irqchip/gic-v4: Don't allow a VMOVP on a dying VPE
Linux
Affected versions
| Product | Affected | Fixed in |
|---|---|---|
| Linux Linux |
|
Linux
| Product | Affected | Fixed in |
|---|---|---|
| Linux Linux |
|
≥ 64edfaa9a2342a3ce34f8cb982c2c2df84db4de3 · < 1442ee0011983f0c5c4b92380e6853afb513841a| No fix yet |
| Linux Linux |
|
|
In the Linux kernel, the following vulnerability has been resolved:
irqchip/gic-v4: Don't allow a VMOVP on a dying VPE
Kunkun Jiang reported that there is a small window of opportunity for userspace to force a change of affinity for a VPE while the VPE has already been unmapped, but the corresponding doorbell interrupt still visible in /proc/irq/.
Plug the race by checking the value of vmapp_count, which tracks whether the VPE is mapped ot not, and returning an error in this case.
This involves making vmapp_count common to both GICv4.1 and its v4.0 ancestor.
From the CVE record · Linux
A race in the Linux kernel's GICv4 interrupt controller handling can leave an unmapped VPE's doorbell interrupt visible in /proc/irq/. During that window, local userspace can force an affinity change for the VPE; the input does not specify whether authentication or additional user interaction is required.
Engineers running Linux kernels with GICv4 or GICv4.1 support should check their systems; affected versions are not specified.