[PATCH v3 00/21] Multifd Migration Compression

[PATCH v3 00/21] Multifd Migration Compression

[Juan Quintela]

[v3]
- rebased on top of upstream + previous multifd cancel series
- split multifd code into its own file (multifd.[ch])
- split zstd/zlib compression methods (multifd-zstd/zlib.c)
- use qemu module feauture to avoid ifdefs
  (my understanding is that zlib needs to be present, but
  we setup zstd only if it is not there or is disabled)
- multifd-method: none|zlib|zstd

  As far as I can see, there is no easy way to convince qapi that zstd
  option could/couldn't be there depending on compliation flags. I
  ended just checking in migrate_parameters_check() if it is enabled
  and giving an error message otherwise.

Questions:
- I am "reusing" the compress-level parameter for both zstd and zlib,
  but it poses a problem:
  * zlib values: 1-9 (default: 6?)
  * zstd values: 1-19 (default: 3)
So, what should I do:
  * create multifd-zstd-level and multifd-zlib-level (easier)
  * reuse compress-level, and change its maximum values depending on
    multifd-method
  * any other good option?

Please, review.

[v2] - rebase on top of previous arguments posted to the list -
introduces zlib compression - introduces zstd compression

Please help if you know anything about zstd/zlib compression.

This puts compression on top of multifd. Advantages about current
compression:

- We copy all pages in a single packet and then compress the whole
  thing.

- We reuse the compression stream for all the packets sent through the
  same channel.

- We can select nocomp/zlib/zstd levels of compression.

Please, review.

Juan Quintela (21):
  migration-test: Use g_free() instead of free()
  multifd: Make sure that we don't do any IO after an error
  qemu-file: Don't do IO after shutdown
  migration-test: Make sure that multifd and cancel works
  migration: Create migration_is_running()
  migration: Don't send data if we have stopped
  migration: Make multifd_save_setup() get an Error parameter
  migration: Make multifd_load_setup() get an Error parameter
  migration: Add multifd-compress parameter
  ram_addr: Split RAMBlock definition
  multifd: multifd_send_pages only needs the qemufile
  multifd: multifd_queue_page only needs the qemufile
  multifd: multifd_send_sync_main only needs the qemufile
  multifd: Use qemu_target_page_size()
  migration: Make checkpatch happy with comments
  migration: Add support for modules
  multifd: Split multifd code into its own file
  migration: Make no compression operations into its own structure
  migration: Add zlib compression multifd support
  configure: Enable test and libs for zstd
  migration: Add zstd compression multifd support

 MAINTAINERS                  |    1 +
 configure                    |   30 +
 hw/core/qdev-properties.c    |   13 +
 include/exec/ram_addr.h      |   40 +-
 include/exec/ramblock.h      |   64 ++
 include/hw/qdev-properties.h |    3 +
 include/qemu/module.h        |    2 +
 migration/Makefile.objs      |    3 +
 migration/migration.c        |   97 +++-
 migration/migration.h        |    4 +-
 migration/multifd-zlib.c     |  289 +++++++++
 migration/multifd-zstd.c     |  304 ++++++++++
 migration/multifd.c          | 1064 ++++++++++++++++++++++++++++++++++
 migration/multifd.h          |  185 ++++++
 migration/qemu-file.c        |   22 +-
 migration/ram.c              | 1006 +-------------------------------
 migration/ram.h              |    7 -
 migration/rdma.c             |    2 +-
 migration/savevm.c           |    4 +-
 monitor/hmp-cmds.c           |   13 +
 qapi/migration.json          |   30 +-
 tests/qtest/migration-test.c |  142 ++++-
 vl.c                         |    1 +
 23 files changed, 2266 insertions(+), 1060 deletions(-)
 create mode 100644 include/exec/ramblock.h
 create mode 100644 migration/multifd-zlib.c
 create mode 100644 migration/multifd-zstd.c
 create mode 100644 migration/multifd.c
 create mode 100644 migration/multifd.h

-- 
2.24.1

[PATCH v3 01/21] migration-test: Use g_free() instead of free()

[Juan Quintela]

Signed-off-by: Juan Quintela <quintela@redhat.com>
---
 tests/qtest/migration-test.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tests/qtest/migration-test.c b/tests/qtest/migration-test.c
index 26e2e77289..b6a74a05ce 100644
--- a/tests/qtest/migration-test.c
+++ b/tests/qtest/migration-test.c
@@ -1291,7 +1291,7 @@ static void test_multifd_tcp(void)
     wait_for_serial("dest_serial");
     wait_for_migration_complete(from);
     test_migrate_end(from, to, true);
-    free(uri);
+    g_free(uri);
 }
 
 int main(int argc, char **argv)
-- 
2.24.1

Re: [PATCH v3 01/21] migration-test: Use g_free() instead of free()

[Dr. David Alan Gilbert]

* Juan Quintela (quintela@redhat.com) wrote:
> Signed-off-by: Juan Quintela <quintela@redhat.com>

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

> ---
>  tests/qtest/migration-test.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/tests/qtest/migration-test.c b/tests/qtest/migration-test.c
> index 26e2e77289..b6a74a05ce 100644
> --- a/tests/qtest/migration-test.c
> +++ b/tests/qtest/migration-test.c
> @@ -1291,7 +1291,7 @@ static void test_multifd_tcp(void)
>      wait_for_serial("dest_serial");
>      wait_for_migration_complete(from);
>      test_migrate_end(from, to, true);
> -    free(uri);
> +    g_free(uri);
>  }
>  
>  int main(int argc, char **argv)
> -- 
> 2.24.1
> 
--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK

Re: [PATCH v3 01/21] migration-test: Use g_free() instead of free()

[Philippe Mathieu-Daudé]

On 1/23/20 12:58 PM, Juan Quintela wrote:
> Signed-off-by: Juan Quintela <quintela@redhat.com>

Nothing changed since v4 (apart it is now v3),
however it misses:

Fixes: b99784ef6c3
Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>

See:
https://www.mail-archive.com/qemu-devel@nongnu.org/msg672805.html
https://www.mail-archive.com/qemu-devel@nongnu.org/msg672853.html

> ---
>   tests/qtest/migration-test.c | 2 +-
>   1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/tests/qtest/migration-test.c b/tests/qtest/migration-test.c
> index 26e2e77289..b6a74a05ce 100644
> --- a/tests/qtest/migration-test.c
> +++ b/tests/qtest/migration-test.c
> @@ -1291,7 +1291,7 @@ static void test_multifd_tcp(void)
>       wait_for_serial("dest_serial");
>       wait_for_migration_complete(from);
>       test_migrate_end(from, to, true);
> -    free(uri);
> +    g_free(uri);
>   }
>   
>   int main(int argc, char **argv)
> 

Re: [PATCH v3 01/21] migration-test: Use g_free() instead of free()

[Daniel P. Berrangé]

On Thu, Jan 23, 2020 at 12:58:11PM +0100, Juan Quintela wrote:
> Signed-off-by: Juan Quintela <quintela@redhat.com>
> ---
>  tests/qtest/migration-test.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/tests/qtest/migration-test.c b/tests/qtest/migration-test.c
> index 26e2e77289..b6a74a05ce 100644
> --- a/tests/qtest/migration-test.c
> +++ b/tests/qtest/migration-test.c
> @@ -1291,7 +1291,7 @@ static void test_multifd_tcp(void)
>      wait_for_serial("dest_serial");
>      wait_for_migration_complete(from);
>      test_migrate_end(from, to, true);
> -    free(uri);
> +    g_free(uri);

Not an objection to this patch, just a general FYI.

Our min glib guarantees that g_malloc/g_free are always using the
system allocator. So using free() is not a correctness problem
these days.

In general I'd suggest eliminating both free() and g_free(), and instead
annotating the variable decl for automatic free. eg

  g_autofree char *uri = NULL;


Regards,
Daniel
-- 
|: https://berrange.com      -o-    https://www.flickr.com/photos/dberrange :|
|: https://libvirt.org         -o-            https://fstop138.berrange.com :|
|: https://entangle-photo.org    -o-    https://www.instagram.com/dberrange :|

Re: [PATCH v3 01/21] migration-test: Use g_free() instead of free()

[Juan Quintela]

Daniel P. Berrangé <berrange@redhat.com> wrote:
> On Thu, Jan 23, 2020 at 12:58:11PM +0100, Juan Quintela wrote:
>> Signed-off-by: Juan Quintela <quintela@redhat.com>
>> ---
>>  tests/qtest/migration-test.c | 2 +-
>>  1 file changed, 1 insertion(+), 1 deletion(-)
>> 
>> diff --git a/tests/qtest/migration-test.c b/tests/qtest/migration-test.c
>> index 26e2e77289..b6a74a05ce 100644
>> --- a/tests/qtest/migration-test.c
>> +++ b/tests/qtest/migration-test.c
>> @@ -1291,7 +1291,7 @@ static void test_multifd_tcp(void)
>>      wait_for_serial("dest_serial");
>>      wait_for_migration_complete(from);
>>      test_migrate_end(from, to, true);
>> -    free(uri);
>> +    g_free(uri);
>
> Not an objection to this patch, just a general FYI.
>
> Our min glib guarantees that g_malloc/g_free are always using the
> system allocator. So using free() is not a correctness problem
> these days.

Ok.  But the rest of the file uses g_malloc/g_free and friends O:-)

> In general I'd suggest eliminating both free() and g_free(), and instead
> annotating the variable decl for automatic free. eg
>
>   g_autofree char *uri = NULL;

I will investigate this, thanks.

Later, Juan.

[PATCH v3 02/21] multifd: Make sure that we don't do any IO after an error

[Juan Quintela]

Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
---
 migration/ram.c | 22 +++++++++++++---------
 1 file changed, 13 insertions(+), 9 deletions(-)

diff --git a/migration/ram.c b/migration/ram.c
index d2208b5534..f95d656c26 100644
--- a/migration/ram.c
+++ b/migration/ram.c
@@ -3445,7 +3445,7 @@ static int ram_save_iterate(QEMUFile *f, void *opaque)
 {
     RAMState **temp = opaque;
     RAMState *rs = *temp;
-    int ret;
+    int ret = 0;
     int i;
     int64_t t0;
     int done = 0;
@@ -3524,12 +3524,14 @@ static int ram_save_iterate(QEMUFile *f, void *opaque)
     ram_control_after_iterate(f, RAM_CONTROL_ROUND);
 
 out:
-    multifd_send_sync_main(rs);
-    qemu_put_be64(f, RAM_SAVE_FLAG_EOS);
-    qemu_fflush(f);
-    ram_counters.transferred += 8;
+    if (ret >= 0) {
+        multifd_send_sync_main(rs);
+        qemu_put_be64(f, RAM_SAVE_FLAG_EOS);
+        qemu_fflush(f);
+        ram_counters.transferred += 8;
 
-    ret = qemu_file_get_error(f);
+        ret = qemu_file_get_error(f);
+    }
     if (ret < 0) {
         return ret;
     }
@@ -3581,9 +3583,11 @@ static int ram_save_complete(QEMUFile *f, void *opaque)
         ram_control_after_iterate(f, RAM_CONTROL_FINISH);
     }
 
-    multifd_send_sync_main(rs);
-    qemu_put_be64(f, RAM_SAVE_FLAG_EOS);
-    qemu_fflush(f);
+    if (ret >= 0) {
+        multifd_send_sync_main(rs);
+        qemu_put_be64(f, RAM_SAVE_FLAG_EOS);
+        qemu_fflush(f);
+    }
 
     return ret;
 }
-- 
2.24.1

[PATCH v3 03/21] qemu-file: Don't do IO after shutdown

[Juan Quintela]

Be sure that we are not doing neither read/write after shutdown of the
QEMUFile.

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

---

Set error in case that there is none (dave)
---
 migration/qemu-file.c | 22 +++++++++++++++++++++-
 1 file changed, 21 insertions(+), 1 deletion(-)

diff --git a/migration/qemu-file.c b/migration/qemu-file.c
index 26fb25ddc1..bbb2b63927 100644
--- a/migration/qemu-file.c
+++ b/migration/qemu-file.c
@@ -53,6 +53,8 @@ struct QEMUFile {
 
     int last_error;
     Error *last_error_obj;
+    /* has the file has been shutdown */
+    bool shutdown;
 };
 
 /*
@@ -61,10 +63,18 @@ struct QEMUFile {
  */
 int qemu_file_shutdown(QEMUFile *f)
 {
+    int ret;
+
+    f->shutdown = true;
     if (!f->ops->shut_down) {
         return -ENOSYS;
     }
-    return f->ops->shut_down(f->opaque, true, true, NULL);
+    ret = f->ops->shut_down(f->opaque, true, true, NULL);
+
+    if (!f->last_error) {
+        qemu_file_set_error(f, -EIO);
+    }
+    return ret;
 }
 
 /*
@@ -214,6 +224,9 @@ void qemu_fflush(QEMUFile *f)
         return;
     }
 
+    if (f->shutdown) {
+        return;
+    }
     if (f->iovcnt > 0) {
         expect = iov_size(f->iov, f->iovcnt);
         ret = f->ops->writev_buffer(f->opaque, f->iov, f->iovcnt, f->pos,
@@ -328,6 +341,10 @@ static ssize_t qemu_fill_buffer(QEMUFile *f)
     f->buf_index = 0;
     f->buf_size = pending;
 
+    if (f->shutdown) {
+        return 0;
+    }
+
     len = f->ops->get_buffer(f->opaque, f->buf + pending, f->pos,
                              IO_BUF_SIZE - pending, &local_error);
     if (len > 0) {
@@ -642,6 +659,9 @@ int64_t qemu_ftell(QEMUFile *f)
 
 int qemu_file_rate_limit(QEMUFile *f)
 {
+    if (f->shutdown) {
+        return 1;
+    }
     if (qemu_file_get_error(f)) {
         return 1;
     }
-- 
2.24.1

[PATCH v3 04/21] migration-test: Make sure that multifd and cancel works

[Juan Quintela]

Test that this sequerce works:

- launch source
- launch target
- start migration
- cancel migration
- relaunch target
- do migration again

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

---

- Wait for 1st trhead to move to cancelled before launching second
  migration
- Add 'to2' parameter to diferentiate 1st and second target.
- Use g_free() instead of free()
---
 tests/qtest/migration-test.c | 112 ++++++++++++++++++++++++++++++++++-
 1 file changed, 111 insertions(+), 1 deletion(-)

diff --git a/tests/qtest/migration-test.c b/tests/qtest/migration-test.c
index b6a74a05ce..cf27ebbc9d 100644
--- a/tests/qtest/migration-test.c
+++ b/tests/qtest/migration-test.c
@@ -424,6 +424,14 @@ static void migrate_recover(QTestState *who, const char *uri)
     qobject_unref(rsp);
 }
 
+static void migrate_cancel(QTestState *who)
+{
+    QDict *rsp;
+
+    rsp = wait_command(who, "{ 'execute': 'migrate_cancel' }");
+    qobject_unref(rsp);
+}
+
 static void migrate_set_capability(QTestState *who, const char *capability,
                                    bool value)
 {
@@ -456,6 +464,8 @@ static void migrate_postcopy_start(QTestState *from, QTestState *to)
 typedef struct {
     bool hide_stderr;
     bool use_shmem;
+    /* only launch the target process */
+    bool only_target;
     char *opts_source;
     char *opts_target;
 } MigrateStart;
@@ -571,7 +581,9 @@ static int test_migrate_start(QTestState **from, QTestState **to,
                                  arch_source, shmem_opts, args->opts_source,
                                  ignore_stderr);
     g_free(arch_source);
-    *from = qtest_init(cmd_source);
+    if (!args->only_target) {
+        *from = qtest_init(cmd_source);
+    }
     g_free(cmd_source);
 
     cmd_target = g_strdup_printf("-accel kvm -accel tcg%s%s "
@@ -1294,6 +1306,103 @@ static void test_multifd_tcp(void)
     g_free(uri);
 }
 
+/*
+ * This test does:
+ *  source               target
+ *                       migrate_incoming
+ *     migrate
+ *     migrate_cancel
+ *                       launch another target
+ *     migrate
+ *
+ *  And see that it works
+ */
+
+static void test_multifd_tcp_cancel(void)
+{
+    MigrateStart *args = migrate_start_new();
+    QTestState *from, *to, *to2;
+    QDict *rsp;
+    char *uri;
+
+    args->hide_stderr = true;
+
+    if (test_migrate_start(&from, &to, "defer", args)) {
+        return;
+    }
+
+    /*
+     * We want to pick a speed slow enough that the test completes
+     * quickly, but that it doesn't complete precopy even on a slow
+     * machine, so also set the downtime.
+     */
+    /* 1 ms should make it not converge*/
+    migrate_set_parameter_int(from, "downtime-limit", 1);
+    /* 300MB/s */
+    migrate_set_parameter_int(from, "max-bandwidth", 30000000);
+
+    migrate_set_parameter_int(from, "multifd-channels", 16);
+    migrate_set_parameter_int(to, "multifd-channels", 16);
+
+    migrate_set_capability(from, "multifd", "true");
+    migrate_set_capability(to, "multifd", "true");
+
+    /* Start incoming migration from the 1st socket */
+    rsp = wait_command(to, "{ 'execute': 'migrate-incoming',"
+                           "  'arguments': { 'uri': 'tcp:127.0.0.1:0' }}");
+    qobject_unref(rsp);
+
+    /* Wait for the first serial output from the source */
+    wait_for_serial("src_serial");
+
+    uri = migrate_get_socket_address(to, "socket-address");
+
+    migrate_qmp(from, uri, "{}");
+
+    wait_for_migration_pass(from);
+
+    migrate_cancel(from);
+
+    args = migrate_start_new();
+    args->only_target = true;
+
+    if (test_migrate_start(&from, &to2, "defer", args)) {
+        return;
+    }
+
+    migrate_set_parameter_int(to2, "multifd-channels", 16);
+
+    migrate_set_capability(to2, "multifd", "true");
+
+    /* Start incoming migration from the 1st socket */
+    rsp = wait_command(to2, "{ 'execute': 'migrate-incoming',"
+                            "  'arguments': { 'uri': 'tcp:127.0.0.1:0' }}");
+    qobject_unref(rsp);
+
+    uri = migrate_get_socket_address(to2, "socket-address");
+
+    wait_for_migration_status(from, "cancelled", NULL);
+
+    /* 300ms it should converge */
+    migrate_set_parameter_int(from, "downtime-limit", 300);
+    /* 1GB/s */
+    migrate_set_parameter_int(from, "max-bandwidth", 1000000000);
+
+    migrate_qmp(from, uri, "{}");
+
+    wait_for_migration_pass(from);
+
+    if (!got_stop) {
+        qtest_qmp_eventwait(from, "STOP");
+    }
+    qtest_qmp_eventwait(to2, "RESUME");
+
+    wait_for_serial("dest_serial");
+    wait_for_migration_complete(from);
+    test_migrate_end(from, to2, true);
+    g_free(uri);
+}
+
 int main(int argc, char **argv)
 {
     char template[] = "/tmp/migration-test-XXXXXX";
@@ -1359,6 +1468,7 @@ int main(int argc, char **argv)
 
     qtest_add_func("/migration/auto_converge", test_migrate_auto_converge);
     qtest_add_func("/migration/multifd/tcp", test_multifd_tcp);
+    qtest_add_func("/migration/multifd/tcp/cancel", test_multifd_tcp_cancel);
 
     ret = g_test_run();
 
-- 
2.24.1

[PATCH v3 05/21] migration: Create migration_is_running()

[Juan Quintela]

This function returns if we are in the middle of a migration.
It is like migration_is_setup_or_active() with CANCELLING and COLO.
Adapt all calers that are needed.

Signed-off-by: Juan Quintela <quintela@redhat.com>
---
 migration/migration.c | 29 ++++++++++++++++++++++++-----
 migration/migration.h |  1 +
 migration/savevm.c    |  4 +---
 3 files changed, 26 insertions(+), 8 deletions(-)

diff --git a/migration/migration.c b/migration/migration.c
index 990bff00c0..1fb0aab44d 100644
--- a/migration/migration.c
+++ b/migration/migration.c
@@ -829,6 +829,27 @@ bool migration_is_setup_or_active(int state)
     }
 }
 
+bool migration_is_running(int state)
+{
+    switch (state) {
+    case MIGRATION_STATUS_ACTIVE:
+    case MIGRATION_STATUS_POSTCOPY_ACTIVE:
+    case MIGRATION_STATUS_POSTCOPY_PAUSED:
+    case MIGRATION_STATUS_POSTCOPY_RECOVER:
+    case MIGRATION_STATUS_SETUP:
+    case MIGRATION_STATUS_PRE_SWITCHOVER:
+    case MIGRATION_STATUS_DEVICE:
+    case MIGRATION_STATUS_WAIT_UNPLUG:
+    case MIGRATION_STATUS_CANCELLING:
+    case MIGRATION_STATUS_COLO:
+        return true;
+
+    default:
+        return false;
+
+    }
+}
+
 static void populate_time_info(MigrationInfo *info, MigrationState *s)
 {
     info->has_status = true;
@@ -1077,7 +1098,7 @@ void qmp_migrate_set_capabilities(MigrationCapabilityStatusList *params,
     MigrationCapabilityStatusList *cap;
     bool cap_list[MIGRATION_CAPABILITY__MAX];
 
-    if (migration_is_setup_or_active(s->state)) {
+    if (migration_is_running(s->state)) {
         error_setg(errp, QERR_MIGRATION_ACTIVE);
         return;
     }
@@ -1590,7 +1611,7 @@ static void migrate_fd_cancel(MigrationState *s)
 
     do {
         old_state = s->state;
-        if (!migration_is_setup_or_active(old_state)) {
+        if (!migration_is_running(old_state)) {
             break;
         }
         /* If the migration is paused, kick it out of the pause */
@@ -1888,9 +1909,7 @@ static bool migrate_prepare(MigrationState *s, bool blk, bool blk_inc,
         return true;
     }
 
-    if (migration_is_setup_or_active(s->state) ||
-        s->state == MIGRATION_STATUS_CANCELLING ||
-        s->state == MIGRATION_STATUS_COLO) {
+    if (migration_is_running(s->state)) {
         error_setg(errp, QERR_MIGRATION_ACTIVE);
         return false;
     }
diff --git a/migration/migration.h b/migration/migration.h
index aa9ff6f27b..44b1d56929 100644
--- a/migration/migration.h
+++ b/migration/migration.h
@@ -279,6 +279,7 @@ void migrate_fd_error(MigrationState *s, const Error *error);
 void migrate_fd_connect(MigrationState *s, Error *error_in);
 
 bool migration_is_setup_or_active(int state);
+bool migration_is_running(int state);
 
 void migrate_init(MigrationState *s);
 bool migration_is_blocked(Error **errp);
diff --git a/migration/savevm.c b/migration/savevm.c
index adfdca26ac..f19cb9ec7a 100644
--- a/migration/savevm.c
+++ b/migration/savevm.c
@@ -1531,9 +1531,7 @@ static int qemu_savevm_state(QEMUFile *f, Error **errp)
     MigrationState *ms = migrate_get_current();
     MigrationStatus status;
 
-    if (migration_is_setup_or_active(ms->state) ||
-        ms->state == MIGRATION_STATUS_CANCELLING ||
-        ms->state == MIGRATION_STATUS_COLO) {
+    if (migration_is_running(ms->state)) {
         error_setg(errp, QERR_MIGRATION_ACTIVE);
         return -EINVAL;
     }
-- 
2.24.1

Re: [PATCH v3 05/21] migration: Create migration_is_running()

[Dr. David Alan Gilbert]

* Juan Quintela (quintela@redhat.com) wrote:
> This function returns if we are in the middle of a migration.
> It is like migration_is_setup_or_active() with CANCELLING and COLO.
> Adapt all calers that are needed.
> 
> Signed-off-by: Juan Quintela <quintela@redhat.com>

Yes; the inclusion of cancelling is interesting because it's a state
where you don't want to be running but it still is.

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

> ---
>  migration/migration.c | 29 ++++++++++++++++++++++++-----
>  migration/migration.h |  1 +
>  migration/savevm.c    |  4 +---
>  3 files changed, 26 insertions(+), 8 deletions(-)
> 
> diff --git a/migration/migration.c b/migration/migration.c
> index 990bff00c0..1fb0aab44d 100644
> --- a/migration/migration.c
> +++ b/migration/migration.c
> @@ -829,6 +829,27 @@ bool migration_is_setup_or_active(int state)
>      }
>  }
>  
> +bool migration_is_running(int state)
> +{
> +    switch (state) {
> +    case MIGRATION_STATUS_ACTIVE:
> +    case MIGRATION_STATUS_POSTCOPY_ACTIVE:
> +    case MIGRATION_STATUS_POSTCOPY_PAUSED:
> +    case MIGRATION_STATUS_POSTCOPY_RECOVER:
> +    case MIGRATION_STATUS_SETUP:
> +    case MIGRATION_STATUS_PRE_SWITCHOVER:
> +    case MIGRATION_STATUS_DEVICE:
> +    case MIGRATION_STATUS_WAIT_UNPLUG:
> +    case MIGRATION_STATUS_CANCELLING:
> +    case MIGRATION_STATUS_COLO:
> +        return true;
> +
> +    default:
> +        return false;
> +
> +    }
> +}
> +
>  static void populate_time_info(MigrationInfo *info, MigrationState *s)
>  {
>      info->has_status = true;
> @@ -1077,7 +1098,7 @@ void qmp_migrate_set_capabilities(MigrationCapabilityStatusList *params,
>      MigrationCapabilityStatusList *cap;
>      bool cap_list[MIGRATION_CAPABILITY__MAX];
>  
> -    if (migration_is_setup_or_active(s->state)) {
> +    if (migration_is_running(s->state)) {
>          error_setg(errp, QERR_MIGRATION_ACTIVE);
>          return;
>      }
> @@ -1590,7 +1611,7 @@ static void migrate_fd_cancel(MigrationState *s)
>  
>      do {
>          old_state = s->state;
> -        if (!migration_is_setup_or_active(old_state)) {
> +        if (!migration_is_running(old_state)) {
>              break;
>          }
>          /* If the migration is paused, kick it out of the pause */
> @@ -1888,9 +1909,7 @@ static bool migrate_prepare(MigrationState *s, bool blk, bool blk_inc,
>          return true;
>      }
>  
> -    if (migration_is_setup_or_active(s->state) ||
> -        s->state == MIGRATION_STATUS_CANCELLING ||
> -        s->state == MIGRATION_STATUS_COLO) {
> +    if (migration_is_running(s->state)) {
>          error_setg(errp, QERR_MIGRATION_ACTIVE);
>          return false;
>      }
> diff --git a/migration/migration.h b/migration/migration.h
> index aa9ff6f27b..44b1d56929 100644
> --- a/migration/migration.h
> +++ b/migration/migration.h
> @@ -279,6 +279,7 @@ void migrate_fd_error(MigrationState *s, const Error *error);
>  void migrate_fd_connect(MigrationState *s, Error *error_in);
>  
>  bool migration_is_setup_or_active(int state);
> +bool migration_is_running(int state);
>  
>  void migrate_init(MigrationState *s);
>  bool migration_is_blocked(Error **errp);
> diff --git a/migration/savevm.c b/migration/savevm.c
> index adfdca26ac..f19cb9ec7a 100644
> --- a/migration/savevm.c
> +++ b/migration/savevm.c
> @@ -1531,9 +1531,7 @@ static int qemu_savevm_state(QEMUFile *f, Error **errp)
>      MigrationState *ms = migrate_get_current();
>      MigrationStatus status;
>  
> -    if (migration_is_setup_or_active(ms->state) ||
> -        ms->state == MIGRATION_STATUS_CANCELLING ||
> -        ms->state == MIGRATION_STATUS_COLO) {
> +    if (migration_is_running(ms->state)) {
>          error_setg(errp, QERR_MIGRATION_ACTIVE);
>          return -EINVAL;
>      }
> -- 
> 2.24.1
> 
--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK

[PATCH v3 06/21] migration: Don't send data if we have stopped

[Juan Quintela]

If we do a cancel, we got out without one error, but we can't do the
rest of the output as in a normal situation.

Signed-off-by: Juan Quintela <quintela@redhat.com>
---
 migration/ram.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/migration/ram.c b/migration/ram.c
index f95d656c26..3fd7fdffcf 100644
--- a/migration/ram.c
+++ b/migration/ram.c
@@ -3524,7 +3524,8 @@ static int ram_save_iterate(QEMUFile *f, void *opaque)
     ram_control_after_iterate(f, RAM_CONTROL_ROUND);
 
 out:
-    if (ret >= 0) {
+    if (ret >= 0
+        && migration_is_setup_or_active(migrate_get_current()->state)) {
         multifd_send_sync_main(rs);
         qemu_put_be64(f, RAM_SAVE_FLAG_EOS);
         qemu_fflush(f);
-- 
2.24.1

Re: [PATCH v3 06/21] migration: Don't send data if we have stopped

[Dr. David Alan Gilbert]

* Juan Quintela (quintela@redhat.com) wrote:
> If we do a cancel, we got out without one error, but we can't do the
> rest of the output as in a normal situation.
> 
> Signed-off-by: Juan Quintela <quintela@redhat.com>

I think it's the sync that's the main problem being avoided here rather
than actually the problem of sending the data.

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

> ---
>  migration/ram.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/migration/ram.c b/migration/ram.c
> index f95d656c26..3fd7fdffcf 100644
> --- a/migration/ram.c
> +++ b/migration/ram.c
> @@ -3524,7 +3524,8 @@ static int ram_save_iterate(QEMUFile *f, void *opaque)
>      ram_control_after_iterate(f, RAM_CONTROL_ROUND);
>  
>  out:
> -    if (ret >= 0) {
> +    if (ret >= 0
> +        && migration_is_setup_or_active(migrate_get_current()->state)) {
>          multifd_send_sync_main(rs);
>          qemu_put_be64(f, RAM_SAVE_FLAG_EOS);
>          qemu_fflush(f);
> -- 
> 2.24.1
> 
--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK

[PATCH v3 07/21] migration: Make multifd_save_setup() get an Error parameter

[Juan Quintela]

Signed-off-by: Juan Quintela <quintela@redhat.com>

---

We can't trust that error_in is not NULL.  Create a local_error.
---
 migration/migration.c | 4 +++-
 migration/ram.c       | 2 +-
 migration/ram.h       | 2 +-
 3 files changed, 5 insertions(+), 3 deletions(-)

diff --git a/migration/migration.c b/migration/migration.c
index 1fb0aab44d..7140d1e040 100644
--- a/migration/migration.c
+++ b/migration/migration.c
@@ -3367,6 +3367,7 @@ static void *migration_thread(void *opaque)
 
 void migrate_fd_connect(MigrationState *s, Error *error_in)
 {
+    Error *local_err = NULL;
     int64_t rate_limit;
     bool resume = s->state == MIGRATION_STATUS_POSTCOPY_PAUSED;
 
@@ -3415,7 +3416,8 @@ void migrate_fd_connect(MigrationState *s, Error *error_in)
         return;
     }
 
-    if (multifd_save_setup() != 0) {
+    if (multifd_save_setup(&local_err) != 0) {
+        error_report_err(local_err);
         migrate_set_state(&s->state, MIGRATION_STATUS_SETUP,
                           MIGRATION_STATUS_FAILED);
         migrate_fd_cleanup(s);
diff --git a/migration/ram.c b/migration/ram.c
index 3fd7fdffcf..d537264ba5 100644
--- a/migration/ram.c
+++ b/migration/ram.c
@@ -1243,7 +1243,7 @@ static void multifd_new_send_channel_async(QIOTask *task, gpointer opaque)
     }
 }
 
-int multifd_save_setup(void)
+int multifd_save_setup(Error **errp)
 {
     int thread_count;
     uint32_t page_count = MULTIFD_PACKET_SIZE / qemu_target_page_size();
diff --git a/migration/ram.h b/migration/ram.h
index bd0eee79b6..da22a417ea 100644
--- a/migration/ram.h
+++ b/migration/ram.h
@@ -41,7 +41,7 @@ int xbzrle_cache_resize(int64_t new_size, Error **errp);
 uint64_t ram_bytes_remaining(void);
 uint64_t ram_bytes_total(void);
 
-int multifd_save_setup(void);
+int multifd_save_setup(Error **errp);
 void multifd_save_cleanup(void);
 int multifd_load_setup(void);
 int multifd_load_cleanup(Error **errp);
-- 
2.24.1

Re: [PATCH v3 07/21] migration: Make multifd_save_setup() get an Error parameter

[Dr. David Alan Gilbert]

* Juan Quintela (quintela@redhat.com) wrote:
> Signed-off-by: Juan Quintela <quintela@redhat.com>
> 

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

> ---
> 
> We can't trust that error_in is not NULL.  Create a local_error.
> ---
>  migration/migration.c | 4 +++-
>  migration/ram.c       | 2 +-
>  migration/ram.h       | 2 +-
>  3 files changed, 5 insertions(+), 3 deletions(-)
> 
> diff --git a/migration/migration.c b/migration/migration.c
> index 1fb0aab44d..7140d1e040 100644
> --- a/migration/migration.c
> +++ b/migration/migration.c
> @@ -3367,6 +3367,7 @@ static void *migration_thread(void *opaque)
>  
>  void migrate_fd_connect(MigrationState *s, Error *error_in)
>  {
> +    Error *local_err = NULL;
>      int64_t rate_limit;
>      bool resume = s->state == MIGRATION_STATUS_POSTCOPY_PAUSED;
>  
> @@ -3415,7 +3416,8 @@ void migrate_fd_connect(MigrationState *s, Error *error_in)
>          return;
>      }
>  
> -    if (multifd_save_setup() != 0) {
> +    if (multifd_save_setup(&local_err) != 0) {
> +        error_report_err(local_err);
>          migrate_set_state(&s->state, MIGRATION_STATUS_SETUP,
>                            MIGRATION_STATUS_FAILED);
>          migrate_fd_cleanup(s);
> diff --git a/migration/ram.c b/migration/ram.c
> index 3fd7fdffcf..d537264ba5 100644
> --- a/migration/ram.c
> +++ b/migration/ram.c
> @@ -1243,7 +1243,7 @@ static void multifd_new_send_channel_async(QIOTask *task, gpointer opaque)
>      }
>  }
>  
> -int multifd_save_setup(void)
> +int multifd_save_setup(Error **errp)
>  {
>      int thread_count;
>      uint32_t page_count = MULTIFD_PACKET_SIZE / qemu_target_page_size();
> diff --git a/migration/ram.h b/migration/ram.h
> index bd0eee79b6..da22a417ea 100644
> --- a/migration/ram.h
> +++ b/migration/ram.h
> @@ -41,7 +41,7 @@ int xbzrle_cache_resize(int64_t new_size, Error **errp);
>  uint64_t ram_bytes_remaining(void);
>  uint64_t ram_bytes_total(void);
>  
> -int multifd_save_setup(void);
> +int multifd_save_setup(Error **errp);
>  void multifd_save_cleanup(void);
>  int multifd_load_setup(void);
>  int multifd_load_cleanup(Error **errp);
> -- 
> 2.24.1
> 
--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK

[PATCH v3 08/21] migration: Make multifd_load_setup() get an Error parameter

[Juan Quintela]

We need to change the full chain to pass the Error parameter.

Signed-off-by: Juan Quintela <quintela@redhat.com>

---

Always use a local_err, and in case of error propagate/print it as needed.
---
 migration/migration.c | 35 +++++++++++++++++++++++++++++------
 migration/migration.h |  2 +-
 migration/ram.c       |  2 +-
 migration/ram.h       |  2 +-
 migration/rdma.c      |  2 +-
 5 files changed, 33 insertions(+), 10 deletions(-)

diff --git a/migration/migration.c b/migration/migration.c
index 7140d1e040..ef17b9f3f2 100644
--- a/migration/migration.c
+++ b/migration/migration.c
@@ -518,13 +518,23 @@ fail:
     exit(EXIT_FAILURE);
 }
 
-static void migration_incoming_setup(QEMUFile *f)
+/**
+ * @migration_incoming_setup: Setup incoming migration
+ *
+ * Returns 0 for no error or 1 for error
+ *
+ * @f: file for main migration channel
+ * @errp: where to put errors
+ */
+static int migration_incoming_setup(QEMUFile *f, Error **errp)
 {
     MigrationIncomingState *mis = migration_incoming_get_current();
+    Error *local_err = NULL;
 
-    if (multifd_load_setup() != 0) {
+    if (multifd_load_setup(&local_err) != 0) {
         /* We haven't been able to create multifd threads
            nothing better to do */
+        error_report_err(local_err);
         exit(EXIT_FAILURE);
     }
 
@@ -532,6 +542,7 @@ static void migration_incoming_setup(QEMUFile *f)
         mis->from_src_file = f;
     }
     qemu_file_set_blocking(f, false);
+    return 0;
 }
 
 void migration_incoming_process(void)
@@ -572,19 +583,27 @@ static bool postcopy_try_recover(QEMUFile *f)
     return false;
 }
 
-void migration_fd_process_incoming(QEMUFile *f)
+void migration_fd_process_incoming(QEMUFile *f, Error **errp)
 {
+    Error *local_err = NULL;
+
     if (postcopy_try_recover(f)) {
         return;
     }
 
-    migration_incoming_setup(f);
+    if (migration_incoming_setup(f, &local_err)) {
+        if (local_err) {
+            error_propagate(errp, local_err);
+        }
+        return;
+    }
     migration_incoming_process();
 }
 
 void migration_ioc_process_incoming(QIOChannel *ioc, Error **errp)
 {
     MigrationIncomingState *mis = migration_incoming_get_current();
+    Error *local_err = NULL;
     bool start_migration;
 
     if (!mis->from_src_file) {
@@ -596,7 +615,12 @@ void migration_ioc_process_incoming(QIOChannel *ioc, Error **errp)
             return;
         }
 
-        migration_incoming_setup(f);
+        if (migration_incoming_setup(f, &local_err)) {
+            if (local_err) {
+                error_propagate(errp, local_err);
+            }
+            return;
+        }
 
         /*
          * Common migration only needs one channel, so we can start
@@ -604,7 +628,6 @@ void migration_ioc_process_incoming(QIOChannel *ioc, Error **errp)
          */
         start_migration = !migrate_use_multifd();
     } else {
-        Error *local_err = NULL;
         /* Multiple connections */
         assert(migrate_use_multifd());
         start_migration = multifd_recv_new_channel(ioc, &local_err);
diff --git a/migration/migration.h b/migration/migration.h
index 44b1d56929..8473ddfc88 100644
--- a/migration/migration.h
+++ b/migration/migration.h
@@ -265,7 +265,7 @@ struct MigrationState
 
 void migrate_set_state(int *state, int old_state, int new_state);
 
-void migration_fd_process_incoming(QEMUFile *f);
+void migration_fd_process_incoming(QEMUFile *f, Error **errp);
 void migration_ioc_process_incoming(QIOChannel *ioc, Error **errp);
 void migration_incoming_process(void);
 
diff --git a/migration/ram.c b/migration/ram.c
index d537264ba5..125c6d0f60 100644
--- a/migration/ram.c
+++ b/migration/ram.c
@@ -1463,7 +1463,7 @@ static void *multifd_recv_thread(void *opaque)
     return NULL;
 }
 
-int multifd_load_setup(void)
+int multifd_load_setup(Error **errp)
 {
     int thread_count;
     uint32_t page_count = MULTIFD_PACKET_SIZE / qemu_target_page_size();
diff --git a/migration/ram.h b/migration/ram.h
index da22a417ea..42be471d52 100644
--- a/migration/ram.h
+++ b/migration/ram.h
@@ -43,7 +43,7 @@ uint64_t ram_bytes_total(void);
 
 int multifd_save_setup(Error **errp);
 void multifd_save_cleanup(void);
-int multifd_load_setup(void);
+int multifd_load_setup(Error **errp);
 int multifd_load_cleanup(Error **errp);
 bool multifd_recv_all_channels_created(void);
 bool multifd_recv_new_channel(QIOChannel *ioc, Error **errp);
diff --git a/migration/rdma.c b/migration/rdma.c
index e241dcb992..2379b8345b 100644
--- a/migration/rdma.c
+++ b/migration/rdma.c
@@ -4004,7 +4004,7 @@ static void rdma_accept_incoming_migration(void *opaque)
     }
 
     rdma->migration_started_on_destination = 1;
-    migration_fd_process_incoming(f);
+    migration_fd_process_incoming(f, errp);
 }
 
 void rdma_start_incoming_migration(const char *host_port, Error **errp)
-- 
2.24.1

Re: [PATCH v3 08/21] migration: Make multifd_load_setup() get an Error parameter

[Dr. David Alan Gilbert]

* Juan Quintela (quintela@redhat.com) wrote:
> We need to change the full chain to pass the Error parameter.
> 
> Signed-off-by: Juan Quintela <quintela@redhat.com>

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

> 
> ---
> 
> Always use a local_err, and in case of error propagate/print it as needed.
> ---
>  migration/migration.c | 35 +++++++++++++++++++++++++++++------
>  migration/migration.h |  2 +-
>  migration/ram.c       |  2 +-
>  migration/ram.h       |  2 +-
>  migration/rdma.c      |  2 +-
>  5 files changed, 33 insertions(+), 10 deletions(-)
> 
> diff --git a/migration/migration.c b/migration/migration.c
> index 7140d1e040..ef17b9f3f2 100644
> --- a/migration/migration.c
> +++ b/migration/migration.c
> @@ -518,13 +518,23 @@ fail:
>      exit(EXIT_FAILURE);
>  }
>  
> -static void migration_incoming_setup(QEMUFile *f)
> +/**
> + * @migration_incoming_setup: Setup incoming migration
> + *
> + * Returns 0 for no error or 1 for error
> + *
> + * @f: file for main migration channel
> + * @errp: where to put errors
> + */
> +static int migration_incoming_setup(QEMUFile *f, Error **errp)
>  {
>      MigrationIncomingState *mis = migration_incoming_get_current();
> +    Error *local_err = NULL;
>  
> -    if (multifd_load_setup() != 0) {
> +    if (multifd_load_setup(&local_err) != 0) {
>          /* We haven't been able to create multifd threads
>             nothing better to do */
> +        error_report_err(local_err);
>          exit(EXIT_FAILURE);
>      }
>  
> @@ -532,6 +542,7 @@ static void migration_incoming_setup(QEMUFile *f)
>          mis->from_src_file = f;
>      }
>      qemu_file_set_blocking(f, false);
> +    return 0;
>  }
>  
>  void migration_incoming_process(void)
> @@ -572,19 +583,27 @@ static bool postcopy_try_recover(QEMUFile *f)
>      return false;
>  }
>  
> -void migration_fd_process_incoming(QEMUFile *f)
> +void migration_fd_process_incoming(QEMUFile *f, Error **errp)
>  {
> +    Error *local_err = NULL;
> +
>      if (postcopy_try_recover(f)) {
>          return;
>      }
>  
> -    migration_incoming_setup(f);
> +    if (migration_incoming_setup(f, &local_err)) {
> +        if (local_err) {
> +            error_propagate(errp, local_err);
> +        }
> +        return;
> +    }
>      migration_incoming_process();
>  }
>  
>  void migration_ioc_process_incoming(QIOChannel *ioc, Error **errp)
>  {
>      MigrationIncomingState *mis = migration_incoming_get_current();
> +    Error *local_err = NULL;
>      bool start_migration;
>  
>      if (!mis->from_src_file) {
> @@ -596,7 +615,12 @@ void migration_ioc_process_incoming(QIOChannel *ioc, Error **errp)
>              return;
>          }
>  
> -        migration_incoming_setup(f);
> +        if (migration_incoming_setup(f, &local_err)) {
> +            if (local_err) {
> +                error_propagate(errp, local_err);
> +            }
> +            return;
> +        }
>  
>          /*
>           * Common migration only needs one channel, so we can start
> @@ -604,7 +628,6 @@ void migration_ioc_process_incoming(QIOChannel *ioc, Error **errp)
>           */
>          start_migration = !migrate_use_multifd();
>      } else {
> -        Error *local_err = NULL;
>          /* Multiple connections */
>          assert(migrate_use_multifd());
>          start_migration = multifd_recv_new_channel(ioc, &local_err);
> diff --git a/migration/migration.h b/migration/migration.h
> index 44b1d56929..8473ddfc88 100644
> --- a/migration/migration.h
> +++ b/migration/migration.h
> @@ -265,7 +265,7 @@ struct MigrationState
>  
>  void migrate_set_state(int *state, int old_state, int new_state);
>  
> -void migration_fd_process_incoming(QEMUFile *f);
> +void migration_fd_process_incoming(QEMUFile *f, Error **errp);
>  void migration_ioc_process_incoming(QIOChannel *ioc, Error **errp);
>  void migration_incoming_process(void);
>  
> diff --git a/migration/ram.c b/migration/ram.c
> index d537264ba5..125c6d0f60 100644
> --- a/migration/ram.c
> +++ b/migration/ram.c
> @@ -1463,7 +1463,7 @@ static void *multifd_recv_thread(void *opaque)
>      return NULL;
>  }
>  
> -int multifd_load_setup(void)
> +int multifd_load_setup(Error **errp)
>  {
>      int thread_count;
>      uint32_t page_count = MULTIFD_PACKET_SIZE / qemu_target_page_size();
> diff --git a/migration/ram.h b/migration/ram.h
> index da22a417ea..42be471d52 100644
> --- a/migration/ram.h
> +++ b/migration/ram.h
> @@ -43,7 +43,7 @@ uint64_t ram_bytes_total(void);
>  
>  int multifd_save_setup(Error **errp);
>  void multifd_save_cleanup(void);
> -int multifd_load_setup(void);
> +int multifd_load_setup(Error **errp);
>  int multifd_load_cleanup(Error **errp);
>  bool multifd_recv_all_channels_created(void);
>  bool multifd_recv_new_channel(QIOChannel *ioc, Error **errp);
> diff --git a/migration/rdma.c b/migration/rdma.c
> index e241dcb992..2379b8345b 100644
> --- a/migration/rdma.c
> +++ b/migration/rdma.c
> @@ -4004,7 +4004,7 @@ static void rdma_accept_incoming_migration(void *opaque)
>      }
>  
>      rdma->migration_started_on_destination = 1;
> -    migration_fd_process_incoming(f);
> +    migration_fd_process_incoming(f, errp);
>  }
>  
>  void rdma_start_incoming_migration(const char *host_port, Error **errp)
> -- 
> 2.24.1
> 
--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK

[PATCH v3 09/21] migration: Add multifd-compress parameter

[Juan Quintela]

Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>

---
Rename it to NONE
Fix typos (dave)
We don't need to chek values returned by visit_type_MultifdCompress (markus)
Fix yet more typos (wei)
---
 hw/core/qdev-properties.c    | 13 +++++++++++++
 include/hw/qdev-properties.h |  3 +++
 migration/migration.c        | 13 +++++++++++++
 monitor/hmp-cmds.c           | 13 +++++++++++++
 qapi/migration.json          | 30 +++++++++++++++++++++++++++---
 tests/qtest/migration-test.c | 14 ++++++++++----
 6 files changed, 79 insertions(+), 7 deletions(-)

diff --git a/hw/core/qdev-properties.c b/hw/core/qdev-properties.c
index 6ca7697599..ff6a752b19 100644
--- a/hw/core/qdev-properties.c
+++ b/hw/core/qdev-properties.c
@@ -8,6 +8,7 @@
 #include "qapi/qmp/qerror.h"
 #include "qemu/ctype.h"
 #include "qemu/error-report.h"
+#include "qapi/qapi-types-migration.h"
 #include "hw/block/block.h"
 #include "net/hub.h"
 #include "qapi/visitor.h"
@@ -641,6 +642,18 @@ const PropertyInfo qdev_prop_fdc_drive_type = {
     .set_default_value = set_default_value_enum,
 };
 
+/* --- MultifdCompress --- */
+
+const PropertyInfo qdev_prop_multifd_compress = {
+    .name = "MultifdCompress",
+    .description = "multifd_compress values, "
+                   "none",
+    .enum_table = &MultifdCompress_lookup,
+    .get = get_enum,
+    .set = set_enum,
+    .set_default_value = set_default_value_enum,
+};
+
 /* --- pci address --- */
 
 /*
diff --git a/include/hw/qdev-properties.h b/include/hw/qdev-properties.h
index a90a9cec80..72f21a74c6 100644
--- a/include/hw/qdev-properties.h
+++ b/include/hw/qdev-properties.h
@@ -20,6 +20,7 @@ extern const PropertyInfo qdev_prop_chr;
 extern const PropertyInfo qdev_prop_tpm;
 extern const PropertyInfo qdev_prop_macaddr;
 extern const PropertyInfo qdev_prop_on_off_auto;
+extern const PropertyInfo qdev_prop_multifd_compress;
 extern const PropertyInfo qdev_prop_losttickpolicy;
 extern const PropertyInfo qdev_prop_blockdev_on_error;
 extern const PropertyInfo qdev_prop_bios_chs_trans;
@@ -184,6 +185,8 @@ extern const PropertyInfo qdev_prop_pcie_link_width;
     DEFINE_PROP(_n, _s, _f, qdev_prop_macaddr, MACAddr)
 #define DEFINE_PROP_ON_OFF_AUTO(_n, _s, _f, _d) \
     DEFINE_PROP_SIGNED(_n, _s, _f, _d, qdev_prop_on_off_auto, OnOffAuto)
+#define DEFINE_PROP_MULTIFD_COMPRESS(_n, _s, _f, _d) \
+    DEFINE_PROP_SIGNED(_n, _s, _f, _d, qdev_prop_multifd_compress, MultifdCompress)
 #define DEFINE_PROP_LOSTTICKPOLICY(_n, _s, _f, _d) \
     DEFINE_PROP_SIGNED(_n, _s, _f, _d, qdev_prop_losttickpolicy, \
                         LostTickPolicy)
diff --git a/migration/migration.c b/migration/migration.c
index ef17b9f3f2..ecb56afd50 100644
--- a/migration/migration.c
+++ b/migration/migration.c
@@ -87,6 +87,7 @@
 /* The delay time (in ms) between two COLO checkpoints */
 #define DEFAULT_MIGRATE_X_CHECKPOINT_DELAY (200 * 100)
 #define DEFAULT_MIGRATE_MULTIFD_CHANNELS 2
+#define DEFAULT_MIGRATE_MULTIFD_COMPRESS MULTIFD_COMPRESS_NONE
 
 /* Background transfer rate for postcopy, 0 means unlimited, note
  * that page requests can still exceed this limit.
@@ -797,6 +798,8 @@ MigrationParameters *qmp_query_migrate_parameters(Error **errp)
     params->block_incremental = s->parameters.block_incremental;
     params->has_multifd_channels = true;
     params->multifd_channels = s->parameters.multifd_channels;
+    params->has_multifd_compress = true;
+    params->multifd_compress = s->parameters.multifd_compress;
     params->has_xbzrle_cache_size = true;
     params->xbzrle_cache_size = s->parameters.xbzrle_cache_size;
     params->has_max_postcopy_bandwidth = true;
@@ -1314,6 +1317,9 @@ static void migrate_params_test_apply(MigrateSetParameters *params,
     if (params->has_multifd_channels) {
         dest->multifd_channels = params->multifd_channels;
     }
+    if (params->has_multifd_compress) {
+        dest->multifd_compress = params->multifd_compress;
+    }
     if (params->has_xbzrle_cache_size) {
         dest->xbzrle_cache_size = params->xbzrle_cache_size;
     }
@@ -1410,6 +1416,9 @@ static void migrate_params_apply(MigrateSetParameters *params, Error **errp)
     if (params->has_multifd_channels) {
         s->parameters.multifd_channels = params->multifd_channels;
     }
+    if (params->has_multifd_compress) {
+        s->parameters.multifd_compress = params->multifd_compress;
+    }
     if (params->has_xbzrle_cache_size) {
         s->parameters.xbzrle_cache_size = params->xbzrle_cache_size;
         xbzrle_cache_resize(params->xbzrle_cache_size, errp);
@@ -3514,6 +3523,9 @@ static Property migration_properties[] = {
     DEFINE_PROP_UINT8("multifd-channels", MigrationState,
                       parameters.multifd_channels,
                       DEFAULT_MIGRATE_MULTIFD_CHANNELS),
+    DEFINE_PROP_MULTIFD_COMPRESS("multifd-compress", MigrationState,
+                      parameters.multifd_compress,
+                      DEFAULT_MIGRATE_MULTIFD_COMPRESS),
     DEFINE_PROP_SIZE("xbzrle-cache-size", MigrationState,
                       parameters.xbzrle_cache_size,
                       DEFAULT_MIGRATE_XBZRLE_CACHE_SIZE),
@@ -3604,6 +3616,7 @@ static void migration_instance_init(Object *obj)
     params->has_x_checkpoint_delay = true;
     params->has_block_incremental = true;
     params->has_multifd_channels = true;
+    params->has_multifd_compress = true;
     params->has_xbzrle_cache_size = true;
     params->has_max_postcopy_bandwidth = true;
     params->has_max_cpu_throttle = true;
diff --git a/monitor/hmp-cmds.c b/monitor/hmp-cmds.c
index d0e0af893a..ba75a609a7 100644
--- a/monitor/hmp-cmds.c
+++ b/monitor/hmp-cmds.c
@@ -39,6 +39,7 @@
 #include "qapi/qapi-commands-tpm.h"
 #include "qapi/qapi-commands-ui.h"
 #include "qapi/qapi-visit-net.h"
+#include "qapi/qapi-visit-migration.h"
 #include "qapi/qmp/qdict.h"
 #include "qapi/qmp/qerror.h"
 #include "qapi/string-input-visitor.h"
@@ -447,6 +448,9 @@ void hmp_info_migrate_parameters(Monitor *mon, const QDict *qdict)
         monitor_printf(mon, "%s: %u\n",
             MigrationParameter_str(MIGRATION_PARAMETER_MULTIFD_CHANNELS),
             params->multifd_channels);
+        monitor_printf(mon, "%s: %s\n",
+            MigrationParameter_str(MIGRATION_PARAMETER_MULTIFD_COMPRESS),
+            MultifdCompress_str(params->multifd_compress));
         monitor_printf(mon, "%s: %" PRIu64 "\n",
             MigrationParameter_str(MIGRATION_PARAMETER_XBZRLE_CACHE_SIZE),
             params->xbzrle_cache_size);
@@ -1738,6 +1742,7 @@ void hmp_migrate_set_parameter(Monitor *mon, const QDict *qdict)
     MigrateSetParameters *p = g_new0(MigrateSetParameters, 1);
     uint64_t valuebw = 0;
     uint64_t cache_size;
+    MultifdCompress compress_type;
     Error *err = NULL;
     int val, ret;
 
@@ -1823,6 +1828,14 @@ void hmp_migrate_set_parameter(Monitor *mon, const QDict *qdict)
         p->has_multifd_channels = true;
         visit_type_int(v, param, &p->multifd_channels, &err);
         break;
+    case MIGRATION_PARAMETER_MULTIFD_COMPRESS:
+        p->has_multifd_compress = true;
+        visit_type_MultifdCompress(v, param, &compress_type, &err);
+        if (err) {
+            break;
+        }
+        p->multifd_compress = compress_type;
+        break;
     case MIGRATION_PARAMETER_XBZRLE_CACHE_SIZE:
         p->has_xbzrle_cache_size = true;
         visit_type_size(v, param, &cache_size, &err);
diff --git a/qapi/migration.json b/qapi/migration.json
index b7348d0c8b..c2891e6ebf 100644
--- a/qapi/migration.json
+++ b/qapi/migration.json
@@ -488,6 +488,19 @@
 ##
 { 'command': 'query-migrate-capabilities', 'returns':   ['MigrationCapabilityStatus']}
 
+##
+# @MultifdCompress:
+#
+# An enumeration of multifd compression.
+#
+# @none: no compression.
+#
+# Since: 4.1
+#
+##
+{ 'enum': 'MultifdCompress',
+  'data': [ 'none' ] }
+
 ##
 # @MigrationParameter:
 #
@@ -586,6 +599,9 @@
 # @max-cpu-throttle: maximum cpu throttle percentage.
 #                    Defaults to 99. (Since 3.1)
 #
+# @multifd-compress: Which compression method to use.
+#                    Defaults to none. (Since 5.0)
+#
 # Since: 2.4
 ##
 { 'enum': 'MigrationParameter',
@@ -598,7 +614,7 @@
            'downtime-limit', 'x-checkpoint-delay', 'block-incremental',
            'multifd-channels',
            'xbzrle-cache-size', 'max-postcopy-bandwidth',
-           'max-cpu-throttle' ] }
+           'max-cpu-throttle', 'multifd-compress' ] }
 
 ##
 # @MigrateSetParameters:
@@ -688,6 +704,9 @@
 # @max-cpu-throttle: maximum cpu throttle percentage.
 #                    The default value is 99. (Since 3.1)
 #
+# @multifd-compress: Which compression method to use.
+#                    Defaults to none. (Since 5.0)
+#
 # Since: 2.4
 ##
 # TODO either fuse back into MigrationParameters, or make
@@ -713,7 +732,8 @@
             '*multifd-channels': 'int',
             '*xbzrle-cache-size': 'size',
             '*max-postcopy-bandwidth': 'size',
-	    '*max-cpu-throttle': 'int' } }
+            '*max-cpu-throttle': 'int',
+            '*multifd-compress': 'MultifdCompress' } }
 
 ##
 # @migrate-set-parameters:
@@ -823,6 +843,9 @@
 #                    Defaults to 99.
 #                     (Since 3.1)
 #
+# @multifd-compress: Which compression method to use.
+#                    Defaults to none. (Since 5.0)
+#
 # Since: 2.4
 ##
 { 'struct': 'MigrationParameters',
@@ -846,7 +869,8 @@
             '*multifd-channels': 'uint8',
             '*xbzrle-cache-size': 'size',
 	    '*max-postcopy-bandwidth': 'size',
-            '*max-cpu-throttle':'uint8'} }
+            '*max-cpu-throttle': 'uint8',
+            '*multifd-compress': 'MultifdCompress' } }
 
 ##
 # @query-migrate-parameters:
diff --git a/tests/qtest/migration-test.c b/tests/qtest/migration-test.c
index cf27ebbc9d..3d5d2aba8c 100644
--- a/tests/qtest/migration-test.c
+++ b/tests/qtest/migration-test.c
@@ -378,7 +378,6 @@ static void migrate_check_parameter_str(QTestState *who, const char *parameter,
     g_free(result);
 }
 
-__attribute__((unused))
 static void migrate_set_parameter_str(QTestState *who, const char *parameter,
                                       const char *value)
 {
@@ -1251,7 +1250,7 @@ static void test_migrate_auto_converge(void)
     test_migrate_end(from, to, true);
 }
 
-static void test_multifd_tcp(void)
+static void test_multifd_tcp(const char *method)
 {
     MigrateStart *args = migrate_start_new();
     QTestState *from, *to;
@@ -1275,6 +1274,9 @@ static void test_multifd_tcp(void)
     migrate_set_parameter_int(from, "multifd-channels", 16);
     migrate_set_parameter_int(to, "multifd-channels", 16);
 
+    migrate_set_parameter_str(from, "multifd-compress", method);
+    migrate_set_parameter_str(to, "multifd-compress", method);
+
     migrate_set_capability(from, "multifd", "true");
     migrate_set_capability(to, "multifd", "true");
 
@@ -1306,6 +1308,11 @@ static void test_multifd_tcp(void)
     g_free(uri);
 }
 
+static void test_multifd_tcp_none(void)
+{
+    test_multifd_tcp("none");
+}
+
 /*
  * This test does:
  *  source               target
@@ -1317,7 +1324,6 @@ static void test_multifd_tcp(void)
  *
  *  And see that it works
  */
-
 static void test_multifd_tcp_cancel(void)
 {
     MigrateStart *args = migrate_start_new();
@@ -1467,7 +1473,7 @@ int main(int argc, char **argv)
                    test_validate_uuid_dst_not_set);
 
     qtest_add_func("/migration/auto_converge", test_migrate_auto_converge);
-    qtest_add_func("/migration/multifd/tcp", test_multifd_tcp);
+    qtest_add_func("/migration/multifd/tcp/none", test_multifd_tcp_none);
     qtest_add_func("/migration/multifd/tcp/cancel", test_multifd_tcp_cancel);
 
     ret = g_test_run();
-- 
2.24.1

Re: [PATCH v3 09/21] migration: Add multifd-compress parameter

[Eric Blake]

On 1/23/20 5:58 AM, Juan Quintela wrote:
> Signed-off-by: Juan Quintela <quintela@redhat.com>
> Reviewed-by: Markus Armbruster <armbru@redhat.com>
> Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
> 
> ---
> Rename it to NONE
> Fix typos (dave)
> We don't need to chek values returned by visit_type_MultifdCompress (markus)
> Fix yet more typos (wei)
> ---
>   hw/core/qdev-properties.c    | 13 +++++++++++++

> +++ b/qapi/migration.json
> @@ -488,6 +488,19 @@
>   ##
>   { 'command': 'query-migrate-capabilities', 'returns':   ['MigrationCapabilityStatus']}
>   
> +##
> +# @MultifdCompress:
> +#
> +# An enumeration of multifd compression.
> +#
> +# @none: no compression.
> +#
> +# Since: 4.1

5.0, now

> +#
> +##
> +{ 'enum': 'MultifdCompress',
> +  'data': [ 'none' ] }
> +
>   ##
>   # @MigrationParameter:
>   #
> @@ -586,6 +599,9 @@
>   # @max-cpu-throttle: maximum cpu throttle percentage.
>   #                    Defaults to 99. (Since 3.1)
>   #
> +# @multifd-compress: Which compression method to use.
> +#                    Defaults to none. (Since 5.0)
> +#
>   # Since: 2.4
>   ##
>   { 'enum': 'MigrationParameter',
> @@ -598,7 +614,7 @@
>              'downtime-limit', 'x-checkpoint-delay', 'block-incremental',
>              'multifd-channels',
>              'xbzrle-cache-size', 'max-postcopy-bandwidth',
> -           'max-cpu-throttle' ] }
> +           'max-cpu-throttle', 'multifd-compress' ] }
-- 
Eric Blake, Principal Software Engineer
Red Hat, Inc.           +1-919-301-3226
Virtualization:  qemu.org | libvirt.org

[PATCH v3 10/21] ram_addr: Split RAMBlock definition

[Juan Quintela]

We need some of the fields without having to poison everything else.

Signed-off-by: Juan Quintela <quintela@redhat.com>
---
 MAINTAINERS             |  1 +
 include/exec/ram_addr.h | 40 +-------------------------
 include/exec/ramblock.h | 64 +++++++++++++++++++++++++++++++++++++++++
 3 files changed, 66 insertions(+), 39 deletions(-)
 create mode 100644 include/exec/ramblock.h

diff --git a/MAINTAINERS b/MAINTAINERS
index 2c768ed3d8..3732f746b3 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -1965,6 +1965,7 @@ F: ioport.c
 F: include/exec/memop.h
 F: include/exec/memory.h
 F: include/exec/ram_addr.h
+F: include/exec/ramblock.h
 F: memory.c
 F: include/exec/memory-internal.h
 F: exec.c
diff --git a/include/exec/ram_addr.h b/include/exec/ram_addr.h
index 5adebb0bc7..5e59a3d8d7 100644
--- a/include/exec/ram_addr.h
+++ b/include/exec/ram_addr.h
@@ -24,45 +24,7 @@
 #include "hw/xen/xen.h"
 #include "sysemu/tcg.h"
 #include "exec/ramlist.h"
-
-struct RAMBlock {
-    struct rcu_head rcu;
-    struct MemoryRegion *mr;
-    uint8_t *host;
-    uint8_t *colo_cache; /* For colo, VM's ram cache */
-    ram_addr_t offset;
-    ram_addr_t used_length;
-    ram_addr_t max_length;
-    void (*resized)(const char*, uint64_t length, void *host);
-    uint32_t flags;
-    /* Protected by iothread lock.  */
-    char idstr[256];
-    /* RCU-enabled, writes protected by the ramlist lock */
-    QLIST_ENTRY(RAMBlock) next;
-    QLIST_HEAD(, RAMBlockNotifier) ramblock_notifiers;
-    int fd;
-    size_t page_size;
-    /* dirty bitmap used during migration */
-    unsigned long *bmap;
-    /* bitmap of already received pages in postcopy */
-    unsigned long *receivedmap;
-
-    /*
-     * bitmap to track already cleared dirty bitmap.  When the bit is
-     * set, it means the corresponding memory chunk needs a log-clear.
-     * Set this up to non-NULL to enable the capability to postpone
-     * and split clearing of dirty bitmap on the remote node (e.g.,
-     * KVM).  The bitmap will be set only when doing global sync.
-     *
-     * NOTE: this bitmap is different comparing to the other bitmaps
-     * in that one bit can represent multiple guest pages (which is
-     * decided by the `clear_bmap_shift' variable below).  On
-     * destination side, this should always be NULL, and the variable
-     * `clear_bmap_shift' is meaningless.
-     */
-    unsigned long *clear_bmap;
-    uint8_t clear_bmap_shift;
-};
+#include "exec/ramblock.h"
 
 /**
  * clear_bmap_size: calculate clear bitmap size
diff --git a/include/exec/ramblock.h b/include/exec/ramblock.h
new file mode 100644
index 0000000000..07d50864d8
--- /dev/null
+++ b/include/exec/ramblock.h
@@ -0,0 +1,64 @@
+/*
+ * Declarations for cpu physical memory functions
+ *
+ * Copyright 2011 Red Hat, Inc. and/or its affiliates
+ *
+ * Authors:
+ *  Avi Kivity <avi@redhat.com>
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or
+ * later.  See the COPYING file in the top-level directory.
+ *
+ */
+
+/*
+ * This header is for use by exec.c and memory.c ONLY.  Do not include it.
+ * The functions declared here will be removed soon.
+ */
+
+#ifndef QEMU_EXEC_RAMBLOCK_H
+#define QEMU_EXEC_RAMBLOCK_H
+
+#ifndef CONFIG_USER_ONLY
+#include "cpu-common.h"
+
+struct RAMBlock {
+    struct rcu_head rcu;
+    struct MemoryRegion *mr;
+    uint8_t *host;
+    uint8_t *colo_cache; /* For colo, VM's ram cache */
+    ram_addr_t offset;
+    ram_addr_t used_length;
+    ram_addr_t max_length;
+    void (*resized)(const char*, uint64_t length, void *host);
+    uint32_t flags;
+    /* Protected by iothread lock.  */
+    char idstr[256];
+    /* RCU-enabled, writes protected by the ramlist lock */
+    QLIST_ENTRY(RAMBlock) next;
+    QLIST_HEAD(, RAMBlockNotifier) ramblock_notifiers;
+    int fd;
+    size_t page_size;
+    /* dirty bitmap used during migration */
+    unsigned long *bmap;
+    /* bitmap of already received pages in postcopy */
+    unsigned long *receivedmap;
+
+    /*
+     * bitmap to track already cleared dirty bitmap.  When the bit is
+     * set, it means the corresponding memory chunk needs a log-clear.
+     * Set this up to non-NULL to enable the capability to postpone
+     * and split clearing of dirty bitmap on the remote node (e.g.,
+     * KVM).  The bitmap will be set only when doing global sync.
+     *
+     * NOTE: this bitmap is different comparing to the other bitmaps
+     * in that one bit can represent multiple guest pages (which is
+     * decided by the `clear_bmap_shift' variable below).  On
+     * destination side, this should always be NULL, and the variable
+     * `clear_bmap_shift' is meaningless.
+     */
+    unsigned long *clear_bmap;
+    uint8_t clear_bmap_shift;
+};
+#endif
+#endif
-- 
2.24.1

Re: [PATCH v3 10/21] ram_addr: Split RAMBlock definition

[Dr. David Alan Gilbert]

* Juan Quintela (quintela@redhat.com) wrote:
> We need some of the fields without having to poison everything else.
> 
> Signed-off-by: Juan Quintela <quintela@redhat.com>

OK.
(I guess the copyright matches the file we're splitting from; would be
good to update the date some time).

I wondered if any of the RAMBlock related function declarations should
move as wlel; but it's not obvious which ones.

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

> ---
>  MAINTAINERS             |  1 +
>  include/exec/ram_addr.h | 40 +-------------------------
>  include/exec/ramblock.h | 64 +++++++++++++++++++++++++++++++++++++++++
>  3 files changed, 66 insertions(+), 39 deletions(-)
>  create mode 100644 include/exec/ramblock.h
> 
> diff --git a/MAINTAINERS b/MAINTAINERS
> index 2c768ed3d8..3732f746b3 100644
> --- a/MAINTAINERS
> +++ b/MAINTAINERS
> @@ -1965,6 +1965,7 @@ F: ioport.c
>  F: include/exec/memop.h
>  F: include/exec/memory.h
>  F: include/exec/ram_addr.h
> +F: include/exec/ramblock.h
>  F: memory.c
>  F: include/exec/memory-internal.h
>  F: exec.c
> diff --git a/include/exec/ram_addr.h b/include/exec/ram_addr.h
> index 5adebb0bc7..5e59a3d8d7 100644
> --- a/include/exec/ram_addr.h
> +++ b/include/exec/ram_addr.h
> @@ -24,45 +24,7 @@
>  #include "hw/xen/xen.h"
>  #include "sysemu/tcg.h"
>  #include "exec/ramlist.h"
> -
> -struct RAMBlock {
> -    struct rcu_head rcu;
> -    struct MemoryRegion *mr;
> -    uint8_t *host;
> -    uint8_t *colo_cache; /* For colo, VM's ram cache */
> -    ram_addr_t offset;
> -    ram_addr_t used_length;
> -    ram_addr_t max_length;
> -    void (*resized)(const char*, uint64_t length, void *host);
> -    uint32_t flags;
> -    /* Protected by iothread lock.  */
> -    char idstr[256];
> -    /* RCU-enabled, writes protected by the ramlist lock */
> -    QLIST_ENTRY(RAMBlock) next;
> -    QLIST_HEAD(, RAMBlockNotifier) ramblock_notifiers;
> -    int fd;
> -    size_t page_size;
> -    /* dirty bitmap used during migration */
> -    unsigned long *bmap;
> -    /* bitmap of already received pages in postcopy */
> -    unsigned long *receivedmap;
> -
> -    /*
> -     * bitmap to track already cleared dirty bitmap.  When the bit is
> -     * set, it means the corresponding memory chunk needs a log-clear.
> -     * Set this up to non-NULL to enable the capability to postpone
> -     * and split clearing of dirty bitmap on the remote node (e.g.,
> -     * KVM).  The bitmap will be set only when doing global sync.
> -     *
> -     * NOTE: this bitmap is different comparing to the other bitmaps
> -     * in that one bit can represent multiple guest pages (which is
> -     * decided by the `clear_bmap_shift' variable below).  On
> -     * destination side, this should always be NULL, and the variable
> -     * `clear_bmap_shift' is meaningless.
> -     */
> -    unsigned long *clear_bmap;
> -    uint8_t clear_bmap_shift;
> -};
> +#include "exec/ramblock.h"
>  
>  /**
>   * clear_bmap_size: calculate clear bitmap size
> diff --git a/include/exec/ramblock.h b/include/exec/ramblock.h
> new file mode 100644
> index 0000000000..07d50864d8
> --- /dev/null
> +++ b/include/exec/ramblock.h
> @@ -0,0 +1,64 @@
> +/*
> + * Declarations for cpu physical memory functions
> + *
> + * Copyright 2011 Red Hat, Inc. and/or its affiliates
> + *
> + * Authors:
> + *  Avi Kivity <avi@redhat.com>
> + *
> + * This work is licensed under the terms of the GNU GPL, version 2 or
> + * later.  See the COPYING file in the top-level directory.
> + *
> + */
> +
> +/*
> + * This header is for use by exec.c and memory.c ONLY.  Do not include it.
> + * The functions declared here will be removed soon.
> + */
> +
> +#ifndef QEMU_EXEC_RAMBLOCK_H
> +#define QEMU_EXEC_RAMBLOCK_H
> +
> +#ifndef CONFIG_USER_ONLY
> +#include "cpu-common.h"
> +
> +struct RAMBlock {
> +    struct rcu_head rcu;
> +    struct MemoryRegion *mr;
> +    uint8_t *host;
> +    uint8_t *colo_cache; /* For colo, VM's ram cache */
> +    ram_addr_t offset;
> +    ram_addr_t used_length;
> +    ram_addr_t max_length;
> +    void (*resized)(const char*, uint64_t length, void *host);
> +    uint32_t flags;
> +    /* Protected by iothread lock.  */
> +    char idstr[256];
> +    /* RCU-enabled, writes protected by the ramlist lock */
> +    QLIST_ENTRY(RAMBlock) next;
> +    QLIST_HEAD(, RAMBlockNotifier) ramblock_notifiers;
> +    int fd;
> +    size_t page_size;
> +    /* dirty bitmap used during migration */
> +    unsigned long *bmap;
> +    /* bitmap of already received pages in postcopy */
> +    unsigned long *receivedmap;
> +
> +    /*
> +     * bitmap to track already cleared dirty bitmap.  When the bit is
> +     * set, it means the corresponding memory chunk needs a log-clear.
> +     * Set this up to non-NULL to enable the capability to postpone
> +     * and split clearing of dirty bitmap on the remote node (e.g.,
> +     * KVM).  The bitmap will be set only when doing global sync.
> +     *
> +     * NOTE: this bitmap is different comparing to the other bitmaps
> +     * in that one bit can represent multiple guest pages (which is
> +     * decided by the `clear_bmap_shift' variable below).  On
> +     * destination side, this should always be NULL, and the variable
> +     * `clear_bmap_shift' is meaningless.
> +     */
> +    unsigned long *clear_bmap;
> +    uint8_t clear_bmap_shift;
> +};
> +#endif
> +#endif
> -- 
> 2.24.1
> 
--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK

[PATCH v3 11/21] multifd: multifd_send_pages only needs the qemufile

[Juan Quintela]

Signed-off-by: Juan Quintela <quintela@redhat.com>
---
 migration/ram.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/migration/ram.c b/migration/ram.c
index 125c6d0f60..19caf5ed4d 100644
--- a/migration/ram.c
+++ b/migration/ram.c
@@ -929,7 +929,7 @@ struct {
  * false.
  */
 
-static int multifd_send_pages(RAMState *rs)
+static int multifd_send_pages(QEMUFile *f)
 {
     int i;
     static int next_channel;
@@ -965,7 +965,7 @@ static int multifd_send_pages(RAMState *rs)
     multifd_send_state->pages = p->pages;
     p->pages = pages;
     transferred = ((uint64_t) pages->used) * TARGET_PAGE_SIZE + p->packet_len;
-    qemu_file_update_transfer(rs->f, transferred);
+    qemu_file_update_transfer(f, transferred);
     ram_counters.multifd_bytes += transferred;
     ram_counters.transferred += transferred;;
     qemu_mutex_unlock(&p->mutex);
@@ -993,7 +993,7 @@ static int multifd_queue_page(RAMState *rs, RAMBlock *block, ram_addr_t offset)
         }
     }
 
-    if (multifd_send_pages(rs) < 0) {
+    if (multifd_send_pages(rs->f) < 0) {
         return -1;
     }
 
@@ -1090,7 +1090,7 @@ static void multifd_send_sync_main(RAMState *rs)
         return;
     }
     if (multifd_send_state->pages->used) {
-        if (multifd_send_pages(rs) < 0) {
+        if (multifd_send_pages(rs->f) < 0) {
             error_report("%s: multifd_send_pages fail", __func__);
             return;
         }
-- 
2.24.1

Re: [PATCH v3 11/21] multifd: multifd_send_pages only needs the qemufile

[Dr. David Alan Gilbert]

* Juan Quintela (quintela@redhat.com) wrote:
> Signed-off-by: Juan Quintela <quintela@redhat.com>

OK, although this is a side effect of multifd_send_state being a global
rather than part of RAMState which might have been cleaner.


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

> ---
>  migration/ram.c | 8 ++++----
>  1 file changed, 4 insertions(+), 4 deletions(-)
> 
> diff --git a/migration/ram.c b/migration/ram.c
> index 125c6d0f60..19caf5ed4d 100644
> --- a/migration/ram.c
> +++ b/migration/ram.c
> @@ -929,7 +929,7 @@ struct {
>   * false.
>   */
>  
> -static int multifd_send_pages(RAMState *rs)
> +static int multifd_send_pages(QEMUFile *f)
>  {
>      int i;
>      static int next_channel;
> @@ -965,7 +965,7 @@ static int multifd_send_pages(RAMState *rs)
>      multifd_send_state->pages = p->pages;
>      p->pages = pages;
>      transferred = ((uint64_t) pages->used) * TARGET_PAGE_SIZE + p->packet_len;
> -    qemu_file_update_transfer(rs->f, transferred);
> +    qemu_file_update_transfer(f, transferred);
>      ram_counters.multifd_bytes += transferred;
>      ram_counters.transferred += transferred;;
>      qemu_mutex_unlock(&p->mutex);
> @@ -993,7 +993,7 @@ static int multifd_queue_page(RAMState *rs, RAMBlock *block, ram_addr_t offset)
>          }
>      }
>  
> -    if (multifd_send_pages(rs) < 0) {
> +    if (multifd_send_pages(rs->f) < 0) {
>          return -1;
>      }
>  
> @@ -1090,7 +1090,7 @@ static void multifd_send_sync_main(RAMState *rs)
>          return;
>      }
>      if (multifd_send_state->pages->used) {
> -        if (multifd_send_pages(rs) < 0) {
> +        if (multifd_send_pages(rs->f) < 0) {
>              error_report("%s: multifd_send_pages fail", __func__);
>              return;
>          }
> -- 
> 2.24.1
> 
--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK

[PATCH v3 12/21] multifd: multifd_queue_page only needs the qemufile

[Juan Quintela]

Signed-off-by: Juan Quintela <quintela@redhat.com>
---
 migration/ram.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/migration/ram.c b/migration/ram.c
index 19caf5ed4d..d4c829bc77 100644
--- a/migration/ram.c
+++ b/migration/ram.c
@@ -974,7 +974,7 @@ static int multifd_send_pages(QEMUFile *f)
     return 1;
 }
 
-static int multifd_queue_page(RAMState *rs, RAMBlock *block, ram_addr_t offset)
+static int multifd_queue_page(QEMUFile *f, RAMBlock *block, ram_addr_t offset)
 {
     MultiFDPages_t *pages = multifd_send_state->pages;
 
@@ -993,12 +993,12 @@ static int multifd_queue_page(RAMState *rs, RAMBlock *block, ram_addr_t offset)
         }
     }
 
-    if (multifd_send_pages(rs->f) < 0) {
+    if (multifd_send_pages(f) < 0) {
         return -1;
     }
 
     if (pages->block != block) {
-        return  multifd_queue_page(rs, block, offset);
+        return  multifd_queue_page(f, block, offset);
     }
 
     return 1;
@@ -2128,7 +2128,7 @@ static int ram_save_page(RAMState *rs, PageSearchStatus *pss, bool last_stage)
 static int ram_save_multifd_page(RAMState *rs, RAMBlock *block,
                                  ram_addr_t offset)
 {
-    if (multifd_queue_page(rs, block, offset) < 0) {
+    if (multifd_queue_page(rs->f, block, offset) < 0) {
         return -1;
     }
     ram_counters.normal++;
-- 
2.24.1

Re: [PATCH v3 12/21] multifd: multifd_queue_page only needs the qemufile

[Dr. David Alan Gilbert]

* Juan Quintela (quintela@redhat.com) wrote:
> Signed-off-by: Juan Quintela <quintela@redhat.com>

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

> ---
>  migration/ram.c | 8 ++++----
>  1 file changed, 4 insertions(+), 4 deletions(-)
> 
> diff --git a/migration/ram.c b/migration/ram.c
> index 19caf5ed4d..d4c829bc77 100644
> --- a/migration/ram.c
> +++ b/migration/ram.c
> @@ -974,7 +974,7 @@ static int multifd_send_pages(QEMUFile *f)
>      return 1;
>  }
>  
> -static int multifd_queue_page(RAMState *rs, RAMBlock *block, ram_addr_t offset)
> +static int multifd_queue_page(QEMUFile *f, RAMBlock *block, ram_addr_t offset)
>  {
>      MultiFDPages_t *pages = multifd_send_state->pages;
>  
> @@ -993,12 +993,12 @@ static int multifd_queue_page(RAMState *rs, RAMBlock *block, ram_addr_t offset)
>          }
>      }
>  
> -    if (multifd_send_pages(rs->f) < 0) {
> +    if (multifd_send_pages(f) < 0) {
>          return -1;
>      }
>  
>      if (pages->block != block) {
> -        return  multifd_queue_page(rs, block, offset);
> +        return  multifd_queue_page(f, block, offset);
>      }
>  
>      return 1;
> @@ -2128,7 +2128,7 @@ static int ram_save_page(RAMState *rs, PageSearchStatus *pss, bool last_stage)
>  static int ram_save_multifd_page(RAMState *rs, RAMBlock *block,
>                                   ram_addr_t offset)
>  {
> -    if (multifd_queue_page(rs, block, offset) < 0) {
> +    if (multifd_queue_page(rs->f, block, offset) < 0) {
>          return -1;
>      }
>      ram_counters.normal++;
> -- 
> 2.24.1
> 
--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK

[PATCH v3 13/21] multifd: multifd_send_sync_main only needs the qemufile

[Juan Quintela]

Signed-off-by: Juan Quintela <quintela@redhat.com>
---
 migration/ram.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/migration/ram.c b/migration/ram.c
index d4c829bc77..2783dc60f4 100644
--- a/migration/ram.c
+++ b/migration/ram.c
@@ -1082,7 +1082,7 @@ void multifd_save_cleanup(void)
     multifd_send_state = NULL;
 }
 
-static void multifd_send_sync_main(RAMState *rs)
+static void multifd_send_sync_main(QEMUFile *f)
 {
     int i;
 
@@ -1090,7 +1090,7 @@ static void multifd_send_sync_main(RAMState *rs)
         return;
     }
     if (multifd_send_state->pages->used) {
-        if (multifd_send_pages(rs->f) < 0) {
+        if (multifd_send_pages(f) < 0) {
             error_report("%s: multifd_send_pages fail", __func__);
             return;
         }
@@ -1111,7 +1111,7 @@ static void multifd_send_sync_main(RAMState *rs)
         p->packet_num = multifd_send_state->packet_num++;
         p->flags |= MULTIFD_FLAG_SYNC;
         p->pending_job++;
-        qemu_file_update_transfer(rs->f, p->packet_len);
+        qemu_file_update_transfer(f, p->packet_len);
         ram_counters.multifd_bytes += p->packet_len;
         ram_counters.transferred += p->packet_len;
         qemu_mutex_unlock(&p->mutex);
@@ -3426,7 +3426,7 @@ static int ram_save_setup(QEMUFile *f, void *opaque)
     ram_control_before_iterate(f, RAM_CONTROL_SETUP);
     ram_control_after_iterate(f, RAM_CONTROL_SETUP);
 
-    multifd_send_sync_main(*rsp);
+    multifd_send_sync_main(f);
     qemu_put_be64(f, RAM_SAVE_FLAG_EOS);
     qemu_fflush(f);
 
@@ -3526,7 +3526,7 @@ static int ram_save_iterate(QEMUFile *f, void *opaque)
 out:
     if (ret >= 0
         && migration_is_setup_or_active(migrate_get_current()->state)) {
-        multifd_send_sync_main(rs);
+        multifd_send_sync_main(rs->f);
         qemu_put_be64(f, RAM_SAVE_FLAG_EOS);
         qemu_fflush(f);
         ram_counters.transferred += 8;
@@ -3585,7 +3585,7 @@ static int ram_save_complete(QEMUFile *f, void *opaque)
     }
 
     if (ret >= 0) {
-        multifd_send_sync_main(rs);
+        multifd_send_sync_main(rs->f);
         qemu_put_be64(f, RAM_SAVE_FLAG_EOS);
         qemu_fflush(f);
     }
-- 
2.24.1

Re: [PATCH v3 13/21] multifd: multifd_send_sync_main only needs the qemufile

[Dr. David Alan Gilbert]

* Juan Quintela (quintela@redhat.com) wrote:
> Signed-off-by: Juan Quintela <quintela@redhat.com>

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

> ---
>  migration/ram.c | 12 ++++++------
>  1 file changed, 6 insertions(+), 6 deletions(-)
> 
> diff --git a/migration/ram.c b/migration/ram.c
> index d4c829bc77..2783dc60f4 100644
> --- a/migration/ram.c
> +++ b/migration/ram.c
> @@ -1082,7 +1082,7 @@ void multifd_save_cleanup(void)
>      multifd_send_state = NULL;
>  }
>  
> -static void multifd_send_sync_main(RAMState *rs)
> +static void multifd_send_sync_main(QEMUFile *f)
>  {
>      int i;
>  
> @@ -1090,7 +1090,7 @@ static void multifd_send_sync_main(RAMState *rs)
>          return;
>      }
>      if (multifd_send_state->pages->used) {
> -        if (multifd_send_pages(rs->f) < 0) {
> +        if (multifd_send_pages(f) < 0) {
>              error_report("%s: multifd_send_pages fail", __func__);
>              return;
>          }
> @@ -1111,7 +1111,7 @@ static void multifd_send_sync_main(RAMState *rs)
>          p->packet_num = multifd_send_state->packet_num++;
>          p->flags |= MULTIFD_FLAG_SYNC;
>          p->pending_job++;
> -        qemu_file_update_transfer(rs->f, p->packet_len);
> +        qemu_file_update_transfer(f, p->packet_len);
>          ram_counters.multifd_bytes += p->packet_len;
>          ram_counters.transferred += p->packet_len;
>          qemu_mutex_unlock(&p->mutex);
> @@ -3426,7 +3426,7 @@ static int ram_save_setup(QEMUFile *f, void *opaque)
>      ram_control_before_iterate(f, RAM_CONTROL_SETUP);
>      ram_control_after_iterate(f, RAM_CONTROL_SETUP);
>  
> -    multifd_send_sync_main(*rsp);
> +    multifd_send_sync_main(f);
>      qemu_put_be64(f, RAM_SAVE_FLAG_EOS);
>      qemu_fflush(f);
>  
> @@ -3526,7 +3526,7 @@ static int ram_save_iterate(QEMUFile *f, void *opaque)
>  out:
>      if (ret >= 0
>          && migration_is_setup_or_active(migrate_get_current()->state)) {
> -        multifd_send_sync_main(rs);
> +        multifd_send_sync_main(rs->f);
>          qemu_put_be64(f, RAM_SAVE_FLAG_EOS);
>          qemu_fflush(f);
>          ram_counters.transferred += 8;
> @@ -3585,7 +3585,7 @@ static int ram_save_complete(QEMUFile *f, void *opaque)
>      }
>  
>      if (ret >= 0) {
> -        multifd_send_sync_main(rs);
> +        multifd_send_sync_main(rs->f);
>          qemu_put_be64(f, RAM_SAVE_FLAG_EOS);
>          qemu_fflush(f);
>      }
> -- 
> 2.24.1
> 
--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK

[PATCH v3 14/21] multifd: Use qemu_target_page_size()

[Juan Quintela]

We will make it cpu independent.

Signed-off-by: Juan Quintela <quintela@redhat.com>
---
 migration/ram.c | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

diff --git a/migration/ram.c b/migration/ram.c
index 2783dc60f4..14b7cbdbc9 100644
--- a/migration/ram.c
+++ b/migration/ram.c
@@ -882,14 +882,14 @@ static int multifd_recv_unfill_packet(MultiFDRecvParams *p, Error **errp)
     for (i = 0; i < p->pages->used; i++) {
         uint64_t offset = be64_to_cpu(packet->offset[i]);
 
-        if (offset > (block->used_length - TARGET_PAGE_SIZE)) {
+        if (offset > (block->used_length - qemu_target_page_size())) {
             error_setg(errp, "multifd: offset too long %" PRIu64
                        " (max " RAM_ADDR_FMT ")",
                        offset, block->max_length);
             return -1;
         }
         p->pages->iov[i].iov_base = block->host + offset;
-        p->pages->iov[i].iov_len = TARGET_PAGE_SIZE;
+        p->pages->iov[i].iov_len = qemu_target_page_size();
     }
 
     return 0;
@@ -964,7 +964,8 @@ static int multifd_send_pages(QEMUFile *f)
     p->packet_num = multifd_send_state->packet_num++;
     multifd_send_state->pages = p->pages;
     p->pages = pages;
-    transferred = ((uint64_t) pages->used) * TARGET_PAGE_SIZE + p->packet_len;
+    transferred = ((uint64_t) pages->used) * qemu_target_page_size()
+                + p->packet_len;
     qemu_file_update_transfer(f, transferred);
     ram_counters.multifd_bytes += transferred;
     ram_counters.transferred += transferred;;
@@ -985,7 +986,7 @@ static int multifd_queue_page(QEMUFile *f, RAMBlock *block, ram_addr_t offset)
     if (pages->block == block) {
         pages->offset[pages->used] = offset;
         pages->iov[pages->used].iov_base = block->host + offset;
-        pages->iov[pages->used].iov_len = TARGET_PAGE_SIZE;
+        pages->iov[pages->used].iov_len = qemu_target_page_size();
         pages->used++;
 
         if (pages->used < pages->allocated) {
-- 
2.24.1

Re: [PATCH v3 14/21] multifd: Use qemu_target_page_size()

[Dr. David Alan Gilbert]

* Juan Quintela (quintela@redhat.com) wrote:
> We will make it cpu independent.
> 
> Signed-off-by: Juan Quintela <quintela@redhat.com>

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

> ---
>  migration/ram.c | 9 +++++----
>  1 file changed, 5 insertions(+), 4 deletions(-)
> 
> diff --git a/migration/ram.c b/migration/ram.c
> index 2783dc60f4..14b7cbdbc9 100644
> --- a/migration/ram.c
> +++ b/migration/ram.c
> @@ -882,14 +882,14 @@ static int multifd_recv_unfill_packet(MultiFDRecvParams *p, Error **errp)
>      for (i = 0; i < p->pages->used; i++) {
>          uint64_t offset = be64_to_cpu(packet->offset[i]);
>  
> -        if (offset > (block->used_length - TARGET_PAGE_SIZE)) {
> +        if (offset > (block->used_length - qemu_target_page_size())) {
>              error_setg(errp, "multifd: offset too long %" PRIu64
>                         " (max " RAM_ADDR_FMT ")",
>                         offset, block->max_length);
>              return -1;
>          }
>          p->pages->iov[i].iov_base = block->host + offset;
> -        p->pages->iov[i].iov_len = TARGET_PAGE_SIZE;
> +        p->pages->iov[i].iov_len = qemu_target_page_size();
>      }
>  
>      return 0;
> @@ -964,7 +964,8 @@ static int multifd_send_pages(QEMUFile *f)
>      p->packet_num = multifd_send_state->packet_num++;
>      multifd_send_state->pages = p->pages;
>      p->pages = pages;
> -    transferred = ((uint64_t) pages->used) * TARGET_PAGE_SIZE + p->packet_len;
> +    transferred = ((uint64_t) pages->used) * qemu_target_page_size()
> +                + p->packet_len;
>      qemu_file_update_transfer(f, transferred);
>      ram_counters.multifd_bytes += transferred;
>      ram_counters.transferred += transferred;;
> @@ -985,7 +986,7 @@ static int multifd_queue_page(QEMUFile *f, RAMBlock *block, ram_addr_t offset)
>      if (pages->block == block) {
>          pages->offset[pages->used] = offset;
>          pages->iov[pages->used].iov_base = block->host + offset;
> -        pages->iov[pages->used].iov_len = TARGET_PAGE_SIZE;
> +        pages->iov[pages->used].iov_len = qemu_target_page_size();
>          pages->used++;
>  
>          if (pages->used < pages->allocated) {
> -- 
> 2.24.1
> 
--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK

[PATCH v3 15/21] migration: Make checkpatch happy with comments

[Juan Quintela]

Signed-off-by: Juan Quintela <quintela@redhat.com>
---
 migration/ram.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

diff --git a/migration/ram.c b/migration/ram.c
index 14b7cbdbc9..c24b4cc771 100644
--- a/migration/ram.c
+++ b/migration/ram.c
@@ -1312,10 +1312,12 @@ static void multifd_recv_terminate_threads(Error *err)
 
         qemu_mutex_lock(&p->mutex);
         p->quit = true;
-        /* We could arrive here for two reasons:
-           - normal quit, i.e. everything went fine, just finished
-           - error quit: We close the channels so the channel threads
-             finish the qio_channel_read_all_eof() */
+        /*
+         * We could arrive here for two reasons:
+         *  - normal quit, i.e. everything went fine, just finished
+         *  - error quit: We close the channels so the channel threads
+         *    finish the qio_channel_read_all_eof()
+         */
         if (p->c) {
             qio_channel_shutdown(p->c, QIO_CHANNEL_SHUTDOWN_BOTH, NULL);
         }
-- 
2.24.1

Re: [PATCH v3 15/21] migration: Make checkpatch happy with comments

[Dr. David Alan Gilbert]

* Juan Quintela (quintela@redhat.com) wrote:
> Signed-off-by: Juan Quintela <quintela@redhat.com>

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

> ---
>  migration/ram.c | 10 ++++++----
>  1 file changed, 6 insertions(+), 4 deletions(-)
> 
> diff --git a/migration/ram.c b/migration/ram.c
> index 14b7cbdbc9..c24b4cc771 100644
> --- a/migration/ram.c
> +++ b/migration/ram.c
> @@ -1312,10 +1312,12 @@ static void multifd_recv_terminate_threads(Error *err)
>  
>          qemu_mutex_lock(&p->mutex);
>          p->quit = true;
> -        /* We could arrive here for two reasons:
> -           - normal quit, i.e. everything went fine, just finished
> -           - error quit: We close the channels so the channel threads
> -             finish the qio_channel_read_all_eof() */
> +        /*
> +         * We could arrive here for two reasons:
> +         *  - normal quit, i.e. everything went fine, just finished
> +         *  - error quit: We close the channels so the channel threads
> +         *    finish the qio_channel_read_all_eof()
> +         */
>          if (p->c) {
>              qio_channel_shutdown(p->c, QIO_CHANNEL_SHUTDOWN_BOTH, NULL);
>          }
> -- 
> 2.24.1
> 
--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK

[PATCH v3 16/21] migration: Add support for modules

[Juan Quintela]

So we don't have to compile everything in, or have ifdefs

Signed-off-by: Juan Quintela <quintela@redhat.com>
---
 include/qemu/module.h | 2 ++
 vl.c                  | 1 +
 2 files changed, 3 insertions(+)

diff --git a/include/qemu/module.h b/include/qemu/module.h
index 65ba596e46..907cb5c0a5 100644
--- a/include/qemu/module.h
+++ b/include/qemu/module.h
@@ -40,6 +40,7 @@ static void __attribute__((constructor)) do_qemu_init_ ## function(void)    \
 #endif
 
 typedef enum {
+    MODULE_INIT_MIGRATION,
     MODULE_INIT_BLOCK,
     MODULE_INIT_OPTS,
     MODULE_INIT_QOM,
@@ -56,6 +57,7 @@ typedef enum {
 #define xen_backend_init(function) module_init(function, \
                                                MODULE_INIT_XEN_BACKEND)
 #define libqos_init(function) module_init(function, MODULE_INIT_LIBQOS)
+#define migration_init(function) module_init(function, MODULE_INIT_MIGRATION)
 
 #define block_module_load_one(lib) module_load_one("block-", lib)
 #define ui_module_load_one(lib) module_load_one("ui-", lib)
diff --git a/vl.c b/vl.c
index 71d3e7eefb..f331a3921f 100644
--- a/vl.c
+++ b/vl.c
@@ -2873,6 +2873,7 @@ int main(int argc, char **argv, char **envp)
     qemu_init_exec_dir(argv[0]);
 
     module_call_init(MODULE_INIT_QOM);
+    module_call_init(MODULE_INIT_MIGRATION);
 
     qemu_add_opts(&qemu_drive_opts);
     qemu_add_drive_opts(&qemu_legacy_drive_opts);
-- 
2.24.1

Re: [PATCH v3 16/21] migration: Add support for modules

[Dr. David Alan Gilbert]

* Juan Quintela (quintela@redhat.com) wrote:
> So we don't have to compile everything in, or have ifdefs

Can you explain to me what this is allowing us to do?

Dave

> Signed-off-by: Juan Quintela <quintela@redhat.com>
> ---
>  include/qemu/module.h | 2 ++
>  vl.c                  | 1 +
>  2 files changed, 3 insertions(+)
> 
> diff --git a/include/qemu/module.h b/include/qemu/module.h
> index 65ba596e46..907cb5c0a5 100644
> --- a/include/qemu/module.h
> +++ b/include/qemu/module.h
> @@ -40,6 +40,7 @@ static void __attribute__((constructor)) do_qemu_init_ ## function(void)    \
>  #endif
>  
>  typedef enum {
> +    MODULE_INIT_MIGRATION,
>      MODULE_INIT_BLOCK,
>      MODULE_INIT_OPTS,
>      MODULE_INIT_QOM,
> @@ -56,6 +57,7 @@ typedef enum {
>  #define xen_backend_init(function) module_init(function, \
>                                                 MODULE_INIT_XEN_BACKEND)
>  #define libqos_init(function) module_init(function, MODULE_INIT_LIBQOS)
> +#define migration_init(function) module_init(function, MODULE_INIT_MIGRATION)
>  
>  #define block_module_load_one(lib) module_load_one("block-", lib)
>  #define ui_module_load_one(lib) module_load_one("ui-", lib)
> diff --git a/vl.c b/vl.c
> index 71d3e7eefb..f331a3921f 100644
> --- a/vl.c
> +++ b/vl.c
> @@ -2873,6 +2873,7 @@ int main(int argc, char **argv, char **envp)
>      qemu_init_exec_dir(argv[0]);
>  
>      module_call_init(MODULE_INIT_QOM);
> +    module_call_init(MODULE_INIT_MIGRATION);
>  
>      qemu_add_opts(&qemu_drive_opts);
>      qemu_add_drive_opts(&qemu_legacy_drive_opts);
> -- 
> 2.24.1
> 
--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK

Re: [PATCH v3 16/21] migration: Add support for modules

[Juan Quintela]

"Dr. David Alan Gilbert" <dgilbert@redhat.com> wrote:
> * Juan Quintela (quintela@redhat.com) wrote:
>> So we don't have to compile everything in, or have ifdefs
>
> Can you explain to me what this is allowing us to do?

See the zstd support.
We don't need to do anything in any other file.  If you compile-in the
multifd-zstd.c code, you have it.  It is to avoid this kind of
constructs:

    if (strstart(uri, "tcp:", &p)) {
        tcp_start_outgoing_migration(s, p, &local_err);
#ifdef CONFIG_RDMA
    } else if (strstart(uri, "rdma:", &p)) {
        rdma_start_outgoing_migration(s, p, &local_err);
#endif
    } else if (strstart(uri, "exec:", &p)) {
        exec_start_outgoing_migration(s, p, &local_err);
    } else if (strstart(uri, "unix:", &p)) {
        unix_start_outgoing_migration(s, p, &local_err);
    } else if (strstart(uri, "fd:", &p)) {
        fd_start_outgoing_migration(s, p, &local_err);
    } else {
        error_setg(errp, QERR_INVALID_PARAMETER_VALUE, "uri",
                   "a valid migration protocol");
        migrate_set_state(&s->state, MIGRATION_STATUS_SETUP,
                          MIGRATION_STATUS_FAILED);
        block_cleanup_parameters(s);
        return;
    }

This is the following bit that I am going to try, put all the code of
rdma/exec/... in its own file, being able to compile it out and not
having ifdefs left and right.

(In the case of zstd, we still have some code in because I don't know
how to convince qapi of doing:

#ifdef CONFIG_ZSTD
{ 'enum': 'MultiFDMethod',
#ifdef CONFIG_ZSTD
  'data': [ 'none', 'zlib', 'zstd' ] }
#else
  'data': [ 'none', 'zlib' ] }
#endif

If I am told how to fix this, I will change

static MultiFDMethods *multifd_ops[MULTIFD_METHOD__MAX] = {
    [MULTIFD_METHOD_NONE] = &multifd_nocomp_ops,
};

And remove the last dependency.

Thanks, Juan.

[PATCH v3 17/21] multifd: Split multifd code into its own file

[Juan Quintela]

Signed-off-by: Juan Quintela <quintela@redhat.com>
---
 migration/Makefile.objs |   1 +
 migration/migration.c   |   1 +
 migration/multifd.c     | 891 ++++++++++++++++++++++++++++++++++++
 migration/multifd.h     | 139 ++++++
 migration/ram.c         | 980 +---------------------------------------
 migration/ram.h         |   7 -
 6 files changed, 1033 insertions(+), 986 deletions(-)
 create mode 100644 migration/multifd.c
 create mode 100644 migration/multifd.h

diff --git a/migration/Makefile.objs b/migration/Makefile.objs
index a4f3bafd86..d3623d5f9b 100644
--- a/migration/Makefile.objs
+++ b/migration/Makefile.objs
@@ -7,6 +7,7 @@ common-obj-y += qemu-file-channel.o
 common-obj-y += xbzrle.o postcopy-ram.o
 common-obj-y += qjson.o
 common-obj-y += block-dirty-bitmap.o
+common-obj-y += multifd.o
 
 common-obj-$(CONFIG_RDMA) += rdma.o
 
diff --git a/migration/migration.c b/migration/migration.c
index ecb56afd50..3501bc3353 100644
--- a/migration/migration.c
+++ b/migration/migration.c
@@ -53,6 +53,7 @@
 #include "monitor/monitor.h"
 #include "net/announce.h"
 #include "qemu/queue.h"
+#include "multifd.h"
 
 #define MAX_THROTTLE  (32 << 20)      /* Migration transfer speed throttling */
 
diff --git a/migration/multifd.c b/migration/multifd.c
new file mode 100644
index 0000000000..1875bb3aaa
--- /dev/null
+++ b/migration/multifd.c
@@ -0,0 +1,891 @@
+/*
+ * Multifd common code
+ *
+ * Copyright (c) 2019-2020 Red Hat Inc
+ *
+ * Authors:
+ *  Juan Quintela <quintela@redhat.com>
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ */
+
+#include "qemu/osdep.h"
+#include "qemu/rcu.h"
+#include "exec/target_page.h"
+#include "sysemu/sysemu.h"
+#include "exec/ramblock.h"
+#include "qemu/error-report.h"
+#include "qapi/error.h"
+#include "ram.h"
+#include "migration.h"
+#include "socket.h"
+#include "qemu-file.h"
+#include "trace.h"
+#include "multifd.h"
+
+/* Multiple fd's */
+
+#define MULTIFD_MAGIC 0x11223344U
+#define MULTIFD_VERSION 1
+
+typedef struct {
+    uint32_t magic;
+    uint32_t version;
+    unsigned char uuid[16]; /* QemuUUID */
+    uint8_t id;
+    uint8_t unused1[7];     /* Reserved for future use */
+    uint64_t unused2[4];    /* Reserved for future use */
+} __attribute__((packed)) MultiFDInit_t;
+
+static int multifd_send_initial_packet(MultiFDSendParams *p, Error **errp)
+{
+    MultiFDInit_t msg = {};
+    int ret;
+
+    msg.magic = cpu_to_be32(MULTIFD_MAGIC);
+    msg.version = cpu_to_be32(MULTIFD_VERSION);
+    msg.id = p->id;
+    memcpy(msg.uuid, &qemu_uuid.data, sizeof(msg.uuid));
+
+    ret = qio_channel_write_all(p->c, (char *)&msg, sizeof(msg), errp);
+    if (ret != 0) {
+        return -1;
+    }
+    return 0;
+}
+
+static int multifd_recv_initial_packet(QIOChannel *c, Error **errp)
+{
+    MultiFDInit_t msg;
+    int ret;
+
+    ret = qio_channel_read_all(c, (char *)&msg, sizeof(msg), errp);
+    if (ret != 0) {
+        return -1;
+    }
+
+    msg.magic = be32_to_cpu(msg.magic);
+    msg.version = be32_to_cpu(msg.version);
+
+    if (msg.magic != MULTIFD_MAGIC) {
+        error_setg(errp, "multifd: received packet magic %x "
+                   "expected %x", msg.magic, MULTIFD_MAGIC);
+        return -1;
+    }
+
+    if (msg.version != MULTIFD_VERSION) {
+        error_setg(errp, "multifd: received packet version %d "
+                   "expected %d", msg.version, MULTIFD_VERSION);
+        return -1;
+    }
+
+    if (memcmp(msg.uuid, &qemu_uuid, sizeof(qemu_uuid))) {
+        char *uuid = qemu_uuid_unparse_strdup(&qemu_uuid);
+        char *msg_uuid = qemu_uuid_unparse_strdup((const QemuUUID *)msg.uuid);
+
+        error_setg(errp, "multifd: received uuid '%s' and expected "
+                   "uuid '%s' for channel %hhd", msg_uuid, uuid, msg.id);
+        g_free(uuid);
+        g_free(msg_uuid);
+        return -1;
+    }
+
+    if (msg.id > migrate_multifd_channels()) {
+        error_setg(errp, "multifd: received channel version %d "
+                   "expected %d", msg.version, MULTIFD_VERSION);
+        return -1;
+    }
+
+    return msg.id;
+}
+
+static MultiFDPages_t *multifd_pages_init(size_t size)
+{
+    MultiFDPages_t *pages = g_new0(MultiFDPages_t, 1);
+
+    pages->allocated = size;
+    pages->iov = g_new0(struct iovec, size);
+    pages->offset = g_new0(ram_addr_t, size);
+
+    return pages;
+}
+
+static void multifd_pages_clear(MultiFDPages_t *pages)
+{
+    pages->used = 0;
+    pages->allocated = 0;
+    pages->packet_num = 0;
+    pages->block = NULL;
+    g_free(pages->iov);
+    pages->iov = NULL;
+    g_free(pages->offset);
+    pages->offset = NULL;
+    g_free(pages);
+}
+
+static void multifd_send_fill_packet(MultiFDSendParams *p)
+{
+    MultiFDPacket_t *packet = p->packet;
+    int i;
+
+    packet->flags = cpu_to_be32(p->flags);
+    packet->pages_alloc = cpu_to_be32(p->pages->allocated);
+    packet->pages_used = cpu_to_be32(p->pages->used);
+    packet->next_packet_size = cpu_to_be32(p->next_packet_size);
+    packet->packet_num = cpu_to_be64(p->packet_num);
+
+    if (p->pages->block) {
+        strncpy(packet->ramblock, p->pages->block->idstr, 256);
+    }
+
+    for (i = 0; i < p->pages->used; i++) {
+        /* there are architectures where ram_addr_t is 32 bit */
+        uint64_t temp = p->pages->offset[i];
+
+        packet->offset[i] = cpu_to_be64(temp);
+    }
+}
+
+static int multifd_recv_unfill_packet(MultiFDRecvParams *p, Error **errp)
+{
+    MultiFDPacket_t *packet = p->packet;
+    uint32_t pages_max = MULTIFD_PACKET_SIZE / qemu_target_page_size();
+    RAMBlock *block;
+    int i;
+
+    packet->magic = be32_to_cpu(packet->magic);
+    if (packet->magic != MULTIFD_MAGIC) {
+        error_setg(errp, "multifd: received packet "
+                   "magic %x and expected magic %x",
+                   packet->magic, MULTIFD_MAGIC);
+        return -1;
+    }
+
+    packet->version = be32_to_cpu(packet->version);
+    if (packet->version != MULTIFD_VERSION) {
+        error_setg(errp, "multifd: received packet "
+                   "version %d and expected version %d",
+                   packet->version, MULTIFD_VERSION);
+        return -1;
+    }
+
+    p->flags = be32_to_cpu(packet->flags);
+
+    packet->pages_alloc = be32_to_cpu(packet->pages_alloc);
+    /*
+     * If we received a packet that is 100 times bigger than expected
+     * just stop migration.  It is a magic number.
+     */
+    if (packet->pages_alloc > pages_max * 100) {
+        error_setg(errp, "multifd: received packet "
+                   "with size %d and expected a maximum size of %d",
+                   packet->pages_alloc, pages_max * 100) ;
+        return -1;
+    }
+    /*
+     * We received a packet that is bigger than expected but inside
+     * reasonable limits (see previous comment).  Just reallocate.
+     */
+    if (packet->pages_alloc > p->pages->allocated) {
+        multifd_pages_clear(p->pages);
+        p->pages = multifd_pages_init(packet->pages_alloc);
+    }
+
+    p->pages->used = be32_to_cpu(packet->pages_used);
+    if (p->pages->used > packet->pages_alloc) {
+        error_setg(errp, "multifd: received packet "
+                   "with %d pages and expected maximum pages are %d",
+                   p->pages->used, packet->pages_alloc) ;
+        return -1;
+    }
+
+    p->next_packet_size = be32_to_cpu(packet->next_packet_size);
+    p->packet_num = be64_to_cpu(packet->packet_num);
+
+    if (p->pages->used == 0) {
+        return 0;
+    }
+
+    /* make sure that ramblock is 0 terminated */
+    packet->ramblock[255] = 0;
+    block = qemu_ram_block_by_name(packet->ramblock);
+    if (!block) {
+        error_setg(errp, "multifd: unknown ram block %s",
+                   packet->ramblock);
+        return -1;
+    }
+
+    for (i = 0; i < p->pages->used; i++) {
+        uint64_t offset = be64_to_cpu(packet->offset[i]);
+
+        if (offset > (block->used_length - qemu_target_page_size())) {
+            error_setg(errp, "multifd: offset too long %" PRIu64
+                       " (max " RAM_ADDR_FMT ")",
+                       offset, block->max_length);
+            return -1;
+        }
+        p->pages->iov[i].iov_base = block->host + offset;
+        p->pages->iov[i].iov_len = qemu_target_page_size();
+    }
+
+    return 0;
+}
+
+struct {
+    MultiFDSendParams *params;
+    /* array of pages to sent */
+    MultiFDPages_t *pages;
+    /* global number of generated multifd packets */
+    uint64_t packet_num;
+    /* send channels ready */
+    QemuSemaphore channels_ready;
+    /*
+     * Have we already run terminate threads.  There is a race when it
+     * happens that we got one error while we are exiting.
+     * We will use atomic operations.  Only valid values are 0 and 1.
+     */
+    int exiting;
+} *multifd_send_state;
+
+/*
+ * How we use multifd_send_state->pages and channel->pages?
+ *
+ * We create a pages for each channel, and a main one.  Each time that
+ * we need to send a batch of pages we interchange the ones between
+ * multifd_send_state and the channel that is sending it.  There are
+ * two reasons for that:
+ *    - to not have to do so many mallocs during migration
+ *    - to make easier to know what to free at the end of migration
+ *
+ * This way we always know who is the owner of each "pages" struct,
+ * and we don't need any locking.  It belongs to the migration thread
+ * or to the channel thread.  Switching is safe because the migration
+ * thread is using the channel mutex when changing it, and the channel
+ * have to had finish with its own, otherwise pending_job can't be
+ * false.
+ */
+
+static int multifd_send_pages(QEMUFile *f)
+{
+    int i;
+    static int next_channel;
+    MultiFDSendParams *p = NULL; /* make happy gcc */
+    MultiFDPages_t *pages = multifd_send_state->pages;
+    uint64_t transferred;
+
+    if (atomic_read(&multifd_send_state->exiting)) {
+        return -1;
+    }
+
+    qemu_sem_wait(&multifd_send_state->channels_ready);
+    for (i = next_channel;; i = (i + 1) % migrate_multifd_channels()) {
+        p = &multifd_send_state->params[i];
+
+        qemu_mutex_lock(&p->mutex);
+        if (p->quit) {
+            error_report("%s: channel %d has already quit!", __func__, i);
+            qemu_mutex_unlock(&p->mutex);
+            return -1;
+        }
+        if (!p->pending_job) {
+            p->pending_job++;
+            next_channel = (i + 1) % migrate_multifd_channels();
+            break;
+        }
+        qemu_mutex_unlock(&p->mutex);
+    }
+    assert(!p->pages->used);
+    assert(!p->pages->block);
+
+    p->packet_num = multifd_send_state->packet_num++;
+    multifd_send_state->pages = p->pages;
+    p->pages = pages;
+    transferred = ((uint64_t) pages->used) * qemu_target_page_size()
+                + p->packet_len;
+    qemu_file_update_transfer(f, transferred);
+    ram_counters.multifd_bytes += transferred;
+    ram_counters.transferred += transferred;;
+    qemu_mutex_unlock(&p->mutex);
+    qemu_sem_post(&p->sem);
+
+    return 1;
+}
+
+int multifd_queue_page(QEMUFile *f, RAMBlock *block, ram_addr_t offset)
+{
+    MultiFDPages_t *pages = multifd_send_state->pages;
+
+    if (!pages->block) {
+        pages->block = block;
+    }
+
+    if (pages->block == block) {
+        pages->offset[pages->used] = offset;
+        pages->iov[pages->used].iov_base = block->host + offset;
+        pages->iov[pages->used].iov_len = qemu_target_page_size();
+        pages->used++;
+
+        if (pages->used < pages->allocated) {
+            return 1;
+        }
+    }
+
+    if (multifd_send_pages(f) < 0) {
+        return -1;
+    }
+
+    if (pages->block != block) {
+        return  multifd_queue_page(f, block, offset);
+    }
+
+    return 1;
+}
+
+static void multifd_send_terminate_threads(Error *err)
+{
+    int i;
+
+    trace_multifd_send_terminate_threads(err != NULL);
+
+    if (err) {
+        MigrationState *s = migrate_get_current();
+        migrate_set_error(s, err);
+        if (s->state == MIGRATION_STATUS_SETUP ||
+            s->state == MIGRATION_STATUS_PRE_SWITCHOVER ||
+            s->state == MIGRATION_STATUS_DEVICE ||
+            s->state == MIGRATION_STATUS_ACTIVE) {
+            migrate_set_state(&s->state, s->state,
+                              MIGRATION_STATUS_FAILED);
+        }
+    }
+
+    /*
+     * We don't want to exit each threads twice.  Depending on where
+     * we get the error, or if there are two independent errors in two
+     * threads at the same time, we can end calling this function
+     * twice.
+     */
+    if (atomic_xchg(&multifd_send_state->exiting, 1)) {
+        return;
+    }
+
+    for (i = 0; i < migrate_multifd_channels(); i++) {
+        MultiFDSendParams *p = &multifd_send_state->params[i];
+
+        qemu_mutex_lock(&p->mutex);
+        p->quit = true;
+        qemu_sem_post(&p->sem);
+        qemu_mutex_unlock(&p->mutex);
+    }
+}
+
+void multifd_save_cleanup(void)
+{
+    int i;
+
+    if (!migrate_use_multifd()) {
+        return;
+    }
+    multifd_send_terminate_threads(NULL);
+    for (i = 0; i < migrate_multifd_channels(); i++) {
+        MultiFDSendParams *p = &multifd_send_state->params[i];
+
+        if (p->running) {
+            qemu_thread_join(&p->thread);
+        }
+    }
+    for (i = 0; i < migrate_multifd_channels(); i++) {
+        MultiFDSendParams *p = &multifd_send_state->params[i];
+
+        socket_send_channel_destroy(p->c);
+        p->c = NULL;
+        qemu_mutex_destroy(&p->mutex);
+        qemu_sem_destroy(&p->sem);
+        qemu_sem_destroy(&p->sem_sync);
+        g_free(p->name);
+        p->name = NULL;
+        multifd_pages_clear(p->pages);
+        p->pages = NULL;
+        p->packet_len = 0;
+        g_free(p->packet);
+        p->packet = NULL;
+    }
+    qemu_sem_destroy(&multifd_send_state->channels_ready);
+    g_free(multifd_send_state->params);
+    multifd_send_state->params = NULL;
+    multifd_pages_clear(multifd_send_state->pages);
+    multifd_send_state->pages = NULL;
+    g_free(multifd_send_state);
+    multifd_send_state = NULL;
+}
+
+void multifd_send_sync_main(QEMUFile *f)
+{
+    int i;
+
+    if (!migrate_use_multifd()) {
+        return;
+    }
+    if (multifd_send_state->pages->used) {
+        if (multifd_send_pages(f) < 0) {
+            error_report("%s: multifd_send_pages fail", __func__);
+            return;
+        }
+    }
+    for (i = 0; i < migrate_multifd_channels(); i++) {
+        MultiFDSendParams *p = &multifd_send_state->params[i];
+
+        trace_multifd_send_sync_main_signal(p->id);
+
+        qemu_mutex_lock(&p->mutex);
+
+        if (p->quit) {
+            error_report("%s: channel %d has already quit", __func__, i);
+            qemu_mutex_unlock(&p->mutex);
+            return;
+        }
+
+        p->packet_num = multifd_send_state->packet_num++;
+        p->flags |= MULTIFD_FLAG_SYNC;
+        p->pending_job++;
+        qemu_file_update_transfer(f, p->packet_len);
+        ram_counters.multifd_bytes += p->packet_len;
+        ram_counters.transferred += p->packet_len;
+        qemu_mutex_unlock(&p->mutex);
+        qemu_sem_post(&p->sem);
+    }
+    for (i = 0; i < migrate_multifd_channels(); i++) {
+        MultiFDSendParams *p = &multifd_send_state->params[i];
+
+        trace_multifd_send_sync_main_wait(p->id);
+        qemu_sem_wait(&p->sem_sync);
+    }
+    trace_multifd_send_sync_main(multifd_send_state->packet_num);
+}
+
+static void *multifd_send_thread(void *opaque)
+{
+    MultiFDSendParams *p = opaque;
+    Error *local_err = NULL;
+    int ret = 0;
+    uint32_t flags = 0;
+
+    trace_multifd_send_thread_start(p->id);
+    rcu_register_thread();
+
+    if (multifd_send_initial_packet(p, &local_err) < 0) {
+        ret = -1;
+        goto out;
+    }
+    /* initial packet */
+    p->num_packets = 1;
+
+    while (true) {
+        qemu_sem_wait(&p->sem);
+
+        if (atomic_read(&multifd_send_state->exiting)) {
+            break;
+        }
+        qemu_mutex_lock(&p->mutex);
+
+        if (p->pending_job) {
+            uint32_t used = p->pages->used;
+            uint64_t packet_num = p->packet_num;
+            flags = p->flags;
+
+            p->next_packet_size = used * qemu_target_page_size();
+            multifd_send_fill_packet(p);
+            p->flags = 0;
+            p->num_packets++;
+            p->num_pages += used;
+            p->pages->used = 0;
+            p->pages->block = NULL;
+            qemu_mutex_unlock(&p->mutex);
+
+            trace_multifd_send(p->id, packet_num, used, flags,
+                               p->next_packet_size);
+
+            ret = qio_channel_write_all(p->c, (void *)p->packet,
+                                        p->packet_len, &local_err);
+            if (ret != 0) {
+                break;
+            }
+
+            if (used) {
+                ret = qio_channel_writev_all(p->c, p->pages->iov,
+                                             used, &local_err);
+                if (ret != 0) {
+                    break;
+                }
+            }
+
+            qemu_mutex_lock(&p->mutex);
+            p->pending_job--;
+            qemu_mutex_unlock(&p->mutex);
+
+            if (flags & MULTIFD_FLAG_SYNC) {
+                qemu_sem_post(&p->sem_sync);
+            }
+            qemu_sem_post(&multifd_send_state->channels_ready);
+        } else if (p->quit) {
+            qemu_mutex_unlock(&p->mutex);
+            break;
+        } else {
+            qemu_mutex_unlock(&p->mutex);
+            /* sometimes there are spurious wakeups */
+        }
+    }
+
+out:
+    if (local_err) {
+        trace_multifd_send_error(p->id);
+        multifd_send_terminate_threads(local_err);
+    }
+
+    /*
+     * Error happen, I will exit, but I can't just leave, tell
+     * who pay attention to me.
+     */
+    if (ret != 0) {
+        qemu_sem_post(&p->sem_sync);
+        qemu_sem_post(&multifd_send_state->channels_ready);
+    }
+
+    qemu_mutex_lock(&p->mutex);
+    p->running = false;
+    qemu_mutex_unlock(&p->mutex);
+
+    rcu_unregister_thread();
+    trace_multifd_send_thread_end(p->id, p->num_packets, p->num_pages);
+
+    return NULL;
+}
+
+static void multifd_new_send_channel_async(QIOTask *task, gpointer opaque)
+{
+    MultiFDSendParams *p = opaque;
+    QIOChannel *sioc = QIO_CHANNEL(qio_task_get_source(task));
+    Error *local_err = NULL;
+
+    trace_multifd_new_send_channel_async(p->id);
+    if (qio_task_propagate_error(task, &local_err)) {
+        migrate_set_error(migrate_get_current(), local_err);
+        multifd_save_cleanup();
+    } else {
+        p->c = QIO_CHANNEL(sioc);
+        qio_channel_set_delay(p->c, false);
+        p->running = true;
+        qemu_thread_create(&p->thread, p->name, multifd_send_thread, p,
+                           QEMU_THREAD_JOINABLE);
+    }
+}
+
+int multifd_save_setup(Error **errp)
+{
+    int thread_count;
+    uint32_t page_count = MULTIFD_PACKET_SIZE / qemu_target_page_size();
+    uint8_t i;
+
+    if (!migrate_use_multifd()) {
+        return 0;
+    }
+    thread_count = migrate_multifd_channels();
+    multifd_send_state = g_malloc0(sizeof(*multifd_send_state));
+    multifd_send_state->params = g_new0(MultiFDSendParams, thread_count);
+    multifd_send_state->pages = multifd_pages_init(page_count);
+    qemu_sem_init(&multifd_send_state->channels_ready, 0);
+    atomic_set(&multifd_send_state->exiting, 0);
+
+    for (i = 0; i < thread_count; i++) {
+        MultiFDSendParams *p = &multifd_send_state->params[i];
+
+        qemu_mutex_init(&p->mutex);
+        qemu_sem_init(&p->sem, 0);
+        qemu_sem_init(&p->sem_sync, 0);
+        p->quit = false;
+        p->pending_job = 0;
+        p->id = i;
+        p->pages = multifd_pages_init(page_count);
+        p->packet_len = sizeof(MultiFDPacket_t)
+                      + sizeof(uint64_t) * page_count;
+        p->packet = g_malloc0(p->packet_len);
+        p->packet->magic = cpu_to_be32(MULTIFD_MAGIC);
+        p->packet->version = cpu_to_be32(MULTIFD_VERSION);
+        p->name = g_strdup_printf("multifdsend_%d", i);
+        socket_send_channel_create(multifd_new_send_channel_async, p);
+    }
+    return 0;
+}
+
+struct {
+    MultiFDRecvParams *params;
+    /* number of created threads */
+    int count;
+    /* syncs main thread and channels */
+    QemuSemaphore sem_sync;
+    /* global number of generated multifd packets */
+    uint64_t packet_num;
+} *multifd_recv_state;
+
+static void multifd_recv_terminate_threads(Error *err)
+{
+    int i;
+
+    trace_multifd_recv_terminate_threads(err != NULL);
+
+    if (err) {
+        MigrationState *s = migrate_get_current();
+        migrate_set_error(s, err);
+        if (s->state == MIGRATION_STATUS_SETUP ||
+            s->state == MIGRATION_STATUS_ACTIVE) {
+            migrate_set_state(&s->state, s->state,
+                              MIGRATION_STATUS_FAILED);
+        }
+    }
+
+    for (i = 0; i < migrate_multifd_channels(); i++) {
+        MultiFDRecvParams *p = &multifd_recv_state->params[i];
+
+        qemu_mutex_lock(&p->mutex);
+        p->quit = true;
+        /*
+         * We could arrive here for two reasons:
+         *  - normal quit, i.e. everything went fine, just finished
+         *  - error quit: We close the channels so the channel threads
+         *    finish the qio_channel_read_all_eof()
+         */
+        if (p->c) {
+            qio_channel_shutdown(p->c, QIO_CHANNEL_SHUTDOWN_BOTH, NULL);
+        }
+        qemu_mutex_unlock(&p->mutex);
+    }
+}
+
+int multifd_load_cleanup(Error **errp)
+{
+    int i;
+    int ret = 0;
+
+    if (!migrate_use_multifd()) {
+        return 0;
+    }
+    multifd_recv_terminate_threads(NULL);
+    for (i = 0; i < migrate_multifd_channels(); i++) {
+        MultiFDRecvParams *p = &multifd_recv_state->params[i];
+
+        if (p->running) {
+            p->quit = true;
+            /*
+             * multifd_recv_thread may hung at MULTIFD_FLAG_SYNC handle code,
+             * however try to wakeup it without harm in cleanup phase.
+             */
+            qemu_sem_post(&p->sem_sync);
+            qemu_thread_join(&p->thread);
+        }
+    }
+    for (i = 0; i < migrate_multifd_channels(); i++) {
+        MultiFDRecvParams *p = &multifd_recv_state->params[i];
+
+        object_unref(OBJECT(p->c));
+        p->c = NULL;
+        qemu_mutex_destroy(&p->mutex);
+        qemu_sem_destroy(&p->sem_sync);
+        g_free(p->name);
+        p->name = NULL;
+        multifd_pages_clear(p->pages);
+        p->pages = NULL;
+        p->packet_len = 0;
+        g_free(p->packet);
+        p->packet = NULL;
+    }
+    qemu_sem_destroy(&multifd_recv_state->sem_sync);
+    g_free(multifd_recv_state->params);
+    multifd_recv_state->params = NULL;
+    g_free(multifd_recv_state);
+    multifd_recv_state = NULL;
+
+    return ret;
+}
+
+void multifd_recv_sync_main(void)
+{
+    int i;
+
+    if (!migrate_use_multifd()) {
+        return;
+    }
+    for (i = 0; i < migrate_multifd_channels(); i++) {
+        MultiFDRecvParams *p = &multifd_recv_state->params[i];
+
+        trace_multifd_recv_sync_main_wait(p->id);
+        qemu_sem_wait(&multifd_recv_state->sem_sync);
+    }
+    for (i = 0; i < migrate_multifd_channels(); i++) {
+        MultiFDRecvParams *p = &multifd_recv_state->params[i];
+
+        qemu_mutex_lock(&p->mutex);
+        if (multifd_recv_state->packet_num < p->packet_num) {
+            multifd_recv_state->packet_num = p->packet_num;
+        }
+        qemu_mutex_unlock(&p->mutex);
+        trace_multifd_recv_sync_main_signal(p->id);
+        qemu_sem_post(&p->sem_sync);
+    }
+    trace_multifd_recv_sync_main(multifd_recv_state->packet_num);
+}
+
+static void *multifd_recv_thread(void *opaque)
+{
+    MultiFDRecvParams *p = opaque;
+    Error *local_err = NULL;
+    int ret;
+
+    trace_multifd_recv_thread_start(p->id);
+    rcu_register_thread();
+
+    while (true) {
+        uint32_t used;
+        uint32_t flags;
+
+        if (p->quit) {
+            break;
+        }
+
+        ret = qio_channel_read_all_eof(p->c, (void *)p->packet,
+                                       p->packet_len, &local_err);
+        if (ret == 0) {   /* EOF */
+            break;
+        }
+        if (ret == -1) {   /* Error */
+            break;
+        }
+
+        qemu_mutex_lock(&p->mutex);
+        ret = multifd_recv_unfill_packet(p, &local_err);
+        if (ret) {
+            qemu_mutex_unlock(&p->mutex);
+            break;
+        }
+
+        used = p->pages->used;
+        flags = p->flags;
+        trace_multifd_recv(p->id, p->packet_num, used, flags,
+                           p->next_packet_size);
+        p->num_packets++;
+        p->num_pages += used;
+        qemu_mutex_unlock(&p->mutex);
+
+        if (used) {
+            ret = qio_channel_readv_all(p->c, p->pages->iov,
+                                        used, &local_err);
+            if (ret != 0) {
+                break;
+            }
+        }
+
+        if (flags & MULTIFD_FLAG_SYNC) {
+            qemu_sem_post(&multifd_recv_state->sem_sync);
+            qemu_sem_wait(&p->sem_sync);
+        }
+    }
+
+    if (local_err) {
+        multifd_recv_terminate_threads(local_err);
+    }
+    qemu_mutex_lock(&p->mutex);
+    p->running = false;
+    qemu_mutex_unlock(&p->mutex);
+
+    rcu_unregister_thread();
+    trace_multifd_recv_thread_end(p->id, p->num_packets, p->num_pages);
+
+    return NULL;
+}
+
+int multifd_load_setup(Error **errp)
+{
+    int thread_count;
+    uint32_t page_count = MULTIFD_PACKET_SIZE / qemu_target_page_size();
+    uint8_t i;
+
+    if (!migrate_use_multifd()) {
+        return 0;
+    }
+    thread_count = migrate_multifd_channels();
+    multifd_recv_state = g_malloc0(sizeof(*multifd_recv_state));
+    multifd_recv_state->params = g_new0(MultiFDRecvParams, thread_count);
+    atomic_set(&multifd_recv_state->count, 0);
+    qemu_sem_init(&multifd_recv_state->sem_sync, 0);
+
+    for (i = 0; i < thread_count; i++) {
+        MultiFDRecvParams *p = &multifd_recv_state->params[i];
+
+        qemu_mutex_init(&p->mutex);
+        qemu_sem_init(&p->sem_sync, 0);
+        p->quit = false;
+        p->id = i;
+        p->pages = multifd_pages_init(page_count);
+        p->packet_len = sizeof(MultiFDPacket_t)
+                      + sizeof(uint64_t) * page_count;
+        p->packet = g_malloc0(p->packet_len);
+        p->name = g_strdup_printf("multifdrecv_%d", i);
+    }
+    return 0;
+}
+
+bool multifd_recv_all_channels_created(void)
+{
+    int thread_count = migrate_multifd_channels();
+
+    if (!migrate_use_multifd()) {
+        return true;
+    }
+
+    return thread_count == atomic_read(&multifd_recv_state->count);
+}
+
+/*
+ * Try to receive all multifd channels to get ready for the migration.
+ * - Return true and do not set @errp when correctly receving all channels;
+ * - Return false and do not set @errp when correctly receiving the current one;
+ * - Return false and set @errp when failing to receive the current channel.
+ */
+bool multifd_recv_new_channel(QIOChannel *ioc, Error **errp)
+{
+    MultiFDRecvParams *p;
+    Error *local_err = NULL;
+    int id;
+
+    id = multifd_recv_initial_packet(ioc, &local_err);
+    if (id < 0) {
+        multifd_recv_terminate_threads(local_err);
+        error_propagate_prepend(errp, local_err,
+                                "failed to receive packet"
+                                " via multifd channel %d: ",
+                                atomic_read(&multifd_recv_state->count));
+        return false;
+    }
+    trace_multifd_recv_new_channel(id);
+
+    p = &multifd_recv_state->params[id];
+    if (p->c != NULL) {
+        error_setg(&local_err, "multifd: received id '%d' already setup'",
+                   id);
+        multifd_recv_terminate_threads(local_err);
+        error_propagate(errp, local_err);
+        return false;
+    }
+    p->c = ioc;
+    object_ref(OBJECT(ioc));
+    /* initial packet */
+    p->num_packets = 1;
+
+    p->running = true;
+    qemu_thread_create(&p->thread, p->name, multifd_recv_thread, p,
+                       QEMU_THREAD_JOINABLE);
+    atomic_inc(&multifd_recv_state->count);
+    return atomic_read(&multifd_recv_state->count) ==
+           migrate_multifd_channels();
+}
+
diff --git a/migration/multifd.h b/migration/multifd.h
new file mode 100644
index 0000000000..d8b0205977
--- /dev/null
+++ b/migration/multifd.h
@@ -0,0 +1,139 @@
+/*
+ * Multifd common functions
+ *
+ * Copyright (c) 2019-2020 Red Hat Inc
+ *
+ * Authors:
+ *  Juan Quintela <quintela@redhat.com>
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ */
+
+#ifndef QEMU_MIGRATION_MULTIFD_H
+#define QEMU_MIGRATION_MULTIFD_H
+
+int multifd_save_setup(Error **errp);
+void multifd_save_cleanup(void);
+int multifd_load_setup(Error **errp);
+int multifd_load_cleanup(Error **errp);
+bool multifd_recv_all_channels_created(void);
+bool multifd_recv_new_channel(QIOChannel *ioc, Error **errp);
+void multifd_recv_sync_main(void);
+void multifd_send_sync_main(QEMUFile *f);
+int multifd_queue_page(QEMUFile *f, RAMBlock *block, ram_addr_t offset);
+
+#define MULTIFD_FLAG_SYNC (1 << 0)
+
+/* This value needs to be a multiple of qemu_target_page_size() */
+#define MULTIFD_PACKET_SIZE (512 * 1024)
+
+typedef struct {
+    uint32_t magic;
+    uint32_t version;
+    uint32_t flags;
+    /* maximum number of allocated pages */
+    uint32_t pages_alloc;
+    uint32_t pages_used;
+    /* size of the next packet that contains pages */
+    uint32_t next_packet_size;
+    uint64_t packet_num;
+    uint64_t unused[4];    /* Reserved for future use */
+    char ramblock[256];
+    uint64_t offset[];
+} __attribute__((packed)) MultiFDPacket_t;
+
+typedef struct {
+    /* number of used pages */
+    uint32_t used;
+    /* number of allocated pages */
+    uint32_t allocated;
+    /* global number of generated multifd packets */
+    uint64_t packet_num;
+    /* offset of each page */
+    ram_addr_t *offset;
+    /* pointer to each page */
+    struct iovec *iov;
+    RAMBlock *block;
+} MultiFDPages_t;
+
+typedef struct {
+    /* this fields are not changed once the thread is created */
+    /* channel number */
+    uint8_t id;
+    /* channel thread name */
+    char *name;
+    /* channel thread id */
+    QemuThread thread;
+    /* communication channel */
+    QIOChannel *c;
+    /* sem where to wait for more work */
+    QemuSemaphore sem;
+    /* this mutex protects the following parameters */
+    QemuMutex mutex;
+    /* is this channel thread running */
+    bool running;
+    /* should this thread finish */
+    bool quit;
+    /* thread has work to do */
+    int pending_job;
+    /* array of pages to sent */
+    MultiFDPages_t *pages;
+    /* packet allocated len */
+    uint32_t packet_len;
+    /* pointer to the packet */
+    MultiFDPacket_t *packet;
+    /* multifd flags for each packet */
+    uint32_t flags;
+    /* size of the next packet that contains pages */
+    uint32_t next_packet_size;
+    /* global number of generated multifd packets */
+    uint64_t packet_num;
+    /* thread local variables */
+    /* packets sent through this channel */
+    uint64_t num_packets;
+    /* pages sent through this channel */
+    uint64_t num_pages;
+    /* syncs main thread and channels */
+    QemuSemaphore sem_sync;
+}  MultiFDSendParams;
+
+typedef struct {
+    /* this fields are not changed once the thread is created */
+    /* channel number */
+    uint8_t id;
+    /* channel thread name */
+    char *name;
+    /* channel thread id */
+    QemuThread thread;
+    /* communication channel */
+    QIOChannel *c;
+    /* this mutex protects the following parameters */
+    QemuMutex mutex;
+    /* is this channel thread running */
+    bool running;
+    /* should this thread finish */
+    bool quit;
+    /* array of pages to receive */
+    MultiFDPages_t *pages;
+    /* packet allocated len */
+    uint32_t packet_len;
+    /* pointer to the packet */
+    MultiFDPacket_t *packet;
+    /* multifd flags for each packet */
+    uint32_t flags;
+    /* global number of generated multifd packets */
+    uint64_t packet_num;
+    /* thread local variables */
+    /* size of the next packet that contains pages */
+    uint32_t next_packet_size;
+    /* packets sent through this channel */
+    uint64_t num_packets;
+    /* pages sent through this channel */
+    uint64_t num_pages;
+    /* syncs main thread and channels */
+    QemuSemaphore sem_sync;
+} MultiFDRecvParams;
+
+#endif
+
diff --git a/migration/ram.c b/migration/ram.c
index c24b4cc771..ed23ed1c7c 100644
--- a/migration/ram.c
+++ b/migration/ram.c
@@ -36,7 +36,6 @@
 #include "xbzrle.h"
 #include "ram.h"
 #include "migration.h"
-#include "socket.h"
 #include "migration/register.h"
 #include "migration/misc.h"
 #include "qemu-file.h"
@@ -53,9 +52,9 @@
 #include "migration/colo.h"
 #include "block.h"
 #include "sysemu/sysemu.h"
-#include "qemu/uuid.h"
 #include "savevm.h"
 #include "qemu/iov.h"
+#include "multifd.h"
 
 /***********************************************************/
 /* ram save/restore */
@@ -575,983 +574,6 @@ exit:
     return -1;
 }
 
-/* Multiple fd's */
-
-#define MULTIFD_MAGIC 0x11223344U
-#define MULTIFD_VERSION 1
-
-#define MULTIFD_FLAG_SYNC (1 << 0)
-
-/* This value needs to be a multiple of qemu_target_page_size() */
-#define MULTIFD_PACKET_SIZE (512 * 1024)
-
-typedef struct {
-    uint32_t magic;
-    uint32_t version;
-    unsigned char uuid[16]; /* QemuUUID */
-    uint8_t id;
-    uint8_t unused1[7];     /* Reserved for future use */
-    uint64_t unused2[4];    /* Reserved for future use */
-} __attribute__((packed)) MultiFDInit_t;
-
-typedef struct {
-    uint32_t magic;
-    uint32_t version;
-    uint32_t flags;
-    /* maximum number of allocated pages */
-    uint32_t pages_alloc;
-    uint32_t pages_used;
-    /* size of the next packet that contains pages */
-    uint32_t next_packet_size;
-    uint64_t packet_num;
-    uint64_t unused[4];    /* Reserved for future use */
-    char ramblock[256];
-    uint64_t offset[];
-} __attribute__((packed)) MultiFDPacket_t;
-
-typedef struct {
-    /* number of used pages */
-    uint32_t used;
-    /* number of allocated pages */
-    uint32_t allocated;
-    /* global number of generated multifd packets */
-    uint64_t packet_num;
-    /* offset of each page */
-    ram_addr_t *offset;
-    /* pointer to each page */
-    struct iovec *iov;
-    RAMBlock *block;
-} MultiFDPages_t;
-
-typedef struct {
-    /* this fields are not changed once the thread is created */
-    /* channel number */
-    uint8_t id;
-    /* channel thread name */
-    char *name;
-    /* channel thread id */
-    QemuThread thread;
-    /* communication channel */
-    QIOChannel *c;
-    /* sem where to wait for more work */
-    QemuSemaphore sem;
-    /* this mutex protects the following parameters */
-    QemuMutex mutex;
-    /* is this channel thread running */
-    bool running;
-    /* should this thread finish */
-    bool quit;
-    /* thread has work to do */
-    int pending_job;
-    /* array of pages to sent */
-    MultiFDPages_t *pages;
-    /* packet allocated len */
-    uint32_t packet_len;
-    /* pointer to the packet */
-    MultiFDPacket_t *packet;
-    /* multifd flags for each packet */
-    uint32_t flags;
-    /* size of the next packet that contains pages */
-    uint32_t next_packet_size;
-    /* global number of generated multifd packets */
-    uint64_t packet_num;
-    /* thread local variables */
-    /* packets sent through this channel */
-    uint64_t num_packets;
-    /* pages sent through this channel */
-    uint64_t num_pages;
-    /* syncs main thread and channels */
-    QemuSemaphore sem_sync;
-}  MultiFDSendParams;
-
-typedef struct {
-    /* this fields are not changed once the thread is created */
-    /* channel number */
-    uint8_t id;
-    /* channel thread name */
-    char *name;
-    /* channel thread id */
-    QemuThread thread;
-    /* communication channel */
-    QIOChannel *c;
-    /* this mutex protects the following parameters */
-    QemuMutex mutex;
-    /* is this channel thread running */
-    bool running;
-    /* should this thread finish */
-    bool quit;
-    /* array of pages to receive */
-    MultiFDPages_t *pages;
-    /* packet allocated len */
-    uint32_t packet_len;
-    /* pointer to the packet */
-    MultiFDPacket_t *packet;
-    /* multifd flags for each packet */
-    uint32_t flags;
-    /* global number of generated multifd packets */
-    uint64_t packet_num;
-    /* thread local variables */
-    /* size of the next packet that contains pages */
-    uint32_t next_packet_size;
-    /* packets sent through this channel */
-    uint64_t num_packets;
-    /* pages sent through this channel */
-    uint64_t num_pages;
-    /* syncs main thread and channels */
-    QemuSemaphore sem_sync;
-} MultiFDRecvParams;
-
-static int multifd_send_initial_packet(MultiFDSendParams *p, Error **errp)
-{
-    MultiFDInit_t msg = {};
-    int ret;
-
-    msg.magic = cpu_to_be32(MULTIFD_MAGIC);
-    msg.version = cpu_to_be32(MULTIFD_VERSION);
-    msg.id = p->id;
-    memcpy(msg.uuid, &qemu_uuid.data, sizeof(msg.uuid));
-
-    ret = qio_channel_write_all(p->c, (char *)&msg, sizeof(msg), errp);
-    if (ret != 0) {
-        return -1;
-    }
-    return 0;
-}
-
-static int multifd_recv_initial_packet(QIOChannel *c, Error **errp)
-{
-    MultiFDInit_t msg;
-    int ret;
-
-    ret = qio_channel_read_all(c, (char *)&msg, sizeof(msg), errp);
-    if (ret != 0) {
-        return -1;
-    }
-
-    msg.magic = be32_to_cpu(msg.magic);
-    msg.version = be32_to_cpu(msg.version);
-
-    if (msg.magic != MULTIFD_MAGIC) {
-        error_setg(errp, "multifd: received packet magic %x "
-                   "expected %x", msg.magic, MULTIFD_MAGIC);
-        return -1;
-    }
-
-    if (msg.version != MULTIFD_VERSION) {
-        error_setg(errp, "multifd: received packet version %d "
-                   "expected %d", msg.version, MULTIFD_VERSION);
-        return -1;
-    }
-
-    if (memcmp(msg.uuid, &qemu_uuid, sizeof(qemu_uuid))) {
-        char *uuid = qemu_uuid_unparse_strdup(&qemu_uuid);
-        char *msg_uuid = qemu_uuid_unparse_strdup((const QemuUUID *)msg.uuid);
-
-        error_setg(errp, "multifd: received uuid '%s' and expected "
-                   "uuid '%s' for channel %hhd", msg_uuid, uuid, msg.id);
-        g_free(uuid);
-        g_free(msg_uuid);
-        return -1;
-    }
-
-    if (msg.id > migrate_multifd_channels()) {
-        error_setg(errp, "multifd: received channel version %d "
-                   "expected %d", msg.version, MULTIFD_VERSION);
-        return -1;
-    }
-
-    return msg.id;
-}
-
-static MultiFDPages_t *multifd_pages_init(size_t size)
-{
-    MultiFDPages_t *pages = g_new0(MultiFDPages_t, 1);
-
-    pages->allocated = size;
-    pages->iov = g_new0(struct iovec, size);
-    pages->offset = g_new0(ram_addr_t, size);
-
-    return pages;
-}
-
-static void multifd_pages_clear(MultiFDPages_t *pages)
-{
-    pages->used = 0;
-    pages->allocated = 0;
-    pages->packet_num = 0;
-    pages->block = NULL;
-    g_free(pages->iov);
-    pages->iov = NULL;
-    g_free(pages->offset);
-    pages->offset = NULL;
-    g_free(pages);
-}
-
-static void multifd_send_fill_packet(MultiFDSendParams *p)
-{
-    MultiFDPacket_t *packet = p->packet;
-    int i;
-
-    packet->flags = cpu_to_be32(p->flags);
-    packet->pages_alloc = cpu_to_be32(p->pages->allocated);
-    packet->pages_used = cpu_to_be32(p->pages->used);
-    packet->next_packet_size = cpu_to_be32(p->next_packet_size);
-    packet->packet_num = cpu_to_be64(p->packet_num);
-
-    if (p->pages->block) {
-        strncpy(packet->ramblock, p->pages->block->idstr, 256);
-    }
-
-    for (i = 0; i < p->pages->used; i++) {
-        /* there are architectures where ram_addr_t is 32 bit */
-        uint64_t temp = p->pages->offset[i];
-
-        packet->offset[i] = cpu_to_be64(temp);
-    }
-}
-
-static int multifd_recv_unfill_packet(MultiFDRecvParams *p, Error **errp)
-{
-    MultiFDPacket_t *packet = p->packet;
-    uint32_t pages_max = MULTIFD_PACKET_SIZE / qemu_target_page_size();
-    RAMBlock *block;
-    int i;
-
-    packet->magic = be32_to_cpu(packet->magic);
-    if (packet->magic != MULTIFD_MAGIC) {
-        error_setg(errp, "multifd: received packet "
-                   "magic %x and expected magic %x",
-                   packet->magic, MULTIFD_MAGIC);
-        return -1;
-    }
-
-    packet->version = be32_to_cpu(packet->version);
-    if (packet->version != MULTIFD_VERSION) {
-        error_setg(errp, "multifd: received packet "
-                   "version %d and expected version %d",
-                   packet->version, MULTIFD_VERSION);
-        return -1;
-    }
-
-    p->flags = be32_to_cpu(packet->flags);
-
-    packet->pages_alloc = be32_to_cpu(packet->pages_alloc);
-    /*
-     * If we received a packet that is 100 times bigger than expected
-     * just stop migration.  It is a magic number.
-     */
-    if (packet->pages_alloc > pages_max * 100) {
-        error_setg(errp, "multifd: received packet "
-                   "with size %d and expected a maximum size of %d",
-                   packet->pages_alloc, pages_max * 100) ;
-        return -1;
-    }
-    /*
-     * We received a packet that is bigger than expected but inside
-     * reasonable limits (see previous comment).  Just reallocate.
-     */
-    if (packet->pages_alloc > p->pages->allocated) {
-        multifd_pages_clear(p->pages);
-        p->pages = multifd_pages_init(packet->pages_alloc);
-    }
-
-    p->pages->used = be32_to_cpu(packet->pages_used);
-    if (p->pages->used > packet->pages_alloc) {
-        error_setg(errp, "multifd: received packet "
-                   "with %d pages and expected maximum pages are %d",
-                   p->pages->used, packet->pages_alloc) ;
-        return -1;
-    }
-
-    p->next_packet_size = be32_to_cpu(packet->next_packet_size);
-    p->packet_num = be64_to_cpu(packet->packet_num);
-
-    if (p->pages->used == 0) {
-        return 0;
-    }
-
-    /* make sure that ramblock is 0 terminated */
-    packet->ramblock[255] = 0;
-    block = qemu_ram_block_by_name(packet->ramblock);
-    if (!block) {
-        error_setg(errp, "multifd: unknown ram block %s",
-                   packet->ramblock);
-        return -1;
-    }
-
-    for (i = 0; i < p->pages->used; i++) {
-        uint64_t offset = be64_to_cpu(packet->offset[i]);
-
-        if (offset > (block->used_length - qemu_target_page_size())) {
-            error_setg(errp, "multifd: offset too long %" PRIu64
-                       " (max " RAM_ADDR_FMT ")",
-                       offset, block->max_length);
-            return -1;
-        }
-        p->pages->iov[i].iov_base = block->host + offset;
-        p->pages->iov[i].iov_len = qemu_target_page_size();
-    }
-
-    return 0;
-}
-
-struct {
-    MultiFDSendParams *params;
-    /* array of pages to sent */
-    MultiFDPages_t *pages;
-    /* global number of generated multifd packets */
-    uint64_t packet_num;
-    /* send channels ready */
-    QemuSemaphore channels_ready;
-    /*
-     * Have we already run terminate threads.  There is a race when it
-     * happens that we got one error while we are exiting.
-     * We will use atomic operations.  Only valid values are 0 and 1.
-     */
-    int exiting;
-} *multifd_send_state;
-
-/*
- * How we use multifd_send_state->pages and channel->pages?
- *
- * We create a pages for each channel, and a main one.  Each time that
- * we need to send a batch of pages we interchange the ones between
- * multifd_send_state and the channel that is sending it.  There are
- * two reasons for that:
- *    - to not have to do so many mallocs during migration
- *    - to make easier to know what to free at the end of migration
- *
- * This way we always know who is the owner of each "pages" struct,
- * and we don't need any locking.  It belongs to the migration thread
- * or to the channel thread.  Switching is safe because the migration
- * thread is using the channel mutex when changing it, and the channel
- * have to had finish with its own, otherwise pending_job can't be
- * false.
- */
-
-static int multifd_send_pages(QEMUFile *f)
-{
-    int i;
-    static int next_channel;
-    MultiFDSendParams *p = NULL; /* make happy gcc */
-    MultiFDPages_t *pages = multifd_send_state->pages;
-    uint64_t transferred;
-
-    if (atomic_read(&multifd_send_state->exiting)) {
-        return -1;
-    }
-
-    qemu_sem_wait(&multifd_send_state->channels_ready);
-    for (i = next_channel;; i = (i + 1) % migrate_multifd_channels()) {
-        p = &multifd_send_state->params[i];
-
-        qemu_mutex_lock(&p->mutex);
-        if (p->quit) {
-            error_report("%s: channel %d has already quit!", __func__, i);
-            qemu_mutex_unlock(&p->mutex);
-            return -1;
-        }
-        if (!p->pending_job) {
-            p->pending_job++;
-            next_channel = (i + 1) % migrate_multifd_channels();
-            break;
-        }
-        qemu_mutex_unlock(&p->mutex);
-    }
-    assert(!p->pages->used);
-    assert(!p->pages->block);
-
-    p->packet_num = multifd_send_state->packet_num++;
-    multifd_send_state->pages = p->pages;
-    p->pages = pages;
-    transferred = ((uint64_t) pages->used) * qemu_target_page_size()
-                + p->packet_len;
-    qemu_file_update_transfer(f, transferred);
-    ram_counters.multifd_bytes += transferred;
-    ram_counters.transferred += transferred;;
-    qemu_mutex_unlock(&p->mutex);
-    qemu_sem_post(&p->sem);
-
-    return 1;
-}
-
-static int multifd_queue_page(QEMUFile *f, RAMBlock *block, ram_addr_t offset)
-{
-    MultiFDPages_t *pages = multifd_send_state->pages;
-
-    if (!pages->block) {
-        pages->block = block;
-    }
-
-    if (pages->block == block) {
-        pages->offset[pages->used] = offset;
-        pages->iov[pages->used].iov_base = block->host + offset;
-        pages->iov[pages->used].iov_len = qemu_target_page_size();
-        pages->used++;
-
-        if (pages->used < pages->allocated) {
-            return 1;
-        }
-    }
-
-    if (multifd_send_pages(f) < 0) {
-        return -1;
-    }
-
-    if (pages->block != block) {
-        return  multifd_queue_page(f, block, offset);
-    }
-
-    return 1;
-}
-
-static void multifd_send_terminate_threads(Error *err)
-{
-    int i;
-
-    trace_multifd_send_terminate_threads(err != NULL);
-
-    if (err) {
-        MigrationState *s = migrate_get_current();
-        migrate_set_error(s, err);
-        if (s->state == MIGRATION_STATUS_SETUP ||
-            s->state == MIGRATION_STATUS_PRE_SWITCHOVER ||
-            s->state == MIGRATION_STATUS_DEVICE ||
-            s->state == MIGRATION_STATUS_ACTIVE) {
-            migrate_set_state(&s->state, s->state,
-                              MIGRATION_STATUS_FAILED);
-        }
-    }
-
-    /*
-     * We don't want to exit each threads twice.  Depending on where
-     * we get the error, or if there are two independent errors in two
-     * threads at the same time, we can end calling this function
-     * twice.
-     */
-    if (atomic_xchg(&multifd_send_state->exiting, 1)) {
-        return;
-    }
-
-    for (i = 0; i < migrate_multifd_channels(); i++) {
-        MultiFDSendParams *p = &multifd_send_state->params[i];
-
-        qemu_mutex_lock(&p->mutex);
-        p->quit = true;
-        qemu_sem_post(&p->sem);
-        qemu_mutex_unlock(&p->mutex);
-    }
-}
-
-void multifd_save_cleanup(void)
-{
-    int i;
-
-    if (!migrate_use_multifd()) {
-        return;
-    }
-    multifd_send_terminate_threads(NULL);
-    for (i = 0; i < migrate_multifd_channels(); i++) {
-        MultiFDSendParams *p = &multifd_send_state->params[i];
-
-        if (p->running) {
-            qemu_thread_join(&p->thread);
-        }
-    }
-    for (i = 0; i < migrate_multifd_channels(); i++) {
-        MultiFDSendParams *p = &multifd_send_state->params[i];
-
-        socket_send_channel_destroy(p->c);
-        p->c = NULL;
-        qemu_mutex_destroy(&p->mutex);
-        qemu_sem_destroy(&p->sem);
-        qemu_sem_destroy(&p->sem_sync);
-        g_free(p->name);
-        p->name = NULL;
-        multifd_pages_clear(p->pages);
-        p->pages = NULL;
-        p->packet_len = 0;
-        g_free(p->packet);
-        p->packet = NULL;
-    }
-    qemu_sem_destroy(&multifd_send_state->channels_ready);
-    g_free(multifd_send_state->params);
-    multifd_send_state->params = NULL;
-    multifd_pages_clear(multifd_send_state->pages);
-    multifd_send_state->pages = NULL;
-    g_free(multifd_send_state);
-    multifd_send_state = NULL;
-}
-
-static void multifd_send_sync_main(QEMUFile *f)
-{
-    int i;
-
-    if (!migrate_use_multifd()) {
-        return;
-    }
-    if (multifd_send_state->pages->used) {
-        if (multifd_send_pages(f) < 0) {
-            error_report("%s: multifd_send_pages fail", __func__);
-            return;
-        }
-    }
-    for (i = 0; i < migrate_multifd_channels(); i++) {
-        MultiFDSendParams *p = &multifd_send_state->params[i];
-
-        trace_multifd_send_sync_main_signal(p->id);
-
-        qemu_mutex_lock(&p->mutex);
-
-        if (p->quit) {
-            error_report("%s: channel %d has already quit", __func__, i);
-            qemu_mutex_unlock(&p->mutex);
-            return;
-        }
-
-        p->packet_num = multifd_send_state->packet_num++;
-        p->flags |= MULTIFD_FLAG_SYNC;
-        p->pending_job++;
-        qemu_file_update_transfer(f, p->packet_len);
-        ram_counters.multifd_bytes += p->packet_len;
-        ram_counters.transferred += p->packet_len;
-        qemu_mutex_unlock(&p->mutex);
-        qemu_sem_post(&p->sem);
-    }
-    for (i = 0; i < migrate_multifd_channels(); i++) {
-        MultiFDSendParams *p = &multifd_send_state->params[i];
-
-        trace_multifd_send_sync_main_wait(p->id);
-        qemu_sem_wait(&p->sem_sync);
-    }
-    trace_multifd_send_sync_main(multifd_send_state->packet_num);
-}
-
-static void *multifd_send_thread(void *opaque)
-{
-    MultiFDSendParams *p = opaque;
-    Error *local_err = NULL;
-    int ret = 0;
-    uint32_t flags = 0;
-
-    trace_multifd_send_thread_start(p->id);
-    rcu_register_thread();
-
-    if (multifd_send_initial_packet(p, &local_err) < 0) {
-        ret = -1;
-        goto out;
-    }
-    /* initial packet */
-    p->num_packets = 1;
-
-    while (true) {
-        qemu_sem_wait(&p->sem);
-
-        if (atomic_read(&multifd_send_state->exiting)) {
-            break;
-        }
-        qemu_mutex_lock(&p->mutex);
-
-        if (p->pending_job) {
-            uint32_t used = p->pages->used;
-            uint64_t packet_num = p->packet_num;
-            flags = p->flags;
-
-            p->next_packet_size = used * qemu_target_page_size();
-            multifd_send_fill_packet(p);
-            p->flags = 0;
-            p->num_packets++;
-            p->num_pages += used;
-            p->pages->used = 0;
-            p->pages->block = NULL;
-            qemu_mutex_unlock(&p->mutex);
-
-            trace_multifd_send(p->id, packet_num, used, flags,
-                               p->next_packet_size);
-
-            ret = qio_channel_write_all(p->c, (void *)p->packet,
-                                        p->packet_len, &local_err);
-            if (ret != 0) {
-                break;
-            }
-
-            if (used) {
-                ret = qio_channel_writev_all(p->c, p->pages->iov,
-                                             used, &local_err);
-                if (ret != 0) {
-                    break;
-                }
-            }
-
-            qemu_mutex_lock(&p->mutex);
-            p->pending_job--;
-            qemu_mutex_unlock(&p->mutex);
-
-            if (flags & MULTIFD_FLAG_SYNC) {
-                qemu_sem_post(&p->sem_sync);
-            }
-            qemu_sem_post(&multifd_send_state->channels_ready);
-        } else if (p->quit) {
-            qemu_mutex_unlock(&p->mutex);
-            break;
-        } else {
-            qemu_mutex_unlock(&p->mutex);
-            /* sometimes there are spurious wakeups */
-        }
-    }
-
-out:
-    if (local_err) {
-        trace_multifd_send_error(p->id);
-        multifd_send_terminate_threads(local_err);
-    }
-
-    /*
-     * Error happen, I will exit, but I can't just leave, tell
-     * who pay attention to me.
-     */
-    if (ret != 0) {
-        qemu_sem_post(&p->sem_sync);
-        qemu_sem_post(&multifd_send_state->channels_ready);
-    }
-
-    qemu_mutex_lock(&p->mutex);
-    p->running = false;
-    qemu_mutex_unlock(&p->mutex);
-
-    rcu_unregister_thread();
-    trace_multifd_send_thread_end(p->id, p->num_packets, p->num_pages);
-
-    return NULL;
-}
-
-static void multifd_new_send_channel_async(QIOTask *task, gpointer opaque)
-{
-    MultiFDSendParams *p = opaque;
-    QIOChannel *sioc = QIO_CHANNEL(qio_task_get_source(task));
-    Error *local_err = NULL;
-
-    trace_multifd_new_send_channel_async(p->id);
-    if (qio_task_propagate_error(task, &local_err)) {
-        migrate_set_error(migrate_get_current(), local_err);
-        multifd_save_cleanup();
-    } else {
-        p->c = QIO_CHANNEL(sioc);
-        qio_channel_set_delay(p->c, false);
-        p->running = true;
-        qemu_thread_create(&p->thread, p->name, multifd_send_thread, p,
-                           QEMU_THREAD_JOINABLE);
-    }
-}
-
-int multifd_save_setup(Error **errp)
-{
-    int thread_count;
-    uint32_t page_count = MULTIFD_PACKET_SIZE / qemu_target_page_size();
-    uint8_t i;
-
-    if (!migrate_use_multifd()) {
-        return 0;
-    }
-    thread_count = migrate_multifd_channels();
-    multifd_send_state = g_malloc0(sizeof(*multifd_send_state));
-    multifd_send_state->params = g_new0(MultiFDSendParams, thread_count);
-    multifd_send_state->pages = multifd_pages_init(page_count);
-    qemu_sem_init(&multifd_send_state->channels_ready, 0);
-    atomic_set(&multifd_send_state->exiting, 0);
-
-    for (i = 0; i < thread_count; i++) {
-        MultiFDSendParams *p = &multifd_send_state->params[i];
-
-        qemu_mutex_init(&p->mutex);
-        qemu_sem_init(&p->sem, 0);
-        qemu_sem_init(&p->sem_sync, 0);
-        p->quit = false;
-        p->pending_job = 0;
-        p->id = i;
-        p->pages = multifd_pages_init(page_count);
-        p->packet_len = sizeof(MultiFDPacket_t)
-                      + sizeof(uint64_t) * page_count;
-        p->packet = g_malloc0(p->packet_len);
-        p->packet->magic = cpu_to_be32(MULTIFD_MAGIC);
-        p->packet->version = cpu_to_be32(MULTIFD_VERSION);
-        p->name = g_strdup_printf("multifdsend_%d", i);
-        socket_send_channel_create(multifd_new_send_channel_async, p);
-    }
-    return 0;
-}
-
-struct {
-    MultiFDRecvParams *params;
-    /* number of created threads */
-    int count;
-    /* syncs main thread and channels */
-    QemuSemaphore sem_sync;
-    /* global number of generated multifd packets */
-    uint64_t packet_num;
-} *multifd_recv_state;
-
-static void multifd_recv_terminate_threads(Error *err)
-{
-    int i;
-
-    trace_multifd_recv_terminate_threads(err != NULL);
-
-    if (err) {
-        MigrationState *s = migrate_get_current();
-        migrate_set_error(s, err);
-        if (s->state == MIGRATION_STATUS_SETUP ||
-            s->state == MIGRATION_STATUS_ACTIVE) {
-            migrate_set_state(&s->state, s->state,
-                              MIGRATION_STATUS_FAILED);
-        }
-    }
-
-    for (i = 0; i < migrate_multifd_channels(); i++) {
-        MultiFDRecvParams *p = &multifd_recv_state->params[i];
-
-        qemu_mutex_lock(&p->mutex);
-        p->quit = true;
-        /*
-         * We could arrive here for two reasons:
-         *  - normal quit, i.e. everything went fine, just finished
-         *  - error quit: We close the channels so the channel threads
-         *    finish the qio_channel_read_all_eof()
-         */
-        if (p->c) {
-            qio_channel_shutdown(p->c, QIO_CHANNEL_SHUTDOWN_BOTH, NULL);
-        }
-        qemu_mutex_unlock(&p->mutex);
-    }
-}
-
-int multifd_load_cleanup(Error **errp)
-{
-    int i;
-    int ret = 0;
-
-    if (!migrate_use_multifd()) {
-        return 0;
-    }
-    multifd_recv_terminate_threads(NULL);
-    for (i = 0; i < migrate_multifd_channels(); i++) {
-        MultiFDRecvParams *p = &multifd_recv_state->params[i];
-
-        if (p->running) {
-            p->quit = true;
-            /*
-             * multifd_recv_thread may hung at MULTIFD_FLAG_SYNC handle code,
-             * however try to wakeup it without harm in cleanup phase.
-             */
-            qemu_sem_post(&p->sem_sync);
-            qemu_thread_join(&p->thread);
-        }
-    }
-    for (i = 0; i < migrate_multifd_channels(); i++) {
-        MultiFDRecvParams *p = &multifd_recv_state->params[i];
-
-        object_unref(OBJECT(p->c));
-        p->c = NULL;
-        qemu_mutex_destroy(&p->mutex);
-        qemu_sem_destroy(&p->sem_sync);
-        g_free(p->name);
-        p->name = NULL;
-        multifd_pages_clear(p->pages);
-        p->pages = NULL;
-        p->packet_len = 0;
-        g_free(p->packet);
-        p->packet = NULL;
-    }
-    qemu_sem_destroy(&multifd_recv_state->sem_sync);
-    g_free(multifd_recv_state->params);
-    multifd_recv_state->params = NULL;
-    g_free(multifd_recv_state);
-    multifd_recv_state = NULL;
-
-    return ret;
-}
-
-static void multifd_recv_sync_main(void)
-{
-    int i;
-
-    if (!migrate_use_multifd()) {
-        return;
-    }
-    for (i = 0; i < migrate_multifd_channels(); i++) {
-        MultiFDRecvParams *p = &multifd_recv_state->params[i];
-
-        trace_multifd_recv_sync_main_wait(p->id);
-        qemu_sem_wait(&multifd_recv_state->sem_sync);
-    }
-    for (i = 0; i < migrate_multifd_channels(); i++) {
-        MultiFDRecvParams *p = &multifd_recv_state->params[i];
-
-        qemu_mutex_lock(&p->mutex);
-        if (multifd_recv_state->packet_num < p->packet_num) {
-            multifd_recv_state->packet_num = p->packet_num;
-        }
-        qemu_mutex_unlock(&p->mutex);
-        trace_multifd_recv_sync_main_signal(p->id);
-        qemu_sem_post(&p->sem_sync);
-    }
-    trace_multifd_recv_sync_main(multifd_recv_state->packet_num);
-}
-
-static void *multifd_recv_thread(void *opaque)
-{
-    MultiFDRecvParams *p = opaque;
-    Error *local_err = NULL;
-    int ret;
-
-    trace_multifd_recv_thread_start(p->id);
-    rcu_register_thread();
-
-    while (true) {
-        uint32_t used;
-        uint32_t flags;
-
-        if (p->quit) {
-            break;
-        }
-
-        ret = qio_channel_read_all_eof(p->c, (void *)p->packet,
-                                       p->packet_len, &local_err);
-        if (ret == 0) {   /* EOF */
-            break;
-        }
-        if (ret == -1) {   /* Error */
-            break;
-        }
-
-        qemu_mutex_lock(&p->mutex);
-        ret = multifd_recv_unfill_packet(p, &local_err);
-        if (ret) {
-            qemu_mutex_unlock(&p->mutex);
-            break;
-        }
-
-        used = p->pages->used;
-        flags = p->flags;
-        trace_multifd_recv(p->id, p->packet_num, used, flags,
-                           p->next_packet_size);
-        p->num_packets++;
-        p->num_pages += used;
-        qemu_mutex_unlock(&p->mutex);
-
-        if (used) {
-            ret = qio_channel_readv_all(p->c, p->pages->iov,
-                                        used, &local_err);
-            if (ret != 0) {
-                break;
-            }
-        }
-
-        if (flags & MULTIFD_FLAG_SYNC) {
-            qemu_sem_post(&multifd_recv_state->sem_sync);
-            qemu_sem_wait(&p->sem_sync);
-        }
-    }
-
-    if (local_err) {
-        multifd_recv_terminate_threads(local_err);
-    }
-    qemu_mutex_lock(&p->mutex);
-    p->running = false;
-    qemu_mutex_unlock(&p->mutex);
-
-    rcu_unregister_thread();
-    trace_multifd_recv_thread_end(p->id, p->num_packets, p->num_pages);
-
-    return NULL;
-}
-
-int multifd_load_setup(Error **errp)
-{
-    int thread_count;
-    uint32_t page_count = MULTIFD_PACKET_SIZE / qemu_target_page_size();
-    uint8_t i;
-
-    if (!migrate_use_multifd()) {
-        return 0;
-    }
-    thread_count = migrate_multifd_channels();
-    multifd_recv_state = g_malloc0(sizeof(*multifd_recv_state));
-    multifd_recv_state->params = g_new0(MultiFDRecvParams, thread_count);
-    atomic_set(&multifd_recv_state->count, 0);
-    qemu_sem_init(&multifd_recv_state->sem_sync, 0);
-
-    for (i = 0; i < thread_count; i++) {
-        MultiFDRecvParams *p = &multifd_recv_state->params[i];
-
-        qemu_mutex_init(&p->mutex);
-        qemu_sem_init(&p->sem_sync, 0);
-        p->quit = false;
-        p->id = i;
-        p->pages = multifd_pages_init(page_count);
-        p->packet_len = sizeof(MultiFDPacket_t)
-                      + sizeof(uint64_t) * page_count;
-        p->packet = g_malloc0(p->packet_len);
-        p->name = g_strdup_printf("multifdrecv_%d", i);
-    }
-    return 0;
-}
-
-bool multifd_recv_all_channels_created(void)
-{
-    int thread_count = migrate_multifd_channels();
-
-    if (!migrate_use_multifd()) {
-        return true;
-    }
-
-    return thread_count == atomic_read(&multifd_recv_state->count);
-}
-
-/*
- * Try to receive all multifd channels to get ready for the migration.
- * - Return true and do not set @errp when correctly receving all channels;
- * - Return false and do not set @errp when correctly receiving the current one;
- * - Return false and set @errp when failing to receive the current channel.
- */
-bool multifd_recv_new_channel(QIOChannel *ioc, Error **errp)
-{
-    MultiFDRecvParams *p;
-    Error *local_err = NULL;
-    int id;
-
-    id = multifd_recv_initial_packet(ioc, &local_err);
-    if (id < 0) {
-        multifd_recv_terminate_threads(local_err);
-        error_propagate_prepend(errp, local_err,
-                                "failed to receive packet"
-                                " via multifd channel %d: ",
-                                atomic_read(&multifd_recv_state->count));
-        return false;
-    }
-    trace_multifd_recv_new_channel(id);
-
-    p = &multifd_recv_state->params[id];
-    if (p->c != NULL) {
-        error_setg(&local_err, "multifd: received id '%d' already setup'",
-                   id);
-        multifd_recv_terminate_threads(local_err);
-        error_propagate(errp, local_err);
-        return false;
-    }
-    p->c = ioc;
-    object_ref(OBJECT(ioc));
-    /* initial packet */
-    p->num_packets = 1;
-
-    p->running = true;
-    qemu_thread_create(&p->thread, p->name, multifd_recv_thread, p,
-                       QEMU_THREAD_JOINABLE);
-    atomic_inc(&multifd_recv_state->count);
-    return atomic_read(&multifd_recv_state->count) ==
-           migrate_multifd_channels();
-}
-
 /**
  * save_page_header: write page header to wire
  *
diff --git a/migration/ram.h b/migration/ram.h
index 42be471d52..a553d40751 100644
--- a/migration/ram.h
+++ b/migration/ram.h
@@ -41,13 +41,6 @@ int xbzrle_cache_resize(int64_t new_size, Error **errp);
 uint64_t ram_bytes_remaining(void);
 uint64_t ram_bytes_total(void);
 
-int multifd_save_setup(Error **errp);
-void multifd_save_cleanup(void);
-int multifd_load_setup(Error **errp);
-int multifd_load_cleanup(Error **errp);
-bool multifd_recv_all_channels_created(void);
-bool multifd_recv_new_channel(QIOChannel *ioc, Error **errp);
-
 uint64_t ram_pagesize_summary(void);
 int ram_save_queue_pages(const char *rbname, ram_addr_t start, ram_addr_t len);
 void acct_update_position(QEMUFile *f, size_t size, bool zero);
-- 
2.24.1

Re: [PATCH v3 17/21] multifd: Split multifd code into its own file

[Dr. David Alan Gilbert]

* Juan Quintela (quintela@redhat.com) wrote:
> Signed-off-by: Juan Quintela <quintela@redhat.com>

Good, ram.c was getting WAY too big.


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

> ---
>  migration/Makefile.objs |   1 +
>  migration/migration.c   |   1 +
>  migration/multifd.c     | 891 ++++++++++++++++++++++++++++++++++++
>  migration/multifd.h     | 139 ++++++
>  migration/ram.c         | 980 +---------------------------------------
>  migration/ram.h         |   7 -
>  6 files changed, 1033 insertions(+), 986 deletions(-)
>  create mode 100644 migration/multifd.c
>  create mode 100644 migration/multifd.h
> 
> diff --git a/migration/Makefile.objs b/migration/Makefile.objs
> index a4f3bafd86..d3623d5f9b 100644
> --- a/migration/Makefile.objs
> +++ b/migration/Makefile.objs
> @@ -7,6 +7,7 @@ common-obj-y += qemu-file-channel.o
>  common-obj-y += xbzrle.o postcopy-ram.o
>  common-obj-y += qjson.o
>  common-obj-y += block-dirty-bitmap.o
> +common-obj-y += multifd.o
>  
>  common-obj-$(CONFIG_RDMA) += rdma.o
>  
> diff --git a/migration/migration.c b/migration/migration.c
> index ecb56afd50..3501bc3353 100644
> --- a/migration/migration.c
> +++ b/migration/migration.c
> @@ -53,6 +53,7 @@
>  #include "monitor/monitor.h"
>  #include "net/announce.h"
>  #include "qemu/queue.h"
> +#include "multifd.h"
>  
>  #define MAX_THROTTLE  (32 << 20)      /* Migration transfer speed throttling */
>  
> diff --git a/migration/multifd.c b/migration/multifd.c
> new file mode 100644
> index 0000000000..1875bb3aaa
> --- /dev/null
> +++ b/migration/multifd.c
> @@ -0,0 +1,891 @@
> +/*
> + * Multifd common code
> + *
> + * Copyright (c) 2019-2020 Red Hat Inc
> + *
> + * Authors:
> + *  Juan Quintela <quintela@redhat.com>
> + *
> + * This work is licensed under the terms of the GNU GPL, version 2 or later.
> + * See the COPYING file in the top-level directory.
> + */
> +
> +#include "qemu/osdep.h"
> +#include "qemu/rcu.h"
> +#include "exec/target_page.h"
> +#include "sysemu/sysemu.h"
> +#include "exec/ramblock.h"
> +#include "qemu/error-report.h"
> +#include "qapi/error.h"
> +#include "ram.h"
> +#include "migration.h"
> +#include "socket.h"
> +#include "qemu-file.h"
> +#include "trace.h"
> +#include "multifd.h"
> +
> +/* Multiple fd's */
> +
> +#define MULTIFD_MAGIC 0x11223344U
> +#define MULTIFD_VERSION 1
> +
> +typedef struct {
> +    uint32_t magic;
> +    uint32_t version;
> +    unsigned char uuid[16]; /* QemuUUID */
> +    uint8_t id;
> +    uint8_t unused1[7];     /* Reserved for future use */
> +    uint64_t unused2[4];    /* Reserved for future use */
> +} __attribute__((packed)) MultiFDInit_t;
> +
> +static int multifd_send_initial_packet(MultiFDSendParams *p, Error **errp)
> +{
> +    MultiFDInit_t msg = {};
> +    int ret;
> +
> +    msg.magic = cpu_to_be32(MULTIFD_MAGIC);
> +    msg.version = cpu_to_be32(MULTIFD_VERSION);
> +    msg.id = p->id;
> +    memcpy(msg.uuid, &qemu_uuid.data, sizeof(msg.uuid));
> +
> +    ret = qio_channel_write_all(p->c, (char *)&msg, sizeof(msg), errp);
> +    if (ret != 0) {
> +        return -1;
> +    }
> +    return 0;
> +}
> +
> +static int multifd_recv_initial_packet(QIOChannel *c, Error **errp)
> +{
> +    MultiFDInit_t msg;
> +    int ret;
> +
> +    ret = qio_channel_read_all(c, (char *)&msg, sizeof(msg), errp);
> +    if (ret != 0) {
> +        return -1;
> +    }
> +
> +    msg.magic = be32_to_cpu(msg.magic);
> +    msg.version = be32_to_cpu(msg.version);
> +
> +    if (msg.magic != MULTIFD_MAGIC) {
> +        error_setg(errp, "multifd: received packet magic %x "
> +                   "expected %x", msg.magic, MULTIFD_MAGIC);
> +        return -1;
> +    }
> +
> +    if (msg.version != MULTIFD_VERSION) {
> +        error_setg(errp, "multifd: received packet version %d "
> +                   "expected %d", msg.version, MULTIFD_VERSION);
> +        return -1;
> +    }
> +
> +    if (memcmp(msg.uuid, &qemu_uuid, sizeof(qemu_uuid))) {
> +        char *uuid = qemu_uuid_unparse_strdup(&qemu_uuid);
> +        char *msg_uuid = qemu_uuid_unparse_strdup((const QemuUUID *)msg.uuid);
> +
> +        error_setg(errp, "multifd: received uuid '%s' and expected "
> +                   "uuid '%s' for channel %hhd", msg_uuid, uuid, msg.id);
> +        g_free(uuid);
> +        g_free(msg_uuid);
> +        return -1;
> +    }
> +
> +    if (msg.id > migrate_multifd_channels()) {
> +        error_setg(errp, "multifd: received channel version %d "
> +                   "expected %d", msg.version, MULTIFD_VERSION);
> +        return -1;
> +    }
> +
> +    return msg.id;
> +}
> +
> +static MultiFDPages_t *multifd_pages_init(size_t size)
> +{
> +    MultiFDPages_t *pages = g_new0(MultiFDPages_t, 1);
> +
> +    pages->allocated = size;
> +    pages->iov = g_new0(struct iovec, size);
> +    pages->offset = g_new0(ram_addr_t, size);
> +
> +    return pages;
> +}
> +
> +static void multifd_pages_clear(MultiFDPages_t *pages)
> +{
> +    pages->used = 0;
> +    pages->allocated = 0;
> +    pages->packet_num = 0;
> +    pages->block = NULL;
> +    g_free(pages->iov);
> +    pages->iov = NULL;
> +    g_free(pages->offset);
> +    pages->offset = NULL;
> +    g_free(pages);
> +}
> +
> +static void multifd_send_fill_packet(MultiFDSendParams *p)
> +{
> +    MultiFDPacket_t *packet = p->packet;
> +    int i;
> +
> +    packet->flags = cpu_to_be32(p->flags);
> +    packet->pages_alloc = cpu_to_be32(p->pages->allocated);
> +    packet->pages_used = cpu_to_be32(p->pages->used);
> +    packet->next_packet_size = cpu_to_be32(p->next_packet_size);
> +    packet->packet_num = cpu_to_be64(p->packet_num);
> +
> +    if (p->pages->block) {
> +        strncpy(packet->ramblock, p->pages->block->idstr, 256);
> +    }
> +
> +    for (i = 0; i < p->pages->used; i++) {
> +        /* there are architectures where ram_addr_t is 32 bit */
> +        uint64_t temp = p->pages->offset[i];
> +
> +        packet->offset[i] = cpu_to_be64(temp);
> +    }
> +}
> +
> +static int multifd_recv_unfill_packet(MultiFDRecvParams *p, Error **errp)
> +{
> +    MultiFDPacket_t *packet = p->packet;
> +    uint32_t pages_max = MULTIFD_PACKET_SIZE / qemu_target_page_size();
> +    RAMBlock *block;
> +    int i;
> +
> +    packet->magic = be32_to_cpu(packet->magic);
> +    if (packet->magic != MULTIFD_MAGIC) {
> +        error_setg(errp, "multifd: received packet "
> +                   "magic %x and expected magic %x",
> +                   packet->magic, MULTIFD_MAGIC);
> +        return -1;
> +    }
> +
> +    packet->version = be32_to_cpu(packet->version);
> +    if (packet->version != MULTIFD_VERSION) {
> +        error_setg(errp, "multifd: received packet "
> +                   "version %d and expected version %d",
> +                   packet->version, MULTIFD_VERSION);
> +        return -1;
> +    }
> +
> +    p->flags = be32_to_cpu(packet->flags);
> +
> +    packet->pages_alloc = be32_to_cpu(packet->pages_alloc);
> +    /*
> +     * If we received a packet that is 100 times bigger than expected
> +     * just stop migration.  It is a magic number.
> +     */
> +    if (packet->pages_alloc > pages_max * 100) {
> +        error_setg(errp, "multifd: received packet "
> +                   "with size %d and expected a maximum size of %d",
> +                   packet->pages_alloc, pages_max * 100) ;
> +        return -1;
> +    }
> +    /*
> +     * We received a packet that is bigger than expected but inside
> +     * reasonable limits (see previous comment).  Just reallocate.
> +     */
> +    if (packet->pages_alloc > p->pages->allocated) {
> +        multifd_pages_clear(p->pages);
> +        p->pages = multifd_pages_init(packet->pages_alloc);
> +    }
> +
> +    p->pages->used = be32_to_cpu(packet->pages_used);
> +    if (p->pages->used > packet->pages_alloc) {
> +        error_setg(errp, "multifd: received packet "
> +                   "with %d pages and expected maximum pages are %d",
> +                   p->pages->used, packet->pages_alloc) ;
> +        return -1;
> +    }
> +
> +    p->next_packet_size = be32_to_cpu(packet->next_packet_size);
> +    p->packet_num = be64_to_cpu(packet->packet_num);
> +
> +    if (p->pages->used == 0) {
> +        return 0;
> +    }
> +
> +    /* make sure that ramblock is 0 terminated */
> +    packet->ramblock[255] = 0;
> +    block = qemu_ram_block_by_name(packet->ramblock);
> +    if (!block) {
> +        error_setg(errp, "multifd: unknown ram block %s",
> +                   packet->ramblock);
> +        return -1;
> +    }
> +
> +    for (i = 0; i < p->pages->used; i++) {
> +        uint64_t offset = be64_to_cpu(packet->offset[i]);
> +
> +        if (offset > (block->used_length - qemu_target_page_size())) {
> +            error_setg(errp, "multifd: offset too long %" PRIu64
> +                       " (max " RAM_ADDR_FMT ")",
> +                       offset, block->max_length);
> +            return -1;
> +        }
> +        p->pages->iov[i].iov_base = block->host + offset;
> +        p->pages->iov[i].iov_len = qemu_target_page_size();
> +    }
> +
> +    return 0;
> +}
> +
> +struct {
> +    MultiFDSendParams *params;
> +    /* array of pages to sent */
> +    MultiFDPages_t *pages;
> +    /* global number of generated multifd packets */
> +    uint64_t packet_num;
> +    /* send channels ready */
> +    QemuSemaphore channels_ready;
> +    /*
> +     * Have we already run terminate threads.  There is a race when it
> +     * happens that we got one error while we are exiting.
> +     * We will use atomic operations.  Only valid values are 0 and 1.
> +     */
> +    int exiting;
> +} *multifd_send_state;
> +
> +/*
> + * How we use multifd_send_state->pages and channel->pages?
> + *
> + * We create a pages for each channel, and a main one.  Each time that
> + * we need to send a batch of pages we interchange the ones between
> + * multifd_send_state and the channel that is sending it.  There are
> + * two reasons for that:
> + *    - to not have to do so many mallocs during migration
> + *    - to make easier to know what to free at the end of migration
> + *
> + * This way we always know who is the owner of each "pages" struct,
> + * and we don't need any locking.  It belongs to the migration thread
> + * or to the channel thread.  Switching is safe because the migration
> + * thread is using the channel mutex when changing it, and the channel
> + * have to had finish with its own, otherwise pending_job can't be
> + * false.
> + */
> +
> +static int multifd_send_pages(QEMUFile *f)
> +{
> +    int i;
> +    static int next_channel;
> +    MultiFDSendParams *p = NULL; /* make happy gcc */
> +    MultiFDPages_t *pages = multifd_send_state->pages;
> +    uint64_t transferred;
> +
> +    if (atomic_read(&multifd_send_state->exiting)) {
> +        return -1;
> +    }
> +
> +    qemu_sem_wait(&multifd_send_state->channels_ready);
> +    for (i = next_channel;; i = (i + 1) % migrate_multifd_channels()) {
> +        p = &multifd_send_state->params[i];
> +
> +        qemu_mutex_lock(&p->mutex);
> +        if (p->quit) {
> +            error_report("%s: channel %d has already quit!", __func__, i);
> +            qemu_mutex_unlock(&p->mutex);
> +            return -1;
> +        }
> +        if (!p->pending_job) {
> +            p->pending_job++;
> +            next_channel = (i + 1) % migrate_multifd_channels();
> +            break;
> +        }
> +        qemu_mutex_unlock(&p->mutex);
> +    }
> +    assert(!p->pages->used);
> +    assert(!p->pages->block);
> +
> +    p->packet_num = multifd_send_state->packet_num++;
> +    multifd_send_state->pages = p->pages;
> +    p->pages = pages;
> +    transferred = ((uint64_t) pages->used) * qemu_target_page_size()
> +                + p->packet_len;
> +    qemu_file_update_transfer(f, transferred);
> +    ram_counters.multifd_bytes += transferred;
> +    ram_counters.transferred += transferred;;
> +    qemu_mutex_unlock(&p->mutex);
> +    qemu_sem_post(&p->sem);
> +
> +    return 1;
> +}
> +
> +int multifd_queue_page(QEMUFile *f, RAMBlock *block, ram_addr_t offset)
> +{
> +    MultiFDPages_t *pages = multifd_send_state->pages;
> +
> +    if (!pages->block) {
> +        pages->block = block;
> +    }
> +
> +    if (pages->block == block) {
> +        pages->offset[pages->used] = offset;
> +        pages->iov[pages->used].iov_base = block->host + offset;
> +        pages->iov[pages->used].iov_len = qemu_target_page_size();
> +        pages->used++;
> +
> +        if (pages->used < pages->allocated) {
> +            return 1;
> +        }
> +    }
> +
> +    if (multifd_send_pages(f) < 0) {
> +        return -1;
> +    }
> +
> +    if (pages->block != block) {
> +        return  multifd_queue_page(f, block, offset);
> +    }
> +
> +    return 1;
> +}
> +
> +static void multifd_send_terminate_threads(Error *err)
> +{
> +    int i;
> +
> +    trace_multifd_send_terminate_threads(err != NULL);
> +
> +    if (err) {
> +        MigrationState *s = migrate_get_current();
> +        migrate_set_error(s, err);
> +        if (s->state == MIGRATION_STATUS_SETUP ||
> +            s->state == MIGRATION_STATUS_PRE_SWITCHOVER ||
> +            s->state == MIGRATION_STATUS_DEVICE ||
> +            s->state == MIGRATION_STATUS_ACTIVE) {
> +            migrate_set_state(&s->state, s->state,
> +                              MIGRATION_STATUS_FAILED);
> +        }
> +    }
> +
> +    /*
> +     * We don't want to exit each threads twice.  Depending on where
> +     * we get the error, or if there are two independent errors in two
> +     * threads at the same time, we can end calling this function
> +     * twice.
> +     */
> +    if (atomic_xchg(&multifd_send_state->exiting, 1)) {
> +        return;
> +    }
> +
> +    for (i = 0; i < migrate_multifd_channels(); i++) {
> +        MultiFDSendParams *p = &multifd_send_state->params[i];
> +
> +        qemu_mutex_lock(&p->mutex);
> +        p->quit = true;
> +        qemu_sem_post(&p->sem);
> +        qemu_mutex_unlock(&p->mutex);
> +    }
> +}
> +
> +void multifd_save_cleanup(void)
> +{
> +    int i;
> +
> +    if (!migrate_use_multifd()) {
> +        return;
> +    }
> +    multifd_send_terminate_threads(NULL);
> +    for (i = 0; i < migrate_multifd_channels(); i++) {
> +        MultiFDSendParams *p = &multifd_send_state->params[i];
> +
> +        if (p->running) {
> +            qemu_thread_join(&p->thread);
> +        }
> +    }
> +    for (i = 0; i < migrate_multifd_channels(); i++) {
> +        MultiFDSendParams *p = &multifd_send_state->params[i];
> +
> +        socket_send_channel_destroy(p->c);
> +        p->c = NULL;
> +        qemu_mutex_destroy(&p->mutex);
> +        qemu_sem_destroy(&p->sem);
> +        qemu_sem_destroy(&p->sem_sync);
> +        g_free(p->name);
> +        p->name = NULL;
> +        multifd_pages_clear(p->pages);
> +        p->pages = NULL;
> +        p->packet_len = 0;
> +        g_free(p->packet);
> +        p->packet = NULL;
> +    }
> +    qemu_sem_destroy(&multifd_send_state->channels_ready);
> +    g_free(multifd_send_state->params);
> +    multifd_send_state->params = NULL;
> +    multifd_pages_clear(multifd_send_state->pages);
> +    multifd_send_state->pages = NULL;
> +    g_free(multifd_send_state);
> +    multifd_send_state = NULL;
> +}
> +
> +void multifd_send_sync_main(QEMUFile *f)
> +{
> +    int i;
> +
> +    if (!migrate_use_multifd()) {
> +        return;
> +    }
> +    if (multifd_send_state->pages->used) {
> +        if (multifd_send_pages(f) < 0) {
> +            error_report("%s: multifd_send_pages fail", __func__);
> +            return;
> +        }
> +    }
> +    for (i = 0; i < migrate_multifd_channels(); i++) {
> +        MultiFDSendParams *p = &multifd_send_state->params[i];
> +
> +        trace_multifd_send_sync_main_signal(p->id);
> +
> +        qemu_mutex_lock(&p->mutex);
> +
> +        if (p->quit) {
> +            error_report("%s: channel %d has already quit", __func__, i);
> +            qemu_mutex_unlock(&p->mutex);
> +            return;
> +        }
> +
> +        p->packet_num = multifd_send_state->packet_num++;
> +        p->flags |= MULTIFD_FLAG_SYNC;
> +        p->pending_job++;
> +        qemu_file_update_transfer(f, p->packet_len);
> +        ram_counters.multifd_bytes += p->packet_len;
> +        ram_counters.transferred += p->packet_len;
> +        qemu_mutex_unlock(&p->mutex);
> +        qemu_sem_post(&p->sem);
> +    }
> +    for (i = 0; i < migrate_multifd_channels(); i++) {
> +        MultiFDSendParams *p = &multifd_send_state->params[i];
> +
> +        trace_multifd_send_sync_main_wait(p->id);
> +        qemu_sem_wait(&p->sem_sync);
> +    }
> +    trace_multifd_send_sync_main(multifd_send_state->packet_num);
> +}
> +
> +static void *multifd_send_thread(void *opaque)
> +{
> +    MultiFDSendParams *p = opaque;
> +    Error *local_err = NULL;
> +    int ret = 0;
> +    uint32_t flags = 0;
> +
> +    trace_multifd_send_thread_start(p->id);
> +    rcu_register_thread();
> +
> +    if (multifd_send_initial_packet(p, &local_err) < 0) {
> +        ret = -1;
> +        goto out;
> +    }
> +    /* initial packet */
> +    p->num_packets = 1;
> +
> +    while (true) {
> +        qemu_sem_wait(&p->sem);
> +
> +        if (atomic_read(&multifd_send_state->exiting)) {
> +            break;
> +        }
> +        qemu_mutex_lock(&p->mutex);
> +
> +        if (p->pending_job) {
> +            uint32_t used = p->pages->used;
> +            uint64_t packet_num = p->packet_num;
> +            flags = p->flags;
> +
> +            p->next_packet_size = used * qemu_target_page_size();
> +            multifd_send_fill_packet(p);
> +            p->flags = 0;
> +            p->num_packets++;
> +            p->num_pages += used;
> +            p->pages->used = 0;
> +            p->pages->block = NULL;
> +            qemu_mutex_unlock(&p->mutex);
> +
> +            trace_multifd_send(p->id, packet_num, used, flags,
> +                               p->next_packet_size);
> +
> +            ret = qio_channel_write_all(p->c, (void *)p->packet,
> +                                        p->packet_len, &local_err);
> +            if (ret != 0) {
> +                break;
> +            }
> +
> +            if (used) {
> +                ret = qio_channel_writev_all(p->c, p->pages->iov,
> +                                             used, &local_err);
> +                if (ret != 0) {
> +                    break;
> +                }
> +            }
> +
> +            qemu_mutex_lock(&p->mutex);
> +            p->pending_job--;
> +            qemu_mutex_unlock(&p->mutex);
> +
> +            if (flags & MULTIFD_FLAG_SYNC) {
> +                qemu_sem_post(&p->sem_sync);
> +            }
> +            qemu_sem_post(&multifd_send_state->channels_ready);
> +        } else if (p->quit) {
> +            qemu_mutex_unlock(&p->mutex);
> +            break;
> +        } else {
> +            qemu_mutex_unlock(&p->mutex);
> +            /* sometimes there are spurious wakeups */
> +        }
> +    }
> +
> +out:
> +    if (local_err) {
> +        trace_multifd_send_error(p->id);
> +        multifd_send_terminate_threads(local_err);
> +    }
> +
> +    /*
> +     * Error happen, I will exit, but I can't just leave, tell
> +     * who pay attention to me.
> +     */
> +    if (ret != 0) {
> +        qemu_sem_post(&p->sem_sync);
> +        qemu_sem_post(&multifd_send_state->channels_ready);
> +    }
> +
> +    qemu_mutex_lock(&p->mutex);
> +    p->running = false;
> +    qemu_mutex_unlock(&p->mutex);
> +
> +    rcu_unregister_thread();
> +    trace_multifd_send_thread_end(p->id, p->num_packets, p->num_pages);
> +
> +    return NULL;
> +}
> +
> +static void multifd_new_send_channel_async(QIOTask *task, gpointer opaque)
> +{
> +    MultiFDSendParams *p = opaque;
> +    QIOChannel *sioc = QIO_CHANNEL(qio_task_get_source(task));
> +    Error *local_err = NULL;
> +
> +    trace_multifd_new_send_channel_async(p->id);
> +    if (qio_task_propagate_error(task, &local_err)) {
> +        migrate_set_error(migrate_get_current(), local_err);
> +        multifd_save_cleanup();
> +    } else {
> +        p->c = QIO_CHANNEL(sioc);
> +        qio_channel_set_delay(p->c, false);
> +        p->running = true;
> +        qemu_thread_create(&p->thread, p->name, multifd_send_thread, p,
> +                           QEMU_THREAD_JOINABLE);
> +    }
> +}
> +
> +int multifd_save_setup(Error **errp)
> +{
> +    int thread_count;
> +    uint32_t page_count = MULTIFD_PACKET_SIZE / qemu_target_page_size();
> +    uint8_t i;
> +
> +    if (!migrate_use_multifd()) {
> +        return 0;
> +    }
> +    thread_count = migrate_multifd_channels();
> +    multifd_send_state = g_malloc0(sizeof(*multifd_send_state));
> +    multifd_send_state->params = g_new0(MultiFDSendParams, thread_count);
> +    multifd_send_state->pages = multifd_pages_init(page_count);
> +    qemu_sem_init(&multifd_send_state->channels_ready, 0);
> +    atomic_set(&multifd_send_state->exiting, 0);
> +
> +    for (i = 0; i < thread_count; i++) {
> +        MultiFDSendParams *p = &multifd_send_state->params[i];
> +
> +        qemu_mutex_init(&p->mutex);
> +        qemu_sem_init(&p->sem, 0);
> +        qemu_sem_init(&p->sem_sync, 0);
> +        p->quit = false;
> +        p->pending_job = 0;
> +        p->id = i;
> +        p->pages = multifd_pages_init(page_count);
> +        p->packet_len = sizeof(MultiFDPacket_t)
> +                      + sizeof(uint64_t) * page_count;
> +        p->packet = g_malloc0(p->packet_len);
> +        p->packet->magic = cpu_to_be32(MULTIFD_MAGIC);
> +        p->packet->version = cpu_to_be32(MULTIFD_VERSION);
> +        p->name = g_strdup_printf("multifdsend_%d", i);
> +        socket_send_channel_create(multifd_new_send_channel_async, p);
> +    }
> +    return 0;
> +}
> +
> +struct {
> +    MultiFDRecvParams *params;
> +    /* number of created threads */
> +    int count;
> +    /* syncs main thread and channels */
> +    QemuSemaphore sem_sync;
> +    /* global number of generated multifd packets */
> +    uint64_t packet_num;
> +} *multifd_recv_state;
> +
> +static void multifd_recv_terminate_threads(Error *err)
> +{
> +    int i;
> +
> +    trace_multifd_recv_terminate_threads(err != NULL);
> +
> +    if (err) {
> +        MigrationState *s = migrate_get_current();
> +        migrate_set_error(s, err);
> +        if (s->state == MIGRATION_STATUS_SETUP ||
> +            s->state == MIGRATION_STATUS_ACTIVE) {
> +            migrate_set_state(&s->state, s->state,
> +                              MIGRATION_STATUS_FAILED);
> +        }
> +    }
> +
> +    for (i = 0; i < migrate_multifd_channels(); i++) {
> +        MultiFDRecvParams *p = &multifd_recv_state->params[i];
> +
> +        qemu_mutex_lock(&p->mutex);
> +        p->quit = true;
> +        /*
> +         * We could arrive here for two reasons:
> +         *  - normal quit, i.e. everything went fine, just finished
> +         *  - error quit: We close the channels so the channel threads
> +         *    finish the qio_channel_read_all_eof()
> +         */
> +        if (p->c) {
> +            qio_channel_shutdown(p->c, QIO_CHANNEL_SHUTDOWN_BOTH, NULL);
> +        }
> +        qemu_mutex_unlock(&p->mutex);
> +    }
> +}
> +
> +int multifd_load_cleanup(Error **errp)
> +{
> +    int i;
> +    int ret = 0;
> +
> +    if (!migrate_use_multifd()) {
> +        return 0;
> +    }
> +    multifd_recv_terminate_threads(NULL);
> +    for (i = 0; i < migrate_multifd_channels(); i++) {
> +        MultiFDRecvParams *p = &multifd_recv_state->params[i];
> +
> +        if (p->running) {
> +            p->quit = true;
> +            /*
> +             * multifd_recv_thread may hung at MULTIFD_FLAG_SYNC handle code,
> +             * however try to wakeup it without harm in cleanup phase.
> +             */
> +            qemu_sem_post(&p->sem_sync);
> +            qemu_thread_join(&p->thread);
> +        }
> +    }
> +    for (i = 0; i < migrate_multifd_channels(); i++) {
> +        MultiFDRecvParams *p = &multifd_recv_state->params[i];
> +
> +        object_unref(OBJECT(p->c));
> +        p->c = NULL;
> +        qemu_mutex_destroy(&p->mutex);
> +        qemu_sem_destroy(&p->sem_sync);
> +        g_free(p->name);
> +        p->name = NULL;
> +        multifd_pages_clear(p->pages);
> +        p->pages = NULL;
> +        p->packet_len = 0;
> +        g_free(p->packet);
> +        p->packet = NULL;
> +    }
> +    qemu_sem_destroy(&multifd_recv_state->sem_sync);
> +    g_free(multifd_recv_state->params);
> +    multifd_recv_state->params = NULL;
> +    g_free(multifd_recv_state);
> +    multifd_recv_state = NULL;
> +
> +    return ret;
> +}
> +
> +void multifd_recv_sync_main(void)
> +{
> +    int i;
> +
> +    if (!migrate_use_multifd()) {
> +        return;
> +    }
> +    for (i = 0; i < migrate_multifd_channels(); i++) {
> +        MultiFDRecvParams *p = &multifd_recv_state->params[i];
> +
> +        trace_multifd_recv_sync_main_wait(p->id);
> +        qemu_sem_wait(&multifd_recv_state->sem_sync);
> +    }
> +    for (i = 0; i < migrate_multifd_channels(); i++) {
> +        MultiFDRecvParams *p = &multifd_recv_state->params[i];
> +
> +        qemu_mutex_lock(&p->mutex);
> +        if (multifd_recv_state->packet_num < p->packet_num) {
> +            multifd_recv_state->packet_num = p->packet_num;
> +        }
> +        qemu_mutex_unlock(&p->mutex);
> +        trace_multifd_recv_sync_main_signal(p->id);
> +        qemu_sem_post(&p->sem_sync);
> +    }
> +    trace_multifd_recv_sync_main(multifd_recv_state->packet_num);
> +}
> +
> +static void *multifd_recv_thread(void *opaque)
> +{
> +    MultiFDRecvParams *p = opaque;
> +    Error *local_err = NULL;
> +    int ret;
> +
> +    trace_multifd_recv_thread_start(p->id);
> +    rcu_register_thread();
> +
> +    while (true) {
> +        uint32_t used;
> +        uint32_t flags;
> +
> +        if (p->quit) {
> +            break;
> +        }
> +
> +        ret = qio_channel_read_all_eof(p->c, (void *)p->packet,
> +                                       p->packet_len, &local_err);
> +        if (ret == 0) {   /* EOF */
> +            break;
> +        }
> +        if (ret == -1) {   /* Error */
> +            break;
> +        }
> +
> +        qemu_mutex_lock(&p->mutex);
> +        ret = multifd_recv_unfill_packet(p, &local_err);
> +        if (ret) {
> +            qemu_mutex_unlock(&p->mutex);
> +            break;
> +        }
> +
> +        used = p->pages->used;
> +        flags = p->flags;
> +        trace_multifd_recv(p->id, p->packet_num, used, flags,
> +                           p->next_packet_size);
> +        p->num_packets++;
> +        p->num_pages += used;
> +        qemu_mutex_unlock(&p->mutex);
> +
> +        if (used) {
> +            ret = qio_channel_readv_all(p->c, p->pages->iov,
> +                                        used, &local_err);
> +            if (ret != 0) {
> +                break;
> +            }
> +        }
> +
> +        if (flags & MULTIFD_FLAG_SYNC) {
> +            qemu_sem_post(&multifd_recv_state->sem_sync);
> +            qemu_sem_wait(&p->sem_sync);
> +        }
> +    }
> +
> +    if (local_err) {
> +        multifd_recv_terminate_threads(local_err);
> +    }
> +    qemu_mutex_lock(&p->mutex);
> +    p->running = false;
> +    qemu_mutex_unlock(&p->mutex);
> +
> +    rcu_unregister_thread();
> +    trace_multifd_recv_thread_end(p->id, p->num_packets, p->num_pages);
> +
> +    return NULL;
> +}
> +
> +int multifd_load_setup(Error **errp)
> +{
> +    int thread_count;
> +    uint32_t page_count = MULTIFD_PACKET_SIZE / qemu_target_page_size();
> +    uint8_t i;
> +
> +    if (!migrate_use_multifd()) {
> +        return 0;
> +    }
> +    thread_count = migrate_multifd_channels();
> +    multifd_recv_state = g_malloc0(sizeof(*multifd_recv_state));
> +    multifd_recv_state->params = g_new0(MultiFDRecvParams, thread_count);
> +    atomic_set(&multifd_recv_state->count, 0);
> +    qemu_sem_init(&multifd_recv_state->sem_sync, 0);
> +
> +    for (i = 0; i < thread_count; i++) {
> +        MultiFDRecvParams *p = &multifd_recv_state->params[i];
> +
> +        qemu_mutex_init(&p->mutex);
> +        qemu_sem_init(&p->sem_sync, 0);
> +        p->quit = false;
> +        p->id = i;
> +        p->pages = multifd_pages_init(page_count);
> +        p->packet_len = sizeof(MultiFDPacket_t)
> +                      + sizeof(uint64_t) * page_count;
> +        p->packet = g_malloc0(p->packet_len);
> +        p->name = g_strdup_printf("multifdrecv_%d", i);
> +    }
> +    return 0;
> +}
> +
> +bool multifd_recv_all_channels_created(void)
> +{
> +    int thread_count = migrate_multifd_channels();
> +
> +    if (!migrate_use_multifd()) {
> +        return true;
> +    }
> +
> +    return thread_count == atomic_read(&multifd_recv_state->count);
> +}
> +
> +/*
> + * Try to receive all multifd channels to get ready for the migration.
> + * - Return true and do not set @errp when correctly receving all channels;
> + * - Return false and do not set @errp when correctly receiving the current one;
> + * - Return false and set @errp when failing to receive the current channel.
> + */
> +bool multifd_recv_new_channel(QIOChannel *ioc, Error **errp)
> +{
> +    MultiFDRecvParams *p;
> +    Error *local_err = NULL;
> +    int id;
> +
> +    id = multifd_recv_initial_packet(ioc, &local_err);
> +    if (id < 0) {
> +        multifd_recv_terminate_threads(local_err);
> +        error_propagate_prepend(errp, local_err,
> +                                "failed to receive packet"
> +                                " via multifd channel %d: ",
> +                                atomic_read(&multifd_recv_state->count));
> +        return false;
> +    }
> +    trace_multifd_recv_new_channel(id);
> +
> +    p = &multifd_recv_state->params[id];
> +    if (p->c != NULL) {
> +        error_setg(&local_err, "multifd: received id '%d' already setup'",
> +                   id);
> +        multifd_recv_terminate_threads(local_err);
> +        error_propagate(errp, local_err);
> +        return false;
> +    }
> +    p->c = ioc;
> +    object_ref(OBJECT(ioc));
> +    /* initial packet */
> +    p->num_packets = 1;
> +
> +    p->running = true;
> +    qemu_thread_create(&p->thread, p->name, multifd_recv_thread, p,
> +                       QEMU_THREAD_JOINABLE);
> +    atomic_inc(&multifd_recv_state->count);
> +    return atomic_read(&multifd_recv_state->count) ==
> +           migrate_multifd_channels();
> +}
> +
> diff --git a/migration/multifd.h b/migration/multifd.h
> new file mode 100644
> index 0000000000..d8b0205977
> --- /dev/null
> +++ b/migration/multifd.h
> @@ -0,0 +1,139 @@
> +/*
> + * Multifd common functions
> + *
> + * Copyright (c) 2019-2020 Red Hat Inc
> + *
> + * Authors:
> + *  Juan Quintela <quintela@redhat.com>
> + *
> + * This work is licensed under the terms of the GNU GPL, version 2 or later.
> + * See the COPYING file in the top-level directory.
> + */
> +
> +#ifndef QEMU_MIGRATION_MULTIFD_H
> +#define QEMU_MIGRATION_MULTIFD_H
> +
> +int multifd_save_setup(Error **errp);
> +void multifd_save_cleanup(void);
> +int multifd_load_setup(Error **errp);
> +int multifd_load_cleanup(Error **errp);
> +bool multifd_recv_all_channels_created(void);
> +bool multifd_recv_new_channel(QIOChannel *ioc, Error **errp);
> +void multifd_recv_sync_main(void);
> +void multifd_send_sync_main(QEMUFile *f);
> +int multifd_queue_page(QEMUFile *f, RAMBlock *block, ram_addr_t offset);
> +
> +#define MULTIFD_FLAG_SYNC (1 << 0)
> +
> +/* This value needs to be a multiple of qemu_target_page_size() */
> +#define MULTIFD_PACKET_SIZE (512 * 1024)
> +
> +typedef struct {
> +    uint32_t magic;
> +    uint32_t version;
> +    uint32_t flags;
> +    /* maximum number of allocated pages */
> +    uint32_t pages_alloc;
> +    uint32_t pages_used;
> +    /* size of the next packet that contains pages */
> +    uint32_t next_packet_size;
> +    uint64_t packet_num;
> +    uint64_t unused[4];    /* Reserved for future use */
> +    char ramblock[256];
> +    uint64_t offset[];
> +} __attribute__((packed)) MultiFDPacket_t;
> +
> +typedef struct {
> +    /* number of used pages */
> +    uint32_t used;
> +    /* number of allocated pages */
> +    uint32_t allocated;
> +    /* global number of generated multifd packets */
> +    uint64_t packet_num;
> +    /* offset of each page */
> +    ram_addr_t *offset;
> +    /* pointer to each page */
> +    struct iovec *iov;
> +    RAMBlock *block;
> +} MultiFDPages_t;
> +
> +typedef struct {
> +    /* this fields are not changed once the thread is created */
> +    /* channel number */
> +    uint8_t id;
> +    /* channel thread name */
> +    char *name;
> +    /* channel thread id */
> +    QemuThread thread;
> +    /* communication channel */
> +    QIOChannel *c;
> +    /* sem where to wait for more work */
> +    QemuSemaphore sem;
> +    /* this mutex protects the following parameters */
> +    QemuMutex mutex;
> +    /* is this channel thread running */
> +    bool running;
> +    /* should this thread finish */
> +    bool quit;
> +    /* thread has work to do */
> +    int pending_job;
> +    /* array of pages to sent */
> +    MultiFDPages_t *pages;
> +    /* packet allocated len */
> +    uint32_t packet_len;
> +    /* pointer to the packet */
> +    MultiFDPacket_t *packet;
> +    /* multifd flags for each packet */
> +    uint32_t flags;
> +    /* size of the next packet that contains pages */
> +    uint32_t next_packet_size;
> +    /* global number of generated multifd packets */
> +    uint64_t packet_num;
> +    /* thread local variables */
> +    /* packets sent through this channel */
> +    uint64_t num_packets;
> +    /* pages sent through this channel */
> +    uint64_t num_pages;
> +    /* syncs main thread and channels */
> +    QemuSemaphore sem_sync;
> +}  MultiFDSendParams;
> +
> +typedef struct {
> +    /* this fields are not changed once the thread is created */
> +    /* channel number */
> +    uint8_t id;
> +    /* channel thread name */
> +    char *name;
> +    /* channel thread id */
> +    QemuThread thread;
> +    /* communication channel */
> +    QIOChannel *c;
> +    /* this mutex protects the following parameters */
> +    QemuMutex mutex;
> +    /* is this channel thread running */
> +    bool running;
> +    /* should this thread finish */
> +    bool quit;
> +    /* array of pages to receive */
> +    MultiFDPages_t *pages;
> +    /* packet allocated len */
> +    uint32_t packet_len;
> +    /* pointer to the packet */
> +    MultiFDPacket_t *packet;
> +    /* multifd flags for each packet */
> +    uint32_t flags;
> +    /* global number of generated multifd packets */
> +    uint64_t packet_num;
> +    /* thread local variables */
> +    /* size of the next packet that contains pages */
> +    uint32_t next_packet_size;
> +    /* packets sent through this channel */
> +    uint64_t num_packets;
> +    /* pages sent through this channel */
> +    uint64_t num_pages;
> +    /* syncs main thread and channels */
> +    QemuSemaphore sem_sync;
> +} MultiFDRecvParams;
> +
> +#endif
> +
> diff --git a/migration/ram.c b/migration/ram.c
> index c24b4cc771..ed23ed1c7c 100644
> --- a/migration/ram.c
> +++ b/migration/ram.c
> @@ -36,7 +36,6 @@
>  #include "xbzrle.h"
>  #include "ram.h"
>  #include "migration.h"
> -#include "socket.h"
>  #include "migration/register.h"
>  #include "migration/misc.h"
>  #include "qemu-file.h"
> @@ -53,9 +52,9 @@
>  #include "migration/colo.h"
>  #include "block.h"
>  #include "sysemu/sysemu.h"
> -#include "qemu/uuid.h"
>  #include "savevm.h"
>  #include "qemu/iov.h"
> +#include "multifd.h"
>  
>  /***********************************************************/
>  /* ram save/restore */
> @@ -575,983 +574,6 @@ exit:
>      return -1;
>  }
>  
> -/* Multiple fd's */
> -
> -#define MULTIFD_MAGIC 0x11223344U
> -#define MULTIFD_VERSION 1
> -
> -#define MULTIFD_FLAG_SYNC (1 << 0)
> -
> -/* This value needs to be a multiple of qemu_target_page_size() */
> -#define MULTIFD_PACKET_SIZE (512 * 1024)
> -
> -typedef struct {
> -    uint32_t magic;
> -    uint32_t version;
> -    unsigned char uuid[16]; /* QemuUUID */
> -    uint8_t id;
> -    uint8_t unused1[7];     /* Reserved for future use */
> -    uint64_t unused2[4];    /* Reserved for future use */
> -} __attribute__((packed)) MultiFDInit_t;
> -
> -typedef struct {
> -    uint32_t magic;
> -    uint32_t version;
> -    uint32_t flags;
> -    /* maximum number of allocated pages */
> -    uint32_t pages_alloc;
> -    uint32_t pages_used;
> -    /* size of the next packet that contains pages */
> -    uint32_t next_packet_size;
> -    uint64_t packet_num;
> -    uint64_t unused[4];    /* Reserved for future use */
> -    char ramblock[256];
> -    uint64_t offset[];
> -} __attribute__((packed)) MultiFDPacket_t;
> -
> -typedef struct {
> -    /* number of used pages */
> -    uint32_t used;
> -    /* number of allocated pages */
> -    uint32_t allocated;
> -    /* global number of generated multifd packets */
> -    uint64_t packet_num;
> -    /* offset of each page */
> -    ram_addr_t *offset;
> -    /* pointer to each page */
> -    struct iovec *iov;
> -    RAMBlock *block;
> -} MultiFDPages_t;
> -
> -typedef struct {
> -    /* this fields are not changed once the thread is created */
> -    /* channel number */
> -    uint8_t id;
> -    /* channel thread name */
> -    char *name;
> -    /* channel thread id */
> -    QemuThread thread;
> -    /* communication channel */
> -    QIOChannel *c;
> -    /* sem where to wait for more work */
> -    QemuSemaphore sem;
> -    /* this mutex protects the following parameters */
> -    QemuMutex mutex;
> -    /* is this channel thread running */
> -    bool running;
> -    /* should this thread finish */
> -    bool quit;
> -    /* thread has work to do */
> -    int pending_job;
> -    /* array of pages to sent */
> -    MultiFDPages_t *pages;
> -    /* packet allocated len */
> -    uint32_t packet_len;
> -    /* pointer to the packet */
> -    MultiFDPacket_t *packet;
> -    /* multifd flags for each packet */
> -    uint32_t flags;
> -    /* size of the next packet that contains pages */
> -    uint32_t next_packet_size;
> -    /* global number of generated multifd packets */
> -    uint64_t packet_num;
> -    /* thread local variables */
> -    /* packets sent through this channel */
> -    uint64_t num_packets;
> -    /* pages sent through this channel */
> -    uint64_t num_pages;
> -    /* syncs main thread and channels */
> -    QemuSemaphore sem_sync;
> -}  MultiFDSendParams;
> -
> -typedef struct {
> -    /* this fields are not changed once the thread is created */
> -    /* channel number */
> -    uint8_t id;
> -    /* channel thread name */
> -    char *name;
> -    /* channel thread id */
> -    QemuThread thread;
> -    /* communication channel */
> -    QIOChannel *c;
> -    /* this mutex protects the following parameters */
> -    QemuMutex mutex;
> -    /* is this channel thread running */
> -    bool running;
> -    /* should this thread finish */
> -    bool quit;
> -    /* array of pages to receive */
> -    MultiFDPages_t *pages;
> -    /* packet allocated len */
> -    uint32_t packet_len;
> -    /* pointer to the packet */
> -    MultiFDPacket_t *packet;
> -    /* multifd flags for each packet */
> -    uint32_t flags;
> -    /* global number of generated multifd packets */
> -    uint64_t packet_num;
> -    /* thread local variables */
> -    /* size of the next packet that contains pages */
> -    uint32_t next_packet_size;
> -    /* packets sent through this channel */
> -    uint64_t num_packets;
> -    /* pages sent through this channel */
> -    uint64_t num_pages;
> -    /* syncs main thread and channels */
> -    QemuSemaphore sem_sync;
> -} MultiFDRecvParams;
> -
> -static int multifd_send_initial_packet(MultiFDSendParams *p, Error **errp)
> -{
> -    MultiFDInit_t msg = {};
> -    int ret;
> -
> -    msg.magic = cpu_to_be32(MULTIFD_MAGIC);
> -    msg.version = cpu_to_be32(MULTIFD_VERSION);
> -    msg.id = p->id;
> -    memcpy(msg.uuid, &qemu_uuid.data, sizeof(msg.uuid));
> -
> -    ret = qio_channel_write_all(p->c, (char *)&msg, sizeof(msg), errp);
> -    if (ret != 0) {
> -        return -1;
> -    }
> -    return 0;
> -}
> -
> -static int multifd_recv_initial_packet(QIOChannel *c, Error **errp)
> -{
> -    MultiFDInit_t msg;
> -    int ret;
> -
> -    ret = qio_channel_read_all(c, (char *)&msg, sizeof(msg), errp);
> -    if (ret != 0) {
> -        return -1;
> -    }
> -
> -    msg.magic = be32_to_cpu(msg.magic);
> -    msg.version = be32_to_cpu(msg.version);
> -
> -    if (msg.magic != MULTIFD_MAGIC) {
> -        error_setg(errp, "multifd: received packet magic %x "
> -                   "expected %x", msg.magic, MULTIFD_MAGIC);
> -        return -1;
> -    }
> -
> -    if (msg.version != MULTIFD_VERSION) {
> -        error_setg(errp, "multifd: received packet version %d "
> -                   "expected %d", msg.version, MULTIFD_VERSION);
> -        return -1;
> -    }
> -
> -    if (memcmp(msg.uuid, &qemu_uuid, sizeof(qemu_uuid))) {
> -        char *uuid = qemu_uuid_unparse_strdup(&qemu_uuid);
> -        char *msg_uuid = qemu_uuid_unparse_strdup((const QemuUUID *)msg.uuid);
> -
> -        error_setg(errp, "multifd: received uuid '%s' and expected "
> -                   "uuid '%s' for channel %hhd", msg_uuid, uuid, msg.id);
> -        g_free(uuid);
> -        g_free(msg_uuid);
> -        return -1;
> -    }
> -
> -    if (msg.id > migrate_multifd_channels()) {
> -        error_setg(errp, "multifd: received channel version %d "
> -                   "expected %d", msg.version, MULTIFD_VERSION);
> -        return -1;
> -    }
> -
> -    return msg.id;
> -}
> -
> -static MultiFDPages_t *multifd_pages_init(size_t size)
> -{
> -    MultiFDPages_t *pages = g_new0(MultiFDPages_t, 1);
> -
> -    pages->allocated = size;
> -    pages->iov = g_new0(struct iovec, size);
> -    pages->offset = g_new0(ram_addr_t, size);
> -
> -    return pages;
> -}
> -
> -static void multifd_pages_clear(MultiFDPages_t *pages)
> -{
> -    pages->used = 0;
> -    pages->allocated = 0;
> -    pages->packet_num = 0;
> -    pages->block = NULL;
> -    g_free(pages->iov);
> -    pages->iov = NULL;
> -    g_free(pages->offset);
> -    pages->offset = NULL;
> -    g_free(pages);
> -}
> -
> -static void multifd_send_fill_packet(MultiFDSendParams *p)
> -{
> -    MultiFDPacket_t *packet = p->packet;
> -    int i;
> -
> -    packet->flags = cpu_to_be32(p->flags);
> -    packet->pages_alloc = cpu_to_be32(p->pages->allocated);
> -    packet->pages_used = cpu_to_be32(p->pages->used);
> -    packet->next_packet_size = cpu_to_be32(p->next_packet_size);
> -    packet->packet_num = cpu_to_be64(p->packet_num);
> -
> -    if (p->pages->block) {
> -        strncpy(packet->ramblock, p->pages->block->idstr, 256);
> -    }
> -
> -    for (i = 0; i < p->pages->used; i++) {
> -        /* there are architectures where ram_addr_t is 32 bit */
> -        uint64_t temp = p->pages->offset[i];
> -
> -        packet->offset[i] = cpu_to_be64(temp);
> -    }
> -}
> -
> -static int multifd_recv_unfill_packet(MultiFDRecvParams *p, Error **errp)
> -{
> -    MultiFDPacket_t *packet = p->packet;
> -    uint32_t pages_max = MULTIFD_PACKET_SIZE / qemu_target_page_size();
> -    RAMBlock *block;
> -    int i;
> -
> -    packet->magic = be32_to_cpu(packet->magic);
> -    if (packet->magic != MULTIFD_MAGIC) {
> -        error_setg(errp, "multifd: received packet "
> -                   "magic %x and expected magic %x",
> -                   packet->magic, MULTIFD_MAGIC);
> -        return -1;
> -    }
> -
> -    packet->version = be32_to_cpu(packet->version);
> -    if (packet->version != MULTIFD_VERSION) {
> -        error_setg(errp, "multifd: received packet "
> -                   "version %d and expected version %d",
> -                   packet->version, MULTIFD_VERSION);
> -        return -1;
> -    }
> -
> -    p->flags = be32_to_cpu(packet->flags);
> -
> -    packet->pages_alloc = be32_to_cpu(packet->pages_alloc);
> -    /*
> -     * If we received a packet that is 100 times bigger than expected
> -     * just stop migration.  It is a magic number.
> -     */
> -    if (packet->pages_alloc > pages_max * 100) {
> -        error_setg(errp, "multifd: received packet "
> -                   "with size %d and expected a maximum size of %d",
> -                   packet->pages_alloc, pages_max * 100) ;
> -        return -1;
> -    }
> -    /*
> -     * We received a packet that is bigger than expected but inside
> -     * reasonable limits (see previous comment).  Just reallocate.
> -     */
> -    if (packet->pages_alloc > p->pages->allocated) {
> -        multifd_pages_clear(p->pages);
> -        p->pages = multifd_pages_init(packet->pages_alloc);
> -    }
> -
> -    p->pages->used = be32_to_cpu(packet->pages_used);
> -    if (p->pages->used > packet->pages_alloc) {
> -        error_setg(errp, "multifd: received packet "
> -                   "with %d pages and expected maximum pages are %d",
> -                   p->pages->used, packet->pages_alloc) ;
> -        return -1;
> -    }
> -
> -    p->next_packet_size = be32_to_cpu(packet->next_packet_size);
> -    p->packet_num = be64_to_cpu(packet->packet_num);
> -
> -    if (p->pages->used == 0) {
> -        return 0;
> -    }
> -
> -    /* make sure that ramblock is 0 terminated */
> -    packet->ramblock[255] = 0;
> -    block = qemu_ram_block_by_name(packet->ramblock);
> -    if (!block) {
> -        error_setg(errp, "multifd: unknown ram block %s",
> -                   packet->ramblock);
> -        return -1;
> -    }
> -
> -    for (i = 0; i < p->pages->used; i++) {
> -        uint64_t offset = be64_to_cpu(packet->offset[i]);
> -
> -        if (offset > (block->used_length - qemu_target_page_size())) {
> -            error_setg(errp, "multifd: offset too long %" PRIu64
> -                       " (max " RAM_ADDR_FMT ")",
> -                       offset, block->max_length);
> -            return -1;
> -        }
> -        p->pages->iov[i].iov_base = block->host + offset;
> -        p->pages->iov[i].iov_len = qemu_target_page_size();
> -    }
> -
> -    return 0;
> -}
> -
> -struct {
> -    MultiFDSendParams *params;
> -    /* array of pages to sent */
> -    MultiFDPages_t *pages;
> -    /* global number of generated multifd packets */
> -    uint64_t packet_num;
> -    /* send channels ready */
> -    QemuSemaphore channels_ready;
> -    /*
> -     * Have we already run terminate threads.  There is a race when it
> -     * happens that we got one error while we are exiting.
> -     * We will use atomic operations.  Only valid values are 0 and 1.
> -     */
> -    int exiting;
> -} *multifd_send_state;
> -
> -/*
> - * How we use multifd_send_state->pages and channel->pages?
> - *
> - * We create a pages for each channel, and a main one.  Each time that
> - * we need to send a batch of pages we interchange the ones between
> - * multifd_send_state and the channel that is sending it.  There are
> - * two reasons for that:
> - *    - to not have to do so many mallocs during migration
> - *    - to make easier to know what to free at the end of migration
> - *
> - * This way we always know who is the owner of each "pages" struct,
> - * and we don't need any locking.  It belongs to the migration thread
> - * or to the channel thread.  Switching is safe because the migration
> - * thread is using the channel mutex when changing it, and the channel
> - * have to had finish with its own, otherwise pending_job can't be
> - * false.
> - */
> -
> -static int multifd_send_pages(QEMUFile *f)
> -{
> -    int i;
> -    static int next_channel;
> -    MultiFDSendParams *p = NULL; /* make happy gcc */
> -    MultiFDPages_t *pages = multifd_send_state->pages;
> -    uint64_t transferred;
> -
> -    if (atomic_read(&multifd_send_state->exiting)) {
> -        return -1;
> -    }
> -
> -    qemu_sem_wait(&multifd_send_state->channels_ready);
> -    for (i = next_channel;; i = (i + 1) % migrate_multifd_channels()) {
> -        p = &multifd_send_state->params[i];
> -
> -        qemu_mutex_lock(&p->mutex);
> -        if (p->quit) {
> -            error_report("%s: channel %d has already quit!", __func__, i);
> -            qemu_mutex_unlock(&p->mutex);
> -            return -1;
> -        }
> -        if (!p->pending_job) {
> -            p->pending_job++;
> -            next_channel = (i + 1) % migrate_multifd_channels();
> -            break;
> -        }
> -        qemu_mutex_unlock(&p->mutex);
> -    }
> -    assert(!p->pages->used);
> -    assert(!p->pages->block);
> -
> -    p->packet_num = multifd_send_state->packet_num++;
> -    multifd_send_state->pages = p->pages;
> -    p->pages = pages;
> -    transferred = ((uint64_t) pages->used) * qemu_target_page_size()
> -                + p->packet_len;
> -    qemu_file_update_transfer(f, transferred);
> -    ram_counters.multifd_bytes += transferred;
> -    ram_counters.transferred += transferred;;
> -    qemu_mutex_unlock(&p->mutex);
> -    qemu_sem_post(&p->sem);
> -
> -    return 1;
> -}
> -
> -static int multifd_queue_page(QEMUFile *f, RAMBlock *block, ram_addr_t offset)
> -{
> -    MultiFDPages_t *pages = multifd_send_state->pages;
> -
> -    if (!pages->block) {
> -        pages->block = block;
> -    }
> -
> -    if (pages->block == block) {
> -        pages->offset[pages->used] = offset;
> -        pages->iov[pages->used].iov_base = block->host + offset;
> -        pages->iov[pages->used].iov_len = qemu_target_page_size();
> -        pages->used++;
> -
> -        if (pages->used < pages->allocated) {
> -            return 1;
> -        }
> -    }
> -
> -    if (multifd_send_pages(f) < 0) {
> -        return -1;
> -    }
> -
> -    if (pages->block != block) {
> -        return  multifd_queue_page(f, block, offset);
> -    }
> -
> -    return 1;
> -}
> -
> -static void multifd_send_terminate_threads(Error *err)
> -{
> -    int i;
> -
> -    trace_multifd_send_terminate_threads(err != NULL);
> -
> -    if (err) {
> -        MigrationState *s = migrate_get_current();
> -        migrate_set_error(s, err);
> -        if (s->state == MIGRATION_STATUS_SETUP ||
> -            s->state == MIGRATION_STATUS_PRE_SWITCHOVER ||
> -            s->state == MIGRATION_STATUS_DEVICE ||
> -            s->state == MIGRATION_STATUS_ACTIVE) {
> -            migrate_set_state(&s->state, s->state,
> -                              MIGRATION_STATUS_FAILED);
> -        }
> -    }
> -
> -    /*
> -     * We don't want to exit each threads twice.  Depending on where
> -     * we get the error, or if there are two independent errors in two
> -     * threads at the same time, we can end calling this function
> -     * twice.
> -     */
> -    if (atomic_xchg(&multifd_send_state->exiting, 1)) {
> -        return;
> -    }
> -
> -    for (i = 0; i < migrate_multifd_channels(); i++) {
> -        MultiFDSendParams *p = &multifd_send_state->params[i];
> -
> -        qemu_mutex_lock(&p->mutex);
> -        p->quit = true;
> -        qemu_sem_post(&p->sem);
> -        qemu_mutex_unlock(&p->mutex);
> -    }
> -}
> -
> -void multifd_save_cleanup(void)
> -{
> -    int i;
> -
> -    if (!migrate_use_multifd()) {
> -        return;
> -    }
> -    multifd_send_terminate_threads(NULL);
> -    for (i = 0; i < migrate_multifd_channels(); i++) {
> -        MultiFDSendParams *p = &multifd_send_state->params[i];
> -
> -        if (p->running) {
> -            qemu_thread_join(&p->thread);
> -        }
> -    }
> -    for (i = 0; i < migrate_multifd_channels(); i++) {
> -        MultiFDSendParams *p = &multifd_send_state->params[i];
> -
> -        socket_send_channel_destroy(p->c);
> -        p->c = NULL;
> -        qemu_mutex_destroy(&p->mutex);
> -        qemu_sem_destroy(&p->sem);
> -        qemu_sem_destroy(&p->sem_sync);
> -        g_free(p->name);
> -        p->name = NULL;
> -        multifd_pages_clear(p->pages);
> -        p->pages = NULL;
> -        p->packet_len = 0;
> -        g_free(p->packet);
> -        p->packet = NULL;
> -    }
> -    qemu_sem_destroy(&multifd_send_state->channels_ready);
> -    g_free(multifd_send_state->params);
> -    multifd_send_state->params = NULL;
> -    multifd_pages_clear(multifd_send_state->pages);
> -    multifd_send_state->pages = NULL;
> -    g_free(multifd_send_state);
> -    multifd_send_state = NULL;
> -}
> -
> -static void multifd_send_sync_main(QEMUFile *f)
> -{
> -    int i;
> -
> -    if (!migrate_use_multifd()) {
> -        return;
> -    }
> -    if (multifd_send_state->pages->used) {
> -        if (multifd_send_pages(f) < 0) {
> -            error_report("%s: multifd_send_pages fail", __func__);
> -            return;
> -        }
> -    }
> -    for (i = 0; i < migrate_multifd_channels(); i++) {
> -        MultiFDSendParams *p = &multifd_send_state->params[i];
> -
> -        trace_multifd_send_sync_main_signal(p->id);
> -
> -        qemu_mutex_lock(&p->mutex);
> -
> -        if (p->quit) {
> -            error_report("%s: channel %d has already quit", __func__, i);
> -            qemu_mutex_unlock(&p->mutex);
> -            return;
> -        }
> -
> -        p->packet_num = multifd_send_state->packet_num++;
> -        p->flags |= MULTIFD_FLAG_SYNC;
> -        p->pending_job++;
> -        qemu_file_update_transfer(f, p->packet_len);
> -        ram_counters.multifd_bytes += p->packet_len;
> -        ram_counters.transferred += p->packet_len;
> -        qemu_mutex_unlock(&p->mutex);
> -        qemu_sem_post(&p->sem);
> -    }
> -    for (i = 0; i < migrate_multifd_channels(); i++) {
> -        MultiFDSendParams *p = &multifd_send_state->params[i];
> -
> -        trace_multifd_send_sync_main_wait(p->id);
> -        qemu_sem_wait(&p->sem_sync);
> -    }
> -    trace_multifd_send_sync_main(multifd_send_state->packet_num);
> -}
> -
> -static void *multifd_send_thread(void *opaque)
> -{
> -    MultiFDSendParams *p = opaque;
> -    Error *local_err = NULL;
> -    int ret = 0;
> -    uint32_t flags = 0;
> -
> -    trace_multifd_send_thread_start(p->id);
> -    rcu_register_thread();
> -
> -    if (multifd_send_initial_packet(p, &local_err) < 0) {
> -        ret = -1;
> -        goto out;
> -    }
> -    /* initial packet */
> -    p->num_packets = 1;
> -
> -    while (true) {
> -        qemu_sem_wait(&p->sem);
> -
> -        if (atomic_read(&multifd_send_state->exiting)) {
> -            break;
> -        }
> -        qemu_mutex_lock(&p->mutex);
> -
> -        if (p->pending_job) {
> -            uint32_t used = p->pages->used;
> -            uint64_t packet_num = p->packet_num;
> -            flags = p->flags;
> -
> -            p->next_packet_size = used * qemu_target_page_size();
> -            multifd_send_fill_packet(p);
> -            p->flags = 0;
> -            p->num_packets++;
> -            p->num_pages += used;
> -            p->pages->used = 0;
> -            p->pages->block = NULL;
> -            qemu_mutex_unlock(&p->mutex);
> -
> -            trace_multifd_send(p->id, packet_num, used, flags,
> -                               p->next_packet_size);
> -
> -            ret = qio_channel_write_all(p->c, (void *)p->packet,
> -                                        p->packet_len, &local_err);
> -            if (ret != 0) {
> -                break;
> -            }
> -
> -            if (used) {
> -                ret = qio_channel_writev_all(p->c, p->pages->iov,
> -                                             used, &local_err);
> -                if (ret != 0) {
> -                    break;
> -                }
> -            }
> -
> -            qemu_mutex_lock(&p->mutex);
> -            p->pending_job--;
> -            qemu_mutex_unlock(&p->mutex);
> -
> -            if (flags & MULTIFD_FLAG_SYNC) {
> -                qemu_sem_post(&p->sem_sync);
> -            }
> -            qemu_sem_post(&multifd_send_state->channels_ready);
> -        } else if (p->quit) {
> -            qemu_mutex_unlock(&p->mutex);
> -            break;
> -        } else {
> -            qemu_mutex_unlock(&p->mutex);
> -            /* sometimes there are spurious wakeups */
> -        }
> -    }
> -
> -out:
> -    if (local_err) {
> -        trace_multifd_send_error(p->id);
> -        multifd_send_terminate_threads(local_err);
> -    }
> -
> -    /*
> -     * Error happen, I will exit, but I can't just leave, tell
> -     * who pay attention to me.
> -     */
> -    if (ret != 0) {
> -        qemu_sem_post(&p->sem_sync);
> -        qemu_sem_post(&multifd_send_state->channels_ready);
> -    }
> -
> -    qemu_mutex_lock(&p->mutex);
> -    p->running = false;
> -    qemu_mutex_unlock(&p->mutex);
> -
> -    rcu_unregister_thread();
> -    trace_multifd_send_thread_end(p->id, p->num_packets, p->num_pages);
> -
> -    return NULL;
> -}
> -
> -static void multifd_new_send_channel_async(QIOTask *task, gpointer opaque)
> -{
> -    MultiFDSendParams *p = opaque;
> -    QIOChannel *sioc = QIO_CHANNEL(qio_task_get_source(task));
> -    Error *local_err = NULL;
> -
> -    trace_multifd_new_send_channel_async(p->id);
> -    if (qio_task_propagate_error(task, &local_err)) {
> -        migrate_set_error(migrate_get_current(), local_err);
> -        multifd_save_cleanup();
> -    } else {
> -        p->c = QIO_CHANNEL(sioc);
> -        qio_channel_set_delay(p->c, false);
> -        p->running = true;
> -        qemu_thread_create(&p->thread, p->name, multifd_send_thread, p,
> -                           QEMU_THREAD_JOINABLE);
> -    }
> -}
> -
> -int multifd_save_setup(Error **errp)
> -{
> -    int thread_count;
> -    uint32_t page_count = MULTIFD_PACKET_SIZE / qemu_target_page_size();
> -    uint8_t i;
> -
> -    if (!migrate_use_multifd()) {
> -        return 0;
> -    }
> -    thread_count = migrate_multifd_channels();
> -    multifd_send_state = g_malloc0(sizeof(*multifd_send_state));
> -    multifd_send_state->params = g_new0(MultiFDSendParams, thread_count);
> -    multifd_send_state->pages = multifd_pages_init(page_count);
> -    qemu_sem_init(&multifd_send_state->channels_ready, 0);
> -    atomic_set(&multifd_send_state->exiting, 0);
> -
> -    for (i = 0; i < thread_count; i++) {
> -        MultiFDSendParams *p = &multifd_send_state->params[i];
> -
> -        qemu_mutex_init(&p->mutex);
> -        qemu_sem_init(&p->sem, 0);
> -        qemu_sem_init(&p->sem_sync, 0);
> -        p->quit = false;
> -        p->pending_job = 0;
> -        p->id = i;
> -        p->pages = multifd_pages_init(page_count);
> -        p->packet_len = sizeof(MultiFDPacket_t)
> -                      + sizeof(uint64_t) * page_count;
> -        p->packet = g_malloc0(p->packet_len);
> -        p->packet->magic = cpu_to_be32(MULTIFD_MAGIC);
> -        p->packet->version = cpu_to_be32(MULTIFD_VERSION);
> -        p->name = g_strdup_printf("multifdsend_%d", i);
> -        socket_send_channel_create(multifd_new_send_channel_async, p);
> -    }
> -    return 0;
> -}
> -
> -struct {
> -    MultiFDRecvParams *params;
> -    /* number of created threads */
> -    int count;
> -    /* syncs main thread and channels */
> -    QemuSemaphore sem_sync;
> -    /* global number of generated multifd packets */
> -    uint64_t packet_num;
> -} *multifd_recv_state;
> -
> -static void multifd_recv_terminate_threads(Error *err)
> -{
> -    int i;
> -
> -    trace_multifd_recv_terminate_threads(err != NULL);
> -
> -    if (err) {
> -        MigrationState *s = migrate_get_current();
> -        migrate_set_error(s, err);
> -        if (s->state == MIGRATION_STATUS_SETUP ||
> -            s->state == MIGRATION_STATUS_ACTIVE) {
> -            migrate_set_state(&s->state, s->state,
> -                              MIGRATION_STATUS_FAILED);
> -        }
> -    }
> -
> -    for (i = 0; i < migrate_multifd_channels(); i++) {
> -        MultiFDRecvParams *p = &multifd_recv_state->params[i];
> -
> -        qemu_mutex_lock(&p->mutex);
> -        p->quit = true;
> -        /*
> -         * We could arrive here for two reasons:
> -         *  - normal quit, i.e. everything went fine, just finished
> -         *  - error quit: We close the channels so the channel threads
> -         *    finish the qio_channel_read_all_eof()
> -         */
> -        if (p->c) {
> -            qio_channel_shutdown(p->c, QIO_CHANNEL_SHUTDOWN_BOTH, NULL);
> -        }
> -        qemu_mutex_unlock(&p->mutex);
> -    }
> -}
> -
> -int multifd_load_cleanup(Error **errp)
> -{
> -    int i;
> -    int ret = 0;
> -
> -    if (!migrate_use_multifd()) {
> -        return 0;
> -    }
> -    multifd_recv_terminate_threads(NULL);
> -    for (i = 0; i < migrate_multifd_channels(); i++) {
> -        MultiFDRecvParams *p = &multifd_recv_state->params[i];
> -
> -        if (p->running) {
> -            p->quit = true;
> -            /*
> -             * multifd_recv_thread may hung at MULTIFD_FLAG_SYNC handle code,
> -             * however try to wakeup it without harm in cleanup phase.
> -             */
> -            qemu_sem_post(&p->sem_sync);
> -            qemu_thread_join(&p->thread);
> -        }
> -    }
> -    for (i = 0; i < migrate_multifd_channels(); i++) {
> -        MultiFDRecvParams *p = &multifd_recv_state->params[i];
> -
> -        object_unref(OBJECT(p->c));
> -        p->c = NULL;
> -        qemu_mutex_destroy(&p->mutex);
> -        qemu_sem_destroy(&p->sem_sync);
> -        g_free(p->name);
> -        p->name = NULL;
> -        multifd_pages_clear(p->pages);
> -        p->pages = NULL;
> -        p->packet_len = 0;
> -        g_free(p->packet);
> -        p->packet = NULL;
> -    }
> -    qemu_sem_destroy(&multifd_recv_state->sem_sync);
> -    g_free(multifd_recv_state->params);
> -    multifd_recv_state->params = NULL;
> -    g_free(multifd_recv_state);
> -    multifd_recv_state = NULL;
> -
> -    return ret;
> -}
> -
> -static void multifd_recv_sync_main(void)
> -{
> -    int i;
> -
> -    if (!migrate_use_multifd()) {
> -        return;
> -    }
> -    for (i = 0; i < migrate_multifd_channels(); i++) {
> -        MultiFDRecvParams *p = &multifd_recv_state->params[i];
> -
> -        trace_multifd_recv_sync_main_wait(p->id);
> -        qemu_sem_wait(&multifd_recv_state->sem_sync);
> -    }
> -    for (i = 0; i < migrate_multifd_channels(); i++) {
> -        MultiFDRecvParams *p = &multifd_recv_state->params[i];
> -
> -        qemu_mutex_lock(&p->mutex);
> -        if (multifd_recv_state->packet_num < p->packet_num) {
> -            multifd_recv_state->packet_num = p->packet_num;
> -        }
> -        qemu_mutex_unlock(&p->mutex);
> -        trace_multifd_recv_sync_main_signal(p->id);
> -        qemu_sem_post(&p->sem_sync);
> -    }
> -    trace_multifd_recv_sync_main(multifd_recv_state->packet_num);
> -}
> -
> -static void *multifd_recv_thread(void *opaque)
> -{
> -    MultiFDRecvParams *p = opaque;
> -    Error *local_err = NULL;
> -    int ret;
> -
> -    trace_multifd_recv_thread_start(p->id);
> -    rcu_register_thread();
> -
> -    while (true) {
> -        uint32_t used;
> -        uint32_t flags;
> -
> -        if (p->quit) {
> -            break;
> -        }
> -
> -        ret = qio_channel_read_all_eof(p->c, (void *)p->packet,
> -                                       p->packet_len, &local_err);
> -        if (ret == 0) {   /* EOF */
> -            break;
> -        }
> -        if (ret == -1) {   /* Error */
> -            break;
> -        }
> -
> -        qemu_mutex_lock(&p->mutex);
> -        ret = multifd_recv_unfill_packet(p, &local_err);
> -        if (ret) {
> -            qemu_mutex_unlock(&p->mutex);
> -            break;
> -        }
> -
> -        used = p->pages->used;
> -        flags = p->flags;
> -        trace_multifd_recv(p->id, p->packet_num, used, flags,
> -                           p->next_packet_size);
> -        p->num_packets++;
> -        p->num_pages += used;
> -        qemu_mutex_unlock(&p->mutex);
> -
> -        if (used) {
> -            ret = qio_channel_readv_all(p->c, p->pages->iov,
> -                                        used, &local_err);
> -            if (ret != 0) {
> -                break;
> -            }
> -        }
> -
> -        if (flags & MULTIFD_FLAG_SYNC) {
> -            qemu_sem_post(&multifd_recv_state->sem_sync);
> -            qemu_sem_wait(&p->sem_sync);
> -        }
> -    }
> -
> -    if (local_err) {
> -        multifd_recv_terminate_threads(local_err);
> -    }
> -    qemu_mutex_lock(&p->mutex);
> -    p->running = false;
> -    qemu_mutex_unlock(&p->mutex);
> -
> -    rcu_unregister_thread();
> -    trace_multifd_recv_thread_end(p->id, p->num_packets, p->num_pages);
> -
> -    return NULL;
> -}
> -
> -int multifd_load_setup(Error **errp)
> -{
> -    int thread_count;
> -    uint32_t page_count = MULTIFD_PACKET_SIZE / qemu_target_page_size();
> -    uint8_t i;
> -
> -    if (!migrate_use_multifd()) {
> -        return 0;
> -    }
> -    thread_count = migrate_multifd_channels();
> -    multifd_recv_state = g_malloc0(sizeof(*multifd_recv_state));
> -    multifd_recv_state->params = g_new0(MultiFDRecvParams, thread_count);
> -    atomic_set(&multifd_recv_state->count, 0);
> -    qemu_sem_init(&multifd_recv_state->sem_sync, 0);
> -
> -    for (i = 0; i < thread_count; i++) {
> -        MultiFDRecvParams *p = &multifd_recv_state->params[i];
> -
> -        qemu_mutex_init(&p->mutex);
> -        qemu_sem_init(&p->sem_sync, 0);
> -        p->quit = false;
> -        p->id = i;
> -        p->pages = multifd_pages_init(page_count);
> -        p->packet_len = sizeof(MultiFDPacket_t)
> -                      + sizeof(uint64_t) * page_count;
> -        p->packet = g_malloc0(p->packet_len);
> -        p->name = g_strdup_printf("multifdrecv_%d", i);
> -    }
> -    return 0;
> -}
> -
> -bool multifd_recv_all_channels_created(void)
> -{
> -    int thread_count = migrate_multifd_channels();
> -
> -    if (!migrate_use_multifd()) {
> -        return true;
> -    }
> -
> -    return thread_count == atomic_read(&multifd_recv_state->count);
> -}
> -
> -/*
> - * Try to receive all multifd channels to get ready for the migration.
> - * - Return true and do not set @errp when correctly receving all channels;
> - * - Return false and do not set @errp when correctly receiving the current one;
> - * - Return false and set @errp when failing to receive the current channel.
> - */
> -bool multifd_recv_new_channel(QIOChannel *ioc, Error **errp)
> -{
> -    MultiFDRecvParams *p;
> -    Error *local_err = NULL;
> -    int id;
> -
> -    id = multifd_recv_initial_packet(ioc, &local_err);
> -    if (id < 0) {
> -        multifd_recv_terminate_threads(local_err);
> -        error_propagate_prepend(errp, local_err,
> -                                "failed to receive packet"
> -                                " via multifd channel %d: ",
> -                                atomic_read(&multifd_recv_state->count));
> -        return false;
> -    }
> -    trace_multifd_recv_new_channel(id);
> -
> -    p = &multifd_recv_state->params[id];
> -    if (p->c != NULL) {
> -        error_setg(&local_err, "multifd: received id '%d' already setup'",
> -                   id);
> -        multifd_recv_terminate_threads(local_err);
> -        error_propagate(errp, local_err);
> -        return false;
> -    }
> -    p->c = ioc;
> -    object_ref(OBJECT(ioc));
> -    /* initial packet */
> -    p->num_packets = 1;
> -
> -    p->running = true;
> -    qemu_thread_create(&p->thread, p->name, multifd_recv_thread, p,
> -                       QEMU_THREAD_JOINABLE);
> -    atomic_inc(&multifd_recv_state->count);
> -    return atomic_read(&multifd_recv_state->count) ==
> -           migrate_multifd_channels();
> -}
> -
>  /**
>   * save_page_header: write page header to wire
>   *
> diff --git a/migration/ram.h b/migration/ram.h
> index 42be471d52..a553d40751 100644
> --- a/migration/ram.h
> +++ b/migration/ram.h
> @@ -41,13 +41,6 @@ int xbzrle_cache_resize(int64_t new_size, Error **errp);
>  uint64_t ram_bytes_remaining(void);
>  uint64_t ram_bytes_total(void);
>  
> -int multifd_save_setup(Error **errp);
> -void multifd_save_cleanup(void);
> -int multifd_load_setup(Error **errp);
> -int multifd_load_cleanup(Error **errp);
> -bool multifd_recv_all_channels_created(void);
> -bool multifd_recv_new_channel(QIOChannel *ioc, Error **errp);
> -
>  uint64_t ram_pagesize_summary(void);
>  int ram_save_queue_pages(const char *rbname, ram_addr_t start, ram_addr_t len);
>  void acct_update_position(QEMUFile *f, size_t size, bool zero);
> -- 
> 2.24.1
> 
--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK

[PATCH v3 18/21] migration: Make no compression operations into its own structure

[Juan Quintela]

It will be used later.

Signed-off-by: Juan Quintela <quintela@redhat.com>

---
Move setup of ->ops helper to proper place (wei)
Rename s/none/nocomp/ (dave)
Introduce MULTIFD_FLAG_NOCOMP
right order of arguments for print
---
 migration/migration.c |   9 +++
 migration/migration.h |   1 +
 migration/multifd.c   | 183 ++++++++++++++++++++++++++++++++++++++++--
 migration/multifd.h   |  22 +++++
 migration/ram.c       |   1 +
 5 files changed, 208 insertions(+), 8 deletions(-)

diff --git a/migration/migration.c b/migration/migration.c
index 3501bc3353..d25fdb3e6b 100644
--- a/migration/migration.c
+++ b/migration/migration.c
@@ -2245,6 +2245,15 @@ int migrate_multifd_channels(void)
     return s->parameters.multifd_channels;
 }
 
+int migrate_multifd_method(void)
+{
+    MigrationState *s;
+
+    s = migrate_get_current();
+
+    return s->parameters.multifd_compress;
+}
+
 int migrate_use_xbzrle(void)
 {
     MigrationState *s;
diff --git a/migration/migration.h b/migration/migration.h
index 8473ddfc88..3b0b413a93 100644
--- a/migration/migration.h
+++ b/migration/migration.h
@@ -300,6 +300,7 @@ bool migrate_auto_converge(void);
 bool migrate_use_multifd(void);
 bool migrate_pause_before_switchover(void);
 int migrate_multifd_channels(void);
+int migrate_multifd_method(void);
 
 int migrate_use_xbzrle(void);
 int64_t migrate_xbzrle_cache_size(void);
diff --git a/migration/multifd.c b/migration/multifd.c
index 1875bb3aaa..353140cd25 100644
--- a/migration/multifd.c
+++ b/migration/multifd.c
@@ -38,6 +38,132 @@ typedef struct {
     uint64_t unused2[4];    /* Reserved for future use */
 } __attribute__((packed)) MultiFDInit_t;
 
+/* Multifd without compression */
+
+/**
+ * nocomp_send_setup: setup send side
+ *
+ * For no compression this function does nothing.
+ *
+ * Returns 0 for success or -1 for error
+ *
+ * @p: Params for the channel that we are using
+ * @errp: pointer to an error
+ */
+static int nocomp_send_setup(MultiFDSendParams *p, Error **errp)
+{
+    return 0;
+}
+
+/**
+ * nocomp_send_cleanup: cleanup send side
+ *
+ * For no compression this function does nothing.
+ *
+ * @p: Params for the channel that we are using
+ */
+static void nocomp_send_cleanup(MultiFDSendParams *p, Error **errp)
+{
+    return;
+}
+
+/**
+ * nocomp_send_prepare: prepare date to be able to send
+ *
+ * For no compression we just have to calculate the size of the
+ * packet.
+ *
+ * Returns 0 for success or -1 for error
+ *
+ * @p: Params for the channel that we are using
+ * @used: number of pages used
+ * @errp: pointer to an error
+ */
+static int nocomp_send_prepare(MultiFDSendParams *p, uint32_t used,
+                               Error **errp)
+{
+    p->next_packet_size = used * qemu_target_page_size();
+    p->flags |= MULTIFD_FLAG_NOCOMP;
+    return 0;
+}
+
+/**
+ * nocomp_send_write: do the actual write of the data
+ *
+ * For no compression we just have to write the data.
+ *
+ * Returns 0 for success or -1 for error
+ *
+ * @p: Params for the channel that we are using
+ * @used: number of pages used
+ * @errp: pointer to an error
+ */
+static int nocomp_send_write(MultiFDSendParams *p, uint32_t used, Error **errp)
+{
+    return qio_channel_writev_all(p->c, p->pages->iov, used, errp);
+}
+
+/**
+ * nocomp_recv_setup: setup receive side
+ *
+ * For no compression this function does nothing.
+ *
+ * Returns 0 for success or -1 for error
+ *
+ * @p: Params for the channel that we are using
+ * @errp: pointer to an error
+ */
+static int nocomp_recv_setup(MultiFDRecvParams *p, Error **errp)
+{
+    return 0;
+}
+
+/**
+ * nocomp_recv_cleanup: setup receive side
+ *
+ * For no compression this function does nothing.
+ *
+ * @p: Params for the channel that we are using
+ */
+static void nocomp_recv_cleanup(MultiFDRecvParams *p)
+{
+}
+
+/**
+ * nocomp_recv_pages: read the data from the channel into actual pages
+ *
+ * For no compression we just need to read things into the correct place.
+ *
+ * Returns 0 for success or -1 for error
+ *
+ * @p: Params for the channel that we are using
+ * @used: number of pages used
+ * @errp: pointer to an error
+ */
+static int nocomp_recv_pages(MultiFDRecvParams *p, uint32_t used, Error **errp)
+{
+    if (p->flags != MULTIFD_FLAG_NOCOMP) {
+        error_setg(errp, "multifd %d: flags received %x flags expected %x",
+                   p->id, p->flags, MULTIFD_FLAG_NOCOMP);
+        return -1;
+    }
+    return qio_channel_readv_all(p->c, p->pages->iov, used, errp);
+}
+
+static MultiFDMethods multifd_nocomp_ops = {
+    .send_setup = nocomp_send_setup,
+    .send_cleanup = nocomp_send_cleanup,
+    .send_prepare = nocomp_send_prepare,
+    .send_write = nocomp_send_write,
+    .recv_setup = nocomp_recv_setup,
+    .recv_cleanup = nocomp_recv_cleanup,
+    .recv_pages = nocomp_recv_pages
+};
+
+static MultiFDMethods *multifd_ops[MULTIFD_COMPRESS__MAX] = {
+    [MULTIFD_COMPRESS_NONE] = &multifd_nocomp_ops,
+};
+
 static int multifd_send_initial_packet(MultiFDSendParams *p, Error **errp)
 {
     MultiFDInit_t msg = {};
@@ -246,6 +372,8 @@ struct {
      * We will use atomic operations.  Only valid values are 0 and 1.
      */
     int exiting;
+    /* multifd ops */
+    MultiFDMethods *ops;
 } *multifd_send_state;
 
 /*
@@ -397,6 +525,7 @@ void multifd_save_cleanup(void)
     }
     for (i = 0; i < migrate_multifd_channels(); i++) {
         MultiFDSendParams *p = &multifd_send_state->params[i];
+        Error *local_err = NULL;
 
         socket_send_channel_destroy(p->c);
         p->c = NULL;
@@ -410,6 +539,10 @@ void multifd_save_cleanup(void)
         p->packet_len = 0;
         g_free(p->packet);
         p->packet = NULL;
+        multifd_send_state->ops->send_cleanup(p, &local_err);
+        if (local_err) {
+            migrate_set_error(migrate_get_current(), local_err);
+        }
     }
     qemu_sem_destroy(&multifd_send_state->channels_ready);
     g_free(multifd_send_state->params);
@@ -494,7 +627,14 @@ static void *multifd_send_thread(void *opaque)
             uint64_t packet_num = p->packet_num;
             flags = p->flags;
 
-            p->next_packet_size = used * qemu_target_page_size();
+            if (used) {
+                ret = multifd_send_state->ops->send_prepare(p, used,
+                                                            &local_err);
+                if (ret != 0) {
+                    qemu_mutex_unlock(&p->mutex);
+                    break;
+                }
+            }
             multifd_send_fill_packet(p);
             p->flags = 0;
             p->num_packets++;
@@ -513,8 +653,7 @@ static void *multifd_send_thread(void *opaque)
             }
 
             if (used) {
-                ret = qio_channel_writev_all(p->c, p->pages->iov,
-                                             used, &local_err);
+                ret = multifd_send_state->ops->send_write(p, used, &local_err);
                 if (ret != 0) {
                     break;
                 }
@@ -596,6 +735,7 @@ int multifd_save_setup(Error **errp)
     multifd_send_state->pages = multifd_pages_init(page_count);
     qemu_sem_init(&multifd_send_state->channels_ready, 0);
     atomic_set(&multifd_send_state->exiting, 0);
+    multifd_send_state->ops = multifd_ops[migrate_multifd_method()];
 
     for (i = 0; i < thread_count; i++) {
         MultiFDSendParams *p = &multifd_send_state->params[i];
@@ -615,6 +755,18 @@ int multifd_save_setup(Error **errp)
         p->name = g_strdup_printf("multifdsend_%d", i);
         socket_send_channel_create(multifd_new_send_channel_async, p);
     }
+
+    for (i = 0; i < thread_count; i++) {
+        MultiFDSendParams *p = &multifd_send_state->params[i];
+        Error *local_err = NULL;
+        int ret;
+
+        ret = multifd_send_state->ops->send_setup(p, &local_err);
+        if (ret) {
+            error_propagate(errp, local_err);
+            return ret;
+        }
+    }
     return 0;
 }
 
@@ -626,6 +778,8 @@ struct {
     QemuSemaphore sem_sync;
     /* global number of generated multifd packets */
     uint64_t packet_num;
+    /* multifd ops */
+    MultiFDMethods *ops;
 } *multifd_recv_state;
 
 static void multifd_recv_terminate_threads(Error *err)
@@ -665,7 +819,6 @@ static void multifd_recv_terminate_threads(Error *err)
 int multifd_load_cleanup(Error **errp)
 {
     int i;
-    int ret = 0;
 
     if (!migrate_use_multifd()) {
         return 0;
@@ -698,6 +851,7 @@ int multifd_load_cleanup(Error **errp)
         p->packet_len = 0;
         g_free(p->packet);
         p->packet = NULL;
+        multifd_recv_state->ops->recv_cleanup(p);
     }
     qemu_sem_destroy(&multifd_recv_state->sem_sync);
     g_free(multifd_recv_state->params);
@@ -705,7 +859,7 @@ int multifd_load_cleanup(Error **errp)
     g_free(multifd_recv_state);
     multifd_recv_state = NULL;
 
-    return ret;
+    return 0;
 }
 
 void multifd_recv_sync_main(void)
@@ -770,6 +924,8 @@ static void *multifd_recv_thread(void *opaque)
 
         used = p->pages->used;
         flags = p->flags;
+        /* recv methods don't know how to handle the SYNC flag */
+        p->flags &= ~MULTIFD_FLAG_SYNC;
         trace_multifd_recv(p->id, p->packet_num, used, flags,
                            p->next_packet_size);
         p->num_packets++;
@@ -777,8 +933,7 @@ static void *multifd_recv_thread(void *opaque)
         qemu_mutex_unlock(&p->mutex);
 
         if (used) {
-            ret = qio_channel_readv_all(p->c, p->pages->iov,
-                                        used, &local_err);
+            ret = multifd_recv_state->ops->recv_pages(p, used, &local_err);
             if (ret != 0) {
                 break;
             }
@@ -817,6 +972,7 @@ int multifd_load_setup(Error **errp)
     multifd_recv_state->params = g_new0(MultiFDRecvParams, thread_count);
     atomic_set(&multifd_recv_state->count, 0);
     qemu_sem_init(&multifd_recv_state->sem_sync, 0);
+    multifd_recv_state->ops = multifd_ops[migrate_multifd_method()];
 
     for (i = 0; i < thread_count; i++) {
         MultiFDRecvParams *p = &multifd_recv_state->params[i];
@@ -831,6 +987,18 @@ int multifd_load_setup(Error **errp)
         p->packet = g_malloc0(p->packet_len);
         p->name = g_strdup_printf("multifdrecv_%d", i);
     }
+
+    for (i = 0; i < thread_count; i++) {
+        MultiFDRecvParams *p = &multifd_recv_state->params[i];
+        Error *local_err = NULL;
+        int ret;
+
+        ret = multifd_recv_state->ops->recv_setup(p, &local_err);
+        if (ret) {
+            error_propagate(errp, local_err);
+            return ret;
+        }
+    }
     return 0;
 }
 
@@ -888,4 +1056,3 @@ bool multifd_recv_new_channel(QIOChannel *ioc, Error **errp)
     return atomic_read(&multifd_recv_state->count) ==
            migrate_multifd_channels();
 }
-
diff --git a/migration/multifd.h b/migration/multifd.h
index d8b0205977..8edea4fdac 100644
--- a/migration/multifd.h
+++ b/migration/multifd.h
@@ -24,6 +24,7 @@ void multifd_send_sync_main(QEMUFile *f);
 int multifd_queue_page(QEMUFile *f, RAMBlock *block, ram_addr_t offset);
 
 #define MULTIFD_FLAG_SYNC (1 << 0)
+#define MULTIFD_FLAG_NOCOMP (1 << 1)
 
 /* This value needs to be a multiple of qemu_target_page_size() */
 #define MULTIFD_PACKET_SIZE (512 * 1024)
@@ -96,6 +97,8 @@ typedef struct {
     uint64_t num_pages;
     /* syncs main thread and channels */
     QemuSemaphore sem_sync;
+    /* used for compression methods */
+    void *data;
 }  MultiFDSendParams;
 
 typedef struct {
@@ -133,7 +136,26 @@ typedef struct {
     uint64_t num_pages;
     /* syncs main thread and channels */
     QemuSemaphore sem_sync;
+    /* used for de-compression methods */
+    void *data;
 } MultiFDRecvParams;
 
+typedef struct {
+    /* Setup for sending side */
+    int (*send_setup)(MultiFDSendParams *p, Error **errp);
+    /* Cleanup for sending side */
+    void (*send_cleanup)(MultiFDSendParams *p, Error **errp);
+    /* Prepare the send packet */
+    int (*send_prepare)(MultiFDSendParams *p, uint32_t used, Error **errp);
+    /* Write the send packet */
+    int (*send_write)(MultiFDSendParams *p, uint32_t used, Error **errp);
+    /* Setup for receiving side */
+    int (*recv_setup)(MultiFDRecvParams *p, Error **errp);
+    /* Cleanup for receiving side */
+    void (*recv_cleanup)(MultiFDRecvParams *p);
+    /* Read all pages */
+    int (*recv_pages)(MultiFDRecvParams *p, uint32_t used, Error **errp);
+} MultiFDMethods;
+
 #endif
 
diff --git a/migration/ram.c b/migration/ram.c
index ed23ed1c7c..73a141bb60 100644
--- a/migration/ram.c
+++ b/migration/ram.c
@@ -43,6 +43,7 @@
 #include "page_cache.h"
 #include "qemu/error-report.h"
 #include "qapi/error.h"
+#include "qapi/qapi-types-migration.h"
 #include "qapi/qapi-events-migration.h"
 #include "qapi/qmp/qerror.h"
 #include "trace.h"
-- 
2.24.1

Re: [PATCH v3 18/21] migration: Make no compression operations into its own structure

[Dr. David Alan Gilbert]

* Juan Quintela (quintela@redhat.com) wrote:
> It will be used later.
> 
> Signed-off-by: Juan Quintela <quintela@redhat.com>
> 
> ---
> Move setup of ->ops helper to proper place (wei)
> Rename s/none/nocomp/ (dave)
> Introduce MULTIFD_FLAG_NOCOMP
> right order of arguments for print
> ---
>  migration/migration.c |   9 +++
>  migration/migration.h |   1 +
>  migration/multifd.c   | 183 ++++++++++++++++++++++++++++++++++++++++--
>  migration/multifd.h   |  22 +++++
>  migration/ram.c       |   1 +
>  5 files changed, 208 insertions(+), 8 deletions(-)
> 
> diff --git a/migration/migration.c b/migration/migration.c
> index 3501bc3353..d25fdb3e6b 100644
> --- a/migration/migration.c
> +++ b/migration/migration.c
> @@ -2245,6 +2245,15 @@ int migrate_multifd_channels(void)
>      return s->parameters.multifd_channels;
>  }
>  
> +int migrate_multifd_method(void)
> +{
> +    MigrationState *s;
> +
> +    s = migrate_get_current();
> +
> +    return s->parameters.multifd_compress;
> +}

Shouldn't that be a MultifdCompress enum returned?

> +
>  int migrate_use_xbzrle(void)
>  {
>      MigrationState *s;
> diff --git a/migration/migration.h b/migration/migration.h
> index 8473ddfc88..3b0b413a93 100644
> --- a/migration/migration.h
> +++ b/migration/migration.h
> @@ -300,6 +300,7 @@ bool migrate_auto_converge(void);
>  bool migrate_use_multifd(void);
>  bool migrate_pause_before_switchover(void);
>  int migrate_multifd_channels(void);
> +int migrate_multifd_method(void);
>  
>  int migrate_use_xbzrle(void);
>  int64_t migrate_xbzrle_cache_size(void);
> diff --git a/migration/multifd.c b/migration/multifd.c
> index 1875bb3aaa..353140cd25 100644
> --- a/migration/multifd.c
> +++ b/migration/multifd.c
> @@ -38,6 +38,132 @@ typedef struct {
>      uint64_t unused2[4];    /* Reserved for future use */
>  } __attribute__((packed)) MultiFDInit_t;
>  
> +/* Multifd without compression */
> +
> +/**
> + * nocomp_send_setup: setup send side
> + *
> + * For no compression this function does nothing.
> + *
> + * Returns 0 for success or -1 for error
> + *
> + * @p: Params for the channel that we are using
> + * @errp: pointer to an error
> + */
> +static int nocomp_send_setup(MultiFDSendParams *p, Error **errp)
> +{
> +    return 0;
> +}
> +
> +/**
> + * nocomp_send_cleanup: cleanup send side
> + *
> + * For no compression this function does nothing.
> + *
> + * @p: Params for the channel that we are using
> + */
> +static void nocomp_send_cleanup(MultiFDSendParams *p, Error **errp)
> +{
> +    return;
> +}
> +
> +/**
> + * nocomp_send_prepare: prepare date to be able to send
> + *
> + * For no compression we just have to calculate the size of the
> + * packet.
> + *
> + * Returns 0 for success or -1 for error
> + *
> + * @p: Params for the channel that we are using
> + * @used: number of pages used
> + * @errp: pointer to an error
> + */
> +static int nocomp_send_prepare(MultiFDSendParams *p, uint32_t used,
> +                               Error **errp)
> +{
> +    p->next_packet_size = used * qemu_target_page_size();
> +    p->flags |= MULTIFD_FLAG_NOCOMP;
> +    return 0;
> +}
> +
> +/**
> + * nocomp_send_write: do the actual write of the data
> + *
> + * For no compression we just have to write the data.
> + *
> + * Returns 0 for success or -1 for error
> + *
> + * @p: Params for the channel that we are using
> + * @used: number of pages used
> + * @errp: pointer to an error
> + */
> +static int nocomp_send_write(MultiFDSendParams *p, uint32_t used, Error **errp)
> +{
> +    return qio_channel_writev_all(p->c, p->pages->iov, used, errp);
> +}
> +
> +/**
> + * nocomp_recv_setup: setup receive side
> + *
> + * For no compression this function does nothing.
> + *
> + * Returns 0 for success or -1 for error
> + *
> + * @p: Params for the channel that we are using
> + * @errp: pointer to an error
> + */
> +static int nocomp_recv_setup(MultiFDRecvParams *p, Error **errp)
> +{
> +    return 0;
> +}
> +
> +/**
> + * nocomp_recv_cleanup: setup receive side
> + *
> + * For no compression this function does nothing.
> + *
> + * @p: Params for the channel that we are using
> + */
> +static void nocomp_recv_cleanup(MultiFDRecvParams *p)
> +{
> +}
> +
> +/**
> + * nocomp_recv_pages: read the data from the channel into actual pages
> + *
> + * For no compression we just need to read things into the correct place.
> + *
> + * Returns 0 for success or -1 for error
> + *
> + * @p: Params for the channel that we are using
> + * @used: number of pages used
> + * @errp: pointer to an error
> + */
> +static int nocomp_recv_pages(MultiFDRecvParams *p, uint32_t used, Error **errp)
> +{
> +    if (p->flags != MULTIFD_FLAG_NOCOMP) {
> +        error_setg(errp, "multifd %d: flags received %x flags expected %x",
> +                   p->id, p->flags, MULTIFD_FLAG_NOCOMP);
> +        return -1;
> +    }
> +    return qio_channel_readv_all(p->c, p->pages->iov, used, errp);
> +}
> +
> +static MultiFDMethods multifd_nocomp_ops = {
> +    .send_setup = nocomp_send_setup,
> +    .send_cleanup = nocomp_send_cleanup,
> +    .send_prepare = nocomp_send_prepare,
> +    .send_write = nocomp_send_write,
> +    .recv_setup = nocomp_recv_setup,
> +    .recv_cleanup = nocomp_recv_cleanup,
> +    .recv_pages = nocomp_recv_pages
> +};
> +
> +static MultiFDMethods *multifd_ops[MULTIFD_COMPRESS__MAX] = {
> +    [MULTIFD_COMPRESS_NONE] = &multifd_nocomp_ops,
> +};
> +
>  static int multifd_send_initial_packet(MultiFDSendParams *p, Error **errp)
>  {
>      MultiFDInit_t msg = {};
> @@ -246,6 +372,8 @@ struct {
>       * We will use atomic operations.  Only valid values are 0 and 1.
>       */
>      int exiting;
> +    /* multifd ops */
> +    MultiFDMethods *ops;
>  } *multifd_send_state;
>  
>  /*
> @@ -397,6 +525,7 @@ void multifd_save_cleanup(void)
>      }
>      for (i = 0; i < migrate_multifd_channels(); i++) {
>          MultiFDSendParams *p = &multifd_send_state->params[i];
> +        Error *local_err = NULL;
>  
>          socket_send_channel_destroy(p->c);
>          p->c = NULL;
> @@ -410,6 +539,10 @@ void multifd_save_cleanup(void)
>          p->packet_len = 0;
>          g_free(p->packet);
>          p->packet = NULL;
> +        multifd_send_state->ops->send_cleanup(p, &local_err);
> +        if (local_err) {
> +            migrate_set_error(migrate_get_current(), local_err);
> +        }
>      }
>      qemu_sem_destroy(&multifd_send_state->channels_ready);
>      g_free(multifd_send_state->params);
> @@ -494,7 +627,14 @@ static void *multifd_send_thread(void *opaque)
>              uint64_t packet_num = p->packet_num;
>              flags = p->flags;
>  
> -            p->next_packet_size = used * qemu_target_page_size();
> +            if (used) {
> +                ret = multifd_send_state->ops->send_prepare(p, used,
> +                                                            &local_err);
> +                if (ret != 0) {
> +                    qemu_mutex_unlock(&p->mutex);
> +                    break;
> +                }
> +            }
>              multifd_send_fill_packet(p);
>              p->flags = 0;
>              p->num_packets++;
> @@ -513,8 +653,7 @@ static void *multifd_send_thread(void *opaque)
>              }
>  
>              if (used) {
> -                ret = qio_channel_writev_all(p->c, p->pages->iov,
> -                                             used, &local_err);
> +                ret = multifd_send_state->ops->send_write(p, used, &local_err);
>                  if (ret != 0) {
>                      break;
>                  }
> @@ -596,6 +735,7 @@ int multifd_save_setup(Error **errp)
>      multifd_send_state->pages = multifd_pages_init(page_count);
>      qemu_sem_init(&multifd_send_state->channels_ready, 0);
>      atomic_set(&multifd_send_state->exiting, 0);
> +    multifd_send_state->ops = multifd_ops[migrate_multifd_method()];
>  
>      for (i = 0; i < thread_count; i++) {
>          MultiFDSendParams *p = &multifd_send_state->params[i];
> @@ -615,6 +755,18 @@ int multifd_save_setup(Error **errp)
>          p->name = g_strdup_printf("multifdsend_%d", i);
>          socket_send_channel_create(multifd_new_send_channel_async, p);
>      }
> +
> +    for (i = 0; i < thread_count; i++) {
> +        MultiFDSendParams *p = &multifd_send_state->params[i];
> +        Error *local_err = NULL;
> +        int ret;
> +
> +        ret = multifd_send_state->ops->send_setup(p, &local_err);
> +        if (ret) {
> +            error_propagate(errp, local_err);
> +            return ret;
> +        }
> +    }
>      return 0;
>  }
>  
> @@ -626,6 +778,8 @@ struct {
>      QemuSemaphore sem_sync;
>      /* global number of generated multifd packets */
>      uint64_t packet_num;
> +    /* multifd ops */
> +    MultiFDMethods *ops;
>  } *multifd_recv_state;
>  
>  static void multifd_recv_terminate_threads(Error *err)
> @@ -665,7 +819,6 @@ static void multifd_recv_terminate_threads(Error *err)
>  int multifd_load_cleanup(Error **errp)
>  {
>      int i;
> -    int ret = 0;
>  
>      if (!migrate_use_multifd()) {
>          return 0;
> @@ -698,6 +851,7 @@ int multifd_load_cleanup(Error **errp)
>          p->packet_len = 0;
>          g_free(p->packet);
>          p->packet = NULL;
> +        multifd_recv_state->ops->recv_cleanup(p);
>      }
>      qemu_sem_destroy(&multifd_recv_state->sem_sync);
>      g_free(multifd_recv_state->params);
> @@ -705,7 +859,7 @@ int multifd_load_cleanup(Error **errp)
>      g_free(multifd_recv_state);
>      multifd_recv_state = NULL;
>  
> -    return ret;
> +    return 0;
>  }
>  
>  void multifd_recv_sync_main(void)
> @@ -770,6 +924,8 @@ static void *multifd_recv_thread(void *opaque)
>  
>          used = p->pages->used;
>          flags = p->flags;
> +        /* recv methods don't know how to handle the SYNC flag */
> +        p->flags &= ~MULTIFD_FLAG_SYNC;
>          trace_multifd_recv(p->id, p->packet_num, used, flags,
>                             p->next_packet_size);
>          p->num_packets++;
> @@ -777,8 +933,7 @@ static void *multifd_recv_thread(void *opaque)
>          qemu_mutex_unlock(&p->mutex);
>  
>          if (used) {
> -            ret = qio_channel_readv_all(p->c, p->pages->iov,
> -                                        used, &local_err);
> +            ret = multifd_recv_state->ops->recv_pages(p, used, &local_err);
>              if (ret != 0) {
>                  break;
>              }
> @@ -817,6 +972,7 @@ int multifd_load_setup(Error **errp)
>      multifd_recv_state->params = g_new0(MultiFDRecvParams, thread_count);
>      atomic_set(&multifd_recv_state->count, 0);
>      qemu_sem_init(&multifd_recv_state->sem_sync, 0);
> +    multifd_recv_state->ops = multifd_ops[migrate_multifd_method()];
>  
>      for (i = 0; i < thread_count; i++) {
>          MultiFDRecvParams *p = &multifd_recv_state->params[i];
> @@ -831,6 +987,18 @@ int multifd_load_setup(Error **errp)
>          p->packet = g_malloc0(p->packet_len);
>          p->name = g_strdup_printf("multifdrecv_%d", i);
>      }
> +
> +    for (i = 0; i < thread_count; i++) {
> +        MultiFDRecvParams *p = &multifd_recv_state->params[i];
> +        Error *local_err = NULL;
> +        int ret;
> +
> +        ret = multifd_recv_state->ops->recv_setup(p, &local_err);
> +        if (ret) {
> +            error_propagate(errp, local_err);
> +            return ret;
> +        }
> +    }
>      return 0;
>  }
>  
> @@ -888,4 +1056,3 @@ bool multifd_recv_new_channel(QIOChannel *ioc, Error **errp)
>      return atomic_read(&multifd_recv_state->count) ==
>             migrate_multifd_channels();
>  }
> -
> diff --git a/migration/multifd.h b/migration/multifd.h
> index d8b0205977..8edea4fdac 100644
> --- a/migration/multifd.h
> +++ b/migration/multifd.h
> @@ -24,6 +24,7 @@ void multifd_send_sync_main(QEMUFile *f);
>  int multifd_queue_page(QEMUFile *f, RAMBlock *block, ram_addr_t offset);
>  
>  #define MULTIFD_FLAG_SYNC (1 << 0)
> +#define MULTIFD_FLAG_NOCOMP (1 << 1)

I don't think this should be a set of individual flags; in later patches
you define a flag for zlib and another for zstd etc etc - but you can't
combine them - you could never have FLAG_NOCOMP|FLAG_ZSTD|FLAG_ZLIB - so
this should be a 3 or 4 bit field which contains a compression id (0
being none).  The ID can't exactly be the migrate_multifd_method() enum
value - because I don't think that's defined to be stable (?).

Dave

>  /* This value needs to be a multiple of qemu_target_page_size() */
>  #define MULTIFD_PACKET_SIZE (512 * 1024)
> @@ -96,6 +97,8 @@ typedef struct {
>      uint64_t num_pages;
>      /* syncs main thread and channels */
>      QemuSemaphore sem_sync;
> +    /* used for compression methods */
> +    void *data;
>  }  MultiFDSendParams;
>  
>  typedef struct {
> @@ -133,7 +136,26 @@ typedef struct {
>      uint64_t num_pages;
>      /* syncs main thread and channels */
>      QemuSemaphore sem_sync;
> +    /* used for de-compression methods */
> +    void *data;
>  } MultiFDRecvParams;
>  
> +typedef struct {
> +    /* Setup for sending side */
> +    int (*send_setup)(MultiFDSendParams *p, Error **errp);
> +    /* Cleanup for sending side */
> +    void (*send_cleanup)(MultiFDSendParams *p, Error **errp);
> +    /* Prepare the send packet */
> +    int (*send_prepare)(MultiFDSendParams *p, uint32_t used, Error **errp);
> +    /* Write the send packet */
> +    int (*send_write)(MultiFDSendParams *p, uint32_t used, Error **errp);
> +    /* Setup for receiving side */
> +    int (*recv_setup)(MultiFDRecvParams *p, Error **errp);
> +    /* Cleanup for receiving side */
> +    void (*recv_cleanup)(MultiFDRecvParams *p);
> +    /* Read all pages */
> +    int (*recv_pages)(MultiFDRecvParams *p, uint32_t used, Error **errp);
> +} MultiFDMethods;
> +
>  #endif
>  
> diff --git a/migration/ram.c b/migration/ram.c
> index ed23ed1c7c..73a141bb60 100644
> --- a/migration/ram.c
> +++ b/migration/ram.c
> @@ -43,6 +43,7 @@
>  #include "page_cache.h"
>  #include "qemu/error-report.h"
>  #include "qapi/error.h"
> +#include "qapi/qapi-types-migration.h"
>  #include "qapi/qapi-events-migration.h"
>  #include "qapi/qmp/qerror.h"
>  #include "trace.h"

Unintended?


> -- 
> 2.24.1
> 
--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK

Re: [PATCH v3 18/21] migration: Make no compression operations into its own structure

[Juan Quintela]

"Dr. David Alan Gilbert" <dgilbert@redhat.com> wrote:
> * Juan Quintela (quintela@redhat.com) wrote:
>> It will be used later.
>> 
>> Signed-off-by: Juan Quintela <quintela@redhat.com>
>> 

>> +int migrate_multifd_method(void)
>> +{
>> +    MigrationState *s;
>> +
>> +    s = migrate_get_current();
>> +
>> +    return s->parameters.multifd_compress;
>> +}
>
> Shouldn't that be a MultifdCompress enum returned?

You are right here.
>>  
>>  #define MULTIFD_FLAG_SYNC (1 << 0)
>> +#define MULTIFD_FLAG_NOCOMP (1 << 1)
>
> I don't think this should be a set of individual flags; in later patches
> you define a flag for zlib and another for zstd etc etc - but you can't
> combine them - you could never have FLAG_NOCOMP|FLAG_ZSTD|FLAG_ZLIB - so
> this should be a 3 or 4 bit field which contains a compression id (0
> being none).  The ID can't exactly be the migrate_multifd_method() enum
> value - because I don't think that's defined to be stable (?).

The idea is to catch up if we got an incorrect packet with an incorrect
flag.

But yes, I agree that it could be the same expecting a value here.
The problem is that I already have the flags field.

Would it be ok for you if I reserve 3 bits for this?  (right now 2
should be enough).

Thanks, Juan.

Re: [PATCH v3 18/21] migration: Make no compression operations into its own structure

[Dr. David Alan Gilbert]

* Juan Quintela (quintela@redhat.com) wrote:
> "Dr. David Alan Gilbert" <dgilbert@redhat.com> wrote:
> > * Juan Quintela (quintela@redhat.com) wrote:
> >> It will be used later.
> >> 
> >> Signed-off-by: Juan Quintela <quintela@redhat.com>
> >> 
> 
> >> +int migrate_multifd_method(void)
> >> +{
> >> +    MigrationState *s;
> >> +
> >> +    s = migrate_get_current();
> >> +
> >> +    return s->parameters.multifd_compress;
> >> +}
> >
> > Shouldn't that be a MultifdCompress enum returned?
> 
> You are right here.
> >>  
> >>  #define MULTIFD_FLAG_SYNC (1 << 0)
> >> +#define MULTIFD_FLAG_NOCOMP (1 << 1)
> >
> > I don't think this should be a set of individual flags; in later patches
> > you define a flag for zlib and another for zstd etc etc - but you can't
> > combine them - you could never have FLAG_NOCOMP|FLAG_ZSTD|FLAG_ZLIB - so
> > this should be a 3 or 4 bit field which contains a compression id (0
> > being none).  The ID can't exactly be the migrate_multifd_method() enum
> > value - because I don't think that's defined to be stable (?).
> 
> The idea is to catch up if we got an incorrect packet with an incorrect
> flag.
> 
> But yes, I agree that it could be the same expecting a value here.
> The problem is that I already have the flags field.
> 
> Would it be ok for you if I reserve 3 bits for this?  (right now 2
> should be enough).

Yeh that's fine - I was going to suggest 4 bits, but 3 is OK;
just define something like:

  MULTIFD_FLAG_COMP_MASK (7 << 1)

Dave

> Thanks, Juan.
--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK

[PATCH v3 19/21] migration: Add zlib compression multifd support

[Juan Quintela]

Signed-off-by: Juan Quintela <quintela@redhat.com>
---
 hw/core/qdev-properties.c    |   2 +-
 migration/Makefile.objs      |   1 +
 migration/multifd-zlib.c     | 289 +++++++++++++++++++++++++++++++++++
 migration/multifd.c          |   6 +
 migration/multifd.h          |   4 +
 qapi/migration.json          |   2 +-
 tests/qtest/migration-test.c |   6 +
 7 files changed, 308 insertions(+), 2 deletions(-)
 create mode 100644 migration/multifd-zlib.c

diff --git a/hw/core/qdev-properties.c b/hw/core/qdev-properties.c
index ff6a752b19..07ec75d8e3 100644
--- a/hw/core/qdev-properties.c
+++ b/hw/core/qdev-properties.c
@@ -647,7 +647,7 @@ const PropertyInfo qdev_prop_fdc_drive_type = {
 const PropertyInfo qdev_prop_multifd_compress = {
     .name = "MultifdCompress",
     .description = "multifd_compress values, "
-                   "none",
+                   "none/zlib",
     .enum_table = &MultifdCompress_lookup,
     .get = get_enum,
     .set = set_enum,
diff --git a/migration/Makefile.objs b/migration/Makefile.objs
index d3623d5f9b..0308caa5c5 100644
--- a/migration/Makefile.objs
+++ b/migration/Makefile.objs
@@ -8,6 +8,7 @@ common-obj-y += xbzrle.o postcopy-ram.o
 common-obj-y += qjson.o
 common-obj-y += block-dirty-bitmap.o
 common-obj-y += multifd.o
+common-obj-y += multifd-zlib.o
 
 common-obj-$(CONFIG_RDMA) += rdma.o
 
diff --git a/migration/multifd-zlib.c b/migration/multifd-zlib.c
new file mode 100644
index 0000000000..33d7ee6741
--- /dev/null
+++ b/migration/multifd-zlib.c
@@ -0,0 +1,289 @@
+/*
+ * Multifd zlib compression implementation
+ *
+ * Copyright (c) 2020 Red Hat Inc
+ *
+ * Authors:
+ *  Juan Quintela <quintela@redhat.com>
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ */
+
+#include "qemu/osdep.h"
+#include <zlib.h>
+#include "qemu/rcu.h"
+#include "exec/target_page.h"
+#include "qapi/error.h"
+#include "migration.h"
+#include "trace.h"
+#include "multifd.h"
+
+struct zlib_data {
+    /* stream for compression */
+    z_stream zs;
+    /* compressed buffer */
+    uint8_t *zbuff;
+    /* size of compressed buffer */
+    uint32_t zbuff_len;
+};
+
+/* Multifd zlib compression */
+
+/**
+ * zlib_send_setup: setup send side
+ *
+ * Setup each channel with zlib compression.
+ *
+ * Returns 0 for success or -1 for error
+ *
+ * @p: Params for the channel that we are using
+ * @errp: pointer to an error
+ */
+static int zlib_send_setup(MultiFDSendParams *p, Error **errp)
+{
+    uint32_t page_count = MULTIFD_PACKET_SIZE / qemu_target_page_size();
+    struct zlib_data *z = g_malloc0(sizeof(struct zlib_data));
+    z_stream *zs = &z->zs;
+
+    p->data = z;
+    zs->zalloc = Z_NULL;
+    zs->zfree = Z_NULL;
+    zs->opaque = Z_NULL;
+    if (deflateInit(zs, migrate_compress_level()) != Z_OK) {
+        g_free(z);
+        error_setg(errp, "multifd %d: deflate init failed", p->id);
+        return -1;
+    }
+    /* We will never have more than page_count pages */
+    z->zbuff_len = page_count * qemu_target_page_size();
+    z->zbuff_len *= 2;
+    z->zbuff = g_try_malloc(z->zbuff_len);
+    if (!z->zbuff) {
+        g_free(z);
+        error_setg(errp, "multifd %d: out of memory for zbuff", p->id);
+        return -1;
+    }
+    return 0;
+}
+
+/**
+ * zlib_send_cleanup: cleanup send side
+ *
+ * Close the channel and return memory.
+ *
+ * @p: Params for the channel that we are using
+ */
+static void zlib_send_cleanup(MultiFDSendParams *p, Error **errp)
+{
+    struct zlib_data *z = p->data;
+
+    deflateEnd(&z->zs);
+    g_free(z->zbuff);
+    z->zbuff = NULL;
+    g_free(p->data);
+    p->data = NULL;
+}
+
+/**
+ * zlib_send_prepare: prepare date to be able to send
+ *
+ * Create a compressed buffer with all the pages that we are going to
+ * send.
+ *
+ * Returns 0 for success or -1 for error
+ *
+ * @p: Params for the channel that we are using
+ * @used: number of pages used
+ */
+static int zlib_send_prepare(MultiFDSendParams *p, uint32_t used, Error **errp)
+{
+    struct iovec *iov = p->pages->iov;
+    struct zlib_data *z = p->data;
+    z_stream *zs = &z->zs;
+    uint32_t out_size = 0;
+    int ret;
+    uint32_t i;
+
+    for (i = 0; i < used; i++) {
+        uint32_t available = z->zbuff_len - out_size;
+        int flush = Z_NO_FLUSH;
+
+        if (i == used  - 1) {
+            flush = Z_SYNC_FLUSH;
+        }
+
+        zs->avail_in = iov[i].iov_len;
+        zs->next_in = iov[i].iov_base;
+
+        zs->avail_out = available;
+        zs->next_out = z->zbuff + out_size;
+
+        ret = deflate(zs, flush);
+        if (ret != Z_OK) {
+            error_setg(errp, "multifd %d: deflate returned %d instead of Z_OK",
+                       p->id, ret);
+            return -1;
+        }
+        out_size += available - zs->avail_out;
+    }
+    p->next_packet_size = out_size;
+    p->flags |= MULTIFD_FLAG_ZLIB;
+
+    return 0;
+}
+
+/**
+ * zlib_send_write: do the actual write of the data
+ *
+ * Do the actual write of the comprresed buffer.
+ *
+ * Returns 0 for success or -1 for error
+ *
+ * @p: Params for the channel that we are using
+ * @used: number of pages used
+ * @errp: pointer to an error
+ */
+static int zlib_send_write(MultiFDSendParams *p, uint32_t used, Error **errp)
+{
+    struct zlib_data *z = p->data;
+
+    return qio_channel_write_all(p->c, (void *)z->zbuff, p->next_packet_size,
+                                 errp);
+}
+
+/**
+ * zlib_recv_setup: setup receive side
+ *
+ * Create the compressed channel and buffer.
+ *
+ * Returns 0 for success or -1 for error
+ *
+ * @p: Params for the channel that we are using
+ * @errp: pointer to an error
+ */
+static int zlib_recv_setup(MultiFDRecvParams *p, Error **errp)
+{
+    uint32_t page_count = MULTIFD_PACKET_SIZE / qemu_target_page_size();
+    struct zlib_data *z = g_malloc0(sizeof(struct zlib_data));
+    z_stream *zs = &z->zs;
+
+    p->data = z;
+    zs->zalloc = Z_NULL;
+    zs->zfree = Z_NULL;
+    zs->opaque = Z_NULL;
+    zs->avail_in = 0;
+    zs->next_in = Z_NULL;
+    if (inflateInit(zs) != Z_OK) {
+        error_setg(errp, "multifd %d: inflate init failed", p->id);
+        return -1;
+    }
+    /* We will never have more than page_count pages */
+    z->zbuff_len = page_count * qemu_target_page_size();
+    /* We know compression "could" use more space */
+    z->zbuff_len *= 2;
+    z->zbuff = g_try_malloc(z->zbuff_len);
+    if (!z->zbuff) {
+        error_setg(errp, "multifd %d: out of memory for zbuff", p->id);
+        return -1;
+    }
+    return 0;
+}
+
+/**
+ * zlib_recv_cleanup: setup receive side
+ *
+ * For no compression this function does nothing.
+ *
+ * @p: Params for the channel that we are using
+ */
+static void zlib_recv_cleanup(MultiFDRecvParams *p)
+{
+    struct zlib_data *z = p->data;
+
+    inflateEnd(&z->zs);
+    g_free(z->zbuff);
+    z->zbuff = NULL;
+    g_free(p->data);
+    p->data = NULL;
+}
+
+/**
+ * zlib_recv_pages: read the data from the channel into actual pages
+ *
+ * Read the compressed buffer, and uncompress it into the actual
+ * pages.
+ *
+ * Returns 0 for success or -1 for error
+ *
+ * @p: Params for the channel that we are using
+ * @used: number of pages used
+ * @errp: pointer to an error
+ */
+static int zlib_recv_pages(MultiFDRecvParams *p, uint32_t used, Error **errp)
+{
+    uint32_t in_size = p->next_packet_size;
+    uint32_t out_size = 0;
+    uint32_t expected_size = used * qemu_target_page_size();
+    struct zlib_data *z = p->data;
+    z_stream *zs = &z->zs;
+    int ret;
+    int i;
+
+    if (p->flags != MULTIFD_FLAG_ZLIB) {
+        error_setg(errp, "multifd %d: flags received %x flags expected %x",
+                   p->id, MULTIFD_FLAG_ZLIB, p->flags);
+        return -1;
+    }
+    ret = qio_channel_read_all(p->c, (void *)z->zbuff, in_size, errp);
+
+    if (ret != 0) {
+        return ret;
+    }
+
+    zs->avail_in = in_size;
+    zs->next_in = z->zbuff;
+
+    for (i = 0; i < used; i++) {
+        struct iovec *iov = &p->pages->iov[i];
+        int flush = Z_NO_FLUSH;
+
+        if (i == used  - 1) {
+            flush = Z_SYNC_FLUSH;
+        }
+
+        zs->avail_out = iov->iov_len;
+        zs->next_out = iov->iov_base;
+
+        ret = inflate(zs, flush);
+        if (ret != Z_OK) {
+            error_setg(errp, "multifd %d: inflate returned %d instead of Z_OK",
+                       p->id, ret);
+            return ret;
+        }
+        out_size += iov->iov_len;
+    }
+    if (out_size != expected_size) {
+        error_setg(errp, "multifd %d: packet size received %d size expected %d",
+                   p->id, out_size, expected_size);
+        return -1;
+    }
+    return 0;
+}
+
+static MultiFDMethods multifd_zlib_ops = {
+    .send_setup = zlib_send_setup,
+    .send_cleanup = zlib_send_cleanup,
+    .send_prepare = zlib_send_prepare,
+    .send_write = zlib_send_write,
+    .recv_setup = zlib_recv_setup,
+    .recv_cleanup = zlib_recv_cleanup,
+    .recv_pages = zlib_recv_pages
+};
+
+static void multifd_zlib_register(void)
+{
+    multifd_register_ops(MULTIFD_COMPRESS_ZLIB, &multifd_zlib_ops);
+}
+
+migration_init(multifd_zlib_register);
diff --git a/migration/multifd.c b/migration/multifd.c
index 353140cd25..a1fc451d49 100644
--- a/migration/multifd.c
+++ b/migration/multifd.c
@@ -164,6 +164,12 @@ static MultiFDMethods *multifd_ops[MULTIFD_COMPRESS__MAX] = {
     [MULTIFD_COMPRESS_NONE] = &multifd_nocomp_ops,
 };
 
+void multifd_register_ops(int method, MultiFDMethods *ops)
+{
+    assert(0 < method && method < MULTIFD_COMPRESS__MAX);
+    multifd_ops[method] = ops;
+}
+
 static int multifd_send_initial_packet(MultiFDSendParams *p, Error **errp)
 {
     MultiFDInit_t msg = {};
diff --git a/migration/multifd.h b/migration/multifd.h
index 8edea4fdac..85542f3222 100644
--- a/migration/multifd.h
+++ b/migration/multifd.h
@@ -23,8 +23,10 @@ void multifd_recv_sync_main(void);
 void multifd_send_sync_main(QEMUFile *f);
 int multifd_queue_page(QEMUFile *f, RAMBlock *block, ram_addr_t offset);
 
+/* Multifd Compression flags */
 #define MULTIFD_FLAG_SYNC (1 << 0)
 #define MULTIFD_FLAG_NOCOMP (1 << 1)
+#define MULTIFD_FLAG_ZLIB (1 << 2)
 
 /* This value needs to be a multiple of qemu_target_page_size() */
 #define MULTIFD_PACKET_SIZE (512 * 1024)
@@ -157,5 +159,7 @@ typedef struct {
     int (*recv_pages)(MultiFDRecvParams *p, uint32_t used, Error **errp);
 } MultiFDMethods;
 
+void multifd_register_ops(int method, MultiFDMethods *ops);
+
 #endif
 
diff --git a/qapi/migration.json b/qapi/migration.json
index c2891e6ebf..1714ea51e3 100644
--- a/qapi/migration.json
+++ b/qapi/migration.json
@@ -499,7 +499,7 @@
 #
 ##
 { 'enum': 'MultifdCompress',
-  'data': [ 'none' ] }
+  'data': [ 'none', 'zlib' ] }
 
 ##
 # @MigrationParameter:
diff --git a/tests/qtest/migration-test.c b/tests/qtest/migration-test.c
index 3d5d2aba8c..985a7d4b97 100644
--- a/tests/qtest/migration-test.c
+++ b/tests/qtest/migration-test.c
@@ -1313,6 +1313,11 @@ static void test_multifd_tcp_none(void)
     test_multifd_tcp("none");
 }
 
+static void test_multifd_tcp_zlib(void)
+{
+    test_multifd_tcp("zlib");
+}
+
 /*
  * This test does:
  *  source               target
@@ -1475,6 +1480,7 @@ int main(int argc, char **argv)
     qtest_add_func("/migration/auto_converge", test_migrate_auto_converge);
     qtest_add_func("/migration/multifd/tcp/none", test_multifd_tcp_none);
     qtest_add_func("/migration/multifd/tcp/cancel", test_multifd_tcp_cancel);
+    qtest_add_func("/migration/multifd/tcp/zlib", test_multifd_tcp_zlib);
 
     ret = g_test_run();
 
-- 
2.24.1

Re: [PATCH v3 19/21] migration: Add zlib compression multifd support

[Dr. David Alan Gilbert]

* Juan Quintela (quintela@redhat.com) wrote:
> Signed-off-by: Juan Quintela <quintela@redhat.com>
> ---
>  hw/core/qdev-properties.c    |   2 +-
>  migration/Makefile.objs      |   1 +
>  migration/multifd-zlib.c     | 289 +++++++++++++++++++++++++++++++++++
>  migration/multifd.c          |   6 +
>  migration/multifd.h          |   4 +
>  qapi/migration.json          |   2 +-
>  tests/qtest/migration-test.c |   6 +
>  7 files changed, 308 insertions(+), 2 deletions(-)
>  create mode 100644 migration/multifd-zlib.c
> 
> diff --git a/hw/core/qdev-properties.c b/hw/core/qdev-properties.c
> index ff6a752b19..07ec75d8e3 100644
> --- a/hw/core/qdev-properties.c
> +++ b/hw/core/qdev-properties.c
> @@ -647,7 +647,7 @@ const PropertyInfo qdev_prop_fdc_drive_type = {
>  const PropertyInfo qdev_prop_multifd_compress = {
>      .name = "MultifdCompress",
>      .description = "multifd_compress values, "
> -                   "none",
> +                   "none/zlib",
>      .enum_table = &MultifdCompress_lookup,
>      .get = get_enum,
>      .set = set_enum,
> diff --git a/migration/Makefile.objs b/migration/Makefile.objs
> index d3623d5f9b..0308caa5c5 100644
> --- a/migration/Makefile.objs
> +++ b/migration/Makefile.objs
> @@ -8,6 +8,7 @@ common-obj-y += xbzrle.o postcopy-ram.o
>  common-obj-y += qjson.o
>  common-obj-y += block-dirty-bitmap.o
>  common-obj-y += multifd.o
> +common-obj-y += multifd-zlib.o
>  
>  common-obj-$(CONFIG_RDMA) += rdma.o
>  
> diff --git a/migration/multifd-zlib.c b/migration/multifd-zlib.c
> new file mode 100644
> index 0000000000..33d7ee6741
> --- /dev/null
> +++ b/migration/multifd-zlib.c
> @@ -0,0 +1,289 @@
> +/*
> + * Multifd zlib compression implementation
> + *
> + * Copyright (c) 2020 Red Hat Inc
> + *
> + * Authors:
> + *  Juan Quintela <quintela@redhat.com>
> + *
> + * This work is licensed under the terms of the GNU GPL, version 2 or later.
> + * See the COPYING file in the top-level directory.
> + */
> +
> +#include "qemu/osdep.h"
> +#include <zlib.h>
> +#include "qemu/rcu.h"
> +#include "exec/target_page.h"
> +#include "qapi/error.h"
> +#include "migration.h"
> +#include "trace.h"
> +#include "multifd.h"
> +
> +struct zlib_data {
> +    /* stream for compression */
> +    z_stream zs;
> +    /* compressed buffer */
> +    uint8_t *zbuff;
> +    /* size of compressed buffer */
> +    uint32_t zbuff_len;
> +};
> +
> +/* Multifd zlib compression */
> +
> +/**
> + * zlib_send_setup: setup send side
> + *
> + * Setup each channel with zlib compression.
> + *
> + * Returns 0 for success or -1 for error
> + *
> + * @p: Params for the channel that we are using
> + * @errp: pointer to an error
> + */
> +static int zlib_send_setup(MultiFDSendParams *p, Error **errp)
> +{
> +    uint32_t page_count = MULTIFD_PACKET_SIZE / qemu_target_page_size();
> +    struct zlib_data *z = g_malloc0(sizeof(struct zlib_data));
> +    z_stream *zs = &z->zs;
> +
> +    p->data = z;
> +    zs->zalloc = Z_NULL;
> +    zs->zfree = Z_NULL;
> +    zs->opaque = Z_NULL;
> +    if (deflateInit(zs, migrate_compress_level()) != Z_OK) {
> +        g_free(z);
> +        error_setg(errp, "multifd %d: deflate init failed", p->id);
> +        return -1;
> +    }
> +    /* We will never have more than page_count pages */
> +    z->zbuff_len = page_count * qemu_target_page_size();
> +    z->zbuff_len *= 2;
> +    z->zbuff = g_try_malloc(z->zbuff_len);
> +    if (!z->zbuff) {

Does a deflateEnd need to be called here?

> +        g_free(z);
> +        error_setg(errp, "multifd %d: out of memory for zbuff", p->id);
> +        return -1;
> +    }
> +    return 0;

I'd like to understand more aobut the failure path - lets say we exit
through one of those return -1's, p->data is still set to point to z
which is now been free'd.  Will zlib_send_cleanup get called?
Maybe it's safer to move the 'p->data = z' to right at the bottom before
the return 0 ?

> +}
> +
> +/**
> + * zlib_send_cleanup: cleanup send side
> + *
> + * Close the channel and return memory.
> + *
> + * @p: Params for the channel that we are using
> + */
> +static void zlib_send_cleanup(MultiFDSendParams *p, Error **errp)
> +{
> +    struct zlib_data *z = p->data;

As previously asked above, could this ever get called if zlib_send_setup
has failed? If so does this need to check for !z ?

> +    deflateEnd(&z->zs);
> +    g_free(z->zbuff);
> +    z->zbuff = NULL;
> +    g_free(p->data);
> +    p->data = NULL;
> +}
> +
> +/**
> + * zlib_send_prepare: prepare date to be able to send
> + *
> + * Create a compressed buffer with all the pages that we are going to
> + * send.
> + *
> + * Returns 0 for success or -1 for error
> + *
> + * @p: Params for the channel that we are using
> + * @used: number of pages used
> + */
> +static int zlib_send_prepare(MultiFDSendParams *p, uint32_t used, Error **errp)
> +{
> +    struct iovec *iov = p->pages->iov;
> +    struct zlib_data *z = p->data;
> +    z_stream *zs = &z->zs;
> +    uint32_t out_size = 0;
> +    int ret;
> +    uint32_t i;
> +
> +    for (i = 0; i < used; i++) {
> +        uint32_t available = z->zbuff_len - out_size;
> +        int flush = Z_NO_FLUSH;
> +
> +        if (i == used  - 1) {

Odd double space formatting there.

> +            flush = Z_SYNC_FLUSH;
> +        }
> +
> +        zs->avail_in = iov[i].iov_len;
> +        zs->next_in = iov[i].iov_base;
> +
> +        zs->avail_out = available;
> +        zs->next_out = z->zbuff + out_size;
> +
> +        ret = deflate(zs, flush);
> +        if (ret != Z_OK) {
> +            error_setg(errp, "multifd %d: deflate returned %d instead of Z_OK",
> +                       p->id, ret);
> +            return -1;
> +        }
> +        out_size += available - zs->avail_out;
> +    }
> +    p->next_packet_size = out_size;
> +    p->flags |= MULTIFD_FLAG_ZLIB;
> +
> +    return 0;
> +}
> +
> +/**
> + * zlib_send_write: do the actual write of the data
> + *
> + * Do the actual write of the comprresed buffer.
> + *
> + * Returns 0 for success or -1 for error
> + *
> + * @p: Params for the channel that we are using
> + * @used: number of pages used
> + * @errp: pointer to an error
> + */
> +static int zlib_send_write(MultiFDSendParams *p, uint32_t used, Error **errp)
> +{
> +    struct zlib_data *z = p->data;
> +
> +    return qio_channel_write_all(p->c, (void *)z->zbuff, p->next_packet_size,
> +                                 errp);
> +}
> +
> +/**
> + * zlib_recv_setup: setup receive side
> + *
> + * Create the compressed channel and buffer.
> + *
> + * Returns 0 for success or -1 for error
> + *
> + * @p: Params for the channel that we are using
> + * @errp: pointer to an error
> + */
> +static int zlib_recv_setup(MultiFDRecvParams *p, Error **errp)
> +{
> +    uint32_t page_count = MULTIFD_PACKET_SIZE / qemu_target_page_size();
> +    struct zlib_data *z = g_malloc0(sizeof(struct zlib_data));
> +    z_stream *zs = &z->zs;
> +
> +    p->data = z;
> +    zs->zalloc = Z_NULL;
> +    zs->zfree = Z_NULL;
> +    zs->opaque = Z_NULL;
> +    zs->avail_in = 0;
> +    zs->next_in = Z_NULL;
> +    if (inflateInit(zs) != Z_OK) {
> +        error_setg(errp, "multifd %d: inflate init failed", p->id);
> +        return -1;
> +    }
> +    /* We will never have more than page_count pages */
> +    z->zbuff_len = page_count * qemu_target_page_size();
> +    /* We know compression "could" use more space */
> +    z->zbuff_len *= 2;
> +    z->zbuff = g_try_malloc(z->zbuff_len);
> +    if (!z->zbuff) {

inflateEnd and similar question to save?

> +        error_setg(errp, "multifd %d: out of memory for zbuff", p->id);
> +        return -1;
> +    }
> +    return 0;
> +}
> +
> +/**
> + * zlib_recv_cleanup: setup receive side
> + *
> + * For no compression this function does nothing.
> + *
> + * @p: Params for the channel that we are using
> + */
> +static void zlib_recv_cleanup(MultiFDRecvParams *p)
> +{
> +    struct zlib_data *z = p->data;
> +
> +    inflateEnd(&z->zs);
> +    g_free(z->zbuff);
> +    z->zbuff = NULL;
> +    g_free(p->data);
> +    p->data = NULL;
> +}
> +
> +/**
> + * zlib_recv_pages: read the data from the channel into actual pages
> + *
> + * Read the compressed buffer, and uncompress it into the actual
> + * pages.
> + *
> + * Returns 0 for success or -1 for error
> + *
> + * @p: Params for the channel that we are using
> + * @used: number of pages used
> + * @errp: pointer to an error
> + */
> +static int zlib_recv_pages(MultiFDRecvParams *p, uint32_t used, Error **errp)
> +{
> +    uint32_t in_size = p->next_packet_size;
> +    uint32_t out_size = 0;
> +    uint32_t expected_size = used * qemu_target_page_size();
> +    struct zlib_data *z = p->data;
> +    z_stream *zs = &z->zs;
> +    int ret;
> +    int i;
> +
> +    if (p->flags != MULTIFD_FLAG_ZLIB) {
> +        error_setg(errp, "multifd %d: flags received %x flags expected %x",
> +                   p->id, MULTIFD_FLAG_ZLIB, p->flags);
> +        return -1;
> +    }
> +    ret = qio_channel_read_all(p->c, (void *)z->zbuff, in_size, errp);
> +
> +    if (ret != 0) {
> +        return ret;
> +    }
> +
> +    zs->avail_in = in_size;
> +    zs->next_in = z->zbuff;
> +
> +    for (i = 0; i < used; i++) {
> +        struct iovec *iov = &p->pages->iov[i];
> +        int flush = Z_NO_FLUSH;
> +
> +        if (i == used  - 1) {
> +            flush = Z_SYNC_FLUSH;
> +        }
> +
> +        zs->avail_out = iov->iov_len;
> +        zs->next_out = iov->iov_base;
> +
> +        ret = inflate(zs, flush);
> +        if (ret != Z_OK) {
> +            error_setg(errp, "multifd %d: inflate returned %d instead of Z_OK",
> +                       p->id, ret);
> +            return ret;
> +        }
> +        out_size += iov->iov_len;

How do we know that's iov_len ?

> +    }
> +    if (out_size != expected_size) {
> +        error_setg(errp, "multifd %d: packet size received %d size expected %d",
> +                   p->id, out_size, expected_size);
> +        return -1;
> +    }
> +    return 0;
> +}
> +
> +static MultiFDMethods multifd_zlib_ops = {
> +    .send_setup = zlib_send_setup,
> +    .send_cleanup = zlib_send_cleanup,
> +    .send_prepare = zlib_send_prepare,
> +    .send_write = zlib_send_write,
> +    .recv_setup = zlib_recv_setup,
> +    .recv_cleanup = zlib_recv_cleanup,
> +    .recv_pages = zlib_recv_pages
> +};
> +
> +static void multifd_zlib_register(void)
> +{
> +    multifd_register_ops(MULTIFD_COMPRESS_ZLIB, &multifd_zlib_ops);
> +}
> +
> +migration_init(multifd_zlib_register);
> diff --git a/migration/multifd.c b/migration/multifd.c
> index 353140cd25..a1fc451d49 100644
> --- a/migration/multifd.c
> +++ b/migration/multifd.c
> @@ -164,6 +164,12 @@ static MultiFDMethods *multifd_ops[MULTIFD_COMPRESS__MAX] = {
>      [MULTIFD_COMPRESS_NONE] = &multifd_nocomp_ops,
>  };
>  
> +void multifd_register_ops(int method, MultiFDMethods *ops)
> +{
> +    assert(0 < method && method < MULTIFD_COMPRESS__MAX);
> +    multifd_ops[method] = ops;
> +}
> +
>  static int multifd_send_initial_packet(MultiFDSendParams *p, Error **errp)
>  {
>      MultiFDInit_t msg = {};
> diff --git a/migration/multifd.h b/migration/multifd.h
> index 8edea4fdac..85542f3222 100644
> --- a/migration/multifd.h
> +++ b/migration/multifd.h
> @@ -23,8 +23,10 @@ void multifd_recv_sync_main(void);
>  void multifd_send_sync_main(QEMUFile *f);
>  int multifd_queue_page(QEMUFile *f, RAMBlock *block, ram_addr_t offset);
>  
> +/* Multifd Compression flags */
>  #define MULTIFD_FLAG_SYNC (1 << 0)
>  #define MULTIFD_FLAG_NOCOMP (1 << 1)
> +#define MULTIFD_FLAG_ZLIB (1 << 2)
>  
>  /* This value needs to be a multiple of qemu_target_page_size() */
>  #define MULTIFD_PACKET_SIZE (512 * 1024)
> @@ -157,5 +159,7 @@ typedef struct {
>      int (*recv_pages)(MultiFDRecvParams *p, uint32_t used, Error **errp);
>  } MultiFDMethods;
>  
> +void multifd_register_ops(int method, MultiFDMethods *ops);
> +
>  #endif
>  
> diff --git a/qapi/migration.json b/qapi/migration.json
> index c2891e6ebf..1714ea51e3 100644
> --- a/qapi/migration.json
> +++ b/qapi/migration.json
> @@ -499,7 +499,7 @@
>  #
>  ##
>  { 'enum': 'MultifdCompress',
> -  'data': [ 'none' ] }
> +  'data': [ 'none', 'zlib' ] }
>  
>  ##
>  # @MigrationParameter:
> diff --git a/tests/qtest/migration-test.c b/tests/qtest/migration-test.c
> index 3d5d2aba8c..985a7d4b97 100644
> --- a/tests/qtest/migration-test.c
> +++ b/tests/qtest/migration-test.c
> @@ -1313,6 +1313,11 @@ static void test_multifd_tcp_none(void)
>      test_multifd_tcp("none");
>  }
>  
> +static void test_multifd_tcp_zlib(void)
> +{
> +    test_multifd_tcp("zlib");
> +}
> +
>  /*
>   * This test does:
>   *  source               target
> @@ -1475,6 +1480,7 @@ int main(int argc, char **argv)
>      qtest_add_func("/migration/auto_converge", test_migrate_auto_converge);
>      qtest_add_func("/migration/multifd/tcp/none", test_multifd_tcp_none);
>      qtest_add_func("/migration/multifd/tcp/cancel", test_multifd_tcp_cancel);
> +    qtest_add_func("/migration/multifd/tcp/zlib", test_multifd_tcp_zlib);
>  
>      ret = g_test_run();
>  
> -- 
> 2.24.1
> 
--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK

Re: [PATCH v3 19/21] migration: Add zlib compression multifd support

[Juan Quintela]

"Dr. David Alan Gilbert" <dgilbert@redhat.com> wrote:
> * Juan Quintela (quintela@redhat.com) wrote:
>> Signed-off-by: Juan Quintela <quintela@redhat.com>
>> ---

>> +    /* We will never have more than page_count pages */
>> +    z->zbuff_len = page_count * qemu_target_page_size();
>> +    z->zbuff_len *= 2;
>> +    z->zbuff = g_try_malloc(z->zbuff_len);
>> +    if (!z->zbuff) {
>
> Does a deflateEnd need to be called here?

Shouldnt matter, but I think that yes.

>
>> +        g_free(z);
>> +        error_setg(errp, "multifd %d: out of memory for zbuff", p->id);
>> +        return -1;
>> +    }
>> +    return 0;
>
> I'd like to understand more aobut the failure path - lets say we exit
> through one of those return -1's, p->data is still set to point to z
> which is now been free'd.  Will zlib_send_cleanup get called?
> Maybe it's safer to move the 'p->data = z' to right at the bottom before
> the return 0 ?

Just did that. Good catch.

>
>> +}
>> +
>> +/**
>> + * zlib_send_cleanup: cleanup send side
>> + *
>> + * Close the channel and return memory.
>> + *
>> + * @p: Params for the channel that we are using
>> + */
>> +static void zlib_send_cleanup(MultiFDSendParams *p, Error **errp)
>> +{
>> +    struct zlib_data *z = p->data;
>
> As previously asked above, could this ever get called if zlib_send_setup
> has failed? If so does this need to check for !z ?

No.
if multifd_zlib_setup() returns != 0, then multifd_save_setup() returns
!= 0, and then we just signal an error and then we cancel migration in
multifd_fd_connect.

So, if z is NULL, we are already in big trouble.

>> +
>> +    for (i = 0; i < used; i++) {
>> +        uint32_t available = z->zbuff_len - out_size;
>> +        int flush = Z_NO_FLUSH;
>> +
>> +        if (i == used  - 1) {
>
> Odd double space formatting there.

Ouch.  Changed.
>> +    /* We will never have more than page_count pages */
>> +    z->zbuff_len = page_count * qemu_target_page_size();
>> +    /* We know compression "could" use more space */
>> +    z->zbuff_len *= 2;
>> +    z->zbuff = g_try_malloc(z->zbuff_len);
>> +    if (!z->zbuff) {
>
> inflateEnd and similar question to save?

Done.

>> +    for (i = 0; i < used; i++) {
>> +        struct iovec *iov = &p->pages->iov[i];
>> +        int flush = Z_NO_FLUSH;
>> +
>> +        if (i == used  - 1) {
>> +            flush = Z_SYNC_FLUSH;
>> +        }
>> +
>> +        zs->avail_out = iov->iov_len;
>> +        zs->next_out = iov->iov_base;
>> +
>> +        ret = inflate(zs, flush);
>> +        if (ret != Z_OK) {
>> +            error_setg(errp, "multifd %d: inflate returned %d instead of Z_OK",
>> +                       p->id, ret);
>> +            return ret;
>> +        }
>> +        out_size += iov->iov_len;
>
> How do we know that's iov_len ?

Because we never had fails O:-)
(If it is not, we have a corrupted stream or something has gone very
wrong).

You win.  I think that the correct value is:

uint32_t out_size = zs->total_out;
...

out_size += zs->total_out - out_size;

As a bonus, we can do that outside of the loop.

Thanks, Juan.

Re: [PATCH v3 19/21] migration: Add zlib compression multifd support

[Eric Blake]

On 1/23/20 5:58 AM, Juan Quintela wrote:
> Signed-off-by: Juan Quintela <quintela@redhat.com>
> ---

> +++ b/qapi/migration.json
> @@ -499,7 +499,7 @@
>   #
>   ##
>   { 'enum': 'MultifdCompress',
> -  'data': [ 'none' ] }
> +  'data': [ 'none', 'zlib' ] }
>   

Patch 9 added 'none' and documented it; it looks odd if 'zlib' is not 
documented.

-- 
Eric Blake, Principal Software Engineer
Red Hat, Inc.           +1-919-301-3226
Virtualization:  qemu.org | libvirt.org

[PATCH v3 20/21] configure: Enable test and libs for zstd

[Juan Quintela]

Signed-off-by: Juan Quintela <quintela@redhat.com>
---
 configure | 30 ++++++++++++++++++++++++++++++
 1 file changed, 30 insertions(+)

diff --git a/configure b/configure
index 557e4382ea..56ddfd7f8a 100755
--- a/configure
+++ b/configure
@@ -448,6 +448,7 @@ lzo=""
 snappy=""
 bzip2=""
 lzfse=""
+zstd=""
 guest_agent=""
 guest_agent_with_vss="no"
 guest_agent_ntddscsi="no"
@@ -1343,6 +1344,10 @@ for opt do
   ;;
   --disable-lzfse) lzfse="no"
   ;;
+  --disable-zstd) zstd="no"
+  ;;
+  --enable-zstd) zstd="yes"
+  ;;
   --enable-guest-agent) guest_agent="yes"
   ;;
   --disable-guest-agent) guest_agent="no"
@@ -1795,6 +1800,8 @@ disabled with --disable-FEATURE, default is enabled if available:
                   (for reading bzip2-compressed dmg images)
   lzfse           support of lzfse compression library
                   (for reading lzfse-compressed dmg images)
+  zstd            support for zstd compression library
+                  (for migration compression)
   seccomp         seccomp support
   coroutine-pool  coroutine freelist (better performance)
   glusterfs       GlusterFS backend
@@ -2408,6 +2415,24 @@ EOF
     fi
 fi
 
+##########################################
+# zstd check
+
+if test "$zstd" != "no" ; then
+    if $pkg_config --exist libzstd ; then
+        zstd_cflags="$($pkg_config --cflags libzstd)"
+        zstd_libs="$($pkg_config --libs libzstd)"
+        LIBS="$zstd_libs $LIBS"
+        QEMU_CFLAGS="$QEMU_CFLAGS $zstd_cflags"
+        zstd="yes"
+    else
+        if test "$zstd" = "yes" ; then
+            feature_not_found "libzstd" "Install libzstd devel"
+        fi
+        zstd="no"
+    fi
+fi
+
 ##########################################
 # libseccomp check
 
@@ -6555,6 +6580,7 @@ echo "lzo support       $lzo"
 echo "snappy support    $snappy"
 echo "bzip2 support     $bzip2"
 echo "lzfse support     $lzfse"
+echo "zstd support      $zstd"
 echo "NUMA host support $numa"
 echo "libxml2           $libxml2"
 echo "tcmalloc support  $tcmalloc"
@@ -7114,6 +7140,10 @@ if test "$lzfse" = "yes" ; then
   echo "LZFSE_LIBS=-llzfse" >> $config_host_mak
 fi
 
+if test "$zstd" = "yes" ; then
+  echo "CONFIG_ZSTD=y" >> $config_host_mak
+fi
+
 if test "$libiscsi" = "yes" ; then
   echo "CONFIG_LIBISCSI=m" >> $config_host_mak
   echo "LIBISCSI_CFLAGS=$libiscsi_cflags" >> $config_host_mak
-- 
2.24.1

Re: [PATCH v3 20/21] configure: Enable test and libs for zstd

[Dr. David Alan Gilbert]

* Juan Quintela (quintela@redhat.com) wrote:
> Signed-off-by: Juan Quintela <quintela@redhat.com>

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

> ---
>  configure | 30 ++++++++++++++++++++++++++++++
>  1 file changed, 30 insertions(+)
> 
> diff --git a/configure b/configure
> index 557e4382ea..56ddfd7f8a 100755
> --- a/configure
> +++ b/configure
> @@ -448,6 +448,7 @@ lzo=""
>  snappy=""
>  bzip2=""
>  lzfse=""
> +zstd=""
>  guest_agent=""
>  guest_agent_with_vss="no"
>  guest_agent_ntddscsi="no"
> @@ -1343,6 +1344,10 @@ for opt do
>    ;;
>    --disable-lzfse) lzfse="no"
>    ;;
> +  --disable-zstd) zstd="no"
> +  ;;
> +  --enable-zstd) zstd="yes"
> +  ;;
>    --enable-guest-agent) guest_agent="yes"
>    ;;
>    --disable-guest-agent) guest_agent="no"
> @@ -1795,6 +1800,8 @@ disabled with --disable-FEATURE, default is enabled if available:
>                    (for reading bzip2-compressed dmg images)
>    lzfse           support of lzfse compression library
>                    (for reading lzfse-compressed dmg images)
> +  zstd            support for zstd compression library
> +                  (for migration compression)
>    seccomp         seccomp support
>    coroutine-pool  coroutine freelist (better performance)
>    glusterfs       GlusterFS backend
> @@ -2408,6 +2415,24 @@ EOF
>      fi
>  fi
>  
> +##########################################
> +# zstd check
> +
> +if test "$zstd" != "no" ; then
> +    if $pkg_config --exist libzstd ; then
> +        zstd_cflags="$($pkg_config --cflags libzstd)"
> +        zstd_libs="$($pkg_config --libs libzstd)"
> +        LIBS="$zstd_libs $LIBS"
> +        QEMU_CFLAGS="$QEMU_CFLAGS $zstd_cflags"
> +        zstd="yes"
> +    else
> +        if test "$zstd" = "yes" ; then
> +            feature_not_found "libzstd" "Install libzstd devel"
> +        fi
> +        zstd="no"
> +    fi
> +fi
> +
>  ##########################################
>  # libseccomp check
>  
> @@ -6555,6 +6580,7 @@ echo "lzo support       $lzo"
>  echo "snappy support    $snappy"
>  echo "bzip2 support     $bzip2"
>  echo "lzfse support     $lzfse"
> +echo "zstd support      $zstd"
>  echo "NUMA host support $numa"
>  echo "libxml2           $libxml2"
>  echo "tcmalloc support  $tcmalloc"
> @@ -7114,6 +7140,10 @@ if test "$lzfse" = "yes" ; then
>    echo "LZFSE_LIBS=-llzfse" >> $config_host_mak
>  fi
>  
> +if test "$zstd" = "yes" ; then
> +  echo "CONFIG_ZSTD=y" >> $config_host_mak
> +fi
> +
>  if test "$libiscsi" = "yes" ; then
>    echo "CONFIG_LIBISCSI=m" >> $config_host_mak
>    echo "LIBISCSI_CFLAGS=$libiscsi_cflags" >> $config_host_mak
> -- 
> 2.24.1
> 
--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK

[PATCH v3 21/21] migration: Add zstd compression multifd support

[Juan Quintela]

Signed-off-by: Juan Quintela <quintela@redhat.com>
---
 hw/core/qdev-properties.c    |   2 +-
 migration/Makefile.objs      |   1 +
 migration/migration.c        |   6 +
 migration/multifd-zstd.c     | 304 +++++++++++++++++++++++++++++++++++
 migration/multifd.h          |  20 +++
 migration/ram.c              |   1 -
 qapi/migration.json          |   2 +-
 tests/qtest/migration-test.c |  10 ++
 8 files changed, 343 insertions(+), 3 deletions(-)
 create mode 100644 migration/multifd-zstd.c

diff --git a/hw/core/qdev-properties.c b/hw/core/qdev-properties.c
index 07ec75d8e3..c9c56d3123 100644
--- a/hw/core/qdev-properties.c
+++ b/hw/core/qdev-properties.c
@@ -647,7 +647,7 @@ const PropertyInfo qdev_prop_fdc_drive_type = {
 const PropertyInfo qdev_prop_multifd_compress = {
     .name = "MultifdCompress",
     .description = "multifd_compress values, "
-                   "none/zlib",
+                   "none/zlib/zstd",
     .enum_table = &MultifdCompress_lookup,
     .get = get_enum,
     .set = set_enum,
diff --git a/migration/Makefile.objs b/migration/Makefile.objs
index 0308caa5c5..0fc619e380 100644
--- a/migration/Makefile.objs
+++ b/migration/Makefile.objs
@@ -9,6 +9,7 @@ common-obj-y += qjson.o
 common-obj-y += block-dirty-bitmap.o
 common-obj-y += multifd.o
 common-obj-y += multifd-zlib.o
+common-obj-$(CONFIG_ZSTD) += multifd-zstd.o
 
 common-obj-$(CONFIG_RDMA) += rdma.o
 
diff --git a/migration/migration.c b/migration/migration.c
index d25fdb3e6b..f0d5ade1df 100644
--- a/migration/migration.c
+++ b/migration/migration.c
@@ -1256,6 +1256,12 @@ static bool migrate_params_check(MigrationParameters *params, Error **errp)
                    "is invalid, it must be in the range of 1 to 10000 ms");
        return false;
     }
+    if (params->has_multifd_compress &&
+       params->multifd_compress == MULTIFD_COMPRESS_ZSTD &&
+       !multifd_compress_zstd_is_enabled()) {
+        error_setg(errp, "The multifd compression method zstd is compiled out");
+        return false;
+    }
     return true;
 }
 
diff --git a/migration/multifd-zstd.c b/migration/multifd-zstd.c
new file mode 100644
index 0000000000..d13c93d297
--- /dev/null
+++ b/migration/multifd-zstd.c
@@ -0,0 +1,304 @@
+/*
+ * Multifd zlib compression implementation
+ *
+ * Copyright (c) 2020 Red Hat Inc
+ *
+ * Authors:
+ *  Juan Quintela <quintela@redhat.com>
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ */
+
+#include "qemu/osdep.h"
+#include <zstd.h>
+#include "qemu/rcu.h"
+#include "exec/target_page.h"
+#include "qapi/error.h"
+#include "migration.h"
+#include "trace.h"
+#include "multifd.h"
+
+struct zstd_data {
+    /* stream for compression */
+    ZSTD_CStream *zcs;
+    /* stream for decompression */
+    ZSTD_DStream *zds;
+    /* buffers */
+    ZSTD_inBuffer in;
+    ZSTD_outBuffer out;
+    /* compressed buffer */
+    uint8_t *zbuff;
+    /* size of compressed buffer */
+    uint32_t zbuff_len;
+};
+
+/* Multifd zstd compression */
+
+/**
+ * zstd_send_setup: setup send side
+ *
+ * Setup each channel with zstd compression.
+ *
+ * Returns 0 for success or -1 for error
+ *
+ * @p: Params for the channel that we are using
+ * @errp: pointer to an error
+ */
+static int zstd_send_setup(MultiFDSendParams *p, Error **errp)
+{
+    uint32_t page_count = MULTIFD_PACKET_SIZE / qemu_target_page_size();
+    struct zstd_data *z = g_new0(struct zstd_data, 1);
+    int res;
+
+    p->data = z;
+    z->zcs = ZSTD_createCStream();
+    if (!z->zcs) {
+        g_free(z);
+        error_setg(errp, "multifd %d: zstd createCStream failed", p->id);
+        return -1;
+    }
+
+    res = ZSTD_initCStream(z->zcs, migrate_compress_level());
+    if (ZSTD_isError(res)) {
+        ZSTD_freeCStream(z->zcs);
+        g_free(z);
+        error_setg(errp, "multifd %d: initCStream failed", p->id);
+        return -1;
+    }
+    /* We will never have more than page_count pages */
+    z->zbuff_len = page_count * qemu_target_page_size();
+    z->zbuff_len *= 2;
+    z->zbuff = g_try_malloc(z->zbuff_len);
+    if (!z->zbuff) {
+        ZSTD_freeCStream(z->zcs);
+        g_free(z);
+        error_setg(errp, "multifd %d: out of memory for zbuff", p->id);
+        return -1;
+    }
+    return 0;
+}
+
+/**
+ * zstd_send_cleanup: cleanup send side
+ *
+ * Close the channel and return memory.
+ *
+ * @p: Params for the channel that we are using
+ */
+static void zstd_send_cleanup(MultiFDSendParams *p, Error **errp)
+{
+    struct zstd_data *z = p->data;
+
+    ZSTD_freeCStream(z->zcs);
+    z->zcs = NULL;
+    g_free(z->zbuff);
+    z->zbuff = NULL;
+    g_free(p->data);
+    p->data = NULL;
+}
+
+/**
+ * zstd_send_prepare: prepare date to be able to send
+ *
+ * Create a compressed buffer with all the pages that we are going to
+ * send.
+ *
+ * Returns 0 for success or -1 for error
+ *
+ * @p: Params for the channel that we are using
+ * @used: number of pages used
+ */
+static int zstd_send_prepare(MultiFDSendParams *p, uint32_t used, Error **errp)
+{
+    struct iovec *iov = p->pages->iov;
+    struct zstd_data *z = p->data;
+    int ret;
+    uint32_t i;
+
+    z->out.dst = z->zbuff;
+    z->out.size = z->zbuff_len;
+    z->out.pos = 0;
+
+    for (i = 0; i < used; i++) {
+        ZSTD_EndDirective flush = ZSTD_e_continue;
+
+        if (i == used - 1) {
+            flush = ZSTD_e_flush;
+        }
+        z->in.src = iov[i].iov_base;
+        z->in.size = iov[i].iov_len;
+        z->in.pos = 0;
+
+        ret = ZSTD_compressStream2(z->zcs, &z->out, &z->in, flush);
+        if (ZSTD_isError(ret)) {
+            error_setg(errp, "multifd %d: compressStream error %s",
+                       p->id, ZSTD_getErrorName(ret));
+            return -1;
+        }
+    }
+    p->next_packet_size = z->out.pos;
+    p->flags |= MULTIFD_FLAG_ZSTD;
+
+    return 0;
+}
+
+/**
+ * zstd_send_write: do the actual write of the data
+ *
+ * Do the actual write of the comprresed buffer.
+ *
+ * Returns 0 for success or -1 for error
+ *
+ * @p: Params for the channel that we are using
+ * @used: number of pages used
+ * @errp: pointer to an error
+ */
+static int zstd_send_write(MultiFDSendParams *p, uint32_t used, Error **errp)
+{
+    struct zstd_data *z = p->data;
+
+    return qio_channel_write_all(p->c, (void *)z->zbuff, p->next_packet_size,
+                                 errp);
+}
+
+/**
+ * zstd_recv_setup: setup receive side
+ *
+ * Create the compressed channel and buffer.
+ *
+ * Returns 0 for success or -1 for error
+ *
+ * @p: Params for the channel that we are using
+ * @errp: pointer to an error
+ */
+static int zstd_recv_setup(MultiFDRecvParams *p, Error **errp)
+{
+    uint32_t page_count = MULTIFD_PACKET_SIZE / qemu_target_page_size();
+    struct zstd_data *z = g_new0(struct zstd_data, 1);
+    int res;
+
+    p->data = z;
+    z->zds = ZSTD_createDStream();
+    if (!z->zds) {
+        g_free(z);
+        error_setg(errp, "multifd %d: zstd createDStream failed", p->id);
+        return -1;
+    }
+
+    res = ZSTD_initDStream(z->zds);
+    if (ZSTD_isError(res)) {
+        ZSTD_freeDStream(z->zds);
+        g_free(z);
+        error_setg(errp, "multifd %d: initDStream failed", p->id);
+        return -1;
+    }
+
+    /* We will never have more than page_count pages */
+    z->zbuff_len = page_count * qemu_target_page_size();
+    /* We know compression "could" use more space */
+    z->zbuff_len *= 2;
+    z->zbuff = g_try_malloc(z->zbuff_len);
+    if (!z->zbuff) {
+        ZSTD_freeDStream(z->zds);
+        g_free(z);
+        error_setg(errp, "multifd %d: out of memory for zbuff", p->id);
+        return -1;
+    }
+    return 0;
+}
+
+/**
+ * zstd_recv_cleanup: setup receive side
+ *
+ * For no compression this function does nothing.
+ *
+ * @p: Params for the channel that we are using
+ */
+static void zstd_recv_cleanup(MultiFDRecvParams *p)
+{
+    struct zstd_data *z = p->data;
+
+    ZSTD_freeDStream(z->zds);
+    z->zds = NULL;
+    g_free(z->zbuff);
+    z->zbuff = NULL;
+    g_free(p->data);
+    p->data = NULL;
+}
+
+/**
+ * zstd_recv_pages: read the data from the channel into actual pages
+ *
+ * Read the compressed buffer, and uncompress it into the actual
+ * pages.
+ *
+ * Returns 0 for success or -1 for error
+ *
+ * @p: Params for the channel that we are using
+ * @used: number of pages used
+ * @errp: pointer to an error
+ */
+static int zstd_recv_pages(MultiFDRecvParams *p, uint32_t used, Error **errp)
+{
+    uint32_t in_size = p->next_packet_size;
+    uint32_t out_size = 0;
+    uint32_t expected_size = used * qemu_target_page_size();
+    struct zstd_data *z = p->data;
+    int ret;
+    int i;
+
+    if (p->flags != MULTIFD_FLAG_ZSTD) {
+        error_setg(errp, "multifd %d: flags received %x flags expected %x",
+                   p->id, p->flags, MULTIFD_FLAG_ZSTD);
+        return -1;
+    }
+    ret = qio_channel_read_all(p->c, (void *)z->zbuff, in_size, errp);
+
+    if (ret != 0) {
+        return ret;
+    }
+
+    z->in.src = z->zbuff;
+    z->in.size = in_size;
+    z->in.pos = 0;
+
+    for (i = 0; i < used; i++) {
+        struct iovec *iov = &p->pages->iov[i];
+
+        z->out.dst = iov->iov_base;
+        z->out.size = iov->iov_len;
+        z->out.pos = 0;
+
+        ret = ZSTD_decompressStream(z->zds, &z->out, &z->in);
+        if (ZSTD_isError(ret)) {
+            error_setg(errp, "multifd %d: decompressStream returned %s",
+                       p->id, ZSTD_getErrorName(ret));
+            return ret;
+        }
+        out_size += iov->iov_len;
+    }
+    if (out_size != expected_size) {
+        error_setg(errp, "multifd %d: packet size received %d size expected %d",
+                   p->id, out_size, expected_size);
+        return -1;
+    }
+    return 0;
+}
+
+static MultiFDMethods multifd_zstd_ops = {
+    .send_setup = zstd_send_setup,
+    .send_cleanup = zstd_send_cleanup,
+    .send_prepare = zstd_send_prepare,
+    .send_write = zstd_send_write,
+    .recv_setup = zstd_recv_setup,
+    .recv_cleanup = zstd_recv_cleanup,
+    .recv_pages = zstd_recv_pages
+};
+
+static void multifd_zstd_register(void)
+{
+    multifd_register_ops(MULTIFD_COMPRESS_ZSTD, &multifd_zstd_ops);
+}
+
+migration_init(multifd_zstd_register);
diff --git a/migration/multifd.h b/migration/multifd.h
index 85542f3222..7ec55a4534 100644
--- a/migration/multifd.h
+++ b/migration/multifd.h
@@ -27,6 +27,7 @@ int multifd_queue_page(QEMUFile *f, RAMBlock *block, ram_addr_t offset);
 #define MULTIFD_FLAG_SYNC (1 << 0)
 #define MULTIFD_FLAG_NOCOMP (1 << 1)
 #define MULTIFD_FLAG_ZLIB (1 << 2)
+#define MULTIFD_FLAG_ZSTD (1 << 3)
 
 /* This value needs to be a multiple of qemu_target_page_size() */
 #define MULTIFD_PACKET_SIZE (512 * 1024)
@@ -161,5 +162,24 @@ typedef struct {
 
 void multifd_register_ops(int method, MultiFDMethods *ops);
 
+
+/*
+ * This is gross, but we don't want to have ifdefs in migration.h
+ * And we need to know if ZSTD is compiled in to be able to know
+ * if we can setup multifd_compress with that parameter.
+ *
+ * As far as I can see there is no way to convince qapi that the value
+ * of multifd_compress is none/zlib if zstd is compiled out, or
+ * none/zlib/zstd when zstd is compiled in.
+ */
+
+static inline bool multifd_compress_zstd_is_enabled(void)
+{
+#ifdef CONFIG_ZSTD
+    return true;
+#else
+    return false;
+#endif
+}
 #endif
 
diff --git a/migration/ram.c b/migration/ram.c
index 73a141bb60..0ef68798d2 100644
--- a/migration/ram.c
+++ b/migration/ram.c
@@ -28,7 +28,6 @@
 
 #include "qemu/osdep.h"
 #include "cpu.h"
-#include <zlib.h>
 #include "qemu/cutils.h"
 #include "qemu/bitops.h"
 #include "qemu/bitmap.h"
diff --git a/qapi/migration.json b/qapi/migration.json
index 1714ea51e3..65db85970e 100644
--- a/qapi/migration.json
+++ b/qapi/migration.json
@@ -499,7 +499,7 @@
 #
 ##
 { 'enum': 'MultifdCompress',
-  'data': [ 'none', 'zlib' ] }
+  'data': [ 'none', 'zlib', 'zstd' ] }
 
 ##
 # @MigrationParameter:
diff --git a/tests/qtest/migration-test.c b/tests/qtest/migration-test.c
index 985a7d4b97..99eee5d891 100644
--- a/tests/qtest/migration-test.c
+++ b/tests/qtest/migration-test.c
@@ -1318,6 +1318,13 @@ static void test_multifd_tcp_zlib(void)
     test_multifd_tcp("zlib");
 }
 
+#ifdef CONFIG_ZSTD
+static void test_multifd_tcp_zstd(void)
+{
+    test_multifd_tcp("zstd");
+}
+#endif
+
 /*
  * This test does:
  *  source               target
@@ -1481,6 +1488,9 @@ int main(int argc, char **argv)
     qtest_add_func("/migration/multifd/tcp/none", test_multifd_tcp_none);
     qtest_add_func("/migration/multifd/tcp/cancel", test_multifd_tcp_cancel);
     qtest_add_func("/migration/multifd/tcp/zlib", test_multifd_tcp_zlib);
+#ifdef CONFIG_ZSTD
+    qtest_add_func("/migration/multifd/tcp/zstd", test_multifd_tcp_zstd);
+#endif
 
     ret = g_test_run();
 
-- 
2.24.1

Re: [PATCH v3 21/21] migration: Add zstd compression multifd support

[Eric Blake]

On 1/23/20 5:58 AM, Juan Quintela wrote:
> Signed-off-by: Juan Quintela <quintela@redhat.com>
> ---
>   hw/core/qdev-properties.c    |   2 +-
>   migration/Makefile.objs      |   1 +
>   migration/migration.c        |   6 +
>   migration/multifd-zstd.c     | 304 +++++++++++++++++++++++++++++++++++
>   migration/multifd.h          |  20 +++
>   migration/ram.c              |   1 -
>   qapi/migration.json          |   2 +-
>   tests/qtest/migration-test.c |  10 ++
>   8 files changed, 343 insertions(+), 3 deletions(-)
>   create mode 100644 migration/multifd-zstd.c
> 

> +++ b/migration/Makefile.objs
> @@ -9,6 +9,7 @@ common-obj-y += qjson.o
>   common-obj-y += block-dirty-bitmap.o
>   common-obj-y += multifd.o
>   common-obj-y += multifd-zlib.o
> +common-obj-$(CONFIG_ZSTD) += multifd-zstd.o

zstd support is conditional...


> +++ b/qapi/migration.json
> @@ -499,7 +499,7 @@
>   #
>   ##
>   { 'enum': 'MultifdCompress',
> -  'data': [ 'none', 'zlib' ] }
> +  'data': [ 'none', 'zlib', 'zstd' ] }

...so I would expect an 'if' conditional here when declaring the enum. 
It also needs documentation.


-- 
Eric Blake, Principal Software Engineer
Red Hat, Inc.           +1-919-301-3226
Virtualization:  qemu.org | libvirt.org

Re: [PATCH v3 21/21] migration: Add zstd compression multifd support

[Juan Quintela]

Eric Blake <eblake@redhat.com> wrote:
> On 1/23/20 5:58 AM, Juan Quintela wrote:
>> Signed-off-by: Juan Quintela <quintela@redhat.com>
>> ---
>>   hw/core/qdev-properties.c    |   2 +-
>>   migration/Makefile.objs      |   1 +
>>   migration/migration.c        |   6 +
>>   migration/multifd-zstd.c     | 304 +++++++++++++++++++++++++++++++++++
>>   migration/multifd.h          |  20 +++
>>   migration/ram.c              |   1 -
>>   qapi/migration.json          |   2 +-
>>   tests/qtest/migration-test.c |  10 ++
>>   8 files changed, 343 insertions(+), 3 deletions(-)
>>   create mode 100644 migration/multifd-zstd.c
>>
>
>> +++ b/migration/Makefile.objs
>> @@ -9,6 +9,7 @@ common-obj-y += qjson.o
>>   common-obj-y += block-dirty-bitmap.o
>>   common-obj-y += multifd.o
>>   common-obj-y += multifd-zlib.o
>> +common-obj-$(CONFIG_ZSTD) += multifd-zstd.o

Hi

> zstd support is conditional...

yeap.

>> +++ b/qapi/migration.json
>> @@ -499,7 +499,7 @@
>>   #
>>   ##
>>   { 'enum': 'MultifdCompress',
>> -  'data': [ 'none', 'zlib' ] }
>> +  'data': [ 'none', 'zlib', 'zstd' ] }
>
> ...so I would expect an 'if' conditional here when declaring the
> enum. It also needs documentation.

I fully agree about the documentation part.

But if you are setting zstd, and zstd is compiled out, we are giving the
error showed in the code shown below.

I am open to do anyother way.  I don't speak qapi natively O:-)

Later, Juan.



diff --git a/migration/migration.c b/migration/migration.c
index d25fdb3e6b..f0d5ade1df 100644
--- a/migration/migration.c
+++ b/migration/migration.c
@@ -1256,6 +1256,12 @@ static bool migrate_params_check(MigrationParameters *params, Error **errp)
                    "is invalid, it must be in the range of 1 to 10000 ms");
        return false;
     }
+    if (params->has_multifd_compress &&
+       params->multifd_compress == MULTIFD_COMPRESS_ZSTD &&
+       !multifd_compress_zstd_is_enabled()) {
+        error_setg(errp, "The multifd compression method zstd is compiled out");
+        return false;
+    }
     return true;
 }
 

Re: [PATCH v3 00/21] Multifd Migration Compression

[Juan Quintela]

Juan Quintela <quintela@redhat.com> wrote:
> [v3]
> - rebased on top of upstream + previous multifd cancel series
> - split multifd code into its own file (multifd.[ch])
> - split zstd/zlib compression methods (multifd-zstd/zlib.c)
> - use qemu module feauture to avoid ifdefs
>   (my understanding is that zlib needs to be present, but
>   we setup zstd only if it is not there or is disabled)
> - multifd-method: none|zlib|zstd
>
>   As far as I can see, there is no easy way to convince qapi that zstd
>   option could/couldn't be there depending on compliation flags. I
>   ended just checking in migrate_parameters_check() if it is enabled
>   and giving an error message otherwise.
>
> Questions:
> - I am "reusing" the compress-level parameter for both zstd and zlib,
>   but it poses a problem:
>   * zlib values: 1-9 (default: 6?)
>   * zstd values: 1-19 (default: 3)
> So, what should I do:
>   * create multifd-zstd-level and multifd-zlib-level (easier)
>   * reuse compress-level, and change its maximum values depending on
>     multifd-method
>   * any other good option?
>
> Please, review.
>
> [v2] - rebase on top of previous arguments posted to the list -
> introduces zlib compression - introduces zstd compression
>
> Please help if you know anything about zstd/zlib compression.
>
> This puts compression on top of multifd. Advantages about current
> compression:
>
> - We copy all pages in a single packet and then compress the whole
>   thing.
>
> - We reuse the compression stream for all the packets sent through the
>   same channel.
>
> - We can select nocomp/zlib/zstd levels of compression.
>
> Please, review.
>
> Juan Quintela (21):
>   migration-test: Use g_free() instead of free()
>   multifd: Make sure that we don't do any IO after an error
>   qemu-file: Don't do IO after shutdown
>   migration-test: Make sure that multifd and cancel works
>   migration: Create migration_is_running()
>   migration: Don't send data if we have stopped

This patches are from my previous series.  I forgot to pass the -b
option to git-publish.  Please ignore them.

Sorry, Juan.

Re: [PATCH v3 00/21] Multifd Migration Compression

[Markus Armbruster]

Juan Quintela <quintela@redhat.com> writes:

> [v3]
> - rebased on top of upstream + previous multifd cancel series
> - split multifd code into its own file (multifd.[ch])
> - split zstd/zlib compression methods (multifd-zstd/zlib.c)
> - use qemu module feauture to avoid ifdefs
>   (my understanding is that zlib needs to be present, but
>   we setup zstd only if it is not there or is disabled)
> - multifd-method: none|zlib|zstd
>
>   As far as I can see, there is no easy way to convince qapi that zstd
>   option could/couldn't be there depending on compliation flags. I
>   ended just checking in migrate_parameters_check() if it is enabled
>   and giving an error message otherwise.

Wild guess: this is about PATCH 21's

   diff --git a/qapi/migration.json b/qapi/migration.json
   index 1714ea51e3..65db85970e 100644
   --- a/qapi/migration.json
   +++ b/qapi/migration.json
   @@ -499,7 +499,7 @@
    #
    ##
    { 'enum': 'MultifdCompress',
   -  'data': [ 'none', 'zlib' ] }
   +  'data': [ 'none', 'zlib', 'zstd' ] }

    ##
    # @MigrationParameter:

where you want 'zstd' to be #ifdef CONFIG_ZSTD.  If not, please advise.
Else use something like

    { 'enum': 'MultifdCompress',
      'data': [ 'none', 'zlib',
                { 'name': 'zstd', 'if': 'defined(CONFIG_ZSTD)' } ] }

See docs/devel/qapi-code-gen.txt sections "Enumeration types" and
"Configuring the schema".

[...]