Re: [PATCH] exec: remove page_table_lock for mm counters in acct_arg_size()

[Frantisek Hrbata <fhrbata@redhat.com>]

On Thu, Jul 14, 2011 at 03:18:22PM +0200, Oleg Nesterov wrote:
> On 07/14, Frantisek Hrbata wrote:
> >
> > Commit 172703b mm: delete non-atomic mm counter implementation
> > removed non-atomic mm counters implementation. I think there is no need to hold
> > page_table_lock for add_mm_counter() when the splitted page table locks are not
> > used. This patch removes the locking.
> >
> > Signed-off-by: Frantisek Hrbata <fhrbata@redhat.com>
> > ---
> >  fs/exec.c |    6 ------
> >  1 files changed, 0 insertions(+), 6 deletions(-)
> >
> > diff --git a/fs/exec.c b/fs/exec.c
> > index 6075a1e..e6fc5e1 100644
> > --- a/fs/exec.c
> > +++ b/fs/exec.c
> > @@ -182,13 +182,7 @@ static void acct_arg_size(struct linux_binprm *bprm, unsigned long pages)
> >
> >  	bprm->vma_pages = pages;
> >
> > -#ifdef SPLIT_RSS_COUNTING
> >  	add_mm_counter(mm, MM_ANONPAGES, diff);
> > -#else
> > -	spin_lock(&mm->page_table_lock);
> > -	add_mm_counter(mm, MM_ANONPAGES, diff);
> > -	spin_unlock(&mm->page_table_lock);
> > -#endif
> 
> Heh. Thanks Frantisek, you are right. But this patch is already in mm-tree,
> fs-execc-acct_arg_size-ptl-is-no-longer-needed-for-add_mm_counter.patch
> 
> Oleg.
> 

:) thanks. Sorry I should check that first.

-- 
Frantisek Hrbata