Greg Kurz [Wed, 17 Jun 2015 13:23:44 +0000 (15:23 +0200)]
tap: add VNET_LE/VNET_BE operations
The linux tap and macvtap backends can be told to parse vnet headers
according to little or big endian. This is done through the TUNSETVNETLE
and TUNSETVNETBE ioctls.
This patch brings all the plumbing for QEMU to use these APIs.
Greg Kurz [Wed, 17 Jun 2015 13:23:39 +0000 (15:23 +0200)]
vhost: set vring endianness for legacy virtio
Legacy virtio is native endian: if the guest and host endianness differ,
we have to tell vhost so it can swap bytes where appropriate. This is
done through a vhost ring ioctl.
Laszlo Ersek [Thu, 11 Jun 2015 00:37:59 +0000 (02:37 +0200)]
i386/acpi-build: fix PXB workarounds for unsupported BIOSes
The patch
apci: fix PXB behaviour if used with unsupported BIOS
uses the following condition to see if a "PXB mem/IO chunk" has *not* been
configured by the BIOS:
(!range_base || range_base > range_limit)
When this condition evaluates to true, said patch *omits* the
corresponding entry from the _CRS.
Later on the patch checks for the opposite condition (with the intent of
*adding* entries to the _CRS if the "PXB mem/IO chunks" *have* been
configured). Unfortunately, the condition was negated incorrectly: only
the first ! operator was removed, which led to the nonsensical expression
(range_base || range_base > range_limit)
leading to bogus entries in the _CRS, and causing BSOD in Windows Server
2012 R2 when it runs on OVMF.
The correct negative of the condition seen at the top is
Laszlo Ersek [Thu, 11 Jun 2015 00:37:58 +0000 (02:37 +0200)]
i386/acpi-build: more traditional _UID and _HID for PXB root buses
The ACPI specification permits the _HID and _UID objects to evaluate to
strings. (See "6.1.5 _HID (Hardware ID)" and "6.1.12 _UID (Unique ID)" in
the ACPI v6.0 spec.)
With regard to related standards, the UEFI specification can also express
a device address composed from string _HID and _UID identifiers, inside
the Expanded ACPI Device Path Node. (See "9.3.3 ACPI Device Path", Table
49, in the UEFI v2.5 spec.)
However, numeric (integer) contents for both _HID and _UID are more
traditional. They are recommended by the UEFI spec for size reasons:
[...] the ACPI Device Path node is smaller and should be used if
possible to reduce the size of device paths that may potentially be
stored in nonvolatile storage [...]
External tools support them better (for example the --acpi_hid and
--acpi_uid options of "efibootmgr" only take numeric identifiers).
Finally, numeric _HID and _UID contents are existing practice in the QEMU
source.
This patch was tested with a Fedora 20 LiveCD and a preexistent Windows
Server 2012 R2 guest. Using "acpidump" and "iasl" in the Fedora guest, we
get, in the SSDT:
> Scope (\_SB)
> {
> Device (PC04)
> {
> Name (_UID, 0x04) // _UID: Unique ID
> Name (_HID, EisaId ("PNP0A03") /* PCI Bus */) // _HID: Hardware ID
Shannon Zhao [Wed, 10 Jun 2015 15:04:35 +0000 (23:04 +0800)]
virtio-9p-device: move qdev properties into virtio-9p-device.c
As only one place in virtio-9p-device.c uses
DEFINE_VIRTIO_9P_PROPERTIES, there is no need to expose it. Inline it
into virtio-9p-device.c to avoid wrongly use.
Shannon Zhao [Wed, 10 Jun 2015 15:04:34 +0000 (23:04 +0800)]
virtio-serial-bus: move qdev properties into virtio-serial-bus.c
As only one place in virtio-serial-bus.c uses
DEFINE_VIRTIO_SERIAL_PROPERTIES, there is no need to expose it. Inline
it into virtio-serial-bus.c to avoid wrongly use.
Shannon Zhao [Wed, 10 Jun 2015 15:04:32 +0000 (23:04 +0800)]
virtio-scsi: move qdev properties into virtio-scsi.c
As only one place in virtio-scsi.c uses DEFINE_VIRTIO_SCSI_PROPERTIES
and DEFINE_VIRTIO_SCSI_FEATURES, there is no need to expose them. Inline
them into virtio-scsi.c to avoid wrongly use.
Gerd Hoffmann [Thu, 4 Jun 2015 10:34:42 +0000 (12:34 +0200)]
virtio-input: emulated devices [pci]
This patch adds virtio-pci support for the emulated virtio-input
devices. Using them is as simple as adding "-device virtio-tablet-pci"
to your command line. If you want add multiple devices but don't want
waste a pci slot for each you can compose a multifunction device this way:
Gerd Hoffmann [Thu, 4 Jun 2015 10:34:39 +0000 (12:34 +0200)]
virtio-pci: fill VirtIOPCIRegions early.
Initialize the modern bar and the VirtIOPCIRegion fields early, in
realize. Also add a size field to VirtIOPCIRegion and variables for
pci bars to VirtIOPCIProxy.
This allows virtio-pci subclasses to change things before the
device_plugged callback applies them. virtio-vga will use that to
arrange regions in a way that virtio-vga is compatible to both stdvga
(in vga mode) and virtio-gpu-pci (in pci mode).
Jason Wang [Thu, 4 Jun 2015 10:34:30 +0000 (12:34 +0200)]
virtio-pci: correctly set host notifiers for modern bar
Currently, during host notifier set. We only add eventfd for legacy
bar, this is not correct since:
- Non-transitional device does not have legacy bar, so qemu will crash
since proxy->bar was not initialized.
- Modern device uses modern bar and notify cap to notify the device,
we should add eventfd for proxy->notify.
So this patch fixes the above two issues by adding eventfd based on
whether legacy or modern device were supported.
Gerd Hoffmann [Thu, 4 Jun 2015 10:34:28 +0000 (12:34 +0200)]
virtio-pci: change & document virtio pci bar layout.
This patch adds variables for the pci bars (to get rid of the magic
numbers in the code) and moves the modern virtio bar to region 4 so
regions 2+3 are kept free. virtio-vga wants use them.
Gerd Hoffmann [Thu, 4 Jun 2015 10:34:26 +0000 (12:34 +0200)]
virtio-pci: add flags to enable/disable legacy/modern
Add VIRTIO_PCI_FLAG_DISABLE_LEGACY and VIRTIO_PCI_FLAG_DISABLE_MODERN
for VirtIOPCIProxy->flags. Also add properties for them. They can be
used to disable modern (virtio 1.0) or legacy (virtio 0.9) modes.
By default only legacy is advertized, modern will be turned on by
default once all remaining spec compilance issues are addressed.
virtio 1.0 defines config space as LE,
as opposed to pre-1.0 which was native endian.
Add API for transports to execute word/dword accesses in
little endian format - will be useful for mmio
and pci (byte access is also wrapped, for completeness).
For simplicity, we still keep config in host native
endian format, byteswap to LE on guest access.
Cornelia Huck [Thu, 4 Jun 2015 10:34:18 +0000 (12:34 +0200)]
virtio-net: enable virtio 1.0
virtio-net (non-vhost) now should have everything in place to support
virtio 1.0: let's enable the feature bit for it.
Note that VIRTIO_F_VERSION_1 is technically a transport feature; once
every device is ready for virtio 1.0, we can move setting this
feature bit out of the individual devices.
Cornelia Huck [Thu, 4 Jun 2015 10:34:15 +0000 (12:34 +0200)]
virtio: allow to fail setting status
virtio-1 allow setting of the FEATURES_OK status bit to fail if
the negotiated feature bits are inconsistent: let's fail
virtio_set_status() in that case and update virtio-ccw to post an
error to the guest.
Cornelia Huck [Thu, 4 Jun 2015 10:34:12 +0000 (12:34 +0200)]
virtio: allow virtio-1 queue layout
For virtio-1 devices, we allow a more complex queue layout that doesn't
require descriptor table and rings on a physically-contigous memory area:
add virtio_queue_set_rings() to allow transports to set this up.
Cornelia Huck [Thu, 4 Jun 2015 10:34:11 +0000 (12:34 +0200)]
virtio: endianness checks for virtio 1.0 devices
Add code that checks for the VERSION_1 feature bit in order to make
decisions about the device's endianness. This allows us to support
transitional devices.
Jason Wang [Thu, 4 Jun 2015 09:28:46 +0000 (05:28 -0400)]
vhost: logs sharing
Currently we allocate one vhost log per vhost device. This is sub
optimal when:
- Guest has several device with vhost as backend
- Guest has multiqueue devices
In the above cases, we can avoid the memory allocation by sharing a
single vhost log among all the vhost devices. This is done through:
- Introducing a new vhost_log structure with refcnt inside.
- Using a global pointer to vhost_log structure that will be used. And
introduce helper to get the log with expected log size and helper to
- drop the refcnt to the old log.
- Each vhost device still keep track of a pointer to the log that was
used.
With above, if no resize happens, all vhost device will share a single
vhost log. During resize, a new vhost_log structure will be allocated
and made for the global pointer. And each vhost devices will drop the
refcnt to the old log.
Tested by doing scp during migration for a 2 queues virtio-net-pci.
hw/acpi: piix4_pm_init(): take fw_cfg object no more
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.
hw/acpi: move "etc/system-states" fw_cfg file from PIIX4 to core
The acpi_pm1_cnt_init() core function is responsible for setting up the
register block that will ultimately react to S3 and S4 requests (see
acpi_pm1_cnt_write()). It makes sense to advertise this configuration to
the guest firmware via an easy to parse fw_cfg file (ACPI is too complex
for firmware to parse), and indeed PIIX4 does that. However, since
acpi_pm1_cnt_init() is not specific to PIIX4, neither should be the fw_cfg
file.
This patch makes "etc/system-states" appear on all chipsets modified in
the previous patch, not just PIIX4 (assuming they have fw_cfg at all).
hw/acpi: acpi_pm1_cnt_init(): take "disable_s3" and "disable_s4"
This patch only modifies the function prototype and updates all chipset
code that calls acpi_pm1_cnt_init() to pass in their own disable_s3 and
disable_s4 settings. vt82c686 is assumed to be fixed "S3 and S4 enabled".
Igor Mammedov [Wed, 3 Jun 2015 15:10:43 +0000 (17:10 +0200)]
pc-dimm: don't assert if pc-dimm alignment != hotpluggable mem range size
Drop superfluous pc-dimm alignment on hot-pluggable mem
range size assert, since it causes QEMU crash during hotplug
when hotplugging pc-dimm with alignment bigger than
an alignment of hot-pluggable mem range size.
Instead allow pc_dimm_get_free_addr() find free address
and bail out gracefully later in that function during
checking if pc-dimm will fit in hot-pluggable mem range.
apci: fix PXB behaviour if used with unsupported BIOS
PXB does not work with unsupported bioses, but should
not interfere with normal OS operation.
We don't ship them anymore, but it's reasonable
to keep the work-around until we update the bios in qemu.
Fix this by not adding PXB mem/IO chunks to _CRS
if they weren't configured by BIOS.
The bios does not index the pxb slot number when
it computes the IRQ because it resides on bus 0
and not on the current bus.
However Qemu routes the irq through bus 0 and adds
the pxb slot to the IRQ computation of the PXB device.
Synchronize between bios and Qemu by canceling
pxb's effect.
PXB is a "light-weight" host bridge whose purpose is to enable
the main host bridge to support multiple PCI root buses
for pc machines.
As oposed to PCI-2-PCI bridge's secondary bus, PXB's bus
is a primary bus and can be associated with a NUMA node
(different from the main host bridge) allowing the guest OS
to recognize the proximity of a pass-through device to
other resources as RAM and CPUs.
The PXB is composed from:
- A primary PCI bus (can be associated with a NUMA node)
Acts like a normal pci bus and from the functionality point
of view is an "expansion" of the bus behind the
main host bridge.
- A pci-2-pci bridge behind the primary PCI bus where the actual
devices will be attached.
- A host-bridge PCI device
Situated on the bus behind the main host bridge, allows
the BIOS to configure the bus number and IO/mem resources.
It does not have its own config/data register for configuration
cycles, this being handled by the main host bridge.
- A host-bridge sysbus to comply with QEMU current design.
hw/acpi: add support for i440fx 'snooping' root busses
If the machine has extra root busses that are snooping to
the i440fx host bridge, we need to add them to
acpi in order to be properly detected by guests.
Gerd Hoffmann [Wed, 3 Jun 2015 12:47:19 +0000 (14:47 +0200)]
virtio: 64bit features fixups.
Commit "019a3ed virtio: make features 64bit wide" missed a few changes,
as I've noticed while trying to rebase the virtio-1 branch to latest
master. This patch adds them.
Jason Wang [Wed, 27 May 2015 08:26:07 +0000 (16:26 +0800)]
virtio-pci: don't try to mask or unmask vqs without notifiers
We should validate the vq index against nvqs_with_notifiers. Otherwise we may
try to mask or unmask vector for vqs without notifiers (e.g control vq). This
will lead qemu abort on kvm_irqchip_commit_routes() when trying to boot win8.1
guest.
In DSDT FDC0 declares the IO region as IO(Decode16, 0x03F2, 0x03F2, 0x00, 0x04).
Use the same in lpc_ich9 initialization code.
Now the floppy drive is detected correctly on Windows.
Peter Maydell [Mon, 1 Jun 2015 14:22:46 +0000 (15:22 +0100)]
Merge remote-tracking branch 'remotes/mst/tags/for_upstream' into staging
pc, pci, tpm, virtio, vhost enhancements and fixes
A bunch of cleanups and fixes all over the place,
enhancements in TPM, virtio and vhost.
Signed-off-by: Michael S. Tsirkin <[email protected]>
# gpg: Signature made Mon Jun 1 13:19:48 2015 BST using RSA key ID D28D5469
# gpg: Good signature from "Michael S. Tsirkin <[email protected]>"
# gpg: aka "Michael S. Tsirkin <[email protected]>"
* remotes/mst/tags/for_upstream: (60 commits)
vhost-user: add multi queue support
virtio: make features 64bit wide
qdev: add 64bit properties
virtio-mmio: ioeventfd support
hw/acpi/aml-build: Fix memory leak
acpi: add aml_while() term
acpi: add aml_increment() term
acpi: add aml_shiftright() term
acpi: add aml_shiftleft() term
acpi: add aml_index() term
acpi: add aml_lless() term
acpi: add aml_add() term
TPM2 ACPI table support
tpm: Probe for connected TPM 1.2 or TPM 2
Extend TPM TIS interface to support TPM 2
Add stream ID to MSI write
acpi: Simplify printing to dynamic string
i386: drop FDC in pc-q35-2.4+ if neither it nor floppy drives are wanted
i386/pc_q35: don't insist on board FDC if there's no default floppy
i386/pc: '-drive if=floppy' should imply a board-default FDC
...
Ouyang Changchun [Thu, 28 May 2015 01:23:06 +0000 (09:23 +0800)]
vhost-user: add multi queue support
Based on patch by Nikolay Nikolaev:
Vhost-user will implement the multi queue support in a similar way
to what vhost already has - a separate thread for each queue.
To enable the multi queue functionality - a new command line parameter
"queues" is introduced for the vhost-user netdev.
Gerd Hoffmann [Mon, 1 Jun 2015 08:45:40 +0000 (10:45 +0200)]
virtio: make features 64bit wide
Make features 64bit wide everywhere.
On migration a full 64bit guest_features field is sent if one of the
high bits is set, in addition to the lower 32bit guest_features field
which must stay for compatibility reasons. That way we send the lower
32 feature bits twice, but the code is simpler because we don't have
to split and compose the 64bit features into two 32bit fields.
Stefan Berger [Tue, 26 May 2015 20:51:07 +0000 (16:51 -0400)]
TPM2 ACPI table support
Add a TPM2 ACPI table if a TPM 2 is used in the backend.
Also add an SSDT for the TPM 2.
Rename tpm_find() to tpm_get_version() and have this function
return the version of the TPM found, TPMVersion_Unspec if
no TPM is found. Use the version number to build version
specific ACPI tables.
Peter Maydell [Mon, 1 Jun 2015 10:29:37 +0000 (11:29 +0100)]
Merge remote-tracking branch 'remotes/bkoppelmann/tags/pull-tricore-20150530' into staging
TriCore bugfixes
# gpg: Signature made Sat May 30 15:50:49 2015 BST using RSA key ID 6B69CA14
# gpg: Good signature from "Bastian Koppelmann <[email protected]>"
* remotes/bkoppelmann/tags/pull-tricore-20150530:
target-tricore: fix BOL_ST_H_LONGOFF using ld
target-tricore: fix msub32_q producing the wrong overflow bit
target-tricore: fix OPC2_32_RR_DVINIT_HU having write before use on the result
Stefan Berger [Tue, 26 May 2015 20:51:06 +0000 (16:51 -0400)]
tpm: Probe for connected TPM 1.2 or TPM 2
In the TPM passthrough backend driver, modify the probing code so
that we can check whether a TPM 1.2 or TPM 2 is being used
and adapt the behavior of the TPM TIS accordingly.
Move the code that tested for a TPM 1.2 into tpm_utils.c
and extend it with test for probing for TPM 2. Have the
function return the version of TPM found.
Stefan Berger [Tue, 26 May 2015 20:51:05 +0000 (16:51 -0400)]
Extend TPM TIS interface to support TPM 2
Following the recent upgrade to version 1.3, extend the TPM TIS
interface with capabilities introduced for support of a TPM 2.
TPM TIS for TPM 2 introduced the following extensions beyond the
TPM TIS 1.3 (used for TPM 1.2):
- A new 32bit interface Id register was introduced.
- New flags for the status (STS) register were defined.
- New flags for the capability flags were defined.
Support the above if a TPM TIS 1.3 for TPM 2 is used with a TPM 2
on the backend side. Support the old TPM TIS 1.3 configuration if a
TPM 1.2 is being used. A subsequent patch will then determine which
TPM version is being used in the backend.
Pavel Fedin [Wed, 27 May 2015 12:59:59 +0000 (15:59 +0300)]
Add stream ID to MSI write
GICv3 ITS distinguishes between devices by using hardwired device IDs passed on the bus.
This patch implements passing these IDs in qemu.
SMMU is also known to use stream IDs, therefore this addition can also be useful for
implementing platforms with SMMU.
Signed-off-by: Pavel Fedin <[email protected]> Reviewed-by: Michael S. Tsirkin <[email protected]>
Changes from v1:
- Added bus number to the stream ID
- Added stream ID not only to MSI-X, but also to plain MSI. Some common code was made into
msi_send_message() function. Reviewed-by: Michael S. Tsirkin <[email protected]> Signed-off-by: Michael S. Tsirkin <[email protected]>
build_append_namestringv() and aml_string() first calculate the
resulting string's length with vsnprintf(NULL, ...), then allocate,
then print for real. Simply use g_strdup_vprintf() or g_vasprintf()
instead.
Laszlo Ersek [Thu, 28 May 2015 20:04:11 +0000 (22:04 +0200)]
i386: drop FDC in pc-q35-2.4+ if neither it nor floppy drives are wanted
It is Very annoying to carry forward an outdatEd coNtroller with a mOdern
Machine type.
Hence, let us not instantiate the FDC when all of the following apply:
- the machine type is pc-q35-2.4 or later,
- "-device isa-fdc" is not passed on the command line (nor in the config
file),
- no "-drive if=floppy,..." is requested.
Laszlo Ersek [Thu, 28 May 2015 20:04:10 +0000 (22:04 +0200)]
i386/pc_q35: don't insist on board FDC if there's no default floppy
The "no_floppy = 1" machine class setting causes "default_floppy" in
main() to become zero. Consequently, default_drive() will not call
drive_add() and drive_new() for IF_FLOPPY, index=0, meaning that no
default floppy drive will be created for the virtual machine. In that
case, board code should also not insist on the creation of the
board-default FDC.
The board-default FDC will still be created if the user requests a floppy
drive with "-drive if=floppy".
Additionally, separate FDCs can be specified manually with "-device
isa-fdc". They allow the
-device isa-fdc,driveA=...
syntax that is more flexible than the one required by the board-default
FDC:
-global isa-fdc.driveA=...
This patch doesn't change the behavior observably, as all Q35 machine
types have "no_floppy = 0".
Laszlo Ersek [Thu, 28 May 2015 20:04:09 +0000 (22:04 +0200)]
i386/pc: '-drive if=floppy' should imply a board-default FDC
Even if board code decides not to request the creation of the FDC (keyed
off board-level factors, to be determined later), we should create the FDC
nevertheless if the user passes '-drive if=floppy' on the command line.
Otherwise '-drive if=floppy' would break without explicit '-device
isa-fdc' on such boards.
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.
projects / qemu.git / log