Merge tag 'ceph-for-4.17-rc1' of git://github.com/ceph/ceph-client

Pull ceph updates from Ilya Dryomov:
 "The big ticket items are:

   - support for rbd "fancy" striping (myself).

     The striping feature bit is now fully implemented, allowing mapping
     v2 images with non-default striping patterns. This completes
     support for --image-format 2.

   - CephFS quota support (Luis Henriques and Zheng Yan).

     This set is based on the new SnapRealm code in the upcoming v13.y.z
     ("Mimic") release. Quota handling will be rejected on older
     filesystems.

   - memory usage improvements in CephFS (Chengguang Xu).

     Directory specific bits have been split out of ceph_file_info and
     some effort went into improving cap reservation code to avoid OOM
     crashes.

  Also included a bunch of assorted fixes all over the place from
  Chengguang and others"

* tag 'ceph-for-4.17-rc1' of git://github.com/ceph/ceph-client: (67 commits)
  ceph: quota: report root dir quota usage in statfs
  ceph: quota: add counter for snaprealms with quota
  ceph: quota: cache inode pointer in ceph_snap_realm
  ceph: fix root quota realm check
  ceph: don't check quota for snap inode
  ceph: quota: update MDS when max_bytes is approaching
  ceph: quota: support for ceph.quota.max_bytes
  ceph: quota: don't allow cross-quota renames
  ceph: quota: support for ceph.quota.max_files
  ceph: quota: add initial infrastructure to support cephfs quotas
  rbd: remove VLA usage
  rbd: fix spelling mistake: "reregisteration" -> "reregistration"
  ceph: rename function drop_leases() to a more descriptive name
  ceph: fix invalid point dereference for error case in mdsc destroy
  ceph: return proper bool type to caller instead of pointer
  ceph: optimize memory usage
  ceph: optimize mds session register
  libceph, ceph: add __init attribution to init funcitons
  ceph: filter out used flags when printing unused open flags
  ceph: don't wait on writeback when there is no more dirty pages
  ...

1 files changed, 112 insertions, 92 deletions

diff --git a/fs/ceph/dir.c b/fs/ceph/dir.c
index 2bdd561c4c68..1a78dd6f8bf2 100644
--- a/fs/ceph/dir.c
+++ b/fs/ceph/dir.c
@@ -101,18 +101,18 @@ static int fpos_cmp(loff_t l, loff_t r)
  * regardless of what dir changes take place on the
  * server.
  */
-static int note_last_dentry(struct ceph_file_info *fi, const char *name,
+static int note_last_dentry(struct ceph_dir_file_info *dfi, const char *name,
 		            int len, unsigned next_offset)
 {
 	char *buf = kmalloc(len+1, GFP_KERNEL);
 	if (!buf)
 		return -ENOMEM;
-	kfree(fi->last_name);
-	fi->last_name = buf;
-	memcpy(fi->last_name, name, len);
-	fi->last_name[len] = 0;
-	fi->next_offset = next_offset;
-	dout("note_last_dentry '%s'\n", fi->last_name);
+	kfree(dfi->last_name);
+	dfi->last_name = buf;
+	memcpy(dfi->last_name, name, len);
+	dfi->last_name[len] = 0;
+	dfi->next_offset = next_offset;
+	dout("note_last_dentry '%s'\n", dfi->last_name);
 	return 0;
 }
 
@@ -174,7 +174,7 @@ __dcache_find_get_entry(struct dentry *parent, u64 idx,
 static int __dcache_readdir(struct file *file,  struct dir_context *ctx,
 			    int shared_gen)
 {
-	struct ceph_file_info *fi = file->private_data;
+	struct ceph_dir_file_info *dfi = file->private_data;
 	struct dentry *parent = file->f_path.dentry;
 	struct inode *dir = d_inode(parent);
 	struct dentry *dentry, *last = NULL;
@@ -221,7 +221,7 @@ static int __dcache_readdir(struct file *file,  struct dir_context *ctx,
 		bool emit_dentry = false;
 		dentry = __dcache_find_get_entry(parent, idx++, &cache_ctl);
 		if (!dentry) {
-			fi->flags |= CEPH_F_ATEND;
+			dfi->file_info.flags |= CEPH_F_ATEND;
 			err = 0;
 			break;
 		}
@@ -272,33 +272,33 @@ out:
 	if (last) {
 		int ret;
 		di = ceph_dentry(last);
-		ret = note_last_dentry(fi, last->d_name.name, last->d_name.len,
+		ret = note_last_dentry(dfi, last->d_name.name, last->d_name.len,
 				       fpos_off(di->offset) + 1);
 		if (ret < 0)
 			err = ret;
 		dput(last);
 		/* last_name no longer match cache index */
-		if (fi->readdir_cache_idx >= 0) {
-			fi->readdir_cache_idx = -1;
-			fi->dir_release_count = 0;
+		if (dfi->readdir_cache_idx >= 0) {
+			dfi->readdir_cache_idx = -1;
+			dfi->dir_release_count = 0;
 		}
 	}
 	return err;
 }
 
-static bool need_send_readdir(struct ceph_file_info *fi, loff_t pos)
+static bool need_send_readdir(struct ceph_dir_file_info *dfi, loff_t pos)
 {
-	if (!fi->last_readdir)
+	if (!dfi->last_readdir)
 		return true;
 	if (is_hash_order(pos))
-		return !ceph_frag_contains_value(fi->frag, fpos_hash(pos));
+		return !ceph_frag_contains_value(dfi->frag, fpos_hash(pos));
 	else
-		return fi->frag != fpos_frag(pos);
+		return dfi->frag != fpos_frag(pos);
 }
 
 static int ceph_readdir(struct file *file, struct dir_context *ctx)
 {
-	struct ceph_file_info *fi = file->private_data;
+	struct ceph_dir_file_info *dfi = file->private_data;
 	struct inode *inode = file_inode(file);
 	struct ceph_inode_info *ci = ceph_inode(inode);
 	struct ceph_fs_client *fsc = ceph_inode_to_client(inode);
@@ -309,7 +309,7 @@ static int ceph_readdir(struct file *file, struct dir_context *ctx)
 	struct ceph_mds_reply_info_parsed *rinfo;
 
 	dout("readdir %p file %p pos %llx\n", inode, file, ctx->pos);
-	if (fi->flags & CEPH_F_ATEND)
+	if (dfi->file_info.flags & CEPH_F_ATEND)
 		return 0;
 
 	/* always start with . and .. */
@@ -350,15 +350,15 @@ static int ceph_readdir(struct file *file, struct dir_context *ctx)
 	/* proceed with a normal readdir */
 more:
 	/* do we have the correct frag content buffered? */
-	if (need_send_readdir(fi, ctx->pos)) {
+	if (need_send_readdir(dfi, ctx->pos)) {
 		struct ceph_mds_request *req;
 		int op = ceph_snap(inode) == CEPH_SNAPDIR ?
 			CEPH_MDS_OP_LSSNAP : CEPH_MDS_OP_READDIR;
 
 		/* discard old result, if any */
-		if (fi->last_readdir) {
-			ceph_mdsc_put_request(fi->last_readdir);
-			fi->last_readdir = NULL;
+		if (dfi->last_readdir) {
+			ceph_mdsc_put_request(dfi->last_readdir);
+			dfi->last_readdir = NULL;
 		}
 
 		if (is_hash_order(ctx->pos)) {
@@ -372,7 +372,7 @@ more:
 		}
 
 		dout("readdir fetching %llx.%llx frag %x offset '%s'\n",
-		     ceph_vinop(inode), frag, fi->last_name);
+		     ceph_vinop(inode), frag, dfi->last_name);
 		req = ceph_mdsc_create_request(mdsc, op, USE_AUTH_MDS);
 		if (IS_ERR(req))
 			return PTR_ERR(req);
@@ -388,8 +388,8 @@ more:
 			__set_bit(CEPH_MDS_R_DIRECT_IS_HASH, &req->r_req_flags);
 			req->r_inode_drop = CEPH_CAP_FILE_EXCL;
 		}
-		if (fi->last_name) {
-			req->r_path2 = kstrdup(fi->last_name, GFP_KERNEL);
+		if (dfi->last_name) {
+			req->r_path2 = kstrdup(dfi->last_name, GFP_KERNEL);
 			if (!req->r_path2) {
 				ceph_mdsc_put_request(req);
 				return -ENOMEM;
@@ -399,10 +399,10 @@ more:
 				cpu_to_le32(fpos_hash(ctx->pos));
 		}
 
-		req->r_dir_release_cnt = fi->dir_release_count;
-		req->r_dir_ordered_cnt = fi->dir_ordered_count;
-		req->r_readdir_cache_idx = fi->readdir_cache_idx;
-		req->r_readdir_offset = fi->next_offset;
+		req->r_dir_release_cnt = dfi->dir_release_count;
+		req->r_dir_ordered_cnt = dfi->dir_ordered_count;
+		req->r_readdir_cache_idx = dfi->readdir_cache_idx;
+		req->r_readdir_offset = dfi->next_offset;
 		req->r_args.readdir.frag = cpu_to_le32(frag);
 		req->r_args.readdir.flags =
 				cpu_to_le16(CEPH_READDIR_REPLY_BITFLAGS);
@@ -426,35 +426,35 @@ more:
 		if (le32_to_cpu(rinfo->dir_dir->frag) != frag) {
 			frag = le32_to_cpu(rinfo->dir_dir->frag);
 			if (!rinfo->hash_order) {
-				fi->next_offset = req->r_readdir_offset;
+				dfi->next_offset = req->r_readdir_offset;
 				/* adjust ctx->pos to beginning of frag */
 				ctx->pos = ceph_make_fpos(frag,
-							  fi->next_offset,
+							  dfi->next_offset,
 							  false);
 			}
 		}
 
-		fi->frag = frag;
-		fi->last_readdir = req;
+		dfi->frag = frag;
+		dfi->last_readdir = req;
 
 		if (test_bit(CEPH_MDS_R_DID_PREPOPULATE, &req->r_req_flags)) {
-			fi->readdir_cache_idx = req->r_readdir_cache_idx;
-			if (fi->readdir_cache_idx < 0) {
+			dfi->readdir_cache_idx = req->r_readdir_cache_idx;
+			if (dfi->readdir_cache_idx < 0) {
 				/* preclude from marking dir ordered */
-				fi->dir_ordered_count = 0;
+				dfi->dir_ordered_count = 0;
 			} else if (ceph_frag_is_leftmost(frag) &&
-				   fi->next_offset == 2) {
+				   dfi->next_offset == 2) {
 				/* note dir version at start of readdir so
 				 * we can tell if any dentries get dropped */
-				fi->dir_release_count = req->r_dir_release_cnt;
-				fi->dir_ordered_count = req->r_dir_ordered_cnt;
+				dfi->dir_release_count = req->r_dir_release_cnt;
+				dfi->dir_ordered_count = req->r_dir_ordered_cnt;
 			}
 		} else {
-			dout("readdir !did_prepopulate");
+			dout("readdir !did_prepopulate\n");
 			/* disable readdir cache */
-			fi->readdir_cache_idx = -1;
+			dfi->readdir_cache_idx = -1;
 			/* preclude from marking dir complete */
-			fi->dir_release_count = 0;
+			dfi->dir_release_count = 0;
 		}
 
 		/* note next offset and last dentry name */
@@ -463,19 +463,19 @@ more:
 					rinfo->dir_entries + (rinfo->dir_nr-1);
 			unsigned next_offset = req->r_reply_info.dir_end ?
 					2 : (fpos_off(rde->offset) + 1);
-			err = note_last_dentry(fi, rde->name, rde->name_len,
+			err = note_last_dentry(dfi, rde->name, rde->name_len,
 					       next_offset);
 			if (err)
 				return err;
 		} else if (req->r_reply_info.dir_end) {
-			fi->next_offset = 2;
+			dfi->next_offset = 2;
 			/* keep last name */
 		}
 	}
 
-	rinfo = &fi->last_readdir->r_reply_info;
+	rinfo = &dfi->last_readdir->r_reply_info;
 	dout("readdir frag %x num %d pos %llx chunk first %llx\n",
-	     fi->frag, rinfo->dir_nr, ctx->pos,
+	     dfi->frag, rinfo->dir_nr, ctx->pos,
 	     rinfo->dir_nr ? rinfo->dir_entries[0].offset : 0LL);
 
 	i = 0;
@@ -519,52 +519,55 @@ more:
 		ctx->pos++;
 	}
 
-	ceph_mdsc_put_request(fi->last_readdir);
-	fi->last_readdir = NULL;
+	ceph_mdsc_put_request(dfi->last_readdir);
+	dfi->last_readdir = NULL;
 
-	if (fi->next_offset > 2) {
-		frag = fi->frag;
+	if (dfi->next_offset > 2) {
+		frag = dfi->frag;
 		goto more;
 	}
 
 	/* more frags? */
-	if (!ceph_frag_is_rightmost(fi->frag)) {
-		frag = ceph_frag_next(fi->frag);
+	if (!ceph_frag_is_rightmost(dfi->frag)) {
+		frag = ceph_frag_next(dfi->frag);
 		if (is_hash_order(ctx->pos)) {
 			loff_t new_pos = ceph_make_fpos(ceph_frag_value(frag),
-							fi->next_offset, true);
+							dfi->next_offset, true);
 			if (new_pos > ctx->pos)
 				ctx->pos = new_pos;
 			/* keep last_name */
 		} else {
-			ctx->pos = ceph_make_fpos(frag, fi->next_offset, false);
-			kfree(fi->last_name);
-			fi->last_name = NULL;
+			ctx->pos = ceph_make_fpos(frag, dfi->next_offset,
+							false);
+			kfree(dfi->last_name);
+			dfi->last_name = NULL;
 		}
 		dout("readdir next frag is %x\n", frag);
 		goto more;
 	}
-	fi->flags |= CEPH_F_ATEND;
+	dfi->file_info.flags |= CEPH_F_ATEND;
 
 	/*
 	 * if dir_release_count still matches the dir, no dentries
 	 * were released during the whole readdir, and we should have
 	 * the complete dir contents in our cache.
 	 */
-	if (atomic64_read(&ci->i_release_count) == fi->dir_release_count) {
+	if (atomic64_read(&ci->i_release_count) ==
+					dfi->dir_release_count) {
 		spin_lock(&ci->i_ceph_lock);
-		if (fi->dir_ordered_count == atomic64_read(&ci->i_ordered_count)) {
+		if (dfi->dir_ordered_count ==
+				atomic64_read(&ci->i_ordered_count)) {
 			dout(" marking %p complete and ordered\n", inode);
 			/* use i_size to track number of entries in
 			 * readdir cache */
-			BUG_ON(fi->readdir_cache_idx < 0);
-			i_size_write(inode, fi->readdir_cache_idx *
+			BUG_ON(dfi->readdir_cache_idx < 0);
+			i_size_write(inode, dfi->readdir_cache_idx *
 				     sizeof(struct dentry*));
 		} else {
 			dout(" marking %p complete\n", inode);
 		}
-		__ceph_dir_set_complete(ci, fi->dir_release_count,
-					fi->dir_ordered_count);
+		__ceph_dir_set_complete(ci, dfi->dir_release_count,
+					dfi->dir_ordered_count);
 		spin_unlock(&ci->i_ceph_lock);
 	}
 
@@ -572,25 +575,25 @@ more:
 	return 0;
 }
 
-static void reset_readdir(struct ceph_file_info *fi)
+static void reset_readdir(struct ceph_dir_file_info *dfi)
 {
-	if (fi->last_readdir) {
-		ceph_mdsc_put_request(fi->last_readdir);
-		fi->last_readdir = NULL;
+	if (dfi->last_readdir) {
+		ceph_mdsc_put_request(dfi->last_readdir);
+		dfi->last_readdir = NULL;
 	}
-	kfree(fi->last_name);
-	fi->last_name = NULL;
-	fi->dir_release_count = 0;
-	fi->readdir_cache_idx = -1;
-	fi->next_offset = 2;  /* compensate for . and .. */
-	fi->flags &= ~CEPH_F_ATEND;
+	kfree(dfi->last_name);
+	dfi->last_name = NULL;
+	dfi->dir_release_count = 0;
+	dfi->readdir_cache_idx = -1;
+	dfi->next_offset = 2;  /* compensate for . and .. */
+	dfi->file_info.flags &= ~CEPH_F_ATEND;
 }
 
 /*
  * discard buffered readdir content on seekdir(0), or seek to new frag,
  * or seek prior to current chunk
  */
-static bool need_reset_readdir(struct ceph_file_info *fi, loff_t new_pos)
+static bool need_reset_readdir(struct ceph_dir_file_info *dfi, loff_t new_pos)
 {
 	struct ceph_mds_reply_info_parsed *rinfo;
 	loff_t chunk_offset;
@@ -599,10 +602,10 @@ static bool need_reset_readdir(struct ceph_file_info *fi, loff_t new_pos)
 	if (is_hash_order(new_pos)) {
 		/* no need to reset last_name for a forward seek when
 		 * dentries are sotred in hash order */
-	} else if (fi->frag != fpos_frag(new_pos)) {
+	} else if (dfi->frag != fpos_frag(new_pos)) {
 		return true;
 	}
-	rinfo = fi->last_readdir ? &fi->last_readdir->r_reply_info : NULL;
+	rinfo = dfi->last_readdir ? &dfi->last_readdir->r_reply_info : NULL;
 	if (!rinfo || !rinfo->dir_nr)
 		return true;
 	chunk_offset = rinfo->dir_entries[0].offset;
@@ -612,7 +615,7 @@ static bool need_reset_readdir(struct ceph_file_info *fi, loff_t new_pos)
 
 static loff_t ceph_dir_llseek(struct file *file, loff_t offset, int whence)
 {
-	struct ceph_file_info *fi = file->private_data;
+	struct ceph_dir_file_info *dfi = file->private_data;
 	struct inode *inode = file->f_mapping->host;
 	loff_t retval;
 
@@ -630,20 +633,20 @@ static loff_t ceph_dir_llseek(struct file *file, loff_t offset, int whence)
 	}
 
 	if (offset >= 0) {
-		if (need_reset_readdir(fi, offset)) {
+		if (need_reset_readdir(dfi, offset)) {
 			dout("dir_llseek dropping %p content\n", file);
-			reset_readdir(fi);
+			reset_readdir(dfi);
 		} else if (is_hash_order(offset) && offset > file->f_pos) {
 			/* for hash offset, we don't know if a forward seek
 			 * is within same frag */
-			fi->dir_release_count = 0;
-			fi->readdir_cache_idx = -1;
+			dfi->dir_release_count = 0;
+			dfi->readdir_cache_idx = -1;
 		}
 
 		if (offset != file->f_pos) {
 			file->f_pos = offset;
 			file->f_version = 0;
-			fi->flags &= ~CEPH_F_ATEND;
+			dfi->file_info.flags &= ~CEPH_F_ATEND;
 		}
 		retval = offset;
 	}
@@ -824,6 +827,9 @@ static int ceph_mknod(struct inode *dir, struct dentry *dentry,
 	if (ceph_snap(dir) != CEPH_NOSNAP)
 		return -EROFS;
 
+	if (ceph_quota_is_max_files_exceeded(dir))
+		return -EDQUOT;
+
 	err = ceph_pre_init_acls(dir, &mode, &acls);
 	if (err < 0)
 		return err;
@@ -877,6 +883,9 @@ static int ceph_symlink(struct inode *dir, struct dentry *dentry,
 	if (ceph_snap(dir) != CEPH_NOSNAP)
 		return -EROFS;
 
+	if (ceph_quota_is_max_files_exceeded(dir))
+		return -EDQUOT;
+
 	dout("symlink in dir %p dentry %p to '%s'\n", dir, dentry, dest);
 	req = ceph_mdsc_create_request(mdsc, CEPH_MDS_OP_SYMLINK, USE_AUTH_MDS);
 	if (IS_ERR(req)) {
@@ -926,6 +935,12 @@ static int ceph_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode)
 		goto out;
 	}
 
+	if (op == CEPH_MDS_OP_MKDIR &&
+	    ceph_quota_is_max_files_exceeded(dir)) {
+		err = -EDQUOT;
+		goto out;
+	}
+
 	mode |= S_IFDIR;
 	err = ceph_pre_init_acls(dir, &mode, &acls);
 	if (err < 0)
@@ -1065,6 +1080,11 @@ static int ceph_rename(struct inode *old_dir, struct dentry *old_dentry,
 		else
 			return -EROFS;
 	}
+	/* don't allow cross-quota renames */
+	if ((old_dir != new_dir) &&
+	    (!ceph_quota_is_same_realm(old_dir, new_dir)))
+		return -EXDEV;
+
 	dout("rename dir %p dentry %p to dir %p dentry %p\n",
 	     old_dir, old_dentry, new_dir, new_dentry);
 	req = ceph_mdsc_create_request(mdsc, op, USE_AUTH_MDS);
@@ -1351,7 +1371,7 @@ static void ceph_d_prune(struct dentry *dentry)
 static ssize_t ceph_read_dir(struct file *file, char __user *buf, size_t size,
 			     loff_t *ppos)
 {
-	struct ceph_file_info *cf = file->private_data;
+	struct ceph_dir_file_info *dfi = file->private_data;
 	struct inode *inode = file_inode(file);
 	struct ceph_inode_info *ci = ceph_inode(inode);
 	int left;
@@ -1360,12 +1380,12 @@ static ssize_t ceph_read_dir(struct file *file, char __user *buf, size_t size,
 	if (!ceph_test_mount_opt(ceph_sb_to_client(inode->i_sb), DIRSTAT))
 		return -EISDIR;
 
-	if (!cf->dir_info) {
-		cf->dir_info = kmalloc(bufsize, GFP_KERNEL);
-		if (!cf->dir_info)
+	if (!dfi->dir_info) {
+		dfi->dir_info = kmalloc(bufsize, GFP_KERNEL);
+		if (!dfi->dir_info)
 			return -ENOMEM;
-		cf->dir_info_len =
-			snprintf(cf->dir_info, bufsize,
+		dfi->dir_info_len =
+			snprintf(dfi->dir_info, bufsize,
 				"entries:   %20lld\n"
 				" files:    %20lld\n"
 				" subdirs:  %20lld\n"
@@ -1385,10 +1405,10 @@ static ssize_t ceph_read_dir(struct file *file, char __user *buf, size_t size,
 				(long)ci->i_rctime.tv_nsec);
 	}
 
-	if (*ppos >= cf->dir_info_len)
+	if (*ppos >= dfi->dir_info_len)
 		return 0;
-	size = min_t(unsigned, size, cf->dir_info_len-*ppos);
-	left = copy_to_user(buf, cf->dir_info + *ppos, size);
+	size = min_t(unsigned, size, dfi->dir_info_len-*ppos);
+	left = copy_to_user(buf, dfi->dir_info + *ppos, size);
 	if (left == size)
 		return -EFAULT;
 	*ppos += (size - left);