Since we *might* have user emulation with softmmu,
use the clearer 'CONFIG_SYSTEM_ONLY' key to check
for system emulation.
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20230613133347.82210-9-philmd@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Coverity points out that if (!s && !s->impl) isn't really what we intended
to do here. CID 1508131.
Fixes: 0324751272 ("hw/xen: Add emulated implementation of XenStore operations")
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Paul Durrant <paul@xen.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-Id: <20230412185102.441523-6-dwmw2@infradead.org>
Signed-off-by: Anthony PERARD <anthony.perard@citrix.com>
I initially put the basic platform init (overlay pages, grant tables,
event channels) into mc->kvm_type because that was the earliest place
that could sensibly test for xen_mode==XEN_EMULATE.
The intent was to do this early enough that we could then initialise the
XenBus and other parts which would have depended on them, from a generic
location for both Xen and KVM/Xen in the PC-specific code, as seen in
https://lore.kernel.org/qemu-devel/20230116221919.1124201-16-dwmw2@infradead.org/
However, then the Xen on Arm patches came along, and *they* wanted to
do the XenBus init from a 'generic' Xen-specific location instead:
https://lore.kernel.org/qemu-devel/20230210222729.957168-4-sstabellini@kernel.org/
Since there's no generic location that covers all three, I conceded to
do it for XEN_EMULATE mode in pc_basic_devices_init().
And now there's absolutely no point in having some of the platform init
done from pc_machine_kvm_type(); we can move it all up to live in a
single place in pc_basic_devices_init(). This has the added benefit that
we can drop the separate xen_evtchn_connect_gsis() function completely,
and pass just the system GSIs in directly to xen_evtchn_create().
While I'm at it, it does no harm to explicitly pass in the *number* of
said GSIs, because it does make me twitch a bit to pass an array of
impicit size. During the lifetime of the KVM/Xen patchset, that had
already changed (albeit just cosmetically) from GSI_NUM_PINS to
IOAPIC_NUM_PINS.
And document a bit better that this is for the *output* GSI for raising
CPU0's events when the per-CPU vector isn't available. The fact that
we create a whole set of them and then only waggle the one we're told
to, instead of having a single output and only *connecting* it to the
GSI that it should be connected to, is still non-intuitive for me.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Paul Durrant <paul@xen.org>
Message-Id: <20230412185102.441523-2-dwmw2@infradead.org>
Signed-off-by: Anthony PERARD <anthony.perard@citrix.com>
Both TYPE_KVM_I8254 and TYPE_I8254 have their own but same implementation of
the "iobase" property. The storage for the property already resides in
PITCommonState, so also move the property definition there.
Signed-off-by: Bernhard Beschow <shentey@gmail.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-Id: <20230523195608.125820-2-shentey@gmail.com>
Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
All callers now pass is_external=false to aio_set_fd_handler() and
aio_set_event_notifier(). The aio_disable_external() API that
temporarily disables fd handlers that were registered is_external=true
is therefore dead code.
Remove aio_disable_external(), aio_enable_external(), and the
is_external arguments to aio_set_fd_handler() and
aio_set_event_notifier().
The entire test-fdmon-epoll test is removed because its sole purpose was
testing aio_disable_external().
Parts of this patch were generated using the following coccinelle
(https://coccinelle.lip6.fr/) semantic patch:
@@
expression ctx, fd, is_external, io_read, io_write, io_poll, io_poll_ready, opaque;
@@
- aio_set_fd_handler(ctx, fd, is_external, io_read, io_write, io_poll, io_poll_ready, opaque)
+ aio_set_fd_handler(ctx, fd, io_read, io_write, io_poll, io_poll_ready, opaque)
@@
expression ctx, notifier, is_external, io_read, io_poll, io_poll_ready;
@@
- aio_set_event_notifier(ctx, notifier, is_external, io_read, io_poll, io_poll_ready)
+ aio_set_event_notifier(ctx, notifier, io_read, io_poll, io_poll_ready)
Reviewed-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20230516190238.8401-21-stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
There is no need to suspend activity between aio_disable_external() and
aio_enable_external(), which is mainly used for the block layer's drain
operation.
This is part of ongoing work to remove the aio_disable_external() API.
Reviewed-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Paul Durrant <paul@xen.org>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20230516190238.8401-9-stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This had been pulled in via qemu/plugin.h from hw/core/cpu.h,
but that will be removed.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-Id: <20230310195252.210956-5-richard.henderson@linaro.org>
[AJB: add various additional cases shown by CI]
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20230315174331.2959-15-alex.bennee@linaro.org>
Reviewed-by: Emilio Cota <cota@braap.org>
This is only part of it; we will also need to get the PV back end drivers
to tear down their own mappings (or do it for them, but they kind of need
to stop using the pointers too).
Some more work on the actual PV back ends and xen-bus code is going to be
needed to really make soft reset and migration fully functional, and this
part is the basis for that.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Paul Durrant <paul@xen.org>
We don't actually access the guest's page through the grant, because
this isn't real Xen, and we can just use the page we gave it in the
first place. Map the grant anyway, mostly for cosmetic purposes so it
*looks* like it's in use in the guest-visible grant table.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Paul Durrant <paul@xen.org>
Now that we have an internal implementation of XenStore, we can populate
the xenstore_backend_ops to allow PV backends to talk to it.
Watches can't be processed with immediate callbacks because that would
call back into XenBus code recursively. Defer them to a QEMUBH to be run
as appropriate from the main loop. We use a QEMUBH per XS handle, and it
walks all the watches (there shouldn't be many per handle) to fire any
which have pending events. We *could* have done it differently but this
allows us to use the same struct watch_event as we have for the guest
side, and keeps things relatively simple.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Paul Durrant <paul@xen.org>
This is limited to mapping a single grant at a time, because under Xen the
pages are mapped *contiguously* into qemu's address space, and that's very
hard to do when those pages actually come from anonymous mappings in qemu
in the first place.
Eventually perhaps we can look at using shared mappings of actual objects
for system RAM, and then we can make new mappings of the same backing
store (be it deleted files, shmem, whatever). But for now let's stick to
a page at a time.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Paul Durrant <paul@xen.org>
We provided the backend-facing evtchn functions very early on as part of
the core Xen platform support, since things like timers and xenstore need
to use them.
By what may or may not be an astonishing coincidence, those functions
just *happen* all to have exactly the right function prototypes to slot
into the evtchn_backend_ops table and be called by the PV backends.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Paul Durrant <paul@xen.org>
This implements the basic migration support in the back end, with unit
tests that give additional confidence in the node-counting already in
the tree.
However, the existing PV back ends like xen-disk don't support migration
yet. They will reset the ring and fail to continue where they left off.
We will fix that in future, but not in time for the 8.0 release.
Since there's also an open question of whether we want to serialize the
full XenStore or only the guest-owned nodes in /local/domain/${domid},
for now just mark the XenStore device as unmigratable.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Paul Durrant <paul@xen.org>
Store perms as a GList of strings, check permissions.
Signed-off-by: Paul Durrant <pdurrant@amazon.com>
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Paul Durrant <paul@xen.org>
Firing watches on the nodes that still exist is relatively easy; just
walk the tree and look at the nodes with refcount of one.
Firing watches on *deleted* nodes is more fun. We add 'modified_in_tx'
and 'deleted_in_tx' flags to each node. Nodes with those flags cannot
be shared, as they will always be unique to the transaction in which
they were created.
When xs_node_walk would need to *create* a node as scaffolding and it
encounters a deleted_in_tx node, it can resurrect it simply by clearing
its deleted_in_tx flag. If that node originally had any *data*, they're
gone, and the modified_in_tx flag will have been set when it was first
deleted.
We then attempt to send appropriate watches when the transaction is
committed, properly delete the deleted_in_tx nodes, and remove the
modified_in_tx flag from the others.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Paul Durrant <paul@xen.org>
Given that the whole thing supported copy on write from the beginning,
transactions end up being fairly simple. On starting a transaction, just
take a ref of the existing root; swap it back in on a successful commit.
The main tree has a transaction ID too, and we keep a record of the last
transaction ID given out. if the main tree is ever modified when it isn't
the latest, it gets a new transaction ID.
A commit can only succeed if the main tree hasn't moved on since it was
forked. Strictly speaking, the XenStore protocol allows a transaction to
succeed as long as nothing *it* read or wrote has changed in the interim,
but no implementations do that; *any* change is sufficient to abort a
transaction.
This does not yet fire watches on the changed nodes on a commit. That bit
is more fun and will come in a follow-on commit.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Paul Durrant <paul@xen.org>
Starts out fairly simple: a hash table of watches based on the path.
Except there can be multiple watches on the same path, so the watch ends
up being a simple linked list, and the head of that list is in the hash
table. Which makes removal a bit of a PITA but it's not so bad; we just
special-case "I had to remove the head of the list and now I have to
replace it in / remove it from the hash table". And if we don't remove
the head, it's a simple linked-list operation.
We do need to fire watches on *deleted* nodes, so instead of just a simple
xs_node_unref() on the topmost victim, we need to recurse down and fire
watches on them all.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Paul Durrant <paul@xen.org>
This is a fairly simple implementation of a copy-on-write tree.
The node walk function starts off at the root, with 'inplace == true'.
If it ever encounters a node with a refcount greater than one (including
the root node), then that node is shared with other trees, and cannot
be modified in place, so the inplace flag is cleared and we copy on
write from there on down.
Xenstore write has 'mkdir -p' semantics and will create the intermediate
nodes if they don't already exist, so in that case we flip the inplace
flag back to true as we populate the newly-created nodes.
We put a copy of the absolute path into the buffer in the struct walk_op,
with *two* NUL terminators at the end. As xs_node_walk() goes down the
tree, it replaces the next '/' separator with a NUL so that it can use
the 'child name' in place. The next recursion down then puts the '/'
back and repeats the exercise for the next path element... if it doesn't
hit that *second* NUL termination which indicates the true end of the
path.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Paul Durrant <paul@xen.org>
This implements the basic wire protocol for the XenStore commands, punting
all the actual implementation to xs_impl_* functions which all just return
errors for now.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Paul Durrant <paul@xen.org>
The default number of PIRQs is set to 256 to avoid issues with 32-bit MSI
devices. Allow it to be increased if the user desires.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Paul Durrant <paul@xen.org>
The way that Xen handles MSI PIRQs is kind of awful.
There is a special MSI message which targets a PIRQ. The vector in the
low bits of data must be zero. The low 8 bits of the PIRQ# are in the
destination ID field, the extended destination ID field is unused, and
instead the high bits of the PIRQ# are in the high 32 bits of the address.
Using the high bits of the address means that we can't intercept and
translate these messages in kvm_send_msi(), because they won't be caught
by the APIC — addresses like 0x1000fee46000 aren't in the APIC's range.
So we catch them in pci_msi_trigger() instead, and deliver the event
channel directly.
That isn't even the worst part. The worst part is that Xen snoops on
writes to devices' MSI vectors while they are *masked*. When a MSI
message is written which looks like it targets a PIRQ, it remembers
the device and vector for later.
When the guest makes a hypercall to bind that PIRQ# (snooped from a
marked MSI vector) to an event channel port, Xen *unmasks* that MSI
vector on the device. Xen guests using PIRQ delivery of MSI don't
ever actually unmask the MSI for themselves.
Now that this is working we can finally enable XENFEAT_hvm_pirqs and
let the guest use it all.
Tested with passthrough igb and emulated e1000e + AHCI.
CPU0 CPU1
0: 65 0 IO-APIC 2-edge timer
1: 0 14 xen-pirq 1-ioapic-edge i8042
4: 0 846 xen-pirq 4-ioapic-edge ttyS0
8: 1 0 xen-pirq 8-ioapic-edge rtc0
9: 0 0 xen-pirq 9-ioapic-level acpi
12: 257 0 xen-pirq 12-ioapic-edge i8042
24: 9600 0 xen-percpu -virq timer0
25: 2758 0 xen-percpu -ipi resched0
26: 0 0 xen-percpu -ipi callfunc0
27: 0 0 xen-percpu -virq debug0
28: 1526 0 xen-percpu -ipi callfuncsingle0
29: 0 0 xen-percpu -ipi spinlock0
30: 0 8608 xen-percpu -virq timer1
31: 0 874 xen-percpu -ipi resched1
32: 0 0 xen-percpu -ipi callfunc1
33: 0 0 xen-percpu -virq debug1
34: 0 1617 xen-percpu -ipi callfuncsingle1
35: 0 0 xen-percpu -ipi spinlock1
36: 8 0 xen-dyn -event xenbus
37: 0 6046 xen-pirq -msi ahci[0000:00:03.0]
38: 1 0 xen-pirq -msi-x ens4
39: 0 73 xen-pirq -msi-x ens4-rx-0
40: 14 0 xen-pirq -msi-x ens4-rx-1
41: 0 32 xen-pirq -msi-x ens4-tx-0
42: 47 0 xen-pirq -msi-x ens4-tx-1
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Paul Durrant <paul@xen.org>
This wires up the basic infrastructure but the actual interrupts aren't
there yet, so don't advertise it to the guest.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Paul Durrant <paul@xen.org>
Just hook up the basic hypercalls to stubs in xen_evtchn.c for now.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Paul Durrant <paul@xen.org>
Extract requests, return ENOSYS to all of them. This is enough to allow
older Linux guests to boot, as they need *something* back but it doesn't
matter much what.
A full implementation of a single-tentant internal XenStore copy-on-write
tree with transactions and watches is waiting in the wings to be sent in
a subsequent round of patches along with hooking up the actual PV disk
back end in qemu, but this is enough to get guests booting for now.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Paul Durrant <paul@xen.org>
Just the basic shell, with the event channel hookup. It only dumps the
buffer for now; a real ring implmentation will come in a subsequent patch.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Paul Durrant <paul@xen.org>
The provides the QEMU side of interdomain event channels, allowing events
to be sent to/from the guest.
The API mirrors libxenevtchn, and in time both this and the real Xen one
will be available through ops structures so that the PV backend drivers
can use the correct one as appropriate.
For now, this implementation can be used directly by our XenStore which
will be for emulated mode only.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Paul Durrant <paul@xen.org>
Introduce support for one shot and periodic mode of Xen PV timers,
whereby timer interrupts come through a special virq event channel
with deadlines being set through:
1) set_timer_op hypercall (only oneshot)
2) vcpu_op hypercall for {set,stop}_{singleshot,periodic}_timer
hypercalls
Signed-off-by: Joao Martins <joao.m.martins@oracle.com>
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Paul Durrant <paul@xen.org>
The guest is permitted to specify an arbitrary domain/bus/device/function
and INTX pin from which the callback IRQ shall appear to have come.
In QEMU we can only easily do this for devices that actually exist, and
even that requires us "knowing" that it's a PCMachine in order to find
the PCI root bus — although that's OK really because it's always true.
We also don't get to get notified of INTX routing changes, because we
can't do that as a passive observer; if we try to register a notifier
it will overwrite any existing notifier callback on the device.
But in practice, guests using PCI_INTX will only ever use pin A on the
Xen platform device, and won't swizzle the INTX routing after they set
it up. So this is just fine.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Paul Durrant <paul@xen.org>
The GSI callback (and later PCI_INTX) is a level triggered interrupt. It
is asserted when an event channel is delivered to vCPU0, and is supposed
to be cleared when the vcpu_info->evtchn_upcall_pending field for vCPU0
is cleared again.
Thankfully, Xen does *not* assert the GSI if the guest sets its own
evtchn_upcall_pending field; we only need to assert the GSI when we
have delivered an event for ourselves. So that's the easy part, kind of.
There's a slight complexity in that we need to hold the BQL before we
can call qemu_set_irq(), and we definitely can't do that while holding
our own port_lock (because we'll need to take that from the qemu-side
functions that the PV backend drivers will call). So if we end up
wanting to set the IRQ in a context where we *don't* already hold the
BQL, defer to a BH.
However, we *do* need to poll for the evtchn_upcall_pending flag being
cleared. In an ideal world we would poll that when the EOI happens on
the PIC/IOAPIC. That's how it works in the kernel with the VFIO eventfd
pairs — one is used to trigger the interrupt, and the other works in the
other direction to 'resample' on EOI, and trigger the first eventfd
again if the line is still active.
However, QEMU doesn't seem to do that. Even VFIO level interrupts seem
to be supported by temporarily unmapping the device's BARs from the
guest when an interrupt happens, then trapping *all* MMIO to the device
and sending the 'resample' event on *every* MMIO access until the IRQ
is cleared! Maybe in future we'll plumb the 'resample' concept through
QEMU's irq framework but for now we'll do what Xen itself does: just
check the flag on every vmexit if the upcall GSI is known to be
asserted.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Paul Durrant <paul@xen.org>
Specifically add listing, injection of event channels.
Signed-off-by: Joao Martins <joao.m.martins@oracle.com>
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Acked-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Paul Durrant <paul@xen.org>
Add the array of virq ports to each vCPU so that we can deliver timers,
debug ports, etc. Global virqs are allocated against vCPU 0 initially,
but can be migrated to other vCPUs (when we implement that).
The kernel needs to know about VIRQ_TIMER in order to accelerate timers,
so tell it via KVM_XEN_VCPU_ATTR_TYPE_TIMER. Also save/restore the value
of the singleshot timer across migration, as the kernel will handle the
hypercalls automatically now.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Paul Durrant <paul@xen.org>
This finally comes with a mechanism for actually injecting events into
the guest vCPU, with all the atomic-test-and-set that's involved in
setting the bit in the shinfo, then the index in the vcpu_info, and
injecting either the lapic vector as MSI, or letting KVM inject the
bare vector.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Paul Durrant <paul@xen.org>
It calls an internal close_port() helper which will also be used from
EVTCHNOP_reset and will actually do the work to disconnect/unbind a port
once any of that is actually implemented in the first place.
That in turn calls a free_port() internal function which will be in
error paths after allocation.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Paul Durrant <paul@xen.org>
This adds the basic structure for maintaining the port table and reporting
the status of ports therein.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Paul Durrant <paul@xen.org>
Include basic support for setting HVM_PARAM_CALLBACK_IRQ to the global
vector method HVM_PARAM_CALLBACK_TYPE_VECTOR, which is handled in-kernel
by raising the vector whenever the vCPU's vcpu_info->evtchn_upcall_pending
flag is set.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Paul Durrant <paul@xen.org>
Xen will "latch" the guest's 32-bit or 64-bit ("long mode") setting when
the guest writes the MSR to fill in the hypercall page, or when the guest
sets the event channel callback in HVM_PARAM_CALLBACK_IRQ.
KVM handles the former and sets the kernel's long_mode flag accordingly.
The latter will be handled in userspace. Keep them in sync by noticing
when a hypercall is made in a mode that doesn't match qemu's idea of
the guest mode, and resyncing from the kernel. Do that same sync right
before serialization too, in case the guest has set the hypercall page
but hasn't yet made a system call.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Paul Durrant <paul@xen.org>
For the shared info page and for grant tables, Xen shares its own pages
from the "Xen heap" to the guest. The guest requests that a given page
from a certain address space (XENMAPSPACE_shared_info, etc.) be mapped
to a given GPA using the XENMEM_add_to_physmap hypercall.
To support that in qemu when *emulating* Xen, create a memory region
(migratable) and allow it to be mapped as an overlay when requested.
Xen theoretically allows the same page to be mapped multiple times
into the guest, but that's hard to track and reinstate over migration,
so we automatically *unmap* any previous mapping when creating a new
one. This approach has been used in production with.... a non-trivial
number of guests expecting true Xen, without any problems yet being
noticed.
This adds just the shared info page for now. The grant tables will be
a larger region, and will need to be overlaid one page at a time. I
think that means I need to create separate aliases for each page of
the overall grant_frames region, so that they can be mapped individually.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Paul Durrant <paul@xen.org>
