Added functions to save and restore PCI device specific data,
specifically config space of PCI device.
Signed-off-by: Kirti Wankhede <kwankhede@nvidia.com>
Reviewed-by: Neo Jia <cjia@nvidia.com>
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
This function will be used for migration region.
Migration region is mmaped when migration starts and will be unmapped when
migration is complete.
Signed-off-by: Kirti Wankhede <kwankhede@nvidia.com>
Reviewed-by: Neo Jia <cjia@nvidia.com>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Split out walking a directory path to a separate new utility function
do_walk() and use that function in do_mkdir().
The code difference saved this way is not much, but we'll use that new
do_walk() function in the upcoming patches, so it will avoid quite
some code duplication after all.
Signed-off-by: Christian Schoenebeck <qemu_oss@crudebyte.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
Message-Id: <4d7275b2363f122438a443ce079cbb355285e9d6.1603285620.git.qemu_oss@crudebyte.com>
Signed-off-by: Christian Schoenebeck <qemu_oss@crudebyte.com>
fs_readdir_split() isn't a top level test function and thus shouldn't
take the "void *obj, void *data, QGuestAllocator *t_alloc" arguments.
Turn it into a helper to be used by test functions.
Signed-off-by: Greg Kurz <groug@kaod.org>
Message-Id: <160321016084.266767.9501523425012383531.stgit@bahia.lan>
Signed-off-by: Christian Schoenebeck <qemu_oss@crudebyte.com>
Coverity wants the return value of mkdir() to be checked:
/qemu/tests/qtest/libqos/virtio-9p.c: 48 in create_local_test_dir()
42 /* Creates the directory for the 9pfs 'local' filesystem driver to
access. */
43 static void create_local_test_dir(void)
44 {
45 struct stat st;
46
47 g_assert(local_test_path != NULL);
>>> CID 1435963: Error handling issues (CHECKED_RETURN)
>>> Calling "mkdir(local_test_path, 511U)" without checking return value.
This library function may fail and return an error code.
48 mkdir(local_test_path, 0777);
49
50 /* ensure test directory exists now ... */
51 g_assert(stat(local_test_path, &st) == 0);
52 /* ... and is actually a directory */
53 g_assert((st.st_mode & S_IFMT) == S_IFDIR);
So let's just do that and log an info-level message at least, because we
actually only care if the required directory exists and we do have an
existence check for that in place already.
Reported-by: Coverity (CID 1435963)
Signed-off-by: Christian Schoenebeck <qemu_oss@crudebyte.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
Message-Id: <03f68c7ec08064e20f43797f4eb4305ad21e1e8e.1604061839.git.qemu_oss@crudebyte.com>
Signed-off-by: Christian Schoenebeck <qemu_oss@crudebyte.com>
Use mkdtemp() to generate a unique directory for the 9p 'local' tests.
This fixes occasional 9p test failures when running 'make check -jN' if
QEMU was compiled for multiple target architectures, because the individual
architecture's test suites would run in parallel and interfere with each
other's data as the test directory was previously hard coded and hence the
same directory was used by all of them simultaniously.
This also requires a change how the test directory is created and deleted:
As the test path is now randomized and virtio_9p_register_nodes() being
called in a somewhat undeterministic way, that's no longer an appropriate
place to create and remove the test directory. Use a constructor and
destructor function for creating and removing the test directory instead.
Unfortunately libqos currently does not support setup/teardown callbacks
to handle this more cleanly.
The constructor functions needs to be in virtio-9p-test.c, not in
virtio-9p.c, because in the latter location it would cause all apps that
link to libqos (i.e. entirely unrelated test suites) to create a 9pfs
test directory as well, which would even break other test suites.
Signed-off-by: Christian Schoenebeck <qemu_oss@crudebyte.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
Tested-by: Greg Kurz <groug@kaod.org>
Message-Id: <7746f42d8f557593898d3d9d8e57c46e872dfb4f.1604243521.git.qemu_oss@crudebyte.com>
Signed-off-by: Christian Schoenebeck <qemu_oss@crudebyte.com>
pc,pci,vhost,virtio: misc fixes
Just a bunch of bugfixes all over the place.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
# gpg: Signature made Fri 30 Oct 2020 12:44:31 GMT
# gpg: using RSA key 5D09FD0871C8F85B94CA8A0D281F0DB8D28D5469
# gpg: issuer "mst@redhat.com"
# gpg: Good signature from "Michael S. Tsirkin <mst@kernel.org>" [full]
# gpg: aka "Michael S. Tsirkin <mst@redhat.com>" [full]
# Primary key fingerprint: 0270 606B 6F3C DF3D 0B17 0970 C350 3912 AFBE 8E67
# Subkey fingerprint: 5D09 FD08 71C8 F85B 94CA 8A0D 281F 0DB8 D28D 5469
* remotes/mst/tags/for_upstream:
intel_iommu: Fix two misuse of "0x%u" prints
virtio: skip guest index check on device load
vhost-blk: set features before setting inflight feature
pci: Disallow improper BAR registration for type 1
pci: Change error_report to assert(3)
pci: advertise a page aligned ATS
pc: Implement -no-hpet as sugar for -machine hpet=on
vhost: Don't special case vq->used_phys in vhost_get_log_size()
pci: Assert irqnum is between 0 and bus->nirqs in pci_bus_change_irq_level
hw/pci: Extract pci_bus_change_irq_level() from pci_change_irq_level()
hw/virtio/vhost-vdpa: Fix Coverity CID 1432864
acpi/crs: Support ranges > 32b for hosts
acpi/crs: Prevent bad ranges for host bridges
vhost-vsock: set vhostfd to non-blocking mode
vhost-vdpa: negotiate VIRTIO_NET_F_STATUS with driver
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Misc fixes
* Improve socket cnnection failure error reporting
* Fix LGPL version number
# gpg: Signature made Thu 29 Oct 2020 10:00:47 GMT
# gpg: using RSA key DAF3A6FDB26B62912D0E8E3FBE86EBB415104FDF
# gpg: Good signature from "Daniel P. Berrange <dan@berrange.com>" [full]
# gpg: aka "Daniel P. Berrange <berrange@redhat.com>" [full]
# Primary key fingerprint: DAF3 A6FD B26B 6291 2D0E 8E3F BE86 EBB4 1510 4FDF
* remotes/berrange-gitlab/tags/misc-next-pull-request:
util: include the target address in socket connect failures
io: Don't use '#' flag of printf format
authz: Fix Lesser GPL version number
crypt: Fix Lesser GPL version number
io: Fix Lesser GPL version number
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
qemu-sparc queue
# gpg: Signature made Wed 28 Oct 2020 08:17:23 GMT
# gpg: using RSA key CC621AB98E82200D915CC9C45BC2C56FAE0F321F
# gpg: issuer "mark.cave-ayland@ilande.co.uk"
# gpg: Good signature from "Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>" [full]
# Primary key fingerprint: CC62 1AB9 8E82 200D 915C C9C4 5BC2 C56F AE0F 321F
* remotes/mcayland/tags/qemu-sparc-20201028:
hw/pci-host/sabre: Simplify code initializing variable once
hw/pci-host/sabre: Remove superfluous address range check
hw/pci-host/sabre: Update documentation link
sabre: increase number of PCI bus IRQs from 32 to 64
hw/display/tcx: Allow 64-bit accesses to framebuffer stippler and blitter
sabre: don't call sysbus_mmio_map() in sabre_realize()
sparc32-ledma: don't reference nd_table directly within the device
sparc32-espdma: use object_initialize_child() for esp child object
sparc32-ledma: use object_initialize_child() for lance child object
sparc32-dma: use object_initialize_child() for espdma and ledma child objects
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Allow the server to expose an additional metacontext to be requested
by savvy clients. qemu-nbd adds a new option -A to expose the
qemu:allocation-depth metacontext through NBD_CMD_BLOCK_STATUS; this
can also be set via QMP when using block-export-add.
qemu as client is hacked into viewing the key aspects of this new
context by abusing the already-experimental x-dirty-bitmap option to
collapse all depths greater than 2, which results in a tri-state value
visible in the output of 'qemu-img map --output=json' (yes, that means
x-dirty-bitmap is now a bit of a misnomer, but I didn't feel like
renaming it as it would introduce a needless break of back-compat,
even though we make no compat guarantees with x- members):
unallocated (depth 0) => "zero":false, "data":true
local (depth 1) => "zero":false, "data":false
backing (depth 2+) => "zero":true, "data":true
libnbd as client is probably a nicer way to get at the information
without having to decipher such hacks in qemu as client. ;)
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-Id: <20201027050556.269064-11-eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
'qemu-img map' provides a way to determine which extents of an image
come from the top layer vs. inherited from a backing chain. This is
useful information worth exposing over NBD. There is a proposal to
add a QMP command block-dirty-bitmap-populate which can create a dirty
bitmap that reflects allocation information, at which point the
qemu:dirty-bitmap:NAME metadata context can expose that information
via the creation of a temporary bitmap, but we can shorten the effort
by adding a new qemu:allocation-depth metadata context that does the
same thing without an intermediate bitmap (this patch does not
eliminate the need for that proposal, as it will have other uses as
well).
While documenting things, remember that although the NBD protocol has
NBD_OPT_SET_META_CONTEXT, the rest of its documentation refers to
'metadata context', which is a more apt description of what is
actually being used by NBD_CMD_BLOCK_STATUS: the user is requesting
metadata by passing one or more context names. So I also touched up
some existing wording to prefer the term 'metadata context' where it
makes sense.
Note that this patch does not actually enable any way to request a
server to enable this context; that will come in the next patch.
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-Id: <20201027050556.269064-10-eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
When checking for allocation across a chain, it's already easy to
count the depth within the chain at which the allocation is found.
Instead of throwing that information away, return it to the caller.
Existing callers only cared about allocated/non-allocated, but having
a depth available will be used by NBD in the next patch.
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-Id: <20201027050556.269064-9-eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
[eblake: rebase to master]
Signed-off-by: Eric Blake <eblake@redhat.com>
With this, 'qemu-nbd -B b0 -B b1 -f qcow2 img.qcow2' can let you sniff
out multiple bitmaps from one server. qemu-img as client can still
only read one bitmap per client connection, but other NBD clients
(hello libnbd) can now read multiple bitmaps in a single pass.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20201027050556.269064-8-eblake@redhat.com>
Rather than open-code the count of negotiated contexts at several
sites, embed it directly into the struct. This will make it easier
for upcoming commits to support even more simultaneous contexts.
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-Id: <20201027050556.269064-7-eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Each dirty bitmap already knows its name; by reducing the scope of the
places where we construct "qemu:dirty-bitmap:NAME" strings, tracking
the name is more localized, and there are fewer per-export fields to
worry about. This in turn will make it easier for an upcoming patch
to export more than one bitmap at once.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20201027050556.269064-6-eblake@redhat.com>
Since 'block-export-add' is new to 5.2, we can still tweak the
interface; there, allowing 'bitmaps':['str'] is nicer than
'bitmap':'str'. This wires up the qapi and qemu-nbd changes to permit
passing multiple bitmaps as distinct metadata contexts that the NBD
client may request, but the actual support for more than one will
require a further patch to the server.
Note that there are no changes made to the existing deprecated
'nbd-server-add' command; this required splitting the QAPI type
BlockExportOptionsNbd, which fortunately does not affect QMP
introspection.
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-Id: <20201027050556.269064-5-eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Peter Krempa <pkrempa@redhat.com>
Rather than open-coding the translation from the deprecated
NbdServerAddOptions type to the preferred BlockExportOptionsNbd, it's
better to utilize QAPI_CLONE_MEMBERS. This solves a couple of issues:
first, if we do any more refactoring of the base type (which an
upcoming patch plans to do), we don't have to revisit the open-coding.
Second, our assignment to arg->name is fishy: the generated QAPI code
for qapi_free_NbdServerAddOptions does not visit arg->name if
arg->has_name is false, but if it DID visit it, we would have
introduced a double-free situation when arg is finally freed.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20201027050556.269064-4-eblake@redhat.com>
block.c has a useful macro QAPI_LIST_ADD() for inserting at the front
of any QAPI-generated list; move it from block.c to qapi/util.h so
more places can use it, including one earlier place in block.c, and
rename it to something more obvious (since we also have a lot of
places that append, rather than prepend, to a list).
There are many more places in the codebase that can benefit from using
the macro, but converting them will be left to later patches.
In theory, all QAPI list types are child classes of GenericList; but
in practice, that relationship is not explicitly spelled out in the C
type declarations (rather, it is something that happens implicitly due
to C compatible layouts), and the macro does not actually depend on
the GenericList type. We considered moving GenericList from visitor.h
into util.h to group related code; however, such a move would be
awkward if we do not also move GenericAlternate. Unfortunately,
moving GenericAlternate would introduce its own problems of
declaration circularity (qapi-builtin-types.h needs a complete
definition of QEnumLookup from util.h, but GenericAlternate needs a
complete definition of QType from qapi-builtin-types.h).
Suggested-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20201027050556.269064-3-eblake@redhat.com>
[eblake: s/ADD/PREPEND/ per suggestion by Markus]
There is no need to rely on the verbosity of the gcc/clang compiler
extension of g_new(typeof(X), 1) when we can instead use the standard
g_malloc(sizeof(X)). In general, we like g_new over g_malloc for
returning type X rather than void* to let the compiler catch more
potential typing mistakes, but in this particular macro, our other use
of typeof on the same line already ensures we are getting correct
results.
Suggested-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-Id: <20201027050556.269064-2-eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
nbd_server_start_unix_socket() includes an implicit nbd_server_stop(),
but we still need an explicit one at the end of the test (where there
follows no next nbd_server_start_unix_socket()), or qemu-nbd will linger
until the test exits.
This will become important when enabling this test to run on FUSE
exports, because then the export (which is the image used by qemu-nbd)
will go away before qemu-nbd exits, which will lead to qemu-nbd
complaining that it cannot flush the bitmaps in the image.
Signed-off-by: Max Reitz <mreitz@redhat.com>
Message-Id: <20201027164416.144115-3-mreitz@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
We need to let _img_info emit the format-specific information so we get
the list of bitmaps we want, but we do not need anything but the
bitmaps. So filter out everything that is irrelevant to us. (Ideally,
this would be a generalized function in common.filters that takes a list
of things to keep, but that would require implementing an anti-bitmap
filter, which would be hard, and which we do not need here. So that is
why this function is just a local hack.)
This lets 291 pass with qcow2 options like refcount_bits or data_file
again.
Fixes: 14f16bf947
("qemu-img: Support bitmap --merge into backing image")
Signed-off-by: Max Reitz <mreitz@redhat.com>
Message-Id: <20201027164416.144115-2-mreitz@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
Block layer patches:
- qcow2: Skip copy-on-write when allocating a zero cluster
- qemu-img: add support for rate limit in qemu-img convert/commit
- Fix deadlock when deleting a block node during drain_all
# gpg: Signature made Tue 27 Oct 2020 15:14:07 GMT
# gpg: using RSA key DC3DEB159A9AF95D3D7456FE7F09B272C88F2FD6
# gpg: issuer "kwolf@redhat.com"
# 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:
block: End quiescent sections when a BDS is deleted
qcow2: Skip copy-on-write when allocating a zero cluster
qcow2: Report BDRV_BLOCK_ZERO more accurately in bdrv_co_block_status()
qemu-img: add support for rate limit in qemu-img convert
qemu-img: add support for rate limit in qemu-img commit
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
QEMU must be careful when loading device state off migration streams to
prevent a malicious source from exploiting the emulator. Overdoing these
checks has the side effect of allowing a guest to "pin itself" in cloud
environments by messing with state which is entirely in its control.
Similarly to what f3081539 achieved in usb_device_post_load(), this
commit removes such a check from virtio_load(). Worth noting, the result
of a load without this check is the same as if a guest enables a VQ with
invalid indexes to begin with. That is, the virtual device is set in a
broken state (by the datapath handler) and must be reset.
Signed-off-by: Felipe Franciosi <felipe@nutanix.com>
Message-Id: <20201028134643.110698-1-felipe@nutanix.com>
Acked-by: Jason Wang <jasowang@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Prevent future developers working on root complexes, root ports, or
bridges that also wish to implement a BAR for those, from shooting
themselves in the foot. PCI type 1 headers only support 2 base address
registers. It is incorrect and difficult to figure out what is wrong
with the device when this mistake is made. With this, it is immediate
and obvious what has gone wrong.
Signed-off-by: Ben Widawsky <ben.widawsky@intel.com>
Message-Id: <20201015181411.89104-2-ben.widawsky@intel.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Asserts are used for developer bugs. As registering a bar of the wrong
size is not something that should be possible for a user to achieve,
this is a developer bug.
While here, use the more obvious helper function.
Signed-off-by: Ben Widawsky <ben.widawsky@intel.com>
Message-Id: <20201015181411.89104-1-ben.widawsky@intel.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
After Linux kernel commit 61363c1474b1 ("iommu/vt-d: Enable ATS only
if the device uses page aligned address."), ATS will be only enabled
if device advertises a page aligned request.
Unfortunately, vhost-net is the only user and we don't advertise the
aligned request capability in the past since both vhost IOTLB and
address_space_get_iotlb_entry() can support non page aligned request.
Though it's not clear that if the above kernel commit makes
sense. Let's advertise a page aligned ATS here to make vhost device
IOTLB work with Intel IOMMU again.
Note that in the future we may extend pcie_ats_init() to accept
parameters like queue depth and page alignment.
Cc: qemu-stable@nongnu.org
Signed-off-by: Jason Wang <jasowang@redhat.com>
Message-Id: <20200909081731.24688-1-jasowang@redhat.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
The first loop in vhost_get_log_size() computes the size of the dirty log
bitmap so that it allows to track changes in the entire guest memory, in
terms of GPA.
When not using a vIOMMU, the address of the vring's used structure,
vq->used_phys, is a GPA. It is thus already covered by the first loop.
When using a vIOMMU, vq->used_phys is a GIOVA that will be translated
to an HVA when the vhost backend needs to update the used structure. It
will log the corresponding GPAs into the bitmap but it certainly won't
log the GIOVA.
So in any case, vq->used_phys shouldn't be explicitly used to size the
bitmap. Drop the second loop.
This fixes a crash of the source when migrating a guest using in-kernel
vhost-net and iommu_platform=on on POWER, because DMA regions are put
over 0x800000000000000ULL. The resulting insanely huge log size causes
g_malloc0() to abort.
BugLink: https://bugzilla.redhat.com/show_bug.cgi?id=1879349
Signed-off-by: Greg Kurz <groug@kaod.org>
Message-Id: <160208823418.29027.15172801181796272300.stgit@bahia.lan>
Acked-by: Jason Wang <jasowang@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Kirti Wankhede