KVM_CAP_X2APIC_API is a capability for features related to x2APIC
enablement. KVM_X2APIC_API_32BIT_FORMAT feature can be enabled to
extend APIC ID in get/set ioctl and MSI addresses to 32 bits.
Both are needed to support x2APIC.
The feature has to be enableable and disabled by default, because
get/set ioctl shifted and truncated APIC ID to 8 bits by using a
non-standard protocol inspired by xAPIC and the change is not
backward-compatible.
Changes to MSI addresses follow the format used by interrupt remapping
unit. The upper address word, that used to be 0, contains upper 24 bits
of the LAPIC address in its upper 24 bits. Lower 8 bits are reserved as
0. Using the upper address word is not backward-compatible either as we
didn't check that userspace zeroed the word. Reserved bits are still
not explicitly checked, but non-zero data will affect LAPIC addresses,
which will cause a bug.
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
LAPIC is reset in xAPIC mode and the surrounding code expects that.
KVM never resets after initialization. This patch is just for sanity.
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The register is in hardware-compatible format now, so there is not need
to intercept.
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
APIC ID should be set to the initial APIC ID when enabling LAPIC.
This only matters if the guest changes APIC ID. No sane OS does that.
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
We currently always shift APIC ID as if APIC was in xAPIC mode.
x2APIC mode wants to use more bits and storing a hardware-compabible
value is the the sanest option.
KVM API to set the lapic expects that bottom 8 bits of APIC ID are in
top 8 bits of APIC_ID register, so the register needs to be shifted in
x2APIC mode.
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
x2APIC supports up to 2^32-1 LAPICs, but most guest in coming years will
probably has fewer VCPUs. Dynamic size saves memory at the cost of
turning one constant into a variable.
apic_map mutex had to be moved before allocation to avoid races with cpu
hotplug.
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Logical x2APIC IDs map injectively to physical x2APIC IDs, so we can
reuse the physical array for them. This allows us to save space by
sizing the logical maps according to the needs of xAPIC.
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
kvm_irq_delivery_to_apic_fast and kvm_intr_is_single_vcpu_fast both
compute the interrupt destination. Factor the code.
'struct kvm_lapic **dst = NULL' had to be added to silence GCC.
GCC might complain about potential NULL access in the future, because it
missed conditions that avoided uninitialized uses of dst.
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
240 has been well tested by Red Hat.
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
MMU now knows about execute only mappings, so
advertise the feature to L1 hypervisors
Signed-off-by: Bandan Das <bsd@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
To support execute only mappings on behalf of L1 hypervisors,
reuse ACC_USER_MASK to signify if the L1 hypervisor has the R bit
set.
For the nested EPT case, we assumed that the U bit was always set
since there was no equivalent in EPT page tables. Strictly
speaking, this was not necessary because handle_ept_violation
never set PFERR_USER_MASK in the error code (uf=0 in the
parlance of update_permission_bitmask). We now have to set
both U and UF correctly, respectively in FNAME(gpte_access)
and in handle_ept_violation.
Also in handle_ept_violation bit 3 of the exit qualification is
not enough to detect a present PTE; all three bits 3-5 have to
be checked.
Signed-off-by: Bandan Das <bsd@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
To support execute only mappings on behalf of L1
hypervisors, we need to teach set_spte() to honor all three of
L1's XWR bits. As a start, add a new variable "shadow_present_mask"
that will be set for non-EPT shadow paging and clear for EPT.
Signed-off-by: Bandan Das <bsd@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
We have two versions of the above function.
To prevent confusion and bugs in the future, remove
the non-FNAME version entirely and replace all calls
with the actual check.
Signed-off-by: Bandan Das <bsd@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
This is safe because this function is called
on host controlled page table and non-present/non-MMIO
sptes never use bits 1..31. For the EPT case, this
ensures that cases where only the execute bit is set
is marked valid.
Signed-off-by: Bandan Das <bsd@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
There is no reason to read the entry/exit control fields of the
VMCS and immediately write back the same value.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Because the vmcs12 preemption timer is emulated through a separate hrtimer,
we can keep on using the preemption timer in the vmcs02 to emulare L1's
TSC deadline timer.
However, the corresponding bit in the pin-based execution control field
must be kept consistent between vmcs01 and vmcs02. On vmentry we copy
it into the vmcs02; on vmexit the preemption timer must be disabled in
the vmcs01 if a preemption timer vmexit happened while in guest mode.
The preemption timer value in the vmcs02 is set by vmx_vcpu_run, so it
need not be considered in prepare_vmcs02.
Cc: Yunhong Jiang <yunhong.jiang@intel.com>
Cc: Haozhong Zhang <haozhong.zhang@intel.com>
Tested-by: Wanpeng Li <kernellwp@gmail.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The preemption timer for nested VMX is emulated by hrtimer which is started on L2
entry, stopped on L2 exit and evaluated via the check_nested_events hook. However,
nested_vmx_exit_handled is always returning true for preemption timer vmexit. Then,
the L1 preemption timer vmexit is captured and be treated as a L2 preemption
timer vmexit, causing NULL pointer dereferences or worse in the L1 guest's
vmexit handler:
BUG: unable to handle kernel NULL pointer dereference at (null)
IP: [< (null)>] (null)
PGD 0
Oops: 0010 [#1] SMP
Call Trace:
? kvm_lapic_expired_hv_timer+0x47/0x90 [kvm]
handle_preemption_timer+0xe/0x20 [kvm_intel]
vmx_handle_exit+0x169/0x15a0 [kvm_intel]
? kvm_arch_vcpu_ioctl_run+0xd5d/0x19d0 [kvm]
kvm_arch_vcpu_ioctl_run+0xdee/0x19d0 [kvm]
? kvm_arch_vcpu_ioctl_run+0xd5d/0x19d0 [kvm]
? vcpu_load+0x1c/0x60 [kvm]
? kvm_arch_vcpu_load+0x57/0x260 [kvm]
kvm_vcpu_ioctl+0x2d3/0x7c0 [kvm]
do_vfs_ioctl+0x96/0x6a0
? __fget_light+0x2a/0x90
SyS_ioctl+0x79/0x90
do_syscall_64+0x68/0x180
entry_SYSCALL64_slow_path+0x25/0x25
Code: Bad RIP value.
RIP [< (null)>] (null)
RSP <ffff8800b5263c48>
CR2: 0000000000000000
---[ end trace 9c70c48b1a2bc66e ]---
This can be reproduced readily by preemption timer enabled on L0 and disabled
on L1.
Return false since preemption timer vmexits must never be reflected to L2.
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Radim Krčmář <rkrcmar@redhat.com>
Cc: Yunhong Jiang <yunhong.jiang@intel.com>
Cc: Jan Kiszka <jan.kiszka@siemens.com>
Cc: Haozhong Zhang <haozhong.zhang@intel.com>
Signed-off-by: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Simplify cpu_has_vmx_preemption_timer. This is consistent with the
rest of setup_vmcs_config and preparatory for the next patch.
Tested-by: Wanpeng Li <kernellwp@gmail.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Use ARRAY_SIZE instead of dividing sizeof array with sizeof an element
Signed-off-by: Wei Yongjun <yongjun_wei@trendmicro.com.cn>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
INFO: rcu_sched detected stalls on CPUs/tasks:
1-...: (11800 GPs behind) idle=45d/140000000000000/0 softirq=0/0 fqs=21663
(detected by 0, t=65016 jiffies, g=11500, c=11499, q=719)
Task dump for CPU 1:
qemu-system-x86 R running task 0 3529 3525 0x00080808
ffff8802021791a0 ffff880212895040 0000000000000001 00007f1c2c00db40
ffff8801dd20fcd3 ffffc90002b98000 ffff8801dd20fc88 ffff8801dd20fcf8
0000000000000286 ffff8801dd2ac538 ffff8801dd20fcc0 ffffffffc06949c9
Call Trace:
? kvm_write_guest_cached+0xb9/0x160 [kvm]
? __delay+0xf/0x20
? wait_lapic_expire+0x14a/0x200 [kvm]
? kvm_arch_vcpu_ioctl_run+0xcbe/0x1b00 [kvm]
? kvm_arch_vcpu_ioctl_run+0xe34/0x1b00 [kvm]
? kvm_vcpu_ioctl+0x2d3/0x7c0 [kvm]
? __fget+0x5/0x210
? do_vfs_ioctl+0x96/0x6a0
? __fget_light+0x2a/0x90
? SyS_ioctl+0x79/0x90
? do_syscall_64+0x7c/0x1e0
? entry_SYSCALL64_slow_path+0x25/0x25
This can be reproduced readily by running a full dynticks guest(since hrtimer
in guest is heavily used) w/ lapic_timer_advance disabled.
If fail to program hardware preemption timer, we will fallback to hrtimer based
method, however, a previous programmed preemption timer miss to cancel in this
scenario which results in one hardware preemption timer and one hrtimer emulated
tsc deadline timer run simultaneously. So sometimes the target guest deadline
tsc is earlier than guest tsc, which leads to the computation in vmx_set_hv_timer
can underflow and cause delta_tsc to be set a huge value, then host soft lockup
as above.
This patch fix it by cancelling the previous programmed preemption timer if there
is once we failed to program the new preemption timer and fallback to hrtimer
based method.
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Radim Krčmář <rkrcmar@redhat.com>
Cc: Yunhong Jiang <yunhong.jiang@intel.com>
Signed-off-by: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
If the TSC deadline timer is programmed really close to the deadline or
even in the past, the computation in vmx_set_hv_timer can underflow and
cause delta_tsc to be set to a huge value. This generally results
in vmx_set_hv_timer returning -ERANGE, but we can fix it by limiting
delta_tsc to be positive or zero.
Reported-by: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
This gains a few clock cycles per vmexit. On Intel there is no need
anymore to enable the interrupts in vmx_handle_external_intr, since
we are using the "acknowledge interrupt on exit" feature. AMD
needs to do that, and must be careful to avoid the interrupt shadow.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Use the functions from context_tracking.h directly.
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Rik van Riel <riel@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
KVM reads the current boottime value as a struct timespec in order to
calculate the guest wallclock time, resulting in an overflow in 2038
on 32-bit systems.
The data then gets passed as an unsigned 32-bit number to the guest,
and that in turn overflows in 2106.
We cannot do much about the second overflow, which affects both 32-bit
and 64-bit hosts, but we can ensure that they both behave the same
way and don't overflow until 2106, by using getboottime64() to read
a timespec64 value.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
On Intel platforms, this patch adds LMCE to KVM MCE supported
capabilities and handles guest access to LMCE related MSRs.
Signed-off-by: Ashok Raj <ashok.raj@intel.com>
[Haozhong: macro KVM_MCE_CAP_SUPPORTED => variable kvm_mce_cap_supported
Only enable LMCE on Intel platform
Check MSR_IA32_FEATURE_CONTROL when handling guest
access to MSR_IA32_MCG_EXT_CTL]
Signed-off-by: Haozhong Zhang <haozhong.zhang@intel.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
KVM currently does not check the value written to guest
MSR_IA32_FEATURE_CONTROL, though bits corresponding to disabled features
may be set. This patch makes KVM to validate individual bits written to
guest MSR_IA32_FEATURE_CONTROL according to enabled features.
Signed-off-by: Haozhong Zhang <haozhong.zhang@intel.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
msr_ia32_feature_control will be used for LMCE and not depend only on
nested anymore, so move it from struct nested_vmx to struct vcpu_vmx.
Signed-off-by: Haozhong Zhang <haozhong.zhang@intel.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Hook the VMX preemption timer to the "hv timer" functionality added
by the previous patch. This includes: checking if the feature is
supported, if the feature is broken on the CPU, the hooks to
setup/clean the VMX preemption timer, arming the timer on vmentry
and handling the vmexit.
A module parameter states if the VMX preemption timer should be
utilized.
Signed-off-by: Yunhong Jiang <yunhong.jiang@intel.com>
[Move hv_deadline_tsc to struct vcpu_vmx, use -1 as the "unset" value.
Put all VMX bits here. Enable it by default #yolo. - Paolo]
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Prepare to switch from preemption timer to hrtimer in the
vmx_pre/post_block. Current functions are only for posted interrupt,
rename them accordingly.
Signed-off-by: Yunhong Jiang <yunhong.jiang@intel.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The VMX preemption timer can be used to virtualize the TSC deadline timer.
The VMX preemption timer is armed when the vCPU is running, and a VMExit
will happen if the virtual TSC deadline timer expires.
When the vCPU thread is blocked because of HLT, KVM will switch to use
an hrtimer, and then go back to the VMX preemption timer when the vCPU
thread is unblocked.
This solution avoids the complex OS's hrtimer system, and the host
timer interrupt handling cost, replacing them with a little math
(for guest->host TSC and host TSC->preemption timer conversion)
and a cheaper VMexit. This benefits latency for isolated pCPUs.
[A word about performance... Yunhong reported a 30% reduction in average
latency from cyclictest. I made a similar test with tscdeadline_latency
from kvm-unit-tests, and measured
- ~20 clock cycles loss (out of ~3200, so less than 1% but still
statistically significant) in the worst case where the test halts
just after programming the TSC deadline timer
- ~800 clock cycles gain (25% reduction in latency) in the best case
where the test busy waits.
I removed the VMX bits from Yunhong's patch, to concentrate them in the
next patch - Paolo]
Signed-off-by: Yunhong Jiang <yunhong.jiang@intel.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The function to start the tsc deadline timer virtualization will be used
also by the pre_block hook when we use the preemption timer; change it
to a separate function. No logic changes.
Signed-off-by: Yunhong Jiang <yunhong.jiang@intel.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The new created_vcpus field makes it possible to avoid the race between
irqchip and VCPU creation in a much nicer way; just check under kvm->lock
whether a VCPU has already been created.
We can then remove KVM_APIC_ARCHITECTURE too, because at this point the
symbol is only governing the default definition of kvm_vcpu_compatible.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
arch/x86/kvm/iommu.c includes <linux/intel-iommu.h> and <linux/dmar.h>, which
both are unnecessary, in fact incorrect to be here as they are intel specific.
Building kvm on x86 passed after removing above inclusion.
Signed-off-by: Kai Huang <kai.huang@linux.intel.com>
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
Make the function names more similar between KVM_REQ_NMI and KVM_REQ_SMI.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
If the processor exits to KVM while delivering an interrupt,
the hypervisor then requeues the interrupt for the next vmentry.
Trying to enter SMM in this same window causes to enter non-root
mode in emulated SMM (i.e. with IF=0) and with a request to
inject an IRQ (i.e. with a valid VM-entry interrupt info field).
This is invalid guest state (SDM 26.3.1.4 "Check on Guest RIP
and RFLAGS") and the processor fails vmentry.
The fix is to defer the injection from KVM_REQ_SMI to KVM_REQ_EVENT,
like we already do for e.g. NMIs. This patch doesn't change the
name of the process_smi function so that it can be applied to
stable releases. The next patch will modify the names so that
process_nmi and process_smi handle respectively KVM_REQ_NMI and
KVM_REQ_SMI.
This is especially common with Windows, probably due to the
self-IPI trick that it uses to deliver deferred procedure
calls (DPCs).
Reported-by: Laszlo Ersek <lersek@redhat.com>
Reported-by: Michał Zegan <webczat_200@poczta.onet.pl>
Fixes: 64d6067057
Cc: stable@vger.kernel.org
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
Intel CPUs having Turbo Boost feature implement an MSR to provide a
control interface via rdmsr/wrmsr instructions. One could detect the
presence of this feature by issuing one of these instructions and
handling the #GP exception which is generated in case the referenced MSR
is not implemented by the CPU.
KVM's vCPU model behaves exactly as a real CPU in this case by injecting
a fault when MSR_IA32_PERF_CTL is called (which KVM does not support).
However, some operating systems use this register during an early boot
stage in which their kernel is not capable of handling #GP correctly,
causing #DP and finally a triple fault effectively resetting the vCPU.
This patch implements a dummy handler for MSR_IA32_PERF_CTL to avoid the
crashes.
Signed-off-by: Dmitry Bilunov <kmeaw@yandex-team.ru>
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
In theory, nothing prevents the compiler from write-tearing PTEs, or
split PTE writes. These partially-modified PTEs can be fetched by other
cores and cause mayhem. I have not really encountered such case in
real-life, but it does seem possible.
For example, the compiler may try to do something creative for
kvm_set_pte_rmapp() and perform multiple writes to the PTE.
Signed-off-by: Nadav Amit <nadav.amit@gmail.com>
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
Pull UML updates from Richard Weinberger:
"This contains a nice FPU fixup from Eli Cooper for UML"
* 'for-linus-4.7-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rw/uml:
um: add extended processor state save/restore support
um: extend fpstate to _xstate to support YMM registers
um: fix FPU state preservation around signal handlers
The do_brk() and vm_brk() return value was "unsigned long" and returned
the starting address on success, and an error value on failure. The
reasons are entirely historical, and go back to it basically behaving
like the mmap() interface does.
However, nobody actually wanted that interface, and it causes totally
pointless IS_ERR_VALUE() confusion.
What every single caller actually wants is just the simpler integer
return of zero for success and negative error number on failure.
So just convert to that much clearer and more common calling convention,
and get rid of all the IS_ERR_VALUE() uses wrt vm_brk().
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>