Change the cap_rights_t type from uint64_t to a structure that we can extend

in the future in a backward compatible (API and ABI) way.

The cap_rights_t represents capability rights. We used to use one bit to
represent one right, but we are running out of spare bits. Currently the new
structure provides place for 114 rights (so 50 more than the previous
cap_rights_t), but it is possible to grow the structure to hold at least 285
rights, although we can make it even larger if 285 rights won't be enough.

The structure definition looks like this:

	struct cap_rights {
		uint64_t	cr_rights[CAP_RIGHTS_VERSION + 2];
	};

The initial CAP_RIGHTS_VERSION is 0.

The top two bits in the first element of the cr_rights[] array contain total
number of elements in the array - 2. This means if those two bits are equal to
0, we have 2 array elements.

The top two bits in all remaining array elements should be 0.
The next five bits in all array elements contain array index. Only one bit is
used and bit position in this five-bits range defines array index. This means
there can be at most five array elements in the future.

To define new right the CAPRIGHT() macro must be used. The macro takes two
arguments - an array index and a bit to set, eg.

	#define	CAP_PDKILL	CAPRIGHT(1, 0x0000000000000800ULL)

We still support aliases that combine few rights, but the rights have to belong
to the same array element, eg:

	#define	CAP_LOOKUP	CAPRIGHT(0, 0x0000000000000400ULL)
	#define	CAP_FCHMOD	CAPRIGHT(0, 0x0000000000002000ULL)

	#define	CAP_FCHMODAT	(CAP_FCHMOD | CAP_LOOKUP)

There is new API to manage the new cap_rights_t structure:

	cap_rights_t *cap_rights_init(cap_rights_t *rights, ...);
	void cap_rights_set(cap_rights_t *rights, ...);
	void cap_rights_clear(cap_rights_t *rights, ...);
	bool cap_rights_is_set(const cap_rights_t *rights, ...);

	bool cap_rights_is_valid(const cap_rights_t *rights);
	void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src);
	void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src);
	bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little);

Capability rights to the cap_rights_init(), cap_rights_set(),
cap_rights_clear() and cap_rights_is_set() functions are provided by
separating them with commas, eg:

	cap_rights_t rights;

	cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT);

There is no need to terminate the list of rights, as those functions are
actually macros that take care of the termination, eg:

	#define	cap_rights_set(rights, ...)				\
		__cap_rights_set((rights), __VA_ARGS__, 0ULL)
	void __cap_rights_set(cap_rights_t *rights, ...);

Thanks to using one bit as an array index we can assert in those functions that
there are no two rights belonging to different array elements provided
together. For example this is illegal and will be detected, because CAP_LOOKUP
belongs to element 0 and CAP_PDKILL to element 1:

	cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL);

Providing several rights that belongs to the same array's element this way is
correct, but is not advised. It should only be used for aliases definition.

This commit also breaks compatibility with some existing Capsicum system calls,
but I see no other way to do that. This should be fine as Capsicum is still
experimental and this change is not going to 9.x.

Sponsored by:	The FreeBSD Foundation

8 files changed, 32 insertions, 15 deletions

diff --git a/sys/dev/aac/aac_linux.c b/sys/dev/aac/aac_linux.c
index 049e2be..591dfbb 100644
--- a/sys/dev/aac/aac_linux.c
+++ b/sys/dev/aac/aac_linux.c
@@ -75,11 +75,13 @@ MODULE_DEPEND(aac_linux, linux, 1, 1, 1);
 static int
 aac_linux_ioctl(struct thread *td, struct linux_ioctl_args *args)
 {
+	cap_rights_t rights;
 	struct file *fp;
 	u_long cmd;
 	int error;
 
-	if ((error = fget(td, args->fd, CAP_IOCTL, &fp)) != 0)
+	error = fget(td, args->fd, cap_rights_init(&rights, CAP_IOCTL), &fp);
+	if (error != 0)
 		return (error);
 	cmd = args->cmd;
 
diff --git a/sys/dev/amr/amr_linux.c b/sys/dev/amr/amr_linux.c
index 44e858b..5b1a17f 100644
--- a/sys/dev/amr/amr_linux.c
+++ b/sys/dev/amr/amr_linux.c
@@ -72,10 +72,12 @@ MODULE_DEPEND(amr, linux, 1, 1, 1);
 static int
 amr_linux_ioctl(struct thread *p, struct linux_ioctl_args *args)
 {
+	cap_rights_t rights;
 	struct file *fp;
 	int error;
 
-	if ((error = fget(p, args->fd, CAP_IOCTL, &fp)) != 0)
+	error = fget(p, args->fd, cap_rights_init(&rights, CAP_IOCTL), &fp);
+	if (error != 0)
 		return (error);
 	error = fo_ioctl(fp, args->cmd, (caddr_t)args->arg, p->td_ucred, p);
 	fdrop(fp, p);
diff --git a/sys/dev/filemon/filemon.c b/sys/dev/filemon/filemon.c
index ce84e3d..e3fda18 100644
--- a/sys/dev/filemon/filemon.c
+++ b/sys/dev/filemon/filemon.c
@@ -138,12 +138,6 @@ filemon_dtr(void *data)
 	}
 }
 
-#if __FreeBSD_version < 900041
-#define FGET_WRITE(a1, a2, a3) fget_write((a1), (a2), (a3))
-#else
-#define FGET_WRITE(a1, a2, a3) fget_write((a1), (a2), CAP_WRITE | CAP_SEEK, (a3))
-#endif
-
 static int
 filemon_ioctl(struct cdev *dev, u_long cmd, caddr_t data, int flag __unused,
     struct thread *td)
@@ -151,13 +145,21 @@ filemon_ioctl(struct cdev *dev, u_long cmd, caddr_t data, int flag __unused,
 	int error = 0;
 	struct filemon *filemon;
 	struct proc *p;
+#if __FreeBSD_version >= 900041
+	cap_rights_t rights;
+#endif
 
 	devfs_get_cdevpriv((void **) &filemon);
 
 	switch (cmd) {
 	/* Set the output file descriptor. */
 	case FILEMON_SET_FD:
-		if ((error = FGET_WRITE(td, *(int *)data, &filemon->fp)) == 0)
+		error = fget_write(td, *(int *)data,
+#if __FreeBSD_version >= 900041
+		    cap_rights_init(&rights, CAP_PWRITE),
+#endif
+		    &filemon->fp);
+		if (error == 0)
 			/* Write the file header. */
 			filemon_comment(filemon);
 		break;
diff --git a/sys/dev/hwpmc/hwpmc_logging.c b/sys/dev/hwpmc/hwpmc_logging.c
index 880bcaa..a60e096 100644
--- a/sys/dev/hwpmc/hwpmc_logging.c
+++ b/sys/dev/hwpmc/hwpmc_logging.c
@@ -570,6 +570,7 @@ pmclog_configure_log(struct pmc_mdep *md, struct pmc_owner *po, int logfd)
 {
 	int error;
 	struct proc *p;
+	cap_rights_t rights;
 
 	/*
 	 * As long as it is possible to get a LOR between pmc_sx lock and
@@ -593,7 +594,8 @@ pmclog_configure_log(struct pmc_mdep *md, struct pmc_owner *po, int logfd)
 		po->po_file));
 
 	/* get a reference to the file state */
-	error = fget_write(curthread, logfd, CAP_WRITE, &po->po_file);
+	error = fget_write(curthread, logfd,
+	    cap_rights_init(&rights, CAP_WRITE), &po->po_file);
 	if (error)
 		goto error;
 
diff --git a/sys/dev/ipmi/ipmi_linux.c b/sys/dev/ipmi/ipmi_linux.c
index 430bd08..b6b38f2 100644
--- a/sys/dev/ipmi/ipmi_linux.c
+++ b/sys/dev/ipmi/ipmi_linux.c
@@ -89,11 +89,13 @@ MODULE_DEPEND(ipmi_linux, linux, 1, 1, 1);
 static int
 ipmi_linux_ioctl(struct thread *td, struct linux_ioctl_args *args)
 {
+	cap_rights_t rights;
 	struct file *fp;
 	u_long cmd;
 	int error;
 
-	if ((error = fget(td, args->fd, CAP_IOCTL, &fp)) != 0)
+	error = fget(td, args->fd, cap_rights_init(&rights, CAP_IOCTL), &fp);
+	if (error != 0)
 		return (error);
 	cmd = args->cmd;
 
diff --git a/sys/dev/iscsi_initiator/iscsi.c b/sys/dev/iscsi_initiator/iscsi.c
index 4dbf163..4a1cb96 100644
--- a/sys/dev/iscsi_initiator/iscsi.c
+++ b/sys/dev/iscsi_initiator/iscsi.c
@@ -382,16 +382,19 @@ i_ping(struct cdev *dev)
 static int
 i_setsoc(isc_session_t *sp, int fd, struct thread *td)
 {
+     cap_rights_t rights;
      int error = 0;
 
      if(sp->soc != NULL)
 	  isc_stop_receiver(sp);
 
-     error = fget(td, fd, CAP_SOCK_CLIENT, &sp->fp);
+     error = fget(td, fd, cap_rights_init(&rights, CAP_SOCK_CLIENT), &sp->fp);
      if(error)
 	  return error;
 
-     if((error = fgetsock(td, fd, CAP_SOCK_CLIENT, &sp->soc, 0)) == 0) {
+     error = fgetsock(td, fd, cap_rights_init(&rights, CAP_SOCK_CLIENT),
+        &sp->soc, 0);
+     if(error == 0) {
 	  sp->td = td;
 	  isc_start_receiver(sp);
      }
diff --git a/sys/dev/mfi/mfi_linux.c b/sys/dev/mfi/mfi_linux.c
index 3328a66..429d496 100644
--- a/sys/dev/mfi/mfi_linux.c
+++ b/sys/dev/mfi/mfi_linux.c
@@ -84,6 +84,7 @@ MODULE_DEPEND(mfi, linux, 1, 1, 1);
 static int
 mfi_linux_ioctl(struct thread *p, struct linux_ioctl_args *args)
 {
+	cap_rights_t rights;
 	struct file *fp;
 	int error;
 	u_long cmd = args->cmd;
@@ -97,7 +98,8 @@ mfi_linux_ioctl(struct thread *p, struct linux_ioctl_args *args)
 		break;
 	}
 
-	if ((error = fget(p, args->fd, CAP_IOCTL, &fp)) != 0)
+	error = fget(p, args->fd, cap_rights_init(&rights, CAP_IOCTL), &fp);
+	if (error != 0)
 		return (error);
 	error = fo_ioctl(fp, cmd, (caddr_t)args->arg, p->td_ucred, p);
 	fdrop(fp, p);
diff --git a/sys/dev/tdfx/tdfx_linux.c b/sys/dev/tdfx/tdfx_linux.c
index 0b769f0..fa39ab1 100644
--- a/sys/dev/tdfx/tdfx_linux.c
+++ b/sys/dev/tdfx/tdfx_linux.c
@@ -45,6 +45,7 @@ LINUX_IOCTL_SET(tdfx, LINUX_IOCTL_TDFX_MIN, LINUX_IOCTL_TDFX_MAX);
 static int
 linux_ioctl_tdfx(struct thread *td, struct linux_ioctl_args* args)
 {
+   cap_rights_t rights;
    int error = 0;
    u_long cmd = args->cmd & 0xffff;
 
@@ -54,7 +55,8 @@ linux_ioctl_tdfx(struct thread *td, struct linux_ioctl_args* args)
 
    struct file *fp;
 
-   if ((error = fget(td, args->fd, CAP_IOCTL, &fp)) != 0)
+   error = fget(td, args->fd, cap_rights_init(&rights, CAP_IOCTL), &fp);
+   if (error != 0)
 	   return (error);
    /* We simply copy the data and send it right to ioctl */
    copyin((caddr_t)args->arg, &d_pio, sizeof(d_pio));