Merge tag 'ceph-for-5.9-rc3' of git://github.com/ceph/ceph-client

Pull ceph fixes from Ilya Dryomov:
 "We have an inode number handling change, prompted by s390x which is a
  64-bit architecture with a 32-bit ino_t, a patch to disallow leases to
  avoid potential data integrity issues when CephFS is re-exported via
  NFS or CIFS and a fix for the bulk of W=1 compilation warnings"

* tag 'ceph-for-5.9-rc3' of git://github.com/ceph/ceph-client:
  ceph: don't allow setlease on cephfs
  ceph: fix inode number handling on arches with 32-bit ino_t
  libceph: add __maybe_unused to DEFINE_CEPH_FEATURE

fs/ceph/caps.c

@@ -887,8 +887,8 @@ int __ceph_caps_issued_mask(struct ceph_inode_info *ci, int mask, int touch)
 	int have = ci->i_snap_caps;
 
 	if ((have & mask) == mask) {
-		dout("__ceph_caps_issued_mask ino 0x%lx snap issued %s"
-		     " (mask %s)\n", ci->vfs_inode.i_ino,
+		dout("__ceph_caps_issued_mask ino 0x%llx snap issued %s"
+		     " (mask %s)\n", ceph_ino(&ci->vfs_inode),
 		     ceph_cap_string(have),
 		     ceph_cap_string(mask));
 		return 1;
@@ -899,8 +899,8 @@ int __ceph_caps_issued_mask(struct ceph_inode_info *ci, int mask, int touch)
 		if (!__cap_is_valid(cap))
 			continue;
 		if ((cap->issued & mask) == mask) {
-			dout("__ceph_caps_issued_mask ino 0x%lx cap %p issued %s"
-			     " (mask %s)\n", ci->vfs_inode.i_ino, cap,
+			dout("__ceph_caps_issued_mask ino 0x%llx cap %p issued %s"
+			     " (mask %s)\n", ceph_ino(&ci->vfs_inode), cap,
 			     ceph_cap_string(cap->issued),
 			     ceph_cap_string(mask));
 			if (touch)
@@ -911,8 +911,8 @@ int __ceph_caps_issued_mask(struct ceph_inode_info *ci, int mask, int touch)
 		/* does a combination of caps satisfy mask? */
 		have |= cap->issued;
 		if ((have & mask) == mask) {
-			dout("__ceph_caps_issued_mask ino 0x%lx combo issued %s"
-			     " (mask %s)\n", ci->vfs_inode.i_ino,
+			dout("__ceph_caps_issued_mask ino 0x%llx combo issued %s"
+			     " (mask %s)\n", ceph_ino(&ci->vfs_inode),
 			     ceph_cap_string(cap->issued),
 			     ceph_cap_string(mask));
 			if (touch) {
@@ -2872,7 +2872,7 @@ int ceph_get_caps(struct file *filp, int need, int want,
 			struct cap_wait cw;
 			DEFINE_WAIT_FUNC(wait, woken_wake_function);
 
-			cw.ino = inode->i_ino;
+			cw.ino = ceph_ino(inode);
 			cw.tgid = current->tgid;
 			cw.need = need;
 			cw.want = want;

fs/ceph/debugfs.c

@@ -202,7 +202,7 @@ static int caps_show_cb(struct inode *inode, struct ceph_cap *cap, void *p)
 {
 	struct seq_file *s = p;
 
-	seq_printf(s, "0x%-17lx%-17s%-17s\n", inode->i_ino,
+	seq_printf(s, "0x%-17llx%-17s%-17s\n", ceph_ino(inode),
 		   ceph_cap_string(cap->issued),
 		   ceph_cap_string(cap->implemented));
 	return 0;
@@ -247,7 +247,7 @@ static int caps_show(struct seq_file *s, void *p)
 
 	spin_lock(&mdsc->caps_list_lock);
 	list_for_each_entry(cw, &mdsc->cap_wait_list, list) {
-		seq_printf(s, "%-13d0x%-17lx%-17s%-17s\n", cw->tgid, cw->ino,
+		seq_printf(s, "%-13d0x%-17llx%-17s%-17s\n", cw->tgid, cw->ino,
 				ceph_cap_string(cw->need),
 				ceph_cap_string(cw->want));
 	}

fs/ceph/dir.c

@@ -259,9 +259,7 @@ static int __dcache_readdir(struct file *file,  struct dir_context *ctx,
 			     dentry, dentry, d_inode(dentry));
 			ctx->pos = di->offset;
 			if (!dir_emit(ctx, dentry->d_name.name,
-				      dentry->d_name.len,
-				      ceph_translate_ino(dentry->d_sb,
-							 d_inode(dentry)->i_ino),
+				      dentry->d_name.len, ceph_present_inode(d_inode(dentry)),
 				      d_inode(dentry)->i_mode >> 12)) {
 				dput(dentry);
 				err = 0;
@@ -324,18 +322,21 @@ static int ceph_readdir(struct file *file, struct dir_context *ctx)
 	/* always start with . and .. */
 	if (ctx->pos == 0) {
 		dout("readdir off 0 -> '.'\n");
-		if (!dir_emit(ctx, ".", 1, 
-			    ceph_translate_ino(inode->i_sb, inode->i_ino),
+		if (!dir_emit(ctx, ".", 1, ceph_present_inode(inode),
 			    inode->i_mode >> 12))
 			return 0;
 		ctx->pos = 1;
 	}
 	if (ctx->pos == 1) {
-		ino_t ino = parent_ino(file->f_path.dentry);
+		u64 ino;
+		struct dentry *dentry = file->f_path.dentry;
+
+		spin_lock(&dentry->d_lock);
+		ino = ceph_present_inode(dentry->d_parent->d_inode);
+		spin_unlock(&dentry->d_lock);
+
 		dout("readdir off 1 -> '..'\n");
-		if (!dir_emit(ctx, "..", 2,
-			    ceph_translate_ino(inode->i_sb, ino),
-			    inode->i_mode >> 12))
+		if (!dir_emit(ctx, "..", 2, ino, inode->i_mode >> 12))
 			return 0;
 		ctx->pos = 2;
 	}
@@ -507,9 +508,6 @@ static int ceph_readdir(struct file *file, struct dir_context *ctx)
 	}
 	for (; i < rinfo->dir_nr; i++) {
 		struct ceph_mds_reply_dir_entry *rde = rinfo->dir_entries + i;
-		struct ceph_vino vino;
-		ino_t ino;
-		u32 ftype;
 
 		BUG_ON(rde->offset < ctx->pos);
 
@@ -519,13 +517,10 @@ static int ceph_readdir(struct file *file, struct dir_context *ctx)
 		     rde->name_len, rde->name, &rde->inode.in);
 
 		BUG_ON(!rde->inode.in);
-		ftype = le32_to_cpu(rde->inode.in->mode) >> 12;
-		vino.ino = le64_to_cpu(rde->inode.in->ino);
-		vino.snap = le64_to_cpu(rde->inode.in->snapid);
-		ino = ceph_vino_to_ino(vino);
 
 		if (!dir_emit(ctx, rde->name, rde->name_len,
-			      ceph_translate_ino(inode->i_sb, ino), ftype)) {
+			      ceph_present_ino(inode->i_sb, le64_to_cpu(rde->inode.in->ino)),
+			      le32_to_cpu(rde->inode.in->mode) >> 12)) {
 			dout("filldir stopping us...\n");
 			return 0;
 		}
@@ -1161,7 +1156,7 @@ static int ceph_unlink(struct inode *dir, struct dentry *dentry)
 
 	if (try_async && op == CEPH_MDS_OP_UNLINK &&
 	    (req->r_dir_caps = get_caps_for_async_unlink(dir, dentry))) {
-		dout("async unlink on %lu/%.*s caps=%s", dir->i_ino,
+		dout("async unlink on %llu/%.*s caps=%s", ceph_ino(dir),
 		     dentry->d_name.len, dentry->d_name.name,
 		     ceph_cap_string(req->r_dir_caps));
 		set_bit(CEPH_MDS_R_ASYNC, &req->r_req_flags);

fs/ceph/file.c

@@ -630,8 +630,8 @@ static int ceph_finish_async_create(struct inode *dir, struct dentry *dentry,
 	} else {
 		struct dentry *dn;
 
-		dout("%s d_adding new inode 0x%llx to 0x%lx/%s\n", __func__,
-			vino.ino, dir->i_ino, dentry->d_name.name);
+		dout("%s d_adding new inode 0x%llx to 0x%llx/%s\n", __func__,
+			vino.ino, ceph_ino(dir), dentry->d_name.name);
 		ceph_dir_clear_ordered(dir);
 		ceph_init_inode_acls(inode, as_ctx);
 		if (inode->i_state & I_NEW) {
@@ -2507,6 +2507,7 @@ const struct file_operations ceph_file_fops = {
 	.mmap = ceph_mmap,
 	.fsync = ceph_fsync,
 	.lock = ceph_lock,
+	.setlease = simple_nosetlease,
 	.flock = ceph_flock,
 	.splice_read = generic_file_splice_read,
 	.splice_write = iter_file_splice_write,

fs/ceph/inode.c

@@ -41,8 +41,10 @@ static void ceph_inode_work(struct work_struct *work);
  */
 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);
+	struct ceph_inode_info *ci = ceph_inode(inode);
+
+	ci->i_vino = *(struct ceph_vino *)data;
+	inode->i_ino = ceph_vino_to_ino_t(ci->i_vino);
 	inode_set_iversion_raw(inode, 0);
 	return 0;
 }
@@ -50,17 +52,14 @@ static int ceph_set_ino_cb(struct inode *inode, void *data)
 struct inode *ceph_get_inode(struct super_block *sb, struct ceph_vino vino)
 {
 	struct inode *inode;
-	ino_t t = ceph_vino_to_ino(vino);
 
-	inode = iget5_locked(sb, t, ceph_ino_compare, ceph_set_ino_cb, &vino);
+	inode = iget5_locked(sb, (unsigned long)vino.ino, ceph_ino_compare,
+			     ceph_set_ino_cb, &vino);
 	if (!inode)
 		return ERR_PTR(-ENOMEM);
-	if (inode->i_state & I_NEW)
-		dout("get_inode created new inode %p %llx.%llx ino %llx\n",
-		     inode, ceph_vinop(inode), (u64)inode->i_ino);
 
-	dout("get_inode on %lu=%llx.%llx got %p\n", inode->i_ino, vino.ino,
-	     vino.snap, inode);
+	dout("get_inode on %llu=%llx.%llx got %p new %d\n", ceph_present_inode(inode),
+	     ceph_vinop(inode), inode, !!(inode->i_state & I_NEW));
 	return inode;
 }
 
@@ -2378,7 +2377,7 @@ 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);
+	stat->ino = ceph_present_inode(inode);
 
 	/*
 	 * btime on newly-allocated inodes is 0, so if this is still set to

fs/ceph/mds_client.h

@@ -372,7 +372,7 @@ struct ceph_quotarealm_inode {
 
 struct cap_wait {
 	struct list_head	list;
-	unsigned long		ino;
+	u64			ino;
 	pid_t			tgid;
 	int			need;
 	int			want;

fs/ceph/quota.c

@@ -23,12 +23,12 @@ static inline bool ceph_has_realms_with_quotas(struct inode *inode)
 {
 	struct ceph_mds_client *mdsc = ceph_inode_to_client(inode)->mdsc;
 	struct super_block *sb = mdsc->fsc->sb;
+	struct inode *root = d_inode(sb->s_root);
 
 	if (atomic64_read(&mdsc->quotarealms_count) > 0)
 		return true;
 	/* if root is the real CephFS root, we don't have quota realms */
-	if (sb->s_root->d_inode &&
-	    (sb->s_root->d_inode->i_ino == CEPH_INO_ROOT))
+	if (root && ceph_ino(root) == CEPH_INO_ROOT)
 		return false;
 	/* otherwise, we can't know for sure */
 	return true;

fs/ceph/super.h

@@ -457,15 +457,7 @@ ceph_vino(const struct inode *inode)
 	return ceph_inode(inode)->i_vino;
 }
 
-/*
- * ino_t is <64 bits on many architectures, blech.
- *
- *               i_ino (kernel inode)   st_ino (userspace)
- * i386          32                     32
- * x86_64+ino32  64                     32
- * x86_64        64                     64
- */
-static inline u32 ceph_ino_to_ino32(__u64 vino)
+static inline u32 ceph_ino_to_ino32(u64 vino)
 {
 	u32 ino = vino & 0xffffffff;
 	ino ^= vino >> 32;
@@ -475,34 +467,17 @@ static inline u32 ceph_ino_to_ino32(__u64 vino)
 }
 
 /*
- * kernel i_ino value
+ * Inode numbers in cephfs are 64 bits, but inode->i_ino is 32-bits on
+ * some arches. We generally do not use this value inside the ceph driver, but
+ * we do want to set it to something, so that generic vfs code has an
+ * appropriate value for tracepoints and the like.
  */
-static inline ino_t ceph_vino_to_ino(struct ceph_vino vino)
+static inline ino_t ceph_vino_to_ino_t(struct ceph_vino vino)
 {
-#if BITS_PER_LONG == 32
-	return ceph_ino_to_ino32(vino.ino);
-#else
+	if (sizeof(ino_t) == sizeof(u32))
+		return ceph_ino_to_ino32(vino.ino);
 	return (ino_t)vino.ino;
-#endif
-}
-
-/*
- * user-visible ino (stat, filldir)
- */
-#if BITS_PER_LONG == 32
-static inline ino_t ceph_translate_ino(struct super_block *sb, ino_t ino)
-{
-	return ino;
-}
-#else
-static inline ino_t ceph_translate_ino(struct super_block *sb, ino_t ino)
-{
-	if (ceph_test_mount_opt(ceph_sb_to_client(sb), INO32))
-		ino = ceph_ino_to_ino32(ino);
-	return ino;
 }
-#endif
-
 
 /* for printf-style formatting */
 #define ceph_vinop(i) ceph_inode(i)->i_vino.ino, ceph_inode(i)->i_vino.snap
@@ -511,11 +486,34 @@ static inline u64 ceph_ino(struct inode *inode)
 {
 	return ceph_inode(inode)->i_vino.ino;
 }
+
 static inline u64 ceph_snap(struct inode *inode)
 {
 	return ceph_inode(inode)->i_vino.snap;
 }
 
+/**
+ * ceph_present_ino - format an inode number for presentation to userland
+ * @sb: superblock where the inode lives
+ * @ino: inode number to (possibly) convert
+ *
+ * If the user mounted with the ino32 option, then the 64-bit value needs
+ * to be converted to something that can fit inside 32 bits. Note that
+ * internal kernel code never uses this value, so this is entirely for
+ * userland consumption.
+ */
+static inline u64 ceph_present_ino(struct super_block *sb, u64 ino)
+{
+	if (unlikely(ceph_test_mount_opt(ceph_sb_to_client(sb), INO32)))
+		return ceph_ino_to_ino32(ino);
+	return ino;
+}
+
+static inline u64 ceph_present_inode(struct inode *inode)
+{
+	return ceph_present_ino(inode->i_sb, ceph_ino(inode));
+}
+
 static inline int ceph_ino_compare(struct inode *inode, void *data)
 {
 	struct ceph_vino *pvino = (struct ceph_vino *)data;
@@ -524,11 +522,16 @@ static inline int ceph_ino_compare(struct inode *inode, void *data)
 		ci->i_vino.snap == pvino->snap;
 }
 
+
 static inline struct inode *ceph_find_inode(struct super_block *sb,
 					    struct ceph_vino vino)
 {
-	ino_t t = ceph_vino_to_ino(vino);
-	return ilookup5(sb, t, ceph_ino_compare, &vino);
+	/*
+	 * NB: The hashval will be run through the fs/inode.c hash function
+	 * anyway, so there is no need to squash the inode number down to
+	 * 32-bits first. Just use low-order bits on arches with 32-bit long.
+	 */
+	return ilookup5(sb, (unsigned long)vino.ino, ceph_ino_compare, &vino);
 }
 
 

include/linux/ceph/ceph_features.h

@@ -11,14 +11,14 @@
 #define CEPH_FEATURE_INCARNATION_2 (1ull<<57) // CEPH_FEATURE_SERVER_JEWEL
 
 #define DEFINE_CEPH_FEATURE(bit, incarnation, name)			\
-	static const uint64_t CEPH_FEATURE_##name = (1ULL<<bit);		\
-	static const uint64_t CEPH_FEATUREMASK_##name =			\
+	static const uint64_t __maybe_unused CEPH_FEATURE_##name = (1ULL<<bit);		\
+	static const uint64_t __maybe_unused CEPH_FEATUREMASK_##name =			\
 		(1ULL<<bit | CEPH_FEATURE_INCARNATION_##incarnation);
 
 /* this bit is ignored but still advertised by release *when* */
 #define DEFINE_CEPH_FEATURE_DEPRECATED(bit, incarnation, name, when) \
-	static const uint64_t DEPRECATED_CEPH_FEATURE_##name = (1ULL<<bit); \
-	static const uint64_t DEPRECATED_CEPH_FEATUREMASK_##name =		\
+	static const uint64_t __maybe_unused DEPRECATED_CEPH_FEATURE_##name = (1ULL<<bit);	\
+	static const uint64_t __maybe_unused DEPRECATED_CEPH_FEATUREMASK_##name =		\
 		(1ULL<<bit | CEPH_FEATURE_INCARNATION_##incarnation);
 
 /*