Re: [PATCH] bpf: reject stack-argument callback subprograms

[Yonghong Song <yonghong.song@linux.dev>]

On 8/16/26 1:45 PM, Jérémy Jean wrote:
> Helper callbacks enter BPF subprograms through bpf_callback_t, whose
> runtime ABI supplies five arguments. BTF validation nevertheless permits
> static callback subprograms to declare more than five arguments when JIT
> stack arguments are supported.
>
> This lets verifier state for a callback use outgoing stack argument slots
> prepared at the helper call site. The helper does not pass those slots. On
> x86-64, callback loads of arguments seven and later therefore read the
> helper native frame instead of the synthetic values checked by the
> verifier. KASAN reports a slab OOB write.
>
> Reject callback subprograms with incoming stack arguments when processing
> callback calls. Add a verifier regression test using bpf_loop() and a
> nine-argument callback.
>
> Fixes: 0f6bd5e7a804 ("bpf: Support stack arguments for bpf functions")
> Assisted-by: Codex:gpt-5
> Signed-off-by: Jérémy Jean <Jeremy.Jean@oss.cyber.gouv.fr>
> ---
>   kernel/bpf/verifier.c                         |  4 +++
>   .../selftests/bpf/progs/verifier_stack_arg.c  | 33 +++++++++++++++++++
>   2 files changed, 37 insertions(+)
>
> diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
> index fdc5fbb1f78c..29aa4911c7f7 100644
> --- a/kernel/bpf/verifier.c
> +++ b/kernel/bpf/verifier.c
> @@ -9285,6 +9285,10 @@ static int push_callback_call(struct bpf_verifier_env *env, struct bpf_insn *ins
>   	err = btf_check_subprog_call(env, subprog, caller->regs);
>   	if (err == -EFAULT)
>   		return err;
> +	if (bpf_in_stack_arg_cnt(&env->subprog_info[subprog])) {
> +		verbose(env, "callback subprog cannot have stack args\n");
> +		return -EINVAL;
> +	}
>   
>   	/* set_callee_state is used for direct subprog calls, but we are
>   	 * interested in validating only BPF helpers that can call subprogs as
> diff --git a/tools/testing/selftests/bpf/progs/verifier_stack_arg.c b/tools/testing/selftests/bpf/progs/verifier_stack_arg.c
> index 7e0ce5db28a0..5acc3b63ca84 100644
> --- a/tools/testing/selftests/bpf/progs/verifier_stack_arg.c
> +++ b/tools/testing/selftests/bpf/progs/verifier_stack_arg.c
> @@ -27,6 +27,13 @@ static int subprog_7args(int a, int b, int c, int d, int e, int f, int g)
>   	return a + b + c + d + e + f + g;
>   }
>   
> +__noinline __used
> +static int callback_9args(__u32 index, void *ctx, long a3, long a4,
> +			  long a5, long a6, long a7, long a8, long a9)
> +{
> +	return a9;
> +}
> +
>   __noinline __used
>   static long subprog_deref_arg6(long a, long b, long c, long d, long e, long *f)
>   {
> @@ -79,6 +86,32 @@ __naked void stack_arg_two_subprogs(void)
>   	);
>   }
>   
> +SEC("tc")
> +__description("stack_arg: callback with incoming stack args")
> +__failure
> +__msg("callback subprog cannot have stack args")
> +__naked void stack_arg_callback_many_args(void)
> +{
> +	asm volatile (
> +		"r6 = 0;"
> +		"*(u64 *)(r11 - 32) = 0;"
> +		"*(u64 *)(r11 - 24) = 0;"
> +		"*(u64 *)(r11 - 16) = 0;"
> +		"*(u64 *)(r11 - 8) = 0;"
> +		"r1 = 1;"
> +		"r2 = %[callback_9args];"
> +		"r3 = 0;"
> +		"r4 = 0;"
> +		"call %[bpf_loop];"
> +		"r0 = 0;"
> +		"exit;"
> +		:
> +		: __imm_ptr(callback_9args),
> +		  __imm(bpf_loop)
> +		: __clobber_all
> +	);
> +}

I build with latest bpf-next and selftest with llvm23 and I got a selftest build
failure:

progs/verifier_stack_arg.c:96:3: error: inline assembly requires more registers than available
    96 |                 "r6 = 0;"
       |                 ^
   CLNG-BPF [test_progs] verifier_value_or_null.bpf.o
1 error generated.

The reason probably due to '__imm_ptr(callback_9args)'. The following change can fix the build:

index 5acc3b63ca84..f55a433b9202 100644
--- a/tools/testing/selftests/bpf/progs/verifier_stack_arg.c
+++ b/tools/testing/selftests/bpf/progs/verifier_stack_arg.c
@@ -108,7 +108,7 @@ __naked void stack_arg_callback_many_args(void)
                 :
                 : __imm_ptr(callback_9args),
                   __imm(bpf_loop)
-               : __clobber_all
+               : __clobber_common, "r6"
         );
  }

I then tested without your verifier.c change, I get the following test failure:

     All error logs:
     tester_init:PASS:tester_log_buf 0 nsec
     process_subtest:PASS:obj_open_mem 0 nsec
     process_subtest:PASS:specs_alloc 0 nsec
     run_subtest:PASS:obj_open_mem 0 nsec
     libbpf: prog 'stack_arg_callback_many_args': BPF program load failed: -EINVAL
     libbpf: prog 'stack_arg_callback_many_args': failed to load: -EINVAL
     libbpf: failed to load object 'verifier_stack_arg'
     run_subtest:PASS:unexpected_load_success 0 nsec
     validate_msgs:FAIL:979 expect_msg
     VERIFIER LOG:
     =============
     func#0 writes 4 stack arg slots, but calls only require 0
     processed 21 insns (limit 1000000) max_states_per_insn 1 total_states 2 peak_states 2 mark_read 0
     =============
     EXPECTED   SUBSTR: 'callback subprog cannot have stack args'
     Summary: 1/18 PASSED, 0 SKIPPED, 1/1 FAILED

I think the error message "func#0 writes 4 stack arg slots, but calls only require 0"
is good enough. So the extra error message "callback subprog cannot have stack args"
is not necessary in my opinion.

> +
>   SEC("tc")
>   __description("stack_arg: read from uninitialized stack arg slot")
>   __failure