Memory event interface improvements

diff -r 15da109b0c7d -r 8d67f4d5bafa xen/common/memory.c
@@ -226,6 +226,8 @@ static void decrease_reservation(struct 
 
     for ( i = a->nr_done; i < a->nr_extents; i++ )
     {
+        int one_remove_succeeded = 0;
+
         if ( hypercall_preempt_check() )
         {
             a->preempted = 1;

@@ -255,8 +257,33 @@ static void decrease_reservation(struct 
             continue;
 
         for ( j = 0; j < (1 << a->extent_order); j++ )
-            if ( !guest_remove_page(a->domain, gmfn + j) )
+        {
+            int rv = guest_remove_page(a->domain, gmfn + j);
+            /* negative rv is a result of a mem event ring full
+             * in the presence of paging. We preempt the hypercall */
+            if ( rv < 0 )
+            {
+                /* Indicate we're done with this extent */
+                if ((j+1) == (1 << a->extent_order))
+                    i++;
+                a->preempted = 1;
+                raise_softirq(SCHEDULE_SOFTIRQ);
                 goto out;
+            }
+            /* rv of zero means we tried to remove a gfn with no backing
+             * mfn. It can be a bad argument, or, a continuation in the midst
+             * of an extent. Heuristically determine second case. */
+            if ( !rv )
+            {
+                /* Has to be first extent */
+                if ( i != a->nr_done )
+                    goto out;
+                /* No previous remove in this extent must have succeeded */
+                if ( one_remove_succeeded )
+                    goto out;
+            } else
+                one_remove_succeeded = 1;
+        }
     }
 
  out:

diff -r 8d67f4d5bafa -r 43dc614d543c xen/arch/x86/hvm/hvm.c
@@ -4108,7 +4108,6 @@ static int hvm_memory_event_traps(long p
     struct vcpu* v = current;
     struct domain *d = v->domain;
     mem_event_request_t req;
-    int rc;
 
     if ( !(p & HVMPME_MODE_MASK) ) 
         return 0;

@@ -4116,10 +4115,6 @@ static int hvm_memory_event_traps(long p
     if ( (p & HVMPME_onchangeonly) && (value == old) )
         return 1;
     
-    rc = mem_event_check_ring(d, &d->mem_access);
-    if ( rc )
-        return rc;
-    
     memset(&req, 0, sizeof(req));
     req.type = MEM_EVENT_TYPE_ACCESS;
     req.reason = reason;

@@ -4139,7 +4134,20 @@ static int hvm_memory_event_traps(long p
         req.gla_valid = 1;
     }
     
-    mem_event_put_request(d, &d->mem_access, &req);
+    if ( mem_event_put_request(d, &d->mem_access, &req) == -ENOSYS )
+    {
+        /* rc == -ENOSYS
+         * If there was no ring to send the event, then simply unpause the
+         * vcpu (if we had previously paused it). It will retry the 
+         * instruction (with the exception "handled") and go on */
+        if ( req.flags & MEM_EVENT_FLAG_VCPU_PAUSED ) 
+            vcpu_unpause(v);
+        /* rc == -EBUSY
+         * If the ring is full, the vcpu has been marked and paused,
+         *  and the exception has been communicated to the consumer.
+         * Once there is room in the ring, the vcpu will be woken up 
+         * and will retry. */
+    }
     
     return 1;
 }

280 lines not on this page · raw message

diff -r 8d67f4d5bafa -r 43dc614d543c xen/arch/x86/mm/mem_sharing.c
@@ -281,12 +281,19 @@ static struct page_info* mem_sharing_all
     vcpu_pause_nosync(v);
     req.flags |= MEM_EVENT_FLAG_VCPU_PAUSED;
 
-    if(mem_event_check_ring(d, &d->mem_share)) return page;
-
     req.gfn = gfn;
     req.p2mt = p2m_ram_shared;
     req.vcpu_id = v->vcpu_id;
-    mem_event_put_request(d, &d->mem_share, &req);
+
+    /* If there is no ring, and we're out of memory, then
+     * there is no way out. */
+    if ( mem_event_put_request(d, &d->mem_share, &req) == -ENOSYS )
+    {
+        gdprintk(XENLOG_ERR, 
+                 "Failed alloc on unshare path for %hu and no ring "
+                 "to upcall\n", d->domain_id);
+        domain_crash(d);
+    }
 
     return page;
 }

@@ -300,12 +307,16 @@ int mem_sharing_sharing_resume(struct do
 {
     mem_event_response_t rsp;
 
-    /* Get request off the ring */
-    mem_event_get_response(&d->mem_share, &rsp);
+    /* Get all requests off the ring */
+    while ( mem_event_get_response(d, &d->mem_share, &rsp) )
+    {
+        /* Unpause domain/vcpu */
+        if ( rsp.flags & MEM_EVENT_FLAG_VCPU_PAUSED )
+            vcpu_unpause(d->vcpu[rsp.vcpu_id]);
+    }
 
-    /* Unpause domain/vcpu */
-    if( rsp.flags & MEM_EVENT_FLAG_VCPU_PAUSED )
-        vcpu_unpause(d->vcpu[rsp.vcpu_id]);
+    /* Unpause any domains that were paused because the ring was full */
+    mem_event_unpause_vcpus(d, &d->mem_paging);
 
     return 0;
 }

189 lines not on this page · raw message

diff -r 8d67f4d5bafa -r 43dc614d543c xen/common/memory.c
@@ -165,10 +165,13 @@ int guest_remove_page(struct domain *d, 
     mfn = mfn_x(get_gfn(d, gmfn, &p2mt)); 
     if ( unlikely(p2m_is_paging(p2mt)) )
     {
+        int rc; 
         guest_physmap_remove_page(d, gmfn, mfn, 0);
-        p2m_mem_paging_drop_page(d, gmfn);
+        rc = p2m_mem_paging_drop_page(d, gmfn);
         put_gfn(d, gmfn);
-        return 1;
+        /* drop_page returns zero on success, we return 1 on success.
+         * drop_page returns negative on error, never returns 1. */
+        return rc ? rc : 1;
     }
 #else
     mfn = gmfn_to_mfn(d, gmfn);

diff -r 8d67f4d5bafa -r 43dc614d543c xen/include/asm-x86/mem_event.h
@@ -25,12 +25,18 @@
 #define __MEM_EVENT_H__
 
 /* Pauses VCPU while marking pause flag for mem event */
-void mem_event_mark_and_pause(struct vcpu *v);
+void mem_event_mark_and_pause(struct vcpu *v, struct mem_event_domain *med);
 int mem_event_check_ring(struct domain *d, struct mem_event_domain *med);
-void mem_event_put_req_producers(struct mem_event_domain *med);
-void mem_event_put_request(struct domain *d, struct mem_event_domain *med, mem_event_request_t *req);
-void mem_event_get_response(struct mem_event_domain *med, mem_event_response_t *rsp);
-void mem_event_unpause_vcpus(struct domain *d);
+void mem_event_unpause_vcpus(struct domain *d, struct mem_event_domain *med);
+
+/* Returns 0 on success, -ENOSYS if there is no ring, -EBUSY if there is 
+ * a ring or no space. FOr success or -EBUSY. the vCPU is left blocked to 
+ * ensure that the ring does not lose events.  In general, put_request should 
+ * not fail, as it attempts to ensure that each vCPU can put at least one req. */
+int mem_event_put_request(struct domain *d, struct mem_event_domain *med,
+                            mem_event_request_t *req);
+int mem_event_get_response(struct domain *d, struct mem_event_domain *med,
+                            mem_event_response_t *rsp);
 
 int mem_event_domctl(struct domain *d, xen_domctl_mem_event_op_t *mec,
                      XEN_GUEST_HANDLE(void) u_domctl);

diff -r 8d67f4d5bafa -r 43dc614d543c xen/include/asm-x86/p2m.h
@@ -475,7 +475,7 @@ int p2m_mem_paging_nominate(struct domai
 /* Evict a frame */
 int p2m_mem_paging_evict(struct domain *d, unsigned long gfn);
 /* Tell xenpaging to drop a paged out frame */
-void p2m_mem_paging_drop_page(struct domain *d, unsigned long gfn);
+int p2m_mem_paging_drop_page(struct domain *d, unsigned long gfn);
 /* Start populating a paged out frame */
 void p2m_mem_paging_populate(struct domain *d, unsigned long gfn);
 /* Prepare the p2m for paging a frame in */

@@ -483,8 +483,8 @@ int p2m_mem_paging_prep(struct domain *d
 /* Resume normal operation (in case a domain was paused) */
 void p2m_mem_paging_resume(struct domain *d);
 #else
-static inline void p2m_mem_paging_drop_page(struct domain *d, unsigned long gfn)
-{ }
+static inline int p2m_mem_paging_drop_page(struct domain *d, unsigned long gfn)
+{ return 0; }
 static inline void p2m_mem_paging_populate(struct domain *d, unsigned long gfn)
 { }
 #endif

diff -r 8d67f4d5bafa -r 43dc614d543c xen/include/xen/mm.h
@@ -318,6 +318,8 @@ page_list_splice(struct page_list_head *
 
 void scrub_one_page(struct page_info *);
 
+/* Returns 1 on success, 0 on error, negative if the ring
+ * for event propagation is full in the presence of paging */
 int guest_remove_page(struct domain *d, unsigned long gmfn);
 
 #define RAM_TYPE_CONVENTIONAL 0x00000001

diff -r 8d67f4d5bafa -r 43dc614d543c xen/include/xen/sched.h
@@ -183,13 +183,16 @@ struct mem_event_domain
 {
     /* ring lock */
     spinlock_t ring_lock;
-    unsigned int req_producers;
     /* shared page */
     mem_event_shared_page_t *shared_page;
     /* shared ring page */
     void *ring_page;
     /* front-end ring */
     mem_event_front_ring_t front_ring;
+    /* the number of vCPUs blocked */
+    unsigned int blocked;
+    /* The last vcpu woken up */
+    unsigned int last_vcpu_wake_up;
     /* event channel port (vcpu0 only) */
     int xen_port;
 };

diff -r 43dc614d543c -r 96ba0cb6e255 xen/arch/ia64/vmx/vmx_init.c
@@ -377,7 +377,7 @@ vmx_vcpu_initialise(struct vcpu *v)
 {
 	struct vmx_ioreq_page *iorp = &v->domain->arch.hvm_domain.ioreq;
 
-	int rc = alloc_unbound_xen_event_channel(v, 0);
+	int rc = alloc_unbound_xen_event_channel(v, 0, NULL);
 	if (rc < 0)
 		return rc;
 	v->arch.arch_vmx.xen_port = rc;

diff -r 43dc614d543c -r 96ba0cb6e255 xen/arch/x86/hvm/hvm.c
@@ -979,7 +979,7 @@ int hvm_vcpu_initialise(struct vcpu *v)
         goto fail3;
 
     /* Create ioreq event channel. */
-    rc = alloc_unbound_xen_event_channel(v, 0);
+    rc = alloc_unbound_xen_event_channel(v, 0, NULL);
     if ( rc < 0 )
         goto fail4;
 

@@ -989,7 +989,7 @@ int hvm_vcpu_initialise(struct vcpu *v)
     if ( v->vcpu_id == 0 )
     {
         /* Create bufioreq event channel. */
-        rc = alloc_unbound_xen_event_channel(v, 0);
+        rc = alloc_unbound_xen_event_channel(v, 0, NULL);
         if ( rc < 0 )
             goto fail2;
         v->domain->arch.hvm_domain.params[HVM_PARAM_BUFIOREQ_EVTCHN] = rc;

@@ -3551,7 +3551,8 @@ long do_hvm_op(unsigned long op, XEN_GUE
                 for_each_vcpu ( d, v )
                 {
                     int old_port, new_port;
-                    new_port = alloc_unbound_xen_event_channel(v, a.value);
+                    new_port = alloc_unbound_xen_event_channel(
+                        v, a.value, NULL);
                     if ( new_port < 0 )
                     {
                         rc = new_port;

diff -r 43dc614d543c -r 96ba0cb6e255 xen/arch/x86/mm/mem_event.c
@@ -96,7 +96,8 @@ static int mem_event_enable(struct domai
 
     /* Allocate event channel */
     rc = alloc_unbound_xen_event_channel(d->vcpu[0],
-                                         current->domain->domain_id);
+                                         current->domain->domain_id,
+                                         NULL);
     if ( rc < 0 )
         goto err;
 

141 lines not on this page · raw message

diff -r 43dc614d543c -r 96ba0cb6e255 xen/include/xen/event.h
@@ -51,8 +51,11 @@ int evtchn_unmask(unsigned int port);
 void evtchn_move_pirqs(struct vcpu *v);
 
 /* Allocate/free a Xen-attached event channel port. */
+typedef void (*xen_event_channel_notification_t)(
+    struct vcpu *v, unsigned int port);
 int alloc_unbound_xen_event_channel(
-    struct vcpu *local_vcpu, domid_t remote_domid);
+    struct vcpu *local_vcpu, domid_t remote_domid,
+    xen_event_channel_notification_t notification_fn);
 void free_xen_event_channel(
     struct vcpu *local_vcpu, int port);
 

diff -r 43dc614d543c -r 96ba0cb6e255 xen/include/xen/sched.h
@@ -47,7 +47,7 @@ struct evtchn
 #define ECS_VIRQ         5 /* Channel is bound to a virtual IRQ line.        */
 #define ECS_IPI          6 /* Channel is bound to a virtual IPI line.        */
     u8  state;             /* ECS_* */
-    u8  consumer_is_xen;   /* Consumed by Xen or by guest? */
+    u8  xen_consumer;      /* Consumer in Xen, if any? (0 = send to guest) */
     u16 notify_vcpu_id;    /* VCPU for local delivery notification */
     union {
         struct {

diff -r 96ba0cb6e255 -r c2b5e65331ee xen/arch/x86/mm/mem_access.c
@@ -29,13 +29,12 @@ int mem_access_domctl(struct domain *d, 
                       XEN_GUEST_HANDLE(void) u_domctl)
 {
     int rc;
-    struct p2m_domain *p2m = p2m_get_hostp2m(d);
 
     switch( mec->op )
     {
     case XEN_DOMCTL_MEM_EVENT_OP_ACCESS_RESUME:
     {
-        p2m_mem_access_resume(p2m);
+        p2m_mem_access_resume(d);
         rc = 0;
     }
     break;

diff -r 96ba0cb6e255 -r c2b5e65331ee xen/arch/x86/mm/p2m.c
@@ -1194,9 +1194,8 @@ void p2m_mem_access_check(unsigned long 
     /* VCPU paused */
 }
 
-void p2m_mem_access_resume(struct p2m_domain *p2m)
+void p2m_mem_access_resume(struct domain *d)
 {
-    struct domain *d = p2m->domain;
     mem_event_response_t rsp;
 
     /* Pull all responses off the ring */

diff -r 96ba0cb6e255 -r c2b5e65331ee xen/include/asm-x86/p2m.h
@@ -495,7 +495,7 @@ static inline void p2m_mem_paging_popula
 void p2m_mem_access_check(unsigned long gpa, bool_t gla_valid, unsigned long gla, 
                           bool_t access_r, bool_t access_w, bool_t access_x);
 /* Resumes the running of the VCPU, restarting the last instruction */
-void p2m_mem_access_resume(struct p2m_domain *p2m);
+void p2m_mem_access_resume(struct domain *d);
 
 /* Set access type for a region of pfns.
  * If start_pfn == -1ul, sets the default access type */

diff -r c2b5e65331ee -r 29701f5bdd84 xen/arch/x86/mm/mem_event.c
@@ -37,9 +37,11 @@
 #define mem_event_ring_lock(_med)       spin_lock(&(_med)->ring_lock)
 #define mem_event_ring_unlock(_med)     spin_unlock(&(_med)->ring_lock)
 
-static int mem_event_enable(struct domain *d,
-                            xen_domctl_mem_event_op_t *mec,
-                            struct mem_event_domain *med)
+static int mem_event_enable(
+    struct domain *d,
+    xen_domctl_mem_event_op_t *mec,
+    struct mem_event_domain *med,
+    xen_event_channel_notification_t notification_fn)
 {
     int rc;
     struct domain *dom_mem_event = current->domain;

@@ -97,7 +99,7 @@ static int mem_event_enable(struct domai
     /* Allocate event channel */
     rc = alloc_unbound_xen_event_channel(d->vcpu[0],
                                          current->domain->domain_id,
-                                         NULL);
+                                         notification_fn);
     if ( rc < 0 )
         goto err;
 

@@ -330,6 +332,18 @@ int mem_event_check_ring(struct domain *
     return 0;
 }
 
+/* Registered with Xen-bound event channel for incoming notifications. */
+static void mem_paging_notification(struct vcpu *v, unsigned int port)
+{
+    p2m_mem_paging_resume(v->domain);
+}
+
+/* Registered with Xen-bound event channel for incoming notifications. */
+static void mem_access_notification(struct vcpu *v, unsigned int port)
+{
+    p2m_mem_access_resume(v->domain);
+}
+
 int mem_event_domctl(struct domain *d, xen_domctl_mem_event_op_t *mec,
                      XEN_GUEST_HANDLE(void) u_domctl)
 {

@@ -391,7 +405,7 @@ int mem_event_domctl(struct domain *d, x
             if ( p2m->pod.entry_count )
                 break;
 
-            rc = mem_event_enable(d, mec, med);
+            rc = mem_event_enable(d, mec, med, mem_paging_notification);
         }
         break;
 

@@ -430,7 +444,7 @@ int mem_event_domctl(struct domain *d, x
             if ( boot_cpu_data.x86_vendor != X86_VENDOR_INTEL )
                 break;
 
-            rc = mem_event_enable(d, mec, med);
+            rc = mem_event_enable(d, mec, med, mem_access_notification);
         }
         break;