diff mbox series

[1/2] qapi/migration.json: add reason to MIGRATION event

Message ID 20240215082659.1378342-2-rkhapov@yandex-team.ru (mailing list archive)
State New, archived
Headers show
Series Add field reason to event MIGRATION | expand

Commit Message

Roman Khapov Feb. 15, 2024, 8:26 a.m. UTC 
This commit adds the optional field reason for the events, which
contains the string, describing reason of status changing.
For example: reason of migration fail.

Function migrate_set_state now accepts 4th argument: the reason to
pass to event. Every call of this function appended with NULL argument.

Also migrate_set_state_err_reason was added to form reason from Error*

Signed-off-by: Roman Khapov <rkhapov@yandex-team.ru>
---
 migration/colo.c      |  6 +--
 migration/migration.c | 86 ++++++++++++++++++++++++++-----------------
 migration/migration.h |  5 ++-
 migration/multifd.c   |  6 +--
 migration/savevm.c    | 14 +++----
 qapi/migration.json   |  3 +-
 6 files changed, 71 insertions(+), 49 deletions(-)
diff mbox series

Patch

diff --git a/migration/colo.c b/migration/colo.c
index 315e31fe32..3c95291a83 100644
--- a/migration/colo.c
+++ b/migration/colo.c
@@ -102,7 +102,7 @@  static void secondary_vm_do_failover(void)
     }
 
     migrate_set_state(&mis->state, MIGRATION_STATUS_COLO,
-                      MIGRATION_STATUS_COMPLETED);
+                      MIGRATION_STATUS_COMPLETED, NULL);
 
     replication_stop_all(true, &local_err);
     if (local_err) {
@@ -157,7 +157,7 @@  static void primary_vm_do_failover(void)
     Error *local_err = NULL;
 
     migrate_set_state(&s->state, MIGRATION_STATUS_COLO,
-                      MIGRATION_STATUS_COMPLETED);
+                      MIGRATION_STATUS_COMPLETED, NULL);
     /*
      * kick COLO thread which might wait at
      * qemu_sem_wait(&s->colo_checkpoint_sem).
@@ -823,7 +823,7 @@  static void *colo_process_incoming_thread(void *opaque)
     qemu_sem_init(&mis->colo_incoming_sem, 0);
 
     migrate_set_state(&mis->state, MIGRATION_STATUS_ACTIVE,
-                      MIGRATION_STATUS_COLO);
+                      MIGRATION_STATUS_COLO, NULL);
 
     if (get_colo_mode() != COLO_MODE_SECONDARY) {
         error_report("COLO mode must be COLO_MODE_SECONDARY");
diff --git a/migration/migration.c b/migration/migration.c
index ab21de2cad..d28885a55b 100644
--- a/migration/migration.c
+++ b/migration/migration.c
@@ -354,10 +354,10 @@  void migration_incoming_state_destroy(void)
     yank_unregister_instance(MIGRATION_YANK_INSTANCE);
 }
 
-static void migrate_generate_event(int new_state)
+static void migrate_generate_event(int new_state, const char *reason)
 {
     if (migrate_events()) {
-        qapi_event_send_migration(new_state);
+        qapi_event_send_migration(new_state, reason);
     }
 }
 
@@ -598,7 +598,7 @@  static void qemu_start_incoming_migration(const char *uri, bool has_channels,
     }
 
     migrate_set_state(&mis->state, MIGRATION_STATUS_NONE,
-                      MIGRATION_STATUS_SETUP);
+                      MIGRATION_STATUS_SETUP, NULL);
 
     if (addr->transport == MIGRATION_ADDRESS_TYPE_SOCKET) {
         SocketAddress *saddr = &addr->u.socket;
@@ -692,7 +692,7 @@  static void process_incoming_migration_bh(void *opaque)
      * it's ready to use.
      */
     migrate_set_state(&mis->state, MIGRATION_STATUS_ACTIVE,
-                      MIGRATION_STATUS_COMPLETED);
+                      MIGRATION_STATUS_COMPLETED, NULL);
     migration_incoming_state_destroy();
 }
 
@@ -713,7 +713,7 @@  process_incoming_migration_co(void *opaque)
     mis->largest_page_size = qemu_ram_pagesize_largest();
     postcopy_state_set(POSTCOPY_INCOMING_NONE);
     migrate_set_state(&mis->state, MIGRATION_STATUS_SETUP,
-                      MIGRATION_STATUS_ACTIVE);
+                      MIGRATION_STATUS_ACTIVE, NULL);
 
     mis->loadvm_co = qemu_coroutine_self();
     ret = qemu_loadvm_state(mis->from_src_file);
@@ -762,7 +762,7 @@  process_incoming_migration_co(void *opaque)
     return;
 fail:
     migrate_set_state(&mis->state, MIGRATION_STATUS_ACTIVE,
-                      MIGRATION_STATUS_FAILED);
+                      MIGRATION_STATUS_FAILED, NULL);
     qemu_fclose(mis->from_src_file);
 
     multifd_recv_cleanup();
@@ -808,7 +808,7 @@  static bool postcopy_try_recover(void)
         mis->to_src_file = qemu_file_get_return_path(mis->from_src_file);
 
         migrate_set_state(&mis->state, MIGRATION_STATUS_POSTCOPY_PAUSED,
-                          MIGRATION_STATUS_POSTCOPY_RECOVER);
+                          MIGRATION_STATUS_POSTCOPY_RECOVER, NULL);
 
         /*
          * Here, we only wake up the main loading thread (while the
@@ -1308,15 +1308,27 @@  void qmp_migrate_start_postcopy(Error **errp)
 
 /* shared migration helpers */
 
-void migrate_set_state(int *state, int old_state, int new_state)
+void migrate_set_state(int *state, int old_state,
+                       int new_state, const char *reason)
 {
     assert(new_state < MIGRATION_STATUS__MAX);
     if (qatomic_cmpxchg(state, old_state, new_state) == old_state) {
         trace_migrate_set_state(MigrationStatus_str(new_state));
-        migrate_generate_event(new_state);
+        migrate_generate_event(new_state, reason);
     }
 }
 
+void migrate_set_state_err_reason(int *state, int old_state,
+                                  int new_state, const Error *err)
+{
+    const char *reason = NULL;
+    if (likely(err != NULL)) {
+        reason = error_get_pretty(err);
+    }
+
+    migrate_set_state(state, old_state, new_state, reason);
+}
+
 static void migrate_fd_cleanup(MigrationState *s)
 {
     g_free(s->hostname);
@@ -1360,7 +1372,7 @@  static void migrate_fd_cleanup(MigrationState *s)
 
     if (s->state == MIGRATION_STATUS_CANCELLING) {
         migrate_set_state(&s->state, MIGRATION_STATUS_CANCELLING,
-                          MIGRATION_STATUS_CANCELLED);
+                          MIGRATION_STATUS_CANCELLED, NULL);
     }
 
     if (s->error) {
@@ -1406,7 +1418,7 @@  static void migrate_fd_error(MigrationState *s, const Error *error)
     trace_migrate_fd_error(error_get_pretty(error));
     assert(s->to_dst_file == NULL);
     migrate_set_state(&s->state, MIGRATION_STATUS_SETUP,
-                      MIGRATION_STATUS_FAILED);
+                      MIGRATION_STATUS_FAILED, NULL);
     migrate_set_error(s, error);
 }
 
@@ -1432,7 +1444,8 @@  static void migrate_fd_cancel(MigrationState *s)
         if (old_state == MIGRATION_STATUS_PRE_SWITCHOVER) {
             qemu_sem_post(&s->pause_sem);
         }
-        migrate_set_state(&s->state, old_state, MIGRATION_STATUS_CANCELLING);
+        migrate_set_state(&s->state, old_state,
+                          MIGRATION_STATUS_CANCELLING, NULL);
     } while (s->state != MIGRATION_STATUS_CANCELLING);
 
     /*
@@ -1612,7 +1625,8 @@  int migrate_init(MigrationState *s, Error **errp)
     s->error = NULL;
     s->vmdesc = NULL;
 
-    migrate_set_state(&s->state, MIGRATION_STATUS_NONE, MIGRATION_STATUS_SETUP);
+    migrate_set_state(&s->state, MIGRATION_STATUS_NONE,
+                      MIGRATION_STATUS_SETUP, NULL);
 
     s->start_time = qemu_clock_get_ms(QEMU_CLOCK_REALTIME);
     s->total_time = 0;
@@ -2024,7 +2038,7 @@  void qmp_migrate(const char *uri, bool has_channels,
         error_setg(&local_err, QERR_INVALID_PARAMETER_VALUE, "uri",
                    "a valid migration protocol");
         migrate_set_state(&s->state, MIGRATION_STATUS_SETUP,
-                          MIGRATION_STATUS_FAILED);
+                          MIGRATION_STATUS_FAILED, NULL);
         block_cleanup_parameters();
     }
 
@@ -2145,7 +2159,7 @@  static bool migrate_handle_rp_resume_ack(MigrationState *s,
 
     /* Now both sides are active. */
     migrate_set_state(&s->state, MIGRATION_STATUS_POSTCOPY_RECOVER,
-                      MIGRATION_STATUS_POSTCOPY_ACTIVE);
+                      MIGRATION_STATUS_POSTCOPY_ACTIVE, NULL);
 
     /* Notify send thread that time to continue send pages */
     migration_rp_kick(s);
@@ -2422,14 +2436,15 @@  static int postcopy_start(MigrationState *ms, Error **errp)
     if (migrate_postcopy_preempt()) {
         migration_wait_main_channel(ms);
         if (postcopy_preempt_establish_channel(ms)) {
-            migrate_set_state(&ms->state, ms->state, MIGRATION_STATUS_FAILED);
+            migrate_set_state(&ms->state, ms->state,
+                              MIGRATION_STATUS_FAILED, NULL);
             return -1;
         }
     }
 
     if (!migrate_pause_before_switchover()) {
         migrate_set_state(&ms->state, MIGRATION_STATUS_ACTIVE,
-                          MIGRATION_STATUS_POSTCOPY_ACTIVE);
+                          MIGRATION_STATUS_POSTCOPY_ACTIVE, NULL);
     }
 
     trace_postcopy_start();
@@ -2559,7 +2574,7 @@  static int postcopy_start(MigrationState *ms, Error **errp)
     if (ret) {
         error_setg(errp, "postcopy_start: Migration stream errored");
         migrate_set_state(&ms->state, MIGRATION_STATUS_POSTCOPY_ACTIVE,
-                              MIGRATION_STATUS_FAILED);
+                          MIGRATION_STATUS_FAILED, NULL);
     }
 
     trace_postcopy_preempt_enabled(migrate_postcopy_preempt());
@@ -2570,7 +2585,7 @@  fail_closefb:
     qemu_fclose(fb);
 fail:
     migrate_set_state(&ms->state, MIGRATION_STATUS_POSTCOPY_ACTIVE,
-                          MIGRATION_STATUS_FAILED);
+                      MIGRATION_STATUS_FAILED, NULL);
     if (restart_block) {
         /* A failure happened early enough that we know the destination hasn't
          * accessed block devices, so we're safe to recover.
@@ -2618,10 +2633,10 @@  static int migration_maybe_pause(MigrationState *s,
     if (s->state != MIGRATION_STATUS_CANCELLING) {
         bql_unlock();
         migrate_set_state(&s->state, *current_active_state,
-                          MIGRATION_STATUS_PRE_SWITCHOVER);
+                          MIGRATION_STATUS_PRE_SWITCHOVER, NULL);
         qemu_sem_wait(&s->pause_sem);
         migrate_set_state(&s->state, MIGRATION_STATUS_PRE_SWITCHOVER,
-                          new_state);
+                          new_state, NULL);
         *current_active_state = new_state;
         bql_lock();
     }
@@ -2706,7 +2721,7 @@  static void migration_completion_failed(MigrationState *s,
     }
 
     migrate_set_state(&s->state, current_active_state,
-                      MIGRATION_STATUS_FAILED);
+                      MIGRATION_STATUS_FAILED, NULL);
 }
 
 /**
@@ -2744,10 +2759,10 @@  static void migration_completion(MigrationState *s)
     if (migrate_colo() && s->state == MIGRATION_STATUS_ACTIVE) {
         /* COLO does not support postcopy */
         migrate_set_state(&s->state, MIGRATION_STATUS_ACTIVE,
-                          MIGRATION_STATUS_COLO);
+                          MIGRATION_STATUS_COLO, NULL);
     } else {
         migrate_set_state(&s->state, current_active_state,
-                          MIGRATION_STATUS_COMPLETED);
+                          MIGRATION_STATUS_COMPLETED, NULL);
     }
 
     return;
@@ -2785,12 +2800,12 @@  static void bg_migration_completion(MigrationState *s)
     }
 
     migrate_set_state(&s->state, current_active_state,
-                      MIGRATION_STATUS_COMPLETED);
+                      MIGRATION_STATUS_COMPLETED, NULL);
     return;
 
 fail:
     migrate_set_state(&s->state, current_active_state,
-                      MIGRATION_STATUS_FAILED);
+                      MIGRATION_STATUS_FAILED, NULL);
 }
 
 typedef enum MigThrError {
@@ -2901,7 +2916,7 @@  static MigThrError postcopy_pause(MigrationState *s)
         close_return_path_on_source(s);
 
         migrate_set_state(&s->state, s->state,
-                          MIGRATION_STATUS_POSTCOPY_PAUSED);
+                          MIGRATION_STATUS_POSTCOPY_PAUSED, NULL);
 
         error_report("Detected IO failure for postcopy. "
                      "Migration paused.");
@@ -2979,7 +2994,7 @@  static MigThrError migration_detect_error(MigrationState *s)
          * For precopy (or postcopy with error outside IO), we fail
          * with no time.
          */
-        migrate_set_state(&s->state, state, MIGRATION_STATUS_FAILED);
+        migrate_set_state(&s->state, state, MIGRATION_STATUS_FAILED, NULL);
         trace_migration_thread_file_err();
 
         /* Time to stop the migration, now. */
@@ -3279,7 +3294,8 @@  static void qemu_savevm_wait_unplug(MigrationState *s, int old_state,
                                     int new_state)
 {
     if (qemu_savevm_state_guest_unplug_pending()) {
-        migrate_set_state(&s->state, old_state, MIGRATION_STATUS_WAIT_UNPLUG);
+        migrate_set_state(&s->state, old_state,
+                          MIGRATION_STATUS_WAIT_UNPLUG, NULL);
 
         while (s->state == MIGRATION_STATUS_WAIT_UNPLUG &&
                qemu_savevm_state_guest_unplug_pending()) {
@@ -3302,9 +3318,10 @@  static void qemu_savevm_wait_unplug(MigrationState *s, int old_state,
             }
         }
 
-        migrate_set_state(&s->state, MIGRATION_STATUS_WAIT_UNPLUG, new_state);
+        migrate_set_state(&s->state, MIGRATION_STATUS_WAIT_UNPLUG,
+                          new_state, NULL);
     } else {
-        migrate_set_state(&s->state, old_state, new_state);
+        migrate_set_state(&s->state, old_state, new_state, NULL);
     }
 }
 
@@ -3549,7 +3566,7 @@  static void *bg_migration_thread(void *opaque)
 fail:
     if (early_fail) {
         migrate_set_state(&s->state, MIGRATION_STATUS_ACTIVE,
-                MIGRATION_STATUS_FAILED);
+                          MIGRATION_STATUS_FAILED, NULL);
         bql_unlock();
     }
 
@@ -3615,7 +3632,8 @@  void migrate_fd_connect(MigrationState *s, Error *error_in)
     if (migrate_postcopy_ram() || migrate_return_path()) {
         if (open_return_path_on_source(s)) {
             error_setg(&local_err, "Unable to open return-path for postcopy");
-            migrate_set_state(&s->state, s->state, MIGRATION_STATUS_FAILED);
+            migrate_set_state(&s->state, s->state,
+                              MIGRATION_STATUS_FAILED, NULL);
             migrate_set_error(s, local_err);
             error_report_err(local_err);
             migrate_fd_cleanup(s);
@@ -3635,7 +3653,7 @@  void migrate_fd_connect(MigrationState *s, Error *error_in)
     if (resume) {
         /* Wakeup the main migration thread to do the recovery */
         migrate_set_state(&s->state, MIGRATION_STATUS_POSTCOPY_PAUSED,
-                          MIGRATION_STATUS_POSTCOPY_RECOVER);
+                          MIGRATION_STATUS_POSTCOPY_RECOVER, NULL);
         qemu_sem_post(&s->postcopy_pause_sem);
         return;
     }
diff --git a/migration/migration.h b/migration/migration.h
index f2c8b8f286..9150478654 100644
--- a/migration/migration.h
+++ b/migration/migration.h
@@ -468,7 +468,10 @@  struct MigrationState {
     bool rdma_migration;
 };
 
-void migrate_set_state(int *state, int old_state, int new_state);
+void migrate_set_state(int *state, int old_state,
+                       int new_state, const char *reason);
+void migrate_set_state_err_reason(int *state, int old_state,
+                                  int new_state, const Error *err);
 
 void migration_fd_process_incoming(QEMUFile *f);
 void migration_ioc_process_incoming(QIOChannel *ioc, Error **errp);
diff --git a/migration/multifd.c b/migration/multifd.c
index adfe8c9a0a..da3d397642 100644
--- a/migration/multifd.c
+++ b/migration/multifd.c
@@ -595,7 +595,7 @@  static void multifd_send_set_error(Error *err)
             s->state == MIGRATION_STATUS_DEVICE ||
             s->state == MIGRATION_STATUS_ACTIVE) {
             migrate_set_state(&s->state, s->state,
-                              MIGRATION_STATUS_FAILED);
+                              MIGRATION_STATUS_FAILED, NULL);
         }
     }
 }
@@ -1056,7 +1056,7 @@  bool multifd_send_setup(void)
         migrate_set_error(s, local_err);
         error_report_err(local_err);
         migrate_set_state(&s->state, MIGRATION_STATUS_SETUP,
-                          MIGRATION_STATUS_FAILED);
+                          MIGRATION_STATUS_FAILED, NULL);
         return false;
     }
 
@@ -1087,7 +1087,7 @@  static void multifd_recv_terminate_threads(Error *err)
         if (s->state == MIGRATION_STATUS_SETUP ||
             s->state == MIGRATION_STATUS_ACTIVE) {
             migrate_set_state(&s->state, s->state,
-                              MIGRATION_STATUS_FAILED);
+                              MIGRATION_STATUS_FAILED, NULL);
         }
     }
 
diff --git a/migration/savevm.c b/migration/savevm.c
index d612c8a902..be6cce8a51 100644
--- a/migration/savevm.c
+++ b/migration/savevm.c
@@ -1744,7 +1744,7 @@  static int qemu_savevm_state(QEMUFile *f, Error **errp)
     } else {
         status = MIGRATION_STATUS_COMPLETED;
     }
-    migrate_set_state(&ms->state, MIGRATION_STATUS_SETUP, status);
+    migrate_set_state(&ms->state, MIGRATION_STATUS_SETUP, status, NULL);
 
     /* f is outer parameter, it should not stay in global migration state after
      * this function finished */
@@ -1999,7 +1999,7 @@  static void *postcopy_ram_listen_thread(void *opaque)
     object_ref(OBJECT(migr));
 
     migrate_set_state(&mis->state, MIGRATION_STATUS_ACTIVE,
-                                   MIGRATION_STATUS_POSTCOPY_ACTIVE);
+                      MIGRATION_STATUS_POSTCOPY_ACTIVE, NULL);
     qemu_sem_post(&mis->thread_sync_sem);
     trace_postcopy_ram_listen_thread_start();
 
@@ -2037,7 +2037,7 @@  static void *postcopy_ram_listen_thread(void *opaque)
         } else {
             error_report("%s: loadvm failed: %d", __func__, load_res);
             migrate_set_state(&mis->state, MIGRATION_STATUS_POSTCOPY_ACTIVE,
-                                           MIGRATION_STATUS_FAILED);
+                              MIGRATION_STATUS_FAILED, NULL);
         }
     }
     if (load_res >= 0) {
@@ -2062,7 +2062,7 @@  static void *postcopy_ram_listen_thread(void *opaque)
     }
 
     migrate_set_state(&mis->state, MIGRATION_STATUS_POSTCOPY_ACTIVE,
-                                   MIGRATION_STATUS_COMPLETED);
+                      MIGRATION_STATUS_COMPLETED, NULL);
     /*
      * If everything has worked fine, then the main thread has waited
      * for us to start, and we're the last use of the mis.
@@ -2257,7 +2257,7 @@  static int loadvm_postcopy_handle_resume(MigrationIncomingState *mis)
      * This means source VM is ready to resume the postcopy migration.
      */
     migrate_set_state(&mis->state, MIGRATION_STATUS_POSTCOPY_RECOVER,
-                      MIGRATION_STATUS_POSTCOPY_ACTIVE);
+                      MIGRATION_STATUS_POSTCOPY_ACTIVE, NULL);
 
     trace_loadvm_postcopy_handle_resume();
 
@@ -2818,8 +2818,8 @@  static bool postcopy_pause_incoming(MigrationIncomingState *mis)
     }
 
     /* Current state can be either ACTIVE or RECOVER */
-    migrate_set_state(&mis->state, mis->state,
-                      MIGRATION_STATUS_POSTCOPY_PAUSED);
+    migrate_set_state(&mis->state, MIGRATION_STATUS_POSTCOPY_ACTIVE,
+                      MIGRATION_STATUS_POSTCOPY_PAUSED, NULL);
 
     /* Notify the fault thread for the invalidated file handle */
     postcopy_fault_thread_notify(mis);
diff --git a/qapi/migration.json b/qapi/migration.json
index 5a565d9b8d..33bb5b7f50 100644
--- a/qapi/migration.json
+++ b/qapi/migration.json
@@ -1392,6 +1392,7 @@ 
 # Emitted when a migration event happens
 #
 # @status: @MigrationStatus describing the current migration status.
+# @reason: Optional description of status changing reason.
 #
 # Since: 2.4
 #
@@ -1402,7 +1403,7 @@ 
 #     "data": {"status": "completed"} }
 ##
 { 'event': 'MIGRATION',
-  'data': {'status': 'MigrationStatus'}}
+  'data': {'status': 'MigrationStatus', '*reason': 'str'}}
 
 ##
 # @MIGRATION_PASS: