This fixes a livelock in hvmloader with credit2 scheduler, whereby an
AP can be brought online, do its work, and shut itself down, before
the BSP re-emulates the VLAPIC write that sent the SIPI. BSP then ends
up in an endless re-emulation work where it sees the target vcpu is
down, therefore schedules a tasklet, which does no work because the
vcpu is already initialised. The fix is to check v->is_initialised
rather than VPF_down, before scheduling the tasklet.
Signed-off-by: Keir Fraser <keir.fraser@citrix.com>
Bitwise OR operator has higher precedence than conditional
operator. This patch fixes the incorrect associativity and re-enables
2MB paging for AMD family 10h system.
Signed-off-by: Wei Wang <wei.wang2@amd.com>
Acked-by: Wei Huang <wei.huang2@amd.com>
This leads to an erroneous WARN_ON and possibly other side effects. It
seems to me that even multi-function devices ought to enjoy the
privilege of MSI-X capabilities.
Signed-off-by: Gianni Tedesco <gianni.tedesco@citrix.com>
The ia64 specific xc_get_pfn_list doesn't seem any different to the
generic xc_get_pfn_list once the call to xc_ia64_get_pfn_list is
expanded so remove and just use the generic one.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com>
xc_ia64_pv_recv_vcpu_context does not need to lock the ctxt buffer
since it calls xc_ia64_recv_vcpu_context which calls
xc_vcpu_setcontext which takes care of any necessary bouncing.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com>
The function isn't really ia64 specific but since the result isn't
actually used in the only caller and the same info is available via
xc_domain_getinfo simply drop the function.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com>
Current function has heavily overloaded semantics for the various
arguments. Separate out into more specific functions.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com>
Current function has heavily overloaded semantics for the various
arguments. Separate out into more specific functions.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com>
Push responsibility for locking buffers refered to by the memory_op
argument up into the callers (which are now all internal to libxc).
This removes the last of the introspecation from do_memory_op and
generally makes the transistion to hypercall buffers smoother.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com>
Now that all XENMEM_* callers go via an op specific function make
xc_memory_op private to libxc (and rename to do_memory_op for
consistency with other private functions).
Also change the interface to take a size parameter so that
do_memory_op knows how much memory to lock for the top-level argument,
removing some of the introspection.
[ Compatibility note: this breaks qemu-xen before 706bc8c588cb ]
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com>
Currently the wrappers for these hypercalls swallow partial success
and return failure to the caller.
In order to use these functions more widely instead of open-coding
uses of XENMEM_* and xc_memory_op add variants which return the actual
hypercall result.
Therefore add the following functions:
xc_domain_increase_reservation
xc_domain_decrease_reservation
xc_domain_populate_physmap
and implement the existing semantics using these new functions as
xc_domain_increase_reservation_exact
xc_domain_decrease_reservation_exact
xc_domain_populate_physmap_exact
replacing the existing xc_domain_memory_* functions.
Use these new functions to replace all open coded uses of
XENMEM_increase_reservation, XENMEM_decrease_reservation and
XENMEM_populate_physmap.
Also rename xc_domain_memory_*_pod_target to xc_domain_*_pod_target
for consistency.
Temporarily add a compatibility macro for
xc_domain_memory_populate_physmap to allow time for qemu to catch up.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com>
We want the size of the struct not the pointer (although rounding up
to page size in lock_pages probably saves us).
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com>
The double indirection has been unnecessary since 9867:ec61a8c25429,
there is no possibility of the buffer being reallocated now.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com>
Not actually used here but useful to confirm that a handle is passed
down to each location where it will be required once we switch to
hypercall buffers.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com>
entries.
Some bits are reserved in x86 pagetable entries and must be zero; the
MMU should raise a pagefault if it sees them, with bit 3 set in the
error code. Xen's software pagetable walker hasn't been doing this,
which has been OK because no guest OSes actually have invalid bits set
except in error cases where things are already very wrong.
Xen's shadow pagetable code deliberately sets these bits as part of
the not-present-entry fast path, so if we're to support
shadow-on-shadow nested HVM, we need to start checking them.
Signed-off-by: Tim Deeegan <Tim.Deegan@citrix.com>
It isn't actually being used to write a vector at the only callsite
and can easily be implemented using xs_write.
Furthermore the old implementation used to leak both the key and value
strings.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
committer: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
22154:ed8019e672c6 causes migration_domname_r to be initialised as
d_config.c_info.name + "--incoming" immediately after setting
d_config.c_info.name. The intention appears to have been to set it to
common_domname + "--incoming" so do so.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
committer: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Pyxml/xmlproc is being used in tools/xen/xm/xenapi_create.py but is
unmaintained for several years now. xmlproc is used only for validating
XML documents against a DTD file.
This patch replaces the pyxml/xmlproc based XML validation with code
based on lxml, which is actively maintained.
Signed-off-by: Stephan Peijnik <spe@anexia.at>
Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
committer: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
This patch add back some missing VMCS defintions as preparation for
nested VMX.
Signed-off-by: Qing He <qing.he@intel.com>
Signed-off-by: Eddie Dong <eddie.dong@intel.com>
This workaround is taken from Linux, and the main motivation (besides
such workarounds indeed belonging in the hypervisor rather than each
kernel) is to suppress the warnings in the Xen log each Linux guest
would cause due to the disallowed wrmsr.
Signed-off-by: Jan Beulich <jbeulich@novell.com>
Intel requires a write of zeros (hence such writes now get silently
ignored) followed by a cpuid(1) followed by the actual read.
Includes some code redundancy elimination possible after the actual
change.
Signed-off-by: Jan Beulich <jbeulich@novell.com>
This fixes two errors with removing PCI devices from HVM domains. The
first error is that the handling of "pci-rem" device-model command is
erroneously implemented in qemu and difficult (impossible?) to get
right.
For example, during domain shutdown there can be a race where the guest
OS unloads it's drivers and perhaps even shuts down PCI subsystem before
the pci-rem command has been received by qemu. This means that no OS is
present to write to the port which causes the dm command to be
acknowledged.
We fix this by implementing a 'force removal' option to
libxl_device_pci_remove which is always set to 1 during guest shutdown.
It can be optionally enabled on the xl command line for other occasions.
The second error is that if a guest OS doesn't respond to the SCI
interrupt and therefore the pci-rem dm command, which can happen if the
guest OS has no ACPI PCI hotplug support, then device removal bails with
an error but only AFTER removing the device from xenstore. This means
that xenstore gets in to an inconsistent state where an assigned device
also appears to be assignable.
This is fixed by moving xenstore device removal to occur only after the
device has really been removed.
Signed-off-by: Gianni Tedesco <gianni.tedesco@citrix.com>
Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
committer: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
This allows us to map pages from guest physical addresses.
This will be used with nested virtualization.
Signed-off-by: Uwe Dannowski <Uwe.Dannowski@amd.com>
Signed-off-by: Christoph Egger <Christoph.Egger@amd.com>
Signed-off-by: Keir Fraser <keir@xen.org>
The first PV console device has an unusual frontend path for
historical reasons (/local/domain/<domid>/console rather than
/local/domain/<domid>/device/console/0).
Therefore we need to check this additional frontend path as well as
those under /local/domain/<domid>/device when tearing down all
backends or else we miss the backend for the console.
As part of this we need to ensure that the frontend directory has a
valid link to the backend, currently this link does not exist.
Fix this by adding libxl__device_frontend_path to return the correct
frontend path, accounting for the special case, and use it from
libxl__device_add_generic. Also add libxl__device_backend_path for
consistency.
This also allows console 0 setup to follow essentially the same code
path as other consoles within libxl_device_console_add, reducing the
special casing required there.
This also fixes the link from backend to frontend which until now gave
a dangling link to the normal device path instead of the exceptional
one.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
committer: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Walking the pinned-shadows list isn't enough: there could be an
unpinned (but still shadowed) l3 somewhere and if we later try to
unshadow it it'll have an up-pointer of PAGE_LIST_NULL:PAGE_LIST_NULL.
Signed-off-by: Tim Deegan <Tim.Deegan@citrix.com>
The patch is derived from a similar quirk in Linux kernel by David
Woodhouse and Adam Jackson. It checks for VT enabling bit in IGD GGC
register. If VT is not enabled correctly in the IGD, Xen does not
enable VT-d translation for IGD VT-d engine. In case where iommu boot
parameter is set to force, Xen calls panic().
Signed-off-by: Allen Kay <allen.m.kay@intel.com>
With the change to index irq_desc[] by IRQ rather than by vector, the
prior implicit change of the used flow handler when altering the IRQ
routing path to go through the 8259A didn't work anymore, and hence
on boards needing the ExtINT delivery workaround failed to boot.
Make make_8259A_irq() a real function again, thus allowing the flow
handler to be changed there.
Also eliminate the generally superfluous and (at least theoretically)
dangerous hard coded setting of the flow handler for IRQ0: Earlier
code should have set this already based on information coming from
ACPI/MPS, and non-standard systems may e.g. have this IRQ level
triggered.
Signed-off-by: Jan Beulich <jbeulich@novell.com>
Tested-by: Markus Schuster <ml@markus.schuster.name>
Changed vt-d feature boot messages from "supported" to "enabled" since
they reflect what is currently enabled in this Xen boot - not what is
supported by VT-d hardware.
Signed-off-by: Allen Kay <allen.m.kay@intel.com>
Ian Campbell
Keir Fraser
Ian Jackson
Gianni Tedesco
Samuel Thibault
Stephan Peijnik
Christoph Egger