[PATCH 0/4] quota: add project quota support

The following patches propose an implementation of project quota
support for ext4. A project is an aggregate of unrelated inodes
which might scatter in different directories. Inodes belongs to a
project possesses a same identification i.e. 'project ID', just
like every inode has its user/group identification. The following
patches adds project quota as supplement to the former uer/group
quota types.

The semantics of ext4 project quota is consistent with XFS. Each
directory can have EXT4_INODE_PROJINHERIT flag set. When the
EXT4_INODE_PROJINHERIT flag of a parent directory is not set, a
newly created inode under that directory will have a default project
ID (i.e. 0). And its EXT4_INODE_PROJINHERIT flag is not set either.
When this flag is set on a directory, following rules will be kept:

1) The newly created inode under that directory will inherit both
the EXT4_INODE_PROJINHERIT flag and the project ID from its parent
directory.

2) Hard-linking a inode with different project ID into that directory
will fail with errno EXDEV.

3) Renaming a inode with different project ID into that directory
will fail with errno EXDEV. However, 'mv' command will detect this
failure and copy the renamed inode to a new inode in the directory.
Thus, this new inode will inherit both the project ID and
EXT4_INODE_PROJINHERIT flag.

4) If the project quota of that ID is being enforced, statfs() on
that directory will take the quotas as another upper limits along
with the capacity of the file system, i.e. the total block/inode
number will be the minimum of the quota limits and file system
capacity.

Changelog:
* v4 <- v3:
 - Do not check project feature when set/get project ID;
 - Use EXT4_MAXQUOTAS instead of MAXQUOTAS in ext4 patches;
 - Remove unnecessary change of fs/quota/dquot.c;
 - Remove CONFIG_QUOTA_PROJECT.
* v3 <- v2:
 - Add EXT4_INODE_PROJINHERIT semantics.
* v2 <- v1:
 - Add ioctl interface for setting/getting project;
 - Add EXT4_FEATURE_RO_COMPAT_PROJECT;
 - Add get_projid() method in struct dquot_operations;
 - Add error check of ext4_inode_projid_set/get().

v3: http://www.spinics.net/lists/linux-ext4/msg45184.html
v2: http://www.spinics.net/lists/linux-ext4/msg44695.html
v1: http://article.gmane.org/gmane.comp.file-systems.ext4/45153

Any comments or feedbacks are appreciated.

Regards,
                                         - Li Xi

Li Xi (4):
  Adds general codes to enforces project quota limits
  Adds project ID support for ext4
  Adds project quota support for ext4
  Adds ioctl interface support for ext4 project

 Documentation/filesystems/ext4.txt |    4 +
 fs/ext4/ext4.h                     |   31 +++++++--
 fs/ext4/ialloc.c                   |    6 ++
 fs/ext4/inode.c                    |   29 ++++++++-
 fs/ext4/ioctl.c                    |   85 ++++++++++++++++++++++++
 fs/ext4/namei.c                    |   10 +++
 fs/ext4/super.c                    |  124 ++++++++++++++++++++++++++++++++---
 fs/quota/dquot.c                   |   16 ++++-
 fs/quota/quota.c                   |    5 +-
 fs/quota/quotaio_v2.h              |    6 +-
 include/linux/quota.h              |    2 +
 include/uapi/linux/quota.h         |    6 +-
 12 files changed, 299 insertions(+), 25 deletions(-)

This patch adds a new internal field of ext4 inode to save project
identifier. Also a new flag EXT4_INODE_PROJINHERIT is added for
inheriting project ID from parent directory.

Signed-off-by: Li Xi <redacted>
---
 fs/ext4/ext4.h   |   21 +++++++++++++++++----
 fs/ext4/ialloc.c |    6 ++++++
 fs/ext4/inode.c  |   29 ++++++++++++++++++++++++++++-
 fs/ext4/namei.c  |   10 ++++++++++
 fs/ext4/super.c  |    1 +
 5 files changed, 62 insertions(+), 5 deletions(-)

diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index 1eb5b7b..092f60c 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h

@@ -386,16 +386,18 @@ struct flex_groups {
 #define EXT4_EA_INODE_FL	        0x00200000 /* Inode used for large EA */
 #define EXT4_EOFBLOCKS_FL		0x00400000 /* Blocks allocated beyond EOF */
 #define EXT4_INLINE_DATA_FL		0x10000000 /* Inode has inline data. */
+#define EXT4_PROJINHERIT_FL		0x20000000 /* Create with parents projid */
 #define EXT4_RESERVED_FL		0x80000000 /* reserved for ext4 lib */
 
-#define EXT4_FL_USER_VISIBLE		0x004BDFFF /* User visible flags */
-#define EXT4_FL_USER_MODIFIABLE		0x004380FF /* User modifiable flags */
+#define EXT4_FL_USER_VISIBLE		0x204BDFFF /* User visible flags */
+#define EXT4_FL_USER_MODIFIABLE		0x204380FF /* User modifiable flags */
 
 /* Flags that should be inherited by new inodes from their parent. */
 #define EXT4_FL_INHERITED (EXT4_SECRM_FL | EXT4_UNRM_FL | EXT4_COMPR_FL |\
 			   EXT4_SYNC_FL | EXT4_NODUMP_FL | EXT4_NOATIME_FL |\
 			   EXT4_NOCOMPR_FL | EXT4_JOURNAL_DATA_FL |\
-			   EXT4_NOTAIL_FL | EXT4_DIRSYNC_FL)
+			   EXT4_NOTAIL_FL | EXT4_DIRSYNC_FL |\
+			   EXT4_PROJINHERIT_FL)
 
 /* Flags that are appropriate for regular files (all but dir-specific ones). */
 #define EXT4_REG_FLMASK (~(EXT4_DIRSYNC_FL | EXT4_TOPDIR_FL))

@@ -443,6 +445,7 @@ enum {
 	EXT4_INODE_EA_INODE	= 21,	/* Inode used for large EA */
 	EXT4_INODE_EOFBLOCKS	= 22,	/* Blocks allocated beyond EOF */
 	EXT4_INODE_INLINE_DATA	= 28,	/* Data in inode. */
+	EXT4_INODE_PROJINHERIT	= 29,	/* Create with parents projid */
 	EXT4_INODE_RESERVED	= 31,	/* reserved for ext4 lib */
 };
 

@@ -694,6 +697,7 @@ struct ext4_inode {
 	__le32  i_crtime;       /* File Creation time */
 	__le32  i_crtime_extra; /* extra FileCreationtime (nsec << 2 | epoch) */
 	__le32  i_version_hi;	/* high 32 bits for 64-bit version */
+	__le32  i_projid;	/* Project ID */
 };
 
 struct move_extent {

@@ -943,6 +947,7 @@ struct ext4_inode_info {
 
 	/* Precomputed uuid+inum+igen checksum for seeding inode checksums */
 	__u32 i_csum_seed;
+	kprojid_t i_projid;
 };
 
 /*

@@ -1526,6 +1531,7 @@ static inline void ext4_clear_state_flags(struct ext4_inode_info *ei)
  * GDT_CSUM bits are mutually exclusive.
  */
 #define EXT4_FEATURE_RO_COMPAT_METADATA_CSUM	0x0400
+#define EXT4_FEATURE_RO_COMPAT_PROJECT		0x1000 /* Project quota */
 
 #define EXT4_FEATURE_INCOMPAT_COMPRESSION	0x0001
 #define EXT4_FEATURE_INCOMPAT_FILETYPE		0x0002

@@ -1575,7 +1581,8 @@ static inline void ext4_clear_state_flags(struct ext4_inode_info *ei)
 					 EXT4_FEATURE_RO_COMPAT_HUGE_FILE |\
 					 EXT4_FEATURE_RO_COMPAT_BIGALLOC |\
 					 EXT4_FEATURE_RO_COMPAT_METADATA_CSUM|\
-					 EXT4_FEATURE_RO_COMPAT_QUOTA)
+					 EXT4_FEATURE_RO_COMPAT_QUOTA |\
+					 EXT4_FEATURE_RO_COMPAT_PROJECT)
 
 /*
  * Default values for user and/or group using reserved blocks

@@ -1583,6 +1590,11 @@ static inline void ext4_clear_state_flags(struct ext4_inode_info *ei)
 #define	EXT4_DEF_RESUID		0
 #define	EXT4_DEF_RESGID		0
 
+/*
+ * Default project ID
+ */
+#define	EXT4_DEF_PROJID		0
+
 #define EXT4_DEF_INODE_READAHEAD_BLKS	32
 
 /*

@@ -2134,6 +2146,7 @@ extern int ext4_zero_partial_blocks(handle_t *handle, struct inode *inode,
 			     loff_t lstart, loff_t lend);
 extern int ext4_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf);
 extern qsize_t *ext4_get_reserved_space(struct inode *inode);
+extern int ext4_get_projid(struct inode *inode, kprojid_t *projid);
 extern void ext4_da_update_reserve_space(struct inode *inode,
 					int used, int quota_claim);
 

diff --git a/fs/ext4/ialloc.c b/fs/ext4/ialloc.c
index 5b87fc3..b6b7284 100644
--- a/fs/ext4/ialloc.c
+++ b/fs/ext4/ialloc.c

@@ -756,6 +756,12 @@ struct inode *__ext4_new_inode(handle_t *handle, struct inode *dir,
 		inode->i_gid = dir->i_gid;
 	} else
 		inode_init_owner(inode, dir, mode);
+	if (ext4_test_inode_flag(dir, EXT4_INODE_PROJINHERIT)) {
+		ei->i_projid = EXT4_I(dir)->i_projid;
+	} else {
+		ei->i_projid =
+			make_kprojid(&init_user_ns, EXT4_DEF_PROJID);
+	}
 	dquot_initialize(inode);
 
 	if (!goal)

diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index d5dd7d4..8c641eb 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c

@@ -3874,6 +3874,24 @@ static inline void ext4_iget_extra_inode(struct inode *inode,
 		EXT4_I(inode)->i_inline_off = 0;
 }
 
+static projid_t ext4_inode_projid_get(struct inode *inode, struct ext4_inode *raw,
+				 struct ext4_inode_info *ei)
+{
+	return (projid_t)le32_to_cpu(raw->i_projid);
+}
+
+static void ext4_inode_projid_set(struct inode *inode, struct ext4_inode *raw,
+				 struct ext4_inode_info *ei, projid_t projid)
+{
+	raw->i_projid = cpu_to_le32(projid);
+}
+
+int ext4_get_projid(struct inode *inode, kprojid_t *projid)
+{
+	*projid = EXT4_I(inode)->i_projid;
+	return 0;
+}
+
 struct inode *ext4_iget(struct super_block *sb, unsigned long ino)
 {
 	struct ext4_iloc iloc;

@@ -3885,6 +3903,7 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino)
 	int block;
 	uid_t i_uid;
 	gid_t i_gid;
+	projid_t i_projid;
 
 	inode = iget_locked(sb, ino);
 	if (!inode)

@@ -3935,12 +3954,15 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino)
 	inode->i_mode = le16_to_cpu(raw_inode->i_mode);
 	i_uid = (uid_t)le16_to_cpu(raw_inode->i_uid_low);
 	i_gid = (gid_t)le16_to_cpu(raw_inode->i_gid_low);
+	i_projid = ext4_inode_projid_get(inode, raw_inode, ei);
+
 	if (!(test_opt(inode->i_sb, NO_UID32))) {
 		i_uid |= le16_to_cpu(raw_inode->i_uid_high) << 16;
 		i_gid |= le16_to_cpu(raw_inode->i_gid_high) << 16;
 	}
 	i_uid_write(inode, i_uid);
 	i_gid_write(inode, i_gid);
+	ei->i_projid = make_kprojid(&init_user_ns, i_projid);;
 	set_nlink(inode, le16_to_cpu(raw_inode->i_links_count));
 
 	ext4_clear_state_flags(ei);	/* Only relevant on 32-bit archs */

@@ -4163,6 +4185,7 @@ static int ext4_do_update_inode(handle_t *handle,
 	int need_datasync = 0, set_large_file = 0;
 	uid_t i_uid;
 	gid_t i_gid;
+	projid_t i_projid;
 
 	spin_lock(&ei->i_raw_lock);
 

@@ -4175,6 +4198,7 @@ static int ext4_do_update_inode(handle_t *handle,
 	raw_inode->i_mode = cpu_to_le16(inode->i_mode);
 	i_uid = i_uid_read(inode);
 	i_gid = i_gid_read(inode);
+	i_projid = from_kprojid(&init_user_ns, ei->i_projid);
 	if (!(test_opt(inode->i_sb, NO_UID32))) {
 		raw_inode->i_uid_low = cpu_to_le16(low_16_bits(i_uid));
 		raw_inode->i_gid_low = cpu_to_le16(low_16_bits(i_gid));

@@ -4204,7 +4228,8 @@ static int ext4_do_update_inode(handle_t *handle,
 	EXT4_INODE_SET_XTIME(i_atime, inode, raw_inode);
 	EXT4_EINODE_SET_XTIME(i_crtime, ei, raw_inode);
 
-	if (ext4_inode_blocks_set(handle, raw_inode, ei)) {
+	err = ext4_inode_blocks_set(handle, raw_inode, ei);
+	if (err) {
 		spin_unlock(&ei->i_raw_lock);
 		goto out_brelse;
 	}

@@ -4255,6 +4280,8 @@ static int ext4_do_update_inode(handle_t *handle,
 
 	ext4_inode_csum_set(inode, raw_inode, ei);
 
+	ext4_inode_projid_set(inode, raw_inode, ei, i_projid);
+
 	spin_unlock(&ei->i_raw_lock);
 
 	BUFFER_TRACE(bh, "call ext4_handle_dirty_metadata");

diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c
index 51705f8..1a9e557 100644
--- a/fs/ext4/namei.c
+++ b/fs/ext4/namei.c

@@ -2944,6 +2944,11 @@ static int ext4_link(struct dentry *old_dentry,
 	if (inode->i_nlink >= EXT4_LINK_MAX)
 		return -EMLINK;
 
+	if ((ext4_test_inode_flag(dir, EXT4_INODE_PROJINHERIT)) &&
+	    (!projid_eq(EXT4_I(dir)->i_projid,
+			EXT4_I(old_dentry->d_inode)->i_projid)))
+		return -EXDEV;
+
 	dquot_initialize(dir);
 
 retry:

@@ -3186,6 +3191,11 @@ static int ext4_rename(struct inode *old_dir, struct dentry *old_dentry,
 	int force_reread;
 	int retval;
 
+	if ((ext4_test_inode_flag(new_dir, EXT4_INODE_PROJINHERIT)) &&
+	    (!projid_eq(EXT4_I(new_dir)->i_projid,
+		EXT4_I(old_dentry->d_inode)->i_projid)))
+		return -EXDEV;
+
 	dquot_initialize(old.dir);
 	dquot_initialize(new.dir);
 

diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index 2632017..8c3e87b 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c

@@ -1080,6 +1080,7 @@ static const struct dquot_operations ext4_quota_operations = {
 	.write_info	= ext4_write_info,
 	.alloc_dquot	= dquot_alloc,
 	.destroy_dquot	= dquot_destroy,
+	.get_projid	= ext4_get_projid,
 };
 
 static const struct quotactl_ops ext4_qctl_operations = {

-- 
1.7.1

This patch adds ioctl interface for setting/getting project of ext4.

Signed-off-by: Li Xi <redacted>
---
 Documentation/filesystems/ext4.txt |    4 ++
 fs/ext4/ext4.h                     |    2 +
 fs/ext4/ioctl.c                    |   85 ++++++++++++++++++++++++++++++++++++
 3 files changed, 91 insertions(+), 0 deletions(-)

diff --git a/Documentation/filesystems/ext4.txt b/Documentation/filesystems/ext4.txt
index 919a329..9c98e62 100644
--- a/Documentation/filesystems/ext4.txt
+++ b/Documentation/filesystems/ext4.txt

@@ -609,6 +609,10 @@ EXT4_IOC_SWAP_BOOT	      Swap i_blocks and associated attributes
 			      The data blocks of the previous boot loader
 			      will be associated with the given inode.
 
+ EXT4_IOC_GETPROJECT	      Get project ID associated with inode.
+
+ EXT4_IOC_SETPROJECT	      Set Project ID associated with inode.
+
 ..............................................................................
 
 References

diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index f8be9bf..51946fd 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h

@@ -617,6 +617,8 @@ enum {
 #define EXT4_IOC_RESIZE_FS		_IOW('f', 16, __u64)
 #define EXT4_IOC_SWAP_BOOT		_IO('f', 17)
 #define EXT4_IOC_PRECACHE_EXTENTS	_IO('f', 18)
+#define EXT4_IOC_GETPROJECT		_IOR('f', 19, long)
+#define EXT4_IOC_SETPROJECT		_IOW('f', 20, long)
 
 #if defined(__KERNEL__) && defined(CONFIG_COMPAT)
 /*

diff --git a/fs/ext4/ioctl.c b/fs/ext4/ioctl.c
index 0f2252e..93b7ff4 100644
--- a/fs/ext4/ioctl.c
+++ b/fs/ext4/ioctl.c

@@ -14,6 +14,8 @@
 #include <linux/compat.h>
 #include <linux/mount.h>
 #include <linux/file.h>
+#include <linux/quotaops.h>
+#include <linux/quota.h>
 #include <asm/uaccess.h>
 #include "ext4_jbd2.h"
 #include "ext4.h"

@@ -611,6 +613,89 @@ resizefs_out:
 	case EXT4_IOC_PRECACHE_EXTENTS:
 		return ext4_ext_precache(inode);
 
+	case EXT4_IOC_GETPROJECT:
+	{
+		__u32 projid;
+
+		projid = (__u32)from_kprojid(&init_user_ns,
+					     EXT4_I(inode)->i_projid);
+		return put_user(projid, (__u32 __user *) arg);
+	}
+	case EXT4_IOC_SETPROJECT:
+	{
+		__u32 projid;
+		int err;
+		handle_t *handle;
+		kprojid_t kprojid;
+		struct ext4_iloc iloc;
+		struct ext4_inode *raw_inode;
+
+		struct dquot *transfer_to[EXT4_MAXQUOTAS] = { };
+
+		/* Make sure caller can change project. */
+		if (!capable(CAP_SYS_ADMIN))
+			return -EACCES;
+
+		if (get_user(projid, (__u32 __user *) arg))
+			return -EFAULT;
+
+		kprojid = make_kprojid(&init_user_ns, (projid_t)projid);
+
+		if (projid_eq(kprojid, EXT4_I(inode)->i_projid))
+			return 0;
+
+		err = mnt_want_write_file(filp);
+		if (err)
+			return err;
+
+		err = -EPERM;
+		mutex_lock(&inode->i_mutex);
+		/* Is it quota file? Do not allow user to mess with it */
+		if (IS_NOQUOTA(inode))
+			goto project_out;
+
+		dquot_initialize(inode);
+
+		handle = ext4_journal_start(inode, EXT4_HT_QUOTA,
+			EXT4_QUOTA_INIT_BLOCKS(sb) +
+			EXT4_QUOTA_DEL_BLOCKS(sb) + 3);
+		if (IS_ERR(handle)) {
+			err = PTR_ERR(handle);
+			goto project_out;
+		}
+
+		err = ext4_reserve_inode_write(handle, inode, &iloc);
+		if (err)
+			goto project_stop;
+
+		raw_inode = ext4_raw_inode(&iloc);
+		if ((EXT4_INODE_SIZE(inode->i_sb) <=
+		     EXT4_GOOD_OLD_INODE_SIZE) ||
+		    (!EXT4_FITS_IN_INODE(raw_inode, ei, i_projid))) {
+		    	err = -EFBIG;
+		    	goto project_stop;
+		}
+
+		transfer_to[PRJQUOTA] = dqget(sb, make_kqid_projid(kprojid));
+		if (!transfer_to[PRJQUOTA])
+			goto project_set;
+
+		err = __dquot_transfer(inode, transfer_to);
+		dqput(transfer_to[PRJQUOTA]);
+		if (err)
+			goto project_stop;
+
+project_set:
+		EXT4_I(inode)->i_projid = kprojid;
+		inode->i_ctime = ext4_current_time(inode);
+		err = ext4_mark_iloc_dirty(handle, inode, &iloc);
+project_stop:
+		ext4_journal_stop(handle);
+project_out:
+		mutex_unlock(&inode->i_mutex);
+		mnt_drop_write_file(filp);
+		return err;
+	}
 	default:
 		return -ENOTTY;
 	}

-- 
1.7.1

This patch adds support for a new quota type PRJQUOTA for project quota
enforcement. Also a new method get_projid() is added into dquot_operations
structure.

Signed-off-by: Li Xi <redacted>
Signed-off-by: Dmitry Monakhov <redacted>
---
 fs/quota/dquot.c           |   16 ++++++++++++++--
 fs/quota/quota.c           |    5 ++++-
 fs/quota/quotaio_v2.h      |    6 ++++--
 include/linux/quota.h      |    2 ++
 include/uapi/linux/quota.h |    6 ++++--
 5 files changed, 28 insertions(+), 7 deletions(-)

diff --git a/fs/quota/dquot.c b/fs/quota/dquot.c
index f2d0eee..8d0c270 100644
--- a/fs/quota/dquot.c
+++ b/fs/quota/dquot.c

@@ -1154,8 +1154,8 @@ static int need_print_warning(struct dquot_warn *warn)
 			return uid_eq(current_fsuid(), warn->w_dq_id.uid);
 		case GRPQUOTA:
 			return in_group_p(warn->w_dq_id.gid);
-		case PRJQUOTA:	/* Never taken... Just make gcc happy */
-			return 0;
+		case PRJQUOTA:
+			return 1;
 	}
 	return 0;
 }

@@ -1394,6 +1394,9 @@ static void __dquot_initialize(struct inode *inode, int type)
 	/* First get references to structures we might need. */
 	for (cnt = 0; cnt < MAXQUOTAS; cnt++) {
 		struct kqid qid;
+		kprojid_t projid;
+		int rc;
+
 		got[cnt] = NULL;
 		if (type != -1 && cnt != type)
 			continue;

@@ -1413,6 +1416,15 @@ static void __dquot_initialize(struct inode *inode, int type)
 		case GRPQUOTA:
 			qid = make_kqid_gid(inode->i_gid);
 			break;
+		case PRJQUOTA:
+			/* Project ID is not supported */
+			if (!inode->i_sb->dq_op->get_projid)
+				continue;
+			rc = inode->i_sb->dq_op->get_projid(inode, &projid);
+			if (rc)
+				continue;
+			qid = make_kqid_projid(projid);
+			break;
 		}
 		got[cnt] = dqget(sb, qid);
 	}

diff --git a/fs/quota/quota.c b/fs/quota/quota.c
index 7562164..cce7371 100644
--- a/fs/quota/quota.c
+++ b/fs/quota/quota.c

@@ -30,7 +30,10 @@ static int check_quotactl_permission(struct super_block *sb, int type, int cmd,
 	case Q_XGETQSTATV:
 	case Q_XQUOTASYNC:
 		break;
-	/* allow to query information for dquots we "own" */
+	/*
+	 * allow to query information for dquots we "own"
+	 * always allow quota check for project quota
+	 */
 	case Q_GETQUOTA:
 	case Q_XGETQUOTA:
 		if ((type == USRQUOTA && uid_eq(current_euid(), make_kuid(current_user_ns(), id))) ||

diff --git a/fs/quota/quotaio_v2.h b/fs/quota/quotaio_v2.h
index f1966b4..b17af05 100644
--- a/fs/quota/quotaio_v2.h
+++ b/fs/quota/quotaio_v2.h

@@ -13,12 +13,14 @@
  */
 #define V2_INITQMAGICS {\
 	0xd9c01f11,	/* USRQUOTA */\
-	0xd9c01927	/* GRPQUOTA */\
+	0xd9c01927,	/* GRPQUOTA */\
+	0xd9c03f14	/* PRJQUOTA */\
 }
 
 #define V2_INITQVERSIONS {\
 	1,		/* USRQUOTA */\
-	1		/* GRPQUOTA */\
+	1,		/* GRPQUOTA */\
+	1		/* PRJQUOTA */\
 }
 
 /* First generic header */

diff --git a/include/linux/quota.h b/include/linux/quota.h
index 80d345a..f1b25f8 100644
--- a/include/linux/quota.h
+++ b/include/linux/quota.h

@@ -50,6 +50,7 @@
 
 #undef USRQUOTA
 #undef GRPQUOTA
+#undef PRJQUOTA
 enum quota_type {
 	USRQUOTA = 0,		/* element used for user quotas */
 	GRPQUOTA = 1,		/* element used for group quotas */

@@ -312,6 +313,7 @@ struct dquot_operations {
 	/* get reserved quota for delayed alloc, value returned is managed by
 	 * quota code only */
 	qsize_t *(*get_reserved_space) (struct inode *);
+	int (*get_projid) (struct inode *, kprojid_t *);/* Get project ID */
 };
 
 struct path;

diff --git a/include/uapi/linux/quota.h b/include/uapi/linux/quota.h
index 3b6cfbe..b2d9486 100644
--- a/include/uapi/linux/quota.h
+++ b/include/uapi/linux/quota.h

@@ -36,11 +36,12 @@
 #include <linux/errno.h>
 #include <linux/types.h>
 
-#define __DQUOT_VERSION__	"dquot_6.5.2"
+#define __DQUOT_VERSION__	"dquot_6.6.0"
 
-#define MAXQUOTAS 2
+#define MAXQUOTAS 3
 #define USRQUOTA  0		/* element used for user quotas */
 #define GRPQUOTA  1		/* element used for group quotas */
+#define PRJQUOTA  2		/* element used for project quotas */
 
 /*
  * Definitions for the default names of the quotas files.

@@ -48,6 +49,7 @@
 #define INITQFNAMES { \
 	"user",    /* USRQUOTA */ \
 	"group",   /* GRPQUOTA */ \
+	"project", /* PRJQUOTA */ \
 	"undefined", \
 };
 

-- 
1.7.1

This patch adds mount options for enabling/disabling project quota
accounting and enforcement. A new specific inode is also used for
project quota accounting.

Signed-off-by: Li Xi <redacted>
Signed-off-by: Dmitry Monakhov <redacted>
---
 fs/ext4/ext4.h  |    8 +++-
 fs/ext4/super.c |  123 ++++++++++++++++++++++++++++++++++++++++++++++++++-----
 2 files changed, 118 insertions(+), 13 deletions(-)

diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index 092f60c..f8be9bf 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h

@@ -217,6 +217,7 @@ struct ext4_io_submit {
 #define EXT4_UNDEL_DIR_INO	 6	/* Undelete directory inode */
 #define EXT4_RESIZE_INO		 7	/* Reserved group descriptors inode */
 #define EXT4_JOURNAL_INO	 8	/* Journal inode */
+#define EXT4_PRJ_QUOTA_INO	 9	/* Project quota inode */
 
 /* First non-reserved inode for old ext4 filesystems */
 #define EXT4_GOOD_OLD_FIRST_INO	11

@@ -991,6 +992,7 @@ struct ext4_inode_info {
 #define EXT4_MOUNT_DIOREAD_NOLOCK	0x400000 /* Enable support for dio read nolocking */
 #define EXT4_MOUNT_JOURNAL_CHECKSUM	0x800000 /* Journal checksums */
 #define EXT4_MOUNT_JOURNAL_ASYNC_COMMIT	0x1000000 /* Journal Async Commit */
+#define EXT4_MOUNT_PRJQUOTA		0x2000000 /* Project quota support */
 #define EXT4_MOUNT_DELALLOC		0x8000000 /* Delalloc support */
 #define EXT4_MOUNT_DATA_ERR_ABORT	0x10000000 /* Abort on file data write */
 #define EXT4_MOUNT_BLOCK_VALIDITY	0x20000000 /* Block validity checking */

@@ -1166,7 +1168,8 @@ struct ext4_super_block {
 	__le32	s_grp_quota_inum;	/* inode for tracking group quota */
 	__le32	s_overhead_clusters;	/* overhead blocks/clusters in fs */
 	__le32	s_backup_bgs[2];	/* groups with sparse_super2 SBs */
-	__le32	s_reserved[106];	/* Padding to the end of the block */
+	__le32	s_prj_quota_inum;	/* inode for tracking project quota */
+	__le32	s_reserved[105];	/* Padding to the end of the block */
 	__le32	s_checksum;		/* crc32c(superblock) */
 };
 

@@ -1181,7 +1184,7 @@ struct ext4_super_block {
 #define EXT4_MF_FS_ABORTED	0x0002	/* Fatal error detected */
 
 /* Number of quota types we support */
-#define EXT4_MAXQUOTAS 2
+#define EXT4_MAXQUOTAS 3
 
 /*
  * fourth extended-fs super-block data in memory

@@ -1372,6 +1375,7 @@ static inline int ext4_valid_inum(struct super_block *sb, unsigned long ino)
 		ino == EXT4_BOOT_LOADER_INO ||
 		ino == EXT4_JOURNAL_INO ||
 		ino == EXT4_RESIZE_INO ||
+		ino == EXT4_PRJ_QUOTA_INO ||
 		(ino >= EXT4_FIRST_INO(sb) &&
 		 ino <= le32_to_cpu(EXT4_SB(sb)->s_es->s_inodes_count));
 }

diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index 8c3e87b..5cb5055 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c

@@ -1048,8 +1048,8 @@ static int bdev_try_to_free_page(struct super_block *sb, struct page *page,
 }
 
 #ifdef CONFIG_QUOTA
-#define QTYPE2NAME(t) ((t) == USRQUOTA ? "user" : "group")
-#define QTYPE2MOPT(on, t) ((t) == USRQUOTA?((on)##USRJQUOTA):((on)##GRPJQUOTA))
+static char *quotatypes[] = INITQFNAMES;
+#define QTYPE2NAME(t) (quotatypes[t])
 
 static int ext4_write_dquot(struct dquot *dquot);
 static int ext4_acquire_dquot(struct dquot *dquot);

@@ -1160,10 +1160,11 @@ enum {
 	Opt_journal_path, Opt_journal_checksum, Opt_journal_async_commit,
 	Opt_abort, Opt_data_journal, Opt_data_ordered, Opt_data_writeback,
 	Opt_data_err_abort, Opt_data_err_ignore,
-	Opt_usrjquota, Opt_grpjquota, Opt_offusrjquota, Opt_offgrpjquota,
+	Opt_usrjquota, Opt_grpjquota, Opt_prjjquota,
+	Opt_offusrjquota, Opt_offgrpjquota, Opt_offprjjquota,
 	Opt_jqfmt_vfsold, Opt_jqfmt_vfsv0, Opt_jqfmt_vfsv1, Opt_quota,
 	Opt_noquota, Opt_barrier, Opt_nobarrier, Opt_err,
-	Opt_usrquota, Opt_grpquota, Opt_i_version,
+	Opt_usrquota, Opt_grpquota, Opt_prjquota, Opt_i_version,
 	Opt_stripe, Opt_delalloc, Opt_nodelalloc, Opt_mblk_io_submit,
 	Opt_nomblk_io_submit, Opt_block_validity, Opt_noblock_validity,
 	Opt_inode_readahead_blks, Opt_journal_ioprio,

@@ -1214,6 +1215,8 @@ static const match_table_t tokens = {
 	{Opt_usrjquota, "usrjquota=%s"},
 	{Opt_offgrpjquota, "grpjquota="},
 	{Opt_grpjquota, "grpjquota=%s"},
+	{Opt_offprjjquota, "prjjquota="},
+	{Opt_prjjquota, "prjjquota=%s"},
 	{Opt_jqfmt_vfsold, "jqfmt=vfsold"},
 	{Opt_jqfmt_vfsv0, "jqfmt=vfsv0"},
 	{Opt_jqfmt_vfsv1, "jqfmt=vfsv1"},

@@ -1221,6 +1224,7 @@ static const match_table_t tokens = {
 	{Opt_noquota, "noquota"},
 	{Opt_quota, "quota"},
 	{Opt_usrquota, "usrquota"},
+	{Opt_prjquota, "prjquota"},
 	{Opt_barrier, "barrier=%u"},
 	{Opt_barrier, "barrier"},
 	{Opt_nobarrier, "nobarrier"},

@@ -1433,6 +1437,8 @@ static const struct mount_opts {
 							MOPT_SET | MOPT_Q},
 	{Opt_grpquota, EXT4_MOUNT_QUOTA | EXT4_MOUNT_GRPQUOTA,
 							MOPT_SET | MOPT_Q},
+	{Opt_prjquota, EXT4_MOUNT_QUOTA | EXT4_MOUNT_PRJQUOTA,
+							MOPT_SET | MOPT_Q},
 	{Opt_noquota, (EXT4_MOUNT_QUOTA | EXT4_MOUNT_USRQUOTA |
 		       EXT4_MOUNT_GRPQUOTA), MOPT_CLEAR | MOPT_Q},
 	{Opt_usrjquota, 0, MOPT_Q},

@@ -1461,10 +1467,14 @@ static int handle_mount_opt(struct super_block *sb, char *opt, int token,
 		return set_qf_name(sb, USRQUOTA, &args[0]);
 	else if (token == Opt_grpjquota)
 		return set_qf_name(sb, GRPQUOTA, &args[0]);
+	else if (token == Opt_prjjquota)
+		return set_qf_name(sb, PRJQUOTA, &args[0]);
 	else if (token == Opt_offusrjquota)
 		return clear_qf_name(sb, USRQUOTA);
 	else if (token == Opt_offgrpjquota)
 		return clear_qf_name(sb, GRPQUOTA);
+	else if (token == Opt_offprjjquota)
+		return clear_qf_name(sb, PRJQUOTA);
 #endif
 	switch (token) {
 	case Opt_noacl:

@@ -1690,19 +1700,28 @@ static int parse_options(char *options, struct super_block *sb,
 	}
 #ifdef CONFIG_QUOTA
 	if (EXT4_HAS_RO_COMPAT_FEATURE(sb, EXT4_FEATURE_RO_COMPAT_QUOTA) &&
-	    (test_opt(sb, USRQUOTA) || test_opt(sb, GRPQUOTA))) {
+	    (test_opt(sb, USRQUOTA) ||
+	     test_opt(sb, GRPQUOTA) ||
+	     test_opt(sb, PRJQUOTA))) {
 		ext4_msg(sb, KERN_ERR, "Cannot set quota options when QUOTA "
 			 "feature is enabled");
 		return 0;
 	}
-	if (sbi->s_qf_names[USRQUOTA] || sbi->s_qf_names[GRPQUOTA]) {
+	if (sbi->s_qf_names[USRQUOTA] ||
+	    sbi->s_qf_names[GRPQUOTA] ||
+	    sbi->s_qf_names[PRJQUOTA]) {
 		if (test_opt(sb, USRQUOTA) && sbi->s_qf_names[USRQUOTA])
 			clear_opt(sb, USRQUOTA);
 
 		if (test_opt(sb, GRPQUOTA) && sbi->s_qf_names[GRPQUOTA])
 			clear_opt(sb, GRPQUOTA);
 
-		if (test_opt(sb, GRPQUOTA) || test_opt(sb, USRQUOTA)) {
+		if (test_opt(sb, PRJQUOTA) && sbi->s_qf_names[PRJQUOTA])
+			clear_opt(sb, PRJQUOTA);
+
+		if (test_opt(sb, GRPQUOTA) ||
+		    test_opt(sb, USRQUOTA) ||
+		    test_opt(sb, PRJQUOTA)) {
 			ext4_msg(sb, KERN_ERR, "old and new quota "
 					"format mixing");
 			return 0;

@@ -1763,6 +1782,9 @@ static inline void ext4_show_quota_options(struct seq_file *seq,
 
 	if (sbi->s_qf_names[GRPQUOTA])
 		seq_printf(seq, ",grpjquota=%s", sbi->s_qf_names[GRPQUOTA]);
+
+	if (sbi->s_qf_names[PRJQUOTA])
+		seq_printf(seq, ",prjjquota=%s", sbi->s_qf_names[PRJQUOTA]);
 #endif
 }
 

@@ -2833,6 +2855,13 @@ static int ext4_feature_set_ok(struct super_block *sb, int readonly)
 			 "without CONFIG_QUOTA");
 		return 0;
 	}
+	if (EXT4_HAS_RO_COMPAT_FEATURE(sb, EXT4_FEATURE_RO_COMPAT_PROJECT) &&
+	    !readonly) {
+		ext4_msg(sb, KERN_ERR,
+			 "Filesystem with project quota feature cannot be"
+			 "mounted RDWR without CONFIG_QUOTA");
+		return 0;
+	}
 #endif  /* CONFIG_QUOTA */
 	return 1;
 }

@@ -5052,6 +5081,46 @@ restore_opts:
 	return err;
 }
 
+static int ext4_statfs_project(struct super_block *sb,
+			       kprojid_t projid, struct kstatfs *buf)
+{
+	struct kqid qid;
+	struct dquot *dquot;
+	u64 limit;
+	u64 curblock;
+
+	qid = make_kqid_projid(projid);
+	dquot = dqget(sb, qid);
+	if (!dquot)
+		return -ESRCH;
+	spin_lock(&dq_data_lock);
+
+	limit = dquot->dq_dqb.dqb_bsoftlimit ?
+		dquot->dq_dqb.dqb_bsoftlimit :
+		dquot->dq_dqb.dqb_bhardlimit;
+	if (limit && buf->f_blocks * buf->f_bsize > limit) {
+		curblock = dquot->dq_dqb.dqb_curspace / buf->f_bsize;
+		buf->f_blocks = limit / buf->f_bsize;
+		buf->f_bfree = buf->f_bavail =
+			(buf->f_blocks > curblock) ?
+			 (buf->f_blocks - curblock) : 0;
+	}
+
+	limit = dquot->dq_dqb.dqb_isoftlimit ?
+		dquot->dq_dqb.dqb_isoftlimit :
+		dquot->dq_dqb.dqb_ihardlimit;
+	if (limit && buf->f_files > limit) {
+		buf->f_files = limit;
+		buf->f_ffree =
+			(buf->f_files > dquot->dq_dqb.dqb_curinodes) ?
+			 (buf->f_ffree - dquot->dq_dqb.dqb_curinodes) : 0;
+	}
+
+	spin_unlock(&dq_data_lock);
+	dqput(dquot);
+	return 0;
+}
+
 static int ext4_statfs(struct dentry *dentry, struct kstatfs *buf)
 {
 	struct super_block *sb = dentry->d_sb;

@@ -5060,6 +5129,8 @@ static int ext4_statfs(struct dentry *dentry, struct kstatfs *buf)
 	ext4_fsblk_t overhead = 0, resv_blocks;
 	u64 fsid;
 	s64 bfree;
+	struct inode *inode = dentry->d_inode;
+	int err = 0;
 	resv_blocks = EXT4_C2B(sbi, atomic64_read(&sbi->s_resv_clusters));
 
 	if (!test_opt(sb, MINIX_DF))

@@ -5084,7 +5155,18 @@ static int ext4_statfs(struct dentry *dentry, struct kstatfs *buf)
 	buf->f_fsid.val[0] = fsid & 0xFFFFFFFFUL;
 	buf->f_fsid.val[1] = (fsid >> 32) & 0xFFFFFFFFUL;
 
-	return 0;
+	if (EXT4_HAS_RO_COMPAT_FEATURE(sb, EXT4_FEATURE_RO_COMPAT_PROJECT) &&
+	    ext4_test_inode_flag(inode, EXT4_INODE_PROJINHERIT) &&
+	    sb_has_quota_usage_enabled(sb, PRJQUOTA) &&
+	    sb_has_quota_limits_enabled(sb, PRJQUOTA)) {
+	    	err = ext4_statfs_project(sb, EXT4_I(inode)->i_projid, buf);
+	    	if (err) {
+			ext4_warning(sb, "Cannot get quota of project %u\n",
+				     from_kprojid(&init_user_ns,
+						  EXT4_I(inode)->i_projid));
+	    	}
+	}
+	return err;
 }
 
 /* Helper function for writing quotas on sync - we need to start transaction

@@ -5164,7 +5246,9 @@ static int ext4_mark_dquot_dirty(struct dquot *dquot)
 
 	/* Are we journaling quotas? */
 	if (EXT4_HAS_RO_COMPAT_FEATURE(sb, EXT4_FEATURE_RO_COMPAT_QUOTA) ||
-	    sbi->s_qf_names[USRQUOTA] || sbi->s_qf_names[GRPQUOTA]) {
+	    sbi->s_qf_names[USRQUOTA] ||
+	    sbi->s_qf_names[GRPQUOTA] ||
+	    sbi->s_qf_names[PRJQUOTA]) {
 		dquot_mark_dquot_dirty(dquot);
 		return ext4_write_dquot(dquot);
 	} else {

@@ -5248,10 +5332,14 @@ static int ext4_quota_enable(struct super_block *sb, int type, int format_id,
 	struct inode *qf_inode;
 	unsigned long qf_inums[EXT4_MAXQUOTAS] = {
 		le32_to_cpu(EXT4_SB(sb)->s_es->s_usr_quota_inum),
-		le32_to_cpu(EXT4_SB(sb)->s_es->s_grp_quota_inum)
+		le32_to_cpu(EXT4_SB(sb)->s_es->s_grp_quota_inum),
+		le32_to_cpu(EXT4_SB(sb)->s_es->s_prj_quota_inum)
 	};
 
 	BUG_ON(!EXT4_HAS_RO_COMPAT_FEATURE(sb, EXT4_FEATURE_RO_COMPAT_QUOTA));
+	BUG_ON(type == PRJQUOTA &&
+	       (!EXT4_HAS_RO_COMPAT_FEATURE(sb,
+				EXT4_FEATURE_RO_COMPAT_PROJECT)));
 
 	if (!qf_inums[type])
 		return -EPERM;

@@ -5276,11 +5364,16 @@ static int ext4_enable_quotas(struct super_block *sb)
 	int type, err = 0;
 	unsigned long qf_inums[EXT4_MAXQUOTAS] = {
 		le32_to_cpu(EXT4_SB(sb)->s_es->s_usr_quota_inum),
-		le32_to_cpu(EXT4_SB(sb)->s_es->s_grp_quota_inum)
+		le32_to_cpu(EXT4_SB(sb)->s_es->s_grp_quota_inum),
+		le32_to_cpu(EXT4_SB(sb)->s_es->s_prj_quota_inum)
 	};
 
 	sb_dqopt(sb)->flags |= DQUOT_QUOTA_SYS_FILE;
 	for (type = 0; type < EXT4_MAXQUOTAS; type++) {
+		if (type == PRJQUOTA &&
+		    (!EXT4_HAS_RO_COMPAT_FEATURE(sb,
+				EXT4_FEATURE_RO_COMPAT_PROJECT)))
+			continue;
 		if (qf_inums[type]) {
 			err = ext4_quota_enable(sb, type, QFMT_VFS_V1,
 						DQUOT_USAGE_ENABLED);

@@ -5305,6 +5398,10 @@ static int ext4_quota_on_sysfile(struct super_block *sb, int type,
 	if (!EXT4_HAS_RO_COMPAT_FEATURE(sb, EXT4_FEATURE_RO_COMPAT_QUOTA))
 		return -EINVAL;
 
+	if (type == PRJQUOTA &&
+	    (!EXT4_HAS_RO_COMPAT_FEATURE(sb, EXT4_FEATURE_RO_COMPAT_PROJECT)))
+		return -EINVAL;
+
 	/*
 	 * USAGE was enabled at mount time. Only need to enable LIMITS now.
 	 */

@@ -5345,6 +5442,10 @@ static int ext4_quota_off_sysfile(struct super_block *sb, int type)
 	if (!EXT4_HAS_RO_COMPAT_FEATURE(sb, EXT4_FEATURE_RO_COMPAT_QUOTA))
 		return -EINVAL;
 
+	if (type == PRJQUOTA &&
+	    (!EXT4_HAS_RO_COMPAT_FEATURE(sb, EXT4_FEATURE_RO_COMPAT_PROJECT)))
+		return -EINVAL;
+
 	/* Disable only the limits. */
 	return dquot_disable(sb, type, DQUOT_LIMITS_ENABLED);
 }

-- 
1.7.1

On Wed 24-09-14 22:04:27, Li Xi wrote:

  Looking into dquot.c you seem to be missing a change to do_get_dqblk()
where setting of di->d_flags needs to handle PRJQUOTA now as well.

  The last sentense should probably be:
"always allow querying project quota"

+	 */
 	case Q_GETQUOTA:
 	case Q_XGETQUOTA:
 		if ((type == USRQUOTA && uid_eq(current_euid(), make_kuid(current_user_ns(), id))) ||

  But looking at the code, you need to add there

|| type == PRJQUOTA

  don't you?


								Honza
-- 
Jan Kara [off-list ref]
SUSE Labs, CR

On Wed 24-09-14 22:04:30, Li Xi wrote:

This patch adds ioctl interface for setting/getting project of ext4.

  The patch looks good to me. I was just wondering whether it won't be
useful to add an ioctl() which isn't ext4 specific. We could just extend
->setattr() to allow setting of project ID (most filesystems would just
return -EOPNOTSUPP but ext4 and xfs could do the right thing) and then call
->setattr from the generic ioctl. That way userspace won't have to care
about filesystem type when setting project ID... What do others think?

								Honza

-- 
Jan Kara [off-list ref]
SUSE Labs, CR

On Wed, Sep 24, 2014 at 06:25:07PM +0200, Jan Kara wrote:

On Wed 24-09-14 22:04:30, Li Xi wrote:

quoted

This patch adds ioctl interface for setting/getting project of ext4.

  The patch looks good to me. I was just wondering whether it won't be
useful to add an ioctl() which isn't ext4 specific. We could just extend
->setattr() to allow setting of project ID (most filesystems would just
return -EOPNOTSUPP but ext4 and xfs could do the right thing) and then call
->setattr from the generic ioctl. That way userspace won't have to care
about filesystem type when setting project ID... What do others think?

Absolutely.  In general I also wonder why this patch doesn't implement
the full XFS API.  Maybe there is a reason it was considered and
rejected, but it would be helpful to document why.

On Wed 24-09-14 09:26:34, Christoph Hellwig wrote:

On Wed, Sep 24, 2014 at 06:25:07PM +0200, Jan Kara wrote:

quoted

On Wed 24-09-14 22:04:30, Li Xi wrote:

quoted

This patch adds ioctl interface for setting/getting project of ext4.

  The patch looks good to me. I was just wondering whether it won't be
useful to add an ioctl() which isn't ext4 specific. We could just extend
->setattr() to allow setting of project ID (most filesystems would just
return -EOPNOTSUPP but ext4 and xfs could do the right thing) and then call
->setattr from the generic ioctl. That way userspace won't have to care
about filesystem type when setting project ID... What do others think?

Absolutely.  In general I also wonder why this patch doesn't implement
the full XFS API.  Maybe there is a reason it was considered and
rejected, but it would be helpful to document why.

  Do you mean full get/setfsxattr API? That basically contains project ID,
flags (those that are currently get/set with FS_IOC_GETFLAGS/SETFLAGS), and
extent size hint right? That seems workable and it would also make setting
of PROJINHERIT flag fs agnostic. Only we would have to create some generic
flags namespace and merge into that ext4 flags and have a translation
function for the old ext4 flags. Also I'm afraid we may quickly run out of
32 available flags in xflags so we'd need to extend that. But all this
seems to be doable.
								Honza
-- 
Jan Kara [off-list ref]
SUSE Labs, CR

On Wed 24-09-14 22:04:28, Li Xi wrote:

This patch adds a new internal field of ext4 inode to save project
identifier. Also a new flag EXT4_INODE_PROJINHERIT is added for
inheriting project ID from parent directory.

  Nice. Just a few smaller things below.

  Firstly, I think your patch forgets to add a check to EXT4_IOC_SETFLAGS
ioctl so that EXT4_INODE_PROJINHERIT flag can be set only when
EXT4_FEATURE_RO_COMPAT_PROJECT is enabled.

Signed-off-by: Li Xi <redacted>
---
 fs/ext4/ext4.h   |   21 +++++++++++++++++----
 fs/ext4/ialloc.c |    6 ++++++
 fs/ext4/inode.c  |   29 ++++++++++++++++++++++++++++-
 fs/ext4/namei.c  |   10 ++++++++++
 fs/ext4/super.c  |    1 +
 5 files changed, 62 insertions(+), 5 deletions(-)

...

  I don't see a need for these two wrappers. IMHO they just obfuscate the
code. Please just write the code directly in appropriate place.

  This is an unrelated fix, right? Please send as a separate patch. Thanks!

  This is wrong - you have to set project ID before a checksum is computed.

								Honza
-- 
Jan Kara [off-list ref]
SUSE Labs, CR

On Thu, Sep 25, 2014 at 1:11 AM, Jan Kara [off-list ref] wrote:

On Wed 24-09-14 22:04:28, Li Xi wrote:

quoted

This patch adds a new internal field of ext4 inode to save project
identifier. Also a new flag EXT4_INODE_PROJINHERIT is added for
inheriting project ID from parent directory.

  Nice. Just a few smaller things below.

  Firstly, I think your patch forgets to add a check to EXT4_IOC_SETFLAGS
ioctl so that EXT4_INODE_PROJINHERIT flag can be set only when
EXT4_FEATURE_RO_COMPAT_PROJECT is enabled.

PROJINHERIT is a flag which affects the project ID inheriting. Is there
any special reason that this flag can be set/cleared only if
EXT4_FEATURE_RO_COMPAT_PROJECT is enabled? We don't check
that feature even when set/get project ID now, so I am not so sure
that this is consistent.

quoted

Signed-off-by: Li Xi <redacted>
---
 fs/ext4/ext4.h   |   21 +++++++++++++++++----
 fs/ext4/ialloc.c |    6 ++++++
 fs/ext4/inode.c  |   29 ++++++++++++++++++++++++++++-
 fs/ext4/namei.c  |   10 ++++++++++
 fs/ext4/super.c  |    1 +
 5 files changed, 62 insertions(+), 5 deletions(-)

...

quoted

diff --git a/fs/ext4/ialloc.c b/fs/ext4/ialloc.c
index 5b87fc3..b6b7284 100644
--- a/fs/ext4/ialloc.c
+++ b/fs/ext4/ialloc.c

@@ -756,6 +756,12 @@ struct inode *__ext4_new_inode(handle_t *handle, struct inode *dir,
              inode->i_gid = dir->i_gid;
      } else
              inode_init_owner(inode, dir, mode);
+     if (ext4_test_inode_flag(dir, EXT4_INODE_PROJINHERIT)) {
+             ei->i_projid = EXT4_I(dir)->i_projid;
+     } else {
+             ei->i_projid =
+                     make_kprojid(&init_user_ns, EXT4_DEF_PROJID);
+     }
      dquot_initialize(inode);

      if (!goal)

diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index d5dd7d4..8c641eb 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c

@@ -3874,6 +3874,24 @@ static inline void ext4_iget_extra_inode(struct inode *inode,
              EXT4_I(inode)->i_inline_off = 0;
 }

+static projid_t ext4_inode_projid_get(struct inode *inode, struct ext4_inode *raw,
+                              struct ext4_inode_info *ei)
+{
+     return (projid_t)le32_to_cpu(raw->i_projid);
+}
+
+static void ext4_inode_projid_set(struct inode *inode, struct ext4_inode *raw,
+                              struct ext4_inode_info *ei, projid_t projid)
+{
+     raw->i_projid = cpu_to_le32(projid);
+}

  I don't see a need for these two wrappers. IMHO they just obfuscate the
code. Please just write the code directly in appropriate place.

Sure.

quoted

@@ -4204,7 +4228,8 @@ static int ext4_do_update_inode(handle_t *handle,
      EXT4_INODE_SET_XTIME(i_atime, inode, raw_inode);
      EXT4_EINODE_SET_XTIME(i_crtime, ei, raw_inode);

-     if (ext4_inode_blocks_set(handle, raw_inode, ei)) {
+     err = ext4_inode_blocks_set(handle, raw_inode, ei);
+     if (err) {
              spin_unlock(&ei->i_raw_lock);
              goto out_brelse;
      }

  This is an unrelated fix, right? Please send as a separate patch. Thanks!

No problem!

quoted

@@ -4255,6 +4280,8 @@ static int ext4_do_update_inode(handle_t *handle,

      ext4_inode_csum_set(inode, raw_inode, ei);

+     ext4_inode_projid_set(inode, raw_inode, ei, i_projid);
+

  This is wrong - you have to set project ID before a checksum is computed.

Yeah, sorry for this mistake.

Regards,
                                     - Li Xi

On Wed, Sep 24, 2014 at 06:25:07PM +0200, Jan Kara wrote:

On Wed 24-09-14 22:04:30, Li Xi wrote:

quoted

This patch adds ioctl interface for setting/getting project of ext4.

  The patch looks good to me. I was just wondering whether it won't be
useful to add an ioctl() which isn't ext4 specific. We could just extend
->setattr() to allow setting of project ID (most filesystems would just
return -EOPNOTSUPP but ext4 and xfs could do the right thing) and then call
->setattr from the generic ioctl. That way userspace won't have to care
about filesystem type when setting project ID... What do others think?

I've repeatedly said that these ext4 project ID patches should
implement the same interfaces as XFS rather than creating a new set
of incompatible, ext4 specific interfaces to do implement the same
functionality.

There is no good reason for forcing userspace to re-invent tools
that already exist just to manage identical functionality in
different filesystems.

Cheers,

Dave.
-- 
Dave Chinner
david@fromorbit.com

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

On Wed, Sep 24, 2014 at 07:01:05PM +0200, Jan Kara wrote:

On Wed 24-09-14 09:26:34, Christoph Hellwig wrote:

quoted

On Wed, Sep 24, 2014 at 06:25:07PM +0200, Jan Kara wrote:

quoted

On Wed 24-09-14 22:04:30, Li Xi wrote:

quoted

This patch adds ioctl interface for setting/getting project of ext4.

  The patch looks good to me. I was just wondering whether it won't be
useful to add an ioctl() which isn't ext4 specific. We could just extend
->setattr() to allow setting of project ID (most filesystems would just
return -EOPNOTSUPP but ext4 and xfs could do the right thing) and then call
->setattr from the generic ioctl. That way userspace won't have to care
about filesystem type when setting project ID... What do others think?

Absolutely.  In general I also wonder why this patch doesn't implement
the full XFS API.  Maybe there is a reason it was considered and
rejected, but it would be helpful to document why.

  Do you mean full get/setfsxattr API?

That's a good start.

The bigger issue in my mind is that we already have a fully featured
quota API that supports project quotas and userspace tools available
that manipulate it. xfstests already uses those tools and API
for testing project quotas.

This whole patchset reinvents all the quota APIs, and will require
adding support in userspace, and hence require re-inventing all the
test infrastructure we already have because it won't be compatible
with the existing project quota test code.

  That basically contains project ID,
flags (those that are currently get/set with FS_IOC_GETFLAGS/SETFLAGS), and
extent size hint right?

It's a different set of flag definitions. We translate the interface
XFS_XFLAG_* values to/from the inode on-disk XFS_DIFLAG_* inode values, just
like we translate the VFS FS_*_FL flags that get passed through the
FS_IOC_GETFLAGS/SETFLAGS ioctl.

That seems workable and it would also make setting
of PROJINHERIT flag fs agnostic. Only we would have to create some generic
flags namespace and merge into that ext4 flags and have a translation
function for the old ext4 flags.

The XFS_XFLAGS_* are already filesystem agnostic - they are flags
that are only used for interfacing with userspace and hence only
exist at the ioctl copy in/out layer.....

Also I'm afraid we may quickly run out of
32 available flags in xflags so we'd need to extend that. But all this
seems to be doable.

The struct fsxattr was designed to be extensible - it has unused
padding and enough space in the flags field to allow us to
conditionally use that padding....

Cheers,

Dave.
-- 
Dave Chinner
david@fromorbit.com

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

在 2014年9月25日，下午3:59，Dave Chinner [off-list ref] 写道：

On Wed, Sep 24, 2014 at 07:01:05PM +0200, Jan Kara wrote:

quoted

On Wed 24-09-14 09:26:34, Christoph Hellwig wrote:

quoted

On Wed, Sep 24, 2014 at 06:25:07PM +0200, Jan Kara wrote:

quoted

On Wed 24-09-14 22:04:30, Li Xi wrote:

quoted

This patch adds ioctl interface for setting/getting project of ext4.

 The patch looks good to me. I was just wondering whether it won't be
useful to add an ioctl() which isn't ext4 specific. We could just extend
->setattr() to allow setting of project ID (most filesystems would just
return -EOPNOTSUPP but ext4 and xfs could do the right thing) and then call
->setattr from the generic ioctl. That way userspace won't have to care
about filesystem type when setting project ID... What do others think?

Absolutely.  In general I also wonder why this patch doesn't implement
the full XFS API.  Maybe there is a reason it was considered and
rejected, but it would be helpful to document why.

 Do you mean full get/setfsxattr API?

That's a good start.

The bigger issue in my mind is that we already have a fully featured
quota API that supports project quotas and userspace tools available
that manipulate it. xfstests already uses those tools and API
for testing project quotas.

This whole patchset reinvents all the quota APIs, and will require
adding support in userspace, and hence require re-inventing all the
test infrastructure we already have because it won't be compatible
with the existing project quota test code.

quoted

 That basically contains project ID,
flags (those that are currently get/set with FS_IOC_GETFLAGS/SETFLAGS), and
extent size hint right?

It's a different set of flag definitions. We translate the interface
XFS_XFLAG_* values to/from the inode on-disk XFS_DIFLAG_* inode values, just
like we translate the VFS FS_*_FL flags that get passed through the
FS_IOC_GETFLAGS/SETFLAGS ioctl.

quoted

That seems workable and it would also make setting
of PROJINHERIT flag fs agnostic. Only we would have to create some generic
flags namespace and merge into that ext4 flags and have a translation
function for the old ext4 flags.

The XFS_XFLAGS_* are already filesystem agnostic - they are flags
that are only used for interfacing with userspace and hence only
exist at the ioctl copy in/out layer.....

quoted

Also I'm afraid we may quickly run out of
32 available flags in xflags so we'd need to extend that. But all this
seems to be doable.

The struct fsxattr was designed to be extensible - it has unused
padding and enough space in the flags field to allow us to
conditionally use that padding….

Hi Dave,

I was mostly working on the semantics of inherit flag on these patches. And
I didn’t realized that the interface differences would bother you so much. Sorry
for that.

I agree that we should choose a good interface. It would be good that it is
general so that it works well for all file systems. I agree that adding an
ext4 specific ioctl() is far from the best choice. I am willing to change it to
any general interface. A general ioctl() sounds good to me. Extend setattr()
and getattr() for project ID sounds even better, since project ID looks like
UID/GID. And general xattr name is another choice. But it is might be a little
bit confusing if we use xattr actually, since we are not saving project ID as
extended attribute on Ext4. Any choice is fine with me, as long as the
implementation won't introduce nasty codes or inconsistent design.

However, the problem is, I do not quite understand why we should keep
the interface exactly the same with XFS. It would be good if we can. But
as far as I can see, it seems hard. XFS uses a lot interfaces which are
not so standard and used by other file systems. For example, struct
fsxattr is not used by other file systems at all except XFS. I am not sure
why we should introduce this into Ext4 if there are a lot of other better
ways. I would be happy to change to XFS interfaces, if it is general.
However, I don’t think it is general enough.

I know xfstest is using the existing project quota interfaces of XFS. And
maybe there are some applications which are using them too. But
keeping the interfaces exactly the same with XFS would cost so much
effort that I’d like to get enough reasons before start working on it. Is it
really necessary? I am not so sure. It is so easy to change user
space applications comparing to changing a weird interfaces. For
example, I think it won’t cost even more than a day to add xfstest
support for new Ext4 project quota. And since project quota is far from
a widely used feature, I don’t think there is much compatibility problems
for existing applications.  And If the new project interface are general
enough, there won’t be any compatibility problems for new applications
at all.

I might missed something important. So please let me know if you have
any advices.

Regards,
                         - Li Xi--
To unsubscribe from this list: send the line "unsubscribe linux-api" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

On Thu, Sep 25, 2014 at 05:59:12PM +1000, Dave Chinner wrote:

quoted

Also I'm afraid we may quickly run out of
32 available flags in xflags so we'd need to extend that. But all this
seems to be doable.

The struct fsxattr was designed to be extensible - it has unused
padding and enough space in the flags field to allow us to
conditionally use that padding....

I agree that it would be useful for ext4 to support as much of the
XFS_IOC_GETXATTR/XFS_IOC_SETATTR as would make sense for ext4, and to
use that to set/get the project ID.  (And that we should probably do
that as a separate set of patches that we could potentially go into
ext4 ahead of the project quota while it is undergoing testing and
review.)

A few questions of Dave and other XFS folks:

1) If we only implement a partial set of the flags or other
functionality, are there going to be tools that get confused?  i.e.,
are there any userspace programs that will test for whether the ioctl
is supported, and then assume that some minimal set of functionality
must be implemented?

2) Unless I'm missing something, there is nothing that enforces that
fsx_pad must be zero.  I assume that means that the only way you can
expand use of fields into that space is via a flag bit being consumed?

Cheers,

						- Ted

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

On Thu 25-09-14 17:59:12, Dave Chinner wrote:

On Wed, Sep 24, 2014 at 07:01:05PM +0200, Jan Kara wrote:

quoted

On Wed 24-09-14 09:26:34, Christoph Hellwig wrote:

quoted

On Wed, Sep 24, 2014 at 06:25:07PM +0200, Jan Kara wrote:

quoted

On Wed 24-09-14 22:04:30, Li Xi wrote:

quoted

This patch adds ioctl interface for setting/getting project of ext4.

  The patch looks good to me. I was just wondering whether it won't be
useful to add an ioctl() which isn't ext4 specific. We could just extend
->setattr() to allow setting of project ID (most filesystems would just
return -EOPNOTSUPP but ext4 and xfs could do the right thing) and then call
->setattr from the generic ioctl. That way userspace won't have to care
about filesystem type when setting project ID... What do others think?

Absolutely.  In general I also wonder why this patch doesn't implement
the full XFS API.  Maybe there is a reason it was considered and
rejected, but it would be helpful to document why.

  Do you mean full get/setfsxattr API?

That's a good start.

The bigger issue in my mind is that we already have a fully featured
quota API that supports project quotas and userspace tools available
that manipulate it. xfstests already uses those tools and API
for testing project quotas.

  Well, the VFS quota API is trivially extended by adding additional quota
type so I don't really see about which reinventing of quota API are you
speaking here...

This whole patchset reinvents all the quota APIs, and will require
adding support in userspace, and hence require re-inventing all the
test infrastructure we already have because it won't be compatible
with the existing project quota test code.

  Well, quota-tools will have to extended to know about the new quota type.
Yes. But that's easy to do. I think teaching xfs quota tools to work with
ext4 will be a bigger project plus I don't think I want to force sysadmins
which are used to work with quota-tools to switch to other utilities just
because of project quotas.

Regarding xfstests - I've checked and most of the project quota tests in
xfs directory aren't directly usable for ext4 anyway because of other
functionality ext4 doesn't support. So we'll need to distill the least
common denominator from them anyway...

quoted

  That basically contains project ID,
flags (those that are currently get/set with FS_IOC_GETFLAGS/SETFLAGS), and
extent size hint right?

It's a different set of flag definitions. We translate the interface
XFS_XFLAG_* values to/from the inode on-disk XFS_DIFLAG_* inode values, just
like we translate the VFS FS_*_FL flags that get passed through the
FS_IOC_GETFLAGS/SETFLAGS ioctl.

quoted

That seems workable and it would also make setting
of PROJINHERIT flag fs agnostic. Only we would have to create some generic
flags namespace and merge into that ext4 flags and have a translation
function for the old ext4 flags.

The XFS_XFLAGS_* are already filesystem agnostic - they are flags
that are only used for interfacing with userspace and hence only
exist at the ioctl copy in/out layer.....

quoted

Also I'm afraid we may quickly run out of
32 available flags in xflags so we'd need to extend that. But all this
seems to be doable.

The struct fsxattr was designed to be extensible - it has unused
padding and enough space in the flags field to allow us to
conditionally use that padding....

  Yeah, this should all be easy.

								Honza
-- 
Jan Kara [off-list ref]
SUSE Labs, CR

On Thu, Sep 25, 2014 at 03:52:13PM +0200, Jan Kara wrote:

On Thu 25-09-14 17:59:12, Dave Chinner wrote:

quoted

On Wed, Sep 24, 2014 at 07:01:05PM +0200, Jan Kara wrote:

quoted

On Wed 24-09-14 09:26:34, Christoph Hellwig wrote:

quoted

On Wed, Sep 24, 2014 at 06:25:07PM +0200, Jan Kara wrote:

quoted

On Wed 24-09-14 22:04:30, Li Xi wrote:

quoted

This patch adds ioctl interface for setting/getting project of ext4.

  The patch looks good to me. I was just wondering whether it won't be
useful to add an ioctl() which isn't ext4 specific. We could just extend
->setattr() to allow setting of project ID (most filesystems would just
return -EOPNOTSUPP but ext4 and xfs could do the right thing) and then call
->setattr from the generic ioctl. That way userspace won't have to care
about filesystem type when setting project ID... What do others think?

Absolutely.  In general I also wonder why this patch doesn't implement
the full XFS API.  Maybe there is a reason it was considered and
rejected, but it would be helpful to document why.

  Do you mean full get/setfsxattr API?

That's a good start.

The bigger issue in my mind is that we already have a fully featured
quota API that supports project quotas and userspace tools available
that manipulate it. xfstests already uses those tools and API
for testing project quotas.

  Well, the VFS quota API is trivially extended by adding additional quota
type so I don't really see about which reinventing of quota API are you
speaking here...

It doesn't seem quite as trivial as you make out given  all thei
issues so far around increasing MAXQUOTA, the increase in size of
the inode, etc.

quoted

This whole patchset reinvents all the quota APIs, and will require
adding support in userspace, and hence require re-inventing all the
test infrastructure we already have because it won't be compatible
with the existing project quota test code.

  Well, quota-tools will have to extended to know about the new quota type.
Yes. But that's easy to do. I think teaching xfs quota tools to work with
ext4 will be a bigger project plus I don't think I want to force sysadmins
which are used to work with quota-tools to switch to other utilities just
because of project quotas.

Regarding xfstests - I've checked and most of the project quota tests in
xfs directory aren't directly usable for ext4 anyway because of other
functionality ext4 doesn't support. So we'll need to distill the least
common denominator from them anyway...

I just did a quick scan - of the ~13 tests in tests/xfs that
exercise project quotas, only 2 of them test things that are xfs
specific (e.g. use xfs_db to peer at things, or use xfs_admin, etc).
The rest all rely on xfs_quota to manage and configure project
quotas but otherwise don't do anything XFS specific.

We want project quotas to have the same management interface for
administrators regardless of the filesystem they are using. The only
way we can do that is to ensure that the same tools work on either
filesystem, and right now it seems to me that the ext4 NIH syndrome
is winning out over what is best for our users...

Look, I have no problems with extending the existing quota
interfaces to support project quotas, but that should be a
*secondary* improvement as userspace tools are updated. The
primary goal needs to be "works identically to XFS" and so it needs
to implement the interfaces that are currently used for management
so that we can actually test that it does work identically.

Cheers,

Dave.
-- 
Dave Chinner
david@fromorbit.com

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

On Fri, Sep 26, 2014 at 08:42:25AM +1000, Dave Chinner wrote:

Look, I have no problems with extending the existing quota
interfaces to support project quotas, but that should be a
*secondary* improvement as userspace tools are updated. The
primary goal needs to be "works identically to XFS" and so it needs
to implement the interfaces that are currently used for management
so that we can actually test that it does work identically.

I think we're getting a bit too hung up on which is the "primary" and
which is the "secondary" interfaces.  The reality is that we should
make both interfaces work.  An example of this is how we handle the
xfs-specific ioctls that are also exposed via the fallocate(2) system
call.

It's not particularly important to me which is the "primary" interface
just because it's been around for 10 years.  Which should implement
*both* so that users are used to using xfs_io(8) or fallocate(1) can
do what they want.

Similarly, if some users are more used to the quotatools interface
(which has been around for quite a long time, BTW, if we're trying to
count primacy by years of availability to Linux users), and some users
that are used to using xfs_quota, both should work.

Cheers,

						- Ted

On Fri 26-09-14 08:42:25, Dave Chinner wrote:

On Thu, Sep 25, 2014 at 03:52:13PM +0200, Jan Kara wrote:

quoted

On Thu 25-09-14 17:59:12, Dave Chinner wrote:

quoted

On Wed, Sep 24, 2014 at 07:01:05PM +0200, Jan Kara wrote:

quoted

On Wed 24-09-14 09:26:34, Christoph Hellwig wrote:

quoted

On Wed, Sep 24, 2014 at 06:25:07PM +0200, Jan Kara wrote:

quoted

On Wed 24-09-14 22:04:30, Li Xi wrote:

quoted

This patch adds ioctl interface for setting/getting project of ext4.

  The patch looks good to me. I was just wondering whether it won't be
useful to add an ioctl() which isn't ext4 specific. We could just extend
->setattr() to allow setting of project ID (most filesystems would just
return -EOPNOTSUPP but ext4 and xfs could do the right thing) and then call
->setattr from the generic ioctl. That way userspace won't have to care
about filesystem type when setting project ID... What do others think?

Absolutely.  In general I also wonder why this patch doesn't implement
the full XFS API.  Maybe there is a reason it was considered and
rejected, but it would be helpful to document why.

  Do you mean full get/setfsxattr API?

That's a good start.

The bigger issue in my mind is that we already have a fully featured
quota API that supports project quotas and userspace tools available
that manipulate it. xfstests already uses those tools and API
for testing project quotas.

  Well, the VFS quota API is trivially extended by adding additional quota
type so I don't really see about which reinventing of quota API are you
speaking here...

It doesn't seem quite as trivial as you make out given  all thei
issues so far around increasing MAXQUOTA, the increase in size of
the inode, etc.