From: Azat Khuzhin <a3at.mail@gmail.com>
To: Miklos Szeredi <miklos@szeredi.hu>
Cc: viro@ZenIV.linux.org.uk, torvalds@linux-foundation.org,
linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org,
hch@infradead.org, akpm@linux-foundation.org, apw@canonical.com,
nbd@openwrt.org, neilb@suse.de, jordipujolp@gmail.com,
ezk@fsl.cs.sunysb.edu, dhowells@redhat.com,
sedat.dilek@gmail.com, hooanon05@yahoo.co.jp, mszeredi@suse.cz
Subject: Re: [PATCH 07/13] vfs: add RENAME_WHITEOUT
Date: Sat, 24 May 2014 12:02:39 +0400 [thread overview]
Message-ID: <20140524080239.GA26948@azat> (raw)
In-Reply-To: <1400838223-30844-8-git-send-email-miklos@szeredi.hu>
On Fri, May 23, 2014 at 11:43:37AM +0200, Miklos Szeredi wrote:
> From: Miklos Szeredi <mszeredi@suse.cz>
>
> This adds a new RENAME_WHITEOUT flag. This flag makes rename() create a
> whiteout of source. The whiteout creation is atomic relative to the
> rename.
>
> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
> ---
> fs/ext4/namei.c | 99 ++++++++++++++++++++++++++++++++++++++++---------
> fs/namei.c | 8 +++-
> include/linux/fs.h | 7 ++++
> include/uapi/linux/fs.h | 1 +
> 4 files changed, 95 insertions(+), 20 deletions(-)
>
> diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c
> index 1cb84f78909e..f782831b19dc 100644
> --- a/fs/ext4/namei.c
> +++ b/fs/ext4/namei.c
> @@ -3148,6 +3148,39 @@ static void ext4_update_dir_count(handle_t *handle, struct ext4_renament *ent)
> }
> }
>
> +static struct inode *ext4_whiteout_for_rename(struct ext4_renament *ent,
> + int credits, handle_t **h)
> +{
> + struct inode *wh;
> + handle_t *handle;
> + int retries = 0;
> +
> + /*
> + * for inode block, sb block, group summaries,
> + * and inode bitmap
> + */
> + credits += (EXT4_MAXQUOTAS_TRANS_BLOCKS(ent->dir->i_sb) +
> + EXT4_XATTR_TRANS_BLOCKS + 4);
> +retry:
> + wh = ext4_new_inode_start_handle(ent->dir, S_IFCHR | WHITEOUT_MODE,
> + &ent->dentry->d_name, 0, NULL,
> + EXT4_HT_DIR, credits);
> +
> + handle = ext4_journal_current_handle();
> + if (IS_ERR(wh)) {
> + if (handle)
> + ext4_journal_stop(handle);
> + if (PTR_ERR(wh) == -ENOSPC &&
> + ext4_should_retry_alloc(ent->dir->i_sb, &retries))
> + goto retry;
> + } else {
> + *h = handle;
> + init_special_inode(wh, wh->i_mode, WHITEOUT_DEV);
> + wh->i_op = &ext4_special_inode_operations;
> + }
> + return wh;
> +}
> +
> /*
> * Anybody can rename anything with this: the permission checks are left to the
> * higher-level routines.
> @@ -3156,8 +3189,9 @@ static void ext4_update_dir_count(handle_t *handle, struct ext4_renament *ent)
> * while new_{dentry,inode) refers to the destination dentry/inode
> * This comes from rename(const char *oldpath, const char *newpath)
> */
> -static int ext4_rename(struct inode *old_dir, struct dentry *old_dentry,
> - struct inode *new_dir, struct dentry *new_dentry)
> +static int ext4_simple_rename(struct inode *old_dir, struct dentry *old_dentry,
> + struct inode *new_dir, struct dentry *new_dentry,
> + unsigned int flags)
> {
> handle_t *handle = NULL;
> struct ext4_renament old = {
> @@ -3171,6 +3205,9 @@ static int ext4_rename(struct inode *old_dir, struct dentry *old_dentry,
> .inode = new_dentry->d_inode,
> };
> int retval;
> + struct inode *whiteout = NULL;
> + int credits;
> + u8 old_file_type;
>
> dquot_initialize(old.dir);
> dquot_initialize(new.dir);
> @@ -3202,11 +3239,17 @@ static int ext4_rename(struct inode *old_dir, struct dentry *old_dentry,
> if (new.inode && !test_opt(new.dir->i_sb, NO_AUTO_DA_ALLOC))
> ext4_alloc_da_blocks(old.inode);
>
> - handle = ext4_journal_start(old.dir, EXT4_HT_DIR,
> - (2 * EXT4_DATA_TRANS_BLOCKS(old.dir->i_sb) +
> - EXT4_INDEX_EXTRA_TRANS_BLOCKS + 2));
> - if (IS_ERR(handle))
> - return PTR_ERR(handle);
> + credits = (2 * EXT4_DATA_TRANS_BLOCKS(old.dir->i_sb) +
> + EXT4_INDEX_EXTRA_TRANS_BLOCKS + 2);
> + if (!(flags & RENAME_WHITEOUT)) {
> + handle = ext4_journal_start(old.dir, EXT4_HT_DIR, credits);
> + if (IS_ERR(handle))
> + return PTR_ERR(handle);
> + } else {
> + whiteout = ext4_whiteout_for_rename(&old, credits, &handle);
> + if (IS_ERR(whiteout))
> + return PTR_ERR(whiteout);
> + }
>
> if (IS_DIRSYNC(old.dir) || IS_DIRSYNC(new.dir))
> ext4_handle_sync(handle);
> @@ -3225,13 +3268,21 @@ static int ext4_rename(struct inode *old_dir, struct dentry *old_dentry,
> if (retval)
> goto end_rename;
> }
> + old_file_type = old.de->file_type;
> + if (whiteout) {
> + retval = ext4_setent(handle, &old, whiteout->i_ino,
> + EXT4_FT_CHRDEV);
> + if (retval)
> + goto end_rename;
> + ext4_mark_inode_dirty(handle, whiteout);
> + }
> if (!new.bh) {
> retval = ext4_add_entry(handle, new.dentry, old.inode);
> if (retval)
> goto end_rename;
> } else {
> retval = ext4_setent(handle, &new,
> - old.inode->i_ino, old.de->file_type);
> + old.inode->i_ino, old_file_type);
> if (retval)
> goto end_rename;
> }
> @@ -3243,10 +3294,12 @@ static int ext4_rename(struct inode *old_dir, struct dentry *old_dentry,
> old.inode->i_ctime = ext4_current_time(old.inode);
> ext4_mark_inode_dirty(handle, old.inode);
>
> - /*
> - * ok, that's it
> - */
> - ext4_rename_delete(handle, &old);
> + if (!whiteout) {
> + /*
> + * ok, that's it
> + */
> + ext4_rename_delete(handle, &old);
> + }
>
> if (new.inode) {
> ext4_dec_count(handle, new.inode);
> @@ -3282,6 +3335,12 @@ end_rename:
> brelse(old.dir_bh);
> brelse(old.bh);
> brelse(new.bh);
> + if (whiteout) {
> + if (retval)
> + drop_nlink(whiteout);
> + unlock_new_inode(whiteout);
> + iput(whiteout);
> + }
> if (handle)
> ext4_journal_stop(handle);
> return retval;
> @@ -3403,22 +3462,26 @@ end_rename:
> return retval;
> }
>
> +static int ext4_rename(struct inode *old_dir, struct dentry *old_dentry,
> + struct inode *new_dir, struct dentry *new_dentry)
> +{
> + return ext4_simple_rename(old_dir, old_dentry, new_dir, new_dentry, 0);
> +}
> +
> static int ext4_rename2(struct inode *old_dir, struct dentry *old_dentry,
> struct inode *new_dir, struct dentry *new_dentry,
> unsigned int flags)
> {
> - if (flags & ~(RENAME_NOREPLACE | RENAME_EXCHANGE))
> + if (flags & ~(RENAME_NOREPLACE | RENAME_EXCHANGE | RENAME_WHITEOUT))
> return -EINVAL;
>
> if (flags & RENAME_EXCHANGE) {
> return ext4_cross_rename(old_dir, old_dentry,
> new_dir, new_dentry);
> }
> - /*
> - * Existence checking was done by the VFS, otherwise "RENAME_NOREPLACE"
> - * is equivalent to regular rename.
> - */
> - return ext4_rename(old_dir, old_dentry, new_dir, new_dentry);
> +
> + return ext4_simple_rename(old_dir, old_dentry,
> + new_dir, new_dentry, flags);
> }
>
> /*
> diff --git a/fs/namei.c b/fs/namei.c
> index 413d7c138e95..adaa73d91173 100644
> --- a/fs/namei.c
> +++ b/fs/namei.c
> @@ -4186,12 +4186,16 @@ SYSCALL_DEFINE5(renameat2, int, olddfd, const char __user *, oldname,
> bool should_retry = false;
> int error;
>
> - if (flags & ~(RENAME_NOREPLACE | RENAME_EXCHANGE))
> + if (flags & ~(RENAME_NOREPLACE | RENAME_EXCHANGE | RENAME_WHITEOUT))
> return -EINVAL;
>
> - if ((flags & RENAME_NOREPLACE) && (flags & RENAME_EXCHANGE))
> + if ((flags & (RENAME_NOREPLACE | RENAME_WHITEOUT)) &&
> + (flags & RENAME_EXCHANGE))
> return -EINVAL;
>
> + if ((flags & RENAME_WHITEOUT) && !capable(CAP_MKNOD))
> + return -EPERM;
> +
> retry:
> from = user_path_parent(olddfd, oldname, &oldnd, lookup_flags);
> if (IS_ERR(from)) {
> diff --git a/include/linux/fs.h b/include/linux/fs.h
> index 88ec7a2878bc..48d3bd908b5d 100644
> --- a/include/linux/fs.h
> +++ b/include/linux/fs.h
> @@ -228,6 +228,13 @@ typedef void (dio_iodone_t)(struct kiocb *iocb, loff_t offset,
> #define WHITEOUT_DEV 0
>
> /*
> + * Whiteout is represented by a char device. The following constants define the
> + * mode and device number to use.
> + */
> +#define WHITEOUT_MODE 0
> +#define WHITEOUT_DEV 0
Quick note:
I suppose that there is no need in duplicating of WHITEOUT_{MODE,DEV}
Thanks.
> +
> +/*
> * This is the Inode Attributes structure, used for notify_change(). It
> * uses the above definitions as flags, to know which values have changed.
> * Also, in this manner, a Filesystem can look at only the values it cares
> diff --git a/include/uapi/linux/fs.h b/include/uapi/linux/fs.h
> index ca1a11bb4443..3735fa0a6784 100644
> --- a/include/uapi/linux/fs.h
> +++ b/include/uapi/linux/fs.h
> @@ -37,6 +37,7 @@
>
> #define RENAME_NOREPLACE (1 << 0) /* Don't overwrite target */
> #define RENAME_EXCHANGE (1 << 1) /* Exchange source and dest */
> +#define RENAME_WHITEOUT (1 << 2) /* Whiteout source */
>
> struct fstrim_range {
> __u64 start;
> --
> 1.8.1.4
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
--
Respectfully
Azat Khuzhin
next prev parent reply other threads:[~2014-05-24 8:02 UTC|newest]
Thread overview: 83+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-05-23 9:43 [PATCH 00/13] overlay filesystem v22 Miklos Szeredi
2014-05-23 9:43 ` [PATCH 01/13] vfs: add i_op->dentry_open() Miklos Szeredi
2014-05-23 9:43 ` [PATCH 02/13] vfs: export do_splice_direct() to modules Miklos Szeredi
2014-05-23 9:43 ` [PATCH 03/13] vfs: export __inode_permission() " Miklos Szeredi
2014-05-23 9:43 ` [PATCH 04/13] vfs: introduce clone_private_mount() Miklos Szeredi
2014-05-23 9:43 ` [PATCH 05/13] vfs: export check_sticky() Miklos Szeredi
2014-05-23 9:43 ` [PATCH 06/13] vfs: add whiteout support Miklos Szeredi
2014-05-24 8:29 ` Tetsuo Handa
2014-05-23 9:43 ` [PATCH 07/13] vfs: add RENAME_WHITEOUT Miklos Szeredi
2014-05-24 8:02 ` Azat Khuzhin [this message]
2014-05-23 9:43 ` [PATCH 08/13] overlay filesystem Miklos Szeredi
2014-05-23 9:43 ` [PATCH 09/13] overlayfs: add statfs support Miklos Szeredi
2014-05-23 9:43 ` [PATCH 10/13] overlayfs: implement show_options Miklos Szeredi
2014-05-23 9:43 ` [PATCH 11/13] overlay: overlay filesystem documentation Miklos Szeredi
2014-05-23 9:43 ` [PATCH 12/13] fs: limit filesystem stacking depth Miklos Szeredi
2014-05-23 9:43 ` [PATCH 13/13] vfs: dcache: Export d_ancestor to modules Miklos Szeredi
2014-05-26 1:56 ` [PATCH 00/13] overlay filesystem v22 J. R. Okajima
2014-05-28 16:06 ` Miklos Szeredi
2014-05-29 10:26 ` David Howells
2014-05-29 11:26 ` Miklos Szeredi
2014-05-29 10:54 ` Kernel errors with " David Howells
2014-05-29 11:16 ` David Howells
2014-05-29 11:28 ` David Howells
2014-05-29 12:07 ` Miklos Szeredi
2014-05-29 16:06 ` Miklos Szeredi
2014-05-29 16:10 ` Sedat Dilek
2014-05-29 16:23 ` More kernel " David Howells
2014-05-29 16:36 ` Miklos Szeredi
2014-05-29 16:44 ` David Howells
2014-05-29 16:48 ` Unionmount and overlayfs testsuite David Howells
2014-05-29 17:11 ` Sedat Dilek
2014-05-29 17:15 ` David Howells
2014-05-29 17:19 ` Sedat Dilek
2014-05-29 17:24 ` David Howells
2014-05-29 17:36 ` Sedat Dilek
2014-05-29 17:41 ` David Howells
2014-05-29 17:44 ` Sedat Dilek
2014-05-29 18:22 ` David Howells
2014-05-29 18:44 ` Sedat Dilek
2014-05-29 18:53 ` Sedat Dilek
2014-05-29 19:20 ` David Howells
2014-05-29 19:24 ` Sedat Dilek
2014-05-29 19:25 ` David Howells
2014-05-29 19:28 ` Sedat Dilek
2014-05-29 19:35 ` David Howells
2014-05-29 20:00 ` Sedat Dilek
2014-05-29 20:59 ` David Howells
2014-05-30 4:15 ` Sedat Dilek
2014-05-30 4:29 ` Sedat Dilek
2014-05-30 7:54 ` David Howells
2014-05-29 17:35 ` Sedat Dilek
2014-05-29 17:50 ` David Howells
2014-05-29 17:58 ` Sedat Dilek
2014-05-29 18:23 ` Dave Jones
2014-05-29 18:27 ` David Howells
2014-05-29 17:17 ` David Howells
2014-05-29 23:21 ` Dave Chinner
2014-05-30 3:35 ` J. R. Okajima
2014-05-30 4:44 ` J. R. Okajima
2014-05-30 8:49 ` David Howells
2014-05-30 9:15 ` J. R. Okajima
2014-06-03 8:08 ` Sedat Dilek
2014-06-03 9:00 ` David Howells
2014-06-03 9:12 ` Miklos Szeredi
2014-06-03 9:18 ` Sedat Dilek
2014-06-03 9:26 ` Sedat Dilek
2014-06-03 9:39 ` Sedat Dilek
2014-06-03 9:42 ` Miklos Szeredi
2014-06-03 10:15 ` Sedat Dilek
2014-06-03 10:21 ` Sedat Dilek
2014-06-03 10:55 ` Sedat Dilek
2014-06-03 10:33 ` David Howells
2014-06-03 13:21 ` Miklos Szeredi
2014-06-03 14:26 ` Sedat Dilek
2014-06-03 15:30 ` David Howells
2014-06-03 15:53 ` Miklos Szeredi
2014-06-05 22:16 ` David Howells
2014-06-24 16:46 ` Overlayfs rename bug David Howells
2014-07-08 9:29 ` Miklos Szeredi
2014-07-08 9:56 ` David Howells
2014-07-09 14:14 ` Miklos Szeredi
2014-05-30 9:14 ` Unionmount and overlayfs testsuite David Howells
2014-05-29 17:29 ` Kernel errors with overlay filesystem v22 David Howells
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20140524080239.GA26948@azat \
--to=a3at.mail@gmail.com \
--cc=akpm@linux-foundation.org \
--cc=apw@canonical.com \
--cc=dhowells@redhat.com \
--cc=ezk@fsl.cs.sunysb.edu \
--cc=hch@infradead.org \
--cc=hooanon05@yahoo.co.jp \
--cc=jordipujolp@gmail.com \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=miklos@szeredi.hu \
--cc=mszeredi@suse.cz \
--cc=nbd@openwrt.org \
--cc=neilb@suse.de \
--cc=sedat.dilek@gmail.com \
--cc=torvalds@linux-foundation.org \
--cc=viro@ZenIV.linux.org.uk \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).