pci: remove PCIBus::config_reg.

PCIBus::config_reg isn't used anymore, so remove it.

Signed-off-by: Isaku Yamahata <[email protected]>
Signed-off-by: Michael S. Tsirkin <[email protected]>

pci: remove unused macro

PCI_STATUS_DEVSEL is unused, and it also
has a different name in pci_regs.h
Remove.

Signed-off-by: Michael S. Tsirkin <[email protected]>
Acked-by: Juan Quintela <[email protected]>
Acked-by: Glauber Costa <[email protected]>

usb-ohci: symbolic names for pci registers

No functional changes. I verified that the generated binary
does not change.

Signed-off-by: Michael S. Tsirkin <[email protected]>
Acked-by: Juan Quintela <[email protected]>
Acked-by: Glauber Costa <[email protected]>

usb-uhci: symbolic names for pci registers

No functional changes. I verified that the generated binary
does not change.

Signed-off-by: Michael S. Tsirkin <[email protected]>
Acked-by: Juan Quintela <[email protected]>
Acked-by: Glauber Costa <[email protected]>

ac97: symbolic names for pci registers

No functional changes. I verified that the generated binary
does not change.

Signed-off-by: Michael S. Tsirkin <[email protected]>
Acked-by: Juan Quintela <[email protected]>
Acked-by: Glauber Costa <[email protected]>

wdt_i6300esb: symbolic names for pci registers

No functional changes. I verified that the generated binary
does not change.

Signed-off-by: Michael S. Tsirkin <[email protected]>
Acked-by: Juan Quintela <[email protected]>
Acked-by: Glauber Costa <[email protected]>

es1370: symbolic names for pci registers

No functional changes. I verified that the generated binary
does not change.

Signed-off-by: Michael S. Tsirkin <[email protected]>
Acked-by: Juan Quintela <[email protected]>
Acked-by: Glauber Costa <[email protected]>

pci: add another devsel macro

will be used by ensoniq emulation

Signed-off-by: Michael S. Tsirkin <[email protected]>
Acked-by: Juan Quintela <[email protected]>
Acked-by: Glauber Costa <[email protected]>

lsi: symbolic names for pci registers

No functional changes. I verified that the generated binary
does not change.

Signed-off-by: Michael S. Tsirkin <[email protected]>
Acked-by: Juan Quintela <[email protected]>
Acked-by: Glauber Costa <[email protected]>

vmware_vga: symbolic names for pci registers

No functional changes. I verified that the generated binary
does not change.

Signed-off-by: Michael S. Tsirkin <[email protected]>
Acked-by: Juan Quintela <[email protected]>
Acked-by: Glauber Costa <[email protected]>

cmd646: symbolic names for pci registers

No functional changes. I verified that the generated binary
does not change.

Signed-off-by: Michael S. Tsirkin <[email protected]>
Acked-by: Juan Quintela <[email protected]>
Acked-by: Glauber Costa <[email protected]>

piix: symbolic constants

No functional changes. I verified that the generated binary
does not change.

Signed-off-by: Michael S. Tsirkin <[email protected]>
Acked-by: Juan Quintela <[email protected]>
Acked-by: Glauber Costa <[email protected]>

eepro100: symbolic names for pci registers

No functional changes. I verified that the generated binary
does not change in meaningful ways.

Signed-off-by: Michael S. Tsirkin <[email protected]>
Acked-by: Juan Quintela <[email protected]>
Acked-by: Glauber Costa <[email protected]>

pci: add more status bits

will be used by eepro100.

Signed-off-by: Michael S. Tsirkin <[email protected]>
Acked-by: Juan Quintela <[email protected]>
Acked-by: Glauber Costa <[email protected]>

pcnet: switch to symbolic names for pci registers

No functional changes. I verified that the generated binary
does not change.

Signed-off-by: Michael S. Tsirkin <[email protected]>
Acked-by: Juan Quintela <[email protected]>
Acked-by: Glauber Costa <[email protected]>

rtl: switch to symbolic names for pci registers

No functional changes. I verified that the generated binary
does not change.

Signed-off-by: Michael S. Tsirkin <[email protected]>
Acked-by: Juan Quintela <[email protected]>
Acked-by: Glauber Costa <[email protected]>

ne2000: switch to symbolic names for pci registers

No functional changes. I verified that the generated binary
does not change.

Signed-off-by: Michael S. Tsirkin <[email protected]>
Acked-by: Juan Quintela <[email protected]>
Acked-by: Glauber Costa <[email protected]>

e1000: switch to symbolic names for pci registers

No functional changes. I verified that the generated
object binary does not change.

Signed-off-by: Michael S. Tsirkin <[email protected]>
Acked-by: Juan Quintela <[email protected]>
Acked-by: Glauber Costa <[email protected]>

Revert "Revert "pci: interrupt disable bit support""

This reverts commit d587e0787153f0224a6140c5015609963ceaabfb.

USB: Improve usbdevice error messages

When an non-existent USB device is specified on the command line,
print "qemu: could not add USB device 'X'".
Likewise for the usb_{add,del} monitor commands.

Signed-off-by: Scott Tsai <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

target-alpha: Initialize fpcr

Linux, at least, disables exceptions by default.

Signed-off-by: Richard Henderson <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

tcg-sparc: Implement brcond2.

Split out tcg_out_cmp and properly handle immediate arguments.
Fix constraints on brcond to match what SUBCC accepts.
Add tcg_out_brcond2_i32 for 32-bit host.

Signed-off-by: Richard Henderson <[email protected]>
Signed-off-by: Blue Swirl <[email protected]>

tcg-sparc: Use TCG_TARGET_REG_BITS in conditional compilation.

The test TCG_TARGET_REG_BITS==64 is exactly the feature that we
are checking for, whereas something involving __sparc_v9__ or
__sparc_v8plus__ should be reserved for something ISA related,
as with SMULX.

Signed-off-by: Richard Henderson <[email protected]>
Signed-off-by: Blue Swirl <[email protected]>

tcg-sparc: Improve tcg_out_movi for sparc64.

Generate sign-extended 32-bit constants with SETHI+XOR.
Otherwise tidy the routine to avoid the need for
conditional compilation and code duplication with movi_imm32.

Signed-off-by: Richard Henderson <[email protected]>
Signed-off-by: Blue Swirl <[email protected]>

tcg-sparc: Fix imm13 check in movi.

We were unnecessarily restricting imm13 constants to 12 bits.

Signed-off-by: Richard Henderson <[email protected]>
Signed-off-by: Blue Swirl <[email protected]>

ARM PBX-A9 memory map tweaks

Add core tile memeory to the RealView PBX-A9 board.  Document the
memeory maps that are known to work with the qemu bootloader.

Signed-off-by: Paul Brook <[email protected]>

LAN9118 improvements

Implement LAN9118 general purpose timer and PHY interrupts. Fix global
interrupt status bit.

Signed-off-by: Paul Brook <[email protected]>

PPC: Make DCR uint32_t

For what I know DCR is always 32 bits wide, so we should also use uint32_t to
pass it along the stacks.

This fixes a warning when compiling qemu-system-ppc64 with KVM enabled, making
it compile without --disable-werror

Signed-off-by: Alexander Graf <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

PPC64: Fix alternate timebase

Fix the alternate time base the same way as the default timebase. SPR_ATBL
should return a 64-bit value on 64 bit implementations.

Signed-off-by: Aurelien Jarno <[email protected]>

PPC64: Fix timebase

On PPC we have a 64-bit time base. Usually (PPC32) this is accessed using
two separate 32 bit SPR accesses to SPR_TBU and SPR_TBL.

On PPC64 the SPR_TBL register acts as 64 bit though, so we get the full
64 bits as return value. If we only take the lower ones, fine. But Linux
wants to see all 64 bits or it breaks.

This patch makes PPC64 Linux work even after TB crossed the 32-bit boundary,
which usually happened a few seconds after bootup.

Signed-off-by: Alexander Graf <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

target-alpha: Emit tcg debug_insn_start.

Signed-off-by: Richard Henderson <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

linux-user: Add aliases for some Alpha syscalls

Alpha always used 32-bit uids, but never renamed the syscalls
to match i386 when 32-bit uids were added there.  This enables
the proper bits in syscall.c.

Signed-off-by: Richard Henderson <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

Update OpenBIOS PPC image to revision 647

Signed-off-by: Aurelien Jarno <[email protected]>

ppc64: Change default machine to mac99

Also change the New World default CPU to 970FX for ppc64,
since the G4 is a 32-bit CPU.

Signed-off-by: Andreas Färber <[email protected]>
Acked-by: Alexander Graf <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

cpu-all.h: fix cpu_get_real_ticks() #ifdef

Reported-by: Hervé Poussineau <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

Multiboot support: Fix rom_copy

ROMs need to be loaded if they are anywhere in the requested area, not
only at the very beginning. This fixes Multiboot with ELF kernels that
have more than one program header.

Signed-off-by: Kevin Wolf <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

linux-user: add core dump support for SH

Signed-off-by: Nathan Froyd <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

linux-user: add core dump support for M68K

Signed-off-by: Nathan Froyd <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

linux-user: add core dump support for MIPS

Signed-off-by: Nathan Froyd <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

linux-user: add core dump support for PPC

Signed-off-by: Nathan Froyd <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

linux-user: fix ARM core dumps on opposite-endian hosts

Signed-off-by: Nathan Froyd <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

linux-user: commonify definitions of target typedefs

There's no sense in separately declaring target_{elf_greg,uid,gid,pid}_t
for every architecture.  Just declare them once with appropriate
USE_UID16 handling.

Signed-off-by: Nathan Froyd <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

linux-user: fix ELF_USE_CORE_DUMP/USE_ELF_CORE_DUMP confusion

Signed-off-by: Nathan Froyd <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

linux-user: use TARGET_ABI_FMT_lx to print abi_ulong types

Signed-off-by: Aurelien Jarno <[email protected]>

roms: allow roms to be loaded at address 0

It was possible to load roms at address 0, but commit
632cf034b401cdd01dae253a8b577fe518e37654 started to forbid that, which
broke at least ARM versatile.

Signed-off-by: Aurelien Jarno <[email protected]>

linux-user: cleanup force_sig() calls

Force_sig should be always called with TARGET_ signals.
Not that it really matters with SEGV, so this patch is
just for cleanup and improving consistency.

Signed-off-by: Riku Voipio <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

linux-user: Fix mmap_lock ordering

mmap_lock() can be called while tb_lock() is being held. To
avoid deadlock when one thread is holding mmap_lock and another
tb_lock, _always_ lock first tb_lock().

Signed-off-by: Riku Voipio <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

linux-user: enable tb unlinking when compiled with NPTL

Fixes receiving signals when guest code is being executed in a tight
loop. For an example, try interrupting the following code with ctrl-c.

http://nchipin.kos.to/test-loop.c

The tight loop is ofcourse brainless, but it is also exactly how the waitpid* testcases
are implemented.

Signed-off-by: Riku Voipio <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

user: fix compilation without -fpie

Fixes breakage by 299060a06e7c73de5cb9962edbca7f7f9e0b7f25.

Signed-off-by: Blue Swirl <[email protected]>

user_only: compile everything with -fpie

We really need compile _all_ sources for user target with -fpie when
use --enable-user-pie.

It's regression introduced by commit add16157d72454.

Signed-off-by: Kirill A. Shutemov <[email protected]>
Cc: Blue Swirl <[email protected]>
Signed-off-by: Juan Quintela <[email protected]>
Signed-off-by: Blue Swirl <[email protected]>

hw/openpic.c: replace tabs by spaces

Signed-off-by: Aurelien Jarno <[email protected]>

target-ppc: Use interrupts for escc

When using the escc with Linux, we need interrupts. So instead of creating
a dummy device, let's just map them to the openpic we have anyways.

This makes Linux on PPC64 with console=ttyPZ0 work.

Obviously, this change needs to be reflected in openbios. Patch for that
follows this one. Please update the binary then.

Signed-off-by: Alexander Graf <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

target-ppc: Initialize OpenPIC properly

When lowering an IRQ line, we search for the line we're supposed to lower.

Usually we run into an optimization there that queues up interrupts. This
queue ends with -1. Unfortunately we didn't set the first item to -1.

This patch fixes this, making interrupts work on PPC64.

Signed-off-by: Alexander Graf <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

Update version and changelog to sync with stable branch

Signed-off-by: Anthony Liguori <[email protected]>

Update to SeaBIOS 0.5.0

The only change is updating the makefile but that way we're carrying an official
release.

Signed-off-by: Anthony Liguori <[email protected]>

Revert "monitor: Convert do_migrate_set_speed() to QObject"

This reverts commit 3a4921047d514c336531c304bec362c4d138ebcc.

From Luiz:

  do_migrate_set_speed() accepts a suffix for the 'value' argument and this is
  not good for QMP.  We will have to add a new argument type to handle that and
  this will have to wait for 0.13.

Signed-off-by: Anthony Liguori <[email protected]>

e1000: Don't muck with PCI commmand register

Otherwise, the driver does not work in Linux after the INT_DISABLE changes in
PCI.

Michael Tsirkin had a patch to do this, I'm not sure what happened to it.

Signed-off-by: Anthony Liguori <[email protected]>

monitor: do_balloon(): Use 'M' argument type

This makes do_balloon() accept megabyte values from the user
Monitor while accepting byte values for QMP.

Signed-off-by: Luiz Capitulino <[email protected]>
Signed-off-by: Anthony Liguori <[email protected]>

monitor: Introduce 'M' argument type

This is a target long value in megabytes which should be
converted to bytes.

It will be used by handlers which accept a megabyte value
when in "user mode".

Signed-off-by: Luiz Capitulino <[email protected]>
Signed-off-by: Anthony Liguori <[email protected]>

QMP: Update spec file

- Remove "draft" status
- Change default success response to be json-object
- Change error and event data member to be a json-object
- Update examples
- Add new section "Compatibility Considerations"
- Other fixes and clarifications

Signed-off-by: Luiz Capitulino <[email protected]>
Signed-off-by: Anthony Liguori <[email protected]>

QMP: Update README file

- Fix output description
- Fix command-line usage notes
- Minor improvements

Signed-off-by: Luiz Capitulino <[email protected]>
Signed-off-by: Anthony Liguori <[email protected]>

QMP: Assure that returned data is a QDict

This is for debug purposes only.

Signed-off-by: Luiz Capitulino <[email protected]>
Signed-off-by: Anthony Liguori <[email protected]>

QMP: Return an empty dict by default

Currently, when a regular command doesn't have any data to output,
QMP will emit:

{ "return": "OK" }

Returning an empty dict is better though, because dicts can support
some protocol changes in a compatible way.

So, with this commit we will return:

{ "return": {} }

Signed-off-by: Luiz Capitulino <[email protected]>
Signed-off-by: Anthony Liguori <[email protected]>

QMP: Only handle converted commands

Looks like I dropped this check when addressing the 'query-'
commands request.

QMP should only handle converted commands, obviously.

Reported-by: Markus Armbruster <[email protected]>
Signed-off-by: Luiz Capitulino <[email protected]>
Signed-off-by: Anthony Liguori <[email protected]>

target-ppc: fix ppc32 kvm build

My segment sync patch broke compilation on PPC32, because it was trying to
sync the SLB even though ppc32 CPUs don't have an SLB.

So let's only sync it when we're on a PP64 one!

Signed-off-by: Alexander Graf <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

usb-musb: convert fifo to 8bit and add more registers

Convert musb fifo to 8bit to allow 8/16/32bit access

MUSB allows reading and writing to the fifo in 32/16/8 bit
width. The Linux kernel does this sometimes, most usually at
the end of writing the packet to allow packet to end at a
odd bytecount.

Convert the fifo to 8bit allows removing lots of shifts
which shows that the fifo is more natural as 8bit.

While at it, add multiple missing register definitions and
and cleanup debug prints.

Signed-off-by: Riku Voipio <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

handle SD CMD5 without error messages

From: Juha Riihimäki <[email protected]>
Signed-off-by: Juha Riihimäki <[email protected]>
Signed-off-by: Riku Voipio <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

Make USB hid devices self-powered

Simplifies power budget negotiation.

Signed-off-by: Riku Voipio <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

Give a error when running out of iomem areas.

The limit of iomem areas is quite low. Without the
debug print, it is quite hard to figure out why more
devices are not getting registered.

Signed-off-by: Riku Voipio <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

fix networking on win32 host

At least under some mingw compilers slirp networking fails without declaring
these fields packed.

From: Juha Riihimäki <[email protected]>
Signed-off-by: Juha Riihimäki <[email protected]>
Signed-off-by: Riku Voipio <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

fix pidfile option to work in WIN32

Explicit read/write locking pidfile under WIN32 is bit extreme
nobody get the chance to read the pidfile. Convert to a write-only lock.

Also, creating pidfile was disabled along with daemonize under
WIN32. Enable it, but do not enable daemon support which doesn't
exist under WIN32 atm.

From: Juha Riihimäki <[email protected]>
Signed-off-by: Juha Riihimäki <[email protected]>
Signed-off-by: Riku Voipio <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

Fix win32 log file location

/tmp doesn't exist under win32. Ease the pain of win32 development slightly.

From: Juha Riihimäki <[email protected]>
Signed-off-by: Riku Voipio <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

Update SeaBIOS to include PCI based option rom loading

Also remove pcbios from the tree.  It will no longer work.

Signed-off-by: Anthony Liguori <[email protected]>

roms: remove option rom packing logic

Now that we load the option roms via fw_cfg, we can stop copying
them to the 0xc000 -> 0xe000.  The patch does just that.

Also the rom loader gets simplified as all remaining users of the
rom loader load the bits at a fixed address so the packing and
aligning logic can go away.

Signed-off-by: Gerd Hoffmann <[email protected]>
Signed-off-by: Anthony Liguori <[email protected]>

roms: use new fw_cfg file xfer support.

roms: use fw_cfg for vgabios and option rom loading, additionally to
deploying them the traditional way (copy to 0xc0000 -> 0xe0000 range).

Signed-off-by: Gerd Hoffmann <[email protected]>
Signed-off-by: Anthony Liguori <[email protected]>

fw_cfg: add API for file transfer.

This patch adds a file transfer interface to fw_cfg.  Intended to be
used for passing non-pci option roms and vgabios to seabios.  Namespace
is modeled after the existing cbfs filesystem support in seabios.

Reading the new FW_CFG_FILE_DIR entry returns a file list.
Fields there are in network byte order (aka bigendian).

aliguori: fix fw_cfg.h for multiboot.bin, add proper fw_cfg.h declarations,
          quiet fprintf() in fw_cfg.c

Signed-off-by: Gerd Hoffmann <[email protected]>
Signed-off-by: Anthony Liguori <[email protected]>

fw_cfg: make calls typesafe

Signed-off-by: Gerd Hoffmann <[email protected]>
Signed-off-by: Anthony Liguori <[email protected]>

pci romfiles: add property, add default to PCIDeviceInfo

This patch adds a romfile property to the pci bus.  It allows to specify
a romfile to load into the rom bar of the pci device.  The default value
comes from a new field in PCIDeviceInfo.  The property allows to change
the file and also to disable the rom loading using an empty string.

Signed-off-by: Gerd Hoffmann <[email protected]>
Signed-off-by: Anthony Liguori <[email protected]>

Support PCI based option rom loading

Currently, we preload option roms into the option rom space in memory.  This
prevents DDIM from functioning correctly which severely limits the number
of roms we can support.

This patch introduces a pci_add_option_rom() which registers the
PCI_ROM_ADDRESS bar which points to our option rom.  It also converts over
the cirrus vga adapter, the rtl8139, virtio, and the e1000 to use this
new mechanism.

The result is that PXE boot functions even with three unique types of cards.

Signed-off-by: Anthony Liguori <[email protected]>
Signed-off-by: Gerd Hoffmann <[email protected]>
Signed-off-by: Anthony Liguori <[email protected]>

Fix backcompat for hotplug of SCSI controllers

SCSI controllers have no trouble existing without any attached
disks. This could be achieved with the (legacy) monitor syntax

  pci_add pci_addr=auto storage if=scsi

This is now denied with

  scsi requires a backing file/device.
  failed to add if=scsi

There is no need for this denial and it breaks compatability
with existing QEMU usage, so remove the check for presence
of a drive.

Signed-off-by: Daniel P. Berrange <[email protected]>
Signed-off-by: Anthony Liguori <[email protected]>

fdc: fix migration from 0.11

0.11 uses as instance ide io_base, get it back

Signed-off-by: Juan Quintela <[email protected]>
Signed-off-by: Anthony Liguori <[email protected]>

Revert "fdc: fix vmstate variable passed"

Floppy used the io_base address to register savevm region.

This reverts commit 2966b390d0f7cd5e5b971290fcc20685de374228.

Signed-off-by: Juan Quintela <[email protected]>
Signed-off-by: Anthony Liguori <[email protected]>

monitor: Accept input only byte-wise

This allows to suspend command interpretation and execution
synchronously, e.g. during migration.

Signed-off-by: Jan Kiszka <[email protected]>
Signed-off-by: Anthony Liguori <[email protected]>

Revert "kvm: x86: Save/restore exception_index"

This reverts commit ebbc8a3d8e76d0402f8a08c10c0f32e24715d41d.

As suggested by Jan Kiszka,

  "It was obsoleted by d1793b836f8f123b961c613de1bb1c0c185c84cc and now
   saves/restores a useless field."

Signed-off-by: Anthony Liguori <[email protected]>

vmware: increase cursor buffer size.

The cursor pixmap size we calculate later ends up being 4096 dwords
long by the looks of it. This boots an F12 LiveCD now.

Signed-off-by: Dave Airlie <[email protected]>
Signed-off-by: Anthony Liguori <[email protected]>

VMware VGA: Only enable dirty log tracking when fifo is disabled

This patch enables dirty log tracking whenever it's needed and disables it
when it is not.

We unconditionally enable dirty log tracking on reset, restart dirty log
tracking when PCI IO regions are remapped, and disable/enable it based on
commands from the guest.

Rebased-by: Dave Airlie <[email protected]>
Signed-off-by: Anthony Liguori <[email protected]>

Fix VMware VGA depth computation

VMware VGA requires that the depth presented to the guest is the same as the
DisplaySurface that it renders to.  This is because it performs a very simple
memcpy() to blit from one surface to another.

We currently hardcode a 24-bit depth.  The surface allocator for SDL may, and
usually will, allocate a surface with a different depth causing screen
corruption.

This changes the code to allocate the DisplaySurface before initializing the
device which allows the depth of the DisplaySurface to be used instead of
hardcoding something.

Signed-off-by: Anthony Liguori <[email protected]>

Make sure to enable dirty log tracking for VMware VGA

This is needed for VMware VGA to work properly under KVM.

Signed-off-by: Anthony Liguori <[email protected]>

Make sure to enable dirty tracking of VBE vram mapping

Apparently, VBE maps the VGA vram to a fixed physical location.  KVM requires
that all mappings of the VGA vram have dirty tracking enabled on them.  Any
access to the VGA vram through the VBE mapping currently fails to result in
dirty page tracking updates causing a black screen.

This is the true root cause of VMware VGA not working correctly under KVM and
likely also an issue with some of the std-vga black screen issues too.

Cirrus does not enable VBE so it would not be a problem when using Cirrus.

Signed-off-by: Anthony Liguori <[email protected]>
Rebased-by: Dave Airlie <[email protected]>

vmware: setup PCI BAR 2 for FIFO as per vmware spec

Signed-off-by: Anthony Liguori <[email protected]>

qdev: improve property error reporting.

Add a error message in case we fail to parse a qdev property.

Also make qemu not abort() in case setting a global property can't be
set.  This used to be a clear programming error.  The introduction of
the -global switch changed that though, so better exit instead (after
printing the new error message).

Signed-off-by: Gerd Hoffmann <[email protected]>
Signed-off-by: Anthony Liguori <[email protected]>

fix vga names in default_list

Fix mismerge between 64465297 and 556cd098.

Cc: Markus Armbruster <[email protected]>
Signed-off-by: Gerd Hoffmann <[email protected]>
Signed-off-by: Anthony Liguori <[email protected]>

usb-host: check mon before using it.

Signed-off-by: Gerd Hoffmann <[email protected]>
Signed-off-by: Anthony Liguori <[email protected]>

usb-net: use qdev for -usbdevice

Rebased to master, adapted to device renaming by armbru,
no other changes.

Signed-off-by: Gerd Hoffmann <[email protected]>
Signed-off-by: Anthony Liguori <[email protected]>

Check rom_load_all() return value.

Check rom_load_all() return value.
Also don't make option rom loading failure fatal.

Signed-off-by: Gerd Hoffmann <[email protected]>
Signed-off-by: Anthony Liguori <[email protected]>

defaults: update device_list[]

Add isa-fdc (disables default_floppy).
Add ide-drive (disables default_cdrom).

Also walk the -global QemuOpts, so we'll catch
-global isa-fdc.drive{A,B}=<name> too.

Signed-off-by: Gerd Hoffmann <[email protected]>
Signed-off-by: Anthony Liguori <[email protected]>

defaults: split default_drive

Split default_drive into default_{floppy,cdrom,sdcard}.
Also add QEMUMachine flags to disable them per machine.

Signed-off-by: Gerd Hoffmann <[email protected]>
Signed-off-by: Anthony Liguori <[email protected]>

monitor: Catch printing to non-existent monitor

The monitor_vprintf() function now touches the 'mon' pointer
before calling monitor_puts(), this causes block migration
to segfault as its functions call monitor_printf() with a
NULL 'mon'.

To fix the problem this commit moves the 'mon' NULL check
from monitor_puts() to monitor_vprintf().

This can potentially hide bugs, but for some reason this has
been the behavior for a long time.

Signed-off-by: Luiz Capitulino <[email protected]>
Signed-off-by: Anthony Liguori <[email protected]>

monitor: Avoid readline functions in QMP

The monitor_read_command() function is readline specific
and should only be used when readline is available.

Signed-off-by: Luiz Capitulino <[email protected]>
Signed-off-by: Anthony Liguori <[email protected]>

monitor: do_balloon(): Check for errors

do_balloon() should check for ballooning availability as
do_info_balloon() does.

Noted by Daniel P. Berrange <[email protected]>.

Signed-off-by: Luiz Capitulino <[email protected]>
Signed-off-by: Anthony Liguori <[email protected]>