Look up another CVE · open CVE-2026-93277 in the interactive tracker
CVE-2026-93277
RDMA/bnxt_re: Validate udata before executing commands
Published 2026-09-24 (24 days ago) by the Linux kernel CNA. Subsystem drivers/infiniband. Debian urgency: unimportant.
What to do, per Debian release
Debian's own status for CVE-2026-93277 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 destroy callbacks currently zero the udata output after tearing down driver resources. If the userspace access fails, uverbs preserves the uobject and allows the destroy callback to run again, even though the driver resource has already been freed. Call ib_no_udata_io() before teardown so udata failures are detected while the resource is still intact, then return success after teardown completes. As part of this change, move ib_respond_empty_udata() to the start of the create and modify flows. While this is not strictly required for general create flows, as the core layer unwinds uobjects on failure, it is necessary for create AH. In _rdma_create_ah(), the HW object is otherwise leaked.
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/infiniband/hw/bnxt_re/ib_verbs.c
- What this is
- BROADCOM NETXTREME-E ROCE DRIVERMAINTAINERS section matched on drivers/infiniband/hw/bnxt_re/ · linux-rdma@vger.kernel.org · Supported
- Kernel module
- bnxt_re
Check it on the machine
grep -wE 'CONFIG_INFINIBAND|CONFIG_INFINIBAND_BNXT_RE' /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 bnxt_rePrints 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 does not build this code in 1 release. It does build it in 3 releases, so the answer depends on which release the machine runs.
- bookwormn
- trixieMnot built on amd64/cloud-amd64, arm64/cloud-arm64, armel/rpi, armhf/armmp, armhf/armmp-lpae, armhf/rt/armmp
- forky, sidMnot built on amd64/cloud-amd64, arm64/cloud-arm64, armhf/armmp, armhf/armmp-lpae
- n
- not enabled on any flavour - this release does not build the code
- M
- a module where it is built at all; some flavours do not build it
- Build switches
- CONFIG_INFINIBAND, CONFIG_INFINIBAND_BNXT_RE
- 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
- 2026-09-24 (24 days ago)source
- CVSS
- 7.8 high
CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:Hsource recompute it - Attack vector
- local (AV:L)source
- EPSS
- 0.14%, higher than 3.3% of all CVEssource
- CISA KEV
- not listedthe KEV catalogue
- Debian urgency
- unimportantsource
- Subsystem
- drivers/infiniband
Upstream fixes
The versions the CNA records as introducing and fixing this, and the commit for each. source
| Introduced | Fixed in | Commit |
|---|---|---|
| 7.1 | 7.2.6 | c5643ea4f040 |
| 7.1 | 7.3 | d38c835925d4 |
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:LAttack vector, local
- The attacker-controlled input is the UVERBS_ATTR_UHW_OUT pointer that ib_uverbs_ioctl's UVERBS_METHOD_INVOKE_WRITE copies into attrs->driver_udata; bnxt_re_dereg_mr, bnxt_re_dealloc_pd, bnxt_re_destroy_qp, bnxt_re_destroy_cq and bnxt_re_destroy_srq then clear_user that buffer via ib_respond_empty_udata after teardown. RoCE receive never supplies that udata.
- AC:LAttack complexity, low
- Mapping a PROT_READ/PROT_NONE page and passing it as UHW_OUT on INVOKE_WRITE IB_USER_VERBS_CMD_DEREG_MR (or DEALLOC_PD/DESTROY_QP/CQ/SRQ) makes ib_respond_empty_udata's clear_user fail after bnxt_re_dereg_mr kfree(mr); uverbs_destroy_uobject then keeps uobj->object and a second destroy or close() retries destroy_hw. No race or layout the attacker cannot arrange.
- PR:LPrivileges required, low
- uverbs_devnode() creates /dev/infiniband/uverbs* mode 0666 and ib_uverbs_open() only checks rdma_dev_access_netns(), so an unprivileged process can GET_CONTEXT, REG_MR, then INVOKE_WRITE DEREG_MR without CAP_SYS_ADMIN or init-namespace root.
- UI:NUser interaction, none
- The attacker opens their own uverbs fd, creates an MR/PD/QP/CQ/SRQ, then issues INVOKE_WRITE destroy with a non-writable UHW_OUT and a follow-up destroy or close(); no other user must mount, open, or click.
- S:UScope, unchanged
- The double-free is of the host kernel's bnxt_re_mr (kfree in bnxt_re_dereg_mr) and related driver objects (fence->mr, cq->cql, umem). It is local privilege escalation and does not cross a VM, IOMMU, or guest-to-host boundary.
- C:HConfidentiality impact, high
- bnxt_re_dereg_mr kfree(mr) then returns -EFAULT from ib_respond_empty_udata, so uobj->object still points at the freed bnxt_re_mr; the second destroy_hw walks that slab, and reclaiming it yields an arbitrary kernel read.
- I:HIntegrity impact, high
- The same preserved uobject lets the attacker spray the freed bnxt_re_mr (and fence->mr from bnxt_re_destroy_fence_mr on PD destroy, or cq->cql from bnxt_re_destroy_cq) and hijack the second destroy_hw/kfree, which is an arbitrary write and control-flow primitive.
- A:HAvailability impact, high
- Retrying destroy or closing the uverbs fd after the failed ib_respond_empty_udata double-frees mr/cql/umem and re-enters bnxt_qplib_destroy_qp/cq/srq on already-destroyed firmware objects, which oopses or panics the host.