Matthieu Moy wrote:

> I didn't find a satisfying solution to this problem, but at least in
> my precise case, automatically removing the revlib entries for which
> inode signature fails would have been a solution (a better solution
> for this would have been a second signature, based on a hash, that
> would survive backup/restore and rebuild the inode signature on
> demand).
> 
In my case, the corporate box I am using has a cron job pruning old
files in /tmp, where I have to use for my revlib. 
I really don't think "intrusion detection" or "forensic clue" should be
the concern here. inode signatures are very weak in term of security
measures; if people can mess with your revlib you are doomed anyway.
Printing a warning message is reasonable though.
Regarding the atomic operation concerns, are you sure tla currently
support concurrent access to the revlib? 
 
> Matthieu
Derek


_______________________________________________
Gnu-arch-users mailing list
[email protected]
http://lists.gnu.org/mailman/listinfo/gnu-arch-users

GNU arch home page:
http://savannah.gnu.org/projects/gnu-arch/

Reply via email to