A discard request deallocates the selected clusters so they read back
as zeroes. This is done by clearing the cluster offset field and
setting QCOW_OFLAG_ZERO in the L2 entry.
This flag is however only supported when qcow_version >= 3. In older
images the cluster is simply deallocated, exposing any possible stale
data from the backing file.
Since discard is an advisory operation it's safer to simply forbid it
in this scenario.
Note that we are adding this check to qcow2_co_pdiscard() and not to
qcow2_cluster_discard() or discard_in_l2_slice() because the last
two are also used by qcow2_snapshot_create() to discard the clusters
used by the VM state. In this case there's no risk of exposing stale
data to the guest and we really want that the clusters are always
discarded.
Signed-off-by: Alberto Garcia <berto@igalia.com>
Message-Id: <20200331114345.29993-1-berto@igalia.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
As the feature name table can be quite large (over 9k if all 64 bits
of all three feature fields have names; a mere 8 features leaves only
8 bytes for a backing file name in a 512-byte cluster), it is unwise
to emit this optional header in images with small cluster sizes.
Update iotest 036 to skip running on small cluster sizes; meanwhile,
note that iotest 061 never passed on alternative cluster sizes
(however, I limited this patch to tests with output affected by adding
feature names, rather than auditing for other tests that are not
robust to alternative cluster sizes).
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
Message-Id: <20200324174233.1622067-4-eblake@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
The feature table is supposed to advertise the name of all feature
bits that we support; however, we forgot to update the table for
autoclear bits. While at it, move the table to read-only memory in
code, and tweak the qcow2 spec to name the second autoclear bit.
Update iotests that are affected by the longer header length.
Fixes: 88ddffae
Fixes: 93c24936
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20200324174233.1622067-3-eblake@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
Various trivial typos noticed while working on this file.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
Message-Id: <20200324174233.1622067-2-eblake@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
This will allow the reuse of a single generic .bdrv_co_create
implementation for several drivers.
No functional changes.
Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
Message-Id: <20200326011218.29230-2-mlevitsk@redhat.com>
Reviewed-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: Max Reitz <mreitz@redhat.com>
data_file being NULL doesn't seem to be a correct state, but it's
better than dead pointer and simpler to debug.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20200316060631.30052-3-vsementsov@virtuozzo.com>
Reviewed-by: John Snow <jsnow@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
If we fail to get bitmap info, we must not leak the encryption info.
Fixes: b8968c875f
Fixes: Coverity CID 1421894
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-Id: <20200320183620.1112123-1-eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Andrey Shinkevich <andrey.shinkevich@virtuozzo.com>
Tested-by: Andrey Shinkevich <andrey.shinkevich@virtuozzo.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
'crypto_opts' forgot to free in qcow2_close(), this patch fix the bellow leak stack:
Direct leak of 24 byte(s) in 1 object(s) allocated from:
#0 0x7f0edd81f970 in __interceptor_calloc (/lib64/libasan.so.5+0xef970)
#1 0x7f0edc6d149d in g_malloc0 (/lib64/libglib-2.0.so.0+0x5249d)
#2 0x55d7eaede63d in qobject_input_start_struct /mnt/sdb/qemu-new/qemu_test/qemu/qapi/qobject-input-visitor.c:295
#3 0x55d7eaed78b8 in visit_start_struct /mnt/sdb/qemu-new/qemu_test/qemu/qapi/qapi-visit-core.c:49
#4 0x55d7eaf5140b in visit_type_QCryptoBlockOpenOptions qapi/qapi-visit-crypto.c:290
#5 0x55d7eae43af3 in block_crypto_open_opts_init /mnt/sdb/qemu-new/qemu_test/qemu/block/crypto.c:163
#6 0x55d7eacd2924 in qcow2_update_options_prepare /mnt/sdb/qemu-new/qemu_test/qemu/block/qcow2.c:1148
#7 0x55d7eacd33f7 in qcow2_update_options /mnt/sdb/qemu-new/qemu_test/qemu/block/qcow2.c:1232
#8 0x55d7eacd9680 in qcow2_do_open /mnt/sdb/qemu-new/qemu_test/qemu/block/qcow2.c:1512
#9 0x55d7eacdc55e in qcow2_open_entry /mnt/sdb/qemu-new/qemu_test/qemu/block/qcow2.c:1792
#10 0x55d7eacdc8fe in qcow2_open /mnt/sdb/qemu-new/qemu_test/qemu/block/qcow2.c:1819
#11 0x55d7eac3742d in bdrv_open_driver /mnt/sdb/qemu-new/qemu_test/qemu/block.c:1317
#12 0x55d7eac3e990 in bdrv_open_common /mnt/sdb/qemu-new/qemu_test/qemu/block.c:1575
#13 0x55d7eac4442c in bdrv_open_inherit /mnt/sdb/qemu-new/qemu_test/qemu/block.c:3126
#14 0x55d7eac45c3f in bdrv_open /mnt/sdb/qemu-new/qemu_test/qemu/block.c:3219
#15 0x55d7ead8e8a4 in blk_new_open /mnt/sdb/qemu-new/qemu_test/qemu/block/block-backend.c:397
#16 0x55d7eacde74c in qcow2_co_create /mnt/sdb/qemu-new/qemu_test/qemu/block/qcow2.c:3534
#17 0x55d7eacdfa6d in qcow2_co_create_opts /mnt/sdb/qemu-new/qemu_test/qemu/block/qcow2.c:3668
#18 0x55d7eac1c678 in bdrv_create_co_entry /mnt/sdb/qemu-new/qemu_test/qemu/block.c:485
#19 0x55d7eb0024d2 in coroutine_trampoline /mnt/sdb/qemu-new/qemu_test/qemu/util/coroutine-ucontext.c:115
Reported-by: Euler Robot <euler.robot@huawei.com>
Signed-off-by: Pan Nengyuan <pannengyuan@huawei.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-Id: <20200227012950.12256-2-pannengyuan@huawei.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
The qcow2 .bdrv_measure() code calculates the crypto payload offset.
This logic really belongs in crypto/block.c where it can be reused by
other image formats.
The "luks" block driver will need this same logic in order to implement
.bdrv_measure(), so extract the qcrypto_block_calculate_payload_offset()
function now.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-Id: <20200221112522.1497712-2-stefanha@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
The bitmap code requires writing the 'file' child when the qcow2 driver
is reopened in read-write mode.
If the 'file' child is being reopened due to a permissions change, the
modification is commited yet when qcow2_reopen_commit is called. This
means that any attempt to write the 'file' child will end with EBADFD
as the original fd was already closed.
Moving bitmap reopening to the new callback which is called after
permission modifications are commited fixes this as the file descriptor
will be replaced with the correct one.
The above problem manifests itself when reopening 'qcow2' format layer
which uses a 'file-posix' file child which was opened with the
'auto-read-only' property set.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Message-Id: <db118dbafe1955afbc0a18d3dd220931074ce349.1582893284.git.pkrempa@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
When initializing the LUKS header the size with default encryption
parameters will currently be 2068480 bytes. This is rounded up to
a multiple of the cluster size, 2081792, with 64k sectors. If the
end of the header is not the same as the end of the cluster we fill
the extra space with zeros. This was forgetting that not even the
space allocated for the header will be fully initialized, as we
only write key material for the first key slot. The space left
for the other 7 slots is never written to.
An optimization to the ref count checking code:
commit a5fff8d4b4 (refs/bisect/bad)
Author: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Date: Wed Feb 27 16:14:30 2019 +0300
qcow2-refcount: avoid eating RAM
made the assumption that every cluster which was allocated would
have at least some data written to it. This was violated by way
the LUKS header is only partially written, with much space simply
reserved for future use.
Depending on the cluster size this problem was masked by the
logic which wrote zeros between the end of the LUKS header and
the end of the cluster.
$ qemu-img create --object secret,id=cluster_encrypt0,data=123456 \
-f qcow2 -o cluster_size=2k,encrypt.iter-time=1,\
encrypt.format=luks,encrypt.key-secret=cluster_encrypt0 \
cluster_size_check.qcow2 100M
Formatting 'cluster_size_check.qcow2', fmt=qcow2 size=104857600
encrypt.format=luks encrypt.key-secret=cluster_encrypt0
encrypt.iter-time=1 cluster_size=2048 lazy_refcounts=off refcount_bits=16
$ qemu-img check --object secret,id=cluster_encrypt0,data=redhat \
'json:{"driver": "qcow2", "encrypt.format": "luks", \
"encrypt.key-secret": "cluster_encrypt0", \
"file.driver": "file", "file.filename": "cluster_size_check.qcow2"}'
ERROR: counting reference for region exceeding the end of the file by one cluster or more: offset 0x2000 size 0x1f9000
Leaked cluster 4 refcount=1 reference=0
...snip...
Leaked cluster 130 refcount=1 reference=0
1 errors were found on the image.
Data may be corrupted, or further writes to the image may corrupt it.
127 leaked clusters were found on the image.
This means waste of disk space, but no harm to data.
Image end offset: 268288
The problem only exists when the disk image is entirely empty. Writing
data to the disk image payload will solve the problem by causing the
end of the file to be extended further.
The change fixes it by ensuring that the entire allocated LUKS header
region is fully initialized with zeros. The qemu-img check will still
fail for any pre-existing disk images created prior to this change,
unless at least 1 byte of the payload is written to.
Fully writing zeros to the entire LUKS header is a good idea regardless
as it ensures that space has been allocated on the host filesystem (or
whatever block storage backend is used).
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Message-Id: <20200207135520.2669430-1-berrange@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
I/O requests to encrypted media should be aligned to the sector size
used by the underlying encryption method, not to BDRV_SECTOR_SIZE.
Fortunately this doesn't break anything at the moment because
both existing QCRYPTO_BLOCK_*_SECTOR_SIZE have the same value as
BDRV_SECTOR_SIZE.
The checks in qcow2_co_preadv_encrypted() are also unnecessary because
they are repeated immediately afterwards in qcow2_co_encdec().
Signed-off-by: Alberto Garcia <berto@igalia.com>
Message-Id: <20200213171646.15876-1-berto@igalia.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
qemu-img's convert_co_copy_range() operates at the sector level and
block_copy() operates at the cluster level so this condition is always
true, but it is not necessary to restrict this here, so let's leave it
to the driver implementation return an error if there is any.
Signed-off-by: Alberto Garcia <berto@igalia.com>
Message-id: a4264aaee656910c84161a2965f7a501437379ca.1579374329.git.berto@igalia.com
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
qcow2_alloc_cluster_offset() and qcow2_get_cluster_offset() always
return offsets that are cluster-aligned so don't just check that they
are sector-aligned.
The check in qcow2_co_preadv_task() is also replaced by an assertion
for the same reason.
Signed-off-by: Alberto Garcia <berto@igalia.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 558ba339965f858bede4c73ce3f50f0c0493597d.1579374329.git.berto@igalia.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
The L1 table is read from disk using the byte-based bdrv_pread() and
is never accessed beyond its last element, so there's no need to
allocate more memory than that.
Signed-off-by: Alberto Garcia <berto@igalia.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: b2e27214ec7b03a585931bcf383ee1ac3a641a10.1579374329.git.berto@igalia.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
This is a bit more efficient than having to allocate and free memory
for each item.
The default size (60) is enough for all the existing incompatible
features or the "Unknown incompatible feature" message.
Suggested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Alberto Garcia <berto@igalia.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Message-id: 20200115135626.19442-1-berto@igalia.com
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
QEMU currently supports writing compressed data of the size equal to
one cluster. This patch allows writing QCOW2 compressed data that
exceed one cluster. Now, we split buffered data into separate clusters
and write them compressed using the block/aio_task API.
Suggested-by: Pavel Butsykin <pbutsykin@virtuozzo.com>
Suggested-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Signed-off-by: Andrey Shinkevich <andrey.shinkevich@virtuozzo.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 1575288906-551879-3-git-send-email-andrey.shinkevich@virtuozzo.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
The local_err check outside of the if block was necessary
when it was introduced in commit d1258dd0c8 because it needed to be
executed even if qcow2_load_autoloading_dirty_bitmaps() returned false.
After some modifications that all required the error check to remain
where it is, commit 9c98f145df finally moved the
qcow2_load_dirty_bitmaps() call into the if block, so now the error
check should be there, too.
Signed-off-by: Guoyi Tu <tu.guoyi@h3c.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
There's a couple of places left in the qcow2 code that still do the
calculation manually, so let's replace them.
Signed-off-by: Alberto Garcia <berto@igalia.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
In the common case, qcow2_co_pwrite_zeroes() already only modifies
metadata case, so we're fine with or without BDRV_REQ_NO_FALLBACK set.
The only exception is when using an external data file, where the
request is passed down to the block driver of the external data file. We
are forwarding the BDRV_REQ_NO_FALLBACK flag there, though, so this is
fine, too.
Declare the flag supported therefore.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
This is a change in behavior, so all instances need a good
justification. The comments added here should explain my reasoning.
qed already had a comment that suggests it always expected
bdrv_truncate()/blk_truncate() to behave as if exact=true were passed
(c743849bee came eight months before 55b949c847), so it was simply
broken until now.
Signed-off-by: Max Reitz <mreitz@redhat.com>
Message-id: 20190918095144.955-8-mreitz@redhat.com
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
[mreitz: Changed comment in qed.c to explain why a new QED file must be
empty, as requested and suggested by Maxim]
Signed-off-by: Max Reitz <mreitz@redhat.com>
When truncating a format node, the @exact parameter is generally handled
simply by virtue of the format storing the new size in the image
metadata. Such formats do not need to pass on the parameter to their
file nodes.
There are exceptions, though:
- raw and crypto cannot store the image size, and thus must pass on
@exact.
- When using qcow2 with an external data file, it just makes sense to
keep its size in sync with the qcow2 virtual disk (because the
external data file is the virtual disk). Therefore, we should pass
@exact when truncating it.
Signed-off-by: Max Reitz <mreitz@redhat.com>
Message-id: 20190918095144.955-7-mreitz@redhat.com
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
We have two drivers (iscsi and file-posix) that (in some cases) return
success from their .bdrv_co_truncate() implementation if the block
device is larger than the requested offset, but cannot be shrunk. Some
callers do not want that behavior, so this patch adds a new parameter
that they can use to turn off that behavior.
This patch just adds the parameter and lets the block/io.c and
block/block-backend.c functions pass it around. All other callers
always pass false and none of the implementations evaluate it, so that
this patch does not change existing behavior. Future patches take care
of that.
Suggested-by: Maxim Levitsky <mlevitsk@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
Message-id: 20190918095144.955-5-mreitz@redhat.com
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
There is no reason why the format drivers need to truncate the protocol
node when formatting it. When using the old .bdrv_co_create_ops()
interface, the file will be created with no size option anyway, which
generally gives it a size of 0. (Exceptions are block devices, which
cannot be truncated anyway.)
When using blockdev-create, the user must have given the file node some
size anyway, so there is no reason why we should override that.
qed is an exception, it needs the file to start completely empty (as
explained by c743849bee).
Signed-off-by: Max Reitz <mreitz@redhat.com>
Message-id: 20190918095144.955-4-mreitz@redhat.com
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
qcow2_check_read_snapshot_table() can perform consistency checks, but it
cannot fix everything. Specifically, it cannot allocate new clusters,
because that should wait until the refcount structures are known to be
consistent (i.e., after qcow2_check_refcounts()). Thus, it cannot call
qcow2_write_snapshots().
Do that in qcow2_check_fix_snapshot_table(), which is called after
qcow2_check_refcounts().
Currently, there is nothing that would set result->corruptions, so this
is a no-op. A follow-up patch will change that.
Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-id: 20191011152814.14791-10-mreitz@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
Reading the snapshot table can fail. That is a problem when we want to
repair the image.
Therefore, stop reading the snapshot table in qcow2_do_open() in check
mode. Instead, add a new function qcow2_check_read_snapshot_table()
that reads the snapshot table at a later point. In the future, we want
to handle errors here and fix them.
Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-id: 20191011152814.14791-9-mreitz@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
qcow2 v3 requires every snapshot table entry to have two extra data
fields: The 64-bit VM state size, and the virtual disk size. Both are
optional for v2 images, so they may not be present.
qcow2_upgrade() therefore should update the snapshot table to ensure all
entries have these extra data fields.
Buglink: https://bugzilla.redhat.com/show_bug.cgi?id=1727347
Reported-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-id: 20191011152814.14791-8-mreitz@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
This does not make sense right now, but it will make sense once we need
to do more than to just update s->qcow_version.
Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-id: 20191011152814.14791-7-mreitz@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-id: 20191011152814.14791-4-mreitz@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
qcow2_detect_metadata_preallocation() calls qcow2_get_refcount() which
requires s->lock to be taken to protect its accesses to the refcount
table and refcount blocks. However, nothing in this code path actually
took the lock. This could cause the same cache entry to be used by two
requests at the same time, for different tables at different offsets,
resulting in image corruption.
As it would be preferable to base the detection on consistent data (even
though it's just heuristics), let's take the lock not only around the
qcow2_get_refcount() calls, but around the whole function.
This patch takes the lock in qcow2_co_block_status() earlier and asserts
in qcow2_detect_metadata_preallocation() that we hold the lock.
Fixes: 69f47505ee
Cc: qemu-stable@nongnu.org
Reported-by: Michael Weiser <michael.weiser@gmx.de>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Tested-by: Michael Weiser <michael.weiser@gmx.de>
Reviewed-by: Michael Weiser <michael.weiser@gmx.de>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
The only reason I can imagine for this strange code at the very-end of
bdrv_reopen_commit is the fact that bs->read_only updated after
calling drv->bdrv_reopen_commit in bdrv_reopen_commit. And in the same
time, prior to previous commit, qcow2_reopen_bitmaps_rw did a wrong
check for being writable, when actually it only need writable file
child not self.
So, as it's fixed, let's move things to correct place.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: John Snow <jsnow@redhat.com>
Acked-by: Max Reitz <mreitz@redhat.com>
Message-id: 20190927122355.7344-10-vsementsov@virtuozzo.com
Signed-off-by: John Snow <jsnow@redhat.com>
qcow2_reopen_bitmaps_ro wants to store bitmaps and then mark them all
readonly. But the latter don't work, as
qcow2_store_persistent_dirty_bitmaps removes bitmaps after storing.
It's OK for inactivation but bad idea for reopen-ro. And this leads to
the following bug:
Assume we have persistent bitmap 'bitmap0'.
Create external snapshot
bitmap0 is stored and therefore removed
Commit snapshot
now we have no bitmaps
Do some writes from guest (*)
they are not marked in bitmap
Shutdown
Start
bitmap0 is loaded as valid, but it is actually broken! It misses
writes (*)
Incremental backup
it will be inconsistent
So, let's stop removing bitmaps on reopen-ro. But don't rejoice:
reopening bitmaps to rw is broken too, so the whole scenario will not
work after this patch and we can't enable corresponding test cases in
260 iotests still. Reopening bitmaps rw will be fixed in the following
patches.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: John Snow <jsnow@redhat.com>
Message-id: 20190927122355.7344-7-vsementsov@virtuozzo.com
Signed-off-by: John Snow <jsnow@redhat.com>
qmp_block_dirty_bitmap_add and do_block_dirty_bitmap_remove do acquire
aio context since 0a6c86d024. But this is not enough: we also must
lock qcow2 mutex when access in-image metadata. Especially it concerns
freeing qcow2 clusters.
To achieve this, move qcow2_can_store_new_dirty_bitmap and
qcow2_remove_persistent_dirty_bitmap to coroutine context.
Since we work in coroutines in correct aio context, we don't need
context acquiring in blockdev.c anymore, drop it.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: John Snow <jsnow@redhat.com>
Message-id: 20190920082543.23444-4-vsementsov@virtuozzo.com
Signed-off-by: John Snow <jsnow@redhat.com>
It improves performance for fragmented qcow2 images.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-id: 20190916175324.18478-6-vsementsov@virtuozzo.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
Similarly to previous commit, prepare for parallelizing write-loop
iterations.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20190916175324.18478-5-vsementsov@virtuozzo.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
Further patch will run partial requests of iterations of
qcow2_co_preadv in parallel for performance reasons. To prepare for
this, separate part which may be parallelized into separate function
(qcow2_co_preadv_task).
While being here, also separate encrypted clusters reading to own
function, like it is done for compressed reading.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20190916175324.18478-4-vsementsov@virtuozzo.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
* Change the qcow2_co_{encrypt|decrypt} to just receive full host and
guest offsets and use this function directly instead of calling
do_perform_cow_encrypt (which is removed by that patch).
* Adjust qcow2_co_encdec to take full host and guest offsets as well.
* Document the qcow2_co_{encrypt|decrypt} arguments
to prevent the bug fixed in former commit from hopefully
happening again.
Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
Message-id: 20190915203655.21638-3-mlevitsk@redhat.com
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
[mreitz: Let perform_cow() return the error value returned by
qcow2_co_encrypt(), as proposed by Vladimir]
Signed-off-by: Max Reitz <mreitz@redhat.com>
Replace instances of:
(n & (BDRV_SECTOR_SIZE - 1)) == 0
And:
(n & ~BDRV_SECTOR_MASK) == 0
With:
QEMU_IS_ALIGNED(n, BDRV_SECTOR_SIZE)
Which reveals the intent of the code better, and makes it easier to
locate the code checking alignment.
Signed-off-by: Nir Soffer <nsoffer@redhat.com>
Message-id: 20190827185913.27427-2-nsoffer@redhat.com
Reviewed-by: John Snow <jsnow@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
The size of the qcow2 L2 cache defaults to 32 MB, which can be easily
larger than the maximum amount of L2 metadata that the image can have.
For example: with 64 KB clusters the user would need a qcow2 image
with a virtual size of 256 GB in order to have 32 MB of L2 metadata.
Because of that, since commit b749562d98
we forbid the L2 cache to become larger than the maximum amount of L2
metadata for the image, calculated using this formula:
uint64_t max_l2_cache = virtual_disk_size / (s->cluster_size / 8);
The problem with this formula is that the result should be rounded up
to the cluster size because an L2 table on disk always takes one full
cluster.
For example, a 1280 MB qcow2 image with 64 KB clusters needs exactly
160 KB of L2 metadata, but we need 192 KB on disk (3 clusters) even if
the last 32 KB of those are not going to be used.
However QEMU rounds the numbers down and only creates 2 cache tables
(128 KB), which is not enough for the image.
A quick test doing 4KB random writes on a 1280 MB image gives me
around 500 IOPS, while with the correct cache size I get 16K IOPS.
Cc: qemu-stable@nongnu.org
Signed-off-by: Alberto Garcia <berto@igalia.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Implement and use new interface to get rid of hd_qiov.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20190604161514.262241-13-vsementsov@virtuozzo.com
Message-Id: <20190604161514.262241-13-vsementsov@virtuozzo.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Implement and use new interface to get rid of hd_qiov.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20190604161514.262241-12-vsementsov@virtuozzo.com
Message-Id: <20190604161514.262241-12-vsementsov@virtuozzo.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Use buffer based io in encrypted case.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20190604161514.262241-11-vsementsov@virtuozzo.com
Message-Id: <20190604161514.262241-11-vsementsov@virtuozzo.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
If a qcow2 file is preallocated, it can no longer guarantee that it
initially appears as filled with zeroes.
So implement .bdrv_has_zero_init() by checking whether the file is
preallocated; if so, forward the call to the underlying storage node,
except for when it is encrypted: Encrypted preallocated images always
return effectively random data, so .bdrv_has_zero_init() must always
return 0 for them.
.bdrv_has_zero_init_truncate() can remain bdrv_has_zero_init_1(),
because it presupposes PREALLOC_MODE_OFF.
Reported-by: Stefano Garzarella <sgarzare@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
Message-id: 20190724171239.8764-7-mreitz@redhat.com
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
We need to implement .bdrv_has_zero_init_truncate() for every block
driver that supports truncation and has a .bdrv_has_zero_init()
implementation.
Implement it the same way each driver implements .bdrv_has_zero_init().
This is at least not any more unsafe than what we had before.
Signed-off-by: Max Reitz <mreitz@redhat.com>
Message-id: 20190724171239.8764-5-mreitz@redhat.com
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
In my "build everything" tree, changing qemu/main-loop.h triggers a
recompile of some 5600 out of 6600 objects (not counting tests and
objects that don't depend on qemu/osdep.h). It includes block/aio.h,
which in turn includes qemu/event_notifier.h, qemu/notify.h,
qemu/processor.h, qemu/qsp.h, qemu/queue.h, qemu/thread-posix.h,
qemu/thread.h, qemu/timer.h, and a few more.
Include qemu/main-loop.h only where it's needed. Touching it now
recompiles only some 1700 objects. For block/aio.h and
qemu/event_notifier.h, these numbers drop from 5600 to 2800. For the
others, they shrink only slightly.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20190812052359.30071-21-armbru@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Commit b76b4f60 allowed '-o compat=v3' as an alias for the
less-appealing '-o compat=1.1' for 'qemu-img create' since we want to
use the QMP form as much as possible, but forgot to do likewise for
qemu-img amend. Also, it doesn't help that '-o help' doesn't list our
new preferred spellings.
Signed-off-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This patch is used in the 'block/stream: introduce a bottom node'
that is following. Instead of the base node, the caller may pass
the node that has the base as its backing image to the function
bdrv_is_allocated_above() with a new parameter include_base = true
and get rid of the dependency on the base that may change during
commit/stream parallel jobs. Now, if the specified base is not
found in the backing image chain, the QEMU will abort.
Suggested-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Signed-off-by: Andrey Shinkevich <andrey.shinkevich@virtuozzo.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
Message-id: 1559152576-281803-2-git-send-email-andrey.shinkevich@virtuozzo.com
[mreitz: Squashed in the following as a rebase on conflicting patches:]
Message-id: e3cf99ae-62e9-8b6e-5a06-d3c8b9363b85@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
This adds a new parameter to blk_new() which requires its callers to
declare from which AioContext this BlockBackend is going to be used (or
the locks of which AioContext need to be taken anyway).
The given context is only stored and kept up to date when changing
AioContexts. Actually applying the stored AioContext to the root node
is saved for another commit.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
drv_co_block_status digs bs->file for additional, more accurate search
for hole inside region, reported as DATA by bs since 5daa74a6eb.
This accuracy is not free: assume we have qcow2 disk. Actually, qcow2
knows, where are holes and where is data. But every block_status
request calls lseek additionally. Assume a big disk, full of
data, in any iterative copying block job (or img convert) we'll call
lseek(HOLE) on every iteration, and each of these lseeks will have to
iterate through all metadata up to the end of file. It's obviously
ineffective behavior. And for many scenarios we don't need this lseek
at all.
However, lseek is needed when we have metadata-preallocated image.
So, let's detect metadata-preallocation case and don't dig qcow2's
protocol file in other cases.
The idea is to compare allocation size in POV of filesystem with
allocations size in POV of Qcow2 (by refcounts). If allocation in fs is
significantly lower, consider it as metadata-preallocation case.
102 iotest changed, as our detector can't detect shrinked file as
metadata-preallocation, which don't seem to be wrong, as with metadata
preallocation we always have valid file length.
Two other iotests have a slight change in their QMP output sequence:
Active 'block-commit' returns earlier because the job coroutine yields
earlier on a blocking operation. This operation is loading the refcount
blocks in qcow2_detect_metadata_preallocation().
Suggested-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
If COW areas of the newly allocated clusters are zeroes on the backing
image, efficient bdrv_write_zeroes(flags=BDRV_REQ_NO_FALLBACK) can be
used on the whole cluster instead of writing explicit zero buffers later
in perform_cow().
iotest 060:
write to the discarded cluster does not trigger COW anymore.
Use a backing image instead.
Signed-off-by: Anton Nefedov <anton.nefedov@virtuozzo.com>
Message-id: 20190516142749.81019-2-anton.nefedov@virtuozzo.com
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
Do encryption/decryption in threads, like it is already done for
compression. This improves asynchronous encrypted io.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20190506142741.41731-9-vsementsov@virtuozzo.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
Encryption will be done in threads, to take benefit of it, we should
move it out of the lock first.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20190506142741.41731-8-vsementsov@virtuozzo.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
Background: decryption will be done in threads, to take benefit of it,
we should move it out of the lock first.
But let's go further: it turns out, that only
qcow2_get_cluster_offset() needs locking, so reduce locking to it.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-id: 20190506142741.41731-7-vsementsov@virtuozzo.com
Reviewed-by: Alberto Garcia <berto@igalia.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
Move generic part out of qcow2_co_do_compress, to reuse it for
encryption and rename things that would be shared with encryption path.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20190506142741.41731-6-vsementsov@virtuozzo.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
Move compression-on-threads to separate file. Encryption will be in it
too.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20190506142741.41731-3-vsementsov@virtuozzo.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
qcow2.h depends on block_int.h. Compilation isn't broken currently only
due to block_int.h always included before qcow2.h. Though, it seems
better to directly include block_int.h in qcow2.h.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20190506142741.41731-2-vsementsov@virtuozzo.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
When an L2 table entry points to a compressed cluster the space used
by the data is specified in 512-byte sectors. This size is independent
from BDRV_SECTOR_SIZE and is specific to the qcow2 file format.
The QCOW2_COMPRESSED_SECTOR_SIZE constant defined in this patch makes
this explicit.
Signed-off-by: Alberto Garcia <berto@igalia.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
The last user of this field disappeared when we replace the
sector-based bdrv_write() with the byte-based bdrv_pwrite().
Signed-off-by: Alberto Garcia <berto@igalia.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This patch fixes a few things in the way error codes are handled in
the qcow2 compression code:
a) qcow2_co_pwritev_compressed() expects qcow2_co_compress() to only
return -1 or -2 on failure, but this is not correct. Since the
change from qcow2_compress() to qcow2_co_compress() in commit
ceb029cd6f the new code can also return -EINVAL (although
there does not seem to exist any code path that would cause that
error in the current implementation).
b) -1 and -2 are ad-hoc error codes defined in qcow2_compress().
This patch replaces them with standard constants from errno.h.
c) Both qcow2_compress() and qcow2_co_do_compress() return a negative
value on failure, but qcow2_co_pwritev_compressed() stores the
value in an unsigned data type.
Signed-off-by: Alberto Garcia <berto@igalia.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
make_completely_empty() is an optimisated path for bdrv_make_empty()
where completely new metadata is created inside the image file instead
of going through all clusters and discarding them. For an external data
file, however, we actually need to do discard operations on the data
file; just overwriting the qcow2 file doesn't get rid of the data.
The necessary slow path with an explicit discard operation already
exists for other cases. Use it for external data files, too.
Cc: qemu-stable@nongnu.org
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
preallocate_co() already gave the data file the full size without
forwarding the requested preallocation mode to the protocol. When
bdrv_co_truncate() was called later with the preallocation mode, the
file didn't actually grow any more, so the data file stayed unallocated
even if full preallocation was requested.
Pass the right preallocation mode to preallocate_co() and remove the
second bdrv_co_truncate() to fix this. As a side effect, the ugly
one-byte write in preallocate_co() is replaced with a truncate call,
now leaving the last block unallocated on the protocol level as it
should be.
Cc: qemu-stable@nongnu.org
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
We'll add a bdrv_co_truncate() call in the next patch which can return
an Error that we don't want to discard. So add an errp parameter to
preallocate_co().
Cc: qemu-stable@nongnu.org
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Limiting the allocation to INT_MAX bytes isn't particularly clever
because it means that the final cluster will be a partial cluster which
will be completed through a COW operation. This results in unnecessary
data read and write requests which lead to an unwanted non-sparse
filesystem block for metadata preallocation.
Align the maximum allocation size down to the cluster size to avoid this
situation.
Cc: qemu-stable@nongnu.org
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
With an external data file, preallocate_co() must write the final byte
to the external data file, not to the qcow2 image file.
This is harmless for preallocation of newly created images (only the
qcow2 file size is increased to the virtual disk size while it should be
much smaller), but with preallocated resize, it could in theory cause
visible corruption if the metadata of the image is larger than the data
(e.g. lots of bitmaps).
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
We were trying to check whether bdrv_open_blockdev_ref() returned
success, but accidentally checked the wrong variable. Spotted by
Coverity (CID 1399703).
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
If we reopen a BlockDriverState and there is an option that is present
in bs->options but missing from the new set of options then we have to
return an error unless the driver is able to reset it to its default
value.
This patch adds a new 'mutable_opts' field to BlockDriver. This is
a list of runtime options that can be modified during reopen. If an
option in this list is unspecified on reopen then it must be reset (or
return an error).
Signed-off-by: Alberto Garcia <berto@igalia.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Since we now load all bitmaps into memory anyway, we can just truncate
them in-memory and then flush them back to disk. Just in case, we will
still check and enforce that this shortcut is valid -- i.e. that any
bitmap described on-disk is indeed in-memory and can be modified.
If there are any inconsistent bitmaps, we refuse to allow the truncate
as we do not actually load these bitmaps into memory, and it isn't safe
or reasonable to attempt to truncate corrupted data.
Signed-off-by: John Snow <jsnow@redhat.com>
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-id: 20190311185147.52309-4-vsementsov@virtuozzo.com
[vsementsov: drop bitmap flushing, fix block comments style]
Signed-off-by: John Snow <jsnow@redhat.com>
Provide an option to force QEMU to always keep the external data file
consistent as a standalone read-only raw image.
At the moment, this means making sure that write_zeroes requests are
forwarded to the data file instead of just updating the metadata, and
checking that no backing file is used.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Rather than requiring that the external data file node is passed
explicitly when creating the qcow2 node, store the filename in the
designated header extension during .bdrv_create and read it from there
as a default during .bdrv_open.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This changes the qcow2 implementation to direct all guest data I/O to
s->data_file rather than bs->file, while metadata I/O still uses
bs->file. At the moment, this is still always the same, but soon we'll
add options to set s->data_file to an external data file.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Offset 0 cannot be assumed to mean an unallocated cluster any more.
Instead, the cluster type needs to be checked.
*file must refer to the data file instead of the image file if a valid
offset is returned from qcow2_co_block_status().
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
qcow2_alloc_compressed_cluster_offset() used to return the cluster
offset for success and 0 for error. This doesn't only conflict with 0 as
a valid host offset, but also loses the error code.
Similar to the change made to qcow2_alloc_cluster_offset() for
uncompressed clusters in commit 148da7ea9d, make the function return
0/-errno and return the allocated cluster offset in a by-reference
parameter.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This adds basic constants, struct fields and helper function for
external data file support to the implementation.
QCOW2_INCOMPAT_MASK and QCOW2_AUTOCLEAR_MASK are not updated yet so that
opening images with an external data file still fails (we don't handle
them correctly yet).
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Image creation already involves a bdrv_co_truncate() call, which allows
to specify a preallocation mode. Just pass the right mode there and
remove the code that is made redundant by this.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
QEMU 2.12 (commit 1221fe6f63) introduced
a new setting called l2-cache-entry-size that allows making entries on
the qcow2 L2 cache smaller than the cluster size.
I have been performing several tests with different cluster and entry
sizes and all of them show that reducing the entry size (aka L2 slice)
consistently improves I/O performance, notably during random I/O (all
tests done with sequential I/O show similar results). This is to be
expected because loading and evicting an L2 slice is more expensive
the larger the slice is.
Here are some numbers on fully populated 40GB qcow2 images. The
rightmost column represents the maximum L2 cache size in both cases.
Cluster size = 64 KB
|-------------+--------------+--------------+--------------|
| | 1MB L2 cache | 3MB L2 cache | 5MB L2 cache |
|-------------+--------------+--------------+--------------|
| 4KB slices | 6545 IOPS | 12045 IOPS | 55680 IOPS |
| 16KB slices | 5177 IOPS | 9798 IOPS | 56278 IOPS |
| 64KB slices | 2718 IOPS | 5326 IOPS | 57355 IOPS |
|-------------+--------------+--------------+--------------|
Cluster size = 256 KB
|--------------+----------------+--------------+-----------------|
| | 512KB L2 cache | 1MB L2 cache | 1280KB L2 cache |
|--------------+----------------+--------------+-----------------|
| 4KB slices | 8539 IOPS | 21071 IOPS | 55417 IOPS |
| 64KB slices | 3598 IOPS | 9772 IOPS | 57687 IOPS |
| 256KB slices | 1415 IOPS | 4120 IOPS | 58001 IOPS |
|--------------+----------------+--------------+-----------------|
As can be seen in the numbers, the only exception to the rule is when
the cache is large enough to hold all L2 tables. This is also to be
expected because in this case no cache entry is ever evicted so
reducing its size doesn't bring any benefit.
This patch sets the default L2 cache entry size to 4KB except when the
cache is large enough for the whole disk.
Signed-off-by: Alberto Garcia <berto@igalia.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
- Block graph change fixes (avoid loops, cope with non-tree graphs)
- bdrv_set_aio_context() related fixes
- HMP snapshot commands: Use only tag, not the ID to identify snapshots
- qmeu-img, commit: Error path fixes
- block/nvme: Build fix for gcc 9
- MAINTAINERS updates
- Fix various issues with bdrv_refresh_filename()
- Fix various iotests
- Include LUKS overhead in qemu-img measure for qcow2
- A fix for vmdk's image creation interface
-----BEGIN PGP SIGNATURE-----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=gaah
-----END PGP SIGNATURE-----
Merge remote-tracking branch 'remotes/kevin/tags/for-upstream' into staging
Block layer patches:
- Block graph change fixes (avoid loops, cope with non-tree graphs)
- bdrv_set_aio_context() related fixes
- HMP snapshot commands: Use only tag, not the ID to identify snapshots
- qmeu-img, commit: Error path fixes
- block/nvme: Build fix for gcc 9
- MAINTAINERS updates
- Fix various issues with bdrv_refresh_filename()
- Fix various iotests
- Include LUKS overhead in qemu-img measure for qcow2
- A fix for vmdk's image creation interface
# gpg: Signature made Mon 25 Feb 2019 14:18:15 GMT
# gpg: using RSA key 7F09B272C88F2FD6
# gpg: Good signature from "Kevin Wolf <kwolf@redhat.com>" [full]
# Primary key fingerprint: DC3D EB15 9A9A F95D 3D74 56FE 7F09 B272 C88F 2FD6
* remotes/kevin/tags/for-upstream: (71 commits)
iotests: Skip 211 on insufficient memory
vmdk: false positive of compat6 with hwversion not set
iotests: add LUKS payload overhead to 178 qemu-img measure test
qcow2: include LUKS payload overhead in qemu-img measure
iotests.py: s/_/-/g on keys in qmp_log()
iotests: Let 045 be run concurrently
iotests: Filter SSH paths
iotests.py: Filter filename in any string value
iotests.py: Add is_str()
iotests: Fix 207 to use QMP filters for qmp_log
iotests: Fix 232 for LUKS
iotests: Remove superfluous rm from 232
iotests: Fix 237 for Python 2.x
iotests: Re-add filename filters
iotests: Test json:{} filenames of internal BDSs
block: BDS options may lack the "driver" option
block/null: Generate filename even with latency-ns
block/curl: Implement bdrv_refresh_filename()
block/curl: Harmonize option defaults
block/nvme: Fix bdrv_refresh_filename()
...
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
LUKS encryption reserves clusters for its own payload data. The size of
this area must be included in the qemu-img measure calculation so that
we arrive at the correct minimum required image size.
(Ab)use the qcrypto_block_create() API to determine the payload
overhead. We discard the payload data that qcrypto thinks will be
written to the image.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20190218104525.23674-2-stefanha@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
This new field can be set by block drivers to list the runtime options
they accept that may influence the contents of the respective BDS. As of
a follow-up patch, this list will be used by the common
bdrv_refresh_filename() implementation to decide which options to put
into BDS.full_open_options (and consequently whether a JSON filename has
to be created), thus freeing the drivers of having to implement that
logic themselves.
Additionally, this patch adds the field to all of the block drivers that
need it and sets it accordingly.
Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
Message-id: 20190201192935.18394-22-mreitz@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
If the backing file is overridden, this most probably does change the
guest-visible data of a BDS. Therefore, we will need to consider this
in bdrv_refresh_filename().
To see whether it has been overridden, we might want to compare
bs->backing_file and bs->backing->bs->filename. However,
bs->backing_file is changed by bdrv_set_backing_hd() (which is just used
to change the backing child at runtime, without modifying the image
header), so bs->backing_file most of the time simply contains a copy of
bs->backing->bs->filename anyway, so it is useless for such a
comparison.
This patch adds an auto_backing_file BDS field which contains the
backing file path as indicated by the image header, which is not changed
by bdrv_set_backing_hd().
Because of bdrv_refresh_filename() magic, however, a BDS's filename may
differ from what has been specified during bdrv_open(). Then, the
comparison between bs->auto_backing_file and bs->backing->bs->filename
may fail even though bs->backing was opened from bs->auto_backing_file.
To mitigate this, we can copy the real BDS's filename (after the whole
bdrv_open() and bdrv_refresh_filename() process) into
bs->auto_backing_file, if we know the former has been opened based on
the latter. This is only possible if no options modifying the backing
file's behavior have been specified, though. To simplify things, this
patch only copies the filename from the backing file if no options have
been specified for it at all.
Furthermore, there are cases where an overlay is created by qemu which
already contains a BDS's filename (e.g. in blockdev-snapshot-sync). We
do not need to worry about updating the overlay's bs->auto_backing_file
there, because we actually wrote a post-bdrv_refresh_filename() filename
into the image header.
So all in all, there will be false negatives where (as of a future
patch) bdrv_refresh_filename() will assume that the backing file differs
from what was specified in the image header, even though it really does
not. However, these cases should be limited to where (1) the user
actually did override something in the backing chain (e.g. by specifying
options for the backing file), or (2) the user executed a QMP command to
change some node's backing file (e.g. change-backing-file or
block-commit with @backing-file given) where the given filename does not
happen to coincide with qemu's idea of the backing BDS's filename.
Then again, (1) really is limited to -drive. With -blockdev or
blockdev-add, you have to adhere to the schema, so a user cannot give
partial "unimportant" options (e.g. by just setting backing.node-name
and leaving the rest to the image header). Therefore, trying to fix
this would mean trying to fix something for -drive only.
To improve on (2), we would need a full infrastructure to "canonicalize"
an arbitrary filename (+ options), so it can be compared against
another. That seems a bit over the top, considering that filenames
nowadays are there mostly for the user's entertainment.
Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
Message-id: 20190201192935.18394-5-mreitz@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
Use new qemu_iovec_init_buf() instead of
qemu_iovec_init_external( ... , 1), which simplifies the code.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20190218140926.333779-10-vsementsov@virtuozzo.com
Message-Id: <20190218140926.333779-10-vsementsov@virtuozzo.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
In the 'Format specific information' section of the 'qemu-img info'
command output, the supplemental information about existing QCOW2
bitmaps will be shown, such as a bitmap name, flags and granularity:
image: /vz/vmprivate/VM1/harddisk.hdd
file format: qcow2
virtual size: 64G (68719476736 bytes)
disk size: 3.0M
cluster_size: 1048576
Format specific information:
compat: 1.1
lazy refcounts: true
bitmaps:
[0]:
flags:
[0]: in-use
[1]: auto
name: back-up1
granularity: 65536
[1]:
flags:
[0]: in-use
[1]: auto
name: back-up2
granularity: 65536
refcount bits: 16
corrupt: false
Signed-off-by: Andrey Shinkevich <andrey.shinkevich@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <1549638368-530182-3-git-send-email-andrey.shinkevich@virtuozzo.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
Inform a user in case qcow2_get_specific_info fails to obtain
QCOW2 image specific information. This patch is preliminary to
the one "qcow2: Add list of bitmaps to ImageInfoSpecificQCow2".
Signed-off-by: Andrey Shinkevich <andrey.shinkevich@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <1549638368-530182-2-git-send-email-andrey.shinkevich@virtuozzo.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
In the block layer, synchronous APIs are often implemented by creating a
coroutine that calls the asynchronous coroutine-based implementation and
then waiting for completion with BDRV_POLL_WHILE().
For this to work with iothreads (more specifically, when the synchronous
API is called in a thread that is not the home thread of the block
device, so that the coroutine will run in a different thread), we must
make sure to call aio_wait_kick() at the end of the operation. Many
places are missing this, so that BDRV_POLL_WHILE() keeps hanging even if
the condition has long become false.
Note that bdrv_dec_in_flight() involves an aio_wait_kick() call. This
corresponds to the BDRV_POLL_WHILE() in the drain functions, but it is
generally not enough for most other operations because they haven't set
the return value in the coroutine entry stub yet. To avoid race
conditions there, we need to kick after setting the return value.
The race window is small enough that the problem doesn't usually surface
in the common path. However, it does surface and causes easily
reproducible hangs if the operation can return early before even calling
bdrv_inc/dec_in_flight, which many of them do (trivial error or no-op
success paths).
The bug in bdrv_truncate(), bdrv_check() and bdrv_invalidate_cache() is
slightly different: These functions even neglected to schedule the
coroutine in the home thread of the node. This avoids the hang, but is
obviously wrong, too. Fix those to schedule the coroutine in the right
AioContext in addition to adding aio_wait_kick() calls.
Cc: qemu-stable@nongnu.org
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Do decompression in threads, like it is already done for compression.
This improves asynchronous compressed reads performance.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Allocate buffers locally and release qcow2 lock. Than, reads inside
qcow2_co_preadv_compressed may be done in parallel, however all
decompression is still done synchronously. Let's improve it in the
following commit.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
We are gradually moving away from sector-based interfaces, towards
byte-based. Get rid of it here too.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
- make it look more like a pair of qcow2_compress - rename the function
and its parameters
- drop extra out_len variable, check filling of output buffer by strm
structure itself
- fix code style
- add some documentation
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Compression is done in threads in qcow2.c. We want to do decompression
in the same way, so, firstly, move it to the same file.
The only change is braces around if-body in decompress_buffer, to
satisfy checkpatch.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Give explicit size both for source and destination buffers, to make it
similar with decompression path and than cleanly reuse parameter
structure for decompression threads.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Use appropriate macro, corresponding to deflateInit2 spec.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
The two thing that should be handled are cipher and ivgen. For ivgen
the solution is just mutex, as iv calculations should not be long in
comparison with encryption/decryption. And for cipher let's just keep
per-thread ciphers.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
This doesn't have any practical effect at the moment because the
values of BDRV_SECTOR_SIZE, QCRYPTO_BLOCK_LUKS_SECTOR_SIZE and
QCRYPTO_BLOCK_QCOW_SECTOR_SIZE are all the same (512 bytes), but
future encryption methods could have different requirements.
Signed-off-by: Alberto Garcia <berto@igalia.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Taking the address of a field in a packed struct is a bad idea, because
it might not be actually aligned enough for that pointer type (and
thus cause a crash on dereference on some host architectures). Newer
versions of clang warn about this. Avoid the bug by not using the
"modify in place" byte swapping functions.
There are a few places where the in-place swap function is
used on something other than a packed struct field; we convert
those anyway, for consistency.
This patch was produced with the following spatch script
(and hand-editing to fold a few resulting overlength lines):
@@
expression E;
@@
-be16_to_cpus(&E);
+E = be16_to_cpu(E);
@@
expression E;
@@
-be32_to_cpus(&E);
+E = be32_to_cpu(E);
@@
expression E;
@@
-be64_to_cpus(&E);
+E = be64_to_cpu(E);
@@
expression E;
@@
-cpu_to_be16s(&E);
+E = cpu_to_be16(E);
@@
expression E;
@@
-cpu_to_be32s(&E);
+E = cpu_to_be32(E);
@@
expression E;
@@
-cpu_to_be64s(&E);
+E = cpu_to_be64(E);
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Tested-by: John Snow <jsnow@redhat.com>
Reviewed-by: John Snow <jsnow@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This patch aims to bring the following behavior:
1. We don't load bitmaps, when started in inactive mode. It's the case
of incoming migration. In this case we wait for bitmaps migration
through migration channel (if 'dirty-bitmaps' capability is enabled) or
for invalidation (to load bitmaps from the image).
2. We don't remove persistent bitmaps on inactivation. Instead, we only
remove bitmaps after storing. This is the only way to restore bitmaps,
if we decided to resume source after [failed] migration with
'dirty-bitmaps' capability enabled (which means, that bitmaps were not
stored).
3. We load bitmaps on open and any invalidation, it's ok for all cases:
- normal open
- migration target invalidation with dirty-bitmaps capability
(bitmaps are migrating through migration channel, the are not
stored, so they should have IN_USE flag set and will be skipped
when loading. However, it would fail if bitmaps are read-only[1])
- migration target invalidation without dirty-bitmaps capability
(normal load of the bitmaps, if migrated with shared storage)
- source invalidation with dirty-bitmaps capability
(skip because IN_USE)
- source invalidation without dirty-bitmaps capability
(bitmaps were dropped, reload them)
[1]: to accurately handle this, migration of read-only bitmaps is
explicitly forbidden in this patch.
New mechanism for not storing bitmaps when migrate with dirty-bitmaps
capability is introduced: migration filed in BdrvDirtyBitmap.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Signed-off-by: John Snow <jsnow@redhat.com>
This variable doesn't work as it should, because it is actually cleared
in qcow2_co_invalidate_cache() by memset(). Drop it, as the following
patch will introduce new behavior.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: John Snow <jsnow@redhat.com>
Signed-off-by: John Snow <jsnow@redhat.com>