[PATCH v18 07/16] x86/VPMU: Initialize PMU for PV(H) guests

[Boris Ostrovsky <boris.ostrovsky@oracle.com>]

Code for initializing/tearing down PMU for PV guests

Signed-off-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Acked-by: Kevin Tian <kevin.tian@intel.com>
Acked-by: Daniel De Graaf <dgdegra@tycho.nsa.gov>
---
 tools/flask/policy/policy/modules/xen/xen.te |   4 ++
 xen/arch/x86/domain.c                        |   2 +
 xen/arch/x86/hvm/hvm.c                       |   1 +
 xen/arch/x86/hvm/svm/svm.c                   |   4 +-
 xen/arch/x86/hvm/svm/vpmu.c                  |  44 ++++++++----
 xen/arch/x86/hvm/vmx/vmx.c                   |   4 +-
 xen/arch/x86/hvm/vmx/vpmu_core2.c            |  79 +++++++++++++++------
 xen/arch/x86/hvm/vpmu.c                      | 102 +++++++++++++++++++++++++--
 xen/common/event_channel.c                   |   1 +
 xen/include/asm-x86/hvm/vpmu.h               |   2 +
 xen/include/public/pmu.h                     |   2 +
 xen/include/public/xen.h                     |   1 +
 xen/include/xsm/dummy.h                      |   3 +
 xen/xsm/flask/hooks.c                        |   4 ++
 xen/xsm/flask/policy/access_vectors          |   2 +
 15 files changed, 212 insertions(+), 43 deletions(-)

diff --git a/tools/flask/policy/policy/modules/xen/xen.te b/tools/flask/policy/policy/modules/xen/xen.te
index 870ff81..73bbe7b 100644
--- a/tools/flask/policy/policy/modules/xen/xen.te
+++ b/tools/flask/policy/policy/modules/xen/xen.te
@@ -120,6 +120,10 @@ domain_comms(dom0_t, dom0_t)
 # Allow all domains to use (unprivileged parts of) the tmem hypercall
 allow domain_type xen_t:xen tmem_op;
 
+# Allow all domains to use PMU (but not to change its settings --- that's what
+# pmu_ctrl is for)
+allow domain_type xen_t:xen2 pmu_use;
+
 ###############################################################################
 #
 # Domain creation
diff --git a/xen/arch/x86/domain.c b/xen/arch/x86/domain.c
index b0e3c3d..838bbb7 100644
--- a/xen/arch/x86/domain.c
+++ b/xen/arch/x86/domain.c
@@ -437,6 +437,8 @@ int vcpu_initialise(struct vcpu *v)
         vmce_init_vcpu(v);
     }
 
+    spin_lock_init(&v->arch.vpmu.vpmu_lock);
+
     if ( has_hvm_container_domain(d) )
     {
         rc = hvm_vcpu_initialise(v);
diff --git a/xen/arch/x86/hvm/hvm.c b/xen/arch/x86/hvm/hvm.c
index b03ee4e..21f7c35 100644
--- a/xen/arch/x86/hvm/hvm.c
+++ b/xen/arch/x86/hvm/hvm.c
@@ -4873,6 +4873,7 @@ static hvm_hypercall_t *const pvh_hypercall64_table[NR_hypercalls] = {
     HYPERCALL(hvm_op),
     HYPERCALL(sysctl),
     HYPERCALL(domctl),
+    HYPERCALL(xenpmu_op),
     [ __HYPERVISOR_arch_1 ] = (hvm_hypercall_t *)paging_domctl_continuation
 };
 
diff --git a/xen/arch/x86/hvm/svm/svm.c b/xen/arch/x86/hvm/svm/svm.c
index a7655bd..59cca08 100644
--- a/xen/arch/x86/hvm/svm/svm.c
+++ b/xen/arch/x86/hvm/svm/svm.c
@@ -1166,7 +1166,9 @@ static int svm_vcpu_initialise(struct vcpu *v)
         return rc;
     }
 
-    vpmu_initialise(v);
+    /* PVH's VPMU is initialized via hypercall */
+    if ( is_hvm_vcpu(v) )
+        vpmu_initialise(v);
 
     svm_guest_osvw_init(v);
 
diff --git a/xen/arch/x86/hvm/svm/vpmu.c b/xen/arch/x86/hvm/svm/vpmu.c
index 03474a3..7eeefa2 100644
--- a/xen/arch/x86/hvm/svm/vpmu.c
+++ b/xen/arch/x86/hvm/svm/vpmu.c
@@ -379,17 +379,19 @@ static void amd_vpmu_destroy(struct vcpu *v)
 {
     struct vpmu_struct *vpmu = vcpu_vpmu(v);
 
-    if ( has_hvm_container_vcpu(v) && is_msr_bitmap_on(vpmu) )
-        amd_vpmu_unset_msr_bitmap(v);
+    if ( has_hvm_container_vcpu(v) )
+    {
+        if ( is_msr_bitmap_on(vpmu) )
+            amd_vpmu_unset_msr_bitmap(v);
 
-    xfree(vpmu->context);
-    vpmu_reset(vpmu, VPMU_CONTEXT_ALLOCATED);
+        if ( is_hvm_vcpu(v) )
+            xfree(vpmu->context);
 
-    if ( vpmu_is_set(vpmu, VPMU_RUNNING) )
-    {
-        vpmu_reset(vpmu, VPMU_RUNNING);
         release_pmu_ownship(PMU_OWNER_HVM);
     }
+
+    vpmu->context = NULL;
+    vpmu_clear(vpmu);
 }
 
 /* VPMU part of the 'q' keyhandler */
@@ -456,15 +458,19 @@ int svm_vpmu_initialise(struct vcpu *v)
     if ( !counters )
         return -EINVAL;
 
-    ctxt = xzalloc_bytes(sizeof(*ctxt) +
-                         2 * sizeof(uint64_t) * num_counters);
-    if ( !ctxt )
+    if ( is_hvm_vcpu(v) )
     {
-        printk(XENLOG_G_WARNING "Insufficient memory for PMU, "
-               " PMU feature is unavailable on domain %d vcpu %d.\n",
-               v->vcpu_id, v->domain->domain_id);
-        return -ENOMEM;
+        ctxt = xzalloc_bytes(sizeof(*ctxt) +
+                             2 * sizeof(uint64_t) * num_counters);
+        if ( !ctxt )
+        {
+            printk(XENLOG_G_WARNING "%pv: Insufficient memory for PMU, "
+                   " PMU feature is unavailable\n", v);
+            return -ENOMEM;
+        }
     }
+    else
+        ctxt = &v->arch.vpmu.xenpmu_data->pmu.c.amd;
 
     ctxt->counters = sizeof(*ctxt);
     ctxt->ctrls = ctxt->counters + sizeof(uint64_t) * num_counters;
@@ -503,6 +509,16 @@ int __init amd_vpmu_init(void)
         return -EINVAL;
     }
 
+    if ( sizeof(struct xen_pmu_data) +
+         2 * sizeof(uint64_t) * num_counters > PAGE_SIZE )
+    {
+        printk(XENLOG_WARNING
+               "VPMU: Register bank does not fit into VPMU shared page\n");
+        counters = ctrls = NULL;
+        num_counters = 0;
+        return -ENOSPC;
+    }
+
     return 0;
 }
 
diff --git a/xen/arch/x86/hvm/vmx/vmx.c b/xen/arch/x86/hvm/vmx/vmx.c
index 88b7821..65b77ed 100644
--- a/xen/arch/x86/hvm/vmx/vmx.c
+++ b/xen/arch/x86/hvm/vmx/vmx.c
@@ -116,7 +116,9 @@ static int vmx_vcpu_initialise(struct vcpu *v)
         return rc;
     }
 
-    vpmu_initialise(v);
+    /* PVH's VPMU is initialized via hypercall */
+    if ( is_hvm_vcpu(v) )
+        vpmu_initialise(v);
 
     vmx_install_vlapic_mapping(v);
 
diff --git a/xen/arch/x86/hvm/vmx/vpmu_core2.c b/xen/arch/x86/hvm/vmx/vpmu_core2.c
index 6c78323..60f0f69 100644
--- a/xen/arch/x86/hvm/vmx/vpmu_core2.c
+++ b/xen/arch/x86/hvm/vmx/vpmu_core2.c
@@ -378,24 +378,34 @@ static int core2_vpmu_alloc_resource(struct vcpu *v)
     struct xen_pmu_intel_ctxt *core2_vpmu_cxt = NULL;
     uint64_t *p = NULL;
 
-    if ( !acquire_pmu_ownership(PMU_OWNER_HVM) )
-        return 0;
-
-    wrmsrl(MSR_CORE_PERF_GLOBAL_CTRL, 0);
-    if ( vmx_add_host_load_msr(MSR_CORE_PERF_GLOBAL_CTRL) )
+    p = xzalloc(uint64_t);
+    if ( !p )
         goto out_err;
 
-    if ( vmx_add_guest_msr(MSR_CORE_PERF_GLOBAL_CTRL) )
-        goto out_err;
-    vmx_write_guest_msr(MSR_CORE_PERF_GLOBAL_CTRL, 0);
+    if ( has_hvm_container_vcpu(v) )
+    {
+        if ( is_hvm_vcpu(v) && !acquire_pmu_ownership(PMU_OWNER_HVM) )
+            goto out_err;
 
-    core2_vpmu_cxt = xzalloc_bytes(sizeof(*core2_vpmu_cxt) +
-                                   sizeof(uint64_t) * fixed_pmc_cnt +
-                                   sizeof(struct xen_pmu_cntr_pair) *
-                                   arch_pmc_cnt);
-    p = xzalloc(uint64_t);
-    if ( !core2_vpmu_cxt || !p )
-        goto out_err;
+        wrmsrl(MSR_CORE_PERF_GLOBAL_CTRL, 0);
+        if ( vmx_add_host_load_msr(MSR_CORE_PERF_GLOBAL_CTRL) )
+            goto out_err_hvm;
+        if ( vmx_add_guest_msr(MSR_CORE_PERF_GLOBAL_CTRL) )
+            goto out_err_hvm;
+        vmx_write_guest_msr(MSR_CORE_PERF_GLOBAL_CTRL, 0);
+    }
+
+    if ( is_hvm_vcpu(v) )
+    {
+        core2_vpmu_cxt = xzalloc_bytes(sizeof(*core2_vpmu_cxt) +
+                                       sizeof(uint64_t) * fixed_pmc_cnt +
+                                       sizeof(struct xen_pmu_cntr_pair) *
+                                       arch_pmc_cnt);
+        if ( !core2_vpmu_cxt )
+            goto out_err_hvm;
+    }
+    else
+        core2_vpmu_cxt = &v->arch.vpmu.xenpmu_data->pmu.c.intel;
 
     core2_vpmu_cxt->fixed_counters = sizeof(*core2_vpmu_cxt);
     core2_vpmu_cxt->arch_counters = core2_vpmu_cxt->fixed_counters +
@@ -408,10 +418,12 @@ static int core2_vpmu_alloc_resource(struct vcpu *v)
 
     return 1;
 
-out_err:
-    release_pmu_ownship(PMU_OWNER_HVM);
-
+ out_err_hvm:
     xfree(core2_vpmu_cxt);
+    if ( is_hvm_vcpu(v) )
+        release_pmu_ownship(PMU_OWNER_HVM);
+
+ out_err:
     xfree(p);
 
     printk("Failed to allocate VPMU resources for domain %u vcpu %u\n",
@@ -728,12 +740,20 @@ static void core2_vpmu_destroy(struct vcpu *v)
 {
     struct vpmu_struct *vpmu = vcpu_vpmu(v);
 
-    xfree(vpmu->context);
+    if ( has_hvm_container_vcpu(v) )
+    {
+        if ( cpu_has_vmx_msr_bitmap )
+            core2_vpmu_unset_msr_bitmap(v->arch.hvm_vmx.msr_bitmap);
+
+        if ( is_hvm_vcpu(v) )
+            xfree(vpmu->context);
+
+        release_pmu_ownship(PMU_OWNER_HVM);
+    }
+
     xfree(vpmu->priv_context);
-    if ( has_hvm_container_vcpu(v) && cpu_has_vmx_msr_bitmap )
-        core2_vpmu_unset_msr_bitmap(v->arch.hvm_vmx.msr_bitmap);
-    release_pmu_ownship(PMU_OWNER_HVM);
-    vpmu_reset(vpmu, VPMU_CONTEXT_ALLOCATED);
+    vpmu->context = NULL;
+    vpmu_clear(vpmu);
 }
 
 struct arch_vpmu_ops core2_vpmu_ops = {
@@ -844,6 +864,10 @@ int vmx_vpmu_initialise(struct vcpu *v)
     ds_warned = 1;
  func_out:
 
+    /* PV domains can allocate resources immediately */
+    if ( is_pv_vcpu(v) && !core2_vpmu_alloc_resource(v) )
+        return -EIO;
+
     vpmu->arch_vpmu_ops = &core2_vpmu_ops;
 
     return 0;
@@ -906,5 +930,14 @@ int __init core2_vpmu_init(void)
 
     check_pmc_quirk();
 
+    if ( sizeof(struct xen_pmu_data) + sizeof(uint64_t) * fixed_pmc_cnt +
+         sizeof(struct xen_pmu_cntr_pair) * arch_pmc_cnt > PAGE_SIZE )
+    {
+        printk(XENLOG_WARNING
+               "VPMU: Register bank does not fit into VPMU share page\n");
+        arch_pmc_cnt = fixed_pmc_cnt = 0;
+        return -ENOSPC;
+    }
+
     return 0;
 }
diff --git a/xen/arch/x86/hvm/vpmu.c b/xen/arch/x86/hvm/vpmu.c
index f62fa80..bdd66de 100644
--- a/xen/arch/x86/hvm/vpmu.c
+++ b/xen/arch/x86/hvm/vpmu.c
@@ -27,6 +27,7 @@
 #include <asm/types.h>
 #include <asm/msr.h>
 #include <asm/nmi.h>
+#include <asm/p2m.h>
 #include <asm/hvm/support.h>
 #include <asm/hvm/vmx/vmx.h>
 #include <asm/hvm/vmx/vmcs.h>
@@ -252,9 +253,6 @@ void vpmu_initialise(struct vcpu *v)
     uint8_t vendor = current_cpu_data.x86_vendor;
     int ret;
 
-    if ( is_pvh_vcpu(v) )
-        return;
-
     BUILD_BUG_ON(sizeof(struct xen_pmu_intel_ctxt) > XENPMU_CTXT_PAD_SZ);
     BUILD_BUG_ON(sizeof(struct xen_pmu_amd_ctxt) > XENPMU_CTXT_PAD_SZ);
 
@@ -262,6 +260,7 @@ void vpmu_initialise(struct vcpu *v)
         vpmu_destroy(v);
     vpmu_clear(vpmu);
     vpmu->context = NULL;
+    vpmu->hw_lapic_lvtpc = PMU_APIC_VECTOR | APIC_LVT_MASKED;
 
     switch ( vendor )
     {
@@ -313,7 +312,92 @@ void vpmu_destroy(struct vcpu *v)
                          vpmu_clear_last, v, 1);
 
     if ( vpmu->arch_vpmu_ops && vpmu->arch_vpmu_ops->arch_vpmu_destroy )
+    {
+        /* Unload VPMU first. This will stop counters */
+        on_selected_cpus(cpumask_of(vcpu_vpmu(v)->last_pcpu),
+                         vpmu_save_force, v, 1);
+
         vpmu->arch_vpmu_ops->arch_vpmu_destroy(v);
+    }
+}
+
+static int pvpmu_init(struct domain *d, xen_pmu_params_t *params)
+{
+    struct vcpu *v;
+    struct vpmu_struct *vpmu;
+    struct page_info *page;
+    uint64_t gfn = params->val;
+
+    if ( vpmu_mode == XENPMU_MODE_OFF )
+        return -EINVAL;
+
+    if ( (params->vcpu >= d->max_vcpus) || (d->vcpu == NULL) ||
+         (d->vcpu[params->vcpu] == NULL) )
+        return -EINVAL;
+
+    if ( v->arch.vpmu.xenpmu_data )
+        return -EINVAL;
+
+    page = get_page_from_gfn(d, gfn, NULL, P2M_ALLOC);
+    if ( !page )
+        return -EINVAL;
+
+    if ( !get_page_type(page, PGT_writable_page) )
+    {
+        put_page(page);
+        return -EINVAL;
+    }
+
+    v = d->vcpu[params->vcpu];
+    vpmu = vcpu_vpmu(v);
+    spin_lock(&vpmu->vpmu_lock);
+
+    v->arch.vpmu.xenpmu_data = __map_domain_page_global(page);
+    if ( !v->arch.vpmu.xenpmu_data )
+    {
+        put_page_and_type(page);
+        spin_unlock(&vpmu->vpmu_lock);
+        return -EINVAL;
+    }
+
+    vpmu_initialise(v);
+
+    spin_unlock(&vpmu->vpmu_lock);
+
+    return 0;
+}
+
+static void pvpmu_finish(struct domain *d, xen_pmu_params_t *params)
+{
+    struct vcpu *v;
+    struct vpmu_struct *vpmu;
+    uint64_t mfn;
+
+    if ( (params->vcpu >= d->max_vcpus) || (d->vcpu == NULL) ||
+         (d->vcpu[params->vcpu] == NULL) )
+        return;
+
+    v = d->vcpu[params->vcpu];
+    if ( v != current )
+        vcpu_pause(v);
+
+    vpmu = vcpu_vpmu(v);
+    spin_lock(&vpmu->vpmu_lock);
+
+    if ( v->arch.vpmu.xenpmu_data )
+    {
+        mfn = domain_page_map_to_mfn(v->arch.vpmu.xenpmu_data);
+        ASSERT(mfn != 0);
+        unmap_domain_page_global(v->arch.vpmu.xenpmu_data);
+        put_page_and_type(mfn_to_page(mfn));
+        v->arch.vpmu.xenpmu_data = NULL;
+    }
+    vpmu_destroy(v);
+
+    spin_unlock(&vpmu->vpmu_lock);
+
+    if ( v != current )
+        vcpu_unpause(v);
 }
 
 /* Dump some vpmu informations on console. Used in keyhandler dump_domains(). */
@@ -445,6 +529,8 @@ long do_xenpmu_op(unsigned int op, XEN_GUEST_HANDLE_PARAM(xen_pmu_params_t) arg)
     {
     case XENPMU_mode_set:
     case XENPMU_feature_set:
+    case XENPMU_init:
+    case XENPMU_finish:
         if ( copy_from_guest(&pmu_params, arg, 1) )
             return -EFAULT;
 
@@ -504,7 +590,7 @@ long do_xenpmu_op(unsigned int op, XEN_GUEST_HANDLE_PARAM(xen_pmu_params_t) arg)
 
         if ( copy_to_guest(arg, &pmu_params, 1) )
             return -EFAULT;
-        break;
+         break;
 
     case XENPMU_feature_set:
         if ( pmu_params.val & ~XENPMU_FEATURE_INTEL_BTS )
@@ -519,6 +605,14 @@ long do_xenpmu_op(unsigned int op, XEN_GUEST_HANDLE_PARAM(xen_pmu_params_t) arg)
             return -EFAULT;
         break;
 
+    case XENPMU_init:
+        ret = pvpmu_init(current->domain, &pmu_params);
+        break;
+
+    case XENPMU_finish:
+        pvpmu_finish(current->domain, &pmu_params);
+        break;
+
     default:
         ret = -EINVAL;
     }
diff --git a/xen/common/event_channel.c b/xen/common/event_channel.c
index fae242d..310f590 100644
--- a/xen/common/event_channel.c
+++ b/xen/common/event_channel.c
@@ -106,6 +106,7 @@ static int virq_is_global(uint32_t virq)
     case VIRQ_TIMER:
     case VIRQ_DEBUG:
     case VIRQ_XENOPROF:
+    case VIRQ_XENPMU:
         rc = 0;
         break;
     case VIRQ_ARCH_0 ... VIRQ_ARCH_7:
diff --git a/xen/include/asm-x86/hvm/vpmu.h b/xen/include/asm-x86/hvm/vpmu.h
index cf32f82..42a09f9 100644
--- a/xen/include/asm-x86/hvm/vpmu.h
+++ b/xen/include/asm-x86/hvm/vpmu.h
@@ -44,6 +44,8 @@ struct vpmu_struct {
     void *context;      /* May be shared with PV guest */
     void *priv_context; /* hypervisor-only */
     struct arch_vpmu_ops *arch_vpmu_ops;
+    struct xen_pmu_data *xenpmu_data;
+    spinlock_t vpmu_lock;
 };
 
 /* Arch specific operations shared by all vpmus */
diff --git a/xen/include/public/pmu.h b/xen/include/public/pmu.h
index 66cc494..afb4ca1 100644
--- a/xen/include/public/pmu.h
+++ b/xen/include/public/pmu.h
@@ -25,6 +25,8 @@
 #define XENPMU_mode_set        1
 #define XENPMU_feature_get     2
 #define XENPMU_feature_set     3
+#define XENPMU_init            4
+#define XENPMU_finish          5
 /* ` } */
 
 /* Parameters structure for HYPERVISOR_xenpmu_op call */
diff --git a/xen/include/public/xen.h b/xen/include/public/xen.h
index 0dd3c97..a6b26fe 100644
--- a/xen/include/public/xen.h
+++ b/xen/include/public/xen.h
@@ -161,6 +161,7 @@ DEFINE_XEN_GUEST_HANDLE(xen_ulong_t);
 #define VIRQ_MEM_EVENT  10 /* G. (DOM0) A memory event has occured           */
 #define VIRQ_XC_RESERVED 11 /* G. Reserved for XenClient                     */
 #define VIRQ_ENOMEM     12 /* G. (DOM0) Low on heap memory       */
+#define VIRQ_XENPMU     13 /* V.  PMC interrupt                              */
 
 /* Architecture-specific VIRQ definitions. */
 #define VIRQ_ARCH_0    16
diff --git a/xen/include/xsm/dummy.h b/xen/include/xsm/dummy.h
index c637454..ae47135 100644
--- a/xen/include/xsm/dummy.h
+++ b/xen/include/xsm/dummy.h
@@ -665,6 +665,9 @@ static XSM_INLINE int xsm_pmu_op (XSM_DEFAULT_ARG struct domain *d, int op)
     case XENPMU_feature_set:
     case XENPMU_feature_get:
         return xsm_default_action(XSM_PRIV, d, current->domain);
+    case XENPMU_init:
+    case XENPMU_finish: 
+        return xsm_default_action(XSM_HOOK, d, current->domain);
     default:
         return -EPERM;
     }
diff --git a/xen/xsm/flask/hooks.c b/xen/xsm/flask/hooks.c
index 982e879..5011cb9 100644
--- a/xen/xsm/flask/hooks.c
+++ b/xen/xsm/flask/hooks.c
@@ -1518,6 +1518,10 @@ static int flask_pmu_op (struct domain *d, unsigned int op)
     case XENPMU_feature_get:
         return avc_has_perm(dsid, SECINITSID_XEN, SECCLASS_XEN2,
                             XEN2__PMU_CTRL, NULL);
+    case XENPMU_init:
+    case XENPMU_finish:
+        return avc_has_perm(dsid, SECINITSID_XEN, SECCLASS_XEN2,
+                            XEN2__PMU_USE, NULL);
     default:
         return -EPERM;
     }
diff --git a/xen/xsm/flask/policy/access_vectors b/xen/xsm/flask/policy/access_vectors
index 626850d..ef5b867 100644
--- a/xen/xsm/flask/policy/access_vectors
+++ b/xen/xsm/flask/policy/access_vectors
@@ -88,6 +88,8 @@ class xen2
     get_symbol
 # PMU control
     pmu_ctrl
+# PMU use (domains, including unprivileged ones, will be using this operation)
+    pmu_use
 }
 
 # Classes domain and domain2 consist of operations that a domain performs on
-- 
1.8.1.4