diff options
author | Peter Maydell <peter.maydell@linaro.org> | 2024-10-09 20:12:11 +0100 |
---|---|---|
committer | Peter Maydell <peter.maydell@linaro.org> | 2024-10-09 20:12:11 +0100 |
commit | 05adb38839ba656c7383a548b460d95c91e2febe (patch) | |
tree | 43a8616f6267c89a03991baabbb4a1699dc66a11 /tests | |
parent | 838fc0a8769d7cc6edfe50451ba4e3368395f5c1 (diff) | |
parent | 2e49d6a20b2193378e508b01de82eaffc5adc3bc (diff) | |
download | qemu-05adb38839ba656c7383a548b460d95c91e2febe.zip qemu-05adb38839ba656c7383a548b460d95c91e2febe.tar.gz qemu-05adb38839ba656c7383a548b460d95c91e2febe.tar.bz2 |
Merge tag 'migration-20241009-pull-request' of https://gitlab.com/peterx/qemu into staging
Migration pull request
- Ani's patch to complete the memory API on coalesced IO / eventfd notifies
- Fabiano's Coverity fix on using pstrcpy() over strncpy()
- Dave's series on removing/deprecating zero-blocks and uffd cleanups
- Juraj's one more fix on multifd/cancel test where it can fail when
cancellation happens too slow on src
- Dave's one more remove deadcode patch in iova-tree.c
- Yuan's build fix for multifd qpl compressor
# -----BEGIN PGP SIGNATURE-----
#
# iIgEABYKADAWIQS5GE3CDMRX2s990ak7X8zN86vXBgUCZwZ6CBIccGV0ZXJ4QHJl
# ZGhhdC5jb20ACgkQO1/MzfOr1wa3ZwD9HiAN9m7WOfZxXKOVIIwhOjUNTw0FiFeO
# HMxp8A2jeYsBAK+d5lYGX1V2FtQ152YiOJQzRW31MkdAOishJzcHCXgO
# =gBW0
# -----END PGP SIGNATURE-----
# gpg: Signature made Wed 09 Oct 2024 13:41:44 BST
# gpg: using EDDSA key B9184DC20CC457DACF7DD1A93B5FCCCDF3ABD706
# gpg: issuer "peterx@redhat.com"
# gpg: Good signature from "Peter Xu <xzpeter@gmail.com>" [marginal]
# gpg: aka "Peter Xu <peterx@redhat.com>" [marginal]
# gpg: WARNING: This key is not certified with sufficiently trusted signatures!
# gpg: It is not certain that the signature belongs to the owner.
# Primary key fingerprint: B918 4DC2 0CC4 57DA CF7D D1A9 3B5F CCCD F3AB D706
* tag 'migration-20241009-pull-request' of https://gitlab.com/peterx/qemu:
migration/multifd: fix build error when qpl compression is enabled
util/iova-tree: Remove deadcode
tests/migration-test: Wait for cancellation sooner in multifd cancel
util/userfaultfd: Remove unused uffd_poll_events
migration/postcopy: Use uffd helpers
util/userfaultfd: Return -errno on error
migration: Remove unused socket_send_channel_create_sync
migration: Deprecate zero-blocks capability
migration: Remove unused migrate_zero_blocks
migration: Remove migrate_cap_set
migration/multifd: Ensure packet->ramblock is null-terminated
memory: notify hypervisor of all eventfds during listener (de)registration
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Diffstat (limited to 'tests')
-rw-r--r-- | tests/qtest/migration-test.c | 12 |
1 files changed, 10 insertions, 2 deletions
diff --git a/tests/qtest/migration-test.c b/tests/qtest/migration-test.c index 814ec10..95e45b5 100644 --- a/tests/qtest/migration-test.c +++ b/tests/qtest/migration-test.c @@ -3267,6 +3267,16 @@ static void test_multifd_tcp_cancel(void) qtest_wait_qemu(to); qtest_quit(to); + /* + * Ensure the source QEMU finishes its cancellation process before we + * proceed with the setup of the next migration. The test_migrate_start() + * function and others might want to interact with the source in a way that + * is not possible while the migration is not canceled properly. For + * example, setting migration capabilities when the migration is still + * running leads to an error. + */ + wait_for_migration_status(from, "cancelled", NULL); + args = (MigrateStart){ .only_target = true, }; @@ -3282,8 +3292,6 @@ static void test_multifd_tcp_cancel(void) /* Start incoming migration from the 1st socket */ migrate_incoming_qmp(to2, "tcp:127.0.0.1:0", "{}"); - wait_for_migration_status(from, "cancelled", NULL); - migrate_ensure_non_converge(from); migrate_qmp(from, to2, NULL, NULL, "{}"); |