Merge tag 'ceph-for-5.3-rc1' of git://github.com/ceph/ceph-client

Pull ceph updates from Ilya Dryomov:
 "Lots of exciting things this time!

   - support for rbd object-map and fast-diff features (myself). This
     will speed up reads, discards and things like snap diffs on sparse
     images.

   - ceph.snap.btime vxattr to expose snapshot creation time (David
     Disseldorp). This will be used to integrate with "Restore Previous
     Versions" feature added in Windows 7 for folks who reexport ceph
     through SMB.

   - security xattrs for ceph (Zheng Yan). Only selinux is supported for
     now due to the limitations of ->dentry_init_security().

   - support for MSG_ADDR2, FS_BTIME and FS_CHANGE_ATTR features (Jeff
     Layton). This is actually a single feature bit which was missing
     because of the filesystem pieces. With this in, the kernel client
     will finally be reported as "luminous" by "ceph features" -- it is
     still being reported as "jewel" even though all required Luminous
     features were implemented in 4.13.

   - stop NULL-terminating ceph vxattrs (Jeff Layton). The convention
     with xattrs is to not terminate and this was causing
     inconsistencies with ceph-fuse.

   - change filesystem time granularity from 1 us to 1 ns, again fixing
     an inconsistency with ceph-fuse (Luis Henriques).

  On top of this there are some additional dentry name handling and cap
  flushing fixes from Zheng. Finally, Jeff is formally taking over for
  Zheng as the filesystem maintainer"

* tag 'ceph-for-5.3-rc1' of git://github.com/ceph/ceph-client: (71 commits)
  ceph: fix end offset in truncate_inode_pages_range call
  ceph: use generic_delete_inode() for ->drop_inode
  ceph: use ceph_evict_inode to cleanup inode's resource
  ceph: initialize superblock s_time_gran to 1
  MAINTAINERS: take over for Zheng as CephFS kernel client maintainer
  rbd: setallochint only if object doesn't exist
  rbd: support for object-map and fast-diff
  rbd: call rbd_dev_mapping_set() from rbd_dev_image_probe()
  libceph: export osd_req_op_data() macro
  libceph: change ceph_osdc_call() to take page vector for response
  libceph: bump CEPH_MSG_MAX_DATA_LEN (again)
  rbd: new exclusive lock wait/wake code
  rbd: quiescing lock should wait for image requests
  rbd: lock should be quiesced on reacquire
  rbd: introduce copyup state machine
  rbd: rename rbd_obj_setup_*() to rbd_obj_init_*()
  rbd: move OSD request allocation into object request state machines
  rbd: factor out __rbd_osd_setup_discard_ops()
  rbd: factor out rbd_osd_setup_copyup()
  rbd: introduce obj_req->osd_reqs list
  ...

1 files changed, 116 insertions, 92 deletions

diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c
index 761451f36e2d..791f84a13bb8 100644
--- a/fs/ceph/inode.c
+++ b/fs/ceph/inode.c
@@ -13,6 +13,7 @@
 #include <linux/posix_acl.h>
 #include <linux/random.h>
 #include <linux/sort.h>
+#include <linux/iversion.h>
 
 #include "super.h"
 #include "mds_client.h"
@@ -42,6 +43,7 @@ static int ceph_set_ino_cb(struct inode *inode, void *data)
 {
 	ceph_inode(inode)->i_vino = *(struct ceph_vino *)data;
 	inode->i_ino = ceph_vino_to_ino(*(struct ceph_vino *)data);
+	inode_set_iversion_raw(inode, 0);
 	return 0;
 }
 
@@ -509,6 +511,7 @@ struct inode *ceph_alloc_inode(struct super_block *sb)
 
 	INIT_WORK(&ci->i_work, ceph_inode_work);
 	ci->i_work_mask = 0;
+	memset(&ci->i_btime, '\0', sizeof(ci->i_btime));
 
 	ceph_fscache_inode_init(ci);
 
@@ -523,17 +526,20 @@ void ceph_free_inode(struct inode *inode)
 	kmem_cache_free(ceph_inode_cachep, ci);
 }
 
-void ceph_destroy_inode(struct inode *inode)
+void ceph_evict_inode(struct inode *inode)
 {
 	struct ceph_inode_info *ci = ceph_inode(inode);
 	struct ceph_inode_frag *frag;
 	struct rb_node *n;
 
-	dout("destroy_inode %p ino %llx.%llx\n", inode, ceph_vinop(inode));
+	dout("evict_inode %p ino %llx.%llx\n", inode, ceph_vinop(inode));
+
+	truncate_inode_pages_final(&inode->i_data);
+	clear_inode(inode);
 
 	ceph_fscache_unregister_inode_cookie(ci);
 
-	__ceph_remove_caps(inode);
+	__ceph_remove_caps(ci);
 
 	if (__ceph_has_any_quota(ci))
 		ceph_adjust_quota_realms_count(inode, false);
@@ -578,16 +584,6 @@ void ceph_destroy_inode(struct inode *inode)
 	ceph_put_string(rcu_dereference_raw(ci->i_layout.pool_ns));
 }
 
-int ceph_drop_inode(struct inode *inode)
-{
-	/*
-	 * Positve dentry and corresponding inode are always accompanied
-	 * in MDS reply. So no need to keep inode in the cache after
-	 * dropping all its aliases.
-	 */
-	return 1;
-}
-
 static inline blkcnt_t calc_inode_blocks(u64 size)
 {
 	return (size + (1<<9) - 1) >> 9;
@@ -795,6 +791,9 @@ static int fill_inode(struct inode *inode, struct page *locked_page,
 	     le64_to_cpu(info->version) > (ci->i_version & ~1)))
 		new_version = true;
 
+	/* Update change_attribute */
+	inode_set_max_iversion_raw(inode, iinfo->change_attr);
+
 	__ceph_caps_issued(ci, &issued);
 	issued |= __ceph_caps_dirty(ci);
 	new_issued = ~issued & info_caps;
@@ -813,6 +812,8 @@ static int fill_inode(struct inode *inode, struct page *locked_page,
 		dout("%p mode 0%o uid.gid %d.%d\n", inode, inode->i_mode,
 		     from_kuid(&init_user_ns, inode->i_uid),
 		     from_kgid(&init_user_ns, inode->i_gid));
+		ceph_decode_timespec64(&ci->i_btime, &iinfo->btime);
+		ceph_decode_timespec64(&ci->i_snap_btime, &iinfo->snap_btime);
 	}
 
 	if ((new_version || (new_issued & CEPH_CAP_LINK_SHARED)) &&
@@ -887,6 +888,7 @@ static int fill_inode(struct inode *inode, struct page *locked_page,
 			       iinfo->xattr_data, iinfo->xattr_len);
 		ci->i_xattrs.version = le64_to_cpu(info->xattr_version);
 		ceph_forget_all_cached_acls(inode);
+		ceph_security_invalidate_secctx(inode);
 		xattr_blob = NULL;
 	}
 
@@ -1027,59 +1029,38 @@ out:
 }
 
 /*
- * caller should hold session s_mutex.
+ * caller should hold session s_mutex and dentry->d_lock.
  */
-static void update_dentry_lease(struct dentry *dentry,
-				struct ceph_mds_reply_lease *lease,
-				struct ceph_mds_session *session,
-				unsigned long from_time,
-				struct ceph_vino *tgt_vino,
-				struct ceph_vino *dir_vino)
+static void __update_dentry_lease(struct inode *dir, struct dentry *dentry,
+				  struct ceph_mds_reply_lease *lease,
+				  struct ceph_mds_session *session,
+				  unsigned long from_time,
+				  struct ceph_mds_session **old_lease_session)
 {
 	struct ceph_dentry_info *di = ceph_dentry(dentry);
 	long unsigned duration = le32_to_cpu(lease->duration_ms);
 	long unsigned ttl = from_time + (duration * HZ) / 1000;
 	long unsigned half_ttl = from_time + (duration * HZ / 2) / 1000;
-	struct inode *dir;
-	struct ceph_mds_session *old_lease_session = NULL;
 
-	/*
-	 * Make sure dentry's inode matches tgt_vino. NULL tgt_vino means that
-	 * we expect a negative dentry.
-	 */
-	if (!tgt_vino && d_really_is_positive(dentry))
-		return;
-
-	if (tgt_vino && (d_really_is_negative(dentry) ||
-			!ceph_ino_compare(d_inode(dentry), tgt_vino)))
-		return;
-
-	spin_lock(&dentry->d_lock);
 	dout("update_dentry_lease %p duration %lu ms ttl %lu\n",
 	     dentry, duration, ttl);
 
-	dir = d_inode(dentry->d_parent);
-
-	/* make sure parent matches dir_vino */
-	if (!ceph_ino_compare(dir, dir_vino))
-		goto out_unlock;
-
 	/* only track leases on regular dentries */
 	if (ceph_snap(dir) != CEPH_NOSNAP)
-		goto out_unlock;
+		return;
 
 	di->lease_shared_gen = atomic_read(&ceph_inode(dir)->i_shared_gen);
 	if (duration == 0) {
 		__ceph_dentry_dir_lease_touch(di);
-		goto out_unlock;
+		return;
 	}
 
 	if (di->lease_gen == session->s_cap_gen &&
 	    time_before(ttl, di->time))
-		goto out_unlock;  /* we already have a newer lease. */
+		return;  /* we already have a newer lease. */
 
 	if (di->lease_session && di->lease_session != session) {
-		old_lease_session = di->lease_session;
+		*old_lease_session = di->lease_session;
 		di->lease_session = NULL;
 	}
 
@@ -1092,6 +1073,62 @@ static void update_dentry_lease(struct dentry *dentry,
 	di->time = ttl;
 
 	__ceph_dentry_lease_touch(di);
+}
+
+static inline void update_dentry_lease(struct inode *dir, struct dentry *dentry,
+					struct ceph_mds_reply_lease *lease,
+					struct ceph_mds_session *session,
+					unsigned long from_time)
+{
+	struct ceph_mds_session *old_lease_session = NULL;
+	spin_lock(&dentry->d_lock);
+	__update_dentry_lease(dir, dentry, lease, session, from_time,
+			      &old_lease_session);
+	spin_unlock(&dentry->d_lock);
+	if (old_lease_session)
+		ceph_put_mds_session(old_lease_session);
+}
+
+/*
+ * update dentry lease without having parent inode locked
+ */
+static void update_dentry_lease_careful(struct dentry *dentry,
+					struct ceph_mds_reply_lease *lease,
+					struct ceph_mds_session *session,
+					unsigned long from_time,
+					char *dname, u32 dname_len,
+					struct ceph_vino *pdvino,
+					struct ceph_vino *ptvino)
+
+{
+	struct inode *dir;
+	struct ceph_mds_session *old_lease_session = NULL;
+
+	spin_lock(&dentry->d_lock);
+	/* make sure dentry's name matches target */
+	if (dentry->d_name.len != dname_len ||
+	    memcmp(dentry->d_name.name, dname, dname_len))
+		goto out_unlock;
+
+	dir = d_inode(dentry->d_parent);
+	/* make sure parent matches dvino */
+	if (!ceph_ino_compare(dir, pdvino))
+		goto out_unlock;
+
+	/* make sure dentry's inode matches target. NULL ptvino means that
+	 * we expect a negative dentry */
+	if (ptvino) {
+		if (d_really_is_negative(dentry))
+			goto out_unlock;
+		if (!ceph_ino_compare(d_inode(dentry), ptvino))
+			goto out_unlock;
+	} else {
+		if (d_really_is_positive(dentry))
+			goto out_unlock;
+	}
+
+	__update_dentry_lease(dir, dentry, lease, session,
+			      from_time, &old_lease_session);
 out_unlock:
 	spin_unlock(&dentry->d_lock);
 	if (old_lease_session)
@@ -1156,19 +1193,6 @@ static int splice_dentry(struct dentry **pdn, struct inode *in)
 	return 0;
 }
 
-static int d_name_cmp(struct dentry *dentry, const char *name, size_t len)
-{
-	int ret;
-
-	/* take d_lock to ensure dentry->d_name stability */
-	spin_lock(&dentry->d_lock);
-	ret = dentry->d_name.len - len;
-	if (!ret)
-		ret = memcmp(dentry->d_name.name, name, len);
-	spin_unlock(&dentry->d_lock);
-	return ret;
-}
-
 /*
  * Incorporate results into the local cache.  This is either just
  * one inode, or a directory, dentry, and possibly linked-to inode (e.g.,
@@ -1371,10 +1395,9 @@ retry_lookup:
 			} else if (have_lease) {
 				if (d_unhashed(dn))
 					d_add(dn, NULL);
-				update_dentry_lease(dn, rinfo->dlease,
-						    session,
-						    req->r_request_started,
-						    NULL, &dvino);
+				update_dentry_lease(dir, dn,
+						    rinfo->dlease, session,
+						    req->r_request_started);
 			}
 			goto done;
 		}
@@ -1396,11 +1419,9 @@ retry_lookup:
 		}
 
 		if (have_lease) {
-			tvino.ino = le64_to_cpu(rinfo->targeti.in->ino);
-			tvino.snap = le64_to_cpu(rinfo->targeti.in->snapid);
-			update_dentry_lease(dn, rinfo->dlease, session,
-					    req->r_request_started,
-					    &tvino, &dvino);
+			update_dentry_lease(dir, dn,
+					    rinfo->dlease, session,
+					    req->r_request_started);
 		}
 		dout(" final dn %p\n", dn);
 	} else if ((req->r_op == CEPH_MDS_OP_LOOKUPSNAP ||
@@ -1418,27 +1439,20 @@ retry_lookup:
 		err = splice_dentry(&req->r_dentry, in);
 		if (err < 0)
 			goto done;
-	} else if (rinfo->head->is_dentry &&
-		   !d_name_cmp(req->r_dentry, rinfo->dname, rinfo->dname_len)) {
+	} else if (rinfo->head->is_dentry && req->r_dentry) {
+		/* parent inode is not locked, be carefull */
 		struct ceph_vino *ptvino = NULL;
-
-		if ((le32_to_cpu(rinfo->diri.in->cap.caps) & CEPH_CAP_FILE_SHARED) ||
-		    le32_to_cpu(rinfo->dlease->duration_ms)) {
-			dvino.ino = le64_to_cpu(rinfo->diri.in->ino);
-			dvino.snap = le64_to_cpu(rinfo->diri.in->snapid);
-
-			if (rinfo->head->is_target) {
-				tvino.ino = le64_to_cpu(rinfo->targeti.in->ino);
-				tvino.snap = le64_to_cpu(rinfo->targeti.in->snapid);
-				ptvino = &tvino;
-			}
-
-			update_dentry_lease(req->r_dentry, rinfo->dlease,
-				session, req->r_request_started, ptvino,
-				&dvino);
-		} else {
-			dout("%s: no dentry lease or dir cap\n", __func__);
+		dvino.ino = le64_to_cpu(rinfo->diri.in->ino);
+		dvino.snap = le64_to_cpu(rinfo->diri.in->snapid);
+		if (rinfo->head->is_target) {
+			tvino.ino = le64_to_cpu(rinfo->targeti.in->ino);
+			tvino.snap = le64_to_cpu(rinfo->targeti.in->snapid);
+			ptvino = &tvino;
 		}
+		update_dentry_lease_careful(req->r_dentry, rinfo->dlease,
+					    session, req->r_request_started,
+					    rinfo->dname, rinfo->dname_len,
+					    &dvino, ptvino);
 	}
 done:
 	dout("fill_trace done err=%d\n", err);
@@ -1600,7 +1614,7 @@ int ceph_readdir_prepopulate(struct ceph_mds_request *req,
 	/* FIXME: release caps/leases if error occurs */
 	for (i = 0; i < rinfo->dir_nr; i++) {
 		struct ceph_mds_reply_dir_entry *rde = rinfo->dir_entries + i;
-		struct ceph_vino tvino, dvino;
+		struct ceph_vino tvino;
 
 		dname.name = rde->name;
 		dname.len = rde->name_len;
@@ -1701,9 +1715,9 @@ retry_lookup:
 
 		ceph_dentry(dn)->offset = rde->offset;
 
-		dvino = ceph_vino(d_inode(parent));
-		update_dentry_lease(dn, rde->lease, req->r_session,
-				    req->r_request_started, &tvino, &dvino);
+		update_dentry_lease(d_inode(parent), dn,
+				    rde->lease, req->r_session,
+				    req->r_request_started);
 
 		if (err == 0 && skipped == 0 && cache_ctl.index >= 0) {
 			ret = fill_readdir_cache(d_inode(parent), dn,
@@ -2282,7 +2296,7 @@ static int statx_to_caps(u32 want)
 {
 	int mask = 0;
 
-	if (want & (STATX_MODE|STATX_UID|STATX_GID|STATX_CTIME))
+	if (want & (STATX_MODE|STATX_UID|STATX_GID|STATX_CTIME|STATX_BTIME))
 		mask |= CEPH_CAP_AUTH_SHARED;
 
 	if (want & (STATX_NLINK|STATX_CTIME))
@@ -2307,6 +2321,7 @@ int ceph_getattr(const struct path *path, struct kstat *stat,
 {
 	struct inode *inode = d_inode(path->dentry);
 	struct ceph_inode_info *ci = ceph_inode(inode);
+	u32 valid_mask = STATX_BASIC_STATS;
 	int err = 0;
 
 	/* Skip the getattr altogether if we're asked not to sync */
@@ -2319,6 +2334,16 @@ int ceph_getattr(const struct path *path, struct kstat *stat,
 
 	generic_fillattr(inode, stat);
 	stat->ino = ceph_translate_ino(inode->i_sb, inode->i_ino);
+
+	/*
+	 * btime on newly-allocated inodes is 0, so if this is still set to
+	 * that, then assume that it's not valid.
+	 */
+	if (ci->i_btime.tv_sec || ci->i_btime.tv_nsec) {
+		stat->btime = ci->i_btime;
+		valid_mask |= STATX_BTIME;
+	}
+
 	if (ceph_snap(inode) == CEPH_NOSNAP)
 		stat->dev = inode->i_sb->s_dev;
 	else
@@ -2342,7 +2367,6 @@ int ceph_getattr(const struct path *path, struct kstat *stat,
 			stat->nlink = 1 + 1 + ci->i_subdirs;
 	}
 
-	/* Mask off any higher bits (e.g. btime) until we have support */
-	stat->result_mask = request_mask & STATX_BASIC_STATS;
+	stat->result_mask = request_mask & valid_mask;
 	return err;
 }