[PATCH 1/1] file capabilities: remove cap_task_kill()

[PATCH 1/1] file capabilities: remove cap_task_kill()

[serge]

The original justification for cap_task_kill() was as follows:

	check_kill_permission() does appropriate uid equivalence checks.
	However with file capabilities it becomes possible for an
	unprivileged user to execute a file with file capabilities
	resulting in a more privileged task with the same uid.

However now that cap_task_kill() always returns 0 (permission
granted) when p->uid==current->uid, the whole hook is worthless,
and only likely to create more subtle problems in the corner cases
where it might still be called but return -EPERM.  Those cases
are basically when uids are different but euid/suid is equivalent
as per the check in check_kill_permission().

This patch removes cap_task_kill().

Signed-off-by: Serge Hallyn <serge@hallyn.com>
---
 include/linux/security.h   |    3 +--
 security/capability.c      |    1 -
 security/commoncap.c       |   40 ----------------------------------------
 security/smack/smack_lsm.c |    5 -----
 4 files changed, 1 insertions(+), 48 deletions(-)

diff --git a/include/linux/security.h b/include/linux/security.h
index fe52cde..95cb830 100644
--- a/include/linux/security.h
+++ b/include/linux/security.h
@@ -63,7 +63,6 @@ extern int cap_inode_need_killpriv(struct dentry *dentry);
 extern int cap_inode_killpriv(struct dentry *dentry);
 extern int cap_task_post_setuid (uid_t old_ruid, uid_t old_euid, uid_t old_suid, int flags);
 extern void cap_task_reparent_to_init (struct task_struct *p);
-extern int cap_task_kill(struct task_struct *p, struct siginfo *info, int sig, u32 secid);
 extern int cap_task_setscheduler (struct task_struct *p, int policy, struct sched_param *lp);
 extern int cap_task_setioprio (struct task_struct *p, int ioprio);
 extern int cap_task_setnice (struct task_struct *p, int nice);
@@ -2138,7 +2137,7 @@ static inline int security_task_kill (struct task_struct *p,
 				      struct siginfo *info, int sig,
 				      u32 secid)
 {
-	return cap_task_kill(p, info, sig, secid);
+	return 0;
 }
 
 static inline int security_task_wait (struct task_struct *p)
diff --git a/security/capability.c b/security/capability.c
index 9e99f36..2c6e06d 100644
--- a/security/capability.c
+++ b/security/capability.c
@@ -40,7 +40,6 @@ static struct security_operations capability_ops = {
 	.inode_need_killpriv =		cap_inode_need_killpriv,
 	.inode_killpriv =		cap_inode_killpriv,
 
-	.task_kill =			cap_task_kill,
 	.task_setscheduler =		cap_task_setscheduler,
 	.task_setioprio =		cap_task_setioprio,
 	.task_setnice =			cap_task_setnice,
diff --git a/security/commoncap.c b/security/commoncap.c
index bb0c095..06d5c94 100644
--- a/security/commoncap.c
+++ b/security/commoncap.c
@@ -540,41 +540,6 @@ int cap_task_setnice (struct task_struct *p, int nice)
 	return cap_safe_nice(p);
 }
 
-int cap_task_kill(struct task_struct *p, struct siginfo *info,
-				int sig, u32 secid)
-{
-	if (info != SEND_SIG_NOINFO && (is_si_special(info) || SI_FROMKERNEL(info)))
-		return 0;
-
-	/*
-	 * Running a setuid root program raises your capabilities.
-	 * Killing your own setuid root processes was previously
-	 * allowed.
-	 * We must preserve legacy signal behavior in this case.
-	 */
-	if (p->uid == current->uid)
-		return 0;
-
-	/* sigcont is permitted within same session */
-	if (sig == SIGCONT && (task_session_nr(current) == task_session_nr(p)))
-		return 0;
-
-	if (secid)
-		/*
-		 * Signal sent as a particular user.
-		 * Capabilities are ignored.  May be wrong, but it's the
-		 * only thing we can do at the moment.
-		 * Used only by usb drivers?
-		 */
-		return 0;
-	if (cap_issubset(p->cap_permitted, current->cap_permitted))
-		return 0;
-	if (capable(CAP_KILL))
-		return 0;
-
-	return -EPERM;
-}
-
 /*
  * called from kernel/sys.c for prctl(PR_CABSET_DROP)
  * done without task_capability_lock() because it introduces
@@ -605,11 +570,6 @@ int cap_task_setnice (struct task_struct *p, int nice)
 {
 	return 0;
 }
-int cap_task_kill(struct task_struct *p, struct siginfo *info,
-				int sig, u32 secid)
-{
-	return 0;
-}
 #endif
 
 void cap_task_reparent_to_init (struct task_struct *p)
diff --git a/security/smack/smack_lsm.c b/security/smack/smack_lsm.c
index 770eb06..a9ca412 100644
--- a/security/smack/smack_lsm.c
+++ b/security/smack/smack_lsm.c
@@ -1124,11 +1124,6 @@ static int smack_task_movememory(struct task_struct *p)
 static int smack_task_kill(struct task_struct *p, struct siginfo *info,
 			   int sig, u32 secid)
 {
-	int rc;
-
-	rc = cap_task_kill(p, info, sig, secid);
-	if (rc != 0)
-		return rc;
 	/*
 	 * Special cases where signals really ought to go through
 	 * in spite of policy. Stephen Smalley suggests it may
-- 
1.5.2.5

Re: [PATCH 1/1] file capabilities: remove cap_task_kill()

[BuraphaLinux Server]

For 2.6.25-rc3,
Tested-By: John Gatewood Ham <buraphalinuxserver@gmail.com>

This fixes the 'at' command for non-root users.  Thank you.


On 2/29/08, serge@hallyn.com <serge@hallyn.com> wrote:
> The original justification for cap_task_kill() was as follows:
>
> 	check_kill_permission() does appropriate uid equivalence checks.
> 	However with file capabilities it becomes possible for an
> 	unprivileged user to execute a file with file capabilities
> 	resulting in a more privileged task with the same uid.
>
> However now that cap_task_kill() always returns 0 (permission
> granted) when p->uid==current->uid, the whole hook is worthless,
> and only likely to create more subtle problems in the corner cases
> where it might still be called but return -EPERM.  Those cases
> are basically when uids are different but euid/suid is equivalent
> as per the check in check_kill_permission().
>
> This patch removes cap_task_kill().
>
> Signed-off-by: Serge Hallyn <serge@hallyn.com>
> ---
>  include/linux/security.h   |    3 +--
>  security/capability.c      |    1 -
>  security/commoncap.c       |   40 ----------------------------------------
>  security/smack/smack_lsm.c |    5 -----
>  4 files changed, 1 insertions(+), 48 deletions(-)
>
> diff --git a/include/linux/security.h b/include/linux/security.h
> index fe52cde..95cb830 100644
> --- a/include/linux/security.h
> +++ b/include/linux/security.h
> @@ -63,7 +63,6 @@ extern int cap_inode_need_killpriv(struct dentry *dentry);
>  extern int cap_inode_killpriv(struct dentry *dentry);
>  extern int cap_task_post_setuid (uid_t old_ruid, uid_t old_euid, uid_t
> old_suid, int flags);
>  extern void cap_task_reparent_to_init (struct task_struct *p);
> -extern int cap_task_kill(struct task_struct *p, struct siginfo *info, int
> sig, u32 secid);
>  extern int cap_task_setscheduler (struct task_struct *p, int policy, struct
> sched_param *lp);
>  extern int cap_task_setioprio (struct task_struct *p, int ioprio);
>  extern int cap_task_setnice (struct task_struct *p, int nice);
> @@ -2138,7 +2137,7 @@ static inline int security_task_kill (struct
> task_struct *p,
>  				      struct siginfo *info, int sig,
>  				      u32 secid)
>  {
> -	return cap_task_kill(p, info, sig, secid);
> +	return 0;
>  }
>
>  static inline int security_task_wait (struct task_struct *p)
> diff --git a/security/capability.c b/security/capability.c
> index 9e99f36..2c6e06d 100644
> --- a/security/capability.c
> +++ b/security/capability.c
> @@ -40,7 +40,6 @@ static struct security_operations capability_ops = {
>  	.inode_need_killpriv =		cap_inode_need_killpriv,
>  	.inode_killpriv =		cap_inode_killpriv,
>
> -	.task_kill =			cap_task_kill,
>  	.task_setscheduler =		cap_task_setscheduler,
>  	.task_setioprio =		cap_task_setioprio,
>  	.task_setnice =			cap_task_setnice,
> diff --git a/security/commoncap.c b/security/commoncap.c
> index bb0c095..06d5c94 100644
> --- a/security/commoncap.c
> +++ b/security/commoncap.c
> @@ -540,41 +540,6 @@ int cap_task_setnice (struct task_struct *p, int nice)
>  	return cap_safe_nice(p);
>  }
>
> -int cap_task_kill(struct task_struct *p, struct siginfo *info,
> -				int sig, u32 secid)
> -{
> -	if (info != SEND_SIG_NOINFO && (is_si_special(info) ||
> SI_FROMKERNEL(info)))
> -		return 0;
> -
> -	/*
> -	 * Running a setuid root program raises your capabilities.
> -	 * Killing your own setuid root processes was previously
> -	 * allowed.
> -	 * We must preserve legacy signal behavior in this case.
> -	 */
> -	if (p->uid == current->uid)
> -		return 0;
> -
> -	/* sigcont is permitted within same session */
> -	if (sig == SIGCONT && (task_session_nr(current) == task_session_nr(p)))
> -		return 0;
> -
> -	if (secid)
> -		/*
> -		 * Signal sent as a particular user.
> -		 * Capabilities are ignored.  May be wrong, but it's the
> -		 * only thing we can do at the moment.
> -		 * Used only by usb drivers?
> -		 */
> -		return 0;
> -	if (cap_issubset(p->cap_permitted, current->cap_permitted))
> -		return 0;
> -	if (capable(CAP_KILL))
> -		return 0;
> -
> -	return -EPERM;
> -}
> -
>  /*
>   * called from kernel/sys.c for prctl(PR_CABSET_DROP)
>   * done without task_capability_lock() because it introduces
> @@ -605,11 +570,6 @@ int cap_task_setnice (struct task_struct *p, int nice)
>  {
>  	return 0;
>  }
> -int cap_task_kill(struct task_struct *p, struct siginfo *info,
> -				int sig, u32 secid)
> -{
> -	return 0;
> -}
>  #endif
>
>  void cap_task_reparent_to_init (struct task_struct *p)
> diff --git a/security/smack/smack_lsm.c b/security/smack/smack_lsm.c
> index 770eb06..a9ca412 100644
> --- a/security/smack/smack_lsm.c
> +++ b/security/smack/smack_lsm.c
> @@ -1124,11 +1124,6 @@ static int smack_task_movememory(struct task_struct
> *p)
>  static int smack_task_kill(struct task_struct *p, struct siginfo *info,
>  			   int sig, u32 secid)
>  {
> -	int rc;
> -
> -	rc = cap_task_kill(p, info, sig, secid);
> -	if (rc != 0)
> -		return rc;
>  	/*
>  	 * Special cases where signals really ought to go through
>  	 * in spite of policy. Stephen Smalley suggests it may
> --
> 1.5.2.5
>
>

Re: [PATCH 1/1] file capabilities: remove cap_task_kill()

[Serge E. Hallyn]

Quoting BuraphaLinux Server (buraphalinuxserver@gmail.com):
> For 2.6.25-rc3,
> Tested-By: John Gatewood Ham <buraphalinuxserver@gmail.com>
> 
> This fixes the 'at' command for non-root users.  Thank you.

Cool, thanks much for testing.

-serge

> On 2/29/08, serge@hallyn.com <serge@hallyn.com> wrote:
> > The original justification for cap_task_kill() was as follows:
> >
> > 	check_kill_permission() does appropriate uid equivalence checks.
> > 	However with file capabilities it becomes possible for an
> > 	unprivileged user to execute a file with file capabilities
> > 	resulting in a more privileged task with the same uid.
> >
> > However now that cap_task_kill() always returns 0 (permission
> > granted) when p->uid==current->uid, the whole hook is worthless,
> > and only likely to create more subtle problems in the corner cases
> > where it might still be called but return -EPERM.  Those cases
> > are basically when uids are different but euid/suid is equivalent
> > as per the check in check_kill_permission().
> >
> > This patch removes cap_task_kill().
> >
> > Signed-off-by: Serge Hallyn <serge@hallyn.com>
> > ---
> >  include/linux/security.h   |    3 +--
> >  security/capability.c      |    1 -
> >  security/commoncap.c       |   40 ----------------------------------------
> >  security/smack/smack_lsm.c |    5 -----
> >  4 files changed, 1 insertions(+), 48 deletions(-)
> >
> > diff --git a/include/linux/security.h b/include/linux/security.h
> > index fe52cde..95cb830 100644
> > --- a/include/linux/security.h
> > +++ b/include/linux/security.h
> > @@ -63,7 +63,6 @@ extern int cap_inode_need_killpriv(struct dentry *dentry);
> >  extern int cap_inode_killpriv(struct dentry *dentry);
> >  extern int cap_task_post_setuid (uid_t old_ruid, uid_t old_euid, uid_t
> > old_suid, int flags);
> >  extern void cap_task_reparent_to_init (struct task_struct *p);
> > -extern int cap_task_kill(struct task_struct *p, struct siginfo *info, int
> > sig, u32 secid);
> >  extern int cap_task_setscheduler (struct task_struct *p, int policy, struct
> > sched_param *lp);
> >  extern int cap_task_setioprio (struct task_struct *p, int ioprio);
> >  extern int cap_task_setnice (struct task_struct *p, int nice);
> > @@ -2138,7 +2137,7 @@ static inline int security_task_kill (struct
> > task_struct *p,
> >  				      struct siginfo *info, int sig,
> >  				      u32 secid)
> >  {
> > -	return cap_task_kill(p, info, sig, secid);
> > +	return 0;
> >  }
> >
> >  static inline int security_task_wait (struct task_struct *p)
> > diff --git a/security/capability.c b/security/capability.c
> > index 9e99f36..2c6e06d 100644
> > --- a/security/capability.c
> > +++ b/security/capability.c
> > @@ -40,7 +40,6 @@ static struct security_operations capability_ops = {
> >  	.inode_need_killpriv =		cap_inode_need_killpriv,
> >  	.inode_killpriv =		cap_inode_killpriv,
> >
> > -	.task_kill =			cap_task_kill,
> >  	.task_setscheduler =		cap_task_setscheduler,
> >  	.task_setioprio =		cap_task_setioprio,
> >  	.task_setnice =			cap_task_setnice,
> > diff --git a/security/commoncap.c b/security/commoncap.c
> > index bb0c095..06d5c94 100644
> > --- a/security/commoncap.c
> > +++ b/security/commoncap.c
> > @@ -540,41 +540,6 @@ int cap_task_setnice (struct task_struct *p, int nice)
> >  	return cap_safe_nice(p);
> >  }
> >
> > -int cap_task_kill(struct task_struct *p, struct siginfo *info,
> > -				int sig, u32 secid)
> > -{
> > -	if (info != SEND_SIG_NOINFO && (is_si_special(info) ||
> > SI_FROMKERNEL(info)))
> > -		return 0;
> > -
> > -	/*
> > -	 * Running a setuid root program raises your capabilities.
> > -	 * Killing your own setuid root processes was previously
> > -	 * allowed.
> > -	 * We must preserve legacy signal behavior in this case.
> > -	 */
> > -	if (p->uid == current->uid)
> > -		return 0;
> > -
> > -	/* sigcont is permitted within same session */
> > -	if (sig == SIGCONT && (task_session_nr(current) == task_session_nr(p)))
> > -		return 0;
> > -
> > -	if (secid)
> > -		/*
> > -		 * Signal sent as a particular user.
> > -		 * Capabilities are ignored.  May be wrong, but it's the
> > -		 * only thing we can do at the moment.
> > -		 * Used only by usb drivers?
> > -		 */
> > -		return 0;
> > -	if (cap_issubset(p->cap_permitted, current->cap_permitted))
> > -		return 0;
> > -	if (capable(CAP_KILL))
> > -		return 0;
> > -
> > -	return -EPERM;
> > -}
> > -
> >  /*
> >   * called from kernel/sys.c for prctl(PR_CABSET_DROP)
> >   * done without task_capability_lock() because it introduces
> > @@ -605,11 +570,6 @@ int cap_task_setnice (struct task_struct *p, int nice)
> >  {
> >  	return 0;
> >  }
> > -int cap_task_kill(struct task_struct *p, struct siginfo *info,
> > -				int sig, u32 secid)
> > -{
> > -	return 0;
> > -}
> >  #endif
> >
> >  void cap_task_reparent_to_init (struct task_struct *p)
> > diff --git a/security/smack/smack_lsm.c b/security/smack/smack_lsm.c
> > index 770eb06..a9ca412 100644
> > --- a/security/smack/smack_lsm.c
> > +++ b/security/smack/smack_lsm.c
> > @@ -1124,11 +1124,6 @@ static int smack_task_movememory(struct task_struct
> > *p)
> >  static int smack_task_kill(struct task_struct *p, struct siginfo *info,
> >  			   int sig, u32 secid)
> >  {
> > -	int rc;
> > -
> > -	rc = cap_task_kill(p, info, sig, secid);
> > -	if (rc != 0)
> > -		return rc;
> >  	/*
> >  	 * Special cases where signals really ought to go through
> >  	 * in spite of policy. Stephen Smalley suggests it may
> > --
> > 1.5.2.5
> >
> >
> --
> To unsubscribe from this list: send the line "unsubscribe linux-security-module" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH 1/1] file capabilities: remove cap_task_kill()

[Luiz Fernando N. Capitulino]

Em Thu, 28 Feb 2008 11:38:17 -0600
serge@hallyn.com escreveu:

| The original justification for cap_task_kill() was as follows:
| 
| 	check_kill_permission() does appropriate uid equivalence checks.
| 	However with file capabilities it becomes possible for an
| 	unprivileged user to execute a file with file capabilities
| 	resulting in a more privileged task with the same uid.
| 
| However now that cap_task_kill() always returns 0 (permission
| granted) when p->uid==current->uid, the whole hook is worthless,
| and only likely to create more subtle problems in the corner cases
| where it might still be called but return -EPERM.  Those cases
| are basically when uids are different but euid/suid is equivalent
| as per the check in check_kill_permission().
| 
| This patch removes cap_task_kill().

 2.6.24 seems to have the same bug, what about a rediff for it and
submit the patch to -stable team?

 Thanks.

-- 
Luiz Fernando N. Capitulino

Re: [PATCH 1/1] file capabilities: remove cap_task_kill()

[serge]

Quoting Luiz Fernando N. Capitulino (lcapitulino@mandriva.com.br):
> Em Thu, 28 Feb 2008 11:38:17 -0600
> serge@hallyn.com escreveu:
> 
> | The original justification for cap_task_kill() was as follows:
> | 
> | 	check_kill_permission() does appropriate uid equivalence checks.
> | 	However with file capabilities it becomes possible for an
> | 	unprivileged user to execute a file with file capabilities
> | 	resulting in a more privileged task with the same uid.
> | 
> | However now that cap_task_kill() always returns 0 (permission
> | granted) when p->uid==current->uid, the whole hook is worthless,
> | and only likely to create more subtle problems in the corner cases
> | where it might still be called but return -EPERM.  Those cases
> | are basically when uids are different but euid/suid is equivalent
> | as per the check in check_kill_permission().
> | 
> | This patch removes cap_task_kill().
> 
>  2.6.24 seems to have the same bug, what about a rediff for it and
> submit the patch to -stable team?

Luiz, could you confirm that the below works?

thanks,
-serge

>From c77b7d418933c14707383f06a1da61169e84071b Mon Sep 17 00:00:00 2001
From: Serge Hallyn <serge@hallyn.com>
Date: Fri, 29 Feb 2008 15:14:57 +0000
Subject: [PATCH 1/1] file capabilities: remove cap_task_kill()

The original justification for cap_task_kill() was as follows:

	check_kill_permission() does appropriate uid equivalence checks.
	However with file capabilities it becomes possible for an
	unprivileged user to execute a file with file capabilities
	resulting in a more privileged task with the same uid.

However now that cap_task_kill() always returns 0 (permission
granted) when p->uid==current->uid, the whole hook is worthless,
and only likely to create more subtle problems in the corner cases
where it might still be called but return -EPERM.  Those cases
are basically when uids are different but euid/suid is equivalent
as per the check in check_kill_permission().

This patch removes cap_task_kill().

Signed-off-by: Serge Hallyn <serge@hallyn.com>
---
 include/linux/security.h |    3 +--
 security/capability.c    |    1 -
 security/commoncap.c     |   39 ---------------------------------------
 3 files changed, 1 insertions(+), 42 deletions(-)

diff --git a/include/linux/security.h b/include/linux/security.h
index ac05083..d842ee3 100644
--- a/include/linux/security.h
+++ b/include/linux/security.h
@@ -62,7 +62,6 @@ extern int cap_inode_need_killpriv(struct dentry *dentry);
 extern int cap_inode_killpriv(struct dentry *dentry);
 extern int cap_task_post_setuid (uid_t old_ruid, uid_t old_euid, uid_t old_suid, int flags);
 extern void cap_task_reparent_to_init (struct task_struct *p);
-extern int cap_task_kill(struct task_struct *p, struct siginfo *info, int sig, u32 secid);
 extern int cap_task_setscheduler (struct task_struct *p, int policy, struct sched_param *lp);
 extern int cap_task_setioprio (struct task_struct *p, int ioprio);
 extern int cap_task_setnice (struct task_struct *p, int nice);
@@ -2112,7 +2111,7 @@ static inline int security_task_kill (struct task_struct *p,
 				      struct siginfo *info, int sig,
 				      u32 secid)
 {
-	return cap_task_kill(p, info, sig, secid);
+	return 0;
 }
 
 static inline int security_task_wait (struct task_struct *p)
diff --git a/security/capability.c b/security/capability.c
index 9e99f36..2c6e06d 100644
--- a/security/capability.c
+++ b/security/capability.c
@@ -40,7 +40,6 @@ static struct security_operations capability_ops = {
 	.inode_need_killpriv =		cap_inode_need_killpriv,
 	.inode_killpriv =		cap_inode_killpriv,
 
-	.task_kill =			cap_task_kill,
 	.task_setscheduler =		cap_task_setscheduler,
 	.task_setioprio =		cap_task_setioprio,
 	.task_setnice =			cap_task_setnice,
diff --git a/security/commoncap.c b/security/commoncap.c
index ea61bc7..6e9065c 100644
--- a/security/commoncap.c
+++ b/security/commoncap.c
@@ -527,40 +527,6 @@ int cap_task_setnice (struct task_struct *p, int nice)
 	return cap_safe_nice(p);
 }
 
-int cap_task_kill(struct task_struct *p, struct siginfo *info,
-				int sig, u32 secid)
-{
-	if (info != SEND_SIG_NOINFO && (is_si_special(info) || SI_FROMKERNEL(info)))
-		return 0;
-
-	/*
-	 * Running a setuid root program raises your capabilities.
-	 * Killing your own setuid root processes was previously
-	 * allowed.
-	 * We must preserve legacy signal behavior in this case.
-	 */
-	if (p->euid == 0 && p->uid == current->uid)
-		return 0;
-
-	/* sigcont is permitted within same session */
-	if (sig == SIGCONT && (task_session_nr(current) == task_session_nr(p)))
-		return 0;
-
-	if (secid)
-		/*
-		 * Signal sent as a particular user.
-		 * Capabilities are ignored.  May be wrong, but it's the
-		 * only thing we can do at the moment.
-		 * Used only by usb drivers?
-		 */
-		return 0;
-	if (cap_issubset(p->cap_permitted, current->cap_permitted))
-		return 0;
-	if (capable(CAP_KILL))
-		return 0;
-
-	return -EPERM;
-}
 #else
 int cap_task_setscheduler (struct task_struct *p, int policy,
 			   struct sched_param *lp)
@@ -575,11 +541,6 @@ int cap_task_setnice (struct task_struct *p, int nice)
 {
 	return 0;
 }
-int cap_task_kill(struct task_struct *p, struct siginfo *info,
-				int sig, u32 secid)
-{
-	return 0;
-}
 #endif
 
 void cap_task_reparent_to_init (struct task_struct *p)
-- 
1.5.2.5

Re: [PATCH 1/1] file capabilities: remove cap_task_kill()

[Andrew G. Morgan]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Acked-by: Andrew G. Morgan <morgan@kernel.org>

Cheers

Andrew

serge@hallyn.com wrote:
| The original justification for cap_task_kill() was as follows:
|
| 	check_kill_permission() does appropriate uid equivalence checks.
| 	However with file capabilities it becomes possible for an
| 	unprivileged user to execute a file with file capabilities
| 	resulting in a more privileged task with the same uid.
|
| However now that cap_task_kill() always returns 0 (permission
| granted) when p->uid==current->uid, the whole hook is worthless,
| and only likely to create more subtle problems in the corner cases
| where it might still be called but return -EPERM.  Those cases
| are basically when uids are different but euid/suid is equivalent
| as per the check in check_kill_permission().
|
| This patch removes cap_task_kill().
|
| Signed-off-by: Serge Hallyn <serge@hallyn.com>
| ---
|  include/linux/security.h   |    3 +--
|  security/capability.c      |    1 -
|  security/commoncap.c       |   40
- ----------------------------------------
|  security/smack/smack_lsm.c |    5 -----
|  4 files changed, 1 insertions(+), 48 deletions(-)
|
| diff --git a/include/linux/security.h b/include/linux/security.h
| index fe52cde..95cb830 100644
| --- a/include/linux/security.h
| +++ b/include/linux/security.h
| @@ -63,7 +63,6 @@ extern int cap_inode_need_killpriv(struct dentry
*dentry);
|  extern int cap_inode_killpriv(struct dentry *dentry);
|  extern int cap_task_post_setuid (uid_t old_ruid, uid_t old_euid,
uid_t old_suid, int flags);
|  extern void cap_task_reparent_to_init (struct task_struct *p);
| -extern int cap_task_kill(struct task_struct *p, struct siginfo *info,
int sig, u32 secid);
|  extern int cap_task_setscheduler (struct task_struct *p, int policy,
struct sched_param *lp);
|  extern int cap_task_setioprio (struct task_struct *p, int ioprio);
|  extern int cap_task_setnice (struct task_struct *p, int nice);
| @@ -2138,7 +2137,7 @@ static inline int security_task_kill (struct
task_struct *p,
|  				      struct siginfo *info, int sig,
|  				      u32 secid)
|  {
| -	return cap_task_kill(p, info, sig, secid);
| +	return 0;
|  }
|
|  static inline int security_task_wait (struct task_struct *p)
| diff --git a/security/capability.c b/security/capability.c
| index 9e99f36..2c6e06d 100644
| --- a/security/capability.c
| +++ b/security/capability.c
| @@ -40,7 +40,6 @@ static struct security_operations capability_ops = {
|  	.inode_need_killpriv =		cap_inode_need_killpriv,
|  	.inode_killpriv =		cap_inode_killpriv,
|
| -	.task_kill =			cap_task_kill,
|  	.task_setscheduler =		cap_task_setscheduler,
|  	.task_setioprio =		cap_task_setioprio,
|  	.task_setnice =			cap_task_setnice,
| diff --git a/security/commoncap.c b/security/commoncap.c
| index bb0c095..06d5c94 100644
| --- a/security/commoncap.c
| +++ b/security/commoncap.c
| @@ -540,41 +540,6 @@ int cap_task_setnice (struct task_struct *p, int
nice)
|  	return cap_safe_nice(p);
|  }
|
| -int cap_task_kill(struct task_struct *p, struct siginfo *info,
| -				int sig, u32 secid)
| -{
| -	if (info != SEND_SIG_NOINFO && (is_si_special(info) ||
SI_FROMKERNEL(info)))
| -		return 0;
| -
| -	/*
| -	 * Running a setuid root program raises your capabilities.
| -	 * Killing your own setuid root processes was previously
| -	 * allowed.
| -	 * We must preserve legacy signal behavior in this case.
| -	 */
| -	if (p->uid == current->uid)
| -		return 0;
| -
| -	/* sigcont is permitted within same session */
| -	if (sig == SIGCONT && (task_session_nr(current) == task_session_nr(p)))
| -		return 0;
| -
| -	if (secid)
| -		/*
| -		 * Signal sent as a particular user.
| -		 * Capabilities are ignored.  May be wrong, but it's the
| -		 * only thing we can do at the moment.
| -		 * Used only by usb drivers?
| -		 */
| -		return 0;
| -	if (cap_issubset(p->cap_permitted, current->cap_permitted))
| -		return 0;
| -	if (capable(CAP_KILL))
| -		return 0;
| -
| -	return -EPERM;
| -}
| -
|  /*
|   * called from kernel/sys.c for prctl(PR_CABSET_DROP)
|   * done without task_capability_lock() because it introduces
| @@ -605,11 +570,6 @@ int cap_task_setnice (struct task_struct *p, int
nice)
|  {
|  	return 0;
|  }
| -int cap_task_kill(struct task_struct *p, struct siginfo *info,
| -				int sig, u32 secid)
| -{
| -	return 0;
| -}
|  #endif
|
|  void cap_task_reparent_to_init (struct task_struct *p)
| diff --git a/security/smack/smack_lsm.c b/security/smack/smack_lsm.c
| index 770eb06..a9ca412 100644
| --- a/security/smack/smack_lsm.c
| +++ b/security/smack/smack_lsm.c
| @@ -1124,11 +1124,6 @@ static int smack_task_movememory(struct
task_struct *p)
|  static int smack_task_kill(struct task_struct *p, struct siginfo *info,
|  			   int sig, u32 secid)
|  {
| -	int rc;
| -
| -	rc = cap_task_kill(p, info, sig, secid);
| -	if (rc != 0)
| -		return rc;
|  	/*
|  	 * Special cases where signals really ought to go through
|  	 * in spite of policy. Stephen Smalley suggests it may

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.6 (GNU/Linux)

iD8DBQFHydMS+bHCR3gb8jsRAjOsAJ9uFe8/h1uBzFDtxy77haw1E7v4PACgw4kg
P7pDqvlQLP6kPWzj/KmGo00=
=4U5z
-----END PGP SIGNATURE-----

Re: [PATCH 1/1] file capabilities: remove cap_task_kill()

[Luiz Fernando N. Capitulino]

Em Fri, 29 Feb 2008 15:26:34 -0600
serge@hallyn.com escreveu:

| Quoting Luiz Fernando N. Capitulino (lcapitulino@mandriva.com.br):
| > Em Thu, 28 Feb 2008 11:38:17 -0600
| > serge@hallyn.com escreveu:
| > 
| > | The original justification for cap_task_kill() was as follows:
| > | 
| > | 	check_kill_permission() does appropriate uid equivalence checks.
| > | 	However with file capabilities it becomes possible for an
| > | 	unprivileged user to execute a file with file capabilities
| > | 	resulting in a more privileged task with the same uid.
| > | 
| > | However now that cap_task_kill() always returns 0 (permission
| > | granted) when p->uid==current->uid, the whole hook is worthless,
| > | and only likely to create more subtle problems in the corner cases
| > | where it might still be called but return -EPERM.  Those cases
| > | are basically when uids are different but euid/suid is equivalent
| > | as per the check in check_kill_permission().
| > | 
| > | This patch removes cap_task_kill().
| > 
| >  2.6.24 seems to have the same bug, what about a rediff for it and
| > submit the patch to -stable team?
| 
| Luiz, could you confirm that the below works?

 Yes, it does.

 Thanks.

-- 
Luiz Fernando N. Capitulino

Re: [PATCH 1/1] file capabilities: remove cap_task_kill()

[Chris Friedhoff]

It works here against 2.6.24.3.
Originaly it was a fix - if I recall correctly - to allow a self
started X to kill completly. This works with the patch.

Chris


On Fri, 29 Feb 2008 15:26:34 -0600
serge@hallyn.com wrote:

> Quoting Luiz Fernando N. Capitulino (lcapitulino@mandriva.com.br):
> > Em Thu, 28 Feb 2008 11:38:17 -0600
> > serge@hallyn.com escreveu:
> > 
> > | The original justification for cap_task_kill() was as follows:
> > | 
> > | 	check_kill_permission() does appropriate uid equivalence checks.
> > | 	However with file capabilities it becomes possible for an
> > | 	unprivileged user to execute a file with file capabilities
> > | 	resulting in a more privileged task with the same uid.
> > | 
> > | However now that cap_task_kill() always returns 0 (permission
> > | granted) when p->uid==current->uid, the whole hook is worthless,
> > | and only likely to create more subtle problems in the corner cases
> > | where it might still be called but return -EPERM.  Those cases
> > | are basically when uids are different but euid/suid is equivalent
> > | as per the check in check_kill_permission().
> > | 
> > | This patch removes cap_task_kill().
> > 
> >  2.6.24 seems to have the same bug, what about a rediff for it and
> > submit the patch to -stable team?
> 
> Luiz, could you confirm that the below works?
> 
> thanks,
> -serge
> 
> From c77b7d418933c14707383f06a1da61169e84071b Mon Sep 17 00:00:00 2001
> From: Serge Hallyn <serge@hallyn.com>
> Date: Fri, 29 Feb 2008 15:14:57 +0000
> Subject: [PATCH 1/1] file capabilities: remove cap_task_kill()
> 
> The original justification for cap_task_kill() was as follows:
> 
> 	check_kill_permission() does appropriate uid equivalence checks.
> 	However with file capabilities it becomes possible for an
> 	unprivileged user to execute a file with file capabilities
> 	resulting in a more privileged task with the same uid.
> 
> However now that cap_task_kill() always returns 0 (permission
> granted) when p->uid==current->uid, the whole hook is worthless,
> and only likely to create more subtle problems in the corner cases
> where it might still be called but return -EPERM.  Those cases
> are basically when uids are different but euid/suid is equivalent
> as per the check in check_kill_permission().
> 
> This patch removes cap_task_kill().
> 
> Signed-off-by: Serge Hallyn <serge@hallyn.com>
> ---
>  include/linux/security.h |    3 +--
>  security/capability.c    |    1 -
>  security/commoncap.c     |   39 ---------------------------------------
>  3 files changed, 1 insertions(+), 42 deletions(-)
> 
> diff --git a/include/linux/security.h b/include/linux/security.h
> index ac05083..d842ee3 100644
> --- a/include/linux/security.h
> +++ b/include/linux/security.h
> @@ -62,7 +62,6 @@ extern int cap_inode_need_killpriv(struct dentry *dentry);
>  extern int cap_inode_killpriv(struct dentry *dentry);
>  extern int cap_task_post_setuid (uid_t old_ruid, uid_t old_euid, uid_t old_suid, int flags);
>  extern void cap_task_reparent_to_init (struct task_struct *p);
> -extern int cap_task_kill(struct task_struct *p, struct siginfo *info, int sig, u32 secid);
>  extern int cap_task_setscheduler (struct task_struct *p, int policy, struct sched_param *lp);
>  extern int cap_task_setioprio (struct task_struct *p, int ioprio);
>  extern int cap_task_setnice (struct task_struct *p, int nice);
> @@ -2112,7 +2111,7 @@ static inline int security_task_kill (struct task_struct *p,
>  				      struct siginfo *info, int sig,
>  				      u32 secid)
>  {
> -	return cap_task_kill(p, info, sig, secid);
> +	return 0;
>  }
>  
>  static inline int security_task_wait (struct task_struct *p)
> diff --git a/security/capability.c b/security/capability.c
> index 9e99f36..2c6e06d 100644
> --- a/security/capability.c
> +++ b/security/capability.c
> @@ -40,7 +40,6 @@ static struct security_operations capability_ops = {
>  	.inode_need_killpriv =		cap_inode_need_killpriv,
>  	.inode_killpriv =		cap_inode_killpriv,
>  
> -	.task_kill =			cap_task_kill,
>  	.task_setscheduler =		cap_task_setscheduler,
>  	.task_setioprio =		cap_task_setioprio,
>  	.task_setnice =			cap_task_setnice,
> diff --git a/security/commoncap.c b/security/commoncap.c
> index ea61bc7..6e9065c 100644
> --- a/security/commoncap.c
> +++ b/security/commoncap.c
> @@ -527,40 +527,6 @@ int cap_task_setnice (struct task_struct *p, int nice)
>  	return cap_safe_nice(p);
>  }
>  
> -int cap_task_kill(struct task_struct *p, struct siginfo *info,
> -				int sig, u32 secid)
> -{
> -	if (info != SEND_SIG_NOINFO && (is_si_special(info) || SI_FROMKERNEL(info)))
> -		return 0;
> -
> -	/*
> -	 * Running a setuid root program raises your capabilities.
> -	 * Killing your own setuid root processes was previously
> -	 * allowed.
> -	 * We must preserve legacy signal behavior in this case.
> -	 */
> -	if (p->euid == 0 && p->uid == current->uid)
> -		return 0;
> -
> -	/* sigcont is permitted within same session */
> -	if (sig == SIGCONT && (task_session_nr(current) == task_session_nr(p)))
> -		return 0;
> -
> -	if (secid)
> -		/*
> -		 * Signal sent as a particular user.
> -		 * Capabilities are ignored.  May be wrong, but it's the
> -		 * only thing we can do at the moment.
> -		 * Used only by usb drivers?
> -		 */
> -		return 0;
> -	if (cap_issubset(p->cap_permitted, current->cap_permitted))
> -		return 0;
> -	if (capable(CAP_KILL))
> -		return 0;
> -
> -	return -EPERM;
> -}
>  #else
>  int cap_task_setscheduler (struct task_struct *p, int policy,
>  			   struct sched_param *lp)
> @@ -575,11 +541,6 @@ int cap_task_setnice (struct task_struct *p, int nice)
>  {
>  	return 0;
>  }
> -int cap_task_kill(struct task_struct *p, struct siginfo *info,
> -				int sig, u32 secid)
> -{
> -	return 0;
> -}
>  #endif
>  
>  void cap_task_reparent_to_init (struct task_struct *p)
> -- 
> 1.5.2.5
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-security-module" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 


--------------------
Chris Friedhoff
chris@friedhoff.org