Linux kernel mirror (for testing) git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
kernel os linux

bpf: Fix invalid prog->stats access when update_effective_progs fails

Syzkaller triggers an invalid memory access issue following fault
injection in update_effective_progs. The issue can be described as
follows:

__cgroup_bpf_detach
update_effective_progs
compute_effective_progs
bpf_prog_array_alloc <-- fault inject
purge_effective_progs
/* change to dummy_bpf_prog */
array->items[index] = &dummy_bpf_prog.prog

---softirq start---
__do_softirq
...
__cgroup_bpf_run_filter_skb
__bpf_prog_run_save_cb
bpf_prog_run
stats = this_cpu_ptr(prog->stats)
/* invalid memory access */
flags = u64_stats_update_begin_irqsave(&stats->syncp)
---softirq end---

static_branch_dec(&cgroup_bpf_enabled_key[atype])

The reason is that fault injection caused update_effective_progs to fail
and then changed the original prog into dummy_bpf_prog.prog in
purge_effective_progs. Then a softirq came, and accessing the members of
dummy_bpf_prog.prog in the softirq triggers invalid mem access.

To fix it, skip updating stats when stats is NULL.

Fixes: 492ecee892c2 ("bpf: enable program stats")
Signed-off-by: Pu Lehui <pulehui@huawei.com>
Link: https://lore.kernel.org/r/20251115102343.2200727-1-pulehui@huaweicloud.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>

authored by

Pu Lehui and committed by
Alexei Starovoitov
7dc211c1 4722981c

+10 -5
+7 -5
include/linux/filter.h
··· 712 712 ret = dfunc(ctx, prog->insnsi, prog->bpf_func); 713 713 714 714 duration = sched_clock() - start; 715 - stats = this_cpu_ptr(prog->stats); 716 - flags = u64_stats_update_begin_irqsave(&stats->syncp); 717 - u64_stats_inc(&stats->cnt); 718 - u64_stats_add(&stats->nsecs, duration); 719 - u64_stats_update_end_irqrestore(&stats->syncp, flags); 715 + if (likely(prog->stats)) { 716 + stats = this_cpu_ptr(prog->stats); 717 + flags = u64_stats_update_begin_irqsave(&stats->syncp); 718 + u64_stats_inc(&stats->cnt); 719 + u64_stats_add(&stats->nsecs, duration); 720 + u64_stats_update_end_irqrestore(&stats->syncp, flags); 721 + } 720 722 } else { 721 723 ret = dfunc(ctx, prog->insnsi, prog->bpf_func); 722 724 }
+3
kernel/bpf/syscall.c
··· 2463 2463 struct bpf_prog_stats *stats; 2464 2464 unsigned int flags; 2465 2465 2466 + if (unlikely(!prog->stats)) 2467 + return; 2468 + 2466 2469 stats = this_cpu_ptr(prog->stats); 2467 2470 flags = u64_stats_update_begin_irqsave(&stats->syncp); 2468 2471 u64_stats_inc(&stats->misses);