Fix panic when err is nil on retry for k8s executor
What does this MR do?
Fix a panic with the executorkubernetes when the err returns after the script execution is nil.
When err is nil, GitLab Runner panics and exits with the following stacktrace
panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x2 addr=0x0 pc=0x102b18c9c]
goroutine 167 [running]:
gitlab.com/gitlab-org/gitlab-runner/executors/kubernetes.(*executor).checkScriptExecution.func1({0x102b8b89c, 0x15})
/Users/ratchade/projects/main-runner/executors/kubernetes/kubernetes.go:2533 +0xbc
slices.IndexFunc[...]({0x104795980, 0xd, 0xd}, 0x14000f5b948)
/opt/homebrew/Cellar/go/1.22.4/libexec/src/slices/slices.go:107 +0x94
slices.ContainsFunc[...]({0x104795980, 0xd, 0xd}, 0x14000f5b948)
/opt/homebrew/Cellar/go/1.22.4/libexec/src/slices/slices.go:122 +0x54
gitlab.com/gitlab-org/gitlab-runner/executors/kubernetes.(*executor).checkScriptExecution(0x1400140ea08, {0x102b79afd, 0xe}, {0x0, 0x0})
/Users/ratchade/projects/main-runner/executors/kubernetes/kubernetes.go:2529 +0xb8
gitlab.com/gitlab-org/gitlab-runner/executors/kubernetes.(*executor).runInContainer.func1.1()
/Users/ratchade/projects/main-runner/executors/kubernetes/kubernetes.go:2443 +0x210
gitlab.com/gitlab-org/gitlab-runner/helpers/retry.RunFunc.toValueFunc.func1()
/Users/ratchade/projects/main-runner/helpers/retry/retry.go:44 +0x34
gitlab.com/gitlab-org/gitlab-runner/helpers/retry.retryRun[...](0x1400000e138, 0x14000c0ca30)
/Users/ratchade/projects/main-runner/helpers/retry/retry.go:173 +0x64
gitlab.com/gitlab-org/gitlab-runner/helpers/retry.(*NoValueRetry).Run(0x14000cb6100)
/Users/ratchade/projects/main-runner/helpers/retry/retry.go:185 +0x64
gitlab.com/gitlab-org/gitlab-runner/executors/kubernetes.(*executor).runInContainer.func1()
/Users/ratchade/projects/main-runner/executors/kubernetes/kubernetes.go:2446 +0x31c
created by gitlab.com/gitlab-org/gitlab-runner/executors/kubernetes.(*executor).runInContainer in goroutine 8
/Users/ratchade/projects/main-runner/executors/kubernetes/kubernetes.go:2424 +0x1f4
Why was this MR needed?
Prevent GitLab Runner from panic-ing
What's the best way to test this MR?
gitlab-ci
variables:
FF_USE_POWERSHELL_PATH_RESOLVER: "true"
FF_RETRIEVE_POD_WARNING_EVENTS: "true"
FF_PRINT_POD_EVENTS: "true"
FF_SCRIPT_SECTIONS: "true"
CI_DEBUG_SERVICES: "true"
GIT_DEPTH: 5
MY_TEST_VARIABLE_1: gitlab-ci
MY_TEST_VARIABLE_2: gitlab-ci
simple-job:
script:
- echo $MY_TEST_VARIABLE_1
- echo $MY_TEST_VARIABLE_2
after_script:
- echo "this is the after_script running"config.toml
concurrent = 1
check_interval = 1
log_level = "debug"
shutdown_timeout = 0
listen_address = ':9252'
[session_server]
session_timeout = 1800
[[runners]]
name = "investigation"
url = "https://gitlab.com/"
id = 0
token = "glrt_REDACTED"
token_obtained_at = "0001-01-01T00:00:00Z"
token_expires_at = "0001-01-01T00:00:00Z"
executor = "kubernetes"
shell = "bash"
limit = 1
[runners.kubernetes]
host = ""
bearer_token_overwrite_allowed = false
image = "alpine"
pod_termination_grace_period_seconds = 0
namespace = ""
namespace_overwrite_allowed = ""
pod_labels_overwrite_allowed = ""
service_account_overwrite_allowed = ""
pod_annotations_overwrite_allowed = ""
node_selector_overwrite_allowed = ".*"
allow_privilege_escalation = false
[[runners.kubernetes.services]]
[runners.kubernetes.dns_config]
[runners.kubernetes.pod_labels]
user = "ratchade"With the main branch, this job will hang as GitLab Runner exit with error.
With the MR branch, the job succeeds as seen in the job log