CVE-2026-99085: sched_ext: Fix CPU hotplug hang when a dying CPU's tasks sit in the BPF scheduler

Published Oct 23, 2026
·
Updated

In the Linux kernel, the following vulnerability has been resolved:

schedext: Fix CPU hotplug hang when a dying CPU's tasks sit in the BPF scheduler

A CPU going down has to empty its own rq. Its hotplug thread waits in schedcpuwaitempty() until nothing else is left, and the only thing that wakes it is balancepush(), which runs from schedule() on the dying CPU and pushes off the migratable tasks the CPU picks.

schedext breaks this. A task sitting on a user DSQ or held by the BPF scheduler is still counted on its rq, but an inactive CPU no longer calls ops.dispatch() and can't pull the task back. The dying CPU goes idle with the task still counted, and one of two things happens:

- Another CPU consumes the task. The rq empties without a schedule() on the dying CPU, so the hotplug thread is never woken and cpudown() hangs holding cpuhotpluglock. - The task is affine only to the dying CPU. Nothing can move it, and the offline stalls until the watchdog ejects the BPF scheduler.

When the rq goes offline, re-enqueue every task on it that isn't already on the local DSQ. An enqueue on an offline rq lands on the local DSQ, so the dying CPU itself picks the tasks and balancepush() pushes them off, the same as for the other sched classes. From then on no schedext path on another CPU can pull a task off the rq.

ops.dispatch() currently stops as soon as the CPU goes inactive, and CPU hotplug then waits for an RCU grace period before taking the rq offline. A BPF-held task affine only to the dying CPU and preempted inside an RCU read-side critical section would block that grace period, and the rq would never go offline. Test SCXRQONLINE directly so that ops.dispatch() keeps running until the rq goes offline. Only the dying CPU's own hotplug thread clears the flag during teardown, and taskcanrunonremoterq() still keeps other rqs' tasks off the CPU.

Affected Software

144 affected componentsFixes available
Linux Linux kernel
debian/linux<6.12.115-2
6.12.115-2
debian/linux<7.2.12-1
7.2.12-1
debian/linux-6.12<6.12.115-2~deb12u1
6.12.115-2~deb12u1
ubuntu/linux-hwe-edge*
ubuntu/linux-aws-5.0*
ubuntu/linux-aws-5.3*
ubuntu/linux-azure*
ubuntu/linux-azure-5.3*
ubuntu/linux-azure-edge*
ubuntu/linux-gcp*
ubuntu/linux-gcp-5.3*
ubuntu/linux-gke-4.15*
ubuntu/linux-gke-5.4*
ubuntu/linux-gkeop-5.4*
ubuntu/linux-hwe*
ubuntu/linux-hwe-edge*
ubuntu/linux-oem*
ubuntu/linux-oracle-5.0*
ubuntu/linux-oracle-5.3*
ubuntu/linux-aws-5.11*
ubuntu/linux-aws-5.13*
ubuntu/linux-aws-5.8*
ubuntu/linux-azure-5.11*
ubuntu/linux-azure-5.13*
ubuntu/linux-azure-5.8*
ubuntu/linux-azure-fde*
ubuntu/linux-gcp-5.11*
ubuntu/linux-gcp-5.13*
ubuntu/linux-gcp-5.8*
ubuntu/linux-gke*
ubuntu/linux-gke-5.15*
ubuntu/linux-gkeop*
ubuntu/linux-gkeop-5.15*
ubuntu/linux-hwe-5.11*
ubuntu/linux-hwe-5.13*
ubuntu/linux-hwe-5.8*
ubuntu/linux-intel-5.13*
ubuntu/linux-oem-5.10*
ubuntu/linux-oem-5.13*
ubuntu/linux-oem-5.6*
ubuntu/linux-oracle-5.11*
ubuntu/linux-oracle-5.13*
ubuntu/linux-oracle-5.8*
ubuntu/linux-raspi2*
ubuntu/linux-riscv*
ubuntu/linux-riscv-5.11*
ubuntu/linux-riscv-5.8*
ubuntu/linux-allwinner-5.19*
ubuntu/linux-aws-5.19*
ubuntu/linux-aws-6.2*
ubuntu/linux-aws-6.5*
ubuntu/linux-azure-5.19*
ubuntu/linux-azure-6.2*
ubuntu/linux-azure-6.5*
ubuntu/linux-azure-fde-5.19*
ubuntu/linux-azure-fde-6.2*
ubuntu/linux-gcp-5.19*
ubuntu/linux-gcp-6.2*
ubuntu/linux-gcp-6.5*
ubuntu/linux-hwe-5.19*
ubuntu/linux-hwe-6.2*
ubuntu/linux-hwe-6.5*
ubuntu/linux-intel-iot-realtime*
ubuntu/linux-lowlatency-hwe-5.19*
ubuntu/linux-lowlatency-hwe-6.2*
ubuntu/linux-lowlatency-hwe-6.5*
ubuntu/linux-nvidia-6.2*
ubuntu/linux-nvidia-6.5*
ubuntu/linux-oem-5.17*
ubuntu/linux-oem-6.0*
ubuntu/linux-oem-6.1*
ubuntu/linux-oem-6.5*
ubuntu/linux-oracle-6.5*
ubuntu/linux-realtime*
ubuntu/linux-riscv*
ubuntu/linux-riscv-5.19*
ubuntu/linux-riscv-6.5*
ubuntu/linux-starfive-5.19*
ubuntu/linux-starfive-6.2*
ubuntu/linux-starfive-6.5*
ubuntu/linux-aws-6.14*
ubuntu/linux-aws-6.17*
ubuntu/linux-aws-7.0*
ubuntu/linux-azure-6.11*
ubuntu/linux-azure-6.14*
ubuntu/linux-azure-6.17*
ubuntu/linux-azure-7.0*
ubuntu/linux-azure-fde-6.14*
ubuntu/linux-azure-fde-6.17*
ubuntu/linux-azure-fde-7.0*
ubuntu/linux-azure-nvidia-6.14*
ubuntu/linux-gcp-6.11*
ubuntu/linux-gcp-6.14*
ubuntu/linux-gcp-7.0*
ubuntu/linux-hwe-6.11*
ubuntu/linux-hwe-6.14*
ubuntu/linux-hwe-6.17*
ubuntu/linux-hwe-7.0*
ubuntu/linux-lowlatency-hwe-6.11*
ubuntu/linux-nvidia-6.11*
ubuntu/linux-nvidia-6.17*
ubuntu/linux-nvidia-7.0*
ubuntu/linux-oem-6.11*
ubuntu/linux-oem-6.14*
ubuntu/linux-oem-6.17*
ubuntu/linux-oem-6.8*
ubuntu/linux-oracle-6.14*
ubuntu/linux-oracle-6.17*
ubuntu/linux-oracle-7.0*
ubuntu/linux-raspi-realtime*
ubuntu/linux-realtime*
ubuntu/linux-riscv*
ubuntu/linux-riscv-6.14*
ubuntu/linux-riscv-6.17*
ubuntu/linux-riscv-7.0*
ubuntu/linux-realtime-6.14*
ubuntu/linux-realtime-6.17*
ubuntu/linux<7.3.0-10.10
7.3.0-10.10
ubuntu/linux-aws<7.3.0-1004.4
7.3.0-1004.4
ubuntu/linux-azure<7.3.0-1004.4
7.3.0-1004.4
ubuntu/linux-azure-fde<7.3.0-1004.5
7.3.0-1004.5
ubuntu/linux-gcp<7.3.0-1004.4
7.3.0-1004.4
ubuntu/linux-oracle<7.3.0-1004.4
7.3.0-1004.4
ubuntu/linux-raspi<7.3.0-1004.4
7.3.0-1004.4
ubuntu/linux-realtime<7.3.0-10.10.1
7.3.0-10.10.1
ubuntu/linux-riscv<7.3.0-10.10.1
7.3.0-10.10.1
ubuntu/linux*
ubuntu/linux-aws*
ubuntu/linux-azure*
ubuntu/linux-azure-fde*
ubuntu/linux-gcp*
ubuntu/linux-gke*
ubuntu/linux-ibm*
ubuntu/linux-nvidia*
ubuntu/linux-nvidia-bos*
ubuntu/linux-oem-7.0*
ubuntu/linux-oracle*
ubuntu/linux-raspi*
ubuntu/linux-realtime*
ubuntu/linux-riscv*
ubuntu/linux-vmware*
debian/linux<=6.12.107-1
6.1.176-16.1.192-26.12.117-17.3.4-1
debian/linux-6.12
6.12.117-1~deb12u1

Remediation

Recommended actions to resolve this vulnerability, in priority order.

  1. Upgrade

    Upgrade debian/linux to a version that resolves this vulnerability.

    Fixed in 6.12.115-2
  2. Upgrade

    Upgrade debian/linux to a version that resolves this vulnerability.

    Fixed in 7.2.12-1
  3. Upgrade

    Upgrade debian/linux-6.12 to a version that resolves this vulnerability.

    Fixed in 6.12.115-2~deb12u1
  4. Upgrade

    Upgrade ubuntu/linux to a version that resolves this vulnerability.

    Fixed in 7.3.0-10.10
  5. Upgrade

    Upgrade ubuntu/linux-aws to a version that resolves this vulnerability.

    Fixed in 7.3.0-1004.4
  6. Upgrade

    Upgrade ubuntu/linux-azure to a version that resolves this vulnerability.

    Fixed in 7.3.0-1004.4
  7. Upgrade

    Upgrade ubuntu/linux-azure-fde to a version that resolves this vulnerability.

    Fixed in 7.3.0-1004.5
  8. Upgrade

    Upgrade ubuntu/linux-gcp to a version that resolves this vulnerability.

    Fixed in 7.3.0-1004.4
  9. Upgrade

    Upgrade ubuntu/linux-oracle to a version that resolves this vulnerability.

    Fixed in 7.3.0-1004.4
  10. Upgrade

    Upgrade ubuntu/linux-raspi to a version that resolves this vulnerability.

    Fixed in 7.3.0-1004.4
  11. Upgrade

    Upgrade ubuntu/linux-realtime to a version that resolves this vulnerability.

    Fixed in 7.3.0-10.10.1
  12. Upgrade

    Upgrade ubuntu/linux-riscv to a version that resolves this vulnerability.

    Fixed in 7.3.0-10.10.1
  13. Upgrade

    Upgrade debian/linux to a version that resolves this vulnerability.

    Fixed in 6.1.176-1Fixed in 6.1.192-2Fixed in 6.12.117-1Fixed in 7.3.4-1
  14. Upgrade

    Upgrade debian/linux-6.12 to a version that resolves this vulnerability.

    Fixed in 6.12.117-1~deb12u1
  15. Compensating control

    In sched_ext, test SCX_RQ_ONLINE directly so ops.dispatch() continues running until the rq goes offline; when the rq goes offline, re-enqueue every task on it that is not already on the local DSQ.

Event History

Oct 23, 2026
CVE Published
via MITRE·12:45 PM
Data Sourced
via MITRE·12:45 PM
Description
Data Sourced
via OSV·01:17 PM
DescriptionSeverityAffected Software
Event
via OSV·01:17 PM
Data Sourced
via NVD·01:17 PM
Description

Parent advisories

This vulnerability appears in the following advisories.

Peer vulnerabilities

Found alongside the following vulnerabilities.

Frequently Asked Questions

1

Which systems are exposed to this hang?

Linux systems using sched_ext with tasks held in a user DSQ or by the BPF scheduler during CPU offlining are affected. The failure occurs when those tasks remain counted on the dying CPU's run queue after that CPU stops dispatching work.

2

How does task affinity change the observed impact?

If another CPU can consume the task, the run queue can empty without waking the hotplug thread, leaving cpu_down() hung while holding cpu_hotplug_lock. If the task is affine only to the dying CPU, offlining stalls until the watchdog ejects the BPF scheduler.

3

What should be done to remediate this issue?

Use a Linux kernel containing the sched_ext CPU-hotplug fix. The listed stable-kernel commits identify backported fixes, but the available data does not map them to specific kernel release versions.

Contact

SecAlerts Pty Ltd.
132 Wickham Terrace
Fortitude Valley,
QLD 4006, Australia
info@secalerts.co
By using SecAlerts services, you agree to our services end-user license agreement. This website is safeguarded by reCAPTCHA and governed by the Google Privacy Policy and Terms of Service. All names, logos, and brands of products are owned by their respective owners, and any usage of these names, logos, and brands for identification purposes only does not imply endorsement. If you possess any content that requires removal, please get in touch with us.
© 2026 SecAlerts Pty Ltd.
ABN: 70 645 966 203, ACN: 645 966 203