On 01/11/2024 06:50, Ritesh Harjani (IBM) wrote:
This patch adds base support for atomic writes via statx getattr.
On bs < ps systems, we can create FS with say bs of 16k. That means
both atomic write min and max unit can be set to 16k for supporting
atomic writes.
Co-developed-by: Ojaswin Mujoo <ojaswin@xxxxxxxxxxxxx>
Signed-off-by: Ojaswin Mujoo <ojaswin@xxxxxxxxxxxxx>
Signed-off-by: Ritesh Harjani (IBM) <ritesh.list@xxxxxxxxx>
Regardless of nitpicks:
Reviewed-by: John Garry <john.g.garry@xxxxxxxxxx>
---
fs/ext4/ext4.h | 10 ++++++++++
fs/ext4/inode.c | 12 ++++++++++++
fs/ext4/super.c | 31 +++++++++++++++++++++++++++++++
3 files changed, 53 insertions(+)
diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index 44b0d418143c..494d443e9fc9 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -1729,6 +1729,10 @@ struct ext4_sb_info {
*/
struct work_struct s_sb_upd_work;
+ /* Atomic write unit values in bytes */
+ unsigned int s_awu_min;
+ unsigned int s_awu_max;
+
/* Ext4 fast commit sub transaction ID */
atomic_t s_fc_subtid;
@@ -3855,6 +3859,12 @@ static inline int ext4_buffer_uptodate(struct buffer_head *bh)
return buffer_uptodate(bh);
}
+static inline bool ext4_inode_can_atomic_write(struct inode *inode)
+{
+
nit: superfluous blank line
+ return S_ISREG(inode->i_mode) && EXT4_SB(inode->i_sb)->s_awu_min > 0;
I am not sure if the S_ISREG() check is required. Other callers also do
the check (like ext4_getattr() for when calling
ext4_inode_can_atomic_write()) or don't need it (ext4_file_open()). I
say ext4_file_open() doesn't need it as ext4_file_open() is only ever
called for regular files, right?
+}
+
extern int ext4_block_write_begin(handle_t *handle, struct folio *folio,
loff_t pos, unsigned len,
get_block_t *get_block);
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 54bdd4884fe6..3e827cfa762e 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -5578,6 +5578,18 @@ int ext4_getattr(struct mnt_idmap *idmap, const struct path *path,
}
}
+ if ((request_mask & STATX_WRITE_ATOMIC) && S_ISREG(inode->i_mode)) {
nit: maybe you could have factored out the S_ISREG() check with
STATX_DIOALIGN
+ struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
+ unsigned int awu_min = 0, awu_max = 0;
+
+ if (ext4_inode_can_atomic_write(inode)) {
+ awu_min = sbi->s_awu_min;
+ awu_max = sbi->s_awu_max;
+ }
+
+ generic_fill_statx_atomic_writes(stat, awu_min, awu_max);
+ }
+
flags = ei->i_flags & EXT4_FL_USER_VISIBLE;
if (flags & EXT4_APPEND_FL)
stat->attributes |= STATX_ATTR_APPEND;
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index 16a4ce704460..ebe1660bd840 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -4425,6 +4425,36 @@ static int ext4_handle_clustersize(struct super_block *sb)
return 0;
}
+/*
+ * ext4_atomic_write_init: Initializes filesystem min & max atomic write units.
+ * @sb: super block
+ * TODO: Later add support for bigalloc
+ */
+static void ext4_atomic_write_init(struct super_block *sb)
+{
+ struct ext4_sb_info *sbi = EXT4_SB(sb);
+ struct block_device *bdev = sb->s_bdev;
+
+ if (!bdev_can_atomic_write(bdev))
+ return;
+
+ if (!ext4_has_feature_extents(sb))
+ return;
+
+ sbi->s_awu_min = max(sb->s_blocksize,
+ bdev_atomic_write_unit_min_bytes(bdev));
+ sbi->s_awu_max = min(sb->s_blocksize,
+ bdev_atomic_write_unit_max_bytes(bdev));
+ if (sbi->s_awu_min && sbi->s_awu_max &&
+ sbi->s_awu_min <= sbi->s_awu_max) {
+ ext4_msg(sb, KERN_NOTICE, "Supports (experimental) DIO atomic writes awu_min: %u, awu_max: %u",
+ sbi->s_awu_min, sbi->s_awu_max);
+ } else {
+ sbi->s_awu_min = 0;
+ sbi->s_awu_max = 0;
+ }
+}
+
static void ext4_fast_commit_init(struct super_block *sb)
{
struct ext4_sb_info *sbi = EXT4_SB(sb);
@@ -5336,6 +5366,7 @@ static int __ext4_fill_super(struct fs_context *fc, struct super_block *sb)
spin_lock_init(&sbi->s_bdev_wb_lock);
+ ext4_atomic_write_init(sb);
ext4_fast_commit_init(sb);
sb->s_root = NULL;