linux-kernel - Re: [PATCH bpf-next v2] bpf: Return -ENOTSUPP if calls are not allowed in non-JITed programs

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <306f4ceb-0d25-d0a3-fc70-1141b6db06c8@iogearbox.net>
Date: Mon, 8 Jan 2024 16:27:20 +0100
From: Daniel Borkmann <daniel@...earbox.net>
To: Jiri Olsa <olsajiri@...il.com>, Tiezhu Yang <yangtiezhu@...ngson.cn>
Cc: Alexei Starovoitov <ast@...nel.org>, Andrii Nakryiko <andrii@...nel.org>,
 Eduard Zingerman <eddyz87@...il.com>, bpf@...r.kernel.org,
 linux-kernel@...r.kernel.org
Subject: Re: [PATCH bpf-next v2] bpf: Return -ENOTSUPP if calls are not
 allowed in non-JITed programs

On 1/8/24 11:05 AM, Jiri Olsa wrote:
> On Thu, Jan 04, 2024 at 09:08:17PM +0800, Tiezhu Yang wrote:
>> If CONFIG_BPF_JIT_ALWAYS_ON is not set and bpf_jit_enable is 0, there
>> exist 6 failed tests.
>>
>>    [root@...ux bpf]# echo 0 > /proc/sys/net/core/bpf_jit_enable
>>    [root@...ux bpf]# echo 0 > /proc/sys/kernel/unprivileged_bpf_disabled
>>    [root@...ux bpf]# ./test_verifier | grep FAIL
>>    #106/p inline simple bpf_loop call FAIL
>>    #107/p don't inline bpf_loop call, flags non-zero FAIL
>>    #108/p don't inline bpf_loop call, callback non-constant FAIL
>>    #109/p bpf_loop_inline and a dead func FAIL
>>    #110/p bpf_loop_inline stack locations for loop vars FAIL
>>    #111/p inline bpf_loop call in a big program FAIL
>>    Summary: 768 PASSED, 15 SKIPPED, 6 FAILED
>>
>> The test log shows that callbacks are not allowed in non-JITed programs,
>> interpreter doesn't support them yet, thus these tests should be skipped
>> if jit is disabled, just return -ENOTSUPP instead of -EINVAL for pseudo
>> calls in fixup_call_args().
>>
>> With this patch:
>>
>>    [root@...ux bpf]# echo 0 > /proc/sys/net/core/bpf_jit_enable
>>    [root@...ux bpf]# echo 0 > /proc/sys/kernel/unprivileged_bpf_disabled
>>    [root@...ux bpf]# ./test_verifier | grep FAIL
>>    Summary: 768 PASSED, 21 SKIPPED, 0 FAILED
>>
>> Additionally, as Eduard suggested, return -ENOTSUPP instead of -EINVAL
>> for the other three places where "non-JITed" is used in error messages
>> to keep consistent.
>>
>> Signed-off-by: Tiezhu Yang <yangtiezhu@...ngson.cn>
>> ---
>>
>> v2:
>>    -- rebase on the latest bpf-next tree.
>>    -- return -ENOTSUPP instead of -EINVAL for the other three places
>>       where "non-JITed" is used in error messages to keep consistent.
>>    -- update the patch subject and commit message.
>>
>>   kernel/bpf/verifier.c | 8 ++++----
>>   1 file changed, 4 insertions(+), 4 deletions(-)
>>
>> diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
>> index d5f4ff1eb235..99558a5186b2 100644
>> --- a/kernel/bpf/verifier.c
>> +++ b/kernel/bpf/verifier.c
>> @@ -8908,7 +8908,7 @@ static int check_map_func_compatibility(struct bpf_verifier_env *env,
>>   			goto error;
>>   		if (env->subprog_cnt > 1 && !allow_tail_call_in_subprogs(env)) {
>>   			verbose(env, "tail_calls are not allowed in non-JITed programs with bpf-to-bpf calls\n");
>> -			return -EINVAL;
>> +			return -ENOTSUPP;
> 
> FWIW I agree with John review earlier [1], also there's chance (however small)
> we could mess up with some app already checking on that

+1, the ship on this has sailed unfortunately. Tiezhu, it would be good if you could
update the selftest handling instead.

> jirka
> 
> [1] https://lore.kernel.org/bpf/6594a4c15a677_11e86208cd@john.notmuch/