postcopy+vhost-user/shared ram

From: "Dr. David Alan Gilbert" <[email protected]>

Add a vhost feature flag for postcopy support, and use the postcopy notifier to check it before allowing postcopy.

Signed-off-by Dr. David Alan Gilbert <[email protected]>

Reviewed-by Peter Xu <[email protected]>

---
 contrib/libvhost-user/libvhost-user.h |  2 ++
 docs/interop/vhost-user.txt           | 10 +++++++++
 hw/virtio/vhost-user.c                | 41 ++++++++++++++++++++++++++++++++++-
 3 files changed, 52 insertions(+), 1 deletion(-)

contrib/libvhost-user/libvhost-user.h +2 -0

index 18f95f65d7..96db29c4ce 100644
@@ -48,6 +48,8 @@ enum VhostUserProtocolFeature {
     VHOST_USER_PROTOCOL_F_NET_MTU = 4,
     VHOST_USER_PROTOCOL_F_SLAVE_REQ = 5,
     VHOST_USER_PROTOCOL_F_CROSS_ENDIAN = 6,
+    VHOST_USER_PROTOCOL_F_CRYPTO_SESSION = 7,
+    VHOST_USER_PROTOCOL_F_PAGEFAULT = 8,
 
     VHOST_USER_PROTOCOL_F_MAX
 };

docs/interop/vhost-user.txt +10 -0

index cb3a7595aa..91a572d781 100644
@@ -290,6 +290,15 @@ Once the source has finished migration, rings will be stopped by
 the source. No further update must be done before rings are
 restarted.
 
+In postcopy migration the slave is started before all the memory has been
+received from the source host, and care must be taken to avoid accessing pages
+that have yet to be received.  The slave opens a 'userfault'-fd and registers
+the memory with it; this fd is then passed back over to the master.
+The master services requests on the userfaultfd for pages that are accessed
+and when the page is available it performs WAKE ioctl's on the userfaultfd
+to wake the stalled slave.  The client indicates support for this via the
+VHOST_USER_PROTOCOL_F_PAGEFAULT feature.
+
 Memory access
 -------------
 

@@ -369,6 +378,7 @@ Protocol features
 #define VHOST_USER_PROTOCOL_F_SLAVE_REQ      5
 #define VHOST_USER_PROTOCOL_F_CROSS_ENDIAN   6
 #define VHOST_USER_PROTOCOL_F_CRYPTO_SESSION 7
+#define VHOST_USER_PROTOCOL_F_PAGEFAULT      8
 
 Master message types
 --------------------

hw/virtio/vhost-user.c +40 -2

index 41ff5cff41..aab35c4845 100644
@@ -18,6 +18,8 @@
 #include "qemu/error-report.h"
 #include "qemu/sockets.h"
 #include "sysemu/cryptodev.h"
+#include "migration/migration.h"
+#include "migration/postcopy-ram.h"
 
 #include <sys/ioctl.h>
 #include <sys/socket.h>

@@ -41,7 +43,7 @@ enum VhostUserProtocolFeature {
     VHOST_USER_PROTOCOL_F_SLAVE_REQ = 5,
     VHOST_USER_PROTOCOL_F_CROSS_ENDIAN = 6,
     VHOST_USER_PROTOCOL_F_CRYPTO_SESSION = 7,
-
+    VHOST_USER_PROTOCOL_F_PAGEFAULT = 8,
     VHOST_USER_PROTOCOL_F_MAX
 };
 

@@ -164,8 +166,10 @@ static VhostUserMsg m __attribute__ ((unused));
 #define VHOST_USER_VERSION    (0x1)
 
 struct vhost_user {
+    struct vhost_dev *dev;
     CharBackend *chr;
     int slave_fd;
+    NotifierWithReturn postcopy_notifier;
 };
 
 static bool ioeventfd_enabled(void)

@@ -791,6 +795,33 @@ out:
     return ret;
 }
 
+static int vhost_user_postcopy_notifier(NotifierWithReturn *notifier,
+                                        void *opaque)
+{
+    struct PostcopyNotifyData *pnd = opaque;
+    struct vhost_user *u = container_of(notifier, struct vhost_user,
+                                         postcopy_notifier);
+    struct vhost_dev *dev = u->dev;
+
+    switch (pnd->reason) {
+    case POSTCOPY_NOTIFY_PROBE:
+        if (!virtio_has_feature(dev->protocol_features,
+                                VHOST_USER_PROTOCOL_F_PAGEFAULT)) {
+            /* TODO: Get the device name into this error somehow */
+            error_setg(pnd->errp,
+                       "vhost-user backend not capable of postcopy");
+            return -ENOENT;
+        }
+        break;
+
+    default:
+        /* We ignore notifications we don't know */
+        break;
+    }
+
+    return 0;
+}
+
 static int vhost_user_init(struct vhost_dev *dev, void *opaque)
 {
     uint64_t features, protocol_features;

@@ -802,6 +833,7 @@ static int vhost_user_init(struct vhost_dev *dev, void *opaque)
     u = g_new0(struct vhost_user, 1);
     u->chr = opaque;
     u->slave_fd = -1;
+    u->dev = dev;
     dev->opaque = u;
 
     err = vhost_user_get_features(dev, &features);

@@ -858,6 +890,9 @@ static int vhost_user_init(struct vhost_dev *dev, void *opaque)
         return err;
     }
 
+    u->postcopy_notifier.notify = vhost_user_postcopy_notifier;
+    postcopy_add_notifier(&u->postcopy_notifier);
+
     return 0;
 }
 

@@ -868,6 +903,10 @@ static int vhost_user_cleanup(struct vhost_dev *dev)
     assert(dev->vhost_ops->backend_type == VHOST_BACKEND_TYPE_USER);
 
     u = dev->opaque;
+    if (u->postcopy_notifier.notify) {
+        postcopy_remove_notifier(&u->postcopy_notifier);
+        u->postcopy_notifier.notify = NULL;
+    }
     if (u->slave_fd >= 0) {
         qemu_set_fd_handler(u->slave_fd, NULL, NULL, NULL);
         close(u->slave_fd);
--

2.14.3

From: "Dr. David Alan Gilbert" <[email protected]>

Wire up a notifier to send a VHOST_USER_POSTCOPY_ADVISE message on an incoming advise.

Later patches will fill in the behaviour/contents of the message.

Signed-off-by Dr. David Alan Gilbert <[email protected]>

Reviewed-by Marc-André Lureau <[email protected]>

---
 contrib/libvhost-user/libvhost-user.c | 14 ++++++++++
 contrib/libvhost-user/libvhost-user.h |  3 +++
 docs/interop/vhost-user.txt           | 10 ++++++++
 hw/virtio/vhost-user.c                | 48 +++++++++++++++++++++++++++++++++++
 migration/postcopy-ram.h              |  1 +
 migration/savevm.c                    |  6 +++++
 6 files changed, 82 insertions(+)

contrib/libvhost-user/libvhost-user.c +1 -0

index 2e358b5bce..71825d2dde 100644
@@ -86,6 +86,7 @@ vu_request_to_string(unsigned int req)
         REQ(VHOST_USER_SET_VRING_ENDIAN),
         REQ(VHOST_USER_GET_CONFIG),
         REQ(VHOST_USER_SET_CONFIG),
+        REQ(VHOST_USER_POSTCOPY_ADVISE),
         REQ(VHOST_USER_MAX),
     };
 #undef REQ

@@ -856,6 +857,17 @@ vu_set_config(VuDev *dev, VhostUserMsg *vmsg)
     return false;

contrib/libvhost-user/libvhost-user.c +5 -0

+    return true; /* = send a reply */
+}
+
 static bool
 vu_process_message(VuDev *dev, VhostUserMsg *vmsg)
 {

@@ -927,6 +939,8 @@ vu_process_message(VuDev *dev, VhostUserMsg *vmsg)
         return vu_set_config(dev, vmsg);
     case VHOST_USER_NONE:
         break;
+    case VHOST_USER_POSTCOPY_ADVISE:
+        return vu_set_postcopy_advise(dev, vmsg);
     default:
         vmsg_close_fds(vmsg);
         vu_panic(dev, "Unhandled request: %d", vmsg->request);

contrib/libvhost-user/libvhost-user.h +3 -0

index 96db29c4ce..00d78a8810 100644
@@ -83,6 +83,9 @@ typedef enum VhostUserRequest {
     VHOST_USER_SET_VRING_ENDIAN = 23,
     VHOST_USER_GET_CONFIG = 24,
     VHOST_USER_SET_CONFIG = 25,
+    VHOST_USER_CREATE_CRYPTO_SESSION = 26,
+    VHOST_USER_CLOSE_CRYPTO_SESSION = 27,
+    VHOST_USER_POSTCOPY_ADVISE  = 28,
     VHOST_USER_MAX
 } VhostUserRequest;
 

docs/interop/vhost-user.txt +0 -0

index 91a572d781..acc4186c79 100644

docs/interop/vhost-user.txt +6 -0

+
+      When VHOST_USER_PROTOCOL_F_PAGEFAULT is supported, the
+      master advises slave that a migration with postcopy enabled is underway,
+      the slave must open a userfaultfd for later use.
+      Note that at this stage the migration is still in precopy mode.
+
 Slave message types
 -------------------
 

hw/virtio/vhost-user.c +48 -0

index aab35c4845..431858b219 100644
@@ -78,6 +78,7 @@ typedef enum VhostUserRequest {
     VHOST_USER_SET_CONFIG = 25,
     VHOST_USER_CREATE_CRYPTO_SESSION = 26,
     VHOST_USER_CLOSE_CRYPTO_SESSION = 27,
+    VHOST_USER_POSTCOPY_ADVISE  = 28,
     VHOST_USER_MAX
 } VhostUserRequest;
 

@@ -795,6 +796,50 @@ out:
     return ret;
 }
 
+/*
+ * Called at the start of an inbound postcopy on reception of the
+ * 'advise' command.
+ */
+static int vhost_user_postcopy_advise(struct vhost_dev *dev, Error **errp)
+{
+    struct vhost_user *u = dev->opaque;
+    CharBackend *chr = u->chr;
+    int ufd;
+    VhostUserMsg msg = {
+        .hdr.request = VHOST_USER_POSTCOPY_ADVISE,
+        .hdr.flags = VHOST_USER_VERSION,
+    };
+
+    if (vhost_user_write(dev, &msg, NULL, 0) < 0) {
+        error_setg(errp, "Failed to send postcopy_advise to vhost");
+        return -1;
+    }
+
+    if (vhost_user_read(dev, &msg) < 0) {
+        error_setg(errp, "Failed to get postcopy_advise reply from vhost");
+        return -1;
+    }
+
+    if (msg.hdr.request != VHOST_USER_POSTCOPY_ADVISE) {
+        error_setg(errp, "Unexpected msg type. Expected %d received %d",
+                     VHOST_USER_POSTCOPY_ADVISE, msg.hdr.request);
+        return -1;
+    }
+
+    if (msg.hdr.size != sizeof(msg.payload.u64)) {
+        error_setg(errp, "Received bad msg size.");
+        return -1;
+    }
+    ufd = qemu_chr_fe_get_msgfd(chr);
+    if (ufd < 0) {
+        error_setg(errp, "%s: Failed to get ufd", __func__);
+        return -1;
+    }
+
+    /* TODO: register ufd with userfault thread */
+    return 0;
+}
+
 static int vhost_user_postcopy_notifier(NotifierWithReturn *notifier,
                                         void *opaque)
 {

@@ -814,6 +859,9 @@ static int vhost_user_postcopy_notifier(NotifierWithReturn *notifier,
         }
         break;
 
+    case POSTCOPY_NOTIFY_INBOUND_ADVISE:
+        return vhost_user_postcopy_advise(dev, pnd->errp);
+
     default:
         /* We ignore notifications we don't know */
         break;

migration/postcopy-ram.h +1 -0

index 2e879bbacb..0421c98d57 100644
@@ -130,6 +130,7 @@ void postcopy_infrastructure_init(void);
  */
 enum PostcopyNotifyReason {
     POSTCOPY_NOTIFY_PROBE = 0,
+    POSTCOPY_NOTIFY_INBOUND_ADVISE,
 };
 
 struct PostcopyNotifyData {

migration/savevm.c +6 -1

index 358c5b51e2..1f2bf12a28 100644
@@ -1386,6 +1386,7 @@ static int loadvm_postcopy_handle_advise(MigrationIncomingState *mis,
 {
     PostcopyState ps = postcopy_state_set(POSTCOPY_INCOMING_ADVISE);
     uint64_t remote_pagesize_summary, local_pagesize_summary, remote_tps;
+    Error *local_err = NULL;
 
     trace_loadvm_postcopy_handle_advise();
     if (ps != POSTCOPY_INCOMING_NONE) {

@@ -1451,6 +1452,11 @@ static int loadvm_postcopy_handle_advise(MigrationIncomingState *mis,
         return -1;
     }
 
+    if (postcopy_notify(POSTCOPY_NOTIFY_INBOUND_ADVISE, &local_err)) {
+        error_report_err(local_err);
+        return -1;
+    }
+
     if (ram_postcopy_incoming_init(mis)) {
         return -1;
     }
--

2.14.3

Hi

Quoted 35 lines · Dr. David Alan Gilbert (git)

From: "Dr. David Alan Gilbert" <[email protected]>

Wire up a notifier to send a VHOST_USER_POSTCOPY_ADVISE message on an incoming advise.

Later patches will fill in the behaviour/contents of the message.

Signed-off-by Dr. David Alan Gilbert <[email protected]>

Reviewed-by Marc-André Lureau <[email protected]>

---
 contrib/libvhost-user/libvhost-user.c | 14 ++++++++++
 contrib/libvhost-user/libvhost-user.h |  3 +++
 docs/interop/vhost-user.txt           | 10 ++++++++
 hw/virtio/vhost-user.c                | 48 +++++++++++++++++++++++++++++++++++
 migration/postcopy-ram.h              |  1 +
 migration/savevm.c                    |  6 +++++
 6 files changed, 82 insertions(+)

contrib/libvhost-user/libvhost-user.c +1 -0

index 2e358b5bce..71825d2dde 100644
@@ -86,6 +86,7 @@ vu_request_to_string(unsigned int req)
         REQ(VHOST_USER_SET_VRING_ENDIAN),
         REQ(VHOST_USER_GET_CONFIG),
         REQ(VHOST_USER_SET_CONFIG),
+        REQ(VHOST_USER_POSTCOPY_ADVISE),
         REQ(VHOST_USER_MAX),
     };
 #undef REQ

@@ -856,6 +857,17 @@ vu_set_config(VuDev *dev, VhostUserMsg *vmsg)
     return false;
 }
 

contrib/libvhost-user/libvhost-user.c vu_set_config(VuDev *dev, VhostUserMsg *vmsg) · line 867

Quoted 8 lines

contrib/libvhost-user/libvhost-user.c +8 -0

+static bool
+vu_set_postcopy_advise(VuDev *dev, VhostUserMsg *vmsg)
+{
+    /* TODO: Open ufd, pass it back in the request
+     * TODO: Add addresses
+     */
+    vmsg->payload.u64 = 0xcafe;
+    vmsg->size = sizeof(vmsg->payload.u64);

Left over? Why send back a u64 with 0xcafe value?

Quoted 34 lines

contrib/libvhost-user/libvhost-user.c +5 -0

+    return true; /* = send a reply */
+}
+
 static bool
 vu_process_message(VuDev *dev, VhostUserMsg *vmsg)
 {

@@ -927,6 +939,8 @@ vu_process_message(VuDev *dev, VhostUserMsg *vmsg)
         return vu_set_config(dev, vmsg);
     case VHOST_USER_NONE:
         break;
+    case VHOST_USER_POSTCOPY_ADVISE:
+        return vu_set_postcopy_advise(dev, vmsg);
     default:
         vmsg_close_fds(vmsg);
         vu_panic(dev, "Unhandled request: %d", vmsg->request);

contrib/libvhost-user/libvhost-user.h +3 -0

index 96db29c4ce..00d78a8810 100644
@@ -83,6 +83,9 @@ typedef enum VhostUserRequest {
     VHOST_USER_SET_VRING_ENDIAN = 23,
     VHOST_USER_GET_CONFIG = 24,
     VHOST_USER_SET_CONFIG = 25,
+    VHOST_USER_CREATE_CRYPTO_SESSION = 26,
+    VHOST_USER_CLOSE_CRYPTO_SESSION = 27,
+    VHOST_USER_POSTCOPY_ADVISE  = 28,
     VHOST_USER_MAX
 } VhostUserRequest;
 

docs/interop/vhost-user.txt +0 -0

index 91a572d781..acc4186c79 100644

docs/interop/vhost-user.txt Master message types · line 705

Quoted 7 lines

docs/interop/vhost-user.txt +4 -0

@@ -699,6 +699,16 @@ Master message types
      feature has been successfully negotiated.
      It's a required feature for crypto devices.
 
+ * VHOST_USER_POSTCOPY_ADVISE
+      Id: 28
+      Master payload: N/A
+      Slave payload: userfault fd + u64

If the reply is modified, update doc as well

Quoted 110 lines

docs/interop/vhost-user.txt +6 -0

+
+      When VHOST_USER_PROTOCOL_F_PAGEFAULT is supported, the
+      master advises slave that a migration with postcopy enabled is underway,
+      the slave must open a userfaultfd for later use.
+      Note that at this stage the migration is still in precopy mode.
+
 Slave message types
 -------------------
 

hw/virtio/vhost-user.c +48 -0

index aab35c4845..431858b219 100644
@@ -78,6 +78,7 @@ typedef enum VhostUserRequest {
     VHOST_USER_SET_CONFIG = 25,
     VHOST_USER_CREATE_CRYPTO_SESSION = 26,
     VHOST_USER_CLOSE_CRYPTO_SESSION = 27,
+    VHOST_USER_POSTCOPY_ADVISE  = 28,
     VHOST_USER_MAX
 } VhostUserRequest;
 

@@ -795,6 +796,50 @@ out:
     return ret;
 }
 
+/*
+ * Called at the start of an inbound postcopy on reception of the
+ * 'advise' command.
+ */
+static int vhost_user_postcopy_advise(struct vhost_dev *dev, Error **errp)
+{
+    struct vhost_user *u = dev->opaque;
+    CharBackend *chr = u->chr;
+    int ufd;
+    VhostUserMsg msg = {
+        .hdr.request = VHOST_USER_POSTCOPY_ADVISE,
+        .hdr.flags = VHOST_USER_VERSION,
+    };
+
+    if (vhost_user_write(dev, &msg, NULL, 0) < 0) {
+        error_setg(errp, "Failed to send postcopy_advise to vhost");
+        return -1;
+    }
+
+    if (vhost_user_read(dev, &msg) < 0) {
+        error_setg(errp, "Failed to get postcopy_advise reply from vhost");
+        return -1;
+    }
+
+    if (msg.hdr.request != VHOST_USER_POSTCOPY_ADVISE) {
+        error_setg(errp, "Unexpected msg type. Expected %d received %d",
+                     VHOST_USER_POSTCOPY_ADVISE, msg.hdr.request);
+        return -1;
+    }
+
+    if (msg.hdr.size != sizeof(msg.payload.u64)) {
+        error_setg(errp, "Received bad msg size.");
+        return -1;
+    }
+    ufd = qemu_chr_fe_get_msgfd(chr);
+    if (ufd < 0) {
+        error_setg(errp, "%s: Failed to get ufd", __func__);
+        return -1;
+    }
+
+    /* TODO: register ufd with userfault thread */
+    return 0;
+}
+
 static int vhost_user_postcopy_notifier(NotifierWithReturn *notifier,
                                         void *opaque)
 {

@@ -814,6 +859,9 @@ static int vhost_user_postcopy_notifier(NotifierWithReturn *notifier,
         }
         break;
 
+    case POSTCOPY_NOTIFY_INBOUND_ADVISE:
+        return vhost_user_postcopy_advise(dev, pnd->errp);
+
     default:
         /* We ignore notifications we don't know */
         break;

migration/postcopy-ram.h +1 -0

index 2e879bbacb..0421c98d57 100644
@@ -130,6 +130,7 @@ void postcopy_infrastructure_init(void);
  */
 enum PostcopyNotifyReason {
     POSTCOPY_NOTIFY_PROBE = 0,
+    POSTCOPY_NOTIFY_INBOUND_ADVISE,
 };
 
 struct PostcopyNotifyData {

migration/savevm.c +1 -0

index 358c5b51e2..1f2bf12a28 100644
@@ -1386,6 +1386,7 @@ static int loadvm_postcopy_handle_advise(MigrationIncomingState *mis,
 {
     PostcopyState ps = postcopy_state_set(POSTCOPY_INCOMING_ADVISE);
     uint64_t remote_pagesize_summary, local_pagesize_summary, remote_tps;
+    Error *local_err = NULL;
 
     trace_loadvm_postcopy_handle_advise();
     if (ps != POSTCOPY_INCOMING_NONE) {

@@ -1451,6 +1452,11 @@ static int loadvm_postcopy_handle_advise(MigrationIncomingState *mis,
         return -1;
     }
 

Quoted 12 lines

migration/savevm.c +5 -1

+    if (postcopy_notify(POSTCOPY_NOTIFY_INBOUND_ADVISE, &local_err)) {
+        error_report_err(local_err);
+        return -1;
+    }
+
     if (ram_postcopy_incoming_init(mis)) {
         return -1;
     }
--

2.14.3

Signature · 1 line

Marc-André Lureau

From: "Dr. David Alan Gilbert" <[email protected]>

Open a userfaultfd (on a postcopy_advise) and send it back in the reply to the qemu for it to monitor.

Signed-off-by Dr. David Alan Gilbert <[email protected]>

From: "Dr. David Alan Gilbert" <[email protected]>

Open a userfaultfd (on a postcopy_advise) and send it back in the reply to the qemu for it to monitor.

Signed-off-by Dr. David Alan Gilbert <[email protected]>

---
 contrib/libvhost-user/libvhost-user.c | 45 +++++++++++++++++++++++++++++++----
 contrib/libvhost-user/libvhost-user.h |  3 +++
 2 files changed, 44 insertions(+), 4 deletions(-)

contrib/libvhost-user/libvhost-user.c +41 -4

index eb0ab9338c..0b563fc5ae 100644
@@ -26,9 +26,20 @@
 #include <sys/socket.h>
 #include <sys/eventfd.h>
 #include <sys/mman.h>
+#include "qemu/compiler.h"
+
+#if defined(__linux__)
+#include <sys/syscall.h>
+#include <fcntl.h>
+#include <sys/ioctl.h>
 #include <linux/vhost.h>
 
-#include "qemu/compiler.h"
+#ifdef __NR_userfaultfd
+#include <linux/userfaultfd.h>
+#endif
+
+#endif
+
 #include "qemu/atomic.h"
 
 #include "libvhost-user.h"

@@ -888,11 +899,37 @@ vu_set_config(VuDev *dev, VhostUserMsg *vmsg)
 static bool
 vu_set_postcopy_advise(VuDev *dev, VhostUserMsg *vmsg)
 {
-    /* TODO: Open ufd, pass it back in the request
-     * TODO: Add addresses
-     */
+    dev->postcopy_ufd = -1;
+#ifdef UFFDIO_API
+    struct uffdio_api api_struct;
+
+    dev->postcopy_ufd = syscall(__NR_userfaultfd, O_CLOEXEC | O_NONBLOCK);
+    /* TODO: Add addresses */
     vmsg->payload.u64 = 0xcafe;
     vmsg->size = sizeof(vmsg->payload.u64);
+#endif
+
+    if (dev->postcopy_ufd == -1) {
+        vu_panic(dev, "Userfaultfd not available: %s", strerror(errno));
+        goto out;
+    }
+
+#ifdef UFFDIO_API
+    api_struct.api = UFFD_API;
+    api_struct.features = 0;
+    if (ioctl(dev->postcopy_ufd, UFFDIO_API, &api_struct)) {
+        vu_panic(dev, "Failed UFFDIO_API: %s", strerror(errno));
+        close(dev->postcopy_ufd);
+        dev->postcopy_ufd = -1;
+        goto out;
+    }
+    /* TODO: Stash feature flags somewhere */
+#endif
+
+out:
+    /* Return a ufd to the QEMU */
+    vmsg->fd_num = 1;
+    vmsg->fds[0] = dev->postcopy_ufd;
     return true; /* = send a reply */
 }
 

contrib/libvhost-user/libvhost-user.h +3 -1

index 00d78a8810..074b7860f6 100644
@@ -282,6 +282,9 @@ struct VuDev {
      * re-initialize */
     vu_panic_cb panic;
     const VuDevIface *iface;
+
+    /* Postcopy data */
+    int postcopy_ufd;
 };
 
 typedef struct VuVirtqElement {
--

2.14.3

From: "Dr. David Alan Gilbert" <[email protected]>

Register the UFD that comes in as the response to the 'advise' method with the postcopy code.

Signed-off-by Dr. David Alan Gilbert <[email protected]>

From: "Dr. David Alan Gilbert" <[email protected]>

Register the UFD that comes in as the response to the 'advise' method with the postcopy code.

Signed-off-by Dr. David Alan Gilbert <[email protected]>

---
 hw/virtio/vhost-user.c | 20 +++++++++++++++++++-
 1 file changed, 19 insertions(+), 1 deletion(-)

hw/virtio/vhost-user.c +19 -2

index 431858b219..9f74111e92 100644
@@ -171,6 +171,7 @@ struct vhost_user {
     CharBackend *chr;
     int slave_fd;
     NotifierWithReturn postcopy_notifier;
+    struct PostCopyFD  postcopy_fd;
 };
 
 static bool ioeventfd_enabled(void)

@@ -796,6 +797,17 @@ out:
     return ret;
 }
 
+/*
+ * Called back from the postcopy fault thread when a fault is received on our
+ * ufd.
+ * TODO: This is Linux specific
+ */
+static int vhost_user_postcopy_fault_handler(struct PostCopyFD *pcfd,
+                                             void *ufd)
+{
+    return 0;
+}
+
 /*
  * Called at the start of an inbound postcopy on reception of the
  * 'advise' command.

@@ -835,8 +847,14 @@ static int vhost_user_postcopy_advise(struct vhost_dev *dev, Error **errp)
         error_setg(errp, "%s: Failed to get ufd", __func__);
         return -1;
     }
+    fcntl(ufd, F_SETFL, O_NONBLOCK);
 
-    /* TODO: register ufd with userfault thread */
+    /* register ufd with userfault thread */
+    u->postcopy_fd.fd = ufd;
+    u->postcopy_fd.data = dev;
+    u->postcopy_fd.handler = vhost_user_postcopy_fault_handler;
+    u->postcopy_fd.idstr = "vhost-user"; /* Need to find unique name */
+    postcopy_register_shared_ufd(&u->postcopy_fd);
     return 0;
 }
 
--

2.14.3

From: "Dr. David Alan Gilbert" <[email protected]>

Notify the vhost-user client on reception of the 'postcopy-listen' event from the source.

Signed-off-by Dr. David Alan Gilbert <[email protected]>

Reviewed-by Marc-André Lureau <[email protected]>

Reviewed-by Peter Xu <[email protected]>

---
 contrib/libvhost-user/libvhost-user.c | 19 +++++++++++++++++++
 contrib/libvhost-user/libvhost-user.h |  2 ++
 docs/interop/vhost-user.txt           | 11 +++++++++++
 hw/virtio/trace-events                |  3 +++
 hw/virtio/vhost-user.c                | 34 ++++++++++++++++++++++++++++++++++
 migration/postcopy-ram.h              |  1 +
 migration/savevm.c                    |  7 +++++++
 7 files changed, 77 insertions(+)

contrib/libvhost-user/libvhost-user.c +19 -0

index 0b563fc5ae..beec7695a8 100644
@@ -98,6 +98,7 @@ vu_request_to_string(unsigned int req)
         REQ(VHOST_USER_GET_CONFIG),
         REQ(VHOST_USER_SET_CONFIG),
         REQ(VHOST_USER_POSTCOPY_ADVISE),
+        REQ(VHOST_USER_POSTCOPY_LISTEN),
         REQ(VHOST_USER_MAX),
     };
 #undef REQ

@@ -933,6 +934,22 @@ out:
     return true; /* = send a reply */
 }
 
+static bool
+vu_set_postcopy_listen(VuDev *dev, VhostUserMsg *vmsg)
+{
+    vmsg->payload.u64 = -1;
+    vmsg->size = sizeof(vmsg->payload.u64);
+
+    if (dev->nregions) {
+        vu_panic(dev, "Regions already registered at postcopy-listen");
+        return true;
+    }
+    dev->postcopy_listening = true;
+
+    vmsg->flags = VHOST_USER_VERSION |  VHOST_USER_REPLY_MASK;
+    vmsg->payload.u64 = 0; /* Success */
+    return true;
+}
 static bool
 vu_process_message(VuDev *dev, VhostUserMsg *vmsg)
 {

@@ -1006,6 +1023,8 @@ vu_process_message(VuDev *dev, VhostUserMsg *vmsg)
         break;
     case VHOST_USER_POSTCOPY_ADVISE:
         return vu_set_postcopy_advise(dev, vmsg);
+    case VHOST_USER_POSTCOPY_LISTEN:
+        return vu_set_postcopy_listen(dev, vmsg);
     default:
         vmsg_close_fds(vmsg);
         vu_panic(dev, "Unhandled request: %d", vmsg->request);

contrib/libvhost-user/libvhost-user.h +2 -0

index 074b7860f6..ed505cf0c1 100644
@@ -86,6 +86,7 @@ typedef enum VhostUserRequest {
     VHOST_USER_CREATE_CRYPTO_SESSION = 26,
     VHOST_USER_CLOSE_CRYPTO_SESSION = 27,
     VHOST_USER_POSTCOPY_ADVISE  = 28,
+    VHOST_USER_POSTCOPY_LISTEN  = 29,
     VHOST_USER_MAX
 } VhostUserRequest;
 

@@ -285,6 +286,7 @@ struct VuDev {
 
     /* Postcopy data */
     int postcopy_ufd;
+    bool postcopy_listening;
 };
 
 typedef struct VuVirtqElement {

docs/interop/vhost-user.txt +11 -0

index acc4186c79..7cc7006ef3 100644
@@ -709,6 +709,17 @@ Master message types
       the slave must open a userfaultfd for later use.
       Note that at this stage the migration is still in precopy mode.
 
+ * VHOST_USER_POSTCOPY_LISTEN
+      Id: 29
+      Master payload: N/A
+
+      Master advises slave that a transition to postcopy mode has happened.
+      The slave must ensure that shared memory is registered with userfaultfd
+      to cause faulting of non-present pages.
+
+      This is always sent sometime after a VHOST_USER_POSTCOPY_ADVISE, and
+      thus only when VHOST_USER_PROTOCOL_F_PAGEFAULT is supported.
+
 Slave message types
 -------------------
 

hw/virtio/trace-events +3 -0

index 742ff0f90b..06ec03d6e7 100644
@@ -6,6 +6,9 @@ vhost_region_add_section(const char *name, uint64_t gpa, uint64_t size, uint64_t
 vhost_region_add_section_abut(const char *name, uint64_t new_size) "%s: 0x%"PRIx64
 vhost_section(const char *name, int r) "%s:%d"
 
+# hw/virtio/vhost-user.c
+vhost_user_postcopy_listen(void) ""
+
 # hw/virtio/virtio.c
 virtqueue_alloc_element(void *elem, size_t sz, unsigned in_num, unsigned out_num) "elem %p size %zd in_num %u out_num %u"
 virtqueue_fill(void *vq, const void *elem, unsigned int len, unsigned int idx) "vq %p elem %p len %u idx %u"

hw/virtio/vhost-user.c +34 -0

index 9f74111e92..ee200f703e 100644
@@ -20,6 +20,7 @@
 #include "sysemu/cryptodev.h"
 #include "migration/migration.h"
 #include "migration/postcopy-ram.h"
+#include "trace.h"
 
 #include <sys/ioctl.h>
 #include <sys/socket.h>

@@ -79,6 +80,7 @@ typedef enum VhostUserRequest {
     VHOST_USER_CREATE_CRYPTO_SESSION = 26,
     VHOST_USER_CLOSE_CRYPTO_SESSION = 27,
     VHOST_USER_POSTCOPY_ADVISE  = 28,
+    VHOST_USER_POSTCOPY_LISTEN  = 29,
     VHOST_USER_MAX
 } VhostUserRequest;
 

@@ -172,6 +174,8 @@ struct vhost_user {
     int slave_fd;
     NotifierWithReturn postcopy_notifier;
     struct PostCopyFD  postcopy_fd;
+    /* True once we've entered postcopy_listen */
+    bool               postcopy_listen;
 };
 
 static bool ioeventfd_enabled(void)

@@ -858,6 +862,33 @@ static int vhost_user_postcopy_advise(struct vhost_dev *dev, Error **errp)
     return 0;
 }
 
+/*
+ * Called at the switch to postcopy on reception of the 'listen' command.
+ */
+static int vhost_user_postcopy_listen(struct vhost_dev *dev, Error **errp)
+{
+    struct vhost_user *u = dev->opaque;
+    int ret;
+    VhostUserMsg msg = {
+        .hdr.request = VHOST_USER_POSTCOPY_LISTEN,
+        .hdr.flags = VHOST_USER_VERSION | VHOST_USER_NEED_REPLY_MASK,
+    };
+    u->postcopy_listen = true;
+    trace_vhost_user_postcopy_listen();
+    if (vhost_user_write(dev, &msg, NULL, 0) < 0) {
+        error_setg(errp, "Failed to send postcopy_listen to vhost");
+        return -1;
+    }
+
+    ret = process_message_reply(dev, &msg);
+    if (ret) {
+        error_setg(errp, "Failed to receive reply to postcopy_listen");
+        return ret;
+    }
+
+    return 0;
+}
+
 static int vhost_user_postcopy_notifier(NotifierWithReturn *notifier,
                                         void *opaque)
 {

@@ -880,6 +911,9 @@ static int vhost_user_postcopy_notifier(NotifierWithReturn *notifier,
     case POSTCOPY_NOTIFY_INBOUND_ADVISE:
         return vhost_user_postcopy_advise(dev, pnd->errp);
 
+    case POSTCOPY_NOTIFY_INBOUND_LISTEN:
+        return vhost_user_postcopy_listen(dev, pnd->errp);
+
     default:
         /* We ignore notifications we don't know */
         break;

migration/postcopy-ram.h +1 -0

index f21eef6702..c8ced3470b 100644
@@ -131,6 +131,7 @@ void postcopy_infrastructure_init(void);
 enum PostcopyNotifyReason {
     POSTCOPY_NOTIFY_PROBE = 0,
     POSTCOPY_NOTIFY_INBOUND_ADVISE,
+    POSTCOPY_NOTIFY_INBOUND_LISTEN,
 };
 
 struct PostcopyNotifyData {

migration/savevm.c +7 -1

index 1f2bf12a28..305c3ceaf5 100644
@@ -1618,6 +1618,8 @@ static int loadvm_postcopy_handle_listen(MigrationIncomingState *mis)
 {
     PostcopyState ps = postcopy_state_set(POSTCOPY_INCOMING_LISTENING);
     trace_loadvm_postcopy_handle_listen();
+    Error *local_err = NULL;
+
     if (ps != POSTCOPY_INCOMING_ADVISE && ps != POSTCOPY_INCOMING_DISCARD) {
         error_report("CMD_POSTCOPY_LISTEN in wrong postcopy state (%d)", ps);
         return -1;

@@ -1643,6 +1645,11 @@ static int loadvm_postcopy_handle_listen(MigrationIncomingState *mis)
         }
     }
 
+    if (postcopy_notify(POSTCOPY_NOTIFY_INBOUND_LISTEN, &local_err)) {
+        error_report_err(local_err);
+        return -1;
+    }
+
     if (mis->have_listen_thread) {
         error_report("CMD_POSTCOPY_RAM_LISTEN already has a listen thread");
         return -1;
--

2.14.3

Hi

On Thu, Mar 8, 2018 at 8:57 PM, Dr. David Alan Gilbert (git) <[email protected]> wrote:

Quoted 4 lines

From: "Dr. David Alan Gilbert" <[email protected]>

Notify the vhost-user client on reception of the 'postcopy-listen' event from the source.

minor nit, please change with "vhost-user slave", since that's the terminology we use for vhost-user protocol. (server/client is for the socket layer relation)

Quoted 216 lines

Signed-off-by Dr. David Alan Gilbert <[email protected]>

Reviewed-by Marc-André Lureau <[email protected]>

Reviewed-by Peter Xu <[email protected]>

---
 contrib/libvhost-user/libvhost-user.c | 19 +++++++++++++++++++
 contrib/libvhost-user/libvhost-user.h |  2 ++
 docs/interop/vhost-user.txt           | 11 +++++++++++
 hw/virtio/trace-events                |  3 +++
 hw/virtio/vhost-user.c                | 34 ++++++++++++++++++++++++++++++++++
 migration/postcopy-ram.h              |  1 +
 migration/savevm.c                    |  7 +++++++
 7 files changed, 77 insertions(+)

contrib/libvhost-user/libvhost-user.c +19 -0

index 0b563fc5ae..beec7695a8 100644
@@ -98,6 +98,7 @@ vu_request_to_string(unsigned int req)
         REQ(VHOST_USER_GET_CONFIG),
         REQ(VHOST_USER_SET_CONFIG),
         REQ(VHOST_USER_POSTCOPY_ADVISE),
+        REQ(VHOST_USER_POSTCOPY_LISTEN),
         REQ(VHOST_USER_MAX),
     };
 #undef REQ

@@ -933,6 +934,22 @@ out:
     return true; /* = send a reply */
 }
 
+static bool
+vu_set_postcopy_listen(VuDev *dev, VhostUserMsg *vmsg)
+{
+    vmsg->payload.u64 = -1;
+    vmsg->size = sizeof(vmsg->payload.u64);
+
+    if (dev->nregions) {
+        vu_panic(dev, "Regions already registered at postcopy-listen");
+        return true;
+    }
+    dev->postcopy_listening = true;
+
+    vmsg->flags = VHOST_USER_VERSION |  VHOST_USER_REPLY_MASK;
+    vmsg->payload.u64 = 0; /* Success */
+    return true;
+}
 static bool
 vu_process_message(VuDev *dev, VhostUserMsg *vmsg)
 {

@@ -1006,6 +1023,8 @@ vu_process_message(VuDev *dev, VhostUserMsg *vmsg)
         break;
     case VHOST_USER_POSTCOPY_ADVISE:
         return vu_set_postcopy_advise(dev, vmsg);
+    case VHOST_USER_POSTCOPY_LISTEN:
+        return vu_set_postcopy_listen(dev, vmsg);
     default:
         vmsg_close_fds(vmsg);
         vu_panic(dev, "Unhandled request: %d", vmsg->request);

contrib/libvhost-user/libvhost-user.h +2 -0

index 074b7860f6..ed505cf0c1 100644
@@ -86,6 +86,7 @@ typedef enum VhostUserRequest {
     VHOST_USER_CREATE_CRYPTO_SESSION = 26,
     VHOST_USER_CLOSE_CRYPTO_SESSION = 27,
     VHOST_USER_POSTCOPY_ADVISE  = 28,
+    VHOST_USER_POSTCOPY_LISTEN  = 29,
     VHOST_USER_MAX
 } VhostUserRequest;
 

@@ -285,6 +286,7 @@ struct VuDev {
 
     /* Postcopy data */
     int postcopy_ufd;
+    bool postcopy_listening;
 };
 
 typedef struct VuVirtqElement {

docs/interop/vhost-user.txt +11 -0

index acc4186c79..7cc7006ef3 100644
@@ -709,6 +709,17 @@ Master message types
       the slave must open a userfaultfd for later use.
       Note that at this stage the migration is still in precopy mode.
 
+ * VHOST_USER_POSTCOPY_LISTEN
+      Id: 29
+      Master payload: N/A
+
+      Master advises slave that a transition to postcopy mode has happened.
+      The slave must ensure that shared memory is registered with userfaultfd
+      to cause faulting of non-present pages.
+
+      This is always sent sometime after a VHOST_USER_POSTCOPY_ADVISE, and
+      thus only when VHOST_USER_PROTOCOL_F_PAGEFAULT is supported.
+
 Slave message types
 -------------------
 

hw/virtio/trace-events +3 -0

index 742ff0f90b..06ec03d6e7 100644
@@ -6,6 +6,9 @@ vhost_region_add_section(const char *name, uint64_t gpa, uint64_t size, uint64_t
 vhost_region_add_section_abut(const char *name, uint64_t new_size) "%s: 0x%"PRIx64
 vhost_section(const char *name, int r) "%s:%d"
 
+# hw/virtio/vhost-user.c
+vhost_user_postcopy_listen(void) ""
+
 # hw/virtio/virtio.c
 virtqueue_alloc_element(void *elem, size_t sz, unsigned in_num, unsigned out_num) "elem %p size %zd in_num %u out_num %u"
 virtqueue_fill(void *vq, const void *elem, unsigned int len, unsigned int idx) "vq %p elem %p len %u idx %u"

hw/virtio/vhost-user.c +34 -0

index 9f74111e92..ee200f703e 100644
@@ -20,6 +20,7 @@
 #include "sysemu/cryptodev.h"
 #include "migration/migration.h"
 #include "migration/postcopy-ram.h"
+#include "trace.h"
 
 #include <sys/ioctl.h>
 #include <sys/socket.h>

@@ -79,6 +80,7 @@ typedef enum VhostUserRequest {
     VHOST_USER_CREATE_CRYPTO_SESSION = 26,
     VHOST_USER_CLOSE_CRYPTO_SESSION = 27,
     VHOST_USER_POSTCOPY_ADVISE  = 28,
+    VHOST_USER_POSTCOPY_LISTEN  = 29,
     VHOST_USER_MAX
 } VhostUserRequest;
 

@@ -172,6 +174,8 @@ struct vhost_user {
     int slave_fd;
     NotifierWithReturn postcopy_notifier;
     struct PostCopyFD  postcopy_fd;
+    /* True once we've entered postcopy_listen */
+    bool               postcopy_listen;
 };
 
 static bool ioeventfd_enabled(void)

@@ -858,6 +862,33 @@ static int vhost_user_postcopy_advise(struct vhost_dev *dev, Error **errp)
     return 0;
 }
 
+/*
+ * Called at the switch to postcopy on reception of the 'listen' command.
+ */
+static int vhost_user_postcopy_listen(struct vhost_dev *dev, Error **errp)
+{
+    struct vhost_user *u = dev->opaque;
+    int ret;
+    VhostUserMsg msg = {
+        .hdr.request = VHOST_USER_POSTCOPY_LISTEN,
+        .hdr.flags = VHOST_USER_VERSION | VHOST_USER_NEED_REPLY_MASK,
+    };
+    u->postcopy_listen = true;
+    trace_vhost_user_postcopy_listen();
+    if (vhost_user_write(dev, &msg, NULL, 0) < 0) {
+        error_setg(errp, "Failed to send postcopy_listen to vhost");
+        return -1;
+    }
+
+    ret = process_message_reply(dev, &msg);
+    if (ret) {
+        error_setg(errp, "Failed to receive reply to postcopy_listen");
+        return ret;
+    }
+
+    return 0;
+}
+
 static int vhost_user_postcopy_notifier(NotifierWithReturn *notifier,
                                         void *opaque)
 {

@@ -880,6 +911,9 @@ static int vhost_user_postcopy_notifier(NotifierWithReturn *notifier,
     case POSTCOPY_NOTIFY_INBOUND_ADVISE:
         return vhost_user_postcopy_advise(dev, pnd->errp);
 
+    case POSTCOPY_NOTIFY_INBOUND_LISTEN:
+        return vhost_user_postcopy_listen(dev, pnd->errp);
+
     default:
         /* We ignore notifications we don't know */
         break;

migration/postcopy-ram.h +1 -0

index f21eef6702..c8ced3470b 100644
@@ -131,6 +131,7 @@ void postcopy_infrastructure_init(void);
 enum PostcopyNotifyReason {
     POSTCOPY_NOTIFY_PROBE = 0,
     POSTCOPY_NOTIFY_INBOUND_ADVISE,
+    POSTCOPY_NOTIFY_INBOUND_LISTEN,
 };
 
 struct PostcopyNotifyData {

migration/savevm.c +2 -0

index 1f2bf12a28..305c3ceaf5 100644
@@ -1618,6 +1618,8 @@ static int loadvm_postcopy_handle_listen(MigrationIncomingState *mis)
 {
     PostcopyState ps = postcopy_state_set(POSTCOPY_INCOMING_LISTENING);
     trace_loadvm_postcopy_handle_listen();
+    Error *local_err = NULL;
+
     if (ps != POSTCOPY_INCOMING_ADVISE && ps != POSTCOPY_INCOMING_DISCARD) {
         error_report("CMD_POSTCOPY_LISTEN in wrong postcopy state (%d)", ps);
         return -1;

@@ -1643,6 +1645,11 @@ static int loadvm_postcopy_handle_listen(MigrationIncomingState *mis)
         }
     }
 

Quoted 12 lines

migration/savevm.c +5 -1

+    if (postcopy_notify(POSTCOPY_NOTIFY_INBOUND_LISTEN, &local_err)) {
+        error_report_err(local_err);
+        return -1;
+    }
+
     if (mis->have_listen_thread) {
         error_report("CMD_POSTCOPY_RAM_LISTEN already has a listen thread");
         return -1;
--

2.14.3

Signature · 1 line

Marc-André Lureau

From: "Dr. David Alan Gilbert" <[email protected]>

Split the set_mem_table routines in both qemu and libvhost-user because the postcopy versions are going to be quite different once changes in the later patches are added.

Signed-off-by Dr. David Alan Gilbert <[email protected]>

---
 contrib/libvhost-user/libvhost-user.c | 53 ++++++++++++++++++++++++
 hw/virtio/vhost-user.c                | 77 ++++++++++++++++++++++++++++++++++-
 2 files changed, 128 insertions(+), 2 deletions(-)

contrib/libvhost-user/libvhost-user.c +53 -0

index beec7695a8..4922b2c722 100644
@@ -448,6 +448,55 @@ vu_reset_device_exec(VuDev *dev, VhostUserMsg *vmsg)
     return false;
 }
 
+static bool
+vu_set_mem_table_exec_postcopy(VuDev *dev, VhostUserMsg *vmsg)
+{
+    int i;
+    VhostUserMemory *memory = &vmsg->payload.memory;
+    dev->nregions = memory->nregions;
+    /* TODO: Postcopy specific code */
+    DPRINT("Nregions: %d\n", memory->nregions);
+    for (i = 0; i < dev->nregions; i++) {
+        void *mmap_addr;
+        VhostUserMemoryRegion *msg_region = &memory->regions[i];
+        VuDevRegion *dev_region = &dev->regions[i];
+
+        DPRINT("Region %d\n", i);
+        DPRINT("    guest_phys_addr: 0x%016"PRIx64"\n",
+               msg_region->guest_phys_addr);
+        DPRINT("    memory_size:     0x%016"PRIx64"\n",
+               msg_region->memory_size);
+        DPRINT("    userspace_addr   0x%016"PRIx64"\n",
+               msg_region->userspace_addr);
+        DPRINT("    mmap_offset      0x%016"PRIx64"\n",
+               msg_region->mmap_offset);
+
+        dev_region->gpa = msg_region->guest_phys_addr;
+        dev_region->size = msg_region->memory_size;
+        dev_region->qva = msg_region->userspace_addr;
+        dev_region->mmap_offset = msg_region->mmap_offset;
+
+        /* We don't use offset argument of mmap() since the
+         * mapped address has to be page aligned, and we use huge
+         * pages.  */
+        mmap_addr = mmap(0, dev_region->size + dev_region->mmap_offset,
+                         PROT_READ | PROT_WRITE, MAP_SHARED,
+                         vmsg->fds[i], 0);
+
+        if (mmap_addr == MAP_FAILED) {
+            vu_panic(dev, "region mmap error: %s", strerror(errno));
+        } else {
+            dev_region->mmap_addr = (uint64_t)(uintptr_t)mmap_addr;
+            DPRINT("    mmap_addr:       0x%016"PRIx64"\n",
+                   dev_region->mmap_addr);
+        }
+
+        close(vmsg->fds[i]);
+    }
+
+    return false;
+}
+
 static bool
 vu_set_mem_table_exec(VuDev *dev, VhostUserMsg *vmsg)
 {

@@ -464,6 +513,10 @@ vu_set_mem_table_exec(VuDev *dev, VhostUserMsg *vmsg)
     }
     dev->nregions = memory->nregions;
 
+    if (dev->postcopy_listening) {
+        return vu_set_mem_table_exec_postcopy(dev, vmsg);
+    }
+
     DPRINT("Nregions: %d\n", memory->nregions);
     for (i = 0; i < dev->nregions; i++) {
         void *mmap_addr;

hw/virtio/vhost-user.c +71 -0

index ee200f703e..311addc33b 100644
@@ -340,15 +340,86 @@ static int vhost_user_set_log_base(struct vhost_dev *dev, uint64_t base,
     return 0;
 }
 
+static int vhost_user_set_mem_table_postcopy(struct vhost_dev *dev,
+                                             struct vhost_memory *mem)
+{
+    int fds[VHOST_MEMORY_MAX_NREGIONS];
+    int i, fd;
+    size_t fd_num = 0;
+    bool reply_supported = virtio_has_feature(dev->protocol_features,
+                                              VHOST_USER_PROTOCOL_F_REPLY_ACK);
+    /* TODO: Add actual postcopy differences */
+    VhostUserMsg msg = {
+        .hdr.request = VHOST_USER_SET_MEM_TABLE,
+        .hdr.flags = VHOST_USER_VERSION,
+    };
+
+    if (reply_supported) {
+        msg.hdr.flags |= VHOST_USER_NEED_REPLY_MASK;
+    }
+
+    for (i = 0; i < dev->mem->nregions; ++i) {
+        struct vhost_memory_region *reg = dev->mem->regions + i;
+        ram_addr_t offset;
+        MemoryRegion *mr;
+
+        assert((uintptr_t)reg->userspace_addr == reg->userspace_addr);
+        mr = memory_region_from_host((void *)(uintptr_t)reg->userspace_addr,
+                                     &offset);
+        fd = memory_region_get_fd(mr);
+        if (fd > 0) {
+            msg.payload.memory.regions[fd_num].userspace_addr =
+                reg->userspace_addr;
+            msg.payload.memory.regions[fd_num].memory_size  = reg->memory_size;
+            msg.payload.memory.regions[fd_num].guest_phys_addr =
+                reg->guest_phys_addr;
+            msg.payload.memory.regions[fd_num].mmap_offset = offset;
+            assert(fd_num < VHOST_MEMORY_MAX_NREGIONS);
+            fds[fd_num++] = fd;
+        }
+    }
+
+    msg.payload.memory.nregions = fd_num;
+
+    if (!fd_num) {
+        error_report("Failed initializing vhost-user memory map, "
+                     "consider using -object memory-backend-file share=on");
+        return -1;
+    }
+
+    msg.hdr.size = sizeof(msg.payload.memory.nregions);
+    msg.hdr.size += sizeof(msg.payload.memory.padding);
+    msg.hdr.size += fd_num * sizeof(VhostUserMemoryRegion);
+
+    if (vhost_user_write(dev, &msg, fds, fd_num) < 0) {
+        return -1;
+    }
+
+    if (reply_supported) {
+        return process_message_reply(dev, &msg);
+    }
+
+    return 0;
+}
+
 static int vhost_user_set_mem_table(struct vhost_dev *dev,
                                     struct vhost_memory *mem)
 {
+    struct vhost_user *u = dev->opaque;
     int fds[VHOST_MEMORY_MAX_NREGIONS];
     int i, fd;
     size_t fd_num = 0;
+    bool do_postcopy = u->postcopy_listen && u->postcopy_fd.handler;
     bool reply_supported = virtio_has_feature(dev->protocol_features,
                                               VHOST_USER_PROTOCOL_F_REPLY_ACK);
 
+    if (do_postcopy) {
+        /* Postcopy has enough differences that it's best done in it's own
+         * version
+         */
+        return vhost_user_set_mem_table_postcopy(dev, mem);
+    }
+
     VhostUserMsg msg = {
         .hdr.request = VHOST_USER_SET_MEM_TABLE,
         .hdr.flags = VHOST_USER_VERSION,

hw/virtio/vhost-user.c +0 -0

             msg.payload.memory.regions[fd_num].mmap_offset = offset;
             fds[fd_num++] = fd;
         }

Signature · 1 line

2.14.3

Quoted 5 lines · Dr. David Alan Gilbert (git)

From: "Dr. David Alan Gilbert" <[email protected]>

Split the set_mem_table routines in both qemu and libvhost-user because the postcopy versions are going to be quite different once changes in the later patches are added.

Quoted 170 lines

Signed-off-by Dr. David Alan Gilbert <[email protected]>

---
 contrib/libvhost-user/libvhost-user.c | 53 ++++++++++++++++++++++++
 hw/virtio/vhost-user.c                | 77 ++++++++++++++++++++++++++++++++++-
 2 files changed, 128 insertions(+), 2 deletions(-)

contrib/libvhost-user/libvhost-user.c +53 -0

index beec7695a8..4922b2c722 100644
@@ -448,6 +448,55 @@ vu_reset_device_exec(VuDev *dev, VhostUserMsg *vmsg)
     return false;
 }
 
+static bool
+vu_set_mem_table_exec_postcopy(VuDev *dev, VhostUserMsg *vmsg)
+{
+    int i;
+    VhostUserMemory *memory = &vmsg->payload.memory;
+    dev->nregions = memory->nregions;
+    /* TODO: Postcopy specific code */
+    DPRINT("Nregions: %d\n", memory->nregions);
+    for (i = 0; i < dev->nregions; i++) {
+        void *mmap_addr;
+        VhostUserMemoryRegion *msg_region = &memory->regions[i];
+        VuDevRegion *dev_region = &dev->regions[i];
+
+        DPRINT("Region %d\n", i);
+        DPRINT("    guest_phys_addr: 0x%016"PRIx64"\n",
+               msg_region->guest_phys_addr);
+        DPRINT("    memory_size:     0x%016"PRIx64"\n",
+               msg_region->memory_size);
+        DPRINT("    userspace_addr   0x%016"PRIx64"\n",
+               msg_region->userspace_addr);
+        DPRINT("    mmap_offset      0x%016"PRIx64"\n",
+               msg_region->mmap_offset);
+
+        dev_region->gpa = msg_region->guest_phys_addr;
+        dev_region->size = msg_region->memory_size;
+        dev_region->qva = msg_region->userspace_addr;
+        dev_region->mmap_offset = msg_region->mmap_offset;
+
+        /* We don't use offset argument of mmap() since the
+         * mapped address has to be page aligned, and we use huge
+         * pages.  */
+        mmap_addr = mmap(0, dev_region->size + dev_region->mmap_offset,
+                         PROT_READ | PROT_WRITE, MAP_SHARED,
+                         vmsg->fds[i], 0);
+
+        if (mmap_addr == MAP_FAILED) {
+            vu_panic(dev, "region mmap error: %s", strerror(errno));
+        } else {
+            dev_region->mmap_addr = (uint64_t)(uintptr_t)mmap_addr;
+            DPRINT("    mmap_addr:       0x%016"PRIx64"\n",
+                   dev_region->mmap_addr);
+        }
+
+        close(vmsg->fds[i]);
+    }
+
+    return false;
+}
+
 static bool
 vu_set_mem_table_exec(VuDev *dev, VhostUserMsg *vmsg)
 {

@@ -464,6 +513,10 @@ vu_set_mem_table_exec(VuDev *dev, VhostUserMsg *vmsg)
     }
     dev->nregions = memory->nregions;
 
+    if (dev->postcopy_listening) {
+        return vu_set_mem_table_exec_postcopy(dev, vmsg);
+    }
+
     DPRINT("Nregions: %d\n", memory->nregions);
     for (i = 0; i < dev->nregions; i++) {
         void *mmap_addr;

hw/virtio/vhost-user.c +71 -0

index ee200f703e..311addc33b 100644
@@ -340,15 +340,86 @@ static int vhost_user_set_log_base(struct vhost_dev *dev, uint64_t base,
     return 0;
 }
 
+static int vhost_user_set_mem_table_postcopy(struct vhost_dev *dev,
+                                             struct vhost_memory *mem)
+{
+    int fds[VHOST_MEMORY_MAX_NREGIONS];
+    int i, fd;
+    size_t fd_num = 0;
+    bool reply_supported = virtio_has_feature(dev->protocol_features,
+                                              VHOST_USER_PROTOCOL_F_REPLY_ACK);
+    /* TODO: Add actual postcopy differences */
+    VhostUserMsg msg = {
+        .hdr.request = VHOST_USER_SET_MEM_TABLE,
+        .hdr.flags = VHOST_USER_VERSION,
+    };
+
+    if (reply_supported) {
+        msg.hdr.flags |= VHOST_USER_NEED_REPLY_MASK;
+    }
+
+    for (i = 0; i < dev->mem->nregions; ++i) {
+        struct vhost_memory_region *reg = dev->mem->regions + i;
+        ram_addr_t offset;
+        MemoryRegion *mr;
+
+        assert((uintptr_t)reg->userspace_addr == reg->userspace_addr);
+        mr = memory_region_from_host((void *)(uintptr_t)reg->userspace_addr,
+                                     &offset);
+        fd = memory_region_get_fd(mr);
+        if (fd > 0) {
+            msg.payload.memory.regions[fd_num].userspace_addr =
+                reg->userspace_addr;
+            msg.payload.memory.regions[fd_num].memory_size  = reg->memory_size;
+            msg.payload.memory.regions[fd_num].guest_phys_addr =
+                reg->guest_phys_addr;
+            msg.payload.memory.regions[fd_num].mmap_offset = offset;
+            assert(fd_num < VHOST_MEMORY_MAX_NREGIONS);
+            fds[fd_num++] = fd;
+        }
+    }
+
+    msg.payload.memory.nregions = fd_num;
+
+    if (!fd_num) {
+        error_report("Failed initializing vhost-user memory map, "
+                     "consider using -object memory-backend-file share=on");
+        return -1;
+    }
+
+    msg.hdr.size = sizeof(msg.payload.memory.nregions);
+    msg.hdr.size += sizeof(msg.payload.memory.padding);
+    msg.hdr.size += fd_num * sizeof(VhostUserMemoryRegion);
+
+    if (vhost_user_write(dev, &msg, fds, fd_num) < 0) {
+        return -1;
+    }
+
+    if (reply_supported) {
+        return process_message_reply(dev, &msg);
+    }
+
+    return 0;
+}
+
 static int vhost_user_set_mem_table(struct vhost_dev *dev,
                                     struct vhost_memory *mem)
 {
+    struct vhost_user *u = dev->opaque;
     int fds[VHOST_MEMORY_MAX_NREGIONS];
     int i, fd;
     size_t fd_num = 0;
+    bool do_postcopy = u->postcopy_listen && u->postcopy_fd.handler;
     bool reply_supported = virtio_has_feature(dev->protocol_features,
                                               VHOST_USER_PROTOCOL_F_REPLY_ACK);
 
+    if (do_postcopy) {
+        /* Postcopy has enough differences that it's best done in it's own
+         * version
+         */
+        return vhost_user_set_mem_table_postcopy(dev, mem);
+    }
+
     VhostUserMsg msg = {
         .hdr.request = VHOST_USER_SET_MEM_TABLE,
         .hdr.flags = VHOST_USER_VERSION,

hw/virtio/vhost-user.c static int vhost_user_set_mem_table(struct vhost_dev *dev, · line 450

Quoted 10 lines

hw/virtio/vhost-user.c +4 -2

@@ -372,9 +443,11 @@ static int vhost_user_set_mem_table(struct vhost_dev *dev,
                 error_report("Failed preparing vhost-user memory table msg");
                 return -1;
             }
-            msg.payload.memory.regions[fd_num].userspace_addr = reg->userspace_addr;
+            msg.payload.memory.regions[fd_num].userspace_addr =
+                reg->userspace_addr;
             msg.payload.memory.regions[fd_num].memory_size  = reg->memory_size;
-            msg.payload.memory.regions[fd_num].guest_phys_addr = reg->guest_phys_addr;
+            msg.payload.memory.regions[fd_num].guest_phys_addr =
+                reg->guest_phys_addr;

I would still prefer to generalize things out as shared functions, since otherwise we are adding somehow duplicate codes, which is IMHO harder to maintain. And even if to do the duplication, I would keep the original code untouched (so these newlines will still not be needed).

But considering that the series has been there for a long time, I'll try to be less harsh... And yes, I know it's a burden to maintain private trees and keep rebasing.

Feel free to remove the newlines (so at least commit log of those lines won't be modified), and either way I'll offer mine:

Reviewed-by Peter Xu <[email protected]>

Thanks,

Quoted 6 lines

hw/virtio/vhost-user.c +0 -1

             msg.payload.memory.regions[fd_num].mmap_offset = offset;
             fds[fd_num++] = fd;
         }
--

2.14.3

--
Peter Xu

Hi

Quoted 6 lines · Peter Xu

Quoted 5 lines · Dr. David Alan Gilbert (git)

From: "Dr. David Alan Gilbert" <[email protected]>

Split the set_mem_table routines in both qemu and libvhost-user because the postcopy versions are going to be quite different once changes in the later patches are added.

You could add that this does not introduce functional change.

Quoted 194 lines

Quoted 170 lines

Signed-off-by Dr. David Alan Gilbert <[email protected]>

---
 contrib/libvhost-user/libvhost-user.c | 53 ++++++++++++++++++++++++
 hw/virtio/vhost-user.c                | 77 ++++++++++++++++++++++++++++++++++-
 2 files changed, 128 insertions(+), 2 deletions(-)

contrib/libvhost-user/libvhost-user.c +53 -0

index beec7695a8..4922b2c722 100644
@@ -448,6 +448,55 @@ vu_reset_device_exec(VuDev *dev, VhostUserMsg *vmsg)
     return false;
 }
 
+static bool
+vu_set_mem_table_exec_postcopy(VuDev *dev, VhostUserMsg *vmsg)
+{
+    int i;
+    VhostUserMemory *memory = &vmsg->payload.memory;
+    dev->nregions = memory->nregions;
+    /* TODO: Postcopy specific code */
+    DPRINT("Nregions: %d\n", memory->nregions);
+    for (i = 0; i < dev->nregions; i++) {
+        void *mmap_addr;
+        VhostUserMemoryRegion *msg_region = &memory->regions[i];
+        VuDevRegion *dev_region = &dev->regions[i];
+
+        DPRINT("Region %d\n", i);
+        DPRINT("    guest_phys_addr: 0x%016"PRIx64"\n",
+               msg_region->guest_phys_addr);
+        DPRINT("    memory_size:     0x%016"PRIx64"\n",
+               msg_region->memory_size);
+        DPRINT("    userspace_addr   0x%016"PRIx64"\n",
+               msg_region->userspace_addr);
+        DPRINT("    mmap_offset      0x%016"PRIx64"\n",
+               msg_region->mmap_offset);
+
+        dev_region->gpa = msg_region->guest_phys_addr;
+        dev_region->size = msg_region->memory_size;
+        dev_region->qva = msg_region->userspace_addr;
+        dev_region->mmap_offset = msg_region->mmap_offset;
+
+        /* We don't use offset argument of mmap() since the
+         * mapped address has to be page aligned, and we use huge
+         * pages.  */
+        mmap_addr = mmap(0, dev_region->size + dev_region->mmap_offset,
+                         PROT_READ | PROT_WRITE, MAP_SHARED,
+                         vmsg->fds[i], 0);
+
+        if (mmap_addr == MAP_FAILED) {
+            vu_panic(dev, "region mmap error: %s", strerror(errno));
+        } else {
+            dev_region->mmap_addr = (uint64_t)(uintptr_t)mmap_addr;
+            DPRINT("    mmap_addr:       0x%016"PRIx64"\n",
+                   dev_region->mmap_addr);
+        }
+
+        close(vmsg->fds[i]);
+    }
+
+    return false;
+}
+
 static bool
 vu_set_mem_table_exec(VuDev *dev, VhostUserMsg *vmsg)
 {

@@ -464,6 +513,10 @@ vu_set_mem_table_exec(VuDev *dev, VhostUserMsg *vmsg)
     }
     dev->nregions = memory->nregions;
 
+    if (dev->postcopy_listening) {
+        return vu_set_mem_table_exec_postcopy(dev, vmsg);
+    }
+
     DPRINT("Nregions: %d\n", memory->nregions);
     for (i = 0; i < dev->nregions; i++) {
         void *mmap_addr;

hw/virtio/vhost-user.c +71 -0

index ee200f703e..311addc33b 100644
@@ -340,15 +340,86 @@ static int vhost_user_set_log_base(struct vhost_dev *dev, uint64_t base,
     return 0;
 }
 
+static int vhost_user_set_mem_table_postcopy(struct vhost_dev *dev,
+                                             struct vhost_memory *mem)
+{
+    int fds[VHOST_MEMORY_MAX_NREGIONS];
+    int i, fd;
+    size_t fd_num = 0;
+    bool reply_supported = virtio_has_feature(dev->protocol_features,
+                                              VHOST_USER_PROTOCOL_F_REPLY_ACK);
+    /* TODO: Add actual postcopy differences */
+    VhostUserMsg msg = {
+        .hdr.request = VHOST_USER_SET_MEM_TABLE,
+        .hdr.flags = VHOST_USER_VERSION,
+    };
+
+    if (reply_supported) {
+        msg.hdr.flags |= VHOST_USER_NEED_REPLY_MASK;
+    }
+
+    for (i = 0; i < dev->mem->nregions; ++i) {
+        struct vhost_memory_region *reg = dev->mem->regions + i;
+        ram_addr_t offset;
+        MemoryRegion *mr;
+
+        assert((uintptr_t)reg->userspace_addr == reg->userspace_addr);
+        mr = memory_region_from_host((void *)(uintptr_t)reg->userspace_addr,
+                                     &offset);
+        fd = memory_region_get_fd(mr);
+        if (fd > 0) {
+            msg.payload.memory.regions[fd_num].userspace_addr =
+                reg->userspace_addr;
+            msg.payload.memory.regions[fd_num].memory_size  = reg->memory_size;
+            msg.payload.memory.regions[fd_num].guest_phys_addr =
+                reg->guest_phys_addr;
+            msg.payload.memory.regions[fd_num].mmap_offset = offset;
+            assert(fd_num < VHOST_MEMORY_MAX_NREGIONS);
+            fds[fd_num++] = fd;
+        }
+    }
+
+    msg.payload.memory.nregions = fd_num;
+
+    if (!fd_num) {
+        error_report("Failed initializing vhost-user memory map, "
+                     "consider using -object memory-backend-file share=on");
+        return -1;
+    }
+
+    msg.hdr.size = sizeof(msg.payload.memory.nregions);
+    msg.hdr.size += sizeof(msg.payload.memory.padding);
+    msg.hdr.size += fd_num * sizeof(VhostUserMemoryRegion);
+
+    if (vhost_user_write(dev, &msg, fds, fd_num) < 0) {
+        return -1;
+    }
+
+    if (reply_supported) {
+        return process_message_reply(dev, &msg);
+    }
+
+    return 0;
+}
+
 static int vhost_user_set_mem_table(struct vhost_dev *dev,
                                     struct vhost_memory *mem)
 {
+    struct vhost_user *u = dev->opaque;
     int fds[VHOST_MEMORY_MAX_NREGIONS];
     int i, fd;
     size_t fd_num = 0;
+    bool do_postcopy = u->postcopy_listen && u->postcopy_fd.handler;
     bool reply_supported = virtio_has_feature(dev->protocol_features,
                                               VHOST_USER_PROTOCOL_F_REPLY_ACK);
 
+    if (do_postcopy) {
+        /* Postcopy has enough differences that it's best done in it's own
+         * version
+         */
+        return vhost_user_set_mem_table_postcopy(dev, mem);
+    }
+
     VhostUserMsg msg = {
         .hdr.request = VHOST_USER_SET_MEM_TABLE,
         .hdr.flags = VHOST_USER_VERSION,

hw/virtio/vhost-user.c static int vhost_user_set_mem_table(struct vhost_dev *dev, · line 450

Quoted 10 lines

hw/virtio/vhost-user.c +4 -2

@@ -372,9 +443,11 @@ static int vhost_user_set_mem_table(struct vhost_dev *dev,
                 error_report("Failed preparing vhost-user memory table msg");
                 return -1;
             }
-            msg.payload.memory.regions[fd_num].userspace_addr = reg->userspace_addr;
+            msg.payload.memory.regions[fd_num].userspace_addr =
+                reg->userspace_addr;
             msg.payload.memory.regions[fd_num].memory_size  = reg->memory_size;
-            msg.payload.memory.regions[fd_num].guest_phys_addr = reg->guest_phys_addr;
+            msg.payload.memory.regions[fd_num].guest_phys_addr =
+                reg->guest_phys_addr;

I would still prefer to generalize things out as shared functions, since otherwise we are adding somehow duplicate codes, which is IMHO harder to maintain. And even if to do the duplication, I would keep the original code untouched (so these newlines will still not be needed).

But considering that the series has been there for a long time, I'll try to be less harsh... And yes, I know it's a burden to maintain private trees and keep rebasing.

Feel free to remove the newlines (so at least commit log of those lines won't be modified), and either way I'll offer mine:

Quoted 1 line

Reviewed-by Peter Xu <[email protected]>

I have the same feeling as Peter, but otherwise no objection - we can try to factorize again later if possible:

Reviewed-by Marc-André Lureau <[email protected]>

Quoted 11 lines

Thanks,

Quoted 6 lines

hw/virtio/vhost-user.c +0 -1

             msg.payload.memory.regions[fd_num].mmap_offset = offset;
             fds[fd_num++] = fd;
         }
--

2.14.3

Quoted 3 lines

--
Peter Xu

Signature · 1 line

Marc-André Lureau

From: "Dr. David Alan Gilbert" <[email protected]>

Utility for testing the map when you already know the offset in the RAMBlock.

Signed-off-by Dr. David Alan Gilbert <[email protected]>

Reviewed-by Peter Xu <[email protected]>

---
 migration/ram.c | 5 +++++
 migration/ram.h | 1 +
 2 files changed, 6 insertions(+)

migration/ram.c +5 -0

index 3b6c077964..3d754422e8 100644
@@ -169,6 +169,11 @@ int ramblock_recv_bitmap_test(RAMBlock *rb, void *host_addr)
                     rb->receivedmap);
 }
 
+bool ramblock_recv_bitmap_test_byte_offset(RAMBlock *rb, uint64_t byte_offset)
+{
+    return test_bit(byte_offset >> TARGET_PAGE_BITS, rb->receivedmap);
+}
+
 void ramblock_recv_bitmap_set(RAMBlock *rb, void *host_addr)
 {
     set_bit_atomic(ramblock_recv_bitmap_offset(host_addr, rb), rb->receivedmap);

migration/ram.h +1 -1

index 53f0021c51..5030be110a 100644
@@ -60,6 +60,7 @@ int ram_postcopy_incoming_init(MigrationIncomingState *mis);
 void ram_handle_compressed(void *host, uint8_t ch, uint64_t size);
 
 int ramblock_recv_bitmap_test(RAMBlock *rb, void *host_addr);
+bool ramblock_recv_bitmap_test_byte_offset(RAMBlock *rb, uint64_t byte_offset);
 void ramblock_recv_bitmap_set(RAMBlock *rb, void *host_addr);
 void ramblock_recv_bitmap_set_range(RAMBlock *rb, void *host_addr, size_t nr);
 
--

2.14.3

From: "Dr. David Alan Gilbert" <[email protected]>

When new regions are sent to the client using SET_MEM_TABLE, register them with the userfaultfd.

Signed-off-by Dr. David Alan Gilbert <[email protected]>

---
 contrib/libvhost-user/libvhost-user.c | 34 ++++++++++++++++++++++++++++++++++
 1 file changed, 34 insertions(+)

contrib/libvhost-user/libvhost-user.c +2 -0

index 4922b2c722..a18bc74a7c 100644
@@ -494,6 +494,40 @@ vu_set_mem_table_exec_postcopy(VuDev *dev, VhostUserMsg *vmsg)
         close(vmsg->fds[i]);
     }
 
+    /* TODO: Get address back to QEMU */
+    for (i = 0; i < dev->nregions; i++) {

contrib/libvhost-user/libvhost-user.c +22 -0

+        reg_struct.mode = UFFDIO_REGISTER_MODE_MISSING;
+
+        if (ioctl(dev->postcopy_ufd, UFFDIO_REGISTER, &reg_struct)) {
+            vu_panic(dev, "%s: Failed to userfault region %d "
+                          "@%p + size:%zx offset: %zx: (ufd=%d)%s\n",
+                     __func__, i,
+                     dev_region->mmap_addr,
+                     dev_region->size, dev_region->mmap_offset,
+                     dev->postcopy_ufd, strerror(errno));
+            return false;
+        }
+        if (!(reg_struct.ioctls & ((__u64)1 << _UFFDIO_COPY))) {
+            vu_panic(dev, "%s Region (%d) doesn't support COPY",
+                     __func__, i);
+            return false;
+        }
+        DPRINT("%s: region %d: Registered userfault for %llx + %llx\n",
+                __func__, i, reg_struct.range.start, reg_struct.range.len);
+        /* TODO: Stash 'zero' support flags somewhere */
+#endif
+    }
+
     return false;
 }
 

Signature · 1 line

2.14.3

Quoted 18 lines · Dr. David Alan Gilbert (git)

From: "Dr. David Alan Gilbert" <[email protected]>

When new regions are sent to the client using SET_MEM_TABLE, register them with the userfaultfd.

Signed-off-by Dr. David Alan Gilbert <[email protected]>

---
 contrib/libvhost-user/libvhost-user.c | 34 ++++++++++++++++++++++++++++++++++
 1 file changed, 34 insertions(+)

contrib/libvhost-user/libvhost-user.c +0 -0

index 4922b2c722..a18bc74a7c 100644
@@ -494,6 +494,40 @@ vu_set_mem_table_exec_postcopy(VuDev *dev, VhostUserMsg *vmsg)
         close(vmsg->fds[i]);
     }
 

contrib/libvhost-user/libvhost-user.c vu_set_mem_table_exec_postcopy(VuDev *dev, VhostUserMsg *vmsg) · line 508

Quoted 12 lines

contrib/libvhost-user/libvhost-user.c +12 -0

+    /* TODO: Get address back to QEMU */
+    for (i = 0; i < dev->nregions; i++) {
+        VuDevRegion *dev_region = &dev->regions[i];
+#ifdef UFFDIO_REGISTER
+        /* We should already have an open ufd. Mark each memory
+         * range as ufd.
+         * Note: Do we need any madvises? Well it's not been accessed
+         * yet, still probably need no THP to be safe, discard to be safe?
+         */
+        struct uffdio_register reg_struct;
+        reg_struct.range.start = (uintptr_t)dev_region->mmap_addr;
+        reg_struct.range.len = dev_region->size + dev_region->mmap_offset;

Do we really care the page faults between offset zero to mmap_offset?

I'm thinking whether we should add that mmap_offset into range.start instead of range.len.

Also, I see that in current vu_set_mem_table_exec():

/* We don't use offset argument of mmap() since the
* mapped address has to be page aligned, and we use huge
* pages. */
mmap_addr = mmap(0, dev_region->size + dev_region->mmap_offset,
PROT_READ | PROT_WRITE, MAP_SHARED, vmsg->fds[i], 0);

So adding the mmap_offset will help to make sure we'll use huge pages? Could it? Or say, how could we be sure that size+mmap_offset would be page aligned?

Thanks,

Quoted 25 lines

contrib/libvhost-user/libvhost-user.c +22 -0

+        reg_struct.mode = UFFDIO_REGISTER_MODE_MISSING;
+
+        if (ioctl(dev->postcopy_ufd, UFFDIO_REGISTER, &reg_struct)) {
+            vu_panic(dev, "%s: Failed to userfault region %d "
+                          "@%p + size:%zx offset: %zx: (ufd=%d)%s\n",
+                     __func__, i,
+                     dev_region->mmap_addr,
+                     dev_region->size, dev_region->mmap_offset,
+                     dev->postcopy_ufd, strerror(errno));
+            return false;
+        }
+        if (!(reg_struct.ioctls & ((__u64)1 << _UFFDIO_COPY))) {
+            vu_panic(dev, "%s Region (%d) doesn't support COPY",
+                     __func__, i);
+            return false;
+        }
+        DPRINT("%s: region %d: Registered userfault for %llx + %llx\n",
+                __func__, i, reg_struct.range.start, reg_struct.range.len);
+        /* TODO: Stash 'zero' support flags somewhere */
+#endif
+    }
+
     return false;
 }
 

Quoted 3 lines

Signature · 1 line

2.14.3

Signature · 1 line

Peter Xu

Quoted 19 lines · * Peter Xu ([email protected]) wrote:

Quoted 18 lines · Dr. David Alan Gilbert (git)

From: "Dr. David Alan Gilbert" <[email protected]>

When new regions are sent to the client using SET_MEM_TABLE, register them with the userfaultfd.

Signed-off-by Dr. David Alan Gilbert <[email protected]>

---
 contrib/libvhost-user/libvhost-user.c | 34 ++++++++++++++++++++++++++++++++++
 1 file changed, 34 insertions(+)

contrib/libvhost-user/libvhost-user.c +0 -0

index 4922b2c722..a18bc74a7c 100644
@@ -494,6 +494,40 @@ vu_set_mem_table_exec_postcopy(VuDev *dev, VhostUserMsg *vmsg)
         close(vmsg->fds[i]);
     }
 

Quoted 14 lines

contrib/libvhost-user/libvhost-user.c vu_set_mem_table_exec_postcopy(VuDev *dev, VhostUserMsg *vmsg) · line 508

Quoted 12 lines

contrib/libvhost-user/libvhost-user.c +12 -0

+    /* TODO: Get address back to QEMU */
+    for (i = 0; i < dev->nregions; i++) {
+        VuDevRegion *dev_region = &dev->regions[i];
+#ifdef UFFDIO_REGISTER
+        /* We should already have an open ufd. Mark each memory
+         * range as ufd.
+         * Note: Do we need any madvises? Well it's not been accessed
+         * yet, still probably need no THP to be safe, discard to be safe?
+         */
+        struct uffdio_register reg_struct;
+        reg_struct.range.start = (uintptr_t)dev_region->mmap_addr;
+        reg_struct.range.len = dev_region->size + dev_region->mmap_offset;

Do we really care the page faults between offset zero to mmap_offset?

No, but if we saw them we'd think it meant something had gone wrong, so it's good to trap them.

Quoted 13 lines

I'm thinking whether we should add that mmap_offset into range.start instead of range.len.

Also, I see that in current vu_set_mem_table_exec():

/* We don't use offset argument of mmap() since the
* mapped address has to be page aligned, and we use huge
* pages. */
mmap_addr = mmap(0, dev_region->size + dev_region->mmap_offset,
PROT_READ | PROT_WRITE, MAP_SHARED, vmsg->fds[i], 0);

Quoted 3 lines

So adding the mmap_offset will help to make sure we'll use huge pages? Could it? Or say, how could we be sure that size+mmap_offset would be page aligned?

If you look into the set_mem_table_exec (non-postcopy) you'll see that code and comment comes from the non-postcopy version; but it's something which as you say we could probably simplify now.

The problem used to be, before we did the merging as part of this series (0026 vhost Huge page align and merge), we could end up with mappings being passed from the qemu that were for small ranges of memory that weren't aligned to a huge page boundary and thus the mmap would fail. With the merging code that's no longer true, so it means we could simplify as you say; although this way it's a smaller change from the existing code.

Quoted 35 lines · Dr. David Alan Gilbert

Quoted 19 lines · * Peter Xu ([email protected]) wrote:

Quoted 18 lines · Dr. David Alan Gilbert (git)

From: "Dr. David Alan Gilbert" <[email protected]>

When new regions are sent to the client using SET_MEM_TABLE, register them with the userfaultfd.

Signed-off-by Dr. David Alan Gilbert <[email protected]>

---
 contrib/libvhost-user/libvhost-user.c | 34 ++++++++++++++++++++++++++++++++++
 1 file changed, 34 insertions(+)

contrib/libvhost-user/libvhost-user.c +0 -0

index 4922b2c722..a18bc74a7c 100644
@@ -494,6 +494,40 @@ vu_set_mem_table_exec_postcopy(VuDev *dev, VhostUserMsg *vmsg)
         close(vmsg->fds[i]);
     }
 

Quoted 14 lines

contrib/libvhost-user/libvhost-user.c vu_set_mem_table_exec_postcopy(VuDev *dev, VhostUserMsg *vmsg) · line 508

Quoted 12 lines

contrib/libvhost-user/libvhost-user.c +12 -0

+    /* TODO: Get address back to QEMU */
+    for (i = 0; i < dev->nregions; i++) {
+        VuDevRegion *dev_region = &dev->regions[i];
+#ifdef UFFDIO_REGISTER
+        /* We should already have an open ufd. Mark each memory
+         * range as ufd.
+         * Note: Do we need any madvises? Well it's not been accessed
+         * yet, still probably need no THP to be safe, discard to be safe?
+         */
+        struct uffdio_register reg_struct;
+        reg_struct.range.start = (uintptr_t)dev_region->mmap_addr;
+        reg_struct.range.len = dev_region->size + dev_region->mmap_offset;

Do we really care the page faults between offset zero to mmap_offset?

Quoted 2 lines

No, but if we saw them we'd think it meant something had gone wrong, so it's good to trap them.

I'm fine with that, especially since that's now only used in the test codes. However that's a still bit confusing to me, especially if current QEMU won't really handle that page fault (and it seems that should never happen). Maybe at least a comment would help on explaining why we need to explicitly extend the range to listen, just like below code when we do the mapping, though with a different reason.

Quoted 21 lines

Quoted 13 lines

I'm thinking whether we should add that mmap_offset into range.start instead of range.len.

Also, I see that in current vu_set_mem_table_exec():

/* We don't use offset argument of mmap() since the
* mapped address has to be page aligned, and we use huge
* pages. */
mmap_addr = mmap(0, dev_region->size + dev_region->mmap_offset,
PROT_READ | PROT_WRITE, MAP_SHARED, vmsg->fds[i], 0);

Quoted 3 lines

So adding the mmap_offset will help to make sure we'll use huge pages? Could it? Or say, how could we be sure that size+mmap_offset would be page aligned?

If you look into the set_mem_table_exec (non-postcopy) you'll see that code and comment comes from the non-postcopy version; but it's something which as you say we could probably simplify now.

Quoted 7 lines

The problem used to be, before we did the merging as part of this series (0026 vhost Huge page align and merge), we could end up with mappings being passed from the qemu that were for small ranges of memory that weren't aligned to a huge page boundary and thus the mmap would fail. With the merging code that's no longer true, so it means we could simplify as you say; although this way it's a smaller change from the existing code.

I was thinking what if the memory section was e.g. splitted as below:

- range A: [0x0, 0x10): non-RAM range, size 0x10
- range B: [0x10, 0x1ffff0): RAM range, size 0x1fffe0
- range C: [0x1ffff0, 0x200000): non-RAM range, size 0x10

Ranges A+B+C will cover a 2M page while vhost-user master should only send range B to the client. Then even size+mmap_offset (which is 0x1fffe0+0x10=0x1ffff0) shouldn't be aligned with the 2M boundary. If previous mmap() can fail, would this fail too?

For sure this is a question not directly related to current code - it's just something I'm not sure about. So it is not a blocker of current patch. Thanks,

Signature · 1 line

Peter Xu

Quoted 82 lines · * Peter Xu ([email protected]) wrote:

Quoted 35 lines · Dr. David Alan Gilbert

Quoted 19 lines · * Peter Xu ([email protected]) wrote:

Quoted 18 lines · Dr. David Alan Gilbert (git)

From: "Dr. David Alan Gilbert" <[email protected]>

When new regions are sent to the client using SET_MEM_TABLE, register them with the userfaultfd.

Signed-off-by Dr. David Alan Gilbert <[email protected]>

---
 contrib/libvhost-user/libvhost-user.c | 34 ++++++++++++++++++++++++++++++++++
 1 file changed, 34 insertions(+)

contrib/libvhost-user/libvhost-user.c +0 -0

index 4922b2c722..a18bc74a7c 100644
@@ -494,6 +494,40 @@ vu_set_mem_table_exec_postcopy(VuDev *dev, VhostUserMsg *vmsg)
         close(vmsg->fds[i]);
     }
 

Quoted 14 lines

contrib/libvhost-user/libvhost-user.c vu_set_mem_table_exec_postcopy(VuDev *dev, VhostUserMsg *vmsg) · line 508

Quoted 12 lines

contrib/libvhost-user/libvhost-user.c +12 -0

+    /* TODO: Get address back to QEMU */
+    for (i = 0; i < dev->nregions; i++) {
+        VuDevRegion *dev_region = &dev->regions[i];
+#ifdef UFFDIO_REGISTER
+        /* We should already have an open ufd. Mark each memory
+         * range as ufd.
+         * Note: Do we need any madvises? Well it's not been accessed
+         * yet, still probably need no THP to be safe, discard to be safe?
+         */
+        struct uffdio_register reg_struct;
+        reg_struct.range.start = (uintptr_t)dev_region->mmap_addr;
+        reg_struct.range.len = dev_region->size + dev_region->mmap_offset;

Do we really care the page faults between offset zero to mmap_offset?

Quoted 2 lines

No, but if we saw them we'd think it meant something had gone wrong, so it's good to trap them.

I'm fine with that, especially since that's now only used in the test codes. However that's a still bit confusing to me, especially if current QEMU won't really handle that page fault (and it seems that should never happen). Maybe at least a comment would help on explaining why we need to explicitly extend the range to listen, just like below code when we do the mapping, though with a different reason.

Quoted 21 lines

Quoted 13 lines

I'm thinking whether we should add that mmap_offset into range.start instead of range.len.

Also, I see that in current vu_set_mem_table_exec():

/* We don't use offset argument of mmap() since the
* mapped address has to be page aligned, and we use huge
* pages. */
mmap_addr = mmap(0, dev_region->size + dev_region->mmap_offset,
PROT_READ | PROT_WRITE, MAP_SHARED, vmsg->fds[i], 0);

Quoted 3 lines

So adding the mmap_offset will help to make sure we'll use huge pages? Could it? Or say, how could we be sure that size+mmap_offset would be page aligned?

If you look into the set_mem_table_exec (non-postcopy) you'll see that code and comment comes from the non-postcopy version; but it's something which as you say we could probably simplify now.

Quoted 7 lines

The problem used to be, before we did the merging as part of this series (0026 vhost Huge page align and merge), we could end up with mappings being passed from the qemu that were for small ranges of memory that weren't aligned to a huge page boundary and thus the mmap would fail. With the merging code that's no longer true, so it means we could simplify as you say; although this way it's a smaller change from the existing code.

I was thinking what if the memory section was e.g. splitted as below:

- range A: [0x0, 0x10): non-RAM range, size 0x10
- range B: [0x10, 0x1ffff0): RAM range, size 0x1fffe0
- range C: [0x1ffff0, 0x200000): non-RAM range, size 0x10

Quoted 4 lines

Ranges A+B+C will cover a 2M page while vhost-user master should only send range B to the client. Then even size+mmap_offset (which is 0x1fffe0+0x10=0x1ffff0) shouldn't be aligned with the 2M boundary. If previous mmap() can fail, would this fail too?

I think, 'vhost: Huge page align and merge' will round (B) to page boundaries so that the range sent to the slave will be aligned.

From: "Dr. David Alan Gilbert" <[email protected]>

When new regions are sent to the client using SET_MEM_TABLE, register them with the userfaultfd.

Signed-off-by Dr. David Alan Gilbert <[email protected]>

---
 contrib/libvhost-user/libvhost-user.c | 34 ++++++++++++++++++++++++++++++++++
 1 file changed, 34 insertions(+)

contrib/libvhost-user/libvhost-user.c +34 -1

index 4922b2c722..a18bc74a7c 100644
@@ -494,6 +494,40 @@ vu_set_mem_table_exec_postcopy(VuDev *dev, VhostUserMsg *vmsg)
         close(vmsg->fds[i]);
     }
 
+    /* TODO: Get address back to QEMU */
+    for (i = 0; i < dev->nregions; i++) {
+        VuDevRegion *dev_region = &dev->regions[i];
+#ifdef UFFDIO_REGISTER
+        /* We should already have an open ufd. Mark each memory
+         * range as ufd.
+         * Note: Do we need any madvises? Well it's not been accessed
+         * yet, still probably need no THP to be safe, discard to be safe?
+         */
+        struct uffdio_register reg_struct;
+        reg_struct.range.start = (uintptr_t)dev_region->mmap_addr;
+        reg_struct.range.len = dev_region->size + dev_region->mmap_offset;
+        reg_struct.mode = UFFDIO_REGISTER_MODE_MISSING;
+
+        if (ioctl(dev->postcopy_ufd, UFFDIO_REGISTER, &reg_struct)) {
+            vu_panic(dev, "%s: Failed to userfault region %d "
+                          "@%p + size:%zx offset: %zx: (ufd=%d)%s\n",
+                     __func__, i,
+                     dev_region->mmap_addr,
+                     dev_region->size, dev_region->mmap_offset,
+                     dev->postcopy_ufd, strerror(errno));
+            return false;
+        }
+        if (!(reg_struct.ioctls & ((__u64)1 << _UFFDIO_COPY))) {
+            vu_panic(dev, "%s Region (%d) doesn't support COPY",
+                     __func__, i);
+            return false;
+        }
+        DPRINT("%s: region %d: Registered userfault for %llx + %llx\n",
+                __func__, i, reg_struct.range.start, reg_struct.range.len);
+        /* TODO: Stash 'zero' support flags somewhere */
+#endif
+    }
+
     return false;
 }
 
--

2.14.3