x86/S3: Fix cpu pool scheduling after suspend/resume

This review is another S3 scheduler problem with the system_state
variable introduced with the following changeset:
http://xenbits.xen.org/gitweb/?p=xen.git;a=commit;h=269f543ea750ed567d18f2e819e5d5ce58eda5c5

Specifically, the cpu_callback function that takes the CPU down during
suspend, and back up during resume. We were seeing situations where,
after S3, only CPU0 was in cpupool0. Guest performance suffered
greatly, since all vcpus were only on a single pcpu. Guests under high
CPU load showed the problem much more quickly than an idle guest.

Removing this if condition forces the CPUs to go through the expected
online/offline state, and be properly scheduled after S3.

This also includes a necessary partial change proposed earlier by
Tomasz Wroblewski here:
http://lists.xen.org/archives/html/xen-devel/2013-01/msg02206.html

It should also resolve the issues discussed in this thread:
http://lists.xen.org/archives/html/xen-devel/2012-11/msg01801.html

Signed-off-by: Ben Guthro <benjamin.guthro@citrix.com>
Acked-by: Juergen Gross <juergen.gross@ts.fujitsu.com>

1 files changed, 45 insertions, 11 deletions

diff --git a/xen/common/cpupool.c b/xen/common/cpupool.c
index b3d8f2b128..2aaa48610c 100644
--- a/xen/common/cpupool.c
+++ b/xen/common/cpupool.c
@@ -41,16 +41,28 @@ static struct cpupool *alloc_cpupool_struct(void)
 {
     struct cpupool *c = xzalloc(struct cpupool);
 
-    if ( c && zalloc_cpumask_var(&c->cpu_valid) )
-        return c;
-    xfree(c);
-    return NULL;
+    if ( !c || !zalloc_cpumask_var(&c->cpu_valid) )
+    {
+        xfree(c);
+        c = NULL;
+    }
+    else if ( !zalloc_cpumask_var(&c->cpu_suspended) )
+    {
+        free_cpumask_var(c->cpu_valid);
+        xfree(c);
+        c = NULL;
+    }
+
+    return c;
 }
 
 static void free_cpupool_struct(struct cpupool *c)
 {
     if ( c )
+    {
+        free_cpumask_var(c->cpu_suspended);
         free_cpumask_var(c->cpu_valid);
+    }
     xfree(c);
 }
 
@@ -417,14 +429,32 @@ void cpupool_rm_domain(struct domain *d)
 
 /*
  * called to add a new cpu to pool admin
- * we add a hotplugged cpu to the cpupool0 to be able to add it to dom0
+ * we add a hotplugged cpu to the cpupool0 to be able to add it to dom0,
+ * unless we are resuming from S3, in which case we put the cpu back
+ * in the cpupool it was in prior to suspend.
  */
 static void cpupool_cpu_add(unsigned int cpu)
 {
     spin_lock(&cpupool_lock);
     cpumask_clear_cpu(cpu, &cpupool_locked_cpus);
     cpumask_set_cpu(cpu, &cpupool_free_cpus);
-    cpupool_assign_cpu_locked(cpupool0, cpu);
+
+    if ( system_state == SYS_STATE_resume )
+    {
+        struct cpupool **c;
+
+        for_each_cpupool(c)
+        {
+            if ( cpumask_test_cpu(cpu, (*c)->cpu_suspended ) )
+            {
+                cpupool_assign_cpu_locked(*c, cpu);
+                cpumask_clear_cpu(cpu, (*c)->cpu_suspended);
+            }
+        }
+    }
+
+    if ( cpumask_test_cpu(cpu, &cpupool_free_cpus) )
+        cpupool_assign_cpu_locked(cpupool0, cpu);
     spin_unlock(&cpupool_lock);
 }
 
@@ -436,7 +466,7 @@ static void cpupool_cpu_add(unsigned int cpu)
 static int cpupool_cpu_remove(unsigned int cpu)
 {
     int ret = 0;
-	
+
     spin_lock(&cpupool_lock);
     if ( !cpumask_test_cpu(cpu, cpupool0->cpu_valid))
         ret = -EBUSY;
@@ -633,9 +663,14 @@ static int cpu_callback(
     unsigned int cpu = (unsigned long)hcpu;
     int rc = 0;
 
-    if ( (system_state == SYS_STATE_suspend) ||
-         (system_state == SYS_STATE_resume) )
-        goto out;
+    if ( system_state == SYS_STATE_suspend )
+    {
+        struct cpupool **c;
+
+        for_each_cpupool(c)
+            if ( cpumask_test_cpu(cpu, (*c)->cpu_valid ) )
+                cpumask_set_cpu(cpu, (*c)->cpu_suspended);
+    }
 
     switch ( action )
     {
@@ -650,7 +685,6 @@ static int cpu_callback(
         break;
     }
 
-out:
     return !rc ? NOTIFY_DONE : notifier_from_errno(rc);
 }