diff mbox series

[01/13] mm: Return the folio from __read_swap_cache_async()

Message ID 20231213215842.671461-2-willy@infradead.org (mailing list archive)
State New
Headers show
Series More swap folio conversions | expand

Commit Message

Matthew Wilcox Dec. 13, 2023, 9:58 p.m. UTC
Move the folio->page conversion into the callers that actually want
that.  Most of the callers are happier with the folio anyway.
If the page_allocated boolean is set, the folio allocated is of order-0,
so it is safe to pass the page directly to swap_readpage().

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 include/linux/zswap.h |  4 +--
 mm/swap.h             |  7 ++--
 mm/swap_state.c       | 75 ++++++++++++++++++++-----------------------
 mm/zswap.c            | 48 +++++++++++++--------------
 4 files changed, 64 insertions(+), 70 deletions(-)

Comments

Andrew Morton Dec. 13, 2023, 11:14 p.m. UTC | #1
On Wed, 13 Dec 2023 21:58:30 +0000 "Matthew Wilcox (Oracle)" <willy@infradead.org> wrote:

> Move the folio->page conversion into the callers that actually want
> that.  Most of the callers are happier with the folio anyway.
> If the page_allocated boolean is set, the folio allocated is of order-0,
> so it is safe to pass the page directly to swap_readpage().

Lots of rejects in mm-unstable's mm/zswap.c.  I think I got everything,
please check the result.  Pushed out now.
diff mbox series

Patch

diff --git a/include/linux/zswap.h b/include/linux/zswap.h
index 08c240e16a01..e88572d4c720 100644
--- a/include/linux/zswap.h
+++ b/include/linux/zswap.h
@@ -34,7 +34,7 @@  void zswap_swapon(int type);
 void zswap_swapoff(int type);
 void zswap_memcg_offline_cleanup(struct mem_cgroup *memcg);
 void zswap_lruvec_state_init(struct lruvec *lruvec);
-void zswap_page_swapin(struct page *page);
+void zswap_folio_swapin(struct folio *folio);
 #else
 
 struct zswap_lruvec_state {};
@@ -54,7 +54,7 @@  static inline void zswap_swapon(int type) {}
 static inline void zswap_swapoff(int type) {}
 static inline void zswap_memcg_offline_cleanup(struct mem_cgroup *memcg) {}
 static inline void zswap_lruvec_state_init(struct lruvec *lruvec) {}
-static inline void zswap_page_swapin(struct page *page) {}
+static inline void zswap_folio_swapin(struct folio *folio) {}
 #endif
 
 #endif /* _LINUX_ZSWAP_H */
diff --git a/mm/swap.h b/mm/swap.h
index c0dc73e10e91..a60ab1cfcaf2 100644
--- a/mm/swap.h
+++ b/mm/swap.h
@@ -49,10 +49,9 @@  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);
-struct page *__read_swap_cache_async(swp_entry_t entry, gfp_t gfp_mask,
-				     struct mempolicy *mpol, pgoff_t ilx,
-				     bool *new_page_allocated,
-				     bool skip_if_exists);
+struct folio *__read_swap_cache_async(swp_entry_t entry, gfp_t gfp_flags,
+		struct mempolicy *mpol, pgoff_t ilx, bool *new_page_allocated,
+		bool skip_if_exists);
 struct page *swap_cluster_readahead(swp_entry_t entry, gfp_t flag,
 				    struct mempolicy *mpol, pgoff_t ilx);
 struct page *swapin_readahead(swp_entry_t entry, gfp_t flag,
diff --git a/mm/swap_state.c b/mm/swap_state.c
index 7d775d0b1312..d4e25d9b5dc6 100644
--- a/mm/swap_state.c
+++ b/mm/swap_state.c
@@ -419,14 +419,12 @@  struct folio *filemap_get_incore_folio(struct address_space *mapping,
 	return folio;
 }
 
-struct page *__read_swap_cache_async(swp_entry_t entry, gfp_t gfp_mask,
-				     struct mempolicy *mpol, pgoff_t ilx,
-				     bool *new_page_allocated,
-				     bool skip_if_exists)
+struct folio *__read_swap_cache_async(swp_entry_t entry, gfp_t gfp_mask,
+		struct mempolicy *mpol, pgoff_t ilx, bool *new_page_allocated,
+		bool skip_if_exists)
 {
 	struct swap_info_struct *si;
 	struct folio *folio;
-	struct page *page;
 	void *shadow = NULL;
 
 	*new_page_allocated = false;
@@ -443,10 +441,8 @@  struct page *__read_swap_cache_async(swp_entry_t entry, gfp_t gfp_mask,
 		 */
 		folio = filemap_get_folio(swap_address_space(entry),
 						swp_offset(entry));
-		if (!IS_ERR(folio)) {
-			page = folio_file_page(folio, swp_offset(entry));
-			goto got_page;
-		}
+		if (!IS_ERR(folio))
+			goto got_folio;
 
 		/*
 		 * Just skip read ahead for unused swap slot.
@@ -460,7 +456,7 @@  struct page *__read_swap_cache_async(swp_entry_t entry, gfp_t gfp_mask,
 			goto fail_put_swap;
 
 		/*
-		 * Get a new page to read into from swap.  Allocate it now,
+		 * Get a new folio to read into from swap.  Allocate it now,
 		 * before marking swap_map SWAP_HAS_CACHE, when -EEXIST will
 		 * cause any racers to loop around until we add it to cache.
 		 */
@@ -496,13 +492,13 @@  struct page *__read_swap_cache_async(swp_entry_t entry, gfp_t gfp_mask,
 		 * stumble across a swap_map entry whose SWAP_HAS_CACHE
 		 * has not yet been cleared.  Or race against another
 		 * __read_swap_cache_async(), which has set SWAP_HAS_CACHE
-		 * in swap_map, but not yet added its page to swap cache.
+		 * in swap_map, but not yet added its folio to swap cache.
 		 */
 		schedule_timeout_uninterruptible(1);
 	}
 
 	/*
-	 * The swap entry is ours to swap in. Prepare the new page.
+	 * The swap entry is ours to swap in. Prepare the new folio.
 	 */
 
 	__folio_set_locked(folio);
@@ -523,10 +519,9 @@  struct page *__read_swap_cache_async(swp_entry_t entry, gfp_t gfp_mask,
 	/* Caller will initiate read into locked folio */
 	folio_add_lru(folio);
 	*new_page_allocated = true;
-	page = &folio->page;
-got_page:
+got_folio:
 	put_swap_device(si);
-	return page;
+	return folio;
 
 fail_unlock:
 	put_swap_folio(folio, entry);
@@ -554,16 +549,16 @@  struct page *read_swap_cache_async(swp_entry_t entry, gfp_t gfp_mask,
 	bool page_allocated;
 	struct mempolicy *mpol;
 	pgoff_t ilx;
-	struct page *page;
+	struct folio *folio;
 
 	mpol = get_vma_policy(vma, addr, 0, &ilx);
-	page = __read_swap_cache_async(entry, gfp_mask, mpol, ilx,
+	folio = __read_swap_cache_async(entry, gfp_mask, mpol, ilx,
 					&page_allocated, false);
 	mpol_cond_put(mpol);
 
 	if (page_allocated)
-		swap_readpage(page, false, plug);
-	return page;
+		swap_readpage(&folio->page, false, plug);
+	return folio_file_page(folio, swp_offset(entry));
 }
 
 static unsigned int __swapin_nr_pages(unsigned long prev_offset,
@@ -648,7 +643,7 @@  static unsigned long swapin_nr_pages(unsigned long offset)
 struct page *swap_cluster_readahead(swp_entry_t entry, gfp_t gfp_mask,
 				    struct mempolicy *mpol, pgoff_t ilx)
 {
-	struct page *page;
+	struct folio *folio;
 	unsigned long entry_offset = swp_offset(entry);
 	unsigned long offset = entry_offset;
 	unsigned long start_offset, end_offset;
@@ -673,31 +668,31 @@  struct page *swap_cluster_readahead(swp_entry_t entry, gfp_t gfp_mask,
 	blk_start_plug(&plug);
 	for (offset = start_offset; offset <= end_offset ; offset++) {
 		/* Ok, do the async read-ahead now */
-		page = __read_swap_cache_async(
+		folio = __read_swap_cache_async(
 				swp_entry(swp_type(entry), offset),
 				gfp_mask, mpol, ilx, &page_allocated, false);
-		if (!page)
+		if (!folio)
 			continue;
 		if (page_allocated) {
-			swap_readpage(page, false, &splug);
+			swap_readpage(&folio->page, false, &splug);
 			if (offset != entry_offset) {
-				SetPageReadahead(page);
+				folio_set_readahead(folio);
 				count_vm_event(SWAP_RA);
 			}
 		}
-		put_page(page);
+		folio_put(folio);
 	}
 	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 */
-	page = __read_swap_cache_async(entry, gfp_mask, mpol, ilx,
+	folio = __read_swap_cache_async(entry, gfp_mask, mpol, ilx,
 					&page_allocated, false);
 	if (unlikely(page_allocated))
-		swap_readpage(page, false, NULL);
-	zswap_page_swapin(page);
-	return page;
+		swap_readpage(&folio->page, false, NULL);
+	zswap_folio_swapin(folio);
+	return folio_file_page(folio, swp_offset(entry));
 }
 
 int init_swap_address_space(unsigned int type, unsigned long nr_pages)
@@ -815,7 +810,7 @@  static struct page *swap_vma_readahead(swp_entry_t targ_entry, gfp_t gfp_mask,
 {
 	struct blk_plug plug;
 	struct swap_iocb *splug = NULL;
-	struct page *page;
+	struct folio *folio;
 	pte_t *pte = NULL, pentry;
 	unsigned long addr;
 	swp_entry_t entry;
@@ -848,18 +843,18 @@  static struct page *swap_vma_readahead(swp_entry_t targ_entry, gfp_t gfp_mask,
 			continue;
 		pte_unmap(pte);
 		pte = NULL;
-		page = __read_swap_cache_async(entry, gfp_mask, mpol, ilx,
+		folio = __read_swap_cache_async(entry, gfp_mask, mpol, ilx,
 						&page_allocated, false);
-		if (!page)
+		if (!folio)
 			continue;
 		if (page_allocated) {
-			swap_readpage(page, false, &splug);
+			swap_readpage(&folio->page, false, &splug);
 			if (i != ra_info.offset) {
-				SetPageReadahead(page);
+				folio_set_readahead(folio);
 				count_vm_event(SWAP_RA);
 			}
 		}
-		put_page(page);
+		folio_put(folio);
 	}
 	if (pte)
 		pte_unmap(pte);
@@ -867,13 +862,13 @@  static struct page *swap_vma_readahead(swp_entry_t targ_entry, gfp_t gfp_mask,
 	swap_read_unplug(splug);
 	lru_add_drain();
 skip:
-	/* The page was likely read above, so no need for plugging here */
-	page = __read_swap_cache_async(targ_entry, gfp_mask, mpol, targ_ilx,
+	/* The folio was likely read above, so no need for plugging here */
+	folio = __read_swap_cache_async(targ_entry, gfp_mask, mpol, targ_ilx,
 					&page_allocated, false);
 	if (unlikely(page_allocated))
-		swap_readpage(page, false, NULL);
-	zswap_page_swapin(page);
-	return page;
+		swap_readpage(&folio->page, false, NULL);
+	zswap_folio_swapin(folio);
+	return folio_file_page(folio, swp_offset(entry));
 }
 
 /**
diff --git a/mm/zswap.c b/mm/zswap.c
index d49aad0359d2..08b5212aa6b8 100644
--- a/mm/zswap.c
+++ b/mm/zswap.c
@@ -368,12 +368,12 @@  void zswap_lruvec_state_init(struct lruvec *lruvec)
 	atomic_long_set(&lruvec->zswap_lruvec_state.nr_zswap_protected, 0);
 }
 
-void zswap_page_swapin(struct page *page)
+void zswap_folio_swapin(struct folio *folio)
 {
 	struct lruvec *lruvec;
 
-	if (page) {
-		lruvec = folio_lruvec(page_folio(page));
+	if (folio) {
+		lruvec = folio_lruvec(folio);
 		atomic_long_inc(&lruvec->zswap_lruvec_state.nr_zswap_protected);
 	}
 }
@@ -1379,14 +1379,14 @@  static int zswap_enabled_param_set(const char *val,
 * writeback code
 **********************************/
 /*
- * Attempts to free an entry by adding a page to the swap cache,
- * decompressing the entry data into the page, and issuing a
- * bio write to write the page back to the swap device.
+ * Attempts to free an entry by adding a folio to the swap cache,
+ * decompressing the entry data into the folio, and issuing a
+ * bio write to write the folio back to the swap device.
  *
- * This can be thought of as a "resumed writeback" of the page
+ * This can be thought of as a "resumed writeback" of the folio
  * to the swap device.  We are basically resuming the same swap
  * writeback path that was intercepted with the zswap_store()
- * in the first place.  After the page has been decompressed into
+ * in the first place.  After the folio has been decompressed into
  * the swap cache, the compressed version stored by zswap can be
  * freed.
  */
@@ -1394,7 +1394,7 @@  static int zswap_writeback_entry(struct zswap_entry *entry,
 				 struct zswap_tree *tree)
 {
 	swp_entry_t swpentry = entry->swpentry;
-	struct page *page;
+	struct folio *folio;
 	struct mempolicy *mpol;
 	struct scatterlist input, output;
 	struct crypto_acomp_ctx *acomp_ctx;
@@ -1413,18 +1413,18 @@  static int zswap_writeback_entry(struct zswap_entry *entry,
 			return -ENOMEM;
 	}
 
-	/* try to allocate swap cache page */
+	/* try to allocate swap cache folio */
 	mpol = get_task_policy(current);
-	page = __read_swap_cache_async(swpentry, GFP_KERNEL, mpol,
+	folio = __read_swap_cache_async(swpentry, GFP_KERNEL, mpol,
 				NO_INTERLEAVE_INDEX, &page_was_allocated, true);
-	if (!page) {
+	if (!folio) {
 		ret = -ENOMEM;
 		goto fail;
 	}
 
-	/* Found an existing page, we raced with load/swapin */
+	/* Found an existing folio, we raced with load/swapin */
 	if (!page_was_allocated) {
-		put_page(page);
+		folio_put(folio);
 		ret = -EEXIST;
 		goto fail;
 	}
@@ -1434,12 +1434,12 @@  static int zswap_writeback_entry(struct zswap_entry *entry,
 	 * concurrent swapping to and from the slot. Verify that the
 	 * swap entry hasn't been invalidated and recycled behind our
 	 * backs (our zswap_entry reference doesn't prevent that), to
-	 * avoid overwriting a new swap page with old compressed data.
+	 * avoid overwriting a new swap folio with old compressed data.
 	 */
 	spin_lock(&tree->lock);
 	if (zswap_rb_search(&tree->rbroot, swp_offset(entry->swpentry)) != entry) {
 		spin_unlock(&tree->lock);
-		delete_from_swap_cache(page_folio(page));
+		delete_from_swap_cache(folio);
 		ret = -ENOMEM;
 		goto fail;
 	}
@@ -1459,7 +1459,7 @@  static int zswap_writeback_entry(struct zswap_entry *entry,
 	mutex_lock(acomp_ctx->mutex);
 	sg_init_one(&input, src, entry->length);
 	sg_init_table(&output, 1);
-	sg_set_page(&output, page, PAGE_SIZE, 0);
+	sg_set_page(&output, &folio->page, PAGE_SIZE, 0);
 	acomp_request_set_params(acomp_ctx->req, &input, &output, entry->length, dlen);
 	ret = crypto_wait_req(crypto_acomp_decompress(acomp_ctx->req), &acomp_ctx->wait);
 	dlen = acomp_ctx->req->dlen;
@@ -1473,20 +1473,20 @@  static int zswap_writeback_entry(struct zswap_entry *entry,
 	BUG_ON(ret);
 	BUG_ON(dlen != PAGE_SIZE);
 
-	/* page is up to date */
-	SetPageUptodate(page);
+	/* folio is up to date */
+	folio_mark_uptodate(folio);
 
 	/* move it to the tail of the inactive list after end_writeback */
-	SetPageReclaim(page);
+	folio_set_reclaim(folio);
 
-	if (!PageLRU(page)) {
+	if (!folio_test_lru(folio)) {
 		/* drain lru cache to help folio_rotate_reclaimable() */
 		lru_add_drain();
 	}
 
 	/* start writeback */
-	__swap_writepage(page, &wbc);
-	put_page(page);
+	__swap_writepage(&folio->page, &wbc);
+	folio_put(folio);
 
 	return ret;
 
@@ -1495,7 +1495,7 @@  static int zswap_writeback_entry(struct zswap_entry *entry,
 		kfree(tmp);
 
 	/*
-	 * If we get here because the page is already in swapcache, a
+	 * If we get here because the folio is already in swapcache, a
 	 * load may be happening concurrently. It is safe and okay to
 	 * not free the entry. It is also okay to return !0.
 	 */