Zdenek reported a bug where a looping "dmsetup status" eventually hangs
on SMP guests.
The problem is that kvm_mmu_get_page() prepopulates the shadow MMU
before write protecting the guest page tables. By doing so, it leaves a
window open where the guest can mark a pte as present while the host has
shadow cached such pte as "notrap". Accesses to such address will fault
in the guest without the host having a chance to fix the situation.
Fix by moving the write protection before the pte prefetch.
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
If the accessed bit is not set, the guest has never accessed this page
(at least through this spte), so there's no need to mark the page
accessed. This provides more accurate data for the eviction algortithm.
Noted by Andrea Arcangeli.
Signed-off-by: Avi Kivity <avi@qumranet.com>
the main purpose of adding this functions is the abilaty to release the
spinlock that protect the kvm list while still be able to do operations
on a specific kvm in a safe way.
Signed-off-by: Izik Eidus <izike@qumranet.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
The kvm_host.h file for x86 declares the functions kvm_set_cr[0348]. In the
header file their second parameter is named cr0 in all cases. This patch
renames the parameters so that they match the function name.
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
Allow the Linux memory manager to reclaim memory in the kvm shadow cache.
Signed-off-by: Izik Eidus <izike@qumranet.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
Unify slots_lock acquision around vcpu_run(). This is simpler and less
error-prone.
Also fix some callsites that were not grabbing the lock properly.
[avi: drop slots_lock while in guest mode to avoid holding the lock
for indefinite periods]
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
MSR Bitmap controls whether the accessing of an MSR causes VM Exit.
Eliminating exits on automatically saved and restored MSRs yields a
small performance gain.
Signed-off-by: Sheng Yang <sheng.yang@intel.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
This patch implements kvm guest kernel support for paravirtualized devices
and contains two parts:
o a basic virtio stub using virtio_ring and external interrupts and hypercalls
o full hypercall implementation in kvm_para.h
Currently we dont have PCI on s390. Making virtio_pci usable for s390 seems
more complicated that providing an own stub. This virtio stub is similar to
the lguest one, the memory for the descriptors and the device detection is made
via additional mapped memory on top of the guest storage. We use an external
interrupt with extint code 0x2603 for host->guest notification.
The hypercall definition uses the diag instruction for issuing a hypercall. The
parameters are written in R2-R7, the hypercall number is written in R1. This is
similar to the system call ABI (svc) which can use R1 for the number and R2-R6
for the parameters.
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Acked-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Carsten Otte <cotte@de.ibm.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
This patch adds functionality to detect if the kernel runs under the KVM
hypervisor. A macro MACHINE_IS_KVM is exported for device drivers. This
allows drivers to skip device detection if the systems runs non-virtualized.
We also define a preferred console to avoid having the ttyS0, which is a line
mode only console.
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Acked-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Carsten Otte <cotte@de.ibm.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
This patch adds an entry for kvm on s390 to the MAINTAINERS file :-). We intend
to push all patches regarding this via Avi's kvm.git.
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Carsten Otte <cotte@de.ibm.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
This patch adds Documentation/s390/kvm.txt, which describes specifics of kvm's
user interface that are unique to s390 architecture.
Signed-off-by: Carsten Otte <cotte@de.ibm.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
This patch adds the virtualization submenu and the kvm option to the kernel
config. It also defines HAVE_KVM for 64bit kernels.
Acked-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Carsten Otte <cotte@de.ibm.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
This patch introduces interpretation of some diagnose instruction intercepts.
Diagnose is our classic architected way of doing a hypercall. This patch
features the following diagnose codes:
- vm storage size, that tells the guest about its memory layout
- time slice end, which is used by the guest to indicate that it waits
for a lock and thus cannot use up its time slice in a useful way
- ipl functions, which a guest can use to reset and reboot itself
In order to implement ipl functions, we also introduce an exit reason that
causes userspace to perform various resets on the virtual machine. All resets
are described in the principles of operation book, except KVM_S390_RESET_IPL
which causes a reboot of the machine.
Acked-by: Martin Schwidefsky <martin.schwidefsky@de.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Carsten Otte <cotte@de.ibm.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
This patch introduces in-kernel handling of _some_ sigp interprocessor
signals (similar to ipi).
kvm_s390_handle_sigp() decodes the sigp instruction and calls individual
handlers depending on the operation requested:
- sigp sense tries to retrieve information such as existence or running state
of the remote cpu
- sigp emergency sends an external interrupt to the remove cpu
- sigp stop stops a remove cpu
- sigp stop store status stops a remote cpu, and stores its entire internal
state to the cpus lowcore
- sigp set arch sets the architecture mode of the remote cpu. setting to
ESAME (s390x 64bit) is accepted, setting to ESA/S390 (s390, 31 or 24 bit) is
denied, all others are passed to userland
- sigp set prefix sets the prefix register of a remote cpu
For implementation of this, the stop intercept indication starts to get reused
on purpose: a set of action bits defines what to do once a cpu gets stopped:
ACTION_STOP_ON_STOP really stops the cpu when a stop intercept is recognized
ACTION_STORE_ON_STOP stores the cpu status to lowcore when a stop intercept is
recognized
Acked-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Carsten Otte <cotte@de.ibm.com>
Signed-off-by: Carsten Otte <cotte@de.ibm.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
This patch introduces in-kernel handling of some intercepts for privileged
instructions:
handle_set_prefix() sets the prefix register of the local cpu
handle_store_prefix() stores the content of the prefix register to memory
handle_store_cpu_address() stores the cpu number of the current cpu to memory
handle_skey() just decrements the instruction address and retries
handle_stsch() delivers condition code 3 "operation not supported"
handle_chsc() same here
handle_stfl() stores the facility list which contains the
capabilities of the cpu
handle_stidp() stores cpu type/model/revision and such
handle_stsi() stores information about the system topology
Acked-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Carsten Otte <cotte@de.ibm.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
This patch contains the s390 interrupt subsystem (similar to in kernel apic)
including timer interrupts (similar to in-kernel-pit) and enabled wait
(similar to in kernel hlt).
In order to achieve that, this patch also introduces intercept handling
for instruction intercepts, and it implements load control instructions.
This patch introduces an ioctl KVM_S390_INTERRUPT which is valid for both
the vm file descriptors and the vcpu file descriptors. In case this ioctl is
issued against a vm file descriptor, the interrupt is considered floating.
Floating interrupts may be delivered to any virtual cpu in the configuration.
The following interrupts are supported:
SIGP STOP - interprocessor signal that stops a remote cpu
SIGP SET PREFIX - interprocessor signal that sets the prefix register of a
(stopped) remote cpu
INT EMERGENCY - interprocessor interrupt, usually used to signal need_reshed
and for smp_call_function() in the guest.
PROGRAM INT - exception during program execution such as page fault, illegal
instruction and friends
RESTART - interprocessor signal that starts a stopped cpu
INT VIRTIO - floating interrupt for virtio signalisation
INT SERVICE - floating interrupt for signalisations from the system
service processor
struct kvm_s390_interrupt, which is submitted as ioctl parameter when injecting
an interrupt, also carrys parameter data for interrupts along with the interrupt
type. Interrupts on s390 usually have a state that represents the current
operation, or identifies which device has caused the interruption on s390.
kvm_s390_handle_wait() does handle waitpsw in two flavors: in case of a
disabled wait (that is, disabled for interrupts), we exit to userspace. In case
of an enabled wait we set up a timer that equals the cpu clock comparator value
and sleep on a wait queue.
[christian: change virtio interrupt to 0x2603]
Acked-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Carsten Otte <cotte@de.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
This path introduces handling of sie intercepts in three flavors: Intercepts
are either handled completely in-kernel by kvm_handle_sie_intercept(),
or passed to userspace with corresponding data in struct kvm_run in case
kvm_handle_sie_intercept() returns -ENOTSUPP.
In case of partial execution in kernel with the need of userspace support,
kvm_handle_sie_intercept() may choose to set up struct kvm_run and return
-EREMOTE.
The trivial intercept reasons are handled in this patch:
handle_noop() just does nothing for intercepts that don't require our support
at all
handle_stop() is called when a cpu enters stopped state, and it drops out to
userland after updating our vcpu state
handle_validity() faults in the cpu lowcore if needed, or passes the request
to userland
Acked-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Carsten Otte <cotte@de.ibm.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
This patch contains the port of Qumranet's kvm kernel module to IBM zSeries
(aka s390x, mainframe) architecture. It uses the mainframe's virtualization
instruction SIE to run virtual machines with up to 64 virtual CPUs each.
This port is only usable on 64bit host kernels, and can only run 64bit guest
kernels. However, running 31bit applications in guest userspace is possible.
The following source files are introduced by this patch
arch/s390/kvm/kvm-s390.c similar to arch/x86/kvm/x86.c, this implements all
arch callbacks for kvm. __vcpu_run calls back into
sie64a to enter the guest machine context
arch/s390/kvm/sie64a.S assembler function sie64a, which enters guest
context via SIE, and switches world before and after that
include/asm-s390/kvm_host.h contains all vital data structures needed to run
virtual machines on the mainframe
include/asm-s390/kvm.h defines kvm_regs and friends for user access to
guest register content
arch/s390/kvm/gaccess.h functions similar to uaccess to access guest memory
arch/s390/kvm/kvm-s390.h header file for kvm-s390 internals, extended by
later patches
Acked-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Carsten Otte <cotte@de.ibm.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
The address 0x11b8 is used by z/VM for pfault and diag 250 I/O to
provide a 64 bit extint parameter. virtio uses the same address, so
its time to update the lowcore structure.
Acked-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Carsten Otte <cotte@de.ibm.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
This patch changes the s390 memory management defintions to use the pgste field
for dirty and reference bit tracking of host and guest code. Usually on s390,
dirty and referenced are tracked in storage keys, which belong to the physical
page. This changes with virtualization: The guest and host dirty/reference bits
are defined to be the logical OR of the values for the mapping and the physical
page. This patch implements the necessary changes in pgtable.h for s390.
There is a common code change in mm/rmap.c, the call to
page_test_and_clear_young must be moved. This is a no-op for all
architecture but s390. page_referenced checks the referenced bits for
the physiscal page and for all mappings:
o The physical page is checked with page_test_and_clear_young.
o The mappings are checked with ptep_test_and_clear_young and friends.
Without pgstes (the current implementation on Linux s390) the physical page
check is implemented but the mapping callbacks are no-ops because dirty
and referenced are not tracked in the s390 page tables. The pgstes introduces
guest and host dirty and reference bits for s390 in the host mapping. These
mapping must be checked before page_test_and_clear_young resets the reference
bit.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Acked-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Acked-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Carsten Otte <cotte@de.ibm.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
The SIE instruction on s390 uses the 2nd half of the page table page to
virtualize the storage keys of a guest. This patch offers the s390_enable_sie
function, which reorganizes the page tables of a single-threaded process to
reserve space in the page table:
s390_enable_sie makes sure that the process is single threaded and then uses
dup_mm to create a new mm with reorganized page tables. The old mm is freed
and the process has now a page status extended field after every page table.
Code that wants to exploit pgstes should SELECT CONFIG_PGSTE.
This patch has a small common code hit, namely making dup_mm non-static.
Edit (Carsten): I've modified Martin's patch, following Jeremy Fitzhardinge's
review feedback. Now we do have the prototype for dup_mm in
include/linux/sched.h. Following Martin's suggestion, s390_enable_sie() does now
call task_lock() to prevent race against ptrace modification of mm_users.
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Carsten Otte <cotte@de.ibm.com>
Acked-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Avi Kivity <avi@qumranet.com>
This emulates the x86 hardware task switch mechanism in software, as it is
unsupported by either vmx or svm. It allows operating systems which use it,
like freedos, to run as kvm guests.
Signed-off-by: Izik Eidus <izike@qumranet.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
When mmu_set_spte() checks if a page related to spte should be release as
dirty or clean, it check if the shadow pte was writeble, but in case
rmap_write_protect() is called called it is possible for shadow ptes that were
writeble to become readonly and therefor mmu_set_spte will release the pages
as clean.
This patch fix this issue by marking the page as dirty inside
rmap_write_protect().
Signed-off-by: Izik Eidus <izike@qumranet.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
If we populate a shadow pte due to a fault (and not speculatively due to a
pte write) then we can set the accessed bit on it, as we know it will be
set immediately on the next guest instruction. This saves a read-modify-write
operation.
Signed-off-by: Avi Kivity <avi@qumranet.com>
include/linux/kvm.h defines struct kvm_dirty_log to
[...]
union {
void __user *dirty_bitmap; /* one bit per page */
__u64 padding;
};
__user requires compiler.h to compile. Currently, this works on x86
only coincidentally due to other include files. This patch makes
kvm.h compile in all cases.
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
This patch writes 0 (actually, what really matters is that the
LSB is cleared) to the system time msr before shutting down
the machine for kexec.
Without it, we can have a random memory location being written
when the guest comes back
It overrides the functions shutdown, used in the path of kernel_kexec() (sys.c)
and crash_shutdown, used in the path of crash_kexec() (kexec.c)
Signed-off-by: Glauber Costa <gcosta@redhat.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
it will allow external users to call it. It is mainly
useful for routines that will override its machine_ops
field for its own special purposes, but want to call the
normal shutdown routine after they're done
Signed-off-by: Glauber Costa <gcosta@redhat.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
This patch a llows machine_crash_shutdown to
be replaced, just like any of the other functions
in machine_ops
Signed-off-by: Glauber Costa <gcosta@redhat.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
Hypercall based pte updates are faster than faults, and also allow use
of the lazy MMU mode to batch operations.
Don't report the feature if two dimensional paging is enabled.
[avi:
- guest/host split
- fix 32-bit truncation issues
- adjust to mmu_op
- adjust to ->release_*() renamed
- add ->release_pud()]
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
Hypercall based pte updates are faster than faults, and also allow use
of the lazy MMU mode to batch operations.
Don't report the feature if two dimensional paging is enabled.
[avi:
- one mmu_op hypercall instead of one per op
- allow 64-bit gpa on hypercall
- don't pass host errors (-ENOMEM) to guest]
[akpm: warning fix on i386]
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Avi Kivity <avi@qumranet.com>
The patch moves the PIT model from userspace to kernel, and increases
the timer accuracy greatly.
[marcelo: make last_injected_time per-guest]
Signed-off-by: Sheng Yang <sheng.yang@intel.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Tested-and-Acked-by: Alex Davis <alex14641@yahoo.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
Most Intel hosts have a stable tsc, and playing with the offset only
reduces accuracy. By limiting tsc offset adjustment only to forward updates,
we effectively disable tsc offset adjustment on these hosts.
Signed-off-by: Avi Kivity <avi@qumranet.com>
In the current inject_page_fault path KVM only checks if there is another PF
pending and injects a DF then. But it has to check for a pending DF too to
detect a shutdown condition in the VCPU. If this is not detected the VCPU goes
to a PF -> DF -> PF loop when it should triple fault. This patch detects this
condition and handles it with an KVM_SHUTDOWN exit to userspace.
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
Since the size of kvm_regs is too big to allocate from kernel stack on ia64,
use kzalloc to allocate it.
Signed-off-by: Xiantao Zhang <xiantao.zhang@intel.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
Create large pages mappings if the guest PTE's are marked as such and
the underlying memory is hugetlbfs backed. If the largepage contains
write-protected pages, a large pte is not used.
Gives a consistent 2% improvement for data copies on ram mounted
filesystem, without NPT/EPT.
Anthony measures a 4% improvement on 4-way kernbench, with NPT.
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
