Skip to contentSkip to conversation

repost: new sockcreate interface to specify context of socket when they are created

PATCH In review Eric Paris 5 participants 13 messages Jun 21, 2006

Subsystems (2)
lore
Share & tools
lore

Eric Paris

author
#0 lore raw

PATCH 1/1 repost: new sockcreate interface to specify context of socket when they are created

Updated to fix the style problems James pointed out. Retested to make sure it was still working.

Below is a patch to add a new /proc/self/attr/sockcreate A process may
write a context into this interface and all subsequent sockets created will be labeled with that context. This is the same idea as the fscreate interface where a process can specify the label of a file about to be created. At this time one envisioned user of this will be xinetd. It will be able to better label sockets for the actual services. At this time all sockets take the label of the creating process, so all xinitd sockets would just be labeled the same.

I tested this by creating a tcp sender and listener. The sender was able to write to this new proc file and then create sockets with the specified label. I am able to be sure the new label was used since the avc denial messages kicked out by the kernel included both the new security permission setsockcreate and all the socket denials were for the new label, not the label of the running process.

I would like to propose this for 2.6.18

 fs/proc/base.c                               |    6 ++++++
 security/selinux/hooks.c                     |   20 ++++++++++++++++----
 security/selinux/include/av_perm_to_string.h |    1 +
 security/selinux/include/av_permissions.h    |    1 +
 security/selinux/include/objsec.h            |    1 +
 5 files changed, 25 insertions(+), 4 deletions(-)
fs/proc/base.c +6 -0
index 6cc77dc..4f9bf8f 100644
@@ -121,6 +121,7 @@ #ifdef CONFIG_SECURITY
 	PROC_TGID_ATTR_PREV,
 	PROC_TGID_ATTR_EXEC,
 	PROC_TGID_ATTR_FSCREATE,
+	PROC_TGID_ATTR_SOCKCREATE,
 #endif
 #ifdef CONFIG_AUDITSYSCALL
 	PROC_TGID_LOGINUID,
@@ -162,6 +163,7 @@ #ifdef CONFIG_SECURITY
 	PROC_TID_ATTR_PREV,
 	PROC_TID_ATTR_EXEC,
 	PROC_TID_ATTR_FSCREATE,
+	PROC_TID_ATTR_SOCKCREATE,
 #endif
 #ifdef CONFIG_AUDITSYSCALL
 	PROC_TID_LOGINUID,
@@ -275,6 +277,7 @@ static struct pid_entry tgid_attr_stuff[
 	E(PROC_TGID_ATTR_PREV,     "prev",     S_IFREG|S_IRUGO),
 	E(PROC_TGID_ATTR_EXEC,     "exec",     S_IFREG|S_IRUGO|S_IWUGO),
 	E(PROC_TGID_ATTR_FSCREATE, "fscreate", S_IFREG|S_IRUGO|S_IWUGO),
+	E(PROC_TGID_ATTR_SOCKCREATE, "sockcreate", S_IFREG|S_IRUGO|S_IWUGO),
 	{0,0,NULL,0}
 };
 static struct pid_entry tid_attr_stuff[] = {
@@ -282,6 +285,7 @@ static struct pid_entry tid_attr_stuff[]
 	E(PROC_TID_ATTR_PREV,      "prev",     S_IFREG|S_IRUGO),
 	E(PROC_TID_ATTR_EXEC,      "exec",     S_IFREG|S_IRUGO|S_IWUGO),
 	E(PROC_TID_ATTR_FSCREATE,  "fscreate", S_IFREG|S_IRUGO|S_IWUGO),
+	E(PROC_TID_ATTR_SOCKCREATE, "sockcreate", S_IFREG|S_IRUGO|S_IWUGO),
 	{0,0,NULL,0}
 };
 #endif
@@ -1800,6 +1804,8 @@ #ifdef CONFIG_SECURITY
 		case PROC_TGID_ATTR_EXEC:
 		case PROC_TID_ATTR_FSCREATE:
 		case PROC_TGID_ATTR_FSCREATE:
+		case PROC_TID_ATTR_SOCKCREATE:
+		case PROC_TGID_ATTR_SOCKCREATE:
 			inode->i_fop = &proc_pid_attr_operations;
 			break;
 #endif
security/selinux/hooks.c +16 -4
index 90b4cdc..33e035c 100644
@@ -1518,8 +1518,9 @@ static int selinux_bprm_set_security(str
 	/* Default to the current task SID. */
 	bsec->sid = tsec->sid;
 
-	/* Reset create SID on execve. */
+	/* Reset create and sockcreate SID on execve. */
 	tsec->create_sid = 0;
+	tsec->sockcreate_sid = 0;
 
 	if (tsec->exec_sid) {
 		newsid = tsec->exec_sid;
@@ -2571,9 +2572,10 @@ static int selinux_task_alloc_security(s
 	tsec2->osid = tsec1->osid;
 	tsec2->sid = tsec1->sid;
 
-	/* Retain the exec and create SIDs across fork */
+	/* Retain the exec, create, and sock SIDs across fork */
 	tsec2->exec_sid = tsec1->exec_sid;
 	tsec2->create_sid = tsec1->create_sid;
+	tsec2->sockcreate_sid = tsec1->sockcreate_sid;
 
 	/* Retain ptracer SID across fork, if any.
 	   This will be reset by the ptrace hook upon any
@@ -2913,12 +2915,14 @@ static int selinux_socket_create(int fam
 {
 	int err = 0;
 	struct task_security_struct *tsec;
+	u32 newsid;
 
 	if (kern)
 		goto out;
 
 	tsec = current->security;
-	err = avc_has_perm(tsec->sid, tsec->sid,
+	newsid = tsec->sockcreate_sid ? : tsec->sid;
+	err = avc_has_perm(tsec->sid, newsid,
 			   socket_type_to_security_class(family, type,
 			   protocol), SOCKET__CREATE, NULL);
 
@@ -2931,12 +2935,14 @@ static void selinux_socket_post_create(s
 {
 	struct inode_security_struct *isec;
 	struct task_security_struct *tsec;
+	u32 newsid;
 
 	isec = SOCK_INODE(sock)->i_security;
 
 	tsec = current->security;
+	newsid = tsec->sockcreate_sid ? : tsec->sid;
 	isec->sclass = socket_type_to_security_class(family, type, protocol);
-	isec->sid = kern ? SECINITSID_KERNEL : tsec->sid;
+	isec->sid = kern ? SECINITSID_KERNEL : newsid;
 	isec->initialized = 1;
 
 	return;
@@ -4114,6 +4120,8 @@ static int selinux_getprocattr(struct ta
 		sid = tsec->exec_sid;
 	else if (!strcmp(name, "fscreate"))
 		sid = tsec->create_sid;
+	else if (!strcmp(name, "sockcreate"))
+		sid = tsec->sockcreate_sid;
 	else
 		return -EINVAL;
 
@@ -4146,6 +4154,8 @@ static int selinux_setprocattr(struct ta
 		error = task_has_perm(current, p, PROCESS__SETEXEC);
 	else if (!strcmp(name, "fscreate"))
 		error = task_has_perm(current, p, PROCESS__SETFSCREATE);
+	else if (!strcmp(name, "sockcreate"))
+		error = task_has_perm(current, p, PROCESS__SETSOCKCREATE);
 	else if (!strcmp(name, "current"))
 		error = task_has_perm(current, p, PROCESS__SETCURRENT);
 	else
@@ -4175,6 +4185,8 @@ static int selinux_setprocattr(struct ta
 		tsec->exec_sid = sid;
 	else if (!strcmp(name, "fscreate"))
 		tsec->create_sid = sid;
+	else if (!strcmp(name, "sockcreate"))
+		tsec->sockcreate_sid = sid;
 	else if (!strcmp(name, "current")) {
 		struct av_decision avd;
 
security/selinux/include/av_perm_to_string.h +1 -0
index 591e98d..ce3a55d 100644
@@ -63,6 +63,7 @@
    S_(SECCLASS_PROCESS, PROCESS__GETATTR, "getattr")
    S_(SECCLASS_PROCESS, PROCESS__SETEXEC, "setexec")
    S_(SECCLASS_PROCESS, PROCESS__SETFSCREATE, "setfscreate")
+   S_(SECCLASS_PROCESS, PROCESS__SETSOCKCREATE, "setsockcreate")
    S_(SECCLASS_PROCESS, PROCESS__NOATSECURE, "noatsecure")
    S_(SECCLASS_PROCESS, PROCESS__SIGINH, "siginh")
    S_(SECCLASS_PROCESS, PROCESS__SETRLIMIT, "setrlimit")
security/selinux/include/av_permissions.h +1 -0
index d7f02ed..2a9a9b3 100644
@@ -467,6 +467,7 @@ #define PROCESS__SETCURRENT             
 #define PROCESS__EXECMEM                          0x02000000UL
 #define PROCESS__EXECSTACK                        0x04000000UL
 #define PROCESS__EXECHEAP                         0x08000000UL
+#define PROCESS__SETSOCKCREATE                    0x10000000UL
 
 #define IPC__CREATE                               0x00000001UL
 #define IPC__DESTROY                              0x00000002UL
security/selinux/include/objsec.h +1 -0
index 54c0307..1873451 100644
@@ -32,6 +32,7 @@ struct task_security_struct {
 	u32 sid;             /* current SID */
 	u32 exec_sid;        /* exec SID */
 	u32 create_sid;      /* fscreate SID */
+	u32 sockcreate_sid;  /* fscreate SID */
 	u32 ptrace_sid;      /* SID of ptrace parent */
 };
 

--
This message was distributed to subscribers of the selinux mailing list. If you no longer wish to subscribe, send mail to [email protected] with the words "unsubscribe selinux" without quotes as the message.

#1 lore raw
Quoted 3 lines · Eric Paris

S_(SECCLASS_PROCESS, PROCESS__SETEXEC, "setexec") S_(SECCLASS_PROCESS, PROCESS__SETFSCREATE, "setfscreate")
+ S_(SECCLASS_PROCESS, PROCESS__SETSOCKCREATE, "setsockcreate")

Did you manually place this here, or was it done by mkaccess_vector.sh in the policy package?

Signature · 7 lines
James Morris
<[email protected]>

--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to [email protected] with
the words "unsubscribe selinux" without quotes as the message.
author
#2 lore raw
Quoted 10 lines · James Morris
Quoted 3 lines · Eric Paris

S_(SECCLASS_PROCESS, PROCESS__SETEXEC, "setexec") S_(SECCLASS_PROCESS, PROCESS__SETFSCREATE, "setfscreate")
+ S_(SECCLASS_PROCESS, PROCESS__SETSOCKCREATE, "setsockcreate")

Did you manually place this here, or was it done by mkaccess_vector.sh in the policy package?

Manually. /me goes to look at mkaccess_vector.sh

-Eric

--
This message was distributed to subscribers of the selinux mailing list. If you no longer wish to subscribe, send mail to [email protected] with the words "unsubscribe selinux" without quotes as the message.

#3 lore raw
Quoted 7 lines · Eric Paris
Quoted 6 lines · James Morris
Quoted 3 lines · Eric Paris

S_(SECCLASS_PROCESS, PROCESS__SETEXEC, "setexec") S_(SECCLASS_PROCESS, PROCESS__SETFSCREATE, "setfscreate")
+ S_(SECCLASS_PROCESS, PROCESS__SETSOCKCREATE, "setsockcreate")

Quoted 5 lines
Quoted 4 lines

Did you manually place this here, or was it done by mkaccess_vector.sh in the policy package?

Manually. /me goes to look at mkaccess_vector.sh

/* This file is automatically generated. Do not edit. */

What you need to do is patch the refpolicy/policy/flask/access_vectors file (with keycreate patch applied first), then run make in that subdirectory to regenerate the headers, then copy them over to your linux-2.6.x/security/selinux/include tree by hand or by running make LINUXDIR=/path/to/linux-2.6.x tokern. You can also update the libselinux ones if you like (make LIBSEL=/path/to/libselinux tolib), but that is only necessary if the permission definition is being used by some application (userspace object manager, like dbus or X).

Signature · 8 lines
Stephen Smalley
National Security Agency


--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to [email protected] with
the words "unsubscribe selinux" without quotes as the message.
#4 lore raw
In reply to Eric Paris
Quoted 12 lines · Eric Paris

Updated to fix the style problems James pointed out. Retested to make sure it was still working.

Below is a patch to add a new /proc/self/attr/sockcreate A process may
write a context into this interface and all subsequent sockets created will be labeled with that context. This is the same idea as the fscreate interface where a process can specify the label of a file about to be created. At this time one envisioned user of this will be xinetd. It will be able to better label sockets for the actual services. At this time all sockets take the label of the creating process, so all xinitd sockets would just be labeled the same.

Quoted 8 lines

I tested this by creating a tcp sender and listener. The sender was able to write to this new proc file and then create sockets with the specified label. I am able to be sure the new label was used since the avc denial messages kicked out by the kernel included both the new security permission setsockcreate and all the socket denials were for the new label, not the label of the running process.

I would like to propose this for 2.6.18

/proc/self/attr/keycreate is in -mm, so this patch won't apply relative to it. Permission definition also has to be regenerated relative to the keycreate refpolicy patch, as that took the same permission slot.

Quoted 184 lines
 fs/proc/base.c                               |    6 ++++++
 security/selinux/hooks.c                     |   20 ++++++++++++++++----
 security/selinux/include/av_perm_to_string.h |    1 +
 security/selinux/include/av_permissions.h    |    1 +
 security/selinux/include/objsec.h            |    1 +
 5 files changed, 25 insertions(+), 4 deletions(-)
fs/proc/base.c +6 -0
index 6cc77dc..4f9bf8f 100644
@@ -121,6 +121,7 @@ #ifdef CONFIG_SECURITY
 	PROC_TGID_ATTR_PREV,
 	PROC_TGID_ATTR_EXEC,
 	PROC_TGID_ATTR_FSCREATE,
+	PROC_TGID_ATTR_SOCKCREATE,
 #endif
 #ifdef CONFIG_AUDITSYSCALL
 	PROC_TGID_LOGINUID,
@@ -162,6 +163,7 @@ #ifdef CONFIG_SECURITY
 	PROC_TID_ATTR_PREV,
 	PROC_TID_ATTR_EXEC,
 	PROC_TID_ATTR_FSCREATE,
+	PROC_TID_ATTR_SOCKCREATE,
 #endif
 #ifdef CONFIG_AUDITSYSCALL
 	PROC_TID_LOGINUID,
@@ -275,6 +277,7 @@ static struct pid_entry tgid_attr_stuff[
 	E(PROC_TGID_ATTR_PREV,     "prev",     S_IFREG|S_IRUGO),
 	E(PROC_TGID_ATTR_EXEC,     "exec",     S_IFREG|S_IRUGO|S_IWUGO),
 	E(PROC_TGID_ATTR_FSCREATE, "fscreate", S_IFREG|S_IRUGO|S_IWUGO),
+	E(PROC_TGID_ATTR_SOCKCREATE, "sockcreate", S_IFREG|S_IRUGO|S_IWUGO),
 	{0,0,NULL,0}
 };
 static struct pid_entry tid_attr_stuff[] = {
@@ -282,6 +285,7 @@ static struct pid_entry tid_attr_stuff[]
 	E(PROC_TID_ATTR_PREV,      "prev",     S_IFREG|S_IRUGO),
 	E(PROC_TID_ATTR_EXEC,      "exec",     S_IFREG|S_IRUGO|S_IWUGO),
 	E(PROC_TID_ATTR_FSCREATE,  "fscreate", S_IFREG|S_IRUGO|S_IWUGO),
+	E(PROC_TID_ATTR_SOCKCREATE, "sockcreate", S_IFREG|S_IRUGO|S_IWUGO),
 	{0,0,NULL,0}
 };
 #endif
@@ -1800,6 +1804,8 @@ #ifdef CONFIG_SECURITY
 		case PROC_TGID_ATTR_EXEC:
 		case PROC_TID_ATTR_FSCREATE:
 		case PROC_TGID_ATTR_FSCREATE:
+		case PROC_TID_ATTR_SOCKCREATE:
+		case PROC_TGID_ATTR_SOCKCREATE:
 			inode->i_fop = &proc_pid_attr_operations;
 			break;
 #endif
security/selinux/hooks.c +16 -4
index 90b4cdc..33e035c 100644
@@ -1518,8 +1518,9 @@ static int selinux_bprm_set_security(str
 	/* Default to the current task SID. */
 	bsec->sid = tsec->sid;
 
-	/* Reset create SID on execve. */
+	/* Reset create and sockcreate SID on execve. */
 	tsec->create_sid = 0;
+	tsec->sockcreate_sid = 0;
 
 	if (tsec->exec_sid) {
 		newsid = tsec->exec_sid;
@@ -2571,9 +2572,10 @@ static int selinux_task_alloc_security(s
 	tsec2->osid = tsec1->osid;
 	tsec2->sid = tsec1->sid;
 
-	/* Retain the exec and create SIDs across fork */
+	/* Retain the exec, create, and sock SIDs across fork */
 	tsec2->exec_sid = tsec1->exec_sid;
 	tsec2->create_sid = tsec1->create_sid;
+	tsec2->sockcreate_sid = tsec1->sockcreate_sid;
 
 	/* Retain ptracer SID across fork, if any.
 	   This will be reset by the ptrace hook upon any
@@ -2913,12 +2915,14 @@ static int selinux_socket_create(int fam
 {
 	int err = 0;
 	struct task_security_struct *tsec;
+	u32 newsid;
 
 	if (kern)
 		goto out;
 
 	tsec = current->security;
-	err = avc_has_perm(tsec->sid, tsec->sid,
+	newsid = tsec->sockcreate_sid ? : tsec->sid;
+	err = avc_has_perm(tsec->sid, newsid,
 			   socket_type_to_security_class(family, type,
 			   protocol), SOCKET__CREATE, NULL);
 
@@ -2931,12 +2935,14 @@ static void selinux_socket_post_create(s
 {
 	struct inode_security_struct *isec;
 	struct task_security_struct *tsec;
+	u32 newsid;
 
 	isec = SOCK_INODE(sock)->i_security;
 
 	tsec = current->security;
+	newsid = tsec->sockcreate_sid ? : tsec->sid;
 	isec->sclass = socket_type_to_security_class(family, type, protocol);
-	isec->sid = kern ? SECINITSID_KERNEL : tsec->sid;
+	isec->sid = kern ? SECINITSID_KERNEL : newsid;
 	isec->initialized = 1;
 
 	return;
@@ -4114,6 +4120,8 @@ static int selinux_getprocattr(struct ta
 		sid = tsec->exec_sid;
 	else if (!strcmp(name, "fscreate"))
 		sid = tsec->create_sid;
+	else if (!strcmp(name, "sockcreate"))
+		sid = tsec->sockcreate_sid;
 	else
 		return -EINVAL;
 
@@ -4146,6 +4154,8 @@ static int selinux_setprocattr(struct ta
 		error = task_has_perm(current, p, PROCESS__SETEXEC);
 	else if (!strcmp(name, "fscreate"))
 		error = task_has_perm(current, p, PROCESS__SETFSCREATE);
+	else if (!strcmp(name, "sockcreate"))
+		error = task_has_perm(current, p, PROCESS__SETSOCKCREATE);
 	else if (!strcmp(name, "current"))
 		error = task_has_perm(current, p, PROCESS__SETCURRENT);
 	else
@@ -4175,6 +4185,8 @@ static int selinux_setprocattr(struct ta
 		tsec->exec_sid = sid;
 	else if (!strcmp(name, "fscreate"))
 		tsec->create_sid = sid;
+	else if (!strcmp(name, "sockcreate"))
+		tsec->sockcreate_sid = sid;
 	else if (!strcmp(name, "current")) {
 		struct av_decision avd;
 
security/selinux/include/av_perm_to_string.h +1 -0
index 591e98d..ce3a55d 100644
@@ -63,6 +63,7 @@
    S_(SECCLASS_PROCESS, PROCESS__GETATTR, "getattr")
    S_(SECCLASS_PROCESS, PROCESS__SETEXEC, "setexec")
    S_(SECCLASS_PROCESS, PROCESS__SETFSCREATE, "setfscreate")
+   S_(SECCLASS_PROCESS, PROCESS__SETSOCKCREATE, "setsockcreate")
    S_(SECCLASS_PROCESS, PROCESS__NOATSECURE, "noatsecure")
    S_(SECCLASS_PROCESS, PROCESS__SIGINH, "siginh")
    S_(SECCLASS_PROCESS, PROCESS__SETRLIMIT, "setrlimit")
security/selinux/include/av_permissions.h +1 -0
index d7f02ed..2a9a9b3 100644
@@ -467,6 +467,7 @@ #define PROCESS__SETCURRENT             
 #define PROCESS__EXECMEM                          0x02000000UL
 #define PROCESS__EXECSTACK                        0x04000000UL
 #define PROCESS__EXECHEAP                         0x08000000UL
+#define PROCESS__SETSOCKCREATE                    0x10000000UL
 
 #define IPC__CREATE                               0x00000001UL
 #define IPC__DESTROY                              0x00000002UL
security/selinux/include/objsec.h +1 -1
index 54c0307..1873451 100644
@@ -32,6 +32,7 @@ struct task_security_struct {
 	u32 sid;             /* current SID */
 	u32 exec_sid;        /* exec SID */
 	u32 create_sid;      /* fscreate SID */
+	u32 sockcreate_sid;  /* fscreate SID */
 	u32 ptrace_sid;      /* SID of ptrace parent */
 };
 
 
 
 
--

This message was distributed to subscribers of the selinux mailing list. If you no longer wish to subscribe, send mail to [email protected] with the words "unsubscribe selinux" without quotes as the message.

Signature · 8 lines
Stephen Smalley
National Security Agency


--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to [email protected] with
the words "unsubscribe selinux" without quotes as the message.
#5 lore raw
Quoted 4 lines · Stephen Smalley

/proc/self/attr/keycreate is in -mm, so this patch won't apply relative to it. Permission definition also has to be regenerated relative to the keycreate refpolicy patch, as that took the same permission slot.

I'm manually merging with -mm now.

- James

Signature · 7 lines
James Morris
<[email protected]>

--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to [email protected] with
the words "unsubscribe selinux" without quotes as the message.
#6 lore raw

Btw, we only have room for two permissions now in the process class:

#define PROCESS__SETSOCKCREATE 0x20000000UL

Signature · 7 lines
James Morris
<[email protected]>

--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to [email protected] with
the words "unsubscribe selinux" without quotes as the message.
#7 lore raw
In reply to Stephen Smalley
Quoted 2 lines · Stephen Smalley
Quoted 1 line · Eric Paris

Below is a patch to add a new /proc/self/attr/sockcreate

[cut]

Quoted 3 lines

/proc/self/attr/keycreate is in -mm, so this patch won't apply relative to it. Permission definition also has to be regenerated relative to the keycreate refpolicy patch, as that took the same permission slot.

I'm going to need a patch for refpolicy that has the the final ordering of permissions with key and sockcreate patches. I haven't added the key class at all yet since I have been waiting for the perm set to finalize.

Signature · 10 lines
Chris PeBenito
Tresys Technology, LLC
(410) 290-1411 x150



--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to [email protected] with
the words "unsubscribe selinux" without quotes as the message.
#8 lore raw
Quoted 11 lines · Christopher J. PeBenito
Quoted 2 lines · Stephen Smalley
Quoted 1 line · Eric Paris

Below is a patch to add a new /proc/self/attr/sockcreate

[cut]

Quoted 3 lines

/proc/self/attr/keycreate is in -mm, so this patch won't apply relative to it. Permission definition also has to be regenerated relative to the keycreate refpolicy patch, as that took the same permission slot.

I'm going to need a patch for refpolicy that has the the final ordering of permissions with key and sockcreate patches. I haven't added the key class at all yet since I have been waiting for the perm set to finalize.

Apply the key+keycreate patch first. sockcreate refpolicy patch has to be re-based and re-submitted, but the kernel patch has already been re-based and submitted.

Signature · 8 lines
Stephen Smalley
National Security Agency


--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to [email protected] with
the words "unsubscribe selinux" without quotes as the message.
#9 lore raw
Quoted 9 lines · Stephen Smalley
Quoted 8 lines · Christopher J. PeBenito
Quoted 2 lines · Stephen Smalley
Quoted 1 line · Eric Paris

Below is a patch to add a new /proc/self/attr/sockcreate

[cut]

Quoted 3 lines

/proc/self/attr/keycreate is in -mm, so this patch won't apply relative to it. Permission definition also has to be regenerated relative to the keycreate refpolicy patch, as that took the same permission slot.

Quoted 7 lines
Quoted 3 lines

I'm going to need a patch for refpolicy that has the the final ordering of permissions with key and sockcreate patches. I haven't added the key class at all yet since I have been waiting for the perm set to finalize.

Apply the key+keycreate patch first. sockcreate refpolicy patch has to be re-based and re-submitted, but the kernel patch has already been re-based and submitted.

Is the one from June 2nd the right one ([PATCH] refpolicy: Kernel access key retention policy support) to start, and then the one from yesterday that has setkeycreate?

Signature · 10 lines
Chris PeBenito
Tresys Technology, LLC
(410) 290-1411 x150



--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to [email protected] with
the words "unsubscribe selinux" without quotes as the message.
#10 lore raw

Yes, but make sure you use the one from June 2nd that has the class "key", not "retained_key". Thanks!

Quoted 14 lines · Christopher J. PeBenito
Quoted 13 lines · Stephen Smalley
Quoted 11 lines · Christopher J. PeBenito
Quoted 2 lines · Stephen Smalley
Quoted 1 line · Eric Paris

Below is a patch to add a new /proc/self/attr/sockcreate

[cut]

Quoted 3 lines

/proc/self/attr/keycreate is in -mm, so this patch won't apply relative to it. Permission definition also has to be regenerated relative to the keycreate refpolicy patch, as that took the same permission slot.

I'm going to need a patch for refpolicy that has the the final ordering of permissions with key and sockcreate patches. I haven't added the key class at all yet since I have been waiting for the perm set to finalize.

Quoted 8 lines
Quoted 3 lines

Apply the key+keycreate patch first. sockcreate refpolicy patch has to be re-based and re-submitted, but the kernel patch has already been re-based and submitted.

Is the one from June 2nd the right one ([PATCH] refpolicy: Kernel access key retention policy support) to start, and then the one from yesterday that has setkeycreate?

--
This message was distributed to subscribers of the selinux mailing list. If you no longer wish to subscribe, send mail to [email protected] with the words "unsubscribe selinux" without quotes as the message.

#11 lore raw
Quoted 2 lines · Michael LeMay

Yes, but make sure you use the one from June 2nd that has the class "key", not "retained_key". Thanks!

I have committed the attached patch for keys to refpolicy, please doublecheck that this is correct. I changed the raw rules to interfaces too.

Quoted 24 lines · Christopher J. PeBenito
Quoted 22 lines · Stephen Smalley
Quoted 9 lines · Christopher J. PeBenito
Quoted 8 lines · Stephen Smalley
Quoted 2 lines · Eric Paris
Quoted 1 line

Below is a patch to add a new /proc/self/attr/sockcreate

[cut]

Quoted 3 lines

/proc/self/attr/keycreate is in -mm, so this patch won't apply relative to it. Permission definition also has to be regenerated relative to the keycreate refpolicy patch, as that took the same permission slot.

Quoted 7 lines
Quoted 3 lines

I'm going to need a patch for refpolicy that has the the final ordering of permissions with key and sockcreate patches. I haven't added the key class at all yet since I have been waiting for the perm set to finalize.

Apply the key+keycreate patch first. sockcreate refpolicy patch has to be re-based and re-submitted, but the kernel patch has already been re-based and submitted.

Is the one from June 2nd the right one ([PATCH] refpolicy: Kernel access key retention policy support) to start, and then the one from yesterday that has setkeycreate?

Signature · 3 lines
Chris PeBenito
Tresys Technology, LLC
(410) 290-1411 x150
flask/security_classes +3 -0
Index: policy/flask/security_classes
===================================================================
@@ -90,4 +90,7 @@
 
 class packet
 
+# Kernel access key retention
+class key
+
 # FLASK
flask/access_vectors +12 -0
Index: policy/flask/access_vectors
===================================================================
@@ -252,6 +252,7 @@
 	execmem
 	execstack
 	execheap
+	setkeycreate
 }
 
 
@@ -617,3 +618,14 @@
 	recv
 	relabelto
 }
+
+class key
+{
+	view
+	read
+	write
+	search
+	link
+	setattr
+	create
+}
modules/services/xserver.te +2 -1
Index: policy/modules/services/xserver.te
===================================================================
@@ -80,7 +80,7 @@
 #
 
 allow xdm_t self:capability { setgid setuid sys_resource kill sys_tty_config mknod chown dac_override dac_read_search fowner fsetid ipc_owner sys_nice sys_rawio net_bind_service };
-allow xdm_t self:process { setexec setpgid setsched setrlimit signal_perms };
+allow xdm_t self:process { setexec setpgid setsched setrlimit signal_perms setkeycreate };
 allow xdm_t self:fifo_file rw_file_perms;
 allow xdm_t self:shm create_shm_perms;
 allow xdm_t self:sem create_sem_perms;
@@ -214,6 +214,7 @@
 
 userdom_dontaudit_use_unpriv_user_fds(xdm_t)
 userdom_dontaudit_search_sysadm_home_dirs(xdm_t)
+userdom_create_all_users_keys(xdm_t)
 # for .dmrc
 userdom_read_unpriv_users_home_content_files(xdm_t)
 # Search /proc for any user domain processes.
modules/system/locallogin.te +1 -0
Index: policy/modules/system/locallogin.te
===================================================================
@@ -169,6 +169,7 @@
 userdom_search_all_users_home_content(local_login_t)
 userdom_use_unpriv_users_fds(local_login_t)
 userdom_sigchld_all_users(local_login_t)
+userdom_create_all_users_keys(local_login_t)
 
 # Search for mail spool file.
 mta_getattr_spool(local_login_t)
modules/system/userdomain.if +22 -0
Index: policy/modules/system/userdomain.if
===================================================================
@@ -4732,6 +4732,28 @@
 
 ########################################
 ## <summary>
+##	Create keys for all user domains.
+## </summary>
+## <param name="domain">
+##	<summary>
+##	Domain allowed access.
+##	</summary>
+## </param>
+#
+interface(`userdom_create_all_users_keys',`
+	ifdef(`strict_policy',`
+		gen_require(`
+			attribute userdomain;
+		')
+
+		allow $1 userdomain:key create;
+	',`
+		unconfined_create_keys($1)
+	')
+')
+
+########################################
+## <summary>
 ##	Send a dbus message to all user domains.
 ## </summary>
 ## <param name="domain">
modules/system/unconfined.if +18 -0
Index: policy/modules/system/unconfined.if
===================================================================
@@ -360,6 +360,24 @@
 
 ########################################
 ## <summary>
+##	Create keys for the unconfined domain.
+## </summary>
+## <param name="domain">
+##	<summary>
+##	Domain allowed access.
+##	</summary>
+## </param>
+#
+interface(`unconfined_create_keys',`
+	gen_require(`
+		type unconfined_t;
+	')
+
+	allow $1 unconfined_t:key create;
+')
+
+########################################
+## <summary>
 ##	Send messages to the unconfined domain over dbus.
 ## </summary>
 ## <param name="domain">
#12 lore raw

That looks great, thanks for creating the interfaces.

Quoted 32 lines · Christopher J. PeBenito
Quoted 2 lines · Michael LeMay

Yes, but make sure you use the one from June 2nd that has the class "key", not "retained_key". Thanks!

I have committed the attached patch for keys to refpolicy, please doublecheck that this is correct. I changed the raw rules to interfaces too.

Quoted 24 lines · Christopher J. PeBenito
Quoted 22 lines · Stephen Smalley
Quoted 9 lines · Christopher J. PeBenito
Quoted 8 lines · Stephen Smalley
Quoted 2 lines · Eric Paris
Quoted 1 line

Below is a patch to add a new /proc/self/attr/sockcreate

[cut]

Quoted 3 lines

/proc/self/attr/keycreate is in -mm, so this patch won't apply relative to it. Permission definition also has to be regenerated relative to the keycreate refpolicy patch, as that took the same permission slot.

Quoted 7 lines
Quoted 3 lines

I'm going to need a patch for refpolicy that has the the final ordering of permissions with key and sockcreate patches. I haven't added the key class at all yet since I have been waiting for the perm set to finalize.

Apply the key+keycreate patch first. sockcreate refpolicy patch has to be re-based and re-submitted, but the kernel patch has already been re-based and submitted.

Is the one from June 2nd the right one ([PATCH] refpolicy: Kernel access key retention policy support) to start, and then the one from yesterday that has setkeycreate?

--
This message was distributed to subscribers of the selinux mailing list. If you no longer wish to subscribe, send mail to [email protected] with the words "unsubscribe selinux" without quotes as the message.