Add zfs_file_* interface, remove vnodes

Provide a common zfs_file_* interface which can be implemented on all 
platforms to perform normal file access from either the kernel module
or the libzpool library.

This allows all non-portable vnode_t usage in the common code to be 
replaced by the new portable zfs_file_t.  The associated vnode and
kobj compatibility functions, types, and macros have been removed
from the SPL.  Moving forward, vnodes should only be used in platform
specific code when provided by the native operating system.

Reviewed-by: Sean Eric Fagan <[email protected]>
Reviewed-by: Brian Behlendorf <[email protected]>
Reviewed-by: Igor Kozhukhov <[email protected]>
Reviewed-by: Jorgen Lundman <[email protected]>
Signed-off-by: Matt Macy <[email protected]>
Closes #9556 

10 files changed, 80 insertions, 62 deletions

diff --git a/include/sys/Makefile.am b/include/sys/Makefile.am
index 21e85431c..75f7d93e7 100644
--- a/include/sys/Makefile.am
+++ b/include/sys/Makefile.am
@@ -104,6 +104,7 @@ COMMON_H = \
 	$(top_srcdir)/include/sys/zfs_context.h \
 	$(top_srcdir)/include/sys/zfs_debug.h \
 	$(top_srcdir)/include/sys/zfs_delay.h \
+	$(top_srcdir)/include/sys/zfs_file.h \
 	$(top_srcdir)/include/sys/zfs_fuid.h \
 	$(top_srcdir)/include/sys/zfs_project.h \
 	$(top_srcdir)/include/sys/zfs_ratelimit.h \
diff --git a/include/sys/dmu.h b/include/sys/dmu.h
index 44889bb10..24cbb2f7a 100644
--- a/include/sys/dmu.h
+++ b/include/sys/dmu.h
@@ -49,6 +49,7 @@
 #include <sys/zio_compress.h>
 #include <sys/zio_priority.h>
 #include <sys/uio.h>
+#include <sys/zfs_file.h>
 
 #ifdef	__cplusplus
 extern "C" {
@@ -1071,7 +1072,7 @@ void dmu_traverse_objset(objset_t *os, uint64_t txg_start,
     dmu_traverse_cb_t cb, void *arg);
 
 int dmu_diff(const char *tosnap_name, const char *fromsnap_name,
-    struct vnode *vp, offset_t *offp);
+    zfs_file_t *fp, offset_t *offp);
 
 /* CRC64 table */
 #define	ZFS_CRC64_POLY	0xC96C5795D7870F42ULL	/* ECMA-182, reflected form */
diff --git a/include/sys/dmu_recv.h b/include/sys/dmu_recv.h
index 1a7347d66..2cbe49c49 100644
--- a/include/sys/dmu_recv.h
+++ b/include/sys/dmu_recv.h
@@ -62,7 +62,7 @@ typedef struct dmu_recv_cookie {
 	nvlist_t *drc_begin_nvl;
 
 	objset_t *drc_os;
-	vnode_t *drc_vp; /* The vnode to read the stream from */
+	zfs_file_t *drc_fp; /* The file to read the stream from */
 	uint64_t drc_voff; /* The current offset in the stream */
 	uint64_t drc_bytes_read;
 	/*
@@ -82,7 +82,7 @@ typedef struct dmu_recv_cookie {
 int dmu_recv_begin(char *tofs, char *tosnap, dmu_replay_record_t *drr_begin,
     boolean_t force, boolean_t resumable, nvlist_t *localprops,
     nvlist_t *hidden_args, char *origin, dmu_recv_cookie_t *drc,
-    vnode_t *vp, offset_t *voffp);
+    zfs_file_t *fp, offset_t *voffp);
 int dmu_recv_stream(dmu_recv_cookie_t *drc, int cleanup_fd,
     uint64_t *action_handlep, offset_t *voffp);
 int dmu_recv_end(dmu_recv_cookie_t *drc, void *owner);
diff --git a/include/sys/spa.h b/include/sys/spa.h
index cb9157738..cae29a4f4 100644
--- a/include/sys/spa.h
+++ b/include/sys/spa.h
@@ -723,6 +723,12 @@ typedef enum spa_import_type {
 	SPA_IMPORT_ASSEMBLE
 } spa_import_type_t;
 
+typedef enum spa_mode {
+	SPA_MODE_UNINIT = 0,
+	SPA_MODE_READ = 1,
+	SPA_MODE_WRITE = 2,
+} spa_mode_t;
+
 /*
  * Send TRIM commands in-line during normal pool operation while deleting.
  *	OFF: no
@@ -1099,7 +1105,7 @@ extern uint32_t spa_get_hostid(spa_t *spa);
 extern void spa_activate_allocation_classes(spa_t *, dmu_tx_t *);
 extern boolean_t spa_livelist_delete_check(spa_t *spa);
 
-extern int spa_mode(spa_t *spa);
+extern spa_mode_t spa_mode(spa_t *spa);
 extern uint64_t zfs_strtonum(const char *str, char **nptr);
 
 extern char *spa_his_ievent_table[];
@@ -1149,7 +1155,7 @@ extern void vdev_mirror_stat_init(void);
 extern void vdev_mirror_stat_fini(void);
 
 /* Initialization and termination */
-extern void spa_init(int flags);
+extern void spa_init(spa_mode_t mode);
 extern void spa_fini(void);
 extern void spa_boot_init(void);
 
@@ -1189,7 +1195,7 @@ _NOTE(CONSTCOND) } while (0)
 #define	dprintf_bp(bp, fmt, ...)
 #endif
 
-extern int spa_mode_global;			/* mode, e.g. FREAD | FWRITE */
+extern spa_mode_t spa_mode_global;
 extern int zfs_deadman_enabled;
 extern unsigned long zfs_deadman_synctime_ms;
 extern unsigned long zfs_deadman_ziotime_ms;
diff --git a/include/sys/spa_impl.h b/include/sys/spa_impl.h
index c83252bca..5a581214d 100644
--- a/include/sys/spa_impl.h
+++ b/include/sys/spa_impl.h
@@ -362,7 +362,7 @@ struct spa {
 	uint8_t		spa_claiming;		/* pool is doing zil_claim() */
 	boolean_t	spa_is_root;		/* pool is root */
 	int		spa_minref;		/* num refs when first opened */
-	int		spa_mode;		/* FREAD | FWRITE */
+	spa_mode_t	spa_mode;		/* SPA_MODE_{READ|WRITE} */
 	spa_log_state_t spa_log_state;		/* log state */
 	uint64_t	spa_autoexpand;		/* lun expansion on/off */
 	ddt_t		*spa_ddt[ZIO_CHECKSUM_FUNCTIONS]; /* in-core DDTs */
diff --git a/include/sys/vdev_file.h b/include/sys/vdev_file.h
index 9a398c583..1514a44fc 100644
--- a/include/sys/vdev_file.h
+++ b/include/sys/vdev_file.h
@@ -34,7 +34,7 @@ extern "C" {
 #endif
 
 typedef struct vdev_file {
-	vnode_t		*vf_vnode;
+	zfs_file_t	*vf_file;
 } vdev_file_t;
 
 extern void vdev_file_init(void);
diff --git a/include/sys/vdev_impl.h b/include/sys/vdev_impl.h
index ae82b75c0..4f63e1ae5 100644
--- a/include/sys/vdev_impl.h
+++ b/include/sys/vdev_impl.h
@@ -220,8 +220,6 @@ struct vdev {
 	vdev_ops_t	*vdev_ops;	/* vdev operations		*/
 	spa_t		*vdev_spa;	/* spa for this vdev		*/
 	void		*vdev_tsd;	/* type-specific data		*/
-	vnode_t		*vdev_name_vp;	/* vnode for pathname		*/
-	vnode_t		*vdev_devid_vp;	/* vnode for devid		*/
 	vdev_t		*vdev_top;	/* top-level vdev		*/
 	vdev_t		*vdev_parent;	/* parent vdev			*/
 	vdev_t		**vdev_child;	/* array of children		*/
diff --git a/include/sys/zfs_context.h b/include/sys/zfs_context.h
index f221091d8..c14d92a2c 100644
--- a/include/sys/zfs_context.h
+++ b/include/sys/zfs_context.h
@@ -42,7 +42,6 @@
 #include <sys/vmem.h>
 #include <sys/taskq.h>
 #include <sys/param.h>
-#include <sys/kobj.h>
 #include <sys/disp.h>
 #include <sys/debug.h>
 #include <sys/random.h>
@@ -515,16 +514,6 @@ extern void	system_taskq_fini(void);
 #define	XVA_MAPSIZE	3
 #define	XVA_MAGIC	0x78766174
 
-/*
- * vnodes
- */
-typedef struct vnode {
-	uint64_t	v_size;
-	int		v_fd;
-	char		*v_path;
-	int		v_dump_fd;
-} vnode_t;
-
 extern char *vn_dumpdir;
 #define	AV_SCANSTAMP_SZ	32		/* length of anti-virus scanstamp */
 
@@ -593,40 +582,7 @@ typedef struct vsecattr {
 #define	CRCREAT		0
 
 #define	F_FREESP	11
-
-extern int fop_getattr(vnode_t *vp, vattr_t *vap);
-
-#define	VOP_CLOSE(vp, f, c, o, cr, ct)	vn_close(vp)
-#define	VOP_PUTPAGE(vp, of, sz, fl, cr, ct)	0
-#define	VOP_GETATTR(vp, vap, fl, cr, ct)  fop_getattr((vp), (vap));
-
-#define	VOP_FSYNC(vp, f, cr, ct)	fsync((vp)->v_fd)
-
-#if defined(FALLOC_FL_PUNCH_HOLE) && defined(FALLOC_FL_KEEP_SIZE)
-#define	VOP_SPACE(vp, cmd, flck, fl, off, cr, ct) \
-	fallocate((vp)->v_fd, FALLOC_FL_PUNCH_HOLE | FALLOC_FL_KEEP_SIZE, \
-	    (flck)->l_start, (flck)->l_len)
-#else
-#define	VOP_SPACE(vp, cmd, flck, fl, off, cr, ct) (0)
-#endif
-
-#define	VN_RELE(vp)	vn_close(vp)
-
-extern int vn_open(char *path, int x1, int oflags, int mode, vnode_t **vpp,
-    int x2, int x3);
-extern int vn_openat(char *path, int x1, int oflags, int mode, vnode_t **vpp,
-    int x2, int x3, vnode_t *vp, int fd);
-extern int vn_rdwr(int uio, vnode_t *vp, void *addr, ssize_t len,
-    offset_t offset, int x1, int x2, rlim64_t x3, void *x4, ssize_t *residp);
-extern void vn_close(vnode_t *vp);
-
-#define	vn_remove(path, x1, x2)		remove(path)
-#define	vn_rename(from, to, seg)	rename((from), (to))
-#define	vn_is_readonly(vp)		B_FALSE
-
-extern vnode_t *rootdir;
-
-#include <sys/file.h>		/* for FREAD, FWRITE, etc */
+#define	FIGNORECASE	0x80000 /* request case-insensitive lookups */
 
 /*
  * Random stuff
@@ -681,7 +637,7 @@ extern int lowbit64(uint64_t i);
 extern int random_get_bytes(uint8_t *ptr, size_t len);
 extern int random_get_pseudo_bytes(uint8_t *ptr, size_t len);
 
-extern void kernel_init(int);
+extern void kernel_init(int mode);
 extern void kernel_fini(void);
 extern void random_init(void);
 extern void random_fini(void);
@@ -758,11 +714,6 @@ typedef struct ace_object {
 #define	ACE_SYSTEM_AUDIT_OBJECT_ACE_TYPE	0x07
 #define	ACE_SYSTEM_ALARM_OBJECT_ACE_TYPE	0x08
 
-extern struct _buf *kobj_open_file(char *name);
-extern int kobj_read_file(struct _buf *file, char *buf, unsigned size,
-    unsigned off);
-extern void kobj_close_file(struct _buf *file);
-extern int kobj_get_filesize(struct _buf *file, uint64_t *size);
 extern int zfs_secpolicy_snapshot_perms(const char *name, cred_t *cr);
 extern int zfs_secpolicy_rename_perms(const char *from, const char *to,
     cred_t *cr);
diff --git a/include/sys/zfs_file.h b/include/sys/zfs_file.h
new file mode 100644
index 000000000..6b33420e3
--- /dev/null
+++ b/include/sys/zfs_file.h
@@ -0,0 +1,61 @@
+/*
+ * CDDL HEADER START
+ *
+ * The contents of this file are subject to the terms of the
+ * Common Development and Distribution License (the "License").
+ * You may not use this file except in compliance with the License.
+ *
+ * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+ * or http://www.opensolaris.org/os/licensing.
+ * See the License for the specific language governing permissions
+ * and limitations under the License.
+ *
+ * When distributing Covered Code, include this CDDL HEADER in each
+ * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+ * If applicable, add the following below this CDDL HEADER, with the
+ * fields enclosed by brackets "[]" replaced with your own identifying
+ * information: Portions Copyright [yyyy] [name of copyright owner]
+ *
+ * CDDL HEADER END
+ */
+
+#ifndef	_SYS_ZFS_FILE_H
+#define	_SYS_ZFS_FILE_H
+
+#ifndef _KERNEL
+typedef struct zfs_file {
+	int f_fd;
+	int f_dump_fd;
+} zfs_file_t;
+#elif defined(__linux__) || defined(__FreeBSD__)
+typedef struct file zfs_file_t;
+#else
+#error "unknown OS"
+#endif
+
+typedef struct zfs_file_attr {
+	uint64_t	zfa_size;	/* file size */
+	mode_t		zfa_mode;	/* file type */
+} zfs_file_attr_t;
+
+int zfs_file_open(const char *path, int flags, int mode, zfs_file_t **fp);
+void zfs_file_close(zfs_file_t *fp);
+
+int zfs_file_write(zfs_file_t *fp, const void *buf, size_t len, ssize_t *resid);
+int zfs_file_pwrite(zfs_file_t *fp, const void *buf, size_t len, loff_t off,
+    ssize_t *resid);
+int zfs_file_read(zfs_file_t *fp, void *buf, size_t len, ssize_t *resid);
+int zfs_file_pread(zfs_file_t *fp, void *buf, size_t len, loff_t off,
+    ssize_t *resid);
+
+int zfs_file_seek(zfs_file_t *fp, loff_t *offp, int whence);
+int zfs_file_getattr(zfs_file_t *fp, zfs_file_attr_t *zfattr);
+int zfs_file_fsync(zfs_file_t *fp, int flags);
+int zfs_file_fallocate(zfs_file_t *fp, int mode, loff_t offset, loff_t len);
+loff_t zfs_file_off(zfs_file_t *fp);
+int zfs_file_unlink(const char *);
+
+int zfs_file_get(int fd, zfs_file_t **fp);
+void zfs_file_put(int fd);
+
+#endif /* _SYS_ZFS_FILE_H */
diff --git a/include/sys/zil.h b/include/sys/zil.h
index e84fb864a..716b19e56 100644
--- a/include/sys/zil.h
+++ b/include/sys/zil.h
@@ -373,7 +373,7 @@ typedef struct {
  *	- the write occupies only one block
  * WR_COPIED:
  *    If we know we'll immediately be committing the
- *    transaction (FSYNC or FDSYNC), then we allocate a larger
+ *    transaction (O_SYNC or O_DSYNC), then we allocate a larger
  *    log record here for the data and copy the data in.
  * WR_NEED_COPY:
  *    Otherwise we don't allocate a buffer, and *if* we need to