BREW-MOODLE-DL-CVE-2025-48945 (GHSA-5QPG-RH4J-QP35)

Vulnerability from osv_homebrew – Published: 2026-08-13 17:14 – Updated: 2026-08-13 17:14 – Source website
VLAI
Summary
pycares has a Use-After-Free Vulnerability
Details

Summary

pycares is vulnerable to a use-after-free condition that occurs when a Channel object is garbage collected while DNS queries are still pending. This results in a fatal Python error and interpreter crash.

Details

Root Cause

The vulnerability stems from improper handling of callback references when the Channel object is destroyed:

  1. When a DNS query is initiated, pycares stores a callback reference using ffi.new_handle()
  2. If the Channel object is garbage collected while queries are pending, the callback references become invalid
  3. When c-ares attempts to invoke the callback, it accesses freed memory, causing a fatal error

This issue was much more likely to occur when using event_thread=True but could happen without it under the right circumstances.

Technical Details

The core issue is a race condition between Python's garbage collector and c-ares's callback execution:

  1. When __del__ is called from within a c-ares callback context, we cannot immediately call ares_destroy() because c-ares is still executing code after the callback returns
  2. c-ares needs to execute cleanup code after our Python callback returns (specifically at lines 1422-1429 in ares_process.c)
  3. If we destroy the channel too quickly, c-ares accesses freed memory

Impact

Applications using pycares can be crashed remotely by triggering DNS queries that result in Channel objects being garbage collected before query completion. This is particularly problematic in scenarios where:

  • Channel objects are created per-request
  • Multiple failed DNS queries are processed rapidly
  • The application doesn't properly manage Channel lifecycle

The error manifests as:

Fatal Python error: b_from_handle: ffi.from_handle() detected that the address passed points to garbage

Fix

The vulnerability has been fixed in pycares 4.9.0 by implementing a safe channel destruction mechanism

Mitigation

For Application Developers

  1. Upgrade to pycares >= 4.9.0 - This version includes the fix and requires no code changes
  2. Best practices (optional but recommended): ```python # Explicit cleanup channel.close()

# Or use context manager with pycares.Channel() as channel: # ... use channel ... # Automatically closed ``` 3. Avoid creating Channel objects per-request - Prefer long-lived instances for better performance and safety

The fix is completely transparent - no API changes or code modifications are required.

Credit

This vulnerability was reported by @vEpiphyte through the aio-libs security program.

References
To clipboard 

{
  "affected": [
    {
      "ecosystem_specific": {
        "fix": "bump",
        "range_state": "fixed",
        "resource": "pycares",
        "resource_purl": "pkg:pypi/pycares@5.0.1",
        "upstream_fixed_in": "4.9.0"
      },
      "package": {
        "ecosystem": "Homebrew",
        "name": "moodle-dl",
        "purl": "pkg:brew/moodle-dl"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "2.3.13_12"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "database_specific": {
    "confidence": "high",
    "source": "matched",
    "strategy": "registry",
    "upstream_evidence": [
      {
        "ecosystem": "PyPI",
        "key": "pkg:pypi/pycares@5.0.1",
        "name": "pycares",
        "resource": "pycares",
        "strategy": "registry",
        "subject_version": "5.0.1"
      }
    ]
  },
  "details": "## Summary\n\npycares is vulnerable to a use-after-free condition that occurs when a Channel object is garbage collected while DNS queries are still pending. This results in a fatal Python error and interpreter crash.\n\n## Details\n\n### Root Cause\n\nThe vulnerability stems from improper handling of callback references when the Channel object is destroyed:\n\n1. When a DNS query is initiated, pycares stores a callback reference using `ffi.new_handle()`\n2. If the Channel object is garbage collected while queries are pending, the callback references become invalid\n3. When c-ares attempts to invoke the callback, it accesses freed memory, causing a fatal error\n\nThis issue was much more likely to occur when using `event_thread=True` but could happen without it under the right circumstances.\n\n### Technical Details\n\nThe core issue is a race condition between Python\u0027s garbage collector and c-ares\u0027s callback execution:\n\n1. When `__del__` is called from within a c-ares callback context, we cannot immediately call `ares_destroy()` because c-ares is still executing code after the callback returns\n2. c-ares needs to execute cleanup code after our Python callback returns (specifically at lines 1422-1429 in ares_process.c)\n3. If we destroy the channel too quickly, c-ares accesses freed memory\n\n### Impact\n\nApplications using `pycares` can be crashed remotely by triggering DNS queries that result in `Channel` objects being garbage collected before query completion. This is particularly problematic in scenarios where:\n\n- Channel objects are created per-request\n- Multiple failed DNS queries are processed rapidly\n- The application doesn\u0027t properly manage Channel lifecycle\n\nThe error manifests as:\n```\nFatal Python error: b_from_handle: ffi.from_handle() detected that the address passed points to garbage\n```\n\n## Fix\n\nThe vulnerability has been fixed in pycares 4.9.0 by implementing a safe channel destruction mechanism\n\n## Mitigation\n\n### For Application Developers\n\n1. **Upgrade to pycares \u003e= 4.9.0** - This version includes the fix and requires no code changes\n2. **Best practices** (optional but recommended):\n   ```python\n   # Explicit cleanup\n   channel.close()\n   \n   # Or use context manager\n   with pycares.Channel() as channel:\n       # ... use channel ...\n   # Automatically closed\n   ```\n3. **Avoid creating Channel objects per-request** - Prefer long-lived instances for better performance and safety\n\nThe fix is completely transparent - no API changes or code modifications are required.\n\n## Credit\n\nThis vulnerability was reported by @vEpiphyte through the aio-libs security program.",
  "id": "BREW-moodle-dl-CVE-2025-48945",
  "modified": "2026-08-13T17:14:32Z",
  "published": "2026-08-13T17:14:32Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/saghul/pycares/security/advisories/GHSA-5qpg-rh4j-qp35"
    },
    {
      "type": "WEB",
      "url": "https://github.com/saghul/pycares/commit/ebfd7d71eb8e74bc1057a361ea79a5906db510d4"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/saghul/pycares"
    }
  ],
  "schema_version": "1.7.3",
  "summary": "pycares has a Use-After-Free Vulnerability",
  "upstream": [
    "GHSA-5qpg-rh4j-qp35",
    "CVE-2025-48945"
  ]
}



Log in or create an account to share your comment.




Tags
Taxonomy of the tags.




Drag nodes to arrange them. Double-click a related vulnerability to open it, or a source to visit its link. The Physics button on the left tunes the layout and the simulation; the legend at the bottom left hides or shows nodes by kind and by sighting type.

Forecast uses a logistic model when the trend is rising, or an exponential decay model when the trend is falling. Fitted via linearized least squares.

Sightings

Author Source Type Date Other

Nomenclature

  • Seen: The vulnerability was mentioned, discussed, or observed by the user.
  • Confirmed: The vulnerability has been validated from an analyst's perspective.
  • Published Proof of Concept: A public proof of concept is available for this vulnerability.
  • Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
  • Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
  • Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
  • Not confirmed: The user expressed doubt about the validity of the vulnerability.
  • Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.
Vulnerability timeline

Sightings and recorded dates, grouped by day · UTC

JSON
How durations are calculated

Following the VTEM framework (Black Hat USA 2015), time to patch starts when the vendor or a coordinator was informed, total time to patch at discovery, and disclosure to patch at public disclosure; vendor response time runs from the report to its acknowledgement. Time to exploit and time unpatched start at the earliest discovery or report, falling back to public disclosure. Unpatched time ends at the first recorded patch, or the calculation time when no patch is recorded. Exploitation to patch shows the interval between those two events; a published proof of concept is a public exploit, not exploitation.

Each result shows its source dates. A negative interval means the end event came first; when date-only inputs leave the order open, the result says so. Date-only inputs are approximate. The earliest evidence wins: KEV report dates and sightings are labelled as such. Records describing another vulnerability are listed but never calculated from, and a PoC alone does not establish exploitation. Missing patch evidence does not prove that no patch exists. Filters below change the event list only.

Loading timeline…




    Related by attack behaviour

    Vulnerabilities whose description is nearest to this one in the vector space of the CIRCL/vulnerability-attack-technique-biencoder model. This is a similarity search over the bi-encoder space (plain cosine), not a classification, and it has no measured accuracy.