* [PATCH v4 02/20] ext4: add handling for extended mount options
2019-12-24 8:13 [PATCH v4 01/20] ext4: update docs for fast commit feature Harshad Shirwadkar
@ 2019-12-24 8:13 ` Harshad Shirwadkar
2019-12-24 8:13 ` [PATCH v4 03/20] ext4, jbd2: add fast commit initialization routines Harshad Shirwadkar
` (18 subsequent siblings)
19 siblings, 0 replies; 31+ messages in thread
From: Harshad Shirwadkar @ 2019-12-24 8:13 UTC (permalink / raw)
To: linux-ext4; +Cc: Harshad Shirwadkar
We are running out of mount option bits. Add handling for using
s_mount_opt2. Add ability to turn on / off the fast commit
feature and to turn on / off fast commit soft consistency option.
Signed-off-by: Harshad Shirwadkar <harshadshirwadkar@gmail.com>
---
fs/ext4/ext4.h | 7 +++++++
fs/ext4/super.c | 23 +++++++++++++++++++----
2 files changed, 26 insertions(+), 4 deletions(-)
diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index 03db3e71676c..9339eb6bf9b0 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -1170,6 +1170,13 @@ struct ext4_inode_info {
#define EXT4_MOUNT2_EXPLICIT_JOURNAL_CHECKSUM 0x00000008 /* User explicitly
specified journal checksum */
+#define EXT4_MOUNT2_JOURNAL_FAST_COMMIT 0x00000010 /* Journal fast commit */
+
+#define EXT4_MOUNT2_JOURNAL_FC_SOFT_CONSISTENCY 0x00000020 /* Soft consistency
+ * mode for fast
+ * commits
+ */
+
#define clear_opt(sb, opt) EXT4_SB(sb)->s_mount_opt &= \
~EXT4_MOUNT_##opt
#define set_opt(sb, opt) EXT4_SB(sb)->s_mount_opt |= \
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index dd654e53ba3d..d635040f21b9 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -1459,6 +1459,7 @@ enum {
Opt_dioread_nolock, Opt_dioread_lock,
Opt_discard, Opt_nodiscard, Opt_init_itable, Opt_noinit_itable,
Opt_max_dir_size_kb, Opt_nojournal_checksum, Opt_nombcache,
+ Opt_no_fc, Opt_fc_soft_consistency
};
static const match_table_t tokens = {
@@ -1541,6 +1542,8 @@ static const match_table_t tokens = {
{Opt_init_itable, "init_itable=%u"},
{Opt_init_itable, "init_itable"},
{Opt_noinit_itable, "noinit_itable"},
+ {Opt_no_fc, "no_fc"},
+ {Opt_fc_soft_consistency, "fc_soft_consistency"},
{Opt_max_dir_size_kb, "max_dir_size_kb=%u"},
{Opt_test_dummy_encryption, "test_dummy_encryption"},
{Opt_nombcache, "nombcache"},
@@ -1663,6 +1666,7 @@ static int clear_qf_name(struct super_block *sb, int qtype)
#define MOPT_NO_EXT3 0x0200
#define MOPT_EXT4_ONLY (MOPT_NO_EXT2 | MOPT_NO_EXT3)
#define MOPT_STRING 0x0400
+#define MOPT_2 0x0800
static const struct mount_opts {
int token;
@@ -1755,6 +1759,10 @@ static const struct mount_opts {
{Opt_max_dir_size_kb, 0, MOPT_GTE0},
{Opt_test_dummy_encryption, 0, MOPT_GTE0},
{Opt_nombcache, EXT4_MOUNT_NO_MBCACHE, MOPT_SET},
+ {Opt_no_fc, EXT4_MOUNT2_JOURNAL_FAST_COMMIT,
+ MOPT_CLEAR | MOPT_2 | MOPT_EXT4_ONLY},
+ {Opt_fc_soft_consistency, EXT4_MOUNT2_JOURNAL_FC_SOFT_CONSISTENCY,
+ MOPT_SET | MOPT_2 | MOPT_EXT4_ONLY},
{Opt_err, 0, 0}
};
@@ -2038,10 +2046,17 @@ static int handle_mount_opt(struct super_block *sb, char *opt, int token,
WARN_ON(1);
return -1;
}
- if (arg != 0)
- sbi->s_mount_opt |= m->mount_opt;
- else
- sbi->s_mount_opt &= ~m->mount_opt;
+ if (m->flags & MOPT_2) {
+ if (arg != 0)
+ sbi->s_mount_opt2 |= m->mount_opt;
+ else
+ sbi->s_mount_opt2 &= ~m->mount_opt;
+ } else {
+ if (arg != 0)
+ sbi->s_mount_opt |= m->mount_opt;
+ else
+ sbi->s_mount_opt &= ~m->mount_opt;
+ }
}
return 1;
}
--
2.24.1.735.g03f4e72817-goog
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PATCH v4 03/20] ext4, jbd2: add fast commit initialization routines
2019-12-24 8:13 [PATCH v4 01/20] ext4: update docs for fast commit feature Harshad Shirwadkar
2019-12-24 8:13 ` [PATCH v4 02/20] ext4: add handling for extended mount options Harshad Shirwadkar
@ 2019-12-24 8:13 ` Harshad Shirwadkar
2019-12-24 8:13 ` [PATCH v4 04/20] jbd2: add fast commit block tracker variables Harshad Shirwadkar
` (17 subsequent siblings)
19 siblings, 0 replies; 31+ messages in thread
From: Harshad Shirwadkar @ 2019-12-24 8:13 UTC (permalink / raw)
To: linux-ext4; +Cc: Harshad Shirwadkar
Define feature flags for fast commits and add routines to allow ext4 to
initialize fast commits. Note that we allow 128 blocks to be used for
fast commits. As of now, that's the default constant value.
Signed-off-by: Harshad Shirwadkar <harshadshirwadkar@gmail.com>
---
fs/ext4/ext4.h | 2 ++
fs/ext4/ext4_jbd2.c | 6 ++++++
fs/ext4/ext4_jbd2.h | 13 +++++++++++++
fs/ext4/super.c | 5 +++++
fs/jbd2/journal.c | 11 +++++++++++
include/linux/jbd2.h | 19 ++++++++++++++++++-
6 files changed, 55 insertions(+), 1 deletion(-)
diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index 9339eb6bf9b0..16ffe6ed9e74 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -1685,6 +1685,7 @@ static inline bool ext4_verity_in_progress(struct inode *inode)
#define EXT4_FEATURE_COMPAT_RESIZE_INODE 0x0010
#define EXT4_FEATURE_COMPAT_DIR_INDEX 0x0020
#define EXT4_FEATURE_COMPAT_SPARSE_SUPER2 0x0200
+#define EXT4_FEATURE_COMPAT_FAST_COMMIT 0x0400
#define EXT4_FEATURE_RO_COMPAT_SPARSE_SUPER 0x0001
#define EXT4_FEATURE_RO_COMPAT_LARGE_FILE 0x0002
@@ -1786,6 +1787,7 @@ EXT4_FEATURE_COMPAT_FUNCS(xattr, EXT_ATTR)
EXT4_FEATURE_COMPAT_FUNCS(resize_inode, RESIZE_INODE)
EXT4_FEATURE_COMPAT_FUNCS(dir_index, DIR_INDEX)
EXT4_FEATURE_COMPAT_FUNCS(sparse_super2, SPARSE_SUPER2)
+EXT4_FEATURE_COMPAT_FUNCS(fast_commit, FAST_COMMIT)
EXT4_FEATURE_RO_COMPAT_FUNCS(sparse_super, SPARSE_SUPER)
EXT4_FEATURE_RO_COMPAT_FUNCS(large_file, LARGE_FILE)
diff --git a/fs/ext4/ext4_jbd2.c b/fs/ext4/ext4_jbd2.c
index 7c70b08d104c..c2ae21f5049b 100644
--- a/fs/ext4/ext4_jbd2.c
+++ b/fs/ext4/ext4_jbd2.c
@@ -330,3 +330,9 @@ int __ext4_handle_dirty_super(const char *where, unsigned int line,
mark_buffer_dirty(bh);
return err;
}
+void ext4_init_fast_commit(struct super_block *sb, journal_t *journal)
+{
+ if (!ext4_should_fast_commit(sb))
+ return;
+ jbd2_init_fast_commit(journal, EXT4_NUM_FC_BLKS);
+}
diff --git a/fs/ext4/ext4_jbd2.h b/fs/ext4/ext4_jbd2.h
index ef8fcf7d0d3b..e5bd95a088e8 100644
--- a/fs/ext4/ext4_jbd2.h
+++ b/fs/ext4/ext4_jbd2.h
@@ -378,6 +378,17 @@ static inline int ext4_jbd2_inode_add_wait(handle_t *handle,
return 0;
}
+static inline int ext4_should_fast_commit(struct super_block *sb)
+{
+ if (!ext4_has_feature_fast_commit(sb))
+ return 0;
+ if (!test_opt2(sb, JOURNAL_FAST_COMMIT))
+ return 0;
+ if (test_opt(sb, QUOTA))
+ return 0;
+ return 1;
+}
+
static inline void ext4_update_inode_fsync_trans(handle_t *handle,
struct inode *inode,
int datasync)
@@ -459,4 +470,6 @@ static inline int ext4_should_dioread_nolock(struct inode *inode)
return 1;
}
+#define EXT4_NUM_FC_BLKS 128
+void ext4_init_fast_commit(struct super_block *sb, journal_t *journal);
#endif /* _EXT4_JBD2_H */
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index d635040f21b9..28675cd78813 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -3759,6 +3759,9 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
#ifdef CONFIG_EXT4_FS_POSIX_ACL
set_opt(sb, POSIX_ACL);
#endif
+ if (ext4_has_feature_fast_commit(sb))
+ set_opt2(sb, JOURNAL_FAST_COMMIT);
+
/* don't forget to enable journal_csum when metadata_csum is enabled. */
if (ext4_has_metadata_csum(sb))
set_opt(sb, JOURNAL_CHECKSUM);
@@ -4376,6 +4379,7 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
sbi->s_def_mount_opt &= ~EXT4_MOUNT_JOURNAL_CHECKSUM;
clear_opt(sb, JOURNAL_CHECKSUM);
clear_opt(sb, DATA_FLAGS);
+ clear_opt2(sb, JOURNAL_FAST_COMMIT);
sbi->s_journal = NULL;
needs_recovery = 0;
goto no_journal;
@@ -4734,6 +4738,7 @@ static void ext4_init_journal_params(struct super_block *sb, journal_t *journal)
journal->j_commit_interval = sbi->s_commit_interval;
journal->j_min_batch_time = sbi->s_min_batch_time;
journal->j_max_batch_time = sbi->s_max_batch_time;
+ ext4_init_fast_commit(sb, journal);
write_lock(&journal->j_state_lock);
if (test_opt(sb, BARRIER))
diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c
index 1c58859aa592..fa22bdb7d952 100644
--- a/fs/jbd2/journal.c
+++ b/fs/jbd2/journal.c
@@ -1181,6 +1181,14 @@ static journal_t *journal_init_common(struct block_device *bdev,
return NULL;
}
+void jbd2_init_fast_commit(journal_t *journal, int num_fc_blks)
+{
+ journal->j_fc_wbufsize = num_fc_blks;
+ journal->j_wbufsize = journal->j_blocksize / sizeof(journal_block_tag_t)
+ - journal->j_fc_wbufsize;
+ journal->j_fc_wbuf = &journal->j_wbuf[journal->j_wbufsize];
+}
+
/* jbd2_journal_init_dev and jbd2_journal_init_inode:
*
* Create a journal structure assigned some fixed set of disk blocks to
@@ -1682,6 +1690,9 @@ int jbd2_journal_load(journal_t *journal)
return -EFSCORRUPTED;
}
+ if (journal->j_fc_wbufsize > 0)
+ jbd2_journal_set_features(journal, 0, 0,
+ JBD2_FEATURE_INCOMPAT_FAST_COMMIT);
/* OK, we've finished with the dynamic journal bits:
* reinitialise the dynamic contents of the superblock in memory
* and reset them on disk. */
diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h
index 603fbc4e2f70..3495fe1e2c36 100644
--- a/include/linux/jbd2.h
+++ b/include/linux/jbd2.h
@@ -288,6 +288,7 @@ typedef struct journal_superblock_s
#define JBD2_FEATURE_INCOMPAT_ASYNC_COMMIT 0x00000004
#define JBD2_FEATURE_INCOMPAT_CSUM_V2 0x00000008
#define JBD2_FEATURE_INCOMPAT_CSUM_V3 0x00000010
+#define JBD2_FEATURE_INCOMPAT_FAST_COMMIT 0x00000020
/* See "journal feature predicate functions" below */
@@ -298,7 +299,8 @@ typedef struct journal_superblock_s
JBD2_FEATURE_INCOMPAT_64BIT | \
JBD2_FEATURE_INCOMPAT_ASYNC_COMMIT | \
JBD2_FEATURE_INCOMPAT_CSUM_V2 | \
- JBD2_FEATURE_INCOMPAT_CSUM_V3)
+ JBD2_FEATURE_INCOMPAT_CSUM_V3 | \
+ JBD2_FEATURE_INCOMPAT_FAST_COMMIT)
#ifdef __KERNEL__
@@ -1059,6 +1061,12 @@ struct journal_s
*/
struct buffer_head **j_wbuf;
+ /**
+ * @j_fc_wbuf: Array of fast commit bhs for
+ * jbd2_journal_commit_transaction.
+ */
+ struct buffer_head **j_fc_wbuf;
+
/**
* @j_wbufsize:
*
@@ -1066,6 +1074,13 @@ struct journal_s
*/
int j_wbufsize;
+ /**
+ * @j_fc_wbufsize:
+ *
+ * Size of @j_fc_wbuf array.
+ */
+ int j_fc_wbufsize;
+
/**
* @j_last_sync_writer:
*
@@ -1235,6 +1250,7 @@ JBD2_FEATURE_INCOMPAT_FUNCS(64bit, 64BIT)
JBD2_FEATURE_INCOMPAT_FUNCS(async_commit, ASYNC_COMMIT)
JBD2_FEATURE_INCOMPAT_FUNCS(csum2, CSUM_V2)
JBD2_FEATURE_INCOMPAT_FUNCS(csum3, CSUM_V3)
+JBD2_FEATURE_INCOMPAT_FUNCS(fast_commit, FAST_COMMIT)
/*
* Journal flag definitions
@@ -1500,6 +1516,7 @@ void __jbd2_log_wait_for_space(journal_t *journal);
extern void __jbd2_journal_drop_transaction(journal_t *, transaction_t *);
extern int jbd2_cleanup_journal_tail(journal_t *);
+void jbd2_init_fast_commit(journal_t *journal, int num_fc_blks);
/*
* is_journal_abort
*
--
2.24.1.735.g03f4e72817-goog
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PATCH v4 04/20] jbd2: add fast commit block tracker variables
2019-12-24 8:13 [PATCH v4 01/20] ext4: update docs for fast commit feature Harshad Shirwadkar
2019-12-24 8:13 ` [PATCH v4 02/20] ext4: add handling for extended mount options Harshad Shirwadkar
2019-12-24 8:13 ` [PATCH v4 03/20] ext4, jbd2: add fast commit initialization routines Harshad Shirwadkar
@ 2019-12-24 8:13 ` Harshad Shirwadkar
2019-12-24 8:13 ` [PATCH v4 05/20] jbd2: disable fast commits if journal is empty Harshad Shirwadkar
` (16 subsequent siblings)
19 siblings, 0 replies; 31+ messages in thread
From: Harshad Shirwadkar @ 2019-12-24 8:13 UTC (permalink / raw)
To: linux-ext4; +Cc: Harshad Shirwadkar
Add j_first_fc, j_last_fc and j_fc_offset variables to track fast commit
area. j_first_fc and j_last_fc mark the start and the end of the area,
while j_fc_offset points to the last used block in the region.
Signed-off-by: Harshad Shirwadkar <harshadshirwadkar@gmail.com>
---
fs/jbd2/journal.c | 33 ++++++++++++++++++++++++++++-----
include/linux/jbd2.h | 24 ++++++++++++++++++++++++
2 files changed, 52 insertions(+), 5 deletions(-)
diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c
index fa22bdb7d952..32f14be5065a 100644
--- a/fs/jbd2/journal.c
+++ b/fs/jbd2/journal.c
@@ -1163,6 +1163,11 @@ static journal_t *journal_init_common(struct block_device *bdev,
if (!journal->j_wbuf)
goto err_cleanup;
+ if (journal->j_fc_wbufsize > 0) {
+ journal->j_wbufsize = n - journal->j_fc_wbufsize;
+ journal->j_fc_wbuf = &journal->j_wbuf[journal->j_wbufsize];
+ }
+
bh = getblk_unmovable(journal->j_dev, start, journal->j_blocksize);
if (!bh) {
pr_err("%s: Cannot get buffer for journal superblock\n",
@@ -1303,11 +1308,20 @@ static int journal_reset(journal_t *journal)
}
journal->j_first = first;
- journal->j_last = last;
- journal->j_head = first;
- journal->j_tail = first;
- journal->j_free = last - first;
+ if (jbd2_has_feature_fast_commit(journal) &&
+ journal->j_fc_wbufsize > 0) {
+ journal->j_last_fc = last;
+ journal->j_last = last - journal->j_fc_wbufsize;
+ journal->j_first_fc = journal->j_last + 1;
+ journal->j_fc_off = 0;
+ } else {
+ journal->j_last = last;
+ }
+
+ journal->j_head = journal->j_first;
+ journal->j_tail = journal->j_first;
+ journal->j_free = journal->j_last - journal->j_first;
journal->j_tail_sequence = journal->j_transaction_sequence;
journal->j_commit_sequence = journal->j_transaction_sequence - 1;
@@ -1632,9 +1646,18 @@ static int load_superblock(journal_t *journal)
journal->j_tail_sequence = be32_to_cpu(sb->s_sequence);
journal->j_tail = be32_to_cpu(sb->s_start);
journal->j_first = be32_to_cpu(sb->s_first);
- journal->j_last = be32_to_cpu(sb->s_maxlen);
journal->j_errno = be32_to_cpu(sb->s_errno);
+ if (jbd2_has_feature_fast_commit(journal) &&
+ journal->j_fc_wbufsize > 0) {
+ journal->j_last_fc = be32_to_cpu(sb->s_maxlen);
+ journal->j_last = journal->j_last_fc - journal->j_fc_wbufsize;
+ journal->j_first_fc = journal->j_last + 1;
+ journal->j_fc_off = 0;
+ } else {
+ journal->j_last = be32_to_cpu(sb->s_maxlen);
+ }
+
return 0;
}
diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h
index 3495fe1e2c36..7139626992f3 100644
--- a/include/linux/jbd2.h
+++ b/include/linux/jbd2.h
@@ -918,6 +918,30 @@ struct journal_s
*/
unsigned long j_last;
+ /**
+ * @j_first_fc:
+ *
+ * The block number of the first fast commit block in the journal
+ * [j_state_lock].
+ */
+ unsigned long j_first_fc;
+
+ /**
+ * @j_fc_off:
+ *
+ * Number of fast commit blocks currently allocated.
+ * [j_state_lock].
+ */
+ unsigned long j_fc_off;
+
+ /**
+ * @j_last_fc:
+ *
+ * The block number one beyond the last fast commit block in the journal
+ * [j_state_lock].
+ */
+ unsigned long j_last_fc;
+
/**
* @j_dev: Device where we store the journal.
*/
--
2.24.1.735.g03f4e72817-goog
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PATCH v4 05/20] jbd2: disable fast commits if journal is empty
2019-12-24 8:13 [PATCH v4 01/20] ext4: update docs for fast commit feature Harshad Shirwadkar
` (2 preceding siblings ...)
2019-12-24 8:13 ` [PATCH v4 04/20] jbd2: add fast commit block tracker variables Harshad Shirwadkar
@ 2019-12-24 8:13 ` Harshad Shirwadkar
2019-12-24 8:13 ` [PATCH v4 06/20] jbd2: fast commit main commit path changes Harshad Shirwadkar
` (15 subsequent siblings)
19 siblings, 0 replies; 31+ messages in thread
From: Harshad Shirwadkar @ 2019-12-24 8:13 UTC (permalink / raw)
To: linux-ext4; +Cc: Harshad Shirwadkar
If journal is empty, clear the fast commit flag from the on disk
superblock. With this optimization, chances of running into backward
compatibility issues are reduced.
Signed-off-by: Harshad Shirwadkar <harshadshirwadkar@gmail.com>
---
fs/jbd2/journal.c | 12 ++++++++++++
1 file changed, 12 insertions(+)
diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c
index 32f14be5065a..7d91f5204366 100644
--- a/fs/jbd2/journal.c
+++ b/fs/jbd2/journal.c
@@ -1464,6 +1464,7 @@ int jbd2_journal_update_sb_log_tail(journal_t *journal, tid_t tail_tid,
static void jbd2_mark_journal_empty(journal_t *journal, int write_op)
{
journal_superblock_t *sb = journal->j_superblock;
+ bool had_fast_commit = false;
BUG_ON(!mutex_is_locked(&journal->j_checkpoint_mutex));
lock_buffer(journal->j_sb_buffer);
@@ -1477,9 +1478,20 @@ static void jbd2_mark_journal_empty(journal_t *journal, int write_op)
sb->s_sequence = cpu_to_be32(journal->j_tail_sequence);
sb->s_start = cpu_to_be32(0);
+ if (jbd2_has_feature_fast_commit(journal)) {
+ /*
+ * When journal is clean, no need to commit fast commit flag and
+ * make file system incompatible with older kernels.
+ */
+ jbd2_clear_feature_fast_commit(journal);
+ had_fast_commit = true;
+ }
jbd2_write_superblock(journal, write_op);
+ if (had_fast_commit)
+ jbd2_set_feature_fast_commit(journal);
+
/* Log is no longer empty */
write_lock(&journal->j_state_lock);
journal->j_flags |= JBD2_FLUSHED;
--
2.24.1.735.g03f4e72817-goog
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PATCH v4 06/20] jbd2: fast commit main commit path changes
2019-12-24 8:13 [PATCH v4 01/20] ext4: update docs for fast commit feature Harshad Shirwadkar
` (3 preceding siblings ...)
2019-12-24 8:13 ` [PATCH v4 05/20] jbd2: disable fast commits if journal is empty Harshad Shirwadkar
@ 2019-12-24 8:13 ` Harshad Shirwadkar
2019-12-24 8:13 ` [PATCH v4 07/20] ext4: add generic diff tracking routines and range tracking Harshad Shirwadkar
` (14 subsequent siblings)
19 siblings, 0 replies; 31+ messages in thread
From: Harshad Shirwadkar @ 2019-12-24 8:13 UTC (permalink / raw)
To: linux-ext4; +Cc: Harshad Shirwadkar
Add 3 new APIs jbd2_start_async_fc_nowait(),
jbd2_start_async_fc_wait() and jbd2_stop_async_fc(). These APIs can be
used by file systems to indicate to jbd2 that they are starting or
stopping a fast commit.
Signed-off-by: Harshad Shirwadkar <harshadshirwadkar@gmail.com>
---
fs/jbd2/commit.c | 21 +++++++++++
fs/jbd2/journal.c | 85 +++++++++++++++++++++++++++++++++++++++++++-
include/linux/jbd2.h | 21 +++++++++++
3 files changed, 126 insertions(+), 1 deletion(-)
diff --git a/fs/jbd2/commit.c b/fs/jbd2/commit.c
index 132fb92098c7..ffbea070582c 100644
--- a/fs/jbd2/commit.c
+++ b/fs/jbd2/commit.c
@@ -413,6 +413,23 @@ void jbd2_journal_commit_transaction(journal_t *journal)
J_ASSERT(journal->j_running_transaction != NULL);
J_ASSERT(journal->j_committing_transaction == NULL);
+ write_lock(&journal->j_state_lock);
+ journal->j_flags |= JBD2_FULL_COMMIT_ONGOING;
+ while (journal->j_flags & JBD2_FAST_COMMIT_ONGOING) {
+ DEFINE_WAIT(wait);
+
+ prepare_to_wait(&journal->j_wait_async_fc, &wait,
+ TASK_UNINTERRUPTIBLE);
+ write_unlock(&journal->j_state_lock);
+ schedule();
+ write_lock(&journal->j_state_lock);
+ finish_wait(&journal->j_wait_async_fc, &wait);
+ }
+ write_unlock(&journal->j_state_lock);
+
+ if (journal->j_fc_cleanup_callback)
+ journal->j_fc_cleanup_callback(journal);
+
commit_transaction = journal->j_running_transaction;
trace_jbd2_start_commit(journal, commit_transaction);
@@ -420,6 +437,7 @@ void jbd2_journal_commit_transaction(journal_t *journal)
commit_transaction->t_tid);
write_lock(&journal->j_state_lock);
+ journal->j_fc_off = 0;
J_ASSERT(commit_transaction->t_state == T_RUNNING);
commit_transaction->t_state = T_LOCKED;
@@ -1114,6 +1132,8 @@ void jbd2_journal_commit_transaction(journal_t *journal)
journal->j_commit_sequence, journal->j_tail_sequence);
write_lock(&journal->j_state_lock);
+ journal->j_flags &= ~JBD2_FULL_COMMIT_ONGOING;
+ journal->j_flags &= ~JBD2_FAST_COMMIT_ONGOING;
spin_lock(&journal->j_list_lock);
commit_transaction->t_state = T_FINISHED;
/* Check if the transaction can be dropped now that we are finished */
@@ -1125,6 +1145,7 @@ void jbd2_journal_commit_transaction(journal_t *journal)
spin_unlock(&journal->j_list_lock);
write_unlock(&journal->j_state_lock);
wake_up(&journal->j_wait_done_commit);
+ wake_up(&journal->j_wait_async_fc);
/*
* Calculate overall stats
diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c
index 7d91f5204366..bd6ab127f0d9 100644
--- a/fs/jbd2/journal.c
+++ b/fs/jbd2/journal.c
@@ -158,7 +158,9 @@ static void commit_timeout(struct timer_list *t)
*
* 1) COMMIT: Every so often we need to commit the current state of the
* filesystem to disk. The journal thread is responsible for writing
- * all of the metadata buffers to disk.
+ * all of the metadata buffers to disk. If a fast commit is ongoing
+ * journal thread waits until it's done and then copntinues from
+ * there on.
*
* 2) CHECKPOINT: We cannot reuse a used section of the log file until all
* of the data in that part of the log has been rewritten elsewhere on
@@ -715,6 +717,86 @@ int jbd2_log_wait_commit(journal_t *journal, tid_t tid)
return err;
}
+/*
+ * Returns 0 if async fc could be started. Returns -EINVAL if no full
+ * commit has been done yet. Returns -EALREADY if another fast /
+ * full commit is ongoing.
+ */
+int jbd2_start_async_fc_nowait(journal_t *journal, tid_t tid)
+{
+ /*
+ * Fast commits only allowed if at least one full commit has
+ * been processed.
+ */
+ if (!journal->j_stats.ts_tid)
+ return -EINVAL;
+
+ if (tid <= journal->j_commit_sequence)
+ return -EALREADY;
+
+ write_lock(&journal->j_state_lock);
+ if (journal->j_flags &
+ (JBD2_FAST_COMMIT_ONGOING | JBD2_FULL_COMMIT_ONGOING)) {
+ write_unlock(&journal->j_state_lock);
+ return -EALREADY;
+ }
+
+ journal->j_flags |= JBD2_FAST_COMMIT_ONGOING;
+ write_unlock(&journal->j_state_lock);
+
+ return 0;
+}
+
+/*
+ * Same as above but waits for any ongoing fast commits to complete.
+ * If a full commit is ongoing, this function returns with
+ * -EALREADY.
+ */
+int jbd2_start_async_fc_wait(journal_t *journal, tid_t tid)
+{
+ int ret;
+
+ /*
+ * Fast commits only allowed if at least one full commit has
+ * been processed.
+ */
+ if (!journal->j_stats.ts_tid)
+ return -EINVAL;
+
+ if (tid <= journal->j_commit_sequence)
+ return -EALREADY;
+
+ write_lock(&journal->j_state_lock);
+restart:
+ if (journal->j_flags & JBD2_FULL_COMMIT_ONGOING) {
+ ret = -EALREADY;
+ } else if (journal->j_flags & JBD2_FAST_COMMIT_ONGOING) {
+ DEFINE_WAIT(wait);
+
+ prepare_to_wait(&journal->j_wait_async_fc, &wait,
+ TASK_UNINTERRUPTIBLE);
+ write_unlock(&journal->j_state_lock);
+ schedule();
+ write_lock(&journal->j_state_lock);
+ finish_wait(&journal->j_wait_async_fc, &wait);
+ goto restart;
+ } else {
+ journal->j_flags |= JBD2_FAST_COMMIT_ONGOING;
+ ret = 0;
+ }
+ write_unlock(&journal->j_state_lock);
+
+ return ret;
+}
+
+void jbd2_stop_async_fc(journal_t *journal, tid_t tid)
+{
+ write_lock(&journal->j_state_lock);
+ journal->j_flags &= ~JBD2_FAST_COMMIT_ONGOING;
+ write_unlock(&journal->j_state_lock);
+ wake_up(&journal->j_wait_async_fc);
+}
+
/* Return 1 when transaction with given tid has already committed. */
int jbd2_transaction_committed(journal_t *journal, tid_t tid)
{
@@ -1126,6 +1208,7 @@ static journal_t *journal_init_common(struct block_device *bdev,
init_waitqueue_head(&journal->j_wait_commit);
init_waitqueue_head(&journal->j_wait_updates);
init_waitqueue_head(&journal->j_wait_reserved);
+ init_waitqueue_head(&journal->j_wait_async_fc);
mutex_init(&journal->j_barrier);
mutex_init(&journal->j_checkpoint_mutex);
spin_lock_init(&journal->j_revoke_lock);
diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h
index 7139626992f3..99b0f50ceb50 100644
--- a/include/linux/jbd2.h
+++ b/include/linux/jbd2.h
@@ -861,6 +861,13 @@ struct journal_s
*/
wait_queue_head_t j_wait_reserved;
+ /**
+ * @j_wait_async_fc:
+ *
+ * Wait queue to wait for completion of async fast commits.
+ */
+ wait_queue_head_t j_wait_async_fc;
+
/**
* @j_checkpoint_mutex:
*
@@ -1204,6 +1211,14 @@ struct journal_s
*/
struct lockdep_map j_trans_commit_map;
#endif
+
+ /**
+ * @j_fc_cleanup_callback:
+ *
+ * Clean-up after fast commit or full commit. JBD2 calls this function
+ * after every commit operation.
+ */
+ void (*j_fc_cleanup_callback)(struct journal_s *journal);
};
#define jbd2_might_wait_for_commit(j) \
@@ -1289,6 +1304,8 @@ JBD2_FEATURE_INCOMPAT_FUNCS(fast_commit, FAST_COMMIT)
* data write error in ordered
* mode */
#define JBD2_REC_ERR 0x080 /* The errno in the sb has been recorded */
+#define JBD2_FAST_COMMIT_ONGOING 0x100 /* Fast commit is ongoing */
+#define JBD2_FULL_COMMIT_ONGOING 0x200 /* Full commit is ongoing */
/*
* Function declarations for the journaling transaction and buffer
@@ -1540,6 +1557,10 @@ void __jbd2_log_wait_for_space(journal_t *journal);
extern void __jbd2_journal_drop_transaction(journal_t *, transaction_t *);
extern int jbd2_cleanup_journal_tail(journal_t *);
+/* Fast commit related APIs */
+int jbd2_start_async_fc_nowait(journal_t *journal, tid_t tid);
+int jbd2_start_async_fc_wait(journal_t *journal, tid_t tid);
+void jbd2_stop_async_fc(journal_t *journal, tid_t tid);
void jbd2_init_fast_commit(journal_t *journal, int num_fc_blks);
/*
* is_journal_abort
--
2.24.1.735.g03f4e72817-goog
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PATCH v4 07/20] ext4: add generic diff tracking routines and range tracking
2019-12-24 8:13 [PATCH v4 01/20] ext4: update docs for fast commit feature Harshad Shirwadkar
` (4 preceding siblings ...)
2019-12-24 8:13 ` [PATCH v4 06/20] jbd2: fast commit main commit path changes Harshad Shirwadkar
@ 2019-12-24 8:13 ` Harshad Shirwadkar
2019-12-27 11:15 ` kbuild test robot
2019-12-27 11:16 ` [RFC PATCH] ext4: __ext4_fc_track_range() can be static kbuild test robot
2019-12-24 8:13 ` [PATCH v4 08/20] ext4: add directory entry tracking routines Harshad Shirwadkar
` (13 subsequent siblings)
19 siblings, 2 replies; 31+ messages in thread
From: Harshad Shirwadkar @ 2019-12-24 8:13 UTC (permalink / raw)
To: linux-ext4; +Cc: Harshad Shirwadkar
In fast commits, we need to track changes that have been made to the
file system since last full commit. Add generic diff tracking
infrastructure. We use those helpers to track logical block ranges
that have been affected for inodes. The diff tracking helpers are used
in following patches to track directory entry updates as well.
Signed-off-by: Harshad Shirwadkar <harshadshirwadkar@gmail.com>
---
fs/ext4/ext4.h | 33 ++++++++++
fs/ext4/ext4_jbd2.c | 121 ++++++++++++++++++++++++++++++++++++
fs/ext4/ext4_jbd2.h | 3 +
fs/ext4/inode.c | 18 ++++++
fs/ext4/super.c | 5 ++
include/trace/events/ext4.h | 27 ++++++++
6 files changed, 207 insertions(+)
diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index 16ffe6ed9e74..5c40fa4b593c 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -979,6 +979,32 @@ struct ext4_inode_info {
struct list_head i_orphan; /* unlinked but open inodes */
+ struct list_head i_fc_list; /*
+ * inodes that need fast commit
+ * protected by sbi->s_fc_lock.
+ */
+ /*
+ * TID of when this struct was last updated. If fc_tid !=
+ * running transaction tid, then none of the other fields in this
+ * struct are valid. Don't directly modify fields in this struct.
+ * Use wrappers provided in ext4_jbd2.c.
+ */
+ tid_t i_fc_tid;
+
+ /*
+ * Start of logical block range that needs to be committed in
+ * this fast commit.
+ */
+ ext4_lblk_t i_fc_lblk_start;
+
+ /*
+ * End of logical block range that needs to be committed in this fast
+ * commit
+ */
+ ext4_lblk_t i_fc_lblk_end;
+
+ rwlock_t i_fc_lock;
+
/*
* i_disksize keeps track of what the inode size is ON DISK, not
* in memory. During truncate, i_size is set to the new size by
@@ -1100,6 +1126,7 @@ struct ext4_inode_info {
#define EXT4_VALID_FS 0x0001 /* Unmounted cleanly */
#define EXT4_ERROR_FS 0x0002 /* Errors detected */
#define EXT4_ORPHAN_FS 0x0004 /* Orphans being recovered */
+#define EXT4_FC_REPLAY 0x0008 /* Fast commit replay ongoing */
/*
* Misc. filesystem flags
@@ -1558,6 +1585,12 @@ struct ext4_sb_info {
/* Barrier between changing inodes' journal flags and writepages ops. */
struct percpu_rw_semaphore s_journal_flag_rwsem;
struct dax_device *s_daxdev;
+
+ /* Ext4 fast commit stuff */
+ struct list_head s_fc_q; /* Inodes staged for fast commit
+ * that have data changes in them.
+ */
+ spinlock_t s_fc_lock;
};
static inline struct ext4_sb_info *EXT4_SB(struct super_block *sb)
diff --git a/fs/ext4/ext4_jbd2.c b/fs/ext4/ext4_jbd2.c
index c2ae21f5049b..0907b1b91301 100644
--- a/fs/ext4/ext4_jbd2.c
+++ b/fs/ext4/ext4_jbd2.c
@@ -330,6 +330,127 @@ int __ext4_handle_dirty_super(const char *where, unsigned int line,
mark_buffer_dirty(bh);
return err;
}
+
+static inline
+void ext4_reset_inode_fc_info(struct inode *inode)
+{
+ struct ext4_inode_info *ei = EXT4_I(inode);
+
+ ei->i_fc_tid = 0;
+ ei->i_fc_lblk_start = 0;
+ ei->i_fc_lblk_end = 0;
+}
+
+void ext4_init_inode_fc_info(struct inode *inode)
+{
+ struct ext4_inode_info *ei = EXT4_I(inode);
+
+ ext4_reset_inode_fc_info(inode);
+ INIT_LIST_HEAD(&ei->i_fc_list);
+}
+
+static void ext4_fc_enqueue_inode(struct inode *inode)
+{
+ struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
+
+ if (!ext4_should_fast_commit(inode->i_sb) ||
+ (EXT4_SB(inode->i_sb)->s_mount_state & EXT4_FC_REPLAY))
+ return;
+
+ spin_lock(&sbi->s_fc_lock);
+ if (list_empty(&EXT4_I(inode)->i_fc_list))
+ list_add_tail(&EXT4_I(inode)->i_fc_list, &sbi->s_fc_q);
+ spin_unlock(&sbi->s_fc_lock);
+}
+
+static inline tid_t get_running_txn_tid(struct super_block *sb)
+{
+ if (EXT4_SB(sb)->s_journal)
+ return EXT4_SB(sb)->s_journal->j_commit_sequence + 1;
+ return 0;
+}
+
+/*
+ * Generic fast commit tracking function. If this is the first
+ * time this we are called after a full commit, we initialize
+ * fast commit fields and then call __fc_track_fn() with
+ * update = 0. If we have already been called after a full commit,
+ * we pass update = 1. Based on that, the track function can
+ * determine if it needs to track a field for the first time
+ * or if it needs to just update the previously tracked value.
+ */
+static int __ext4_fc_track_template(
+ struct inode *inode,
+ int (*__fc_track_fn)(struct inode *, void *, bool),
+ void *args)
+{
+ tid_t running_txn_tid = get_running_txn_tid(inode->i_sb);
+ bool update = false;
+ struct ext4_inode_info *ei = EXT4_I(inode);
+ int ret;
+
+ if (!ext4_should_fast_commit(inode->i_sb) ||
+ (EXT4_SB(inode->i_sb)->s_mount_state & EXT4_FC_REPLAY))
+ return -EOPNOTSUPP;
+
+ write_lock(&ei->i_fc_lock);
+ if (running_txn_tid == ei->i_fc_tid) {
+ update = true;
+ } else {
+ ext4_reset_inode_fc_info(inode);
+ ei->i_fc_tid = running_txn_tid;
+ }
+ ret = __fc_track_fn(inode, args, update);
+ write_unlock(&ei->i_fc_lock);
+
+ ext4_fc_enqueue_inode(inode);
+
+ return ret;
+}
+struct __ext4_fc_track_range_args {
+ ext4_lblk_t start, end;
+};
+
+#define MIN(__a, __b) ((__a) < (__b) ? (__a) : (__b))
+#define MAX(__a, __b) ((__a) > (__b) ? (__a) : (__b))
+
+int __ext4_fc_track_range(struct inode *inode, void *arg, bool update)
+{
+ struct ext4_inode_info *ei = EXT4_I(inode);
+ struct __ext4_fc_track_range_args *__arg =
+ (struct __ext4_fc_track_range_args *)arg;
+
+ if (inode->i_ino < EXT4_FIRST_INO(inode->i_sb)) {
+ ext4_debug("Special inode %ld being modified\n", inode->i_ino);
+ return -ECANCELED;
+ }
+
+ if (update) {
+ ei->i_fc_lblk_start = MIN(ei->i_fc_lblk_start, __arg->start);
+ ei->i_fc_lblk_end = MAX(ei->i_fc_lblk_end, __arg->end);
+ } else {
+ ei->i_fc_lblk_start = __arg->start;
+ ei->i_fc_lblk_end = __arg->end;
+ }
+
+ return 0;
+}
+
+void ext4_fc_track_range(struct inode *inode, ext4_lblk_t start,
+ ext4_lblk_t end)
+{
+ struct __ext4_fc_track_range_args args;
+ int ret;
+
+ args.start = start;
+ args.end = end;
+
+ ret = __ext4_fc_track_template(inode,
+ __ext4_fc_track_range, &args);
+
+ trace_ext4_fc_track_range(inode, start, end, ret);
+}
+
void ext4_init_fast_commit(struct super_block *sb, journal_t *journal)
{
if (!ext4_should_fast_commit(sb))
diff --git a/fs/ext4/ext4_jbd2.h b/fs/ext4/ext4_jbd2.h
index e5bd95a088e8..d7eca4b9a935 100644
--- a/fs/ext4/ext4_jbd2.h
+++ b/fs/ext4/ext4_jbd2.h
@@ -472,4 +472,7 @@ static inline int ext4_should_dioread_nolock(struct inode *inode)
#define EXT4_NUM_FC_BLKS 128
void ext4_init_fast_commit(struct super_block *sb, journal_t *journal);
+void ext4_init_inode_fc_info(struct inode *inode);
+void ext4_fc_track_range(struct inode *inode, ext4_lblk_t start,
+ ext4_lblk_t end);
#endif /* _EXT4_JBD2_H */
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 516faa280ced..07c8da778368 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -744,6 +744,8 @@ int ext4_map_blocks(handle_t *handle, struct inode *inode,
if (ret)
return ret;
}
+ ext4_fc_track_range(inode, map->m_lblk,
+ map->m_lblk + map->m_len - 1);
}
return retval;
}
@@ -4368,6 +4370,7 @@ int ext4_punch_hole(struct inode *inode, loff_t offset, loff_t length)
up_write(&EXT4_I(inode)->i_data_sem);
}
+ ext4_fc_track_range(inode, first_block, stop_block);
if (IS_SYNC(inode))
ext4_handle_sync(handle);
@@ -4965,6 +4968,7 @@ struct inode *__ext4_iget(struct super_block *sb, unsigned long ino,
for (block = 0; block < EXT4_N_BLOCKS; block++)
ei->i_data[block] = raw_inode->i_block[block];
INIT_LIST_HEAD(&ei->i_orphan);
+ ext4_init_inode_fc_info(&ei->vfs_inode);
/*
* Set transaction id's of transactions that have to be committed
@@ -5628,6 +5632,20 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr)
inode->i_mtime = current_time(inode);
inode->i_ctime = inode->i_mtime;
}
+
+ if (shrink)
+ ext4_fc_track_range(
+ inode, attr->ia_size >>
+ inode->i_sb->s_blocksize_bits,
+ oldsize >>
+ inode->i_sb->s_blocksize_bits);
+ else
+ ext4_fc_track_range(
+ inode, oldsize >>
+ inode->i_sb->s_blocksize_bits,
+ attr->ia_size >>
+ inode->i_sb->s_blocksize_bits);
+
down_write(&EXT4_I(inode)->i_data_sem);
EXT4_I(inode)->i_disksize = attr->ia_size;
rc = ext4_mark_inode_dirty(handle, inode);
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index 28675cd78813..2f922ef522a3 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -1100,6 +1100,8 @@ static struct inode *ext4_alloc_inode(struct super_block *sb)
ei->i_datasync_tid = 0;
atomic_set(&ei->i_unwritten, 0);
INIT_WORK(&ei->i_rsv_conversion_work, ext4_end_io_rsv_work);
+ ext4_init_inode_fc_info(&ei->vfs_inode);
+ rwlock_init(&ei->i_fc_lock);
return &ei->vfs_inode;
}
@@ -1142,6 +1144,7 @@ static void init_once(void *foo)
init_rwsem(&ei->i_data_sem);
init_rwsem(&ei->i_mmap_sem);
inode_init_once(&ei->vfs_inode);
+ ext4_init_inode_fc_info(&ei->vfs_inode);
}
static int __init init_inodecache(void)
@@ -4330,6 +4333,8 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
INIT_LIST_HEAD(&sbi->s_orphan); /* unlinked but open files */
mutex_init(&sbi->s_orphan_lock);
+ INIT_LIST_HEAD(&sbi->s_fc_q);
+ spin_lock_init(&sbi->s_fc_lock);
sb->s_root = NULL;
needs_recovery = (es->s_last_orphan != 0 ||
diff --git a/include/trace/events/ext4.h b/include/trace/events/ext4.h
index d68e9e536814..0f6d43dfd4b2 100644
--- a/include/trace/events/ext4.h
+++ b/include/trace/events/ext4.h
@@ -2703,6 +2703,33 @@ TRACE_EVENT(ext4_error,
__entry->function, __entry->line)
);
+TRACE_EVENT(ext4_fc_track_range,
+ TP_PROTO(struct inode *inode, long start, long end, int ret),
+
+ TP_ARGS(inode, start, end, ret),
+
+ TP_STRUCT__entry(
+ __field(dev_t, dev)
+ __field(int, ino)
+ __field(long, start)
+ __field(long, end)
+ __field(int, error)
+ ),
+
+ TP_fast_assign(
+ __entry->dev = inode->i_sb->s_dev;
+ __entry->ino = inode->i_ino;
+ __entry->start = start;
+ __entry->end = end;
+ __entry->error = ret;
+ ),
+
+ TP_printk("dev %d:%d, inode %d, error %d, start %ld, end %ld",
+ MAJOR(__entry->dev), MINOR(__entry->dev),
+ __entry->ino, __entry->error, __entry->start,
+ __entry->end)
+ );
+
#endif /* _TRACE_EXT4_H */
/* This part must be outside protection */
--
2.24.1.735.g03f4e72817-goog
^ permalink raw reply related [flat|nested] 31+ messages in thread* Re: [PATCH v4 07/20] ext4: add generic diff tracking routines and range tracking
2019-12-24 8:13 ` [PATCH v4 07/20] ext4: add generic diff tracking routines and range tracking Harshad Shirwadkar
@ 2019-12-27 11:15 ` kbuild test robot
2019-12-27 11:16 ` [RFC PATCH] ext4: __ext4_fc_track_range() can be static kbuild test robot
1 sibling, 0 replies; 31+ messages in thread
From: kbuild test robot @ 2019-12-27 11:15 UTC (permalink / raw)
To: Harshad Shirwadkar; +Cc: kbuild-all, linux-ext4, Harshad Shirwadkar
Hi Harshad,
Thank you for the patch! Perhaps something to improve:
[auto build test WARNING on tip/perf/core]
[cannot apply to ext4/dev linus/master v5.5-rc3 next-20191220]
[if your patch is applied to the wrong git tree, please drop us a note to help
improve the system. BTW, we also suggest to use '--base' option to specify the
base tree in git format-patch, please see https://stackoverflow.com/a/37406982]
url: https://github.com/0day-ci/linux/commits/Harshad-Shirwadkar/ext4-update-docs-for-fast-commit-feature/20191225-200339
base: https://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git ceb9e77324fa661b1001a0ae66f061b5fcb4e4e6
reproduce:
# apt-get install sparse
# sparse version: v0.6.1-129-g341daf20-dirty
make ARCH=x86_64 allmodconfig
make C=1 CF='-fdiagnostic-prefix -D__CHECK_ENDIAN__'
If you fix the issue, kindly add following tag
Reported-by: kbuild test robot <lkp@intel.com>
sparse warnings: (new ones prefixed by >>)
>> fs/ext4/ext4_jbd2.c:417:5: sparse: sparse: symbol '__ext4_fc_track_range' was not declared. Should it be static?
Please review and possibly fold the followup patch.
---
0-DAY kernel test infrastructure Open Source Technology Center
https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org Intel Corporation
^ permalink raw reply [flat|nested] 31+ messages in thread* [RFC PATCH] ext4: __ext4_fc_track_range() can be static
2019-12-24 8:13 ` [PATCH v4 07/20] ext4: add generic diff tracking routines and range tracking Harshad Shirwadkar
2019-12-27 11:15 ` kbuild test robot
@ 2019-12-27 11:16 ` kbuild test robot
1 sibling, 0 replies; 31+ messages in thread
From: kbuild test robot @ 2019-12-27 11:16 UTC (permalink / raw)
To: Harshad Shirwadkar; +Cc: kbuild-all, linux-ext4, Harshad Shirwadkar
Fixes: 9b03f2f7eee6 ("ext4: add generic diff tracking routines and range tracking")
Signed-off-by: kbuild test robot <lkp@intel.com>
---
ext4_jbd2.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/fs/ext4/ext4_jbd2.c b/fs/ext4/ext4_jbd2.c
index 0907b1b913013..b65625bbe7146 100644
--- a/fs/ext4/ext4_jbd2.c
+++ b/fs/ext4/ext4_jbd2.c
@@ -414,7 +414,7 @@ struct __ext4_fc_track_range_args {
#define MIN(__a, __b) ((__a) < (__b) ? (__a) : (__b))
#define MAX(__a, __b) ((__a) > (__b) ? (__a) : (__b))
-int __ext4_fc_track_range(struct inode *inode, void *arg, bool update)
+static int __ext4_fc_track_range(struct inode *inode, void *arg, bool update)
{
struct ext4_inode_info *ei = EXT4_I(inode);
struct __ext4_fc_track_range_args *__arg =
^ permalink raw reply related [flat|nested] 31+ messages in thread
* [PATCH v4 08/20] ext4: add directory entry tracking routines
2019-12-24 8:13 [PATCH v4 01/20] ext4: update docs for fast commit feature Harshad Shirwadkar
` (5 preceding siblings ...)
2019-12-24 8:13 ` [PATCH v4 07/20] ext4: add generic diff tracking routines and range tracking Harshad Shirwadkar
@ 2019-12-24 8:13 ` Harshad Shirwadkar
2019-12-24 8:13 ` [PATCH v4 09/20] ext4: add inode tracking and ineligible marking routines Harshad Shirwadkar
` (12 subsequent siblings)
19 siblings, 0 replies; 31+ messages in thread
From: Harshad Shirwadkar @ 2019-12-24 8:13 UTC (permalink / raw)
To: linux-ext4; +Cc: Harshad Shirwadkar
Adds directory entry change tracking routines for fast commits. Use an
in-memory list of directory updates to track directory entry updates.
Signed-off-by: Harshad Shirwadkar <harshadshirwadkar@gmail.com>
---
fs/ext4/ext4.h | 26 +++++++++
fs/ext4/ext4_jbd2.c | 102 ++++++++++++++++++++++++++++++++++++
fs/ext4/ext4_jbd2.h | 4 ++
fs/ext4/super.c | 7 +++
include/trace/events/ext4.h | 28 ++++++++++
5 files changed, 167 insertions(+)
diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index 5c40fa4b593c..6b08c4e2a08c 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -946,6 +946,26 @@ enum {
};
+/* Fast commit tags */
+#define EXT4_FC_TAG_ADD_RANGE 0x1
+#define EXT4_FC_TAG_DEL_RANGE 0x2
+#define EXT4_FC_TAG_CREAT_DENTRY 0x3
+#define EXT4_FC_TAG_ADD_DENTRY 0x4
+#define EXT4_FC_TAG_DEL_DENTRY 0x5
+
+/*
+ * In memory list of dentry updates that are performed on the file
+ * system used by fast commit code.
+ */
+struct ext4_fc_dentry_update {
+ int fcd_op; /* Type of update create / add / del */
+ int fcd_parent; /* Parent inode number */
+ int fcd_ino; /* Inode number */
+ struct qstr fcd_name; /* Dirent name qstr */
+ unsigned char fcd_iname[DNAME_INLINE_LEN]; /* Dirent name string */
+ struct list_head fcd_list;
+};
+
/*
* fourth extended file system inode data in memory
*/
@@ -1005,6 +1025,11 @@ struct ext4_inode_info {
rwlock_t i_fc_lock;
+ /*
+ * Last mdata / dirent update that happened on this inode.
+ */
+ struct ext4_fc_dentry_update *i_fc_mdata_update;
+
/*
* i_disksize keeps track of what the inode size is ON DISK, not
* in memory. During truncate, i_size is set to the new size by
@@ -1590,6 +1615,7 @@ struct ext4_sb_info {
struct list_head s_fc_q; /* Inodes staged for fast commit
* that have data changes in them.
*/
+ struct list_head s_fc_dentry_q;
spinlock_t s_fc_lock;
};
diff --git a/fs/ext4/ext4_jbd2.c b/fs/ext4/ext4_jbd2.c
index 0907b1b91301..f3daa941cba5 100644
--- a/fs/ext4/ext4_jbd2.c
+++ b/fs/ext4/ext4_jbd2.c
@@ -331,6 +331,8 @@ int __ext4_handle_dirty_super(const char *where, unsigned int line,
return err;
}
+static struct kmem_cache *ext4_fc_dentry_cachep;
+
static inline
void ext4_reset_inode_fc_info(struct inode *inode)
{
@@ -339,6 +341,7 @@ void ext4_reset_inode_fc_info(struct inode *inode)
ei->i_fc_tid = 0;
ei->i_fc_lblk_start = 0;
ei->i_fc_lblk_end = 0;
+ ei->i_fc_mdata_update = NULL;
}
void ext4_init_inode_fc_info(struct inode *inode)
@@ -407,6 +410,94 @@ static int __ext4_fc_track_template(
return ret;
}
+
+struct __ext4_dentry_update_args {
+ struct dentry *dentry;
+ int op;
+};
+
+static int __ext4_dentry_update(struct inode *inode, void *arg, bool update)
+{
+ struct ext4_fc_dentry_update *node;
+ struct ext4_inode_info *ei = EXT4_I(inode);
+ struct __ext4_dentry_update_args *dentry_update =
+ (struct __ext4_dentry_update_args *)arg;
+ struct dentry *dentry = dentry_update->dentry;
+
+ write_unlock(&ei->i_fc_lock);
+ node = kmem_cache_alloc(ext4_fc_dentry_cachep, GFP_NOFS);
+ if (!node) {
+ write_lock(&ei->i_fc_lock);
+ return -ENOMEM;
+ }
+
+ node->fcd_op = dentry_update->op;
+ node->fcd_parent = dentry->d_parent->d_inode->i_ino;
+ node->fcd_ino = inode->i_ino;
+ if (dentry->d_name.len > DNAME_INLINE_LEN) {
+ node->fcd_name.name = kmalloc(dentry->d_name.len + 1,
+ GFP_KERNEL);
+ if (!node->fcd_iname) {
+ kmem_cache_free(ext4_fc_dentry_cachep, node);
+ return -ENOMEM;
+ }
+ memcpy((u8 *)node->fcd_name.name, dentry->d_name.name,
+ dentry->d_name.len);
+ } else {
+ memcpy(node->fcd_iname, dentry->d_name.name,
+ dentry->d_name.len);
+ node->fcd_name.name = node->fcd_iname;
+ }
+ node->fcd_name.len = dentry->d_name.len;
+
+ spin_lock(&EXT4_SB(inode->i_sb)->s_fc_lock);
+ list_add_tail(&node->fcd_list, &EXT4_SB(inode->i_sb)->s_fc_dentry_q);
+ spin_unlock(&EXT4_SB(inode->i_sb)->s_fc_lock);
+ write_lock(&ei->i_fc_lock);
+ EXT4_I(inode)->i_fc_mdata_update = node;
+
+ return 0;
+}
+
+void ext4_fc_track_unlink(struct inode *inode, struct dentry *dentry)
+{
+ struct __ext4_dentry_update_args args;
+ int ret;
+
+ args.dentry = dentry;
+ args.op = EXT4_FC_TAG_DEL_DENTRY;
+
+ ret = __ext4_fc_track_template(inode, __ext4_dentry_update,
+ (void *)&args);
+ trace_ext4_fc_track_unlink(inode, dentry, ret);
+}
+
+void ext4_fc_track_link(struct inode *inode, struct dentry *dentry)
+{
+ struct __ext4_dentry_update_args args;
+ int ret;
+
+ args.dentry = dentry;
+ args.op = EXT4_FC_TAG_ADD_DENTRY;
+
+ ret = __ext4_fc_track_template(inode, __ext4_dentry_update,
+ (void *)&args);
+ trace_ext4_fc_track_link(inode, dentry, ret);
+}
+
+void ext4_fc_track_create(struct inode *inode, struct dentry *dentry)
+{
+ struct __ext4_dentry_update_args args;
+ int ret;
+
+ args.dentry = dentry;
+ args.op = EXT4_FC_TAG_CREAT_DENTRY;
+
+ ret = __ext4_fc_track_template(inode, __ext4_dentry_update,
+ (void *)&args);
+ trace_ext4_fc_track_create(inode, dentry, ret);
+}
+
struct __ext4_fc_track_range_args {
ext4_lblk_t start, end;
};
@@ -457,3 +548,14 @@ void ext4_init_fast_commit(struct super_block *sb, journal_t *journal)
return;
jbd2_init_fast_commit(journal, EXT4_NUM_FC_BLKS);
}
+
+int __init ext4_init_fc_dentry_cache(void)
+{
+ ext4_fc_dentry_cachep = KMEM_CACHE(ext4_fc_dentry_update,
+ SLAB_RECLAIM_ACCOUNT);
+
+ if (ext4_fc_dentry_cachep == NULL)
+ return -ENOMEM;
+
+ return 0;
+}
diff --git a/fs/ext4/ext4_jbd2.h b/fs/ext4/ext4_jbd2.h
index d7eca4b9a935..1539e672aec6 100644
--- a/fs/ext4/ext4_jbd2.h
+++ b/fs/ext4/ext4_jbd2.h
@@ -475,4 +475,8 @@ void ext4_init_fast_commit(struct super_block *sb, journal_t *journal);
void ext4_init_inode_fc_info(struct inode *inode);
void ext4_fc_track_range(struct inode *inode, ext4_lblk_t start,
ext4_lblk_t end);
+void ext4_fc_track_unlink(struct inode *inode, struct dentry *dentry);
+void ext4_fc_track_link(struct inode *inode, struct dentry *dentry);
+void ext4_fc_track_create(struct inode *inode, struct dentry *dentry);
+int __init ext4_init_fc_dentry_cache(void);
#endif /* _EXT4_JBD2_H */
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index 2f922ef522a3..71ecca296fe4 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -4334,6 +4334,7 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
mutex_init(&sbi->s_orphan_lock);
INIT_LIST_HEAD(&sbi->s_fc_q);
+ INIT_LIST_HEAD(&sbi->s_fc_dentry_q);
spin_lock_init(&sbi->s_fc_lock);
sb->s_root = NULL;
@@ -6176,6 +6177,11 @@ static int __init ext4_init_fs(void)
err = init_inodecache();
if (err)
goto out1;
+
+ err = ext4_init_fc_dentry_cache();
+ if (err)
+ goto out05;
+
register_as_ext3();
register_as_ext2();
err = register_filesystem(&ext4_fs_type);
@@ -6186,6 +6192,7 @@ static int __init ext4_init_fs(void)
out:
unregister_as_ext2();
unregister_as_ext3();
+out05:
destroy_inodecache();
out1:
ext4_exit_mballoc();
diff --git a/include/trace/events/ext4.h b/include/trace/events/ext4.h
index 0f6d43dfd4b2..02f9fd718d37 100644
--- a/include/trace/events/ext4.h
+++ b/include/trace/events/ext4.h
@@ -2703,6 +2703,34 @@ TRACE_EVENT(ext4_error,
__entry->function, __entry->line)
);
+#define DEFINE_TRACE_DENTRY_EVENT(__type) \
+ TRACE_EVENT(ext4_fc_track_##__type, \
+ TP_PROTO(struct inode *inode, struct dentry *dentry, int ret), \
+ \
+ TP_ARGS(inode, dentry, ret), \
+ \
+ TP_STRUCT__entry( \
+ __field(dev_t, dev) \
+ __field(int, ino) \
+ __field(int, error) \
+ ), \
+ \
+ TP_fast_assign( \
+ __entry->dev = inode->i_sb->s_dev; \
+ __entry->ino = inode->i_ino; \
+ __entry->error = ret; \
+ ), \
+ \
+ TP_printk("dev %d:%d, inode %d, error %d, fc_%s", \
+ MAJOR(__entry->dev), MINOR(__entry->dev), \
+ __entry->ino, __entry->error, \
+ #__type) \
+ )
+
+DEFINE_TRACE_DENTRY_EVENT(create);
+DEFINE_TRACE_DENTRY_EVENT(link);
+DEFINE_TRACE_DENTRY_EVENT(unlink);
+
TRACE_EVENT(ext4_fc_track_range,
TP_PROTO(struct inode *inode, long start, long end, int ret),
--
2.24.1.735.g03f4e72817-goog
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PATCH v4 09/20] ext4: add inode tracking and ineligible marking routines
2019-12-24 8:13 [PATCH v4 01/20] ext4: update docs for fast commit feature Harshad Shirwadkar
` (6 preceding siblings ...)
2019-12-24 8:13 ` [PATCH v4 08/20] ext4: add directory entry tracking routines Harshad Shirwadkar
@ 2019-12-24 8:13 ` Harshad Shirwadkar
2019-12-24 8:13 ` [PATCH v4 10/20] ext4: break ext4_unlink() and ext4_link() Harshad Shirwadkar
` (11 subsequent siblings)
19 siblings, 0 replies; 31+ messages in thread
From: Harshad Shirwadkar @ 2019-12-24 8:13 UTC (permalink / raw)
To: linux-ext4; +Cc: Harshad Shirwadkar
Under certain situations, such as zeroing a range, there are only data
updates and no metadata updates. We need to track such inodes for fast
commits. Also, under some situations, we need to fall back to full
commits because remembering the delta is either not yet supported or
fast commits won't be "fast" enough. In such cases, we need to mark
inodes as ineligible for fast commits. Add routines that allow
tracking just the inodes and marking inodes as well as entire file
system as fast commit ineligible.
Signed-off-by: Harshad Shirwadkar <harshadshirwadkar@gmail.com>
---
fs/ext4/ext4.h | 2 ++
fs/ext4/ext4_jbd2.c | 57 +++++++++++++++++++++++++++++++++++++
fs/ext4/ext4_jbd2.h | 4 +++
fs/ext4/super.c | 1 +
include/trace/events/ext4.h | 22 ++++++++++++++
5 files changed, 86 insertions(+)
diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index 6b08c4e2a08c..b4c32f02071f 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -1152,6 +1152,7 @@ struct ext4_inode_info {
#define EXT4_ERROR_FS 0x0002 /* Errors detected */
#define EXT4_ORPHAN_FS 0x0004 /* Orphans being recovered */
#define EXT4_FC_REPLAY 0x0008 /* Fast commit replay ongoing */
+#define EXT4_FC_INELIGIBLE 0x0010 /* Fast commit ineligible */
/*
* Misc. filesystem flags
@@ -1651,6 +1652,7 @@ enum {
EXT4_STATE_EXT_PRECACHED, /* extents have been precached */
EXT4_STATE_LUSTRE_EA_INODE, /* Lustre-style ea_inode */
EXT4_STATE_VERITY_IN_PROGRESS, /* building fs-verity Merkle tree */
+ EXT4_STATE_FC_ELIGIBLE, /* File is Fast commit eligible */
};
#define EXT4_INODE_BIT_FNS(name, field, offset) \
diff --git a/fs/ext4/ext4_jbd2.c b/fs/ext4/ext4_jbd2.c
index f3daa941cba5..7c27f9284064 100644
--- a/fs/ext4/ext4_jbd2.c
+++ b/fs/ext4/ext4_jbd2.c
@@ -342,6 +342,7 @@ void ext4_reset_inode_fc_info(struct inode *inode)
ei->i_fc_lblk_start = 0;
ei->i_fc_lblk_end = 0;
ei->i_fc_mdata_update = NULL;
+ ext4_clear_inode_state(inode, EXT4_STATE_FC_ELIGIBLE);
}
void ext4_init_inode_fc_info(struct inode *inode)
@@ -373,6 +374,36 @@ static inline tid_t get_running_txn_tid(struct super_block *sb)
return 0;
}
+static bool ext4_is_inode_fc_ineligible(struct inode *inode)
+{
+ if (get_running_txn_tid(inode->i_sb) == EXT4_I(inode)->i_fc_tid)
+ return !ext4_test_inode_state(inode, EXT4_STATE_FC_ELIGIBLE);
+ return false;
+}
+
+void ext4_fc_mark_ineligible(struct inode *inode, int reason)
+{
+ struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
+ struct ext4_inode_info *ei = EXT4_I(inode);
+
+ if (!ext4_should_fast_commit(inode->i_sb) ||
+ (EXT4_SB(inode->i_sb)->s_mount_state & EXT4_FC_REPLAY))
+ return;
+
+ if (sbi->s_journal)
+ ei->i_fc_tid = get_running_txn_tid(inode->i_sb);
+ ext4_clear_inode_state(inode, EXT4_STATE_FC_ELIGIBLE);
+
+ ext4_fc_enqueue_inode(inode);
+}
+
+void ext4_fc_disable(struct super_block *sb, int reason)
+{
+ struct ext4_sb_info *sbi = EXT4_SB(sb);
+
+ sbi->s_mount_state |= EXT4_FC_INELIGIBLE;
+}
+
/*
* Generic fast commit tracking function. If this is the first
* time this we are called after a full commit, we initialize
@@ -398,10 +429,15 @@ static int __ext4_fc_track_template(
write_lock(&ei->i_fc_lock);
if (running_txn_tid == ei->i_fc_tid) {
+ if (!ext4_test_inode_state(inode, EXT4_STATE_FC_ELIGIBLE)) {
+ write_unlock(&ei->i_fc_lock);
+ return -EINVAL;
+ }
update = true;
} else {
ext4_reset_inode_fc_info(inode);
ei->i_fc_tid = running_txn_tid;
+ ext4_set_inode_state(inode, EXT4_STATE_FC_ELIGIBLE);
}
ret = __fc_track_fn(inode, args, update);
write_unlock(&ei->i_fc_lock);
@@ -498,6 +534,27 @@ void ext4_fc_track_create(struct inode *inode, struct dentry *dentry)
trace_ext4_fc_track_create(inode, dentry, ret);
}
+static int __ext4_fc_add_inode(struct inode *inode, void *arg, bool update)
+{
+ struct ext4_inode_info *ei = EXT4_I(inode);
+
+ if (update)
+ return -EEXIST;
+
+ ei->i_fc_lblk_start = (i_size_read(inode) - 1) >> inode->i_blkbits;
+ ei->i_fc_lblk_end = (i_size_read(inode) - 1) >> inode->i_blkbits;
+
+ return 0;
+}
+
+void ext4_fc_track_inode(struct inode *inode)
+{
+ int ret;
+
+ ret = __ext4_fc_track_template(inode, __ext4_fc_add_inode, NULL);
+ trace_ext4_fc_track_inode(inode, ret);
+}
+
struct __ext4_fc_track_range_args {
ext4_lblk_t start, end;
};
diff --git a/fs/ext4/ext4_jbd2.h b/fs/ext4/ext4_jbd2.h
index 1539e672aec6..60f484377c2e 100644
--- a/fs/ext4/ext4_jbd2.h
+++ b/fs/ext4/ext4_jbd2.h
@@ -479,4 +479,8 @@ void ext4_fc_track_unlink(struct inode *inode, struct dentry *dentry);
void ext4_fc_track_link(struct inode *inode, struct dentry *dentry);
void ext4_fc_track_create(struct inode *inode, struct dentry *dentry);
int __init ext4_init_fc_dentry_cache(void);
+void ext4_fc_track_inode(struct inode *inode);
+void ext4_fc_mark_ineligible(struct inode *inode, int reason);
+void ext4_fc_disable(struct super_block *sb, int reason);
+
#endif /* _EXT4_JBD2_H */
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index 71ecca296fe4..538ee986d7f7 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -4335,6 +4335,7 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
INIT_LIST_HEAD(&sbi->s_fc_q);
INIT_LIST_HEAD(&sbi->s_fc_dentry_q);
+ sbi->s_mount_state &= ~EXT4_FC_INELIGIBLE;
spin_lock_init(&sbi->s_fc_lock);
sb->s_root = NULL;
diff --git a/include/trace/events/ext4.h b/include/trace/events/ext4.h
index 02f9fd718d37..0808b62ac108 100644
--- a/include/trace/events/ext4.h
+++ b/include/trace/events/ext4.h
@@ -2731,6 +2731,28 @@ DEFINE_TRACE_DENTRY_EVENT(create);
DEFINE_TRACE_DENTRY_EVENT(link);
DEFINE_TRACE_DENTRY_EVENT(unlink);
+TRACE_EVENT(ext4_fc_track_inode,
+ TP_PROTO(struct inode *inode, int ret),
+
+ TP_ARGS(inode, ret),
+
+ TP_STRUCT__entry(
+ __field(dev_t, dev)
+ __field(int, ino)
+ __field(int, error)
+ ),
+
+ TP_fast_assign(
+ __entry->dev = inode->i_sb->s_dev;
+ __entry->ino = inode->i_ino;
+ __entry->error = ret;
+ ),
+
+ TP_printk("dev %d:%d, inode %d, error %d",
+ MAJOR(__entry->dev), MINOR(__entry->dev),
+ __entry->ino, __entry->error)
+ );
+
TRACE_EVENT(ext4_fc_track_range,
TP_PROTO(struct inode *inode, long start, long end, int ret),
--
2.24.1.735.g03f4e72817-goog
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PATCH v4 10/20] ext4: break ext4_unlink() and ext4_link()
2019-12-24 8:13 [PATCH v4 01/20] ext4: update docs for fast commit feature Harshad Shirwadkar
` (7 preceding siblings ...)
2019-12-24 8:13 ` [PATCH v4 09/20] ext4: add inode tracking and ineligible marking routines Harshad Shirwadkar
@ 2019-12-24 8:13 ` Harshad Shirwadkar
2019-12-24 8:13 ` [PATCH v4 11/20] ext4: add fast commit track points Harshad Shirwadkar
` (10 subsequent siblings)
19 siblings, 0 replies; 31+ messages in thread
From: Harshad Shirwadkar @ 2019-12-24 8:13 UTC (permalink / raw)
To: linux-ext4; +Cc: Harshad Shirwadkar
Break ext4_link() and ext4_unlink() each into 2 parts in order to make
them usable in recovery path as well.
Signed-off-by: Harshad Shirwadkar <harshadshirwadkar@gmail.com>
---
fs/ext4/ext4.h | 4 ++
fs/ext4/namei.c | 138 +++++++++++++++++++++++++++++-------------------
2 files changed, 87 insertions(+), 55 deletions(-)
diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index b4c32f02071f..cc3a1489eae4 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -3262,6 +3262,10 @@ extern int ext4_handle_dirty_dirblock(handle_t *handle, struct inode *inode,
extern int ext4_ci_compare(const struct inode *parent,
const struct qstr *fname,
const struct qstr *entry, bool quick);
+extern int __ext4_unlink(struct inode *dir, const struct qstr *d_name,
+ struct inode *inode);
+extern int __ext4_link(struct inode *dir, struct inode *inode,
+ struct dentry *dentry);
#define S_SHIFT 12
static const unsigned char ext4_type_by_mode[(S_IFMT >> S_SHIFT) + 1] = {
diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c
index a427d2031a8d..a405564ae02f 100644
--- a/fs/ext4/namei.c
+++ b/fs/ext4/namei.c
@@ -3137,39 +3137,36 @@ static int ext4_rmdir(struct inode *dir, struct dentry *dentry)
return retval;
}
-static int ext4_unlink(struct inode *dir, struct dentry *dentry)
+int __ext4_unlink(struct inode *dir, const struct qstr *d_name,
+ struct inode *inode)
{
- int retval;
- struct inode *inode;
struct buffer_head *bh;
struct ext4_dir_entry_2 *de;
handle_t *handle = NULL;
+ int retval = -ENOENT;
+ int skip_remove_dentry = 0;
- if (unlikely(ext4_forced_shutdown(EXT4_SB(dir->i_sb))))
- return -EIO;
-
- trace_ext4_unlink_enter(dir, dentry);
- /* Initialize quotas before so that eventual writes go
- * in separate transaction */
- retval = dquot_initialize(dir);
- if (retval)
- return retval;
- retval = dquot_initialize(d_inode(dentry));
- if (retval)
- return retval;
-
- retval = -ENOENT;
- bh = ext4_find_entry(dir, &dentry->d_name, &de, NULL);
+ bh = ext4_find_entry(dir, d_name, &de, NULL);
if (IS_ERR(bh))
return PTR_ERR(bh);
- if (!bh)
- goto end_unlink;
- inode = d_inode(dentry);
+ if (!bh) {
+ retval = -ENOENT;
+ goto end_unlink;
+ }
retval = -EFSCORRUPTED;
- if (le32_to_cpu(de->inode) != inode->i_ino)
- goto end_unlink;
+ if (le32_to_cpu(de->inode) != inode->i_ino) {
+ /*
+ * It's okay if we find dont find dentry which matches
+ * the inode. That's because it might have gotten
+ * renamed to a different inode number
+ */
+ if (EXT4_SB(inode->i_sb)->s_mount_state & EXT4_FC_REPLAY)
+ skip_remove_dentry = 1;
+ else
+ goto end_unlink;
+ }
handle = ext4_journal_start(dir, EXT4_HT_DIR,
EXT4_DATA_TRANS_BLOCKS(dir->i_sb));
@@ -3184,21 +3181,52 @@ static int ext4_unlink(struct inode *dir, struct dentry *dentry)
if (inode->i_nlink == 0) {
ext4_warning_inode(inode, "Deleting file '%.*s' with no links",
- dentry->d_name.len, dentry->d_name.name);
+ d_name->len, d_name->name);
set_nlink(inode, 1);
}
- retval = ext4_delete_entry(handle, dir, de, bh);
- if (retval)
- goto end_unlink;
- dir->i_ctime = dir->i_mtime = current_time(dir);
- ext4_update_dx_flag(dir);
- ext4_mark_inode_dirty(handle, dir);
+ if (!skip_remove_dentry) {
+ retval = ext4_delete_entry(handle, dir, de, bh);
+ if (retval)
+ goto end_unlink;
+ dir->i_ctime = dir->i_mtime = current_time(dir);
+ ext4_update_dx_flag(dir);
+ ext4_mark_inode_dirty(handle, dir);
+ } else {
+ retval = 0;
+ }
drop_nlink(inode);
if (!inode->i_nlink)
ext4_orphan_add(handle, inode);
inode->i_ctime = current_time(inode);
ext4_mark_inode_dirty(handle, inode);
+end_unlink:
+ brelse(bh);
+ if (handle)
+ ext4_journal_stop(handle);
+ return retval;
+}
+
+static int ext4_unlink(struct inode *dir, struct dentry *dentry)
+{
+ int retval;
+
+ if (unlikely(ext4_forced_shutdown(EXT4_SB(dir->i_sb))))
+ return -EIO;
+
+ trace_ext4_unlink_enter(dir, dentry);
+ /*
+ * Initialize quotas before so that eventual writes go
+ * in separate transaction
+ */
+ retval = dquot_initialize(dir);
+ if (retval)
+ return retval;
+ retval = dquot_initialize(d_inode(dentry));
+ if (retval)
+ return retval;
+
+ retval = __ext4_unlink(dir, &dentry->d_name, d_inode(dentry));
#ifdef CONFIG_UNICODE
/* VFS negative dentries are incompatible with Encoding and
* Case-insensitiveness. Eventually we'll want avoid
@@ -3209,11 +3237,6 @@ static int ext4_unlink(struct inode *dir, struct dentry *dentry)
if (IS_CASEFOLDED(dir))
d_invalidate(dentry);
#endif
-
-end_unlink:
- brelse(bh);
- if (handle)
- ext4_journal_stop(handle);
trace_ext4_unlink_exit(dentry, retval);
return retval;
}
@@ -3348,29 +3371,10 @@ static int ext4_symlink(struct inode *dir,
return err;
}
-static int ext4_link(struct dentry *old_dentry,
- struct inode *dir, struct dentry *dentry)
+int __ext4_link(struct inode *dir, struct inode *inode, struct dentry *dentry)
{
handle_t *handle;
- struct inode *inode = d_inode(old_dentry);
int err, retries = 0;
-
- if (inode->i_nlink >= EXT4_LINK_MAX)
- return -EMLINK;
-
- err = fscrypt_prepare_link(old_dentry, dir, dentry);
- if (err)
- return err;
-
- if ((ext4_test_inode_flag(dir, EXT4_INODE_PROJINHERIT)) &&
- (!projid_eq(EXT4_I(dir)->i_projid,
- EXT4_I(old_dentry->d_inode)->i_projid)))
- return -EXDEV;
-
- err = dquot_initialize(dir);
- if (err)
- return err;
-
retry:
handle = ext4_journal_start(dir, EXT4_HT_DIR,
(EXT4_DATA_TRANS_BLOCKS(dir->i_sb) +
@@ -3404,6 +3408,30 @@ static int ext4_link(struct dentry *old_dentry,
return err;
}
+static int ext4_link(struct dentry *old_dentry,
+ struct inode *dir, struct dentry *dentry)
+{
+ struct inode *inode = d_inode(old_dentry);
+ int err;
+
+ if (inode->i_nlink >= EXT4_LINK_MAX)
+ return -EMLINK;
+
+ err = fscrypt_prepare_link(old_dentry, dir, dentry);
+ if (err)
+ return err;
+
+ if ((ext4_test_inode_flag(dir, EXT4_INODE_PROJINHERIT)) &&
+ (!projid_eq(EXT4_I(dir)->i_projid,
+ EXT4_I(old_dentry->d_inode)->i_projid)))
+ return -EXDEV;
+
+ err = dquot_initialize(dir);
+ if (err)
+ return err;
+ return __ext4_link(dir, inode, dentry);
+}
+
/*
* Try to find buffer head where contains the parent block.
--
2.24.1.735.g03f4e72817-goog
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PATCH v4 11/20] ext4: add fast commit track points
2019-12-24 8:13 [PATCH v4 01/20] ext4: update docs for fast commit feature Harshad Shirwadkar
` (8 preceding siblings ...)
2019-12-24 8:13 ` [PATCH v4 10/20] ext4: break ext4_unlink() and ext4_link() Harshad Shirwadkar
@ 2019-12-24 8:13 ` Harshad Shirwadkar
2019-12-24 8:13 ` [PATCH v4 12/20] ext4: add fast commit on-disk format structs and helpers Harshad Shirwadkar
` (9 subsequent siblings)
19 siblings, 0 replies; 31+ messages in thread
From: Harshad Shirwadkar @ 2019-12-24 8:13 UTC (permalink / raw)
To: linux-ext4; +Cc: Harshad Shirwadkar
Previous patches in the series have added following tracking routines:
- ext4_fc_track_inode() -> tracks just the inode
- ext4_fc_track_create() -> tracks creation of an inode and remembers
its dirent
- ext4_fc_track_unlink() -> tracks inode unlink
- ext4_fc_track_link() -> tracks inode link
- ext4_fc_mark_ineligible() -> marks inode as ineligible for fast
commits
- ext4_fc_disable() -> marks entire file system as fast commit
ineligible
Add these different track points at various points in the file
system. This patch also adds high level stats to remember reasons why
inodes were marked ineligible.
Signed-off-by: Harshad Shirwadkar <harshadshirwadkar@gmail.com>
---
fs/ext4/acl.c | 1 +
fs/ext4/balloc.c | 3 +++
fs/ext4/ext4.h | 26 ++++++++++++++++++++++++++
fs/ext4/ext4_jbd2.c | 5 +++++
fs/ext4/extents.c | 5 +++++
fs/ext4/inline.c | 3 +++
fs/ext4/inode.c | 6 ++++++
fs/ext4/ioctl.c | 3 +++
fs/ext4/namei.c | 32 ++++++++++++++++++++++++++++++++
fs/ext4/super.c | 9 +++++++++
fs/ext4/xattr.c | 6 ++++++
11 files changed, 99 insertions(+)
diff --git a/fs/ext4/acl.c b/fs/ext4/acl.c
index 8c7bbf3e566d..28e9e04a8e96 100644
--- a/fs/ext4/acl.c
+++ b/fs/ext4/acl.c
@@ -257,6 +257,7 @@ ext4_set_acl(struct inode *inode, struct posix_acl *acl, int type)
inode->i_mode = mode;
inode->i_ctime = current_time(inode);
ext4_mark_inode_dirty(handle, inode);
+ ext4_fc_track_inode(inode);
}
out_stop:
ext4_journal_stop(handle);
diff --git a/fs/ext4/balloc.c b/fs/ext4/balloc.c
index 0b202e00d93f..14787065d030 100644
--- a/fs/ext4/balloc.c
+++ b/fs/ext4/balloc.c
@@ -654,6 +654,9 @@ ext4_fsblk_t ext4_new_meta_blocks(handle_t *handle, struct inode *inode,
ar.len = count ? *count : 1;
ar.flags = flags;
+ ext4_fc_mark_ineligible(inode,
+ EXT4_FC_REASON_META_ALLOC);
+
ret = ext4_mb_new_blocks(handle, &ar, errp);
if (count)
*count = ar.len;
diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index cc3a1489eae4..ede039a01bab 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -1436,6 +1436,31 @@ struct ext4_super_block {
#define ext4_has_strict_mode(sbi) \
(sbi->s_encoding_flags & EXT4_ENC_STRICT_MODE_FL)
+/*
+ * Fast commit ineligible reasons.
+ */
+enum {
+ EXT4_FC_REASON_META_ALLOC,
+ EXT4_FC_REASON_QUOTA,
+ EXT4_FC_REASON_XATTR,
+ EXT4_FC_REASON_CROSS_RENAME,
+ EXT4_FC_REASON_FALLOC_RANGE_OP,
+ EXT4_FC_REASON_JOURNAL_FLAG_CHANGE,
+ EXT4_FC_REASON_DELETE,
+ EXT4_FC_REASON_MEM,
+ EXT4_FC_REASON_SWAP_BOOT,
+ EXT4_FC_REASON_RESIZE,
+ EXT4_FC_REASON_RENAME_DIR,
+ EXT4_FC_REASON_MAX
+};
+
+struct ext4_fc_stats {
+ int fc_ineligible_reason_count[EXT4_FC_REASON_MAX];
+ int fc_num_commits;
+ int fc_ineligible_commits;
+ int fc_numblks;
+};
+
/*
* fourth extended-fs super-block data in memory
*/
@@ -1618,6 +1643,7 @@ struct ext4_sb_info {
*/
struct list_head s_fc_dentry_q;
spinlock_t s_fc_lock;
+ struct ext4_fc_stats s_fc_stats;
};
static inline struct ext4_sb_info *EXT4_SB(struct super_block *sb)
diff --git a/fs/ext4/ext4_jbd2.c b/fs/ext4/ext4_jbd2.c
index 7c27f9284064..9e060ba927c1 100644
--- a/fs/ext4/ext4_jbd2.c
+++ b/fs/ext4/ext4_jbd2.c
@@ -390,6 +390,8 @@ void ext4_fc_mark_ineligible(struct inode *inode, int reason)
(EXT4_SB(inode->i_sb)->s_mount_state & EXT4_FC_REPLAY))
return;
+ WARN_ON(reason >= EXT4_FC_REASON_MAX);
+ sbi->s_fc_stats.fc_ineligible_reason_count[reason]++;
if (sbi->s_journal)
ei->i_fc_tid = get_running_txn_tid(inode->i_sb);
ext4_clear_inode_state(inode, EXT4_STATE_FC_ELIGIBLE);
@@ -402,6 +404,8 @@ void ext4_fc_disable(struct super_block *sb, int reason)
struct ext4_sb_info *sbi = EXT4_SB(sb);
sbi->s_mount_state |= EXT4_FC_INELIGIBLE;
+ WARN_ON(reason >= EXT4_FC_REASON_MAX);
+ sbi->s_fc_stats.fc_ineligible_reason_count[reason]++;
}
/*
@@ -463,6 +467,7 @@ static int __ext4_dentry_update(struct inode *inode, void *arg, bool update)
write_unlock(&ei->i_fc_lock);
node = kmem_cache_alloc(ext4_fc_dentry_cachep, GFP_NOFS);
if (!node) {
+ ext4_fc_disable(inode->i_sb, EXT4_FC_REASON_MEM);
write_lock(&ei->i_fc_lock);
return -ENOMEM;
}
diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c
index fb0f99dc8c22..43119ad5970b 100644
--- a/fs/ext4/extents.c
+++ b/fs/ext4/extents.c
@@ -4832,6 +4832,7 @@ static long ext4_zero_range(struct file *file, loff_t offset,
ext4_set_inode_flag(inode, EXT4_INODE_EOFBLOCKS);
}
ext4_mark_inode_dirty(handle, inode);
+ ext4_fc_track_inode(inode);
/* Zero out partial block at the edges of the range */
ret = ext4_zero_partial_blocks(handle, inode, offset, len);
@@ -5563,6 +5564,8 @@ int ext4_collapse_range(struct inode *inode, loff_t offset, loff_t len)
if (IS_SYNC(inode))
ext4_handle_sync(handle);
inode->i_mtime = inode->i_ctime = current_time(inode);
+ ext4_fc_mark_ineligible(inode,
+ EXT4_FC_REASON_FALLOC_RANGE_OP);
ext4_mark_inode_dirty(handle, inode);
ext4_update_inode_fsync_trans(handle, inode, 1);
@@ -5677,6 +5680,8 @@ int ext4_insert_range(struct inode *inode, loff_t offset, loff_t len)
inode->i_size += len;
EXT4_I(inode)->i_disksize += len;
inode->i_mtime = inode->i_ctime = current_time(inode);
+ ext4_fc_mark_ineligible(inode,
+ EXT4_FC_REASON_FALLOC_RANGE_OP);
ret = ext4_mark_inode_dirty(handle, inode);
if (ret)
goto out_stop;
diff --git a/fs/ext4/inline.c b/fs/ext4/inline.c
index 2fec62d764fa..b3a2439a272e 100644
--- a/fs/ext4/inline.c
+++ b/fs/ext4/inline.c
@@ -759,6 +759,7 @@ int ext4_write_inline_data_end(struct inode *inode, loff_t pos, unsigned len,
ext4_write_unlock_xattr(inode, &no_expand);
brelse(iloc.bh);
+ ext4_fc_track_inode(inode);
mark_inode_dirty(inode);
out:
return copied;
@@ -974,6 +975,7 @@ int ext4_da_write_inline_data_end(struct inode *inode, loff_t pos,
* ordering of page lock and transaction start for journaling
* filesystems.
*/
+ ext4_fc_track_inode(inode);
mark_inode_dirty(inode);
return copied;
@@ -1986,6 +1988,7 @@ int ext4_inline_data_truncate(struct inode *inode, int *has_inline)
if (err == 0) {
inode->i_mtime = inode->i_ctime = current_time(inode);
+ ext4_fc_track_inode(inode);
err = ext4_mark_inode_dirty(handle, inode);
if (IS_SYNC(inode))
ext4_handle_sync(handle);
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 07c8da778368..2a61c3a31b74 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -2948,6 +2948,7 @@ static int ext4_writepages(struct address_space *mapping,
out_writepages:
trace_ext4_writepages_result(inode, wbc, ret,
nr_to_write - wbc->nr_to_write);
+ ext4_fc_track_inode(inode);
percpu_up_read(&sbi->s_journal_flag_rwsem);
return ret;
}
@@ -5572,6 +5573,7 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr)
if (attr->ia_valid & ATTR_GID)
inode->i_gid = attr->ia_gid;
error = ext4_mark_inode_dirty(handle, inode);
+ ext4_fc_track_inode(inode);
ext4_journal_stop(handle);
}
@@ -5690,6 +5692,7 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr)
if (!error) {
setattr_copy(inode, attr);
+ ext4_fc_track_inode(inode);
mark_inode_dirty(inode);
}
@@ -6102,6 +6105,7 @@ void ext4_dirty_inode(struct inode *inode, int flags)
goto out;
ext4_mark_inode_dirty(handle, inode);
+ ext4_fc_track_inode(inode);
ext4_journal_stop(handle);
out:
@@ -6187,6 +6191,8 @@ int ext4_change_inode_journal_flag(struct inode *inode, int val)
if (IS_ERR(handle))
return PTR_ERR(handle);
+ ext4_fc_mark_ineligible(inode,
+ EXT4_FC_REASON_JOURNAL_FLAG_CHANGE);
err = ext4_mark_inode_dirty(handle, inode);
ext4_handle_sync(handle);
ext4_journal_stop(handle);
diff --git a/fs/ext4/ioctl.c b/fs/ext4/ioctl.c
index 0b7f316fd30f..2bc655b2164e 100644
--- a/fs/ext4/ioctl.c
+++ b/fs/ext4/ioctl.c
@@ -204,6 +204,8 @@ static long swap_inode_boot_loader(struct super_block *sb,
ext4_discard_preallocations(inode);
+ if (EXT4_SB(sb)->s_journal)
+ ext4_fc_disable(sb, EXT4_FC_REASON_SWAP_BOOT);
err = ext4_mark_inode_dirty(handle, inode);
if (err < 0) {
/* No need to update quota information. */
@@ -1080,6 +1082,7 @@ long ext4_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
err = ext4_resize_fs(sb, n_blocks_count);
if (EXT4_SB(sb)->s_journal) {
+ ext4_fc_disable(sb, EXT4_FC_REASON_RESIZE);
jbd2_journal_lock_updates(EXT4_SB(sb)->s_journal);
err2 = jbd2_journal_flush(EXT4_SB(sb)->s_journal);
jbd2_journal_unlock_updates(EXT4_SB(sb)->s_journal);
diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c
index a405564ae02f..b732c0bb1d51 100644
--- a/fs/ext4/namei.c
+++ b/fs/ext4/namei.c
@@ -2595,6 +2595,8 @@ static int ext4_create(struct inode *dir, struct dentry *dentry, umode_t mode,
err = ext4_add_nondir(handle, dentry, inode);
if (!err && IS_DIRSYNC(dir))
ext4_handle_sync(handle);
+ if (!err)
+ ext4_fc_track_create(inode, dentry);
}
if (handle)
ext4_journal_stop(handle);
@@ -2627,6 +2629,9 @@ static int ext4_mknod(struct inode *dir, struct dentry *dentry,
err = ext4_add_nondir(handle, dentry, inode);
if (!err && IS_DIRSYNC(dir))
ext4_handle_sync(handle);
+ if (!err)
+ ext4_fc_track_create(inode, dentry);
+
}
if (handle)
ext4_journal_stop(handle);
@@ -2661,6 +2666,8 @@ static int ext4_tmpfile(struct inode *dir, struct dentry *dentry, umode_t mode)
err = ext4_orphan_add(handle, inode);
if (err)
goto err_unlock_inode;
+
+ ext4_fc_track_inode(inode);
mark_inode_dirty(inode);
unlock_new_inode(inode);
}
@@ -2784,6 +2791,7 @@ static int ext4_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode)
iput(inode);
goto out_stop;
}
+ ext4_fc_track_create(inode, dentry);
ext4_inc_count(handle, dir);
ext4_update_dx_flag(dir);
err = ext4_mark_inode_dirty(handle, dir);
@@ -3227,6 +3235,8 @@ static int ext4_unlink(struct inode *dir, struct dentry *dentry)
return retval;
retval = __ext4_unlink(dir, &dentry->d_name, d_inode(dentry));
+ if (!retval)
+ ext4_fc_track_unlink(d_inode(dentry), dentry);
#ifdef CONFIG_UNICODE
/* VFS negative dentries are incompatible with Encoding and
* Case-insensitiveness. Eventually we'll want avoid
@@ -3391,6 +3401,7 @@ int __ext4_link(struct inode *dir, struct inode *inode, struct dentry *dentry)
err = ext4_add_entry(handle, dentry, inode);
if (!err) {
+ ext4_fc_track_link(inode, dentry);
ext4_mark_inode_dirty(handle, inode);
/* this can happen only for tmpfile being
* linked the first time
@@ -3839,6 +3850,23 @@ static int ext4_rename(struct inode *old_dir, struct dentry *old_dentry,
ext4_mark_inode_dirty(handle, new.dir);
}
}
+
+ if (S_ISDIR(old.inode->i_mode)) {
+ /*
+ * We disable fast commits here that's because the
+ * replay code is not yet capable of changing dot dot
+ * dirents in directories. Since this is a metadata
+ * update that's ineligible, we need to mark entire fs
+ * as ineligbile.
+ */
+ ext4_fc_disable(old.inode->i_sb, EXT4_FC_REASON_RENAME_DIR);
+ } else {
+ if (new.inode)
+ ext4_fc_track_unlink(new.inode, new.dentry);
+ ext4_fc_track_link(old.inode, new.dentry);
+ ext4_fc_track_unlink(old.inode, old.dentry);
+ }
+
ext4_mark_inode_dirty(handle, old.dir);
if (new.inode) {
ext4_mark_inode_dirty(handle, new.inode);
@@ -3975,7 +4003,11 @@ static int ext4_cross_rename(struct inode *old_dir, struct dentry *old_dentry,
ctime = current_time(old.inode);
old.inode->i_ctime = ctime;
new.inode->i_ctime = ctime;
+ ext4_fc_mark_ineligible(old.inode,
+ EXT4_FC_REASON_CROSS_RENAME);
ext4_mark_inode_dirty(handle, old.inode);
+ ext4_fc_mark_ineligible(new.inode,
+ EXT4_FC_REASON_CROSS_RENAME);
ext4_mark_inode_dirty(handle, new.inode);
if (old.dir_bh) {
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index 538ee986d7f7..23a194d871c6 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -1330,6 +1330,7 @@ static int ext4_set_context(struct inode *inode, const void *ctx, size_t len,
* S_DAX may be disabled
*/
ext4_set_inode_flags(inode);
+ ext4_fc_track_inode(inode);
res = ext4_mark_inode_dirty(handle, inode);
if (res)
EXT4_ERROR_INODE(inode, "Failed to mark inode dirty");
@@ -4337,6 +4338,8 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
INIT_LIST_HEAD(&sbi->s_fc_dentry_q);
sbi->s_mount_state &= ~EXT4_FC_INELIGIBLE;
spin_lock_init(&sbi->s_fc_lock);
+ memset(&sbi->s_fc_stats, 0, sizeof(sbi->s_fc_stats));
+
sb->s_root = NULL;
needs_recovery = (es->s_last_orphan != 0 ||
@@ -5829,6 +5832,8 @@ static int ext4_quota_on(struct super_block *sb, int type, int format_id,
EXT4_I(inode)->i_flags |= EXT4_NOATIME_FL | EXT4_IMMUTABLE_FL;
inode_set_flags(inode, S_NOATIME | S_IMMUTABLE,
S_NOATIME | S_IMMUTABLE);
+ ext4_fc_mark_ineligible(inode,
+ EXT4_FC_REASON_QUOTA);
ext4_mark_inode_dirty(handle, inode);
ext4_journal_stop(handle);
unlock_inode:
@@ -5936,6 +5941,8 @@ static int ext4_quota_off(struct super_block *sb, int type)
EXT4_I(inode)->i_flags &= ~(EXT4_NOATIME_FL | EXT4_IMMUTABLE_FL);
inode_set_flags(inode, 0, S_NOATIME | S_IMMUTABLE);
inode->i_mtime = inode->i_ctime = current_time(inode);
+ ext4_fc_mark_ineligible(inode,
+ EXT4_FC_REASON_QUOTA);
ext4_mark_inode_dirty(handle, inode);
ext4_journal_stop(handle);
out_unlock:
@@ -6042,6 +6049,8 @@ static ssize_t ext4_quota_write(struct super_block *sb, int type,
if (inode->i_size < off + len) {
i_size_write(inode, off + len);
EXT4_I(inode)->i_disksize = inode->i_size;
+ ext4_fc_mark_ineligible(inode,
+ EXT4_FC_REASON_QUOTA);
ext4_mark_inode_dirty(handle, inode);
}
return len;
diff --git a/fs/ext4/xattr.c b/fs/ext4/xattr.c
index 491f9ee4040e..4f1744caf9f7 100644
--- a/fs/ext4/xattr.c
+++ b/fs/ext4/xattr.c
@@ -2433,6 +2433,8 @@ ext4_xattr_set_handle(handle_t *handle, struct inode *inode, int name_index,
if (IS_SYNC(inode))
ext4_handle_sync(handle);
}
+ ext4_fc_mark_ineligible(inode,
+ EXT4_FC_REASON_XATTR);
cleanup:
brelse(is.iloc.bh);
@@ -2510,6 +2512,8 @@ ext4_xattr_set(struct inode *inode, int name_index, const char *name,
if (error == 0)
error = error2;
}
+ ext4_fc_mark_ineligible(inode,
+ EXT4_FC_REASON_XATTR);
return error;
}
@@ -2942,6 +2946,8 @@ int ext4_xattr_delete_inode(handle_t *handle, struct inode *inode,
error);
goto cleanup;
}
+ ext4_fc_mark_ineligible(inode,
+ EXT4_FC_REASON_XATTR);
}
error = 0;
cleanup:
--
2.24.1.735.g03f4e72817-goog
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PATCH v4 12/20] ext4: add fast commit on-disk format structs and helpers
2019-12-24 8:13 [PATCH v4 01/20] ext4: update docs for fast commit feature Harshad Shirwadkar
` (9 preceding siblings ...)
2019-12-24 8:13 ` [PATCH v4 11/20] ext4: add fast commit track points Harshad Shirwadkar
@ 2019-12-24 8:13 ` Harshad Shirwadkar
2019-12-24 8:13 ` [PATCH v4 13/20] jbd2: add new APIs for commit path of fast commits Harshad Shirwadkar
` (8 subsequent siblings)
19 siblings, 0 replies; 31+ messages in thread
From: Harshad Shirwadkar @ 2019-12-24 8:13 UTC (permalink / raw)
To: linux-ext4; +Cc: Harshad Shirwadkar
Add structs representing on-disk format of the commit header and
tlvs in the commit header. Also, add helpers to arrange data
in on-disk fast commit format.
Signed-off-by: Harshad Shirwadkar <harshadshirwadkar@gmail.com>
---
fs/ext4/ext4_jbd2.c | 163 ++++++++++++++++++++++++++++++++++++++++++++
fs/ext4/ext4_jbd2.h | 45 ++++++++++++
2 files changed, 208 insertions(+)
diff --git a/fs/ext4/ext4_jbd2.c b/fs/ext4/ext4_jbd2.c
index 9e060ba927c1..9e34aa560ea1 100644
--- a/fs/ext4/ext4_jbd2.c
+++ b/fs/ext4/ext4_jbd2.c
@@ -4,6 +4,7 @@
*/
#include "ext4_jbd2.h"
+#include "ext4_extents.h"
#include <trace/events/ext4.h>
@@ -604,6 +605,168 @@ void ext4_fc_track_range(struct inode *inode, ext4_lblk_t start,
trace_ext4_fc_track_range(inode, start, end, ret);
}
+
+/*
+ * Adds tag, length and value at memory pointed to by dst. Returns
+ * true if tlv was added. Returns false if there's not enough space.
+ * If successful also updates *dst to point to the end of this tlv.
+ */
+static bool fc_try_add_tlv(u8 **dst, u8 *end, u16 tag, u16 len, u8 *val)
+{
+ struct ext4_fc_tl tl;
+
+ if (*dst + sizeof(tl) + len >= end)
+ return false;
+
+ tl.fc_tag = cpu_to_le16(tag);
+ tl.fc_len = cpu_to_le16(len);
+ memcpy(*dst, &tl, sizeof(tl));
+ memcpy(*dst + sizeof(tl), val, len);
+
+ *dst = *dst + sizeof(tl) + len;
+ return true;
+}
+
+/* Same as above, but tries to add dentry tlv. */
+static bool fc_try_add_dentry_info_tlv(u8 **dst, u8 *end, u16 tag,
+ int parent_ino, int ino, int dlen,
+ const unsigned char *dname)
+{
+ struct ext4_fc_dentry_info fcd;
+ struct ext4_fc_tl tl;
+
+
+ if (*dst + sizeof(tl) + sizeof(fcd) + dlen >= end)
+ return false;
+
+ fcd.fc_parent_ino = cpu_to_le32(parent_ino);
+ fcd.fc_ino = cpu_to_le32(ino);
+ tl.fc_tag = cpu_to_le16(tag);
+ tl.fc_len = cpu_to_le16(sizeof(fcd) + dlen);
+ memcpy(*dst, &tl, sizeof(tl));
+ *dst += sizeof(tl);
+ memcpy(*dst, &fcd, sizeof(fcd));
+ *dst += sizeof(fcd);
+ memcpy(*dst, dname, dlen);
+ *dst += dlen;
+
+ return true;
+}
+
+/* Get length of a particular tlv */
+static int fc_tag_len(struct ext4_fc_tl *tl)
+{
+ return le16_to_cpu(tl->fc_len);
+}
+
+/* Get a pointer to "value" of a tlv */
+static u8 *fc_tag_val(struct ext4_fc_tl *tl)
+{
+ return (u8 *)tl + sizeof(*tl);
+}
+
+/*
+ * Writes fast commit header and inode structure at memory
+ * pointed to by start. Returns 0 on success, error on failure.
+ * If successful, *last is upadated to point to the end of
+ * inode that was copied.
+ */
+static int fc_write_hdr(struct inode *inode, u8 *start, u8 *end,
+ u8 **last)
+{
+ struct ext4_fc_commit_hdr *fc_hdr = (struct ext4_fc_commit_hdr *)start;
+ struct ext4_inode_info *ei = EXT4_I(inode);
+ int inode_len = EXT4_GOOD_OLD_INODE_SIZE;
+ struct ext4_iloc iloc;
+ u8 *cur = start;
+ int ret;
+
+ if (ext4_is_inode_fc_ineligible(inode))
+ return -ECANCELED;
+
+ ret = ext4_get_inode_loc(inode, &iloc);
+ if (ret)
+ return ret;
+
+ fc_hdr->fc_magic = cpu_to_le32(EXT4_FC_MAGIC);
+ fc_hdr->fc_ino = cpu_to_le32(inode->i_ino);
+ fc_hdr->fc_features = 0;
+ fc_hdr->fc_csum = 0;
+
+ cur = (u8 *)(fc_hdr + 1);
+ if (EXT4_INODE_SIZE(inode->i_sb) > EXT4_GOOD_OLD_INODE_SIZE)
+ inode_len += ei->i_extra_isize;
+ if (cur + inode_len >= end)
+ return -ECANCELED;
+
+ memcpy(cur, ext4_raw_inode(&iloc), inode_len);
+ cur += inode_len;
+ *last = cur;
+
+ return 0;
+}
+
+/*
+ * Writes data tags (EXT4_FC_TAG_ADD_RANGE / EXT4_FC_TAG_DEL_RANGE)
+ * at memory pointed to by start. Returns number of TLVs that were
+ * added if successfully. Returns errors otherwise.
+ */
+static int fc_write_data(struct inode *inode, u8 *start, u8 *end,
+ u8 **last)
+{
+ ext4_lblk_t old_blk_size, cur_lblk_off, new_blk_size;
+ struct ext4_inode_info *ei = EXT4_I(inode);
+ struct ext4_map_blocks map;
+ struct ext4_extent extent;
+ struct ext4_fc_lrange lrange;
+ u8 *cur = start;
+ int num_tlvs = 0;
+ int ret;
+
+ write_lock(&ei->i_fc_lock);
+ old_blk_size = ei->i_fc_lblk_start;
+ new_blk_size = ei->i_fc_lblk_end;
+ ei->i_fc_lblk_start = ei->i_fc_lblk_end;
+ write_unlock(&ei->i_fc_lock);
+
+ cur_lblk_off = old_blk_size;
+ jbd_debug(1, "%s: will try writing %ld to %ld for inode %ld\n",
+ __func__, cur_lblk_off, new_blk_size, inode->i_ino);
+ while (cur_lblk_off <= new_blk_size) {
+ map.m_lblk = cur_lblk_off;
+ map.m_len = new_blk_size - cur_lblk_off + 1;
+ ret = ext4_map_blocks(NULL, inode, &map, 0);
+ if (ret < 0)
+ return ret;
+ if (map.m_len == 0)
+ return -ECANCELED;
+ if (map.m_flags & EXT4_MAP_UNWRITTEN)
+ return -ECANCELED;
+
+ cur_lblk_off += map.m_len;
+ if (ret == 0) {
+ lrange.fc_lblk = cpu_to_le32(map.m_lblk);
+ lrange.fc_len = cpu_to_le32(map.m_len);
+ if (!fc_try_add_tlv(&cur, end, EXT4_FC_TAG_DEL_RANGE,
+ sizeof(lrange), (u8 *)&lrange))
+ return -ENOSPC;
+
+ } else {
+ extent.ee_block = cpu_to_le32(map.m_lblk);
+ extent.ee_len = cpu_to_le16(map.m_len);
+ ext4_ext_store_pblock(&extent, map.m_pblk);
+ ext4_ext_mark_initialized(&extent);
+ if (!fc_try_add_tlv(&cur, end, EXT4_FC_TAG_ADD_RANGE,
+ sizeof(struct ext4_extent), (u8 *)&extent))
+ return -ENOSPC;
+ }
+ num_tlvs++;
+ }
+ *last = cur;
+
+ return num_tlvs;
+}
+
void ext4_init_fast_commit(struct super_block *sb, journal_t *journal)
{
if (!ext4_should_fast_commit(sb))
diff --git a/fs/ext4/ext4_jbd2.h b/fs/ext4/ext4_jbd2.h
index 60f484377c2e..d3a951d2e58d 100644
--- a/fs/ext4/ext4_jbd2.h
+++ b/fs/ext4/ext4_jbd2.h
@@ -470,7 +470,52 @@ static inline int ext4_should_dioread_nolock(struct inode *inode)
return 1;
}
+/* Ext4 fast commit related info */
+
+/* Magic of fast commit header */
+#define EXT4_FC_MAGIC 0xE2540090
+
#define EXT4_NUM_FC_BLKS 128
+
+struct ext4_fc_commit_hdr {
+ /* Fast commit magic, should be EXT4_FC_MAGIC */
+ __le32 fc_magic;
+ /* Features used by this fast commit block */
+ __u8 fc_features;
+ /* Number of TLVs in this fast commmit block */
+ __le16 fc_num_tlvs;
+ /* Inode number */
+ __le32 fc_ino;
+ /* Csum(hdr+contents) */
+ __le32 fc_csum;
+};
+
+/* Fast commit on disk tag length structure */
+struct ext4_fc_tl {
+ __le16 fc_tag;
+ __le16 fc_len;
+};
+
+/* On disk fast commit tlv value structure for dirent tags:
+ * - EXT4_FC_TAG_CREATE_DENTRY
+ * - EXT4_FC_TAG_ADD_DENTRY
+ * - EXT4_FC_TAG_DEL_DENTRY
+ */
+struct ext4_fc_dentry_info {
+ __le32 fc_parent_ino;
+ __le32 fc_ino;
+ u8 fc_dname[0];
+};
+
+/*
+ * On disk fast commit tlv value structure for tag
+ * EXT4_FC_TAG_HOLE.
+ */
+struct ext4_fc_lrange {
+ __le32 fc_lblk;
+ __le32 fc_len;
+};
+
void ext4_init_fast_commit(struct super_block *sb, journal_t *journal);
void ext4_init_inode_fc_info(struct inode *inode);
void ext4_fc_track_range(struct inode *inode, ext4_lblk_t start,
--
2.24.1.735.g03f4e72817-goog
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PATCH v4 13/20] jbd2: add new APIs for commit path of fast commits
2019-12-24 8:13 [PATCH v4 01/20] ext4: update docs for fast commit feature Harshad Shirwadkar
` (10 preceding siblings ...)
2019-12-24 8:13 ` [PATCH v4 12/20] ext4: add fast commit on-disk format structs and helpers Harshad Shirwadkar
@ 2019-12-24 8:13 ` Harshad Shirwadkar
2019-12-24 8:13 ` [PATCH v4 14/20] ext4: main commit routine for " Harshad Shirwadkar
` (7 subsequent siblings)
19 siblings, 0 replies; 31+ messages in thread
From: Harshad Shirwadkar @ 2019-12-24 8:13 UTC (permalink / raw)
To: linux-ext4; +Cc: Harshad Shirwadkar
Add following helpers for commit path:
- jbd2_map_fc_buf() - allocates fast commit buffers for caller
- jbd2_wait_on_fc_bufs() - waits on fast commit buffers allocated
using jbd2_map_fc_buf()
- jbd2_submit_inode_data() - submit data buffers for one inode
- jbd2_wait_inode_data() - wait for inode data
Signed-off-by: Harshad Shirwadkar <harshadshirwadkar@gmail.com>
---
fs/jbd2/commit.c | 40 +++++++++++++++++++++++
fs/jbd2/journal.c | 76 ++++++++++++++++++++++++++++++++++++++++++++
include/linux/jbd2.h | 6 ++++
3 files changed, 122 insertions(+)
diff --git a/fs/jbd2/commit.c b/fs/jbd2/commit.c
index ffbea070582c..20a1bd955784 100644
--- a/fs/jbd2/commit.c
+++ b/fs/jbd2/commit.c
@@ -202,6 +202,46 @@ static int journal_submit_inode_data_buffers(struct address_space *mapping,
return ret;
}
+/* Send all the data buffers related to an inode */
+int jbd2_submit_inode_data(journal_t *journal, struct jbd2_inode *jinode)
+{
+ struct address_space *mapping;
+ loff_t dirty_start;
+ loff_t dirty_end;
+ int ret;
+
+ if (!jinode)
+ return 0;
+
+ dirty_start = jinode->i_dirty_start;
+ dirty_end = jinode->i_dirty_end;
+
+ if (!(jinode->i_flags & JI_WRITE_DATA))
+ return 0;
+
+ dirty_start = jinode->i_dirty_start;
+ dirty_end = jinode->i_dirty_end;
+
+ mapping = jinode->i_vfs_inode->i_mapping;
+
+ trace_jbd2_submit_inode_data(jinode->i_vfs_inode);
+ ret = journal_submit_inode_data_buffers(mapping, dirty_start,
+ dirty_end);
+
+ return ret;
+}
+EXPORT_SYMBOL(jbd2_submit_inode_data);
+
+int jbd2_wait_inode_data(journal_t *journal, struct jbd2_inode *jinode)
+{
+ if (!jinode || !(jinode->i_flags & JI_WAIT_DATA))
+ return 0;
+ return filemap_fdatawait_range_keep_errors(
+ jinode->i_vfs_inode->i_mapping, jinode->i_dirty_start,
+ jinode->i_dirty_end);
+}
+EXPORT_SYMBOL(jbd2_wait_inode_data);
+
/*
* Submit all the data buffers of inode associated with the transaction to
* disk.
diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c
index bd6ab127f0d9..e0b2e8934fa4 100644
--- a/fs/jbd2/journal.c
+++ b/fs/jbd2/journal.c
@@ -865,6 +865,82 @@ int jbd2_journal_next_log_block(journal_t *journal, unsigned long long *retp)
return jbd2_journal_bmap(journal, blocknr, retp);
}
+/* Map one fast commit buffer for use by the file system */
+int jbd2_map_fc_buf(journal_t *journal, struct buffer_head **bh_out)
+{
+ unsigned long long pblock;
+ unsigned long blocknr;
+ int ret = 0;
+ struct buffer_head *bh;
+ int fc_off;
+ journal_header_t *jhdr;
+
+ write_lock(&journal->j_state_lock);
+
+ if (journal->j_fc_off + journal->j_first_fc < journal->j_last_fc) {
+ fc_off = journal->j_fc_off;
+ blocknr = journal->j_first_fc + fc_off;
+ journal->j_fc_off++;
+ } else {
+ ret = -EINVAL;
+ }
+ write_unlock(&journal->j_state_lock);
+
+ if (ret)
+ return ret;
+
+ ret = jbd2_journal_bmap(journal, blocknr, &pblock);
+ if (ret)
+ return ret;
+
+ bh = __getblk(journal->j_dev, pblock, journal->j_blocksize);
+ if (!bh)
+ return -ENOMEM;
+
+ lock_buffer(bh);
+ jhdr = (journal_header_t *)bh->b_data;
+ jhdr->h_magic = cpu_to_be32(JBD2_MAGIC_NUMBER);
+ jhdr->h_blocktype = cpu_to_be32(JBD2_FC_BLOCK);
+ jhdr->h_sequence = cpu_to_be32(journal->j_running_transaction->t_tid);
+
+ set_buffer_uptodate(bh);
+ unlock_buffer(bh);
+ journal->j_fc_wbuf[fc_off] = bh;
+
+ *bh_out = bh;
+
+ return 0;
+}
+EXPORT_SYMBOL(jbd2_map_fc_buf);
+
+/*
+ * Wait on fast commit buffers that were allocated by jbd2_map_fc_buf
+ * for completion.
+ */
+int jbd2_wait_on_fc_bufs(journal_t *journal, int num_blks)
+{
+ struct buffer_head *bh;
+ int i, j_fc_off;
+
+ read_lock(&journal->j_state_lock);
+ j_fc_off = journal->j_fc_off;
+ read_unlock(&journal->j_state_lock);
+
+ /*
+ * Wait in reverse order to minimize chances of us being woken up before
+ * all IOs have completed
+ */
+ for (i = j_fc_off - 1; i >= j_fc_off - num_blks; i--) {
+ bh = journal->j_fc_wbuf[i];
+ wait_on_buffer(bh);
+ if (unlikely(!buffer_uptodate(bh)))
+ return -EIO;
+ }
+
+ return 0;
+}
+EXPORT_SYMBOL(jbd2_wait_on_fc_bufs);
+
/*
* Conversion of logical to physical block numbers for the journal
*
diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h
index 99b0f50ceb50..d450dcb93e51 100644
--- a/include/linux/jbd2.h
+++ b/include/linux/jbd2.h
@@ -123,6 +123,7 @@ typedef struct journal_s journal_t; /* Journal control structure */
#define JBD2_SUPERBLOCK_V1 3
#define JBD2_SUPERBLOCK_V2 4
#define JBD2_REVOKE_BLOCK 5
+#define JBD2_FC_BLOCK 6
/*
* Standard header for all descriptor blocks:
@@ -1562,6 +1563,11 @@ int jbd2_start_async_fc_nowait(journal_t *journal, tid_t tid);
int jbd2_start_async_fc_wait(journal_t *journal, tid_t tid);
void jbd2_stop_async_fc(journal_t *journal, tid_t tid);
void jbd2_init_fast_commit(journal_t *journal, int num_fc_blks);
+int jbd2_map_fc_buf(journal_t *journal, struct buffer_head **bh_out);
+int jbd2_wait_on_fc_bufs(journal_t *journal, int num_blks);
+int jbd2_submit_inode_data(journal_t *journal, struct jbd2_inode *jinode);
+int jbd2_wait_inode_data(journal_t *journal, struct jbd2_inode *jinode);
+
/*
* is_journal_abort
*
--
2.24.1.735.g03f4e72817-goog
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PATCH v4 14/20] ext4: main commit routine for fast commits
2019-12-24 8:13 [PATCH v4 01/20] ext4: update docs for fast commit feature Harshad Shirwadkar
` (11 preceding siblings ...)
2019-12-24 8:13 ` [PATCH v4 13/20] jbd2: add new APIs for commit path of fast commits Harshad Shirwadkar
@ 2019-12-24 8:13 ` Harshad Shirwadkar
2019-12-27 16:36 ` kbuild test robot
2019-12-27 16:36 ` [RFC PATCH] ext4: submit_fc_bh() can be static kbuild test robot
2019-12-24 8:13 ` [PATCH v4 15/20] jbd2: add fast commit recovery path support Harshad Shirwadkar
` (6 subsequent siblings)
19 siblings, 2 replies; 31+ messages in thread
From: Harshad Shirwadkar @ 2019-12-24 8:13 UTC (permalink / raw)
To: linux-ext4; +Cc: Harshad Shirwadkar
Add main commit routine for fast commits to perform a fast commit.
Also, handle race between inode deletion which results in inode
being deleted from fast commit list and the commit routine. Add
commit routines for fast commit with hard consistency as well
as for soft consistency.
Signed-off-by: Harshad Shirwadkar <harshadshirwadkar@gmail.com>
---
fs/ext4/ext4.h | 7 +
fs/ext4/ext4_jbd2.c | 563 ++++++++++++++++++++++++++++++++++++
fs/ext4/ext4_jbd2.h | 2 +
fs/ext4/fsync.c | 2 +-
fs/ext4/inode.c | 6 +-
fs/ext4/migrate.c | 1 +
fs/ext4/super.c | 12 +
include/trace/events/ext4.h | 101 +++++++
8 files changed, 691 insertions(+), 3 deletions(-)
diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index ede039a01bab..9c2f67c64b4f 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -1679,6 +1679,10 @@ enum {
EXT4_STATE_LUSTRE_EA_INODE, /* Lustre-style ea_inode */
EXT4_STATE_VERITY_IN_PROGRESS, /* building fs-verity Merkle tree */
EXT4_STATE_FC_ELIGIBLE, /* File is Fast commit eligible */
+ EXT4_STATE_FC_DATA_SUBMIT, /* File is going through fast commit */
+ EXT4_STATE_FC_MDATA_SUBMIT, /* Fast commit block is
+ * being submitted
+ */
};
#define EXT4_INODE_BIT_FNS(name, field, offset) \
@@ -2767,6 +2771,9 @@ extern int ext4_group_extend(struct super_block *sb,
extern int ext4_resize_fs(struct super_block *sb, ext4_fsblk_t n_blocks_count);
/* super.c */
+
+int ext4_fc_async_commit_inode(journal_t *journal, tid_t commit_tid,
+ struct inode *inode);
extern struct buffer_head *ext4_sb_bread(struct super_block *sb,
sector_t block, int op_flags);
extern int ext4_seq_options_show(struct seq_file *seq, void *offset);
diff --git a/fs/ext4/ext4_jbd2.c b/fs/ext4/ext4_jbd2.c
index 9e34aa560ea1..f371f1f0f914 100644
--- a/fs/ext4/ext4_jbd2.c
+++ b/fs/ext4/ext4_jbd2.c
@@ -351,6 +351,8 @@ void ext4_init_inode_fc_info(struct inode *inode)
struct ext4_inode_info *ei = EXT4_I(inode);
ext4_reset_inode_fc_info(inode);
+ ext4_clear_inode_state(inode, EXT4_STATE_FC_DATA_SUBMIT);
+ ext4_clear_inode_state(inode, EXT4_STATE_FC_MDATA_SUBMIT);
INIT_LIST_HEAD(&ei->i_fc_list);
}
@@ -375,6 +377,43 @@ static inline tid_t get_running_txn_tid(struct super_block *sb)
return 0;
}
+void ext4_fc_del(struct inode *inode)
+{
+ if (!ext4_should_fast_commit(inode->i_sb) ||
+ (EXT4_SB(inode->i_sb)->s_mount_state & EXT4_FC_REPLAY))
+ return;
+
+ if (list_empty(&EXT4_I(inode)->i_fc_list))
+ return;
+
+ ext4_fc_disable(inode->i_sb, EXT4_FC_REASON_DELETE);
+
+restart:
+ spin_lock(&EXT4_SB(inode->i_sb)->s_fc_lock);
+ if (ext4_test_inode_state(inode, EXT4_STATE_FC_DATA_SUBMIT)) {
+ struct ext4_inode_info *ei = EXT4_I(inode);
+ wait_queue_head_t *wq;
+#if (BITS_PER_LONG < 64)
+ DEFINE_WAIT_BIT(wait, &ei->i_state_flags,
+ EXT4_STATE_FC_DATA_SUBMIT);
+ wq = bit_waitqueue(&ei->i_state_flags,
+ EXT4_STATE_FC_DATA_SUBMIT);
+#else
+ DEFINE_WAIT_BIT(wait, &ei->i_flags,
+ EXT4_STATE_FC_DATA_SUBMIT);
+ wq = bit_waitqueue(&ei->i_flags,
+ EXT4_STATE_FC_DATA_SUBMIT);
+#endif
+ prepare_to_wait(wq, &wait.wq_entry, TASK_UNINTERRUPTIBLE);
+ spin_unlock(&EXT4_SB(inode->i_sb)->s_fc_lock);
+ schedule();
+ finish_wait(wq, &wait.wq_entry);
+ goto restart;
+ }
+ list_del_init(&EXT4_I(inode)->i_fc_list);
+ spin_unlock(&EXT4_SB(inode->i_sb)->s_fc_lock);
+}
+
static bool ext4_is_inode_fc_ineligible(struct inode *inode)
{
if (get_running_txn_tid(inode->i_sb) == EXT4_I(inode)->i_fc_tid)
@@ -433,6 +472,7 @@ static int __ext4_fc_track_template(
return -EOPNOTSUPP;
write_lock(&ei->i_fc_lock);
+ ext4_clear_inode_state(inode, EXT4_STATE_FC_MDATA_SUBMIT);
if (running_txn_tid == ei->i_fc_tid) {
if (!ext4_test_inode_state(inode, EXT4_STATE_FC_ELIGIBLE)) {
write_unlock(&ei->i_fc_lock);
@@ -605,6 +645,21 @@ void ext4_fc_track_range(struct inode *inode, ext4_lblk_t start,
trace_ext4_fc_track_range(inode, start, end, ret);
}
+static void ext4_end_buffer_io_sync(struct buffer_head *bh, int uptodate)
+{
+ BUFFER_TRACE(bh, "");
+ if (uptodate) {
+ ext4_debug("%s: Block %lld up-to-date",
+ __func__, bh->b_blocknr);
+ set_buffer_uptodate(bh);
+ } else {
+ ext4_debug("%s: Block %lld not up-to-date",
+ __func__, bh->b_blocknr);
+ clear_buffer_uptodate(bh);
+ }
+
+ unlock_buffer(bh);
+}
/*
* Adds tag, length and value at memory pointed to by dst. Returns
@@ -767,10 +822,518 @@ static int fc_write_data(struct inode *inode, u8 *start, u8 *end,
return num_tlvs;
}
+void submit_fc_bh(struct buffer_head *bh)
+{
+ lock_buffer(bh);
+ clear_buffer_dirty(bh);
+ set_buffer_uptodate(bh);
+ bh->b_end_io = ext4_end_buffer_io_sync;
+ submit_bh(REQ_OP_WRITE, REQ_SYNC, bh);
+}
+
+static int fc_commit_data_inode(journal_t *journal, struct inode *inode)
+{
+ struct ext4_fc_commit_hdr *hdr;
+ struct buffer_head *bh;
+ u8 *start, *cur, *end;
+ int ret;
+ int num_tlvs = 0;
+
+ ret = jbd2_map_fc_buf(journal, &bh);
+ if (ret)
+ return -ECANCELED;
+
+ start = cur = ((__u8 *)bh->b_data + sizeof(journal_header_t));
+ end = (__u8 *)bh->b_data + journal->j_blocksize;
+ hdr = (struct ext4_fc_commit_hdr *)start;
+
+ ret = fc_write_hdr(inode, start, end, &cur);
+ if (ret < 0)
+ return ret;
+
+ ret = fc_write_data(inode, cur, end, &cur);
+ if (ret < 0)
+ return ret;
+ memset(cur, 0, end - cur);
+
+ hdr->fc_num_tlvs = cpu_to_le16(num_tlvs + ret);
+ hdr->fc_csum = 0;
+ hdr->fc_csum = cpu_to_le32(ext4_chksum(EXT4_SB(inode->i_sb),
+ 0, start, end - start));
+ submit_fc_bh(bh);
+ ext4_set_inode_state(inode, EXT4_STATE_FC_MDATA_SUBMIT);
+
+ return 1;
+}
+
+static int submit_all_inode_data(journal_t *journal)
+{
+ struct super_block *sb = (struct super_block *)(journal->j_private);
+ struct ext4_sb_info *sbi = EXT4_SB(sb);
+ struct ext4_inode_info *iter;
+ struct list_head *pos;
+ int ret = 0;
+
+ spin_lock(&sbi->s_fc_lock);
+ list_for_each(pos, &sbi->s_fc_q) {
+ iter = list_entry(pos, struct ext4_inode_info, i_fc_list);
+ ext4_set_inode_state(&iter->vfs_inode,
+ EXT4_STATE_FC_DATA_SUBMIT);
+ spin_unlock(&sbi->s_fc_lock);
+ ret = jbd2_submit_inode_data(journal, iter->jinode);
+ if (ret)
+ return ret;
+ spin_lock(&sbi->s_fc_lock);
+ }
+ spin_unlock(&sbi->s_fc_lock);
+
+ return ret;
+}
+
+static int wait_all_inode_data(journal_t *journal)
+{
+ struct super_block *sb = (struct super_block *)(journal->j_private);
+ struct ext4_sb_info *sbi = EXT4_SB(sb);
+ struct ext4_inode_info *pos, *n;
+ int ret = 0;
+
+ spin_lock(&sbi->s_fc_lock);
+ list_for_each_entry_safe(pos, n, &sbi->s_fc_q, i_fc_list) {
+ if (!ext4_test_inode_state(&pos->vfs_inode,
+ EXT4_STATE_FC_DATA_SUBMIT))
+ continue;
+ spin_unlock(&sbi->s_fc_lock);
+ ret = jbd2_wait_inode_data(journal, pos->jinode);
+ if (ret)
+ break;
+ spin_lock(&sbi->s_fc_lock);
+ list_safe_reset_next(pos, n, i_fc_list);
+ list_del_init(&pos->i_fc_list);
+
+ ext4_clear_inode_state(&pos->vfs_inode,
+ EXT4_STATE_FC_DATA_SUBMIT);
+ /* Make sure DATA_SUBMIT bit is set before waking up */
+ smp_mb();
+#if (BITS_PER_LONG < 64)
+ wake_up_bit(&pos->i_state_flags, EXT4_STATE_FC_DATA_SUBMIT);
+#else
+ wake_up_bit(&pos->i_flags, EXT4_STATE_FC_DATA_SUBMIT);
+#endif
+ }
+ spin_unlock(&sbi->s_fc_lock);
+
+ return 0;
+}
+
+static int fc_inode_match(struct inode *inode, void *data)
+{
+ if (inode->i_ino != (long)data)
+ return 0;
+
+ if (inode->i_nlink)
+ return 1;
+
+ /*
+ * Avoid returning a nearly dead inode (withi_nlink == 0).
+ */
+ if (ext4_test_inode_state(inode,
+ EXT4_STATE_FC_DATA_SUBMIT)) {
+ /*
+ * This is a tricky situation, after we
+ * submitted data for this inode, someone
+ * tried to free this. ext4_fc_del() is
+ * waiting on FC_DATA_SUBMIT bit to clear.
+ * Since we are never going to wait for data
+ * just wake the sleeper.
+ * TODO: Even in this case don't fallback to full commits
+ * and indicate the caller that this is a deleted inode.
+ */
+ ext4_clear_inode_state(
+ inode, EXT4_STATE_FC_DATA_SUBMIT);
+ /* Make sure that data_submit bit is set */
+ smp_mb();
+#if (BITS_PER_LONG < 64)
+ wake_up_bit(&EXT4_I(inode)->i_state_flags,
+ EXT4_STATE_FC_DATA_SUBMIT);
+#else
+ wake_up_bit(&EXT4_I(inode)->i_flags,
+ EXT4_STATE_FC_DATA_SUBMIT);
+#endif
+ }
+ return 0;
+}
+
+/*
+ * Commits all the dentry updates and respective inodes till and
+ * including "last".
+ */
+static int fc_commit_dentry_updates(journal_t *journal,
+ struct ext4_fc_dentry_update *last)
+{
+ struct super_block *sb = (struct super_block *)(journal->j_private);
+ struct ext4_sb_info *sbi = EXT4_SB(sb);
+ struct ext4_fc_commit_hdr *hdr;
+ struct ext4_fc_dentry_update *fc_dentry;
+ struct inode *inode;
+ struct buffer_head *bh;
+ u8 *start, *cur, *end;
+ int len, ret;
+ int nblks = 0;
+ int num_tlvs = 0;
+ bool is_last;
+
+ ret = jbd2_map_fc_buf(journal, &bh);
+ if (ret)
+ return -ECANCELED;
+
+ start = cur = ((__u8 *)bh->b_data + sizeof(journal_header_t));
+ end = (__u8 *)bh->b_data + journal->j_blocksize;
+ hdr = (struct ext4_fc_commit_hdr *)start;
+
+ spin_lock(&sbi->s_fc_lock);
+ while (!list_empty(&sbi->s_fc_dentry_q)) {
+ fc_dentry = list_first_entry(
+ &sbi->s_fc_dentry_q, struct ext4_fc_dentry_update,
+ fcd_list);
+ list_del_init(&fc_dentry->fcd_list);
+ spin_unlock(&sbi->s_fc_lock);
+ if (!fc_try_add_dentry_info_tlv(
+ &cur, end, fc_dentry->fcd_op,
+ fc_dentry->fcd_parent, fc_dentry->fcd_ino,
+ fc_dentry->fcd_name.len,
+ fc_dentry->fcd_name.name)) {
+ kmem_cache_free(ext4_fc_dentry_cachep, fc_dentry);
+ return -ENOSPC;
+ }
+ num_tlvs++;
+ inode = ilookup5_nowait(sb, fc_dentry->fcd_ino, fc_inode_match,
+ (void *)(long)fc_dentry->fcd_ino);
+ /*
+ * If this was the last metadata update for this inode, clear
+ * since we are going to handle it now.
+ */
+ if (inode && EXT4_I(inode)->i_fc_mdata_update == fc_dentry)
+ EXT4_I(inode)->i_fc_mdata_update = NULL;
+ if (fc_dentry != last &&
+ fc_dentry->fcd_op != EXT4_FC_TAG_CREAT_DENTRY) {
+ if (inode)
+ iput(inode);
+ spin_lock(&sbi->s_fc_lock);
+ kmem_cache_free(ext4_fc_dentry_cachep, fc_dentry);
+ continue;
+ }
+ is_last = (fc_dentry == last);
+ kmem_cache_free(ext4_fc_dentry_cachep, fc_dentry);
+ if (IS_ERR_OR_NULL(inode))
+ /*
+ * Inode got evicted from memory for some
+ * reason. it's possible that someone deleted
+ * the inode after we started fast commit.
+ * We just abort fast commits in this case.
+ */
+ return -ECANCELED;
+
+ /*
+ * It's either the last dentry update or it's inode
+ * creation. Until now, we have written all the
+ * directory entry updates since the beginning or
+ * the last creation in current fast commit buf.
+ * Move the contents towards the end of the block and
+ * then write header first. We move it to the end
+ * because header size is variable.
+ */
+ len = cur - start;
+ memmove(end - len, start, len);
+ ret = fc_write_hdr(inode, start, end - len, &cur);
+ if (ret < 0) {
+ iput(inode);
+ return ret;
+ }
+ /*
+ * Place directory entry updates right after the
+ * header and the inode and write remaining
+ * tags if any.
+ */
+ memmove(cur, end - len, len);
+ cur = cur + len;
+ if (inode->i_nlink) {
+ ret = fc_write_data(inode, cur, end, &cur);
+ if (ret < 0) {
+ iput(inode);
+ return ret;
+ }
+ }
+ memset(cur, 0, end - cur);
+ hdr->fc_num_tlvs = cpu_to_le16(num_tlvs + ret);
+ hdr->fc_csum = cpu_to_le32(
+ ext4_chksum(sbi, 0, start, end - start));
+ submit_fc_bh(bh);
+ nblks++;
+ if (!inode->i_nlink) {
+ ext4_clear_inode_state(inode,
+ EXT4_STATE_FC_DATA_SUBMIT);
+ smp_mb(); /* Make sure data submit bit is set */
+#if (BITS_PER_LONG < 64)
+ wake_up_bit(&EXT4_I(inode)->i_state_flags,
+ EXT4_STATE_FC_DATA_SUBMIT);
+#else
+ wake_up_bit(&EXT4_I(inode)->i_flags,
+ EXT4_STATE_FC_DATA_SUBMIT);
+#endif
+ } else if (!ext4_test_inode_state(inode,
+ EXT4_STATE_FC_DATA_SUBMIT)) {
+ ret = jbd2_submit_inode_data(
+ journal, EXT4_I(inode)->jinode);
+ if (ret < 0)
+ return ret;
+ ext4_set_inode_state(inode,
+ EXT4_STATE_FC_DATA_SUBMIT);
+ }
+ ext4_set_inode_state(inode, EXT4_STATE_FC_MDATA_SUBMIT);
+ iput(inode);
+ if (is_last) {
+ bh = NULL;
+ goto skip_unlock;
+ }
+ ret = jbd2_map_fc_buf(journal, &bh);
+ if (ret < 0)
+ return ret;
+ start = cur = ((__u8 *)bh->b_data + sizeof(journal_header_t));
+ hdr = (struct ext4_fc_commit_hdr *)start;
+ end = (__u8 *)bh->b_data + journal->j_blocksize;
+ memset(start, 0, end - start);
+ spin_lock(&sbi->s_fc_lock);
+ }
+
+ spin_unlock(&sbi->s_fc_lock);
+skip_unlock:
+ WARN_ON(bh != NULL);
+ return nblks;
+}
+
+static void ext4_journal_fc_cleanup_cb(journal_t *journal)
+{
+ struct super_block *sb = journal->j_private;
+ struct ext4_sb_info *sbi = EXT4_SB(sb);
+ struct ext4_inode_info *iter;
+ struct ext4_fc_dentry_update *fc_dentry;
+
+ spin_lock(&sbi->s_fc_lock);
+ while (!list_empty(&sbi->s_fc_q)) {
+ iter = list_first_entry(&sbi->s_fc_q,
+ struct ext4_inode_info, i_fc_list);
+ iter->i_fc_mdata_update = NULL;
+
+ list_del_init(&iter->i_fc_list);
+ ext4_clear_inode_state(&iter->vfs_inode,
+ EXT4_STATE_FC_DATA_SUBMIT);
+ ext4_clear_inode_state(&iter->vfs_inode,
+ EXT4_STATE_FC_MDATA_SUBMIT);
+ /* Make sure DATA_SUBMIT bit is set */
+ smp_mb();
+ wake_up_bit(&iter->i_flags, EXT4_STATE_FC_DATA_SUBMIT);
+ }
+ INIT_LIST_HEAD(&sbi->s_fc_q);
+ while (!list_empty(&sbi->s_fc_dentry_q)) {
+ fc_dentry = list_first_entry(&sbi->s_fc_dentry_q,
+ struct ext4_fc_dentry_update,
+ fcd_list);
+ list_del_init(&fc_dentry->fcd_list);
+ spin_unlock(&sbi->s_fc_lock);
+
+ if (fc_dentry->fcd_name.name &&
+ fc_dentry->fcd_name.len > DNAME_INLINE_LEN)
+ kfree(fc_dentry->fcd_name.name);
+ kmem_cache_free(ext4_fc_dentry_cachep, fc_dentry);
+ spin_lock(&sbi->s_fc_lock);
+ }
+ INIT_LIST_HEAD(&sbi->s_fc_dentry_q);
+ sbi->s_mount_state &= ~EXT4_FC_INELIGIBLE;
+ spin_unlock(&sbi->s_fc_lock);
+ trace_ext4_journal_fc_stats(sb);
+}
+
+int ext4_fc_perform_hard_commit(journal_t *journal)
+{
+ struct super_block *sb = (struct super_block *)(journal->j_private);
+ struct ext4_sb_info *sbi = EXT4_SB(sb);
+ struct ext4_inode_info *iter;
+ struct list_head *pos;
+ struct inode *inode;
+ int ret = 0, nblks = 0;
+
+ ret = submit_all_inode_data(journal);
+ if (ret < 0)
+ return ret;
+
+ if (!list_empty(&EXT4_SB(sb)->s_fc_dentry_q)) {
+ ret = fc_commit_dentry_updates(
+ journal, list_last_entry(
+ &EXT4_SB(sb)->s_fc_dentry_q,
+ struct ext4_fc_dentry_update,
+ fcd_list));
+ if (ret < 0)
+ return ret;
+ nblks = ret;
+ }
+
+ spin_lock(&sbi->s_fc_lock);
+ list_for_each(pos, &sbi->s_fc_q) {
+ iter = list_entry(pos, struct ext4_inode_info, i_fc_list);
+ inode = &iter->vfs_inode;
+ if (ext4_test_inode_state(
+ inode, EXT4_STATE_FC_MDATA_SUBMIT) ||
+ !ext4_test_inode_state(
+ inode, EXT4_STATE_FC_DATA_SUBMIT))
+ continue;
+
+ spin_unlock(&sbi->s_fc_lock);
+ ret = fc_commit_data_inode(journal, inode);
+ if (ret < 0)
+ return ret;
+ nblks += ret;
+ spin_lock(&sbi->s_fc_lock);
+ }
+ spin_unlock(&sbi->s_fc_lock);
+
+ ret = wait_all_inode_data(journal);
+ if (ret < 0)
+ return ret;
+
+ return nblks;
+}
+int ext4_fc_async_commit_inode(journal_t *journal, tid_t commit_tid,
+ struct inode *inode)
+{
+ struct ext4_inode_info *ei = EXT4_I(inode);
+ struct super_block *sb = inode->i_sb;
+ struct ext4_sb_info *sbi = EXT4_SB(sb);
+ int nblks = 0, ret;
+ int start_jiffies;
+
+ trace_ext4_journal_fc_commit_cb_start(sb);
+ start_jiffies = jiffies;
+
+ if (!ext4_should_fast_commit(sb) ||
+ (sbi->s_mount_state & EXT4_FC_INELIGIBLE)) {
+ sbi->s_fc_stats.fc_ineligible_commits++;
+ trace_ext4_journal_fc_commit_cb_stop(sb, 0, "disabled");
+ trace_ext4_journal_fc_stats(sb);
+ return jbd2_complete_transaction(journal, commit_tid);
+ }
+
+ if (ext4_is_inode_fc_ineligible(inode)) {
+ sbi->s_fc_stats.fc_ineligible_commits++;
+ trace_ext4_journal_fc_commit_cb_stop(sb, 0, "ineligible");
+ trace_ext4_journal_fc_stats(sb);
+ return jbd2_complete_transaction(journal, commit_tid);
+ }
+
+ /*
+ * In case of soft consistency mode, we wait for any parallel
+ * fast commits to complete. In case of hard consistency, if a
+ * parallel fast commit is ongoing, it is going to take care
+ * of us as well, so we don't wait.
+ */
+ if (!test_opt2(sb, JOURNAL_FC_SOFT_CONSISTENCY))
+ ret = jbd2_start_async_fc_nowait(journal, commit_tid);
+ else
+ ret = jbd2_start_async_fc_wait(journal, commit_tid);
+ if (ret == -EALREADY) {
+ trace_ext4_journal_fc_commit_cb_stop(sb, 0, "already");
+ trace_ext4_journal_fc_stats(sb);
+ return 0;
+ }
+
+ if (ret) {
+ sbi->s_fc_stats.fc_ineligible_commits++;
+ trace_ext4_journal_fc_commit_cb_stop(sb, 0, "start");
+ trace_ext4_journal_fc_stats(sb);
+ return jbd2_complete_transaction(journal, commit_tid);
+ }
+
+ if (ext4_test_inode_state(inode, EXT4_STATE_FC_MDATA_SUBMIT)) {
+ jbd2_stop_async_fc(journal, commit_tid);
+ trace_ext4_journal_fc_commit_cb_stop(sb, 0, "committed");
+ trace_ext4_journal_fc_stats(sb);
+ return 0;
+ }
+
+ if (ei->i_fc_tid != commit_tid) {
+ jbd2_stop_async_fc(journal, commit_tid);
+ trace_ext4_journal_fc_commit_cb_stop(sb, 0, "stale");
+ trace_ext4_journal_fc_stats(sb);
+ return 0;
+ }
+
+ if (!test_opt2(sb, JOURNAL_FC_SOFT_CONSISTENCY)) {
+ ret = ext4_fc_perform_hard_commit(journal);
+ nblks = ret;
+ } else if (ei->i_fc_mdata_update) {
+ ret = submit_all_inode_data(journal);
+ if (ret < 0)
+ goto out;
+ nblks = fc_commit_dentry_updates(journal,
+ ei->i_fc_mdata_update);
+ if (nblks < 0) {
+ ret = nblks;
+ goto out;
+ }
+ ret = wait_all_inode_data(journal);
+ } else if (!list_empty(&EXT4_I(inode)->i_fc_list)) {
+ ext4_set_inode_state(inode, EXT4_STATE_FC_DATA_SUBMIT);
+ ret = jbd2_submit_inode_data(journal, EXT4_I(inode)->jinode);
+ if (ret < 0)
+ goto out;
+ nblks = fc_commit_data_inode(journal, inode);
+ if (nblks < 0) {
+ ret = nblks;
+ goto out;
+ }
+ ext4_set_inode_state(inode, EXT4_STATE_FC_MDATA_SUBMIT);
+ ret = jbd2_wait_inode_data(journal, EXT4_I(inode)->jinode);
+ spin_lock(&sbi->s_fc_lock);
+
+ list_del_init(&EXT4_I(inode)->i_fc_list);
+ ext4_clear_inode_state(inode, EXT4_STATE_FC_DATA_SUBMIT);
+ smp_mb(); /* Make sure data submit bit is set */
+#if (BITS_PER_LONG < 64)
+ wake_up_bit(&EXT4_I(inode)->i_state_flags,
+ EXT4_STATE_FC_DATA_SUBMIT);
+#else
+ wake_up_bit(&EXT4_I(inode)->i_flags,
+ EXT4_STATE_FC_DATA_SUBMIT);
+#endif
+ spin_unlock(&sbi->s_fc_lock);
+ }
+
+out:
+ if (ret < 0) {
+ sbi->s_fc_stats.fc_ineligible_commits++;
+ trace_ext4_journal_fc_commit_cb_stop(sb, 0, "fail1");
+ jbd2_stop_async_fc(journal, commit_tid);
+ trace_ext4_journal_fc_stats(sb);
+ sbi->s_mount_state &= ~EXT4_FC_REPLAY;
+ return jbd2_complete_transaction(journal, commit_tid);
+ }
+ jbd2_wait_on_fc_bufs(journal, nblks);
+ jbd2_stop_async_fc(journal, commit_tid);
+
+ EXT4_SB(sb)->s_fc_stats.fc_num_commits++;
+ EXT4_SB(sb)->s_fc_stats.fc_numblks += nblks;
+ trace_ext4_journal_fc_commit_cb_stop(sb,
+ nblks < 0 ? 0 : nblks,
+ nblks >= 0 ? "success" : "fail2");
+ trace_ext4_journal_fc_stats(sb);
+ sbi->s_mount_state &= ~EXT4_FC_REPLAY;
+ return 0;
+}
+
void ext4_init_fast_commit(struct super_block *sb, journal_t *journal)
{
if (!ext4_should_fast_commit(sb))
return;
+ journal->j_fc_cleanup_callback = ext4_journal_fc_cleanup_cb;
jbd2_init_fast_commit(journal, EXT4_NUM_FC_BLKS);
}
diff --git a/fs/ext4/ext4_jbd2.h b/fs/ext4/ext4_jbd2.h
index d3a951d2e58d..4b6c179c02eb 100644
--- a/fs/ext4/ext4_jbd2.h
+++ b/fs/ext4/ext4_jbd2.h
@@ -527,5 +527,7 @@ int __init ext4_init_fc_dentry_cache(void);
void ext4_fc_track_inode(struct inode *inode);
void ext4_fc_mark_ineligible(struct inode *inode, int reason);
void ext4_fc_disable(struct super_block *sb, int reason);
+void ext4_fc_del(struct inode *inode);
+
#endif /* _EXT4_JBD2_H */
diff --git a/fs/ext4/fsync.c b/fs/ext4/fsync.c
index 5508baa11bb6..4be3724d4f84 100644
--- a/fs/ext4/fsync.c
+++ b/fs/ext4/fsync.c
@@ -151,7 +151,7 @@ int ext4_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
if (journal->j_flags & JBD2_BARRIER &&
!jbd2_trans_will_send_data_barrier(journal, commit_tid))
needs_barrier = true;
- ret = jbd2_complete_transaction(journal, commit_tid);
+ ret = ext4_fc_async_commit_inode(journal, commit_tid, inode);
if (needs_barrier) {
issue_flush:
err = blkdev_issue_flush(inode->i_sb->s_bdev, GFP_KERNEL, NULL);
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 2a61c3a31b74..c4cde431d5fa 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -5416,8 +5416,10 @@ int ext4_write_inode(struct inode *inode, struct writeback_control *wbc)
if (wbc->sync_mode != WB_SYNC_ALL || wbc->for_sync)
return 0;
- err = jbd2_complete_transaction(EXT4_SB(inode->i_sb)->s_journal,
- EXT4_I(inode)->i_sync_tid);
+ err = ext4_fc_async_commit_inode(EXT4_SB(inode->i_sb)
+ ->s_journal,
+ EXT4_I(inode)->i_sync_tid,
+ inode);
} else {
struct ext4_iloc iloc;
diff --git a/fs/ext4/migrate.c b/fs/ext4/migrate.c
index b1e4d359f73b..b995690d73ce 100644
--- a/fs/ext4/migrate.c
+++ b/fs/ext4/migrate.c
@@ -513,6 +513,7 @@ int ext4_ext_migrate(struct inode *inode)
* work to orphan_list_cleanup()
*/
ext4_orphan_del(NULL, tmp_inode);
+ ext4_fc_del(inode);
retval = PTR_ERR(handle);
goto out;
}
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index 23a194d871c6..bfd19a127188 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -1111,6 +1111,11 @@ static int ext4_drop_inode(struct inode *inode)
if (!drop)
drop = fscrypt_drop_inode(inode);
+ if (drop) {
+ spin_unlock(&inode->i_lock);
+ ext4_fc_del(inode);
+ spin_lock(&inode->i_lock);
+ }
trace_ext4_drop_inode(inode, drop);
return drop;
@@ -1119,6 +1124,11 @@ static int ext4_drop_inode(struct inode *inode)
static void ext4_free_in_core_inode(struct inode *inode)
{
fscrypt_free_inode(inode);
+ if (!list_empty(&(EXT4_I(inode)->i_fc_list))) {
+ pr_warn("%s: inode %ld still in fc list",
+ __func__, inode->i_ino);
+ ext4_fc_del(inode);
+ }
kmem_cache_free(ext4_inode_cachep, EXT4_I(inode));
}
@@ -1133,6 +1143,8 @@ static void ext4_destroy_inode(struct inode *inode)
true);
dump_stack();
}
+ if (!list_empty(&(EXT4_I(inode)->i_fc_list)))
+ ext4_fc_del(inode);
}
static void init_once(void *foo)
diff --git a/include/trace/events/ext4.h b/include/trace/events/ext4.h
index 0808b62ac108..e47059a02fec 100644
--- a/include/trace/events/ext4.h
+++ b/include/trace/events/ext4.h
@@ -75,6 +75,19 @@ struct partial_cluster;
{ FALLOC_FL_COLLAPSE_RANGE, "COLLAPSE_RANGE"}, \
{ FALLOC_FL_ZERO_RANGE, "ZERO_RANGE"})
+#define show_fc_reason(reason) \
+ __print_symbolic(reason, \
+ { EXT4_FC_REASON_META_ALLOC, "META_ALLOC"}, \
+ { EXT4_FC_REASON_QUOTA, "QUOTA"}, \
+ { EXT4_FC_REASON_XATTR, "XATTR"}, \
+ { EXT4_FC_REASON_CROSS_RENAME, "CROSS_RENAME"}, \
+ { EXT4_FC_REASON_FALLOC_RANGE_OP, "FALLOC_RANGE_OP"}, \
+ { EXT4_FC_REASON_JOURNAL_FLAG_CHANGE, "JOURNAL_FLAG_CHANGE"}, \
+ { EXT4_FC_REASON_MEM, "NO_MEM"}, \
+ { EXT4_FC_REASON_SWAP_BOOT, "SWAP_BOOT"}, \
+ { EXT4_FC_REASON_RESIZE, "RESIZE"}, \
+ { EXT4_FC_REASON_RENAME_DIR, "FALLOC_RANGE_OP"})
+
TRACE_EVENT(ext4_other_inode_update_time,
TP_PROTO(struct inode *inode, ino_t orig_ino),
@@ -2703,6 +2716,94 @@ TRACE_EVENT(ext4_error,
__entry->function, __entry->line)
);
+TRACE_EVENT(ext4_journal_fc_commit_cb_start,
+ TP_PROTO(struct super_block *sb),
+
+ TP_ARGS(sb),
+
+ TP_STRUCT__entry(
+ __field(dev_t, dev)
+ ),
+
+ TP_fast_assign(
+ __entry->dev = sb->s_dev;
+ ),
+
+ TP_printk("fast_commit started on dev %d,%d",
+ MAJOR(__entry->dev), MINOR(__entry->dev))
+);
+
+TRACE_EVENT(ext4_journal_fc_commit_cb_stop,
+ TP_PROTO(struct super_block *sb, int nblks, const char *reason),
+
+ TP_ARGS(sb, nblks, reason),
+
+ TP_STRUCT__entry(
+ __field(dev_t, dev)
+ __field(int, nblks)
+ __field(const char *, reason)
+ __field(int, num_fc)
+ __field(int, num_fc_ineligible)
+ __field(int, nblks_agg)
+ ),
+
+ TP_fast_assign(
+ __entry->dev = sb->s_dev;
+ __entry->nblks = nblks;
+ __entry->reason = reason;
+ __entry->num_fc = EXT4_SB(sb)->s_fc_stats.fc_num_commits;
+ __entry->num_fc_ineligible =
+ EXT4_SB(sb)->s_fc_stats.fc_ineligible_commits;
+ __entry->nblks_agg = EXT4_SB(sb)->s_fc_stats.fc_numblks;
+ ),
+
+ TP_printk("fc on [%d,%d] nblks %d, reason %s, fc = %d, ineligible = %d, agg_nblks %d",
+ MAJOR(__entry->dev), MINOR(__entry->dev),
+ __entry->nblks, __entry->reason, __entry->num_fc,
+ __entry->num_fc_ineligible, __entry->nblks_agg)
+);
+
+#define FC_REASON_NAME_STAT(reason) \
+ show_fc_reason(reason), \
+ __entry->sbi->s_fc_stats.fc_ineligible_reason_count[reason]
+
+TRACE_EVENT(ext4_journal_fc_stats,
+ TP_PROTO(struct super_block *sb),
+
+ TP_ARGS(sb),
+
+ TP_STRUCT__entry(
+ __field(dev_t, dev)
+ __field(struct ext4_sb_info *, sbi)
+ __field(int, count)
+ ),
+
+ TP_fast_assign(
+ __entry->dev = sb->s_dev;
+ __entry->sbi = EXT4_SB(sb);
+ ),
+
+ TP_printk("dev %d:%d fc ineligible reasons:\n"
+ "%s:%d, %s:%d, %s:%d, %s:%d, %s:%d, %s:%d, "
+ "%s:%d, %s:%d, %s:%d, %s:%d; "
+ "num_commits:%d, ineligible: %d, numblks: %d",
+ MAJOR(__entry->dev), MINOR(__entry->dev),
+ FC_REASON_NAME_STAT(EXT4_FC_REASON_META_ALLOC),
+ FC_REASON_NAME_STAT(EXT4_FC_REASON_QUOTA),
+ FC_REASON_NAME_STAT(EXT4_FC_REASON_XATTR),
+ FC_REASON_NAME_STAT(EXT4_FC_REASON_CROSS_RENAME),
+ FC_REASON_NAME_STAT(EXT4_FC_REASON_FALLOC_RANGE_OP),
+ FC_REASON_NAME_STAT(EXT4_FC_REASON_JOURNAL_FLAG_CHANGE),
+ FC_REASON_NAME_STAT(EXT4_FC_REASON_MEM),
+ FC_REASON_NAME_STAT(EXT4_FC_REASON_SWAP_BOOT),
+ FC_REASON_NAME_STAT(EXT4_FC_REASON_RESIZE),
+ FC_REASON_NAME_STAT(EXT4_FC_REASON_RENAME_DIR),
+ __entry->sbi->s_fc_stats.fc_num_commits,
+ __entry->sbi->s_fc_stats.fc_ineligible_commits,
+ __entry->sbi->s_fc_stats.fc_numblks)
+
+);
+
#define DEFINE_TRACE_DENTRY_EVENT(__type) \
TRACE_EVENT(ext4_fc_track_##__type, \
TP_PROTO(struct inode *inode, struct dentry *dentry, int ret), \
--
2.24.1.735.g03f4e72817-goog
^ permalink raw reply related [flat|nested] 31+ messages in thread* Re: [PATCH v4 14/20] ext4: main commit routine for fast commits
2019-12-24 8:13 ` [PATCH v4 14/20] ext4: main commit routine for " Harshad Shirwadkar
@ 2019-12-27 16:36 ` kbuild test robot
2019-12-27 16:36 ` [RFC PATCH] ext4: submit_fc_bh() can be static kbuild test robot
1 sibling, 0 replies; 31+ messages in thread
From: kbuild test robot @ 2019-12-27 16:36 UTC (permalink / raw)
To: Harshad Shirwadkar; +Cc: kbuild-all, linux-ext4, Harshad Shirwadkar
Hi Harshad,
Thank you for the patch! Perhaps something to improve:
[auto build test WARNING on tip/perf/core]
[cannot apply to ext4/dev linus/master v5.5-rc3 next-20191220]
[if your patch is applied to the wrong git tree, please drop us a note to help
improve the system. BTW, we also suggest to use '--base' option to specify the
base tree in git format-patch, please see https://stackoverflow.com/a/37406982]
url: https://github.com/0day-ci/linux/commits/Harshad-Shirwadkar/ext4-update-docs-for-fast-commit-feature/20191225-200339
base: https://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git ceb9e77324fa661b1001a0ae66f061b5fcb4e4e6
reproduce:
# apt-get install sparse
# sparse version: v0.6.1-129-g341daf20-dirty
make ARCH=x86_64 allmodconfig
make C=1 CF='-fdiagnostic-prefix -D__CHECK_ENDIAN__'
If you fix the issue, kindly add following tag
Reported-by: kbuild test robot <lkp@intel.com>
sparse warnings: (new ones prefixed by >>)
fs/ext4/ext4_jbd2.c:611:5: sparse: sparse: symbol '__ext4_fc_track_range' was not declared. Should it be static?
>> fs/ext4/ext4_jbd2.c:825:6: sparse: sparse: symbol 'submit_fc_bh' was not declared. Should it be static?
>> fs/ext4/ext4_jbd2.c:1156:5: sparse: sparse: symbol 'ext4_fc_perform_hard_commit' was not declared. Should it be static?
fs/ext4/ext4_jbd2.c:500:12: sparse: sparse: context imbalance in '__ext4_dentry_update' - unexpected unlock
>> fs/ext4/ext4_jbd2.c:923:20: sparse: sparse: context imbalance in 'wait_all_inode_data' - unexpected unlock
Please review and possibly fold the followup patch.
---
0-DAY kernel test infrastructure Open Source Technology Center
https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org Intel Corporation
^ permalink raw reply [flat|nested] 31+ messages in thread* [RFC PATCH] ext4: submit_fc_bh() can be static
2019-12-24 8:13 ` [PATCH v4 14/20] ext4: main commit routine for " Harshad Shirwadkar
2019-12-27 16:36 ` kbuild test robot
@ 2019-12-27 16:36 ` kbuild test robot
1 sibling, 0 replies; 31+ messages in thread
From: kbuild test robot @ 2019-12-27 16:36 UTC (permalink / raw)
To: Harshad Shirwadkar; +Cc: kbuild-all, linux-ext4, Harshad Shirwadkar
Fixes: 6ae30577ea12 ("ext4: main commit routine for fast commits")
Signed-off-by: kbuild test robot <lkp@intel.com>
---
ext4_jbd2.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/fs/ext4/ext4_jbd2.c b/fs/ext4/ext4_jbd2.c
index f371f1f0f9148..05d5bfa0f355e 100644
--- a/fs/ext4/ext4_jbd2.c
+++ b/fs/ext4/ext4_jbd2.c
@@ -822,7 +822,7 @@ static int fc_write_data(struct inode *inode, u8 *start, u8 *end,
return num_tlvs;
}
-void submit_fc_bh(struct buffer_head *bh)
+static void submit_fc_bh(struct buffer_head *bh)
{
lock_buffer(bh);
clear_buffer_dirty(bh);
@@ -1153,7 +1153,7 @@ static void ext4_journal_fc_cleanup_cb(journal_t *journal)
trace_ext4_journal_fc_stats(sb);
}
-int ext4_fc_perform_hard_commit(journal_t *journal)
+static int ext4_fc_perform_hard_commit(journal_t *journal)
{
struct super_block *sb = (struct super_block *)(journal->j_private);
struct ext4_sb_info *sbi = EXT4_SB(sb);
^ permalink raw reply related [flat|nested] 31+ messages in thread
* [PATCH v4 15/20] jbd2: add fast commit recovery path support
2019-12-24 8:13 [PATCH v4 01/20] ext4: update docs for fast commit feature Harshad Shirwadkar
` (12 preceding siblings ...)
2019-12-24 8:13 ` [PATCH v4 14/20] ext4: main commit routine for " Harshad Shirwadkar
@ 2019-12-24 8:13 ` Harshad Shirwadkar
2019-12-24 8:13 ` [PATCH v4 16/20] ext4: fast commit recovery path preparation Harshad Shirwadkar
` (5 subsequent siblings)
19 siblings, 0 replies; 31+ messages in thread
From: Harshad Shirwadkar @ 2019-12-24 8:13 UTC (permalink / raw)
To: linux-ext4; +Cc: Harshad Shirwadkar
Add fc_do_one_pass to invoke file system specific replay
callback and pass discovered fast commit blocks to let
file system handle those.
Signed-off-by: Harshad Shirwadkar <harshadshirwadkar@gmail.com>
---
fs/jbd2/recovery.c | 67 +++++++++++++++++++++++++++++++++++++++++---
include/linux/jbd2.h | 13 +++++++++
2 files changed, 76 insertions(+), 4 deletions(-)
diff --git a/fs/jbd2/recovery.c b/fs/jbd2/recovery.c
index a4967b27ffb6..09f069e59c36 100644
--- a/fs/jbd2/recovery.c
+++ b/fs/jbd2/recovery.c
@@ -35,7 +35,6 @@ struct recovery_info
int nr_revoke_hits;
};
-enum passtype {PASS_SCAN, PASS_REVOKE, PASS_REPLAY};
static int do_one_pass(journal_t *journal,
struct recovery_info *info, enum passtype pass);
static int scan_revoke_records(journal_t *, struct buffer_head *,
@@ -225,10 +224,63 @@ static int count_tags(journal_t *journal, struct buffer_head *bh)
/* Make sure we wrap around the log correctly! */
#define wrap(journal, var) \
do { \
- if (var >= (journal)->j_last) \
- var -= ((journal)->j_last - (journal)->j_first); \
+ unsigned long _wrap_last = \
+ jbd2_has_feature_fast_commit(journal) ? \
+ (journal)->j_last_fc : (journal)->j_last; \
+ \
+ if (var >= _wrap_last) \
+ var -= (_wrap_last - (journal)->j_first); \
} while (0)
+static int fc_do_one_pass(journal_t *journal,
+ struct recovery_info *info, enum passtype pass)
+{
+ unsigned int expected_commit_id = info->end_transaction;
+ unsigned long next_fc_block;
+ struct buffer_head *bh;
+ unsigned int seq;
+ journal_header_t *jhdr;
+ int err = 0;
+
+ next_fc_block = journal->j_first_fc;
+
+ while (next_fc_block <= journal->j_last_fc) {
+ jbd_debug(3, "Fast commit replay: next block %ld",
+ next_fc_block);
+ err = jread(&bh, journal, next_fc_block);
+ if (err) {
+ jbd_debug(3, "Fast commit replay: read error");
+ break;
+ }
+
+ jhdr = (journal_header_t *)bh->b_data;
+ seq = be32_to_cpu(jhdr->h_sequence);
+ if (be32_to_cpu(jhdr->h_magic) != JBD2_MAGIC_NUMBER ||
+ seq != expected_commit_id) {
+ jbd_debug(3, "Fast commit replay: magic / commitid error [%d / %d / %d]\n",
+ be32_to_cpu(jhdr->h_magic), seq,
+ expected_commit_id);
+ break;
+ }
+ jbd_debug(3, "Processing fast commit blk with seq %d",
+ seq);
+ if (journal->j_fc_replay_callback) {
+ err = journal->j_fc_replay_callback(
+ journal, bh, pass,
+ next_fc_block -
+ journal->j_first_fc);
+ if (err)
+ break;
+ }
+ next_fc_block++;
+ }
+
+ if (err)
+ jbd_debug(3, "Fast commit replay failed, err = %d\n", err);
+
+ return err;
+}
+
/**
* jbd2_journal_recover - recovers a on-disk journal
* @journal: the journal to recover
@@ -470,7 +522,7 @@ static int do_one_pass(journal_t *journal,
break;
jbd_debug(2, "Scanning for sequence ID %u at %lu/%lu\n",
- next_commit_ID, next_log_block, journal->j_last);
+ next_commit_ID, next_log_block, journal->j_last_fc);
/* Skip over each chunk of the transaction looking
* either the next descriptor block or the final commit
@@ -768,6 +820,9 @@ static int do_one_pass(journal_t *journal,
if (err)
goto failed;
continue;
+ case JBD2_FC_BLOCK:
+ pr_warn("Unexpectedly found fast commit block.\n");
+ continue;
default:
jbd_debug(3, "Unrecognised magic %d, end of scan.\n",
@@ -799,6 +854,10 @@ static int do_one_pass(journal_t *journal,
success = -EIO;
}
}
+
+ if (jbd2_has_feature_fast_commit(journal) && pass != PASS_REVOKE)
+ success = fc_do_one_pass(journal, info, pass);
+
if (block_error && success == 0)
success = -EIO;
return success;
diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h
index d450dcb93e51..0b49c8ff0563 100644
--- a/include/linux/jbd2.h
+++ b/include/linux/jbd2.h
@@ -757,6 +757,8 @@ jbd2_time_diff(unsigned long start, unsigned long end)
#define JBD2_NR_BATCH 64
+enum passtype {PASS_SCAN, PASS_REVOKE, PASS_REPLAY};
+
/**
* struct journal_s - The journal_s type is the concrete type associated with
* journal_t.
@@ -1220,6 +1222,17 @@ struct journal_s
* after every commit operation.
*/
void (*j_fc_cleanup_callback)(struct journal_s *journal);
+
+ /*
+ * @j_fc_replay_callback:
+ *
+ * File-system specific function that performs replay of a fast
+ * commit. JBD2 calls this function for each fast commit block found in
+ * the journal.
+ */
+ int (*j_fc_replay_callback)(struct journal_s *journal,
+ struct buffer_head *bh,
+ enum passtype pass, int off);
};
#define jbd2_might_wait_for_commit(j) \
--
2.24.1.735.g03f4e72817-goog
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PATCH v4 16/20] ext4: fast commit recovery path preparation
2019-12-24 8:13 [PATCH v4 01/20] ext4: update docs for fast commit feature Harshad Shirwadkar
` (13 preceding siblings ...)
2019-12-24 8:13 ` [PATCH v4 15/20] jbd2: add fast commit recovery path support Harshad Shirwadkar
@ 2019-12-24 8:13 ` Harshad Shirwadkar
2019-12-24 8:13 ` [PATCH v4 17/20] ext4: add idempotent helpers to manipulate bitmaps Harshad Shirwadkar
` (4 subsequent siblings)
19 siblings, 0 replies; 31+ messages in thread
From: Harshad Shirwadkar @ 2019-12-24 8:13 UTC (permalink / raw)
To: linux-ext4; +Cc: Harshad Shirwadkar
Prepare for making ext4 fast commit recovery path changes. Make a few
existing functions visible. Break and add a wrapper around
ext4_get_inode_loc to allow reading inode from disk without having
a corresponding VFS inode.
Signed-off-by: Harshad Shirwadkar <harshadshirwadkar@gmail.com>
---
fs/ext4/ext4.h | 7 +++++++
fs/ext4/inode.c | 32 ++++++++++++++++++--------------
fs/ext4/ioctl.c | 6 +++---
fs/ext4/namei.c | 2 +-
include/trace/events/ext4.h | 8 ++++----
5 files changed, 33 insertions(+), 22 deletions(-)
diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index 9c2f67c64b4f..f2603deefe51 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -2651,6 +2651,8 @@ extern int ext4_trim_fs(struct super_block *, struct fstrim_range *);
extern void ext4_process_freed_data(struct super_block *sb, tid_t commit_tid);
/* inode.c */
+void ext4_inode_csum_set(struct inode *inode, struct ext4_inode *raw,
+ struct ext4_inode_info *ei);
int ext4_inode_is_fast_symlink(struct inode *inode);
struct buffer_head *ext4_getblk(handle_t *, struct inode *, ext4_lblk_t, int);
struct buffer_head *ext4_bread(handle_t *, struct inode *, ext4_lblk_t, int);
@@ -2699,6 +2701,8 @@ extern int ext4_sync_inode(handle_t *, struct inode *);
extern void ext4_dirty_inode(struct inode *, int);
extern int ext4_change_inode_journal_flag(struct inode *, int);
extern int ext4_get_inode_loc(struct inode *, struct ext4_iloc *);
+extern int ext4_get_fc_inode_loc(struct super_block *sb, unsigned long ino,
+ struct ext4_iloc *iloc);
extern int ext4_inode_attach_jinode(struct inode *inode);
extern int ext4_can_truncate(struct inode *inode);
extern int ext4_truncate(struct inode *);
@@ -2734,12 +2738,15 @@ extern int ext4_ind_remove_space(handle_t *handle, struct inode *inode,
/* ioctl.c */
extern long ext4_ioctl(struct file *, unsigned int, unsigned long);
extern long ext4_compat_ioctl(struct file *, unsigned int, unsigned long);
+extern void ext4_reset_inode_seed(struct inode *inode);
/* migrate.c */
extern int ext4_ext_migrate(struct inode *);
extern int ext4_ind_migrate(struct inode *inode);
/* namei.c */
+extern int ext4_init_new_dir(handle_t *handle, struct inode *dir,
+ struct inode *inode);
extern int ext4_dirblock_csum_verify(struct inode *inode,
struct buffer_head *bh);
extern int ext4_orphan_add(handle_t *, struct inode *);
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index c4cde431d5fa..e902000dac51 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -103,8 +103,8 @@ static int ext4_inode_csum_verify(struct inode *inode, struct ext4_inode *raw,
return provided == calculated;
}
-static void ext4_inode_csum_set(struct inode *inode, struct ext4_inode *raw,
- struct ext4_inode_info *ei)
+void ext4_inode_csum_set(struct inode *inode, struct ext4_inode *raw,
+ struct ext4_inode_info *ei)
{
__u32 csum;
@@ -4548,22 +4548,21 @@ int ext4_truncate(struct inode *inode)
* data in memory that is needed to recreate the on-disk version of this
* inode.
*/
-static int __ext4_get_inode_loc(struct inode *inode,
+static int __ext4_get_inode_loc(struct super_block *sb, unsigned long ino,
struct ext4_iloc *iloc, int in_mem)
{
struct ext4_group_desc *gdp;
struct buffer_head *bh;
- struct super_block *sb = inode->i_sb;
ext4_fsblk_t block;
struct blk_plug plug;
int inodes_per_block, inode_offset;
iloc->bh = NULL;
- if (inode->i_ino < EXT4_ROOT_INO ||
- inode->i_ino > le32_to_cpu(EXT4_SB(sb)->s_es->s_inodes_count))
+ if (ino < EXT4_ROOT_INO ||
+ ino > le32_to_cpu(EXT4_SB(sb)->s_es->s_inodes_count))
return -EFSCORRUPTED;
- iloc->block_group = (inode->i_ino - 1) / EXT4_INODES_PER_GROUP(sb);
+ iloc->block_group = (ino - 1) / EXT4_INODES_PER_GROUP(sb);
gdp = ext4_get_group_desc(sb, iloc->block_group, NULL);
if (!gdp)
return -EIO;
@@ -4572,7 +4571,7 @@ static int __ext4_get_inode_loc(struct inode *inode,
* Figure out the offset within the block group inode table
*/
inodes_per_block = EXT4_SB(sb)->s_inodes_per_block;
- inode_offset = ((inode->i_ino - 1) %
+ inode_offset = ((ino - 1) %
EXT4_INODES_PER_GROUP(sb));
block = ext4_inode_table(sb, gdp) + (inode_offset / inodes_per_block);
iloc->offset = (inode_offset % inodes_per_block) * EXT4_INODE_SIZE(sb);
@@ -4671,15 +4670,14 @@ static int __ext4_get_inode_loc(struct inode *inode,
* has in-inode xattrs, or we don't have this inode in memory.
* Read the block from disk.
*/
- trace_ext4_load_inode(inode);
+ trace_ext4_load_inode(sb, ino);
get_bh(bh);
bh->b_end_io = end_buffer_read_sync;
submit_bh(REQ_OP_READ, REQ_META | REQ_PRIO, bh);
blk_finish_plug(&plug);
wait_on_buffer(bh);
if (!buffer_uptodate(bh)) {
- EXT4_ERROR_INODE_BLOCK(inode, block,
- "unable to read itable block");
+ ext4_error(sb, "unable to read itable block");
brelse(bh);
return -EIO;
}
@@ -4692,10 +4690,16 @@ static int __ext4_get_inode_loc(struct inode *inode,
int ext4_get_inode_loc(struct inode *inode, struct ext4_iloc *iloc)
{
/* We have all inode data except xattrs in memory here. */
- return __ext4_get_inode_loc(inode, iloc,
+ return __ext4_get_inode_loc(inode->i_sb, inode->i_ino, iloc,
!ext4_test_inode_state(inode, EXT4_STATE_XATTR));
}
+int ext4_get_fc_inode_loc(struct super_block *sb, unsigned long ino,
+ struct ext4_iloc *iloc)
+{
+ return __ext4_get_inode_loc(sb, ino, iloc, 0);
+}
+
static bool ext4_should_use_dax(struct inode *inode)
{
if (!test_opt(inode->i_sb, DAX))
@@ -4845,7 +4849,7 @@ struct inode *__ext4_iget(struct super_block *sb, unsigned long ino,
ei = EXT4_I(inode);
iloc.bh = NULL;
- ret = __ext4_get_inode_loc(inode, &iloc, 0);
+ ret = __ext4_get_inode_loc(sb, inode->i_ino, &iloc, 0);
if (ret < 0)
goto bad_inode;
raw_inode = ext4_raw_inode(&iloc);
@@ -5423,7 +5427,7 @@ int ext4_write_inode(struct inode *inode, struct writeback_control *wbc)
} else {
struct ext4_iloc iloc;
- err = __ext4_get_inode_loc(inode, &iloc, 0);
+ err = __ext4_get_inode_loc(inode->i_sb, inode->i_ino, &iloc, 0);
if (err)
return err;
/*
diff --git a/fs/ext4/ioctl.c b/fs/ext4/ioctl.c
index 2bc655b2164e..59ff5f90ed2a 100644
--- a/fs/ext4/ioctl.c
+++ b/fs/ext4/ioctl.c
@@ -86,7 +86,7 @@ static void swap_inode_data(struct inode *inode1, struct inode *inode2)
i_size_write(inode2, isize);
}
-static void reset_inode_seed(struct inode *inode)
+void ext4_reset_inode_seed(struct inode *inode)
{
struct ext4_inode_info *ei = EXT4_I(inode);
struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
@@ -199,8 +199,8 @@ static long swap_inode_boot_loader(struct super_block *sb,
inode->i_generation = prandom_u32();
inode_bl->i_generation = prandom_u32();
- reset_inode_seed(inode);
- reset_inode_seed(inode_bl);
+ ext4_reset_inode_seed(inode);
+ ext4_reset_inode_seed(inode_bl);
ext4_discard_preallocations(inode);
diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c
index b732c0bb1d51..48fea5ce8530 100644
--- a/fs/ext4/namei.c
+++ b/fs/ext4/namei.c
@@ -2710,7 +2710,7 @@ struct ext4_dir_entry_2 *ext4_init_dot_dotdot(struct inode *inode,
return ext4_next_entry(de, blocksize);
}
-static int ext4_init_new_dir(handle_t *handle, struct inode *dir,
+int ext4_init_new_dir(handle_t *handle, struct inode *dir,
struct inode *inode)
{
struct buffer_head *dir_block = NULL;
diff --git a/include/trace/events/ext4.h b/include/trace/events/ext4.h
index e47059a02fec..8da371b38332 100644
--- a/include/trace/events/ext4.h
+++ b/include/trace/events/ext4.h
@@ -1738,9 +1738,9 @@ TRACE_EVENT(ext4_ext_load_extent,
);
TRACE_EVENT(ext4_load_inode,
- TP_PROTO(struct inode *inode),
+ TP_PROTO(struct super_block *sb, unsigned long ino),
- TP_ARGS(inode),
+ TP_ARGS(sb, ino),
TP_STRUCT__entry(
__field( dev_t, dev )
@@ -1748,8 +1748,8 @@ TRACE_EVENT(ext4_load_inode,
),
TP_fast_assign(
- __entry->dev = inode->i_sb->s_dev;
- __entry->ino = inode->i_ino;
+ __entry->dev = sb->s_dev;
+ __entry->ino = ino;
),
TP_printk("dev %d,%d ino %ld",
--
2.24.1.735.g03f4e72817-goog
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PATCH v4 17/20] ext4: add idempotent helpers to manipulate bitmaps
2019-12-24 8:13 [PATCH v4 01/20] ext4: update docs for fast commit feature Harshad Shirwadkar
` (14 preceding siblings ...)
2019-12-24 8:13 ` [PATCH v4 16/20] ext4: fast commit recovery path preparation Harshad Shirwadkar
@ 2019-12-24 8:13 ` Harshad Shirwadkar
2019-12-27 19:36 ` kbuild test robot
2019-12-27 19:36 ` [RFC PATCH] ext4: ext4_free_blocks_simple() can be static kbuild test robot
2019-12-24 8:13 ` [PATCH v4 18/20] ext4: disable certain features in replay path Harshad Shirwadkar
` (3 subsequent siblings)
19 siblings, 2 replies; 31+ messages in thread
From: Harshad Shirwadkar @ 2019-12-24 8:13 UTC (permalink / raw)
To: linux-ext4; +Cc: Harshad Shirwadkar
For fast commit replay path, we need idempotent helpers that mark
inodes used, data blocks as used or free. It's important these are
idempotent and that's because we can crash while we are replaying.
Signed-off-by: Harshad Shirwadkar <harshadshirwadkar@gmail.com>
---
fs/ext4/ext4.h | 1 +
fs/ext4/ialloc.c | 113 ++++++++++++++++++++++++++++++++++++++
fs/ext4/mballoc.c | 135 +++++++++++++++++++++++++++++++++++++++++++++-
fs/ext4/mballoc.h | 2 +
4 files changed, 250 insertions(+), 1 deletion(-)
diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index f2603deefe51..1957f30a7f2e 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -2601,6 +2601,7 @@ extern int ext4fs_dirhash(const struct inode *dir, const char *name, int len,
struct dx_hash_info *hinfo);
/* ialloc.c */
+extern int ext4_mark_inode_used(struct super_block *sb, int ino);
extern struct inode *__ext4_new_inode(handle_t *, struct inode *, umode_t,
const struct qstr *qstr, __u32 goal,
uid_t *owner, __u32 i_flags,
diff --git a/fs/ext4/ialloc.c b/fs/ext4/ialloc.c
index 764ff4c56233..f1a1432f9ffa 100644
--- a/fs/ext4/ialloc.c
+++ b/fs/ext4/ialloc.c
@@ -729,6 +729,119 @@ static int find_inode_bit(struct super_block *sb, ext4_group_t group,
return 1;
}
+int ext4_mark_inode_used(struct super_block *sb, int ino)
+{
+ unsigned long max_ino = le32_to_cpu(EXT4_SB(sb)->s_es->s_inodes_count);
+ struct buffer_head *inode_bitmap_bh = NULL, *group_desc_bh = NULL;
+ struct ext4_group_desc *gdp;
+ ext4_group_t group;
+ int bit;
+ int err = -EFSCORRUPTED;
+
+ if (ino < EXT4_FIRST_INO(sb) || ino > max_ino)
+ goto out;
+
+ group = (ino - 1) / EXT4_INODES_PER_GROUP(sb);
+ bit = (ino - 1) % EXT4_INODES_PER_GROUP(sb);
+ inode_bitmap_bh = ext4_read_inode_bitmap(sb, group);
+ if (IS_ERR(inode_bitmap_bh))
+ return PTR_ERR(inode_bitmap_bh);
+
+ if (ext4_test_bit(bit, inode_bitmap_bh->b_data)) {
+ err = -EEXIST;
+ goto out;
+ }
+
+ gdp = ext4_get_group_desc(sb, group, &group_desc_bh);
+ if (!gdp || !group_desc_bh) {
+ err = -EINVAL;
+ goto out;
+ }
+
+ ext4_set_bit(bit, inode_bitmap_bh->b_data);
+
+ BUFFER_TRACE(inode_bitmap_bh, "call ext4_handle_dirty_metadata");
+ err = ext4_handle_dirty_metadata(NULL, NULL, inode_bitmap_bh);
+ if (err) {
+ ext4_std_error(sb, err);
+ goto out;
+ }
+ sync_dirty_buffer(inode_bitmap_bh);
+ BUFFER_TRACE(group_desc_bh, "get_write_access");
+
+ /* We may have to initialize the block bitmap if it isn't already */
+ if (ext4_has_group_desc_csum(sb) &&
+ gdp->bg_flags & cpu_to_le16(EXT4_BG_BLOCK_UNINIT)) {
+ struct buffer_head *block_bitmap_bh;
+
+ block_bitmap_bh = ext4_read_block_bitmap(sb, group);
+ if (IS_ERR(block_bitmap_bh)) {
+ err = PTR_ERR(block_bitmap_bh);
+ goto out;
+ }
+
+ BUFFER_TRACE(block_bitmap_bh, "dirty block bitmap");
+ err = ext4_handle_dirty_metadata(NULL, NULL, block_bitmap_bh);
+ sync_dirty_buffer(block_bitmap_bh);
+
+ /* recheck and clear flag under lock if we still need to */
+ ext4_lock_group(sb, group);
+ if (ext4_has_group_desc_csum(sb) &&
+ (gdp->bg_flags & cpu_to_le16(EXT4_BG_BLOCK_UNINIT))) {
+ gdp->bg_flags &= cpu_to_le16(~EXT4_BG_BLOCK_UNINIT);
+ ext4_free_group_clusters_set(sb, gdp,
+ ext4_free_clusters_after_init(sb, group, gdp));
+ ext4_block_bitmap_csum_set(sb, group, gdp,
+ block_bitmap_bh);
+ ext4_group_desc_csum_set(sb, group, gdp);
+ }
+ ext4_unlock_group(sb, group);
+ brelse(block_bitmap_bh);
+
+ if (err) {
+ ext4_std_error(sb, err);
+ goto out;
+ }
+ }
+
+ /* Update the relevant bg descriptor fields */
+ if (ext4_has_group_desc_csum(sb)) {
+ int free;
+
+ ext4_lock_group(sb, group); /* while we modify the bg desc */
+ free = EXT4_INODES_PER_GROUP(sb) -
+ ext4_itable_unused_count(sb, gdp);
+ if (gdp->bg_flags & cpu_to_le16(EXT4_BG_INODE_UNINIT)) {
+ gdp->bg_flags &= cpu_to_le16(~EXT4_BG_INODE_UNINIT);
+ free = 0;
+ }
+
+ /*
+ * Check the relative inode number against the last used
+ * relative inode number in this group. if it is greater
+ * we need to update the bg_itable_unused count
+ */
+ if (bit >= free)
+ ext4_itable_unused_set(sb, gdp,
+ (EXT4_INODES_PER_GROUP(sb) - bit - 1));
+ } else {
+ ext4_lock_group(sb, group);
+ }
+
+ ext4_free_inodes_set(sb, gdp, ext4_free_inodes_count(sb, gdp) - 1);
+ if (ext4_has_group_desc_csum(sb)) {
+ ext4_inode_bitmap_csum_set(sb, group, gdp, inode_bitmap_bh,
+ EXT4_INODES_PER_GROUP(sb) / 8);
+ ext4_group_desc_csum_set(sb, group, gdp);
+ }
+
+ ext4_unlock_group(sb, group);
+ err = ext4_handle_dirty_metadata(NULL, NULL, group_desc_bh);
+ sync_dirty_buffer(group_desc_bh);
+out:
+ return err;
+}
+
/*
* There are two policies for allocating an inode. If the new inode is
* a directory, then a forward search is made for a block group with both
diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
index a3e2767bdf2f..05ca9001f8fa 100644
--- a/fs/ext4/mballoc.c
+++ b/fs/ext4/mballoc.c
@@ -3033,6 +3033,92 @@ ext4_mb_mark_diskspace_used(struct ext4_allocation_context *ac,
return err;
}
+void ext4_mb_mark_used(struct super_block *sb, ext4_fsblk_t block,
+ int len)
+{
+ struct buffer_head *bitmap_bh = NULL;
+ struct ext4_group_desc *gdp;
+ struct buffer_head *gdp_bh;
+ struct ext4_sb_info *sbi = EXT4_SB(sb);
+ ext4_group_t group;
+ ext4_fsblk_t cluster;
+ ext4_grpblk_t blkoff;
+ int i, clen, err;
+ int already_allocated_count;
+
+ cluster = EXT4_B2C(sbi, block);
+ clen = EXT4_B2C(sbi, len);
+
+ ext4_get_group_no_and_offset(sb, block, &group, &blkoff);
+ bitmap_bh = ext4_read_block_bitmap(sb, group);
+ if (IS_ERR(bitmap_bh)) {
+ err = PTR_ERR(bitmap_bh);
+ bitmap_bh = NULL;
+ goto out_err;
+ }
+
+ err = -EIO;
+ gdp = ext4_get_group_desc(sb, group, &gdp_bh);
+ if (!gdp)
+ goto out_err;
+
+ if (!ext4_data_block_valid(sbi, block, len)) {
+ ext4_error(sb, "Allocating blks %llu-%llu which overlap mdata",
+ cluster, cluster+clen);
+ /* File system mounted not to panic on error
+ * Fix the bitmap and return EFSCORRUPTED
+ * We leak some of the blocks here.
+ */
+ ext4_lock_group(sb, group);
+ ext4_set_bits(bitmap_bh->b_data, blkoff, clen);
+ ext4_unlock_group(sb, group);
+ err = ext4_handle_dirty_metadata(NULL, NULL, bitmap_bh);
+ if (!err)
+ err = -EFSCORRUPTED;
+ sync_dirty_buffer(bitmap_bh);
+ goto out_err;
+ }
+
+ ext4_lock_group(sb, group);
+ already_allocated_count = 0;
+ for (i = 0; i < clen; i++)
+ if (mb_test_bit(blkoff + i, bitmap_bh->b_data))
+ already_allocated_count++;
+
+ ext4_set_bits(bitmap_bh->b_data, blkoff, clen);
+ if (ext4_has_group_desc_csum(sb) &&
+ (gdp->bg_flags & cpu_to_le16(EXT4_BG_BLOCK_UNINIT))) {
+ gdp->bg_flags &= cpu_to_le16(~EXT4_BG_BLOCK_UNINIT);
+ ext4_free_group_clusters_set(sb, gdp,
+ ext4_free_clusters_after_init(sb,
+ group, gdp));
+ }
+ clen = ext4_free_group_clusters(sb, gdp) - clen +
+ already_allocated_count;
+ ext4_free_group_clusters_set(sb, gdp, clen);
+ ext4_block_bitmap_csum_set(sb, group, gdp, bitmap_bh);
+ ext4_group_desc_csum_set(sb, group, gdp);
+
+ ext4_unlock_group(sb, group);
+
+ if (sbi->s_log_groups_per_flex) {
+ ext4_group_t flex_group = ext4_flex_group(sbi, group);
+
+ atomic64_sub(len,
+ &sbi->s_flex_groups[flex_group].free_clusters);
+ }
+
+ err = ext4_handle_dirty_metadata(NULL, NULL, bitmap_bh);
+ if (err)
+ goto out_err;
+ sync_dirty_buffer(bitmap_bh);
+ err = ext4_handle_dirty_metadata(NULL, NULL, gdp_bh);
+ sync_dirty_buffer(gdp_bh);
+
+out_err:
+ brelse(bitmap_bh);
+}
+
/*
* here we normalize request for locality group
* Group request are normalized to s_mb_group_prealloc, which goes to
@@ -4692,6 +4778,47 @@ ext4_mb_free_metadata(handle_t *handle, struct ext4_buddy *e4b,
return 0;
}
+void ext4_free_blocks_simple(struct inode *inode, ext4_fsblk_t block,
+ unsigned long count)
+{
+ struct buffer_head *bitmap_bh;
+ struct super_block *sb = inode->i_sb;
+ struct ext4_group_desc *gdp;
+ struct buffer_head *gdp_bh;
+ ext4_group_t group;
+ ext4_grpblk_t blkoff;
+ int already_freed = 0, err, i;
+
+ ext4_get_group_no_and_offset(sb, block, &group, &blkoff);
+ bitmap_bh = ext4_read_block_bitmap(sb, group);
+ if (IS_ERR(bitmap_bh)) {
+ err = PTR_ERR(bitmap_bh);
+ pr_warn("Failed to read block bitmap\n");
+ return;
+ }
+ gdp = ext4_get_group_desc(sb, group, &gdp_bh);
+ if (!gdp)
+ return;
+
+ for (i = 0; i < count; i++) {
+ if (!mb_test_bit(blkoff + i, bitmap_bh->b_data))
+ already_freed++;
+ }
+ mb_clear_bits(bitmap_bh->b_data, blkoff, count);
+ err = ext4_handle_dirty_metadata(NULL, NULL, bitmap_bh);
+ if (err)
+ return;
+ ext4_free_group_clusters_set(
+ sb, gdp, ext4_free_group_clusters(sb, gdp) +
+ count - already_freed);
+ ext4_block_bitmap_csum_set(sb, group, gdp, bitmap_bh);
+ ext4_group_desc_csum_set(sb, group, gdp);
+ ext4_handle_dirty_metadata(NULL, NULL, gdp_bh);
+ sync_dirty_buffer(bitmap_bh);
+ sync_dirty_buffer(gdp_bh);
+ brelse(bitmap_bh);
+}
+
/**
* ext4_free_blocks() -- Free given blocks and update quota
* @handle: handle for this transaction
@@ -4718,6 +4845,13 @@ void ext4_free_blocks(handle_t *handle, struct inode *inode,
int err = 0;
int ret;
+ sbi = EXT4_SB(sb);
+
+ if (sbi->s_mount_state & EXT4_FC_REPLAY) {
+ ext4_free_blocks_simple(inode, block, count);
+ return;
+ }
+
might_sleep();
if (bh) {
if (block)
@@ -4726,7 +4860,6 @@ void ext4_free_blocks(handle_t *handle, struct inode *inode,
block = bh->b_blocknr;
}
- sbi = EXT4_SB(sb);
if (!(flags & EXT4_FREE_BLOCKS_VALIDATED) &&
!ext4_data_block_valid(sbi, block, count)) {
ext4_error(sb, "Freeing blocks not in datazone - "
diff --git a/fs/ext4/mballoc.h b/fs/ext4/mballoc.h
index 88c98f17e3d9..1881710041b6 100644
--- a/fs/ext4/mballoc.h
+++ b/fs/ext4/mballoc.h
@@ -215,4 +215,6 @@ ext4_mballoc_query_range(
ext4_mballoc_query_range_fn formatter,
void *priv);
+void ext4_mb_mark_used(struct super_block *sb, ext4_fsblk_t block,
+ int len);
#endif
--
2.24.1.735.g03f4e72817-goog
^ permalink raw reply related [flat|nested] 31+ messages in thread* Re: [PATCH v4 17/20] ext4: add idempotent helpers to manipulate bitmaps
2019-12-24 8:13 ` [PATCH v4 17/20] ext4: add idempotent helpers to manipulate bitmaps Harshad Shirwadkar
@ 2019-12-27 19:36 ` kbuild test robot
2019-12-27 19:36 ` [RFC PATCH] ext4: ext4_free_blocks_simple() can be static kbuild test robot
1 sibling, 0 replies; 31+ messages in thread
From: kbuild test robot @ 2019-12-27 19:36 UTC (permalink / raw)
To: Harshad Shirwadkar; +Cc: kbuild-all, linux-ext4, Harshad Shirwadkar
Hi Harshad,
Thank you for the patch! Perhaps something to improve:
[auto build test WARNING on tip/perf/core]
[cannot apply to ext4/dev linus/master v5.5-rc3 next-20191220]
[if your patch is applied to the wrong git tree, please drop us a note to help
improve the system. BTW, we also suggest to use '--base' option to specify the
base tree in git format-patch, please see https://stackoverflow.com/a/37406982]
url: https://github.com/0day-ci/linux/commits/Harshad-Shirwadkar/ext4-update-docs-for-fast-commit-feature/20191225-200339
base: https://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git ceb9e77324fa661b1001a0ae66f061b5fcb4e4e6
reproduce:
# apt-get install sparse
# sparse version: v0.6.1-129-g341daf20-dirty
make ARCH=x86_64 allmodconfig
make C=1 CF='-fdiagnostic-prefix -D__CHECK_ENDIAN__'
If you fix the issue, kindly add following tag
Reported-by: kbuild test robot <lkp@intel.com>
sparse warnings: (new ones prefixed by >>)
>> fs/ext4/mballoc.c:4781:6: sparse: sparse: symbol 'ext4_free_blocks_simple' was not declared. Should it be static?
Please review and possibly fold the followup patch.
---
0-DAY kernel test infrastructure Open Source Technology Center
https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org Intel Corporation
^ permalink raw reply [flat|nested] 31+ messages in thread* [RFC PATCH] ext4: ext4_free_blocks_simple() can be static
2019-12-24 8:13 ` [PATCH v4 17/20] ext4: add idempotent helpers to manipulate bitmaps Harshad Shirwadkar
2019-12-27 19:36 ` kbuild test robot
@ 2019-12-27 19:36 ` kbuild test robot
1 sibling, 0 replies; 31+ messages in thread
From: kbuild test robot @ 2019-12-27 19:36 UTC (permalink / raw)
To: Harshad Shirwadkar; +Cc: kbuild-all, linux-ext4, Harshad Shirwadkar
Fixes: bb863fa8f074 ("ext4: add idempotent helpers to manipulate bitmaps")
Signed-off-by: kbuild test robot <lkp@intel.com>
---
mballoc.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
index 05ca9001f8fa0..d5e9cc938338f 100644
--- a/fs/ext4/mballoc.c
+++ b/fs/ext4/mballoc.c
@@ -4778,8 +4778,8 @@ ext4_mb_free_metadata(handle_t *handle, struct ext4_buddy *e4b,
return 0;
}
-void ext4_free_blocks_simple(struct inode *inode, ext4_fsblk_t block,
- unsigned long count)
+static void ext4_free_blocks_simple(struct inode *inode, ext4_fsblk_t block,
+ unsigned long count)
{
struct buffer_head *bitmap_bh;
struct super_block *sb = inode->i_sb;
^ permalink raw reply related [flat|nested] 31+ messages in thread
* [PATCH v4 18/20] ext4: disable certain features in replay path
2019-12-24 8:13 [PATCH v4 01/20] ext4: update docs for fast commit feature Harshad Shirwadkar
` (15 preceding siblings ...)
2019-12-24 8:13 ` [PATCH v4 17/20] ext4: add idempotent helpers to manipulate bitmaps Harshad Shirwadkar
@ 2019-12-24 8:13 ` Harshad Shirwadkar
2019-12-24 8:13 ` [PATCH v4 19/20] ext4: add fast commit " Harshad Shirwadkar
` (2 subsequent siblings)
19 siblings, 0 replies; 31+ messages in thread
From: Harshad Shirwadkar @ 2019-12-24 8:13 UTC (permalink / raw)
To: linux-ext4; +Cc: Harshad Shirwadkar
Replay path uses similar code paths for replaying committed changes.
But since it runs before full initialization of the file system and
also since we don't have to be super careful about performance, we
can and need to disable certain file system features during the replay
path. More specifically, we disable most of the extent status tree
stuff, mballoc and some places where we mark file system with errors.
Signed-off-by: Harshad Shirwadkar <harshadshirwadkar@gmail.com>
---
fs/ext4/balloc.c | 7 +++++-
fs/ext4/ext4_jbd2.c | 2 +-
fs/ext4/extents_status.c | 24 +++++++++++++++++++
fs/ext4/ialloc.c | 52 +++++++++++++++++++++++++++-------------
fs/ext4/inode.c | 15 ++++++++----
fs/ext4/mballoc.c | 22 +++++++++++------
6 files changed, 92 insertions(+), 30 deletions(-)
diff --git a/fs/ext4/balloc.c b/fs/ext4/balloc.c
index 14787065d030..2cf39a6b9f7a 100644
--- a/fs/ext4/balloc.c
+++ b/fs/ext4/balloc.c
@@ -360,7 +360,12 @@ static int ext4_validate_block_bitmap(struct super_block *sb,
struct buffer_head *bh)
{
ext4_fsblk_t blk;
- struct ext4_group_info *grp = ext4_get_group_info(sb, block_group);
+ struct ext4_group_info *grp;
+
+ if (EXT4_SB(sb)->s_mount_state & EXT4_FC_REPLAY)
+ return 0;
+
+ grp = ext4_get_group_info(sb, block_group);
if (buffer_verified(bh))
return 0;
diff --git a/fs/ext4/ext4_jbd2.c b/fs/ext4/ext4_jbd2.c
index f371f1f0f914..7c6cdbc63aa6 100644
--- a/fs/ext4/ext4_jbd2.c
+++ b/fs/ext4/ext4_jbd2.c
@@ -77,7 +77,7 @@ handle_t *__ext4_journal_start_sb(struct super_block *sb, unsigned int line,
return ERR_PTR(err);
journal = EXT4_SB(sb)->s_journal;
- if (!journal)
+ if (!journal || (EXT4_SB(sb)->s_mount_state & EXT4_FC_REPLAY))
return ext4_get_nojournal();
return jbd2__journal_start(journal, blocks, rsv_blocks, GFP_NOFS,
type, line);
diff --git a/fs/ext4/extents_status.c b/fs/ext4/extents_status.c
index d996b44d2265..69c16ac7416e 100644
--- a/fs/ext4/extents_status.c
+++ b/fs/ext4/extents_status.c
@@ -311,6 +311,9 @@ void ext4_es_find_extent_range(struct inode *inode,
ext4_lblk_t lblk, ext4_lblk_t end,
struct extent_status *es)
{
+ if (EXT4_SB(inode->i_sb)->s_mount_state & EXT4_FC_REPLAY)
+ return;
+
trace_ext4_es_find_extent_range_enter(inode, lblk);
read_lock(&EXT4_I(inode)->i_es_lock);
@@ -361,6 +364,9 @@ bool ext4_es_scan_range(struct inode *inode,
{
bool ret;
+ if (EXT4_SB(inode->i_sb)->s_mount_state & EXT4_FC_REPLAY)
+ return false;
+
read_lock(&EXT4_I(inode)->i_es_lock);
ret = __es_scan_range(inode, matching_fn, lblk, end);
read_unlock(&EXT4_I(inode)->i_es_lock);
@@ -404,6 +410,9 @@ bool ext4_es_scan_clu(struct inode *inode,
{
bool ret;
+ if (EXT4_SB(inode->i_sb)->s_mount_state & EXT4_FC_REPLAY)
+ return false;
+
read_lock(&EXT4_I(inode)->i_es_lock);
ret = __es_scan_clu(inode, matching_fn, lblk);
read_unlock(&EXT4_I(inode)->i_es_lock);
@@ -812,6 +821,9 @@ int ext4_es_insert_extent(struct inode *inode, ext4_lblk_t lblk,
int err = 0;
struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
+ if (EXT4_SB(inode->i_sb)->s_mount_state & EXT4_FC_REPLAY)
+ return 0;
+
es_debug("add [%u/%u) %llu %x to extent status tree of inode %lu\n",
lblk, len, pblk, status, inode->i_ino);
@@ -873,6 +885,9 @@ void ext4_es_cache_extent(struct inode *inode, ext4_lblk_t lblk,
struct extent_status newes;
ext4_lblk_t end = lblk + len - 1;
+ if (EXT4_SB(inode->i_sb)->s_mount_state & EXT4_FC_REPLAY)
+ return;
+
newes.es_lblk = lblk;
newes.es_len = len;
ext4_es_store_pblock_status(&newes, pblk, status);
@@ -908,6 +923,9 @@ int ext4_es_lookup_extent(struct inode *inode, ext4_lblk_t lblk,
struct rb_node *node;
int found = 0;
+ if (EXT4_SB(inode->i_sb)->s_mount_state & EXT4_FC_REPLAY)
+ return 0;
+
trace_ext4_es_lookup_extent_enter(inode, lblk);
es_debug("lookup extent in block %u\n", lblk);
@@ -1419,6 +1437,9 @@ int ext4_es_remove_extent(struct inode *inode, ext4_lblk_t lblk,
int err = 0;
int reserved = 0;
+ if (EXT4_SB(inode->i_sb)->s_mount_state & EXT4_FC_REPLAY)
+ return 0;
+
trace_ext4_es_remove_extent(inode, lblk, len);
es_debug("remove [%u/%u) from extent status tree of inode %lu\n",
lblk, len, inode->i_ino);
@@ -1969,6 +1990,9 @@ int ext4_es_insert_delayed_block(struct inode *inode, ext4_lblk_t lblk,
struct extent_status newes;
int err = 0;
+ if (EXT4_SB(inode->i_sb)->s_mount_state & EXT4_FC_REPLAY)
+ return 0;
+
es_debug("add [%u/1) delayed to extent status tree of inode %lu\n",
lblk, inode->i_ino);
diff --git a/fs/ext4/ialloc.c b/fs/ext4/ialloc.c
index f1a1432f9ffa..24a2d7171cd4 100644
--- a/fs/ext4/ialloc.c
+++ b/fs/ext4/ialloc.c
@@ -82,7 +82,12 @@ static int ext4_validate_inode_bitmap(struct super_block *sb,
struct buffer_head *bh)
{
ext4_fsblk_t blk;
- struct ext4_group_info *grp = ext4_get_group_info(sb, block_group);
+ struct ext4_group_info *grp;
+
+ if (EXT4_SB(sb)->s_mount_state & EXT4_FC_REPLAY)
+ return 0;
+
+ grp = ext4_get_group_info(sb, block_group);
if (buffer_verified(bh))
return 0;
@@ -287,15 +292,17 @@ void ext4_free_inode(handle_t *handle, struct inode *inode)
bit = (ino - 1) % EXT4_INODES_PER_GROUP(sb);
bitmap_bh = ext4_read_inode_bitmap(sb, block_group);
/* Don't bother if the inode bitmap is corrupt. */
- grp = ext4_get_group_info(sb, block_group);
if (IS_ERR(bitmap_bh)) {
fatal = PTR_ERR(bitmap_bh);
bitmap_bh = NULL;
goto error_return;
}
- if (unlikely(EXT4_MB_GRP_IBITMAP_CORRUPT(grp))) {
- fatal = -EFSCORRUPTED;
- goto error_return;
+ if (!(sbi->s_mount_state & EXT4_FC_REPLAY)) {
+ grp = ext4_get_group_info(sb, block_group);
+ if (unlikely(EXT4_MB_GRP_IBITMAP_CORRUPT(grp))) {
+ fatal = -EFSCORRUPTED;
+ goto error_return;
+ }
}
BUFFER_TRACE(bitmap_bh, "get_write_access");
@@ -871,7 +878,7 @@ struct inode *__ext4_new_inode(handle_t *handle, struct inode *dir,
struct inode *ret;
ext4_group_t i;
ext4_group_t flex_group;
- struct ext4_group_info *grp;
+ struct ext4_group_info *grp = NULL;
int encrypt = 0;
/* Cannot create files in a deleted directory */
@@ -1009,15 +1016,21 @@ struct inode *__ext4_new_inode(handle_t *handle, struct inode *dir,
if (ext4_free_inodes_count(sb, gdp) == 0)
goto next_group;
- grp = ext4_get_group_info(sb, group);
- /* Skip groups with already-known suspicious inode tables */
- if (EXT4_MB_GRP_IBITMAP_CORRUPT(grp))
- goto next_group;
+ if (!(sbi->s_mount_state & EXT4_FC_REPLAY)) {
+ grp = ext4_get_group_info(sb, group);
+ /*
+ * Skip groups with already-known suspicious inode
+ * tables
+ */
+ if (EXT4_MB_GRP_IBITMAP_CORRUPT(grp))
+ goto next_group;
+ }
brelse(inode_bitmap_bh);
inode_bitmap_bh = ext4_read_inode_bitmap(sb, group);
/* Skip groups with suspicious inode tables */
- if (EXT4_MB_GRP_IBITMAP_CORRUPT(grp) ||
+ if (((!(sbi->s_mount_state & EXT4_FC_REPLAY))
+ && EXT4_MB_GRP_IBITMAP_CORRUPT(grp)) ||
IS_ERR(inode_bitmap_bh)) {
inode_bitmap_bh = NULL;
goto next_group;
@@ -1036,7 +1049,7 @@ struct inode *__ext4_new_inode(handle_t *handle, struct inode *dir,
goto next_group;
}
- if (!handle) {
+ if ((!(sbi->s_mount_state & EXT4_FC_REPLAY)) && !handle) {
BUG_ON(nblocks <= 0);
handle = __ext4_journal_start_sb(dir->i_sb, line_no,
handle_type, nblocks,
@@ -1140,9 +1153,15 @@ struct inode *__ext4_new_inode(handle_t *handle, struct inode *dir,
/* Update the relevant bg descriptor fields */
if (ext4_has_group_desc_csum(sb)) {
int free;
- struct ext4_group_info *grp = ext4_get_group_info(sb, group);
-
- down_read(&grp->alloc_sem); /* protect vs itable lazyinit */
+ struct ext4_group_info *grp = NULL;
+
+ if (!(sbi->s_mount_state & EXT4_FC_REPLAY)) {
+ grp = ext4_get_group_info(sb, group);
+ down_read(&grp->alloc_sem); /*
+ * protect vs itable
+ * lazyinit
+ */
+ }
ext4_lock_group(sb, group); /* while we modify the bg desc */
free = EXT4_INODES_PER_GROUP(sb) -
ext4_itable_unused_count(sb, gdp);
@@ -1158,7 +1177,8 @@ struct inode *__ext4_new_inode(handle_t *handle, struct inode *dir,
if (ino > free)
ext4_itable_unused_set(sb, gdp,
(EXT4_INODES_PER_GROUP(sb) - ino));
- up_read(&grp->alloc_sem);
+ if (!(sbi->s_mount_state & EXT4_FC_REPLAY))
+ up_read(&grp->alloc_sem);
} else {
ext4_lock_group(sb, group);
}
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index e902000dac51..f8090f5ce5e4 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -527,7 +527,8 @@ int ext4_map_blocks(handle_t *handle, struct inode *inode,
return -EFSCORRUPTED;
/* Lookup extent status tree firstly */
- if (ext4_es_lookup_extent(inode, map->m_lblk, NULL, &es)) {
+ if (!(EXT4_SB(inode->i_sb)->s_mount_state & EXT4_FC_REPLAY) &&
+ ext4_es_lookup_extent(inode, map->m_lblk, NULL, &es)) {
if (ext4_es_is_written(&es) || ext4_es_is_unwritten(&es)) {
map->m_pblk = ext4_es_pblock(&es) +
map->m_lblk - es.es_lblk;
@@ -969,7 +970,8 @@ struct buffer_head *ext4_getblk(handle_t *handle, struct inode *inode,
int create = map_flags & EXT4_GET_BLOCKS_CREATE;
int err;
- J_ASSERT(handle != NULL || create == 0);
+ J_ASSERT((EXT4_SB(inode->i_sb)->s_mount_state | EXT4_FC_REPLAY)
+ || handle != NULL || create == 0);
map.m_lblk = block;
map.m_len = 1;
@@ -985,7 +987,8 @@ struct buffer_head *ext4_getblk(handle_t *handle, struct inode *inode,
return ERR_PTR(-ENOMEM);
if (map.m_flags & EXT4_MAP_NEW) {
J_ASSERT(create != 0);
- J_ASSERT(handle != NULL);
+ J_ASSERT((EXT4_SB(inode->i_sb)->s_mount_state & EXT4_FC_REPLAY)
+ || (handle != NULL));
/*
* Now that we do not always journal data, we should
@@ -4896,8 +4899,10 @@ struct inode *__ext4_iget(struct super_block *sb, unsigned long ino,
}
if (!ext4_inode_csum_verify(inode, raw_inode, ei)) {
- ext4_error_inode(inode, function, line, 0,
- "iget: checksum invalid");
+
+ if (!(EXT4_SB(sb)->s_mount_state & EXT4_FC_REPLAY))
+ ext4_error_inode(inode, function, line, 0,
+ "iget: checksum invalid");
ret = -EFSBADCRC;
goto bad_inode;
}
diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
index 05ca9001f8fa..4f2d611e5a75 100644
--- a/fs/ext4/mballoc.c
+++ b/fs/ext4/mballoc.c
@@ -1449,14 +1449,17 @@ static void mb_free_blocks(struct inode *inode, struct ext4_buddy *e4b,
blocknr = ext4_group_first_block_no(sb, e4b->bd_group);
blocknr += EXT4_C2B(sbi, block);
- ext4_grp_locked_error(sb, e4b->bd_group,
- inode ? inode->i_ino : 0,
- blocknr,
- "freeing already freed block "
- "(bit %u); block bitmap corrupt.",
- block);
- ext4_mark_group_bitmap_corrupted(sb, e4b->bd_group,
+ if (!(sbi->s_mount_state & EXT4_FC_REPLAY)) {
+ ext4_grp_locked_error(sb, e4b->bd_group,
+ inode ? inode->i_ino : 0,
+ blocknr,
+ "freeing already freed block "
+ "(bit %u); block bitmap corrupt.",
+ block);
+ ext4_mark_group_bitmap_corrupted(
+ sb, e4b->bd_group,
EXT4_GROUP_INFO_BBITMAP_CORRUPT);
+ }
mb_regenerate_buddy(e4b);
goto done;
}
@@ -4088,6 +4091,9 @@ void ext4_discard_preallocations(struct inode *inode)
return;
}
+ if (EXT4_SB(sb)->s_mount_state & EXT4_FC_REPLAY)
+ return;
+
mb_debug(1, "discard preallocation for inode %lu\n", inode->i_ino);
trace_ext4_discard_preallocations(inode);
@@ -4561,6 +4567,8 @@ ext4_fsblk_t ext4_mb_new_blocks(handle_t *handle,
sb = ar->inode->i_sb;
sbi = EXT4_SB(sb);
+ WARN_ON(sbi->s_mount_state & EXT4_FC_REPLAY);
+
trace_ext4_request_blocks(ar);
/* Allow to use superuser reservation for quota file */
--
2.24.1.735.g03f4e72817-goog
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PATCH v4 19/20] ext4: add fast commit replay path
2019-12-24 8:13 [PATCH v4 01/20] ext4: update docs for fast commit feature Harshad Shirwadkar
` (16 preceding siblings ...)
2019-12-24 8:13 ` [PATCH v4 18/20] ext4: disable certain features in replay path Harshad Shirwadkar
@ 2019-12-24 8:13 ` Harshad Shirwadkar
2019-12-27 22:51 ` kbuild test robot
2019-12-24 8:13 ` [PATCH v4 20/20] ext4: add debug mount option to test fast commit replay Harshad Shirwadkar
2020-01-09 4:29 ` [PATCH v4 01/20] ext4: update docs for fast commit feature xiaohui li
19 siblings, 1 reply; 31+ messages in thread
From: Harshad Shirwadkar @ 2019-12-24 8:13 UTC (permalink / raw)
To: linux-ext4; +Cc: Harshad Shirwadkar
Add main routine for replaying fast commit blocks. Fast commit replay
routine should be idempotent; so that if we crash while replaying,
we can restart from the beginning and won't result in a corrupted
file system.
Signed-off-by: Harshad Shirwadkar <harshadshirwadkar@gmail.com>
---
fs/ext4/ext4.h | 9 +
fs/ext4/ext4_jbd2.c | 459 ++++++++++++++++++++++++++++++++++++
include/trace/events/ext4.h | 22 ++
3 files changed, 490 insertions(+)
diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index 1957f30a7f2e..e12900d77673 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -1436,6 +1436,14 @@ struct ext4_super_block {
#define ext4_has_strict_mode(sbi) \
(sbi->s_encoding_flags & EXT4_ENC_STRICT_MODE_FL)
+/*
+ * Fast commit replay state.
+ */
+struct ext4_fc_replay_state {
+ int fc_replay_error;
+ int fc_replay_expected_off;
+};
+
/*
* Fast commit ineligible reasons.
*/
@@ -1642,6 +1650,7 @@ struct ext4_sb_info {
* that have data changes in them.
*/
struct list_head s_fc_dentry_q;
+ struct ext4_fc_replay_state s_fc_replay_state;
spinlock_t s_fc_lock;
struct ext4_fc_stats s_fc_stats;
};
diff --git a/fs/ext4/ext4_jbd2.c b/fs/ext4/ext4_jbd2.c
index 7c6cdbc63aa6..ef36a973ed8b 100644
--- a/fs/ext4/ext4_jbd2.c
+++ b/fs/ext4/ext4_jbd2.c
@@ -5,6 +5,7 @@
#include "ext4_jbd2.h"
#include "ext4_extents.h"
+#include "mballoc.h"
#include <trace/events/ext4.h>
@@ -1153,6 +1154,457 @@ static void ext4_journal_fc_cleanup_cb(journal_t *journal)
trace_ext4_journal_fc_stats(sb);
}
+struct dentry_info_args {
+ int parent_ino, dname_len, ino, inode_len;
+ char *dname;
+};
+
+static int fc_replay_add_link(struct super_block *sb, struct inode *inode,
+ struct dentry_info_args *darg)
+{
+ struct inode *dir = NULL;
+ struct dentry *dentry_dir = NULL, *dentry_inode = NULL;
+ struct qstr qstr_dname = QSTR_INIT(darg->dname, darg->dname_len);
+ int ret = 0;
+
+ dir = ext4_iget(sb, darg->parent_ino, EXT4_IGET_NORMAL);
+ if (IS_ERR(dir)) {
+ jbd_debug(1, "Dir with inode %d not found.", darg->parent_ino);
+ ret = PTR_ERR(dir);
+ dir = NULL;
+ goto out;
+ }
+
+ dentry_dir = d_obtain_alias(dir);
+ if (IS_ERR(dentry_dir)) {
+ jbd_debug(1, "Failed to obtain dentry");
+ ret = PTR_ERR(dentry_dir);
+ dentry_dir = NULL;
+ goto out;
+ }
+
+ dentry_inode = d_alloc(dentry_dir, &qstr_dname);
+ if (!dentry_inode) {
+ jbd_debug(1, "Inode dentry not created.");
+ ret = -ENOMEM;
+ goto out;
+ }
+
+ ret = __ext4_link(dir, inode, dentry_inode);
+ if (ret && ret != -EEXIST) {
+ jbd_debug(1, "Failed to link\n");
+ goto out;
+ }
+
+ /*
+ * It's possible that link already existed since data blocks
+ * for the dir in question got persisted before we crashed.
+ */
+ if (ret == -EEXIST)
+ ret = 0;
+out:
+ if (dentry_dir) {
+ d_drop(dentry_dir);
+ dput(dentry_dir);
+ } else if (dir) {
+ iput(dir);
+ }
+ if (dentry_inode) {
+ d_drop(dentry_inode);
+ dput(dentry_inode);
+ }
+
+ return ret;
+}
+
+static int fc_replay_create_inode(struct super_block *sb,
+ struct ext4_inode *raw_inode,
+ struct dentry_info_args *darg)
+{
+ int ret = 0;
+ struct ext4_iloc iloc;
+ int orig_nlink = 0;
+ struct inode *inode = NULL;
+ struct inode *dir = NULL;
+
+ /*
+ * First let's setup the on-disk inode using the one found in
+ * the journal
+ */
+ ret = ext4_get_fc_inode_loc(sb, darg->ino, &iloc);
+ if (ret)
+ goto out;
+
+ orig_nlink = le16_to_cpu(ext4_raw_inode(&iloc)->i_links_count);
+ memcpy(ext4_raw_inode(&iloc), raw_inode, darg->inode_len);
+ ret = ext4_handle_dirty_metadata(NULL, NULL, iloc.bh);
+ if (ret)
+ goto out;
+ sync_dirty_buffer(iloc.bh);
+ brelse(iloc.bh);
+ iloc.bh = NULL;
+
+ /* This takes care of update group descriptor and other metadata */
+ ret = ext4_mark_inode_used(sb, darg->ino);
+ if (ret)
+ goto out;
+
+ inode = ext4_iget(sb, darg->ino, EXT4_IGET_NORMAL);
+ if (IS_ERR(inode)) {
+ jbd_debug(1, "inode %d not found.", darg->ino);
+ ret = PTR_ERR(inode);
+ goto out;
+ }
+
+ if (S_ISDIR(inode->i_mode)) {
+ dir = ext4_iget(sb, darg->parent_ino, EXT4_IGET_NORMAL);
+ if (IS_ERR_OR_NULL(dir)) {
+ iput(inode);
+ ret = PTR_ERR(dir);
+ goto out;
+ }
+ ret = ext4_init_new_dir(NULL, dir, inode);
+ iput(dir);
+ if (ret)
+ goto out;
+ }
+ ret = fc_replay_add_link(sb, inode, darg);
+ if (ret)
+ goto out;
+ set_nlink(inode, orig_nlink + 1);
+out:
+ if (inode)
+ iput(inode);
+ if (iloc.bh)
+ brelse(iloc.bh);
+ return ret;
+}
+
+static int fc_replay_dentries(journal_t *journal,
+ struct ext4_fc_commit_hdr *fc_hdr)
+{
+ struct dentry_info_args darg = {0};
+ struct super_block *sb = journal->j_private;
+ struct ext4_fc_tl *tl;
+ __u8 *start;
+ int inode_len = EXT4_GOOD_OLD_INODE_SIZE;
+ struct inode *old_parent;
+ struct inode *inode;
+ int ret;
+ int i;
+ struct ext4_fc_dentry_info *fcd;
+
+ if (EXT4_INODE_SIZE(sb) > EXT4_GOOD_OLD_INODE_SIZE)
+ inode_len +=
+ le16_to_cpu(((struct ext4_inode *)
+ (fc_hdr + 1))->i_extra_isize);
+ tl = (struct ext4_fc_tl *)((u8 *)fc_hdr +
+ sizeof(struct ext4_fc_commit_hdr) +
+ inode_len);
+ start = (__u8 *)tl;
+ for (i = 0; i < le16_to_cpu(fc_hdr->fc_num_tlvs); i++) {
+ fcd = (struct ext4_fc_dentry_info *)fc_tag_val(tl);
+
+ darg.parent_ino = le32_to_cpu(fcd->fc_parent_ino);
+ darg.ino = le32_to_cpu(fcd->fc_ino);
+ darg.dname = fcd->fc_dname;
+ darg.dname_len = fc_tag_len(tl) -
+ sizeof(struct ext4_fc_dentry_info);
+ if (le16_to_cpu(tl->fc_tag) == EXT4_FC_TAG_ADD_DENTRY) {
+ inode = ext4_iget(sb, darg.ino, EXT4_IGET_NORMAL);
+ if (IS_ERR_OR_NULL(inode)) {
+ jbd_debug(1, "Inode not found.");
+ return PTR_ERR(inode);
+ }
+ ret = fc_replay_add_link(sb, inode, &darg);
+ iput(inode);
+ if (ret)
+ return ret;
+ } else if (le16_to_cpu(tl->fc_tag) == EXT4_FC_TAG_DEL_DENTRY) {
+ const struct qstr entry = {
+ .name = darg.dname,
+ .len = darg.dname_len
+ };
+ inode = ext4_iget(sb, darg.ino, EXT4_IGET_NORMAL);
+
+ if (IS_ERR_OR_NULL(inode))
+ return -ECANCELED;
+
+ old_parent = ext4_iget(sb, darg.parent_ino,
+ EXT4_IGET_NORMAL);
+ if (IS_ERR_OR_NULL(old_parent)) {
+ iput(inode);
+ return -ECANCELED;
+ }
+
+ ret = __ext4_unlink(old_parent, &entry, inode);
+ /* -ENOENT ok coz it might not exist anymore. */
+ if (ret == -ENOENT)
+ ret = 0;
+ iput(old_parent);
+ iput(inode);
+ if (ret)
+ return ret;
+ } else if (le16_to_cpu(tl->fc_tag) ==
+ EXT4_FC_TAG_CREAT_DENTRY) {
+ darg.inode_len = inode_len;
+ ret = fc_replay_create_inode(
+ sb, (struct ext4_inode *)(fc_hdr + 1), &darg);
+ if (ret) {
+ jbd_debug(1, "Failed to create ext4 inode.");
+ return ret;
+ }
+ }
+ tl = (struct ext4_fc_tl *)((__u8 *)tl +
+ le16_to_cpu(tl->fc_len) +
+ sizeof(*tl));
+ }
+ return 0;
+}
+
+static int ext4_journal_fc_replay_scan(journal_t *journal,
+ struct buffer_head *bh, int off)
+{
+ struct super_block *sb = journal->j_private;
+ struct ext4_sb_info *sbi = EXT4_SB(sb);
+ struct ext4_fc_replay_state *state;
+ struct ext4_fc_commit_hdr *fc_hdr;
+ __u32 csum, old_csum;
+ __u8 *start, *end;
+
+ state = &sbi->s_fc_replay_state;
+ fc_hdr = (struct ext4_fc_commit_hdr *)
+ ((__u8 *)bh->b_data + sizeof(journal_header_t));
+
+ start = (u8 *)fc_hdr;
+ end = (__u8 *)bh->b_data + journal->j_blocksize;
+
+ /* Check if we already concluded that this fast commit is not useful */
+ if (state->fc_replay_expected_off && state->fc_replay_error)
+ goto out_err;
+
+ if (le32_to_cpu(fc_hdr->fc_magic) != EXT4_FC_MAGIC) {
+ state->fc_replay_error = -ENOENT;
+ goto out_err;
+ }
+
+ if (off != state->fc_replay_expected_off) {
+ state->fc_replay_error = -EFSCORRUPTED;
+ goto out_err;
+ }
+
+ state->fc_replay_expected_off++;
+
+ if (le16_to_cpu(fc_hdr->fc_features)) {
+ state->fc_replay_error = -EOPNOTSUPP;
+ goto out_err;
+ }
+
+ old_csum = fc_hdr->fc_csum;
+ fc_hdr->fc_csum = 0;
+ csum = ext4_chksum(sbi, 0, start, end - start);
+ fc_hdr->fc_csum = old_csum;
+
+ if (csum != le32_to_cpu(fc_hdr->fc_csum)) {
+ state->fc_replay_error = -EFSBADCRC;
+ goto out_err;
+ }
+
+ trace_ext4_journal_fc_replay_scan(sb, state->fc_replay_error, off);
+ return 0;
+
+out_err:
+ trace_ext4_journal_fc_replay_scan(sb, state->fc_replay_error, off);
+ return state->fc_replay_error;
+}
+
+static int fc_add_range(struct inode *inode, struct ext4_extent *ex)
+{
+ struct ext4_extent newex;
+ ext4_lblk_t start, cur;
+ int remaining, len;
+ ext4_fsblk_t start_pblk;
+ struct ext4_map_blocks map;
+ struct ext4_ext_path *path = NULL;
+ int ret;
+
+ start = le32_to_cpu(ex->ee_block);
+ start_pblk = ext4_ext_pblock(ex);
+ len = ext4_ext_get_actual_len(ex);
+
+ cur = start;
+ remaining = len;
+
+ jbd_debug(1, "Adding extent %ld:%ld to inode %ld\n",
+ start, len, inode->i_ino);
+
+ while (remaining > 0) {
+ map.m_lblk = cur;
+ map.m_len = remaining;
+
+ ret = ext4_map_blocks(NULL, inode, &map, 0);
+ if (ret < 0)
+ return -ECANCELED;
+ if (ret > 0) {
+ if (!!(ext4_ext_is_unwritten(ex)) ==
+ !!(map.m_flags & EXT4_MAP_UNWRITTEN)) {
+ remaining -= ret;
+ cur += ret;
+ ext4_mb_mark_used(inode->i_sb,
+ ext4_ext_pblock(ex),
+ map.m_len);
+ continue;
+ }
+
+ /* handle change of state */
+ map.m_lblk = cur;
+ map.m_len = ret;
+ map.m_flags = 0;
+ ret = ext4_map_blocks(
+ NULL, inode, &map,
+ EXT4_GET_BLOCKS_IO_CONVERT_EXT);
+ if (ret <= 0)
+ return -ECANCELED;
+ remaining -= ret;
+ cur += ret;
+ } else if (ret == 0) {
+ path = ext4_find_extent(inode, cur, NULL, 0);
+ if (!path)
+ continue;
+ memset(&newex, 0, sizeof(newex));
+ newex.ee_block = cpu_to_le32(cur);
+ ext4_ext_store_pblock(
+ &newex, start_pblk + cur - start);
+ newex.ee_len = cpu_to_le16(map.m_len);
+ if (ext4_ext_is_unwritten(ex))
+ ext4_ext_mark_unwritten(&newex);
+ down_write(&EXT4_I(inode)->i_data_sem);
+
+ ret = ext4_ext_insert_extent(
+ NULL, inode, &path, &newex, 0);
+ ext4_mb_mark_used(
+ inode->i_sb, ext4_ext_pblock(&newex),
+ map.m_len);
+ up_write((&EXT4_I(inode)->i_data_sem));
+ kfree(path);
+ if (ret)
+ return -ECANCELED;
+ cur += map.m_len;
+ remaining -= map.m_len;
+ }
+ }
+ return 0;
+}
+
+static int ext4_journal_fc_replay_cb(journal_t *journal, struct buffer_head *bh,
+ enum passtype pass, int off)
+{
+ struct super_block *sb = journal->j_private;
+ struct ext4_sb_info *sbi = EXT4_SB(sb);
+ struct ext4_fc_commit_hdr *fc_hdr;
+ struct ext4_fc_tl *tl;
+ struct ext4_iloc iloc;
+ struct ext4_extent *ex;
+ struct ext4_fc_lrange *lrange;
+ struct inode *inode;
+
+ int i, ret;
+ int inode_len = EXT4_GOOD_OLD_INODE_SIZE;
+
+ if (pass == PASS_SCAN)
+ return ext4_journal_fc_replay_scan(journal, bh, off);
+
+ if (sbi->s_fc_replay_state.fc_replay_error) {
+ jbd_debug(1, "FC replay error set = %d\n",
+ sbi->s_fc_replay_state.fc_replay_error);
+ return sbi->s_fc_replay_state.fc_replay_error;
+ }
+
+ sbi->s_mount_state |= EXT4_FC_REPLAY;
+ fc_hdr = (struct ext4_fc_commit_hdr *)
+ ((__u8 *)bh->b_data + sizeof(journal_header_t));
+
+ jbd_debug(3, "%s: Got FC block for inode %d at [%d,%d]", __func__,
+ le32_to_cpu(fc_hdr->fc_ino),
+ be32_to_cpu(((journal_header_t *)bh->b_data)->h_sequence));
+
+ if (EXT4_INODE_SIZE(sb) > EXT4_GOOD_OLD_INODE_SIZE)
+ inode_len += le16_to_cpu(((struct ext4_inode *)
+ (fc_hdr + 1))->i_extra_isize);
+
+ ret = fc_replay_dentries(journal, fc_hdr);
+
+ inode = ext4_iget(sb, le32_to_cpu(fc_hdr->fc_ino), EXT4_IGET_NORMAL);
+ if (IS_ERR(inode))
+ return 0;
+
+ ret = ext4_get_inode_loc(inode, &iloc);
+ if (ret)
+ return ret;
+
+ inode_lock(inode);
+ tl = (struct ext4_fc_tl *)((u8 *)fc_hdr +
+ sizeof(struct ext4_fc_commit_hdr) +
+ inode_len);
+ for (i = 0; i < le16_to_cpu(fc_hdr->fc_num_tlvs); i++) {
+ switch (le16_to_cpu(tl->fc_tag)) {
+ case EXT4_FC_TAG_ADD_RANGE:
+ ex = (struct ext4_extent *)(tl + 1);
+ ret = fc_add_range(inode, ex);
+ break;
+ case EXT4_FC_TAG_DEL_RANGE:
+ lrange = (struct ext4_fc_lrange *)(tl + 1);
+ inode_unlock(inode);
+ ret = ext4_punch_hole(inode,
+ le32_to_cpu(lrange->fc_lblk) <<
+ sb->s_blocksize_bits,
+ le32_to_cpu(lrange->fc_len) <<
+ sb->s_blocksize_bits);
+ inode_lock(inode);
+ break;
+ case EXT4_FC_TAG_ADD_DENTRY:
+ break;
+ default:
+ jbd_debug(1, "Unknown tag found.\n");
+ }
+ tl = (struct ext4_fc_tl *)((__u8 *)tl +
+ le16_to_cpu(tl->fc_len) +
+ sizeof(*tl));
+ }
+ ext4_reserve_inode_write(NULL, inode, &iloc);
+ inode_unlock(inode);
+
+ /*
+ * Unless inode contains inline data, copy everything except
+ * i_blocks. i_blocks would have been set alright by ext4_fc_add_block
+ * call above.
+ */
+ if (ext4_has_inline_data(inode)) {
+ memcpy(ext4_raw_inode(&iloc), fc_hdr + 1, inode_len);
+ } else {
+ memcpy(ext4_raw_inode(&iloc), fc_hdr + 1,
+ offsetof(struct ext4_inode, i_block));
+ memcpy(&ext4_raw_inode(&iloc)->i_generation,
+ &((struct ext4_inode *)(fc_hdr + 1))->i_generation,
+ inode_len -
+ offsetof(struct ext4_inode, i_generation));
+ }
+ inode->i_generation = le32_to_cpu(ext4_raw_inode(&iloc)->i_generation);
+ ext4_reset_inode_seed(inode);
+
+ ext4_inode_csum_set(inode, ext4_raw_inode(&iloc), EXT4_I(inode));
+ ret = ext4_handle_dirty_metadata(NULL, NULL, iloc.bh);
+ sync_dirty_buffer(iloc.bh);
+ brelse(iloc.bh);
+ iput(inode);
+ if (!ret)
+ ret = blkdev_issue_flush(sb->s_bdev, GFP_KERNEL, NULL);
+
+ sbi->s_mount_state &= ~EXT4_FC_REPLAY;
+
+ return ret;
+}
+
int ext4_fc_perform_hard_commit(journal_t *journal)
{
struct super_block *sb = (struct super_block *)(journal->j_private);
@@ -1202,6 +1654,7 @@ int ext4_fc_perform_hard_commit(journal_t *journal)
return nblks;
}
+
int ext4_fc_async_commit_inode(journal_t *journal, tid_t commit_tid,
struct inode *inode)
{
@@ -1331,6 +1784,12 @@ int ext4_fc_async_commit_inode(journal_t *journal, tid_t commit_tid,
void ext4_init_fast_commit(struct super_block *sb, journal_t *journal)
{
+ /*
+ * We set replay callback even if fast commit disabled because we may
+ * could still have fast commit blocks that need to be replayed even if
+ * fast commit has now been turned off.
+ */
+ journal->j_fc_replay_callback = ext4_journal_fc_replay_cb;
if (!ext4_should_fast_commit(sb))
return;
journal->j_fc_cleanup_callback = ext4_journal_fc_cleanup_cb;
diff --git a/include/trace/events/ext4.h b/include/trace/events/ext4.h
index 8da371b38332..c455f8384def 100644
--- a/include/trace/events/ext4.h
+++ b/include/trace/events/ext4.h
@@ -2716,6 +2716,28 @@ TRACE_EVENT(ext4_error,
__entry->function, __entry->line)
);
+TRACE_EVENT(ext4_journal_fc_replay_scan,
+ TP_PROTO(struct super_block *sb, int error, int off),
+
+ TP_ARGS(sb, error, off),
+
+ TP_STRUCT__entry(
+ __field(dev_t, dev)
+ __field(int, error)
+ __field(int, off)
+ ),
+
+ TP_fast_assign(
+ __entry->dev = sb->s_dev;
+ __entry->error = error;
+ __entry->off = off;
+ ),
+
+ TP_printk("FC scan pass on dev %d,%d: error %d, off %d",
+ MAJOR(__entry->dev), MINOR(__entry->dev),
+ __entry->error, __entry->off)
+);
+
TRACE_EVENT(ext4_journal_fc_commit_cb_start,
TP_PROTO(struct super_block *sb),
--
2.24.1.735.g03f4e72817-goog
^ permalink raw reply related [flat|nested] 31+ messages in thread* Re: [PATCH v4 19/20] ext4: add fast commit replay path
2019-12-24 8:13 ` [PATCH v4 19/20] ext4: add fast commit " Harshad Shirwadkar
@ 2019-12-27 22:51 ` kbuild test robot
0 siblings, 0 replies; 31+ messages in thread
From: kbuild test robot @ 2019-12-27 22:51 UTC (permalink / raw)
To: Harshad Shirwadkar; +Cc: kbuild-all, linux-ext4, Harshad Shirwadkar
Hi Harshad,
Thank you for the patch! Perhaps something to improve:
[auto build test WARNING on tip/perf/core]
[cannot apply to ext4/dev linus/master v5.5-rc3 next-20191220]
[if your patch is applied to the wrong git tree, please drop us a note to help
improve the system. BTW, we also suggest to use '--base' option to specify the
base tree in git format-patch, please see https://stackoverflow.com/a/37406982]
url: https://github.com/0day-ci/linux/commits/Harshad-Shirwadkar/ext4-update-docs-for-fast-commit-feature/20191225-200339
base: https://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git ceb9e77324fa661b1001a0ae66f061b5fcb4e4e6
reproduce:
# apt-get install sparse
# sparse version: v0.6.1-129-g341daf20-dirty
make ARCH=x86_64 allmodconfig
make C=1 CF='-fdiagnostic-prefix -D__CHECK_ENDIAN__'
If you fix the issue, kindly add following tag
Reported-by: kbuild test robot <lkp@intel.com>
sparse warnings: (new ones prefixed by >>)
fs/ext4/ext4_jbd2.c:612:5: sparse: sparse: symbol '__ext4_fc_track_range' was not declared. Should it be static?
fs/ext4/ext4_jbd2.c:826:6: sparse: sparse: symbol 'submit_fc_bh' was not declared. Should it be static?
>> fs/ext4/ext4_jbd2.c:1398:13: sparse: sparse: cast to restricted __le16
>> fs/ext4/ext4_jbd2.c:1403:18: sparse: sparse: incorrect type in assignment (different base types)
>> fs/ext4/ext4_jbd2.c:1403:18: sparse: expected unsigned int [usertype] old_csum
>> fs/ext4/ext4_jbd2.c:1403:18: sparse: got restricted __le32 [usertype] fc_csum
fs/ext4/ext4_jbd2.c:1406:25: sparse: sparse: incorrect type in assignment (different base types)
>> fs/ext4/ext4_jbd2.c:1406:25: sparse: expected restricted __le32 [usertype] fc_csum
>> fs/ext4/ext4_jbd2.c:1406:25: sparse: got unsigned int [usertype] old_csum
fs/ext4/ext4_jbd2.c:1608:5: sparse: sparse: symbol 'ext4_fc_perform_hard_commit' was not declared. Should it be static?
fs/ext4/ext4_jbd2.c:501:12: sparse: sparse: context imbalance in '__ext4_dentry_update' - unexpected unlock
fs/ext4/ext4_jbd2.c:924:20: sparse: sparse: context imbalance in 'wait_all_inode_data' - unexpected unlock
vim +1398 fs/ext4/ext4_jbd2.c
1364
1365 static int ext4_journal_fc_replay_scan(journal_t *journal,
1366 struct buffer_head *bh, int off)
1367 {
1368 struct super_block *sb = journal->j_private;
1369 struct ext4_sb_info *sbi = EXT4_SB(sb);
1370 struct ext4_fc_replay_state *state;
1371 struct ext4_fc_commit_hdr *fc_hdr;
1372 __u32 csum, old_csum;
1373 __u8 *start, *end;
1374
1375 state = &sbi->s_fc_replay_state;
1376 fc_hdr = (struct ext4_fc_commit_hdr *)
1377 ((__u8 *)bh->b_data + sizeof(journal_header_t));
1378
1379 start = (u8 *)fc_hdr;
1380 end = (__u8 *)bh->b_data + journal->j_blocksize;
1381
1382 /* Check if we already concluded that this fast commit is not useful */
1383 if (state->fc_replay_expected_off && state->fc_replay_error)
1384 goto out_err;
1385
1386 if (le32_to_cpu(fc_hdr->fc_magic) != EXT4_FC_MAGIC) {
1387 state->fc_replay_error = -ENOENT;
1388 goto out_err;
1389 }
1390
1391 if (off != state->fc_replay_expected_off) {
1392 state->fc_replay_error = -EFSCORRUPTED;
1393 goto out_err;
1394 }
1395
1396 state->fc_replay_expected_off++;
1397
> 1398 if (le16_to_cpu(fc_hdr->fc_features)) {
1399 state->fc_replay_error = -EOPNOTSUPP;
1400 goto out_err;
1401 }
1402
> 1403 old_csum = fc_hdr->fc_csum;
1404 fc_hdr->fc_csum = 0;
1405 csum = ext4_chksum(sbi, 0, start, end - start);
> 1406 fc_hdr->fc_csum = old_csum;
1407
1408 if (csum != le32_to_cpu(fc_hdr->fc_csum)) {
1409 state->fc_replay_error = -EFSBADCRC;
1410 goto out_err;
1411 }
1412
1413 trace_ext4_journal_fc_replay_scan(sb, state->fc_replay_error, off);
1414 return 0;
1415
1416 out_err:
1417 trace_ext4_journal_fc_replay_scan(sb, state->fc_replay_error, off);
1418 return state->fc_replay_error;
1419 }
1420
---
0-DAY kernel test infrastructure Open Source Technology Center
https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org Intel Corporation
^ permalink raw reply [flat|nested] 31+ messages in thread
* [PATCH v4 20/20] ext4: add debug mount option to test fast commit replay
2019-12-24 8:13 [PATCH v4 01/20] ext4: update docs for fast commit feature Harshad Shirwadkar
` (17 preceding siblings ...)
2019-12-24 8:13 ` [PATCH v4 19/20] ext4: add fast commit " Harshad Shirwadkar
@ 2019-12-24 8:13 ` Harshad Shirwadkar
2020-01-09 4:29 ` [PATCH v4 01/20] ext4: update docs for fast commit feature xiaohui li
19 siblings, 0 replies; 31+ messages in thread
From: Harshad Shirwadkar @ 2019-12-24 8:13 UTC (permalink / raw)
To: linux-ext4; +Cc: Harshad Shirwadkar
Add a debug mount option to simulate errors while replaying. If
fc_debug_max_replay is set, ext4 will replay only as many fast
commit blocks as passed as an argument.
Signed-off-by: Harshad Shirwadkar <harshadshirwadkar@gmail.com>
---
fs/ext4/ext4.h | 3 +++
fs/ext4/ext4_jbd2.c | 6 ++++++
fs/ext4/super.c | 12 +++++++++++-
3 files changed, 20 insertions(+), 1 deletion(-)
diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index e12900d77673..62d72e7005ad 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -1652,6 +1652,9 @@ struct ext4_sb_info {
struct list_head s_fc_dentry_q;
struct ext4_fc_replay_state s_fc_replay_state;
spinlock_t s_fc_lock;
+#ifdef EXT4_FC_DEBUG
+ int s_fc_debug_max_replay;
+#endif
struct ext4_fc_stats s_fc_stats;
};
diff --git a/fs/ext4/ext4_jbd2.c b/fs/ext4/ext4_jbd2.c
index ef36a973ed8b..1f8ba23912ba 100644
--- a/fs/ext4/ext4_jbd2.c
+++ b/fs/ext4/ext4_jbd2.c
@@ -1520,6 +1520,12 @@ static int ext4_journal_fc_replay_cb(journal_t *journal, struct buffer_head *bh,
return sbi->s_fc_replay_state.fc_replay_error;
}
+#ifdef EXT4_FC_DEBUG
+ if (sbi->s_fc_debug_max_replay && off >= sbi->s_fc_debug_max_replay) {
+ pr_warn("Dropping fc block %d because max_replay set\n", off);
+ return -EINVAL;
+ }
+#endif
sbi->s_mount_state |= EXT4_FC_REPLAY;
fc_hdr = (struct ext4_fc_commit_hdr *)
((__u8 *)bh->b_data + sizeof(journal_header_t));
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index bfd19a127188..117ccde1a7c1 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -1475,7 +1475,7 @@ enum {
Opt_dioread_nolock, Opt_dioread_lock,
Opt_discard, Opt_nodiscard, Opt_init_itable, Opt_noinit_itable,
Opt_max_dir_size_kb, Opt_nojournal_checksum, Opt_nombcache,
- Opt_no_fc, Opt_fc_soft_consistency
+ Opt_no_fc, Opt_fc_soft_consistency, Opt_fc_debug_max_replay
};
static const match_table_t tokens = {
@@ -1560,6 +1560,9 @@ static const match_table_t tokens = {
{Opt_noinit_itable, "noinit_itable"},
{Opt_no_fc, "no_fc"},
{Opt_fc_soft_consistency, "fc_soft_consistency"},
+#ifdef EXT4_FC_DEBUG
+ {Opt_fc_debug_max_replay, "fc_debug_max_replay=%u"},
+#endif
{Opt_max_dir_size_kb, "max_dir_size_kb=%u"},
{Opt_test_dummy_encryption, "test_dummy_encryption"},
{Opt_nombcache, "nombcache"},
@@ -1779,6 +1782,9 @@ static const struct mount_opts {
MOPT_CLEAR | MOPT_2 | MOPT_EXT4_ONLY},
{Opt_fc_soft_consistency, EXT4_MOUNT2_JOURNAL_FC_SOFT_CONSISTENCY,
MOPT_SET | MOPT_2 | MOPT_EXT4_ONLY},
+#ifdef EXT4_FC_DEBUG
+ {Opt_fc_debug_max_replay, 0, MOPT_GTE0},
+#endif
{Opt_err, 0, 0}
};
@@ -1931,6 +1937,10 @@ static int handle_mount_opt(struct super_block *sb, char *opt, int token,
sbi->s_li_wait_mult = arg;
} else if (token == Opt_max_dir_size_kb) {
sbi->s_max_dir_size_kb = arg;
+#ifdef EXT4_FC_DEBUG
+ } else if (token == Opt_fc_debug_max_replay) {
+ sbi->s_fc_debug_max_replay = arg;
+#endif
} else if (token == Opt_stripe) {
sbi->s_stripe = arg;
} else if (token == Opt_resuid) {
--
2.24.1.735.g03f4e72817-goog
^ permalink raw reply related [flat|nested] 31+ messages in thread* Re: [PATCH v4 01/20] ext4: update docs for fast commit feature
2019-12-24 8:13 [PATCH v4 01/20] ext4: update docs for fast commit feature Harshad Shirwadkar
` (18 preceding siblings ...)
2019-12-24 8:13 ` [PATCH v4 20/20] ext4: add debug mount option to test fast commit replay Harshad Shirwadkar
@ 2020-01-09 4:29 ` xiaohui li
2020-01-10 9:49 ` xiaohui li
2020-01-12 3:45 ` Theodore Y. Ts'o
19 siblings, 2 replies; 31+ messages in thread
From: xiaohui li @ 2020-01-09 4:29 UTC (permalink / raw)
To: Harshad Shirwadkar; +Cc: Ext4 Developers List, Theodore Y. Ts'o
hi Harshad
cc ted
sorry, but i have some idea about this fast commit which i want to
share with you.
there are nearly 20 patches about this v4 fast commit , so many patches.
I wonder if necessary to make this fast commit function so complexly.
maybe i have not understand the difficulty of the fast commit coding work.
so I appreciate it very much if you give some more detailed
descriptions about the patches correlationship of v4 fast commit,
especially the reason why need have so many patches.
from my viewpoint, the purpose of doing this fast commit function is
to resolve the ext4 fsync time-cost-so-much problem.
firstly we need to resolve some actual customer problems which exist
in ext4 filesystems when doing this fast commit function.
so the first release version of fast commit is just only to accomplish
the goal of reducing the time cost of fsync because of jbd2 order
shortcoming described in ijournal paper from my opinion.
it need not do so many other unnecessary things.
if i have free time , I will review these patches continually.
thank you for your reply.
On Tue, Dec 24, 2019 at 4:14 PM Harshad Shirwadkar
<harshadshirwadkar@gmail.com> wrote:
>
> This patch series adds support for fast commits which is a simplified
> version of the scheme proposed by Park and Shin, in their paper,
> "iJournaling: Fine-Grained Journaling for Improving the Latency of
> Fsync System Call"[1]. The basic idea of fast commits is to make JBD2
> give the client file system an opportunity to perform a faster
> commit. Only if the file system cannot perform such a commit
> operation, then JBD2 should fall back to traditional commits.
>
> Because JBD2 operates at block granularity, for every file system
> metadata update it commits all the changed blocks are written to the
> journal at commit time. This is inefficient because updates to some
> blocks that JBD2 commits are derivable from some other blocks. For
> example, if a new extent is added to an inode, then corresponding
> updates to the inode table, the block bitmap, the group descriptor and
> the superblock can be derived based on just the extent information and
> the corresponding inode information. So, if we take this relationship
> between blocks into account and replay the journalled blocks smartly,
> we could increase performance of file system commits significantly.
>
> Fast commits introduced in this patch have two main contributions:
>
> (1) Making JBD2 fast commit aware, so that clients of JBD2 can
> implement fast commits
>
> (2) Add support in ext4 to use JBD2's new interfaces and implement
> fast commits.
>
> Ext4 supports two modes of fast commits: 1) fast commits with hard
> consistency guarantees 2) fast commits with soft consistency guarantees
>
> When hard consistency is enabled, fast commit guarantees that all the
> updates will be committed. After a successful replay of fast commits
> blocks in hard consistency mode, the entire file system would be in
> the same state as that when fsync() returned before crash. This
> guarantee is similar to what jbd2 gives with full commits.
>
> With soft consistency, file system only guarantees consistency for the
> inode in question. In this mode, file system will try to write as less
> data to the backend as possible during the commit time. To be precise,
> file system records all the data updates for the inode in question and
> directory updates that are required for guaranteeing consistency of the
> inode in question.
>
> In our evaluations, fast commits with hard consistency performed
> better than fast commits with soft consistency. That's because with
> hard consistency, a fast commit often ends up committing other inodes
> together, while with soft consistency commits get serialized. Future
> work can look at creating hybrid approach between the two extremes
> that are there in this patchset.
>
> Testing
> -------
>
> e2fsprogs was updated to set fast commit feature flag and to ignore
> fast commit blocks during e2fsck.
>
> https://github.com/harshadjs/e2fsprogs.git
>
> After applying all the patches in this series, following runs of
> xfstests were performed:
>
> - kvm-xfstest.sh -g log -c 4k
> - kvm-xfstests.sh smoke
>
> All the log tests were successful and smoke tests didn't introduce any
> additional failures.
>
> Performance Evaluation
> ----------------------
>
> Ext4 file system performance was tested with full commits, with fast
> commits with soft consistency and with fast commits with hard
> consistency. fs_mark benchmark showed that depending on the file size,
> performance improvement was seen up to 50%. Soft fast commits performed
> slightly worse than hard fast commits. But soft fast commits ended up
> writing slightly lesser number of blocks on disk.
>
> Changes since V3:
>
> - Removed invocation of fast commits from the jbd2 thread.
>
> - Removed sub transaction ID from journal_t.
>
> - Added rename, truncate, punch hole support.
>
> - Added soft consistency mode and hard consistency mode.
>
> - More bug fixes and refactoring.
>
> - Added better debugging support: more tracepoints and debug mount
> options.
>
> Harshad Shirwadkar(20):
> ext4: add debug mount option to test fast commit replay
> ext4: add fast commit replay path
> ext4: disable certain features in replay path
> ext4: add idempotent helpers to manipulate bitmaps
> ext4: fast commit recovery path preparation
> jbd2: add fast commit recovery path support
> ext4: main commit routine for fast commits
> jbd2: add new APIs for commit path of fast commits
> ext4: add fast commit on-disk format structs and helpers
> ext4: add fast commit track points
> ext4: break ext4_unlink() and ext4_link()
> ext4: add inode tracking and ineligible marking routines
> ext4: add directory entry tracking routines
> ext4: add generic diff tracking routines and range tracking
> jbd2: fast commit main commit path changes
> jbd2: disable fast commits if journal is empty
> jbd2: add fast commit block tracker variables
> ext4, jbd2: add fast commit initialization routines
> ext4: add handling for extended mount options
> ext4: update docs for fast commit feature
>
> Documentation/filesystems/ext4/journal.rst | 127 ++-
> Documentation/filesystems/journalling.rst | 18 +
> fs/ext4/acl.c | 1 +
> fs/ext4/balloc.c | 10 +-
> fs/ext4/ext4.h | 127 +++
> fs/ext4/ext4_jbd2.c | 1484 +++++++++++++++++++++++++++-
> fs/ext4/ext4_jbd2.h | 71 ++
> fs/ext4/extents.c | 5 +
> fs/ext4/extents_status.c | 24 +
> fs/ext4/fsync.c | 2 +-
> fs/ext4/ialloc.c | 165 +++-
> fs/ext4/inline.c | 3 +
> fs/ext4/inode.c | 77 +-
> fs/ext4/ioctl.c | 9 +-
> fs/ext4/mballoc.c | 157 ++-
> fs/ext4/mballoc.h | 2 +
> fs/ext4/migrate.c | 1 +
> fs/ext4/namei.c | 172 ++--
> fs/ext4/super.c | 72 +-
> fs/ext4/xattr.c | 6 +
> fs/jbd2/commit.c | 61 ++
> fs/jbd2/journal.c | 217 +++-
> fs/jbd2/recovery.c | 67 +-
> include/linux/jbd2.h | 83 +-
> include/trace/events/ext4.h | 208 +++-
> 25 files changed, 3037 insertions(+), 132 deletions(-)
> ---
> Documentation/filesystems/ext4/journal.rst | 127 ++++++++++++++++++++-
> Documentation/filesystems/journalling.rst | 18 +++
> 2 files changed, 139 insertions(+), 6 deletions(-)
>
> diff --git a/Documentation/filesystems/ext4/journal.rst b/Documentation/filesystems/ext4/journal.rst
> index ea613ee701f5..f94e66f2f8c4 100644
> --- a/Documentation/filesystems/ext4/journal.rst
> +++ b/Documentation/filesystems/ext4/journal.rst
> @@ -29,10 +29,10 @@ safest. If ``data=writeback``, dirty data blocks are not flushed to the
> disk before the metadata are written to disk through the journal.
>
> The journal inode is typically inode 8. The first 68 bytes of the
> -journal inode are replicated in the ext4 superblock. The journal itself
> -is normal (but hidden) file within the filesystem. The file usually
> -consumes an entire block group, though mke2fs tries to put it in the
> -middle of the disk.
> +journal inode are replicated in the ext4 superblock. The journal
> +itself is normal (but hidden) file within the filesystem. The file
> +usually consumes an entire block group, though mke2fs tries to put it
> +in the middle of the disk.
>
> All fields in jbd2 are written to disk in big-endian order. This is the
> opposite of ext4.
> @@ -42,22 +42,74 @@ NOTE: Both ext4 and ocfs2 use jbd2.
> The maximum size of a journal embedded in an ext4 filesystem is 2^32
> blocks. jbd2 itself does not seem to care.
>
> +Fast Commits
> +~~~~~~~~~~~~
> +
> +Ext4 also implements fast commits and integrates it with JBD2 journalling.
> +Fast commits store metadata changes made to the file system as inode level
> +diff. In other words, each fast commit block identifies updates made to
> +a particular inode and collectively they represent total changes made to
> +the file system.
> +
> +A fast commit is valid only if there is no full commit after that particular
> +fast commit. Because of this feature, fast commit blocks can be reused by
> +the following transactions.
> +
> +Each fast commit block stores updates to 1 particular inode. Updates in each
> +fast commit block are one of the 2 types:
> +- Data updates (add range / delete range)
> +- Directory entry updates (Add / remove links)
> +
> +Fast commit blocks must be replayed in the order in which they appear on disk.
> +That's because directory entry updates are written in fast commit blocks
> +in the order in which they are applied on the file system before crash.
> +Changing the order of replaying for directory entry updates may result
> +in inconsistent file system. Note that only directory entry updates need
> +ordering, data updates, since they apply to only one inode, do not require
> +ordered replay. Also, fast commits guarantee that file system is in consistent
> +state after replay of each fast commit block as long as order of replay has
> +been followed.
> +
> +Note that directory inode updates are never directly recorded in fast commits.
> +Just like other file system level metaata, updates to directories are always
> +implied based on directory entry updates stored in fast commit blocks.
> +
> +Based on which directory entry updates are committed with an inode, fast
> +commits have two modes of operation:
> +
> +- Hard Consistency (default)
> +- Soft Consistency (can be enabled by setting mount flag "fc_soft_consistency")
> +
> +When hard consistency is enabled, fast commit guarantees that all the updates
> +will be committed. After a successful replay of fast commits blocks
> +in hard consistency mode, the entire file system would be in the same state as
> +that when fsync() returned before crash. This guarantee is similar to what
> +jbd2 gives.
> +
> +With soft consistency, file system only guarantees consistency for the
> +inode in question. In this mode, file system will try to write as less data
> +to the backed as possible during the commit time. To be precise, file system
> +records all the data updates for the inode in question and directory updates
> +that are required for guaranteeing consistency of the inode in question.
> +
> Layout
> ~~~~~~
>
> Generally speaking, the journal has this format:
>
> .. list-table::
> - :widths: 16 48 16
> + :widths: 16 48 16 18
> :header-rows: 1
>
> * - Superblock
> - descriptor\_block (data\_blocks or revocation\_block) [more data or
> revocations] commmit\_block
> - [more transactions...]
> + - [Fast commits...]
> * -
> - One transaction
> -
> + -
>
> Notice that a transaction begins with either a descriptor and some data,
> or a block revocation list. A finished transaction always ends with a
> @@ -76,7 +128,7 @@ The journal superblock will be in the next full block after the
> superblock.
>
> .. list-table::
> - :widths: 12 12 12 32 12
> + :widths: 12 12 12 32 12 12
> :header-rows: 1
>
> * - 1024 bytes of padding
> @@ -85,11 +137,13 @@ superblock.
> - descriptor\_block (data\_blocks or revocation\_block) [more data or
> revocations] commmit\_block
> - [more transactions...]
> + - [Fast commits...]
> * -
> -
> -
> - One transaction
> -
> + -
>
> Block Header
> ~~~~~~~~~~~~
> @@ -609,3 +663,64 @@ bytes long (but uses a full block):
> - h\_commit\_nsec
> - Nanoseconds component of the above timestamp.
>
> +Fast Commit Block
> +~~~~~~~~~~~~~~~~~
> +
> +The fast commit block indicates an append to the last commit block
> +that was written to the journal. One fast commit block records updates
> +to one inode. So, typically you would find as many fast commit blocks
> +as the number of inodes that got changed since the last commit. A fast
> +commit block is valid only if there is no commit block present with
> +transaction ID greater than that of the fast commit block. If such a
> +block a present, then there is no need to replay the fast commit
> +block.
> +
> +.. list-table::
> + :widths: 8 8 24 40
> + :header-rows: 1
> +
> + * - Offset
> + - Type
> + - Name
> + - Descriptor
> + * - 0x0
> + - journal\_header\_s
> + - (open coded)
> + - Common block header.
> + * - 0xC
> + - \_\_le32
> + - fc\_magic
> + - Magic value which should be set to 0xE2540090. This identifies
> + that this block is a fast commit block.
> + * - 0x10
> + - \_\_u8
> + - fc\_features
> + - Features used by this fast commit block.
> + * - 0x11
> + - \_\_le16
> + - fc_num_tlvs
> + - Number of TLVs contained in this fast commit block
> + * - 0x13
> + - \_\_le32
> + - \_\_fc\_len
> + - Length of the fast commit block in terms of number of blocks
> + * - 0x17
> + - \_\_le32
> + - fc\_ino
> + - Inode number of the inode that will be recovered using this fast commit
> + * - 0x2B
> + - struct ext4\_inode
> + - inode
> + - On-disk copy of the inode at the commit time
> + * - <Variable based on inode size>
> + - struct ext4\_fc\_tl
> + - Array of struct ext4\_fc\_tl
> + - The actual delta with the last commit. Starting at this offset,
> + there is an array of TLVs that indicates which all extents
> + should be present in the corresponding inode. Currently,
> + following tags are supported: EXT4\_FC\_TAG\_EXT (extent that
> + should be present in the inode), EXT4\_FC\_TAG\_HOLE (extent
> + that should be removed from the inode), EXT4\_FC\_TAG\_ADD\_DENTRY
> + (dentry that should be linked), EXT4\_FC\_TAG\_DEL\_DENTRY
> + (dentry that should be unlinked), EXT4\_FC\_TAG\_CREATE\_DENTRY
> + (dentry that for the file that should be created for the first time).
> diff --git a/Documentation/filesystems/journalling.rst b/Documentation/filesystems/journalling.rst
> index 58ce6b395206..1cb116ab27ab 100644
> --- a/Documentation/filesystems/journalling.rst
> +++ b/Documentation/filesystems/journalling.rst
> @@ -115,6 +115,24 @@ called after each transaction commit. You can also use
> ``transaction->t_private_list`` for attaching entries to a transaction
> that need processing when the transaction commits.
>
> +JBD2 also allows client file systems to implement file system specific
> +commits which are called as ``fast commits``. Fast commits are
> +asynchronous in nature i.e. file systems can call their own commit
> +functions at any time. In order to avoid the race with kjournald
> +thread and other possible fast commits that may be happening in
> +parallel, file systems should first call
> +:c:func:`jbd2_start_async_fc()`. File system can call
> +:c:func:`jbd2_map_fc_buf()` to get buffers reserved for fast
> +commits. Once a fast commit is completed, file system should call
> +:c:func:`jbd2_stop_async_fc()` to indicate and unblock other
> +committers and the kjournald thread. After performing either a fast
> +or a full commit, JBD2 calls ``journal->j_fc_cleanup_cb`` to allow
> +file systems to perform cleanups for their internal fast commit
> +related data structures. At the replay time, JBD2 passes each and
> +every fast commit block to the file system via
> +``journal->j_fc_replay_cb``. Ext4 effectively uses this fast commit
> +mechanism to improve journal commit performance.
> +
> JBD2 also provides a way to block all transaction updates via
> :c:func:`jbd2_journal_lock_updates()` /
> :c:func:`jbd2_journal_unlock_updates()`. Ext4 uses this when it wants a
> --
> 2.24.1.735.g03f4e72817-goog
>
^ permalink raw reply [flat|nested] 31+ messages in thread* Re: [PATCH v4 01/20] ext4: update docs for fast commit feature
2020-01-09 4:29 ` [PATCH v4 01/20] ext4: update docs for fast commit feature xiaohui li
@ 2020-01-10 9:49 ` xiaohui li
2020-01-11 2:13 ` harshad shirwadkar
2020-01-12 3:45 ` Theodore Y. Ts'o
1 sibling, 1 reply; 31+ messages in thread
From: xiaohui li @ 2020-01-10 9:49 UTC (permalink / raw)
To: Harshad Shirwadkar; +Cc: Ext4 Developers List, Theodore Y. Ts'o
hi Harshad：
I apologize for my some direct and improper speaking in my last email.
what i want to say in my last email is that maybe an iterative
software development method can be better for patches application.
for the first release version, we can not do everything. it is good
enough if we can have finished just one major function in ext4 fast
commit field.
I have known that develop work of this fast commit function is more
difficult and more complex.
so i am very grateful for your work on field of ext4 fast commit development. :)
On Thu, Jan 9, 2020 at 12:29 PM xiaohui li
<lixiaohui1@xiaomi.corp-partner.google.com> wrote:
>
> hi Harshad
> cc ted
>
> sorry, but i have some idea about this fast commit which i want to
> share with you.
>
> there are nearly 20 patches about this v4 fast commit , so many patches.
> I wonder if necessary to make this fast commit function so complexly.
>
> maybe i have not understand the difficulty of the fast commit coding work.
> so I appreciate it very much if you give some more detailed
> descriptions about the patches correlationship of v4 fast commit,
> especially the reason why need have so many patches.
>
> from my viewpoint, the purpose of doing this fast commit function is
> to resolve the ext4 fsync time-cost-so-much problem.
> firstly we need to resolve some actual customer problems which exist
> in ext4 filesystems when doing this fast commit function.
>
> so the first release version of fast commit is just only to accomplish
> the goal of reducing the time cost of fsync because of jbd2 order
> shortcoming described in ijournal paper from my opinion.
> it need not do so many other unnecessary things.
>
> if i have free time , I will review these patches continually.
> thank you for your reply.
>
>
>
>
>
> On Tue, Dec 24, 2019 at 4:14 PM Harshad Shirwadkar
> <harshadshirwadkar@gmail.com> wrote:
> >
> > This patch series adds support for fast commits which is a simplified
> > version of the scheme proposed by Park and Shin, in their paper,
> > "iJournaling: Fine-Grained Journaling for Improving the Latency of
> > Fsync System Call"[1]. The basic idea of fast commits is to make JBD2
> > give the client file system an opportunity to perform a faster
> > commit. Only if the file system cannot perform such a commit
> > operation, then JBD2 should fall back to traditional commits.
> >
> > Because JBD2 operates at block granularity, for every file system
> > metadata update it commits all the changed blocks are written to the
> > journal at commit time. This is inefficient because updates to some
> > blocks that JBD2 commits are derivable from some other blocks. For
> > example, if a new extent is added to an inode, then corresponding
> > updates to the inode table, the block bitmap, the group descriptor and
> > the superblock can be derived based on just the extent information and
> > the corresponding inode information. So, if we take this relationship
> > between blocks into account and replay the journalled blocks smartly,
> > we could increase performance of file system commits significantly.
> >
> > Fast commits introduced in this patch have two main contributions:
> >
> > (1) Making JBD2 fast commit aware, so that clients of JBD2 can
> > implement fast commits
> >
> > (2) Add support in ext4 to use JBD2's new interfaces and implement
> > fast commits.
> >
> > Ext4 supports two modes of fast commits: 1) fast commits with hard
> > consistency guarantees 2) fast commits with soft consistency guarantees
> >
> > When hard consistency is enabled, fast commit guarantees that all the
> > updates will be committed. After a successful replay of fast commits
> > blocks in hard consistency mode, the entire file system would be in
> > the same state as that when fsync() returned before crash. This
> > guarantee is similar to what jbd2 gives with full commits.
> >
> > With soft consistency, file system only guarantees consistency for the
> > inode in question. In this mode, file system will try to write as less
> > data to the backend as possible during the commit time. To be precise,
> > file system records all the data updates for the inode in question and
> > directory updates that are required for guaranteeing consistency of the
> > inode in question.
> >
> > In our evaluations, fast commits with hard consistency performed
> > better than fast commits with soft consistency. That's because with
> > hard consistency, a fast commit often ends up committing other inodes
> > together, while with soft consistency commits get serialized. Future
> > work can look at creating hybrid approach between the two extremes
> > that are there in this patchset.
> >
> > Testing
> > -------
> >
> > e2fsprogs was updated to set fast commit feature flag and to ignore
> > fast commit blocks during e2fsck.
> >
> > https://github.com/harshadjs/e2fsprogs.git
> >
> > After applying all the patches in this series, following runs of
> > xfstests were performed:
> >
> > - kvm-xfstest.sh -g log -c 4k
> > - kvm-xfstests.sh smoke
> >
> > All the log tests were successful and smoke tests didn't introduce any
> > additional failures.
> >
> > Performance Evaluation
> > ----------------------
> >
> > Ext4 file system performance was tested with full commits, with fast
> > commits with soft consistency and with fast commits with hard
> > consistency. fs_mark benchmark showed that depending on the file size,
> > performance improvement was seen up to 50%. Soft fast commits performed
> > slightly worse than hard fast commits. But soft fast commits ended up
> > writing slightly lesser number of blocks on disk.
> >
> > Changes since V3:
> >
> > - Removed invocation of fast commits from the jbd2 thread.
> >
> > - Removed sub transaction ID from journal_t.
> >
> > - Added rename, truncate, punch hole support.
> >
> > - Added soft consistency mode and hard consistency mode.
> >
> > - More bug fixes and refactoring.
> >
> > - Added better debugging support: more tracepoints and debug mount
> > options.
> >
> > Harshad Shirwadkar(20):
> > ext4: add debug mount option to test fast commit replay
> > ext4: add fast commit replay path
> > ext4: disable certain features in replay path
> > ext4: add idempotent helpers to manipulate bitmaps
> > ext4: fast commit recovery path preparation
> > jbd2: add fast commit recovery path support
> > ext4: main commit routine for fast commits
> > jbd2: add new APIs for commit path of fast commits
> > ext4: add fast commit on-disk format structs and helpers
> > ext4: add fast commit track points
> > ext4: break ext4_unlink() and ext4_link()
> > ext4: add inode tracking and ineligible marking routines
> > ext4: add directory entry tracking routines
> > ext4: add generic diff tracking routines and range tracking
> > jbd2: fast commit main commit path changes
> > jbd2: disable fast commits if journal is empty
> > jbd2: add fast commit block tracker variables
> > ext4, jbd2: add fast commit initialization routines
> > ext4: add handling for extended mount options
> > ext4: update docs for fast commit feature
> >
> > Documentation/filesystems/ext4/journal.rst | 127 ++-
> > Documentation/filesystems/journalling.rst | 18 +
> > fs/ext4/acl.c | 1 +
> > fs/ext4/balloc.c | 10 +-
> > fs/ext4/ext4.h | 127 +++
> > fs/ext4/ext4_jbd2.c | 1484 +++++++++++++++++++++++++++-
> > fs/ext4/ext4_jbd2.h | 71 ++
> > fs/ext4/extents.c | 5 +
> > fs/ext4/extents_status.c | 24 +
> > fs/ext4/fsync.c | 2 +-
> > fs/ext4/ialloc.c | 165 +++-
> > fs/ext4/inline.c | 3 +
> > fs/ext4/inode.c | 77 +-
> > fs/ext4/ioctl.c | 9 +-
> > fs/ext4/mballoc.c | 157 ++-
> > fs/ext4/mballoc.h | 2 +
> > fs/ext4/migrate.c | 1 +
> > fs/ext4/namei.c | 172 ++--
> > fs/ext4/super.c | 72 +-
> > fs/ext4/xattr.c | 6 +
> > fs/jbd2/commit.c | 61 ++
> > fs/jbd2/journal.c | 217 +++-
> > fs/jbd2/recovery.c | 67 +-
> > include/linux/jbd2.h | 83 +-
> > include/trace/events/ext4.h | 208 +++-
> > 25 files changed, 3037 insertions(+), 132 deletions(-)
> > ---
> > Documentation/filesystems/ext4/journal.rst | 127 ++++++++++++++++++++-
> > Documentation/filesystems/journalling.rst | 18 +++
> > 2 files changed, 139 insertions(+), 6 deletions(-)
> >
> > diff --git a/Documentation/filesystems/ext4/journal.rst b/Documentation/filesystems/ext4/journal.rst
> > index ea613ee701f5..f94e66f2f8c4 100644
> > --- a/Documentation/filesystems/ext4/journal.rst
> > +++ b/Documentation/filesystems/ext4/journal.rst
> > @@ -29,10 +29,10 @@ safest. If ``data=writeback``, dirty data blocks are not flushed to the
> > disk before the metadata are written to disk through the journal.
> >
> > The journal inode is typically inode 8. The first 68 bytes of the
> > -journal inode are replicated in the ext4 superblock. The journal itself
> > -is normal (but hidden) file within the filesystem. The file usually
> > -consumes an entire block group, though mke2fs tries to put it in the
> > -middle of the disk.
> > +journal inode are replicated in the ext4 superblock. The journal
> > +itself is normal (but hidden) file within the filesystem. The file
> > +usually consumes an entire block group, though mke2fs tries to put it
> > +in the middle of the disk.
> >
> > All fields in jbd2 are written to disk in big-endian order. This is the
> > opposite of ext4.
> > @@ -42,22 +42,74 @@ NOTE: Both ext4 and ocfs2 use jbd2.
> > The maximum size of a journal embedded in an ext4 filesystem is 2^32
> > blocks. jbd2 itself does not seem to care.
> >
> > +Fast Commits
> > +~~~~~~~~~~~~
> > +
> > +Ext4 also implements fast commits and integrates it with JBD2 journalling.
> > +Fast commits store metadata changes made to the file system as inode level
> > +diff. In other words, each fast commit block identifies updates made to
> > +a particular inode and collectively they represent total changes made to
> > +the file system.
> > +
> > +A fast commit is valid only if there is no full commit after that particular
> > +fast commit. Because of this feature, fast commit blocks can be reused by
> > +the following transactions.
> > +
> > +Each fast commit block stores updates to 1 particular inode. Updates in each
> > +fast commit block are one of the 2 types:
> > +- Data updates (add range / delete range)
> > +- Directory entry updates (Add / remove links)
> > +
> > +Fast commit blocks must be replayed in the order in which they appear on disk.
> > +That's because directory entry updates are written in fast commit blocks
> > +in the order in which they are applied on the file system before crash.
> > +Changing the order of replaying for directory entry updates may result
> > +in inconsistent file system. Note that only directory entry updates need
> > +ordering, data updates, since they apply to only one inode, do not require
> > +ordered replay. Also, fast commits guarantee that file system is in consistent
> > +state after replay of each fast commit block as long as order of replay has
> > +been followed.
> > +
> > +Note that directory inode updates are never directly recorded in fast commits.
> > +Just like other file system level metaata, updates to directories are always
> > +implied based on directory entry updates stored in fast commit blocks.
> > +
> > +Based on which directory entry updates are committed with an inode, fast
> > +commits have two modes of operation:
> > +
> > +- Hard Consistency (default)
> > +- Soft Consistency (can be enabled by setting mount flag "fc_soft_consistency")
> > +
> > +When hard consistency is enabled, fast commit guarantees that all the updates
> > +will be committed. After a successful replay of fast commits blocks
> > +in hard consistency mode, the entire file system would be in the same state as
> > +that when fsync() returned before crash. This guarantee is similar to what
> > +jbd2 gives.
> > +
> > +With soft consistency, file system only guarantees consistency for the
> > +inode in question. In this mode, file system will try to write as less data
> > +to the backed as possible during the commit time. To be precise, file system
> > +records all the data updates for the inode in question and directory updates
> > +that are required for guaranteeing consistency of the inode in question.
> > +
> > Layout
> > ~~~~~~
> >
> > Generally speaking, the journal has this format:
> >
> > .. list-table::
> > - :widths: 16 48 16
> > + :widths: 16 48 16 18
> > :header-rows: 1
> >
> > * - Superblock
> > - descriptor\_block (data\_blocks or revocation\_block) [more data or
> > revocations] commmit\_block
> > - [more transactions...]
> > + - [Fast commits...]
> > * -
> > - One transaction
> > -
> > + -
> >
> > Notice that a transaction begins with either a descriptor and some data,
> > or a block revocation list. A finished transaction always ends with a
> > @@ -76,7 +128,7 @@ The journal superblock will be in the next full block after the
> > superblock.
> >
> > .. list-table::
> > - :widths: 12 12 12 32 12
> > + :widths: 12 12 12 32 12 12
> > :header-rows: 1
> >
> > * - 1024 bytes of padding
> > @@ -85,11 +137,13 @@ superblock.
> > - descriptor\_block (data\_blocks or revocation\_block) [more data or
> > revocations] commmit\_block
> > - [more transactions...]
> > + - [Fast commits...]
> > * -
> > -
> > -
> > - One transaction
> > -
> > + -
> >
> > Block Header
> > ~~~~~~~~~~~~
> > @@ -609,3 +663,64 @@ bytes long (but uses a full block):
> > - h\_commit\_nsec
> > - Nanoseconds component of the above timestamp.
> >
> > +Fast Commit Block
> > +~~~~~~~~~~~~~~~~~
> > +
> > +The fast commit block indicates an append to the last commit block
> > +that was written to the journal. One fast commit block records updates
> > +to one inode. So, typically you would find as many fast commit blocks
> > +as the number of inodes that got changed since the last commit. A fast
> > +commit block is valid only if there is no commit block present with
> > +transaction ID greater than that of the fast commit block. If such a
> > +block a present, then there is no need to replay the fast commit
> > +block.
> > +
> > +.. list-table::
> > + :widths: 8 8 24 40
> > + :header-rows: 1
> > +
> > + * - Offset
> > + - Type
> > + - Name
> > + - Descriptor
> > + * - 0x0
> > + - journal\_header\_s
> > + - (open coded)
> > + - Common block header.
> > + * - 0xC
> > + - \_\_le32
> > + - fc\_magic
> > + - Magic value which should be set to 0xE2540090. This identifies
> > + that this block is a fast commit block.
> > + * - 0x10
> > + - \_\_u8
> > + - fc\_features
> > + - Features used by this fast commit block.
> > + * - 0x11
> > + - \_\_le16
> > + - fc_num_tlvs
> > + - Number of TLVs contained in this fast commit block
> > + * - 0x13
> > + - \_\_le32
> > + - \_\_fc\_len
> > + - Length of the fast commit block in terms of number of blocks
> > + * - 0x17
> > + - \_\_le32
> > + - fc\_ino
> > + - Inode number of the inode that will be recovered using this fast commit
> > + * - 0x2B
> > + - struct ext4\_inode
> > + - inode
> > + - On-disk copy of the inode at the commit time
> > + * - <Variable based on inode size>
> > + - struct ext4\_fc\_tl
> > + - Array of struct ext4\_fc\_tl
> > + - The actual delta with the last commit. Starting at this offset,
> > + there is an array of TLVs that indicates which all extents
> > + should be present in the corresponding inode. Currently,
> > + following tags are supported: EXT4\_FC\_TAG\_EXT (extent that
> > + should be present in the inode), EXT4\_FC\_TAG\_HOLE (extent
> > + that should be removed from the inode), EXT4\_FC\_TAG\_ADD\_DENTRY
> > + (dentry that should be linked), EXT4\_FC\_TAG\_DEL\_DENTRY
> > + (dentry that should be unlinked), EXT4\_FC\_TAG\_CREATE\_DENTRY
> > + (dentry that for the file that should be created for the first time).
> > diff --git a/Documentation/filesystems/journalling.rst b/Documentation/filesystems/journalling.rst
> > index 58ce6b395206..1cb116ab27ab 100644
> > --- a/Documentation/filesystems/journalling.rst
> > +++ b/Documentation/filesystems/journalling.rst
> > @@ -115,6 +115,24 @@ called after each transaction commit. You can also use
> > ``transaction->t_private_list`` for attaching entries to a transaction
> > that need processing when the transaction commits.
> >
> > +JBD2 also allows client file systems to implement file system specific
> > +commits which are called as ``fast commits``. Fast commits are
> > +asynchronous in nature i.e. file systems can call their own commit
> > +functions at any time. In order to avoid the race with kjournald
> > +thread and other possible fast commits that may be happening in
> > +parallel, file systems should first call
> > +:c:func:`jbd2_start_async_fc()`. File system can call
> > +:c:func:`jbd2_map_fc_buf()` to get buffers reserved for fast
> > +commits. Once a fast commit is completed, file system should call
> > +:c:func:`jbd2_stop_async_fc()` to indicate and unblock other
> > +committers and the kjournald thread. After performing either a fast
> > +or a full commit, JBD2 calls ``journal->j_fc_cleanup_cb`` to allow
> > +file systems to perform cleanups for their internal fast commit
> > +related data structures. At the replay time, JBD2 passes each and
> > +every fast commit block to the file system via
> > +``journal->j_fc_replay_cb``. Ext4 effectively uses this fast commit
> > +mechanism to improve journal commit performance.
> > +
> > JBD2 also provides a way to block all transaction updates via
> > :c:func:`jbd2_journal_lock_updates()` /
> > :c:func:`jbd2_journal_unlock_updates()`. Ext4 uses this when it wants a
> > --
> > 2.24.1.735.g03f4e72817-goog
> >
^ permalink raw reply [flat|nested] 31+ messages in thread
* Re: [PATCH v4 01/20] ext4: update docs for fast commit feature
2020-01-10 9:49 ` xiaohui li
@ 2020-01-11 2:13 ` harshad shirwadkar
0 siblings, 0 replies; 31+ messages in thread
From: harshad shirwadkar @ 2020-01-11 2:13 UTC (permalink / raw)
To: xiaohui li; +Cc: Ext4 Developers List, Theodore Y. Ts'o
Hi Xiaohui,
No worries. I think the reason we have so many patches is that this
feature requires on-disk format change and it's better to get
everything in one go, so that the size of testing matrix doesn't
become too big :)
- Harshad
On Fri, Jan 10, 2020 at 1:50 AM xiaohui li
<lixiaohui1@xiaomi.corp-partner.google.com> wrote:
>
> hi Harshad：
>
> I apologize for my some direct and improper speaking in my last email.
>
> what i want to say in my last email is that maybe an iterative
> software development method can be better for patches application.
> for the first release version, we can not do everything. it is good
> enough if we can have finished just one major function in ext4 fast
> commit field.
>
> I have known that develop work of this fast commit function is more
> difficult and more complex.
> so i am very grateful for your work on field of ext4 fast commit development. :)
>
> On Thu, Jan 9, 2020 at 12:29 PM xiaohui li
> <lixiaohui1@xiaomi.corp-partner.google.com> wrote:
> >
> > hi Harshad
> > cc ted
> >
> > sorry, but i have some idea about this fast commit which i want to
> > share with you.
> >
> > there are nearly 20 patches about this v4 fast commit , so many patches.
> > I wonder if necessary to make this fast commit function so complexly.
> >
> > maybe i have not understand the difficulty of the fast commit coding work.
> > so I appreciate it very much if you give some more detailed
> > descriptions about the patches correlationship of v4 fast commit,
> > especially the reason why need have so many patches.
> >
> > from my viewpoint, the purpose of doing this fast commit function is
> > to resolve the ext4 fsync time-cost-so-much problem.
> > firstly we need to resolve some actual customer problems which exist
> > in ext4 filesystems when doing this fast commit function.
> >
> > so the first release version of fast commit is just only to accomplish
> > the goal of reducing the time cost of fsync because of jbd2 order
> > shortcoming described in ijournal paper from my opinion.
> > it need not do so many other unnecessary things.
> >
> > if i have free time , I will review these patches continually.
> > thank you for your reply.
> >
> >
> >
> >
> >
> > On Tue, Dec 24, 2019 at 4:14 PM Harshad Shirwadkar
> > <harshadshirwadkar@gmail.com> wrote:
> > >
> > > This patch series adds support for fast commits which is a simplified
> > > version of the scheme proposed by Park and Shin, in their paper,
> > > "iJournaling: Fine-Grained Journaling for Improving the Latency of
> > > Fsync System Call"[1]. The basic idea of fast commits is to make JBD2
> > > give the client file system an opportunity to perform a faster
> > > commit. Only if the file system cannot perform such a commit
> > > operation, then JBD2 should fall back to traditional commits.
> > >
> > > Because JBD2 operates at block granularity, for every file system
> > > metadata update it commits all the changed blocks are written to the
> > > journal at commit time. This is inefficient because updates to some
> > > blocks that JBD2 commits are derivable from some other blocks. For
> > > example, if a new extent is added to an inode, then corresponding
> > > updates to the inode table, the block bitmap, the group descriptor and
> > > the superblock can be derived based on just the extent information and
> > > the corresponding inode information. So, if we take this relationship
> > > between blocks into account and replay the journalled blocks smartly,
> > > we could increase performance of file system commits significantly.
> > >
> > > Fast commits introduced in this patch have two main contributions:
> > >
> > > (1) Making JBD2 fast commit aware, so that clients of JBD2 can
> > > implement fast commits
> > >
> > > (2) Add support in ext4 to use JBD2's new interfaces and implement
> > > fast commits.
> > >
> > > Ext4 supports two modes of fast commits: 1) fast commits with hard
> > > consistency guarantees 2) fast commits with soft consistency guarantees
> > >
> > > When hard consistency is enabled, fast commit guarantees that all the
> > > updates will be committed. After a successful replay of fast commits
> > > blocks in hard consistency mode, the entire file system would be in
> > > the same state as that when fsync() returned before crash. This
> > > guarantee is similar to what jbd2 gives with full commits.
> > >
> > > With soft consistency, file system only guarantees consistency for the
> > > inode in question. In this mode, file system will try to write as less
> > > data to the backend as possible during the commit time. To be precise,
> > > file system records all the data updates for the inode in question and
> > > directory updates that are required for guaranteeing consistency of the
> > > inode in question.
> > >
> > > In our evaluations, fast commits with hard consistency performed
> > > better than fast commits with soft consistency. That's because with
> > > hard consistency, a fast commit often ends up committing other inodes
> > > together, while with soft consistency commits get serialized. Future
> > > work can look at creating hybrid approach between the two extremes
> > > that are there in this patchset.
> > >
> > > Testing
> > > -------
> > >
> > > e2fsprogs was updated to set fast commit feature flag and to ignore
> > > fast commit blocks during e2fsck.
> > >
> > > https://github.com/harshadjs/e2fsprogs.git
> > >
> > > After applying all the patches in this series, following runs of
> > > xfstests were performed:
> > >
> > > - kvm-xfstest.sh -g log -c 4k
> > > - kvm-xfstests.sh smoke
> > >
> > > All the log tests were successful and smoke tests didn't introduce any
> > > additional failures.
> > >
> > > Performance Evaluation
> > > ----------------------
> > >
> > > Ext4 file system performance was tested with full commits, with fast
> > > commits with soft consistency and with fast commits with hard
> > > consistency. fs_mark benchmark showed that depending on the file size,
> > > performance improvement was seen up to 50%. Soft fast commits performed
> > > slightly worse than hard fast commits. But soft fast commits ended up
> > > writing slightly lesser number of blocks on disk.
> > >
> > > Changes since V3:
> > >
> > > - Removed invocation of fast commits from the jbd2 thread.
> > >
> > > - Removed sub transaction ID from journal_t.
> > >
> > > - Added rename, truncate, punch hole support.
> > >
> > > - Added soft consistency mode and hard consistency mode.
> > >
> > > - More bug fixes and refactoring.
> > >
> > > - Added better debugging support: more tracepoints and debug mount
> > > options.
> > >
> > > Harshad Shirwadkar(20):
> > > ext4: add debug mount option to test fast commit replay
> > > ext4: add fast commit replay path
> > > ext4: disable certain features in replay path
> > > ext4: add idempotent helpers to manipulate bitmaps
> > > ext4: fast commit recovery path preparation
> > > jbd2: add fast commit recovery path support
> > > ext4: main commit routine for fast commits
> > > jbd2: add new APIs for commit path of fast commits
> > > ext4: add fast commit on-disk format structs and helpers
> > > ext4: add fast commit track points
> > > ext4: break ext4_unlink() and ext4_link()
> > > ext4: add inode tracking and ineligible marking routines
> > > ext4: add directory entry tracking routines
> > > ext4: add generic diff tracking routines and range tracking
> > > jbd2: fast commit main commit path changes
> > > jbd2: disable fast commits if journal is empty
> > > jbd2: add fast commit block tracker variables
> > > ext4, jbd2: add fast commit initialization routines
> > > ext4: add handling for extended mount options
> > > ext4: update docs for fast commit feature
> > >
> > > Documentation/filesystems/ext4/journal.rst | 127 ++-
> > > Documentation/filesystems/journalling.rst | 18 +
> > > fs/ext4/acl.c | 1 +
> > > fs/ext4/balloc.c | 10 +-
> > > fs/ext4/ext4.h | 127 +++
> > > fs/ext4/ext4_jbd2.c | 1484 +++++++++++++++++++++++++++-
> > > fs/ext4/ext4_jbd2.h | 71 ++
> > > fs/ext4/extents.c | 5 +
> > > fs/ext4/extents_status.c | 24 +
> > > fs/ext4/fsync.c | 2 +-
> > > fs/ext4/ialloc.c | 165 +++-
> > > fs/ext4/inline.c | 3 +
> > > fs/ext4/inode.c | 77 +-
> > > fs/ext4/ioctl.c | 9 +-
> > > fs/ext4/mballoc.c | 157 ++-
> > > fs/ext4/mballoc.h | 2 +
> > > fs/ext4/migrate.c | 1 +
> > > fs/ext4/namei.c | 172 ++--
> > > fs/ext4/super.c | 72 +-
> > > fs/ext4/xattr.c | 6 +
> > > fs/jbd2/commit.c | 61 ++
> > > fs/jbd2/journal.c | 217 +++-
> > > fs/jbd2/recovery.c | 67 +-
> > > include/linux/jbd2.h | 83 +-
> > > include/trace/events/ext4.h | 208 +++-
> > > 25 files changed, 3037 insertions(+), 132 deletions(-)
> > > ---
> > > Documentation/filesystems/ext4/journal.rst | 127 ++++++++++++++++++++-
> > > Documentation/filesystems/journalling.rst | 18 +++
> > > 2 files changed, 139 insertions(+), 6 deletions(-)
> > >
> > > diff --git a/Documentation/filesystems/ext4/journal.rst b/Documentation/filesystems/ext4/journal.rst
> > > index ea613ee701f5..f94e66f2f8c4 100644
> > > --- a/Documentation/filesystems/ext4/journal.rst
> > > +++ b/Documentation/filesystems/ext4/journal.rst
> > > @@ -29,10 +29,10 @@ safest. If ``data=writeback``, dirty data blocks are not flushed to the
> > > disk before the metadata are written to disk through the journal.
> > >
> > > The journal inode is typically inode 8. The first 68 bytes of the
> > > -journal inode are replicated in the ext4 superblock. The journal itself
> > > -is normal (but hidden) file within the filesystem. The file usually
> > > -consumes an entire block group, though mke2fs tries to put it in the
> > > -middle of the disk.
> > > +journal inode are replicated in the ext4 superblock. The journal
> > > +itself is normal (but hidden) file within the filesystem. The file
> > > +usually consumes an entire block group, though mke2fs tries to put it
> > > +in the middle of the disk.
> > >
> > > All fields in jbd2 are written to disk in big-endian order. This is the
> > > opposite of ext4.
> > > @@ -42,22 +42,74 @@ NOTE: Both ext4 and ocfs2 use jbd2.
> > > The maximum size of a journal embedded in an ext4 filesystem is 2^32
> > > blocks. jbd2 itself does not seem to care.
> > >
> > > +Fast Commits
> > > +~~~~~~~~~~~~
> > > +
> > > +Ext4 also implements fast commits and integrates it with JBD2 journalling.
> > > +Fast commits store metadata changes made to the file system as inode level
> > > +diff. In other words, each fast commit block identifies updates made to
> > > +a particular inode and collectively they represent total changes made to
> > > +the file system.
> > > +
> > > +A fast commit is valid only if there is no full commit after that particular
> > > +fast commit. Because of this feature, fast commit blocks can be reused by
> > > +the following transactions.
> > > +
> > > +Each fast commit block stores updates to 1 particular inode. Updates in each
> > > +fast commit block are one of the 2 types:
> > > +- Data updates (add range / delete range)
> > > +- Directory entry updates (Add / remove links)
> > > +
> > > +Fast commit blocks must be replayed in the order in which they appear on disk.
> > > +That's because directory entry updates are written in fast commit blocks
> > > +in the order in which they are applied on the file system before crash.
> > > +Changing the order of replaying for directory entry updates may result
> > > +in inconsistent file system. Note that only directory entry updates need
> > > +ordering, data updates, since they apply to only one inode, do not require
> > > +ordered replay. Also, fast commits guarantee that file system is in consistent
> > > +state after replay of each fast commit block as long as order of replay has
> > > +been followed.
> > > +
> > > +Note that directory inode updates are never directly recorded in fast commits.
> > > +Just like other file system level metaata, updates to directories are always
> > > +implied based on directory entry updates stored in fast commit blocks.
> > > +
> > > +Based on which directory entry updates are committed with an inode, fast
> > > +commits have two modes of operation:
> > > +
> > > +- Hard Consistency (default)
> > > +- Soft Consistency (can be enabled by setting mount flag "fc_soft_consistency")
> > > +
> > > +When hard consistency is enabled, fast commit guarantees that all the updates
> > > +will be committed. After a successful replay of fast commits blocks
> > > +in hard consistency mode, the entire file system would be in the same state as
> > > +that when fsync() returned before crash. This guarantee is similar to what
> > > +jbd2 gives.
> > > +
> > > +With soft consistency, file system only guarantees consistency for the
> > > +inode in question. In this mode, file system will try to write as less data
> > > +to the backed as possible during the commit time. To be precise, file system
> > > +records all the data updates for the inode in question and directory updates
> > > +that are required for guaranteeing consistency of the inode in question.
> > > +
> > > Layout
> > > ~~~~~~
> > >
> > > Generally speaking, the journal has this format:
> > >
> > > .. list-table::
> > > - :widths: 16 48 16
> > > + :widths: 16 48 16 18
> > > :header-rows: 1
> > >
> > > * - Superblock
> > > - descriptor\_block (data\_blocks or revocation\_block) [more data or
> > > revocations] commmit\_block
> > > - [more transactions...]
> > > + - [Fast commits...]
> > > * -
> > > - One transaction
> > > -
> > > + -
> > >
> > > Notice that a transaction begins with either a descriptor and some data,
> > > or a block revocation list. A finished transaction always ends with a
> > > @@ -76,7 +128,7 @@ The journal superblock will be in the next full block after the
> > > superblock.
> > >
> > > .. list-table::
> > > - :widths: 12 12 12 32 12
> > > + :widths: 12 12 12 32 12 12
> > > :header-rows: 1
> > >
> > > * - 1024 bytes of padding
> > > @@ -85,11 +137,13 @@ superblock.
> > > - descriptor\_block (data\_blocks or revocation\_block) [more data or
> > > revocations] commmit\_block
> > > - [more transactions...]
> > > + - [Fast commits...]
> > > * -
> > > -
> > > -
> > > - One transaction
> > > -
> > > + -
> > >
> > > Block Header
> > > ~~~~~~~~~~~~
> > > @@ -609,3 +663,64 @@ bytes long (but uses a full block):
> > > - h\_commit\_nsec
> > > - Nanoseconds component of the above timestamp.
> > >
> > > +Fast Commit Block
> > > +~~~~~~~~~~~~~~~~~
> > > +
> > > +The fast commit block indicates an append to the last commit block
> > > +that was written to the journal. One fast commit block records updates
> > > +to one inode. So, typically you would find as many fast commit blocks
> > > +as the number of inodes that got changed since the last commit. A fast
> > > +commit block is valid only if there is no commit block present with
> > > +transaction ID greater than that of the fast commit block. If such a
> > > +block a present, then there is no need to replay the fast commit
> > > +block.
> > > +
> > > +.. list-table::
> > > + :widths: 8 8 24 40
> > > + :header-rows: 1
> > > +
> > > + * - Offset
> > > + - Type
> > > + - Name
> > > + - Descriptor
> > > + * - 0x0
> > > + - journal\_header\_s
> > > + - (open coded)
> > > + - Common block header.
> > > + * - 0xC
> > > + - \_\_le32
> > > + - fc\_magic
> > > + - Magic value which should be set to 0xE2540090. This identifies
> > > + that this block is a fast commit block.
> > > + * - 0x10
> > > + - \_\_u8
> > > + - fc\_features
> > > + - Features used by this fast commit block.
> > > + * - 0x11
> > > + - \_\_le16
> > > + - fc_num_tlvs
> > > + - Number of TLVs contained in this fast commit block
> > > + * - 0x13
> > > + - \_\_le32
> > > + - \_\_fc\_len
> > > + - Length of the fast commit block in terms of number of blocks
> > > + * - 0x17
> > > + - \_\_le32
> > > + - fc\_ino
> > > + - Inode number of the inode that will be recovered using this fast commit
> > > + * - 0x2B
> > > + - struct ext4\_inode
> > > + - inode
> > > + - On-disk copy of the inode at the commit time
> > > + * - <Variable based on inode size>
> > > + - struct ext4\_fc\_tl
> > > + - Array of struct ext4\_fc\_tl
> > > + - The actual delta with the last commit. Starting at this offset,
> > > + there is an array of TLVs that indicates which all extents
> > > + should be present in the corresponding inode. Currently,
> > > + following tags are supported: EXT4\_FC\_TAG\_EXT (extent that
> > > + should be present in the inode), EXT4\_FC\_TAG\_HOLE (extent
> > > + that should be removed from the inode), EXT4\_FC\_TAG\_ADD\_DENTRY
> > > + (dentry that should be linked), EXT4\_FC\_TAG\_DEL\_DENTRY
> > > + (dentry that should be unlinked), EXT4\_FC\_TAG\_CREATE\_DENTRY
> > > + (dentry that for the file that should be created for the first time).
> > > diff --git a/Documentation/filesystems/journalling.rst b/Documentation/filesystems/journalling.rst
> > > index 58ce6b395206..1cb116ab27ab 100644
> > > --- a/Documentation/filesystems/journalling.rst
> > > +++ b/Documentation/filesystems/journalling.rst
> > > @@ -115,6 +115,24 @@ called after each transaction commit. You can also use
> > > ``transaction->t_private_list`` for attaching entries to a transaction
> > > that need processing when the transaction commits.
> > >
> > > +JBD2 also allows client file systems to implement file system specific
> > > +commits which are called as ``fast commits``. Fast commits are
> > > +asynchronous in nature i.e. file systems can call their own commit
> > > +functions at any time. In order to avoid the race with kjournald
> > > +thread and other possible fast commits that may be happening in
> > > +parallel, file systems should first call
> > > +:c:func:`jbd2_start_async_fc()`. File system can call
> > > +:c:func:`jbd2_map_fc_buf()` to get buffers reserved for fast
> > > +commits. Once a fast commit is completed, file system should call
> > > +:c:func:`jbd2_stop_async_fc()` to indicate and unblock other
> > > +committers and the kjournald thread. After performing either a fast
> > > +or a full commit, JBD2 calls ``journal->j_fc_cleanup_cb`` to allow
> > > +file systems to perform cleanups for their internal fast commit
> > > +related data structures. At the replay time, JBD2 passes each and
> > > +every fast commit block to the file system via
> > > +``journal->j_fc_replay_cb``. Ext4 effectively uses this fast commit
> > > +mechanism to improve journal commit performance.
> > > +
> > > JBD2 also provides a way to block all transaction updates via
> > > :c:func:`jbd2_journal_lock_updates()` /
> > > :c:func:`jbd2_journal_unlock_updates()`. Ext4 uses this when it wants a
> > > --
> > > 2.24.1.735.g03f4e72817-goog
> > >
^ permalink raw reply [flat|nested] 31+ messages in thread
* Re: [PATCH v4 01/20] ext4: update docs for fast commit feature
2020-01-09 4:29 ` [PATCH v4 01/20] ext4: update docs for fast commit feature xiaohui li
2020-01-10 9:49 ` xiaohui li
@ 2020-01-12 3:45 ` Theodore Y. Ts'o
1 sibling, 0 replies; 31+ messages in thread
From: Theodore Y. Ts'o @ 2020-01-12 3:45 UTC (permalink / raw)
To: xiaohui li; +Cc: Harshad Shirwadkar, Ext4 Developers List
On Thu, Jan 09, 2020 at 12:29:01PM +0800, xiaohui li wrote:
> maybe i have not understand the difficulty of the fast commit coding work.
> so I appreciate it very much if you give some more detailed
> descriptions about the patches correlationship of v4 fast commit,
> especially the reason why need have so many patches.
>
> from my viewpoint, the purpose of doing this fast commit function is
> to resolve the ext4 fsync time-cost-so-much problem.
> firstly we need to resolve some actual customer problems which exist
> in ext4 filesystems when doing this fast commit function.
>
> so the first release version of fast commit is just only to accomplish
> the goal of reducing the time cost of fsync because of jbd2 order
> shortcoming described in ijournal paper from my opinion.
> it need not do so many other unnecessary things.
As Harshad has mentioned, one of the reasons why an incremental
approach does not make sense is that once we release a version of fast
commit into a mainline kernel, we have to worry about what happens if
users start trying to use it, and we have to provide backwards
compatibility for it. So if we were to break up fast commit into 5
parts, then we would have to allocate 5 feature bits, and we would
have to support each version of fast commit --- essentially forever.
As far as why are we doing this, we absolutely have a specific use
case in mind, and that's to improve ext4's performance when used on a
NFS server. The NFS protocol requires that any file system operation
requested by a client is persisted before the server sends an
acknowledgement back to the client. For the workloads that are heavy
with metadata updates, avoiding the need to do a full jbd2 commit for
every NFS RPC request which modifies metadata will a big difference to
the NFS server's performance.
This is why we are interested in making things like renames to be fast
commit eligible, and not just the smaller set of system calls needed
by (for example) SQLite.
Regards,
- Ted
^ permalink raw reply [flat|nested] 31+ messages in thread