starred/act

Fork 0

mirror of https://github.com/nektos/act.git synced 2026-04-26 09:25:54 +03:00

[PR #5782] Fix segmentation fault in dry-run mode with service containers #2604

New issue

Open

opened 2026-03-01 22:37:24 +03:00 by kerem · 0 comments

kerem commented

2026-03-01 22:37:24 +03:00

Owner

📋 Pull Request Information

Original PR: https://github.com/nektos/act/pull/5782
Author: @TheFynx
Created: 6/21/2025
Status: 🔄 Open

Base: master ← Head: 2607-services-dry-run

📝 Commits (4)

d1314b3 fixed issue with service container dry-run usage
c8becd3 cleanup test
3b4ce3a Merge branch 'master' into 2607-services-dry-run
d704c91 Merge branch 'nektos:master' into 2607-services-dry-run

📊 Changes

3 files changed (+75 additions, -1 deletions)

View changed files

📝 pkg/container/docker_run.go (+19 -1)
📝 pkg/runner/run_context.go (+8 -0)
➕ pkg/runner/run_context_dryrun_test.go (+48 -0)

📄 Description

Description

Fixes segmentation fault that occurs when running act -n (dry-run mode) with GitHub Actions workflows containing service containers.

Fixes #2607

Problem

When running workflows with service containers in dry-run mode, Act would crash with a segmentation fault at pkg/container/docker_run.go:173 in the GetHealth() method. This happened because:

In dry-run mode, service containers are not actually created (no Docker client connection)
The health checking code still attempted to inspect non-existent containers via cr.cli.ContainerInspect(ctx, cr.id)
This resulted in a nil pointer dereference when cr.cli was nil

Solution

Added dry-run mode protection at two critical points:

GetHealth() method - Returns HealthHealthy immediately in dry-run mode instead of attempting Docker inspection
waitForServiceContainer() function - Skips all health checking in dry-run mode

The fix ensures that act's dry-run mode works consistently with service containers by providing appropriate mock responses instead of attempting real Docker operations.

Testing

Added regression test TestServiceContainerWorkflowDryRun to prevent future regressions
Manually validated that act -n with service containers no longer crashes
All existing service container tests continue to pass:
- TestRunEvent/services
- TestRunEvent/services-empty-image
- TestRunEvent/services-host-network
- TestRunEvent/services-with-container
- TestRunEvent/mysql-service-container-with-health-check

Compliance

Code formatted with go fmt ./...
Ran go mod tidy (why go.mod was altered)
Includes test coverage of change

_{🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.}

## 📋 Pull Request Information **Original PR:** https://github.com/nektos/act/pull/5782 **Author:** [@TheFynx](https://github.com/TheFynx) **Created:** 6/21/2025 **Status:** 🔄 Open **Base:** `master` ← **Head:** `2607-services-dry-run` --- ### 📝 Commits (4) - [`d1314b3`](https://github.com/nektos/act/commit/d1314b32526378ca45fead6df8193f0ed4261007) fixed issue with service container dry-run usage - [`c8becd3`](https://github.com/nektos/act/commit/c8becd3717f3ef51fc94b4b2d0d1b93ed43cb4ee) cleanup test - [`3b4ce3a`](https://github.com/nektos/act/commit/3b4ce3a266200f6f7871b3c334b59e733583d4aa) Merge branch 'master' into 2607-services-dry-run - [`d704c91`](https://github.com/nektos/act/commit/d704c91a46e36c56291ed4f0c6352386ae33fc76) Merge branch 'nektos:master' into 2607-services-dry-run ### 📊 Changes **3 files changed** (+75 additions, -1 deletions) <details> <summary>View changed files</summary> 📝 `pkg/container/docker_run.go` (+19 -1) 📝 `pkg/runner/run_context.go` (+8 -0) ➕ `pkg/runner/run_context_dryrun_test.go` (+48 -0) </details> ### 📄 Description ## Description Fixes segmentation fault that occurs when running `act -n` (dry-run mode) with GitHub Actions workflows containing service containers. Fixes #2607 ## Problem When running workflows with service containers in dry-run mode, Act would crash with a segmentation fault at `pkg/container/docker_run.go:173` in the `GetHealth()` method. This happened because: 1. In dry-run mode, service containers are not actually created (no Docker client connection) 2. The health checking code still attempted to inspect non-existent containers via `cr.cli.ContainerInspect(ctx, cr.id)` 3. This resulted in a nil pointer dereference when `cr.cli` was nil ## Solution Added dry-run mode protection at two critical points: 1. **`GetHealth()` method** - Returns `HealthHealthy` immediately in dry-run mode instead of attempting Docker inspection 2. **`waitForServiceContainer()` function** - Skips all health checking in dry-run mode The fix ensures that act's dry-run mode works consistently with service containers by providing appropriate mock responses instead of attempting real Docker operations. ## Testing - Added regression test `TestServiceContainerWorkflowDryRun` to prevent future regressions - Manually validated that `act -n` with service containers no longer crashes - All existing service container tests continue to pass: - `TestRunEvent/services` - `TestRunEvent/services-empty-image` - `TestRunEvent/services-host-network` - `TestRunEvent/services-with-container` - `TestRunEvent/mysql-service-container-with-health-check` ## Compliance - [x] Code formatted with `go fmt ./...` - [x] Ran `go mod tidy` (why go.mod was altered) - [x] Includes test coverage of change --- <sub>🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.</sub>

kerem added the

pull-request

label

2026-03-01 22:37:24 +03:00

No milestone

No project

No assignees

1 participant

Notifications

Due date

The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference

starred/act#2604

No description provided.

Rows
Columns