On Tue, 21 Nov 2000 12:47:43 +0200 (SAST), Alan Barrett wrote:

>  If we are thinking of new syscalls, then I think that
>  open_if_stat_matches() would be more useful than open_as().  It
>  would be used like this: stat(path, statbuf); check attributes in
>  statbuf; open_if_stat_matches(path, statbuf, flags, mode).  There
>  is no race vulnerability between stat() and open_if_stat_matches()
>  because the kernel would refuse to open the file if the results from
>  fstat() after the open would be different from the results of stat()
>  before the open.  There should probably be a way of saying that
>  certain fields in the stat structure (such as access time) are not
>  important.
>  
>  Note that an open_as() library function could be implemented using
>  an open_if_stat_matches() syscall.
>  
>  --apb (Alan Barrett)

I fully agree with you - indeed, I had the same idea, and had gone from the
stage of using an md function on the contents of a file and passing that to
the open(2), to the stage of expanding that idea to use a stat struct to
capture the info. I suspect the st_atime field would have to be ignored by
the open_if_stat_matches(2) call (:-)), and you could also zero out parts of
the stat struct that you didn't want compared, or use a separate stat
structre as a bit mask for relevant fields to check.

Regards,
Alistair


--
Alistair Crooks (agc@pkgsrc.org)





_______________________________________________________
Tired of slow Internet? Get @Home Broadband Internet
http://www.home.com/xinbox/signup.html