Merge branch 'next-evm' of git://git.kernel.org/pub/scm/linux/kernel/git/zohar/ima...
[linux-3.10.git] / fs / xfs / linux-2.6 / xfs_iops.c
index 27a3658..3719460 100644 (file)
@@ -39,6 +39,7 @@
 #include "xfs_buf_item.h"
 #include "xfs_utils.h"
 #include "xfs_vnodeops.h"
+#include "xfs_inode_item.h"
 #include "xfs_trace.h"
 
 #include <linux/capability.h>
@@ -202,9 +203,9 @@ xfs_vn_mknod(
 
        if (default_acl) {
                error = -xfs_inherit_acl(inode, default_acl);
+               default_acl = NULL;
                if (unlikely(error))
                        goto out_cleanup_inode;
-               posix_acl_release(default_acl);
        }
 
 
@@ -498,12 +499,442 @@ xfs_vn_getattr(
        return 0;
 }
 
+int
+xfs_setattr_nonsize(
+       struct xfs_inode        *ip,
+       struct iattr            *iattr,
+       int                     flags)
+{
+       xfs_mount_t             *mp = ip->i_mount;
+       struct inode            *inode = VFS_I(ip);
+       int                     mask = iattr->ia_valid;
+       xfs_trans_t             *tp;
+       int                     error;
+       uid_t                   uid = 0, iuid = 0;
+       gid_t                   gid = 0, igid = 0;
+       struct xfs_dquot        *udqp = NULL, *gdqp = NULL;
+       struct xfs_dquot        *olddquot1 = NULL, *olddquot2 = NULL;
+
+       trace_xfs_setattr(ip);
+
+       if (mp->m_flags & XFS_MOUNT_RDONLY)
+               return XFS_ERROR(EROFS);
+
+       if (XFS_FORCED_SHUTDOWN(mp))
+               return XFS_ERROR(EIO);
+
+       error = -inode_change_ok(inode, iattr);
+       if (error)
+               return XFS_ERROR(error);
+
+       ASSERT((mask & ATTR_SIZE) == 0);
+
+       /*
+        * If disk quotas is on, we make sure that the dquots do exist on disk,
+        * before we start any other transactions. Trying to do this later
+        * is messy. We don't care to take a readlock to look at the ids
+        * in inode here, because we can't hold it across the trans_reserve.
+        * If the IDs do change before we take the ilock, we're covered
+        * because the i_*dquot fields will get updated anyway.
+        */
+       if (XFS_IS_QUOTA_ON(mp) && (mask & (ATTR_UID|ATTR_GID))) {
+               uint    qflags = 0;
+
+               if ((mask & ATTR_UID) && XFS_IS_UQUOTA_ON(mp)) {
+                       uid = iattr->ia_uid;
+                       qflags |= XFS_QMOPT_UQUOTA;
+               } else {
+                       uid = ip->i_d.di_uid;
+               }
+               if ((mask & ATTR_GID) && XFS_IS_GQUOTA_ON(mp)) {
+                       gid = iattr->ia_gid;
+                       qflags |= XFS_QMOPT_GQUOTA;
+               }  else {
+                       gid = ip->i_d.di_gid;
+               }
+
+               /*
+                * We take a reference when we initialize udqp and gdqp,
+                * so it is important that we never blindly double trip on
+                * the same variable. See xfs_create() for an example.
+                */
+               ASSERT(udqp == NULL);
+               ASSERT(gdqp == NULL);
+               error = xfs_qm_vop_dqalloc(ip, uid, gid, xfs_get_projid(ip),
+                                        qflags, &udqp, &gdqp);
+               if (error)
+                       return error;
+       }
+
+       tp = xfs_trans_alloc(mp, XFS_TRANS_SETATTR_NOT_SIZE);
+       error = xfs_trans_reserve(tp, 0, XFS_ICHANGE_LOG_RES(mp), 0, 0, 0);
+       if (error)
+               goto out_dqrele;
+
+       xfs_ilock(ip, XFS_ILOCK_EXCL);
+
+       /*
+        * Change file ownership.  Must be the owner or privileged.
+        */
+       if (mask & (ATTR_UID|ATTR_GID)) {
+               /*
+                * These IDs could have changed since we last looked at them.
+                * But, we're assured that if the ownership did change
+                * while we didn't have the inode locked, inode's dquot(s)
+                * would have changed also.
+                */
+               iuid = ip->i_d.di_uid;
+               igid = ip->i_d.di_gid;
+               gid = (mask & ATTR_GID) ? iattr->ia_gid : igid;
+               uid = (mask & ATTR_UID) ? iattr->ia_uid : iuid;
+
+               /*
+                * Do a quota reservation only if uid/gid is actually
+                * going to change.
+                */
+               if (XFS_IS_QUOTA_RUNNING(mp) &&
+                   ((XFS_IS_UQUOTA_ON(mp) && iuid != uid) ||
+                    (XFS_IS_GQUOTA_ON(mp) && igid != gid))) {
+                       ASSERT(tp);
+                       error = xfs_qm_vop_chown_reserve(tp, ip, udqp, gdqp,
+                                               capable(CAP_FOWNER) ?
+                                               XFS_QMOPT_FORCE_RES : 0);
+                       if (error)      /* out of quota */
+                               goto out_trans_cancel;
+               }
+       }
+
+       xfs_trans_ijoin(tp, ip);
+
+       /*
+        * Change file ownership.  Must be the owner or privileged.
+        */
+       if (mask & (ATTR_UID|ATTR_GID)) {
+               /*
+                * CAP_FSETID overrides the following restrictions:
+                *
+                * The set-user-ID and set-group-ID bits of a file will be
+                * cleared upon successful return from chown()
+                */
+               if ((ip->i_d.di_mode & (S_ISUID|S_ISGID)) &&
+                   !capable(CAP_FSETID))
+                       ip->i_d.di_mode &= ~(S_ISUID|S_ISGID);
+
+               /*
+                * Change the ownerships and register quota modifications
+                * in the transaction.
+                */
+               if (iuid != uid) {
+                       if (XFS_IS_QUOTA_RUNNING(mp) && XFS_IS_UQUOTA_ON(mp)) {
+                               ASSERT(mask & ATTR_UID);
+                               ASSERT(udqp);
+                               olddquot1 = xfs_qm_vop_chown(tp, ip,
+                                                       &ip->i_udquot, udqp);
+                       }
+                       ip->i_d.di_uid = uid;
+                       inode->i_uid = uid;
+               }
+               if (igid != gid) {
+                       if (XFS_IS_QUOTA_RUNNING(mp) && XFS_IS_GQUOTA_ON(mp)) {
+                               ASSERT(!XFS_IS_PQUOTA_ON(mp));
+                               ASSERT(mask & ATTR_GID);
+                               ASSERT(gdqp);
+                               olddquot2 = xfs_qm_vop_chown(tp, ip,
+                                                       &ip->i_gdquot, gdqp);
+                       }
+                       ip->i_d.di_gid = gid;
+                       inode->i_gid = gid;
+               }
+       }
+
+       /*
+        * Change file access modes.
+        */
+       if (mask & ATTR_MODE) {
+               umode_t mode = iattr->ia_mode;
+
+               if (!in_group_p(inode->i_gid) && !capable(CAP_FSETID))
+                       mode &= ~S_ISGID;
+
+               ip->i_d.di_mode &= S_IFMT;
+               ip->i_d.di_mode |= mode & ~S_IFMT;
+
+               inode->i_mode &= S_IFMT;
+               inode->i_mode |= mode & ~S_IFMT;
+       }
+
+       /*
+        * Change file access or modified times.
+        */
+       if (mask & ATTR_ATIME) {
+               inode->i_atime = iattr->ia_atime;
+               ip->i_d.di_atime.t_sec = iattr->ia_atime.tv_sec;
+               ip->i_d.di_atime.t_nsec = iattr->ia_atime.tv_nsec;
+               ip->i_update_core = 1;
+       }
+       if (mask & ATTR_CTIME) {
+               inode->i_ctime = iattr->ia_ctime;
+               ip->i_d.di_ctime.t_sec = iattr->ia_ctime.tv_sec;
+               ip->i_d.di_ctime.t_nsec = iattr->ia_ctime.tv_nsec;
+               ip->i_update_core = 1;
+       }
+       if (mask & ATTR_MTIME) {
+               inode->i_mtime = iattr->ia_mtime;
+               ip->i_d.di_mtime.t_sec = iattr->ia_mtime.tv_sec;
+               ip->i_d.di_mtime.t_nsec = iattr->ia_mtime.tv_nsec;
+               ip->i_update_core = 1;
+       }
+
+       xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
+
+       XFS_STATS_INC(xs_ig_attrchg);
+
+       if (mp->m_flags & XFS_MOUNT_WSYNC)
+               xfs_trans_set_sync(tp);
+       error = xfs_trans_commit(tp, 0);
+
+       xfs_iunlock(ip, XFS_ILOCK_EXCL);
+
+       /*
+        * Release any dquot(s) the inode had kept before chown.
+        */
+       xfs_qm_dqrele(olddquot1);
+       xfs_qm_dqrele(olddquot2);
+       xfs_qm_dqrele(udqp);
+       xfs_qm_dqrele(gdqp);
+
+       if (error)
+               return XFS_ERROR(error);
+
+       /*
+        * XXX(hch): Updating the ACL entries is not atomic vs the i_mode
+        *           update.  We could avoid this with linked transactions
+        *           and passing down the transaction pointer all the way
+        *           to attr_set.  No previous user of the generic
+        *           Posix ACL code seems to care about this issue either.
+        */
+       if ((mask & ATTR_MODE) && !(flags & XFS_ATTR_NOACL)) {
+               error = -xfs_acl_chmod(inode);
+               if (error)
+                       return XFS_ERROR(error);
+       }
+
+       return 0;
+
+out_trans_cancel:
+       xfs_trans_cancel(tp, 0);
+       xfs_iunlock(ip, XFS_ILOCK_EXCL);
+out_dqrele:
+       xfs_qm_dqrele(udqp);
+       xfs_qm_dqrele(gdqp);
+       return error;
+}
+
+/*
+ * Truncate file.  Must have write permission and not be a directory.
+ */
+int
+xfs_setattr_size(
+       struct xfs_inode        *ip,
+       struct iattr            *iattr,
+       int                     flags)
+{
+       struct xfs_mount        *mp = ip->i_mount;
+       struct inode            *inode = VFS_I(ip);
+       int                     mask = iattr->ia_valid;
+       struct xfs_trans        *tp;
+       int                     error;
+       uint                    lock_flags;
+       uint                    commit_flags = 0;
+
+       trace_xfs_setattr(ip);
+
+       if (mp->m_flags & XFS_MOUNT_RDONLY)
+               return XFS_ERROR(EROFS);
+
+       if (XFS_FORCED_SHUTDOWN(mp))
+               return XFS_ERROR(EIO);
+
+       error = -inode_change_ok(inode, iattr);
+       if (error)
+               return XFS_ERROR(error);
+
+       ASSERT(S_ISREG(ip->i_d.di_mode));
+       ASSERT((mask & (ATTR_MODE|ATTR_UID|ATTR_GID|ATTR_ATIME|ATTR_ATIME_SET|
+                       ATTR_MTIME_SET|ATTR_KILL_SUID|ATTR_KILL_SGID|
+                       ATTR_KILL_PRIV|ATTR_TIMES_SET)) == 0);
+
+       lock_flags = XFS_ILOCK_EXCL;
+       if (!(flags & XFS_ATTR_NOLOCK))
+               lock_flags |= XFS_IOLOCK_EXCL;
+       xfs_ilock(ip, lock_flags);
+
+       /*
+        * Short circuit the truncate case for zero length files.
+        */
+       if (iattr->ia_size == 0 &&
+           ip->i_size == 0 && ip->i_d.di_nextents == 0) {
+               if (!(mask & (ATTR_CTIME|ATTR_MTIME)))
+                       goto out_unlock;
+
+               /*
+                * Use the regular setattr path to update the timestamps.
+                */
+               xfs_iunlock(ip, lock_flags);
+               iattr->ia_valid &= ~ATTR_SIZE;
+               return xfs_setattr_nonsize(ip, iattr, 0);
+       }
+
+       /*
+        * Make sure that the dquots are attached to the inode.
+        */
+       error = xfs_qm_dqattach_locked(ip, 0);
+       if (error)
+               goto out_unlock;
+
+       /*
+        * Now we can make the changes.  Before we join the inode to the
+        * transaction, take care of the part of the truncation that must be
+        * done without the inode lock.  This needs to be done before joining
+        * the inode to the transaction, because the inode cannot be unlocked
+        * once it is a part of the transaction.
+        */
+       if (iattr->ia_size > ip->i_size) {
+               /*
+                * Do the first part of growing a file: zero any data in the
+                * last block that is beyond the old EOF.  We need to do this
+                * before the inode is joined to the transaction to modify
+                * i_size.
+                */
+               error = xfs_zero_eof(ip, iattr->ia_size, ip->i_size);
+               if (error)
+                       goto out_unlock;
+       }
+       xfs_iunlock(ip, XFS_ILOCK_EXCL);
+       lock_flags &= ~XFS_ILOCK_EXCL;
+
+       /*
+        * We are going to log the inode size change in this transaction so
+        * any previous writes that are beyond the on disk EOF and the new
+        * EOF that have not been written out need to be written here.  If we
+        * do not write the data out, we expose ourselves to the null files
+        * problem.
+        *
+        * Only flush from the on disk size to the smaller of the in memory
+        * file size or the new size as that's the range we really care about
+        * here and prevents waiting for other data not within the range we
+        * care about here.
+        */
+       if (ip->i_size != ip->i_d.di_size && iattr->ia_size > ip->i_d.di_size) {
+               error = xfs_flush_pages(ip, ip->i_d.di_size, iattr->ia_size,
+                                       XBF_ASYNC, FI_NONE);
+               if (error)
+                       goto out_unlock;
+       }
+
+       /*
+        * Wait for all I/O to complete.
+        */
+       xfs_ioend_wait(ip);
+
+       error = -block_truncate_page(inode->i_mapping, iattr->ia_size,
+                                    xfs_get_blocks);
+       if (error)
+               goto out_unlock;
+
+       tp = xfs_trans_alloc(mp, XFS_TRANS_SETATTR_SIZE);
+       error = xfs_trans_reserve(tp, 0, XFS_ITRUNCATE_LOG_RES(mp), 0,
+                                XFS_TRANS_PERM_LOG_RES,
+                                XFS_ITRUNCATE_LOG_COUNT);
+       if (error)
+               goto out_trans_cancel;
+
+       truncate_setsize(inode, iattr->ia_size);
+
+       commit_flags = XFS_TRANS_RELEASE_LOG_RES;
+       lock_flags |= XFS_ILOCK_EXCL;
+
+       xfs_ilock(ip, XFS_ILOCK_EXCL);
+
+       xfs_trans_ijoin(tp, ip);
+
+       /*
+        * Only change the c/mtime if we are changing the size or we are
+        * explicitly asked to change it.  This handles the semantic difference
+        * between truncate() and ftruncate() as implemented in the VFS.
+        *
+        * The regular truncate() case without ATTR_CTIME and ATTR_MTIME is a
+        * special case where we need to update the times despite not having
+        * these flags set.  For all other operations the VFS set these flags
+        * explicitly if it wants a timestamp update.
+        */
+       if (iattr->ia_size != ip->i_size &&
+           (!(mask & (ATTR_CTIME | ATTR_MTIME)))) {
+               iattr->ia_ctime = iattr->ia_mtime =
+                       current_fs_time(inode->i_sb);
+               mask |= ATTR_CTIME | ATTR_MTIME;
+       }
+
+       if (iattr->ia_size > ip->i_size) {
+               ip->i_d.di_size = iattr->ia_size;
+               ip->i_size = iattr->ia_size;
+       } else if (iattr->ia_size <= ip->i_size ||
+                  (iattr->ia_size == 0 && ip->i_d.di_nextents)) {
+               error = xfs_itruncate_data(&tp, ip, iattr->ia_size);
+               if (error)
+                       goto out_trans_abort;
+
+               /*
+                * Truncated "down", so we're removing references to old data
+                * here - if we delay flushing for a long time, we expose
+                * ourselves unduly to the notorious NULL files problem.  So,
+                * we mark this inode and flush it when the file is closed,
+                * and do not wait the usual (long) time for writeout.
+                */
+               xfs_iflags_set(ip, XFS_ITRUNCATED);
+       }
+
+       if (mask & ATTR_CTIME) {
+               inode->i_ctime = iattr->ia_ctime;
+               ip->i_d.di_ctime.t_sec = iattr->ia_ctime.tv_sec;
+               ip->i_d.di_ctime.t_nsec = iattr->ia_ctime.tv_nsec;
+               ip->i_update_core = 1;
+       }
+       if (mask & ATTR_MTIME) {
+               inode->i_mtime = iattr->ia_mtime;
+               ip->i_d.di_mtime.t_sec = iattr->ia_mtime.tv_sec;
+               ip->i_d.di_mtime.t_nsec = iattr->ia_mtime.tv_nsec;
+               ip->i_update_core = 1;
+       }
+
+       xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
+
+       XFS_STATS_INC(xs_ig_attrchg);
+
+       if (mp->m_flags & XFS_MOUNT_WSYNC)
+               xfs_trans_set_sync(tp);
+
+       error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES);
+out_unlock:
+       if (lock_flags)
+               xfs_iunlock(ip, lock_flags);
+       return error;
+
+out_trans_abort:
+       commit_flags |= XFS_TRANS_ABORT;
+out_trans_cancel:
+       xfs_trans_cancel(tp, commit_flags);
+       goto out_unlock;
+}
+
 STATIC int
 xfs_vn_setattr(
        struct dentry   *dentry,
        struct iattr    *iattr)
 {
-       return -xfs_setattr(XFS_I(dentry->d_inode), iattr, 0);
+       if (iattr->ia_valid & ATTR_SIZE)
+               return -xfs_setattr_size(XFS_I(dentry->d_inode), iattr, 0);
+       return -xfs_setattr_nonsize(XFS_I(dentry->d_inode), iattr, 0);
 }
 
 #define XFS_FIEMAP_FLAGS       (FIEMAP_FLAG_SYNC|FIEMAP_FLAG_XATTR)
@@ -592,7 +1023,7 @@ xfs_vn_fiemap(
 }
 
 static const struct inode_operations xfs_inode_operations = {
-       .check_acl              = xfs_check_acl,
+       .get_acl                = xfs_get_acl,
        .getattr                = xfs_vn_getattr,
        .setattr                = xfs_vn_setattr,
        .setxattr               = generic_setxattr,
@@ -618,7 +1049,7 @@ static const struct inode_operations xfs_dir_inode_operations = {
        .rmdir                  = xfs_vn_unlink,
        .mknod                  = xfs_vn_mknod,
        .rename                 = xfs_vn_rename,
-       .check_acl              = xfs_check_acl,
+       .get_acl                = xfs_get_acl,
        .getattr                = xfs_vn_getattr,
        .setattr                = xfs_vn_setattr,
        .setxattr               = generic_setxattr,
@@ -643,7 +1074,7 @@ static const struct inode_operations xfs_dir_ci_inode_operations = {
        .rmdir                  = xfs_vn_unlink,
        .mknod                  = xfs_vn_mknod,
        .rename                 = xfs_vn_rename,
-       .check_acl              = xfs_check_acl,
+       .get_acl                = xfs_get_acl,
        .getattr                = xfs_vn_getattr,
        .setattr                = xfs_vn_setattr,
        .setxattr               = generic_setxattr,
@@ -656,7 +1087,7 @@ static const struct inode_operations xfs_symlink_inode_operations = {
        .readlink               = generic_readlink,
        .follow_link            = xfs_vn_follow_link,
        .put_link               = xfs_vn_put_link,
-       .check_acl              = xfs_check_acl,
+       .get_acl                = xfs_get_acl,
        .getattr                = xfs_vn_getattr,
        .setattr                = xfs_vn_setattr,
        .setxattr               = generic_setxattr,
@@ -764,6 +1195,15 @@ xfs_setup_inode(
                break;
        }
 
+       /*
+        * If there is no attribute fork no ACL can exist on this inode,
+        * and it can't have any file capabilities attached to it either.
+        */
+       if (!XFS_IFORK_Q(ip)) {
+               inode_has_no_xattr(inode);
+               cache_no_acl(inode);
+       }
+
        xfs_iflags_clear(ip, XFS_INEW);
        barrier();