[linux.git] / kernel / seccomp.c

/*
 * linux/kernel/seccomp.c
 *
 * Copyright 2004-2005  Andrea Arcangeli <[email protected]>
 *
 * Copyright (C) 2012 Google, Inc.
 * Will Drewry <[email protected]>
 *
 * This defines a simple but solid secure-computing facility.
 *
 * Mode 1 uses a fixed list of allowed system calls.
 * Mode 2 allows user-defined system call filters in the form
 *        of Berkeley Packet Filters/Linux Socket Filters.
 */

#include <linux/atomic.h>
#include <linux/audit.h>
#include <linux/compat.h>
#include <linux/sched.h>
#include <linux/seccomp.h>

/* #define SECCOMP_DEBUG 1 */

#ifdef CONFIG_SECCOMP_FILTER
#include <asm/syscall.h>
#include <linux/filter.h>
#include <linux/ptrace.h>
#include <linux/security.h>
#include <linux/slab.h>
#include <linux/tracehook.h>
#include <linux/uaccess.h>

/**
 * struct seccomp_filter - container for seccomp BPF programs
 *
 * @usage: reference count to manage the object lifetime.
 *         get/put helpers should be used when accessing an instance
 *         outside of a lifetime-guarded section.  In general, this
 *         is only needed for handling filters shared across tasks.
 * @prev: points to a previously installed, or inherited, filter
 * @len: the number of instructions in the program
 * @insns: the BPF program instructions to evaluate
 *
 * seccomp_filter objects are organized in a tree linked via the @prev
 * pointer.  For any task, it appears to be a singly-linked list starting
 * with current->seccomp.filter, the most recently attached or inherited filter.
 * However, multiple filters may share a @prev node, by way of fork(), which
 * results in a unidirectional tree existing in memory.  This is similar to
 * how namespaces work.
 *
 * seccomp_filter objects should never be modified after being attached
 * to a task_struct (other than @usage).
 */
struct seccomp_filter {
	atomic_t usage;
	struct seccomp_filter *prev;
	unsigned short len;  /* Instruction count */
	struct sock_filter insns[];
};

/* Limit any path through the tree to 256KB worth of instructions. */
#define MAX_INSNS_PER_PATH ((1 << 18) / sizeof(struct sock_filter))

/**
 * get_u32 - returns a u32 offset into data
 * @data: a unsigned 64 bit value
 * @index: 0 or 1 to return the first or second 32-bits
 *
 * This inline exists to hide the length of unsigned long.  If a 32-bit
 * unsigned long is passed in, it will be extended and the top 32-bits will be
 * 0. If it is a 64-bit unsigned long, then whatever data is resident will be
 * properly returned.
 *
 * Endianness is explicitly ignored and left for BPF program authors to manage
 * as per the specific architecture.
 */
static inline u32 get_u32(u64 data, int index)
{
	return ((u32 *)&data)[index];
}

/* Helper for bpf_load below. */
#define BPF_DATA(_name) offsetof(struct seccomp_data, _name)
/**
 * bpf_load: checks and returns a pointer to the requested offset
 * @off: offset into struct seccomp_data to load from
 *
 * Returns the requested 32-bits of data.
 * seccomp_check_filter() should assure that @off is 32-bit aligned
 * and not out of bounds.  Failure to do so is a BUG.
 */
u32 seccomp_bpf_load(int off)
{
	struct pt_regs *regs = task_pt_regs(current);
	if (off == BPF_DATA(nr))
		return syscall_get_nr(current, regs);
	if (off == BPF_DATA(arch))
		return syscall_get_arch(current, regs);
	if (off >= BPF_DATA(args[0]) && off < BPF_DATA(args[6])) {
		unsigned long value;
		int arg = (off - BPF_DATA(args[0])) / sizeof(u64);
		int index = !!(off % sizeof(u64));
		syscall_get_arguments(current, regs, arg, 1, &value);
		return get_u32(value, index);
	}
	if (off == BPF_DATA(instruction_pointer))
		return get_u32(KSTK_EIP(current), 0);
	if (off == BPF_DATA(instruction_pointer) + sizeof(u32))
		return get_u32(KSTK_EIP(current), 1);
	/* seccomp_check_filter should make this impossible. */
	BUG();
}

/**
 *	seccomp_check_filter - verify seccomp filter code
 *	@filter: filter to verify
 *	@flen: length of filter
 *
 * Takes a previously checked filter (by sk_chk_filter) and
 * redirects all filter code that loads struct sk_buff data
 * and related data through seccomp_bpf_load.  It also
 * enforces length and alignment checking of those loads.
 *
 * Returns 0 if the rule set is legal or -EINVAL if not.
 */
static int seccomp_check_filter(struct sock_filter *filter, unsigned int flen)
{
	int pc;
	for (pc = 0; pc < flen; pc++) {
		struct sock_filter *ftest = &filter[pc];
		u16 code = ftest->code;
		u32 k = ftest->k;

		switch (code) {
		case BPF_S_LD_W_ABS:
			ftest->code = BPF_S_ANC_SECCOMP_LD_W;
			/* 32-bit aligned and not out of bounds. */
			if (k >= sizeof(struct seccomp_data) || k & 3)
				return -EINVAL;
			continue;
		case BPF_S_LD_W_LEN:
			ftest->code = BPF_S_LD_IMM;
			ftest->k = sizeof(struct seccomp_data);
			continue;
		case BPF_S_LDX_W_LEN:
			ftest->code = BPF_S_LDX_IMM;
			ftest->k = sizeof(struct seccomp_data);
			continue;
		/* Explicitly include allowed calls. */
		case BPF_S_RET_K:
		case BPF_S_RET_A:
		case BPF_S_ALU_ADD_K:
		case BPF_S_ALU_ADD_X:
		case BPF_S_ALU_SUB_K:
		case BPF_S_ALU_SUB_X:
		case BPF_S_ALU_MUL_K:
		case BPF_S_ALU_MUL_X:
		case BPF_S_ALU_DIV_X:
		case BPF_S_ALU_AND_K:
		case BPF_S_ALU_AND_X:
		case BPF_S_ALU_OR_K:
		case BPF_S_ALU_OR_X:
		case BPF_S_ALU_LSH_K:
		case BPF_S_ALU_LSH_X:
		case BPF_S_ALU_RSH_K:
		case BPF_S_ALU_RSH_X:
		case BPF_S_ALU_NEG:
		case BPF_S_LD_IMM:
		case BPF_S_LDX_IMM:
		case BPF_S_MISC_TAX:
		case BPF_S_MISC_TXA:
		case BPF_S_ALU_DIV_K:
		case BPF_S_LD_MEM:
		case BPF_S_LDX_MEM:
		case BPF_S_ST:
		case BPF_S_STX:
		case BPF_S_JMP_JA:
		case BPF_S_JMP_JEQ_K:
		case BPF_S_JMP_JEQ_X:
		case BPF_S_JMP_JGE_K:
		case BPF_S_JMP_JGE_X:
		case BPF_S_JMP_JGT_K:
		case BPF_S_JMP_JGT_X:
		case BPF_S_JMP_JSET_K:
		case BPF_S_JMP_JSET_X:
			continue;
		default:
			return -EINVAL;
		}
	}
	return 0;
}

/**
 * seccomp_run_filters - evaluates all seccomp filters against @syscall
 * @syscall: number of the current system call
 *
 * Returns valid seccomp BPF response codes.
 */
static u32 seccomp_run_filters(int syscall)
{
	struct seccomp_filter *f;
	u32 ret = SECCOMP_RET_ALLOW;

	/* Ensure unexpected behavior doesn't result in failing open. */
	if (WARN_ON(current->seccomp.filter == NULL))
		return SECCOMP_RET_KILL;

	/*
	 * All filters in the list are evaluated and the lowest BPF return
	 * value always takes priority (ignoring the DATA).
	 */
	for (f = current->seccomp.filter; f; f = f->prev) {
		u32 cur_ret = sk_run_filter(NULL, f->insns);
		if ((cur_ret & SECCOMP_RET_ACTION) < (ret & SECCOMP_RET_ACTION))
			ret = cur_ret;
	}
	return ret;
}

/**
 * seccomp_attach_filter: Attaches a seccomp filter to current.
 * @fprog: BPF program to install
 *
 * Returns 0 on success or an errno on failure.
 */
static long seccomp_attach_filter(struct sock_fprog *fprog)
{
	struct seccomp_filter *filter;
	unsigned long fp_size = fprog->len * sizeof(struct sock_filter);
	unsigned long total_insns = fprog->len;
	long ret;

	if (fprog->len == 0 || fprog->len > BPF_MAXINSNS)
		return -EINVAL;

	for (filter = current->seccomp.filter; filter; filter = filter->prev)
		total_insns += filter->len + 4;  /* include a 4 instr penalty */
	if (total_insns > MAX_INSNS_PER_PATH)
		return -ENOMEM;

	/*
	 * Installing a seccomp filter requires that the task have
	 * CAP_SYS_ADMIN in its namespace or be running with no_new_privs.
	 * This avoids scenarios where unprivileged tasks can affect the
	 * behavior of privileged children.
	 */
	if (!current->no_new_privs &&
	    security_capable_noaudit(current_cred(), current_user_ns(),
				     CAP_SYS_ADMIN) != 0)
		return -EACCES;

	/* Allocate a new seccomp_filter */
	filter = kzalloc(sizeof(struct seccomp_filter) + fp_size,
			 GFP_KERNEL|__GFP_NOWARN);
	if (!filter)
		return -ENOMEM;
	atomic_set(&filter->usage, 1);
	filter->len = fprog->len;

	/* Copy the instructions from fprog. */
	ret = -EFAULT;
	if (copy_from_user(filter->insns, fprog->filter, fp_size))
		goto fail;

	/* Check and rewrite the fprog via the skb checker */
	ret = sk_chk_filter(filter->insns, filter->len);
	if (ret)
		goto fail;

	/* Check and rewrite the fprog for seccomp use */
	ret = seccomp_check_filter(filter->insns, filter->len);
	if (ret)
		goto fail;

	/*
	 * If there is an existing filter, make it the prev and don't drop its
	 * task reference.
	 */
	filter->prev = current->seccomp.filter;
	current->seccomp.filter = filter;
	return 0;
fail:
	kfree(filter);
	return ret;
}

/**
 * seccomp_attach_user_filter - attaches a user-supplied sock_fprog
 * @user_filter: pointer to the user data containing a sock_fprog.
 *
 * Returns 0 on success and non-zero otherwise.
 */
long seccomp_attach_user_filter(char __user *user_filter)
{
	struct sock_fprog fprog;
	long ret = -EFAULT;

#ifdef CONFIG_COMPAT
	if (is_compat_task()) {
		struct compat_sock_fprog fprog32;
		if (copy_from_user(&fprog32, user_filter, sizeof(fprog32)))
			goto out;
		fprog.len = fprog32.len;
		fprog.filter = compat_ptr(fprog32.filter);
	} else /* falls through to the if below. */
#endif
	if (copy_from_user(&fprog, user_filter, sizeof(fprog)))
		goto out;
	ret = seccomp_attach_filter(&fprog);
out:
	return ret;
}

/* get_seccomp_filter - increments the reference count of the filter on @tsk */
void get_seccomp_filter(struct task_struct *tsk)
{
	struct seccomp_filter *orig = tsk->seccomp.filter;
	if (!orig)
		return;
	/* Reference count is bounded by the number of total processes. */
	atomic_inc(&orig->usage);
}

/* put_seccomp_filter - decrements the ref count of tsk->seccomp.filter */
void put_seccomp_filter(struct task_struct *tsk)
{
	struct seccomp_filter *orig = tsk->seccomp.filter;
	/* Clean up single-reference branches iteratively. */
	while (orig && atomic_dec_and_test(&orig->usage)) {
		struct seccomp_filter *freeme = orig;
		orig = orig->prev;
		kfree(freeme);
	}
}

/**
 * seccomp_send_sigsys - signals the task to allow in-process syscall emulation
 * @syscall: syscall number to send to userland
 * @reason: filter-supplied reason code to send to userland (via si_errno)
 *
 * Forces a SIGSYS with a code of SYS_SECCOMP and related sigsys info.
 */
static void seccomp_send_sigsys(int syscall, int reason)
{
	struct siginfo info;
	memset(&info, 0, sizeof(info));
	info.si_signo = SIGSYS;
	info.si_code = SYS_SECCOMP;
	info.si_call_addr = (void __user *)KSTK_EIP(current);
	info.si_errno = reason;
	info.si_arch = syscall_get_arch(current, task_pt_regs(current));
	info.si_syscall = syscall;
	force_sig_info(SIGSYS, &info, current);
}
#endif	/* CONFIG_SECCOMP_FILTER */

/*
 * Secure computing mode 1 allows only read/write/exit/sigreturn.
 * To be fully secure this must be combined with rlimit
 * to limit the stack allocations too.
 */
static int mode1_syscalls[] = {
	__NR_seccomp_read, __NR_seccomp_write, __NR_seccomp_exit, __NR_seccomp_sigreturn,
	0, /* null terminated */
};

#ifdef CONFIG_COMPAT
static int mode1_syscalls_32[] = {
	__NR_seccomp_read_32, __NR_seccomp_write_32, __NR_seccomp_exit_32, __NR_seccomp_sigreturn_32,
	0, /* null terminated */
};
#endif

int __secure_computing(int this_syscall)
{
	int mode = current->seccomp.mode;
	int exit_sig = 0;
	int *syscall;
	u32 ret;

	switch (mode) {
	case SECCOMP_MODE_STRICT:
		syscall = mode1_syscalls;
#ifdef CONFIG_COMPAT
		if (is_compat_task())
			syscall = mode1_syscalls_32;
#endif
		do {
			if (*syscall == this_syscall)
				return 0;
		} while (*++syscall);
		exit_sig = SIGKILL;
		ret = SECCOMP_RET_KILL;
		break;
#ifdef CONFIG_SECCOMP_FILTER
	case SECCOMP_MODE_FILTER: {
		int data;
		ret = seccomp_run_filters(this_syscall);
		data = ret & SECCOMP_RET_DATA;
		ret &= SECCOMP_RET_ACTION;
		switch (ret) {
		case SECCOMP_RET_ERRNO:
			/* Set the low-order 16-bits as a errno. */
			syscall_set_return_value(current, task_pt_regs(current),
						 -data, 0);
			goto skip;
		case SECCOMP_RET_TRAP:
			/* Show the handler the original registers. */
			syscall_rollback(current, task_pt_regs(current));
			/* Let the filter pass back 16 bits of data. */
			seccomp_send_sigsys(this_syscall, data);
			goto skip;
		case SECCOMP_RET_TRACE:
			/* Skip these calls if there is no tracer. */
			if (!ptrace_event_enabled(current, PTRACE_EVENT_SECCOMP))
				goto skip;
			/* Allow the BPF to provide the event message */
			ptrace_event(PTRACE_EVENT_SECCOMP, data);
			/*
			 * The delivery of a fatal signal during event
			 * notification may silently skip tracer notification.
			 * Terminating the task now avoids executing a system
			 * call that may not be intended.
			 */
			if (fatal_signal_pending(current))
				break;
			return 0;
		case SECCOMP_RET_ALLOW:
			return 0;
		case SECCOMP_RET_KILL:
		default:
			break;
		}
		exit_sig = SIGSYS;
		break;
	}
#endif
	default:
		BUG();
	}

#ifdef SECCOMP_DEBUG
	dump_stack();
#endif
	audit_seccomp(this_syscall, exit_sig, ret);
	do_exit(exit_sig);
#ifdef CONFIG_SECCOMP_FILTER
skip:
	audit_seccomp(this_syscall, exit_sig, ret);
#endif
	return -1;
}

long prctl_get_seccomp(void)
{
	return current->seccomp.mode;
}

/**
 * prctl_set_seccomp: configures current->seccomp.mode
 * @seccomp_mode: requested mode to use
 * @filter: optional struct sock_fprog for use with SECCOMP_MODE_FILTER
 *
 * This function may be called repeatedly with a @seccomp_mode of
 * SECCOMP_MODE_FILTER to install additional filters.  Every filter
 * successfully installed will be evaluated (in reverse order) for each system
 * call the task makes.
 *
 * Once current->seccomp.mode is non-zero, it may not be changed.
 *
 * Returns 0 on success or -EINVAL on failure.
 */
long prctl_set_seccomp(unsigned long seccomp_mode, char __user *filter)
{
	long ret = -EINVAL;

	if (current->seccomp.mode &&
	    current->seccomp.mode != seccomp_mode)
		goto out;

	switch (seccomp_mode) {
	case SECCOMP_MODE_STRICT:
		ret = 0;
#ifdef TIF_NOTSC
		disable_TSC();
#endif
		break;
#ifdef CONFIG_SECCOMP_FILTER
	case SECCOMP_MODE_FILTER:
		ret = seccomp_attach_user_filter(filter);
		if (ret)
			goto out;
		break;
#endif
	default:
		goto out;
	}

	current->seccomp.mode = seccomp_mode;
	set_thread_flag(TIF_SECCOMP);
out:
	return ret;
}
Commit	Line	Data
1da177e4 LT	1	/*
	2	* linux/kernel/seccomp.c
	3	*
	4	* Copyright 2004-2005 Andrea Arcangeli <[email protected]>
	5	*
e2cfabdf WD	6	* Copyright (C) 2012 Google, Inc.
	7	* Will Drewry <[email protected]>
	8	*
	9	* This defines a simple but solid secure-computing facility.
	10	*
	11	* Mode 1 uses a fixed list of allowed system calls.
	12	* Mode 2 allows user-defined system call filters in the form
	13	* of Berkeley Packet Filters/Linux Socket Filters.
1da177e4 LT	14	*/
1da177e4 LT	15
e2cfabdf	16	#include <linux/atomic.h>
85e7bac3	17	#include <linux/audit.h>
5b101740	18	#include <linux/compat.h>
e2cfabdf WD	19	#include <linux/sched.h>
e2cfabdf WD	20	#include <linux/seccomp.h>
1da177e4 LT	21
1da177e4 LT	22	/* #define SECCOMP_DEBUG 1 */
e2cfabdf WD	23
	24	#ifdef CONFIG_SECCOMP_FILTER
	25	#include <asm/syscall.h>
	26	#include <linux/filter.h>
fb0fadf9	27	#include <linux/ptrace.h>
e2cfabdf WD	28	#include <linux/security.h>
	29	#include <linux/slab.h>
	30	#include <linux/tracehook.h>
	31	#include <linux/uaccess.h>
	32
	33	/**
	34	* struct seccomp_filter - container for seccomp BPF programs
	35	*
	36	* @usage: reference count to manage the object lifetime.
	37	* get/put helpers should be used when accessing an instance
	38	* outside of a lifetime-guarded section. In general, this
	39	* is only needed for handling filters shared across tasks.
	40	* @prev: points to a previously installed, or inherited, filter
	41	* @len: the number of instructions in the program
	42	* @insns: the BPF program instructions to evaluate
	43	*
	44	* seccomp_filter objects are organized in a tree linked via the @prev
	45	* pointer. For any task, it appears to be a singly-linked list starting
	46	* with current->seccomp.filter, the most recently attached or inherited filter.
	47	* However, multiple filters may share a @prev node, by way of fork(), which
	48	* results in a unidirectional tree existing in memory. This is similar to
	49	* how namespaces work.
	50	*
	51	* seccomp_filter objects should never be modified after being attached
	52	* to a task_struct (other than @usage).
	53	*/
	54	struct seccomp_filter {
	55	atomic_t usage;
	56	struct seccomp_filter *prev;
	57	unsigned short len; /* Instruction count */
	58	struct sock_filter insns[];
	59	};
	60
	61	/* Limit any path through the tree to 256KB worth of instructions. */
	62	#define MAX_INSNS_PER_PATH ((1 << 18) / sizeof(struct sock_filter))
	63
e2cfabdf WD	64	/**
	65	* get_u32 - returns a u32 offset into data
	66	* @data: a unsigned 64 bit value
	67	* @index: 0 or 1 to return the first or second 32-bits
	68	*
	69	* This inline exists to hide the length of unsigned long. If a 32-bit
	70	* unsigned long is passed in, it will be extended and the top 32-bits will be
	71	* 0. If it is a 64-bit unsigned long, then whatever data is resident will be
	72	* properly returned.
	73	*
	74	* Endianness is explicitly ignored and left for BPF program authors to manage
	75	* as per the specific architecture.
	76	*/
	77	static inline u32 get_u32(u64 data, int index)
	78	{
	79	return ((u32 *)&data)[index];
	80	}
	81
	82	/* Helper for bpf_load below. */
	83	#define BPF_DATA(_name) offsetof(struct seccomp_data, _name)
	84	/**
	85	* bpf_load: checks and returns a pointer to the requested offset
	86	* @off: offset into struct seccomp_data to load from
	87	*
	88	* Returns the requested 32-bits of data.
	89	* seccomp_check_filter() should assure that @off is 32-bit aligned
	90	* and not out of bounds. Failure to do so is a BUG.
	91	*/
	92	u32 seccomp_bpf_load(int off)
	93	{
	94	struct pt_regs *regs = task_pt_regs(current);
	95	if (off == BPF_DATA(nr))
	96	return syscall_get_nr(current, regs);
	97	if (off == BPF_DATA(arch))
	98	return syscall_get_arch(current, regs);
	99	if (off >= BPF_DATA(args[0]) && off < BPF_DATA(args[6])) {
	100	unsigned long value;
	101	int arg = (off - BPF_DATA(args[0])) / sizeof(u64);
	102	int index = !!(off % sizeof(u64));
	103	syscall_get_arguments(current, regs, arg, 1, &value);
	104	return get_u32(value, index);
	105	}
	106	if (off == BPF_DATA(instruction_pointer))
	107	return get_u32(KSTK_EIP(current), 0);
	108	if (off == BPF_DATA(instruction_pointer) + sizeof(u32))
	109	return get_u32(KSTK_EIP(current), 1);
	110	/* seccomp_check_filter should make this impossible. */
	111	BUG();
	112	}
	113
	114	/**
	115	* seccomp_check_filter - verify seccomp filter code
	116	* @filter: filter to verify
	117	* @flen: length of filter
	118	*
	119	* Takes a previously checked filter (by sk_chk_filter) and
	120	* redirects all filter code that loads struct sk_buff data
	121	* and related data through seccomp_bpf_load. It also
	122	* enforces length and alignment checking of those loads.
	123	*
	124	* Returns 0 if the rule set is legal or -EINVAL if not.
	125	*/
	126	static int seccomp_check_filter(struct sock_filter *filter, unsigned int flen)
	127	{
128	int pc;
129	for (pc = 0; pc < flen; pc++) {
130	struct sock_filter *ftest = &filter[pc];
131	u16 code = ftest->code;
132	u32 k = ftest->k;
133
134	switch (code) {
135	case BPF_S_LD_W_ABS:
136	ftest->code = BPF_S_ANC_SECCOMP_LD_W;
137	/* 32-bit aligned and not out of bounds. */
138	if (k >= sizeof(struct seccomp_data) \|\| k & 3)
139	return -EINVAL;
140	continue;
141	case BPF_S_LD_W_LEN:
142	ftest->code = BPF_S_LD_IMM;
143	ftest->k = sizeof(struct seccomp_data);
144	continue;
145	case BPF_S_LDX_W_LEN:
146	ftest->code = BPF_S_LDX_IMM;
147	ftest->k = sizeof(struct seccomp_data);
148	continue;
149	/* Explicitly include allowed calls. */
150	case BPF_S_RET_K:
151	case BPF_S_RET_A:
152	case BPF_S_ALU_ADD_K:
153	case BPF_S_ALU_ADD_X:
154	case BPF_S_ALU_SUB_K:
155	case BPF_S_ALU_SUB_X:
156	case BPF_S_ALU_MUL_K:
157	case BPF_S_ALU_MUL_X:
158	case BPF_S_ALU_DIV_X:
159	case BPF_S_ALU_AND_K:
160	case BPF_S_ALU_AND_X:
161	case BPF_S_ALU_OR_K:
162	case BPF_S_ALU_OR_X:
163	case BPF_S_ALU_LSH_K:
164	case BPF_S_ALU_LSH_X:
165	case BPF_S_ALU_RSH_K:
166	case BPF_S_ALU_RSH_X:
167	case BPF_S_ALU_NEG:
168	case BPF_S_LD_IMM:
169	case BPF_S_LDX_IMM:
170	case BPF_S_MISC_TAX:
171	case BPF_S_MISC_TXA:
172	case BPF_S_ALU_DIV_K:
173	case BPF_S_LD_MEM:
174	case BPF_S_LDX_MEM:
175	case BPF_S_ST:
176	case BPF_S_STX:
177	case BPF_S_JMP_JA:
178	case BPF_S_JMP_JEQ_K:
179	case BPF_S_JMP_JEQ_X:
180	case BPF_S_JMP_JGE_K:
181	case BPF_S_JMP_JGE_X:
182	case BPF_S_JMP_JGT_K:
183	case BPF_S_JMP_JGT_X:
184	case BPF_S_JMP_JSET_K:
185	case BPF_S_JMP_JSET_X:
186	continue;
187	default:
188	return -EINVAL;
189	}
190	}
191	return 0;
192	}
193
194	/**
195	* seccomp_run_filters - evaluates all seccomp filters against @syscall
196	* @syscall: number of the current system call
197	*
198	* Returns valid seccomp BPF response codes.
199	*/
200	static u32 seccomp_run_filters(int syscall)
201	{
202	struct seccomp_filter *f;
acf3b2c7 WD	203	u32 ret = SECCOMP_RET_ALLOW;
	204
	205	/* Ensure unexpected behavior doesn't result in failing open. */
	206	if (WARN_ON(current->seccomp.filter == NULL))
	207	return SECCOMP_RET_KILL;
	208
e2cfabdf WD	209	/*
e2cfabdf WD	210	* All filters in the list are evaluated and the lowest BPF return
acf3b2c7	211	* value always takes priority (ignoring the DATA).
e2cfabdf WD	212	*/
e2cfabdf WD	213	for (f = current->seccomp.filter; f; f = f->prev) {
acf3b2c7 WD	214	u32 cur_ret = sk_run_filter(NULL, f->insns);
	215	if ((cur_ret & SECCOMP_RET_ACTION) < (ret & SECCOMP_RET_ACTION))
	216	ret = cur_ret;
e2cfabdf WD	217	}
	218	return ret;
	219	}
	220
	221	/**
	222	* seccomp_attach_filter: Attaches a seccomp filter to current.
	223	* @fprog: BPF program to install
	224	*
	225	* Returns 0 on success or an errno on failure.
	226	*/
	227	static long seccomp_attach_filter(struct sock_fprog *fprog)
	228	{
	229	struct seccomp_filter *filter;
	230	unsigned long fp_size = fprog->len * sizeof(struct sock_filter);
	231	unsigned long total_insns = fprog->len;
	232	long ret;
	233
	234	if (fprog->len == 0 \|\| fprog->len > BPF_MAXINSNS)
	235	return -EINVAL;
	236
	237	for (filter = current->seccomp.filter; filter; filter = filter->prev)
	238	total_insns += filter->len + 4; /* include a 4 instr penalty */
	239	if (total_insns > MAX_INSNS_PER_PATH)
	240	return -ENOMEM;
	241
	242	/*
	243	* Installing a seccomp filter requires that the task have
	244	* CAP_SYS_ADMIN in its namespace or be running with no_new_privs.
	245	* This avoids scenarios where unprivileged tasks can affect the
	246	* behavior of privileged children.
	247	*/
	248	if (!current->no_new_privs &&
	249	security_capable_noaudit(current_cred(), current_user_ns(),
	250	CAP_SYS_ADMIN) != 0)
	251	return -EACCES;
	252
	253	/* Allocate a new seccomp_filter */
	254	filter = kzalloc(sizeof(struct seccomp_filter) + fp_size,
	255	GFP_KERNEL\|__GFP_NOWARN);
	256	if (!filter)
	257	return -ENOMEM;
	258	atomic_set(&filter->usage, 1);
	259	filter->len = fprog->len;
	260
	261	/* Copy the instructions from fprog. */
	262	ret = -EFAULT;
	263	if (copy_from_user(filter->insns, fprog->filter, fp_size))
	264	goto fail;
	265
	266	/* Check and rewrite the fprog via the skb checker */
	267	ret = sk_chk_filter(filter->insns, filter->len);
	268	if (ret)
	269	goto fail;
	270
	271	/* Check and rewrite the fprog for seccomp use */
	272	ret = seccomp_check_filter(filter->insns, filter->len);
	273	if (ret)
	274	goto fail;
	275
	276	/*
	277	* If there is an existing filter, make it the prev and don't drop its
	278	* task reference.
	279	*/
	280	filter->prev = current->seccomp.filter;
281	current->seccomp.filter = filter;
282	return 0;
283	fail:
284	kfree(filter);
285	return ret;
286	}
287
288	/**
289	* seccomp_attach_user_filter - attaches a user-supplied sock_fprog
290	* @user_filter: pointer to the user data containing a sock_fprog.
291	*
292	* Returns 0 on success and non-zero otherwise.
293	*/
294	long seccomp_attach_user_filter(char __user *user_filter)
295	{
296	struct sock_fprog fprog;
297	long ret = -EFAULT;
298
299	#ifdef CONFIG_COMPAT
300	if (is_compat_task()) {
301	struct compat_sock_fprog fprog32;
302	if (copy_from_user(&fprog32, user_filter, sizeof(fprog32)))
303	goto out;
304	fprog.len = fprog32.len;
305	fprog.filter = compat_ptr(fprog32.filter);
306	} else /* falls through to the if below. */
307	#endif
308	if (copy_from_user(&fprog, user_filter, sizeof(fprog)))
309	goto out;
310	ret = seccomp_attach_filter(&fprog);
311	out:
312	return ret;
313	}
314
315	/* get_seccomp_filter - increments the reference count of the filter on @tsk */
316	void get_seccomp_filter(struct task_struct *tsk)
317	{
318	struct seccomp_filter *orig = tsk->seccomp.filter;
319	if (!orig)
320	return;
321	/* Reference count is bounded by the number of total processes. */
322	atomic_inc(&orig->usage);
323	}
324
325	/* put_seccomp_filter - decrements the ref count of tsk->seccomp.filter */
326	void put_seccomp_filter(struct task_struct *tsk)
327	{
328	struct seccomp_filter *orig = tsk->seccomp.filter;
329	/* Clean up single-reference branches iteratively. */
330	while (orig && atomic_dec_and_test(&orig->usage)) {
331	struct seccomp_filter *freeme = orig;
332	orig = orig->prev;
333	kfree(freeme);
334	}
335	}
bb6ea430 WD	336
	337	/**
	338	* seccomp_send_sigsys - signals the task to allow in-process syscall emulation
	339	* @syscall: syscall number to send to userland
	340	* @reason: filter-supplied reason code to send to userland (via si_errno)
	341	*
	342	* Forces a SIGSYS with a code of SYS_SECCOMP and related sigsys info.
	343	*/
	344	static void seccomp_send_sigsys(int syscall, int reason)
	345	{
	346	struct siginfo info;
	347	memset(&info, 0, sizeof(info));
	348	info.si_signo = SIGSYS;
	349	info.si_code = SYS_SECCOMP;
	350	info.si_call_addr = (void __user *)KSTK_EIP(current);
	351	info.si_errno = reason;
	352	info.si_arch = syscall_get_arch(current, task_pt_regs(current));
	353	info.si_syscall = syscall;
	354	force_sig_info(SIGSYS, &info, current);
	355	}
e2cfabdf	356	#endif /* CONFIG_SECCOMP_FILTER */
1da177e4 LT	357
	358	/*
	359	* Secure computing mode 1 allows only read/write/exit/sigreturn.
	360	* To be fully secure this must be combined with rlimit
	361	* to limit the stack allocations too.
	362	*/
	363	static int mode1_syscalls[] = {
	364	__NR_seccomp_read, __NR_seccomp_write, __NR_seccomp_exit, __NR_seccomp_sigreturn,
	365	0, /* null terminated */
	366	};
	367
5b101740	368	#ifdef CONFIG_COMPAT
1da177e4 LT	369	static int mode1_syscalls_32[] = {
	370	__NR_seccomp_read_32, __NR_seccomp_write_32, __NR_seccomp_exit_32, __NR_seccomp_sigreturn_32,
	371	0, /* null terminated */
	372	};
	373	#endif
	374
acf3b2c7	375	int __secure_computing(int this_syscall)
1da177e4 LT	376	{
1da177e4 LT	377	int mode = current->seccomp.mode;
e2cfabdf WD	378	int exit_sig = 0;
e2cfabdf WD	379	int *syscall;
8156b451	380	u32 ret;
1da177e4 LT	381
1da177e4 LT	382	switch (mode) {
e2cfabdf	383	case SECCOMP_MODE_STRICT:
1da177e4	384	syscall = mode1_syscalls;
5b101740 RM	385	#ifdef CONFIG_COMPAT
5b101740 RM	386	if (is_compat_task())
1da177e4 LT	387	syscall = mode1_syscalls_32;
	388	#endif
	389	do {
	390	if (*syscall == this_syscall)
acf3b2c7	391	return 0;
1da177e4	392	} while (*++syscall);
e2cfabdf	393	exit_sig = SIGKILL;
8156b451	394	ret = SECCOMP_RET_KILL;
1da177e4	395	break;
e2cfabdf	396	#ifdef CONFIG_SECCOMP_FILTER
8156b451 WD	397	case SECCOMP_MODE_FILTER: {
8156b451 WD	398	int data;
acf3b2c7 WD	399	ret = seccomp_run_filters(this_syscall);
acf3b2c7 WD	400	data = ret & SECCOMP_RET_DATA;
8156b451 WD	401	ret &= SECCOMP_RET_ACTION;
8156b451 WD	402	switch (ret) {
acf3b2c7 WD	403	case SECCOMP_RET_ERRNO:
	404	/* Set the low-order 16-bits as a errno. */
	405	syscall_set_return_value(current, task_pt_regs(current),
	406	-data, 0);
	407	goto skip;
bb6ea430 WD	408	case SECCOMP_RET_TRAP:
	409	/* Show the handler the original registers. */
	410	syscall_rollback(current, task_pt_regs(current));
	411	/* Let the filter pass back 16 bits of data. */
	412	seccomp_send_sigsys(this_syscall, data);
	413	goto skip;
fb0fadf9 WD	414	case SECCOMP_RET_TRACE:
	415	/* Skip these calls if there is no tracer. */
	416	if (!ptrace_event_enabled(current, PTRACE_EVENT_SECCOMP))
	417	goto skip;
	418	/* Allow the BPF to provide the event message */
	419	ptrace_event(PTRACE_EVENT_SECCOMP, data);
	420	/*
	421	* The delivery of a fatal signal during event
	422	* notification may silently skip tracer notification.
	423	* Terminating the task now avoids executing a system
	424	* call that may not be intended.
	425	*/
	426	if (fatal_signal_pending(current))
	427	break;
	428	return 0;
acf3b2c7 WD	429	case SECCOMP_RET_ALLOW:
	430	return 0;
	431	case SECCOMP_RET_KILL:
	432	default:
	433	break;
	434	}
e2cfabdf WD	435	exit_sig = SIGSYS;
e2cfabdf WD	436	break;
8156b451	437	}
e2cfabdf	438	#endif
1da177e4 LT	439	default:
	440	BUG();
	441	}
	442
	443	#ifdef SECCOMP_DEBUG
	444	dump_stack();
	445	#endif
acf3b2c7	446	audit_seccomp(this_syscall, exit_sig, ret);
e2cfabdf	447	do_exit(exit_sig);
8156b451	448	#ifdef CONFIG_SECCOMP_FILTER
acf3b2c7 WD	449	skip:
acf3b2c7 WD	450	audit_seccomp(this_syscall, exit_sig, ret);
8156b451	451	#endif
acf3b2c7	452	return -1;
1da177e4	453	}
1d9d02fe AA	454
	455	long prctl_get_seccomp(void)
	456	{
	457	return current->seccomp.mode;
	458	}
	459
e2cfabdf WD	460	/**
	461	* prctl_set_seccomp: configures current->seccomp.mode
	462	* @seccomp_mode: requested mode to use
	463	* @filter: optional struct sock_fprog for use with SECCOMP_MODE_FILTER
	464	*
	465	* This function may be called repeatedly with a @seccomp_mode of
	466	* SECCOMP_MODE_FILTER to install additional filters. Every filter
	467	* successfully installed will be evaluated (in reverse order) for each system
	468	* call the task makes.
	469	*
	470	* Once current->seccomp.mode is non-zero, it may not be changed.
	471	*
	472	* Returns 0 on success or -EINVAL on failure.
	473	*/
	474	long prctl_set_seccomp(unsigned long seccomp_mode, char __user *filter)
1d9d02fe	475	{
e2cfabdf	476	long ret = -EINVAL;
1d9d02fe	477
e2cfabdf WD	478	if (current->seccomp.mode &&
e2cfabdf WD	479	current->seccomp.mode != seccomp_mode)
1d9d02fe AA	480	goto out;
1d9d02fe AA	481
e2cfabdf WD	482	switch (seccomp_mode) {
	483	case SECCOMP_MODE_STRICT:
	484	ret = 0;
cf99abac AA	485	#ifdef TIF_NOTSC
	486	disable_TSC();
	487	#endif
e2cfabdf WD	488	break;
	489	#ifdef CONFIG_SECCOMP_FILTER
	490	case SECCOMP_MODE_FILTER:
	491	ret = seccomp_attach_user_filter(filter);
	492	if (ret)
	493	goto out;
	494	break;
	495	#endif
	496	default:
	497	goto out;
1d9d02fe AA	498	}
1d9d02fe AA	499
e2cfabdf WD	500	current->seccomp.mode = seccomp_mode;
	501	set_thread_flag(TIF_SECCOMP);
	502	out:
1d9d02fe AA	503	return ret;
1d9d02fe AA	504	}