ci: fix the gate to assert the code, not the runner image

kv named the wrong server binary (kv-server, not valkey-server) and assumed a C
compiler the reusable installs only for Go callers; forum assumed a ruby the
reusable never provisions; rocm-rs left gpu-sort on, whose nested nightly
build-std step no stock runner can do; insights-rs asked for --all-features,
which selects the live-endpoint e2e feature.

Co-authored-by: Hanzo Dev <[email protected]>
This commit is contained in:
2026-08-06 02:39:59 -07:00
co-authored by Hanzo Dev
parent 9260fbc462
commit 5626ba6e6a
+7 -1
View File
@@ -6,8 +6,14 @@
# what CI proves is that the Rust in this repo compiles against the bindings this # what CI proves is that the Rust in this repo compiles against the bindings this
# repo ships. Regenerating them needs ROCm headers and an AMD host; that is a # repo ships. Regenerating them needs ROCm headers and an AMD host; that is a
# hardware lane, not a push gate. # hardware lane, not a push gate.
#
# Features are the default set minus `gpu-sort`, which Cargo.toml documents as
# the one to turn off where "the kernel's nested nightly+build-std cargo build is
# unavailable" — a stock runner is precisely that host. Leaving it on would make
# this gate report on the absence of a nightly amdgcn toolchain rather than on
# the code.
test: test:
- name: build - name: build
run: | run: |
set -euo pipefail set -euo pipefail
SKIP_BINDGEN=1 cargo build --locked --all-features SKIP_BINDGEN=1 cargo build --locked --no-default-features --features macros,miopen