Yes, seriously. There is no migration support at all for usb devices.
They loose state, especially the device address, and stop responding
because of that. Oops.
Luckily there is so much broken usb hardware out there that the guest
usually just kicks the device hard (via port reset and
reinitialization), then continues without a hitch. So we got away with
that in a surprising high number of cases.
The arrival of remote wakeup (which enables autosuspend support) changes
that picture though. The usb devices also forget that it they are
supposed to wakeup, so they don't do that. The host also doesn't notice
the device stopped working in case it suspended the device and thus
expects it waking up instead of polling it. Result is that your mouse
is dead.
Lets start fixing that. Add a vmstate struct for USBDevice.
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
This patch makes the numlock+capslock tracking logic only look at
keydown events. Without this patch the vnc server will insert
bogous capslock keypress in case it sees the following key sequence:
shift down --- 'A' down --- shift up --- 'A' up
^ here
It doesn't hurt with a PS/2 keyboard, but it disturbs the USB Keyboard.
And with the key event queue just added to the usb keyboard the guest
will actually notice.
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
This patch moves the 'head' and 'n' fields from USBMouseState and
USBKeyboardState to the common USBHIDState struct.
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
This patch adds a event queue to the usb keyboard. This makes sure the
guest will see all key events even if they come in bursts. With this
patch applied sending Ctrl-Alt-Del using vncviewer's F8 menu works.
Also with autosuspend enabled the first keypress on a suspended keyboard
takes a little longer to be delivered to the guest because the usb bus
must be resumed first. Without event queue this easily gets lost.
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
The polling nature of the USB HID device makes it very hard to double
click or drag while on a high-latency VNC connection. This patch,
based on work done in the Xen qemu-dm tree by Ian Jackson, fixes this
bug by adding an event queue to the device. The event queue associates
each movement with the correct button state, and remembers all button
presses and releases as well.
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com>
Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Gerd Hoffman <kraxel@redhat.com>
The backing format should be honored during image creation. For some
reason we currently use the image format to open the backing file. This
fails when the backing file has a different format than the image being
created. Keep the image and backing format drivers completely separate.
Also print the backing filename if there is an error opening the backing
file instead of the image filename.
Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
Acked-by: Jes Sorensen <Jes.Sorensen@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Watch this:
(qemu) drive_add 0 if=none,file=tmp.img
OK
(qemu) info block
none0: type=hd removable=0 file=tmp.img ro=0 drv=raw encrypted=0
(qemu) drive_del none0
Segmentation fault (core dumped)
do_drive_del()'s code to clean up the pointer from a qdev using the
drive back to the drive needs to check whether such a device exists.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This makes the errors point to the error location, and fixes drive_add
to report errors in the monitor instead of stderr.
While there, tweak a few error messages for consistency.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
When cyls, heads or secs are out of range, the error message prints
buf, which points to the value of option "if". Bogus, may even be
null. Drop that.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
b02bea3a85 added a check on the return
value of bdrv_write and aborts migration when it fails. However, if the
size of the block device to migrate is not a multiple of BLOCK_SIZE
(currently 1 MB), the last bdrv_write will fail with -EIO.
Fixed by calling bdrv_write with the correct size of the last block.
Signed-off-by: Pierre Riteau <Pierre.Riteau@irisa.fr>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
QED relies on the underlying filesystem to extend the file and maintain
its size. Check that images are not created on a block device.
Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
With bm == NULL, other code in the same function would crash.
This bug was reported by cppcheck:
hw/ide/pci.c:280: error: Possible null pointer dereference: bm
Cc: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Stefan Weil <weil@mail.berlios.de>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
qcow2 calls bdrv_flush() after performing COW in order to ensure that the
L2 table change is never written before the copy is safe on disk. Now that the
L2 table is cached, we can wait with flushing until we write out the next L2
table.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Adding a chardev backend for spice, where spice determines what
to do with it based on the name attribute given during chardev creation.
For usage by spice vdagent in conjunction with a properly named
virtio-serial device, and future smartcard channel usage.
Example usage:
qemu -device virtio-serial -chardev spicevmc,name=vdagent,id=vdagent \
-device virtserialport,chardev=vdagent,name=com.redhat.spice.0
v4->v5:
* add tracing events
* fix missing comma
* fix help string to show debug is optional
v3->v4:
* updated commit message
v1->v3 changes: (v2 had a wrong commit message)
* removed spice-qemu-char.h, folded into ui/qemu-spice.h
* removed dead IOCTL code
* removed comment
* removed ifdef CONFIG_SPICE from qemu-config.c and qemu-options.hx help.
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Non-existent $pkgconfig instead of $pkg_config was used when configure
probes for spice availability.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Live migration from and to spice 0.4 qxl devices isn't going to work.
Rip out the bits which attempt to support that. Zap the subsection
logic which is obsolete now. Bumb the version to make a clean cut.
This should obviously go in before 0.14 is released.
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Handle spice client migration, i.e. inform a spice client connected
about the new host and connection parameters, so it can move over the
connection automatically.
The monitor command has a not-yet used protocol argument simliar to
set_password and expire_password commands. This allows to add a simliar
feature to vnc in the future. Daniel Berrange plans to work on this.
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
This patch adds functions to register and unregister notifiers for
migration state changes and a function to query the migration state.
The notifier is called on every state change. Once after establishing a
new migration object (which is in active state then) and once when the
state changes from active to completed, canceled or error.
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
PCIDevice::rom_file is leaked.
PCIDevice::rom_file is allocated in pci_qdev_init(), but not freed anywhere.
free it in qemu_unregister_device().
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
This adds some new cache functions to qcow2 which can be used for caching
refcount blocks and L2 tables. When used with cache=writethrough they work
like the old caching code which is spread all over qcow2, so for this case we
have merely a cleanup.
The interesting case is with writeback caching (this includes cache=none) where
data isn't written to disk immediately but only kept in cache initially. This
leads to some form of metadata write batching which avoids the current "write
to refcount block, flush, write to L2 table" pattern for each single request
when a lot of cluster allocations happen. Instead, cache entries are only
written out if its required to maintain the right order. In the pure cluster
allocation case this means that all metadata updates for requests are done in
memory initially and on sync, first the refcount blocks are written to disk,
then fsync, then L2 tables.
This improves performance of scenarios with lots of cluster allocations
noticably (e.g. installation or after taking a snapshot).
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Merge ide_dma_submit_check into it's only caller. Also use tail recursion
using a goto instead of a real recursion - this avoid overflowing the
stack in the pathological situation of an recurring error that is ignored.
We'll still be busy looping in ide_dma_cb, but at least won't eat up
all stack space after this.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Currenly the code only resets the io_buffer_index field for reads,
but the code seems to expect this for all types of I/O. I guess
we simply don't hit large enough transfers that would require this
often enough.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Factor the DMA I/O path that is duplicated between read and write
commands, into common helpers using the s->is_read flag added for
the macio ATA controller.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
When block migration is requested and no read-write block device is
present, a divide by zero exception is triggered because
total_sector_sum equals zero.
Signed-off-by: Pierre Riteau <Pierre.Riteau@irisa.fr>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
strtosz() needs to return a 64 bit type even on 32 bit
architectures. Otherwise qemu-img will fail to create disk
images >= 2GB
Signed-off-by: Jes Sorensen <Jes.Sorensen@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Current code does not support snapshot internally to the running
image. Error in case no snapshot_file is specified.
Signed-off-by: Jes Sorensen <Jes.Sorensen@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
None of the other qemu-img subcommands uses writethrough, and there's no reason
why snapshot should be special.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
cpu_to_be64w() is called with an obviously non-aligned pointer. Use
cpu_to_be64wu() instead. It fixes unaligned accesses errors on IA64
hosts.
Cc: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
The upper memory size field should exclude the first MB of RAM.
Signed-off-by: Kevin Wolf <mail@kevin-wolf.de>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
Include a header to get the declaration for xml_builtin. This
avoids a warning from sparse:
CC m68k-softmmu/gdbstub-xml.o
gdbstub-xml.c:244:12: warning: symbol 'xml_builtin' was not declared. Should it be static?
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
If the kernel does not support KVM_CAP_ASYNC_PF, it also does not know
about the related MSR. So skip it during state synchronization in that
case. Fixes annoying kernel warnings.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
All CPUX86State variables before CPU_COMMON are automatically cleared on
reset. Reorder nmi_injected and nmi_pending to avoid having to touch
them explicitly.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
The imbalance in the hold time of qemu_global_mutex only exists in TCG
mode. In contrast to TCG VCPUs, KVM drops the global lock during guest
execution. We already avoid touching the fairness lock from the
IO-thread in KVM mode, so also stop using it from the VCPU threads.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
We must flush pending mmio writes if we leave kvm_cpu_exec for an IO
window. Otherwise we risk to loose those requests when migrating to a
different host during that window.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
In order to support loading BIOSes > 256K, reorder the code, adjusting
the base if the kernel supports moving the identity map.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Instead of splattering the code with #ifdefs and runtime checks for
capabilities we cannot work without anyway, provide central test
infrastructure for verifying their availability both at build and
runtime.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
If we lack kvm_para.h, MSR_KVM_ASYNC_PF_EN is not defined. The change in
kvm_arch_init_vcpu is just for consistency reasons.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Make sure to write the cleared MSR_KVM_SYSTEM_TIME, MSR_KVM_WALL_CLOCK,
and MSR_KVM_ASYNC_PF_EN to the kernel state so that a freshly booted
guest cannot be disturbed by old values.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
CC: Glauber Costa <glommer@redhat.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Simplify kvm_has_msr_star/hsave_pa to booleans and push their one-time
initialization into kvm_arch_init. Also handle potential errors of that
setup procedure.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
For unknown reasons, xcr0 reset ended up in kvm_arch_update_guest_debug
on upstream merge. Fix this and also remove the misleading comment (1 is
THE reset value).
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
kvm_arch_reset_vcpu initializes mp_state, and that function is invoked
right after kvm_arch_init_vcpu.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
This code path will not yet be taken as we still lack in-kernel irqchip
support. But qemu-kvm can already make use of it and drop its own
mp_state access services.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
The ordering doesn't matter in this case, but better keep it consistent.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
