Skip to contentSkip to conversation

f2fs: fix wrong error injection for evict_inode

PATCH In review Jaegeuk Kim 2 participants 9 messages Mar 7, 2017 – Mar 9, 2017

Subsystems (2)
VFS F2FS file system
lore
Share & tools
lore

Jaegeuk Kim

author maintainer
#0 lore raw

PATCH 1/4 f2fs: fix wrong error injection for evict_inode

The previous one was not a proper location to inject an error, since there is no point to get errors. Instead, we can emulate EIO during truncation, and the below logic should handle it correctly.

Signed-off-by Jaegeuk Kim <[email protected]>
---
 fs/f2fs/inode.c | 13 ++++++-------
 1 file changed, 6 insertions(+), 7 deletions(-)
fs/f2fs/inode.c +6 -7
index ef8610bf950f..2520fa72b23f 100644
@@ -372,13 +372,6 @@ void f2fs_evict_inode(struct inode *inode)
 	if (inode->i_nlink || is_bad_inode(inode))
 		goto no_delete;
 
-#ifdef CONFIG_F2FS_FAULT_INJECTION
-	if (time_to_inject(sbi, FAULT_EVICT_INODE)) {
-		f2fs_show_injection_info(FAULT_EVICT_INODE);
-		goto no_delete;
-	}
-#endif
-
 	remove_ino_entry(sbi, inode->i_ino, APPEND_INO);
 	remove_ino_entry(sbi, inode->i_ino, UPDATE_INO);
 

@@ -389,6 +382,12 @@ void f2fs_evict_inode(struct inode *inode)
 	if (F2FS_HAS_BLOCKS(inode))
 		err = f2fs_truncate(inode);
 
+#ifdef CONFIG_F2FS_FAULT_INJECTION
+	if (time_to_inject(sbi, FAULT_EVICT_INODE)) {
+		f2fs_show_injection_info(FAULT_EVICT_INODE);
+		err = -EIO;
+	}
+#endif
 	if (!err) {
 		f2fs_lock_op(sbi);
 		err = remove_inode_page(inode);
Signature · 9 lines
2.11.0


------------------------------------------------------------------------------
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
author maintainer
#1 lore raw

PATCH 2/4 f2fs: build stat_info before orphan inode recovery

f2fs_sync_fs() -> write_checkpoint() calls stat_inc_cp_count(sbi->stat_info), which needs stat_info allocation.
Otherwise, we can hit:

[254042.598623]  ? count_shadow_nodes+0xa0/0xa0
[254042.598633]  f2fs_sync_fs+0x65/0xd0 [f2fs]
[254042.598645]  f2fs_balance_fs_bg+0xe4/0x1c0 [f2fs]
[254042.598657]  f2fs_write_node_pages+0x34/0x1a0 [f2fs]
[254042.598664]  ? pagevec_lookup_entries+0x1e/0x30
[254042.598673]  do_writepages+0x1e/0x30
[254042.598682]  __writeback_single_inode+0x45/0x330
[254042.598688]  writeback_single_inode+0xd7/0x190
[254042.598694]  write_inode_now+0x86/0xa0
[254042.598699]  iput+0x122/0x200
[254042.598709]  f2fs_fill_super+0xd4a/0x14d0 [f2fs]
[254042.598717]  mount_bdev+0x184/0x1c0
[254042.598934]  ? f2fs_commit_super+0x100/0x100 [f2fs]
[254042.599142]  f2fs_mount+0x15/0x20 [f2fs]
[254042.599349]  mount_fs+0x39/0x160
[254042.599554]  ? __alloc_percpu+0x15/0x20
[254042.599759]  vfs_kern_mount+0x67/0x110
[254042.599972]  do_mount+0x1bb/0xc80
[254042.600175]  ? memdup_user+0x42/0x60
[254042.600380]  SyS_mount+0x83/0xd0
[254042.600583]  entry_SYSCALL_64_fastpath+0x1e/0xad
Signed-off-by Jaegeuk Kim <[email protected]>
---
 fs/f2fs/super.c | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)
fs/f2fs/super.c +5 -5
index 96fe8ed73100..cfb40d3fd875 100644
@@ -2022,6 +2022,10 @@ static int f2fs_fill_super(struct super_block *sb, void *data, int silent)
 
 	f2fs_join_shrinker(sbi);
 
+	err = f2fs_build_stats(sbi);
+	if (err)
+		goto free_nm;
+
 	/* if there are nt orphan nodes free them */
 	err = recover_orphan_inodes(sbi);
 	if (err)

@@ -2046,10 +2050,6 @@ static int f2fs_fill_super(struct super_block *sb, void *data, int silent)
 		goto free_root_inode;
 	}
 
-	err = f2fs_build_stats(sbi);
-	if (err)
-		goto free_root_inode;
-
 	if (f2fs_proc_root)
 		sbi->s_proc = proc_mkdir(sb->s_id, f2fs_proc_root);
 

@@ -2143,7 +2143,6 @@ static int f2fs_fill_super(struct super_block *sb, void *data, int silent)
 		remove_proc_entry("segment_bits", sbi->s_proc);
 		remove_proc_entry(sb->s_id, f2fs_proc_root);
 	}
-	f2fs_destroy_stats(sbi);
 free_root_inode:
 	dput(sb->s_root);
 	sb->s_root = NULL;

@@ -2161,6 +2160,7 @@ static int f2fs_fill_super(struct super_block *sb, void *data, int silent)
 	truncate_inode_pages_final(META_MAPPING(sbi));
 	iput(sbi->node_inode);
 	mutex_unlock(&sbi->umount_mutex);
+	f2fs_destroy_stats(sbi);
 free_nm:
 	destroy_node_manager(sbi);
 free_sm:
Signature · 9 lines
2.11.0


------------------------------------------------------------------------------
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
#2 lore raw

Re: [f2fs-dev] [PATCH 2/4] f2fs: build stat_info before orphan inode recovery

Quoted 36 lines · Jaegeuk Kim

f2fs_sync_fs() -> write_checkpoint() calls stat_inc_cp_count(sbi->stat_info), which needs stat_info allocation.
Otherwise, we can hit:

[254042.598623]  ? count_shadow_nodes+0xa0/0xa0
[254042.598633]  f2fs_sync_fs+0x65/0xd0 [f2fs]
[254042.598645]  f2fs_balance_fs_bg+0xe4/0x1c0 [f2fs]
[254042.598657]  f2fs_write_node_pages+0x34/0x1a0 [f2fs]
[254042.598664]  ? pagevec_lookup_entries+0x1e/0x30
[254042.598673]  do_writepages+0x1e/0x30
[254042.598682]  __writeback_single_inode+0x45/0x330
[254042.598688]  writeback_single_inode+0xd7/0x190
[254042.598694]  write_inode_now+0x86/0xa0
[254042.598699]  iput+0x122/0x200
[254042.598709]  f2fs_fill_super+0xd4a/0x14d0 [f2fs]
[254042.598717]  mount_bdev+0x184/0x1c0
[254042.598934]  ? f2fs_commit_super+0x100/0x100 [f2fs]
[254042.599142]  f2fs_mount+0x15/0x20 [f2fs]
[254042.599349]  mount_fs+0x39/0x160
[254042.599554]  ? __alloc_percpu+0x15/0x20
[254042.599759]  vfs_kern_mount+0x67/0x110
[254042.599972]  do_mount+0x1bb/0xc80
[254042.600175]  ? memdup_user+0x42/0x60
[254042.600380]  SyS_mount+0x83/0xd0
[254042.600583]  entry_SYSCALL_64_fastpath+0x1e/0xad
Signed-off-by Jaegeuk Kim <[email protected]>
---
 fs/f2fs/super.c | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)
fs/f2fs/super.c +0 -0
index 96fe8ed73100..cfb40d3fd875 100644

fs/f2fs/super.c static int f2fs_fill_super(struct super_block *sb, void *data, int silent) · line 2027

Quoted 6 lines
fs/f2fs/super.c +3 -0
@@ -2022,6 +2022,10 @@ static int f2fs_fill_super(struct super_block *sb, void *data, int silent)
 
 	f2fs_join_shrinker(sbi);
 
+	err = f2fs_build_stats(sbi);
+	if (err)
+		goto free_nm;

Both node_inode and shrinker need to be released and detached.

How about relocate f2fs_build_stats in front of build_segment_manager?

Thanks,

Quoted 32 lines
fs/f2fs/super.c +2 -5
+
 	/* if there are nt orphan nodes free them */
 	err = recover_orphan_inodes(sbi);
 	if (err)

@@ -2046,10 +2050,6 @@ static int f2fs_fill_super(struct super_block *sb, void *data, int silent)
 		goto free_root_inode;
 	}
 
-	err = f2fs_build_stats(sbi);
-	if (err)
-		goto free_root_inode;
-
 	if (f2fs_proc_root)
 		sbi->s_proc = proc_mkdir(sb->s_id, f2fs_proc_root);
 

@@ -2143,7 +2143,6 @@ static int f2fs_fill_super(struct super_block *sb, void *data, int silent)
 		remove_proc_entry("segment_bits", sbi->s_proc);
 		remove_proc_entry(sb->s_id, f2fs_proc_root);
 	}
-	f2fs_destroy_stats(sbi);
 free_root_inode:
 	dput(sb->s_root);
 	sb->s_root = NULL;

@@ -2161,6 +2160,7 @@ static int f2fs_fill_super(struct super_block *sb, void *data, int silent)
 	truncate_inode_pages_final(META_MAPPING(sbi));
 	iput(sbi->node_inode);
 	mutex_unlock(&sbi->umount_mutex);
+	f2fs_destroy_stats(sbi);
 free_nm:
 	destroy_node_manager(sbi);
 free_sm:
author maintainer
#3 lore raw

Re: [f2fs-dev] [PATCH 2/4] f2fs: build stat_info before orphan inode recovery

Quoted 40 lines · Chao Yu
Quoted 39 lines · Jaegeuk Kim

f2fs_sync_fs() -> write_checkpoint() calls stat_inc_cp_count(sbi->stat_info), which needs stat_info allocation.
Otherwise, we can hit:

[254042.598623]  ? count_shadow_nodes+0xa0/0xa0
[254042.598633]  f2fs_sync_fs+0x65/0xd0 [f2fs]
[254042.598645]  f2fs_balance_fs_bg+0xe4/0x1c0 [f2fs]
[254042.598657]  f2fs_write_node_pages+0x34/0x1a0 [f2fs]
[254042.598664]  ? pagevec_lookup_entries+0x1e/0x30
[254042.598673]  do_writepages+0x1e/0x30
[254042.598682]  __writeback_single_inode+0x45/0x330
[254042.598688]  writeback_single_inode+0xd7/0x190
[254042.598694]  write_inode_now+0x86/0xa0
[254042.598699]  iput+0x122/0x200
[254042.598709]  f2fs_fill_super+0xd4a/0x14d0 [f2fs]
[254042.598717]  mount_bdev+0x184/0x1c0
[254042.598934]  ? f2fs_commit_super+0x100/0x100 [f2fs]
[254042.599142]  f2fs_mount+0x15/0x20 [f2fs]
[254042.599349]  mount_fs+0x39/0x160
[254042.599554]  ? __alloc_percpu+0x15/0x20
[254042.599759]  vfs_kern_mount+0x67/0x110
[254042.599972]  do_mount+0x1bb/0xc80
[254042.600175]  ? memdup_user+0x42/0x60
[254042.600380]  SyS_mount+0x83/0xd0
[254042.600583]  entry_SYSCALL_64_fastpath+0x1e/0xad
Signed-off-by Jaegeuk Kim <[email protected]>
---
 fs/f2fs/super.c | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)
fs/f2fs/super.c +0 -0
index 96fe8ed73100..cfb40d3fd875 100644
@@ -2022,6 +2022,10 @@ static int f2fs_fill_super(struct super_block *sb, void *data, int silent)
 
 	f2fs_join_shrinker(sbi);
Quoted 7 lines

fs/f2fs/super.c static int f2fs_fill_super(struct super_block *sb, void *data, int silent) · line 2027

Quoted 3 lines
fs/f2fs/super.c +3 -0
+	err = f2fs_build_stats(sbi);
+	if (err)
+		goto free_nm;

Both node_inode and shrinker need to be released and detached.

How about relocate f2fs_build_stats in front of build_segment_manager?

I just got another panic to try this, since sys/../f2fs/status needs them.

Thanks,

Quoted 35 lines

Thanks,

Quoted 32 lines
fs/f2fs/super.c +2 -5
+
 	/* if there are nt orphan nodes free them */
 	err = recover_orphan_inodes(sbi);
 	if (err)

@@ -2046,10 +2050,6 @@ static int f2fs_fill_super(struct super_block *sb, void *data, int silent)
 		goto free_root_inode;
 	}
 
-	err = f2fs_build_stats(sbi);
-	if (err)
-		goto free_root_inode;
-
 	if (f2fs_proc_root)
 		sbi->s_proc = proc_mkdir(sb->s_id, f2fs_proc_root);
 

@@ -2143,7 +2143,6 @@ static int f2fs_fill_super(struct super_block *sb, void *data, int silent)
 		remove_proc_entry("segment_bits", sbi->s_proc);
 		remove_proc_entry(sb->s_id, f2fs_proc_root);
 	}
-	f2fs_destroy_stats(sbi);
 free_root_inode:
 	dput(sb->s_root);
 	sb->s_root = NULL;

@@ -2161,6 +2160,7 @@ static int f2fs_fill_super(struct super_block *sb, void *data, int silent)
 	truncate_inode_pages_final(META_MAPPING(sbi));
 	iput(sbi->node_inode);
 	mutex_unlock(&sbi->umount_mutex);
+	f2fs_destroy_stats(sbi);
 free_nm:
 	destroy_node_manager(sbi);
 free_sm:
author maintainer
#4 lore raw

Re: [PATCH 2/4 v2] f2fs: build stat_info before orphan inode recovery

In reply to Jaegeuk Kim

Change log from v1:
o move its location up more

Quoted 1 line

From e993253fef6c4a07a65e6038c2c0cae684844047 Mon Sep 17 00:00:00 2001

From: Jaegeuk Kim <[email protected]>
Date: Tue, 7 Mar 2017 13:41:22 -0800
Subject: [PATCH] f2fs: build stat_info before orphan inode recovery

f2fs_sync_fs() -> write_checkpoint() calls stat_inc_cp_count(sbi->stat_info), which needs stat_info allocation.
Otherwise, we can hit:

[254042.598623]  ? count_shadow_nodes+0xa0/0xa0
[254042.598633]  f2fs_sync_fs+0x65/0xd0 [f2fs]
[254042.598645]  f2fs_balance_fs_bg+0xe4/0x1c0 [f2fs]
[254042.598657]  f2fs_write_node_pages+0x34/0x1a0 [f2fs]
[254042.598664]  ? pagevec_lookup_entries+0x1e/0x30
[254042.598673]  do_writepages+0x1e/0x30
[254042.598682]  __writeback_single_inode+0x45/0x330
[254042.598688]  writeback_single_inode+0xd7/0x190
[254042.598694]  write_inode_now+0x86/0xa0
[254042.598699]  iput+0x122/0x200
[254042.598709]  f2fs_fill_super+0xd4a/0x14d0 [f2fs]
[254042.598717]  mount_bdev+0x184/0x1c0
[254042.598934]  ? f2fs_commit_super+0x100/0x100 [f2fs]
[254042.599142]  f2fs_mount+0x15/0x20 [f2fs]
[254042.599349]  mount_fs+0x39/0x160
[254042.599554]  ? __alloc_percpu+0x15/0x20
[254042.599759]  vfs_kern_mount+0x67/0x110
[254042.599972]  do_mount+0x1bb/0xc80
[254042.600175]  ? memdup_user+0x42/0x60
[254042.600380]  SyS_mount+0x83/0xd0
[254042.600583]  entry_SYSCALL_64_fastpath+0x1e/0xad
Signed-off-by Jaegeuk Kim <[email protected]>
---
 fs/f2fs/super.c | 16 ++++++++--------
 1 file changed, 8 insertions(+), 8 deletions(-)
fs/f2fs/super.c +8 -8
index 96fe8ed73100..939cb7bfa7bc 100644
@@ -789,9 +789,6 @@ static void f2fs_put_super(struct super_block *sb)
 	/* be sure to wait for any on-going discard commands */
 	f2fs_wait_discard_bio(sbi, NULL_ADDR);
 
-	/* write_checkpoint can update stat informaion */
-	f2fs_destroy_stats(sbi);
-
 	/*
 	 * normally superblock is clean, so we need to release this.
 	 * In addition, EIO will skip do checkpoint, we need this as well.

@@ -811,6 +808,9 @@ static void f2fs_put_super(struct super_block *sb)
 	destroy_node_manager(sbi);
 	destroy_segment_manager(sbi);
 
+	/* write_checkpoint can update stat informaion */
+	f2fs_destroy_stats(sbi);
+
 	kfree(sbi->ckpt);
 	kobject_put(&sbi->s_kobj);
 	wait_for_completion(&sbi->s_kobj_unregister);

@@ -1985,6 +1985,10 @@ static int f2fs_fill_super(struct super_block *sb, void *data, int silent)
 
 	init_ino_entry_info(sbi);
 
+	err = f2fs_build_stats(sbi);
+	if (err)
+		goto free_devices;
+
 	/* setup f2fs internal modules */
 	err = build_segment_manager(sbi);
 	if (err) {

@@ -2046,10 +2050,6 @@ static int f2fs_fill_super(struct super_block *sb, void *data, int silent)
 		goto free_root_inode;
 	}
 
-	err = f2fs_build_stats(sbi);
-	if (err)
-		goto free_root_inode;
-
 	if (f2fs_proc_root)
 		sbi->s_proc = proc_mkdir(sb->s_id, f2fs_proc_root);
 

@@ -2143,7 +2143,6 @@ static int f2fs_fill_super(struct super_block *sb, void *data, int silent)
 		remove_proc_entry("segment_bits", sbi->s_proc);
 		remove_proc_entry(sb->s_id, f2fs_proc_root);
 	}
-	f2fs_destroy_stats(sbi);
 free_root_inode:
 	dput(sb->s_root);
 	sb->s_root = NULL;

@@ -2165,6 +2164,7 @@ static int f2fs_fill_super(struct super_block *sb, void *data, int silent)
 	destroy_node_manager(sbi);
 free_sm:
 	destroy_segment_manager(sbi);
+	f2fs_destroy_stats(sbi);
 free_devices:
 	destroy_device_list(sbi);
 	kfree(sbi->ckpt);
Signature · 1 line
2.11.0
author maintainer
#5 lore raw

PATCH 3/4 f2fs: show more precise message on orphan recovery failure

In reply to Jaegeuk Kim

This case is not caused by fsck.f2fs. User needs to retry mount.

Signed-off-by Jaegeuk Kim <[email protected]>
---
 fs/f2fs/checkpoint.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
fs/f2fs/checkpoint.c +1 -1
index 91734cc13f71..36ccc30ed54b 100644
@@ -568,7 +568,7 @@ static int recover_orphan_inode(struct f2fs_sb_info *sbi, nid_t ino)
 	if (ni.blk_addr != NULL_ADDR) {
 		set_sbi_flag(sbi, SBI_NEED_FSCK);
 		f2fs_msg(sbi->sb, KERN_WARNING,
-			"%s: orphan failed (ino=%x), run fsck to fix.",
+			"%s: orphan failed (ino=%x) by kernel, retry mount.",
 				__func__, ino);
 		return -EIO;
 	}
Signature · 9 lines
2.11.0


------------------------------------------------------------------------------
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
author maintainer
#6 lore raw

PATCH 4/4 f2fs: use __set{__claer}_bit_le

In reply to Jaegeuk Kim

This patch uses __set{__clear}_bit_le for highter speed.

Signed-off-by Jaegeuk Kim <[email protected]>
---
 fs/f2fs/dir.c  |  2 +-
 fs/f2fs/node.c | 16 ++++++++--------
 2 files changed, 9 insertions(+), 9 deletions(-)
fs/f2fs/dir.c +1 -1
index 4650c9b85de7..8d5c62b07b28 100644
@@ -750,7 +750,7 @@ void f2fs_delete_entry(struct f2fs_dir_entry *dentry, struct page *page,
 	dentry_blk = page_address(page);
 	bit_pos = dentry - dentry_blk->dentry;
 	for (i = 0; i < slots; i++)
-		clear_bit_le(bit_pos + i, &dentry_blk->dentry_bitmap);
+		__clear_bit_le(bit_pos + i, &dentry_blk->dentry_bitmap);
 
 	/* Let's check and deallocate this dentry page */
 	bit_pos = find_next_bit_le(&dentry_blk->dentry_bitmap,
fs/f2fs/node.c +8 -8
index 342742754a09..3377a512e299 100644
@@ -1825,9 +1825,9 @@ void update_free_nid_bitmap(struct f2fs_sb_info *sbi, nid_t nid,
 		return;
 
 	if (set)
-		set_bit_le(nid_ofs, nm_i->free_nid_bitmap[nat_ofs]);
+		__set_bit_le(nid_ofs, nm_i->free_nid_bitmap[nat_ofs]);
 	else
-		clear_bit_le(nid_ofs, nm_i->free_nid_bitmap[nat_ofs]);
+		__clear_bit_le(nid_ofs, nm_i->free_nid_bitmap[nat_ofs]);
 
 	spin_lock(&nm_i->free_nid_lock);
 	if (set)

@@ -1849,7 +1849,7 @@ static void scan_nat_page(struct f2fs_sb_info *sbi,
 	if (test_bit_le(nat_ofs, nm_i->nat_block_bitmap))
 		return;
 
-	set_bit_le(nat_ofs, nm_i->nat_block_bitmap);
+	__set_bit_le(nat_ofs, nm_i->nat_block_bitmap);
 
 	i = start_nid % NAT_ENTRY_PER_BLOCK;
 

@@ -2339,16 +2339,16 @@ void __update_nat_bits(struct f2fs_sb_info *sbi, nid_t start_nid,
 			valid++;
 	}
 	if (valid == 0) {
-		set_bit_le(nat_index, nm_i->empty_nat_bits);
-		clear_bit_le(nat_index, nm_i->full_nat_bits);
+		__set_bit_le(nat_index, nm_i->empty_nat_bits);
+		__clear_bit_le(nat_index, nm_i->full_nat_bits);
 		return;
 	}
 
-	clear_bit_le(nat_index, nm_i->empty_nat_bits);
+	__clear_bit_le(nat_index, nm_i->empty_nat_bits);
 	if (valid == NAT_ENTRY_PER_BLOCK)
-		set_bit_le(nat_index, nm_i->full_nat_bits);
+		__set_bit_le(nat_index, nm_i->full_nat_bits);
 	else
-		clear_bit_le(nat_index, nm_i->full_nat_bits);
+		__clear_bit_le(nat_index, nm_i->full_nat_bits);
 }
 
 static void __flush_nat_entry_set(struct f2fs_sb_info *sbi,
Signature · 9 lines
2.11.0


------------------------------------------------------------------------------
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
Reviewed-by on v1 4/4
#7 lore raw

Re: [PATCH 4/4] f2fs: use __set{__claer}_bit_le

Quoted 3 lines · Jaegeuk Kim

This patch uses __set{__clear}_bit_le for highter speed.

Signed-off-by Jaegeuk Kim <[email protected]>
Reviewed-by Chao Yu <[email protected]>

------------------------------------------------------------------------------
Announcing the Oxford Dictionaries API! The API offers world-renowned dictionary content that is easy and intuitive to access. Sign up for an account today to start using our lexical data to power your apps and projects. Get started today and enter our developer competition.
http://sdm.link/oxford

#8 lore raw
In reply to Jaegeuk Kim

Hi Jaegeuk,

Quoted 28 lines · Jaegeuk Kim

The previous one was not a proper location to inject an error, since there is no point to get errors. Instead, we can emulate EIO during truncation, and the below logic should handle it correctly.

Signed-off-by Jaegeuk Kim <[email protected]>
---
 fs/f2fs/inode.c | 13 ++++++-------
 1 file changed, 6 insertions(+), 7 deletions(-)
fs/f2fs/inode.c +0 -7
index ef8610bf950f..2520fa72b23f 100644
@@ -372,13 +372,6 @@ void f2fs_evict_inode(struct inode *inode)
 	if (inode->i_nlink || is_bad_inode(inode))
 		goto no_delete;
 
-#ifdef CONFIG_F2FS_FAULT_INJECTION
-	if (time_to_inject(sbi, FAULT_EVICT_INODE)) {
-		f2fs_show_injection_info(FAULT_EVICT_INODE);
-		goto no_delete;
-	}
-#endif
-
 	remove_ino_entry(sbi, inode->i_ino, APPEND_INO);
 	remove_ino_entry(sbi, inode->i_ino, UPDATE_INO);

fs/f2fs/inode.c void f2fs_evict_inode(struct inode *inode) · line 382

Quoted 1 line
fs/f2fs/inode.c +0 -0
@@ -389,6 +382,12 @@ void f2fs_evict_inode(struct inode *inode)
 	if (F2FS_HAS_BLOCKS(inode))

In addition, how about add below code to emulate error came from f2fs_truncate?

	if (F2FS_HAS_BLOCKS(inode)) {
#ifdef CONFIG_F2FS_FAULT_INJECTION
		if (time_to_inject(sbi, FAULT_EVICT_INODE)) {
			f2fs_show_injection_info(FAULT_EVICT_INODE);
			err = -EIO;
		}
#endif
		if (!err)
			err = f2fs_truncate(inode);
	}

Thanks,

Quoted 12 lines
fs/f2fs/inode.c +6 -0
 		err = f2fs_truncate(inode);
 
+#ifdef CONFIG_F2FS_FAULT_INJECTION
+	if (time_to_inject(sbi, FAULT_EVICT_INODE)) {
+		f2fs_show_injection_info(FAULT_EVICT_INODE);
+		err = -EIO;
+	}
+#endif
 	if (!err) {
 		f2fs_lock_op(sbi);
 		err = remove_inode_page(inode);