Turn the if()s into switch()es, as we're going to need AMD sections.
Move the RTM adjustments into the Intel section, where they ought to live.
No functional change.
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
(cherry picked from commit 7c5025394d)
The check of this_softirq_pending must be performed with irqs disabled, but
this property was broken by an attempt to optimise entry/exit latency.
Commit c227233ad64c in Linux (which we copied into Xen) was fixed up by
edc8fc01f608 in Linux, which we have so far missed.
Going to sleep without waking on interrupts is nonsensical outside of
play_dead(), so overload this to select between two possible MWAITs, the
second using the STI shadow to cover MWAIT for exactly the same reason as we
do in safe_halt().
Fixes: b17e0ec72e ("x86/mwait-idle: enable interrupts before C1 on Xeons")
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
(cherry picked from commit 9b0f0f6e23)
In order elide IPIs, we must be able to identify whether a target CPU is in
MWAIT at the point it is woken up. i.e. the store to wake it up must also
identify the state.
Create a new in_mwait variable beside __softirq_pending, so we can use a
CMPXCHG to set the softirq while also observing the status safely. Implement
an x86 version of arch_pend_softirq() which does this.
In mwait_idle_with_hints(), advertise in_mwait, with an explanation of
precisely what it means. X86_BUG_MONITOR can be accounted for simply by not
advertising in_mwait.
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Acked-by: Roger Pau Monné <roger.pau@citrix.com>
(cherry picked from commit 3e0bc4b503)
x86 is the only architecture wanting an optimisation here, but the
test_and_set_bit() is a store into the monitored line (i.e. will wake up the
target) and, prior to the removal of the broken IPI-elision algorithm, was
racy, causing unnecessary IPIs to be sent.
To do this in a race-free way, the store to the monited line needs to also
sample the status of the target in one atomic action. Implement a new arch
helper with different semantics; to make the softirq pending and decide about
IPIs together. For now, implement the default helper. It will be overridden
by x86 in a subsequent change.
No functional change.
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
(cherry picked from commit b473e5e212)
With the recent simplifications, it becomes obvious that smp_mb() isn't the
right barrier. Strictly speaking, MONITOR is ordered as a load, but smp_rmb()
isn't correct either, as this only pertains to local ordering. All we need is
a compiler barrier().
Merge the barier() into the monitor() itself, along with an explantion.
No functional change.
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Acked-by: Roger Pau Monné <roger.pau@citrix.com>
(cherry picked from commit e7710dd843)
cpuidle_wakeup_mwait() is a TOCTOU race. The cpumask_and() sampling
cpuidle_mwait_flags can take a arbitrary period of time, and there's no
guarantee that the target CPUs are still in MWAIT when writing into
mwait_wakeup(cpu).
The consequence of the race is that we'll fail to IPI certain targets. Also,
there's no guarantee that mwait_idle_with_hints() will raise a TIMER_SOFTIRQ
on it's way out.
The fundamental bug is that the "in_mwait" variable needs to be in the
monitored line, and not in a separate cpuidle_mwait_flags variable, in order
to do this in a race-free way.
Arranging to fix this while keeping the old implementation is prohibitive, so
strip the current one out in order to implement the new one cleanly. In the
interim, this causes IPIs to be used unconditionally which is safe albeit
suboptimal.
Fixes: 3d521e933e ("cpuidle: mwait on softirq_pending & remove wakeup ipis")
Fixes: 1adb34ea84 ("CPUIDLE: re-implement mwait wakeup process")
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
(cherry picked from commit 3faf0866a3)
For features which are unconditionally set in the max policies, making the
default policy to match the host can be done with a conditional clear.
This is simpler than the unconditional clear, conditional set currently
performed.
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
(cherry picked from commit 30f8fed68f)
Most of the patch (handling of CPUIDLE_FLAG_IBRS) is fine, but the
adjustements to mwait_idle() are not; spec_ctrl_enter_idle() does more than
just alter MSR_SPEC_CTRL.IBRS.
The only reason this doesn't need an XSA is because the unconditional
spec_ctrl_{enter,exit}_idle() in mwait_idle_with_hints() were left unaltered,
and thus the MWAIT remained properly protected.
There (would have been) two problems. In the ibrs_disable (== deep C) case:
* On entry, VERW and RSB-stuffing are architecturally skipped.
* On exit, there's a branch crossing the WRMSR which reinstates the
speculative safety for indirect branches.
All this change did was double up the expensive operations in the deep C case,
and fail to optimise the intended case.
I have an idea of how to plumb this more nicely, but it requires larger
changes to legacy IBRS handling to not make spec_ctrl_enter_idle() vulnerable
in other ways. In the short term, simply take out the perf hit.
Fixes: 08acdf9a26 ("x86/mwait-idle: disable IBRS during long idle")
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
(cherry picked from commit 07d7163334)
Commit 48d32458bc ("x86, idle: add barriers to CLFLUSH workaround") was
inherited from Linux and added MFENCEs around the AAI65 errata fix.
The SDM now states:
Executions of the CLFLUSH instruction are ordered with respect to each
other and with respect to writes, locked read-modify-write instructions,
and fence instructions[1].
with footnote 1 reading:
Earlier versions of this manual specified that executions of the CLFLUSH
instruction were ordered only by the MFENCE instruction. All processors
implementing the CLFLUSH instruction also order it relative to the other
operations enumerated above.
I.e. the MFENCEs came about because of an incorrect statement in the SDM.
The Spec Update (no longer available on Intel's website) simply says "issue a
CLFLUSH", with no mention of MFENCEs.
As this erratum is specific to Intel, it's fine to remove the the MFENCEs; AMD
CPUs of a similar vintage do sport otherwise-unordered CLFLUSHs.
Move the feature bit into the BUG range (rather than FEATURE), and move the
workaround into monitor() itself.
The erratum check itself must use setup_force_cpu_cap(). It needs activating
if any CPU needs it, not if all of them need it.
Fixes: 48d32458bc ("x86, idle: add barriers to CLFLUSH workaround")
Fixes: 96d1b237ae ("x86/Intel: work around Xeon 7400 series erratum AAI65")
Link: https://web.archive.org/web/20090219054841/http://download.intel.com/design/xeon/specupdt/32033601.pdf
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
(cherry picked from commit f77ef34435)
They're not used by any other translation unit, so shouldn't live in
asm/processor.h, which is included almost everywhere.
Our new toolchain baseline knows the MONITOR/MWAIT instructions, so use them
directly rather than using raw hex.
Change the hint/extention parameters from long to int. They're specified to
remain 32bit operands even 64-bit mode.
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
(cherry picked from commit 61e10fc28c)
LFENCE_DISPATCH used to be a synthetic feature, but was given a real CPUID bit
by AMD. The define wasn't moved when this was changed.
NSCB has always been a real CPUID bit, and was misplaced when introduced in
the synthetic block alongside LFENCE_DISPATCH.
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
(cherry picked from commit 6a039b0500)
When support was originally introduced, ebx, ecx and edx were reserved and
should have been zeroed in recalculate_misc() to avoid leaking into guests.
Since then, fields have been added into ebx. Guests can't load microcode, so
shouldn't see ucode_size, and while in principle we do want to support larger
RAP sizes in guests, virtualising this for guests depends on AMD procuding any
official documentation for ERAPS, which is long overdue and with no ETA.
This patch will cause a difference in guests on Zen5 CPUs, but as the main
ERAPS feature is hidden, guests should be ignoring the rap_size field too.
Fixes: e9b4fe2636 ("x86/cpuid: support LFENCE always serialising CPUID bit")
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
master commit: 10dc35c516
master date: 2025-07-02 18:25:03 +0100
See the respective bullet point in the Caveats section of
https://gcc.gnu.org/gcc-15/changes.html.
While I'm unaware of us currently relying on the pre-gcc15 behavior,
let's still play safe and retain what unknowingly we may have been
relying upon.
According to my observations, on x86 generated code changes
- somewhere deep in modify_bars(), presumably from the struct map_data
initializer in apply_map() (a new MOVQ),
- in vpci_process_pending(), apparently again from the struct map_data
initializer (and again a new MOVQ),
- near the top of find_cpio_data(), presumably from the struct cpio_data
initializer (a MOVW changing to a MOVQ).
Requested-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: 86da14178d
master date: 2025-07-01 11:23:59 +0200
If the total returned by Xen is more than the number of elements
allocated, it means that the buffer was too small and so the data is
incomplete. Retry to get all the data.
Signed-off-by: Ross Lagerwall <ross.lagerwall@citrix.com>
Reviewed-by: Anthony PERARD <anthony.perard@vates.tech>
master commit: 1ff7f87e8f
master date: 2025-06-11 11:07:00 +0200
xc_pm_get_pxstat() requires the caller to allocate the pt and trans_pt
buffers but then calls xc_pm_get_max_px() to determine how big they are
(and hence how much Xen will copy into them). This is susceptible to
races if xc_pm_get_max_px() changes so avoid the problem by requiring
the caller to also pass in the size of the buffers.
Suggested-by: Jan Beulich <jbeulich@suse.com>
Signed-off-by: Ross Lagerwall <ross.lagerwall@citrix.com>
Reviewed-by: Anthony PERARD <anthony.perard@vates.tech>
master commit: de6a05a8a0
master date: 2025-06-11 11:06:45 +0200
For exception recovery in the stubs, the registered address for fixup is the
return address of the CALL entering the stub.
In invoke_stub(), the '.Lret%=:' label is the wrong side of the 'post'
parameter. The 'post' parameter is non-empty in cases where the arithmetic
flags of the operation need recovering.
Split the line to separate 'pre' and 'post', making it more obvious that the
return address label was in the wrong position.
However, in the case that an exception did occur, we want to skip 'post' as
it's logically part of the operation which had already failed. Therefore, add
a new skip label and use that for the exception recovery path.
This is XSA-470 / CVE-2025-27465
Fixes: 79903e50db ("x86emul: catch exceptions occurring in stubs")
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
(cherry picked from commit b9f8311975)
The 2 missed ones are: register_runstate_phys_area and
register_vcpu_time_phys_area.
Fixes: d5df44275e ("domain: introduce GADDR based runstate area registration alternative")
Fixes: 60e544a8c5 ("x86: introduce GADDR based secondary time area registration alternative")
Signed-off-by: Gang Ji <gang.ji@cloud.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
master commit: ed939591e4
master date: 2025-06-11 11:08:11 +0200
If set_px_pminfo is called a second time with a larger state_count than
the first call, calls to PMSTAT_get_pxstat will read beyond the end of
the pt and trans_pt buffers allocated in cpufreq_statistic_init() since
they would have been allocated with the original state_count.
Secondly, the states array leaks on each subsequent call of
set_px_pminfo.
Fix both these issues by ignoring subsequent calls to set_px_pminfo if
it completed successfully previously. Return success rather than an
error to avoid errors in the dom0 kernel log when reloading the
xen_acpi_processor module.
At the same time, fix a leak of the states array on error.
Signed-off-by: Ross Lagerwall <ross.lagerwall@citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
master commit: 2f36bef3d8
master date: 2025-06-11 11:06:24 +0200
Check that the total number of states passed in and hence the size of
buffers is sufficient to avoid writing more than the caller has
allocated.
The interface is not explicit about whether getpx.total is expected to
be set by the caller in this case but since it is always set in
libxenctrl it seems reasonable to check it and make it explicit.
Fixes: c06a7db0c5 ("X86 and IA64: Update cpufreq statistic logic for supporting both x86 and ia64")
Signed-off-by: Ross Lagerwall <ross.lagerwall@citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
# Commit fb16c7411d
# Date 2025-06-18 09:25:09 +0200
# Author Jan Beulich <jbeulich@suse.com>
# Committer Jan Beulich <jbeulich@suse.com>
x86/pmstat: correct PMSTAT_get_pxstat buffer size checking
min(pmpt->perf.state_count, op->u.getpx.total) == op->u.getpx.total can
be expressed differently as pmpt->perf.state_count >= op->u.getpx.total.
Copying when the two are equal is fine; (partial) copying when the state
count is larger than the number of array elements that a buffer was
allocated to hold is what - as per the comment - we mean to avoid. Drop
the use of min() again, but retain its effect for the subsequent copying
from pxpt->u.pt.
Fixes: aa70996a68 ("x86/pmstat: Check size of PMSTAT_get_pxstat buffers")
Reported-by: Andrew Cooper <andrew.cooper3@citrix.com>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Ross Lagerwall <ross.lagerwall@citrix.com>
master commit: aa70996a68
master date: 2025-06-11 11:05:42 +0200
master commit: fb16c7411d
master date: 2025-06-18 09:25:09 +0200
Only access the HVM union b_info->u.hvm on HVM guests. The union
access is not guarded, so this reads and sets the default even on
non-HVM guests. Usually this doesn't matter as PV and PVH unions are
smaller and zero-initialized, but the zero default will be re-written as
a -1 boolean. Generally, it could incorrectly set b_info->altp2m
through aliased data.
Fixes: 0291089f6e ("xen: enable altp2m at create domain domctl")
Signed-off-by: Jason Andryuk <jason.andryuk@amd.com>
Acked-by: Anthony PERARD <anthony.perard@vates.tech>
master commit: 01dfd25509
master date: 2025-06-02 10:31:21 +0200
x86_merge_dr6() is not a no-op when 0 is passed in; it will discard the
previously latched breakpoint bits.
The combination of do_debug()'s manual call to x86_merge_dr6() for external
debuggers, and pv_inject_DB() calling pv_inject_event(), results in two
x86_merge_dr6() calls.
Feed the same pending_dbg in the second time. This makes pv_inject_event()'s
update of dr6 effectively a no-op, retaining the correct breakpoint bits.
Fixes: db39fa4b27 ("x86/pv: Fix merging of new status bits into %dr6")
Reported-by: Manuel Bouyer <bouyer@antioche.eu.org>
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Tested-by: Manuel Bouyer <bouyer@antioche.eu.org>
Acked-by: Roger Pau Monné <roger.pau@citrix.com>
master commit: d965e2ee07
master date: 2025-05-27 12:21:38 +0100
Paging Writeable depends on EPT so must be disabled in non-EPT guests like the
other EPT dependent features. Otherwise, VMEntry fails with bad control
state.
Drop a piece of trailing whitepsace in context.
Fixes: ff10aa9d8f ("x86: Add Support for Paging-Write Feature")
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
master commit: d779e3f98c
master date: 2025-05-27 12:21:38 +0100
For once the message printed when a BAR overlaps with a non-hole regions is
not accurate on x86. While the BAR won't be mapped by the vPCI logic, it
is quite likely overlapping with a reserved region in the memory map, and
already mapped as by default all reserved regions are identity mapped in
the p2m. Avoid printing the warning message in modify_bars(), and instead
print a more lax message in the x86 implementation of pci_check_bar() to
note the current BAR position overlaps with non-hole region(s).
Secondly, when an overlap is detected the BAR 'enabled' field is not set,
hence other vPCI code that depends on it like vPCI MSI-X handling won't
function properly, as it sees the BAR as disabled, even when memory
decoding is enabled for the device and the BAR is likely mapped in the
p2m. Change the handling of BARs that overlap non-hole regions to instead
remove any overlapped regions from the rangeset, so the resulting ranges to
map just contain the hole regions. This requires introducing a new
pci_sanitize_bar_memory() that's implemented per-arch and sanitizes the
address range to add to the p2m.
For x86 pci_sanitize_bar_memory() removes any regions present in the host
memory map, for ARM this is currently left as a dummy handler to not change
existing behavior.
Ultimately the above changes should fix the vPCI MSI-X handlers not working
correctly when the BAR that contains the MSI-X table overlaps with a
non-hole region, as then the 'enabled' BAR bit won't be set and the MSI-X
traps won't handle accesses as expected.
Reported-by: Stefano Stabellini <stefano.stabellini@amd.com>
Fixes: 53d9133638 ('pci: do not disable memory decoding for devices')
Signed-off-by: Roger Pau Monné <roger.pau@citrix.com>
Tested-by: Victor M Lira <victorm.lira@amd.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Julien Grall <jgrall@amazon.com>
x86/vpci: fix off-by-one in pci_sanitize_bar_memory()
rangeset_remove_range() uses inclusive ranges, and hence the end of the
range should be calculated using PFN_DOWN(), not PFN_UP().
Fixes: 4acab25a93 ('x86/vpci: fix handling of BAR overlaps with non-hole regions')
Signed-off-by: Roger Pau Monné <roger.pau@citrix.com>
Acked-by: Andrew Cooper <andrew.cooper3@citrix.com>
master commit: 4acab25a93
master date: 2025-05-21 18:29:55 +0200
master commit: 7ab4b392b7
master date: 2025-05-22 16:17:12 +0200
The current emulation of wb{,no}invd is bogus for PV guests: it will only
flush the current pCPU cache, without taking into account pCPUs where the
vCPU had run previously. Resort to flushing the cache on all host pCPUs to
make it correct.
Fixes: 799fed0a7c ("Priv-op emulation in Xen, for RDMSR/WRMSR/WBINVD")
Signed-off-by: Roger Pau Monné <roger.pau@citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
[backport: no FLUSH_CACHE_WRITEBACK]
master commit: b0f8bf8605
master date: 2025-05-20 16:35:52 +0200
This is reported as a MISRA R16.3 (missing break) violation, but turns out to
be substantially more complicated than expected.
In commit a8fe4ec532 ("x86emul: support RDRAND/RDSEED"), the switch()
statement had a default case going to cannot_emulate, with both the case 6 and
case 7 labels being fully within #ifdef HAVE_GAS_RD{RAND,SEED}.
Therefore, when the toolchain didn't understand the RDRAND/RDSEED
instructions, attempts to emulate them suffered #UD. (In principle, this is a
problem as there's no interlock to prevent RDRAND/RDSEED being advertised to
the guest, but as instructions with only register encodings, they can only
legitimately be emulated when VM Introspection is in use.)
In commit 58f1bba440 ("x86emul: support RDPID"), case 7 was taken outside of
HAVE_GAS_RDSEED, meaning that emulating an RDSEED instruction no longer hit
the default case when the toolchain was too old.
Instead, it would fall out of the switch statement and be completed normally,
behaving as a NOP to the guest.
Retrofit a "return X86EMUL_UNIMPLEMENTED" in the case that the toolchain
doesn't know the RDRAND instruction, matching how RDRAND work.
Note that this has been fixed differently in Xen 4.21. Commit
05bf9f1f0f ("x86/emulate: Remove HAVE_AS_RDRAND and HAVE_AS_RDSEED") has
removed the problematic condition due to the toolchain baseline upgrade.
Fixes: 58f1bba440 ("x86emul: support RDPID")
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
master commit: ddec00769c
master date: 2025-05-16 21:34:19 +0100
If init_one_irq_desc() fails, ->arch.creator_domid won't be set to the
expected value, and hence the assertion may trigger. Limit it to just
the success case of that function call.
Fixes: 92d9101eab ("x86: allow stubdom access to irq created for msi")
Reported-by: Marek Marczykowski-Górecki <marmarek@invisiblethingslab.com>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
master commit: d7127e06f6
master date: 2025-05-14 11:00:00 +0200
The implementation of MMUEXT_FLUSH_CACHE is bogus, as it doesn't account to
flush the cache of any previous pCPU where the current vCPU might have run,
and hence is likely to not work as expected.
Fix this by resorting to use the same logic as MMUEXT_FLUSH_CACHE_GLOBAL,
which will be correct in all cases.
Adjust MMUEXT_FLUSH_CACHE_GLOBAL return code in case cache flush is not
permitted for the domain to use -EACCES instead of -EINVAL, as that's more
accurate and also matches the error code used by MMUEXT_FLUSH_CACHE.
Fixes: 8e90e37e6d ("Fix WBINVD by adding a new hypercall.")
Signed-off-by: Roger Pau Monné <roger.pau@citrix.com>
Acked-by: Jan Beulich <jbeulich@suse.com>
master commit: bbaea9878b
master date: 2025-05-13 19:43:48 +0200
Error paths of cpufreq_statistic_init() correctly free the base
structure pointer, but the per-CPU variable would still hold it, mis-
guiding e.g. cpufreq_statistic_update(). Defer installing of the pointer
there until the structure was fully populated.
Fixes: 755af07edb ("x86/cpufreq: don't use static array for large per-CPU data structures")
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Andrew Cooper <andrew.cooper3@citrix.com>
master commit: a1ce987411
master date: 2025-04-30 08:47:49 +0200
16-bit quantities, no matter whether expressed as uint16_t or as
bitfield, will be promoted to plain int before doing any arithmetic on
them. Shifting such values by 16 will therefore shift into the sign bit,
which is UB if that bit becomes set. To account for all reads and all
writes accessing opposite members of the same union, introduce yet more
local variables to reduce the shift counts to 12.
Fixes: be55ed744e ("x86emul: support FLDENV and FRSTOR")
Reported-by: Fabian Specht <f.specht@tum.de>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Jason Andryuk <jason.andryuk@amd.com>
master commit: d00fc019b6
master date: 2025-04-30 08:46:21 +0200
Building Xen with Clang-17 yields the following warning:
ld: warning: orphan section `.debug_str_offsets' from `prelink.o' being placed in section `.debug_str_offsets'
ld: ./.xen.efi.0xffff82d040000000.0:/4: section below image base
ld: ./.xen.efi.0xffff82d040000000.1:/4: section below image base
ld: warning: orphan section `.debug_str_offsets' from `prelink.o' being placed in section `.debug_str_offsets'
ld: xen.efi:/4: section below image base
Set the alignment to 4 as it holds 4-byte values, despite the fact that Clang
appears to only use 1.
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
master commit: a88b99300a
master date: 2025-04-28 19:46:46 +0100
It is easier to express feature word 17 in terms of word 16 + [32, 64) as
that's how the layout is given in documentation.
This is part of XSA-469 / CVE-2024-28956
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
(cherry picked from commit f6042f38e6)
The Indirect Target Selection speculative vulnerability means that indirect
branches (including RETs) are unsafe when in the first half of a cacheline.
In order to mitigate this, build with return thunks and arrange for
__x86_return_thunk to be (mis)aligned in the same manner as
__x86_indirect_thunk_* so the RET instruction is placed in a safe location.
place_ret() needs to conditionally emit JMP __x86_return_thunk instead of RET.
This is part of XSA-469 / CVE-2024-28956
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
(cherry picked from commit afcb4a06c7)
The Indirect Target Selection speculative vulnerability means that indirect
branches (including RETs) are unsafe when in the first half of a cacheline.
This means it's not safe for logic using the stubs to write raw 0xc3's.
Introduce place_ret() which, for now, writes a raw 0xc3 but will contain
additional logic when return thunks are in use.
stub_selftest() doesn't strictly need to be converted as they only run on
boot, but doing so gets us a partial test of place_ret() too.
No functional change.
This is part of XSA-469 / CVE-2024-28956
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
(cherry picked from commit 2eb1132f79)
The Indirect Target Selection speculative vulnerability means that indirect
branches (including RETs) are unsafe when in the first half of a cacheline.
clear_bhb_loops() has a precise layout of branches. The alignment for
performance cause the RETs to always be in an unsafe position, and converting
those to return thunks changes the branching pattern. While such a conversion
is believed to be safe, clear_bhb_loops() is also a performance-relevant
fastpath, so (mis)align the RETs to be in a safe position.
No functional change.
This is part of XSA-469 / CVE-2024-28956
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
(cherry picked from commit c0db07031a)
The Indirect Target Selection speculative vulnerability means that indirect
branches (including RETs) are unsafe when in the first half of a cacheline.
Arrange for __x86_indirect_thunk_* to always be in the second half.
This is part of XSA-469 / CVE-2024-28956
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
(cherry picked from commit d293cc9da9)
In order to protect against ITS, Xen needs to start using return thunks.
Therefore the advice in XSA-466 becomes relevant, and the hypercall_page needs
to be removed.
Implement early_hypercall(), with infrastructure to figure out the correct
instruction on first use. Use ALTERNATIVE()s to result in inline hypercalls,
including the ALT_NOT() form so we only need a single synthetic feature bit.
No overall change.
This is part of XSA-469 / CVE-2024-28956
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
(cherry picked from commit ef30ffe0a0)
Use the top bit of a->cpuid to express inverted polarity. This requires
stripping the top bit back out when performing the sanity checks.
Despite only being used once, create a replace boolean to express the decision
more clearly in _apply_alternatives().
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
(cherry picked from commit 328ed39c59)
Use the new test-artifacts which provide rootfs.cpio.gz rather than
initrd.tar.gz. rootfs.cpio.gz also has all the necessary top-level
directories, and includes the rc_verbose setting, so these modifications can
be dropped.
Having that, do not repack the whole initrd, but only pack modified
files and rely on Linux handling of concatenated archives.
This allows packing just test-related files (which includes the whole
toolstack), instead of the whole initrd.
For xilinx-smoke-dom0-x86_64.sh, this involves instructing grub not to unzip
the archive, as doing so corrupts it.
Signed-off-by: Marek Marczykowski-Górecki <marmarek@invisiblethingslab.com>
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Anthony PERARD <anthony.perard@vates.tech>
(cherry picked from commit 786f323809)
Right now, we have initrd.cpio.gz as domU, and initrd.tar.gz as the base for
dom0.
Rename initrd.cpio.gz to domU-rootfs.cpio.gz, and xen-rootfs.cpio.gz to
dom0-rootfs.cpio.gz to make it clearer which is which. Rename the VM from
test to domU.
No functional change.
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Anthony PERARD <anthony.perard@vates.tech>
(cherry picked from commit 9caca83404)
Upcoming changes won't unpack original rootfs anymore, so sed on
existing file cannot be used. Override the whole file instead.
No functional change intended.
Signed-off-by: Marek Marczykowski-Górecki <marmarek@invisiblethingslab.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
(cherry picked from commit f401eae6ec)
It appears as sometimes it takes more time for Xen even start booting,
mostly due to firmware and fetching large boot files by grub. In some
jobs the current timeout is pretty close to the actual time needed, and
sometimes (rarely for now) test fails due to timeout expiring in the
middle of dom0 booting. This will be happening more often if the
initramfs will grow (and with more complex tests).
This has been observed on some dom0pvh-hvm jobs, at least on runners hw3
and hw11.
Switch to using expect (console.exp) for more robust test output
handling. This allows waiting separately for Xen starting to boot and
then for the test to complete. For now, set both of those to 120s, which
pessimistically bumps timeout for the whole test to 240s (from 120s).
Add S3 handling to console.exp via SUSPEND_MSG + WAKEUP_CMD.
Signed-off-by: Marek Marczykowski-Górecki <marmarek@invisiblethingslab.com>
(cherry picked from commit d604d774a2)
Expect normally discards initial part of its buffer after matching the
patter, before looking for the next one. If both PASSED and LOG_MSG
happen to be in the buffer at the same time, depending on their order,
only one will be matched and the waiting for the other will timeout.
Example expect -d output of this happening (parts eclipsed for brevity):
expect: does "\r\r\r\nWelcome to Alpine Linux 3.18\r\r\r\n...\r\r\r\r\n(domU) + echo 'pci test passed'\r\r\r\r\n(domU) pci test passed\r\r\r\r..." (spawn_id exp4) match regular expression "pci test passed"? Gate "pci test passed"? gate=yes re=yes
...
Gate keeper glob pattern for '\nWelcome to Alpine Linux' is '
Welcome to Alpine Linux'. Activating booster.
expect: does "'\r\r\r\r\n(domU) pci test passed\r\r\r\r\n(domU) [ ok ]\r\r\r\r\n(domU) [ ok ]\r\r\r\r\n(domU) \r\r\r\r\r\n(domU) domU Welcome to Alpine Linux 3.18\r\r\r\r\n(domU) \rKernel 6.6.56 on an x86_64 (/dev/hvc0)\r\r\r\r\n(domU) \r\r\r\r\r\n" (spawn_id exp4) match regular expression "\nWelcome to Alpine Linux"? Gate "\nWelcome to Alpine Linux"? gate=no
Fix this by using -notransfer flag to keep matched part in the buffer.
Signed-off-by: Marek Marczykowski-Górecki <marmarek@invisiblethingslab.com>
Acked-by: Andrew Cooper <andrew.cooper3@citix.com>
(cherry picked from commit 1e12cbd6af)
CI: fix waiting for final test message (again)
The previous attempt has correct diagnosis, but added -notransfer flag
in a wrong place - it should be used in the first (outer) match out of
two, not the second (inner) one.
Fixes: 1e12cbd6af ("CI: fix waiting for final test message")
Signed-off-by: Marek Marczykowski-Górecki <marmarek@invisiblethingslab.com>
Acked-by: Andrew Cooper <andrew.cooper3@citrix.com>
(cherry picked from commit 79de6507c5)
Add additional stage in console output parsing - wait for first message
from Xen. The message is defined via BOOT_MSG variable. This has two
effects:
- distinguishes failing Xen to load at all from later test failures
- resets timeout when Xen starts loading
The latter is especially relevant for hardware tests where firmware +
network boot may take some time before Xen starts booting. The two-stage
timeout is more robust solution than increasing the overall timeout.
The issue has been observed on some dom0pvh-hvm jobs, at least on
runners hw3 and hw11. This patch is a first stage before qubes-x86-64.sh
is switched to use expect in the next stage.
While at it, consistently use 'expect -re' for all matches. This
especially allows matching newlines ("\n"), which will become relevant
in the next patch. And document variables used in console.exp.
Signed-off-by: Marek Marczykowski-Górecki <marmarek@invisiblethingslab.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
(cherry picked from commit 7c96a27e37)
This should have been dropped too, as the authoritative copy lives in another
repo now.
Fixes: babe11b46c ("CI: Drop alpine-3.18-rootfs-export and use test-artefacts")
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Marek Marczykowski-Górecki <marmarek@invisiblethingslab.com>
(cherry picked from commit 424d20afae)
Linux 6.6.56 was already added to test-artifacts for the argo testing, and
this removes one moving part while cleaning things up.
Drop the associated export job, and dockerfile.
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Marek Marczykowski-Górecki <marmarek@invisiblethingslab.com>
(cherry picked from commit f858837664)
Change how PCI device lookup is done to handle also USB devices, in
which case get the USB controller. Instead of taking basename of the
'device' symlink, resolve the full path (example:
/sys/devices/pci0000:00/0000:00:09.0/usb4/4-7/4-7:1.0) and take the
first part after pci0000:00. Theoretically it could be a bridge, but VM
has flat PCI topology.
Signed-off-by: Marek Marczykowski-Górecki <marmarek@invisiblethingslab.com>
Reviewed-by: Stefano Stabellini <sstabellini@kernel.org>
(cherry picked from commit 930279d96c)
Andrew Cooper