home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.unix.aix
- Path: sparky!uunet!ddssuprs!lpc
- From: lpc@dickens.com (Luis P Caamano)
- Subject: Re: process numbers
- Message-ID: <1993Jan11.231826.25685@dickens.com>
- Date: Mon, 11 Jan 1993 23:18:26 GMT
- References: <1h6hv0INNrsl@life.ai.mit.edu> <C0EB6D.174w@austin.ibm.com> <C0Ly9M.nB6@austin.ibm.com>
- Organization: Dickens Data Systems, Inc.
- Lines: 34
-
- In article <C0Ly9M.nB6@austin.ibm.com> jfh@greenber.austin.ibm.com (John F Haugh II) writes:
- >
- >>>This is actually fairly useful. When I want to spy on a process with
- >>>`crash', it's easy to calculate the index into the process table which
- >>>I need for the `u' command.
- >>
- >> I always looked at the output from the 'proc' subcommand, then
- >> used the slot number for 'u'.....
- >
- >Charles is right on this one.
- >
- >% expr PID / 256
- >
- >is a LOT easier than wading through hundreds of lines of output from "proc".
-
-
- The covert channel is pretty interesting (not being in the cracking-
- hacking-spy-security business, I never thought of that one :)
-
- Anyway what I wanted to say is that I thought the new proc number
- thing was done for performance. With the AIX scheme, it's easier
- to get to the proc slot in the proc table from the proc number
- than using the sequential numbering scheme in classic unix since
- in the latter the proc number doesn't have any information about
- the slot. Does it make sense to you? I might be missing
- something ... :-}
-
-
- --
- Luis P. Caamano | lpc@dickens.com
- Dickens Data Systems, Inc. Atlanta, GA | uunet!dickens.com!lpc
- ---------------------------------------------------------------------------
- If I think I know it all, I'll stop learning. -myself
- The more I learn, the more I know I know nothing. -somebody else
-