Re: [PATCHv5 01/28] mm, proc: adjust PSS calculation

From: Kirill A. Shutemov
Date: Fri May 15 2015 - 06:57:00 EST


On Thu, May 14, 2015 at 04:12:29PM +0200, Vlastimil Babka wrote:
> On 04/23/2015 11:03 PM, Kirill A. Shutemov wrote:
> >With new refcounting all subpages of the compound page are not nessessary
> >have the same mapcount. We need to take into account mapcount of every
> >sub-page.
> >
> >Signed-off-by: Kirill A. Shutemov <kirill.shutemov@xxxxxxxxxxxxxxx>
> >Tested-by: Sasha Levin <sasha.levin@xxxxxxxxxx>
>
> Acked-by: Vlastimil Babka <vbabka@xxxxxxx>
>
> (some nitpicks below)
>
> >---
> > fs/proc/task_mmu.c | 43 ++++++++++++++++++++++---------------------
> > 1 file changed, 22 insertions(+), 21 deletions(-)
> >
> >diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c
> >index 956b75d61809..95bc384ee3f7 100644
> >--- a/fs/proc/task_mmu.c
> >+++ b/fs/proc/task_mmu.c
> >@@ -449,9 +449,10 @@ struct mem_size_stats {
> > };
> >
> > static void smaps_account(struct mem_size_stats *mss, struct page *page,
> >- unsigned long size, bool young, bool dirty)
> >+ bool compound, bool young, bool dirty)
> > {
> >- int mapcount;
> >+ int i, nr = compound ? hpage_nr_pages(page) : 1;
>
> Why not just HPAGE_PMD_NR instead of hpage_nr_pages(page)?

Okay, makes sense. Compiler is smart enough to optimize away HPAGE_PMD_NR
for THP=n. (HPAGE_PMD_NR is BUILD_BUG() for THP=n)

> We already came here through a pmd mapping. Even if the page stopped
> being a hugepage meanwhile (I'm not sure if any locking prevents that or
> not?),

We're under ptl here. PMD will not go away under us.

> it would be more accurate to continue assuming it's a hugepage,
> otherwise we account only the base page (formerly head) and skip the 511
> formerly tail pages?
>
> Also, is there some shortcut way to tell us that we are the only one mapping
> the whole compound page, and nobody has any base pages, so we don't need to
> loop on each tail page? I guess not under the new design, right...

No, we don't have shortcut here.

> >+ unsigned long size = nr * PAGE_SIZE;
> >
> > if (PageAnon(page))
> > mss->anonymous += size;
> >@@ -460,23 +461,23 @@ static void smaps_account(struct mem_size_stats *mss, struct page *page,
> > /* Accumulate the size in pages that have been accessed. */
> > if (young || PageReferenced(page))
> > mss->referenced += size;
> >- mapcount = page_mapcount(page);
> >- if (mapcount >= 2) {
> >- u64 pss_delta;
> >
> >- if (dirty || PageDirty(page))
> >- mss->shared_dirty += size;
> >- else
> >- mss->shared_clean += size;
> >- pss_delta = (u64)size << PSS_SHIFT;
> >- do_div(pss_delta, mapcount);
> >- mss->pss += pss_delta;
> >- } else {
> >- if (dirty || PageDirty(page))
> >- mss->private_dirty += size;
> >- else
> >- mss->private_clean += size;
> >- mss->pss += (u64)size << PSS_SHIFT;
> >+ for (i = 0; i < nr; i++) {
> >+ int mapcount = page_mapcount(page + i);
> >+
> >+ if (mapcount >= 2) {
> >+ if (dirty || PageDirty(page + i))
> >+ mss->shared_dirty += PAGE_SIZE;
> >+ else
> >+ mss->shared_clean += PAGE_SIZE;
> >+ mss->pss += (PAGE_SIZE << PSS_SHIFT) / mapcount;
> >+ } else {
> >+ if (dirty || PageDirty(page + i))
> >+ mss->private_dirty += PAGE_SIZE;
> >+ else
> >+ mss->private_clean += PAGE_SIZE;
> >+ mss->pss += PAGE_SIZE << PSS_SHIFT;
> >+ }
>
> That's 3 instances of "page + i", why not just use page and do a page++ in
> the for loop?

Okay.

--
Kirill A. Shutemov
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/