Skip to content

Build TensorFlow Lite targets in one Bazel batch invocation - #63

Merged
Koen Kooi (koenkooi) merged 1 commit into
mainfrom
koenkooi-studious-garbanzo
Oct 8, 2026
Merged

Koen Kooi (koenkooi) merged 1 commit into
mainfrom
koenkooi-studious-garbanzo

Conversation

@koenkooi

Copy link
Copy Markdown
Contributor

The arm64 TensorFlow Lite build intermittently stalls under QEMU after Bazel reports "no actions running", causing CI to hit the container build timeout.

Build all five targets in a single bazel --batch build invocation to avoid the persistent Bazel server, remove the duplicate GPU delegate build, and propagate failures instead of masking them with semicolon-separated commands.

This is a proposed workaround, not a confirmed fix. Several consecutive successful CI builds are needed to establish whether it prevents the intermittent stall. Diff and RUN shell syntax checks passed; the full container build has not been run locally.

Refs #62.

@koenkooi
Koen Kooi (koenkooi) force-pushed the koenkooi-studious-garbanzo branch from c412a45 to 11db312 Compare October 8, 2026 04:45
Avoid the persistent Bazel server under QEMU, remove the duplicate GPU delegate build, and propagate build failures.

Signed-off-by: Koen Kooi <koen.kooi@oss.qualcomm.com>
Co-authored-by: Copilot App <223556219+Copilot@users.noreply.github.com>
@koenkooi
Koen Kooi (koenkooi) force-pushed the koenkooi-studious-garbanzo branch from 11db312 to ea09809 Compare October 8, 2026 04:46
@koenkooi
Koen Kooi (koenkooi) merged commit 76bf666 into main Oct 8, 2026
15 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant