There are a few improvements that can be made to the TSC offset
matching code. First, we don't need to call the 128-bit multiply
(especially on a constant number), the code works much nicer to
do computation in nanosecond units.
Second, the way everything is setup with software TSC rate scaling,
we currently have per-cpu rates. Obviously this isn't too desirable
to use in practice, but if for some reason we do change the rate of
all VCPUs at runtime, then reset the TSCs, we will only want to
match offsets for VCPUs running at the same rate.
Finally, for the case where we have an unstable host TSC, but
rate scaling is being done in hardware, we should call the platform
code to compute the TSC offset, so the math is reorganized to recompute
the base instead, then transform the base into an offset using the
existing API.
[avi: fix 64-bit division on i386]
Signed-off-by: Zachary Amsden <zamsden@gmail.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
KVM: Fix 64-bit division in kvm_write_tsc()
Breaks i386 build.
Signed-off-by: Avi Kivity <avi@redhat.com>
This requires some restructuring; rather than use 'virtual_tsc_khz'
to indicate whether hardware rate scaling is in effect, we consider
each VCPU to always have a virtual TSC rate. Instead, there is new
logic above the vendor-specific hardware scaling that decides whether
it is even necessary to use and updates all rate variables used by
common code. This means we can simply query the virtual rate at
any point, which is needed for software rate scaling.
There is also now a threshold added to the TSC rate scaling; minor
differences and variations of measured TSC rate can accidentally
provoke rate scaling to be used when it is not needed. Instead,
we have a tolerance variable called tsc_tolerance_ppm, which is
the maximum variation from user requested rate at which scaling
will be used. The default is 250ppm, which is the half the
threshold for NTP adjustment, allowing for some hardware variation.
In the event that hardware rate scaling is not available, we can
kludge a bit by forcing TSC catchup to turn on when a faster than
hardware speed has been requested, but there is nothing available
yet for the reverse case; this requires a trap and emulate software
implementation for RDTSC, which is still forthcoming.
[avi: fix 64-bit division on i386]
Signed-off-by: Zachary Amsden <zamsden@gmail.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Increase recommended max vcpus from 64 to 160 (tested internally
at Red Hat).
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
In some cases guests should not provide workarounds for errata even when the
physical processor is affected. For example, because of erratum 400 on family
10h processors a Linux guest will read an MSR (resulting in VMEXIT) before
going to idle in order to avoid getting stuck in a non-C0 state. This is not
necessary: HLT and IO instructions are intercepted and therefore there is no
reason for erratum 400 workaround in the guest.
This patch allows us to present a guest with certain errata as fixed,
regardless of the state of actual hardware.
Signed-off-by: Boris Ostrovsky <boris.ostrovsky@amd.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Add a helper function that emulates the RDPMC instruction operation.
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Use perf_events to emulate an architectural PMU, version 2.
Based on PMU version 1 emulation by Avi Kivity.
[avi: adjust for cpuid.c]
[jan: fix anonymous field initialization for older gcc]
Signed-off-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Move the mmu code in kvm_arch_vcpu_init() to kvm_mmu_create()
Signed-off-by: Xiao Guangrong <xiaoguangrong@linux.vnet.ibm.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Currently, write protecting a slot needs to walk all the shadow pages
and checks ones which have a pte mapping a page in it.
The walk is overly heavy when dirty pages in that slot are not so many
and checking the shadow pages would result in unwanted cache pollution.
To mitigate this problem, we use rmap_write_protect() and check only
the sptes which can be reached from gfns marked in the dirty bitmap
when the number of dirty pages are less than that of shadow pages.
This criterion is reasonable in its meaning and worked well in our test:
write protection became some times faster than before when the ratio of
dirty pages are low and was not worse even when the ratio was near the
criterion.
Note that the locking for this write protection becomes fine grained.
The reason why this is safe is descripted in the comments.
Signed-off-by: Takuya Yoshikawa <yoshikawa.takuya@oss.ntt.co.jp>
Signed-off-by: Avi Kivity <avi@redhat.com>
The host side pv mmu support has been marked for feature removal in
January 2011. It's not in use, is slower than shadow or hardware
assisted paging, and a maintenance burden. It's November 2011, time to
remove it.
Signed-off-by: Chris Wright <chrisw@redhat.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Detecting write-flooding does not work well, when we handle page written, if
the last speculative spte is not accessed, we treat the page is
write-flooding, however, we can speculative spte on many path, such as pte
prefetch, page synced, that means the last speculative spte may be not point
to the written page and the written page can be accessed via other sptes, so
depends on the Accessed bit of the last speculative spte is not enough
Instead of detected page accessed, we can detect whether the spte is accessed
after it is written, if the spte is not accessed but it is written frequently,
we treat is not a page table or it not used for a long time
Signed-off-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Fast prefetch spte for the unsync shadow page on invlpg path
Signed-off-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
In current code, the accessed bit is always set when page fault occurred,
do not need to set it on pte write path
Signed-off-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
If the emulation is caused by #PF and it is non-page_table writing instruction,
it means the VM-EXIT is caused by shadow page protected, we can zap the shadow
page and retry this instruction directly
The idea is from Avi
Signed-off-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
If simultaneous NMIs happen, we're supposed to queue the second
and next (collapsing them), but currently we sometimes collapse
the second into the first.
Fix by using a counter for pending NMIs instead of a bool; since
the counter limit depends on whether the processor is currently
in an NMI handler, which can only be checked in vcpu context
(via the NMI mask), we add a new KVM_REQ_NMI to request recalculation
of the counter.
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
KVM assumed in several places that reading the TSC MSR returns the value for
L1. This is incorrect, because when L2 is running, the correct TSC read exit
emulation is to return L2's value.
We therefore add a new x86_ops function, read_l1_tsc, to use in places that
specifically need to read the L1 TSC, NOT the TSC of the current level of
guest.
Note that one change, of one line in kvm_arch_vcpu_load, is made redundant
by a different patch sent by Zachary Amsden (and not yet applied):
kvm_arch_vcpu_load() should not read the guest TSC, and if it didn't, of
course we didn't have to change the call of kvm_get_msr() to read_l1_tsc().
[avi: moved callback to kvm_x86_ops tsc block]
Signed-off-by: Nadav Har'El <nyh@il.ibm.com>
Acked-by: Zachary Amsdem <zamsden@gmail.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Architecturally, PDPTEs are cached in the PDPTRs when CR3 is reloaded.
On SVM, it is not possible to implement this, but on VMX this is possible
and was indeed implemented until nested SVM changed this to unconditionally
read PDPTEs dynamically. This has noticable impact when running PAE guests.
Fix by changing the MMU to read PDPTRs from the cache, falling back to
reading from memory for the nested MMU.
Signed-off-by: Avi Kivity <avi@redhat.com>
Tested-by: Joerg Roedel <joerg.roedel@amd.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
The vmexit tracepoints format the exit_reason to make it human-readable.
Since the exit_reason depends on the instruction set (vmx or svm),
formatting is handled with ftrace_print_symbols_seq() by referring to
the appropriate exit reason table.
However, the ftrace_print_symbols_seq() function is not meant to be used
directly in tracepoints since it does not export the formatting table
which userspace tools like trace-cmd and perf use to format traces.
In practice perf dies when formatting vmexit-related events and
trace-cmd falls back to printing the numeric value (with extra
formatting code in the kvm plugin to paper over this limitation). Other
userspace consumers of vmexit-related tracepoints would be in similar
trouble.
To avoid significant changes to the kvm_exit tracepoint, this patch
moves the vmx and svm exit reason tables into arch/x86/kvm/trace.h and
selects the right table with __print_symbolic() depending on the
instruction set. Note that __print_symbolic() is designed for exporting
the formatting table to userspace and allows trace-cmd and perf to work.
Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
The patch raises the hard limit of VCPU count to 254.
This will allow developers to easily work on scalability
and will allow users to test high VCPU setups easily without
patching the kernel.
To prevent possible issues with current setups, KVM_CAP_NR_VCPUS
now returns the recommended VCPU limit (which is still 64) - this
should be a safe value for everybody, while a new KVM_CAP_MAX_VCPUS
returns the hard limit which is now 254.
Cc: Avi Kivity <avi@redhat.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Cc: Pekka Enberg <penberg@kernel.org>
Suggested-by: Pekka Enberg <penberg@cs.helsinki.fi>
Signed-off-by: Sasha Levin <levinsasha928@gmail.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Use rcu to protect shadow pages table to be freed, so we can safely walk it,
it should run fastly and is needed by mmio page fault
Signed-off-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
The idea is from Avi:
| Maybe it's time to kill off bypass_guest_pf=1. It's not as effective as
| it used to be, since unsync pages always use shadow_trap_nonpresent_pte,
| and since we convert between the two nonpresent_ptes during sync and unsync.
Signed-off-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
If the page fault is caused by mmio, we can cache the mmio info, later, we do
not need to walk guest page table and quickly know it is a mmio fault while we
emulate the mmio instruction
Signed-off-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
To implement steal time, we need the hypervisor to pass the guest
information about how much time was spent running other processes
outside the VM, while the vcpu had meaningful work to do - halt
time does not count.
This information is acquired through the run_delay field of
delayacct/schedstats infrastructure, that counts time spent in a
runqueue but not running.
Steal time is a per-cpu information, so the traditional MSR-based
infrastructure is used. A new msr, KVM_MSR_STEAL_TIME, holds the
memory area address containing information about steal time
This patch contains the hypervisor part of the steal time infrasructure,
and can be backported independently of the guest portion.
[avi, yongjie: export delayacct_on, to avoid build failures in some configs]
Signed-off-by: Glauber Costa <glommer@redhat.com>
Tested-by: Eric B Munson <emunson@mgebm.net>
CC: Rik van Riel <riel@redhat.com>
CC: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
CC: Peter Zijlstra <peterz@infradead.org>
CC: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Yongjie Ren <yongjie.ren@intel.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
When CR0.WP=0, we sometimes map user pages as kernel pages (to allow
the kernel to write to them). Unfortunately this also allows the kernel
to fetch from these pages, even if CR4.SMEP is set.
Adjust for this by also setting NX on the spte in these circumstances.
Signed-off-by: Avi Kivity <avi@redhat.com>
This patch allows the guest to enable the VMXE bit in CR4, which is a
prerequisite to running VMXON.
Whether to allow setting the VMXE bit now depends on the architecture (svm
or vmx), so its checking has moved to kvm_x86_ops->set_cr4(). This function
now returns an int: If kvm_x86_ops->set_cr4() returns 1, __kvm_set_cr4()
will also return 1, and this will cause kvm_set_cr4() will throw a #GP.
Turning on the VMXE bit is allowed only when the nested VMX feature is
enabled, and turning it off is forbidden after a vmxon.
Signed-off-by: Nadav Har'El <nyh@il.ibm.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Parent pte rmap and page rmap are very similar, so use the same arithmetic
for them
Signed-off-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Abstract the operation of rmap to spte_list, then we can use it for the
reverse mapping of parent pte in the later patch
Signed-off-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Simply return from kvm_mmu_pte_write path if no shadow page is
write-protected, then we can avoid to walk all shadow pages and hold
mmu-lock
Signed-off-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
We clean up a failed VMREAD by clearing the output register. Do
it in the exception handler instead of unconditionally. This is
worthwhile since there are more than a hundred call sites.
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Since the emulator now checks segment limits and access rights, it
generates a lot more accesses to the vmcs segment fields. Undo some
of the performance hit by cacheing those fields in a read-only cache
(the entire cache is invalidated on any write, or on guest exit).
Signed-off-by: Avi Kivity <avi@redhat.com>
Avoid using ctxt->vcpu; we can do everything with ->get_cr() and ->set_cr().
A side effect is that we no longer activate the fpu on emulated CLTS; but that
should be very rare.
Signed-off-by: Avi Kivity <avi@redhat.com>
The last_guest_tsc is used in vcpu_load to adjust the
tsc_offset since tsc-scaling is merged. So the
last_guest_tsc needs to be updated in vcpu_put instead of
the the last_host_tsc. This is fixed with this patch.
Reported-by: Jan Kiszka <jan.kiszka@web.de>
Tested-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Currently we sync registers back and forth before/after exiting
to userspace for IO, but during IO device model shouldn't need to
read/write the registers, so we can as well skip those sync points. The
only exaception is broken vmware backdor interface. The new code sync
registers content during IO only if registers are read from/written to
by userspace in the middle of the IO operation and this almost never
happens in practise.
Signed-off-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
This patch implements two new vm-ioctls to get and set the
virtual_tsc_khz if the machine supports tsc-scaling. Setting
the tsc-frequency is only possible before userspace creates
any vcpu.
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
With TSC scaling in SVM the tsc-offset needs to be
calculated differently. This patch propagates this
calculation into the architecture specific modules so that
this complexity can be handled there.
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
This patch implements a call-back into the architecture code
to allow the propagation of changes to the virtual tsc_khz
of the vcpu.
On SVM it updates the tsc_ratio variable, on VMX it does
nothing.
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
This patch changes the kvm_guest_time_update function to use
TSC frequency the guest actually has for updating its clock.
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
The mmu_seq verification can be removed since we get the pfn in the
protection of mmu_lock.
Signed-off-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
This patch adds all necessary intercept checks for
instructions that access the crX registers.
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
This patch adds a callback into kvm_x86_ops so that svm and
vmx code can do intercept checks on emulated instructions.
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Since sse instructions can issue 16-byte mmios, we need to support them. We
can't increase the kvm_run mmio buffer size to 16 bytes without breaking
compatibility, so instead we break the large mmios into two smaller 8-byte
ones. Since the bus is 64-bit we aren't breaking any atomicity guarantees.
Signed-off-by: Avi Kivity <avi@redhat.com>
We may read the cpl quite often in the same vmexit (instruction privilege
check, memory access checks for instruction and operands), so we gain
a bit if we cache the value.
Signed-off-by: Avi Kivity <avi@redhat.com>
Zachary Amsden