Add support for shared memory regions to vhost-user. This is adding
support for a front-end message to query for necessary shared memory
regions plus back-end message to support mapping/unmapping files from
the shared memory region.
go/vvu-shared-memory
BUG=b:201745804
TEST=compiles
Change-Id: I35c5d260ee09175b68f6778b81883e0070ee0265
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3716344
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
Commit-Queue: David Stevens <stevensd@chromium.org>
Reviewed-by: Alexandre Courbot <acourbot@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Make Listener type depend on Endpoint, instead of Endpoint type
depending on Listener. This will allow Endpoints to be used for the
slave_req_fd.
BUG=b:201745804
TEST=compiles
Change-Id: I1a0732bf7477a4e768f348f78e352a3b53607633
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3716342
Commit-Queue: David Stevens <stevensd@chromium.org>
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
Reviewed-by: Vikram Auradkar <auradkar@google.com>
Tested-by: kokoro <noreply+kokoro@google.com>
The existing code confuses the length of the request with the length of
the reply in recv_reply_with_payload. This makes it impossible to use
for any requests where the reply differs in size. Fix this by
determining payload size after reading the reply header.
BUG=b:201745804
TEST=compiles
Change-Id: Iec21918e45b81d299f3178e7893e146cbd4c7db7
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3716343
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Commit-Queue: David Stevens <stevensd@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>
These files are not used, and libslirp-rs is not a functional crate.
Only context.rs is used in heavily forked form as a part of crosvm's `net_util`.
BUG=b:213151463
TEST=n/a
Change-Id: I6536cde7ff2beed655e83c3d920c6940ed4969a6
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3650467
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
Commit-Queue: Noah Gold <nkgold@google.com>
The set_nonblocking() implementation of the VFIO Listener was marked as
unimplemented and caused a crash. However, given that accept() on a VFIO
listener never blocks, it is just as safe to mark the operation as
successful, since the behavior of accept() won't be impacted and it will
behave properly both in blocking and non-blocking modes.
Removing this panic will allow us to safely use the Listener's interface
set_nonblocking() method in generic code.
BUG=b:229554679
TEST=vhost-user console device works.
TEST=vvu console device works.
Change-Id: Ibbde7bcd048505fc3a84eb4881a74285bae6bf54
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3591111
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Commit-Queue: Alexandre Courbot <acourbot@chromium.org>
The SlaveReqHandler should be the sole owner of a VhostUserBackend since
it calls its mutable methods, so we shouldn't need an Arc here. If
sharing is necessary users can use their own locking mechanism.
Single ownership will help to retrieve the backend after a client
disconnects, to reuse it for another connection if needed.
BUG=b:229554679
BUG=b:216407443
TEST=cargo test -p vmm_vhost
TEST=vhost-user console device works.
TEST=cargo test -p devices
Change-Id: I2358c807ac3ddb1ee4b29d97df0ade5a3e30a85a
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3591108
Commit-Queue: Alexandre Courbot <acourbot@chromium.org>
Reviewed-by: Richard Zhang <rizhang@google.com>
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
For cross-domain, virgl_renderer needs to be able to import a dmabuf fd
allocated by another component.
BUG=b:230100768
TEST=gfxbench, android games in arcvm
Change-Id: I9d6f965589ec24f06a83a6d921a9c00db1f057f6
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3501130
Reviewed-by: Daniel Verkamp <dverkamp@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Auto-Submit: Rob Clark <robdclark@chromium.org>
Commit-Queue: Rob Clark <robdclark@chromium.org>
Reviewed-by: Ryan Neph <ryanneph@google.com>
Pull in commit 77c59db8ff9a ("rust: add minijail_run_env to
minijail-sys") so the minijail-sys build script does not keep making the
submodule dirty on every build.
BUG=None
TEST=cargo build && git status
Change-Id: I662a4a9801eb27245691f69644508026a69e8a11
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3631409
Reviewed-by: Anton Romanov <romanton@google.com>
Tested-by: kokoro <noreply+kokoro@google.com>
Commit-Queue: Daniel Verkamp <dverkamp@chromium.org>
Use the PollToken derive and the build_with() function to make the
example more representative of actual uses.
BUG=None
TEST=cargo doc
TEST=cargo test -p base --doc wait_context
Change-Id: I1f154dfd6247786024dc94035ea63db11968e85a
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3630424
Reviewed-by: Dennis Kempin <denniskempin@google.com>
Reviewed-by: Noah Gold <nkgold@google.com>
Tested-by: kokoro <noreply+kokoro@google.com>
Commit-Queue: Daniel Verkamp <dverkamp@chromium.org>
Pull in ff062d54f9a4 ("rust: remove redundant format in unreachable") to
fix warnings in Rust 1.60.
BUG=None
TEST=cargo build # no warnings with Rust 1.60.0
Change-Id: Ib2bf2445bc064fe3f81589e426de104565805bbf
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3624801
Tested-by: kokoro <noreply+kokoro@google.com>
Commit-Queue: Daniel Verkamp <dverkamp@chromium.org>
Reviewed-by: Dennis Kempin <denniskempin@google.com>
Having a default protocol implementation is dangerous, as implementors
can just ignore this method and return an erroneous type.
BUG=None
TEST=cargo build
Change-Id: Iee7dbcf563be1e5b15903b0fd7bdfb7ff5823545
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3565619
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Noah Gold <nkgold@google.com>
Commit-Queue: Alexandre Courbot <acourbot@chromium.org>
Also fixes few warning introduced during upstream which linux clippy
run does not catch.
test: py tools\clippy
bug: 226966790
Change-Id: I6979f87be5d6d34727afabafcaccb57631975567
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3553401
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Dennis Kempin <denniskempin@google.com>
Commit-Queue: Vikram Auradkar <auradkar@google.com>
Does couple of things
- move tube under right directory
- cleanup cros_async dependency
- export tube module from base
- enable base windows tests
- add verbose flag to test runner
test: built on window and presubmit
bug: b:213151429
bug: b:213153157
Change-Id: I0d1906d7edcc3352db53325ea7527550eaf00895
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3550296
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Dennis Kempin <denniskempin@google.com>
Commit-Queue: Vikram Auradkar <auradkar@google.com>
This is follow-up of CL:3468235.
For vhost-user, a device (i.e. slave) works like a server while a
vmm (i.e., master) does like a client. So the devices should allow the
vmm to close the connection. An error handling code added by CL:3468235
was opposite. It should have been added to `slave_req_handler.rs`
instead of `master_req_handler.rs`.
BUG=b:219674197
BUG=b:220639724
TEST=run tast.vm.Audio*.vhost_user_cras and tast.vm.VhostUserNet
Change-Id: Ie661c02bed8287d9d770c6c6c0c658ef6681db10
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3533239
Reviewed-by: Alexandre Courbot <acourbot@chromium.org>
Auto-Submit: Keiichi Watanabe <keiichiw@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Commit-Queue: Keiichi Watanabe <keiichiw@chromium.org>
Use the cross-platform base::Event type to replace the Linux-specific
base::EventFd.
BUG=b:221882601
BUG=b:219522861
TEST=tools/presubmit
Change-Id: I0503c130306b508af9dd5044653408bd316c7ef1
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3533618
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Commit-Queue: Daniel Verkamp <dverkamp@chromium.org>
Windows vmm_vhost differs from linux vmm_vhost mostly on the structures
that it depends on. For example windows uses Tube for linux'
UnixStream.
The patch aliases some of these types in a platform specific file. Rest
of the platform independent code depends on those aliases.
Bug: b:213151429
Test: presubmit, manually run cargo test on windows.
Upstream-Crate: third_party/vmm_vhost
Change-Id: I3f98fbec1ffdd8914dc5befddce3a8483a5fd94f
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3496082
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
Commit-Queue: Vikram Auradkar <auradkar@google.com>
When a client sends 0-sized data, the device regards that the client
closed the connection and will exit properly.
BUG=b:219674197
TEST=check if no error is shown when a vhost-user blk vmm disconnects
Change-Id: Icd84fb241c39b6dc54a8af41a068a6dd95ce4c2b
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3468235
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Alexandre Courbot <acourbot@chromium.org>
Commit-Queue: Keiichi Watanabe <keiichiw@chromium.org>
Removed `vmm_vhost::Error::should_reconnect()` which we're not using at
all. We don't support reconnect feature at this point.
This function exists even in rust-vmm/vhost, but neither
cloud-hypervisor or rust-vmm/vhost-user-backend is using it.
BUG=none
TEST=cargo check
Change-Id: I4fda22d927573acfedcb189ad786aeb2cf7496e1
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3468234
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Vikram Auradkar <auradkar@google.com>
Reviewed-by: Noah Gold <nkgold@google.com>
Reviewed-by: Alexandre Courbot <acourbot@chromium.org>
Commit-Queue: Keiichi Watanabe <keiichiw@chromium.org>
Run tests for sys_util_core, poll_token_derive and balloon_control on
windows.
Using dotfiles to disable/serialize test runs of a subset of crates does
not work well with third party crates as it forces us to commit the dot
file to the crate.
The patch modifies and uses the script that runs linux tests.
This patch also allows us to
- build/test child crate even if parent crate has disabled build/test.
- avoid building crosvm if it is not explicitly specified.
RIP short lived .windows_build_test_skip. You allowed us to run noop
kokoro tests.
Test: py .\tools\impl\test_runner.py --arch x86_64
Bug: b:215610772
Change-Id: Icc6d04ffd7c0c33d4f60aeac16fc7d23881c387d
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3459809
Reviewed-by: Dennis Kempin <denniskempin@google.com>
Tested-by: kokoro <noreply+kokoro@google.com>
Commit-Queue: Vikram Auradkar <auradkar@google.com>
The patch prepares backend.rs and connection.rs to be platform
independent by
- depending on base instead on sys_util
- using base::Event over sys_util::EventFd
- replacing RawFd usage with RawDescriptor and derivatives
To keep the noise low, the patch
- aliases structs/traits (Event as EventFd)
- does not rename variables (say from fd to rd).
Note: With this patch the crate/files are not completely platform
independent.
Test: Built, clippy and fmt
Bug: b:213151429
Change-Id: Ib57528ef1a951c3d083cf345c878ec1417b7ce3e
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3460428
Reviewed-by: Dennis Kempin <denniskempin@google.com>
Tested-by: kokoro <noreply+kokoro@google.com>
Commit-Queue: Vikram Auradkar <auradkar@google.com>
The patch prepares backend.rs and connection.rs to be platform
independent by
- depending on base instead on sys_util
- using base::Event over sys_util::EventFd
- replacing RawFd usage with RawDescriptor and derivatives
To keep the noise low, the patch
- aliases structs/traits (Event as EventFd)
- does not rename variables (say from fd to rd).
Note: With this patch the crate/files are not completely platform
independent.
Test: Built, clippy and fmt
Bug: b:213151429
Change-Id: Id93d4a70db627578e6577017f8ca4d69f892d69a
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3460427
Reviewed-by: Dennis Kempin <denniskempin@google.com>
Tested-by: kokoro <noreply+kokoro@google.com>
Commit-Queue: Vikram Auradkar <auradkar@google.com>
The patch also adds files to skip building and testing crates on
windows. When we run
```
tools/windows/build_test.py --skip_file_name .windows_build_test_skip
```
the build/test succeeds without actually doing anything as build/test
for all crates is skipped by creating '.windows_build_test_skip'.
Bug: 213170957
Test: Ran script on downstream repo
Change-Id: Iebd2cea463ee722be4feaed88229e1fb5e9fd6c5
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3417918
Reviewed-by: Dennis Kempin <denniskempin@google.com>
Tested-by: kokoro <noreply+kokoro@google.com>
Commit-Queue: Vikram Auradkar <auradkar@google.com>
This change caches the parameters received in the activate method. It
spawns a Worker thread to process Vhost master communication.
Furthermore, it forwards all non-action messages to the Vhost slave.
BUG=b:194136484
TEST=Compile.
Change-Id: Ie853add605ea4daa1c62ec9ce93cbac083e57de7
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3288870
Reviewed-by: Chirantan Ekbote <chirantan@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Commit-Queue: Abhishek Bhardwaj <abhishekbh@chromium.org>
The type parameter `S: SlaveReqHandler` for `SlaveReqHelper` was
introduced to check if it's vvu or not.
Howver, SlaveReqHelper shouldn't take a type parameter of
`SlaveReqHandler` because the helper will be used in the VVU proxy
device which doesn't have its own req_handler.
So, instead of taking the type parameter, we should have a new bool
field which indicates if it's VVU or not.
BUG=b:194136484
TEST=cargo test --all-features in vmm_vhost
TEST=kokoro
Change-Id: I35f20ef7c42cd74678ab5d87b68bbb58586d2fa5
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3320122
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Chirantan Ekbote <chirantan@chromium.org>
Commit-Queue: Keiichi Watanabe <keiichiw@chromium.org>
Allow having a Master struct with a backend other than socket. This will
be helpful for multi-platform support.
BUG=b:194137301
TEST=cargo test --all-features in vmm_vhost
Change-Id: Ie1bd8949ddb708c87a964544c6ae2fe7391b1fb5
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3320904
Reviewed-by: Chirantan Ekbote <chirantan@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Commit-Queue: Keiichi Watanabe <keiichiw@chromium.org>
Expose connection::{socket, vfio} modules so that we don't need to have
prefix for structs there.
e.g. SocketEndpoint -> connection::socket::Endpoint
BUG=b:194137301
TEST=cargo test --all-features in vmm_vhost
TEST=cargo check in crosvm
Change-Id: I7d992a0df5a838fa6a726f366e9595b20405ec3b
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3317324
Reviewed-by: Chirantan Ekbote <chirantan@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Commit-Queue: Keiichi Watanabe <keiichiw@chromium.org>
Make `send_bufs_with_fd(s)` take `IoSlice` instead of `&[u8]`.
Note that no Chrome OS-side is needed for this API change because these
methods aren't used in any other Chrome OS projects.
BUG=b:204720423
TEST=cargo test in sys_utils and vmm_vhost
Change-Id: I02701d9c4e964ea95f0746b17254fa029fee7531
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3302696
Tested-by: kokoro <noreply+kokoro@google.com>
Commit-Queue: Keiichi Watanabe <keiichiw@chromium.org>
Reviewed-by: Chirantan Ekbote <chirantan@chromium.org>
send_ack_message used to take a module local Result as a parameter. It
doesn't make sense for other modules to use this. This change makes the
method take a bool instead of the aforementioned Result.
BUG=b:194136484
TEST=Compile.
Change-Id: I475cc39826db94ef2c686e754e13f478f8109d0a
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3285629
Tested-by: kokoro <noreply+kokoro@google.com>
Commit-Queue: Abhishek Bhardwaj <abhishekbh@chromium.org>
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
This change adds a helper class that talks to the Vhost User Master.
This way the communication logic can be used by a Virtio Vhost User
Slave implementation as well as a Vhost Virtio User proxy device.
BUG=b:194136484
TEST=Compile.
Change-Id: I7d675c6739f7c8675320d956175c63c8d1190b79
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3285628
Commit-Queue: Abhishek Bhardwaj <abhishekbh@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Chirantan Ekbote <chirantan@chromium.org>
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
This change exposes members from the vhost_user implementation to be
used by the crosvm virtio-vhost-user proxy device.
BUG=b:194136484
TEST=Compile.
Change-Id: Ib2748f4e19138e08baaa6b2e8ae45f1b44c42cb1
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3285627
Commit-Queue: Abhishek Bhardwaj <abhishekbh@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Auto-Submit: Abhishek Bhardwaj <abhishekbh@chromium.org>
Reviewed-by: Keiichi Watanabe <keiichiw@chromium.org>
Since vhost_user is the only module in vmm_vhost now, remove this module
and have everything in the top-level module.
BUG=b:206900656
TEST=cargo check
Change-Id: I6663a93654e5d692efc7d9904fff0d207144c352
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3295219
Tested-by: kokoro <noreply+kokoro@google.com>
Commit-Queue: Keiichi Watanabe <keiichiw@chromium.org>
Reviewed-by: Chirantan Ekbote <chirantan@chromium.org>
Remove the previous vmm_vhost::Error enum, as most of its variants are
no longer used. Instead, expose vhost_user::Error as the top-level one.
BUG=b:206900656
TEST=cargo check
Change-Id: Ic5d9f7dcc6e2986d7b51885e2c0f44aa4e012e5f
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/3295218
Tested-by: kokoro <noreply+kokoro@google.com>
Commit-Queue: Keiichi Watanabe <keiichiw@chromium.org>
Reviewed-by: Chirantan Ekbote <chirantan@chromium.org>
