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

Pull x86/mce changes from Ingo Molnar:
 "This tree improves the AMD thresholding bank code and includes a
  memory fault signal handling fixlet."

* 'x86-mce-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  x86/mce: Fix siginfo_t->si_addr value for non-recoverable memory faults
  x86, MCE, AMD: Update copyrights and boilerplate
  x86, MCE, AMD: Give proper names to the thresholding banks
  x86, MCE, AMD: Make error_count read only
  x86, MCE, AMD: Cleanup reading of error_count
  x86, MCE, AMD: Print decimal thresholding values
  x86, MCE, AMD: Move shared bank to node descriptor
  x86, MCE, AMD: Remove local_allocate_... wrapper
  x86, MCE, AMD: Remove shared banks sysfs linking
  x86, amd_nb: Export model 0x10 and later PCI id

8 files changed, 160 insertions, 153 deletions

diff --git a/arch/x86/include/asm/amd_nb.h b/arch/x86/include/asm/amd_nb.h
index 49ad773f4b9f..b3341e9cd8fd 100644
--- a/arch/x86/include/asm/amd_nb.h
+++ b/arch/x86/include/asm/amd_nb.h
@@ -26,10 +26,31 @@ struct amd_l3_cache {
 	u8	 subcaches[4];
 };
 
+struct threshold_block {
+	unsigned int		block;
+	unsigned int		bank;
+	unsigned int		cpu;
+	u32			address;
+	u16			interrupt_enable;
+	bool			interrupt_capable;
+	u16			threshold_limit;
+	struct kobject		kobj;
+	struct list_head	miscj;
+};
+
+struct threshold_bank {
+	struct kobject		*kobj;
+	struct threshold_block	*blocks;
+
+	/* initialized to the number of CPUs on the node sharing this bank */
+	atomic_t		cpus;
+};
+
 struct amd_northbridge {
 	struct pci_dev *misc;
 	struct pci_dev *link;
 	struct amd_l3_cache l3_cache;
+	struct threshold_bank *bank4;
 };
 
 struct amd_northbridge_info {
diff --git a/arch/x86/kernel/amd_nb.c b/arch/x86/kernel/amd_nb.c
index f29f6dd6bc08..aadf3359e2a7 100644
--- a/arch/x86/kernel/amd_nb.c
+++ b/arch/x86/kernel/amd_nb.c
@@ -19,6 +19,7 @@ const struct pci_device_id amd_nb_misc_ids[] = {
 	{ PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_K8_NB_MISC) },
 	{ PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_10H_NB_MISC) },
 	{ PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_15H_NB_F3) },
+	{ PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_15H_M10H_F3) },
 	{}
 };
 EXPORT_SYMBOL(amd_nb_misc_ids);
diff --git a/arch/x86/kernel/cpu/mcheck/mce.c b/arch/x86/kernel/cpu/mcheck/mce.c
index 5a5a5dc1ff15..9473e8772fd1 100644
--- a/arch/x86/kernel/cpu/mcheck/mce.c
+++ b/arch/x86/kernel/cpu/mcheck/mce.c
@@ -1190,6 +1190,7 @@ void mce_notify_process(void)
 {
 	unsigned long pfn;
 	struct mce_info *mi = mce_find_info();
+	int flags = MF_ACTION_REQUIRED;
 
 	if (!mi)
 		mce_panic("Lost physical address for unconsumed uncorrectable error", NULL, NULL);
@@ -1204,8 +1205,9 @@ void mce_notify_process(void)
 	 * doomed. We still need to mark the page as poisoned and alert any
 	 * other users of the page.
 	 */
-	if (memory_failure(pfn, MCE_VECTOR, MF_ACTION_REQUIRED) < 0 ||
-			   mi->restartable == 0) {
+	if (!mi->restartable)
+		flags |= MF_MUST_KILL;
+	if (memory_failure(pfn, MCE_VECTOR, flags) < 0) {
 		pr_err("Memory error not recovered");
 		force_sig(SIGBUS, current);
 	}
diff --git a/arch/x86/kernel/cpu/mcheck/mce_amd.c b/arch/x86/kernel/cpu/mcheck/mce_amd.c
index f4873a64f46d..671b95a2ffb5 100644
--- a/arch/x86/kernel/cpu/mcheck/mce_amd.c
+++ b/arch/x86/kernel/cpu/mcheck/mce_amd.c
@@ -1,15 +1,17 @@
 /*
- *  (c) 2005, 2006 Advanced Micro Devices, Inc.
+ *  (c) 2005-2012 Advanced Micro Devices, Inc.
  *  Your use of this code is subject to the terms and conditions of the
  *  GNU general public license version 2. See "COPYING" or
  *  http://www.gnu.org/licenses/gpl.html
  *
  *  Written by Jacob Shin - AMD, Inc.
  *
- *  Support : jacob.shin@amd.com
+ *  Support: borislav.petkov@amd.com
  *
  *  April 2006
  *     - added support for AMD Family 0x10 processors
+ *  May 2012
+ *     - major scrubbing
  *
  *  All MC4_MISCi registers are shared between multi-cores
  */
@@ -25,6 +27,7 @@
 #include <linux/cpu.h>
 #include <linux/smp.h>
 
+#include <asm/amd_nb.h>
 #include <asm/apic.h>
 #include <asm/idle.h>
 #include <asm/mce.h>
@@ -45,23 +48,15 @@
 #define MASK_BLKPTR_LO    0xFF000000
 #define MCG_XBLK_ADDR     0xC0000400
 
-struct threshold_block {
-	unsigned int		block;
-	unsigned int		bank;
-	unsigned int		cpu;
-	u32			address;
-	u16			interrupt_enable;
-	bool			interrupt_capable;
-	u16			threshold_limit;
-	struct kobject		kobj;
-	struct list_head	miscj;
+static const char * const th_names[] = {
+	"load_store",
+	"insn_fetch",
+	"combined_unit",
+	"",
+	"northbridge",
+	"execution_unit",
 };
 
-struct threshold_bank {
-	struct kobject		*kobj;
-	struct threshold_block	*blocks;
-	cpumask_var_t		cpus;
-};
 static DEFINE_PER_CPU(struct threshold_bank * [NR_BANKS], threshold_banks);
 
 static unsigned char shared_bank[NR_BANKS] = {
@@ -84,6 +79,26 @@ struct thresh_restart {
 	u16			old_limit;
 };
 
+static const char * const bank4_names(struct threshold_block *b)
+{
+	switch (b->address) {
+	/* MSR4_MISC0 */
+	case 0x00000413:
+		return "dram";
+
+	case 0xc0000408:
+		return "ht_links";
+
+	case 0xc0000409:
+		return "l3_cache";
+
+	default:
+		WARN(1, "Funny MSR: 0x%08x\n", b->address);
+		return "";
+	}
+};
+
+
 static bool lvt_interrupt_supported(unsigned int bank, u32 msr_high_bits)
 {
 	/*
@@ -224,8 +239,6 @@ void mce_amd_feature_init(struct cpuinfo_x86 *c)
 
 			if (!block)
 				per_cpu(bank_map, cpu) |= (1 << bank);
-			if (shared_bank[bank] && c->cpu_core_id)
-				break;
 
 			memset(&b, 0, sizeof(b));
 			b.cpu			= cpu;
@@ -326,7 +339,7 @@ struct threshold_attr {
 #define SHOW_FIELDS(name)						\
 static ssize_t show_ ## name(struct threshold_block *b, char *buf)	\
 {									\
-	return sprintf(buf, "%lx\n", (unsigned long) b->name);		\
+	return sprintf(buf, "%lu\n", (unsigned long) b->name);		\
 }
 SHOW_FIELDS(interrupt_enable)
 SHOW_FIELDS(threshold_limit)
@@ -377,38 +390,21 @@ store_threshold_limit(struct threshold_block *b, const char *buf, size_t size)
 	return size;
 }
 
-struct threshold_block_cross_cpu {
-	struct threshold_block	*tb;
-	long			retval;
-};
-
-static void local_error_count_handler(void *_tbcc)
-{
-	struct threshold_block_cross_cpu *tbcc = _tbcc;
-	struct threshold_block *b = tbcc->tb;
-	u32 low, high;
-
-	rdmsr(b->address, low, high);
-	tbcc->retval = (high & 0xFFF) - (THRESHOLD_MAX - b->threshold_limit);
-}
-
 static ssize_t show_error_count(struct threshold_block *b, char *buf)
 {
-	struct threshold_block_cross_cpu tbcc = { .tb = b, };
+	u32 lo, hi;
 
-	smp_call_function_single(b->cpu, local_error_count_handler, &tbcc, 1);
-	return sprintf(buf, "%lx\n", tbcc.retval);
-}
+	rdmsr_on_cpu(b->cpu, b->address, &lo, &hi);
 
-static ssize_t store_error_count(struct threshold_block *b,
-				 const char *buf, size_t count)
-{
-	struct thresh_restart tr = { .b = b, .reset = 1, .old_limit = 0 };
-
-	smp_call_function_single(b->cpu, threshold_restart_bank, &tr, 1);
-	return 1;
+	return sprintf(buf, "%u\n", ((hi & THRESHOLD_MAX) -
+				     (THRESHOLD_MAX - b->threshold_limit)));
 }
 
+static struct threshold_attr error_count = {
+	.attr = {.name = __stringify(error_count), .mode = 0444 },
+	.show = show_error_count,
+};
+
 #define RW_ATTR(val)							\
 static struct threshold_attr val = {					\
 	.attr	= {.name = __stringify(val), .mode = 0644 },		\
@@ -418,7 +414,6 @@ static struct threshold_attr val = {					\
 
 RW_ATTR(interrupt_enable);
 RW_ATTR(threshold_limit);
-RW_ATTR(error_count);
 
 static struct attribute *default_attrs[] = {
 	&threshold_limit.attr,
@@ -517,7 +512,7 @@ static __cpuinit int allocate_threshold_blocks(unsigned int cpu,
 
 	err = kobject_init_and_add(&b->kobj, &threshold_ktype,
 				   per_cpu(threshold_banks, cpu)[bank]->kobj,
-				   "misc%i", block);
+				   (bank == 4 ? bank4_names(b) : th_names[bank]));
 	if (err)
 		goto out_free;
 recurse:
@@ -548,98 +543,91 @@ out_free:
 	return err;
 }
 
-static __cpuinit long
-local_allocate_threshold_blocks(int cpu, unsigned int bank)
+static __cpuinit int __threshold_add_blocks(struct threshold_bank *b)
 {
-	return allocate_threshold_blocks(cpu, bank, 0,
-					 MSR_IA32_MC0_MISC + bank * 4);
+	struct list_head *head = &b->blocks->miscj;
+	struct threshold_block *pos = NULL;
+	struct threshold_block *tmp = NULL;
+	int err = 0;
+
+	err = kobject_add(&b->blocks->kobj, b->kobj, b->blocks->kobj.name);
+	if (err)
+		return err;
+
+	list_for_each_entry_safe(pos, tmp, head, miscj) {
+
+		err = kobject_add(&pos->kobj, b->kobj, pos->kobj.name);
+		if (err) {
+			list_for_each_entry_safe_reverse(pos, tmp, head, miscj)
+				kobject_del(&pos->kobj);
+
+			return err;
+		}
+	}
+	return err;
 }
 
-/* symlinks sibling shared banks to first core.  first core owns dir/files. */
 static __cpuinit int threshold_create_bank(unsigned int cpu, unsigned int bank)
 {
-	int i, err = 0;
-	struct threshold_bank *b = NULL;
 	struct device *dev = per_cpu(mce_device, cpu);
-	char name[32];
-
-	sprintf(name, "threshold_bank%i", bank);
+	struct amd_northbridge *nb = NULL;
+	struct threshold_bank *b = NULL;
+	const char *name = th_names[bank];
+	int err = 0;
 
-#ifdef CONFIG_SMP
-	if (cpu_data(cpu).cpu_core_id && shared_bank[bank]) {	/* symlink */
-		i = cpumask_first(cpu_llc_shared_mask(cpu));
+	if (shared_bank[bank]) {
 
-		/* first core not up yet */
-		if (cpu_data(i).cpu_core_id)
-			goto out;
+		nb = node_to_amd_nb(amd_get_nb_id(cpu));
+		WARN_ON(!nb);
 
-		/* already linked */
-		if (per_cpu(threshold_banks, cpu)[bank])
-			goto out;
+		/* threshold descriptor already initialized on this node? */
+		if (nb->bank4) {
+			/* yes, use it */
+			b = nb->bank4;
+			err = kobject_add(b->kobj, &dev->kobj, name);
+			if (err)
+				goto out;
 
-		b = per_cpu(threshold_banks, i)[bank];
+			per_cpu(threshold_banks, cpu)[bank] = b;
+			atomic_inc(&b->cpus);
 
-		if (!b)
-			goto out;
+			err = __threshold_add_blocks(b);
 
-		err = sysfs_create_link(&dev->kobj, b->kobj, name);
-		if (err)
 			goto out;
-
-		cpumask_copy(b->cpus, cpu_llc_shared_mask(cpu));
-		per_cpu(threshold_banks, cpu)[bank] = b;
-
-		goto out;
+		}
 	}
-#endif
 
 	b = kzalloc(sizeof(struct threshold_bank), GFP_KERNEL);
 	if (!b) {
 		err = -ENOMEM;
 		goto out;
 	}
-	if (!zalloc_cpumask_var(&b->cpus, GFP_KERNEL)) {
-		kfree(b);
-		err = -ENOMEM;
-		goto out;
-	}
 
 	b->kobj = kobject_create_and_add(name, &dev->kobj);
-	if (!b->kobj)
+	if (!b->kobj) {
+		err = -EINVAL;
 		goto out_free;
-
-#ifndef CONFIG_SMP
-	cpumask_setall(b->cpus);
-#else
-	cpumask_set_cpu(cpu, b->cpus);
-#endif
+	}
 
 	per_cpu(threshold_banks, cpu)[bank] = b;
 
-	err = local_allocate_threshold_blocks(cpu, bank);
-	if (err)
-		goto out_free;
-
-	for_each_cpu(i, b->cpus) {
-		if (i == cpu)
-			continue;
+	if (shared_bank[bank]) {
+		atomic_set(&b->cpus, 1);
 
-		dev = per_cpu(mce_device, i);
-		if (dev)
-			err = sysfs_create_link(&dev->kobj,b->kobj, name);
-		if (err)
-			goto out;
-
-		per_cpu(threshold_banks, i)[bank] = b;
+		/* nb is already initialized, see above */
+		WARN_ON(nb->bank4);
+		nb->bank4 = b;
 	}
 
-	goto out;
+	err = allocate_threshold_blocks(cpu, bank, 0,
+					MSR_IA32_MC0_MISC + bank * 4);
+	if (!err)
+		goto out;
 
-out_free:
-	per_cpu(threshold_banks, cpu)[bank] = NULL;
-	free_cpumask_var(b->cpus);
+ out_free:
 	kfree(b);
-out:
+
+ out:
 	return err;
 }
 
@@ -660,12 +648,6 @@ static __cpuinit int threshold_create_device(unsigned int cpu)
 	return err;
 }
 
-/*
- * let's be hotplug friendly.
- * in case of multiple core processors, the first core always takes ownership
- *   of shared sysfs dir/files, and rest of the cores will be symlinked to it.
- */
-
 static void deallocate_threshold_block(unsigned int cpu,
 						 unsigned int bank)
 {
@@ -686,41 +668,42 @@ static void deallocate_threshold_block(unsigned int cpu,
 	per_cpu(threshold_banks, cpu)[bank]->blocks = NULL;
 }
 
+static void __threshold_remove_blocks(struct threshold_bank *b)
+{
+	struct threshold_block *pos = NULL;
+	struct threshold_block *tmp = NULL;
+
+	kobject_del(b->kobj);
+
+	list_for_each_entry_safe(pos, tmp, &b->blocks->miscj, miscj)
+		kobject_del(&pos->kobj);
+}
+
 static void threshold_remove_bank(unsigned int cpu, int bank)
 {
+	struct amd_northbridge *nb;
 	struct threshold_bank *b;
-	struct device *dev;
-	char name[32];
-	int i = 0;
 
 	b = per_cpu(threshold_banks, cpu)[bank];
 	if (!b)
 		return;
+
 	if (!b->blocks)
 		goto free_out;
 
-	sprintf(name, "threshold_bank%i", bank);
-
-#ifdef CONFIG_SMP
-	/* sibling symlink */
-	if (shared_bank[bank] && b->blocks->cpu != cpu) {
-		dev = per_cpu(mce_device, cpu);
-		sysfs_remove_link(&dev->kobj, name);
-		per_cpu(threshold_banks, cpu)[bank] = NULL;
-
-		return;
-	}
-#endif
-
-	/* remove all sibling symlinks before unregistering */
-	for_each_cpu(i, b->cpus) {
-		if (i == cpu)
-			continue;
-
-		dev = per_cpu(mce_device, i);
-		if (dev)
-			sysfs_remove_link(&dev->kobj, name);
-		per_cpu(threshold_banks, i)[bank] = NULL;
+	if (shared_bank[bank]) {
+		if (!atomic_dec_and_test(&b->cpus)) {
+			__threshold_remove_blocks(b);
+			per_cpu(threshold_banks, cpu)[bank] = NULL;
+			return;
+		} else {
+			/*
+			 * the last CPU on this node using the shared bank is
+			 * going away, remove that bank now.
+			 */
+			nb = node_to_amd_nb(amd_get_nb_id(cpu));
+			nb->bank4 = NULL;
+		}
 	}
 
 	deallocate_threshold_block(cpu, bank);
@@ -728,7 +711,6 @@ static void threshold_remove_bank(unsigned int cpu, int bank)
 free_out:
 	kobject_del(b->kobj);
 	kobject_put(b->kobj);
-	free_cpumask_var(b->cpus);
 	kfree(b);
 	per_cpu(threshold_banks, cpu)[bank] = NULL;
 }
diff --git a/drivers/hwmon/k10temp.c b/drivers/hwmon/k10temp.c
index 7356b5ec8f67..f2fe8078633b 100644
--- a/drivers/hwmon/k10temp.c
+++ b/drivers/hwmon/k10temp.c
@@ -33,9 +33,6 @@ static bool force;
 module_param(force, bool, 0444);
 MODULE_PARM_DESC(force, "force loading on processors with erratum 319");
 
-/* PCI-IDs for Northbridge devices not used anywhere else */
-#define PCI_DEVICE_ID_AMD_15H_M10H_NB_F3	0x1403
-
 /* CPUID function 0x80000001, ebx */
 #define CPUID_PKGTYPE_MASK	0xf0000000
 #define CPUID_PKGTYPE_F		0x00000000
@@ -213,7 +210,7 @@ static DEFINE_PCI_DEVICE_TABLE(k10temp_id_table) = {
 	{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_11H_NB_MISC) },
 	{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_CNB17H_F3) },
 	{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_15H_NB_F3) },
-	{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_15H_M10H_NB_F3) },
+	{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_15H_M10H_F3) },
 	{}
 };
 MODULE_DEVICE_TABLE(pci, k10temp_id_table);
diff --git a/include/linux/mm.h b/include/linux/mm.h
index b36d08ce5c57..f9f279cf5b1b 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -1591,6 +1591,7 @@ void vmemmap_populate_print_last(void);
 enum mf_flags {
 	MF_COUNT_INCREASED = 1 << 0,
 	MF_ACTION_REQUIRED = 1 << 1,
+	MF_MUST_KILL = 1 << 2,
 };
 extern int memory_failure(unsigned long pfn, int trapno, int flags);
 extern void memory_failure_queue(unsigned long pfn, int trapno, int flags);
diff --git a/include/linux/pci_ids.h b/include/linux/pci_ids.h
index 5f187026b812..fc3526077348 100644
--- a/include/linux/pci_ids.h
+++ b/include/linux/pci_ids.h
@@ -517,6 +517,7 @@
 #define PCI_DEVICE_ID_AMD_11H_NB_DRAM	0x1302
 #define PCI_DEVICE_ID_AMD_11H_NB_MISC	0x1303
 #define PCI_DEVICE_ID_AMD_11H_NB_LINK	0x1304
+#define PCI_DEVICE_ID_AMD_15H_M10H_F3	0x1403
 #define PCI_DEVICE_ID_AMD_15H_NB_F0	0x1600
 #define PCI_DEVICE_ID_AMD_15H_NB_F1	0x1601
 #define PCI_DEVICE_ID_AMD_15H_NB_F2	0x1602
diff --git a/mm/memory-failure.c b/mm/memory-failure.c
index ab1e7145e290..de4ce7058450 100644
--- a/mm/memory-failure.c
+++ b/mm/memory-failure.c
@@ -345,14 +345,14 @@ static void add_to_kill(struct task_struct *tsk, struct page *p,
  * Also when FAIL is set do a force kill because something went
  * wrong earlier.
  */
-static void kill_procs(struct list_head *to_kill, int doit, int trapno,
+static void kill_procs(struct list_head *to_kill, int forcekill, int trapno,
 			  int fail, struct page *page, unsigned long pfn,
 			  int flags)
 {
 	struct to_kill *tk, *next;
 
 	list_for_each_entry_safe (tk, next, to_kill, nd) {
-		if (doit) {
+		if (forcekill) {
 			/*
 			 * In case something went wrong with munmapping
 			 * make sure the process doesn't catch the
@@ -858,7 +858,7 @@ static int hwpoison_user_mappings(struct page *p, unsigned long pfn,
 	struct address_space *mapping;
 	LIST_HEAD(tokill);
 	int ret;
-	int kill = 1;
+	int kill = 1, forcekill;
 	struct page *hpage = compound_head(p);
 	struct page *ppage;
 
@@ -888,7 +888,7 @@ static int hwpoison_user_mappings(struct page *p, unsigned long pfn,
 	 * be called inside page lock (it's recommended but not enforced).
 	 */
 	mapping = page_mapping(hpage);
-	if (!PageDirty(hpage) && mapping &&
+	if (!(flags & MF_MUST_KILL) && !PageDirty(hpage) && mapping &&
 	    mapping_cap_writeback_dirty(mapping)) {
 		if (page_mkclean(hpage)) {
 			SetPageDirty(hpage);
@@ -965,12 +965,14 @@ static int hwpoison_user_mappings(struct page *p, unsigned long pfn,
 	 * Now that the dirty bit has been propagated to the
 	 * struct page and all unmaps done we can decide if
 	 * killing is needed or not.  Only kill when the page
-	 * was dirty, otherwise the tokill list is merely
+	 * was dirty or the process is not restartable,
+	 * otherwise the tokill list is merely
 	 * freed.  When there was a problem unmapping earlier
 	 * use a more force-full uncatchable kill to prevent
 	 * any accesses to the poisoned memory.
 	 */
-	kill_procs(&tokill, !!PageDirty(ppage), trapno,
+	forcekill = PageDirty(ppage) || (flags & MF_MUST_KILL);
+	kill_procs(&tokill, forcekill, trapno,
 		      ret != SWAP_SUCCESS, p, pfn, flags);
 
 	return ret;