this patch tries to optimize zero write requests
by automatically using bdrv_write_zeroes if it is
supported by the format.
This significantly speeds up file system initialization and
should speed zero write test used to test backend storage
performance.
I ran the following 2 tests on my internal SSD with a
50G QCOW2 container and on an attached iSCSI storage.
a) mkfs.ext4 -E lazy_itable_init=0,lazy_journal_init=0 /dev/vdX
QCOW2 [off] [on] [unmap]
-----
runtime: 14secs 1.1secs 1.1secs
filesize: 937M 18M 18M
iSCSI [off] [on] [unmap]
----
runtime: 9.3s 0.9s 0.9s
b) dd if=/dev/zero of=/dev/vdX bs=1M oflag=direct
QCOW2 [off] [on] [unmap]
-----
runtime: 246secs 18secs 18secs
filesize: 51G 192K 192K
throughput: 203M/s 2.3G/s 2.3G/s
iSCSI* [off] [on] [unmap]
----
runtime: 8mins 45secs 33secs
throughput: 106M/s 1.2G/s 1.6G/s
allocated: 100% 100% 0%
* The storage was connected via an 1Gbit interface.
It seems to internally handle writing zeroes
via WRITESAME16 very fast.
Signed-off-by: Peter Lieven <pl@kamp.de>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
this adds a generic function to recover the enum id of a parameter
given as a string.
Signed-off-by: Peter Lieven <pl@kamp.de>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
* remotes/qmp-unstable/queue/qmp:
monitor: fix qmp_getfd() fd leak in error case
HMP: support specifying dump format for dump-guest-memory
HMP: fix doc of dump-guest-memory
qmp: object-add: Validate class before creating object
monitor: Add device_add and device_del completion.
monitor: Add command_completion callback to mon_cmd_t.
monitor: Fix drive_del id argument type completion.
error: Remove some unused headers
qerror.h: Replace QERR_NOT_SUPPORTED with QERR_UNSUPPORTED
qerror.h: Remove QERR defines that are only used once
qerror.h: Remove unused error classes
error: Print error_report() to stderr if using qmp
monitor: Remove unused monitor_print_filename
error: Privatize error_print_loc
vnc: Remove default_mon usage
slirp: Remove default_mon usage
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Using error_is_set(ERRP) to find out whether a function failed is
either wrong, fragile, or unnecessarily opaque. It's wrong when ERRP
may be null, because errors go undetected when it is. It's fragile
when proving ERRP non-null involves a non-local argument. Else, it's
unnecessarily opaque (see commit 84d18f0).
The error_is_set(errp) in internal_snapshot_prepare() is merely
fragile, because the caller never passes a null errp argument.
Make the code more robust and more obviously correct: receive the
error in a local variable, then propagate it through the parameter.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Just hardcode them in the callers
Cc: Luiz Capitulino <lcapitulino@redhat.com>
Cc: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Cole Robinson <crobinso@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Since commit f298d071, block devices added with blockdev-add don't have
a QemuOpts around in dinfo->opts. Consequently, we can't rely any more
on QemuOpts catching duplicate IDs for block devices.
This patch adds a new check for duplicate IDs to bdrv_new(), and moves
the existing check that the ID isn't already taken for a node-name there
as well.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
This patch adds an errp parameter to bdrv_new() and updates all its
callers. The next patches will make use of this in order to check for
duplicate IDs. Most of the callers know that their ID is fine, so they
can simply assert that there is no error.
Behaviour doesn't change with this patch yet as bdrv_new() doesn't
actually assign errors to errp.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
As speed is an optional parameter for the QMP block-commit command, it
should be set to 0 if not given (as it is undefined if has_speed is
false), that is, the speed should not be limited.
Cc: qemu-stable@nongnu.org
Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
If aio=native, we check that cache.direct is set as well. If however
cache wasn't specified at all, qemu just segfaulted.
The old condition didn't make any sense anyway because it effectively
only checked for the default cache mode case, but not for an explicitly
set cache.direct=off mode.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Benoit Canet <benoit@irqsave.net>
Reviewed-by: Eric Blake <eblake@redhat.com>
Encrypted images need a password before they can be used, and we don't
want blockdev-add to create BDSes that aren't fully initialised. So for
now simply forbid encrypted images; we can come back to it later if we
need the functionality.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Allow bdrv_open() to handle references to existing block devices just as
bdrv_file_open() is already capable of.
Signed-off-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Make bdrv_open() take a pointer to a BDS pointer, similarly to
bdrv_file_open(). If a pointer to a NULL pointer is given, bdrv_open()
will create a new BDS with an empty name; if the BDS pointer is not
NULL, that existing BDS will be reused (in the same way as bdrv_open()
already did).
Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
* QTest cleanups and test cases for PCI NICs
* NAND fix for "info qtree"
* Cleanup and extension of QOM machine tests
* IndustryPack test cases and conversion to QOM realize
* I2C cleanups
* Cleanups of legacy qdev properties
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.22 (GNU/Linux)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=KRcc
-----END PGP SIGNATURE-----
Merge remote-tracking branch 'remotes/afaerber/tags/qom-devices-for-peter' into staging
QOM infrastructure fixes and device conversions
* QTest cleanups and test cases for PCI NICs
* NAND fix for "info qtree"
* Cleanup and extension of QOM machine tests
* IndustryPack test cases and conversion to QOM realize
* I2C cleanups
* Cleanups of legacy qdev properties
# gpg: Signature made Mon 17 Feb 2014 22:15:37 GMT using RSA key ID 3E7E013F
# gpg: Good signature from "Andreas Färber <afaerber@suse.de>"
# gpg: aka "Andreas Färber <afaerber@suse.com>"
* remotes/afaerber/tags/qom-devices-for-peter: (49 commits)
qtest: Include system headers before user headers
qapi: Refine human printing of sizes
qdev: Use QAPI type names for properties
qdev: Add enum property types to QAPI schema
block: Handle "rechs" and "large" translation options
qdev: Remove hex8/32/64 property types
qdev: Remove most legacy printers
qdev: Use human mode in "info qtree"
qapi: Add human mode to StringOutputVisitor
qdev: Inline qdev_prop_parse()
qdev: Legacy properties are just strings
qdev: Legacy properties are now read-only
qdev: Remove legacy parsers for hex8/32/64
qdev: Sizes are now parsed by StringInputVisitor
qapi: Add size parser to StringInputVisitor
qtest: Don't segfault with invalid -qtest option
ipack: Move IndustryPack out of hw/char/
ipoctal232: QOM parent field cleanup
ipack: QOM parent field cleanup for IPackDevice
ipack: QOM parent field cleanup for IPackBus
...
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
error_is_set(&var) is the same as var != NULL, but it takes
whole-program analysis to figure that out. Unnecessarily hard for
optimizers, static checkers, and human readers. Dumb it down to
obvious.
Gets rid of several dozen Coverity false positives.
Note that the obvious form is already used in many places.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Andreas Färber <afaerber@suse.de>
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Sure, CHS translation is an obscure topic, and legacy options for
hard-disk geometries are obscure as well. But since QEMU does nothing
with it except telling the BIOS, and since there "large" and "rechs"
are listed in the enums, parsing them seems to be the bare minimum.
Acked-by: Stefan Hajnoczi <stefanha@gmail.com>
Reviewed-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Andreas Färber <afaerber@suse.de>
Since we introduced node_name for named bs of the graph modify the opening by
reference to use it as a fallback.
This patch also enforce the separation of the device id and graph node
namespaces.
Signed-off-by: Benoit Canet <benoit@irqsave.net>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
As bdrv_open() documentation states:
"The reference to the QDict belongs to the block layer
* after the call (even on failure), so if the caller intends to reuse the
* dictionary, it needs to use QINCREF() before calling bdrv_open."
the optional options dict will not be reused after bdrv_open() and should
belong to the block layer so remove the extra QDECREF(options).
Signed-off-by: Benoit Canet <benoit@irqsave.net>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
blockdev-add doesn't know about the device that the backend will be
attached to, this is a legacy -drive concept. Move the remaining checks
that use it to drive_init().
[Fam Zheng <famz@redhat.com> suggested line-wrapping to 80 chars as
required by the coding standard. I have fixed this.
--Stefan]
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Benoit Canet <benoit@irqsave.net>
Reviewed-by: Fam Zheng <famz@redhat.com>
There was two candidate ways to implement named node manipulation:
1)
{ 'command': 'block_passwd', 'data': {'*device': 'str',
'*node-name': 'str', 'password': 'str'}
}
2)
{ 'command': 'block_passwd', 'data': {'device': 'str',
'*device-is-node': 'bool',
'password': 'str'} }
Luiz proposed 1 and says 2 was an abuse of the QMP interface and proposed to
rewrite the QMP block interface for 2.0.
Luiz does not like in 1 the fact that 2 fields are optional but one of them must
be specified leading to an abuse of the QMP semantic.
Kevin argumented that 2 what a clear abuse of the device field and would not be
practical when reading fast some log file because the user would read "device"
and think that a device is manipulated when it's in fact a node name.
Documentation of 1 make it pretty clear what to do for the user.
Kevin argued that all bs are node including devices ones so 2 does not make
sense.
Kevin also argued that rewriting the QMP block interface would not make disapear
the current one.
Kevin pushed the argument that making the QAPI generator compatible with the
semantic of the operation would need a rewrite that no one has done yet.
A vote has been done on the list to elect the version to use and 1 won.
For reference the complete thread is:
"[Qemu-devel] [PATCH V4 4/7] qmp: Allow to change password on names block driver
states."
Signed-off-by: Benoit Canet <benoit@irqsave.net>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Specifying the image filename through the "file" option is a legacy
option and should not be supported by blockdev-add (in that case, giving
a string for "file" references an existing block device).
Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This is a boiler-plate _nofail variant of qemu_opts_create. Remove and
use error_abort in call sites.
null/0 arguments needs to be added for the id and fail_if_exists fields
in affected callsites due to argument inconsistency between the normal and
no_fail variants.
Signed-off-by: Peter Crosthwaite <peter.crosthwaite@xilinx.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
If active is top, it will be mirrored to base, (with block/mirror.c
code), then the image is switched when user completes the block job.
QMP documentation is updated.
Signed-off-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
For "none" sync mode in "absolute-paths" mode, the current image should
be used as the backing file for the newly created image.
The current behavior is:
a) If the image to be mirrored has a backing file, use that (which is
wrong, since the operations recorded by "none" are applied to the
image itself, not to its backing file).
b) If the image to be mirrored lacks a backing file, the target doesn't
have one either (which is not really wrong, but not really right,
either; "none" records a set of operations executed on the image
file, therefore having no backing file to apply these operations on
seems rather pointless).
For a, this is clearly a bugfix. For b, it is still a bugfix, although
it might break existing API - but since that case crashed qemu just
three weeks ago (before 1452686495), we
can safely assume there is no such API relying on that case yet.
Suggested-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-id: 1385407736-13941-2-git-send-email-mreitz@redhat.com
Signed-off-by: Anthony Liguori <aliguori@amazon.com>
Currently we have three QemuOptsList (qemu_common_drive_opts,
qemu_legacy_drive_opts, and qemu_drive_opts), only qemu_drive_opts
is added to vm_config_groups[].
This patch changes query-command-line-options to access three local
QemuOptsLists for drive option, and merge the description items
together.
Signed-off-by: Amos Kong <akong@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
It should be possible to execute the QMP "drive-mirror" command in
"none" sync mode and "absolute-paths" mode even for block devices
lacking a backing file.
"absolute-paths" does in fact not require a backing file to be present,
as can be seen from the "top" sync mode code path. "top" basically
states that the device should indeed have a backing file - however, the
current code catches the case if it doesn't and then simply treats it as
"full" sync mode, creating a target image without a backing file (in
"absolute-paths" mode). Thus, "absolute-paths" does not imply the target
file must indeed have a backing file.
Therefore, the target file may be left unbacked in case of "none" sync
mode as well, if the specified device is not backed either. Currently,
qemu will crash trying to dereference the backing file pointer since it
assumes that it will always be non-NULL in that case ("none" with
"absolute-paths").
Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Wenchao Xia <xiawenc@linux.vnet.ibm.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
These memory leaks also make drive_add if=none,id=drive0 without a file=
option leak the options list. This keeps ID "drive0" around forever.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Since 0ebd24e0, cdrom doesn't have read-only on by default, which will
error out when using an read only image. Fix it by setting the default
value when parsing opts.
Reported-by: Edivaldo de Araujo Pereira <edivaldoapereira@yahoo.com.br>
Signed-off-by: Fam Zheng <famz@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
This gives us meaningful error messages for the blockdev-add QMP
command.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
If a read-only device is configured with copy-on-read=on, the old code
only prints a warning and automatically disables copy on read. Make it
a real error for blockdev-add.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
The remaining users shouldn't be there with blockdev-add and are easy to
move to drive_init().
Bonus bug fix: As a side effect, CD-ROM drives can now use block drivers
on the read-only whitelist without explicitly specifying read-only=on,
even if a format is explicitly specified.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
IF_NONE allows read-only, which makes forbidding it in this place
for other types pretty much pointless.
Instead, make sure that all devices for which the check would have
errored out check in their init function that they don't get a read-only
BlockDriverState. This catches even cases where IF_NONE and -device is
used.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
This requires moving the automatic ID generation at the same time, so
let's do that as well.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
It's already ignored and only prints a deprecation message. No use in
making it available in new interfaces.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
This moves all of the geometry options (cyls/heads/secs/trans) to
drive_init so that they can only be accessed using legacy functions, but
never with anything blockdev-add related.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
It's always IF_NONE for blockdev-add.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Benoit Canet <benoit@irqsave.net>
Reviewed-by: Eric Blake <eblake@redhat.com>
This moves as much as possible of the processing of the 'media' option
to drive_init so that it can only be accessed using legacy functions,
but never with anything blockdev-add related.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Benoit Canet <benoit@irqsave.net>
Reviewed-by: Eric Blake <eblake@redhat.com>
Working on a QDict instead of a QemuOpts that accepts anything is more
in line with bdrv_open(). A QDict is what qmp_blockdev_add() already has
anyway, so this saves additional conversions. And last, but not least,
it allows later patches to easily extract legacy options into a
separate, typed QemuOpts for drive_init() (the untyped QemuOpts that
drive_init already has doesn't allow access to numbers, only strings,
and is therefore useless without conversion).
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Benoit Canet <benoit@irqsave.net>
Reviewed-by: Eric Blake <eblake@redhat.com>
blockdev-add shouldn't automatically generate IDs, but will keep most of
the DriveInfo creation code.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Wenchao Xia <xiawenc@linux.vnet.ibm.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
BlockDriverStates shouldn't be affected by an unplugged guest device,
except if created with the legacy -drive command line option or the
drive_add HMP command.
Make the automatic deletion as well as cancelling of jobs conditional on
an enable_auto_del boolean that is only set in drive_init().
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Wenchao Xia <xiawenc@linux.vnet.ibm.com>
The main intent of this patch is to consolidate the whitelist checks to
a single point in the code instead of spreading it everywhere. This adds
a nicer error message for read-only whitelisting, too, in places where
it was still missing.
The patch also contains a bonus bug fix: By finding the format first in
bdrv_open() and then independently checking against the whitelist only
later, we avoid the case that use of a non-whitelisted format results in
probing rather than an error message. Previously, this could happen when
using the driver=... option.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
This field is used by blkverify to disable external snapshots creation.
It will also be used by block filters like quorum to disable external
snapshot creation.
Signed-off-by: Benoit Canet <benoit@irqsave.net>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
