2013-04-06 10:04:52

by Marco Stornelli

[permalink] [raw]
Subject: [PATCH 2/4] fsfreeze: manage kill signal when sb_start_write is called

In every place where sb_start_write was called now we must manage
the error code and return -EINTR.

Signed-off-by: Marco Stornelli <[email protected]>
---
fs/btrfs/file.c | 9 +++++++--
fs/cifs/file.c | 4 +++-
fs/ext4/mmp.c | 3 ++-
fs/ext4/super.c | 4 +++-
fs/fuse/file.c | 4 +++-
fs/namespace.c | 8 ++++++--
fs/ntfs/file.c | 4 +++-
fs/ocfs2/file.c | 4 +++-
fs/open.c | 8 ++++++--
fs/splice.c | 4 +++-
fs/xfs/xfs_file.c | 4 +++-
include/linux/fs.h | 6 ++++--
mm/filemap.c | 4 +++-
mm/filemap_xip.c | 4 +++-
14 files changed, 52 insertions(+), 18 deletions(-)

diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index ade03e6..4891fda 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -329,7 +329,9 @@ static int __btrfs_run_defrag_inode(struct btrfs_fs_info *fs_info,
range.len = (u64)-1;
range.start = defrag->last_offset;

- sb_start_write(fs_info->sb);
+ ret = sb_start_write(fs_info->sb);
+ if (ret)
+ goto cleanup;
num_defrag = btrfs_defrag_file(inode, NULL, &range, defrag->transid,
BTRFS_DEFRAG_BATCH);
sb_end_write(fs_info->sb);
@@ -1514,7 +1516,9 @@ static ssize_t btrfs_file_aio_write(struct kiocb *iocb,
size_t count, ocount;
bool sync = (file->f_flags & O_DSYNC) || IS_SYNC(file->f_mapping->host);

- sb_start_write(inode->i_sb);
+ err = sb_start_write(inode->i_sb);
+ if (err)
+ goto out2;

mutex_lock(&inode->i_mutex);

@@ -1618,6 +1622,7 @@ static ssize_t btrfs_file_aio_write(struct kiocb *iocb,
atomic_dec(&BTRFS_I(inode)->sync_writers);
out:
sb_end_write(inode->i_sb);
+out2:
current->backing_dev_info = NULL;
return num_written ? num_written : err;
}
diff --git a/fs/cifs/file.c b/fs/cifs/file.c
index 7a0dd99..1613eb9 100644
--- a/fs/cifs/file.c
+++ b/fs/cifs/file.c
@@ -2520,7 +2520,9 @@ cifs_writev(struct kiocb *iocb, const struct iovec *iov,

BUG_ON(iocb->ki_pos != pos);

- sb_start_write(inode->i_sb);
+ rc = sb_start_write(inode->i_sb);
+ if (rc)
+ return rc;

/*
* We need to hold the sem to be sure nobody modifies lock list
diff --git a/fs/ext4/mmp.c b/fs/ext4/mmp.c
index f9b5515..dbdfc6d 100644
--- a/fs/ext4/mmp.c
+++ b/fs/ext4/mmp.c
@@ -48,7 +48,8 @@ static int write_mmp_block(struct super_block *sb, struct buffer_head *bh)
* We protect against freezing so that we don't create dirty buffers
* on frozen filesystem.
*/
- sb_start_write(sb);
+ if (sb_start_write(sb))
+ return 1;
ext4_mmp_csum_set(sb, mmp);
mark_buffer_dirty(bh);
lock_buffer(bh);
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index 5d6d535..e6962be 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -2712,7 +2712,9 @@ static int ext4_run_li_request(struct ext4_li_request *elr)
sb = elr->lr_super;
ngroups = EXT4_SB(sb)->s_groups_count;

- sb_start_write(sb);
+ ret = sb_start_write(sb);
+ if (ret)
+ return 1;
for (group = elr->lr_next_group; group < ngroups; group++) {
gdp = ext4_get_group_desc(sb, group, NULL);
if (!gdp) {
diff --git a/fs/fuse/file.c b/fs/fuse/file.c
index 34b80ba..57194eb 100644
--- a/fs/fuse/file.c
+++ b/fs/fuse/file.c
@@ -971,7 +971,9 @@ static ssize_t fuse_file_aio_write(struct kiocb *iocb, const struct iovec *iov,
return err;

count = ocount;
- sb_start_write(inode->i_sb);
+ err = sb_start_write(inode->i_sb);
+ if (err)
+ return err;
mutex_lock(&inode->i_mutex);

/* We can write back this queue in page reclaim */
diff --git a/fs/namespace.c b/fs/namespace.c
index d581e45..2d2b054 100644
--- a/fs/namespace.c
+++ b/fs/namespace.c
@@ -343,7 +343,9 @@ int mnt_want_write(struct vfsmount *m)
{
int ret;

- sb_start_write(m->mnt_sb);
+ ret = sb_start_write(m->mnt_sb);
+ if (ret)
+ return ret;
ret = __mnt_want_write(m);
if (ret)
sb_end_write(m->mnt_sb);
@@ -403,7 +405,9 @@ int mnt_want_write_file(struct file *file)
{
int ret;

- sb_start_write(file->f_path.mnt->mnt_sb);
+ ret = sb_start_write(file->f_path.mnt->mnt_sb);
+ if (ret)
+ return ret;
ret = __mnt_want_write_file(file);
if (ret)
sb_end_write(file->f_path.mnt->mnt_sb);
diff --git a/fs/ntfs/file.c b/fs/ntfs/file.c
index 5b2d4f0..a414c53 100644
--- a/fs/ntfs/file.c
+++ b/fs/ntfs/file.c
@@ -2129,7 +2129,9 @@ static ssize_t ntfs_file_aio_write(struct kiocb *iocb, const struct iovec *iov,

BUG_ON(iocb->ki_pos != pos);

- sb_start_write(inode->i_sb);
+ ret = sb_start_write(inode->i_sb);
+ if (ret)
+ return ret;
mutex_lock(&inode->i_mutex);
ret = ntfs_file_aio_write_nolock(iocb, iov, nr_segs, &iocb->ki_pos);
mutex_unlock(&inode->i_mutex);
diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c
index 6474cb4..8b68176 100644
--- a/fs/ocfs2/file.c
+++ b/fs/ocfs2/file.c
@@ -2248,7 +2248,9 @@ static ssize_t ocfs2_file_aio_write(struct kiocb *iocb,
if (iocb->ki_left == 0)
return 0;

- sb_start_write(inode->i_sb);
+ ret = sb_start_write(inode->i_sb);
+ if (ret)
+ return ret;

appending = file->f_flags & O_APPEND ? 1 : 0;
direct_io = file->f_flags & O_DIRECT ? 1 : 0;
diff --git a/fs/open.c b/fs/open.c
index 6835446..4589d88 100644
--- a/fs/open.c
+++ b/fs/open.c
@@ -182,7 +182,9 @@ static long do_sys_ftruncate(unsigned int fd, loff_t length, int small)
if (IS_APPEND(inode))
goto out_putf;

- sb_start_write(inode->i_sb);
+ error = sb_start_write(inode->i_sb);
+ if (error)
+ goto out_putf;
error = locks_verify_truncate(inode, f.file, length);
if (!error)
error = security_path_truncate(&f.file->f_path);
@@ -293,7 +295,9 @@ int do_fallocate(struct file *file, int mode, loff_t offset, loff_t len)
if (!file->f_op->fallocate)
return -EOPNOTSUPP;

- sb_start_write(inode->i_sb);
+ ret = sb_start_write(inode->i_sb);
+ if (ret)
+ return ret;
ret = file->f_op->fallocate(file, mode, offset, len);
sb_end_write(inode->i_sb);
return ret;
diff --git a/fs/splice.c b/fs/splice.c
index 29e394e..4b5c365 100644
--- a/fs/splice.c
+++ b/fs/splice.c
@@ -1000,7 +1000,9 @@ generic_file_splice_write(struct pipe_inode_info *pipe, struct file *out,
};
ssize_t ret;

- sb_start_write(inode->i_sb);
+ ret = sb_start_write(inode->i_sb);
+ if (ret)
+ return ret;

pipe_lock(pipe);

diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c
index f03bf1a..6ba4c0d 100644
--- a/fs/xfs/xfs_file.c
+++ b/fs/xfs/xfs_file.c
@@ -775,7 +775,9 @@ xfs_file_aio_write(
if (ocount == 0)
return 0;

- sb_start_write(inode->i_sb);
+ ret = sb_start_write(inode->i_sb);
+ if (ret)
+ return ret;

if (XFS_FORCED_SHUTDOWN(ip->i_mount)) {
ret = -EIO;
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 2bfb88d..03921d6 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -1376,6 +1376,8 @@ static inline void sb_end_intwrite(struct super_block *sb)
* sb_start_write - get write access to a superblock
* @sb: the super we write to
*
+ * It returns zero when no error occured, the error code otherwise.
+ *
* When a process wants to write data or metadata to a file system (i.e. dirty
* a page or an inode), it should embed the operation in a sb_start_write() -
* sb_end_write() pair to get exclusion against file system freezing. This
@@ -1391,9 +1393,9 @@ static inline void sb_end_intwrite(struct super_block *sb)
* -> i_mutex (write path, truncate, directory ops, ...)
* -> s_umount (freeze_super, thaw_super)
*/
-static inline void sb_start_write(struct super_block *sb)
+static inline int sb_start_write(struct super_block *sb)
{
- __sb_start_write_wait(sb, SB_FREEZE_WRITE, false);
+ return __sb_start_write_wait(sb, SB_FREEZE_WRITE, true);
}

static inline int sb_start_write_trylock(struct super_block *sb)
diff --git a/mm/filemap.c b/mm/filemap.c
index e1979fd..b238671 100644
--- a/mm/filemap.c
+++ b/mm/filemap.c
@@ -2528,7 +2528,9 @@ ssize_t generic_file_aio_write(struct kiocb *iocb, const struct iovec *iov,

BUG_ON(iocb->ki_pos != pos);

- sb_start_write(inode->i_sb);
+ ret = sb_start_write(inode->i_sb);
+ if (ret)
+ return ret;
mutex_lock(&inode->i_mutex);
ret = __generic_file_aio_write(iocb, iov, nr_segs, &iocb->ki_pos);
mutex_unlock(&inode->i_mutex);
diff --git a/mm/filemap_xip.c b/mm/filemap_xip.c
index a912da6..4cde643 100644
--- a/mm/filemap_xip.c
+++ b/mm/filemap_xip.c
@@ -404,7 +404,9 @@ xip_file_write(struct file *filp, const char __user *buf, size_t len,
loff_t pos;
ssize_t ret;

- sb_start_write(inode->i_sb);
+ ret = sb_start_write(inode->i_sb);
+ if (ret)
+ return ret;

mutex_lock(&inode->i_mutex);

--
1.7.3.4

------------------------------------------------------------------------------
Minimize network downtime and maximize team effectiveness.
Reduce network management and security costs.Learn how to hire
the most talented Cisco Certified professionals. Visit the
Employer Resources Portal
http://www.cisco.com/web/learning/employer_resources/index.html


2013-04-06 13:17:03

by Matthew Wilcox

[permalink] [raw]
Subject: Re: [PATCH 2/4] fsfreeze: manage kill signal when sb_start_write is called

On Sat, Apr 06, 2013 at 12:04:52PM +0200, Marco Stornelli wrote:
> In every place where sb_start_write was called now we must manage
> the error code and return -EINTR.

If we must manage the error code, then these functions should be marked
__must_check.

--
Matthew Wilcox Intel Open Source Technology Centre
"Bill, look, we understand that you're interested in selling us this
operating system, but compare it to ours. We can't possibly take such
a retrograde step."

2013-04-06 14:56:25

by Marco Stornelli

[permalink] [raw]
Subject: Re: [PATCH 2/4] fsfreeze: manage kill signal when sb_start_write is called

Il 06/04/2013 15:17, Matthew Wilcox ha scritto:
> On Sat, Apr 06, 2013 at 12:04:52PM +0200, Marco Stornelli wrote:
>> In every place where sb_start_write was called now we must manage
>> the error code and return -EINTR.
>
> If we must manage the error code, then these functions should be marked
> __must_check.
>

Yep, good point.

Marco

2013-04-06 21:50:33

by Al Viro

[permalink] [raw]
Subject: Re: [PATCH 2/4] fsfreeze: manage kill signal when sb_start_write is called

On Sat, Apr 06, 2013 at 12:04:52PM +0200, Marco Stornelli wrote:
> In every place where sb_start_write was called now we must manage
> the error code and return -EINTR.

Now go and look for callers of mnt_want_write() ;-/ The really
painful one is in do_last(), but kern_path_create() is not much
better; mq_open() would be bad too, had it not been about something
you are not going to freeze anyway...