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>

vmstate: remove ps2_kbd_load_old()

Now that we have version_id on post_load() we don't need the old load
function

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

vmstate: remove const from pre_save() functions

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

vmstate: add version_id argument to post_load

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

vmstate: remove const for put operations

In a later patch, we introduce pre_save() and post_save() functions.
The whole point of that operation is to change things in the state.
Without this patch, we have to remove the const qualifier in each
use with a cast

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

vga: move back dirty_log functions to vga.c

They are needed there on qemu-kvm.

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

cirrus_vga: also assign gr0/1 when writting shadow_gr0/1

This is a regression/bug caused by previous vga_cleanup.  This fixes
Ubuntu installer reported by:
 Pierre Riteau

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

Revert "eepro100: Remove unused device status entries"

This reverts commit 3031efabd0bb744126a53f32c5426580b5d394d5.

target-i386: Fix exceptions for fxsave/fxrstor

This patch corrects the following aspects of exception generation in
fxsave/fxrstor:

* Generate #GP if the operand is not aligned to a 16 byte boundary
* Generate #UD if the LOCK prefix is used
* For CR0.EM = 1 #NM is generated, not #UD

Signed-off-by: Kevin Wolf <mail@kevin-wolf.de>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

tcg: allocate s->op_dead_iargs dynamically

Similarly to what is already done in tcg_liveness_analysis() when
USE_LIVENESS_ANALYSIS is not set.

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

tcg: remove dead code

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

bsd-user: fix "#if 0"'d printf()

Make an "#if 0"'d printf() in load_elf_binary() reflect what the actual
code does (see commit 3bc0bdcaadef1100ce2413af818d9c8e2f6319fc).

Signed-off-by: Paul Bolle <pebolle@tiscali.nl>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

eepro100: Add more i825xx devices

The new devices added here are still not functional -
partially because some patches are still missing,
partially because I cannot test them. Nevertheless
they belong to the same family and will be supported
by this driver some day.

As soon as they work, they will also be added to hw/pci.c.

Signed-off-by: Stefan Weil <weil@mail.berlios.de>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

eepro100: Remove unused device status entries

A lot of entries are unused (they were added by copy + paste
from other drivers during development of eepro100.c).

Removing them from nic_save, nic_load makes any
old saved status incompatible, so a new version
for the virtual machine data was needed, too.

Signed-off-by: Stefan Weil <weil@mail.berlios.de>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

target-i386: add RDTSCP support

RDTSCP reads the time stamp counter and atomically also the content
of a 32-bit MSR, which can be freely set by the OS. This allows CPU
local data to be queried by userspace.
Linux uses this to allow a fast implementation of the getcpu()
syscall, which uses the vsyscall page to avoid a context switch.
AMD CPUs since K8RevF and Intel CPUs since Nehalem support this
instruction.
RDTSCP is guarded by the RDTSCP CPUID bit (Fn8000_0001:EDX[27]).

Signed-off-by: Andre Przywara <andre.przywara@amd.com>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

target-i386: add SSE4a instruction support

This adds support for the AMD Phenom/Barcelona's SSE4a instructions.
Those include insertq and extrq, which are doing shift and mask on
XMM registers, in two versions (immediate shift/length values and
stored in another XMM register).
Additionally it implements movntss, movntsd, which are scalar
non-temporal stores (avoiding cache trashing). These are implemented
as normal stores, though.
SSE4a is guarded by the SSE4A CPUID bit (Fn8000_0001:ECX[6]).

Signed-off-by: Andre Przywara <andre.przywara@amd.com>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

target-i386: add lock mov cr0 = cr8

AMD CPUs featuring a shortcut to access CR8 even from 32-bit mode.
If you use the LOCK prefix with "mov CR0", it accesses CR8 instead.
This behavior is guarded by the CR8_LEGACY CPUID bit
(Fn8000_0001:ECX[1]).

Signed-off-by: Andre Przywara <andre.przywara@amd.com>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

tcg/i386: add support for ext{8,16}u_i32 TCG ops

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

tcg/x86_64: add support for ext{8,16,32}u_i{32,64} TCG ops

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

tcg: add ext{8,16,32}u_i{32,64} TCG ops

Currently zero extensions ops are implemented by a and op with a
constant. This is then catched in some backend, and replaced by
a zero extension instruction. While this works well on RISC
machines, this adds a useless register move on non-RISC machines.

Example on x86:
  ext16u_i32 r1, r2
is translated into
  mov    %eax,%ebx
  movzwl %bx, %ebx
while the optimized version should be:
  movzwl %ax, %ebx

This patch adds ext{8,16,32}u_i{32,64} TCG ops that can be
implemented in the backends to avoid emitting useless register
moves.

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

linux-user: fix "#if 0"'d printf()

Make an "#if 0"'d printf() in load_elf_binary(), probably left to aid in
debugging, reflect what the actual code does. The current printf() will
only confuse those who "#if 1" it (it certainly confused me enough to
write this trivial patch).

Signed-off-by: Paul Bolle <pebolle@tiscali.nl>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

Check availability of uuid header / library

If available, the Universally Unique Identifier library
is used by the vdi block driver.

Other parts of QEMU (vl.c) could also use it.

Signed-off-by: Stefan Weil <weil@mail.berlios.de>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

hw/omap_dma: add matching {} in if 0

MULTI_REQ is never defined, so it doesn't matter much, but since
we have an if statement there, let's add {} to clarify what it
should do if it's uncommented, and indent the code properly.

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

oss/alsa: Do not invoke UB described in 7.15.1.1 (this time for ADC)

Signed-off-by: malc <av1474@comtv.ru>

libuser is a generated directory

Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>

Use GNU ld to link roms on OpenSolaris

OpenSolaris ld seems to miss the equivalent of GNU ld's "-Ttext 0".

Signed-off-by: Blue Swirl <blauwirbel@gmail.com>

Fix warning about undefined madvise() on OpenSolaris

OpenSolaris headers can't export madvise() with a sane set of #defines.
For background, see MySQL bug #7156 (http://bugs.mysql.com/bug.php?id=7156)
for discussion about Solaris header problems.

Signed-off-by: Blue Swirl <blauwirbel@gmail.com>

linux-user: fix up oversealous nitpicking

Looks like linux-user code was correct, just unreadable: what it wanted
to do with "-=" was really assign a negative number, not decrement.  Fix
up accordingly.

Reported-by: Laurent Desnogues <laurent.desnogues@gmail.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>

tests: missing ; in if 0

Fix missing ; in commented out code

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>

alsa: Change default buffer/period size

Increase buffer size but do not rely on ALSA picking up default period
size.

Signed-off-by: malc <av1474@comtv.ru>

oss/alsa: Do not invoke UB described in 7.15.1.1

Additional argument (whether to try poll mode) is only passed with
VOICE_ENABLE command.

Thanks to Markus Armbruster for noticing the potential breakage.

Revert "Get rid of _t suffix"

In the very least, a change like this requires discussion on the list.

The naming convention is goofy and it causes a massive merge problem.  Something
like this _must_ be presented on the list first so people can provide input
and cope with it.

This reverts commit 99a0949b720a0936da2052cb9a46db04ffc6db29.

Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>

Get rid of _t suffix

Some not so obvious bits, slirp and Xen were left alone for the time
being.

Signed-off-by: malc <av1474@comtv.ru>

Include microblaze binaries in tarbin.

Signed-off-by: Edgar E. Iglesias <edgar.iglesias@gmail.com>

target-mips: make sure constants are in the second argument

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

x86: use globals for CPU registers

Use globals for the 8 or 16 CPU registers on i386 and x86_64.

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

mips: Fix spelling in comment

inofficial -> unofficial

Thanks to Blue Swirl.

Signed-off-by: Stefan Weil <weil@mail.berlios.de>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>

escc: fix another coding style nit

Fix another place with =- to be "= -".
to avoid confusion with old-style "-="
(which we also have, and needs to be fixed).

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>

slirp: fix unmatched bracket in if 0

Fix unmatched bracket in commented out code

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>

net: fix coding style nit

Put space between = and - assigning a negative number
to avoid confusion with old-style "-="
(which we also have, and need to be fixed).

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>

linux-user: fix coding style nit

Put space between = and & when taking a pointer,
to avoid confusion with old-style "&=".

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>

target-mips: unmatched brackets in if 0

Fix unmatched braket in commented out code

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>

linux-user: fix old style decrement usage

Modern compilers do not parse "=-" as decrement:
you must use "-=" for that.

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>

twl92230: fix old style increment/decrement usage

Modern compilers do not parse "=-" as decrement:
you must use "-=" for that. Same for "=+"/"+=".

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>

omap_dma: fix unbalanced { in commented out code

Fix unbalanced {} in commented out code.

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>

escc: fix coding style nit

Put space between = and - assigning a negative number
to avoid confusion with old-style "-="
(which we also have, and needs to be fixed).

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>

gdbstub: fix coding style nit

Put space between = and * when dereferencing a pointer,
to avoid confusion with old-style "*="

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>

elfload: fix coding style nit

Put space between = and * when dereferencing a pointer,
to avoid confusion with old-style "*="

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>

vvfat: fix coding style nit

Put space between = and & when taking a pointer,
to avoid confusion with old-style "&=".

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>

sdlaudio: make it suck less

Signed-off-by: malc <av1474@comtv.ru>

Revert part of 6692b043198d58a12317009edb98654c6839f043

Committed by accident.

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

Fix build with profiler enabled

Broken by 4a1418e07bdcfaa3177739e04707ecaec75d89e1

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

TCG: fix DEF2 macro

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

target-ppc: log instructions start in TCG code

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

target-mips: log instructions start in TCG code

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

Win32: avoid a warning

GetLastError() returns a DWORD.

Signed-off-by: Blue Swirl <blauwirbel@gmail.com>

BSD user: suppress a warning

Signed-off-by: Blue Swirl <blauwirbel@gmail.com>

BSD user: implement GUEST_BASE

Based on 379f6698d73f476de38682b3ff96ecb226728c43.

Signed-off-by: Blue Swirl <blauwirbel@gmail.com>

tcg/i386: generates dec/inc instead of sub/add when possible

We must take care that dec/inc do not compute CF, which is needed by
add2/sub2.

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

tcg/i386: optimize and $0xff(ff), reg

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

Compile some user files only once for all targets

Signed-off-by: Blue Swirl <blauwirbel@gmail.com>

tcg/x86_64: generated dec/inc instead of sub/add when possible

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

Don't compile roms if not building system targets

Signed-off-by: Blue Swirl <blauwirbel@gmail.com>

Fix user emulator breakage

Fix breakage in the following conditions:
- use in-tree building
- build user targets after system targets

Signed-off-by: Blue Swirl <blauwirbel@gmail.com>

tcg/ppc: always use tcg_out_call

Signed-off-by: malc <av1474@comtv.ru>

vl: Add failure check for SetEvent

Signed-off-by: malc <av1474@comtv.ru>

vl: Do not use perror after failed Win32 API calls

Signed-off-by: malc <av1474@comtv.ru>

posix-aio-compat: avoid signal race when spawning a thread

Signed-off-by: malc <av1474@comtv.ru>

target-i386: kill a tmp register

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

target-i386: use subfi instead of sub with a non-freed constant

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

eepro100: Fix format strings in debug messages

size_t arguments need %zu instead of %d.

Signed-off-by: Stefan Weil <weil@mail.berlios.de>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>

eepro100: Replace sprintf by snprintf

Signed-off-by: Stefan Weil <weil@mail.berlios.de>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>