Re: [PATCH -V3 3/8] hugetlb: add charge/uncharge calls for HugeTLB alloc/free

[Aneesh Kumar K.V]

On Tue, 13 Mar 2012 21:20:21 +0800, Hillf Danton <dhillf@xxxxxxxxx> wrote:
> On Tue, Mar 13, 2012 at 3:07 PM, Aneesh Kumar K.V
> <aneesh.kumar@xxxxxxxxxxxxxxxxxx> wrote:
> > From: "Aneesh Kumar K.V" <aneesh.kumar@xxxxxxxxxxxxxxxxxx>
> >
> > This adds necessary charge/uncharge calls in the HugeTLB code
> >
> > Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@xxxxxxxxxxxxxxxxxx>
> > ---
> > Âmm/hugetlb.c  Â|  21 ++++++++++++++++++++-
> > Âmm/memcontrol.c | Â Â5 +++++
> > Â2 files changed, 25 insertions(+), 1 deletions(-)
> >
> > diff --git a/mm/hugetlb.c b/mm/hugetlb.c
> > index fe7aefd..b7152d1 100644
> > --- a/mm/hugetlb.c
> > +++ b/mm/hugetlb.c
> > @@ -21,6 +21,8 @@
> > Â#include <linux/rmap.h>
> > Â#include <linux/swap.h>
> > Â#include <linux/swapops.h>
> > +#include <linux/memcontrol.h>
> > +#include <linux/page_cgroup.h>
> >
> > Â#include <asm/page.h>
> > Â#include <asm/pgtable.h>
> > @@ -542,6 +544,9 @@ static void free_huge_page(struct page *page)
> > Â Â Â ÂBUG_ON(page_mapcount(page));
> > Â Â Â ÂINIT_LIST_HEAD(&page->lru);
> >
> > + Â Â Â if (mapping)
> > + Â Â Â Â Â Â Â mem_cgroup_hugetlb_uncharge_page(h - hstates,
> > + Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Âpages_per_huge_page(h), page);
> > Â Â Â Âspin_lock(&hugetlb_lock);
> > Â Â Â Âif (h->surplus_huge_pages_node[nid] && huge_page_order(h) < MAX_ORDER) {
> > Â Â Â Â Â Â Â Âupdate_and_free_page(h, page);
> > @@ -1019,12 +1024,15 @@ static void vma_commit_reservation(struct hstate *h,
> > Âstatic struct page *alloc_huge_page(struct vm_area_struct *vma,
> > Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Âunsigned long addr, int avoid_reserve)
> > Â{
> > + Â Â Â int ret, idx;
> > Â Â Â Âstruct hstate *h = hstate_vma(vma);
> > Â Â Â Âstruct page *page;
> > + Â Â Â struct mem_cgroup *memcg = NULL;
> > Â Â Â Âstruct address_space *mapping = vma->vm_file->f_mapping;
> > Â Â Â Âstruct inode *inode = mapping->host;
> > Â Â Â Âlong chg;
> >
> > + Â Â Â idx = h - hstates;
> 
> Better if hstate index is computed with a tiny inline helper?
> Other than that,

Will update in the next iteration.

> 
> Acked-by: Hillf Danton <dhillf@xxxxxxxxx>
> 

-aneesh

--
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/