WINA-1275: add agent_loaded_modules.json with module metadata and fla…

[gouri-yerra]

What does this PR do?

Adds a new Windows-only flare artifact, agent_loaded_modules.json, which shows the Agent’s loaded DLLs along with relevant metadata. The existing _open_files.txt file still exists for backward compatibility.

For each loaded module, the flare includes:
• Full path and module name
• File timestamp, size, and permissions
• Windows version info fields (CompanyName, ProductName, FileVersion, ProductVersion, OriginalFilename, InternalName)

Implementation details:
• Introduces a helper for reading Windows version info strings (pkg/util/winutil)
• Adds a JSON builder for loaded modules (pkg/util/lsof)
• Wires this into the Windows lsof flare provider (comp/core/lsof/impl)

Motivation

• Addresses WINA-1275: the current Windows flare output is misnamed and not machine-parsable, making it harder for support to look at loaded modules.
• Unblocks WINA-1274 by providing a structured view that can be matched against known antimalware/interference lists (e.g., Panda, SentinelOne, Carbon Black).

Describe how you validated your changes

With Roberto
Done:

• Build Agent locally on Windows vm and make multiple flares to validate
• check if agent_loaded_modules.json is included in the flare output and existing _open_files.txt is the same for backward compatibility.
  Verify JSON schema:
• generated_at timestamp present and correctly populated.
• modules is an array with expected fields (path, name, size, timestamps, permissions, and version info).
• Try without embedded Windows version resources:
• DLLs w/o versioninfo are version fields empty/null
• No flare gen fails or partial output.
  Check scope
• Logic executes only during flare creation; no impact to normal Agent runtime or startup.
• Non-Windows builds are unaffected

agent_loaded_modules.json

Additional Notes

•This PR intentionally limits scope to Windows version info fields. Other details can be added incrementally in a follow-up PR
• The logic only runs when generating a flare; normal Agent runtime is unaffected.
• The change is additive and backward compatible with existing flare consumers.

[agent-platform-auto-pr]

Static quality checks

✅ Please find below the results from static quality gates
Comparison made with ancestor 969d924
📊 Static Quality Gates Dashboard

Successful checks

Info

	Quality gate	Change	Size (prev → curr → max)
✅	agent_deb_amd64	-17.4 KiB (0.00% reduction)	705.290 → 705.273 → 708.410
✅	agent_deb_amd64_fips	-17.4 KiB (0.00% reduction)	700.567 → 700.550 → 704.000
✅	agent_heroku_amd64	-17.4 KiB (0.01% reduction)	326.933 → 326.916 → 329.530
✅	agent_msi	-3.49 KiB (0.00% reduction)	571.545 → 571.541 → 982.080
✅	agent_rpm_amd64	-17.4 KiB (0.00% reduction)	705.276 → 705.259 → 708.380
✅	agent_rpm_amd64_fips	-17.4 KiB (0.00% reduction)	700.554 → 700.537 → 703.990
✅	agent_suse_amd64	-17.4 KiB (0.00% reduction)	705.276 → 705.259 → 708.380
✅	agent_suse_amd64_fips	-17.4 KiB (0.00% reduction)	700.554 → 700.537 → 703.990
✅	docker_agent_amd64	-34.85 KiB (0.00% reduction)	767.485 → 767.451 → 770.720
✅	docker_agent_arm64	-18.87 KiB (0.00% reduction)	773.621 → 773.602 → 780.200
✅	docker_agent_jmx_amd64	-34.85 KiB (0.00% reduction)	958.364 → 958.329 → 961.600
✅	docker_agent_jmx_arm64	-18.86 KiB (0.00% reduction)	953.219 → 953.200 → 959.800
✅	docker_cluster_agent_amd64	-17.45 KiB (0.01% reduction)	180.715 → 180.698 → 181.080
✅	docker_cluster_agent_arm64	-17.44 KiB (0.01% reduction)	196.557 → 196.540 → 198.490

17 successful checks with minimal change (< 2 KiB)

	Quality gate	Current Size
✅	agent_rpm_arm64	686.832 MiB
✅	agent_rpm_arm64_fips	682.938 MiB
✅	agent_suse_arm64	686.832 MiB
✅	agent_suse_arm64_fips	682.938 MiB
✅	docker_cws_instrumentation_amd64	7.135 MiB
✅	docker_cws_instrumentation_arm64	6.689 MiB
✅	docker_dogstatsd_amd64	38.808 MiB
✅	docker_dogstatsd_arm64	37.128 MiB
✅	dogstatsd_deb_amd64	30.027 MiB
✅	dogstatsd_deb_arm64	28.176 MiB
✅	dogstatsd_rpm_amd64	30.027 MiB
✅	dogstatsd_suse_amd64	30.027 MiB
✅	iot_agent_deb_amd64	42.987 MiB
✅	iot_agent_deb_arm64	40.112 MiB
✅	iot_agent_deb_armhf	40.689 MiB
✅	iot_agent_rpm_amd64	42.987 MiB
✅	iot_agent_suse_amd64	42.987 MiB

On-wire sizes (compressed)

	Quality gate	Change	Size (prev → curr → max)
✅	agent_deb_amd64	-41.69 KiB (0.02% reduction)	173.329 → 173.288 → 174.490
✅	agent_deb_amd64_fips	-19.11 KiB (0.01% reduction)	172.252 → 172.233 → 173.750
✅	agent_heroku_amd64	+6.81 KiB (0.01% increase)	87.115 → 87.122 → 88.450
✅	agent_msi	-36.0 KiB (0.02% reduction)	142.945 → 142.910 → 143.020
✅	agent_rpm_amd64	+30.89 KiB (0.02% increase)	176.137 → 176.167 → 177.660
✅	agent_rpm_amd64_fips	-22.07 KiB (0.01% reduction)	174.930 → 174.908 → 176.600
✅	agent_rpm_arm64	-50.71 KiB (0.03% reduction)	159.358 → 159.308 → 161.260
✅	agent_rpm_arm64_fips	-21.75 KiB (0.01% reduction)	158.770 → 158.749 → 160.550
✅	agent_suse_amd64	+30.89 KiB (0.02% increase)	176.137 → 176.167 → 177.660
✅	agent_suse_amd64_fips	-22.07 KiB (0.01% reduction)	174.930 → 174.908 → 176.600
✅	agent_suse_arm64	-50.71 KiB (0.03% reduction)	159.358 → 159.308 → 161.260
✅	agent_suse_arm64_fips	-21.75 KiB (0.01% reduction)	158.770 → 158.749 → 160.550
✅	docker_agent_amd64	+4.51 KiB (0.00% increase)	261.048 → 261.052 → 262.450
✅	docker_agent_arm64	+12.05 KiB (0.00% increase)	250.048 → 250.060 → 252.630
✅	docker_agent_jmx_amd64	+13.97 KiB (0.00% increase)	329.677 → 329.691 → 331.080
✅	docker_agent_jmx_arm64	+10.69 KiB (0.00% increase)	314.668 → 314.678 → 317.270
✅	docker_cluster_agent_amd64	+11.82 KiB (0.02% increase)	63.835 → 63.847 → 64.490
✅	docker_cluster_agent_arm64	+8.7 KiB (0.01% increase)	60.121 → 60.130 → 61.170
✅	docker_cws_instrumentation_amd64	neutral	2.994 MiB
✅	docker_cws_instrumentation_arm64	neutral	2.726 MiB
✅	docker_dogstatsd_amd64	neutral	15.026 MiB
✅	docker_dogstatsd_arm64	neutral	14.354 MiB
✅	dogstatsd_deb_amd64	neutral	7.945 MiB
✅	dogstatsd_deb_arm64	neutral	6.824 MiB
✅	dogstatsd_rpm_amd64	neutral	7.956 MiB
✅	dogstatsd_suse_amd64	neutral	7.956 MiB
✅	iot_agent_deb_amd64	neutral	11.258 MiB
✅	iot_agent_deb_arm64	neutral	9.630 MiB
✅	iot_agent_deb_armhf	+2.6 KiB (0.03% increase)	9.823 → 9.825 → 10.620
✅	iot_agent_rpm_amd64	-3.09 KiB (0.03% reduction)	11.279 → 11.276 → 12.060
✅	iot_agent_suse_amd64	-3.09 KiB (0.03% reduction)	11.279 → 11.276 → 12.060

[cit-pr-commenter]

Regression Detector

Regression Detector Results

Metrics dashboard
Target profiles
Run ID: 86e70328-4c45-4282-bd46-355db7bd48b4

Baseline: 969d924
Comparison: 36c8a87
Diff

Optimization Goals: ✅ No significant changes detected

Experiments ignored for regressions

Regressions in experiments with settings containing erratic: true are ignored.

perf	experiment	goal	Δ mean %	Δ mean % CI	trials	links
➖	docker_containers_cpu	% cpu utilization	+0.65	[-2.28, +3.59]	1	Logs

Fine details of change detection per experiment

perf	experiment	goal	Δ mean %	Δ mean % CI	trials	links
➖	quality_gate_metrics_logs	memory utilization	+2.36	[+2.14, +2.57]	1	Logs bounds checks dashboard
➖	quality_gate_logs	% cpu utilization	+0.93	[-0.54, +2.40]	1	Logs bounds checks dashboard
➖	ddot_metrics_sum_cumulativetodelta_exporter	memory utilization	+0.73	[+0.50, +0.96]	1	Logs
➖	docker_containers_cpu	% cpu utilization	+0.65	[-2.28, +3.59]	1	Logs
➖	file_tree	memory utilization	+0.12	[+0.07, +0.17]	1	Logs
➖	docker_containers_memory	memory utilization	+0.09	[+0.02, +0.16]	1	Logs
➖	ddot_metrics_sum_cumulative	memory utilization	+0.09	[-0.08, +0.25]	1	Logs
➖	quality_gate_idle	memory utilization	+0.06	[+0.01, +0.11]	1	Logs bounds checks dashboard
➖	file_to_blackhole_1000ms_latency	egress throughput	+0.05	[-0.36, +0.46]	1	Logs
➖	file_to_blackhole_0ms_latency	egress throughput	+0.02	[-0.51, +0.56]	1	Logs
➖	file_to_blackhole_100ms_latency	egress throughput	+0.01	[-0.03, +0.05]	1	Logs
➖	quality_gate_idle_all_features	memory utilization	+0.01	[-0.03, +0.04]	1	Logs bounds checks dashboard
➖	uds_dogstatsd_to_api_v3	ingress throughput	+0.00	[-0.13, +0.13]	1	Logs
➖	uds_dogstatsd_to_api	ingress throughput	+0.00	[-0.14, +0.14]	1	Logs
➖	tcp_dd_logs_filter_exclude	ingress throughput	+0.00	[-0.09, +0.09]	1	Logs
➖	file_to_blackhole_500ms_latency	egress throughput	-0.05	[-0.44, +0.34]	1	Logs
➖	otlp_ingest_metrics	memory utilization	-0.10	[-0.25, +0.06]	1	Logs
➖	ddot_logs	memory utilization	-0.27	[-0.33, -0.21]	1	Logs
➖	otlp_ingest_logs	memory utilization	-0.30	[-0.41, -0.19]	1	Logs
➖	uds_dogstatsd_20mb_12k_contexts_20_senders	memory utilization	-0.30	[-0.35, -0.25]	1	Logs
➖	tcp_syslog_to_blackhole	ingress throughput	-0.52	[-0.61, -0.43]	1	Logs
➖	ddot_metrics_sum_delta	memory utilization	-0.57	[-0.78, -0.36]	1	Logs
➖	ddot_metrics	memory utilization	-0.62	[-0.85, -0.38]	1	Logs

Bounds Checks: ✅ Passed

perf	experiment	bounds_check_name	replicates_passed	links
✅	docker_containers_cpu	simple_check_run	10/10
✅	docker_containers_memory	memory_usage	10/10
✅	docker_containers_memory	simple_check_run	10/10
✅	file_to_blackhole_0ms_latency	lost_bytes	10/10
✅	file_to_blackhole_0ms_latency	memory_usage	10/10
✅	file_to_blackhole_1000ms_latency	lost_bytes	10/10
✅	file_to_blackhole_1000ms_latency	memory_usage	10/10
✅	file_to_blackhole_100ms_latency	lost_bytes	10/10
✅	file_to_blackhole_100ms_latency	memory_usage	10/10
✅	file_to_blackhole_500ms_latency	lost_bytes	10/10
✅	file_to_blackhole_500ms_latency	memory_usage	10/10
✅	quality_gate_idle	intake_connections	10/10	bounds checks dashboard
✅	quality_gate_idle	memory_usage	10/10	bounds checks dashboard
✅	quality_gate_idle_all_features	intake_connections	10/10	bounds checks dashboard
✅	quality_gate_idle_all_features	memory_usage	10/10	bounds checks dashboard
✅	quality_gate_logs	intake_connections	10/10	bounds checks dashboard
✅	quality_gate_logs	lost_bytes	10/10	bounds checks dashboard
✅	quality_gate_logs	memory_usage	10/10	bounds checks dashboard
✅	quality_gate_metrics_logs	cpu_usage	10/10	bounds checks dashboard
✅	quality_gate_metrics_logs	intake_connections	10/10	bounds checks dashboard
✅	quality_gate_metrics_logs	lost_bytes	10/10	bounds checks dashboard
✅	quality_gate_metrics_logs	memory_usage	10/10	bounds checks dashboard

Explanation

Confidence level: 90.00%
Effect size tolerance: |Δ mean %| ≥ 5.00%

Performance changes are noted in the perf column of each table:

• ✅ = significantly better comparison variant performance
• ❌ = significantly worse comparison variant performance
• ➖ = no significant change in performance

A regression test is an A/B test of target performance in a repeatable rig, where "performance" is measured as "comparison variant minus baseline variant" for an optimization goal (e.g., ingress throughput). Due to intrinsic variability in measuring that goal, we can only estimate its mean value for each experiment; we report uncertainty in that value as a 90.00% confidence interval denoted "Δ mean % CI".

For each experiment, we decide whether a change in performance is a "regression" -- a change worth investigating further -- if all of the following criteria are true:

1. Its estimated |Δ mean %| ≥ 5.00%, indicating the change is big enough to merit a closer look.

2. Its 90.00% confidence interval "Δ mean % CI" does not contain zero, indicating that if our statistical model is accurate, there is at least a 90.00% chance there is a difference in performance between baseline and comparison variants.

3. Its configuration does not mark it "erratic".

Replicate Execution Details

We run multiple replicates for each experiment/variant. However, we allow replicates to be automatically retried if there are any failures, up to 8 times, at which point the replicate is marked dead and we are unable to run analysis for the entire experiment. We call each of these attempts at running replicates a replicate execution. This section lists all replicate executions that failed due to the target crashing or being oom killed.

Note: In the below tables we bucket failures by experiment, variant, and failure type. For each of these buckets we list out the replicate indexes that failed with an annotation signifying how many times said replicate failed with the given failure mode. In the below example the baseline variant of the experiment named experiment_with_failures had two replicates that failed by oom kills. Replicate 0, which failed 8 executions, and replicate 1 which failed 6 executions, all with the same failure mode.

Experiment	Variant	Replicates	Failure	Logs	Debug Dashboard
experiment_with_failures	baseline	0 (x8) 1 (x6)	Oom killed		Debug Dashboard

The debug dashboard links will take you to a debugging dashboard specifically designed to investigate replicate execution failures.

❌ Retried Profiling Replicate Execution Failures (target internal profiling)

Note: Profiling replicas may still be executing. See the debug dashboard for up to date status.

Experiment	Variant	Replicates	Failure	Debug Dashboard
quality_gate_idle_all_features	baseline	11 (x4)	Oom killed	Debug Dashboard
quality_gate_idle_all_features	comparison	11 (x4)	Oom killed	Debug Dashboard

CI Pass/Fail Decision

✅ Passed. All Quality Gates passed.

• quality_gate_metrics_logs, bounds check memory_usage: 10/10 replicas passed. Gate passed.
• quality_gate_metrics_logs, bounds check lost_bytes: 10/10 replicas passed. Gate passed.
• quality_gate_metrics_logs, bounds check intake_connections: 10/10 replicas passed. Gate passed.
• quality_gate_metrics_logs, bounds check cpu_usage: 10/10 replicas passed. Gate passed.
• quality_gate_idle_all_features, bounds check intake_connections: 10/10 replicas passed. Gate passed.
• quality_gate_idle_all_features, bounds check memory_usage: 10/10 replicas passed. Gate passed.
• quality_gate_logs, bounds check intake_connections: 10/10 replicas passed. Gate passed.
• quality_gate_logs, bounds check memory_usage: 10/10 replicas passed. Gate passed.
• quality_gate_logs, bounds check lost_bytes: 10/10 replicas passed. Gate passed.
• quality_gate_idle, bounds check intake_connections: 10/10 replicas passed. Gate passed.
• quality_gate_idle, bounds check memory_usage: 10/10 replicas passed. Gate passed.

[iglendd]

We do not need pid and process name for each process, they are identical, right? Also are we collecting it for all Agent processes or only core.

[gouri-yerra]

You’re right, they’re the same. I moved process metadata to the report header and removed it from each module entry.
Change (schema):
Report header now has ProcessName/ProcessPID
ModuleEntry no longer includes process fields. And for core processes!

[iglendd]

What size is it"? In memory or on disk (I wonder).

[gouri-yerra]

On-disk file size. I renamed the field to SizeBytes and documented it in the struct tags.

[iglendd]

This is crazy mislabeling. ListOpenFilesFromSelf does not return open files by loaded module files. And files var does not even reveal that (loadModFiles e.g. probably would).

[gouri-yerra]

Ok. I renamed vars and the helper to reflect “loaded modules,” and removed the “files".

[iglendd]

Interesting question. DLL name is in the DLL path and seemingly redundant. But you probably right it is easier to read.

[gouri-yerra]

Ok! I removed DLLName from the JSON

[iglendd]

How is the time reported for other Flare artifacts. Is it the same way or different? Here and lower in the file.

[gouri-yerra]

Using UTC RFC3339, which matches other flare sections. I added a short comment too.

[iglendd]

I wonder would we even need perm. @clarkb7?

[iglendd]

Why not a plain function?

[gouri-yerra]

Good point. I took out the inner closure to a small helper so it’s reusable and clearer.

[iglendd]

Also looks like some of the information also retrieved slightly differently in https://github.com/DataDog/datadog-agent/pull/44786/changes#diff-111b15151741e55b83953b0b56365d0ff53083208a1df35aca7108e82d8d4a26R113. Potentially there are other Live Process code may benefit from the refactoring in GetFileDescription (DataDog\datadog-agent\pkg\util\winutil\process.go)

[gouri-yerra]

I added a small shared helper in winutil and switched both call sites.