swapin needs gfp_mask for loop on tmpfs

Building in a filesystem on a loop device on a tmpfs file can hang when
swapping, the loop thread caught in that infamous throttle_vm_writeout.

In theory this is a long standing problem, which I've either never seen in
practice, or long ago suppressed the recollection, after discounting my load
and my tmpfs size as unrealistically high.  But now, with the new aops, it has
become easy to hang on one machine.

Loop used to grab_cache_page before the old prepare_write to tmpfs, which
seems to have been enough to free up some memory for any swapin needed; but
the new write_begin lets tmpfs find or allocate the page (much nicer, since
grab_cache_page missed tmpfs pages in swapcache).

When allocating a fresh page, tmpfs respects loop's mapping_gfp_mask, which
has __GFP_IO|__GFP_FS stripped off, and throttle_vm_writeout is designed to
break out when __GFP_IO or GFP_FS is unset; but when tmfps swaps in,
read_swap_cache_async allocates with GFP_HIGHUSER_MOVABLE regardless of the
mapping_gfp_mask - hence the hang.

So, pass gfp_mask down the line from shmem_getpage to shmem_swapin to
swapin_readahead to read_swap_cache_async to add_to_swap_cache.

Signed-off-by: Hugh Dickins <[email protected]>
Acked-by: Rik van Riel <[email protected]>
Acked-by: Peter Zijlstra <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>

include/linux/swap.h

@@ -228,9 +228,9 @@ extern int move_from_swap_cache(struct page *, unsigned long,
 extern void free_page_and_swap_cache(struct page *);
 extern void free_pages_and_swap_cache(struct page **, int);
 extern struct page *lookup_swap_cache(swp_entry_t);
-extern struct page *read_swap_cache_async(swp_entry_t,
+extern struct page *read_swap_cache_async(swp_entry_t, gfp_t,
 			struct vm_area_struct *vma, unsigned long addr);
-extern struct page *swapin_readahead(swp_entry_t,
+extern struct page *swapin_readahead(swp_entry_t, gfp_t,
 			struct vm_area_struct *vma, unsigned long addr);
 
 /* linux/mm/swapfile.c */
@@ -306,7 +306,7 @@ static inline void swap_free(swp_entry_t swp)
 {
 }
 
-static inline struct page *swapin_readahead(swp_entry_t swp,
+static inline struct page *swapin_readahead(swp_entry_t swp, gfp_t gfp_mask,
 			struct vm_area_struct *vma, unsigned long addr)
 {
 	return NULL;

mm/memory.c

@@ -2007,7 +2007,8 @@ static int do_swap_page(struct mm_struct *mm, struct vm_area_struct *vma,
 	page = lookup_swap_cache(entry);
 	if (!page) {
 		grab_swap_token(); /* Contend for token _before_ read-in */
-		page = swapin_readahead(entry, vma, address);
+		page = swapin_readahead(entry,
+					GFP_HIGHUSER_MOVABLE, vma, address);
 		if (!page) {
 			/*
 			 * Back out if somebody else faulted in this pte

mm/shmem.c

@@ -1025,8 +1025,8 @@ static inline int shmem_parse_mpol(char *value, int *policy, nodemask_t *policy_
 	return err;
 }
 
-static struct page *shmem_swapin(struct shmem_inode_info *info,
-				       swp_entry_t entry, unsigned long idx)
+static struct page *shmem_swapin(swp_entry_t entry, gfp_t gfp,
+			struct shmem_inode_info *info, unsigned long idx)
 {
 	struct vm_area_struct pvma;
 	struct page *page;
@@ -1036,13 +1036,13 @@ static struct page *shmem_swapin(struct shmem_inode_info *info,
 	pvma.vm_pgoff = idx;
 	pvma.vm_ops = NULL;
 	pvma.vm_policy = mpol_shared_policy_lookup(&info->policy, idx);
-	page = swapin_readahead(entry, &pvma, 0);
+	page = swapin_readahead(entry, gfp, &pvma, 0);
 	mpol_free(pvma.vm_policy);
 	return page;
 }
 
-static struct page *shmem_alloc_page(gfp_t gfp, struct shmem_inode_info *info,
-					unsigned long idx)
+static struct page *shmem_alloc_page(gfp_t gfp,
+			struct shmem_inode_info *info, unsigned long idx)
 {
 	struct vm_area_struct pvma;
 	struct page *page;
@@ -1063,14 +1063,14 @@ static inline int shmem_parse_mpol(char *value, int *policy,
 	return 1;
 }
 
-static inline struct page *
-shmem_swapin(struct shmem_inode_info *info,swp_entry_t entry,unsigned long idx)
+static inline struct page *shmem_swapin(swp_entry_t entry, gfp_t gfp,
+			struct shmem_inode_info *info, unsigned long idx)
 {
-	return swapin_readahead(entry, NULL, 0);
+	return swapin_readahead(entry, gfp, NULL, 0);
 }
 
-static inline struct page *
-shmem_alloc_page(gfp_t gfp,struct shmem_inode_info *info, unsigned long idx)
+static inline struct page *shmem_alloc_page(gfp_t gfp,
+			struct shmem_inode_info *info, unsigned long idx)
 {
 	return alloc_page(gfp);
 }
@@ -1093,6 +1093,7 @@ static int shmem_getpage(struct inode *inode, unsigned long idx,
 	struct page *swappage;
 	swp_entry_t *entry;
 	swp_entry_t swap;
+	gfp_t gfp;
 	int error;
 
 	if (idx >= SHMEM_MAX_INDEX)
@@ -1117,6 +1118,7 @@ static int shmem_getpage(struct inode *inode, unsigned long idx,
 	error = 0;
 	if (sgp == SGP_QUICK)
 		goto failed;
+	gfp = mapping_gfp_mask(mapping);
 
 	spin_lock(&info->lock);
 	shmem_recalc_inode(inode);
@@ -1139,7 +1141,7 @@ static int shmem_getpage(struct inode *inode, unsigned long idx,
 				*type |= VM_FAULT_MAJOR;
 			}
 			spin_unlock(&info->lock);
-			swappage = shmem_swapin(info, swap, idx);
+			swappage = shmem_swapin(swap, gfp, info, idx);
 			if (!swappage) {
 				spin_lock(&info->lock);
 				entry = shmem_swp_alloc(info, idx, sgp);
@@ -1251,9 +1253,7 @@ static int shmem_getpage(struct inode *inode, unsigned long idx,
 
 		if (!filepage) {
 			spin_unlock(&info->lock);
-			filepage = shmem_alloc_page(mapping_gfp_mask(mapping),
-						    info,
-						    idx);
+			filepage = shmem_alloc_page(gfp, info, idx);
 			if (!filepage) {
 				shmem_unacct_blocks(info->flags, 1);
 				shmem_free_blocks(inode, 1);

mm/swap_state.c

@@ -96,7 +96,8 @@ static int __add_to_swap_cache(struct page *page, swp_entry_t entry,
 	return error;
 }
 
-static int add_to_swap_cache(struct page *page, swp_entry_t entry)
+static int add_to_swap_cache(struct page *page, swp_entry_t entry,
+				gfp_t gfp_mask)
 {
 	int error;
 
@@ -106,7 +107,7 @@ static int add_to_swap_cache(struct page *page, swp_entry_t entry)
 		return -ENOENT;
 	}
 	SetPageLocked(page);
-	error = __add_to_swap_cache(page, entry, GFP_KERNEL);
+	error = __add_to_swap_cache(page, entry, gfp_mask & GFP_KERNEL);
 	/*
 	 * Anon pages are already on the LRU, we don't run lru_cache_add here.
 	 */
@@ -318,7 +319,7 @@ struct page * lookup_swap_cache(swp_entry_t entry)
  * A failure return means that either the page allocation failed or that
  * the swap entry is no longer in use.
  */
-struct page *read_swap_cache_async(swp_entry_t entry,
+struct page *read_swap_cache_async(swp_entry_t entry, gfp_t gfp_mask,
 			struct vm_area_struct *vma, unsigned long addr)
 {
 	struct page *found_page, *new_page = NULL;
@@ -338,8 +339,7 @@ struct page *read_swap_cache_async(swp_entry_t entry,
 		 * Get a new page to read into from swap.
 		 */
 		if (!new_page) {
-			new_page = alloc_page_vma(GFP_HIGHUSER_MOVABLE,
-								vma, addr);
+			new_page = alloc_page_vma(gfp_mask, vma, addr);
 			if (!new_page)
 				break;		/* Out of memory */
 		}
@@ -354,7 +354,7 @@ struct page *read_swap_cache_async(swp_entry_t entry,
 		 * the just freed swap entry for an existing page.
 		 * May fail (-ENOMEM) if radix-tree node allocation failed.
 		 */
-		err = add_to_swap_cache(new_page, entry);
+		err = add_to_swap_cache(new_page, entry, gfp_mask);
 		if (!err) {
 			/*
 			 * Initiate read into locked page and return.
@@ -388,7 +388,7 @@ struct page *read_swap_cache_async(swp_entry_t entry,
  *
  * Caller must hold down_read on the vma->vm_mm if vma is not NULL.
  */
-struct page *swapin_readahead(swp_entry_t entry,
+struct page *swapin_readahead(swp_entry_t entry, gfp_t gfp_mask,
 			struct vm_area_struct *vma, unsigned long addr)
 {
 	int nr_pages;
@@ -407,11 +407,11 @@ struct page *swapin_readahead(swp_entry_t entry,
 	for (end_offset = offset + nr_pages; offset < end_offset; offset++) {
 		/* Ok, do the async read-ahead now */
 		page = read_swap_cache_async(swp_entry(swp_type(entry), offset),
-						vma, addr);
+						gfp_mask, vma, addr);
 		if (!page)
 			break;
 		page_cache_release(page);
 	}
 	lru_add_drain();	/* Push any new pages onto the LRU now */
-	return read_swap_cache_async(entry, vma, addr);
+	return read_swap_cache_async(entry, gfp_mask, vma, addr);
 }

mm/swapfile.c

@@ -730,7 +730,8 @@ static int try_to_unuse(unsigned int type)
 		 */
 		swap_map = &si->swap_map[i];
 		entry = swp_entry(type, i);
-		page = read_swap_cache_async(entry, NULL, 0);
+		page = read_swap_cache_async(entry,
+					GFP_HIGHUSER_MOVABLE, NULL, 0);
 		if (!page) {
 			/*
 			 * Either swap_duplicate() failed because entry