Message-id: <1386545325.2101.58.camel@kjgkr>
Date:	Mon, 09 Dec 2013 08:28:45 +0900
From:	Jaegeuk Kim <jaegeuk.kim@...sung.com>
To:	Christoph Hellwig <hch@...radead.org>
Cc:	viro@...iv.linux.org.uk, linux-fsdevel@...r.kernel.org,
	linux-btrfs@...r.kernel.org, linux-ext4@...r.kernel.org,
	linux-f2fs-devel@...ts.sourceforge.net,
	linux-mtd@...ts.infradead.org, Mark Fasheh <mfasheh@...e.com>,
	Joel Becker <jlbec@...lplan.org>,
	reiserfs-devel@...r.kernel.org, xfs@....sgi.com,
	jfs-discussion@...ts.sourceforge.net, cluster-devel@...hat.com,
	linux-nfs@...r.kernel.org
Subject: Re: [PATCH 09/18] f2fs: use generic posix ACL infrastructure

2013-12-08 (일), 01:14 -0800, Christoph Hellwig:
> On Fri, Dec 06, 2013 at 10:37:34AM +0900, Jaegeuk Kim wrote:
> > f2fs caches a new mode bit for a while to make the consistency between
> > xattr's acl mode and the inode mode.
> 
> Can you explain what exactly you're trying to do there?  I've been
> trying to unwrap what's going on and can't really see the point:
> 
>  - i_acl_mode and FI_ACL_MODE get set in __setattr_copy, but right
>    after that call, still under i_mutex and before marking the inode
>    dirty f2fs_acl_chmod makes use of it, and it gets cleared right
>    after. Is there any race that could happen with a locked inode
>    not marked dirty yet on f2fs?

As you guess, there is no race problem, but the problem is on acl
consistency between xattr->mode and inode->mode.

Previously, f2fs_setattr triggers:
              new_mode inode->mode xattr->mode iblock->mode
f2fs_setattr     x    ->    x           y         y       
[update_inode]              x    ---  [ y ]  ---> x
[checkpoint]                x           y         x
__f2fs_setxattr             x     ->    x         x

In this flow, f2fs is able to break the consistency between xattr->mode
and iblock->mode after checkpoint followed by sudden-power-off.

So, fi->mode was introduced to address the problem.
The new f2fs_setattr triggers:
              new_mode inode->mode fi->mode xattr->mode iblock->mode
f2fs_setattr     x    ---  [y]  --->   x          y          y
[update_inode]              y          x          y          y
[checkpoint]                y          x          y          y
__f2fs_setxattr             x    <-    x    ->    x     ->   x

Finally, __f2fs_setxattr synchronizes inode->mode, xattr->mode, and
iblock->mode all together.

The root question is "is it possible to call update_inode in the
i_mutex-covered region like f2fs_setattr?".
The update_inode of f2fs is called from a bunch of places so currently
I'm not sure it can be impossible.

Thanks,

> We could pass a mode argument
>    to posix_acl_create, but I'd prefer to avoid that if we can.
>  - on the set_acl side it gets set in __f2fs_set_acl, and then
>    i_mode is update in __f2fs_setxattr which could easily done with
>    a stack variable.
> 
> RFC patch below:
> 
> 
> diff --git a/fs/f2fs/acl.c b/fs/f2fs/acl.c
> index 4f52fe0f..6647545 100644
> --- a/fs/f2fs/acl.c
> +++ b/fs/f2fs/acl.c
> @@ -17,9 +17,6 @@
>  #include "xattr.h"
>  #include "acl.h"
>  
> -#define get_inode_mode(i)	((is_inode_flag_set(F2FS_I(i), FI_ACL_MODE)) ? \
> -					(F2FS_I(i)->i_acl_mode) : ((i)->i_mode))
> -
>  static inline size_t f2fs_acl_size(int count)
>  {
>  	if (count <= 4) {
> @@ -209,11 +206,11 @@ static int __f2fs_set_acl(struct inode *inode, int type,
>  			struct posix_acl *acl, struct page *ipage)
>  {
>  	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
> -	struct f2fs_inode_info *fi = F2FS_I(inode);
>  	int name_index;
>  	void *value = NULL;
>  	size_t size = 0;
>  	int error;
> +	umode_t mode = 0;
>  
>  	if (!test_opt(sbi, POSIX_ACL))
>  		return 0;
> @@ -224,10 +221,10 @@ static int __f2fs_set_acl(struct inode *inode, int type,
>  	case ACL_TYPE_ACCESS:
>  		name_index = F2FS_XATTR_INDEX_POSIX_ACL_ACCESS;
>  		if (acl) {
> -			error = posix_acl_equiv_mode(acl, &inode->i_mode);
> +			mode = inode->i_mode;
> +			error = posix_acl_equiv_mode(acl, &mode);
>  			if (error < 0)
>  				return error;
> -			set_acl_inode(fi, inode->i_mode);
>  			if (error == 0)
>  				acl = NULL;
>  		}
> @@ -245,19 +242,15 @@ static int __f2fs_set_acl(struct inode *inode, int type,
>  
>  	if (acl) {
>  		value = f2fs_acl_to_disk(acl, &size);
> -		if (IS_ERR(value)) {
> -			cond_clear_inode_flag(fi, FI_ACL_MODE);
> +		if (IS_ERR(value))
>  			return (int)PTR_ERR(value);
> -		}
>  	}
>  
> -	error = f2fs_setxattr(inode, name_index, "", value, size, ipage);
> +	error = f2fs_setxattr(inode, name_index, "", value, size, ipage, mode);
>  
>  	kfree(value);
>  	if (!error)
>  		set_cached_acl(inode, type, acl);
> -
> -	cond_clear_inode_flag(fi, FI_ACL_MODE);
>  	return error;
>  }
>  
> @@ -289,28 +282,3 @@ int f2fs_init_acl(struct inode *inode, struct inode *dir, struct page *ipage)
>  
>  	return error;
>  }
> -
> -int f2fs_acl_chmod(struct inode *inode)
> -{
> -	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
> -	struct posix_acl *acl;
> -	int error;
> -	umode_t mode = get_inode_mode(inode);
> -
> -	if (!test_opt(sbi, POSIX_ACL))
> -		return 0;
> -	if (S_ISLNK(mode))
> -		return -EOPNOTSUPP;
> -
> -	acl = f2fs_get_acl(inode, ACL_TYPE_ACCESS);
> -	if (IS_ERR(acl) || !acl)
> -		return PTR_ERR(acl);
> -
> -	error = __posix_acl_chmod(&acl, GFP_KERNEL, mode);
> -	if (error)
> -		return error;
> -
> -	error = __f2fs_set_acl(inode, ACL_TYPE_ACCESS, acl, NULL);
> -	posix_acl_release(acl);
> -	return error;
> -}
> diff --git a/fs/f2fs/acl.h b/fs/f2fs/acl.h
> index 2af31fe..e086465 100644
> --- a/fs/f2fs/acl.h
> +++ b/fs/f2fs/acl.h
> @@ -38,18 +38,12 @@ struct f2fs_acl_header {
>  
>  extern struct posix_acl *f2fs_get_acl(struct inode *, int);
>  extern int f2fs_set_acl(struct inode *inode, struct posix_acl *acl, int type);
> -extern int f2fs_acl_chmod(struct inode *);
>  extern int f2fs_init_acl(struct inode *, struct inode *, struct page *);
>  #else
>  #define f2fs_check_acl	NULL
>  #define f2fs_get_acl	NULL
>  #define f2fs_set_acl	NULL
>  
> -static inline int f2fs_acl_chmod(struct inode *inode)
> -{
> -	return 0;
> -}
> -
>  static inline int f2fs_init_acl(struct inode *inode, struct inode *dir,
>  							struct page *page)
>  {
> diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h
> index 89dc750..1e774e6 100644
> --- a/fs/f2fs/f2fs.h
> +++ b/fs/f2fs/f2fs.h
> @@ -181,7 +181,6 @@ struct f2fs_inode_info {
>  	unsigned char i_advise;		/* use to give file attribute hints */
>  	unsigned int i_current_depth;	/* use only in directory structure */
>  	unsigned int i_pino;		/* parent inode number */
> -	umode_t i_acl_mode;		/* keep file acl mode temporarily */
>  
>  	/* Use below internally in f2fs*/
>  	unsigned long flags;		/* use to pass per-file flags */
> @@ -872,7 +871,6 @@ enum {
>  	FI_NEW_INODE,		/* indicate newly allocated inode */
>  	FI_DIRTY_INODE,		/* indicate inode is dirty or not */
>  	FI_INC_LINK,		/* need to increment i_nlink */
> -	FI_ACL_MODE,		/* indicate acl mode */
>  	FI_NO_ALLOC,		/* should not allocate any blocks */
>  	FI_UPDATE_DIR,		/* should update inode block for consistency */
>  	FI_DELAY_IPUT,		/* used for the recovery */
> @@ -894,21 +892,6 @@ static inline void clear_inode_flag(struct f2fs_inode_info *fi, int flag)
>  	clear_bit(flag, &fi->flags);
>  }
>  
> -static inline void set_acl_inode(struct f2fs_inode_info *fi, umode_t mode)
> -{
> -	fi->i_acl_mode = mode;
> -	set_inode_flag(fi, FI_ACL_MODE);
> -}
> -
> -static inline int cond_clear_inode_flag(struct f2fs_inode_info *fi, int flag)
> -{
> -	if (is_inode_flag_set(fi, FI_ACL_MODE)) {
> -		clear_inode_flag(fi, FI_ACL_MODE);
> -		return 1;
> -	}
> -	return 0;
> -}
> -
>  static inline void get_inline_info(struct f2fs_inode_info *fi,
>  					struct f2fs_inode *ri)
>  {
> diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c
> index 13eff60..80ef669 100644
> --- a/fs/f2fs/file.c
> +++ b/fs/f2fs/file.c
> @@ -339,41 +339,9 @@ int f2fs_getattr(struct vfsmount *mnt,
>  	return 0;
>  }
>  
> -#ifdef CONFIG_F2FS_FS_POSIX_ACL
> -static void __setattr_copy(struct inode *inode, const struct iattr *attr)
> -{
> -	struct f2fs_inode_info *fi = F2FS_I(inode);
> -	unsigned int ia_valid = attr->ia_valid;
> -
> -	if (ia_valid & ATTR_UID)
> -		inode->i_uid = attr->ia_uid;
> -	if (ia_valid & ATTR_GID)
> -		inode->i_gid = attr->ia_gid;
> -	if (ia_valid & ATTR_ATIME)
> -		inode->i_atime = timespec_trunc(attr->ia_atime,
> -						inode->i_sb->s_time_gran);
> -	if (ia_valid & ATTR_MTIME)
> -		inode->i_mtime = timespec_trunc(attr->ia_mtime,
> -						inode->i_sb->s_time_gran);
> -	if (ia_valid & ATTR_CTIME)
> -		inode->i_ctime = timespec_trunc(attr->ia_ctime,
> -						inode->i_sb->s_time_gran);
> -	if (ia_valid & ATTR_MODE) {
> -		umode_t mode = attr->ia_mode;
> -
> -		if (!in_group_p(inode->i_gid) && !capable(CAP_FSETID))
> -			mode &= ~S_ISGID;
> -		set_acl_inode(fi, mode);
> -	}
> -}
> -#else
> -#define __setattr_copy setattr_copy
> -#endif
> -
>  int f2fs_setattr(struct dentry *dentry, struct iattr *attr)
>  {
>  	struct inode *inode = dentry->d_inode;
> -	struct f2fs_inode_info *fi = F2FS_I(inode);
>  	int err;
>  
>  	err = inode_change_ok(inode, attr);
> @@ -387,15 +355,9 @@ int f2fs_setattr(struct dentry *dentry, struct iattr *attr)
>  		f2fs_balance_fs(F2FS_SB(inode->i_sb));
>  	}
>  
> -	__setattr_copy(inode, attr);
> -
> -	if (attr->ia_valid & ATTR_MODE) {
> -		err = f2fs_acl_chmod(inode);
> -		if (err || is_inode_flag_set(fi, FI_ACL_MODE)) {
> -			inode->i_mode = fi->i_acl_mode;
> -			clear_inode_flag(fi, FI_ACL_MODE);
> -		}
> -	}
> +	setattr_copy(inode, attr);
> +	if (attr->ia_valid & ATTR_MODE)
> +		err = posix_acl_chmod(inode);
>  
>  	mark_inode_dirty(inode);
>  	return err;
> diff --git a/fs/f2fs/xattr.c b/fs/f2fs/xattr.c
> index e2b9299..8820857 100644
> --- a/fs/f2fs/xattr.c
> +++ b/fs/f2fs/xattr.c
> @@ -108,7 +108,7 @@ static int f2fs_xattr_generic_set(struct dentry *dentry, const char *name,
>  	if (strcmp(name, "") == 0)
>  		return -EINVAL;
>  
> -	return f2fs_setxattr(dentry->d_inode, type, name, value, size, NULL);
> +	return f2fs_setxattr(dentry->d_inode, type, name, value, size, NULL, 0);
>  }
>  
>  static size_t f2fs_xattr_advise_list(struct dentry *dentry, char *list,
> @@ -157,7 +157,7 @@ static int f2fs_xattr_advise_set(struct dentry *dentry, const char *name,
>  #ifdef CONFIG_F2FS_FS_SECURITY
>  static int __f2fs_setxattr(struct inode *inode, int name_index,
>  			const char *name, const void *value, size_t value_len,
> -			struct page *ipage);
> +			struct page *ipage, umode_t mode);
>  static int f2fs_initxattrs(struct inode *inode, const struct xattr *xattr_array,
>  		void *page)
>  {
> @@ -167,7 +167,7 @@ static int f2fs_initxattrs(struct inode *inode, const struct xattr *xattr_array,
>  	for (xattr = xattr_array; xattr->name != NULL; xattr++) {
>  		err = __f2fs_setxattr(inode, F2FS_XATTR_INDEX_SECURITY,
>  				xattr->name, xattr->value,
> -				xattr->value_len, (struct page *)page);
> +				xattr->value_len, (struct page *)page, 0);
>  		if (err < 0)
>  			break;
>  	}
> @@ -475,9 +475,8 @@ cleanup:
>  
>  static int __f2fs_setxattr(struct inode *inode, int name_index,
>  			const char *name, const void *value, size_t value_len,
> -			struct page *ipage)
> +			struct page *ipage, umode_t mode)
>  {
> -	struct f2fs_inode_info *fi = F2FS_I(inode);
>  	struct f2fs_xattr_entry *here, *last;
>  	void *base_addr;
>  	int found, newsize;
> @@ -566,10 +565,9 @@ static int __f2fs_setxattr(struct inode *inode, int name_index,
>  	if (error)
>  		goto exit;
>  
> -	if (is_inode_flag_set(fi, FI_ACL_MODE)) {
> -		inode->i_mode = fi->i_acl_mode;
> +	if (mode) {
> +		inode->i_mode = mode;
>  		inode->i_ctime = CURRENT_TIME;
> -		clear_inode_flag(fi, FI_ACL_MODE);
>  	}
>  
>  	if (ipage)
> @@ -582,7 +580,8 @@ exit:
>  }
>  
>  int f2fs_setxattr(struct inode *inode, int name_index, const char *name,
> -			const void *value, size_t value_len, struct page *ipage)
> +			const void *value, size_t value_len, struct page *ipage,
> +			umode_t mode)
>  {
>  	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
>  	int err;
> @@ -590,7 +589,8 @@ int f2fs_setxattr(struct inode *inode, int name_index, const char *name,
>  	f2fs_balance_fs(sbi);
>  
>  	f2fs_lock_op(sbi);
> -	err = __f2fs_setxattr(inode, name_index, name, value, value_len, ipage);
> +	err = __f2fs_setxattr(inode, name_index, name, value, value_len, ipage,
> +			      mode);
>  	f2fs_unlock_op(sbi);
>  
>  	return err;
> diff --git a/fs/f2fs/xattr.h b/fs/f2fs/xattr.h
> index b21d9eb..c73588a 100644
> --- a/fs/f2fs/xattr.h
> +++ b/fs/f2fs/xattr.h
> @@ -114,14 +114,15 @@ extern const struct xattr_handler f2fs_xattr_security_handler;
>  extern const struct xattr_handler *f2fs_xattr_handlers[];
>  
>  extern int f2fs_setxattr(struct inode *, int, const char *,
> -				const void *, size_t, struct page *);
> +				const void *, size_t, struct page *, umode_t);
>  extern int f2fs_getxattr(struct inode *, int, const char *, void *, size_t);
>  extern ssize_t f2fs_listxattr(struct dentry *, char *, size_t);
>  #else
>  
>  #define f2fs_xattr_handlers	NULL
>  static inline int f2fs_setxattr(struct inode *inode, int name_index,
> -		const char *name, const void *value, size_t value_len)
> +		const char *name, const void *value, size_t value_len,
> +		umode_t mode)
>  {
>  	return -EOPNOTSUPP;
>  }
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
> the body of a message to majordomo@...r.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

-- 
Jaegeuk Kim
Samsung

--
To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Powered by blists - more mailing lists