modifying hardware ) Pardo
--

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