Merge Capsicum overhaul:

- Capability is no longer separate descriptor type. Now every descriptor
  has set of its own capability rights.

- The cap_new(2) system call is left, but it is no longer documented and
  should not be used in new code.

- The new syscall cap_rights_limit(2) should be used instead of
  cap_new(2), which limits capability rights of the given descriptor
  without creating a new one.

- The cap_getrights(2) syscall is renamed to cap_rights_get(2).

- If CAP_IOCTL capability right is present we can further reduce allowed
  ioctls list with the new cap_ioctls_limit(2) syscall. List of allowed
  ioctls can be retrived with cap_ioctls_get(2) syscall.

- If CAP_FCNTL capability right is present we can further reduce fcntls
  that can be used with the new cap_fcntls_limit(2) syscall and retrive
  them with cap_fcntls_get(2).

- To support ioctl and fcntl white-listing the filedesc structure was
  heavly modified.

- The audit subsystem, kdump and procstat tools were updated to
  recognize new syscalls.

- Capability rights were revised and eventhough I tried hard to provide
  backward API and ABI compatibility there are some incompatible changes
  that are described in detail below:

	CAP_CREATE old behaviour:
	- Allow for openat(2)+O_CREAT.
	- Allow for linkat(2).
	- Allow for symlinkat(2).
	CAP_CREATE new behaviour:
	- Allow for openat(2)+O_CREAT.

	Added CAP_LINKAT:
	- Allow for linkat(2). ABI: Reuses CAP_RMDIR bit.
	- Allow to be target for renameat(2).

	Added CAP_SYMLINKAT:
	- Allow for symlinkat(2).

	Removed CAP_DELETE. Old behaviour:
	- Allow for unlinkat(2) when removing non-directory object.
	- Allow to be source for renameat(2).

	Removed CAP_RMDIR. Old behaviour:
	- Allow for unlinkat(2) when removing directory.

	Added CAP_RENAMEAT:
	- Required for source directory for the renameat(2) syscall.

	Added CAP_UNLINKAT (effectively it replaces CAP_DELETE and CAP_RMDIR):
	- Allow for unlinkat(2) on any object.
	- Required if target of renameat(2) exists and will be removed by this
	  call.

	Removed CAP_MAPEXEC.

	CAP_MMAP old behaviour:
	- Allow for mmap(2) with any combination of PROT_NONE, PROT_READ and
	  PROT_WRITE.
	CAP_MMAP new behaviour:
	- Allow for mmap(2)+PROT_NONE.

	Added CAP_MMAP_R:
	- Allow for mmap(PROT_READ).
	Added CAP_MMAP_W:
	- Allow for mmap(PROT_WRITE).
	Added CAP_MMAP_X:
	- Allow for mmap(PROT_EXEC).
	Added CAP_MMAP_RW:
	- Allow for mmap(PROT_READ | PROT_WRITE).
	Added CAP_MMAP_RX:
	- Allow for mmap(PROT_READ | PROT_EXEC).
	Added CAP_MMAP_WX:
	- Allow for mmap(PROT_WRITE | PROT_EXEC).
	Added CAP_MMAP_RWX:
	- Allow for mmap(PROT_READ | PROT_WRITE | PROT_EXEC).

	Renamed CAP_MKDIR to CAP_MKDIRAT.
	Renamed CAP_MKFIFO to CAP_MKFIFOAT.
	Renamed CAP_MKNODE to CAP_MKNODEAT.

	CAP_READ old behaviour:
	- Allow pread(2).
	- Disallow read(2), readv(2) (if there is no CAP_SEEK).
	CAP_READ new behaviour:
	- Allow read(2), readv(2).
	- Disallow pread(2) (CAP_SEEK was also required).

	CAP_WRITE old behaviour:
	- Allow pwrite(2).
	- Disallow write(2), writev(2) (if there is no CAP_SEEK).
	CAP_WRITE new behaviour:
	- Allow write(2), writev(2).
	- Disallow pwrite(2) (CAP_SEEK was also required).

	Added convinient defines:

	#define	CAP_PREAD		(CAP_SEEK | CAP_READ)
	#define	CAP_PWRITE		(CAP_SEEK | CAP_WRITE)
	#define	CAP_MMAP_R		(CAP_MMAP | CAP_SEEK | CAP_READ)
	#define	CAP_MMAP_W		(CAP_MMAP | CAP_SEEK | CAP_WRITE)
	#define	CAP_MMAP_X		(CAP_MMAP | CAP_SEEK | 0x0000000000000008ULL)
	#define	CAP_MMAP_RW		(CAP_MMAP_R | CAP_MMAP_W)
	#define	CAP_MMAP_RX		(CAP_MMAP_R | CAP_MMAP_X)
	#define	CAP_MMAP_WX		(CAP_MMAP_W | CAP_MMAP_X)
	#define	CAP_MMAP_RWX		(CAP_MMAP_R | CAP_MMAP_W | CAP_MMAP_X)
	#define	CAP_RECV		CAP_READ
	#define	CAP_SEND		CAP_WRITE

	#define	CAP_SOCK_CLIENT \
		(CAP_CONNECT | CAP_GETPEERNAME | CAP_GETSOCKNAME | CAP_GETSOCKOPT | \
		 CAP_PEELOFF | CAP_RECV | CAP_SEND | CAP_SETSOCKOPT | CAP_SHUTDOWN)
	#define	CAP_SOCK_SERVER \
		(CAP_ACCEPT | CAP_BIND | CAP_GETPEERNAME | CAP_GETSOCKNAME | \
		 CAP_GETSOCKOPT | CAP_LISTEN | CAP_PEELOFF | CAP_RECV | CAP_SEND | \
		 CAP_SETSOCKOPT | CAP_SHUTDOWN)

	Added defines for backward API compatibility:

	#define	CAP_MAPEXEC		CAP_MMAP_X
	#define	CAP_DELETE		CAP_UNLINKAT
	#define	CAP_MKDIR		CAP_MKDIRAT
	#define	CAP_RMDIR		CAP_UNLINKAT
	#define	CAP_MKFIFO		CAP_MKFIFOAT
	#define	CAP_MKNOD		CAP_MKNODAT
	#define	CAP_SOCK_ALL		(CAP_SOCK_CLIENT | CAP_SOCK_SERVER)

Sponsored by:	The FreeBSD Foundation
Reviewed by:	Christoph Mallon <christoph.mallon@gmx.de>
Many aspects discussed with:	rwatson, benl, jonathan
ABI compatibility discussed with:	kib

1 files changed, 41 insertions, 47 deletions

diff --git a/sys/kern/vfs_syscalls.c b/sys/kern/vfs_syscalls.c
index bd44a3a..787399a 100644
--- a/sys/kern/vfs_syscalls.c
+++ b/sys/kern/vfs_syscalls.c
@@ -970,6 +970,8 @@ flags_to_rights(int flags)
 			/* FALLTHROUGH */
 		case O_WRONLY:
 			rights |= CAP_WRITE;
+			if (!(flags & O_APPEND))
+				rights |= CAP_SEEK;
 			break;
 		}
 	}
@@ -1143,19 +1145,22 @@ success:
 	 * If we haven't already installed the FD (for dupfdopen), do so now.
 	 */
 	if (indx == -1) {
+		struct filecaps *fcaps;
+
 #ifdef CAPABILITIES
-		if (nd.ni_strictrelative == 1) {
-			/*
-			 * We are doing a strict relative lookup; wrap the
-			 * result in a capability.
-			 */
-			if ((error = kern_capwrap(td, fp, nd.ni_baserights,
-			    &indx)) != 0)
-				goto bad;
-		} else
+		if (nd.ni_strictrelative == 1)
+			fcaps = &nd.ni_filecaps;
+		else
 #endif
-			if ((error = finstall(td, fp, &indx, flags)) != 0)
-				goto bad;
+			fcaps = NULL;
+		error = finstall(td, fp, &indx, flags, fcaps);
+		/* On success finstall() consumes fcaps. */
+		if (error != 0) {
+			filecaps_free(&nd.ni_filecaps);
+			goto bad;
+		}
+	} else {
+		filecaps_free(&nd.ni_filecaps);
 	}
 
 	/*
@@ -1279,7 +1284,7 @@ kern_mknodat(struct thread *td, int fd, char *path, enum uio_seg pathseg,
 restart:
 	bwillwrite();
 	NDINIT_ATRIGHTS(&nd, CREATE, LOCKPARENT | SAVENAME | AUDITVNODE1,
-	    pathseg, path, fd, CAP_MKNOD, td);
+	    pathseg, path, fd, CAP_MKNODAT, td);
 	if ((error = namei(&nd)) != 0)
 		return (error);
 	vp = nd.ni_vp;
@@ -1399,7 +1404,7 @@ kern_mkfifoat(struct thread *td, int fd, char *path, enum uio_seg pathseg,
 restart:
 	bwillwrite();
 	NDINIT_ATRIGHTS(&nd, CREATE, LOCKPARENT | SAVENAME | AUDITVNODE1,
-	    pathseg, path, fd, CAP_MKFIFO, td);
+	    pathseg, path, fd, CAP_MKFIFOAT, td);
 	if ((error = namei(&nd)) != 0)
 		return (error);
 	if (nd.ni_vp != NULL) {
@@ -1553,7 +1558,7 @@ kern_linkat(struct thread *td, int fd1, int fd2, char *path1, char *path2,
 		return (error);
 	}
 	NDINIT_ATRIGHTS(&nd, CREATE, LOCKPARENT | SAVENAME | AUDITVNODE2,
-	    segflg, path2, fd2, CAP_CREATE, td);
+	    segflg, path2, fd2, CAP_LINKAT, td);
 	if ((error = namei(&nd)) == 0) {
 		if (nd.ni_vp != NULL) {
 			if (nd.ni_dvp == nd.ni_vp)
@@ -1646,7 +1651,7 @@ kern_symlinkat(struct thread *td, char *path1, int fd, char *path2,
 restart:
 	bwillwrite();
 	NDINIT_ATRIGHTS(&nd, CREATE, LOCKPARENT | SAVENAME | AUDITVNODE1,
-	    segflg, path2, fd, CAP_CREATE, td);
+	    segflg, path2, fd, CAP_SYMLINKAT, td);
 	if ((error = namei(&nd)) != 0)
 		goto out;
 	if (nd.ni_vp) {
@@ -1798,7 +1803,7 @@ kern_unlinkat(struct thread *td, int fd, char *path, enum uio_seg pathseg,
 restart:
 	bwillwrite();
 	NDINIT_ATRIGHTS(&nd, DELETE, LOCKPARENT | LOCKLEAF | AUDITVNODE1,
-	    pathseg, path, fd, CAP_DELETE, td);
+	    pathseg, path, fd, CAP_UNLINKAT, td);
 	if ((error = namei(&nd)) != 0)
 		return (error == EINVAL ? EPERM : error);
 	vp = nd.ni_vp;
@@ -3502,10 +3507,10 @@ kern_renameat(struct thread *td, int oldfd, char *old, int newfd, char *new,
 	bwillwrite();
 #ifdef MAC
 	NDINIT_ATRIGHTS(&fromnd, DELETE, LOCKPARENT | LOCKLEAF | SAVESTART |
-	    AUDITVNODE1, pathseg, old, oldfd, CAP_DELETE, td);
+	    AUDITVNODE1, pathseg, old, oldfd, CAP_RENAMEAT, td);
 #else
 	NDINIT_ATRIGHTS(&fromnd, DELETE, WANTPARENT | SAVESTART | AUDITVNODE1,
-	    pathseg, old, oldfd, CAP_DELETE, td);
+	    pathseg, old, oldfd, CAP_RENAMEAT, td);
 #endif
 
 	if ((error = namei(&fromnd)) != 0)
@@ -3527,7 +3532,7 @@ kern_renameat(struct thread *td, int oldfd, char *old, int newfd, char *new,
 		goto out1;
 	}
 	NDINIT_ATRIGHTS(&tond, RENAME, LOCKPARENT | LOCKLEAF | NOCACHE |
-	    SAVESTART | AUDITVNODE2, pathseg, new, newfd, CAP_CREATE, td);
+	    SAVESTART | AUDITVNODE2, pathseg, new, newfd, CAP_LINKAT, td);
 	if (fromnd.ni_vp->v_type == VDIR)
 		tond.ni_cnd.cn_flags |= WILLBEDIR;
 	if ((error = namei(&tond)) != 0) {
@@ -3550,6 +3555,15 @@ kern_renameat(struct thread *td, int oldfd, char *old, int newfd, char *new,
 			error = EISDIR;
 			goto out;
 		}
+#ifdef CAPABILITIES
+		/*
+		 * If the target already exists we require CAP_UNLINKAT
+		 * from 'newfd'.
+		 */
+		error = cap_check(tond.ni_filecaps.fc_rights, CAP_UNLINKAT);
+		if (error != 0)
+			goto out;
+#endif
 	}
 	if (fvp == tdvp) {
 		error = EINVAL;
@@ -3650,7 +3664,7 @@ kern_mkdirat(struct thread *td, int fd, char *path, enum uio_seg segflg,
 restart:
 	bwillwrite();
 	NDINIT_ATRIGHTS(&nd, CREATE, LOCKPARENT | SAVENAME | AUDITVNODE1,
-	    segflg, path, fd, CAP_MKDIR, td);
+	    segflg, path, fd, CAP_MKDIRAT, td);
 	nd.ni_cnd.cn_flags |= WILLBEDIR;
 	if ((error = namei(&nd)) != 0)
 		return (error);
@@ -3734,7 +3748,7 @@ kern_rmdirat(struct thread *td, int fd, char *path, enum uio_seg pathseg)
 restart:
 	bwillwrite();
 	NDINIT_ATRIGHTS(&nd, DELETE, LOCKPARENT | LOCKLEAF | AUDITVNODE1,
-	    pathseg, path, fd, CAP_RMDIR, td);
+	    pathseg, path, fd, CAP_UNLINKAT, td);
 	if ((error = namei(&nd)) != 0)
 		return (error);
 	vp = nd.ni_vp;
@@ -3987,8 +4001,7 @@ kern_getdirentries(struct thread *td, int fd, char *buf, u_int count,
 	if (count > IOSIZE_MAX)
 		return (EINVAL);
 	auio.uio_resid = count;
-	if ((error = getvnode(td->td_proc->p_fd, fd, CAP_READ | CAP_SEEK,
-	    &fp)) != 0)
+	if ((error = getvnode(td->td_proc->p_fd, fd, CAP_READ, &fp)) != 0)
 		return (error);
 	if ((fp->f_flag & FREAD) == 0) {
 		fdrop(fp, td);
@@ -4151,33 +4164,14 @@ out:
  * entry is held upon returning.
  */
 int
-getvnode(struct filedesc *fdp, int fd, cap_rights_t rights,
-    struct file **fpp)
+getvnode(struct filedesc *fdp, int fd, cap_rights_t rights, struct file **fpp)
 {
 	struct file *fp;
-#ifdef CAPABILITIES
-	struct file *fp_fromcap;
 	int error;
-#endif
 
-	if (fdp == NULL || (fp = fget_unlocked(fdp, fd)) == NULL)
-		return (EBADF);
-#ifdef CAPABILITIES
-	/*
-	 * If the file descriptor is for a capability, test rights and use the
-	 * file descriptor referenced by the capability.
-	 */
-	error = cap_funwrap(fp, rights, &fp_fromcap);
-	if (error) {
-		fdrop(fp, curthread);
+	error = fget_unlocked(fdp, fd, rights, 0, &fp, NULL);
+	if (error != 0)
 		return (error);
-	}
-	if (fp != fp_fromcap) {
-		fhold(fp_fromcap);
-		fdrop(fp, curthread);
-		fp = fp_fromcap;
-	}
-#endif /* CAPABILITIES */
 
 	/*
 	 * The file could be not of the vnode type, or it may be not
@@ -4361,7 +4355,7 @@ sys_fhopen(td, uap)
 			goto bad;
 	}
 
-	error = finstall(td, fp, &indx, fmode);
+	error = finstall(td, fp, &indx, fmode, NULL);
 bad:
 	fdrop(fp, td);
 	td->td_retval[0] = indx;
@@ -4614,7 +4608,7 @@ kern_posix_fadvise(struct thread *td, int fd, off_t offset, off_t len,
 		return (EINVAL);
 	}
 	/* XXX: CAP_POSIX_FADVISE? */
-	error = fget(td, fd, 0, &fp);
+	error = fget(td, fd, CAP_NONE, &fp);
 	if (error != 0)
 		goto out;