linux-stable.git/drivers/xen, branch v4.4.271

xen-pciback: reconfigure also from backend watch handler

2021-05-26T09:27:32+00:00

commit c81d3d24602540f65256f98831d0a25599ea6b87 upstream.

When multiple PCI devices get assigned to a guest right at boot, libxl
incrementally populates the backend tree. The writes for the first of
the devices trigger the backend watch. In turn xen_pcibk_setup_backend()
will set the XenBus state to Initialised, at which point no further
reconfigures would happen unless a device got hotplugged. Arrange for
reconfigure to also get triggered from the backend watch handler.

Signed-off-by: Jan Beulich 
Cc: stable@vger.kernel.org
Reviewed-by: Boris Ostrovsky 
Link: https://lore.kernel.org/r/2337cbd6-94b9-4187-9862-c03ea12e0c61@suse.com
Signed-off-by: Juergen Gross 
Signed-off-by: Greg Kroah-Hartman

xen/events: fix setting irq affinity

2021-04-16T10:00:23+00:00

The backport of upstream patch 25da4618af240fbec61 ("xen/events: don't
unmask an event channel when an eoi is pending") introduced a
regression for stable kernels 5.10 and older: setting IRQ affinity for
IRQs related to interdomain events would no longer work, as moving the
IRQ to its new cpu was not included in the irq_ack callback for those
events.

Fix that by adding the needed call.

Note that kernels 5.11 and later don't need the explicit moving of the
IRQ to the target cpu in the irq_ack callback, due to a rework of the
affinity setting in kernel 5.11.

Signed-off-by: Juergen Gross 
Signed-off-by: Greg Kroah-Hartman

xen/evtchn: Change irq_info lock to raw_spinlock_t

2021-04-16T10:00:20+00:00

commit d120198bd5ff1d41808b6914e1eb89aff937415c upstream.

Unmask operation must be called with interrupt disabled,
on preempt_rt spin_lock_irqsave/spin_unlock_irqrestore
don't disable/enable interrupts, so use raw_* implementation
and change lock variable in struct irq_info from spinlock_t
to raw_spinlock_t

Cc: stable@vger.kernel.org
Fixes: 25da4618af24 ("xen/events: don't unmask an event channel when an eoi is pending")
Signed-off-by: Luca Fancellu 
Reviewed-by: Julien Grall 
Reviewed-by: Wei Liu 
Link: https://lore.kernel.org/r/20210406105105.10141-1-luca.fancellu@arm.com
Signed-off-by: Boris Ostrovsky 
Signed-off-by: Greg Kroah-Hartman

xen/events: avoid handling the same event on two cpus at the same time

2021-03-17T15:07:24+00:00

commit b6622798bc50b625a1e62f82c7190df40c1f5b21 upstream.

When changing the cpu affinity of an event it can happen today that
(with some unlucky timing) the same event will be handled on the old
and the new cpu at the same time.

Avoid that by adding an "event active" flag to the per-event data and
call the handler only if this flag isn't set.

Cc: stable@vger.kernel.org
Reported-by: Julien Grall 
Signed-off-by: Juergen Gross 
Reviewed-by: Julien Grall 
Link: https://lore.kernel.org/r/20210306161833.4552-4-jgross@suse.com
Signed-off-by: Boris Ostrovsky 
Signed-off-by: Greg Kroah-Hartman

xen/events: don't unmask an event channel when an eoi is pending

2021-03-17T15:07:24+00:00

commit 25da4618af240fbec6112401498301a6f2bc9702 upstream.

An event channel should be kept masked when an eoi is pending for it.
When being migrated to another cpu it might be unmasked, though.

In order to avoid this keep three different flags for each event channel
to be able to distinguish "normal" masking/unmasking from eoi related
masking/unmasking and temporary masking. The event channel should only
be able to generate an interrupt if all flags are cleared.

Cc: stable@vger.kernel.org
Fixes: 54c9de89895e ("xen/events: add a new "late EOI" evtchn framework")
Reported-by: Julien Grall 
Signed-off-by: Juergen Gross 
Reviewed-by: Julien Grall 
Reviewed-by: Boris Ostrovsky 
Tested-by: Ross Lagerwall 
Link: https://lore.kernel.org/r/20210306161833.4552-3-jgross@suse.com

[boris -- corrected Fixed tag format]

Signed-off-by: Boris Ostrovsky 
Signed-off-by: Greg Kroah-Hartman

xen/events: reset affinity of 2-level event when tearing it down

2021-03-17T15:07:24+00:00

commit 9e77d96b8e2724ed00380189f7b0ded61113b39f upstream.

When creating a new event channel with 2-level events the affinity
needs to be reset initially in order to avoid using an old affinity
from earlier usage of the event channel port. So when tearing an event
channel down reset all affinity bits.

The same applies to the affinity when onlining a vcpu: all old
affinity settings for this vcpu must be reset. As percpu events get
initialized before the percpu event channel hook is called,
resetting of the affinities happens after offlining a vcpu (this is
working, as initial percpu memory is zeroed out).

Cc: stable@vger.kernel.org
Reported-by: Julien Grall 
Signed-off-by: Juergen Gross 
Reviewed-by: Julien Grall 
Link: https://lore.kernel.org/r/20210306161833.4552-2-jgross@suse.com
Signed-off-by: Boris Ostrovsky 
Signed-off-by: Greg Kroah-Hartman

xen-scsiback: don't "handle" error by BUG()

2021-02-23T12:58:14+00:00

commit 7c77474b2d22176d2bfb592ec74e0f2cb71352c9 upstream.

In particular -ENOMEM may come back here, from set_foreign_p2m_mapping().
Don't make problems worse, the more that handling elsewhere (together
with map's status fields now indicating whether a mapping wasn't even
attempted, and hence has to be considered failed) doesn't require this
odd way of dealing with errors.

This is part of XSA-362.

Signed-off-by: Jan Beulich 
Cc: stable@vger.kernel.org
Reviewed-by: Juergen Gross 
Signed-off-by: Juergen Gross 
Signed-off-by: Greg Kroah-Hartman

Xen/gntdev: correct error checking in gntdev_map_grant_pages()

2021-02-23T12:58:14+00:00

commit ebee0eab08594b2bd5db716288a4f1ae5936e9bc upstream.

Failure of the kernel part of the mapping operation should also be
indicated as an error to the caller, or else it may assume the
respective kernel VA is okay to access.

Furthermore gnttab_map_refs() failing still requires recording
successfully mapped handles, so they can be unmapped subsequently. This
in turn requires there to be a way to tell full hypercall failure from
partial success - preset map_op status fields such that they won't
"happen" to look as if the operation succeeded.

Also again use GNTST_okay instead of implying its value (zero).

This is part of XSA-361.

Signed-off-by: Jan Beulich 
Cc: stable@vger.kernel.org
Reviewed-by: Juergen Gross 
Signed-off-by: Juergen Gross 
Signed-off-by: Greg Kroah-Hartman

Xen/gntdev: correct dev_bus_addr handling in gntdev_map_grant_pages()

2021-02-23T12:58:14+00:00

commit dbe5283605b3bc12ca45def09cc721a0a5c853a2 upstream.

We may not skip setting the field in the unmap structure when
GNTMAP_device_map is in use - such an unmap would fail to release the
respective resources (a page ref in the hypervisor). Otoh the field
doesn't need setting at all when GNTMAP_device_map is not in use.

To record the value for unmapping, we also better don't use our local
p2m: In particular after a subsequent change it may not have got updated
for all the batch elements. Instead it can simply be taken from the
respective map's results.

We can additionally avoid playing this game altogether for the kernel
part of the mappings in (x86) PV mode.

This is part of XSA-361.

Signed-off-by: Jan Beulich 
Cc: stable@vger.kernel.org
Reviewed-by: Stefano Stabellini 
Signed-off-by: Juergen Gross 
Signed-off-by: Greg Kroah-Hartman

xenbus/xenbus_backend: Disallow pending watch messages

2020-12-29T12:42:45+00:00

commit 9996bd494794a2fe393e97e7a982388c6249aa76 upstream.

'xenbus_backend' watches 'state' of devices, which is writable by
guests.  Hence, if guests intensively updates it, dom0 will have lots of
pending events that exhausting memory of dom0.  In other words, guests
can trigger dom0 memory pressure.  This is known as XSA-349.  However,
the watch callback of it, 'frontend_changed()', reads only 'state', so
doesn't need to have the pending events.

To avoid the problem, this commit disallows pending watch messages for
'xenbus_backend' using the 'will_handle()' watch callback.

This is part of XSA-349

Cc: stable@vger.kernel.org
Signed-off-by: SeongJae Park 
Reported-by: Michael Kurth 
Reported-by: Pawel Wieczorkiewicz 
Reviewed-by: Juergen Gross 
Signed-off-by: Juergen Gross 
Signed-off-by: Greg Kroah-Hartman