[santhoshg6]

• Publish client disassociation stats over EM bus
• Include traffic counters and reason code in JSON payload
• Add basic MLO affiliated STA metrics support
• Improve validation and error handling
• Fix return codes and memory handling

[Copilot]

Pull request overview

This PR adds a new EasyMesh (EM) bus event to publish client disassociation statistics, including per-STA traffic counters, the disassociation reason code, and a basic placeholder for Wi-Fi 7 MLO “Affiliated STA Metrics”.

Changes:

• Introduces a new EM bus event name (Device.WiFi.EM.ClientDisassocStats) and registers it in em_init().
• Publishes a JSON payload on STA disassociation, using cached per-STA counters when available.
• Adds initial (placeholder) MLO affiliated STA metrics emission and improves some validation/return codes.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 3 comments.

File	Description
source/apps/em/wifi_em.h	Adds the bus event name constant for client disassociation stats.
source/apps/em/wifi_em.c	Publishes the new disassociation stats event and registers it on the EM bus.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Nikita-Hakai]

Hi @santhoshg6 , could you address the comments.

[dkyncu]

Hi @santhoshg6, the unified-wifi-mesh side of this is up for review:
rdkcentral/unified-wifi-mesh#806. The agent subscribes to Device.WiFi.EM.ClientDisassocStats and sends the Client Disassociation Stats message from it, and the controller publishes DisassociationEvent.Disassociated!. We tested both changes together on a device.

#806 depends on this PR. Could you address the open review comments so it can move forward?

@Nikita-Hakai @mateuszCieslak-GL FYI.

[github-actions]

clang-format suggests the formatting changes below. Use Commit suggestion to apply them.

[Copilot]

Copilot review overview

🟡 Changes recommended

Unresolved moderate issues affect MLO metrics, JSON truncation handling, and function declaration order.

Get a fresh assessment by requesting another Copilot review.

Review effort: Lite
Findings: 1

Open (1)

• MLO disassociation metadata is lost before affiliated metrics check · New

[Copilot]

Copilot review overview

🟡 Changes recommended

Moderate issues remain with compilation, unverified counters, and MLO metric data.

Review effort: Lite
Findings: 2

Open (2)

• Use each link's STA MAC in affiliated MLO entries · New
• MLO disassociation metadata is lost before affiliated metrics check

[github-actions]

clang-format suggests the formatting changes below. Use Commit suggestion to apply them.

[Copilot]

Copilot review overview

🔵 Needs a closer look

Moderate issues remain in MLO data propagation, address serialization, declaration order, and metric schema consistency.

Review effort: Lite
Findings: 2

Open (2)

• Use each link's STA MAC in affiliated MLO entries
• MLO disassociation metadata is lost before affiliated metrics check

[Copilot]

Copilot review overview

🔵 Needs a closer look

Unresolved moderate issues affect compilation and correctness of MLO data and schema output.

Review effort: Lite
Findings: 2

Open (2)

• Use each link's STA MAC in affiliated MLO entries
• MLO disassociation metadata is lost before affiliated metrics check

[Copilot]

Copilot review overview

🔵 Needs a closer look

Unresolved compilation, MLO serialization, schema, and error-propagation issues remain.

Review effort: Lite
Findings: 2

Open (2)

• Use each link's STA MAC in affiliated MLO entries
• MLO disassociation metadata is lost before affiliated metrics check

Previously missed (1)

In code that hasn't changed since last review

MLO affiliated entries use duplicate MLD addresses

source/​apps/​em/​wifi_em.c:1827

Even when MLO metadata is available, this uses the MLD address for every valid link, producing duplicate STA identifiers. cli_LinkInfo[link_idx].cli_LinkAddress is the per-link address already populated/used by the monitor (source/stats/wifi_monitor.c:2166); serialize that address for each affiliated entry instead.

[Copilot]

🔵 Needs a closer look

The reviewed implementation has build-breaking declaration issues and correctness problems in published traffic statistics and schema fields.

No open issues 🎉

_{Review effort: Lite}

[Copilot]

🟡 Changes recommended

Unresolved compilation, concurrency, and error-propagation issues block approval.

1 open issue

• Protect client-type map access from concurrent freeing · New

_{Review effort: Lite}

[Copilot]

🔵 Needs a closer look

Add a forward declaration or move the helper definition before its first use to resolve the compilation issue.

1 open issue

• Protect client-type map access from concurrent freeing

_{Review effort: Lite}

[Copilot]

🔵 Needs a closer look

Two moderate issues remain unresolved, including a potential compilation failure and inconsistent affiliated-STA metric field names.

No open issues 🎉

1 resolved since last review

• Protect client-type map access from concurrent freeing

_{Review effort: Lite}

[Copilot]

🔵 Needs a closer look

Add the missing helper declaration and preserve/report failed event publications instead of losing data and returning success.

No open issues 🎉

_{Review effort: Lite}

[Copilot]

🟡 Changes recommended

Two critical correctness and memory-safety issues, plus a moderate counter-publication issue, remain unresolved.

2 open issues

• MLO disassociation cache removal uses mismatched STA and MLD keys · New
• Client-type cache cleanup dereferences freed iterator after removal · New

_{Review effort: Lite}

[Copilot]

🟡 Changes recommended

The unreachable MLO fallback can publish zeroed statistics because the cached entry is not removed correctly.

1 open issue

• Correct hash_map_remove comment about value ownership · New

2 resolved since last review

• Client-type cache cleanup dereferences freed iterator after removal
• MLO disassociation cache removal uses mismatched STA and MLD keys

_{Review effort: Lite}

[Copilot]

🔵 Needs a closer look

The unresolved MLO cache lookup issue can cause disassociation events to publish incorrect zero counters and discard cached MLO metadata.

No open issues 🎉

1 resolved since last review

• Correct hash_map_remove comment about value ownership

_{Review effort: Lite}

[Copilot]

🟡 Changes recommended

Address the three moderate issues in wifi_em.c before approval.

2 open issues

• Handle reverse-index allocation and insertion failures · New
• Destroy per-VAP reverse-index maps during deinitialization · New

_{Review effort: Lite}

[Copilot]

🟡 Changes recommended

Unresolved memory-management and cache-consistency issues include a critical double-free risk.

2 open issues

• Double-free on hash_map_put failure · New
• Double-free map values during stats cleanup · New

2 resolved since last review

• Destroy per-VAP reverse-index maps during deinitialization
• Handle reverse-index allocation and insertion failures

_{Review effort: Lite}

[Copilot]

🟡 Changes recommended

It contains a compilation error plus unresolved cache ownership, timer teardown, and failure-path memory-safety issues.

3 open issues

• Stop AP report task before destroying report maps · New
• Missing semicolon prevents wifi_em.c from compiling · New
• Double-free on hash_map_put failure

1 resolved since last review

• Double-free map values during stats cleanup

_{Review effort: Lite}

[Copilot]

🔵 Needs a closer look

Six unresolved moderate issues affect memory safety, cache consistency, error handling, and MLO statistics publication.

2 open issues

• Stop AP report task before destroying report maps
• Double-free on hash_map_put failure

1 previously missed issue

In code that hasn't changed since last review

Stale reverse-index entries survive MLO link replacement

source/​apps/​em/​wifi_em.c:400

When the sampled MLO link changes, this removes only the mapping for the new link_key; older link keys for the same primary MLD remain in link_mac_map. After the primary stats entry is removed on one disassociation, a later event using one of those stale keys resolves to a deleted primary key and is published without cached counters. Remove or otherwise invalidate all reverse-index entries that point to the primary key when replacing/removing that client.

This issue also appears on line 1781 of the same file.

1 resolved since last review

• Missing semicolon prevents wifi_em.c from compiling

_{Review effort: Lite}

[Copilot]

🟡 Changes recommended

Unresolved moderate memory-safety, event-publication, cache-consistency, and teardown issues must be addressed before approval.

2 open issues

• Stop AP report task before destroying report maps
• Stale reverse-index entries remain after primary client removal · New

1 resolved since last review

• Double-free on hash_map_put failure

_{Review effort: Lite}

[Copilot]

🔵 Needs a closer look

Unresolved teardown race, double-free risks, and stale or misleading disassociation statistics can affect correctness and stability.

No open issues 🎉

2 resolved since last review

• Stop AP report task before destroying report maps
• Stale reverse-index entries remain after primary client removal

_{Review effort: Lite}

[Copilot]

🟡 Changes recommended

Unresolved critical shutdown deadlock and moderate cache ownership, cleanup, MLO indexing, and client-type consistency issues remain.

1 open issue

• Deadlock from inconsistent teardown and callback lock ordering · New

_{Review effort: Lite}

[Copilot]

🔵 Needs a closer look

Three unresolved moderate correctness and memory-management issues block approval.

No open issues 🎉

1 resolved since last review

• Deadlock from inconsistent teardown and callback lock ordering

_{Review effort: Lite}

[Copilot]

🔵 Needs a closer look

Two unresolved moderate issues affect cache ownership and failure handling in the disassociation statistics implementation.

No open issues 🎉

_{Review effort: Lite}

[santhoshg6]

Hi @Nikita-Hakai

I have addressed all the review comments and updated the PR accordingly. Could you please review it and merge it if everything looks good?