CVE-2026-99092
NoneCVSS
—
EPSS
0.2%
Published
Oct 26, 2026
CISA due date
—
Cloudanix analysis
A recent Linux kernel change on RISC-V processors can stop certain low-level programs from running at full speed, requiring a software update to restore normal performance.
- Exploitation
- Local Privilege
- Attacker needs
- Low Privilege
- Complexity
- Low
- Impact
- Denial Of Service
- Component
- OS kernel
- Class
- Memory Corruption
- Remediation
- Patch
- Effort
- Low
- Patch risk
- Low
What it actually does
On RISC-V with CONFIG_STRICT_MODULE_RWX=n, patch_map() returns a read-only vmalloc address for BPF JIT text; copy_to_kernel_nofault then faults, bpf_arch_text_copy converts the fault to -EINVAL, and bpf_int_jit_compile emits a WARN_ON leaving the program un-JITed.
Why it matters to the business
Systems running recent RISC-V kernels may experience degraded packet processing or observability performance when BPF programs fail to JIT, and operators must apply kernel updates before the next scheduled maintenance window.
How an attack unfolds
An attacker loads a BPF program via bpf() syscall; the JIT path calls alloc_new_pack which sets the memory read-only, patch_map then hands back the read-only address, the subsequent write faults, and the program is rejected with a warning, preventing JIT execution without requiring special privileges beyond normal BPF loading.
How to fix it
Upgrade to a kernel containing the referenced fix for riscv: patch: fix handling of bpf-jit execmem addresses; rebuild with CONFIG_EXECMEM enabled so patch_map creates the required temporary writable alias for execmem regions.
How to confirm the fix worked
After patching, load a BPF program that exercises the JIT path and confirm absence of the WARN_ON in bpf_int_jit_compile; verify that /proc/kallsyms or bpftool shows the program as JITed rather than interpreted.
Typical attack path
- Obtain ability to load BPF programs (CAP_BPF or unprivileged BPF)
- Invoke bpf_prog_load or equivalent with a program that triggers JIT compilation
- Trigger patch_map() on the vmalloc-backed BPF text region
- Observe page fault in copy_to_kernel_nofault and resulting -EINVAL / WARN_ON
How to detect it
Search kernel logs and dmesg for the WARN_ON emitted by bpf_int_jit_compile after a failed bpf_arch_text_copy; inspect audit logs for bpf() syscalls that return -EINVAL during JIT finalization.
What your security team should do
- Prioritize kernel update on all RISC-V systems within the next change window
- Audit BPF program loading paths and ensure fallback to interpreter does not violate performance SLAs
- Monitor kernel logs for the specific WARN_ON string from bpf_int_jit_compile
The record
In the Linux kernel, the following vulnerability has been resolved: riscv: patch: fix handling of bpf-jit execmem addresses Commit 8718e5a3090b ("riscv: patch: skip fixmap mapping when kernel text is already writable") gated the core_kernel_text() branch of patch_map() on CONFIG_STRICT_KERNEL_RWX, and explicitly left the vmalloc branch unchanged. That branch has a separate problem. That branch only creates a temporary writable fixmap alias when CONFIG_STRICT_MODULE_RWX is enabled, and otherwise assumes the target is directly writable and returns the address unchanged. That assumption no longer holds: bpf_prog_pack_alloc calls set_memory_rox() on every pack unconditionally in alloc_new_pack(), independently of any CONFIG_STRICT_*_RWX option, so BPF text in the vmalloc area is read-only regardless of what CONFIG_STRICT_MODULE_RWX says. With CONFIG_STRICT_MODULE_RWX=n, patch_map() returns the read-only address directly, the subsequent copy_to_kernel_nofault() takes a page fault and returns -EFAULT. bpf_arch_text_copy() turns that into -EINVAL, which propagates through bpf_jit_binary_pack_finalize() to the WARN_ON() in bpf_int_jit_compile() and leaves the program un-JITed. Note that BPF cannot be fixed the way kprobes was in commit bdc46e507b59 ("riscv: mm: make EXECMEM_KPROBES writable without CONFIG_STRICT_MODULE_RWX"). EXECMEM_BPF is already PAGE_KERNEL, i.e. writable at allocation time; the read-only mapping is established afterwards by generic code in alloc_new_pack(), which arch code cannot influence. The only place this can be handled is patch_map(). This is the same problem that was fixed on arm64 by commit b1480ed230ac ("arm64: patching: fix handling of execmem addresses"), and the fix is the same: CONFIG_EXECMEM is what actually tracks whether the vmalloc area can hold executable memory that needs a temporary alias to be written. CONFIG_BPF_JIT, CONFIG_KPROBES and CONFIG_MODULES all select it. Note that this is not limited to CONFIG_MODULES=n. Unlike arm64, riscv selects CONFIG_ARCH_OPTIONAL_KERNEL_RWX, so CONFIG_STRICT_MODULE_RWX can be disabled with CONFIG_MODULES=y as well, and the bug is reachable in that configuration too. The !CONFIG_MMU build is unaffected: it has its own __patch_insn_set() and __patch_insn_write() implementations and never calls patch_map().
Affected packages
| Ecosystem | Package | Introduced | Fixed |
|---|