Merge remote-tracking branch 'remotes/stefanha/tags/block-pull-request' into staging

Block pull request

# gpg: Signature made Fri 14 Feb 2014 17:26:30 GMT using RSA key ID 81AB73C8
# gpg: Good signature from "Stefan Hajnoczi <stefanha@redhat.com>"
# gpg:                 aka "Stefan Hajnoczi <stefanha@gmail.com>"
# gpg: WARNING: This key is not certified with a trusted signature!
# gpg:          There is no indication that the signature belongs to the owner.
# Primary key fingerprint: 8695 A8BF D3F9 7CDA AC35  775A 9CA4 ABB3 81AB 73C8

* remotes/stefanha/tags/block-pull-request:
  block: Open by reference will try device then node_name.
  block: Relax bdrv_lookup_bs constraints.
  blockdev: Fix wrong usage of QDECREF causing snapshoted quorum to crash on close.
  block: mirror - use local_err to avoid NULL errp
  qemu-iotests: Don't run 005 on vmdk split formats
  block: qemu-iotests - add vhdx log replay tests for qemu-img
  block: qemu-iotests - fix test 070 (vhdx)
  block: Don't throw away errno via error_setg
  block: Add notes to iSCSI's .bdrv_open and .bdrv_reopen_prepare
  blockdev: Remove 'type' parameter from blockdev_init()
  sdhci: Drop unnecessary #include

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

Merge remote-tracking branch 'remotes/jliu/or32-ld-st' into staging

* remotes/jliu/or32-ld-st:
  target-openrisc: Use new qemu_ld/st opcodes

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

Merge remote-tracking branch 'remotes/jovanovic/mips-ufrp' into staging

* remotes/jovanovic/mips-ufrp:
  target-mips: add user-mode FR switch support for MIPS32r5
  target-mips: add support for CP0_Config5
  target-mips: add support for CP0_Config4
  target-mips: add CPU definition for MIPS32R5

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

block: Open by reference will try device then node_name.

Since we introduced node_name for named bs of the graph modify the opening by
reference to use it as a fallback.

This patch also enforce the separation of the device id and graph node
namespaces.

Signed-off-by: Benoit Canet <benoit@irqsave.net>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

block: Relax bdrv_lookup_bs constraints.

The following patch will reuse bdrv_lookup_bs in order to open images by
references so the rules of usage of bdrv_lookup_bs must be relaxed a bit.

Signed-off-by: Benoit Canet <benoit@irqsave.net>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

blockdev: Fix wrong usage of QDECREF causing snapshoted quorum to crash on close.

As bdrv_open() documentation states:
"The reference to the QDict belongs to the block layer
 * after the call (even on failure), so if the caller intends to reuse the
 * dictionary, it needs to use QINCREF() before calling bdrv_open."

the optional options dict will not be reused after bdrv_open() and should
belong to the block layer so remove the extra QDECREF(options).

Signed-off-by: Benoit Canet <benoit@irqsave.net>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

block: mirror - use local_err to avoid NULL errp

When starting a block job, commit_active_start() relies on whether *errp
is set by mirror_start_job.  This allows it to determine if the mirror
job start failed, so that it can clean up any changes to open flags from
the bdrv_reopen().  If errp is NULL, then it will not be able to
determine if mirror_start_job failed or not.

To avoid this, use a local Error variable, and then propagate the error
(if any) to errp.

Reported-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Jeff Cody <jcody@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

qemu-iotests: Don't run 005 on vmdk split formats

There would be too many extents that VMDK driver can't open all of them:

    005 0s ... - output mismatch (see 005.out.bad)
    --- 005.out     2013-12-24 09:27:27.608181030 +0800
    +++ 005.out.bad 2014-02-13 10:00:15.282184557 +0800
    @@ -4,10 +4,10 @@
     Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=5368709120000

     small read
    -read 4096/4096 bytes at offset 1024
    -4 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
    +qemu-io: can't open device /tmp/qemu-iotests/t.vmdk: Could not open '/tmp/qemu-iotests/t-s1016.vmdk': Too many open files
    +no file open, try 'help open'

     small write
    -wrote 4096/4096 bytes at offset 8192
    -4 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
    +qemu-io: can't open device /tmp/qemu-iotests/t.vmdk: Could not open '/tmp/qemu-iotests/t-s1016.vmdk': Too many open files
    +no file open, try 'help open'
     *** done

So disable the two subformats.

Signed-off-by: Fam Zheng <famz@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

block: qemu-iotests - add vhdx log replay tests for qemu-img

VHDX logs can now be replayed via 'qemu-img check -r all'.  Add
tests to verify that the log replay is successful when using qemu-img.

Signed-off-by: Jeff Cody <jcody@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

block: qemu-iotests - fix test 070 (vhdx)

VHDX test 070 failed, due to different output from qemu-io / qemu
when opening an image read-only that contains a log file.  Filter
the output, and update the expected results to match the correct
output.

Signed-off-by: Jeff Cody <jcody@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

block: Don't throw away errno via error_setg

There are a handful of places in the block layer where a failure path
has a valid -errno value, yet error_setg() is used.  Those instances
should instead use error_setg_errno(), to preserve as much error
information as possible.

This patch replaces those instances with error_setg_errno(), so that
errno is passed up the stack in the error message.

Reported-By: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Jeff Cody <jcody@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

block: Add notes to iSCSI's .bdrv_open and .bdrv_reopen_prepare

iSCSI currently does not need to do any actions to support the
current usage of bdrv_reopen().  However, it is important to note
a couple of things: 1.) A connection will not be re-established to
an iSCSI target, and 2.) If iscsi_open() is changed to parse 'flags',
then iscsi_reopen_prepare() may need to be more than a stub.

In light of the above, this commit adds comments above both of the
functions to bring attention to these facts.

Signed-off-by: Jeff Cody <jcody@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

blockdev: Remove 'type' parameter from blockdev_init()

blockdev-add doesn't know about the device that the backend will be
attached to, this is a legacy -drive concept. Move the remaining checks
that use it to drive_init().

[Fam Zheng <famz@redhat.com> suggested line-wrapping to 80 chars as
required by the coding standard.  I have fixed this.
--Stefan]

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

sdhci: Drop unnecessary #include

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

PPC: KVM: Add missing address space to ldl_phys helper

We now have to pass an address space to our _phys helpers. During the
transition apparently the EPR exit path missed out, so let's put it there.

Signed-off-by: Alexander Graf <agraf@suse.de>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

Merge remote-tracking branch 'remotes/mst/tags/for_upstream' into staging

acpi,pc,pci fixes and enhancements

Most changes here are hotplug related:

This merges hotplug infrastructure changes by Igor,
some acpi related fixes, and PC fixes.

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
# gpg: Signature made Mon 10 Feb 2014 09:13:26 GMT using RSA key ID D28D5469
# gpg: Good signature from "Michael S. Tsirkin <mst@kernel.org>"
# gpg:                 aka "Michael S. Tsirkin <mst@redhat.com>"
# gpg: WARNING: This key is not certified with a trusted signature!
# gpg:          There is no indication that the signature belongs to the owner.
# Primary key fingerprint: 0270 606B 6F3C DF3D 0B17  0970 C350 3912 AFBE 8E67
#      Subkey fingerprint: 5D09 FD08 71C8 F85B 94CA  8A0D 281F 0DB8 D28D 5469

* remotes/mst/tags/for_upstream:
  ACPI: Remove commented-out code from HPET._CRS
  hw/pci: switch to a generic hotplug handling for PCIDevice
  pci/pcie: convert PCIE hotplug to use hotplug-handler API
  pci/shpc: convert SHPC hotplug to use hotplug-handler API
  acpi/piix4pm: convert ACPI PCI hotplug to use hotplug-handler API
  qdev:pci: refactor PCIDevice to use generic "hotpluggable" property
  hw/acpi: move typeinfo to the file end
  qdev: add "hotpluggable" property to Device
  qdev: add to BusState "hotplug-handler" link
  define hotplug interface
  loader: document that errno is set
  pc.c: better error message on initrd sizing failure
  pc_piix: enable legacy hotplug for Xen
  qtest: don't report signals if qtest driver enabled
  hw:piix4:acpi: reuse pcihp code for legacy PCI hotplug
  pcihp: remove unused AcpiPciHpPciStatus.device_present field
  pcihp: make pci_read() mmio calback compatible with legacy ACPI hotplug
  pcihp: make PCI hotplug mmio handlers indifferent to PCI_HOTPLUG_ADDR
  pcihp: replace enable|disable_device() with oneliners
  pcihp: reduce number of device check events

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

Merge remote-tracking branch 'remotes/afaerber/tags/prep-for-upstream' into staging

PReP machine and devices

* Cleanups for Raven PCI host bridge
* Removal of PReP machine and devices from qemu-system-ppcemb

# gpg: Signature made Mon 10 Feb 2014 16:19:03 GMT using RSA key ID 3E7E013F
# gpg: Good signature from "Andreas Färber <afaerber@suse.de>"
# gpg:                 aka "Andreas Färber <afaerber@suse.com>"

* remotes/afaerber/tags/prep-for-upstream:
  prep: Drop from ppcemb-softmmu
  raven: Use constant PCI_NUM_PINS instead of 4
  prep: Kill get_system_io() usage

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

Merge remote-tracking branch 'remotes/alon/pull-libcacard.glusterfs' into staging

* remotes/alon/pull-libcacard.glusterfs:
  libcacard: Don't link with all libraries QEMU links to

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

Merge remote-tracking branch 'remotes/kevin/tags/for-anthony' into staging

Block patches

# gpg: Signature made Sun 09 Feb 2014 08:12:51 GMT using RSA key ID C88F2FD6
# gpg: Good signature from "Kevin Wolf <kwolf@redhat.com>"

* remotes/kevin/tags/for-anthony:
  block: Fix 32 bit truncation in mark_request_serialising()
  blkdebug: Don't leak bs->file on failure
  block: Don't call ROUND_UP with negative values
  block: bdrv_aligned_pwritev: Assert overlap range
  block: Fix memory leaks in bdrv_co_do_pwritev()
  raw: Fix BlockLimits passthrough
  qemu-iotests: add test for qcow2 preallocation with different cluster sizes
  qcow2: check for NULL l2meta
  qcow2: fix offset overflow in qcow2_alloc_clusters_at()
  qcow2: remove n_start and n_end of qcow2_alloc_cluster_offset()
  block/iscsi: always fill bs->bl.opt_transfer_length
  block: Fail gracefully with missing filename
  qemu-iotests: enable support for NFS protocol
  qemu-iotests: enable test 016 and 025 to work with NFS protocol
  qemu-iotests: blacklist test 020 for NFS protocol
  qemu-iotests: change _supported_proto to file for various tests
  block: add native support for NFS
  qemu-iotest: Make 077 raw-only

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target-openrisc: Use new qemu_ld/st opcodes

Signed-off-by: Richard Henderson <rth@twiddle.net>
Acked-by: Jia Liu <proljc@gmail.com>
Signed-off-by: Jia Liu <proljc@gmail.com>

petalogix-ml605: Create the CPU with object_new()

This is to allow future patches to set properties before cpu::realize().

Reviewed-by: Andreas Färber <afaerber@suse.de>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

exec: Make cpu_memory_rw_debug use the CPUs AS

Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

exec: Make cpu_physical_memory_write_rom input an AS

Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

exec: Make stb_phys input an AddressSpace

Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

exec: Make stw_*_phys input an AddressSpace

Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

exec: Make stl_phys_notdirty input an AddressSpace

Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

exec: Make stl_*_phys input an AddressSpace

Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

exec: Make stq_*_phys input an AddressSpace

Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

exec: Make lduw_*_phys input an AddressSpace

Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

exec: Make ldq/ldub_*_phys input an AddressSpace

Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

exec: Make ldl_*_phys input an AddressSpace

Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

exec: On AS changes, only flush affected CPU TLBs

Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

cpu: Add per-cpu address space

Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

memory: Add MemoryListener to typedefs.h

Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

exec: Make memory_region_section_get_iotlb use section AS

Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

exec: Always initialize MemorySection address spaces

Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

exec: Make iotlb_to_region input an AS

Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

exec: Make tb_invalidate_phys_addr input an AS

No functional change.

Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>

Merge remote-tracking branch 'remotes/pmaydell/tags/pull-target-arm-20140208' into staging

target-arm queue:
 * more A64 Neon instructions
 * AArch32 VCVTB and VCVTT ARMv8 instructions
 * fixes to inaccuracies in GIC emulation
 * libvixl disassembler for A64
 * Allwinner SoC ethernet controller
 * zynq software system reset support

# gpg: Signature made Sat 08 Feb 2014 15:53:05 GMT using RSA key ID 14360CDE
# gpg: Good signature from "Peter Maydell <peter.maydell@linaro.org>"

* remotes/pmaydell/tags/pull-target-arm-20140208: (29 commits)
  arm/zynq: Add software system reset via SCLR
  hw/arm/allwinner-a10: initialize EMAC
  hw/net: add support for Allwinner EMAC Fast Ethernet controller
  util/fifo8: clear fifo head upon reset
  util/fifo8: implement push/pop of multiple bytes
  disas: Implement disassembly output for A64
  disas/libvixl: Fix upstream libvixl compilation issues
  disas: Add subset of libvixl sources for A64 disassembler
  rules.mak: Link with C++ if we have a C++ compiler
  rules.mak: Support .cc as a C++ source file suffix
  arm_gic: Add GICC_APRn state to the GICState
  vmstate: Add uint32 2D-array support
  arm_gic: Support setting/getting binary point reg
  arm_gic: Keep track of SGI sources
  arm_gic: Fix GIC pending behavior
  target-arm: Add support for AArch32 64bit VCVTB and VCVTT
  target-arm: A64: Add FNEG and FABS to the SIMD 2-reg-misc group
  target-arm: A64: Add 2-reg-misc REV* instructions
  target-arm: A64: Add narrowing 2-reg-misc instructions
  target-arm: A64: Implement 2-reg-misc CNT, NOT and RBIT
  ...

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

Merge remote-tracking branch 'remotes/kvaneesh/for-upstream' into staging

* remotes/kvaneesh/for-upstream:
  hw/9pfs: fix P9_STATS_GEN handling
  hw/9pfs: make get_st_gen() return ENOTTY error on special files
  hw/9pfs: handle undefined FS_IOC_GETVERSION case in handle_ioc_getversion()
  hw/9pfs: fix error handing in local_ioc_getversion()

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

memory: fix limiting of translation at a page boundary

Commit 360e607 (address_space_translate: do not cross page boundaries,
2014-01-30) broke MMIO accesses in cases where the section is shorter
than the full register width.  This can happen for example with the
Bochs DISPI registers, which are 16 bits wide but have only a 1-byte
long MemoryRegion (if you write to the "second byte" of the register
your access is discarded; it doesn't write only to half of the register).

Restrict the action of commit 360e607 to direct RAM accesses.  This
is enough for Xen, since MMIO will not go through the mapcache.

Reported-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Cc: qemu-stable@nongnu.org
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Tested-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

prep: Drop from ppcemb-softmmu

ppcemb covers only embedded processors, which does not include PReP.

Signed-off-by: Andreas Färber <andreas.faerber@web.de>

target-mips: add user-mode FR switch support for MIPS32r5

Description of UFR feature:

Required in MIPS32r5 if floating point is implemented and user-mode FR
switching is supported. The UFR register allows user-mode to clear StatusFR
by executing a CTC1 to UFR with GPR[0] as input, and read StatusFR by
executing a CFC1 to UFR.

helper_ctc1 has been extended with an additional parameter rt to check
requirements for UFR feature.
Definition of mips32r5-generic has been modified to include support for UFR.

Signed-off-by: Petar Jovanovic <petar.jovanovic@imgtec.com>
Reviewed-by: Eric Johnson <eric.johnson@imgtec.com>

target-mips: add support for CP0_Config5

Add CP0_Config5, define rw_bitmask and enable modifications.

Signed-off-by: Petar Jovanovic <petar.jovanovic@imgtec.com>
Reviewed-by: Eric Johnson <eric.johnson@imgtec.com>

target-mips: add support for CP0_Config4

Add CP0_Config4, define rw_bitmask.

Signed-off-by: Petar Jovanovic <petar.jovanovic@imgtec.com>
Reviewed-by: Eric Johnson <eric.johnson@imgtec.com>

target-mips: add CPU definition for MIPS32R5

Add mips32r5-generic among CPU definitions for MIPS.
Define ISA_MIPS32R3 and ISA_MIPS32R5.

Signed-off-by: Petar Jovanovic <petar.jovanovic@imgtec.com>
Reviewed-by: Eric Johnson <eric.johnson@imgtec.com>

ACPI: Remove commented-out code from HPET._CRS

IRQNoFlags on HPET._CRS crashes WinXP because it causes the HPET
to conflict with the system timer and/or the RTC. It only occurs
on Apple hardware, and even there it is exposed fully only when
OS X is detected (via _OSI). Recent OS X versions work on QEMU
without this statement, so at this time there is no need to find
a better way to conditionally include the statement. This patch
removes the commented out (and wrong, should have been {0, 8})
statement from HPET._CRS.

Signed-off-by: Gabriel Somlo <somlo@cmu.edu>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-By: Igor Mammedov <imammedo@redhat.com>

hw/pci: switch to a generic hotplug handling for PCIDevice

make qdev_unplug()/device_set_realized() to call hotplug handler's
plug/unplug methods if available and remove not needed anymore
hot(un)plug handling from PCIDevice.

In case if hotplug handler is not available, revert to the legacy
hotplug method for compatibility with not yet converted buses.

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

pci/pcie: convert PCIE hotplug to use hotplug-handler API

Split pcie_cap_slot_hotplug() into hotplug/unplug callbacks
and register them as "hotplug-handler" interface implementation of
PCIE_SLOT device.

Replace pci_bus_hotplug() wiring with setting link on PCI BUS
"hotplug-handler" property to PCI_BRIDGE_DEV device.

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

pci/shpc: convert SHPC hotplug to use hotplug-handler API

Split shpc_device_hotplug() into hotplug/unplug callbacks
and register them as "hotplug-handler" interface implementation of
PCI_BRIDGE_DEV device.

Replace pci_bus_hotplug() wiring with setting link on PCI BUS
"hotplug-handler" property to PCI_BRIDGE_DEV device.

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

acpi/piix4pm: convert ACPI PCI hotplug to use hotplug-handler API

Split piix4_device_hotplug() into hotplug/unplug callbacks
and register them as "hotplug-handler" interface implementation of
PIIX4_PM device.

Replace pci_bus_hotplug() wiring with setting link on
PCI BUS "hotplug-handler" property to PIIX4_PM device.

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

qdev:pci: refactor PCIDevice to use generic "hotpluggable" property

Get rid of PCIDevice specific PCIDeviceClass.no_hotplug and use
generic DeviceClass.hotpluggable field instead.

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

hw/acpi: move typeinfo to the file end

do so to avoid not necessary forward declarations and
place typeinfo registration at the file end where it's
usually expected.

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

qdev: add "hotpluggable" property to Device

Currently it's possible to make PCIDevice not hotpluggable
by using no_hotplug field of PCIDeviceClass. However it
limits this only to PCI devices and prevents from
generalizing hotplug code.

So add similar field to DeviceClass so it could be reused
with other Devices and would allow to replace PCI specific
hotplug callbacks with generic implementation. Following
patches will replace PCIDeviceClass.no_hotplug with this
new property.

In addition expose field as "hotpluggable" readonly property,
to make it possible to read its value via QOM interface.

Make DeviceClass hotpluggable by default as it was assumed
before.

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

qdev: add to BusState "hotplug-handler" link

It will allow to reuse field with different BUSes,
reducing code duplication. Field is intended for
replacing 'hotplug_qdev' field in PCIBus and also
will allow to avoid adding equivalent field to
DimmBus with possiblitity to refactor other BUSes
to use it instead of custom field.
In addition once all users of allow_hotplug field
are converted to new API, link could replace
allow_hotplug field in qdev hotplug code.

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

define hotplug interface

Provide a generic hotplug interface for hotplug handlers.
Intended for replacing hotplug mechanism used by
PCI/PCIE/SHPC code and will be used for memory hotplug.

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

loader: document that errno is set

Document that get_image_size sets errno
on failure.

Suggested-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

raven: Use constant PCI_NUM_PINS instead of 4

Signed-off-by: Hervé Poussineau <hpoussin@reactos.org>
Signed-off-by: Andreas Färber <andreas.faerber@web.de>

prep: Kill get_system_io() usage

While ISA address space in prep machine is currently the one returned
by get_system_io(), this depends of the implementation of i82378/raven
devices, and this may not be the case forever.

Use the right ISA address space when adding some more ports to it.
We can use whatever ISA device on the right ISA bus, as all ISA devices
on the same ISA bus share the same ISA address space.

Signed-off-by: Hervé Poussineau <hpoussin@reactos.org>
Signed-off-by: Andreas Färber <andreas.faerber@web.de>

libcacard: Don't link with all libraries QEMU links to

As described in https://bugzilla.redhat.com/show_bug.cgi?id=987441 ,
libcacard currently links to all the libraries QEMU is linking to,
including glusterfs libraries, libiscsi, ... libcacard does not need all of
these. This patch ensures it's only linked with the libraries it needs.

Signed-off-by: Christophe Fergeau <cfergeau@redhat.com>
Signed-off-by: Alon Levy <alevy@redhat.com>

block: Fix 32 bit truncation in mark_request_serialising()

On 32 bit hosts, size_t is too small for align as the bitmask
~(align - 1) will zero out the higher 32 bits of the offset.

While at it, change the local overlap_bytes variable to unsigned to
match the field in BdrvTrackedRequest.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>

blkdebug: Don't leak bs->file on failure

Reported-by: Laszlo Ersek <lersek@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>

block: Don't call ROUND_UP with negative values

The behaviour of the ROUND_UP macro with negative numbers isn't obvious.
It happens to do the right thing in this please, but better avoid it.

Suggested-by: Laszlo Ersek <lersek@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>

block: bdrv_aligned_pwritev: Assert overlap range

This adds assertions that the request that we actually end up passing to
the block driver (which includes RMW data and has therefore potentially
been rounded to alignment boundaries) is fully covered by the
overlap_{offset,size} fields of the associated BdrvTrackedRequest.

Suggested-by: Laszlo Ersek <lersek@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>

block: Fix memory leaks in bdrv_co_do_pwritev()

The error path for a failure in one of the two bdrv_aligned_preadv()
calls leaked head_buf or tail_buf, respectively. This fixes the memory
leak.

Reported-by: Laszlo Ersek <lersek@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>

raw: Fix BlockLimits passthrough

raw copies over the BlockLimits of bs->file during bdrv_open().
However, since commit d34682cd it is immediately overwritten during
bdrv_refresh_limits(). This caused all fields except for
opt_transfer_length and opt_mem_alignment (which happen to be correctly
inherited in generic code) to be zeroed.

Move the BlockLimit assignment to a .bdrv_refresh_limits() callback to
make it work again for all fields.

Reported-by: Laszlo Ersek <lersek@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>

qemu-iotests: add test for qcow2 preallocation with different cluster sizes

Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Hu Tao <hutao@cn.fujitsu.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

qcow2: check for NULL l2meta

In the case of a metadata preallocation with a large cluster size,
qcow2_alloc_cluster_offset() can allocate nothing and returns a
NULL l2meta. This patch checks for it and link2 l2 with only valid
l2meta.

Replace 9 and 512 with BDRV_SECTOR_BITS, BDRV_SECTOR_SIZE
respectively while at the function.

Signed-off-by: Hu Tao <hutao@cn.fujitsu.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Benoit Canet <benoit@irqsave.net>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

qcow2: fix offset overflow in qcow2_alloc_clusters_at()

When cluster size is big enough it can lead to an offset overflow
in qcow2_alloc_clusters_at(). This patch fixes it.

The allocation is stopped each time at L2 table boundary
(see handle_alloc()), so the possible maximum bytes could be

  2^(cluster_bits - 3 + cluster_bits)

cluster_bits - 3 is used to compute the number of entry by L2
and the additional cluster_bits is to take into account each
clusters referenced by the L2 entries.

so int is safe for cluster_bits<=17, unsafe otherwise.

Signed-off-by: Hu Tao <hutao@cn.fujitsu.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Benoit Canet <benoit@irqsave.net>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

qcow2: remove n_start and n_end of qcow2_alloc_cluster_offset()

n_start can be actually calculated from offset. The number of
sectors to be allocated(n_end - n_start) can be passed in in
num. By removing n_start and n_end, we can save two parameters.

The side effect is there is a bug in qcow2.c:preallocate() that
passes incorrect n_start to qcow2_alloc_cluster_offset() is
fixed. The bug can be triggerred by a larger cluster size than
the default value(65536), for example:

./qemu-img create -f qcow2 \
  -o 'cluster_size=131072,preallocation=metadata' file.img 4G

Signed-off-by: Hu Tao <hutao@cn.fujitsu.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Benoit Canet <benoit@irqsave.net>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

block/iscsi: always fill bs->bl.opt_transfer_length

the opt_transfer_length has nothing to do with logical
block provisioning stuff so always copy it from
the block limits VPD page.

Reported-By: Benoit Canet <benoit@irqsave.net>
Signed-off-by: Peter Lieven <pl@kamp.de>
Reviewed-by: Benoit Canet <benoit@irqsave.net>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

block: Fail gracefully with missing filename

This fixes a regression introduced in commit 2a05cbe42 ('block: Allow
block devices without files'):

$ qemu-system-x86_64 -drive driver=file
qemu-system-x86_64: block.c:892: bdrv_open_common: Assertion
`!drv->bdrv_needs_filename || filename != ((void *)0)' failed.

Now the respective check must be performed not only in bdrv_file_open(),
but also in bdrv_open().

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>

qemu-iotests: enable support for NFS protocol

Signed-off-by: Peter Lieven <pl@kamp.de>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

qemu-iotests: enable test 016 and 025 to work with NFS protocol

Signed-off-by: Peter Lieven <pl@kamp.de>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

qemu-iotests: blacklist test 020 for NFS protocol

reopening is currently not supported.

Signed-off-by: Peter Lieven <pl@kamp.de>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

qemu-iotests: change _supported_proto to file for various tests

all these tests do anything of the following and thus fail with any
protocol other than file:
 - the tests use rm, cp or mv shell commands which only work on file
 - the tests use qcow2.py
 - the images construct new filenames (e.g. backing file names) and
   the logic is broken for anything else than file

Signed-off-by: Peter Lieven <pl@kamp.de>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

block: add native support for NFS

This patch adds native support for accessing images on NFS
shares without the requirement to actually mount the entire
NFS share on the host.

NFS Images can simply be specified by an url of the form:
nfs://<host>/<export>/<filename>[?param=value[&param2=value2[&...]]]

For example:
qemu-img create -f qcow2 nfs://10.0.0.1/qemu-images/test.qcow2

You need LibNFS from Ronnie Sahlberg available at:
   git://github.com/sahlberg/libnfs.git
for this to work.

During configure it is automatically probed for libnfs and support
is enabled on-the-fly. You can forbid or enforce libnfs support
with --disable-libnfs or --enable-libnfs respectively.

Due to NFS restrictions you might need to execute your binaries
as root, allow them to open priviledged ports (<1024) or specify
insecure option on the NFS server.

For additional information on ROOT vs. non-ROOT operation and URL
format + parameters see:
   https://raw.github.com/sahlberg/libnfs/master/README

Supported by qemu are the uid, gid and tcp-syncnt URL parameters.

LibNFS currently support NFS version 3 only.

Signed-off-by: Peter Lieven <pl@kamp.de>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

qemu-iotest: Make 077 raw-only

The qemu-io command sequences make the assumption that an unaligned
request on the format layer will be unaligned on the blkdebug layer as
well. This doesn't necessarily hold true for drivers other than raw.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Benoît Canet <benoit@irqsave.net>

Merge remote-tracking branch 'remotes/mwalle/tags/lm32-fixes/20140204' into staging

target-lm32: fixes

# gpg: Signature made Tue 04 Feb 2014 18:47:56 GMT using DSA key ID 3F98A378
# gpg: Can't check signature: public key not found

* remotes/mwalle/tags/lm32-fixes/20140204:
  hw/lm32: print error if cpu model is not found
  target-lm32: stop VM on illegal or unknown instruction
  lm32_sys: dump cpu state if test case fails
  lm32_sys: print test result on stderr
  target-lm32: add breakpoint/watchpoint support
  target-lm32: move model features to LM32CPU
  target-lm32: kill cpu_abort() calls
  milkymist-vgafb: swap pixel data in source buffer
  lm32_uart/lm32_juart: use qemu_chr_fe_write_all()
  milkymist-uart: use qemu_chr_fe_write_all() instead of qemu_chr_fe_write()
  tests: lm32: new rule for single test cases
  lm32_sys: increase test case name length limit

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

arm/zynq: Add software system reset via SCLR

Support software-driven system reset via the register in the SCLR.

Signed-off-by: Sebastian Huber <sebastian.huber@embedded-brains.de>
Reviewed-by: Peter Crosthwaite <peter.crosthwaite@xilinx.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

hw/arm/allwinner-a10: initialize EMAC

Signed-off-by: Beniamino Galvani <b.galvani@gmail.com>
Reviewed-by: Peter Crosthwaite <peter.crosthwaite@xilinx.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

hw/net: add support for Allwinner EMAC Fast Ethernet controller

This patch adds support for the Fast Ethernet MAC found on Allwinner
SoCs, together with a basic emulation of Realtek RTL8201CP PHY.

Since there is no public documentation of the Allwinner controller, the
implementation is based on Linux kernel driver.

Signed-off-by: Beniamino Galvani <b.galvani@gmail.com>
Reviewed-by: Peter Crosthwaite <peter.crosthwaite@xilinx.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

util/fifo8: clear fifo head upon reset

To improve the predictability of fifo8_pop_buf(), the fifo head is set
to the start of data buffer upon a reset so that the first call to the
function will be able to retrieve all data in the fifo.

Signed-off-by: Beniamino Galvani <b.galvani@gmail.com>
Reviewed-by: Peter Crosthwaite <peter.crosthwaite@xilinx.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

util/fifo8: implement push/pop of multiple bytes

The patch adds functions fifo8_push_all() and fifo8_pop_buf() which
can be used respectively to push the content of a memory buffer to the
fifo and to pop multiple bytes obtaining a pointer to the fifo backing
buffer.

In addition, it implements fifo8_num_free() and fifo8_num_used() which
allow to check if a multi-byte operation can be performed.

Signed-off-by: Beniamino Galvani <b.galvani@gmail.com>
Reviewed-by: Peter Crosthwaite <peter.crosthwaite@xilinx.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

disas: Implement disassembly output for A64

Use libvixl to implement disassembly output in debug
logs for A64, for use with both AArch64 hosts and targets.

Signed-off-by: Claudio Fontana <claudio.fontana@linaro.org>
[PMM:
 * added support for target disassembly
 * switched to custom QEMUDisassembler so the output format
   matches what QEMU expects
 * make sure we correctly fall back to "just print hex"
   if we didn't build the AArch64 disassembler because of
   lack of a C++ compiler
 * rename from 'aarch64' to 'arm-a64' because this is a
   disassembler for the A64 instruction set
 * merge aarch64.c and aarch64-cxx.cc into one C++ file
 * simplify the aarch64.c<->aarch64-cxx.cc interface]
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

disas/libvixl: Fix upstream libvixl compilation issues

Fix various minor issues with upstream libvixl so that it will compile
successfully on the platforms QEMU cares about:
 * remove unused GBytes constant (it clashes with the glib headers)
 * fix suffixes on constants to use 'LL' for 64 bit constants so
   we can compile on 32 bit hosts

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <rth@twiddle.net>

disas: Add subset of libvixl sources for A64 disassembler

Add the subset of the libvixl sources that are needed for the
A64 disassembler support. These sources come from
https://github.com/armvixl/vixl commit 578645f14e122d2b
which is VIXL release 1.1.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <rth@twiddle.net>

rules.mak: Link with C++ if we have a C++ compiler

If we have a C++ compiler available, link with it, because we might be
linking some C++ files in. This allows us to include C++ object files
in the QEMU binary proper.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <rth@twiddle.net>

rules.mak: Support .cc as a C++ source file suffix

The A64 disassembler libvixl uses .cc as its suffix for
C++ source files, so add support for it (we already support
.cpp).

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <rth@twiddle.net>

arm_gic: Add GICC_APRn state to the GICState

The GICC_APRn registers are not currently supported by the ARM GIC v2.0
emulation.  This patch adds the missing state.

Note that we also change the number of APRs to use a define GIC_NR_APRS
based on the maximum number of preemption levels.  This patch also adds
RAZ/WI accessors for the four registers on the emulated CPU interface.

Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

vmstate: Add uint32 2D-array support

Add support for saving VMState of 2D arrays of uint32 values.

Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

arm_gic: Support setting/getting binary point reg

Add a binary_point field to the gic emulation structure and support
setting/getting this register now when we have it.  We don't actually
support interrupt grouping yet, oh well.

Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

arm_gic: Keep track of SGI sources

Right now the arm gic emulation doesn't keep track of the source of an
SGI (which apparently Linux guests don't use, or they're fine with
assuming CPU 0 always).

Add the necessary matrix on the GICState structure and maintain the data
when setting and clearing the pending state of an IRQ and make the state
visible to the guest.

Note that we always choose to present the source as the lowest-numbered
CPU in case multiple cores have signalled the same SGI number to a core
on the system.

Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

arm_gic: Fix GIC pending behavior

The existing implementation of the pending behavior in gic_set_irq,
gic_complete_irq, and the distributor pending set/clear registers does
not follow the semantics of the GICv2.0 specs, but may implement the
11MPCore support.  Therefore, maintain the existing semantics for
11MPCore and v7M NVIC and change the behavior to be in accordance with
the GICv2.0 specs for "generic implementations" (s->revision == 1 ||
s->revision == 2).

Generic implementations distinguish between setting a level-triggered
interrupt pending through writes to the GICD_ISPENDR and when hardware
raises the interrupt line.  Writing to the GICD_ICPENDR will not cause
the interrupt to become non-pending if the line is still active, and
conversely, if the line is deactivated but the interrupt is marked as
pending through a write to GICD_ISPENDR, the interrupt remains pending.
Handle this situation in the GIC_TEST_PENDING (which now becomes a
static inline named gic_test_pending) and let the 'pending' field
correspond only to the latched state of the D-flip flop in the GICv2.0
specs Figure 4-10.

The following changes are added:

gic_test_pending:
Make this a static inline and split out the 11MPCore from the generic
behavior.  For the generic behavior, consider interrupts pending if:
    ((s->irq_state[irq].pending & (cm) != 0) ||
       (!GIC_TEST_EDGE_TRIGGER(irq) && GIC_TEST_LEVEL(irq, cm))

gic_set_irq:
Split out the 11MPCore from the generic behavior.  For the generic
behavior, always GIC_SET_LEVEL() on positive level, but only
GIC_SET_PENDING for edge-triggered interrupts and always simply
GIC_CLEAR_LEVEL() on negative level.

gic_complete_irq:
Only resample the line for line-triggered interrupts on an 11MPCore.
Generic implementations will sample the line directly in
gic_test_pending().

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target-arm: Add support for AArch32 64bit VCVTB and VCVTT

Add support for the AArch32 floating-point half-precision to double-
precision conversion VCVTB and VCVTT instructions.

Signed-off-by: Will Newton <will.newton@linaro.org>
[PMM: fixed a minor missing-braces style issue]
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target-arm: A64: Add FNEG and FABS to the SIMD 2-reg-misc group

Add the SIMD FNEG and FABS instructions in the SIMD 2-reg-misc group.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <rth@twiddle.net>

target-arm: A64: Add 2-reg-misc REV* instructions

Add the byte-reverse operations REV64, REV32 and REV16 from the
two-reg-misc group.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <rth@twiddle.net>

target-arm: A64: Add narrowing 2-reg-misc instructions

Add the narrowing integer instructions in the 2-reg-misc class.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <rth@twiddle.net>

target-arm: A64: Implement 2-reg-misc CNT, NOT and RBIT

Implement the 2-reg-misc CNT, NOT and RBIT instructions.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <rth@twiddle.net>

target-arm: A64: Implement 2-register misc compares, ABS, NEG

Implement the simple 2-register-misc operations we can share
with the scalar-two-register-misc code. (SUQADD, USQADD, SQABS,
SQNEG also fall into this category, but aren't implemented in
the scalar-2-register case yet either.)

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <rth@twiddle.net>