10 files changed, 398 insertions, 9 deletions

diff --git a/arch/x86/include/asm/xen/interface.h b/arch/x86/include/asm/xen/interface.h
index 3b88eeacdbda..62ca03ef5c65 100644
--- a/arch/x86/include/asm/xen/interface.h
+++ b/arch/x86/include/asm/xen/interface.h
@@ -250,6 +250,129 @@ struct vcpu_guest_context {
 #endif
 };
 DEFINE_GUEST_HANDLE_STRUCT(vcpu_guest_context);
+
+/* AMD PMU registers and structures */
+struct xen_pmu_amd_ctxt {
+        /*
+         * Offsets to counter and control MSRs (relative to xen_pmu_arch.c.amd).
+         * For PV(H) guests these fields are RO.
+         */
+        uint32_t counters;
+        uint32_t ctrls;
+
+        /* Counter MSRs */
+#if defined(__STDC_VERSION__) && __STDC_VERSION__ >= 199901L
+        uint64_t regs[];
+#elif defined(__GNUC__)
+        uint64_t regs[0];
+#endif
+};
+
+/* Intel PMU registers and structures */
+struct xen_pmu_cntr_pair {
+        uint64_t counter;
+        uint64_t control;
+};
+
+struct xen_pmu_intel_ctxt {
+        /*
+         * Offsets to fixed and architectural counter MSRs (relative to
+         * xen_pmu_arch.c.intel).
+         * For PV(H) guests these fields are RO.
+         */
+        uint32_t fixed_counters;
+        uint32_t arch_counters;
+
+        /* PMU registers */
+        uint64_t global_ctrl;
+        uint64_t global_ovf_ctrl;
+        uint64_t global_status;
+        uint64_t fixed_ctrl;
+        uint64_t ds_area;
+        uint64_t pebs_enable;
+        uint64_t debugctl;
+
+        /* Fixed and architectural counter MSRs */
+#if defined(__STDC_VERSION__) && __STDC_VERSION__ >= 199901L
+        uint64_t regs[];
+#elif defined(__GNUC__)
+        uint64_t regs[0];
+#endif
+};
+
+/* Sampled domain's registers */
+struct xen_pmu_regs {
+        uint64_t ip;
+        uint64_t sp;
+        uint64_t flags;
+        uint16_t cs;
+        uint16_t ss;
+        uint8_t cpl;
+        uint8_t pad[3];
+};
+
+/* PMU flags */
+#define PMU_CACHED         (1<<0) /* PMU MSRs are cached in the context */
+#define PMU_SAMPLE_USER    (1<<1) /* Sample is from user or kernel mode */
+#define PMU_SAMPLE_REAL    (1<<2) /* Sample is from realmode */
+#define PMU_SAMPLE_PV      (1<<3) /* Sample from a PV guest */
+
+/*
+ * Architecture-specific information describing state of the processor at
+ * the time of PMU interrupt.
+ * Fields of this structure marked as RW for guest should only be written by
+ * the guest when PMU_CACHED bit in pmu_flags is set (which is done by the
+ * hypervisor during PMU interrupt). Hypervisor will read updated data in
+ * XENPMU_flush hypercall and clear PMU_CACHED bit.
+ */
+struct xen_pmu_arch {
+        union {
+                /*
+                 * Processor's registers at the time of interrupt.
+                 * WO for hypervisor, RO for guests.
+                 */
+                struct xen_pmu_regs regs;
+                /*
+                 * Padding for adding new registers to xen_pmu_regs in
+                 * the future
+                 */
+#define XENPMU_REGS_PAD_SZ  64
+                uint8_t pad[XENPMU_REGS_PAD_SZ];
+        } r;
+
+        /* WO for hypervisor, RO for guest */
+        uint64_t pmu_flags;
+
+        /*
+         * APIC LVTPC register.
+         * RW for both hypervisor and guest.
+         * Only APIC_LVT_MASKED bit is loaded by the hypervisor into hardware
+         * during XENPMU_flush or XENPMU_lvtpc_set.
+         */
+        union {
+                uint32_t lapic_lvtpc;
+                uint64_t pad;
+        } l;
+
+        /*
+         * Vendor-specific PMU registers.
+         * RW for both hypervisor and guest (see exceptions above).
+         * Guest's updates to this field are verified and then loaded by the
+         * hypervisor into hardware during XENPMU_flush
+         */
+        union {
+                struct xen_pmu_amd_ctxt amd;
+                struct xen_pmu_intel_ctxt intel;
+
+                /*
+                 * Padding for contexts (fixed parts only, does not include
+                 * MSR banks that are specified by offsets)
+                 */
+#define XENPMU_CTXT_PAD_SZ  128
+                uint8_t pad[XENPMU_CTXT_PAD_SZ];
+        } c;
+};
+
 #endif  /* !__ASSEMBLY__ */
 
 /*
diff --git a/arch/x86/xen/Makefile b/arch/x86/xen/Makefile
index 4b6e29ac0968..e47e52787d32 100644
--- a/arch/x86/xen/Makefile
+++ b/arch/x86/xen/Makefile
@@ -13,7 +13,7 @@ CFLAGS_mmu.o			:= $(nostackp)
 obj-y           := enlighten.o setup.o multicalls.o mmu.o irq.o \
                         time.o xen-asm.o xen-asm_$(BITS).o \
                         grant-table.o suspend.o platform-pci-unplug.o \
-                        p2m.o apic.o
+                        p2m.o apic.o pmu.o
 
 obj-$(CONFIG_EVENT_TRACING) += trace.o
 
diff --git a/arch/x86/xen/apic.c b/arch/x86/xen/apic.c
index 70e060ad879a..d03ebfa89b9f 100644
--- a/arch/x86/xen/apic.c
+++ b/arch/x86/xen/apic.c
@@ -72,6 +72,9 @@ static u32 xen_apic_read(u32 reg)
 
 static void xen_apic_write(u32 reg, u32 val)
 {
+        if (reg == APIC_LVTPC)
+                return;
+
         /* Warn to see if there's any stray references */
         WARN(1,"register: %x, value: %x\n", reg, val);
 }
diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c
index 373dbc9810d1..19072f91a8e2 100644
--- a/arch/x86/xen/enlighten.c
+++ b/arch/x86/xen/enlighten.c
@@ -84,6 +84,7 @@
 #include "mmu.h"
 #include "smp.h"
 #include "multicalls.h"
+#include "pmu.h"
 
 EXPORT_SYMBOL_GPL(hypercall_page);
 
@@ -1082,6 +1083,11 @@ static int xen_write_msr_safe(unsigned int msr, unsigned low, unsigned high)
         return ret;
 }
 
+unsigned long long xen_read_pmc(int counter)
+{
+        return 0;
+}
+
 void xen_setup_shared_info(void)
 {
         if (!xen_feature(XENFEAT_auto_translated_physmap)) {
@@ -1216,7 +1222,7 @@ static const struct pv_cpu_ops xen_cpu_ops __initconst = {
         .write_msr = xen_write_msr_safe,
 
         .read_tsc = native_read_tsc,
-        .read_pmc = native_read_pmc,
+        .read_pmc = xen_read_pmc,
 
         .read_tscp = native_read_tscp,
 
@@ -1267,6 +1273,10 @@ static const struct pv_apic_ops xen_apic_ops __initconst = {
 static void xen_reboot(int reason)
 {