sched: add hypercall option to override and restore vcpu affinity

Some hardware (e.g. Dell studio 1555 laptops) require SMIs to be
called on physical cpu 0 only. Linux drivers like dcdbas or i8k try
to achieve this by pinning the running thread to cpu 0, but in Dom0
this is not enough: the vcpu must be pinned to physical cpu 0 via
Xen, too.

Add a stable hypercall option SCHEDOP_pin_override to the sched_op
hypercall to achieve this. It is taking a physical cpu number as
parameter. If pinning is possible (the calling domain has the
privilege to make the call and the cpu is available in the domain's
cpupool) the calling vcpu is pinned to the specified cpu. The old
cpu affinity is saved. To undo the override pinning a negative cpu
value is specified. This will restore the original cpu affinity of
the vcpu.

Signed-off-by: Juergen Gross <jgross@suse.com>
Acked-by: Dario Faggioli <dario.faggioli@citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>

xen/common/compat/schedule.c

@@ -10,6 +10,10 @@
 
 #define do_sched_op compat_sched_op
 
+#define xen_sched_pin_override sched_pin_override
+CHECK_sched_pin_override;
+#undef xen_sched_pin_override
+
 #define xen_sched_shutdown sched_shutdown
 CHECK_sched_shutdown;
 #undef xen_sched_shutdown

xen/common/schedule.c

@@ -271,6 +271,12 @@ int sched_move_domain(struct domain *d, struct cpupool *c)
     struct scheduler *old_ops;
     void *old_domdata;
 
+    for_each_vcpu ( d, v )
+    {
+        if ( v->affinity_broken )
+            return -EBUSY;
+    }
+
     domdata = SCHED_OP(c->sched, alloc_domdata, d);
     if ( domdata == NULL )
         return -ENOMEM;
@@ -669,6 +675,14 @@ int cpu_disable_scheduler(unsigned int cpu)
             if ( cpumask_empty(&online_affinity) &&
                  cpumask_test_cpu(cpu, v->cpu_hard_affinity) )
             {
+                if ( v->affinity_broken )
+                {
+                    /* The vcpu is temporarily pinned, can't move it. */
+                    vcpu_schedule_unlock_irqrestore(lock, flags, v);
+                    ret = -EBUSY;
+                    break;
+                }
+
                 if (system_state == SYS_STATE_suspend)
                 {
                     cpumask_copy(v->cpu_hard_affinity_saved,
@@ -752,14 +766,22 @@ static int vcpu_set_affinity(
     struct vcpu *v, const cpumask_t *affinity, cpumask_t *which)
 {
     spinlock_t *lock;
+    int ret = 0;
 
     lock = vcpu_schedule_lock_irq(v);
 
-    cpumask_copy(which, affinity);
+    if ( v->affinity_broken )
+        ret = -EBUSY;
+    else
+    {
+        cpumask_copy(which, affinity);
 
-    /* Always ask the scheduler to re-evaluate placement
-     * when changing the affinity */
-    set_bit(_VPF_migrating, &v->pause_flags);
+        /*
+         * Always ask the scheduler to re-evaluate placement
+         * when changing the affinity.
+         */
+        set_bit(_VPF_migrating, &v->pause_flags);
+    }
 
     vcpu_schedule_unlock_irq(lock, v);
 
@@ -771,7 +793,7 @@ static int vcpu_set_affinity(
         vcpu_migrate(v);
     }
 
-    return 0;
+    return ret;
 }
 
 int vcpu_set_hard_affinity(struct vcpu *v, const cpumask_t *affinity)
@@ -982,6 +1004,50 @@ void watchdog_domain_destroy(struct domain *d)
         kill_timer(&d->watchdog_timer[i]);
 }
 
+int vcpu_pin_override(struct vcpu *v, int cpu)
+{
+    spinlock_t *lock;
+    int ret = -EINVAL;
+
+    lock = vcpu_schedule_lock_irq(v);
+
+    if ( cpu < 0 )
+    {
+        if ( v->affinity_broken )
+        {
+            cpumask_copy(v->cpu_hard_affinity, v->cpu_hard_affinity_saved);
+            v->affinity_broken = 0;
+            set_bit(_VPF_migrating, &v->pause_flags);
+            ret = 0;
+        }
+    }
+    else if ( cpu < nr_cpu_ids )
+    {
+        if ( v->affinity_broken )
+            ret = -EBUSY;
+        else if ( cpumask_test_cpu(cpu, VCPU2ONLINE(v)) )
+        {
+            cpumask_copy(v->cpu_hard_affinity_saved, v->cpu_hard_affinity);
+            v->affinity_broken = 1;
+            cpumask_copy(v->cpu_hard_affinity, cpumask_of(cpu));
+            set_bit(_VPF_migrating, &v->pause_flags);
+            ret = 0;
+        }
+    }
+
+    vcpu_schedule_unlock_irq(lock, v);
+
+    domain_update_node_affinity(v->domain);
+
+    if ( v->pause_flags & VPF_migrating )
+    {
+        vcpu_sleep_nosync(v);
+        vcpu_migrate(v);
+    }
+
+    return ret;
+}
+
 typedef long ret_t;
 
 #endif /* !COMPAT */
@@ -1091,6 +1157,23 @@ ret_t do_sched_op(int cmd, XEN_GUEST_HANDLE_PARAM(void) arg)
         break;
     }
 
+    case SCHEDOP_pin_override:
+    {
+        struct sched_pin_override sched_pin_override;
+
+        ret = -EPERM;
+        if ( !is_hardware_domain(current->domain) )
+            break;
+
+        ret = -EFAULT;
+        if ( copy_from_guest(&sched_pin_override, arg, 1) )
+            break;
+
+        ret = vcpu_pin_override(current, sched_pin_override.pcpu);
+
+        break;
+    }
+
     default:
         ret = -ENOSYS;
     }

xen/include/public/sched.h

@@ -118,6 +118,18 @@
  * With id != 0 and timeout != 0, poke watchdog timer and set new timeout.
  */
 #define SCHEDOP_watchdog    6
+
+/*
+ * Override the current vcpu affinity by pinning it to one physical cpu or
+ * undo this override restoring the previous affinity.
+ * @arg == pointer to sched_pin_override_t structure.
+ *
+ * A negative pcpu value will undo a previous pin override and restore the
+ * previous cpu affinity.
+ * This call is allowed for the hardware domain only and requires the cpu
+ * to be part of the domain's cpupool.
+ */
+#define SCHEDOP_pin_override 7
 /* ` } */
 
 struct sched_shutdown {
@@ -148,6 +160,12 @@ struct sched_watchdog {
 typedef struct sched_watchdog sched_watchdog_t;
 DEFINE_XEN_GUEST_HANDLE(sched_watchdog_t);
 
+struct sched_pin_override {
+    int32_t pcpu;
+};
+typedef struct sched_pin_override sched_pin_override_t;
+DEFINE_XEN_GUEST_HANDLE(sched_pin_override_t);
+
 /*
  * Reason codes for SCHEDOP_shutdown. These may be interpreted by control
  * software to determine the appropriate action. For the most part, Xen does

xen/include/xen/sched.h

@@ -811,6 +811,7 @@ int cpu_disable_scheduler(unsigned int cpu);
 int vcpu_set_hard_affinity(struct vcpu *v, const cpumask_t *affinity);
 int vcpu_set_soft_affinity(struct vcpu *v, const cpumask_t *affinity);
 void restore_vcpu_affinity(struct domain *d);
+int vcpu_pin_override(struct vcpu *v, int cpu);
 
 void vcpu_runstate_get(struct vcpu *v, struct vcpu_runstate_info *runstate);
 uint64_t get_cpu_idle_time(unsigned int cpu);

xen/include/xlat.lst

@@ -104,6 +104,7 @@
 ?	pmu_data			pmu.h
 ?	pmu_params			pmu.h
 !	sched_poll			sched.h
+?	sched_pin_override		sched.h
 ?	sched_remote_shutdown		sched.h
 ?	sched_shutdown			sched.h
 ?	tmem_oid			tmem.h