home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!world!ksr!jfw
- From: jfw@ksr.com (John F. Woods)
- Newsgroups: comp.std.c
- Subject: Re: Must an object remain in place? (was: fwrite+fread of pointer)
- Message-ID: <19173@ksr.com>
- Date: 23 Nov 92 11:28:37 EST
- References: <1992Nov18.104542.8499@thunder.mcrcim.mcgill.edu> <1992Nov19.061601.13164@sq.sq.com> <7951@charon.cwi.nl> <1992Nov21.110226.24705@sq.sq.com>
- Sender: news@ksr.com
- Lines: 22
-
- msb@sq.sq.com (Mark Brader) writes:
- >> Well, examine the *as if* rule. I think a run-time system is able to
- >> move data under the proviso that it is able to identify all pointers
- >> and update them accordingly; even if they are stored in non-pointer type
- >> variables. (I know of one Algol 68 garbage collector that did just that,
- >> but it cheated. On a tagged architecture it would certainly be possible.)
- >In theory maybe, but not in practice. Consider that you can transform
- >a pointer to a character string with "%p" in sprintf(), do arbitrarily
- >complex operations on the characters which end up reproducing the same
- >string, then use "%p" in sscanf() to reconstruct the pointer, and the
- >result must compare equal to the original pointer.
-
- That's a pretty strong requirement, I think, but still not quite enough to
- (by itself) ensure that the contents of pointer variables may not change
- during execution:
-
- printf("%p\n", ptr);
-
- could yield something like "<The Object Named Fred-0007>", if the architecture
- supported some other constant naming scheme for objects other than their
- virtual bus addresses.
-
-