Currently if gluster AIO callback thread fails to notify the QEMU thread about
AIO completion, we try graceful recovery by marking the disk drive as
inaccessible. This error recovery code is race-prone as found by Asias and
Stefan. However as found out by Paolo, this kind of error is impossible and
hence simplify the code that handles this error recovery.
Signed-off-by: Bharata B Rao <bharata@linux.vnet.ibm.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
"Incoming" function prototypes and "outgoing" function calls must match
reality. Implemented using the "struct BlockDriver" definition in
"include/block/block_int.h", and gcc errors & warnings.
v1->v2:
On 08/20/13 09:51, Kevin Wolf wrote:
> Am 18.08.2013 um 16:29 hat Paolo Bonzini geschrieben:
>> Il 16/08/2013 16:15, Laszlo Ersek ha scritto:
>>> +static int raw_reopen_prepare(BDRVReopenState *reopen_state,
>>> + BlockReopenQueue *queue, Error **errp)
>>> {
>>> - return bdrv_reopen_prepare(bs->file);
>>> + BDRVReopenState tmp = *reopen_state;
>>> +
>>> + tmp.bs = tmp.bs->file;
>>> + return bdrv_reopen_prepare(&tmp, queue, errp);
>>> }
>>
>> This should just return zero, my fault.
>
> Which is because bdrv_reopen_queue() already queues bs->file for reopen.
> The simple return 0; implementation is shared by all other format drivers
> that support reopening images.
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
On 08/05/13 15:03, Paolo Bonzini wrote:
>
> [...]
>
> 5) Formats are registered with bdrv_register (takes a BlockDriver*). You
> also need to pass the caller of bdrv_register to block_init.
Fill in the BlockDriver structure with the raw_*() functions that have
been added to "block/raw_bsd.c", in the order the fields are defined in
"include/block/block_int.h".
I needed more explanation / naming examples for registering the driver
than what Paolo gave me, so I copied / adapted from "block/qcow2.c". The
parts I took as basis for modification are blamed on
commit 5efa9d5a8b
Author: Anthony Liguori <aliguori@us.ibm.com>
Date: Sat May 9 17:03:42 2009 -0500
Convert block infrastructure to use new module init functionality
commit 20d97356c9
Author: Blue Swirl <blauwirbel@gmail.com>
Date: Fri Apr 23 20:19:47 2010 +0000
Fix OpenBSD build
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
On 08/05/13 15:03, Paolo Bonzini wrote:
>
> [...]
>
> 4) There is another member, .create_options, which is an array of
> QEMUOptionParameter structs, terminated by an all-zero item. The only
> option you need is for the virtual disk size. You will find something
> to copy from in other block drivers, for example block/qcow2.c.
Code taken and adapted from "block/qcow2.c", as suggested. The code being
copied/modified is blamed on
commit 20d97356c9
Author: Blue Swirl <blauwirbel@gmail.com>
Date: Fri Apr 23 20:19:47 2010 +0000
Fix OpenBSD build
and
commit 7c80ab3f21
Author: Jes Sorensen <Jes.Sorensen@redhat.com>
Date: Fri Dec 17 16:02:39 2010 +0100
block/qcow2.c: rename qcow_ functions to qcow2_
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
On 08/05/13 15:03, Paolo Bonzini wrote:
>
> [...]
>
> 3) These members are special
>
> .format_name is the string "raw"
> .bdrv_open raw_open should set bs->sg to bs->file->sg and return 0
> .bdrv_close raw_close should do nothing
> .bdrv_probe raw_probe should just return 1.
v1->v2:
On 08/20/13 10:11, Kevin Wolf wrote:
> Am 16.08.2013 um 16:15 hat Laszlo Ersek geschrieben:
>> +static int raw_probe(void)
>> +{
>> + return 1;
>> +}
>
> Maybe add a comment here like "smallest possible positive score so that
> raw is used if and only if no other block driver works".
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
On 08/05/13 15:03, Paolo Bonzini wrote:
>
> [...]
>
> 2) This is also a simple forwarder function:
>
> .bdrv_create
>
> but there is no BlockDriverState argument so the forwarded-to function
> does not have a bs->file argument either. The forwarded-to function is
> bdrv_create_file.
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
On 08/05/13 15:03, Paolo Bonzini wrote:
>
> [...]
>
> 1) BlockDriver is a struct in which these function members are
> interesting:
>
> .bdrv_reopen_prepare
> .bdrv_co_readv
> .bdrv_co_writev
> .bdrv_co_is_allocated
> .bdrv_co_write_zeroes
> .bdrv_co_discard
> .bdrv_getlength
> .bdrv_get_info
> .bdrv_truncate
> .bdrv_is_inserted
> .bdrv_media_changed
> .bdrv_eject
> .bdrv_lock_medium
> .bdrv_ioctl
> .bdrv_aio_ioctl
> .bdrv_has_zero_init
>
> They should be implemented as simple forwarders (see above). There are
> 16 functions listed here, you can easily see how this already accounts
> for 100+ SLOC roughly...
>
> The implementations of bdrv_co_readv and bdrv_co_writev should also call
> BLKDBG_EVENT on bs->file too, before forwarding to bs->file. The events
> to be generated are BLKDBG_READ_AIO and BLKDBG_WRITE_AIO.
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
On 08/05/13 15:03, Paolo Bonzini wrote:
>
>
> ----- Original Message -----
>> From: "Laszlo Ersek" <lersek@redhat.com>
>> To: "Paolo Bonzini" <pbonzini@redhat.com>
>> Sent: Monday, August 5, 2013 2:43:46 PM
>> Subject: Re: [PATCH 1/2] raw: add license header
>>
>> On 08/02/13 00:27, Paolo Bonzini wrote:
>>> On 08/01/2013 10:13 AM, Christoph Hellwig wrote:
>>>> On Wed, Jul 31, 2013 at 08:19:51AM +0200, Paolo Bonzini wrote:
>>>>> Most of the block layer is under the BSD license, thus it is
>>>>> reasonable to license block/raw.c the same way. CCed people should
>>>>> ACK by replying with a Signed-off-by line.
>>>>
>>>> The coded was intended to be GPLv2.
>>>
>>> Laszlo, would you be willing to do clean-room reverse engineering?
>>>
>>> (No rants, please. :))
>>
>> What's the scope exactly?
>
> It's quite small, it's a file full of forwarders like
>
> static void raw_foo(BlockDriverState *bs)
> {
> return bdrv_foo(bs->file);
> }
>
> It's 170 lines of code, all as boring as this. I only picked you
> because I'm quite certain you have never seen the file (and the answer
> confirmed it).
>
> Basically:
>
> 1) BlockDriver is a struct in which these function members are
> interesting:
>
> .bdrv_reopen_prepare
> .bdrv_co_readv
> .bdrv_co_writev
> .bdrv_co_is_allocated
> .bdrv_co_write_zeroes
> .bdrv_co_discard
> .bdrv_getlength
> .bdrv_get_info
> .bdrv_truncate
> .bdrv_is_inserted
> .bdrv_media_changed
> .bdrv_eject
> .bdrv_lock_medium
> .bdrv_ioctl
> .bdrv_aio_ioctl
> .bdrv_has_zero_init
>
> They should be implemented as simple forwarders (see above).
> There are 16 functions listed here, you can easily see how this
> already accounts for 100+ SLOC roughly...
>
> The implementations of bdrv_co_readv and bdrv_co_writev should also
> call BLKDBG_EVENT on bs->file too, before forwarding to bs->file. The
> events to be generated are BLKDBG_READ_AIO and BLKDBG_WRITE_AIO.
>
> 2) This is also a simple forwarder function:
>
> .bdrv_create
>
> but there is no BlockDriverState argument so the forwarded-to function
> does not have a bs->file argument either. The forwarded-to function
> is bdrv_create_file.
>
> 3) These members are special
>
> .format_name is the string "raw"
> .bdrv_open raw_open should set bs->sg to bs->file->sg and return 0
> .bdrv_close raw_close should do nothing
> .bdrv_probe raw_probe should just return 1.
>
> 4) There is another member, .create_options, which is an array of
> QEMUOptionParameter structs, terminated by an all-zero item. The only
> option you need is for the virtual disk size. You will find something
> to copy from in other block drivers, for example block/qcow2.c.
>
> 5) Formats are registered with bdrv_register (takes a BlockDriver*).
> You also need to pass the caller of bdrv_register to block_init.
>
> 6) I'm not sure how to organize the patch series, so I'll leave this to
> your creativity. I guess in this case move/copy detection of git should
> be disabled. I would definitely include this spec in the commit
> message as a proof of clean-room reverse engineering.
>
> 7) Remember a BSD header like the one in block.c.
>
> Paolo
This patch implements the email up to the paragraph ending with "100+ SLOC
roughly". The skeleton is generated from the list there, with a simple
shell loop using "sed" and the raw_foo() template.
The BSD license block is copied (and reflowed) from
"util/qemu-progress.c".
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
The expression "1LL << 63" tries to shift the 1 into the sign bit of a
'long long', which provokes a clang sanitizer warning:
runtime error: left shift of 1 by 63 places cannot be represented in type 'long long'
Use "1ULL << 63" as the definition of QCOW_OFLAG_COPIED instead
to avoid this. For consistency, we also update the other QCOW_OFLAG
definitions to use the ULL suffix rather than LL, though only the
shift by 63 is undefined behaviour.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
By the time that qemu 1.7 will be released, enough time will have passed
since qemu 1.1, which is the first version to understand version 3
images, that changing the default shouldn't hurt many people any more
and the benefits of using the new format outweigh the pain.
qemu-iotests already runs with compat=1.1 by default.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
The io_flush argument to qemu_aio_set_event_notifier() has been removed
since the block layer learnt to drain requests by itself. Fix the
Windows build for win32-aio.o by updating the
qemu_aio_set_event_notifier() call and dropping win32_aio_flush_cb().
Reviewed-by: Stefan Weil <sw@weilnetz.de>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
This is an autogenerated patch using scripts/switch-timer-api.
Switch the entire code base to using the new timer API.
Note this patch may introduce some line length issues.
Signed-off-by: Alex Bligh <alex@alex.org.uk>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Convert block_job_sleep_ns and co_sleep_ns to use the new timer
API.
Signed-off-by: Alex Bligh <alex@alex.org.uk>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
VMware ESX hosts also use different create and extent types for flat
files, respectively "vmfs" and "VMFS". This is not documented, but it
can be found at http://kb.vmware.com/kb/10002511 (Recreating a missing
virtual machine disk (VMDK) descriptor file).
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
VMware ESX hosts use a variant of the VMDK3 format, identified by the
vmfsSparse create type ad the VMFSSPARSE extent type.
It has 16 KB grain tables (L2) and a variable-size grain directory (L1).
In addition, the grain size is always 512, but that is not a problem
because it is included in the header.
The format of the extents is documented in the VMDK spec. The format
of the descriptor file is not documented precisely, but it can be
found at http://kb.vmware.com/kb/10026353 (Recreating a missing virtual
machine disk (VMDK) descriptor file for delta disks).
With these patches, vmfsSparse files only work if opened through the
descriptor file. Data files without descriptor files, as far as I
could understand, are not supported by ESX.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Fam Zheng <famz@redhat.com>
--
v2: Rebase to patch 01.
Change le64_to_cpu to le32_to_cpu.
Rename vmdk_open_vmdk3 to vmdk_open_vmfs_sparse, which represents the
current usage of this format.
Signed-off-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
VMDK3 header has the field l1dir_size, but vmdk_open_vmdk3 hardcoded the
value. This patch honors the header field.
And the L2 table size is 4096 according to VMDK spec[1], instead of
1 << 9 (512).
[1]:
http://www.vmware.com/support/developer/vddk/vmdk_50_technote.pdf?src=vmdk
Signed-off-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
This header check is common to VMDK3 and VMDK4, so move it into
vmdk_add_extent().
Signed-off-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
In 4146b46c42e0989cb5842e04d88ab6ccb1713a48 (block: Produce zeros when
protocols reading beyond end of file), we break qemu-iotests ./check
-qcow2 022. This happens because qcow2 temporarily sets ->growable = 1
for vmstate accesses (which are stored beyond the end of regular image
data).
We introduce the bs->zero_beyond_eof to allow qcow2_load_vmstate() to
disable ->zero_beyond_eof temporarily in addition to enable ->growable.
[Since the broken patch "block: Produce zeros when protocols reading
beyond end of file" has not been merged yet, I have applied this fix
*first* and will then apply the next patch to keep the tree bisectable.
-- Stefan]
Suggested-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Asias He <asias@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
By the time that qemu 1.7 will be released, enough time will have passed
since qemu 1.1, which is the first version to understand version 3
images, that changing the default shouldn't hurt many people any more
and the benefits of using the new format outweigh the pain.
qemu-iotests already runs with compat=1.1 by default.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
The .io_flush() handler no longer exists and has no users. Drop the
io_flush argument to aio_set_fd_handler() and related functions.
The AioFlushEventNotifierHandler and AioFlushHandler typedefs are no
longer used and are dropped too.
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
.io_flush() is no longer called so drop qemu_rbd_aio_flush_cb().
qemu_aio_count is unused now so drop it too.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
.io_flush() is no longer called so drop nbd_have_request(). We cannot
drop in_flight since it is still used by other block/nbd.c code.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
.io_flush() is no longer called so drop qemu_laio_completion_cb(). It
turns out that count is now unused so drop that too.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Since .io_flush() is no longer called we do not need
qemu_gluster_aio_flush_cb() anymore. It turns out that qemu_aio_count
is unused now and can be dropped.
Thanks to Bharata B Rao <bharata@linux.vnet.ibm.com> for catching a
build failure with CONFIG_GLUSTERFS_DISCARD, which has been fixed.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
.io_flush() is no longer called so drop curl_aio_flush(). The acb[]
array that the function checks is still used in other parts of
block/curl.c. Therefore we cannot remove acb[], it is needed.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
If a block driver has no file descriptors to monitor but there are still
active requests, it can return 1 from .io_flush(). This is used to spin
during synchronous I/O.
Stop relying on .io_flush() and instead check
QLIST_EMPTY(&bs->tracked_requests) to decide whether there are active
requests.
This is the first step in removing .io_flush() so that event loops no
longer need to have the concept of synchronous I/O. Eventually we may
be able to kill synchronous I/O completely by running everything in a
coroutine, but that is future work.
Note this patch moves bs->throttled_reqs initialization to bdrv_new() so
that bdrv_requests_pending(bs) can safely access it. In practice bs is
g_malloc0() so the memory is already zeroed but it's safer to initialize
the queue properly.
We also need to fix up block/stream.c:close_unused_images() to prevent
traversing a dangling pointer while it rearranges the backing file
chain. This is necessary since the new bdrv_drain_all() traverses the
backing file chain.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Most of the block layer is under the BSD license, thus it is reasonable
to license block/raw.c the same way. CCed people should ACK by replying
with a Signed-off-by line.
Cc: Christoph Hellwig <hch@lst.de>
Cc: Kevin Wolf <kwolf@redhat.com>
Cc: Anthony Liguori <aliguori@us.ibm.com>
Cc: Markus Armbruster <armbru@redhat.com>
Cc: Stefan Hajnoczi <stefanha@redhat.com>
Cc: Luiz Capitulino <lcapitulino@redhat.com>
Cc: Jeff Cody <jcody@redhat.com>
Cc: Peter Lieven <pl@kamp.de>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Peter Lieven <pl@kamp.de>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Jeff Cody <jcody@redhat.com>
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 1375251592-2537-2-git-send-email-pbonzini@redhat.com
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
num_gtes_per_gte is a historical typo, rename it to a more sensible
name. It means "number of GrainTableEntries per GrainTable".
Signed-off-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
We should never grow the stack beyond 1 MB, otherwise we'll fall off the
end. Thread stacks and coroutine stacks (1 MB) do not grow.
get_cluster_offset() allocates a big stack offset, it will fail for big
cluster images, change to heap allocated buffer.
Signed-off-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
L1 table size is calculated from capacity, granularity and l2 table
size. If capacity is too big or later two are too small, the L1 table
will be too big to allocate in memory. Limit it to a reasonable range.
Signed-off-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
header.num_gtes_per_gte determines size for L2 table. Check for too big
value before using it. Limit to 512M entries (2GB per one L2 table).
Signed-off-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Granularity is used to calculate the cluster size and allocate r/w
buffer. Check the value from image before using it, so we don't abort()
for unbounded memory allocation.
Signed-off-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
The size and offset fields are all non-negative values, use uint64_t for
them to avoid getting negative in memory value by int overflow.
Signed-off-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
It's best to make it consistent that all on disk structures are
QEMU_PACKED.
Signed-off-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Commit 3ac21627 changed the behaviour of bdrv_has_zero_init() to default
to 0. In the review for Sheepdog it turned out that enabling it is safe,
so that commit updated one BlockDriver definition of sheepdog to use
bdrv_has_zero_init_1, missed however that there are more BlockDrivers in
the driver. Fix these now.
Cc: Kevin Wolf <kwolf@redhat.com>
Cc: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Liu Yuan <namei.unix@gmail.com>
Reviewed-by: MORITA Kazutaka <morita.kazutaka@lab.ntt.co.jp>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
The comment was truncated. Add the missing parts, especially explain why
we need zero_dry_run.
Signed-off-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
The error on armv7hl was:
block/iscsi.c: In function ‘is_request_lun_aligned’:
block/iscsi.c:251:26: error: format ‘%ld’ expects argument of type ‘long int’, but argument 3 has type ‘int64_t’ [-Werror=format=]
iscsilun->block_size, sector_num, nb_sectors);
^
This also splits the long line to comply with qemu coding guidelines.
Signed-off-by: Richard W.M. Jones <rjones@redhat.com>
Reviewed-by: Stefan Weil <sw@weilnetz.de>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
'dprintf' is the name of a POSIX standard function so we should not be
stealing it for our debug macro. Rename to 'DPRINTF' (in line with
a number of other source files.)
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Andreas Färber <afaerber@suse.de>
Reviewed-by: MORITA Kazutaka <morita.kazutaka@lab.ntt.co.jp>
Acked-by: Richard Henderson <rth@twiddle.net>
Acked-by: Kevin Wolf <kwolf@redhat.com>
Message-id: 1375100199-13934-2-git-send-email-peter.maydell@linaro.org
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
# By Stefan Hajnoczi (4) and others
# Via Stefan Hajnoczi
* stefanha/block:
dataplane: refuse to start if device is already in use
dataplane: enable virtio-blk x-data-plane=on live migration
migration: fix spice migration
migration: notify migration state before starting thread
block: Repair the throttling code.
gluster: Add image resize support
Message-id: 1375112172-24863-1-git-send-email-stefanha@redhat.com
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Implement .bdrv_truncate in GlusterFS block driver so that GlusterFS backend
can support image resizing.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Bharata B Rao <bharata@linux.vnet.ibm.com>
Tested-by: Bharata B Rao <bharata@linux.vnet.ibm.com>
Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
All these typos were found by codespell.
sould -> should
emperical -> empirical
intialization -> initialization
successfuly -> successfully
gaurantee -> guarantee
Fix also another error (before before) in the same context.
Signed-off-by: Stefan Weil <sw@weilnetz.de>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
This patch adds sync-modes to the drive-backup interface and
implements the FULL, NONE and TOP modes of synchronization.
FULL performs as before copying the entire contents of the drive
while preserving the point-in-time using CoW.
NONE only copies new writes to the target drive.
TOP copies changes to the topmost drive image and preserves the
point-in-time using CoW.
For sync mode TOP are creating a new target image using the same backing
file as the original disk image. Then any new data that has been laid
on top of it since creation is copied in the main backup_run() loop.
There is an extra check in the 'TOP' case so that we don't bother to copy
all the data of the backing file as it already exists in the target.
This is where the bdrv_co_is_allocated() is used to determine if the
data exists in the topmost layer or below.
Also any new data being written is intercepted via the write_notifier
hook which ends up calling backup_do_cow() to copy old data out before
it gets overwritten.
For mode 'NONE' we create the new target image and only copy in the
original data from the disk image starting from the time the call was
made. This preserves the point in time data by only copying the parts
that are *going to change* to the target image. This way we can
reconstruct the final image by checking to see if the given block exists
in the new target image first, and if it does not, you can get it from
the original image. This is basically an optimization allowing you to
do point-in-time snapshots with low overhead vs the 'FULL' version.
Since there is no old data to copy out the loop in backup_run() for the
NONE case just calls qemu_coroutine_yield() which only wakes up after
an event (usually cancel in this case). The rest is handled by the
before_write notifier which again calls backup_do_cow() to write out
the old data so it can be preserved.
Signed-off-by: Ian Main <imain@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This is what QMP wants to use. The options haven't been enabled in any
release yet, so we're still free to change them.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
s->qcow and s->qcow_filename are allocated but not freed on error. Fix the
possible leaks, remove unnecessary check for bdrv_new(), propagate ret code of
bdrv_create() and also the one of enable_write_target().
Signed-off-by: Fam Zheng <famz@redhat.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Implement bdrv_aio_discard for gluster.
Signed-off-by: Bharata B Rao <bharata@linux.vnet.ibm.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
if the blocksize of an iSCSI LUN is bigger than the BDRV_SECTOR_SIZE
it is possible that sector_num or nb_sectors are not correctly
aligned.
to avoid corruption we fail requests which are misaligned.
Signed-off-by: Peter Lieven <pl@kamp.de>
Cc: qemu-stable@nongnu.org
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
this hask is not working (anymore). support for misaligned offsets should
be handled at the block layer.
Signed-off-by: Peter Lieven <pl@kamp.de>
Cc: qemu-stable@nongnu.org
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
the -ENOPSC case did not work due to the missing goto.
Reported-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Peter Lieven <pl@kamp.de>
Cc: qemu-stable@nongnu.org
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Don't assume that SG_IO is always invoked with a simple buffer,
check the iovec_count and if it is >= 1 then we need to pass an array
of iovectors to libiscsi instead of just a plain buffer.
Signed-off-by: Ronnie Sahlberg <ronniesahlberg@gmail.com>
Cc: qemu-stable@nongnu.org
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
One of the major reasons for doing something new for -blockdev and
blockdev-add was that the old block layer code parses filenames instead
of just taking them literally. So we should really leave it untouched
when it's passing using the new interfaces (like -drive
file.filename=...).
This allows opening relative file names that contain a colon.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
CURL driver requests partial data from server on guest IO req. For HTTP
and HTTPS, it uses "Range: ***" in requests, and this will not work if
server not accepting range. This patch does this check when open.
* Removed curl_size_cb, which is not used: On one hand it's registered to
libcurl as CURLOPT_WRITEFUNCTION, instead of CURLOPT_HEADERFUNCTION,
which will get called with *data*, not *header*. On the other hand the
s->len is assigned unconditionally later.
In this gone function, the sscanf for "Content-Length: %zd", on
(void *)ptr, which is not guaranteed to be zero-terminated, is
potentially a security bug. So this patch fixes it as a side-effect. The
bug is reported as: https://bugs.launchpad.net/qemu/+bug/1188943
(Note the bug is marked "private" so you might not be able to see it)
* Introduced curl_header_cb, which is used to parse header and mark the
server as accepting range if "Accept-Ranges: bytes" line is seen from
response header. If protocol is HTTP or HTTPS, but server response has
no not this support, refuse to open this URL.
Note that python builtin module SimpleHTTPServer is an example of not
supporting range, if you need to test this driver, get a better server
or use internet URLs.
Signed-off-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Depending on the subformat, has_zero_init queries underlying storage for
flat extent. If it has a flat extent and its underlying storage doesn't
have zero init, return 0. Otherwise return 1.
Aligns the operator assignments.
Signed-off-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
.has_zero_init defaults to 1 for all formats and protocols.
this is a dangerous default since this means that all
new added drivers need to manually overwrite it to 0 if
they do not ensure that a device is zero initialized
after bdrv_create().
if a driver needs to explicitly set this value to
1 its easier to verify the correctness in the review process.
during review of the existing drivers it turned out
that ssh and gluster had a wrong default of 1.
both protocols support host_devices as backend
which are not by default zero initialized. this
wrong assumption will lead to possible corruption
if qemu-img convert is used to write to such a backend.
vpc and vmdk also defaulted to 1 altough they support
fixed respectively flat extends. this has to be addresses
in separate patches. both formats as well as the mentioned
ssh and gluster are turned to the default of 0 with this
patch for safety.
a similar problem with the wrong default existed for
iscsi most likely because the driver developer did
oversee the default value of 1.
Signed-off-by: Peter Lieven <pl@kamp.de>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Depending on the subformat, has_zero_init on VHD must behave like raw
and query the underlying storage (fixed) or like other sparse formats
that can always return 1 (dynamic, differencing).
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
When creating image with backing file, the driver tries to calculate the
relative path from created image file to backing file, but the path
computation is incorrect. e.g.:
$ qemu-img create -f vmdk -b vmdk-data-disk.vmdk vmdk-data-snapshot1
Formatting 'vmdk-data-snapshot1', fmt=vmdk size=10737418240
backing_file='vmdk-data-disk.vmdk' compat6=off zeroed_grain=off
$ qemu-img info vmdk-data-snapshot1
image: vmdk-data-snapshot1
file format: vmdk
virtual size: 10G (10737418240 bytes)
disk size: 12K
-> backing file: disk.vmdk
The common part in file names, "vmdk-data-", is incorrectly forgotten by
relative_path(). As the VMDK specification has no restriction on
parentNameHint to be relative path, we simply remove this by using the
backing_file option.
Cc: qemu-stable@nongnu.org
Signed-off-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
GlusterFS volumes can be backed by block devices, in which case
bdrv_create() doesn't make sure that the image is zeroed out. It is
currently not possibly to detect whether a given image is backed by a
file or a block device, and incorrectly assuming that it is zeroed
corrupts images during qemu-img convert, so let's err on the side of
caution and always return 0.
Cc: qemu-stable@nongnu.org
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
If the remote is a regular file, set it to true (ie. reads of
uninitialized areas in a newly created file will return zeroes).
If we can't prove that, return false (a safe default).
Tested by adding a debugging print statement [not part of this commit]
and creating a remote file and a remote block device:
$ ./qemu-img create ssh://localhost/tmp/new 100M
Formatting 'ssh://localhost/tmp/new', fmt=raw size=104857600
filename ssh://localhost/tmp/new: has_zero_init = 1
$ sudo lvcreate -L 1G -n tmp /dev/fedora
Logical volume "tmp" created
$ ./qemu-img create ssh://localhost/dev/fedora/tmp 1G
Formatting 'ssh://localhost/dev/fedora/tmp', fmt=raw size=1073741824
filename ssh://localhost/dev/fedora/tmp: has_zero_init = 0
Cc: Kevin Wolf <kwolf@redhat.com>
Cc: qemu-stable@nongnu.org
Signed-off-by: Richard W.M. Jones <rjones@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
backup_start() creates a block job that copies a point-in-time snapshot
of a block device to a target block device.
We call backup_do_cow() for each write during backup. That function
reads the original data from the block device before it gets
overwritten. The data is then written to the target device.
Currently backup cluster size is hardcoded to 65536 bytes.
[I made a number of changes to Dietmar's original patch and folded them
in to make code review easy. Here is the full list:
* Drop BackupDumpFunc interface in favor of a target block device
* Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes()
* Use 0 delay instead of 1us, like other block jobs
* Unify creation/start functions into backup_start()
* Simplify cleanup, free bitmap in backup_run() instead of cb
* function
* Use HBitmap to avoid duplicating bitmap code
* Use bdrv_getlength() instead of accessing ->total_sectors
* directly
* Delete the backup.h header file, it is no longer necessary
* Move ./backup.c to block/backup.c
* Remove #ifdefed out code
* Coding style and whitespace cleanups
* Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks
* Keep our own in-flight CowRequest list instead of using block.c
tracked requests. This means a little code duplication but is much
simpler than trying to share the tracked requests list and use the
backup block size.
* Add on_source_error and on_target_error error handling.
* Use trace events instead of DPRINTF()
-- stefanha]
Signed-off-by: Dietmar Maurer <dietmar@proxmox.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
The raw-posix driver has code to provide a /dev/cdrom on OS X even
though it doesn't really exist. However, since commit c66a6157 the real
filename is dismissed after finding it, so opening /dev/cdrom fails.
Put the filename back into the options QDict to make this work again.
Cc: qemu-stable@nongnu.org
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Refuse to open higher version for safety.
Although we try to be compatible with published VMDK spec, VMware has
newer version from ESXi 5.1 exported OVF/OVA, which we have no knowledge
what's changed in it. And it is very likely to have more new versions in
the future, so it's not safe to open them blindly.
Signed-off-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
This optimises the discard operation for freed clusters by batching
discard requests (both snapshot deletion and bdrv_discard end up
updating the refcounts cluster by cluster).
Note that we don't discard asynchronously, but keep s->lock held. This
is to avoid that a freed cluster is reallocated and written to while the
discard is still in flight.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Deleted snapshots are discarded in the image file by default, discard
requests take their default from the -drive discard=... option and other
places that free clusters must always be enabled explicitly.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
This adds a refcount update reason to all callers of update_refcounts(),
so that a follow-up patch can use this information to decide whether
clusters that reach a refcount of 0 should be discarded in the image
file.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
# By Paolo Bonzini (3) and others
# Via Paolo Bonzini
* bonzini/scsi-next:
iscsi: reorganize iscsi_readcapacity_sync
iscsi: simplify freeing of tasks
vhost-scsi: fix k->set_guest_notifiers() NULL dereference
scsi-disk: scsi-block device for scsi pass-through should not be removable
scsi-generic: check the return value of bdrv_aio_ioctl in execute_command
scsi-generic: fix sign extension of READ CAPACITY(10) data
scsi: reset cdrom tray statuses on scsi_disk_reset
Message-id: 1371565016-2643-1-git-send-email-pbonzini@redhat.com
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Avoid the goto, and use the same retry logic for the 10- and 16-
byte versions.
Cc: qemu-stable@nongnu.org
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Always free them in the iscsi_aio_*_acb functions and remove the
checks in their callers. Remove ifs when the task struct was
previously dereferenced (spotted by Coverity).
Cc: qemu-stable@nongnu.org
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Otherwise they would get passed to getaddrinfo and fail with:
address resolution failed for [::1]🔢 Name or service not known
(Broken by commit v1.4.0-736-gf17c90b)
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Cc: qemu-stable@nongnu.org
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
# By Luiz Capitulino
# Via Luiz Capitulino
* luiz/queue/qmp:
qerror: drop QERR_OPEN_FILE_FAILED macro
block: bdrv_reopen_prepare(): don't use QERR_OPEN_FILE_FAILED
savevm: qmp_xen_save_devices_state(): use error_setg_file_open()
dump: qmp_dump_guest_memory(): use error_setg_file_open()
cpus: use error_setg_file_open()
blockdev: use error_setg_file_open()
block: mirror_complete(): use error_setg_file_open()
rng-random: use error_setg_file_open()
error: add error_setg_file_open() helper
Message-id: 1371484631-29510-1-git-send-email-lcapitulino@redhat.com
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
the hard-coded 2k buffer on the stack won't allow reading big descriptor
files which can be generated when storing big images. For example 500G
vmdk splitted to 2G chunks.
Signed-off-by: Evgeny Budilovsky <evgeny.budilovsky@ravellosystems.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
(Found by Kamil Dudka)
Signed-off-by: Richard W.M. Jones <rjones@redhat.com>
Cc: Michael Tokarev <mjt@tls.msk.ru>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Remember to byteswap VMDK4Header.desc_offset on big-endian machines.
Cc: qemu-stable@nongnu.org
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Just call sd_create_branch() in the snapshot_goto to rollback the image is good
enough. With this patch, 'loadvm' process for sheepdog is modified:
Suppose we have a snapshot chain A --> B --> C, we do 'loadvm A' so as to get
a new chain,
A --> B
|
V
C1
in the old code:
1 reload inode of A (in snapshot_goto)
2 read vmstate via A's vdi_id (loadvm_state)
3 delete C and create C1, reload inode of C1 (sd_create_branch on write)
with this patch applied:
1 reload inode of A, delete C and create C1 (in snapshot_goto)
2 read vmstate via C1's parent, that is A's vdi_id (loadvm_state)
This will fix the possible bug that QEMU exit between 2 and 3 in the old code
Cc: qemu-devel@nongnu.org
Cc: MORITA Kazutaka <morita.kazutaka@lab.ntt.co.jp>
Cc: Kevin Wolf <kwolf@redhat.com>
Cc: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Liu Yuan <namei.unix@gmail.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This is an old and obvious bug. We should pass snapshot_id to the
tag. Or simple command like 'qemu-img snapshot -a tag sheepdog:image' will fail
Cc: qemu-devel@nongnu.org
Cc: MORITA Kazutaka <morita.kazutaka@lab.ntt.co.jp>
Cc: Kevin Wolf <kwolf@redhat.com>
Cc: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Liu Yuan <namei.unix@gmail.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Now image info will be retrieved as an embbed json object inside
BlockDeviceInfo, backing chain info and all related internal snapshot
info can be got in the enhanced recursive structure of ImageInfo. New
recursive member *backing-image is added to reflect the backing chain
status.
Signed-off-by: Wenchao Xia <xiawenc@linux.vnet.ibm.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
This patch adds function bdrv_query_image_info(), which will
retrieve image info in qmp object format. The implementation is
based on the code moved from qemu-img.c, but uses block layer
function to get snapshot info.
Signed-off-by: Wenchao Xia <xiawenc@linux.vnet.ibm.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
This patch adds function bdrv_query_snapshot_info_list(), which will
retrieve snapshot info of an image in qmp object format. The implementation
is based on the code moved from qemu-img.c with modification to fit more
for qmp based block layer API.
Signed-off-by: Wenchao Xia <xiawenc@linux.vnet.ibm.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
bdrv_snapshot_dump() and bdrv_image_info_dump() do not dump to a buffer now,
some internal buffers are still used for format control, which have no
chance to be truncated. As a result, these two functions have no more issue
of truncation, and they can be used by both qemu and qemu-img with correct
parameter specified.
Signed-off-by: Wenchao Xia <xiawenc@linux.vnet.ibm.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This patch is a pure code move patch, except following modification:
1 get_human_readable_size() is changed to static function.
2 dump_human_image_info() is renamed to bdrv_image_info_dump().
3 in qmp_query_block() and qmp_query_blockstats, use bdrv_next(bs)
instead of direct traverse of global array 'bdrv_states'.
4 collect_snapshots() and collect_image_info() are renamed, unused parameter
*fmt in collect_image_info() is removed.
5 code style fix.
To avoid conflict and tip better, macro in header file is BLOCK_QAPI_H
instead of QAPI_H. Now block.h and snapshot.h are at the same level in
include path, block_int.h and qapi.h will both include them.
Signed-off-by: Wenchao Xia <xiawenc@linux.vnet.ibm.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
All snapshot related code, except bdrv_snapshot_dump() and
bdrv_is_snapshot(), is moved to block/snapshot.c. bdrv_snapshot_dump()
will be moved to another file later. bdrv_is_snapshot() is not related
with internal snapshot. It also fixes small code style errors reported
by check script.
Signed-off-by: Wenchao Xia <xiawenc@linux.vnet.ibm.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This patch is used to remove twice include of "qemu-common.h" in
block/win32-aio.c
Signed-off-by: Qiao Nuohan <qiaonuohan@cn.fujitsu.com>
Reviewed-by: Stefan Weil <sw@weilnetz.de>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
This catches the situation that is described in the bug report at
https://bugs.launchpad.net/qemu/+bug/865518 and goes like this:
$ qemu-img create -f qcow2 huge.qcow2 $((1024*1024))T
Formatting 'huge.qcow2', fmt=qcow2 size=1152921504606846976 encryption=off cluster_size=65536 lazy_refcounts=off
$ qemu-io /tmp/huge.qcow2 -c "write $((1024*1024*1024*1024*1024*1024 - 1024)) 512"
Segmentation fault
With this patch applied the segfault will be avoided, however the case
will still fail, though gracefully:
$ qemu-img create -f qcow2 /tmp/huge.qcow2 $((1024*1024))T
Formatting 'huge.qcow2', fmt=qcow2 size=1152921504606846976 encryption=off cluster_size=65536 lazy_refcounts=off
qemu-img: The image size is too large for file format 'qcow2'
Note that even long before these overflow checks kick in, you get
insanely high memory usage (up to INT_MAX * sizeof(uint64_t) = 16 GB for
the L1 table), so with somewhat smaller image sizes you'll probably see
qemu aborting for a failed g_malloc().
If you need huge image sizes, you should increase the cluster size to
the maximum of 2 MB in order to get higher limits.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Use special offset to write zeroes efficiently, when zeroed-grain GTE is
available. If zero-write an allocated cluster, cluster is leaked because
its offset pointer is overwritten by "0x1".
Signed-off-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Previously VmdkMetaData.offset is stored little endian while other
fields are cpu endian. This changes offset to cpu endian and convert
before writing to image.
Signed-off-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Add image create option "zeroed-grain" to enable zeroed-grain GTE
feature of vmdk sparse extents. When this option is on, header version
of newly created extent will be 2 and VMDK4_FLAG_ZERO_GRAIN flag bit
will be set.
Signed-off-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Introduced support for zeroed-grain GTE, as specified in Virtual Disk
Format 5.0[1].
Recent VMware hosted platform products support a new “zeroed‐grain”
grain table entry (GTE). The zeroed‐grain GTE returns all zeros on
read. In other words, the zeroed‐grain GTE indicates that a grain
in the child disk is zero‐filled but does not actually occupy space
in storage. A sparse extent with zeroed‐grain GTE has the following
in its header:
* SparseExtentHeader.version = 2
* SparseExtentHeader.flags has bit 2 set
Other than the new flag and the possibly zeroed‐grain GTE, version 2
sparse extents are identical to version 1. Also, a zeroed‐grain GTE
has value 0x1 in the GT table.
[1] Virtual Disk Format 5.0, http://www.vmware.com/support/developer/vddk/vmdk_50_technote.pdf?src=vmdk
Signed-off-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Internal routines in vmdk.c previously return -1 on error and 0 on
success. More return values are useful for future changes such as
zeroed-grain GTE. Change all the magic `return 0` and `return -1` to
macro names:
* VMDK_OK 0
* VMDK_ERROR (-1)
* VMDK_UNALLOC (-2)
* VMDK_ZEROED (-3)
Signed-off-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
This adds in read-only support to the VHDX image format. This supports
reads for fixed-size, and dynamic sized VHDX images.
Differencing files are still unsupported.
The image must be opened without BDRV_O_RDWR set, because we do not
yet update the headers. I.e., pass 'readonly=on' in the drive image
options from the QEMU commandline.
Signed-off-by: Jeff Cody <jcody@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
This is the initial block driver framework for VHDX image support
(i.e. Hyper-V image file formats), that supports opening VHDX files, and
parsing the headers.
This commit does not yet enable:
- reading
- writing
- updating the header
- differencing files (images with parents)
- log replay / dirty logs (only clean images)
This is based on Microsoft's VHDX specification:
"VHDX Format Specification v0.95", published 4/12/2012
https://www.microsoft.com/en-us/download/details.aspx?id=29681
Signed-off-by: Jeff Cody <jcody@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
This is based on Microsoft's VHDX specification:
"VHDX Format Specification v0.95", published 4/12/2012
https://www.microsoft.com/en-us/download/details.aspx?id=29681
These structures define the various header, metadata, and other
block structures defined in the VHDX specification.
Signed-off-by: Jeff Cody <jcody@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Currently the 'loadvm' opertaion works as following:
1. switch to the snapshot
2. mark current working VDI as a snapshot
3. rely on sd_create_branch to create a new working VDI based on the snapshot
This works not the same as other format as QCOW2. For e.g,
qemu > savevm # get a live snapshot snap1
qemu > savevm # snap2
qemu > loadvm 1 # This will steally create snap3 of the working VDI
Which will result in following snapshot chain:
base <-- snap1 <-- snap2 <-- snap3
^
|
working VDI
snap3 was unnecessarily created and might be annoying users.
This patch discard the unnecessary 'snap3' creation. and implement
rollback(loadvm) operation to the specified snapshot by
1. switch to the snapshot
2. delete working VDI
3. rely on sd_create_branch to create a new working VDI based on the snapshot
The snapshot chain for above example will be:
base <-- snap1 <-- snap2
^
|
working VDI
Cc: qemu-devel@nongnu.org
Cc: MORITA Kazutaka <morita.kazutaka@lab.ntt.co.jp>
Cc: Kevin Wolf <kwolf@redhat.com>
Cc: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Liu Yuan <tailai.ly@taobao.com>
Reviewed-by: MORITA Kazutaka <morita.kazutaka@lab.ntt.co.jp>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
When a snapshot is taken from out side of qemu (e.g. qemu-img
snapshot), write requests to the current vdi return SD_RES_READONLY.
In this case, the sheepdog block driver needs to update the current
inode to the latest one and resend the write requests.
Signed-off-by: MORITA Kazutaka <morita.kazutaka@lab.ntt.co.jp>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
This adds a helper function to update the current inode state with the
specified vdi object.
Signed-off-by: MORITA Kazutaka <morita.kazutaka@lab.ntt.co.jp>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Sheepdog returns SD_RES_READONLY when qemu sends write requests to the
snapshot vdi. This adds the result code and makes sd_strerror() print
its error reason.
Signed-off-by: MORITA Kazutaka <morita.kazutaka@lab.ntt.co.jp>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
This makes 'filename' and 'tag' constant variables, and renames
'for_snapshot' to 'lock' to clear how it works.
Signed-off-by: MORITA Kazutaka <morita.kazutaka@lab.ntt.co.jp>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Commit a9ccedc3 frees the QemuOpts for the driver-specific options
immediately, even though it still needs the filename string that is
contained there. This doesn't work. Move the deletion of the QemuOpts to
the end of the function where its content isn't needed any more.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Cc: MORITA Kazutaka <morita.kazutaka@lab.ntt.co.jp>
Cc: Kevin Wolf <kwolf@redhat.com>
Cc: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Liu Yuan <tailai.ly@taobao.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Cc: MORITA Kazutaka <morita.kazutaka@lab.ntt.co.jp>
Cc: Kevin Wolf <kwolf@redhat.com>
Cc: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Liu Yuan <tailai.ly@taobao.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
The 'TRIM' command from VM that is to release underlying data storage for
better thin-provision is already supported by the Sheepdog.
This patch adds the TRIM support at QEMU part.
For older Sheepdog that doesn't support it, we return 0(success) to upper layer.
Cc: MORITA Kazutaka <morita.kazutaka@lab.ntt.co.jp>
Cc: Kevin Wolf <kwolf@redhat.com>
Cc: Stefan Hajnoczi <stefanha@redhat.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Liu Yuan <tailai.ly@taobao.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
# By Kevin Wolf (16) and Stefan Hajnoczi (4)
# Via Kevin Wolf
* kwolf/for-anthony:
qemu-iotests: add 053 unaligned compressed image size test
block: Allow overriding backing.file.filename
block: Remove filename parameter from .bdrv_file_open()
vvfat: Use bdrv_open options instead of filename
sheepdog: Use bdrv_open options instead of filename
rbd: Use bdrv_open options instead of filename
iscsi: Use bdrv_open options instead of filename
gluster: Use bdrv_open options instead of filename
curl: Use bdrv_open options instead of filename
blkverify: Use bdrv_open options instead of filename
blkdebug: Use bdrv_open options instead of filename
raw-win32: Use bdrv_open options instead of filename
raw-posix: Use bdrv_open options instead of filename
block: Enable filename option
block: Add driver-specific options for backing files
block: Fail gracefully when using a format driver on protocol level
qemu-iotests: Fix _filter_qemu
qemu-img: do not zero-pad the compressed write buffer
qcow: allow sub-cluster compressed write to last cluster
qcow2: allow sub-cluster compressed write to last cluster
Message-id: 1366630294-18984-1-git-send-email-kwolf@redhat.com
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
# By Stefan Hajnoczi
# Via Paolo Bonzini
* bonzini/nbd-next:
nbd: set TCP_NODELAY
nbd: use TCP_CORK in nbd_co_send_request()
nbd: unlock mutex in nbd_co_send_request() error path
Message-id: 1366381830-11267-1-git-send-email-pbonzini@redhat.com
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This is only to convert the internal interface that is used for passing
the "filename" to be parsed, but converting to actual fine grained
options is left for another day, as it doesn't look trivial.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
This is only to convert the internal interface that is used for passing
the "filename" to be parsed, but converting to actual fine grained
options is left for another day, as it doesn't look trivial.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
This is only to convert the internal interface that is used for passing
the "filename" to be parsed, but converting to actual fine grained
options is left for another day, as it doesn't look trivial.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
This is only to convert the internal interface that is used for passing
the "filename" to be parsed, but converting to actual fine grained
options is left for another day, as it doesn't look trivial.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
As a bonus, going through the QemuOpts QEMU_OPT_SIZE parser for the
readahead option gives us proper error reporting that the previous use
of atoi() lacked.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Options starting in "backing." are passed to the backing file now. If
you don't need to specify the filename for the backing file, you can add
it on the command line instead of in the image file:
$ qemu-nbd -t /tmp/test.img
$ qemu-img create -f qcow2 empty.qcow2 1G
$ qemu-system-x86_64 -drive file=empty.qcow2,backing.file.driver=nbd,\
backing.file.host=localhost
Note that this doesn't override the backing filename from the image. If
the image has one, this will fail because NBD doesn't want the options
and a filename at the same time.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Compression in qcow requires image length to be a multiple of the
cluster size. Lift this requirement by zero-padding the final cluster
when necessary. The virtual disk size is still not cluster-aligned, so
the guest cannot access the zero sectors.
Note that this is almost identical to the qcow2 version of this code.
qcow2's compression code is drawn from qcow.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Compression in qcow2 requires image length to be a multiple of the
cluster size. Lift this requirement by zero-padding the final cluster
when necessary. The virtual disk size is still not cluster-aligned, so
the guest cannot access the zero sectors.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Now gcc will check whether format string and variable arguments match.
Signed-off-by: Stefan Weil <sw@weilnetz.de>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Disable the Nagle algorithm to reduce latency. Note this means we must
also use TCP_CORK when sending header followed by payload to avoid
fragmenting lots of little packets. The previous patch took care of
that.
Suggested-by: Nick Thomas <nick@bytemark.co.uk>
Tested-by: Nick Thomas <nick@bytemark.co.uk>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Use TCP_CORK to defer packet transmission until both the header and the
payload have been written.
Suggested-by: Nick Thomas <nick@bytemark.co.uk>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The existing bdrv_co_flush_to_disk implementation uses rbd_flush(),
which is sychronous and causes the main qemu thread to block until it
is complete. This results in unresponsiveness and extra latency for
the guest.
Fix this by using an asynchronous version of flush. This was added to
librbd with a special #define to indicate its presence, since it will
be backported to stable versions. Thus, there is no need to check the
version of librbd.
Implement this as bdrv_aio_flush, since it matches other aio functions
in the rbd block driver, and leave out bdrv_co_flush_to_disk when the
asynchronous version is available.
Reported-by: Oliver Francke <oliver@filoo.de>
Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
libssh2_sftp_fsync is an extension to libssh2 to support fsync(2) over
sftp, which is itself an extension of OpenSSH.
If both libssh2 and the ssh daemon support it, this will allow
bdrv_flush_to_disk to commit changes through to disk on the remote
server.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
qemu-system-x86_64 -drive file=ssh://hostname/some/image
QEMU will ssh into 'hostname' and open '/some/image' which is made
available as a standard block device.
You can specify a username (ssh://user@host/...) and/or a port number
(ssh://host:port/...). You can also use an alternate syntax using
properties (file.user, file.host, file.port, file.path).
Current limitations:
- Authentication must be done without passwords or passphrases, using
ssh-agent. Other authentication methods are not supported.
- Uses a single connection, instead of concurrent AIO with multiple
SSH connections.
This is implemented using libssh2 on the client side. The server just
requires a regular ssh daemon with sftp-server support. Most ssh
daemons on Unix/Linux systems will work out of the box.
Signed-off-by: Richard W.M. Jones <rjones@redhat.com>
Cc: Stefan Hajnoczi <stefanha@gmail.com>
Cc: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Directly pass the QEMUIOVector on instead of linearising it.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Move aes.h from include/block to include/qemu to show it can be reused
by other subsystems.
Cc: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Edgar E. Iglesias <edgar.iglesias@gmail.com>
Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
Many of these should be cleaned up with proper qdev-/QOM-ification.
Right now there are many catch-all headers in include/hw/ARCH depending
on cpu.h, and this makes it necessary to compile these files per-target.
However, fixing this does not belong in these patches.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
It ignored the error code, and at least the 'goto fail' is obvious
nonsense as it creates an endless loop (if the next attempt doesn't
magically succeed) and leaves the in-memory L1 table in big-endian
instead of converting it back.
In error cases, there's no point in writing an updated L1 table, so
skip this part for them.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
The fcntl(fd, F_SETFL, O_NONBLOCK) flag is not specific to sockets.
Rename to qemu_set_nonblock() just like qemu_set_cloexec().
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Instead of just checking once in exactly this order if there are
dependendies, non-COW clusters and new allocation, this starts looping
around these. This way we can, for example, gather non-COW clusters after
new allocations as long as the host cluster offsets stay contiguous.
Once handle_dependencies() is extended so that COW areas of in-flight
allocations can be overwritten, this allows to continue with gathering
other clusters (we wouldn't be able to do that without this change
because we would have missed a possible second dependency in one of the
next clusters).
This means that in the typical sequential write case, we can combine the
COW overwrite of one cluster with the allocation of the next cluster as
soon as something like Delayed COW gets actually implemented. It is only
by avoiding splitting requests this way that Delayed COW actually starts
improving performance noticably.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
This patch is mainly to separate the indentation change from the
semantic changes. All that really changes here is that everything moves
into a while loop, all 'goto done' become 'break' and at the end of the
loop a new 'break is inserted.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Instead of expecting a single l2meta, have a list of them. This allows
to still have a single I/O request for the guest data, even though
multiple l2meta may be needed in order to describe both a COW overwrite
and a new cluster allocation (typical sequential write case).
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
This gets rid of the nb_clusters and keep_clusters and the associated
complicated calculations. Just advance the number of bytes that have
been processed and everything is fine.
This patch advances the variables even after the last operation even
though they aren't used any more afterwards to make things look more
uniform. A later patch will turn the whole thing into a loop and then
it actually starts making sense.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
This makes handle_alloc() and handle_copied() return byte-granularity
host offsets instead of returning always the cluster start. This is
required so that qcow2_alloc_cluster_offset() can stop aligning
everything to cluster boundaries.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Look only for clusters that start at a given physical offset.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Now *bytes is used to return the length of the area that can be written
to without performing an allocation or COW.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
