Our __smp_store_release() and __smp_load_acquire() macros use inline
assembly, which is opaque to kasan. This means that kasan can't catch
erroneous use of these.
This patch adds kasan instrumentation to both.
Cc: Catalin Marinas <catalin.marinas@arm.com>
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
[will: consistently use *p as argument to sizeof]
Signed-off-by: Will Deacon <will.deacon@arm.com>
This change uses the original virt_to_page() (the one with __pa()) to
check the given virtual address if CONFIG_DEBUG_VIRTUAL=y.
Recently, I worked on a bug: a driver passes a symbol address to
dma_map_single() and the virt_to_page() (called by dma_map_single())
does not work for non-linear addresses after commit 9f2875912d
("arm64: mm: restrict virt_to_page() to the linear mapping").
I tried to trap the bug by enabling CONFIG_DEBUG_VIRTUAL but it
did not work - bacause the commit removes the __pa() from
virt_to_page() but CONFIG_DEBUG_VIRTUAL checks the virtual address
in __pa()/__virt_to_phys().
A simple solution is to use the original virt_to_page()
(the one with__pa()) if CONFIG_DEBUG_VIRTUAL=y.
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Signed-off-by: Miles Chen <miles.chen@mediatek.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
Advertise ARM64_HAS_DCPODP when both DC CVAP and DC CVADP are supported.
Even though we don't use this feature now, we provide it for consistency
with DCPOP and anticipate it being used in the future.
Signed-off-by: Andrew Murray <andrew.murray@arm.com>
Reviewed-by: Dave Martin <Dave.Martin@arm.com>
Reviewed-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
ARMv8.5 builds upon the ARMv8.2 DC CVAP instruction by introducing a DC
CVADP instruction which cleans the data cache to the point of deep
persistence. Let's expose this support via the arm64 ELF hwcaps.
Signed-off-by: Andrew Murray <andrew.murray@arm.com>
Reviewed-by: Dave Martin <Dave.Martin@arm.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
The ARMv8.5 DC CVADP instruction may be trapped to EL1 via
SCTLR_EL1.UCI therefore let's provide a handler for it.
Just like the CVAP instruction we use a 'sys' instruction instead of
the 'dc' alias to avoid build issues with older toolchains.
Signed-off-by: Andrew Murray <andrew.murray@arm.com>
Reviewed-by: Mark Rutland <mark.rutland@arm.com>
Reviewed-by: Dave Martin <Dave.Martin@arm.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
The introduction of AT_HWCAP2 introduced accessors which ensure that
hwcap features are set and tested appropriately.
Let's now mandate access to elf_hwcap via these accessors by making
elf_hwcap static within cpufeature.c.
Signed-off-by: Andrew Murray <andrew.murray@arm.com>
Reviewed-by: Dave Martin <Dave.Martin@arm.com>
Reviewed-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
As we will exhaust the first 32 bits of AT_HWCAP let's start
exposing AT_HWCAP2 to userspace to give us up to 64 caps.
Whilst it's possible to use the remaining 32 bits of AT_HWCAP, we
prefer to expand into AT_HWCAP2 in order to provide a consistent
view to userspace between ILP32 and LP64. However internal to the
kernel we prefer to continue to use the full space of elf_hwcap.
To reduce complexity and allow for future expansion, we now
represent hwcaps in the kernel as ordinals and use a
KERNEL_HWCAP_ prefix. This allows us to support automatic feature
based module loading for all our hwcaps.
We introduce cpu_set_feature to set hwcaps which complements the
existing cpu_have_feature helper. These helpers allow us to clean
up existing direct uses of elf_hwcap and reduce any future effort
required to move beyond 64 caps.
For convenience we also introduce cpu_{have,set}_named_feature which
makes use of the cpu_feature macro to allow providing a hwcap name
without a {KERNEL_}HWCAP_ prefix.
Signed-off-by: Andrew Murray <andrew.murray@arm.com>
[will: use const_ilog2() and tweak documentation]
Signed-off-by: Will Deacon <will.deacon@arm.com>
Add the io_uring and pidfd_send_signal system calls to all architectures.
These system calls are designed to handle both native and compat tasks,
so all entries are the same across architectures, only arm-compat and
the generic tale still use an old format.
Acked-by: Michael Ellerman <mpe@ellerman.id.au> (powerpc)
Acked-by: Heiko Carstens <heiko.carstens@de.ibm.com> (s390)
Acked-by: Geert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Add regs_get_argument() which returns N th argument of the function
call. On arm64, it supports up to 8th argument.
Note that this chooses most probably assignment, in some case
it can be incorrect (e.g. passing data structure or floating
point etc.)
This enables ftrace kprobe events to access kernel function
arguments via $argN syntax.
Signed-off-by: Masami Hiramatsu <mhiramat@kernel.org>
[will: tidied up the comment a bit]
Signed-off-by: Will Deacon <will.deacon@arm.com>
Rather embarrassingly, our futex() FUTEX_WAKE_OP implementation doesn't
explicitly set the return value on the non-faulting path and instead
leaves it holding the result of the underlying atomic operation. This
means that any FUTEX_WAKE_OP atomic operation which computes a non-zero
value will be reported as having failed. Regrettably, I wrote the buggy
code back in 2011 and it was upstreamed as part of the initial arm64
support in 2012.
The reasons we appear to get away with this are:
1. FUTEX_WAKE_OP is rarely used and therefore doesn't appear to get
exercised by futex() test applications
2. If the result of the atomic operation is zero, the system call
behaves correctly
3. Prior to version 2.25, the only operation used by GLIBC set the
futex to zero, and therefore worked as expected. From 2.25 onwards,
FUTEX_WAKE_OP is not used by GLIBC at all.
Fix the implementation by ensuring that the return value is either 0
to indicate that the atomic operation completed successfully, or -EFAULT
if we encountered a fault when accessing the user mapping.
Cc: <stable@kernel.org>
Fixes: 6170a97460 ("arm64: Atomic operations")
Signed-off-by: Will Deacon <will.deacon@arm.com>
Currently, compat tasks running on arm64 can allocate memory up to
TASK_SIZE_32 (UL(0x100000000)).
This means that mmap() allocations, if we treat them as returning an
array, are not compliant with the sections 6.5.8 of the C standard
(C99) which states that: "If the expression P points to an element of
an array object and the expression Q points to the last element of the
same array object, the pointer expression Q+1 compares greater than P".
Redefine TASK_SIZE_32 to address the issue.
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Jann Horn <jannh@google.com>
Cc: <stable@vger.kernel.org>
Reported-by: Jann Horn <jannh@google.com>
Signed-off-by: Vincenzo Frascino <vincenzo.frascino@arm.com>
[will: fixed typo in comment]
Signed-off-by: Will Deacon <will.deacon@arm.com>
Switch from per mm_struct to per pmd page table lock by enabling
ARCH_ENABLE_SPLIT_PMD_PTLOCK. This provides better granularity for
large system.
I'm not sure if there is contention on mm->page_table_lock. Given
the option comes at no cost (apart from initializing more spin
locks), why not enable it now.
We only do so when pmd is not folded, so we don't mistakenly call
pgtable_pmd_page_ctor() on pud or p4d in pgd_pgtable_alloc().
Signed-off-by: Yu Zhao <yuzhao@google.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
ARM64 standard pgtable functions are going to use pgtable_page_[ctor|dtor]
or pgtable_pmd_page_[ctor|dtor] constructs. At present KVM guest stage-2
PUD|PMD|PTE level page tabe pages are allocated with __get_free_page()
via mmu_memory_cache_alloc() but released with standard pud|pmd_free() or
pte_free_kernel(). These will fail once they start calling into pgtable_
[pmd]_page_dtor() for pages which never originally went through respective
constructor functions. Hence convert all stage-2 page table page release
functions to call buddy directly while freeing pages.
Reviewed-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Acked-by: Yu Zhao <yuzhao@google.com>
Acked-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
kprobes and uprobes reserve some BRK immediates for installing their
probes. Define these along with the other reservations in brk-imm.h
and rename the ESR definitions to be consistent with the others that we
already have.
Reviewed-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
Kprobes bypasses our debug hook registration code so that it doesn't
get tangled up with recursive debug exceptions from things like lockdep:
http://lists.infradead.org/pipermail/linux-arm-kernel/2015-February/324385.html
However, since then, (a) the hook list has become RCU protected and (b)
the kprobes hooks were found not to filter out exceptions from userspace
correctly. On top of that, the step handler is invoked directly from
single_step_handler(), which *does* use the debug hook list, so it's
clearly not the end of the world.
For now, have kprobes use the debug hook registration API like everybody
else. We can revisit this in the future if this is found to limit
coverage significantly.
Reviewed-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
Mixing kernel and user debug hooks together is highly error-prone as it
relies on all of the hooks to figure out whether the exception came from
kernel or user, and then to act accordingly.
Make our debug hook code a little more robust by maintaining separate
hook lists for user and kernel, with separate registration functions
to force callers to be explicit about the exception levels that they
care about.
Reviewed-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
The ftrace trampoline code (which deals with modules loaded out of
BL range of the core kernel) uses plt_entries_equal() to check whether
the per-module trampoline equals a zero buffer, to decide whether the
trampoline has already been initialized.
This triggers a BUG() in the opcode manipulation code, since we end
up checking the ADRP offset of a 0x0 opcode, which is not an ADRP
instruction.
So instead, add a helper to check whether a PLT is initialized, and
call that from the frace code.
Cc: <stable@vger.kernel.org> # v5.0
Fixes: bdb85cd1d2 ("arm64/module: switch to ADRP/ADD sequences for PLT entries")
Acked-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Will Deacon <will.deacon@arm.com>
arm64 includes asm-generic/io.h, which provides a dummy definition of
mmiowb() if one isn't already provided by the architecture.
Remove the useless definition.
Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Will Deacon <will.deacon@arm.com>
Following assembly code is not trivial; make it slightly easier to read by
replacing some of the magic numbers with the defines which are already
present in sysreg.h.
Reviewed-by: Dave Martin <Dave.Martin@arm.com>
Signed-off-by: Alexandru Elisei <alexandru.elisei@arm.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
As the generic rwsem-xadd code is using the appropriate acquire and
release versions of the atomic operations, the arch specific rwsem.h
files will not be that much faster than the generic code as long as the
atomic functions are properly implemented. So we can remove those arch
specific rwsem.h and stop building asm/rwsem.h to reduce maintenance
effort.
Currently, only x86, alpha and ia64 have implemented architecture
specific fast paths. I don't have access to alpha and ia64 systems for
testing, but they are legacy systems that are not likely to be updated
to the latest kernel anyway.
By using a rwsem microbenchmark, the total locking rates on a 4-socket
56-core 112-thread x86-64 system before and after the patch were as
follows (mixed means equal # of read and write locks):
Before Patch After Patch
# of Threads wlock rlock mixed wlock rlock mixed
------------ ----- ----- ----- ----- ----- -----
1 29,201 30,143 29,458 28,615 30,172 29,201
2 6,807 13,299 1,171 7,725 15,025 1,804
4 6,504 12,755 1,520 7,127 14,286 1,345
8 6,762 13,412 764 6,826 13,652 726
16 6,693 15,408 662 6,599 15,938 626
32 6,145 15,286 496 5,549 15,487 511
64 5,812 15,495 60 5,858 15,572 60
There were some run-to-run variations for the multi-thread tests. For
x86-64, using the generic C code fast path seems to be a little bit
faster than the assembly version with low lock contention. Looking at
the assembly version of the fast paths, there are assembly to/from C
code wrappers that save and restore all the callee-clobbered registers
(7 registers on x86-64). The assembly generated from the generic C
code doesn't need to do that. That may explain the slight performance
gain here.
The generic asm rwsem.h can also be merged into kernel/locking/rwsem.h
with no code change as no other code other than those under
kernel/locking needs to access the internal rwsem macros and functions.
Signed-off-by: Waiman Long <longman@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Davidlohr Bueso <dave@stgolabs.net>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Tim Chen <tim.c.chen@linux.intel.com>
Cc: Will Deacon <will.deacon@arm.com>
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-c6x-dev@linux-c6x.org
Cc: linux-m68k@lists.linux-m68k.org
Cc: linux-riscv@lists.infradead.org
Cc: linux-um@lists.infradead.org
Cc: linux-xtensa@linux-xtensa.org
Cc: linuxppc-dev@lists.ozlabs.org
Cc: nios2-dev@lists.rocketboards.org
Cc: openrisc@lists.librecores.org
Cc: uclinux-h8-devel@lists.sourceforge.jp
Link: https://lkml.kernel.org/r/20190322143008.21313-2-longman@redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Provide a generic tlb_flush() implementation that relies on
flush_tlb_range(). This is a little awkward because flush_tlb_range()
assumes a VMA for range invalidation, but we no longer have one.
Audit of all flush_tlb_range() implementations shows only vma->vm_mm
and vma->vm_flags are used, and of the latter only VM_EXEC (I-TLB
invalidates) and VM_HUGETLB (large TLB invalidate) are used.
Therefore, track VM_EXEC and VM_HUGETLB in two more bits, and create a
'fake' VMA.
This allows architectures that have a reasonably efficient
flush_tlb_range() to not require any additional effort.
No change in behavior intended.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Will Deacon <will.deacon@arm.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Nick Piggin <npiggin@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@surriel.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Now that all the pieces are in place, this patch offers a new flag
KVM_ARM_VCPU_SVE that userspace can pass to KVM_ARM_VCPU_INIT to
turn on SVE for the guest, on a per-vcpu basis.
As part of this, support for initialisation and reset of the SVE
vector length set and registers is added in the appropriate places,
as well as finally setting the KVM_ARM64_GUEST_HAS_SVE vcpu flag,
to turn on the SVE support code.
Allocation of the SVE register storage in vcpu->arch.sve_state is
deferred until the SVE configuration is finalized, by which time
the size of the registers is known.
Setting the vector lengths supported by the vcpu is considered
configuration of the emulated hardware rather than runtime
configuration, so no support is offered for changing the vector
lengths available to an existing vcpu across reset.
Signed-off-by: Dave Martin <Dave.Martin@arm.com>
Reviewed-by: Julien Thierry <julien.thierry@arm.com>
Tested-by: zhang.lei <zhang.lei@jp.fujitsu.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
This patch adds a new pseudo-register KVM_REG_ARM64_SVE_VLS to
allow userspace to set and query the set of vector lengths visible
to the guest.
In the future, multiple register slices per SVE register may be
visible through the ioctl interface. Once the set of slices has
been determined we would not be able to allow the vector length set
to be changed any more, in order to avoid userspace seeing
inconsistent sets of registers. For this reason, this patch adds
support for explicit finalization of the SVE configuration via the
KVM_ARM_VCPU_FINALIZE ioctl.
Finalization is the proper place to allocate the SVE register state
storage in vcpu->arch.sve_state, so this patch adds that as
appropriate. The data is freed via kvm_arch_vcpu_uninit(), which
was previously a no-op on arm64.
To simplify the logic for determining what vector lengths can be
supported, some code is added to KVM init to work this out, in the
kvm_arm_init_arch_resources() hook.
The KVM_REG_ARM64_SVE_VLS pseudo-register is not exposed yet.
Subsequent patches will allow SVE to be turned on for guest vcpus,
making it visible.
Signed-off-by: Dave Martin <Dave.Martin@arm.com>
Reviewed-by: Julien Thierry <julien.thierry@arm.com>
Tested-by: zhang.lei <zhang.lei@jp.fujitsu.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Some aspects of vcpu configuration may be too complex to be
completed inside KVM_ARM_VCPU_INIT. Thus, there may be a
requirement for userspace to do some additional configuration
before various other ioctls will work in a consistent way.
In particular this will be the case for SVE, where userspace will
need to negotiate the set of vector lengths to be made available to
the guest before the vcpu becomes fully usable.
In order to provide an explicit way for userspace to confirm that
it has finished setting up a particular vcpu feature, this patch
adds a new ioctl KVM_ARM_VCPU_FINALIZE.
When userspace has opted into a feature that requires finalization,
typically by means of a feature flag passed to KVM_ARM_VCPU_INIT, a
matching call to KVM_ARM_VCPU_FINALIZE is now required before
KVM_RUN or KVM_GET_REG_LIST is allowed. Individual features may
impose additional restrictions where appropriate.
No existing vcpu features are affected by this, so current
userspace implementations will continue to work exactly as before,
with no need to issue KVM_ARM_VCPU_FINALIZE.
As implemented in this patch, KVM_ARM_VCPU_FINALIZE is currently a
placeholder: no finalizable features exist yet, so ioctl is not
required and will always yield EINVAL. Subsequent patches will add
the finalization logic to make use of this ioctl for SVE.
No functional change for existing userspace.
Signed-off-by: Dave Martin <Dave.Martin@arm.com>
Reviewed-by: Julien Thierry <julien.thierry@arm.com>
Tested-by: zhang.lei <zhang.lei@jp.fujitsu.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
This patch adds a kvm_arm_init_arch_resources() hook to perform
subarch-specific initialisation when starting up KVM.
This will be used in a subsequent patch for global SVE-related
setup on arm64.
No functional change.
Signed-off-by: Dave Martin <Dave.Martin@arm.com>
Reviewed-by: Julien Thierry <julien.thierry@arm.com>
Tested-by: zhang.lei <zhang.lei@jp.fujitsu.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
KVM will need to interrogate the set of SVE vector lengths
available on the system.
This patch exposes the relevant bits to the kernel, along with a
sve_vq_available() helper to check whether a particular vector
length is supported.
__vq_to_bit() and __bit_to_vq() are not intended for use outside
these functions: now that these are exposed outside fpsimd.c, they
are prefixed with __ in order to provide an extra hint that they
are not intended for general-purpose use.
Signed-off-by: Dave Martin <Dave.Martin@arm.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Tested-by: zhang.lei <zhang.lei@jp.fujitsu.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
This patch adds the following registers for access via the
KVM_{GET,SET}_ONE_REG interface:
* KVM_REG_ARM64_SVE_ZREG(n, i) (n = 0..31) (in 2048-bit slices)
* KVM_REG_ARM64_SVE_PREG(n, i) (n = 0..15) (in 256-bit slices)
* KVM_REG_ARM64_SVE_FFR(i) (in 256-bit slices)
In order to adapt gracefully to future architectural extensions,
the registers are logically divided up into slices as noted above:
the i parameter denotes the slice index.
This allows us to reserve space in the ABI for future expansion of
these registers. However, as of today the architecture does not
permit registers to be larger than a single slice, so no code is
needed in the kernel to expose additional slices, for now. The
code can be extended later as needed to expose them up to a maximum
of 32 slices (as carved out in the architecture itself) if they
really exist someday.
The registers are only visible for vcpus that have SVE enabled.
They are not enumerated by KVM_GET_REG_LIST on vcpus that do not
have SVE.
Accesses to the FPSIMD registers via KVM_REG_ARM_CORE is not
allowed for SVE-enabled vcpus: SVE-aware userspace can use the
KVM_REG_ARM64_SVE_ZREG() interface instead to access the same
register state. This avoids some complex and pointless emulation
in the kernel to convert between the two views of these aliased
registers.
Signed-off-by: Dave Martin <Dave.Martin@arm.com>
Reviewed-by: Julien Thierry <julien.thierry@arm.com>
Tested-by: zhang.lei <zhang.lei@jp.fujitsu.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
In order to give each vcpu its own view of the SVE registers, this
patch adds context storage via a new sve_state pointer in struct
vcpu_arch. An additional member sve_max_vl is also added for each
vcpu, to determine the maximum vector length visible to the guest
and thus the value to be configured in ZCR_EL2.LEN while the vcpu
is active. This also determines the layout and size of the storage
in sve_state, which is read and written by the same backend
functions that are used for context-switching the SVE state for
host tasks.
On SVE-enabled vcpus, SVE access traps are now handled by switching
in the vcpu's SVE context and disabling the trap before returning
to the guest. On other vcpus, the trap is not handled and an exit
back to the host occurs, where the handle_sve() fallback path
reflects an undefined instruction exception back to the guest,
consistently with the behaviour of non-SVE-capable hardware (as was
done unconditionally prior to this patch).
No SVE handling is added on non-VHE-only paths, since VHE is an
architectural and Kconfig prerequisite of SVE.
Signed-off-by: Dave Martin <Dave.Martin@arm.com>
Reviewed-by: Julien Thierry <julien.thierry@arm.com>
Tested-by: zhang.lei <zhang.lei@jp.fujitsu.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
This patch adds the necessary support for context switching ZCR_EL1
for each vcpu.
ZCR_EL1 is trapped alongside the FPSIMD/SVE registers, so it makes
sense for it to be handled as part of the guest FPSIMD/SVE context
for context switch purposes instead of handling it as a general
system register. This means that it can be switched in lazily at
the appropriate time. No effort is made to track host context for
this register, since SVE requires VHE: thus the hosts's value for
this register lives permanently in ZCR_EL2 and does not alias the
guest's value at any time.
The Hyp switch and fpsimd context handling code is extended
appropriately.
Accessors are added in sys_regs.c to expose the SVE system
registers and ID register fields. Because these need to be
conditionally visible based on the guest configuration, they are
implemented separately for now rather than by use of the generic
system register helpers. This may be abstracted better later on
when/if there are more features requiring this model.
ID_AA64ZFR0_EL1 is RO-RAZ for MRS/MSR when SVE is disabled for the
guest, but for compatibility with non-SVE aware KVM implementations
the register should not be enumerated at all for KVM_GET_REG_LIST
in this case. For consistency we also reject ioctl access to the
register. This ensures that a non-SVE-enabled guest looks the same
to userspace, irrespective of whether the kernel KVM implementation
supports SVE.
Signed-off-by: Dave Martin <Dave.Martin@arm.com>
Reviewed-by: Julien Thierry <julien.thierry@arm.com>
Tested-by: zhang.lei <zhang.lei@jp.fujitsu.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Since SVE will be enabled or disabled on a per-vcpu basis, a flag
is needed in order to track which vcpus have it enabled.
This patch adds a suitable flag and a helper for checking it.
Signed-off-by: Dave Martin <Dave.Martin@arm.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Tested-by: zhang.lei <zhang.lei@jp.fujitsu.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
The current FPSIMD/SVE context handling support for non-task (i.e.,
KVM vcpu) contexts does not take SVE into account. This means that
only task contexts can safely use SVE at present.
In preparation for enabling KVM guests to use SVE, it is necessary
to keep track of SVE state for non-task contexts too.
This patch adds the necessary support, removing assumptions from
the context switch code about the location of the SVE context
storage.
When binding a vcpu context, its vector length is arbitrarily
specified as SVE_VL_MIN for now. In any case, because TIF_SVE is
presently cleared at vcpu context bind time, the specified vector
length will not be used for anything yet. In later patches TIF_SVE
will be set here as appropriate, and the appropriate maximum vector
length for the vcpu will be passed when binding.
Signed-off-by: Dave Martin <Dave.Martin@arm.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Julien Grall <julien.grall@arm.com>
Tested-by: zhang.lei <zhang.lei@jp.fujitsu.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Due to the way the effective SVE vector length is controlled and
trapped at different exception levels, certain mismatches in the
sets of vector lengths supported by different physical CPUs in the
system may prevent straightforward virtualisation of SVE at parity
with the host.
This patch analyses the extent to which SVE can be virtualised
safely without interfering with migration of vcpus between physical
CPUs, and rejects late secondary CPUs that would erode the
situation further.
It is left up to KVM to decide what to do with this information.
Signed-off-by: Dave Martin <Dave.Martin@arm.com>
Reviewed-by: Julien Thierry <julien.thierry@arm.com>
Tested-by: zhang.lei <zhang.lei@jp.fujitsu.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
kvm_host.h uses some declarations from other headers that are
currently included by accident, without an explicit #include.
This patch adds a few #includes that are clearly missing. Although
the header builds without them today, this should help to avoid
future surprises.
Signed-off-by: Dave Martin <Dave.Martin@arm.com>
Acked-by: Mark Rutland <mark.rutland@arm.com>
Tested-by: zhang.lei <zhang.lei@jp.fujitsu.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
KVM/ARM fixes for 5.1
- Fix THP handling in the presence of pre-existing PTEs
- Honor request for PTE mappings even when THPs are available
- GICv4 performance improvement
- Take the srcu lock when writing to guest-controlled ITS data structures
- Reset the virtual PMU in preemptible context
- Various cleanups
When halting a guest, QEMU flushes the virtual ITS caches, which
amounts to writing to the various tables that the guest has allocated.
When doing this, we fail to take the srcu lock, and the kernel
shouts loudly if running a lockdep kernel:
[ 69.680416] =============================
[ 69.680819] WARNING: suspicious RCU usage
[ 69.681526] 5.1.0-rc1-00008-g600025238f51-dirty #18 Not tainted
[ 69.682096] -----------------------------
[ 69.682501] ./include/linux/kvm_host.h:605 suspicious rcu_dereference_check() usage!
[ 69.683225]
[ 69.683225] other info that might help us debug this:
[ 69.683225]
[ 69.683975]
[ 69.683975] rcu_scheduler_active = 2, debug_locks = 1
[ 69.684598] 6 locks held by qemu-system-aar/4097:
[ 69.685059] #0: 0000000034196013 (&kvm->lock){+.+.}, at: vgic_its_set_attr+0x244/0x3a0
[ 69.686087] #1: 00000000f2ed935e (&its->its_lock){+.+.}, at: vgic_its_set_attr+0x250/0x3a0
[ 69.686919] #2: 000000005e71ea54 (&vcpu->mutex){+.+.}, at: lock_all_vcpus+0x64/0xd0
[ 69.687698] #3: 00000000c17e548d (&vcpu->mutex){+.+.}, at: lock_all_vcpus+0x64/0xd0
[ 69.688475] #4: 00000000ba386017 (&vcpu->mutex){+.+.}, at: lock_all_vcpus+0x64/0xd0
[ 69.689978] #5: 00000000c2c3c335 (&vcpu->mutex){+.+.}, at: lock_all_vcpus+0x64/0xd0
[ 69.690729]
[ 69.690729] stack backtrace:
[ 69.691151] CPU: 2 PID: 4097 Comm: qemu-system-aar Not tainted 5.1.0-rc1-00008-g600025238f51-dirty #18
[ 69.691984] Hardware name: rockchip evb_rk3399/evb_rk3399, BIOS 2019.04-rc3-00124-g2feec69fb1 03/15/2019
[ 69.692831] Call trace:
[ 69.694072] lockdep_rcu_suspicious+0xcc/0x110
[ 69.694490] gfn_to_memslot+0x174/0x190
[ 69.694853] kvm_write_guest+0x50/0xb0
[ 69.695209] vgic_its_save_tables_v0+0x248/0x330
[ 69.695639] vgic_its_set_attr+0x298/0x3a0
[ 69.696024] kvm_device_ioctl_attr+0x9c/0xd8
[ 69.696424] kvm_device_ioctl+0x8c/0xf8
[ 69.696788] do_vfs_ioctl+0xc8/0x960
[ 69.697128] ksys_ioctl+0x8c/0xa0
[ 69.697445] __arm64_sys_ioctl+0x28/0x38
[ 69.697817] el0_svc_common+0xd8/0x138
[ 69.698173] el0_svc_handler+0x38/0x78
[ 69.698528] el0_svc+0x8/0xc
The fix is to obviously take the srcu lock, just like we do on the
read side of things since bf308242ab. One wonders why this wasn't
fixed at the same time, but hey...
Fixes: bf308242ab ("KVM: arm/arm64: VGIC/ITS: protect kvm_read_guest() calls with SRCU lock")
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Fujitsu erratum 010001 applies to A64FX v0r0 and v1r0, and we try to
handle either by masking MIDR with MIDR_FUJITSU_ERRATUM_010001_MASK
before comparing it to MIDR_FUJITSU_ERRATUM_010001.
Unfortunately, MIDR_FUJITSU_ERRATUM_010001 is constructed incorrectly
using MIDR_VARIANT(), which is intended to extract the variant field
from MIDR_EL1, rather than generate the field in-place. This results in
MIDR_FUJITSU_ERRATUM_010001 being all-ones, and we only match A64FX
v0r0.
This patch uses MIDR_CPU_VAR_REV() to generate an in-place mask for the
variant field, ensuring the we match both v0r0 and v1r0.
Fixes: 3e32131abc ("arm64: Add workaround for Fujitsu A64FX erratum 010001")
Reported-by: "Okamoto, Takayuki" <tokamoto@jp.fujitsu.com>
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
[catalin.marinas@arm.com: fixed the patch author]
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Pull more Kbuild updates from Masahiro Yamada:
- add more Build-Depends to Debian source package
- prefix header search paths with $(srctree)/
- make modpost show verbose section mismatch warnings
- avoid hard-coded CROSS_COMPILE for h8300
- fix regression for Debian make-kpkg command
- add semantic patch to detect missing put_device()
- fix some warnings of 'make deb-pkg'
- optimize NOSTDINC_FLAGS evaluation
- add warnings about redundant generic-y
- clean up Makefiles and scripts
* tag 'kbuild-v5.1-2' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild:
kconfig: remove stale lxdialog/.gitignore
kbuild: force all architectures except um to include mandatory-y
kbuild: warn redundant generic-y
Revert "modsign: Abort modules_install when signing fails"
kbuild: Make NOSTDINC_FLAGS a simply expanded variable
kbuild: deb-pkg: avoid implicit effects
coccinelle: semantic code search for missing put_device()
kbuild: pkg: grep include/config/auto.conf instead of $KCONFIG_CONFIG
kbuild: deb-pkg: introduce is_enabled and if_enabled_echo to builddeb
kbuild: deb-pkg: add CONFIG_ prefix to kernel config options
kbuild: add workaround for Debian make-kpkg
kbuild: source include/config/auto.conf instead of ${KCONFIG_CONFIG}
unicore32: simplify linker script generation for decompressor
h8300: use cc-cross-prefix instead of hardcoding h8300-unknown-linux-
kbuild: move archive command to scripts/Makefile.lib
modpost: always show verbose warning for section mismatch
ia64: prefix header search path with $(srctree)/
libfdt: prefix header search paths with $(srctree)/
deb-pkg: generate correct build dependencies
Currently, every arch/*/include/uapi/asm/Kbuild explicitly includes
the common Kbuild.asm file. Factor out the duplicated include directives
to scripts/Makefile.asm-generic so that no architecture would opt out
of the mandatory-y mechanism.
um is not forced to include mandatory-y since it is a very exceptional
case which does not support UAPI.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Pull KVM updates from Paolo Bonzini:
"ARM:
- some cleanups
- direct physical timer assignment
- cache sanitization for 32-bit guests
s390:
- interrupt cleanup
- introduction of the Guest Information Block
- preparation for processor subfunctions in cpu models
PPC:
- bug fixes and improvements, especially related to machine checks
and protection keys
x86:
- many, many cleanups, including removing a bunch of MMU code for
unnecessary optimizations
- AVIC fixes
Generic:
- memcg accounting"
* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm: (147 commits)
kvm: vmx: fix formatting of a comment
KVM: doc: Document the life cycle of a VM and its resources
MAINTAINERS: Add KVM selftests to existing KVM entry
Revert "KVM/MMU: Flush tlb directly in the kvm_zap_gfn_range()"
KVM: PPC: Book3S: Add count cache flush parameters to kvmppc_get_cpu_char()
KVM: PPC: Fix compilation when KVM is not enabled
KVM: Minor cleanups for kvm_main.c
KVM: s390: add debug logging for cpu model subfunctions
KVM: s390: implement subfunction processor calls
arm64: KVM: Fix architecturally invalid reset value for FPEXC32_EL2
KVM: arm/arm64: Remove unused timer variable
KVM: PPC: Book3S: Improve KVM reference counting
KVM: PPC: Book3S HV: Fix build failure without IOMMU support
Revert "KVM: Eliminate extra function calls in kvm_get_dirty_log_protect()"
x86: kvmguest: use TSC clocksource if invariant TSC is exposed
KVM: Never start grow vCPU halt_poll_ns from value below halt_poll_ns_grow_start
KVM: Expose the initial start value in grow_halt_poll_ns() as a module parameter
KVM: grow_halt_poll_ns() should never shrink vCPU halt_poll_ns
KVM: x86/mmu: Consolidate kvm_mmu_zap_all() and kvm_mmu_zap_mmio_sptes()
KVM: x86/mmu: WARN if zapping a MMIO spte results in zapping children
...
Pull DMA mapping updates from Christoph Hellwig:
- add debugfs support for dumping dma-debug information (Corentin
Labbe)
- Kconfig cleanups (Andy Shevchenko and me)
- debugfs cleanups (Greg Kroah-Hartman)
- improve dma_map_resource and use it in the media code
- arch_setup_dma_ops / arch_teardown_dma_ops cleanups
- various small cleanups and improvements for the per-device coherent
allocator
- make the DMA mask an upper bound and don't fail "too large" dma mask
in the remaning two architectures - this will allow big driver
cleanups in the following merge windows
* tag 'dma-mapping-5.1' of git://git.infradead.org/users/hch/dma-mapping: (21 commits)
Documentation/DMA-API-HOWTO: update dma_mask sections
sparc64/pci_sun4v: allow large DMA masks
sparc64/iommu: allow large DMA masks
sparc64: refactor the ali DMA quirk
ccio: allow large DMA masks
dma-mapping: remove the DMA_MEMORY_EXCLUSIVE flag
dma-mapping: remove dma_mark_declared_memory_occupied
dma-mapping: move CONFIG_DMA_CMA to kernel/dma/Kconfig
dma-mapping: improve selection of dma_declare_coherent availability
dma-mapping: remove an incorrect __iommem annotation
of: select OF_RESERVED_MEM automatically
device.h: dma_mem is only needed for HAVE_GENERIC_DMA_COHERENT
mfd/sm501: depend on HAS_DMA
dma-mapping: add a kconfig symbol for arch_teardown_dma_ops availability
dma-mapping: add a kconfig symbol for arch_setup_dma_ops availability
dma-mapping: move debug configuration options to kernel/dma
dma-debug: add dumping facility via debugfs
dma: debug: no need to check return value of debugfs_create functions
videobuf2: replace a layering violation with dma_map_resource
dma-mapping: don't BUG when calling dma_map_resource on RAM
...
Pull arm64 updates from Catalin Marinas:
- Pseudo NMI support for arm64 using GICv3 interrupt priorities
- uaccess macros clean-up (unsafe user accessors also merged but
reverted, waiting for objtool support on arm64)
- ptrace regsets for Pointer Authentication (ARMv8.3) key management
- inX() ordering w.r.t. delay() on arm64 and riscv (acks in place by
the riscv maintainers)
- arm64/perf updates: PMU bindings converted to json-schema, unused
variable and misleading comment removed
- arm64/debug fixes to ensure checking of the triggering exception
level and to avoid the propagation of the UNKNOWN FAR value into the
si_code for debug signals
- Workaround for Fujitsu A64FX erratum 010001
- lib/raid6 ARM NEON optimisations
- NR_CPUS now defaults to 256 on arm64
- Minor clean-ups (documentation/comments, Kconfig warning, unused
asm-offsets, clang warnings)
- MAINTAINERS update for list information to the ARM64 ACPI entry
* tag 'arm64-upstream' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux: (54 commits)
arm64: mmu: drop paging_init comments
arm64: debug: Ensure debug handlers check triggering exception level
arm64: debug: Don't propagate UNKNOWN FAR into si_code for debug signals
Revert "arm64: uaccess: Implement unsafe accessors"
arm64: avoid clang warning about self-assignment
arm64: Kconfig.platforms: fix warning unmet direct dependencies
lib/raid6: arm: optimize away a mask operation in NEON recovery routine
lib/raid6: use vdupq_n_u8 to avoid endianness warnings
arm64: io: Hook up __io_par() for inX() ordering
riscv: io: Update __io_[p]ar() macros to take an argument
asm-generic/io: Pass result of I/O accessor to __io_[p]ar()
arm64: Add workaround for Fujitsu A64FX erratum 010001
arm64: Rename get_thread_info()
arm64: Remove documentation about TIF_USEDFPU
arm64: irqflags: Fix clang build warnings
arm64: Enable the support of pseudo-NMIs
arm64: Skip irqflags tracing for NMI in IRQs disabled context
arm64: Skip preemption when exiting an NMI
arm64: Handle serror in NMI context
irqchip/gic-v3: Allow interrupts to be set as pseudo-NMI
...
