For VMEXITs caused by external interrupts, vmx_handle_external_intr()
indirectly calls into the interrupt handlers through the host's IDT.
It follows that these interrupts get accounted for in the
kvm_cpu_l1tf_flush_l1d per-cpu flag.
The subsequently executed vmx_l1d_flush() will thus be aware that some
interrupts have happened and conduct a L1d flush anyway.
Setting l1tf_flush_l1d from vmx_handle_external_intr() isn't needed
anymore. Drop it.
Signed-off-by: Nicolai Stange <nstange@suse.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
The last missing piece to having vmx_l1d_flush() take interrupts after
VMEXIT into account is to set the kvm_cpu_l1tf_flush_l1d per-cpu flag on
irq entry.
Issue calls to kvm_set_cpu_l1tf_flush_l1d() from entering_irq(),
ipi_entering_ack_irq(), smp_reschedule_interrupt() and
uv_bau_message_interrupt().
Suggested-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Nicolai Stange <nstange@suse.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
The next patch in this series will have to make the definition of
irq_cpustat_t available to entering_irq().
Inclusion of asm/hardirq.h into asm/apic.h would cause circular header
dependencies like
asm/smp.h
asm/apic.h
asm/hardirq.h
linux/irq.h
linux/topology.h
linux/smp.h
asm/smp.h
or
linux/gfp.h
linux/mmzone.h
asm/mmzone.h
asm/mmzone_64.h
asm/smp.h
asm/apic.h
asm/hardirq.h
linux/irq.h
linux/irqdesc.h
linux/kobject.h
linux/sysfs.h
linux/kernfs.h
linux/idr.h
linux/gfp.h
and others.
This causes compilation errors because of the header guards becoming
effective in the second inclusion: symbols/macros that had been defined
before wouldn't be available to intermediate headers in the #include chain
anymore.
A possible workaround would be to move the definition of irq_cpustat_t
into its own header and include that from both, asm/hardirq.h and
asm/apic.h.
However, this wouldn't solve the real problem, namely asm/harirq.h
unnecessarily pulling in all the linux/irq.h cruft: nothing in
asm/hardirq.h itself requires it. Also, note that there are some other
archs, like e.g. arm64, which don't have that #include in their
asm/hardirq.h.
Remove the linux/irq.h #include from x86' asm/hardirq.h.
Fix resulting compilation errors by adding appropriate #includes to *.c
files as needed.
Note that some of these *.c files could be cleaned up a bit wrt. to their
set of #includes, but that should better be done from separate patches, if
at all.
Signed-off-by: Nicolai Stange <nstange@suse.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Part of the L1TF mitigation for vmx includes flushing the L1D cache upon
VMENTRY.
L1D flushes are costly and two modes of operations are provided to users:
"always" and the more selective "conditional" mode.
If operating in the latter, the cache would get flushed only if a host side
code path considered unconfined had been traversed. "Unconfined" in this
context means that it might have pulled in sensitive data like user data
or kernel crypto keys.
The need for L1D flushes is tracked by means of the per-vcpu flag
l1tf_flush_l1d. KVM exit handlers considered unconfined set it. A
vmx_l1d_flush() subsequently invoked before the next VMENTER will conduct a
L1d flush based on its value and reset that flag again.
Currently, interrupts delivered "normally" while in root operation between
VMEXIT and VMENTER are not taken into account. Part of the reason is that
these don't leave any traces and thus, the vmx code is unable to tell if
any such has happened.
As proposed by Paolo Bonzini, prepare for tracking all interrupts by
introducing a new per-cpu flag, "kvm_cpu_l1tf_flush_l1d". It will be in
strong analogy to the per-vcpu ->l1tf_flush_l1d.
A later patch will make interrupt handlers set it.
For the sake of cache locality, group kvm_cpu_l1tf_flush_l1d into x86'
per-cpu irq_cpustat_t as suggested by Peter Zijlstra.
Provide the helpers kvm_set_cpu_l1tf_flush_l1d(),
kvm_clear_cpu_l1tf_flush_l1d() and kvm_get_cpu_l1tf_flush_l1d(). Make them
trivial resp. non-existent for !CONFIG_KVM_INTEL as appropriate.
Let vmx_l1d_flush() handle kvm_cpu_l1tf_flush_l1d in the same way as
l1tf_flush_l1d.
Suggested-by: Paolo Bonzini <pbonzini@redhat.com>
Suggested-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Nicolai Stange <nstange@suse.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
An upcoming patch will extend KVM's L1TF mitigation in conditional mode
to also cover interrupts after VMEXITs. For tracking those, stores to a
new per-cpu flag from interrupt handlers will become necessary.
In order to improve cache locality, this new flag will be added to x86's
irq_cpustat_t.
Make some space available there by shrinking the ->softirq_pending bitfield
from 32 to 16 bits: the number of bits actually used is only NR_SOFTIRQS,
i.e. 10.
Suggested-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Nicolai Stange <nstange@suse.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Currently, vmx_vcpu_run() checks if l1tf_flush_l1d is set and invokes
vmx_l1d_flush() if so.
This test is unncessary for the "always flush L1D" mode.
Move the check to vmx_l1d_flush()'s conditional mode code path.
Notes:
- vmx_l1d_flush() is likely to get inlined anyway and thus, there's no
extra function call.
- This inverts the (static) branch prediction, but there hadn't been any
explicit likely()/unlikely() annotations before and so it stays as is.
Signed-off-by: Nicolai Stange <nstange@suse.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
The vmx_l1d_flush_always static key is only ever evaluated if
vmx_l1d_should_flush is enabled. In that case however, there are only two
L1d flushing modes possible: "always" and "conditional".
The "conditional" mode's implementation tends to require more sophisticated
logic than the "always" mode.
Avoid inverted logic by replacing the 'vmx_l1d_flush_always' static key
with a 'vmx_l1d_flush_cond' one.
There is no change in functionality.
Signed-off-by: Nicolai Stange <nstange@suse.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
vmx_l1d_flush() gets invoked only if l1tf_flush_l1d is true. There's no
point in setting l1tf_flush_l1d to true from there again.
Signed-off-by: Nicolai Stange <nstange@suse.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Pull KVM fixes from Paolo Bonzini:
"Two vmx bugfixes"
* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
kvm: x86: vmx: fix vpid leak
KVM: vmx: use local variable for current_vmptr when emulating VMPTRST
Pull powerpc fixes from Michael Ellerman:
"One fix for a regression in a recent TLB flush optimisation, which
caused us to incorrectly not send TLB invalidations to coprocessors.
Thanks to Frederic Barrat, Nicholas Piggin, Vaibhav Jain"
* tag 'powerpc-4.18-5' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:
powerpc/64s/radix: Fix missing global invalidations when removing copro
Future Intel processors will support "Enhanced IBRS" which is an "always
on" mode i.e. IBRS bit in SPEC_CTRL MSR is enabled once and never
disabled.
From the specification [1]:
"With enhanced IBRS, the predicted targets of indirect branches
executed cannot be controlled by software that was executed in a less
privileged predictor mode or on another logical processor. As a
result, software operating on a processor with enhanced IBRS need not
use WRMSR to set IA32_SPEC_CTRL.IBRS after every transition to a more
privileged predictor mode. Software can isolate predictor modes
effectively simply by setting the bit once. Software need not disable
enhanced IBRS prior to entering a sleep state such as MWAIT or HLT."
If Enhanced IBRS is supported by the processor then use it as the
preferred spectre v2 mitigation mechanism instead of Retpoline. Intel's
Retpoline white paper [2] states:
"Retpoline is known to be an effective branch target injection (Spectre
variant 2) mitigation on Intel processors belonging to family 6
(enumerated by the CPUID instruction) that do not have support for
enhanced IBRS. On processors that support enhanced IBRS, it should be
used for mitigation instead of retpoline."
The reason why Enhanced IBRS is the recommended mitigation on processors
which support it is that these processors also support CET which
provides a defense against ROP attacks. Retpoline is very similar to ROP
techniques and might trigger false positives in the CET defense.
If Enhanced IBRS is selected as the mitigation technique for spectre v2,
the IBRS bit in SPEC_CTRL MSR is set once at boot time and never
cleared. Kernel also has to make sure that IBRS bit remains set after
VMEXIT because the guest might have cleared the bit. This is already
covered by the existing x86_spec_ctrl_set_guest() and
x86_spec_ctrl_restore_host() speculation control functions.
Enhanced IBRS still requires IBPB for full mitigation.
[1] Speculative-Execution-Side-Channel-Mitigations.pdf
[2] Retpoline-A-Branch-Target-Injection-Mitigation.pdf
Both documents are available at:
https://bugzilla.kernel.org/show_bug.cgi?id=199511
Originally-by: David Woodhouse <dwmw@amazon.co.uk>
Signed-off-by: Sai Praneeth Prakhya <sai.praneeth.prakhya@intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Tim C Chen <tim.c.chen@intel.com>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Ravi Shankar <ravi.v.shankar@intel.com>
Link: https://lkml.kernel.org/r/1533148945-24095-1-git-send-email-sai.praneeth.prakhya@intel.com
Some Intel processors have an EPT feature whereby the accessed & dirty bits
in EPT entries can be updated by HW. MSR IA32_VMX_EPT_VPID_CAP exposes the
presence of this capability.
There is no point in trying to use that new feature bit in the VMX code as
VMX needs to read the MSR anyway to access other bits, but having the
feature bit for EPT_AD in place helps virtualization management as it
exposes "ept_ad" in /proc/cpuinfo/$proc/flags if the feature is present.
[ tglx: Amended changelog ]
Signed-off-by: Peter Feiner <pfeiner@google.com>
Signed-off-by: Peter Shier <pshier@google.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Jim Mattson <jmattson@google.com>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Borislav Petkov <bp@suse.de>
Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Cc: David Woodhouse <dwmw@amazon.co.uk>
Link: https://lkml.kernel.org/r/20180801180657.138051-1-pshier@google.com
We've encountered a performance issue when multiple processors stress
{get,put}_mmio_atsd_reg(). These functions contend for
mmio_atsd_usage, an unsigned long used as a bitmask.
The accesses to mmio_atsd_usage are done using test_and_set_bit_lock()
and clear_bit_unlock(). As implemented, both of these will require
a (successful) stwcx to that same cache line.
What we end up with is thread A, attempting to unlock, being slowed by
other threads repeatedly attempting to lock. A's stwcx instructions
fail and retry because the memory reservation is lost every time a
different thread beats it to the punch.
There may be a long-term way to fix this at a larger scale, but for
now resolve the immediate problem by gating our call to
test_and_set_bit_lock() with one to test_bit(), which is obviously
implemented without using a store.
Fixes: 1ab66d1fba ("powerpc/powernv: Introduce address translation services for Nvlink2")
Signed-off-by: Reza Arbab <arbab@linux.ibm.com>
Acked-by: Alistair Popple <alistair@popple.id.au>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
kernel/dma/Kconfig already defines NEED_DMA_MAP_STATE, just select it
from CONFIG_PPC using the same condition as an if guard.
Signed-off-by: Christoph Hellwig <hch@lst.de>
[mpe: Move it under PPC]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
The 4-way ChaCha20 NEON code implements 16-bit rotates with vrev32.16,
but the one-way code (used on remainder blocks) implements it with
vshl + vsri, which is slower. Switch the one-way code to vrev32.16 too.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Acked-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Commit 9c7b185ab2 ("powernv/cpuidle: Parse dt idle properties into
global structure") parses dt idle states into structs, but never marks
them valid. This results in all idle states being lost.
Fixes: 9c7b185ab2 ("powernv/cpuidle: Parse dt idle properties into global structure")
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Acked-by: Akshay Adiga <akshay.adiga@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
We also split up the regmap address space into two, for the TM and SROT
registers. This was required to deal with different address offsets for the
TM and SROT registers across different SoC families.
8996 has two TSENS IP blocks, initialise the second one too.
Since tsens-common.c/init_common() currently only registers one address
space, the order is important (TM before SROT). This is OK since the code
doesn't really use the SROT functionality yet.
Signed-off-by: Amit Kucheria <amit.kucheria@linaro.org>
Reviewed-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Tested-by: Matthias Kaehlcke <mka@chromium.org>
Reviewed-by: Matthias Kaehlcke <mka@chromium.org>
Reviewed-by: Douglas Anderson <dianders@chromium.org>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
The BTF conflicts were simple overlapping changes.
The virtio_net conflict was an overlap of a fix of statistics counter,
happening alongisde a move over to a bonafide statistics structure
rather than counting value on the stack.
Signed-off-by: David S. Miller <davem@davemloft.net>
Pull arm64 regression fix from Will Deacon:
"Ard found a nasty arm64 regression in 4.18 where the AES ghash/gcm
code doesn't notify the kernel about its use of the vector registers,
therefore potentially corrupting live user state.
The fix is straightforward and Herbert agreed for it to go via arm64"
* tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux:
crypto/arm64: aes-ce-gcm - add missing kernel_neon_begin/end pair
generic_defconfig explicitly disables CONFIG_INPUT_MOUSEDEV,
CONFIG_INPUT_KEYBOARD & CONFIG_INPUT_MOUSE which results in warnings
when merging board config fragments if any of them require these
options. This is the case for the ranchu board, which means we've had
the following warning when configuring for generic platform targets
since commit f2d0b0d5c1 ("MIPS: ranchu: Add Ranchu as a new
generic-based board"):
$ make ARCH=mips 32r2el_defconfig
Using ./arch/mips/configs/generic_defconfig as base
Merging arch/mips/configs/generic/32r2.config
Merging arch/mips/configs/generic/el.config
Merging ./arch/mips/configs/generic/board-sead-3.config
Merging ./arch/mips/configs/generic/board-ranchu.config
Value of CONFIG_INPUT_KEYBOARD is redefined by fragment ./arch/mips/configs/generic/board-ranchu.config:
Previous value: # CONFIG_INPUT_KEYBOARD is not set
New value: CONFIG_INPUT_KEYBOARD=y
Merging ./arch/mips/configs/generic/board-ni169445.config
Merging ./arch/mips/configs/generic/board-boston.config
Merging ./arch/mips/configs/generic/board-ocelot.config
Merging ./arch/mips/configs/generic/board-xilfpga.config
scripts/kconfig/conf --olddefconfig Kconfig
#
# configuration written to .config
#
Resolve this by removing mention of the CONFIG_INPUT_* Kconfig symbols
from generic_defconfig, allowing them to take their default values &
allowing board config fragments to enable them without warnings.
This may be problematic if CONFIG_ARCH_MIGHT_HAVE_PC_SERIO is ever
enabled for CONFIG_MIPS_GENERIC=y configurations, but for now that isn't
the case so we can worry about that if & when it happens.
Signed-off-by: Paul Burton <paul.burton@mips.com>
Patchwork: https://patchwork.linux-mips.org/patch/20109/
Cc: James Hogan <jhogan@kernel.org>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: linux-mips@linux-mips.org
Spectre variant 1 attacks are about this sequence of pseudo-code:
index = load(user-manipulated pointer);
access(base + index * stride);
In order for the cache side-channel to work, the access() must me made
to memory which userspace can detect whether cache lines have been
loaded. On 32-bit ARM, this must be either user accessible memory, or
a kernel mapping of that same user accessible memory.
The problem occurs when the load() speculatively loads privileged data,
and the subsequent access() is made to user accessible memory.
Any load() which makes use of a user-maniplated pointer is a potential
problem if the data it has loaded is used in a subsequent access. This
also applies for the access() if the data loaded by that access is used
by a subsequent access.
Harden the get_user() accessors against Spectre attacks by forcing out
of bounds addresses to a NULL pointer. This prevents get_user() being
used as the load() step above. As a side effect, put_user() will also
be affected even though it isn't implicated.
Also harden copy_from_user() by redoing the bounds check within the
arm_copy_from_user() code, and NULLing the pointer if out of bounds.
Acked-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
Fixing __get_user() for spectre variant 1 is not sane: we would have to
add address space bounds checking in order to validate that the location
should be accessed, and then zero the address if found to be invalid.
Since __get_user() is supposed to avoid the bounds check, and this is
exactly what get_user() does, there's no point having two different
implementations that are doing the same thing. So, when the Spectre
workarounds are required, make __get_user() an alias of get_user().
Acked-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
Borrow the x86 implementation of __inttype() to use in get_user() to
select an integer type suitable to temporarily hold the result value.
This is necessary to avoid propagating the volatile nature of the
result argument, which can cause the following warning:
lib/iov_iter.c:413:5: warning: optimization may eliminate reads and/or writes to register variables [-Wvolatile-register-var]
Acked-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
__get_user_error() is used as a fast accessor to make copying structure
members as efficient as possible. However, with software PAN and the
recent Spectre variant 1, the efficiency is reduced as these are no
longer fast accessors.
In the case of software PAN, it has to switch the domain register around
each access, and with Spectre variant 1, it would have to repeat the
access_ok() check for each access.
Rather than using __get_user_error() to copy each semops element member,
copy each semops element in full using __copy_from_user().
Acked-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
__get_user_error() is used as a fast accessor to make copying structure
members in the signal handling path as efficient as possible. However,
with software PAN and the recent Spectre variant 1, the efficiency is
reduced as these are no longer fast accessors.
In the case of software PAN, it has to switch the domain register around
each access, and with Spectre variant 1, it would have to repeat the
access_ok() check for each access.
Use __copy_from_user() rather than __get_user_err() for individual
members when restoring VFP state.
Acked-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
KVM: s390: Features for 4.19
- initial version for host large page support. Must be enabled with
module parameter hpage=1 and will conflict with the nested=1
parameter.
- enable etoken facility for guests
- Fixes
Switch to the generic noncoherent direct mapping implementation.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: Yoshinori Sato <ysato@users.sourceforge.jp>
Half of the file just contains platform device memory setup code which
is required for all builds, and half contains helpers for dma coherent
allocation, which is only needed if CONFIG_DMA_NONCOHERENT is enabled.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: Yoshinori Sato <ysato@users.sourceforge.jp>
This is a slight change in behavior as we avoid the detour through the
virtual mapping for the coherent allocator, but if this CPU really is
coherent that should be the right thing to do.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: Yoshinori Sato <ysato@users.sourceforge.jp>
And use it in the maple bus code to avoid a dma API dependency.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: Yoshinori Sato <ysato@users.sourceforge.jp>
Remove the indirection through the dma_ops variable, and just return
nommu_dma_ops directly from get_arch_dma_ops.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: Yoshinori Sato <ysato@users.sourceforge.jp>
