The next patch adds section footers; but we don't want to
break migration compatibility so disable them on older
machine types
Signed-off-by: NDr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: NJuan Quintela <quintela@redhat.com>
Signed-off-by: NJuan Quintela <quintela@redhat.com>

This PIIX4 init function has no more reason to receive a pointer to the
FwCfg object. Remove the parameter from the prototype, and update callers.

As a result, the pc_init1() function no longer needs to save the return
value of pc_memory_init() and xen_load_linux(), which makes it more
similar to pc_q35_init().

The return type & value of pc_memory_init() and xen_load_linux() are not
changed themselves; maybe we'll need their return values sometime later.

RHBZ: https://bugzilla.redhat.com/show_bug.cgi?id=1204696
Cc: Amit Shah <amit.shah@redhat.com>
Cc: "Michael S. Tsirkin" <mst@redhat.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Richard Henderson <rth@twiddle.net>
Cc: Eduardo Habkost <ehabkost@redhat.com>
Cc: Aurelien Jarno <aurelien@aurel32.net>
Cc: Leon Alrae <leon.alrae@imgtec.com>
Signed-off-by: NLaszlo Ersek <lersek@redhat.com>
Reviewed-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NLaszlo Ersek <lersek@redhat.com>

Since pc_allocate_cpu_irq only requests one irq, so let it just call
qemu_allocate_irq.
Signed-off-by: NShannon Zhao <zhaoshenglong@huawei.com>
Signed-off-by: NShannon Zhao <shannon.zhao@linaro.org>
Signed-off-by: NMichael Tokarev <mjt@tls.msk.ru>

valgrind complains about:
==16447== 8 bytes in 1 blocks are definitely lost in loss record 552 of 3,310
==16447==    at 0x4C2845D: malloc (in /usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so)
==16447==    by 0x2E4FD7: malloc_and_trace (vl.c:2546)
==16447==    by 0x64C770E: g_malloc (in /usr/lib64/libglib-2.0.so.0.3600.3)
==16447==    by 0x36FB47: qemu_extend_irqs (irq.c:55)
==16447==    by 0x36FBD3: qemu_allocate_irqs (irq.c:64)
==16447==    by 0x24E622: pc_init1 (pc_piix.c:287)
==16447==    by 0x24E76A: pc_init_pci (pc_piix.c:310)
==16447==    by 0x2E9360: main (vl.c:4226)

==16447== 128 bytes in 1 blocks are definitely lost in loss record 2,569 of 3,310
==16447==    at 0x4C2845D: malloc (in /usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so)
==16447==    by 0x2E4FD7: malloc_and_trace (vl.c:2546)
==16447==    by 0x64C770E: g_malloc (in /usr/lib64/libglib-2.0.so.0.3600.3)
==16447==    by 0x36FB47: qemu_extend_irqs (irq.c:55)
==16447==    by 0x36FBD3: qemu_allocate_irqs (irq.c:64)
==16447==    by 0x25BEB2: kvm_i8259_init (i8259.c:133)
==16447==    by 0x24E1F1: pc_init1 (pc_piix.c:219)
==16447==    by 0x24E76A: pc_init_pci (pc_piix.c:310)
==16447==    by 0x2E9360: main (vl.c:4226)
Signed-off-by: NShannon Zhao <zhaoshenglong@huawei.com>
Signed-off-by: NShannon Zhao <shannon.zhao@linaro.org>
Reviewed-by: NMarcel Apfelbaum <marcel@redhat.com>
Signed-off-by: NMichael Tokarev <mjt@tls.msk.ru>

This patch introduces no observable change, but it allows the callers of
pc_basic_device_init(), ie. pc_init1() and pc_q35_init(), to request (or
not request) the creation of the FDC explicitly.

At the moment both callers pass constant create_fdctrl=true (hence no
observable change).

Assuming a board passes create_fdctrl=false, "floppy" will be NULL on
output, and (beyond the FDC not being created) that NULL will be passed on
to pc_cmos_init(). Luckily, pc_cmos_init() already handles that case.

Cc: Markus Armbruster <armbru@redhat.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Gerd Hoffmann <kraxel@redhat.com>
Cc: John Snow <jsnow@redhat.com>
Cc: "Gabriel L. Somlo" <gsomlo@gmail.com>
Cc: "Michael S. Tsirkin" <mst@redhat.com>
Cc: Kevin Wolf <kwolf@redhat.com>
Cc: qemu-block@nongnu.org
Signed-off-by: NLaszlo Ersek <lersek@redhat.com>
Reviewed-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>
Reviewed-by: NMarkus Armbruster <armbru@redhat.com>

All pc-i440fx and pc-q35 init functions simply call the corresponding
compat function and then call the main init function. Use a macro to
generate that code.
Signed-off-by: NEduardo Habkost <ehabkost@redhat.com>
Reviewed-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

The function is not needed anymore, we can simply call pc_init1()
directly.
Signed-off-by: NEduardo Habkost <ehabkost@redhat.com>
Reviewed-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

This looks like a step backwards, but it will allow pc-0.1[0123] and
isapc to follow the same compat+init pattern used by the other
machine-types, allowing us to generate all init function using the same
macro later.
Signed-off-by: NEduardo Habkost <ehabkost@redhat.com>
Reviewed-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

Now that we have a DEFINE_PC_MACHINE helper macro that just requires an
initialization function, it is trivial to convert them to register a QOM
machine class directly, instead of using QEMUMachine.
Signed-off-by: NEduardo Habkost <ehabkost@redhat.com>
Reviewed-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

This will simplify the DEFINE_PC_MACHINE macro, and will help us to
implement reuse of PC_COMPAT_* macros through class_init function reuse,
in the future.
Signed-off-by: NEduardo Habkost <ehabkost@redhat.com>
Reviewed-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

By now the new functions will get QEMUMachine as argument, but they will
be later converted to initialize a MachineClass struct directly.
Signed-off-by: NEduardo Habkost <ehabkost@redhat.com>
Reviewed-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

This will automatically generate the existing QEMUMachine structs based
on the *_MACHINE_OPTIONS macros, and automatically add registration code
for them.
Signed-off-by: NEduardo Habkost <ehabkost@redhat.com>
Reviewed-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

Define a MACHINE_OPTIONS macro for each PC machine, and move every field
inside the QEMUMachine structs to the macros, except for name, init, and
compat_props.

This also ensures that all MACHINE_OPTIONS inherit the fields from the
next version, so their definitions carry only the changes that exist
between one version and the next one.

Comments about specific cases:

pc-*-2.1:

  Existing PC_*_2_1_MACHINE_OPTIONS macros were defined as:
      PC_*_MACHINE_OPTIONS,
      .default_machine_opts = "firmware=bios-256k.bin"

  PC_*_2_2_MACHINE_OPTIONS is:
      PC_*_2_3_MACHINE_OPTIONS
  which is expanded to:
      PC_*_MACHINE_OPTIONS,
      .default_machine_opts = "firmware=bios-256k.bin",
      .default_display = "std"

  The only difference between 2_1 and 2_2 is .default_display, that's why
  we didn't reuse PC_*_2_2_MACHINE_OPTIONS. The good news is that having
  multiple initializers for a field is allowed by C99, and the last
  initializer overrides the previous ones.

  So we can reuse the 2_2 macro in 2_1 and define PC_*_2_1_MACHINE_OPTIONS
  as:
      PC_*_2_2_MACHINE_OPTIONS,
      .default_display = NULL

pc-*-1.7:

  PC_*_1_7_MACHINE_OPTIONS was defined as:
      PC_*_MACHINE_OPTIONS

  PC_*_2_0_MACHINE_OPTIONS is defined as:
      PC_*_2_1_MACHINE_OPTIONS
  which is expanded to:
      PC_*_2_2_MACHINE_OPTIONS,
      .default_display = NULL
  which is expanded to:
      PC_*_2_3_MACHINE_OPTIONS,
      .default_display = NULL
  which is expanded to:
      PC_*_MACHINE_OPTIONS,
      .default_machine_opts = "firmware=bios-256k.bin",
      .default_display = "std",
      .default_display = NULL  /* overrides the previous line */

  So, the only difference between PC_*_1_7_MACHINE_OPTIONS and
  PC_*_2_0_MACHINE_OPTIONS is .default_machine_opts (as .default_display
  is not explicitly set by PC_*_MACHINE_OPTIONS so it is NULL).

  So we can keep the macro reuse pattern and define
  PC_*_2_0_MACHINE_OPTIONS as:
      PC_*_2_0_MACHINE_OPTIONS,
      .default_machine_opts = NULL

pc-*-2.4 (alias and is_default fields):

  Set alias and is_default fields inside the 2.4 MACHINE_OPTIONS macro,
  and clear it in the 2.3 macro (that reuses the 2.4 macro).

hw_machine:

  As all the machines older than v1.0 set hw_version explicitly, we can
  safely move the field to the MACHINE_OPTIONS macros without affecting
  the other versions that reuse them.

init function:

  Some machines had the init function set inside the MACHINE_OPTIONS
  macro. Move it to the QEMUMachine declaration, to keep it consistent
  with the other machines.
Signed-off-by: NEduardo Habkost <ehabkost@redhat.com>
Reviewed-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

Move compat_props from pc-0.10 to the macro, to make it consistent with
the other machines.
Signed-off-by: NEduardo Habkost <ehabkost@redhat.com>
Reviewed-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

The VGA and vmware-svga rombar compat properties were added by commit
281a26b1, but only to pc-0.13 and
pc-0.12. This breaks the PC_COMPAT_* nesting pattern we currently
follow.

The new variables will now be inherited by pc-0.11 and older, but
pc-0.11 and pc-0.10 already have PCI.rombar=0 on compat_props, so they
shouldn't be affected at all.

Cc: Stefan Weil <sw@weilnetz.de>
Cc: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: NEduardo Habkost <ehabkost@redhat.com>
Reviewed-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

The compat property was added by commit
9dbcca5a, and the pc-0.12 and older
machine-types were not changed because virtio-9p-pci was introduced on QEMU
0.13 (commit 9f107513). The only problem is
that this breaks the PC_COMPAT_* nesting pattern we currently use.

So, move the property to PC_COMPAT_0_13. This make pc-0.12 and older inherit
it, but that shouldn't be an issue as QEMU 0.12 didn't have virtio-9p-pci.

Cc: Gerd Hoffmann <kraxel@redhat.com>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: NEduardo Habkost <ehabkost@redhat.com>
Reviewed-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

The current code setting ide-drive.ver and scsi-disk.ver on pc-0.11
breaks the PC_COMPAT_* nesting pattern we currently use.

As those variables are overwritten in pc-0.10 too, they can be inherited
by pc-0.10 with no side-effects at all.

Cc: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: NEduardo Habkost <ehabkost@redhat.com>
Reviewed-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

Those properties were introduced by commit
3827cdb1. They were not duplicated into
pc-0.13 and older because 0.14 was the first QEMU version supporting
qxl. The only problem is that this breaks the PC_COMPAT_* nesting
pattern we currently use.

So, move the properties to PC_COMPAT_0_14. This makes pc-0.13 and older
inherit them, but that shouldn't be an issue as QEMU 0.13 didn't support
qxl.

Cc: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: NEduardo Habkost <ehabkost@redhat.com>
Reviewed-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

Once we start adding compat code for pc-2.3, the usage of HW_COMPAT_2_1
in pc-*-2.2 won't be enough, as it also has to include PC_COMPAT_2_3
inside it. To ensure that, define PC_COMPAT_2_3, PC_COMPAT_2_2, and
PC_COMPAT_2_1 macros.
Signed-off-by: NEduardo Habkost <ehabkost@redhat.com>
Reviewed-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

Changing the convention to include commas inside the macros will allow
macros containing empty lists to be defined and used without compilation
errors.
Signed-off-by: NEduardo Habkost <ehabkost@redhat.com>
Reviewed-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

Changing the convention to include commas inside the macros will allow
macros containing empty lists to be defined and used without compilation
errors.
Signed-off-by: NEduardo Habkost <ehabkost@redhat.com>
Reviewed-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

Coding style change only.
Signed-off-by: NEduardo Habkost <ehabkost@redhat.com>
Reviewed-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

The following patches will limit the following things to legacy
machine type:

- maximum number of virtqueues for virtio-pci were limited to 64
- auto msix bar size for virtio-net-pci were disabled by default

Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Richard Henderson <rth@twiddle.net>
Cc: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: NJason Wang <jasowang@redhat.com>
Reviewed-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

This reverts commit 13704e4c.

With the Intel microcode update that removed HLE and RTM, there will be
different kinds of Haswell and Broadwell CPUs out there: some that still
have the HLE and RTM features, and some that don't have the HLE and RTM
features. On both cases people may be willing to use the pc-*-2.3
machine-types.

So instead of making the CPU model results confusing by making it depend
on the machine-type, keep HLE and RTM on the existing Haswell and
Broadwell CPU models. The plan is to introduce "Haswell-noTSX" and
"Broadwell-noTSX" CPU models later, for people who have CPUs that don't
have TSX feature available.
Reviewed-by: NDaniel P. Berrange <berrange@redhat.com>
Signed-off-by: NEduardo Habkost <ehabkost@redhat.com>

Older PC machine types might by accident be backwards live migration compatible,
but with the new vmdesc self-describing blob in our live migration stream we
would break that compatibility.

Also users wouldn't expect massive behaviorial differences when updating to a
new version of QEMU while retaining their old machine type, especially not
potential breakage in tooling around live migration.

So disable vmdesc submission for old PC machine types.
Signed-off-by: NAlexander Graf <agraf@suse.de>
Reviewed-by: NDr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: NJuan Quintela <quintela@redhat.com>

As comment in acpi-build.c notes, RSDP is not really immutable.  So it's
really a question of whether it's in RAM, name the variable accordingly.
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>
Reviewed-by: NIgor Mammedov <imammedo@redhat.com>

Makes sure that RSDP stays the same
/i.e. matches ACPI tables blob in source/
if guest is migrated during RSDP reading or
has been already shadowed by firmware.

Fix applies only to new machine types starting
from 2.3, so it won't break migration for old
machine types.
Signed-off-by: NIgor Mammedov <imammedo@redhat.com>
Reviewed-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>
Reviewed-by: NMarcel Apfelbaum <marcel@redhat.com>

Currently, keep current behaviour by always using get_system_memory().

Also use QOM casts when possible.
Signed-off-by: NHervé Poussineau <hpoussin@reactos.org>
Reviewed-by: NPaolo Bonzini <pbonzini@redhat.com>
Signed-off-by: NLeon Alrae <leon.alrae@imgtec.com>

All Haswell CPUs and some Broadwell CPUs were updated by Intel to have
the HLE and RTM features disabled. This will prevent
"-cpu Haswell,enforce" and "-cpu Broadwell,enforce" from running out of
the box on those CPUs.

Disable those features by default on Broadwell and Haswell CPU models,
starting on pc-*-2.3. Users who want to use those features can enable
them explicitly on the command-line.
Signed-off-by: NEduardo Habkost <ehabkost@redhat.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

Due to a typo, instead of disabling KVM_FEATURE_PV_EOI (bit
6) these machine types are disabling bits 1 and 2, which are
KVM_FEATURE_NOP_IO_DELAY and KVM_FEATURE_MMU_OP.  Not a big deal
because they aren't very important and KVM_FEATURE_MMU_OP is
disabled anyway.  The worst part is actually that KVM_FEATURE_PV_EOI
is remaining enabled.
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

The argument is not longer used and the implementation
uses now QOM instead of QemuOpts.
Signed-off-by: NMarcel Apfelbaum <marcel@redhat.com>
Reviewed-by: NPaolo Bonzini <pbonzini@redhat.com>
Reviewed-by: NStefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: NAlexander Graf <agraf@suse.de>
Message-id: 1420550957-22337-4-git-send-email-marcel@redhat.com
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>

The argument is not longer used and the implementation
uses now QOM instead of QemuOpts.
Signed-off-by: NMarcel Apfelbaum <marcel@redhat.com>
Reviewed-by: NStefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: NPaolo Bonzini <pbonzini@redhat.com>
Signed-off-by: NAlexander Graf <agraf@suse.de>

Both were added in Ivy Bridge (for which we do not have a CPU model
yet!).
Reviewed-by: NEduardo Habkost <ehabkost@redhat.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

vm86 mode extensions date back to the 486.  All models should have
them.
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>
Signed-off-by: NEduardo Habkost <ehabkost@redhat.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

The next patch will differentiate them.
Reviewed-by: NEduardo Habkost <ehabkost@redhat.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

c/s 9b23cfb7

or

c/s b154537a

moved the testing of xen_enabled() from pc_init1() to
pc_machine_initfn().

xen_enabled() does not return the correct value in
pc_machine_initfn().

Changed vmport from a bool to an enum.  Added the value "auto" to do
the old way.  Move check of xen_enabled() back to pc_init1().
Acked-by: NEric Blake <eblake@redhat.com>
Reviewed-by: NEduardo Habkost <ehabkost@redhat.com>
Signed-off-by: NDon Slutz <dslutz@verizon.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>

Performance wise it's better to align GVA by the backend's
page size.

Also do not allow to create DIMM device with suboptimal
size (i.e. not aligned to backends page size) to aviod
memory loss.

Do above only for 2.2 and newer machine types to avoid
breaking working configs with 2.1 machine type.
Signed-off-by: NIgor Mammedov <imammedo@redhat.com>
Reviewed-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

If user starts QEMU with "-machine pc,accel=xen", then
compat property in xenfv won't work and it would cause error:
"Unsupported bus. Bus doesn't have property 'acpi-pcihp-bsel' set"
when PCI device is added with -device on QEMU CLI.

From: Igor Mammedov <imammedo@redhat.com>

In case of Xen instead of using compat property, just use the fact
that xen doesn't use QEMU's fw_cfg/acpi tables to switch piix4_pm
into legacy PCI hotplug mode when Xen is enabled.
Signed-off-by: NIgor Mammedov <imammedo@redhat.com>
Signed-off-by: NLi Liang <liang.z.li@intel.com>
Signed-off-by: NStefano Stabellini <stefano.stabellini@eu.citrix.com>
Acked-by: NPaolo Bonzini <pbonzini@redhat.com>

Make SVM be disabled by default on all CPU models when in KVM mode.
Nested SVM is enabled by default in the KVM kernel module, but it is
probably less stable than nested VMX (which is already disabled by
default).

Add a new compat function, x86_cpu_compat_kvm_no_autodisable(), to keep
compatibility on previous machine-types.
Suggested-by: NPaolo Bonzini <pbonzini@redhat.com>
Signed-off-by: NEduardo Habkost <ehabkost@redhat.com>
Signed-off-by: NAndreas Färber <afaerber@suse.de>

TCG doesn't support VMX, and nested VMX is not enabled by default in the
KVM kernel module.

So, there's no reason to have VMX enabled by default on the core2duo and
coreduo CPU models, today. Even the newer Intel CPU model definitions
don't have it enabled.

In this case, we need machine-type compat code, as people may be running
the older machine-types on hosts that had VMX nesting enabled.
Signed-off-by: NEduardo Habkost <ehabkost@redhat.com>
Signed-off-by: NAndreas Färber <afaerber@suse.de>