[deliverable/linux.git] / arch / x86 / entry / common.c

/*
 * common.c - C code for kernel entry and exit
 * Copyright (c) 2015 Andrew Lutomirski
 * GPL v2
 *
 * Based on asm and ptrace code by many authors.  The code here originated
 * in ptrace.c and signal.c.
 */

#include <linux/kernel.h>
#include <linux/sched.h>
#include <linux/mm.h>
#include <linux/smp.h>
#include <linux/errno.h>
#include <linux/ptrace.h>
#include <linux/tracehook.h>
#include <linux/audit.h>
#include <linux/seccomp.h>
#include <linux/signal.h>
#include <linux/export.h>
#include <linux/context_tracking.h>
#include <linux/user-return-notifier.h>
#include <linux/uprobes.h>

#include <asm/desc.h>
#include <asm/traps.h>
#include <asm/vdso.h>
#include <asm/uaccess.h>

#define CREATE_TRACE_POINTS
#include <trace/events/syscalls.h>

#ifdef CONFIG_CONTEXT_TRACKING
/* Called on entry from user mode with IRQs off. */
__visible void enter_from_user_mode(void)
{
	CT_WARN_ON(ct_state() != CONTEXT_USER);
	user_exit();
}
#endif

static void do_audit_syscall_entry(struct pt_regs *regs, u32 arch)
{
#ifdef CONFIG_X86_64
	if (arch == AUDIT_ARCH_X86_64) {
		audit_syscall_entry(regs->orig_ax, regs->di,
				    regs->si, regs->dx, regs->r10);
	} else
#endif
	{
		audit_syscall_entry(regs->orig_ax, regs->bx,
				    regs->cx, regs->dx, regs->si);
	}
}

/*
 * We can return 0 to resume the syscall or anything else to go to phase
 * 2.  If we resume the syscall, we need to put something appropriate in
 * regs->orig_ax.
 *
 * NB: We don't have full pt_regs here, but regs->orig_ax and regs->ax
 * are fully functional.
 *
 * For phase 2's benefit, our return value is:
 * 0:			resume the syscall
 * 1:			go to phase 2; no seccomp phase 2 needed
 * anything else:	go to phase 2; pass return value to seccomp
 */
unsigned long syscall_trace_enter_phase1(struct pt_regs *regs, u32 arch)
{
	unsigned long ret = 0;
	u32 work;

	if (IS_ENABLED(CONFIG_DEBUG_ENTRY))
		BUG_ON(regs != task_pt_regs(current));

	work = ACCESS_ONCE(current_thread_info()->flags) &
		_TIF_WORK_SYSCALL_ENTRY;

#ifdef CONFIG_CONTEXT_TRACKING
	/*
	 * If TIF_NOHZ is set, we are required to call user_exit() before
	 * doing anything that could touch RCU.
	 */
	if (work & _TIF_NOHZ) {
		enter_from_user_mode();
		work &= ~_TIF_NOHZ;
	}
#endif

#ifdef CONFIG_SECCOMP
	/*
	 * Do seccomp first -- it should minimize exposure of other
	 * code, and keeping seccomp fast is probably more valuable
	 * than the rest of this.
	 */
	if (work & _TIF_SECCOMP) {
		struct seccomp_data sd;

		sd.arch = arch;
		sd.nr = regs->orig_ax;
		sd.instruction_pointer = regs->ip;
#ifdef CONFIG_X86_64
		if (arch == AUDIT_ARCH_X86_64) {
			sd.args[0] = regs->di;
			sd.args[1] = regs->si;
			sd.args[2] = regs->dx;
			sd.args[3] = regs->r10;
			sd.args[4] = regs->r8;
			sd.args[5] = regs->r9;
		} else
#endif
		{
			sd.args[0] = regs->bx;
			sd.args[1] = regs->cx;
			sd.args[2] = regs->dx;
			sd.args[3] = regs->si;
			sd.args[4] = regs->di;
			sd.args[5] = regs->bp;
		}

		BUILD_BUG_ON(SECCOMP_PHASE1_OK != 0);
		BUILD_BUG_ON(SECCOMP_PHASE1_SKIP != 1);

		ret = seccomp_phase1(&sd);
		if (ret == SECCOMP_PHASE1_SKIP) {
			regs->orig_ax = -1;
			ret = 0;
		} else if (ret != SECCOMP_PHASE1_OK) {
			return ret;  /* Go directly to phase 2 */
		}

		work &= ~_TIF_SECCOMP;
	}
#endif

	/* Do our best to finish without phase 2. */
	if (work == 0)
		return ret;  /* seccomp and/or nohz only (ret == 0 here) */

#ifdef CONFIG_AUDITSYSCALL
	if (work == _TIF_SYSCALL_AUDIT) {
		/*
		 * If there is no more work to be done except auditing,
		 * then audit in phase 1.  Phase 2 always audits, so, if
		 * we audit here, then we can't go on to phase 2.
		 */
		do_audit_syscall_entry(regs, arch);
		return 0;
	}
#endif

	return 1;  /* Something is enabled that we can't handle in phase 1 */
}

/* Returns the syscall nr to run (which should match regs->orig_ax). */
long syscall_trace_enter_phase2(struct pt_regs *regs, u32 arch,
				unsigned long phase1_result)
{
	long ret = 0;
	u32 work = ACCESS_ONCE(current_thread_info()->flags) &
		_TIF_WORK_SYSCALL_ENTRY;

	if (IS_ENABLED(CONFIG_DEBUG_ENTRY))
		BUG_ON(regs != task_pt_regs(current));

	/*
	 * If we stepped into a sysenter/syscall insn, it trapped in
	 * kernel mode; do_debug() cleared TF and set TIF_SINGLESTEP.
	 * If user-mode had set TF itself, then it's still clear from
	 * do_debug() and we need to set it again to restore the user
	 * state.  If we entered on the slow path, TF was already set.
	 */
	if (work & _TIF_SINGLESTEP)
		regs->flags |= X86_EFLAGS_TF;

#ifdef CONFIG_SECCOMP
	/*
	 * Call seccomp_phase2 before running the other hooks so that
	 * they can see any changes made by a seccomp tracer.
	 */
	if (phase1_result > 1 && seccomp_phase2(phase1_result)) {
		/* seccomp failures shouldn't expose any additional code. */
		return -1;
	}
#endif

	if (unlikely(work & _TIF_SYSCALL_EMU))
		ret = -1L;

	if ((ret || test_thread_flag(TIF_SYSCALL_TRACE)) &&
	    tracehook_report_syscall_entry(regs))
		ret = -1L;

	if (unlikely(test_thread_flag(TIF_SYSCALL_TRACEPOINT)))
		trace_sys_enter(regs, regs->orig_ax);

	do_audit_syscall_entry(regs, arch);

	return ret ?: regs->orig_ax;
}

long syscall_trace_enter(struct pt_regs *regs)
{
	u32 arch = is_ia32_task() ? AUDIT_ARCH_I386 : AUDIT_ARCH_X86_64;
	unsigned long phase1_result = syscall_trace_enter_phase1(regs, arch);

	if (phase1_result == 0)
		return regs->orig_ax;
	else
		return syscall_trace_enter_phase2(regs, arch, phase1_result);
}

static struct thread_info *pt_regs_to_thread_info(struct pt_regs *regs)
{
	unsigned long top_of_stack =
		(unsigned long)(regs + 1) + TOP_OF_KERNEL_STACK_PADDING;
	return (struct thread_info *)(top_of_stack - THREAD_SIZE);
}

/* Called with IRQs disabled. */
__visible void prepare_exit_to_usermode(struct pt_regs *regs)
{
	if (IS_ENABLED(CONFIG_PROVE_LOCKING) && WARN_ON(!irqs_disabled()))
		local_irq_disable();

	lockdep_sys_exit();

	/*
	 * In order to return to user mode, we need to have IRQs off with
	 * none of _TIF_SIGPENDING, _TIF_NOTIFY_RESUME, _TIF_USER_RETURN_NOTIFY,
	 * _TIF_UPROBE, or _TIF_NEED_RESCHED set.  Several of these flags
	 * can be set at any time on preemptable kernels if we have IRQs on,
	 * so we need to loop.  Disabling preemption wouldn't help: doing the
	 * work to clear some of the flags can sleep.
	 */
	while (true) {
		u32 cached_flags =
			READ_ONCE(pt_regs_to_thread_info(regs)->flags);

		if (!(cached_flags & (_TIF_SIGPENDING | _TIF_NOTIFY_RESUME |
				      _TIF_UPROBE | _TIF_NEED_RESCHED |
				      _TIF_USER_RETURN_NOTIFY)))
			break;

		/* We have work to do. */
		local_irq_enable();

		if (cached_flags & _TIF_NEED_RESCHED)
			schedule();

		if (cached_flags & _TIF_UPROBE)
			uprobe_notify_resume(regs);

		/* deal with pending signal delivery */
		if (cached_flags & _TIF_SIGPENDING)
			do_signal(regs);

		if (cached_flags & _TIF_NOTIFY_RESUME) {
			clear_thread_flag(TIF_NOTIFY_RESUME);
			tracehook_notify_resume(regs);
		}

		if (cached_flags & _TIF_USER_RETURN_NOTIFY)
			fire_user_return_notifiers();

		/* Disable IRQs and retry */
		local_irq_disable();
	}

	user_enter();
}

/*
 * Called with IRQs on and fully valid regs.  Returns with IRQs off in a
 * state such that we can immediately switch to user mode.
 */
__visible void syscall_return_slowpath(struct pt_regs *regs)
{
	struct thread_info *ti = pt_regs_to_thread_info(regs);
	u32 cached_flags = READ_ONCE(ti->flags);
	bool step;

	CT_WARN_ON(ct_state() != CONTEXT_KERNEL);

	if (IS_ENABLED(CONFIG_PROVE_LOCKING) &&
	    WARN(irqs_disabled(), "syscall %ld left IRQs disabled", regs->orig_ax))
		local_irq_enable();

	/*
	 * First do one-time work.  If these work items are enabled, we
	 * want to run them exactly once per syscall exit with IRQs on.
	 */
	if (cached_flags & (_TIF_SYSCALL_TRACE | _TIF_SYSCALL_AUDIT |
			    _TIF_SINGLESTEP | _TIF_SYSCALL_TRACEPOINT)) {
		audit_syscall_exit(regs);

		if (cached_flags & _TIF_SYSCALL_TRACEPOINT)
			trace_sys_exit(regs, regs->ax);

		/*
		 * If TIF_SYSCALL_EMU is set, we only get here because of
		 * TIF_SINGLESTEP (i.e. this is PTRACE_SYSEMU_SINGLESTEP).
		 * We already reported this syscall instruction in
		 * syscall_trace_enter().
		 */
		step = unlikely(
			(cached_flags & (_TIF_SINGLESTEP | _TIF_SYSCALL_EMU))
			== _TIF_SINGLESTEP);
		if (step || cached_flags & _TIF_SYSCALL_TRACE)
			tracehook_report_syscall_exit(regs, step);
	}

#ifdef CONFIG_COMPAT
	/*
	 * Compat syscalls set TS_COMPAT.  Make sure we clear it before
	 * returning to user mode.
	 */
	ti->status &= ~TS_COMPAT;
#endif

	local_irq_disable();
	prepare_exit_to_usermode(regs);
}

#if defined(CONFIG_X86_32) || defined(CONFIG_IA32_EMULATION)
/*
 * Does a 32-bit syscall.  Called with IRQs on and does all entry and
 * exit work and returns with IRQs off.  This function is extremely hot
 * in workloads that use it, and it's usually called from
 * do_fast_syscall_32, so forcibly inline it to improve performance.
 */
static __always_inline void do_syscall_32_irqs_on(struct pt_regs *regs)
{
	struct thread_info *ti = pt_regs_to_thread_info(regs);
	unsigned int nr = (unsigned int)regs->orig_ax;

#ifdef CONFIG_IA32_EMULATION
	ti->status |= TS_COMPAT;
#endif

	if (READ_ONCE(ti->flags) & _TIF_WORK_SYSCALL_ENTRY) {
		/*
		 * Subtlety here: if ptrace pokes something larger than
		 * 2^32-1 into orig_ax, this truncates it.  This may or
		 * may not be necessary, but it matches the old asm
		 * behavior.
		 */
		nr = syscall_trace_enter(regs);
	}

	if (likely(nr < IA32_NR_syscalls)) {
		/*
		 * It's possible that a 32-bit syscall implementation
		 * takes a 64-bit parameter but nonetheless assumes that
		 * the high bits are zero.  Make sure we zero-extend all
		 * of the args.
		 */
		regs->ax = ia32_sys_call_table[nr](
			(unsigned int)regs->bx, (unsigned int)regs->cx,
			(unsigned int)regs->dx, (unsigned int)regs->si,
			(unsigned int)regs->di, (unsigned int)regs->bp);
	}

	syscall_return_slowpath(regs);
}

/* Handles int $0x80 */
__visible void do_int80_syscall_32(struct pt_regs *regs)
{
	local_irq_enable();
	do_syscall_32_irqs_on(regs);
}

/* Returns 0 to return using IRET or 1 to return using SYSEXIT/SYSRETL. */
__visible long do_fast_syscall_32(struct pt_regs *regs)
{
	/*
	 * Called using the internal vDSO SYSENTER/SYSCALL32 calling
	 * convention.  Adjust regs so it looks like we entered using int80.
	 */

	unsigned long landing_pad = (unsigned long)current->mm->context.vdso +
		vdso_image_32.sym_int80_landing_pad;

	/*
	 * SYSENTER loses EIP, and even SYSCALL32 needs us to skip forward
	 * so that 'regs->ip -= 2' lands back on an int $0x80 instruction.
	 * Fix it up.
	 */
	regs->ip = landing_pad;

	/*
	 * Fetch ECX from where the vDSO stashed it.
	 *
	 * WARNING: We are in CONTEXT_USER and RCU isn't paying attention!
	 */
	local_irq_enable();
	if (
#ifdef CONFIG_X86_64
		/*
		 * Micro-optimization: the pointer we're following is explicitly
		 * 32 bits, so it can't be out of range.
		 */
		__get_user(*(u32 *)&regs->cx,
			    (u32 __user __force *)(unsigned long)(u32)regs->sp)
#else
		get_user(*(u32 *)&regs->cx,
			 (u32 __user __force *)(unsigned long)(u32)regs->sp)
#endif
		) {

		/* User code screwed up. */
		local_irq_disable();
		regs->ax = -EFAULT;
#ifdef CONFIG_CONTEXT_TRACKING
		enter_from_user_mode();
#endif
		prepare_exit_to_usermode(regs);
		return 0;	/* Keep it simple: use IRET. */
	}

	/* Now this is just like a normal syscall. */
	do_syscall_32_irqs_on(regs);

#ifdef CONFIG_X86_64
	/*
	 * Opportunistic SYSRETL: if possible, try to return using SYSRETL.
	 * SYSRETL is available on all 64-bit CPUs, so we don't need to
	 * bother with SYSEXIT.
	 *
	 * Unlike 64-bit opportunistic SYSRET, we can't check that CX == IP,
	 * because the ECX fixup above will ensure that this is essentially
	 * never the case.
	 */
	return regs->cs == __USER32_CS && regs->ss == __USER_DS &&
		regs->ip == landing_pad &&
		(regs->flags & (X86_EFLAGS_RF | X86_EFLAGS_TF)) == 0;
#else
	/*
	 * Opportunistic SYSEXIT: if possible, try to return using SYSEXIT.
	 *
	 * Unlike 64-bit opportunistic SYSRET, we can't check that CX == IP,
	 * because the ECX fixup above will ensure that this is essentially
	 * never the case.
	 *
	 * We don't allow syscalls at all from VM86 mode, but we still
	 * need to check VM, because we might be returning from sys_vm86.
	 */
	return static_cpu_has(X86_FEATURE_SEP) &&
		regs->cs == __USER_CS && regs->ss == __USER_DS &&
		regs->ip == landing_pad &&
		(regs->flags & (X86_EFLAGS_RF | X86_EFLAGS_TF | X86_EFLAGS_VM)) == 0;
#endif
}
#endif
Commit	Line	Data
1f484aa6 AL	1	/*
	2	* common.c - C code for kernel entry and exit
	3	* Copyright (c) 2015 Andrew Lutomirski
	4	* GPL v2
	5	*
	6	* Based on asm and ptrace code by many authors. The code here originated
	7	* in ptrace.c and signal.c.
	8	*/
	9
	10	#include <linux/kernel.h>
	11	#include <linux/sched.h>
	12	#include <linux/mm.h>
	13	#include <linux/smp.h>
	14	#include <linux/errno.h>
	15	#include <linux/ptrace.h>
	16	#include <linux/tracehook.h>
	17	#include <linux/audit.h>
	18	#include <linux/seccomp.h>
	19	#include <linux/signal.h>
	20	#include <linux/export.h>
	21	#include <linux/context_tracking.h>
	22	#include <linux/user-return-notifier.h>
	23	#include <linux/uprobes.h>
	24
	25	#include <asm/desc.h>
	26	#include <asm/traps.h>
710246df AL	27	#include <asm/vdso.h>
710246df AL	28	#include <asm/uaccess.h>
1f484aa6 AL	29
	30	#define CREATE_TRACE_POINTS
	31	#include <trace/events/syscalls.h>
	32
feed36cd AL	33	#ifdef CONFIG_CONTEXT_TRACKING
	34	/* Called on entry from user mode with IRQs off. */
	35	__visible void enter_from_user_mode(void)
	36	{
	37	CT_WARN_ON(ct_state() != CONTEXT_USER);
	38	user_exit();
	39	}
	40	#endif
	41
1f484aa6 AL	42	static void do_audit_syscall_entry(struct pt_regs *regs, u32 arch)
	43	{
	44	#ifdef CONFIG_X86_64
	45	if (arch == AUDIT_ARCH_X86_64) {
	46	audit_syscall_entry(regs->orig_ax, regs->di,
	47	regs->si, regs->dx, regs->r10);
	48	} else
	49	#endif
	50	{
	51	audit_syscall_entry(regs->orig_ax, regs->bx,
	52	regs->cx, regs->dx, regs->si);
	53	}
	54	}
	55
	56	/*
	57	* We can return 0 to resume the syscall or anything else to go to phase
	58	* 2. If we resume the syscall, we need to put something appropriate in
	59	* regs->orig_ax.
	60	*
	61	* NB: We don't have full pt_regs here, but regs->orig_ax and regs->ax
	62	* are fully functional.
	63	*
	64	* For phase 2's benefit, our return value is:
	65	* 0: resume the syscall
	66	* 1: go to phase 2; no seccomp phase 2 needed
	67	* anything else: go to phase 2; pass return value to seccomp
	68	*/
	69	unsigned long syscall_trace_enter_phase1(struct pt_regs *regs, u32 arch)
	70	{
	71	unsigned long ret = 0;
	72	u32 work;
	73
4aabd140 AL	74	if (IS_ENABLED(CONFIG_DEBUG_ENTRY))
4aabd140 AL	75	BUG_ON(regs != task_pt_regs(current));
1f484aa6 AL	76
	77	work = ACCESS_ONCE(current_thread_info()->flags) &
	78	_TIF_WORK_SYSCALL_ENTRY;
	79
feed36cd	80	#ifdef CONFIG_CONTEXT_TRACKING
1f484aa6 AL	81	/*
	82	* If TIF_NOHZ is set, we are required to call user_exit() before
	83	* doing anything that could touch RCU.
	84	*/
	85	if (work & _TIF_NOHZ) {
feed36cd	86	enter_from_user_mode();
1f484aa6 AL	87	work &= ~_TIF_NOHZ;
1f484aa6 AL	88	}
feed36cd	89	#endif
1f484aa6 AL	90
	91	#ifdef CONFIG_SECCOMP
	92	/*
	93	* Do seccomp first -- it should minimize exposure of other
	94	* code, and keeping seccomp fast is probably more valuable
	95	* than the rest of this.
	96	*/
	97	if (work & _TIF_SECCOMP) {
	98	struct seccomp_data sd;
	99
	100	sd.arch = arch;
	101	sd.nr = regs->orig_ax;
	102	sd.instruction_pointer = regs->ip;
	103	#ifdef CONFIG_X86_64
	104	if (arch == AUDIT_ARCH_X86_64) {
	105	sd.args[0] = regs->di;
	106	sd.args[1] = regs->si;
	107	sd.args[2] = regs->dx;
	108	sd.args[3] = regs->r10;
	109	sd.args[4] = regs->r8;
	110	sd.args[5] = regs->r9;
	111	} else
	112	#endif
	113	{
	114	sd.args[0] = regs->bx;
	115	sd.args[1] = regs->cx;
	116	sd.args[2] = regs->dx;
	117	sd.args[3] = regs->si;
	118	sd.args[4] = regs->di;
	119	sd.args[5] = regs->bp;
	120	}
	121
	122	BUILD_BUG_ON(SECCOMP_PHASE1_OK != 0);
	123	BUILD_BUG_ON(SECCOMP_PHASE1_SKIP != 1);
	124
	125	ret = seccomp_phase1(&sd);
	126	if (ret == SECCOMP_PHASE1_SKIP) {
	127	regs->orig_ax = -1;
	128	ret = 0;
	129	} else if (ret != SECCOMP_PHASE1_OK) {
	130	return ret; /* Go directly to phase 2 */
	131	}
	132
	133	work &= ~_TIF_SECCOMP;
	134	}
	135	#endif
	136
	137	/* Do our best to finish without phase 2. */
	138	if (work == 0)
	139	return ret; /* seccomp and/or nohz only (ret == 0 here) */
	140
	141	#ifdef CONFIG_AUDITSYSCALL
	142	if (work == _TIF_SYSCALL_AUDIT) {
	143	/*
	144	* If there is no more work to be done except auditing,
	145	* then audit in phase 1. Phase 2 always audits, so, if
	146	* we audit here, then we can't go on to phase 2.
	147	*/
	148	do_audit_syscall_entry(regs, arch);
	149	return 0;
	150	}
	151	#endif
	152
	153	return 1; /* Something is enabled that we can't handle in phase 1 */
154	}
155
156	/* Returns the syscall nr to run (which should match regs->orig_ax). */
157	long syscall_trace_enter_phase2(struct pt_regs *regs, u32 arch,
158	unsigned long phase1_result)
159	{
160	long ret = 0;
161	u32 work = ACCESS_ONCE(current_thread_info()->flags) &
162	_TIF_WORK_SYSCALL_ENTRY;
163
4aabd140 AL	164	if (IS_ENABLED(CONFIG_DEBUG_ENTRY))
4aabd140 AL	165	BUG_ON(regs != task_pt_regs(current));
1f484aa6 AL	166
	167	/*
	168	* If we stepped into a sysenter/syscall insn, it trapped in
	169	* kernel mode; do_debug() cleared TF and set TIF_SINGLESTEP.
	170	* If user-mode had set TF itself, then it's still clear from
	171	* do_debug() and we need to set it again to restore the user
	172	* state. If we entered on the slow path, TF was already set.
	173	*/
	174	if (work & _TIF_SINGLESTEP)
	175	regs->flags \|= X86_EFLAGS_TF;
	176
	177	#ifdef CONFIG_SECCOMP
	178	/*
	179	* Call seccomp_phase2 before running the other hooks so that
	180	* they can see any changes made by a seccomp tracer.
	181	*/
	182	if (phase1_result > 1 && seccomp_phase2(phase1_result)) {
	183	/* seccomp failures shouldn't expose any additional code. */
	184	return -1;
	185	}
	186	#endif
	187
	188	if (unlikely(work & _TIF_SYSCALL_EMU))
	189	ret = -1L;
	190
	191	if ((ret \|\| test_thread_flag(TIF_SYSCALL_TRACE)) &&
	192	tracehook_report_syscall_entry(regs))
	193	ret = -1L;
	194
	195	if (unlikely(test_thread_flag(TIF_SYSCALL_TRACEPOINT)))
	196	trace_sys_enter(regs, regs->orig_ax);
	197
	198	do_audit_syscall_entry(regs, arch);
	199
	200	return ret ?: regs->orig_ax;
	201	}
	202
	203	long syscall_trace_enter(struct pt_regs *regs)
	204	{
	205	u32 arch = is_ia32_task() ? AUDIT_ARCH_I386 : AUDIT_ARCH_X86_64;
	206	unsigned long phase1_result = syscall_trace_enter_phase1(regs, arch);
	207
	208	if (phase1_result == 0)
	209	return regs->orig_ax;
	210	else
	211	return syscall_trace_enter_phase2(regs, arch, phase1_result);
	212	}
	213
c5c46f59 AL	214	static struct thread_info pt_regs_to_thread_info(struct pt_regs regs)
	215	{
	216	unsigned long top_of_stack =
	217	(unsigned long)(regs + 1) + TOP_OF_KERNEL_STACK_PADDING;
	218	return (struct thread_info *)(top_of_stack - THREAD_SIZE);
	219	}
	220
	221	/* Called with IRQs disabled. */
	222	__visible void prepare_exit_to_usermode(struct pt_regs *regs)
	223	{
460d1245	224	if (IS_ENABLED(CONFIG_PROVE_LOCKING) && WARN_ON(!irqs_disabled()))
c5c46f59 AL	225	local_irq_disable();
c5c46f59 AL	226
72f92478 AL	227	lockdep_sys_exit();
72f92478 AL	228
c5c46f59 AL	229	/*
	230	* In order to return to user mode, we need to have IRQs off with
	231	* none of _TIF_SIGPENDING, _TIF_NOTIFY_RESUME, _TIF_USER_RETURN_NOTIFY,
	232	* _TIF_UPROBE, or _TIF_NEED_RESCHED set. Several of these flags
	233	* can be set at any time on preemptable kernels if we have IRQs on,
	234	* so we need to loop. Disabling preemption wouldn't help: doing the
	235	* work to clear some of the flags can sleep.
	236	*/
	237	while (true) {
	238	u32 cached_flags =
	239	READ_ONCE(pt_regs_to_thread_info(regs)->flags);
	240
	241	if (!(cached_flags & (_TIF_SIGPENDING \| _TIF_NOTIFY_RESUME \|
d132803e AL	242	_TIF_UPROBE \| _TIF_NEED_RESCHED \|
d132803e AL	243	_TIF_USER_RETURN_NOTIFY)))
c5c46f59 AL	244	break;
	245
	246	/* We have work to do. */
	247	local_irq_enable();
	248
	249	if (cached_flags & _TIF_NEED_RESCHED)
	250	schedule();
	251
	252	if (cached_flags & _TIF_UPROBE)
	253	uprobe_notify_resume(regs);
	254
	255	/* deal with pending signal delivery */
	256	if (cached_flags & _TIF_SIGPENDING)
	257	do_signal(regs);
	258
	259	if (cached_flags & _TIF_NOTIFY_RESUME) {
	260	clear_thread_flag(TIF_NOTIFY_RESUME);
	261	tracehook_notify_resume(regs);
	262	}
	263
	264	if (cached_flags & _TIF_USER_RETURN_NOTIFY)
	265	fire_user_return_notifiers();
	266
	267	/* Disable IRQs and retry */
	268	local_irq_disable();
	269	}
	270
	271	user_enter();
	272	}
	273
	274	/*
	275	* Called with IRQs on and fully valid regs. Returns with IRQs off in a
	276	* state such that we can immediately switch to user mode.
	277	*/
	278	__visible void syscall_return_slowpath(struct pt_regs *regs)
	279	{
	280	struct thread_info *ti = pt_regs_to_thread_info(regs);
	281	u32 cached_flags = READ_ONCE(ti->flags);
	282	bool step;
	283
	284	CT_WARN_ON(ct_state() != CONTEXT_KERNEL);
	285
460d1245 AL	286	if (IS_ENABLED(CONFIG_PROVE_LOCKING) &&
460d1245 AL	287	WARN(irqs_disabled(), "syscall %ld left IRQs disabled", regs->orig_ax))
c5c46f59 AL	288	local_irq_enable();
	289
	290	/*
	291	* First do one-time work. If these work items are enabled, we
	292	* want to run them exactly once per syscall exit with IRQs on.
	293	*/
	294	if (cached_flags & (_TIF_SYSCALL_TRACE \| _TIF_SYSCALL_AUDIT \|
	295	_TIF_SINGLESTEP \| _TIF_SYSCALL_TRACEPOINT)) {
	296	audit_syscall_exit(regs);
	297
	298	if (cached_flags & _TIF_SYSCALL_TRACEPOINT)
	299	trace_sys_exit(regs, regs->ax);
	300
	301	/*
	302	* If TIF_SYSCALL_EMU is set, we only get here because of
	303	* TIF_SINGLESTEP (i.e. this is PTRACE_SYSEMU_SINGLESTEP).
	304	* We already reported this syscall instruction in
	305	* syscall_trace_enter().
	306	*/
	307	step = unlikely(
	308	(cached_flags & (_TIF_SINGLESTEP \| _TIF_SYSCALL_EMU))
	309	== _TIF_SINGLESTEP);
	310	if (step \|\| cached_flags & _TIF_SYSCALL_TRACE)
	311	tracehook_report_syscall_exit(regs, step);
	312	}
	313
	314	#ifdef CONFIG_COMPAT
	315	/*
	316	* Compat syscalls set TS_COMPAT. Make sure we clear it before
	317	* returning to user mode.
	318	*/
	319	ti->status &= ~TS_COMPAT;
	320	#endif
	321
	322	local_irq_disable();
	323	prepare_exit_to_usermode(regs);
	324	}
bd2d3a3b AL	325
	326	#if defined(CONFIG_X86_32) \|\| defined(CONFIG_IA32_EMULATION)
	327	/*
8b13c255	328	* Does a 32-bit syscall. Called with IRQs on and does all entry and
33c52129 AL	329	* exit work and returns with IRQs off. This function is extremely hot
	330	* in workloads that use it, and it's usually called from
	331	* do_fast_syscall_32, so forcibly inline it to improve performance.
bd2d3a3b	332	*/
33c52129	333	static __always_inline void do_syscall_32_irqs_on(struct pt_regs *regs)
bd2d3a3b AL	334	{
	335	struct thread_info *ti = pt_regs_to_thread_info(regs);
	336	unsigned int nr = (unsigned int)regs->orig_ax;
	337
	338	#ifdef CONFIG_IA32_EMULATION
	339	ti->status \|= TS_COMPAT;
	340	#endif
	341
bd2d3a3b AL	342	if (READ_ONCE(ti->flags) & _TIF_WORK_SYSCALL_ENTRY) {
	343	/*
	344	* Subtlety here: if ptrace pokes something larger than
	345	* 2^32-1 into orig_ax, this truncates it. This may or
	346	* may not be necessary, but it matches the old asm
	347	* behavior.
	348	*/
	349	nr = syscall_trace_enter(regs);
	350	}
	351
33c52129	352	if (likely(nr < IA32_NR_syscalls)) {
bd2d3a3b AL	353	/*
	354	* It's possible that a 32-bit syscall implementation
	355	* takes a 64-bit parameter but nonetheless assumes that
	356	* the high bits are zero. Make sure we zero-extend all
	357	* of the args.
	358	*/
	359	regs->ax = ia32_sys_call_table[nr](
	360	(unsigned int)regs->bx, (unsigned int)regs->cx,
	361	(unsigned int)regs->dx, (unsigned int)regs->si,
	362	(unsigned int)regs->di, (unsigned int)regs->bp);
	363	}
	364
	365	syscall_return_slowpath(regs);
	366	}
710246df	367
8b13c255 AL	368	/* Handles int $0x80 */
	369	__visible void do_int80_syscall_32(struct pt_regs *regs)
	370	{
	371	local_irq_enable();
	372	do_syscall_32_irqs_on(regs);
	373	}
	374
5f310f73	375	/* Returns 0 to return using IRET or 1 to return using SYSEXIT/SYSRETL. */
7841b408	376	__visible long do_fast_syscall_32(struct pt_regs *regs)
710246df AL	377	{
	378	/*
	379	* Called using the internal vDSO SYSENTER/SYSCALL32 calling
	380	* convention. Adjust regs so it looks like we entered using int80.
	381	*/
	382
	383	unsigned long landing_pad = (unsigned long)current->mm->context.vdso +
	384	vdso_image_32.sym_int80_landing_pad;
	385
	386	/*
	387	* SYSENTER loses EIP, and even SYSCALL32 needs us to skip forward
	388	* so that 'regs->ip -= 2' lands back on an int $0x80 instruction.
	389	* Fix it up.
	390	*/
	391	regs->ip = landing_pad;
	392
	393	/*
	394	* Fetch ECX from where the vDSO stashed it.
	395	*
	396	* WARNING: We are in CONTEXT_USER and RCU isn't paying attention!
	397	*/
	398	local_irq_enable();
c68ca678 AL	399	if (
	400	#ifdef CONFIG_X86_64
	401	/*
	402	* Micro-optimization: the pointer we're following is explicitly
	403	* 32 bits, so it can't be out of range.
	404	*/
	405	__get_user((u32 )&regs->cx,
	406	(u32 __user __force *)(unsigned long)(u32)regs->sp)
	407	#else
	408	get_user((u32 )&regs->cx,
	409	(u32 __user __force *)(unsigned long)(u32)regs->sp)
	410	#endif
	411	) {
	412
710246df AL	413	/* User code screwed up. */
	414	local_irq_disable();
	415	regs->ax = -EFAULT;
	416	#ifdef CONFIG_CONTEXT_TRACKING
	417	enter_from_user_mode();
	418	#endif
	419	prepare_exit_to_usermode(regs);
7841b408	420	return 0; /* Keep it simple: use IRET. */
710246df	421	}
710246df AL	422
710246df AL	423	/* Now this is just like a normal syscall. */
8b13c255	424	do_syscall_32_irqs_on(regs);
7841b408 AL	425
	426	#ifdef CONFIG_X86_64
	427	/*
	428	* Opportunistic SYSRETL: if possible, try to return using SYSRETL.
	429	* SYSRETL is available on all 64-bit CPUs, so we don't need to
	430	* bother with SYSEXIT.
	431	*
	432	* Unlike 64-bit opportunistic SYSRET, we can't check that CX == IP,
	433	* because the ECX fixup above will ensure that this is essentially
	434	* never the case.
	435	*/
	436	return regs->cs == __USER32_CS && regs->ss == __USER_DS &&
	437	regs->ip == landing_pad &&
	438	(regs->flags & (X86_EFLAGS_RF \| X86_EFLAGS_TF)) == 0;
	439	#else
5f310f73 AL	440	/*
	441	* Opportunistic SYSEXIT: if possible, try to return using SYSEXIT.
	442	*
	443	* Unlike 64-bit opportunistic SYSRET, we can't check that CX == IP,
	444	* because the ECX fixup above will ensure that this is essentially
	445	* never the case.
	446	*
	447	* We don't allow syscalls at all from VM86 mode, but we still
	448	* need to check VM, because we might be returning from sys_vm86.
	449	*/
	450	return static_cpu_has(X86_FEATURE_SEP) &&
	451	regs->cs == __USER_CS && regs->ss == __USER_DS &&
	452	regs->ip == landing_pad &&
	453	(regs->flags & (X86_EFLAGS_RF \| X86_EFLAGS_TF \| X86_EFLAGS_VM)) == 0;
7841b408	454	#endif
710246df	455	}
bd2d3a3b	456	#endif