linux/arch/x86/mm/fault.c

943 lines
23 KiB
C
Raw Normal View History

/*
* Copyright (C) 1995 Linus Torvalds
* Copyright (C) 2001,2002 Andi Kleen, SuSE Labs.
*/
#include <linux/signal.h>
#include <linux/sched.h>
#include <linux/kernel.h>
#include <linux/errno.h>
#include <linux/string.h>
#include <linux/types.h>
#include <linux/ptrace.h>
x86: mmiotrace full patch, preview 1 kmmio.c handles the list of mmio probes with callbacks, list of traced pages, and attaching into the page fault handler and die notifier. It arms, traps and disarms the given pages, this is the core of mmiotrace. mmio-mod.c is a user interface, hooking into ioremap functions and registering the mmio probes. It also decodes the required information from trapped mmio accesses via the pre and post callbacks in each probe. Currently, hooking into ioremap functions works by redefining the symbols of the target (binary) kernel module, so that it calls the traced versions of the functions. The most notable changes done since the last discussion are: - kmmio.c is a built-in, not part of the module - direct call from fault.c to kmmio.c, removing all dynamic hooks - prepare for unregistering probes at any time - make kmmio re-initializable and accessible to more than one user - rewrite kmmio locking to remove all spinlocks from page fault path Can I abuse call_rcu() like I do in kmmio.c:unregister_kmmio_probe() or is there a better way? The function called via call_rcu() itself calls call_rcu() again, will this work or break? There I need a second grace period for RCU after the first grace period for page faults. Mmiotrace itself (mmio-mod.c) is still a module, I am going to attack that next. At some point I will start looking into how to make mmiotrace a tracer component of ftrace (thanks for the hint, Ingo). Ftrace should make the user space part of mmiotracing as simple as 'cat /debug/trace/mmio > dump.txt'. Signed-off-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2008-05-12 19:20:57 +00:00
#include <linux/mmiotrace.h>
#include <linux/mman.h>
#include <linux/mm.h>
#include <linux/smp.h>
#include <linux/interrupt.h>
#include <linux/init.h>
#include <linux/tty.h>
#include <linux/vt_kern.h> /* For unblank_screen() */
#include <linux/compiler.h>
#include <linux/highmem.h>
#include <linux/bootmem.h> /* for max_low_pfn */
#include <linux/vmalloc.h>
#include <linux/module.h>
#include <linux/kprobes.h>
#include <linux/uaccess.h>
#include <linux/kdebug.h>
#include <asm/system.h>
#include <asm/desc.h>
#include <asm/segment.h>
#include <asm/pgalloc.h>
#include <asm/smp.h>
#include <asm/tlbflush.h>
#include <asm/proto.h>
#include <asm-generic/sections.h>
#include <asm/traps.h>
/*
* Page fault error code bits
* bit 0 == 0 means no page found, 1 means protection fault
* bit 1 == 0 means read, 1 means write
* bit 2 == 0 means kernel, 1 means user-mode
* bit 3 == 1 means use of reserved bit detected
* bit 4 == 1 means fault was an instruction fetch
*/
#define PF_PROT (1<<0)
#define PF_WRITE (1<<1)
#define PF_USER (1<<2)
#define PF_RSVD (1<<3)
#define PF_INSTR (1<<4)
x86: mmiotrace full patch, preview 1 kmmio.c handles the list of mmio probes with callbacks, list of traced pages, and attaching into the page fault handler and die notifier. It arms, traps and disarms the given pages, this is the core of mmiotrace. mmio-mod.c is a user interface, hooking into ioremap functions and registering the mmio probes. It also decodes the required information from trapped mmio accesses via the pre and post callbacks in each probe. Currently, hooking into ioremap functions works by redefining the symbols of the target (binary) kernel module, so that it calls the traced versions of the functions. The most notable changes done since the last discussion are: - kmmio.c is a built-in, not part of the module - direct call from fault.c to kmmio.c, removing all dynamic hooks - prepare for unregistering probes at any time - make kmmio re-initializable and accessible to more than one user - rewrite kmmio locking to remove all spinlocks from page fault path Can I abuse call_rcu() like I do in kmmio.c:unregister_kmmio_probe() or is there a better way? The function called via call_rcu() itself calls call_rcu() again, will this work or break? There I need a second grace period for RCU after the first grace period for page faults. Mmiotrace itself (mmio-mod.c) is still a module, I am going to attack that next. At some point I will start looking into how to make mmiotrace a tracer component of ftrace (thanks for the hint, Ingo). Ftrace should make the user space part of mmiotracing as simple as 'cat /debug/trace/mmio > dump.txt'. Signed-off-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2008-05-12 19:20:57 +00:00
static inline int kmmio_fault(struct pt_regs *regs, unsigned long addr)
{
#ifdef CONFIG_MMIOTRACE_HOOKS
x86: mmiotrace full patch, preview 1 kmmio.c handles the list of mmio probes with callbacks, list of traced pages, and attaching into the page fault handler and die notifier. It arms, traps and disarms the given pages, this is the core of mmiotrace. mmio-mod.c is a user interface, hooking into ioremap functions and registering the mmio probes. It also decodes the required information from trapped mmio accesses via the pre and post callbacks in each probe. Currently, hooking into ioremap functions works by redefining the symbols of the target (binary) kernel module, so that it calls the traced versions of the functions. The most notable changes done since the last discussion are: - kmmio.c is a built-in, not part of the module - direct call from fault.c to kmmio.c, removing all dynamic hooks - prepare for unregistering probes at any time - make kmmio re-initializable and accessible to more than one user - rewrite kmmio locking to remove all spinlocks from page fault path Can I abuse call_rcu() like I do in kmmio.c:unregister_kmmio_probe() or is there a better way? The function called via call_rcu() itself calls call_rcu() again, will this work or break? There I need a second grace period for RCU after the first grace period for page faults. Mmiotrace itself (mmio-mod.c) is still a module, I am going to attack that next. At some point I will start looking into how to make mmiotrace a tracer component of ftrace (thanks for the hint, Ingo). Ftrace should make the user space part of mmiotracing as simple as 'cat /debug/trace/mmio > dump.txt'. Signed-off-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2008-05-12 19:20:57 +00:00
if (unlikely(is_kmmio_active()))
if (kmmio_handler(regs, addr) == 1)
return -1;
#endif
x86: mmiotrace full patch, preview 1 kmmio.c handles the list of mmio probes with callbacks, list of traced pages, and attaching into the page fault handler and die notifier. It arms, traps and disarms the given pages, this is the core of mmiotrace. mmio-mod.c is a user interface, hooking into ioremap functions and registering the mmio probes. It also decodes the required information from trapped mmio accesses via the pre and post callbacks in each probe. Currently, hooking into ioremap functions works by redefining the symbols of the target (binary) kernel module, so that it calls the traced versions of the functions. The most notable changes done since the last discussion are: - kmmio.c is a built-in, not part of the module - direct call from fault.c to kmmio.c, removing all dynamic hooks - prepare for unregistering probes at any time - make kmmio re-initializable and accessible to more than one user - rewrite kmmio locking to remove all spinlocks from page fault path Can I abuse call_rcu() like I do in kmmio.c:unregister_kmmio_probe() or is there a better way? The function called via call_rcu() itself calls call_rcu() again, will this work or break? There I need a second grace period for RCU after the first grace period for page faults. Mmiotrace itself (mmio-mod.c) is still a module, I am going to attack that next. At some point I will start looking into how to make mmiotrace a tracer component of ftrace (thanks for the hint, Ingo). Ftrace should make the user space part of mmiotracing as simple as 'cat /debug/trace/mmio > dump.txt'. Signed-off-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2008-05-12 19:20:57 +00:00
return 0;
}
static inline int notify_page_fault(struct pt_regs *regs)
[PATCH] Notify page fault call chain for x86_64 Currently in the do_page_fault() code path, we call notify_die(DIE_PAGE_FAULT, ...) to notify the page fault. Since notify_die() is highly overloaded, this page fault notification is currently being sent to all the components registered with register_die_notification() which uses the same die_chain to loop for all the registered components which is unnecessary. In order to optimize the do_page_fault() code path, this critical page fault notification is now moved to different call chain and the test results showed great improvements. And the kprobes which is interested in this notifications, now registers onto this new call chain only when it need to, i.e Kprobes now registers for page fault notification only when their are an active probes and unregisters from this page fault notification when no probes are active. I have incorporated all the feedback given by Ananth and Keith and everyone, and thanks for all the review feedback. This patch: Overloading of page fault notification with the notify_die() has performance issues(since the only interested components for page fault is kprobes and/or kdb) and hence this patch introduces the new notifier call chain exclusively for page fault notifications their by avoiding notifying unnecessary components in the do_page_fault() code path. Signed-off-by: Anil S Keshavamurthy <anil.s.keshavamurthy@intel.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-06-26 07:25:25 +00:00
{
#ifdef CONFIG_KPROBES
int ret = 0;
/* kprobe_running() needs smp_processor_id() */
if (!user_mode_vm(regs)) {
preempt_disable();
if (kprobe_running() && kprobe_fault_handler(regs, 14))
ret = 1;
preempt_enable();
}
[PATCH] Notify page fault call chain for x86_64 Currently in the do_page_fault() code path, we call notify_die(DIE_PAGE_FAULT, ...) to notify the page fault. Since notify_die() is highly overloaded, this page fault notification is currently being sent to all the components registered with register_die_notification() which uses the same die_chain to loop for all the registered components which is unnecessary. In order to optimize the do_page_fault() code path, this critical page fault notification is now moved to different call chain and the test results showed great improvements. And the kprobes which is interested in this notifications, now registers onto this new call chain only when it need to, i.e Kprobes now registers for page fault notification only when their are an active probes and unregisters from this page fault notification when no probes are active. I have incorporated all the feedback given by Ananth and Keith and everyone, and thanks for all the review feedback. This patch: Overloading of page fault notification with the notify_die() has performance issues(since the only interested components for page fault is kprobes and/or kdb) and hence this patch introduces the new notifier call chain exclusively for page fault notifications their by avoiding notifying unnecessary components in the do_page_fault() code path. Signed-off-by: Anil S Keshavamurthy <anil.s.keshavamurthy@intel.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-06-26 07:25:25 +00:00
return ret;
#else
return 0;
#endif
}
[PATCH] Notify page fault call chain for x86_64 Currently in the do_page_fault() code path, we call notify_die(DIE_PAGE_FAULT, ...) to notify the page fault. Since notify_die() is highly overloaded, this page fault notification is currently being sent to all the components registered with register_die_notification() which uses the same die_chain to loop for all the registered components which is unnecessary. In order to optimize the do_page_fault() code path, this critical page fault notification is now moved to different call chain and the test results showed great improvements. And the kprobes which is interested in this notifications, now registers onto this new call chain only when it need to, i.e Kprobes now registers for page fault notification only when their are an active probes and unregisters from this page fault notification when no probes are active. I have incorporated all the feedback given by Ananth and Keith and everyone, and thanks for all the review feedback. This patch: Overloading of page fault notification with the notify_die() has performance issues(since the only interested components for page fault is kprobes and/or kdb) and hence this patch introduces the new notifier call chain exclusively for page fault notifications their by avoiding notifying unnecessary components in the do_page_fault() code path. Signed-off-by: Anil S Keshavamurthy <anil.s.keshavamurthy@intel.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-06-26 07:25:25 +00:00
/*
* X86_32
* Sometimes AMD Athlon/Opteron CPUs report invalid exceptions on prefetch.
* Check that here and ignore it.
*
* X86_64
* Sometimes the CPU reports invalid exceptions on prefetch.
* Check that here and ignore it.
*
* Opcode checker based on code by Richard Brunner
*/
static int is_prefetch(struct pt_regs *regs, unsigned long addr,
unsigned long error_code)
{
unsigned char *instr;
int scan_more = 1;
int prefetch = 0;
unsigned char *max_instr;
/*
* If it was a exec (instruction fetch) fault on NX page, then
* do not ignore the fault:
*/
if (error_code & PF_INSTR)
return 0;
instr = (unsigned char *)convert_ip_to_linear(current, regs);
max_instr = instr + 15;
if (user_mode(regs) && instr >= (unsigned char *)TASK_SIZE)
return 0;
while (scan_more && instr < max_instr) {
unsigned char opcode;
unsigned char instr_hi;
unsigned char instr_lo;
if (probe_kernel_address(instr, opcode))
break;
instr_hi = opcode & 0xf0;
instr_lo = opcode & 0x0f;
instr++;
switch (instr_hi) {
case 0x20:
case 0x30:
/*
* Values 0x26,0x2E,0x36,0x3E are valid x86 prefixes.
* In X86_64 long mode, the CPU will signal invalid
* opcode if some of these prefixes are present so
* X86_64 will never get here anyway
*/
scan_more = ((instr_lo & 7) == 0x6);
break;
#ifdef CONFIG_X86_64
case 0x40:
/*
* In AMD64 long mode 0x40..0x4F are valid REX prefixes
* Need to figure out under what instruction mode the
* instruction was issued. Could check the LDT for lm,
* but for now it's good enough to assume that long
* mode only uses well known segments or kernel.
*/
scan_more = (!user_mode(regs)) || (regs->cs == __USER_CS);
break;
#endif
case 0x60:
/* 0x64 thru 0x67 are valid prefixes in all modes. */
scan_more = (instr_lo & 0xC) == 0x4;
break;
case 0xF0:
/* 0xF0, 0xF2, 0xF3 are valid prefixes in all modes. */
scan_more = !instr_lo || (instr_lo>>1) == 1;
break;
case 0x00:
/* Prefetch instruction is 0x0F0D or 0x0F18 */
scan_more = 0;
if (probe_kernel_address(instr, opcode))
break;
prefetch = (instr_lo == 0xF) &&
(opcode == 0x0D || opcode == 0x18);
break;
default:
scan_more = 0;
break;
}
}
return prefetch;
}
static void force_sig_info_fault(int si_signo, int si_code,
unsigned long address, struct task_struct *tsk)
{
siginfo_t info;
info.si_signo = si_signo;
info.si_errno = 0;
info.si_code = si_code;
info.si_addr = (void __user *)address;
force_sig_info(si_signo, &info, tsk);
}
#ifdef CONFIG_X86_64
static int bad_address(void *p)
{
unsigned long dummy;
return probe_kernel_address((unsigned long *)p, dummy);
}
#endif
static void dump_pagetable(unsigned long address)
{
#ifdef CONFIG_X86_32
__typeof__(pte_val(__pte(0))) page;
page = read_cr3();
page = ((__typeof__(page) *) __va(page))[address >> PGDIR_SHIFT];
#ifdef CONFIG_X86_PAE
printk("*pdpt = %016Lx ", page);
if ((page >> PAGE_SHIFT) < max_low_pfn
&& page & _PAGE_PRESENT) {
page &= PAGE_MASK;
page = ((__typeof__(page) *) __va(page))[(address >> PMD_SHIFT)
& (PTRS_PER_PMD - 1)];
printk(KERN_CONT "*pde = %016Lx ", page);
page &= ~_PAGE_NX;
}
#else
printk("*pde = %08lx ", page);
#endif
/*
* We must not directly access the pte in the highpte
* case if the page table is located in highmem.
* And let's rather not kmap-atomic the pte, just in case
* it's allocated already.
*/
if ((page >> PAGE_SHIFT) < max_low_pfn
&& (page & _PAGE_PRESENT)
&& !(page & _PAGE_PSE)) {
page &= PAGE_MASK;
page = ((__typeof__(page) *) __va(page))[(address >> PAGE_SHIFT)
& (PTRS_PER_PTE - 1)];
printk("*pte = %0*Lx ", sizeof(page)*2, (u64)page);
}
printk("\n");
#else /* CONFIG_X86_64 */
pgd_t *pgd;
pud_t *pud;
pmd_t *pmd;
pte_t *pte;
pgd = (pgd_t *)read_cr3();
pgd = __va((unsigned long)pgd & PHYSICAL_PAGE_MASK);
pgd += pgd_index(address);
if (bad_address(pgd)) goto bad;
printk("PGD %lx ", pgd_val(*pgd));
if (!pgd_present(*pgd)) goto ret;
pud = pud_offset(pgd, address);
if (bad_address(pud)) goto bad;
printk("PUD %lx ", pud_val(*pud));
if (!pud_present(*pud) || pud_large(*pud))
goto ret;
pmd = pmd_offset(pud, address);
if (bad_address(pmd)) goto bad;
printk("PMD %lx ", pmd_val(*pmd));
if (!pmd_present(*pmd) || pmd_large(*pmd)) goto ret;
pte = pte_offset_kernel(pmd, address);
if (bad_address(pte)) goto bad;
printk("PTE %lx", pte_val(*pte));
ret:
printk("\n");
return;
bad:
printk("BAD\n");
#endif
}
#ifdef CONFIG_X86_32
static inline pmd_t *vmalloc_sync_one(pgd_t *pgd, unsigned long address)
{
unsigned index = pgd_index(address);
pgd_t *pgd_k;
pud_t *pud, *pud_k;
pmd_t *pmd, *pmd_k;
pgd += index;
pgd_k = init_mm.pgd + index;
if (!pgd_present(*pgd_k))
return NULL;
/*
* set_pgd(pgd, *pgd_k); here would be useless on PAE
* and redundant with the set_pmd() on non-PAE. As would
* set_pud.
*/
pud = pud_offset(pgd, address);
pud_k = pud_offset(pgd_k, address);
if (!pud_present(*pud_k))
return NULL;
pmd = pmd_offset(pud, address);
pmd_k = pmd_offset(pud_k, address);
if (!pmd_present(*pmd_k))
return NULL;
if (!pmd_present(*pmd)) {
set_pmd(pmd, *pmd_k);
arch_flush_lazy_mmu_mode();
} else
BUG_ON(pmd_page(*pmd) != pmd_page(*pmd_k));
return pmd_k;
}
#endif
#ifdef CONFIG_X86_64
static const char errata93_warning[] =
KERN_ERR "******* Your BIOS seems to not contain a fix for K8 errata #93\n"
KERN_ERR "******* Working around it, but it may cause SEGVs or burn power.\n"
KERN_ERR "******* Please consider a BIOS update.\n"
KERN_ERR "******* Disabling USB legacy in the BIOS may also help.\n";
#endif
/* Workaround for K8 erratum #93 & buggy BIOS.
BIOS SMM functions are required to use a specific workaround
to avoid corruption of the 64bit RIP register on C stepping K8.
A lot of BIOS that didn't get tested properly miss this.
The OS sees this as a page fault with the upper 32bits of RIP cleared.
Try to work around it here.
Note we only handle faults in kernel here.
Does nothing for X86_32
*/
static int is_errata93(struct pt_regs *regs, unsigned long address)
{
#ifdef CONFIG_X86_64
static int warned;
if (address != regs->ip)
return 0;
if ((address >> 32) != 0)
return 0;
address |= 0xffffffffUL << 32;
if ((address >= (u64)_stext && address <= (u64)_etext) ||
(address >= MODULES_VADDR && address <= MODULES_END)) {
if (!warned) {
printk(errata93_warning);
warned = 1;
}
regs->ip = address;
return 1;
}
#endif
return 0;
}
/*
* Work around K8 erratum #100 K8 in compat mode occasionally jumps to illegal
* addresses >4GB. We catch this in the page fault handler because these
* addresses are not reachable. Just detect this case and return. Any code
* segment in LDT is compatibility mode.
*/
static int is_errata100(struct pt_regs *regs, unsigned long address)
{
#ifdef CONFIG_X86_64
if ((regs->cs == __USER32_CS || (regs->cs & (1<<2))) &&
(address >> 32))
return 1;
#endif
return 0;
}
static int is_f00f_bug(struct pt_regs *regs, unsigned long address)
{
#ifdef CONFIG_X86_F00F_BUG
unsigned long nr;
/*
* Pentium F0 0F C7 C8 bug workaround.
*/
if (boot_cpu_data.f00f_bug) {
nr = (address - idt_descr.address) >> 3;
if (nr == 6) {
do_invalid_op(regs, 0);
return 1;
}
}
#endif
return 0;
}
static void show_fault_oops(struct pt_regs *regs, unsigned long error_code,
unsigned long address)
{
#ifdef CONFIG_X86_32
if (!oops_may_print())
return;
#endif
#ifdef CONFIG_X86_PAE
if (error_code & PF_INSTR) {
unsigned int level;
pte_t *pte = lookup_address(address, &level);
if (pte && pte_present(*pte) && !pte_exec(*pte))
printk(KERN_CRIT "kernel tried to execute "
"NX-protected page - exploit attempt? "
"(uid: %d)\n", current->uid);
}
#endif
printk(KERN_ALERT "BUG: unable to handle kernel ");
if (address < PAGE_SIZE)
printk(KERN_CONT "NULL pointer dereference");
else
printk(KERN_CONT "paging request");
printk(KERN_CONT " at %p\n", (void *) address);
printk(KERN_ALERT "IP:");
printk_address(regs->ip, 1);
dump_pagetable(address);
}
#ifdef CONFIG_X86_64
static noinline void pgtable_bad(unsigned long address, struct pt_regs *regs,
unsigned long error_code)
{
unsigned long flags = oops_begin();
struct task_struct *tsk;
printk(KERN_ALERT "%s: Corrupted page table at address %lx\n",
current->comm, address);
dump_pagetable(address);
tsk = current;
tsk->thread.cr2 = address;
tsk->thread.trap_no = 14;
tsk->thread.error_code = error_code;
if (__die("Bad pagetable", regs, error_code))
regs = NULL;
oops_end(flags, regs, SIGKILL);
}
#endif
static int spurious_fault_check(unsigned long error_code, pte_t *pte)
{
if ((error_code & PF_WRITE) && !pte_write(*pte))
return 0;
if ((error_code & PF_INSTR) && !pte_exec(*pte))
return 0;
return 1;
}
/*
* Handle a spurious fault caused by a stale TLB entry. This allows
* us to lazily refresh the TLB when increasing the permissions of a
* kernel page (RO -> RW or NX -> X). Doing it eagerly is very
* expensive since that implies doing a full cross-processor TLB
* flush, even if no stale TLB entries exist on other processors.
* There are no security implications to leaving a stale TLB when
* increasing the permissions on a page.
*/
static int spurious_fault(unsigned long address,
unsigned long error_code)
{
pgd_t *pgd;
pud_t *pud;
pmd_t *pmd;
pte_t *pte;
/* Reserved-bit violation or user access to kernel space? */
if (error_code & (PF_USER | PF_RSVD))
return 0;
pgd = init_mm.pgd + pgd_index(address);
if (!pgd_present(*pgd))
return 0;
pud = pud_offset(pgd, address);
if (!pud_present(*pud))
return 0;
if (pud_large(*pud))
return spurious_fault_check(error_code, (pte_t *) pud);
pmd = pmd_offset(pud, address);
if (!pmd_present(*pmd))
return 0;
if (pmd_large(*pmd))
return spurious_fault_check(error_code, (pte_t *) pmd);
pte = pte_offset_kernel(pmd, address);
if (!pte_present(*pte))
return 0;
return spurious_fault_check(error_code, pte);
}
/*
* X86_32
* Handle a fault on the vmalloc or module mapping area
*
* X86_64
* Handle a fault on the vmalloc area
*
* This assumes no large pages in there.
*/
static int vmalloc_fault(unsigned long address)
{
#ifdef CONFIG_X86_32
unsigned long pgd_paddr;
pmd_t *pmd_k;
pte_t *pte_k;
x86: fix endless page faults in mount_block_root for Linux 2.6 Page faults in kernel address space between PAGE_OFFSET up to VMALLOC_START should not try to map as vmalloc. Fix rarely endless page faults inside mount_block_root for root filesystem at boot time. All 32bit kernels up to 2.6.25 can fail into this hole. I can not present this under native linux kernel. I see, that the 64bit has fixed the problem. I copied the same lines into 32bit part. Recorded debugs are from coLinux kernel 2.6.22.18 (virtualisation): http://www.henrynestler.com/colinux/testing/pfn-check-0.7.3/20080410-antinx/bug16-recursive-page-fault-endless.txt The physicaly memory was trimmed down to 192MB to better catch the bug. More memory gets the bug more rarely. Details, how every x86 32bit system can fail: Start from "mount_block_root", http://lxr.linux.no/linux/init/do_mounts.c#L297 There the variable "fs_names" got one memory page with 4096 bytes. Variable "p" walks through the existing file system types. The first string is no problem. But, with the second loop in mount_block_root the offset of "p" is not at beginning of page, the offset is for example +9, if "reiserfs" is the first in list. Than calls do_mount_root, and lands in sys_mount. Remember: Variable "type_page" contains now "fs_type+9" and not contains a full page. The sys_mount copies 4096 bytes with function "exact_copy_from_user()": http://lxr.linux.no/linux/fs/namespace.c#L1540 Mostly exist pages after the buffer "fs_names+4096+9" and the page fault handler was not called. No problem. In the case, if the page after "fs_names+4096" is not mapped, the page fault handler was called from http://lxr.linux.no/linux/fs/namespace.c#L1320 The do_page_fault gots an address 0xc03b4000. It's kernel address, address >= TASK_SIZE, but not from vmalloc! It's from "__getname()" alias "kmem_cache_alloc". The "error_code" is 0. "vmalloc_fault" will be call: http://lxr.linux.no/linux/arch/i386/mm/fault.c#L332 "vmalloc_fault" tryed to find the physical page for a non existing virtual memory area. The macro "pte_present" in vmalloc_fault() got a next page fault for 0xc0000ed0 at: http://lxr.linux.no/linux/arch/i386/mm/fault.c#L282 No PTE exist for such virtual address. The page fault handler was trying to sync the physical page for the PTE lockup. This called vmalloc_fault() again for address 0xc000000, and that also was not existing. The endless began... In normal case the cpu would still loop with disabled interrrupts. Under coLinux this was catched by a stack overflow inside printk debugs. Signed-off-by: Henry Nestler <henry.nestler@gmail.com> Signed-off-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2008-05-12 13:44:39 +00:00
/* Make sure we are in vmalloc area */
if (!(address >= VMALLOC_START && address < VMALLOC_END))
return -1;
/*
* Synchronize this task's top level page-table
* with the 'reference' page table.
*
* Do _not_ use "current" here. We might be inside
* an interrupt in the middle of a task switch..
*/
pgd_paddr = read_cr3();
pmd_k = vmalloc_sync_one(__va(pgd_paddr), address);
if (!pmd_k)
return -1;
pte_k = pte_offset_kernel(pmd_k, address);
if (!pte_present(*pte_k))
return -1;
return 0;
#else
pgd_t *pgd, *pgd_ref;
pud_t *pud, *pud_ref;
pmd_t *pmd, *pmd_ref;
pte_t *pte, *pte_ref;
/* Make sure we are in vmalloc area */
if (!(address >= VMALLOC_START && address < VMALLOC_END))
return -1;
/* Copy kernel mappings over when needed. This can also
happen within a race in page table update. In the later
case just flush. */
pgd = pgd_offset(current->mm ?: &init_mm, address);
pgd_ref = pgd_offset_k(address);
if (pgd_none(*pgd_ref))
return -1;
if (pgd_none(*pgd))
set_pgd(pgd, *pgd_ref);
else
BUG_ON(pgd_page_vaddr(*pgd) != pgd_page_vaddr(*pgd_ref));
/* Below here mismatches are bugs because these lower tables
are shared */
pud = pud_offset(pgd, address);
pud_ref = pud_offset(pgd_ref, address);
if (pud_none(*pud_ref))
return -1;
if (pud_none(*pud) || pud_page_vaddr(*pud) != pud_page_vaddr(*pud_ref))
BUG();
pmd = pmd_offset(pud, address);
pmd_ref = pmd_offset(pud_ref, address);
if (pmd_none(*pmd_ref))
return -1;
if (pmd_none(*pmd) || pmd_page(*pmd) != pmd_page(*pmd_ref))
BUG();
pte_ref = pte_offset_kernel(pmd_ref, address);
if (!pte_present(*pte_ref))
return -1;
pte = pte_offset_kernel(pmd, address);
/* Don't use pte_page here, because the mappings can point
outside mem_map, and the NUMA hash lookup cannot handle
that. */
if (!pte_present(*pte) || pte_pfn(*pte) != pte_pfn(*pte_ref))
BUG();
return 0;
#endif
}
int show_unhandled_signals = 1;
/*
* This routine handles page faults. It determines the address,
* and the problem, and then passes it off to one of the appropriate
* routines.
*/
#ifdef CONFIG_X86_64
asmlinkage
#endif
void __kprobes do_page_fault(struct pt_regs *regs, unsigned long error_code)
{
struct task_struct *tsk;
struct mm_struct *mm;
struct vm_area_struct *vma;
unsigned long address;
int write, si_code;
int fault;
#ifdef CONFIG_X86_64
unsigned long flags;
#endif
tsk = current;
mm = tsk->mm;
prefetchw(&mm->mmap_sem);
/* get the address */
address = read_cr2();
si_code = SEGV_MAPERR;
if (notify_page_fault(regs))
return;
x86: mmiotrace full patch, preview 1 kmmio.c handles the list of mmio probes with callbacks, list of traced pages, and attaching into the page fault handler and die notifier. It arms, traps and disarms the given pages, this is the core of mmiotrace. mmio-mod.c is a user interface, hooking into ioremap functions and registering the mmio probes. It also decodes the required information from trapped mmio accesses via the pre and post callbacks in each probe. Currently, hooking into ioremap functions works by redefining the symbols of the target (binary) kernel module, so that it calls the traced versions of the functions. The most notable changes done since the last discussion are: - kmmio.c is a built-in, not part of the module - direct call from fault.c to kmmio.c, removing all dynamic hooks - prepare for unregistering probes at any time - make kmmio re-initializable and accessible to more than one user - rewrite kmmio locking to remove all spinlocks from page fault path Can I abuse call_rcu() like I do in kmmio.c:unregister_kmmio_probe() or is there a better way? The function called via call_rcu() itself calls call_rcu() again, will this work or break? There I need a second grace period for RCU after the first grace period for page faults. Mmiotrace itself (mmio-mod.c) is still a module, I am going to attack that next. At some point I will start looking into how to make mmiotrace a tracer component of ftrace (thanks for the hint, Ingo). Ftrace should make the user space part of mmiotracing as simple as 'cat /debug/trace/mmio > dump.txt'. Signed-off-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2008-05-12 19:20:57 +00:00
if (unlikely(kmmio_fault(regs, address)))
return;
/*
* We fault-in kernel-space virtual memory on-demand. The
* 'reference' page table is init_mm.pgd.
*
* NOTE! We MUST NOT take any locks for this case. We may
* be in an interrupt or a critical region, and should
* only copy the information from the master page table,
* nothing more.
*
* This verifies that the fault happens in kernel space
* (error_code & 4) == 0, and that the fault was not a
* protection error (error_code & 9) == 0.
*/
#ifdef CONFIG_X86_32
if (unlikely(address >= TASK_SIZE)) {
#else
if (unlikely(address >= TASK_SIZE64)) {
#endif
if (!(error_code & (PF_RSVD|PF_USER|PF_PROT)) &&
vmalloc_fault(address) >= 0)
return;
/* Can handle a stale RO->RW TLB */
if (spurious_fault(address, error_code))
return;
/*
* Don't take the mm semaphore here. If we fixup a prefetch
* fault we could otherwise deadlock.
*/
goto bad_area_nosemaphore;
}
/*
* It's safe to allow irq's after cr2 has been saved and the
* vmalloc fault has been handled.
*
* User-mode registers count as a user access even for any
* potential system fault or CPU buglet.
*/
if (user_mode_vm(regs)) {
local_irq_enable();
error_code |= PF_USER;
} else if (regs->flags & X86_EFLAGS_IF)
local_irq_enable();
#ifdef CONFIG_X86_64
if (unlikely(error_code & PF_RSVD))
pgtable_bad(address, regs, error_code);
#endif
/*
* If we're in an interrupt, have no user context or are running in an
* atomic region then we must not take the fault.
*/
if (unlikely(in_atomic() || !mm))
goto bad_area_nosemaphore;
again:
/*
* When running in the kernel we expect faults to occur only to
* addresses in user space. All other faults represent errors in the
* kernel and should generate an OOPS. Unfortunately, in the case of an
* erroneous fault occurring in a code path which already holds mmap_sem
* we will deadlock attempting to validate the fault against the
* address space. Luckily the kernel only validly references user
* space from well defined areas of code, which are listed in the
* exceptions table.
*
* As the vast majority of faults will be valid we will only perform
* the source reference check when there is a possibility of a deadlock.
* Attempt to lock the address space, if we cannot we then validate the
* source. If this is invalid we can skip the address space check,
* thus avoiding the deadlock.
*/
if (!down_read_trylock(&mm->mmap_sem)) {
if ((error_code & PF_USER) == 0 &&
!search_exception_tables(regs->ip))
goto bad_area_nosemaphore;
down_read(&mm->mmap_sem);
}
vma = find_vma(mm, address);
if (!vma)
goto bad_area;
if (vma->vm_start <= address)
goto good_area;
if (!(vma->vm_flags & VM_GROWSDOWN))
goto bad_area;
if (error_code & PF_USER) {
/*
* Accessing the stack below %sp is always a bug.
* The large cushion allows instructions like enter
* and pusha to work. ("enter $65535,$31" pushes
* 32 pointers and then decrements %sp by 65535.)
*/
if (address + 65536 + 32 * sizeof(unsigned long) < regs->sp)
goto bad_area;
}
if (expand_stack(vma, address))
goto bad_area;
/*
* Ok, we have a good vm_area for this memory access, so
* we can handle it..
*/
good_area:
si_code = SEGV_ACCERR;
write = 0;
switch (error_code & (PF_PROT|PF_WRITE)) {
default: /* 3: write, present */
/* fall through */
case PF_WRITE: /* write, not present */
if (!(vma->vm_flags & VM_WRITE))
goto bad_area;
write++;
break;
case PF_PROT: /* read, present */
goto bad_area;
case 0: /* read, not present */
if (!(vma->vm_flags & (VM_READ | VM_EXEC | VM_WRITE)))
goto bad_area;
}
/*
* If for any reason at all we couldn't handle the fault,
* make sure we exit gracefully rather than endlessly redo
* the fault.
*/
mm: fault feedback #2 This patch completes Linus's wish that the fault return codes be made into bit flags, which I agree makes everything nicer. This requires requires all handle_mm_fault callers to be modified (possibly the modifications should go further and do things like fault accounting in handle_mm_fault -- however that would be for another patch). [akpm@linux-foundation.org: fix alpha build] [akpm@linux-foundation.org: fix s390 build] [akpm@linux-foundation.org: fix sparc build] [akpm@linux-foundation.org: fix sparc64 build] [akpm@linux-foundation.org: fix ia64 build] Signed-off-by: Nick Piggin <npiggin@suse.de> Cc: Richard Henderson <rth@twiddle.net> Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ian Molton <spyro@f2s.com> Cc: Bryan Wu <bryan.wu@analog.com> Cc: Mikael Starvik <starvik@axis.com> Cc: David Howells <dhowells@redhat.com> Cc: Yoshinori Sato <ysato@users.sourceforge.jp> Cc: "Luck, Tony" <tony.luck@intel.com> Cc: Hirokazu Takata <takata@linux-m32r.org> Cc: Geert Uytterhoeven <geert@linux-m68k.org> Cc: Roman Zippel <zippel@linux-m68k.org> Cc: Greg Ungerer <gerg@uclinux.org> Cc: Matthew Wilcox <willy@debian.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Paul Mundt <lethal@linux-sh.org> Cc: Kazumoto Kojima <kkojima@rr.iij4u.or.jp> Cc: Richard Curnow <rc@rc0.org.uk> Cc: William Lee Irwin III <wli@holomorphy.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Jeff Dike <jdike@addtoit.com> Cc: Paolo 'Blaisorblade' Giarrusso <blaisorblade@yahoo.it> Cc: Miles Bader <uclinux-v850@lsi.nec.co.jp> Cc: Chris Zankel <chris@zankel.net> Acked-by: Kyle McMartin <kyle@mcmartin.ca> Acked-by: Haavard Skinnemoen <hskinnemoen@atmel.com> Acked-by: Ralf Baechle <ralf@linux-mips.org> Acked-by: Andi Kleen <ak@muc.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> [ Still apparently needs some ARM and PPC loving - Linus ] Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-19 08:47:05 +00:00
fault = handle_mm_fault(mm, vma, address, write);
if (unlikely(fault & VM_FAULT_ERROR)) {
if (fault & VM_FAULT_OOM)
goto out_of_memory;
else if (fault & VM_FAULT_SIGBUS)
goto do_sigbus;
BUG();
}
mm: fault feedback #2 This patch completes Linus's wish that the fault return codes be made into bit flags, which I agree makes everything nicer. This requires requires all handle_mm_fault callers to be modified (possibly the modifications should go further and do things like fault accounting in handle_mm_fault -- however that would be for another patch). [akpm@linux-foundation.org: fix alpha build] [akpm@linux-foundation.org: fix s390 build] [akpm@linux-foundation.org: fix sparc build] [akpm@linux-foundation.org: fix sparc64 build] [akpm@linux-foundation.org: fix ia64 build] Signed-off-by: Nick Piggin <npiggin@suse.de> Cc: Richard Henderson <rth@twiddle.net> Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ian Molton <spyro@f2s.com> Cc: Bryan Wu <bryan.wu@analog.com> Cc: Mikael Starvik <starvik@axis.com> Cc: David Howells <dhowells@redhat.com> Cc: Yoshinori Sato <ysato@users.sourceforge.jp> Cc: "Luck, Tony" <tony.luck@intel.com> Cc: Hirokazu Takata <takata@linux-m32r.org> Cc: Geert Uytterhoeven <geert@linux-m68k.org> Cc: Roman Zippel <zippel@linux-m68k.org> Cc: Greg Ungerer <gerg@uclinux.org> Cc: Matthew Wilcox <willy@debian.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Paul Mundt <lethal@linux-sh.org> Cc: Kazumoto Kojima <kkojima@rr.iij4u.or.jp> Cc: Richard Curnow <rc@rc0.org.uk> Cc: William Lee Irwin III <wli@holomorphy.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Jeff Dike <jdike@addtoit.com> Cc: Paolo 'Blaisorblade' Giarrusso <blaisorblade@yahoo.it> Cc: Miles Bader <uclinux-v850@lsi.nec.co.jp> Cc: Chris Zankel <chris@zankel.net> Acked-by: Kyle McMartin <kyle@mcmartin.ca> Acked-by: Haavard Skinnemoen <hskinnemoen@atmel.com> Acked-by: Ralf Baechle <ralf@linux-mips.org> Acked-by: Andi Kleen <ak@muc.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> [ Still apparently needs some ARM and PPC loving - Linus ] Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-19 08:47:05 +00:00
if (fault & VM_FAULT_MAJOR)
tsk->maj_flt++;
else
tsk->min_flt++;
#ifdef CONFIG_X86_32
/*
* Did it hit the DOS screen memory VA from vm86 mode?
*/
if (v8086_mode(regs)) {
unsigned long bit = (address - 0xA0000) >> PAGE_SHIFT;
if (bit < 32)
tsk->thread.screen_bitmap |= 1 << bit;
}
#endif
up_read(&mm->mmap_sem);
return;
/*
* Something tried to access memory that isn't in our memory map..
* Fix it, but check if it's kernel or user first..
*/
bad_area:
up_read(&mm->mmap_sem);
bad_area_nosemaphore:
/* User mode accesses just cause a SIGSEGV */
if (error_code & PF_USER) {
/*
* It's possible to have interrupts off here.
*/
local_irq_enable();
/*
* Valid to do another page fault here because this one came
* from user space.
*/
if (is_prefetch(regs, address, error_code))
return;
if (is_errata100(regs, address))
return;
if (show_unhandled_signals && unhandled_signal(tsk, SIGSEGV) &&
printk_ratelimit()) {
printk(
"%s%s[%d]: segfault at %lx ip %p sp %p error %lx",
task_pid_nr(tsk) > 1 ? KERN_INFO : KERN_EMERG,
tsk->comm, task_pid_nr(tsk), address,
(void *) regs->ip, (void *) regs->sp, error_code);
print_vma_addr(" in ", regs->ip);
printk("\n");
}
tsk->thread.cr2 = address;
/* Kernel addresses are always protection faults */
tsk->thread.error_code = error_code | (address >= TASK_SIZE);
tsk->thread.trap_no = 14;
force_sig_info_fault(SIGSEGV, si_code, address, tsk);
return;
}
if (is_f00f_bug(regs, address))
return;
no_context:
/* Are we prepared to handle this kernel fault? */
if (fixup_exception(regs))
return;
/*
* X86_32
* Valid to do another page fault here, because if this fault
* had been triggered by is_prefetch fixup_exception would have
* handled it.
*
* X86_64
* Hall of shame of CPU/BIOS bugs.
*/
if (is_prefetch(regs, address, error_code))
return;
if (is_errata93(regs, address))
return;
/*
* Oops. The kernel tried to access some bad page. We'll have to
* terminate things with extreme prejudice.
*/
#ifdef CONFIG_X86_32
bust_spinlocks(1);
#else
flags = oops_begin();
#endif
show_fault_oops(regs, error_code, address);
tsk->thread.cr2 = address;
tsk->thread.trap_no = 14;
tsk->thread.error_code = error_code;
#ifdef CONFIG_X86_32
die("Oops", regs, error_code);
bust_spinlocks(0);
do_exit(SIGKILL);
#else
if (__die("Oops", regs, error_code))
regs = NULL;
/* Executive summary in case the body of the oops scrolled away */
printk(KERN_EMERG "CR2: %016lx\n", address);
oops_end(flags, regs, SIGKILL);
#endif
/*
* We ran out of memory, or some other thing happened to us that made
* us unable to handle the page fault gracefully.
*/
out_of_memory:
up_read(&mm->mmap_sem);
if (is_global_init(tsk)) {
yield();
/*
* Re-lookup the vma - in theory the vma tree might
* have changed:
*/
goto again;
}
printk("VM: killing process %s\n", tsk->comm);
if (error_code & PF_USER)
do_group_exit(SIGKILL);
goto no_context;
do_sigbus:
up_read(&mm->mmap_sem);
/* Kernel mode? Handle exceptions or die */
if (!(error_code & PF_USER))
goto no_context;
#ifdef CONFIG_X86_32
/* User space => ok to do another page fault */
if (is_prefetch(regs, address, error_code))
return;
#endif
tsk->thread.cr2 = address;
tsk->thread.error_code = error_code;
tsk->thread.trap_no = 14;
force_sig_info_fault(SIGBUS, BUS_ADRERR, address, tsk);
}
DEFINE_SPINLOCK(pgd_lock);
LIST_HEAD(pgd_list);
void vmalloc_sync_all(void)
{
unsigned long address;
#ifdef CONFIG_X86_32
if (SHARED_KERNEL_PMD)
return;
for (address = VMALLOC_START & PMD_MASK;
address >= TASK_SIZE && address < FIXADDR_TOP;
address += PMD_SIZE) {
unsigned long flags;
struct page *page;
spin_lock_irqsave(&pgd_lock, flags);
list_for_each_entry(page, &pgd_list, lru) {
if (!vmalloc_sync_one(page_address(page),
address))
break;
}
spin_unlock_irqrestore(&pgd_lock, flags);
}
#else /* CONFIG_X86_64 */
for (address = VMALLOC_START & PGDIR_MASK; address <= VMALLOC_END;
address += PGDIR_SIZE) {
const pgd_t *pgd_ref = pgd_offset_k(address);
unsigned long flags;
struct page *page;
if (pgd_none(*pgd_ref))
continue;
spin_lock_irqsave(&pgd_lock, flags);
list_for_each_entry(page, &pgd_list, lru) {
pgd_t *pgd;
pgd = (pgd_t *)page_address(page) + pgd_index(address);
if (pgd_none(*pgd))
set_pgd(pgd, *pgd_ref);
else
BUG_ON(pgd_page_vaddr(*pgd) != pgd_page_vaddr(*pgd_ref));
}
spin_unlock_irqrestore(&pgd_lock, flags);
}
#endif
}