If we have a system with xenforeignmemory_map2() implemented
we don't need to save/restore physmap on suspend/restore
anymore. In case we resume a VM without physmap - try to
recreate the physmap during memory region restore phase and
remap map cache entries accordingly. The old code is left
for compatibility reasons.
Signed-off-by: Igor Druzhinin <igor.druzhinin@citrix.com>
Reviewed-by: Paul Durrant <paul.durrant@citrix.com>
Reviewed-by: Stefano Stabellini <sstabellini@kernel.org>
Signed-off-by: Stefano Stabellini <sstabellini@kernel.org>
This new call is trying to update a requested map cache entry
according to the changes in the physmap. The call is searching
for the entry, unmaps it and maps again at the same place using
a new guest address. If the mapping is dummy this call will
make it real.
This function makes use of a new xenforeignmemory_map2() call
with an extended interface that was recently introduced in
libxenforeignmemory [1].
[1] https://www.mail-archive.com/xen-devel@lists.xen.org/msg113007.html
Signed-off-by: Igor Druzhinin <igor.druzhinin@citrix.com>
Reviewed-by: Paul Durrant <paul.durrant@citrix.com>
Reviewed-by: Stefano Stabellini <sstabellini@kernel.org>
Signed-off-by: Stefano Stabellini <sstabellini@kernel.org>
Dummys are simple anonymous mappings that are placed instead
of regular foreign mappings in certain situations when we need
to postpone the actual mapping but still have to give a
memory region to QEMU to play with.
This is planned to be used for restore on Xen.
Signed-off-by: Igor Druzhinin <igor.druzhinin@citrix.com>
Reviewed-by: Paul Durrant <paul.durrant@citrix.com>
Reviewed-by: Stefano Stabellini <sstabellini@kernel.org>
Commit 090fa1c8 "add support for unplugging NVMe disks..." extended the
existing disk unplug flag to cover NVMe disks as well as IDE and SCSI.
The recent thread on the xen-devel mailing list [1] has highlighted that
this is not desirable behaviour: PV frontends should be able to distinguish
NVMe disks from other types of disk and should have separate control over
whether they are unplugged.
This patch defines a new bit in the unplug mask for this purpose (see Xen
commit [2]) and also tidies up the definitions of, and improves the
comments regarding, the previously exiting bits in the protocol.
[1] https://lists.xen.org/archives/html/xen-devel/2017-03/msg02924.html
[2] http://xenbits.xen.org/gitweb/?p=xen.git;a=commit;h=1096aa02
Signed-off-by: Paul Durrant <paul.durrant@citrix.com>
Reviewed-by: Stefano Stabellini <sstabellini@kernel.org>
Signed-off-by: Stefano Stabellini <sstabellini@kernel.org>
Check the return status of the xen_host_pci_get_* functions we call in
xen_pt_msix_init(), and fail device init if the reads failed rather than
ploughing ahead. (Spotted by Coverity: CID 777338.)
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Stefano Stabellini <sstabellini@kernel.org>
Signed-off-by: Stefano Stabellini <sstabellini@kernel.org>
In igd passthrough environment, guest could only access opregion at the
first bootup time. Once guest shutdown, later guest couldn't access
opregion anymore.
This is because qemu set emulated guest opregion base address to host
register. Later guest get a wrong host opregion base address, and couldn't
access it anymore.
This patch set emu_mask for igd_opregion register, so guest won't set
guest opregion base address to host.
Signed-off-by: Xiong Zhang <xiong.y.zhang@intel.com>
Acked-by: Anthony PERARD <anthony.perard@citrix.com>
Signed-off-by: Stefano Stabellini <sstabellini@kernel.org>
-----BEGIN PGP SIGNATURE-----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=Vt6Z
-----END PGP SIGNATURE-----
Merge remote-tracking branch 'remotes/ehabkost/tags/x86-and-machine-pull-request' into staging
x86 and machine queue, 2017-07-17
# gpg: Signature made Mon 17 Jul 2017 19:46:14 BST
# gpg: using RSA key 0x2807936F984DC5A6
# gpg: Good signature from "Eduardo Habkost <ehabkost@redhat.com>"
# Primary key fingerprint: 5A32 2FD5 ABC4 D3DB ACCF D1AA 2807 936F 984D C5A6
* remotes/ehabkost/tags/x86-and-machine-pull-request:
qmp: Include parent type on 'qom-list-types' output
qmp: Include 'abstract' field on 'qom-list-types' output
tests: Simplify abstract-interfaces check with a helper
i386: add Skylake-Server cpu model
i386: Update comment about XSAVES on Skylake-Client
i386: expose "TCGTCGTCGTCG" in the 0x40000000 CPUID leaf
fw_cfg: move QOM type defines and fw_cfg types into fw_cfg.h
fw_cfg: move qdev_init_nofail() from fw_cfg_init1() to callers
fw_cfg: switch fw_cfg_find() to locate the fw_cfg device by type rather than path
qom: Fix ambiguous path detection when ambiguous=NULL
Revert "machine: Convert abstract typename on compat_props to subclass names"
test-qdev-global-props: Test global property ordering
qdev: fix the order compat and global properties are applied
tests: Test case for object_resolve_path*()
device-crash-test: Fix regexp on whitelist
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
* new model of the ARM MPS2/MPS2+ FPGA based development board
* clean up DISAS_* exit conditions and fix various regressions
since commits e75449a3468a6b28c7b5 (in particular including
ones which broke OP-TEE guests)
* make Cortex-M3 and M4 correctly default to 8 PMSA regions
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1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=gVxi
-----END PGP SIGNATURE-----
Merge remote-tracking branch 'remotes/pmaydell/tags/pull-target-arm-20170717' into staging
target-arm queue:
* new model of the ARM MPS2/MPS2+ FPGA based development board
* clean up DISAS_* exit conditions and fix various regressions
since commits e75449a3468a6b28c7b5 (in particular including
ones which broke OP-TEE guests)
* make Cortex-M3 and M4 correctly default to 8 PMSA regions
# gpg: Signature made Mon 17 Jul 2017 13:43:45 BST
# gpg: using RSA key 0x3C2525ED14360CDE
# 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>"
# Primary key fingerprint: E1A5 C593 CD41 9DE2 8E83 15CF 3C25 25ED 1436 0CDE
* remotes/pmaydell/tags/pull-target-arm-20170717:
MAINTAINERS: Add entries for MPS2 board
hw/arm/mps2: Add ethernet
hw/arm/mps2: Add SCC
hw/misc/mps2_scc: Implement MPS2 Serial Communication Controller
hw/arm/mps2: Add timers
hw/char/cmsdk-apb-timer: Implement CMSDK APB timer device
hw/arm/mps2: Add UARTs
hw/char/cmsdk-apb-uart.c: Implement CMSDK APB UART
hw/arm/mps2: Implement skeleton mps2-an385 and mps2-an511 board models
target/arm: use DISAS_EXIT for eret handling
target/arm: use gen_goto_tb for ISB handling
target/arm/translate: ensure gen_goto_tb sets exit flags
target/arm/translate.h: expand comment on DISAS_EXIT
target/arm/translate: make DISAS_UPDATE match declared semantics
include/exec/exec-all: document common exit conditions
target/arm: Make Cortex-M3 and M4 default to 8 PMSA regions
qdev: support properties which don't set a default value
qdev-properties.h: Explicitly set the default value for arraylen properties
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Version: GnuPG v1
iQEcBAABAgAGBQJZbKrNAAoJEO8Ells5jWIRzWkH/0GgEkT5XqSPR8gTixxR5+aT
1+LvlqI861/oR3aZ/1+6nzbFF4RBHO0TJb9v8HovfaOMU/tjaVMOGOD98+rqToa7
2P2BTQo5jfsQhzGj2GBWnjpTqYunUjXdT0jjZAdERGqrNjoFOGhAjFXPvTKL23d5
haDgRQgTh2z4w+rvuHNQ79S8tCDtUGvH1i9fIpWNnVLlv4Lea8XJlm7p2+jNQslF
W2ysoQ6PR/3HihtqMwsh4ZBJAQfhEpJcrcLeq5wWEdg40U2JVA1MjpX0q58X6fRJ
YQ36K0vxmdnxdCK6NnoMLkGqI12aRqJnFEq0Avc3dC2U0OWIfNk4mp8X0Vr8o+s=
=z3QF
-----END PGP SIGNATURE-----
Merge remote-tracking branch 'remotes/jasowang/tags/net-pull-request' into staging
# gpg: Signature made Mon 17 Jul 2017 13:17:17 BST
# gpg: using RSA key 0xEF04965B398D6211
# gpg: Good signature from "Jason Wang (Jason Wang on RedHat) <jasowang@redhat.com>"
# gpg: WARNING: This key is not certified with a trusted signature!
# gpg: There is no indication that the signature belongs to the owner.
# Primary key fingerprint: 215D 46F4 8246 689E C77F 3562 EF04 965B 398D 6211
* remotes/jasowang/tags/net-pull-request:
virtio-net: fix offload ctrl endian
virtion-net: Prefer is_power_of_2()
docs/colo-proxy.txt: Update colo-proxy usage of net driver with vnet_header
net/filter-rewriter.c: Make filter-rewriter support vnet_hdr_len
net/colo-compare.c: Add vnet packet's tcp/udp/icmp compare
net/colo.c: Add vnet packet parse feature in colo-proxy
net/colo-compare.c: Make colo-compare support vnet_hdr_len
net/colo-compare.c: Introduce parameter for compare_chr_send()
net/colo.c: Make vnet_hdr_len as packet property
net/filter-mirror.c: Add new option to enable vnet support for filter-redirector
net/filter-mirror.c: Make filter mirror support vnet support.
net/filter-mirror.c: Introduce parameter for filter_send()
net/net.c: Add vnet_hdr support in SocketReadState
net: Add vnet_hdr_len arguments in NetClientState
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
By exposing FWCfgIoState and FWCfgMemState internals we allow the possibility
for the internal MemoryRegion fields to be mapped by name for boards that wish
to wire up the fw_cfg device themselves.
Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Message-Id: <1500025208-14827-4-git-send-email-mark.cave-ayland@ilande.co.uk>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
When looking to instantiate a TYPE_FW_CFG_MEM or TYPE_FW_CFG_IO device to be
able to wire it up differently, it is much more convenient for the caller to
instantiate the device and have the fw_cfg default files already preloaded
during realize.
Move fw_cfg_init1() to the end of both the fw_cfg_mem_realize() and
fw_cfg_io_realize() functions so it no longer needs to be called manually
when instantiating the device, and also rename it to fw_cfg_common_realize()
which better describes its new purpose.
Since it is now the responsibility of the machine to wire up the fw_cfg device
it is necessary to introduce a object_property_add_child() call into
fw_cfg_init_io() and fw_cfg_init_mem() to link the fw_cfg device to the root
machine object as before.
Finally with the previous change to fw_cfg_find() we can now remove the
assert() preventing multiple fw_cfg devices being instantiated and replace
them with a simple call to fw_cfg_find() at realize time instead. This allows
us to remove FW_CFG_NAME and FW_CFG_PATH since they are no longer required.
Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Reviewed-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <1500025208-14827-3-git-send-email-mark.cave-ayland@ilande.co.uk>
Reviewed-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
This will enable the fw_cfg device to be placed anywhere within the QOM tree
regardless of its machine location.
Note that we also add a comment to document the behaviour that we return NULL to
indicate failure where either no fw_cfg device or multiple fw_cfg devices are
found.
Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Message-Id: <1500025208-14827-2-git-send-email-mark.cave-ayland@ilande.co.uk>
Reviewed-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
This reverts commit 0bcba41fe3.
The bug addressed by that commit is now fixed in a better way by the
commit "qdev: fix the order compat and global properties are applied".
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Message-Id: <20170711004303.3902-4-ehabkost@redhat.com>
Acked-by: Cornelia Huck <cohuck@redhat.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
Reviewed-by: Halil Pasic <pasic@linux.vnet.ibm.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
The current code recursively applies global properties from child up to
parent types. This can cause properties passed with the -global option to
be silently overridden by internal compat properties.
This is exactly what happened with virtio-*-pci drivers since commit:
"9a4c0e220d8a hw/virtio-pci: fix virtio behaviour"
Passing -device virtio-blk-pci.disable-modern=off had no effect on 2.6
machine types because the internal virtio-pci.disable-modern=on compat
property always prevailed.
A workaround for this was included with commit 0bcba41f ("machine:
Convert abstract typename on compat_props to subclass names").
This patch fixes the issue properly by reversing the logic: we now go
through the global property list and, for each property, we check if it
is applicable to the device.
This results in compat properties being applied first, in the order they
appear in the HW_COMPAT_* macros, followed by global properties, in the
order they appear on the command line.
Signed-off-by: Greg Kurz <groug@kaod.org>
Message-Id: <148103887228.22326.478406873609299999.stgit@bahia.lab.toulouse-stg.fr.ibm.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Message-Id: <20170711004303.3902-2-ehabkost@redhat.com>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Reviewed-by: Halil Pasic <pasic@linux.vnet.ibm.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
At the moment VFIO PCI device initialization works as follows:
vfio_realize
vfio_get_group
vfio_connect_container
register memory listeners (1)
update QEMU groups lists
vfio_kvm_device_add_group
Then (example for pseries) the machine reset hook triggers region_add()
for all regions where listeners from (1) are listening:
ppc_spapr_reset
spapr_phb_reset
spapr_tce_table_enable
memory_region_add_subregion
vfio_listener_region_add
vfio_spapr_create_window
This scheme works fine until we need to handle VFIO PCI device hotplug
and we want to enable PPC64/sPAPR in-kernel TCE acceleration on,
i.e. after PCI hotplug we need a place to call
ioctl(vfio_kvm_device_fd, KVM_DEV_VFIO_GROUP_SET_SPAPR_TCE).
Since the ioctl needs a LIOBN fd (from sPAPRTCETable) and a IOMMU group fd
(from VFIOGroup), vfio_listener_region_add() seems to be the only place
for this ioctl().
However this only works during boot time because the machine reset
happens strictly after all devices are finalized. When hotplug happens,
vfio_listener_region_add() is called when a memory listener is registered
but when this happens:
1. new group is not added to the container->group_list yet;
2. VFIO KVM device is unaware of the new IOMMU group.
This moves bits around to have all necessary VFIO infrastructure
in place for both initial startup and hotplug cases.
[aw: ie, register vfio groups with kvm prior to memory listener
registration such that kvm-vfio pseudo device ioctls are available
during the region_add callback]
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
The MPS2 FPGA images support ethernet via a LAN9220. We use
QEMU's LAN9118 model, which is software compatible except
that it is missing the checksum-offload feature.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 1500029487-14822-9-git-send-email-peter.maydell@linaro.org
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Implement a model of the Serial Communication Controller (SCC) found
in MPS2 FPGA images.
The primary purpose of this device is to communicate with the
Motherboard Configuration Controller (MCC) which is located on
the MPS board itself, outside the FPGA image. This is used
for programming the MPS clock generators. The SCC also has
some basic ID registers and an output for the board LEDs.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-id: 1500029487-14822-7-git-send-email-peter.maydell@linaro.org
Implement a model of the simple timer device found in the CMSDK.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-id: 1500029487-14822-5-git-send-email-peter.maydell@linaro.org
Add the UARTs to the MPS2 board models.
Unfortunately the details of the wiring of the interrupts through
various OR gates differ between AN511 and AN385 so this can't
be purely a data-driven difference.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Alistair Francis <alistair.francis@xilinx.com>
Message-id: 1500029487-14822-4-git-send-email-peter.maydell@linaro.org
Implement a model of the simple "APB UART" provided in
the Cortex-M System Design Kit (CMSDK).
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 1500029487-14822-3-git-send-email-peter.maydell@linaro.org
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Model the ARM MPS2/MPS2+ FPGA based development board.
The MPS2 and MPS2+ dev boards are FPGA based (the 2+ has a bigger
FPGA but is otherwise the same as the 2). Since the CPU itself
and most of the devices are in the FPGA, the details of the board
as seen by the guest depend significantly on the FPGA image.
We model the following FPGA images:
"mps2_an385" -- Cortex-M3 as documented in ARM Application Note AN385
"mps2_an511" -- Cortex-M3 'DesignStart' as documented in AN511
They are fairly similar but differ in the details for some
peripherals.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 1500029487-14822-2-git-send-email-peter.maydell@linaro.org
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
In some situations it's useful to have a qdev property which doesn't
automatically set its default value when qdev_property_add_static is
called (for instance when the default value is not constant).
Support this by adding a flag to the Property struct indicating
whether to set the default value. This replaces the existing test
for whether the PropertyInfo set_default_value function pointer is
NULL, and we set the .set_default field to true for all those cases
of struct Property which use a PropertyInfo with a non-NULL
set_default_value, so behaviour remains the same as before.
This gives us the semantics of:
* if .set_default is true, then .info->set_default_value must
be not NULL, and .defval is used as the the default value of
the property
* otherwise, the property system does not set any default, and
the field will retain whatever initial value it was given by
the device's .instance_init method
We define two new macros DEFINE_PROP_SIGNED_NODEFAULT and
DEFINE_PROP_UNSIGNED_NODEFAULT, to cover the most plausible use cases
of wanting to set an integer property with no default value.
Suggested-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Message-id: 1499788408-10096-3-git-send-email-peter.maydell@linaro.org
Spec said offloads should be le64, so use virtio_ldq_p() to guarantee
valid endian.
Fixes: 644c98587d ("virtio-net: dynamic network offloads configuration")
Cc: qemu-stable@nongnu.org
Cc: Dmitry Fleytman <dfleytma@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
We have a function that checks if given number is power of two.
We should prefer it instead of expanding the check on our own.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
This pull requests supersedes the one from 2017-07-14. That one had a
couple of subtle regressions: there was a build error for mingw32, and
an instance_size which was theoretically wrong everywhere, but only
actually bit on the Travis OSX build.
There are two major batches in this set, rather than the usual
collection of assorted fixes.
* More DRC cleanup. This gets the state management into a state
which should fix many of the hotplug+migration problems we've
had. Plus it gets the migration stream format into something
well defined and pretty minimal which we can reasonably support
into the future.
* Hashed Page Table resizing. It's been a while since this was
posted, but it's been through several previous rounds of review.
The kernel parts (both guest and host) are merged in 4.11, so
this is the only remaining piece left to allow resizing of the
HPT in a running guest.
There are also a handful of unrelated fixes.
-----BEGIN PGP SIGNATURE-----
iQIzBAABCAAdFiEEdfRlhq5hpmzETofcbDjKyiDZs5IFAllsWwQACgkQbDjKyiDZ
s5LMnA//dpoqWrTPiEmx2DsXMkjLefn/2Yl1dkQDzhyb7v+tNGFYmxpbb7nPRfJE
tfvcKu1Tz23NPOp6+1VC9eTyTO1YOXTgvQrNSbF1MmIg4PGN6s2DHrLviAqCS15M
29x6+RdRaeLUSCsk8elsViiWb8h7cISDuN0SMA0WWjWP3bO/drz5nq5z5dRgdVFe
Z5O0qwDNoN0NypJ68Cld+riP1uDAYMONPxA0QOWCLx8qowoJ3hYMuyNnqBQU5OJn
PpAA3EfdxkN6rtaBjDt7xHkJfm9Xkm9SsT8qTcj/R2JjkENef8EbzrdjFE+pSVz0
7c9C4evgYgmhUCUFvnZfgN+VBL1lS/p5UGnFPyNQ7KbSXDE71OAgWH/f/7kzsJPy
MxbJWM6eUN9Ny0APxM8olLV1FM4GzEoCSLfDVhStrdJ6P5wBmjLSugqSOLB8aMtd
8NwBY06nTpmo9xXGz9enLUWlpSeoReKU3TxvQvY+JcOWWpasDZOO4zD8B3bdLbA/
I8jdkH5Vs0pyPLaWD+1FxlQvlF45CuwpwoiAz00V2XkkMu8jKCGsQ0iuqXorSqvs
/7tQ1pHlUybAX+5W9raaJmphgc4gk33P3PlQCjhgYzxRu4yzRsEzS9hahoO/TAmq
Y70CooZaaeGNOBEDcKLZEzJdBr52cqW4MM8t1xHWTg3VCHJGeYI=
=O6NQ
-----END PGP SIGNATURE-----
Merge remote-tracking branch 'remotes/dgibson/tags/ppc-for-2.10-20170717' into staging
ppc patch queue 2017-07-17
This pull requests supersedes the one from 2017-07-14. That one had a
couple of subtle regressions: there was a build error for mingw32, and
an instance_size which was theoretically wrong everywhere, but only
actually bit on the Travis OSX build.
There are two major batches in this set, rather than the usual
collection of assorted fixes.
* More DRC cleanup. This gets the state management into a state
which should fix many of the hotplug+migration problems we've
had. Plus it gets the migration stream format into something
well defined and pretty minimal which we can reasonably support
into the future.
* Hashed Page Table resizing. It's been a while since this was
posted, but it's been through several previous rounds of review.
The kernel parts (both guest and host) are merged in 4.11, so
this is the only remaining piece left to allow resizing of the
HPT in a running guest.
There are also a handful of unrelated fixes.
# gpg: Signature made Mon 17 Jul 2017 07:36:52 BST
# gpg: using RSA key 0x6C38CACA20D9B392
# 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: aka "David Gibson (kernel.org) <dwg@kernel.org>"
# Primary key fingerprint: 75F4 6586 AE61 A66C C44E 87DC 6C38 CACA 20D9 B392
* remotes/dgibson/tags/ppc-for-2.10-20170717: (21 commits)
target/ppc: fix CPU hotplug when radix is enabled (TCG)
spapr: fix memory leak in spapr_core_pre_plug()
pseries: Allow HPT resizing with KVM
pseries: Use smaller default hash page tables when guest can resize
pseries: Enable HPT resizing for 2.10
pseries: Implement HPT resizing
pseries: Stubs for HPT resizing
ppc/pnv: Remove unused XICSState reference
spapr: fix potential memory leak in spapr_core_plug()
spapr: Implement DR-indicator for physical DRCs only
spapr: Remove sPAPRConfigureConnectorState sub-structure
spapr: Consolidate DRC state variables
spapr: Cleanups relating to DRC awaiting_release field
spapr: Refactor spapr_drc_detach()
spapr: Abort on delete failure in spapr_drc_release()
spapr: Simplify unplug path
spapr: Remove 'awaiting_allocation' DRC flag
spapr: Treat devices added before inbound migration as coldplugged
spapr: Minor cleanups to events handling
spapr: migrate pending_events of spapr state
...
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
If the cursor resource id isn't set the guest didn't define a cursor.
Skip the cursor update in post_load in that that case.
Reported-by: wanghaibin <wanghaibin.wang@huawei.com>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Tested-by: wanghaibin <wanghaibin.wang@huawei.com>
Message-id: 20170710070432.856-1-kraxel@redhat.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
In case of error, we must ensure the dynamically allocated base_core_type
is freed, like it is done everywhere else in this function.
This is a regression introduced in QEMU 2.9 by commit 8149e2992f.
Signed-off-by: Greg Kurz <groug@kaod.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
So far, qemu implements the PAPR Hash Page Table (HPT) resizing extension
with TCG. The same implementation will work with KVM PR, but we don't
currently allow that. For KVM HV we can only implement resizing with the
assistance of the host kernel, which needs a new capability and ioctl()s.
This patch adds support for testing the new KVM capability and implementing
the resize in terms of KVM facilities when necessary. If we're running on
a kernel which doesn't have the new capability flag at all, we fall back to
testing for PR vs. HV KVM using the same hack that we already use in a
number of places for older kernels.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
We've now implemented a PAPR extension allowing PAPR guest to resize
their hash page table (HPT) during runtime.
This patch makes use of that facility to allocate smaller HPTs by default.
Specifically when a guest is aware of the HPT resize facility, qemu sizes
the HPT to the initial memory size, rather than the maximum memory size on
the assumption that the guest will resize its HPT if necessary for hot
plugged memory.
When the initial memory size is much smaller than the maximum memory size
(a common configuration with e.g. oVirt / RHEV) then this can save
significant memory on the HPT.
If the guest does *not* advertise HPT resize awareness when it makes the
ibm,client-architecture-support call, qemu resizes the HPT for maxmimum
memory size (unless it's been configured not to allow such guests at all).
For now we make that reallocation assuming the guest has not yet used the
HPT at all. That's true in practice, but not, strictly, an architectural
or PAPR requirement. If we need to in future we can fix this by having
the client-architecture-support call reboot the guest with the revised
HPT size (the client-architecture-support call is explicitly permitted to
trigger a reboot in this way).
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Suraj Jitindar Singh <sjitindarsingh@gmail.com>
We've now implemented a PAPR extensions which allows PAPR guests (i.e.
"pseries" machine type) to resize their hash page table during runtime.
However, that extension is only enabled if explicitly chosen on the
command line. This patch enables it by default for spapr-2.10, but leaves
it disabled (by default) for older machine types.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Laurent Vivier <lvivier@redhat.com>
This patch implements hypercalls allowing a PAPR guest to resize its own
hash page table. This will eventually allow for more flexible memory
hotplug.
The implementation is partially asynchronous, handled in a special thread
running the hpt_prepare_thread() function. The state of a pending resize
is stored in SPAPR_MACHINE->pending_hpt.
The H_RESIZE_HPT_PREPARE hypercall will kick off creation of a new HPT, or,
if one is already in progress, monitor it for completion. If there is an
existing HPT resize in progress that doesn't match the size specified in
the call, it will cancel it, replacing it with a new one matching the
given size.
The H_RESIZE_HPT_COMMIT completes transition to a resized HPT, and can only
be called successfully once H_RESIZE_HPT_PREPARE has successfully
completed initialization of a new HPT. The guest must ensure that there
are no concurrent accesses to the existing HPT while this is called (this
effectively means stop_machine() for Linux guests).
For now H_RESIZE_HPT_COMMIT goes through the whole old HPT, rehashing each
HPTE into the new HPT. This can have quite high latency, but it seems to
be of the order of typical migration downtime latencies for HPTs of size
up to ~2GiB (which would be used in a 256GiB guest).
In future we probably want to move more of the rehashing to the "prepare"
phase, by having H_ENTER and other hcalls update both current and
pending HPTs. That's a project for another day, but should be possible
without any changes to the guest interface.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
This introduces stub implementations of the H_RESIZE_HPT_PREPARE and
H_RESIZE_HPT_COMMIT hypercalls which we hope to add in a PAPR
extension to allow run time resizing of a guest's hash page table. It
also adds a new machine property for controlling whether this new
facility is available.
For now we only allow resizing with TCG, allowing it with KVM will require
kernel changes as well.
Finally, it adds a new string to the hypertas property in the device
tree, advertising to the guest the availability of the HPT resizing
hypercalls. This is a tentative suggested value, and would need to be
standardized by PAPR before being merged.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Suraj Jitindar Singh <sjitindarsingh@gmail.com>
Reviewed-by: Laurent Vivier <lvivier@redhat.com>
Since commit 5c1da81215 ("spapr: Remove unnecessary differences between
hotplug and coldplug paths"), the CPU DT for the DRC is always allocated.
This causes a memory leak for pseries-2.6 and older machine types, that
don't support CPU hotplug and don't allocate DRCs for CPUs.
Reported-by: Bharata B Rao <bharata@linux.vnet.ibm.com>
Signed-off-by: Greg Kurz <groug@kaod.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
According to PAPR, the DR-indicator should only be valid for physical DRCs,
not logical DRCs. At the moment we implement it for all DRCs, so restrict
it to physical ones only.
We move the state to the physical DRC subclass, which means adding some
QOM boilerplate to handle the newly distinct type.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Daniel Barboza <danielhb@linux.vnet.ibm.com>
Tested-by: Daniel Barboza <danielhb@linux.vnet.ibm.com>
Most of the time, the state of a DRC object is contained in the single
'state' variable. However, during the transition from UNISOLATE to
CONFIGURED state requires multiple calls to the ibm,configure-connector
RTAS call to retrieve the device tree for the attached device. We need
some extra state to keep track of where we're up to in delivering the
device tree information to the guest.
Currently that extra state is in a sPAPRConfigureConnectorState
substructure which is only allocated when we're in the middle of the
configure connector process. That sounds like a good idea, but the extra
state is only two integers - on many platforms that will take up the same
room as the (maybe NULL) ccs pointer even before malloc() overhead. Plus
it's another object whose lifetime we need to manage. In short, it's not
worth it.
So, fold the sPAPRConfigureConnectorState substructure directly into the
DRC object.
Previously the structure was allocated lazily when the configure-connector
call discovers it's not there. Now, we need to initialize the subfields
pre-emptively, as soon as we enter UNISOLATE state.
Although it's not strictly necessary (the field values should only ever
be consulted when in UNISOLATE state), we try to keep them at -1 when in
other states, as a debugging aid.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Daniel Barboza <danielhb@linux.vnet.ibm.com>
Tested-by: Daniel Barboza <danielhb@linux.vnet.ibm.com>
Each DRC has three fields describing its state: isolation_state,
allocation_state and configured. At first this seems like a reasonable
representation, since its based directly on the PAPR defined
isolation-state and allocation-state indicators. However:
* Only a few combinations of the two fields' values are permitted
* allocation_state isn't used at all for physical DRCs
* The indicators are write only so they don't really have a well
defined current value independent of each other
This replaces these variables with a single state variable, whose names
and numbers are based on the diagram in LoPAPR section 13.4. Along with
this we add code to check the current state on various operations and make
sure the requested transition is permitted.
Strictly speaking, this makes guest visible changes to behaviour (since we
probably allowed some transitions we shouldn't have before). However, a
hypothetical guest broken by that wasn't PAPR compliant, and probably
wouldn't have worked under PowerVM.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Daniel Barboza <danielhb@linux.vnet.ibm.com>
Tested-by: Daniel Barboza <danielhb@linux.vnet.ibm.com>
'awaiting_release' indicates that the host has requested an unplug of the
device attached to the DRC, but the guest has not (yet) put the device
into a state where it is safe to complete removal.
1. Rename it to 'unplug_requested' which to me at least is clearer
2. Remove the ->release_pending() method used to check this from outside
spapr_drc.c. The method only plausibly has one implementation, so use
a plain function (spapr_drc_unplug_requested()) instead.
3. Remove it from the migration stream. Attempting to migrate mid-unplug
is broken not just for spapr - in general management has no good way to
determine if the device should be present on the destination or not. So,
until that's fixed, there's no point adding extra things to the stream.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Greg Kurz <groug@kaod.org>
Tested-by: Daniel Barboza <danielhb@linux.vnet.ibm.com>
This function has two unused parameters - remove them.
It also sets awaiting_release on all paths, except one. On that path
setting it is harmless, since it will be immediately cleared by
spapr_drc_release(). So factor it out of the if statements.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Greg Kurz <groug@kaod.org>
Tested-by: Daniel Barboza <danielhb@linux.vnet.ibm.com>
We currently ignore errors from the object_property_del() in
spapr_drc_release(). But the only way that could fail is if the property
doesn't exist, in which case it's a bug that we're in spapr_drc_release()
at all. So change from ignoring to abort()ing on errors.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
spapr_lmb_release() and spapr_core_release() call hotplug_handler_unplug()
which after a bunch of indirection calls spapr_memory_unplug() or
spapr_core_unplug(). But we already know which is the appropriate thing
to call here, so we can just fold it directly into the release function.
Once that's done, there's no need for an hc->unplug method in the spapr
machine at all: since we also have an hc->unplug_request method, the
hotplug core will never use ->unplug.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Greg Kurz <groug@kaod.org>
Tested-by: Daniel Barboza <danielhb@linux.vnet.ibm.com>
Igor Druzhinin
Stefano Stabellini
Peter Maydell
Xiong Zhang
Mark Cave-Ayland
Eduardo Habkost
Greg Kurz
Alexey Kardashevskiy
Jason Wang
Michal Privoznik
Gerd Hoffmann
Hervé Poussineau
David Gibson