Re: [RFC] situation with fput() locking (was Re: [PULL REQUEST] : ima-appraisal patches)

[Linus Torvalds <torvalds@linux-foundation.org>]

On Fri, Apr 20, 2012 at 12:58 PM, Al Viro <viro@zeniv.linux.org.uk> wrote:
>
> So does a bunch of other places.  Let me dig out the call graph circa
> 3.3.0...  Here is the relevant part:

Yes, but a lot of those would actually be helped by a helper function
that does all of:
 - grab mmap_sem
 - call do_m[un]map()
 - release mmap_sem

and that would actually clean them up even in the current case.

And then we could do the cleanup in just the helper function.

Not all, no. But a preparatory patch that just creates the helper
functions for doing brk/mmap/munmap would get rid of a fairly big
chunk of them.

You can visualize how many of them do that by just doing

    git grep -5 do_m[un]*map

and then high-lghting '_write(' (to visually show the
down_write/up_write pairs that surround most of them) by searching for
it.

Are they all like that? No. But most of the ones outside of mm/ do fit
that simple pattern and should probably be fixed up just to have them
not contain VM locking details in them *anyway*.

                    Linus