Commit Graph

722808 Commits

Author SHA1 Message Date
Christoffer Dall
9b062471e5 KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl
Move the calls to vcpu_load() and vcpu_put() in to the architecture
specific implementations of kvm_arch_vcpu_ioctl() which dispatches
further architecture-specific ioctls on to other functions.

Some architectures support asynchronous vcpu ioctls which cannot call
vcpu_load() or take the vcpu->mutex, because that would prevent
concurrent execution with a running VCPU, which is the intended purpose
of these ioctls, for example because they inject interrupts.

We repeat the separate checks for these specifics in the architecture
code for MIPS, S390 and PPC, and avoid taking the vcpu->mutex and
calling vcpu_load for these ioctls.

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:26:58 +01:00
Christoffer Dall
6a96bc7fa0 KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_fpu
Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_set_fpu().

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:26:57 +01:00
Christoffer Dall
1393123e1e KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_fpu
Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_get_fpu().

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:26:56 +01:00
Christoffer Dall
66b5656222 KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_guest_debug
Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_set_guest_debug().

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:26:56 +01:00
Christoffer Dall
1da5b61dac KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_translate
Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_translate().

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:26:55 +01:00
Christoffer Dall
e83dff5edf KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_mpstate
Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_set_mpstate().

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:26:54 +01:00
Christoffer Dall
fd2325612c KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_mpstate
Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_get_mpstate().

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:26:54 +01:00
Christoffer Dall
b4ef9d4e8c KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_sregs
Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_set_sregs().

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
Reviewed-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:26:53 +01:00
Christoffer Dall
bcdec41cef KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_sregs
Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_get_sregs().

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
Reviewed-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:26:52 +01:00
Christoffer Dall
875656fe0c KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_regs
Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_set_regs().

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
Reviewed-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:26:52 +01:00
Christoffer Dall
1fc9b76b3d KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_regs
Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_get_regs().

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
Reviewed-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:26:51 +01:00
Christoffer Dall
accb757d79 KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_run
Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_run().

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
Reviewed-by: Christian Borntraeger <borntraeger@de.ibm.com> # s390 parts
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
[Rebased. - Paolo]
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:26:50 +01:00
Christoffer Dall
8a32dd60ec KVM: Prepare for moving vcpu_load/vcpu_put into arch specific code
In preparation for moving calls to vcpu_load() and vcpu_put() into the
architecture specific implementations of the KVM vcpu ioctls, move the
calls in the main kvm_vcpu_ioctl() dispatcher function to each case
of the ioctl select statement.  This allows us to move the vcpu_load()
and vcpu_put() calls into architecture specific implementations of vcpu
ioctls, one by one.

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:26:50 +01:00
Christoffer Dall
ec7660ccdd KVM: Take vcpu->mutex outside vcpu_load
As we're about to call vcpu_load() from architecture-specific
implementations of the KVM vcpu ioctls, but yet we access data
structures protected by the vcpu->mutex in the generic code, factor
this logic out from vcpu_load().

x86 is the only architecture which calls vcpu_load() outside of the main
vcpu ioctl function, and these calls will no longer take the vcpu mutex
following this patch.  However, with the exception of
kvm_arch_vcpu_postcreate (see below), the callers are either in the
creation or destruction path of the VCPU, which means there cannot be
any concurrent access to the data structure, because the file descriptor
is not yet accessible, or is already gone.

kvm_arch_vcpu_postcreate makes the newly created vcpu potentially
accessible by other in-kernel threads through the kvm->vcpus array, and
we therefore take the vcpu mutex in this case directly.

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:26:49 +01:00
Quan Xu
8eb73e2d41 KVM: VMX: drop I/O permission bitmaps
Since KVM removes the only I/O port 0x80 bypass on Intel hosts,
clear CPU_BASED_USE_IO_BITMAPS and set CPU_BASED_UNCOND_IO_EXITING
bit. Then these I/O permission bitmaps are not used at all, so
drop I/O permission bitmaps.

Signed-off-by: Jim Mattson <jmattson@google.com>
Signed-off-by: Radim KrÄmář <rkrcmar@redhat.com>
Signed-off-by: Quan Xu <quan.xu0@gmail.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:26:48 +01:00
Wanpeng Li
9c48d517ce KVM: X86: Reduce the overhead when lapic_timer_advance is disabled
When I run ebizzy in a 32 vCPUs guest on a 32 pCPUs Xeon box, I can observe
~8000 kvm_wait_lapic_expire CurAvg/s through kvm_stat tool even if the advance
tscdeadline hrtimer expiration is disabled. Each call to wait_lapic_expire()
will consume ~70 cycles when a timer fires since apic_timer_expire() will
set expired_tscdeadline and then wait_lapic_expire() will do some caculation
before bailing out. So total ~175us per second is lost on this 3.2Ghz machine.
This patch reduces the overhead by skipping the function wait_lapic_expire()
when lapic_timer_advance is disabled.

Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Radim Krčmář <rkrcmar@redhat.com>
Signed-off-by: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:26:48 +01:00
Wanpeng Li
74c55931c7 KVM: VMX: Cache IA32_DEBUGCTL in memory
MSR_IA32_DEBUGCTLMSR is zeroed on VMEXIT, so it is saved/restored each
time during world switch.  This patch caches the host IA32_DEBUGCTL MSR
and saves/restores the host IA32_DEBUGCTL msr when guest/host switches
to avoid to save/restore each time during world switch.  This saves
about 100 clock cycles per vmexit.

Suggested-by: Jim Mattson <jmattson@google.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Radim Krčmář <rkrcmar@redhat.com>
Cc: Jim Mattson <jmattson@google.com>
Signed-off-by: Wanpeng Li <wanpeng.li@hotmail.com>
Reviewed-by: Jim Mattson <jmattson@google.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
2017-12-14 09:26:47 +01:00
Mark Kanda
276c796cfe KVM: nVMX: Add a WARN for freeing a loaded VMCS02
When attempting to free a loaded VMCS02, add a WARN and avoid
freeing it (to avoid use-after-free situations).

Suggested-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Mark Kanda <mark.kanda@oracle.com>
Reviewed-by: Ameya More <ameya.more@oracle.com>
Reviewed-by: Krish Sadhukhan <krish.sadhukhan@oracle.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
2017-12-14 09:26:46 +01:00
Jim Mattson
00647b4494 KVM: nVMX: Eliminate vmcs02 pool
The potential performance advantages of a vmcs02 pool have never been
realized. To simplify the code, eliminate the pool. Instead, a single
vmcs02 is allocated per VCPU when the VCPU enters VMX operation.

Signed-off-by: Jim Mattson <jmattson@google.com>
Signed-off-by: Mark Kanda <mark.kanda@oracle.com>
Reviewed-by: Ameya More <ameya.more@oracle.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
2017-12-14 09:26:46 +01:00
Yang Zhong
80fef315a7 KVM: Expose new cpu features to guest
Intel IceLake cpu has added new cpu features,AVX512_VBMI2/GFNI/
VAES/VPCLMULQDQ/AVX512_VNNI/AVX512_BITALG. Those new cpu features
need expose to guest VM.

The bit definition:
CPUID.(EAX=7,ECX=0):ECX[bit 06] AVX512_VBMI2
CPUID.(EAX=7,ECX=0):ECX[bit 08] GFNI
CPUID.(EAX=7,ECX=0):ECX[bit 09] VAES
CPUID.(EAX=7,ECX=0):ECX[bit 10] VPCLMULQDQ
CPUID.(EAX=7,ECX=0):ECX[bit 11] AVX512_VNNI
CPUID.(EAX=7,ECX=0):ECX[bit 12] AVX512_BITALG

The release document ref below link:
https://software.intel.com/sites/default/files/managed/c5/15/\
architecture-instruction-set-extensions-programming-reference.pdf

The kernel dependency commit in kvm.git:
(c128dbfa0f)

Signed-off-by: Yang Zhong <yang.zhong@intel.com>
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
2017-12-14 09:26:45 +01:00
Markus Elfring
7858833506 kvm_main: Use common error handling code in kvm_dev_ioctl_create_vm()
Add a jump target so that a bit of exception handling can be better reused
at the end of this function.

This issue was detected by using the Coccinelle software.

Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
2017-12-14 09:26:44 +01:00
Liran Alon
52797bf9a8 KVM: x86: Add emulation of MSR_SMI_COUNT
This MSR returns the number of #SMIs that occurred on CPU since
boot.

It was seen to be used frequently by ESXi guest.

Patch adds a new vcpu-arch specific var called smi_count to
save the number of #SMIs which occurred on CPU since boot.
It is exposed as a read-only MSR to guest (causing #GP
on wrmsr) in RDMSR/WRMSR emulation code.
MSR_SMI_COUNT is also added to emulated_msrs[] to make sure
user-space can save/restore it for migration purposes.

Signed-off-by: Liran Alon <liran.alon@oracle.com>
Suggested-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Nikita Leshenko <nikita.leshchenko@oracle.com>
Reviewed-by: Bhavesh Davda <bhavesh.davda@oracle.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
2017-12-14 09:26:44 +01:00
Radim Krčmář
431f5d4443 KVM: x86: simplify kvm_mwait_in_guest()
If Intel/AMD implements MWAIT, we expect that it works well and only
reject known bugs;  no reason to do it the other way around for minor
vendors.  (Not that they are relevant ATM.)

This allows further simplification of kvm_mwait_in_guest().
And use boot_cpu_has() instead of "cpu_has(&boot_cpu_data," while at it.

Reviewed-by: Alexander Graf <agraf@suse.de>
Acked-by: Borislav Petkov <bp@suse.de>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:26:43 +01:00
Radim Krčmář
346f48fa31 KVM: x86: drop bogus MWAIT check
The check was added in some iteration while trying to fix a reported OS
X on Core 2 bug, but that bug is elsewhere.

The comment is misleading because the guest can call MWAIT with ECX = 0
even if we enforce CPUID5_ECX_INTERRUPT_BREAK;  the call would have the
exactly the same effect as if the host didn't have the feature.

A problem is that a QEMU feature exposes CPUID5_ECX_INTERRUPT_BREAK on
CPUs that do not support it.  Removing the check changes behavior on
last Pentium 4 lines (Presler, Dempsey, and Tulsa, which had VMX and
MONITOR while missing INTERRUPT_BREAK) when running a guest OS that uses
MWAIT without checking for its presence (QEMU doesn't expose MONITOR).

The only known OS that ignores the MONITOR flag is old Mac OS X and we
allowed it to bug on Core 2 (MWAIT used to throw #UD and only that OS
noticed), so we can save another 20 lines letting it bug on even older
CPUs.  Alternatively, we can return MWAIT exiting by default and let
userspace toggle it.

Reviewed-by: Alexander Graf <agraf@suse.de>
Acked-by: Borislav Petkov <bp@suse.de>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:26:43 +01:00
Radim Krčmář
2a140f3b6e KVM: x86: prevent MWAIT in guest with buggy MONITOR
The bug prevents MWAIT from waking up after a write to the monitored
cache line.
KVM might emulate a CPU model that shouldn't have the bug, so the guest
would not employ a workaround and possibly miss wakeups.
Better to avoid the situation.

Reviewed-by: Alexander Graf <agraf@suse.de>
Acked-by: Borislav Petkov <bp@suse.de>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:26:42 +01:00
Colin Ian King
0d37d26f11 KVM: x86: MMU: make array audit_point_name static
The array audit_point_name is local to the source and does not need to
be in global scope, so make it static.

Cleans up sparse warning:
arch/x86/kvm/mmu_audit.c:22:12: warning: symbol 'audit_point_name' was
not declared. Should it be static?

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:26:41 +01:00
Gimcuan Hui
858ac87fc2 x86: kvm: mmu: make kvm_mmu_clear_all_pte_masks static
The kvm_mmu_clear_all_pte_masks interface is only used by kvm_mmu_module_init
locally, and does not need to be called by other module, make it static.

This patch cleans up sparse warning:
symbol 'kvm_mmu_clear_all_pte_masks' was not declared. Should it be static?

Signed-off-by: Gimcuan Hui <gimcuan@gmail.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:26:41 +01:00
Paolo Bonzini
fb6d4d340e KVM: x86: emulate RDPID
This is encoded as F3 0F C7 /7 with a register argument.  The register
argument is the second array in the group9 GroupDual, while F3 is the
fourth element of a Prefix.

Reviewed-by: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:26:40 +01:00
Paolo Bonzini
0367f205a3 KVM: vmx: add support for emulating UMIP
UMIP can be emulated almost perfectly on Intel processor by enabling
descriptor-table exits.  SMSW does not cause a vmexit and hence it
cannot be changed into a #GP fault, but all in all it's the most
"innocuous" of the unprivileged instructions that UMIP blocks.

In fact, Linux is _also_ emulating SMSW instructions on behalf of the
program that executes them, because some 16-bit programs expect to use
SMSW to detect vm86 mode, so this is an even smaller issue.

Reviewed-by: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:26:40 +01:00
Paolo Bonzini
66336cab35 KVM: x86: add support for emulating UMIP
The User-Mode Instruction Prevention feature present in recent Intel
processor prevents a group of instructions (sgdt, sidt, sldt, smsw, and
str) from being executed with CPL > 0. Otherwise, a general protection
fault is issued.

UMIP instructions in general are also able to trigger vmexits, so we can
actually emulate UMIP on older processors.  This commit sets up the
infrastructure so that kvm-intel.ko and kvm-amd.ko can set the UMIP
feature bit for CPUID even if the feature is not actually available
in hardware.

Reviewed-by: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:26:39 +01:00
Paolo Bonzini
dd307d017b KVM: x86: emulate sldt and str
These are needed to handle the descriptor table vmexits when emulating
UMIP.

Reviewed-by: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:26:38 +01:00
Paolo Bonzini
ae3e61e1c2 KVM: x86: add support for UMIP
Add the CPUID bits, make the CR4.UMIP bit not reserved anymore, and
add UMIP support for instructions that are already emulated by KVM.

Reviewed-by: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:26:38 +01:00
Stefan Raspl
cf656c7661 tools/kvm_stat: add line for totals
Add a line for the total number of events and current average at the
bottom of the body.
Note that both values exclude child trace events. I.e. if drilldown is
activated via interactive command 'x', only the totals are accounted, or
we'd be counting these twice (see previous commit "tools/kvm_stat: fix
child trace events accounting").

Signed-off-by: Stefan Raspl <raspl@linux.vnet.ibm.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:25:47 +01:00
Stefan Raspl
73fab6ffbd tools/kvm_stat: stop ignoring unhandled arguments
Unhandled arguments, which could easily include typos, are simply
ignored. We should be strict to avoid undetected typos.
To reproduce start kvm_stat with an extra argument, e.g.
'kvm_stat -d bnuh5ol' and note that this will actually work.

Signed-off-by: Stefan Raspl <raspl@linux.vnet.ibm.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:25:46 +01:00
Stefan Raspl
822cfe3e48 tools/kvm_stat: suppress usage information on command line errors
Errors while parsing the '-g' command line argument result in display of
usage information prior to the error message. This is a bit confusing,
as the command line is syntactically correct.
To reproduce, run 'kvm_stat -g' and specify a non-existing or inactive
guest.

Signed-off-by: Stefan Raspl <raspl@linux.vnet.ibm.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:25:46 +01:00
Stefan Raspl
08e20a6300 tools/kvm_stat: handle invalid regular expressions
Passing an invalid regular expression on the command line results in a
traceback. Note that interactive specification of invalid regular
expressions is not affected
To reproduce, run "kvm_stat -f '*'".

Signed-off-by: Stefan Raspl <raspl@linux.vnet.ibm.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:25:45 +01:00
Stefan Raspl
f3d11b0e86 tools/kvm_stat: add hint on '-f help' to man page
The man page update for this new functionality was omitted.

Signed-off-by: Stefan Raspl <raspl@linux.vnet.ibm.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:25:44 +01:00
Stefan Raspl
fff8c9eb48 tools/kvm_stat: fix child trace events accounting
Child trace events were included in calculation of the overall total,
which is used for calculation of the percentages of the '%Total' column.
However, the parent trace envents' stats summarize the child trace
events, hence we'd incorrectly account for them twice, leading to
slightly wrong stats.
With this fix, we use the correct total. Consequently, the sum of the
child trace events' '%Total' column values is identical to the
respective value of the respective parent event. However, this also
means that the sum of the '%Total' column values will aggregate to more
than 100 percent.

Signed-off-by: Stefan Raspl <raspl@linux.vnet.ibm.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:25:44 +01:00
Stefan Raspl
b74faa930d tools/kvm_stat: fix extra handling of 'help' with fields filter
Commit 67fbcd62f5 ("tools/kvm_stat: add '-f help' to get the available
event list") added support for '-f help'. However, the extra handling of
'help' will also take effect when 'help' is specified as a regex in
interactive mode via 'f'. This results in display of all events while
only those matching this regex should be shown.

Signed-off-by: Stefan Raspl <raspl@linux.vnet.ibm.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:25:43 +01:00
Stefan Raspl
67c162b089 tools/kvm_stat: fix missing field update after filter change
When updating the fields filter, tracepoint events of fields previously
not visible were not enabled, as TracepointProvider.update_fields()
updated the member variable directly instead of using the setter, which
triggers the event enable/disable.
To reproduce, run 'kvm_stat -f kvm_exit', press 'c' to remove the
filter, and notice that no add'l fields that do not match the regex
'kvm_exit' will appear.
This issue was introduced by commit c469117df0 ("tools/kvm_stat:
simplify initializers").

Signed-off-by: Stefan Raspl <raspl@linux.vnet.ibm.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:25:42 +01:00
Stefan Raspl
faa0665041 tools/kvm_stat: fix drilldown in events-by-guests mode
When displaying debugfs events listed by guests, an attempt to switch to
reporting of stats for individual child trace events results in garbled
output. Reason is that when toggling drilldown, the update of the stats
doesn't honor when events are displayed by guests, as indicated by
Tui._display_guests.
To reproduce, run 'kvm_stat -d' and press 'b' followed by 'x'.

Signed-off-by: Stefan Raspl <raspl@linux.vnet.ibm.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:25:42 +01:00
Stefan Raspl
19e8e54f43 tools/kvm_stat: fix command line option '-g'
Specifying a guest via '-g foo' always results in an error:
  $ kvm_stat -g foo
  Usage: kvm_stat [options]

  kvm_stat: error: Error while searching for guest "foo", use "-p" to
  specify a pid instead

Reason is that Tui.get_pid_from_gname() is not static, as it is supposed
to be.

Signed-off-by: Stefan Raspl <raspl@linux.vnet.ibm.com>
Tested-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:25:41 +01:00
Peter Xu
5663d8f9bb kvm: x86: fix WARN due to uninitialized guest FPU state
------------[ cut here ]------------
 Bad FPU state detected at kvm_put_guest_fpu+0xd8/0x2d0 [kvm], reinitializing FPU registers.
 WARNING: CPU: 1 PID: 4594 at arch/x86/mm/extable.c:103 ex_handler_fprestore+0x88/0x90
 CPU: 1 PID: 4594 Comm: qemu-system-x86 Tainted: G    B      OE    4.15.0-rc2+ #10
 RIP: 0010:ex_handler_fprestore+0x88/0x90
 Call Trace:
  fixup_exception+0x4e/0x60
  do_general_protection+0xff/0x270
  general_protection+0x22/0x30
 RIP: 0010:kvm_put_guest_fpu+0xd8/0x2d0 [kvm]
 RSP: 0018:ffff8803d5627810 EFLAGS: 00010246
  kvm_vcpu_reset+0x3b4/0x3c0 [kvm]
  kvm_apic_accept_events+0x1c0/0x240 [kvm]
  kvm_arch_vcpu_ioctl_run+0x1658/0x2fb0 [kvm]
  kvm_vcpu_ioctl+0x479/0x880 [kvm]
  do_vfs_ioctl+0x142/0x9a0
  SyS_ioctl+0x74/0x80
  do_syscall_64+0x15f/0x600

where kvm_put_guest_fpu is called without a prior kvm_load_guest_fpu.
To fix it, move kvm_load_guest_fpu to the very beginning of
kvm_arch_vcpu_ioctl_run.

Cc: stable@vger.kernel.org
Fixes: f775b13eed
Signed-off-by: Peter Xu <peterx@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:24:35 +01:00
Wanpeng Li
d73235d17b KVM: X86: Fix load RFLAGS w/o the fixed bit
*** Guest State ***
 CR0: actual=0x0000000000000030, shadow=0x0000000060000010, gh_mask=fffffffffffffff7
 CR4: actual=0x0000000000002050, shadow=0x0000000000000000, gh_mask=ffffffffffffe871
 CR3 = 0x00000000fffbc000
 RSP = 0x0000000000000000  RIP = 0x0000000000000000
 RFLAGS=0x00000000         DR7 = 0x0000000000000400
        ^^^^^^^^^^

The failed vmentry is triggered by the following testcase when ept=Y:

    #include <unistd.h>
    #include <sys/syscall.h>
    #include <string.h>
    #include <stdint.h>
    #include <linux/kvm.h>
    #include <fcntl.h>
    #include <sys/ioctl.h>

    long r[5];
    int main()
    {
    	r[2] = open("/dev/kvm", O_RDONLY);
    	r[3] = ioctl(r[2], KVM_CREATE_VM, 0);
    	r[4] = ioctl(r[3], KVM_CREATE_VCPU, 7);
    	struct kvm_regs regs = {
    		.rflags = 0,
    	};
    	ioctl(r[4], KVM_SET_REGS, &regs);
    	ioctl(r[4], KVM_RUN, 0);
    }

X86 RFLAGS bit 1 is fixed set, userspace can simply clearing bit 1
of RFLAGS with KVM_SET_REGS ioctl which results in vmentry fails.
This patch fixes it by oring X86_EFLAGS_FIXED during ioctl.

Cc: stable@vger.kernel.org
Suggested-by: Jim Mattson <jmattson@google.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Quan Xu <quan.xu0@gmail.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Radim Krčmář <rkrcmar@redhat.com>
Cc: Jim Mattson <jmattson@google.com>
Cc: stable@vger.kernel.org
Signed-off-by: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:24:26 +01:00
Wanpeng Li
ed52870f46 KVM: MMU: Fix infinite loop when there is no available mmu page
The below test case can cause infinite loop in kvm when ept=0.

    #include <unistd.h>
    #include <sys/syscall.h>
    #include <string.h>
    #include <stdint.h>
    #include <linux/kvm.h>
    #include <fcntl.h>
    #include <sys/ioctl.h>

    long r[5];
    int main()
    {
    	r[2] = open("/dev/kvm", O_RDONLY);
    	r[3] = ioctl(r[2], KVM_CREATE_VM, 0);
    	r[4] = ioctl(r[3], KVM_CREATE_VCPU, 7);
    	ioctl(r[4], KVM_RUN, 0);
    }

It doesn't setup the memory regions, mmu_alloc_shadow/direct_roots() in
kvm return 1 when kvm fails to allocate root page table which can result
in beblow infinite loop:

    vcpu_run() {
    	for (;;) {
	    	r = vcpu_enter_guest()::kvm_mmu_reload() returns 1
	    	if (r <= 0)
	    		break;
	    	if (need_resched())
	    		cond_resched();
      }
    }

This patch fixes it by returning -ENOSPC when there is no available kvm mmu
page for root page table.

Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Radim Krčmář <rkrcmar@redhat.com>
Cc: stable@vger.kernel.org
Fixes: 26eeb53cf0 (KVM: MMU: Bail out immediately if there is no available mmu page)
Signed-off-by: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2017-12-14 09:24:14 +01:00
Linus Torvalds
50c4c4e268 Linux 4.15-rc3 2017-12-10 17:56:26 -08:00
Jeff Layton
98087c05b9 hpfs: don't bother with the i_version counter or f_version
HPFS does not set SB_I_VERSION and does not use the i_version counter
internally.

Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Mikulas Patocka <mikulas@twibright.com>
Reviewed-by: Mikulas Patocka <mikulas@twibright.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2017-12-10 12:58:18 -08:00
Jiri Slaby
d70ef22892 futex: futex_wake_op, fix sign_extend32 sign bits
sign_extend32 counts the sign bit parameter from 0, not from 1.  So we
have to use "11" for 12th bit, not "12".

This mistake means we have not allowed negative op and cmp args since
commit 30d6e0a419 ("futex: Remove duplicated code and fix undefined
behaviour") till now.

Fixes: 30d6e0a419 ("futex: Remove duplicated code and fix undefined behaviour")
Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Darren Hart <dvhart@infradead.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2017-12-10 12:50:57 -08:00
Linus Torvalds
51090c5d6d for-4.15-rc3-tag
-----BEGIN PGP SIGNATURE-----
 
 iQIzBAABCgAdFiEE8rQSAMVO+zA4DBdWxWXV+ddtWDsFAlosIf4ACgkQxWXV+ddt
 WDspsw//YPhztOkAM7L37Lcv6PuMIBm7AsZax+iUctx9GlE9Yb9dYX+yIGjk3N44
 M6oHANP/Af70lGn3jaNlH+BeQre+RFD2KnT+Yyvp/0DV5+v+Bb6wqzrVqeYf9NIr
 lf6yc925gX10+DM6UXpYopTmdB8zXXO8xnqmFuT1jC/PrW/g+Hpxi7UtFFcoXwnE
 uucdih1LnNC/2pwp4ygQAxMkLnU2foWRsEP9lqsv83ecKDBfVxHUidzEZLTO7L+c
 ePc74AcyuPZ7DobuSDyDF4e0Ru5YtY5Zf+KR7RZHag5BNF2YLJE/XtN+hd3YhOQA
 7VniaPzUEG74ukvkL3L2oqxrMEavE0IFJtmzT4CM8DlRsGsDnn5n45sGHfo5clr8
 33XOq8aiGtbG1vwVbBJOuNQI2SWJxwe1OyAZoV/o1UVrltSCRf+dYL8Yf3IO2K0M
 DRnRNqEcZQGfqrVO5Iblw7VzVqY9LKiRESScS0Btvrys+DTVZAgC9CJDwN446E5v
 i56PrmT8OcC9MzP9wFIZtg27jiC0ndNwkqUhFrt1LBvC+BtvZvshAnFLhLfSRyZo
 0gqp2GoP6CFaUd5Ok+osALWF2VG8cpMJ7urdX0O5zXEYKioLwiXUS9Z7sldfHsJr
 Uiy1uh70UIOM96ZcsXyjLr0LO5vmgkV2kyDNbR5DtrJhfFai4Gs=
 =YaZE
 -----END PGP SIGNATURE-----

Merge tag 'for-4.15-rc3-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux

Pull btrfs fixes from David Sterba:
 "This contains a few fixes (error handling, quota leak, FUA vs
  nobarrier mount option).

  There's one one worth mentioning separately - an off-by-one fix that
  leads to overwriting first byte of an adjacent page with 0, out of
  bounds of the memory allocated by an ioctl. This is under a privileged
  part of the ioctl, can be triggerd in some subvolume layouts"

* tag 'for-4.15-rc3-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux:
  btrfs: Fix possible off-by-one in btrfs_search_path_in_tree
  Btrfs: disable FUA if mounted with nobarrier
  btrfs: fix missing error return in btrfs_drop_snapshot
  btrfs: handle errors while updating refcounts in update_ref_for_cow
  btrfs: Fix quota reservation leak on preallocated files
2017-12-10 08:30:04 -08:00
Linus Torvalds
9c02e0601b ARM: SoC fixes for 4.15-rc
ARM SoC fixes for this merge window:
 
  - A revert of all SCPI changes from the 4.15 merge window. They had
    regressions on the Amlogic platforms, and the submaintainer isn't
    around to fix these bugs due to vacation, etc. So we agreed to revert
    and revisit in next release cycle.
 
  - A series fixing a number of bugs for ARM CCN interconnect, around
    module unload, smp_processor_id() in preemptable context, and fixing
    some memory allocation failure checks.
 
  - A handful of devicetree fixes for different platforms, fixing
    warnings and errors that were previously ignored by the compiler.
 
  - The usual set of mostly minor fixes for different platforms.
 -----BEGIN PGP SIGNATURE-----
 
 iQJDBAABCAAtFiEElf+HevZ4QCAJmMQ+jBrnPN6EHHcFAloswBYPHG9sb2ZAbGl4
 b20ubmV0AAoJEIwa5zzehBx3bFoP/R6fAth5zVuUveykTbHYSs1Oz5XsLw4X77mA
 vg9Y2bBiz82/R7/weISxn28iXd29FQk0inbohI3bGsqJ+OInKiBNJ4UDrvuvrG39
 F2ifH7bjg6x9aooXx0VRXdECckO6klVhe8kjRovRZv712bWUA5uRG/6YUxnyC5LX
 6smrhFUAhj5StUqaY8rFlF8Pob9ftee0aM8/C3LAggQBhwqU5RXY8HgL0jLb919q
 ewOuDgO3FtfrigtOlDWkrQLRe+sY2b5D97Z4amIe4rojqvD6i7grRFBfkfDb4gR2
 7Vc0FZmk7OaaOLsLChv8H8atCxpDJ6KPpg5NAfXk4KM2kbWLbinkLPMbXxSXB4bC
 Q26PPZhKP1OupsSl/9fewHbZeaMZSY0kQXHxBIXCtyMV110TsLVmdFTksbV0gVKk
 x0lBAttX2RZiRU0bFxOHACEXnRTS4/uDCxBvxh5othQZg5EqAEqcFoium2psMvsJ
 XzFPNetyqPE2KhOUMfwUbOtWs2js5E4HaxwyF9xduw32g10NxtRTKGUMeDrxdtW3
 dHQZMr33yLkSkk08S4LMdwUrlONieyJZYeVZ2Jxd+Lv4bn4ZX8mDbLU24BCtLwUI
 wR+8zS2YSVfer/t5JC+dIJzKAiufeOFovNurlsoFJsE2yNLfekZcZ/x63ZdVQnve
 8f0mibwa
 =DIHP
 -----END PGP SIGNATURE-----

Merge tag 'armsoc-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc

Pull ARM SoC fixes from Olof Johansson:

 - A revert of all SCPI changes from the 4.15 merge window. They had
   regressions on the Amlogic platforms, and the submaintainer isn't
   around to fix these bugs due to vacation, etc. So we agreed to revert
   and revisit in next release cycle.

 - A series fixing a number of bugs for ARM CCN interconnect, around
   module unload, smp_processor_id() in preemptable context, and fixing
   some memory allocation failure checks.

 - A handful of devicetree fixes for different platforms, fixing
   warnings and errors that were previously ignored by the compiler.

 - The usual set of mostly minor fixes for different platforms.

* tag 'armsoc-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc: (42 commits)
  ARM64: dts: meson-gx: fix UART pclk clock name
  ARM: omap2: hide omap3_save_secure_ram on non-OMAP3 builds
  arm: dts: nspire: Add missing #phy-cells to usb-nop-xceiv
  ARM: dts: Fix dm814x missing phy-cells property
  ARM: dts: Fix elm interrupt compiler warning
  bus: arm-ccn: fix module unloading Error: Removing state 147 which has instances left.
  bus: arm-cci: Fix use of smp_processor_id() in preemptible context
  bus: arm-ccn: Fix use of smp_processor_id() in preemptible context
  bus: arm-ccn: Simplify code
  bus: arm-ccn: Check memory allocation failure
  bus: arm-ccn: constify attribute_group structures.
  firmware: arm_scpi: Revert updates made during v4.15 merge window
  arm: dts: marvell: Add missing #phy-cells to usb-nop-xceiv
  arm64: dts: sort vendor subdirectories in Makefile alphabetically
  meson-gx-socinfo: Fix package id parsing
  ARM: meson: fix spelling mistake: "Couln't" -> "Couldn't"
  ARM: dts: meson: fix the memory region of the GPIO interrupt controller
  ARM: dts: meson: correct the sort order for the the gpio_intc node
  MAINTAINERS: exclude other Socionext SoC DT files from ARM/UNIPHIER entry
  arm64: dts: uniphier: remove unnecessary interrupt-parent
  ...
2017-12-10 08:26:59 -08:00