[deliverable/linux.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 <linux/context_tracking.h>
#include <linux/uaccess.h>

#include <asm/page.h>
#include <asm/pgtable.h>
#include <asm/openprom.h>
#include <asm/oplib.h>
#include <asm/asi.h>
#include <asm/lsu.h>
#include <asm/sections.h>
#include <asm/mmu_context.h>
#include <asm/setup.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)) {
		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);
}

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)
{
	enum ctx_state prev_state = exception_enter();
	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))
		goto exit_exception;

	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);
			goto exit_exception;
		}
	} else
		flags |= FAULT_FLAG_USER;

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

	if (fault_code & FAULT_CODE_BAD_RA)
		goto do_sigbus;

	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)) {
		WARN(address != regs->tpc,
		     "address (%lx) != regs->tpc (%lx)\n", address, regs->tpc);
		WARN_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))
		goto exit_exception;

	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
exit_exception:
	exception_exit(prev_state);
	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);
	goto exit_exception;

/*
 * 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();
		goto exit_exception;
	}
	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>
812cb83a	24	#include <linux/context_tracking.h>
70ffdb93	25	#include <linux/uaccess.h>
1da177e4 LT	26
	27	#include <asm/page.h>
	28	#include <asm/pgtable.h>
	29	#include <asm/openprom.h>
	30	#include <asm/oplib.h>
1da177e4 LT	31	#include <asm/asi.h>
	32	#include <asm/lsu.h>
	33	#include <asm/sections.h>
7a1ac526	34	#include <asm/mmu_context.h>
8df52620	35	#include <asm/setup.h>
1da177e4	36
4b177647 DM	37	int show_unhandled_signals = 1;
4b177647 DM	38
4ed5d5e4	39	static inline __kprobes int notify_page_fault(struct pt_regs *regs)
d98f8f05	40	{
127cda1e DM	41	int ret = 0;
	42
	43	/* kprobe_running() needs smp_processor_id() */
135d0821	44	if (kprobes_built_in() && !user_mode(regs)) {
127cda1e DM	45	preempt_disable();
	46	if (kprobe_running() && kprobe_fault_handler(regs, 0))
	47	ret = 1;
	48	preempt_enable();
	49	}
	50	return ret;
d98f8f05	51	}
d98f8f05	52
05e14cb3 PP	53	static void __kprobes unhandled_fault(unsigned long address,
	54	struct task_struct *tsk,
	55	struct pt_regs *regs)
1da177e4 LT	56	{
	57	if ((unsigned long) address < PAGE_SIZE) {
	58	printk(KERN_ALERT "Unable to handle kernel NULL "
	59	"pointer dereference\n");
	60	} else {
	61	printk(KERN_ALERT "Unable to handle kernel paging request "
	62	"at virtual address %016lx\n", (unsigned long)address);
	63	}
	64	printk(KERN_ALERT "tsk->{mm,active_mm}->context = %016lx\n",
	65	(tsk->mm ?
	66	CTX_HWBITS(tsk->mm->context) :
	67	CTX_HWBITS(tsk->active_mm->context)));
	68	printk(KERN_ALERT "tsk->{mm,active_mm}->pgd = %016lx\n",
	69	(tsk->mm ? (unsigned long) tsk->mm->pgd :
	70	(unsigned long) tsk->active_mm->pgd));
1da177e4 LT	71	die_if_kernel("Oops", regs);
	72	}
	73
4ed5d5e4	74	static void __kprobes bad_kernel_pc(struct pt_regs *regs, unsigned long vaddr)
1da177e4	75	{
1da177e4 LT	76	printk(KERN_CRIT "OOPS: Bogus kernel PC [%016lx] in fault handler\n",
1da177e4 LT	77	regs->tpc);
eb398d10	78	printk(KERN_CRIT "OOPS: RPC [%016lx]\n", regs->u_regs[15]);
4fe3ebec	79	printk("OOPS: RPC <%pS>\n", (void *) regs->u_regs[15]);
bf941d6c	80	printk(KERN_CRIT "OOPS: Fault was to vaddr[%lx]\n", vaddr);
c1f193a7	81	dump_stack();
1da177e4 LT	82	unhandled_fault(regs->tpc, current, regs);
	83	}
	84
	85	/*
	86	* We now make sure that mmap_sem is held in all paths that call
	87	* this. Additionally, to prevent kswapd from ripping ptes from
	88	* under us, raise interrupts around the time that we look at the
	89	* pte, kswapd will have to wait to get his smp ipi response from
da160546	90	* us. vmtruncate likewise. This saves us having to get pte lock.
1da177e4 LT	91	*/
	92	static unsigned int get_user_insn(unsigned long tpc)
	93	{
	94	pgd_t *pgdp = pgd_offset(current->mm, tpc);
	95	pud_t *pudp;
	96	pmd_t *pmdp;
	97	pte_t *ptep, pte;
	98	unsigned long pa;
	99	u32 insn = 0;
1da177e4	100
70ffc6eb DM	101	if (pgd_none(pgdp) \|\| unlikely(pgd_bad(pgdp)))
70ffc6eb DM	102	goto out;
1da177e4	103	pudp = pud_offset(pgdp, tpc);
70ffc6eb DM	104	if (pud_none(pudp) \|\| unlikely(pud_bad(pudp)))
70ffc6eb DM	105	goto out;
1da177e4 LT	106
1da177e4 LT	107	/* This disables preemption for us as well. */
70ffc6eb	108	local_irq_disable();
1da177e4	109
70ffc6eb DM	110	pmdp = pmd_offset(pudp, tpc);
	111	if (pmd_none(pmdp) \|\| unlikely(pmd_bad(pmdp)))
	112	goto out_irq_enable;
	113
	114	#ifdef CONFIG_TRANSPARENT_HUGEPAGE
	115	if (pmd_trans_huge(*pmdp)) {
70ffc6eb DM	116	pa = pmd_pfn(*pmdp) << PAGE_SHIFT;
70ffc6eb DM	117	pa += tpc & ~HPAGE_MASK;
1da177e4	118
70ffc6eb DM	119	/* Use phys bypass so we don't pollute dtlb/dcache. */
	120	__asm__ __volatile__("lduwa [%1] %2, %0"
	121	: "=r" (insn)
	122	: "r" (pa), "i" (ASI_PHYS_USE_EC));
	123	} else
	124	#endif
	125	{
	126	ptep = pte_offset_map(pmdp, tpc);
	127	pte = *ptep;
	128	if (pte_present(pte)) {
	129	pa = (pte_pfn(pte) << PAGE_SHIFT);
	130	pa += (tpc & ~PAGE_MASK);
	131
	132	/* Use phys bypass so we don't pollute dtlb/dcache. */
	133	__asm__ __volatile__("lduwa [%1] %2, %0"
	134	: "=r" (insn)
	135	: "r" (pa), "i" (ASI_PHYS_USE_EC));
	136	}
	137	pte_unmap(ptep);
	138	}
	139	out_irq_enable:
	140	local_irq_enable();
1da177e4	141	out:
1da177e4 LT	142	return insn;
	143	}
	144
4b177647 DM	145	static inline void
	146	show_signal_msg(struct pt_regs *regs, int sig, int code,
	147	unsigned long address, struct task_struct *tsk)
	148	{
	149	if (!unhandled_signal(tsk, sig))
	150	return;
	151
	152	if (!printk_ratelimit())
	153	return;
	154
	155	printk("%s%s[%d]: segfault at %lx ip %p (rpc %p) sp %p error %x",
	156	task_pid_nr(tsk) > 1 ? KERN_INFO : KERN_EMERG,
	157	tsk->comm, task_pid_nr(tsk), address,
	158	(void )regs->tpc, (void )regs->u_regs[UREG_I7],
	159	(void *)regs->u_regs[UREG_FP], code);
	160
	161	print_vma_addr(KERN_CONT " in ", regs->tpc);
	162
	163	printk(KERN_CONT "\n");
	164	}
	165
1da177e4	166	static void do_fault_siginfo(int code, int sig, struct pt_regs *regs,
70ffc6eb DM	167	unsigned long fault_addr, unsigned int insn,
70ffc6eb DM	168	int fault_code)
1da177e4	169	{
4b177647	170	unsigned long addr;
1da177e4 LT	171	siginfo_t info;
	172
	173	info.si_code = code;
	174	info.si_signo = sig;
	175	info.si_errno = 0;
70ffc6eb	176	if (fault_code & FAULT_CODE_ITLB) {
4b177647	177	addr = regs->tpc;
70ffc6eb DM	178	} else {
	179	/* If we were able to probe the faulting instruction, use it
	180	* to compute a precise fault address. Otherwise use the fault
	181	* time provided address which may only have page granularity.
	182	*/
	183	if (insn)
	184	addr = compute_effective_address(regs, insn, 0);
	185	else
	186	addr = fault_addr;
	187	}
4b177647	188	info.si_addr = (void __user *) addr;
1da177e4	189	info.si_trapno = 0;
4b177647 DM	190
	191	if (unlikely(show_unhandled_signals))
	192	show_signal_msg(regs, sig, code, addr, current);
	193
1da177e4 LT	194	force_sig_info(sig, &info, current);
	195	}
	196
1da177e4 LT	197	static unsigned int get_fault_insn(struct pt_regs *regs, unsigned int insn)
	198	{
	199	if (!insn) {
	200	if (!regs->tpc \|\| (regs->tpc & 0x3))
	201	return 0;
	202	if (regs->tstate & TSTATE_PRIV) {
	203	insn = (unsigned int ) regs->tpc;
	204	} else {
	205	insn = get_user_insn(regs->tpc);
	206	}
	207	}
	208	return insn;
	209	}
	210
4ed5d5e4 DM	211	static void __kprobes do_kernel_fault(struct pt_regs *regs, int si_code,
	212	int fault_code, unsigned int insn,
	213	unsigned long address)
1da177e4	214	{
1da177e4 LT	215	unsigned char asi = ASI_P;
	216
	217	if ((!insn) && (regs->tstate & TSTATE_PRIV))
	218	goto cannot_handle;
	219
	220	/* If user insn could be read (thus insn is zero), that
	221	* is fine. We will just gun down the process with a signal
	222	* in that case.
	223	*/
	224
	225	if (!(fault_code & (FAULT_CODE_WRITE\|FAULT_CODE_ITLB)) &&
	226	(insn & 0xc0800000) == 0xc0800000) {
	227	if (insn & 0x2000)
	228	asi = (regs->tstate >> 24);
	229	else
	230	asi = (insn >> 5);
	231	if ((asi & 0xf2) == 0x82) {
	232	if (insn & 0x1000000) {
	233	handle_ldf_stq(insn, regs);
	234	} else {
	235	/* This was a non-faulting load. Just clear the
	236	* destination register(s) and continue with the next
	237	* instruction. -jj
	238	*/
	239	handle_ld_nf(insn, regs);
	240	}
	241	return;
	242	}
	243	}
	244
1da177e4 LT	245	/* Is this in ex_table? */
1da177e4 LT	246	if (regs->tstate & TSTATE_PRIV) {
8cf14af0	247	const struct exception_table_entry *entry;
1da177e4	248
622eaec6 DM	249	entry = search_exception_tables(regs->tpc);
622eaec6 DM	250	if (entry) {
8cf14af0	251	regs->tpc = entry->fixup;
1da177e4	252	regs->tnpc = regs->tpc + 4;
1da177e4 LT	253	return;
	254	}
	255	} else {
	256	/* The si_code was set to make clear whether
	257	* this was a SEGV_MAPERR or SEGV_ACCERR fault.
	258	*/
70ffc6eb	259	do_fault_siginfo(si_code, SIGSEGV, regs, address, insn, fault_code);
1da177e4 LT	260	return;
	261	}
	262
	263	cannot_handle:
	264	unhandled_fault (address, current, regs);
	265	}
	266
4ed5d5e4	267	static void noinline __kprobes bogus_32bit_fault_tpc(struct pt_regs *regs)
9b026058 DM	268	{
	269	static int times;
	270
	271	if (times++ < 10)
	272	printk(KERN_ERR "FAULT[%s:%d]: 32-bit process reports "
	273	"64-bit TPC [%lx]\n",
	274	current->comm, current->pid,
	275	regs->tpc);
	276	show_regs(regs);
	277	}
	278
05e14cb3	279	asmlinkage void __kprobes do_sparc64_fault(struct pt_regs *regs)
1da177e4	280	{
812cb83a	281	enum ctx_state prev_state = exception_enter();
1da177e4 LT	282	struct mm_struct *mm = current->mm;
	283	struct vm_area_struct *vma;
	284	unsigned int insn = 0;
83c54070	285	int si_code, fault_code, fault;
7a1ac526	286	unsigned long address, mm_rss;
7358e510	287	unsigned int flags = FAULT_FLAG_ALLOW_RETRY \| FAULT_FLAG_KILLABLE;
1da177e4 LT	288
	289	fault_code = get_thread_fault_code();
	290
127cda1e	291	if (notify_page_fault(regs))
812cb83a	292	goto exit_exception;
1da177e4 LT	293
	294	si_code = SEGV_MAPERR;
	295	address = current_thread_info()->fault_address;
	296
	297	if ((fault_code & FAULT_CODE_ITLB) &&
	298	(fault_code & FAULT_CODE_DTLB))
	299	BUG();
	300
eeabac73	301	if (test_thread_flag(TIF_32BIT)) {
9b026058 DM	302	if (!(regs->tstate & TSTATE_PRIV)) {
	303	if (unlikely((regs->tpc >> 32) != 0)) {
	304	bogus_32bit_fault_tpc(regs);
	305	goto intr_or_no_mm;
	306	}
	307	}
e5c460f4	308	if (unlikely((address >> 32) != 0))
9b026058	309	goto intr_or_no_mm;
eeabac73 DM	310	}
eeabac73 DM	311
1da177e4	312	if (regs->tstate & TSTATE_PRIV) {
9b026058	313	unsigned long tpc = regs->tpc;
1da177e4 LT	314
1da177e4 LT	315	/* Sanity check the PC. */
be71716e	316	if ((tpc >= KERNBASE && tpc < (unsigned long) __init_end) \|\|
1da177e4 LT	317	(tpc >= MODULES_VADDR && tpc < MODULES_END)) {
	318	/* Valid, no problems... */
	319	} else {
bf941d6c	320	bad_kernel_pc(regs, address);
812cb83a	321	goto exit_exception;
1da177e4	322	}
759496ba JW	323	} else
759496ba JW	324	flags \|= FAULT_FLAG_USER;
1da177e4 LT	325
	326	/*
	327	* If we're in an interrupt or have no user
	328	* context, we must not take the fault..
	329	*/
70ffdb93	330	if (faulthandler_disabled() \|\| !mm)
1da177e4 LT	331	goto intr_or_no_mm;
1da177e4 LT	332
a8b0ca17	333	perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS, 1, regs, address);
a084b667	334
1da177e4 LT	335	if (!down_read_trylock(&mm->mmap_sem)) {
	336	if ((regs->tstate & TSTATE_PRIV) &&
	337	!search_exception_tables(regs->tpc)) {
	338	insn = get_fault_insn(regs, insn);
	339	goto handle_kernel_fault;
	340	}
7358e510 KC	341
7358e510 KC	342	retry:
1da177e4 LT	343	down_read(&mm->mmap_sem);
	344	}
	345
4ccb9272	346	if (fault_code & FAULT_CODE_BAD_RA)
	347	goto do_sigbus;
	348
1da177e4 LT	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)) {
2bf7c3ef DA	413	WARN(address != regs->tpc,
	414	"address (%lx) != regs->tpc (%lx)\n", address, regs->tpc);
	415	WARN_ON(regs->tstate & TSTATE_PRIV);
1da177e4 LT	416	goto bad_area;
	417	}
	418
	419	if (fault_code & FAULT_CODE_WRITE) {
	420	if (!(vma->vm_flags & VM_WRITE))
	421	goto bad_area;
	422
	423	/* Spitfire has an icache which does not snoop
	424	* processor stores. Later processors do...
	425	*/
	426	if (tlb_type == spitfire &&
	427	(vma->vm_flags & VM_EXEC) != 0 &&
	428	vma->vm_file != NULL)
	429	set_thread_fault_code(fault_code \|
	430	FAULT_CODE_BLKCOMMIT);
759496ba JW	431
759496ba JW	432	flags \|= FAULT_FLAG_WRITE;
1da177e4 LT	433	} else {
	434	/* Allow reads even for write-only mappings */
	435	if (!(vma->vm_flags & (VM_READ \| VM_EXEC)))
	436	goto bad_area;
	437	}
	438
7358e510 KC	439	fault = handle_mm_fault(mm, vma, address, flags);
	440
	441	if ((fault & VM_FAULT_RETRY) && fatal_signal_pending(current))
812cb83a	442	goto exit_exception;
7358e510	443
83c54070 NP	444	if (unlikely(fault & VM_FAULT_ERROR)) {
	445	if (fault & VM_FAULT_OOM)
	446	goto out_of_memory;
33692f27 LT	447	else if (fault & VM_FAULT_SIGSEGV)
33692f27 LT	448	goto bad_area;
83c54070 NP	449	else if (fault & VM_FAULT_SIGBUS)
83c54070 NP	450	goto do_sigbus;
1da177e4 LT	451	BUG();
1da177e4 LT	452	}
7358e510 KC	453
	454	if (flags & FAULT_FLAG_ALLOW_RETRY) {
	455	if (fault & VM_FAULT_MAJOR) {
	456	current->maj_flt++;
	457	perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MAJ,
	458	1, regs, address);
	459	} else {
	460	current->min_flt++;
	461	perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MIN,
	462	1, regs, address);
	463	}
	464	if (fault & VM_FAULT_RETRY) {
	465	flags &= ~FAULT_FLAG_ALLOW_RETRY;
45cac65b	466	flags \|= FAULT_FLAG_TRIED;
7358e510 KC	467
	468	/* No need to up_read(&mm->mmap_sem) as we would
	469	* have already released it in __lock_page_or_retry
	470	* in mm/filemap.c.
	471	*/
	472
	473	goto retry;
	474	}
a084b667	475	}
1da177e4	476	up_read(&mm->mmap_sem);
7a1ac526 DM	477
7a1ac526 DM	478	mm_rss = get_mm_rss(mm);
9e695d2e	479	#if defined(CONFIG_HUGETLB_PAGE) \|\| defined(CONFIG_TRANSPARENT_HUGEPAGE)
dcc1e8dd DM	480	mm_rss -= (mm->context.huge_pte_count * (HPAGE_SIZE / PAGE_SIZE));
dcc1e8dd DM	481	#endif
7bebd83d	482	if (unlikely(mm_rss >
dcc1e8dd DM	483	mm->context.tsb_block[MM_TSB_BASE].tsb_rss_limit))
dcc1e8dd DM	484	tsb_grow(mm, MM_TSB_BASE, mm_rss);
9e695d2e	485	#if defined(CONFIG_HUGETLB_PAGE) \|\| defined(CONFIG_TRANSPARENT_HUGEPAGE)
dcc1e8dd	486	mm_rss = mm->context.huge_pte_count;
7bebd83d	487	if (unlikely(mm_rss >
0fbebed6 DM	488	mm->context.tsb_block[MM_TSB_HUGE].tsb_rss_limit)) {
	489	if (mm->context.tsb_block[MM_TSB_HUGE].tsb)
	490	tsb_grow(mm, MM_TSB_HUGE, mm_rss);
	491	else
	492	hugetlb_setup(regs);
	493
	494	}
dcc1e8dd	495	#endif
812cb83a KT	496	exit_exception:
812cb83a KT	497	exception_exit(prev_state);
efdc1e20	498	return;
1da177e4 LT	499
	500	/*
	501	* Something tried to access memory that isn't in our memory map..
	502	* Fix it, but check if it's kernel or user first..
	503	*/
	504	bad_area:
	505	insn = get_fault_insn(regs, insn);
	506	up_read(&mm->mmap_sem);
	507
	508	handle_kernel_fault:
	509	do_kernel_fault(regs, si_code, fault_code, insn, address);
812cb83a	510	goto exit_exception;
1da177e4 LT	511
	512	/*
	513	* We ran out of memory, or some other thing happened to us that made
	514	* us unable to handle the page fault gracefully.
	515	*/
	516	out_of_memory:
	517	insn = get_fault_insn(regs, insn);
	518	up_read(&mm->mmap_sem);
a923c28f DM	519	if (!(regs->tstate & TSTATE_PRIV)) {
a923c28f DM	520	pagefault_out_of_memory();
812cb83a	521	goto exit_exception;
a923c28f	522	}
1da177e4 LT	523	goto handle_kernel_fault;
	524
	525	intr_or_no_mm:
	526	insn = get_fault_insn(regs, 0);
	527	goto handle_kernel_fault;
	528
	529	do_sigbus:
	530	insn = get_fault_insn(regs, insn);
	531	up_read(&mm->mmap_sem);
	532
	533	/*
	534	* Send a sigbus, regardless of whether we were in kernel
	535	* or user mode.
	536	*/
70ffc6eb	537	do_fault_siginfo(BUS_ADRERR, SIGBUS, regs, address, insn, fault_code);
1da177e4 LT	538
	539	/* Kernel mode? Handle exceptions or die */
	540	if (regs->tstate & TSTATE_PRIV)
	541	goto handle_kernel_fault;
1da177e4	542	}