- Introduce a new struct xvfsconf, the userland version of struct vfsconf.

- Make getvfsbyname() take a struct xvfsconf *.
- Convert several consumers of getvfsbyname() to use struct xvfsconf.
- Correct the getvfsbyname.3 manpage.
- Create a new vfs.conflist sysctl to dump all the struct xvfsconf in the
  kernel, and rewrite getvfsbyname() to use this instead of the weird
  existing API.
- Convert some {set,get,end}vfsent() consumers to use the new vfs.conflist
  sysctl.
- Convert a vfsload() call in nfsiod.c to kldload() and remove the useless
  vfsisloadable() and endvfsent() calls.
- Add a warning printf() in vfs_sysctl() to tell people they are using
  an old userland.

After these changes, it's possible to modify struct vfsconf without
breaking the binary compatibility.  Please note that these changes don't
break this compatibility either.

When bp will have updated mount_smbfs(8) with the patch I sent him, there
will be no more consumers of the {set,get,end}vfsent(), vfsisloadable()
and vfsload() API, and I will promptly delete it.

1 files changed, 22 insertions, 18 deletions

diff --git a/lib/libc/gen/getvfsbyname.c b/lib/libc/gen/getvfsbyname.c
index eb6f26f..3ab03f4 100644
--- a/lib/libc/gen/getvfsbyname.c
+++ b/lib/libc/gen/getvfsbyname.c
@@ -41,39 +41,43 @@ __FBSDID("$FreeBSD$");
 #include <sys/mount.h>
 #include <sys/sysctl.h>
 #include <errno.h>
+#include <stdlib.h>
+#include <string.h>
 
 /*
  * Given a filesystem name, determine if it is resident in the kernel,
- * and if it is resident, return its vfsconf structure.
+ * and if it is resident, return its xvfsconf structure.
  */
+int
 getvfsbyname(fsname, vfcp)
 	const char *fsname;
-	struct vfsconf *vfcp;
+	struct xvfsconf *vfcp;
 {
 #ifdef	__NETBSD_SYSCALLS
 	errno = ENOSYS;
 #else
-	int name[4], maxtypenum, cnt;
+	struct xvfsconf *xvfsp;
 	size_t buflen;
+	int cnt, i;
 
-	name[0] = CTL_VFS;
-	name[1] = VFS_GENERIC;
-	name[2] = VFS_MAXTYPENUM;
-	buflen = 4;
-	if (sysctl(name, 3, &maxtypenum, &buflen, (void *)0, (size_t)0) < 0)
+	if (sysctlbyname("vfs.conflist", NULL, &buflen, NULL, 0) < 0)
 		return (-1);
-	name[2] = VFS_CONF;
-	buflen = sizeof *vfcp;
-	for (cnt = 0; cnt < maxtypenum; cnt++) {
-		name[3] = cnt;
-		if (sysctl(name, 4, vfcp, &buflen, (void *)0, (size_t)0) < 0) {
-			if (errno != EOPNOTSUPP)
-				return (-1);
-			continue;
-		}
-		if (!strcmp(fsname, vfcp->vfc_name))
+	xvfsp = malloc(buflen);
+	if (xvfsp == NULL)
+		return (-1);
+	if (sysctlbyname("vfs.conflist", xvfsp, &buflen, NULL, 0) < 0) {
+		free(xvfsp);
+		return (-1);
+	}
+	cnt = buflen / sizeof(struct xvfsconf);
+	for (i = 0; i < cnt; i++) {
+		if (strcmp(fsname, xvfsp[i].vfc_name) == 0) {
+			memcpy(vfcp, xvfsp + i, sizeof(struct xvfsconf));
+			free(xvfsp);
 			return (0);
+		}
 	}
+	free(xvfsp);
 	errno = ENOENT;
 #endif
 	return (-1);