Look up another CVE · open CVE-2026-90371 in the interactive tracker
CVE-2026-90371
wifi: mt76: fix RXDMAD_C buffer recycling race
Published 2026-09-17 (31 days ago) by the Linux kernel CNA. Subsystem drivers/net. Debian urgency: unimportant.
What to do, per Debian release
Debian's own status for CVE-2026-90371 in every release this site tracks, and what each one means in practice. Every status below is the one the Debian Security Team publishes for this CVE. source
| Release | Status | Version | What this means |
|---|---|---|---|
| Debian 12bookworm | not affected | 6.1.190-1 | Nothing to do. bookworm never shipped the vulnerable code. |
| Debian 13trixie | not affected | 6.12.113-1 | Nothing to do. trixie never shipped the vulnerable code. |
| Debian 14forky | fixed | 7.2.6-1 | Update the kernel and reboot. Fixed in 7.2.6-1. A running kernel keeps the old code until the machine restarts. |
| Debian unstablesid | fixed | 7.2.6-1 | Update the kernel and reboot. Fixed in 7.2.6-1. A running kernel keeps the old code until the machine restarts. |
The vulnerability
What the bug is
The RXDMAD_C buffers come from the RRO data queues' page pools, which are bound to a different NAPI, so the direct page-pool recycle used here could race the owning NAPI; take the non-direct path as is already done for WED RX queues.
Where it lives
The fix touches these files. If the subsystem is one you do not use, the practical exposure is lower, though the package is still the vulnerable one.
drivers/net/wireless/mediatek/mt76/dma.c
- What this is
- MEDIATEK MT76 WIRELESS LAN DRIVERMAINTAINERS section matched on drivers/net/wireless/mediatek/mt76/ · linux-wireless@vger.kernel.org · Maintained
- Kernel module
- mt76
Check it on the machine
grep -wE 'CONFIG_MT76_CORE|CONFIG_WLAN|CONFIG_WLAN_VENDOR_MEDIATEK' /boot/config-$(uname -r)Reads the booted kernel's own configuration, on the machine and nowhere else. =y is built in, =m is a module, and an is not set line means the code is not there.
lsmod | grep -w mt76Prints a line if that module is loaded right now, and nothing if it is not. One that is not loaded can still be loaded later by anything that needs it.
Show whether Debian builds this, per release
Debian builds this, but not on every flavour. At least one release below builds the code on some of its kernel flavours and not on others, so the answer depends on the architecture and flavour installed.
- bookwormMnot built on amd64/cloud-amd64, arm64/cloud-arm64
- trixie, forky, sidMnot built on amd64/cloud-amd64, arm64/cloud-arm64, s390x/s390x
- M
- a module where it is built at all; some flavours do not build it
- Build switches
- CONFIG_MT76_CORE, CONFIG_WLAN, CONFIG_WLAN_VENDOR_MEDIATEK
- Reads as
- all on means the file is compiled; any one off means it is not.
From the configuration Debian publishes for every kernel flavour it builds. It covers only the architectures Debian ships a linux-config package for, and nothing about a kernel you built yourself. Debian's own status above stays authoritative about the source package.MAINTAINERS kbuild Makefiles Debian linux-config packages
Signals
- Published
- 2026-09-17 (31 days ago)source
- CVSS
- 8.8 high
CVSS:3.1/AV:A/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:Hsource recompute it - Attack vector
- adjacent network (AV:A)source
- EPSS
- 0.38%, higher than 29.9% of all CVEssource
- CISA KEV
- not listedthe KEV catalogue
- Debian urgency
- unimportantsource
- Subsystem
- drivers/net
Upstream fixes
The versions the CNA records as introducing and fixing this, and the commit for each. source
| Introduced | Fixed in | Commit |
|---|---|---|
| 6.18 | 6.18.52 | 7ac05ed3a50e |
| 6.18 | 7.2.6 | a273dc3b86a7 |
| 6.18 | 7.3 | e1f97c10a4ec |
Why the CNA scored it this way
Quoted from the kernel CNA's own scoring file, word for word. A highlighted metric carries the most severe value CVSS v3.1 defines for it, which is what pushes the score up. source
- AV:AAttack vector, adjacent network
- Received 802.11 frames are completed by HW-RRO v3.1 into the RXDMAD_C ring; mt7996_irq_handler/mt7996_irq_tasklet schedules napi[MT_RXQ_RRO_RXDMAD_C] and mt76_dma_rx_poll→mt76_dma_rx_process→mt76_dma_get_rxdmad_c_buf recycles those payload pages. A radio-range peer supplies the frames; WiFi is adjacent, not a routable IP transport.
- AC:LAttack complexity, low
- The attacker drives both NAPIs by flooding 802.11 traffic: the RRO data-queue NAPI runs mt76_dma_rx_fill→page_pool_alloc_frag on the owning pool, while RXDMAD_C drops completions (SDL0 vs SKB_WITH_OVERHEAD of MT7996_RX_BUF_SIZE, or extra frags in mt76_add_fragment) and calls mt76_put_page_pool_buf(..., true) into the same unlocked alloc.cache.
- PR:NPrivileges required, none
- mt76_dma_get_rxdmad_c_buf runs on the WiFi RX DMA completion path with no host credential check; a nearby unauthenticated station can send frames that HW-RRO v3.1 writes into RXDMAD_C (rx_token_id/SDL0) without any kernel account.
- UI:NUser interaction, none
- With the mt7996/mt7992 interface already up in HWRRO v3.1, the attacker triggers mt76_dma_rx_process on MT_RXQ_RRO_RXDMAD_C solely by transmitting frames; no mount, click, or other victim action is required.
- S:UScope, unchanged
- The race corrupts the host kernel page_pool alloc.cache for the RRO data queues (dev->q_rx[t->qid]->page_pool) in the same kernel; it does not cross a VM, IOMMU, or guest/host boundary.
- C:HConfidentiality impact, high
- Concurrent lockless updates of pool->alloc.cache/count in page_pool_recycle_in_cache versus page_pool_alloc_frag can issue the same RX page twice or return a stale pointer, a use-after-free of packet buffers that enables disclosure of kernel or packet memory.
- I:HIntegrity impact, high
- The same unlocked recycle can bump alloc.count past PP_ALLOC_CACHE_SIZE so a later page_pool_recycle_in_cache writes a netmem pointer out of bounds, and a double-issued page lets attacker-controlled 802.11 DMA clobber reused kernel memory.
- A:HAvailability impact, high
- A corrupted page_pool alloc.cache yields invalid page pointers and use-after-free in mt76_dma_rx_fill/mt76_dma_rx_process, which oopses or panics the kernel.