subsys: net: lib: dns: resolve Fix/dns resolve cleanup inactive state

[zafersn]

dns_resolve_close() fails to clean up DNS server slots when the context state is DNS_RESOLVE_CONTEXT_INACTIVE, leaving stale server configuration data that causes subsequent dns_resolve_init() calls to incorrectly detect "server already exists" and skip socket creation, breaking DNS resolution.

#99089

Signed-off-by: Zafer SEN [email protected]

[sonarqubecloud]

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud

[zafersn]

Closing PR: Root Cause Was Application-Level

After thorough debugging, we've determined that this PR is not needed. The issue we encountered was caused by our application code directly manipulating the DNS context internal state, which is not the intended usage of the DNS resolver API.

What happened:

Our modem driver was setting dnsCtx->state = DNS_RESOLVE_CONTEXT_INACTIVE directly during modem power cycles, bypassing the normal cleanup flow. This left resources (socket dispatcher registrations) in an inconsistent state.

Proper fix:

The correct approach is to:

Call dns_resolve_close() before shutting down the modem/socket
Never directly modify internal dns_resolve_context state fields
Let the DNS resolver manage its own state machine
After fixing our application code to properly use the DNS resolver API, the issue is resolved without any changes needed to Zephyr.

Apologies:

Sorry for the unnecessary PR. This was a case of improper API usage on our end, not a Zephyr bug. We've learned to always respect the abstraction boundaries and use the provided APIs correctly.

Closing this PR.