Related articles |
---|
real-time defragmentation, etc.... joe@erix.ericsson.se (1993-09-29) |
Newsgroups: | comp.compilers |
From: | joe@erix.ericsson.se (Joe Armstrong) |
Keywords: | GC |
Organization: | Ellemtel Telecom Systems Labs, Stockholm, Sweden |
Date: | Wed, 29 Sep 1993 10:29:36 GMT |
I'm interested in real-time garbage collection techniques for
functional languages. It seems to me that RT GC is a combination of 4
problems:
1) RT deallocation of memory
2) RT allocation of memory
3) RT defragmentation
4) RT address space condensation
1) and 2) are well covered in the literature. Are their any papers in
the literature about 3 and 4?.
Specifically I'm thinking about schemes which do not move data in the
alloc/decalloc process - this produces fragmentation which I'd like to
(concurrently) remove!
I'd also like to condense the address space.
(Yes I *know* that real-time baker schemes do all of the above but
they *move* lots of data - I'm interested in keeping the data in the same
place most of the time (possibly moving it when defragmenting or
condensing))
Joe
--
Return to the
comp.compilers page.
Search the
comp.compilers archives again.