1 files changed, 247 insertions, 127 deletions

diff --git a/arch/s390/mm/fault.c b/arch/s390/mm/fault.c
index 354dd39073e..3f3b35403d0 100644
--- a/arch/s390/mm/fault.c
+++ b/arch/s390/mm/fault.c
@@ -1,8 +1,6 @@
 /*
- *  arch/s390/mm/fault.c
- *
  *  S390 version
- *    Copyright (C) 1999 IBM Deutschland Entwicklung GmbH, IBM Corporation
+ *    Copyright IBM Corp. 1999
  *    Author(s): Hartmut Penner (hp@de.ibm.com)
  *               Ulrich Weigand (uweigand@de.ibm.com)
  *
@@ -32,11 +30,10 @@
 #include <linux/uaccess.h>
 #include <linux/hugetlb.h>
 #include <asm/asm-offsets.h>
-#include <asm/system.h>
 #include <asm/pgtable.h>
 #include <asm/irq.h>
 #include <asm/mmu_context.h>
-#include <asm/compat.h>
+#include <asm/facility.h>
 #include "../kernel/entry.h"
 
 #ifndef CONFIG_64BIT
@@ -52,14 +49,20 @@
 #define VM_FAULT_BADCONTEXT	0x010000
 #define VM_FAULT_BADMAP		0x020000
 #define VM_FAULT_BADACCESS	0x040000
+#define VM_FAULT_SIGNAL		0x080000
+#define VM_FAULT_PFAULT		0x100000
 
-static unsigned long store_indication;
+static unsigned long store_indication __read_mostly;
 
-void fault_init(void)
+#ifdef CONFIG_64BIT
+static int __init fault_init(void)
 {
-	if (test_facility(2) && test_facility(75))
+	if (test_facility(75))
 		store_indication = 0xc00;
+	return 0;
 }
+early_initcall(fault_init);
+#endif
 
 static inline int notify_page_fault(struct pt_regs *regs)
 {
@@ -103,26 +106,151 @@ void bust_spinlocks(int yes)
  * Returns the address space associated with the fault.
  * Returns 0 for kernel space and 1 for user space.
  */
-static inline int user_space_fault(unsigned long trans_exc_code)
+static inline int user_space_fault(struct pt_regs *regs)
 {
+	unsigned long trans_exc_code;
+
 	/*
 	 * The lowest two bits of the translation exception
 	 * identification indicate which paging table was used.
 	 */
-	trans_exc_code &= 3;
-	if (trans_exc_code == 2)
-		/* Access via secondary space, set_fs setting decides */
+	trans_exc_code = regs->int_parm_long & 3;
+	if (trans_exc_code == 3) /* home space -> kernel */
+		return 0;
+	if (user_mode(regs))
+		return 1;
+	if (trans_exc_code == 2) /* secondary space -> set_fs */
 		return current->thread.mm_segment.ar4;
-	if (user_mode == HOME_SPACE_MODE)
-		/* User space if the access has been done via home space. */
-		return trans_exc_code == 3;
-	/*
-	 * If the user space is not the home space the kernel runs in home
-	 * space. Access via secondary space has already been covered,
-	 * access via primary space or access register is from user space
-	 * and access via home space is from the kernel.
-	 */
-	return trans_exc_code != 3;
+	if (current->flags & PF_VCPU)
+		return 1;
+	return 0;
+}
+
+static int bad_address(void *p)
+{
+	unsigned long dummy;
+
+	return probe_kernel_address((unsigned long *)p, dummy);
+}
+
+#ifdef CONFIG_64BIT
+static void dump_pagetable(unsigned long asce, unsigned long address)
+{
+	unsigned long *table = __va(asce & PAGE_MASK);
+
+	pr_alert("AS:%016lx ", asce);
+	switch (asce & _ASCE_TYPE_MASK) {
+	case _ASCE_TYPE_REGION1:
+		table = table + ((address >> 53) & 0x7ff);
+		if (bad_address(table))
+			goto bad;
+		pr_cont("R1:%016lx ", *table);
+		if (*table & _REGION_ENTRY_INVALID)
+			goto out;
+		table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN);
+		/* fallthrough */
+	case _ASCE_TYPE_REGION2:
+		table = table + ((address >> 42) & 0x7ff);
+		if (bad_address(table))
+			goto bad;
+		pr_cont("R2:%016lx ", *table);
+		if (*table & _REGION_ENTRY_INVALID)
+			goto out;
+		table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN);
+		/* fallthrough */
+	case _ASCE_TYPE_REGION3:
+		table = table + ((address >> 31) & 0x7ff);
+		if (bad_address(table))
+			goto bad;
+		pr_cont("R3:%016lx ", *table);
+		if (*table & (_REGION_ENTRY_INVALID | _REGION3_ENTRY_LARGE))
+			goto out;
+		table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN);
+		/* fallthrough */
+	case _ASCE_TYPE_SEGMENT:
+		table = table + ((address >> 20) & 0x7ff);
+		if (bad_address(table))
+			goto bad;
+		pr_cont(KERN_CONT "S:%016lx ", *table);
+		if (*table & (_SEGMENT_ENTRY_INVALID | _SEGMENT_ENTRY_LARGE))
+			goto out;
+		table = (unsigned long *)(*table & _SEGMENT_ENTRY_ORIGIN);
+	}
+	table = table + ((address >> 12) & 0xff);
+	if (bad_address(table))
+		goto bad;
+	pr_cont("P:%016lx ", *table);
+out:
+	pr_cont("\n");
+	return;
+bad:
+	pr_cont("BAD\n");
+}
+
+#else /* CONFIG_64BIT */
+
+static void dump_pagetable(unsigned long asce, unsigned long address)
+{
+	unsigned long *table = __va(asce & PAGE_MASK);
+
+	pr_alert("AS:%08lx ", asce);
+	table = table + ((address >> 20) & 0x7ff);
+	if (bad_address(table))
+		goto bad;
+	pr_cont("S:%08lx ", *table);
+	if (*table & _SEGMENT_ENTRY_INVALID)
+		goto out;
+	table = (unsigned long *)(*table & _SEGMENT_ENTRY_ORIGIN);
+	table = table + ((address >> 12) & 0xff);
+	if (bad_address(table))
+		goto bad;
+	pr_cont("P:%08lx ", *table);
+out:
+	pr_cont("\n");
+	return;
+bad:
+	pr_cont("BAD\n");
+}
+
+#endif /* CONFIG_64BIT */
+
+static void dump_fault_info(struct pt_regs *regs)
+{
+	unsigned long asce;
+
+	pr_alert("Fault in ");
+	switch (regs->int_parm_long & 3) {
+	case 3:
+		pr_cont("home space ");
+		break;
+	case 2:
+		pr_cont("secondary space ");
+		break;
+	case 1:
+		pr_cont("access register ");
+		break;
+	case 0:
+		pr_cont("primary space ");
+		break;
+	}
+	pr_cont("mode while using ");
+	if (!user_space_fault(regs)) {
+		asce = S390_lowcore.kernel_asce;
+		pr_cont("kernel ");
+	}
+#ifdef CONFIG_PGSTE
+	else if ((current->flags & PF_VCPU) && S390_lowcore.gmap) {
+		struct gmap *gmap = (struct gmap *)S390_lowcore.gmap;
+		asce = gmap->asce;
+		pr_cont("gmap ");
+	}
+#endif
+	else {
+		asce = S390_lowcore.user_asce;
+		pr_cont("user ");
+	}
+	pr_cont("ASCE.\n");
+	dump_pagetable(asce, regs->int_parm_long & __FAIL_ADDR_MASK);
 }
 
 static inline void report_user_fault(struct pt_regs *regs, long signr)
@@ -137,8 +265,9 @@ static inline void report_user_fault(struct pt_regs *regs, long signr)
 	       regs->int_code);
 	print_vma_addr(KERN_CONT "in ", regs->psw.addr & PSW_ADDR_INSN);
 	printk(KERN_CONT "\n");
-	printk(KERN_ALERT "failing address: %lX\n",
-	       regs->int_parm_long & __FAIL_ADDR_MASK);
+	printk(KERN_ALERT "failing address: %016lx TEID: %016lx\n",
+	       regs->int_parm_long & __FAIL_ADDR_MASK, regs->int_parm_long);
+	dump_fault_info(regs);
 	show_regs(regs);
 }
 
@@ -165,7 +294,7 @@ static noinline void do_no_context(struct pt_regs *regs)
 	/* Are we prepared to handle this kernel fault?  */
 	fixup = search_exception_tables(regs->psw.addr & PSW_ADDR_INSN);
 	if (fixup) {
-		regs->psw.addr = fixup->fixup | PSW_ADDR_AMODE;
+		regs->psw.addr = extable_fixup(fixup) | PSW_ADDR_AMODE;
 		return;
 	}
 
@@ -174,13 +303,15 @@ static noinline void do_no_context(struct pt_regs *regs)
 	 * terminate things with extreme prejudice.
 	 */
 	address = regs->int_parm_long & __FAIL_ADDR_MASK;
-	if (!user_space_fault(regs->int_parm_long))
+	if (!user_space_fault(regs))
 		printk(KERN_ALERT "Unable to handle kernel pointer dereference"
-		       " at virtual kernel address %p\n", (void *)address);
+		       " in virtual kernel address space\n");
 	else
 		printk(KERN_ALERT "Unable to handle kernel paging request"
-		       " at virtual user address %p\n", (void *)address);
-
+		       " in virtual user address space\n");
+	printk(KERN_ALERT "failing address: %016lx TEID: %016lx\n",
+	       regs->int_parm_long & __FAIL_ADDR_MASK, regs->int_parm_long);
+	dump_fault_info(regs);
 	die(regs, "Oops");
 	do_exit(SIGKILL);
 }
@@ -222,7 +353,7 @@ static noinline void do_fault_error(struct pt_regs *regs, int fault)
 	case VM_FAULT_BADACCESS:
 	case VM_FAULT_BADMAP:
 		/* Bad memory access. Check if it is kernel or user space. */
-		if (regs->psw.mask & PSW_MASK_PSTATE) {
+		if (user_mode(regs)) {
 			/* User mode accesses just cause a SIGSEGV */
 			si_code = (fault == VM_FAULT_BADMAP) ?
 				SEGV_MAPERR : SEGV_ACCERR;
@@ -230,17 +361,22 @@ static noinline void do_fault_error(struct pt_regs *regs, int fault)
 			return;
 		}
 	case VM_FAULT_BADCONTEXT:
+	case VM_FAULT_PFAULT:
 		do_no_context(regs);
 		break;
+	case VM_FAULT_SIGNAL:
+		if (!user_mode(regs))
+			do_no_context(regs);
+		break;
 	default: /* fault & VM_FAULT_ERROR */
 		if (fault & VM_FAULT_OOM) {
-			if (!(regs->psw.mask & PSW_MASK_PSTATE))
+			if (!user_mode(regs))
 				do_no_context(regs);
 			else
 				pagefault_out_of_memory();
 		} else if (fault & VM_FAULT_SIGBUS) {
 			/* Kernel mode? Handle exceptions or die */
-			if (!(regs->psw.mask & PSW_MASK_PSTATE))
+			if (!user_mode(regs))
 				do_no_context(regs);
 			else
 				do_sigbus(regs);
@@ -263,6 +399,9 @@ static noinline void do_fault_error(struct pt_regs *regs, int fault)
  */
 static inline int do_exception(struct pt_regs *regs, int access)
 {
+#ifdef CONFIG_PGSTE
+	struct gmap *gmap;
+#endif
 	struct task_struct *tsk;
 	struct mm_struct *mm;
 	struct vm_area_struct *vma;
@@ -271,10 +410,16 @@ static inline int do_exception(struct pt_regs *regs, int access)
 	unsigned int flags;
 	int fault;
 
+	tsk = current;
+	/*
+	 * The instruction that caused the program check has
+	 * been nullified. Don't signal single step via SIGTRAP.
+	 */
+	clear_pt_regs_flag(regs, PIF_PER_TRAP);
+
 	if (notify_page_fault(regs))
 		return 0;
 
-	tsk = current;
 	mm = tsk->mm;
 	trans_exc_code = regs->int_parm_long;
 
@@ -284,20 +429,23 @@ static inline int do_exception(struct pt_regs *regs, int access)
 	 * user context.
 	 */
 	fault = VM_FAULT_BADCONTEXT;
-	if (unlikely(!user_space_fault(trans_exc_code) || in_atomic() || !mm))
+	if (unlikely(!user_space_fault(regs) || in_atomic() || !mm))
 		goto out;
 
 	address = trans_exc_code & __FAIL_ADDR_MASK;
 	perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS, 1, regs, address);
-	flags = FAULT_FLAG_ALLOW_RETRY;
+	flags = FAULT_FLAG_ALLOW_RETRY | FAULT_FLAG_KILLABLE;
+	if (user_mode(regs))
+		flags |= FAULT_FLAG_USER;
 	if (access == VM_WRITE || (trans_exc_code & store_indication) == 0x400)
 		flags |= FAULT_FLAG_WRITE;
 	down_read(&mm->mmap_sem);
 
 #ifdef CONFIG_PGSTE
-	if (test_tsk_thread_flag(current, TIF_SIE) && S390_lowcore.gmap) {
-		address = __gmap_fault(address,
-				     (struct gmap *) S390_lowcore.gmap);
+	gmap = (struct gmap *)
+		((current->flags & PF_VCPU) ? S390_lowcore.gmap : 0);
+	if (gmap) {
+		address = __gmap_fault(address, gmap);
 		if (address == -EFAULT) {
 			fault = VM_FAULT_BADMAP;
 			goto out_up;
@@ -306,6 +454,8 @@ static inline int do_exception(struct pt_regs *regs, int access)
 			fault = VM_FAULT_OOM;
 			goto out_up;
 		}
+		if (gmap->pfault_enabled)
+			flags |= FAULT_FLAG_RETRY_NOWAIT;
 	}
 #endif
 
@@ -338,6 +488,11 @@ retry:
 	 * the fault.
 	 */
 	fault = handle_mm_fault(mm, vma, address, flags);
+	/* No reason to continue if interrupted by SIGKILL. */
+	if ((fault & VM_FAULT_RETRY) && fatal_signal_pending(current)) {
+		fault = VM_FAULT_SIGNAL;
+		goto out;
+	}
 	if (unlikely(fault & VM_FAULT_ERROR))
 		goto out_up;
 
@@ -357,18 +512,24 @@ retry:
 				      regs, address);
 		}
 		if (fault & VM_FAULT_RETRY) {
+#ifdef CONFIG_PGSTE
+			if (gmap && (flags & FAULT_FLAG_RETRY_NOWAIT)) {
+				/* FAULT_FLAG_RETRY_NOWAIT has been set,
+				 * mmap_sem has not been released */
+				current->thread.gmap_pfault = 1;
+				fault = VM_FAULT_PFAULT;
+				goto out_up;
+			}
+#endif
 			/* Clear FAULT_FLAG_ALLOW_RETRY to avoid any risk
 			 * of starvation. */
-			flags &= ~FAULT_FLAG_ALLOW_RETRY;
+			flags &= ~(FAULT_FLAG_ALLOW_RETRY |
+				   FAULT_FLAG_RETRY_NOWAIT);
+			flags |= FAULT_FLAG_TRIED;
 			down_read(&mm->mmap_sem);
 			goto retry;
 		}
 	}
-	/*
-	 * The instruction that caused the program check will
-	 * be repeated. Don't signal single step via SIGTRAP.
-	 */
-	clear_tsk_thread_flag(tsk, TIF_PER_TRAP);
 	fault = 0;
 out_up:
 	up_read(&mm->mmap_sem);
@@ -382,8 +543,13 @@ void __kprobes do_protection_exception(struct pt_regs *regs)
 	int fault;
 
 	trans_exc_code = regs->int_parm_long;
-	/* Protection exception is suppressing, decrement psw address. */
-	regs->psw.addr = __rewind_psw(regs->psw, regs->int_code >> 16);
+	/*
+	 * Protection exceptions are suppressing, decrement psw address.
+	 * The exception to this rule are aborted transactions, for these
+	 * the PSW already points to the correct location.
+	 */
+	if (!(regs->int_code & 0x200))
+		regs->psw.addr = __rewind_psw(regs->psw, regs->int_code >> 16);
 	/*
 	 * Check for low-address protection.  This needs to be treated
 	 * as a special case because the translation exception code
@@ -408,60 +574,6 @@ void __kprobes do_dat_exception(struct pt_regs *regs)
 		do_fault_error(regs, fault);
 }
 
-#ifdef CONFIG_64BIT
-void __kprobes do_asce_exception(struct pt_regs *regs)
-{
-	struct mm_struct *mm = current->mm;
-	struct vm_area_struct *vma;
-	unsigned long trans_exc_code;
-
-	trans_exc_code = regs->int_parm_long;
-	if (unlikely(!user_space_fault(trans_exc_code) || in_atomic() || !mm))
-		goto no_context;
-
-	down_read(&mm->mmap_sem);
-	vma = find_vma(mm, trans_exc_code & __FAIL_ADDR_MASK);
-	up_read(&mm->mmap_sem);
-
-	if (vma) {
-		update_mm(mm, current);
-		return;
-	}
-
-	/* User mode accesses just cause a SIGSEGV */
-	if (regs->psw.mask & PSW_MASK_PSTATE) {
-		do_sigsegv(regs, SEGV_MAPERR);
-		return;
-	}
-
-no_context:
-	do_no_context(regs);
-}
-#endif
-
-int __handle_fault(unsigned long uaddr, unsigned long pgm_int_code, int write)
-{
-	struct pt_regs regs;
-	int access, fault;
-
-	regs.psw.mask = psw_kernel_bits | PSW_MASK_DAT | PSW_MASK_MCHECK;
-	if (!irqs_disabled())
-		regs.psw.mask |= PSW_MASK_IO | PSW_MASK_EXT;
-	regs.psw.addr = (unsigned long) __builtin_return_address(0);
-	regs.psw.addr |= PSW_ADDR_AMODE;
-	regs.int_code = pgm_int_code;
-	regs.int_parm_long = (uaddr & PAGE_MASK) | 2;
-	access = write ? VM_WRITE : VM_READ;
-	fault = do_exception(&regs, access);
-	if (unlikely(fault)) {
-		if (fault & VM_FAULT_OOM)
-			return -EFAULT;
-		else if (fault & VM_FAULT_SIGBUS)
-			do_sigbus(&regs);
-	}
-	return fault ? -EFAULT : 0;
-}
-
 #ifdef CONFIG_PFAULT 
 /*
  * 'pfault' pseudo page faults routines.
@@ -533,7 +645,7 @@ void pfault_fini(void)
 static DEFINE_SPINLOCK(pfault_lock);
 static LIST_HEAD(pfault_list);
 
-static void pfault_interrupt(unsigned int ext_int_code,
+static void pfault_interrupt(struct ext_code ext_code,
 			     unsigned int param32, unsigned long param64)
 {
 	struct task_struct *tsk;
@@ -546,23 +658,19 @@ static void pfault_interrupt(unsigned int ext_int_code,
 	 * in the 'cpu address' field associated with the
          * external interrupt. 
 	 */
-	subcode = ext_int_code >> 16;
+	subcode = ext_code.subcode;
 	if ((subcode & 0xff00) != __SUBCODE_MASK)
 		return;
-	kstat_cpu(smp_processor_id()).irqs[EXTINT_PFL]++;
-	if (subcode & 0x0080) {
-		/* Get the token (= pid of the affected task). */
-		pid = sizeof(void *) == 4 ? param32 : param64;
-		rcu_read_lock();
-		tsk = find_task_by_pid_ns(pid, &init_pid_ns);
-		if (tsk)
-			get_task_struct(tsk);
-		rcu_read_unlock();
-		if (!tsk)
-			return;
-	} else {
-		tsk = current;
-	}
+	inc_irq_stat(IRQEXT_PFL);
+	/* Get the token (= pid of the affected task). */
+	pid = sizeof(void *) == 4 ? param32 : param64;
+	rcu_read_lock();
+	tsk = find_task_by_pid_ns(pid, &init_pid_ns);
+	if (tsk)
+		get_task_struct(tsk);
+	rcu_read_unlock();
+	if (!tsk)
+		return;
 	spin_lock(&pfault_lock);
 	if (subcode & 0x0080) {
 		/* signal bit is set -> a page has been swapped in by VM */
@@ -575,6 +683,7 @@ static void pfault_interrupt(unsigned int ext_int_code,
 			tsk->thread.pfault_wait = 0;
 			list_del(&tsk->thread.list);
 			wake_up_process(tsk);
+			put_task_struct(tsk);
 		} else {
 			/* Completion interrupt was faster than initial
 			 * interrupt. Set pfault_wait to -1 so the initial
@@ -586,41 +695,52 @@ static void pfault_interrupt(unsigned int ext_int_code,
 			if (tsk->state == TASK_RUNNING)
 				tsk->thread.pfault_wait = -1;
 		}
-		put_task_struct(tsk);
 	} else {
 		/* signal bit not set -> a real page is missing. */
-		if (tsk->thread.pfault_wait == -1) {
+		if (WARN_ON_ONCE(tsk != current))
+			goto out;
+		if (tsk->thread.pfault_wait == 1) {
+			/* Already on the list with a reference: put to sleep */
+			__set_task_state(tsk, TASK_UNINTERRUPTIBLE);
+			set_tsk_need_resched(tsk);
+		} else if (tsk->thread.pfault_wait == -1) {
 			/* Completion interrupt was faster than the initial
 			 * interrupt (pfault_wait == -1). Set pfault_wait
 			 * back to zero and exit. */
 			tsk->thread.pfault_wait = 0;
 		} else {
 			/* Initial interrupt arrived before completion
-			 * interrupt. Let the task sleep. */
+			 * interrupt. Let the task sleep.
+			 * An extra task reference is needed since a different
+			 * cpu may set the task state to TASK_RUNNING again
+			 * before the scheduler is reached. */
+			get_task_struct(tsk);
 			tsk->thread.pfault_wait = 1;
 			list_add(&tsk->thread.list, &pfault_list);
-			set_task_state(tsk, TASK_UNINTERRUPTIBLE);
+			__set_task_state(tsk, TASK_UNINTERRUPTIBLE);
 			set_tsk_need_resched(tsk);
 		}
 	}
+out:
 	spin_unlock(&pfault_lock);
+	put_task_struct(tsk);
 }
 
-static int __cpuinit pfault_cpu_notify(struct notifier_block *self,
-				       unsigned long action, void *hcpu)
+static int pfault_cpu_notify(struct notifier_block *self, unsigned long action,
+			     void *hcpu)
 {
 	struct thread_struct *thread, *next;
 	struct task_struct *tsk;
 
-	switch (action) {
+	switch (action & ~CPU_TASKS_FROZEN) {
 	case CPU_DEAD:
-	case CPU_DEAD_FROZEN:
 		spin_lock_irq(&pfault_lock);
 		list_for_each_entry_safe(thread, next, &pfault_list, list) {
 			thread->pfault_wait = 0;
 			list_del(&thread->list);
 			tsk = container_of(thread, struct task_struct, thread);
 			wake_up_process(tsk);
+			put_task_struct(tsk);
 		}
 		spin_unlock_irq(&pfault_lock);
 		break;
@@ -634,18 +754,18 @@ static int __init pfault_irq_init(void)
 {
 	int rc;
 
-	rc = register_external_interrupt(0x2603, pfault_interrupt);
+	rc = register_external_irq(EXT_IRQ_CP_SERVICE, pfault_interrupt);
 	if (rc)
 		goto out_extint;
 	rc = pfault_init() == 0 ? 0 : -EOPNOTSUPP;
 	if (rc)
 		goto out_pfault;
-	service_subclass_irq_register();
+	irq_subclass_register(IRQ_SUBCLASS_SERVICE_SIGNAL);
 	hotcpu_notifier(pfault_cpu_notify, 0);
 	return 0;
 
 out_pfault:
-	unregister_external_interrupt(0x2603, pfault_interrupt);
+	unregister_external_irq(EXT_IRQ_CP_SERVICE, pfault_interrupt);
 out_extint:
 	pfault_disable = 1;
 	return rc;