When migrating the 'pseries' machine type with KVM, we use a special fd
to access the hash page table stored within KVM. Usually, this fd is
opened at the beginning of migration, and kept open until the migration
is complete.
However, if there is a guest reset during the migration, the fd can become
stale and we need to re-open it. At the moment we use an 'htab_fd_stale'
flag in sPAPRMachineState to signal this, which is checked in the migration
iterators.
But that's rather ugly. It's simpler to just close and invalidate the
fd on reset, and lazily re-open it in migration if necessary. This patch
implements that change.
This requires a small addition to the machine state's instance_init,
so that htab_fd is initialized to -1 (telling the migration code it
needs to open it) instead of 0, which could be a valid fd.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Alexey Kardashevskiy <aik@ozlabs.ru>
The HMP command "info registers" produces somewhat different information on
different ppc cpu variants. For those with a hash MMU it's supposed to
include the SDR1, DAR and DSISR registers related to the MMU. However,
the switch is missing a couple of MMU model variants, meaning we will
miss out this information on certain CPUs which should have it.
This patch corrects the oversight. (Really these MMU model IDs need a big
cleanup, but we might as well fix the bug in the interim).
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Alexey Kardashevskiy <aik@ozlabs.ru>
This KVM stub implementation isn't used anywhere.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Alexey Kardashevskiy <aik@ozlabs.ru>
* qemu-char fixes from Daniel and Marc-André
* Bug fixes that break qemu-iotests
* Changes to fix reset from panicked state
* checkpatch false positives for designated initializers
* TLS support in the NBD servers and clients
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2
iQEcBAABCAAGBQJWw03lAAoJEL/70l94x66D/0MH/3Nctz5y1GKgAX0i6rKErV3/
hvPt6JHdWd7uBtowzO5kOy3fyOnVVST6jNHMQPAmJplUC40s6Ca0hycw9TjdJUdu
ULq0Ba7tQ1TAXowDqibtEn+iTkzSrocTJLfEglNscKzJ4y5w0vc5Bt5PgPB65mbn
oTo/YR8KyRWS6rXjyNnKb0PCaYEQziBndjuIxp9yJUsLcw1UgQJVcrUNEIiciOWu
SlWDxvJJQt5cCrTPnUXeBdjJVGaLxbcpe2llEJnIuf6Pjq4u7J0y+DJ40y0DCi9q
v3V4r16HhAGmBZNIlCtbClfhb/sRTVhONQiS9ehhROo8QCaL1psc11HWvMJ0tx0=
=CDoZ
-----END PGP SIGNATURE-----
Merge remote-tracking branch 'remotes/bonzini/tags/for-upstream' into staging
* Coverity fixes for IPMI and mptsas
* qemu-char fixes from Daniel and Marc-André
* Bug fixes that break qemu-iotests
* Changes to fix reset from panicked state
* checkpatch false positives for designated initializers
* TLS support in the NBD servers and clients
# gpg: Signature made Tue 16 Feb 2016 16:27:17 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)
nbd: enable use of TLS with nbd-server-start command
nbd: enable use of TLS with qemu-nbd server
nbd: enable use of TLS with NBD block driver
nbd: implement TLS support in the protocol negotiation
nbd: use "" as a default export name if none provided
nbd: always query export list in fixed new style protocol
nbd: allow setting of an export name for qemu-nbd server
nbd: make client request fixed new style if advertised
nbd: make server compliant with fixed newstyle spec
nbd: invert client logic for negotiating protocol version
nbd: convert to using I/O channels for actual socket I/O
nbd: convert blockdev NBD server to use I/O channels for connection setup
nbd: convert qemu-nbd server to use I/O channels for connection setup
nbd: convert block client to use I/O channels for connection setup
qemu-nbd: add support for --object command line arg
qom: add helpers for UserCreatable object types
ipmi: sensor number should not exceed MAX_SENSORS
mptsas: fix wrong formula
mptsas: fix memory leak
mptsas: add missing va_end
...
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
This modifies the nbd-server-start QMP command so that it
is possible to request use of TLS. This is done by adding
a new optional parameter "tls-creds" which provides the ID
of a previously created QCryptoTLSCreds object instance.
TLS is only supported when using an IPv4/IPv6 socket listener.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-Id: <1455129674-17255-17-git-send-email-berrange@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
This modifies the qemu-nbd program so that it is possible to
request the use of TLS with the server. It simply adds a new
command line option --tls-creds which is used to provide the
ID of a QCryptoTLSCreds object previously created via the
--object command line option.
For example
qemu-nbd --object tls-creds-x509,id=tls0,endpoint=server,\
dir=/home/berrange/security/qemutls \
--tls-creds tls0 \
--exportname default
TLS requires the new style NBD protocol, so if no export name
is set (via --export-name), then we use the default NBD protocol
export name ""
TLS is only supported when using an IPv4/IPv6 socket listener.
It is not possible to use with UNIX sockets, which includes
when connecting the NBD server to a host device.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-Id: <1455129674-17255-16-git-send-email-berrange@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
This modifies the NBD driver so that it is possible to request
use of TLS. This is done by providing the 'tls-creds' parameter
with the ID of a previously created QCryptoTLSCreds object.
For example
$QEMU -object tls-creds-x509,id=tls0,endpoint=client,\
dir=/home/berrange/security/qemutls \
-drive driver=nbd,host=localhost,port=9000,tls-creds=tls0
The client will drop the connection if the NBD server does not
provide TLS.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-Id: <1455129674-17255-15-git-send-email-berrange@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
This extends the NBD protocol handling code so that it is capable
of negotiating TLS support during the connection setup. This involves
requesting the STARTTLS protocol option before any other NBD options.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-Id: <1455129674-17255-14-git-send-email-berrange@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
If the user does not provide an export name and the server
is running the new style protocol, where export names are
mandatory, use "" as the default export name if the user
has not specified any. "" is defined in the NBD protocol
as the default name to use in such scenarios.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-Id: <1455129674-17255-13-git-send-email-berrange@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
With the new style protocol, the NBD client will currenetly
send NBD_OPT_EXPORT_NAME as the first (and indeed only)
option it wants. The problem is that the NBD protocol spec
does not allow for returning an error message with the
NBD_OPT_EXPORT_NAME option. So if the server mandates use
of TLS, the client will simply see an immediate connection
close after issuing NBD_OPT_EXPORT_NAME which is not user
friendly.
To improve this situation, if we have the fixed new style
protocol, we can sent NBD_OPT_LIST as the first option
to query the list of server exports. We can check for our
named export in this list and raise an error if it is not
found, instead of going ahead and sending NBD_OPT_EXPORT_NAME
with a name that we know will be rejected.
This improves the error reporting both in the case that the
server required TLS, and in the case that the client requested
export name does not exist on the server.
If the server does not support NBD_OPT_LIST, we just ignore
that and carry on with NBD_OPT_EXPORT_NAME as before.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-Id: <1455129674-17255-12-git-send-email-berrange@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The qemu-nbd server currently always uses the old style protocol
since it never sets any export name. This is a problem because
future TLS support will require use of the new style protocol
negotiation.
This adds "--exportname NAME" / "-x NAME" arguments to qemu-nbd
which allow the user to set an explicit export name. When an
export name is set the server will always use the new style
NBD protocol.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-Id: <1455129674-17255-11-git-send-email-berrange@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
If the server advertises support for the fixed new style
negotiation, the client should in turn enable new style.
This will allow the client to negotiate further NBD
options besides the export name.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-Id: <1455129674-17255-10-git-send-email-berrange@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
If the client does not request the fixed new style protocol,
then we should only accept NBD_OPT_EXPORT_NAME. All other
options are only valid when fixed new style has been activated.
The qemu-nbd client doesn't currently request fixed new style
protocol, but this change won't break qemu-nbd, because it
fortunately only ever uses NBD_OPT_EXPORT_NAME, so was never
triggering the non-compliant server behaviour.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-Id: <1455129674-17255-9-git-send-email-berrange@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The nbd_receive_negotiate() method takes different code
paths based on whether 'name == NULL', and then checks
the expected protocol version in each branch.
This patch inverts the logic, so that it takes different
code paths based on what protocol version it receives and
then checks if name is NULL or not as needed.
This facilitates later code which allows the client to
be capable of using the new style protocol regardless
of whether an export name is listed or not.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-Id: <1455129674-17255-8-git-send-email-berrange@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Now that all callers are converted to use I/O channels for
initial connection setup, it is possible to switch the core
NBD protocol handling core over to use QIOChannel APIs for
actual sockets I/O.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-Id: <1455129674-17255-7-git-send-email-berrange@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
This converts the blockdev NBD server to use the QIOChannelSocket
class for initial listener socket setup and accepting of client
connections. Actual I/O is still being performed against the
socket file descriptor using the POSIX socket APIs.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-Id: <1455129674-17255-6-git-send-email-berrange@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
This converts the qemu-nbd server to use the QIOChannelSocket
class for initial listener socket setup and accepting of client
connections. Actual I/O is still being performed against the
socket file descriptor using the POSIX socket APIs.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-Id: <1455129674-17255-5-git-send-email-berrange@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
This converts the NBD block driver client to use the QIOChannelSocket
class for initial connection setup. The NbdClientSession struct has
two pointers, one to the master QIOChannelSocket providing the raw
data channel, and one to a QIOChannel which is the current channel
used for I/O. Initially the two point to the same object, but when
TLS support is added, they will point to different objects.
The qemu-img & qemu-io tools now need to use MODULE_INIT_QOM to
ensure the QIOChannel object classes are registered. The qemu-nbd
tool already did this.
In this initial conversion though, all I/O is still actually done
using the raw POSIX sockets APIs.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-Id: <1455129674-17255-4-git-send-email-berrange@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Allow creation of user creatable object types with qemu-nbd
via a new --object command line arg. This will be used to supply
passwords and/or encryption keys to the various block driver
backends via the recently added 'secret' object type.
# printf letmein > mypasswd.txt
# qemu-nbd --object secret,id=sec0,file=mypasswd.txt \
...other nbd args...
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-Id: <1455129674-17255-3-git-send-email-berrange@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The QMP monitor code has two helper methods object_add
and qmp_object_del that are called from several places
in the code (QMP, HMP and main emulator startup).
The HMP and main emulator startup code also share
further logic that extracts the qom-type & id
values from a qdict.
We soon need to use this logic from qemu-img, qemu-io
and qemu-nbd too, but don't want those to depend on
the monitor, nor do we want to duplicate the code.
To avoid this, move some code out of qmp.c and hmp.c
adding new methods to qom/object_interfaces.c
- user_creatable_add - takes a QDict holding a full
object definition & instantiates it
- user_creatable_add_type - takes an ID, type name,
and QDict holding object properties & instantiates
it
- user_creatable_add_opts - takes a QemuOpts holding
a full object definition & instantiates it
- user_creatable_add_opts_foreach - variant on
user_creatable_add_opts which can be directly used
in conjunction with qemu_opts_foreach.
- user_creatable_del - takes an ID and deletes the
corresponding object
The existing code is updated to use these new methods.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-Id: <1455129674-17255-2-git-send-email-berrange@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Fix a number of off-by-ones, one of them spotted by Coverity.
Signed-off-by: Cédric Le Goater <clg@fr.ibm.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
MPI_DOORBELL_WHO_INIT_SHIFT is being repeated twice. Reported
by Coverity.
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
This can cause various segmentation faults or aborts in qemu-iotests
test 091.
Fixes: 5b82b703b6
Cc: Dave Gilbert <dgilbert@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The right place for "work around issues with system headers" code
is osdep.h. Move the workaround for OSX's stdlib.h emitting a
deprecation warning for daemon() to that header.
This also fixes a problem where running clean-includes on
oslib-posix.c would erroneously remove the #include <stdlib.h>
from it, breaking the workaround.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Eric Blake <eblake@redhat.com>
Clean up includes so that osdep.h is included first and headers
which it implies are not included manually.
This commit was created with scripts/clean-includes.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Eric Blake <eblake@redhat.com>
Include qemu/osdep.h as the first include in generated .c files,
so they don't implicitly rely on some other included header
to pull it in.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Eric Blake <eblake@redhat.com>
In the .c files generated by this script, include qemu/osdep.h
as the first included header, not config.h.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Eric Blake <eblake@redhat.com>
As a followup to commit cbf2115, clean up the includes in files
generated by QAPI so that osdep.h is included first in .c files,
and headers which it implies are not included manually. This
patch is done manually, since Coccinelle (and therefore
scripts/clean-includes) doesn't see into the generator scripts.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Eric Blake <eblake@redhat.com>
Clean up includes so that osdep.h is included first and headers
which it implies are not included manually.
This commit was created with scripts/clean-includes.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Eric Blake <eblake@redhat.com>
Tested-by: Eric Blake <eblake@redhat.com>
Don't define ARRAY_SIZE locally; instead include osdep.h for it.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Clean up includes so that osdep.h is included first and headers
which it implies are not included manually.
This commit was created with scripts/clean-includes.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Eric Blake <eblake@redhat.com>
Clean up includes so that osdep.h is included first and headers
which it implies are not included manually.
This commit was created with scripts/clean-includes.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Eric Blake <eblake@redhat.com>
mmu.c is only built for CONFIG_SOFTMMU targets, so there is
no need to redundantly surround the whole file contents with
an #ifndef CONFIG_USER_ONLY. The ifdef also confuses the
Coccinelle tool.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Eric Blake <eblake@redhat.com>
Include osdep.h as the first header in nand.c; this has to be
done manually because coccinelle gets confused by the way that
this C file includes itself.
We fix some odd spacing in #includes while we are in the area.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Eric Blake <eblake@redhat.com>
Actively redefining 'inline' is wrong for C++, where gcc has an
extension 'inline namespace' which fails to compile if the
keyword 'inline' is replaced by a macro expansion. This will
matter once we start to include "qemu/osdep.h" first from C++
files, depending also on whether the system headers are new
enough to be using the gcc extension.
But rather than just guard things by __cplusplus, let's look at
the overall picture. Commit df2542c737 in 2007 defined 'inline'
to the gcc attribute __always_inline__, with the rationale "To
avoid discarded inlining bug". But compilers have improved since
then, and we are probably better off trusting the compiler rather
than trying to force its hand.
So just nuke our craziness.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 1455043788-28112-1-git-send-email-eblake@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Since it can`t fail. Also modify the callers.
Signed-off-by: Cao jin <caoj.fnst@cn.fujitsu.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Marcel Apfelbaum <marcel@redhat.com>
Also because pci_bridge_initfn() can`t fail.
Signed-off-by: Cao jin <caoj.fnst@cn.fujitsu.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Marcel Apfelbaum <marcel@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
The test is based on bios-tables-test.c. It creates a file with
the boot sector image and loads it into a guest using PXE and TFTP
functionality.
Cc: Jason Wang <jasowang@redhat.com>
Signed-off-by: Victor Kaplansky <victork@redhat.com>
Suggested-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
commit 428c3ece97 ("fix MSI injection on Xen")
inadvertently enabled the xen-specific logic unconditionally.
Limit it to only when xen is enabled.
Additionally, msix data should be read with pci_get_log
since the format is pci little-endian.
Reported-by: "Daniel P. Berrange" <berrange@redhat.com>
Cc: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
When adding cross-endian support, we introduced the TARGET_IS_BIENDIAN macro
and the virtio_access_is_big_endian() helper to have a branchless fast path
in the virtio memory accessors for targets that don't switch endian.
This was considered as a strong requirement at the time.
Now we have added a runtime check for virtio 1.0, which ruins the benefit
of the virtio_access_is_big_endian() helper for always little-endian targets.
With this patch, always little-endian targets stop checking for virtio 1.0,
since the result is little-endian in all cases.
Reviewed-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Reviewed-by: Laurent Vivier <lvivier@redhat.com>
Signed-off-by: Greg Kurz <gkurz@linux.vnet.ibm.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Laurent Vivier <lvivier@redhat.com>
After the call to virtio_vdev_has_feature(), we only care for legacy
devices, so we don't need the extra check in virtio_is_big_endian().
Also the device_endian field is always set (VIRTIO_DEVICE_ENDIAN_UNKNOWN
may only happen on a virtio_load() path that cannot lead here), so we
don't need the assert() either.
This open codes the device_endian checking in vhost_needs_vring_endian().
It also adds a comment to explain the logic, as recent reviews showed the
cross-endian tweaks aren't that obvious.
Signed-off-by: Greg Kurz <gkurz@linux.vnet.ibm.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Laurent Vivier <lvivier@redhat.com>
Reviewed-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Indeed vhost doesn't need to ask for vring endian fixing if the device is
virtio 1.0, since it is already handled by the in-kernel vhost driver. This
patch simply consolidates the logic into the existing helper.
Reviewed-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: Greg Kurz <gkurz@linux.vnet.ibm.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Laurent Vivier <lvivier@redhat.com>
If target is bi-endian (ppc64, arm), the virtio_legacy_is_cross_endian()
indeed returns the runtime state of the virtio device. However, it returns
false unconditionally in the general case. This sounds a bit strange
given the name of the function.
This helper is only useful for vhost actually, where indeed non bi-endian
targets don't have to deal with cross-endian issues.
This patch moves the helper to vhost.c and gives it a more appropriate name.
Reviewed-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Reviewed-by: Laurent Vivier <lvivier@redhat.com>
Signed-off-by: Greg Kurz <gkurz@linux.vnet.ibm.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Laurent Vivier <lvivier@redhat.com>
Cross-endian is now handled by the core virtio-net code.
This patch reverts:
commit 5be7d9f1b1
vhost-net: tell tap backend about the vnet endianness
and
commit cf0a628f6e81bfc9b7a944fa0b80c3594836df56
net: set endianness on all backend devices
Reviewed-by: Laurent Vivier <lvivier@redhat.com>
Signed-off-by: Greg Kurz <gkurz@linux.vnet.ibm.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Laurent Vivier <lvivier@redhat.com>
When running a fully emulated device in cross-endian conditions, including
a virtio 1.0 device offered to a big endian guest, we need to fix the vnet
headers. This is currently handled by the virtio_net_hdr_swap() function
in the core virtio-net code but it should actually be handled by the net
backend.
With this patch, virtio-net now tries to configure the backend to do the
endian fixing when the device starts (i.e. drivers sets the CONFIG_OK bit).
If the backend cannot support the requested endiannes, we have to fallback
onto virtio_net_hdr_swap(): this is recorded in the needs_vnet_hdr_swap flag,
to be used in the TX and RX paths.
Note that we reset the backend to the default behaviour (guest native
endianness) when the device stops (i.e. device status had CONFIG_OK bit and
driver unsets it). This is needed, with the linux tap backend at least,
otherwise the guest may lose network connectivity if rebooted into a
different endianness.
The current vhost-net code also tries to configure net backends. This will
be no more needed and will be reverted in a subsequent patch.
Reviewed-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Reviewed-by: Laurent Vivier <lvivier@redhat.com>
Signed-off-by: Greg Kurz <gkurz@linux.vnet.ibm.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Laurent Vivier <lvivier@redhat.com>
Reproducer is simply to migrate a virtual machine that was started with -S,
or that was already migrated.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
David Gibson