Merge branch 'x86-mce-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

Pull MCE changes from Ingo Molnar.

* 'x86-mce-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  x86/mce: Fix return value of mce_chrdev_read() when erst is disabled
  x86/mce: Convert static array of pointers to per-cpu variables
  x86/mce: Replace hard coded hex constants with symbolic defines
  x86/mce: Recognise machine check bank signature for data path error
  x86/mce: Handle "action required" errors
  x86/mce: Add mechanism to safely save information in MCE handler
  x86/mce: Create helper function to save addr/misc when needed
  HWPOISON: Add code to handle "action required" errors.
  HWPOISON: Clean up memory_failure() vs. __memory_failure()

3 files changed, 164 insertions, 64 deletions

diff --git a/arch/x86/kernel/cpu/mcheck/mce-severity.c b/arch/x86/kernel/cpu/mcheck/mce-severity.c
index 7395d5f4272d..0c82091b1652 100644
--- a/arch/x86/kernel/cpu/mcheck/mce-severity.c
+++ b/arch/x86/kernel/cpu/mcheck/mce-severity.c
@@ -54,7 +54,14 @@ static struct severity {
 #define  MASK(x, y)	.mask = x, .result = y
 #define MCI_UC_S (MCI_STATUS_UC|MCI_STATUS_S)
 #define MCI_UC_SAR (MCI_STATUS_UC|MCI_STATUS_S|MCI_STATUS_AR)
+#define	MCI_ADDR (MCI_STATUS_ADDRV|MCI_STATUS_MISCV)
 #define MCACOD 0xffff
+/* Architecturally defined codes from SDM Vol. 3B Chapter 15 */
+#define MCACOD_SCRUB	0x00C0	/* 0xC0-0xCF Memory Scrubbing */
+#define MCACOD_SCRUBMSK	0xfff0
+#define MCACOD_L3WB	0x017A	/* L3 Explicit Writeback */
+#define MCACOD_DATA	0x0134	/* Data Load */
+#define MCACOD_INSTR	0x0150	/* Instruction Fetch */
 
 	MCESEV(
 		NO, "Invalid",
@@ -102,11 +109,24 @@ static struct severity {
 		SER, BITCLR(MCI_STATUS_S)
 		),
 
-	/* AR add known MCACODs here */
 	MCESEV(
 		PANIC, "Action required with lost events",
 		SER, BITSET(MCI_STATUS_OVER|MCI_UC_SAR)
 		),
+
+	/* known AR MCACODs: */
+#ifdef	CONFIG_MEMORY_FAILURE
+	MCESEV(
+		KEEP, "HT thread notices Action required: data load error",
+		SER, MASK(MCI_STATUS_OVER|MCI_UC_SAR|MCI_ADDR|MCACOD, MCI_UC_SAR|MCI_ADDR|MCACOD_DATA),
+		MCGMASK(MCG_STATUS_EIPV, 0)
+		),
+	MCESEV(
+		AR, "Action required: data load error",
+		SER, MASK(MCI_STATUS_OVER|MCI_UC_SAR|MCI_ADDR|MCACOD, MCI_UC_SAR|MCI_ADDR|MCACOD_DATA),
+		USER
+		),
+#endif
 	MCESEV(
 		PANIC, "Action required: unknown MCACOD",
 		SER, MASK(MCI_STATUS_OVER|MCI_UC_SAR, MCI_UC_SAR)
@@ -115,11 +135,11 @@ static struct severity {
 	/* known AO MCACODs: */
 	MCESEV(
 		AO, "Action optional: memory scrubbing error",
-		SER, MASK(MCI_STATUS_OVER|MCI_UC_SAR|0xfff0, MCI_UC_S|0x00c0)
+		SER, MASK(MCI_STATUS_OVER|MCI_UC_SAR|MCACOD_SCRUBMSK, MCI_UC_S|MCACOD_SCRUB)
 		),
 	MCESEV(
 		AO, "Action optional: last level cache writeback error",
-		SER, MASK(MCI_STATUS_OVER|MCI_UC_SAR|MCACOD, MCI_UC_S|0x017a)
+		SER, MASK(MCI_STATUS_OVER|MCI_UC_SAR|MCACOD, MCI_UC_S|MCACOD_L3WB)
 		),
 	MCESEV(
 		SOME, "Action optional: unknown MCACOD",
diff --git a/arch/x86/kernel/cpu/mcheck/mce.c b/arch/x86/kernel/cpu/mcheck/mce.c
index 5a11ae2e9e91..c614bd4de0f3 100644
--- a/arch/x86/kernel/cpu/mcheck/mce.c
+++ b/arch/x86/kernel/cpu/mcheck/mce.c
@@ -540,6 +540,27 @@ static void mce_report_event(struct pt_regs *regs)
 	irq_work_queue(&__get_cpu_var(mce_irq_work));
 }
 
+/*
+ * Read ADDR and MISC registers.
+ */
+static void mce_read_aux(struct mce *m, int i)
+{
+	if (m->status & MCI_STATUS_MISCV)
+		m->misc = mce_rdmsrl(MSR_IA32_MCx_MISC(i));
+	if (m->status & MCI_STATUS_ADDRV) {
+		m->addr = mce_rdmsrl(MSR_IA32_MCx_ADDR(i));
+
+		/*
+		 * Mask the reported address by the reported granularity.
+		 */
+		if (mce_ser && (m->status & MCI_STATUS_MISCV)) {
+			u8 shift = MCI_MISC_ADDR_LSB(m->misc);
+			m->addr >>= shift;
+			m->addr <<= shift;
+		}
+	}
+}
+
 DEFINE_PER_CPU(unsigned, mce_poll_count);
 
 /*
@@ -590,10 +611,7 @@ void machine_check_poll(enum mcp_flags flags, mce_banks_t *b)
 		    (m.status & (mce_ser ? MCI_STATUS_S : MCI_STATUS_UC)))
 			continue;
 
-		if (m.status & MCI_STATUS_MISCV)
-			m.misc = mce_rdmsrl(MSR_IA32_MCx_MISC(i));
-		if (m.status & MCI_STATUS_ADDRV)
-			m.addr = mce_rdmsrl(MSR_IA32_MCx_ADDR(i));
+		mce_read_aux(&m, i);
 
 		if (!(flags & MCP_TIMESTAMP))
 			m.tsc = 0;
@@ -917,6 +935,49 @@ static void mce_clear_state(unsigned long *toclear)
 }
 
 /*
+ * Need to save faulting physical address associated with a process
+ * in the machine check handler some place where we can grab it back
+ * later in mce_notify_process()
+ */
+#define	MCE_INFO_MAX	16
+
+struct mce_info {
+	atomic_t		inuse;
+	struct task_struct	*t;
+	__u64			paddr;
+} mce_info[MCE_INFO_MAX];
+
+static void mce_save_info(__u64 addr)
+{
+	struct mce_info *mi;
+
+	for (mi = mce_info; mi < &mce_info[MCE_INFO_MAX]; mi++) {
+		if (atomic_cmpxchg(&mi->inuse, 0, 1) == 0) {
+			mi->t = current;
+			mi->paddr = addr;
+			return;
+		}
+	}
+
+	mce_panic("Too many concurrent recoverable errors", NULL, NULL);
+}
+
+static struct mce_info *mce_find_info(void)
+{
+	struct mce_info *mi;
+
+	for (mi = mce_info; mi < &mce_info[MCE_INFO_MAX]; mi++)
+		if (atomic_read(&mi->inuse) && mi->t == current)
+			return mi;
+	return NULL;
+}
+
+static void mce_clear_info(struct mce_info *mi)
+{
+	atomic_set(&mi->inuse, 0);
+}
+
+/*
  * The actual machine check handler. This only handles real
  * exceptions when something got corrupted coming in through int 18.
  *
@@ -969,7 +1030,9 @@ 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;
@@ -1023,16 +1086,7 @@ void do_machine_check(struct pt_regs *regs, long error_code)
 			continue;
 		}
 
-		/*
-		 * Kill on action required.
-		 */
-		if (severity == MCE_AR_SEVERITY)
-			kill_it = 1;
-
-		if (m.status & MCI_STATUS_MISCV)
-			m.misc = mce_rdmsrl(MSR_IA32_MCx_MISC(i));
-		if (m.status & MCI_STATUS_ADDRV)
-			m.addr = mce_rdmsrl(MSR_IA32_MCx_ADDR(i));
+		mce_read_aux(&m, i);
 
 		/*
 		 * Action optional error. Queue address for later processing.
@@ -1052,6 +1106,9 @@ 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);
 
@@ -1063,27 +1120,22 @@ 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.
-	 */
-	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.
+	 * 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 (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);
@@ -1094,34 +1146,57 @@ out:
 }
 EXPORT_SYMBOL_GPL(do_machine_check);
 
-/* dummy to break dependency. actual code is in mm/memory-failure.c */
-void __attribute__((weak)) memory_failure(unsigned long pfn, int vector)
+#ifndef CONFIG_MEMORY_FAILURE
+int memory_failure(unsigned long pfn, int vector, int flags)
 {
-	printk(KERN_ERR "Action optional memory failure at %lx ignored\n", pfn);
+	/* 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);
+
+	return 0;
 }
+#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);
+	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
@@ -1211,8 +1286,6 @@ 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);
@@ -1541,6 +1614,12 @@ static int __mce_read_apei(char __user **ubuf, size_t usize)
 	/* Error or no more MCE record */
 	if (rc <= 0) {
 		mce_apei_read_done = 1;
+		/*
+		 * When ERST is disabled, mce_chrdev_read() should return
+		 * "no record" instead of "no device."
+		 */
+		if (rc == -ENODEV)
+			return 0;
 		return rc;
 	}
 	rc = -EFAULT;
@@ -1859,7 +1938,7 @@ static struct bus_type mce_subsys = {
 	.dev_name	= "machinecheck",
 };
 
-struct device *mce_device[CONFIG_NR_CPUS];
+DEFINE_PER_CPU(struct device *, mce_device);
 
 __cpuinitdata
 void (*threshold_cpu_callback)(unsigned long action, unsigned int cpu);
@@ -2038,7 +2117,7 @@ static __cpuinit int mce_device_create(unsigned int cpu)
 			goto error2;
 	}
 	cpumask_set_cpu(cpu, mce_device_initialized);
-	mce_device[cpu] = dev;
+	per_cpu(mce_device, cpu) = dev;
 
 	return 0;
 error2:
@@ -2055,7 +2134,7 @@ error:
 
 static __cpuinit void mce_device_remove(unsigned int cpu)
 {
-	struct device *dev = mce_device[cpu];
+	struct device *dev = per_cpu(mce_device, cpu);
 	int i;
 
 	if (!cpumask_test_cpu(cpu, mce_device_initialized))
@@ -2069,7 +2148,7 @@ static __cpuinit void mce_device_remove(unsigned int cpu)
 
 	device_unregister(dev);
 	cpumask_clear_cpu(cpu, mce_device_initialized);
-	mce_device[cpu] = NULL;
+	per_cpu(mce_device, cpu) = NULL;
 }
 
 /* Make sure there are no machine checks on offlined CPUs. */
diff --git a/arch/x86/kernel/cpu/mcheck/mce_amd.c b/arch/x86/kernel/cpu/mcheck/mce_amd.c
index e4eeaaf58a47..99b57179f912 100644
--- a/arch/x86/kernel/cpu/mcheck/mce_amd.c
+++ b/arch/x86/kernel/cpu/mcheck/mce_amd.c
@@ -523,7 +523,7 @@ static __cpuinit int threshold_create_bank(unsigned int cpu, unsigned int bank)
 {
 	int i, err = 0;
 	struct threshold_bank *b = NULL;
-	struct device *dev = mce_device[cpu];
+	struct device *dev = per_cpu(mce_device, cpu);
 	char name[32];
 
 	sprintf(name, "threshold_bank%i", bank);
@@ -587,7 +587,7 @@ static __cpuinit int threshold_create_bank(unsigned int cpu, unsigned int bank)
 		if (i == cpu)
 			continue;
 
-		dev = mce_device[i];
+		dev = per_cpu(mce_device, i);
 		if (dev)
 			err = sysfs_create_link(&dev->kobj,b->kobj, name);
 		if (err)
@@ -667,7 +667,8 @@ static void threshold_remove_bank(unsigned int cpu, int bank)
 #ifdef CONFIG_SMP
 	/* sibling symlink */
 	if (shared_bank[bank] && b->blocks->cpu != cpu) {
-		sysfs_remove_link(&mce_device[cpu]->kobj, name);
+		dev = per_cpu(mce_device, cpu);
+		sysfs_remove_link(&dev->kobj, name);
 		per_cpu(threshold_banks, cpu)[bank] = NULL;
 
 		return;
@@ -679,7 +680,7 @@ static void threshold_remove_bank(unsigned int cpu, int bank)
 		if (i == cpu)
 			continue;
 
-		dev = mce_device[i];
+		dev = per_cpu(mce_device, i);
 		if (dev)
 			sysfs_remove_link(&dev->kobj, name);
 		per_cpu(threshold_banks, i)[bank] = NULL;