pve-qemu/debian/patches/pve/0010-internal-snapshot-async.patch

From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Wolfgang Bumiller <w.bumiller@proxmox.com>
Date: Wed, 9 Dec 2015 16:04:32 +0100
Subject: [PATCH] internal snapshot async

---
 Makefile.objs           |   1 +
 hmp-commands-info.hx    |  13 ++
 hmp-commands.hx         |  32 +++
 hmp.c                   |  57 ++++++
 hmp.h                   |   5 +
 include/sysemu/sysemu.h |   5 +-
 migration/savevm.c      |  12 +-
 qapi-schema.json        |  68 +++++++
 qemu-options.hx         |  13 ++
 savevm-async.c          | 523 ++++++++++++++++++++++++++++++++++++++++++++++++
 vl.c                    |   8 +
 11 files changed, 730 insertions(+), 7 deletions(-)
 create mode 100644 savevm-async.c

diff --git a/Makefile.objs b/Makefile.objs
index 6167e7b17d..fbfbbb7f70 100644
--- a/Makefile.objs
+++ b/Makefile.objs
@@ -50,6 +50,7 @@ common-obj-$(CONFIG_LINUX) += fsdev/
 
 common-obj-y += migration/
 common-obj-y += page_cache.o #aio.o
+common-obj-y += savevm-async.o
 
 common-obj-$(CONFIG_SPICE) += spice-qemu-char.o
 
diff --git a/hmp-commands-info.hx b/hmp-commands-info.hx
index a53f105c52..5fc57a2210 100644
--- a/hmp-commands-info.hx
+++ b/hmp-commands-info.hx
@@ -560,6 +560,19 @@ Show current migration xbzrle cache size.
 ETEXI
 
     {
+        .name       = "savevm",
+        .args_type  = "",
+        .params     = "",
+        .help       = "show savevm status",
+        .cmd = hmp_info_savevm,
+    },
+
+STEXI
+@item info savevm
+show savevm status
+ETEXI
+
+    {
         .name       = "balloon",
         .args_type  = "",
         .params     = "",
diff --git a/hmp-commands.hx b/hmp-commands.hx
index 88192817b2..58940a762b 100644
--- a/hmp-commands.hx
+++ b/hmp-commands.hx
@@ -1777,3 +1777,35 @@ ETEXI
 STEXI
 @end table
 ETEXI
+
+    {
+        .name       = "savevm-start",
+        .args_type  = "statefile:s?",
+        .params     = "[statefile]",
+        .help       = "Prepare for snapshot and halt VM. Save VM state to statefile.",
+        .cmd = hmp_savevm_start,
+    },
+
+    {
+        .name       = "snapshot-drive",
+        .args_type  = "device:s,name:s",
+        .params     = "device name",
+        .help       = "Create internal snapshot.",
+        .cmd = hmp_snapshot_drive,
+    },
+
+    {
+        .name       = "delete-drive-snapshot",
+        .args_type  = "device:s,name:s",
+        .params     = "device name",
+        .help       = "Delete internal snapshot.",
+        .cmd = hmp_delete_drive_snapshot,
+    },
+
+    {
+        .name       = "savevm-end",
+        .args_type  = "",
+        .params     = "",
+        .help       = "Resume VM after snaphot.",
+        .cmd = hmp_savevm_end,
+    },
diff --git a/hmp.c b/hmp.c
index 904542d026..f725d061e6 100644
--- a/hmp.c
+++ b/hmp.c
@@ -2207,6 +2207,63 @@ void hmp_info_memory_devices(Monitor *mon, const QDict *qdict)
     qapi_free_MemoryDeviceInfoList(info_list);
 }
 
+void hmp_savevm_start(Monitor *mon, const QDict *qdict)
+{
+    Error *errp = NULL;
+    const char *statefile = qdict_get_try_str(qdict, "statefile");
+
+    qmp_savevm_start(statefile != NULL, statefile, &errp);
+    hmp_handle_error(mon, &errp);
+}
+
+void hmp_snapshot_drive(Monitor *mon, const QDict *qdict)
+{
+    Error *errp = NULL;
+    const char *name = qdict_get_str(qdict, "name");
+    const char *device = qdict_get_str(qdict, "device");
+
+    qmp_snapshot_drive(device, name, &errp);
+    hmp_handle_error(mon, &errp);
+}
+
+void hmp_delete_drive_snapshot(Monitor *mon, const QDict *qdict)
+{
+    Error *errp = NULL;
+    const char *name = qdict_get_str(qdict, "name");
+    const char *device = qdict_get_str(qdict, "device");
+
+    qmp_delete_drive_snapshot(device, name, &errp);
+    hmp_handle_error(mon, &errp);
+}
+
+void hmp_savevm_end(Monitor *mon, const QDict *qdict)
+{
+    Error *errp = NULL;
+
+    qmp_savevm_end(&errp);
+    hmp_handle_error(mon, &errp);
+}
+
+void hmp_info_savevm(Monitor *mon, const QDict *qdict)
+{
+    SaveVMInfo *info;
+    info = qmp_query_savevm(NULL);
+
+    if (info->has_status) {
+        monitor_printf(mon, "savevm status: %s\n", info->status);
+        monitor_printf(mon, "total time: %" PRIu64 " milliseconds\n",
+                       info->total_time);
+    } else {
+        monitor_printf(mon, "savevm status: not running\n");
+    }
+    if (info->has_bytes) {
+        monitor_printf(mon, "Bytes saved: %"PRIu64"\n", info->bytes);
+    }
+    if (info->has_error) {
+        monitor_printf(mon, "Error: %s\n", info->error);
+    }
+}
+
 void hmp_info_iothreads(Monitor *mon, const QDict *qdict)
 {
     IOThreadInfoList *info_list = qmp_query_iothreads(NULL);
diff --git a/hmp.h b/hmp.h
index 799fd371fa..0497afbf65 100644
--- a/hmp.h
+++ b/hmp.h
@@ -26,6 +26,7 @@ void hmp_info_status(Monitor *mon, const QDict *qdict);
 void hmp_info_uuid(Monitor *mon, const QDict *qdict);
 void hmp_info_chardev(Monitor *mon, const QDict *qdict);
 void hmp_info_mice(Monitor *mon, const QDict *qdict);
+void hmp_info_savevm(Monitor *mon, const QDict *qdict);
 void hmp_info_migrate(Monitor *mon, const QDict *qdict);
 void hmp_info_migrate_capabilities(Monitor *mon, const QDict *qdict);
 void hmp_info_migrate_parameters(Monitor *mon, const QDict *qdict);
@@ -92,6 +93,10 @@ void hmp_netdev_add(Monitor *mon, const QDict *qdict);
 void hmp_netdev_del(Monitor *mon, const QDict *qdict);
 void hmp_getfd(Monitor *mon, const QDict *qdict);
 void hmp_closefd(Monitor *mon, const QDict *qdict);
+void hmp_savevm_start(Monitor *mon, const QDict *qdict);
+void hmp_snapshot_drive(Monitor *mon, const QDict *qdict);
+void hmp_delete_drive_snapshot(Monitor *mon, const QDict *qdict);
+void hmp_savevm_end(Monitor *mon, const QDict *qdict);
 void hmp_sendkey(Monitor *mon, const QDict *qdict);
 void hmp_screendump(Monitor *mon, const QDict *qdict);
 void hmp_nbd_server_start(Monitor *mon, const QDict *qdict);
diff --git a/include/sysemu/sysemu.h b/include/sysemu/sysemu.h
index 576c7ce640..74623de16c 100644
--- a/include/sysemu/sysemu.h
+++ b/include/sysemu/sysemu.h
@@ -78,6 +78,7 @@ void qemu_remove_machine_init_done_notifier(Notifier *notify);
 void hmp_savevm(Monitor *mon, const QDict *qdict);
 int save_vmstate(Monitor *mon, const char *name);
 int load_vmstate(const char *name);
+int load_state_from_blockdev(const char *filename);
 void hmp_delvm(Monitor *mon, const QDict *qdict);
 void hmp_info_snapshots(Monitor *mon, const QDict *qdict);
 
@@ -105,13 +106,13 @@ enum qemu_vm_cmd {
 #define MAX_VM_CMD_PACKAGED_SIZE (1ul << 24)
 
 bool qemu_savevm_state_blocked(Error **errp);
-void qemu_savevm_state_begin(QEMUFile *f,
+int qemu_savevm_state_begin(QEMUFile *f,
                              const MigrationParams *params);
 void qemu_savevm_state_header(QEMUFile *f);
 int qemu_savevm_state_iterate(QEMUFile *f, bool postcopy);
 void qemu_savevm_state_cleanup(void);
 void qemu_savevm_state_complete_postcopy(QEMUFile *f);
-void qemu_savevm_state_complete_precopy(QEMUFile *f, bool iterable_only);
+int qemu_savevm_state_complete_precopy(QEMUFile *f, bool iterable_only);
 void qemu_savevm_state_pending(QEMUFile *f, uint64_t max_size,
                                uint64_t *res_non_postcopiable,
                                uint64_t *res_postcopiable);
diff --git a/migration/savevm.c b/migration/savevm.c
index 3b19a4a274..feb0dc6834 100644
--- a/migration/savevm.c
+++ b/migration/savevm.c
@@ -970,11 +970,11 @@ void qemu_savevm_state_header(QEMUFile *f)
 
 }
 
-void qemu_savevm_state_begin(QEMUFile *f,
+int qemu_savevm_state_begin(QEMUFile *f,
                              const MigrationParams *params)
 {
     SaveStateEntry *se;
-    int ret;
+    int ret = 0;
 
     trace_savevm_state_begin();
     QTAILQ_FOREACH(se, &savevm_state.handlers, entry) {
@@ -1002,6 +1002,7 @@ void qemu_savevm_state_begin(QEMUFile *f,
             break;
         }
     }
+    return ret;
 }
 
 /*
@@ -1105,7 +1106,7 @@ void qemu_savevm_state_complete_postcopy(QEMUFile *f)
     qemu_fflush(f);
 }
 
-void qemu_savevm_state_complete_precopy(QEMUFile *f, bool iterable_only)
+int qemu_savevm_state_complete_precopy(QEMUFile *f, bool iterable_only)
 {
     QJSON *vmdesc;
     int vmdesc_len;
@@ -1139,12 +1140,12 @@ void qemu_savevm_state_complete_precopy(QEMUFile *f, bool iterable_only)
         save_section_footer(f, se);
         if (ret < 0) {
             qemu_file_set_error(f, ret);
-            return;
+            return ret;
         }
     }
 
     if (iterable_only) {
-        return;
+        return ret;
     }
 
     vmdesc = qjson_new();
@@ -1191,6 +1192,7 @@ void qemu_savevm_state_complete_precopy(QEMUFile *f, bool iterable_only)
     qjson_destroy(vmdesc);
 
     qemu_fflush(f);
+    return qemu_file_get_error(f);
 }
 
 /* Give an estimate of the amount left to be transferred,
diff --git a/qapi-schema.json b/qapi-schema.json
index 1b14ff2476..361700d37c 100644
--- a/qapi-schema.json
+++ b/qapi-schema.json
@@ -723,6 +723,40 @@
            '*error-desc': 'str'} }
 
 ##
+# @SaveVMInfo:
+#
+# Information about current migration process.
+#
+# @status: string describing the current savevm status.
+#          This can be 'active', 'completed', 'failed'.
+#          If this field is not returned, no savevm process
+#          has been initiated
+#
+# @error: string containing error message is status is failed.
+#
+# @total-time: total amount of milliseconds since savevm started.
+#        If savevm has ended, it returns the total save time
+#
+# @bytes: total amount of data transfered
+#
+# Since: 1.3
+##
+{ 'struct': 'SaveVMInfo',
+  'data': {'*status': 'str', '*error': 'str',
+           '*total-time': 'int', '*bytes': 'int'} }
+
+##
+# @query-savevm:
+#
+# Returns information about current savevm process.
+#
+# Returns: @SaveVMInfo
+#
+# Since: 1.3
+##
+{ 'command': 'query-savevm', 'returns': 'SaveVMInfo' }
+
+##
 # @query-migrate:
 #
 # Returns information about current migration process. If migration
@@ -4735,9 +4769,43 @@
 #
 # Since: 1.2.0
 ##
+
 { 'command': 'query-target', 'returns': 'TargetInfo' }
 
 ##
+# @savevm-start:
+#
+# Prepare for snapshot and halt VM. Save VM state to statefile.
+#
+##
+{ 'command': 'savevm-start', 'data': { '*statefile': 'str' } }
+
+##
+# @snapshot-drive:
+#
+# Create an internal drive snapshot.
+#
+##
+{ 'command': 'snapshot-drive', 'data': { 'device': 'str', 'name': 'str' } }
+
+##
+# @delete-drive-snapshot:
+#
+# Delete a drive snapshot.
+#
+##
+{ 'command': 'delete-drive-snapshot', 'data': { 'device': 'str', 'name': 'str' } }
+
+##
+# @savevm-end:
+#
+# Resume VM after a snapshot.
+#
+##
+{ 'command': 'savevm-end' }
+
+
+##
 # @QKeyCode:
 #
 # An enumeration of key name.
diff --git a/qemu-options.hx b/qemu-options.hx
index 48dfffd86a..cbcb27da9a 100644
--- a/qemu-options.hx
+++ b/qemu-options.hx
@@ -3396,6 +3396,19 @@ STEXI
 Start right away with a saved state (@code{loadvm} in monitor)
 ETEXI
 
+DEF("loadstate", HAS_ARG, QEMU_OPTION_loadstate, \
+    "-loadstate file\n" \
+    "                start right away with a saved state\n",
+    QEMU_ARCH_ALL)
+STEXI
+@item -loadstate @var{file}
+@findex -loadstate
+Start right away with a saved state. This option does not rollback
+disk state like @code{loadvm}, so user must make sure that disk
+have correct state. @var{file} can be any valid device URL. See the section
+for "Device URL Syntax" for more information.
+ETEXI
+
 #ifndef _WIN32
 DEF("daemonize", 0, QEMU_OPTION_daemonize, \
     "-daemonize      daemonize QEMU after initializing\n", QEMU_ARCH_ALL)
diff --git a/savevm-async.c b/savevm-async.c
new file mode 100644
index 0000000000..5fcb56d373
--- /dev/null
+++ b/savevm-async.c
@@ -0,0 +1,523 @@
+#include "qemu/osdep.h"
+#include "qemu-common.h"
+#include "qapi/qmp/qerror.h"
+#include "qemu/error-report.h"
+#include "sysemu/sysemu.h"
+#include "qmp-commands.h"
+#include "qemu-options.h"
+#include "migration/qemu-file.h"
+#include "qom/qom-qobject.h"
+#include "migration/migration.h"
+#include "block/snapshot.h"
+#include "block/qapi.h"
+#include "block/block.h"
+#include "qemu/timer.h"
+#include "sysemu/block-backend.h"
+#include "qapi/qmp/qstring.h"
+#include "qemu/rcu.h"
+#include "qemu/thread.h"
+#include "qemu/cutils.h"
+
+/* #define DEBUG_SAVEVM_STATE */
+
+#define NOT_DONE 0x7fffffff /* used while emulated sync operation in progress */
+
+#ifdef DEBUG_SAVEVM_STATE
+#define DPRINTF(fmt, ...) \
+    do { printf("savevm-async: " fmt, ## __VA_ARGS__); } while (0)
+#else
+#define DPRINTF(fmt, ...) \
+    do { } while (0)
+#endif
+
+enum {
+    SAVE_STATE_DONE,
+    SAVE_STATE_ERROR,
+    SAVE_STATE_ACTIVE,
+    SAVE_STATE_COMPLETED,
+    SAVE_STATE_CANCELLED
+};
+
+
+static struct SnapshotState {
+    BlockBackend *target;
+    size_t bs_pos;
+    int state;
+    Error *error;
+    Error *blocker;
+    int saved_vm_running;
+    QEMUFile *file;
+    int64_t total_time;
+} snap_state;
+
+SaveVMInfo *qmp_query_savevm(Error **errp)
+{
+    SaveVMInfo *info = g_malloc0(sizeof(*info));
+    struct SnapshotState *s = &snap_state;
+
+    if (s->state != SAVE_STATE_DONE) {
+        info->has_bytes = true;
+        info->bytes = s->bs_pos;
+        switch (s->state) {
+        case SAVE_STATE_ERROR:
+            info->has_status = true;
+            info->status = g_strdup("failed");
+            info->has_total_time = true;
+            info->total_time = s->total_time;
+            if (s->error) {
+                info->has_error = true;
+                info->error = g_strdup(error_get_pretty(s->error));
+            }
+            break;
+        case SAVE_STATE_ACTIVE:
+            info->has_status = true;
+            info->status = g_strdup("active");
+            info->has_total_time = true;
+            info->total_time = qemu_clock_get_ms(QEMU_CLOCK_REALTIME)
+                - s->total_time;
+            break;
+        case SAVE_STATE_COMPLETED:
+            info->has_status = true;
+            info->status = g_strdup("completed");
+            info->has_total_time = true;
+            info->total_time = s->total_time;
+            break;
+        }
+    }
+
+    return info;
+}
+
+static int save_snapshot_cleanup(void)
+{
+    int ret = 0;
+
+    DPRINTF("save_snapshot_cleanup\n");
+
+    snap_state.total_time = qemu_clock_get_ms(QEMU_CLOCK_REALTIME) -
+        snap_state.total_time;
+
+    if (snap_state.file) {
+        ret = qemu_fclose(snap_state.file);
+    }
+
+    if (snap_state.target) {
+        /* try to truncate, but ignore errors (will fail on block devices).
+         * note: bdrv_read() need whole blocks, so we round up
+         */
+        size_t size = (snap_state.bs_pos + BDRV_SECTOR_SIZE) & BDRV_SECTOR_MASK;
+        blk_truncate(snap_state.target, size, NULL);
+        blk_op_unblock_all(snap_state.target, snap_state.blocker);
+        error_free(snap_state.blocker);
+        snap_state.blocker = NULL;
+        blk_unref(snap_state.target);
+        snap_state.target = NULL;
+    }
+
+    return ret;
+}
+
+static void save_snapshot_error(const char *fmt, ...)
+{
+    va_list ap;
+    char *msg;
+
+    va_start(ap, fmt);
+    msg = g_strdup_vprintf(fmt, ap);
+    va_end(ap);
+
+    DPRINTF("save_snapshot_error: %s\n", msg);
+
+    if (!snap_state.error) {
+        error_set(&snap_state.error, ERROR_CLASS_GENERIC_ERROR, "%s", msg);
+    }
+
+    g_free (msg);
+
+    snap_state.state = SAVE_STATE_ERROR;
+
+    save_snapshot_cleanup();
+}
+
+static void save_snapshot_completed(void)
+{
+    DPRINTF("save_snapshot_completed\n");
+
+    if (save_snapshot_cleanup() < 0) {
+        snap_state.state = SAVE_STATE_ERROR;
+    } else {
+        snap_state.state = SAVE_STATE_COMPLETED;
+    }
+}
+
+static int block_state_close(void *opaque)
+{
+    snap_state.file = NULL;
+    return blk_flush(snap_state.target);
+}
+
+static ssize_t block_state_writev_buffer(void *opaque, struct iovec *iov,
+                                         int iovcnt, int64_t pos)
+{
+    int ret;
+    QEMUIOVector qiov;
+
+    qemu_iovec_init_external(&qiov, iov, iovcnt);
+    ret = blk_co_pwritev(snap_state.target, pos, qiov.size, &qiov, 0);
+    if (ret < 0) {
+        return ret;
+    }
+    snap_state.bs_pos += qiov.size;
+    return qiov.size;
+}
+
+static int store_and_stop(void) {
+    if (global_state_store()) {
+        save_snapshot_error("Error saving global state");
+        return 1;
+    }
+    if (runstate_is_running()) {
+        vm_stop(RUN_STATE_SAVE_VM);
+    }
+    return 0;
+}
+
+static void process_savevm_co(void *opaque)
+{
+    int ret;
+    int64_t maxlen;
+    MigrationParams params = {
+        .blk = 0,
+        .shared = 0
+    };
+
+    snap_state.state = SAVE_STATE_ACTIVE;
+
+    qemu_mutex_unlock_iothread();
+    qemu_savevm_state_header(snap_state.file);
+    ret = qemu_savevm_state_begin(snap_state.file, &params);
+    qemu_mutex_lock_iothread();
+
+    if (ret < 0) {
+        save_snapshot_error("qemu_savevm_state_begin failed");
+        return;
+    }
+
+    while (snap_state.state == SAVE_STATE_ACTIVE) {
+        uint64_t pending_size, pend_post, pend_nonpost;
+
+        qemu_savevm_state_pending(snap_state.file, 0, &pend_nonpost, &pend_post);
+        pending_size = pend_post + pend_nonpost;
+
+        if (pending_size) {
+                ret = qemu_savevm_state_iterate(snap_state.file, false);
+                if (ret < 0) {
+                    save_snapshot_error("qemu_savevm_state_iterate error %d", ret);
+                    break;
+                }
+                DPRINTF("savevm inerate pending size %lu ret %d\n", pending_size, ret);
+        } else {
+            DPRINTF("done iterating\n");
+            if (store_and_stop())
+                break;
+            DPRINTF("savevm inerate finished\n");
+            qemu_savevm_state_complete_precopy(snap_state.file, false);
+            DPRINTF("save complete\n");
+            save_snapshot_completed();
+            break;
+        }
+
+        /* stop the VM if we get to the end of available space,
+         * or if pending_size is just a few MB
+         */
+        maxlen = blk_getlength(snap_state.target) - 30*1024*1024;
+        if ((pending_size < 100000) ||
+            ((snap_state.bs_pos + pending_size) >= maxlen)) {
+            if (store_and_stop())
+                break;
+        }
+    }
+
+    if(snap_state.state == SAVE_STATE_CANCELLED) {
+        save_snapshot_completed();
+        Error *errp = NULL;
+        qmp_savevm_end(&errp);
+    }
+
+}
+
+static const QEMUFileOps block_file_ops = {
+    .writev_buffer =  block_state_writev_buffer,
+    .close =          block_state_close,
+};
+
+
+void qmp_savevm_start(bool has_statefile, const char *statefile, Error **errp)
+{
+    Error *local_err = NULL;
+
+    int bdrv_oflags = BDRV_O_RDWR | BDRV_O_RESIZE | BDRV_O_NO_FLUSH;
+
+    if (snap_state.state != SAVE_STATE_DONE) {
+        error_set(errp, ERROR_CLASS_GENERIC_ERROR,
+                  "VM snapshot already started\n");
+        return;
+    }
+
+    /* initialize snapshot info */
+    snap_state.saved_vm_running = runstate_is_running();
+    snap_state.bs_pos = 0;
+    snap_state.total_time = qemu_clock_get_ms(QEMU_CLOCK_REALTIME);
+    snap_state.blocker = NULL;
+
+    if (snap_state.error) {
+        error_free(snap_state.error);
+        snap_state.error = NULL;
+    }
+
+    if (!has_statefile) {
+        vm_stop(RUN_STATE_SAVE_VM);
+        snap_state.state = SAVE_STATE_COMPLETED;
+        return;
+    }
+
+    if (qemu_savevm_state_blocked(errp)) {
+        return;
+    }
+
+    /* Open the image */
+    QDict *options = NULL;
+    options = qdict_new();
+    qdict_put(options, "driver", qstring_from_str("raw"));
+    snap_state.target = blk_new_open(statefile, NULL, options, bdrv_oflags, &local_err);
+    if (!snap_state.target) {
+        error_set(errp, ERROR_CLASS_GENERIC_ERROR, "failed to open '%s'", statefile);
+        goto restart;
+    }
+
+    snap_state.file = qemu_fopen_ops(&snap_state, &block_file_ops);
+
+    if (!snap_state.file) {
+        error_set(errp, ERROR_CLASS_GENERIC_ERROR, "failed to open '%s'", statefile);
+        goto restart;
+    }
+
+
+    error_setg(&snap_state.blocker, "block device is in use by savevm");
+    blk_op_block_all(snap_state.target, snap_state.blocker);
+
+    Coroutine *co = qemu_coroutine_create(process_savevm_co, NULL);
+    qemu_coroutine_enter(co);
+
+    return;
+
+restart:
+
+    save_snapshot_error("setup failed");
+
+    if (snap_state.saved_vm_running) {
+        vm_start();
+    }
+}
+
+void qmp_savevm_end(Error **errp)
+{
+    if (snap_state.state == SAVE_STATE_DONE) {
+        error_set(errp, ERROR_CLASS_GENERIC_ERROR,
+                  "VM snapshot not started\n");
+        return;
+    }
+
+    if (snap_state.state == SAVE_STATE_ACTIVE) {
+        snap_state.state = SAVE_STATE_CANCELLED;
+        return;
+    }
+
+    if (snap_state.saved_vm_running) {
+        vm_start();
+    }
+
+    snap_state.state = SAVE_STATE_DONE;
+}
+
+void qmp_snapshot_drive(const char *device, const char *name, Error **errp)
+{
+    BlockBackend *blk;
+    BlockDriverState *bs;
+    QEMUSnapshotInfo sn1, *sn = &sn1;
+    int ret;
+#ifdef _WIN32
+    struct _timeb tb;
+#else
+    struct timeval tv;
+#endif
+
+    if (snap_state.state != SAVE_STATE_COMPLETED) {
+        error_set(errp, ERROR_CLASS_GENERIC_ERROR,
+                  "VM snapshot not ready/started\n");
+        return;
+    }
+
+    blk = blk_by_name(device);
+    if (!blk) {
+        error_set(errp, ERROR_CLASS_DEVICE_NOT_FOUND,
+                  "Device '%s' not found", device);
+        return;
+    }
+
+    bs = blk_bs(blk);
+    if (!bdrv_is_inserted(bs)) {
+        error_setg(errp, QERR_DEVICE_HAS_NO_MEDIUM, device);
+        return;
+    }
+
+    if (bdrv_is_read_only(bs)) {
+        error_setg(errp, "Node '%s' is read only", device);
+        return;
+    }
+
+    if (!bdrv_can_snapshot(bs)) {
+        error_setg(errp, QERR_UNSUPPORTED);
+        return;
+    }
+
+    if (bdrv_snapshot_find(bs, sn, name) >= 0) {
+        error_set(errp, ERROR_CLASS_GENERIC_ERROR,
+                  "snapshot '%s' already exists", name);
+        return;
+    }
+
+    sn = &sn1;
+    memset(sn, 0, sizeof(*sn));
+
+#ifdef _WIN32
+    _ftime(&tb);
+    sn->date_sec = tb.time;
+    sn->date_nsec = tb.millitm * 1000000;
+#else
+    gettimeofday(&tv, NULL);
+    sn->date_sec = tv.tv_sec;
+    sn->date_nsec = tv.tv_usec * 1000;
+#endif
+    sn->vm_clock_nsec = qemu_clock_get_ns(QEMU_CLOCK_VIRTUAL);
+
+    pstrcpy(sn->name, sizeof(sn->name), name);
+
+    sn->vm_state_size = 0; /* do not save state */
+
+    ret = bdrv_snapshot_create(bs, sn);
+    if (ret < 0) {
+        error_set(errp, ERROR_CLASS_GENERIC_ERROR,
+                  "Error while creating snapshot on '%s'\n", device);
+        return;
+    }
+}
+
+void qmp_delete_drive_snapshot(const char *device, const char *name,
+                               Error **errp)
+{
+    BlockBackend *blk;
+    BlockDriverState *bs;
+    QEMUSnapshotInfo sn1, *sn = &sn1;
+    Error *local_err = NULL;
+
+    int ret;
+
+    blk = blk_by_name(device);
+    if (!blk) {
+        error_set(errp, ERROR_CLASS_DEVICE_NOT_FOUND,
+                  "Device '%s' not found", device);
+        return;
+    }
+
+    bs = blk_bs(blk);
+    if (bdrv_is_read_only(bs)) {
+        error_setg(errp, "Node '%s' is read only", device);
+        return;
+    }
+
+    if (!bdrv_can_snapshot(bs)) {
+        error_setg(errp, QERR_UNSUPPORTED);
+        return;
+    }
+
+    if (bdrv_snapshot_find(bs, sn, name) < 0) {
+        /* return success if snapshot does not exists */
+        return;
+    }
+
+    ret = bdrv_snapshot_delete(bs, NULL, name, &local_err);
+    if (ret < 0) {
+        error_set(errp, ERROR_CLASS_GENERIC_ERROR,
+                  "Error while deleting snapshot on '%s'\n", device);
+        return;
+    }
+}
+
+static ssize_t loadstate_get_buffer(void *opaque, uint8_t *buf, int64_t pos,
+                                    size_t size)
+{
+    BlockBackend *be = opaque;
+    int64_t maxlen = blk_getlength(be);
+    if (pos > maxlen) {
+        return -EIO;
+    }
+    if ((pos + size) > maxlen) {
+        size = maxlen - pos - 1;
+    }
+    if (size == 0) {
+        return 0;
+    }
+    return blk_pread(be, pos, buf, size);
+}
+
+static const QEMUFileOps loadstate_file_ops = {
+    .get_buffer = loadstate_get_buffer,
+};
+
+int load_state_from_blockdev(const char *filename)
+{
+    BlockBackend *be;
+    Error *local_err = NULL;
+    Error *blocker = NULL;
+
+    QEMUFile *f;
+    int ret = -EINVAL;
+
+    be = blk_new_open(filename, NULL, NULL, 0, &local_err);
+
+    if (!be) {
+        error_report("Could not open VM state file");
+        goto the_end;
+    }
+
+    error_setg(&blocker, "block device is in use by load state");
+    blk_op_block_all(be, blocker);
+
+    /* restore the VM state */
+    f = qemu_fopen_ops(be, &loadstate_file_ops);
+    if (!f) {
+        error_report("Could not open VM state file");
+        goto the_end;
+    }
+
+    qemu_system_reset(VMRESET_SILENT);
+    ret = qemu_loadvm_state(f);
+
+    qemu_fclose(f);
+    migration_incoming_state_destroy();
+    if (ret < 0) {
+        error_report("Error %d while loading VM state", ret);
+        goto the_end;
+    }
+
+    ret = 0;
+
+ the_end:
+    if (be) {
+        blk_op_unblock_all(be, blocker);
+        error_free(blocker);
+        blk_unref(be);
+    }
+    return ret;
+}
diff --git a/vl.c b/vl.c
index 46de1b9087..2132a77129 100644
--- a/vl.c
+++ b/vl.c
@@ -2960,6 +2960,7 @@ int main(int argc, char **argv, char **envp)
     int optind;
     const char *optarg;
     const char *loadvm = NULL;
+    const char *loadstate = NULL;
     MachineClass *machine_class;
     const char *cpu_model;
     const char *vga_model = NULL;
@@ -3635,6 +3636,9 @@ int main(int argc, char **argv, char **envp)
             case QEMU_OPTION_loadvm:
                 loadvm = optarg;
                 break;
+            case QEMU_OPTION_loadstate:
+                loadstate = optarg;
+                break;
             case QEMU_OPTION_full_screen:
                 full_screen = 1;
                 break;
@@ -4693,6 +4697,10 @@ int main(int argc, char **argv, char **envp)
         if (load_vmstate(loadvm) < 0) {
             autostart = 0;
         }
+    } else if (loadstate) {
+        if (load_state_from_blockdev(loadstate) < 0) {
+            autostart = 0;
+        }
     }
 
     qdev_prop_check_globals();
-- 
2.11.0