x86/mce: Handle "action required" errors

All non-urgent actions (reporting low severity errors and handling
"action-optional" errors) are now handled by a work queue. This
means that TIF_MCE_NOTIFY can be used to block execution for a
thread experiencing an "action-required" fault until we get all
cpus out of the machine check handler (and the thread that hit
the fault into mce_notify_process().

We use the new mce_{save,find,clear}_info() API to get information
from do_machine_check() to mce_notify_process(), and then use the
newly improved memory_failure(..., MF_ACTION_REQUIRED) to handle
the error (possibly signalling the process).

Update some comments to make the new code flows clearer.

Signed-off-by: Tony Luck <tony.luck@intel.com>

arch/x86/kernel/cpu/mcheck/mce.c

@@ -982,7 +982,9 @@ void do_machine_check(struct pt_regs *regs, long error_code)
 	barrier();
 
 	/*
-	 * When no restart IP must always kill or panic.
+	 * When no restart IP might need to kill or panic.
+	 * Assume the worst for now, but if we find the
+	 * severity is MCE_AR_SEVERITY we have other options.
 	 */
 	if (!(m.mcgstatus & MCG_STATUS_RIPV))
 		kill_it = 1;
@@ -1036,12 +1038,6 @@ void do_machine_check(struct pt_regs *regs, long error_code)
 			continue;
 		}
 
-		/*
-		 * Kill on action required.
-		 */
-		if (severity == MCE_AR_SEVERITY)
-			kill_it = 1;
-
 		mce_read_aux(&m, i);
 
 		/*
@@ -1062,6 +1058,9 @@ void do_machine_check(struct pt_regs *regs, long error_code)
 		}
 	}
 
+	/* mce_clear_state will clear *final, save locally for use later */
+	m = *final;
+
 	if (!no_way_out)
 		mce_clear_state(toclear);
 
@@ -1073,27 +1072,22 @@ void do_machine_check(struct pt_regs *regs, long error_code)
 		no_way_out = worst >= MCE_PANIC_SEVERITY;
 
 	/*
-	 * If we have decided that we just CAN'T continue, and the user
-	 * has not set tolerant to an insane level, give up and die.
-	 *
-	 * This is mainly used in the case when the system doesn't
-	 * support MCE broadcasting or it has been disabled.
+	 * At insane "tolerant" levels we take no action. Otherwise
+	 * we only die if we have no other choice. For less serious
+	 * issues we try to recover, or limit damage to the current
+	 * process.
 	 */
-	if (no_way_out && tolerant < 3)
-		mce_panic("Fatal machine check on current CPU", final, msg);
-
-	/*
-	 * If the error seems to be unrecoverable, something should be
-	 * done.  Try to kill as little as possible.  If we can kill just
-	 * one task, do that.  If the user has set the tolerance very
-	 * high, don't try to do anything at all.
-	 */
-
-	if (kill_it && tolerant < 3)
-		force_sig(SIGBUS, current);
-
-	/* notify userspace ASAP */
-	set_thread_flag(TIF_MCE_NOTIFY);
+	if (tolerant < 3) {
+		if (no_way_out)
+			mce_panic("Fatal machine check on current CPU", &m, msg);
+		if (worst == MCE_AR_SEVERITY) {
+			/* schedule action before return to userland */
+			mce_save_info(m.addr);
+			set_thread_flag(TIF_MCE_NOTIFY);
+		} else if (kill_it) {
+			force_sig(SIGBUS, current);
+		}
+	}
 
 	if (worst > 0)
 		mce_report_event(regs);
@@ -1107,6 +1101,8 @@ EXPORT_SYMBOL_GPL(do_machine_check);
 #ifndef CONFIG_MEMORY_FAILURE
 int memory_failure(unsigned long pfn, int vector, int flags)
 {
+	/* mce_severity() should not hand us an ACTION_REQUIRED error */
+	BUG_ON(flags & MF_ACTION_REQUIRED);
 	printk(KERN_ERR "Uncorrected memory error in page 0x%lx ignored\n"
 		"Rebuild kernel with CONFIG_MEMORY_FAILURE=y for smarter handling\n", pfn);
 
@@ -1115,27 +1111,44 @@ int memory_failure(unsigned long pfn, int vector, int flags)
 #endif
 
 /*
- * Called after mce notification in process context. This code
- * is allowed to sleep. Call the high level VM handler to process
- * any corrupted pages.
- * Assume that the work queue code only calls this one at a time
- * per CPU.
- * Note we don't disable preemption, so this code might run on the wrong
- * CPU. In this case the event is picked up by the scheduled work queue.
- * This is merely a fast path to expedite processing in some common
- * cases.
+ * Called in process context that interrupted by MCE and marked with
+ * TIF_MCE_NOTIFY, just before returning to erroneous userland.
+ * This code is allowed to sleep.
+ * Attempt possible recovery such as calling the high level VM handler to
+ * process any corrupted pages, and kill/signal current process if required.
+ * Action required errors are handled here.
  */
 void mce_notify_process(void)
 {
 	unsigned long pfn;
-	mce_notify_irq();
-	while (mce_ring_get(&pfn))
-		memory_failure(pfn, MCE_VECTOR, 0);
+	struct mce_info *mi = mce_find_info();
+
+	if (!mi)
+		mce_panic("Lost physical address for unconsumed uncorrectable error", NULL, NULL);
+	pfn = mi->paddr >> PAGE_SHIFT;
+
+	clear_thread_flag(TIF_MCE_NOTIFY);
+
+	pr_err("Uncorrected hardware memory error in user-access at %llx",
+		 mi->paddr);
+	if (memory_failure(pfn, MCE_VECTOR, MF_ACTION_REQUIRED) < 0) {
+		pr_err("Memory error not recovered");
+		force_sig(SIGBUS, current);
+	}
+	mce_clear_info(mi);
 }
 
+/*
+ * Action optional processing happens here (picking up
+ * from the list of faulting pages that do_machine_check()
+ * placed into the "ring").
+ */
 static void mce_process_work(struct work_struct *dummy)
 {
-	mce_notify_process();
+	unsigned long pfn;
+
+	while (mce_ring_get(&pfn))
+		memory_failure(pfn, MCE_VECTOR, 0);
 }
 
 #ifdef CONFIG_X86_MCE_INTEL
@@ -1225,8 +1238,6 @@ int mce_notify_irq(void)
 	/* Not more than two messages every minute */
 	static DEFINE_RATELIMIT_STATE(ratelimit, 60*HZ, 2);
 
-	clear_thread_flag(TIF_MCE_NOTIFY);
-
 	if (test_and_clear_bit(0, &mce_need_notify)) {
 		/* wake processes polling /dev/mcelog */
 		wake_up_interruptible(&mce_chrdev_wait);