Merge branch 'x86-core-for-linus' of git://git.kernel.org/pub/scm/lin…

…ux/kernel/git/tip/tip

Pull misc x86 updates from Ingo Molnar:
 "Misc changes:

   - Enhance #GP fault printouts by distinguishing between canonical and
     non-canonical address faults, and also add KASAN fault decoding.

   - Fix/enhance the x86 NMI handler by putting the duration check into
     a direct function call instead of an irq_work which we know to be
     broken in some cases.

   - Clean up do_general_protection() a bit"

* 'x86-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  x86/nmi: Remove irq_work from the long duration NMI handler
  x86/traps: Cleanup do_general_protection()
  x86/kasan: Print original address on #GP
  x86/dumpstack: Introduce die_addr() for die() with #GP fault address
  x86/traps: Print address on #GP
  x86/insn-eval: Add support for 64-bit kernel mode

arch/x86/include/asm/kdebug.h

@@ -33,6 +33,7 @@ enum show_regs_mode {
 };
 
 extern void die(const char *, struct pt_regs *,long);
+void die_addr(const char *str, struct pt_regs *regs, long err, long gp_addr);
 extern int __must_check __die(const char *, struct pt_regs *, long);
 extern void show_stack_regs(struct pt_regs *regs);
 extern void __show_regs(struct pt_regs *regs, enum show_regs_mode);

arch/x86/include/asm/nmi.h

@@ -41,7 +41,6 @@ struct nmiaction {
 	struct list_head	list;
 	nmi_handler_t		handler;
 	u64			max_duration;
-	struct irq_work		irq_work;
 	unsigned long		flags;
 	const char		*name;
 };

arch/x86/include/asm/ptrace.h

@@ -159,6 +159,19 @@ static inline bool user_64bit_mode(struct pt_regs *regs)
 #endif
 }
 
+/*
+ * Determine whether the register set came from any context that is running in
+ * 64-bit mode.
+ */
+static inline bool any_64bit_mode(struct pt_regs *regs)
+{
+#ifdef CONFIG_X86_64
+	return !user_mode(regs) || user_64bit_mode(regs);
+#else
+	return false;
+#endif
+}
+
 #ifdef CONFIG_X86_64
 #define current_user_stack_pointer()	current_pt_regs()->sp
 #define compat_user_stack_pointer()	current_pt_regs()->sp

arch/x86/kernel/dumpstack.c

@@ -365,7 +365,7 @@ void oops_end(unsigned long flags, struct pt_regs *regs, int signr)
 }
 NOKPROBE_SYMBOL(oops_end);
 
-int __die(const char *str, struct pt_regs *regs, long err)
+static void __die_header(const char *str, struct pt_regs *regs, long err)
 {
 	const char *pr = "";
 
@@ -384,7 +384,11 @@ int __die(const char *str, struct pt_regs *regs, long err)
 	       IS_ENABLED(CONFIG_KASAN)   ? " KASAN"           : "",
 	       IS_ENABLED(CONFIG_PAGE_TABLE_ISOLATION) ?
 	       (boot_cpu_has(X86_FEATURE_PTI) ? " PTI" : " NOPTI") : "");
+}
+NOKPROBE_SYMBOL(__die_header);
 
+static int __die_body(const char *str, struct pt_regs *regs, long err)
+{
 	show_regs(regs);
 	print_modules();
 
@@ -394,6 +398,13 @@ int __die(const char *str, struct pt_regs *regs, long err)
 
 	return 0;
 }
+NOKPROBE_SYMBOL(__die_body);
+
+int __die(const char *str, struct pt_regs *regs, long err)
+{
+	__die_header(str, regs, err);
+	return __die_body(str, regs, err);
+}
 NOKPROBE_SYMBOL(__die);
 
 /*
@@ -410,6 +421,19 @@ void die(const char *str, struct pt_regs *regs, long err)
 	oops_end(flags, regs, sig);
 }
 
+void die_addr(const char *str, struct pt_regs *regs, long err, long gp_addr)
+{
+	unsigned long flags = oops_begin();
+	int sig = SIGSEGV;
+
+	__die_header(str, regs, err);
+	if (gp_addr)
+		kasan_non_canonical_hook(gp_addr);
+	if (__die_body(str, regs, err))
+		sig = 0;
+	oops_end(flags, regs, sig);
+}
+
 void show_regs(struct pt_regs *regs)
 {
 	show_regs_print_info(KERN_DEFAULT);

arch/x86/kernel/nmi.c

@@ -104,18 +104,22 @@ static int __init nmi_warning_debugfs(void)
 }
 fs_initcall(nmi_warning_debugfs);
 
-static void nmi_max_handler(struct irq_work *w)
+static void nmi_check_duration(struct nmiaction *action, u64 duration)
 {
-	struct nmiaction *a = container_of(w, struct nmiaction, irq_work);
+	u64 whole_msecs = READ_ONCE(action->max_duration);
 	int remainder_ns, decimal_msecs;
-	u64 whole_msecs = READ_ONCE(a->max_duration);
+
+	if (duration < nmi_longest_ns || duration < action->max_duration)
+		return;
+
+	action->max_duration = duration;
 
 	remainder_ns = do_div(whole_msecs, (1000 * 1000));
 	decimal_msecs = remainder_ns / 1000;
 
 	printk_ratelimited(KERN_INFO
 		"INFO: NMI handler (%ps) took too long to run: %lld.%03d msecs\n",
-		a->handler, whole_msecs, decimal_msecs);
+		action->handler, whole_msecs, decimal_msecs);
 }
 
 static int nmi_handle(unsigned int type, struct pt_regs *regs)
@@ -142,11 +146,7 @@ static int nmi_handle(unsigned int type, struct pt_regs *regs)
 		delta = sched_clock() - delta;
 		trace_nmi_handler(a->handler, (int)delta, thishandled);
 
-		if (delta < nmi_longest_ns || delta < a->max_duration)
-			continue;
-
-		a->max_duration = delta;
-		irq_work_queue(&a->irq_work);
+		nmi_check_duration(a, delta);
 	}
 
 	rcu_read_unlock();
@@ -164,8 +164,6 @@ int __register_nmi_handler(unsigned int type, struct nmiaction *action)
 	if (!action->handler)
 		return -EINVAL;
 
-	init_irq_work(&action->irq_work, nmi_max_handler);
-
 	raw_spin_lock_irqsave(&desc->lock, flags);
 
 	/*

arch/x86/kernel/traps.c

@@ -56,6 +56,8 @@
 #include <asm/mpx.h>
 #include <asm/vm86.h>
 #include <asm/umip.h>
+#include <asm/insn.h>
+#include <asm/insn-eval.h>
 
 #ifdef CONFIG_X86_64
 #include <asm/x86_init.h>
@@ -518,11 +520,57 @@ dotraplinkage void do_bounds(struct pt_regs *regs, long error_code)
 	do_trap(X86_TRAP_BR, SIGSEGV, "bounds", regs, error_code, 0, NULL);
 }
 
-dotraplinkage void
-do_general_protection(struct pt_regs *regs, long error_code)
+enum kernel_gp_hint {
+	GP_NO_HINT,
+	GP_NON_CANONICAL,
+	GP_CANONICAL
+};
+
+/*
+ * When an uncaught #GP occurs, try to determine the memory address accessed by
+ * the instruction and return that address to the caller. Also, try to figure
+ * out whether any part of the access to that address was non-canonical.
+ */
+static enum kernel_gp_hint get_kernel_gp_address(struct pt_regs *regs,
+						 unsigned long *addr)
 {
-	const char *desc = "general protection fault";
+	u8 insn_buf[MAX_INSN_SIZE];
+	struct insn insn;
+
+	if (probe_kernel_read(insn_buf, (void *)regs->ip, MAX_INSN_SIZE))
+		return GP_NO_HINT;
+
+	kernel_insn_init(&insn, insn_buf, MAX_INSN_SIZE);
+	insn_get_modrm(&insn);
+	insn_get_sib(&insn);
+
+	*addr = (unsigned long)insn_get_addr_ref(&insn, regs);
+	if (*addr == -1UL)
+		return GP_NO_HINT;
+
+#ifdef CONFIG_X86_64
+	/*
+	 * Check that:
+	 *  - the operand is not in the kernel half
+	 *  - the last byte of the operand is not in the user canonical half
+	 */
+	if (*addr < ~__VIRTUAL_MASK &&
+	    *addr + insn.opnd_bytes - 1 > __VIRTUAL_MASK)
+		return GP_NON_CANONICAL;
+#endif
+
+	return GP_CANONICAL;
+}
+
+#define GPFSTR "general protection fault"
+
+dotraplinkage void do_general_protection(struct pt_regs *regs, long error_code)
+{
+	char desc[sizeof(GPFSTR) + 50 + 2*sizeof(unsigned long) + 1] = GPFSTR;
+	enum kernel_gp_hint hint = GP_NO_HINT;
 	struct task_struct *tsk;
+	unsigned long gp_addr;
+	int ret;
 
 	RCU_LOCKDEP_WARN(!rcu_is_watching(), "entry code didn't wake RCU");
 	cond_local_irq_enable(regs);
@@ -539,34 +587,56 @@ do_general_protection(struct pt_regs *regs, long error_code)
 	}
 
 	tsk = current;
-	if (!user_mode(regs)) {
-		if (fixup_exception(regs, X86_TRAP_GP, error_code, 0))
-			return;
 
+	if (user_mode(regs)) {
 		tsk->thread.error_code = error_code;
 		tsk->thread.trap_nr = X86_TRAP_GP;
 
-		/*
-		 * To be potentially processing a kprobe fault and to
-		 * trust the result from kprobe_running(), we have to
-		 * be non-preemptible.
-		 */
-		if (!preemptible() && kprobe_running() &&
-		    kprobe_fault_handler(regs, X86_TRAP_GP))
-			return;
+		show_signal(tsk, SIGSEGV, "", desc, regs, error_code);
+		force_sig(SIGSEGV);
 
-		if (notify_die(DIE_GPF, desc, regs, error_code,
-			       X86_TRAP_GP, SIGSEGV) != NOTIFY_STOP)
-			die(desc, regs, error_code);
 		return;
 	}
 
+	if (fixup_exception(regs, X86_TRAP_GP, error_code, 0))
+		return;
+
 	tsk->thread.error_code = error_code;
 	tsk->thread.trap_nr = X86_TRAP_GP;
 
-	show_signal(tsk, SIGSEGV, "", desc, regs, error_code);
+	/*
+	 * To be potentially processing a kprobe fault and to trust the result
+	 * from kprobe_running(), we have to be non-preemptible.
+	 */
+	if (!preemptible() &&
+	    kprobe_running() &&
+	    kprobe_fault_handler(regs, X86_TRAP_GP))
+		return;
+
+	ret = notify_die(DIE_GPF, desc, regs, error_code, X86_TRAP_GP, SIGSEGV);
+	if (ret == NOTIFY_STOP)
+		return;
+
+	if (error_code)
+		snprintf(desc, sizeof(desc), "segment-related " GPFSTR);
+	else
+		hint = get_kernel_gp_address(regs, &gp_addr);
+
+	if (hint != GP_NO_HINT)
+		snprintf(desc, sizeof(desc), GPFSTR ", %s 0x%lx",
+			 (hint == GP_NON_CANONICAL) ? "probably for non-canonical address"
+						    : "maybe for address",
+			 gp_addr);
+
+	/*
+	 * KASAN is interested only in the non-canonical case, clear it
+	 * otherwise.
+	 */
+	if (hint != GP_NON_CANONICAL)
+		gp_addr = 0;
+
+	die_addr(desc, regs, error_code, gp_addr);
 
-	force_sig(SIGSEGV);
 }
 NOKPROBE_SYMBOL(do_general_protection);
 

arch/x86/lib/insn-eval.c

@@ -155,7 +155,7 @@ static bool check_seg_overrides(struct insn *insn, int regoff)
  */
 static int resolve_default_seg(struct insn *insn, struct pt_regs *regs, int off)
 {
-	if (user_64bit_mode(regs))
+	if (any_64bit_mode(regs))
 		return INAT_SEG_REG_IGNORE;
 	/*
 	 * Resolve the default segment register as described in Section 3.7.4
@@ -266,7 +266,7 @@ static int resolve_seg_reg(struct insn *insn, struct pt_regs *regs, int regoff)
 	 * which may be invalid at this point.
 	 */
 	if (regoff == offsetof(struct pt_regs, ip)) {
-		if (user_64bit_mode(regs))
+		if (any_64bit_mode(regs))
 			return INAT_SEG_REG_IGNORE;
 		else
 			return INAT_SEG_REG_CS;
@@ -289,7 +289,7 @@ static int resolve_seg_reg(struct insn *insn, struct pt_regs *regs, int regoff)
 	 * In long mode, segment override prefixes are ignored, except for
 	 * overrides for FS and GS.
 	 */
-	if (user_64bit_mode(regs)) {
+	if (any_64bit_mode(regs)) {
 		if (idx != INAT_SEG_REG_FS &&
 		    idx != INAT_SEG_REG_GS)
 			idx = INAT_SEG_REG_IGNORE;
@@ -646,23 +646,27 @@ unsigned long insn_get_seg_base(struct pt_regs *regs, int seg_reg_idx)
 		 */
 		return (unsigned long)(sel << 4);
 
-	if (user_64bit_mode(regs)) {
+	if (any_64bit_mode(regs)) {
 		/*
 		 * Only FS or GS will have a base address, the rest of
 		 * the segments' bases are forced to 0.
 		 */
 		unsigned long base;
 
-		if (seg_reg_idx == INAT_SEG_REG_FS)
+		if (seg_reg_idx == INAT_SEG_REG_FS) {
 			rdmsrl(MSR_FS_BASE, base);
-		else if (seg_reg_idx == INAT_SEG_REG_GS)
+		} else if (seg_reg_idx == INAT_SEG_REG_GS) {
 			/*
 			 * swapgs was called at the kernel entry point. Thus,
 			 * MSR_KERNEL_GS_BASE will have the user-space GS base.
 			 */
-			rdmsrl(MSR_KERNEL_GS_BASE, base);
-		else
+			if (user_mode(regs))
+				rdmsrl(MSR_KERNEL_GS_BASE, base);
+			else
+				rdmsrl(MSR_GS_BASE, base);
+		} else {
 			base = 0;
+		}
 		return base;
 	}
 
@@ -703,7 +707,7 @@ static unsigned long get_seg_limit(struct pt_regs *regs, int seg_reg_idx)
 	if (sel < 0)
 		return 0;
 
-	if (user_64bit_mode(regs) || v8086_mode(regs))
+	if (any_64bit_mode(regs) || v8086_mode(regs))
 		return -1L;
 
 	if (!sel)
@@ -948,7 +952,7 @@ static int get_eff_addr_modrm(struct insn *insn, struct pt_regs *regs,
 	 * following instruction.
 	 */
 	if (*regoff == -EDOM) {
-		if (user_64bit_mode(regs))
+		if (any_64bit_mode(regs))
 			tmp = regs->ip + insn->length;
 		else
 			tmp = 0;
@@ -1250,7 +1254,7 @@ static void __user *get_addr_ref_32(struct insn *insn, struct pt_regs *regs)
 	 * After computed, the effective address is treated as an unsigned
 	 * quantity.
 	 */
-	if (!user_64bit_mode(regs) && ((unsigned int)eff_addr > seg_limit))
+	if (!any_64bit_mode(regs) && ((unsigned int)eff_addr > seg_limit))
 		goto out;
 
 	/*