[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
- To: slug@Warbucks.AI.SRI.COM
- Subject: GC questions
- From: firstname.lastname@example.org (Anton Beschta)
- Date: Fri, 22 Feb 1991 09:00:00 -0500
- Fcc: L4:>toni>fcc.mail.newest
- Illegal-object: Syntax error in Message-Id: value found on neat.cs.toronto.edu: Message-Id: <19910222135500.4.TONI@email@example.com> ^-illegal end of message identification
I have some questions about gc on 3600-family machines.
I ran into the problem that static space grows very fast. Using Peek I
tracked the problem down to a particular area. [In my case it was
TV:BIT-ARRAY-AREA. This happend due to really stupid programming :-(.
The problem is fixed now but I still want the following questions
to be answered.]
I tried si:full-gc since it is the only way to gc static areas but
without success. The documentation says that si:full-gc gcs only
"some" static areas. [TV:BIT-ARRAY-AREA seems not to be included.]
Which static areas are gced by si:full-gc?
Is it possible to add areas to the above set?
But one not always wants to perform a si:full-gc with all the stuff on
its initialization list.
Is there a way to change the gc behavior of an area from :static to
:dynamic or are there implementation dependent or theoretical reasons
which makes this impossible?
The documentation claims that make-area is the way to dynamically change
characteristics of areas. Once more I weren't successful.
Are there restrictions to make-area for changing characteristics?
And if above questions from your point of view are all garbage then
refer to the following :-)
How to selectively gc static areas?
Thanks in advance,