GHSA-W4Q3-279C-XVRV
Vulnerability from github – Published: 2026-08-26 21:31 – Updated: 2026-08-26 21:31A heap-based buffer overflow vulnerability exists in the Dia diagram editor WPG file format importer.
In plug-ins/wpg/wpg-import.c, the WPG import renderer allocates a fixed palette with:
ren->pPal = g_new0(WPGColorRGB, 256);
When handling a WPG_COLORMAP record, the parser reads a start index (i16) and number of colors (iNum16) from the file and reads palette data with:
bRet &= (iNum16 == (int)fread(&ren->pPal[i16], sizeof(WPGColorRGB), iNum16, f));
The only bounds-related check is if (i16 >= 0 && i16 <= iSize), where iSize is the WPG record size—not the palette capacity. There is no validation that i16 is less than 256 or that i16 + iNum16 does not exceed 256.
A malicious WPG file can supply i16=256 and iNum16=264. That causes fread() to write 792 bytes starting at &pPal[256], while the palette buffer is only 768 bytes (256 entries × 3 bytes). This overflows into adjacent heap metadata and can crash Dia (SIGABRT / malloc corruption errors) or, depending on heap layout and exploit primitives, potentially lead to arbitrary code execution.
Exploitation requires convincing a user to open a crafted WPG file via Dia's file dialog, command line, or file association. No special privileges are required to deliver the file to the victim.
Affected component: WPG parser (plug-ins/wpg/wpg-import.c). Affected versions: all Dia versions containing this code path (reporter tested Dia 0.98+git20260221-1; issue present on upstream master as of 2026-08-21).
{
"affected": [],
"aliases": [
"CVE-2026-77652"
],
"database_specific": {
"cwe_ids": [
"CWE-122"
],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2026-08-26T20:18:02Z",
"severity": "HIGH"
},
"details": "A heap-based buffer overflow vulnerability exists in the Dia diagram editor WPG file format importer.\n\nIn plug-ins/wpg/wpg-import.c, the WPG import renderer allocates a fixed palette with:\n\n ren-\u003epPal = g_new0(WPGColorRGB, 256);\n\nWhen handling a WPG_COLORMAP record, the parser reads a start index (i16) and number of colors (iNum16) from the file and reads palette data with:\n\n bRet \u0026= (iNum16 == (int)fread(\u0026ren-\u003epPal[i16], sizeof(WPGColorRGB), iNum16, f));\n\nThe only bounds-related check is `if (i16 \u003e= 0 \u0026\u0026 i16 \u003c= iSize)`, where iSize is the WPG record size\u2014not the palette capacity. There is no validation that i16 is less than 256 or that i16 + iNum16 does not exceed 256.\n\nA malicious WPG file can supply i16=256 and iNum16=264. That causes fread() to write 792 bytes starting at \u0026pPal[256], while the palette buffer is only 768 bytes (256 entries \u00d7 3 bytes). This overflows into adjacent heap metadata and can crash Dia (SIGABRT / malloc corruption errors) or, depending on heap layout and exploit primitives, potentially lead to arbitrary code execution.\n\nExploitation requires convincing a user to open a crafted WPG file via Dia\u0027s file dialog, command line, or file association. No special privileges are required to deliver the file to the victim.\n\nAffected component: WPG parser (plug-ins/wpg/wpg-import.c).\nAffected versions: all Dia versions containing this code path (reporter tested Dia 0.98+git20260221-1; issue present on upstream master as of 2026-08-21).",
"id": "GHSA-w4q3-279c-xvrv",
"modified": "2026-08-26T21:31:45Z",
"published": "2026-08-26T21:31:45Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2026-77652"
},
{
"type": "WEB",
"url": "https://access.redhat.com/security/cve/CVE-2026-77652"
},
{
"type": "WEB",
"url": "https://bugzilla.redhat.com/show_bug.cgi?id=2520884"
},
{
"type": "WEB",
"url": "https://gitlab.gnome.org/GNOME/dia/-/blob/master/plug-ins/wpg/wpg-import.c"
},
{
"type": "WEB",
"url": "https://gitlab.gnome.org/GNOME/dia/-/issues/580"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H",
"type": "CVSS_V3"
}
]
}
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.
The approach is described in our paper Mapping CVEs to MITRE ATT&CK Techniques: A Curated Gold-Set Classifier and the Limits of LLM-Assisted Label Expansion.