1 files changed, 244 insertions, 69 deletions

diff --git a/arch/x86/kernel/dumpstack.c b/arch/x86/kernel/dumpstack.c
index dbce3cca94cb..b10684dedc58 100644
--- a/arch/x86/kernel/dumpstack.c
+++ b/arch/x86/kernel/dumpstack.c
@@ -17,17 +17,18 @@
 #include <linux/bug.h>
 #include <linux/nmi.h>
 #include <linux/sysfs.h>
+#include <linux/kasan.h>
 
+#include <asm/cpu_entry_area.h>
 #include <asm/stacktrace.h>
 #include <asm/unwind.h>
 
-int panic_on_unrecovered_nmi;
-int panic_on_io_nmi;
-unsigned int code_bytes = 64;
 static int die_counter;
 
-bool in_task_stack(unsigned long *stack, struct task_struct *task,
-		   struct stack_info *info)
+static struct pt_regs exec_summary_regs;
+
+bool noinstr in_task_stack(unsigned long *stack, struct task_struct *task,
+			   struct stack_info *info)
 {
 	unsigned long *begin = task_stack_page(task);
 	unsigned long *end   = task_stack_page(task) + THREAD_SIZE;
@@ -43,25 +44,157 @@ bool in_task_stack(unsigned long *stack, struct task_struct *task,
 	return true;
 }
 
+/* Called from get_stack_info_noinstr - so must be noinstr too */
+bool noinstr in_entry_stack(unsigned long *stack, struct stack_info *info)
+{
+	struct entry_stack *ss = cpu_entry_stack(smp_processor_id());
+
+	void *begin = ss;
+	void *end = ss + 1;
+
+	if ((void *)stack < begin || (void *)stack >= end)
+		return false;
+
+	info->type	= STACK_TYPE_ENTRY;
+	info->begin	= begin;
+	info->end	= end;
+	info->next_sp	= NULL;
+
+	return true;
+}
+
 static void printk_stack_address(unsigned long address, int reliable,
-				 char *log_lvl)
+				 const char *log_lvl)
 {
 	touch_nmi_watchdog();
-	printk("%s %s%pB\n", log_lvl, reliable ? "" : "? ", (void *)address);
+	printk("%s %s%pBb\n", log_lvl, reliable ? "" : "? ", (void *)address);
+}
+
+static int copy_code(struct pt_regs *regs, u8 *buf, unsigned long src,
+		     unsigned int nbytes)
+{
+	if (!user_mode(regs))
+		return copy_from_kernel_nofault(buf, (u8 *)src, nbytes);
+
+	/* The user space code from other tasks cannot be accessed. */
+	if (regs != task_pt_regs(current))
+		return -EPERM;
+
+	/*
+	 * Even if named copy_from_user_nmi() this can be invoked from
+	 * other contexts and will not try to resolve a pagefault, which is
+	 * the correct thing to do here as this code can be called from any
+	 * context.
+	 */
+	return copy_from_user_nmi(buf, (void __user *)src, nbytes);
+}
+
+/*
+ * There are a couple of reasons for the 2/3rd prologue, courtesy of Linus:
+ *
+ * In case where we don't have the exact kernel image (which, if we did, we can
+ * simply disassemble and navigate to the RIP), the purpose of the bigger
+ * prologue is to have more context and to be able to correlate the code from
+ * the different toolchains better.
+ *
+ * In addition, it helps in recreating the register allocation of the failing
+ * kernel and thus make sense of the register dump.
+ *
+ * What is more, the additional complication of a variable length insn arch like
+ * x86 warrants having longer byte sequence before rIP so that the disassembler
+ * can "sync" up properly and find instruction boundaries when decoding the
+ * opcode bytes.
+ *
+ * Thus, the 2/3rds prologue and 64 byte OPCODE_BUFSIZE is just a random
+ * guesstimate in attempt to achieve all of the above.
+ */
+void show_opcodes(struct pt_regs *regs, const char *loglvl)
+{
+#define PROLOGUE_SIZE 42
+#define EPILOGUE_SIZE 21
+#define OPCODE_BUFSIZE (PROLOGUE_SIZE + 1 + EPILOGUE_SIZE)
+	u8 opcodes[OPCODE_BUFSIZE];
+	unsigned long prologue = regs->ip - PROLOGUE_SIZE;
+
+	switch (copy_code(regs, opcodes, prologue, sizeof(opcodes))) {
+	case 0:
+		printk("%sCode: %" __stringify(PROLOGUE_SIZE) "ph <%02x> %"
+		       __stringify(EPILOGUE_SIZE) "ph\n", loglvl, opcodes,
+		       opcodes[PROLOGUE_SIZE], opcodes + PROLOGUE_SIZE + 1);
+		break;
+	case -EPERM:
+		/* No access to the user space stack of other tasks. Ignore. */
+		break;
+	default:
+		printk("%sCode: Unable to access opcode bytes at 0x%lx.\n",
+		       loglvl, prologue);
+		break;
+	}
+}
+
+void show_ip(struct pt_regs *regs, const char *loglvl)
+{
+#ifdef CONFIG_X86_32
+	printk("%sEIP: %pS\n", loglvl, (void *)regs->ip);
+#else
+	printk("%sRIP: %04x:%pS\n", loglvl, (int)regs->cs, (void *)regs->ip);
+#endif
+	show_opcodes(regs, loglvl);
+}
+
+void show_iret_regs(struct pt_regs *regs, const char *log_lvl)
+{
+	show_ip(regs, log_lvl);
+	printk("%sRSP: %04x:%016lx EFLAGS: %08lx", log_lvl, (int)regs->ss,
+		regs->sp, regs->flags);
 }
 
-void show_trace_log_lvl(struct task_struct *task, struct pt_regs *regs,
-			unsigned long *stack, char *log_lvl)
+static void show_regs_if_on_stack(struct stack_info *info, struct pt_regs *regs,
+				  bool partial, const char *log_lvl)
+{
+	/*
+	 * These on_stack() checks aren't strictly necessary: the unwind code
+	 * has already validated the 'regs' pointer.  The checks are done for
+	 * ordering reasons: if the registers are on the next stack, we don't
+	 * want to print them out yet.  Otherwise they'll be shown as part of
+	 * the wrong stack.  Later, when show_trace_log_lvl() switches to the
+	 * next stack, this function will be called again with the same regs so
+	 * they can be printed in the right context.
+	 */
+	if (!partial && on_stack(info, regs, sizeof(*regs))) {
+		__show_regs(regs, SHOW_REGS_SHORT, log_lvl);
+
+	} else if (partial && on_stack(info, (void *)regs + IRET_FRAME_OFFSET,
+				       IRET_FRAME_SIZE)) {
+		/*
+		 * When an interrupt or exception occurs in entry code, the
+		 * full pt_regs might not have been saved yet.  In that case
+		 * just print the iret frame.
+		 */
+		show_iret_regs(regs, log_lvl);
+	}
+}
+
+/*
+ * This function reads pointers from the stack and dereferences them. The
+ * pointers may not have their KMSAN shadow set up properly, which may result
+ * in false positive reports. Disable instrumentation to avoid those.
+ */
+__no_kmsan_checks
+static void __show_trace_log_lvl(struct task_struct *task, struct pt_regs *regs,
+				 unsigned long *stack, const char *log_lvl)
 {
 	struct unwind_state state;
 	struct stack_info stack_info = {0};
 	unsigned long visit_mask = 0;
 	int graph_idx = 0;
+	bool partial = false;
 
 	printk("%sCall Trace:\n", log_lvl);
 
 	unwind_start(&state, task, regs, stack);
-	stack = stack ? : get_stack_pointer(task, regs);
+	stack = stack ?: get_stack_pointer(task, regs);
+	regs = unwind_get_entry_regs(&state, &partial);
 
 	/*
 	 * Iterate through the stacks, starting with the current stack pointer.
@@ -71,29 +204,38 @@ void show_trace_log_lvl(struct task_struct *task, struct pt_regs *regs,
 	 * - task stack
 	 * - interrupt stack
 	 * - HW exception stacks (double fault, nmi, debug, mce)
+	 * - entry stack
 	 *
-	 * x86-32 can have up to three stacks:
+	 * x86-32 can have up to four stacks:
 	 * - task stack
 	 * - softirq stack
 	 * - hardirq stack
+	 * - entry stack
 	 */
-	for (regs = NULL; stack; stack = PTR_ALIGN(stack_info.next_sp, sizeof(long))) {
+	for (; stack; stack = stack_info.next_sp) {
 		const char *stack_name;
 
-		/*
-		 * If we overflowed the task stack into a guard page, jump back
-		 * to the bottom of the usable stack.
-		 */
-		if (task_stack_page(task) - (void *)stack < PAGE_SIZE)
-			stack = task_stack_page(task);
+		stack = PTR_ALIGN(stack, sizeof(long));
 
-		if (get_stack_info(stack, task, &stack_info, &visit_mask))
-			break;
+		if (get_stack_info(stack, task, &stack_info, &visit_mask)) {
+			/*
+			 * We weren't on a valid stack.  It's possible that
+			 * we overflowed a valid stack into a guard page.
+			 * See if the next page up is valid so that we can
+			 * generate some kind of backtrace if this happens.
+			 */
+			stack = (unsigned long *)PAGE_ALIGN((unsigned long)stack);
+			if (get_stack_info(stack, task, &stack_info, &visit_mask))
+				break;
+		}
 
 		stack_name = stack_type_name(stack_info.type);
 		if (stack_name)
 			printk("%s <%s>\n", log_lvl, stack_name);
 
+		if (regs)
+			show_regs_if_on_stack(&stack_info, regs, partial, log_lvl);
+
 		/*
 		 * Scan the stack, printing any text addresses we find.  At the
 		 * same time, follow proper stack frames with the unwinder.
@@ -116,12 +258,10 @@ void show_trace_log_lvl(struct task_struct *task, struct pt_regs *regs,
 
 			/*
 			 * Don't print regs->ip again if it was already printed
-			 * by __show_regs() below.
+			 * by show_regs_if_on_stack().
 			 */
-			if (regs && stack == &regs->ip) {
-				unwind_next_frame(&state);
-				continue;
-			}
+			if (regs && stack == &regs->ip)
+				goto next;
 
 			if (stack == ret_addr_p)
 				reliable = 1;
@@ -144,6 +284,7 @@ void show_trace_log_lvl(struct task_struct *task, struct pt_regs *regs,
 			if (!reliable)
 				continue;
 
+next:
 			/*
 			 * Get the next frame from the unwinder.  No need to
 			 * check for an error: if anything goes wrong, the rest
@@ -152,9 +293,9 @@ void show_trace_log_lvl(struct task_struct *task, struct pt_regs *regs,
 			unwind_next_frame(&state);
 
 			/* if the frame has entry regs, print them */
-			regs = unwind_get_entry_regs(&state);
+			regs = unwind_get_entry_regs(&state, &partial);
 			if (regs)
-				__show_regs(regs, 0);
+				show_regs_if_on_stack(&stack_info, regs, partial, log_lvl);
 		}
 
 		if (stack_name)
@@ -162,7 +303,27 @@ void show_trace_log_lvl(struct task_struct *task, struct pt_regs *regs,
 	}
 }
 
-void show_stack(struct task_struct *task, unsigned long *sp)
+static void show_trace_log_lvl(struct task_struct *task, struct pt_regs *regs,
+			       unsigned long *stack, const char *log_lvl)
+{
+	/*
+	 * Disable KASAN to avoid false positives during walking another
+	 * task's stacks, as values on these stacks may change concurrently
+	 * with task execution.
+	 */
+	bool disable_kasan = task && task != current;
+
+	if (disable_kasan)
+		kasan_disable_current();
+
+	__show_trace_log_lvl(task, regs, stack, log_lvl);
+
+	if (disable_kasan)
+		kasan_enable_current();
+}
+
+void show_stack(struct task_struct *task, unsigned long *sp,
+		       const char *loglvl)
 {
 	task = task ? : current;
 
@@ -173,7 +334,7 @@ void show_stack(struct task_struct *task, unsigned long *sp)
 	if (!sp && task == current)
 		sp = get_stack_pointer(current, NULL);
 
-	show_trace_log_lvl(task, NULL, sp, KERN_DEFAULT);
+	show_trace_log_lvl(task, NULL, sp, loglvl);
 }
 
 void show_stack_regs(struct pt_regs *regs)
@@ -207,10 +368,9 @@ unsigned long oops_begin(void)
 	bust_spinlocks(1);
 	return flags;
 }
-EXPORT_SYMBOL_GPL(oops_begin);
 NOKPROBE_SYMBOL(oops_begin);
 
-void __noreturn rewind_stack_do_exit(int signr);
+void __noreturn rewind_stack_and_make_dead(int signr);
 
 void oops_end(unsigned long flags, struct pt_regs *regs, int signr)
 {
@@ -227,6 +387,9 @@ void oops_end(unsigned long flags, struct pt_regs *regs, int signr)
 	raw_local_irq_restore(flags);
 	oops_exit();
 
+	/* Executive summary in case the oops scrolled away */
+	__show_regs(&exec_summary_regs, SHOW_REGS_ALL, KERN_DEFAULT);
+
 	if (!signr)
 		return;
 	if (in_interrupt())
@@ -238,46 +401,49 @@ void oops_end(unsigned long flags, struct pt_regs *regs, int signr)
 	 * We're not going to return, but we might be on an IST stack or
 	 * have very little stack space left.  Rewind the stack and kill
 	 * the task.
+	 * Before we rewind the stack, we have to tell KASAN that we're going to
+	 * reuse the task stack and that existing poisons are invalid.
 	 */
-	rewind_stack_do_exit(signr);
+	kasan_unpoison_task_stack(current);
+	rewind_stack_and_make_dead(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)
 {
-#ifdef CONFIG_X86_32
-	unsigned short ss;
-	unsigned long sp;
-#endif
+	/* Save the regs of the first oops for the executive summary later. */
+	if (!die_counter)
+		exec_summary_regs = *regs;
+
 	printk(KERN_DEFAULT
-	       "%s: %04lx [#%d]%s%s%s%s\n", str, err & 0xffff, ++die_counter,
-	       IS_ENABLED(CONFIG_PREEMPT) ? " PREEMPT"         : "",
+	       "Oops: %s: %04lx [#%d]%s%s%s%s\n", str, err & 0xffff,
+	       ++die_counter,
 	       IS_ENABLED(CONFIG_SMP)     ? " SMP"             : "",
 	       debug_pagealloc_enabled()  ? " DEBUG_PAGEALLOC" : "",
-	       IS_ENABLED(CONFIG_KASAN)   ? " KASAN"           : "");
+	       IS_ENABLED(CONFIG_KASAN)   ? " KASAN"           : "",
+	       IS_ENABLED(CONFIG_MITIGATION_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();
 
 	if (notify_die(DIE_OOPS, str, regs, err,
 			current->thread.trap_nr, SIGSEGV) == NOTIFY_STOP)
 		return 1;
 
-	print_modules();
-	show_regs(regs);
-#ifdef CONFIG_X86_32
-	if (user_mode(regs)) {
-		sp = regs->sp;
-		ss = regs->ss & 0xffff;
-	} else {
-		sp = kernel_stack_pointer(regs);
-		savesegment(ss, ss);
-	}
-	printk(KERN_EMERG "EIP: %pS SS:ESP: %04x:%08lx\n",
-	       (void *)regs->ip, ss, sp);
-#else
-	/* Executive summary in case the oops scrolled away */
-	printk(KERN_ALERT "RIP: %pS RSP: %016lx\n", (void *)regs->ip, regs->sp);
-#endif
 	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);
 
 /*
@@ -294,22 +460,31 @@ void die(const char *str, struct pt_regs *regs, long err)
 	oops_end(flags, regs, sig);
 }
 
-static int __init code_bytes_setup(char *s)
+void die_addr(const char *str, struct pt_regs *regs, long err, long gp_addr)
 {
-	ssize_t ret;
-	unsigned long val;
+	unsigned long flags = oops_begin();
+	int sig = SIGSEGV;
 
-	if (!s)
-		return -EINVAL;
+	__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)
+{
+	enum show_regs_mode print_kernel_regs;
 
-	ret = kstrtoul(s, 0, &val);
-	if (ret)
-		return ret;
+	show_regs_print_info(KERN_DEFAULT);
 
-	code_bytes = val;
-	if (code_bytes > 8192)
-		code_bytes = 8192;
+	print_kernel_regs = user_mode(regs) ? SHOW_REGS_USER : SHOW_REGS_ALL;
+	__show_regs(regs, print_kernel_regs, KERN_DEFAULT);
 
-	return 1;
+	/*
+	 * When in-kernel, we also print out the stack at the time of the fault..
+	 */
+	if (!user_mode(regs))
+		show_trace_log_lvl(current, regs, NULL, KERN_DEFAULT);
 }
-__setup("code_bytes=", code_bytes_setup);