Before introducing a pmu.h header for them, make the naming more
consistent.
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

Based on Intel's SDM, mapping huge page which do not have consistent
memory cache for each 4k page will cause undefined behavior

In order to avoiding this kind of undefined behavior, we force to use
4k pages under this case
Signed-off-by: NXiao Guangrong <guangrong.xiao@linux.intel.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

mtrr_for_each_mem_type() is ready now, use it to simplify
kvm_mtrr_get_guest_memory_type()
Signed-off-by: NXiao Guangrong <guangrong.xiao@linux.intel.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

It walks all MTRRs and gets all the memory cache type setting for the
specified range also it checks if the range is fully covered by MTRRs
Signed-off-by: NXiao Guangrong <guangrong.xiao@linux.intel.com>
[Adjust for range_size->range_shift change. - Paolo]
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

Two functions are introduced:
- fixed_mtrr_addr_to_seg() translates the address to the fixed
  MTRR segment

- fixed_mtrr_addr_seg_to_range_index() translates the address to
  the index of kvm_mtrr.fixed_ranges[]

They will be used in the later patch
Signed-off-by: NXiao Guangrong <guangrong.xiao@linux.intel.com>
[Adjust for range_size->range_shift change. - Paolo]
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

Sort all valid variable MTRRs based on its base address, it will help us to
check a range to see if it's fully contained in variable MTRRs
Signed-off-by: NXiao Guangrong <guangrong.xiao@linux.intel.com>
[Fix list insertion sort, simplify var_mtrr_range_is_valid to just
 test the V bit. - Paolo]
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

It gets the range for the specified variable MTRR
Signed-off-by: NXiao Guangrong <guangrong.xiao@linux.intel.com>
[Simplify boolean operations. - Paolo]
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

This table summarizes the information of fixed MTRRs and introduce some APIs
to abstract its operation which helps us to clean up the code and will be
used in later patches
Signed-off-by: NXiao Guangrong <guangrong.xiao@linux.intel.com>
[Change range_size to range_shift, in order to avoid udivdi3 errors.
 - Paolo]
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

 - kvm_mtrr_get_guest_memory_type() only checks one page in MTRRs so
   that it's unnecessary to check to see if the range is partially
   covered in MTRR

 - optimize the check of overlap memory type and add some comments
   to explain the precedence
Signed-off-by: NXiao Guangrong <guangrong.xiao@linux.intel.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

Variable MTRR MSRs are 64 bits which are directly accessed with full length,
no reason to split them to two 32 bits
Signed-off-by: NXiao Guangrong <guangrong.xiao@linux.intel.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

Drop kvm_mtrr->enable, omit the decode/code workload and get rid of
all the hard code
Signed-off-by: NXiao Guangrong <guangrong.xiao@linux.intel.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

vMTRR does not depend on any host MTRR feature and fixed MTRRs have always
been implemented, so drop this field
Signed-off-by: NXiao Guangrong <guangrong.xiao@linux.intel.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

MSR_MTRRcap is a MTRR msr so move the handler to the common place, also
add some comments to make the hard code more readable
Signed-off-by: NXiao Guangrong <guangrong.xiao@linux.intel.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

MTRR code locates in x86.c and mmu.c so that move them to a separate file to
make the organization more clearer and it will be the place where we fully
implement vMTRR
Signed-off-by: NXiao Guangrong <guangrong.xiao@linux.intel.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

Currently, CR0.CD is not checked when we virtualize memory cache type for
noncoherent_dma guests, this patch fixes it by :

- setting UC for all memory if CR0.CD = 1
- zapping all the last sptes in MMU if CR0.CD is changed
Signed-off-by: NXiao Guangrong <guangrong.xiao@linux.intel.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

If hardware doesn't support DecodeAssist - a feature that provides
more information about the intercept in the VMCB, KVM decodes the
instruction and then updates the next_rip vmcb control field.
However, NRIP support itself depends on cpuid Fn8000_000A_EDX[NRIPS].
Since skip_emulated_instruction() doesn't verify nrip support
before accepting control.next_rip as valid, avoid writing this
field if support isn't present.
Signed-off-by: NBandan Das <bsd@redhat.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

Follow up to commit e194bbdf.
Suggested-by: NBandan Das <bsd@redhat.com>
Suggested-by: NAlex Williamson <alex.williamson@redhat.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

... and we're done. :)

Because SMBASE is usually relocated above 1M on modern chipsets, and
SMM handlers might indeed rely on 4G segment limits, we only expose it
if KVM is able to run the guest in big real mode.  This includes any
of VMX+emulate_invalid_guest_state, VMX+unrestricted_guest, or SVM.
Reviewed-by: NRadim Krčmář <rkrcmar@redhat.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

This is now very simple to do.  The only interesting part is a simple
trick to find the right memslot in gfn_to_rmap, retrieving the address
space from the spte role word.  The same trick is used in the auditing
code.

The comment on top of union kvm_mmu_page_role has been stale forever,
so remove it.  Speaking of stale code, remove pad_for_nice_hex_output
too: it was splitting the "access" bitfield across two bytes and thus
had effectively turned into pad_for_ugly_hex_output.
Reviewed-by: NRadim Krčmář <rkrcmar@redhat.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

This patch has no semantic change, but it prepares for the introduction
of a second address space for system management mode.

A new function x86_set_memory_region (and the "slots_lock taken"
counterpart __x86_set_memory_region) is introduced in order to
operate on all address spaces when adding or deleting private
memory slots.
Reviewed-by: NRadim Krčmář <rkrcmar@redhat.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

We need to hide SMRAM from guests not running in SMM.  Therefore,
all uses of kvm_read_guest* and kvm_write_guest* must be changed to
check whether the VCPU is in system management mode and use a
different set of memslots.  Switch from kvm_* to the newly-introduced
kvm_vcpu_*, which call into kvm_arch_vcpu_memslots_id.
Reviewed-by: NRadim Krčmář <rkrcmar@redhat.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

This is always available (with one exception in the auditing code),
and with the same auditing exception the level was coming from
sp->role.level.

Later, the spte's role will also be used to look up the right memslots
array.
Reviewed-by: NRadim Krčmář <rkrcmar@redhat.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

The big ugly one.  This patch adds support for switching in and out of
system management mode, respectively upon receiving KVM_REQ_SMI and upon
executing a RSM instruction.  Both 32- and 64-bit formats are supported
for the SMM state save area.
Reviewed-by: NRadim Krčmář <rkrcmar@redhat.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

Do not process INITs immediately while in system management mode, keep
it instead in apic->pending_events.  Tell userspace if an INIT is
pending when they issue GET_VCPU_EVENTS, and similarly handle the
new field in SET_VCPU_EVENTS.

Note that the same treatment should be done while in VMX non-root mode.
Reviewed-by: NRadim Krčmář <rkrcmar@redhat.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

This patch adds the interface between x86.c and the emulator: the
SMBASE register, a new emulator flag, the RSM instruction.  It also
adds a new request bit that will be used by the KVM_SMI ioctl.
Reviewed-by: NRadim Krčmář <rkrcmar@redhat.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

This patch includes changes to the external API for SMM support.
Userspace can predicate the availability of the new fields and
ioctls on a new capability, KVM_CAP_X86_SMM, which is added at the end
of the patch series.
Reviewed-by: NRadim Krčmář <rkrcmar@redhat.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

The hflags field will contain information about system management mode
and will be useful for the emulator.  Pass the entire field rather than
just the guest-mode information.
Reviewed-by: NRadim Krčmář <rkrcmar@redhat.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

SMBASE is only readable from SMM for the VCPU, but it must be always
accessible if userspace is accessing it.  Thus, all functions that
read MSRs are changed to accept a struct msr_data; the host_initiated
and index fields are pre-initialized, while the data field is filled
on return.
Reviewed-by: NRadim Krčmář <rkrcmar@redhat.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

We will want to filter away MSR_IA32_SMBASE from the emulated_msrs if
the host CPU does not support SMM virtualization.  Introduce the
logic to do that, and also move paravirt MSRs to emulated_msrs for
simplicity and to get rid of KVM_SAVE_MSRS_BEGIN.
Reviewed-by: NRadim Krčmář <rkrcmar@redhat.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

This was noticed by Radim while reviewing the implementation of
system management mode.
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

Malicious (or egregiously buggy) userspace can trigger it, but it
should never happen in normal operation.
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

VFIO has proved itself a much better option than KVM's built-in
device assignment.  It is mature, provides better isolation because
it enforces ACS, and even the userspace code is being tested on
a wider variety of hardware these days than the legacy support.

Disable legacy device assignment by default.
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

Initialize kvmclock base, on kvmclock system MSR write time,
so that the guest sees kvmclock counting from zero.

This matches baremetal behaviour when kvmclock in guest
sets sched clock stable.
Signed-off-by: NMarcelo Tosatti <mtosatti@redhat.com>
[Remove unnecessary comment. - Paolo]
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

arch/x86/kvm/mmu.c: In function 'kvm_mmu_pte_write':
arch/x86/kvm/mmu.c:4256: error: unknown field 'cr0_wp' specified in initializer
arch/x86/kvm/mmu.c:4257: error: unknown field 'cr4_pae' specified in initializer
arch/x86/kvm/mmu.c:4257: warning: excess elements in union initializer
...

gcc-4.4.4 (at least) has issues when using anonymous unions in
initializers.

Fixes: edc90b7d ("KVM: MMU: fix SMAP virtualization")
Cc: Xiao Guangrong <guangrong.xiao@linux.intel.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

There is no reason to deny this feature to guests. We are emulating the
APIC timer, thus are exposing it without stops in power-saving states.
Signed-off-by: NJan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

Logical x2APIC stops working if we rewrite it with zeros.
The best references are SDM April 2015: 10.12.10.1 Logical Destination
Mode in x2APIC Mode

  [...], the LDR are initialized by hardware based on the value of
  x2APIC ID upon x2APIC state transitions.

and SDM April 2015: 10.12.10.2 Deriving Logical x2APIC ID from the Local
x2APIC ID

  The LDR initialization occurs whenever the x2APIC mode is enabled
Signed-off-by: NRadim Krčmář <rkrcmar@redhat.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

SDM April 2015, 10.12.5 State Changes From xAPIC Mode to x2APIC Mode
• Any APIC ID value written to the memory-mapped local APIC ID register
  is not preserved.

Fix it by sourcing vcpu_id (= initial APIC ID) instead of memory-mapped
APIC ID.  Proper use of apic functions would result in two calls to
recalculate_apic_map(), so this patch makes a new helper.
Signed-off-by: NRadim Krčmář <rkrcmar@redhat.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

The periodic kvmclock sync can be an undesired source of latencies.

When running cyclictest on a guest, a latency spike is visible.
With kvmclock periodic sync disabled, the spike is gone.

Guests should use ntp which means the propagations of ntp corrections
from the host clock are unnecessary.

v2:
-> Make parameter read-only (Radim)
-> Return early on kvmclock_sync_fn (Andrew)
Reported-and-tested-by: NLuiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: NMarcelo Tosatti <mtosatti@redhat.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

Prepare for multiple address spaces this way, since a VCPU is not available
where unaccount_shadowed is called.  We will get to the right kvm_memslots
struct through the role field in struct kvm_mmu_page.
Reviewed-by: NTakuya Yoshikawa <yoshikawa_takuya_b1@lab.ntt.co.jp>
Reviewed-by: NRadim Krcmar <rkrcmar@redhat.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

The memory slot is already available from gfn_to_memslot_dirty_bitmap.
Isn't it a shame to look it up again?  Plus, it makes gfn_to_page_many_atomic
agnostic of multiple VCPU address spaces.
Reviewed-by: NRadim Krcmar <rkrcmar@redhat.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>