aboutsummaryrefslogtreecommitdiff
path: root/arch/x86_64/kernel
diff options
context:
space:
mode:
Diffstat (limited to 'arch/x86_64/kernel')
-rw-r--r--arch/x86_64/kernel/i8259.c2
-rw-r--r--arch/x86_64/kernel/kprobes.c183
-rw-r--r--arch/x86_64/kernel/process.c69
-rw-r--r--arch/x86_64/kernel/ptrace.c43
-rw-r--r--arch/x86_64/kernel/smpboot.c3
5 files changed, 127 insertions, 173 deletions
diff --git a/arch/x86_64/kernel/i8259.c b/arch/x86_64/kernel/i8259.c
index 5de30035e54..6e5101ad3d1 100644
--- a/arch/x86_64/kernel/i8259.c
+++ b/arch/x86_64/kernel/i8259.c
@@ -516,7 +516,7 @@ void i8254_timer_resume(void)
}
static struct sysdev_class timer_sysclass = {
- set_kset_name("timer"),
+ set_kset_name("timer_pit"),
.resume = timer_resume,
};
diff --git a/arch/x86_64/kernel/kprobes.c b/arch/x86_64/kernel/kprobes.c
index 76a28b007be..dddeb678b44 100644
--- a/arch/x86_64/kernel/kprobes.c
+++ b/arch/x86_64/kernel/kprobes.c
@@ -34,7 +34,6 @@
#include <linux/config.h>
#include <linux/kprobes.h>
#include <linux/ptrace.h>
-#include <linux/spinlock.h>
#include <linux/string.h>
#include <linux/slab.h>
#include <linux/preempt.h>
@@ -44,17 +43,10 @@
#include <asm/kdebug.h>
static DECLARE_MUTEX(kprobe_mutex);
-
-static struct kprobe *current_kprobe;
-static unsigned long kprobe_status, kprobe_old_rflags, kprobe_saved_rflags;
-static struct kprobe *kprobe_prev;
-static unsigned long kprobe_status_prev, kprobe_old_rflags_prev, kprobe_saved_rflags_prev;
-static struct pt_regs jprobe_saved_regs;
-static long *jprobe_saved_rsp;
void jprobe_return_end(void);
-/* copy of the kernel stack at the probe fire time */
-static kprobe_opcode_t jprobes_stack[MAX_STACK_SIZE];
+DEFINE_PER_CPU(struct kprobe *, current_kprobe) = NULL;
+DEFINE_PER_CPU(struct kprobe_ctlblk, kprobe_ctlblk);
/*
* returns non-zero if opcode modifies the interrupt flag.
@@ -236,29 +228,30 @@ void __kprobes arch_remove_kprobe(struct kprobe *p)
up(&kprobe_mutex);
}
-static inline void save_previous_kprobe(void)
+static inline void save_previous_kprobe(struct kprobe_ctlblk *kcb)
{
- kprobe_prev = current_kprobe;
- kprobe_status_prev = kprobe_status;
- kprobe_old_rflags_prev = kprobe_old_rflags;
- kprobe_saved_rflags_prev = kprobe_saved_rflags;
+ kcb->prev_kprobe.kp = kprobe_running();
+ kcb->prev_kprobe.status = kcb->kprobe_status;
+ kcb->prev_kprobe.old_rflags = kcb->kprobe_old_rflags;
+ kcb->prev_kprobe.saved_rflags = kcb->kprobe_saved_rflags;
}
-static inline void restore_previous_kprobe(void)
+static inline void restore_previous_kprobe(struct kprobe_ctlblk *kcb)
{
- current_kprobe = kprobe_prev;
- kprobe_status = kprobe_status_prev;
- kprobe_old_rflags = kprobe_old_rflags_prev;
- kprobe_saved_rflags = kprobe_saved_rflags_prev;
+ __get_cpu_var(current_kprobe) = kcb->prev_kprobe.kp;
+ kcb->kprobe_status = kcb->prev_kprobe.status;
+ kcb->kprobe_old_rflags = kcb->prev_kprobe.old_rflags;
+ kcb->kprobe_saved_rflags = kcb->prev_kprobe.saved_rflags;
}
-static inline void set_current_kprobe(struct kprobe *p, struct pt_regs *regs)
+static inline void set_current_kprobe(struct kprobe *p, struct pt_regs *regs,
+ struct kprobe_ctlblk *kcb)
{
- current_kprobe = p;
- kprobe_saved_rflags = kprobe_old_rflags
+ __get_cpu_var(current_kprobe) = p;
+ kcb->kprobe_saved_rflags = kcb->kprobe_old_rflags
= (regs->eflags & (TF_MASK | IF_MASK));
if (is_IF_modifier(p->ainsn.insn))
- kprobe_saved_rflags &= ~IF_MASK;
+ kcb->kprobe_saved_rflags &= ~IF_MASK;
}
static void __kprobes prepare_singlestep(struct kprobe *p, struct pt_regs *regs)
@@ -272,6 +265,7 @@ static void __kprobes prepare_singlestep(struct kprobe *p, struct pt_regs *regs)
regs->rip = (unsigned long)p->ainsn.insn;
}
+/* Called with kretprobe_lock held */
void __kprobes arch_prepare_kretprobe(struct kretprobe *rp,
struct pt_regs *regs)
{
@@ -292,32 +286,30 @@ void __kprobes arch_prepare_kretprobe(struct kretprobe *rp,
}
}
-/*
- * Interrupts are disabled on entry as trap3 is an interrupt gate and they
- * remain disabled thorough out this function.
- */
int __kprobes kprobe_handler(struct pt_regs *regs)
{
struct kprobe *p;
int ret = 0;
kprobe_opcode_t *addr = (kprobe_opcode_t *)(regs->rip - sizeof(kprobe_opcode_t));
+ struct kprobe_ctlblk *kcb;
- /* We're in an interrupt, but this is clear and BUG()-safe. */
+ /*
+ * We don't want to be preempted for the entire
+ * duration of kprobe processing
+ */
preempt_disable();
+ kcb = get_kprobe_ctlblk();
/* Check we're not actually recursing */
if (kprobe_running()) {
- /* We *are* holding lock here, so this is safe.
- Disarm the probe we just hit, and ignore it. */
p = get_kprobe(addr);
if (p) {
- if (kprobe_status == KPROBE_HIT_SS &&
+ if (kcb->kprobe_status == KPROBE_HIT_SS &&
*p->ainsn.insn == BREAKPOINT_INSTRUCTION) {
regs->eflags &= ~TF_MASK;
- regs->eflags |= kprobe_saved_rflags;
- unlock_kprobes();
+ regs->eflags |= kcb->kprobe_saved_rflags;
goto no_kprobe;
- } else if (kprobe_status == KPROBE_HIT_SSDONE) {
+ } else if (kcb->kprobe_status == KPROBE_HIT_SSDONE) {
/* TODO: Provide re-entrancy from
* post_kprobes_handler() and avoid exception
* stack corruption while single-stepping on
@@ -325,6 +317,7 @@ int __kprobes kprobe_handler(struct pt_regs *regs)
*/
arch_disarm_kprobe(p);
regs->rip = (unsigned long)p->addr;
+ reset_current_kprobe();
ret = 1;
} else {
/* We have reentered the kprobe_handler(), since
@@ -334,27 +327,24 @@ int __kprobes kprobe_handler(struct pt_regs *regs)
* of the new probe without calling any user
* handlers.
*/
- save_previous_kprobe();
- set_current_kprobe(p, regs);
+ save_previous_kprobe(kcb);
+ set_current_kprobe(p, regs, kcb);
p->nmissed++;
prepare_singlestep(p, regs);
- kprobe_status = KPROBE_REENTER;
+ kcb->kprobe_status = KPROBE_REENTER;
return 1;
}
} else {
- p = current_kprobe;
+ p = __get_cpu_var(current_kprobe);
if (p->break_handler && p->break_handler(p, regs)) {
goto ss_probe;
}
}
- /* If it's not ours, can't be delete race, (we hold lock). */
goto no_kprobe;
}
- lock_kprobes();
p = get_kprobe(addr);
if (!p) {
- unlock_kprobes();
if (*addr != BREAKPOINT_INSTRUCTION) {
/*
* The breakpoint instruction was removed right
@@ -372,8 +362,8 @@ int __kprobes kprobe_handler(struct pt_regs *regs)
goto no_kprobe;
}
- kprobe_status = KPROBE_HIT_ACTIVE;
- set_current_kprobe(p, regs);
+ set_current_kprobe(p, regs, kcb);
+ kcb->kprobe_status = KPROBE_HIT_ACTIVE;
if (p->pre_handler && p->pre_handler(p, regs))
/* handler has already set things up, so skip ss setup */
@@ -381,7 +371,7 @@ int __kprobes kprobe_handler(struct pt_regs *regs)
ss_probe:
prepare_singlestep(p, regs);
- kprobe_status = KPROBE_HIT_SS;
+ kcb->kprobe_status = KPROBE_HIT_SS;
return 1;
no_kprobe:
@@ -409,9 +399,10 @@ int __kprobes trampoline_probe_handler(struct kprobe *p, struct pt_regs *regs)
struct kretprobe_instance *ri = NULL;
struct hlist_head *head;
struct hlist_node *node, *tmp;
- unsigned long orig_ret_address = 0;
+ unsigned long flags, orig_ret_address = 0;
unsigned long trampoline_address =(unsigned long)&kretprobe_trampoline;
+ spin_lock_irqsave(&kretprobe_lock, flags);
head = kretprobe_inst_table_head(current);
/*
@@ -450,13 +441,14 @@ int __kprobes trampoline_probe_handler(struct kprobe *p, struct pt_regs *regs)
BUG_ON(!orig_ret_address || (orig_ret_address == trampoline_address));
regs->rip = orig_ret_address;
- unlock_kprobes();
+ reset_current_kprobe();
+ spin_unlock_irqrestore(&kretprobe_lock, flags);
preempt_enable_no_resched();
/*
* By returning a non-zero value, we are telling
- * kprobe_handler() that we have handled unlocking
- * and re-enabling preemption.
+ * kprobe_handler() that we don't want the post_handler
+ * to run (and have re-enabled preemption)
*/
return 1;
}
@@ -483,7 +475,8 @@ int __kprobes trampoline_probe_handler(struct kprobe *p, struct pt_regs *regs)
* that is atop the stack is the address following the copied instruction.
* We need to make it the address following the original instruction.
*/
-static void __kprobes resume_execution(struct kprobe *p, struct pt_regs *regs)
+static void __kprobes resume_execution(struct kprobe *p,
+ struct pt_regs *regs, struct kprobe_ctlblk *kcb)
{
unsigned long *tos = (unsigned long *)regs->rsp;
unsigned long next_rip = 0;
@@ -498,7 +491,7 @@ static void __kprobes resume_execution(struct kprobe *p, struct pt_regs *regs)
switch (*insn) {
case 0x9c: /* pushfl */
*tos &= ~(TF_MASK | IF_MASK);
- *tos |= kprobe_old_rflags;
+ *tos |= kcb->kprobe_old_rflags;
break;
case 0xc3: /* ret/lret */
case 0xcb:
@@ -537,30 +530,28 @@ static void __kprobes resume_execution(struct kprobe *p, struct pt_regs *regs)
}
}
-/*
- * Interrupts are disabled on entry as trap1 is an interrupt gate and they
- * remain disabled thoroughout this function. And we hold kprobe lock.
- */
int __kprobes post_kprobe_handler(struct pt_regs *regs)
{
- if (!kprobe_running())
+ struct kprobe *cur = kprobe_running();
+ struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
+
+ if (!cur)
return 0;
- if ((kprobe_status != KPROBE_REENTER) && current_kprobe->post_handler) {
- kprobe_status = KPROBE_HIT_SSDONE;
- current_kprobe->post_handler(current_kprobe, regs, 0);
+ if ((kcb->kprobe_status != KPROBE_REENTER) && cur->post_handler) {
+ kcb->kprobe_status = KPROBE_HIT_SSDONE;
+ cur->post_handler(cur, regs, 0);
}
- resume_execution(current_kprobe, regs);
- regs->eflags |= kprobe_saved_rflags;
+ resume_execution(cur, regs, kcb);
+ regs->eflags |= kcb->kprobe_saved_rflags;
/* Restore the original saved kprobes variables and continue. */
- if (kprobe_status == KPROBE_REENTER) {
- restore_previous_kprobe();
+ if (kcb->kprobe_status == KPROBE_REENTER) {
+ restore_previous_kprobe(kcb);
goto out;
- } else {
- unlock_kprobes();
}
+ reset_current_kprobe();
out:
preempt_enable_no_resched();
@@ -575,18 +566,19 @@ out:
return 1;
}
-/* Interrupts disabled, kprobe_lock held. */
int __kprobes kprobe_fault_handler(struct pt_regs *regs, int trapnr)
{
- if (current_kprobe->fault_handler
- && current_kprobe->fault_handler(current_kprobe, regs, trapnr))
+ struct kprobe *cur = kprobe_running();
+ struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
+
+ if (cur->fault_handler && cur->fault_handler(cur, regs, trapnr))
return 1;
- if (kprobe_status & KPROBE_HIT_SS) {
- resume_execution(current_kprobe, regs);
- regs->eflags |= kprobe_old_rflags;
+ if (kcb->kprobe_status & KPROBE_HIT_SS) {
+ resume_execution(cur, regs, kcb);
+ regs->eflags |= kcb->kprobe_old_rflags;
- unlock_kprobes();
+ reset_current_kprobe();
preempt_enable_no_resched();
}
return 0;
@@ -599,39 +591,41 @@ int __kprobes kprobe_exceptions_notify(struct notifier_block *self,
unsigned long val, void *data)
{
struct die_args *args = (struct die_args *)data;
+ int ret = NOTIFY_DONE;
+
switch (val) {
case DIE_INT3:
if (kprobe_handler(args->regs))
- return NOTIFY_STOP;
+ ret = NOTIFY_STOP;
break;
case DIE_DEBUG:
if (post_kprobe_handler(args->regs))
- return NOTIFY_STOP;
+ ret = NOTIFY_STOP;
break;
case DIE_GPF:
- if (kprobe_running() &&
- kprobe_fault_handler(args->regs, args->trapnr))
- return NOTIFY_STOP;
- break;
case DIE_PAGE_FAULT:
+ /* kprobe_running() needs smp_processor_id() */
+ preempt_disable();
if (kprobe_running() &&
kprobe_fault_handler(args->regs, args->trapnr))
- return NOTIFY_STOP;
+ ret = NOTIFY_STOP;
+ preempt_enable();
break;
default:
break;
}
- return NOTIFY_DONE;
+ return ret;
}
int __kprobes setjmp_pre_handler(struct kprobe *p, struct pt_regs *regs)
{
struct jprobe *jp = container_of(p, struct jprobe, kp);
unsigned long addr;
+ struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
- jprobe_saved_regs = *regs;
- jprobe_saved_rsp = (long *) regs->rsp;
- addr = (unsigned long)jprobe_saved_rsp;
+ kcb->jprobe_saved_regs = *regs;
+ kcb->jprobe_saved_rsp = (long *) regs->rsp;
+ addr = (unsigned long)(kcb->jprobe_saved_rsp);
/*
* As Linus pointed out, gcc assumes that the callee
* owns the argument space and could overwrite it, e.g.
@@ -639,7 +633,8 @@ int __kprobes setjmp_pre_handler(struct kprobe *p, struct pt_regs *regs)
* we also save and restore enough stack bytes to cover
* the argument area.
*/
- memcpy(jprobes_stack, (kprobe_opcode_t *) addr, MIN_STACK_SIZE(addr));
+ memcpy(kcb->jprobes_stack, (kprobe_opcode_t *)addr,
+ MIN_STACK_SIZE(addr));
regs->eflags &= ~IF_MASK;
regs->rip = (unsigned long)(jp->entry);
return 1;
@@ -647,36 +642,40 @@ int __kprobes setjmp_pre_handler(struct kprobe *p, struct pt_regs *regs)
void __kprobes jprobe_return(void)
{
- preempt_enable_no_resched();
+ struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
+
asm volatile (" xchg %%rbx,%%rsp \n"
" int3 \n"
" .globl jprobe_return_end \n"
" jprobe_return_end: \n"
" nop \n"::"b"
- (jprobe_saved_rsp):"memory");
+ (kcb->jprobe_saved_rsp):"memory");
}
int __kprobes longjmp_break_handler(struct kprobe *p, struct pt_regs *regs)
{
+ struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
u8 *addr = (u8 *) (regs->rip - 1);
- unsigned long stack_addr = (unsigned long)jprobe_saved_rsp;
+ unsigned long stack_addr = (unsigned long)(kcb->jprobe_saved_rsp);
struct jprobe *jp = container_of(p, struct jprobe, kp);
if ((addr > (u8 *) jprobe_return) && (addr < (u8 *) jprobe_return_end)) {
- if ((long *)regs->rsp != jprobe_saved_rsp) {
+ if ((long *)regs->rsp != kcb->jprobe_saved_rsp) {
struct pt_regs *saved_regs =
- container_of(jprobe_saved_rsp, struct pt_regs, rsp);
+ container_of(kcb->jprobe_saved_rsp,
+ struct pt_regs, rsp);
printk("current rsp %p does not match saved rsp %p\n",
- (long *)regs->rsp, jprobe_saved_rsp);
+ (long *)regs->rsp, kcb->jprobe_saved_rsp);
printk("Saved registers for jprobe %p\n", jp);
show_registers(saved_regs);
printk("Current registers\n");
show_registers(regs);
BUG();
}
- *regs = jprobe_saved_regs;
- memcpy((kprobe_opcode_t *) stack_addr, jprobes_stack,
+ *regs = kcb->jprobe_saved_regs;
+ memcpy((kprobe_opcode_t *) stack_addr, kcb->jprobes_stack,
MIN_STACK_SIZE(stack_addr));
+ preempt_enable_no_resched();
return 1;
}
return 0;
diff --git a/arch/x86_64/kernel/process.c b/arch/x86_64/kernel/process.c
index bc44c92ca19..5afd63e8cef 100644
--- a/arch/x86_64/kernel/process.c
+++ b/arch/x86_64/kernel/process.c
@@ -86,12 +86,22 @@ EXPORT_SYMBOL(enable_hlt);
*/
void default_idle(void)
{
+ local_irq_enable();
+
if (!atomic_read(&hlt_counter)) {
- local_irq_disable();
- if (!need_resched())
- safe_halt();
- else
- local_irq_enable();
+ clear_thread_flag(TIF_POLLING_NRFLAG);
+ smp_mb__after_clear_bit();
+ while (!need_resched()) {
+ local_irq_disable();
+ if (!need_resched())
+ safe_halt();
+ else
+ local_irq_enable();
+ }
+ set_thread_flag(TIF_POLLING_NRFLAG);
+ } else {
+ while (!need_resched())
+ cpu_relax();
}
}
@@ -102,30 +112,16 @@ void default_idle(void)
*/
static void poll_idle (void)
{
- int oldval;
-
local_irq_enable();
- /*
- * Deal with another CPU just having chosen a thread to
- * run here:
- */
- oldval = test_and_clear_thread_flag(TIF_NEED_RESCHED);
-
- if (!oldval) {
- set_thread_flag(TIF_POLLING_NRFLAG);
- asm volatile(
- "2:"
- "testl %0,%1;"
- "rep; nop;"
- "je 2b;"
- : :
- "i" (_TIF_NEED_RESCHED),
- "m" (current_thread_info()->flags));
- clear_thread_flag(TIF_POLLING_NRFLAG);
- } else {
- set_need_resched();
- }
+ asm volatile(
+ "2:"
+ "testl %0,%1;"
+ "rep; nop;"
+ "je 2b;"
+ : :
+ "i" (_TIF_NEED_RESCHED),
+ "m" (current_thread_info()->flags));
}
void cpu_idle_wait(void)
@@ -188,6 +184,8 @@ static inline void play_dead(void)
*/
void cpu_idle (void)
{
+ set_thread_flag(TIF_POLLING_NRFLAG);
+
/* endless idle loop with no priority at all */
while (1) {
while (!need_resched()) {
@@ -205,7 +203,9 @@ void cpu_idle (void)
idle();
}
+ preempt_enable_no_resched();
schedule();
+ preempt_disable();
}
}
@@ -220,15 +220,12 @@ static void mwait_idle(void)
{
local_irq_enable();
- if (!need_resched()) {
- set_thread_flag(TIF_POLLING_NRFLAG);
- do {
- __monitor((void *)&current_thread_info()->flags, 0, 0);
- if (need_resched())
- break;
- __mwait(0, 0);
- } while (!need_resched());
- clear_thread_flag(TIF_POLLING_NRFLAG);
+ while (!need_resched()) {
+ __monitor((void *)&current_thread_info()->flags, 0, 0);
+ smp_mb();
+ if (need_resched())
+ break;
+ __mwait(0, 0);
}
}
diff --git a/arch/x86_64/kernel/ptrace.c b/arch/x86_64/kernel/ptrace.c
index bbf64b59a21..a87b6cebe80 100644
--- a/arch/x86_64/kernel/ptrace.c
+++ b/arch/x86_64/kernel/ptrace.c
@@ -313,48 +313,11 @@ static unsigned long getreg(struct task_struct *child, unsigned long regno)
}
-asmlinkage long sys_ptrace(long request, long pid, unsigned long addr, long data)
+long arch_ptrace(struct task_struct *child, long request, long addr, long data)
{
- struct task_struct *child;
long i, ret;
unsigned ui;
- /* This lock_kernel fixes a subtle race with suid exec */
- lock_kernel();
- ret = -EPERM;
- if (request == PTRACE_TRACEME) {
- /* are we already being traced? */
- if (current->ptrace & PT_PTRACED)
- goto out;
- ret = security_ptrace(current->parent, current);
- if (ret)
- goto out;
- /* set the ptrace bit in the process flags. */
- current->ptrace |= PT_PTRACED;
- ret = 0;
- goto out;
- }
- ret = -ESRCH;
- read_lock(&tasklist_lock);
- child = find_task_by_pid(pid);
- if (child)
- get_task_struct(child);
- read_unlock(&tasklist_lock);
- if (!child)
- goto out;
-
- ret = -EPERM;
- if (pid == 1) /* you may not mess with init */
- goto out_tsk;
-
- if (request == PTRACE_ATTACH) {
- ret = ptrace_attach(child);
- goto out_tsk;
- }
- ret = ptrace_check_attach(child, request == PTRACE_KILL);
- if (ret < 0)
- goto out_tsk;
-
switch (request) {
/* when I and D space are separate, these will need to be fixed. */
case PTRACE_PEEKTEXT: /* read word at location addr. */
@@ -608,10 +571,6 @@ asmlinkage long sys_ptrace(long request, long pid, unsigned long addr, long data
ret = ptrace_request(child, request, addr, data);
break;
}
-out_tsk:
- put_task_struct(child);
-out:
- unlock_kernel();
return ret;
}
diff --git a/arch/x86_64/kernel/smpboot.c b/arch/x86_64/kernel/smpboot.c
index 2b9ddba61b3..683c33f7b96 100644
--- a/arch/x86_64/kernel/smpboot.c
+++ b/arch/x86_64/kernel/smpboot.c
@@ -66,8 +66,6 @@ int smp_num_siblings = 1;
u8 phys_proc_id[NR_CPUS] __read_mostly = { [0 ... NR_CPUS-1] = BAD_APICID };
/* core ID of each logical CPU */
u8 cpu_core_id[NR_CPUS] __read_mostly = { [0 ... NR_CPUS-1] = BAD_APICID };
-EXPORT_SYMBOL(phys_proc_id);
-EXPORT_SYMBOL(cpu_core_id);
/* Bitmask of currently online CPUs */
cpumask_t cpu_online_map __read_mostly;
@@ -507,6 +505,7 @@ void __cpuinit start_secondary(void)
* things done here to the most necessary things.
*/
cpu_init();
+ preempt_disable();
smp_callin();
/* otherwise gcc will move up the smp_processor_id before the cpu_init */
generated by cgit v1.2.3 (git 2.25.1) at 2024-08-04 16:03:19 +0000