This patch adds logic to kvm/x86 which allows to mark an
injected exception as reinjected. This allows to remove an
ugly hack from svm_complete_interrupts that prevented
exceptions from being reinjected at all in the nested case.
The hack was necessary because an reinjected exception into
the nested guest could cause a nested vmexit emulation. But
reinjected exceptions must not intercept. The downside of
the hack is that a exception that in injected could get
lost.
This patch fixes the problem and puts the code for it into
generic x86 files because. Nested-VMX will likely have the
same problem and could reuse the code.
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
This patch adds the get_supported_cpuid callback to
kvm_x86_ops. It will be used in do_cpuid_ent to delegate the
decission about some supported cpuid bits to the
architecture modules.
Cc: stable@kernel.org
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Since commit bf47a760f6, we no longer handle ptes with the global bit
set specially, so there is no reason to distinguish between shadow pages
created with cr4.gpe set and clear.
Such tracking is expensive when the guest toggles cr4.pge, so drop it.
Signed-off-by: Avi Kivity <avi@redhat.com>
emulator_task_switch() should return -1 for failure and 0 for success to
the caller, just like x86_emulate_insn() does.
Signed-off-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
When a fault triggers a task switch, the error code, if existent, has to
be pushed on the new task's stack. Implement the missing bits.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
We can call kvm_mmu_pte_write() directly from
emulator_cmpxchg_emulated() instead of passing mmu_only down to
emulator_write_emulated_onepage() and call it there.
Signed-off-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Currently both SVM and VMX have their own DR handling code. Move it to
x86.c.
Acked-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
MAXPHYADDR is derived from cpuid 0x80000008, but when that isn't present, we
get some random value.
Fix by checking first that cpuid 0x80000008 is supported.
Acked-by: Pekka Enberg <penberg@cs.helsinki.fi>
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Log emulated instructions in ftrace, especially if they failed.
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Make sure that rflags is committed only after successful instruction
emulation.
Signed-off-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Unify all conditions that get us back into emulator after returning from
userspace.
Signed-off-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Currently when string instruction is only partially complete we go back
to a guest mode, guest tries to reexecute instruction and exits again
and at this point emulation continues. Avoid all of this by restarting
instruction without going back to a guest mode, but return to a guest
mode each 1024 iterations to allow interrupt injection. Pending
exception causes immediate guest entry too.
Signed-off-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Currently emulation is done outside of emulator so things like doing
ins/outs to/from mmio are broken it also makes it hard (if not impossible)
to implement single stepping in the future. The implementation in this
patch is not efficient since it exits to userspace for each IO while
previous implementation did 'ins' in batches. Further patch that
implements pio in string read ahead address this problem.
Signed-off-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
in/out emulation is broken now. The breakage is different depending
on where IO device resides. If it is in userspace emulator reports
emulation failure since it incorrectly interprets kvm_emulate_pio()
return value. If IO device is in the kernel emulation of 'in' will do
nothing since kvm_emulate_pio() stores result directly into vcpu
registers, so emulator will overwrite result of emulation during
commit of shadowed register.
Signed-off-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Eliminate the need to call back into KVM to get it from emulator.
Signed-off-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Use this callback instead of directly call kvm function. Also rename
realmode_(set|get)_cr to emulator_(set|get)_cr since function has nothing
to do with real mode.
Signed-off-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Mov reg, cr instruction doesn't change flags in any meaningful way, so
no need to update rflags after instruction execution.
Signed-off-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Currently emulated atomic operations are immediately followed by a non-atomic
operation, so that kvm_mmu_pte_write() can be invoked. This updates the mmu
but undoes the whole point of doing things atomically.
Fix by only performing the atomic operation and the mmu update, and avoiding
the non-atomic write.
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Once upon a time, locked operations were emulated while holding the mmu mutex.
Since mmu pages were write protected, it was safe to emulate the writes in
a non-atomic manner, since there could be no other writer, either in the
guest or in the kernel.
These days emulation takes place without holding the mmu spinlock, so the
write could be preempted by an unshadowing event, which exposes the page
to writes by the guest. This may cause corruption of guest page tables.
Fix by using an atomic cmpxchg for these operations.
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
If no irq chip in kernel, ioctl KVM_IRQ_LINE will return -EFAULT.
But I see in other place such as KVM_[GET|SET]IRQCHIP, -ENXIO is
return. So this patch used -ENXIO instead of -EFAULT.
Signed-off-by: Wei Yongjun <yjwei@cn.fujitsu.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
This patch does:
- no need call tracepoint_synchronize_unregister() when kvm module
is unloaded since ftrace can handle it
- cleanup ftrace's macro
Signed-off-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
fix segment_base() to properly check for null segment selector and
avoid accessing NULL pointer if ldt selector in null.
Signed-off-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Linux now has native_store_gdt() to do the same. Use it instead of
kvm local version.
Signed-off-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
The vcpu->arch.cr0 variable is already set in the
architecture specific set_cr0 callbacks. There is no need to
set it in the common code.
This allows the architecture code to keep the old arch.cr0
value if it wants. This is required for nested svm to decide
if a selective_cr0 exit needs to be injected.
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Hyper-V as a guest wants to write this bit. This patch
ignores it.
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
This patch adds a tracepoint to get information about the
most important intercept bitmasks from the nested vmcb.
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Call directly into the vendor services for getting/setting rflags in
emulate_instruction to ensure injected TF survives the emulation.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
RF is not required for injecting TF as the latter will trigger only
after an instruction execution anyway. So do not touch RF when arming or
disarming guest single-step mode.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Based on Gleb's suggestion: Add a helper kvm_is_linear_rip that matches
a given linear RIP against the current one. Use this for guest
single-stepping, more users will follow.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
So far user space was not able to save and restore debug registers for
migration or after reset. Plug this hole.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
The interrupt shadow created by STI or MOV-SS-like operations is part of
the VCPU state and must be preserved across migration. Transfer it in
the spare padding field of kvm_vcpu_events.interrupt.
As a side effect we now have to make vmx_set_interrupt_shadow robust
against both shadow types being set. Give MOV SS a higher priority and
skip STI in that case to avoid that VMX throws a fault on next entry.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
To avoid that user space migrates a pending software exception or
interrupt, mask them out on KVM_GET_VCPU_EVENTS. Without this, user
space would try to reinject them, and we would have to reconstruct the
proper instruction length for VMX event injection. Now the pending event
will be reinjected via executing the triggering instruction again.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
vcpu->run is initialized on vcpu creation and can never be NULL
here.
Signed-off-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
x86 arch defines desc_ptr for idt/gdt pointers, no need to define
another structure in kvm code.
Signed-off-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
A 16-bit TSS is only 44 bytes long. So make sure to test for the correct
size on task switch.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Int is not long enough to store the size of a dirty bitmap.
This patch fixes this problem with the introduction of a wrapper
function to calculate the sizes of dirty bitmaps.
Note: in mark_page_dirty(), we have to consider the fact that
__set_bit() takes the offset as int, not long.
Signed-off-by: Takuya Yoshikawa <yoshikawa.takuya@oss.ntt.co.jp>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
There is a quirk for AMD K8 CPUs in many Linux kernels (see
arch/x86/kernel/cpu/mcheck/mce.c:__mcheck_cpu_apply_quirks()) that
clears bit 10 in that MCE related MSR. KVM can only cope with all
zeros or all ones, so it will inject a #GP into the guest, which
will let it panic.
So lets add a quirk to the quirk and ignore this single cleared bit.
This fixes -cpu kvm64 on all machines and -cpu host on K8 machines
with some guest Linux kernels.
Signed-off-by: Andre Przywara <andre.przywara@amd.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
complete_pio() may use slot table which is protected by srcu.
Signed-off-by: Gleb Natapov <gleb@redhat.com>
Cc: stable@kernel.org
Signed-off-by: Avi Kivity <avi@redhat.com>
Joerg Roedel
Avi Kivity
Lai Jiangshan
Gleb Natapov
Jan Kiszka
Jan Kiszka
Wei Yongjun
Xiao Guangrong
Takuya Yoshikawa
Andre Przywara
Zhang, Yanmin