No state is left to protect. It being the last field, drop the struct
itself as well. Similarly for then ending up empty, drop the .complete
handler.
This is part of XSA-463 / CVE-2024-45818
Suggested-by: Andrew Cooper <andrew.cooper3@citrix.com>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
(cherry picked from commit b180a50326)
No uses are left, hence its setup, teardown, and the field itself can
also go away. stdvga_deinit() is then empty and can be dropped as well.
This is part of XSA-463 / CVE-2024-45818
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
(cherry picked from commit 3beb4baf2a)
No consumers are left, hence the producer and the fields themselves can
also go away. stdvga_outb() is then useless, rendering stdvga_out()
useless as well. Hence the entire I/O port intercept can go away.
This is part of XSA-463 / CVE-2024-45818
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
(cherry picked from commit 86c03372e1)
No consumers are left, hence the producer and the array itself can also
go away. The static sr_mask[] is then orphaned and hence needs dropping,
too.
This is part of XSA-463 / CVE-2024-45818
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
(cherry picked from commit 7aba44bdd7)
No consumers are left, hence the producer and the array itself can also
go away. The static gr_mask[] is then orphaned and hence needs dropping,
too.
This is part of XSA-463 / CVE-2024-45818
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
(cherry picked from commit b16c0966a1)
All read accesses are rejected by the ->accept handler, while writes
bypass the bulk of the function body. Drop the dead code, leaving an
assertion in the read handler.
A number of other static items (and a macro) are then unreferenced and
hence also need (want) dropping. The same applies to the "latch" field
of the state structure.
This is part of XSA-463 / CVE-2024-45818
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
(cherry picked from commit 89108547af)
Two of its consumers are dead (in compile-time constant conditionals)
and the only remaining ones are merely controlling debug logging. Hence
the field is now pointless to set, which in particular allows to get rid
of the questionable conditional from which the field's value was
established (afaict 551ceee975 ["x86, hvm: stdvga cache always on"]
had dropped too much of the earlier extra check that was there, and
quite likely further checks were missing).
This is part of XSA-463 / CVE-2024-45818
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
(cherry picked from commit b740a9369e)
While ->count will only be different from 1 for "indirect" (data in
guest memory) accesses, it being 1 does not exclude the request being an
"indirect" one. Check both to be on the safe side, and bring the ->count
part also in line with what ioreq_send_buffered() actually refuses to
handle.
This is part of XSA-463 / CVE-2024-45818
Fixes: 3bbaaec09b ("x86/hvm: unify stdvga mmio intercept with standard mmio intercept")
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
(cherry picked from commit eb7cd0593d)
Since 68e1183411 ("libxc: introduce a xc_dom_arch for hvm-3.0-x86_32
guests"), HVM guests are built using XEN_DOMCTL_sethvmcontext, which
ends up disabling stdvga caching because of arch_hvm_load() being
involved in the processing of the request. With that the field is
useless, and can be dropped. Drop the helper functions manipulating /
checking as well right away, but leave the use sites of
stdvga_cache_is_enabled() with the hard-coded result the function would
have produced, to aid validation of subsequent dropping of further code.
This is part of XSA-463 / CVE-2024-45818
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
(cherry picked from commit 53b7246bdf)
With the error vector set to an illegal value, the function invoking
vlapic_set_irq() would bring execution back here, with the non-recursive
lock already held. Avoid the call in this case, merely further updating
ESR (if necessary).
This is XSA-462 / CVE-2024-45817.
Fixes: 5f32d186a8 ("x86/vlapic: don't silently accept bad vectors")
Reported-by: Federico Serafini <federico.serafini@bugseng.com>
Reported-by: Andrew Cooper <andrew.cooper3@citrix.com>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
master commit: c42d9ec61f
master date: 2024-09-24 14:23:29 +0200
When multiple devices share resources and one of them is to be passed
through to a guest, security of the entire system and of respective
guests individually cannot really be guaranteed without knowing
internals of any of the involved guests. Therefore such a configuration
cannot really be security-supported, yet making that explicit was so far
missing.
This is XSA-461 / CVE-2024-31146.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Juergen Gross <jgross@suse.com>
master commit: 9c94eda1e3
master date: 2024-08-13 16:37:25 +0200
Forever since its introduction the main loop in the function dealing
with multi-vector MSI had error exit points ("break") with different
properties: In one case no IRQ descriptor lock is being held.
Nevertheless the subsequent error cleanup path assumed such a lock would
uniformly need releasing. Identify the case by setting "desc" to NULL,
thus allowing the unlock to be skipped as necessary.
This is CVE-2024-31143 / XSA-458.
Coverity ID: 1605298
Fixes: d1b6d0a024 ("x86: enable multi-vector MSI")
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
master commit: 57338346f2
master date: 2024-07-16 14:09:14 +0200
This is a prereq to enabling e.g. the MSRLIST feature.
Note that the PROCBASED_CTLS3 MSR is different from other VMX feature
reporting MSRs, in that all 64 bits report allowed 1-settings.
vVMX code is left alone, though, for the time being.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
(cherry picked from commit 878159bf25)
Hypercalls in particular can be the subject of continuations, and logic
there checks updated state against incoming register values. If the
guest manufactured a suitable argument register with a non-zero upper
half before entering compatibility mode and issuing a hypercall from
there, checks in hypercall_xlat_continuation() might trip.
Since for HVM we want to also be sure to not hit a corner case in the
emulator, initiate the clipping right from the top of
{svm,vmx}_vmexit_handler(). Also rename the invoked function, as it no
longer does only invalidation of fields.
Note that architecturally the upper halves of registers are undefined
after a switch between compatibility and 64-bit mode (either direction).
Hence once having entered compatibility mode, the guest can't assume
the upper half of any register to retain its value.
This is part of XSA-454 / CVE-2023-46842.
Fixes: b8a7efe852 ("Enable compatibility mode operation for HYPERVISOR_memory_op")
Reported-by: Manuel Andreas <manuel.andreas@tum.de>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
(cherry picked from commit 6a98383b08)
Dealing with exceptions raised from within emulation stubs involves
discarding return address (replaced by exception related information).
Such discarding of course also requires removing the corresponding entry
from the shadow stack.
Also amend the comment in fixup_exception_return(), to further clarify
why use of ptr[1] can't be an out-of-bounds access.
This is CVE-2023-46841 / XSA-451.
Fixes: 209fb9919b ("x86/extable: Adjust extable handling to be shadow stack compatible")
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
master commit: 91f5f7a915
master date: 2024-02-27 13:49:22 +0100
sh_set_toplevel_shadow() re-pinning the top-level shadow we may be
running on is not enough (and at the same time unnecessary when the
shadow isn't what we're running on): That shadow becomes eligible for
blowing away (from e.g. shadow_prealloc()) immediately after the
paging lock was dropped. Yet it needs to remain valid until the actual
page table switch occurred.
Propagate up the call chain the shadow entry that needs releasing
eventually, and carry out the release immediately after switching page
tables. Handle update_cr3() failures by switching to idle pagetables.
Note that various further uses of update_cr3() are HVM-only or only act
on paused vCPU-s, in which case sh_set_toplevel_shadow() will not defer
releasing of the reference.
While changing the update_cr3() hook, also convert the "do_locking"
parameter to boolean.
This is CVE-2023-34322 / XSA-438.
Reported-by: Tim Deegan <tim@xen.org>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: George Dunlap <george.dunlap@cloud.com>
(cherry picked from commit fb0ff49fe9)
Reportedly the AMD Custom APU 0405 found on SteamDeck, models 0x90 and
0x91, (quoting the respective Linux commit) is similarly affected. Put
another instance of our Zen1 vs Zen2 distinction checks in
amd_check_zenbleed(), forcing use of the chickenbit irrespective of
ucode version (building upon real hardware never surfacing a version of
0xffffffff).
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
(cherry picked from commit 145a69c094)
We can't use initializers of unnamed struct/union members just yet.
Fixes: d638fe233c ("libxl: use the cpuid feature names from cpufeatureset.h")
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Juergen Gross <jgross@suse.com>
Acked-by: Andrew Cooper <andrew.cooper3@citrix.com>
(cherry picked from commit 1aa5acbbec)
Because of -Wshadow the build otherwise fails with old enough glibc.
While there also obey line length limits for msr_add().
Fixes: 6d21cedbaa ("libxl: add support for parsing MSR features")
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Juergen Gross <jgross@suse.com>
Acked-by: Andrew Cooper <andrew.cooper3@citrix.com>
(cherry picked from commit 4f6afde88b)
Loading the VGA base address involves sym_esi(), i.e. %esi still needs
to hold the relocation base address. Therefore the address of the
message to output cannot be "passed" in %esi. Put the message offset in
%ecx instead, adding it into %esi _after_ its last use as base address.
Fixes: b28044226e ("x86: make Xen early boot code relocatable")
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Andrew Cooper <andrew.cooper3@citrix.com>
master commit: b1c16800e5
master date: 2023-07-19 10:22:56 +0200
While I was the one to introduce it, I don't think it is correct: A
bogus continuation call issued by a tool stack domain may find another
continuation in progress. IOW we've been asserting caller controlled
state (which is reachable only via a domctl), and the early (lock-less)
check in paging_domctl() helps in a limited way only.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Roger Pau Monné <roger.pau@citrix.com>
master commit: 0956aa2219
master date: 2023-05-03 13:38:30 +0200
In COVERAGE=y but DEBUG=n builds (observed by randconfig testing) gcc12
takes issue with the subtraction of 1 from __stop___pre_ex_table[],
considering this an out of bounds access. Not being able to know that
the symbol actually marks the end of an array, the compiler is kind of
right with this diagnosis. Move the subtraction into the function.
Reported-by: Anthony Perard <anthony.perard@citrix.com>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Andrew Cooper <andrew.cooper3@citrix.com>
master commit: 353b8cc568
master date: 2023-02-22 13:50:20 +0100
Based on observations on a fair range of hardware from both primary
vendors even zero-iteration-count instances of these insns perform the
port related permission checking first.
Fixes: fe30060046 ("x86: Fix emulation of REP prefix")
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
master commit: f41c88a6fc
master date: 2023-03-30 13:07:16 +0200
First of all these were inverted: "bridge=" caused the port coordinates
to be established, while "port=" controlled the bridge coordinates. And
then the error messages being identical also wasn't helpful. While
correcting this also move both case blocks close together.
Fixes: 97fd49a7e0 ("ns16550: add support for UART parameters to be specifed with name-value pairs")
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Andrew Cooper <andrew.cooper3@citrix.com>
master commit: e692b22230
master date: 2023-03-29 14:55:37 +0200
While no caller currently invokes the function without first making sure
there is at least one shadow [1], we'd better eliminate UB here:
find_first_set_bit() requires input to be non-zero to return a well-
defined result.
Further, using find_first_set_bit() isn't very efficient in the first
place for the intended purpose.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
[1] The function has exactly two uses, and both are from OOS code, which
is HVM-only. For HVM (but not for PV) sh_mfn_is_a_page_table(),
guarding the call to sh_unsync(), guarantees at least one shadow.
Hence even if sh_page_has_multiple_shadows() returned a bogus value
when invoked for a PV domain, the subsequent is_hvm_vcpu() and
oos_active checks (the former being redundant with the latter) will
compensate. (Arguably that oos_active check should come first, for
both clarity and efficiency reasons.)
master commit: 2896224a4e
master date: 2023-03-24 11:07:08 +0100
An earlier change with the same title (commit 1ba66a870e) altered only
the path where x2apic_phys was already set to false (perhaps from the
command line). The same of course needs applying when the variable
wasn't modified yet from its initial value.
Reported-by: Elliott Mitchell <ehem+xen@m5p.com>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
master commit: 0d2686f6b6
master date: 2023-03-21 09:23:25 +0100
Marking a DRHD as controlling an IGD isn't very sensible without
checking that at the very least it's a graphics device that lives at
0000:00:02.0. Re-use the reading of the class-code to control both the
clearing of "gfx_only" and the setting of "igd_drhd_address".
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
master commit: f8c4317295
master date: 2023-03-14 10:44:08 +0100
Switches of altp2m-s always expect a valid altp2m to be in place (and
indeed altp2m_vcpu_initialise() sets the active one to be at index 0).
The compiler, however, cannot know that, and hence it cannot eliminate
p2m_get_altp2m()'s case of returnin (literal) NULL. If then the compiler
decides to special case that code path in the caller, the dereference in
instances of
atomic_dec(&p2m_get_altp2m(v)->active_vcpus);
can, to the code generator, appear to be NULL dereferences, leading to
In function 'atomic_dec',
inlined from '...' at ...:
./arch/x86/include/asm/atomic.h:182:5: error: array subscript 0 is outside array bounds of 'int[0]' [-Werror=array-bounds=]
Aid the compiler by adding a BUG_ON() checking the return value of the
problematic p2m_get_altp2m(). Since with the use of the local variable
the 2nd p2m_get_altp2m() each will look questionable at the first glance
(Why is the local variable not used here?), open-code the only relevant
piece of p2m_get_altp2m() there.
To avoid repeatedly doing these transformations, and also to limit how
"bad" the open-coding really is, convert the entire operation to an
inline helper, used by all three instances (and accepting the redundant
BUG_ON(idx >= MAX_ALTP2M) in two of the three cases).
Reported-by: Charles Arnold <carnold@suse.com>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Andrew Cooper <andrew.cooper3@citrix.com>
master commit: be62b1fc2a
master date: 2023-03-13 15:16:21 +0100
Gcc12 takes issue with core_parking_remove()'s
for ( ; i < cur_idle_nums; ++i )
core_parking_cpunum[i] = core_parking_cpunum[i + 1];
complaining that the right hand side array access is past the bounds of
1. Clearly the compiler can't know that cur_idle_nums can only ever be
zero in this case (as the sole CPU cannot be parked).
Arrange for core_parking.c's contents to not be needed altogether, and
then disable its building when NR_CPUS == 1.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Andrew Cooper <andrew.cooper3@citrix.com>
master commit: 4b0422f70f
master date: 2023-03-13 15:15:42 +0100
While the RCU variants of list insertion and removal allow lockless list
traversal (with RCU just read-locked), insertions and removals still
need serializing amongst themselves. To keep things simple, use the
domain lock for this purpose.
This is CVE-2022-42334 / part of XSA-428.
Fixes: 642123c512 ("x86/hvm: provide XEN_DMOP_pin_memory_cacheattr")
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Julien Grall <jgrall@amazon.com>
(cherry picked from commit 829ec245cf66560e3b50d140ccb3168e7fb7c945)
This is exposed via DMOP, i.e. to potentially not fully privileged
device models. With that we may not permit registration of an (almost)
unbounded amount of such regions.
This is CVE-2022-42333 / part of XSA-428.
Fixes: 642123c512 ("x86/hvm: provide XEN_DMOP_pin_memory_cacheattr")
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
(cherry picked from commit a5e768640f786b681063f4e08af45d0c4e91debf)
Pre-allocation is intended to ensure that in the course of constructing
or updating shadows there won't be any risk of just made shadows or
shadows being acted upon can disappear under our feet. The amount of
pages pre-allocated then, however, needs to account for all possible
subsequent allocations. While the use in sh_page_fault() accounts for
all shadows which may need making, so far it didn't account for
allocations coming from log-dirty tracking (which piggybacks onto the
P2M allocation functions).
Since shadow_prealloc() takes a count of shadows (or other data
structures) rather than a count of pages, putting the adjustment at the
call site of this function won't work very well: We simply can't express
the correct count that way in all cases. Instead take care of this in
the function itself, by "snooping" for L1 type requests. (While not
applicable right now, future new request sites of L1 tables would then
also be covered right away.)
It is relevant to note here that pre-allocations like the one done from
shadow_alloc_p2m_page() are benign when they fall in the "scope" of an
earlier pre-alloc which already included that count: The inner call will
simply find enough pages available then; it'll bail right away.
This is CVE-2022-42332 / XSA-427.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Tim Deegan <tim@xen.org>
(cherry picked from commit 91767a71061035ae42be93de495cd976f863a41a)
Clearly within the for_each_vcpu() the vCPU of this loop is meant, not
the (loop invariant) one the fault occurred on.
Fixes: 3d5e6a3ff3 ("x86 hvm: implement HVMOP_pagetable_dying")
Fixes: ef3b0d8d2c ("x86/shadow: shadow_table[] needs only one entry for PV-only configs")
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
master commit: f8fdceefbb
master date: 2023-01-20 09:23:42 +0100
Linux'es relatively new use of HVMOP_set_evtchn_upcall_vector has
exposed a problem with the marking of the respective vector as
pending: For quite some time Linux has been checking whether any stale
ISR or IRR bits would still be set while preparing the LAPIC for use.
This check is now triggering on the upcall vector, as the registration,
at least for APs, happens before the LAPIC is actually enabled.
In software-disabled state an LAPIC would not accept any interrupt
requests and hence no IRR bit would newly become set while in this
state. As a result it is also wrong for us to mark the upcall vector as
having a pending request when the vLAPIC is in this state.
To compensate for the "enabled" check added to the assertion logic, add
logic to (conditionally) mark the upcall vector as having a request
pending at the time the LAPIC is being software-enabled by the guest.
Note however that, like for the pt_may_unmask_irq() we already have
there, long term we may need to find a different solution. This will be
especially relevant in case yet better LAPIC acceleration would
eliminate notifications of guest writes to this and other registers.
Fixes: 7b5b8ca7df ("x86/upcall: inject a spurious event after setting upcall vector")
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Juergen Gross <jgross@suse.com>
master commit: f5d0279839
master date: 2022-12-06 13:51:49 +0100
In software-disabled state an LAPIC does not accept any interrupt
requests and hence no IRR bit would newly become set while in this
state. As a result it is also wrong for us to mark Viridian IPI or timer
vectors as having a pending request when the vLAPIC is in this state.
Such interrupts are simply lost.
Introduce a local variable in send_ipi() to help readability.
Fixes: fda96b7382 ("viridian: add implementation of the HvSendSyntheticClusterIpi hypercall")
Fixes: 26fba3c855 ("viridian: add implementation of synthetic timers")
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Paul Durrant <paul@xen.org>
master commit: 831419f829
master date: 2022-12-02 10:35:32 +0100
In software-disabled state an LAPIC does not accept any interrupt
requests and hence no IRR bit would newly become set while in this
state. As a result it is also wrong for us to mark IO-APIC or MSI
originating vectors as having a pending request when the vLAPIC is in
this state. Such interrupts are simply lost.
Introduce (IO-APIC) or re-use (MSI) a local variable to help
readability.
Fixes: 4fe21ad371 ("This patch add virtual IOAPIC support for VMX guest")
Fixes: 85715f4bc7 ("MSI 5/6: add MSI support to passthrough HVM domain")
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Andrew Cooper <andrew.cooper3@citrix.com>
master commit: f1d7aac1e3
master date: 2022-12-02 10:35:01 +0100
Not passing P2M_UNSHARE to get_page_from_gfn() means there won't even be
an attempt to unshare the referenced page, without any indication to the
caller (e.g. -EAGAIN). Note that guests have no direct control over
which of their pages are shared (or paged out), and hence they have no
way to make sure all on their own that the subsequent obtaining of a
writable type reference can actually succeed.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
Acked-by: Julien Grall <jgrall@amazon.com>
master commit: 48980cf24d
master date: 2022-10-28 11:38:32 +0200
Just like domain_soft_reset() properly zaps runstate area handles, the
secondary time area ones also need discarding to prevent guest memory
corruption once the guest is re-started.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
master commit: b80d4f8d2e
master date: 2022-10-27 11:49:09 +0200
The addition of a call to shadow_blow_tables() from shadow_teardown()
has resulted in the "no vcpus" related assertion becoming triggerable:
If domain_create() fails with at least one page successfully allocated
in the course of shadow_enable(), or if domain_create() succeeds and
the domain is then killed without ever invoking XEN_DOMCTL_max_vcpus.
Note that in-tree tests (test-resource and test-tsx) do exactly the
latter of these two.
The assertion's comment was bogus anyway: Shadow mode has been getting
enabled before allocation of vCPU-s for quite some time. Convert the
assertion to a conditional: As long as there are no vCPU-s, there's
nothing to blow away.
Fixes: e7aa55c0aa ("x86/p2m: free the paging memory pool preemptively")
Reported-by: Andrew Cooper <andrew.cooper3@citrix.com>
A similar assertion/comment pair exists in _shadow_prealloc(); the
comment is similarly bogus, and the assertion could in principle trigger
e.g. when shadow_alloc_p2m_page() is called early enough. Replace those
at the same time by a similar early return, here indicating failure to
the caller (which will generally lead to the domain being crashed in
shadow_prealloc()).
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Roger Pau Monné <roger.pau@citrix.com>
Acked-by: Andrew Cooper <andrew.cooper3@citrix.com>
master commit: a92dc2bb30
master date: 2022-10-24 15:46:11 +0200
efi_init_memory() in both relevant places is treating EFI_MEMORY_RUNTIME
higher priority than the type of the range. To avoid accessing memory at
runtime which was re-used for other purposes, make
efi_arch_process_memory_map() follow suit. While in theory the same would
apply to EfiACPIReclaimMemory, we don't actually "reclaim" or clobber
that memory (converted to E820_ACPI on x86) there (and it would be a bug
if the Dom0 kernel tried to reclaim the range, bypassing Xen's memory
management, plus it would be at least bogus if it clobbered that space),
hence that type's handling can be left alone.
Fixes: bf6501a62e ("x86-64: EFI boot code")
Fixes: facac0af87 ("x86-64: EFI runtime code")
Fixes: 6d70ea10d4 ("Add ARM EFI boot support")
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Roger Pau Monné <roger.pau@citrix.com>
Reviewed-by: Julien Grall <jgrall@amazon.com>
master commit: f324300c83
master date: 2022-10-21 12:30:24 +0200
With the addition of vmx_add_msr() calls to construct_vmcs() there are
now cases where simply freeing the VMCS isn't enough: The MSR bitmap
page as well as one of the MSR area ones (if it's the 2nd vmx_add_msr()
which fails) may also need freeing. Switch to using vmx_destroy_vmcs()
instead.
Fixes: 3bd36952da ("x86/spec-ctrl: Introduce an option to control L1D_FLUSH for HVM HAP guests")
Fixes: 53a570b285 ("x86/spec-ctrl: Support IBPB-on-entry")
Reported-by: Andrew Cooper <andrew.cooper3@citrix.com>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
master commit: 448d28309f
master date: 2022-10-12 17:57:56 +0200
protmode_load_seg() would better adhere to that "feature" of clearing
base (and limit) during NULL selector loads.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
master commit: 87a20c98d9
master date: 2022-10-11 14:30:41 +0200
Forever sinced its introduction VCPUOP_register_vcpu_time_memory_area
was available only to native domains. Linux, for example, would attempt
to use it irrespective of guest bitness (including in its so called
PVHVM mode) as long as it finds XEN_PVCLOCK_TSC_STABLE_BIT set (which we
set only for clocksource=tsc, which in turn needs engaging via command
line option).
Fixes: a5d39947cb ("Allow guests to register secondary vcpu_time_info")
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Roger Pau Monné <roger.pau@citrix.com>
master commit: b726541d94
master date: 2022-09-29 14:47:45 +0200
I haven't been able to find evidence of "-nopie" ever having been a
supported compiler option. The correct spelling is "-no-pie".
Furthermore like "-pie" this is an option which is solely passed to the
linker. The compiler only recognizes "-fpie" / "-fPIE" / "-fno-pie", and
it doesn't infer these options from "-pie" / "-no-pie".
Add the compiler recognized form, but for the possible case of the
variable also being used somewhere for linking keep the linker option as
well (with corrected spelling).
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Julien Grall <jgrall@amazon.com>
Build: Drop -no-pie from EMBEDDED_EXTRA_CFLAGS
This breaks all Clang builds, as demostrated by Gitlab CI.
Contrary to the description in ecd6b97599, -no-pie is not even an option
passed to the linker. GCC's actual behaviour is to inhibit the passing of
-pie to the linker, as well as selecting different cr0 artefacts to be linked.
EMBEDDED_EXTRA_CFLAGS is not used for $(CC)-doing-linking, and not liable to
gain such a usecase.
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Acked-by: Jan Beulich <jbeulich@suse.com>
Tested-by: Stefano Stabellini <sstabellini@kernel.org>
Fixes: ecd6b97599 ("Config.mk: correct PIE-related option(s) in EMBEDDED_EXTRA_CFLAGS")
master commit: ecd6b97599
master date: 2022-09-08 09:25:26 +0200
master commit: 13a7c0074a
master date: 2022-09-27 15:40:42 -0700
While the SDM isn't very clear about this, our present behavior make
Linux 5.19 unhappy. As of commit 8ad7e8f69695 ("x86/fpu/xsave: Support
XSAVEC in the kernel") they're using this CPUID output also to size
the compacted area used by XSAVEC. Getting back zero there isn't really
liked, yet for PV that's the default on capable hardware: XSAVES isn't
exposed to PV domains.
Considering that the size reported is that of the compacted save area,
I view Linux'es assumption as appropriate (short of the SDM properly
considering the case). Therefore we need to populate the field also when
only XSAVEC is supported for a guest.
Fixes: 460b9a4b36 ("x86/xsaves: enable xsaves/xrstors for hvm guest")
Fixes: 8d050ed109 ("x86: don't expose XSAVES capability to PV guests")
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Andrew Cooper <andrew.cooper3@citrix.com>
master commit: c3bd0b83ea
master date: 2022-08-24 14:23:59 +0200
While the comment next to the lock dropping in preparation of
recursively calling acquire_grant_for_copy() mistakenly talks about the
rd == td case (excluded a few lines further up), the same concerns apply
to the calling of release_grant_for_copy() on a subsequent error path.
This is CVE-2022-33748 / XSA-411.
Fixes: ad48fb963d ("gnttab: fix transitive grant handling")
Signed-off-by: Jan Beulich <jbeulich@suse.com>
master commit: 6e3aab858e
master date: 2022-10-11 14:29:30 +0200
Subsequently sh_set_toplevel_shadow() will be adjusted to install a
blank entry in case prealloc fails. There are, in fact, pre-existing
error paths which would put in place a blank entry. The 4- and 2-level
code in sh_update_cr3(), however, assume the top level entry to be
valid.
Hence bail from the function in the unlikely event that it's not. Note
that 3-level logic works differently: In particular a guest is free to
supply a PDPTR pointing at 4 non-present (or otherwise deemed invalid)
entries. The guest will crash, but we already cope with that.
Really mfn_valid() is likely wrong to use in sh_set_toplevel_shadow(),
and it should instead be !mfn_eq(gmfn, INVALID_MFN). Avoid such a change
in security context, but add a respective assertion.
This is part of CVE-2022-33746 / XSA-410.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Tim Deegan <tim@xen.org>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
master commit: eac000978c
master date: 2022-10-11 14:22:24 +0200
hap_make_monitor_table() will return INVALID_MFN if it encounters an
error condition, but hap_update_paging_modes() wasn’t handling this
value, resulting in an inappropriate value being stored in
monitor_table. This would subsequently misguide at least
hap_vcpu_teardown(). Avoid this by bailing early.
Further, when a domain has/was already crashed or (perhaps less
important as there's no such path known to lead here) is already dying,
avoid calling domain_crash() on it again - that's at best confusing.
This is part of CVE-2022-33746 / XSA-410.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
master commit: 5b44a61180
master date: 2022-10-11 14:21:56 +0200
The last "wildcard" use of either function went away with f591755823
("IOMMU/PCI: don't let domain cleanup continue when device de-assignment
failed"). Don't allow them to be called this way anymore. Besides
simplifying the code this also fixes two bugs:
1) When seg != -1, the outer loops should have been terminated after the
first iteration, or else a device with the same BDF but on another
segment could be found / returned.
Reported-by: Rahul Singh <rahul.singh@arm.com>
2) When seg == -1 calling get_pseg() is bogus. The function (taking a
u16) would look for segment 0xffff, which might exist. If it exists,
we might then find / return a wrong device.
In pci_get_pdev_by_domain() also switch from using the per-segment list
to using the per-domain one, with the exception of the hardware domain
(see the code comment there).
While there also constify "pseg" and drop "pdev"'s already previously
unnecessary initializer.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Rahul Singh <rahul.singh@arm.com>
Tested-by: Rahul Singh <rahul.singh@arm.com>
master commit: 8cf6e07389
master date: 2022-08-12 08:34:33 +0200
Commit 68f5aac012 ("build: suppress future GNU ld warning about RWX
load segments") didn't quite cover all the cases: Apparently I missed
ones in the building of 32-bit helper objects because of only looking at
incremental builds (where those wouldn't normally be re-built). Clone
the workaround there to the specific Makefile in question.
Reported-by: Andrew Cooper <andrew.cooper3@citrix.com>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Andrew Cooper <andrew.cooper3@citrix.com>
master commit: 3eb1865ae3
master date: 2022-08-11 17:45:12 +0200
Passing -mno-sse alone is not enough: The compiler may still find
(questionable) reasons to use MMX insns. In particular with gcc12 use
of MOVD+PUNPCKLDQ+MOVQ was observed in an apparent attempt to auto-
vectorize the storing of two adjacent zeroes, 32 bits each.
Reported-by: ChrisD <chris@dalessio.org>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Andrew Cooper <andrew.cooper3@citrix.com>
master commit: 6fe2e39a02
master date: 2022-07-20 15:48:49 +0200
Already ISE rev 044 added text to this effect; rev 045 further dropped
leftover earlier text indicating the contrary:
- ENQCMD requires the low 32 bits of the memory operand to be clear,
- ENDCMDS requires bits 20...30 of the memory operand to be clear.
Fixes: d273859687 ("x86emul: support ENQCMD insns")
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Andrew Cooper <andrew.cooper3@citrix.com>
master commit: d620c66bdb
master date: 2022-07-20 15:46:48 +0200
While a number of issues we previously had with pre-release gcc12 were
fixed in the final release, we continue to have one issue (with multiple
instances) when doing release builds (i.e. at higher optimization
levels): The compiler takes issue with subtracting (always 1 in our
case) from artifical labels (expressed as array) marking the end of
certain regions. This isn't an unreasonable position to take. Simply
hide the "array-ness" by casting to an integer type. To keep things
looking consistently, apply the same cast also on the respective
expressions dealing with the starting addresses. (Note how
efi_arch_memory_setup()'s l2_table_offset() invocations avoid a similar
issue by already having the necessary casts.) In is_xen_fixed_mfn()
further switch from __pa() to virt_to_maddr() to better match the left
sides of the <= operators.
Reported-by: Charles Arnold <carnold@suse.com>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Andrew Cooper <andrew.cooper3@citrix.com>
master commit: 9723507daf
master date: 2022-07-19 08:37:29 +0200
There are command line controls for this and the default also isn't "always
enable when hardware supports it", which logging should take into account.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
master commit: fdbf8bdfeb
master date: 2022-07-19 08:36:53 +0200
Move the "credit expired" loop exit to the middle of the loop,
immediately after "return true". This way having reached the goal on the
last iteration would be reported as success to the caller, rather than
as "timed out".
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Anthony PERARD <anthony.perard@citrix.com>
master commit: d8f8cb8bdd
master date: 2022-07-18 17:48:18 +0200
While in principle possible also under other conditions as long as other
parallel operations potentially consuming memory aren't "locked out", in
particular with IOMMU large page mappings used in Dom0 (for PV when in
strict mode; for PVH when not sharing page tables with HAP) ballooning
out of individual pages can actually lead to less free memory available
afterwards. This is because to split a large page, one or more page
table pages are necessary (one per level that is split).
When rebooting a guest I've observed freemem() to fail: A single page
was required to be ballooned out (presumably because of heap
fragmentation in the hypervisor). This ballooning out of a single page
of course went fast, but freemem() then found that it would require to
balloon out another page. This repeating just another time leads to the
function to signal failure to the caller - without having come anywhere
near the designated 30s that the whole process is allowed to not make
any progress at all.
Convert from a simple retry count to actually calculating elapsed time,
subtracting from an initial credit of 30s. Don't go as far as limiting
the "wait_secs" value passed to libxl_wait_for_memory_target(), though.
While this leads to the overall process now possibly taking longer (if
the previous iteration ended very close to the intended 30s), this
compensates to some degree for the value passed really meaning "allowed
to run for this long without making progress".
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Anthony PERARD <anthony.perard@citrix.com>
master commit: e58370df76
master date: 2022-07-12 15:25:00 +0200
When this logic was moved, it was moved across the point where nx is
updated to hold the new type for the page. IOW originally it was
equivalent to using x (and perhaps x would better have been used), but
now it isn't anymore. Switch to using x, which then brings things in
line again with the slightly earlier comment there (now) talking about
transitions _from_ writable.
I have to confess though that I cannot make a direct connection between
the reported observed behavior of guests leaving several pages around
with pending general references and the change here. Repeated testing,
nevertheless, confirms the reported issue is no longer there.
This is CVE-2022-33745 / XSA-408.
Reported-by: Charles Arnold <carnold@suse.com>
Fixes: 8cc5036bc3 ("x86/pv: Fix ABAC cmpxchg() race in _get_page_type()")
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
master commit: a9949efb28
master date: 2022-07-26 14:54:34 +0200
As per [1] the expansion of the pirq_dpci() macro causes a -Waddress
controlled warning (enabled implicitly in our builds, if not by default)
tying the middle part of the involved conditional expression to the
surrounding boolean context. Work around this by introducing a local
inline function in the affected source file.
Reported-by: Andrew Cooper <andrew.cooper3@citrix.com>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Roger Pau Monné <roger.pau@citrix.com>
[1] https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102967
master commit: 80ad8db8a4
master date: 2022-06-15 10:19:32 +0200
While PGT_pae_xen_l2 will be zapped once the type refcount of an L2 page
reaches zero, it'll be retained as long as the type refcount is non-
zero. Hence any checking against the requested type needs to either zap
the bit from the type or include it in the used mask.
Fixes: 9186e96b19 ("x86/pv: Clean up _get_page_type()")
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
master commit: c2095ac76b
master date: 2022-06-10 10:21:06 +0200
IOMMU code mapping / unmapping devices and interrupts will misbehave if
a wrong command line option declared a function "phantom" when there's a
real device at that position. Warn about this and adjust the specified
stride (in the worst case ignoring the option altogether).
Requested-by: Andrew Cooper <andrew.cooper3@citrix.com>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
master commit: 444b555dc9
master date: 2022-05-20 12:20:35 +0200
While for C files the compiler is supposed to arrange for emitting
respective information, for assembly sources we're responsible ourselves.
Present GNU ld master started warning about such, and hence 2.39 is
anticipated to have this warning.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Andrew Cooper <andrew.cooper3@citrix.com>
Acked-by: Julien Grall <jgrall@amazon.com>
master commit: 62d22296a9
master date: 2022-05-18 11:18:45 +0200
We cannot really avoid such and we're also not really at risk because of
them, as we control page table permissions ourselves rather than relying
on a loader of some sort. Present GNU ld master started warning about
such, and hence 2.39 is anticipated to have this warning.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Andrew Cooper <andrew.cooper3@citrix.com>
Acked-by: Julien Grall <jgrall@amazon.com>
master commit: 68f5aac012
master date: 2022-05-18 11:17:19 +0200
While it is okay for IOMMU page tables to be set up for guests starting
in PoD mode, actual device assignment may only occur once all PoD
entries have been removed from the P2M. So far this was enforced only
for boot-time assignment, and only in the tool stack.
Also use the new function to replace p2m_pod_entry_count(): Its unlocked
access to p2m->pod.entry_count wasn't really okay (irrespective of the
result being stale by the time the caller gets to see it). Nor was the
use of that function in line with the immediately preceding comment: A
PoD guest isn't just one with a non-zero entry count, but also one with
a non-empty cache (e.g. prior to actually launching the guest).
To allow the tool stack to see a consistent snapshot of PoD state, move
the tail of XENMEM_{get,set}_pod_target handling into a function, adding
proper locking there.
In libxl take the liberty to use the new local variable r also for a
pre-existing call into libxc.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
master commit: ad4312d764
master date: 2022-05-02 08:48:02 +0200
Besides the reporter's issue of hitting a NULL deref when !CONFIG_GDBSX,
XEN_DOMCTL_test_assign_device can legitimately end up having NULL passed
here, when the domctl was passed DOMID_INVALID.
Fixes: 71e617a6b8 ("use is_iommu_enabled() where appropriate...")
Reported-by: Cheyenne Wills <cheyenne.wills@gmail.com>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Paul Durrant <paul@xen.org>
Reviewed-by: Juergen Gross <jgross@suse.com>
master commit: fa4d84e6dd
master date: 2022-04-26 10:25:54 +0200
The field taking the value 7 (resulting in 18-bit DIDs when using the
calculation in cap_ndoms(), when the DID fields are only 16 bits wide)
is reserved. Instead of misbehaving in case we would encounter such an
IOMMU, refuse to use it.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
master commit: a1545fbf45
master date: 2022-04-20 10:54:26 +0200
%pp wants the address of an SBDF, not that of a PCI device.
Fixes: b4f2116060 ("vpci/msix: fix PBA accesses")
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
master commit: d3f61beea4
master date: 2022-04-07 18:01:24 +0200
Despite the comment there infinite recursion was still possible, by
flip-flopping between two domains. This is because prev_dom is derived
from the DID found in the context entry, which was already updated by
the time error recovery is invoked. Simply introduce yet another mode
flag to prevent rolling back an in-progress roll-back of a prior
mapping attempt.
Also drop the existing recursion prevention for having been dead anyway:
Earlier in the function we already bail when prev_dom == domain.
Fixes: 8f41e481b4 ("VT-d: re-assign devices directly")
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
master commit: 99d829dba1
master date: 2022-04-07 12:31:16 +0200
First there's a printk() which actually wrongly uses pdev in the first
place: We want to log the coordinates of the (perhaps fake) device
acted upon, which may not be pdev.
Then it was quite pointless for eb19326a32 ("VT-d: prepare for per-
device quarantine page tables (part I)") to add a domid_t parameter to
domain_context_unmap_one(): It's only used to pass back here via
me_wifi_quirk() -> map_me_phantom_function(). Drop the parameter again.
Finally there's the invocation of domain_context_mapping_one(), which
needs to be passed the correct domain ID. Avoid taking that path when
pdev is NULL and the quarantine state is what would need restoring to.
This means we can't security-support non-PCI-Express devices with RMRRs
(if such exist in practice) any longer; note that as of trhe 1st of the
two commits referenced below assigning them to DomU-s is unsupported
anyway.
Fixes: 8f41e481b4 ("VT-d: re-assign devices directly")
Fixes: 14dd241aad ("IOMMU/x86: use per-device page tables for quarantining")
Coverity ID: 1503784
Reported-by: Andrew Cooper <andrew.cooper3@citrix.com>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
master commit: 608394b906
master date: 2022-04-07 12:30:19 +0200
If get_iommu_domid() in domain_context_unmap_one() fails, we better
wouldn't clear the context entry in the first place, as we're then unable
to issue the corresponding flush. However, we have no need to look up the
DID in the first place: What needs flushing is very specifically the DID
that was in the context entry before our clearing of it.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
master commit: 445ab9852d
master date: 2022-04-07 12:29:03 +0200
Devices with RMRRs / unity mapped regions, due to it being unspecified
how/when these memory regions may be accessed, may not be left
disconnected from the mappings of these regions (as long as it's not
certain that the device has been fully quiesced). Hence even the page
tables used when quarantining such devices need to have mappings of
those regions. This implies installing page tables in the first place
even when not in scratch-page quarantining mode.
This is CVE-2022-26361 / part of XSA-400.
While for the purpose here it would be sufficient to have devices with
RMRRs / unity mapped regions use per-device page tables, extend this to
all devices (in scratch-page quarantining mode). This allows the leaf
pages to be mapped r/w, thus covering also memory writes (rather than
just reads) issued by non-quiescent devices.
Set up quarantine page tables as late as possible, yet early enough to
not encounter failure during de-assign. This means setup generally
happens in assign_device(), while (for now) the one in deassign_device()
is there mainly to be on the safe side.
As to the removal of QUARANTINE_SKIP() from domain_context_unmap_one():
I think this was never really needed there, as the function explicitly
deals with finding a non-present context entry. Leaving it there would
require propagating pgd_maddr into the function (like was done by "VT-d:
prepare for per-device quarantine page tables" for
domain_context_mapping_one()).
In VT-d's DID allocation function don't require the IOMMU lock to be
held anymore: All involved code paths hold pcidevs_lock, so this way we
avoid the need to acquire the IOMMU lock around the new call to
context_set_domain_id().
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Paul Durrant <paul@xen.org>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
master commit: 14dd241aad
master date: 2022-04-05 14:24:18 +0200
We will want to use the constant elsewhere.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Paul Durrant <paul@xen.org>
master commit: a038b514c1
master date: 2022-04-05 14:20:04 +0200
The page tables just created aren't hooked up yet anywhere, so there's
nothing that could be present in any TLB, and hence nothing to flush.
Dropping this flush is, at least on the VT-d side, a prereq to per-
device domain ID use when quarantining devices, as dom_io isn't going
to be assigned a DID anymore: The warning in get_iommu_did() would
trigger.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Paul Durrant <paul@xen.org>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
master commit: 54c5cef492
master date: 2022-04-05 14:19:42 +0200
In order to subsequently enable per-device quarantine page tables, we'll
need domain-ID-like identifiers to be inserted in the respective device
(AMD) or context (Intel) table entries alongside the per-device page
table root addresses.
Make use of "real" domain IDs occupying only half of the value range
coverable by domid_t.
Note that in VT-d's iommu_alloc() I didn't want to introduce new memory
leaks in case of error, but existing ones don't get plugged - that'll be
the subject of a later change.
The VT-d changes are slightly asymmetric, but this way we can avoid
assigning pseudo domain IDs to devices which would never be mapped while
still avoiding to add a new parameter to domain_context_unmap().
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Paul Durrant <paul@xen.org>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
master commit: 97af062b89
master date: 2022-04-05 14:19:10 +0200
Replace the passing of struct domain * by domid_t in preparation of
per-device quarantine page tables also requiring per-device pseudo
domain IDs, which aren't going to be associated with any struct domain
instances.
No functional change intended (except for slightly adjusted log message
text).
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Paul Durrant <paul@xen.org>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
master commit: 7131163c48
master date: 2022-04-05 14:18:48 +0200
Arrange for domain ID and page table root to be passed around, the latter in
particular to domain_pgd_maddr() such that taking it from the per-domain
fields can be overridden.
No functional change intended.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Paul Durrant <paul@xen.org>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
master commit: eb19326a32
master date: 2022-04-05 14:18:26 +0200
Devices with unity map ranges, due to it being unspecified how/when
these memory ranges may get accessed, may not be left disconnected from
their unity mappings (as long as it's not certain that the device has
been fully quiesced). Hence rather than tearing down the old root page
table pointer and then establishing the new one, re-assignment needs to
be done in a single step.
This is CVE-2022-26360 / part of XSA-400.
Reported-by: Roger Pau Monné <roger.pau@citrix.com>
Similarly quarantining scratch-page mode relies on page tables to be
continuously wired up.
To avoid complicating things more than necessary, treat all devices
mostly equally, i.e. regardless of their association with any unity map
ranges. The main difference is when it comes to updating DTEs, which need
to be atomic when there are unity mappings. Yet atomicity can only be
achieved with CMPXCHG16B, availability of which we can't take for given.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Paul Durrant <paul@xen.org>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
master commit: 1fa6e9aa36
master date: 2022-04-05 14:18:04 +0200
Devices with RMRRs, due to it being unspecified how/when the specified
memory regions may get accessed, may not be left disconnected from their
respective mappings (as long as it's not certain that the device has
been fully quiesced). Hence rather than unmapping the old context and
then mapping the new one, re-assignment needs to be done in a single
step.
This is CVE-2022-26359 / part of XSA-400.
Reported-by: Roger Pau Monné <roger.pau@citrix.com>
Similarly quarantining scratch-page mode relies on page tables to be
continuously wired up.
To avoid complicating things more than necessary, treat all devices
mostly equally, i.e. regardless of their association with any RMRRs. The
main difference is when it comes to updating context entries, which need
to be atomic when there are RMRRs. Yet atomicity can only be achieved
with CMPXCHG16B, availability of which we can't take for given.
The seemingly complicated choice of non-negative return values for
domain_context_mapping_one() is to limit code churn: This way callers
passing NULL for pdev don't need fiddling with.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
master commit: 8f41e481b4
master date: 2022-04-05 14:17:42 +0200
Despite putting in quite a bit of effort it was not possible to
establish why exactly this code exists (beyond possibly sanity
checking). Instead of a subsequent change further complicating this
logic, simply get rid of it.
Take the opportunity and move the respective unmap_vtd_domain_page() out
of the locked region.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
Reviewed-by: Paul Durrant <paul@xen.org>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
master commit: a680b8134b
master date: 2022-04-05 14:17:21 +0200
This is to make more obvious that nothing outside of domain_iommu(d)
actually changes or is otherwise needed by the function.
No functional change intended.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
Reviewed-by: Paul Durrant <paul@xen.org>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
master commit: eba09b9dd7
master date: 2022-04-05 14:16:46 +0200
In the event that the RMRR mappings are essential for device operation,
they should be established before updating the device's context entry,
while they should be torn down only after the device's context entry was
successfully cleared.
Also switch to %pd in related log messages.
Fixes: fa88cfadf9 ("vt-d: Map RMRR in intel_iommu_add_device() if the device has RMRR")
Fixes: 8b99f4400b ("VT-d: fix RMRR related error handling")
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
master commit: 3221f270cf
master date: 2022-04-05 14:16:10 +0200
In the event that the RMRR mappings are essential for device operation,
they should be established before updating the device's context entry,
while they should be torn down only after the device's context entry was
successfully updated.
Also adjust a related log message.
This is CVE-2022-26358 / part of XSA-400.
Fixes: 8b99f4400b ("VT-d: fix RMRR related error handling")
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
Reviewed-by: Paul Durrant <paul@xen.org>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
master commit: 78a40f8b5d
master date: 2022-04-05 14:15:33 +0200
The function may be called without any locks held (leaving aside the
domctl one, which we surely don't want to depend on here), so needs to
play safe wrt other accesses to domid_map[] and domid_bitmap[]. This is
to avoid context_set_domain_id()'s writing of domid_map[] to be reset to
zero right away in the case of it racing the freeing of a DID.
For the interaction with context_set_domain_id() and did_to_domain_id()
see the code comment.
{check_,}cleanup_domid_map() are called with pcidevs_lock held or during
domain cleanup only (and pcidevs_lock is also held around
context_set_domain_id()), i.e. racing calls with the same (dom, iommu)
tuple cannot occur.
domain_iommu_domid(), besides its use by cleanup_domid_map(), has its
result used only to control flushing, and hence a stale result would
only lead to a stray extra flush.
This is CVE-2022-26357 / XSA-399.
Fixes: b9c20c7878 ("VT-d: per-iommu domain-id")
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
master commit: d9eca7bb6c
master date: 2022-04-05 14:12:27 +0200
While not triggered by the trivial xen_nop in-tree patch on
staging/master, that patch exposes a problem on the stable trees, where
all functions have ENDBR inserted. When NOP-ing out a range, we need to
account for this. Handle this right in livepatch_insn_len().
This requires livepatch_insn_len() to be called _after_ ->patch_offset
was set.
Fixes: 6974c75180 ("xen/x86: Livepatch: support patching CET-enhanced functions")
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
master commit: 8a87b9a0fb
master date: 2022-03-31 10:45:46 +0200
For an unknown reason I added back the operator while backporting,
despite 4.16 having c06e3d8103 ("VT-d: per-domain IOMMU bitmap needs
to have dynamic size"). I can only assume that I mistakenly took the
4.15 backport as basis and/or reference.
Fixes: fa45f6b556 ("VT-d: split domid map cleanup check into a function")
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
Element size for this opcode depends on EVEX.W, not the low opcode bit.
Make use of AVX512BW being a prereq to AVX512_BITALG and move the case
label there, adding an AVX512BW feature check.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
master commit: eddf13b5e9
master date: 2022-02-14 10:08:38 +0100
While the problem report was for extreme errors, even smaller ones would
better be avoided: The calculated period to run calibration loops over
can (and usually will) be shorter than the actual time elapsed between
first and last platform timer and TSC reads. Adjust values returned from
the init functions accordingly.
On a Skylake system I've tested this on accuracy (using HPET) went from
detecting in some cases more than 220kHz too high a value to about
±2kHz. On other systems (or on this system, but with PMTMR) the original
error range was much smaller, with less (in some cases only very little)
improvement.
Reported-by: James Dingwall <james-xen@dingwall.me.uk>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
master commit: a5c9a80af3
master date: 2022-01-13 14:31:52 +0100
Jan Beulich