[GitHub/mt8127/android_kernel_alcatel_ttab.git] / arch / sparc / mm / fault_64.c

/*
 * arch/sparc64/mm/fault.c: Page fault handlers for the 64-bit Sparc.
 *
 * Copyright (C) 1996, 2008 David S. Miller (davem@davemloft.net)
 * Copyright (C) 1997, 1999 Jakub Jelinek (jj@ultra.linux.cz)
 */

#include <asm/head.h>

#include <linux/string.h>
#include <linux/types.h>
#include <linux/sched.h>
#include <linux/ptrace.h>
#include <linux/mman.h>
#include <linux/signal.h>
#include <linux/mm.h>
#include <linux/module.h>
#include <linux/init.h>
#include <linux/perf_event.h>
#include <linux/interrupt.h>
#include <linux/kprobes.h>
#include <linux/kdebug.h>
#include <linux/percpu.h>

#include <asm/page.h>
#include <asm/pgtable.h>
#include <asm/openprom.h>
#include <asm/oplib.h>
#include <asm/uaccess.h>
#include <asm/asi.h>
#include <asm/lsu.h>
#include <asm/sections.h>
#include <asm/mmu_context.h>

int show_unhandled_signals = 1;

static inline __kprobes int notify_page_fault(struct pt_regs *regs)
{
	int ret = 0;

	/* kprobe_running() needs smp_processor_id() */
	if (kprobes_built_in() && !user_mode(regs)) {
		preempt_disable();
		if (kprobe_running() && kprobe_fault_handler(regs, 0))
			ret = 1;
		preempt_enable();
	}
	return ret;
}

static void __kprobes unhandled_fault(unsigned long address,
				      struct task_struct *tsk,
				      struct pt_regs *regs)
{
	if ((unsigned long) address < PAGE_SIZE) {
		printk(KERN_ALERT "Unable to handle kernel NULL "
		       "pointer dereference\n");
	} else {
		printk(KERN_ALERT "Unable to handle kernel paging request "
		       "at virtual address %016lx\n", (unsigned long)address);
	}
	printk(KERN_ALERT "tsk->{mm,active_mm}->context = %016lx\n",
	       (tsk->mm ?
		CTX_HWBITS(tsk->mm->context) :
		CTX_HWBITS(tsk->active_mm->context)));
	printk(KERN_ALERT "tsk->{mm,active_mm}->pgd = %016lx\n",
	       (tsk->mm ? (unsigned long) tsk->mm->pgd :
		          (unsigned long) tsk->active_mm->pgd));
	die_if_kernel("Oops", regs);
}

static void __kprobes bad_kernel_pc(struct pt_regs *regs, unsigned long vaddr)
{
	printk(KERN_CRIT "OOPS: Bogus kernel PC [%016lx] in fault handler\n",
	       regs->tpc);
	printk(KERN_CRIT "OOPS: RPC [%016lx]\n", regs->u_regs[15]);
	printk("OOPS: RPC <%pS>\n", (void *) regs->u_regs[15]);
	printk(KERN_CRIT "OOPS: Fault was to vaddr[%lx]\n", vaddr);
	dump_stack();
	unhandled_fault(regs->tpc, current, regs);
}

/*
 * We now make sure that mmap_sem is held in all paths that call 
 * this. Additionally, to prevent kswapd from ripping ptes from
 * under us, raise interrupts around the time that we look at the
 * pte, kswapd will have to wait to get his smp ipi response from
 * us. vmtruncate likewise. This saves us having to get pte lock.
 */
static unsigned int get_user_insn(unsigned long tpc)
{
	pgd_t *pgdp = pgd_offset(current->mm, tpc);
	pud_t *pudp;
	pmd_t *pmdp;
	pte_t *ptep, pte;
	unsigned long pa;
	u32 insn = 0;

	if (pgd_none(*pgdp) || unlikely(pgd_bad(*pgdp)))
		goto out;
	pudp = pud_offset(pgdp, tpc);
	if (pud_none(*pudp) || unlikely(pud_bad(*pudp)))
		goto out;

	/* This disables preemption for us as well. */
	local_irq_disable();

	pmdp = pmd_offset(pudp, tpc);
	if (pmd_none(*pmdp) || unlikely(pmd_bad(*pmdp)))
		goto out_irq_enable;

#ifdef CONFIG_TRANSPARENT_HUGEPAGE
	if (pmd_trans_huge(*pmdp)) {
		if (pmd_trans_splitting(*pmdp))
			goto out_irq_enable;

		pa  = pmd_pfn(*pmdp) << PAGE_SHIFT;
		pa += tpc & ~HPAGE_MASK;

		/* Use phys bypass so we don't pollute dtlb/dcache. */
		__asm__ __volatile__("lduwa [%1] %2, %0"
				     : "=r" (insn)
				     : "r" (pa), "i" (ASI_PHYS_USE_EC));
	} else
#endif
	{
		ptep = pte_offset_map(pmdp, tpc);
		pte = *ptep;
		if (pte_present(pte)) {
			pa  = (pte_pfn(pte) << PAGE_SHIFT);
			pa += (tpc & ~PAGE_MASK);

			/* Use phys bypass so we don't pollute dtlb/dcache. */
			__asm__ __volatile__("lduwa [%1] %2, %0"
					     : "=r" (insn)
					     : "r" (pa), "i" (ASI_PHYS_USE_EC));
		}
		pte_unmap(ptep);
	}
out_irq_enable:
	local_irq_enable();
out:
	return insn;
}

static inline void
show_signal_msg(struct pt_regs *regs, int sig, int code,
		unsigned long address, struct task_struct *tsk)
{
	if (!unhandled_signal(tsk, sig))
		return;

	if (!printk_ratelimit())
		return;

	printk("%s%s[%d]: segfault at %lx ip %p (rpc %p) sp %p error %x",
	       task_pid_nr(tsk) > 1 ? KERN_INFO : KERN_EMERG,
	       tsk->comm, task_pid_nr(tsk), address,
	       (void *)regs->tpc, (void *)regs->u_regs[UREG_I7],
	       (void *)regs->u_regs[UREG_FP], code);

	print_vma_addr(KERN_CONT " in ", regs->tpc);

	printk(KERN_CONT "\n");
}

static void do_fault_siginfo(int code, int sig, struct pt_regs *regs,
			     unsigned long fault_addr, unsigned int insn,
			     int fault_code)
{
	unsigned long addr;
	siginfo_t info;

	info.si_code = code;
	info.si_signo = sig;
	info.si_errno = 0;
	if (fault_code & FAULT_CODE_ITLB) {
		addr = regs->tpc;
	} else {
		/* If we were able to probe the faulting instruction, use it
		 * to compute a precise fault address.  Otherwise use the fault
		 * time provided address which may only have page granularity.
		 */
		if (insn)
			addr = compute_effective_address(regs, insn, 0);
		else
			addr = fault_addr;
	}
	info.si_addr = (void __user *) addr;
	info.si_trapno = 0;

	if (unlikely(show_unhandled_signals))
		show_signal_msg(regs, sig, code, addr, current);

	force_sig_info(sig, &info, current);
}

extern int handle_ldf_stq(u32, struct pt_regs *);
extern int handle_ld_nf(u32, struct pt_regs *);

static unsigned int get_fault_insn(struct pt_regs *regs, unsigned int insn)
{
	if (!insn) {
		if (!regs->tpc || (regs->tpc & 0x3))
			return 0;
		if (regs->tstate & TSTATE_PRIV) {
			insn = *(unsigned int *) regs->tpc;
		} else {
			insn = get_user_insn(regs->tpc);
		}
	}
	return insn;
}

static void __kprobes do_kernel_fault(struct pt_regs *regs, int si_code,
				      int fault_code, unsigned int insn,
				      unsigned long address)
{
	unsigned char asi = ASI_P;
 
	if ((!insn) && (regs->tstate & TSTATE_PRIV))
		goto cannot_handle;

	/* If user insn could be read (thus insn is zero), that
	 * is fine.  We will just gun down the process with a signal
	 * in that case.
	 */

	if (!(fault_code & (FAULT_CODE_WRITE|FAULT_CODE_ITLB)) &&
	    (insn & 0xc0800000) == 0xc0800000) {
		if (insn & 0x2000)
			asi = (regs->tstate >> 24);
		else
			asi = (insn >> 5);
		if ((asi & 0xf2) == 0x82) {
			if (insn & 0x1000000) {
				handle_ldf_stq(insn, regs);
			} else {
				/* This was a non-faulting load. Just clear the
				 * destination register(s) and continue with the next
				 * instruction. -jj
				 */
				handle_ld_nf(insn, regs);
			}
			return;
		}
	}
		
	/* Is this in ex_table? */
	if (regs->tstate & TSTATE_PRIV) {
		const struct exception_table_entry *entry;

		entry = search_exception_tables(regs->tpc);
		if (entry) {
			regs->tpc = entry->fixup;
			regs->tnpc = regs->tpc + 4;
			return;
		}
	} else {
		/* The si_code was set to make clear whether
		 * this was a SEGV_MAPERR or SEGV_ACCERR fault.
		 */
		do_fault_siginfo(si_code, SIGSEGV, regs, address, insn, fault_code);
		return;
	}

cannot_handle:
	unhandled_fault (address, current, regs);
}

static void noinline __kprobes bogus_32bit_fault_tpc(struct pt_regs *regs)
{
	static int times;

	if (times++ < 10)
		printk(KERN_ERR "FAULT[%s:%d]: 32-bit process reports "
		       "64-bit TPC [%lx]\n",
		       current->comm, current->pid,
		       regs->tpc);
	show_regs(regs);
}

asmlinkage void __kprobes do_sparc64_fault(struct pt_regs *regs)
{
	struct mm_struct *mm = current->mm;
	struct vm_area_struct *vma;
	unsigned int insn = 0;
	int si_code, fault_code, fault;
	unsigned long address, mm_rss;
	unsigned int flags = FAULT_FLAG_ALLOW_RETRY | FAULT_FLAG_KILLABLE;

	fault_code = get_thread_fault_code();

	if (notify_page_fault(regs))
		return;

	si_code = SEGV_MAPERR;
	address = current_thread_info()->fault_address;

	if ((fault_code & FAULT_CODE_ITLB) &&
	    (fault_code & FAULT_CODE_DTLB))
		BUG();

	if (test_thread_flag(TIF_32BIT)) {
		if (!(regs->tstate & TSTATE_PRIV)) {
			if (unlikely((regs->tpc >> 32) != 0)) {
				bogus_32bit_fault_tpc(regs);
				goto intr_or_no_mm;
			}
		}
		if (unlikely((address >> 32) != 0))
			goto intr_or_no_mm;
	}

	if (regs->tstate & TSTATE_PRIV) {
		unsigned long tpc = regs->tpc;

		/* Sanity check the PC. */
		if ((tpc >= KERNBASE && tpc < (unsigned long) __init_end) ||
		    (tpc >= MODULES_VADDR && tpc < MODULES_END)) {
			/* Valid, no problems... */
		} else {
			bad_kernel_pc(regs, address);
			return;
		}
	} else
		flags |= FAULT_FLAG_USER;

	/*
	 * If we're in an interrupt or have no user
	 * context, we must not take the fault..
	 */
	if (in_atomic() || !mm)
		goto intr_or_no_mm;

	perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS, 1, regs, address);

	if (!down_read_trylock(&mm->mmap_sem)) {
		if ((regs->tstate & TSTATE_PRIV) &&
		    !search_exception_tables(regs->tpc)) {
			insn = get_fault_insn(regs, insn);
			goto handle_kernel_fault;
		}

retry:
		down_read(&mm->mmap_sem);
	}

	vma = find_vma(mm, address);
	if (!vma)
		goto bad_area;

	/* Pure DTLB misses do not tell us whether the fault causing
	 * load/store/atomic was a write or not, it only says that there
	 * was no match.  So in such a case we (carefully) read the
	 * instruction to try and figure this out.  It's an optimization
	 * so it's ok if we can't do this.
	 *
	 * Special hack, window spill/fill knows the exact fault type.
	 */
	if (((fault_code &
	      (FAULT_CODE_DTLB | FAULT_CODE_WRITE | FAULT_CODE_WINFIXUP)) == FAULT_CODE_DTLB) &&
	    (vma->vm_flags & VM_WRITE) != 0) {
		insn = get_fault_insn(regs, 0);
		if (!insn)
			goto continue_fault;
		/* All loads, stores and atomics have bits 30 and 31 both set
		 * in the instruction.  Bit 21 is set in all stores, but we
		 * have to avoid prefetches which also have bit 21 set.
		 */
		if ((insn & 0xc0200000) == 0xc0200000 &&
		    (insn & 0x01780000) != 0x01680000) {
			/* Don't bother updating thread struct value,
			 * because update_mmu_cache only cares which tlb
			 * the access came from.
			 */
			fault_code |= FAULT_CODE_WRITE;
		}
	}
continue_fault:

	if (vma->vm_start <= address)
		goto good_area;
	if (!(vma->vm_flags & VM_GROWSDOWN))
		goto bad_area;
	if (!(fault_code & FAULT_CODE_WRITE)) {
		/* Non-faulting loads shouldn't expand stack. */
		insn = get_fault_insn(regs, insn);
		if ((insn & 0xc0800000) == 0xc0800000) {
			unsigned char asi;

			if (insn & 0x2000)
				asi = (regs->tstate >> 24);
			else
				asi = (insn >> 5);
			if ((asi & 0xf2) == 0x82)
				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;

	/* If we took a ITLB miss on a non-executable page, catch
	 * that here.
	 */
	if ((fault_code & FAULT_CODE_ITLB) && !(vma->vm_flags & VM_EXEC)) {
		BUG_ON(address != regs->tpc);
		BUG_ON(regs->tstate & TSTATE_PRIV);
		goto bad_area;
	}

	if (fault_code & FAULT_CODE_WRITE) {
		if (!(vma->vm_flags & VM_WRITE))
			goto bad_area;

		/* Spitfire has an icache which does not snoop
		 * processor stores.  Later processors do...
		 */
		if (tlb_type == spitfire &&
		    (vma->vm_flags & VM_EXEC) != 0 &&
		    vma->vm_file != NULL)
			set_thread_fault_code(fault_code |
					      FAULT_CODE_BLKCOMMIT);

		flags |= FAULT_FLAG_WRITE;
	} else {
		/* Allow reads even for write-only mappings */
		if (!(vma->vm_flags & (VM_READ | VM_EXEC)))
			goto bad_area;
	}

	fault = handle_mm_fault(mm, vma, address, flags);

	if ((fault & VM_FAULT_RETRY) && fatal_signal_pending(current))
		return;

	if (unlikely(fault & VM_FAULT_ERROR)) {
		if (fault & VM_FAULT_OOM)
			goto out_of_memory;
		else if (fault & VM_FAULT_SIGSEGV)
			goto bad_area;
		else if (fault & VM_FAULT_SIGBUS)
			goto do_sigbus;
		BUG();
	}

	if (flags & FAULT_FLAG_ALLOW_RETRY) {
		if (fault & VM_FAULT_MAJOR) {
			current->maj_flt++;
			perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MAJ,
				      1, regs, address);
		} else {
			current->min_flt++;
			perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MIN,
				      1, regs, address);
		}
		if (fault & VM_FAULT_RETRY) {
			flags &= ~FAULT_FLAG_ALLOW_RETRY;
			flags |= FAULT_FLAG_TRIED;

			/* No need to up_read(&mm->mmap_sem) as we would
			 * have already released it in __lock_page_or_retry
			 * in mm/filemap.c.
			 */

			goto retry;
		}
	}
	up_read(&mm->mmap_sem);

	mm_rss = get_mm_rss(mm);
#if defined(CONFIG_HUGETLB_PAGE) || defined(CONFIG_TRANSPARENT_HUGEPAGE)
	mm_rss -= (mm->context.huge_pte_count * (HPAGE_SIZE / PAGE_SIZE));
#endif
	if (unlikely(mm_rss >
		     mm->context.tsb_block[MM_TSB_BASE].tsb_rss_limit))
		tsb_grow(mm, MM_TSB_BASE, mm_rss);
#if defined(CONFIG_HUGETLB_PAGE) || defined(CONFIG_TRANSPARENT_HUGEPAGE)
	mm_rss = mm->context.huge_pte_count;
	if (unlikely(mm_rss >
		     mm->context.tsb_block[MM_TSB_HUGE].tsb_rss_limit)) {
		if (mm->context.tsb_block[MM_TSB_HUGE].tsb)
			tsb_grow(mm, MM_TSB_HUGE, mm_rss);
		else
			hugetlb_setup(regs);

	}
#endif
	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:
	insn = get_fault_insn(regs, insn);
	up_read(&mm->mmap_sem);

handle_kernel_fault:
	do_kernel_fault(regs, si_code, fault_code, insn, address);
	return;

/*
 * 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:
	insn = get_fault_insn(regs, insn);
	up_read(&mm->mmap_sem);
	if (!(regs->tstate & TSTATE_PRIV)) {
		pagefault_out_of_memory();
		return;
	}
	goto handle_kernel_fault;

intr_or_no_mm:
	insn = get_fault_insn(regs, 0);
	goto handle_kernel_fault;

do_sigbus:
	insn = get_fault_insn(regs, insn);
	up_read(&mm->mmap_sem);

	/*
	 * Send a sigbus, regardless of whether we were in kernel
	 * or user mode.
	 */
	do_fault_siginfo(BUS_ADRERR, SIGBUS, regs, address, insn, fault_code);

	/* Kernel mode? Handle exceptions or die */
	if (regs->tstate & TSTATE_PRIV)
		goto handle_kernel_fault;
}
Commit	Line	Data
b00dc837	1	/*
1da177e4 LT	2	* arch/sparc64/mm/fault.c: Page fault handlers for the 64-bit Sparc.
1da177e4 LT	3	*
4fe3ebec	4	* Copyright (C) 1996, 2008 David S. Miller (davem@davemloft.net)
1da177e4 LT	5	* Copyright (C) 1997, 1999 Jakub Jelinek (jj@ultra.linux.cz)
	6	*/
	7
	8	#include <asm/head.h>
	9
	10	#include <linux/string.h>
	11	#include <linux/types.h>
	12	#include <linux/sched.h>
	13	#include <linux/ptrace.h>
	14	#include <linux/mman.h>
	15	#include <linux/signal.h>
	16	#include <linux/mm.h>
	17	#include <linux/module.h>
1da177e4	18	#include <linux/init.h>
a084b667	19	#include <linux/perf_event.h>
1da177e4	20	#include <linux/interrupt.h>
05e14cb3	21	#include <linux/kprobes.h>
1eeb66a1	22	#include <linux/kdebug.h>
eeabac73	23	#include <linux/percpu.h>
1da177e4 LT	24
	25	#include <asm/page.h>
	26	#include <asm/pgtable.h>
	27	#include <asm/openprom.h>
	28	#include <asm/oplib.h>
	29	#include <asm/uaccess.h>
	30	#include <asm/asi.h>
	31	#include <asm/lsu.h>
	32	#include <asm/sections.h>
7a1ac526	33	#include <asm/mmu_context.h>
1da177e4	34
4b177647 DM	35	int show_unhandled_signals = 1;
4b177647 DM	36
4ed5d5e4	37	static inline __kprobes int notify_page_fault(struct pt_regs *regs)
d98f8f05	38	{
127cda1e DM	39	int ret = 0;
	40
	41	/* kprobe_running() needs smp_processor_id() */
135d0821	42	if (kprobes_built_in() && !user_mode(regs)) {
127cda1e DM	43	preempt_disable();
	44	if (kprobe_running() && kprobe_fault_handler(regs, 0))
	45	ret = 1;
	46	preempt_enable();
	47	}
	48	return ret;
d98f8f05	49	}
d98f8f05	50
05e14cb3 PP	51	static void __kprobes unhandled_fault(unsigned long address,
	52	struct task_struct *tsk,
	53	struct pt_regs *regs)
1da177e4 LT	54	{
	55	if ((unsigned long) address < PAGE_SIZE) {
	56	printk(KERN_ALERT "Unable to handle kernel NULL "
	57	"pointer dereference\n");
	58	} else {
	59	printk(KERN_ALERT "Unable to handle kernel paging request "
	60	"at virtual address %016lx\n", (unsigned long)address);
	61	}
	62	printk(KERN_ALERT "tsk->{mm,active_mm}->context = %016lx\n",
	63	(tsk->mm ?
	64	CTX_HWBITS(tsk->mm->context) :
	65	CTX_HWBITS(tsk->active_mm->context)));
	66	printk(KERN_ALERT "tsk->{mm,active_mm}->pgd = %016lx\n",
	67	(tsk->mm ? (unsigned long) tsk->mm->pgd :
	68	(unsigned long) tsk->active_mm->pgd));
1da177e4 LT	69	die_if_kernel("Oops", regs);
	70	}
	71
4ed5d5e4	72	static void __kprobes bad_kernel_pc(struct pt_regs *regs, unsigned long vaddr)
1da177e4	73	{
1da177e4 LT	74	printk(KERN_CRIT "OOPS: Bogus kernel PC [%016lx] in fault handler\n",
1da177e4 LT	75	regs->tpc);
eb398d10	76	printk(KERN_CRIT "OOPS: RPC [%016lx]\n", regs->u_regs[15]);
4fe3ebec	77	printk("OOPS: RPC <%pS>\n", (void *) regs->u_regs[15]);
bf941d6c	78	printk(KERN_CRIT "OOPS: Fault was to vaddr[%lx]\n", vaddr);
c1f193a7	79	dump_stack();
1da177e4 LT	80	unhandled_fault(regs->tpc, current, regs);
	81	}
	82
	83	/*
	84	* We now make sure that mmap_sem is held in all paths that call
	85	* this. Additionally, to prevent kswapd from ripping ptes from
	86	* under us, raise interrupts around the time that we look at the
	87	* pte, kswapd will have to wait to get his smp ipi response from
da160546	88	* us. vmtruncate likewise. This saves us having to get pte lock.
1da177e4 LT	89	*/
	90	static unsigned int get_user_insn(unsigned long tpc)
	91	{
	92	pgd_t *pgdp = pgd_offset(current->mm, tpc);
	93	pud_t *pudp;
	94	pmd_t *pmdp;
	95	pte_t *ptep, pte;
	96	unsigned long pa;
	97	u32 insn = 0;
1da177e4	98
f2701cd6 DM	99	if (pgd_none(pgdp) \|\| unlikely(pgd_bad(pgdp)))
f2701cd6 DM	100	goto out;
1da177e4	101	pudp = pud_offset(pgdp, tpc);
f2701cd6	102	if (pud_none(pudp) \|\| unlikely(pud_bad(pudp)))
1da177e4 LT	103	goto out;
1da177e4 LT	104
f2701cd6 DM	105	/* This disables preemption for us as well. */
f2701cd6 DM	106	local_irq_disable();
1da177e4	107
f2701cd6 DM	108	pmdp = pmd_offset(pudp, tpc);
	109	if (pmd_none(pmdp) \|\| unlikely(pmd_bad(pmdp)))
	110	goto out_irq_enable;
	111
	112	#ifdef CONFIG_TRANSPARENT_HUGEPAGE
	113	if (pmd_trans_huge(*pmdp)) {
	114	if (pmd_trans_splitting(*pmdp))
	115	goto out_irq_enable;
	116
	117	pa = pmd_pfn(*pmdp) << PAGE_SHIFT;
	118	pa += tpc & ~HPAGE_MASK;
	119
	120	/* Use phys bypass so we don't pollute dtlb/dcache. */
	121	__asm__ __volatile__("lduwa [%1] %2, %0"
	122	: "=r" (insn)
	123	: "r" (pa), "i" (ASI_PHYS_USE_EC));
	124	} else
	125	#endif
	126	{
	127	ptep = pte_offset_map(pmdp, tpc);
	128	pte = *ptep;
	129	if (pte_present(pte)) {
	130	pa = (pte_pfn(pte) << PAGE_SHIFT);
	131	pa += (tpc & ~PAGE_MASK);
	132
	133	/* Use phys bypass so we don't pollute dtlb/dcache. */
	134	__asm__ __volatile__("lduwa [%1] %2, %0"
	135	: "=r" (insn)
	136	: "r" (pa), "i" (ASI_PHYS_USE_EC));
	137	}
	138	pte_unmap(ptep);
	139	}
	140	out_irq_enable:
	141	local_irq_enable();
1da177e4	142	out:
1da177e4 LT	143	return insn;
	144	}
	145
4b177647 DM	146	static inline void
	147	show_signal_msg(struct pt_regs *regs, int sig, int code,
	148	unsigned long address, struct task_struct *tsk)
	149	{
	150	if (!unhandled_signal(tsk, sig))
	151	return;
	152
	153	if (!printk_ratelimit())
	154	return;
	155
	156	printk("%s%s[%d]: segfault at %lx ip %p (rpc %p) sp %p error %x",
	157	task_pid_nr(tsk) > 1 ? KERN_INFO : KERN_EMERG,
	158	tsk->comm, task_pid_nr(tsk), address,
	159	(void )regs->tpc, (void )regs->u_regs[UREG_I7],
	160	(void *)regs->u_regs[UREG_FP], code);
	161
	162	print_vma_addr(KERN_CONT " in ", regs->tpc);
	163
	164	printk(KERN_CONT "\n");
	165	}
	166
1da177e4	167	static void do_fault_siginfo(int code, int sig, struct pt_regs *regs,
f2701cd6 DM	168	unsigned long fault_addr, unsigned int insn,
f2701cd6 DM	169	int fault_code)
1da177e4	170	{
4b177647	171	unsigned long addr;
1da177e4 LT	172	siginfo_t info;
	173
	174	info.si_code = code;
	175	info.si_signo = sig;
	176	info.si_errno = 0;
f2701cd6	177	if (fault_code & FAULT_CODE_ITLB) {
4b177647	178	addr = regs->tpc;
f2701cd6 DM	179	} else {
	180	/* If we were able to probe the faulting instruction, use it
	181	* to compute a precise fault address. Otherwise use the fault
	182	* time provided address which may only have page granularity.
	183	*/
	184	if (insn)
	185	addr = compute_effective_address(regs, insn, 0);
	186	else
	187	addr = fault_addr;
	188	}
4b177647	189	info.si_addr = (void __user *) addr;
1da177e4	190	info.si_trapno = 0;
4b177647 DM	191
	192	if (unlikely(show_unhandled_signals))
	193	show_signal_msg(regs, sig, code, addr, current);
	194
1da177e4 LT	195	force_sig_info(sig, &info, current);
	196	}
	197
	198	extern int handle_ldf_stq(u32, struct pt_regs *);
	199	extern int handle_ld_nf(u32, struct pt_regs *);
	200
	201	static unsigned int get_fault_insn(struct pt_regs *regs, unsigned int insn)
	202	{
	203	if (!insn) {
	204	if (!regs->tpc \|\| (regs->tpc & 0x3))
	205	return 0;
	206	if (regs->tstate & TSTATE_PRIV) {
	207	insn = (unsigned int ) regs->tpc;
	208	} else {
	209	insn = get_user_insn(regs->tpc);
	210	}
	211	}
	212	return insn;
	213	}
	214
4ed5d5e4 DM	215	static void __kprobes do_kernel_fault(struct pt_regs *regs, int si_code,
	216	int fault_code, unsigned int insn,
	217	unsigned long address)
1da177e4	218	{
1da177e4 LT	219	unsigned char asi = ASI_P;
	220
	221	if ((!insn) && (regs->tstate & TSTATE_PRIV))
	222	goto cannot_handle;
	223
	224	/* If user insn could be read (thus insn is zero), that
	225	* is fine. We will just gun down the process with a signal
	226	* in that case.
	227	*/
	228
	229	if (!(fault_code & (FAULT_CODE_WRITE\|FAULT_CODE_ITLB)) &&
	230	(insn & 0xc0800000) == 0xc0800000) {
	231	if (insn & 0x2000)
	232	asi = (regs->tstate >> 24);
	233	else
	234	asi = (insn >> 5);
	235	if ((asi & 0xf2) == 0x82) {
	236	if (insn & 0x1000000) {
	237	handle_ldf_stq(insn, regs);
	238	} else {
	239	/* This was a non-faulting load. Just clear the
	240	* destination register(s) and continue with the next
	241	* instruction. -jj
	242	*/
	243	handle_ld_nf(insn, regs);
	244	}
	245	return;
	246	}
	247	}
	248
1da177e4 LT	249	/* Is this in ex_table? */
1da177e4 LT	250	if (regs->tstate & TSTATE_PRIV) {
8cf14af0	251	const struct exception_table_entry *entry;
1da177e4	252
622eaec6 DM	253	entry = search_exception_tables(regs->tpc);
622eaec6 DM	254	if (entry) {
8cf14af0	255	regs->tpc = entry->fixup;
1da177e4	256	regs->tnpc = regs->tpc + 4;
1da177e4 LT	257	return;
	258	}
	259	} else {
	260	/* The si_code was set to make clear whether
	261	* this was a SEGV_MAPERR or SEGV_ACCERR fault.
	262	*/
f2701cd6	263	do_fault_siginfo(si_code, SIGSEGV, regs, address, insn, fault_code);
1da177e4 LT	264	return;
	265	}
	266
	267	cannot_handle:
	268	unhandled_fault (address, current, regs);
	269	}
	270
4ed5d5e4	271	static void noinline __kprobes bogus_32bit_fault_tpc(struct pt_regs *regs)
9b026058 DM	272	{
	273	static int times;
	274
	275	if (times++ < 10)
	276	printk(KERN_ERR "FAULT[%s:%d]: 32-bit process reports "
	277	"64-bit TPC [%lx]\n",
	278	current->comm, current->pid,
	279	regs->tpc);
	280	show_regs(regs);
	281	}
	282
05e14cb3	283	asmlinkage void __kprobes do_sparc64_fault(struct pt_regs *regs)
1da177e4 LT	284	{
	285	struct mm_struct *mm = current->mm;
	286	struct vm_area_struct *vma;
	287	unsigned int insn = 0;
83c54070	288	int si_code, fault_code, fault;
7a1ac526	289	unsigned long address, mm_rss;
7358e510	290	unsigned int flags = FAULT_FLAG_ALLOW_RETRY \| FAULT_FLAG_KILLABLE;
1da177e4 LT	291
	292	fault_code = get_thread_fault_code();
	293
127cda1e	294	if (notify_page_fault(regs))
1da177e4 LT	295	return;
	296
	297	si_code = SEGV_MAPERR;
	298	address = current_thread_info()->fault_address;
	299
	300	if ((fault_code & FAULT_CODE_ITLB) &&
	301	(fault_code & FAULT_CODE_DTLB))
	302	BUG();
	303
eeabac73	304	if (test_thread_flag(TIF_32BIT)) {
9b026058 DM	305	if (!(regs->tstate & TSTATE_PRIV)) {
	306	if (unlikely((regs->tpc >> 32) != 0)) {
	307	bogus_32bit_fault_tpc(regs);
	308	goto intr_or_no_mm;
	309	}
	310	}
e5bd62b9	311	if (unlikely((address >> 32) != 0))
9b026058	312	goto intr_or_no_mm;
eeabac73 DM	313	}
eeabac73 DM	314
1da177e4	315	if (regs->tstate & TSTATE_PRIV) {
9b026058	316	unsigned long tpc = regs->tpc;
1da177e4 LT	317
1da177e4 LT	318	/* Sanity check the PC. */
be71716e	319	if ((tpc >= KERNBASE && tpc < (unsigned long) __init_end) \|\|
1da177e4 LT	320	(tpc >= MODULES_VADDR && tpc < MODULES_END)) {
	321	/* Valid, no problems... */
	322	} else {
bf941d6c	323	bad_kernel_pc(regs, address);
1da177e4 LT	324	return;
1da177e4 LT	325	}
e2ec2c2b JW	326	} else
e2ec2c2b JW	327	flags \|= FAULT_FLAG_USER;
1da177e4 LT	328
	329	/*
	330	* If we're in an interrupt or have no user
	331	* context, we must not take the fault..
	332	*/
	333	if (in_atomic() \|\| !mm)
	334	goto intr_or_no_mm;
	335
a8b0ca17	336	perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS, 1, regs, address);
a084b667	337
1da177e4 LT	338	if (!down_read_trylock(&mm->mmap_sem)) {
	339	if ((regs->tstate & TSTATE_PRIV) &&
	340	!search_exception_tables(regs->tpc)) {
	341	insn = get_fault_insn(regs, insn);
	342	goto handle_kernel_fault;
	343	}
7358e510 KC	344
7358e510 KC	345	retry:
1da177e4 LT	346	down_read(&mm->mmap_sem);
	347	}
	348
	349	vma = find_vma(mm, address);
	350	if (!vma)
	351	goto bad_area;
	352
	353	/* Pure DTLB misses do not tell us whether the fault causing
	354	* load/store/atomic was a write or not, it only says that there
	355	* was no match. So in such a case we (carefully) read the
	356	* instruction to try and figure this out. It's an optimization
	357	* so it's ok if we can't do this.
	358	*
	359	* Special hack, window spill/fill knows the exact fault type.
	360	*/
	361	if (((fault_code &
	362	(FAULT_CODE_DTLB \| FAULT_CODE_WRITE \| FAULT_CODE_WINFIXUP)) == FAULT_CODE_DTLB) &&
	363	(vma->vm_flags & VM_WRITE) != 0) {
	364	insn = get_fault_insn(regs, 0);
	365	if (!insn)
	366	goto continue_fault;
73c50a27 DM	367	/* All loads, stores and atomics have bits 30 and 31 both set
	368	* in the instruction. Bit 21 is set in all stores, but we
	369	* have to avoid prefetches which also have bit 21 set.
	370	*/
1da177e4	371	if ((insn & 0xc0200000) == 0xc0200000 &&
73c50a27	372	(insn & 0x01780000) != 0x01680000) {
1da177e4 LT	373	/* Don't bother updating thread struct value,
	374	* because update_mmu_cache only cares which tlb
	375	* the access came from.
	376	*/
	377	fault_code \|= FAULT_CODE_WRITE;
	378	}
	379	}
	380	continue_fault:
	381
	382	if (vma->vm_start <= address)
	383	goto good_area;
	384	if (!(vma->vm_flags & VM_GROWSDOWN))
	385	goto bad_area;
	386	if (!(fault_code & FAULT_CODE_WRITE)) {
	387	/* Non-faulting loads shouldn't expand stack. */
	388	insn = get_fault_insn(regs, insn);
	389	if ((insn & 0xc0800000) == 0xc0800000) {
	390	unsigned char asi;
	391
	392	if (insn & 0x2000)
	393	asi = (regs->tstate >> 24);
	394	else
	395	asi = (insn >> 5);
	396	if ((asi & 0xf2) == 0x82)
	397	goto bad_area;
	398	}
	399	}
	400	if (expand_stack(vma, address))
	401	goto bad_area;
	402	/*
	403	* Ok, we have a good vm_area for this memory access, so
	404	* we can handle it..
	405	*/
	406	good_area:
	407	si_code = SEGV_ACCERR;
	408
	409	/* If we took a ITLB miss on a non-executable page, catch
	410	* that here.
	411	*/
	412	if ((fault_code & FAULT_CODE_ITLB) && !(vma->vm_flags & VM_EXEC)) {
	413	BUG_ON(address != regs->tpc);
	414	BUG_ON(regs->tstate & TSTATE_PRIV);
	415	goto bad_area;
	416	}
	417
	418	if (fault_code & FAULT_CODE_WRITE) {
	419	if (!(vma->vm_flags & VM_WRITE))
	420	goto bad_area;
	421
	422	/* Spitfire has an icache which does not snoop
	423	* processor stores. Later processors do...
	424	*/
	425	if (tlb_type == spitfire &&
	426	(vma->vm_flags & VM_EXEC) != 0 &&
	427	vma->vm_file != NULL)
	428	set_thread_fault_code(fault_code \|
	429	FAULT_CODE_BLKCOMMIT);
e2ec2c2b JW	430
e2ec2c2b JW	431	flags \|= FAULT_FLAG_WRITE;
1da177e4 LT	432	} else {
	433	/* Allow reads even for write-only mappings */
	434	if (!(vma->vm_flags & (VM_READ \| VM_EXEC)))
	435	goto bad_area;
	436	}
	437
7358e510 KC	438	fault = handle_mm_fault(mm, vma, address, flags);
	439
	440	if ((fault & VM_FAULT_RETRY) && fatal_signal_pending(current))
	441	return;
	442
83c54070 NP	443	if (unlikely(fault & VM_FAULT_ERROR)) {
	444	if (fault & VM_FAULT_OOM)
	445	goto out_of_memory;
0c42d1fb LT	446	else if (fault & VM_FAULT_SIGSEGV)
0c42d1fb LT	447	goto bad_area;
83c54070 NP	448	else if (fault & VM_FAULT_SIGBUS)
83c54070 NP	449	goto do_sigbus;
1da177e4 LT	450	BUG();
1da177e4 LT	451	}
7358e510 KC	452
	453	if (flags & FAULT_FLAG_ALLOW_RETRY) {
	454	if (fault & VM_FAULT_MAJOR) {
	455	current->maj_flt++;
	456	perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MAJ,
	457	1, regs, address);
	458	} else {
	459	current->min_flt++;
	460	perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MIN,
	461	1, regs, address);
	462	}
	463	if (fault & VM_FAULT_RETRY) {
	464	flags &= ~FAULT_FLAG_ALLOW_RETRY;
45cac65b	465	flags \|= FAULT_FLAG_TRIED;
7358e510 KC	466
	467	/* No need to up_read(&mm->mmap_sem) as we would
	468	* have already released it in __lock_page_or_retry
	469	* in mm/filemap.c.
	470	*/
	471
	472	goto retry;
	473	}
a084b667	474	}
1da177e4	475	up_read(&mm->mmap_sem);
7a1ac526 DM	476
7a1ac526 DM	477	mm_rss = get_mm_rss(mm);
9e695d2e	478	#if defined(CONFIG_HUGETLB_PAGE) \|\| defined(CONFIG_TRANSPARENT_HUGEPAGE)
dcc1e8dd DM	479	mm_rss -= (mm->context.huge_pte_count * (HPAGE_SIZE / PAGE_SIZE));
dcc1e8dd DM	480	#endif
7bebd83d	481	if (unlikely(mm_rss >
dcc1e8dd DM	482	mm->context.tsb_block[MM_TSB_BASE].tsb_rss_limit))
dcc1e8dd DM	483	tsb_grow(mm, MM_TSB_BASE, mm_rss);
9e695d2e	484	#if defined(CONFIG_HUGETLB_PAGE) \|\| defined(CONFIG_TRANSPARENT_HUGEPAGE)
dcc1e8dd	485	mm_rss = mm->context.huge_pte_count;
7bebd83d	486	if (unlikely(mm_rss >
0fbebed6 DM	487	mm->context.tsb_block[MM_TSB_HUGE].tsb_rss_limit)) {
	488	if (mm->context.tsb_block[MM_TSB_HUGE].tsb)
	489	tsb_grow(mm, MM_TSB_HUGE, mm_rss);
	490	else
	491	hugetlb_setup(regs);
	492
	493	}
dcc1e8dd	494	#endif
efdc1e20	495	return;
1da177e4 LT	496
	497	/*
	498	* Something tried to access memory that isn't in our memory map..
	499	* Fix it, but check if it's kernel or user first..
	500	*/
	501	bad_area:
	502	insn = get_fault_insn(regs, insn);
	503	up_read(&mm->mmap_sem);
	504
	505	handle_kernel_fault:
	506	do_kernel_fault(regs, si_code, fault_code, insn, address);
efdc1e20	507	return;
1da177e4 LT	508
	509	/*
	510	* We ran out of memory, or some other thing happened to us that made
	511	* us unable to handle the page fault gracefully.
	512	*/
	513	out_of_memory:
	514	insn = get_fault_insn(regs, insn);
	515	up_read(&mm->mmap_sem);
a923c28f DM	516	if (!(regs->tstate & TSTATE_PRIV)) {
	517	pagefault_out_of_memory();
	518	return;
	519	}
1da177e4 LT	520	goto handle_kernel_fault;
	521
	522	intr_or_no_mm:
	523	insn = get_fault_insn(regs, 0);
	524	goto handle_kernel_fault;
	525
	526	do_sigbus:
	527	insn = get_fault_insn(regs, insn);
	528	up_read(&mm->mmap_sem);
	529
	530	/*
	531	* Send a sigbus, regardless of whether we were in kernel
	532	* or user mode.
	533	*/
f2701cd6	534	do_fault_siginfo(BUS_ADRERR, SIGBUS, regs, address, insn, fault_code);
1da177e4 LT	535
	536	/* Kernel mode? Handle exceptions or die */
	537	if (regs->tstate & TSTATE_PRIV)
	538	goto handle_kernel_fault;
1da177e4	539	}