RDKEMW-24270:Avoid race condition between memcr and stop cont

[Sonajeya31]

What does this PR change/fix and why?

This PR fixes a race between container stop and in-progress hibernation.

When a container is being hibernated, MEMCR can still hold an active ptrace/seize on the in-flight PID while Dobby is checkpointing it. If a stop request arrives in that window, the code could attempt to kill the container before MEMCR has fully unwound the process, leading to a crash in the restore/checkpoint path.

This change adds a safe abort path for in-flight hibernation:

detect the PID currently being checkpointed
move the container state from Hibernating to Awakening
stop the hibernation loop from continuing
issue WakeupProcess for the active PID
only then continue with the stop path
This prevents a stop/kill from racing the MEMCR restore/checkpoint flow while the PID is still held under ptrace.

Test Procedure
Refer Ticket

Type of Change

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• Other (doesn't fit into the above categories - e.g. documentation updates)

Requires Bitbake Recipe changes?

• The base Bitbake recipe (meta-rdk-ext/recipes-containers/dobby/dobby.bb) must be modified to support the changes in this PR (beyond updating SRC_REV)

[Copilot]

Copilot review overview

🟡 Changes recommended

The abort path does not reliably stop in-flight checkpointing before the container is killed.

Get a fresh assessment by requesting another Copilot review.

Review effort: Lite
Findings: 1

Open (1)

• Abort Awakening checkpoint before killing the container · New

What changed in this PR

This PR addresses a race between container stopping and in-progress hibernation.

Changes:

• Extends stop handling for awakening containers with active hibernation PIDs.
• Adds diagnostic logging for hibernation abort transitions.

File	Summary
daemon/​lib/​source/​DobbyManager.cpp	Updates hibernation abort handling and logging.

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

[Copilot]

Copilot review overview

🟡 Changes recommended

Container identity must be revalidated after reacquiring the lock to avoid stopping a replacement container.

Review effort: Balanced
Findings: 1

Open (1)

• Revalidate container identity before stopping after lock release · New

Resolved since last review (1)

• Abort Awakening checkpoint before killing the container

[Copilot]

Copilot review overview

🟡 Changes recommended

The stop path can issue a duplicate concurrent restore, and the new race handling lacks regression coverage.

Review effort: Balanced
Findings: 1 · 1 · 2

Open (4)

• Stop can trigger duplicate restore during container wakeup · New
• Add regression coverage for stop during concurrent wakeup · New
• Update dobby.bb recipe or correct PR metadata · New
• Document blocking stop behavior during container wakeup · New

Resolved since last review (1)

• Revalidate container identity before stopping after lock release