bpf: Keep refcount_acquire nullable for borrowed RCU kptrs
Linux
Affected versions
| Product | Affected | Fixed in |
|---|---|---|
| Linux Linux |
| No fix yet |
| Linux Linux |
|
Linux
| Product | Affected | Fixed in |
|---|---|---|
| Linux Linux |
| No fix yet |
| Linux Linux |
|
|
In the Linux kernel, the following vulnerability has been resolved:
bpf: Keep refcount_acquire nullable for borrowed RCU kptrs
bpf_refcount_acquire() is fallible for a borrowed reference because the object may have reached a zero refcount. The verifier therefore keeps KF_RET_NULL on the return value unless the argument is an owning reference.
An RCU-protected load of a local kptr is marked MEM_ALLOC, but it only receives NON_OWN_REF when the pointee contains a graph node. A refcounted object without a graph node consequently looks like an owning reference even though the loaded register has no acquired reference state. If the program drops the last real reference while remaining in the RCU critical section, refcount_inc_not_zero() returns NULL while the verifier treats the result as non-NULL.
Only classify the argument as owning when it is backed by a verifier-tracked reference. This retains the non-NULL return for pointers from bpf_obj_new(), bpf_kptr_xchg(), or an earlier successful acquisition, while requiring a NULL check for borrowed RCU kptrs.
[ kkd: Rewrote commit log ]
From the CVE record · Linux
The Linux kernel BPF verifier can incorrectly treat the result of bpf_refcount_acquire() as non-NULL for a borrowed RCU kptr. If a program drops the last real reference during the RCU critical section, the call can return NULL despite that assumption. The input does not specify an attacker impact or the access conditions needed to trigger this issue.
Engineers running the Linux kernel should check their versions; affected versions are not specified in the input.