The supervisor binary runs inside sandbox images whose libc and glibc
version are unknown at build time, so it must be statically linked. Add
SUPERVISOR_LIBC to select between the default musl variant and a new
glibc-static variant that builds the GNU target with +crt-static.
glibc-static has no cross-compile path: zig cc accepts -static for
*-linux-gnu targets and emits a dynamically linked binary anyway. The
staging script therefore refuses a cross-arch request for that variant
rather than silently degrading linkage, and requires a native
per-architecture build.
Add verify-static-binary.sh, run after every supervisor build in both the
staging script and CI so linkage cannot regress unnoticed for either
variant. It inspects via readelf (or greadelf/llvm-readelf) and fails closed
rather than trusting the tool's exit status: every inspection must produce no
diagnostics, the input must be an executable ELF (ET_EXEC, or ET_DYN with
DF_1_PIE) whose PT_LOAD segments all lie within the file, whose dynamic table
agrees with PT_DYNAMIC, and which carries no PT_INTERP and no DT_NEEDED. That
rejects a dynamically linked, truncated, corrupt, non-ELF, or shared-object
input that naive parsing would misread as static. Hosts without any inspector
(e.g. macOS, which ships no binutils) skip with a warning; Linux, including
CI, requires one and fails closed.
No image or release workflow builds the glibc-static variant, so add a
dedicated supervisor-static-validate workflow that builds it on both
architectures and runs the verifier. rust-native-build.yml uses self-hosted
runners, which reject pull_request-triggered jobs, so it validates in the merge
queue and on pushes to main that touch the build inputs, plus a nightly
schedule, so the GNU + crt-static build branch cannot regress unnoticed.
The default is unchanged, so image, release, and CI behavior is identical.
Selecting glibc-static statically links LGPL glibc into a redistributed
binary, which is why it is opt-in.
Signed-off-by: Mrunal Patel <mrunalp@gmail.com>
Co-authored-by: Mrunal Patel <mrunalp@gmail.com>
Remove the production-crate dev-settings feature so e2e tests build the same gateway, CLI, and supervisor binaries that we ship. Move settings coverage from dummy test-only keys to the production ocsf_json_enabled setting.
Signed-off-by: Simon Scatton <sscatton@nvidia.com>
* wip
* refactor(docker): use native rust builds for split gateway/supervisor images
Drop the in-Docker BUILD_FROM_SOURCE path so both images consume only
prebuilt binaries staged natively via tasks/scripts/stage-prebuilt-binaries.sh.
This mirrors what CI does and reuses the host's cargo target cache and
sccache across rebuilds.
- Dockerfile.gateway: nvcr.io/nvidia/distroless/cc:v4.0.4 base (the 4.0.0
tag does not exist on nvcr.io; the registry uses a v prefix). GNU-linked
binary copied to /usr/local/bin.
- Dockerfile.supervisor: scratch base, static musl binary. Static linkage
lets the image stay scratch while still being executable as a Kubernetes
init container.
- skaffold.yaml: each artifact invokes tasks/scripts/docker-build-image.sh,
which stages the binary natively (cargo / cargo-zigbuild) and then builds
the image. Drops the cross-build.sh dependency from the supervisor build.
- seccomp.rs: add a local SYS_kexec_file_load constant for musl/aarch64.
libc 0.2.185 omits the symbol from its musl/aarch64 bindings, so the
supervisor's seccomp filter previously failed to compile for that target.
- architecture/build.md: describe the native-first pipeline and per-image
runtime choices.
Local validation: gateway image 101MB (was 194MB), supervisor image 21.7MB.
helm:skaffold:run deploys cleanly; the static musl supervisor binary runs
correctly in a non-glibc agent container.
* refactor(docker): tighten binary perms via --chown + 0550
Replace `COPY --chmod=755` with `COPY --chown=<user> --chmod=0550` in
the gateway and supervisor Dockerfiles. The binary is no longer
world-readable or world-executable; ownership is pinned to the runtime
user.
- Gateway uses `--chown=nvs:nvs` + `USER nvs:nvs`, matching the only
non-root user defined in `nvcr.io/nvidia/distroless/cc` (UID 1000) and
the Helm chart's `securityContext.runAsUser: 1000`, which overrides
the Dockerfile USER at runtime.
- Supervisor uses numeric `--chown=65534:65534` because the scratch base
has no `/etc/passwd` for name resolution. The supervisor image is
only consumed by the init-container copy-self path; the destination
pod's runAsUser governs execute access.
Validated by deploying to a local k3d cluster via `helm:skaffold:run`
and confirming the gateway StatefulSet reaches 1/1 Running.
* ci(gpu): repoint GPU probe image lookup at Dockerfile.gateway
The previous awk parsed `FROM <image> AS gateway` from the now-deleted
`Dockerfile.images`. The new `Dockerfile.gateway` uses an ARG with a
default (`ARG GATEWAY_BASE_IMAGE=nvcr.io/nvidia/distroless/cc:v4.0.4`)
and `FROM ${GATEWAY_BASE_IMAGE} AS gateway`, so the old script returns
nothing.
Parse the ARG default value directly so the GPU prerequisites check
keeps using the gateway base image as a `nvidia-smi` probe target.
* ci(gpu): pin GPU probe to nvcr.io/nvidia/base/ubuntu:noble
The previous probe parsed the gateway base image out of the Dockerfile,
relying on the fact that the gateway ran on `nvcr.io/nvidia/base/ubuntu`
and that NVIDIA Container Toolkit CDI injection would populate
`nvidia-smi` and the supporting libs at runtime. The new gateway base
(`nvcr.io/nvidia/distroless/cc`) lacks `ldconfig`, a populated
`/usr/bin`, and the broader filesystem layout CDI injection assumes,
so it cannot serve as a GPU probe.
Pin the probe image explicitly to the NVIDIA-managed Ubuntu base. The
probe is independent of the gateway runtime and survives future base
swaps.
* fix(e2e-gpu): pass GPU probe image via env, drop Dockerfile.images parse
The Rust e2e test `gpu_request_for_each_discovered_device_matches_plain_container`
was still parsing the deleted `Dockerfile.images` to derive its GPU probe
image, panicking with `No such file or directory` after this PR's
Dockerfile split.
Move the probe image to a single source of truth in the workflow
(`OPENSHELL_E2E_GPU_PROBE_IMAGE` env at the job level) and require the
e2e test to read it from there. No silent codebase default — the test
panics with a pointer at the workflow if the env is missing, so it
fails loudly rather than drifting from CI.
The prereq probe step in the workflow now consumes the same env, so
the probe image is declared exactly once.
* fix(docker): keep supervisor binary root-owned for rootless Podman
The Podman driver mounts the supervisor image read-only into the sandbox
container at /opt/openshell/bin and runs that container as UID 0, but
deliberately drops DAC_OVERRIDE for hardening (container.rs:419). With
--chown=65534:65534 --chmod=0550 the binary was r-xr-x--- owned by UID
65534, so the container's UID 0 fell into "other" with no read or exec
access and the supervisor crashed on start (ContainerExited code 1).
Docker and Kubernetes both retain DAC_OVERRIDE, so root could still
exec the file — which is why this regression only surfaced in the
Podman e2e job.
Drop --chown from the supervisor COPY so the binary stays root-owned.
Keep --chmod=0550: the security win was dropping world-execute, not
changing the owner. The chown bought nothing here because the container
is always UID 0 regardless of driver, but it actively broke the only
driver that drops DAC_OVERRIDE.
---------
Co-authored-by: Drew Newberry <anewberry@nvidia.com>
Simon Scatton