CRED: Make execve() take advantage of copy-on-write credentials

Make execve() take advantage of copy-on-write credentials, allowing it to set
up the credentials in advance, and then commit the whole lot after the point
of no return.

This patch and the preceding patches have been tested with the LTP SELinux
testsuite.

This patch makes several logical sets of alteration:

 (1) execve().

     The credential bits from struct linux_binprm are, for the most part,
     replaced with a single credentials pointer (bprm->cred).  This means that
     all the creds can be calculated in advance and then applied at the point
     of no return with no possibility of failure.

     I would like to replace bprm->cap_effective with:

	cap_isclear(bprm->cap_effective)

     but this seems impossible due to special behaviour for processes of pid 1
     (they always retain their parent's capability masks where normally they'd
     be changed - see cap_bprm_set_creds()).

     The following sequence of events now happens:

     (a) At the start of do_execve, the current task's cred_exec_mutex is
     	 locked to prevent PTRACE_ATTACH from obsoleting the calculation of
     	 creds that we make.

     (a) prepare_exec_creds() is then called to make a copy of the current
     	 task's credentials and prepare it.  This copy is then assigned to
     	 bprm->cred.

  	 This renders security_bprm_alloc() and security_bprm_free()
     	 unnecessary, and so they've been removed.

     (b) The determination of unsafe execution is now performed immediately
     	 after (a) rather than later on in the code.  The result is stored in
     	 bprm->unsafe for future reference.

     (c) prepare_binprm() is called, possibly multiple times.

     	 (i) This applies the result of set[ug]id binaries to the new creds
     	     attached to bprm->cred.  Personality bit clearance is recorded,
     	     but now deferred on the basis that the exec procedure may yet
     	     fail.

         (ii) This then calls the new security_bprm_set_creds().  This should
	     calculate the new LSM and capability credentials into *bprm->cred.

	     This folds together security_bprm_set() and parts of
	     security_bprm_apply_creds() (these two have been removed).
	     Anything that might fail must be done at this point.

         (iii) bprm->cred_prepared is set to 1.

	     bprm->cred_prepared is 0 on the first pass of the security
	     calculations, and 1 on all subsequent passes.  This allows SELinux
	     in (ii) to base its calculations only on the initial script and
	     not on the interpreter.

     (d) flush_old_exec() is called to commit the task to execution.  This
     	 performs the following steps with regard to credentials:

	 (i) Clear pdeath_signal and set dumpable on certain circumstances that
	     may not be covered by commit_creds().

         (ii) Clear any bits in current->personality that were deferred from
             (c.i).

     (e) install_exec_creds() [compute_creds() as was] is called to install the
     	 new credentials.  This performs the following steps with regard to
     	 credentials:

         (i) Calls security_bprm_committing_creds() to apply any security
             requirements, such as flushing unauthorised files in SELinux, that
             must be done before the credentials are changed.

	     This is made up of bits of security_bprm_apply_creds() and
	     security_bprm_post_apply_creds(), both of which have been removed.
	     This function is not allowed to fail; anything that might fail
	     must have been done in (c.ii).

         (ii) Calls commit_creds() to apply the new credentials in a single
             assignment (more or less).  Possibly pdeath_signal and dumpable
             should be part of struct creds.

	 (iii) Unlocks the task's cred_replace_mutex, thus allowing
	     PTRACE_ATTACH to take place.

         (iv) Clears The bprm->cred pointer as the credentials it was holding
             are now immutable.

         (v) Calls security_bprm_committed_creds() to apply any security
             alterations that must be done after the creds have been changed.
             SELinux uses this to flush signals and signal handlers.

     (f) If an error occurs before (d.i), bprm_free() will call abort_creds()
     	 to destroy the proposed new credentials and will then unlock
     	 cred_replace_mutex.  No changes to the credentials will have been
     	 made.

 (2) LSM interface.

     A number of functions have been changed, added or removed:

     (*) security_bprm_alloc(), ->bprm_alloc_security()
     (*) security_bprm_free(), ->bprm_free_security()

     	 Removed in favour of preparing new credentials and modifying those.

     (*) security_bprm_apply_creds(), ->bprm_apply_creds()
     (*) security_bprm_post_apply_creds(), ->bprm_post_apply_creds()

     	 Removed; split between security_bprm_set_creds(),
     	 security_bprm_committing_creds() and security_bprm_committed_creds().

     (*) security_bprm_set(), ->bprm_set_security()

     	 Removed; folded into security_bprm_set_creds().

     (*) security_bprm_set_creds(), ->bprm_set_creds()

     	 New.  The new credentials in bprm->creds should be checked and set up
     	 as appropriate.  bprm->cred_prepared is 0 on the first call, 1 on the
     	 second and subsequent calls.

     (*) security_bprm_committing_creds(), ->bprm_committing_creds()
     (*) security_bprm_committed_creds(), ->bprm_committed_creds()

     	 New.  Apply the security effects of the new credentials.  This
     	 includes closing unauthorised files in SELinux.  This function may not
     	 fail.  When the former is called, the creds haven't yet been applied
     	 to the process; when the latter is called, they have.

 	 The former may access bprm->cred, the latter may not.

 (3) SELinux.

     SELinux has a number of changes, in addition to those to support the LSM
     interface changes mentioned above:

     (a) The bprm_security_struct struct has been removed in favour of using
     	 the credentials-under-construction approach.

     (c) flush_unauthorized_files() now takes a cred pointer and passes it on
     	 to inode_has_perm(), file_has_perm() and dentry_open().

Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: James Morris <jmorris@namei.org>
Acked-by: Serge Hallyn <serue@us.ibm.com>
Signed-off-by: James Morris <jmorris@namei.org>

1 files changed, 118 insertions, 165 deletions

diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c
index c71bba78872f..21a592184633 100644
--- a/security/selinux/hooks.c
+++ b/security/selinux/hooks.c
@@ -2029,59 +2029,45 @@ static int selinux_vm_enough_memory(struct mm_struct *mm, long pages)
 
 /* binprm security operations */
 
-static int selinux_bprm_alloc_security(struct linux_binprm *bprm)
+static int selinux_bprm_set_creds(struct linux_binprm *bprm)
 {
-	struct bprm_security_struct *bsec;
-
-	bsec = kzalloc(sizeof(struct bprm_security_struct), GFP_KERNEL);
-	if (!bsec)
-		return -ENOMEM;
-
-	bsec->sid = SECINITSID_UNLABELED;
-	bsec->set = 0;
-
-	bprm->security = bsec;
-	return 0;
-}
-
-static int selinux_bprm_set_security(struct linux_binprm *bprm)
-{
-	struct task_security_struct *tsec;
-	struct inode *inode = bprm->file->f_path.dentry->d_inode;
+	const struct task_security_struct *old_tsec;
+	struct task_security_struct *new_tsec;
 	struct inode_security_struct *isec;
-	struct bprm_security_struct *bsec;
-	u32 newsid;
 	struct avc_audit_data ad;
+	struct inode *inode = bprm->file->f_path.dentry->d_inode;
 	int rc;
 
-	rc = secondary_ops->bprm_set_security(bprm);
+	rc = secondary_ops->bprm_set_creds(bprm);
 	if (rc)
 		return rc;
 
-	bsec = bprm->security;
-
-	if (bsec->set)
+	/* SELinux context only depends on initial program or script and not
+	 * the script interpreter */
+	if (bprm->cred_prepared)
 		return 0;
 
-	tsec = current_security();
+	old_tsec = current_security();
+	new_tsec = bprm->cred->security;
 	isec = inode->i_security;
 
 	/* Default to the current task SID. */
-	bsec->sid = tsec->sid;
+	new_tsec->sid = old_tsec->sid;
+	new_tsec->osid = old_tsec->sid;
 
 	/* Reset fs, key, and sock SIDs on execve. */
-	tsec->create_sid = 0;
-	tsec->keycreate_sid = 0;
-	tsec->sockcreate_sid = 0;
+	new_tsec->create_sid = 0;
+	new_tsec->keycreate_sid = 0;
+	new_tsec->sockcreate_sid = 0;
 
-	if (tsec->exec_sid) {
-		newsid = tsec->exec_sid;
+	if (old_tsec->exec_sid) {
+		new_tsec->sid = old_tsec->exec_sid;
 		/* Reset exec SID on execve. */
-		tsec->exec_sid = 0;
+		new_tsec->exec_sid = 0;
 	} else {
 		/* Check for a default transition on this program. */
-		rc = security_transition_sid(tsec->sid, isec->sid,
-					     SECCLASS_PROCESS, &newsid);
+		rc = security_transition_sid(old_tsec->sid, isec->sid,
+					     SECCLASS_PROCESS, &new_tsec->sid);
 		if (rc)
 			return rc;
 	}
@@ -2090,33 +2076,63 @@ static int selinux_bprm_set_security(struct linux_binprm *bprm)
 	ad.u.fs.path = bprm->file->f_path;
 
 	if (bprm->file->f_path.mnt->mnt_flags & MNT_NOSUID)
-		newsid = tsec->sid;
+		new_tsec->sid = old_tsec->sid;
 
-	if (tsec->sid == newsid) {
-		rc = avc_has_perm(tsec->sid, isec->sid,
+	if (new_tsec->sid == old_tsec->sid) {
+		rc = avc_has_perm(old_tsec->sid, isec->sid,
 				  SECCLASS_FILE, FILE__EXECUTE_NO_TRANS, &ad);
 		if (rc)
 			return rc;
 	} else {
 		/* Check permissions for the transition. */
-		rc = avc_has_perm(tsec->sid, newsid,
+		rc = avc_has_perm(old_tsec->sid, new_tsec->sid,
 				  SECCLASS_PROCESS, PROCESS__TRANSITION, &ad);
 		if (rc)
 			return rc;
 
-		rc = avc_has_perm(newsid, isec->sid,
+		rc = avc_has_perm(new_tsec->sid, isec->sid,
 				  SECCLASS_FILE, FILE__ENTRYPOINT, &ad);
 		if (rc)
 			return rc;
 
-		/* Clear any possibly unsafe personality bits on exec: */
-		current->personality &= ~PER_CLEAR_ON_SETID;
+		/* Check for shared state */
+		if (bprm->unsafe & LSM_UNSAFE_SHARE) {
+			rc = avc_has_perm(old_tsec->sid, new_tsec->sid,
+					  SECCLASS_PROCESS, PROCESS__SHARE,
+					  NULL);
+			if (rc)
+				return -EPERM;
+		}
+
+		/* Make sure that anyone attempting to ptrace over a task that
+		 * changes its SID has the appropriate permit */
+		if (bprm->unsafe &
+		    (LSM_UNSAFE_PTRACE | LSM_UNSAFE_PTRACE_CAP)) {
+			struct task_struct *tracer;
+			struct task_security_struct *sec;
+			u32 ptsid = 0;
+
+			rcu_read_lock();
+			tracer = tracehook_tracer_task(current);
+			if (likely(tracer != NULL)) {
+				sec = __task_cred(tracer)->security;
+				ptsid = sec->sid;
+			}
+			rcu_read_unlock();
+
+			if (ptsid != 0) {
+				rc = avc_has_perm(ptsid, new_tsec->sid,
+						  SECCLASS_PROCESS,
+						  PROCESS__PTRACE, NULL);
+				if (rc)
+					return -EPERM;
+			}
+		}
 
-		/* Set the security field to the new SID. */
-		bsec->sid = newsid;
+		/* Clear any possibly unsafe personality bits on exec: */
+		bprm->per_clear |= PER_CLEAR_ON_SETID;
 	}
 
-	bsec->set = 1;
 	return 0;
 }
 
@@ -2125,7 +2141,6 @@ static int selinux_bprm_check_security(struct linux_binprm *bprm)
 	return secondary_ops->bprm_check_security(bprm);
 }
 
-
 static int selinux_bprm_secureexec(struct linux_binprm *bprm)
 {
 	const struct cred *cred = current_cred();
@@ -2141,19 +2156,13 @@ static int selinux_bprm_secureexec(struct linux_binprm *bprm)
 		   the noatsecure permission is granted between
 		   the two SIDs, i.e. ahp returns 0. */
 		atsecure = avc_has_perm(osid, sid,
-					 SECCLASS_PROCESS,
-					 PROCESS__NOATSECURE, NULL);
+					SECCLASS_PROCESS,
+					PROCESS__NOATSECURE, NULL);
 	}
 
 	return (atsecure || secondary_ops->bprm_secureexec(bprm));
 }
 
-static void selinux_bprm_free_security(struct linux_binprm *bprm)
-{
-	kfree(bprm->security);
-	bprm->security = NULL;
-}
-
 extern struct vfsmount *selinuxfs_mount;
 extern struct dentry *selinux_null;
 
@@ -2252,108 +2261,78 @@ static inline void flush_unauthorized_files(const struct cred *cred,
 	spin_unlock(&files->file_lock);
 }
 
-static int selinux_bprm_apply_creds(struct linux_binprm *bprm, int unsafe)
+/*
+ * Prepare a process for imminent new credential changes due to exec
+ */
+static void selinux_bprm_committing_creds(struct linux_binprm *bprm)
 {
-	struct task_security_struct *tsec;
-	struct bprm_security_struct *bsec;
-	struct cred *new;
-	u32 sid;
-	int rc;
-
-	rc = secondary_ops->bprm_apply_creds(bprm, unsafe);
-	if (rc < 0)
-		return rc;
-
-	new = prepare_creds();
-	if (!new)
-		return -ENOMEM;
+	struct task_security_struct *new_tsec;
+	struct rlimit *rlim, *initrlim;
+	int rc, i;
 
-	tsec = new->security;
+	secondary_ops->bprm_committing_creds(bprm);
 
-	bsec = bprm->security;
-	sid = bsec->sid;
-
-	tsec->osid = tsec->sid;
-	bsec->unsafe = 0;
-	if (tsec->sid != sid) {
-		/* Check for shared state.  If not ok, leave SID
-		   unchanged and kill. */
-		if (unsafe & LSM_UNSAFE_SHARE) {
-			rc = avc_has_perm(tsec->sid, sid, SECCLASS_PROCESS,
-					PROCESS__SHARE, NULL);
-			if (rc) {
-				bsec->unsafe = 1;
-				goto out;
-			}
-		}
+	new_tsec = bprm->cred->security;
+	if (new_tsec->sid == new_tsec->osid)
+		return;
 
-		/* Check for ptracing, and update the task SID if ok.
-		   Otherwise, leave SID unchanged and kill. */
-		if (unsafe & (LSM_UNSAFE_PTRACE | LSM_UNSAFE_PTRACE_CAP)) {
-			struct task_struct *tracer;
-			struct task_security_struct *sec;
-			u32 ptsid = 0;
+	/* Close files for which the new task SID is not authorized. */
+	flush_unauthorized_files(bprm->cred, current->files);
 
-			rcu_read_lock();
-			tracer = tracehook_tracer_task(current);
-			if (likely(tracer != NULL)) {
-				sec = __task_cred(tracer)->security;
-				ptsid = sec->sid;
-			}
-			rcu_read_unlock();
+	/* Always clear parent death signal on SID transitions. */
+	current->pdeath_signal = 0;
 
-			if (ptsid != 0) {
-				rc = avc_has_perm(ptsid, sid, SECCLASS_PROCESS,
-						  PROCESS__PTRACE, NULL);
-				if (rc) {
-					bsec->unsafe = 1;
-					goto out;
-				}
-			}
+	/* Check whether the new SID can inherit resource limits from the old
+	 * SID.  If not, reset all soft limits to the lower of the current
+	 * task's hard limit and the init task's soft limit.
+	 *
+	 * Note that the setting of hard limits (even to lower them) can be
+	 * controlled by the setrlimit check.  The inclusion of the init task's
+	 * soft limit into the computation is to avoid resetting soft limits
+	 * higher than the default soft limit for cases where the default is
+	 * lower than the hard limit, e.g. RLIMIT_CORE or RLIMIT_STACK.
+	 */
+	rc = avc_has_perm(new_tsec->osid, new_tsec->sid, SECCLASS_PROCESS,
+			  PROCESS__RLIMITINH, NULL);
+	if (rc) {
+		for (i = 0; i < RLIM_NLIMITS; i++) {
+			rlim = current->signal->rlim + i;
+			initrlim = init_task.signal->rlim + i;
+			rlim->rlim_cur = min(rlim->rlim_max, initrlim->rlim_cur);
 		}
-		tsec->sid = sid;
+		update_rlimit_cpu(rlim->rlim_cur);
 	}
-
-out:
-	commit_creds(new);
-	return 0;
 }
 
 /*
- * called after apply_creds without the task lock held
+ * Clean up the process immediately after the installation of new credentials
+ * due to exec
  */
-static void selinux_bprm_post_apply_creds(struct linux_binprm *bprm)
+static void selinux_bprm_committed_creds(struct linux_binprm *bprm)
 {
-	const struct cred *cred = current_cred();
-	struct task_security_struct *tsec;
-	struct rlimit *rlim, *initrlim;
+	const struct task_security_struct *tsec = current_security();
 	struct itimerval itimer;
-	struct bprm_security_struct *bsec;
 	struct sighand_struct *psig;
+	u32 osid, sid;
 	int rc, i;
 	unsigned long flags;
 
-	tsec = current_security();
-	bsec = bprm->security;
+	secondary_ops->bprm_committed_creds(bprm);
 
-	if (bsec->unsafe) {
-		force_sig_specific(SIGKILL, current);
-		return;
-	}
-	if (tsec->osid == tsec->sid)
+	osid = tsec->osid;
+	sid = tsec->sid;
+
+	if (sid == osid)
 		return;
 
-	/* Close files for which the new task SID is not authorized. */
-	flush_unauthorized_files(cred, current->files);
-
-	/* Check whether the new SID can inherit signal state
-	   from the old SID.  If not, clear itimers to avoid
-	   subsequent signal generation and flush and unblock
-	   signals. This must occur _after_ the task SID has
-	  been updated so that any kill done after the flush
-	  will be checked against the new SID. */
-	rc = avc_has_perm(tsec->osid, tsec->sid, SECCLASS_PROCESS,
-			  PROCESS__SIGINH, NULL);
+	/* Check whether the new SID can inherit signal state from the old SID.
+	 * If not, clear itimers to avoid subsequent signal generation and
+	 * flush and unblock signals.
+	 *
+	 * This must occur _after_ the task SID has been updated so that any
+	 * kill done after the flush will be checked against the new SID.
+	 */
+	rc = avc_has_perm(osid, sid, SECCLASS_PROCESS, PROCESS__SIGINH, NULL);
 	if (rc) {
 		memset(&itimer, 0, sizeof itimer);
 		for (i = 0; i < 3; i++)
@@ -2366,32 +2345,8 @@ static void selinux_bprm_post_apply_creds(struct linux_binprm *bprm)
 		spin_unlock_irq(&current->sighand->siglock);
 	}
 
-	/* Always clear parent death signal on SID transitions. */
-	current->pdeath_signal = 0;
-
-	/* Check whether the new SID can inherit resource limits
-	   from the old SID.  If not, reset all soft limits to
-	   the lower of the current task's hard limit and the init
-	   task's soft limit.  Note that the setting of hard limits
-	   (even to lower them) can be controlled by the setrlimit
-	   check. The inclusion of the init task's soft limit into
-	   the computation is to avoid resetting soft limits higher
-	   than the default soft limit for cases where the default
-	   is lower than the hard limit, e.g. RLIMIT_CORE or
-	   RLIMIT_STACK.*/
-	rc = avc_has_perm(tsec->osid, tsec->sid, SECCLASS_PROCESS,
-			  PROCESS__RLIMITINH, NULL);
-	if (rc) {
-		for (i = 0; i < RLIM_NLIMITS; i++) {
-			rlim = current->signal->rlim + i;
-			initrlim = init_task.signal->rlim+i;
-			rlim->rlim_cur = min(rlim->rlim_max, initrlim->rlim_cur);
-		}
-		update_rlimit_cpu(rlim->rlim_cur);
-	}
-
-	/* Wake up the parent if it is waiting so that it can
-	   recheck wait permission to the new task SID. */
+	/* Wake up the parent if it is waiting so that it can recheck
+	 * wait permission to the new task SID. */
 	read_lock_irq(&tasklist_lock);
 	psig = current->parent->sighand;
 	spin_lock_irqsave(&psig->siglock, flags);
@@ -5556,12 +5511,10 @@ static struct security_operations selinux_ops = {
 	.netlink_send =			selinux_netlink_send,
 	.netlink_recv =			selinux_netlink_recv,
 
-	.bprm_alloc_security =		selinux_bprm_alloc_security,
-	.bprm_free_security =		selinux_bprm_free_security,
-	.bprm_apply_creds =		selinux_bprm_apply_creds,
-	.bprm_post_apply_creds =	selinux_bprm_post_apply_creds,
-	.bprm_set_security =		selinux_bprm_set_security,
+	.bprm_set_creds =		selinux_bprm_set_creds,
 	.bprm_check_security =		selinux_bprm_check_security,
+	.bprm_committing_creds =	selinux_bprm_committing_creds,
+	.bprm_committed_creds =		selinux_bprm_committed_creds,
 	.bprm_secureexec =		selinux_bprm_secureexec,
 
 	.sb_alloc_security =		selinux_sb_alloc_security,