* qemu-kvm/memory/core: (30 commits)
memory: allow phys_map tree paths to terminate early
memory: unify PhysPageEntry::node and ::leaf
memory: change phys_page_set() to set multiple pages
memory: switch phys_page_set() to a recursive implementation
memory: replace phys_page_find_alloc() with phys_page_set()
memory: simplify multipage/subpage registration
memory: give phys_page_find() its own tree search loop
memory: make phys_page_find() return a MemoryRegionSection
memory: move tlb flush to MemoryListener commit callback
memory: unify the two branches of cpu_register_physical_memory_log()
memory: fix RAM subpages in newly initialized pages
memory: compress phys_map node pointers to 16 bits
memory: store MemoryRegionSection pointers in phys_map
memory: unify phys_map last level with intermediate levels
memory: remove first level of l1_phys_map
memory: change memory registration to rebuild the memory map on each change
memory: support stateless memory listeners
memory: split memory listener for the two address spaces
xen: ignore I/O memory regions
memory: allow MemoryListeners to observe a specific address space
...
kvm_set_phys_mem() may be passed sections that are not aligned to a page
boundary. The current code simply brute-forces the alignment which leads
to an inconsistency and an abort().
Fix by aligning the start and the end of the section correctly, discarding
and unaligned head or tail.
This was triggered by a guest sizing a 64-bit BAR that is smaller than a page
with PCI_COMMAND_MEMORY enabled and the upper dword clear.
Signed-off-by: Avi Kivity <avi@redhat.com>
Current memory listeners are incremental; that is, they are expected to
maintain their own state, and receive callbacks for changes to that state.
This patch adds support for stateless listeners; these work by receiving
a ->begin() callback (which tells them that new state is coming), a
sequence of ->region_add() and ->region_nop() callbacks, and then a
->commit() callback which signifies the end of the new state. They should
ignore ->region_del() callbacks.
Signed-off-by: Avi Kivity <avi@redhat.com>
This allows reverse iteration, which in turns allows consistent ordering
among multiple listeners:
l1->add
l2->add
l2->del
l1->del
Signed-off-by: Avi Kivity <avi@redhat.com>
Reviewed-by: Richard Henderson <rth@twiddle.net>
As we have thread-local cpu_single_env now and KVM uses exactly one
thread per VCPU, we can drop the cpu_single_env updates from the loop
and initialize this variable only once during setup.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
To both avoid that kvm_irqchip_in_kernel always has to be paired with
kvm_enabled and that the former ends up in a function call, implement it
like the latter. This means keeping the state in a global variable and
defining kvm_irqchip_in_kernel as a preprocessor macro.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Commit 84b058d broke compilation for KVM on non-x86 targets, which
don't have KVM_CAP_IRQ_ROUTING defined.
Fix by not using the unavailable constant when it's not around.
Signed-off-by: Alexander Graf <agraf@suse.de>
Instead of each target knowing or guessing the guest page size,
just pass the desired size of dirtied memory area.
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
* qemu-kvm/uq/master:
kvm: Activate in-kernel irqchip support
kvm: x86: Add user space part for in-kernel IOAPIC
kvm: x86: Add user space part for in-kernel i8259
kvm: x86: Add user space part for in-kernel APIC
kvm: x86: Establish IRQ0 override control
kvm: Introduce core services for in-kernel irqchip support
memory: Introduce memory_region_init_reservation
ioapic: Factor out base class for KVM reuse
ioapic: Drop post-load irr initialization
i8259: Factor out base class for KVM reuse
i8259: Completely privatize PicState
apic: Open-code timer save/restore
apic: Factor out base class for KVM reuse
apic: Introduce apic_report_irq_delivered
apic: Inject external NMI events via LINT1
apic: Stop timer on reset
kvm: Move kvmclock into hw/kvm folder
msi: Generalize msix_supported to msi_supported
hyper-v: initialize Hyper-V CPUID leaves.
hyper-v: introduce Hyper-V support infrastructure.
Conflicts:
Makefile.target
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
The kvm_get_dirty_pages_log_range() function uses two address
variables to step through the monitored memory region to update the
dirty log. However, these variables have type unsigned long, which
can overflow if running a 64-bit guest with a 32-bit qemu binary.
This patch changes these to target_phys_addr_t which will have the
correct size.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Alexander Graf <agraf@suse.de>
KVM is forced to disable the IRQ0 override when we run with in-kernel
irqchip but without IRQ routing support of the kernel. Set the fwcfg
value correspondingly. This aligns us with qemu-kvm.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Add the basic infrastructure to active in-kernel irqchip support, inject
interrupts into these models, and maintain IRQ routes.
Routing is optional and depends on the host arch supporting
KVM_CAP_IRQ_ROUTING. When it's not available on x86, we looe the HPET as
we can't route GSI0 to IOAPIC pin 2.
In-kernel irqchip support will once be controlled by the machine
property 'kernel_irqchip', but this is not yet wired up.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Otherwise, the dirty log information is lost in the kernel forever.
Fixes opensuse-12.1 boot screen, which changes the vga windows rapidly.
Signed-off-by: Avi Kivity <avi@redhat.com>
Drop the use of cpu_register_phys_memory_client() in favour of the new
MemoryListener API. The new API simplifies the caller, since there is no
need to deal with splitting and merging slots; however this is not exploited
in this patch.
Signed-off-by: Avi Kivity <avi@redhat.com>
It's a little unfriendly to call abort() without printing any sort of
error message. So turn the DPRINTK() into an fprintf(stderr, ...).
Signed-off-by: Michael Ellerman <michael@ellerman.id.au>
Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
And kvm_ioctl(s, KVM_CREATE_VM, 0)'s return value can be < -1,
so change the check of vmfd at label 'err'.
Signed-off-by: Xu He Jie <xuhj@linux.vnet.ibm.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
mmio callbacks invoked by kvm_flush_coalesced_mmio_buffer() may
themselves indirectly call kvm_flush_coalesced_mmio_buffer().
Prevent reentering the function by checking a flag that indicates
we're processing coalesced mmio requests.
Signed-off-by: Avi Kivity <avi@redhat.com>
Next commit will convert the query-status command to use the
RunState type as generated by the QAPI.
In order to "transparently" replace the current enum by the QAPI
one, we have to make some changes to some enum values.
As the changes are simple renames, I'll do them in one shot. The
changes are:
- Rename the prefix from RSTATE_ to RUN_STATE_
- RUN_STATE_SAVEVM to RUN_STATE_SAVE_VM
- RUN_STATE_IN_MIGRATE to RUN_STATE_INMIGRATE
- RUN_STATE_PANICKED to RUN_STATE_INTERNAL_ERROR
- RUN_STATE_POST_MIGRATE to RUN_STATE_POSTMIGRATE
- RUN_STATE_PRE_LAUNCH to RUN_STATE_PRELAUNCH
- RUN_STATE_PRE_MIGRATE to RUN_STATE_PREMIGRATE
- RUN_STATE_RESTORE to RUN_STATE_RESTORE_VM
- RUN_STATE_PRE_MIGRATE to RUN_STATE_FINISH_MIGRATE
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Today, when notifying a VM state change with vm_state_notify(),
we pass a VMSTOP macro as the 'reason' argument. This is not ideal
because the VMSTOP macros tell why qemu stopped and not exactly
what the current VM state is.
One example to demonstrate this problem is that vm_start() calls
vm_state_notify() with reason=0, which turns out to be VMSTOP_USER.
This commit fixes that by replacing the VMSTOP macros with a proper
state type called RunState.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Enabling the I/O thread by default seems like an important part of declaring
1.0. Besides allowing true SMP support with KVM, the I/O thread means that the
TCG VCPU doesn't have to multiplex itself with the I/O dispatch routines which
currently requires a (racey) signal based alarm system.
I know there have been concerns about performance. I think so far the ones that
have come up (virtio-net) are most likely due to secondary reasons like
decreased batching.
I think we ought to force enabling I/O thread early in 1.0 development and
commit to resolving any lingering issues.
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
No longer needed with accompanied kernel headers. We are only left with
build dependencies that are controlled by kvm arch headers.
CC: Alexander Graf <agraf@suse.de>
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
On PPC, the default PAGE_SIZE is 64kb. Unfortunately, the hardware
alignments don't match here: There are RAM and MMIO regions within
a single page when it's 64kb in size.
So the only way out for now is to tell the user that he should use 4k
PAGE_SIZE.
This patch gives the user a hint on that, telling him that failing to
register a prefix slot is most likely to be caused by mismatching PAGE_SIZE.
This way it's also more future-proof, as bigger PAGE_SIZE can easily be
supported by other machines then, as long as they stick to 64kb granularities.
Signed-off-by: Alexander Graf <agraf@suse.de>
This change fixes a long-standing immediate crash (memory corruption
and abort in glibc malloc code) in migration on 32bits.
The bug is present since this commit:
commit 692d9aca97b865b0f7903565274a52606910f129
Author: Bruce Rogers <brogers@novell.com>
Date: Wed Sep 23 16:13:18 2009 -0600
qemu-kvm: allocate correct size for dirty bitmap
The dirty bitmap copied out to userspace is stored in a long array,
and gets copied out to userspace accordingly. This patch accounts
for that correctly. Currently I'm seeing kvm crashing due to writing
beyond the end of the alloc'd dirty bitmap memory, because the buffer
has the wrong size.
Signed-off-by: Bruce Rogers
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
--- a/qemu-kvm.c
+++ b/qemu-kvm.c
@@ int kvm_get_dirty_pages_range(kvm_context_t kvm, unsigned long phys_addr,
- buf = qemu_malloc((slots[i].len / 4096 + 7) / 8 + 2);
+ buf = qemu_malloc(BITMAP_SIZE(slots[i].len));
r = kvm_get_map(kvm, KVM_GET_DIRTY_LOG, i, buf);
BITMAP_SIZE is now open-coded in that function, like this:
size = ALIGN(((mem->memory_size) >> TARGET_PAGE_BITS), HOST_LONG_BITS) / 8;
The problem is that HOST_LONG_BITS in 32bit userspace is 32
but it's 64 in 64bit kernel. So userspace aligns this to
32, and kernel to 64, but since no length is passed from
userspace to kernel on ioctl, kernel uses its size calculation
and copies 4 extra bytes to userspace, corrupting memory.
Here's how it looks like during migrate execution:
our=20, kern=24
our=4, kern=8
...
our=4, kern=8
our=4064, kern=4064
our=512, kern=512
our=4, kern=8
our=20, kern=24
our=4, kern=8
...
our=4, kern=8
our=4064, kern=4064
*** glibc detected *** ./x86_64-softmmu/qemu-system-x86_64: realloc(): invalid next size: 0x08f20528 ***
(our is userspace size above, kern is the size as calculated
by the kernel).
Fix this by always aligning to 64 in a hope that no platform will
have sizeof(long)>8 any time soon, and add a comment describing it
all. It's a small price to pay for bad kernel design.
Alternatively it's possible to fix that in the kernel by using
different size calculation depending on the current process.
But this becomes quite ugly.
Special thanks goes to Stefan Hajnoczi for spotting the fundamental
cause of the issue, and to Alexander Graf for his support in #qemu.
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
CC: Bruce Rogers <brogers@novell.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
KVM only requires to set the raised IRQ in CPUState and to kick the
receiving vcpu if it is remote. Installing a specialized handler allows
potential future changes to the TCG code path without risking KVM side
effects.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
use the new api to reduce the number of these (expensive)
system calls.
Note: using this API, we should be able to
get rid of vga_dirty_log_xxx APIs. Using them doesn't
affect the performance though because we detects
the log_dirty flag set and ignores the call.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
There are no generic bits remaining in the handling of KVM_EXIT_DEBUG.
So push its logic completely into arch hands, i.e. only x86 so far.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Make the return code of kvm_arch_handle_exit directly usable for
kvm_cpu_exec. This is straightforward for x86 and ppc, just s390
would require more work. Avoid this for now by pushing the return code
translation logic into s390's kvm_arch_handle_exit.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
CC: Alexander Graf <agraf@suse.de>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Let kvm_cpu_exec return EXCP_* values consistently and generate those
codes already inside its inner loop. This means we will now re-enter the
kernel while ret == 0.
Update kvm_handle_internal_error accordingly, but keep
kvm_arch_handle_exit untouched, it will be converted in a separate step.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Test for general errors first as this is the slower path.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Avoid using 'ret' both for the return value of KVM_RUN as well as the
code kvm_cpu_exec is supposed to return. Both have no direct relation.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Without KVM_CAP_SET_GUEST_DEBUG, we neither motivate the kernel to
report KVM_EXIT_DEBUG nor do we expect such exits. So fall through to
the arch code which will simply report an unknown exit reason.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
This avoids that early cpu_synchronize_state calls try to retrieve an
uninitialized state from the kernel. That even causes a deadlock if
io-thread is enabled.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
We will broaden the scope of this function on x86 beyond irqchip events.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Original fix by David Gibson.
CC: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
KVM-assisted devices need access to it but we have no clean channel to
distribute a reference. As a workaround until there is a better
solution, export kvm_state for global use, though use should remain
restricted to the mentioned scenario.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
In order to use log_start/log_stop with Xen as well in the vga code,
this two operations have been put in CPUPhysMemoryClient.
The two new functions cpu_physical_log_start,cpu_physical_log_stop are
used in hw/vga.c and replace the kvm_log_start/stop. With this, vga does
no longer depends on kvm header.
[ Jan: rebasing and style fixlets ]
Signed-off-by: Anthony PERARD <anthony.perard@citrix.com>
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
The number of slots and the location of private ones changed several
times in KVM's early days. However, it's stable since 2.6.29 (our
required baseline), and slots 8..11 are no longer reserved since then.
So remove this unneeded restriction.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
CC: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Mixing up TCG bits with KVM already led to problems around eflags
emulation on x86. Moreover, quite some code that TCG requires on cpu
enty/exit is useless for KVM. So dispatch between tcg_cpu_exec and
kvm_cpu_exec as early as possible.
The core logic of cpu_halted from cpu_exec is added to
kvm_arch_process_irqchip_events. Moving away from cpu_exec makes
exception_index meaningless for KVM, we can simply pass the exit reason
directly (only "EXCP_DEBUG vs. rest" is relevant).
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
