Pull KVM updates from Paolo Bonzini:
"One NULL pointer dereference, and two fixes for regressions introduced
during the merge window.
The rest are fixes for MIPS, s390 and nested VMX"
* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
kvm: x86: Check memopp before dereference (CVE-2016-8630)
kvm: nVMX: VMCLEAR an active shadow VMCS after last use
KVM: x86: drop TSC offsetting kvm_x86_ops to fix KVM_GET/SET_CLOCK
KVM: x86: fix wbinvd_dirty_mask use-after-free
kvm/x86: Show WRMSR data is in hex
kvm: nVMX: Fix kernel panics induced by illegal INVEPT/INVVPID types
KVM: document lock orders
KVM: fix OOPS on flush_work
KVM: s390: Fix STHYI buffer alignment for diag224
KVM: MIPS: Precalculate MMIO load resume PC
KVM: MIPS: Make ERET handle ERL before EXL
KVM: MIPS: Fix lazy user ASID regenerate for SMP
When low memory doesn't reach HIGHMEM_START (e.g. up to 256MB at PA=0 is
common) and highmem is present above HIGHMEM_START (e.g. on Malta the
RAM overlayed by the IO region is aliased at PA=0x90000000), max_low_pfn
will be initially calculated very large and then clipped down to
HIGHMEM_START.
This causes crashes when reading /sys/kernel/mm/page_idle/bitmap
(i.e. CONFIG_IDLE_PAGE_TRACKING=y) when highmem is disabled. pfn_valid()
will compare against max_mapnr which is derived from max_low_pfn when
there is no highend_pfn set up, and will return true for PFNs right up
to HIGHMEM_START, even though they are beyond the end of low memory and
no page structs will actually exist for these PFNs.
This is fixed by skipping high memory regions when initially calculating
max_low_pfn if highmem is disabled, so it doesn't get clipped too high.
We also clip regions which overlap the highmem boundary when highmem is
disabled, so that max_pfn doesn't extend into highmem either.
Signed-off-by: James Hogan <james.hogan@imgtec.com>
Cc: Paul Burton <paul.burton@imgtec.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/14490/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Complement commit 80cbfad790 ("MIPS: Correct MIPS I FP context
layout") and correct the way Floating Point General registers are stored
in a signal context with MIPS I hardware.
Use the S.D and L.D assembly macros to have pairs of SWC1 instructions
and pairs of LWC1 instructions produced, respectively, in an arrangement
which makes the memory representation of floating-point data passed
compatible with that used by hardware SDC1 and LDC1 instructions, where
available, regardless of the hardware endianness used. This matches the
layout used by r4k_fpu.S, ensuring run-time compatibility for MIPS I
software across all o32 hardware platforms.
Define an EX2 macro to handle exceptions from both hardware instructions
implicitly produced from S.D and L.D assembly macros.
Signed-off-by: Maciej W. Rozycki <macro@imgtec.com>
Cc: linux-mips@linux-mips.org
Cc: linux-kernel@vger.kernel.org
Patchwork: https://patchwork.linux-mips.org/patch/14477/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Fix a regression introduced with commit 2db9ca0a35 ("MIPS: Use struct
mips_abi offsets to save FP context") for MIPS I/I FP signal contexts,
by converting save/restore code to the updated internal API. Start FGR
offsets from 0 rather than SC_FPREGS from $a0 and use $a1 rather than
the offset of SC_FPC_CSR from $a0 for the Floating Point Control/Status
Register (FCSR).
Document the new internal API and adjust assembly code formatting for
consistency.
Signed-off-by: Maciej W. Rozycki <macro@imgtec.com>
Cc: Paul Burton <paul.burton@imgtec.com>
Cc: linux-mips@linux-mips.org
Cc: linux-kernel@vger.kernel.org
Patchwork: https://patchwork.linux-mips.org/patch/14476/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Complement commit e50c0a8fa6 ("Support the MIPS32 / MIPS64 DSP ASE.")
and remove the Floating Point Implementation Register (FIR) from the FP
register set recorded in a signal context with MIPS I processors too, in
line with the change applied to r4k_fpu.S.
The `sc_fpc_eir' slot is unused according to our current ABI and the FIR
register is read-only and always directly accessible from user software.
[ralf@linux-mips.org: This is also required because the next commit depends
on it.]
Signed-off-by: Maciej W. Rozycki <macro@imgtec.com>
Cc: linux-mips@linux-mips.org
Cc: linux-kernel@vger.kernel.org
Patchwork: https://patchwork.linux-mips.org/patch/14475/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Complement commit 0ae8dceaebe3 ("Merge with 2.3.10.") and use the local
`fault' handler to recover from FP sigcontext access violation faults,
like corresponding code does in r4k_fpu.S. The `bad_stack' handler is
in syscall.c and is not suitable here as we want to propagate the error
condition up through the caller rather than killing the thread outright.
Signed-off-by: Maciej W. Rozycki <macro@imgtec.com>
Cc: linux-mips@linux-mips.org
Cc: linux-kernel@vger.kernel.org
Patchwork: https://patchwork.linux-mips.org/patch/14474/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Sanitize FCSR Cause bit handling, following a trail of past attempts:
* commit 4249548454 ("MIPS: ptrace: Fix FP context restoration FCSR
regression"),
* commit 443c44032a ("MIPS: Always clear FCSR cause bits after
emulation"),
* commit 64bedffe49 ("MIPS: Clear [MSA]FPE CSR.Cause after
notify_die()"),
* commit b1442d39fa ("MIPS: Prevent user from setting FCSR cause
bits"),
* commit b54d2901517d ("Properly handle branch delay slots in connection
with signals.").
Specifically do not mask these bits out in ptrace(2) processing and send
a SIGFPE signal instead whenever a matching pair of an FCSR Cause and
Enable bit is seen as execution of an affected context is about to
resume. Only then clear Cause bits, and even then do not clear any bits
that are set but masked with the respective Enable bits. Adjust Cause
bit clearing throughout code likewise, except within the FPU emulator
proper where they are set according to IEEE 754 exceptions raised as the
operation emulated executed. Do so so that any IEEE 754 exceptions
subject to their default handling are recorded like with operations
executed by FPU hardware.
Signed-off-by: Maciej W. Rozycki <macro@imgtec.com>
Cc: Paul Burton <paul.burton@imgtec.com>
Cc: James Hogan <james.hogan@imgtec.com>
Cc: linux-mips@linux-mips.org
Cc: linux-kernel@vger.kernel.org
Patchwork: https://patchwork.linux-mips.org/patch/14460/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Since commit 4bcc595ccd ("printk: reinstate KERN_CONT for printing
continuation lines") the output from TLB dumps on MIPS has been
pretty unreadable due to the lack of KERN_CONT markers. Use pr_cont to
provide the appropriate markers & restore the expected output.
Continuation is also used for the second line of each TLB entry printed
in dump_tlb.c even though it has a newline, since it is a continuation
of the interpretation of the same TLB entry. For example:
[ 46.371884] Index: 0 pgmask=16kb va=77654000 asid=73 gid=00
[ri=0 xi=0 pa=ffc18000 c=5 d=0 v=1 g=0] [ri=0 xi=0 pa=ffc1c000 c=5 d=0 v=1 g=0]
[ 46.385380] Index: 12 pgmask=16kb va=004b4000 asid=73 gid=00
[ri=0 xi=0 pa=00000000 c=0 d=0 v=0 g=0] [ri=0 xi=0 pa=ffb00000 c=5 d=1 v=1 g=0]
Signed-off-by: James Hogan <james.hogan@imgtec.com>
Cc: Maciej W. Rozycki <macro@imgtec.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/14444/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
The KASLR code requires that the plat_get_fdt() function return the
address of the device tree, and it must be available early in the boot,
before prom_init() is called. Move the code determining the address of
the device tree into plat_get_fdt, and call that from prom_init().
The fdt pointer will be set up by plat_get_fdt() called from
relocate_kernel initially and once the relocated kernel has started,
prom_init() will use it again to determine the address in the relocated
image.
Fixes: eed0eabd12 ("MIPS: generic: Introduce generic DT-based board support")
Signed-off-by: Matt Redfearn <matt.redfearn@imgtec.com>
Reviewed-by: James Hogan <james.hogan@imgtec.com>
Reviewed-by: Paul Burton <paul.burton@imgtec.com>
Cc: linux-mips@linux-mips.org
Cc: linux-kernel@vger.kernel.org
Patchwork: https://patchwork.linux-mips.org/patch/14415/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Commit 10b6ea0959 ("MIPS: Malta: Use syscon-reboot driver to reboot")
converted the Malta board to use the generic syscon-reboot driver to
handle reboots, but incorrectly used the value 0x4d rather than 0x42 as
the magic to write to the reboot register.
I also incorrectly believed that syscon/regmap would default to native
endianness, but this isn't the case. Force this by specifying with a
native-endian property in the devicetree.
Signed-off-by: Paul Burton <paul.burton@imgtec.com>
Fixes: 10b6ea0959 ("MIPS: Malta: Use syscon-reboot driver to reboot")
Reported-by: Guenter Roeck <linux@roeck-us.net>
Cc: linux-mips@linux-mips.org
Tested-by: Guenter Roeck <linux@roeck-us.net>
Tested-by: Maciej W. Rozycki <macro@imgtec.com>
Patchwork: https://patchwork.linux-mips.org/patch/14396/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Provide a default implementation of mips_cpc_default_phys_base() which
simply returns 0, and adjust mips_cpc_phys_base() to allow for
mips_cpc_default_phys_base() returning 0. This allows kernels which
include CPC support to be built without platform code & simply ignore
the CPC if it wasn't already enabled by the bootloader.
This fixes link failures such as the following from generic defconfigs:
arch/mips/built-in.o: In function `mips_cpc_phys_base':
arch/mips/kernel/mips-cpc.c:47: undefined reference to `mips_cpc_default_phys_base'
[ralf@linux-mips.org: changed prototype for coding style compliance.]
Signed-off-by: Paul Burton <paul.burton@imgtec.com>
Reported-by: kbuild test robot <fengguang.wu@intel.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/14401/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
The advancing of the PC when completing an MMIO load is done before
re-entering the guest, i.e. before restoring the guest ASID. However if
the load is in a branch delay slot it may need to access guest code to
read the prior branch instruction. This isn't safe in TLB mapped code at
the moment, nor in the future when we'll access unmapped guest segments
using direct user accessors too, as it could read the branch from host
user memory instead.
Therefore calculate the resume PC in advance while we're still in the
right context and save it in the new vcpu->arch.io_pc (replacing the no
longer needed vcpu->arch.pending_load_cause), and restore it on MMIO
completion.
Fixes: e685c689f3 ("KVM/MIPS32: Privileged instruction/target branch emulation.")
Signed-off-by: James Hogan <james.hogan@imgtec.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: "Radim Krčmář <rkrcmar@redhat.com>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: linux-mips@linux-mips.org
Cc: kvm@vger.kernel.org
Cc: <stable@vger.kernel.org> # 3.10.x-
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The ERET instruction to return from exception is used for returning from
exception level (Status.EXL) and error level (Status.ERL). If both bits
are set however we should be returning from ERL first, as ERL can
interrupt EXL, for example when an NMI is taken. KVM however checks EXL
first.
Fix the order of the checks to match the pseudocode in the instruction
set manual.
Fixes: e685c689f3 ("KVM/MIPS32: Privileged instruction/target branch emulation.")
Signed-off-by: James Hogan <james.hogan@imgtec.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: "Radim Krčmář <rkrcmar@redhat.com>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: linux-mips@linux-mips.org
Cc: kvm@vger.kernel.org
Cc: <stable@vger.kernel.org> # 3.10.x-
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
kvm_mips_check_asids() runs before entering the guest and performs lazy
regeneration of host ASID for guest usermode, using last_user_gasid to
track the last guest ASID in the VCPU that was used by guest usermode on
any host CPU.
last_user_gasid is reset after performing the lazy ASID regeneration on
the current CPU, and by kvm_arch_vcpu_load() if the host ASID for guest
usermode is regenerated due to staleness (to cancel outstanding lazy
ASID regenerations). Unfortunately neither case handles SMP hosts
correctly:
- When the lazy ASID regeneration is performed it should apply to all
CPUs (as last_user_gasid does), so reset the ASID on other CPUs to
zero to trigger regeneration when the VCPU is next loaded on those
CPUs.
- When the ASID is found to be stale on the current CPU, we should not
cancel lazy ASID regenerations globally, so drop the reset of
last_user_gasid altogether here.
Both cases would require a guest ASID change and two host CPU migrations
(and in the latter case one of the CPUs to start a new ASID cycle)
before guest usermode could potentially access stale user pages from a
previously running ASID in the same VCPU.
Fixes: 25b08c7fb0 ("KVM: MIPS: Invalidate TLB by regenerating ASIDs")
Signed-off-by: James Hogan <james.hogan@imgtec.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: "Radim Krčmář <rkrcmar@redhat.com>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: linux-mips@linux-mips.org
Cc: kvm@vger.kernel.org
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Merge the gup_flags cleanups from Lorenzo Stoakes:
"This patch series adjusts functions in the get_user_pages* family such
that desired FOLL_* flags are passed as an argument rather than
implied by flags.
The purpose of this change is to make the use of FOLL_FORCE explicit
so it is easier to grep for and clearer to callers that this flag is
being used. The use of FOLL_FORCE is an issue as it overrides missing
VM_READ/VM_WRITE flags for the VMA whose pages we are reading
from/writing to, which can result in surprising behaviour.
The patch series came out of the discussion around commit 38e0885465
("mm: check VMA flags to avoid invalid PROT_NONE NUMA balancing"),
which addressed a BUG_ON() being triggered when a page was faulted in
with PROT_NONE set but having been overridden by FOLL_FORCE.
do_numa_page() was run on the assumption the page _must_ be one marked
for NUMA node migration as an actual PROT_NONE page would have been
dealt with prior to this code path, however FOLL_FORCE introduced a
situation where this assumption did not hold.
See
https://marc.info/?l=linux-mm&m=147585445805166
for the patch proposal"
Additionally, there's a fix for an ancient bug related to FOLL_FORCE and
FOLL_WRITE by me.
[ This branch was rebased recently to add a few more acked-by's and
reviewed-by's ]
* gup_flag-cleanups:
mm: replace access_process_vm() write parameter with gup_flags
mm: replace access_remote_vm() write parameter with gup_flags
mm: replace __access_remote_vm() write parameter with gup_flags
mm: replace get_user_pages_remote() write/force parameters with gup_flags
mm: replace get_user_pages() write/force parameters with gup_flags
mm: replace get_vaddr_frames() write/force parameters with gup_flags
mm: replace get_user_pages_locked() write/force parameters with gup_flags
mm: replace get_user_pages_unlocked() write/force parameters with gup_flags
mm: remove write/force parameters from __get_user_pages_unlocked()
mm: remove write/force parameters from __get_user_pages_locked()
mm: remove gup_flags FOLL_WRITE games from __get_user_pages()
This removes the 'write' argument from access_process_vm() and replaces
it with 'gup_flags' as use of this function previously silently implied
FOLL_FORCE, whereas after this patch callers explicitly pass this flag.
We make this explicit as use of FOLL_FORCE can result in surprising
behaviour (and hence bugs) within the mm subsystem.
Signed-off-by: Lorenzo Stoakes <lstoakes@gmail.com>
Acked-by: Jesper Nilsson <jesper.nilsson@axis.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Acked-by: Michael Ellerman <mpe@ellerman.id.au>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
MIPS KVM uses user memory accessors but mips.c doesn't directly include
uaccess.h, so include it now.
This wasn't too much of a problem before v4.9-rc1 as asm/module.h
included asm/uaccess.h, however since commit 29abfbd9cb ("mips:
separate extable.h, switch module.h to it") this is no longer the case.
This resulted in build failures when trace points were disabled, as
trace/define_trace.h includes trace/trace_events.h only ifdef
TRACEPOINTS_ENABLED, which goes on to include asm/uaccess.h via a couple
of other headers.
Fixes: 29abfbd9cb ("mips: separate extable.h, switch module.h to it")
Signed-off-by: James Hogan <james.hogan@imgtec.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: "Radim Krčmář" <rkrcmar@redhat.com>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: linux-mips@linux-mips.org
Cc: kvm@vger.kernel.org
Linus Torvalds
James Hogan
Maciej W. Rozycki
Paul Burton
Matt Redfearn
Lorenzo Stoakes