virtio-9p: Bugfix to send correct iounit

LCREATE function packs address of iounit in the pdu, fix that to send
actual iounit itself.

Signed-off-by: M. Mohan Kumar <[email protected]>
Acked-by: Aneesh Kumar K.V <[email protected]>
Signed-off-by: Venkateswararao Jujjuri <[email protected]>

virtio-9p: removexattr on default acl should return 0

If we don't have default acl, removexattr on default acl
should return 0

Signed-off-by: Aneesh Kumar K.V <[email protected]>
Signed-off-by: Venkateswararao Jujjuri <[email protected]>

virtio-9p: Print the pdu details on return

Signed-off-by: Aneesh Kumar K.V <[email protected]>
Signed-off-by: Venkateswararao Jujjuri <[email protected]>

virtio-9p: move 9p files around

Now that we start adding more files related to 9pfs
it make sense to move them to a separate directory

Signed-off-by: Aneesh Kumar K.V <[email protected]>
Signed-off-by: Venkateswararao Jujjuri <[email protected]>

pflash: Restore & fix lazy ROMD switching

Commit 5145b3d1cc revealed a bug in the lazy ROMD switch-back logic, but
resolved it by breaking that feature. This approach addresses the issue
by switching back to ROMD after a certain amount of read accesses
without further unlock sequences.

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

darwin-user: Remove unneeded null pointer check

cppcheck reports this error:

commpage.c:223: error: Possible null pointer dereference:
value - otherwise it is redundant to check if value is null at line 214

The null pointer check in line 214 is indeed not needed.
If value were null, the code would crash in line 223.
See do_compare_and_swap64 were for a reference.

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

Merge branch 'for-anthony' of git://repo.or.cz/qemu/kevin

* 'for-anthony' of git://repo.or.cz/qemu/kevin:
  Remove obsolete 'enabled' variable from progress state
  Add dd-style SIGUSR1 progress reporting
  qed: Fix consistency check on 32-bit hosts
  ide/atapi: Introduce CHECK_READY flag for commands
  ide/atapi: Replace bdrv_get_geometry calls by s->nb_sectors
  ide/atapi: Use table instead of switch for commands
  ide/atapi: Factor commands out
  ide: Split atapi.c out
  Improve accuracy of block migration bandwidth calculation
  atapi: Add 'medium ready' to 'medium not ready' transition on cd change
  qemu-img: allow rebase to a NULL backing file when unsafe

rtl8139: Fix compilation for w32/w64

Compilation for Windows needs a different declaration for the
printf format attribute, so use the macro which was defined for
this purpose.

Cc: Benjamin Poirier <[email protected]>
Signed-off-by: Stefan Weil <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

Remove obsolete 'enabled' variable from progress state

Signed-off-by: Jes Sorensen <[email protected]>
Signed-off-by: Kevin Wolf <[email protected]>

Add dd-style SIGUSR1 progress reporting

This introduces support for dd-style progress reporting on POSIX
systems, if the user hasn't specified -p to report progress. If sent a
SIGUSR1, qemu-img will report current progress for commands that
support progress reporting.

Signed-off-by: Jes Sorensen <[email protected]>
Signed-off-by: Kevin Wolf <[email protected]>

qed: Fix consistency check on 32-bit hosts

The qed_bytes_to_clusters() function is normally used with size_t
lengths.  Consistency check used it with file size length and therefore
failed on 32-bit hosts when the image file is 4 GB or more.

Make qed_bytes_to_clusters() explicitly 64-bit and update consistency
check to keep 64-bit cluster counts.

Reported-by: Michael Tokarev <[email protected]>
Signed-off-by: Stefan Hajnoczi <[email protected]>
Signed-off-by: Kevin Wolf <[email protected]>

ide/atapi: Introduce CHECK_READY flag for commands

Some commands are supposed to report a Not Ready Condition (i.e. they require
a medium to be present in order to execute successfully). Instead of
duplicating the check in each command implementation, let's add a flag and
check it before calling the command.

This patch only converts existing checks, it does not introduce new checks for
any of the other commands that can/should report a Not Ready Condition.

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

ide/atapi: Replace bdrv_get_geometry calls by s->nb_sectors

The disk size can only change when the medium is changed, and the change
callback takes care of updating s->nb_sectors in this case.

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

ide/atapi: Use table instead of switch for commands

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

ide/atapi: Factor commands out

In preparation for a table of function pointers, factor each command out from
ide_atapi_cmd() into its own function.

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

ide: Split atapi.c out

Besides moving code, this patch only fixes some whitespace issues in the moved
code and makes all functions in atapi.c static which can be static.

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

Improve accuracy of block migration bandwidth calculation

block_mig_state.total_time is currently the sum of the read request
latencies.  This is not very accurate because block migration uses aio and
so several requests can be submitted at once.  Bandwidth should be computed
with wall-clock time, not by adding the latencies.  In this case,
"total_time" has a higher value than it should, and so the computed
bandwidth is lower than it is in reality.  This means that migration can
take longer than it needs to.
However, we don't want to use pure wall-clock time here.  We are computing
bandwidth in the asynchronous phase, where the migration repeatedly wakes
up and sends some aio requests.  The computed bandwidth will be used for
synchronous transfer.

Signed-off-by: Avishay Traeger <[email protected]>
Reviewed-by: Michael Roth <[email protected]>
Signed-off-by: Kevin Wolf <[email protected]>

atapi: Add 'medium ready' to 'medium not ready' transition on cd change

MMC-5 Table F.1 lists errors that can be thrown for the TEST_UNIT_READY
command.  Going from medium not ready to medium ready states is
communicated by throwing an error.

This adds the missing 'tray opened' event that we fail to report to
guests.  After doing this, older Linux guests properly revalidate a disc
on the change command.  HSM violation errors, which caused Linux guests
to do a soft-reset of the link, also go away:

ata2.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6
sr 1:0:0:0: CDB: Test Unit Ready: 00 00 00 00 00 00
ata2.00: cmd a0/00:00:00:00:00/00:00:00:00:00/a0 tag 0
         res 01/60:00:00:00:00/00:00:00:00:00/a0 Emask 0x3 (HSM violation)
ata2.00: status: { ERR }
ata2: soft resetting link
ata2.00: configured for MWDMA2
ata2: EH complete

Signed-off-by: Amit Shah <[email protected]>
Acked-by: Jes Sorensen <[email protected]>
Tested-by: Markus Armbruster <[email protected]>
Signed-off-by: Kevin Wolf <[email protected]>

qemu-timer: Fix timers for w32

Commit 68c23e5520e8286d79d96ab47c0ea722ceb75041 removed the
multimedia timer, but this timer is needed for certain
Linux kernels. Otherwise Linux boot stops with this error:

    MP-BIOS bug: 8254 timer not connected to IO-APIC

So the multimedia timer is added again here.

Cc: Paolo Bonzini <[email protected]>
Signed-off-by: Stefan Weil <[email protected]>

qemu-timer: Avoid type casts

The type casts are no longer needed after some small changes
in struct qemu_alarm_timer. This also improves readability
of the code.

Signed-off-by: Stefan Weil <[email protected]>

qemu-timer: Remove unneeded include statement (w32)

mmsystem.h is not needed in qemu-timer.h, so remove it.

Signed-off-by: Stefan Weil <[email protected]>

qemu-timer: Add and use new function qemu_timer_expired_ns

This simply moves code which is used three times
into a new function thus improving readability.

Signed-off-by: Stefan Weil <[email protected]>

qemu-img: allow rebase to a NULL backing file when unsafe

QEMU can drop a backing file so that an image file no longer depends on
the backing file, but this feature has not been exposed in qemu-img.
This is useful in an image streaming usecase or when an image file has
been fully allocated and no reads can hit the backing file anymore.

Since the dropping the backing file can make the image unusable, only
allow this when the unsafe flag has been set.

Signed-off-by: Anthony Liguori <[email protected]>
Signed-off-by: Stefan Hajnoczi <[email protected]>
Signed-off-by: Kevin Wolf <[email protected]>

configure: reenable opengl by default

Because the opengl library is only linked to for the lm32 target, we can
now safely enable opengl by default again.

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

configure: support target dependent linking

This patch is the first attempt to make configure more intelligent with
regard to how it links to libraries. It divides the softmmu libraries into
two lists, a general one and a list which depends on the target
architecture.

Signed-off-by: Michael Walle <[email protected]>
Reviewed-by: Aurelien Jarno <[email protected]>
Acked-by: Edgar E. Iglesias <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

Merge branch 'linux-user-for-upstream' of git://gitorious.org/qemu-maemo/qemu

* 'linux-user-for-upstream' of git://gitorious.org/qemu-maemo/qemu:
  linux-user: untie syscalls from UID16
  linux-user: add s390x to llseek list
  linux-user: add ioctl(SIOCGIWNAME, ...) support.
  linux-user: convert ioctl(SIOCGIFCONF, ...) result.
  linux-user: improve traces
  [v2] linux-user: bigger default stack

configure: Make epoll_create1 test work around SPARC glibc bug

Work around a SPARC glibc bug which caused the epoll_create1 configure
test to wrongly claim that the function was present. Some versions of
SPARC glibc provided the function in the library but didn't declare
it in the include file; the result is that gcc warns about an implicit
declaration but a link succeeds. So we reference the function as a
value rather than a function call to induce a compile time error
if the declaration was not present.

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

Merge remote-tracking branch 'stefanha/tracing' into staging

vl: trivial spelling fix

Signed-off-by: Brad Hards <[email protected]>
Signed-off-by: Stefan Hajnoczi <[email protected]>

trace: [trace-events] fix print formats in some events

Signed-off-by: Lluís Vilanova <[email protected]>
Signed-off-by: Stefan Hajnoczi <[email protected]>

trace: [ust] fix generation of 'trace.c' on events without args

Signed-off-by: Lluís Vilanova <[email protected]>
Signed-off-by: Stefan Hajnoczi <[email protected]>

docs/tracing.txt: minor documentation fixes

Signed-off-by: Lluís Vilanova <[email protected]>
Signed-off-by: Stefan Hajnoczi <[email protected]>

docs: Trace events must not expect pointer dereferencing

Signed-off-by: Stefan Hajnoczi <[email protected]>

trace: Remove %s in grlib trace events

Trace events cannot use %s in their format strings because trace
backends vary in how they can deference pointers (if at all).  Recording
const char * values is not meaningful if their contents are not recorded
too.

Change grlib trace events that rely on strings so that they communicate
similar information without using strings.

A follow-up patch explains this limitation and updates docs/tracing.txt.

Signed-off-by: Stefan Hajnoczi <[email protected]>

tracetool: allow ) in trace output string

Be greedy in matching the trailing "\)*" pattern.  Otherwise, all the
text in the trace string up to the last closed parenthesis is taken as
part of the prototype.

Signed-off-by: Paolo Bonzini <[email protected]>
Signed-off-by: Stefan Hajnoczi <[email protected]>

linux-user: untie syscalls from UID16

Quite a number of uid/gid related syscalls are only defined on systems
with USE_UID16 defined. This is apperently based on the idea that these
system calls would never be called on non-UID16 systems. Make these
syscalls available for all architectures that define them.

drop alpha hack to support selected UID16 syscalls. MIPS and PowerPC
were also defined as UID16, to get uid/gid syscalls available, drop
this error as well.

Change QEMU to reflect this.

Cc: Ulrich Hecht <[email protected]>
Cc: Richard Henderson <[email protected]>
Cc: Alexander Graf <[email protected]>
Signed-off-by: Riku Voipio <[email protected]>

linux-user: add s390x to llseek list

We keep a list of host architectures that do llseek with the same
syscall as lseek. S390x is one of them, so let's add it to the list.

Original-patch-by: Ulrich Hecht <[email protected]>
Signed-off-by: Alexander Graf <[email protected]>
Signed-off-by: Riku Voipio <[email protected]>

linux-user: add ioctl(SIOCGIWNAME, ...) support.

Allow to run properly following program from linux-user:

/* cc -o wifi wifi.c */

 #include <stdio.h>
 #include <sys/ioctl.h>
 #include <sys/types.h>
 #include <sys/socket.h>
 #include <linux/wireless.h>
 #include <netinet/in.h>
 #include <arpa/inet.h>
 #include <string.h>

int main(int argc, char **argv)
{
    int ret;
    struct ifreq req;
    struct sockaddr_in *addr;
    int s;

    if (argc != 2) {
        fprintf(stderr, "Need an interface name (like wlan0)\n");
return 1;
    }

    s = socket( AF_INET, SOCK_DGRAM, 0 );
    if (s < 0) {
        perror("Cannot open socket");
        return 1;
    }
    strncpy(req.ifr_name, argv[1], sizeof(req.ifr_name));
    ret = ioctl( s, SIOCGIWNAME, &req );
    if (ret < 0) {
fprintf(stderr, "No wireless extension\n");
        return 1;
    }

    printf("%s\n", req.ifr_name);
    printf("%s\n", req.ifr_newname);
    return 0;
}

$ ./wifi eth0
No wireless extension

$ ./wifi wlan0
wlan0
IEEE 802.11bg

Signed-off-by: Laurent Vivier <[email protected]>
Signed-off-by: Riku Voipio <[email protected]>

linux-user: convert ioctl(SIOCGIFCONF, ...) result.

The result needs to be converted as it is stored in an array of struct
ifreq and sizeof(struct ifreq) differs according to target and host
alignment rules.

This patch allows to execute correctly the following program on arm
and m68k:

 #include <stdio.h>
 #include <sys/ioctl.h>
 #include <net/if.h>
 #include <alloca.h>
 #include <string.h>
 #include <sys/socket.h>
 #include <netinet/in.h>
 #include <arpa/inet.h>

int main(void)
{
    int s, ret;
    struct ifconf ifc;
    int i;

    memset( &ifc, 0, sizeof( struct ifconf ) );
    ifc.ifc_len = 8 * sizeof(struct ifreq);
    ifc.ifc_buf = alloca(ifc.ifc_len);

    s = socket( AF_INET, SOCK_DGRAM, 0 );
    if (s < 0) {
        perror("Cannot open socket");
        return 1;
    }
    ret = ioctl( s, SIOCGIFCONF, &ifc );
    if (s < 0) {
        perror("ioctl() failed");
        return 1;
    }

    for (i = 0; i < ifc.ifc_len / sizeof(struct ifreq) ; i ++) {
        struct sockaddr_in *s;
        s = (struct sockaddr_in*)&ifc.ifc_req[i].ifr_addr;
        printf("%s\n", ifc.ifc_req[i].ifr_name);
        printf("%s\n", inet_ntoa(s->sin_addr));
    }
}

Signed-off-by: Laurent Vivier <[email protected]>
Signed-off-by: Riku Voipio <[email protected]>

linux-user: improve traces

Add trace details for getpid(), kill(), _llseek(), rt_sigaction(),
rt_sigprocmask(), clone().

Signed-off-by: Laurent Vivier <[email protected]>
Signed-off-by: Riku Voipio <[email protected]>

[v2] linux-user: bigger default stack

PTHREAD_STACK_MIN (16KB) is somewhat inadequate for a new stack for new
QEMU threads. Set new limit to 256K which should be enough, yet doesn't
increase memory pressure significantly.

Signed-off-by: Riku Voipio <[email protected]>
Reviewed-by: Nathan Froyd <[email protected]>

doc: fix slirp description

net/slirp.c says:
    /* default settings according to historic slirp */
    struct in_addr net  = { .s_addr = htonl(0x0a000200) }; /* 10.0.2.0 */
    struct in_addr mask = { .s_addr = htonl(0xffffff00) }; /* 255.255.255.0 */
    struct in_addr host = { .s_addr = htonl(0x0a000202) }; /* 10.0.2.2 */
    struct in_addr dhcp = { .s_addr = htonl(0x0a00020f) }; /* 10.0.2.15 */
    struct in_addr dns  = { .s_addr = htonl(0x0a000203) }; /* 10.0.2.3 */

Which I think is not what the documentation says.

Signed-off-by: Brad Hards <[email protected]>
Reviewed-by: Stefan Hajnoczi <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

rtl8139: add format attribute to DPRINTF

gcc can check the format string for correctness even when debugging output is
not enabled.
Have to make sure arguments are always available. They are optimized out if
unneeded.

Signed-off-by: Benjamin Poirier <[email protected]>
Cc: Igor V. Kovalenko <[email protected]>
Reviewed-by: Stefan Hajnoczi <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

rtl8139: use variadic macro for debug statements

Removes double (( )) to make DEBUG_PRINT compatible with real function calls.
Change the name to DPRINTF to be consistent with other DPRINTF macros
throughout qemu.
Include the "RTL8139: " prefix in the macro. This changes some debug output
slightly since the prefix wasn't present on all lines.

Part of the change was done using the "coccinelle" tool with the following
small semantic match:
    @@ expression E; @@

    - DEBUG_PRINT((E))
    + DPRINTF(E)

Signed-off-by: Benjamin Poirier <[email protected]>
Cc: Igor V. Kovalenko <[email protected]>
Reviewed-by: Stefan Hajnoczi <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

rtl8139: use TARGET_FMT_plx in debug messages

Prevents a compilation failure when DEBUG_RTL8139 is defined:

CC    libhw32/rtl8139.o
cc1: warnings being treated as errors
hw/rtl8139.c: In function ‘rtl8139_cplus_transmit_one’:
hw/rtl8139.c:1960: error: format ‘%8lx’ expects type ‘long unsigned int’, but argument 5 has type ‘target_phys_addr_t’
make[1]: *** [rtl8139.o] Error 1

Signed-off-by: Benjamin Poirier <[email protected]>
Cc: Igor V. Kovalenko <[email protected]>
Reviewed-by: Stefan Hajnoczi <[email protected]>
Signed-off-by: Aurelien Jarno <[email protected]>

linux-user/arm/nwfpe: rename REG_PC to ARM_REG_PC

The REG_PC constant used in the ARM nwfpe code is fine in the kernel
but when used in qemu can clash with a definition in the host system
include files (in particular on Ubuntu Lucid SPARC, including signal.h
will define a REG_PC). Rename the constant to avoid this issue.

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

target-arm: Handle UNDEF cases for Neon VLD/VST multiple-structures

Correctly UNDEF for Neon VLD/VST "multiple structures" forms where the
align field is not valid.

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

target-arm: Handle UNDEFs for Neon single element load/stores

Handle the UNDEF and UNPREDICTABLE cases for Neon "single element to
one lane" VLD and "single element from one lane" VST.

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

Merge remote-tracking branch 'awilliam/ipxe' into staging

target-i386: switch to softfloat

This increase the correctness (precision, NaN values, corner cases) on
non-x86 machines, and add the possibility to handle the exception
correctly.

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

target-i386: fix constants wrt softfloat

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

target-i386: fix helper_fprem() and helper_fprem1() wrt softfloat

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

target-i386: fix logarithmic and trigonometric helpers wrt softfloat

Use the new CPU86_LDouble <-> double conversion functions to make logarithmic
and trigonometric helpers working with softfloat.

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

target-i386: add CPU86_LDouble <-> double conversion functions

Add functions to convert CPU86_LDouble to double and vice versa. They
are going to be used to implement logarithmic and trigonometric function
until softfloat implement them.

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

target-i386: replace approx_rsqrt and approx_rcp by softfloat ops

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

target-i386: fix helper_fsqrt() wrt softfloat

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

target-i386: fix helper_fdiv() wrt softfloat

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

target-i386: fix helper_fxtract() wrt softfloat

With softfloat it's not possible to play with the overflow of an
unsigned value to get the 0 case partially correct. Use a special case
for that. Using a division to generate an infinity is the easiest way
that works for both softfloat and softfloat-native.

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

target-i386: fix helper_fbld_ST0() wrt softfloat

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

target-i386: fix helper_fscale() wrt softfloat

Use the scalbn softfloat function to implement helper_fscale(). This
fixes corner cases (e.g. NaN) and makes a few more GNU libc math tests
to pass.

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

softfloat-native: add float*_is_any_nan() functions

Add float*_is_any_nan() functions to match the softfloat API.

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

softfloat-native: fix float*_scalbn() functions

float*_scalbn() should be able to take a status parameter. Fix that.

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

softfloat: fix float*_scalnb() corner cases

float*_scalnb() were not taking into account all cases. This patch fixes
some corner cases:
- NaN values in input were not properly propagated and the invalid flag
  not correctly raised. Use propagateFloat*NaN() for that.
- NaN or infinite values in input of floatx80_scalnb() were not correctly
  detected due to a typo.
- The sum of exponent and n could overflow, leading to strange results.
  Additionally having int16 defined to int make that happening for a very
  small range of values. Fix that by saturating n to the maximum exponent
  range, and using an explicit wider type if needed.

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

softfloat: add floatx80_compare*() functions

Add floatx80_compare() and floatx80_compare_quiet() functions to match
the softfloat-native ones.

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

softfloat-native: add a few constant values

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

softfloat: add pi constants

Add a pi constant for float32, float64, floatx80. It will be used by
target-i386 and later by the trigonometric functions.

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

softfloat: add floatx80 constants

Add floatx80 constants similarly to float32 or float64.

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

softfloat: fix floatx80_is_infinity()

With floatx80, the explicit bit is set for infinity.

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

softfloat: fix floatx80 handling of NaN

The floatx80 format uses an explicit bit that should be taken into account
when converting to and from commonNaN format.

When converting to commonNaN, the explicit bit should be removed if it is
a 1, and a default NaN should be used if it is 0.

When converting from commonNan, the explicit bit should be added.

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

vmstate: port mac_dbdma

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

mac_dbdma: create DBDMAState instead of passing one array around

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

vmstate: port piix4

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

piix4: create PIIX4State

It only contains a PCIDevice by know, but it makes easy to use migration code

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

vmstate: port mac_nvram

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

mac_nvram: size is a size, no need to be a target dependent type

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

vmstate: port nand

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

nand: pin values are uint8_t

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

vmstate: port max111x

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

max111x: input field is only used as uint8_t

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

vmstate: port pxa2xx_lcd

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

pxa2xx_lcd: up field is used as a bool and migrated as an uint8_t

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

pxa2xx_lcd: name anonymous struct

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

vmstate: stellaris use unused for placeholder entries

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

vmstate: port stellaris gamepad

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

vmstate: port syborg_keyboard

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

vmstate: port syborg_serial

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

vmstate: port stellaris_adc

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

vmstate: port syborg_pointer

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

vmstate: port ppc4xx_pci

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

vmstate: port ppce500_pci

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

vmstate: port pxa2xx_pm

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

vmstate: port pxa2xx_mm

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

vmstate: port pxa2xx_cm

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

vmstate: port pxa2xx_i2s

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

vmstate: port stellaris gptm

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

vmstate: port cuda

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

vmstate: port heathrow_pic

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

vmstate: port pl022 ssp

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

vmstate: port stellaris sys

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

vmstate: port stellaris ssi bus

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