From: Kevin Wolf <kwolf@redhat.com>
To: qemu-block@nongnu.org
Cc: kwolf@redhat.com, qemu-devel@nongnu.org
Subject: [PULL 12/25] migration/block-active: Remove global active flag
Date: Mon, 10 Feb 2025 17:10:21 +0100 [thread overview]
Message-ID: <20250210161034.76494-13-kwolf@redhat.com> (raw)
In-Reply-To: <20250210161034.76494-1-kwolf@redhat.com>
Block devices have an individual active state, a single global flag
can't cover this correctly. This becomes more important as we allow
users to manually manage which nodes are active or inactive.
Now that it's allowed to call bdrv_inactivate_all() even when some
nodes are already inactive, we can remove the flag and just
unconditionally call bdrv_inactivate_all() and, more importantly,
bdrv_activate_all() before we make use of the nodes.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Acked-by: Fabiano Rosas <farosas@suse.de>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-ID: <20250204211407.381505-5-kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
migration/migration.h | 3 ---
migration/block-active.c | 46 ----------------------------------------
migration/migration.c | 8 -------
3 files changed, 57 deletions(-)
diff --git a/migration/migration.h b/migration/migration.h
index 4c1fafc2b5..8b24cbedd1 100644
--- a/migration/migration.h
+++ b/migration/migration.h
@@ -554,7 +554,4 @@ void migration_bitmap_sync_precopy(bool last_stage);
void dirty_bitmap_mig_init(void);
bool should_send_vmdesc(void);
-/* migration/block-active.c */
-void migration_block_active_setup(bool active);
-
#endif
diff --git a/migration/block-active.c b/migration/block-active.c
index d477cf8182..40e986aade 100644
--- a/migration/block-active.c
+++ b/migration/block-active.c
@@ -12,51 +12,12 @@
#include "qemu/error-report.h"
#include "trace.h"
-/*
- * Migration-only cache to remember the block layer activation status.
- * Protected by BQL.
- *
- * We need this because..
- *
- * - Migration can fail after block devices are invalidated (during
- * switchover phase). When that happens, we need to be able to recover
- * the block drive status by re-activating them.
- *
- * - Currently bdrv_inactivate_all() is not safe to be invoked on top of
- * invalidated drives (even if bdrv_activate_all() is actually safe to be
- * called any time!). It means remembering this could help migration to
- * make sure it won't invalidate twice in a row, crashing QEMU. It can
- * happen when we migrate a PAUSED VM from host1 to host2, then migrate
- * again to host3 without starting it. TODO: a cleaner solution is to
- * allow safe invoke of bdrv_inactivate_all() at anytime, like
- * bdrv_activate_all().
- *
- * For freshly started QEMU, the flag is initialized to TRUE reflecting the
- * scenario where QEMU owns block device ownerships.
- *
- * For incoming QEMU taking a migration stream, the flag is initialized to
- * FALSE reflecting that the incoming side doesn't own the block devices,
- * not until switchover happens.
- */
-static bool migration_block_active;
-
-/* Setup the disk activation status */
-void migration_block_active_setup(bool active)
-{
- migration_block_active = active;
-}
-
bool migration_block_activate(Error **errp)
{
ERRP_GUARD();
assert(bql_locked());
- if (migration_block_active) {
- trace_migration_block_activation("active-skipped");
- return true;
- }
-
trace_migration_block_activation("active");
bdrv_activate_all(errp);
@@ -65,7 +26,6 @@ bool migration_block_activate(Error **errp)
return false;
}
- migration_block_active = true;
return true;
}
@@ -75,11 +35,6 @@ bool migration_block_inactivate(void)
assert(bql_locked());
- if (!migration_block_active) {
- trace_migration_block_activation("inactive-skipped");
- return true;
- }
-
trace_migration_block_activation("inactive");
ret = bdrv_inactivate_all();
@@ -89,6 +44,5 @@ bool migration_block_inactivate(void)
return false;
}
- migration_block_active = false;
return true;
}
diff --git a/migration/migration.c b/migration/migration.c
index 74c50cc72c..95e05bbc3b 100644
--- a/migration/migration.c
+++ b/migration/migration.c
@@ -1895,12 +1895,6 @@ void qmp_migrate_incoming(const char *uri, bool has_channels,
return;
}
- /*
- * Newly setup incoming QEMU. Mark the block active state to reflect
- * that the src currently owns the disks.
- */
- migration_block_active_setup(false);
-
once = false;
}
@@ -3992,8 +3986,6 @@ static void migration_instance_init(Object *obj)
ms->state = MIGRATION_STATUS_NONE;
ms->mbps = -1;
ms->pages_per_second = -1;
- /* Freshly started QEMU owns all the block devices */
- migration_block_active_setup(true);
qemu_sem_init(&ms->pause_sem, 0);
qemu_mutex_init(&ms->error_mutex);
--
2.48.1
next prev parent reply other threads:[~2025-02-10 16:14 UTC|newest]
Thread overview: 27+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-02-10 16:10 [PULL 00/25] Block layer patches Kevin Wolf
2025-02-10 16:10 ` [PULL 01/25] vpc: Split off vpc_ignore_current_size() helper Kevin Wolf
2025-02-10 16:10 ` [PULL 02/25] vpc: Read images exported from Azure correctly Kevin Wolf
2025-02-10 16:10 ` [PULL 03/25] block: Improve blk_get_attached_dev_id() docstring Kevin Wolf
2025-02-10 16:10 ` [PULL 04/25] block: Fix leak in send_qmp_error_event Kevin Wolf
2025-02-10 16:10 ` [PULL 05/25] scripts/qemu-gdb: Always do full stack dump for python errors Kevin Wolf
2025-02-10 16:10 ` [PULL 06/25] scripts/qemu-gdb: Simplify fs_base fetching for coroutines Kevin Wolf
2025-02-10 16:10 ` [PULL 07/25] scripts/qemu-gdb: Support coroutine dumps in coredumps Kevin Wolf
2025-02-10 16:10 ` [PULL 08/25] block-backend: Fix argument order when calling 'qapi_event_send_block_io_error()' Kevin Wolf
2025-02-10 16:10 ` [PULL 09/25] block: Add 'active' field to BlockDeviceInfo Kevin Wolf
2025-02-10 16:10 ` [PULL 10/25] block: Allow inactivating already inactive nodes Kevin Wolf
2025-02-10 16:10 ` [PULL 11/25] block: Inactivate external snapshot overlays when necessary Kevin Wolf
2025-02-10 16:10 ` Kevin Wolf [this message]
2025-02-10 16:10 ` [PULL 13/25] block: Don't attach inactive child to active node Kevin Wolf
2025-02-10 16:10 ` [PULL 14/25] block: Fix crash on block_resize on inactive node Kevin Wolf
2025-02-10 16:10 ` [PULL 15/25] block: Add option to create inactive nodes Kevin Wolf
2025-02-10 16:10 ` [PULL 16/25] block: Add blockdev-set-active QMP command Kevin Wolf
2025-02-10 16:10 ` [PULL 17/25] block: Support inactive nodes in blk_insert_bs() Kevin Wolf
2025-02-10 16:10 ` [PULL 18/25] block/export: Don't ignore image activation error in blk_exp_add() Kevin Wolf
2025-02-10 16:10 ` [PULL 19/25] block: Drain nodes before inactivating them Kevin Wolf
2025-02-10 16:10 ` [PULL 20/25] block/export: Add option to allow export of inactive nodes Kevin Wolf
2025-02-10 16:10 ` [PULL 21/25] nbd/server: Support " Kevin Wolf
2025-02-10 16:10 ` [PULL 22/25] iotests: Add filter_qtest() Kevin Wolf
2025-02-10 16:10 ` [PULL 23/25] iotests: Add qsd-migrate case Kevin Wolf
2025-02-10 16:10 ` [PULL 24/25] iotests: Add (NBD-based) tests for inactive nodes Kevin Wolf
2025-02-10 16:10 ` [PULL 25/25] block: remove unused BLOCK_OP_TYPE_DATAPLANE Kevin Wolf
2025-02-11 2:37 ` [PULL 00/25] Block layer patches Stefan Hajnoczi
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=20250210161034.76494-13-kwolf@redhat.com \
--to=kwolf@redhat.com \
--cc=qemu-block@nongnu.org \
--cc=qemu-devel@nongnu.org \
/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).