Merge remote-tracking branch 'remotes/dgilbert/tags/pull-migration-20200601a' into staging

Migration/virtio/hmp pull 2020-06-01

A mixed pull with:
  - RDMA migration fix (CID 1428762)
  - HMP qom-get addition and qom-set cleanup
  - a virtiofsd fix
  - COLO fixes

Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>

# gpg: Signature made Mon 01 Jun 2020 19:37:15 BST
# gpg:                using RSA key 45F5C71B4A0CB7FB977A9FA90516331EBC5BFDE7
# gpg: Good signature from "Dr. David Alan Gilbert (RH2) <dgilbert@redhat.com>" [full]
# Primary key fingerprint: 45F5 C71B 4A0C B7FB 977A  9FA9 0516 331E BC5B FDE7

* remotes/dgilbert/tags/pull-migration-20200601a:
  migration/migration.c: Fix hang in ram_save_host_page
  migration/colo.c: Move colo_notify_compares_event to the right place
  migration/colo.c: Relaunch failover even if there was an error
  migration/colo.c: Flush ram cache only after receiving device state
  migration/colo.c: Use cpu_synchronize_all_states()
  migration/colo.c: Use event instead of semaphore
  migration/vmstate: Remove unnecessary MemoryRegion forward declaration
  virtiofsd: remove symlink fallbacks
  hmp: Simplify qom-set
  hmp: Implement qom-get HMP command
  migration/rdma: cleanup rdma context before g_free to avoid memleaks
  migration/rdma: fix potential nullptr access in rdma_start_incoming_migration

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

12 files changed, 86 insertions, 207 deletions

diff --git a/hmp-commands.hx b/hmp-commands.hx
index 7f0f397..2825620 100644
--- a/hmp-commands.hx
+++ b/hmp-commands.hx
@@ -1791,8 +1791,22 @@ SRST
 ERST
 
     {
+        .name       = "qom-get",
+        .args_type  = "path:s,property:s",
+        .params     = "path property",
+        .help       = "print QOM property",
+        .cmd        = hmp_qom_get,
+        .flags      = "p",
+    },
+
+SRST
+``qom-get`` *path* *property*
+  Print QOM property *property* of object at location *path*
+ERST
+
+    {
         .name       = "qom-set",
-        .args_type  = "path:s,property:s,value:s",
+        .args_type  = "path:s,property:s,value:S",
         .params     = "path property value",
         .help       = "set QOM property",
         .cmd        = hmp_qom_set,
diff --git a/include/migration/vmstate.h b/include/migration/vmstate.h
index 3066763..eafa39f 100644
--- a/include/migration/vmstate.h
+++ b/include/migration/vmstate.h
@@ -1199,7 +1199,6 @@ static inline int vmstate_register(VMStateIf *obj, int instance_id,
 void vmstate_unregister(VMStateIf *obj, const VMStateDescription *vmsd,
                         void *opaque);
 
-struct MemoryRegion;
 void vmstate_register_ram(struct MemoryRegion *memory, DeviceState *dev);
 void vmstate_unregister_ram(struct MemoryRegion *memory, DeviceState *dev);
 void vmstate_register_ram_global(struct MemoryRegion *memory);
diff --git a/include/monitor/hmp.h b/include/monitor/hmp.h
index e33ca5a..c986cfd 100644
--- a/include/monitor/hmp.h
+++ b/include/monitor/hmp.h
@@ -96,6 +96,7 @@ void hmp_info_memdev(Monitor *mon, const QDict *qdict);
 void hmp_info_numa(Monitor *mon, const QDict *qdict);
 void hmp_info_memory_devices(Monitor *mon, const QDict *qdict);
 void hmp_qom_list(Monitor *mon, const QDict *qdict);
+void hmp_qom_get(Monitor *mon, const QDict *qdict);
 void hmp_qom_set(Monitor *mon, const QDict *qdict);
 void hmp_info_qom_tree(Monitor *mon, const QDict *dict);
 void object_add_completion(ReadLineState *rs, int nb_args, const char *str);
diff --git a/migration/colo.c b/migration/colo.c
index d015d4f..ea7d1e9 100644
--- a/migration/colo.c
+++ b/migration/colo.c
@@ -436,11 +436,6 @@ static int colo_do_checkpoint_transaction(MigrationState *s,
         goto out;
     }
 
-    colo_notify_compares_event(NULL, COLO_EVENT_CHECKPOINT, &local_err);
-    if (local_err) {
-        goto out;
-    }
-
     /* Disable block migration */
     migrate_set_block_enabled(false, &local_err);
     if (local_err) {
@@ -502,6 +497,12 @@ static int colo_do_checkpoint_transaction(MigrationState *s,
         goto out;
     }
 
+    qemu_event_reset(&s->colo_checkpoint_event);
+    colo_notify_compares_event(NULL, COLO_EVENT_CHECKPOINT, &local_err);
+    if (local_err) {
+        goto out;
+    }
+
     colo_receive_check_message(s->rp_state.from_dst_file,
                        COLO_MESSAGE_VMSTATE_LOADED, &local_err);
     if (local_err) {
@@ -589,7 +590,7 @@ static void colo_process_checkpoint(MigrationState *s)
             goto out;
         }
 
-        qemu_sem_wait(&s->colo_checkpoint_sem);
+        qemu_event_wait(&s->colo_checkpoint_event);
 
         if (s->state != MIGRATION_STATUS_COLO) {
             goto out;
@@ -637,7 +638,7 @@ out:
     colo_compare_unregister_notifier(&packets_compare_notifier);
     timer_del(s->colo_delay_timer);
     timer_free(s->colo_delay_timer);
-    qemu_sem_destroy(&s->colo_checkpoint_sem);
+    qemu_event_destroy(&s->colo_checkpoint_event);
 
     /*
      * Must be called after failover BH is completed,
@@ -654,7 +655,7 @@ void colo_checkpoint_notify(void *opaque)
     MigrationState *s = opaque;
     int64_t next_notify_time;
 
-    qemu_sem_post(&s->colo_checkpoint_sem);
+    qemu_event_set(&s->colo_checkpoint_event);
     s->colo_checkpoint_time = qemu_clock_get_ms(QEMU_CLOCK_HOST);
     next_notify_time = s->colo_checkpoint_time +
                     s->parameters.x_checkpoint_delay;
@@ -664,7 +665,7 @@ void colo_checkpoint_notify(void *opaque)
 void migrate_start_colo_process(MigrationState *s)
 {
     qemu_mutex_unlock_iothread();
-    qemu_sem_init(&s->colo_checkpoint_sem, 0);
+    qemu_event_init(&s->colo_checkpoint_event, false);
     s->colo_delay_timer =  timer_new_ms(QEMU_CLOCK_HOST,
                                 colo_checkpoint_notify, s);
 
@@ -704,7 +705,7 @@ static void colo_incoming_process_checkpoint(MigrationIncomingState *mis,
     }
 
     qemu_mutex_lock_iothread();
-    cpu_synchronize_all_pre_loadvm();
+    cpu_synchronize_all_states();
     ret = qemu_loadvm_state_main(mis->from_src_file, mis);
     qemu_mutex_unlock_iothread();
 
@@ -747,9 +748,11 @@ static void colo_incoming_process_checkpoint(MigrationIncomingState *mis,
 
     qemu_mutex_lock_iothread();
     vmstate_loading = true;
+    colo_flush_ram_cache();
     ret = qemu_load_device_state(fb);
     if (ret < 0) {
         error_setg(errp, "COLO: load device state failed");
+        vmstate_loading = false;
         qemu_mutex_unlock_iothread();
         return;
     }
@@ -758,6 +761,7 @@ static void colo_incoming_process_checkpoint(MigrationIncomingState *mis,
     replication_get_error_all(&local_err);
     if (local_err) {
         error_propagate(errp, local_err);
+        vmstate_loading = false;
         qemu_mutex_unlock_iothread();
         return;
     }
@@ -766,6 +770,7 @@ static void colo_incoming_process_checkpoint(MigrationIncomingState *mis,
     replication_do_checkpoint_all(&local_err);
     if (local_err) {
         error_propagate(errp, local_err);
+        vmstate_loading = false;
         qemu_mutex_unlock_iothread();
         return;
     }
@@ -777,6 +782,7 @@ static void colo_incoming_process_checkpoint(MigrationIncomingState *mis,
 
     if (local_err) {
         error_propagate(errp, local_err);
+        vmstate_loading = false;
         qemu_mutex_unlock_iothread();
         return;
     }
@@ -787,9 +793,6 @@ static void colo_incoming_process_checkpoint(MigrationIncomingState *mis,
     qemu_mutex_unlock_iothread();
 
     if (failover_get_state() == FAILOVER_STATUS_RELAUNCH) {
-        failover_set_state(FAILOVER_STATUS_RELAUNCH,
-                        FAILOVER_STATUS_NONE);
-        failover_request_active(NULL);
         return;
     }
 
@@ -888,6 +891,14 @@ void *colo_process_incoming_thread(void *opaque)
             error_report_err(local_err);
             break;
         }
+
+        if (failover_get_state() == FAILOVER_STATUS_RELAUNCH) {
+            failover_set_state(FAILOVER_STATUS_RELAUNCH,
+                            FAILOVER_STATUS_NONE);
+            failover_request_active(NULL);
+            break;
+        }
+
         if (failover_get_state() != FAILOVER_STATUS_NONE) {
             error_report("failover request");
             break;
@@ -895,8 +906,6 @@ void *colo_process_incoming_thread(void *opaque)
     }
 
 out:
-    vmstate_loading = false;
-
     /*
      * There are only two reasons we can get here, some error happened
      * or the user triggered failover.
diff --git a/migration/migration.c b/migration/migration.c
index 0bb042a..b63ad91 100644
--- a/migration/migration.c
+++ b/migration/migration.c
@@ -3361,6 +3361,10 @@ bool migration_rate_limit(void)
     bool urgent = false;
     migration_update_counters(s, now);
     if (qemu_file_rate_limit(s->to_dst_file)) {
+
+        if (qemu_file_get_error(s->to_dst_file)) {
+            return false;
+        }
         /*
          * Wait for a delay to do rate limiting OR
          * something urgent to post the semaphore.
diff --git a/migration/migration.h b/migration/migration.h
index 507284e..f617960 100644
--- a/migration/migration.h
+++ b/migration/migration.h
@@ -215,8 +215,8 @@ struct MigrationState
     /* The semaphore is used to notify COLO thread that failover is finished */
     QemuSemaphore colo_exit_sem;
 
-    /* The semaphore is used to notify COLO thread to do checkpoint */
-    QemuSemaphore colo_checkpoint_sem;
+    /* The event is used to notify COLO thread to do checkpoint */
+    QemuEvent colo_checkpoint_event;
     int64_t colo_checkpoint_time;
     QEMUTimer *colo_delay_timer;
 
diff --git a/migration/ram.c b/migration/ram.c
index 859f835..41cc530 100644
--- a/migration/ram.c
+++ b/migration/ram.c
@@ -3360,7 +3360,7 @@ static bool postcopy_is_running(void)
  * Flush content of RAM cache into SVM's memory.
  * Only flush the pages that be dirtied by PVM or SVM or both.
  */
-static void colo_flush_ram_cache(void)
+void colo_flush_ram_cache(void)
 {
     RAMBlock *block = NULL;
     void *dst_host;
@@ -3632,9 +3632,6 @@ static int ram_load(QEMUFile *f, void *opaque, int version_id)
     }
     trace_ram_load_complete(ret, seq_iter);
 
-    if (!ret  && migration_incoming_in_colo_state()) {
-        colo_flush_ram_cache();
-    }
     return ret;
 }
 
diff --git a/migration/ram.h b/migration/ram.h
index 5ceaff7..2eeaacf 100644
--- a/migration/ram.h
+++ b/migration/ram.h
@@ -65,6 +65,7 @@ int ram_dirty_bitmap_reload(MigrationState *s, RAMBlock *rb);
 
 /* ram cache */
 int colo_init_ram_cache(void);
+void colo_flush_ram_cache(void);
 void colo_release_ram_cache(void);
 void colo_incoming_start_dirty_log(void);
 
diff --git a/migration/rdma.c b/migration/rdma.c
index 967fda5..ec45d33 100644
--- a/migration/rdma.c
+++ b/migration/rdma.c
@@ -4056,7 +4056,9 @@ void rdma_start_incoming_migration(const char *host_port, Error **errp)
     return;
 err:
     error_propagate(errp, local_err);
-    g_free(rdma->host);
+    if (rdma) {
+        g_free(rdma->host);
+    }
     g_free(rdma);
     g_free(rdma_return_path);
 }
@@ -4092,20 +4094,20 @@ void rdma_start_outgoing_migration(void *opaque,
         rdma_return_path = qemu_rdma_data_init(host_port, errp);
 
         if (rdma_return_path == NULL) {
-            goto err;
+            goto return_path_err;
         }
 
         ret = qemu_rdma_source_init(rdma_return_path,
             s->enabled_capabilities[MIGRATION_CAPABILITY_RDMA_PIN_ALL], errp);
 
         if (ret) {
-            goto err;
+            goto return_path_err;
         }
 
         ret = qemu_rdma_connect(rdma_return_path, errp);
 
         if (ret) {
-            goto err;
+            goto return_path_err;
         }
 
         rdma->return_path = rdma_return_path;
@@ -4118,6 +4120,8 @@ void rdma_start_outgoing_migration(void *opaque,
     s->to_dst_file = qemu_fopen_rdma(rdma, "wb");
     migrate_fd_connect(s, NULL);
     return;
+return_path_err:
+    qemu_rdma_cleanup(rdma);
 err:
     g_free(rdma);
     g_free(rdma_return_path);
diff --git a/qom/qom-hmp-cmds.c b/qom/qom-hmp-cmds.c
index cd08233..f704b69 100644
--- a/qom/qom-hmp-cmds.c
+++ b/qom/qom-hmp-cmds.c
@@ -12,6 +12,8 @@
 #include "qapi/error.h"
 #include "qapi/qapi-commands-qom.h"
 #include "qapi/qmp/qdict.h"
+#include "qapi/qmp/qjson.h"
+#include "qapi/qmp/qstring.h"
 #include "qom/object.h"
 
 void hmp_qom_list(Monitor *mon, const QDict *qdict)
@@ -46,19 +48,29 @@ void hmp_qom_set(Monitor *mon, const QDict *qdict)
     const char *property = qdict_get_str(qdict, "property");
     const char *value = qdict_get_str(qdict, "value");
     Error *err = NULL;
-    bool ambiguous = false;
-    Object *obj;
+    QObject *obj;
 
-    obj = object_resolve_path(path, &ambiguous);
-    if (obj == NULL) {
-        error_set(&err, ERROR_CLASS_DEVICE_NOT_FOUND,
-                  "Device '%s' not found", path);
-    } else {
-        if (ambiguous) {
-            monitor_printf(mon, "Warning: Path '%s' is ambiguous\n", path);
-        }
-        object_property_parse(obj, value, property, &err);
+    obj = qobject_from_json(value, &err);
+    if (err == NULL) {
+        qmp_qom_set(path, property, obj, &err);
+    }
+
+    hmp_handle_error(mon, err);
+}
+
+void hmp_qom_get(Monitor *mon, const QDict *qdict)
+{
+    const char *path = qdict_get_str(qdict, "path");
+    const char *property = qdict_get_str(qdict, "property");
+    Error *err = NULL;
+    QObject *obj = qmp_qom_get(path, property, &err);
+
+    if (err == NULL) {
+        QString *str = qobject_to_json_pretty(obj);
+        monitor_printf(mon, "%s\n", qstring_get_str(str));
+        qobject_unref(str);
     }
+
     hmp_handle_error(mon, err);
 }
 
diff --git a/tests/qtest/test-hmp.c b/tests/qtest/test-hmp.c
index f8aa5f9..b8b1271 100644
--- a/tests/qtest/test-hmp.c
+++ b/tests/qtest/test-hmp.c
@@ -61,6 +61,7 @@ static const char *hmp_cmds[] = {
     "p $pc + 8",
     "qom-list /",
     "qom-set /machine initrd test",
+    "qom-get /machine initrd",
     "screendump /dev/null",
     "sendkey x",
     "singlestep on",
diff --git a/tools/virtiofsd/passthrough_ll.c b/tools/virtiofsd/passthrough_ll.c
index 3ba1d90..2ce7c96 100644
--- a/tools/virtiofsd/passthrough_ll.c
+++ b/tools/virtiofsd/passthrough_ll.c
@@ -140,7 +140,6 @@ enum {
 struct lo_data {
     pthread_mutex_t mutex;
     int debug;
-    int norace;
     int writeback;
     int flock;
     int posix_lock;
@@ -176,7 +175,6 @@ static const struct fuse_opt lo_opts[] = {
     { "cache=none", offsetof(struct lo_data, cache), CACHE_NONE },
     { "cache=auto", offsetof(struct lo_data, cache), CACHE_AUTO },
     { "cache=always", offsetof(struct lo_data, cache), CACHE_ALWAYS },
-    { "norace", offsetof(struct lo_data, norace), 1 },
     { "readdirplus", offsetof(struct lo_data, readdirplus_set), 1 },
     { "no_readdirplus", offsetof(struct lo_data, readdirplus_clear), 1 },
     FUSE_OPT_END
@@ -592,136 +590,6 @@ static void lo_getattr(fuse_req_t req, fuse_ino_t ino,
     fuse_reply_attr(req, &buf, lo->timeout);
 }
 
-/*
- * Increments parent->nlookup and caller must release refcount using
- * lo_inode_put(&parent).
- */
-static int lo_parent_and_name(struct lo_data *lo, struct lo_inode *inode,
-                              char path[PATH_MAX], struct lo_inode **parent)
-{
-    char procname[64];
-    char *last;
-    struct stat stat;
-    struct lo_inode *p;
-    int retries = 2;
-    int res;
-
-retry:
-    sprintf(procname, "%i", inode->fd);
-
-    res = readlinkat(lo->proc_self_fd, procname, path, PATH_MAX);
-    if (res < 0) {
-        fuse_log(FUSE_LOG_WARNING, "%s: readlink failed: %m\n", __func__);
-        goto fail_noretry;
-    }
-
-    if (res >= PATH_MAX) {
-        fuse_log(FUSE_LOG_WARNING, "%s: readlink overflowed\n", __func__);
-        goto fail_noretry;
-    }
-    path[res] = '\0';
-
-    last = strrchr(path, '/');
-    if (last == NULL) {
-        /* Shouldn't happen */
-        fuse_log(
-            FUSE_LOG_WARNING,
-            "%s: INTERNAL ERROR: bad path read from proc\n", __func__);
-        goto fail_noretry;
-    }
-    if (last == path) {
-        p = &lo->root;
-        pthread_mutex_lock(&lo->mutex);
-        p->nlookup++;
-        g_atomic_int_inc(&p->refcount);
-        pthread_mutex_unlock(&lo->mutex);
-    } else {
-        *last = '\0';
-        res = fstatat(AT_FDCWD, last == path ? "/" : path, &stat, 0);
-        if (res == -1) {
-            if (!retries) {
-                fuse_log(FUSE_LOG_WARNING,
-                         "%s: failed to stat parent: %m\n", __func__);
-            }
-            goto fail;
-        }
-        p = lo_find(lo, &stat);
-        if (p == NULL) {
-            if (!retries) {
-                fuse_log(FUSE_LOG_WARNING,
-                         "%s: failed to find parent\n", __func__);
-            }
-            goto fail;
-        }
-    }
-    last++;
-    res = fstatat(p->fd, last, &stat, AT_SYMLINK_NOFOLLOW);
-    if (res == -1) {
-        if (!retries) {
-            fuse_log(FUSE_LOG_WARNING,
-                     "%s: failed to stat last\n", __func__);
-        }
-        goto fail_unref;
-    }
-    if (stat.st_dev != inode->key.dev || stat.st_ino != inode->key.ino) {
-        if (!retries) {
-            fuse_log(FUSE_LOG_WARNING,
-                     "%s: failed to match last\n", __func__);
-        }
-        goto fail_unref;
-    }
-    *parent = p;
-    memmove(path, last, strlen(last) + 1);
-
-    return 0;
-
-fail_unref:
-    unref_inode_lolocked(lo, p, 1);
-    lo_inode_put(lo, &p);
-fail:
-    if (retries) {
-        retries--;
-        goto retry;
-    }
-fail_noretry:
-    errno = EIO;
-    return -1;
-}
-
-static int utimensat_empty(struct lo_data *lo, struct lo_inode *inode,
-                           const struct timespec *tv)
-{
-    int res;
-    struct lo_inode *parent;
-    char path[PATH_MAX];
-
-    if (S_ISLNK(inode->filetype)) {
-        res = utimensat(inode->fd, "", tv, AT_EMPTY_PATH);
-        if (res == -1 && errno == EINVAL) {
-            /* Sorry, no race free way to set times on symlink. */
-            if (lo->norace) {
-                errno = EPERM;
-            } else {
-                goto fallback;
-            }
-        }
-        return res;
-    }
-    sprintf(path, "%i", inode->fd);
-
-    return utimensat(lo->proc_self_fd, path, tv, 0);
-
-fallback:
-    res = lo_parent_and_name(lo, inode, path, &parent);
-    if (res != -1) {
-        res = utimensat(parent->fd, path, tv, AT_SYMLINK_NOFOLLOW);
-        unref_inode_lolocked(lo, parent, 1);
-        lo_inode_put(lo, &parent);
-    }
-
-    return res;
-}
-
 static int lo_fi_fd(fuse_req_t req, struct fuse_file_info *fi)
 {
     struct lo_data *lo = lo_data(req);
@@ -828,7 +696,8 @@ static void lo_setattr(fuse_req_t req, fuse_ino_t ino, struct stat *attr,
         if (fi) {
             res = futimens(fd, tv);
         } else {
-            res = utimensat_empty(lo, inode, tv);
+            sprintf(procname, "%i", inode->fd);
+            res = utimensat(lo->proc_self_fd, procname, tv, 0);
         }
         if (res == -1) {
             goto out_err;
@@ -1129,41 +998,6 @@ static void lo_symlink(fuse_req_t req, const char *link, fuse_ino_t parent,
     lo_mknod_symlink(req, parent, name, S_IFLNK, 0, link);
 }
 
-static int linkat_empty_nofollow(struct lo_data *lo, struct lo_inode *inode,
-                                 int dfd, const char *name)
-{
-    int res;
-    struct lo_inode *parent;
-    char path[PATH_MAX];
-
-    if (S_ISLNK(inode->filetype)) {
-        res = linkat(inode->fd, "", dfd, name, AT_EMPTY_PATH);
-        if (res == -1 && (errno == ENOENT || errno == EINVAL)) {
-            /* Sorry, no race free way to hard-link a symlink. */
-            if (lo->norace) {
-                errno = EPERM;
-            } else {
-                goto fallback;
-            }
-        }
-        return res;
-    }
-
-    sprintf(path, "%i", inode->fd);
-
-    return linkat(lo->proc_self_fd, path, dfd, name, AT_SYMLINK_FOLLOW);
-
-fallback:
-    res = lo_parent_and_name(lo, inode, path, &parent);
-    if (res != -1) {
-        res = linkat(parent->fd, path, dfd, name, 0);
-        unref_inode_lolocked(lo, parent, 1);
-        lo_inode_put(lo, &parent);
-    }
-
-    return res;
-}
-
 static void lo_link(fuse_req_t req, fuse_ino_t ino, fuse_ino_t parent,
                     const char *name)
 {
@@ -1172,6 +1006,7 @@ static void lo_link(fuse_req_t req, fuse_ino_t ino, fuse_ino_t parent,
     struct lo_inode *parent_inode;
     struct lo_inode *inode;
     struct fuse_entry_param e;
+    char procname[64];
     int saverr;
 
     if (!is_safe_path_component(name)) {
@@ -1190,7 +1025,9 @@ static void lo_link(fuse_req_t req, fuse_ino_t ino, fuse_ino_t parent,
     e.attr_timeout = lo->timeout;
     e.entry_timeout = lo->timeout;
 
-    res = linkat_empty_nofollow(lo, inode, parent_inode->fd, name);
+    sprintf(procname, "%i", inode->fd);
+    res = linkat(lo->proc_self_fd, procname, parent_inode->fd, name,
+                 AT_SYMLINK_FOLLOW);
     if (res == -1) {
         goto out_err;
     }