Skip to content
This repository has been archived by the owner on Jan 4, 2024. It is now read-only.

AmbiML/sparrow-manifest

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

20 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Project Sparrow

Sparrow is a project to build a low-power secure embeded platform for Ambient ML applications. The target platform leverages RISC-V and OpenTitan. The Sparrow software includes a home-grown operating system named CantripOS, that runs on top of seL4 and (ignoring the seL4 kernel) is written almost entirely in Rust.

Sparrow (and CantripOS) are definitely a work in progress. The CantripOS components are based on an augmented version of seL4's CAmkES framework. Critical system services are CAmkES components that are statically configured. Applications are developed using an AmbiML-focused SDK and dynamically loaded by the system services.

Sparrow software repositories (what's included here).

Sparrow consists of multiple git repositories stitched together with the repo tool. The following git repositories are currently available:

  • camkes-tool: seL4's camkes-tool repository with additions to support CantripOS services
  • capdl: seL4's capdl repository with addition for CantripOS services and the CantripOS rootserver (a replacement for capdl-loader-app that is written in Rust and supports hand-off of system resources to the CantripOS MemoryManager service)
  • kernel: seL4's kernel with drivers for Sparrow's RISC-V platform and support for reclaiming the memory used by the CantripOS rootserver
  • cantrip-full frameworks for developing in Rust and the CantripOS system services
  • scripts: support scripts including build-sparrow.sh

[More software will be published as we deem it ready for sharing until eventually all of Sparrow (software and hardware designs) will be available.]

Most CantripOS Rust crates are in the cantrip/apps/system/components directory. Common/shared code is in cantrip-os-common:

  • allocator: a heap allocator built on the linked-list-allocator crate
  • camkes: support for writing CAmkES components in Rust
  • capdl: support for reading the capDL specification generated by capDL-tool
  • copyregion: a helper for temporarily mapping physical pages into a thread's VSpace
  • cspace-slot: an RAII helper for the slot-allocator
  • logger: seL4 integration with the Rust logger crate
  • model: support for processing capDL; used by the cantrip-os-rootserver
  • panic: an seL4-specific panic handler
  • sel4-config: build glue for seL4 kernel configuration
  • sel4-sys: seL4 system interfaces & glue
  • slot-allocator: an allocator for slots in the top-level CNode

CantripOS system services make up the remaining components:

  • DebugConsole: a command line interface intended for debug builds
  • MailboxDriver: a driver for the Mailbox interface used to communicate betwen the security and management cores [sparow-only]
  • MemoryManager: the memory / object manager service that supports dynamic memory management
  • MlCoordinator: a service that manages running ML jobs [requires support that is currently on on sparrow]
  • OpenTitanUARTDriver: a driver for the UART on the management core [sparrow-only]
  • ProcessManager: the service that creates & manages execution of applications
  • SDKRuntime: the service that handles application runtime requests
  • SecurityCoordinator: the service that provides an interface to the security core (using the MailboxDriver)
  • TimerService: a service built on top of the management core timer hardware [requires hardware timer support]

The components used depends on the target platform. At the moment two platforms are buildable: sparrow and rpi3 (Raspberry Pi BCM2837 running in 64-bit mode). The sparrow platform is not useful other than for reference as building it requires toolchain & simulator support that is not yet released. The rpi3 platform is the intended target platform for public consumption. Contribution of additional platform support is welcomed (e.g. a timer driver for the TimerService).

All platform-supported CantripOS services are included in both debug and release builds. Production builds replace the DebugConsole with a more limited interface and drop the UART driver when there is no serial console. The SDKRuntime is more a proof of concept than anything else. There are test applications written in C and Rust in the apps tree that exercise the experimental api. Production systems may have their own SDK and associated runtime tailored to their needs.

The other main Rust piece is the rootserver application that is located in projects/capdl/cantrip-os-rootserver. This depends on the capdl and model submodules of cantrip-os-common. It is possible to select either cantrip-os-rootserver or the C-based capdl-loader-app with a CMake setting in the CAmkES project's easy-settings.cmake file; e.g. projects/cantrip/easy-settings.cmake has:

#set(CAPDL_LOADER_APP "capdl-loader-app" CACHE STRING "")
set(CAPDL_LOADER_APP "cantrip-os-rootserver" CACHE STRING "")

But using capdl-loader-app is not advised because it lacks important functionality found only in cantrip-os-rootserver.

How we do software development for the Sparrow platform.

Our primary development environment uses Renode for simulation of our Sparrow hardware design. Renode allows us to do rapid software/hardware co-design of our multi-core RISC-V target platform. The software environment is derived from those provided by seL4 & OpenTitan. Debugging leverages Renode facilites and gdb talking to Renode for system software and application developement.

How to do software development on non-Sparrow platforms.

For public use CantripOS works on the seL4 rpi3 platform with aarch64 (64-bit ARM) enabled (as opposed to 32-bit aarch32). Instead of Renode, qemu is used to simulate the hardware platform. In theory the software should run on real hardware too but it's not been tried. Support for other platforms can be added.

Getting started with repo & the build system.

This repository includes a multi-platform build framework for use with the Sparrow software. This framwork leverages make, cmake, and cargo. To get started follow these steps:

  1. Clone the Sparrow project from GitHub using the repo tool We assume below this lands in a top-level directory named "sparrow".
  2. Download, build, and boot the system to the Cantrip shell prompt. For now the only target platform that works is "rpi3" (for a raspi3b machine running in simulation on qemu).
mkdir sparrow
cd sparrow
repo init -u https://github.com/AmbiML/sparrow-manifest -m sparrow-manifest.xml
repo sync -j$(nproc)
export PLATFORM=rpi3
source build/setup.sh
m prereqs
m simulate

[Beware that if your repo tool is out of date you may need to supply -b main to the init request as older versions of repo only check for a master branch.]

Note the above assumes you have the follow prerequisites installed on your system and in your shell's search path:

  1. Gcc (or clang) for the target architecture
  2. Rust; any nightly build >=nightly-2021-11-05 should work. A default version is set in the build/setup.sh script; if that is not what you are using either edit the shell script or export CANTRIP_RUST_VERSION in each shell where you work. Beware that we use various nightly-only features that are not supported by stable versions of Rust (e.g. to override the default TLS model).
  3. Whichever simulator seL4 expects for your target architecture; e.g. for aarch64 this is qemu-system-aarch64.

Because Sparrow is a CAmkES project you also need CAmkES dependencies.

Sparrow uses repo to download and piece together Sparrow git repositories as well as dependent projects / repositories such as seL4.

$ repo init -u https://github.com/AmbiML/sparrow-manifest -m sparrow-manifest.xml
Downloading Repo source from https://gerrit.googlesource.com/git-repo

repo has been initialized in <your-directory>/sparrow/
If this is not the directory in which you want to initialize repo, please run:
   rm -r <your-directory>/sparrow//.repo
and try again.
$ repo sync -j12
Fetching: 100% (23/23), done in 9.909s
Garbage collecting: 100% (23/23), done in 0.218s
Checking out: 100% (23/23), done in 0.874s
repo sync has finished successfully.
$ export PLATFORM=rpi3
$ source build/setup.sh
========================================
ROOTDIR=/<your-directory>/sparrow
OUT=/<your-directory>/sparrow/out
PLATFORM=rpi3
PYTHON_SPARROW_ENV=<your-directory>/sparrow/cache/rpi3-venv
========================================

Type 'm [target]' to build.

Targets available are:

...
cantrip cantrip-build-debug-prepare cantrip-build-release-prepare cantrip-builtins
cantrip-builtins-debug cantrip-builtins-release cantrip-bundle-debug cantrip-bundle-release
cantrip-clean cantrip-clean-headers cantrip-clippy cantrip-component-headers
...

To get more information on a target, use 'hmm [target]'

$ m prereqs
<your-directory>/sparrow/scripts/install-prereqs.sh \
        -p "<your-directory>/sparrow/scripts/python-requirements.txt \
                 " \
        -a ""
Installing apt package dependencies...
Installing python package dependencies...
Creating virtual python environment <your-directory>/sparrow/cache/rpi3-venv
...
Installation complete.
$ m simulate
...
info: component 'rust-std' for target 'aarch64-unknown-none' is up to date
loading initial cache file <your-directory>/sparrow/cantrip/projects/camkes/settings.cmake
-- Set platform details from PLATFORM=rpi3
--   KernelPlatform: bcm2837
--   KernelARMPlatform: rpi3
-- Setting from flags KernelSel4Arch: aarch64
-- Found seL4: <your-directory>/sparrow/kernel
-- The C compiler identification is GNU 11.2.1
...
[291/291] Generating images/capdl-loader-image-arm-bcm2837
...
qemu-system-aarch64 -machine raspi3b -nographic -serial null -serial mon:stdio -m size=1024M -s \
-kernel /<your-directory>/sparrow/out/cantrip/aarch64-unknown-elf/release/capdl-loader-image \
--mem-path /<your-directory>/sparrow/out/cantrip/aarch64-unknown-elf/release/cantrip.mem
ELF-loader started on CPU: ARM Ltd. Cortex-A53 r0p4
  paddr=[8bd000..fed0ff]
No DTB passed in from boot loader.
Looking for DTB in CPIO archive...found at 9b3ef8.
Loaded DTB from 9b3ef8.
   paddr=[23c000..23ffff]
ELF-loading image 'kernel' to 0
  paddr=[0..23bfff]
  vaddr=[ffffff8000000000..ffffff800023bfff]
  virt_entry=ffffff8000000000
ELF-loading image 'capdl-loader' to 240000
  paddr=[240000..4c0fff]
  vaddr=[400000..680fff]
  virt_entry=4009e8
Enabling MMU and paging
Jumping to kernel-image entry point...

Warning:  gpt_cntfrq 62500000, expected 19200000
Bootstrapping kernel
Booting all finished, dropped to user space
cantrip_os_rootserver::Bootinfo: (1969, 131072) empty slots 1 nodes (15, 83) untyped 131072 cnode slots
cantrip_os_rootserver::Model: 1821 objects 1 irqs 0 untypeds 2 asids
cantrip_os_rootserver::capDL spec: 0.39 Mbytes
cantrip_os_rootserver::CAmkES components: 5.85 Mbytes
cantrip_os_rootserver::Rootserver executable: 1.07 Mbytes
<<seL4(CPU 0) [decodeARMFrameInvocation/2137 T0xffffff80004c7400 "rootserver" @44373c]: ARMPageMap: Attempting to remap a frame that does not belong to the passed address space>>
...
<<seL4(CPU 0) [decodeCNodeInvocation/107 T0xffffff80009a3400 "rootserver" @4268a0]: CNode Copy/Mint/Move/Mutate: Source slot invalid or empty.>>
...
CANTRIP> cantrip_memory_manager::Global memory: 0 allocated 124501760 free, reserved: 2334720 kernel 7340032 user
...

The m simulate command can be run repeatedly. If you need to reset your setup just remove the build tree and re-run m simulate; e.g.

cd sparrow
m clean
m simulate

Depending on Rust crates

To use crates from Sparrow you can reference them from a local repository or directly from GitHub using git; e.g. in a Config.toml:

cantrip-os-common = { path = "../system/components/cantrip-os-common" }
cantrip-os-common = { git = "https://github.com/AmbiML/sparrow/cantrip-full" }

NB: the git usage depends on cargo's support for searching for a crate named "cantrip-os-common" in the cantrip repo. When using a git dependency a git tag can be used to lock the crate version.

Note that many Sparrow crates need the seL4 kernel configuration (e.g. to know whether MCS is configured). This is handled by the cantrip-os-common/sel4-config crate that is used by a build.rs to import kernel configuration parameters as Cargo features. In a Cargo.toml create a features manifest with the kernel parameters you need e.g.

[features]
default = []
# Used by sel4-config to extract kernel config
CONFIG_PRINTING = []

then specify build-dependencies:

[build-dependencies]
# build.rs depends on SEL4_OUT_DIR = "${ROOTDIR}/out/cantrip/kernel"
sel4-config = { path = "../../cantrip/apps/system/components/cantrip-os-common/src/sel4-config" }

and use a build.rs that includes at least:

extern crate sel4_config;
use std::env;

fn main() {
    // If SEL4_OUT_DIR is not set we expect the kernel build at a fixed
    // location relative to the ROOTDIR env variable.
    println!("SEL4_OUT_DIR {:?}", env::var("SEL4_OUT_DIR"));
    let sel4_out_dir = env::var("SEL4_OUT_DIR")
        .unwrap_or_else(|_| format!("{}/out/cantrip/kernel", env::var("ROOTDIR").unwrap()));
    println!("sel4_out_dir {}", sel4_out_dir);

    // Dredge seL4 kernel config for settings we need as features to generate
    // correct code: e.g. CONFIG_KERNEL_MCS enables MCS support which changes
    // the system call numbering.
    let features = sel4_config::get_sel4_features(&sel4_out_dir);
    println!("features={:?}", features);
    for feature in features {
        println!("cargo:rustc-cfg=feature=\"{}\"", feature);
    }
}

Note how build.rs expects an SEL4_OUT_DIR environment variable that has the path to the top of the kernel build area. The build/cantrip.mk build glue sets this for you but, for example, if you choose to run ninja directly you will need it set in your environment.

Similar to SEL4_OUT_DIR the cantrip-os-common/src/sel4-sys crate that has the seL4 system call wrappers for Rust programs requires an SEL4_DIR envronment variable that has the path to the top of the kernel sources. This also is set by build/cantrip.mk.

Source Code Headers

Every file containing source code includes copyright and license information. For dependent / non-Google code these are inherited from the upstream repositories. If there are Google modifications you may find the Google Apache license found below.

Apache header:

Copyright 2022 Google LLC

Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at

    https://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.

Releases

No releases published

Packages

No packages published

Contributors 4

  •  
  •  
  •  
  •