[PATCH 1/6] f2fs: introduce f2fs_set_page_dirty_nobuffer

[PATCH 1/6] f2fs: introduce f2fs_set_page_dirty_nobuffer

[Jaegeuk Kim]

This patch adds f2fs_set_page_dirty_nobuffer() copied from __set_page_dirty_buffer.
When appending 4KB blocks in f2fs on pmem with multiple cores, this improves the
overall performance.

Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
---
 fs/f2fs/checkpoint.c |  2 +-
 fs/f2fs/data.c       | 33 ++++++++++++++++++++++++++++++++-
 fs/f2fs/f2fs.h       |  1 +
 fs/f2fs/node.c       |  2 +-
 4 files changed, 35 insertions(+), 3 deletions(-)

diff --git a/fs/f2fs/checkpoint.c b/fs/f2fs/checkpoint.c
index 2b43d40..2755ef7 100644
--- a/fs/f2fs/checkpoint.c
+++ b/fs/f2fs/checkpoint.c
@@ -366,7 +366,7 @@ static int f2fs_set_meta_page_dirty(struct page *page)
 
 	SetPageUptodate(page);
 	if (!PageDirty(page)) {
-		__set_page_dirty_nobuffers(page);
+		f2fs_set_page_dirty_nobuffers(page);
 		inc_page_count(F2FS_P_SB(page), F2FS_DIRTY_META);
 		SetPagePrivate(page);
 		f2fs_trace_pid(page);
diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c
index 3d93cf1..4a2e97d 100644
--- a/fs/f2fs/data.c
+++ b/fs/f2fs/data.c
@@ -19,6 +19,8 @@
 #include <linux/bio.h>
 #include <linux/prefetch.h>
 #include <linux/uio.h>
+#include <linux/mm.h>
+#include <linux/memcontrol.h>
 #include <linux/cleancache.h>
 
 #include "f2fs.h"
@@ -1775,6 +1777,35 @@ int f2fs_release_page(struct page *page, gfp_t wait)
 	return 1;
 }
 
+/*
+ * This was copied from __set_page_dirty_buffers which gives higher performance
+ * in very high speed storages. (e.g., pmem)
+ */
+void f2fs_set_page_dirty_nobuffers(struct page *page)
+{
+	struct address_space *mapping = page->mapping;
+	unsigned long flags;
+
+	if (unlikely(!mapping))
+		return;
+
+	spin_lock(&mapping->private_lock);
+	lock_page_memcg(page);
+	SetPageDirty(page);
+	spin_unlock(&mapping->private_lock);
+
+	spin_lock_irqsave(&mapping->tree_lock, flags);
+	WARN_ON_ONCE(!PageUptodate(page));
+	account_page_dirtied(page, mapping);
+	radix_tree_tag_set(&mapping->page_tree,
+			page_index(page), PAGECACHE_TAG_DIRTY);
+	spin_unlock_irqrestore(&mapping->tree_lock, flags);
+	unlock_page_memcg(page);
+
+	__mark_inode_dirty(mapping->host, I_DIRTY_PAGES);
+	return;
+}
+
 static int f2fs_set_data_page_dirty(struct page *page)
 {
 	struct address_space *mapping = page->mapping;
@@ -1797,7 +1828,7 @@ static int f2fs_set_data_page_dirty(struct page *page)
 	}
 
 	if (!PageDirty(page)) {
-		__set_page_dirty_nobuffers(page);
+		f2fs_set_page_dirty_nobuffers(page);
 		update_dirty_page(inode, page);
 		return 1;
 	}
diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h
index bd82b6d..a05f65b 100644
--- a/fs/f2fs/f2fs.h
+++ b/fs/f2fs/f2fs.h
@@ -2097,6 +2097,7 @@ struct page *get_new_data_page(struct inode *, struct page *, pgoff_t, bool);
 int do_write_data_page(struct f2fs_io_info *);
 int f2fs_map_blocks(struct inode *, struct f2fs_map_blocks *, int, int);
 int f2fs_fiemap(struct inode *inode, struct fiemap_extent_info *, u64, u64);
+void f2fs_set_page_dirty_nobuffers(struct page *);
 void f2fs_invalidate_page(struct page *, unsigned int, unsigned int);
 int f2fs_release_page(struct page *, gfp_t);
 
diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c
index 69171ce..db73f3c 100644
--- a/fs/f2fs/node.c
+++ b/fs/f2fs/node.c
@@ -1646,7 +1646,7 @@ static int f2fs_set_node_page_dirty(struct page *page)
 
 	SetPageUptodate(page);
 	if (!PageDirty(page)) {
-		__set_page_dirty_nobuffers(page);
+		f2fs_set_page_dirty_nobuffers(page);
 		inc_page_count(F2FS_P_SB(page), F2FS_DIRTY_NODES);
 		SetPagePrivate(page);
 		f2fs_trace_pid(page);
-- 
2.8.3

[PATCH 1/6] f2fs: introduce f2fs_set_page_dirty_nobuffer

[Jaegeuk Kim]

This patch adds f2fs_set_page_dirty_nobuffer() copied from __set_page_dirty_buffer.
When appending 4KB blocks in f2fs on pmem with multiple cores, this improves the
overall performance.

Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
---
 fs/f2fs/checkpoint.c |  2 +-
 fs/f2fs/data.c       | 33 ++++++++++++++++++++++++++++++++-
 fs/f2fs/f2fs.h       |  1 +
 fs/f2fs/node.c       |  2 +-
 4 files changed, 35 insertions(+), 3 deletions(-)

diff --git a/fs/f2fs/checkpoint.c b/fs/f2fs/checkpoint.c
index 2b43d40..2755ef7 100644
--- a/fs/f2fs/checkpoint.c
+++ b/fs/f2fs/checkpoint.c
@@ -366,7 +366,7 @@ static int f2fs_set_meta_page_dirty(struct page *page)
 
 	SetPageUptodate(page);
 	if (!PageDirty(page)) {
-		__set_page_dirty_nobuffers(page);
+		f2fs_set_page_dirty_nobuffers(page);
 		inc_page_count(F2FS_P_SB(page), F2FS_DIRTY_META);
 		SetPagePrivate(page);
 		f2fs_trace_pid(page);
diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c
index 3d93cf1..4a2e97d 100644
--- a/fs/f2fs/data.c
+++ b/fs/f2fs/data.c
@@ -19,6 +19,8 @@
 #include <linux/bio.h>
 #include <linux/prefetch.h>
 #include <linux/uio.h>
+#include <linux/mm.h>
+#include <linux/memcontrol.h>
 #include <linux/cleancache.h>
 
 #include "f2fs.h"
@@ -1775,6 +1777,35 @@ int f2fs_release_page(struct page *page, gfp_t wait)
 	return 1;
 }
 
+/*
+ * This was copied from __set_page_dirty_buffers which gives higher performance
+ * in very high speed storages. (e.g., pmem)
+ */
+void f2fs_set_page_dirty_nobuffers(struct page *page)
+{
+	struct address_space *mapping = page->mapping;
+	unsigned long flags;
+
+	if (unlikely(!mapping))
+		return;
+
+	spin_lock(&mapping->private_lock);
+	lock_page_memcg(page);
+	SetPageDirty(page);
+	spin_unlock(&mapping->private_lock);
+
+	spin_lock_irqsave(&mapping->tree_lock, flags);
+	WARN_ON_ONCE(!PageUptodate(page));
+	account_page_dirtied(page, mapping);
+	radix_tree_tag_set(&mapping->page_tree,
+			page_index(page), PAGECACHE_TAG_DIRTY);
+	spin_unlock_irqrestore(&mapping->tree_lock, flags);
+	unlock_page_memcg(page);
+
+	__mark_inode_dirty(mapping->host, I_DIRTY_PAGES);
+	return;
+}
+
 static int f2fs_set_data_page_dirty(struct page *page)
 {
 	struct address_space *mapping = page->mapping;
@@ -1797,7 +1828,7 @@ static int f2fs_set_data_page_dirty(struct page *page)
 	}
 
 	if (!PageDirty(page)) {
-		__set_page_dirty_nobuffers(page);
+		f2fs_set_page_dirty_nobuffers(page);
 		update_dirty_page(inode, page);
 		return 1;
 	}
diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h
index bd82b6d..a05f65b 100644
--- a/fs/f2fs/f2fs.h
+++ b/fs/f2fs/f2fs.h
@@ -2097,6 +2097,7 @@ struct page *get_new_data_page(struct inode *, struct page *, pgoff_t, bool);
 int do_write_data_page(struct f2fs_io_info *);
 int f2fs_map_blocks(struct inode *, struct f2fs_map_blocks *, int, int);
 int f2fs_fiemap(struct inode *inode, struct fiemap_extent_info *, u64, u64);
+void f2fs_set_page_dirty_nobuffers(struct page *);
 void f2fs_invalidate_page(struct page *, unsigned int, unsigned int);
 int f2fs_release_page(struct page *, gfp_t);
 
diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c
index 69171ce..db73f3c 100644
--- a/fs/f2fs/node.c
+++ b/fs/f2fs/node.c
@@ -1646,7 +1646,7 @@ static int f2fs_set_node_page_dirty(struct page *page)
 
 	SetPageUptodate(page);
 	if (!PageDirty(page)) {
-		__set_page_dirty_nobuffers(page);
+		f2fs_set_page_dirty_nobuffers(page);
 		inc_page_count(F2FS_P_SB(page), F2FS_DIRTY_NODES);
 		SetPagePrivate(page);
 		f2fs_trace_pid(page);
-- 
2.8.3


------------------------------------------------------------------------------
Attend Shape: An AT&T Tech Expo July 15-16. Meet us at AT&T Park in San
Francisco, CA to explore cutting-edge tech and listen to tech luminaries
present their vision of the future. This family event has something for
everyone, including kids. Get more information and register today.
http://sdm.link/attshape

[PATCH 2/6] f2fs: call SetPageUptodate if needed

[Jaegeuk Kim]

SetPageUptodate() issues memory barrier, resulting in performance degrdation.
Let's avoid that.

Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
---
 fs/f2fs/checkpoint.c |  6 ++++--
 fs/f2fs/data.c       | 18 ++++++++++++------
 fs/f2fs/file.c       |  3 ++-
 fs/f2fs/inline.c     |  9 ++++++---
 fs/f2fs/node.c       |  9 ++++++---
 5 files changed, 30 insertions(+), 15 deletions(-)

diff --git a/fs/f2fs/checkpoint.c b/fs/f2fs/checkpoint.c
index 2755ef7..8ea8953 100644
--- a/fs/f2fs/checkpoint.c
+++ b/fs/f2fs/checkpoint.c
@@ -48,7 +48,8 @@ repeat:
 		goto repeat;
 	}
 	f2fs_wait_on_page_writeback(page, META, true);
-	SetPageUptodate(page);
+	if (!PageUptodate(page))
+		SetPageUptodate(page);
 	return page;
 }
 
@@ -364,7 +365,8 @@ static int f2fs_set_meta_page_dirty(struct page *page)
 {
 	trace_f2fs_set_page_dirty(page, META);
 
-	SetPageUptodate(page);
+	if (!PageUptodate(page))
+		SetPageUptodate(page);
 	if (!PageDirty(page)) {
 		f2fs_set_page_dirty_nobuffers(page);
 		inc_page_count(F2FS_P_SB(page), F2FS_DIRTY_META);
diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c
index 4a2e97d..b6fd5bd 100644
--- a/fs/f2fs/data.c
+++ b/fs/f2fs/data.c
@@ -47,7 +47,8 @@ static void f2fs_read_end_io(struct bio *bio)
 		struct page *page = bvec->bv_page;
 
 		if (!bio->bi_error) {
-			SetPageUptodate(page);
+			if (!PageUptodate(page))
+				SetPageUptodate(page);
 		} else {
 			ClearPageUptodate(page);
 			SetPageError(page);
@@ -443,7 +444,8 @@ got_it:
 	 */
 	if (dn.data_blkaddr == NEW_ADDR) {
 		zero_user_segment(page, 0, PAGE_SIZE);
-		SetPageUptodate(page);
+		if (!PageUptodate(page))
+			SetPageUptodate(page);
 		unlock_page(page);
 		return page;
 	}
@@ -554,7 +556,8 @@ struct page *get_new_data_page(struct inode *inode,
 
 	if (dn.data_blkaddr == NEW_ADDR) {
 		zero_user_segment(page, 0, PAGE_SIZE);
-		SetPageUptodate(page);
+		if (!PageUptodate(page))
+			SetPageUptodate(page);
 	} else {
 		f2fs_put_page(page, 1);
 
@@ -1065,7 +1068,8 @@ got_it:
 			}
 		} else {
 			zero_user_segment(page, 0, PAGE_SIZE);
-			SetPageUptodate(page);
+			if (!PageUptodate(page))
+				SetPageUptodate(page);
 			unlock_page(page);
 			goto next_page;
 		}
@@ -1659,7 +1663,8 @@ repeat:
 		}
 	}
 out_update:
-	SetPageUptodate(page);
+	if (!PageUptodate(page))
+		SetPageUptodate(page);
 out_clear:
 	clear_cold_data(page);
 	return 0;
@@ -1813,7 +1818,8 @@ static int f2fs_set_data_page_dirty(struct page *page)
 
 	trace_f2fs_set_page_dirty(page, DATA);
 
-	SetPageUptodate(page);
+	if (!PageUptodate(page))
+		SetPageUptodate(page);
 
 	if (f2fs_is_atomic_file(inode)) {
 		if (!IS_ATOMIC_WRITTEN_PAGE(page)) {
diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c
index 4e220ba..952cbcf 100644
--- a/fs/f2fs/file.c
+++ b/fs/f2fs/file.c
@@ -83,7 +83,8 @@ static int f2fs_vm_page_mkwrite(struct vm_area_struct *vma,
 		zero_user_segment(page, offset, PAGE_SIZE);
 	}
 	set_page_dirty(page);
-	SetPageUptodate(page);
+	if (!PageUptodate(page))
+		SetPageUptodate(page);
 
 	trace_f2fs_vm_page_mkwrite(page, DATA);
 mapped:
diff --git a/fs/f2fs/inline.c b/fs/f2fs/inline.c
index e10e958..2cd0edc 100644
--- a/fs/f2fs/inline.c
+++ b/fs/f2fs/inline.c
@@ -59,7 +59,8 @@ void read_inline_data(struct page *page, struct page *ipage)
 	memcpy(dst_addr, src_addr, MAX_INLINE_DATA);
 	flush_dcache_page(page);
 	kunmap_atomic(dst_addr);
-	SetPageUptodate(page);
+	if (!PageUptodate(page))
+		SetPageUptodate(page);
 }
 
 bool truncate_inline_inode(struct page *ipage, u64 from)
@@ -97,7 +98,8 @@ int f2fs_read_inline_data(struct inode *inode, struct page *page)
 	else
 		read_inline_data(page, ipage);
 
-	SetPageUptodate(page);
+	if (!PageUptodate(page))
+		SetPageUptodate(page);
 	f2fs_put_page(ipage, 1);
 	unlock_page(page);
 	return 0;
@@ -370,7 +372,8 @@ static int f2fs_move_inline_dirents(struct inode *dir, struct page *ipage,
 					NR_INLINE_DENTRY * F2FS_SLOT_LEN);
 
 	kunmap_atomic(dentry_blk);
-	SetPageUptodate(page);
+	if (!PageUptodate(page))
+		SetPageUptodate(page);
 	set_page_dirty(page);
 
 	/* clear inline dir and flag after data writeback */
diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c
index db73f3c..ca1bb3c 100644
--- a/fs/f2fs/node.c
+++ b/fs/f2fs/node.c
@@ -1045,7 +1045,8 @@ struct page *new_node_page(struct dnode_of_data *dn,
 	f2fs_wait_on_page_writeback(page, NODE, true);
 	fill_node_footer(page, dn->nid, dn->inode->i_ino, ofs, true);
 	set_cold_node(dn->inode, page);
-	SetPageUptodate(page);
+	if (!PageUptodate(page))
+		SetPageUptodate(page);
 	if (set_page_dirty(page))
 		dn->node_changed = true;
 
@@ -1644,7 +1645,8 @@ static int f2fs_set_node_page_dirty(struct page *page)
 {
 	trace_f2fs_set_page_dirty(page, NODE);
 
-	SetPageUptodate(page);
+	if (!PageUptodate(page))
+		SetPageUptodate(page);
 	if (!PageDirty(page)) {
 		f2fs_set_page_dirty_nobuffers(page);
 		inc_page_count(F2FS_P_SB(page), F2FS_DIRTY_NODES);
@@ -2015,7 +2017,8 @@ int recover_inode_page(struct f2fs_sb_info *sbi, struct page *page)
 	/* Should not use this inode from free nid list */
 	remove_free_nid(NM_I(sbi), ino);
 
-	SetPageUptodate(ipage);
+	if (!PageUptodate(ipage))
+		SetPageUptodate(ipage);
 	fill_node_footer(ipage, ino, ino, 0, true);
 
 	src = F2FS_INODE(page);
-- 
2.8.3

[PATCH 2/6] f2fs: call SetPageUptodate if needed

[Jaegeuk Kim]

SetPageUptodate() issues memory barrier, resulting in performance degrdation.
Let's avoid that.

Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
---
 fs/f2fs/checkpoint.c |  6 ++++--
 fs/f2fs/data.c       | 18 ++++++++++++------
 fs/f2fs/file.c       |  3 ++-
 fs/f2fs/inline.c     |  9 ++++++---
 fs/f2fs/node.c       |  9 ++++++---
 5 files changed, 30 insertions(+), 15 deletions(-)

diff --git a/fs/f2fs/checkpoint.c b/fs/f2fs/checkpoint.c
index 2755ef7..8ea8953 100644
--- a/fs/f2fs/checkpoint.c
+++ b/fs/f2fs/checkpoint.c
@@ -48,7 +48,8 @@ repeat:
 		goto repeat;
 	}
 	f2fs_wait_on_page_writeback(page, META, true);
-	SetPageUptodate(page);
+	if (!PageUptodate(page))
+		SetPageUptodate(page);
 	return page;
 }
 
@@ -364,7 +365,8 @@ static int f2fs_set_meta_page_dirty(struct page *page)
 {
 	trace_f2fs_set_page_dirty(page, META);
 
-	SetPageUptodate(page);
+	if (!PageUptodate(page))
+		SetPageUptodate(page);
 	if (!PageDirty(page)) {
 		f2fs_set_page_dirty_nobuffers(page);
 		inc_page_count(F2FS_P_SB(page), F2FS_DIRTY_META);
diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c
index 4a2e97d..b6fd5bd 100644
--- a/fs/f2fs/data.c
+++ b/fs/f2fs/data.c
@@ -47,7 +47,8 @@ static void f2fs_read_end_io(struct bio *bio)
 		struct page *page = bvec->bv_page;
 
 		if (!bio->bi_error) {
-			SetPageUptodate(page);
+			if (!PageUptodate(page))
+				SetPageUptodate(page);
 		} else {
 			ClearPageUptodate(page);
 			SetPageError(page);
@@ -443,7 +444,8 @@ got_it:
 	 */
 	if (dn.data_blkaddr == NEW_ADDR) {
 		zero_user_segment(page, 0, PAGE_SIZE);
-		SetPageUptodate(page);
+		if (!PageUptodate(page))
+			SetPageUptodate(page);
 		unlock_page(page);
 		return page;
 	}
@@ -554,7 +556,8 @@ struct page *get_new_data_page(struct inode *inode,
 
 	if (dn.data_blkaddr == NEW_ADDR) {
 		zero_user_segment(page, 0, PAGE_SIZE);
-		SetPageUptodate(page);
+		if (!PageUptodate(page))
+			SetPageUptodate(page);
 	} else {
 		f2fs_put_page(page, 1);
 
@@ -1065,7 +1068,8 @@ got_it:
 			}
 		} else {
 			zero_user_segment(page, 0, PAGE_SIZE);
-			SetPageUptodate(page);
+			if (!PageUptodate(page))
+				SetPageUptodate(page);
 			unlock_page(page);
 			goto next_page;
 		}
@@ -1659,7 +1663,8 @@ repeat:
 		}
 	}
 out_update:
-	SetPageUptodate(page);
+	if (!PageUptodate(page))
+		SetPageUptodate(page);
 out_clear:
 	clear_cold_data(page);
 	return 0;
@@ -1813,7 +1818,8 @@ static int f2fs_set_data_page_dirty(struct page *page)
 
 	trace_f2fs_set_page_dirty(page, DATA);
 
-	SetPageUptodate(page);
+	if (!PageUptodate(page))
+		SetPageUptodate(page);
 
 	if (f2fs_is_atomic_file(inode)) {
 		if (!IS_ATOMIC_WRITTEN_PAGE(page)) {
diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c
index 4e220ba..952cbcf 100644
--- a/fs/f2fs/file.c
+++ b/fs/f2fs/file.c
@@ -83,7 +83,8 @@ static int f2fs_vm_page_mkwrite(struct vm_area_struct *vma,
 		zero_user_segment(page, offset, PAGE_SIZE);
 	}
 	set_page_dirty(page);
-	SetPageUptodate(page);
+	if (!PageUptodate(page))
+		SetPageUptodate(page);
 
 	trace_f2fs_vm_page_mkwrite(page, DATA);
 mapped:
diff --git a/fs/f2fs/inline.c b/fs/f2fs/inline.c
index e10e958..2cd0edc 100644
--- a/fs/f2fs/inline.c
+++ b/fs/f2fs/inline.c
@@ -59,7 +59,8 @@ void read_inline_data(struct page *page, struct page *ipage)
 	memcpy(dst_addr, src_addr, MAX_INLINE_DATA);
 	flush_dcache_page(page);
 	kunmap_atomic(dst_addr);
-	SetPageUptodate(page);
+	if (!PageUptodate(page))
+		SetPageUptodate(page);
 }
 
 bool truncate_inline_inode(struct page *ipage, u64 from)
@@ -97,7 +98,8 @@ int f2fs_read_inline_data(struct inode *inode, struct page *page)
 	else
 		read_inline_data(page, ipage);
 
-	SetPageUptodate(page);
+	if (!PageUptodate(page))
+		SetPageUptodate(page);
 	f2fs_put_page(ipage, 1);
 	unlock_page(page);
 	return 0;
@@ -370,7 +372,8 @@ static int f2fs_move_inline_dirents(struct inode *dir, struct page *ipage,
 					NR_INLINE_DENTRY * F2FS_SLOT_LEN);
 
 	kunmap_atomic(dentry_blk);
-	SetPageUptodate(page);
+	if (!PageUptodate(page))
+		SetPageUptodate(page);
 	set_page_dirty(page);
 
 	/* clear inline dir and flag after data writeback */
diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c
index db73f3c..ca1bb3c 100644
--- a/fs/f2fs/node.c
+++ b/fs/f2fs/node.c
@@ -1045,7 +1045,8 @@ struct page *new_node_page(struct dnode_of_data *dn,
 	f2fs_wait_on_page_writeback(page, NODE, true);
 	fill_node_footer(page, dn->nid, dn->inode->i_ino, ofs, true);
 	set_cold_node(dn->inode, page);
-	SetPageUptodate(page);
+	if (!PageUptodate(page))
+		SetPageUptodate(page);
 	if (set_page_dirty(page))
 		dn->node_changed = true;
 
@@ -1644,7 +1645,8 @@ static int f2fs_set_node_page_dirty(struct page *page)
 {
 	trace_f2fs_set_page_dirty(page, NODE);
 
-	SetPageUptodate(page);
+	if (!PageUptodate(page))
+		SetPageUptodate(page);
 	if (!PageDirty(page)) {
 		f2fs_set_page_dirty_nobuffers(page);
 		inc_page_count(F2FS_P_SB(page), F2FS_DIRTY_NODES);
@@ -2015,7 +2017,8 @@ int recover_inode_page(struct f2fs_sb_info *sbi, struct page *page)
 	/* Should not use this inode from free nid list */
 	remove_free_nid(NM_I(sbi), ino);
 
-	SetPageUptodate(ipage);
+	if (!PageUptodate(ipage))
+		SetPageUptodate(ipage);
 	fill_node_footer(ipage, ino, ino, 0, true);
 
 	src = F2FS_INODE(page);
-- 
2.8.3


------------------------------------------------------------------------------
Attend Shape: An AT&T Tech Expo July 15-16. Meet us at AT&T Park in San
Francisco, CA to explore cutting-edge tech and listen to tech luminaries
present their vision of the future. This family event has something for
everyone, including kids. Get more information and register today.
http://sdm.link/attshape

[PATCH 3/6] f2fs: shrink critical region in spin_lock

[Jaegeuk Kim]

This patch shrinks the critical region in spin_lock.

Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
---
 fs/f2fs/f2fs.h | 23 ++++++++---------------
 1 file changed, 8 insertions(+), 15 deletions(-)

diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h
index a05f65b..a51f230 100644
--- a/fs/f2fs/f2fs.h
+++ b/fs/f2fs/f2fs.h
@@ -1156,30 +1156,23 @@ static inline void f2fs_i_blocks_write(struct inode *, blkcnt_t, bool);
 static inline bool inc_valid_block_count(struct f2fs_sb_info *sbi,
 				 struct inode *inode, blkcnt_t *count)
 {
-	block_t	valid_block_count;
-
-	spin_lock(&sbi->stat_lock);
 #ifdef CONFIG_F2FS_FAULT_INJECTION
-	if (time_to_inject(FAULT_BLOCK)) {
-		spin_unlock(&sbi->stat_lock);
+	if (time_to_inject(FAULT_BLOCK))
 		return false;
-	}
 #endif
-	valid_block_count =
-		sbi->total_valid_block_count + (block_t)(*count);
-	if (unlikely(valid_block_count > sbi->user_block_count)) {
-		*count = sbi->user_block_count - sbi->total_valid_block_count;
+	spin_lock(&sbi->stat_lock);
+	sbi->total_valid_block_count += (block_t)(*count);
+	if (unlikely(sbi->total_valid_block_count > sbi->user_block_count)) {
+		*count -= sbi->total_valid_block_count - sbi->user_block_count;
+		sbi->total_valid_block_count = sbi->user_block_count;
 		if (!*count) {
 			spin_unlock(&sbi->stat_lock);
 			return false;
 		}
 	}
-	/* *count can be recalculated */
-	f2fs_i_blocks_write(inode, *count, true);
-	sbi->total_valid_block_count =
-		sbi->total_valid_block_count + (block_t)(*count);
 	spin_unlock(&sbi->stat_lock);
 
+	f2fs_i_blocks_write(inode, *count, true);
 	percpu_counter_add(&sbi->alloc_valid_block_count, (*count));
 	return true;
 }
@@ -1191,9 +1184,9 @@ static inline void dec_valid_block_count(struct f2fs_sb_info *sbi,
 	spin_lock(&sbi->stat_lock);
 	f2fs_bug_on(sbi, sbi->total_valid_block_count < (block_t) count);
 	f2fs_bug_on(sbi, inode->i_blocks < count);
-	f2fs_i_blocks_write(inode, count, false);
 	sbi->total_valid_block_count -= (block_t)count;
 	spin_unlock(&sbi->stat_lock);
+	f2fs_i_blocks_write(inode, count, false);
 }
 
 static inline void inc_page_count(struct f2fs_sb_info *sbi, int count_type)
-- 
2.8.3

[PATCH 3/6] f2fs: shrink critical region in spin_lock

[Jaegeuk Kim]

This patch shrinks the critical region in spin_lock.

Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
---
 fs/f2fs/f2fs.h | 23 ++++++++---------------
 1 file changed, 8 insertions(+), 15 deletions(-)

diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h
index a05f65b..a51f230 100644
--- a/fs/f2fs/f2fs.h
+++ b/fs/f2fs/f2fs.h
@@ -1156,30 +1156,23 @@ static inline void f2fs_i_blocks_write(struct inode *, blkcnt_t, bool);
 static inline bool inc_valid_block_count(struct f2fs_sb_info *sbi,
 				 struct inode *inode, blkcnt_t *count)
 {
-	block_t	valid_block_count;
-
-	spin_lock(&sbi->stat_lock);
 #ifdef CONFIG_F2FS_FAULT_INJECTION
-	if (time_to_inject(FAULT_BLOCK)) {
-		spin_unlock(&sbi->stat_lock);
+	if (time_to_inject(FAULT_BLOCK))
 		return false;
-	}
 #endif
-	valid_block_count =
-		sbi->total_valid_block_count + (block_t)(*count);
-	if (unlikely(valid_block_count > sbi->user_block_count)) {
-		*count = sbi->user_block_count - sbi->total_valid_block_count;
+	spin_lock(&sbi->stat_lock);
+	sbi->total_valid_block_count += (block_t)(*count);
+	if (unlikely(sbi->total_valid_block_count > sbi->user_block_count)) {
+		*count -= sbi->total_valid_block_count - sbi->user_block_count;
+		sbi->total_valid_block_count = sbi->user_block_count;
 		if (!*count) {
 			spin_unlock(&sbi->stat_lock);
 			return false;
 		}
 	}
-	/* *count can be recalculated */
-	f2fs_i_blocks_write(inode, *count, true);
-	sbi->total_valid_block_count =
-		sbi->total_valid_block_count + (block_t)(*count);
 	spin_unlock(&sbi->stat_lock);
 
+	f2fs_i_blocks_write(inode, *count, true);
 	percpu_counter_add(&sbi->alloc_valid_block_count, (*count));
 	return true;
 }
@@ -1191,9 +1184,9 @@ static inline void dec_valid_block_count(struct f2fs_sb_info *sbi,
 	spin_lock(&sbi->stat_lock);
 	f2fs_bug_on(sbi, sbi->total_valid_block_count < (block_t) count);
 	f2fs_bug_on(sbi, inode->i_blocks < count);
-	f2fs_i_blocks_write(inode, count, false);
 	sbi->total_valid_block_count -= (block_t)count;
 	spin_unlock(&sbi->stat_lock);
+	f2fs_i_blocks_write(inode, count, false);
 }
 
 static inline void inc_page_count(struct f2fs_sb_info *sbi, int count_type)
-- 
2.8.3


------------------------------------------------------------------------------
Attend Shape: An AT&T Tech Expo July 15-16. Meet us at AT&T Park in San
Francisco, CA to explore cutting-edge tech and listen to tech luminaries
present their vision of the future. This family event has something for
everyone, including kids. Get more information and register today.
http://sdm.link/attshape

[PATCH 4/6] f2fs: skip to check the block address of node page

[Jaegeuk Kim]

If the node page is up-to-date, it should be alive.

Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
---
 fs/f2fs/node.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c
index ca1bb3c..9d994b9 100644
--- a/fs/f2fs/node.c
+++ b/fs/f2fs/node.c
@@ -1080,6 +1080,9 @@ static int read_node_page(struct page *page, int rw)
 		.encrypted_page = NULL,
 	};
 
+	if (PageUptodate(page))
+		return LOCKED_PAGE;
+
 	get_node_info(sbi, page->index, &ni);
 
 	if (unlikely(ni.blk_addr == NULL_ADDR)) {
@@ -1087,9 +1090,6 @@ static int read_node_page(struct page *page, int rw)
 		return -ENOENT;
 	}
 
-	if (PageUptodate(page))
-		return LOCKED_PAGE;
-
 	fio.new_blkaddr = fio.old_blkaddr = ni.blk_addr;
 	return f2fs_submit_page_bio(&fio);
 }
-- 
2.8.3

[PATCH 4/6] f2fs: skip to check the block address of node page

[Jaegeuk Kim]

If the node page is up-to-date, it should be alive.

Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
---
 fs/f2fs/node.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c
index ca1bb3c..9d994b9 100644
--- a/fs/f2fs/node.c
+++ b/fs/f2fs/node.c
@@ -1080,6 +1080,9 @@ static int read_node_page(struct page *page, int rw)
 		.encrypted_page = NULL,
 	};
 
+	if (PageUptodate(page))
+		return LOCKED_PAGE;
+
 	get_node_info(sbi, page->index, &ni);
 
 	if (unlikely(ni.blk_addr == NULL_ADDR)) {
@@ -1087,9 +1090,6 @@ static int read_node_page(struct page *page, int rw)
 		return -ENOENT;
 	}
 
-	if (PageUptodate(page))
-		return LOCKED_PAGE;
-
 	fio.new_blkaddr = fio.old_blkaddr = ni.blk_addr;
 	return f2fs_submit_page_bio(&fio);
 }
-- 
2.8.3


------------------------------------------------------------------------------
Attend Shape: An AT&T Tech Expo July 15-16. Meet us at AT&T Park in San
Francisco, CA to explore cutting-edge tech and listen to tech luminaries
present their vision of the future. This family event has something for
everyone, including kids. Get more information and register today.
http://sdm.link/attshape

[PATCH 5/6] f2fs: use percpu_rw_semaphore

[Jaegeuk Kim]

This patch replaces rw_semaphore with percpu_rw_semaphore for:
sbi->cp_rwsem
nm_i->nat_tree_lock

Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
---
 fs/f2fs/f2fs.h  | 12 ++++++------
 fs/f2fs/node.c  | 47 ++++++++++++++++++++++++-----------------------
 fs/f2fs/super.c |  6 +++++-
 3 files changed, 35 insertions(+), 30 deletions(-)

diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h
index a51f230..2f59888 100644
--- a/fs/f2fs/f2fs.h
+++ b/fs/f2fs/f2fs.h
@@ -547,7 +547,7 @@ struct f2fs_nm_info {
 	/* NAT cache management */
 	struct radix_tree_root nat_root;/* root of the nat entry cache */
 	struct radix_tree_root nat_set_root;/* root of the nat set cache */
-	struct rw_semaphore nat_tree_lock;	/* protect nat_tree_lock */
+	struct percpu_rw_semaphore nat_tree_lock;	/* protect nat_tree_lock */
 	struct list_head nat_entries;	/* cached nat entry list (clean) */
 	unsigned int nat_cnt;		/* the # of cached nat entries */
 	unsigned int dirty_nat_cnt;	/* total num of nat entries in set */
@@ -795,7 +795,7 @@ struct f2fs_sb_info {
 	struct f2fs_checkpoint *ckpt;		/* raw checkpoint pointer */
 	struct inode *meta_inode;		/* cache meta blocks */
 	struct mutex cp_mutex;			/* checkpoint procedure lock */
-	struct rw_semaphore cp_rwsem;		/* blocking FS operations */
+	struct percpu_rw_semaphore cp_rwsem;		/* blocking FS operations */
 	struct rw_semaphore node_write;		/* locking node writes */
 	wait_queue_head_t cp_wait;
 	unsigned long last_time[MAX_TIME];	/* to store time in jiffies */
@@ -1082,22 +1082,22 @@ static inline void clear_ckpt_flags(struct f2fs_checkpoint *cp, unsigned int f)
 
 static inline void f2fs_lock_op(struct f2fs_sb_info *sbi)
 {
-	down_read(&sbi->cp_rwsem);
+	percpu_down_read(&sbi->cp_rwsem);
 }
 
 static inline void f2fs_unlock_op(struct f2fs_sb_info *sbi)
 {
-	up_read(&sbi->cp_rwsem);
+	percpu_up_read(&sbi->cp_rwsem);
 }
 
 static inline void f2fs_lock_all(struct f2fs_sb_info *sbi)
 {
-	down_write(&sbi->cp_rwsem);
+	percpu_down_write(&sbi->cp_rwsem);
 }
 
 static inline void f2fs_unlock_all(struct f2fs_sb_info *sbi)
 {
-	up_write(&sbi->cp_rwsem);
+	percpu_up_write(&sbi->cp_rwsem);
 }
 
 static inline int __get_cp_reason(struct f2fs_sb_info *sbi)
diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c
index 9d994b9..b841c43 100644
--- a/fs/f2fs/node.c
+++ b/fs/f2fs/node.c
@@ -206,14 +206,14 @@ int need_dentry_mark(struct f2fs_sb_info *sbi, nid_t nid)
 	struct nat_entry *e;
 	bool need = false;
 
-	down_read(&nm_i->nat_tree_lock);
+	percpu_down_read(&nm_i->nat_tree_lock);
 	e = __lookup_nat_cache(nm_i, nid);
 	if (e) {
 		if (!get_nat_flag(e, IS_CHECKPOINTED) &&
 				!get_nat_flag(e, HAS_FSYNCED_INODE))
 			need = true;
 	}
-	up_read(&nm_i->nat_tree_lock);
+	percpu_up_read(&nm_i->nat_tree_lock);
 	return need;
 }
 
@@ -223,11 +223,11 @@ bool is_checkpointed_node(struct f2fs_sb_info *sbi, nid_t nid)
 	struct nat_entry *e;
 	bool is_cp = true;
 
-	down_read(&nm_i->nat_tree_lock);
+	percpu_down_read(&nm_i->nat_tree_lock);
 	e = __lookup_nat_cache(nm_i, nid);
 	if (e && !get_nat_flag(e, IS_CHECKPOINTED))
 		is_cp = false;
-	up_read(&nm_i->nat_tree_lock);
+	percpu_up_read(&nm_i->nat_tree_lock);
 	return is_cp;
 }
 
@@ -237,13 +237,13 @@ bool need_inode_block_update(struct f2fs_sb_info *sbi, nid_t ino)
 	struct nat_entry *e;
 	bool need_update = true;
 
-	down_read(&nm_i->nat_tree_lock);
+	percpu_down_read(&nm_i->nat_tree_lock);
 	e = __lookup_nat_cache(nm_i, ino);
 	if (e && get_nat_flag(e, HAS_LAST_FSYNC) &&
 			(get_nat_flag(e, IS_CHECKPOINTED) ||
 			 get_nat_flag(e, HAS_FSYNCED_INODE)))
 		need_update = false;
-	up_read(&nm_i->nat_tree_lock);
+	percpu_up_read(&nm_i->nat_tree_lock);
 	return need_update;
 }
 
@@ -284,7 +284,7 @@ static void set_node_addr(struct f2fs_sb_info *sbi, struct node_info *ni,
 	struct f2fs_nm_info *nm_i = NM_I(sbi);
 	struct nat_entry *e;
 
-	down_write(&nm_i->nat_tree_lock);
+	percpu_down_write(&nm_i->nat_tree_lock);
 	e = __lookup_nat_cache(nm_i, ni->nid);
 	if (!e) {
 		e = grab_nat_entry(nm_i, ni->nid);
@@ -334,7 +334,7 @@ static void set_node_addr(struct f2fs_sb_info *sbi, struct node_info *ni,
 			set_nat_flag(e, HAS_FSYNCED_INODE, true);
 		set_nat_flag(e, HAS_LAST_FSYNC, fsync_done);
 	}
-	up_write(&nm_i->nat_tree_lock);
+	percpu_up_write(&nm_i->nat_tree_lock);
 }
 
 int try_to_free_nats(struct f2fs_sb_info *sbi, int nr_shrink)
@@ -342,8 +342,7 @@ int try_to_free_nats(struct f2fs_sb_info *sbi, int nr_shrink)
 	struct f2fs_nm_info *nm_i = NM_I(sbi);
 	int nr = nr_shrink;
 
-	if (!down_write_trylock(&nm_i->nat_tree_lock))
-		return 0;
+	percpu_down_write(&nm_i->nat_tree_lock);
 
 	while (nr_shrink && !list_empty(&nm_i->nat_entries)) {
 		struct nat_entry *ne;
@@ -352,7 +351,7 @@ int try_to_free_nats(struct f2fs_sb_info *sbi, int nr_shrink)
 		__del_from_nat_cache(nm_i, ne);
 		nr_shrink--;
 	}
-	up_write(&nm_i->nat_tree_lock);
+	percpu_up_write(&nm_i->nat_tree_lock);
 	return nr - nr_shrink;
 }
 
@@ -374,13 +373,13 @@ void get_node_info(struct f2fs_sb_info *sbi, nid_t nid, struct node_info *ni)
 	ni->nid = nid;
 
 	/* Check nat cache */
-	down_read(&nm_i->nat_tree_lock);
+	percpu_down_read(&nm_i->nat_tree_lock);
 	e = __lookup_nat_cache(nm_i, nid);
 	if (e) {
 		ni->ino = nat_get_ino(e);
 		ni->blk_addr = nat_get_blkaddr(e);
 		ni->version = nat_get_version(e);
-		up_read(&nm_i->nat_tree_lock);
+		percpu_up_read(&nm_i->nat_tree_lock);
 		return;
 	}
 
@@ -404,11 +403,11 @@ void get_node_info(struct f2fs_sb_info *sbi, nid_t nid, struct node_info *ni)
 	node_info_from_raw_nat(ni, &ne);
 	f2fs_put_page(page, 1);
 cache:
-	up_read(&nm_i->nat_tree_lock);
+	percpu_up_read(&nm_i->nat_tree_lock);
 	/* cache nat entry */
-	down_write(&nm_i->nat_tree_lock);
+	percpu_down_write(&nm_i->nat_tree_lock);
 	cache_nat_entry(sbi, nid, &ne);
-	up_write(&nm_i->nat_tree_lock);
+	percpu_up_write(&nm_i->nat_tree_lock);
 }
 
 /*
@@ -1783,7 +1782,7 @@ void build_free_nids(struct f2fs_sb_info *sbi)
 	ra_meta_pages(sbi, NAT_BLOCK_OFFSET(nid), FREE_NID_PAGES,
 							META_NAT, true);
 
-	down_read(&nm_i->nat_tree_lock);
+	percpu_down_read(&nm_i->nat_tree_lock);
 
 	while (1) {
 		struct page *page = get_current_nat_page(sbi, nid);
@@ -1815,7 +1814,7 @@ void build_free_nids(struct f2fs_sb_info *sbi)
 			remove_free_nid(nm_i, nid);
 	}
 	up_read(&curseg->journal_rwsem);
-	up_read(&nm_i->nat_tree_lock);
+	percpu_up_read(&nm_i->nat_tree_lock);
 
 	ra_meta_pages(sbi, NAT_BLOCK_OFFSET(nm_i->next_scan_nid),
 					nm_i->ra_nid_pages, META_NAT, false);
@@ -2204,7 +2203,7 @@ void flush_nat_entries(struct f2fs_sb_info *sbi)
 	if (!nm_i->dirty_nat_cnt)
 		return;
 
-	down_write(&nm_i->nat_tree_lock);
+	percpu_down_write(&nm_i->nat_tree_lock);
 
 	/*
 	 * if there are no enough space in journal to store dirty nat
@@ -2227,7 +2226,7 @@ void flush_nat_entries(struct f2fs_sb_info *sbi)
 	list_for_each_entry_safe(set, tmp, &sets, set_list)
 		__flush_nat_entry_set(sbi, set);
 
-	up_write(&nm_i->nat_tree_lock);
+	percpu_up_write(&nm_i->nat_tree_lock);
 
 	f2fs_bug_on(sbi, nm_i->dirty_nat_cnt);
 }
@@ -2263,7 +2262,8 @@ static int init_node_manager(struct f2fs_sb_info *sbi)
 
 	mutex_init(&nm_i->build_lock);
 	spin_lock_init(&nm_i->free_nid_list_lock);
-	init_rwsem(&nm_i->nat_tree_lock);
+	if (percpu_init_rwsem(&nm_i->nat_tree_lock))
+		return -ENOMEM;
 
 	nm_i->next_scan_nid = le32_to_cpu(sbi->ckpt->next_free_nid);
 	nm_i->bitmap_size = __bitmap_size(sbi, NAT_BITMAP);
@@ -2320,7 +2320,7 @@ void destroy_node_manager(struct f2fs_sb_info *sbi)
 	spin_unlock(&nm_i->free_nid_list_lock);
 
 	/* destroy nat cache */
-	down_write(&nm_i->nat_tree_lock);
+	percpu_down_write(&nm_i->nat_tree_lock);
 	while ((found = __gang_lookup_nat_cache(nm_i,
 					nid, NATVEC_SIZE, natvec))) {
 		unsigned idx;
@@ -2345,8 +2345,9 @@ void destroy_node_manager(struct f2fs_sb_info *sbi)
 			kmem_cache_free(nat_entry_set_slab, setvec[idx]);
 		}
 	}
-	up_write(&nm_i->nat_tree_lock);
+	percpu_up_write(&nm_i->nat_tree_lock);
 
+	percpu_free_rwsem(&nm_i->nat_tree_lock);
 	kfree(nm_i->nat_bitmap);
 	sbi->nm_info = NULL;
 	kfree(nm_i);
diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c
index e0a975d..2bac917 100644
--- a/fs/f2fs/super.c
+++ b/fs/f2fs/super.c
@@ -695,6 +695,8 @@ static void destroy_percpu_info(struct f2fs_sb_info *sbi)
 		percpu_counter_destroy(&sbi->nr_pages[i]);
 	percpu_counter_destroy(&sbi->alloc_valid_block_count);
 	percpu_counter_destroy(&sbi->total_valid_inode_count);
+
+	percpu_free_rwsem(&sbi->cp_rwsem);
 }
 
 static void f2fs_put_super(struct super_block *sb)
@@ -1471,6 +1473,9 @@ static int init_percpu_info(struct f2fs_sb_info *sbi)
 {
 	int i, err;
 
+	if (percpu_init_rwsem(&sbi->cp_rwsem))
+		return -ENOMEM;
+
 	for (i = 0; i < NR_COUNT_TYPE; i++) {
 		err = percpu_counter_init(&sbi->nr_pages[i], 0, GFP_KERNEL);
 		if (err)
@@ -1671,7 +1676,6 @@ try_onemore:
 		sbi->write_io[i].bio = NULL;
 	}
 
-	init_rwsem(&sbi->cp_rwsem);
 	init_waitqueue_head(&sbi->cp_wait);
 	init_sb_info(sbi);
 
-- 
2.8.3

[PATCH 5/6] f2fs: use percpu_rw_semaphore

[Jaegeuk Kim]

This patch replaces rw_semaphore with percpu_rw_semaphore for:
sbi->cp_rwsem
nm_i->nat_tree_lock

Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
---
 fs/f2fs/f2fs.h  | 12 ++++++------
 fs/f2fs/node.c  | 47 ++++++++++++++++++++++++-----------------------
 fs/f2fs/super.c |  6 +++++-
 3 files changed, 35 insertions(+), 30 deletions(-)

diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h
index a51f230..2f59888 100644
--- a/fs/f2fs/f2fs.h
+++ b/fs/f2fs/f2fs.h
@@ -547,7 +547,7 @@ struct f2fs_nm_info {
 	/* NAT cache management */
 	struct radix_tree_root nat_root;/* root of the nat entry cache */
 	struct radix_tree_root nat_set_root;/* root of the nat set cache */
-	struct rw_semaphore nat_tree_lock;	/* protect nat_tree_lock */
+	struct percpu_rw_semaphore nat_tree_lock;	/* protect nat_tree_lock */
 	struct list_head nat_entries;	/* cached nat entry list (clean) */
 	unsigned int nat_cnt;		/* the # of cached nat entries */
 	unsigned int dirty_nat_cnt;	/* total num of nat entries in set */
@@ -795,7 +795,7 @@ struct f2fs_sb_info {
 	struct f2fs_checkpoint *ckpt;		/* raw checkpoint pointer */
 	struct inode *meta_inode;		/* cache meta blocks */
 	struct mutex cp_mutex;			/* checkpoint procedure lock */
-	struct rw_semaphore cp_rwsem;		/* blocking FS operations */
+	struct percpu_rw_semaphore cp_rwsem;		/* blocking FS operations */
 	struct rw_semaphore node_write;		/* locking node writes */
 	wait_queue_head_t cp_wait;
 	unsigned long last_time[MAX_TIME];	/* to store time in jiffies */
@@ -1082,22 +1082,22 @@ static inline void clear_ckpt_flags(struct f2fs_checkpoint *cp, unsigned int f)
 
 static inline void f2fs_lock_op(struct f2fs_sb_info *sbi)
 {
-	down_read(&sbi->cp_rwsem);
+	percpu_down_read(&sbi->cp_rwsem);
 }
 
 static inline void f2fs_unlock_op(struct f2fs_sb_info *sbi)
 {
-	up_read(&sbi->cp_rwsem);
+	percpu_up_read(&sbi->cp_rwsem);
 }
 
 static inline void f2fs_lock_all(struct f2fs_sb_info *sbi)
 {
-	down_write(&sbi->cp_rwsem);
+	percpu_down_write(&sbi->cp_rwsem);
 }
 
 static inline void f2fs_unlock_all(struct f2fs_sb_info *sbi)
 {
-	up_write(&sbi->cp_rwsem);
+	percpu_up_write(&sbi->cp_rwsem);
 }
 
 static inline int __get_cp_reason(struct f2fs_sb_info *sbi)
diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c
index 9d994b9..b841c43 100644
--- a/fs/f2fs/node.c
+++ b/fs/f2fs/node.c
@@ -206,14 +206,14 @@ int need_dentry_mark(struct f2fs_sb_info *sbi, nid_t nid)
 	struct nat_entry *e;
 	bool need = false;
 
-	down_read(&nm_i->nat_tree_lock);
+	percpu_down_read(&nm_i->nat_tree_lock);
 	e = __lookup_nat_cache(nm_i, nid);
 	if (e) {
 		if (!get_nat_flag(e, IS_CHECKPOINTED) &&
 				!get_nat_flag(e, HAS_FSYNCED_INODE))
 			need = true;
 	}
-	up_read(&nm_i->nat_tree_lock);
+	percpu_up_read(&nm_i->nat_tree_lock);
 	return need;
 }
 
@@ -223,11 +223,11 @@ bool is_checkpointed_node(struct f2fs_sb_info *sbi, nid_t nid)
 	struct nat_entry *e;
 	bool is_cp = true;
 
-	down_read(&nm_i->nat_tree_lock);
+	percpu_down_read(&nm_i->nat_tree_lock);
 	e = __lookup_nat_cache(nm_i, nid);
 	if (e && !get_nat_flag(e, IS_CHECKPOINTED))
 		is_cp = false;
-	up_read(&nm_i->nat_tree_lock);
+	percpu_up_read(&nm_i->nat_tree_lock);
 	return is_cp;
 }
 
@@ -237,13 +237,13 @@ bool need_inode_block_update(struct f2fs_sb_info *sbi, nid_t ino)
 	struct nat_entry *e;
 	bool need_update = true;
 
-	down_read(&nm_i->nat_tree_lock);
+	percpu_down_read(&nm_i->nat_tree_lock);
 	e = __lookup_nat_cache(nm_i, ino);
 	if (e && get_nat_flag(e, HAS_LAST_FSYNC) &&
 			(get_nat_flag(e, IS_CHECKPOINTED) ||
 			 get_nat_flag(e, HAS_FSYNCED_INODE)))
 		need_update = false;
-	up_read(&nm_i->nat_tree_lock);
+	percpu_up_read(&nm_i->nat_tree_lock);
 	return need_update;
 }
 
@@ -284,7 +284,7 @@ static void set_node_addr(struct f2fs_sb_info *sbi, struct node_info *ni,
 	struct f2fs_nm_info *nm_i = NM_I(sbi);
 	struct nat_entry *e;
 
-	down_write(&nm_i->nat_tree_lock);
+	percpu_down_write(&nm_i->nat_tree_lock);
 	e = __lookup_nat_cache(nm_i, ni->nid);
 	if (!e) {
 		e = grab_nat_entry(nm_i, ni->nid);
@@ -334,7 +334,7 @@ static void set_node_addr(struct f2fs_sb_info *sbi, struct node_info *ni,
 			set_nat_flag(e, HAS_FSYNCED_INODE, true);
 		set_nat_flag(e, HAS_LAST_FSYNC, fsync_done);
 	}
-	up_write(&nm_i->nat_tree_lock);
+	percpu_up_write(&nm_i->nat_tree_lock);
 }
 
 int try_to_free_nats(struct f2fs_sb_info *sbi, int nr_shrink)
@@ -342,8 +342,7 @@ int try_to_free_nats(struct f2fs_sb_info *sbi, int nr_shrink)
 	struct f2fs_nm_info *nm_i = NM_I(sbi);
 	int nr = nr_shrink;
 
-	if (!down_write_trylock(&nm_i->nat_tree_lock))
-		return 0;
+	percpu_down_write(&nm_i->nat_tree_lock);
 
 	while (nr_shrink && !list_empty(&nm_i->nat_entries)) {
 		struct nat_entry *ne;
@@ -352,7 +351,7 @@ int try_to_free_nats(struct f2fs_sb_info *sbi, int nr_shrink)
 		__del_from_nat_cache(nm_i, ne);
 		nr_shrink--;
 	}
-	up_write(&nm_i->nat_tree_lock);
+	percpu_up_write(&nm_i->nat_tree_lock);
 	return nr - nr_shrink;
 }
 
@@ -374,13 +373,13 @@ void get_node_info(struct f2fs_sb_info *sbi, nid_t nid, struct node_info *ni)
 	ni->nid = nid;
 
 	/* Check nat cache */
-	down_read(&nm_i->nat_tree_lock);
+	percpu_down_read(&nm_i->nat_tree_lock);
 	e = __lookup_nat_cache(nm_i, nid);
 	if (e) {
 		ni->ino = nat_get_ino(e);
 		ni->blk_addr = nat_get_blkaddr(e);
 		ni->version = nat_get_version(e);
-		up_read(&nm_i->nat_tree_lock);
+		percpu_up_read(&nm_i->nat_tree_lock);
 		return;
 	}
 
@@ -404,11 +403,11 @@ void get_node_info(struct f2fs_sb_info *sbi, nid_t nid, struct node_info *ni)
 	node_info_from_raw_nat(ni, &ne);
 	f2fs_put_page(page, 1);
 cache:
-	up_read(&nm_i->nat_tree_lock);
+	percpu_up_read(&nm_i->nat_tree_lock);
 	/* cache nat entry */
-	down_write(&nm_i->nat_tree_lock);
+	percpu_down_write(&nm_i->nat_tree_lock);
 	cache_nat_entry(sbi, nid, &ne);
-	up_write(&nm_i->nat_tree_lock);
+	percpu_up_write(&nm_i->nat_tree_lock);
 }
 
 /*
@@ -1783,7 +1782,7 @@ void build_free_nids(struct f2fs_sb_info *sbi)
 	ra_meta_pages(sbi, NAT_BLOCK_OFFSET(nid), FREE_NID_PAGES,
 							META_NAT, true);
 
-	down_read(&nm_i->nat_tree_lock);
+	percpu_down_read(&nm_i->nat_tree_lock);
 
 	while (1) {
 		struct page *page = get_current_nat_page(sbi, nid);
@@ -1815,7 +1814,7 @@ void build_free_nids(struct f2fs_sb_info *sbi)
 			remove_free_nid(nm_i, nid);
 	}
 	up_read(&curseg->journal_rwsem);
-	up_read(&nm_i->nat_tree_lock);
+	percpu_up_read(&nm_i->nat_tree_lock);
 
 	ra_meta_pages(sbi, NAT_BLOCK_OFFSET(nm_i->next_scan_nid),
 					nm_i->ra_nid_pages, META_NAT, false);
@@ -2204,7 +2203,7 @@ void flush_nat_entries(struct f2fs_sb_info *sbi)
 	if (!nm_i->dirty_nat_cnt)
 		return;
 
-	down_write(&nm_i->nat_tree_lock);
+	percpu_down_write(&nm_i->nat_tree_lock);
 
 	/*
 	 * if there are no enough space in journal to store dirty nat
@@ -2227,7 +2226,7 @@ void flush_nat_entries(struct f2fs_sb_info *sbi)
 	list_for_each_entry_safe(set, tmp, &sets, set_list)
 		__flush_nat_entry_set(sbi, set);
 
-	up_write(&nm_i->nat_tree_lock);
+	percpu_up_write(&nm_i->nat_tree_lock);
 
 	f2fs_bug_on(sbi, nm_i->dirty_nat_cnt);
 }
@@ -2263,7 +2262,8 @@ static int init_node_manager(struct f2fs_sb_info *sbi)
 
 	mutex_init(&nm_i->build_lock);
 	spin_lock_init(&nm_i->free_nid_list_lock);
-	init_rwsem(&nm_i->nat_tree_lock);
+	if (percpu_init_rwsem(&nm_i->nat_tree_lock))
+		return -ENOMEM;
 
 	nm_i->next_scan_nid = le32_to_cpu(sbi->ckpt->next_free_nid);
 	nm_i->bitmap_size = __bitmap_size(sbi, NAT_BITMAP);
@@ -2320,7 +2320,7 @@ void destroy_node_manager(struct f2fs_sb_info *sbi)
 	spin_unlock(&nm_i->free_nid_list_lock);
 
 	/* destroy nat cache */
-	down_write(&nm_i->nat_tree_lock);
+	percpu_down_write(&nm_i->nat_tree_lock);
 	while ((found = __gang_lookup_nat_cache(nm_i,
 					nid, NATVEC_SIZE, natvec))) {
 		unsigned idx;
@@ -2345,8 +2345,9 @@ void destroy_node_manager(struct f2fs_sb_info *sbi)
 			kmem_cache_free(nat_entry_set_slab, setvec[idx]);
 		}
 	}
-	up_write(&nm_i->nat_tree_lock);
+	percpu_up_write(&nm_i->nat_tree_lock);
 
+	percpu_free_rwsem(&nm_i->nat_tree_lock);
 	kfree(nm_i->nat_bitmap);
 	sbi->nm_info = NULL;
 	kfree(nm_i);
diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c
index e0a975d..2bac917 100644
--- a/fs/f2fs/super.c
+++ b/fs/f2fs/super.c
@@ -695,6 +695,8 @@ static void destroy_percpu_info(struct f2fs_sb_info *sbi)
 		percpu_counter_destroy(&sbi->nr_pages[i]);
 	percpu_counter_destroy(&sbi->alloc_valid_block_count);
 	percpu_counter_destroy(&sbi->total_valid_inode_count);
+
+	percpu_free_rwsem(&sbi->cp_rwsem);
 }
 
 static void f2fs_put_super(struct super_block *sb)
@@ -1471,6 +1473,9 @@ static int init_percpu_info(struct f2fs_sb_info *sbi)
 {
 	int i, err;
 
+	if (percpu_init_rwsem(&sbi->cp_rwsem))
+		return -ENOMEM;
+
 	for (i = 0; i < NR_COUNT_TYPE; i++) {
 		err = percpu_counter_init(&sbi->nr_pages[i], 0, GFP_KERNEL);
 		if (err)
@@ -1671,7 +1676,6 @@ try_onemore:
 		sbi->write_io[i].bio = NULL;
 	}
 
-	init_rwsem(&sbi->cp_rwsem);
 	init_waitqueue_head(&sbi->cp_wait);
 	init_sb_info(sbi);
 
-- 
2.8.3


------------------------------------------------------------------------------
Attend Shape: An AT&T Tech Expo July 15-16. Meet us at AT&T Park in San
Francisco, CA to explore cutting-edge tech and listen to tech luminaries
present their vision of the future. This family event has something for
everyone, including kids. Get more information and register today.
http://sdm.link/attshape

[PATCH 6/6] f2fs: avoid mark_inode_dirty and update_inode

[Jaegeuk Kim]

Let's check inode's dirtiness before calling mark_inode_dirty and update_inode.

Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
---
 fs/f2fs/acl.c          |  2 +-
 fs/f2fs/dir.c          |  6 +++---
 fs/f2fs/extent_cache.c |  2 +-
 fs/f2fs/f2fs.h         | 25 +++++++++++++------------
 fs/f2fs/file.c         |  8 ++++----
 fs/f2fs/inline.c       |  2 +-
 fs/f2fs/inode.c        | 10 +++++++++-
 fs/f2fs/namei.c        |  6 +++---
 fs/f2fs/super.c        |  6 ++++--
 fs/f2fs/xattr.c        |  4 ++--
 10 files changed, 41 insertions(+), 30 deletions(-)

diff --git a/fs/f2fs/acl.c b/fs/f2fs/acl.c
index 6a414e7..4dcc9e2 100644
--- a/fs/f2fs/acl.c
+++ b/fs/f2fs/acl.c
@@ -384,7 +384,7 @@ int f2fs_init_acl(struct inode *inode, struct inode *dir, struct page *ipage,
 	if (error)
 		return error;
 
-	mark_inode_dirty_sync(inode);
+	f2fs_mark_inode_dirty_sync(inode);
 
 	if (default_acl) {
 		error = __f2fs_set_acl(inode, ACL_TYPE_DEFAULT, default_acl,
diff --git a/fs/f2fs/dir.c b/fs/f2fs/dir.c
index 7ba52a0..db4022f 100644
--- a/fs/f2fs/dir.c
+++ b/fs/f2fs/dir.c
@@ -300,7 +300,7 @@ void f2fs_set_link(struct inode *dir, struct f2fs_dir_entry *de,
 	set_page_dirty(page);
 
 	dir->i_mtime = dir->i_ctime = CURRENT_TIME;
-	mark_inode_dirty_sync(dir);
+	f2fs_mark_inode_dirty_sync(dir);
 	f2fs_put_page(page, 1);
 }
 
@@ -452,7 +452,7 @@ void update_parent_metadata(struct inode *dir, struct inode *inode,
 		clear_inode_flag(inode, FI_NEW_INODE);
 	}
 	dir->i_mtime = dir->i_ctime = CURRENT_TIME;
-	mark_inode_dirty_sync(dir);
+	f2fs_mark_inode_dirty_sync(dir);
 
 	if (F2FS_I(dir)->i_current_depth != current_depth)
 		f2fs_i_depth_write(dir, current_depth);
@@ -704,7 +704,7 @@ void f2fs_delete_entry(struct f2fs_dir_entry *dentry, struct page *page,
 	set_page_dirty(page);
 
 	dir->i_ctime = dir->i_mtime = CURRENT_TIME;
-	mark_inode_dirty_sync(dir);
+	f2fs_mark_inode_dirty_sync(dir);
 
 	if (inode)
 		f2fs_drop_nlink(dir, inode);
diff --git a/fs/f2fs/extent_cache.c b/fs/f2fs/extent_cache.c
index e858869..5b4b6d4 100644
--- a/fs/f2fs/extent_cache.c
+++ b/fs/f2fs/extent_cache.c
@@ -172,7 +172,7 @@ static void __drop_largest_extent(struct inode *inode,
 
 	if (fofs < largest->fofs + largest->len && fofs + len > largest->fofs) {
 		largest->len = 0;
-		mark_inode_dirty_sync(inode);
+		f2fs_mark_inode_dirty_sync(inode);
 	}
 }
 
diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h
index 2f59888..ee3ac0b 100644
--- a/fs/f2fs/f2fs.h
+++ b/fs/f2fs/f2fs.h
@@ -526,12 +526,13 @@ static inline bool __is_front_mergeable(struct extent_info *cur,
 	return __is_extent_mergeable(cur, front);
 }
 
+extern void f2fs_mark_inode_dirty_sync(struct inode *);
 static inline void __try_update_largest_extent(struct inode *inode,
 			struct extent_tree *et, struct extent_node *en)
 {
 	if (en->ei.len > et->largest.len) {
 		et->largest = en->ei;
-		mark_inode_dirty_sync(inode);
+		f2fs_mark_inode_dirty_sync(inode);
 	}
 }
 
@@ -1571,7 +1572,7 @@ static inline void __mark_inode_dirty_flag(struct inode *inode,
 			return;
 	case FI_DATA_EXIST:
 	case FI_INLINE_DOTS:
-		mark_inode_dirty_sync(inode);
+		f2fs_mark_inode_dirty_sync(inode);
 	}
 }
 
@@ -1598,7 +1599,7 @@ static inline void set_acl_inode(struct inode *inode, umode_t mode)
 {
 	F2FS_I(inode)->i_acl_mode = mode;
 	set_inode_flag(inode, FI_ACL_MODE);
-	mark_inode_dirty_sync(inode);
+	f2fs_mark_inode_dirty_sync(inode);
 }
 
 static inline void f2fs_i_links_write(struct inode *inode, bool inc)
@@ -1607,7 +1608,7 @@ static inline void f2fs_i_links_write(struct inode *inode, bool inc)
 		inc_nlink(inode);
 	else
 		drop_nlink(inode);
-	mark_inode_dirty_sync(inode);
+	f2fs_mark_inode_dirty_sync(inode);
 }
 
 static inline void f2fs_i_blocks_write(struct inode *inode,
@@ -1618,7 +1619,7 @@ static inline void f2fs_i_blocks_write(struct inode *inode,
 
 	inode->i_blocks = add ? inode->i_blocks + diff :
 				inode->i_blocks - diff;
-	mark_inode_dirty_sync(inode);
+	f2fs_mark_inode_dirty_sync(inode);
 	if (clean || recover)
 		set_inode_flag(inode, FI_AUTO_RECOVER);
 }
@@ -1632,7 +1633,7 @@ static inline void f2fs_i_size_write(struct inode *inode, loff_t i_size)
 		return;
 
 	i_size_write(inode, i_size);
-	mark_inode_dirty_sync(inode);
+	f2fs_mark_inode_dirty_sync(inode);
 	if (clean || recover)
 		set_inode_flag(inode, FI_AUTO_RECOVER);
 }
@@ -1647,19 +1648,19 @@ static inline bool f2fs_skip_inode_update(struct inode *inode)
 static inline void f2fs_i_depth_write(struct inode *inode, unsigned int depth)
 {
 	F2FS_I(inode)->i_current_depth = depth;
-	mark_inode_dirty_sync(inode);
+	f2fs_mark_inode_dirty_sync(inode);
 }
 
 static inline void f2fs_i_xnid_write(struct inode *inode, nid_t xnid)
 {
 	F2FS_I(inode)->i_xattr_nid = xnid;
-	mark_inode_dirty_sync(inode);
+	f2fs_mark_inode_dirty_sync(inode);
 }
 
 static inline void f2fs_i_pino_write(struct inode *inode, nid_t pino)
 {
 	F2FS_I(inode)->i_pino = pino;
-	mark_inode_dirty_sync(inode);
+	f2fs_mark_inode_dirty_sync(inode);
 }
 
 static inline void get_inline_info(struct inode *inode, struct f2fs_inode *ri)
@@ -1787,13 +1788,13 @@ static inline int is_file(struct inode *inode, int type)
 static inline void set_file(struct inode *inode, int type)
 {
 	F2FS_I(inode)->i_advise |= type;
-	mark_inode_dirty_sync(inode);
+	f2fs_mark_inode_dirty_sync(inode);
 }
 
 static inline void clear_file(struct inode *inode, int type)
 {
 	F2FS_I(inode)->i_advise &= ~type;
-	mark_inode_dirty_sync(inode);
+	f2fs_mark_inode_dirty_sync(inode);
 }
 
 static inline int f2fs_readonly(struct super_block *sb)
@@ -1940,7 +1941,7 @@ static inline int f2fs_add_link(struct dentry *dentry, struct inode *inode)
 /*
  * super.c
  */
-void f2fs_inode_synced(struct inode *);
+int f2fs_inode_synced(struct inode *);
 int f2fs_commit_super(struct f2fs_sb_info *, bool);
 int f2fs_sync_fs(struct super_block *, int);
 extern __printf(3, 4)
diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c
index 952cbcf..e6b770d 100644
--- a/fs/f2fs/file.c
+++ b/fs/f2fs/file.c
@@ -265,7 +265,7 @@ sync_nodes:
 	}
 
 	if (need_inode_block_update(sbi, ino)) {
-		mark_inode_dirty_sync(inode);
+		f2fs_mark_inode_dirty_sync(inode);
 		f2fs_write_inode(inode, NULL);
 		goto sync_nodes;
 	}
@@ -670,7 +670,7 @@ int f2fs_truncate(struct inode *inode)
 		return err;
 
 	inode->i_mtime = inode->i_ctime = CURRENT_TIME;
-	mark_inode_dirty_sync(inode);
+	f2fs_mark_inode_dirty_sync(inode);
 	return 0;
 }
 
@@ -760,7 +760,7 @@ int f2fs_setattr(struct dentry *dentry, struct iattr *attr)
 		}
 	}
 
-	mark_inode_dirty_sync(inode);
+	f2fs_mark_inode_dirty_sync(inode);
 	return err;
 }
 
@@ -1315,7 +1315,7 @@ static long f2fs_fallocate(struct file *file, int mode,
 
 	if (!ret) {
 		inode->i_mtime = inode->i_ctime = CURRENT_TIME;
-		mark_inode_dirty_sync(inode);
+		f2fs_mark_inode_dirty_sync(inode);
 		f2fs_update_time(F2FS_I_SB(inode), REQ_TIME);
 	}
 
diff --git a/fs/f2fs/inline.c b/fs/f2fs/inline.c
index 2cd0edc..d411ab6 100644
--- a/fs/f2fs/inline.c
+++ b/fs/f2fs/inline.c
@@ -569,7 +569,7 @@ void f2fs_delete_inline_entry(struct f2fs_dir_entry *dentry, struct page *page,
 	f2fs_put_page(page, 1);
 
 	dir->i_ctime = dir->i_mtime = CURRENT_TIME;
-	mark_inode_dirty_sync(dir);
+	f2fs_mark_inode_dirty_sync(dir);
 
 	if (inode)
 		f2fs_drop_nlink(dir, inode);
diff --git a/fs/f2fs/inode.c b/fs/f2fs/inode.c
index 9221ca2..a6e8eb0 100644
--- a/fs/f2fs/inode.c
+++ b/fs/f2fs/inode.c
@@ -18,6 +18,13 @@
 
 #include <trace/events/f2fs.h>
 
+void f2fs_mark_inode_dirty_sync(struct inode *inode)
+{
+	if (is_inode_flag_set(inode, FI_DIRTY_INODE))
+		return;
+	mark_inode_dirty_sync(inode);
+}
+
 void f2fs_set_inode_flags(struct inode *inode)
 {
 	unsigned int flags = F2FS_I(inode)->i_flags;
@@ -231,7 +238,8 @@ int update_inode(struct inode *inode, struct page *node_page)
 {
 	struct f2fs_inode *ri;
 
-	f2fs_inode_synced(inode);
+	if (f2fs_inode_synced(inode))
+		return 0;
 
 	f2fs_wait_on_page_writeback(node_page, NODE, true);
 
diff --git a/fs/f2fs/namei.c b/fs/f2fs/namei.c
index 4460400..c9ba6d7 100644
--- a/fs/f2fs/namei.c
+++ b/fs/f2fs/namei.c
@@ -757,7 +757,7 @@ static int f2fs_rename(struct inode *old_dir, struct dentry *old_dentry,
 	up_write(&F2FS_I(old_inode)->i_sem);
 
 	old_inode->i_ctime = CURRENT_TIME;
-	mark_inode_dirty_sync(old_inode);
+	f2fs_mark_inode_dirty_sync(old_inode);
 
 	f2fs_delete_entry(old_entry, old_page, old_dir, NULL);
 
@@ -909,7 +909,7 @@ static int f2fs_cross_rename(struct inode *old_dir, struct dentry *old_dentry,
 		f2fs_i_links_write(old_dir, old_nlink > 0);
 		up_write(&F2FS_I(old_dir)->i_sem);
 	}
-	mark_inode_dirty_sync(old_dir);
+	f2fs_mark_inode_dirty_sync(old_dir);
 
 	/* update directory entry info of new dir inode */
 	f2fs_set_link(new_dir, new_entry, new_page, old_inode);
@@ -924,7 +924,7 @@ static int f2fs_cross_rename(struct inode *old_dir, struct dentry *old_dentry,
 		f2fs_i_links_write(new_dir, new_nlink > 0);
 		up_write(&F2FS_I(new_dir)->i_sem);
 	}
-	mark_inode_dirty_sync(new_dir);
+	f2fs_mark_inode_dirty_sync(new_dir);
 
 	f2fs_unlock_op(sbi);
 
diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c
index 2bac917..e8756aa 100644
--- a/fs/f2fs/super.c
+++ b/fs/f2fs/super.c
@@ -658,14 +658,14 @@ static void f2fs_dirty_inode(struct inode *inode, int flags)
 	spin_unlock(&sbi->inode_lock[DIRTY_META]);
 }
 
-void f2fs_inode_synced(struct inode *inode)
+int f2fs_inode_synced(struct inode *inode)
 {
 	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
 
 	spin_lock(&sbi->inode_lock[DIRTY_META]);
 	if (!is_inode_flag_set(inode, FI_DIRTY_INODE)) {
 		spin_unlock(&sbi->inode_lock[DIRTY_META]);
-		return;
+		return 1;
 	}
 	list_del_init(&F2FS_I(inode)->gdirty_list);
 	clear_inode_flag(inode, FI_DIRTY_INODE);
@@ -673,6 +673,8 @@ void f2fs_inode_synced(struct inode *inode)
 	dec_page_count(sbi, F2FS_DIRTY_IMETA);
 	stat_dec_dirty_inode(F2FS_I_SB(inode), DIRTY_META);
 	spin_unlock(&sbi->inode_lock[DIRTY_META]);
+
+	return 0;
 }
 
 static void f2fs_i_callback(struct rcu_head *head)
diff --git a/fs/f2fs/xattr.c b/fs/f2fs/xattr.c
index e530254..5e9aced 100644
--- a/fs/f2fs/xattr.c
+++ b/fs/f2fs/xattr.c
@@ -106,7 +106,7 @@ static int f2fs_xattr_advise_set(const struct xattr_handler *handler,
 		return -EINVAL;
 
 	F2FS_I(inode)->i_advise |= *(char *)value;
-	mark_inode_dirty_sync(inode);
+	f2fs_mark_inode_dirty_sync(inode);
 	return 0;
 }
 
@@ -555,7 +555,7 @@ static int __f2fs_setxattr(struct inode *inode, int index,
 		else
 			file_clear_key(inode);
 	}
-	mark_inode_dirty_sync(inode);
+	f2fs_mark_inode_dirty_sync(inode);
 exit:
 	kzfree(base_addr);
 	return error;
-- 
2.8.3

[PATCH 6/6] f2fs: avoid mark_inode_dirty and update_inode

[Jaegeuk Kim]

Let's check inode's dirtiness before calling mark_inode_dirty and update_inode.

Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
---
 fs/f2fs/acl.c          |  2 +-
 fs/f2fs/dir.c          |  6 +++---
 fs/f2fs/extent_cache.c |  2 +-
 fs/f2fs/f2fs.h         | 25 +++++++++++++------------
 fs/f2fs/file.c         |  8 ++++----
 fs/f2fs/inline.c       |  2 +-
 fs/f2fs/inode.c        | 10 +++++++++-
 fs/f2fs/namei.c        |  6 +++---
 fs/f2fs/super.c        |  6 ++++--
 fs/f2fs/xattr.c        |  4 ++--
 10 files changed, 41 insertions(+), 30 deletions(-)

diff --git a/fs/f2fs/acl.c b/fs/f2fs/acl.c
index 6a414e7..4dcc9e2 100644
--- a/fs/f2fs/acl.c
+++ b/fs/f2fs/acl.c
@@ -384,7 +384,7 @@ int f2fs_init_acl(struct inode *inode, struct inode *dir, struct page *ipage,
 	if (error)
 		return error;
 
-	mark_inode_dirty_sync(inode);
+	f2fs_mark_inode_dirty_sync(inode);
 
 	if (default_acl) {
 		error = __f2fs_set_acl(inode, ACL_TYPE_DEFAULT, default_acl,
diff --git a/fs/f2fs/dir.c b/fs/f2fs/dir.c
index 7ba52a0..db4022f 100644
--- a/fs/f2fs/dir.c
+++ b/fs/f2fs/dir.c
@@ -300,7 +300,7 @@ void f2fs_set_link(struct inode *dir, struct f2fs_dir_entry *de,
 	set_page_dirty(page);
 
 	dir->i_mtime = dir->i_ctime = CURRENT_TIME;
-	mark_inode_dirty_sync(dir);
+	f2fs_mark_inode_dirty_sync(dir);
 	f2fs_put_page(page, 1);
 }
 
@@ -452,7 +452,7 @@ void update_parent_metadata(struct inode *dir, struct inode *inode,
 		clear_inode_flag(inode, FI_NEW_INODE);
 	}
 	dir->i_mtime = dir->i_ctime = CURRENT_TIME;
-	mark_inode_dirty_sync(dir);
+	f2fs_mark_inode_dirty_sync(dir);
 
 	if (F2FS_I(dir)->i_current_depth != current_depth)
 		f2fs_i_depth_write(dir, current_depth);
@@ -704,7 +704,7 @@ void f2fs_delete_entry(struct f2fs_dir_entry *dentry, struct page *page,
 	set_page_dirty(page);
 
 	dir->i_ctime = dir->i_mtime = CURRENT_TIME;
-	mark_inode_dirty_sync(dir);
+	f2fs_mark_inode_dirty_sync(dir);
 
 	if (inode)
 		f2fs_drop_nlink(dir, inode);
diff --git a/fs/f2fs/extent_cache.c b/fs/f2fs/extent_cache.c
index e858869..5b4b6d4 100644
--- a/fs/f2fs/extent_cache.c
+++ b/fs/f2fs/extent_cache.c
@@ -172,7 +172,7 @@ static void __drop_largest_extent(struct inode *inode,
 
 	if (fofs < largest->fofs + largest->len && fofs + len > largest->fofs) {
 		largest->len = 0;
-		mark_inode_dirty_sync(inode);
+		f2fs_mark_inode_dirty_sync(inode);
 	}
 }
 
diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h
index 2f59888..ee3ac0b 100644
--- a/fs/f2fs/f2fs.h
+++ b/fs/f2fs/f2fs.h
@@ -526,12 +526,13 @@ static inline bool __is_front_mergeable(struct extent_info *cur,
 	return __is_extent_mergeable(cur, front);
 }
 
+extern void f2fs_mark_inode_dirty_sync(struct inode *);
 static inline void __try_update_largest_extent(struct inode *inode,
 			struct extent_tree *et, struct extent_node *en)
 {
 	if (en->ei.len > et->largest.len) {
 		et->largest = en->ei;
-		mark_inode_dirty_sync(inode);
+		f2fs_mark_inode_dirty_sync(inode);
 	}
 }
 
@@ -1571,7 +1572,7 @@ static inline void __mark_inode_dirty_flag(struct inode *inode,
 			return;
 	case FI_DATA_EXIST:
 	case FI_INLINE_DOTS:
-		mark_inode_dirty_sync(inode);
+		f2fs_mark_inode_dirty_sync(inode);
 	}
 }
 
@@ -1598,7 +1599,7 @@ static inline void set_acl_inode(struct inode *inode, umode_t mode)
 {
 	F2FS_I(inode)->i_acl_mode = mode;
 	set_inode_flag(inode, FI_ACL_MODE);
-	mark_inode_dirty_sync(inode);
+	f2fs_mark_inode_dirty_sync(inode);
 }
 
 static inline void f2fs_i_links_write(struct inode *inode, bool inc)
@@ -1607,7 +1608,7 @@ static inline void f2fs_i_links_write(struct inode *inode, bool inc)
 		inc_nlink(inode);
 	else
 		drop_nlink(inode);
-	mark_inode_dirty_sync(inode);
+	f2fs_mark_inode_dirty_sync(inode);
 }
 
 static inline void f2fs_i_blocks_write(struct inode *inode,
@@ -1618,7 +1619,7 @@ static inline void f2fs_i_blocks_write(struct inode *inode,
 
 	inode->i_blocks = add ? inode->i_blocks + diff :
 				inode->i_blocks - diff;
-	mark_inode_dirty_sync(inode);
+	f2fs_mark_inode_dirty_sync(inode);
 	if (clean || recover)
 		set_inode_flag(inode, FI_AUTO_RECOVER);
 }
@@ -1632,7 +1633,7 @@ static inline void f2fs_i_size_write(struct inode *inode, loff_t i_size)
 		return;
 
 	i_size_write(inode, i_size);
-	mark_inode_dirty_sync(inode);
+	f2fs_mark_inode_dirty_sync(inode);
 	if (clean || recover)
 		set_inode_flag(inode, FI_AUTO_RECOVER);
 }
@@ -1647,19 +1648,19 @@ static inline bool f2fs_skip_inode_update(struct inode *inode)
 static inline void f2fs_i_depth_write(struct inode *inode, unsigned int depth)
 {
 	F2FS_I(inode)->i_current_depth = depth;
-	mark_inode_dirty_sync(inode);
+	f2fs_mark_inode_dirty_sync(inode);
 }
 
 static inline void f2fs_i_xnid_write(struct inode *inode, nid_t xnid)
 {
 	F2FS_I(inode)->i_xattr_nid = xnid;
-	mark_inode_dirty_sync(inode);
+	f2fs_mark_inode_dirty_sync(inode);
 }
 
 static inline void f2fs_i_pino_write(struct inode *inode, nid_t pino)
 {
 	F2FS_I(inode)->i_pino = pino;
-	mark_inode_dirty_sync(inode);
+	f2fs_mark_inode_dirty_sync(inode);
 }
 
 static inline void get_inline_info(struct inode *inode, struct f2fs_inode *ri)
@@ -1787,13 +1788,13 @@ static inline int is_file(struct inode *inode, int type)
 static inline void set_file(struct inode *inode, int type)
 {
 	F2FS_I(inode)->i_advise |= type;
-	mark_inode_dirty_sync(inode);
+	f2fs_mark_inode_dirty_sync(inode);
 }
 
 static inline void clear_file(struct inode *inode, int type)
 {
 	F2FS_I(inode)->i_advise &= ~type;
-	mark_inode_dirty_sync(inode);
+	f2fs_mark_inode_dirty_sync(inode);
 }
 
 static inline int f2fs_readonly(struct super_block *sb)
@@ -1940,7 +1941,7 @@ static inline int f2fs_add_link(struct dentry *dentry, struct inode *inode)
 /*
  * super.c
  */
-void f2fs_inode_synced(struct inode *);
+int f2fs_inode_synced(struct inode *);
 int f2fs_commit_super(struct f2fs_sb_info *, bool);
 int f2fs_sync_fs(struct super_block *, int);
 extern __printf(3, 4)
diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c
index 952cbcf..e6b770d 100644
--- a/fs/f2fs/file.c
+++ b/fs/f2fs/file.c
@@ -265,7 +265,7 @@ sync_nodes:
 	}
 
 	if (need_inode_block_update(sbi, ino)) {
-		mark_inode_dirty_sync(inode);
+		f2fs_mark_inode_dirty_sync(inode);
 		f2fs_write_inode(inode, NULL);
 		goto sync_nodes;
 	}
@@ -670,7 +670,7 @@ int f2fs_truncate(struct inode *inode)
 		return err;
 
 	inode->i_mtime = inode->i_ctime = CURRENT_TIME;
-	mark_inode_dirty_sync(inode);
+	f2fs_mark_inode_dirty_sync(inode);
 	return 0;
 }
 
@@ -760,7 +760,7 @@ int f2fs_setattr(struct dentry *dentry, struct iattr *attr)
 		}
 	}
 
-	mark_inode_dirty_sync(inode);
+	f2fs_mark_inode_dirty_sync(inode);
 	return err;
 }
 
@@ -1315,7 +1315,7 @@ static long f2fs_fallocate(struct file *file, int mode,
 
 	if (!ret) {
 		inode->i_mtime = inode->i_ctime = CURRENT_TIME;
-		mark_inode_dirty_sync(inode);
+		f2fs_mark_inode_dirty_sync(inode);
 		f2fs_update_time(F2FS_I_SB(inode), REQ_TIME);
 	}
 
diff --git a/fs/f2fs/inline.c b/fs/f2fs/inline.c
index 2cd0edc..d411ab6 100644
--- a/fs/f2fs/inline.c
+++ b/fs/f2fs/inline.c
@@ -569,7 +569,7 @@ void f2fs_delete_inline_entry(struct f2fs_dir_entry *dentry, struct page *page,
 	f2fs_put_page(page, 1);
 
 	dir->i_ctime = dir->i_mtime = CURRENT_TIME;
-	mark_inode_dirty_sync(dir);
+	f2fs_mark_inode_dirty_sync(dir);
 
 	if (inode)
 		f2fs_drop_nlink(dir, inode);
diff --git a/fs/f2fs/inode.c b/fs/f2fs/inode.c
index 9221ca2..a6e8eb0 100644
--- a/fs/f2fs/inode.c
+++ b/fs/f2fs/inode.c
@@ -18,6 +18,13 @@
 
 #include <trace/events/f2fs.h>
 
+void f2fs_mark_inode_dirty_sync(struct inode *inode)
+{
+	if (is_inode_flag_set(inode, FI_DIRTY_INODE))
+		return;
+	mark_inode_dirty_sync(inode);
+}
+
 void f2fs_set_inode_flags(struct inode *inode)
 {
 	unsigned int flags = F2FS_I(inode)->i_flags;
@@ -231,7 +238,8 @@ int update_inode(struct inode *inode, struct page *node_page)
 {
 	struct f2fs_inode *ri;
 
-	f2fs_inode_synced(inode);
+	if (f2fs_inode_synced(inode))
+		return 0;
 
 	f2fs_wait_on_page_writeback(node_page, NODE, true);
 
diff --git a/fs/f2fs/namei.c b/fs/f2fs/namei.c
index 4460400..c9ba6d7 100644
--- a/fs/f2fs/namei.c
+++ b/fs/f2fs/namei.c
@@ -757,7 +757,7 @@ static int f2fs_rename(struct inode *old_dir, struct dentry *old_dentry,
 	up_write(&F2FS_I(old_inode)->i_sem);
 
 	old_inode->i_ctime = CURRENT_TIME;
-	mark_inode_dirty_sync(old_inode);
+	f2fs_mark_inode_dirty_sync(old_inode);
 
 	f2fs_delete_entry(old_entry, old_page, old_dir, NULL);
 
@@ -909,7 +909,7 @@ static int f2fs_cross_rename(struct inode *old_dir, struct dentry *old_dentry,
 		f2fs_i_links_write(old_dir, old_nlink > 0);
 		up_write(&F2FS_I(old_dir)->i_sem);
 	}
-	mark_inode_dirty_sync(old_dir);
+	f2fs_mark_inode_dirty_sync(old_dir);
 
 	/* update directory entry info of new dir inode */
 	f2fs_set_link(new_dir, new_entry, new_page, old_inode);
@@ -924,7 +924,7 @@ static int f2fs_cross_rename(struct inode *old_dir, struct dentry *old_dentry,
 		f2fs_i_links_write(new_dir, new_nlink > 0);
 		up_write(&F2FS_I(new_dir)->i_sem);
 	}
-	mark_inode_dirty_sync(new_dir);
+	f2fs_mark_inode_dirty_sync(new_dir);
 
 	f2fs_unlock_op(sbi);
 
diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c
index 2bac917..e8756aa 100644
--- a/fs/f2fs/super.c
+++ b/fs/f2fs/super.c
@@ -658,14 +658,14 @@ static void f2fs_dirty_inode(struct inode *inode, int flags)
 	spin_unlock(&sbi->inode_lock[DIRTY_META]);
 }
 
-void f2fs_inode_synced(struct inode *inode)
+int f2fs_inode_synced(struct inode *inode)
 {
 	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
 
 	spin_lock(&sbi->inode_lock[DIRTY_META]);
 	if (!is_inode_flag_set(inode, FI_DIRTY_INODE)) {
 		spin_unlock(&sbi->inode_lock[DIRTY_META]);
-		return;
+		return 1;
 	}
 	list_del_init(&F2FS_I(inode)->gdirty_list);
 	clear_inode_flag(inode, FI_DIRTY_INODE);
@@ -673,6 +673,8 @@ void f2fs_inode_synced(struct inode *inode)
 	dec_page_count(sbi, F2FS_DIRTY_IMETA);
 	stat_dec_dirty_inode(F2FS_I_SB(inode), DIRTY_META);
 	spin_unlock(&sbi->inode_lock[DIRTY_META]);
+
+	return 0;
 }
 
 static void f2fs_i_callback(struct rcu_head *head)
diff --git a/fs/f2fs/xattr.c b/fs/f2fs/xattr.c
index e530254..5e9aced 100644
--- a/fs/f2fs/xattr.c
+++ b/fs/f2fs/xattr.c
@@ -106,7 +106,7 @@ static int f2fs_xattr_advise_set(const struct xattr_handler *handler,
 		return -EINVAL;
 
 	F2FS_I(inode)->i_advise |= *(char *)value;
-	mark_inode_dirty_sync(inode);
+	f2fs_mark_inode_dirty_sync(inode);
 	return 0;
 }
 
@@ -555,7 +555,7 @@ static int __f2fs_setxattr(struct inode *inode, int index,
 		else
 			file_clear_key(inode);
 	}
-	mark_inode_dirty_sync(inode);
+	f2fs_mark_inode_dirty_sync(inode);
 exit:
 	kzfree(base_addr);
 	return error;
-- 
2.8.3


------------------------------------------------------------------------------
Attend Shape: An AT&T Tech Expo July 15-16. Meet us at AT&T Park in San
Francisco, CA to explore cutting-edge tech and listen to tech luminaries
present their vision of the future. This family event has something for
everyone, including kids. Get more information and register today.
http://sdm.link/attshape