11 Feb, 2015
21 commits
-
For each device, We need to include the vendor capabilities to demark
where virtio common, notification and ISR regions are (we put them
all in BAR0).We need to handle the switching of the virtqueues using the accessors.
Signed-off-by: Rusty Russell
-
This handles ioport 0xCF8 and 0xCFC accesses, which are used to
read/write PCI device config space.Signed-off-by: Rusty Russell
-
We don't do anything with them yet (emulate_mmio_write and
emulate_mmio_read are stubs), but we decode the instructions and
search for the device they're hitting.Signed-off-by: Rusty Russell
-
This is where we point our PCI BARs, so that we can intercept MMIO
accesses. We tell the kernel about it so any faults in this area are
directed to us.Signed-off-by: Rusty Russell
-
This lets us deliver interrupts for our emulated PCI devices using our
dumb PIC, and not emulate an 8259 and PCI irq mapping tables or whatever.Signed-off-by: Rusty Russell
-
Once we add PCI, it starts trying to manage our interrupts.
Signed-off-by: Rusty Russell
-
This lets us implement PCI.
Signed-off-by: Rusty Russell
-
This no longer speeds up boot (IDE got better, I guess), but it does stop
us probing for a PCI bus.Signed-off-by: Rusty Russell
-
While hacking on getting I/O out to the lguest launcher, I noticed
that returning 0xFF for the PS/2 keyboard status made it spin for a
while thinking there was a key pending. Fix this by returning 1
instead of 0xFF.Signed-off-by: Rusty Russell
-
We copy 7 bytes at eip for userspace's instruction decode; we have to
carefully handle the case where eip is at the end of a page. We can't
leave this to userspace since kernel has all the page table decode
logic.The decode logic moves to userspace, basically unchanged.
Signed-off-by: Rusty Russell
-
We normally abort the guest unconditionally when it gives us a bad address,
but in the next patch we want to copy some bytes which may not be mapped.Signed-off-by: Rusty Russell
-
This is required for instruction emulation to move to userspace.
Signed-off-by: Rusty Russell
-
This is preparation for userspace handling MMIO and ioport accesses.
Signed-off-by: Rusty Russell
-
We use the ptrace API struct, and we currently don't let them set
anything but the normal registers (we'd have to filter the others).Signed-off-by: Rusty Russell
-
Theoretical debates aside, now it boots.
Signed-off-by: Rusty Russell
-
Strictly, it's only needed when we have features (size or multiport).
Signed-off-by: Rusty Russell
-
Since PCI is little endian, 8-bit access might work, but the spec section
is very clear on this:4.1.3.1 Driver Requirements: PCI Device Layout
The driver MUST access each field using the “natural” access method,
i.e. 32-bit accesses for 32-bit fields, 16-bit accesses for 16-bit
fields and 8-bit accesses for 8-bit fields.Signed-off-by: Rusty Russell
Acked-by: Michael S. Tsirkin -
The VIRTIO_F_ANY_LAYOUT and VIRTIO_F_NOTIFY_ON_EMPTY features are pre-1.0
only.Signed-off-by: Rusty Russell
Acked-by: Michael S. Tsirkin -
This allows modern implementations to ensure they don't use legacy
feature bits or SCSI commands (which are not used in v1.0 non-legacy).Signed-off-by: Rusty Russell
Acked-by: Michael S. Tsirkin -
This provides backdoor access to the device MMIOs, and every device should
have one. From the virtio 1.0 spec (CS03):4.1.4.7.1 Device Requirements: PCI configuration access capability
The device MUST present at least one VIRTIO_PCI_CAP_PCI_CFG capability.
Signed-off-by: Rusty Russell
Acked-by: Michael S. Tsirkin -
The virtqueue_add() calls START_USE() upon entry. The virtqueue_kick() is
called if vq->num_added == (1 << 16) - 1 before calling END_USE().
The virtqueue_kick_prepare() called via virtqueue_kick() calls START_USE()
upon entry, and will call panic() if DEBUG is enabled.
Move this virtqueue_kick() call to after END_USE() call.Signed-off-by: Tetsuo Handa
Signed-off-by: Rusty Russell
23 Jan, 2015
1 commit
-
This patch add a support for second version of the virtio-mmio device,
which follows OASIS "Virtual I/O Device (VIRTIO) Version 1.0"
specification.Main changes:
1. The control register symbolic names use the new device/driver
nomenclature rather than the old guest/host one.2. The driver detect the device version (version 1 is the pre-OASIS
spec, version 2 is compatible with fist revision of the OASIS spec)
and drives the device accordingly.3. New version uses direct addressing (64 bit address split into two
low/high register) instead of the guest page size based one,
and addresses each part of the queue (descriptors, available, used)
separately.4. The device activity is now explicitly triggered by writing to the
"queue ready" register.5. Whole 64 bit features are properly handled now (both ways).
Signed-off-by: Pawel Moll
Acked-by: Michael S. Tsirkin
Signed-off-by: Rusty Russell
21 Jan, 2015
18 commits
-
release function in modern driver is unused:
it's a left-over from when each driver had
to have its own release.Signed-off-by: Michael S. Tsirkin
-
If set, try legacy interface first, modern one if that fails. Useful to
work around device/driver bugs, and for compatibility testing.Signed-off-by: Michael S. Tsirkin
Signed-off-by: Rusty Russell -
Useful for testing device virtio 1 compatibility.
Based on patch by Rusty - couldn't resist putting
that flying car joke in there!Signed-off-by: Michael S. Tsirkin
Signed-off-by: Rusty Russell -
The ABI *is* stable, and has been for a while now.
Drop Kconfig warning saying that it's not guaranteed
to work.Signed-off-by: Michael S. Tsirkin
Signed-off-by: Rusty Russell -
This drivers -> this driver.
Signed-off-by: Michael S. Tsirkin
Signed-off-by: Rusty Russell -
makes code look a bit prettier.
Signed-off-by: Michael S. Tsirkin
Signed-off-by: Rusty Russell -
Most of our code has
struct foo {
}Fix one instances where ring is inconsistent.
Signed-off-by: Michael S. Tsirkin
Signed-off-by: Rusty Russell -
Most of our code has
struct foo {
}Fix two instances where blk is inconsistent.
Signed-off-by: Michael S. Tsirkin
Signed-off-by: Rusty Russell -
Most of our code has
struct foo {
}Fix two instances where balloon is inconsistent.
Signed-off-by: Michael S. Tsirkin
Signed-off-by: Rusty Russell -
Virtio 1.0 spec lists device config as optional.
Set get/set callbacks to NULL. Drivers can check that
and fail gracefully.Signed-off-by: Michael S. Tsirkin
Signed-off-by: Rusty Russell -
We don't know the # of VQs that drivers are going to use so it's hard to
predict how much memory we'll need to map. However, the relevant
capability does give us an upper limit.
If that's below a page, we can reduce the number of required
mappings by mapping it all once ahead of the time.Signed-off-by: Michael S. Tsirkin
Signed-off-by: Rusty Russell -
QEMU wants it, so why not? Trust, but verify.
Signed-off-by: Rusty Russell
Signed-off-by: Michael S. Tsirkin -
Lightly tested against qemu.
One thing *not* implemented here is separate mappings
for descriptor/avail/used rings. That's nice to have,
will be done later after we have core support.This also exposes the PCI layout to userspace, and
adds macros for PCI layout offsets:QEMU wants it, so why not? Trust, but verify.
Signed-off-by: Rusty Russell
Signed-off-by: Michael S. Tsirkin -
Based on patches by Michael S. Tsirkin , but I found it
hard to follow so changed to use structures which are more
self-documenting.Signed-off-by: Rusty Russell
Signed-off-by: Michael S. Tsirkin -
Most of initialization is device-independent.
Let's move it to common.Signed-off-by: Michael S. Tsirkin
Signed-off-by: Rusty Russell -
Device VQs were getting freed twice: once in every device's removal
functions, and then again in virtio_pci_legacy_remove(). The ones in
devices are called first, so drop the useless second call.Signed-off-by: Sasha Levin
Signed-off-by: Michael S. Tsirkin
Signed-off-by: Rusty Russell -
Virtio drivers should map the part of the range they need, not
necessarily all of it.
To this end, support mapping ranges within BAR on s390.
Since multiple ranges can now be mapped within a BAR, we keep track of
the number of mappings created, and only clear out the mapping for a BAR
when this number reaches 0.Cc: Bjorn Helgaas
Cc: linux-pci@vger.kernel.org
Tested-by: Sebastian Ott
Signed-off-by: Michael S. Tsirkin
Signed-off-by: Rusty Russell -
Virtio drivers should map the part of the BAR they need, not necessarily
all of it.Cc: Bjorn Helgaas
Cc: linux-pci@vger.kernel.org
Acked-by: Arnd Bergmann
Signed-off-by: Michael S. Tsirkin
Signed-off-by: Rusty Russell
