The patch titled
Subject: mm, page_alloc: pass preferred nid instead of zonelist to allocator
has been added to the -mm tree. Its filename is
mm-page_alloc-pass-preferred-nid-instead-of-zonelist-to-allocator.patch
This patch should soon appear at
http://ozlabs.org/~akpm/mmots/broken-out/mm-page_alloc-pass-preferred-nid-instead-of-zonelist-to-allocator.patch
and later at
http://ozlabs.org/~akpm/mmotm/broken-out/mm-page_alloc-pass-preferred-nid-instead-of-zonelist-to-allocator.patch
Before you just go and hit "reply", please:
a) Consider who else should be cc'ed
b) Prefer to cc a suitable mailing list as well
c) Ideally: find the original patch on the mailing list and do a
reply-to-all to that, adding suitable additional cc's
*** Remember to use Documentation/SubmitChecklist when testing your code ***
The -mm tree is included into linux-next and is updated
there every 3-4 working days
------------------------------------------------------
From: Vlastimil Babka <vbabka@xxxxxxx>
Subject: mm, page_alloc: pass preferred nid instead of zonelist to allocator
The main allocator function __alloc_pages_nodemask() takes a zonelist
pointer as one of its parameters. All of its callers directly or
indirectly obtain the zonelist via node_zonelist() using a preferred node
id and gfp_mask. We can make the code a bit simpler by doing the zonelist
lookup in __alloc_pages_nodemask(), passing it a preferred node id instead
(gfp_mask is already another parameter).
There are some code size benefits thanks to removal of inlined
node_zonelist():
bloat-o-meter add/remove: 2/2 grow/shrink: 4/36 up/down: 399/-1351 (-952)
This will also make things simpler if we proceed with converting cpusets
to zonelists.
Link: http://lkml.kernel.org/r/20170517081140.30654-4-vbabka@xxxxxxx
Signed-off-by: Vlastimil Babka <vbabka@xxxxxxx>
Reviewed-by: Christoph Lameter <cl@xxxxxxxxx>
Acked-by: Michal Hocko <mhocko@xxxxxxxx>
Cc: Dimitri Sivanich <sivanich@xxxxxxx>
Cc: "Kirill A. Shutemov" <kirill.shutemov@xxxxxxxxxxxxxxx>
Cc: Andrea Arcangeli <aarcange@xxxxxxxxxx>
Cc: Anshuman Khandual <khandual@xxxxxxxxxxxxxxxxxx>
Cc: David Rientjes <rientjes@xxxxxxxxxx>
Cc: Hugh Dickins <hughd@xxxxxxxxxx>
Cc: Li Zefan <lizefan@xxxxxxxxxx>
Cc: Mel Gorman <mgorman@xxxxxxxxxxxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
include/linux/gfp.h | 11 ++++-----
include/linux/mempolicy.h | 6 ++---
mm/hugetlb.c | 15 +++++++------
mm/memory_hotplug.c | 6 +----
mm/mempolicy.c | 41 ++++++++++++++++--------------------
mm/page_alloc.c | 10 ++++----
6 files changed, 43 insertions(+), 46 deletions(-)
diff -puN include/linux/gfp.h~mm-page_alloc-pass-preferred-nid-instead-of-zonelist-to-allocator include/linux/gfp.h
--- a/include/linux/gfp.h~mm-page_alloc-pass-preferred-nid-instead-of-zonelist-to-allocator
+++ a/include/linux/gfp.h
@@ -432,14 +432,13 @@ static inline void arch_alloc_page(struc
#endif
struct page *
-__alloc_pages_nodemask(gfp_t gfp_mask, unsigned int order,
- struct zonelist *zonelist, nodemask_t *nodemask);
+__alloc_pages_nodemask(gfp_t gfp_mask, unsigned int order, int preferred_nid,
+ nodemask_t *nodemask);
static inline struct page *
-__alloc_pages(gfp_t gfp_mask, unsigned int order,
- struct zonelist *zonelist)
+__alloc_pages(gfp_t gfp_mask, unsigned int order, int preferred_nid)
{
- return __alloc_pages_nodemask(gfp_mask, order, zonelist, NULL);
+ return __alloc_pages_nodemask(gfp_mask, order, preferred_nid, NULL);
}
/*
@@ -452,7 +451,7 @@ __alloc_pages_node(int nid, gfp_t gfp_ma
VM_BUG_ON(nid < 0 || nid >= MAX_NUMNODES);
VM_WARN_ON(!node_online(nid));
- return __alloc_pages(gfp_mask, order, node_zonelist(nid, gfp_mask));
+ return __alloc_pages(gfp_mask, order, nid);
}
/*
diff -puN include/linux/mempolicy.h~mm-page_alloc-pass-preferred-nid-instead-of-zonelist-to-allocator include/linux/mempolicy.h
--- a/include/linux/mempolicy.h~mm-page_alloc-pass-preferred-nid-instead-of-zonelist-to-allocator
+++ a/include/linux/mempolicy.h
@@ -146,7 +146,7 @@ extern void mpol_rebind_task(struct task
enum mpol_rebind_step step);
extern void mpol_rebind_mm(struct mm_struct *mm, nodemask_t *new);
-extern struct zonelist *huge_zonelist(struct vm_area_struct *vma,
+extern int huge_node(struct vm_area_struct *vma,
unsigned long addr, gfp_t gfp_flags,
struct mempolicy **mpol, nodemask_t **nodemask);
extern bool init_nodemask_of_mempolicy(nodemask_t *mask);
@@ -269,13 +269,13 @@ static inline void mpol_rebind_mm(struct
{
}
-static inline struct zonelist *huge_zonelist(struct vm_area_struct *vma,
+static inline int huge_node(struct vm_area_struct *vma,
unsigned long addr, gfp_t gfp_flags,
struct mempolicy **mpol, nodemask_t **nodemask)
{
*mpol = NULL;
*nodemask = NULL;
- return node_zonelist(0, gfp_flags);
+ return 0;
}
static inline bool init_nodemask_of_mempolicy(nodemask_t *m)
diff -puN mm/hugetlb.c~mm-page_alloc-pass-preferred-nid-instead-of-zonelist-to-allocator mm/hugetlb.c
--- a/mm/hugetlb.c~mm-page_alloc-pass-preferred-nid-instead-of-zonelist-to-allocator
+++ a/mm/hugetlb.c
@@ -920,6 +920,8 @@ static struct page *dequeue_huge_page_vm
struct page *page = NULL;
struct mempolicy *mpol;
nodemask_t *nodemask;
+ gfp_t gfp_mask;
+ int nid;
struct zonelist *zonelist;
struct zone *zone;
struct zoneref *z;
@@ -940,12 +942,13 @@ static struct page *dequeue_huge_page_vm
retry_cpuset:
cpuset_mems_cookie = read_mems_allowed_begin();
- zonelist = huge_zonelist(vma, address,
- htlb_alloc_mask(h), &mpol, &nodemask);
+ gfp_mask = htlb_alloc_mask(h);
+ nid = huge_node(vma, address, gfp_mask, &mpol, &nodemask);
+ zonelist = node_zonelist(nid, gfp_mask);
for_each_zone_zonelist_nodemask(zone, z, zonelist,
MAX_NR_ZONES - 1, nodemask) {
- if (cpuset_zone_allowed(zone, htlb_alloc_mask(h))) {
+ if (cpuset_zone_allowed(zone, gfp_mask)) {
page = dequeue_huge_page_node(h, zone_to_nid(zone));
if (page) {
if (avoid_reserve)
@@ -1558,13 +1561,13 @@ static struct page *__hugetlb_alloc_budd
do {
struct page *page;
struct mempolicy *mpol;
- struct zonelist *zl;
+ int nid;
nodemask_t *nodemask;
cpuset_mems_cookie = read_mems_allowed_begin();
- zl = huge_zonelist(vma, addr, gfp, &mpol, &nodemask);
+ nid = huge_node(vma, addr, gfp, &mpol, &nodemask);
mpol_cond_put(mpol);
- page = __alloc_pages_nodemask(gfp, order, zl, nodemask);
+ page = __alloc_pages_nodemask(gfp, order, nid, nodemask);
if (page)
return page;
} while (read_mems_allowed_retry(cpuset_mems_cookie));
diff -puN mm/memory_hotplug.c~mm-page_alloc-pass-preferred-nid-instead-of-zonelist-to-allocator mm/memory_hotplug.c
--- a/mm/memory_hotplug.c~mm-page_alloc-pass-preferred-nid-instead-of-zonelist-to-allocator
+++ a/mm/memory_hotplug.c
@@ -1446,11 +1446,9 @@ static struct page *new_node_page(struct
gfp_mask |= __GFP_HIGHMEM;
if (!nodes_empty(nmask))
- new_page = __alloc_pages_nodemask(gfp_mask, 0,
- node_zonelist(nid, gfp_mask), &nmask);
+ new_page = __alloc_pages_nodemask(gfp_mask, 0, nid, &nmask);
if (!new_page)
- new_page = __alloc_pages(gfp_mask, 0,
- node_zonelist(nid, gfp_mask));
+ new_page = __alloc_pages(gfp_mask, 0, nid);
return new_page;
}
diff -puN mm/mempolicy.c~mm-page_alloc-pass-preferred-nid-instead-of-zonelist-to-allocator mm/mempolicy.c
--- a/mm/mempolicy.c~mm-page_alloc-pass-preferred-nid-instead-of-zonelist-to-allocator
+++ a/mm/mempolicy.c
@@ -1669,9 +1669,9 @@ static nodemask_t *policy_nodemask(gfp_t
return NULL;
}
-/* Return a zonelist indicated by gfp for node representing a mempolicy */
-static struct zonelist *policy_zonelist(gfp_t gfp, struct mempolicy *policy,
- int nd)
+/* Return the node id preferred by the given mempolicy, or the given id */
+static int policy_node(gfp_t gfp, struct mempolicy *policy,
+ int nd)
{
if (policy->mode == MPOL_PREFERRED && !(policy->flags & MPOL_F_LOCAL))
nd = policy->v.preferred_node;
@@ -1684,7 +1684,7 @@ static struct zonelist *policy_zonelist(
WARN_ON_ONCE(policy->mode == MPOL_BIND && (gfp & __GFP_THISNODE));
}
- return node_zonelist(nd, gfp);
+ return nd;
}
/* Do dynamic interleaving for a process */
@@ -1791,38 +1791,37 @@ static inline unsigned interleave_nid(st
#ifdef CONFIG_HUGETLBFS
/*
- * huge_zonelist(@vma, @addr, @gfp_flags, @mpol)
+ * huge_node(@vma, @addr, @gfp_flags, @mpol)
* @vma: virtual memory area whose policy is sought
* @addr: address in @vma for shared policy lookup and interleave policy
* @gfp_flags: for requested zone
* @mpol: pointer to mempolicy pointer for reference counted mempolicy
* @nodemask: pointer to nodemask pointer for MPOL_BIND nodemask
*
- * Returns a zonelist suitable for a huge page allocation and a pointer
+ * Returns a nid suitable for a huge page allocation and a pointer
* to the struct mempolicy for conditional unref after allocation.
* If the effective policy is 'BIND, returns a pointer to the mempolicy's
* @nodemask for filtering the zonelist.
*
* Must be protected by read_mems_allowed_begin()
*/
-struct zonelist *huge_zonelist(struct vm_area_struct *vma, unsigned long addr,
- gfp_t gfp_flags, struct mempolicy **mpol,
- nodemask_t **nodemask)
+int huge_node(struct vm_area_struct *vma, unsigned long addr, gfp_t gfp_flags,
+ struct mempolicy **mpol, nodemask_t **nodemask)
{
- struct zonelist *zl;
+ int nid;
*mpol = get_vma_policy(vma, addr);
*nodemask = NULL; /* assume !MPOL_BIND */
if (unlikely((*mpol)->mode == MPOL_INTERLEAVE)) {
- zl = node_zonelist(interleave_nid(*mpol, vma, addr,
- huge_page_shift(hstate_vma(vma))), gfp_flags);
+ nid = interleave_nid(*mpol, vma, addr,
+ huge_page_shift(hstate_vma(vma)));
} else {
- zl = policy_zonelist(gfp_flags, *mpol, numa_node_id());
+ nid = policy_node(gfp_flags, *mpol, numa_node_id());
if ((*mpol)->mode == MPOL_BIND)
*nodemask = &(*mpol)->v.nodes;
}
- return zl;
+ return nid;
}
/*
@@ -1924,12 +1923,10 @@ out:
static struct page *alloc_page_interleave(gfp_t gfp, unsigned order,
unsigned nid)
{
- struct zonelist *zl;
struct page *page;
- zl = node_zonelist(nid, gfp);
- page = __alloc_pages(gfp, order, zl);
- if (page && page_zone(page) == zonelist_zone(&zl->_zonerefs[0]))
+ page = __alloc_pages(gfp, order, nid);
+ if (page && page_to_nid(page) == nid)
inc_zone_page_state(page, NUMA_INTERLEAVE_HIT);
return page;
}
@@ -1963,8 +1960,8 @@ alloc_pages_vma(gfp_t gfp, int order, st
{
struct mempolicy *pol;
struct page *page;
+ int preferred_nid;
unsigned int cpuset_mems_cookie;
- struct zonelist *zl;
nodemask_t *nmask;
retry_cpuset:
@@ -2007,8 +2004,8 @@ retry_cpuset:
}
nmask = policy_nodemask(gfp, pol);
- zl = policy_zonelist(gfp, pol, node);
- page = __alloc_pages_nodemask(gfp, order, zl, nmask);
+ preferred_nid = policy_node(gfp, pol, node);
+ page = __alloc_pages_nodemask(gfp, order, preferred_nid, nmask);
mpol_cond_put(pol);
out:
if (unlikely(!page && read_mems_allowed_retry(cpuset_mems_cookie)))
@@ -2055,7 +2052,7 @@ retry_cpuset:
page = alloc_page_interleave(gfp, order, interleave_nodes(pol));
else
page = __alloc_pages_nodemask(gfp, order,
- policy_zonelist(gfp, pol, numa_node_id()),
+ policy_node(gfp, pol, numa_node_id()),
policy_nodemask(gfp, pol));
if (unlikely(!page && read_mems_allowed_retry(cpuset_mems_cookie)))
diff -puN mm/page_alloc.c~mm-page_alloc-pass-preferred-nid-instead-of-zonelist-to-allocator mm/page_alloc.c
--- a/mm/page_alloc.c~mm-page_alloc-pass-preferred-nid-instead-of-zonelist-to-allocator
+++ a/mm/page_alloc.c
@@ -3969,12 +3969,12 @@ got_pg:
}
static inline bool prepare_alloc_pages(gfp_t gfp_mask, unsigned int order,
- struct zonelist *zonelist, nodemask_t *nodemask,
+ int preferred_nid, nodemask_t *nodemask,
struct alloc_context *ac, gfp_t *alloc_mask,
unsigned int *alloc_flags)
{
ac->high_zoneidx = gfp_zone(gfp_mask);
- ac->zonelist = zonelist;
+ ac->zonelist = node_zonelist(preferred_nid, gfp_mask);
ac->nodemask = nodemask;
ac->migratetype = gfpflags_to_migratetype(gfp_mask);
@@ -4019,8 +4019,8 @@ static inline void finalise_ac(gfp_t gfp
* This is the 'heart' of the zoned buddy allocator.
*/
struct page *
-__alloc_pages_nodemask(gfp_t gfp_mask, unsigned int order,
- struct zonelist *zonelist, nodemask_t *nodemask)
+__alloc_pages_nodemask(gfp_t gfp_mask, unsigned int order, int preferred_nid,
+ nodemask_t *nodemask)
{
struct page *page;
unsigned int alloc_flags = ALLOC_WMARK_LOW;
@@ -4028,7 +4028,7 @@ __alloc_pages_nodemask(gfp_t gfp_mask, u
struct alloc_context ac = { };
gfp_mask &= gfp_allowed_mask;
- if (!prepare_alloc_pages(gfp_mask, order, zonelist, nodemask, &ac, &alloc_mask, &alloc_flags))
+ if (!prepare_alloc_pages(gfp_mask, order, preferred_nid, nodemask, &ac, &alloc_mask, &alloc_flags))
return NULL;
finalise_ac(gfp_mask, order, &ac);
_
Patches currently in -mm which might be from vbabka@xxxxxxx are
mm-page_alloc-fix-more-premature-oom-due-to-race-with-cpuset-update.patch
mm-mempolicy-stop-adjusting-current-il_next-in-mpol_rebind_nodemask.patch
mm-page_alloc-pass-preferred-nid-instead-of-zonelist-to-allocator.patch
mm-mempolicy-simplify-rebinding-mempolicies-when-updating-cpusets.patch
mm-cpuset-always-use-seqlock-when-changing-tasks-nodemask.patch
mm-mempolicy-dont-check-cpuset-seqlock-where-it-doesnt-matter.patch
--
To unsubscribe from this list: send the line "unsubscribe mm-commits" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
