Due to the way we allow arguments to be specified I was not able to
reuse Config structure for cmdline arguments. (eg we allow multiple ways
to specify same config)
There is still some usage of arguments mod that we need to cleanup
later.
Change-Id: Ia0c12d508d0a10d17f58944de71276e5984d69ee
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3684144
Auto-Submit: Anton Romanov <romanton@google.com>
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Daniel Verkamp <dverkamp@chromium.org>
Commit-Queue: Anton Romanov <romanton@google.com>
Use `cfg_if` to refactor test imports to provide test only definition
for `__cpuid_count`. Otherwise use the definition from
`std::arch::x86_64`.
BUG=b:234508273
TEST=Compiled.
Change-Id: I0d93f6a86522451c540af9d520bfdeffd4f8dbf9
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3701441
Commit-Queue: Vaibhav Nagarnaik <vnagarnaik@google.com>
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Daniel Verkamp <dverkamp@chromium.org>
This change will allow the guest to start a new VVU device backend
after an old VVU device backend process exited due to sibling
disconnection.
More specifically, this CL will allow the vvu-proxy to support scenarios
like follows:
1. A device backend starts in the guest and a sibling VM connects to
the socket.
2. The sibling VM shuts down. Then, the device backend process exists in
the guest. And, the vvu-proxy's state is changed from `Running` to
`Activated`.
3. Return to 1.
Note that we don't support more complicated reconnection scenario such
as:
* reconnection after sibling's unexpected crash
* restart after device backend's unexpected crash
To support the reconnection feature, the vvu-proxy device needs to clean
its status when a sibling disconnected. Specifically, it needs to
* reset virtqueues' state,
* unregister memory regions that it registered via VmMemoryRequest, and
* update the device state from the worker thread.
BUG=b:216407443
TEST=run on workstation; sibling could reconnect after the first
instance exits.
Change-Id: I4c01e6069484ff74a0d643edd6a3b3231fb5c2d6
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3602361
Reviewed-by: Alexandre Courbot <acourbot@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Commit-Queue: Keiichi Watanabe <keiichiw@chromium.org>
Reworks the way we organize parameters to this module, and
adds CPU manufacturer detection.
BUG=213152505
TEST=builds
Change-Id: Ibb96f08cacbdfdaff5ef971c3d6d7b901b5d73fa
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3690369
Reviewed-by: Daniel Verkamp <dverkamp@chromium.org>
Commit-Queue: Noah Gold <nkgold@google.com>
Tested-by: kokoro <noreply+kokoro@google.com>
This APIC is used by the userspace irqchip (to be upstreamed soon).
Original author: Steven Richman <srichman@google.com>
BUG=213152505
TEST=builds
Change-Id: I51b89e96af88f2f1c08998b70053fc37bd3cfcaf
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3696535
Commit-Queue: Noah Gold <nkgold@google.com>
Reviewed-by: Daniel Verkamp <dverkamp@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
This function is used to generate cpuid leaf 15H in some situations.
BUG=213152505
TEST=builds
Change-Id: Ia5a7c46b32f1bdba366a500caa650edb2e3ae99f
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3696534
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Daniel Verkamp <dverkamp@chromium.org>
Commit-Queue: Noah Gold <nkgold@google.com>
This CL will:
* Fix the test in ipc_memory_mapper for Windows
* Conditionally compile out the Vfio error for Windows
On Windows, when one end of a Tube is dropped, the underlying named pipe
connection will be severed. `response_tx` could be dropped due to a race
condition, so this CL fixes that.
BUG=b:234497712
TEST=built and ran test
Change-Id: I4a8b475e117639fd3615cbf9e958aac8af1d00f9
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3696672
Reviewed-by: Noah Gold <nkgold@google.com>
Tested-by: kokoro <noreply+kokoro@google.com>
Commit-Queue: Richard Zhang <rizhang@google.com>
Add `tsc` as a device. Includes operations to calibrate and sync the
TSC values across the physical cores in order to minimize variations
found when moving vcpu threads across cores.
BUG=b:234508273
TEST=Compiled.
Change-Id: I5c79deebc9471dacf9d6596384ee77a8811a58bc
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3682118
Reviewed-by: Daniel Verkamp <dverkamp@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Commit-Queue: Vaibhav Nagarnaik <vnagarnaik@google.com>
There's some ring related wrapping math for interrupt suppression
(aka VIRTIO_F_EVENT_IDX) that had some documentation, but was still a little
tricky to follow. This CL adds some diagrams & further details.
BUG=none
TEST=n/a
Change-Id: I591661a676eb6506fbfea2afa20da28aaa1722f9
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3689972
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Commit-Queue: Noah Gold <nkgold@google.com>
This doesn't split iommu submodules into their own sys directory. That
will be done in subsequent CLs.
BUG=b:213149162
TEST=built and presubmits
Change-Id: I66c4e9b7dec730f259b9f94adac5a39cb0d2ff48
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3692397
Tested-by: kokoro <noreply+kokoro@google.com>
Commit-Queue: Richard Zhang <rizhang@google.com>
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
http://crrev.com/c/3592633 make `run_with_listener` generic, but did not
update the cras vhost-user device to use this new version.
BUG=None
TEST=Chrome OS build succeeds.
Change-Id: I82342e3a2505dc97e5cdaa1749b9091fb986fa35
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3694880
Commit-Queue: Alexandre Courbot <acourbot@chromium.org>
Auto-Submit: Alexandre Courbot <acourbot@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
These fields were reported as never read by the compiler.
BUG=None
TEST=cargo clippy --features "video-decoder,video-encoder,libvda"
Change-Id: I7580fa9161ba8239398bb7a5b547d67c0c2591e0
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3685675
Commit-Queue: Alexandre Courbot <acourbot@chromium.org>
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Since `Self` implements Copy, these methods can and should take their
parameter by value.
BUG=None
TEST=cargo clippy --features "video-decoder,video-encoder,libvda"
Change-Id: I13c4ed627e4cddc1873cf8fa6e3d5c9ac89eccd5
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3685676
Tested-by: kokoro <noreply+kokoro@google.com>
Commit-Queue: Alexandre Courbot <acourbot@chromium.org>
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
The run_with_listener() method can now be made generic and be used for
both the socket and VVU handlers.
BUG=b:229554679
TEST=vhost-user console device works.
TEST=vvu console device works.
Change-Id: Ic1ae286e573383d6b66c8a406227d6cf51cbd578
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3592633
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
Commit-Queue: Alexandre Courbot <acourbot@chromium.org>
The async methods of the socket VVU handler were calling the non-async
accept() on a dedicated thread so the async method does not get blocked
awaiting for a connection. This socket-specific behavior forces us to
have dedicated code for the socket and VVU handlers.
Fix this by adding a poll_descriptor() method to vmm_vhost's Listener
trait that returns a descriptor that callers can poll on if a call to
accept() can block. This way we can wait for the connection
asynchronously and avoid using a thread for that. It also opens the way
towards factorizing the socket and VVU specific code.
BUG=b:229554679
TEST=vhost-user console device works.
Change-Id: I3fbc96d3904f777c6165b04564f577c553ce55d2
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3592632
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Commit-Queue: Alexandre Courbot <acourbot@chromium.org>
The virtio video decoder device is currently only available under very
drastic conditions: a build linked against libvda (a ChromeOS-only
library that needs the cros chroot to be built and linked against), and
a ChromeOS-flavored Chrome instance running alongside crosvm, so the
browser can provide the video decoding service through Mojo.
This makes the decoder device very difficult to develop on for
non-Chromies, and also for Chromies actually since they will always need
a DUT to test it on.
This patch introduces an alternative decoder backend based on
ffmpeg's libraries that performs decoding on the host's CPU. It supports both
guest pages and virtio objects as target, and can be considered a
reliable and predictable way to test the decoder in any environment.
We introduce our own ffmpeg bindings after a quick state of the art
revealed that the existing ones were all unsuitable, either for
technical or licensing reasons. Doing so is also not a big effort and
does not add any new external crate dependency to crosvm.
BUG=b:169295147
TEST=cargo test --features "video-decoder,ffmpeg" -p devices ffmpeg
TEST=v4l2r's simple_decoder example decodes test-25fps.h264 properly with the
following command:
./simple_decoder test-25fps.h264 /dev/video0 --input_format h264 --save test-25fps.nv12
TEST=ARCVM Android youtube plays videos correctly when the ffmpeg
backend is used.
Change-Id: Ic9c586193f7939f2a3fe59d009c3666585a8bbc7
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3026355
Tested-by: kokoro <noreply+kokoro@google.com>
Commit-Queue: Alexandre Courbot <acourbot@chromium.org>
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
This change enables memory mapping of the vfio device doorbell regions so
we can access the vfio bar more efficiently by limiting the scope of the vvu
device interactions with the guest kernel.
BUG=b:231938067
TEST=built and run, verified vvu communication happens as expected
Change-Id: I16925d99a467d44ce7f74316770f1c665de75433
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3670459
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
Auto-Submit: Morg <morg@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Commit-Queue: Morg <morg@chromium.org>
The outer set of parentheses is redundant.
BUG=None
TEST=emerge-kevin crosvm
Change-Id: I482e04601b0f7a3ef22221a77c00ef746bbd288a
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3691969
Reviewed-by: Junichi Uekawa <uekawa@chromium.org>
Commit-Queue: Daniel Verkamp <dverkamp@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
The flexible_array_impl! macro should not rely on the caller to import
FlexibleArray; use $crate::FlexibleArray to refer to it instead.
BUG=None
TEST=tools/presubmit
Change-Id: I9743a4cbc01deb6cdfadd98e63f4803a5572421c
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3687064
Reviewed-by: Alexandre Courbot <acourbot@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Commit-Queue: Daniel Verkamp <dverkamp@chromium.org>
Introduce `State` enum to handle vvu-proxy's lifecycle.
This state will be used to support recovery from sibling disconnection in the follow-up CL.
BUG=b:216407443
TEST=run VVU
Change-Id: I47f9c69abb9d08520ec90a4a9a87fc4d230c96a1
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3591105
Reviewed-by: Alexandre Courbot <acourbot@chromium.org>
Commit-Queue: Keiichi Watanabe <keiichiw@chromium.org>
Reviewed-by: Abhishek Bhardwaj <abhishekbh@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
vhost-user implementations of the net device can currently not
specify a MAC for the guest, even though they would support this.
This adds the corresponding feature to the allowlist and increases
the queue size to 1k to avoid situations where the device runs out
of available descriptors.
BUG=None
TEST=./tools/run_tests
Change-Id: Iefe8ce21215b36c65ee2516428249c41bf13f285
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3683380
Reviewed-by: Noah Gold <nkgold@google.com>
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
Commit-Queue: Kevin Hamacher <hamacher@google.com>
Tested-by: kokoro <noreply+kokoro@google.com>
Pull the declaration of the `EventToken` enum up to the top level of the
base crate, replacing the identical implementations inside sys/windows
and sys/unix.
Use the `EventToken` name consistently throughout the tree to remove the
unix-flavored "poll" nomenclature.
BUG=b:213153157
TEST=tools/dev_container tools/presubmit --all
Change-Id: I0ba42037b533b796797a7a3f6d8d7e71a5592aba
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3642673
Tested-by: kokoro <noreply+kokoro@google.com>
Commit-Queue: Daniel Verkamp <dverkamp@chromium.org>
Reviewed-by: Noah Gold <nkgold@google.com>
The main distinction between PollContext and EpollContext was that the
latter was safe to use from multiple threads. This was not true of the
more widely-used PollContext for two reasons:
1. The `events` array was stored inside the `PollContext` structure,
whereas `EpollContext` required the caller to pass their own storage
for events.
2. `PollContext` had a hangup detection feature used to debug busy
looping cases that result from a failure to remove hungup file
descriptors from the context.
Point 1 is resolved by returning a `SmallVec` of events that avoids
allocation for the normal case (and in fact it should avoid allocation
in all cases on Linux, where the maximum number of events returned from
a single `epoll_wait()` call is limited to the same size as the
preallocated `SmallVec`). This simplifies the API and also means that
there is no need for per-context storage.
Point 2 can't easily be resolved, since it would require the `wait` call
to mutate shared state (this could be done by adding a mutex around the
shared data, but that seems like too much overhead for the value of the
feature). Instead, this patch just removes the hangup detection code.
BUG=b:213153157
TEST=tools/dev_container tools/presubmit --all
Change-Id: Ia48c46de96976da27cb5387e3e5e8fcf92d0e85b
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3633111
Commit-Queue: Daniel Verkamp <dverkamp@chromium.org>
Reviewed-by: Noah Gold <nkgold@google.com>
Tested-by: kokoro <noreply+kokoro@google.com>
Example help output (note that in main crosvm device help is broken)
```
Usage: crosvm device <command> [<args>]
Devices
Options:
--help display usage information
Commands:
block Block device
net Net device
console Console device
fs FS Device
gpu GPU device
vsock Vsock device
wl Wayland device
```
crosvm device wl help:
```
Usage: crosvm device wl --socket <PATH> --vm-socket <PATH> [--wayland-sock <PATH[,name=NAME]...>] [--resource-bridge <PATH>]
Wayland device
Options:
--socket path to bind a listening vhost-user socket
--vm-socket path to a socket for wayland-specific messages
--wayland-sock path to one or more Wayland sockets. The unnamed socket is
used for displaying virtual screens while the named ones are
used for IPC
--resource-bridge path to the GPU resource bridge
--help display usage information
```
Change-Id: Ie976b5545644c10b24cb3cdfcb17b213641064f9
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3668920
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Daniel Verkamp <dverkamp@chromium.org>
Reviewed-by: Vikram Auradkar <auradkar@google.com>
Reviewed-by: Alexandre Courbot <acourbot@chromium.org>
Commit-Queue: Anton Romanov <romanton@google.com>
This reverts commit e84d9c1d7c.
Reason for revert: vfio-pci fails on ManaTEE with an out-of-range BAR index (8), so further investigation is required
Original change's description:
> devices: pci: PciBarIndex API cleanup
>
> Everything that takes a PCI BAR index should use the appropriate
> PciBarIndex type rather than u8 (or usize) for better type safety and
> descriptiveness.
>
> BUG=b:232838930
> TEST=tools/presubmit
>
> Change-Id: I0a918a6da6327c963fb297ba765563774a7e5e49
> Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3664282
> Tested-by: kokoro <noreply+kokoro@google.com>
> Commit-Queue: Daniel Verkamp <dverkamp@chromium.org>
> Reviewed-by: Anton Romanov <romanton@google.com>
Bug: b:232838930
Change-Id: Id5d19e779289e47045af67fa51c301ba2b79ee7d
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3687516
Commit-Queue: Daniel Verkamp <dverkamp@chromium.org>
Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Once we have opened /dev/vfio/vfio we can leverage new_from_container().
BUG=None
TEST=cargo build
Change-Id: Ib309ab708cf561ec5dde8a606164a7df9a1bc4aa
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3675886
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
Commit-Queue: Alexandre Courbot <acourbot@chromium.org>
This reverts commit 56c0d02760.
Reason for revert: Break ARCVM after suspend/resume
BUG=b:234067421
TEST=Checked that ARCVM kept working after suspend/resume
Original change's description:
> crosvm: remove balloon stats request timeout
>
> It was done to avoid deadlock when stats are requested before guest is
> up. Implement a stub BalloonStats::NotReady replier until host is up so
> that timeout is no longer necessary.
>
> BUG=b:232289535
> TEST=few tast crostini/arc tests
>
> Change-Id: I6731b4ee9eaecdd65aebdd3f530f0932b0660c85
> Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3652887
> Reviewed-by: Daniel Verkamp <dverkamp@chromium.org>
> Tested-by: kokoro <noreply+kokoro@google.com>
> Commit-Queue: Anton Romanov <romanton@google.com>
> Auto-Submit: Anton Romanov <romanton@google.com>
Bug: b:232289535
Change-Id: Iffafaef1e1136ebc050d69d001d384eedd0a2319
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3686789
Commit-Queue: Keiichi Watanabe <keiichiw@chromium.org>
Auto-Submit: Hikaru Nishida <hikalium@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
Reviewed-by: Hikaru Nishida <hikalium@chromium.org>
Move this function to a place closer from its call site - this will make
the following CL less cluttered.
BUG=None
TEST=cargo build
Change-Id: I2837af9bb1f721967d7b7676140a8bd11973a34b
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3600171
Commit-Queue: Alexandre Courbot <acourbot@chromium.org>
Reviewed-by: Daniel Verkamp <dverkamp@chromium.org>
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Use the newly introduced async queue structs to manage the vhost-user
console device queues. This allows the device to restart its queues
after they are stopped or reset. The previous implementation was losing
the input and output streams as the queue was stopped.
BUG=b:228385297
BUG=b:228912920
TEST=vhost-user console device is working (with input) on Linux.
Change-Id: Ie968af7cfa77734c86c088da3a52f4d6937c3d0d
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3600170
Commit-Queue: Alexandre Courbot <acourbot@chromium.org>
Reviewed-by: Daniel Verkamp <dverkamp@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
Add a structure that facilitates managing the state and resources of a
queue running on an async executor. This is to be first used by the
console device, but the goal is also to extend these helpers and use
them on as many devices as possible.
BUG=b:228385297
BUG=b:228912920
TEST=cargo build
Change-Id: I72a3df7d9bf900cff38b189380d3c8ec4f098a0b
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3600169
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
Reviewed-by: Daniel Verkamp <dverkamp@chromium.org>
Commit-Queue: Alexandre Courbot <acourbot@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
With this and some pending hacks, dEQP works inside sibling
guest with VVU.
Device can be started with "crosvm device gpu --vfio <PCI ID>"
BUG=b:213532598
TEST=manual - run dEQP from a sibling guest when using VVU.
Change-Id: I97fa0e44c36cd5be13b6d31592878bdc19e48aed
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3682201
Auto-Submit: Lepton Wu <lepton@chromium.org>
Commit-Queue: Lepton Wu <lepton@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
This has only a tiny change outside of unix-specific code so shouldn't
complicate windows upstreaming
Change-Id: Iae68e761d730ed7497b4586b9b9a948524224d8e
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3671093
Reviewed-by: Daniel Verkamp <dverkamp@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Vikram Auradkar <auradkar@google.com>
Commit-Queue: Anton Romanov <romanton@google.com>
Reviewed-by: Alexandre Courbot <acourbot@chromium.org>
Skip logging an error when the translate request tube closes. That
happens whenever the process on the other side of the tube closes, and
isn't an error in and of itself.
BUG=None
TEST=exit crosvm and observe no virtio-iommu error message
Change-Id: I5f0460182a96d00f8c348224c838554710fc5451
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3659815
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Alexandre Courbot <acourbot@chromium.org>
Commit-Queue: David Stevens <stevensd@chromium.org>
Make sure the endpoint is removed from the endpoint and domain maps when
a vfio device is removed.
BUG=b:232161488
TEST=Boot a guest, hotplug in and out a vfio-pci device repeatedly
Change-Id: Ie9838b412cac22a9e14bb6e986be39303aeb01e3
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3676692
Tested-by: kokoro <noreply+kokoro@google.com>
Commit-Queue: David Stevens <stevensd@chromium.org>
Reviewed-by: Alexandre Courbot <acourbot@chromium.org>
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
Refactor shared state so that it can be shared between multiple async
callbacks. This moves all state from the Worker to a dedicated State
struct, which allows it to be passed via Rc<RefCell<>> to the various
async callbacks. The collection of endpoints can also be passed as part
of the shared state, rather than as its own dedicated argument.
There should be no behavioral changes in this CL.
BUG=b:232161488
TEST=Boot a guest, hotplug in and out a vfio-pci device repeatedly
Change-Id: I1e65ef3e4048bf0889fdee0f18fa5b58be264666
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3659554
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
Reviewed-by: Alexandre Courbot <acourbot@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Commit-Queue: David Stevens <stevensd@chromium.org>
Check that the same mapper isn't used for multiple domains. This ensures
that there won't be mapping collisions/leaks if the guest tries to use
the same mapper with multiple domains.
BUG=None
TEST=Boot ManaTEE with devices managed by viommu
Change-Id: I45889e6fe020c3a24b355fc05a8f74734bece78a
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3641221
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Commit-Queue: David Stevens <stevensd@chromium.org>
Remove incomplete support for detaching a virtio endpoint from domains
with multiple endpoints. We could technically handle this by checking
domain mappings when handling TranslateRequests, but there are no use
cases for this at the moment.
BUG=None
TEST=Boot ManaTEE with devices managed by viommu
Change-Id: I9d1cb86f246f7d8639058eaaf158f6538f1c1c57
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3676697
Reviewed-by: Alexandre Courbot <acourbot@chromium.org>
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Commit-Queue: David Stevens <stevensd@chromium.org>
Refactor create_vfio_device so that it returns the created viommu
related mapper, if any. This allows add_vfio_device to be written more
cleanly.
BUG=None
TEST=Boot ManaTEE with devices managed by viommu
Change-Id: Ieaa345ed5f9146fde8e305fc7bc37001ff708683
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3676698
Commit-Queue: David Stevens <stevensd@chromium.org>
Reviewed-by: Alexandre Courbot <acourbot@chromium.org>
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
For vfio devices which we expect to have granular, dynamic IOMMU
mappings (i.e. passthrough devices using viommu/coiommu), use the
ChromeOS specific VFIO_TYPE1v3_IOMMU if available, as it is optimized
for that use case.
BUG=b:217479627
TEST=boot ManaTEE, observe reduced hypervisor memory usage after IO
Change-Id: I7b3a04adb5049c165885917fa29cedf2ccdee1d0
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3675885
Commit-Queue: David Stevens <stevensd@chromium.org>
Reviewed-by: Alexandre Courbot <acourbot@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
When creating shared memory, 'name' can have one of two meanings. It
could either be a debugging tag with no semantic meaning, or it could
uniquely identify a shared memory object within some namespace. Linux
memfd uses name in the first meaning, whereas Windows (and Linux shm)
uses it in the second meaning.
Currently, crosvm has no use cases for the named shared memory of the
second type, so it is not supported. Make it clear that the SharedMemory
APIs treats name as a debugging-only name. Remove the "anon" and "named"
constructors, since they had no semantic meaning. Also require a name
when constructing a SharedMemory, since there's no reason not to provide
one to make debugging easier.
The only semantic change is setting the name of GuestMemory's underlying
shmem to "crosvm_guest", which it was until recently. This fixes some
ManaTEE tests which use the name to determine CrOS guest memory usage.
BUG=None
TEST=compiles
Change-Id: I78d5046df04d6f19640abbbc67af6bd433a177b0
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3676695
Commit-Queue: David Stevens <stevensd@chromium.org>
Reviewed-by: Daniel Verkamp <dverkamp@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Change the comparison from an `if let` into a regular `if ... ==` by
implementing `PartialEq` for the relevant types. This makes the lifetime
of the `wl_state.borrow_mut()` end before calling
`kick_evt.next_val().await()`.
Fixes a new clippy warning from Rust 1.61.0:
<https://rust-lang.github.io/rust-clippy/master/index.html#await_holding_refcell_ref>
BUG=None
TEST=tools/clippy # with Rust 1.61.0
Change-Id: I067c2ab532d753306cd5334df5bb4f2c7c2e6584
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3671130
Reviewed-by: Alexandre Courbot <acourbot@chromium.org>
Reviewed-by: Anton Romanov <romanton@google.com>
Commit-Queue: Daniel Verkamp <dverkamp@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
The Worker struct is storing a Descriptor obtained from a timer for the
sole purpose of adding it to a WaitContext when the worker thread starts
running. Instead of doing this, create the WaitContext and pass it as
part of the worker data, so all ioctls involving that descriptor are
performed directly on the timer, guaranteeing that the descriptor is
valid.
BUG=233968702
TEST=Linux VM starts and boots with `--split-irqchip` specified.
Change-Id: Ie5da1e9866742108f52c526100e936d292167ab7
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3670105
Reviewed-by: Michael Hoyle <mikehoyle@google.com>
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Daniel Verkamp <dverkamp@chromium.org>
Reviewed-by: Dennis Kempin <denniskempin@google.com>
Commit-Queue: Alexandre Courbot <acourbot@chromium.org>
