mempolicy: alloc_pages_mpol() for NUMA policy without vma

Shrink shmem's stack usage by eliminating the pseudo-vma from its folio
allocation.  alloc_pages_mpol(gfp, order, pol, ilx, nid) becomes the
principal actor for passing mempolicy choice down to __alloc_pages(),
rather than vma_alloc_folio(gfp, order, vma, addr, hugepage).

vma_alloc_folio() and alloc_pages() remain, but as wrappers around
alloc_pages_mpol().  alloc_pages_bulk_*() untouched, except to provide the
additional args to policy_nodemask(), which subsumes policy_node(). 
Cleanup throughout, cutting out some unhelpful "helpers".

It would all be much simpler without MPOL_INTERLEAVE, but that adds a
dynamic to the constant mpol: complicated by v3.6 commit 09c231cb
("tmpfs: distribute interleave better across nodes"), which added ino bias
to the interleave, hidden from mm/mempolicy.c until this commit.

Hence "ilx" throughout, the "interleave index".  Originally I thought it
could be done just with nid, but that's wrong: the nodemask may come from
the shared policy layer below a shmem vma, or it may come from the task
layer above a shmem vma; and without the final nodemask then nodeid cannot
be decided.  And how ilx is applied depends also on page order.

The interleave index is almost always irrelevant unless MPOL_INTERLEAVE:
with one exception in alloc_pages_mpol(), where the NO_INTERLEAVE_INDEX
passed down from vma-less alloc_pages() is also used as hint not to use
THP-style hugepage allocation - to avoid the overhead of a hugepage arg
(though I don't understand why we never just added a GFP bit for THP - if
it actually needs a different allocation strategy from other pages of the
same order).  vma_alloc_folio() still carries its hugepage arg here, but
it is not used, and should be removed when agreed.

get_vma_policy() no longer allows a NULL vma: over time I believe we've
eradicated all the places which used to need it e.g.  swapoff and madvise
used to pass NULL vma to read_swap_cache_async(), but now know the vma.

[hughd@google.com: handle NULL mpol being passed to __read_swap_cache_async()]
  Link: https://lkml.kernel.org/r/ea419956-4751-0102-21f7-9c93cb957892@google.com
Link: https://lkml.kernel.org/r/74e34633-6060-f5e3-aee-7040d43f2e93@google.com
Link: https://lkml.kernel.org/r/1738368e-bac0-fd11-ed7f-b87142a939fe@google.comSigned-off-by: Hugh Dickins <hughd@google.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Christoph Lameter <cl@linux.com>
Cc: David Hildenbrand <david@redhat.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Huang Ying <ying.huang@intel.com>
Cc: Kefeng Wang <wangkefeng.wang@huawei.com>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Mel Gorman <mgorman@techsingularity.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Nhat Pham <nphamcs@gmail.com>
Cc: Sidhartha Kumar <sidhartha.kumar@oracle.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Tejun heo <tj@kernel.org>
Cc: Vishal Moola (Oracle) <vishal.moola@gmail.com>
Cc: Yang Shi <shy828301@gmail.com>
Cc: Yosry Ahmed <yosryahmed@google.com>
Cc: Domenico Cerasuolo <mimmocerasuolo@gmail.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

fs/proc/task_mmu.c

@@ -2673,8 +2673,9 @@ static int show_numa_map(struct seq_file *m, void *v)
 	struct numa_maps *md = &numa_priv->md;
 	struct file *file = vma->vm_file;
 	struct mm_struct *mm = vma->vm_mm;
-	struct mempolicy *pol;
 	char buffer[64];
+	struct mempolicy *pol;
+	pgoff_t ilx;
 	int nid;
 
 	if (!mm)
@@ -2683,7 +2684,7 @@ static int show_numa_map(struct seq_file *m, void *v)
 	/* Ensure we start with an empty set of numa_maps statistics. */
 	memset(md, 0, sizeof(*md));
 
-	pol = __get_vma_policy(vma, vma->vm_start);
+	pol = __get_vma_policy(vma, vma->vm_start, &ilx);
 	if (pol) {
 		mpol_to_str(buffer, sizeof(buffer), pol);
 		mpol_cond_put(pol);

include/linux/gfp.h

@@ -8,6 +8,7 @@
 #include <linux/topology.h>
 
 struct vm_area_struct;
+struct mempolicy;
 
 /* Convert GFP flags to their corresponding migrate type */
 #define GFP_MOVABLE_MASK (__GFP_RECLAIMABLE|__GFP_MOVABLE)
@@ -262,7 +263,9 @@ static inline struct page *alloc_pages_node(int nid, gfp_t gfp_mask,
 
 #ifdef CONFIG_NUMA
 struct page *alloc_pages(gfp_t gfp, unsigned int order);
-struct folio *folio_alloc(gfp_t gfp, unsigned order);
+struct page *alloc_pages_mpol(gfp_t gfp, unsigned int order,
+		struct mempolicy *mpol, pgoff_t ilx, int nid);
+struct folio *folio_alloc(gfp_t gfp, unsigned int order);
 struct folio *vma_alloc_folio(gfp_t gfp, int order, struct vm_area_struct *vma,
 		unsigned long addr, bool hugepage);
 #else
@@ -270,6 +273,11 @@ static inline struct page *alloc_pages(gfp_t gfp_mask, unsigned int order)
 {
 	return alloc_pages_node(numa_node_id(), gfp_mask, order);
 }
+static inline struct page *alloc_pages_mpol(gfp_t gfp, unsigned int order,
+		struct mempolicy *mpol, pgoff_t ilx, int nid)
+{
+	return alloc_pages(gfp, order);
+}
 static inline struct folio *folio_alloc(gfp_t gfp, unsigned int order)
 {
 	return __folio_alloc_node(gfp, order, numa_node_id());

include/linux/mempolicy.h

@@ -17,6 +17,8 @@
 
 struct mm_struct;
 
+#define NO_INTERLEAVE_INDEX (-1UL)	/* use task il_prev for interleaving */
+
 #ifdef CONFIG_NUMA
 
 /*
@@ -126,7 +128,9 @@ struct mempolicy *mpol_shared_policy_lookup(struct shared_policy *sp,
 
 struct mempolicy *get_task_policy(struct task_struct *p);
 struct mempolicy *__get_vma_policy(struct vm_area_struct *vma,
-		unsigned long addr);
+		unsigned long addr, pgoff_t *ilx);
+struct mempolicy *get_vma_policy(struct vm_area_struct *vma,
+		unsigned long addr, int order, pgoff_t *ilx);
 bool vma_policy_mof(struct vm_area_struct *vma);
 
 extern void numa_default_policy(void);
@@ -140,8 +144,6 @@ extern int huge_node(struct vm_area_struct *vma,
 extern bool init_nodemask_of_mempolicy(nodemask_t *mask);
 extern bool mempolicy_in_oom_domain(struct task_struct *tsk,
 				const nodemask_t *mask);
-extern nodemask_t *policy_nodemask(gfp_t gfp, struct mempolicy *policy);
-
 extern unsigned int mempolicy_slab_node(void);
 
 extern enum zone_type policy_zone;
@@ -179,6 +181,11 @@ extern bool apply_policy_zone(struct mempolicy *policy, enum zone_type zone);
 
 struct mempolicy {};
 
+static inline struct mempolicy *get_task_policy(struct task_struct *p)
+{
+	return NULL;
+}
+
 static inline bool mpol_equal(struct mempolicy *a, struct mempolicy *b)
 {
 	return true;
@@ -213,6 +220,13 @@ mpol_shared_policy_lookup(struct shared_policy *sp, pgoff_t idx)
 	return NULL;
 }
 
+static inline struct mempolicy *get_vma_policy(struct vm_area_struct *vma,
+				unsigned long addr, int order, pgoff_t *ilx)
+{
+	*ilx = 0;
+	return NULL;
+}
+
 static inline int
 vma_dup_policy(struct vm_area_struct *src, struct vm_area_struct *dst)
 {

include/linux/mm.h

@@ -619,7 +619,7 @@ struct vm_operations_struct {
 	 * policy.
 	 */
 	struct mempolicy *(*get_policy)(struct vm_area_struct *vma,
-					unsigned long addr);
+					unsigned long addr, pgoff_t *ilx);
 #endif
 	/*
 	 * Called by vm_normal_page() for special PTEs to find the

ipc/shm.c

@@ -562,30 +562,25 @@ static unsigned long shm_pagesize(struct vm_area_struct *vma)
 }
 
 #ifdef CONFIG_NUMA
-static int shm_set_policy(struct vm_area_struct *vma, struct mempolicy *new)
+static int shm_set_policy(struct vm_area_struct *vma, struct mempolicy *mpol)
 {
-	struct file *file = vma->vm_file;
-	struct shm_file_data *sfd = shm_file_data(file);
+	struct shm_file_data *sfd = shm_file_data(vma->vm_file);
 	int err = 0;
 
 	if (sfd->vm_ops->set_policy)
-		err = sfd->vm_ops->set_policy(vma, new);
+		err = sfd->vm_ops->set_policy(vma, mpol);
 	return err;
 }
 
 static struct mempolicy *shm_get_policy(struct vm_area_struct *vma,
-					unsigned long addr)
+					unsigned long addr, pgoff_t *ilx)
 {
-	struct file *file = vma->vm_file;
-	struct shm_file_data *sfd = shm_file_data(file);
-	struct mempolicy *pol = NULL;
+	struct shm_file_data *sfd = shm_file_data(vma->vm_file);
+	struct mempolicy *mpol = vma->vm_policy;
 
 	if (sfd->vm_ops->get_policy)
-		pol = sfd->vm_ops->get_policy(vma, addr);
-	else if (vma->vm_policy)
-		pol = vma->vm_policy;
-
-	return pol;
+		mpol = sfd->vm_ops->get_policy(vma, addr, ilx);
+	return mpol;
 }
 #endif
 

mm/shmem.c

@@ -1544,38 +1544,20 @@ static inline struct mempolicy *shmem_get_sbmpol(struct shmem_sb_info *sbinfo)
 	return NULL;
 }
 #endif /* CONFIG_NUMA && CONFIG_TMPFS */
-#ifndef CONFIG_NUMA
-#define vm_policy vm_private_data
-#endif
-
-static void shmem_pseudo_vma_init(struct vm_area_struct *vma,
-		struct shmem_inode_info *info, pgoff_t index)
-{
-	/* Create a pseudo vma that just contains the policy */
-	vma_init(vma, NULL);
-	/* Bias interleave by inode number to distribute better across nodes */
-	vma->vm_pgoff = index + info->vfs_inode.i_ino;
-	vma->vm_policy = mpol_shared_policy_lookup(&info->policy, index);
-}
 
-static void shmem_pseudo_vma_destroy(struct vm_area_struct *vma)
-{
-	/* Drop reference taken by mpol_shared_policy_lookup() */
-	mpol_cond_put(vma->vm_policy);
-}
+static struct mempolicy *shmem_get_pgoff_policy(struct shmem_inode_info *info,
+			pgoff_t index, unsigned int order, pgoff_t *ilx);
 
-static struct folio *shmem_swapin(swp_entry_t swap, gfp_t gfp,
+static struct folio *shmem_swapin_cluster(swp_entry_t swap, gfp_t gfp,
 			struct shmem_inode_info *info, pgoff_t index)
 {
-	struct vm_area_struct pvma;
+	struct mempolicy *mpol;
+	pgoff_t ilx;
 	struct page *page;
-	struct vm_fault vmf = {
-		.vma = &pvma,
-	};
 
-	shmem_pseudo_vma_init(&pvma, info, index);
-	page = swap_cluster_readahead(swap, gfp, &vmf);
-	shmem_pseudo_vma_destroy(&pvma);
+	mpol = shmem_get_pgoff_policy(info, index, 0, &ilx);
+	page = swap_cluster_readahead(swap, gfp, mpol, ilx);
+	mpol_cond_put(mpol);
 
 	if (!page)
 		return NULL;
@@ -1609,27 +1591,29 @@ static gfp_t limit_gfp_mask(gfp_t huge_gfp, gfp_t limit_gfp)
 static struct folio *shmem_alloc_hugefolio(gfp_t gfp,
 		struct shmem_inode_info *info, pgoff_t index)
 {
-	struct vm_area_struct pvma;
-	struct folio *folio;
+	struct mempolicy *mpol;
+	pgoff_t ilx;
+	struct page *page;
 
-	shmem_pseudo_vma_init(&pvma, info, index);
-	folio = vma_alloc_folio(gfp, HPAGE_PMD_ORDER, &pvma, 0, true);
-	shmem_pseudo_vma_destroy(&pvma);
+	mpol = shmem_get_pgoff_policy(info, index, HPAGE_PMD_ORDER, &ilx);
+	page = alloc_pages_mpol(gfp, HPAGE_PMD_ORDER, mpol, ilx, numa_node_id());
+	mpol_cond_put(mpol);
 
-	return folio;
+	return page_rmappable_folio(page);
 }
 
 static struct folio *shmem_alloc_folio(gfp_t gfp,
 		struct shmem_inode_info *info, pgoff_t index)
 {
-	struct vm_area_struct pvma;
-	struct folio *folio;
+	struct mempolicy *mpol;
+	pgoff_t ilx;
+	struct page *page;
 
-	shmem_pseudo_vma_init(&pvma, info, index);
-	folio = vma_alloc_folio(gfp, 0, &pvma, 0, false);
-	shmem_pseudo_vma_destroy(&pvma);
+	mpol = shmem_get_pgoff_policy(info, index, 0, &ilx);
+	page = alloc_pages_mpol(gfp, 0, mpol, ilx, numa_node_id());
+	mpol_cond_put(mpol);
 
-	return folio;
+	return (struct folio *)page;
 }
 
 static struct folio *shmem_alloc_and_add_folio(gfp_t gfp,
@@ -1883,7 +1867,7 @@ static int shmem_swapin_folio(struct inode *inode, pgoff_t index,
 			count_memcg_event_mm(fault_mm, PGMAJFAULT);
 		}
 		/* Here we actually start the io */
-		folio = shmem_swapin(swap, gfp, info, index);
+		folio = shmem_swapin_cluster(swap, gfp, info, index);
 		if (!folio) {
 			error = -ENOMEM;
 			goto failed;
@@ -2334,15 +2318,41 @@ static int shmem_set_policy(struct vm_area_struct *vma, struct mempolicy *mpol)
 }
 
 static struct mempolicy *shmem_get_policy(struct vm_area_struct *vma,
-					  unsigned long addr)
+					  unsigned long addr, pgoff_t *ilx)
 {
 	struct inode *inode = file_inode(vma->vm_file);
 	pgoff_t index;
 
+	/*
+	 * Bias interleave by inode number to distribute better across nodes;
+	 * but this interface is independent of which page order is used, so
+	 * supplies only that bias, letting caller apply the offset (adjusted
+	 * by page order, as in shmem_get_pgoff_policy() and get_vma_policy()).
+	 */
+	*ilx = inode->i_ino;
 	index = ((addr - vma->vm_start) >> PAGE_SHIFT) + vma->vm_pgoff;
 	return mpol_shared_policy_lookup(&SHMEM_I(inode)->policy, index);
 }
-#endif
+
+static struct mempolicy *shmem_get_pgoff_policy(struct shmem_inode_info *info,
+			pgoff_t index, unsigned int order, pgoff_t *ilx)
+{
+	struct mempolicy *mpol;
+
+	/* Bias interleave by inode number to distribute better across nodes */
+	*ilx = info->vfs_inode.i_ino + (index >> order);
+
+	mpol = mpol_shared_policy_lookup(&info->policy, index);
+	return mpol ? mpol : get_task_policy(current);
+}
+#else
+static struct mempolicy *shmem_get_pgoff_policy(struct shmem_inode_info *info,
+			pgoff_t index, unsigned int order, pgoff_t *ilx)
+{
+	*ilx = 0;
+	return NULL;
+}
+#endif /* CONFIG_NUMA */
 
 int shmem_lock(struct file *file, int lock, struct ucounts *ucounts)
 {

mm/swap.h

@@ -2,6 +2,8 @@
 #ifndef _MM_SWAP_H
 #define _MM_SWAP_H
 
+struct mempolicy;
+
 #ifdef CONFIG_SWAP
 #include <linux/blk_types.h> /* for bio_end_io_t */
 
@@ -48,11 +50,10 @@ struct page *read_swap_cache_async(swp_entry_t entry, gfp_t gfp_mask,
 				   unsigned long addr,
 				   struct swap_iocb **plug);
 struct page *__read_swap_cache_async(swp_entry_t entry, gfp_t gfp_mask,
-				     struct vm_area_struct *vma,
-				     unsigned long addr,
+				     struct mempolicy *mpol, pgoff_t ilx,
 				     bool *new_page_allocated);
 struct page *swap_cluster_readahead(swp_entry_t entry, gfp_t flag,
-				    struct vm_fault *vmf);
+				    struct mempolicy *mpol, pgoff_t ilx);
 struct page *swapin_readahead(swp_entry_t entry, gfp_t flag,
 			      struct vm_fault *vmf);
 
@@ -80,7 +81,7 @@ static inline void show_swap_cache_info(void)
 }
 
 static inline struct page *swap_cluster_readahead(swp_entry_t entry,
-				gfp_t gfp_mask, struct vm_fault *vmf)
+			gfp_t gfp_mask, struct mempolicy *mpol, pgoff_t ilx)
 {
 	return NULL;
 }

mm/swap_state.c

@@ -10,6 +10,7 @@
 #include <linux/mm.h>
 #include <linux/gfp.h>
 #include <linux/kernel_stat.h>
+#include <linux/mempolicy.h>
 #include <linux/swap.h>
 #include <linux/swapops.h>
 #include <linux/init.h>
@@ -410,8 +411,8 @@ struct folio *filemap_get_incore_folio(struct address_space *mapping,
 }
 
 struct page *__read_swap_cache_async(swp_entry_t entry, gfp_t gfp_mask,
-			struct vm_area_struct *vma, unsigned long addr,
-			bool *new_page_allocated)
+				     struct mempolicy *mpol, pgoff_t ilx,
+				     bool *new_page_allocated)
 {
 	struct swap_info_struct *si;
 	struct folio *folio;
@@ -453,7 +454,8 @@ struct page *__read_swap_cache_async(swp_entry_t entry, gfp_t gfp_mask,
 		 * before marking swap_map SWAP_HAS_CACHE, when -EEXIST will
 		 * cause any racers to loop around until we add it to cache.
 		 */
-		folio = vma_alloc_folio(gfp_mask, 0, vma, addr, false);
+		folio = (struct folio *)alloc_pages_mpol(gfp_mask, 0,
+						mpol, ilx, numa_node_id());
 		if (!folio)
                         goto fail_put_swap;
 
@@ -528,14 +530,19 @@ struct page *read_swap_cache_async(swp_entry_t entry, gfp_t gfp_mask,
 				   struct vm_area_struct *vma,
 				   unsigned long addr, struct swap_iocb **plug)
 {
-	bool page_was_allocated;
-	struct page *retpage = __read_swap_cache_async(entry, gfp_mask,
-			vma, addr, &page_was_allocated);
+	bool page_allocated;
+	struct mempolicy *mpol;
+	pgoff_t ilx;
+	struct page *page;
 
-	if (page_was_allocated)
-		swap_readpage(retpage, false, plug);
+	mpol = get_vma_policy(vma, addr, 0, &ilx);
+	page = __read_swap_cache_async(entry, gfp_mask, mpol, ilx,
+					&page_allocated);
+	mpol_cond_put(mpol);
 
-	return retpage;
+	if (page_allocated)
+		swap_readpage(page, false, plug);
+	return page;
 }
 
 static unsigned int __swapin_nr_pages(unsigned long prev_offset,
@@ -603,7 +610,8 @@ static unsigned long swapin_nr_pages(unsigned long offset)
  * swap_cluster_readahead - swap in pages in hope we need them soon
  * @entry: swap entry of this memory
  * @gfp_mask: memory allocation flags
- * @vmf: fault information
+ * @mpol: NUMA memory allocation policy to be applied
+ * @ilx: NUMA interleave index, for use only when MPOL_INTERLEAVE
  *
  * Returns the struct page for entry and addr, after queueing swapin.
  *
@@ -612,13 +620,12 @@ static unsigned long swapin_nr_pages(unsigned long offset)
  * because it doesn't cost us any seek time.  We also make sure to queue
  * the 'original' request together with the readahead ones...
  *
- * This has been extended to use the NUMA policies from the mm triggering
- * the readahead.
- *
- * Caller must hold read mmap_lock if vmf->vma is not NULL.
+ * Note: it is intentional that the same NUMA policy and interleave index
+ * are used for every page of the readahead: neighbouring pages on swap
+ * are fairly likely to have been swapped out from the same node.
  */
 struct page *swap_cluster_readahead(swp_entry_t entry, gfp_t gfp_mask,
-				struct vm_fault *vmf)
+				    struct mempolicy *mpol, pgoff_t ilx)
 {
 	struct page *page;
 	unsigned long entry_offset = swp_offset(entry);
@@ -629,8 +636,6 @@ struct page *swap_cluster_readahead(swp_entry_t entry, gfp_t gfp_mask,
 	struct blk_plug plug;
 	struct swap_iocb *splug = NULL;
 	bool page_allocated;
-	struct vm_area_struct *vma = vmf->vma;
-	unsigned long addr = vmf->address;
 
 	mask = swapin_nr_pages(offset) - 1;
 	if (!mask)
@@ -648,8 +653,8 @@ struct page *swap_cluster_readahead(swp_entry_t entry, gfp_t gfp_mask,
 	for (offset = start_offset; offset <= end_offset ; offset++) {
 		/* Ok, do the async read-ahead now */
 		page = __read_swap_cache_async(
-			swp_entry(swp_type(entry), offset),
-			gfp_mask, vma, addr, &page_allocated);
+				swp_entry(swp_type(entry), offset),
+				gfp_mask, mpol, ilx, &page_allocated);
 		if (!page)
 			continue;
 		if (page_allocated) {
@@ -663,11 +668,14 @@ struct page *swap_cluster_readahead(swp_entry_t entry, gfp_t gfp_mask,
 	}
 	blk_finish_plug(&plug);
 	swap_read_unplug(splug);
-
 	lru_add_drain();	/* Push any new pages onto the LRU now */
 skip:
 	/* The page was likely read above, so no need for plugging here */
-	return read_swap_cache_async(entry, gfp_mask, vma, addr, NULL);
+	page = __read_swap_cache_async(entry, gfp_mask, mpol, ilx,
+					&page_allocated);
+	if (unlikely(page_allocated))
+		swap_readpage(page, false, NULL);
+	return page;
 }
 
 int init_swap_address_space(unsigned int type, unsigned long nr_pages)
@@ -765,8 +773,10 @@ static void swap_ra_info(struct vm_fault *vmf,
 
 /**
  * swap_vma_readahead - swap in pages in hope we need them soon
- * @fentry: swap entry of this memory
+ * @targ_entry: swap entry of the targeted memory
  * @gfp_mask: memory allocation flags
+ * @mpol: NUMA memory allocation policy to be applied
+ * @targ_ilx: NUMA interleave index, for use only when MPOL_INTERLEAVE
  * @vmf: fault information
  *
  * Returns the struct page for entry and addr, after queueing swapin.
@@ -777,16 +787,17 @@ static void swap_ra_info(struct vm_fault *vmf,
  * Caller must hold read mmap_lock if vmf->vma is not NULL.
  *
  */
-static struct page *swap_vma_readahead(swp_entry_t fentry, gfp_t gfp_mask,
+static struct page *swap_vma_readahead(swp_entry_t targ_entry, gfp_t gfp_mask,
+				       struct mempolicy *mpol, pgoff_t targ_ilx,
 				       struct vm_fault *vmf)
 {
 	struct blk_plug plug;
 	struct swap_iocb *splug = NULL;
-	struct vm_area_struct *vma = vmf->vma;
 	struct page *page;
 	pte_t *pte = NULL, pentry;
 	unsigned long addr;
 	swp_entry_t entry;
+	pgoff_t ilx;
 	unsigned int i;
 	bool page_allocated;
 	struct vma_swap_readahead ra_info = {
@@ -798,9 +809,10 @@ static struct page *swap_vma_readahead(swp_entry_t fentry, gfp_t gfp_mask,
 		goto skip;
 
 	addr = vmf->address - (ra_info.offset * PAGE_SIZE);
+	ilx = targ_ilx - ra_info.offset;
 
 	blk_start_plug(&plug);
-	for (i = 0; i < ra_info.nr_pte; i++, addr += PAGE_SIZE) {
+	for (i = 0; i < ra_info.nr_pte; i++, ilx++, addr += PAGE_SIZE) {
 		if (!pte++) {
 			pte = pte_offset_map(vmf->pmd, addr);
 			if (!pte)
@@ -814,8 +826,8 @@ static struct page *swap_vma_readahead(swp_entry_t fentry, gfp_t gfp_mask,
 			continue;
 		pte_unmap(pte);
 		pte = NULL;
-		page = __read_swap_cache_async(entry, gfp_mask, vma,
-					       addr, &page_allocated);
+		page = __read_swap_cache_async(entry, gfp_mask, mpol, ilx,
+						&page_allocated);
 		if (!page)
 			continue;
 		if (page_allocated) {
@@ -834,8 +846,11 @@ static struct page *swap_vma_readahead(swp_entry_t fentry, gfp_t gfp_mask,
 	lru_add_drain();
 skip:
 	/* The page was likely read above, so no need for plugging here */
-	return read_swap_cache_async(fentry, gfp_mask, vma, vmf->address,
-				     NULL);
+	page = __read_swap_cache_async(targ_entry, gfp_mask, mpol, targ_ilx,
+					&page_allocated);
+	if (unlikely(page_allocated))
+		swap_readpage(page, false, NULL);
+	return page;
 }
 
 /**
@@ -853,9 +868,16 @@ static struct page *swap_vma_readahead(swp_entry_t fentry, gfp_t gfp_mask,
 struct page *swapin_readahead(swp_entry_t entry, gfp_t gfp_mask,
 				struct vm_fault *vmf)
 {
-	return swap_use_vma_readahead() ?
-			swap_vma_readahead(entry, gfp_mask, vmf) :
-			swap_cluster_readahead(entry, gfp_mask, vmf);
+	struct mempolicy *mpol;
+	pgoff_t ilx;
+	struct page *page;
+
+	mpol = get_vma_policy(vmf->vma, vmf->address, 0, &ilx);
+	page = swap_use_vma_readahead() ?
+		swap_vma_readahead(entry, gfp_mask, mpol, ilx, vmf) :
+		swap_cluster_readahead(entry, gfp_mask, mpol, ilx);
+	mpol_cond_put(mpol);
+	return page;
 }
 
 #ifdef CONFIG_SYSFS

mm/zswap.c

@@ -24,6 +24,7 @@
 #include <linux/swap.h>
 #include <linux/crypto.h>
 #include <linux/scatterlist.h>
+#include <linux/mempolicy.h>
 #include <linux/mempool.h>
 #include <linux/zpool.h>
 #include <crypto/acompress.h>
@@ -1057,6 +1058,7 @@ static int zswap_writeback_entry(struct zswap_entry *entry,
 {
 	swp_entry_t swpentry = entry->swpentry;
 	struct page *page;
+	struct mempolicy *mpol;
 	struct scatterlist input, output;
 	struct crypto_acomp_ctx *acomp_ctx;
 	struct zpool *pool = zswap_find_zpool(entry);
@@ -1075,8 +1077,9 @@ static int zswap_writeback_entry(struct zswap_entry *entry,
 	}
 
 	/* try to allocate swap cache page */
-	page = __read_swap_cache_async(swpentry, GFP_KERNEL, NULL, 0,
-				       &page_was_allocated);
+	mpol = get_task_policy(current);
+	page = __read_swap_cache_async(swpentry, GFP_KERNEL, mpol,
+				NO_INTERLEAVE_INDEX, &page_was_allocated);
 	if (!page) {
 		ret = -ENOMEM;
 		goto fail;