[Qemu-devel] Re: [PATCH 0/6] Save state error handling (kill off no_migrate)

[Alex Williamson]

On Tue, 2010-11-09 at 10:44 -0700, Alex Williamson wrote:
> On Tue, 2010-11-09 at 18:49 +0200, Michael S. Tsirkin wrote:
> > On Tue, Nov 09, 2010 at 09:30:45AM -0700, Alex Williamson wrote:
> > > On Tue, 2010-11-09 at 18:15 +0200, Michael S. Tsirkin wrote:
> > > > On Tue, Nov 09, 2010 at 08:47:00AM -0700, Alex Williamson wrote:
> > > > > > > But it could.  What if ivshmem is acting in a peer role, but has 
> > > > > > > no
> > > > > > > clients, could it migrate?  What if ivshmem is migratable when the
> > > > > > > migration begins, but while the migration continues, a connection 
> > > > > > > is
> > > > > > > setup and it becomes unmigratable.
> > > > > > 
> > > > > > Sounds like something we should work to prevent, not support :)
> > > > > 
> > > > > s/:)/:(/  why?
> > > > 
> > > > It will just confuse everyone. Also if it happens after sending
> > > > all of memory, it's pretty painful.
> > > 
> > > It happens after sending all of memory with no_migrate, and I think
> > > pushing that earlier might introduce some races around when
> > > register_device_unmigratable() can be called.
> > 
> > Good point. I guess we could check it twice just to speed things up.
> > 
> > > > > > >  Using this series, ivshmem would
> > > > > > > have multiple options how to support this.  It could a) NAK the
> > > > > > > migration, b) drop connections and prevent new connections until 
> > > > > > > the
> > > > > > > migration finishes, c) detect that new connections have happened 
> > > > > > > since
> > > > > > > the migration started and cancel.  And probably more.  no_migrate 
> > > > > > > can
> > > > > > > only do a).  And in fact, we can only test no_migrate after the 
> > > > > > > VM is
> > > > > > > stopped (after all memory is migrated) because otherwise it could 
> > > > > > > race
> > > > > > > with devices setting no_migrate during migration.
> > > > > > 
> > > > > > We really want no_migrate to be static. changing it is abusing
> > > > > > the infrastructure.
> > > > > 
> > > > > You call it abusing, I call it making use of the infrastructure.  Why
> > > > > unnecessarily restrict ourselves?  Is return 0/-1 really that scary,
> > > > > unmaintainable, undebuggable?  I don't understand the resistance.
> > > > > 
> > > > > Alex
> > > > 
> > > > management really does not know how to handle unexpected
> > > > migration failures. They must be avoided.
> > > > 
> > > > There are some very special cases that fail migration. They are
> > > > currently easy to find with grep register_device_unmigratable.
> > > > I prefer to keep it that way.
> > > 
> > > How can management tools be improved to better handle unexpected
> > > migration failures when the only way for qemu to fail is an abort?
> > > We need the infrastructure to at least return an error first.  Do we just
> > > need to add some fprintfs to the save core to print the id string of the
> > > device that failed to save?  I just can't buy the "code is easier to
> > > grep" as an argument against adding better error handling to the save
> > > code path.
> > 
> > I just don't buy the 'we'll return meaningless error codes at random
> > point in time and management will figure it out' as an argument :)
> 
> Why is the error code meaningless?  The error code stops the migration
> in qemu and hopefully prints an error message (we could easily add an
> fprintf to the core save code to ensure we know the device responsible
> for the NAK).  From there we can figure out how to return the error to
> monitors and management tools, but we need to have a way to know there's
> an error first.
> 
> > >  Anyone else want to chime in?
> > > 
> > > Alex
> > 
> > Maybe try coding up some user using the new infrastructure to do
> > something useful, that register_device_unmigratable can't do.
> 
> With the number of people I hear complaining about how qemu has too many
> aborts, no error checking, and no way to return errors, I'm a little
> dumbfounded that there's such a roadblock to actually add some simple
> error handling.  Is it the error handling you're opposed to, or the way
> I'm using it to NAK a migration?

Would something like this in place of 6/6 ease your grep'ability and
debug'ability concerns?  Unfortunately it adds an assert, but if we
stomp on the vmsd, then the save state entry can't be unregistered.

Signed-off-by: Alex Williamson <address@hidden>
---

diff --git a/savevm.c b/savevm.c
index 521edc8..b0aaa46 100644
--- a/savevm.c
+++ b/savevm.c
@@ -1039,7 +1039,6 @@ typedef struct SaveStateEntry {
     const VMStateDescription *vmsd;
     void *opaque;
     CompatEntry *compat;
-    int no_migrate;
 } SaveStateEntry;
 
 
@@ -1103,7 +1102,6 @@ int register_savevm_live(DeviceState *dev,
     se->load_state = load_state;
     se->opaque = opaque;
     se->vmsd = NULL;
-    se->no_migrate = 0;
 
     if (dev && dev->parent_bus && dev->parent_bus->info->get_dev_path) {
         char *id = dev->parent_bus->info->get_dev_path(dev);
@@ -1170,6 +1168,22 @@ void unregister_savevm(DeviceState *dev, const char 
*idstr, void *opaque)
     }
 }
 
+static int nomigrate_set_params(int blk_enable, int shared, void *opaque)
+{
+    return -EFAULT;
+}
+
+static int nomigrate_save_live_state(Monitor *mon, QEMUFile *f, int stage,
+                                     void *opaque)
+{
+    return -EFAULT;
+}
+
+static int nomigrate_save_state(QEMUFile *f, void *opaque)
+{
+    return -EFAULT;
+}
+
 /* mark a device as not to be migrated, that is the device should be
    unplugged before migration */
 void register_device_unmigratable(DeviceState *dev, const char *idstr,
@@ -1190,7 +1204,10 @@ void register_device_unmigratable(DeviceState *dev, 
const char *idstr,
 
     QTAILQ_FOREACH(se, &savevm_handlers, entry) {
         if (strcmp(se->idstr, id) == 0 && se->opaque == opaque) {
-            se->no_migrate = 1;
+            se->set_params = nomigrate_set_params;
+            se->save_live_state = nomigrate_save_live_state;
+            se->save_state = nomigrate_save_state;
+            assert(!se->vmsd);
         }
     }
 }
@@ -1410,10 +1427,6 @@ static int vmstate_load(QEMUFile *f, SaveStateEntry *se, 
int version_id)
 
 static int vmstate_save(QEMUFile *f, SaveStateEntry *se)
 {
-    if (se->no_migrate) {
-        return -1;
-    }
-
     if (!se->vmsd) {         /* Old style */
         return se->save_state(f, se->opaque);
     }
@@ -1443,6 +1456,9 @@ int qemu_savevm_state_begin(Monitor *mon, QEMUFile *f, 
int blk_enable,
        }
        ret = se->set_params(blk_enable, shared, se->opaque);
         if (ret < 0) {
+            monitor_printf(mon,
+                           "Save state begin blocked by device '%s', error:"
+                           " %s\n", se->idstr, strerror(-ret));
             return ret;
         }
     }
@@ -1470,6 +1486,9 @@ int qemu_savevm_state_begin(Monitor *mon, QEMUFile *f, 
int blk_enable,
 
         ret = se->save_live_state(mon, f, QEMU_VM_SECTION_START, se->opaque);
         if (ret < 0) {
+            monitor_printf(mon,
+                           "Save state begin blocked by device '%s', error:"
+                           " %s\n", se->idstr, strerror(-ret));
             return ret;
         }
     }
@@ -1503,6 +1522,9 @@ int qemu_savevm_state_iterate(Monitor *mon, QEMUFile *f)
                synchronized over and over again. */
             break;
         } else if (ret < 0) {
+            monitor_printf(mon,
+                           "Save state iterate blocked by device '%s', error:"
+                           " %s\n", se->idstr, strerror(-ret));
             return ret;
         }
     }
@@ -1535,6 +1557,9 @@ int qemu_savevm_state_complete(Monitor *mon, QEMUFile *f)
 
         r = se->save_live_state(mon, f, QEMU_VM_SECTION_END, se->opaque);
         if (r < 0) {
+            monitor_printf(mon,
+                           "Save state complete blocked by device '%s', error:"
+                           " %s\n", se->idstr, strerror(-r));
             return r;
         }
     }
@@ -1559,7 +1584,9 @@ int qemu_savevm_state_complete(Monitor *mon, QEMUFile *f)
 
         r = vmstate_save(f, se);
         if (r < 0) {
-            monitor_printf(mon, "cannot migrate with device '%s'\n", 
se->idstr);
+            monitor_printf(mon,
+                           "Save state complete blocked by device '%s', error:"
+                           " %s\n", se->idstr, strerror(-r));
             return r;
         }
     }