Re: [f2fs-dev] [PATCH] f2fs: change i_compr_blocks of inode to atomic value

From: Daeho Jeong
Date: Wed Sep 02 2020 - 21:43:36 EST


And how about this atomic64 to just atomic?
As you mentioned before, we don't need a 64bit value for
i_compr_blocks of inode.

2020년 9월 3일 (목) 오전 10:08, Chao Yu <yuchao0@xxxxxxxxxx>님이 작성:
>
> On 2020/9/2 14:29, Daeho Jeong wrote:
> > From: Daeho Jeong <daehojeong@xxxxxxxxxx>
> >
> > writepages() can be concurrently invoked for the same file by different
> > threads such as a thread fsyncing the file and a kworker kernel thread.
> > So, changing i_compr_blocks without protection is racy and we need to
> > protect it by changing it with atomic type value.
> >
> > Signed-off-by: Daeho Jeong <daehojeong@xxxxxxxxxx>
> > ---
> > fs/f2fs/f2fs.h | 15 +++++++++------
> > fs/f2fs/file.c | 18 ++++++++++--------
> > fs/f2fs/inode.c | 11 +++++++----
> > fs/f2fs/super.c | 1 +
> > 4 files changed, 27 insertions(+), 18 deletions(-)
> >
> > diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h
> > index f60414805e05..633957e14944 100644
> > --- a/fs/f2fs/f2fs.h
> > +++ b/fs/f2fs/f2fs.h
> > @@ -801,7 +801,7 @@ struct f2fs_inode_info {
> > struct timespec64 i_disk_time[4];/* inode disk times */
> >
> > /* for file compress */
> > - u64 i_compr_blocks; /* # of compressed blocks */
> > + atomic64_t i_compr_blocks; /* # of compressed blocks */
> > unsigned char i_compress_algorithm; /* algorithm type */
> > unsigned char i_log_cluster_size; /* log of cluster size */
> > unsigned int i_cluster_size; /* cluster size */
> > @@ -3933,14 +3933,16 @@ static inline void set_compress_context(struct inode *inode)
> > static inline u64 f2fs_disable_compressed_file(struct inode *inode)
> > {
> > struct f2fs_inode_info *fi = F2FS_I(inode);
> > + u64 i_compr_blocks;
> >
> > if (!f2fs_compressed_file(inode))
> > return 0;
> > if (S_ISREG(inode->i_mode)) {
> > if (get_dirty_pages(inode))
> > return 1;
> > - if (fi->i_compr_blocks)
> > - return fi->i_compr_blocks;
> Caller doesn't care how many compr_blocks this inode has saved, so we can just
> return 1 here.
>
> > + i_compr_blocks = atomic64_read(&fi->i_compr_blocks);
> > + if (i_compr_blocks)
> > + return i_compr_blocks;
> > }
> >
> > fi->i_flags &= ~F2FS_COMPR_FL;
> > @@ -4057,16 +4059,17 @@ static inline void f2fs_i_compr_blocks_update(struct inode *inode,
> > u64 blocks, bool add)
> > {
> > int diff = F2FS_I(inode)->i_cluster_size - blocks;
> > + struct f2fs_inode_info *fi = F2FS_I(inode);
> >
> > /* don't update i_compr_blocks if saved blocks were released */
> > - if (!add && !F2FS_I(inode)->i_compr_blocks)
> > + if (!add && !atomic64_read(&fi->i_compr_blocks))
> > return;
> >
> > if (add) {
> > - F2FS_I(inode)->i_compr_blocks += diff;
> > + atomic64_add(diff, &fi->i_compr_blocks);
> > stat_add_compr_blocks(inode, diff);
> > } else {
> > - F2FS_I(inode)->i_compr_blocks -= diff;
> > + atomic64_sub(diff, &fi->i_compr_blocks);
> > stat_sub_compr_blocks(inode, diff);
> > }
> > f2fs_mark_inode_dirty_sync(inode, true);
> > diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c
> > index cc7f5670390f..4e4e369b97e5 100644
> > --- a/fs/f2fs/file.c
> > +++ b/fs/f2fs/file.c
> > @@ -564,7 +564,7 @@ void f2fs_truncate_data_blocks_range(struct dnode_of_data *dn, int count)
> > bool compressed_cluster = false;
> > int cluster_index = 0, valid_blocks = 0;
> > int cluster_size = F2FS_I(dn->inode)->i_cluster_size;
> > - bool released = !F2FS_I(dn->inode)->i_compr_blocks;
> > + bool released = !atomic64_read(&F2FS_I(dn->inode)->i_compr_blocks);
> >
> > if (IS_INODE(dn->node_page) && f2fs_has_extra_attr(dn->inode))
> > base = get_extra_isize(dn->inode);
> > @@ -3436,7 +3436,7 @@ static int f2fs_get_compress_blocks(struct file *filp, unsigned long arg)
> > if (!f2fs_compressed_file(inode))
> > return -EINVAL;
> >
> > - blocks = F2FS_I(inode)->i_compr_blocks;
> > + blocks = atomic64_read(&F2FS_I(inode)->i_compr_blocks);
> > return put_user(blocks, (u64 __user *)arg);
> > }
> >
> > @@ -3535,7 +3535,7 @@ static int f2fs_release_compress_blocks(struct file *filp, unsigned long arg)
> > if (ret)
> > goto out;
> >
> > - if (!F2FS_I(inode)->i_compr_blocks)
> > + if (!atomic64_read(&F2FS_I(inode)->i_compr_blocks))
> > goto out;
> >
> > F2FS_I(inode)->i_flags |= F2FS_IMMUTABLE_FL;
> > @@ -3588,14 +3588,15 @@ static int f2fs_release_compress_blocks(struct file *filp, unsigned long arg)
> >
> > if (ret >= 0) {
> > ret = put_user(released_blocks, (u64 __user *)arg);
> > - } else if (released_blocks && F2FS_I(inode)->i_compr_blocks) {
> > + } else if (released_blocks &&
> > + atomic64_read(&F2FS_I(inode)->i_compr_blocks)) {
> > set_sbi_flag(sbi, SBI_NEED_FSCK);
> > f2fs_warn(sbi, "%s: partial blocks were released i_ino=%lx "
> > "iblocks=%llu, released=%u, compr_blocks=%llu, "
> > "run fsck to fix.",
> > __func__, inode->i_ino, inode->i_blocks,
> > released_blocks,
> > - F2FS_I(inode)->i_compr_blocks);
> > + atomic64_read(&F2FS_I(inode)->i_compr_blocks));
> > }
> >
> > return ret;
> > @@ -3683,7 +3684,7 @@ static int f2fs_reserve_compress_blocks(struct file *filp, unsigned long arg)
> > if (ret)
> > return ret;
> >
> > - if (F2FS_I(inode)->i_compr_blocks)
> > + if (atomic64_read(&F2FS_I(inode)->i_compr_blocks))
> > goto out;
> >
> > f2fs_balance_fs(F2FS_I_SB(inode), true);
> > @@ -3747,14 +3748,15 @@ static int f2fs_reserve_compress_blocks(struct file *filp, unsigned long arg)
> >
> > if (ret >= 0) {
> > ret = put_user(reserved_blocks, (u64 __user *)arg);
> > - } else if (reserved_blocks && F2FS_I(inode)->i_compr_blocks) {
> > + } else if (reserved_blocks &&
> > + atomic64_read(&F2FS_I(inode)->i_compr_blocks)) {
> > set_sbi_flag(sbi, SBI_NEED_FSCK);
> > f2fs_warn(sbi, "%s: partial blocks were released i_ino=%lx "
> > "iblocks=%llu, reserved=%u, compr_blocks=%llu, "
> > "run fsck to fix.",
> > __func__, inode->i_ino, inode->i_blocks,
> > reserved_blocks,
> > - F2FS_I(inode)->i_compr_blocks);
> > + atomic64_read(&F2FS_I(inode)->i_compr_blocks));
> > }
> >
> > return ret;
> > diff --git a/fs/f2fs/inode.c b/fs/f2fs/inode.c
> > index 66969ae852b9..e86857af9699 100644
> > --- a/fs/f2fs/inode.c
> > +++ b/fs/f2fs/inode.c
> > @@ -442,7 +442,8 @@ static int do_read_inode(struct inode *inode)
> > (fi->i_flags & F2FS_COMPR_FL)) {
> > if (F2FS_FITS_IN_INODE(ri, fi->i_extra_isize,
> > i_log_cluster_size)) {
> > - fi->i_compr_blocks = le64_to_cpu(ri->i_compr_blocks);
> > + atomic64_set(&fi->i_compr_blocks,
> > + le64_to_cpu(ri->i_compr_blocks));
> > fi->i_compress_algorithm = ri->i_compress_algorithm;
> > fi->i_log_cluster_size = ri->i_log_cluster_size;
> > fi->i_cluster_size = 1 << fi->i_log_cluster_size;
> > @@ -460,7 +461,7 @@ static int do_read_inode(struct inode *inode)
> > stat_inc_inline_inode(inode);
> > stat_inc_inline_dir(inode);
> > stat_inc_compr_inode(inode);
> > - stat_add_compr_blocks(inode, F2FS_I(inode)->i_compr_blocks);
> > + stat_add_compr_blocks(inode, atomic64_read(&fi->i_compr_blocks));
> >
> > return 0;
> > }
> > @@ -619,7 +620,8 @@ void f2fs_update_inode(struct inode *inode, struct page *node_page)
> > F2FS_FITS_IN_INODE(ri, F2FS_I(inode)->i_extra_isize,
> > i_log_cluster_size)) {
> > ri->i_compr_blocks =
> > - cpu_to_le64(F2FS_I(inode)->i_compr_blocks);
> > + cpu_to_le64(atomic64_read(
> > + &F2FS_I(inode)->i_compr_blocks));
> > ri->i_compress_algorithm =
> > F2FS_I(inode)->i_compress_algorithm;
> > ri->i_log_cluster_size =
> > @@ -768,7 +770,8 @@ void f2fs_evict_inode(struct inode *inode)
> > stat_dec_inline_dir(inode);
> > stat_dec_inline_inode(inode);
> > stat_dec_compr_inode(inode);
> > - stat_sub_compr_blocks(inode, F2FS_I(inode)->i_compr_blocks);
> > + stat_sub_compr_blocks(inode,
> > + atomic64_read(&F2FS_I(inode)->i_compr_blocks));
> >
> > if (likely(!f2fs_cp_error(sbi) &&
> > !is_sbi_flag_set(sbi, SBI_CP_DISABLED)))
> > diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c
> > index 83bf9a02f83f..65613500e1ab 100644
> > --- a/fs/f2fs/super.c
> > +++ b/fs/f2fs/super.c
> > @@ -1011,6 +1011,7 @@ static struct inode *f2fs_alloc_inode(struct super_block *sb)
> >
> > /* Initialize f2fs-specific inode info */
> > atomic_set(&fi->dirty_pages, 0);
> > + atomic64_set(&fi->i_compr_blocks, 0);
> > init_rwsem(&fi->i_sem);
> > spin_lock_init(&fi->i_size_lock);
> > INIT_LIST_HEAD(&fi->dirty_list);
> >