home *** CD-ROM | disk | FTP | other *** search
- From: Torsten Scherer <itschere@techfak.uni-bielefeld.de>
- Subject: XATTR structure for BIOS filesystem #2
- Date: Sun, 21 Nov 93 13:37:40 +0100
-
- Hi people!
-
- During my "research" for implementing a complete XATTR structure for
- entries in the BIOS filesystem (see mail some days ago), I've now - after
- having some discussion about security problems and so - strumbled about the
- datime() function each device driver has got to provide and I wonder if this
- one could be used to report the time/datestamps needed.
-
- So far, it does not support all m/a-time/date fields, but I think this
- could perhaps be easily fixed by the following idea:
-
- Most device drivers just seem to test if the rwflag is non-zero to check
- if a write call is wanted, although my documentation (german Profibuch) says
- explicitly a write call has to set rwflag=1. If one could rely on this fact,
- it would be easy to implement modes 2-3 to report the m/a-t/d fields (c-t/d
- is unlikely to change :-), if not, you'd have to face the fact that the
- device driver would interpret this as a write call.
-
- This may not be too much disturbing, since this call would be used only for
- internal kernel use and would only be done for biosfs entries. So if you use
- a new-style driver everything's fine, and if your driver doesn't understand
- this it either reports an error - in which case you still can use the current
- time just like the biosfs does up to now - or it would set its values to the
- ones given, which mightn't do any harm if the caller preset's them with the
- current time fields. The device would than act exactly like if it was running
- on an older kernel which doesn't know about all these.
-
- This would only involve changes in the biosfs.c code and would _not_ offer
- the driver the possibility to manipulate its own XATTR structure in order to
- do strange things like gaining superuper priority.
-
- Does this sound like a good idea? It _seems_ to run ok on my setup.
-
- so long,
- TeSche
- ---
- PS: If the above written looks weird, then that's probably because it _is_.
- WhoDunnIt: Torsten Scherer (Schiller, Tesche, ...)
- Where: Faculty of Technology, University of Bielefeld, Germany
- EMail: itschere@techfak.uni-bielefeld.de / tesche@hrz.uni-bielefeld.de
-