It's like bdrv_parse_cache_flags(), except that writethrough mode isn't
included in the flags, but returned as a separate bool.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
This patch introduces block driver that implement recording
and replaying of block devices' operations.
All block completion operations are added to the queue.
Queue is flushed at checkpoints and information about processed requests
is recorded to the log. In replay phase the queue is matched with
events read from the log. Therefore block devices requests are processed
deterministically.
Signed-off-by: Pavel Dovgalyuk <pavel.dovgaluk@ispras.ru>
[ kwolf: Rebased onto modified and already applied part of the series ]
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This patch fixes error message in saving loop of the asynchronous events queue.
Signed-off-by: Pavel Dovgalyuk <pavel.dovgaluk@ispras.ru>
[ kwolf: Fixed format string to use PRId64 instead of %d ]
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This patch fixes scheduling of bottom halves when record/replay is enabled.
Now BH are not added to replay queue when asynchronous events are disabled.
This may happen in startup and loadvm/savevm phases of execution.
Signed-off-by: Pavel Dovgalyuk <pavel.dovgaluk@ispras.ru>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This patch adds callback for flush request. This callback is responsible
for flushing whole block devices stack. bdrv_flush function does not
proceed to underlying devices. It should be performed by this callback
function, if needed.
Signed-off-by: Pavel Dovgalyuk <pavel.dovgaluk@ispras.ru>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
It is important that the QEMU luks implementation retains 100%
compatibility with the reference implementation provided by
the combination of the linux kernel dm-crypt module and cryptsetup
userspace tools.
There is a matrix of tests to be performed with different sets
of encryption settings. For each matrix entry, two tests will
be performed. One will create a LUKS image with the cryptsetup
tool and then do I/O with both cryptsetup & qemu-io. The other
will create the image with qemu-img and then again do I/O with
both cryptsetup and qemu-io.
The new I/O test 149 performs interoperability testing between
QEMU and the reference implementation. Such testing inherantly
requires elevated privileges, so to this this the user must have
configured passwordless sudo access. The test will automatically
skip if sudo is not available.
The test has to be run explicitly thus:
cd tests/qemu-iotests
./check -luks 149
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
For a couple of releases we have been warning
Encrypted images are deprecated
Support for them will be removed in a future release.
You can use 'qemu-img convert' to convert your image to an unencrypted one.
This warning was issued by system emulators, qemu-img, qemu-nbd
and qemu-io. Such a broad warning was issued because the original
intention was to rip out all the code for dealing with encryption
inside the QEMU block layer APIs.
The new block encryption framework used for the LUKS driver does
not rely on the unloved block layer API for encryption keys,
instead using the QOM 'secret' object type. It is thus no longer
appropriate to warn about encryption unconditionally.
When the qcow/qcow2 drivers are converted to use the new encryption
framework too, it will be practical to keep AES-CBC support present
for use in qemu-img, qemu-io & qemu-nbd to allow for interoperability
with older QEMU versions and liberation of data from existing encrypted
qcow2 files.
This change moves the warning out of the generic block code and
into the qcow/qcow2 drivers. Further, the warning is set to only
appear when running the system emulators, since qemu-img, qemu-io,
qemu-nbd are expected to support qcow2 encryption long term now that
the maint burden has been eliminated.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Add a block driver that is capable of supporting any full disk
encryption format. This utilizes the previously added block
encryption code, and at this time supports the LUKS format.
The driver code is capable of supporting any format supported
by the QCryptoBlock module, so it registers one block driver
for each format. This patch only registers the "luks" driver
since the "qcow" driver is there only for back-compatibility
with existing qcow built-in encryption.
New LUKS compatible volumes can be formatted using qemu-img
with defaults for all settings.
$ qemu-img create --object secret,data=123456,id=sec0 \
-f luks -o key-secret=sec0 demo.luks 10G
Alternatively the cryptographic settings can be explicitly
set
$ qemu-img create --object secret,data=123456,id=sec0 \
-f luks -o key-secret=sec0,cipher-alg=aes-256,\
cipher-mode=cbc,ivgen-alg=plain64,hash-alg=sha256 \
demo.luks 10G
And query its size
$ qemu-img info demo.img
image: demo.img
file format: luks
virtual size: 10G (10737418240 bytes)
disk size: 132K
encrypted: yes
Note that it was not necessary to provide the password
when querying info for the volume. The password is only
required when performing I/O on the volume
All volumes created by this new 'luks' driver should be
capable of being opened by the kernel dm-crypt driver.
The only algorithms listed in the LUKS spec that are
not currently supported by this impl are sha512 and
ripemd160 hashes and cast6 cipher.
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
[ kwolf - Added #include to resolve conflict with da34e65c ]
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Add a 'log' method to iotests.py which prints messages to
stdout, with optional filtering of data. Port over some
standard filters already present in the shell common.filter
code to be usable in python too.
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
The iotests.py helper provides a main() method for running
tests via the python unit test framework. Not all tests
will want to use this, so refactor it to split the testing
of compatible formats and platforms into separate helper
methods
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
The python I/O tests helper for running qemu-img/qemu-io
setup stdout to be captured to a pipe, but left stderr
untouched. As a result, if something failed in qemu-img/
qemu-io, data written to stderr would get output directly
and not line up with data on the test stdout due to
buffering. If we explicitly redirect stderr to the same
pipe as stdout, things are much clearer when they go
wrong.
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
The qemu-img/qemu-io tools prompt for disk encryption passwords
regardless of whether any are actually required. Adding a check
on bdrv_key_required() avoids this prompt for disk formats which
have been converted to the QCryptoSecret APIs.
This is just a temporary hack to ensure the block I/O tests
continue to work after each patch, since the last patch will
completely delete all the password prompting code.
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
When opening an image it is useful to know whether the caller
intends to perform I/O on the image or not. In the case of
encrypted images this will allow the block driver to avoid
having to prompt for decryption keys when we merely want to
query header metadata about the image. eg qemu-img info
This flag is enforced at the top level only, since even if
we don't want todo I/O on the 'qcow2' file payload, the
underlying 'file' driver will still need todo I/O to read
the qcow2 header, for example.
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
bdrv_query_blk_stats() does not need access to all of BlockStats,
BlockDeviceStats is enough and is what this function is actually
supposed to fill.
Signed-off-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This is the only instance of bdrv_query_blk_stats() accessing anything
in the BlockStats structure other than s->stats, so let us move it to
its caller (where it makes just as much sense) allowing us to make
bdrv_query_blk_stats() take a pointer to the BlockDeviceStats instead of
BlockStats.
Signed-off-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Using heap instead of stack for better safety.
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Fix two places to use literal printf format when possible.
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
The function is unused since commit f21d96d0 ('block: Use BdrvChild in
BlockBackend').
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Mac OS X can be picky when it comes to allowing the user
to use physical devices in QEMU. Most mounted volumes
appear to be off limits to QEMU. If an issue is detected,
a message is displayed showing the user how to unmount a
volume. Now QEMU uses both CD and DVD media.
Signed-off-by: John Arbuckle <programmingkidx@gmail.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Writethrough mode is going to become a BlockBackend feature rather than
a BDS one, so forbid it in places where we won't be able to support it
when the code finally matches the envisioned design.
We only allowed setting the cache mode of non-root nodes after the 2.5
release, so we're still free to make this change.
The target of block jobs is now always opened in a writeback mode
because it doesn't have a BlockBackend attached. This makes more sense
anyway because block jobs know when to flush. If the graph is modified
on job completion, the original cache mode moves to the new root, so
for the guest device writethough always stays enabled if it was
configured this way.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
First of all, we're generally not writing to backing files, but when we
do, it's in the context of block jobs which know very well when to flush
the image.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
The WCE bit is a frontend property and should not be part of the backend
configuration. This is especially important because the same BDS can be
used by different users with different WCE requirements.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This patch changes dirty bitmaps from following a BlockBackend in graph
changes to sticking with the node they were created at. For the full
discussion, read the following mailing list thread:
[Qemu-block] block: Dirty bitmaps and COR in bdrv_move_feature_fields()
https://lists.nongnu.org/archive/html/qemu-block/2016-02/msg00745.html
In summary, the justification for this change is:
* When moving the dirty bitmap to the top of the tree was introduced in
bdrv_append() in commit a9fc4408, it didn't actually have any effect
because there could never be a bitmap in use when bdrv_append() was
called (op blockers would prevent this). This is still true today for
all internal uses of dirty bitmaps.
* Support for user-defined dirty bitmaps was introduced in 2.4, but we
discouraged users from using it because we didn't consider it ready
yet.
Moreover, in 2.5, the bdrv_swap() removal introduced a bug that left
dangling pointers if a dirty bitmap was present (the anchors of the
dirty bitmap were swapped, but the back link in the first element
wasn't updated), so it didn't even work correctly.
* block-dirty-bitmap-add takes an arbitrary node name, even if no
BlockBackend is attached. This suggests that it is a node level
operation and not a BlockBackend one. Consequently, there is no reason
for dirty bitmaps to stay with a BlockBackend that was attached to the
node they were created for.
* It was suggested that block-dirty-bitmap-add could track the node if a
node name was specified, and track the BlockBackend if the device name
was specified. This would however be inconsistent with other QMP
commands. Commands that accept both device and node names currently
interpret the device name just as an alias for the current root node
of that BlockBackend.
* Dirty bitmaps have a name that is only unique amongst the bitmaps in a
specific node. Moving bitmaps could lead to name clashes. Automatic
renaming would involve too much magic.
* Persistent bitmaps are stored in a specific node. Moving them around
automatically might be at least surprising, but it would probably also
become a real problem because that would have to happen atomically
without the management tool knowing of the operation.
At the end of the day it seems to be very clear that it was a mistake to
include dirty bitmaps in bdrv_move_feature_fields(). The functionality
of moving bitmaps and/or attaching them to a BlockBackend instead will
probably be needed, but it should be done with a new explicit QMP
command or option.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Ever since we first introduced bdrv_append() in commit 8802d1fd ('qapi:
Introduce blockdev-group-snapshot-sync command'), the copy-on-read flag
was moved to the new top layer when taking a snapshot. The only problem
is that it doesn't make a whole lot of sense.
The use case for manually enabled CoR is to avoid reading data twice
from a slow remote image, so we want to save it to a local overlay, say
an ISO image accessed via HTTP to a local qcow2 overlay. When taking a
snapshot, we end up with a backing chain like this:
http <- local.qcow2 <- snap_overlay.qcow2
There is no point in doing CoR from local.qcow2 into snap_overlay.qcow2,
we just want to keep copying data from the remote source into
local.qcow2.
The other use case of CoR is in the context of streaming, which isn't
very interesting for bdrv_move_feature_fields() because op blockers
prevent this combination.
This patch makes the copy-on-read flag stay on the image for which it
was originally set and prevents it from being propagated to the new
overlay. It is no longer intended to move CoR to the BlockBackend level.
In order for this to make sense, we also need to keep the respective
image read-write.
As a side effect of these changes, creating a live snapshot image (as
opposed to using an existing externally created one) on top of a COR
block device works now. It used to fail because it tried to open its
backing file both read-only and with COR.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
The call in hmp_drive_del() is dead code because blk_remove_bs() is
called a few lines above. The only other remaining user is
bdrv_delete(), which only abuses bdrv_make_anon() to remove it from the
named nodes list. This path inlines the list entry removal into
bdrv_delete() and removes bdrv_make_anon().
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
- Use 128bit math to avoid asserts with IOMMU regions (Bandan Das)
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.14 (GNU/Linux)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=DpB8
-----END PGP SIGNATURE-----
Merge remote-tracking branch 'remotes/awilliam/tags/vfio-update-20160328.0' into staging
VFIO updates 2016-03-28
- Use 128bit math to avoid asserts with IOMMU regions (Bandan Das)
# gpg: Signature made Mon 28 Mar 2016 23:16:52 BST using RSA key ID 3BB08B22
# gpg: Good signature from "Alex Williamson <alex.williamson@redhat.com>"
# gpg: aka "Alex Williamson <alex@shazbot.org>"
# gpg: aka "Alex Williamson <alwillia@redhat.com>"
# gpg: aka "Alex Williamson <alex.l.williamson@gmail.com>"
* remotes/awilliam/tags/vfio-update-20160328.0:
vfio: convert to 128 bit arithmetic calculations when adding mem regions
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Rename the recently-added ip6-foo options into ipv6-foo options, to make
them coherent with other ipv6 options.
Also rework the documentation.
Signed-off-by: Samuel Thibault <samuel.thibault@ens-lyon.org>
Reviewed-by: Eric Blake <eblake@redhat.com>
vfio_listener_region_add for a iommu mr results in
an overflow assert since iommu memory region is initialized
with UINT64_MAX. Convert calculations to 128 bit arithmetic
for iommu memory regions and let int128_get64 assert for non iommu
regions if there's an overflow.
Suggested-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Bandan Das <bsd@redhat.com>
[missed (end - 1) on 2nd trace call, move llsize closer to use]
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
This patch tests that in a partial block-stream operation, no data is
ever copied from the base image.
Signed-off-by: Alberto Garcia <berto@igalia.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 5272a2aa57bc0b3f981f8b3e0c813e58a88c974b.1458566441.git.berto@igalia.com
Signed-off-by: Jeff Cody <jcody@redhat.com>
This test is streaming to the top layer using the intermediate image
as the base. This is a mistake since block-stream never copies data
from the base image and its backing chain, so this is effectively a
no-op.
In addition to fixing the base parameter, this patch also writes some
data to the intermediate image before the test, so there's something
to copy and the test is meaningful.
Signed-off-by: Alberto Garcia <berto@igalia.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 2efa304da38b32d47c120ce728568a589c5a3afc.1458566441.git.berto@igalia.com
Signed-off-by: Jeff Cody <jcody@redhat.com>
We need to call stream_complete() in order to do all the necessary
clean-ups, even if there's an early failure. At the moment it's only
useful to make sure that s->backing_file_str is not leaked, but it
will become more important if we introduce support for streaming to
any intermediate node.
Signed-off-by: Alberto Garcia <berto@igalia.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 2abedf2debc65c250560237f31a8e6756883c8fc.1458566441.git.berto@igalia.com
Signed-off-by: Jeff Cody <jcody@redhat.com>
* Chardev fix from Marc-André
* config.status tweak from David
* Header file tweaks from Markus, myself and Veronia (Outreachy candidate)
* get_ticks_per_sec() removal from Rutuja (Outreachy candidate)
* Coverity fix from myself
* PKE implementation from myself, based on rth's XSAVE support
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2
iQEcBAABCAAGBQJW9ErPAAoJEL/70l94x66DJfEH/A/QkMpAhrgNdyVsahzsGrzE
wx5gHFIc1nBYxyr62w4apUb5jPB7zaXu0LA7EAWDeAe0pyP8hZzLT9kJyOEDsuJu
zwKN2QeLSNMtPbnbKN0I/YQ2za2xX1V5ruhSeOJoVslUI214hgnAURaGshhQNzuZ
2CluDT9KgL5cQifAnKs5kJrwhIYShYNQB+1eDC/7wk28dd/EH+sPALIoF+rqrSmt
Zu4Mdqd+9Ns+oKOjA6br9ULq/Hzg0aDfY82J+XLVVqfF3PXQe8rTDmuMf/7jTn+M
Un7ZOcei9oZF2/9vfAfKQpDCcgD9HvOUSbgqV/ubmkPPmN/LNJzeKj0fBhrRN+Y=
=K12D
-----END PGP SIGNATURE-----
Merge remote-tracking branch 'remotes/bonzini/tags/for-upstream' into staging
* Log filtering from Alex and Peter
* Chardev fix from Marc-André
* config.status tweak from David
* Header file tweaks from Markus, myself and Veronia (Outreachy candidate)
* get_ticks_per_sec() removal from Rutuja (Outreachy candidate)
* Coverity fix from myself
* PKE implementation from myself, based on rth's XSAVE support
# gpg: Signature made Thu 24 Mar 2016 20:15:11 GMT using RSA key ID 78C7AE83
# gpg: Good signature from "Paolo Bonzini <bonzini@gnu.org>"
# gpg: aka "Paolo Bonzini <pbonzini@redhat.com>"
* remotes/bonzini/tags/for-upstream: (28 commits)
target-i386: implement PKE for TCG
config.status: Pass extra parameters
char: translate from QIOChannel error to errno
exec: fix error handling in file_ram_alloc
cputlb: modernise the debug support
qemu-log: support simple pid substitution for logs
target-arm: dfilter support for in_asm
qemu-log: dfilter-ise exec, out_asm, op and opt_op
qemu-log: new option -dfilter to limit output
qemu-log: Improve the "exec" TB execution logging
qemu-log: Avoid function call for disabled qemu_log_mask logging
qemu-log: correct help text for -d cpu
tcg: pass down TranslationBlock to tcg_code_gen
util: move declarations out of qemu-common.h
Replaced get_tick_per_sec() by NANOSECONDS_PER_SECOND
hw: explicitly include qemu-common.h and cpu.h
include/crypto: Include qapi-types.h or qemu/bswap.h instead of qemu-common.h
isa: Move DMA_transfer_handler from qemu-common.h to hw/isa/isa.h
Move ParallelIOArg from qemu-common.h to sysemu/char.h
Move QEMU_ALIGN_*() from qemu-common.h to qemu/osdep.h
...
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Conflicts:
scripts/clean-includes
Accumulated patches for target-ppc, pseries machine type and related
devices.
* Preliminary patches from BenH & Cédric Le Goater's powernv code
* We don't want the full machine type before 2.7
* Adding some of the SPRs also fixes migration corner cases for
spapr (when qemu has no knowledge of the registers, they're
obviously not migrated)
* We include some patches that aren't strictly fixes, but make
applying the others easier, and they're low risk
* Fix to buffer management which significantly improves throughput in
the spapr-llan virtual network device
* Start with 64-bit mode enabled on spapr. This is the way it's
supposed to be but we broke it a while back and didn't notice
because Linux guests cope anyway.
* Picked up by kvm-unit-tests
* Still some bugs here that I'm working on
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1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=AuUu
-----END PGP SIGNATURE-----
Merge remote-tracking branch 'remotes/dgibson/tags/ppc-for-2.6-20160324' into staging
ppc patch queue for 2016-03-24
Accumulated patches for target-ppc, pseries machine type and related
devices.
* Preliminary patches from BenH & Cédric Le Goater's powernv code
* We don't want the full machine type before 2.7
* Adding some of the SPRs also fixes migration corner cases for
spapr (when qemu has no knowledge of the registers, they're
obviously not migrated)
* We include some patches that aren't strictly fixes, but make
applying the others easier, and they're low risk
* Fix to buffer management which significantly improves throughput in
the spapr-llan virtual network device
* Start with 64-bit mode enabled on spapr. This is the way it's
supposed to be but we broke it a while back and didn't notice
because Linux guests cope anyway.
* Picked up by kvm-unit-tests
* Still some bugs here that I'm working on
# gpg: Signature made Thu 24 Mar 2016 04:29:42 GMT using RSA key ID 20D9B392
# gpg: Good signature from "David Gibson <david@gibson.dropbear.id.au>"
# gpg: aka "David Gibson (Red Hat) <dgibson@redhat.com>"
# gpg: aka "David Gibson (ozlabs.org) <dgibson@ozlabs.org>"
# gpg: WARNING: This key is not certified with sufficiently trusted signatures!
# gpg: It is not certain that the signature belongs to the owner.
# Primary key fingerprint: 75F4 6586 AE61 A66C C44E 87DC 6C38 CACA 20D9 B392
* remotes/dgibson/tags/ppc-for-2.6-20160324:
ppc: move POWER8 Book4 regs in their own routine
hw/net/spapr_llan: Enable the RX buffer pools by default for new machines
hw/net/spapr_llan: Fix receive buffer handling for better performance
hw/net/spapr_llan: Extract rx buffer code into separate functions
ppc: A couple more dummy POWER8 Book4 regs
ppc: Add dummy CIABR SPR
ppc: Add POWER8 IAMR register
ppc: Fix writing to AMR/UAMOR
ppc: Initialize AMOR in PAPR mode
ppc: Add dummy SPR_IC for POWER8
ppc: Create cpu_ppc_set_papr() helper
ppc: Add a bunch of hypervisor SPRs to Book3s
ppc: Add macros to register hypervisor mode SPRs
ppc: Update SPR definitions
spapr/target-ppc/kvm: Only add hcall-instructions if KVM supports it
ppc64: set MSR_SF bit
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
* update cocoa UI front end to use QKeyCodes
* fix the help menu documentation links to actually work
(with both an installed and an uninstalled QEMU)
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1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=RVzw
-----END PGP SIGNATURE-----
Merge remote-tracking branch 'remotes/pmaydell/tags/pull-cocoa-20160323-1' into staging
cocoa queue:
* update cocoa UI front end to use QKeyCodes
* fix the help menu documentation links to actually work
(with both an installed and an uninstalled QEMU)
# gpg: Signature made Wed 23 Mar 2016 14:31:01 GMT using RSA key ID 14360CDE
# gpg: Good signature from "Peter Maydell <peter.maydell@linaro.org>"
# gpg: aka "Peter Maydell <pmaydell@gmail.com>"
# gpg: aka "Peter Maydell <pmaydell@chiark.greenend.org.uk>"
* remotes/pmaydell/tags/pull-cocoa-20160323-1:
ui/cocoa.m: switch to QKeyCode
qapi-schema.json: Add power and keypad equal keys
ui/cocoa.m: fix help menus
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
This allows you to do:
./config.status --the-option-you-forgot
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Message-Id: <1452599928-7471-1-git-send-email-dgilbert@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Currently, virgl support has to go through a local unix socket, trying
to connect to a VM using -spice gl through spice://localhost:5900 will
only result in a black screen.
This commit errors out when the user tries to start a VM with both GL
support and a port/tls-port set.
This would fit better in spice-server, but currently QEMU does not call
into spice-server when parsing 'gl' on its command line, so we have to
do this check in QEMU instead.
Signed-off-by: Christophe Fergeau <cfergeau@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-id: 1457955672-28758-1-git-send-email-cfergeau@redhat.com
[ applied codestyle fix: break long line ]
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
This patches makes input-linux use -object instead of a new command line
switch. So, instead of the switch ...
-input-linux /dev/input/event$nr
... you must create an object this way:
-object input-linux,id=$name,evdev=/dev/input/event$nr
Bonus is that you can hot-add and hot-remove them via monitor now.
Suggested-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-id: 1457681901-30916-1-git-send-email-kraxel@redhat.com
commit fce55481360d "ppc: A couple more dummy POWER8 Book4 regs"
squashed in to rapidly a set of POWER8 Book4 regs in the wrong
routine. This patch introduces the missing gen_spr_power8_book4()
routine to fix their location.
Signed-off-by: Cédric Le Goater <clg@fr.ibm.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
RX buffer pools are now enabled by default for new machine types.
For older machine types, they are still disabled to avoid breaking
migration.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Laurent Vivier <lvivier@redhat.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
tl;dr:
This patch introduces an alternate way of handling the receive
buffers of the spapr-vlan device, resulting in much better
receive performance for the guest.
Full story:
One of our testers recently discovered that the performance of the
spapr-vlan device is very poor compared to other NICs, and that
a simple "ping -i 0.2 -s 65507 someip" in the guest can result
in more than 50% lost ping packets (especially with older guest
kernels < 3.17).
After doing some analysis, it was clear that there is a problem
with the way we handle the receive buffers in spapr_llan.c: The
ibmveth driver of the guest Linux kernel tries to add a lot of
buffers into several buffer pools (with 512, 2048 and 65536 byte
sizes by default, but it can be changed via the entries in the
/sys/devices/vio/1000/pool* directories of the guest). However,
the spapr-vlan device of QEMU only tries to squeeze all receive
buffer descriptors into one single page which has been supplied
by the guest during the H_REGISTER_LOGICAL_LAN call, without
taking care of different buffer sizes. This has two bad effects:
First, only a very limited number of buffer descriptors is accepted
at all. Second, we also hand 64k buffers to the guest even if
the 2k buffers would fit better - and this results in dropped packets
in the IP layer of the guest since too much skbuf memory is used.
Though it seems at a first glance like PAPR says that we should store
the receive buffer descriptors in the page that is supplied during
the H_REGISTER_LOGICAL_LAN call, chapter 16.4.1.2 in the LoPAPR spec
declares that "the contents of these descriptors are architecturally
opaque, none of these descriptors are manipulated by code above
the architected interfaces". That means we don't have to store
the RX buffer descriptors in this page, but can also manage the
receive buffers at the hypervisor level only. This is now what we
are doing here: Introducing proper RX buffer pools which are also
sorted by size of the buffers, so we can hand out a buffer with
the best fitting size when a packet has been received.
To avoid problems with migration from/to older version of QEMU,
the old behavior is also retained and enabled by default. The new
buffer management has to be enabled via a new "use-rx-buffer-pools"
property.
Now with the new buffer pool management enabled, the problem with
"ping -s 65507" is fixed for me, and the throughput of a simple
test with wget increases from creeping 3MB/s up to 20MB/s!
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Kevin Wolf
Pavel Dovgalyuk
Daniel P. Berrange
Max Reitz
Peter Xu
Programmingkid
Peter Maydell
Samuel Thibault
Bandan Das
Alberto Garcia
Paolo Bonzini
Dr. David Alan Gilbert
Christophe Fergeau
Gerd Hoffmann
Cédric Le Goater
Thomas Huth