[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[PATCH 07/11] multifd: Prepare to send a packet without the mutex held
From: |
Juan Quintela |
Subject: |
[PATCH 07/11] multifd: Prepare to send a packet without the mutex held |
Date: |
Mon, 28 Nov 2022 11:04:18 +0100 |
We do the send_prepare() and the fill of the head packet without the
mutex held. It will help a lot for compression and later in the
series for zero pages.
Notice that we can use p->pages without holding p->mutex because
p->pending_job == 1.
Signed-off-by: Juan Quintela <quintela@redhat.com>
---
migration/multifd.h | 2 ++
migration/multifd.c | 12 ++++++------
2 files changed, 8 insertions(+), 6 deletions(-)
diff --git a/migration/multifd.h b/migration/multifd.h
index a67cefc0a2..cd389d18d2 100644
--- a/migration/multifd.h
+++ b/migration/multifd.h
@@ -109,7 +109,9 @@ typedef struct {
/* array of pages to sent.
* The owner of 'pages' depends of 'pending_job' value:
* pending_job == 0 -> migration_thread can use it.
+ * No need for mutex lock.
* pending_job != 0 -> multifd_channel can use it.
+ * No need for mutex lock.
*/
MultiFDPages_t *pages;
diff --git a/migration/multifd.c b/migration/multifd.c
index 77196a55b4..7ebaca6e55 100644
--- a/migration/multifd.c
+++ b/migration/multifd.c
@@ -672,6 +672,8 @@ static void *multifd_send_thread(void *opaque)
p->flags |= MULTIFD_FLAG_SYNC;
p->sync_needed = false;
}
+ qemu_mutex_unlock(&p->mutex);
+
p->normal_num = 0;
if (use_zero_copy_send) {
@@ -688,16 +690,10 @@ static void *multifd_send_thread(void *opaque)
if (p->normal_num) {
ret = multifd_send_state->ops->send_prepare(p, &local_err);
if (ret != 0) {
- qemu_mutex_unlock(&p->mutex);
break;
}
}
multifd_send_fill_packet(p);
- p->num_packets++;
- p->total_normal_pages += p->normal_num;
- p->pages->num = 0;
- p->pages->block = NULL;
- qemu_mutex_unlock(&p->mutex);
trace_multifd_send(p->id, packet_num, p->normal_num, p->flags,
p->next_packet_size);
@@ -722,6 +718,10 @@ static void *multifd_send_thread(void *opaque)
}
qemu_mutex_lock(&p->mutex);
+ p->num_packets++;
+ p->total_normal_pages += p->normal_num;
+ p->pages->num = 0;
+ p->pages->block = NULL;
p->sent_bytes += p->packet_len;
p->sent_bytes += p->next_packet_size;
p->pending_job--;
--
2.38.1
- [PATCH 00/11] Multifd zero page support, Juan Quintela, 2022/11/28
- [PATCH 01/11] migration: Update atomic stats out of the mutex, Juan Quintela, 2022/11/28
- [PATCH 03/11] multifd: We already account for this packet on the multifd thread, Juan Quintela, 2022/11/28
- [PATCH 02/11] migration: Make multifd_bytes atomic, Juan Quintela, 2022/11/28
- [PATCH 05/11] migration: Make ram_save_target_page() a pointer, Juan Quintela, 2022/11/28
- [PATCH 04/11] multifd: Count the number of bytes sent correctly, Juan Quintela, 2022/11/28
- [PATCH 07/11] multifd: Prepare to send a packet without the mutex held,
Juan Quintela <=
- [PATCH 06/11] multifd: Make flags field thread local, Juan Quintela, 2022/11/28
- [PATCH 08/11] multifd: Add capability to enable/disable zero_page, Juan Quintela, 2022/11/28
- [PATCH 09/11] multifd: Support for zero pages transmission, Juan Quintela, 2022/11/28
- [PATCH 10/11] multifd: Zero pages transmission, Juan Quintela, 2022/11/28
- [PATCH 11/11] So we use multifd to transmit zero pages., Juan Quintela, 2022/11/28