mbox

[PULL,00/20] migration queue

Message ID 20210701141545.193571-1-dgilbert@redhat.com (mailing list archive)
State New, archived
Headers show

Pull-request

https://gitlab.com/dagrh/qemu.git tags/pull-migration-20210701a

Message

Dr. David Alan Gilbert July 1, 2021, 2:15 p.m. UTC
From: "Dr. David Alan Gilbert" <dgilbert@redhat.com>

The following changes since commit 1ec2cd0ce2ca94292ce237becc2c21b4eb9edca0:

  Merge remote-tracking branch 'remotes/nvme/tags/nvme-next-pull-request' into staging (2021-06-30 21:09:27 +0100)

are available in the Git repository at:

  https://gitlab.com/dagrh/qemu.git tags/pull-migration-20210701a

for you to fetch changes up to 9c21e61435e09a7a67f951a096b93183a7a5ad89:

  migration/rdma: Use error_report to suppress errno message (2021-07-01 12:21:32 +0100)

----------------------------------------------------------------
Migration and virtiofs pull 2021-07-01

A bunch of small fixes and improvements; two particular to note:
  a) Peter's fix to migration-test for uffd, means that a lot of
migration tests will start running in a lot of places again when they'd
previously been skipped.
  b) Vivek's 'Fix fuse_setxattr...' fixes our build against an API
breakage in the kernel headers.

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

----------------------------------------------------------------
Daniel P. Berrangé (2):
      virtiofsd: use GDateTime for formatting timestamp for debug messages
      docs: describe the security considerations with virtiofsd xattr mapping

Feng Lin (1):
      migration: fix the memory overwriting risk in add_to_iovec

Greg Kurz (1):
      virtiofsd: Don't allow file creation with FUSE_OPEN

Hyman Huang(黄勇) (2):
      tests/migration: parse the thread-id key of CpuInfoFast
      tests/migration: fix "downtime_limit" type when "migrate-set-parameters"

Laurent Vivier (2):
      migration: move wait-unplug loop to its own function
      migration: failover: continue to wait card unplug on error

Li Zhijian (1):
      migration/rdma: Use error_report to suppress errno message

Peter Xu (4):
      tests: migration-test: Still run the rest even if uffd missing
      tests: migration-test: Add dirty ring test
      migration: Move yank outside qemu_start_incoming_migration()
      migration: Allow reset of postcopy_recover_triggered when failed

Vivek Goyal (7):
      virtiofsd: Fix fuse setxattr() API change issue
      virtiofsd: Fix xattr operations overwriting errno
      virtiofsd: Add support for extended setxattr
      virtiofsd: Add umask to seccom allow list
      virtiofsd: Add capability to change/restore umask
      virtiofsd: Switch creds, drop FSETID for system.posix_acl_access xattr
      virtiofsd: Add an option to enable/disable posix acls

 docs/tools/virtiofsd.rst              |  58 +++++++-
 migration/migration.c                 |  89 +++++++-----
 migration/qemu-file.c                 |   5 +
 migration/rdma.c                      |   4 +-
 tests/migration/guestperf/engine.py   |   4 +-
 tests/qtest/migration-test.c          |  69 +++++++--
 tools/virtiofsd/fuse_common.h         |   5 +
 tools/virtiofsd/fuse_lowlevel.c       |  24 +++-
 tools/virtiofsd/fuse_lowlevel.h       |   3 +-
 tools/virtiofsd/helper.c              |   1 +
 tools/virtiofsd/passthrough_ll.c      | 254 +++++++++++++++++++++++++++++-----
 tools/virtiofsd/passthrough_seccomp.c |   1 +
 12 files changed, 428 insertions(+), 89 deletions(-)

Comments

Peter Maydell July 5, 2021, 8:57 a.m. UTC | #1
On Thu, 1 Jul 2021 at 15:19, Dr. David Alan Gilbert (git)
<dgilbert@redhat.com> wrote:
>
> From: "Dr. David Alan Gilbert" <dgilbert@redhat.com>
>
> The following changes since commit 1ec2cd0ce2ca94292ce237becc2c21b4eb9edca0:
>
>   Merge remote-tracking branch 'remotes/nvme/tags/nvme-next-pull-request' into staging (2021-06-30 21:09:27 +0100)
>
> are available in the Git repository at:
>
>   https://gitlab.com/dagrh/qemu.git tags/pull-migration-20210701a
>
> for you to fetch changes up to 9c21e61435e09a7a67f951a096b93183a7a5ad89:
>
>   migration/rdma: Use error_report to suppress errno message (2021-07-01 12:21:32 +0100)
>
> ----------------------------------------------------------------
> Migration and virtiofs pull 2021-07-01
>
> A bunch of small fixes and improvements; two particular to note:
>   a) Peter's fix to migration-test for uffd, means that a lot of
> migration tests will start running in a lot of places again when they'd
> previously been skipped.
>   b) Vivek's 'Fix fuse_setxattr...' fixes our build against an API
> breakage in the kernel headers.
>
> Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
>

This seems to reliably cause 'make check' to hang in the openbsd VM.
(The NetBSD and FreeBSD VMs are OK.)

-- PMM
Daniel P. Berrangé July 5, 2021, 9:03 a.m. UTC | #2
On Mon, Jul 05, 2021 at 09:57:20AM +0100, Peter Maydell wrote:
> On Thu, 1 Jul 2021 at 15:19, Dr. David Alan Gilbert (git)
> <dgilbert@redhat.com> wrote:
> >
> > From: "Dr. David Alan Gilbert" <dgilbert@redhat.com>
> >
> > The following changes since commit 1ec2cd0ce2ca94292ce237becc2c21b4eb9edca0:
> >
> >   Merge remote-tracking branch 'remotes/nvme/tags/nvme-next-pull-request' into staging (2021-06-30 21:09:27 +0100)
> >
> > are available in the Git repository at:
> >
> >   https://gitlab.com/dagrh/qemu.git tags/pull-migration-20210701a
> >
> > for you to fetch changes up to 9c21e61435e09a7a67f951a096b93183a7a5ad89:
> >
> >   migration/rdma: Use error_report to suppress errno message (2021-07-01 12:21:32 +0100)
> >
> > ----------------------------------------------------------------
> > Migration and virtiofs pull 2021-07-01
> >
> > A bunch of small fixes and improvements; two particular to note:
> >   a) Peter's fix to migration-test for uffd, means that a lot of
> > migration tests will start running in a lot of places again when they'd
> > previously been skipped.

Presumably this change is the cause of....

> >   b) Vivek's 'Fix fuse_setxattr...' fixes our build against an API
> > breakage in the kernel headers.
> >
> > Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
> >
> 
> This seems to reliably cause 'make check' to hang in the openbsd VM.
> (The NetBSD and FreeBSD VMs are OK.)

...the openbsd VM failure

Guess there was a pre-existing bug affect openbsd we've now exposed.

Regards,
Daniel