Merge branch 'x86/urgent' into x86-mm

Merge reason: Pick up the following two fix commits.

  2be19102b7: x86, NUMA: Fix empty memblk detection in numa_cleanup_meminfo()
  765af22da8: x86-32, NUMA: Fix ACPI NUMA init broken by recent x86-64 change

Scheduled NUMA init 32/64bit unification changes depend on these.

Signed-off-by: Tejun Heo <tj@kernel.org>

11 files changed, 126 insertions, 29 deletions

diff --git a/arch/x86/kernel/aperture_64.c b/arch/x86/kernel/aperture_64.c
index 86d1ad4962a7..73fb469908c6 100644
--- a/arch/x86/kernel/aperture_64.c
+++ b/arch/x86/kernel/aperture_64.c
@@ -499,7 +499,7 @@ out:
 		 * Don't enable translation yet but enable GART IO and CPU
 		 * accesses and set DISTLBWALKPRB since GART table memory is UC.
 		 */
-		u32 ctl = DISTLBWALKPRB | aper_order << 1;
+		u32 ctl = aper_order << 1;
 
 		bus = amd_nb_bus_dev_ranges[i].bus;
 		dev_base = amd_nb_bus_dev_ranges[i].dev_base;
diff --git a/arch/x86/kernel/apic/io_apic.c b/arch/x86/kernel/apic/io_apic.c
index 68df09bba92e..45fd33d1fd3a 100644
--- a/arch/x86/kernel/apic/io_apic.c
+++ b/arch/x86/kernel/apic/io_apic.c
@@ -128,8 +128,8 @@ static int __init parse_noapic(char *str)
 }
 early_param("noapic", parse_noapic);
 
-static int io_apic_setup_irq_pin_once(unsigned int irq, int node,
-				      struct io_apic_irq_attr *attr);
+static int io_apic_setup_irq_pin(unsigned int irq, int node,
+				 struct io_apic_irq_attr *attr);
 
 /* Will be called in mpparse/acpi/sfi codes for saving IRQ info */
 void mp_save_irq(struct mpc_intsrc *m)
@@ -3570,7 +3570,7 @@ int arch_setup_ht_irq(unsigned int irq, struct pci_dev *dev)
 }
 #endif /* CONFIG_HT_IRQ */
 
-int
+static int
 io_apic_setup_irq_pin(unsigned int irq, int node, struct io_apic_irq_attr *attr)
 {
 	struct irq_cfg *cfg = alloc_irq_and_cfg_at(irq, node);
@@ -3585,8 +3585,8 @@ io_apic_setup_irq_pin(unsigned int irq, int node, struct io_apic_irq_attr *attr)
 	return ret;
 }
 
-static int io_apic_setup_irq_pin_once(unsigned int irq, int node,
-				      struct io_apic_irq_attr *attr)
+int io_apic_setup_irq_pin_once(unsigned int irq, int node,
+			       struct io_apic_irq_attr *attr)
 {
 	unsigned int id = attr->ioapic, pin = attr->ioapic_pin;
 	int ret;
diff --git a/arch/x86/kernel/apm_32.c b/arch/x86/kernel/apm_32.c
index 0b4be431c620..adee12e0da1f 100644
--- a/arch/x86/kernel/apm_32.c
+++ b/arch/x86/kernel/apm_32.c
@@ -228,6 +228,7 @@
 #include <linux/kthread.h>
 #include <linux/jiffies.h>
 #include <linux/acpi.h>
+#include <linux/syscore_ops.h>
 
 #include <asm/system.h>
 #include <asm/uaccess.h>
@@ -1238,6 +1239,7 @@ static int suspend(int vetoable)
 
 	local_irq_disable();
 	sysdev_suspend(PMSG_SUSPEND);
+	syscore_suspend();
 
 	local_irq_enable();
 
@@ -1255,6 +1257,7 @@ static int suspend(int vetoable)
 		apm_error("suspend", err);
 	err = (err == APM_SUCCESS) ? 0 : -EIO;
 
+	syscore_resume();
 	sysdev_resume();
 	local_irq_enable();
 
@@ -1280,6 +1283,7 @@ static void standby(void)
 
 	local_irq_disable();
 	sysdev_suspend(PMSG_SUSPEND);
+	syscore_suspend();
 	local_irq_enable();
 
 	err = set_system_power_state(APM_STATE_STANDBY);
@@ -1287,6 +1291,7 @@ static void standby(void)
 		apm_error("standby", err);
 
 	local_irq_disable();
+	syscore_resume();
 	sysdev_resume();
 	local_irq_enable();
 
diff --git a/arch/x86/kernel/cpu/amd.c b/arch/x86/kernel/cpu/amd.c
index 3ecece0217ef..bb9eb29a52dd 100644
--- a/arch/x86/kernel/cpu/amd.c
+++ b/arch/x86/kernel/cpu/amd.c
@@ -615,6 +615,25 @@ static void __cpuinit init_amd(struct cpuinfo_x86 *c)
 	/* As a rule processors have APIC timer running in deep C states */
 	if (c->x86 >= 0xf && !cpu_has_amd_erratum(amd_erratum_400))
 		set_cpu_cap(c, X86_FEATURE_ARAT);
+
+	/*
+	 * Disable GART TLB Walk Errors on Fam10h. We do this here
+	 * because this is always needed when GART is enabled, even in a
+	 * kernel which has no MCE support built in.
+	 */
+	if (c->x86 == 0x10) {
+		/*
+		 * BIOS should disable GartTlbWlk Errors themself. If
+		 * it doesn't do it here as suggested by the BKDG.
+		 *
+		 * Fixes: https://bugzilla.kernel.org/show_bug.cgi?id=33012
+		 */
+		u64 mask;
+
+		rdmsrl(MSR_AMD64_MCx_MASK(4), mask);
+		mask |= (1 << 10);
+		wrmsrl(MSR_AMD64_MCx_MASK(4), mask);
+	}
 }
 
 #ifdef CONFIG_X86_32
@@ -679,7 +698,7 @@ cpu_dev_register(amd_cpu_dev);
  */
 
 const int amd_erratum_400[] =
-	AMD_OSVW_ERRATUM(1, AMD_MODEL_RANGE(0xf, 0x41, 0x2, 0xff, 0xf),
+	AMD_OSVW_ERRATUM(1, AMD_MODEL_RANGE(0x0f, 0x4, 0x2, 0xff, 0xf),
 			    AMD_MODEL_RANGE(0x10, 0x2, 0x1, 0xff, 0xf));
 EXPORT_SYMBOL_GPL(amd_erratum_400);
 
diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c
index eed3673a8656..e638689279d3 100644
--- a/arch/x86/kernel/cpu/perf_event.c
+++ b/arch/x86/kernel/cpu/perf_event.c
@@ -586,8 +586,12 @@ static int x86_setup_perfctr(struct perf_event *event)
 			return -EOPNOTSUPP;
 	}
 
+	/*
+	 * Do not allow config1 (extended registers) to propagate,
+	 * there's no sane user-space generalization yet:
+	 */
 	if (attr->type == PERF_TYPE_RAW)
-		return x86_pmu_extra_regs(event->attr.config, event);
+		return 0;
 
 	if (attr->type == PERF_TYPE_HW_CACHE)
 		return set_ext_hw_attr(hwc, event);
@@ -609,8 +613,8 @@ static int x86_setup_perfctr(struct perf_event *event)
 	/*
 	 * Branch tracing:
 	 */
-	if ((attr->config == PERF_COUNT_HW_BRANCH_INSTRUCTIONS) &&
-	    (hwc->sample_period == 1)) {
+	if (attr->config == PERF_COUNT_HW_BRANCH_INSTRUCTIONS &&
+	    !attr->freq && hwc->sample_period == 1) {
 		/* BTS is not supported by this architecture. */
 		if (!x86_pmu.bts_active)
 			return -EOPNOTSUPP;
@@ -1284,6 +1288,16 @@ static int x86_pmu_handle_irq(struct pt_regs *regs)
 
 	cpuc = &__get_cpu_var(cpu_hw_events);
 
+	/*
+	 * Some chipsets need to unmask the LVTPC in a particular spot
+	 * inside the nmi handler.  As a result, the unmasking was pushed
+	 * into all the nmi handlers.
+	 *
+	 * This generic handler doesn't seem to have any issues where the
+	 * unmasking occurs so it was left at the top.
+	 */
+	apic_write(APIC_LVTPC, APIC_DM_NMI);
+
 	for (idx = 0; idx < x86_pmu.num_counters; idx++) {
 		if (!test_bit(idx, cpuc->active_mask)) {
 			/*
@@ -1370,8 +1384,6 @@ perf_event_nmi_handler(struct notifier_block *self,
 		return NOTIFY_DONE;
 	}
 
-	apic_write(APIC_LVTPC, APIC_DM_NMI);
-
 	handled = x86_pmu.handle_irq(args->regs);
 	if (!handled)
 		return NOTIFY_DONE;
diff --git a/arch/x86/kernel/cpu/perf_event_amd.c b/arch/x86/kernel/cpu/perf_event_amd.c
index 461f62bbd774..cf4e369cea67 100644
--- a/arch/x86/kernel/cpu/perf_event_amd.c
+++ b/arch/x86/kernel/cpu/perf_event_amd.c
@@ -8,7 +8,7 @@ static __initconst const u64 amd_hw_cache_event_ids
  [ C(L1D) ] = {
 	[ C(OP_READ) ] = {
 		[ C(RESULT_ACCESS) ] = 0x0040, /* Data Cache Accesses        */
-		[ C(RESULT_MISS)   ] = 0x0041, /* Data Cache Misses          */
+		[ C(RESULT_MISS)   ] = 0x0141, /* Data Cache Misses          */
 	},
 	[ C(OP_WRITE) ] = {
 		[ C(RESULT_ACCESS) ] = 0x0142, /* Data Cache Refills :system */
@@ -427,7 +427,9 @@ static __initconst const struct x86_pmu amd_pmu = {
  *
  * Exceptions:
  *
+ * 0x000	FP	PERF_CTL[3], PERF_CTL[5:3] (*)
  * 0x003	FP	PERF_CTL[3]
+ * 0x004	FP	PERF_CTL[3], PERF_CTL[5:3] (*)
  * 0x00B	FP	PERF_CTL[3]
  * 0x00D	FP	PERF_CTL[3]
  * 0x023	DE	PERF_CTL[2:0]
@@ -448,6 +450,8 @@ static __initconst const struct x86_pmu amd_pmu = {
  * 0x0DF	LS	PERF_CTL[5:0]
  * 0x1D6	EX	PERF_CTL[5:0]
  * 0x1D8	EX	PERF_CTL[5:0]
+ *
+ * (*) depending on the umask all FPU counters may be used
  */
 
 static struct event_constraint amd_f15_PMC0  = EVENT_CONSTRAINT(0, 0x01, 0);
@@ -460,18 +464,28 @@ static struct event_constraint amd_f15_PMC53 = EVENT_CONSTRAINT(0, 0x38, 0);
 static struct event_constraint *
 amd_get_event_constraints_f15h(struct cpu_hw_events *cpuc, struct perf_event *event)
 {
-	unsigned int event_code = amd_get_event_code(&event->hw);
+	struct hw_perf_event *hwc = &event->hw;
+	unsigned int event_code = amd_get_event_code(hwc);
 
 	switch (event_code & AMD_EVENT_TYPE_MASK) {
 	case AMD_EVENT_FP:
 		switch (event_code) {
+		case 0x000:
+			if (!(hwc->config & 0x0000F000ULL))
+				break;
+			if (!(hwc->config & 0x00000F00ULL))
+				break;
+			return &amd_f15_PMC3;
+		case 0x004:
+			if (hweight_long(hwc->config & ARCH_PERFMON_EVENTSEL_UMASK) <= 1)
+				break;
+			return &amd_f15_PMC3;
 		case 0x003:
 		case 0x00B:
 		case 0x00D:
 			return &amd_f15_PMC3;
-		default:
-			return &amd_f15_PMC53;
 		}
+		return &amd_f15_PMC53;
 	case AMD_EVENT_LS:
 	case AMD_EVENT_DC:
 	case AMD_EVENT_EX_LS:
diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c
index 8fc2b2cee1da..e61539b07d2c 100644
--- a/arch/x86/kernel/cpu/perf_event_intel.c
+++ b/arch/x86/kernel/cpu/perf_event_intel.c
@@ -25,7 +25,7 @@ struct intel_percore {
 /*
  * Intel PerfMon, used on Core and later.
  */
-static const u64 intel_perfmon_event_map[] =
+static u64 intel_perfmon_event_map[PERF_COUNT_HW_MAX] __read_mostly =
 {
   [PERF_COUNT_HW_CPU_CYCLES]		= 0x003c,
   [PERF_COUNT_HW_INSTRUCTIONS]		= 0x00c0,
@@ -391,12 +391,12 @@ static __initconst const u64 nehalem_hw_cache_event_ids
 {
  [ C(L1D) ] = {
 	[ C(OP_READ) ] = {
-		[ C(RESULT_ACCESS) ] = 0x0f40, /* L1D_CACHE_LD.MESI            */
-		[ C(RESULT_MISS)   ] = 0x0140, /* L1D_CACHE_LD.I_STATE         */
+		[ C(RESULT_ACCESS) ] = 0x010b, /* MEM_INST_RETIRED.LOADS       */
+		[ C(RESULT_MISS)   ] = 0x0151, /* L1D.REPL                     */
 	},
 	[ C(OP_WRITE) ] = {
-		[ C(RESULT_ACCESS) ] = 0x0f41, /* L1D_CACHE_ST.MESI            */
-		[ C(RESULT_MISS)   ] = 0x0141, /* L1D_CACHE_ST.I_STATE         */
+		[ C(RESULT_ACCESS) ] = 0x020b, /* MEM_INST_RETURED.STORES      */
+		[ C(RESULT_MISS)   ] = 0x0251, /* L1D.M_REPL                   */
 	},
 	[ C(OP_PREFETCH) ] = {
 		[ C(RESULT_ACCESS) ] = 0x014e, /* L1D_PREFETCH.REQUESTS        */
@@ -933,6 +933,16 @@ static int intel_pmu_handle_irq(struct pt_regs *regs)
 
 	cpuc = &__get_cpu_var(cpu_hw_events);
 
+	/*
+	 * Some chipsets need to unmask the LVTPC in a particular spot
+	 * inside the nmi handler.  As a result, the unmasking was pushed
+	 * into all the nmi handlers.
+	 *
+	 * This handler doesn't seem to have any issues with the unmasking
+	 * so it was left at the top.
+	 */
+	apic_write(APIC_LVTPC, APIC_DM_NMI);
+
 	intel_pmu_disable_all();
 	handled = intel_pmu_drain_bts_buffer();
 	status = intel_pmu_get_status();
@@ -998,6 +1008,9 @@ intel_bts_constraints(struct perf_event *event)
 	struct hw_perf_event *hwc = &event->hw;
 	unsigned int hw_event, bts_event;
 
+	if (event->attr.freq)
+		return NULL;
+
 	hw_event = hwc->config & INTEL_ARCH_EVENT_MASK;
 	bts_event = x86_pmu.event_map(PERF_COUNT_HW_BRANCH_INSTRUCTIONS);
 
@@ -1305,7 +1318,7 @@ static void intel_clovertown_quirks(void)
 	 * AJ106 could possibly be worked around by not allowing LBR
 	 *       usage from PEBS, including the fixup.
 	 * AJ68  could possibly be worked around by always programming
-	 * 	 a pebs_event_reset[0] value and coping with the lost events.
+	 *	 a pebs_event_reset[0] value and coping with the lost events.
 	 *
 	 * But taken together it might just make sense to not enable PEBS on
 	 * these chips.
@@ -1409,6 +1422,18 @@ static __init int intel_pmu_init(void)
 		x86_pmu.percore_constraints = intel_nehalem_percore_constraints;
 		x86_pmu.enable_all = intel_pmu_nhm_enable_all;
 		x86_pmu.extra_regs = intel_nehalem_extra_regs;
+
+		if (ebx & 0x40) {
+			/*
+			 * Erratum AAJ80 detected, we work it around by using
+			 * the BR_MISP_EXEC.ANY event. This will over-count
+			 * branch-misses, but it's still much better than the
+			 * architectural event which is often completely bogus:
+			 */
+			intel_perfmon_event_map[PERF_COUNT_HW_BRANCH_MISSES] = 0x7f89;
+
+			pr_cont("erratum AAJ80 worked around, ");
+		}
 		pr_cont("Nehalem events, ");
 		break;
 
@@ -1425,6 +1450,7 @@ static __init int intel_pmu_init(void)
 
 	case 37: /* 32 nm nehalem, "Clarkdale" */
 	case 44: /* 32 nm nehalem, "Gulftown" */
+	case 47: /* 32 nm Xeon E7 */
 		memcpy(hw_cache_event_ids, westmere_hw_cache_event_ids,
 		       sizeof(hw_cache_event_ids));
 		memcpy(hw_cache_extra_regs, nehalem_hw_cache_extra_regs,
diff --git a/arch/x86/kernel/cpu/perf_event_p4.c b/arch/x86/kernel/cpu/perf_event_p4.c
index c2520e178d32..e93fcd55fae1 100644
--- a/arch/x86/kernel/cpu/perf_event_p4.c
+++ b/arch/x86/kernel/cpu/perf_event_p4.c
@@ -947,14 +947,23 @@ static int p4_pmu_handle_irq(struct pt_regs *regs)
 		if (!x86_perf_event_set_period(event))
 			continue;
 		if (perf_event_overflow(event, 1, &data, regs))
-			p4_pmu_disable_event(event);
+			x86_pmu_stop(event, 0);
 	}
 
-	if (handled) {
-		/* p4 quirk: unmask it again */
-		apic_write(APIC_LVTPC, apic_read(APIC_LVTPC) & ~APIC_LVT_MASKED);
+	if (handled)
 		inc_irq_stat(apic_perf_irqs);
-	}
+
+	/*
+	 * When dealing with the unmasking of the LVTPC on P4 perf hw, it has
+	 * been observed that the OVF bit flag has to be cleared first _before_
+	 * the LVTPC can be unmasked.
+	 *
+	 * The reason is the NMI line will continue to be asserted while the OVF
+	 * bit is set.  This causes a second NMI to generate if the LVTPC is
+	 * unmasked before the OVF bit is cleared, leading to unknown NMI
+	 * messages.
+	 */
+	apic_write(APIC_LVTPC, APIC_DM_NMI);
 
 	return handled;
 }
diff --git a/arch/x86/kernel/devicetree.c b/arch/x86/kernel/devicetree.c
index 706a9fb46a58..e90f08458e6b 100644
--- a/arch/x86/kernel/devicetree.c
+++ b/arch/x86/kernel/devicetree.c
@@ -391,7 +391,7 @@ static int ioapic_xlate(struct irq_domain *id, const u32 *intspec, u32 intsize,
 
 	set_io_apic_irq_attr(&attr, idx, line, it->trigger, it->polarity);
 
-	return io_apic_setup_irq_pin(*out_hwirq, cpu_to_node(0), &attr);
+	return io_apic_setup_irq_pin_once(*out_hwirq, cpu_to_node(0), &attr);
 }
 
 static void __init ioapic_add_ofnode(struct device_node *np)
diff --git a/arch/x86/kernel/pci-gart_64.c b/arch/x86/kernel/pci-gart_64.c
index 82ada01625b9..b117efd24f71 100644
--- a/arch/x86/kernel/pci-gart_64.c
+++ b/arch/x86/kernel/pci-gart_64.c
@@ -81,6 +81,9 @@ static u32 gart_unmapped_entry;
 #define AGPEXTERN
 #endif
 
+/* GART can only remap to physical addresses < 1TB */
+#define GART_MAX_PHYS_ADDR	(1ULL << 40)
+
 /* backdoor interface to AGP driver */
 AGPEXTERN int agp_memory_reserved;
 AGPEXTERN __u32 *agp_gatt_table;
@@ -212,9 +215,13 @@ static dma_addr_t dma_map_area(struct device *dev, dma_addr_t phys_mem,
 				size_t size, int dir, unsigned long align_mask)
 {
 	unsigned long npages = iommu_num_pages(phys_mem, size, PAGE_SIZE);
-	unsigned long iommu_page = alloc_iommu(dev, npages, align_mask);
+	unsigned long iommu_page;
 	int i;
 
+	if (unlikely(phys_mem + size > GART_MAX_PHYS_ADDR))
+		return bad_dma_addr;
+
+	iommu_page = alloc_iommu(dev, npages, align_mask);
 	if (iommu_page == -1) {
 		if (!nonforced_iommu(dev, phys_mem, size))
 			return phys_mem;
diff --git a/arch/x86/kernel/setup.c b/arch/x86/kernel/setup.c
index 5a0484a95ad6..4be9b398470e 100644
--- a/arch/x86/kernel/setup.c
+++ b/arch/x86/kernel/setup.c
@@ -976,6 +976,11 @@ void __init setup_arch(char **cmdline_p)
 	paging_init();
 	x86_init.paging.pagetable_setup_done(swapper_pg_dir);
 
+	if (boot_cpu_data.cpuid_level >= 0) {
+		/* A CPU has %cr4 if and only if it has CPUID */
+		mmu_cr4_features = read_cr4();
+	}
+
 #ifdef CONFIG_X86_32
 	/* sync back kernel address range */
 	clone_pgd_range(initial_page_table + KERNEL_PGD_BOUNDARY,