[PATCH] ARM: LPAE: Invalidate the TLB for module addresses during translation fault

[Catalin Marinas]

On Fri, Feb 24, 2012 at 10:12:00AM +0000, Russell King - ARM Linux wrote:
> On Fri, Feb 24, 2012 at 09:58:06AM +0000, Catalin Marinas wrote:
> > On Thu, Feb 23, 2012 at 10:13:43PM +0000, Catalin Marinas wrote:
> > > On Thu, Feb 23, 2012 at 07:50:03PM +0000, Russell King - ARM Linux wrote:
> > > > On Thu, Feb 23, 2012 at 05:59:41PM +0000, Catalin Marinas wrote:
> > > > > During the free_pgtables() call all user and modules/pkmap entries are
> > > > > removed.
> > > > 
> > > > Err, no.  free_pgtables() should only ever touch up to TASK_SIZE, which
> > > > is the maximum size of userspace.  Moreover, because TASK_SIZE does not
> > > > cover an entire pud, free_pgtables() should not be removing the pud
> > > > table.
> > > 
> > > Are you sure? exit_mmap() calls free_pgtables() with ceiling == 0.  If
> > > free_pmd_range() is called for a range in the top 1GB of the task
> > > address space (which includes modules), it also calls pud_clear()
> > > because ceiling is 0.
> > 
> > BTW, an alternative patch but which needs wider acknowledgement as it
> > touches generic code (I can post it to linux-mm):
> 
> I think that's the right solution.

Something similar is needed in shift_arg_pages() when calling
free_pgd_range().

But my concern - are there other architectures that rely on
free_pgtables() to remove page tables corresponding to vmas beyond
TASK_SIZE (e.g. vsyscall)?

-- 
Catalin