Existing KVM statistics are either just counters (kvm_stat) reported for
KVM generally or trace based aproaches like kvm_trace.
For KVM on powerpc we had the need to track the timings of the different exit
types. While this could be achieved parsing data created with a kvm_trace
extension this adds too much overhead (at least on embedded PowerPC) slowing
down the workloads we wanted to measure.
Therefore this patch adds a in-kernel exit timing statistic to the powerpc kvm
code. These statistic is available per vm&vcpu under the kvm debugfs directory.
As this statistic is low, but still some overhead it can be enabled via a
.config entry and should be off by default.
Since this patch touched all powerpc kvm_stat code anyway this code is now
merged and simplified together with the exit timing statistic code (still
working with exit timing disabled in .config).
Signed-off-by: Christian Ehrhardt <ehrhardt@linux.vnet.ibm.com>
Signed-off-by: Hollis Blanchard <hollisb@us.ibm.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Store shadow TLB entries in memory, but only use it on host context switch
(instead of every guest entry). This improves performance for most workloads on
440 by reducing the guest TLB miss rate.
Signed-off-by: Hollis Blanchard <hollisb@us.ibm.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Formerly, we used to maintain a per-vcpu shadow TLB and on every entry to the
guest would load this array into the hardware TLB. This consumed 1280 bytes of
memory (64 entries of 16 bytes plus a struct page pointer each), and also
required some assembly to loop over the array on every entry.
Instead of saving a copy in memory, we can just store shadow mappings directly
into the hardware TLB, accepting that the host kernel will clobber these as
part of the normal 440 TLB round robin. When we do that we need less than half
the memory, and we have decreased the exit handling time for all guest exits,
at the cost of increased number of TLB misses because the host overwrites some
guest entries.
These savings will be increased on processors with larger TLBs or which
implement intelligent flush instructions like tlbivax (which will avoid the
need to walk arrays in software).
In addition to that and to the code simplification, we have a greater chance of
leaving other host userspace mappings in the TLB, instead of forcing all
subsequent tasks to re-fault all their mappings.
Signed-off-by: Hollis Blanchard <hollisb@us.ibm.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
KVM currently ignores the host's round robin TLB eviction selection, instead
maintaining its own TLB state and its own round robin index. However, by
participating in the normal 44x TLB selection, we can drop the alternate TLB
processing in KVM. This results in a significant performance improvement,
since that processing currently must be done on *every* guest exit.
Accordingly, KVM needs to be able to access and increment tlb_44x_index.
(KVM on 440 cannot be a module, so there is no need to export this symbol.)
Signed-off-by: Hollis Blanchard <hollisb@us.ibm.com>
Acked-by: Josh Boyer <jwboyer@linux.vnet.ibm.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
KVM on 440 has always been able to handle large guest mappings with 4K host
pages -- we must, since the guest kernel uses 256MB mappings.
This patch makes KVM work when the host has large pages too (tested with 64K).
Signed-off-by: Hollis Blanchard <hollisb@us.ibm.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Impact: make global function static
arch/x86/kvm/vmx.c:134:3: warning: symbol 'vmx_capability' was not declared. Should it be static?
Signed-off-by: Hannes Eder <hannes@hanneseder.net>
Signed-off-by: Avi Kivity <avi@redhat.com>
The s390 backend of kvm never calls kvm_vcpu_uninit. This causes
a memory leak of vcpu->run pages.
Lets call kvm_vcpu_uninit in kvm_arch_vcpu_destroy to free
the vcpu->run.
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Acked-by: Carsten Otte <cotte@de.ibm.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Currently it is impossible to unload the kvm module on s390.
This patch fixes kvm_arch_destroy_vm to release all cpus.
This make it possible to unload the module.
In addition we stop messing with the module refcount in arch code.
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Acked-by: Carsten Otte <cotte@de.ibm.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
The write protect verification in set_spte is unnecessary for page sync.
Its guaranteed that, if the unsync spte was writable, the target page
does not have a write protected shadow (if it had, the spte would have
been write protected under mmu_lock by rmap_write_protect before).
Same reasoning applies to mark_page_dirty: the gfn has been marked as
dirty via the pagefault path.
The cost of hash table and memslot lookups are quite significant if the
workload is pagetable write intensive resulting in increased mmu_lock
contention.
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Currently, we only set the KVM paravirt signature in case
of CONFIG_KVM_GUEST. However, it is possible to have it turned
off, while CONFIG_KVM_CLOCK is turned on. This is also a paravirt
case, and should be shown accordingly.
Signed-off-by: Glauber Costa <glommer@redhat.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
If we're injecting an interrupt, and another one is pending, request
an interrupt window notification so we don't have excess latency on the
second interrupt.
This shouldn't happen in practice since an EOI will be issued, giving a second
chance to request an interrupt window, but...
Signed-off-by: Avi Kivity <avi@redhat.com>
Since vmm runs in an isolated address space and it is just a copy
of host's kvm-intel module, so once vmm crashes, we just crash all guests
running on it instead of crashing whole kernel.
Signed-off-by: Xiantao Zhang <xiantao.zhang@intel.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Use printk infrastructure to print out some debug info once VM crashes.
Signed-off-by: Xiantao Zhang <xiantao.zhang@intel.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
kvm-intel module is relocated to an isolated address space
with kernel, so it can't call host kernel's printk for debug
purpose. In the module, we implement the printk to output debug
info of vmm.
Signed-off-by: Xiantao Zhang <xiantao.zhang@intel.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
On emergency_restart, we may need to use an NMI to disable virtualization
on all CPUs. We do that using nmi_shootdown_cpus() if VMX is enabled.
Note: With this patch, we will run the NMI stuff only when the CPU where
emergency_restart() was called has VMX enabled. This should work on most
cases because KVM enables VMX on all CPUs, but we may miss the small
window where KVM is doing that. Also, I don't know if all code using
VMX out there always enable VMX on all CPUs like KVM does. We have two
other alternatives for that:
a) Have an API that all code that enables VMX on any CPU should use
to tell the kernel core that it is going to enable VMX on the CPUs.
b) Always call nmi_shootdown_cpus() if the CPU supports VMX. This is
a bit intrusive and more risky, as it would run nmi_shootdown_cpus()
on emergency_reboot() even on systems where virtualization is never
enabled.
Finding a proper point to hook the nmi_shootdown_cpus() call isn't
trivial, as the non-emergency machine_restart() (that doesn't need the
NMI tricks) uses machine_emergency_restart() directly.
The solution to make this work without adding a new function or argument
to machine_ops was setting a 'reboot_emergency' flag that tells if
native_machine_emergency_restart() needs to do the virt cleanup or not.
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
We need to disable virtualization extensions on all CPUs before booting
the kdump kernel, otherwise the kdump kernel booting will fail, and
rebooting after the kdump kernel did its task may also fail.
We do it using cpu_emergency_vmxoff() and cpu_emergency_svm_disable(),
that should always work, because those functions check if the CPUs
support SVM or VMX before doing their tasks.
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
This function can be used by the reboot or kdump code to forcibly
disable SVM on the CPU.
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Use a trick to keep the printk()s on has_svm() working as before. gcc
will take care of not generating code for the 'msg' stuff when the
function is called with a NULL msg argument.
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Add cpu_emergency_vmxoff() and its friends: cpu_vmx_enabled() and
__cpu_emergency_vmxoff().
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Along with some comments on why it is different from the core cpu_vmxoff()
function.
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Unfortunately we can't use exactly the same code from vmx
hardware_disable(), because the KVM function uses the
__kvm_handle_fault_on_reboot() tricks.
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
It will be used by core code on kdump and reboot, to disable
vmx if needed.
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Those definitions will be used by code outside KVM, so move it outside
of a KVM-specific source file.
Those definitions are used only on kvm/vmx.c, that already includes
asm/vmx.h, so they can be moved safely.
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
svm.h will be used by core code that is independent of KVM, so I am
moving it outside the arch/x86/kvm directory.
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
vmx.h will be used by core code that is independent of KVM, so I am
moving it outside the arch/x86/kvm directory.
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
We used to defer invalidating userspace TLB entries until jumping out of the
kernel. This was causing MMU weirdness most easily triggered by using a pipe in
the guest, e.g. "dmesg | tail". I believe the problem was that after the guest
kernel changed the PID (part of context switch), the old process's mappings
were still present, and so copy_to_user() on the "return to new process" path
ended up using stale mappings.
Testing with large pages (64K) exposed the problem, probably because with 4K
pages, pressure on the TLB faulted all process A's mappings out before the
guest kernel could insert any for process B.
Signed-off-by: Hollis Blanchard <hollisb@us.ibm.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Bare metal Linux on 440 can "overmap" RAM in the kernel linear map, so that it
can use large (256MB) mappings even if memory isn't a multiple of 256MB. To
prevent the hardware prefetcher from loading from an invalid physical address
through that mapping, it's marked Guarded.
However, KVM must ensure that all guest mappings are backed by real physical
RAM (since a deliberate access through a guarded mapping could still cause a
machine check). Accordingly, we don't need to make our mappings guarded, so
let's allow prefetching as the designers intended.
Curiously this patch didn't affect performance at all on the quick test I
tried, but it's clearly the right thing to do anyways and may improve other
workloads.
Signed-off-by: Hollis Blanchard <hollisb@us.ibm.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Make sure that CONFIG_KVM cannot be selected without processor support
(currently, 440 is the only processor implementation available).
Signed-off-by: Hollis Blanchard <hollisb@us.ibm.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
The code to traverse the cpuid data array list for counting type of leaves is
currently broken.
This patches fixes the 2 things in it.
1. Set the 1st counting entry's flag KVM_CPUID_FLAG_STATE_READ_NEXT. Without
it the code will never find a valid entry.
2. Also the stop condition in the for loop while looking for the next unflaged
entry is broken. It needs to stop when it find one matching entry;
and in the case of count of 1, it will be the same entry found in this
iteration.
Signed-Off-By: Nitin A Kamble <nitin.a.kamble@intel.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
For cpuid leaf 0xb the bits 8-15 in ECX register define the end of counting
leaf. The previous code was using bits 0-7 for this purpose, which is
a bug.
Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
set ESR[PTR] when emulating a guest trap. This allows Linux guests to
properly handle WARN_ON() (i.e. detect that it's a non-fatal trap).
Also remove debugging printk in trap emulation.
Signed-off-by: Hollis Blanchard <hollisb@us.ibm.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
In kvmppc_deliver_interrupt is just one case left in the switch and it is a
rare one (less than 8%) when looking at the exit numbers. Therefore we can
at least drop the switch/case and if an if. I inserted an unlikely too, but
that's open for discussion.
In kvmppc_can_deliver_interrupt all frequent cases are in the default case.
I know compilers are smart but we can make it easier for them. By writing
down all options and removing the default case combined with the fact that
ithe values are constants 0..15 should allow the compiler to write an easy
jump table.
Modifying kvmppc_can_deliver_interrupt pointed me to the fact that gcc seems
to be unable to reduce priority_exception[x] to a build time constant.
Therefore I changed the usage of the translation arrays in the interrupt
delivery path completely. It is now using priority without translation to irq
on the full irq delivery path.
To be able to do that ivpr regs are stored by their priority now.
Additionally the decision made in kvmppc_can_deliver_interrupt is already
sufficient to get the value of interrupt_msr_mask[x]. Therefore we can replace
the 16x4byte array used here with a single 4byte variable (might still be one
miss, but the chance to find this in cache should be better than the right
entry of the whole array).
Signed-off-by: Christian Ehrhardt <ehrhardt@linux.vnet.ibm.com>
Signed-off-by: Hollis Blanchard <hollisb@us.ibm.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Since we use a unsigned long here anyway we can use the optimized __ffs.
Signed-off-by: Christian Ehrhardt <ehrhardt@linux.vnet.ibm.com>
Signed-off-by: Hollis Blanchard <hollisb@us.ibm.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Currently we use an unnecessary if&switch to detect some cases.
To be honest we don't need the ligh_exits counter anyway, because we can
calculate it out of others. Sum_exits can also be calculated, so we can
remove that too.
MMIO, DCR and INTR can be counted on other places without these
additional control structures (The INTR case was never hit anyway).
The handling of BOOKE_INTERRUPT_EXTERNAL/BOOKE_INTERRUPT_DECREMENTER is
similar, but we can avoid the additional if when copying 3 lines of code.
I thought about a goto there to prevent duplicate lines, but rewriting three
lines should be better style than a goto cross switch/case statements (its
also not enough code to justify a new inline function).
Signed-off-by: Christian Ehrhardt <ehrhardt@linux.vnet.ibm.com>
Signed-off-by: Hollis Blanchard <hollisb@us.ibm.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
When changing some msr bits e.g. problem state we need to take special
care of that. We call the function in our mtmsr emulation (not needed for
wrtee[i]), but we don't call kvmppc_set_msr if we change msr via set_regs
ioctl.
It's a corner case we never hit so far, but I assume it should be
kvmppc_set_msr in our arch set regs function (I found it because it is also
a corner case when using pv support which would miss the update otherwise).
Signed-off-by: Christian Ehrhardt <ehrhardt@linux.vnet.ibm.com>
Signed-off-by: Hollis Blanchard <hollisb@us.ibm.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
However, some of these fields could be split into separate per-core structures
in the future.
Signed-off-by: Hollis Blanchard <hollisb@us.ibm.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
This patch doesn't yet move all 44x-specific data into the new structure, but
is the first step down that path. In the future we may also want to create a
struct kvm_vcpu_booke.
Based on patch from Liu Yu <yu.liu@freescale.com>.
Signed-off-by: Hollis Blanchard <hollisb@us.ibm.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Hollis Blanchard