| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Date: Tue, 10 Sep 2013 17:53:29 -0400
From: Naoya Horiguchi <n-horiguchi@...jp.nec.com>
To: Mel Gorman <mgorman@...e.de>
Cc: Andrew Morton <akpm@...ux-foundation.org>, linux-mm@...ck.org,
Hugh Dickins <hughd@...gle.com>,
KOSAKI Motohiro <kosaki.motohiro@...fujitsu.com>,
Andi Kleen <andi@...stfloor.org>,
Hillf Danton <dhillf@...il.com>, Michal Hocko <mhocko@...e.cz>,
Rik van Riel <riel@...hat.com>,
"Aneesh Kumar K.V" <aneesh.kumar@...ux.vnet.ibm.com>,
Wanpeng Li <liwanp@...ux.vnet.ibm.com>,
linux-kernel@...r.kernel.org,
Naoya Horiguchi <nao.horiguchi@...il.com>
Subject: Re: [PATCH 5/9] mbind: add hugepage migration code to mbind()
On Tue, Sep 10, 2013 at 03:41:09PM +0100, Mel Gorman wrote:
> On Fri, Aug 09, 2013 at 01:21:38AM -0400, Naoya Horiguchi wrote:
> > This patch extends do_mbind() to handle vma with VM_HUGETLB set.
> > We will be able to migrate hugepage with mbind(2) after
> > applying the enablement patch which comes later in this series.
> >
> > ChangeLog v3:
> > - revert introducing migrate_movable_pages
> > - added alloc_huge_page_noerr free from ERR_VALUE
> >
> > ChangeLog v2:
> > - updated description and renamed patch title
> >
> > Signed-off-by: Naoya Horiguchi <n-horiguchi@...jp.nec.com>
> > Acked-by: Andi Kleen <ak@...ux.intel.com>
> > Reviewed-by: Wanpeng Li <liwanp@...ux.vnet.ibm.com>
> > Acked-by: Hillf Danton <dhillf@...il.com>
> > ---
> > include/linux/hugetlb.h | 3 +++
> > mm/hugetlb.c | 14 ++++++++++++++
> > mm/mempolicy.c | 4 +++-
> > 3 files changed, 20 insertions(+), 1 deletion(-)
> >
> > diff --git v3.11-rc3.orig/include/linux/hugetlb.h v3.11-rc3/include/linux/hugetlb.h
> > index bc8d837..d1db007 100644
> > --- v3.11-rc3.orig/include/linux/hugetlb.h
> > +++ v3.11-rc3/include/linux/hugetlb.h
> > @@ -265,6 +265,8 @@ struct huge_bootmem_page {
> > };
> >
> > struct page *alloc_huge_page_node(struct hstate *h, int nid);
> > +struct page *alloc_huge_page_noerr(struct vm_area_struct *vma,
> > + unsigned long addr, int avoid_reserve);
> >
> > /* arch callback */
> > int __init alloc_bootmem_huge_page(struct hstate *h);
> > @@ -378,6 +380,7 @@ static inline pgoff_t basepage_index(struct page *page)
> > #else /* CONFIG_HUGETLB_PAGE */
> > struct hstate {};
> > #define alloc_huge_page_node(h, nid) NULL
> > +#define alloc_huge_page_noerr(v, a, r) NULL
> > #define alloc_bootmem_huge_page(h) NULL
> > #define hstate_file(f) NULL
> > #define hstate_sizelog(s) NULL
> > diff --git v3.11-rc3.orig/mm/hugetlb.c v3.11-rc3/mm/hugetlb.c
> > index 649771c..ee764b0 100644
> > --- v3.11-rc3.orig/mm/hugetlb.c
> > +++ v3.11-rc3/mm/hugetlb.c
> > @@ -1195,6 +1195,20 @@ static struct page *alloc_huge_page(struct vm_area_struct *vma,
> > return page;
> > }
> >
> > +/*
> > + * alloc_huge_page()'s wrapper which simply returns the page if allocation
> > + * succeeds, otherwise NULL. This function is called from new_vma_page(),
> > + * where no ERR_VALUE is expected to be returned.
> > + */
> > +struct page *alloc_huge_page_noerr(struct vm_area_struct *vma,
> > + unsigned long addr, int avoid_reserve)
> > +{
> > + struct page *page = alloc_huge_page(vma, addr, avoid_reserve);
> > + if (IS_ERR(page))
> > + page = NULL;
> > + return page;
> > +}
> > +
> > int __weak alloc_bootmem_huge_page(struct hstate *h)
> > {
> > struct huge_bootmem_page *m;
> > diff --git v3.11-rc3.orig/mm/mempolicy.c v3.11-rc3/mm/mempolicy.c
> > index d96afc1..4a03c14 100644
> > --- v3.11-rc3.orig/mm/mempolicy.c
> > +++ v3.11-rc3/mm/mempolicy.c
> > @@ -1183,6 +1183,8 @@ static struct page *new_vma_page(struct page *page, unsigned long private, int *
> > vma = vma->vm_next;
> > }
> >
> > + if (PageHuge(page))
> > + return alloc_huge_page_noerr(vma, address, 1);
> > /*
> > * if !vma, alloc_page_vma() will use task or system default policy
> > */
>
> It's interesting to note that it will be tricky to configure a system to
> allow this sort of migration to succeed.
>
> This call correctly uses avoid_reserve but that does mean that for it
> to work that there there must be free pages statically allocated in the
> hugepage pool of the destination node or hugepage dynamic pool resizing
> must be enabled. The former option is going to waste memory because pages
> allocated to the static pool cannot be used for any other purpose and
> using dynamic hugepage pool resizing may fail.
Yes, that's interesting because it's important to make page migration
more likely to succeed. I guess that dynamic pool resizing can affect
the pool configuration without administrators' knowing, so allocating
surplus hugepages directly from buddy seems more preferable.
> It makes me wonder how actually useful generic hugetlbfs page migration
> will be in practice. Are there really usecases where the system
> administrator is willing to create unused hugepage pools on each node
> just to enable migration?
Maybe most users don't want it.
Thanks,
Naoya Horiguchi
> > @@ -1293,7 +1295,7 @@ static long do_mbind(unsigned long start, unsigned long len,
> > (unsigned long)vma,
> > MIGRATE_SYNC, MR_MEMPOLICY_MBIND);
> > if (nr_failed)
> > - putback_lru_pages(&pagelist);
> > + putback_movable_pages(&pagelist);
> > }
> >
> > if (nr_failed && (flags & MPOL_MF_STRICT))
> > --
> > 1.8.3.1
> >
>
> --
> Mel Gorman
> SUSE Labs
>
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists