* Rusty Russell).
* 2004-July Suparna Bhattacharya <suparna@in.ibm.com> added jumper probes
* interface to access function arguments.
+ * 2005-May Hien Nguyen <hien@us.ibm.com>, Jim Keniston
+ * <jkenisto@us.ibm.com> and Prasanna S Panchamukhi
+ * <prasanna@in.ibm.com> added function-return probes.
*/
#include <linux/config.h>
#include <linux/kprobes.h>
#include <linux/ptrace.h>
-#include <linux/spinlock.h>
#include <linux/preempt.h>
+#include <asm/cacheflush.h>
#include <asm/kdebug.h>
#include <asm/desc.h>
-/* kprobe_status settings */
-#define KPROBE_HIT_ACTIVE 0x00000001
-#define KPROBE_HIT_SS 0x00000002
-
-static struct kprobe *current_kprobe;
-static unsigned long kprobe_status, kprobe_old_eflags, kprobe_saved_eflags;
-static struct pt_regs jprobe_saved_regs;
-static long *jprobe_saved_esp;
-/* copy of the kernel stack at the probe fire time */
-static kprobe_opcode_t jprobes_stack[MAX_STACK_SIZE];
void jprobe_return_end(void);
+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.
*/
return 0;
}
-int arch_prepare_kprobe(struct kprobe *p)
+int __kprobes arch_prepare_kprobe(struct kprobe *p)
{
+ /* insn: must be on special executable page on i386. */
+ p->ainsn.insn = get_insn_slot();
+ if (!p->ainsn.insn)
+ return -ENOMEM;
+
+ memcpy(p->ainsn.insn, p->addr, MAX_INSN_SIZE * sizeof(kprobe_opcode_t));
+ p->opcode = *p->addr;
return 0;
}
-void arch_copy_kprobe(struct kprobe *p)
+void __kprobes arch_arm_kprobe(struct kprobe *p)
{
- memcpy(p->ainsn.insn, p->addr, MAX_INSN_SIZE * sizeof(kprobe_opcode_t));
+ *p->addr = BREAKPOINT_INSTRUCTION;
+ flush_icache_range((unsigned long) p->addr,
+ (unsigned long) p->addr + sizeof(kprobe_opcode_t));
}
-void arch_remove_kprobe(struct kprobe *p)
+void __kprobes arch_disarm_kprobe(struct kprobe *p)
{
+ *p->addr = p->opcode;
+ flush_icache_range((unsigned long) p->addr,
+ (unsigned long) p->addr + sizeof(kprobe_opcode_t));
}
-static inline void disarm_kprobe(struct kprobe *p, struct pt_regs *regs)
+void __kprobes arch_remove_kprobe(struct kprobe *p)
{
- *p->addr = p->opcode;
- regs->eip = (unsigned long)p->addr;
+ down(&kprobe_mutex);
+ free_insn_slot(p->ainsn.insn);
+ up(&kprobe_mutex);
+}
+
+static inline void save_previous_kprobe(struct kprobe_ctlblk *kcb)
+{
+ kcb->prev_kprobe.kp = kprobe_running();
+ kcb->prev_kprobe.status = kcb->kprobe_status;
+ kcb->prev_kprobe.old_eflags = kcb->kprobe_old_eflags;
+ kcb->prev_kprobe.saved_eflags = kcb->kprobe_saved_eflags;
+}
+
+static inline void restore_previous_kprobe(struct kprobe_ctlblk *kcb)
+{
+ __get_cpu_var(current_kprobe) = kcb->prev_kprobe.kp;
+ kcb->kprobe_status = kcb->prev_kprobe.status;
+ kcb->kprobe_old_eflags = kcb->prev_kprobe.old_eflags;
+ kcb->kprobe_saved_eflags = kcb->prev_kprobe.saved_eflags;
+}
+
+static inline void set_current_kprobe(struct kprobe *p, struct pt_regs *regs,
+ struct kprobe_ctlblk *kcb)
+{
+ __get_cpu_var(current_kprobe) = p;
+ kcb->kprobe_saved_eflags = kcb->kprobe_old_eflags
+ = (regs->eflags & (TF_MASK | IF_MASK));
+ if (is_IF_modifier(p->opcode))
+ kcb->kprobe_saved_eflags &= ~IF_MASK;
}
static inline void prepare_singlestep(struct kprobe *p, struct pt_regs *regs)
{
regs->eflags |= TF_MASK;
regs->eflags &= ~IF_MASK;
- regs->eip = (unsigned long)&p->ainsn.insn;
+ /*single step inline if the instruction is an int3*/
+ if (p->opcode == BREAKPOINT_INSTRUCTION)
+ regs->eip = (unsigned long)p->addr;
+ else
+ regs->eip = (unsigned long)p->ainsn.insn;
+}
+
+/* Called with kretprobe_lock held */
+void __kprobes arch_prepare_kretprobe(struct kretprobe *rp,
+ struct pt_regs *regs)
+{
+ unsigned long *sara = (unsigned long *)®s->esp;
+ struct kretprobe_instance *ri;
+
+ if ((ri = get_free_rp_inst(rp)) != NULL) {
+ ri->rp = rp;
+ ri->task = current;
+ ri->ret_addr = (kprobe_opcode_t *) *sara;
+
+ /* Replace the return addr with trampoline addr */
+ *sara = (unsigned long) &kretprobe_trampoline;
+
+ add_rp_inst(ri);
+ } else {
+ rp->nmissed++;
+ }
}
/*
* Interrupts are disabled on entry as trap3 is an interrupt gate and they
* remain disabled thorough out this function.
*/
-static int kprobe_handler(struct pt_regs *regs)
+static int __kprobes kprobe_handler(struct pt_regs *regs)
{
struct kprobe *p;
int ret = 0;
kprobe_opcode_t *addr = NULL;
unsigned long *lp;
+ 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 if the application is using LDT entry for its code segment and
* calculate the address by reading the base address from the LDT entry.
*/
}
/* 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) {
- disarm_kprobe(p, regs);
- ret = 1;
+ if (kcb->kprobe_status == KPROBE_HIT_SS &&
+ *p->ainsn.insn == BREAKPOINT_INSTRUCTION) {
+ regs->eflags &= ~TF_MASK;
+ regs->eflags |= kcb->kprobe_saved_eflags;
+ goto no_kprobe;
+ }
+ /* We have reentered the kprobe_handler(), since
+ * another probe was hit while within the handler.
+ * We here save the original kprobes variables and
+ * just single step on the instruction of the new probe
+ * without calling any user handlers.
+ */
+ save_previous_kprobe(kcb);
+ set_current_kprobe(p, regs, kcb);
+ kprobes_inc_nmissed_count(p);
+ prepare_singlestep(p, regs);
+ kcb->kprobe_status = KPROBE_REENTER;
+ return 1;
} else {
- p = current_kprobe;
+ if (regs->eflags & VM_MASK) {
+ /* We are in virtual-8086 mode. Return 0 */
+ goto no_kprobe;
+ }
+ if (*addr != BREAKPOINT_INSTRUCTION) {
+ /* The breakpoint instruction was removed by
+ * another cpu right after we hit, no further
+ * handling of this interrupt is appropriate
+ */
+ regs->eip -= sizeof(kprobe_opcode_t);
+ ret = 1;
+ goto no_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 (regs->eflags & VM_MASK) {
/* We are in virtual-8086 mode. Return 0 */
goto no_kprobe;
* either a probepoint or a debugger breakpoint
* at this address. In either case, no further
* handling of this interrupt is appropriate.
+ * Back up over the (now missing) int3 and run
+ * the original instruction.
*/
+ regs->eip -= sizeof(kprobe_opcode_t);
ret = 1;
}
/* Not one of ours: let kernel handle it */
goto no_kprobe;
}
- kprobe_status = KPROBE_HIT_ACTIVE;
- current_kprobe = p;
- kprobe_saved_eflags = kprobe_old_eflags
- = (regs->eflags & (TF_MASK | IF_MASK));
- if (is_IF_modifier(p->opcode))
- kprobe_saved_eflags &= ~IF_MASK;
+ set_current_kprobe(p, regs, kcb);
+ kcb->kprobe_status = KPROBE_HIT_ACTIVE;
- if (p->pre_handler(p, regs)) {
+ if (p->pre_handler && p->pre_handler(p, regs))
/* handler has already set things up, so skip ss setup */
return 1;
- }
- ss_probe:
+ss_probe:
prepare_singlestep(p, regs);
- kprobe_status = KPROBE_HIT_SS;
+ kcb->kprobe_status = KPROBE_HIT_SS;
return 1;
- no_kprobe:
+no_kprobe:
preempt_enable_no_resched();
return ret;
}
+/*
+ * For function-return probes, init_kprobes() establishes a probepoint
+ * here. When a retprobed function returns, this probe is hit and
+ * trampoline_probe_handler() runs, calling the kretprobe's handler.
+ */
+ void kretprobe_trampoline_holder(void)
+ {
+ asm volatile ( ".global kretprobe_trampoline\n"
+ "kretprobe_trampoline: \n"
+ "nop\n");
+ }
+
+/*
+ * Called when we hit the probe point at kretprobe_trampoline
+ */
+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 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);
+
+ /*
+ * It is possible to have multiple instances associated with a given
+ * task either because an multiple functions in the call path
+ * have a return probe installed on them, and/or more then one return
+ * return probe was registered for a target function.
+ *
+ * We can handle this because:
+ * - instances are always inserted at the head of the list
+ * - when multiple return probes are registered for the same
+ * function, the first instance's ret_addr will point to the
+ * real return address, and all the rest will point to
+ * kretprobe_trampoline
+ */
+ hlist_for_each_entry_safe(ri, node, tmp, head, hlist) {
+ if (ri->task != current)
+ /* another task is sharing our hash bucket */
+ continue;
+
+ if (ri->rp && ri->rp->handler)
+ ri->rp->handler(ri, regs);
+
+ orig_ret_address = (unsigned long)ri->ret_addr;
+ recycle_rp_inst(ri);
+
+ if (orig_ret_address != trampoline_address)
+ /*
+ * This is the real return address. Any other
+ * instances associated with this task are for
+ * other calls deeper on the call stack
+ */
+ break;
+ }
+
+ BUG_ON(!orig_ret_address || (orig_ret_address == trampoline_address));
+ regs->eip = orig_ret_address;
+
+ 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 don't want the post_handler
+ * to run (and have re-enabled preemption)
+ */
+ return 1;
+}
+
/*
* Called after single-stepping. p->addr is the address of the
* instruction whose first byte has been replaced by the "int 3"
* 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 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 *)®s->esp;
unsigned long next_eip = 0;
- unsigned long copy_eip = (unsigned long)&p->ainsn.insn;
+ unsigned long copy_eip = (unsigned long)p->ainsn.insn;
unsigned long orig_eip = (unsigned long)p->addr;
switch (p->ainsn.insn[0]) {
case 0x9c: /* pushfl */
*tos &= ~(TF_MASK | IF_MASK);
- *tos |= kprobe_old_eflags;
+ *tos |= kcb->kprobe_old_eflags;
break;
+ case 0xc3: /* ret/lret */
+ case 0xcb:
+ case 0xc2:
+ case 0xca:
+ regs->eflags &= ~TF_MASK;
+ /* eip is already adjusted, no more changes required*/
+ return;
case 0xe8: /* call relative - Fix return addr */
*tos = orig_eip + (*tos - copy_eip);
break;
/*
* Interrupts are disabled on entry as trap1 is an interrupt gate and they
- * remain disabled thoroughout this function. And we hold kprobe lock.
+ * remain disabled thoroughout this function.
*/
static inline int 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 (current_kprobe->post_handler)
- 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_eflags;
+ resume_execution(cur, regs, kcb);
+ regs->eflags |= kcb->kprobe_saved_eflags;
- unlock_kprobes();
+ /*Restore back the original saved kprobes variables and continue. */
+ if (kcb->kprobe_status == KPROBE_REENTER) {
+ restore_previous_kprobe(kcb);
+ goto out;
+ }
+ reset_current_kprobe();
+out:
preempt_enable_no_resched();
/*
return 1;
}
-/* Interrupts disabled, kprobe_lock held. */
static inline int 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_eflags;
+ if (kcb->kprobe_status & KPROBE_HIT_SS) {
+ resume_execution(cur, regs, kcb);
+ regs->eflags |= kcb->kprobe_old_eflags;
- unlock_kprobes();
+ reset_current_kprobe();
preempt_enable_no_resched();
}
return 0;
/*
* Wrapper routine to for handling exceptions.
*/
-int kprobe_exceptions_notify(struct notifier_block *self, unsigned long val,
- void *data)
+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 setjmp_pre_handler(struct kprobe *p, struct pt_regs *regs)
+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_esp = ®s->esp;
- addr = (unsigned long)jprobe_saved_esp;
+ kcb->jprobe_saved_regs = *regs;
+ kcb->jprobe_saved_esp = ®s->esp;
+ addr = (unsigned long)(kcb->jprobe_saved_esp);
/*
* TBD: As Linus pointed out, gcc assumes that the callee
* 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->eip = (unsigned long)(jp->entry);
return 1;
}
-void jprobe_return(void)
+void __kprobes jprobe_return(void)
{
- preempt_enable_no_resched();
+ struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
+
asm volatile (" xchgl %%ebx,%%esp \n"
" int3 \n"
" .globl jprobe_return_end \n"
" jprobe_return_end: \n"
" nop \n"::"b"
- (jprobe_saved_esp):"memory");
+ (kcb->jprobe_saved_esp):"memory");
}
-int longjmp_break_handler(struct kprobe *p, struct pt_regs *regs)
+int __kprobes longjmp_break_handler(struct kprobe *p, struct pt_regs *regs)
{
+ struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
u8 *addr = (u8 *) (regs->eip - 1);
- unsigned long stack_addr = (unsigned long)jprobe_saved_esp;
+ unsigned long stack_addr = (unsigned long)(kcb->jprobe_saved_esp);
struct jprobe *jp = container_of(p, struct jprobe, kp);
if ((addr > (u8 *) jprobe_return) && (addr < (u8 *) jprobe_return_end)) {
- if (®s->esp != jprobe_saved_esp) {
+ if (®s->esp != kcb->jprobe_saved_esp) {
struct pt_regs *saved_regs =
- container_of(jprobe_saved_esp, struct pt_regs, esp);
+ container_of(kcb->jprobe_saved_esp,
+ struct pt_regs, esp);
printk("current esp %p does not match saved esp %p\n",
- ®s->esp, jprobe_saved_esp);
+ ®s->esp, kcb->jprobe_saved_esp);
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;
}
+
+static struct kprobe trampoline_p = {
+ .addr = (kprobe_opcode_t *) &kretprobe_trampoline,
+ .pre_handler = trampoline_probe_handler
+};
+
+int __init arch_init_kprobes(void)
+{
+ return register_kprobe(&trampoline_p);
+}
git://git.onelab.eu / linux-2.6.git / blobdiff