[mirror_ubuntu-focal-kernel.git] / arch / sparc / mm / fault_64.c

// SPDX-License-Identifier: GPL-2.0
/*
 * 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/sched/debug.h>
#include <linux/ptrace.h>
#include <linux/mman.h>
#include <linux/signal.h>
#include <linux/mm.h>
#include <linux/extable.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;

#if defined(CONFIG_HUGETLB_PAGE) || defined(CONFIG_TRANSPARENT_HUGEPAGE)
	if (is_hugetlb_pmd(*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 %px (rpc %px) sp %px 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(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_TRANSPARENT_HUGEPAGE)
	mm_rss -= (mm->context.thp_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.hugetlb_pte_count + mm->context.thp_pte_count;
	mm_rss *= REAL_HPAGE_PER_HPAGE;
	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
b2441318	1	// SPDX-License-Identifier: GPL-2.0
b00dc837	2	/*
1da177e4 LT	3	* arch/sparc64/mm/fault.c: Page fault handlers for the 64-bit Sparc.
1da177e4 LT	4	*
4fe3ebec	5	* Copyright (C) 1996, 2008 David S. Miller (davem@davemloft.net)
1da177e4 LT	6	* Copyright (C) 1997, 1999 Jakub Jelinek (jj@ultra.linux.cz)
	7	*/
	8
	9	#include <asm/head.h>
	10
	11	#include <linux/string.h>
	12	#include <linux/types.h>
	13	#include <linux/sched.h>
b17b0153	14	#include <linux/sched/debug.h>
1da177e4 LT	15	#include <linux/ptrace.h>
	16	#include <linux/mman.h>
	17	#include <linux/signal.h>
	18	#include <linux/mm.h>
cdd4f4c7	19	#include <linux/extable.h>
1da177e4	20	#include <linux/init.h>
a084b667	21	#include <linux/perf_event.h>
1da177e4	22	#include <linux/interrupt.h>
05e14cb3	23	#include <linux/kprobes.h>
1eeb66a1	24	#include <linux/kdebug.h>
eeabac73	25	#include <linux/percpu.h>
812cb83a	26	#include <linux/context_tracking.h>
70ffdb93	27	#include <linux/uaccess.h>
1da177e4 LT	28
	29	#include <asm/page.h>
	30	#include <asm/pgtable.h>
	31	#include <asm/openprom.h>
	32	#include <asm/oplib.h>
1da177e4 LT	33	#include <asm/asi.h>
	34	#include <asm/lsu.h>
	35	#include <asm/sections.h>
7a1ac526	36	#include <asm/mmu_context.h>
8df52620	37	#include <asm/setup.h>
1da177e4	38
4b177647 DM	39	int show_unhandled_signals = 1;
4b177647 DM	40
4ed5d5e4	41	static inline __kprobes int notify_page_fault(struct pt_regs *regs)
d98f8f05	42	{
127cda1e DM	43	int ret = 0;
	44
	45	/* kprobe_running() needs smp_processor_id() */
135d0821	46	if (kprobes_built_in() && !user_mode(regs)) {
127cda1e DM	47	preempt_disable();
	48	if (kprobe_running() && kprobe_fault_handler(regs, 0))
	49	ret = 1;
	50	preempt_enable();
	51	}
	52	return ret;
d98f8f05	53	}
d98f8f05	54
05e14cb3 PP	55	static void __kprobes unhandled_fault(unsigned long address,
	56	struct task_struct *tsk,
	57	struct pt_regs *regs)
1da177e4 LT	58	{
	59	if ((unsigned long) address < PAGE_SIZE) {
	60	printk(KERN_ALERT "Unable to handle kernel NULL "
	61	"pointer dereference\n");
	62	} else {
	63	printk(KERN_ALERT "Unable to handle kernel paging request "
	64	"at virtual address %016lx\n", (unsigned long)address);
	65	}
	66	printk(KERN_ALERT "tsk->{mm,active_mm}->context = %016lx\n",
	67	(tsk->mm ?
	68	CTX_HWBITS(tsk->mm->context) :
	69	CTX_HWBITS(tsk->active_mm->context)));
	70	printk(KERN_ALERT "tsk->{mm,active_mm}->pgd = %016lx\n",
	71	(tsk->mm ? (unsigned long) tsk->mm->pgd :
	72	(unsigned long) tsk->active_mm->pgd));
1da177e4 LT	73	die_if_kernel("Oops", regs);
	74	}
	75
4ed5d5e4	76	static void __kprobes bad_kernel_pc(struct pt_regs *regs, unsigned long vaddr)
1da177e4	77	{
1da177e4 LT	78	printk(KERN_CRIT "OOPS: Bogus kernel PC [%016lx] in fault handler\n",
1da177e4 LT	79	regs->tpc);
eb398d10	80	printk(KERN_CRIT "OOPS: RPC [%016lx]\n", regs->u_regs[15]);
4fe3ebec	81	printk("OOPS: RPC <%pS>\n", (void *) regs->u_regs[15]);
bf941d6c	82	printk(KERN_CRIT "OOPS: Fault was to vaddr[%lx]\n", vaddr);
c1f193a7	83	dump_stack();
1da177e4 LT	84	unhandled_fault(regs->tpc, current, regs);
	85	}
	86
	87	/*
	88	* We now make sure that mmap_sem is held in all paths that call
	89	* this. Additionally, to prevent kswapd from ripping ptes from
	90	* under us, raise interrupts around the time that we look at the
	91	* pte, kswapd will have to wait to get his smp ipi response from
da160546	92	* us. vmtruncate likewise. This saves us having to get pte lock.
1da177e4 LT	93	*/
	94	static unsigned int get_user_insn(unsigned long tpc)
	95	{
	96	pgd_t *pgdp = pgd_offset(current->mm, tpc);
	97	pud_t *pudp;
	98	pmd_t *pmdp;
	99	pte_t *ptep, pte;
	100	unsigned long pa;
	101	u32 insn = 0;
1da177e4	102
70ffc6eb DM	103	if (pgd_none(pgdp) \|\| unlikely(pgd_bad(pgdp)))
70ffc6eb DM	104	goto out;
1da177e4	105	pudp = pud_offset(pgdp, tpc);
70ffc6eb DM	106	if (pud_none(pudp) \|\| unlikely(pud_bad(pudp)))
70ffc6eb DM	107	goto out;
1da177e4 LT	108
1da177e4 LT	109	/* This disables preemption for us as well. */
70ffc6eb	110	local_irq_disable();
1da177e4	111
70ffc6eb DM	112	pmdp = pmd_offset(pudp, tpc);
	113	if (pmd_none(pmdp) \|\| unlikely(pmd_bad(pmdp)))
	114	goto out_irq_enable;
	115
7bc3777c NG	116	#if defined(CONFIG_HUGETLB_PAGE) \|\| defined(CONFIG_TRANSPARENT_HUGEPAGE)
7bc3777c NG	117	if (is_hugetlb_pmd(*pmdp)) {
70ffc6eb DM	118	pa = pmd_pfn(*pmdp) << PAGE_SHIFT;
70ffc6eb DM	119	pa += tpc & ~HPAGE_MASK;
1da177e4	120
70ffc6eb DM	121	/* Use phys bypass so we don't pollute dtlb/dcache. */
	122	__asm__ __volatile__("lduwa [%1] %2, %0"
	123	: "=r" (insn)
	124	: "r" (pa), "i" (ASI_PHYS_USE_EC));
	125	} else
	126	#endif
	127	{
	128	ptep = pte_offset_map(pmdp, tpc);
	129	pte = *ptep;
	130	if (pte_present(pte)) {
	131	pa = (pte_pfn(pte) << PAGE_SHIFT);
	132	pa += (tpc & ~PAGE_MASK);
	133
	134	/* Use phys bypass so we don't pollute dtlb/dcache. */
	135	__asm__ __volatile__("lduwa [%1] %2, %0"
	136	: "=r" (insn)
	137	: "r" (pa), "i" (ASI_PHYS_USE_EC));
	138	}
	139	pte_unmap(ptep);
	140	}
	141	out_irq_enable:
	142	local_irq_enable();
1da177e4	143	out:
1da177e4 LT	144	return insn;
	145	}
	146
4b177647 DM	147	static inline void
	148	show_signal_msg(struct pt_regs *regs, int sig, int code,
	149	unsigned long address, struct task_struct *tsk)
	150	{
	151	if (!unhandled_signal(tsk, sig))
	152	return;
	153
	154	if (!printk_ratelimit())
	155	return;
	156
10a7e9d8	157	printk("%s%s[%d]: segfault at %lx ip %px (rpc %px) sp %px error %x",
4b177647 DM	158	task_pid_nr(tsk) > 1 ? KERN_INFO : KERN_EMERG,
	159	tsk->comm, task_pid_nr(tsk), address,
	160	(void )regs->tpc, (void )regs->u_regs[UREG_I7],
	161	(void *)regs->u_regs[UREG_FP], code);
	162
	163	print_vma_addr(KERN_CONT " in ", regs->tpc);
	164
	165	printk(KERN_CONT "\n");
	166	}
	167
1da177e4	168	static void do_fault_siginfo(int code, int sig, struct pt_regs *regs,
70ffc6eb DM	169	unsigned long fault_addr, unsigned int insn,
70ffc6eb DM	170	int fault_code)
1da177e4	171	{
4b177647	172	unsigned long addr;
1da177e4 LT	173	siginfo_t info;
	174
	175	info.si_code = code;
	176	info.si_signo = sig;
	177	info.si_errno = 0;
70ffc6eb	178	if (fault_code & FAULT_CODE_ITLB) {
4b177647	179	addr = regs->tpc;
70ffc6eb DM	180	} else {
	181	/* If we were able to probe the faulting instruction, use it
	182	* to compute a precise fault address. Otherwise use the fault
	183	* time provided address which may only have page granularity.
	184	*/
	185	if (insn)
	186	addr = compute_effective_address(regs, insn, 0);
	187	else
	188	addr = fault_addr;
	189	}
4b177647	190	info.si_addr = (void __user *) addr;
1da177e4	191	info.si_trapno = 0;
4b177647 DM	192
	193	if (unlikely(show_unhandled_signals))
	194	show_signal_msg(regs, sig, code, addr, current);
	195
1da177e4 LT	196	force_sig_info(sig, &info, current);
	197	}
	198
1da177e4 LT	199	static unsigned int get_fault_insn(struct pt_regs *regs, unsigned int insn)
	200	{
	201	if (!insn) {
	202	if (!regs->tpc \|\| (regs->tpc & 0x3))
	203	return 0;
	204	if (regs->tstate & TSTATE_PRIV) {
	205	insn = (unsigned int ) regs->tpc;
	206	} else {
	207	insn = get_user_insn(regs->tpc);
	208	}
	209	}
	210	return insn;
	211	}
	212
4ed5d5e4 DM	213	static void __kprobes do_kernel_fault(struct pt_regs *regs, int si_code,
	214	int fault_code, unsigned int insn,
	215	unsigned long address)
1da177e4	216	{
1da177e4 LT	217	unsigned char asi = ASI_P;
	218
	219	if ((!insn) && (regs->tstate & TSTATE_PRIV))
	220	goto cannot_handle;
	221
	222	/* If user insn could be read (thus insn is zero), that
	223	* is fine. We will just gun down the process with a signal
	224	* in that case.
	225	*/
	226
	227	if (!(fault_code & (FAULT_CODE_WRITE\|FAULT_CODE_ITLB)) &&
	228	(insn & 0xc0800000) == 0xc0800000) {
	229	if (insn & 0x2000)
	230	asi = (regs->tstate >> 24);
	231	else
	232	asi = (insn >> 5);
	233	if ((asi & 0xf2) == 0x82) {
	234	if (insn & 0x1000000) {
	235	handle_ldf_stq(insn, regs);
	236	} else {
	237	/* This was a non-faulting load. Just clear the
	238	* destination register(s) and continue with the next
	239	* instruction. -jj
	240	*/
	241	handle_ld_nf(insn, regs);
	242	}
	243	return;
	244	}
	245	}
	246
1da177e4 LT	247	/* Is this in ex_table? */
1da177e4 LT	248	if (regs->tstate & TSTATE_PRIV) {
8cf14af0	249	const struct exception_table_entry *entry;
1da177e4	250
622eaec6 DM	251	entry = search_exception_tables(regs->tpc);
622eaec6 DM	252	if (entry) {
8cf14af0	253	regs->tpc = entry->fixup;
1da177e4	254	regs->tnpc = regs->tpc + 4;
1da177e4 LT	255	return;
	256	}
	257	} else {
	258	/* The si_code was set to make clear whether
	259	* this was a SEGV_MAPERR or SEGV_ACCERR fault.
	260	*/
70ffc6eb	261	do_fault_siginfo(si_code, SIGSEGV, regs, address, insn, fault_code);
1da177e4 LT	262	return;
	263	}
	264
	265	cannot_handle:
	266	unhandled_fault (address, current, regs);
	267	}
	268
4ed5d5e4	269	static void noinline __kprobes bogus_32bit_fault_tpc(struct pt_regs *regs)
9b026058 DM	270	{
	271	static int times;
	272
	273	if (times++ < 10)
	274	printk(KERN_ERR "FAULT[%s:%d]: 32-bit process reports "
	275	"64-bit TPC [%lx]\n",
	276	current->comm, current->pid,
	277	regs->tpc);
	278	show_regs(regs);
	279	}
	280
05e14cb3	281	asmlinkage void __kprobes do_sparc64_fault(struct pt_regs *regs)
1da177e4	282	{
812cb83a	283	enum ctx_state prev_state = exception_enter();
1da177e4 LT	284	struct mm_struct *mm = current->mm;
	285	struct vm_area_struct *vma;
	286	unsigned int insn = 0;
83c54070	287	int si_code, fault_code, fault;
7a1ac526	288	unsigned long address, mm_rss;
7358e510	289	unsigned int flags = FAULT_FLAG_ALLOW_RETRY \| FAULT_FLAG_KILLABLE;
1da177e4 LT	290
	291	fault_code = get_thread_fault_code();
	292
127cda1e	293	if (notify_page_fault(regs))
812cb83a	294	goto exit_exception;
1da177e4 LT	295
	296	si_code = SEGV_MAPERR;
	297	address = current_thread_info()->fault_address;
	298
	299	if ((fault_code & FAULT_CODE_ITLB) &&
	300	(fault_code & FAULT_CODE_DTLB))
	301	BUG();
	302
eeabac73	303	if (test_thread_flag(TIF_32BIT)) {
9b026058 DM	304	if (!(regs->tstate & TSTATE_PRIV)) {
	305	if (unlikely((regs->tpc >> 32) != 0)) {
	306	bogus_32bit_fault_tpc(regs);
	307	goto intr_or_no_mm;
	308	}
	309	}
e5c460f4	310	if (unlikely((address >> 32) != 0))
9b026058	311	goto intr_or_no_mm;
eeabac73 DM	312	}
eeabac73 DM	313
1da177e4	314	if (regs->tstate & TSTATE_PRIV) {
9b026058	315	unsigned long tpc = regs->tpc;
1da177e4 LT	316
1da177e4 LT	317	/* Sanity check the PC. */
be71716e	318	if ((tpc >= KERNBASE && tpc < (unsigned long) __init_end) \|\|
1da177e4 LT	319	(tpc >= MODULES_VADDR && tpc < MODULES_END)) {
	320	/* Valid, no problems... */
	321	} else {
bf941d6c	322	bad_kernel_pc(regs, address);
812cb83a	323	goto exit_exception;
1da177e4	324	}
759496ba JW	325	} else
759496ba JW	326	flags \|= FAULT_FLAG_USER;
1da177e4 LT	327
	328	/*
	329	* If we're in an interrupt or have no user
	330	* context, we must not take the fault..
	331	*/
70ffdb93	332	if (faulthandler_disabled() \|\| !mm)
1da177e4 LT	333	goto intr_or_no_mm;
1da177e4 LT	334
a8b0ca17	335	perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS, 1, regs, address);
a084b667	336
1da177e4 LT	337	if (!down_read_trylock(&mm->mmap_sem)) {
	338	if ((regs->tstate & TSTATE_PRIV) &&
	339	!search_exception_tables(regs->tpc)) {
	340	insn = get_fault_insn(regs, insn);
	341	goto handle_kernel_fault;
	342	}
7358e510 KC	343
7358e510 KC	344	retry:
1da177e4 LT	345	down_read(&mm->mmap_sem);
	346	}
	347
4ccb9272	348	if (fault_code & FAULT_CODE_BAD_RA)
	349	goto do_sigbus;
	350
1da177e4 LT	351	vma = find_vma(mm, address);
	352	if (!vma)
	353	goto bad_area;
	354
	355	/* Pure DTLB misses do not tell us whether the fault causing
	356	* load/store/atomic was a write or not, it only says that there
	357	* was no match. So in such a case we (carefully) read the
	358	* instruction to try and figure this out. It's an optimization
	359	* so it's ok if we can't do this.
	360	*
	361	* Special hack, window spill/fill knows the exact fault type.
	362	*/
	363	if (((fault_code &
	364	(FAULT_CODE_DTLB \| FAULT_CODE_WRITE \| FAULT_CODE_WINFIXUP)) == FAULT_CODE_DTLB) &&
	365	(vma->vm_flags & VM_WRITE) != 0) {
	366	insn = get_fault_insn(regs, 0);
	367	if (!insn)
	368	goto continue_fault;
73c50a27 DM	369	/* All loads, stores and atomics have bits 30 and 31 both set
	370	* in the instruction. Bit 21 is set in all stores, but we
	371	* have to avoid prefetches which also have bit 21 set.
	372	*/
1da177e4	373	if ((insn & 0xc0200000) == 0xc0200000 &&
73c50a27	374	(insn & 0x01780000) != 0x01680000) {
1da177e4 LT	375	/* Don't bother updating thread struct value,
	376	* because update_mmu_cache only cares which tlb
	377	* the access came from.
	378	*/
	379	fault_code \|= FAULT_CODE_WRITE;
	380	}
	381	}
	382	continue_fault:
	383
	384	if (vma->vm_start <= address)
	385	goto good_area;
	386	if (!(vma->vm_flags & VM_GROWSDOWN))
	387	goto bad_area;
	388	if (!(fault_code & FAULT_CODE_WRITE)) {
	389	/* Non-faulting loads shouldn't expand stack. */
	390	insn = get_fault_insn(regs, insn);
	391	if ((insn & 0xc0800000) == 0xc0800000) {
	392	unsigned char asi;
	393
	394	if (insn & 0x2000)
	395	asi = (regs->tstate >> 24);
	396	else
	397	asi = (insn >> 5);
	398	if ((asi & 0xf2) == 0x82)
	399	goto bad_area;
	400	}
	401	}
	402	if (expand_stack(vma, address))
	403	goto bad_area;
	404	/*
	405	* Ok, we have a good vm_area for this memory access, so
	406	* we can handle it..
	407	*/
	408	good_area:
	409	si_code = SEGV_ACCERR;
	410
	411	/* If we took a ITLB miss on a non-executable page, catch
	412	* that here.
	413	*/
	414	if ((fault_code & FAULT_CODE_ITLB) && !(vma->vm_flags & VM_EXEC)) {
2bf7c3ef DA	415	WARN(address != regs->tpc,
	416	"address (%lx) != regs->tpc (%lx)\n", address, regs->tpc);
	417	WARN_ON(regs->tstate & TSTATE_PRIV);
1da177e4 LT	418	goto bad_area;
	419	}
	420
	421	if (fault_code & FAULT_CODE_WRITE) {
	422	if (!(vma->vm_flags & VM_WRITE))
	423	goto bad_area;
	424
	425	/* Spitfire has an icache which does not snoop
	426	* processor stores. Later processors do...
	427	*/
	428	if (tlb_type == spitfire &&
	429	(vma->vm_flags & VM_EXEC) != 0 &&
	430	vma->vm_file != NULL)
	431	set_thread_fault_code(fault_code \|
	432	FAULT_CODE_BLKCOMMIT);
759496ba JW	433
759496ba JW	434	flags \|= FAULT_FLAG_WRITE;
1da177e4 LT	435	} else {
	436	/* Allow reads even for write-only mappings */
	437	if (!(vma->vm_flags & (VM_READ \| VM_EXEC)))
	438	goto bad_area;
	439	}
	440
dcddffd4	441	fault = handle_mm_fault(vma, address, flags);
7358e510 KC	442
7358e510 KC	443	if ((fault & VM_FAULT_RETRY) && fatal_signal_pending(current))
812cb83a	444	goto exit_exception;
7358e510	445
83c54070 NP	446	if (unlikely(fault & VM_FAULT_ERROR)) {
	447	if (fault & VM_FAULT_OOM)
	448	goto out_of_memory;
33692f27 LT	449	else if (fault & VM_FAULT_SIGSEGV)
33692f27 LT	450	goto bad_area;
83c54070 NP	451	else if (fault & VM_FAULT_SIGBUS)
83c54070 NP	452	goto do_sigbus;
1da177e4 LT	453	BUG();
1da177e4 LT	454	}
7358e510 KC	455
	456	if (flags & FAULT_FLAG_ALLOW_RETRY) {
	457	if (fault & VM_FAULT_MAJOR) {
	458	current->maj_flt++;
	459	perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MAJ,
	460	1, regs, address);
	461	} else {
	462	current->min_flt++;
	463	perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MIN,
	464	1, regs, address);
	465	}
	466	if (fault & VM_FAULT_RETRY) {
	467	flags &= ~FAULT_FLAG_ALLOW_RETRY;
45cac65b	468	flags \|= FAULT_FLAG_TRIED;
7358e510 KC	469
	470	/* No need to up_read(&mm->mmap_sem) as we would
	471	* have already released it in __lock_page_or_retry
	472	* in mm/filemap.c.
	473	*/
	474
	475	goto retry;
	476	}
a084b667	477	}
1da177e4	478	up_read(&mm->mmap_sem);
7a1ac526 DM	479
7a1ac526 DM	480	mm_rss = get_mm_rss(mm);
af1b1a9b MK	481	#if defined(CONFIG_TRANSPARENT_HUGEPAGE)
af1b1a9b MK	482	mm_rss -= (mm->context.thp_pte_count * (HPAGE_SIZE / PAGE_SIZE));
dcc1e8dd	483	#endif
7bebd83d	484	if (unlikely(mm_rss >
dcc1e8dd DM	485	mm->context.tsb_block[MM_TSB_BASE].tsb_rss_limit))
dcc1e8dd DM	486	tsb_grow(mm, MM_TSB_BASE, mm_rss);
9e695d2e	487	#if defined(CONFIG_HUGETLB_PAGE) \|\| defined(CONFIG_TRANSPARENT_HUGEPAGE)
af1b1a9b	488	mm_rss = mm->context.hugetlb_pte_count + mm->context.thp_pte_count;
1e953d84	489	mm_rss *= REAL_HPAGE_PER_HPAGE;
7bebd83d	490	if (unlikely(mm_rss >
0fbebed6 DM	491	mm->context.tsb_block[MM_TSB_HUGE].tsb_rss_limit)) {
	492	if (mm->context.tsb_block[MM_TSB_HUGE].tsb)
	493	tsb_grow(mm, MM_TSB_HUGE, mm_rss);
	494	else
	495	hugetlb_setup(regs);
	496
	497	}
dcc1e8dd	498	#endif
812cb83a KT	499	exit_exception:
812cb83a KT	500	exception_exit(prev_state);
efdc1e20	501	return;
1da177e4 LT	502
	503	/*
	504	* Something tried to access memory that isn't in our memory map..
	505	* Fix it, but check if it's kernel or user first..
	506	*/
	507	bad_area:
	508	insn = get_fault_insn(regs, insn);
	509	up_read(&mm->mmap_sem);
	510
	511	handle_kernel_fault:
	512	do_kernel_fault(regs, si_code, fault_code, insn, address);
812cb83a	513	goto exit_exception;
1da177e4 LT	514
	515	/*
	516	* We ran out of memory, or some other thing happened to us that made
	517	* us unable to handle the page fault gracefully.
	518	*/
	519	out_of_memory:
	520	insn = get_fault_insn(regs, insn);
	521	up_read(&mm->mmap_sem);
a923c28f DM	522	if (!(regs->tstate & TSTATE_PRIV)) {
a923c28f DM	523	pagefault_out_of_memory();
812cb83a	524	goto exit_exception;
a923c28f	525	}
1da177e4 LT	526	goto handle_kernel_fault;
	527
	528	intr_or_no_mm:
	529	insn = get_fault_insn(regs, 0);
	530	goto handle_kernel_fault;
	531
	532	do_sigbus:
	533	insn = get_fault_insn(regs, insn);
	534	up_read(&mm->mmap_sem);
	535
	536	/*
	537	* Send a sigbus, regardless of whether we were in kernel
	538	* or user mode.
	539	*/
70ffc6eb	540	do_fault_siginfo(BUS_ADRERR, SIGBUS, regs, address, insn, fault_code);
1da177e4 LT	541
	542	/* Kernel mode? Handle exceptions or die */
	543	if (regs->tstate & TSTATE_PRIV)
	544	goto handle_kernel_fault;
1da177e4	545	}