qemu: clean up target page usage in msix

Since cpu_register_phys_memory does not require size to be a multiple of
target page size, simply make msix page size 0x1000.  Do this in msix,
reverting part of 5e520a7d500ec2569d22d80f9ef4272a34cb3c80, as we no
longer have to pass target page around.

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

fix comment on cpu_register_physical_memory_offset

We don't require full pages in cpu_register_physical_memory,
except for RAM.

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

qemu/pci: reset device registers on bus reset

Reset BARs and a couple of other registers on bus reset, as per PCI
spec.

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

qemu/pci: refactor code/symbolic constants

refactor code slightly, adding symbolic constants and functions, and
using macros where possible.  This will also make following reset
patches easier.

No functional changes.

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

qemu/virtio: fix reset with device removal

virtio pci registers its own reset handler, but fails to unregister it,
which will lead to crashes after device removal.  Solve this problem by
switching to qdev reset handler, which is automatically unregistered.

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

qemu/qdev: type safety in reset handler

Add type safety to qdev reset handlers, by declaring them as
DeviceState * rather than void *.

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

serial: convert isa to qdev

Everything using standard isa I/O ports and IRQ windup is considerd
being an actual isa device.  That are all serial_init() users except
mips_mipssim() which seems to have a non-standard IRQ windup.

baud rate is fixed at 115200 now as no caller passed in something else.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

qdev: don't crash on unset drive properties.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

floppy: move dma setup + drive connect to fdctrl_init_common()

isa-fdc is completely qdev-ified with this patch applied, all
configuration uses properties.

sysbus-fdc needs dma_channel config fixed.  There is only one user
(mips_jazz) which uses dma channel 0.  Can anyone knowing this
platform suggest how to deal with it?  Is hardcoding fine?

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

floppy: add drive properties.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

store a pointer to QemuOpts in DeviceState, release it when zapping a device.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

allow if=none for drive_add

Allow adding unconnected host drives by specifying if=none like it is
possible with -drive.  They can be put in use with drive attributes,
like this:

  drive_add dummy if=none,id=mydisk,file=/some/disk.img
  device_add virtio-blk-pci,drive=mydisk

which is the monitor aequivalent to these command line switches:

  -drive if=none,id=mydisk,file=/some/disk.img
  -device virtio-blk-pci,drive=mydisk

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

refactor drive_hot_add

move pci device lookup into the "case IF_SCSI" section, so we
can do something else for other interface types.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

drive cleanup fixes.

Changes:
  * drive_uninit() wants a DriveInfo now.
  * drive_uninit() also calls bdrv_delete(),
    so callers don't need to do that.
  * drive_uninit() calls are moved over to the ->exit()
    callbacks, destroy_bdrvs() is zapped.
  * setting bdrv->private is not needed any more as the
    only user (destroy_bdrvs) is gone.
  * usb-storage needs no drive_uninit, scsi-disk will
    handle that.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

pci: windup acpi-based hotplug

Switch over acpi-based PCI hotplug for pc over to the new
qdev-based pci hotplugging.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

pci: hotplug windup

Create qdev infrastructure for pci hotplug.  PCI bus implementations
must register a handler for hotplug.  Creating a new PCI device will
automagically hot-plug it in case the PCI bus in question supports this.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

scsi: hotplug windup

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

usb: hotplug windup

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

qdev hotplug: infrastructure and monitor commands.

Adds device_add and device_del commands.  device_add accepts accepts
the same syntax like the -device command line switch.  device_del
expects a device id.  So you should tag your devices with ids if you
want to remove them later on, like this:

  device_add pci-ohci,id=ohci
  device_del ohci

Unplugging via pci_del or usb_del works too.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

switch qemu-config to qemu_error

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

usb: hook unplug into qdev, cleanups + fixes.

Hook into DeviceInfo->exit().

handle_destroy() must not free the state struct, this is handled
by the new usb_qdev_exit() function now.

qdev_free(usb_device) works now.

Fix usb hub to qdev_free() all connected devices on unplug.
Unplugging a usb hub works now.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

pci: move unregister from PCIDevice to PCIDeviceInfo

One more cleanup while being at it ;)

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

pci: use qdev for device destruction.

pci_unregister_device is static now and hooked into Devicestate->exit.
qdev_free(pci_device) works now.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

Implement scsi device destruction

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

Add exit callback to DeviceInfo.

This adds a exit callback for device destruction to DeviceInfo, so
we can hook cleanups into qdev device destruction.

Followup patches will put that into use.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

qdev: device free fixups.

Two bug fixes:
 * When freeing a device we unregister even stuff we didn't register in
   the first place because the ->init() callback failed.
 * When freeing a device with child busses attached, we fail to zap the
   child bus (and the devices attached to it).

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

parallel: convert isa to qdev

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

update pc-bios/bios.bin

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

update pcbios submodule to current master

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

piix_pci: kill PIIX3IrqState

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

convert pci bridge to qdev

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

support inplace allocation for pci bus, split irq init.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

switch ide bus to inplace allocation.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

switch usb bus to inplace allocation.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

switch scsi bus to inplace allocation.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

allow qdev busses allocations be inplace

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

temporary fix for on_vcpu

Recent changes made on_vcpu hit the abort() path, even with the IO thread
disabled. This is because cpu_single_env is no longer set when we call this
function. Although the correct fix is a little bit more complicated that that,
the recent thread in which I proposed qemu_queue_work (which fixes that, btw),
is likely to go on a quite different direction.

So for the benefit of those using guest debugging, I'm proposing this simple
fix in the interim.

Signed-off-by: Glauber Costa <glommer@redhat.com>
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

kvm: Fix guest single-stepping

Hopefully the last regression of 4c0960c0: KVM_SET_GUEST_DEBUG requires
properly synchronized guest registers (on x86: eflags) on entry.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

gdbstub: x86: Switch 64/32 bit registers dynamically

Commit 56aebc891674cd2d07b3f64183415697be200084 changed gdbstub in way
that debugging 32 or 16-bit guest code is no longer possible with qemu
for x86_64 guest CPUs. Since that commit, qemu only provides registers
sets for 64-bit, forcing current and foreseeable gdb to also switch its
architecture to 64-bit. And this breaks if the inferior is 32 or 16 bit.

No question, this is a gdb issue. But, as it was confirmed in several
discusssions with gdb people, it is a non-trivial thing to fix. So until
qemu finds a gdb version attach with a rework x86 support, we have to
work around it by switching the register layout as the guest switches
its execution mode between 16/32 and 64 bit.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

Enable host-clock-based RTC

Switch RTC emulations to the new host_clock instead of vm_clock by
default. This has the advantage that the emulated RTC will follow
automatically the host time while it might be tuned via NTP. vm_clock
can still be selected by passing '-rtc clock=vm' on the command line.

Note that some RTC emulations (at least M48T59) already use the host
time unconditionally while others (namely MC146818) do not. This patch
introduces the required infrastructure for selecting the base clock but
only converts MC146818 for now.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

Refactor RTC command line switches

Deprecate -localtime, -setdate and -rtc-td-hack in favor of a new
unified command line switch:

    -rtc [base=utc|localtime|date][,driftfix=none|slew]

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

Introduce QEMU_CLOCK_HOST

Despite its name QEMU_CLOCK_REALTIME is (normally) not using
CLOCK_REALTIME / the host system time as base. In order to allow also
non-trivial RTC emulations (MC146818) to follow the host time instead of
the virtual guest time, introduce the new clock type QEMU_CLOCK_HOST. It
is unconditionally based on CLOCK_REALTIME, thus will follow system time
changes of the host.

The only limitation of its current implementation is that pending
host_clock timers may not fire early if the host time is pushed forward
beyond their expiry. So far no urgent need to overcome this limitation
was identified, so it's left as simple as it is (expiry on next alarm
timer tick).

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

win32: Drop dead dyntick timer code

nearest_delta_us is calculated but not used. Drop it.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

Rename QEMU_TIMER_* to QEMU_CLOCK_*

These constants select clocks, not timers. And init_timers initializes
clocks.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

Fix exit on 'pci_add' Monitor command

If the user issues one of the following commands to the Monitor:

pci_add pci_addr=auto nic model=None
pci_add pci_addr=auto nic model=?

QEMU will exit, because the function used to perform sanity
checks (qemu_check_nic_model_list()) exits on error.

This function is used by the startup code, where it makes
sense to exit on error, but in the Monitor it doesn't.

Changing qemu_check_nic_model_list() to not exit on error
is not possible though, as it's used by the board init
code (the PC one), where all board specific code must have
void return.

The way I've chosen to fix this was to introduce a new function
called pci_nic_supported(), which checks if the NIC is supported
and returns true or false accordingly.

The new function is used only by the Monitor, it performs the
necessary check and returns an error in case the NIC is not
supported, thus qemu_check_nic_model_list()'s exit is never trigged.

The following should be observed:

1. Only the specified NIC is checked, the default one is assumed
to be supported

2. The NIC query command (model=?) won't work with pci_add, the
right way to do this with the Monitor is to add a new command

Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

Correctly free nd structure

When we "free" a NICInfo structure, we can leak pointers, since we don't do
much more than setting used = 0.

We free() the model parameter, but we don't set it to NULL. This means that
a new user of this structure will see garbage in there. It was not noticed
before because reusing a NICInfo is not that common, but it can be, for
users of device pci hotplug.

A user hit it, described at https://bugzilla.redhat.com/show_bug.cgi?id=524022

This patch memset's the whole structure, guaranteeing that anyone reusing it
will see a fresh NICinfo. Also, we free some other strings that are currently
leaking.

This codebase is quite old, so this patch should feed all stable trees.

Signed-off-by: Glauber Costa <glommer@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

do proper cpu_self check

Currently, our check for qemu_cpu_self only checks if there is a cpu
currently in execution (represented by cpu_single_env being set). While
this might be okay for tcg, it is certainly not okay for kvm, since multiple
cpus might be executing.

Instead, I propose we use pthread primitives to test if the caller thread is
the same as env->thread.

For tcg, it will have the same semantics as before, since all CPUStates will
point to the same thread, and we'll only have one in execution at a time.

Signed-off-by: Glauber Costa <glommer@mothafucka.localdomain>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

bdf: Remove last users of FALSE/TRUE

This brings bfd_boolean to the wonderful world of <stdbool.h>, it is needed
because it defines bdf_boolean as an enum with values true and false,
and some architectures use TRUE, FALSE and give problems when you try to use
<stdbool.h>

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

Add -Wold-style-* flags

This time, I add them in configure only if target compiler supports it

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

slirp: It needs to use QEMU_CFLAGS not CFLAGS

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

musicpal: Add VMState support

Register all relevant fields of Musicpal device states with the VMState
framework. This involves a few type changes of state variables.

Signed-off-by: Jan Kiszka <jan.kiszka@web.de>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

musicpal: Make PIT emulation more robust

Stop the periodic timers of the PIT on reset, disabling via the control
register and invalid parameters.

Signed-off-by: Jan Kiszka <jan.kiszka@web.de>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

musicpal: True reset support for audio device

Signed-off-by: Jan Kiszka <jan.kiszka@web.de>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

musicpal: True reset support for GPIO

Signed-off-by: Jan Kiszka <jan.kiszka@web.de>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

musicpal: Coding style fixes

Signed-off-by: Jan Kiszka <jan.kiszka@web.de>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

musicpal: Clean up typecasts

Signed-off-by: Jan Kiszka <jan.kiszka@web.de>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

musicpal: Rework GPIO input events

The qdev_gpio conversion of 343ec8e caused come polarity mismatch of key
event pins and left an overly complex solution behind. Take this chance
and refactor the GPIO input system of the Musicpal, moving it closer to
reality:
 - Instantiate all 32 GPIO input pins and do the routing only via
   qdev_connect_gpio_out.
 - Implement IMR and IER registers. They manage the GPIO pin IRQ. IMR
   seems to enable IRQs on rising edges, IER on falling ones. At least
   this matches what the Musicpal fireware require.
 - Move key pin logic inversion from the GPIO layer to musicpal_key.

Signed-off-by: Jan Kiszka <jan.kiszka@web.de>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

musicpal: Catch null TX qeueues

They likely represent invalid queues that should be skipped. We already
do this for RX queues. Wish I had a spec...

Credits go to malc for analyzing the issue and suggesting this fix.

Signed-off-by: Jan Kiszka <jan.kiszka@web.de>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

x86: port cpu to vmstate

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

vmstate: Add suppot for field_exist() test

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

x86: factor out cpu_get/put_fpreg()

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

x86: factor out cpu_get/put_mttr_var()

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

x86: factor out cpu_get/put_xmm_reg()

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

x86: factor out cpu_pre/post_load()

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

x86: factor out cpu_pre_save()

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

x86: port segments to vmstate

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

Add *TL functions to vmstate

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

x86: split MTRRVar union

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

x86: split FPReg union

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

x86: mcg_cap is never 0

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

x86: send mce_banks as an array

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

x86: mce_banks always have the same size

mce_banks is always MCE_BANKS_DEF * 4 in size, value never change

CC: Huang Ying <ying.huang@intel.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

x86: add fpregs_format_vmstate

Don't even ask, being able to load/save between 64<->80bit floats should be forbidden

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

x86: add pending_irq_vmstate to the state

It is needed to save the interrupt_bitmap

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

x86: add fptag_vmstate to the state

It is needed to store fptags

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

x86: fpus is uint16_t not unsigned int

We save more that fpus on that 16 bits (fpstt), we need an additional field

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

x86: fpuc is uint16_t not unsigned int

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

x86: make a20_mask int32_t

This makes the savevm code correct, and sign extensins gives us exactly
what we need (namely, sign extend to 64 bits when used with 64bit addresess.

Once there, change 0x100000 for 1 << 20, that maks all a20 use the same syntax.

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

x86: hflags is not modified at all, just save it directly

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

vmstate: remove i2c_slave_load/save

All its users moved to vmstate

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

vmstate: port lm832x device

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

lm832x: make fields to have the same types that they are saved/loaded

They were saved as uint8_t already.  To make things simpler, I just
reg == -1 used to indicate an error, I create LM832x_GENERAL_ERROR
with vale 0xff to represet it

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

vmstate: add support for arrays of pointers

We need this to send arrays of timers

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

vmstate: port twl92230 device

Just don't look.  struct tm members are ints' and they are sent as uint16_t.
VMState code complains as it should.  Have to create hacky int32_as_uint16
type.  Don't ever think about copying it

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

twl92230: change pwrbtn_state to uint8_t

its value is always the level of an interrupt, 0 or 1

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

vmstate: port tmp105 device

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

tmp105: change len and alorm to uint8_t

They were using only with very small integers, and they are sent/read as
bytes.  They can't become negative as far as I can see

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

vmstate: create VMSTATE_INT16_ARRAY

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

vmstate: port ssd0303 device

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

vmstate: port pxa2xx_i2c device

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

vmstate: create VMSTATE_STRUCT_POINTER

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

vmstate: port max7310 device

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

vmstate: port wm8750 device

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

vmstate: create VMSTATE_I2C_SLAVE

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

vmstate: add uint8 array

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

vmstate: port i2c_slave device

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

vmstate: port i2c_bus device

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

i2c: addresses are load/save as uint8_t values, change types to reflect this

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

qdev: Add support for uint8_t

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

vmstate: Add support for sending partial arrays

This one is needed for changees happening on fdc.  It allows you to send
arrays of structs whose size we want to send it is another field with type
uint8_t.  (If you have been able to read the whole sentence without
stoping for breathing, you can use it.

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>