Add the initial vestigates of vnode support

git-svn-id: https://outreach.scidac.gov/svn/spl/trunk@30 7e1ea52c-4ff2-0310-8f11-9dd32ca42a1c

4 files changed, 267 insertions, 0 deletions

diff --git a/include/sys/sysmacros.h b/include/sys/sysmacros.h
index 3bc9f7a37..b65a5797c 100644
--- a/include/sys/sysmacros.h
+++ b/include/sys/sysmacros.h
@@ -131,6 +131,7 @@ extern int highbit(unsigned long i);
 #define makedevice(maj,min) makedev(maj,min)
 #define zone_dataset_visible(x, y)			(1)
 #define INGLOBALZONE(z)					(1)
+#define utsname						system_utsname
 
 /* XXX - Borrowed from zfs project libsolcompat/include/sys/sysmacros.h */
 /* common macros */
diff --git a/include/sys/vnode.h b/include/sys/vnode.h
new file mode 100644
index 000000000..9afac4c25
--- /dev/null
+++ b/include/sys/vnode.h
@@ -0,0 +1,97 @@
+#ifndef _SPL_VNODE_H
+#define _SPL_VNODE_H
+
+#define XVA_MAPSIZE     3
+#define XVA_MAGIC       0x78766174
+
+typedef struct vnode {
+        uint64_t        v_size;
+        int             v_fd;
+        mode_t          v_mode;
+        char            *v_path;
+} vnode_t;
+
+
+typedef struct xoptattr {
+        timestruc_t     xoa_createtime; /* Create time of file */
+        uint8_t         xoa_archive;
+        uint8_t         xoa_system;
+        uint8_t         xoa_readonly;
+        uint8_t         xoa_hidden;
+        uint8_t         xoa_nounlink;
+        uint8_t         xoa_immutable;
+        uint8_t         xoa_appendonly;
+        uint8_t         xoa_nodump;
+        uint8_t         xoa_settable;
+        uint8_t         xoa_opaque;
+        uint8_t         xoa_av_quarantined;
+        uint8_t         xoa_av_modified;
+} xoptattr_t;
+
+typedef struct vattr {
+        uint_t          va_mask;        /* bit-mask of attributes */
+        u_offset_t      va_size;        /* file size in bytes */
+} vattr_t;
+
+
+typedef struct xvattr {
+        vattr_t         xva_vattr;      /* Embedded vattr structure */
+        uint32_t        xva_magic;      /* Magic Number */
+        uint32_t        xva_mapsize;    /* Size of attr bitmap (32-bit words) */
+        uint32_t        *xva_rtnattrmapp;       /* Ptr to xva_rtnattrmap[] */
+        uint32_t        xva_reqattrmap[XVA_MAPSIZE];    /* Requested attrs */
+        uint32_t        xva_rtnattrmap[XVA_MAPSIZE];    /* Returned attrs */
+        xoptattr_t      xva_xoptattrs;  /* Optional attributes */
+} xvattr_t;
+
+typedef struct vsecattr {
+        uint_t          vsa_mask;       /* See below */
+        int             vsa_aclcnt;     /* ACL entry count */
+        void            *vsa_aclentp;   /* pointer to ACL entries */
+        int             vsa_dfaclcnt;   /* default ACL entry count */
+        void            *vsa_dfaclentp; /* pointer to default ACL entries */
+        size_t          vsa_aclentsz;   /* ACE size in bytes of vsa_aclentp */
+} vsecattr_t;
+
+#define AT_TYPE         0x00001
+#define AT_MODE         0x00002
+// #define AT_UID          0x00004	/* Conflicts with linux/auxvec.h */
+// #define AT_GID          0x00008	/* Conflicts with linux/auxvec.h */
+#define AT_FSID         0x00010
+#define AT_NODEID       0x00020
+#define AT_NLINK        0x00040
+#define AT_SIZE         0x00080
+#define AT_ATIME        0x00100
+#define AT_MTIME        0x00200
+#define AT_CTIME        0x00400
+#define AT_RDEV         0x00800
+#define AT_BLKSIZE      0x01000
+#define AT_NBLOCKS      0x02000
+#define AT_SEQ          0x08000
+#define AT_XVATTR       0x10000
+
+#define CRCREAT         0
+
+#define VOP_CLOSE(vp, f, c, o, cr, ct)  0
+#define VOP_PUTPAGE(vp, of, sz, fl, cr, ct)     0
+#define VOP_GETATTR(vp, vap, fl, cr, ct)  ((vap)->va_size = (vp)->v_size, 0)
+
+#define VOP_FSYNC(vp, f, cr, ct)        fsync((vp)->v_fd)
+
+#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;
+
+#endif /* SPL_VNODE_H */
diff --git a/modules/spl/Makefile.in b/modules/spl/Makefile.in
index 09e934e4f..667858aa9 100644
--- a/modules/spl/Makefile.in
+++ b/modules/spl/Makefile.in
@@ -13,6 +13,7 @@ spl-objs += spl-kmem.o
 spl-objs += spl-thread.o
 spl-objs += spl-taskq.o
 spl-objs += spl-rwlock.o
+spl-objs += spl-vnode.o
 spl-objs += spl-generic.o
 
 splmodule := spl.ko
diff --git a/modules/spl/spl-vnode.c b/modules/spl/spl-vnode.c
new file mode 100644
index 000000000..5089f8567
--- /dev/null
+++ b/modules/spl/spl-vnode.c
@@ -0,0 +1,168 @@
+#include <sys/sysmacros.h>
+#include "config.h"
+
+/*
+ * XXX: currently borrrowed from libsolcompat until this
+ * can be adapted to the linux kernel interfaces.
+ */
+#if 0
+/*
+ * =========================================================================
+ * vnode operations
+ * =========================================================================
+ */
+/*
+ * Note: for the xxxat() versions of these functions, we assume that the
+ * starting vp is always rootdir (which is true for spa_directory.c, the only
+ * ZFS consumer of these interfaces).  We assert this is true, and then emulate
+ * them by adding '/' in front of the path.
+ */
+
+/*ARGSUSED*/
+int
+vn_open(char *path, int x1, int flags, int mode, vnode_t **vpp, int x2, int x3)
+{
+	int fd;
+	vnode_t *vp;
+	int old_umask;
+	char realpath[MAXPATHLEN];
+	struct stat64 st;
+
+	/*
+	 * If we're accessing a real disk from userland, we need to use
+	 * the character interface to avoid caching.  This is particularly
+	 * important if we're trying to look at a real in-kernel storage
+	 * pool from userland, e.g. via zdb, because otherwise we won't
+	 * see the changes occurring under the segmap cache.
+	 * On the other hand, the stupid character device returns zero
+	 * for its size.  So -- gag -- we open the block device to get
+	 * its size, and remember it for subsequent VOP_GETATTR().
+	 */
+#if defined(__sun__) || defined(__sun)
+	if (strncmp(path, "/dev/", 5) == 0) {
+#else
+	if (0) {
+#endif
+		char *dsk;
+		fd = open64(path, O_RDONLY);
+		if (fd == -1)
+			return (errno);
+		if (fstat64(fd, &st) == -1) {
+			close(fd);
+			return (errno);
+		}
+		close(fd);
+		(void) sprintf(realpath, "%s", path);
+		dsk = strstr(path, "/dsk/");
+		if (dsk != NULL)
+			(void) sprintf(realpath + (dsk - path) + 1, "r%s",
+			    dsk + 1);
+	} else {
+		(void) sprintf(realpath, "%s", path);
+		if (!(flags & FCREAT) && stat64(realpath, &st) == -1)
+			return (errno);
+	}
+
+#ifdef __linux__
+	if (!(flags & FCREAT) && S_ISBLK(st.st_mode)) {
+		flags |= O_DIRECT;
+		if (flags & FWRITE)
+			flags |= O_EXCL;
+	}
+#endif
+
+	if (flags & FCREAT)
+		old_umask = umask(0);
+
+	/*
+	 * The construct 'flags - FREAD' conveniently maps combinations of
+	 * FREAD and FWRITE to the corresponding O_RDONLY, O_WRONLY, and O_RDWR.
+	 */
+	fd = open64(realpath, flags - FREAD, mode);
+
+	if (flags & FCREAT)
+		(void) umask(old_umask);
+
+	if (fd == -1)
+		return (errno);
+
+	if (fstat64(fd, &st) == -1) {
+		close(fd);
+		return (errno);
+	}
+
+	(void) fcntl(fd, F_SETFD, FD_CLOEXEC);
+
+	*vpp = vp = umem_zalloc(sizeof (vnode_t), UMEM_NOFAIL);
+
+	vp->v_fd = fd;
+	vp->v_size = st.st_size;
+	vp->v_mode = st.st_mode;
+	vp->v_path = spa_strdup(path);
+
+	return (0);
+}
+
+/*ARGSUSED*/
+int
+vn_openat(char *path, int x1, int flags, int mode, vnode_t **vpp, int x2,
+    int x3, vnode_t *startvp, int fd)
+{
+	char *realpath = umem_alloc(strlen(path) + 2, UMEM_NOFAIL);
+	int ret;
+
+	ASSERT(startvp == rootdir);
+	(void) sprintf(realpath, "/%s", path);
+
+	/* fd ignored for now, need if want to simulate nbmand support */
+	ret = vn_open(realpath, x1, flags, mode, vpp, x2, x3);
+
+	umem_free(realpath, strlen(path) + 2);
+
+	return (ret);
+}
+
+/*ARGSUSED*/
+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)
+{
+	ssize_t iolen, split;
+
+	if (uio == UIO_READ) {
+		iolen = pread64(vp->v_fd, addr, len, offset);
+	} else {
+		/*
+		 * To simulate partial disk writes, we split writes into two
+		 * system calls so that the process can be killed in between.
+		 */
+#ifdef ZFS_DEBUG
+		if (!S_ISBLK(vp->v_mode) && !S_ISCHR(vp->v_mode)) {
+			split = (len > 0 ? rand() % len : 0);
+			iolen = pwrite64(vp->v_fd, addr, split, offset);
+			iolen += pwrite64(vp->v_fd, (char *)addr + split,
+			    len - split, offset + split);
+		} else
+			iolen = pwrite64(vp->v_fd, addr, len, offset);
+#else
+		iolen = pwrite64(vp->v_fd, addr, len, offset);
+#endif
+	}
+
+	if (iolen < 0)
+		return (errno);
+	if (residp)
+		*residp = len - iolen;
+	else if (iolen != len)
+		return (EIO);
+	return (0);
+}
+
+void
+vn_close(vnode_t *vp)
+{
+	close(vp->v_fd);
+	spa_strfree(vp->v_path);
+	umem_free(vp, sizeof (vnode_t));
+}
+#endif