qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Juan Quintela <quintela@redhat.com>
To: qemu-devel@nongnu.org
Cc: qemu-block@nongnu.org,
	"Marcel Apfelbaum" <marcel.apfelbaum@gmail.com>,
	"Philippe Mathieu-Daudé" <philmd@linaro.org>,
	"Eric Blake" <eblake@redhat.com>,
	"Juan Quintela" <quintela@redhat.com>,
	"Peter Xu" <peterx@redhat.com>, "John Snow" <jsnow@redhat.com>,
	"Yanan Wang" <wangyanan55@huawei.com>,
	"Stefan Hajnoczi" <stefanha@redhat.com>,
	"Leonardo Bras" <leobras@redhat.com>,
	"Fam Zheng" <fam@euphon.net>,
	"Eduardo Habkost" <eduardo@habkost.net>,
	"Vladimir Sementsov-Ogievskiy" <vsementsov@yandex-team.ru>,
	"Dr . David Alan Gilbert" <dgilbert@redhat.com>
Subject: [PULL 13/18] multifd: Create property multifd-flush-after-each-section
Date: Thu, 27 Apr 2023 17:22:29 +0200	[thread overview]
Message-ID: <20230427152234.25400-14-quintela@redhat.com> (raw)
In-Reply-To: <20230427152234.25400-1-quintela@redhat.com>

We used to flush all channels at the end of each RAM section
sent.  That is not needed, so preparing to only flush after a full
iteration through all the RAM.

Default value of the property is false.  But we return "true" in
migrate_multifd_flush_after_each_section() until we implement the code
in following patches.

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

---

Rename each-iteration to after-each-section
Rename multifd-sync-after-each-section to
       multifd-flush-after-each-section
Move to machine-8.0 (peter)
---
 hw/core/machine.c     |  4 +++-
 migration/migration.h | 12 ++++++++++++
 migration/options.c   | 13 +++++++++++++
 migration/options.h   |  1 +
 4 files changed, 29 insertions(+), 1 deletion(-)

diff --git a/hw/core/machine.c b/hw/core/machine.c
index 2ce97a5d3b..47a34841a5 100644
--- a/hw/core/machine.c
+++ b/hw/core/machine.c
@@ -39,7 +39,9 @@
 #include "hw/virtio/virtio.h"
 #include "hw/virtio/virtio-pci.h"
 
-GlobalProperty hw_compat_8_0[] = {};
+GlobalProperty hw_compat_8_0[] = {
+    { "migration", "multifd-flush-after-each-section", "on"},
+};
 const size_t hw_compat_8_0_len = G_N_ELEMENTS(hw_compat_8_0);
 
 GlobalProperty hw_compat_7_2[] = {
diff --git a/migration/migration.h b/migration/migration.h
index 2b71df8617..e2247d708f 100644
--- a/migration/migration.h
+++ b/migration/migration.h
@@ -404,6 +404,18 @@ struct MigrationState {
      */
     bool preempt_pre_7_2;
 
+    /*
+     * flush every channel after each section sent.
+     *
+     * This assures that we can't mix pages from one iteration through
+     * ram pages with pages for the following iteration.  We really
+     * only need to do this flush after we have go through all the
+     * dirty pages.  For historical reasons, we do that after each
+     * section.  This is suboptimal (we flush too many times).
+     * Default value is false.  Setting this property has no effect
+     * until the patch that removes this comment.  (since 8.1)
+     */
+    bool multifd_flush_after_each_section;
     /*
      * This decides the size of guest memory chunk that will be used
      * to track dirty bitmap clearing.  The size of memory chunk will
diff --git a/migration/options.c b/migration/options.c
index f2a77cdef1..690c8f718e 100644
--- a/migration/options.c
+++ b/migration/options.c
@@ -88,6 +88,8 @@ Property migration_properties[] = {
                      send_section_footer, true),
     DEFINE_PROP_BOOL("decompress-error-check", MigrationState,
                       decompress_error_check, true),
+    DEFINE_PROP_BOOL("multifd-flush-after-each-section", MigrationState,
+                      multifd_flush_after_each_section, true),
     DEFINE_PROP_UINT8("x-clear-bitmap-shift", MigrationState,
                       clear_bitmap_shift, CLEAR_BITMAP_SHIFT_DEFAULT),
     DEFINE_PROP_BOOL("x-preempt-pre-7-2", MigrationState,
@@ -335,6 +337,17 @@ bool migrate_zero_copy_send(void)
 
 /* pseudo capabilities */
 
+bool migrate_multifd_flush_after_each_section(void)
+{
+    MigrationState *s = migrate_get_current();
+
+    /*
+     * Until the patch that remove this comment, we always return that
+     * the property is enabled.
+     */
+    return true || s->multifd_flush_after_each_section;
+}
+
 bool migrate_postcopy(void)
 {
     return migrate_postcopy_ram() || migrate_dirty_bitmaps();
diff --git a/migration/options.h b/migration/options.h
index 00854d60f9..3c322867cd 100644
--- a/migration/options.h
+++ b/migration/options.h
@@ -59,6 +59,7 @@ bool migrate_zero_copy_send(void);
  * check, but they are not a capability.
  */
 
+bool migrate_multifd_flush_after_each_section(void);
 bool migrate_postcopy(void);
 bool migrate_tls(void);
 
-- 
2.40.0



  parent reply	other threads:[~2023-04-27 15:26 UTC|newest]

Thread overview: 28+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-04-27 15:22 [PULL 00/18] Migration 20230427 patches Juan Quintela
2023-04-27 15:22 ` [PULL 01/18] multifd: Fix the number of channels ready Juan Quintela
2023-04-27 15:22 ` [PULL 02/18] migration: Create migrate_params_init() function Juan Quintela
2023-04-27 15:22 ` [PULL 03/18] migration: Make all functions check have the same format Juan Quintela
2023-04-27 15:22 ` [PULL 04/18] migration: Create migrate_downtime_limit() function Juan Quintela
2023-04-27 15:22 ` [PULL 05/18] migration: Move migrate_set_block_incremental() to options.c Juan Quintela
2023-04-27 15:22 ` [PULL 06/18] migration: Move block_cleanup_parameters() " Juan Quintela
2023-04-27 15:22 ` [PULL 07/18] migration: Remove MigrationState from block_cleanup_parameters() Juan Quintela
2023-04-27 15:22 ` [PULL 08/18] migration: Create migrate_tls_creds() function Juan Quintela
2023-04-27 15:22 ` [PULL 09/18] migration: Create migrate_tls_authz() function Juan Quintela
2023-04-27 15:22 ` [PULL 10/18] migration: Create migrate_tls_hostname() function Juan Quintela
2023-04-27 15:22 ` [PULL 11/18] migration: Create migrate_block_bitmap_mapping() function Juan Quintela
2023-05-03 14:33   ` Kevin Wolf
2023-05-03 14:53     ` Juan Quintela
2023-05-03 17:10       ` Kevin Wolf
2023-05-03 17:17         ` Juan Quintela
2023-05-03 17:15     ` Juan Quintela
2023-05-03 19:00       ` Vladimir Sementsov-Ogievskiy
2023-05-04  8:38       ` Kevin Wolf
2023-05-04 10:02         ` Juan Quintela
2023-04-27 15:22 ` [PULL 12/18] migration: Move migration_properties to options.c Juan Quintela
2023-04-27 15:22 ` Juan Quintela [this message]
2023-04-27 15:22 ` [PULL 14/18] multifd: Protect multifd_send_sync_main() calls Juan Quintela
2023-04-27 15:22 ` [PULL 15/18] multifd: Only flush once each full round of memory Juan Quintela
2023-04-27 15:22 ` [PULL 16/18] stat64: Add stat64_set() operation Juan Quintela
2023-04-27 15:22 ` [PULL 17/18] migration: Make dirty_pages_rate atomic Juan Quintela
2023-04-27 15:22 ` [PULL 18/18] migration: Make dirty_bytes_last_sync atomic Juan Quintela
2023-04-28  9:55 ` [PULL 00/18] Migration 20230427 patches Richard Henderson

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20230427152234.25400-14-quintela@redhat.com \
    --to=quintela@redhat.com \
    --cc=dgilbert@redhat.com \
    --cc=eblake@redhat.com \
    --cc=eduardo@habkost.net \
    --cc=fam@euphon.net \
    --cc=jsnow@redhat.com \
    --cc=leobras@redhat.com \
    --cc=marcel.apfelbaum@gmail.com \
    --cc=peterx@redhat.com \
    --cc=philmd@linaro.org \
    --cc=qemu-block@nongnu.org \
    --cc=qemu-devel@nongnu.org \
    --cc=stefanha@redhat.com \
    --cc=vsementsov@yandex-team.ru \
    --cc=wangyanan55@huawei.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).