linux_dsm_epyc7002/arch/arm/kernel/kprobes.c

652 lines
18 KiB
C
Raw Normal View History

/*
* arch/arm/kernel/kprobes.c
*
* Kprobes on ARM
*
* Abhishek Sagar <sagar.abhishek@gmail.com>
* Copyright (C) 2006, 2007 Motorola Inc.
*
* Nicolas Pitre <nico@marvell.com>
* Copyright (C) 2007 Marvell Ltd.
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License version 2 as
* published by the Free Software Foundation.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* General Public License for more details.
*/
#include <linux/kernel.h>
#include <linux/kprobes.h>
#include <linux/module.h>
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h percpu.h is included by sched.h and module.h and thus ends up being included when building most .c files. percpu.h includes slab.h which in turn includes gfp.h making everything defined by the two files universally available and complicating inclusion dependencies. percpu.h -> slab.h dependency is about to be removed. Prepare for this change by updating users of gfp and slab facilities include those headers directly instead of assuming availability. As this conversion needs to touch large number of source files, the following script is used as the basis of conversion. http://userweb.kernel.org/~tj/misc/slabh-sweep.py The script does the followings. * Scan files for gfp and slab usages and update includes such that only the necessary includes are there. ie. if only gfp is used, gfp.h, if slab is used, slab.h. * When the script inserts a new include, it looks at the include blocks and try to put the new include such that its order conforms to its surrounding. It's put in the include block which contains core kernel includes, in the same order that the rest are ordered - alphabetical, Christmas tree, rev-Xmas-tree or at the end if there doesn't seem to be any matching order. * If the script can't find a place to put a new include (mostly because the file doesn't have fitting include block), it prints out an error message indicating which .h file needs to be added to the file. The conversion was done in the following steps. 1. The initial automatic conversion of all .c files updated slightly over 4000 files, deleting around 700 includes and adding ~480 gfp.h and ~3000 slab.h inclusions. The script emitted errors for ~400 files. 2. Each error was manually checked. Some didn't need the inclusion, some needed manual addition while adding it to implementation .h or embedding .c file was more appropriate for others. This step added inclusions to around 150 files. 3. The script was run again and the output was compared to the edits from #2 to make sure no file was left behind. 4. Several build tests were done and a couple of problems were fixed. e.g. lib/decompress_*.c used malloc/free() wrappers around slab APIs requiring slab.h to be added manually. 5. The script was run on all .h files but without automatically editing them as sprinkling gfp.h and slab.h inclusions around .h files could easily lead to inclusion dependency hell. Most gfp.h inclusion directives were ignored as stuff from gfp.h was usually wildly available and often used in preprocessor macros. Each slab.h inclusion directive was examined and added manually as necessary. 6. percpu.h was updated not to include slab.h. 7. Build test were done on the following configurations and failures were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my distributed build env didn't work with gcov compiles) and a few more options had to be turned off depending on archs to make things build (like ipr on powerpc/64 which failed due to missing writeq). * x86 and x86_64 UP and SMP allmodconfig and a custom test config. * powerpc and powerpc64 SMP allmodconfig * sparc and sparc64 SMP allmodconfig * ia64 SMP allmodconfig * s390 SMP allmodconfig * alpha SMP allmodconfig * um on x86_64 SMP allmodconfig 8. percpu.h modifications were reverted so that it could be applied as a separate patch and serve as bisection point. Given the fact that I had only a couple of failures from tests on step 6, I'm fairly confident about the coverage of this conversion patch. If there is a breakage, it's likely to be something in one of the arch headers which should be easily discoverable easily on most builds of the specific arch. Signed-off-by: Tejun Heo <tj@kernel.org> Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 15:04:11 +07:00
#include <linux/slab.h>
#include <linux/stop_machine.h>
#include <linux/stringify.h>
#include <asm/traps.h>
#include <asm/cacheflush.h>
#include "kprobes.h"
#define MIN_STACK_SIZE(addr) \
min((unsigned long)MAX_STACK_SIZE, \
(unsigned long)current_thread_info() + THREAD_START_SP - (addr))
#define flush_insns(addr, size) \
flush_icache_range((unsigned long)(addr), \
(unsigned long)(addr) + \
(size))
/* Used as a marker in ARM_pc to note when we're in a jprobe. */
#define JPROBE_MAGIC_ADDR 0xffffffff
DEFINE_PER_CPU(struct kprobe *, current_kprobe) = NULL;
DEFINE_PER_CPU(struct kprobe_ctlblk, kprobe_ctlblk);
int __kprobes arch_prepare_kprobe(struct kprobe *p)
{
kprobe_opcode_t insn;
kprobe_opcode_t tmp_insn[MAX_INSN_SIZE];
unsigned long addr = (unsigned long)p->addr;
kprobe_decode_insn_t *decode_insn;
int is;
if (in_exception_text(addr))
return -EINVAL;
#ifdef CONFIG_THUMB2_KERNEL
addr &= ~1; /* Bit 0 would normally be set to indicate Thumb code */
insn = ((u16 *)addr)[0];
if (is_wide_instruction(insn)) {
insn <<= 16;
insn |= ((u16 *)addr)[1];
decode_insn = thumb32_kprobe_decode_insn;
} else
decode_insn = thumb16_kprobe_decode_insn;
#else /* !CONFIG_THUMB2_KERNEL */
if (addr & 0x3)
return -EINVAL;
insn = *p->addr;
decode_insn = arm_kprobe_decode_insn;
#endif
p->opcode = insn;
p->ainsn.insn = tmp_insn;
switch ((*decode_insn)(insn, &p->ainsn)) {
case INSN_REJECTED: /* not supported */
return -EINVAL;
case INSN_GOOD: /* instruction uses slot */
p->ainsn.insn = get_insn_slot();
if (!p->ainsn.insn)
return -ENOMEM;
for (is = 0; is < MAX_INSN_SIZE; ++is)
p->ainsn.insn[is] = tmp_insn[is];
flush_insns(p->ainsn.insn,
sizeof(p->ainsn.insn[0]) * MAX_INSN_SIZE);
break;
case INSN_GOOD_NO_SLOT: /* instruction doesn't need insn slot */
p->ainsn.insn = NULL;
break;
}
return 0;
}
#ifdef CONFIG_THUMB2_KERNEL
/*
* For a 32-bit Thumb breakpoint spanning two memory words we need to take
* special precautions to insert the breakpoint atomically, especially on SMP
* systems. This is achieved by calling this arming function using stop_machine.
*/
static int __kprobes set_t32_breakpoint(void *addr)
{
((u16 *)addr)[0] = KPROBE_THUMB32_BREAKPOINT_INSTRUCTION >> 16;
((u16 *)addr)[1] = KPROBE_THUMB32_BREAKPOINT_INSTRUCTION & 0xffff;
flush_insns(addr, 2*sizeof(u16));
return 0;
}
void __kprobes arch_arm_kprobe(struct kprobe *p)
{
uintptr_t addr = (uintptr_t)p->addr & ~1; /* Remove any Thumb flag */
if (!is_wide_instruction(p->opcode)) {
*(u16 *)addr = KPROBE_THUMB16_BREAKPOINT_INSTRUCTION;
flush_insns(addr, sizeof(u16));
} else if (addr & 2) {
/* A 32-bit instruction spanning two words needs special care */
stop_machine(set_t32_breakpoint, (void *)addr, &cpu_online_map);
} else {
/* Word aligned 32-bit instruction can be written atomically */
u32 bkp = KPROBE_THUMB32_BREAKPOINT_INSTRUCTION;
#ifndef __ARMEB__ /* Swap halfwords for little-endian */
bkp = (bkp >> 16) | (bkp << 16);
#endif
*(u32 *)addr = bkp;
flush_insns(addr, sizeof(u32));
}
}
#else /* !CONFIG_THUMB2_KERNEL */
void __kprobes arch_arm_kprobe(struct kprobe *p)
{
kprobe_opcode_t insn = p->opcode;
kprobe_opcode_t brkp = KPROBE_ARM_BREAKPOINT_INSTRUCTION;
if (insn >= 0xe0000000)
brkp |= 0xe0000000; /* Unconditional instruction */
else
brkp |= insn & 0xf0000000; /* Copy condition from insn */
*p->addr = brkp;
flush_insns(p->addr, sizeof(p->addr[0]));
}
#endif /* !CONFIG_THUMB2_KERNEL */
/*
* The actual disarming is done here on each CPU and synchronized using
* stop_machine. This synchronization is necessary on SMP to avoid removing
* a probe between the moment the 'Undefined Instruction' exception is raised
* and the moment the exception handler reads the faulting instruction from
* memory. It is also needed to atomically set the two half-words of a 32-bit
* Thumb breakpoint.
*/
int __kprobes __arch_disarm_kprobe(void *p)
{
struct kprobe *kp = p;
#ifdef CONFIG_THUMB2_KERNEL
u16 *addr = (u16 *)((uintptr_t)kp->addr & ~1);
kprobe_opcode_t insn = kp->opcode;
unsigned int len;
if (is_wide_instruction(insn)) {
((u16 *)addr)[0] = insn>>16;
((u16 *)addr)[1] = insn;
len = 2*sizeof(u16);
} else {
((u16 *)addr)[0] = insn;
len = sizeof(u16);
}
flush_insns(addr, len);
#else /* !CONFIG_THUMB2_KERNEL */
*kp->addr = kp->opcode;
flush_insns(kp->addr, sizeof(kp->addr[0]));
#endif
return 0;
}
void __kprobes arch_disarm_kprobe(struct kprobe *p)
{
stop_machine(__arch_disarm_kprobe, p, &cpu_online_map);
}
void __kprobes arch_remove_kprobe(struct kprobe *p)
{
if (p->ainsn.insn) {
free_insn_slot(p->ainsn.insn, 0);
p->ainsn.insn = NULL;
}
}
static void __kprobes save_previous_kprobe(struct kprobe_ctlblk *kcb)
{
kcb->prev_kprobe.kp = kprobe_running();
kcb->prev_kprobe.status = kcb->kprobe_status;
}
static void __kprobes restore_previous_kprobe(struct kprobe_ctlblk *kcb)
{
__get_cpu_var(current_kprobe) = kcb->prev_kprobe.kp;
kcb->kprobe_status = kcb->prev_kprobe.status;
}
static void __kprobes set_current_kprobe(struct kprobe *p)
{
__get_cpu_var(current_kprobe) = p;
}
static void __kprobes
singlestep_skip(struct kprobe *p, struct pt_regs *regs)
{
#ifdef CONFIG_THUMB2_KERNEL
regs->ARM_cpsr = it_advance(regs->ARM_cpsr);
if (is_wide_instruction(p->opcode))
regs->ARM_pc += 4;
else
regs->ARM_pc += 2;
#else
regs->ARM_pc += 4;
#endif
}
static inline void __kprobes
singlestep(struct kprobe *p, struct pt_regs *regs, struct kprobe_ctlblk *kcb)
{
p->ainsn.insn_singlestep(p, regs);
}
/*
* Called with IRQs disabled. IRQs must remain disabled from that point
* all the way until processing this kprobe is complete. The current
* kprobes implementation cannot process more than one nested level of
* kprobe, and that level is reserved for user kprobe handlers, so we can't
* risk encountering a new kprobe in an interrupt handler.
*/
void __kprobes kprobe_handler(struct pt_regs *regs)
{
struct kprobe *p, *cur;
struct kprobe_ctlblk *kcb;
kcb = get_kprobe_ctlblk();
cur = kprobe_running();
#ifdef CONFIG_THUMB2_KERNEL
/*
* First look for a probe which was registered using an address with
* bit 0 set, this is the usual situation for pointers to Thumb code.
* If not found, fallback to looking for one with bit 0 clear.
*/
p = get_kprobe((kprobe_opcode_t *)(regs->ARM_pc | 1));
if (!p)
p = get_kprobe((kprobe_opcode_t *)regs->ARM_pc);
#else /* ! CONFIG_THUMB2_KERNEL */
p = get_kprobe((kprobe_opcode_t *)regs->ARM_pc);
#endif
if (p) {
if (cur) {
/* Kprobe is pending, so we're recursing. */
switch (kcb->kprobe_status) {
case KPROBE_HIT_ACTIVE:
case KPROBE_HIT_SSDONE:
/* A pre- or post-handler probe got us here. */
kprobes_inc_nmissed_count(p);
save_previous_kprobe(kcb);
set_current_kprobe(p);
kcb->kprobe_status = KPROBE_REENTER;
singlestep(p, regs, kcb);
restore_previous_kprobe(kcb);
break;
default:
/* impossible cases */
BUG();
}
} else if (p->ainsn.insn_check_cc(regs->ARM_cpsr)) {
/* Probe hit and conditional execution check ok. */
set_current_kprobe(p);
kcb->kprobe_status = KPROBE_HIT_ACTIVE;
/*
* If we have no pre-handler or it returned 0, we
* continue with normal processing. If we have a
* pre-handler and it returned non-zero, it prepped
* for calling the break_handler below on re-entry,
* so get out doing nothing more here.
*/
if (!p->pre_handler || !p->pre_handler(p, regs)) {
kcb->kprobe_status = KPROBE_HIT_SS;
singlestep(p, regs, kcb);
if (p->post_handler) {
kcb->kprobe_status = KPROBE_HIT_SSDONE;
p->post_handler(p, regs, 0);
}
reset_current_kprobe();
}
} else {
/*
* Probe hit but conditional execution check failed,
* so just skip the instruction and continue as if
* nothing had happened.
*/
singlestep_skip(p, regs);
}
} else if (cur) {
/* We probably hit a jprobe. Call its break handler. */
if (cur->break_handler && cur->break_handler(cur, regs)) {
kcb->kprobe_status = KPROBE_HIT_SS;
singlestep(cur, regs, kcb);
if (cur->post_handler) {
kcb->kprobe_status = KPROBE_HIT_SSDONE;
cur->post_handler(cur, regs, 0);
}
}
reset_current_kprobe();
} else {
/*
* The probe was removed and a race is in progress.
* There is nothing we can do about it. Let's restart
* the instruction. By the time we can restart, the
* real instruction will be there.
*/
}
}
static int __kprobes kprobe_trap_handler(struct pt_regs *regs, unsigned int instr)
{
unsigned long flags;
local_irq_save(flags);
kprobe_handler(regs);
local_irq_restore(flags);
return 0;
}
int __kprobes kprobe_fault_handler(struct pt_regs *regs, unsigned int fsr)
{
struct kprobe *cur = kprobe_running();
struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
switch (kcb->kprobe_status) {
case KPROBE_HIT_SS:
case KPROBE_REENTER:
/*
* We are here because the instruction being single
* stepped caused a page fault. We reset the current
* kprobe and the PC to point back to the probe address
* and allow the page fault handler to continue as a
* normal page fault.
*/
regs->ARM_pc = (long)cur->addr;
if (kcb->kprobe_status == KPROBE_REENTER) {
restore_previous_kprobe(kcb);
} else {
reset_current_kprobe();
}
break;
case KPROBE_HIT_ACTIVE:
case KPROBE_HIT_SSDONE:
/*
* We increment the nmissed count for accounting,
* we can also use npre/npostfault count for accounting
* these specific fault cases.
*/
kprobes_inc_nmissed_count(cur);
/*
* We come here because instructions in the pre/post
* handler caused the page_fault, this could happen
* if handler tries to access user space by
* copy_from_user(), get_user() etc. Let the
* user-specified handler try to fix it.
*/
if (cur->fault_handler && cur->fault_handler(cur, regs, fsr))
return 1;
break;
default:
break;
}
return 0;
}
int __kprobes kprobe_exceptions_notify(struct notifier_block *self,
unsigned long val, void *data)
{
/*
* notify_die() is currently never called on ARM,
* so this callback is currently empty.
*/
return NOTIFY_DONE;
}
/*
* When a retprobed function returns, trampoline_handler() is called,
* calling the kretprobe's handler. We construct a struct pt_regs to
* give a view of registers r0-r11 to the user return-handler. This is
* not a complete pt_regs structure, but that should be plenty sufficient
* for kretprobe handlers which should normally be interested in r0 only
* anyway.
*/
void __naked __kprobes kretprobe_trampoline(void)
{
__asm__ __volatile__ (
"stmdb sp!, {r0 - r11} \n\t"
"mov r0, sp \n\t"
"bl trampoline_handler \n\t"
"mov lr, r0 \n\t"
"ldmia sp!, {r0 - r11} \n\t"
#ifdef CONFIG_THUMB2_KERNEL
"bx lr \n\t"
#else
"mov pc, lr \n\t"
#endif
: : : "memory");
}
/* Called from kretprobe_trampoline */
static __used __kprobes void *trampoline_handler(struct pt_regs *regs)
{
struct kretprobe_instance *ri = NULL;
struct hlist_head *head, empty_rp;
struct hlist_node *node, *tmp;
unsigned long flags, orig_ret_address = 0;
unsigned long trampoline_address = (unsigned long)&kretprobe_trampoline;
INIT_HLIST_HEAD(&empty_rp);
kprobes: improve kretprobe scalability with hashed locking Currently list of kretprobe instances are stored in kretprobe object (as used_instances,free_instances) and in kretprobe hash table. We have one global kretprobe lock to serialise the access to these lists. This causes only one kretprobe handler to execute at a time. Hence affects system performance, particularly on SMP systems and when return probe is set on lot of functions (like on all systemcalls). Solution proposed here gives fine-grain locks that performs better on SMP system compared to present kretprobe implementation. Solution: 1) Instead of having one global lock to protect kretprobe instances present in kretprobe object and kretprobe hash table. We will have two locks, one lock for protecting kretprobe hash table and another lock for kretporbe object. 2) We hold lock present in kretprobe object while we modify kretprobe instance in kretprobe object and we hold per-hash-list lock while modifying kretprobe instances present in that hash list. To prevent deadlock, we never grab a per-hash-list lock while holding a kretprobe lock. 3) We can remove used_instances from struct kretprobe, as we can track used instances of kretprobe instances using kretprobe hash table. Time duration for kernel compilation ("make -j 8") on a 8-way ppc64 system with return probes set on all systemcalls looks like this. cacheline non-cacheline Un-patched kernel aligned patch aligned patch =============================================================================== real 9m46.784s 9m54.412s 10m2.450s user 40m5.715s 40m7.142s 40m4.273s sys 2m57.754s 2m58.583s 3m17.430s =========================================================== Time duration for kernel compilation ("make -j 8) on the same system, when kernel is not probed. ========================= real 9m26.389s user 40m8.775s sys 2m7.283s ========================= Signed-off-by: Srinivasa DS <srinivasa@in.ibm.com> Signed-off-by: Jim Keniston <jkenisto@us.ibm.com> Acked-by: Ananth N Mavinakayanahalli <ananth@in.ibm.com> Cc: Anil S Keshavamurthy <anil.s.keshavamurthy@intel.com> Cc: David S. Miller <davem@davemloft.net> Cc: Masami Hiramatsu <mhiramat@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-25 15:46:04 +07:00
kretprobe_hash_lock(current, &head, &flags);
/*
* It is possible to have multiple instances associated with a given
* task either because multiple functions in the call path have
* a return probe installed on them, and/or more than one 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) {
__get_cpu_var(current_kprobe) = &ri->rp->kp;
get_kprobe_ctlblk()->kprobe_status = KPROBE_HIT_ACTIVE;
ri->rp->handler(ri, regs);
__get_cpu_var(current_kprobe) = NULL;
}
orig_ret_address = (unsigned long)ri->ret_addr;
recycle_rp_inst(ri, &empty_rp);
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;
}
kretprobe_assert(ri, orig_ret_address, trampoline_address);
kprobes: improve kretprobe scalability with hashed locking Currently list of kretprobe instances are stored in kretprobe object (as used_instances,free_instances) and in kretprobe hash table. We have one global kretprobe lock to serialise the access to these lists. This causes only one kretprobe handler to execute at a time. Hence affects system performance, particularly on SMP systems and when return probe is set on lot of functions (like on all systemcalls). Solution proposed here gives fine-grain locks that performs better on SMP system compared to present kretprobe implementation. Solution: 1) Instead of having one global lock to protect kretprobe instances present in kretprobe object and kretprobe hash table. We will have two locks, one lock for protecting kretprobe hash table and another lock for kretporbe object. 2) We hold lock present in kretprobe object while we modify kretprobe instance in kretprobe object and we hold per-hash-list lock while modifying kretprobe instances present in that hash list. To prevent deadlock, we never grab a per-hash-list lock while holding a kretprobe lock. 3) We can remove used_instances from struct kretprobe, as we can track used instances of kretprobe instances using kretprobe hash table. Time duration for kernel compilation ("make -j 8") on a 8-way ppc64 system with return probes set on all systemcalls looks like this. cacheline non-cacheline Un-patched kernel aligned patch aligned patch =============================================================================== real 9m46.784s 9m54.412s 10m2.450s user 40m5.715s 40m7.142s 40m4.273s sys 2m57.754s 2m58.583s 3m17.430s =========================================================== Time duration for kernel compilation ("make -j 8) on the same system, when kernel is not probed. ========================= real 9m26.389s user 40m8.775s sys 2m7.283s ========================= Signed-off-by: Srinivasa DS <srinivasa@in.ibm.com> Signed-off-by: Jim Keniston <jkenisto@us.ibm.com> Acked-by: Ananth N Mavinakayanahalli <ananth@in.ibm.com> Cc: Anil S Keshavamurthy <anil.s.keshavamurthy@intel.com> Cc: David S. Miller <davem@davemloft.net> Cc: Masami Hiramatsu <mhiramat@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-25 15:46:04 +07:00
kretprobe_hash_unlock(current, &flags);
hlist_for_each_entry_safe(ri, node, tmp, &empty_rp, hlist) {
hlist_del(&ri->hlist);
kfree(ri);
}
return (void *)orig_ret_address;
}
void __kprobes arch_prepare_kretprobe(struct kretprobe_instance *ri,
struct pt_regs *regs)
{
ri->ret_addr = (kprobe_opcode_t *)regs->ARM_lr;
/* Replace the return addr with trampoline addr. */
regs->ARM_lr = (unsigned long)&kretprobe_trampoline;
}
int __kprobes setjmp_pre_handler(struct kprobe *p, struct pt_regs *regs)
{
struct jprobe *jp = container_of(p, struct jprobe, kp);
struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
long sp_addr = regs->ARM_sp;
long cpsr;
kcb->jprobe_saved_regs = *regs;
memcpy(kcb->jprobes_stack, (void *)sp_addr, MIN_STACK_SIZE(sp_addr));
regs->ARM_pc = (long)jp->entry;
cpsr = regs->ARM_cpsr | PSR_I_BIT;
#ifdef CONFIG_THUMB2_KERNEL
/* Set correct Thumb state in cpsr */
if (regs->ARM_pc & 1)
cpsr |= PSR_T_BIT;
else
cpsr &= ~PSR_T_BIT;
#endif
regs->ARM_cpsr = cpsr;
preempt_disable();
return 1;
}
void __kprobes jprobe_return(void)
{
struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
__asm__ __volatile__ (
/*
* Setup an empty pt_regs. Fill SP and PC fields as
* they're needed by longjmp_break_handler.
ARM: 6005/1: arm: kprobes: fix register corruption with jprobes Current implementation of jprobes allocates empty pt_regs from the stack which is then passed to kprobe_handler() and eventually to singlestep(). Now when instruction being simulated is STMFD (like in normal function prologues without CONFIG_FRAME_POINTER), stores using SP actually write over top of the fabricated pt_regs structure. This can be reproduced for example by using LKDTM module: # modprobe lkdtm # mount -t debugfs none /sys/kernel/debug # echo PANIC > /sys/kernel/debug/provoke-crash/INT_HW_IRQ_EN after this, it fails with corrupted registers (before the requested crash would occur): lkdtm: Crash point INT_HW_IRQ_EN of type PANIC hit, trigger in 9 rounds lkdtm: Crash point INT_HW_IRQ_EN of type PANIC hit, trigger in 8 rounds Internal error: Oops - undefined instruction: 0 [#1] last sysfs file: /sys/devices/platform/serial8250.0/sleep_timeout Modules linked in: lkdtm CPU: 0 Not tainted (2.6.34-rc2 #69) PC is at irq_desc+0x1638/0xeeb0 LR is at 0x25 pc : [<c050b428>] lr : [<00000025>] psr: c80a0013 sp : ce94bd60 ip : c050b3e8 fp : a0000013 r10: c0aa453c r9 : cf5d4000 r8 : ce9a1822 r7 : c050b424 r6 : 00000025 r5 : c039d8f8 r4 : c050b3e8 r3 : 00000001 r2 : cf4d0440 r1 : c039d8f8 r0 : 00000020 Flags: NZcv IRQs on FIQs on Mode SVC_32 ISA ARM Segment user Control: 10c5387d Table: 8e804019 DAC: 00000015 Process sh (pid: 496, stack limit = 0xce94a2e8) Stack: (0xce94bd60 to 0xce94c000) [...] Code: 000002cd 00000000 00000000 00000001 (dead4ead) ---[ end trace 2b46d5f2b682f370 ]--- Kernel panic - not syncing: Fatal exception in interrupt This patch allocates enough space (2 * sizeof(struct pt_regs)) from the stack to prevent such corruption. Signed-off-by: Mika Westerberg <ext-mika.1.westerberg@nokia.com> Acked-by: Nicolas Pitre <nico@marvell.com> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
2010-03-29 12:59:16 +07:00
*
* We allocate some slack between the original SP and start of
* our fabricated regs. To be precise we want to have worst case
* covered which is STMFD with all 16 regs so we allocate 2 *
* sizeof(struct_pt_regs)).
*
* This is to prevent any simulated instruction from writing
* over the regs when they are accessing the stack.
*/
#ifdef CONFIG_THUMB2_KERNEL
"sub r0, %0, %1 \n\t"
"mov sp, r0 \n\t"
#else
"sub sp, %0, %1 \n\t"
#endif
"ldr r0, ="__stringify(JPROBE_MAGIC_ADDR)"\n\t"
"str %0, [sp, %2] \n\t"
"str r0, [sp, %3] \n\t"
"mov r0, sp \n\t"
"bl kprobe_handler \n\t"
/*
* Return to the context saved by setjmp_pre_handler
* and restored by longjmp_break_handler.
*/
#ifdef CONFIG_THUMB2_KERNEL
"ldr lr, [sp, %2] \n\t" /* lr = saved sp */
"ldrd r0, r1, [sp, %5] \n\t" /* r0,r1 = saved lr,pc */
"ldr r2, [sp, %4] \n\t" /* r2 = saved psr */
"stmdb lr!, {r0, r1, r2} \n\t" /* push saved lr and */
/* rfe context */
"ldmia sp, {r0 - r12} \n\t"
"mov sp, lr \n\t"
"ldr lr, [sp], #4 \n\t"
"rfeia sp! \n\t"
#else
"ldr r0, [sp, %4] \n\t"
"msr cpsr_cxsf, r0 \n\t"
"ldmia sp, {r0 - pc} \n\t"
#endif
:
: "r" (kcb->jprobe_saved_regs.ARM_sp),
ARM: 6005/1: arm: kprobes: fix register corruption with jprobes Current implementation of jprobes allocates empty pt_regs from the stack which is then passed to kprobe_handler() and eventually to singlestep(). Now when instruction being simulated is STMFD (like in normal function prologues without CONFIG_FRAME_POINTER), stores using SP actually write over top of the fabricated pt_regs structure. This can be reproduced for example by using LKDTM module: # modprobe lkdtm # mount -t debugfs none /sys/kernel/debug # echo PANIC > /sys/kernel/debug/provoke-crash/INT_HW_IRQ_EN after this, it fails with corrupted registers (before the requested crash would occur): lkdtm: Crash point INT_HW_IRQ_EN of type PANIC hit, trigger in 9 rounds lkdtm: Crash point INT_HW_IRQ_EN of type PANIC hit, trigger in 8 rounds Internal error: Oops - undefined instruction: 0 [#1] last sysfs file: /sys/devices/platform/serial8250.0/sleep_timeout Modules linked in: lkdtm CPU: 0 Not tainted (2.6.34-rc2 #69) PC is at irq_desc+0x1638/0xeeb0 LR is at 0x25 pc : [<c050b428>] lr : [<00000025>] psr: c80a0013 sp : ce94bd60 ip : c050b3e8 fp : a0000013 r10: c0aa453c r9 : cf5d4000 r8 : ce9a1822 r7 : c050b424 r6 : 00000025 r5 : c039d8f8 r4 : c050b3e8 r3 : 00000001 r2 : cf4d0440 r1 : c039d8f8 r0 : 00000020 Flags: NZcv IRQs on FIQs on Mode SVC_32 ISA ARM Segment user Control: 10c5387d Table: 8e804019 DAC: 00000015 Process sh (pid: 496, stack limit = 0xce94a2e8) Stack: (0xce94bd60 to 0xce94c000) [...] Code: 000002cd 00000000 00000000 00000001 (dead4ead) ---[ end trace 2b46d5f2b682f370 ]--- Kernel panic - not syncing: Fatal exception in interrupt This patch allocates enough space (2 * sizeof(struct pt_regs)) from the stack to prevent such corruption. Signed-off-by: Mika Westerberg <ext-mika.1.westerberg@nokia.com> Acked-by: Nicolas Pitre <nico@marvell.com> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
2010-03-29 12:59:16 +07:00
"I" (sizeof(struct pt_regs) * 2),
"J" (offsetof(struct pt_regs, ARM_sp)),
"J" (offsetof(struct pt_regs, ARM_pc)),
"J" (offsetof(struct pt_regs, ARM_cpsr)),
"J" (offsetof(struct pt_regs, ARM_lr))
: "memory", "cc");
}
int __kprobes longjmp_break_handler(struct kprobe *p, struct pt_regs *regs)
{
struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
long stack_addr = kcb->jprobe_saved_regs.ARM_sp;
long orig_sp = regs->ARM_sp;
struct jprobe *jp = container_of(p, struct jprobe, kp);
if (regs->ARM_pc == JPROBE_MAGIC_ADDR) {
if (orig_sp != stack_addr) {
struct pt_regs *saved_regs =
(struct pt_regs *)kcb->jprobe_saved_regs.ARM_sp;
printk("current sp %lx does not match saved sp %lx\n",
orig_sp, stack_addr);
printk("Saved registers for jprobe %p\n", jp);
show_regs(saved_regs);
printk("Current registers\n");
show_regs(regs);
BUG();
}
*regs = kcb->jprobe_saved_regs;
memcpy((void *)stack_addr, kcb->jprobes_stack,
MIN_STACK_SIZE(stack_addr));
preempt_enable_no_resched();
return 1;
}
return 0;
}
int __kprobes arch_trampoline_kprobe(struct kprobe *p)
{
return 0;
}
#ifdef CONFIG_THUMB2_KERNEL
static struct undef_hook kprobes_thumb16_break_hook = {
.instr_mask = 0xffff,
.instr_val = KPROBE_THUMB16_BREAKPOINT_INSTRUCTION,
.cpsr_mask = MODE_MASK,
.cpsr_val = SVC_MODE,
.fn = kprobe_trap_handler,
};
static struct undef_hook kprobes_thumb32_break_hook = {
.instr_mask = 0xffffffff,
.instr_val = KPROBE_THUMB32_BREAKPOINT_INSTRUCTION,
.cpsr_mask = MODE_MASK,
.cpsr_val = SVC_MODE,
.fn = kprobe_trap_handler,
};
#else /* !CONFIG_THUMB2_KERNEL */
static struct undef_hook kprobes_arm_break_hook = {
.instr_mask = 0x0fffffff,
.instr_val = KPROBE_ARM_BREAKPOINT_INSTRUCTION,
.cpsr_mask = MODE_MASK,
.cpsr_val = SVC_MODE,
.fn = kprobe_trap_handler,
};
#endif /* !CONFIG_THUMB2_KERNEL */
int __init arch_init_kprobes()
{
arm_kprobe_decode_init();
#ifdef CONFIG_THUMB2_KERNEL
register_undef_hook(&kprobes_thumb16_break_hook);
register_undef_hook(&kprobes_thumb32_break_hook);
#else
register_undef_hook(&kprobes_arm_break_hook);
#endif
return 0;
}