* [CVE-2026-99184][LOW] accel/rocket: search every core slot when a core is removed
@ 2026-10-30 16:13 21% AL-KERNEL
0 siblings, 0 replies; 1+ results
From: AL-KERNEL @ 2026-10-30 16:13 UTC (permalink / raw)
To: kernel-cve
CVE: CVE-2026-99184
Priority: LOW
AL-KERNEL base severity: LOW
KPANIC flag: NO
Patch: accel/rocket: search every core slot when a core is removed
Commit: 5078c91f5c84bb383d0a10ed39856f98575e39cf
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=5078c91f5c84bb383d0a10ed39856f98575e39cf
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-99184
Analysis date: Fri, 30 Oct 2026 12:13:45 -0400
ActionableScore: 1
ActionableScore lower bound: 0
Actionable bucket: Ordinary Moderate
Manual review required: NO
Summary:
Repeated host-admin device unbind/rebind leaves a stale core count that causes a fixed-slot heap OOB write and later invalid structured-state consumption.
======================================================================
ABOUT THIS REPORT
======================================================================
The original Linux kernel CVE announcement for CVE-2026-99184 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-99184
The original announcement does not normally provide a security severity
estimate, CVSS assessment, or enough information to determine whether the
reported kernel bug represents a practically relevant security issue.
This report was generated by AL-KERNEL, an AI-assisted Linux kernel
vulnerability analysis system developed by Alexander Larkin. It combines
an autonomous classifier with LLM-assisted technical analysis and a
separate ActionableScore mechanism.
The purpose of this report is to prioritize Linux kernel CVEs before
manual review, identify cases that require prompt investigation, and
support automatic closure of issues that are unlikely to have meaningful
security impact.
Published priority for this report: LOW
Manual review required: NO
A detailed explanation of the methodology and priority rules is included
at the end of this message.
======================================================================
AL-KERNEL CLASSIFICATION RESULT
======================================================================
CVE-2026-99184 LOW https://www.kernelcve.org/list/?q=CVE-2026-99184 CHECK WITH IMPACT FROM ORIG NN LOW | AEGIS IMPACT MODERATE_KPANIC AEGIS_AS=0 LEGACY=LOW RESULT=DIFF TEMP_DISTANCE_MERGE_SKIP=AEGIS_NO_POSITIVE_SEVERITY_EVIDENCE (final Legacy and Aegis differ; FINAL: LOW, Aegis: MODERATE_KPANIC) Maybe valid. Check manually. Hints by AL-KERNEL: The best (paranoid) CVSS is 'AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H';*CWE-787;CWE-122;*CWE-672;Other CVSS 'AV:L/AC:L/PR:H/UI:N/S:U/C:N/I:L/A:H';BEST CVSS score: '7.8';DESCR 'The Rocket driver can fail to locate the final core during ordered device removal because the search range shrinks with num_cores. The incomplete teardown leaves stale rdev state, and a later bind can use the stale count as an array index and write rdev cores 3 past a three element allocation. This is a concrete kernel heap out-of-bounds write with plausible integrity and availability impact, although no controlled arbitrary write or information disclosure primitive is demonstrated. For the CVSS the PR:L value in the paranoid vector reflects systems where device rebinding is delegated to a restricted service or reduced capability administrator. Normal sysfs unbind and bind access instead supports PR:H for the base vector. Automated impact decision path. The original SimpleNet classifier selected LOW. The primary LLM estimated paranoid CVSS 7.8 and the production ActionableScore returned 1 with recommendation Ordinary Moderate. The production rules then evaluated CVSS thresholds and ActionableScore thresholds. Recorded rule transitions were final reconciliation stage LOW to LOW because KP specialist did not establish a qualifying panic and the cross model evidence reconciliation rule was applied. The legacy KP specialist has no review artifact for this run. The legacy FD or DH downgrade specialist has no review artifact for this run. Exact triggered rule markers were KPANIC_UNSUPPORTED_REMOVED and V64_AEGIS_UNSUPPORTED_PROMOTION_BLOCKED. The independent weaker Aegis shadow evaluated CVSS 0.0 and ActionableScore 0 with primitive class CLASS_B and manual review REQUIRED. Its independently reconciled impact was MODERATE_KPANIC. The models disagreed but the aggregation rules retained the production result. The final reported impact is LOW.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Ordinary Moderate (with actual score 1) MAYBE KASAN LEAK HARDWARE LINUS TEST KPANIC_UNSUPPORTED_REMOVED V64_AEGIS_UNSUPPORTED_PROMOTION_BLOCKED - - checked
======================================================================
ACTIONABLESCORE ANALYSIS
======================================================================
ActionableScore=1
ActionableScoreLower=0
PrimitiveClass=CLASS_B
RealUAF=NO
HostAdminRequired=YES
NamespaceScopedPrivilege=NO
ImportantCandidate=NO
ManualReview=RECOMMENDED
AutoDowngradeAllowed=YES
RequiresHostUID0=UNKNOWN
AuthorizationDecisionBypass=NO
ProtectedResourceImpact=NONE
B4PendingReview=NO
B4PendingConfirmed=NO
1. ActionableScore
* Raw base score: 0.
* Conservative score: 0.
* Raw paranoid score: 1.
* Paranoid score: 1.
* Final recommended bucket: **Ordinary Moderate**::
2. Signal breakdown
* +1 constrained heap overwrite: stale `num_cores` makes the second probe round write `rdev->cores[3]` beyond a three-element allocation. The index sequence is controllable through repeated unbind/rebind, but the destination is a fixed adjacent slot and arbitrary payload or target selection is not shown.
* +1 corrupted structured-state consumption: the supplied devfreq test context shows a worker subsequently reading the overrun entry, producing an invalid `_Bool` load.
* -2 base privilege: the evidenced trigger uses host platform-device sysfs unbind/rebind, described as requiring normal host-administrative access.
* Paranoid difference: use -1 instead of -2 only for the supplied conditional deployment where rebinding is delegated to a reduced-privilege service.
* No hardware-rarity penalty: the path is hardware-dependent, but the evidence does not establish sufficiently rare hardware or physical attacker access.
* No crash point: UBSAN detected an invalid load, but a reliable panic or host-wide DoS is not shown.
* No arbitrary write, controlled adjacent victim, disclosure, LPE, or protected-state integrity impact is established.
B5-R is not confirmed: conditions (1), (2), and corrupted-state consumption are supported, but the complete evidenced retry trigger is host-admin-only. Conditional delegation does not prove a concrete below-host-admin path.
3. Reachability analysis
The demonstrated actor repeatedly unbinds and rebinds all Rockchip NPU cores through host sysfs. These platform-device operations are not namespace-scoped. Standard access is host administrative; numeric UID 0 specifically is not proven because ACL/service delegation is possible. No render-node or ordinary-user ioctl trigger is supplied. The sequence is deterministic rather than a race and requires supported Rockchip NPU hardware.
4. Primitive and severity reading
This is a confirmed retry-accumulated heap OOB write into an adjacent `rocket_core` slot, followed in the supplied devfreq context by consumption of corrupted structured state. It is CLASS_B because the write has a fixed index and kernel/device-derived contents, with no directed target, arbitrary payload, or independently useful corrupt primitive shown. It is not a UAF. Host-admin-only demonstrated reachability and lack of a protected boundary keep it from the B5-R Important-preserving profile. Auto-downgrade remains allowed, but source review is recommended because the exact adjacent heap effects are not characterized.
5. One-sentence report phrase
Repeated host-admin device unbind/rebind leaves a stale core count that causes a fixed-slot heap OOB write and later invalid structured-state consumption.
6. Manual review advice
MANUAL CHECK RECOMMENDED
Confirm whether any product delegates platform-device rebinding below host-admin and whether an in-tree or downstream worker consumes additional overwritten pointer fields.
7. Call-site confidence
Call-site confidence: medium
======================================================================
UPSTREAM PATCH SUMMARY
======================================================================
Patch: accel/rocket: search every core slot when a core is removed
Commit: 5078c91f5c84bb383d0a10ed39856f98575e39cf
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=5078c91f5c84bb383d0a10ed39856f98575e39cf
Commit description:
rocket_remove() decrements rdev->num_cores for each core it removes,
while find_core_for_dev() searches slots 0 to num_cores - 1. Unbinding
the cores in the order they were bound therefore loses the last one: by
the time it is removed the search range has already shrunk past its
slot, so find_core_for_dev() returns -1 and rocket_remove() gives up
without doing anything.
num_cores never reaches zero, rocket_device_fini() never runs, and the
file-scoped rdev keeps pointing at a device that is going away. Binding
the cores again starts from that stale count, because rocket_probe()
takes rdev->num_cores as the slot to fill. On an RK3588, which describes
three cores, the second round lands on slots 1, 2 and 3 while rdev->cores
was allocated with room for three:
rocket fdab0000.npu: Rockchip NPU core 1 version: 1179210309
rocket fdac0000.npu: Rockchip NPU core 2 version: 1179210309
rocket fdad0000.npu: Rockchip NPU core 3 version: 1179210309
The write to rdev->cores[3] is past the end of the array.
Nothing in tree reads the core array often enough to notice, so the
overrun is silent today. It turned up while testing a devfreq series on
top of this, where a worker walks every core a few times a second, and
UBSAN caught the first bool it read out of the overrun entry:
UBSAN: invalid-load in drivers/accel/rocket/rocket_devfreq.c:47:10
load of value 5 is not a valid value for type '_Bool'
Workqueue: devfreq_wq devfreq_monitor
Record how many slots were allocated and search all of them. Every core
is then found on removal, num_cores reaches zero, the device is torn down
and a later bind starts from a clean rdev.
This does not make unbinding a single core out of several work. probe
still takes num_cores as the slot to fill, so rebinding one core while
its siblings stay bound would write over a slot that is already in use,
and rocket_open() still reaches for cores[0] whether or not anything is
there. Both of those want more thought than a fix should carry.
Found by unbinding and rebinding all three cores on an Orange Pi 5 Plus.
With this applied, 25 unbind/rebind rounds and 5 module unload/reload
rounds run clean there: the cores land in slots 0, 1 and 2 every time,
whichever order they are bound in, and the shared supply goes back to a
single user after each round.
Fixes: ed98261 ("accel/rocket: Add a new driver for Rockchip's NPU")
Cc: [email protected]
Signed-off-by: Igor Paunovic <[email protected]>
Assisted-by: LLM sparse checkpatch
Tested-by: Jiaxing Hu <[email protected]> # RK3576, two cores
Tested-by: Sidong Yang <[email protected]> # RK3588, 3 cores
Signed-off-by: Heiko Stuebner <[email protected]>
Link: https://patch.msgid.link/[email protected]
Signed-off-by: Greg Kroah-Hartman <[email protected]>
Changed files:
drivers/accel/rocket/rocket_devfreq.c
drivers/accel/rocket/rocket_device.c
drivers/accel/rocket/rocket_device.h
drivers/accel/rocket/rocket_drv.c
Diff excerpt:
Not included in this email. See the upstream URL for the full patch.
Full patch:
https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=5078c91f5c84bb383d0a10ed39856f98575e39cf
======================================================================
DETAILED REPORT METHODOLOGY
======================================================================
The original Linux kernel CVE announcement for CVE-2026-99184 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-99184
The original CVE announcement normally does not include a security-level
estimate. In particular, it may not contain a CVSS assessment, an impact
level, or enough information to determine whether the reported bug is a
practically relevant security issue. One purpose of this parallel CVE list
is to provide that missing technical and prioritization information.
The original goal of the AL-KERNEL project was to prioritize Linux kernel
CVE analysis automatically before manual review. The system can also help
identify non-security issues that may be suitable for automatic closure.
This report was generated by AL-KERNEL, an AI-assisted Linux kernel
vulnerability analysis system developed by Alexander Larkin.
The first analysis stage combines an autonomous classifier with additional
LLM-based analysis. The autonomous classifier runs locally on a CPU and is
based on a backpropagation neural network. Together, these mechanisms
produce a technical vulnerability description, identify likely weakness
types, estimate CVSS severity, and provide input for ActionableScore.
Two CVSS estimates are retained because incomplete kernel vulnerability
information often permits more than one defensible interpretation:
Conservative CVSS vector: AV:L/AC:L/PR:H/UI:N/S:U/C:N/I:L/A:H
The Best / paranoid CVSS vector: AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H
The Best / paranoid CVSS score: 7.8
The conservative vector represents a lower-impact interpretation.
The Best/paranoid vector intentionally represents a plausible upper-bound
interpretation and should not automatically be treated as demonstrated
real-world impact.
CVSS may also need to be adjusted for a particular Linux deployment,
because actual reachability, privileges, enabled kernel configuration,
hardware, namespaces, exposed device nodes, and other environmental
conditions can differ significantly between systems.
A separate ActionableScore mechanism evaluates practical remediation
urgency. Its analysis may include reachability, attack prerequisites,
subsystem exposure, memory-corruption characteristics, denial-of-service
reliability, and possible confidentiality, integrity, or
privilege-escalation impact.
Conservative ActionableScore: 0
Paranoid ActionableScore: 1
The final base severity is taken directly from the second tab-separated
field of the AL-KERNEL classification result. ActionableScore does not
replace or independently override that final AL-KERNEL decision, and
if ActionableScore adjusted impact level of ALKERNEL, then you would see
self-readable flags above like INCREASED_TO_HIGH_BASED_ON_ACTIONABLESCOREHIGHEREQTHAN7.
For an AL-KERNEL result of MODERATE, this report uses the following
additional presentation split:
ActionableScore below 5 -> MODERATE REGULAR
ActionableScore 5 or more -> MODERATE 7.0
The distinction between MODERATE REGULAR and MODERATE 7.0 makes it
possible to identify Moderate issues that should receive manual analysis
and fixes before lower-priority MODERATE REGULAR issues. In many cases,
MODERATE REGULAR fixes may wait for a later rebase or routine update.
There is one override in which MODERATE REGULAR becomes MODERATE 7.0
even when the ActionableScore is below 5. When the AL-KERNEL result
contains the KPANIC flag, a MODERATE result is always presented as
MODERATE 7.0. The KPANIC flag selected with few regexps without
usage of AI at all, so it helps to detect cases when Kernel Crash happens
and similar (to filter False-Negative results from the LLM usage).
KPANIC indicates that a reliable kernel crash, kernel panic, or similarly
serious kernel availability impact was identified by the classification
workflow.
AL-KERNEL base severity for this report: LOW
KPANIC detected for this report: NO
Published priority for this report (same as in Subject): LOW
These results are intended to support engineering triage. They are
machine-generated estimates, and cases marked for manual review should
be validated by a human security engineer before final disposition.
For more info read docs linked from here: https://kernelcve.org/
(and you can submit you own patch there to generate such a report
for non-existant CVE-id yet).
Note that in many cases this AI tool selects higher severity, than
real is (means you can expect Importants instead of Moderate 7.0 or
Moderates 7.0 instead of regular Moderates). If you see such cases,
please use reply email interface to add additional manual analyses
info to this particular CVE.
And please, please, let me know when you see Lows instead of Importants
or Important instead of Low (because particular for such cases I
need to tune this AI tool to make it better for this one and next similar).
My contact email for such notifications is [email protected] (and both
send reply to CVE record itself too and see "reply" button below for howto reply).
^ permalink raw reply [relevance 21%]
Results 1-1 of 1 | reverse | sort options + mbox downloads above
-- links below jump to the message on this page --
2026-10-30 16:13 21% [CVE-2026-99184][LOW] accel/rocket: search every core slot when a core is removed AL-KERNEL
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox