Re: [PATCH 1/4] Cache xattr security drop check for write v2

[Steven Whitehouse <swhiteho@redhat.com>]

Hi,

On Tue, 2011-05-31 at 11:06 -0700, Andi Kleen wrote:
> > It sounds like a good idea, but cluster filesystems will need to clear
> > the flag when they update their in-core inodes. Without that we could
> > have:
> > 
> >  Node A looks up inode and sets S_NOSEC since its not suid
> >  Node B does chmod +s on the inode
> >  Node A now has S_NOSEC set, but inode is suid, so writes don't clear
> > suid
> 
> Good point. I assume that's also true for network file systems.
> 
> This would essentially argue that for those putting the helper
> into the inode read paths is not optional. I'll look into this
> later.
> 
> > -	flags &= ~(S_SYNC|S_APPEND|S_IMMUTABLE|S_NOATIME|S_DIRSYNC);
> > +	flags &= ~(S_SYNC|S_APPEND|S_IMMUTABLE|S_NOATIME|S_DIRSYNC|S_NOSEC);
> > +	if (!is_sxid(inode->i_mode))
> > +		flags |= S_NOSEC;
> 
> Doesn't that need a check for no xattr too? or do you not support
> those currently?
> 
> Note I added a helper for this in the latest version:
> inode_has_no_xattr()
> 
> -Andi

Yes, it should test for xattr too,

Steve.