History | View | Annotate | Download (98.9 kB)
vfio: cleanup includes
Starting to get messy, put the back in alphabetical order.
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
vfio: Add bootindex support
vfio-pci: Move devices to D0 on reset
Guests may leave devices in a low power state at reboot, but we expectdevices to be woken up for the next boot. Make this happen.
vfio-pci: Add extra debugging
Often when debugging it's useful to be able to disable bypass pathsso no interactions with the device are missed. Add some extra debugoptions to do this. Also add device info on read/write BAR accesses,which is useful when debugging more than one assigned device. A...
qemu vfio-pci: Graphics device quirks
Graphics cards have a number of different backdoors. Some of theseare alternative ways to get PCI BAR addresses, some of them arecomplete mirrors of PCI config space available through MMIO andI/O port access. These quirks cover a number of ATI Radeon and...
qemu vfio-pci: Add support for VGA MMIO and I/O port access
Most VGA cards need some kind of quirk to fully operate since theyhide backdoors to get to other registers outside of PCI config spacewithin the registers, but this provides the base infrastructure. If...
vfio-pci: Generalize PCI config mangling
Kernel-side vfio virtualizes all of config space, but some parts areunique to Qemu. For instance we may or may not expose the ROM BAR,Qemu manages MSI/MSIX, and Qemu manages the multi-function bit so thatsingle function devices can appear as multi-function and vica versa....
vfio-pci: Add PCIe capability mangling based on bus type
Windows seems to pay particular interest to the PCIe header type ofdevices and will fail to load drivers if we attach Endpoint devices orLegacy Endpoint devices to the Root Complex. We can usepci_bus_is_express and pci_bus_is_root to determine the bus type and...
hw: include hw header files with full paths
Done with this script:
cd hwfor i in `find . -name '*.h' | sed 's/^..//'`; do echo '\,^#.*include.*["<]'$i'[">], s,'$i',hw/&,'done | sed -i -f - `find . -type f`
This is so that paths remain valid as files are moved....
error: Strip trailing '\n' from error string arguments (again)
Commit 6daf194d and be62a2eb got rid of a bunch, but they keep comingback. Tracked down with this Coccinelle semantic patch:
r expression err, eno, cls, fmt; position p; @@ (...
r
error: Clean up abuse of error_report() for help
Use error_printf() instead, so the help gets presented more nicely.
Signed-off-by: Markus Armbruster <armbru@redhat.com>Message-id: 1360354939-10994-3-git-send-email-armbru@redhat.comSigned-off-by: Anthony Liguori <aliguori@us.ibm.com>
vfio-pci: Enable PCIe extended config space
We don't know pre-init time whether the device we're exposing is PCIeor legacy PCI. We could ask for it to be specified via a deviceoption, but that seems like too much to ask of the user. Instead wecan assume everything will be PCIe, which makes PCI-core allocate...
vfio-pci: Loosen sanity checks to allow future features
VFIO_PCI_NUM_REGIONS and VFIO_PCI_NUM_IRQS should never have beenused in this manner as it locks a specific kernel implementation.Future features may introduce new regions or interrupt entries(VGA may add legacy ranges, AER might add an IRQ for error...
vfio-pci: Make host MSI-X enable track guest
Guests typically enable MSI-X with all of the vectors in the MSI-Xvector table masked. Only when the vector is enabled does the vectorget unmasked, resulting in a vector_use callback. These two points,enable and unmask, correspond to pci_enable_msix() and request_irq()...
msi: add API to get notified about pending bit poll
Update all users.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
softmmu: move include files to include/sysemu/
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
misc: move include files to include/qemu/
exec: move include files to include/exec/
Merge commit '1dd3a74d2ee2d873cde0b390b536e45420b3fe05' into HEAD
pci: update all users to look in pci/
update all users so we can remove the makefile hack.
vfio-pci: Don't use kvm_irqchip_in_kernel
kvm_irqchip_in_kernel() has an architecture specific meaning, sowe shouldn't be using it to determine whether to enabled KVM INTxbypass. kvm_irqfds_enabled() seems most appropriate. Also use thisto protect our other call to kvm_check_extension() as that explodes...
vfio-pci: Add KVM INTx acceleration
This makes use of the new level irqfd support enabling bypass of qemuuserspace both on INTx injection and unmask. This significantlyboosts the performance of devices making use of legacy interrupts (ex.~60% better netperf TCP_RR scores for an e1000e assigned to a Linux...
vfio-pci: Use common msi_get_message
We can get rid of our local version now that a helper exists.
Rename target_phys_addr_t to hwaddr
target_phys_addr_t is unwieldly, violates the C standard (_t suffixes arereserved) and its purpose doesn't match the name (most target_phys_addr_taddresses are not target specific). Replace it with a finger-friendly,...
Merge remote-tracking branch 'awilliam/tags/vfio-pci-for-qemu-20121017.0' into staging
memory: use AddressSpace for MemoryListener filtering
Using the AddressSpace type reduces confusion, as you can't accidentallysupply the MemoryRegion you're interested in.
Reviewed-by: Anthony Liguori <aliguori@us.ibm.com>Signed-off-by: Avi Kivity <avi@redhat.com>
vfio-pci: Mark non-migratable
We haven't magically fixed this yet. Toss in a description too.
vfio-pci: Fix debug build
Stray variable from before MSI-X rework
vfio: drop no-op MemoryListener callbacks
Removes quite a bit of useless code.
Signed-off-by: Avi Kivity <avi@redhat.com>
vfio-pci: Clang cleanup
Blue Swirl reports that Clang doesn't like the structure we define toavoid dynamic allocation for a number of calls to VFIO_DEVICE_SET_IRQS.Adding an element after a variable sized type is a GNU extension.Switch back to dynamic allocation, which really isn't a problem since...
vfio-pci: Fix BAR->VFIODevice translation in
DO_UPCAST is supposed to translate from the first member of a struct tothat struct, not from arbitrary ones. And it (usually) breaks the buildwhen neglecting this rule. Use container_of to fix the build breakage...
vfio-pci: Roll the header into the .c file
It's only ~100 lines and nobody else should be using this.Suggested by Michael Tsirkin.
vfio-pci: Don't peak at msi_supported
Let the init function fail, just don't warn for -ENOTSUP.
vfio-pci: Use uintptr_t for void* cast
We don't seem to run into any sign extension problems, butunsigned looks more correct.
Signed-off-by: Alex williamson <alex.williamson@redhat.com>
vfio-pci: Remove setting of MSI qsize
This was a misinterpretation of the spec, hardware doesn't get tospecify how many were actually enabled through this field.
vfio-pci: Extend reset
Take what we've learned from pci-assign and apply it to vfio-pci.On reset, disable previous interrupt config, perform a devicereset if available, re-enable INTx, and disable memory regions onthe device to prevent continuing DMA....
vfio-pci: Cleanup on INTx setup failure
Missing some unwind code.
vfio-pci: Update slow path INTx algorithm
We can't afford the overhead of switching out and back into mmap modearound each interrupt, but we can do it lazily via a timer. On INTxinterrupt, disable the mmap'd memory regions and set a timer. Onevery interrupt, push the timer out. If the timer expires and the...
vfio-pci: Re-order map/unmap
This cleans up the next patch that calls unmap from map.
vfio-pci: Unmap and retry DMA mapping
Occasionally we get regions added that overlap with existing mappings.These always seems to be in the VGA ROM range. VFIO returns EBUSYfor these mapping attempts. We can try a little harder and assumethat the latest mapping is correct by removing any overlapping ranges...
vfio-pci: Rework MSIX setup/teardown
We try to do lazy initialization of MSIX since we don't actually needto setup anything until MSIX vectors start getting used. This leadsto problems if MSIX is enabled, but never used (we can end up tryingto re-enable INTx while it's still enabled). We also run into...
vfio-pci: No spurious MSIs
FreeBSD doesn't like these spurious MSIs, remove them as they'remostly paranoia anyway.
vfio_pci: fix build on 32-bit systems
We cannot cast directly from pointer to uint64.
Cc: Alex Williamson <alex.williamson@redhat.com>Cc: Alex Barcelo <abarcelo@ac.upc.edu>Reported-by: Alex Barcelo <abarcelo@ac.upc.edu>Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
vfio: vfio-pci device assignment driver
This adds the core of the QEMU VFIO-based PCI device assignment driver.To make use of this driver, enable CONFIG_VFIO, CONFIG_VFIO_IOMMU_TYPE1,and CONFIG_VFIO_PCI in your host Linux kernel config. Load the vfio-pci...