The QUORUM_REPORT_BAD event is emitted whenever there's an I/O error
in a child of a Quorum device. This event is emitted at a maximum rate
of 1 per second. This means that an error in one of the children will
mask errors in the other children if they happen within the same 1
second interval.

This patch modifies qapi_event_throttle_equal() so QUORUM_REPORT_BAD
events are kept separately if they come from different children.
Signed-off-by: NAlberto Garcia <berto@igalia.com>
Reviewed-by: NMax Reitz <mreitz@redhat.com>
Message-id: b989c0cb3755bc4b6696e796fa8ed2ef6c56606a.1457610443.git.berto@igalia.com
Reviewed-by: NEric Blake <eblake@redhat.com>
Signed-off-by: NMax Reitz <mreitz@redhat.com>

quorum_aio_cb() emits the QUORUM_REPORT_BAD event if there's
an I/O error in a Quorum child. However sacb->aiocb must be
correctly initialized for this to happen. read_quorum_children() and
read_fifo_child() are not doing this, which results in a QEMU crash.
Signed-off-by: NAlberto Garcia <berto@igalia.com>
Reviewed-by: NMax Reitz <mreitz@redhat.com>
Message-id: 8138570d071ba7e25db3736979234a1fd71dbd05.1457610443.git.berto@igalia.com
Signed-off-by: NMax Reitz <mreitz@redhat.com>

The newly added type parameter for the QUORUM_REPORT_BAD event changed
the output of iotest 081, so the reference should be amended
accordingly.
Signed-off-by: NMax Reitz <mreitz@redhat.com>
Message-id: 1457705687-27122-1-git-send-email-mreitz@redhat.com
Reviewed-by: NAlberto Garcia <berto@igalia.com>

Signed-off-by: NFam Zheng <famz@redhat.com>
Reviewed-by: NJohn Snow <jsnow@redhat.com>
Message-id: 1457412306-18940-6-git-send-email-famz@redhat.com
Signed-off-by: NMax Reitz <mreitz@redhat.com>

The only code change is making bdrv_dirty_bitmap_truncate public. It is
used in block.c.

Also two long lines (bdrv_get_dirty) are wrapped.
Signed-off-by: NFam Zheng <famz@redhat.com>
Reviewed-by: NJohn Snow <jsnow@redhat.com>
Message-id: 1457412306-18940-5-git-send-email-famz@redhat.com
Signed-off-by: NMax Reitz <mreitz@redhat.com>

Following patches to refactor and move block dirty bitmap code could use
this.
Signed-off-by: NFam Zheng <famz@redhat.com>
Reviewed-by: NJohn Snow <jsnow@redhat.com>
Message-id: 1457412306-18940-4-git-send-email-famz@redhat.com
Signed-off-by: NMax Reitz <mreitz@redhat.com>

Signed-off-by: NFam Zheng <famz@redhat.com>
Reviewed-by: NJohn Snow <jsnow@redhat.com>
Message-id: 1457412306-18940-3-git-send-email-famz@redhat.com
Signed-off-by: NMax Reitz <mreitz@redhat.com>

"s->bitmap" tracks done sectors, we only check bit states without using any
iterator which HBitmap is good for. Switch to "Bitmap" which is simpler and
more memory efficient.

Meanwhile, rename it to done_bitmap, to reflect the intention.
Signed-off-by: NFam Zheng <famz@redhat.com>
Reviewed-by: NJohn Snow <jsnow@redhat.com>
Message-id: 1457412306-18940-2-git-send-email-famz@redhat.com
Signed-off-by: NMax Reitz <mreitz@redhat.com>

All users of the block layers are supposed to go through a BlockBackend.
The .bdrv_create() implementation is one such user, so this patch
converts it.
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

All users of the block layers are supposed to go through a BlockBackend.
The .bdrv_create() implementation is one such user, so this patch
converts it.
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

All users of the block layers are supposed to go through a BlockBackend.
The .bdrv_create() implementation is one such user, so this patch
converts it.
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

All users of the block layers are supposed to go through a BlockBackend.
The .bdrv_create() implementation is one such user, so this patch
converts it.
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

All users of the block layers are supposed to go through a BlockBackend.
The .bdrv_create() implementation is one such user, so this patch
converts it.
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

All users of the block layers are supposed to go through a BlockBackend.
The .bdrv_create() implementation is one such user, so this patch
converts it.
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

All users of the block layers are supposed to go through a BlockBackend.
The .bdrv_create() implementation is one such user, so this patch
converts it.
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

All users of the block layers are supposed to go through a BlockBackend.
The .bdrv_create() implementation is one such user, so this patch
converts it.
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

All users of the block layers are supposed to go through a BlockBackend.
The .bdrv_create() implementation is one such user, so this patch
converts it.
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

We check that the guest can't write beyond the end of its disk, but for
other internal users it can make sense to allow growing a file.
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

There's no reason to use a writethrough cache mode while creating an
image.
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

Now that we can use drive_add to create new nodes without a BB, we also
want to be able to delete such nodes again.
Signed-off-by: NKevin Wolf <kwolf@redhat.com>
Reviewed-by: NMax Reitz <mreitz@redhat.com>

This patch adds an option to the drive_add HMP command to create only a
BlockDriverState without a BlockBackend on top.

The motivation for this is that libvirt needs to specify options to a
migration target (specifically, detect-zeroes). drive-mirror doesn't
allow specifying options, and the proper way to do this is to create the
target BDS separately with blockdev-add (where you can specify options)
and then use blockdev-mirror to that BDS.

However, libvirt can't use blockdev-add as long as it is still
experimental, and we're expecting that it will still take some time, so
we need to resort to drive_add.

The problem with drive_add is that so far it always created a BB, and
BDSes with a BB can't be used as a mirroring target as long as we don't
support multiple BBs per BDS - and while we're working towards that
goal, it's another thing that will still take some time.

So to achieve the goal, the simplest solution to provide the
functionality now without adding one-off options to the mirror QMP
commands is to extend drive_add to create nodes without BBs.
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

Signed-off-by: NFam Zheng <famz@redhat.com>
Reviewed-by: NPaolo Bonzini <pbonzini@redhat.com>
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

Signed-off-by: NFam Zheng <famz@redhat.com>
Reviewed-by: NPaolo Bonzini <pbonzini@redhat.com>
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

It is only called once for each opened image, so we can do it the easy
way.
Reviewed-by: NPeter Xu <peterx@redhat.com>
Signed-off-by: NFam Zheng <famz@redhat.com>
Reviewed-by: NPaolo Bonzini <pbonzini@redhat.com>
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

Without setting explicit defaults in the options, blockdev-add without
an ID ended up defaulting to writethrough. It should be writeback as
documented.
Signed-off-by: NKevin Wolf <kwolf@redhat.com>
Reviewed-by: NEric Blake <eblake@redhat.com>

Since commit 91a097e7, we end up with a somewhat weird cache mode
configuration with snapshot=on: The commit broke the cache mode
inheritance for the snapshot overlay so that it is opened as
writethrough instead of unsafe now. The following bdrv_append() call to
put it on top of the tree swaps the WCE flag with the snapshot's backing
file (i.e. the originally given file), so what we eventually get is
cache=writeback on the temporary overlay and
cache=writethrough,cache.no-flush=on on the real image file.

This patch changes things so that the temporary overlay gets
cache=unsafe again like it used to, and the real images get whatever the
user specified. This means that cache.direct is now respected even with
snapshot=on, and in the case of committing changes, the final flush is
no longer ignored except explicitly requested by the user.
Signed-off-by: NKevin Wolf <kwolf@redhat.com>
Reviewed-by: NMax Reitz <mreitz@redhat.com>

Calling bdrv_img_create() with a size of -1 means that it determines the
size automatically by opening the backing file. However, in the case of
live snapshots, the backing file is already opened and we must avoid
opening the same image twice at the same time. Apart from that, just
getting the size from the already existing BDS is a lot less overhead
than opening a new instance.
Signed-off-by: NKevin Wolf <kwolf@redhat.com>
Reviewed-by: NJeff Cody <jcody@redhat.com>

Keep flush interface the same logic as quorum read/write, Otherwise in
following scenario, we'll encounter unexpected errors.

Quorum has two children(A, B). A do flush sucessfully, but B flush failed.
This cause the filesystem of guest become read-only with following errors:

end_request: I/O error, dev vda, sector 11159960
Aborting journal on device vda3-8
EXT4-fs error (device vda3): ext4_journal_start_sb:327: Detected abort journal
EXT4-fs (vda3): Remounting filesystem read-only

Cc: Dr. David Alan Gilbert <dgilbert@redhat.com>
Cc: Wen Congyang <wency@cn.fujitsu.com>
Signed-off-by: NWen Congyang <wency@cn.fujitsu.com>
Signed-off-by: NChanglong Xie <xiecl.fnst@cn.fujitsu.com>
Reviewed-by: NAlberto Garcia <berto@igalia.com>
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

Introduce QuorumOpType, and make QUORUM_REPORT_BAD compatible
with it.

Cc: Dr. David Alan Gilbert <dgilbert@redhat.com>
Cc: Wen Congyang <wency@cn.fujitsu.com>
Signed-off-by: NWen Congyang <wency@cn.fujitsu.com>
Signed-off-by: NChanglong Xie <xiecl.fnst@cn.fujitsu.com>
Reviewed-by: NAlberto Garcia <berto@igalia.com>
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

Cc: Dr. David Alan Gilbert <dgilbert@redhat.com>
Cc: Wen Congyang <wency@cn.fujitsu.com>
Signed-off-by: NWen Congyang <wency@cn.fujitsu.com>
Signed-off-by: NChanglong Xie <xiecl.fnst@cn.fujitsu.com>
Reviewed-by: NAlberto Garcia <berto@igalia.com>
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

Signed-off-by: NJeff Cody <jcody@redhat.com>
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

When QEMU creates a VHD image, it goes by the original spec,
calculating the current_size based on the nearest CHS geometry (with an
exception for disks > 127GB).

Apparently, Azure will only allow images that are sized to the nearest
MB, and the current_size as calculated from CHS cannot guarantee that.

Allow QEMU to create images similar to how Hyper-V creates images, by
setting current_size to the specified virtual disk size.  This
introduces an option, force_size, to be passed to the vpc format during
image creation, e.g.:

    qemu-img convert -f raw -o force_size -O vpc test.img test.vhd

When using the "force_size" option, the creator app field used by
QEMU will be "qem2" instead of "qemu", to indicate the difference.
In light of this, we also add parsing of the "qem2" field during
vpc_open.

Bug reference: https://bugs.launchpad.net/qemu/+bug/1490611Signed-off-by: NJeff Cody <jcody@redhat.com>
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

This tests auto-detection, and overrides, of VHD image sizes created
by Virtual PC, Hyper-V, and Disk2vhd.

This adds three sample images:

hyperv2012r2-dynamic.vhd.bz2 - dynamic VHD image created with Hyper-V
virtualpc-dynamic.vhd.bz2    - dynamic VHD image created with Virtual PC
d2v-zerofilled.vhd.bz2       - dynamic VHD image created with Disk2vhd
Signed-off-by: NJeff Cody <jcody@redhat.com>
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

The VHD file format is used by both Virtual PC, and Hyper-V.  However,
how the virtual disk size is calculated varies between the two.

Virtual PC uses the CHS drive parameters to determine the drive size.
Hyper-V, on the other hand, uses the current_size field in the footer
when determining image size.

This is problematic for a few reasons:

* VHD images from Hyper-V, using CHS calculations, will likely be
  trunctated.

* If we just rely always on current_size, then QEMU may have data
  compatibility issues with Virtual PC (we may write too much data
  into a VHD file to be used by Virtual PC, for instance).

* Existing VHD images created by QEMU have used the CHS calculations,
  except for images exceeding the 127GB limit.  We want to remain
  compatible with our own generated images.

Luckily, the VHD specification defines a 'Creator App' field, that is
used to indicate what software created the VHD file.

This patch does two things:

    1. Uses the 'Creator App' field to help determine how to calculate
       size, and

    2. Adds a VPC format option 'force_size_calc', so that the user can
       override the 'Creator App' auto-detection, in case there exist
       VHD images with unknown or contradictory 'Creator App' entries.

N.B.: We currently use the maximum CHS value as an indication to use the
current_size field.  This patch does not change that, even with the
'force_size_calc' option.
Signed-off-by: NJeff Cody <jcody@redhat.com>
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

Since commit 5ec18f8c, query-blockstats didn't return the statistics of
drives without media any more because such drives have only a BB now,
but not a BDS any more.

This patch fixes the regression so that query-blockstats iterates over
BBs by default and empty drives are displayed again.
Signed-off-by: NKevin Wolf <kwolf@redhat.com>
Reviewed-by: NMax Reitz <mreitz@redhat.com>

The new functions handles the data that is taken from the
BlockDriverState.
Signed-off-by: NKevin Wolf <kwolf@redhat.com>
Reviewed-by: NMax Reitz <mreitz@redhat.com>

The new functions handles the data that is taken from the BlockBackend.
Signed-off-by: NKevin Wolf <kwolf@redhat.com>
Reviewed-by: NMax Reitz <mreitz@redhat.com>

Not particularly important since qemu-img exits immediately after
calling img_rebase, but easily fixed.  Coverity says thanks.
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>
Reviewed-by: NAlberto Garcia <berto@igalia.com>
Signed-off-by: NKevin Wolf <kwolf@redhat.com>

migration:
 - postcopy is no longer experimental
 - fix a use-after-free in postcopy
 - fix a compile warning

# gpg: Signature made Fri 11 Mar 2016 12:29:33 GMT using RSA key ID 854083B6
# gpg: Good signature from "Amit Shah <amit@amitshah.net>"
# gpg:                 aka "Amit Shah <amit@kernel.org>"
# gpg:                 aka "Amit Shah <amitshah@gmx.net>"

* remotes/amit-migration/tags/migration-for-2.6-7:
  postcopy: Remove the x-
  postcopy: listen thread is never joined
  migration: fix use-after-free in loadvm_postcopy_handle_run_bh
  migration: fix warning for source_return_path_thread
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>

Merge I/O fixes for win32

# gpg: Signature made Fri 11 Mar 2016 10:03:20 GMT using RSA key ID 15104FDF
# gpg: Good signature from "Daniel P. Berrange <dan@berrange.com>"
# gpg:                 aka "Daniel P. Berrange <berrange@redhat.com>"

* remotes/berrange/tags/pull-io-win32-2016-03-11-1:
  osdep: remove use of socket_error() from all code
  osdep: add wrappers for socket functions
  char: remove qemu_chr_open_socket_fd method
  char: remove socket_try_connect method
  char: remove qemu_chr_finish_socket_connection method
  io: implement socket watch for win32 using WSAEventSelect+select
  io: remove checking of EWOULDBLOCK
  io: use qemu_accept to ensure SOCK_CLOEXEC is set
  io: introduce qio_channel_create_socket_watch
  io: pass HANDLE to g_source_add_poll on Win32
  io: fix copy+paste mistake in socket error message
  io: assert errors before asserting content in I/O test
  io: set correct error object in background reader test thread
  io: wait for incoming client in socket test
  io: bind to socket before creating QIOChannelSocket
  io: initialize sockets in test program
  io: use bind() to check for IPv4/6 availability
  osdep: fix socket_error() to work with Mingw64
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>