This per-device notifier shall be triggered by any interrupt router
along the path of a device's legacy interrupt signal on routing changes.
For simplicity reasons and as this is a slow path anyway, no further
details on the routing changes are provided. Instead, the callback is
expected to use pci_device_route_intx_to_irq to check the effect of the
change.
Will be used by KVM PCI device assignment and VFIO.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Device assigned on KVM needs to know the mode
(enabled/inverted/disabled) and the IRQ number that a given device
triggers in the attached interrupt controller.
Add a PCI IRQ path discovery function that walks from a given device to
the host bridge, and gets this information. For
this purpose, a host bridge callback function is introduced:
route_intx_to_irq. It is so far only implemented by the PIIX3, other
host bridges can be added later on as required.
Will be used for KVM PCI device assignment and VFIO.
Based on patch by Jan Kiszka, with minor tweaks.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
BARs are registered in init functions from memory regions created
by the drivers. Exit functions destroy those memory regions.
By unregistering the io regions after exit(), we're calling
memory_region_del_subregion on freed memory. Don't do that. The
option rom comes along for the ride because it's more symmetric
to how it's created.
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Not a single driver has any possibility of failure on their
exit function, let's keep it that way.
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Finally, complete the fully specified interface. msix_add_config()
gets folded into msix_init() because we now have quite a few parameters
to pass and rolling it in let's us error earlier, avoiding the ugly
unwind exit path. msix_mmio_setup() also gets rolled in, just because
it's redundant to rediscover offsets when we already have them for
such a tiny function.
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
These don't have to be contiguous. Size them to only what
they need and use separate MemoryRegions for the vector
table and PBA.
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
MSIX, like PCI, is little endian. Specifying native is wrong here,
but we need to check the rest of the file to determine if it's
as simple as flipping this macro.
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
What's this doing so far from msix_mmio_ops?
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Trivial conversion, failed to have an uninit before and after.
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
msi_init() takes over a BAR without really specifying or allowing
specification of how it does so. Instead, let's split it into
two interfaces, one fully specified, and one trivially easy. This
implements the latter. msix_init_exclusive_bar() takes over
allocating and filling a PCI BAR _exclusively_ for the use of MSIX.
When used, the matching msi_uninit_exclusive_bar() should be used
to tear it down.
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
msix.h calls the PCIDevice * parameter "dev" almost everywhere except
the msix_write_config declaration. Fix the inconsistency.
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
No user in sight for msix_bar_size.
bar_size for all users is aligned, let's simply
require this instead of trying to fix up invalid input.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Currently, we do not properly cleanup, if pci_bridge_dev_initfn
fails to initialize properly. Make sure to call pci_bridge_exitfn()
in the error path.
Signed-off-by: Jason Baron <jbaron@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
While looking into hot-plugging bridges, I can create a qemu segfault via:
$ device_add pci-bridge
Bridge chassis not specified. Each bridge is required to be assigned a unique chassis id > 0.
**
ERROR:qom/object.c:389:object_delete: assertion failed: (obj->ref == 0)
I'm proposing to fix this by adding a call to 'object_unparent()', before the
call to qdev_free(). I see there is already a precedent for this usage pattern as
seen in qdev_simple_unplug_cb():
/* can be used as ->unplug() callback for the simple cases */
int qdev_simple_unplug_cb(DeviceState *dev)
{
/* just zap it */
object_unparent(OBJECT(dev));
qdev_free(dev);
return 0;
}
Signed-off-by: Jason Baron <jbaron@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Also this functions is better invoked by the core than by each and every
device. This allows to drop the config_write callbacks from ich and
intel-hda.
CC: Alexander Graf <agraf@suse.de>
CC: Gerd Hoffmann <kraxel@redhat.com>
CC: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Terminate msi/msix_write_config early if support is not enabled. This
allows to remove checks at the caller site if MSI is optional.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
There is no point in pushing this burden to the devices, they tend to
forget to call them (like intel-hda, ahci, xhci did). Instead, reset
functions are now called from pci_device_reset. They do nothing if
MSI/MSI-X is not in use.
CC: Alexander Graf <agraf@suse.de>
CC: Gerd Hoffmann <kraxel@redhat.com>
CC: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Properly register reset functions via the device class.
CC: Alexander Graf <agraf@suse.de>
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Call msi_reset on device reset as still required by the core.
CC: Gerd Hoffmann <kraxel@redhat.com>
CC: qemu-stable@nongnu.org
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Call msi_reset on device reset as still required by the core.
CC: Alexander Graf <agraf@suse.de>
CC: qemu-stable@nongnu.org
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Some drivers (Linux' 8139too among them) rely on the NIC
injecting an interrupt in the event of a receive buffer overflow
and, accordingly, set the RxOverflow bit in the interrupt
mask. Unfortunately rtl8139's can_receive method ignores the
RxOverflow flag, which may lead to a situation where rtl8139
stops receiving packets (can_receive returns 0) when the receive
buffer becomes full.
If the driver eventually read from the receive buffer or reset
the card the emulator could recover from this situation. However
some implementations only do this upon receiving an interrupt
with either RxOK or RxOverflow set in the ISR; interrupt that
will never come because QEMU's flow control mechanisms would
prevent rtl8139 from receiving any packet.
Letting packets go through when the overflow interrupt is enabled
makes the QEMU emulator compliant to the spec and solves the
problem.
This patch should fix a relatively common (in our experience)
network stall observed when running enterprise distros with
rtl8139 as the NIC; in some cases the 8139too device driver gets
loaded and when under heavy load the network eventually stops
working.
Reported-by: Hayato Kakuta <kakuta.hayato@oss.ntt.co.jp>
Tested-by: Hayato Kakuta <kakuta.hayato@oss.ntt.co.jp>
Acked-by: Igor Kovalenko <igor.v.kovalenko@gmail.com>
Signed-off-by: Fernando Luis Vazquez Cao <fernando@oss.ntt.co.jp>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
This reverts commit ff71f2e8ca. This is because
the linux 8139cp driver would leave the card in "Config Register Write Enable"
mode after the eeprom were read or write ( which is unexpected in the spec
). Also a physical 8139 card can still DMA into host memory in modes other than
Normal mode, so we need revert this commit to align with the behavior of
physical card.
The issue of 8139cp driver should be fixed in linux seperately.
Signed-off-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
* qemu-kvm/uq/master:
virtio/vhost: Add support for KVM in-kernel MSI injection
msix: Add msix_nr_vectors_allocated
kvm: Enable use of kvm_irqchip_in_kernel in hwlib code
kvm: Introduce kvm_irqchip_add/remove_irqfd
kvm: Make kvm_irqchip_commit_routes an internal service
kvm: Publicize kvm_irqchip_release_virq
kvm: Introduce kvm_irqchip_add_msi_route
kvm: Rename kvm_irqchip_add_route to kvm_irqchip_add_irq_route
msix: Introduce vector notifiers
msix: Invoke msix_handle_mask_update on msix_mask_all
msix: Factor out msix_get_message
kvm: update vmxcap for EPT A/D, INVPCID, RDRAND, VMFUNC
kvm: Enable in-kernel irqchip support by default
kvm: Add support for direct MSI injections
kvm: Update kernel headers
kvm: x86: Wire up MSI support for in-kernel irqchip
pc: Enable MSI support at APIC level
kvm: Introduce basic MSI support for in-kernel irqchips
Introduce MSIMessage structure
kvm: Refactor KVMState::max_gsi to gsi_count
* kwolf/for-anthony:
ahci: SATA FIS is 20 bytes, not 0x20
virtio-blk: Fix geometry sector calculation
block: prevent snapshot mode $TMPDIR symlink attack
sheepdog: fix return value of do_load_save_vm_state
virtio: Fix compiler warning for non Linux hosts
As in the SATA and AHCI specifications, a FIS is 5 Dwords of 4 bytes
each, which comes to 20 bytes (decimal), not 0x20.
Signed-off-by: Daniel Verkamp <daniel@drv.nu>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Currently the sector value for the geometry is masked, even if the
user usesa command line parameter that explicitely gives a number.
This breaks dasd devices on s390. A dasd device can have
a physical block size of 4096 (== same for logical block size)
and a typcial geometry of 15 heads and 12 sectors per cyl.
The ibm partition detection relies on a correct geometry
reported by the device. Unfortunately the current code changes
12 to 8. This would be necessary if the total size is
not a multiple of logical sector size, but for dasd this
is not the case.
This patch checks the device size and only applies sector
mask if necessary.
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
CC: Christoph Hellwig <hch@lst.de>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
In snapshot mode, bdrv_open creates an empty temporary file without
checking for mkstemp or close failure, and ignoring the possibility
of a buffer overrun given a surprisingly long $TMPDIR.
Change the get_tmp_filename function to return int (not void),
so that it can inform its two callers of those failures.
Also avoid the risk of buffer overrun and do not ignore mkstemp
or close failure.
Update both callers (in block.c and vvfat.c) to propagate
temp-file-creation failure to their callers.
get_tmp_filename creates and closes an empty file, while its
callers later open that presumed-existing file with O_CREAT.
The problem was that a malicious user could provoke mkstemp failure
and race to create a symlink with the selected temporary file name,
thus causing the qemu process (usually root owned) to open through
the symlink, overwriting an attacker-chosen file.
This addresses CVE-2012-2652.
http://bugzilla.redhat.com/CVE-2012-2652
Signed-off-by: Jim Meyering <meyering@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
bdrv_save_vmstate and bdrv_load_vmstate should return the vmstate size
on success, and -errno on error.
Signed-off-by: MORITA Kazutaka <morita.kazutaka@lab.ntt.co.jp>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
The local variables ret, i are only used if __linux__ is defined.
Signed-off-by: Stefan Weil <sw@weilnetz.de>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
In snapshot mode, bdrv_open creates an empty temporary file without
checking for mkstemp or close failure, and ignoring the possibility
of a buffer overrun given a surprisingly long $TMPDIR.
Change the get_tmp_filename function to return int (not void),
so that it can inform its two callers of those failures.
Also avoid the risk of buffer overrun and do not ignore mkstemp
or close failure.
Update both callers (in block.c and vvfat.c) to propagate
temp-file-creation failure to their callers.
get_tmp_filename creates and closes an empty file, while its
callers later open that presumed-existing file with O_CREAT.
The problem was that a malicious user could provoke mkstemp failure
and race to create a symlink with the selected temporary file name,
thus causing the qemu process (usually root owned) to open through
the symlink, overwriting an attacker-chosen file.
This addresses CVE-2012-2652.
http://bugzilla.redhat.com/CVE-2012-2652
Reviewed-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
Signed-off-by: Jim Meyering <meyering@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
NULL pointer dereference in case no vnc server is configured.
Catch this and return -EINVAL like vnc_display_password() does.
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Changes v2 -> v3;
- Check for kvm_enabled() before setting cpuid_7_0_ebx_features
Changes v1 -> v2:
- Use kvm_arch_get_supported_cpuid() instead of host_cpuid() on
cpu_x86_fill_host().
We should use GET_SUPPORTED_CPUID for all bits on "-cpu host"
eventually, but I am not changing all the other CPUID leaves because
we may not be able to test such an intrusive change in time for 1.1.
Description of the bug:
Since QEMU 0.15, the CPUID information on CPUID[EAX=7,ECX=0] is being
returned unfiltered to the guest, directly from the GET_SUPPORTED_CPUID
return value.
The problem is that this makes the resulting CPU feature flags
unpredictable and dependent on the host CPU and kernel version. This
breaks live-migration badly if migrating from a host CPU that supports
some features on that CPUID leaf (running a recent kernel) to a kernel
or host CPU that doesn't support it.
Migration also is incorrect (the virtual CPU changes under the guest's
feet) if you migrate in the opposite direction (from an old CPU/kernel
to a new CPU/kernel), but with less serious consequences (guests
normally query CPUID information only once on boot).
Fortunately, the bug affects only users using cpudefs with level >= 7.
The right behavior should be to explicitly enable those features on
[cpudef] config sections or on the "-cpu" command-line arguments. Right
now there is no predefined CPU model on QEMU that has those features:
the latest Intel model we have is Sandy Bridge.
I would like to get this fixed on 1.1, so I am submitting this patch,
that enables those features only if "-cpu host" is being used (as we
don't have any pre-defined CPU model that actually have those features).
After 1.1 is released, we can make those features properly configurable
on [cpudef] and -cpu configuration.
One problem is: with this patch, users with the following setup:
- Running QEMU 1.0;
- Using a cpudef having level >= 7;
- Running a kernel that supports the features on CPUID leaf 7; and
- Running on a CPU that supports some features on CPUID leaf 7
won't be able to live-migrate to QEMU 1.1. But for these users
live-migration is already broken (they can't live-migrate to hosts with
older CPUs or older kernels, already), I don't see how to avoid this
problem.
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Currently we re-read/re-process /etc/mtab to get an updated list of
mounts when guest-fsfreeze-thaw is called. This can cause an atime
update on /etc/mtab, which will block if we're in a frozen state.
Instead, use /proc's version of mtab, which may not be up-to-date with
options passed via -o remount, but is compatible for our use cases since
we only care about the filesystem type.
Reported-by: Matsuda, Daiki <matsudadik@intellilink.co.jp>
Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com>
Use _NSGetEnviron() helper to access the environment.
Signed-off-by: Andreas Färber <andreas.faerber@web.de>
Cc: Charlie Somerville <charlie@charliesomerville.com>
Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com>
Start VM with 8 multiple-function block devs, hot-removing
those block devs by 'device_del ...' would cause qemu abort.
| (qemu) device_del virti0-0-0
| (qemu) **
|ERROR:qom/object.c:389:object_delete: assertion failed: (obj->ref == 0)
It's a regression introduced by commit 57c9fafe
The whole PCI slot should be removed once. Currently only one func
is cleaned in pci_unplug_device(), if you try to remove a single
func by monitor cmd.
free_qdev() are called for all functions in slot,
but unparent_delete() is only called for one
function.
Signed-off-by: XXXX
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
The previous multiboot load code did not treat the case where
load_end_addr was 0 specially. The multiboot specification says the
following:
* load_end_addr
Contains the physical address of the end of the data segment.
(load_end_addr - load_addr) specifies how much data to load. This
implies that the text and data segments must be consecutive in the
OS image; this is true for existing a.out executable formats. If
this field is zero, the boot loader assumes that the text and data
segments occupy the whole OS image file.
Signed-off-by: Scott Moser <smoser@ubuntu.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
With pc-0.12, we map the video RAM both through the PCI BAR (the guest does
this) and through a fixed mapping at 0xe0000000. The memory API doesn't allow
this double map, and aborts.
Fix by using an alias.
Reported-by: Michael Tokarev <mjt@tls.msk.ru>
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
