home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.os.vms
- Path: sparky!uunet!zaphod.mps.ohio-state.edu!srvr1.engin.umich.edu!deccrl!news.crl.dec.com!news!nntpd.lkg.dec.com!Nephi.Enet.DEC.Com!coar
- From: coar@Nephi.Enet.DEC.Com (Rodent of Unusual Size)
- Subject: Re: re: Re: HELP!!! Security problem for gurus. [Directories]
- Message-ID: <1993Jan22.203904.12819@nntpd.lkg.dec.com>
- Summary: It's in there..
- Keywords: ACLEVEL, access mode, security, protection, file system
- Sender: usenet@nntpd.lkg.dec.com (USENET News System)
- Organisation: Employed by (*NOT* speaker for) DEC, Boxborough, MA, U.S.A.
- Organization: Digital Equipment Corp.
- References: <9212290242.AA00964@uu3.psi.com>
- Distribution: comp
- Date: Fri, 22 Jan 1993 20:16:35 GMT
- Lines: 52
-
-
- In article <9212290242.AA00964@uu3.psi.com>, leichter@lrw.com (Jerry Leichter) writes...
- >
- >(ODS-II has several other "spec'ed but never implemented" features, including
- >a hashed file type and something I gather will finally appear in VMS V6:
- >Access-mode restricted files (i.e., files that can only be accessed from
- >inner-mode code, regardless of the usual protections on the file).
-
- The access-mode restriction is in there now, and has worked since at least
- 1984. Unfortunately, file accesses are generally performed by RMS, which
- runs in executive mode, so the only really useful settings involve kernel
- restrictions. It was possible to set up a write-only file this way..
-
- Back in them olde days, the checks for the access mode field were rather
- orphaned; you could change the access-mode protection on a file and BYPASS
- wouldn't let someone at it.
-
- A lot of things changed in this area around the time of late VMS V4 or VMS
- V5.0. The situation with BYPASS and READALL was corrected, and
- BYPASS will now let you access files regardless of the contents of the
- ACLEVEL field. The nesting of write-access as a proper superset of
- read-access appears to have been implemented at the same time; you can no
- longer write to a file that has read access denied to your access mode.
- In addition, the mode-checking seems to have been changed to check against
- PRVMOD rather than CURMOD, so protecting a file against user, supervisor,
- and executive-mode access is now meaningful.
-
- Unfortunately, this entire area is semi-documented, and not supported.
- The behaviour is not exactly right, but I'm not going to spend the time at
- the moment to figure out the details of the deviations. If you're
- interested, the thing to check out is the ATR$C_ACLEVEL file attribute in
- the file system chapter of the I/O Abuser's Guide. I have an antique
- BLISS-32 programme that frobs the field.
-
- >It's not clear to me what BACKUP would do with a hacked directory entry. It
- >might complain, making the hacked files VERY visible during the next system
- >incremental.
-
- Hacked directory entries are processed silently and as-is on image
- backups, but not for non-image activities. (I hacked a directory long ago
- with PATCH /ABSOLUTE to produce filenames with spaces, escape sequences,
- and other oddities. DIRECTORY displayed them verbatim, but BACKUP had a
- fit.)
-
- > -- Jerry
-
- #ken :-)} Jeratol the Chaotic
-
- Coar@Nephi.Enet.DEC.Com | All opinions herein contained, stated or implied,
- Coar@DECUS.Org | are solely those of the author. And he's fullovem.
- Coar@Eisner.DECUS.Org | `... it was mine art, ... that made gape the pine
- Massachusetts, USA | and let thee out.' - Prospero (_The Tempest_)
-