incremental compilation

Jecel Mattos Assumpcao Jr vme131!lsiserv1!jecel at uunet.UU.NET
Tue Apr 23 12:52:38 UTC 1991


I don't know if this idea has already been mentioned. If so, here it
is again :

What if the first time a method was called it was compiled by a quick
and dirty compiler and placed in the compiled method cache ? Then, if
it turned out not to be used very often it would be kicked out of the
cache to make room for others. If, on the other hand, it was called a
lot another, better, compiler would generate an optimized code for the
method and place it in the cache. This could be exetended to any number
of levels of optimization, but I doubt that more than two would be useful.

The idea here is to make the compiler work harder on the methods which
really make a difference, not to waste too much time with initialization
code or stuff like that. Since compilation happens at run time in SELF
this might make interactive applications seem more responsive without
degrading benchmarks and other compute-intensive programs too much.

- Jecel



More information about the Self-interest mailing list