On 2025-10-30 18:49, Peter Xu wrote:
> On Thu, Oct 30, 2025 at 10:49:08PM +0100, Juraj Marcin wrote:
> > From: Juraj Marcin <[email protected]>
> >
> > Currently, there are two functions that are responsible for calling the
> > cleanup of the incoming migration state. With successful precopy, it's
> > the incoming migration coroutine, and with successful postcopy it's the
> > postcopy listen thread. However, if postcopy fails during in the device
> > load, both functions will try to do the cleanup.
> >
> > This patch refactors all cleanup that needs to be done on the incoming
> > side into a common function and defines a clear boundary, who is
> > responsible for the cleanup. The incoming migration coroutine is
> > responsible for calling the cleanup function, unless the listen thread
> > has been started, in which case the postcopy listen thread runs the
> > incoming migration cleanup in its BH.
> >
> > Signed-off-by: Juraj Marcin <[email protected]>
> > ---
> > migration/migration.c | 44 +++++++++-------------------
> > migration/migration.h | 1 +
> > migration/postcopy-ram.c | 63 +++++++++++++++++++++-------------------
> > migration/trace-events | 2 +-
> > 4 files changed, 49 insertions(+), 61 deletions(-)
> >
> > diff --git a/migration/migration.c b/migration/migration.c
> > index 9a367f717e..637be71bfe 100644
> > --- a/migration/migration.c
> > +++ b/migration/migration.c
> > @@ -438,10 +438,15 @@ void
> > migration_incoming_transport_cleanup(MigrationIncomingState *mis)
> >
> > void migration_incoming_state_destroy(void)
> > {
> > - struct MigrationIncomingState *mis = migration_incoming_get_current();
> > + MigrationIncomingState *mis = migration_incoming_get_current();
> > + PostcopyState ps = postcopy_state_get();
> >
> > multifd_recv_cleanup();
> >
> > + if (ps != POSTCOPY_INCOMING_NONE) {
> > + postcopy_incoming_cleanup(mis);
> > + }
> > +
> > /*
> > * RAM state cleanup needs to happen after multifd cleanup, because
> > * multifd threads can use some of its states (receivedmap).
> > @@ -866,7 +871,6 @@ process_incoming_migration_co(void *opaque)
> > {
> > MigrationState *s = migrate_get_current();
> > MigrationIncomingState *mis = migration_incoming_get_current();
> > - PostcopyState ps;
> > int ret;
> > Error *local_err = NULL;
> >
> > @@ -883,25 +887,14 @@ process_incoming_migration_co(void *opaque)
> >
> > trace_vmstate_downtime_checkpoint("dst-precopy-loadvm-completed");
> >
> > - ps = postcopy_state_get();
> > - trace_process_incoming_migration_co_end(ret, ps);
> > - if (ps != POSTCOPY_INCOMING_NONE) {
> > - if (ps == POSTCOPY_INCOMING_ADVISE) {
> > - /*
> > - * Where a migration had postcopy enabled (and thus went to
> > advise)
> > - * but managed to complete within the precopy period, we can
> > use
> > - * the normal exit.
> > - */
> > - postcopy_incoming_cleanup(mis);
> > - } else if (ret >= 0) {
> > - /*
> > - * Postcopy was started, cleanup should happen at the end of
> > the
> > - * postcopy thread.
> > - */
> > - trace_process_incoming_migration_co_postcopy_end_main();
> > - goto out;
> > - }
> > - /* Else if something went wrong then just fall out of the normal
> > exit */
> > + trace_process_incoming_migration_co_end(ret);
> > + if (mis->have_listen_thread) {
> > + /*
> > + * Postcopy was started, cleanup should happen at the end of the
> > + * postcopy listen thread.
> > + */
> > + trace_process_incoming_migration_co_postcopy_end_main();
> > + goto out;
> > }
> >
> > if (ret < 0) {
> > @@ -933,15 +926,6 @@ fail:
> > }
> >
> > exit(EXIT_FAILURE);
> > - } else {
> > - /*
> > - * Report the error here in case that QEMU abruptly exits
> > - * when postcopy is enabled.
> > - */
> > - WITH_QEMU_LOCK_GUARD(&s->error_mutex) {
> > - error_report_err(s->error);
> > - s->error = NULL;
> > - }
>
> The patch looks all good itself. Here a pure question: is the old code
> wrong? If user sets exit_on_error=false, then this path seems to be
> releasing the error object, then query-migrate will see nothing?
I have tested this, and it is indeed what happens with the old code.
There is no "error-desc" in the response of "query-migrate" command,
just "status": "failed".
>
> > }
> > out:
> > /* Pairs with the refcount taken in qmp_migrate_incoming() */
> > diff --git a/migration/migration.h b/migration/migration.h
> > index 01329bf824..4a37f7202c 100644
> > --- a/migration/migration.h
> > +++ b/migration/migration.h
> > @@ -254,6 +254,7 @@ struct MigrationIncomingState {
> > MigrationIncomingState *migration_incoming_get_current(void);
> > void migration_incoming_state_destroy(void);
> > void migration_incoming_transport_cleanup(MigrationIncomingState *mis);
> > +void migration_incoming_qemu_exit(void);
> > /*
> > * Functions to work with blocktime context
> > */
> > diff --git a/migration/postcopy-ram.c b/migration/postcopy-ram.c
> > index b47c955763..48cbb46c27 100644
> > --- a/migration/postcopy-ram.c
> > +++ b/migration/postcopy-ram.c
> > @@ -2078,6 +2078,24 @@ bool postcopy_is_paused(MigrationStatus status)
> > status == MIGRATION_STATUS_POSTCOPY_RECOVER_SETUP;
> > }
> >
> > +static void postcopy_listen_thread_bh(void *opaque)
> > +{
> > + MigrationIncomingState *mis = migration_incoming_get_current();
> > +
> > + migration_incoming_state_destroy();
> > +
> > + if (mis->state == MIGRATION_STATUS_FAILED) {
> > + /*
> > + * If something went wrong then we have a bad state so exit;
> > + * we only could have gotten here if something failed before
> > + * POSTCOPY_INCOMING_RUNNING (for example device load), otherwise
> > + * postcopy migration would pause inside qemu_loadvm_state_main().
> > + * Failing dirty-bitmaps won't fail the whole migration.
> > + */
> > + exit(1);
> > + }
> > +}
> > +
> > /*
> > * Triggered by a postcopy_listen command; this thread takes over reading
> > * the input stream, leaving the main thread free to carry on loading the
> > rest
> > @@ -2131,53 +2149,38 @@ static void *postcopy_listen_thread(void *opaque)
> > "bitmaps are correctly migrated and valid.",
> > __func__, load_res, error_get_pretty(local_err));
> > g_clear_pointer(&local_err, error_free);
> > - load_res = 0; /* prevent further exit() */
> > } else {
> > + /*
> > + * Something went fatally wrong and we have a bad state, QEMU
> > will
> > + * exit depending on if postcopy-exit-on-error is true, but the
> > + * migration cannot be recovered.
> > + */
> > error_prepend(&local_err,
> > "loadvm failed during postcopy: %d: ", load_res);
> > migrate_set_error(migr, local_err);
> > g_clear_pointer(&local_err, error_report_err);
> > migrate_set_state(&mis->state,
> > MIGRATION_STATUS_POSTCOPY_ACTIVE,
> > MIGRATION_STATUS_FAILED);
> > + goto out;
> > }
> > }
> > - if (load_res >= 0) {
> > - /*
> > - * This looks good, but it's possible that the device loading in
> > the
> > - * main thread hasn't finished yet, and so we might not be in 'RUN'
> > - * state yet; wait for the end of the main thread.
> > - */
> > - qemu_event_wait(&mis->main_thread_load_event);
> > - }
> > - postcopy_incoming_cleanup(mis);
> > -
> > - if (load_res < 0) {
> > - /*
> > - * If something went wrong then we have a bad state so exit;
> > - * depending how far we got it might be possible at this point
> > - * to leave the guest running and fire MCEs for pages that never
> > - * arrived as a desperate recovery step.
> > - */
> > - rcu_unregister_thread();
> > - exit(EXIT_FAILURE);
> > - }
> > + /*
> > + * This looks good, but it's possible that the device loading in the
> > + * main thread hasn't finished yet, and so we might not be in 'RUN'
> > + * state yet; wait for the end of the main thread.
> > + */
> > + qemu_event_wait(&mis->main_thread_load_event);
> >
> > migrate_set_state(&mis->state, MIGRATION_STATUS_POSTCOPY_ACTIVE,
> > MIGRATION_STATUS_COMPLETED);
> > - /*
> > - * If everything has worked fine, then the main thread has waited
> > - * for us to start, and we're the last use of the mis.
> > - * (If something broke then qemu will have to exit anyway since it's
> > - * got a bad migration state).
> > - */
> > - bql_lock();
> > - migration_incoming_state_destroy();
> > - bql_unlock();
> >
> > +out:
> > rcu_unregister_thread();
> > mis->have_listen_thread = false;
> > postcopy_state_set(POSTCOPY_INCOMING_END);
> >
> > + migration_bh_schedule(postcopy_listen_thread_bh, NULL);
> > +
> > object_unref(OBJECT(migr));
> >
> > return NULL;
> > diff --git a/migration/trace-events b/migration/trace-events
> > index e8edd1fbba..772636f3ac 100644
> > --- a/migration/trace-events
> > +++ b/migration/trace-events
> > @@ -193,7 +193,7 @@ source_return_path_thread_resume_ack(uint32_t v)
> > "%"PRIu32
> > source_return_path_thread_switchover_acked(void) ""
> > migration_thread_low_pending(uint64_t pending) "%" PRIu64
> > migrate_transferred(uint64_t transferred, uint64_t time_spent, uint64_t
> > bandwidth, uint64_t avail_bw, uint64_t size) "transferred %" PRIu64 "
> > time_spent %" PRIu64 " bandwidth %" PRIu64 " switchover_bw %" PRIu64 "
> > max_size %" PRId64
> > -process_incoming_migration_co_end(int ret, int ps) "ret=%d
> > postcopy-state=%d"
> > +process_incoming_migration_co_end(int ret) "ret=%d"
> > process_incoming_migration_co_postcopy_end_main(void) ""
> > postcopy_preempt_enabled(bool value) "%d"
> > migration_precopy_complete(void) ""
> > --
> > 2.51.0
> >
>
> --
> Peter Xu
>
