Commit fecd264 added a number of fall-throughs, but neglected to
properly document them as intentional. Commit d922445 cleaned that up
for many, but not all cases. Take care of the remaining ones.
Spotted by Coverity.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.22 (GNU/Linux)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=RmXU
-----END PGP SIGNATURE-----
Merge remote-tracking branch 'remotes/kevin/tags/for-upstream' into staging
Block patches for 2.3
# gpg: Signature made Fri 06 Feb 2015 17:14:10 GMT using RSA key ID C88F2FD6
# gpg: Good signature from "Kevin Wolf <kwolf@redhat.com>"
* remotes/kevin/tags/for-upstream: (47 commits)
block/raw-posix.c: Fix raw_getlength() on Mac OS X block devices
block: Eliminate silly QERR_ macros used for encryption keys
block: New bdrv_add_key(), convert monitor to use it
blockdev: Eliminate silly QERR_BLOCK_JOB_NOT_ACTIVE macro
blockdev: Give find_block_job() an Error ** parameter
qcow2: Rewrite qcow2_alloc_bytes()
block: Give always priority to unused entries in the qcow2 L2 cache
nbd: fix max_discard/max_transfer_length
block: introduce BDRV_REQUEST_MAX_SECTORS
nbd: Improve error messages
iotests: Fix 104 for NBD
iotests: Fix 100 for nbd
iotests: Fix 083
block: fix off-by-one error in qcow and qcow2
qemu-iotests: add 116 invalid QED input file tests
qed: check for header size overflow
block/dmg: improve zeroes handling
block/dmg: support bzip2 block entry types
block/dmg: factor out block type check
block/dmg: use SectorNumber from BLKX header
...
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
This patch replaces the dummy code in raw_getlength() for block devices
on OS X, which always returned LLONG_MAX, with a real implementation
that returns the actual block device size.
Signed-off-by: John Arbuckle <programmingkidx@gmail.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Tested-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
* mreitz/block:
block: Eliminate silly QERR_ macros used for encryption keys
block: New bdrv_add_key(), convert monitor to use it
blockdev: Eliminate silly QERR_BLOCK_JOB_NOT_ACTIVE macro
blockdev: Give find_block_job() an Error ** parameter
The QERR_ macros are leftovers from the days of "rich" error objects.
They're used with error_set() and qerror_report(), and expand into the
first *two* arguments. This trickiness has become pointless. Clean
up QERR_DEVICE_ENCRYPTED and QERR_DEVICE_NOT_ENCRYPTED.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-id: 1422524221-8566-5-git-send-email-armbru@redhat.com
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-id: 1422524221-8566-4-git-send-email-armbru@redhat.com
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
The QERR_ macros are leftovers from the days of "rich" error objects.
They're used with error_set() and qerror_report(), and expand into the
first *two* arguments. This trickiness has become pointless. Clean
this one up.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-id: 1422524221-8566-3-git-send-email-armbru@redhat.com
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
When find_block_job() fails, all its callers build the same Error
object. Build it in find_block_job() instead.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-id: 1422524221-8566-2-git-send-email-armbru@redhat.com
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
qcow2_alloc_bytes() is a function with insufficient error handling and
an unnecessary goto. This patch rewrites it.
Signed-off-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
The current algorithm to replace entries from the L2 cache gives
priority to newer hits by dividing the hit count of all existing
entries by two everytime there is a cache miss.
However, if there are several cache misses the hit count of the
existing entries can easily go down to 0. This will result in those
entries being replaced even when there are others that have never been
used.
This problem is more noticeable with larger disk images and cache
sizes, since the chances of having several misses before the cache is
full are higher.
If we make sure that the hit count can never go down to 0 again,
unused entries will always have priority.
Signed-off-by: Alberto Garcia <berto@igalia.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
nbd_co_discard calls nbd_client_session_co_discard which uses uint32_t
as the length in bytes of the data to discard due to the following
definition:
struct nbd_request {
uint32_t magic;
uint32_t type;
uint64_t handle;
uint64_t from;
uint32_t len; <-- the length of data to be discarded, in bytes
} QEMU_PACKED;
Thus we should limit bl_max_discard to UINT32_MAX >> BDRV_SECTOR_BITS to
avoid overflow.
NBD read/write code uses the same structure for transfers. Fix
max_transfer_length accordingly.
Signed-off-by: Denis V. Lunev <den@openvz.org>
CC: Peter Lieven <pl@kamp.de>
CC: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
we check and adjust request sizes at several places with
sometimes inconsistent checks or default values:
INT_MAX
INT_MAX >> BDRV_SECTOR_BITS
UINT_MAX >> BDRV_SECTOR_BITS
SIZE_MAX >> BDRV_SECTOR_BITS
This patches introdocues a macro for the maximal allowed sectors
per request and uses it at several places.
Signed-off-by: Peter Lieven <pl@kamp.de>
Reviewed-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This patch makes use of the Error object for nbd_receive_negotiate() so
that errors during negotiation look nicer.
Furthermore, this patch adds an additional error message if the received
magic was wrong, but would be correct for the other protocol version,
respectively: So if an export name was specified, but the NBD server
magic corresponds to an old handshake, this condition is explicitly
signaled to the user, and vice versa.
As these messages are now part of the "Could not open image" error
message, additional filtering has to be employed in iotest 083, which
this patch does as well.
Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
_make_test_img sets up an NBD server, _cleanup_test_img shuts it down;
thus, _cleanup_test_img has to be called before _make_test_img is
invoked another time.
Furthermore, the pipe through _filter_test_img was unnecessary;
_make_test_img already takes care of that.
And finally, a filter is added to _filter_img_info to replace
"nbd://127.0.0.1:10810" by "TEST_DIR/t.IMGFMT", since the former is the
way to express the full image path (normally the latter) for NBD tests.
Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
In case of NBD, _make_test_img starts a new NBD server. Therefore,
_cleanup_test_img (which shuts that server down) has to be invoked
before the next _make_test_img call in order to make 100 work for NBD.
Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
As of 8f9e835fd2, probing should be
disabled in the qemu-iotests (at least when using qemu-io). This broke
083's reference output (which consisted mostly of "Could not read image
for determining its format").
This patch fixes it.
Note that one case which failed before is now successful: Disconnect
after data. This is due to qemu having read twice before (once for
probing, once for the qemu-io read command), but only once now (the
qemu-io read command). Therefore, reading is successful (which is
correct).
Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This fixes an off-by-one error introduced in 9a29e18. Both qcow and
qcow2 need to make sure to leave room for string terminator '\0' for
the backing file, so the max length of the non-terminated string is
either 1023 or PATH_MAX - 1.
Reported-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Jeff Cody <jcody@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
These tests exercise error code paths in the QED image format. The
tests are very simple, they just prove that the error path exits
cleanly.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 1421065893-18875-3-git-send-email-stefanha@redhat.com
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Header size is denoted in clusters. The maximum cluster size is 64 MB
but there is no limit on header size. Check for uint32_t overflow in
case the header size field has a whacky value.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 1421065893-18875-2-git-send-email-stefanha@redhat.com
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Disk images may contain large all-zeroes gaps (1.66k sectors or 812 MiB
is seen in the real world). These blocks (type 2) do not need to be
extracted into a temporary buffer, there is no need to allocate memory
for these blocks nor to check its length.
(For the test image, the maximum uncompressed size is 1054371 bytes,
probably for a bzip2-compressed block.)
Signed-off-by: Peter Wu <peter@lekensteyn.nl>
Reviewed-by: John Snow <jsnow@redhat.com>
Message-id: 1420566495-13284-13-git-send-email-peter@lekensteyn.nl
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This patch adds support for bzip2-compressed block entries as introduced
with OS X 10.4 (source: https://en.wikipedia.org/wiki/Apple_Disk_Image).
It was tested against a 5.2G "OS X Yosemite" installation image which
stores the BLXX block in the XML property list (instead of resource
forks) and has over 5k chunks.
New configure entries are added (--enable-bzip2 / --disable-bzip2) to
control inclusion of bzip2 functionality (which requires linking against
libbz2). The help message suggests that this option is needed for DMG
files, but the tests are generic enough that other parts of QEMU can use
bzip2 if needed.
The identifiers are based on http://newosxbook.com/DMG.html.
The decompression routines are based on the zlib case, but as there is
no way to reset the decompression state (unlike zlib), memory is
allocated and deallocated for every decompression. This should not be
problematic as the decompression takes most of the time and as blocks
are typically about/over 1 MiB in size, only one allocation is done
every 2000 sectors.
Signed-off-by: Peter Wu <peter@lekensteyn.nl>
Reviewed-by: John Snow <jsnow@redhat.com>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Message-id: 1420566495-13284-12-git-send-email-peter@lekensteyn.nl
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
In preparation for adding bzip2 support, split the type check into a
separate function. Make all offsets relative to the begin of a chunk
such that it is easier to recognize the position without having to
add up all offsets. Some comments are added to describe the fields.
There is no functional change.
Signed-off-by: Peter Wu <peter@lekensteyn.nl>
Reviewed-by: John Snow <jsnow@redhat.com>
Message-id: 1420566495-13284-11-git-send-email-peter@lekensteyn.nl
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Previously the sector table parsing relied on the previous offset of
the DMG file. Now it uses the sector number from the BLKX header
(see http://newosxbook.com/DMG.html).
The implementation of dmg2img (from vu1tur) does not base the output
sector on the location of the terminator (0xffffffff) either so it
should be safe to drop this dependency on the previous state.
(It makes somehow makes sense, a terminator should halt further
processing of a block and is perhaps used to preallocate some space.)
Signed-off-by: Peter Wu <peter@lekensteyn.nl>
Reviewed-by: John Snow <jsnow@redhat.com>
Message-id: 1420566495-13284-10-git-send-email-peter@lekensteyn.nl
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This patch addresses two issues:
- The data fork offset was not taken into account, resulting in failure
to read an InstallESD.dmg file (5164763151 bytes) which had a
non-zero DataForkOffset field.
- The offset of the previous block ("partition") was unconditionally
added to the current block because older files would start the input
offset of a new block at zero. Newer files (including vlc-2.1.5.dmg,
tuxpaint-0.9.15-macosx.dmg and OS X Yosemite [MAS].dmg) failed in
reads because these files have chunk offsets, relative to the begin
of a data fork.
Now the data offset of the mish is taken into account. While we could
check that the data_offset is within the data fork, let's not do that
here as it would only result in parse failures on invalid files (rather
than gracefully handling such bad files). dmg_read will error out if
the offset is incorrect.
Signed-off-by: Peter Wu <peter@lekensteyn.nl>
Reviewed-by: John Snow <jsnow@redhat.com>
Message-id: 1420566495-13284-9-git-send-email-peter@lekensteyn.nl
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Right now the virtual size is always reported as zero which makes it
impossible to convert between formats.
After this patch, the number of sectors will be read from the trailer
("koly" block).
To verify the behavior, the output of `dmg2img foo.dmg foo.img` was
compared against `qemu-img convert -f dmg -O raw foo.dmg foo.raw`. The
tests showed that the file contents are exactly the same, except that
QEMU creates a slightly larger file (it matches the total sectors
count).
Signed-off-by: Peter Wu <peter@lekensteyn.nl>
Reviewed-by: John Snow <jsnow@redhat.com>
Message-id: 1420566495-13284-8-git-send-email-peter@lekensteyn.nl
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
The format is simple enough to avoid using a full-blown XML parser. It
assumes that all BLKX items begin with the "mish" magic word, therefore
it is not a problem if other values get matched which are not a BLKX
block.
The offsets are based on the description at
http://newosxbook.com/DMG.html
For compatibility with glib 2.12, use g_base64_decode (which
additionally requires an extra buffer allocation) instead of
g_base64_decode_inplace (which is only available since glib 2.20).
Signed-off-by: Peter Wu <peter@lekensteyn.nl>
Reviewed-by: John Snow <jsnow@redhat.com>
Message-id: 1420566495-13284-7-git-send-email-peter@lekensteyn.nl
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Previously the chunk size was not checked, allowing for a large memory
allocation. This patch checks whether the chunks size is within the
resource fork length, and whether the resource fork is below the
trailer of the dmg file.
Signed-off-by: Peter Wu <peter@lekensteyn.nl>
Reviewed-by: John Snow <jsnow@redhat.com>
Message-id: 1420566495-13284-6-git-send-email-peter@lekensteyn.nl
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
As the decoded plist XML is not a pointer in the file,
dmg_read_mish_block must be able to process a buffer instead of a file
pointer. Since the full buffer must be processed, let's change the
return value again to just a success flag.
Signed-off-by: Peter Wu <peter@lekensteyn.nl>
Reviewed-by: John Snow <jsnow@redhat.com>
Message-id: 1420566495-13284-5-git-send-email-peter@lekensteyn.nl
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Besides the offset, also read the resource length. This length is now
used in the extracted function to verify the end of the resource fork
against "count" from the resource fork.
Instead of relying on the value of offset to conclude whether the
resource fork is available or not (info_begin==0), check the
rsrc_fork_length instead. This would allow a dmg file to begin with a
resource fork. This seemingly unnecessary restriction was found while
trying to craft a DMG file by hand.
Other changes:
- Do not require resource data offset to be 0x100 (but check that it
is within bounds though).
- Further improve boundary checking (resource data must be within
the resource fork).
- Use correct value for resource data length (spotted by John Snow)
- Consider the resource data offset when determining info_end.
This fixes an EINVAL on the tuxpaint dmg example.
The resource fork format is documented at
https://developer.apple.com/legacy/library/documentation/mac/pdf/MoreMacintoshToolbox.pdf#page=151
Signed-off-by: Peter Wu <peter@lekensteyn.nl>
Reviewed-by: John Snow <jsnow@redhat.com>
Message-id: 1420566495-13284-4-git-send-email-peter@lekensteyn.nl
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Extract the mish block decoder such that this can be used for other
formats in the future. A new DmgHeaderState struct is introduced to
share state while decoding.
The code is kept unchanged as much as possible, a "fail" label is added
for example where a simple return would probably do. In dmg_open, the
variable "tmp" is renamed to "rsrc_data_offset" for clarity and comments
have been added explaining various data.
Note that this patch has one subtle difference with the previous
version which should not affect functionality. In the previous code,
the end of a resource was inferred from the mish block (the offsets
would be increased by the fields). In this patch, the resource length
is used instead to avoid the need to rely on the previous offsets.
Signed-off-by: Peter Wu <peter@lekensteyn.nl>
Reviewed-by: John Snow <jsnow@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 1420566495-13284-3-git-send-email-peter@lekensteyn.nl
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
DMG files have a variable length with a UDIF trailer at the end of a
file. This UDIF trailer is essential as it describes the contents of
the image. At the moment however, the start of this trailer is almost
always incorrect as bdrv_getlength() returns a multiple of the block
size (rounded up). This results in a failure to recognize DMG files,
resulting in Invalid argument (EINVAL) errors.
As there is no API to retrieve the real file size, look for the magic
header in the last two sectors to find the start of this 512-byte UDIF
trailer (the "koly" block).
The resource fork offset ("info_begin") has its offset adjusted as the
initial value of offset does not mean "end of file" anymore, but "begin
of UDIF trailer".
[Replaced error_set(errp, ERROR_CLASS_GENERIC_ERROR, ...) with
error_setg(errp, ...) as discussed with Peter.
--Stefan]
Signed-off-by: Peter Wu <peter@lekensteyn.nl>
Reviewed-by: John Snow <jsnow@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 1420566495-13284-2-git-send-email-peter@lekensteyn.nl
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Managing applications, like oVirt (http://www.ovirt.org), make extensive
use of thin-provisioned disk images.
To let the guest run smoothly and be not unnecessarily paused, oVirt sets
a disk usage threshold (so called 'high water mark') based on the occupation
of the device, and automatically extends the image once the threshold
is reached or exceeded.
In order to detect the crossing of the threshold, oVirt has no choice but
aggressively polling the QEMU monitor using the query-blockstats command.
This lead to unnecessary system load, and is made even worse under scale:
deployments with hundreds of VMs are no longer rare.
To fix this, this patch adds:
* A new monitor command `block-set-write-threshold', to set a mark for
a given block device.
* A new event `BLOCK_WRITE_THRESHOLD', to report if a block device
usage exceeds the threshold.
* A new `write_threshold' field into the `BlockDeviceInfo' structure,
to report the configured threshold.
This will allow the managing application to use smarter and more
efficient monitoring, greatly reducing the need of polling.
[Updated qemu-iotests 067 output to add the new 'write_threshold'
property. --Stefan]
[Changed g_assert_false() to !g_assert() to fix the build on older glib
versions. --Kevin]
Signed-off-by: Francesco Romani <fromani@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-id: 1421068273-692-1-git-send-email-fromani@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This patch is necessary to suppress the "probed raw" warning when
running raw over nbd tests.
Signed-off-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
There is a bug in the recently added sys.platform test, and we no longer
run python tests, because "linux2" is the value to compare here. So do a
prefix match. According to python doc [1], the way to use sys.platform
is "unless you want to test for a specific system version, it is
therefore recommended to use the following idiom":
if sys.platform.startswith('freebsd'):
# FreeBSD-specific code here...
elif sys.platform.startswith('linux'):
# Linux-specific code here...
[1]: https://docs.python.org/2.7/library/sys.html#sys.platform
Signed-off-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
this adds a knob to disable request merging for debugging or benchmarks if dedired.
Signed-off-by: Peter Lieven <pl@kamp.de>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
this patch finally introduces multiread support to virtio-blk. While
multiwrite support was there for a long time, read support was missing.
The complete merge logic is moved into virtio-blk.c which has
been the only user of request merging ever since. This is required
to be able to merge chunks of requests and immediately invoke callbacks
for those requests. Secondly, this is required to switch to
direct invocation of coroutines which is planned at a later stage.
The following benchmarks show the performance of running fio with
4 worker threads on a local ram disk. The numbers show the average
of 10 test runs after 1 run as warmup phase.
| 4k | 64k | 4k
MB/s | rd seq | rd rand | rd seq | rd rand | wr seq | wr rand
--------------+--------+---------+--------+---------+--------+--------
master | 1221 | 1187 | 4178 | 4114 | 1745 | 1213
multiread | 1829 | 1189 | 4639 | 4110 | 1894 | 1216
Signed-off-by: Peter Lieven <pl@kamp.de>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
As it was not obvious (at least for me) where the 32 comes from;
add a constant for it.
Signed-off-by: Peter Lieven <pl@kamp.de>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Peter Lieven <pl@kamp.de>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
The commit 533ffb17a that removed qed_aiocb_info.cancel said to remove
this but didn't do it.
Signed-off-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This is the same way vl.c handles this.
Signed-off-by: Don Slutz <dslutz@verizon.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
do not trim requests if the driver does not supply a limit
through BlockLimits. For write zeroes we still keep a limit
for the unsupported path to avoid allocating a big bounce buffer.
Suggested-by: Kevin Wolf <kwolf@redhat.com>
Suggested-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: Peter Lieven <pl@kamp.de>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This sequence works efficiently if FALLOC_FL_ZERO_RANGE is not supported.
Unfortunately, FALLOC_FL_ZERO_RANGE is supported on really modern systems
and only for a couple of filesystems. FALLOC_FL_PUNCH_HOLE is much more
mature.
The sequence of 2 operations FALLOC_FL_PUNCH_HOLE and 0 is necessary due
to the following reasons:
- FALLOC_FL_PUNCH_HOLE creates a hole in the file, the file becomes
sparse. In order to retain original functionality we must allocate
disk space afterwards. This is done using fallocate(0) call
- fallocate(0) without preceeding FALLOC_FL_PUNCH_HOLE will do nothing
if called above already allocated areas of the file, i.e. the content
will not be zeroed
This should increase the performance a bit for not-so-modern kernels.
CC: Max Reitz <mreitz@redhat.com>
CC: Kevin Wolf <kwolf@redhat.com>
CC: Stefan Hajnoczi <stefanha@redhat.com>
CC: Peter Lieven <pl@kamp.de>
CC: Fam Zheng <famz@redhat.com>
Signed-off-by: Denis V. Lunev <den@openvz.org>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
There is a possibility that we are extending our image and thus writing
zeroes beyond the end of the file. In this case we do not need to care
about the hole to make sure that there is no data in the file under
this offset (pre-condition to fallocate(0) to work). We could simply call
fallocate(0).
This improves the performance of writing zeroes even on really old
platforms which do not have even FALLOC_FL_PUNCH_HOLE.
Before the patch do_fallocate was used when either
CONFIG_FALLOCATE_PUNCH_HOLE or CONFIG_FALLOCATE_ZERO_RANGE are defined.
Now the story is different. CONFIG_FALLOCATE is defined when Linux
fallocate is defined, posix_fallocate is completely different story
(CONFIG_POSIX_FALLOCATE). CONFIG_FALLOCATE is mandatory prerequite
for both CONFIG_FALLOCATE_PUNCH_HOLE and CONFIG_FALLOCATE_ZERO_RANGE
thus we are on the safe side.
CC: Max Reitz <mreitz@redhat.com>
CC: Kevin Wolf <kwolf@redhat.com>
CC: Stefan Hajnoczi <stefanha@redhat.com>
CC: Peter Lieven <pl@kamp.de>
CC: Fam Zheng <famz@redhat.com>
Signed-off-by: Denis V. Lunev <den@openvz.org>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This efficiently writes zeroes on Linux if the kernel is capable enough.
FALLOC_FL_ZERO_RANGE correctly handles all cases, including and not
including file expansion.
CC: Kevin Wolf <kwolf@redhat.com>
CC: Stefan Hajnoczi <stefanha@redhat.com>
CC: Peter Lieven <pl@kamp.de>
CC: Fam Zheng <famz@redhat.com>
Signed-off-by: Denis V. Lunev <den@openvz.org>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
move code dealing with a block device to a separate function. This will
allow to implement additional processing for ordinary files.
Please note, that xfs_code has been moved before checking for
s->has_write_zeroes as xfs_write_zeroes does not touch this flag inside.
This makes code a bit more consistent.
CC: Kevin Wolf <kwolf@redhat.com>
CC: Stefan Hajnoczi <stefanha@redhat.com>
CC: Peter Lieven <pl@kamp.de>
CC: Fam Zheng <famz@redhat.com>
Signed-off-by: Denis V. Lunev <den@openvz.org>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
The pattern
do {
if (fallocate(s->fd, mode, offset, len) == 0) {
return 0;
}
} while (errno == EINTR);
ret = translate_err(-errno);
will be commonly useful in next patches. Create helper for it.
CC: Kevin Wolf <kwolf@redhat.com>
CC: Stefan Hajnoczi <stefanha@redhat.com>
CC: Peter Lieven <pl@kamp.de>
CC: Fam Zheng <famz@redhat.com>
Signed-off-by: Denis V. Lunev <den@openvz.org>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Peter Lieven <pl@kamp.de>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
actually the code
if (ret == -ENODEV || ret == -ENOSYS || ret == -EOPNOTSUPP ||
ret == -ENOTTY) {
ret = -ENOTSUP;
}
is present twice and will be added a couple more times. Create helper
for this.
CC: Kevin Wolf <kwolf@redhat.com>
CC: Stefan Hajnoczi <stefanha@redhat.com>
CC: Peter Lieven <pl@kamp.de>
CC: Fam Zheng <famz@redhat.com>
Signed-off-by: Denis V. Lunev <den@openvz.org>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Peter Lieven <pl@kamp.de>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
(With the previous atapi_dma flag recovery)
If migration happens between the ATAPI command being written and the
bmdma being started, the DMA is dropped. Eventually the guest times
out and recovers, but that can take many seconds.
(This is rare, on a pingpong reading the CD continuously I hit
this about ~1/30-1/50 migrates)
I don't think we've got enough state to be able to recover safely
at this point, so I throw a 'medium error, no seek complete'
that I'm assuming guests will try and recover from an apparently
dirty CD.
OK, it's a hack, the real solution is probably to push a lot of
ATAPI state into the migration stream, but this is a fix that
works with no stream changes. Tested only on Linux (both RHEL5
(pre-libata) and RHEL7).
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: John Snow <jsnow@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
