Quoting Andrew Morgan (morgan@kernel.org ):
quoted text > -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Attached is what I consider only an RFC patch.
>
> I've not really thought through (to my satisfaction) the re-purposing of
> current->keep_capabilities in the non-filesystem-supporting-capability
> configuration, but this is basically the code I'm thinking about. (I'm
> typing this email from a system running this patch over 2.6.23-rc3-mm1
> so its not 'obviously' broken.)
>
> Adrian Bunk wrote:
> >> The user would be userspace...
> >>
> >> Unless by 'the user' you actually mean the patch itself which will allow
> >> the setting of secure_noroot per-process. I don't know for sure, but
> >> suspect Andrew might like to wait until file capabilities make it into
> >> and stabilize in Linus' tree before going on with that.
> >
> > That's what I am talking about.
> >
> > This patch should be submitted and discussed together with the changes
> > Andrew has for securebits.
>
> Cheers
>
> Andrew
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.2.6 (GNU/Linux)
>
> iD8DBQFG08y0QheEq9QabfIRAnUhAKCEHyUko292kULNTkRqQOGki2NohgCdGXvV
> bc+bHzBbI6sPimdf4UTAzGY=
> =vB0u
> -----END PGP SIGNATURE-----
quoted text > >From a8366ab7a12ed4283e24096e891ac13c1d471756 Mon Sep 17 00:00:00 2001
> From: Andrew Morgan <morgan@kernel.org>
> Date: Mon, 27 Aug 2007 23:09:20 -0700
> Subject: [PATCH] Remove global securebits from the kernel.
>
> In the absence of filesystem capability support, there is no
> longer any internal support for disabling root. The plain fact is
> that while there was internal support for it, the userspace API for
> enabling it was not implemented, and having a global personality
> of this sort was almost impossible to use.
>
> With filesystem capability support, a new prctl(PR_[GS]ET_CAPONLY)
> per-process flag is available. Once set (to 1) a process and all
> of its children will irrevocably lose root-is-all-capable privilege.
>
> NOTE: just because root within a given process tree is not capable,
> it does not mean root is impotent. Most Linux systems are riddled
> with critical system files that are owned by root (uid=0), so much
> care should be used relying on the security provided by this support.
> It is likely that the best way to use this feature is within some
> sort of restricted (chroot etc.) environment. Be especially careful
> of where you mount /proc/....
> ---
> include/linux/capability.h | 10 +++-
> include/linux/prctl.h | 4 +
> include/linux/sched.h | 1 -
> include/linux/securebits.h | 30 ----------
> include/linux/security.h | 17 ++++--
> kernel/sys.c | 14 +----
> security/capability.c | 1 +
> security/commoncap.c | 132 +++++++++++++++++++++++++++++---------------
> security/dummy.c | 5 +-
> security/security.c | 5 +-
> security/selinux/hooks.c | 6 +-
> 11 files changed, 124 insertions(+), 101 deletions(-)
> delete mode 100644 include/linux/securebits.h
>
> diff --git a/include/linux/capability.h b/include/linux/capability.h
> index 7a8d7ad..0fcef09 100644
> --- a/include/linux/capability.h
> +++ b/include/linux/capability.h
> @@ -147,8 +147,14 @@ typedef __u32 kernel_cap_t;
> ** Linux-specific capabilities
> **/
>
> -/* Transfer any capability in your permitted set to any pid,
> - remove any capability in your permitted set from any pid */
> +/* With filesystem support for capabilities configured:
> + * Permit the current process to raise any capability in its inheritable set.
> + * Permit the current process to lock the process into capability-only mode
> + * - prctl(PR_SET_CAPONLY, 1, ...)
> + * Otherwise:
> + * Transfer any capability in your permitted set to any pid,
> + * remove any capability in your permitted set from any pid
> + */
>
> #define CAP_SETPCAP 8
>
> diff --git a/include/linux/prctl.h b/include/linux/prctl.h
> index e2eff90..e3f49a7 100644
> --- a/include/linux/prctl.h
> +++ b/include/linux/prctl.h
> @@ -63,4 +63,8 @@
> #define PR_GET_SECCOMP 21
> #define PR_SET_SECCOMP 22
>
> +/* Get/Set process personality to secure-by-capabilities alone */
> +#define PR_GET_CAPONLY 23
> +#define PR_SET_CAPONLY 24
> +
> #endif /* _LINUX_PRCTL_H */
> diff --git a/include/linux/sched.h b/include/linux/sched.h
> index 2b3c936..be2e9c4 100644
> --- a/include/linux/sched.h
> +++ b/include/linux/sched.h
> @@ -67,7 +67,6 @@ struct sched_param {
> #include <linux/smp.h>
> #include <linux/sem.h>
> #include <linux/signal.h>
> -#include <linux/securebits.h>
> #include <linux/fs_struct.h>
> #include <linux/compiler.h>
> #include <linux/completion.h>
> diff --git a/include/linux/securebits.h b/include/linux/securebits.h
> deleted file mode 100644
> index 5b06178..0000000
> --- a/include/linux/securebits.h
> +++ /dev/null
> @@ -1,30 +0,0 @@
> -#ifndef _LINUX_SECUREBITS_H
> -#define _LINUX_SECUREBITS_H 1
> -
> -#define SECUREBITS_DEFAULT 0x00000000
> -
> -extern unsigned securebits;
> -
> -/* When set UID 0 has no special privileges. When unset, we support
> - inheritance of root-permissions and suid-root executable under
> - compatibility mode. We raise the effective and inheritable bitmasks
> - *of the executable file* if the effective uid of the new process is
> - 0. If the real uid is 0, we raise the inheritable bitmask of the
> - executable file. */
> -#define SECURE_NOROOT 0
> -
> -/* When set, setuid to/from uid 0 does not trigger capability-"fixes"
> - to be compatible with old programs relying on set*uid to loose
> - privileges. When unset, setuid doesn't change privileges. */
> -#define SECURE_NO_SETUID_FIXUP 2
> -
> -/* Each securesetting is implemented using two bits. One bit specify
> - whether the setting is on or off. The other bit specify whether the
> - setting is fixed or not. A setting which is fixed cannot be changed
> - from user-level. */
> -
> -#define issecure(X) ( (1 << (X+1)) & SECUREBITS_DEFAULT ? \
> - (1 << (X)) & SECUREBITS_DEFAULT : \
> - (1 << (X)) & securebits )
> -
> -#endif /* !_LINUX_SECUREBITS_H */
> diff --git a/include/linux/security.h b/include/linux/security.h
> index 13d48fd..4227464 100644
> --- a/include/linux/security.h
> +++ b/include/linux/security.h
> @@ -660,13 +660,18 @@ struct request_sock;
> * Return 0 if permission is granted.
> * @task_prctl:
> * Check permission before performing a process control operation on the
> - * current process.
> + * current process. This hook can also be used to provide LSM specific
> + * process control functions.
> * @option contains the operation.
> * @arg2 contains a argument.
> * @arg3 contains a argument.
> * @arg4 contains a argument.
> * @arg5 contains a argument.
> - * Return 0 if permission is granted.
> + * @errorp contains a pointer to the kernel variable holding the return
> + * value.
> + * Return 0 if permission is granted (pass-through); return 1
> + * (and set *errorp to the appropriate value, if the LSM is
> + * overriding default behavior.
> * @task_reparent_to_init:
> * Set the security attributes in @p->security for a kernel thread that
> * is being reparented to the init task.
> @@ -1304,7 +1309,7 @@ struct security_operations {
> int (*task_wait) (struct task_struct * p);
> int (*task_prctl) (int option, unsigned long arg2,
> unsigned long arg3, unsigned long arg4,
> - unsigned long arg5);
> + unsigned long arg5, long *errorp);
> void (*task_reparent_to_init) (struct task_struct * p);
> void (*task_to_inode)(struct task_struct *p, struct inode *inode);
>
> @@ -1550,7 +1555,8 @@ int security_task_kill(struct task_struct *p, struct siginfo *info,
> int sig, u32 secid);
> int security_task_wait(struct task_struct *p);
> int security_task_prctl(int option, unsigned long arg2, unsigned long arg3,
> - unsigned long arg4, unsigned long arg5);
> + unsigned long arg4, unsigned long arg5,
> + long *errorp);
> void security_task_reparent_to_init(struct task_struct *p);
> void security_task_to_inode(struct task_struct *p, struct inode *inode);
> int security_ipc_permission(struct kern_ipc_perm *ipcp, short flag);
> @@ -2096,7 +2102,8 @@ static inline int security_task_wait (struct task_struct *p)
> static inline int security_task_prctl (int option, unsigned long arg2,
> unsigned long arg3,
> unsigned long arg4,
> - unsigned long arg5)
> + unsigned long arg5,
> + long *errorp)
> {
> return 0;
> }
> diff --git a/kernel/sys.c b/kernel/sys.c
> index c7c4fa4..42f26fd 100644
> --- a/kernel/sys.c
> +++ b/kernel/sys.c
> @@ -1639,8 +1639,7 @@ asmlinkage long sys_prctl(int option, unsigned long arg2, unsigned long arg3,
> {
> long error;
>
> - error = security_task_prctl(option, arg2, arg3, arg4, arg5);
> - if (error)
> + if (security_task_prctl(option, arg2, arg3, arg4, arg5, &error))
> return error;
>
> switch (option) {
> @@ -1693,17 +1692,6 @@ asmlinkage long sys_prctl(int option, unsigned long arg2, unsigned long arg3,
> error = -EINVAL;
> break;
>
> - case PR_GET_KEEPCAPS:
> - if (current->keep_capabilities)
> - error = 1;
> - break;
> - case PR_SET_KEEPCAPS:
> - if (arg2 != 0 && arg2 != 1) {
> - error = -EINVAL;
> - break;
> - }
> - current->keep_capabilities = arg2;
> - break;
> case PR_SET_NAME: {
> struct task_struct *me = current;
> unsigned char ncomm[sizeof(me->comm)];
> diff --git a/security/capability.c b/security/capability.c
> index 9e99f36..8340655 100644
> --- a/security/capability.c
> +++ b/security/capability.c
> @@ -45,6 +45,7 @@ static struct security_operations capability_ops = {
> .task_setioprio = cap_task_setioprio,
> .task_setnice = cap_task_setnice,
> .task_post_setuid = cap_task_post_setuid,
> + .task_prctl = cap_task_prctl,
> .task_reparent_to_init = cap_task_reparent_to_init,
>
> .syslog = cap_syslog,
> diff --git a/security/commoncap.c b/security/commoncap.c
> index d65ddd3..44bf3fc 100644
> --- a/security/commoncap.c
> +++ b/security/commoncap.c
> @@ -24,6 +24,7 @@
> #include <linux/hugetlb.h>
> #include <linux/mount.h>
> #include <linux/sched.h>
> +#include <linux/prctl.h>
>
> #ifdef CONFIG_SECURITY_FILE_CAPABILITIES
> /*
> @@ -39,11 +40,6 @@
> kernel_cap_t cap_bset = CAP_INIT_BSET; /* systemwide capability bound */
> EXPORT_SYMBOL(cap_bset);
>
> -/* Global security state */
> -
> -unsigned securebits = SECUREBITS_DEFAULT; /* systemwide security settings */
> -EXPORT_SYMBOL(securebits);
> -
> int cap_netlink_send(struct sock *sk, struct sk_buff *skb)
> {
> NETLINK_CB(skb).eff_cap = current->cap_effective;
> @@ -164,6 +160,27 @@ static inline void bprm_clear_caps(struct linux_binprm *bprm)
> bprm->cap_effective = false;
> }
>
> +static inline void bprm_force_uid0_caps(struct linux_binprm *bprm)
> +{
> + /*
> + * To support inheritance of root-permissions and suid-root
> + * executables under compatibility mode, we raise all three
> + * capability sets for the file.
> + *
> + * If only the real uid is 0, we only raise the inheritable
> + * and permitted sets of the executable file.
> + */
> + if (bprm->e_uid == 0 || current->uid == 0) {
> + cap_set_full (bprm->cap_inheritable);
> + cap_set_full (bprm->cap_permitted);
> + }
> + if (bprm->e_uid == 0) {
> + bprm->cap_effective = true;
> + }
> +
> + return;
> +}
> +
> #ifdef CONFIG_SECURITY_FILE_CAPABILITIES
>
> int cap_inode_need_killpriv(struct dentry *dentry)
> @@ -215,7 +232,7 @@ static inline int cap_from_disk(__le32 *caps, struct linux_binprm *bprm,
> }
>
> /* Locate any VFS capabilities: */
> -static int get_file_caps(struct linux_binprm *bprm)
> +int cap_bprm_set_security(struct linux_binprm *bprm)
> {
> struct dentry *dentry;
> int rc = 0;
> @@ -223,8 +240,7 @@ static int get_file_caps(struct linux_binprm *bprm)
> struct inode *inode;
>
> if (bprm->file->f_vfsmnt->mnt_flags & MNT_NOSUID) {
> - bprm_clear_caps(bprm);
> - return 0;
> + goto out_no_dentry;
> }
>
> dentry = dget(bprm->file->f_dentry);
> @@ -249,8 +265,16 @@ static int get_file_caps(struct linux_binprm *bprm)
>
> out:
> dput(dentry);
> - if (rc)
> +
> + if (rc) {
> +out_no_dentry:
> bprm_clear_caps(bprm);
> + }
> +
> + if (! current->keep_capabilities) {
> + bprm_force_uid0_caps(bprm);
> + rc = 0;
> + }
what about a process tree wanting to maintain the current
behavior - that is !SECURE_NOROOT, but able to keep_caps
across setuid, then regain full privs on executing a setuid
binary? That is no longer possible when file capabilities
are enabled. I think it should be, given just how long that
was the expected way to use capabilities.
Of course that means keep_capabilities can't be multiplexed
like this. But that really doesn't seem like a big loss.
Trying to be too clever probably means we'll get it wrong,
and heck, the name is completely wrong in this sense :)
To summarize more clearly, I think that so long as we support
process trees with a sort of !SECURE_NOROOT support, that
support should include the ability to use prctl(KEEP_CAPS) the
way one uses it now.
When a process tree is in strict capability mode,
prctl(PR_{G,S}ET_KEEP_CAPS) should return -EINVAL.
quoted text >
> return rc;
> }
> @@ -266,42 +290,15 @@ int cap_inode_killpriv(struct dentry *dentry)
> return 0;
> }
>
> -static inline int get_file_caps(struct linux_binprm *bprm)
> +int cap_bprm_set_security(struct linux_binprm *bprm)
> {
> bprm_clear_caps(bprm);
> + bprm_force_uid0_caps(bprm);
> + current->keep_capabilities = 0;
This is being moved from bprm_apply to bprm_set, which moves it
earlier. If exec fails later on, keep_capabilities might be set
to 0 even though exec failed.
quoted text > return 0;
> }
> #endif
>
> -int cap_bprm_set_security (struct linux_binprm *bprm)
> -{
> - int ret;
> -
> - ret = get_file_caps(bprm);
> - if (ret)
> - printk(KERN_NOTICE "%s: get_file_caps returned %d for %s\n",
> - __FUNCTION__, ret, bprm->filename);
> -
> - /* To support inheritance of root-permissions and suid-root
> - * executables under compatibility mode, we raise all three
> - * capability sets for the file.
> - *
> - * If only the real uid is 0, we only raise the inheritable
> - * and permitted sets of the executable file.
> - */
> -
> - if (!issecure (SECURE_NOROOT)) {
> - if (bprm->e_uid == 0 || current->uid == 0) {
> - cap_set_full (bprm->cap_inheritable);
> - cap_set_full (bprm->cap_permitted);
> - }
> - if (bprm->e_uid == 0)
> - bprm->cap_effective = true;
> - }
> -
> - return ret;
> -}
> -
> void cap_bprm_apply_creds (struct linux_binprm *bprm, int unsafe)
> {
> /* Derived from fs/exec.c:compute_creds. */
> @@ -341,8 +338,6 @@ void cap_bprm_apply_creds (struct linux_binprm *bprm, int unsafe)
> }
>
> /* AUD: Audit candidate if current->cap_effective is set */
> -
> - current->keep_capabilities = 0;
> }
>
> int cap_bprm_secureexec (struct linux_binprm *bprm)
> @@ -441,8 +436,7 @@ int cap_task_post_setuid (uid_t old_ruid, uid_t old_euid, uid_t old_suid,
> case LSM_SETID_RE:
> case LSM_SETID_ID:
> case LSM_SETID_RES:
> - /* Copied from kernel/sys.c:setreuid/setuid/setresuid. */
> - if (!issecure (SECURE_NO_SETUID_FIXUP)) {
> + if (! current->keep_capabilities) {
> cap_emulate_setxuid (old_ruid, old_euid, old_suid);
> }
> break;
> @@ -457,7 +451,7 @@ int cap_task_post_setuid (uid_t old_ruid, uid_t old_euid, uid_t old_suid,
> * if not, we might be a bit too harsh here.
> */
>
> - if (!issecure (SECURE_NO_SETUID_FIXUP)) {
> + if (! current->keep_capabilities) {
> if (old_fsuid == 0 && current->fsuid != 0) {
> cap_t (current->cap_effective) &=
> ~CAP_FS_MASK;
> @@ -579,3 +573,53 @@ int cap_vm_enough_memory(struct mm_struct *mm, long pages)
> return __vm_enough_memory(mm, pages, cap_sys_admin);
> }
>
> +int cap_task_prctl(int option, unsigned long arg2, unsigned long arg3,
> + unsigned long arg4, unsigned long arg5, long *errorp)
> +{
> + switch (option) {
> +
> +#ifdef CONFIG_SECURITY_FILE_CAPABILITIES
> +
> + case PR_GET_CAPONLY:
> + *errorp = current->keep_capabilities;
> + break;
> +
> + case PR_SET_CAPONLY:
> + /*
> + * Only permit setting to 1
> + */
> + if ((arg2 != 1) || !cap_capable(current, CAP_SETPCAP)) {
> + *errorp = -EPERM;
> + } else {
> + /*
> + * Once locked, no unlocking for this process
> + * and its children
> + */
> + current->keep_capabilities = 1;
> + *errorp = 0;
> + }
> + break;
> +
> +#else /* ie. ndef CONFIG_SECURITY_FILE_CAPABILITIES */
> +
> + case PR_GET_KEEPCAPS:
> + if (current->keep_capabilities)
> + *errorp = 1;
> + break;
> +
> + case PR_SET_KEEPCAPS:
> + if (arg2 != 0 && arg2 != 1) {
> + *errorp = -EINVAL;
> + break;
> + }
> + current->keep_capabilities = arg2;
> + break;
> +
> +#endif /* def CONFIG_SECURITY_FILE_CAPABILITIES */
> +
> + default:
> + return 0; /* pass-through */
> + }
> +
> + return 1; /* overridden by capability LSM */
> +}
> diff --git a/security/dummy.c b/security/dummy.c
> index 88bb1bc..0497dc8 100644
> --- a/security/dummy.c
> +++ b/security/dummy.c
> @@ -566,8 +566,9 @@ static int dummy_task_kill (struct task_struct *p, struct siginfo *info,
> return 0;
> }
>
> -static int dummy_task_prctl (int option, unsigned long arg2, unsigned long arg3,
> - unsigned long arg4, unsigned long arg5)
> +static int dummy_task_prctl (int option, unsigned long arg2,
> + unsigned long arg3, unsigned long arg4,
> + unsigned long arg5, long *errorp)
> {
> return 0;
> }
> diff --git a/security/security.c b/security/security.c
> index e6d53b3..e3a4285 100644
> --- a/security/security.c
> +++ b/security/security.c
> @@ -665,9 +665,10 @@ int security_task_wait(struct task_struct *p)
> }
>
> int security_task_prctl(int option, unsigned long arg2, unsigned long arg3,
> - unsigned long arg4, unsigned long arg5)
> + unsigned long arg4, unsigned long arg5, long *errorp)
> {
> - return security_ops->task_prctl(option, arg2, arg3, arg4, arg5);
> + return security_ops->task_prctl(option, arg2, arg3, arg4, arg5,
> + errorp);
> }
>
> void security_task_reparent_to_init(struct task_struct *p)
> diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c
> index dc196b9..c8aac50 100644
> --- a/security/selinux/hooks.c
> +++ b/security/selinux/hooks.c
> @@ -2913,12 +2913,14 @@ static int selinux_task_prctl(int option,
> unsigned long arg2,
> unsigned long arg3,
> unsigned long arg4,
> - unsigned long arg5)
> + unsigned long arg5,
> + long *errorp)
> {
> /* The current prctl operations do not appear to require
> any SELinux controls since they merely observe or modify
> the state of the current process. */
> - return 0;
> + return secondary_ops->task_prctl(option, arg2, arg3, arg4, arg5,
> + errorp);
> }
>
> static int selinux_task_wait(struct task_struct *p)
> --
> 1.5.1.3
>
-
unsubscribe notice To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to
majordomo@vger.kernel.org
More majordomo info at
http://vger.kernel.org/majordomo-info.html
Please read the FAQ at
http://www.tux.org/lkml/ Messages in current thread:
Re: [2.6 patch] remove securebits , Serge E. Hallyn , (Tue Aug 28, 2:19 pm)
