#include <linux/module.h>
#include <linux/security.h>
#include <linux/syscalls.h>
+#include <linux/pid_namespace.h>
#include <asm/uaccess.h>
/*
*/
static DEFINE_SPINLOCK(task_capability_lock);
+/*
+ * Leveraged for setting/resetting capabilities
+ */
+
+const kernel_cap_t __cap_empty_set = CAP_EMPTY_SET;
+const kernel_cap_t __cap_full_set = CAP_FULL_SET;
+const kernel_cap_t __cap_init_eff_set = CAP_INIT_EFF_SET;
+
+EXPORT_SYMBOL(__cap_empty_set);
+EXPORT_SYMBOL(__cap_full_set);
+EXPORT_SYMBOL(__cap_init_eff_set);
+
+/*
+ * More recent versions of libcap are available from:
+ *
+ * http://www.kernel.org/pub/linux/libs/security/linux-privs/
+ */
+
+static void warn_legacy_capability_use(void)
+{
+ static int warned;
+ if (!warned) {
+ char name[sizeof(current->comm)];
+
+ printk(KERN_INFO "warning: `%s' uses 32-bit capabilities"
+ " (legacy support in use)\n",
+ get_task_comm(name, current));
+ warned = 1;
+ }
+}
+
+/*
+ * Version 2 capabilities worked fine, but the linux/capability.h file
+ * that accompanied their introduction encouraged their use without
+ * the necessary user-space source code changes. As such, we have
+ * created a version 3 with equivalent functionality to version 2, but
+ * with a header change to protect legacy source code from using
+ * version 2 when it wanted to use version 1. If your system has code
+ * that trips the following warning, it is using version 2 specific
+ * capabilities and may be doing so insecurely.
+ *
+ * The remedy is to either upgrade your version of libcap (to 2.10+,
+ * if the application is linked against it), or recompile your
+ * application with modern kernel headers and this warning will go
+ * away.
+ */
+
+static void warn_deprecated_v2(void)
+{
+ static int warned;
+
+ if (!warned) {
+ char name[sizeof(current->comm)];
+
+ printk(KERN_INFO "warning: `%s' uses deprecated v2"
+ " capabilities in a way that may be insecure.\n",
+ get_task_comm(name, current));
+ warned = 1;
+ }
+}
+
+/*
+ * Version check. Return the number of u32s in each capability flag
+ * array, or a negative value on error.
+ */
+static int cap_validate_magic(cap_user_header_t header, unsigned *tocopy)
+{
+ __u32 version;
+
+ if (get_user(version, &header->version))
+ return -EFAULT;
+
+ switch (version) {
+ case _LINUX_CAPABILITY_VERSION_1:
+ warn_legacy_capability_use();
+ *tocopy = _LINUX_CAPABILITY_U32S_1;
+ break;
+ case _LINUX_CAPABILITY_VERSION_2:
+ warn_deprecated_v2();
+ /*
+ * fall through - v3 is otherwise equivalent to v2.
+ */
+ case _LINUX_CAPABILITY_VERSION_3:
+ *tocopy = _LINUX_CAPABILITY_U32S_3;
+ break;
+ default:
+ if (put_user((u32)_KERNEL_CAPABILITY_VERSION, &header->version))
+ return -EFAULT;
+ return -EINVAL;
+ }
+
+ return 0;
+}
+
/*
* For sys_getproccap() and sys_setproccap(), any of the three
* capability set pointers may be NULL -- indicating that that set is
{
int ret = 0;
pid_t pid;
- __u32 version;
struct task_struct *target;
- struct __user_cap_data_struct data;
-
- if (get_user(version, &header->version))
- return -EFAULT;
+ unsigned tocopy;
+ kernel_cap_t pE, pI, pP;
- if (version != _LINUX_CAPABILITY_VERSION) {
- if (put_user(_LINUX_CAPABILITY_VERSION, &header->version))
- return -EFAULT;
- return -EINVAL;
- }
+ ret = cap_validate_magic(header, &tocopy);
+ if (ret != 0)
+ return ret;
if (get_user(pid, &header->pid))
return -EFAULT;
spin_lock(&task_capability_lock);
read_lock(&tasklist_lock);
- if (pid && pid != current->pid) {
- target = find_task_by_pid(pid);
+ if (pid && pid != task_pid_vnr(current)) {
+ target = find_task_by_vpid(pid);
if (!target) {
ret = -ESRCH;
goto out;
} else
target = current;
- ret = security_capget(target, &data.effective, &data.inheritable, &data.permitted);
+ ret = security_capget(target, &pE, &pI, &pP);
out:
read_unlock(&tasklist_lock);
spin_unlock(&task_capability_lock);
- if (!ret && copy_to_user(dataptr, &data, sizeof data))
- return -EFAULT;
+ if (!ret) {
+ struct __user_cap_data_struct kdata[_KERNEL_CAPABILITY_U32S];
+ unsigned i;
+
+ for (i = 0; i < tocopy; i++) {
+ kdata[i].effective = pE.cap[i];
+ kdata[i].permitted = pP.cap[i];
+ kdata[i].inheritable = pI.cap[i];
+ }
+
+ /*
+ * Note, in the case, tocopy < _KERNEL_CAPABILITY_U32S,
+ * we silently drop the upper capabilities here. This
+ * has the effect of making older libcap
+ * implementations implicitly drop upper capability
+ * bits when they perform a: capget/modify/capset
+ * sequence.
+ *
+ * This behavior is considered fail-safe
+ * behavior. Upgrading the application to a newer
+ * version of libcap will enable access to the newer
+ * capabilities.
+ *
+ * An alternative would be to return an error here
+ * (-ERANGE), but that causes legacy applications to
+ * unexpectidly fail; the capget/modify/capset aborts
+ * before modification is attempted and the application
+ * fails.
+ */
+
+ if (copy_to_user(dataptr, kdata, tocopy
+ * sizeof(struct __user_cap_data_struct))) {
+ return -EFAULT;
+ }
+ }
return ret;
}
int found = 0;
struct pid *pgrp;
- pgrp = find_pid(pgrp_nr);
+ pgrp = find_vpid(pgrp_nr);
do_each_pid_task(pgrp, PIDTYPE_PGID, g) {
target = g;
while_each_thread(g, target) {
int found = 0;
do_each_thread(g, target) {
- if (target == current || is_init(target))
+ if (target == current || is_container_init(target->group_leader))
continue;
found = 1;
if (security_capset_check(target, effective, inheritable,
*/
asmlinkage long sys_capset(cap_user_header_t header, const cap_user_data_t data)
{
+ struct __user_cap_data_struct kdata[_KERNEL_CAPABILITY_U32S];
+ unsigned i, tocopy;
kernel_cap_t inheritable, permitted, effective;
- __u32 version;
struct task_struct *target;
int ret;
pid_t pid;
- if (get_user(version, &header->version))
- return -EFAULT;
-
- if (version != _LINUX_CAPABILITY_VERSION) {
- if (put_user(_LINUX_CAPABILITY_VERSION, &header->version))
- return -EFAULT;
- return -EINVAL;
- }
+ ret = cap_validate_magic(header, &tocopy);
+ if (ret != 0)
+ return ret;
if (get_user(pid, &header->pid))
return -EFAULT;
- if (pid && pid != current->pid && !capable(CAP_SETPCAP))
+ if (pid && pid != task_pid_vnr(current) && !capable(CAP_SETPCAP))
return -EPERM;
- if (copy_from_user(&effective, &data->effective, sizeof(effective)) ||
- copy_from_user(&inheritable, &data->inheritable, sizeof(inheritable)) ||
- copy_from_user(&permitted, &data->permitted, sizeof(permitted)))
+ if (copy_from_user(&kdata, data, tocopy
+ * sizeof(struct __user_cap_data_struct))) {
return -EFAULT;
+ }
+
+ for (i = 0; i < tocopy; i++) {
+ effective.cap[i] = kdata[i].effective;
+ permitted.cap[i] = kdata[i].permitted;
+ inheritable.cap[i] = kdata[i].inheritable;
+ }
+ while (i < _KERNEL_CAPABILITY_U32S) {
+ effective.cap[i] = 0;
+ permitted.cap[i] = 0;
+ inheritable.cap[i] = 0;
+ i++;
+ }
spin_lock(&task_capability_lock);
read_lock(&tasklist_lock);
- if (pid > 0 && pid != current->pid) {
- target = find_task_by_pid(pid);
+ if (pid > 0 && pid != task_pid_vnr(current)) {
+ target = find_task_by_vpid(pid);
if (!target) {
ret = -ESRCH;
goto out;