The guest kernel will allocate from PCI range of the virtgpu
device, and send physical addresses via a hypercall.
Right now, only support buffers that can be mmap'ed. We could
add optimizations for GBM buffers later if needed.
BUG=chromium:924405
TEST=compile
Change-Id: I094de96a2c35bcd2e18c8a6a2d8cdc39bb392e36
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/1626794
Tested-by: kokoro <noreply+kokoro@google.com>
Tested-by: Gurchetan Singh <gurchetansingh@chromium.org>
Commit-Queue: Gurchetan Singh <gurchetansingh@chromium.org>
Reviewed-by: Zach Reizner <zachr@chromium.org>
Adds support for virtio-pmem device as an alternative for virtio-blk.
Exposing disk image to guest as virtio-blk device results in both guest
and host independently caching the disk I/O. Using virtio-pmem device
allows to mount disk image as direct access (DAX) in the guest and thus
bypass the guest cache. This will reduce memory foodprint of the VMs.
BUG=None
TEST=cargo test
TEST=Boot patched termina kernel in crosvm; mount virtio-pmem device as
DAX and run xfstests.
Change-Id: I935fc8fc7527f79e5169f07ec7927e4ea4fa6027
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/crosvm/+/1605517
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Zach Reizner <zachr@chromium.org>
Commit-Queue: Jakub Staroń <jstaron@google.com>
AddressAllocator now maintains a HashMap<Alloc, (u64, u64, u64)>,
which uniquely maps a Allocation enum (e.g: PciBar(bus, dev, bar),
GpuRenderNode, etc...) to it's address, size, and human-readable tag
/ description.
The interface has also been modified to use Error instead of Option.
Aside from improving debugging, tracking allocations will have
numerous uses in the future. For example, when allocating guest memory
over VmControl sockets, it will be possible to restrict allocations to
pre-allocated slices of memory owned by the requesting device.
To plumb through PCI information to PCI devices, this CL necessitated
the addition of a PciDevice method called `assign_bus_dev`, which
notifies PCI devices of their uniquely assigned Bus and Device numbers.
BUG=chromium:936567
TEST=cargo test -p resources && cargo build --features="gpu gpu-forward"
Change-Id: I8b4b0e32c6f3168138739249ede53d03143ee5c3
Reviewed-on: https://chromium-review.googlesource.com/1536207
Commit-Ready: Daniel Prilik <prilik@google.com>
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Zach Reizner <zachr@chromium.org>
AddressRanges' name doesn't suggest that it's a SystemAllocator builder.
This CL renames it to SystemAllocatorBuilder, and adds a
SystemAllocator::builder() that removes the need to have a separate
import for the Builder.
A minor change, but it cleans up the interface a bit.
BUG=chromium:936567
TEST=cargo test -p resources && cargo build
Change-Id: I6d14368490c0d3c4018858f541e4ae5390995878
Reviewed-on: https://chromium-review.googlesource.com/1540398
Commit-Ready: Daniel Prilik <prilik@google.com>
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Zach Reizner <zachr@chromium.org>
In Rust 2018 edition, `extern crate` is no longer required for importing
from other crates. Instead of writing:
extern crate dep;
use dep::Thing;
we write:
use dep::Thing;
In this approach, macros are imported individually from the declaring
crate rather than through #[macro_use]. Before:
#[macro_use]
extern crate sys_util;
After:
use sys_util::{debug, error};
The only place that `extern crate` continues to be required is in
importing the compiler's proc_macro API into a procedural macro crate.
This will hopefully be fixed in a future Rust release.
extern crate proc_macro;
TEST=cargo check
TEST=cargo check --all-features
TEST=cargo check --target aarch64-unknown-linux-gnu
TEST=local kokoro
Change-Id: I0b43768c0d81f2a250b1959fb97ba35cbac56293
Reviewed-on: https://chromium-review.googlesource.com/1565302
Commit-Ready: David Tolnay <dtolnay@chromium.org>
Commit-Ready: ChromeOS CL Exonerator Bot <chromiumos-cl-exonerator@appspot.gserviceaccount.com>
Tested-by: David Tolnay <dtolnay@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: David Tolnay <dtolnay@chromium.org>
This may help reduce cases of conflicts between independent CLs each
appending a dependency at the bottom of the list, of which I hit two
today rebasing some of my open CLs.
TEST=cargo check --all-features
Change-Id: Ief10bb004cc7b44b107dc3841ce36c6b23632aed
Reviewed-on: https://chromium-review.googlesource.com/1557172
Commit-Ready: David Tolnay <dtolnay@chromium.org>
Tested-by: David Tolnay <dtolnay@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Daniel Verkamp <dverkamp@chromium.org>
Found by running: `cargo rustc -- -D bare_trait_objects`
Bare trait objects like `&Trait` and `Box<Trait>` are soft-deprecated in
2018 edition and will start warning at some point.
As part of this, I replaced `Box<Trait + 'static>` with `Box<dyn Trait>`
because the 'static bound is implied for boxed trait objects.
TEST=cargo check --all-features
TEST=cargo check --target aarch64-unknown-linux-gnu
TEST=local kokoro
Change-Id: I41c4f13530bece8a34a8ed1c1afd7035b8f86f19
Reviewed-on: https://chromium-review.googlesource.com/1513059
Commit-Ready: ChromeOS CL Exonerator Bot <chromiumos-cl-exonerator@appspot.gserviceaccount.com>
Tested-by: David Tolnay <dtolnay@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: David Tolnay <dtolnay@chromium.org>
Separated out of CL:1513058 to make it possible to land parts
individually while the affected crate has no other significant CLs
pending. This avoids repeatedly introducing non-textual conflicts with
new code that adds `use` statements.
TEST=cargo check
TEST=cargo check --all-features
TEST=cargo check --target aarch64-unknown-linux-gnu
Change-Id: I87103777763ab9c00e6053605c5168f43243fc25
Reviewed-on: https://chromium-review.googlesource.com/1520066
Commit-Ready: David Tolnay <dtolnay@chromium.org>
Tested-by: David Tolnay <dtolnay@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: David Tolnay <dtolnay@chromium.org>
This is an easy step toward adopting 2018 edition eventually, and will
make any future CL that sets `edition = "2018"` this much smaller.
The module system changes in Rust 2018 are described here:
https://doc.rust-lang.org/edition-guide/rust-2018/module-system/path-clarity.html
Generated by running:
cargo fix --edition --all
in each workspace, followed by bin/fmt.
TEST=cargo check
TEST=cargo check --all-features
TEST=cargo check --target aarch64-unknown-linux-gnu
Change-Id: I000ab5e69d69aa222c272fae899464bbaf65f6d8
Reviewed-on: https://chromium-review.googlesource.com/1513054
Commit-Ready: ChromeOS CL Exonerator Bot <chromiumos-cl-exonerator@appspot.gserviceaccount.com>
Tested-by: David Tolnay <dtolnay@chromium.org>
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: David Tolnay <dtolnay@chromium.org>
When wl-dmabuf is not enabled, rustc complains about unused imports and
enum values. Add compiler directives to silence the warnings.
BUG=None
TEST='cargo build', 'emerge-nami crosvm'
Change-Id: Ib39735d329f8aa835c0b5842b10bfe78d0e578d9
Signed-off-by: Daniel Verkamp <dverkamp@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/1327827
Refactor existing code to use msg_socket.
BUG=None
TEST=local build and run
Change-Id: Iee72326b330e035303f679e1aedd6e5d18ad4f8a
Reviewed-on: https://chromium-review.googlesource.com/1260260
Commit-Ready: Jingkui Wang <jkwang@google.com>
Tested-by: Jingkui Wang <jkwang@google.com>
Reviewed-by: Zach Reizner <zachr@chromium.org>
Now that cargo fmt has landed, run it over everything at once to bring
rust source to the standard formatting.
TEST=cargo test
BUG=None
Change-Id: Ic95a48725e5a40dcbd33ba6d5aef2bd01e91865b
Reviewed-on: https://chromium-review.googlesource.com/1259287
Commit-Ready: ChromeOS CL Exonerator Bot <chromiumos-cl-exonerator@appspot.gserviceaccount.com>
Tested-by: Zach Reizner <zachr@chromium.org>
Reviewed-by: Zach Reizner <zachr@chromium.org>
Combine GPU buffer allocation with the system resource allocator making
life easier as only one allocator needs to get passed to the execute
function.
Change-Id: I199eb0fd6b99b629aaec1ae3295e8a1942da5309
Signed-off-by: Dylan Reid <dgreid@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/1099856
Allow IRQs to be assigned before creating device manager.
For PCI, we need to add devices with interrupts before MMIO setup. Add
the ability to tell the architecture device manager about IRQs that we
have stolen.
There was only one function in device_manager and all of its state is
now delegated to the resource allocator, remove it.
Change-Id: I9afa0e3081a20cb024551ef18ae34fe76a1ef39d
Reviewed-on: https://chromium-review.googlesource.com/1089720
Commit-Ready: Dylan Reid <dgreid@chromium.org>
Tested-by: Dylan Reid <dgreid@chromium.org>
Reviewed-by: Sonny Rao <sonnyrao@chromium.org>
Start a system resource allocator that will be able to manage the
resources specific to each architecture.
Change-Id: I98cf35c280fefd7b0000801eb7405a236373b753
Signed-off-by: Dylan Reid <dgreid@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/1089719
Reviewed-by: Sonny Rao <sonnyrao@chromium.org>
Add the AddressAllocator module that will be used by both architectures
to manage distributing address ranges to devices. This will make the
addition of PCI devices easier as now both MMIO and PCI will need to
share address space. Add this to a new resources crate.
Change-Id: I6a971dd795f2118bd6cfec7dc34a65b0d4a32f9b
Signed-off-by: Dylan Reid <dgreid@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/1072570
Reviewed-by: Sonny Rao <sonnyrao@chromium.org>
