Summary

Fix the kernel build helper scripts so Docker build failures are not masked by the following cleanup step.

Currently, if docker run fails, the scripts still run docker rm kernelbuild and can continue into package movement and repository update steps. In a forced-failure test this allowed the script to finish with exit code 0, even
though no package was built.

This change adds a small run_kernelbuild() wrapper that:

• runs the selected Docker build image
• stores the original docker run exit status
• removes the kernelbuild container
• returns the original build status

With set -euo pipefail, a failed Docker build now stops the script before package movement or repo update.

Changed scripts:

• script.sh
• script-v3-v4.sh
• script-znver4.sh

Note about set -euo pipefail

This enables stricter Bash error handling for the build helper scripts:

• set -e: stop the script when a command returns a non-zero exit code
• set -u: fail on unset variables instead of silently expanding them to an empty string
• set -o pipefail: make pipelines fail if any command in the pipeline fails, not only the last command

The important behavior for this fix is set -e: run_kernelbuild() returns the original docker run exit code, so a failed Docker build now stops the script before package movement or repo-manage-util update.