Re: [PATCH -tip] remove the BKL: Replace BKL in mount/umount syscalls with a mutex

[Al Viro <viro@ZenIV.linux.org.uk>]

On Fri, Apr 17, 2009 at 07:04:45PM +0200, Peter Zijlstra wrote:
> On Fri, 2009-04-17 at 18:56 +0200, Ingo Molnar wrote:
> 
> > Stupid question regarding c): wouldnt such data structures go via 
> > the VFS - which you said was free of BKL constraints? Or are there 
> > interconnected private data structures between certain types of 
> > closely related filesystems that the VFS does not know about? (and 
> > hence might have BKL assumptions)
> 
> The VFS is stuffed with ->private like pointers for filesystems to flesh
> out, and I could well imagine some implicit serialization between the
> various (4?) VFS hooks that are currently still under BKL.

Explicit one, TYVM...  Anyway, any fs that dares to use private data of
objects from another filesystem deserves everything it gets; I'm not
aware of any that would do that and I have no sympathy whatsoever to
any that would try.

That's not what I'd been talking about, though; what we want to watch out
for is sharing of data structures by some (or all) fs instances of given
type.  I.e. internals of particular fs driver, not cross-driver problems.