As pointed out by avi the vgabios update is guest-visible and thus has
migration implications.

One change is that the vga has a valid pci rom bar now.  We already have
a pci bus property to enable/disable the rom bar and we'll load the bios
via fw_cfg as fallback for the no-rom-bar case.  So we just have to add
compat properties to handle this case.

A second change is that the magic bochs lfb @ 0xe0000000 is gone.  When
live-migrating a guest from a older qemu version it might be using the
lfb though, so we have to keep it for the old machine types.  The patch
enables the bochs lfb in case we don't have the pci rom bar enabled
(i.e. we are in 0.13+older compat mode).

This patch depends on these patches which add (and use) the pc-0.13
machine type:
  http://patchwork.ozlabs.org/patch/70797/
  http://patchwork.ozlabs.org/patch/70798/Signed-off-by: NGerd Hoffmann <kraxel@redhat.com>
Cc: avi@redhat.com
Signed-off-by: NAnthony Liguori <aliguori@us.ibm.com>

While not explicitly stated in the spec, it was observed on real systems
that enabling loopback testing on the pcnet controller disables
reception of external frames. And some legacy software relies on it, so
provide this behavior.
Signed-off-by: NJan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: NAnthony Liguori <aliguori@us.ibm.com>

Acked-by: NAnthony Liguori <aliguori@us.ibm.com>
Signed-off-by: NAvi Kivity <avi@redhat.com>
Signed-off-by: NAnthony Liguori <aliguori@us.ibm.com>

The current ioport callbacks are not type-safe, in that they accept an "opaque"
pointer as an argument whose type must match the argument to the registration
function; this is not checked by the compiler.

This patch adds an alternative that is type-safe.  Instead of an opaque
argument, both registation and the callback use a new IOPort type.  The
callback then uses container_of() to access its main structures.

Currently the old and new methods exist side by side; once the old way is gone,
we can also save a bunch of memory since the new method requires one pointer
per ioport instead of 6.
Acked-by: NAnthony Liguori <aliguori@us.ibm.com>
Signed-off-by: NAvi Kivity <avi@redhat.com>
Signed-off-by: NAnthony Liguori <aliguori@us.ibm.com>

VM state change notifications are invoked from vm_start()/vm_stop().
Trace these state changes so we can reason about the state of the VM
from trace output.
Signed-off-by: NStefan Hajnoczi <stefanha@linux.vnet.ibm.com>
Signed-off-by: NAnthony Liguori <aliguori@us.ibm.com>

This patch enables MSI-X for virtfs-9p-pci.  It also adds a
compat property to pc-0.13 which turns it of there to stay
compatible to 0.13-stable.
Signed-off-by: NGerd Hoffmann <kraxel@redhat.com>
Signed-off-by: NAnthony Liguori <aliguori@us.ibm.com>

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

fprintf_function adds format checking with GCC_FMT_ATTR.

Cc: Blue Swirl <blauwirbel@gmail.com>
Signed-off-by: NStefan Weil <weil@mail.berlios.de>
Signed-off-by: NAnthony Liguori <aliguori@us.ibm.com>

Neither DECLARE_SPRINTF nor BAD_SPRINTF are needed for QEMU.

QEMU won't support systems with missing or bad declarations
for sprintf. The unused code was detected while looking for
functions with missing format checking. Instead of adding
GCC_FMT_ATTR, the unused code was removed.

Cc: Blue Swirl <blauwirbel@gmail.com>
Signed-off-by: NStefan Weil <weil@mail.berlios.de>
Signed-off-by: NAnthony Liguori <aliguori@us.ibm.com>

We have an OS which writes to port 0x400 when probing for special hardware.
This causes an exit of the VM. With SeaBIOS this port isn't used anyway.
Signed-off-by: NAlexander Graf <agraf@suse.de>
Reviewed-By: NPaolo Bonzini <pbonzini@redhat.com>
Signed-off-by: NBernhard Kohl <bernhard.kohl@nsn.com>
Signed-off-by: NAnthony Liguori <aliguori@us.ibm.com>

signrom.sh has multiple bugs:

- the last byte is considered when calculating the existing checksum, but not
  when computing the correction
- apprently the 'expr' expression overflows and produces incorrect results with
  larger roms
- if the checksum happened to be zero, we calculated the correction byte to be
  256

Instead of rewriting this in half a line of python, this patch fixes the bugs.
Signed-off-by: NAvi Kivity <avi@redhat.com>
Signed-off-by: NAnthony Liguori <aliguori@us.ibm.com>

Commit b152aa84 broke the unit-tests
build, fix it.
Signed-off-by: NLuiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: NAnthony Liguori <aliguori@us.ibm.com>

Block migration can submit multiple AIO reads for the same sector/chunk, but
completion of such reads can happen out of order:

migration               guest
- get_dirty(N)
- aio_read(N)
- clear_dirty(N)
                        write(N)
                        set_dirty(N)
- get_dirty(N)
- aio_read(N)

If the first aio_read completes after the second, stale data will be
migrated to the destination.

Fix by not allowing multiple AIOs inflight for the same sector.
Signed-off-by: NMarcelo Tosatti <mtosatti@redhat.com>
Signed-off-by: NAnthony Liguori <aliguori@us.ibm.com>

Sectors are marked dirty in the bitmap on AIO submission. This is wrong
since data has not reached storage.

Set a given sector as dirty in the dirty bitmap on AIO completion, so that
reading a sector marked as dirty is guaranteed to return uptodate data.
Reviewed-by: NKevin Wolf <kwolf@redhat.com>
Signed-off-by: NMarcelo Tosatti <mtosatti@redhat.com>
Signed-off-by: NAnthony Liguori <aliguori@us.ibm.com>

Otherwise upper 32 bits of bitmap entries are not correctly calculated.
Reviewed-by: NKevin Wolf <kwolf@redhat.com>
Signed-off-by: NMarcelo Tosatti <mtosatti@redhat.com>
Signed-off-by: NAnthony Liguori <aliguori@us.ibm.com>

This introduces generation of a qemu.stp/qemu-system-XXX.stp
files which provides tapsets with friendly names for static
probes & their arguments. Instead of

    probe process("qemu").mark("qemu_malloc") {
        printf("Malloc %d %p\n", $arg1, $arg2);
    }

It is now possible todo

    probe qemu.system.i386.qemu_malloc {
        printf("Malloc %d %p\n", size, ptr);
    }

There is one tapset defined per target arch, for both
user and system emulators.

* Makefile.target: Generate stp files for each target
* tracetool: Support for generating systemtap tapsets
* configure: Check for whether systemtap is available
  with the DTrace backend
Reviewed-by: NStefan Hajnoczi <stefanha@linux.vnet.ibm.com>
Signed-off-by: NDaniel P. Berrange <berrange@redhat.com>
Signed-off-by: NAnthony Liguori <aliguori@us.ibm.com>

This introduces a new tracing backend that targets the SystemTAP
implementation of DTrace userspace tracing. The core functionality
should be applicable and standard across any DTrace implementation
on Solaris, OS-X, *BSD, but the Makefile rules will likely need
some small additional changes to cope with OS specific build
requirements.

This backend builds a little differently from the other tracing
backends. Specifically there is no 'trace.c' file, because the
'dtrace' command line tool generates a '.o' file directly from
the dtrace probe definition file. The probe definition is usually
named with a '.d' extension but QEMU uses '.d' files for its
external makefile dependancy tracking, so this uses '.dtrace' as
the extension for the probe definition file.

The 'tracetool' program gains the ability to generate a trace.h
file for DTrace, and also to generate the trace.d file containing
the dtrace probe definition.

Example usage of a dtrace probe in systemtap looks like:

  probe process("qemu").mark("qemu_malloc") {
    printf("Malloc %d %p\n", $arg1, $arg2);
  }

* .gitignore: Ignore trace-dtrace.*
* Makefile: Extra rules for generating DTrace files
* Makefile.obj: Don't build trace.o for DTrace, use
  trace-dtrace.o generated by 'dtrace' instead
* tracetool: Support for generating DTrace data files
Reviewed-by: NStefan Hajnoczi <stefanha@linux.vnet.ibm.com>
Signed-off-by: NDaniel P. Berrange <berrange@redhat.com>
Signed-off-by: NAnthony Liguori <aliguori@us.ibm.com>

This reverts commit 4addb112.

This reverts commit 2834c3e0.

Conflicts:

	Makefile.target

Originally proposed by Gerd Hoffmann.
Signed-off-by: Nmalc <av1474@comtv.ru>
Acked-by: NGerd Hoffmann <kraxel@redhat.com>

Signed-off-by: NAlex Williamson <alex.williamson@redhat.com>
Signed-off-by: NAnthony Liguori <aliguori@us.ibm.com>

We can't let the compiler define the alignment for qemu_cfg data.
Signed-off-by: NAlex Williamson <alex.williamson@redhat.com>
Signed-off-by: NAnthony Liguori <aliguori@us.ibm.com>

Fix a makefile error that meant that qemu would not compile if
the source and object directories were the same.
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>
Signed-off-by: NAnthony Liguori <aliguori@us.ibm.com>

Since commit 4bed9837 an .fd_read()
handler that deletes its IOHandler is exposed to .fd_write() being
called on the deleted IOHandler.

This patch fixes deletion so that .fd_read() and .fd_write() are never
called on an IOHandler that is marked for deletion.
Signed-off-by: NStefan Hajnoczi <stefanha@linux.vnet.ibm.com>
Signed-off-by: NAnthony Liguori <aliguori@us.ibm.com>

Some devices seem to choke on receiving a USB_REQ_GET_CONFIGURATION ctrl msg
(witnessed with a digital picture frame usb id 1908:1320).
When usb_fs_type == USB_FS_SYS, the active configuration can be read directly
from sysfs, which allows using this device through qemu's usb redirection.
More in general it seems a good idea to not send needless control msg's to
devices, esp. as the code in question is called every time a set_interface
is done. Which happens multiple times during virtual machine startup, and
when device drivers are activating the usb device.
Signed-off-by: NAnthony Liguori <aliguori@us.ibm.com>

The next patch in this series introduces multiple ways to get the
configuration dependent upon usb_fs_type, it is cleaner to put this
into its own function.
Signed-off-by: NAnthony Liguori <aliguori@us.ibm.com>

This allows us to recreate the sysfspath used during scanning later
(which will be used in a later patch in this series).
Signed-off-by: NAnthony Liguori <aliguori@us.ibm.com>

This patch adds missing braces around if/else statements that call
macros which are likely to result in errors if the macro is
changed. It also makes the code comply better with CODING_STYLE.
Signed-off-by: NJes Sorensen <Jes.Sorensen@redhat.com>
Signed-off-by: NAnthony Liguori <aliguori@us.ibm.com>

Properly check array bounds before accessing array element.
Signed-off-by: NGleb Natapov <gleb@redhat.com>
Signed-off-by: NAnthony Liguori <aliguori@us.ibm.com>

This introduces generation of a qemu.stp/qemu-system-XXX.stp
files which provides tapsets with friendly names for static
probes & their arguments. Instead of

    probe process("qemu").mark("qemu_malloc") {
        printf("Malloc %d %p\n", $arg1, $arg2);
    }

It is now possible todo

    probe qemu.system.i386.qemu_malloc {
        printf("Malloc %d %p\n", size, ptr);
    }

There is one tapset defined per target arch.

* Makefile: Generate a qemu.stp file for systemtap
* tracetool: Support for generating systemtap tapsets
Signed-off-by: NDaniel P. Berrange <berrange@redhat.com>
Signed-off-by: NAnthony Liguori <aliguori@us.ibm.com>

This introduces a new tracing backend that targets the SystemTAP
implementation of DTrace userspace tracing. The core functionality
should be applicable and standard across any DTrace implementation
on Solaris, OS-X, *BSD, but the Makefile rules will likely need
some small additional changes to cope with OS specific build
requirements.

This backend builds a little differently from the other tracing
backends. Specifically there is no 'trace.c' file, because the
'dtrace' command line tool generates a '.o' file directly from
the dtrace probe definition file. The probe definition is usually
named with a '.d' extension but QEMU uses '.d' files for its
external makefile dependancy tracking, so this uses '.dtrace' as
the extension for the probe definition file.

The 'tracetool' program gains the ability to generate a trace.h
file for DTrace, and also to generate the trace.d file containing
the dtrace probe definition.

Example usage of a dtrace probe in systemtap looks like:

  probe process("qemu").mark("qemu_malloc") {
    printf("Malloc %d %p\n", $arg1, $arg2);
  }

* .gitignore: Ignore trace-dtrace.*
* Makefile: Extra rules for generating DTrace files
* Makefile.obj: Don't build trace.o for DTrace, use
  trace-dtrace.o generated by 'dtrace' instead
* tracetool: Support for generating DTrace data files
Signed-off-by: NDaniel P. Berrange <berrange@redhat.com>
Signed-off-by: NAnthony Liguori <aliguori@us.ibm.com>

A via -kernel supplied x86_64 ELF image is being started in 32bit mode.
Detect and exit if a 64bit image has been supplied.
Signed-off-by: NAdam Lackorzynski <adam@os.inf.tu-dresden.de>
Acked-by: NAlexander Graf <agraf@suse.de>
Signed-off-by: NAnthony Liguori <aliguori@us.ibm.com>

local_apics are allocated sequentially and never removed, so
we can stop any iterations that go to MAX_APICS as soon as we
hit the first NULL.  Looking at a small guest running a virtio-net
workload with oprofile, this drops apic_get_delivery_bitmask()
from #3 in the profile to down in the noise.
Signed-off-by: NAlex Williamson <alex.williamson@redhat.com>
Signed-off-by: NAnthony Liguori <aliguori@us.ibm.com>

This patch fixes hot unplug of cold plugged devices
(those present at system start), which got broken by
5beb8ad5 .
Signed-off-by: NIsaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>
Acked-by: NCam Macdonell <cam@cs.ualberta.ca>
Tested-by: NCam Macdonell <cam@cs.ualberta.ca>
Reported-by: Cam Macdonell <cam@cs.ualberta.ca>.

pcibus_dev_print() was erroneously retrieving the device bus
number from the secondary bus number offset of the device
instead of the bridge above the device.  This ends of landing
in the 2nd byte of the 3rd BAR for devices, which thankfully
is usually zero.

Note: pcibus_get_dev_path() copied this code,
inheriting the same bug.  pcibus_get_dev_path() is used for
ramblock naming, so changing it can effect migration.  However,
I've only seen this byte be non-zero for an assigned device,
which can't migrate anyway, so hopefully we won't run into
any issues.

This patch does not touch pcibus_get_dev_path, as
bus number is guest assigned for nested buses,
so using it for migration is broken anyway.
Fix it properly later.
Signed-off-by: NAlex Williamson <alex.williamson@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

When adding the length to the pseudo header, we're not properly
accounting for overflow.

From: Mark Wu <dwu@redhat.com>
Signed-off-by: NAlex Williamson <alex.williamson@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

virtio-net expects set_offload to succeed after
peer cleanup.
Since we don't have an open fd anymore, make it so.
Fixes warning about the failure of offload setting.
Reported-by: NJason Wang <jasowang@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

Add save/restore of MSR for migration and cpuid bit.
Signed-off-by: NGleb Natapov <gleb@redhat.com>
Signed-off-by: NMarcelo Tosatti <mtosatti@redhat.com>

Frontends calling tap_get_vhost_net get an invalid pointer after the
peer backend has been deleted. Jason Wang <jasowang@redhat.com> reports
this leading to a crash in ack_features when we remove the vhost-net
bakend of a virtio nic.

The fix is simply to clear the backend pointer.
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>