Merge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm

Pull KVM updates from Paolo Bonzini:
 "Small release overall.

  x86:
   - miscellaneous fixes
   - AVIC support (local APIC virtualization, AMD version)

  s390:
   - polling for interrupts after a VCPU goes to halted state is now
     enabled for s390
   - use hardware provided information about facility bits that do not
     need any hypervisor activity, and other fixes for cpu models and
     facilities
   - improve perf output
   - floating interrupt controller improvements.

  MIPS:
   - miscellaneous fixes

  PPC:
   - bugfixes only

  ARM:
   - 16K page size support
   - generic firmware probing layer for timer and GIC

  Christoffer Dall (KVM-ARM maintainer) says:
    "There are a few changes in this pull request touching things
     outside KVM, but they should all carry the necessary acks and it
     made the merge process much easier to do it this way."

  though actually the irqchip maintainers' acks didn't make it into the
  patches.  Marc Zyngier, who is both irqchip and KVM-ARM maintainer,
  later acked at http://mid.gmane.org/573351D1.4060303@arm.com ('more
  formally and for documentation purposes')"

* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm: (82 commits)
  KVM: MTRR: remove MSR 0x2f8
  KVM: x86: make hwapic_isr_update and hwapic_irr_update look the same
  svm: Manage vcpu load/unload when enable AVIC
  svm: Do not intercept CR8 when enable AVIC
  svm: Do not expose x2APIC when enable AVIC
  KVM: x86: Introducing kvm_x86_ops.apicv_post_state_restore
  svm: Add VMEXIT handlers for AVIC
  svm: Add interrupt injection via AVIC
  KVM: x86: Detect and Initialize AVIC support
  svm: Introduce new AVIC VMCB registers
  KVM: split kvm_vcpu_wake_up from kvm_vcpu_kick
  KVM: x86: Introducing kvm_x86_ops VCPU blocking/unblocking hooks
  KVM: x86: Introducing kvm_x86_ops VM init/destroy hooks
  KVM: x86: Rename kvm_apic_get_reg to kvm_lapic_get_reg
  KVM: x86: Misc LAPIC changes to expose helper functions
  KVM: shrink halt polling even more for invalid wakeups
  KVM: s390: set halt polling to 80 microseconds
  KVM: halt_polling: provide a way to qualify wakeups during poll
  KVM: PPC: Book3S HV: Re-enable XICS fast path for irqfd-generated interrupts
  kvm: Conditionally register IRQ bypass consumer
  ...

4 files changed, 246 insertions, 34 deletions

diff --git a/drivers/irqchip/irq-gic-common.c b/drivers/irqchip/irq-gic-common.c
index 97c0028e8388..89e7423f0ebb 100644
--- a/drivers/irqchip/irq-gic-common.c
+++ b/drivers/irqchip/irq-gic-common.c
@@ -21,6 +21,19 @@
 
 #include "irq-gic-common.h"
 
+static const struct gic_kvm_info *gic_kvm_info;
+
+const struct gic_kvm_info *gic_get_kvm_info(void)
+{
+	return gic_kvm_info;
+}
+
+void gic_set_kvm_info(const struct gic_kvm_info *info)
+{
+	BUG_ON(gic_kvm_info != NULL);
+	gic_kvm_info = info;
+}
+
 void gic_enable_quirks(u32 iidr, const struct gic_quirk *quirks,
 		void *data)
 {
diff --git a/drivers/irqchip/irq-gic-common.h b/drivers/irqchip/irq-gic-common.h
index fff697db8e22..205e5fddf6da 100644
--- a/drivers/irqchip/irq-gic-common.h
+++ b/drivers/irqchip/irq-gic-common.h
@@ -19,6 +19,7 @@
 
 #include <linux/of.h>
 #include <linux/irqdomain.h>
+#include <linux/irqchip/arm-gic-common.h>
 
 struct gic_quirk {
 	const char *desc;
@@ -35,4 +36,6 @@ void gic_cpu_config(void __iomem *base, void (*sync_access)(void));
 void gic_enable_quirks(u32 iidr, const struct gic_quirk *quirks,
 		void *data);
 
+void gic_set_kvm_info(const struct gic_kvm_info *info);
+
 #endif /* _IRQ_GIC_COMMON_H */
diff --git a/drivers/irqchip/irq-gic-v3.c b/drivers/irqchip/irq-gic-v3.c
index 1a1ea4f733c1..fb042ba9a3db 100644
--- a/drivers/irqchip/irq-gic-v3.c
+++ b/drivers/irqchip/irq-gic-v3.c
@@ -15,6 +15,8 @@
  * along with this program.  If not, see <http://www.gnu.org/licenses/>.
  */
 
+#define pr_fmt(fmt)	"GICv3: " fmt
+
 #include <linux/acpi.h>
 #include <linux/cpu.h>
 #include <linux/cpu_pm.h>
@@ -28,6 +30,7 @@
 #include <linux/slab.h>
 
 #include <linux/irqchip.h>
+#include <linux/irqchip/arm-gic-common.h>
 #include <linux/irqchip/arm-gic-v3.h>
 #include <linux/irqchip/irq-partition-percpu.h>
 
@@ -59,6 +62,8 @@ struct gic_chip_data {
 static struct gic_chip_data gic_data __read_mostly;
 static struct static_key supports_deactivate = STATIC_KEY_INIT_TRUE;
 
+static struct gic_kvm_info gic_v3_kvm_info;
+
 #define gic_data_rdist()		(this_cpu_ptr(gic_data.rdists.rdist))
 #define gic_data_rdist_rd_base()	(gic_data_rdist()->rd_base)
 #define gic_data_rdist_sgi_base()	(gic_data_rdist_rd_base() + SZ_64K)
@@ -1002,7 +1007,7 @@ static int get_cpu_number(struct device_node *dn)
 }
 
 /* Create all possible partitions at boot time */
-static void gic_populate_ppi_partitions(struct device_node *gic_node)
+static void __init gic_populate_ppi_partitions(struct device_node *gic_node)
 {
 	struct device_node *parts_node, *child_part;
 	int part_idx = 0, i;
@@ -1089,6 +1094,30 @@ static void gic_populate_ppi_partitions(struct device_node *gic_node)
 	}
 }
 
+static void __init gic_of_setup_kvm_info(struct device_node *node)
+{
+	int ret;
+	struct resource r;
+	u32 gicv_idx;
+
+	gic_v3_kvm_info.type = GIC_V3;
+
+	gic_v3_kvm_info.maint_irq = irq_of_parse_and_map(node, 0);
+	if (!gic_v3_kvm_info.maint_irq)
+		return;
+
+	if (of_property_read_u32(node, "#redistributor-regions",
+				 &gicv_idx))
+		gicv_idx = 1;
+
+	gicv_idx += 3;	/* Also skip GICD, GICC, GICH */
+	ret = of_address_to_resource(node, gicv_idx, &r);
+	if (!ret)
+		gic_v3_kvm_info.vcpu = r;
+
+	gic_set_kvm_info(&gic_v3_kvm_info);
+}
+
 static int __init gic_of_init(struct device_node *node, struct device_node *parent)
 {
 	void __iomem *dist_base;
@@ -1144,6 +1173,7 @@ static int __init gic_of_init(struct device_node *node, struct device_node *pare
 		goto out_unmap_rdist;
 
 	gic_populate_ppi_partitions(node);
+	gic_of_setup_kvm_info(node);
 	return 0;
 
 out_unmap_rdist:
@@ -1159,19 +1189,25 @@ out_unmap_dist:
 IRQCHIP_DECLARE(gic_v3, "arm,gic-v3", gic_of_init);
 
 #ifdef CONFIG_ACPI
-static void __iomem *dist_base;
-static struct redist_region *redist_regs __initdata;
-static u32 nr_redist_regions __initdata;
-static bool single_redist;
+static struct
+{
+	void __iomem *dist_base;
+	struct redist_region *redist_regs;
+	u32 nr_redist_regions;
+	bool single_redist;
+	u32 maint_irq;
+	int maint_irq_mode;
+	phys_addr_t vcpu_base;
+} acpi_data __initdata;
 
 static void __init
 gic_acpi_register_redist(phys_addr_t phys_base, void __iomem *redist_base)
 {
 	static int count = 0;
 
-	redist_regs[count].phys_base = phys_base;
-	redist_regs[count].redist_base = redist_base;
-	redist_regs[count].single_redist = single_redist;
+	acpi_data.redist_regs[count].phys_base = phys_base;
+	acpi_data.redist_regs[count].redist_base = redist_base;
+	acpi_data.redist_regs[count].single_redist = acpi_data.single_redist;
 	count++;
 }
 
@@ -1199,7 +1235,7 @@ gic_acpi_parse_madt_gicc(struct acpi_subtable_header *header,
 {
 	struct acpi_madt_generic_interrupt *gicc =
 				(struct acpi_madt_generic_interrupt *)header;
-	u32 reg = readl_relaxed(dist_base + GICD_PIDR2) & GIC_PIDR2_ARCH_MASK;
+	u32 reg = readl_relaxed(acpi_data.dist_base + GICD_PIDR2) & GIC_PIDR2_ARCH_MASK;
 	u32 size = reg == GIC_PIDR2_ARCH_GICv4 ? SZ_64K * 4 : SZ_64K * 2;
 	void __iomem *redist_base;
 
@@ -1216,7 +1252,7 @@ static int __init gic_acpi_collect_gicr_base(void)
 	acpi_tbl_entry_handler redist_parser;
 	enum acpi_madt_type type;
 
-	if (single_redist) {
+	if (acpi_data.single_redist) {
 		type = ACPI_MADT_TYPE_GENERIC_INTERRUPT;
 		redist_parser = gic_acpi_parse_madt_gicc;
 	} else {
@@ -1267,14 +1303,14 @@ static int __init gic_acpi_count_gicr_regions(void)
 	count = acpi_table_parse_madt(ACPI_MADT_TYPE_GENERIC_REDISTRIBUTOR,
 				      gic_acpi_match_gicr, 0);
 	if (count > 0) {
-		single_redist = false;
+		acpi_data.single_redist = false;
 		return count;
 	}
 
 	count = acpi_table_parse_madt(ACPI_MADT_TYPE_GENERIC_INTERRUPT,
 				      gic_acpi_match_gicc, 0);
 	if (count > 0)
-		single_redist = true;
+		acpi_data.single_redist = true;
 
 	return count;
 }
@@ -1294,36 +1330,117 @@ static bool __init acpi_validate_gic_table(struct acpi_subtable_header *header,
 	if (count <= 0)
 		return false;
 
-	nr_redist_regions = count;
+	acpi_data.nr_redist_regions = count;
 	return true;
 }
 
+static int __init gic_acpi_parse_virt_madt_gicc(struct acpi_subtable_header *header,
+						const unsigned long end)
+{
+	struct acpi_madt_generic_interrupt *gicc =
+		(struct acpi_madt_generic_interrupt *)header;
+	int maint_irq_mode;
+	static int first_madt = true;
+
+	/* Skip unusable CPUs */
+	if (!(gicc->flags & ACPI_MADT_ENABLED))
+		return 0;
+
+	maint_irq_mode = (gicc->flags & ACPI_MADT_VGIC_IRQ_MODE) ?
+		ACPI_EDGE_SENSITIVE : ACPI_LEVEL_SENSITIVE;
+
+	if (first_madt) {
+		first_madt = false;
+
+		acpi_data.maint_irq = gicc->vgic_interrupt;
+		acpi_data.maint_irq_mode = maint_irq_mode;
+		acpi_data.vcpu_base = gicc->gicv_base_address;
+
+		return 0;
+	}
+
+	/*
+	 * The maintenance interrupt and GICV should be the same for every CPU
+	 */
+	if ((acpi_data.maint_irq != gicc->vgic_interrupt) ||
+	    (acpi_data.maint_irq_mode != maint_irq_mode) ||
+	    (acpi_data.vcpu_base != gicc->gicv_base_address))
+		return -EINVAL;
+
+	return 0;
+}
+
+static bool __init gic_acpi_collect_virt_info(void)
+{
+	int count;
+
+	count = acpi_table_parse_madt(ACPI_MADT_TYPE_GENERIC_INTERRUPT,
+				      gic_acpi_parse_virt_madt_gicc, 0);
+
+	return (count > 0);
+}
+
 #define ACPI_GICV3_DIST_MEM_SIZE (SZ_64K)
+#define ACPI_GICV2_VCTRL_MEM_SIZE	(SZ_4K)
+#define ACPI_GICV2_VCPU_MEM_SIZE	(SZ_8K)
+
+static void __init gic_acpi_setup_kvm_info(void)
+{
+	int irq;
+
+	if (!gic_acpi_collect_virt_info()) {
+		pr_warn("Unable to get hardware information used for virtualization\n");
+		return;
+	}
+
+	gic_v3_kvm_info.type = GIC_V3;
+
+	irq = acpi_register_gsi(NULL, acpi_data.maint_irq,
+				acpi_data.maint_irq_mode,
+				ACPI_ACTIVE_HIGH);
+	if (irq <= 0)
+		return;
+
+	gic_v3_kvm_info.maint_irq = irq;
+
+	if (acpi_data.vcpu_base) {
+		struct resource *vcpu = &gic_v3_kvm_info.vcpu;
+
+		vcpu->flags = IORESOURCE_MEM;
+		vcpu->start = acpi_data.vcpu_base;
+		vcpu->end = vcpu->start + ACPI_GICV2_VCPU_MEM_SIZE - 1;
+	}
+
+	gic_set_kvm_info(&gic_v3_kvm_info);
+}
 
 static int __init
 gic_acpi_init(struct acpi_subtable_header *header, const unsigned long end)
 {
 	struct acpi_madt_generic_distributor *dist;
 	struct fwnode_handle *domain_handle;
+	size_t size;
 	int i, err;
 
 	/* Get distributor base address */
 	dist = (struct acpi_madt_generic_distributor *)header;
-	dist_base = ioremap(dist->base_address, ACPI_GICV3_DIST_MEM_SIZE);
-	if (!dist_base) {
+	acpi_data.dist_base = ioremap(dist->base_address,
+				      ACPI_GICV3_DIST_MEM_SIZE);
+	if (!acpi_data.dist_base) {
 		pr_err("Unable to map GICD registers\n");
 		return -ENOMEM;
 	}
 
-	err = gic_validate_dist_version(dist_base);
+	err = gic_validate_dist_version(acpi_data.dist_base);
 	if (err) {
-		pr_err("No distributor detected at @%p, giving up", dist_base);
+		pr_err("No distributor detected at @%p, giving up",
+		       acpi_data.dist_base);
 		goto out_dist_unmap;
 	}
 
-	redist_regs = kzalloc(sizeof(*redist_regs) * nr_redist_regions,
-			      GFP_KERNEL);
-	if (!redist_regs) {
+	size = sizeof(*acpi_data.redist_regs) * acpi_data.nr_redist_regions;
+	acpi_data.redist_regs = kzalloc(size, GFP_KERNEL);
+	if (!acpi_data.redist_regs) {
 		err = -ENOMEM;
 		goto out_dist_unmap;
 	}
@@ -1332,29 +1449,31 @@ gic_acpi_init(struct acpi_subtable_header *header, const unsigned long end)
 	if (err)
 		goto out_redist_unmap;
 
-	domain_handle = irq_domain_alloc_fwnode(dist_base);
+	domain_handle = irq_domain_alloc_fwnode(acpi_data.dist_base);
 	if (!domain_handle) {
 		err = -ENOMEM;
 		goto out_redist_unmap;
 	}
 
-	err = gic_init_bases(dist_base, redist_regs, nr_redist_regions, 0,
-			     domain_handle);
+	err = gic_init_bases(acpi_data.dist_base, acpi_data.redist_regs,
+			     acpi_data.nr_redist_regions, 0, domain_handle);
 	if (err)
 		goto out_fwhandle_free;
 
 	acpi_set_irq_model(ACPI_IRQ_MODEL_GIC, domain_handle);
+	gic_acpi_setup_kvm_info();
+
 	return 0;
 
 out_fwhandle_free:
 	irq_domain_free_fwnode(domain_handle);
 out_redist_unmap:
-	for (i = 0; i < nr_redist_regions; i++)
-		if (redist_regs[i].redist_base)
-			iounmap(redist_regs[i].redist_base);
-	kfree(redist_regs);
+	for (i = 0; i < acpi_data.nr_redist_regions; i++)
+		if (acpi_data.redist_regs[i].redist_base)
+			iounmap(acpi_data.redist_regs[i].redist_base);
+	kfree(acpi_data.redist_regs);
 out_dist_unmap:
-	iounmap(dist_base);
+	iounmap(acpi_data.dist_base);
 	return err;
 }
 IRQCHIP_ACPI_DECLARE(gic_v3, ACPI_MADT_TYPE_GENERIC_DISTRIBUTOR,
diff --git a/drivers/irqchip/irq-gic.c b/drivers/irqchip/irq-gic.c
index 1de20e14a721..b4e647179346 100644
--- a/drivers/irqchip/irq-gic.c
+++ b/drivers/irqchip/irq-gic.c
@@ -105,6 +105,8 @@ static struct static_key supports_deactivate = STATIC_KEY_INIT_TRUE;
 
 static struct gic_chip_data gic_data[CONFIG_ARM_GIC_MAX_NR] __read_mostly;
 
+static struct gic_kvm_info gic_v2_kvm_info;
+
 #ifdef CONFIG_GIC_NON_BANKED
 static void __iomem *gic_get_percpu_base(union gic_base *base)
 {
@@ -1248,7 +1250,7 @@ static bool gic_check_eoimode(struct device_node *node, void __iomem **base)
 	return true;
 }
 
-static int gic_of_setup(struct gic_chip_data *gic, struct device_node *node)
+static int __init gic_of_setup(struct gic_chip_data *gic, struct device_node *node)
 {
 	if (!gic || !node)
 		return -EINVAL;
@@ -1272,6 +1274,29 @@ error:
 	return -ENOMEM;
 }
 
+static void __init gic_of_setup_kvm_info(struct device_node *node)
+{
+	int ret;
+	struct resource *vctrl_res = &gic_v2_kvm_info.vctrl;
+	struct resource *vcpu_res = &gic_v2_kvm_info.vcpu;
+
+	gic_v2_kvm_info.type = GIC_V2;
+
+	gic_v2_kvm_info.maint_irq = irq_of_parse_and_map(node, 0);
+	if (!gic_v2_kvm_info.maint_irq)
+		return;
+
+	ret = of_address_to_resource(node, 2, vctrl_res);
+	if (ret)
+		return;
+
+	ret = of_address_to_resource(node, 3, vcpu_res);
+	if (ret)
+		return;
+
+	gic_set_kvm_info(&gic_v2_kvm_info);
+}
+
 int __init
 gic_of_init(struct device_node *node, struct device_node *parent)
 {
@@ -1303,8 +1328,10 @@ gic_of_init(struct device_node *node, struct device_node *parent)
 		return ret;
 	}
 
-	if (!gic_cnt)
+	if (!gic_cnt) {
 		gic_init_physaddr(node);
+		gic_of_setup_kvm_info(node);
+	}
 
 	if (parent) {
 		irq = irq_of_parse_and_map(node, 0);
@@ -1330,7 +1357,14 @@ IRQCHIP_DECLARE(pl390, "arm,pl390", gic_of_init);
 #endif
 
 #ifdef CONFIG_ACPI
-static phys_addr_t cpu_phy_base __initdata;
+static struct
+{
+	phys_addr_t cpu_phys_base;
+	u32 maint_irq;
+	int maint_irq_mode;
+	phys_addr_t vctrl_base;
+	phys_addr_t vcpu_base;
+} acpi_data __initdata;
 
 static int __init
 gic_acpi_parse_madt_cpu(struct acpi_subtable_header *header,
@@ -1350,10 +1384,16 @@ gic_acpi_parse_madt_cpu(struct acpi_subtable_header *header,
 	 * All CPU interface addresses have to be the same.
 	 */
 	gic_cpu_base = processor->base_address;
-	if (cpu_base_assigned && gic_cpu_base != cpu_phy_base)
+	if (cpu_base_assigned && gic_cpu_base != acpi_data.cpu_phys_base)
 		return -EINVAL;
 
-	cpu_phy_base = gic_cpu_base;
+	acpi_data.cpu_phys_base = gic_cpu_base;
+	acpi_data.maint_irq = processor->vgic_interrupt;
+	acpi_data.maint_irq_mode = (processor->flags & ACPI_MADT_VGIC_IRQ_MODE) ?
+				    ACPI_EDGE_SENSITIVE : ACPI_LEVEL_SENSITIVE;
+	acpi_data.vctrl_base = processor->gich_base_address;
+	acpi_data.vcpu_base = processor->gicv_base_address;
+
 	cpu_base_assigned = 1;
 	return 0;
 }
@@ -1384,6 +1424,41 @@ static bool __init gic_validate_dist(struct acpi_subtable_header *header,
 
 #define ACPI_GICV2_DIST_MEM_SIZE	(SZ_4K)
 #define ACPI_GIC_CPU_IF_MEM_SIZE	(SZ_8K)
+#define ACPI_GICV2_VCTRL_MEM_SIZE	(SZ_4K)
+#define ACPI_GICV2_VCPU_MEM_SIZE	(SZ_8K)
+
+static void __init gic_acpi_setup_kvm_info(void)
+{
+	int irq;
+	struct resource *vctrl_res = &gic_v2_kvm_info.vctrl;
+	struct resource *vcpu_res = &gic_v2_kvm_info.vcpu;
+
+	gic_v2_kvm_info.type = GIC_V2;
+
+	if (!acpi_data.vctrl_base)
+		return;
+
+	vctrl_res->flags = IORESOURCE_MEM;
+	vctrl_res->start = acpi_data.vctrl_base;
+	vctrl_res->end = vctrl_res->start + ACPI_GICV2_VCTRL_MEM_SIZE - 1;
+
+	if (!acpi_data.vcpu_base)
+		return;
+
+	vcpu_res->flags = IORESOURCE_MEM;
+	vcpu_res->start = acpi_data.vcpu_base;
+	vcpu_res->end = vcpu_res->start + ACPI_GICV2_VCPU_MEM_SIZE - 1;
+
+	irq = acpi_register_gsi(NULL, acpi_data.maint_irq,
+				acpi_data.maint_irq_mode,
+				ACPI_ACTIVE_HIGH);
+	if (irq <= 0)
+		return;
+
+	gic_v2_kvm_info.maint_irq = irq;
+
+	gic_set_kvm_info(&gic_v2_kvm_info);
+}
 
 static int __init gic_v2_acpi_init(struct acpi_subtable_header *header,
 				   const unsigned long end)
@@ -1401,7 +1476,7 @@ static int __init gic_v2_acpi_init(struct acpi_subtable_header *header,
 		return -EINVAL;
 	}
 
-	gic->raw_cpu_base = ioremap(cpu_phy_base, ACPI_GIC_CPU_IF_MEM_SIZE);
+	gic->raw_cpu_base = ioremap(acpi_data.cpu_phys_base, ACPI_GIC_CPU_IF_MEM_SIZE);
 	if (!gic->raw_cpu_base) {
 		pr_err("Unable to map GICC registers\n");
 		return -ENOMEM;
@@ -1447,6 +1522,8 @@ static int __init gic_v2_acpi_init(struct acpi_subtable_header *header,
 	if (IS_ENABLED(CONFIG_ARM_GIC_V2M))
 		gicv2m_init(NULL, gic_data[0].domain);
 
+	gic_acpi_setup_kvm_info();
+
 	return 0;
 }
 IRQCHIP_ACPI_DECLARE(gic_v2, ACPI_MADT_TYPE_GENERIC_DISTRIBUTOR,