OSDN Git Service

xfs: remove the kuid/kgid conversion wrappers
authorChristoph Hellwig <hch@lst.de>
Fri, 21 Feb 2020 16:31:27 +0000 (08:31 -0800)
committerDarrick J. Wong <darrick.wong@oracle.com>
Tue, 3 Mar 2020 04:55:50 +0000 (20:55 -0800)
Remove the XFS wrappers for converting from and to the kuid/kgid types.
Mostly this means switching to VFS i_{u,g}id_{read,write} helpers, but
in a few spots the calls to the conversion functions is open coded.
To match the use of sb->s_user_ns in the helpers and other file systems,
sb->s_user_ns is also used in the quota code.  The ACL code already does
the conversion in a grotty layering violation in the VFS xattr code,
so it keeps using init_user_ns for the identity mapping.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
fs/xfs/libxfs/xfs_inode_buf.c
fs/xfs/xfs_acl.c
fs/xfs/xfs_dquot.c
fs/xfs/xfs_inode_item.c
fs/xfs/xfs_itable.c
fs/xfs/xfs_linux.h
fs/xfs/xfs_qm.c

index bc72b57..17e88a8 100644 (file)
@@ -222,8 +222,8 @@ xfs_inode_from_disk(
        }
 
        to->di_format = from->di_format;
-       inode->i_uid = xfs_uid_to_kuid(be32_to_cpu(from->di_uid));
-       inode->i_gid = xfs_gid_to_kgid(be32_to_cpu(from->di_gid));
+       i_uid_write(inode, be32_to_cpu(from->di_uid));
+       i_gid_write(inode, be32_to_cpu(from->di_gid));
        to->di_flushiter = be16_to_cpu(from->di_flushiter);
 
        /*
@@ -276,8 +276,8 @@ xfs_inode_to_disk(
 
        to->di_version = from->di_version;
        to->di_format = from->di_format;
-       to->di_uid = cpu_to_be32(xfs_kuid_to_uid(inode->i_uid));
-       to->di_gid = cpu_to_be32(xfs_kgid_to_gid(inode->i_gid));
+       to->di_uid = cpu_to_be32(i_uid_read(inode));
+       to->di_gid = cpu_to_be32(i_gid_read(inode));
        to->di_projid_lo = cpu_to_be16(from->di_projid & 0xffff);
        to->di_projid_hi = cpu_to_be16(from->di_projid >> 16);
 
index cd743fa..e7314b5 100644 (file)
@@ -67,10 +67,12 @@ xfs_acl_from_disk(
 
                switch (acl_e->e_tag) {
                case ACL_USER:
-                       acl_e->e_uid = xfs_uid_to_kuid(be32_to_cpu(ace->ae_id));
+                       acl_e->e_uid = make_kuid(&init_user_ns,
+                                                be32_to_cpu(ace->ae_id));
                        break;
                case ACL_GROUP:
-                       acl_e->e_gid = xfs_gid_to_kgid(be32_to_cpu(ace->ae_id));
+                       acl_e->e_gid = make_kgid(&init_user_ns,
+                                                be32_to_cpu(ace->ae_id));
                        break;
                case ACL_USER_OBJ:
                case ACL_GROUP_OBJ:
@@ -103,10 +105,12 @@ xfs_acl_to_disk(struct xfs_acl *aclp, const struct posix_acl *acl)
                ace->ae_tag = cpu_to_be32(acl_e->e_tag);
                switch (acl_e->e_tag) {
                case ACL_USER:
-                       ace->ae_id = cpu_to_be32(xfs_kuid_to_uid(acl_e->e_uid));
+                       ace->ae_id = cpu_to_be32(
+                                       from_kuid(&init_user_ns, acl_e->e_uid));
                        break;
                case ACL_GROUP:
-                       ace->ae_id = cpu_to_be32(xfs_kgid_to_gid(acl_e->e_gid));
+                       ace->ae_id = cpu_to_be32(
+                                       from_kgid(&init_user_ns, acl_e->e_gid));
                        break;
                default:
                        ace->ae_id = cpu_to_be32(ACL_UNDEFINED_ID);
index 3579de9..711376c 100644 (file)
@@ -829,9 +829,9 @@ xfs_qm_id_for_quotatype(
 {
        switch (type) {
        case XFS_DQ_USER:
-               return xfs_kuid_to_uid(VFS_I(ip)->i_uid);
+               return i_uid_read(VFS_I(ip));
        case XFS_DQ_GROUP:
-               return xfs_kgid_to_gid(VFS_I(ip)->i_gid);
+               return i_gid_read(VFS_I(ip));
        case XFS_DQ_PROJ:
                return ip->i_d.di_projid;
        }
index 83d7914..f021b55 100644 (file)
@@ -308,8 +308,8 @@ xfs_inode_to_log_dinode(
 
        to->di_version = from->di_version;
        to->di_format = from->di_format;
-       to->di_uid = xfs_kuid_to_uid(inode->i_uid);
-       to->di_gid = xfs_kgid_to_gid(inode->i_gid);
+       to->di_uid = i_uid_read(inode);
+       to->di_gid = i_gid_read(inode);
        to->di_projid_lo = from->di_projid & 0xffff;
        to->di_projid_hi = from->di_projid >> 16;
 
index 497db41..d106604 100644 (file)
@@ -86,8 +86,8 @@ xfs_bulkstat_one_int(
         */
        buf->bs_projectid = ip->i_d.di_projid;
        buf->bs_ino = ino;
-       buf->bs_uid = xfs_kuid_to_uid(inode->i_uid);
-       buf->bs_gid = xfs_kgid_to_gid(inode->i_gid);
+       buf->bs_uid = i_uid_read(inode);
+       buf->bs_gid = i_gid_read(inode);
        buf->bs_size = dic->di_size;
 
        buf->bs_nlink = inode->i_nlink;
index 8738bb0..bc43cd9 100644 (file)
@@ -163,32 +163,6 @@ struct xstats {
 
 extern struct xstats xfsstats;
 
-/* Kernel uid/gid conversion. These are used to convert to/from the on disk
- * uid_t/gid_t types to the kuid_t/kgid_t types that the kernel uses internally.
- * The conversion here is type only, the value will remain the same since we
- * are converting to the init_user_ns. The uid is later mapped to a particular
- * user namespace value when crossing the kernel/user boundary.
- */
-static inline uint32_t xfs_kuid_to_uid(kuid_t uid)
-{
-       return from_kuid(&init_user_ns, uid);
-}
-
-static inline kuid_t xfs_uid_to_kuid(uint32_t uid)
-{
-       return make_kuid(&init_user_ns, uid);
-}
-
-static inline uint32_t xfs_kgid_to_gid(kgid_t gid)
-{
-       return from_kgid(&init_user_ns, gid);
-}
-
-static inline kgid_t xfs_gid_to_kgid(uint32_t gid)
-{
-       return make_kgid(&init_user_ns, gid);
-}
-
 static inline dev_t xfs_to_linux_dev_t(xfs_dev_t dev)
 {
        return MKDEV(sysv_major(dev) & 0x1ff, sysv_minor(dev));
index 54dda7d..de1d2c6 100644 (file)
@@ -326,8 +326,7 @@ xfs_qm_dqattach_locked(
        ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL));
 
        if (XFS_IS_UQUOTA_ON(mp) && !ip->i_udquot) {
-               error = xfs_qm_dqattach_one(ip,
-                               xfs_kuid_to_uid(VFS_I(ip)->i_uid),
+               error = xfs_qm_dqattach_one(ip, i_uid_read(VFS_I(ip)),
                                XFS_DQ_USER, doalloc, &ip->i_udquot);
                if (error)
                        goto done;
@@ -335,8 +334,7 @@ xfs_qm_dqattach_locked(
        }
 
        if (XFS_IS_GQUOTA_ON(mp) && !ip->i_gdquot) {
-               error = xfs_qm_dqattach_one(ip,
-                               xfs_kgid_to_gid(VFS_I(ip)->i_gid),
+               error = xfs_qm_dqattach_one(ip, i_gid_read(VFS_I(ip)),
                                XFS_DQ_GROUP, doalloc, &ip->i_gdquot);
                if (error)
                        goto done;
@@ -1625,6 +1623,7 @@ xfs_qm_vop_dqalloc(
 {
        struct xfs_mount        *mp = ip->i_mount;
        struct inode            *inode = VFS_I(ip);
+       struct user_namespace   *user_ns = inode->i_sb->s_user_ns;
        struct xfs_dquot        *uq = NULL;
        struct xfs_dquot        *gq = NULL;
        struct xfs_dquot        *pq = NULL;
@@ -1664,7 +1663,7 @@ xfs_qm_vop_dqalloc(
                         * holding ilock.
                         */
                        xfs_iunlock(ip, lockflags);
-                       error = xfs_qm_dqget(mp, xfs_kuid_to_uid(uid),
+                       error = xfs_qm_dqget(mp, from_kuid(user_ns, uid),
                                        XFS_DQ_USER, true, &uq);
                        if (error) {
                                ASSERT(error != -ENOENT);
@@ -1688,7 +1687,7 @@ xfs_qm_vop_dqalloc(
        if ((flags & XFS_QMOPT_GQUOTA) && XFS_IS_GQUOTA_ON(mp)) {
                if (!gid_eq(inode->i_gid, gid)) {
                        xfs_iunlock(ip, lockflags);
-                       error = xfs_qm_dqget(mp, xfs_kgid_to_gid(gid),
+                       error = xfs_qm_dqget(mp, from_kgid(user_ns, gid),
                                        XFS_DQ_GROUP, true, &gq);
                        if (error) {
                                ASSERT(error != -ENOENT);
@@ -1815,8 +1814,7 @@ xfs_qm_vop_chown_reserve(
                        XFS_QMOPT_RES_RTBLKS : XFS_QMOPT_RES_REGBLKS;
 
        if (XFS_IS_UQUOTA_ON(mp) && udqp &&
-           xfs_kuid_to_uid(VFS_I(ip)->i_uid) !=
-                       be32_to_cpu(udqp->q_core.d_id)) {
+           i_uid_read(VFS_I(ip)) != be32_to_cpu(udqp->q_core.d_id)) {
                udq_delblks = udqp;
                /*
                 * If there are delayed allocation blocks, then we have to
@@ -1829,8 +1827,7 @@ xfs_qm_vop_chown_reserve(
                }
        }
        if (XFS_IS_GQUOTA_ON(ip->i_mount) && gdqp &&
-           xfs_kgid_to_gid(VFS_I(ip)->i_gid) !=
-                       be32_to_cpu(gdqp->q_core.d_id)) {
+           i_gid_read(VFS_I(ip)) != be32_to_cpu(gdqp->q_core.d_id)) {
                gdq_delblks = gdqp;
                if (delblks) {
                        ASSERT(ip->i_gdquot);
@@ -1927,16 +1924,14 @@ xfs_qm_vop_create_dqattach(
 
        if (udqp && XFS_IS_UQUOTA_ON(mp)) {
                ASSERT(ip->i_udquot == NULL);
-               ASSERT(xfs_kuid_to_uid(VFS_I(ip)->i_uid) ==
-                       be32_to_cpu(udqp->q_core.d_id));
+               ASSERT(i_uid_read(VFS_I(ip)) == be32_to_cpu(udqp->q_core.d_id));
 
                ip->i_udquot = xfs_qm_dqhold(udqp);
                xfs_trans_mod_dquot(tp, udqp, XFS_TRANS_DQ_ICOUNT, 1);
        }
        if (gdqp && XFS_IS_GQUOTA_ON(mp)) {
                ASSERT(ip->i_gdquot == NULL);
-               ASSERT(xfs_kgid_to_gid(VFS_I(ip)->i_gid) ==
-                       be32_to_cpu(gdqp->q_core.d_id));
+               ASSERT(i_gid_read(VFS_I(ip)) == be32_to_cpu(gdqp->q_core.d_id));
 
                ip->i_gdquot = xfs_qm_dqhold(gdqp);
                xfs_trans_mod_dquot(tp, gdqp, XFS_TRANS_DQ_ICOUNT, 1);