* Juan Quintela (quint...@redhat.com) wrote:
> Each time that we sync the bitmap, it is a possiblity that we receive
> a page that is being processed by a different thread. We fix this
> problem just making sure that we wait for all receiving threads to
> finish its work before we procedeed with the next stage.
>
> We are low on page flags, so we use a combination that is not valid to
> emit that message: MULTIFD_PAGE and COMPRESSED.
>
> I tried to make a migration command for it, but it don't work because
> we sync the bitmap sometimes when we have already sent the beggining
> of the section, so I just added a new page flag.
>
> Signed-off-by: Juan Quintela <quint...@redhat.com>
>
> --
> Create RAM_SAVE_FLAG_MULTIFD_SYNC (dave suggestion)
> Move the set of need_flush to inside the bitmap_sync code (peter suggestion)
> ---
> migration/ram.c | 55 +++++++++++++++++++++++++++++++++++++++++++++++++++++++
> 1 file changed, 55 insertions(+)
>
> diff --git a/migration/ram.c b/migration/ram.c
> index 20f3726909..b1ad7b2730 100644
> --- a/migration/ram.c
> +++ b/migration/ram.c
> @@ -74,6 +74,14 @@
> #define RAM_SAVE_FLAG_COMPRESS_PAGE 0x100
> #define RAM_SAVE_FLAG_MULTIFD_PAGE 0x200
>
> +/* We are getting low on pages flags, so we start using combinations
> + When we need to flush a page, we sent it as
> + RAM_SAVE_FLAG_MULTIFD_PAGE | RAM_SAVE_FLAG_ZERO
> + We don't allow that combination
> +*/
> +#define RAM_SAVE_FLAG_MULTIFD_SYNC \
> + (RAM_SAVE_FLAG_MULTIFD_PAGE | RAM_SAVE_FLAG_ZERO)
> +
> static inline bool is_zero_range(uint8_t *p, uint64_t size)
> {
> return buffer_is_zero(p, size);
> @@ -226,6 +234,9 @@ struct RAMState {
> uint64_t iterations_prev;
> /* Iterations since start */
> uint64_t iterations;
> + /* Indicates if we have synced the bitmap and we need to assure that
> + target has processeed all previous pages */
> + bool multifd_needs_flush;
> /* number of dirty bits in the bitmap */
> uint64_t migration_dirty_pages;
> /* protects modification of the bitmap */
> @@ -675,11 +686,13 @@ struct MultiFDRecvParams {
> QIOChannel *c;
> QemuSemaphore ready;
> QemuSemaphore sem;
> + QemuCond cond_sync;
> QemuMutex mutex;
> /* proteced by param mutex */
> bool quit;
> /* how many patches has sent this channel */
> uint32_t packets_recv;
> + bool sync;
> multifd_pages_t *pages;
> bool done;
> };
> @@ -734,6 +747,7 @@ int multifd_load_cleanup(Error **errp)
> qemu_thread_join(&p->thread);
> qemu_mutex_destroy(&p->mutex);
> qemu_sem_destroy(&p->sem);
> + qemu_cond_destroy(&p->cond_sync);
> socket_recv_channel_unref(p->c);
> g_free(p->name);
> p->name = NULL;
> @@ -779,6 +793,10 @@ static void *multifd_recv_thread(void *opaque)
> qemu_mutex_lock(&p->mutex);
> p->done = true;
> p->packets_recv++;
> + if (p->sync) {
> + qemu_cond_signal(&p->cond_sync);
> + p->sync = false;
> + }
> qemu_mutex_unlock(&p->mutex);
> qemu_sem_post(&p->ready);
> continue;
> @@ -845,9 +863,11 @@ void multifd_recv_new_channel(QIOChannel *ioc)
> qemu_mutex_init(&p->mutex);
> qemu_sem_init(&p->sem, 0);
> qemu_sem_init(&p->ready, 0);
> + qemu_cond_init(&p->cond_sync);
> p->quit = false;
> p->id = msg.id;
> p->done = false;
> + p->sync = false;
> multifd_pages_init(&p->pages, migrate_multifd_page_count());
> p->c = ioc;
> multifd_recv_state->count++;
> @@ -891,6 +911,27 @@ static void multifd_recv_page(RAMBlock *block,
> ram_addr_t offset,
> qemu_sem_post(&p->sem);
> }
>
> +static int multifd_flush(void)
> +{
> + int i, thread_count;
> +
> + if (!migrate_use_multifd()) {
> + return 0;
> + }
> + thread_count = migrate_multifd_channels();
> + for (i = 0; i < thread_count; i++) {
> + MultiFDRecvParams *p = &multifd_recv_state->params[i];
> +
> + qemu_mutex_lock(&p->mutex);
> + while (!p->done) {
> + p->sync = true;
> + qemu_cond_wait(&p->cond_sync, &p->mutex);
> + }
I think this can get stuck if there's an error at just the wrong time
in the recv_thread; perhaps if you make sure the
terminate_multifd_recev_threads wakes all the cond_sync's it would work?
Dave
> + qemu_mutex_unlock(&p->mutex);
> + }
> + return 0;
> +}
> +
> /**
> * save_page_header: write page header to wire
> *
> @@ -908,6 +949,12 @@ static size_t save_page_header(RAMState *rs, QEMUFile
> *f, RAMBlock *block,
> {
> size_t size, len;
>
> + if (rs->multifd_needs_flush &&
> + (offset & RAM_SAVE_FLAG_MULTIFD_PAGE)) {
> + offset |= RAM_SAVE_FLAG_ZERO;
> + rs->multifd_needs_flush = false;
> + }
> +
> if (block == rs->last_sent_block) {
> offset |= RAM_SAVE_FLAG_CONTINUE;
> }
> @@ -1196,6 +1243,9 @@ static void migration_bitmap_sync(RAMState *rs)
> if (migrate_use_events()) {
> qapi_event_send_migration_pass(ram_counters.dirty_sync_count, NULL);
> }
> + if (!rs->ram_bulk_stage && migrate_use_multifd()) {
> + rs->multifd_needs_flush = true;
> + }
> }
>
> /**
> @@ -3201,6 +3251,11 @@ static int ram_load(QEMUFile *f, void *opaque, int
> version_id)
> break;
> }
>
> + if ((flags & RAM_SAVE_FLAG_MULTIFD_SYNC)
> + == RAM_SAVE_FLAG_MULTIFD_SYNC) {
> + multifd_flush();
> + flags = flags & ~RAM_SAVE_FLAG_ZERO;
OK, so it's worth a comment saying that a MULTIFD_SYNC is not just a
sync, it's a sync and a page.
Dave
> + }
> if (flags & (RAM_SAVE_FLAG_ZERO | RAM_SAVE_FLAG_PAGE |
> RAM_SAVE_FLAG_COMPRESS_PAGE | RAM_SAVE_FLAG_XBZRLE |
> RAM_SAVE_FLAG_MULTIFD_PAGE)) {
> --
> 2.14.3
>
--
Dr. David Alan Gilbert / dgilb...@redhat.com / Manchester, UK
