Re: [PATCH] vfs: fix IMA lockdep circular locking dependency

[Mimi Zohar <zohar@linux.vnet.ibm.com>]

On Thu, 2012-05-31 at 01:56 +0100, Al Viro wrote:
> On Wed, May 30, 2012 at 05:40:54PM -0700, Linus Torvalds wrote:
> > On Wed, May 30, 2012 at 5:28 PM, Al Viro <viro@zeniv.linux.org.uk> wrote:
> > >
> > > FWIW, I think it's cleaner to take the whole thing into an inlined helper.
> > 
> > Even better.
> > 
> > I notice that your inlined helper doesn't do what I did: if PROT_EXEC
> > is already set, stop all the stupid games. IOW, the first test in that
> > function could as well be
> > 
> >     if (prot & (PROT_READ | PROT_EXEC) != PROT_READ)
> >         return prot;
> > 
> > because if PROT_EXEC is already set, or if PROT_READ wasn't set, none
> > of the rest of the checks make any sense at all.
> 
> Point...  OK, done, pushed and the whole thing thrown into #for-next.

Patch "split cap_mmap_addr() out of cap_file_mmap()" contains the typo
'cat_mmap_addr', which is then removed in
"split ->file_mmap() into ->mmap_addr()/->mmap_file()".

> Probably too late for today's linux-next, but tomorrow one should pick
> that.  BTW, tomorrow there'll be a signal.git pull request as well,
> with task_work_add() series in it.
> 
> Now for the Miklos' stuff...
>